Spontaneous slow replication fork progression elicits mitosis alterations in homologous recombination-deficient mammalian cells

Significance

Faithful genome duplication requires the precise coordination of DNA replication, repair/recombination and chromosome segregation. Homologous recombination (HR) plays a pivotal role in the resumption of blocked replication forks, and HR⁻ cells exhibit both spontaneous slower genome-wide replication fork speed and mitotic extra centrosomes (MECs). We show that MECs result from slow replication kinetics and that MECs are associated with multipolar mitosis leading to general unbalanced chromosome segregation. Thus, low levels of replication stress, which are not detected by cell surveillance, allow cells to progress through the cell cycle, resulting in aberrant mitosis and chromosome instability. These data underline the essential role of HR facing endogenous stress at the interface between replication and mitosis for protection against spontaneous general chromosome instability.

Abstract

Homologous recombination deficient (HR⁻) mammalian cells spontaneously display reduced replication fork (RF) movement and mitotic extra centrosomes. We show here that these cells present a complex mitotic phenotype, including prolonged metaphase arrest, anaphase bridges, and multipolar segregations. We then asked whether the replication and the mitotic phenotypes are interdependent. First, we determined low doses of hydroxyurea that did not affect the cell cycle distribution or activate CHK1 phosphorylation but did slow the replication fork movement of wild-type cells to the same level than in HR⁻ cells. Remarkably, these low hydroxyurea doses generated the same mitotic defects (and to the same extent) in wild-type cells as observed in unchallenged HR⁻ cells. Reciprocally, supplying nucleotide precursors to HR⁻ cells suppressed both their replication deceleration and mitotic extra centrosome phenotypes. Therefore, subtle replication stress that escapes to surveillance pathways and, thus, fails to prevent cells from entering mitosis alters metaphase progression and centrosome number, resulting in multipolar mitosis. Importantly, multipolar mitosis results in global unbalanced chromosome segregation involving the whole genome, even fully replicated chromosomes. These data highlight the cross-talk between chromosome replication and segregation, and the importance of HR at the interface of these two processes for protection against general genome instability.

DNA is continuously subjected to injury by exogenous and endogenous sources. The faithful transmission of genetic material relies on the DNA damage response (DDR), which coordinates a network of pathways, including DNA replication-repair-recombination, the cell cycle checkpoint, and chromosome segregation. A defect in any of these pathways causes genetic instability and cancer predisposition. Strikingly, both spontaneous DDR activation as a consequence of endogenous replication stress and centrosome abnormalities, which cause uneven chromosome segregation, have been reported in precancerous and early-stage malignancies (1–10). Therefore, endogenous stresses must play a key role in spontaneous chromosome instability and in cancer etiology.

Homologous recombination (HR) is an evolutionarily conserved process that controls the balance between genetic stability and diversity. Specifically, HR plays a pivotal role in the reactivation of replication forks that have been blocked, contributing to DNA replication accuracy (11–16). Replication forks are routinely inactivated by endogenous stress (17, 18); therefore, HR should play an essential role to protect cells against these types of stresses, and HR deficiency should reveal endogenous replication stress. Remarkably, unchallenged HR-deficient (HR⁻) cells display both a genome-wide decrease in replication fork speed (19) and a spontaneous increase in the frequency of cells containing extra centrosomes (20–28). Two hypotheses may account for these two phenotypes. First, replication stress leads to chromosome alteration at incomplete replicated regions and chromosome rearrangements (29). However, centrosomes do not contain DNA, and if extra centrosomes at mitosis [mitotic extra centrosome (MEC)] are active, unbalanced chromosome segregation should lead to global chromosome instability, even for fully replicated chromosomes. Second, HR proteins are associated with supernumerary centrosomes; therefore, centrosome duplication defects may directly result from HR misregulation (30, 31).

In this study, we addressed whether spontaneous MECs result from slow replication fork movement in HR⁻ cells. The data presented here underline the importance of HR at the molecular interface between replication and chromosome segregation to protect against spontaneous genomic instability.

