1 Division of Genetic Resources, National Institute of Infectious Diseases Japan, 1-23-1, Toyama, Shin-juku-ku, Tokyo, 162-8640, Japan
2 Department of Radiology, Jikei University of School of Medicine, 3-25-8, Nishi-Simbashi, Minato-ku, Tokyo, 116, Japan
* Corresponding Address: Division of Genetic Resources, National Institute of Infectious Diseases Japan, 1-23-1, Toyama, Shin-juku-ku, Tokyo, 162-8640, Japan
Key words: chromosome condensation/compaction, chromosome structure, DNA replication, cell cycle, mitosis, S-phase, premature chromosome condensation (PCC), prematurely condensed chromosomes (PCCs), calyculin A, beads loading method
A most miracle mysterious and profound event in eukaryote cell is how DNA folds to chromosomes. In human diploid cell (2n), for example, the total DNA (~6x109 nucleotide base pairs, a meter of length when fully relaxed) is packed to 46 chromosomes (22 pairs of autosomes and 1 pair of sex chromosomes) and contained in nuclei size of ~5 μm in diameter (Alberts et al., 1989). It is quite difficult to imagine how such long length thin fibrous linear molecule is folded in small sized chromosomes without entangling in a narrow nucleus space. Very earlier, the concept about the chromosome architecture formation during cell cycling was conceived as follows: (1) chromosomes are diffused over nucleus as decondensed form in G1-phase (Gap 1 phase), (2) DNA synthesis starts and chromosome replicates in S-phase (Synthesis of DNA phase), (3) DNA synthesis finished, the resulted chromosomes are duplicated and ready for cell division in G2-phase (Gap 2 phase) and then (4) chromosomes condense: separation/segregation and cell division occurs in M-phase (Mitotic phase). This traditional concept seems to tell that DNA replication and chromosome condensation are independent events that proceed in S- and M- cell cycle stage, respectively.
Recently, number of accumulated evidences suggests a close relationship between DNA replication and chromosome condensation. Premature chromosome condensation (PCC) technique was introduced in the 1970's as a useful technique that allows the interphase nuclei to be visualized as condensed mitotic chromosome (Johnson and Rao, 1970; Johnson et al., 1970; Sperling and Rao, 1974). Since then, a lot of studies including DNA replication and chromosome packaging have been archived using the PCC method (Hittelman and Rao, 1976; Rao et al., 1977; Hanks and Rao, 1980; Mullinger and Johnson, 1980; Lau and Arrighi, 1981; Mullinger and Johnson, 1983). These studies seem to teach that the different DNA packaging appearance in different sub-phase of S-phase suggest that the degree of chromosome condensation might be tightly coupled with the progressing of DNA replication. However, the limited available methodologies at that time did not allow the precise mechanism to be cleared.
More recently, accumulated evidences have further concrete that eukaryote DNA replication/transcription is involved in compaction of chromosomes (Zink et al., 1998; Manders et al., 1999; Samaniego et al., 2002, Pflumm, 2002). Molecular genetic studies have also provided supporting evidence for the idea that mutation (in genes as HIRA/Tuple1, XCDT1, cdt1, Orc2, Orc3, Orc5, MCM2, MCM4, MCM10, RECQL4, required for DNA replication) showed abnormal phenotype in chromosome condensation (Loupart et al., 2000; Maiorano et al., 2000; Nishitani et al., 2000; Pflumm and Botchan, 2001; Christensen and Tye, 2003; McHugh and Heck, 2003; Prasanth et al., 2004), inherited diseases (D'Antoni et al., 2004; Sangrithi et al., 2005), genomic instability or prone to cancer (Tatsumi et al., 2006; Pruitt et al., 2007; Shima et al., 2007), or aberrant replication timing causes abnormal chromosome condensation (Loupart et al., 2000; Marheineke and Hyrien, 2001; Pliss et al., 2009). For more detailed knowledge about the DNA replication, see also the following excellent reviews; Bell and Dutta, 2002 and Masai et al., 2010.
