| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
The pattern of replication at a human telomeric region (16p13.3): its relationship to chromosome structure and gene expression
Introduction
Results
Evaluation of the modified FISH-based assay to study the pattern and timing of DNA replication
The normal pattern of replication throughout the terminal region of 16p13.3
The effect of chromosomal deletions on the pattern of replication
The influence of the [alpha]-globin regulatory element on the pattern of replication
Replication of segments of 16p13.3 in transgenic mice
Discussion
The relationship between the pattern of replication and chromatin structure
Where does replication initiate in the GC-rich isochore?
A paradigm for replication in GC-rich and -poor regions of the genome
Conclusion
Materials And Methods
Cell culture and preparation of nuclei
Nick translation labelling of probes
FISH
Slide scoring
Acknowledgements
References
The pattern of replication at a human telomeric region (16p13.3): its relationship to chromosome structure and gene expression
Received March 30, 1999; Revised and Accepted May 6, 1999
We have studied replication throughout 325 kb of the telomeric region of a human chromosome (16p13.3) and related the findings to various aspects of chromosome structure and function (DNA sequence organization, nuclease-hypersensitive sites, nuclear matrix attachment sites, patterns of methylation and gene expression). The GC-rich isochore lying adjacent to the telomere, which contains the [alpha]-globin locus and many widely expressed genes, replicates early in the cell cycle regardless of the pattern of gene expression. In subtelomeric DNA, replication occurs later in the cell cycle and the most telomeric region (20 kb) is late replicating. Juxtaposition of early replicating DNA next to the telomere causes it to replicate later in S-phase. Analysis of the timing of replication in chromosomes with deletions, or in transgenes containing various segments of this telomeric region, suggests that there are no critical origins or zones that initiate replication, rather the pattern of replication appears to be related to the underlying chromatin structure which may restrict or facilitate access to multiple, redundant origins. These results contrast with the pattern of replication at the human [beta]-globin locus and this may similarly reflect the different chromosomal environments containing these gene clusters.
INTRODUCTION
An important goal in molecular and cellular biology is to integrate our knowledge of nuclear organization, chromosome structure and primary DNA sequence with epigenetic modifications (e.g. DNA methylation and histone acetylation) and the key nuclear processes of replication, transcription, recombination and repair. In higher eukaryotes, the mechanisms regulating replication and their relationships to other nuclear processes are proving amongst the most difficult to elucidate.
Mammalian chromosomes can be subdivided at low (5-10 Mb) resolution into early and late replicating bands, suggesting that large chromosomal regions of the genome are coordinately replicated during S-phase of the cell cycle (1,2). At a higher level of resolution, replication bands may be subdivided into large (50-300 kb) segments of DNA (replicons) each copied once during the cell cycle from a single, bi-directional origin (3,4). Within the nucleus, DNA replication takes place in several hundred discrete sub-domains referred to as `replication foci' which may be associated with the nuclear matrix. These sites first appear in late G1, and it has been estimated that each one contains a cluster of simultaneously active replicons associated with many of the proteins required for replication. As S-phase proceeds, new foci appear as adjacent clusters of replicons are duplicated in synchronized waves, suggesting that the temporally coordinated activation of different sets of replicons is in some way related to chromosome organization (reviewed in ref. 5).
