The cytoplasmic linker protein (CLIP)-170, an outer kinetochore protein, has a role in kinetochore–microtubule attachment and chromosome alignment during mitosis. However, the mechanism by which CLIP-170 is involved in chromosome alignment is not known. Here, we show that CLIP-170 colocalizes with Polo-like kinase 1 (PLK1) at kinetochores during early mitosis. Depletion of CLIP-170 results in a significant reduction in PLK1 recruitment to kinetochores and causes kinetochore-fiber (K-fiber) instability and defects in chromosome alignment at the metaphase plate. These phenotypes are dependent on the phosphorylation of CLIP-170 at a CDK1-dependent site, T287, as ectopic expression of wild-type CLIP-170, but not the expression of a non-phosphorylatable mutant, CLIP-170-T287A, restores PLK1 localization at kinetochores and rescues K-fiber stability and chromosome alignment in CLIP-170-depleted cells. These data suggest that CLIP-170 acts as a novel recruiter and spatial regulator of PLK1 at kinetochores during early mitosis, promoting K-fiber stability and chromosome alignment for error-free chromosome segregation.
During cell division, faithful chromosome segregation is dependent on the efficient and correct formation of kinetochore–microtubule (KT–MT) attachments on the mitotic spindle, which facilitate the progress of chromosome alignment (Tanaka et al., 2005; Tanaka, 2012; Tanaka, 2013). Polo-like kinase 1 (PLK1) is a key regulator of mitotic division (Strebhardt, 2010). Disruption of PLK1 activity, either by RNAi or by chemical inhibition, prevents the formation of stable KT–MT attachments, leading to defects in chromosome alignment (Sumara et al., 2004; Hanisch et al., 2006; Peters et al., 2006; Lénárt et al., 2007; Petronczki et al., 2007). A more recent study has shown that PLK1 accumulates at kinetochores to stabilize the initial KT–MT attachments during prometaphase by suppressing kinetochore-microtubule dynamics, and that its levels decline markedly at kinetochores during metaphase, thus allowing stable kinetochore-microtubules to become dynamic (Liu et al., 2012).
Cytoplasmic linker protein (CLIP)-170, a microtubule plus-end-tracking and outer kinetochore protein, plays an essential role in mitosis (Dujardin et al., 1998; Wieland et al., 2004; Tanenbaum et al., 2006). Depletion of CLIP-170 in human cells leads to a mitotic block and defects in chromosome alignment (Wieland et al., 2004), due to a failure in the formation of KT–MT attachments (Tanenbaum et al., 2006). Thus, CLIP-170 has an important role in mitotic progression. CLIP-170 interacts with, and is directly phosphorylated by, PLK1 at its N-terminal region and casein kinase 2 (CK2) at its C-terminal region. Both phosphorylation events are required for the dynactin-dependent kinetochore localization of CLIP-170 and the formation of KT–MT attachments (Li et al., 2010). Moreover, CLIP-170 is phosphorylated on T287 by cyclin-dependent kinase 1 (CDK1) – an event that is essential for cell cycle progression (Yang et al., 2009). However, the mechanism of the role of CLIP-170 in chromosome alignment remains unknown. Here, we report that CLIP-170 functions in chromosome alignment by recruiting PLK1 to kinetochores during early mitosis and, in turn, by stabilizing kinetochore (K)-fibers – functions that are both mediated by phosphorylation of its CDK1-dependent site, T287.
RESULTS AND DISCUSSION
CLIP-170 promotes the localization of PLK1 at kinetochores during early mitosis, and their localization is interdependent
Live-cell imaging showed that CLIP-170 fused with mCherry appeared at kinetochores during early mitosis and disappeared from kinetochores immediately after KT–MT attachment (supplementary material Fig. S1A; Movie 1). Immunofluorescence of fixed cells showed that a strong kinetochore-associated CLIP-170 signal appeared in early mitosis and declined markedly during metaphase (supplementary material Fig. S1B,C). Immunofluorescence data also showed that CLIP-170 colocalized with PLK1 during prometaphase (Fig. 1A). Depletion of CLIP-170, confirmed by both western blotting and immunostaining (supplementary material Fig. S2A–C), caused a significant reduction in the amount of PLK1 recruitment to kinetochores compared with that observed during prometaphase, and even metaphase, in mock-treated cells (Fig. 1B,D). However, total PLK1 levels were unaffected (supplementary material Fig. S3A,B). This reduction in PLK1 recruitment to kinetochores was independent of microtubule attachment, as the level of kinetochore-associated PLK1 was also significantly reduced in CLIP-170-depleted nocodazole-treated cells compared with that of mock-treated cells (Fig. 1C,E). Furthermore, the level of CLIP-170 localization at kinetochores was also reduced in cells treated with nocodazole and a PLK1 inhibitor (Fig. 1F,G), consistent with a previous report (Tanenbaum et al., 2006). These data indicate that the localization of CLIP-170 and PLK1 at kinetochores is interdependent during early mitosis.