In the present article, using drug-induced PCC technique and direct Cy3-dUTP fluorescent replicating DNA by beads loading method, we demonstrate the dynamics of chromosome structure, formation and transition during the S-phase progression in which tight-coupled relation between DNA replication and chromosome condensation /compaction. Possible hypothetical chromosome condensation/compaction model involving the role of DNA replication will be suggested.
Under a quite stringent and higher ordered mechanism, chromosomes condense during mitosis within a very short lapse of time in the mitotic phase. Mitotic phase is further divided into several subphases (preprophase, prophase, prometaphase, metaphase, anaphase and telophase), followed by cytokinesis. In the course of Mitotic phase, number of sequential drastic conformational transactions are proceeded as following: (1) chromatins condense to well-defined visible chromosomes under the microscope (2) mitotic spindle assemble begin, nuclear envelope breakdown into membrane vesicles, centriole and mitotic spindle formation followed by spindle attaches to chromosome centromeres (3) kinetocore microtubules align the chromosomes at metaphase plate (4) chromosome separation segregates to spindle poles (5) separated daughter chromatids reach the poles followed by the nuclear envelope re-forms (6) formation of contractile ring and cleavage furrows which constrict the cell center, cytokinesis, cell dividing into two daughter cells, chromosome decondensation in divided cells and finally re-entering the cells in the G1 phase (Alberts et al., 1989). The detailed of the whole mechanism is still almost unclear. However, number of molecules which involved in the mitotic events have been identified such as SMC proteins, including condensin (chromosome condensation), cohesion (chromosome cohesion of replicated chromosomes) (Swedlow and Hirano, 2003), NuMA protein for spindle pole formation (Chang et al., 2009; Haren et al., 2009; Silk et al., 2009; Torres et al., 2010), nuclear lamins (Moir et al., 2000), aurora kinases in centromere function (Tanno et al., 2006; Meyer et al., 2010; Tanno et al., 2010), shugoshin and protein protein phosphatase 2A in chromosome cohesion (Kitajima et al., 2006; Tanno et al., 2010), cdk1 in chromosome condensation, chromosome bi-orientation (Tsukahara et al., 2010), cyclin B, cdc2, cdc25 in chromosome condensation (Masui, 1974; Draetta and Beach, 1988; Dunphy et al., 1988; Kumagai and Dunphy, 1992), Polo and Rho in cytokinesis (Burkard et al., 2009; Wolfe et al., 2009; Li et al., 2010) and many other proteins. Chromosome dynamic consists of such number of various elements. Regarding these dynamics, numerous visualizing studies reported or in progression are achieved through the mitosis events, relatively easy to observe under microscope. However, visualizing approaches in chromosome dynamics, coupled with DNA replication, is still limited. This restraint is due to difficulties as to observe the chromosomes in the S-phase: chromosomes are usually invisible at this stage, since they are decondensed. In the present review, we simply focus on visualizing the chromosome dynamics coupled with DNA replication during the S-phase progression and we show how replicating DNA is folded into higher order chromosomes.
