HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 39, 36639-36646, September 28, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/39/36639    most recent M104501200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Anglana, M. Articles by Debatisse, M. Search for Related Content PubMed PubMed Citation Articles by Anglana, M. Articles by Debatisse, M. Social Bookmarking                      What's this?

Dual Control of Replication Timing

STOCHASTIC ONSET BUT PROGRAMMED COMPLETION OF MAMMALIAN CHROMOSOME DUPLICATION^*

Mauro Anglana and Michelle Debatisse^§

From the UMR147, Batiment Trouillet-Rossignol, Institut Curie/CNRS, 26 Rue d'Ulm, 75248 Paris, France

Received for publication, May 17, 2001, and in revised form, July 2, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In mammalian cells, DNA replication proceeds according to a precise temporal order during the S phase, but how this program is controlled remains poorly understood. We analyzed the replication-dependent bromodeoxyuridine banding of chromosomes in Chinese hamster cells treated with the spindle poison nocodazole. In these cells, nocodazole induces a transient mitotic arrest, followed by DNA re-replication without intervening cell division. Nuclear fragmentation is often observed in tetraploid derivatives, and previous studies suggest that replication timing of chromosomes could be affected when they are segregated into different micronuclei. Here we show that the onset of replication is frequently asynchronous on individual chromosomes during the re-replication process. Moreover, fluorescence in situ hybridization analysis revealed that replication synchrony is equally altered in fragmented and non-fragmented nuclei, indicating that asynchronous onset of replication is not dependent on physical separation of the chromosomes into isolated compartments. We also show that the ordered program of replication is always preserved along individual chromosomes. Our results demonstrate that the onset of replication of individual chromosomes in the same nuclear compartment can be uncoupled from the time of S-phase entry and from the programmed replication of chromosome sub-domains, revealing that multi-level controls contribute to establish replication timing in mammalian cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA replication of eukaryotic chromosomes initiates at multiple sites and at various times during S phase, according to an orderly program (1). After pulse-labeling of proliferating cells with thymidine analogs, foci of ongoing replication can be visualized in interphase nuclei by fluorescence microscopy. Early and late replication patterns have been identified, revealing highly organized nuclear compartments, well conserved throughout successive cell generations (2-7). Studies with living cells showed that replication foci do not change their positions in interphase nuclei but assemble and disassemble continuously during S phase, giving rise to highly dynamic patterns (8). All these findings are consistent with the hypothesis that the DNA replication process is strictly associated with and functionally linked to the topological organization of DNA in interphase nuclei (9, 10).

The DNA replication process has been primarily analyzed using the yeast Saccharomyces cerevisiae as a model. These and other studies performed with different eukaryotic cells revealed that initiation of replication depends both on the sequential recruitment of specific factors on replication origins and on the activity of trans-acting protein kinases (11). However, the mechanisms that regulate the establishment of early and late origins are still poorly understood. Interestingly, it has been demonstrated that the programming of one late replication origin occurs during the G₁ phase between mitosis and START (12). In good agreement with this latter finding, ex vivo analysis of DNA replication of isolated Chinese hamster nuclei introduced in Xenopus egg extracts suggested that replication timing is determined during early G₁ in mammalian cells, at a stage called the "timing decision point," after completion of the positioning of chromosomal domains within the nucleus (13). However, some evidence indicates that early and late replicating chromosomal domains are unaffected in abnormal nuclear structures, such as micronuclei (2), or during chromosome repositioning that occurs when quiescent human cells re-enter the cell cycle (14).

Recent work suggested that a passage through the S and M phases plays a key role in correct repositioning of chromosomes during the next interphase (14). Interestingly, treatment with spindle poisons affected the synchrony of replication of individual chromosomes (15). After prolonged metaphase block, mammalian cells progressively exit mitosis without an intervening cell division, a phenomenon called adaptation or mitotic slippage, giving rise to tetraploid derivatives (16-19). Normal human and rodent cells undergo a second block in G₁, which prevents such cells from further cycling (17, 20-23), whereas transformed and immortalized cells generally re-enter S phase. Mitotic slippage is often accompanied by abnormal nuclear budding and micro-nucleation (15, 24, 25). Our study was aimed to investigate the consequences of nuclear alteration on the temporal programming of chromosome duplication of mammalian cells upon treatment with the spindle poison nocodazole.

We studied the timing of replication in nocodazole-treated and untreated cells of the Chinese hamster line GMA32, which are permissive for S-phase entry after mitotic slippage. We demonstrate that the synchrony of the onset of replication on different chromosomes is impaired during the reduplication process and that this phenomenon is not dependent on the segregation of chromosomes into different micronuclei. We show that chromosomes that tend to cluster after adaptation also tend to maintain replication synchrony, suggesting a correlation between nuclear spatial organization and replication timing control. Nevertheless, the temporal program of replication of chromosomal sub-regions is unaltered. Besides providing new evidence that DNA replication timing can be affected by changes in the general structure of the nucleus, we identify a phenotype characterized by a stochastic onset of replication from chromosome to chromosome within a cell, whereas the relative order of replication of chromosomal sub-domains remains unaffected.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture, FACS¹ Analysis, and Chromosome Preparation-- Chinese hamster GMA32 cells (26) and GMA32-T cells, a tetraploid sub-clone of GMA32, were grown in minimal essential medium supplemented with antibiotics, amino acids, vitamins, and10% horse serum (Life Technologies, Inc.). In general 4 × 10⁵ GMA32 cells were plated in a 10-cm diameter dish. Nocodazole (Sigma) was added at the final concentration of 10 µM. Thymidine analog 5-bromodeoxyuridine (BrdUrd) (Sigma) was added at the final concentration of 30 µM. When pulse-labeled with BrdUrd, cells were washed with fresh medium and grown in fresh medium supplemented with thymidine at the final concentration of 100 µM.

