• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
Cell. Author manuscript; available in PMC Jan 12, 2011.
Published in final edited form as:
PMCID: PMC3019746

Live-cell imaging reveals replication of individual replicons in eukaryotic replication factories


Faithful DNA replication ensures genetic integrity in eukaryotic cells, but it is still obscure how replication is organized in space and time within the nucleus. Using time-lapse microscopy, we have developed a new assay to analyse the dynamics of DNA replication both spatially and temporally in individual Saccharomyces cerevisiae cells. This allowed us to visualize replication factories, nuclear foci consisting of replication proteins where the bulk of DNA synthesis occurs. We show that the formation of replication factories is a consequence of DNA replication itself. Our analyses of replication at specific DNA sequences support a long-standing hypothesis that sister replication forks generated from the same origin stay associated with each other within a replication factory whilst the entire replicon is replicated. This assay system allows replication to be studied at extremely high temporal resolution in individual cells, thereby opening a window onto how replication dynamics vary from cell to cell.


Faithful DNA replication is essential for all prokaryotic and eukaryotic cells to maintain their genetic integrity. DNA replication is initiated at replication origins and proceeds as sister forks from the same origin move along parental DNA in a bidirectional manner. DNA polymerases involved in replicating both leading and lagging strands, together with their accessory proteins such as replication factor C (RFC) and proliferating cell nuclear antigen (PCNA), are thought to form a large complex (called the ‘replisome’) that moves along with each replication fork (Baker and Bell, 1998; Waga and Stillman, 1998; Johnson and O’Donnell, 2005).

It was originally thought that the two replisomes at sister forks (i.e. initiated from the same origin) would behave independently since they travel in opposite directions along parental DNA. However, it was found that on bacterial circular chromosomes where DNA replication starts from a single defined origin, sister forks move along DNA and normally complete DNA replication with similar timing at a defined region on the chromosome (studied in Eschericia coli, Bacillus subtilis etc: Bussiere and Bastia, 1999). To explain this coordinated termination of DNA replication, it was proposed that two replisomes at sister forks (sister replisomes) remain attached during DNA replication (Dingman, 1974; Falaschi, 2000). This model predicts that template DNA moves into two associated replisomes and newly-replicated sister DNA strands are extruded as replication proceeds. Such DNA motion relative to centrally-located stationary replisomes (Lemon and Grossman, 1998) was indeed recently confirmed in bacteria (Lemon and Grossman, 2000; Jensen et al., 2001; Lau et al., 2003; Migocki et al., 2004).

In simian virus 40, the large tumour antigen (T-antigen) forms a hexamer that works as a DNA helicase at replication forks (Herendeen and Kelly, 1996). An electron microscopy study revealed that unwound DNA from the viral replication origin forms two single-strand loops, both of which were pinched by the same pair of associated T-antigen hexamers (Wessel et al., 1992). This is also consistent with the model of associated sister replisomes.

In contrast to viruses and bacteria, DNA replication in eukaryotes initiates at multiple replication origins along linear chromosomes. The DNA region replicated by sister forks from a single origin is called a replicon. In contrast to bacterial circular chromosomes, sister forks terminate at two different loci along eukaryotic linear chromosomes, meeting the replication forks of adjacent replicons. In such circumstances, two sister replisomes may operate independently of each other. So, it is sill unclear whether the two sister replisomes are associated with each other in eukaryotic cells.

How is DNA replication spatially organised in the eukaryotic cell nucleus? During DNA replication in vertebrate cells, replisome components such as PCNA and DNA polymerases assemble into dozens of globular foci called replication factories in the nucleus, and it has been shown that new DNA replication takes place within these replication factories (Nakamura et al., 1986; Hozak et al., 1993; Newport and Yan, 1996; Berezney et al., 2000; Frouin et al., 2003). To account for the number of DNA replication forks generated during S phase, a single replication factory must contain 20-200 DNA replication forks (Berezney et al., 2000). A simple model is that two sister replisomes localize inside the same factory during replication. However, if this is the case, it is still unclear whether sister replisomes are closely associated with each other or stay at a distance within a replication factory, which may have a diameter of up to 1 μm in vertebrate cells (Leonhardt et al., 2000; Somanathan et al., 2001; Sporbert et al., 2002). Moreover, recent data indicate that replication factories continuously assemble and disassemble during S phase (Leonhardt et al., 2000; Somanathan et al., 2001; Sporbert et al., 2002); thus sister replisomes may sometimes, if not always, be redistributed between different replication factories.

Here we describe dynamics of both replication factories and individual replicons in budding yeast S. cerevisiae, using time-lapse microscopy. Replication factories are much smaller in yeast cells than in vertebrate cells, and show dynamic behaviour during S phase. Crucially, we show that the formation of replication factories is a consequence of DNA replication itself. We have established an assay system to analyse DNA replication of chromosomal loci both spatially and temporally in individual live cells. Using this system, we find evidence that while a replicon is in the process of DNA replication, sister replication forks remain closely associated with each other within a replication factory.


Visualizing replication factories during S phase of budding yeast

Polymerases α, ε and PCNA are components of replisomes (Baker and Bell, 1998; Waga and Stillman, 1998; Johnson and O’Donnell, 2005). To visualize replisomes in budding yeast, we tagged POL1, POL2 (the catalytic subunits of polymerase α, ε, respectively) with 4 tandem copies of green fluorescent protein (GFP), and PCNA with yellow fluorescent protein (YFP). All tagging was done at their original gene loci and all tagged genes were expressed with their authentic promoters.

