The inhibitor of apoptosis protein survivin is implicated in two key biological events: in the control of cell proliferation and in the regulation of cell lifespan. Although the details of mitotic roles of survivin are unclear, the protein appears to modulate microtubule function and might participate in regulating the spindle checkpoint. Survivin physically associates with Aurora B, a serine-threonine kinase involved in microtubule attachment to centromeres and regulation of chromosome segregation. Here we have examined the dynamics and localization of a survivin-GFP chimera using high-resolution fluorescence microscopy and photobleaching. Survivin forms a bi-partite structure at the inner centromere that undergoes significant stretching during mitosis. Photobleaching experiments revealed marked changes in rates of survivin turnover at centromeres. These were regulated by stage of the cell cycle, microtubule attachment, and Aurora B kinase activity. We hypothesize that changes in the turnover of survivin at centromeres influence the stability of kinetochore-microtubule attachment and signaling of the spindle checkpoint.

Introduction

Unbalanced chromosome segregation in mitosis contributes to cancer malignancy. The pathways that regulate the fidelity of chromosome segregation are incompletely understood. Survivin, a human member of the inhibitor of apoptosis protein (IAP) family, is undetectable in terminally differentiated adult tissues, but is expressed in dividing cells (Chiou et al., 2003; Altieri, 2003). Survivin is highly overexpressed in transformed cell lines and in human neoplasms of lung, colon, pancreas, prostate and breast (Chiou et al., 2003; Altieri, 2003). Survivin is closely related to the BIR motif proteins of budding yeast, fission yeast (BIR1 and bir1, respectively), and Caenorhabditis elegans (BIR-1 and BIR-2) (Reed and Bischoff, 2000; Uren et al., 1998). In contrast to other IAPs, which have primary roles in protection from apoptosis, members of this distinct subgroup of BIR motif proteins have essential roles in spindle function and cell cleavage (Li et al., 1998; Fraser et al., 1999; Giodini et al., 2002).

The mitotic localization of one subclass of survivin is consistent with the proteins described as chromosomal passenger proteins (Skoufias et al., 2000; Adams et al., 2001a). Survivin physically interacts with two other passenger proteins, Aurora B kinase and inner centromere protein INCENP in vitro (Wheatley et al., 2001; Honda et al., 2003) and in vivo (Bolton et al., 2002). Recently published data suggest that vertebrate Aurora B and its yeast homolog Ipl1 are required for the release of inappropriate microtubule-kinetochore interactions (Biggins and Murray, 2001; Tanaka et al., 2002; Kallio et al., 2002b). Vertebrate Aurora B also has a distinct role in direct signaling of the spindle checkpoint (Kallio et al., 2002b). Cells in which survivin, Aurora B, or INCENP are perturbed exhibit phenotypic similarities suggesting that these proteins function in common pathways (Uren et al., 2000; Adams et al., 2001a). Moreover, deletion of one member from the complex results in mislocalization and malfunction of the other two proteins (Speliotes et al., 2000; Adams et al., 2001b; Wheatley et al., 2001).

Many of the regulatory proteins implicated in the control of cell division associate with centromeres and centrosomes, often at specific times in the cell cycle (reviewed by Musacchio and Hardwick, 2002). To understand the mitotic functions of survivin we examined dynamics of a survivin-GFP chimera at different subcellular locations and at different cell-cycle phases in HeLa and LLC-PK cells using high-resolution fluorescence microscopy and fluorescence recovery after photobleaching (FRAP).

Materials and Methods

Materials and cell culture

All chemicals were purchased from Sigma and tissue culture reagents from Gibco unless stated otherwise. Cell culture was performed as described previously (Kallio et al., 2002b). For some experiments cells were treated with nocodazole (8 μM) or taxol (3 μM) from 15 minutes to 14 hours or with 2 μM ZM-447439 (a kind gift from Astra-Zeneca), a small molecule inhibitor of Aurora B (Ditchfield et al., 2003), for 10-60 minutes in the presence of 20 μM MG132, a proteasome inhibitor. In order to synchronize cells at S phase, aphidicolin was used at 1 μg/ml for 16 hours after which the cells were released from the arrest by extensive buffer washes. Cells growing on coverslips were fixed at hourly time points (0-12 hours) after aphidicolin wash out and were then processed for immunofluorescence.

Transfection and expression of survivin-GFP

Wild-type survivin was inserted into pEGFPc1 (Clontech Laboratories). LLC-PK and HeLa cells were transiently transfected with plasmids encoding full-length survivin-GFP or empty GFP vector with Fugene 6 Reagent (Roche Diagnostics) according to the manufacturer's recommendation. At 24-72 hours after initiation of the transfection protocol, the cells were fixed and processed for microscopic analysis, subjected to biochemical studies, or live cell video analysis. To assess the extent to which expression of survivin-GFP increased the total pool of survivin, we compared the transfected cells with non-transfected cells by quantitative immunofluorescence with anti-survivin antibody. We measured the average integrated intensity of total cellular survivin fluorescence from 10 non-tranfected cells and from 10 transiently transfected cells at mitosis exhibiting a low level of survivin-GFP fluorescence using the imaging system described below.

Fixation and immunofluorescence

LLC-PK and HeLa cells expressing survivin-GFP were fixed with 2.5% formaldehyde in PHEM (60 mM Pipes, 25 mM Hepes, pH 6.9, 10 mM EDTA, 4 mM MgCl2) for 15 minutes. After washes with MBST (10 mM Mops, 150 mM NaCl, pH 7.3, 0.05% Tween 20) DNA was stained with DAPI and the cells on coverslips were mounted in Vectashield (Vector Laboratories). For immunofluorescence, cells on coverslips were extracted before fixation or simultaneously fixed and extracted for 15 minutes in 0.25% CHAPS in PHEM containing 2.5% formaldehyde. Cells were labeled with antibodies against Crest centromere marker, survivin, hCdc20, Aurora B, or tubulin as previously described (Fortugno et al., 2002; Kallio et al., 2002a). DNA was stained with DAPI. Imaging was performed using a Zeiss Axioplan II microscope equipped with 63× (N.A. 1.4) and 100× objectives (N.A. 1.4), a Hamamatsu Orca 2 camera (Hamamatsu Photonics), and Metamorph imaging software (Universal Imaging).