For FACS analysis, cells were collected and resuspended in 1 ml of cold GM buffer (6 mM glucose, 137 mM NaCl, 5.4 mM KCl, 1.1 mM Na₂HPO₄, 1.1 mM KH₂PO₄, 0.5 mM EDTA) and 3 ml of 100% ethanol. After 2 h at 4 °C, cells were washed once in 1× PBS and incubated in propidium iodide solution (25 µg/ml propidium iodide, 25 µg/ml RNase A in 1× PBS) for 30 min at room temperature. Cellular DNA content was measured with a FACScan analyzer (Cellquest software; Becton Dickinson). Metaphase plates were prepared as already described (27).

Fluorescence Plus Giemsa Detection-- Slides were stained as described (28) with some modifications. They were immersed in a 150 µg/ml Hoechst 33258 (Sigma) solution in water for 15 min and then washed in distilled water. After washing, coverslips were mounted in distilled water, and slides were put on aluminum foil on ice and exposed to excitation light (UV light, 300 W) for 20 min at a distance of 16 cm. Next, slides were stained with Giemsa 2% (Sigma) for 10 min and observed with an Axiophot (Zeiss) microscope. Images were assembled and annotated using Adobe Photoshop 5.0 software on Power Macintosh computers.

In Situ Hybridization and BrdUrd Immunodetection-- FISH was performed as described (29, 30) with minor modifications (31). Cosmids were biotinylated by nick translation (BioNick labeling system kit; Life Technologies, Inc.) and used as probes. Cosmids 61W14, 61WG1, 56W11, and 56A1 (32) are specific for the AMPD2 region on chromosome 1. Cosmid P3C3 (isolated from a Chinese hamster chromosome 1-specific library)² is a marker for chromosome 1. Signals were developed with alternating layers of fluorescein-conjugated avidin (Vector) or Texas Red-conjugated avidin (Molecular Probes) and biotin-conjugated goat anti-avidin antibody (Vector). DNA was counterstained with VECTASHIELD mounting medium with propidium iodide and observed with an Axiophot (Zeiss) fluorescence light microscope.

BrdUrd was immunodetected following described procedures (33, 34) with some modifications. Chromosomes were denaturated as in FISH experiments and incorporated BrdUrd was detected with two alternating layers of fluorescein-conjugated mouse anti-BrdUrd antibody (Becton Dickinson) and fluorescein-conjugated donkey anti-mouse antibody (Jackson ImmunoResearch). Chromosomes were counterstained and observed as described above. Images were assembled and annotated as described in the previous section.

Nuclear Envelope and Lamin Detection-- For nuclear envelope analysis, cells were incubated with the lipophilic dye DiOC₆ (Molecular Probes) at the final concentration of 0.5 µg/ml for 15 min at 37 °C, washed three times for 10 min with warm medium, and fixed with 3.5% paraformaldehyde in PBS for 10 min. Cells were then rinsed in PBS, and nuclei were counterstained with VECTASHIELD mounting medium with 4,6-diamidino-2-phenylindole. Signals were observed with an Axiophot (Zeiss) fluorescence light microscope. Nuclear section images were obtained with a Bio-Rad MRC 1024 confocal laser-scanning microscope with a ×63 objective (Micro Nikon) and a 4.15 zoom factor. Images were assembled and annotated using ImageJ 1.17y and Adobe Photoshop 5.0 software on Power Macintosh computers.

For lamin detection, cells were fixed with 3.5% paraformaldehyde in PBS for 10 min and permeabilized with 0.2% Triton X-100, PBS for 10 min. Lamins were detected with two alternating layers of polyclonal goat anti-lamin B plus polyclonal goat anti-lamin A/C antibodies (Santa Cruz Biotechnology Inc.) and Alexa Fluor 488 donkey anti-goat antibody (Molecular Probes). Nuclei were counterstained with VECTASHIELD mounting medium with propidium iodide. Signals were observed, and images were assembled as described above.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