When diploid cells were not in S phase, only weak globular POL1-4GFP, POL2-4GFP and YFP-PCNA signals were found in the nucleus (Fig 1A, unbudded cells or cells with medium or large buds). In S phase (small-budded) cells, however, each construct, showed up to 12 bright globular signals in the diploid nucleus. The size of each globular signal was 180-280 nm in diameter (230 ± 50 nm: mean ± SD), much smaller than the replication factories of vertebrate cells (Leonhardt et al., 2000; Somanathan et al., 2001; Sporbert et al., 2002). After DNA replication (cells with large buds), Pol1-4GFP, but not Pol2-4GFP or YFP-PCNA, re-localized along the nuclear envelope (see Supplementary Note 1). We obtained similar results using haploid cells with POL1-4GFP and POL2-4GFP, but the number of bright globular signals was up to 8 per nucleus. We demonstrate below that these bright globules of replication proteins represent replication factories where the bulk of DNA synthesis takes place.

Fig 1
Replication factories in budding yeast

Formation of replication factories is a consequence of DNA replication

We next addressed whether formation of the clusters of replication proteins is directly related to DNA replication. When B-type cyclins CLB5 and 6 are deleted, DNA replication is significantly delayed relative to bud emergence (Schwob and Nasmyth, 1993). In clb5,6-deleted cells, the formation of bright globular signals of Pol1-4GFP was also delayed relative to bud emergence (Fig 1B), suggesting that they are formed as a consequence of B-type cyclin/ cyclin-dependent kinase (CDK) activation in S phase. This notion is reminiscent of observations in vertebrate cells showing that cyclin E-CDK is required for the globular signals of replication protein A (RPA) becoming tightly bound to DNA during S phase (Jackson et al., 1995; Dimitrova et al., 1999).

We next studied Pol1-4GFP and YFP-PCNA localization in Cdc6-depleted cells. Cdc6 associates with DNA replication origins (Tanaka et al., 1997) and is required for replication initiation (Blow and Tanaka, 2005). Cdc6-depleted cells do not replicate DNA but still undergo other cell cycle events such as bud emergence and CDK activation (Piatti et al., 1995). We inhibited CDC6 expression in cells where the only functional CDC6 was under control of a galactose-inducible promoter (Piatti et al., 1996). After CDC6 expression was suppressed, small-budded cells never showed bright globular signals of Pol1-4GFP (Fig (Fig1C,1C, S1). Similar results were obtained with YFP-PCNA (data not shown). Taken together, our data indicate that bright globular foci of replication proteins are formed as a consequence of DNA replication.

We then treated cells with the DNA synthesis inhibitor hydroxyurea (HU) from G1 phase (Fig S2). When HU-treated cells started budding (S phase), globular Pol1-4GFP signals became brighter than in G1 phase, but remained dimmer than those in small-budded cells in the absence of HU (Fig S2, pink bars). The number of such Pol1-4GFP globules was generally not more than 2-3 in HU-treated cells. The results are consistent with HU limiting DNA replication to the vicinity of early-firing origins (Santocanale and Diffley, 1998).

Subsequently, we followed the behaviour of bright globular signals of Pol1-4GFP with time-lapse microscopy. During observation over 30 seconds, the globular signals showed significant changes in their shape and location, indicating dynamic behaviour of replication factories (Fig S3, S4; see Supplementary Note 2).

DNA replication of chromosomal loci observed in individual live cells

We next established an assay system for the replication of specific chromosomal loci in individual live cells. In budding yeast, a genome-wide study previously revealed the replication timing of all chromosome loci (Raghuraman et al., 2001). We chose for further study a single large replicon on chromosome IV. We selected two loci, 60 kb apart and on opposite sides of the replication origin, which are replicated at similar times (Fig 2A). We then marked one of these loci with an array of 224 tet operators (tetOs-1) and marked the other with an array of 256 lac operators (lacOs) (Straight et al., 1996; Michaelis et al., 1997). To minimize potential perturbation of the replication profile, only the tet and lac operator arrays, each ~11 kb in size, were inserted, but no other DNA sequences such as plasmid backbones. The tetO array can be bound by a fusion protein of the Tet repressor plus three tandem copies of cyan fluorescent protein (TetR-3CFP: Bressan et al., 2004), whilst the lacO array can be associated with a fusion protein of the lac repressor plus GFP (GFP-lacI: Straight et al., 1996). Using strains expressing TetR-3CFP and GFP-lacI, we could visualize these chromosomal loci as small CFP and GFP dots.

Fig 2
Determining replication timing by time-lapse microscopy

When these dots were observed in asynchronous cell cultures, their fluorescence intensity was approximately doubled in cells with medium-size buds (before sister chromatid separation in anaphase) when compared to unbudded cells (Fig S5). The dot intensity did not increase if DNA replication of these loci was inhibited by HU (Fig S6). We reasoned that DNA replication doubled the number of tet and lac operators, which led to recruitment of more TetR-3CFP and GFP-lacI fusion proteins. Consistent with this idea, both loci in individual cells doubled in intensity at approximately the same time during the cell cycle (Fig 2B). A slight difference in replication timing of the two loci, which was seen in a few cells, is expected to occur due to differences in fork velocity or if the replication origin between the two loci failed to fire.

How long does it take for each dot to become brighter after the locus replicates? To address this, we compared the kinetics of changes in the dot intensity (Fig 2B) with incorporation of the thymidine analogue 5-bromo-deoxyuridine (BrdU), into DNA at each locus, during the cell cycle (Fig 2 C-D). BrdU incorporation was analysed by chromatin immunoprecipitation of sheared and denatured DNA, using an antibody against BrdU (Pichler et al., 1997), followed by PCR amplification of the two loci plus the replication origin lying between them (Fig 2C-D).