It has been shown that initiation of replication in mammalian chromosomes requires an intact nucleus (6). Beyond this, relatively little is known about how or where replication starts or the mechanisms by which the timing of replication is regulated during the cell cycle. The prevailing models are based largely on data from bacteria (7), eukaryotic DNA viruses (8) and yeast (9). In each of these organisms, origins of replication are defined by specific cis-acting sequences that are bound and regulated by well-characterized complexes of trans-acting factors (reviewed in refs 10-12). However, observations on mammalian origins do not conform to such simple models. Some sites of initiation are discrete (0.5-2 kb) whereas others are spread over considerable distances (up to 11 kb) and, in some cases, replication bubbles may be detected throughout large (up to 65 kb) `initiation zones' (13). For most mammalian origins, no specific cis-acting sequence can be defined, although one such element recently was localized (14) and characterized (15) in the human [beta]-globin cluster. In some experimental systems, it appears that almost any large segment of mammalian DNA can initiate replication (16,17). A popular, but unproven, interpretation of these findings is that, in mammalian chromosomes, many sequence motifs potentially can act as origins although some sequences may be preferred (18). The choice of origins may be determined by their accessibility to the trans-acting complexes which initiate DNA replication (19). In general, such access is thought to be defined by features such as nuclear sub-localization, chromatin structure and the association with chromatin re-modelling factors.
Since transcription also depends on chromatin-restricted accessibility (20,21), this hypothesis might explain some of the observed correlations between patterns of replication and gene expression. For example, early replicating bands correspond to cytogenetically defined R and T bands which in turn correspond to GC-rich, Alu-rich regions of the genome that contain a high density of CpG islands and genes. In contrast, late replicating bands correspond to segments of heterochromatin and the relatively gene-poor, AT-rich, Giemsa (G) bands (22,23). Although there are exceptions, many expressed genes are replicated early in S-phase whereas repressed genes are replicated late; for some tissue- and developmental stage-specific genes, transcription and replication timing are tightly coupled (23-25). It is interesting that replication foci active earliest in S-phase are located near, but distinct from, transcription foci active at the G1/S transition (26). Despite these observations, the precise relationship (cause or effect) between gene expression and replication in higher eukaryotes remains unknown.
Progress in understanding the mechanisms by which mammalian chromosomes replicate and how this relates to other nuclear processes has been hampered by the lack of adequate experimental assays and the fact that very few chromosomal loci have been analysed in detail. The telomeric region of the human G light band, 16p13.3, provides a uniquely well-characterized model to study these issues (summarized in Fig. 1). DNA sequence analysis of a contiguous 325 kb segment of DNA extending from the terminal (TTAGGG)n repeats has shown that immediately adjacent to the subtelomeric region lies an Alu-rich, GC-rich (~54%) isochore. The majority of this region can be described as `open' chromatin, containing many constitutively expressed DNase I-hypersensitive sites; but it is surrounded by segments of apparently `closed' chromatin including the telomere, the subtelomeric region and a long segment of nuclease-insensitive chromatin between coordinates ~180 000 and 218 000. The entire 325 kb region contains many unmethylated CpG islands associated with both widely expressed and tissue-specific genes including the [alpha]-globin gene family (27-29). Expression of this gene cluster and its dependence on local and long-range regulatory elements has been studied extensively, while analyses of individuals with [alpha]-thalassaemia has identified >100 natural mutants affecting this chromosomal region (30).
Figure 1. The structure of the terminal segment of 16p13.3 and the cloned resources spanning this region. (A) A summary of the genomic data of the terminal 325 kb of 16p with the telomere (oval) on the left. The graph at the top represents the frequency (%) of CpG dinucleotides. Previously described genes [3-20 (27)] are shown as black boxes above (transcribed towards the centromere) or below (transcribed towards the telomere) the line. The [alpha]-globin regulatory element (HS -40) is shown as a red box. The embryonic ([zeta]) and fetal/adult ([alpha]) genes are indicated. Above the chromosome, matches to the consensus sequence for mammalian origins of replication 5[prime]-WAWTTDDWWWDHWGWHMAWTT-3[prime] where W = A or T, D = A, G or T, H = A, C or T, M = A or C (63) are shown (orange). Above this, the positions of SINES (green), LINES (blue) and CpG islands (yellow) are shown. The graph displays the frequency of CpG dinucleotides. Below the chromosome, the positions of previously described cosmids used as probes are shown. Below this, the transgenic constructs GG1/GG2 and [alpha]PAC100 are shown. (B) Details of the terminal 50 kb with annotations as above. A large region of internal telomeric repeats, which may subdivide the subtelomere into separate domains (75) is indicated (grey box). The scale is in kb with 1 representing the most telomeric nucleotide.