GMA32 Cells Undergo Re-replication and Exhibit Abnormal Nuclear Shape following Mitotic Slippage-- Using the Chinese hamster cell line GMA32, we studied the re-replication process that follows mitotic slippage. An unsynchronized cell population was grown in the presence of nocodazole, and aliquots were analyzed at different times after drug addition. FACS analysis (Fig. 1) showed that cells with a 4C (4 chromatids per cell) DNA content accumulated progressively, reaching a maximum 8-10 h after drug addition. Cytogenetic analysis revealed that the population contained up to 57% of mitotic cells after 10 h in drug-containing medium as compared with some 4% in untreated exponentially growing cells. Thus, GMA32 cells were able to delay mitotic exit in response to spindle disruption. After 12 h of growth in the presence of nocodazole, some cells with more than 4C DNA content appeared, and the rest of the cells progressively engaged re-replication during the following 10 h. Twenty-two hours after drug addition, most of the cells had an 8C DNA content, and cytogenetic analysis indicated that some 35% of them were tetraploid or, more frequently, slightly hypo-tetraploid mitotic cells. Fig. 1 shows that nocodazole-induced lengthening of mitosis indeed perturbs the normal distribution of the cells in the different phases of cell cycle. However, synchronization of the cells does not result from this treatment, since all the cells of the population are similarly delayed before mitotic slippage.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   GMA32 cells exit mitosis and engage re-replication in response to nocodazole treatment. Cells were grown in nocodazole-containing medium for up to 22 h, and aliquots were analyzed by FACS during this treatment. 2C, 4C, and 8C indicate number of chromatids per cell. The number of hours in nocodazole is indicated in the boxes. The profile at 0 h refers to untreated cells.

Nuclear budding and micronucleation have been repeatedly observed in cells undergoing mitotic slippage. In our experimental model, micronucleation was observed in 32, 48, and 68% of the cells from populations grown, respectively, for 10, 12, and 22 h in the presence of nocodazole (see Fig. 3 and 4 for examples of micronucleation). The micronuclei observed in these cells were distinct from apoptotic bodies, since FACS profiles indicated that very few cells undergo apoptosis in the populations treated with nocodazole (Fig. 1) and since micronucleated cells were not terminal dUTP nick-end labeling-positive (not shown).

The Onset of DNA Replication Is Altered in Cells Undergoing Re-replication following Mitotic Slippage-- To investigate whether DNA replication timing is affected in tetraploid cells, which were obtained after prolonged growth in the presence of nocodazole, we analyzed the replication-dependant banding of chromosomes in cells that were pulse-labeled with BrdUrd. As a control, exponentially growing GMA32-T cells, a tetraploid sub-clone of the GMA32 cell line, were pulse-labeled with BrdUrd for 1 h in independent plates, 5, 6, 7, 8, 9, 10, or 11 h before harvesting at a same end point. Such a protocol was required to allow the observation of metaphase plates with replication banding patterns representing all possible stages of S phase. We examined a total of 620 BrdUrd-labeled metaphase plates, about 90 from each aliquot. In all of them, we found that each chromosome was labeled, which indicates that replication starts synchronously on all chromosomes of a cell at the onset of S phase and proceeds on each of them for all the duration of the S phase (Fig. 2a). Nocodazole-treated cells were then analyzed. BrdUrd was added to the medium 12, 13, or 14 h after nocodazole addition and removed 1 h later. For all time points, the cells were collected 22 h after nocodazole addition when, as mentioned above, around 35% of them were mitotic, tetraploid, or slightly hypo-tetraploid, and very few, if any, had a DNA content that was more than 8C (Fig. 1). We examined a total of 279 labeled metaphase plates, corresponding roughly to 90 examples from each experimental condition. In 23 of them, we observed that BrdUrd was incorporated only in a variable subset of the chromosomes, and the presence of these abnormally labeled metaphase plates was independent of the time of BrdUrd addition (Fig. 2, b-c). To determine whether asynchrony results from impaired replication rates or rather from altered replication onset on individual chromosomes, nocodazole-treated cells were pulse-labeled with BrdUrd for various times (2-6 h) starting from 10 h upon nocodazole addition, a time at which cells had not yet entered re-replication (Fig. 1). We focused on cells pulse-labeled for 5 and 6 h, and we observed that 11 out of 63 and 3 out of 38 metaphase plates, respectively, displayed asynchronous patterns as described above (Fig. 2d). Thus, in cells challenged with nocodazole, the onset of replication of some chromosomes is asynchronous with respect to the progression of S phase. To determine whether nocodazole per se affects the replication synchrony of chromosomes, we analyzed the replication-dependent banding in 600 metaphase plates from cells grown in nocodazole for 10 or 11 h and pulse-labeled for 1 h with BrdUrd at different time points before recovery. We never observed an asynchronous onset of replication among the individual chromosomes from each metaphase plate (not shown), and we concluded that nocodazole treatment does not directly affect replication timing.

View larger version (62K): [in this window] [in a new window]   Fig. 2.   Variable subsets of chromosomes from nocodazole-treated cells engage DNA replication asynchronously with respect to the progression of S phase. Cells grown with or without nocodazole for 22 h were pulse-labeled with BrdUrd, and replication-dependent chromosome bands were detected by immunofluorescence. Shown are metaphase plates from control GMA32-T tetraploid cells (a) and nocodazole-treated GMA32 cells (b-c) after a 1-h BrdUrd pulse at different time points during the cell cycle. A metaphase plate from treated GMA32 cells after a BrdUrd pulse from 10 to 15 h upon nocodazole addition is shown in panel d. In b-d, the absence of label on some chromosomes (arrows) suggests that they were fully replicated at a time interval different from when BrdUrd was supplied. Bar, 5 µm.