As expected from the previously published timing data (Raghuraman et al., 2001), the tetOs-1 and lacOs sites showed BrdU incorporation in late S phase about 20-min later than the replication origin (Fig 2 D-E). Notably, BrdU incorporation and the increase in the dot intensity showed almost identical kinetics for both tetOs-1 and lacOs integration sites (Fig 2B, D), suggesting that tetOs-1-CFP and lacOs-GFP dots became brighter very soon after they were replicated. We can therefore use the intensity of tetOs-1-CFP and lacOs-GFP dots as a read-out of replication of these loci.

DNA replication of chromosomal loci observed at replication factories

Having established techniques for visualizing replication factories and for visualization of the replication of specific chromosomal loci, we were in a position to test whether replication actually occurs within the factories. To this end, we created a strain combining Pol1-4GFP, TetR-3CFP and the tetO array (without the lacOs dot). If DNA at the tetO array replicates in one of replication factories, the intensity of the dot would be expected to increase when it colocalizes with a bright globular Pol1-4GFP signal (Fig 3A). To test this hypothesis, we measured the intensity of the tetOs-1-CFP dot and its three-dimensional position relative to the Pol1-4GFP signal in cells undergoing replication of this locus.

Fig 3
DNA replication of chromosomal loci in replication factories.

We observed 11 cells in which their tetOs-1-CFP dot increased in its intensity during observation, and in all these cells the dot colocalized with one of the bright globular Pol1-4GFP signals for a period of 2-7 min when the dot intensity started to increase (Fig (Fig3B,3B, S7). The peak of colocalization frequency was found 1 min prior to replication time, defined as the mid-point of the increase in the dot intensity on the regression curve (Fig 3C). The number and size of replication factories did not increase when the tetO array was replicated (Fig S8; the gradual decrease of their signals is probably due to photo-bleaching during time-lapse microscopy), suggesting a specific co-localization of the two structures. Colocalization was occasionally observed at times when the tetO array was not in the process of replication (Fig 3C, Fig S7), but this occurred at a frequency below that expected by chance, given that the globular Pol1-4GFP signal occupies about a third of the nuclear volume (32.8 ± 3.8 %). This suggests that when the tetO array was not actually in the process of DNA replication, it was excluded from replication factories. Similarly, when replication of the tetO array was not observed during the time-lapse experiment, the tetO array only occasionally colocalized with bright globular Pol1-4GFP signals (Fig S9).

These results suggest that bright globular Pol1-4GFP signals, found specifically during S phase, are the sites of DNA replication and indeed represent replication factories. We can therefore visualize replication of chromosomal loci at replication factories by time-lapse microscopy.

Evidence that sister replication forks are closely associated during replication

The method we describe here allows us to determine which chromosomal loci are replicated in close proximity to one another at particular replication factories. In particular, we wanted to test the hypothesis that sister replication forks, generated from the same origin, are associated with each other during replication (Dingman, 1974; Falaschi, 2000). This hypothesis predicts that, as replication proceeds, template DNA enters an associated double replisome located within a replication factory and newly-synthesized DNA is then extruded from it (Fig 4A). To address this point we used the strain described above, where the tetO and lacO arrays were integrated on either side of a replication origin in a single large replicon (Fig 4A; tetOs-1 and lacOs). If the two sister forks from the central origin stay together at an associated double replisome during replication, then tetO and lacO arrays would be brought together when they are replicated and would then separate as they are extruded from the replisomes (Fig 4A).

Fig 4
Behaviour of two chromosomal loci on the opposite sides of a replicon with similar replication timing

With time-lapse microscopy, we observed 12 yeast cells in which the tetOs-1 and lacOs dots increased their intensity during the observation period (Fig 4B-C, S10). As expected from Fig 2, the two loci replicated at very similar times, with a difference of only 2.0±1.4 min (mean±SD). We standardized the observation times to the mid-replication time, defined as the mean replication time of the two dots. An example is shown in Fig 4B; 2 to 3 minutes before the mid-replication time, the two loci moved very close together (120-150 nm) for two consecutive time points just before the dot intensity started to increase.

All cells observed showed a close localization of the two dots (separation ≤ 350 nm) for 2 minutes or longer, 0-3 min before the mid-replication time (Fig 4C). Although close localization of the two dots was occasionally found at other times, a peak of frequency in close localization of the two dots was evidently at −2 to 0 min relative to mid-replication time of the two dots (Fig 4C).

When more stringent thresholds were set to judge close localization, the peak of frequency in close localization of the two dots was still found at −2 to 0 min, but the height of the peak became lower (Fig S11A). We assume that the lacOs-GFP and tetOs-1-CFP dots did not always come very close (e.g. < 200 nm) around mid-replication time because their replication time was never exactly the same, and was more different in some cells than others. In fact, there was a correlation between the synchrony of replication and the proximity of the loci (95% confidence, r=0.58), so that if two dots showed larger differences in replication time they tended to show larger minimum distances around their mid-replication time (see Supplementary Note 3 for more details).

To confirm further the close localization of the two dots upon replication, we also measured their median distance at each time point (relative to mid-replication time) among the 12 cells (Fig 5A). Although taking median distance among different cells significantly masked the minimum separation at any particular time, the median distance was indeed smaller, especially at −2 to +1 min relative to mid-replication time. These data suggest that the tetOs-1 and lacOs dots are in closer proximity when DNA replication takes place at these loci, and move apart from each other after replication.

Fig 5
Median distance between tetOs-1 and lacOs dots is reduced around mid-replication time.

