Article

Plk1 and Mps1 Cooperatively Regulate the Spindle Assembly Checkpoint in Human Cells Graphical Abstract

Authors Conrad von Schubert, Fabien Cubizolles, Jasmine M. Bracher, Tale Sliedrecht, Geert J.P.L. Kops, Erich A. Nigg

Correspondence [email protected]

In Brief Faithful chromosome segregation depends on dynamic kinetochore phosphorylation by the spindle assembly checkpoint kinase Mps1. von Schubert et al. show that Plk1 supports Mps1 function by phosphorylation of shared sites on Mps1 and KNL-1, leading to potentiation of Mps1 kinase activity and robust kinetochore recruitment of checkpoint components, respectively.

Highlights d

Plk1 promotes the spindle checkpoint through a substrate preference shared with Mps1

d

Plk1 and Mps1 cooperatively phosphorylate KNL-1 MELT motifs and Mps1

d

Plk1 potentiates Bub3:BubR1 and Mad1:C-Mad2 kinetochore recruitment and Mps1 activity

von Schubert et al., 2015, Cell Reports 12, 66–78 July 7, 2015 ª2015 The Authors http://dx.doi.org/10.1016/j.celrep.2015.06.007

Cell Reports

Article Plk1 and Mps1 Cooperatively Regulate the Spindle Assembly Checkpoint in Human Cells Conrad von Schubert,1 Fabien Cubizolles,1 Jasmine M. Bracher,1,5 Tale Sliedrecht,2,3,4 Geert J.P.L. Kops,2,3,4 and Erich A. Nigg1,* 1Biozentrum,

University of Basel, Klingelbergstrasse 50/70, 4056 Basel, Switzerland Cancer Research 3Center for Molecular Medicine 4Cancer Genomics Netherlands University Medical Center Utrecht, 3584 CG, Utrecht, the Netherlands 5Present address: Department of Biotechnology, Delft University of Technology, Julianalaan 67, 2628 BC Delft, the Netherlands *Correspondence: [email protected] http://dx.doi.org/10.1016/j.celrep.2015.06.007 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 2Molecular

SUMMARY

Equal mitotic chromosome segregation is critical for genome integrity and is monitored by the spindle assembly checkpoint (SAC). We have previously shown that the consensus phosphorylation motif of the essential SAC kinase Monopolar spindle 1 (Mps1) is very similar to that of Polo-like kinase 1 (Plk1). This prompted us to ask whether human Plk1 cooperates with Mps1 in SAC signaling. Here, we demonstrate that Plk1 promotes checkpoint signaling at kinetochores through the phosphorylation of at least two Mps1 substrates, including KNL-1 and Mps1 itself. As a result, Plk1 activity enhances Mps1 catalytic activity as well as the recruitment of the SAC components Mad1:C-Mad2 and Bub3:BubR1 to kinetochores. We conclude that Plk1 strengthens the robustness of SAC establishment at the onset of mitosis and supports SAC maintenance during prolonged mitotic arrest. INTRODUCTION The spindle assembly checkpoint (SAC) ensures equal chromosome segregation during eukaryotic cell division (Lara-Gonzalez et al., 2012; Musacchio and Salmon, 2007; Vleugel et al., 2012). Consequently, its deregulation results in aneuploidy, loss of genome integrity, and increased tumorigenic potential. The SAC delays progression through mitosis until all sister chromatids are attached via their kinetochores (KTs) to microtubules (MTs) emerging from opposite spindle poles. Central to the SAC is a diffusible effector, known as the mitotic checkpoint complex (MCC), that inhibits the anaphase-promoting complex/cyclosome (APC/C), a ubiquitin ligase required for the degradation of early mitotic proteins. In turn, APC/C controls the onset of sister chromatid separation and mitotic exit through proteolytic degradation of the Separase inhibitor Securin and the Cdk1 activator Cyclin B, respectively. The estab66 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

lishment and maintenance of the MCC-based inhibitory signal requires the protein kinase Monopolar spindle 1 (Mps1) (Liu and Winey, 2012), which accumulates and autoactivates at unattached or misaligned KTs. Active Mps1 is important for the recruitment of numerous checkpoint components, including Mad1:C-Mad2 and Bub3:BubR1 complexes (Funabiki and Wynne, 2013), that then cooperate to form MCC: upon binding to KTs, Mad1:C-Mad2 catalyzes the conformational activation of additional Mad2 molecules, resulting in the closed conformer C-Mad2. C-Mad2 associates in an inhibitory sub-complex with the APC/C coactivator Cdc20. In a parallel process, Mps1 phosphorylates an array of Met-Glu-Leu-Thr (MELT) motifs within the KT-associated protein KNL-1, which then acts as a recruitment platform for Bub3:Bub1 and Bub3:BubR1 complexes (Krenn et al., 2014; London et al., 2012; Overlack et al., 2015; Primorac et al., 2013; Shepperd et al., 2012; Vleugel et al., 2013, 2015; Yamagishi et al., 2012). Accumulating Bub3:BubR1 and C-Mad2:Cdc20 sub-complexes finally combine to form the APC/C-inhibitory MCC (Lara-Gonzalez et al., 2012). The aforementioned functions place Mps1 at the core of SAC signaling, but surprisingly few physiological substrates have, so far, been described for this kinase. In human cells, these include Mps1 itself, the KNL-1/Mis12/Ndc80 complex (KMN) network component KNL-1, and Borealin (Funabiki and Wynne, 2013; Liu and Winey, 2012). Mass spectrometric analysis of in vivo autophosphorylation sites (Dou et al., 2011), as well in vitro phosphorylated peptide libraries (Hennrich et al., 2013), revealed a strong preference of Mps1 for acidic residues in the
2 position relative to the phosphoacceptor site. Remarkably, this motif preference is very similar to that described for Polo-like kinase 1 (Plk1) (Nakajima et al., 2003; Santamaria et al., 2011), a key regulator of mitotic progression (Petronczki et al., 2008). This raised the question of whether the two kinases might cooperate in SAC signaling through phosphorylation of common substrates. Similar to Mps1, Plk1 accumulates at unaligned KTs, but its role in SAC signaling remains controversial (Petronczki et al., 2008): in many instances, inhibition or depletion of Plk1 causes a SAC-dependent arrest, arguing that Plk1 is not strictly required for SAC signaling, but, in some species, Plk1 contributes to