Results

The Impact of Low Hydroxyurea Doses on the Cell Cycle and RF Speed.

We used cell lines derived from hamster V79 cells because several extensively characterized HR⁻ cell lines exist in this background. Notably, we used cells that were affected by a loss-of-function mutation in the endogenous BRCA2 gene (V-C8 cells) (26) and cells that express a dominant negative form of RAD51 (V79SM24), which specifically inhibits gene conversion without affecting cell viability (20, 32–34).

In an attempt to reproduce the level of replication fork (RF) slowing observed in untreated V-C8 and V79SM24 (HR⁻) cells (19), we analyzed genome-wide RF speed following treatment with increasing doses of hydroxyurea (HU), an inhibitor of the ribonucleotide reductase that depletes deoxyribonucleotide (dNTP) pools (35). Doses between 1 and 5 mM are commonly used to arrest cells in S phase, but we expected that lower doses would decelerate rather than block RF progression.

Flow cytometry analysis demonstrated that increasing the HU dose progressively leads to the accumulation of cells in S phase, which was accompanied by fewer cells in the G₁ phase (Fig. S1). Overall, 1 mM HU provoked a strong accumulation of cells in S phase, which was less pronounced with 250 or 500 μM HU. Remarkably, 100 μM HU moderately affected the cell cycle distribution, and 10 μM HU did not significantly affect the cell cycle distribution (Fig. S1), indicating that the cells were actively replicating their DNA and efficiently progressing through S phase. Consistently, the level of CHK1 phosphorylation (pCHK1) progressively increased in a HU dose-dependent manner (Fig. S2A). Ten micromolar HU did not impact the pCHK1 status in WT and V-C8 and V79SM24 (HR⁻) cells. CHK1 phosphorylation was detectable in all cell lines except V-C8#13 upon treatment with 50 μM HU, and high HU doses strongly activated CHK1 (Fig. S2A). Taken together, these data suggest that HU treatment progressively affects replication in a dose-dependent manner. Interestingly, untreated V-C8 and V79SM24 (HR⁻) cells exhibited higher levels of phosphorylated CHK1 than their respective controls (Fig. S2 A and B), revealing endogenous replication stress in this genetic context.

RF speed was monitored by using single molecule analysis (molecular combing). We found that the replication tracts are much shorter in cells grown in the presence of 100 μM HU than in untreated cells, demonstrating that genome-wide RF progression was strongly affected by this treatment (Fig. 1). However, interorigin distance was strongly reduced (Fig. S3), showing that more replication origins were active, partially compensating for the forks slowing, consistently with the moderate effect on cell cycle distribution. These data demonstrate that at least 10-fold lower HU doses than the levels classically used (1–5 mM) strongly impair fork progression without causing S phase arrest.

Fig. 1.

The impact of HR deficiency or hydroxyurea on replication fork speed. (A) Examples of combed DNA fibers with replication tracts: IdU (green), CldU (red), and ssDNA (blue) in nontreated (NT) conditions or after HU exposure. (B) RF speed distribution in V79 cells and derivatives (Left) and V-C8 cells and derivatives (Right). HR-deficient cells are monitored in red. Median and P values are indicated (*P < 0.05; **P < 0.01; ***P < 0.001). Median values are represented as horizontal black lines. Approximately, 100–120 fibers were scored per condition. ns, not significant.

Surprisingly, 10 μM HU reduced the RF speed in HR⁻ proficient cells (Fig. 1), although this concentration did not significantly induce CHK1 phosphorylation (Fig. S2A) or significantly affect the cell cycle distribution (Fig. S1). Even a lower HU dose (5 μM) affected genome-wide RF elongation kinetics (Fig. 1). Interestingly, the RF speed observed in WT cells treated with 5 or 10 μM HU was similar to that of untreated V-C8 and V79SM24 (HR⁻) cells (Fig. 1B). Interorigin distances were also significantly reduced in 5 or 10 μM HU-treated WT cells to values similar to that of V-C8 and V79SM24 (HR⁻) cells (Fig. S3B). These data show that the treatment of WT cells with 5–10 μM HU mimics the replication dynamics of unchallenged V-C8 and V79SM24 (HR⁻) cells.