Replication Asynchrony Does Not Depend on Physical Segregation of Chromosomes in Different Nuclear Compartments-- To confirm our observations and to establish whether replication synchrony is disturbed specifically in subsets of chromosomes that have been physically separated by the micronucleation process, we performed FISH experiments to study DNA replication in interphase nuclei. The rationale for using FISH has been established previously by work showing that the replication status of a chromosomal locus can be determined after hybridization with a probe specific for this locus, by scoring the number of the hybridization spots and studying their arrangement in a nucleus. Indeed, for each homologue, a single spot (singlet) is revealed before replication, whereas a couple of close spots (doublet), corresponding to the presence of hybridization signal on both chromatids, is observed after replication (35). We performed hybridization with a combination of probes specific for the AMPD2 region, which lies on the q arm of Chinese hamster chromosome 1 (36). These probes were chosen because previous experiments established that in GMA32 cells the AMPD2 locus replicates synchronously on both homologues, early in S phase.³ We first determined the frequency of abnormal patterns in control GMA32-T cells (Table I, Control). In metaphase plates, a spot is expected to be present on each chromatid of the fully replicated chromosomes 1. Thus, the frequency of missing spots on metaphase plates allows an unambiguous determination of the hybridization efficiency. We observed that 77% of the tetraploid metaphase plates displayed eight spots and that 21% of them exhibited seven spots. The absence of more than one spot was very rarely observed. The same slides were also used to analyze interphase nuclei. As expected for an unsynchronized cell population, cells exhibiting four singlets, in which the AMPD2 locus had not yet replicated, were observed in interphase nuclei. However, we excluded these cells from the counts since they had no equivalent in the metaphase plates and would have biased the comparison of the frequencies of cells with abnormal patterns. As shown in Table I, the very same percentage of cells displayed four doublets or three doublets and one singlet in interphase nuclei and in metaphase plates (see Fig. 3a for an example of a nucleus with four doublets). The proportion of cells exhibiting two doublets and two singlets was roughly similar in interphase nuclei and in metaphase plates, although the low number of cases observed precluded a reliable statistical comparison. The pattern "one doublet plus three singlets" was never observed in control cells. Taken together, our results indicate that the hybridization efficiency is similar in interphase nuclei and metaphase plates.

View larger version (88K): [in this window] [in a new window]   Fig. 3.   Replication of the AMPD2 region occurs asynchronously in the distinct copies of chromosome 1 in nocodazole-induced tetraploid cells. FISH with probes specific for the AMPD2 region was performed on spreads from GMA32-T control cells and from GMA32 cells treated for 22 h with nocodazole. Different arrangements of hybridization spots in interphase nuclei are shown. An example of interphase nucleus from control cells with four couples of hybridization spots (four doublets) is shown in panel a. Each spot corresponds to the AMPD2 region lying on one single chromatid. Examples of interphase nuclei from nocodazole-treated cells are shown (b-f): non-fragmented nucleus with four doublets (b), fragmented nucleus with four doublets (c), non-fragmented nucleus with two doublets and two single hybridization spots (two singlets) (d), fragmented nucleus with two doublets and two singlets (e), and non-fragmented nucleus with three doublets and one singlet (f). Note that in nuclei from nocodazole-treated cells, signals (singlets or doublets) from single chromosomes tend to be clustered two by two. Bar, 5 µm.

The same analysis was performed with GMA32 cells grown for 22 h in the presence of nocodazole. From comparison of FACS profiles and mitotic indexes, we estimated that at that time the proportion of cells in S and G₂ phase was roughly similar to that of untreated control cells. A total of 165 metaphase plates and 278 nuclei were scored (Table I, Nocodazole). In this experiment, the hybridization efficiency was higher than in the control experiment since 88% of the plates exhibited the expected eight spots. However, in striking contrast with the results obtained with control cells, the frequency of interphase nuclei exhibiting four doublets was lower (53%). On the contrary, patterns such as three doublets plus one singlet and two doublets plus two singlets were, respectively, observed in 22 and 25% of the nuclei, a frequency significantly higher than in metaphase plates, the increase being more pronounced for the latter category (see Fig. 3, b-f, for examples of nuclei with different patterns). This reveals that the replication of homologues can be asynchronous in nocodazole-treated cells and indicates that DNA replication is preferentially altered in couples of chromosomes (two-thirds of the abnormalities taking into account the bias resulting from hybridization failure) rather than in single ones (one-third). Interestingly, the different types of replication patterns were found at rather similar frequencies in non-fragmented and fragmented nuclei (Table I, Nocodazole).

As a control, we verified that nocodazole treatment per se does not affect the separation of sister chromatids in a way that could increase the frequency of unresolved double spots. The same probes were used to study cells that replicated their DNA in the presence of nocodazole before mitotic block. GMA32 cells were treated with nocodazole for 7 or 8 h before recovery and FISH analysis. The diploid cells in interphase observed in these conditions entered and completed S phase in the presence of the drug. Indeed, cells that were in S phase when nocodazole was added already reached mitosis 7 or 8 h later (Fig. 1). Slides were also prepared from untreated GMA32 cells as control. A total of 191 nuclei from treated cells and 99 nuclei from control cells were scored, and no significant differences in the relative frequencies of nuclei exhibiting patterns with one singlet and one doublet or two doublets were observed (not shown). We concluded that nocodazole treatment does not affect the reliability of spot-counting. Moreover, these results support our previous conclusion that nocodazole does not directly affect the synchrony of replication.

