Segregation of Transcription and Replication Sites Into Higher Order Domains

See allHide authors and affiliations

Science  04 Sep 1998:
Vol. 281, Issue 5382, pp. 1502-1505
DOI: 10.1126/science.281.5382.1502


Microscopy shows that individual sites of DNA replication and transcription of mammalian nuclei segregate into sets of roughly 22 and 16 higher order domains, respectively. Each domain set displayed a distinct network-like appearance, including regions of individual domains and interdigitation of domains between the two networks. These data support a dynamic mosaic model for the higher order arrangement of genomic function inside the cell nuclei.

Genomic processes have an underlying structural organization in the cell nucleus (1, 2). The genome itself is arranged into discrete chromosome-specific territories (3), and there is an emerging view that the genome and its associated functional domains are dynamically linked in an overall nuclear architecture termed the nuclear matrix (1, 2,4). Numerous studies have demonstrated that the sites of DNA replication, transcription, RNA splicing factors, and RNA tracks visualized in the nucleus of the intact cell are spatially maintained following extraction for nuclear matrix (5–7).

Another approach to investigating the relationship of nuclear architecture to genomic function is to determine whether the individual sites of replication or transcription, or both, are arranged into higher order domains in the cell nucleus. We directly addressed this issue by simultaneously labeling sites of replication and transcription in permeabilized mouse 3T3 or human diploid NHF1 fibroblasts (8). Fluorescence laser scanning confocal microscopy and three-dimensional image analysis were then used to visualize the individual sites of replication and transcription, and the spatial relationships between these sites (9). The extranucleolar transcription sites that are visualized with this procedure are predominantly, if not exclusively, transcribed by RNA polymerase II (8).

We found that the individual sites of DNA replication and transcription are spatially distinct (separate red and green colors) during all periods of the S phase (Fig. 1) (10). More than 95% of the replication sites activated early in the S phase do not coincide with transcription sites (11). A yellow color represents an overlap between red and green signals. The very small percentage of yellow sites observed (arrows in Fig. 1I) is likely due to the occasional overlay of the green replication and red transcription sites at different levels in three-dimensional space, rather than concurrence of red and green fluorescence at a single site. We routinely observe this phenomenon in optical sections of mixed red and green fluorescence beads. These results are consistent with those of Wansink et al. (12) but in direct contradiction to those of Hassanet al. (13).

Figure 1

Spatial separation of replication and transcription sites throughout the S phase. DNA replication and transcription sites are visualized in green and red, respectively, in mouse 3T3 cells. Merged images of replication and transcription sites are displayed in the right panels. (A through I) Early S phase. Enlarged areas are displayed in (D) through (I). (F) Thick and thin arrows indicate clusters of transcription and replication sites, respectively. Examples of apparent overlap between transcription and replication sites are indicated with arrows in (I). (J through L) Middle to late S phase. (M through O) Late S phase.

We observed that replication (thin arrows; Fig. 1F) and transcription sites (thick arrows; Fig. 1F), aside from being separate from one another, were grouped into separate and distinct clusters. To study the overall distribution of individual sites, we performed contour analysis on individual optical sections at mid-plane (Fig. 2, A and B) in nuclei showing typical early S phase replication patterns (14). Of the contoured area in the extranucleolar regions (15), 77.4% were occupied by separate clusters of replication and transcription sites (Fig. 2, C and E). Because this analysis was performed on exponentially dividing cells, the findings should reflect the entire early S period (approximately 4 hours), when most active genes are replicated (10). Similar results were obtained in cells labeled immediately after release from the G1/S border (16). This indicates preferential grouping, because only 36.9% of the extranucleolar area is predicted to form separate replication and transcription clusters by a random clustering model (Fig. 2, D and E) (17).