Western blotting and immunoprecipitation

HeLa cells expressing full-length survivin-GFP or control GFP alone were treated with nocodazole for 12 hours before harvest of mitotic cells. Cell extracts and supernatants were prepared and used for immunoprecipitation as described previously (Kallio et al., 1998). In brief, for immunoprecipitation 5 μg of polyclonal anti-GFP antibody (Abcam) was prebound to 25 μl of protein A beads for 2 hours at +4°C. Immunoprecipitation was performed from 8×105 cells overnight. For western blotting of HeLa and LLC-PK extracts, monoclonal anti-survivin antibody 60.II was used at 2 μg/ml. For western blotting of anti-GFP immunoprecipitations, monoclonal anti-human Aurora B antibody (anti-AIM1, Transduction Labs) was used at 1 μg/ml.

Analysis of living cells, laser photobleaching and FRAP

For live cell experiments, cells were incubated in phenol red-free DMEM supplemented with 10 mM Hepes, 10% fetal bovine serum, and antibiotics (penicillin and streptomycin). The cells were analysed using a Zeiss Axiovert 200M microscope equipped with 63× (N.A. 1.4) and 100× objectives (N.A. 1.4) and Hamamatsu Orca ER CCD camera (Hamamatsu Photonics). Images were captured using Metamorph software (Universal Imaging). The average signal intensities of chromosome arm (n=5) and centromere (n=8) bound survivin-GFP at different mitotic phases were analysed from live cell sequences by measuring the integrated fluorescence intensities minus the background per sample and time point.

Photobleaching experiments were performed using a Micropoint Laser System (Phototonic Instruments) affixed to the epi-illumination port of the microscope. Laser light was attenuated with a neutral density filter to a level that did not affect cell viability or cell-cycle progression. The laser was aligned to the target region using phase contrast optics and a set of pre-bleach images was captured. A short exposure of laser irradiation (5-10 pulses each 4 nanoseconds) was used to reduce the fluorescence intensity of the target. Time-lapse sequences (time intervals ranging from 1 frame/second to 1 frame every 60 seconds) were captured using 30 to 500 millisecond exposure times.

Statistical analysis of survivin-GFP dynamics at different subcellular compartments

Measurement procedures were adapted from those reported previously (Maddox et al., 2000; Howell et al., 2000; Kallio et al., 2002a). A small region (Rtar) slightly larger than the target area was marked and the integrated fluorescence intensity within the region (Ftar) was logged into an Excel spreadsheet from each time point. The background fluorescence was subtracted by placing a larger region (Rback) over the smaller target area. Integrated fluorescence within the Rback was measured (Fback). The following equation was used to obtain integrated fluorescence intensity of a kinetochore minus the background (Fint):
\[\ \mathrm{F}_{\mathrm{int}}=\frac{\mathrm{F}_{\mathrm{tar}}-[(\mathrm{F}_{\mathrm{back}}-\mathrm{F}_{\mathrm{tar}}){\times}\mathrm{R}_{\mathrm{tar}}]}{(\mathrm{R}_{\mathrm{back}}-\mathrm{R}_{\mathrm{tar}})}.\ \]

The half-life of survivin-GFP fluorescence recovery was calculated using the equation: t1/2=ln2/k. The constant, k, was derived from fitting the data to a curve generated with the non-linear perturbation relaxation equation within GraphPad Prism software. The percentage of fluorescence recovery was calculated using the equation: 100% [Fmax-F(0)]/[Fpre-F(0)]; where Fpre is the intensity of the region before laser irradiation, Fmax is the intensity of the same region after maximal recovery and F(0) is the intensity at the first time point after photobleaching. Measurements were adjusted to control for any overall decrease in cellular fluorescence caused by photobleaching or imaging.

Results

Survivin accumulates near centromeres at early G2 and forms a bi-partite flexible structure between sister kinetochores by prophase. The subcellular localization of survivin has remained controversial, with debate as to whether it is a chromosomal passenger protein (Skoufias et al., 2000; Uren et al., 2000; Wheatley et al., 2001), a microtubule-associated protein (Li et al., 1998) or whether immunochemically distinct pools of survivin exist (Fortugno et al., 2002). To resolve these conflicts, we transiently transfected HeLa and LLC-PK cells with a plasmid encoding full-length survivin-GFP chimera and examined distribution of both exogenous and endogenous survivin at different cell-cycle phases. The ectopically expressed survivin-GFP was present throughout the cell cycle but its subcellular distribution appeared to be cell cycle regulated. We investigated cell-cycle-dependent changes in survivin localization. In G1 cells, identified by the presence of a remnant midbody from the previous cytokinesis, most of survivin-GFP was diffusely distributed in the cytosol and to a lesser extent in the nucleus (Fig. 1A). We synchronized HeLa cells expressing survivin-GFP at S phase by treatment with aphidicolin and tracked changes in survivin localization with progression through the cell cycle after release from the block. In early S phase (Fig. 1A, time point 0 hours) the majority of survivin-GFP was found within the nucleus where it remained diffuse but was excluded from the nucleoli. Three hours after the release from S phase block, faint traces of survivin-GFP-positive accumulations were observed in the nucleus, some of which were adjacent to centromeres detected with human autoimmune sera (Fig. 1A). Two hours later, all of the survivin-GFP foci were in close proximity to centromeres indicating the initial accumulation of the protein near the pre-kinetochores of G2 cells (Fig. 1A). This result was confirmed by immunofluorescence in HeLa (data not shown) and LLC-PK cells utilizing monoclonal anti-survivin antibody (Fig. 2). In prophase cells, survivin-GFP was found diffusely distributed in the cytosol and nucleus excluding the nucleoli, and accumulated at centromeres with some label at chromosome arms, especially during late prophase (Fig. 1A; time point +10 hours). After nuclear envelope breakdown (NEB), survivin accumulated further at inner centromeres and chromosome arms, and to some extent along the spindle microtubules (Fig. 1A; time point +11 hours, Fig. 2).

Fig. 1.