SAC signaling through positive effects on the KT recruitment of SAC components. The present study aims at clarifying the role of Plk1 in SAC signaling in human cells. We show that Plk1, indeed, strengthens SAC signaling but that this role is generally masked by additional functions of Plk1—notably, its contribution to bipolar spindle formation. Our data support a model according to which Plk1 cooperates with Mps1 through a shared preference for similar phosphorylation consensus motifs within substrates. Specifically, we identify Mps1 autophosphorylation sites, as well as MELT motifs within KNL-1, as common substrates for both kinases. Collectively, not only do our results demonstrate a synergistic role of Plk1 and Mps1 in the regulation of SAC signaling, but they also provide a molecular basis for the observed synergy. RESULTS Human Plk1 and Mps1 Cooperatively Regulate Checkpoint Establishment and Maintenance The similarity between the preferred phosphorylation sites of Mps1 and Plk1 raised the possibility of functional cooperation. We reasoned that a role for Plk1 in SAC signaling might previously have escaped detection because of the essential roles of Plk1 in bipolar spindle formation and KT-fiber stabilization. Therefore, we carried out initial experiments in cells that had been treated with high doses of the MT-depolymerizing drug nocodazole, thereby maximizing the strength of the SAC signal. Moreover, we reasoned that a potential synergy between Plk1 and Mps1 in SAC signaling might be more readily detectable under conditions of partially impaired Mps1 activity. Asynchronously growing human RPE-1 cells were, therefore, treated with nocodazole and analyzed by time-lapse imaging in the presence or absence of kinase inhibitors (Figure 1A). In the absence of nocodazole, inhibition of Plk1 by TAL (Santamaria et al., 2007) caused spindle collapse and mitotic arrest, while inhibition of Mps1 by Reversine (Santaguida et al., 2010) caused a SAC override, regardless of the activity state of Plk1, as expected (Figure 1A). At a low dose of nocodazole (0.1 mg/ml), the effect of single Plk1 inhibition on SAC signaling was difficult to discern, presumably reflecting residual MTs interacting with KTs (Yang et al., 2009), but synergy with Mps1 inhibition could be observed (Figure 1A). When nocodazole was used at a dose (1 mg/ml) causing complete depolymerization of MTs (Yang et al., 2009), inhibition of Plk1 reduced the duration of checkpoint arrest from 31 hr to 21 hr, inhibition of Mps1 reduced this time to 11 hr, and simultaneous inhibition of both kinases clearly produced a strong synergistic effect, causing almost immediate slippage (Figures 1A, S1A, and S1B). Synergy between Plk1 and Mps1 could also be demonstrated using small interfering RNA (siRNA) to deplete Plk1 (Figure S1C) or using ATP analogs for kinase inhibition in cell lines harboring corresponding analogsensitive versions of the two kinases (Burkard et al., 2007; Maciejowski et al., 2010) (Figures S1D and S1E). Furthermore, similar results were obtained for a range of transformed and untransformed cell lines, representing different tissues (cervix, colon, and retina) and different degrees of karyotypic complexity (Figure S1F). We conclude that, under conditions of maximal SAC signal strength, simultaneous interference with Plk1 and

Mps1 causes more drastic weakening of the SAC than inhibition of Mps1 alone. Next, we examined the consequences of Plk1 inhibition on key SAC components. No significant effects could be seen at the level of protein expression (Figure S1G). In contrast, Plk1 inhibition had a major impact on the KT recruitment of Mad1, as visualized in RPE-1 cells that had been arrested in the M phase by treatment with the proteasome inhibitor MG132 (Figure 1B). While Mad1 was absent from KTs in control cells, indicating completion of bipolar spindle formation and silencing of the SAC, inhibition of Plk1 triggered collapse of the spindle and concomitant re-recruitment of Mad1 to KTs. As expected, this restoration of KT Mad1 was dependent on Mps1 activity (Figure 1B). Strikingly, however, when analogous experiments were carried out in the complete absence of MTs, maximal KT recruitment of Mad1 was only seen when both Plk1 and Mps1 were active (Figure 1B). Similarly, inhibition of Plk1 led to an additive effect on Mad1 KT recruitment when analog-sensitive Mps1 was inhibited with the ATP analog 3-MB-PP1 (Figure S1H). These results demonstrate that Plk1 activity is required for maximal recruitment of the SAC component Mad1 to KTs. To extend the aforementioned findings to Mad2, the major binding partner of Mad1, we used an RPE-1 cell line harboring an EGFP-tagged version of Mad2 that had been genetically engineered through introduction of the EGFP coding sequence into the 50 end of one of the two endogenous MAD2L1 loci (Figure S1I). In these cells, KT recruitment of Mad2 was as dependent on both Plk1 and Mps1 as that of Mad1 (Figure S1J). Virtually identical observations were made when using an antibody specific for the closed conformation of Mad2 (C-Mad2) (data not shown). Taken together, the aforementioned results demonstrate that Plk1 and Mps1 cooperate in the KT recruitment of the Mad1:C-Mad2 complex. Having shown that Plk1 cooperates with Mps1 in the establishment of the SAC, we asked whether this kinase also contributes to SAC maintenance. RPE-1 cells expressing endogenously tagged Mad2 were arrested in the M phase with nocodazole, and their ability to maintain a prolonged SAC arrest was determined by time-lapse imaging (Figure 1C; Movie S1; since TAL displays auto-fluorescence, the chemically distinct Plk1 inhibitor BI-2536 [Petronczki et al., 2008] was used for these experiments). While treatment with either Plk1 or Mps1 inhibitor resulted in a progressive loss of EGFP-Mad2 from KTs (Figure 1C) and a concomitant shortening of mitotic arrest (Figure 1C, right panel), the simultaneous inhibition of both kinases produced additive effects. This demonstrates that cooperation between Plk1 and Mps1 also occurs during maintenance of the SAC. Next, we considered it important to extend our conclusions to cells undergoing unperturbed mitosis. To avoid possible complications from the role of Plk1 in KT-fiber stability, we focused attention on cells at the early stages of mitosis (prophase and prometaphase), which are characterized by unstable lateral interactions between MTs and KTs, or a complete lack thereof, rather than end-on attachments (Magidson et al., 2011). Specifically, we quantified the KT recruitment of C-Mad2 at the onset of nuclear envelope breakdown (late prophase, visualized by staining of the nuclear pore complex protein TPR), at the time of chromosome ring formation (early prometaphase), and at the Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 67

Figure 1. Human Plk1 and Mps1 Cooperatively Regulate Checkpoint Establishment and Maintenance (A) Asynchronously growing RPE-1 cells were treated with the anti-MT drug nocodazole and solvent only (control), the Mps1 inhibitor Reversine, the Plk1 inhibitor TAL, or both inhibitors combined and analyzed by time-lapse imaging. Scatterplots show time in mitosis; 10–30 cells per conditions. aver, average; h, hours. (B) RPE-1 cells pretreated with MG132 were incubated for 1 hr in the presence or absence of nocodazole and indicated inhibitors. Cells were analyzed by immunofluorescence (IF) microscopy, and human CREST antiserum was used to identify KTs. Left: chromosome alignment states at the end of incubation; percentages indicate the frequencies of shown spindle morphologies. Right: scatterplot depicts Mad1 staining intensity at KTs; 28–39 cells per condition. Rev, Reversine; rel. signal int., relative signal intensity. (C) RPE-1 cells expressing endogenously EGFP-tagged Mad2 (Figure S1I) were synchronized in mitosis by incubation with nocodazole, followed by the addition of the indicated kinase inhibitors and time-lapse imaging. Left: stills from live recordings (Movie S1). Middle: traces illustrating loss of EGFP-Mad2 from KTs; n = 8 cells per condition, shaded areas represent SEM. Right: quantification of the time in mitosis; n = 41–49 cells per condition. int., intensity. Scale bars, 5 mm. ***p < 0.001 (from Student’s t tests). See also Figure S1I and Movie S1.

beginning of chromosome biorientation (late prometaphase). In both late prophase and early prometaphase cells, inhibition of Plk1 strongly reduced the KT recruitment of both C-Mad2 (Figure 2; Figure S2A) and Mad1 (Figure S2B), demonstrating that Plk1 promotes SAC establishment during the early stages of an otherwise unperturbed mitosis. Plk1 Inhibition Acts on the KT Branch of the SAC Elegant recent studies have demonstrated that MCC generated during late interphase and early mitosis originates at two distinct sources (Rodriguez-Bravo et al., 2014). The first pool of MCC is assembled at nuclear envelopes already prior to the onset of mitosis, thereby ensuring a minimal level of anaphase-onset 68 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

inhibitory SAC signal until the second pool of MCC can be generated from unattached KTs. Mps1 is implicated in the generation of both pools of MCC (Maciejowski et al., 2010), raising the question of whether Plk1 cooperates with Mps1 already during interphase. Inhibition of Mps1 was previously shown to substantially reduce the levels of Cyclin B1 that are detectable within the nucleus at the onset of mitosis, providing a convenient readout for the generation of interphase MCC (Maciejowski et al., 2010). Moreover, examination of the rate of Cyclin B1 destruction in cells undergoing mitosis in the presence of a high dose of nocodazole provided a readout for the activity of the KT-borne branch of MCC. For both measurements, we used RPE-1 cells harboring a Venus-tagged version of the CCNB1 gene (Collin et al., 2013) to