Tools to visualize the dynamic chromosomes
Cytogenetic analysis studies are usually performed on chromosomes. As condensed in mitosis, chromosomes are usually visible, but as they decondensed in the interphase, they are invisible (Manders et al., 1996; Gotoh and Durante, 2006). Therefore, it is practically difficult or even impossible to analyze the dynamics of chromosome condensation during the interphase by conventional chromosome methods such as colcemid block. Premature chromosome condensation (PCC) is a useful and a unique technique that allows the interphase nuclei to be visualized as a condensed form of mitotic chromosome (Johnson and Rao, 1970; Rao and Johnson, 1970). Conventional PCC has been carried out by cell fusion using either fusogenic viruses (i.e. Sendai virus) (Johnson and Rao, 1970) or polyethylene glycol (PEG) (Pantelias and Maillie, 1983) (cell fusion-mediated PCC). But these protocols are usually technically demanding and keenly depend on the activity of the virus or PEG. Virus-mediated PCC might be also problematic because of infectious viruses use. Moreover, resulting chromosomes are mixture of those inducer and recipient cells (Gotoh and Durante, 2006). Due to these restrictions, conventional PCC has been used in limited institutions. These drawbacks of the conventional PCC technique have been recently overcome with a much easier and more rapid technique using calyculin A or okadaic acid, specific inhibitors of protein phosphatases (drug-induced PCC technique) (Gotoh et al., 1995; Gotoh and Asakawa, 1996; Asakawa and Gotoh, 1997; Durante et al., 1998; Gotoh and Durante, 2006). Drug-induced PCC is becoming now 'popular' and has been used in a wide range of cytogenetic applications (Gotoh and Asakawa, 1996; Asakawa and Gotoh, 1997; Gotoh et al., 1999; Ito et al., 2002; Terzoudi et al., 2003; El Achkar et al., 2005; Gotoh and Tanno, 2005; Gotoh et al., 2005; Srebniak et al., 2005; Terzoudi et al., 2005; Deckbar et al., 2007; Gotoh, 2007; Beucher et al., 2009; van Harn et al., 2010). Thus, drug-induced PCC technique is suitable to visualize dynamic chromosomes particularly in interphase nuclei. This technique will be also useful and applicable in many fields of cytogenetic approaches including traditional chromosome analysis study, because the technique is very simple and much easier even than the conventional colcemid blocking method (Gotoh, 2009).
Cytogenetically visualization of the replicating DNA is certainly a most direct approach to identify the DNA replication dynamics. In the very earlier studies, the fibre autoradiography of DNA had been labeled with 3H-thymidine (Fakan and Hancock, 1974; Edenberg and Huberman, 1975; Hand, 1978). The spatial resolution of fibre autoradiography is, however, limited because the location of the silver grains, developed in photosensitive emulsion layer and covered the specimens, do not correctly reflect the actual regions of the foci incorporating the 3H-thymidine and the size of grains are not enough tiny to determine the precise location of replicating regions. More precise localization and measure the replication foci were then done using thymidine analog BrdU (Bromodeoxy Uridine) labeling and its antibodies (Nakamura et al., 1986; Mills et al., 1989; Nakayasu and Berezney, 1989). However, the resolution is still limited presumably because it is based on accessibility problems or size of immunocomplex (antigen/antibodies). Recently, the replication regions and chromosome formation in living cells were visualized using Cy5-dUTP directly labeled fluorescent DNA (Manders et al., 1999) by beads loading methods (McNeil and Warder, 1987). The procedure facilitate the analogues (Cy3-dUTP or Cy5-dUTP) to be incorporated in the cell nucleus in a very short time whereby transiently permeabilizes the cell membranes. This method allows the replicating DNA to be Cy3 fluorescently imaged within very short lapse of time. The obtaining fluorescence signal reflects the real incorporated site of analogue replicating DNA with a very fine signal resolution. Combined with the beads loading method and drug-induced PCC, dynamic study of chromosome condensation, involving DNA replication, has been realized (Gotoh, 2007).
Chemicals and Instruments
Visualize the dynamics of chromosome structure formation coupled with DNA replication during S-phase
Many studies, for visualizing the dynamics of chromosome condensation during cell division in mitosis, have been achieved and well documented. However, the visualizing study on the relationship between chromosome condensation and DNA replication is still limited. Several studies tried to define fairly well the replication foci distribution in interphase nuclei (Nakamura et al., 1986), but little is yet known about how replicating DNA is folded to higher order chromosomes (since chromosomes are invisible in interphase stage as they decondensed).