Survivin concentrates near centromeres in G2 and forms an extendable bipartite structure between the sister kinetochores in mitosis. (A) Accumulation of survivin-GFP adjacent to centromeres of early G2 cells. See text for details. The arrow and arrowheads denote the midbody and the spindle poles, respectively. The merge shows overlay of survivin-GFP (green) and autoimmune anti-centromere marker (red). The insets show higher magnification of selected areas. (B) Aurora B does not co-localize with survivin-GFP foci (thin arrows) until M phase (short arrows). (C) Survivin (red) undergoes extensive stretching between the sister kinetochores (anti-hCdc20, green) in a manner dependent on microtubule attachment and chromosome bipolarity. The arrows indicate a gap separating the two survivin accumulations on chromosomes that are under robust tension. (a) prophase, (b) unattached prometaphase, (c) attached prometaphase, (d) metaphase, (e) metaphaseanaphase transition. The numbers show the average length of the survivin segments under each condition (n=10-20 chromosomes per each category). The three linescan graphs show the average gray values for the above prometaphase (f), metaphase (g), and metaphase-anaphase transition (h) chromosomes (red survivin; green kinetochores). (D) Aurora B kinase associates with survivin-GFP in mitotic HeLa cells. Immunoprecipitations were performed with rabbit anti-GFP antibody utilizing HeLa cell populations expressing either survivin-GFP or control GFP alone. The arrow denotes a ∼42 kDa band detected with anti-hAurora B antibody and the asterisk the antibody heavy chain at 55 kDa. IP, immunoprecipitate; S, supernatant. Bars = 10 μm (panels A,B) and 0.5 μm (panel C).

Fig. 1.

Survivin concentrates near centromeres in G2 and forms an extendable bipartite structure between the sister kinetochores in mitosis. (A) Accumulation of survivin-GFP adjacent to centromeres of early G2 cells. See text for details. The arrow and arrowheads denote the midbody and the spindle poles, respectively. The merge shows overlay of survivin-GFP (green) and autoimmune anti-centromere marker (red). The insets show higher magnification of selected areas. (B) Aurora B does not co-localize with survivin-GFP foci (thin arrows) until M phase (short arrows). (C) Survivin (red) undergoes extensive stretching between the sister kinetochores (anti-hCdc20, green) in a manner dependent on microtubule attachment and chromosome bipolarity. The arrows indicate a gap separating the two survivin accumulations on chromosomes that are under robust tension. (a) prophase, (b) unattached prometaphase, (c) attached prometaphase, (d) metaphase, (e) metaphaseanaphase transition. The numbers show the average length of the survivin segments under each condition (n=10-20 chromosomes per each category). The three linescan graphs show the average gray values for the above prometaphase (f), metaphase (g), and metaphase-anaphase transition (h) chromosomes (red survivin; green kinetochores). (D) Aurora B kinase associates with survivin-GFP in mitotic HeLa cells. Immunoprecipitations were performed with rabbit anti-GFP antibody utilizing HeLa cell populations expressing either survivin-GFP or control GFP alone. The arrow denotes a ∼42 kDa band detected with anti-hAurora B antibody and the asterisk the antibody heavy chain at 55 kDa. IP, immunoprecipitate; S, supernatant. Bars = 10 μm (panels A,B) and 0.5 μm (panel C).

Fig. 2.

Indirect immunolocalization of endogenous survivin in LLCPK cells. (A) Survivin accumulates near the centromeres at G2 phase (arrows). (B) In late prophase cells, survivin is between the sister kinetochores (arrows). (C) During prometaphase, survivin changes its morphology as microtubules attach to the kinetochores and chromosomes become bipolarly oriented (the short arrow shows a `relaxed' chromosome and the long arrow denotes a chromosome under tension from both spindle poles). (D) Metaphase. (E) Early anaphase. (F) Anaphase B. The arrowheads denote survivin strands at the central spindle. (G) In telophase cells, survivin concentrates at the cleavage furrow (arrowheads) before final accumulation at the midbody (arrow). Merge images of survivin (red, detected with the monoclonal anti-survivin antibody), kinetochores (green, detected with anti-hCdc20 antibodies or Crest anti-centromere serum), and DNA (blue, stained with DAPI) are shown. (H) The specificity of monoclonal anti-survivin antibody. A single band corresponding to a 16 kDa protein was detected in interphase (I) and mitotic (M, 12 h taxol treatment) HeLa and LLC-PK cells. Bars, 5 μm.

Fig. 2.

Indirect immunolocalization of endogenous survivin in LLCPK cells. (A) Survivin accumulates near the centromeres at G2 phase (arrows). (B) In late prophase cells, survivin is between the sister kinetochores (arrows). (C) During prometaphase, survivin changes its morphology as microtubules attach to the kinetochores and chromosomes become bipolarly oriented (the short arrow shows a `relaxed' chromosome and the long arrow denotes a chromosome under tension from both spindle poles). (D) Metaphase. (E) Early anaphase. (F) Anaphase B. The arrowheads denote survivin strands at the central spindle. (G) In telophase cells, survivin concentrates at the cleavage furrow (arrowheads) before final accumulation at the midbody (arrow). Merge images of survivin (red, detected with the monoclonal anti-survivin antibody), kinetochores (green, detected with anti-hCdc20 antibodies or Crest anti-centromere serum), and DNA (blue, stained with DAPI) are shown. (H) The specificity of monoclonal anti-survivin antibody. A single band corresponding to a 16 kDa protein was detected in interphase (I) and mitotic (M, 12 h taxol treatment) HeLa and LLC-PK cells. Bars, 5 μm.

The other two chromosome passenger proteins, INCENP and Aurora B, have not been reported to concentrate at centromeres until prophase in Xenopus and porcine tissue culture cells (Kallio et al., 2002b; MacKay et al., 1998) although both proteins are found diffusely distributed in late G2 nuclei. To examine the distribution of survivin and Aurora B in human cells we synchronized HeLa cells expressing survivin-GFP at S phase. The cells were fixed and processed for immunofluorescence at 0, 5, 7, 9 and 10 hours after release from the S phase block. No Aurora B was found in interphase cells at time points between 5-10 hours (Fig. 1B). Only in M phase cells did Aurora B co-localize with survivin-GFP at the inner centromere (Fig. 1B) consistent with the idea that survivin might recruit Aurora B to mitotic chromosomes.