To address whether close localization of the two loci upon replication is dependent on their particular positions within the replicon, we made a control strain where the array of tetOs was at a different site, 60 kb from the array of lacOs, but in a direction away from the replication origin (tetOs-2: Fig 4A). From the previously published genome-wide replication study (Raghuraman et al., 2001), we would expect tetOs-2 to replicate significantly earlier than lacOs. We observed 9 cells in which the tetOs-2 and lacOs dots increased their intensity during the observation period (Fig S12). As expected, the tetOs-2 dot showed an increase in its intensity 14-20 min earlier than the lacOs dot. In contrast to the tetOs-1 dot, the tetOs-2 dot did not show frequent close localization with the lacOs dot around the mid-replication time (Fig (Fig4D,4D, S11B). The median distance between the two dots was not significantly smaller around the mid-replication time (Fig 5B).

Localization and replication of two chromosomal loci on the same or on separate homologous chromosomes in diploid cells

The above experiments revealed that two chromosomal loci on opposite sides of a single replicon with similar replication timing, show close localization as they replicate. We next addressed whether such close localization was found even if the two loci were on separate homologous chromosomes. For this purpose, we made two kinds of diploid yeast strains. In diploid strain A, the tetOs-1 and lacOs dots were on separate homologous chromosomes, while in diploid strain B they were on the same chromosome (Fig 6A). We analysed 7 diploid A cells and 5 diploid B cells where the two dots increased in intensity during the observation period (Fig (Fig6B,6B, S13).

Fig 6
Behaviour of two loci on the same or on homologous chromosomes in diploids.

The tetOs-1 and lacOs loci showed similar replication timing in diploid A cells (difference: 3.8 ± 2.8 min) and in diploid B cells (difference: 2.4 ± 1.0 min). However, in 3 diploid A cells (cells #2, 3 and 5), the difference in replication timing was 5 min or larger, which seldom happened in either diploid B cells or in haploid cells (Fig (Fig4C,4C, ,6B).6B). This was probably due to differences in the timing of replication initiation from the two relevant origins on the homologous chromosomes.

In all diploid B cells, as in the haploid cells, the tetOs-1 and lacOs loci closely localized at about the time both loci replicated (Fig 6B). By contrast, in none of diploid A cells was close localization of the two dots observed around the mid-replication time of the two dots. Even when the two loci replicated at a very similar time in the diploid A (cells #1, 6 and 7 with a replication timing difference of 1-2 min), their close localization was not observed. Although some diploid A cells had a larger distance between the tetOs-1 and lacOs dots before DNA replication than did diploid B cells, this was not the sole reason for the lack of close localization upon replication of the two dots in diploid A. In fact, a couple of diploid A cells (Fig S13, diploid A cells # 2 and 3) had a separation of the two dots as small as diploid B cells (400-800 nm) before DNA replication, but still did not show close localization around mid-replication time.

These results show that only when the two loci are on the same chromosome (and thus within the same replicon) do they consistently show close localization as they replicate. Replication of two chromosomal loci with a similar timing does not necessarily lead to their close localization upon replication.

Global sites of DNA replication associated with replication factories

We then studied how newly synthesized DNA globally localizes in the nucleus relative to replication factories. Previous studies showed that new DNA synthesis, visualized by incorporation of BrdU etc, takes place at multiple foci in the budding yeast nucleus (Pasero et al., 1997; Lengronne et al., 2001; Hiraga et al., 2005). We incubated POL1-4GFP cells with BrdU for 3 minutes, the minimum incubation time after which BrdU incorporation into chromosomes can be detected. After the lysis of cells, chromosomes were fixed on glass slides, and Pol1-4GFP and incorporated BrdU were visualized by immuno-staining. BrdU incorporation was observed in a subset of nuclei with bright globular Pol1-4GFP signals, which correspond to S-phase nuclei. In such nuclei, most, if not all, of the BrdU signals were found colocalized or closely associated with bright (and some weak) globular Pol1 signals (Fig S14). The BrdU signals, which were adjacent to, rather than co-localized at, globular Pol1 signals, may be a consequence of newly synthesized DNA spooling out of replication factories. If so, newly synthesized DNA must have moved up to 350 nm or even more in 3 min after replication. This is consistent with our estimation, based on time-lapse analyses of the two chromosomal loci, that replicated DNA can move away from a replication factory at maximum at a speed of 250 nm/min (see Fig 4B, Supplementary Note 4).


By marking individual chromosome loci with bacteria-derived operators, we have developed a new assay to study the dynamics of DNA replication with time-lapse microscopy. For temporal studies of DNA replication, conventional methods (e.g. FACS DNA content analyses, Amberg et al., 2005; density-shift analyses of isotope-labelled DNA, Raghuraman et al., 2001; and immunoprecipitation of BrdU-incorporated DNA, Pichler et al., 1997) have used cells cultured synchronously in the cell cycle. However, it is not technically possible to obtain cells in perfect synchrony. Further, even in perfectly synchronized cells, many replication processes may have a significant stochastic element. For example, because not all origins fire in every cell cycle some origins may fire in some cells but not others; similarly, particular origins may fire at different times in S phase in different cell cycles (Schwob, 2004; Machida et al., 2005). This inter-cell variation is a major obstacle to understanding the dynamics of DNA replication using population-based assays. Because our assay analyses timing of DNA replication in individual cells, much greater temporal resolution can be obtained than by the conventional methods, and inter-cell variation can be directly monitored. Moreover, by simultaneously following intensity and localization of marked chromosome loci, chromosome duplication can be studied both spatially and temporally in individual live cells.