As several kinetochore-associated proteins, such as centromere protein E (CENP-E), depend on PLK1 for their localization at kinetochores (Ahonen et al., 2005), we checked whether CENP-E intensity at kinetochores was affected upon CLIP-170 depletion. Immunofluorescence data showed that kinetochore-associated CENP-E intensity was significantly reduced in nocodazole-treated and CLIP-170-depleted cells, compared with mock-treated cells (Fig. 1H,I). However, the total CENP-E levels remained unaffected by CLIP-170 depletion (supplementary material Fig. S3A,B). Kinetochore localization of two other outer kinetochore proteins, Hec1 (also known as NDC80) and BubR1 (also known as BUB1B), was not affected in CLIP-170-depleted cells, showing that outer kinetochore structure per se has not been altered (supplementary material Fig. S3C,D). By contrast, the kinetochore localization of CLIP-170 remained unchanged in CENP-E-depleted cells (supplementary material Fig. S3F,G). These data indicate that CLIP-170 is important for the localization of CENP-E at kinetochores.
The effect of the depletion of CLIP-170 on the kinetochore localization of PLK1 during early mitosis was confirmed by live-cell imaging analyses of HeLa cells expressing H2B–mCherry and YFP–PLK1 (Fig. 2A,B). In mock-treated cells, the kinetochore-associated PLK1 signal was strong from the time of nuclear envelope breakdown (NEBD) until the initial formation of bipolar spindles (∼8 minutes), and then the signal declined (Fig. 2A; supplementary material Movie 2). By contrast, in CLIP-170-depleted cells, the kinetochore-associated PLK1 signal was lower starting from NEBD, although the PLK1 signal at the centrosomes and nuclear envelope remained unchanged (Fig. 2B; supplementary material Movie 3). These data indicate that CLIP-170 is crucial for PLK1 localization at kinetochores during early mitosis. The reason why the kinetochore localization of PLK1 is already reduced at NEBD, when CLIP-170 does not localize to kinetochores, is currently unknown.
PLK1 recruitment to kinetochores is dependent on CDK1 phosphorylation of CLIP-170 at T287
CDK1 interacts with CLIP-170 and phosphorylates it on T287, and this is essential for cell cycle progression (Yang et al., 2009). A series of elegant studies has demonstrated that, by phosphorylating proteins, CDK1 generates PLK1-docking sites in early mitosis when CDK1 activity is high (Elia et al., 2003a; Elia et al., 2003b; Barr et al., 2004). Because the binding of PLK1 to substrates in early mitosis is often dependent on priming phosphorylation mediated by CDK1 (Kang et al., 2006; Qi et al., 2006), we checked whether the non-phosphorylatable mCherry–CLIP-170-T287A mutant influenced the recruitment of PLK1 to kinetochores in early mitosis. Rescue experiments and immunofluorescence data showed that wild-type mCherry-tagged CLIP-170 (mCherry–CLIP-170-WT), but not the mCherry–CLIP-170-T287A mutant, rescued the localization of PLK1 at kinetochores in CLIP-170-depleted cells (Fig. 3B,C). Kinetochore localization of mCherry–CLIP-170-T287A, which cannot be efficiently bound or phosphorylated by PLK1, was also reduced, compared with that of mCherry–CLIP-170-WT (Fig. 3D,E). This is consistent with the reduction in the kinetochore localization of CLIP-170 in PLK1-inhibited cells. Western blotting analyses demonstrated the efficiency of CLIP-170 depletion and the expression of RNAi-resistant rescue constructs (Fig. 3A). As CLIP-170 interacts with PLK1 in early mitosis (Li et al., 2010), we further tested whether this interaction was affected in cells expressing the