Figure 2. Plk1 Inhibition Delays SAC Activation in Unperturbed Mitosis Left: asynchronously growing RPE-1 cells were incubated with kinase inhibitors or solvent for 2 hr and analyzed by IF, using the indicated antibodies. Mitotic phases were determined as described in Results (see also Figure S2A). Right: KT recruitment of C-Mad2 was measured at different stages of mitosis; scatterplot shows results for n = 377–456 KT pairs from 10–13 cells per condition. fluor. int., fluorescence intensity. Scale bar, 5 mm. ***p < 0.001 (from Student’s t tests). See also Figure S2A.

monitor Cyclin B1 levels by fluorescence time-lapse imaging (Figures 3A and 3B; Movie S2). In agreement with previous observations (Maciejowski et al., 2010), cells treated with the Mps1 inhibitor Reversine showed a marked reduction of Venus fluorescence in the nuclei of prophase cells, suggesting reduced activity of interphase MCC (Figures 3A and 3B; t = 0). Remarkably, however, addition of the Plk1 inhibitor BI-2536 did not detectably lower the nuclear levels of Cyclin B1 (Figures 3A and 3B; t = 0), suggesting that Plk1 does not contribute to the generation of interphase MCC. One caveat to this conclusion is that high levels of Cyclin B could also reflect, at least in part, a role of Plk1 in the activation of APC/C (Petronczki et al., 2008). The activity of the KT-borne branch of MCC could be assessed in the same experiment by monitoring the kinetics of Cyclin B1 degradation (Figure 3B; t = 20–200). While inhibition of either Plk1 or Mps1 caused a marked acceleration of mitotic exit, the simultaneous inhibition of both kinases resulted in a precipitous drop of Cyclin B1 levels. These results led us to conclude that Plk1 contributes primarily to the KT-borne branch of SAC signaling. This conclusion is further supported by experiments showing that Mps1 inhibition causes only partial removal of Plk1 from KTs (Figure S3A) and no detectable effect on global Plk1 activity (Figure S3B). Moreover, Plk1 needs to be associated with KTs for SAC signaling (Figures S3C and S3D). The Effect of Plk1 Inhibition on SAC Signaling Does Not Involve Delocalization or Reduced Catalytic Activity of Aurora B We considered the possibility that the effect of Plk1 on SAC signaling might be mediated by Aurora B kinase, which promotes Mps1 KT recruitment and reduces KT-associated phosphatase 1 activity (Hewitt et al., 2010; Maciejowski et al., 2010; Maldonado and Kapoor, 2011; Rodriguez-Bravo et al., 2014; Santaguida et al., 2010, 2011; Vigneron et al., 2004). Aurora

B is recruited to centromeres in response to phosphorylation of histone H2A (on Thr120) and histone H3 (on Thr3) by the kinases Bub1 and Haspin, respectively (van der Horst and Lens, 2014). Although inhibition of Plk1 in mitotically arrested RPE-1 cells showed a marginal reduction of Thr3 phosphorylation (by about 20%), consistent with a reported role for Plk1 in Haspin activation (Zhou et al., 2014), this was not sufficient to detectably reduce Aurora B levels at centromeres (Figure 4A; Figure S4A). In contrast, Aurora B centromere levels dropped by 40% when histone H3 Thr3 phosphorylation was almost completely abolished in response to Haspin inhibition by 5-iodotubercidin (5-ITu), as expected (De Antoni et al., 2012). The lack of significant effects of Plk1 inhibition on Aurora B localization was also confirmed in RPE-1 cells expressing endogenously EGFP-tagged Aurora B, generated by the introduction of an EGFP tag into one allele of the AURKB gene (Figure S4B). Live-cell imaging revealed that single or combined inhibition of Plk1 and/or Mps1 did not detectably impair Aurora B recruitment, while inhibition of Haspin produced the expected inhibitory effect (Figure 4B). Conversely, the concomitant inhibition of Plk1 and Mps1 strongly reduced Mad1 levels at KTs, whereas the inhibition of Haspin had no effect (Figures 4C and S4C). These results demonstrate that Plk1 regulates SAC signaling, as monitored by Mad1 recruitment, through mechanisms that are independent of any potential effects on Aurora B localization. To extend the aforementioned conclusion to the catalytic activity of Aurora B, we took advantage of the Aurora B biosensors tethered to either chromatin (histone H2B) or centromeres (centromere protein B [CENP-B]) (Fuller et al., 2008). Inhibition of Plk1 or Mps1, either individually or in combination, did not detectably alter fluorescence resonance energy transfer (FRET) efficiency (Figure 4D; Movie S3). In contrast, direct inhibition of Aurora B by ZM-447439 completely abolished sensor Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 69

phosphorylation, as expected (Figure 4D). While combined inhibition of Plk1 and Mps1 resulted in a more rapid SAC override than inhibition of Mps1 alone (Figure 4D), Aurora B inhibition did not cause mitotic slippage within the observation period, supporting our conclusion that loss of Aurora B activity cannot explain the observed effect of combined Plk1 and Mps1 inhibition. Fully consistent results were obtained in experiments using a centromere-targeted FRET sensor (Figure S4D); phospho-specific antibodies against histone H3 (Ser10) and CENP-A (Thr7), two well-characterized Aurora B substrates (De Antoni et al., 2012) (Figures 4E and S4E); and a Mis12Mps1 construct conferring constitutive KT localization (Figure S4F). Collectively, these results strongly argue that Plk1 inhibition does not affect Aurora B localization or activity to an extent that could explain the cooperation between Plk1 and Mps1 in SAC signaling.

Figure 3. Plk1 Inhibition Acts on the KT Branch of the SAC Asynchronous RPE-1 cells expressing endogenously Venus-tagged Cyclin B1 were incubated with nocodazole and indicated inhibitors, followed by timelapse imaging. (A) Scatterplots depict levels of Cyclin B1-Venus at the indicated cell cycle stages; time is indicated relative to nuclear translocation of Cyclin B1-Venus (n = 20–29 cells per condition). Prometaph., prometaphase. (B) Upper: micrographs represent still images from live recordings (Movie S2); time stamps indicate time relative to nuclear translocation of Cyclin B1-Venus. Scale bar, 5 mm. Lower: tracings illustrating levels of Cyclin B1 before nuclear envelope breakdown (time = 0) as well as kinetics of subsequent degradation (n = 7–12 cells per condition). Shaded areas represent SEM. Rev, Reversine; int., intensity. See also Movie S2.

70 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

Plk1 and Mps1 Both Target KNL-1 MELT Motifs MCC production from KTs results from the convergence of two major pathways. One involves the KT recruitment of a Mad1:C-Mad2 complex, which then allows the formation of CMad2:Cdc20; the other is centered on the KMN complex component KNL-1, which has emerged as an important hub for the assembly of Bub3:Bub1 and Bub3:BubR1 complexes (see Introduction). Our data clearly indicate that Plk1 cooperates with Mps1 in the KT recruitment of the Mad1:C-Mad2 complex, but this does not exclude additional levels of cooperation. To explore this possibility, we asked whether Plk1 inhibition would still cooperate with Mps1 inhibition when Mad1 was constitutively targeted to KTs through fusion to Mis12. As shown in Figure S5A, combined inhibition of Plk1 and Mps1 produced a synergistic effect on the timing of mitotic exit even in these cells, arguing that Plk1 does more than simply cooperate with Mps1 in the KT recruitment of Mad1. The aforementioned result led us to focus on the second arm of KT-dependent SAC signaling, the one centered on KNL-1. Mps1 is known to phosphorylate KNL-1 on a series of MELT motifs, which then triggers the recruitment of Bub3:Bub1 and Bub3:BubR1 complexes, prompting us to ask whether these phosphorylation sites might also be targeted by Plk1. Synthetic peptides representing all 19 MELT motifs of human KNL-1 (Vleugel et al., 2013) were spotted onto membranes and exposed to recombinant Plk1 in the presence of g-32P-ATP. Remarkably, Plk1 readily and specifically phosphorylated 13 of these sites (Figure 5A; Figure S5B), including motifs that have recently been shown to be targeted by Mps1 (Thr875, NDMDITKSY) (Yamagishi et al., 2012) and to be sufficient for KNL-1 function (DDMEITRSHTTA) (Vleugel et al., 2013). To extend these in vitro results to living cells, we used the KNL-1 Thr875 site to generate a chromatin (H2B)-targeted FRET reporter (Figure S5C; Movie S4). After expression of this FRET reporter in HeLa cells, we could readily demonstrate reduced phosphorylation in response to Plk1 inhibition (Figure 5B), arguing that Plk1 is, indeed, capable of phosphorylating KNL-1 MELT motifs in vivo. Sensitivity to Mps1 inhibition could not be demonstrated in this experiment, presumably reflecting low abundance of cytosolic Mps1 (Saurin et al., 2011). However, we emphasize that this same reporter was similarly sensitive to Plk1 or Mps1 inhibition