Using this novel assay, we have studied how individual replicons are processed for DNA replication. We discovered that two chromosomal loci, having the same replication timing but located at the opposite sides of the same replicon, move close together upon DNA replication and separate from each other afterwards. Our discovery strongly supports a long-standing hypothesis, which has however lacked firm evidence in eukaryotic cells, that sister replication forks generated from the same origin stay associated with each other during replication of the relevant replicon (Dingman, 1974; Falaschi, 2000). In prokaryotic cells, recent evidence also supports the idea that the two replisomes stay together in the middle of the cell during chromosomal DNA replication (Lemon and Grossman, 2000; Jensen et al., 2001; Lau et al., 2003; Migocki et al., 2004).

The second advance demonstrated in our study is visualization of replication factories in S. cerevisiae. Previous studies using fixed nuclei have suggested that replication proteins in budding yeast are clustered together at potential sites of DNA synthesis (Ohya et al., 2002; Hiraga et al., 2005). Similarly, observations using a cell-free replication system suggested that DNA synthesis occurs in small globular foci within yeast nuclei (Pasero et al., 1997). We show here that replication factories are formed in globular patterns specifically during S phase of the budding yeast, and colocalize with sites of DNA synthesis. This is similar to factories observed in vertebrate cells (Hozak et al., 1993; Frouin et al., 2003), though the yeast foci have a smaller size. A recent report also suggested that fission yeast GFP-PCNA shows bright globular signals specifically in S-phase nuclei (Meister et al., 2005).

We show that factory formation in yeast is dependent on S phase-cyclin CDKs (S-CDKs) as it is in vertebrate cells (Jackson et al., 1995; Dimitrova et al., 1999). We also addressed whether replication factory formation is dependent on DNA replication itself, rather than being independent events both regulated by S-CDKs. We found that factory formation was dependent on Cdc6, which is required for DNA replication initiation but not for progression of a cell cycle (CDK activation etc). Therefore, replication factory formation is likely to be a consequence of DNA replication in budding yeast. This is probably also the case in vertebrate cells because the loading of replication fork proteins onto DNA during S phase in Xenopus is dependent on the DNA having previously been licensed, which in turn is dependent on Cdc6 activity (Jares and Blow, 2000; Mimura et al., 2000). Because replisomes are assembled at replication origins dependent upon S-CDK activation, Cdc6 and origin unwinding (Tanaka and Nasmyth, 1998), we suggest that replication factories are mainly, if not exclusively, composed of replisomes at advancing replication forks. Our data is not readily consistent with formation of factories by replisome components prior to DNA replication.

In contrast to our results, a previous study using immuno-staining of fixed nuclei detected globular patterns of Pol2 and Pol3 (a catalytic subunit of polymerase δ) in the nucleus throughout cell cycle (Ohya et al., 2002). Such signals might correspond to the weak globular Pol2 (and Pol1, PCNA) signals that we also observed throughout the cell cycle. In fact, we found that after fixing cells for immuno-staining, the S phase-specific bright globular signals of Pol1, Pol2 and PCNA were less evident (data not shown). On the other hand, it is unlikely that the bright globular POL1-4GFP, POL2-4GFP and YFP-PCNA signals in S phase of live cells were artefacts of GFP/YFP tagging because 1) such signals were strictly dependent on DNA replication, 2) new DNA synthesis occurred specifically at these signals, and 3) POL2-4GFP bright globular signals were reduced in DNA primase mutants (data not shown).

Our observations of replication factories in budding yeast raise questions about whether multiple replicons are processed in a single replication factory at a given time during S phase, as in vertebrate cells, and, if so, how many replicons are processed simultaneously. A genome-wide replication profile in S. cerevisiae suggested that up to 80 replicons are undergoing replication at any given time in S phase of haploid cells (Raghuraman et al., 2001). Because we observed up to 8 replication factories in haploid cells, this suggest that an average of 10 replicons are processed in each factory at any given time. However, this number could be highly variable as factories have a wide range of brightness when visualized with GFP.

Obviously, budding yeast cells process a lower number of replicons in a single replication factory than do vertebrate cells (where 10-100 replicons are processed per factory). This difference accounts for the different size of replication factories between budding yeast (180-280 nm) and vertebrate cells (up to 500 nm-1 μm; Leonhardt et al., 2000; Somanathan et al., 2001; Sporbert et al., 2002) when visualized with GFP. Replication factories show more dynamic behaviour in budding yeast, presumably because factories containing a smaller number of replicons can change their shapes and sizes more readily as replicons join and leave factories as they initiate and terminate replication. It will be intriguing to address which replicons are processed in the same factory, the timing by which these replicons join the factory, and how these processes could differ from cell to cell, using our experimental system.

Why should cells keep sister replisomes closely associated and why do they assemble groups of replisomes in factories? One possible benefit might be to avoid only a half of replicon being replicated. Provided that replication origins lose their ability to initiate replication once the origin is unwound and replication forks are generated (Blow and Dutta, 2005), a half replicon might fail to replicate if one replisome could initiate replication without waiting for the other replisome to be loaded onto the origin. In addition, associated sister forks may coordinate the DNA polymerase operation for two leading and two lagging strands to avoid chromosome entanglement and to facilitate smooth reeling-in and –out of unreplicated and replicated DNA strands (Falaschi, 2000). Moreover, the assembly of multiple replisome pairs into a replication factory may be a way to increase the local concentration of DNA components (e.g. deoxyribonucleotides) and/or replication accessory proteins (e.g. PCNA and RPA), which could lead to more efficient DNA replication. Furthermore, cohesion between sister chromatids is also established during S phase (Uhlmann and Nasmyth, 1998). Proteins involved in the establishment of cohesion might be also concentrated in replication factories, which may account for coupled DNA replication and cohesion establishment.