non-phosphorylatable mCherry–CLIP-170-T287A mutant. Immunoprecipitation and western blot data showed that mCherry–CLIP-170-WT, but not the mCherry–CLIP-170-T287A mutant, could interact with PLK1 in cells arrested in mitosis (Fig. 3F). This interaction is thought to be through the Polo-box domain (PBD) in PLK1, as CLIP-170 was co-immunoprecipitated with FLAG-tagged wild-type PBD, but not with its binding-deficient mutant (Fig. 3G; Oshimori et al., 2006). The sequence context of T287 in CLIP-170 (STT287P) partially satisfies the optimal sequence motif recognized by the PBD, which is Ser-[pSer/pThr]-[Pro/X]. These data suggest that PLK1 recruitment to kinetochores is dependent on CLIP-170 phosphorylation on the CDK1-dependent site, T287, during early mitosis. We also found that the reduced kinetochore localization of CENP-E was restored by expression of wild-type CLIP-170, but not by that of the T287A mutant (supplementary material Fig. S3E). It has been reported that kinetochore localization of two other microtubule plus-end-tracking proteins, CLASP1 and CLASP2, is dependent on CENP-E (Maffini et al., 2009). Consistent with this, the kinetochore localization of CLASP1 and CLASP2 was also reduced in CLIP-170-depleted cells, and this was rescued by expressing wild-type CLIP-170, but not by the expression of the T287A mutant (supplementary material Fig. S3H,I). As CLASP2 has also been reported to recruit PLK1 to kinetochores (Maia et al., 2012), CLIP-170 might recruit PLK1 to kinetochores both directly and also indirectly through the recruitment of CLASP2.
CLIP-170 phosphorylation on a CDK1-dependent site, T287, is required for K-fiber stability and chromosome alignment
Knockdown of CLIP-170 causes defects in KT–MT attachment and chromosome alignment in human cells (supplementary material Fig. S2C; Tanenbaum et al., 2006). We confirmed that the formation of cold-stable K-fibers was compromised in CLIP-170-depleted cells in nocodazole-washout assays (supplementary material Fig. S4). To test whether CLIP-170 phosphorylation on T287 might play a role in KT–MT attachment, we examined the stability of K-fibers during cold treatment using an approach described previously (Itoh et al., 2011; Itoh et al., 2013). Immunofluorescence data showed that cells expressing mCherry–CLIP-170-T287A had unstable K-fibers and unattached chromosomes, whereas cells expressing mCherry–CLIP-170-WT had robust and stable K-fibers after cold treatment (Fig. 4A,B). These data suggest that CLIP-170 phosphorylation on T287 is required for the formation of stable K-fibers.
Next, we examined whether the phosphorylation of CLIP-170 on T287 had an essential function in chromosome alignment. In the presence of the proteasome inhibitor MG132, the defects in chromosome alignment at the metaphase plate were rescued in >60% of cells expressing mCherry–CLIP-170-WT; approximately similar to results from mock-treated cells (>70%; Fig. 4C). By contrast, in total, ∼60% of mitotic cells expressing the mCherry–CLIP170-T287A mutant had misaligned chromosomes (Fig. 4C). We further confirmed the data obtained from fixed cells by using live-cell imaging analyses. HeLa cells expressing mCherry–CLIP-170-WT underwent chromosome alignment and segregation within a standard period (∼30 minutes for alignment), whereas cells expressing the mCherry–CLIP-170-T287A mutant could not align chromosomes even after 3 hours of recording (Fig. 4D; supplementary material Movies 4–6). Thus, it can be concluded that the phosphorylation of CLIP-170 on the CDK1-dependent site, T287, is required for chromosome alignment.