In fixed mitotic HeLa and LLC-PK cells immunolabeled with anti-survivin antibodies or expressing survivin-GFP, we noticed that survivin accumulated in an oval-shaped region between the sister kinetochores. This region changed morphology as microtubules attached to the kinetochores and chromosomes became bipolarly oriented. Owing to the larger size of porcine chromosomes, we used LLC-PK cells for more detailed analyses. We measured the average length of survivin labeling between sister kinetochores at different stages of cell division and in chromosomes that were either under tension from bipolar attachment or were in a relaxed state (Fig. 1C). In prophase cells, the length of survivin labeling between sister kinetochores was relatively short (0.45±0.1 μm, n=20) and showed little variation among different chromosomes. In prometaphase cells, the length of survivin labeling varied considerably depending on chromosome attachment and orientation. In unattached or mono-oriented chromosomes the average length of survivin signal was 0.3 μm (Fig. 1C panel b, n=10). In bi-oriented chromosomes the distance increased to 0.79±0.13 μm (n=20, Fig. 1C panel c). At metaphase, survivin showed an average length of 1.01±0.16 μm (n=20, Fig. 1C panel d). In this highly extended state the survivin at each centromere formed two accumulations separated by a ∼0.4 μm gap. At the metaphaseanaphase transition, this gap further extended to a width of ∼0.7 μm and survivin showed its most lengthened arrangement (1.77±0.31 μm, n=10, Fig. 1C panel e).

To determine whether survivin-GFP is incorporated into the chromosome passenger protein complex, we immunoprecipitated GFP-associated proteins from mitotic HeLa cells expressing either full-length survivin-GFP or control GFP alone. We found that survivin-GFP but not GFP co-precipitated with a portion of the Aurora B kinase in the cell extracts (Fig. 1D). This is in accordance with previously published data (Wheatley et al., 2001).

From the localization data we conclude that among those proteins that are not constitutive components of the centromeres, survivin is one of the earliest G2/M phase proteins to concentrate at the maturing centromere-kinetochore complexes of interphase cells at early G2. Moreover, during mitosis survivin forms two separate accumulations between the sister kinetochores, revealing that the inner centromere, like the outer centromere and kinetochore, is arranged in a bi-partite fashion. Survivin accumulations undergo notable stretching after microtubule attachment and interkinetochore tension. We speculate that changes in centromere structure upon microtubule attachment and interkinetochore tension are important in regulating structural arrangement of survivin and its association with other proteins at inner centromeres. These effects on survivin might, in turn, regulate the activities of centromere-associated Aurora B kinase.

Dynamic redistribution of survivin during mitosis

To investigate the effects of survivin-GFP expression on cell division we transiently transfected populations of HeLa cells and monitored their progression through mitosis using phase contrast and fluorescence time-lapse microscopy (data not shown). We selected transfected cells exhibiting a low level of GFP fluorescence for analysis. Transfected cells in mitosis showed an average increase in survivin expression of 19.9% (n=10) compared with endogenous survivin levels in nontransfected cells (n=10). A slight cell-cycle delay was observed at metaphase where survivin-GFP transfected cells (n=30) remained for an average of 43±39 minutes while GFP transfected cells remained for 22±15 minutes (n=15).

To characterize the cell-cycle-regulated changes in accumulation of survivin, we quantified the changes in accumulation of survivin-GFP at different stages of mitosis. Survivin first concentrates at chromosome arms and at centromeres during late prophase. We measured the intensities of the survivin-GFP pools at chromosome arms (n=5) and at inner centromeres (n=8) in living LLC-PK cells as they progressed from late prophase to early anaphase (Fig. 3A,B and Movie 1).

Fig. 3.

Redistribution of survivin-GFP during cell division. (A) Time-lapse sequence of an LLC-PK cell expressing survivin-GFP during prophase, nuclear envelope breakdown (NEB) and prometaphase. The amount of survivin-GFP at the inner centromeres increases substantially around NEB (time point 0 minutes) and shows further accumulation during prometaphase. The white circle denotes a centromere that remained in focus throughout the observation period. In contrast to inner centromere labeling, the intensity of chromosome arm labeling diminishes during prometaphase after peaking at NEB. A video corresponding to the still images is available as supplementary information (see Movie 1). (B) Changes in the average integrated fluorescence of chromosome arm (▪, n=5) and inner centromere localized survivin-GFP (▵, n=8) from late prophase to onset of anaphase. (C) Co-localization of survivin-GFP with the spindle microtubules in anaphase cells is lost by early telophase when survivin-GFP forms separate filamentous accumulations (arrows) that run between the microtubules of the central spindle. The area inside the white boxes is shown in higher magnification. mts, microtubules; surv, survivin-GFP. The merge shows an overlay of DNA (blue), survivin-GFP (green), and spindle microtubules (red). Bars = 10 μm (panel A) and 3 μm (panel C).

Fig. 3.

Redistribution of survivin-GFP during cell division. (A) Time-lapse sequence of an LLC-PK cell expressing survivin-GFP during prophase, nuclear envelope breakdown (NEB) and prometaphase. The amount of survivin-GFP at the inner centromeres increases substantially around NEB (time point 0 minutes) and shows further accumulation during prometaphase. The white circle denotes a centromere that remained in focus throughout the observation period. In contrast to inner centromere labeling, the intensity of chromosome arm labeling diminishes during prometaphase after peaking at NEB. A video corresponding to the still images is available as supplementary information (see Movie 1). (B) Changes in the average integrated fluorescence of chromosome arm (▪, n=5) and inner centromere localized survivin-GFP (▵, n=8) from late prophase to onset of anaphase. (C) Co-localization of survivin-GFP with the spindle microtubules in anaphase cells is lost by early telophase when survivin-GFP forms separate filamentous accumulations (arrows) that run between the microtubules of the central spindle. The area inside the white boxes is shown in higher magnification. mts, microtubules; surv, survivin-GFP. The merge shows an overlay of DNA (blue), survivin-GFP (green), and spindle microtubules (red). Bars = 10 μm (panel A) and 3 μm (panel C).

The average fluorescence intensity of survivin-GFP at both the chromosome arm and the inner centromere increased noticeably 2-3 minutes before NEB. Survivin at chromosome arms achieved its maximum concentration ∼2 minutes after NEB and then faint traces diminished gradually to near background levels during prometaphase (Fig. 3B, P<0.01). In contrast, the average fluorescence intensity of survivin-GFP at inner centromeres increased markedly through prometaphase (P<0.01) and reached its maximum level at metaphase (Fig. 3A,B).