Slowing the Replication Forks Generates MECs.

Next, we analyzed the impact of low HU doses on mitosis in WT cells. We first monitored the number of centrosomes per mitotic cells that were identified by chromosome condensation (Fig. 2A and Fig. S4). The centrosomes were monitored by immunofluorescence by using antibodies raised against two different centrosomal markers, namely, γ-tubulin (Fig. 2A) and centrobin (Fig. S4), the latter of which is a centriole-associated protein that is required for centriole elongation and stability and regulates functional mitotic spindles (36–38). We first confirmed with both antibodies that V-C8 and V79SM24 (HR⁻) mutants displayed a spontaneous increase in the frequency of mitotic cells bearing MECs compared with untreated HR⁻ proficient cells. Strikingly, the treatment of WT cells with 5 or 10 μM HU increased the frequency of cells with MECs by 2.7- to 5.5-fold, which is comparable to the frequency in V-C8 and V79SM24 (HR⁻) cells (Fig. 2B and Fig. S4B). Remarkably, the centrobin analysis suggested that these MECs may be functional. Importantly, interphase cells did not display abnormal centrosome number (Fig. S5), suggesting that extra centrosomes accumulate specifically during mitosis. The analysis of the MEC distribution demonstrated that three centrosomes were more frequently observed both in untreated V-C8 and V79SM24 (HR⁻) cells and WT cells treated with low doses of HU (Fig. 2C and Fig. S4C). In addition, although treatment with 10 μM HU generated 15–30% of cells with MECs (Fig. 2B), these values increased to 50–70% upon exposure to 100 μM HU (Fig. S6). These data argue that the frequency of MECs is correlated with the intensity of replication stress.

Fig. 2.

The impact of HR deficiency or very low HU doses (5 or 10 μM) on metaphases with aberrant centrosome number monitored with γ-tubulin antibody. (A) Examples of labeled centrosomes in mitotic cells (chromosomal DAPI staining). (Left) Normal centrosome number (= 2); columns 2–5, aberrant centrosomes number (unequal 2), causing metaphase alterations (see DNA labeling). (Scale bars: 10 μm.) (B) Frequency of mitotic cells with aberrant centrosome number. Left histograms, V79 cells and derivatives; right histograms, V-C8 cells and derivatives. The mean value ± SD was calculated from at least three independent experiments: *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. In total, 150 mitoses were scored for each experiment and condition. (C) Centrosome number distribution in mitotic cells. Upper histograms, V79 cells and derivatives; lower histograms, V-C8 cells and derivatives. In total, 100–150 mitotic cells per cell line were analyzed per condition.

Finally, exposure to low doses of aphidicolin (an inhibitor of DNA polymerases α, ε, and δ) also increased the frequency of MEC-positive cells, specifically mitotic cells with three centrosomes (Fig. S7). This result demonstrates that MECs result from different types of replication stresses, which excludes the possibility that HU may exert an unexpected side effect on MECs.

Low Replication Stress Elicits Prolonged Metaphase Arrest.

Extra centrosomes are frequently but not systematically associated with aberrant mitosis; therefore, we determined whether moderate reductions of RF speed that still allows progression through S and G₂ phases actually cause altered mitosis and aberrant chromosome segregation in V-C8 and V79SM24 (HR⁻) cells and in WT cells exposed to low HU doses.