To verify that the so-called non-fragmented nuclei are devoid of internal abnormalities, which could have escaped detection in the experiments described above, we analyzed the spatial organization of the nuclear envelope and of the nuclear lamina. Cells treated with nocodazole for 22 h and control cells were incubated with the lipophilic dye DiOC₆, which specifically labels the membranes, including the endoplasmic reticulum, of which the nuclear envelope is a specialized compartment (37). We analyzed 251 nocodazole-treated and 278 control cells using conventional fluorescence light microscopy. The DiOC₆ signal showed that the nuclear envelope is essentially unaffected by nocodazole treatment in normally shaped nuclei (not shown). This point was verified by analyzing the same cell preparations by confocal microscopy. Although each micronucleus was found individually embedded in nuclear envelope, nuclear compartmentalization was not observed in normally shaped nuclei of nocodazole-treated or untreated cells (Fig. 4). We also studied the distribution of the lamina, which associates with the inner membrane of the nuclear envelope (38) by indirect immuno-fluorescence with a combination of anti-lamin A, B, and C antibodies, the major components of the nuclear lamina. We observed 284 nocodazole-treated and 304 control cells by conventional microscopy. Perinuclear and internal lamin structures were detected in both types of cells, and confocal microscopy analysis confirmed the similarity of the lamin distribution in treated and untreated cells.

View larger version (70K): [in this window] [in a new window]   Fig. 4.   In non-fragmented nuclei from nocodazole-treated cells, the nuclear envelope organizes as in nuclei from untreated control cells. Cells grown with or without nocodazole for 22 h were labeled with DiOC6 (green) and fixed with paraformaldehyde. Shown are serial optical sections of nuclei obtained by confocal microscopy at 0.6-µm intervals; the image at the top is the top of the nucleus. a, normal nucleus from untreated cells; b, non-fragmented nucleus; c, fragmented nucleus from nocodazole-treated cells. Bar,10 µm.

The whole set of results indicated that the onset of replication is asynchronous from chromosome to chromosome during the S phase that follows mitotic slippage and that this phenomenon is not restricted to cells with a fragmented nucleus. Indeed, treated cells that contain a single nuclear compartment that exhibits an apparently normal spatial distribution of either the nuclear envelope or the nuclear lamina also fail to maintain chromosome synchrony. Thus, asynchrony might derive from a subtle perturbation of nuclear organization induced directly or indirectly by disruption of the microtubule network.

Relative Replication Order of Chromosome Sub-regions Is Not Affected in Treated Cells-- Early and late replicating regions are visualized in metaphase plates using different banding techniques, relying often on continuous or discontinuous incorporation of thymidine analogs (Ref. 39 and references therein). To determine whether the asynchrony observed from chromosome to chromosome after mitotic slippage correlates with abnormal timing of replication of sub-regions within the chromosomes, we compared the dynamic banding of chromosomes in metaphase plates of control GMA32-T cells and nocodazole-treated GMA32 cells. In both cases, the cells were labeled continuously with BrdUrd, which was added to the medium 4, 5, or 6 h before cell recovery. This protocol allowed us to reveal the banding patterns typical of the second part of S phase. Cytogenetic analysis was performed focusing on two chromosomes; these are the only chromosome X present in these cells, easily identified by its highly condensed long arm, and chromosome 1. Indeed, chromosome 1 was of special interest in these cells because a pericentric inversion has occurred on one homologue (referred to as 1B, whereas the normal one was named 1A; Refs. 27 and 31), allowing us to distinguish both of them. The replication timing of X has been studied previously in various mammalian cells, and its orderly program of replication was partially re-determined here. In Fig. 5A a diagram shows some informative banding profiles of chromosome X obtained with GMA32-T control cells. The temporal order of replication of the sub-regions of chromosomes 1A and 1B, which were identified unambiguously by FISH with the probe P3C3, was reconstituted in parallel (Fig. 5A). When reconstituting these replication patterns we did not take into account the relative intensity of individual BrdUrd bands, which was highly variable, but we considered the presence or the absence of signal on whole chromosome domains (bands a, b, c, d, and e in Fig. 5A). Indeed, these patterns were similar on the two copies of each pair of chromosomes X, 1A and 1B. Then, we analyzed the banding patterns of these chromosomes in metaphase plates from GMA32 cells grown for 22 h in nocodazole-containing medium. We studied 117 such metaphase plates and never found abnormalities in the relative order of replication of the sub-regions defined upon study of control cells, even in the metaphases that display asynchronous homologues (Fig. 5B). This point will be discussed in further detail in the following paragraph.