Figure 4. The Effect of Plk1 Inhibition on SAC Signaling Does Not Involve Delocalization or Reduced Catalytic Activity of Aurora B In (A)–(C) and (E), RPE-1 cells were preincubated for 2 hr with MG132 and then treated with nocodazole and the indicated kinase inhibitors for 1 hr. (A) Histogram shows the phosphorylation state of histone H3 Thr3 and centromere levels of Aurora B as determined by IF (n = 25–35 cells per condition); micrographs are shown in Figure S4A. rel. signal int., relative signal intensity. (B) Left: EGFP centromere signals captured from living cells expressing endogenously EGFP-tagged Aurora B (Figure S4B). Right: scatterplot showing EGFP signals for n = 120 centromeres from 15 cells per condition. Rev., Reversine. (C) Histogram shows Mad1 KT levels as determined by IF analysis (n = 31–39 cells per condition; micrographs are shown in Figure S4C). (D) Asynchronously growing HeLa cells stably expressing a chromatin-targeted FRET sensor for Aurora B activity were incubated with nocodazole and kinase inhibitors and monitored by time-lapse ratio imaging (Movie S3). Left: heatmap representation of reporter phosphorylation; time is indicated relative to mitotic entry. Right: traces illustrating the results from n = 14–22 cells per condition; shaded areas represent SEM. (E) Histogram shows the quantification of IF signals produced by phospho-specific antibodies detecting Aurora B targets. Signals were quantified over chromosome arms for phospho-Ser10 on histone H3 and over centromeres for phospho-Thr7 on CENP-A (n = 24–29 cells per condition; micrographs are shown in Figure S4E). Scale bars, 5 mm. Error bars indicate SEM. ***p < 0.001 (from Student’s t tests); ns, not significant. See also Figure S4 and Movie S3.

when tethered to KTs, confirming cooperativity between the two kinases (data not shown). Cooperativity in the phosphorylation of Thr875 could also be demonstrated using a phospho-specific antibody (Yamagishi et al., 2012) to stain RPE-1 cells: while inhibition of either Plk1 or Mps1 significantly reduced signal intensity at KTs, combined kinase inhibition clearly produced additive effects (Figures 5C and S5D). The most straightforward interpretation of these results is that Plk1 and Mps1 cooperate to target endogenous KNL-1 on MELT motifs. To rule out the alternative explanation that Plk1 might enhance KNL-1 phos-

phorylation through indirect effects on phosphatases antagonistic to Mps1, we treated RPE-1 cells with high doses of okadaic acid (500 nM), which are known to inhibit both PP1 and PP2A (De Antoni et al., 2012). While okadaic acid enhanced Thr875 phosphorylation, as expected, it did not abolish the effect of Plk1 inhibition (Figure S5E), arguing that Plk1 phosphorylates this site directly. Likewise, okadaic acid did not affect the impact of Plk1 inhibition on the KT recruitment of Mad1 (Figure S5E). To demonstrate that the phosphorylation of KNL-1 MELT motifs by Plk1 resulted in important physiological consequences, Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 71

Figure 5. Plk1 and Mps1 Both Target KNL-1 MELT Motifs (A) Autoradiography shows the results of in vitro peptide spotting kinase assays. Peptides containing the MELT motif (shaded areas) were synthesized carrying the Thr phosphoacceptor (T, arrowhead) or an Ala (A) in the corresponding position and incubated with g-32P-ATP in the presence (+Plk1) or absence (no enzyme) of recombinant Plk1. Additional peptides carrying Ala in place of potentially interfering phosphoacceptors were also analyzed. Equal synthesis of peptides was monitored by hydration of membranes prior to side-chain de-protection (peptide). (B) The motif around KNL-1 Thr875 was used to design a MELT-based FRET sensor (Figure S5C) targeted to chromatin by H2B fusion. To determine the specificity of MELT phosphorylation, Thr875 was substituted with Ala. HeLa S3 cells expressing the FRET reporters were enriched in mitosis with nocodazole and additionally treated with MG132 and 100 nM BI-2536 before CFP/FRET ratios were determined by live-cell microscopy. Left: heatmap representation of reporter phosphorylation. Right: scatterplot shows emission ratios from n = 17–31 cells per condition. (C) RPE-1 cells were synchronized in metaphase using MG132 and subsequently treated with nocodazole and kinase inhibitors. Left: KNL-1 phosphorylation site occupancy, as determined by IF staining with an antibody specific for phospho-Thr875. Right: histogram shows quantification of P-Thr875 IF staining (n = 50–62 cells per condition). Rev., Reversine; rel. signal int., relative signal intensity. Error bars indicate SEM. (D and E) Plk1 inhibition reduces Bub3:BubR1 complex recruitment to KTs. RPE-1 cells (D) or RPE-1 cells expressing endogenously EGFP-tagged Bub3 (E; Figure S5H) were treated as described in (C) and analyzed by IF, using the indicated antibodies; scatterplots show results from quantitative IF (n = 65–69 cells per condition; micrographs are shown in Figure S5I). Scale bars, 5 mm. ***p < 0.001 (from Student’s t tests). See also Figure S5.

we sought to correlate the effect of Plk1 inhibition on KNL-1 phosphorylation with the KT recruitment of the Bub3:BubR1 complex. Indeed, inhibition of Plk1 lowered not only Thr875 phosphorylation but also BubR1 (Figures 5D, S5F, and S5G) as 72 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

well as EGFP-Bub3 signals at KTs (Figures 5E, S5H, and S5I). Notably, quantification revealed a strong correlation between the reductions in MELT motif phosphorylation and KT recruitment of these SAC signaling components (Figures 5E and S5I).

Figure 6. Plk1 Targets Mps1 Autophosphorylation Sites In Vitro (A) Mixtures of recombinant Plk1 and Mps1 were incubated for 2 min with g-32P-ATP and with kinase inhibitors as indicated. Phosphorylation of the Plk1 substrate casein was monitored for control. Phosphorylation of Mps1 and casein was detected by autoradiography (32P), levels of Plk1 and Mps1 were monitored by western blotting (WB), and casein was visualized by Ponceau S (PS) staining. Rev., Reversine. (B) Autoradiography and WB illustrate the time course of in vitro Mps1 phosphorylation in the presence or absence of Plk1 and Reversine (left). Quantitative data (right) were averaged from two distinct experiments. (C) Autoradiography shows the results of peptide spotting kinase assays performed as described in the legend for Figure 5A. Peptides were synthesized to represent phosphorylation sites within Mps1 that conform to the Plk1/Mps1 consensus or were previously annotated as Mps1 autophosphorylation sites (marked by asterisks). T/S, Thr/Ser phosphoacceptor; A, Ala in position of the phosphoacceptor. (D) Mixtures of recombinant Plk1 and kinase-dead Mps1 were subjected to in vitro kinase assays before phosphorylations on Mps1 (pS/pT) were mapped by mass spectrometry. Phosphopeptides were identified either with Mascot/Scaffold (M/S) or MaxQuant/Andromeda (MQ/A) with a decoy FDR of 0.01 on peptide spectrum matches as significance cutoff (Table S1). Asterisks indicate Mps1 autophosphorylation sites. (E) Schematic summarizes Plk1-dependent phosphorylations on Mps1, as detected in the assays shown in (C) and (D), as well as Table S1. Annotated Mps1 autophosphorylation sites are highlighted in blue. See also Table S1.