To address possible benefits of colocalizing replisomes in factories, we need to intervene in these structures without interfering with other processes, and to follow the outcome. Such studies will be facilitated by use of the technology that we describe here.

Experimental Procedures

Yeast genetics and molecular biology

Yeast strain background (W303), methods for yeast culture and for cell cycle synchronization using α factor treatment or elutriation, and the TetR-GFP/tet and GFP-lacI/lac operator system were described previously (Piatti et al., 1996; Straight et al., 1996; Michaelis et al., 1997; Tanaka et al., 1997; Amberg et al., 2005). Cells were cultured at 25°C in YP medium containing glucose unless otherwise stated.

POL1 and POL2 were tagged with 4 tandem copies of GFP at their C-termini, and PCNA with YFP at its N-terminus. All tagging was done at the original gene loci by PCR methods as previously described (Prein et al., 2000; Maekawa et al., 2003), and all tagged genes were expressed with their authentic promoters. POL1, POL2 and PCNA are all essential for DNA replication and cell growth (Sugino, 1995; Waga and Stillman, 1998). Haploid cells containing POL1-4GFP and POL2-4GFP grew normally, suggesting that these tagged genes were functional. YFP-PCNA severely retarded haploid cell growth, indicating that the tagged construct was not fully functional. Therefore, we used haploid or homozygous diploid cells containing POL1-4GFP or POL2-4GFP, and diploid cells containing heterozygous YFP-PCNA (i.e. PCNA on the other homologous chromosome is not tagged), all of which showed normal growth.

To integrate tetO×224 and lacO×256 arrays into chromosomal loci (tetOs-1, tetOs-2 and lacOs) as shown in Fig Fig2A2A and and4A,4A, K. lactis URA3 was first inserted at non-coding regions between YDL089w and YDL088c, YDL020c and YDL019c, YDL055c and YDL054c, respectively, using a one-step PCR method (Amberg et al., 2005). Two 650-700 bp DNA fragments flanking the insertion sites were amplified by PCR and cloned into the pUC18 plasmid (GenBank/EMBL accession number L09136). tetO×224 and lacO×256 arrays were inserted between the two DNA fragments. Using these constructs, K. lactis URA3 was replaced with tetO×224 and lacO×256 arrays by negative selection against URA3 on culture plates containing 5-fluoroorotic acid (Amberg et al., 2005). Correct insertion of URA3 and subsequent replacement were confirmed by PCR amplification of relevant chromosomal regions. TetR-3CFP and GFP-lacI constructs were as described previously (Straight et al., 1996; Bressan et al., 2004). MATa/a diploid cells were made by expressing the HO gene from the GAL1-10 promoter in MATa/α diploid cells (Herskowitz and Jensen, 1991).

To facilitate BrdU incorporation (Vernis et al., 2003), 5 copies of the herpes simplex thymidine kinase gene were expressed from GPD1 promoters (Dahmann et al., 1995), and the human equilibrative nucleoside transporter 1 (ENT1) gene (its codon usage was optimized for yeast) was expressed from ADH1 promoter (K Shirahige et al, unpublished).

Microscopy and image analyses

The general procedures for time-lapse microscopy were described previously (Tanaka et al., 2005). Time-lapse images were collected at indicated time intervals at 23°C (ambient temperature) unless otherwise stated. Using the Deltavision microscope (Applied Precision), we acquired 5-9 (0.7 μm apart) z-sections, unless otherwise stated, which were subsequently deconvoluted and projected to two-dimensional images using SoftWoRx (Applied Precision) and Volocity (Improvision) software. To distinguish GFP and CFP signals, the JP4 filter set (Chroma) was used. GFP and CFP signals were quantified using Volocity. Replication factories were judged to colocalize with tetOs dots when Pol1-4GFP bright globular signals overlapped with tetOs-CFP dots on the focal plane for tetOs-CFP dots for two minutes or longer. To measure the distance between tetOs-CFP and lacOs-GFP dots, the distance between their centres was quantified in three dimensional space.

Other methods

FACS DNA content analyses, BrdU incorporation and subsequent chromatin immunoprecipitation, indirect immunostaining of epitope tags and incorporated BrdU were as described previously (Pichler et al., 1997; Dimitrova et al., 1999; Lengronne et al., 2001; Amberg et al., 2005). Find more methods in Supplementary Note 5.

Supplementary Material

Suppl inf


We thank M. K. Raghuraman, A. D. Donaldson, S. Hiraga, E. Schwob, J. R. Swedlow and all members of the Tanaka and Blow laboratories for discussions; L. Clayton for reading the manuscript; K. Shirahige, K. Nasmyth, R. Ciosk, S. Piatti, A. W. Murray, A. F. Straight, J. E. Haber, E. Schiebel, R. Tsien, Euroscarf, the Yeast Resource Centre for reagents; Y. Kitamura and S. Swift for technical help; F. Wheatley and the media kitchen for media preparation. This work was supported by Cancer Research UK grants C303/A3289 and C303/A3135. TUT was also supported by The Wellcome Trust, Lister Research Prize and Human Frontier Science Program.


Supplementary Data Supplementary Data including 5 Supplementary Notes, References and 14 Supplementary Figures can be found with this article online at……….