Here we have used a modified fluorescence in situ hybridization (FISH) assay to study the pattern and timing of replication throughout the terminal region of 16p13.3. We have looked for evidence of essential initiation zones or origins of replication by analysing mutant chromosomes (lacking defined regions of the chromosome) and transgenic mice (including defined regions of the chromosome). Finally, we have related the resulting patterns of replication to those of chromosome organization, chromatin structure and gene expression in this region. This analysis thus provides a comprehensive description of the pattern of replication timing throughout a telomeric region and addresses the current models for how replication of the human genome is related to chromosome structure.
RESULTS
Evaluation of the modified FISH-based assay to study the pattern and timing of DNA replication
In the past, the average pattern of replication timing at specific loci within a population of cells was determined by analysing newly replicated, bromodeoxyuridine (BrdU)-labelled DNA from cells isolated by elutriation or flow cytometry at sequential stages of S-phase. More recently, FISH has been used to study DNA replication at specific loci, in individual cells, exploiting the observation that before replication, hybridization to each chromatid appears as single signals whereas, after replication, duplicated signals are seen (Fig. 2). The proportion of duplicated signals in a population of unsynchronized nuclei reflects the average timing of replication during S-phase. In general, for non-imprinted autosomal loci, the results of FISH and elutriation-based assays have been in good agreement (14,31).
Figure 2. FISH-based analysis of replication. (A) A field from an asynchronous culture of a normal EBV-transformed lymphoblastoid cell line (TA). In this example, the slide was hybridized with cosmid probe c419C1 (Fig. 1) labelled with DIG-11-UTP and detected with FITC (green). The centre of the figure shows an S-phase nucleus (BrdU-positive, detected with Texas red) where both alleles of 16p have replicated (duplicated signals). On the left is a G1 nucleus (BrdU negative), where both alleles are unreplicated (single signals). Bottom right is a metaphase spread typical of those used to assess hybridization efficiency. DNA is counterstained with DAPI (blue). (B) S-phase nuclei (red) from a normal, diploid lymphoblastoid cell line (TA), hybridized to a 16p-specific cosmid probe. (i) Left-hand nucleus: an early stage of S-phase when both alleles are unreplicated (single signals); (ii) centre nucleus: replication of one allele has occurred (top signal duplicated) but not of the other (lower single signal); (iii) right-hand nucleus: later in S-phase when both alleles have duplicated.
Although FISH analysis provides a simple way to analyse the pattern of replication timing, considerable care is required to establish a robust, reproducible scoring method using slides with unequivocal hybridization signals and low background (see Materials and Methods). It has been noted previously that single signals in interphase nuclei can occur artefactually as a result of inefficient hybridization; this can be internally controlled by simultaneously scoring the efficiency of hybridization (%) in metaphase spreads on each slide analysed (Fig. 2). In addition, it has been shown that a small proportion of artefactual doublet signals can result from separation of denatured chromatin strands in G1 and G2 cells (32). By limiting analysis to BrdU-labelled S-phase nuclei (i.e. those cells that are really replicating), this problem is minimized (see Fig. 2 and Materials and Methods).
In extensive preliminary studies, incorporating these internal controls, we established that routinely scoring the proportions of single or duplicated hybridization signals in 200 S-phase nuclei gave accurate and reproducible results. By re-analysing several `early' and `late' replicating human and murine loci that had been characterized previously using FISH and/or elutriation methods (see Table 1, its footnotes and the experiments described below), we confirmed that the FISH assay, as used here, truly reflects the timing of replication in the cell cycle.