Another change in survivin distribution occurs at early anaphase. Previous immunofluorescent studies suggest that the protein moves from chromosomes to microtubules of the central spindle at early anaphase and then concentrates to the midbody at late telophase (Skoufias et al., 2000; Uren et al., 2000; Wheatley et al., 2001). In contrast, recent live cell experiments proposed a microtubule-independent translocation of a survivin-dsRed chimera from chromosomes to the cytosol at anaphase and then to the ends of polar microtubules by early telophase (Temme et al., 2003). To investigate this, we analyzed the distribution of endogenous survivin and survivin-GFP in fixed HeLa and LLC-PK cells immunolabeled with anti-tubulin antibodies. In these cells, survivin co-localized with microtubules of the central spindle from early stages of anaphase A to the start of anaphase B (Fig. 2E,F, Fig. 3C). However, later in anaphase the accumulations of survivin in the central spindle were displaced from microtubules and relocated to narrow 3-6 μm long strands between the microtubule bundles before finally co-localizing with the dense midbody in late telophase (Fig. 2G, Fig. 3C). We conclude that survivin populations concentrated at chromosome arms and inner centromeres experience dynamic changes in affinity during early stages of mitosis resulting in reduction in chromosome arm labeling and accumulation of the protein to inner centromeres.

Turnover of inner centromere-bound survivin is cell-cycle regulated

Changes in survivin concentration at mitotic organelles may be related to changes in dynamic protein-protein interactions such as the interaction between the passenger proteins. To study survivin dynamics more directly we performed FRAP experiments at different phases of cell cycle on HeLa and LLCPK cells expressing survivin-GFP. Our pull-down assays (Fig. 1D) and in vitro binding experiments performed by others (Wheatley et al., 2001) suggest that tagging survivin with GFP might affect the ability of the protein to interact fully with Aurora B and INCENP. It is possible that less survivin-GFP is in a complex with Aurora B and INCENP compared with endogenous survivin in vivo. Thus our results regarding the mobility of survivin-GFP might reflect to some extent the behavior of unbound survivin. As a counterpoint, the observed subcellular localization and dynamic translocation of survivin-GFP are indistinguishable from those of the endogenous protein. For this reason we believe that the behavior of survivin-GFP reflects that of the endogenous protein in vivo.

In interphase cells at G2, we targeted the initial survivin-GFP accumulation near the centromeres. The turnover was slow at this point in both HeLa and LLC-PK cells showing average half time for recovery of 1860±600 seconds and 2940±480 seconds, respectively (Fig. 4A, Table 1 and Movie 2). The total recovery of fluorescence was high in both HeLa (84±11%) and LLC-PK (89±12%) cells. After entry into M phase, the turnover rate of survivin at inner centromeres increased sharply (P<0.01) in both cell types investigated. We observed no significant differences in the average recovery half times between prophase, prometaphase and metaphase centromeres. Both cell lines showed the same range of 3-12 seconds turnover at late prophase and metaphase (Fig. 4B,C, Table 1, and Movies 3 and 4). The average turnover rates and the total recovery of survivin-GFP fluorescence at chromosome arms were similar to those of inner centromeres (Table 1). After the onset of anaphase, survivin turnover in the central spindle (Fig. 4D) and midbody became slow and limited in extent (Fig. 4E, Table 1, and Movie 5). In conclusion, survivin turns over slowly near centromeres of interphase cells, becomes highly dynamic on chromosome arms and at centromeres during early phases of mitosis, and becomes more stably associated with the central spindle after the onset of anaphase.

Fig. 4.

The turnover rate of survivin peaks at early M phase and is partially dependent on microtubule-kinetochore attachment. (A-G) Target areas (indicated with white circles) in HeLa cells (panel A) or LLC-PK cells (panels B-G) were photo-bleached and followed by fluorescence time-lapse microscopy. Pre-bleach and post-bleach images representing partial and maximal recovery are shown. The insets show higher magnification views of the target area. At the end of each row are corresponding graphs of survivin-GFP recovery (arrows indicate the pre-bleach fluorescence level of the target area). Proportions of fluorescence recovery (recf) and half time of recovery (t1/2) are shown for each graph, as percentages. Supplemental Movies 2-6 corresponding to the still images of panels A, B, C, E and F, respectively, are available online (Supplementary Information).

Fig. 4.

The turnover rate of survivin peaks at early M phase and is partially dependent on microtubule-kinetochore attachment. (A-G) Target areas (indicated with white circles) in HeLa cells (panel A) or LLC-PK cells (panels B-G) were photo-bleached and followed by fluorescence time-lapse microscopy. Pre-bleach and post-bleach images representing partial and maximal recovery are shown. The insets show higher magnification views of the target area. At the end of each row are corresponding graphs of survivin-GFP recovery (arrows indicate the pre-bleach fluorescence level of the target area). Proportions of fluorescence recovery (recf) and half time of recovery (t1/2) are shown for each graph, as percentages. Supplemental Movies 2-6 corresponding to the still images of panels A, B, C, E and F, respectively, are available online (Supplementary Information).

Table 1.

Dynamics of survivin at different cell cycle phases and subcellular locations in HeLa and LLC-PK cells

Cell cycle phase/location Treatment n t1/2 (seconds)* % Rec*
HeLa      
G2 phase/centromere   -   5   1860±600  84±11  
Prometaphase/inner centromere   -   10   7.2±4.2   79±8  
Prometaphase/inner centromere   Nocodazole   10   31.3±9.8   71±10  
Telophase/midbody   -   5   144±42   9±5  
LLC-PK      
G2 phase/centromere   -   5   2940±480  89±12  
Prophase/inner centromere   -   7   7.2±4.4   92±9  
Prophase/chromosome arm   -   5   6.8±3.1   82±9  
Metaphase/inner centromere   -   10   5.3±3.5   93±8  
Prometaphase/inner centromere   -   5   4.6±1.1   79±8  
Prometaphase/inner centromere   Nocodazole   10   74.2±18.1§  77±11  
Prometaphase/inner centromere   ZM447439/MG132   8   67.1±27.2   18±13  
Prometaphase/inner centromere   MG132   5   4.4±2.1   83±8  
Metaphase/inner centromere   Taxol   10   10.3±8.0   78±9  
Anaphase/central spindle   -   3   87±23   8±4  
Telophase/midbody   -   5   204±108  11±3  
Cell cycle phase/location Treatment n t1/2 (seconds)* % Rec*
HeLa      
G2 phase/centromere   -   5   1860±600  84±11  
Prometaphase/inner centromere   -   10   7.2±4.2   79±8  
Prometaphase/inner centromere   Nocodazole   10   31.3±9.8   71±10  
Telophase/midbody   -   5   144±42   9±5  
LLC-PK      
G2 phase/centromere   -   5   2940±480  89±12  
Prophase/inner centromere   -   7   7.2±4.4   92±9  
Prophase/chromosome arm   -   5   6.8±3.1   82±9  
Metaphase/inner centromere   -   10   5.3±3.5   93±8  
Prometaphase/inner centromere   -   5   4.6±1.1   79±8  
Prometaphase/inner centromere   Nocodazole   10   74.2±18.1§  77±11  
Prometaphase/inner centromere   ZM447439/MG132   8   67.1±27.2   18±13  
Prometaphase/inner centromere   MG132   5   4.4±2.1   83±8  
Metaphase/inner centromere   Taxol   10   10.3±8.0   78±9  
Anaphase/central spindle   -   3   87±23   8±4  
Telophase/midbody   -   5   204±108  11±3  