Monitoring mitotic cells using phosphorylated histone H3 antibody or condensed chromosomes by DAPI staining demonstrated that the mitotic index increased in HR-proficient cells exposed to 5 μM HU as well as in untreated V-C8 and V79SM24 (HR⁻) cells (Fig. S8). This finding is consistent with the delay in chromosome condensation resulting from delayed replication timing (39). To determine which stage of mitosis was most impacted, we performed time-lapse video microscopy analyses of mitotic progression by using cells expressing histone H2B-GFP (Movies S1–S3). This analysis revealed that the mitotic delay specifically results from prolonged metaphase arrest both in V-C8 and V79SM24 (HR⁻) cells (Fig. 3 A and B and Movie S1) and in WT cells treated with HU doses that mimic the fork speed of V-C8 and V79SM24 (HR⁻) cells (Fig. 3B).

Fig. 3.

The impact of HR deficiency or 5 μM HU on mitosis duration. (A) Chromosome segregation kinetics. Example of time-lapse video microscopy of y-H2B-GFP–tagged wild-type (Top) and HR defective (Middle and Bottom) cells during a complete mitotic cycle. (Scale bars: 10 μm.) (B) Median kinetics of the different mitosis phase as measured by time-lapse video microscopy. For analysis, mitosis was clustered into prophase to metaphase and metaphase to anaphase. Left histogram, V79 cells and derivatives; right histogram, V-C8 cells and derivatives. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. At least 100 cells were analyzed per condition.

Low Replication Stress Generates Anaphase Chromatin Bridges and Aberrant Chromosome Segregation.

Chromatin bridges at anaphase have been described in other HR⁻ cell systems (27, 40, 41). Here, we confirmed these observations and, in addition, showed that the percentage of cells displaying anaphase chromatin bridges increased similarly in V-C8 and V79SM24 (HR⁻) cells and in WT cells exposed to 10 μM HU (Fig. 4A). Moreover, we also observed aberrant mitosis and uneven chromosome segregation (Fig. 4B), which primarily corresponded to multipolar mitosis (Fig. 3A, Bottom and Movie S3), at similar frequencies in both V-C8 and V79SM24 (HR⁻) cells, and WT cells exposed to low levels of HU (Fig. 4B). Remarkably, multipolar mitosis was generally associated with chromatin bridges (see examples in Fig. 3A, Bottom and Movie S3). Taken together, the data demonstrate that the presence of MECs is associated with abnormal mitosis both in V-C8 and V79SM24 (HR⁻) cells and in WT cells treated with low HU doses. Consistent with the centrobin analysis, the increased multipolar segregations in the situations analyzed here demonstrate that the MECs are functional. Thus, MECs represent a tractable predictive marker of aberrant mitosis upon moderate levels of replication stress.

Fig. 4.

The impact of HR deficiency or very low HU doses (5 or 10 μM) on chromosome segregation. (A) Chromatin bridges. (A Upper) Example of anaphase chromatin bridge (see also Fig. 3A, Middle). (Scale bars: 10 μm.) (A Lower) Frequency of mitotic cells with chromatin bridges. Shown are V79 cells and derivatives (A Left) and V-C8 cells and derivatives (A Right). The mean value ± SD from three independent experiments was calculated. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. In total, 150 mitoses were scored per condition. (B) The frequency of aberrant mitosis in V79 cells and derivatives (Left) and V-C8 cells and derivatives (Right). *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. The mean value ± SD of three independent experiments was calculated. In total, 150 mitoses were scored per condition.

The Addition of Deoxynucleotide Precursor to V-C8 and V79SM24 (HR⁻) Cells Restores Both RF Speed and MEC Frequency.