  • Amberg DC, Burke DJ, Strathern JN. Methods in yeast genetics. CSHL press; 2005.
  • Baker TA, Bell SP. Polymerases and the replisome: machines within machines. Cell. 1998;92:295–305. [PubMed]
  • Berezney R, Dubey DD, Huberman JA. Heterogeneity of eukaryotic replicons, replicon clusters, and replication foci. Chromosoma. 2000;108:471–484. [PubMed]
  • Blow JJ, Dutta A. Preventing re-replication of chromosomal DNA. Nat Rev Mol Cell Biol. 2005;6:476–486. [PMC free article] [PubMed]
  • Blow JJ, Tanaka TU. The chromosome cycle: coordinating replication and segregation. Second in the cycles review series. EMBO Rep. 2005;6:1028–1034. [PMC free article] [PubMed]
  • Bressan DA, Vazquez J, Haber JE. Mating type-dependent constraints on the mobility of the left arm of yeast chromosome III. J Cell Biol. 2004;164:361–371. [PMC free article] [PubMed]
  • Bussiere DE, Bastia D. Termination of DNA replication of bacterial and plasmid chromosomes. Mol Microbiol. 1999;31:1611–1618. [PubMed]
  • Dahmann C, Diffley JF, Nasmyth KA. S-phase-promoting cyclin-dependent kinases prevent re-replication by inhibiting the transition of replication origins to a pre-replicative state. Curr Biol. 1995;5:1257–1269. [PubMed]
  • Dimitrova DS, Todorov IT, Melendy T, Gilbert DM. Mcm2, but not RPA, is a component of the mammalian early G1-phase prereplication complex. J Cell Biol. 1999;146:709–722. [PMC free article] [PubMed]
  • Dingman CW. Bidirectional chromosome replication: some topological considerations. J Theor Biol. 1974;43:187–195. [PubMed]
  • Falaschi A. Eukaryotic DNA replication: a model for a fixed double replisome. Trends Genet. 2000;16:88–92. [PubMed]
  • Frouin I, Montecucco A, Spadari S, Maga G. DNA replication: a complex matter. EMBO Rep. 2003;4:666–670. [PMC free article] [PubMed]
  • Herendeen D, Kelly TJ. SV40 DNA replication. In: Blow JJ, editor. Eukaryotic DNA Replication. IRL press; 1996. pp. 29–65.
  • Herskowitz I, Jensen RE. Putting the HO Gene to work: practical uses for mating-type switching. In: Guthrie C, Fink GR, editors. Methods in Enzymology. Academic press; 1991. pp. 132–146. [PubMed]
  • Hiraga S, Hagihara-Hayashi A, Ohya T, Sugino A. DNA polymerases alpha, delta, and epsilon localize and function together at replication forks in Saccharomyces cerevisiae. Genes Cells. 2005;10:297–309. [PubMed]
  • Hozak P, Hassan AB, Jackson DA, Cook PR. Visualization of replication factories attached to nucleoskeleton. Cell. 1993;73:361–373. [PubMed]
  • Jackson PK, Chevalier S, Philippe M, Kirschner MW. Early events in DNA replication require cyclin E and are blocked by p21CIP1. J Cell Biol. 1995;130:755–769. [PMC free article] [PubMed]
  • Jares P, Blow JJ. Xenopus cdc7 function is dependent on licensing but not on XORC, XCdc6, or CDK activity and is required for XCdc45 loading. Genes Dev. 2000;14:1528–1540. [PMC free article] [PubMed]
  • Jensen RB, Wang SC, Shapiro L. A moving DNA replication factory in Caulobacter crescentus. Embo J. 2001;20:4952–4963. [PMC free article] [PubMed]
  • Johnson A, O’Donnell M. Cellular DNA replicases: components and dynamics at the replication fork. Annu Rev Biochem. 2005;74:283–315. [PubMed]
  • Lau IF, Filipe SR, Soballe B, Okstad OA, Barre FX, Sherratt DJ. Spatial and temporal organization of replicating Escherichia coli chromosomes. Mol Microbiol. 2003;49:731–743. [PubMed]
  • Lemon KP, Grossman AD. Localization of bacterial DNA polymerase: evidence for a factory model of replication. Science. 1998;282:1516–1519. [PubMed]
  • Lemon KP, Grossman AD. Movement of replicating DNA through a stationary replisome. Mol Cell. 2000;6:1321–1330. [PubMed]
  • Lengronne A, Pasero P, Bensimon A, Schwob E. Monitoring S phase progression globally and locally using BrdU incorporation in TK(+) yeast strains. Nucleic Acids Res. 2001;29:1433–1442. [PMC free article] [PubMed]
  • Leonhardt H, Rahn HP, Weinzierl P, Sporbert A, Cremer T, Zink D, Cardoso MC. Dynamics of DNA replication factories in living cells. J Cell Biol. 2000;149:271–280. [PMC free article] [PubMed]
  • Machida YJ, Hamlin JL, Dutta A. Right place, right time, and only once: replication initiation in metazoans. Cell. 2005;123:13–24. [PubMed]
  • Maekawa H, Usui T, Knop M, Schiebel E. Yeast Cdk1 translocates to the plus end of cytoplasmic microtubules to regulate bud cortex interactions. Embo J. 2003;22:438–449. [PMC free article] [PubMed]
  • Meister P, Taddei A, Vernis L, Poidevin M, Gasser SM, Baldacci G. Temporal separation of replication and recombination requires the intra-S checkpoint. J Cell Biol. 2005;168:537–544. [PMC free article] [PubMed]
  • Michaelis C, Ciosk R, Nasmyth K. Cohesins: chromosomal proteins that prevent premature separation of sister chromatids. Cell. 1997;91:35–45. [PubMed]