Table 1. The pattern of replication using control probes
| Cell line | CW44 | cJ21 | Cos Me2.5 | PYGM | Reference |
| K562 | 21 | 26 | 61 | 68 | 31 |
| Manca | 18 | 13 | 4 | 77 | 31 |
| Caco-2 | 62 | 66 | 15 | 64 | 31 |
| Lymphoblastoid 1 | 24 | 22 | 29 | 69 | This study |
| Lymphoblastoid 2 | 20 | 22 | 23 | 53 | This study |
| Lymphoblastoid 3 | 20 | 19 | 19 | 57 | This study |
The normal pattern of replication throughout the terminal region of 16p13.3
Since the entire terminal segment of 16p13.3 has been cloned in overlapping recombinants (Fig. 1), we could use the FISH assay to analyse the temporal order of replication throughout this telomeric region. For each experiment, ~30 slides were prepared from a single aliquot of cells to avoid any batch-to-batch variation. From each batch, nuclei on individual slides were hybridized to the appropriate early and late replicating controls (Table 1) and to probes from the 16p region (Fig. 1).
In diploid cells, the pattern of replication may differ on homologous chromosomes, e.g. due to imprinting (see below). Therefore, initially, we studied three, independently derived, interspecific hybrids (JY5-4, JG7 and CW3-6) each containing a single normal copy of human chromosome 16 in a mouse erythroleukaemia (MEL) cell background. We have shown previously that the terminal region of 16p assumes its normal pattern of DNase I hypersensitivity, CpG island methylation and gene expression in this experimental system (33). Others have shown that this (34,35) and other heterologous vertebrate systems (36,37) faithfully replicate mammalian chromosomes (23,24). FISH signals were scored easily in these hybrids, and the timing of replication of each probe could be assigned unequivocally to individual chromosomes. Since proterminal repeats are shared by the ends of many human chromosomes (38), specific subtelomeric regions cannot be studied in normal cell lines. However, these repeats rapidly diverge during evolution and share little homology with the subtelomeric sequences of mouse. Using interspecific hybrids therefore also allowed us to study replication of the 16p subtelomeric region using FISH analysis. The proportions of duplicated signals were thus determined for early (mouse: p48) and late (mouse: cSamD4) replicating controls, and each recombinant probe lying between the telomeric repeats and coordinate 325 000 of chromosome 16p (Table 2 and Fig. 3).
Figure 3. Replication timing throughout the terminal 325 kb of human 16p. (A) Each point is the mean value recorded for each cosmid, plotted at the midpoint coordinate of the cosmid along the 16p telomere. Error bars represent 1 SD. The greater the percentage doublet score, the earlier the region replicates. Horizontal black dashed lines indicate the mean percentage doublet scores for the early (top line) and late (bottom line) control. Results for the mean of all erythroid lines (Table 2) are shown in grey (closed diamond) and the non-erythroid cell line (TA) in black (open squares) dashed lines. (B) Replication timing of the control cosmids. Values represented by black bars were calculated by pooling scores obtained from all EBV lines in this study (±1 SD). PYGM is early replicating; cJ21, CW44 (CF locus) and ME2.5 ([beta]-globin) are late replicating. Values represented by hatched bars were calculated by pooling scores from all interspecific hybrids in this study (±1 SD). P48 is early replicating; cSamD4 is late replicating.