Plk1 Targets Mps1 Autophosphorylation Sites Our identification of KNL-1 as a common substrate for Plk1 and Mps1 readily explains the observed effects of kinase inhibition on Bub3:Bub1/BubR1 localization. This left open the question of how Plk1 and Mps1 cooperate in the KT recruitment of the Mad1:C-Mad2 complex. A priori, Plk1 could either phosphorylate Mad1 or its interaction partners, or, alternatively, enhance the ability of Mps1 to recruit Mad1. To explore this latter possibility, we asked whether Plk1 is capable of phosphorylating Mps1. When assayed in vitro, Plk1 readily phosphorylated kinase-dead Mps1

(Figure S6A), confirming earlier data (Dou et al., 2011). Moreover, Plk1 was able to rapidly stimulate wild-type Mps1 autophosphorylation (Figures 6A and 6B). To identify the sites targeted by Plk1, we carried out two types of experiments. First, peptides representing all annotated Mps1 phosphorylation sites, as well as putative Plk1 sites, were spotted onto membranes and tested as substrates in kinase assays (Figure 6C). We found that 12 sites within Mps1 were efficiently phosphorylated by Plk1 and, notably, 7 of these had previously been annotated as Mps1 autophosphorylation sites (Dou et al., 2011; Jelluma et al., 2008; Tyler Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 73

(legend on next page)

74 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

et al., 2009; Xu et al., 2009). Second, mass spectrometry was used to map phosphorylation sites in kinase-dead Mps1 that had been phosphorylated in vitro by Plk1; this resulted in the detection of 11 sites, of which 7 had previously been attributed to Mps1 autophosphorylation (Figure 6D; Table S1). In total, we identified 17 distinct sites targeted by Plk1, including 9 purported Mps1 autophosphorylation sites (Figures 6E and S6B). Most of these 17 sites were highly conserved in vertebrates, and 12 could also be detected in immunoprecipitated Mps1, arguing that these sites are also phosphorylated in vivo (Figure S6B; Table S1). Taken together, these results demonstrate that Plk1 is capable of phosphorylating Mps1 on residues previously characterized as autophosphorylation sites. This conclusion is further supported by data identifying three of the aforementioned sites (Thr33, Ser37 and Ser363) as being both responsive to Plk1 inhibition in vivo (Dou et al., 2011; Oppermann et al., 2012) and capable of serving as Mps1 substrates in vitro (Jelluma et al., 2008; Xu et al., 2009). Plk1 Enhances Mps1 Kinase Activity and KT Dissociation Most of the residues identified as common targets for Plk1 and Mps1 are located within or near the N-terminal KT-binding domain of Mps1 (Figure 6E), and several of these, including Thr33, Ser37, Ser321, Thr363, and Thr371, are implicated in Mps1 turnover at KTs (Wang et al., 2014). Moreover, previous studies have shown that Mps1 promotes its own dissociation from KTs (Hewitt et al., 2010; Jelluma et al., 2010; Santaguida et al., 2010). To explore the possibility that Plk1 might contribute to modulate Mps1 turnover at KTs, we monitored the KT dynamics of GFP-tagged wild-type Mps1 expressed in RPE-1 Mps1 null cells (Maciejowski et al., 2010). As predicted, KT levels of Mps1 increased significantly in the presence of a Plk1 inhibitor, and this increase was additive to the effect of Mps1 inhibition (Figure 7A). In contrast, no marked changes in the KT levels of Hec1, a putative binding partner of Mps1, were detected (Figure S7A). To corroborate and extend these results, we also carried out FRAP experiments (Figure 7B; Movie S5). In control cells, Mps1 showed a short recovery half-time of 1 s after photobleaching and a small immobile fraction (4%), in excellent agreement with previous data (Jelluma et al., 2010). Inhibition of Plk1

caused a nearly 2-fold increase in the half-time and a 3-fold increase in the immobile fraction (to 13%). In comparison, Mps1 inhibition increased the half-time to 2 s and the immobile fraction to 17%. Most important, the combined inhibition of both Plk1 and Mps1 produced additive effects, resulting in a half-time of 4 s and an immobile fraction of 25%. Furthermore, we noticed that combined inhibition of the two kinases led to premature KT recruitment of GFP-Mps1 during the G2 phase (Figure 7C; Movie S6). Since Plk1 localizes to KTs already during the S and G2 phases (McKinley and Cheeseman, 2014) and is at least partially active several hours before mitotic onset (Mac urek et al., 2008), it is attractive to speculate that this kinase helps to increase Mps1 turnover at KTs during the G2/M transition. The aforementioned data concur to suggest that phosphorylation of Mps1 by Plk1 contributes to stimulate Mps1 activity and strengthen SAC signaling. To further support this conclusion, we used phospho-specific antibodies against Thr676, an autophosphorylation site within the activation loop of Mps1 (Jelluma et al., 2008), to monitor Mps1 activity in response to Plk1 phosphorylation. While inhibition of either Plk1 or Mps1 during SAC establishment partially suppressed Thr676 phosphorylation at KTs, simultaneous inactivation of both kinases reduced staining below the detection limit (Figures 7D and S7B). Identical results were also observed in the absence of nocodazole (Figure 7E), and similar results were obtained with an antibody that detects phosphorylated Thr686 (Tyler et al., 2009) (Figure S7C), an autophosphorylation site essential for activity located within the P + 1 loop (Jelluma et al., 2008; Mattison et al., 2007; Tyler et al., 2009). Thus, use of phospho-specific antibodies strongly supports the conclusion that Plk1 potentiates Mps1 functionality in the SAC through direct phosphorylation of Mps1, which, in turn, increases Mps1 activity and turnover at KTs. DISCUSSION The specificity of kinases and phosphatases for their physiological substrates stays at the heart of all signaling pathways. This specificity is largely determined by two key parameters that have been referred to as motif space and localization space (Alexander et al., 2011). According to this model, the target range

Figure 7. Plk1 Enhances Mps1 Kinase Activity and KT Dissociation (A) Mps1 null cells ectopically expressing GFP-tagged Mps1 were treated with nocodazole, MG132, and kinase inhibitors (100 nM BI-2536, 500 nM Reversine [Rev], 5 mM ZM-447439) before GFP-Mps1 KT levels were recorded by live-cell microscopy. Top: scatterplot shows intensity of GFP signals (n = 157–241 KTs from 45–50 cells per condition). Bottom: representative GFP KT signals. (B) Cells described in (A) were incubated with nocodazole, MG132, and kinase inhibitors, and GFP-Mps1 KT levels were monitored by 500-ms time-lapse microscopy (Movie S5). After 2 s, a single KT pair was bleached (top left), and fluorescence recovery was measured (bottom). Top right: traces illustrate fluorescence recovery (n = 17–20 KT pairs per condition); shaded areas represent SEM. t1/2, half-time; plat., plateau. (C) Asynchronously growing cells, as used in (A and B), were incubated with nocodazole and kinase inhibitors before fluorescence at KTs was monitored by timelapse microscopy (Movie S6). Left: stills show data obtained from
60 min (G2 phase) to +10 min (M phase) relative to the time of mitotic entry (t = 0). Arrowheads indicate premature KT association of GFP-Mps1. Right: scatterplot shows GFP-Mps1 intensity at KTs (50 KTs from ten cells per condition). (D) RPE-1 cells were synchronized in metaphase using MG132 and subsequently treated with nocodazole and kinase inhibitors. Mps1 activation state was assessed by staining with an antibody directed at phospho-Thr676. Histogram shows staining intensity (n = 75–79 cells per condition; micrographs are shown in Figure S7B). Error bars indicate SEM. rel. signal int., relative signal intensity. (E) Asynchronously growing RPE-1 cells were incubated for 2 hr with kinase inhibitors in the absence of nocodazole and then stained for Mps1 activation loop phosphorylation (P-Thr676). IF staining (right) shows KT levels of P-Thr676; scatterplot (left) indicates P-Thr676 staining intensities at KTs, determined at the indicated mitotic stages (n = 19–38 cells per condition). Mitotic phases were classified as described in the legend for Figure 2. fluor. int., fluorescence intensity. (F) Schematic illustrating the proposed involvement of Plk1 in SAC signaling at KTs. P, phosphorylation. Scale bars, 5 mm. *p = 0.0115; **p = 0.0013; ***p < 0.001 (from Student’s t tests). See also Figure S7B and Movies S5 and S6.

Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 75

of kinases depends on, first, consensus motifs within substrate proteins and, second, subcellular localization. Recently, two key mitotic kinases, Plk1 and Mps1, were found to display an intriguing overlap in both localization and motif space (Dou et al., 2011; Hennrich et al., 2013). In this study, we have addressed the question of how to reconcile this discovery with the apparently different physiological functions of these kinases. We show that Plk1 cooperates with Mps1 in SAC signaling, although this role is generally obscured by other functions of Plk1—notably, its role in bipolar spindle formation. We further demonstrate that Plk1 makes positive contributions to two major KT-associated branches of the SAC signaling pathway. In particular, we have identified two key SAC components that are phosphorylated on identical sites by Plk1 and Mps1—notably, KNL-1 and Mps1 itself. While phosphorylation of KNL-1 on MELT motifs triggers the recruitment of the Bub3:BubR1 complex, phosphorylation of Mps1 contributes to enhance Mps1 activity and turnover at KTs. This Mps1 activation, in turn, is essential for the recruitment of the Mad1:C-Mad1 complex to KTs. While additional substrates shared between Plk1 and Mps1 are likely to await discovery, our present data suggest cooperation between the two kinases in both KT pathways that are central to the production of the APC/C-inhibitory MCC (Figure 7F). It has been argued that the SAC operates as a bi-stable switch (Rieder and Maiato, 2004), but recent studies demonstrate that the strength of SAC signaling is proportional to the number of unattached KTs (Collin et al., 2013; Dick and Gerlich, 2013; Heinrich et al., 2013). As shown here, Plk1 contributes to the augmentation of the strength of SAC signaling, which is expected to sensitize cellular responses to unattached KTs. This may be of particular importance at very early stages of mitosis, when MCC production needs to ramp up, or shortly before anaphase onset, when only a few unattached KTs remain available for MCC production. A similar function has been proposed for Aurora B, which potentiates SAC signaling at the onset of mitosis by recruiting Mps1 to KTs (Saurin et al., 2011). It is interesting to consider the cooperation between Plk1 and Mps1 from an evolutionary perspective. One plausible scenario is that the relative contributions of Mps1 and Plk1 to SAC signaling differ between organisms. An extreme example for this can be seen in Caenorhabditis elegans, whose genome lacks a recognizable gene for Mps1, although both Mad1:C-Mad2 and Bub3:Bub1/BubR1, as well as KNL-1 and MELT motifs, are clearly present (Vleugel et al., 2012). An independent study demonstrates that Plk1 phosphorylates the KNL-1 MELT repeats in C. elegans, suggesting that functions commonly attributed to Mps1 in vertebrates are performed by Plk1 in this nematode (Espeut et al., 2015), fully in line with the data reported here. Still another situation holds in Drosophila melanogaster, where KNL-1 MELT motifs are mostly degenerate, unphosphorylatable, and largely dispensable for SAC function (Schittenhelm et al., 2009). However, in Drosophila, Mad1 localization also depends on Mps1, and Polo contributes to the recruitment of Mps1 to KTs (Althoff et al., 2012; Conde et al., 2013). Collectively, these data suggest that the contribution of Plk1 to SAC signaling is likely to be auxiliary to Mps1 in many organisms or cell types but apparently indispensable in others. This consideration may also be relevant to tumorigenesis. Although many human tumors show chromosomal insta76 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

bility, aneuploid tumor cells still depend on SAC functionality to prevent lethality (Holland and Cleveland, 2009). It is remarkable, therefore, that high levels of Mps1 were shown to protect aneuploid cells (Daniel et al., 2011) and that elevated expression of Plk1 is often observed in human cancers (Strebhardt, 2010). Therefore, it is tempting to speculate that the cooperation between Mps1 and Plk1 may be important to strengthen the SAC in tumors, which may prove particularly critical in cells that undergo mitosis in the presence of hyperdiploid numbers of chromosomes. If this were the case, the simultaneous inhibition of Mps1 and Plk1 might constitute an attractive strategy for anti-cancer therapy. EXPERIMENTAL PROCEDURES rAAV-Mediated Gene Targeting Homology arms to human Mad2 (MAD2L1), Aurora B (AURKB), and Bub3 (BUB3L) genes were amplified from RPE-1 cell genomic DNA. Targeting constructs were assembled by four-piece ligation in a NotI-digested pAAV vector. Recombinant adenovirus-associated virus (rAAV) particles were generated as previously described (Berdougo et al., 2009). RPE-1 cells were infected with 3 ml of viral supernatant for 48 hr and then expanded into fresh medium for an additional 48 hr. They were synchronized with RO-3306 (10 mM) for 18 hr, released into nocodazole (50 nM) for 2 hr, and then subjected to fluorescence-activated cell sorting (FACS). Cells were trypsinized and kept in PBS supplemented with nocodazole (10 nM) during sorting. Infected cells were filtered (30 mm, Partec), and EGFP-positive cells (488 excitation, 514/30 emission filter) were isolated on an Aria III (BD Biosciences) cell sorter, as previously described (Collin et al., 2013). In brief, cells were selected using the 514/30 channel against a 585/42 filter detecting auto-fluorescence of uninfected cells. Single cells were sorted into 96-well plates filled with conditioned medium, and positive clones were identified by fluorescence microscopy. FRET Reporter Cloning FRET reporters used in this study are based on a sensor design in which phosphorylation of a target sequence triggers the binding of a phosphate-binding (FHA2) domain and concomitant changes in intra-molecular FRET between CFP and yellow fluorescent protein (YFP). Sensors harboring the KNL-1 MELT motif around Thr875 (or Ala for control) were constructed by QuickChange mutagenesis (Agilent Technologies) of a plasmid encoding Aurora B H2B-fused biosensors (Fuller et al., 2008). The Aurora B sequence in the reporter was replaced with the sequence SEDDKNDMDI(T/A)KSI; the Ile in position +3 relative to the phospho-acceptor was included to promote FHA2 domain binding. Standard Procedures Cell culture, transfections, western blotting, immunoprecipitation, fluorescence microscopy, in vitro kinase assays, and mass spectrometry were carried out as described in detail in the Supplemental Experimental Procedures. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, seven figures, one table, and six movies and can be found with this article online at http://dx.doi.org/10.1016/j.celrep.2015.06.007. AUTHOR CONTRIBUTIONS C.v.S., G.J.P.L.K., and E.A.N. conceived experiments; C.v.S., F.C., J.M.B., and T.S. contributed data; and C.v.S. and E.A.N. wrote the manuscript. ACKNOWLEDGMENTS We thank the Proteomics Core Facility (A. Schmidt and T. Glatter) and the Imaging Core Facility (O. Biehlmaier, A. Loynton-Ferrand, and N. Ehrenfeuchter) of the Biozentrum for expert advice. We also thank P. Jallepalli, M. Burkard, J.