View larger version (58K): [in this window] [in a new window]   Fig. 5.   Temporal program of replication of chromosomal sub-domains is maintained in nocodazole-treated cells. GMA32-T control cells and GMA32 cells grown in nocodazole-containing medium for 22 h were labeled continuously with BrdUrd for 4, 5, or 6 h before recovery, and replication-dependent chromosome bands were detected by immunofluorescence. FISH with the cosmid P3C3 probe was performed to mark chromosome 1 and allow the identification of the two chromosome 1 homologues. A, four informative banding profiles of active chromosome X (top) and chromosomes 1A and 1B (bottom) are drawn to the side of the corresponding chromosome image. The amount of incorporated BrdUrd is proportional to the duration of its incorporation during S phase. Therefore, as an example, chromosomes from cells that were at an earlier stage of S phase, when the BrdUrd was added (left end of the diagram), display more numerous and large BrdUrd-labeled domains than the chromosomes from cells at a later stage (right end). The arrow indicates the direction of S phase progression. Banding profiles of active X chromosome are shown above the synchronous patterns of the chromosome 1 homologues. Five BrdUrd-labeled chromosomal bands or regions (a, b, c, d, and e) were chosen for both X chromosome and chromosome 1. Band a in chromosome 1 colocalizes with the portion involved in the pericentromeric inversion. Region c of chromosome 1 seems to be composed of close multiple bands. We considered them in a unique region since they were synchronous. The red dots represent the target region of the FISH probe. A blue arrow points to chromosome X, whereas a white arrow and a green arrow point to chromosomes 1A and 1B, respectively, at the level of the red dots. B, images of metaphase plates from nocodazole-treated GMA32 cells obtained with two different microscope filters in order to show P3C3 FISH signal (above) and the BrdUrd-containing regions (below). Chromosomes X are indicated with a blue arrowhead, whereas chromosomes 1A and 1B are shown with a white or a green arrowhead, respectively. Note that in b and c, the two couples of chromosome 1 show two different banding profiles. Bar, 5 µm.

Replication Alterations Preferentially Affect Couples of Sister Chromosomes-- Analysis of the AMPD2 hybridization signals in interphasic nuclei also showed that in nocodazole-treated cells homologues tend to cluster two by two, whereas in untreated GMA32-T control cells they do not (Fig. 3). This was confirmed by analysis of 96 non-fragmented and 98 fragmented nuclei from cells treated with nocodazole for 22 h and 101 nuclei from GMA32-T control cells after in situ hybridization with another chromosome marker, P3C3 cosmid (not shown). We reasoned that such a chromosome distribution in nocodazole-treated cells might have derived from a defect in the separation of sister chromatids at anaphase. Indeed, upon prolonged spindle disruption by nocodazole, the cells enter a G₁-like state without passage through anaphase (16-19). As mentioned, interphase nuclei displaying two couples plus two single spots from AMPD2 markers were the most frequent among the nuclei exhibiting an altered replication profile (Table I, nocodazole). We also noticed that the two doublets and the two singlets tend to lie close to each other (Fig. 3, d and e), suggesting that clustering favors the maintenance of replication synchrony. This interpretation implies that synchrony should be preferentially conserved for couples of sister chromosomes (we named "sister chromosomes" the chromosomes that derive from the replication of sister chromatids). To verify this model, we took advantage of the possibility of distinguishing unambiguously between chromosome 1 homologues in metaphase plates after hybridization with probe P3C3. Replication patterns revealing asynchrony between couples of chromosomes 1 were observed in 13 out of 117 metaphase plates (Fig. 5B, b and c). In good agreement with our model, the two X chromosomes always showed identical patterns (Fig. 5B, a-c). As mentioned in the previous paragraph, both types of banding patterns identified on chromosome 1 in these abnormal metaphase plates were consistent with banding patterns observed in control cells. We found that in 13 out of 13 cases, the two 1A chromosomes were synchronous as were the two 1B. We obtained similar results with cells that had been grown in nocodazole for 22 h and pulse-labeled with BrdUrd (not shown). This indicates that, as expected, synchrony was preferentially retained for pairs of sister chromosomes. We conclude that the close spatial positioning of sister chromosomes in the nuclei of cells that undergo mitotic slippage directly or indirectly maintains the synchrony of their replication onset. This provides further evidence of a direct link between nuclear organization and DNA replication timing.

Chromosomes Complete Two Rounds of Replication in Nocodazole-treated Cells-- As mentioned above, cells that escaped mitotic arrest and underwent re-replication were frequently found to be hypo-tetraploid rather than tetraploid when chromosomes were counted at the first mitosis after mitotic slippage. This could result either from chromosome loss or from non-duplication of some chromosomes. Indeed, the absence of a functional mitotic spindle might occasionally preserve a strict cohesion between the two sister chromatids, which in turn may prevent DNA replication. To determine whether some chromosomes escape replication during the S phase that follows mitotic slippage, cells were grown in medium supplemented with BrdUrd for 24 h, about twice the doubling time of GMA32 cells. Nocodazole was added 2 h after BrdUrd addition, and chromosome spreads were prepared 22 h later. As a control, cells were grown for 24 h in the presence of BrdUrd without nocodazole. Upon two rounds of replication in the presence of BrdUrd, each chromosome was expected to have one chromatid in which thymidine is replaced by BrdUrd on both DNA strands, the other one being substituted on a single strand. After fluorescence plus Giemsa staining, the chromatid with a double-strand substitution should appear light, whereas the hemi-substituted chromatid should be dark (28). Consistently, in all metaphase plates from control cells we observed chromosomes with differentially stained chromatids (Fig. 6a), and sister chromatid exchanges were frequent, as expected in cycling cells (40). We analyzed 112 metaphase plates of nocodazole-treated cells, all of them displayed patterns indistinguishable from those of control cells (Fig. 6b). This result indicates that at least in cells that reach mitosis after mitotic slippage, all chromosomes have completed two rounds of replication during the treatment with nocodazole and, thus, suggests that chromosome loss is responsible for aneuploidy.