Table 2. The pattern of replication of erythroid cell lines
| Cosmid | % Doubletsa (% hybridization efficiency) | |||
| JY5-4 | JG7 | Meanb | SD | |
| cSamD4: mouse | 29 (81) | 33 (98) | ||
| cNFG9 | 56 (88) | 56 (92) | 56 | 0.4 |
| cRA36 | 62 (100) | 56 (93) | 60 | 3.8 |
| cGG4 | 75 (100) | 78 (100) | 76 | 2.5 |
| cNFG7 | 66 (95) | 69 (95) | 67 | 1.8 |
| cGG2 | 67 (100) | 70 (90) | 68 | 2.5 |
| cRN24 | 67 (100) | 73 (100) | 70 | 4.6 |
| cGG1 | 76 (100) | 67 (95) | 72 | 3.7 |
| cos12 | 61 (95) | 63 (90) | 67 | 8.9 |
| cRJ14 | 79 (97) | 81 (100) | 80 | 1.8 |
| c310H5 | 68 (100) | 74 (100) | 71 | 4.2 |
| c314G4 | 74 (100) | 82 (100) | 78 | 5.7 |
| c419C1 | 73 (100) | 76 (100) | 74 | 1.4 |
| P48: mouse | 63 (100) | 55 (100) | ||
| cSamD4: mouse | 29 (81) | 33 (98) | ||
| c419C1 | 73 (96) | 80 (96) | Control | |
| cNFG9 | 56 (88) | 56 (92) | 56 | 0.4 |
| pBam probes | 43 (95) | 47 (100) | 45 | 2.5 |
| pTel probes | 20 (86) | 29 (80) | 25 | 6.4 |
| P48: mouse | 63 (100) | 55 (100) | ||
bMean of the full sets of data shown, plus additional data (incomplete) from a third hybrid (CW3-6) not included in the table.
The proportion of duplicated signals (~70%) scored for most recombinant probes across this region (Table 2) was similar to the value (~60%) for the early replicating control (mouse: p48) and it was therefore considered early replicating. We noted that the region of Alu-dense, DNase I-insensitive chromatin lying between coordinates ~180 000 and ~218 000 (27) always replicated later than its flanking regions (albeit still relatively early in S-phase) and, in contrast to all other areas, duplicated signals from this segment (detected with Cos12; Fig. 1) often remained closely paired in many nuclei, suggesting that more time is taken for this region to separate than other regions along the duplicated chromatids. Subdivision of the cosmid into three EcoRI fragments (8.6, 9.0 and 16.0 kb) and further FISH analysis (data not shown) showed that this phenomenon was restricted mainly to the most Alu-dense region of Cos12. Still lower proportions of duplicated signals were found for the most telomeric cosmid (cNFG9, coordinates ~29 000-61 000) which extends into the subtelomeric region. The proportion of duplicated signals progressively decreased using probes lying closer to the 16p telomere [allele A (39)] (Fig. 1B and Table 2). The values obtained with the most telomeric probes (pTel: ~20-29%) were similar to those from the late replicating controls (~29-33%). Thus it appears that in an erythroid cell background, most of this 325 kb segment is duplicated early in S-phase but replication occurs progressively later throughout the subtelomeric region and is unequivocally late within 20 kb of the telomeric repeats.
We next determined the pattern of replication in a non-erythroid cell line [Epstein-Barr virus (EBV)-transformed B lymphocytes]. The overall pattern is remarkably similar to that observed in the MEL×16 hybrids (Fig. 3 and Table 3). Most of the 325 kb region replicates early in the cell cycle (although somewhat later than in MEL×16 hybrids), but again becomes progressively later replicating towards the telomere. For reasons set out above, it was not possible to analyse the subtelomeric region in these lines.