  • Migocki MD, Lewis PJ, Wake RG, Harry EJ. The midcell replication factory in Bacillus subtilis is highly mobile: implications for coordinating chromosome replication with other cell cycle events. Mol Microbiol. 2004;54:452–463. [PubMed]
  • Mimura S, Masuda T, Matsui T, Takisawa H. Central role for cdc45 in establishing an initiation complex of DNA replication in Xenopus egg extracts. Genes Cells. 2000;5:439–452. [PubMed]
  • Nakamura H, Morita T, Sato C. Structural organizations of replicon domains during DNA synthetic phase in the mammalian nucleus. Exp Cell Res. 1986;165:291–297. [PubMed]
  • Newport J, Yan H. Organization of DNA into foci during replication. Curr Opin Cell Biol. 1996;8:365–368. [PubMed]
  • Ohya T, Kawasaki Y, Hiraga S, Kanbara S, Nakajo K, Nakashima N, Suzuki A, Sugino A. The DNA polymerase domain of pol(epsilon) is required for rapid, efficient, and highly accurate chromosomal DNA replication, telomere length maintenance, and normal cell senescence in Saccharomyces cerevisiae. J Biol Chem. 2002;277:28099–28108. [PubMed]
  • Pasero P, Braguglia D, Gasser SM. ORC-dependent and origin-specific initiation of DNA replication at defined foci in isolated yeast nuclei. Genes Dev. 1997;11:1504–1518. [PubMed]
  • Piatti S, Bohm T, Cocker JH, Diffley JF, Nasmyth K. Activation of S-phase-promoting CDKs in late G1 defines a “point of no return” after which Cdc6 synthesis cannot promote DNA replication in yeast. Genes Dev. 1996;10:1516–1531. [PubMed]
  • Piatti S, Lengauer C, Nasmyth K. Cdc6 is an unstable protein whose de novo synthesis in G1 is important for the onset of S phase and for preventing a ‘reductional’ anaphase in the budding yeast Saccharomyces cerevisiae. Embo J. 1995;14:3788–3799. [PMC free article] [PubMed]
  • Pichler S, Piatti S, Nasmyth K. Is the yeast anaphase promoting complex needed to prevent re-replication during G2 and M phases? Embo J. 1997;16:5988–5997. [PMC free article] [PubMed]
  • Prein B, Natter K, Kohlwein SD. A novel strategy for constructing N-terminal chromosomal fusions to green fluorescent protein in the yeast Saccharomyces cerevisiae. FEBS Lett. 2000;485:29–34. [PubMed]
  • Raghuraman MK, Winzeler EA, Collingwood D, Hunt S, Wodicka L, Conway A, Lockhart DJ, Davis RW, Brewer BJ, Fangman WL. Replication dynamics of the yeast genome. Science. 2001;294:115–121. [PubMed]
  • Santocanale C, Diffley JF. A Mec1- and Rad53-dependent checkpoint controls late-firing origins of DNA replication. Nature. 1998;395:615–618. [PubMed]
  • Schwob E. Flexibility and governance in eukaryotic DNA replication. Curr Opin Microbiol. 2004;7:680–690. [PubMed]
  • Schwob E, Nasmyth K. CLB5 and CLB6, a new pair of B cyclins involved in DNA replication in Saccharomyces cerevisiae. Genes Dev. 1993;7:1160–1175. [PubMed]
  • Somanathan S, Suchyna TM, Siegel AJ, Berezney R. Targeting of PCNA to sites of DNA replication in the mammalian cell nucleus. J Cell Biochem. 2001;81:56–67. [PubMed]
  • Sporbert A, Gahl A, Ankerhold R, Leonhardt H, Cardoso MC. DNA polymerase clamp shows little turnover at established replication sites but sequential de novo assembly at adjacent origin clusters. Mol Cell. 2002;10:1355–1365. [PubMed]
  • Straight AF, Belmont AS, Robinett CC, Murray AW. GFP tagging of budding yeast chromosomes reveals that protein-protein interactions can mediate sister chromatid cohesion. Curr Biol. 1996;6:1599–1608. [PubMed]
  • Sugino A. Yeast DNA polymerases and their role at the replication fork. Trends Biochem Sci. 1995;20:319–323. [PubMed]
  • Tanaka K, Mukae N, Dewar H, van Breugel M, James EK, Prescott AR, Antony C, Tanaka TU. Molecular mechanisms of kinetochore capture by spindle microtubules. Nature. 2005;434:987–994. [PubMed]
  • Tanaka T, Knapp D, Nasmyth K. Loading of an Mcm protein onto DNA replication origins is regulated by Cdc6p and CDKs. Cell. 1997;90:649–660. [PubMed]
  • Tanaka T, Nasmyth K. Association of RPA with chromosomal replication origins requires an Mcm protein, and is regulated by Rad53, and cyclin- and Dbf4-dependent kinases. Embo J. 1998;17:5182–5191. [PMC free article] [PubMed]
  • Uhlmann F, Nasmyth K. Cohesion between sister chromatids must be established during DNA replication. Curr Biol. 1998;8:1095–1101. [PubMed]
  • Vernis L, Piskur J, Diffley JF. Reconstitution of an efficient thymidine salvage pathway in Saccharomyces cerevisiae. Nucleic Acids Res. 2003;31:e120. [PMC free article] [PubMed]
  • Waga S, Stillman B. The DNA replication fork in eukaryotic cells. Annu Rev Biochem. 1998;67:721–751. [PubMed]
  • Wessel R, Schweizer J, Stahl H. Simian virus 40 T-antigen DNA helicase is a hexamer which forms a binary complex during bidirectional unwinding from the viral origin of DNA replication. J Virol. 1992;66:804–815. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...