View larger version (66K): [in this window] [in a new window]   Fig. 6.   In nocodazole-treated cells that have reached mitosis after adaptation, all chromosomes are fully re-replicated. Cells were grown in the presence BrdUrd for 24 h with or without nocodazole for the last 22 h, and chromosomes were differentially stained by the fluorescence plus Giemsa technique. Shown are metaphase plates from GMA32 control cells (a) and from GMA32 cells that were grown in nocodazole for 22 h and engaged re-replication (b). Note that sister chromatid exchanges are evident, as expected in normal cycling cells. Bar, 5 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The establishment of the DNA replication timing is a complex phenomenon at least partially correlated to the positioning of the chromosomal domains in the nuclei of mammalian cells (13). Here we have studied Chinese hamster cells in which the nuclear organization has been perturbed by treatment with nocodazole, a well known spindle poison. This category of drugs impairs the progression of cells through mitosis, but mammalian cells gradually overcome the block and enter G₁, a process often accompanied by nuclear distortion or fragmentation. Although normal cells fail to enter S phase after mitotic slippage, most transformed and immortalized cells, including cells of the GMA32 line, do undergo DNA replication (Fig. 1). Analysis of the tetraploid or slightly hypo-tetraploid metaphase plates resulting from re-replication shows that during this S phase all chromosomes of cells that reach mitosis are completely re-replicated (Fig. 6), but the patterns of replication-dependent chromosome banding observed in nocodazole-treated cells indicate that various subsets of chromosomes replicate asynchronously (Fig. 2). Moreover, the observed asynchrony results from an uncoupled onset of replication on different chromosomes. Indeed, in some experiments, nocodazole-treated cells were pulse-labeled with BrdUrd during a time lapse covering the G₁/S border after mitotic slippage, and a few completely unlabeled chromosomes were observed in metaphase plates (Fig. 2d). However, we cannot rule out the possibility that some variations in the replication rate accompany the chromosome asynchrony. In contrast, the temporal order of replication of syntenic sequences is not affected, since no abnormalities were detected when the banding patterns of chromosomes from nocodazole-treated cells were compared with those of untreated control cells. Significantly, a normal temporal program of replication is maintained also in the chromosomes that replicate asynchronously (Fig. 5).

The existence of structural relationships between nuclear organization and the replication process has been often described (9, 10). In our work, FISH experiments focusing on the AMPD2 and P3C3 loci indicated that the four copies of chromosome 1 distribute differently in the interphase nuclei of cells of a long established tetraploid line and of diploid cells treated with nocodazole and allowed to undergo a second S phase. Only in this latter case, the homologues tend to cluster two by two (Fig. 3, b-f), and the replication of the AMPD2 locus was frequently altered in pairs of clustered chromosomes (Fig. 3, d and e). Analysis of the replication-dependent banding of pairs of asynchronous 1A and 1B chromosomes revealed that sister chromosomes shared the very same replication pattern, whereas non-sister homologues are often replicated asynchronously (Fig. 5B, b and c). Consistently, tetraploid cells contain a single pair of sister chromosomes X that replicate synchronously (Fig. 5B, a-c). Hence, it is tempting to assume that sister chromatids, from which the sister chromosomes derive, remained close to each other in interphase, giving rise to the observed clustering of homologues two by two. Such an absence of sister chromatid disjunction is easily accounted for by the nocodazole-induced disruption of the mitotic spindle (41). These observations indicate that replication synchrony is preferentially preserved on chromosomes that are maintained in close proximity and suggest that nuclear organization plays an essential role to ensure a coordinated onset of replication on the different chromosomes. Asynchrony of chromosome replication was observed nearly four decades ago by Stubblefield (15) after prolonged treatment with colcemid of Chinese hamster cells labeled with tritiated thymidine. He suggested that this phenomenon results from the confinement of different chromosomes in different micronuclei (15, 42). More recently, in vitro analysis of the replication timing of pseudo-nuclei formed in Xenopus oocyte extracts upon reconstitution of a nuclear envelope around demembranated nuclei supported the confinement hypothesis (43). In this experimental system, different pseudo-nuclei replicate asynchronously, whereas individual demembranated nuclei embedded within the same nuclear envelope replicate synchronously. Asynchronous replication of chromosomes segregated in different micronuclei was also observed during Xenopus early development (44). On the contrary, the physical separation of two sets of chromosomes within two nuclei in mammalian heterokaryons does not correlate per se with an asynchrony of replication or of other important cell cycle events (45). Here we show that asynchronous replication is not limited to the chromosomes that are physically isolated into separate nuclear compartments in somatic mammalian cells. Indeed, FISH experiments indicated that this phenomenon occurs as well in cells that exhibit an apparently normal or only slightly abnormally shaped nucleus (Fig. 3). Careful analysis of normally shaped nuclei of nocodazole-treated cells reveals that they are devoid of cryptic internal compartmentalization (Fig. 4).