To visualize the chromosome compaction dynamics coupled with DNA replication, more precisely in S-phase nucleus, the drug-induced PCC method was used (Gotoh et al., 1995; Asakawa and Gotoh, 1997; Johnson et al., 1999; Ito et al., 2002). The cells were unsynchronized because cell synchronization using DNA synthesis inhibitor such as thymidine may give some bias in DNA replication and consequently all phases of replication can be observed. Individual substage of S-phase can be easily identified by typical diagnostic appearances seen in different phases of S-PCCs (Mullinger and Johnson, 1983; Gollin et al., 1984; Hameister and Sperling, 1984; Savage et al., 1984; Gotoh et al., 1995; Gotoh and Durante, 2006). A drastic conformational change of chromosome structure formation along with the proceed of DNA replication, as shown in Fig. 1 (reproduced from Chromosoma. Gotoh, 2007; 116(5):453-462), is clearly revealed in PCCs following Cy3-dUTP loading. Cy3-dUTP loading procedure takes 10 minutes followed by 10 minutes of PCC induction and fixation (for details, see Materials and Methods in Chromosoma. Gotoh, 2007; 116(5):453-462). Accordingly, only replicated DNA in this short lapse will be fluoresced. Thus, the observed S-PCCs in the present study reflected the replication stages at most 20 minutes before the cell fixation. (i) In early S-phase, PCCs showed a cloudy spreading mass of thin fibres like a 'nebula', where numerous fine granular foci homogeneously distributed on overall the fibres (Fig. 1I), showing 'beads on a string' or 'particles on a string': these structures are observed under an electron microscope (Olins and Olins, 1974; Thoma et al., 1979). (ii) In the middle of S-phase, typical 'pulverized' PCCs were recognized; the size of foci was increased while the number of foci, unevenly distributed on chromosomes, was decreased. As shown in Fig. 1J, the foci become brighter. (iii) In the late S-phase, chromosomes were mostly condensed like mitotic chromosomes. Cy3-dUTP incorporated regions were recognized as band arrays inserted in the condensed chromosome (Fig. 1K, indicated by arrows). The similar appearance of replication foci along longitudinally on chromosomes were previously reported on metaphase of kangaroo-rat kidney PtK1 cells (Ma et al., 1998). The size of foci is still up and their number is still down to the point that they could be easily scored. (iv) In the very late S-phase, the number of foci is further reduced and predominantly they are localized at centromeric or telomeric regions (Fig. 1L, indicated by arrows). These regions are actually known as satellite heterochromatic DNA regions where DNA replicates at very late S (O'Keefe et al., 1992).
Figure 1: (1) DNA replication regions on prematurely condensed chromosomes (PCCs) of different substages of S-phase. Ten minutes after Cy3-dUTP loading, cells were condensed prematurely using 50 nM of calyculin A (Gotoh et al., 1995). From left to right column, (A,B,C) early S-phase PCCs, (D,E,F) middle S-PCCs, (G,H,I) late S-PCCs and (J,K,L) very late S-PCCs. (A,D,G,J) DAPI counterstained DNA, (B,E,H,K) Cy3-dUTP labelled DNA replication region and (C,F,I,L) Merged image of DAPI and Cy3. (L) Centromeric region (arrow) or telomeric region (arrowhead) replicates in very-late S-phase are indicated. (I,L) Late S- and very late S-PCCs already condensed like as mitotic chromosomes, but these PCCs were actually S-phase chromosomes because they incorporated Cy3-dUTP. G2/M chromosomes are easily distinguished from late or very late S chromosomes as G2/M chromosomes do not incorporate Cy3-dUTP (data not shown). Inset in (C) is higher magnification of the boxed portion. Scale bar, 10 μm. (2) DNA replication regions seen on prominent fibre of PCCs. (M) early-S-phase and (N) middle S-phase. Replication foci are clearly seen as 'beads on a string' structure, some of these are indicated by arrowhead. Scale bar, 10 μm. (Figure reproduced from Figure 2 of Chromosoma 2007; 116(5):453-462. By Gotoh).
In the present review, the dynamics of chromosomal conformation change, which is tightly coupled with DNA replication during S-phase, was clearly seen on PCCs of different sub S-phase using drug-induced PCC method (Gotoh and Durante, 2006) and on Cy3-dUTP direct labeling method (McNeil and Warder, 1987). Drug-induced PCC would be, therefore, a useful tool that provides new insights of the dynamics of chromosome formation and DNA replication.