Pines, S. Taylor, M. Lampson, D. Gerlich, I. Cheeseman, Y. Watanabe, P. Eyers, and the late L. Johnson for reagents and cell lines. Finally, we thank Bayer Pharma AG for providing TAL; P. Jallepalli and J. Pines for advice on AAV-mediated gene targeting; and the E.A.N. lab for helpful discussions. This work was supported by the University of Basel, the Swiss National Science Foundation (310030B_149641), and the Netherlands Organisation for Scientific Research (NWO-Vici 865.12.004).

nal SCX chromatography and stable isotope dimethyl labeling provides highdefinition kinase consensus motifs for PKA and human Mps1. J. Proteome Res. 12, 2214–2224.

Received: March 23, 2015 Revised: May 14, 2015 Accepted: June 1, 2015 Published: June 25, 2015

Holland, A.J., and Cleveland, D.W. (2009). Boveri revisited: chromosomal instability, aneuploidy and tumorigenesis. Nat. Rev. Mol. Cell Biol. 10, 478– 487.

REFERENCES

Hewitt, L., Tighe, A., Santaguida, S., White, A.M., Jones, C.D., Musacchio, A., Green, S., and Taylor, S.S. (2010). Sustained Mps1 activity is required in mitosis to recruit O-Mad2 to the Mad1-C-Mad2 core complex. J. Cell Biol. 190, 25–34.

Jelluma, N., Brenkman, A.B., McLeod, I., Yates, J.R., 3rd, Cleveland, D.W., Medema, R.H., and Kops, G.J.P.L. (2008). Chromosomal instability by inefficient Mps1 auto-activation due to a weakened mitotic checkpoint and lagging chromosomes. PLoS ONE 3, e2415.

Alexander, J., Lim, D., Joughin, B.A., Hegemann, B., Hutchins, J.R.A., Ehrenberger, T., Ivins, F., Sessa, F., Hudecz, O., Nigg, E.A., et al. (2011). Spatial exclusivity combined with positive and negative selection of phosphorylation motifs is the basis for context-dependent mitotic signaling. Sci. Signal. 4, ra42.

Jelluma, N., Dansen, T.B., Sliedrecht, T., Kwiatkowski, N.P., and Kops, G.J.P.L. (2010). Release of Mps1 from kinetochores is crucial for timely anaphase onset. J. Cell Biol. 191, 281–290.

Althoff, F., Karess, R.E., and Lehner, C.F. (2012). Spindle checkpoint-independent inhibition of mitotic chromosome segregation by Drosophila Mps1. Mol. Biol. Cell 23, 2275–2291.

Krenn, V., Overlack, K., Primorac, I., van Gerwen, S., and Musacchio, A. (2014). KI motifs of human Knl1 enhance assembly of comprehensive spindle checkpoint complexes around MELT repeats. Curr. Biol. 24, 29–39.

Berdougo, E., Terret, M.-E., and Jallepalli, P.V. (2009). Functional dissection of mitotic regulators through gene targeting in human somatic cells. Methods Mol. Biol. 545, 21–37.

Lara-Gonzalez, P., Westhorpe, F.G., and Taylor, S.S. (2012). The spindle assembly checkpoint. Curr. Biol. 22, R966–R980.

Burkard, M.E., Randall, C.L., Larochelle, S., Zhang, C., Shokat, K.M., Fisher, R.P., and Jallepalli, P.V. (2007). Chemical genetics reveals the requirement for Polo-like kinase 1 activity in positioning RhoA and triggering cytokinesis in human cells. Proc. Natl. Acad. Sci. USA 104, 4383–4388. Collin, P., Nashchekina, O., Walker, R., and Pines, J. (2013). The spindle assembly checkpoint works like a rheostat rather than a toggle switch. Nat. Cell Biol. 15, 1378–1385. Conde, C., Osswald, M., Barbosa, J., Moutinho-Santos, T., Pinheiro, D., Guimara˜es, S., Matos, I., Maiato, H., and Sunkel, C.E. (2013). Drosophila Polo regulates the spindle assembly checkpoint through Mps1-dependent BubR1 phosphorylation. EMBO J. 32, 1761–1777. Daniel, J., Coulter, J., Woo, J.-H., Wilsbach, K., and Gabrielson, E. (2011). High levels of the Mps1 checkpoint protein are protective of aneuploidy in breast cancer cells. Proc. Natl. Acad. Sci. USA 108, 5384–5389. De Antoni, A., Maffini, S., Knapp, S., Musacchio, A., and Santaguida, S. (2012). A small-molecule inhibitor of Haspin alters the kinetochore functions of Aurora B. J. Cell Biol. 199, 269–284.

Liu, X., and Winey, M. (2012). The MPS1 family of protein kinases. Annu. Rev. Biochem. 81, 561–585. London, N., Ceto, S., Ranish, J.A., and Biggins, S. (2012). Phosphoregulation of Spc105 by Mps1 and PP1 regulates Bub1 localization to kinetochores. Curr. Biol. 22, 900–906. Maciejowski, J., George, K.A., Terret, M.-E., Zhang, C., Shokat, K.M., and Jallepalli, P.V. (2010). Mps1 directs the assembly of Cdc20 inhibitory complexes during interphase and mitosis to control M phase timing and spindle checkpoint signaling. J. Cell Biol. 190, 89–100. Mac urek, L., Lindqvist, A., Lim, D., Lampson, M.A., Klompmaker, R., Freire, R., Clouin, C., Taylor, S.S., Yaffe, M.B., and Medema, R.H. (2008). Polo-like kinase-1 is activated by aurora A to promote checkpoint recovery. Nature 455, 119–123. arek, J., Paul, R., Mogilner, A., and KhodMagidson, V., O’Connell, C.B., Lonc jakov, A. (2011). The spatial arrangement of chromosomes during prometaphase facilitates spindle assembly. Cell 146, 555–567.

Dick, A.E., and Gerlich, D.W. (2013). Kinetic framework of spindle assembly checkpoint signalling. Nat. Cell Biol. 15, 1370–1377.

Maldonado, M., and Kapoor, T.M. (2011). Constitutive Mad1 targeting to kinetochores uncouples checkpoint signalling from chromosome biorientation. Nat. Cell Biol. 13, 475–482.

Dou, Z., von Schubert, C., Ko¨rner, R., Santamaria, A., Elowe, S., and Nigg, E.A. (2011). Quantitative mass spectrometry analysis reveals similar substrate consensus motif for human Mps1 kinase and Plk1. PLoS ONE 6, e18793.

Mattison, C.P., Old, W.M., Steiner, E., Huneycutt, B.J., Resing, K.A., Ahn, N.G., and Winey, M. (2007). Mps1 activation loop autophosphorylation enhances kinase activity. J. Biol. Chem. 282, 30553–30561.

Espeut, J., Lara-Gonzalez, P., Sassine, M., Shiau, A.K., Desai, A., and Abrieu, A. (2015). Natural loss of Mps1 kinase in nematodes uncovers a role for Pololike kinase 1 in spindle checkpoint initiation. Cell Rep. 12, Published online June 25, 2015. http://dx.doi.org/10.1016/j.celrep.2015.05.039.

McKinley, K.L., and Cheeseman, I.M. (2014). Polo-like kinase 1 licenses CENP-A deposition at centromeres. Cell 158, 397–411. Musacchio, A., and Salmon, E.D. (2007). The spindle-assembly checkpoint in space and time. Nat. Rev. Mol. Cell Biol. 8, 379–393.