*Mean ± s.d.; significantly (P<0.01) different from mitotic cells; §significantly (P<0.01) different from untreated and taxol-treated cells; significantly (P<0.01) different from interphase and early mitotic cells.

Survivin dynamics at inner centromeres is partially regulated by microtubule attachment

Because microtubule attachment and mechanical tension at kinetochores regulate the spindle checkpoint and progression through M phase, we sought to determine whether microtubule attachment affects survivin dynamics at centromeres. We treated survivin-GFP cells with nocodazole, which destabilizes microtubules eliminating microtubule attachment and tension or with taxol, a microtubule hyperstabilizing drug, which maintains many microtubule-kinetochore attachments but reduces inter-kinetochore tension. We then measured survivin turnover using FRAP. Cells treated with nocodazole showed significantly (P<0.01) slower survivin turnover compared with that of normal cells or of cells treated with taxol in both cell lines examined (Fig. 4F, Table 1). The average recovery half time of survivin fluorescence was seven times longer (74.2±18.1 seconds) in nocodazole treated LLC-PK cells (see Movie 6) compared with taxol-treated cells (10.3±8.0 seconds, Fig. 4G, Table 1). These results are consistent with the notion that survivin turnover at inner centromeres of prometaphase cells is increased by microtubule occupancy at kinetochores but is not markedly affected by tension. We recognize that overexpression of survivin or the use of a chimeric protein might influence survivin turnover. However, the consistent changes in dynamics seen with cell-cycle progression and microtubule attachment suggest that the observations reflect physiologically significant regulation of the survivin protein.

Aurora B kinase activity is required for maintenance of survivin dynamics at inner centromeres

Aurora B has been proposed to modulate microtubule-kinetochore interactions to promote proper bi-orientation of chromosomes. The absence of normal Aurora B function leads to defects in chromosome movement and alignment (Kallio et al., 2002b; Ditchfield et al., 2003; Hauf et al., 2003). Activity of Aurora B is also essential for spindle checkpoint signaling as perturbation of its functions leads to rapid exit from M phase even in presence of unaligned chromosomes (Kallio et al., 2002b). To investigate how Aurora B kinase activity affects survivin dynamics we applied ZM447439, a small molecule inhibitor of Aurora B (Ditchfield et al., 2003), to LLC-PK cells expressing survivin-GFP and performed time-lapse and FRAP analysis. ZM447439 has been reported to inhibit Aurora A and B with IC50 values of 110 and 130 nM, respectively, while the majority of other protein kinases are not affected by the drug (Ditchfield et al., 2003). We cannot formally exclude that effects of ZM447439 on survivin behavior are due, at least in part, to consequences of inhibition of other kinases, particularly other Aurora kinases. However, ZM447439-treated cells exhibit a very similar phenotype to Aurora B RNAi- (Hauf et al., 2003; Adams et al., 2001b) but not to Aurora A RNAi- (Hirota et al., 2003; Giet et al., 2002; Hannak et al., 2001) treated cells. In our assays, we first added MG132, a drug that blocks the onset of anaphase by inhibiting proteasome activity, to prevent cells from exiting M phase prematurely due to the inhibitory effects of ZM447439 on the spindle checkpoint. In the presence of MG132 alone survivin-GFP fluorescence and its average recovery half time at inner centromeres was unaltered compared with untreated metaphase cells (Table 1, Fig. 5). In contrast, in cells that were incubated in 2 μM ZM447439 in the continued presence MG132, survivin-GFP became stably associated with centromeres showing very slow and limited recovery. The total recovery was only 18±13%, and the average turnover time to achieve this low extent of recovery was 67±27 seconds (Fig. 5, Table 1 and Movie 7). Consistent with previous observations (Ditchfield et al., 2003; Hauf et al., 2003) we noticed that concentration of survivin was lost from inner centromeres within 30 minutes after addition of ZM447439 and MG132 (data not shown). These results are consistent with a model in which the recruitment of survivin to inner centromeres is dependent on Aurora B kinase activity.

Fig. 5.

Aurora B regulates survivin turnover at inner centromeres. (Top) FRAP analysis in an LLC-PK cell shows that survivin-GFP turnover at metaphase centromeres occurs at its normally rapid rate in a cell treated with the proteaseome inhibitor MG132. (Bottom) In a cell pretreated for 30 minutes with MG132 then treated for 10 minutes with ZM447439 in the continued presence of MG132, survivin-GFP at the inner centromeres shows slow and limited turnover. Pre-bleach and post-bleach images representing partial and maximal recovery are shown. The white circles denote the target area. The graphs show survivin-GFP recovery (arrows indicates the pre-bleach fluorescence level of the target area). Proportion of fluorescence recovery (recf) and half time of recovery (t1/2) are shown as percentages. Movie 7 corresponding to the still images of panel B is available online (Supplementary Information).

Fig. 5.