Table 3. The pattern of replication in non-erythroid cell lines
| Cosmid | TA1 | TA2 | TA only | BO | JG | |||
| % Doublets | % Asynchrony | % Doublets | % Asynchrony | Mean | SD | % Doublets | % Doublets | |
| CW44 | 16 (79) | 21 | 31 (82) | 22 | 20 (89) | 20 | ||
| cJ21 | 23 | 21 | 21 | 22 | 19 | 22 (90) | ||
| cosME2.5 | 30 | 18 | 29 (80) | 27 | 19 (93) | 23 | ||
| cNFG9 | 35 | 28 | 50 (93) | 42 | 42 | 11.0 | 35 | 44 (100) |
| cRA36 | 43 | 48 | 52 (85) | 41 | 47 | 6.0 | 45 | 42 (100) |
| cNFG7 | 40 | 26 | 47 | 28 | 43 | 5.0 | 39 | 48 |
| cGG4 | 56 | 30 | 59 | 26 | 58 | 2.1 | 46 | 51 |
| cGG2 | 49 | 50 | 56 (93) | 39 | 52 | 4.6 | 34 | 51 (95) |
| cRN24 | 46 | 30 | 60 (94) | 39 | 53 | 9.5 | 40 | 49 |
| cGG1 | 54 (90) | 31 | 66 (93) | 26 | 60 | 8.5 | 46 | 55 |
| cos12 | 52 | 36 | 41 | 30 | 47 | 7.8 | 49 | 46 |
| cRJ14 | 67 (95) | 28 | 66 (94) | 27 | 66 | 1.1 | 62 (97) | 69 (100) |
| c310H5 | 64 (88) | 35 | 58 (85) | 34 | 61 | 4.2 | 51 | 68 (93) |
| c314G4 | 70 (89) | 42 | 61 (95) | 31 | 65 | 6.0 | 53 | 62 |
| c419C1 | 71 | 10 | 70 (94) | 30 | 73 | 5.8 | 68 | 63 |
| PYGM | 67 (92) | 36 | 62 (96) | 30 | 57 | 53 (96) |
The temporal pattern of replication in these non-erythroid cell lines, measured by the proportion of duplicated hybridization signals, represents an average for the two homologous chromosomes. However, by documenting the presence of 2:1 signals (Table 3), it appears that many loci replicate asynchronously in individual cells. More extreme examples of such patterns have been observed previously in imprinted regions of the genome. To investigate this further, we analysed two lymphoblastoid cell lines derived from patients who are monosomic for 16p13.3. A maternally derived chromosome is deleted in one case [JG (unpublished data)] and a paternal chromosome is deleted in the other [BO (40)]. The data in Table 3 and Figure 4 show that, on average, most homologous regions replicate at the same time during the cell cycle (scores within 10% of each other). These findings indicate that although homologous regions may replicate asynchronously in some single cells, in a population of cells the temporal pattern of replication timing proceeds in a synchronous manner. The largest differences were seen for the regions spanned by cGG2 (16.5% difference) and c310H5 (17.5% difference) (Fig. 4). It is interesting that, when the maternally and paternally derived chromosomes are added back together (Fig. 4), the replication pattern is very similar to that seen from a normal diploid lymphoblastoid cell line. These data suggest that there may be subtle differences in the replication pattern of maternally and paternally derived chromosomes, but they are consistent with previous genetic (41,42) and methylation (43) data indicating that this region of 16p is not classically imprinted.
Figure 4. Replication characteristics of the individual alleles of 16p. Replication timing of the 325 kb region was measured in cell lines derived from patients monosomic for this region. (A) Results from the cell line containing the paternally inherited copy (JG) are shown as closed squares and the maternally inherited copy (BO) is shown as open circles. Horizontal black dashed lines indicate the mean values of the early (top line) and late (lower line) replicating controls in EBV lines as set out in Figure 3 and its legend. (B) Comparison of the temporal order of replication of the 325 kb region in the normal diploid EBV line (TA), shown as a black dashed line (closed diamonds) and the mean of the two monosomic EBV lines (JG and BO) shown as open squares. Mean values of percentage doublets are shown for the early (top dashed horizontal line) and late replicating (lower dashed horizontal line) control cosmids.
The effect of chromosomal deletions on the pattern of replication
The consistently similar patterns of replication (peaks and troughs; Figs 3 and 4) observed on both homologues of 16p in independently derived erythroid and non-erythroid cell lines suggested that the origin(s) and timing of replication are established and maintained by a stably inherited genetic or epigenetic programme. The pattern is determined by the number and position of origins and the time in the cell cycle when they are activated. Since DNA is replicated at ~2 kb/min (reviewed in refs 19,44), it could take up to 3 h (~40-50% of S-phase) to duplicate the 325 kb region from a single origin lying outside of this region and observe a complete transition from all single to all double hybridization signals. Since we see no major differences in the proportion of doublets in unsynchronized S-phase cells using probes at the centromeric and telomeric ends of the GC-rich isochore, this suggests that the 325 kb region contains at least one origin of replication. This in turn would be consistent with previous studies showing that mammalian replicons are 50-300 kb in size, predicting that this region should contain between one and six origins of replication (4).