Next, we attempted to rescue the fork speed in V-C8 and V79SM24 (HR⁻) cells. In various mutant cell types, spontaneous RF slowing has been corrected by adding deoxynucleotide precursors (dNs) to the culture media (42–45). Particularly in colorectal cancer cells, supplying dNs corrects slow replication and complex chromosome segregation errors, which were generated by the silencing of CIN (cancer chromosomal instability) suppressor genes, although centrosome number was not affected (46). Here, in our cell lines, the addition of dNs did not alter RF progression in WT cells but rescued RF speed in V-C8 and V79SM24 (HR⁻) cells (Fig. 5A). Remarkably, exogenous dNs did not affect the frequency of cells with MECs in unchallenged WT cells, but they decreased their frequency in V-C8 and V79SM24 (HR⁻) cells (Fig. 5B). Moreover, dNs specifically decreased the frequency of mitotic cells with three centrosomes (Fig. 5C). Treatments with low HU doses or aphidicolin primarily induced cells with three centrosomes; therefore, these data support the conclusion that MECs result from RF slowing.

Fig. 5.

The effect of dNs addition on replication fork speed and the frequency of mitosis with aberrant centrosome number. (A) Replication fork speed distribution in V79 cells and derivatives (Left) and V-C8 cells and derivatives (Right) is presented. The numbers correspond to the median replication speed. Median and P values are indicated (*P < 0.05; **P < 0.01; ***P < 0.001). Median values are represented as horizontal black lines. Seventy to 145 fibers were scored per condition. (B) The quantification of mitotic cells with aberrant centrosome number (unequal 2) using γ-tubulin labeling. Left histogram, V79 cells and derivatives; right histogram, V-C8 cells and derivatives. The mean value ± SD of three independent experiments was calculated. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. In total, 150 mitoses were scored per condition. (C) Centrosome number distribution in mitotic cells monitored by immunofluorescence using γ-tubulin labeling. V79 cells and derivatives (Left); V-C8 cells and derivatives (Right). At least 100 mitotic cells were analyzed per condition.

Discussion

The data presented here demonstrate that moderate decreases in RF movement do not detectably delay the progression in S and G₂ phases and mitotic entry; however, they cause severe mitosis defects via MEC formation, resulting in multipolar mitosis and uneven chromosome segregation. The model in Fig. 6 summarizes the mechanisms that may link HR, replication, and mitosis to protect against global chromosome instability: Different causes decelerate RF (HR defect, dNTP shortage, polymerase inhibition); the firing of origins compensates for RF slowing upon moderate levels of stress, allowing maintenance of the global replication rate and normal cell cycle progression (47); however, the origins cannot be activated in several genomic regions, resulting in mitotic entry with incomplete replicated chromosomes (48). This process should be facilitated by the fact that low levels of replication stress escape to the surveillance pathway. Incomplete replicated regions impair chromatid segregation (29), generating chromatin bridges that activate mitotic arrest. Indeed, in this study, we demonstrate the increased frequency of mitotic cells and metaphase delay. The capacity to bypass mitotic arrest without resolving the causes of the arrest has been described in yeast in a process called adaptation (49, 50). We propose that anaphase chromatin bridges are revealed in cells bypassing the mitotic arrest in a process that is reminiscent of the adaptation process described in yeast. In addition, it is tempting to speculate that during the mitotic arrest bypass, opposite forces between chromosome migration and metaphase arrest lead to the splitting of centrosomes, resulting in active MECs and multipolar segregations. This model is consistent with the fact that extra centrosomes are detectable during mitosis but not in interphase; that these centrosomes are functional, causing multipolar segregations; and that almost all multipolar segregations are associated with chromatin bridges. Incomplete replication results in local chromosome abnormalities during mitosis (29). Here, very moderate replication stress lead to MEC formation, which do not contain DNA but amplify the signal to the whole genome by generating multipolar mitosis and aberrant chromosome segregation. Therefore, very low replication stress not detected by the cell cycle checkpoint leads to global chromosome missegregation involving even fully replicated chromosomes.

Fig. 6.

RF deceleration alters global chromosome segregation. Different causes such as dNTP shortage, polymerase inhibition, and HR deficiency cause RF deceleration, which elicits the firing of RFs. At such low or endogenous stresses, cells do not arrest and, thus, progress through the G₂ phase with incompletely replicated DNA. RF deceleration exacerbates these processes leading to chromatin bridges in nonreplicated regions. However, these cells are blocked at metaphase. Bypassing this arrest (in an “adaptation-like” process) causes abnormal mitosis including MEC, anaphase chromatin bridges and multipolar cells, which results in uneven chromosome segregation and aneuploidy in the whole genome, even for replicated chromosomes.