Our data show that CLIP-170 is required for targeting PLK1 to kinetochores during early mitosis, and they thus reveal a mechanism by which the loss of CLIP-170 induces defects in the stability of the KT–MT attachments and in chromosome alignment. We propose the following model to explain the roles of CLIP-170 phosphorylation during early mitosis (Fig. 4E). After entry into mitosis, CDK1 activity is high, and it induces phosphorylation of CLIP-170 and other mitotic proteins, such as Bub1 (Qi et al., 2006), INCENP (Goto et al., 2006), MCAK (also known as KIF2C) (Rosasco-Nitcher et al., 2008) and CLASP2 (Maia et al., 2012). Then, PLK1 interacts with phosphorylated (phospho)-CLIP-170, followed by recruitment of the resulting PLK1–phospho-CLIP-170 complex to outer kinetochores through dynactin (Li et al., 2010). PLK1 in the complex possibly phosphorylates CLIP-170 at S195, which is required for the kinetochore localization of CLIP-170 (Li et al., 2010). The primary candidate for regulation by the PLK1 that is recruited to kinetochores through CLIP-170 is CLIP-170 itself. CLIP-170 is thought to facilitate KT–MT attachment by binding to microtubules through its N-terminus, which contains cytoskeleton-associated protein-glycine-rich (CAP-Gly) domains (Tanenbaum et al., 2006). Recently, it was reported that PLK1 also phosphorylates CLIP-170 at S312, which regulates the binding of CLIP-170 to microtubules and is pivotal for chromosome alignment (Kakeno et al., 2014). Therefore, PLK1 recruitment by CLIP-170 is expected to promote KT–MT attachment directly through CLIP-170 by regulating the kinetochore localization and microtubule-binding activity of the latter. Other nearby molecules, e.g. CENP-E, might also be regulated by CLIP-170 through PLK1 recruitment to kinetochores. As CENP-E (a microtubule-dependent plus-end-directed motor protein) is known to be essential for stable KT–MT interactions, as well as for aligning chromosomes that localize close to spindle poles to the metaphase plate (Wood et al., 1997; Kapoor et al., 2006; Kim et al., 2008; Kim et al., 2010; Cai et al., 2009), it might cooperate with CLIP-170 to facilitate the formation of KT–MT attachments.
Localized PLK1 activity is essential for the formation of stable KT–MT attachments and for chromosome alignment (Sumara et al., 2004; Hanisch et al., 2006; Matsumura et al., 2007). Therefore, it is not surprising that the kinetochore localization of PLK1 is regulated by multiple factors, as is already known (Qi et al., 2006; Kang et al., 2006; Nishino et al., 2006; Maia et al., 2012; Yeh et al., 2013). In this study, we have revealed a further layer of the elaborate network that regulates PLK1 function. Interestingly, CLIP-170 might contribute to the efficient recruitment of PLK1 to kinetochores not only directly but also indirectly, by recruiting CLASP2 to kinetochores. Residual PLK1 signal on kinetochores in CLIP-170-depleted cells might represent the contribution of other pathways of PLK1 recruitment to kinetochores. The relative contributions of different pathways for PLK1 recruitment to kinetochores to facilitate KT–MT attachment and chromosome alignment will be an important topic for future studies.
MATERIALS AND METHODS
Cell culture and drug treatments
HeLa cells were grown at 37°C in a humidified atmosphere under 5% CO2, in DMEM supplemented with 10% fetal bovine serum. Nocodazole (2 µM) and MG132 (10 µM) were used to arrest cells in prometaphase and metaphase, respectively. The PLK1 inhibitor BI2536 was used at 100 nM for 1 hour.
The cDNA coding for human CLIP-170 was PCR-amplified from a MegaManTM library (Agilent). The PCR product for CLIP-170 was subcloned into the pmCherry-C1 (Clontech) expression vector. The RNAi-resistant vectors, such as those encoding mCherry–CLIP-170-WT and mCherry–CLIP-170-T287A, were prepared using the vector encoding mCherry–CLIP-170 as a template vector.
The following primary antibodies were used for immunofluorescence: rabbit anti-CLIP-170 (Santa Cruz Biotechnology), mouse and rabbit anti-α-tubulin (Sigma-Aldrich), mouse anti-PLK1 (Santa Cruz Biotechnology), rabbit anti-GFP (Invitrogen), rabbit anti-CLASP1 (Epitomics), rabbit anti-CLASP2 (Santa Cruz Biotechnology), mouse anti-mCherry (Clontech), rabbit anti-mCherry (Abnova), mouse anti-CENP-E (Abcam), rabbit anti-BubR1 (Novus), mouse anti-Hec1 (Abcam) and mouse anti-FLAG (Sigma-Aldrich).
RNAi and rescue experiments
The targeted sequences for human CLIP-170 and CENP-E were 5′-AUUUCUAGCAGCAUGGACUGUUCCC-3′ and 5′-UUAUAUUACAGCCUUCCUUGAGCCG-3′, respectively (Invitrogen). HeLa cells were transfected with 150 nM duplexed siRNA by using RNAi MAX (Invitrogen) according to the manufacturer's instructions. As a control, a mock transfection reaction was performed using H2O instead of siRNA oligonucleotides. Cells were prepared for analysis after 72 hours of transfection. For rescue assays, RNAi-resistant CLIP-170 constructs, prepared by introducing six silent mutations using the Mutagenesis Basal Kit (Takara), were transfected into cells after 24 hours of siRNA transfection using Lipofectamine 2000 (Life Technologies) according to the manufacturer's protocol. Cells were prepared for analysis after 48 hours of plasmid transfection.