As described in the previous section, number of accumulated evidences suggested the role of DNA replication in chromosome condensation/compaction (Pflumm, 2002). As previously reported, evidence and results of this study show that: (i) The different appearance of condensation in different sub-phase of S-PCCs is thought to be depended on the different degrees of chromosome conformation at the time of PCC induction (Johnson and Rao, 1970; Rao, 1977; Rao et al., 1977). In the late or very late S phase, particularly, chromosome conformation already changes like mitotic chromosomes (Fig. 1F). (ii) Chromosomes condense asynchronous and the different degree of condensation depend on the time of chromatin replication (Kuroiwa, 1971). (iii) Chromosomes are not fully diffused nor nonrandomely positioned in the nucleus, but are separately compartmentalized in interphase nuclei (Cremer et al., 1993; Ferreira et al., 1997; Berezney et al., 2000). These chromosomes, occupying the 'territory', do not intermingle (Hadlaczky et al., 1986; Cremer et al., 1993; Swedlow and Hirano, 2003; Cremer et al., 2006; Heard and Bickmore, 2007). (iv) Late replication foci were prealigned during interphase. They moved subtly to generate recognizable chromosomes presumably due to shortening of the longitudinal chromosome axis (Manders et al., 1999). (v) The gross structure of an interphase chromosome territories is directly related to that of the prophase chromosomes (Manders et al., 1999). (vi) The structure of mitotic chromosomes and the nuclear chromosome territories are closely related (Manders et al., 1999) and the different bands of mitotic chromosomes are presented as distinct domains regarded subchromosomal foci within chromosome territories (Zink et al., 1999). (vii) During the cell cycling, the global chromosome territories are conserved. Although some conflicts still remains, several studies reported that chromosome territories are transmitted through mitosis (Manders et al., 1999; Gerlich et al., 2003; Gerlich and Ellenberg, 2003) whereas others reported that positional relations of chromosome territories are lost either at mitosis (Walter et al., 2003) or at early G1 (Essers et al., 2005). (viii) The spatio-temporal organization of DNA replication is determined by the specific nuclear order of these stable chromosomal units (Sadoni et al., 2004). (ix) Chromatin domains with the dimension of replication foci may be fundamental units of chromosomal architecture (Berezney et al., 2000). (x) DNA replication occurs at fixed sites and replicated DNA move through replication center (Berezney and Coffey, 1975; Pardoll et al., 1980; Hozak et al., 1993). (xi) DNA replication contributes to a longitudinal contraction of the chromosome axis (Hearst et al., 1998). (xii) Functional replication origins are a critical requirement for longitudinal condensation of the chromosome axis (Pflumm and Botchan, 2001). The results presented in this review and previous findings strongly suggest that DNA replication, nuclear organization and chromosome condensation are mutually integrated to construct a higher ordered structure of eukaryote chromosomes.
Number of models for eukaryote chromosome architecture have been proposed (Marsden and Laemmli, 1979; Woodcock et al., 1984; Woodcock and Dimitrov, 2001; Swedlow and Hirano, 2003; Kireeva et al., 2004), but they are controversial and many aspects are still unclear. In addition, these models do not take account of the involvement of DNA replication/transcription in chromosome packaging. DNA/RNA polymerase are known to be tightly immobilized to the replication/transcription factories (Cook, 1999; Frouin et al., 2003). In the proposed model, DNA polymerase is thought to be a 'reel in DNA template and extrude replicated DNA' (Hozak et al., 1996; Cook, 1999) rather than an enzyme track along DNA template, which is proposed in many conventional models. In the context of Cook's model, some kinds of mechanical tension force should be generated in the DNA template along with DNA replication goes on because the factory is not freely suspended in the nucleus but attached to nucleoskeleton. Consequently, this force may pull and aggregate the replication foci of both sides as to release the tension in DNA strands, which may result in the formation of the chromosomes as seen in mitosis. Based on the above mechanism and the observed findings obtained from chromosome structure dynamics coupled with DNA replication, Fig. 2 shows a hypothetical model for the relationship between DNA replication and chromosomal conformation changes, and it shows too how the interphase chromatin is constructed into chromosomes (Figure reproduced from Chromosoma. Gotoh, 2007; 116(5): 453-462). During the S-phase, chromosomal conformation changes and the chromosome formation would be mostly completed at the end of DNA replication (Fig. 2A,B,C). From G2 to prophase, chromosomes are still more elastic, less condensed, folded only several times and prealigned in interphase nuclei (Manders et al., 1999). At these phases, the chromosomes would be observed as chromosome territories (Cremer et al., 1993; Berezney et al., 2000) (Fig. 2D). Entering in mitosis, these chromosomes would condense even more as shortening the longitudinal axis to form solid and rod shape appearance of recognizable mitotic chromosomes (Manders et al., 1999) (Fig. 2E).