Centrosome reduplication arises when centrosome duplication is uncoupled from cell cycle progression when replication is completely arrested by high doses of HU for a cell division cycle (51). Centrosome reduplication should primarily cause a paired number of extra centrosomes; however, we essentially observed cells with three centrosomes. In addition, interphase cells did not display an increased number of centrosomes, which would be expected for centrosome reduplication. Finally, WT cells treated with low doses of HU and V-C8 and V79SM24 (HR⁻) mutant cells cycled normally; therefore, centrosome reduplication does not account for the phenomenon we observed here. Alternatively, centrosome splitting has been associated with DNA damage accumulation (52).

HR proteins are associated with centrosomes, and centrosome duplication defects are thought to directly result from HR misregulation (30, 31). In addition, CHK1 activation controls centrosome alterations and amplification (25, 53–55). However, we observed MECs in the HR proficient cells treated with HU or aphidicolin doses that did not trigger CHK1 phosphorylation, which argues against these hypotheses. Rather, the data presented here demonstrate that HR protects against endogenous stress arising upon RF slowing. Thus, deciphering the molecular mechanism at the origin of the endogenous replication stress in HR⁻ cells represents an exciting challenge.

Theodor Boveri hypothesized one century ago that tumors originate from improper chromosome segregation, thus creating aneuploid cells that undergo clonal expansion. He proposed that in some cases, abnormal chromosome segregation and aneuploidy are caused by extracentrosome generation, leading to multipolar cells (56, 57). Our data suggest that modest RF slowing imbalances general chromosome segregation. Among the stresses responsible for RF slowing, dNTP pools appear to be very sensitive limiting factors. Two hypothesis can account for the data in HR⁻ deficient cells: (i) the nucleotide pools are spontaneously affected by processes that remain to be identified; and (ii) part of the dNTP production can be channeled to the endogenous damages, which should be more persistent in HR⁻ deficient cells, resulting in a dNTP shortage for replication. Because the nucleotide pool is a tight limiting factor for replication but is not expendable in mammalian cells after genotoxic stress, this outcome might lead to replication fork deceleration without affecting the global intracellular nucleotide pool (58–60). Remarkably, deoxynucleotide deficiencies promote genomic instability at early oncogenic stages (43). Consistent with our data, both replication stress and centrosome abnormalities have also been reported at early stages of malignancy. Interestingly, HR protects against spontaneous endogenous replication stress, is affected in most familial breast cancers (61), and likely occurs in a high frequency of sporadic cases (62). Therefore, the data presented here shed light on the importance of HR at the molecular interface between replication and mitosis when cells face spontaneous endogenous or low genotoxic stresses that do not trigger cell cycle checkpoints or prevent the cells from entering mitosis.

Materials and Methods

Cells and Treatments.

Cell lines were cultured at 37 °C in 5% CO₂ in Eagle’s minimal essential medium supplemented with 10% (vol/vol) FBS, 2 mM glutamine, 200 IU/mL penicillin, and 200 µg/mL streptomycin. V79SM24 is a derivative of the V79 cell line that stably expresses the RAD51 dominant negative SMRAD51; V79puro are V79 cells that were transfected with the empty expression vector (21, 32). We also used the V-C8 BRCA2-defective cell line and its V-C8#13 counterpart in which BRCA2 is complemented by human chromosome 13 (26, 63). V-C8#13 cells were used as a second WT cell line.

Cells were exposed to the indicated HU or aphidicolin doses for 48 h. Nucleotide precursor supply analysis was performed by adding a mix of the four dNs (20 μM each) for 48 h. Deoxycytidine (Sigma D0776) was solubilized in 1 M NaOH (100 mM). Deoxyadenosine (Sigma D8668) was solubilized in 0.1 M NaOH (20 mM). Thymidine (Sigma T1895) was solubilized in H₂O (50 mM). Deoxyguanosine (Sigma D7145) was solubilized in 1 M NH₄OH (100 mM).