HeLa cells fixed in cold methanol or 4% paraformaldehyde at room temperature were blocked with 3% bovine serum albumin (BSA) in PBS, and incubated with the specified primary antibodies at a dilution of 1∶300 (see ‘Antibodies’, above). The primary antibodies were detected by using Alexa-Fluor-labeled secondary antibodies at a dilution of 1∶500 (Molecular Probes), and DNA was counterstained with 1 mg/ml DAPI. Analyses of cold-stable microtubules were performed as reported previously (Itoh et al., 2011; Itoh et al., 2013).
Image acquisition and analysis
Image acquisition was performed on a DeltaVision Personal DV microscope (Applied Precision) equipped with a camera (CoolSNAP HQ2, Photometrics) and a 100× or 60× NA 1.40 Plan Apochromat objective lens (Olympus), and SoftWoRx acquisition software (Applied Precision). For fixed-cell experiments, images were acquired as z-stacks at 0.2-µm intervals. To measure the fluorescence intensity, we selected manually and measured individual kinetochores by quantification of the pixel gray levels of the focused z plane in a region of interest (ROI) using ImageJ (NIH). After subtracting the background, which was measured outside of the ROI, intensities were averaged over all kinetochores in a single cell. The average intensity was normalized to the expression level by dividing by the peak intensity for that cell. For the measurement of K-fiber intensity, we converted identically scaled images into TIFF files, selected ROIs at the equatorial position of the metaphase plate and measured microtubule intensity as described above. For live-cell imaging, HeLa cells, grown in chambered coverslips (Thermo Fisher Scientific), were imaged at 24 or 48 hours after DNA transfection in Leibovitz's L-15 medium (Invitrogen), supplemented with 20% FBS and 20 mM HEPES pH 7.0 at 37°C within an environmental chamber (Precision Control). Z-stacks were collected at 0.8–1.2-µm intervals every 1–6 minutes. All images represent maximum-intensity projections of all z planes. All images and movies presented were processed in ImageJ, and the statistical tests performed are specified in the figure legends.
Immunoprecipitation and western blotting
Cell lysates were prepared with lysis buffer [20 mM Tris-HCl pH 7.5, 150 mM NaCl, 20 mM β-glycerophosphate, 5 mM MgCl2, 0.1% NP-40, 5% glycerol, 30 µg/ml DNase, 30 µg/ml RNase, complete protease inhibitor cocktail (Roche) and complete phosphatase inhibitor (Nacalai Tesque)]. Western blotting was performed as described previously (Itoh, et al., 2013). For immunoprecipitation assays, pre-cleared native protein extracts (1 mg of total protein) were incubated overnight with mouse anti-mCherry at a dilution of 1∶100 (Clontech) or mouse nonspecific IgG at a dilution of 1∶100, followed by incubation with 40 µl of anti-mouse-IgG magnetic Dynabeads (Invitrogen) for 2 hours at 4°C. After washing three times with lysis buffer and once with cold PBS, precipitated proteins were eluted by boiling for 5 minutes in 2× SDS sample buffer and were then analyzed by electrophoresis and western blotting.
The authors thank Kensaku Mizuno (Tohoku University, Sendai, Japan) for YFP-PLK1 construct, Miho Ohsugi (University of Tokyo, Japan) for FLAG-PBD constructs and Aoi Harata (Tohoku University, Sendai, Japan) for technical assistance.
M.A.A. and K.T. designed the study. M.A.A. performed the experiments and analyzed the data. M.A.A. wrote the paper. G.I. prepared some constructs; K.I. and M.I. prepared some parts of the microscopy data. K.T. supervised and revised all experimental data and the entire manuscript.
This work was supported by a Grant-in-Aid for Scientific Research and Japan Society for the Promotion of Science fellows from the Ministry of Education, Culture, Sports, Science and Technology of Japan; and grants from the Mitsubishi Foundation; the Naito Foundation; Takeda Science Foundation; Princess Takamatsu Cancer Research Fund [grant number 10-24210]; Daiichi-Sankyo Foundation of Life Science; and Gonryo Medical Foundation.
The authors declare no competing interests.