Fuller, B.G., Lampson, M.A., Foley, E.A., Rosasco-Nitcher, S., Le, K.V., Tobelmann, P., Brautigan, D.L., Stukenberg, P.T., and Kapoor, T.M. (2008). Midzone activation of aurora B in anaphase produces an intracellular phosphorylation gradient. Nature 453, 1132–1136.

Nakajima, H., Toyoshima-Morimoto, F., Taniguchi, E., and Nishida, E. (2003). Identification of a consensus motif for Plk (Polo-like kinase) phosphorylation reveals Myt1 as a Plk1 substrate. J. Biol. Chem. 278, 25277–25280.

Funabiki, H., and Wynne, D.J. (2013). Making an effective switch at the kinetochore by phosphorylation and dephosphorylation. Chromosoma 122, 135–158.

Oppermann, F.S., Grundner-Culemann, K., Kumar, C., Gruss, O.J., Jallepalli, P.V., and Daub, H. (2012). Combination of chemical genetics and phosphoproteomics for kinase signaling analysis enables confident identification of cellular downstream targets. Mol. Cell. Proteomics 11, O111.012351.

Heinrich, S., Geissen, E.-M., Kamenz, J., Trautmann, S., Widmer, C., Drewe, P., Knop, M., Radde, N., Hasenauer, J., and Hauf, S. (2013). Determinants of robustness in spindle assembly checkpoint signalling. Nat. Cell Biol. 15, 1328–1339.

Overlack, K., Primorac, I., Vleugel, M., Krenn, V., Maffini, S., Hoffmann, I., Kops, G.J.P.L., and Musacchio, A. (2015). A molecular basis for the differential roles of Bub1 and BubR1 in the spindle assembly checkpoint. eLife 4, e05269.

Hennrich, M.L., Marino, F., Groenewold, V., Kops, G.J.P.L., Mohammed, S., and Heck, A.J.R. (2013). Universal quantitative kinase assay based on diago-

Petronczki, M., Le´na´rt, P., and Peters, J.-M. (2008). Polo on the rise-from mitotic entry to cytokinesis with Plk1. Dev. Cell 14, 646–659.

Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors 77

Primorac, I., Weir, J.R., Chiroli, E., Gross, F., Hoffmann, I., van Gerwen, S., Ciliberto, A., and Musacchio, A. (2013). Bub3 reads phosphorylated MELT repeats to promote spindle assembly checkpoint signaling. eLife 2, e01030. Rieder, C.L., and Maiato, H. (2004). Stuck in division or passing through: what happens when cells cannot satisfy the spindle assembly checkpoint. Dev. Cell 7, 637–651. Rodriguez-Bravo, V., Maciejowski, J., Corona, J., Buch, H.K., Collin, P., Kanemaki, M.T., Shah, J.V., and Jallepalli, P.V. (2014). Nuclear pores protect genome integrity by assembling a premitotic and Mad1-dependent anaphase inhibitor. Cell 156, 1017–1031. Santaguida, S., Tighe, A., D’Alise, A.M., Taylor, S.S., and Musacchio, A. (2010). Dissecting the role of MPS1 in chromosome biorientation and the spindle checkpoint through the small molecule inhibitor reversine. J. Cell Biol. 190, 73–87. Santaguida, S., Vernieri, C., Villa, F., Ciliberto, A., and Musacchio, A. (2011). Evidence that Aurora B is implicated in spindle checkpoint signalling independently of error correction. EMBO J. 30, 1508–1519. Santamaria, A., Neef, R., Eberspa¨cher, U., Eis, K., Husemann, M., Mumberg, D., Prechtl, S., Schulze, V., Siemeister, G., Wortmann, L., et al. (2007). Use of the novel Plk1 inhibitor ZK-thiazolidinone to elucidate functions of Plk1 in early and late stages of mitosis. Mol. Biol. Cell 18, 4024–4036. Santamaria, A., Wang, B., Elowe, S., Malik, R., Zhang, F., Bauer, M., Schmidt, A., Sillje´, H.H.W., Ko¨rner, R., and Nigg, E.A. (2011). The Plk1-dependent phosphoproteome of the early mitotic spindle. Mol. Cell. Proteomics 10, M110.004457. Saurin, A.T., van der Waal, M.S., Medema, R.H., Lens, S.M.A., and Kops, G.J.P.L. (2011). Aurora B potentiates Mps1 activation to ensure rapid checkpoint establishment at the onset of mitosis. Nat. Commun. 2, 316. Schittenhelm, R.B., Chaleckis, R., and Lehner, C.F. (2009). Intrakinetochore localization and essential functional domains of Drosophila Spc105. EMBO J. 28, 2374–2386. Shepperd, L.A., Meadows, J.C., Sochaj, A.M., Lancaster, T.C., Zou, J., Buttrick, G.J., Rappsilber, J., Hardwick, K.G., and Millar, J.B.A. (2012). Phosphodependent recruitment of Bub1 and Bub3 to Spc7/KNL1 by Mph1 kinase maintains the spindle checkpoint. Curr. Biol. 22, 891–899.

78 Cell Reports 12, 66–78, July 7, 2015 ª2015 The Authors

Strebhardt, K. (2010). Multifaceted polo-like kinases: drug targets and antitargets for cancer therapy. Nat. Rev. Drug Discov. 9, 1–18. Tyler, R.K., Chu, M.L.H., Johnson, H., McKenzie, E.A., Gaskell, S.J., and Eyers, P.A. (2009). Phosphoregulation of human Mps1 kinase. Biochem. J. 417, 173–181. van der Horst, A., and Lens, S.M.A. (2014). Cell division: control of the chromosomal passenger complex in time and space. Chromosoma 123, 25–42. Vigneron, S., Prieto, S., Bernis, C., Labbe´, J.-C., Castro, A., and Lorca, T. (2004). Kinetochore localization of spindle checkpoint proteins: who controls whom? Mol. Biol. Cell 15, 4584–4596. Vleugel, M., Hoogendoorn, E., Snel, B., and Kops, G.J.P.L. (2012). Evolution and function of the mitotic checkpoint. Dev. Cell 23, 239–250. Vleugel, M., Tromer, E., Omerzu, M., Groenewold, V., Nijenhuis, W., Snel, B., and Kops, G.J.P.L. (2013). Arrayed BUB recruitment modules in the kinetochore scaffold KNL1 promote accurate chromosome segregation. J. Cell Biol. 203, 943–955. Vleugel, M., Omerzu, M., Groenewold, V., Hadders, M.A., Lens, S.M.A., and Kops, G.J.P.L. (2015). Sequential multisite phospho-regulation of KNL1BUB3 interfaces at mitotic kinetochores. Mol. Cell 57, 824–835. Wang, X., Yu, H., Xu, L., Zhu, T., Zheng, F., Fu, C., Wang, Z., and Dou, Z. (2014). Dynamic autophosphorylation of mps1 kinase is required for faithful mitotic progression. PLoS ONE 9, e104723. Xu, Q., Zhu, S., Wang, W., Zhang, X., Old, W., Ahn, N., and Liu, X. (2009). Regulation of kinetochore recruitment of two essential mitotic spindle checkpoint proteins by Mps1 phosphorylation. Mol. Biol. Cell 20, 10–20. Yamagishi, Y., Yang, C.-H., Tanno, Y., and Watanabe, Y. (2012). MPS1/Mph1 phosphorylates the kinetochore protein KNL1/Spc7 to recruit SAC components. Nat. Cell Biol. 14, 746–752. Yang, Z., Kenny, A.E., Brito, D.A., and Rieder, C.L. (2009). Cells satisfy the mitotic checkpoint in Taxol, and do so faster in concentrations that stabilize syntelic attachments. J. Cell Biol. 186, 675–684. Zhou, L., Tian, X., Zhu, C., Wang, F., and Higgins, J.M.G. (2014). Polo-like kinase-1 triggers histone phosphorylation by Haspin in mitosis. EMBO Rep. 15, 273–281.