Molecular Combing.

Molecular combing was performed as described (64). Iododeoxyuridine (IdU) labeling and chlorodeoxyuridine (CldU) labeling (20-min pulse labeling for each) were performed as described (42). Briefly, chromatin fibers were incubated with rat anti-BrdU (1:25, OBT0030, clone BU1/75 ICR1; AbD Serotec) and FITC-conjugated mouse anti-BrdU (1:5, 347583, clone B44; BD Biosciences) for 1 h at 37 °C. After a short wash step (0.5 M NaCl, 20 mM Tris at pH 7.8, and 0.5% Tween 20), the fibers were incubated with goat anti-mouse Alexa 488 (1:50, A-11029; Invitrogen) and goat anti-rat Alexa 594 (1:50, A-11007; Invitrogen) for 1 h. DNA counterstaining was performed by using anti-DNA antibody (1:20, MAB3034, clone 16–19; Millipore) and the following two secondary antibodies: goat-anti mouse Cy5.5 (1:100, ab6947; Abcam) and donkey anti-goat Cy5.5 (1:100, ab6951; Abcam). For each condition, ∼100 fibers with symmetrical green-red labeling were analyzed. Images were captured by using a 63× oil immersion objective of a motorized Axio Imager.Z2 epifluorescence microscope (Carl Zeiss) that had been equipped with a high-sensitivity cooled interline CCD camera (Cool SNAP HQ2; Roper Scientific) and a PIEZO stage (Physik Instrumente). Image acquisition was performed with MetaMorph software (Molecular Devices).

Centrosome Analysis.

Cells were fixed in methanol for 15 min at –20°C and permeabilized with acetone. After three washes in PBS, the samples were incubated in 2% (wt/vol) BSA for 30 min and washed three times in PBS. Centrosomes were stained by using γ-tubulin antibody (Sigma T3559, 1/300) or centrobin antibody (Abcam 70448, 1/40) diluted in 0.5% BSA and 0.05% Tween 20 for 1 h at 37°C. The cells were washed three times in 0.05% Tween 20 in PBS followed by incubation with Alexa Fluor 594-conjugated goat anti-rabbit antibodies (Molecular Probes), which were diluted 1:400 in 0.5% BSA and 0.05% Tween 20 for 1 h at 37 °C. After three washes in 0.05% Tween 20 in PBS, the cells were incubated with DAPI (1 μg/mL). In each case, 100–150 metaphase cells were analyzed per condition. Images were captured by using a 63× oil immersion objective of a motorized Axio Imager.Z2 epifluorescence microscope (Carl Zeiss) that had been equipped with a high-sensitivity cooled interline CCD camera (Cool SNAP HQ2; Roper Scientific) and a PIEZO stage (Physik Instrumente). Image acquisition was performed with MetaMorph software (Molecular Devices).

Time-Lapse Video Microscopy.

Cells lines expressing GFP-tagged H2B were cultured for 48 h before video microscopy on glass coverslips. The cells were enclosed in a Ludin chamber, and a photo was taken every 2 min with a 40× objective for at least 16 h. Image acquisition was performed with a motorized inverted microscope (Olympus IX81) that had been wired to a CoolSNAP HQ camera (Princeton Instruments). During image capture, the carbon dioxide levels were 5% and the temperature was 37 °C, which are typical cell culture conditions. MetaMorph software (Molecular Devices) was used for image capture and analysis.

Statistical Analysis.

RF speed, interorigin distance, and mitosis duration were compared by using the Mann–Whitney test. Mitotic cells and cells with MEC percentages were analyzed by Fisher’s exact test. All of the tests were two-sided, and P < 0.05 was considered statistically significant.