It was repeatedly suggested that the role of the nuclear envelope with respect to the control of the DNA replication process is to modulate the concentration of positive or negative regulatory proteins involved in the onset or the progression of S phase (46-48). In the experimental system based on Xenopus oocytes described above (43), synchrony of replication within each pseudo-nucleus might be favored by the high concentration of replication factors in the oocyte extract, which in turn could allow an efficient and homogenous distribution of positive regulators in the pseudo-nuclei. Yet, in mammalian nuclei replication factors are far less concentrated, and the generation of gradients throughout the nuclear volume might explain the asynchrony observed here. In this hypothesis, initiation and/or subsequent elongation at early origins on a given chromosome can be blocked until enough factors become available to allow all early origins in cis to fire simultaneously. Indeed a block at the onset of the replication was correlated with an abnormal distribution of elongation factors in cells with an altered lamin organization (49). We failed to detect strong abnormalities in the general distribution of lamins in nocodazole-treated cells (not shown); however, our analysis is probably insufficient to definitely rule out such a possibility. On the other hand, specialized sub-nuclear compartments with different concentrations of transcription and replication factors are present in the nuclei of untreated cells (50, 51). The existence of preferred localization of chromosomal sub-domains and whole chromosomes within the nucleus (52, 53) may be correlated with the uneven distribution of such factors, and abnormal chromosome positioning might interfere with their proper replication. Indeed, chromosome positioning in interphase seems to require the passage through the S and M phases (14). Thus, the position of each chromosome in interphase nuclei may be, at least in part, determined during metaphase and anaphase of the previous cell cycle.⁴ In the work presented here, chromosome positioning was not studied, but we observed that chromosome distribution is perturbed upon disruption of the mitotic spindle and mitotic slippage. In support to this hypothesis, experiments in yeast correlated nocodazole treatment to a facilitated chromatin diffusion in interphase, which in turn leads to an expanded confinement region of chromosomes (54). Finally, we cannot exclude that the nocodazole-mediated mitotic block could also alter the chromosome structure, for example their condensation, and consequently perturb their metabolism.

Although our analysis was restricted to the mid- to late S-phase banding patterns of chromosomes 1 and X, our observations indicate that changes in nuclear organization induced by nocodazole treatment do not affect the timing of replication of syntenic sequences within a chromosome. Such results are consistent with the stability of the replication patterns of chromosomal sub-domains observed in cultured human cells undergoing chromosome repositioning (14) and with the similar distribution of early and late replication domains observed in non-fragmented and fragmented nuclei of hamster cells (2). Our results are also in good agreement with experiments showing that when S-phase onset is delayed in yeast cells, the temporal program of replication of chromosome sub-domains is maintained (55). The sequential replication program of chromosome sub-domains may be essentially dictated by the chromatin status. For example, in yeast cells, the late replication program, established in the vicinity of telomeres, depends on the silencing chromatin component Sir3 (56), and studies of the human - and -globin loci have revealed a correlation between early replication and an open chromatin conformation (57, 58).

In conclusion, our results disclose the existence of two independent levels of control on replication timing of mammalian chromosomes; one ensures the synchrony of the onset of replication on the whole chromosome population of a nuclear compartment, and the second establishes the sequential replication timing of sub-regions along each chromosome. This also indicates that the completion of the replication program along a chromosome is independent of the replication status of the chromosome population as a whole.

	ACKNOWLEDGEMENTS

We thank G. Buttin, M. C. Weiss, M. Mechali, and B. Dutrillaux for helpful discussions and critical reading of the manuscript.

	FOOTNOTES

^* This work was supported in part by the Université Pierre et Marie Curie, the Ligue Nationale Française contre le Cancer (Comité de Paris), and the Association pour la Recherche sur le Cancer.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by grants from the Association pour la Recherche sur le Cancer and the Fondation pour la Recherche Médicale.

^§ To whom correspondence should be addressed. Tel.: 33 (0)1 42346725; Fax: 33 (0)1 42346674; E-mail: Michelle.Debatisse@curie.fr.

Published, JBC Papers in Press, July 12, 2001, DOI 10.1074/jbc.M104501200

² M. Debatisse, B. Labidi, and P. Metezeau, unpublished data.

³ M. Debatisse, unpublished results.

⁴ E. Manders and R. Van Driel, personal communication.

	ABBREVIATIONS

The abbreviations used are: FACS, fluorescence-activated cell sorter; BrdUrd, 5-bromodeoxyuridine; PBS, phosphate-buffered saline; FISH, fluorescence in situ hybridization; DiOC₆, 3,3'-dihexyloxacarbocyanine iodide.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Kar, D. Ghosh, and J. Sengupta Histochemical and morphological examination of proliferation and apoptosis in human first trimester villous trophoblast Hum. Reprod., November 1, 2007; 22(11): 2814 - 2823. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/39/36639    most recent M104501200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Anglana, M. Articles by Debatisse, M. Search for Related Content PubMed PubMed Citation Articles by Anglana, M. Articles by Debatisse, M. Social Bookmarking                      What's this?