Aurora B regulates survivin turnover at inner centromeres. (Top) FRAP analysis in an LLC-PK cell shows that survivin-GFP turnover at metaphase centromeres occurs at its normally rapid rate in a cell treated with the proteaseome inhibitor MG132. (Bottom) In a cell pretreated for 30 minutes with MG132 then treated for 10 minutes with ZM447439 in the continued presence of MG132, survivin-GFP at the inner centromeres shows slow and limited turnover. Pre-bleach and post-bleach images representing partial and maximal recovery are shown. The white circles denote the target area. The graphs show survivin-GFP recovery (arrows indicates the pre-bleach fluorescence level of the target area). Proportion of fluorescence recovery (recf) and half time of recovery (t1/2) are shown as percentages. Movie 7 corresponding to the still images of panel B is available online (Supplementary Information).

Discussion

We extend previous observations of survivin behavior as a chromosomal passenger protein that exhibits dynamic changes in its localization, and turnover in a cell-cycle-dependent manner. The ∼20-fold increase in the turnover of survivin at inner centromeres of prophase and prometaphase cells compared with interphase and anaphase cells suggests an active role for survivin dynamics in regulation of early mitotic events. We hypothesize that changes in survivin dynamics at centromeres might regulate Aurora B kinase activity, which, in turn, governs kinetochore-microtubule interactions and signaling of the spindle checkpoint. Rapid modulation of Aurora B might be required for release of inappropriate microtubule connections.

There is controversy about whether the spindle checkpoint is initiated at kinetochores in response to the loss of microtubule occupancy, the loss of mechanical tension, or both (reviewed by Millband et al., 2002). We detect a slower turnover of survivin in nocodazole-treated cells lacking microtubule-kinetochore interactions compared with cells treated with taxol expected to have a normal or near-normal number of microtubules bound to kinetochores (Waters et al., 1996; McEwen et al., 1997) and reduced tension (Waters et al., 1998). We propose that the longer residence time of survivin at the centromeres of chromosomes lacking microtubule-kinetochore associations increases Aurora B kinase activity and results in stronger activation of the spindle checkpoint. This model is consistent with the reports that small molecule inhibitors of Aurora kinases more readily override the M phase arrest induced in mammalian cells by taxol compared with the arrest induced by nocodazole or other microtubule destabilizers (Ditchfield et al., 2003; Hauf et al., 2003). Thus we suggest that one mechanism for dynamic regulation of Aurora B kinase is modulation of the residence time of survivin at centromeres, influenced by microtubule occupancy at the kinetochores. We do not exclude the possibility that mechanical tension might provide an additional mechanism for regulating Aurora B activity.

Suppression of Aurora B activity by addition of a small molecule inhibitor ZM447439 to the cells caused hyperstabilation of survivin-GFP at inner centromeres. This was accompanied with a gradual loss of survivin from this location even in the presence of a proteasome inhibitor that prevents cells from entering anaphase. This finding suggests that Aurora B activity is required for maintenance of proper survivin dynamics at inner centromeres. Further work will be required to identify downstream targets of Aurora B kinase that affect microtubule dynamics and cell-cycle regulation in M phase.

Movies available online

Acknowledgements

We are grateful to Dario Altieri who provided anti-survivin antibodies and survivin-GFP plasmids. We thank Astra-Zeneca for the ZM447439 drug. This study was supported by grant R01-GM50412 from the National Institute of General Medical Sciences and from the Presbyterian Health Foundation.