Figure 2

Preferential clustering of replication and transcription sites in early S phase. (A) Clustered replication sites are contoured with light green lines. (B) Clustered transcription sites are contoured with pink lines. (C) Areas occupied by replication site clusters and transcription site clusters are filled with green and red, respectively. Areas occupied with replication and transcription sites in a mixed pattern are filled with yellow. Nuclear areas with sparsely distributed replication and transcription sites are not contoured and appear as black regions (14). (D) The same contour analysis was applied to a computer-generated random sampling model (17). (E) Bar plot of percentage distribution of separate clustered (black bars) versus mixed (striped bars) areas in the extranucleolar regions of experimental samples and random sampling (16). Error bars denote the standard error of the mean.

We next examined the replication and transcription clusters in three dimensions. Contour analysis was performed on individual optical sections for each nucleus (Fig. 3) followed by three-dimensional reconstruction (Fig. 4) (18). Individual contours corresponding to specific clusters of replication (Fig. 4A) or transcription (Fig. 4C) sites typically extended into several optical sections (0.5 μm per section). An average of 22 ± 2.7 and 16 ± 1.6 higher order domains for replication and transcription were calculated at a given instant in early S phase. Each set of higher order domains was further arranged into a discontinuous network-like appearance that extended throughout the nuclear volume (Fig. 4, A and C). Small regions of individual domains within each set were often found in close apposition with neighboring domains (Fig. 4, A through D). Although the network patterns corresponding to each set of domains appeared similar, they occupied completely separate regions of the nuclear volume (Fig. 4E). Individual domains from corresponding sets, however, were strongly juxtaposed throughout the overall three-dimensional arrangement (Fig. 4).

Figure 3

Cluster distribution of replication and transcription sites in early S phase extending into several optical sections. To study the cluster distribution in three dimensions, the same contour analysis was applied to a series of optical sections (A through I) of mouse fibroblast nuclei labeled for replication (light green contours) and transcription (pink contours) sites. Three nucleolar transcription clusters are also contoured in pink (indicated with arrows).

Figure 4

Replication and transcription sites in early S phase separate into distinct higher order domains in three dimensions. Replication and transcription site clusters drawn on individual optical sections were reconstructed to visualize the spatial relationship of clusters between different sections. In this particular mouse fibroblast nucleus, three-dimensional observation shows that replication site clusters form 24 higher order domains, and transcription site clusters form 20 higher order domains, three of which are located in nucleolar regions (indicated with arrows). (A) Stereopair of three-dimensional reconstructed DNA replication site clusters (light green contours). (B) Enlargement of the lower left part of (A). (C) Stereopair of three-dimensional reconstructed RNA transcription site clusters (pink contours). (D) Enlargement of the lower left part of (C). (E) Stereopair of merged image. (F) Enlargement of the lower left part of (E). All stereoimages were constructed with a 10° angle.

Previous studies indicate that individual sites of replication and transcription in the cell nucleus are composed of numerous replicons and genes, respectively (5, 6). Our results demonstrate an even higher order arrangement of these genomic sites. The extranucleolar region of the nucleus is segmented into a mosaic of spatially juxtaposed replication and transcription domains. Each individual domain contains numerous individual sites of replication or transcription, which are under common temporal control.

We propose that different domains of replication and transcription are progressively activated and inactivated as the cell transverses the S phase and that the changing spatial patterns of these higher order domains correlate with temporal programs of replication and transcription in the cell. In this way, a domain may function in replication at one time and in transcription at another time in S phase. Studies of replication timing of specific gene sequences provide further support for this conclusion (10). The interdigitation of replication and transcription domains that we observed in three dimensions may, therefore, be an indication of dynamic cross-talk between the replication and transcription domains or temporal transitions from one functional state to the other.

Arrangement of separate replication and transcription domains into even higher order network patterns suggests an underlying architectural basis. The notable preservation of replication and transcription sites on the nuclear matrix following in situ extraction (5, 6) implicates a dynamic nuclear architecture as a basis for these global spatial properties of genomic function. The recent identification of a specific nuclear matrix targeting sequence provides a possible direct approach for further study of higher order functional domains and nuclear architecture (19).

  • * Present address: Life Imaging Systems, 195 Dufferin Avenue, Suite 300, London, Ontario N6A1K7, Canada.

  • To whom correspondence should be addressed. E-mail: berezney{at}


View Abstract

Stay Connected to Science

Navigate This Article