Figure 2: A hypothetical two-dimensional model for chromosome conformational change involving DNA replication based on the models proposed by Cook (Cook, 1995) or Pflumm (Pflumm, 2002). (A) Early S-phase. DNA replication starts at multiple origins and proceeds bi-directionally. Early S-PCCs are seen as 'beads on a string' appearance. (B) Middle S-phase. As DNA replication proceeds, replicated DNA pass through replication factory and some tension are generated. The generated tension may pull back the replication factories close together so as to release the tension. Replication factories may in turn fuse together and chromosomes compact. Middle S-PCCs are seen as well known 'pulverized chromosomes' appearance. (C) Late S-phase. Most of DNA finished replication and conformation was changed. Late S-PCCs are seen as 'tandem band arrayed structured chromosomes' like as mitotic chromosomes. (D) G2 to prophase. After finishing of DNA replication, chromosome conformation changed like as mitotic chromosomes, but still so elastic that packed in nucleus. Before fixation, each chromosome occupies individual chromosome territory (CT) in interphase nucleus, thus observed as compartment regions (colorized). (E) Mitosis. After prophase, chromosomes further shortening in longitudinal axis of chromosomes, consequently a straight rod shaped recognizable chromosome formed as usually seen by cytologists under a microscope. For simplicity, the model is shown as two-dimensional and the scaling is arbitrary. The model intends not to depict actual events of chromosome conformation change but to help imagine how DNA replication is involved in chromosomal conformation. As the real chromosomes condense as three-dimensionally, other elements such as coiling and helical winding should be considered together to construct a stereoscopic hierarchical structure of eukaryote chromosomes (Woodcock and Dimitrov, 2001; Swedlow and Hirano, 2003). (Figure reproduced from Figure 3 of Chromosoma 2007; 116(5):453-462. By Gotoh).
A basic and principle question of cell biology is: how DNA folds to chromosome? Numbers of evidence have suggested the involvement of DNA replication in chromosome structure formation. To visualize the dynamics of chromosome structure formation coupled with DNA replication, Cy3-dUTP direct-labeled active replicating DNA was observed in prematurely condensed chromosomes (PCCs) utilized with drug-induced premature chromosome condensation technique, which facilitates the visualization of interphase chromatin as well as the condensed chromosome form. S-phase PCCs revealed clearly the drastic dynamic transition of chromosome formation during S-phase along with the progress of DNA replication: from a 'cloudy nebula' structure in early S-phase to numerous number of 'beads on a string' in middle S-phase and finally to 'striped arrays of banding structured chromosome' in the late S-phase as usual observed in mitotic chromosomes. The drug-induced PCC is clearly provided a new insight that the eukaryote DNA replication is tightly coupled with the dynamics of chromosome condensation/compaction for the construction of eukaryote higher ordered chromosome structure. Based on these findings, a hypothetical model for chromosome compaction involved the role of DNA replication is proposed. In this model, conformational change is simply illustrated as two-dimensional but the real architecture is a three-dimensionally chromosome constructs, with much more complex fashion. It is mostly unclear how DNA replication/transcription conducts to make up a three-dimensional hierarchical structure of chromosomes coupled with twisting/folding/winding or other factors. It should be a most principle challenge in cell science.