References

Adams, R. R., Carmena, M. and Earnshaw, W. C. (
2001a
). Chromosomal passengers and the (aurora) ABCs of mitosis.
Trends Cell Biol.
11
,
49
-54.
Adams, R. R., Maiato, H., Earnshaw, W. C. and Carmena, M. (
2001b
). Essential roles of Drosophila inner centromere protein (INCENP) and Aurora B in histone H3 phosphorylation, metaphase chromosome alignment, kinetochore disjunction, and chromosome segregation.
J. Cell Biol.
153
,
865
-880.
Altieri, D. C. (
2003
). Validating survivin as a cancer therapeutic target.
Nat. Rev. Cancer
3
,
46
-54.
Biggins, S. and Murray, A. W. (
2001
). The budding yeast protein kinase Ipl1/Aurora allows the absence of tension to activate the spindle checkpoint.
Genes Dev.
15
,
3118
-3129.
Bolton, M. A., Lan, W., Powers, S. E., McCleland, M. L., Kuang, J. and Stukenberg, P. T. (
2002
). Aurora B kinase exists in a complex with survivin and INCENP and its kinase activity is stimulated by survivin binding and phosphorylation.
Mol. Biol. Cell
13
,
3064
-3077.
Chiou, S. K., Jones, M. K. and Tarnawski, A. S. (
2003
). Survivin - an anti-apoptosis protein: its biological roles and implications for cancer and beyond.
Med. Sci. Monit.
9
,
125
-129.
Ditchfield, C., Johnson, V. L., Tighe, A., Ellston, R., Haworth, C., Johnson, T., Mortlock, A., Keen, N. and Taylor, S. S. (
2003
). Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores.
J. Cell Biol.
161
,
267
-280.
Fortugno, P., Wall, N. R., Giodini, A., O'Connor, D. S., Plescia, J., Padgett, K. M., Tognin, S., Marchisio, P. C. and Altieri, D. C. (
2002
). Survivin exists in immunochemically distinct subcellular pools and is involved in spindle microtubule function.
J. Cell Sci.
115
,
575
-585.
Fraser, A. G., James, C., Evan, G. I. and Hentgartner, M. O. (
1999
). Caenorhabditis elegans inhibitor of apoptosis protein (IAP) homologue BIR-1 plays a conserved role in cytokinesis.
Curr. Biol.
9
,
292
-301.
Giet, R., McLean, D., Descamps, S., Lee, M. J., Raff, J. W., Prigent, C. and Glover, D. M. (
2002
). Drosophila Aurora A kinase is required to localize D-TACC to centrosomes and to regulate astral microtubules.
J. Cell Biol.
156
,
437
-451.
Giodini, A., Kallio, M. J., Wall, N. R., Gorbsky, G. J., Tognin, S., Marchisio, P. C., Symons, M. and Altieri, D. C. (
2002
). Regulation of microtubule stability and mitotic progression by survivin.
Cancer Res.
62
,
2462
-2467.
Hannak, E., Kirkham, M., Hyman, A. A. and Oegema, K. (
2001
). Aurora- A kinase is required for centrosome maturation in Caenorhabditis elegans.
J. Cell Biol.
155
,
1109
-1116.
Hauf, S., Cole, R. W., LaTerra, S., Zimmer, C., Schnapp, G., Walter, R., Heckel, A., van Meel, J., Rieder, C. L. and Peters, J. M. (
2003
). The small molecule Hesperadin reveals a role for Aurora B in correcting kinetochore-microtubule attachment and in maintaining the spindle assembly checkpoint.
J. Cell Biol.
161
,
281
-294.
Hirota, T., Kunitoku, N., Sasayama, T., Marumoto, T., Zhang, D., Nitta, M., Hatakeyama, K. and Saya, H. (
2003
). Aurora-A and an interacting activator, the LIM protein Ajuba, are required for mitotic commitment in human cells.
Cell
114
,
585
-598.
Honda, R., Korner, R. and Nigg, E. A. (
2003
). Exploring the functional interactions between Aurora B, INCENP, and survivin in mitosis.
Mol. Biol. Cell
14
,
3325
-3341.
Howell, B. J., Hoffman, D. B., Fang, G., Murray, A. W. and Salmon, E. D. (
2000
). Visualization of Mad2 dynamics at kinetochores, along spindle fibers, and at spindle poles in living cells.
J. Cell Biol.
150
,
1233
-1250.
Kallio, M., Weinstein, J., Daum, J. R., Burke, D. J. and Gorbsky, G. J. (
1998
). Mammalian p55CDC mediates association of the spindle checkpoint protein Mad2 with the cyclosome/anaphase-promoting complex, and is involved in regulating anaphase onset and late mitotic events.
J. Cell Biol.
141
,
1393
-1406.
Kallio, M. J., Beardmore, V. A., Weinstein, J. and Gorbsky, G. J. (
2002a
). Rapid microtubule-independent dynamics of Cdc20 at kinetochores and centrosomes in mammalian cells.
J. Cell Biol.
158
,
841
-847.
Kallio, M. J., McCleland, M. L., Stukenberg, P. T. and Gorbsky, G. J. (
2002b
). Inhibition of Aurora B kinase blocks chromosome segregation, overrides the spindle checkpoint, and perturbs microtubule dynamics in mitosis.
Curr. Biol.
12
,
900
-905.
Li, F., Ambrosini, G., Chu, E. Y., Plescia, J., Tognin, S., Marchisio, P. C. and Altieri, D. C. (
1998
). Control of apoptosis and mitotic spindle checkpoint by survivin.
Nature
396
,
580
-584.
Mackay, A. M., Ainsztein, A. M., Eckley, D. M. and Earnshaw, W. C. (
1998
). A dominant mutant of inner centromere protein (INCENP), a chromosomal protein, disrupts prometaphase congression and cytokinesis.
J. Cell Biol.
140
,
991
-1002.
Maddox, P. S., Bloom, K. S. and Salmon, E. D. (
2000
). The polarity and dynamics of microtubule assembly in the budding yeast Saccharomyces cerevisiae.
Nat. Cell Biol.
2
,
36
-41.
McEwen, B. F., Heagle, A. H., Cassels, G. O., Buttle, K. F. and Rieder, C. L. (
1997
). Kinetochore fiber maturation in PtK1 cells and its implication for the mechanisms of chromosome congression and anaphase onset.
J. Cell Biol.
137
,
1567
-1580.
Millband, D. N., Campbell, L. and Hardwick, K. G. (
2002
). The awesome power of multiple model systems: interpreting the complex nature of spindle checkpoint signaling.
Trends Cell Biol.
12
,
205
-209.
Musacchio, A. and Hardwick, K. G. (
2002
). The spindle checkpoint: structural insights into dynamic signalling.
Nat. Rev. Mol. Cell Biol.
3
,
731
-741.
Reed, J. C. and Bischoff, J. R. (
2000
). BIRinging chromosomes through cell division- and survivin' the experience.
Cell
102
,
545
-548.
Skoufias, D. A., Mollinari, C., Lacroix, F. B. and Margolis, R. L. (
2000
). Human survivin is a kinetochore-associated passenger protein.
J. Cell Biol.
151
,
1575
-1582.
Speliotes, E. K., Uren, A., Vaux, D. and Horvitz, H. R. (
2000
). The survivin-like C. elegans BIR-1 protein acts with the Aurora-like kinase AIR-2 to affect chromosomes and the spindle midzone.
Mol. Cell
6
,
211
-223.
Tanaka, T. U., Rachidi, N., Janke, C., Pereira, G., Galova, M., Schiebel, E., Stark, M. J. and Nasmyth, K. (
2002
). Evidence that the Ipl1-Sli15 (Aurora kinase-INCENP) complex promotes chromosome bi-orientation by altering kinetochore-spindle pole connections.
Cell
108
,
317
-329.
Temme, A., Rieger, M., Reber, F., Lindemann, D., Weigle, B., Diestelkoetter-Bachert, P., Ehninger, G., Tatsuka, M., Terada, Y. and Rieber, E. P. (
2003
). Localization, dynamics and function of survivin revealed by expression of functional survivinDsRed fusion proteins in the living cell.
Mol. Biol. Cell
14
,
78
-92.
Uren, A. G., Coulson, E. J. and Vaux, D. L. (
1998
). Conservation of baculovirus inhibitor of apoptosis repeat proteins (BIRPs) in viruses, nematodes, vertebrates and yeasts.
Trends Biochem Sci.
23
,
159
-162.
Uren, A. G., Wong, L., Pakusch, M., Fowler, K. J., Burrows, F. J., Vaux, D. L. and Choo, K. H. A. (
2000
). Survivin and the inner centromere protein INCENP show similar cell-cycle localization and gene knockout phenotype.
Curr. Biol.
10
,
1319
-1328.
Waters, J. C., Mitchison, T. J., Rieder, C. L. and Salmon, E. D. (
1996
). Kinetochore microtubule poleward flux produces a force that can do work.
Mol. Biol. Cell
7
,
1547
-1558.
Waters, J. C., Chen, R. H., Murray, A. W. and Salmon, E. D. (
1998
). Localization of Mad2 to kinetochores depends on microtubule attachment, not tension.
J. Cell Biol.
141
,
1181
-1191.
Wheatley, S. P., Carvalho, A., Vagnarello, P. and Earnshaw, W. C. (
2001
). INCENP is required for proper targeting of survivin to the centromeres and the anaphase spindle during mitosis.
Curr. Biol.
11
,
886
-890.

Supplementary information