If origin(s) are discrete, sequence-dependent elements, deletions that remove them should increase the time required to replicate the chromosomal segment and this would be detectable in the FISH assay. To look for essential, specific zones or origins of initiation within the 325 kb region, we examined the pattern of replication in three previously characterized chromosomes (IC, CL and TAT) from which large segments of DNA had been deleted. The first (CL) is an ~113 kb deletion spanning coordinates ~142 000-255 000, identified in an individual with [alpha]-thalassaemia (45); in effect, it removes the centromeric portion of the 325 kb segment (Fig. 5). Analysis of an MEL×16 hybrid containing this abnormal chromosome (CL2-9) showed that the timing of replication detected by the remaining cosmids was largely unaffected by the deletion (Fig. 5 and Table 4), implying that the deleted segment of DNA does not contain any sequences that are essential for replication of this region.
Figure 5. The pattern and timing of replication in chromosomes with deletions from 16p13.3. In all panels, the black dashed horizontal lines represent the mean values of early (top) and late (bottom) replicating controls. In each right-hand panel, the mean values for early and late replicating controls (as set out in Figure 3 and its legend) are shown, as are individual control values from the cell line studied. (A) Replication profile of the hybrid line CL2-9 (green) and a normal MEL×16 hybrid (black dashed line). (B) Replication profile of the hybrid lines TAT (green), IC (orange) and a normal MEL×16 hybrid (black dashed lines). (C) Replication profile of the hybrid line JJ523 (green) and a normal MEL×16 hybrid (black). (D) Replication profile of the hybrid lines C40 (green), C10 (orange) and a normal MEL×16 hybrid (black dashed lines). (E) The normal chromosome as set out in Figure 1 and, below, the extent of each chromosomal deletion described in the text. The segments of DNA in the transgenic lines (GG1/GG2 and [alpha]PAC) are represented by black lines.
Table 4. The percentage doublets for three MEL×16 hybrids made from chromosomes (CL2-9, IC and TAT) deleted for various segments of 16p13.3
| % Doublets | ||||
| CL2 9 | IC | TAT | Normala | |
| cSam D4: mouse | 20 | 24 | 30 (90) | |
| CNFG9 | 46 | Deleted | Deleted | 56 |
| CRA36 | 58 | Deleted | Deleted | 60 |
| CGG4 | 61 | Deleted | Deleted | 76 |
| CNFG7 | 66 | Deleted | Deleted | 67 |
| CGG2 | 77 | Deleted | Deleted | 68 |
| CRN24 | 60 | Deleted | Deleted | 70 |
| CGG1 | Deleted | 41 (90) | 52 (90) | 72 |
| Cos12 | Deleted | 49 | 64 (93) | 67 |
| CRJ14 | Deleted | 63 (100) | 67 (100) | 80 |
| 310H5 | Deleted | 61 (95) | 69 (96) | 71 |
| 314G4 | 70 | 60 (100) | 70 (95) | 78 |
| 419C1 | 70 | 60 (96) | 71 (93) | 74 |
| P48: mouse | 55 | 57 | 60 (99) |
aMean data from Table 2.
We next analysed two MEL×16 hybrids containing previously described truncated copies of chromosome 16p, originally identified because they are associated with [alpha]-thalassaemia. One (TAT) removes all sequences between coordinates 1 and 142 819, the other (IC) removes sequences between coordinates 1 and 121 194. In both cases, the broken, truncated chromosome has been `healed' by the direct addition of telomeric repeats (46). These deletions effectively remove the telomeric half of the 325 kb segment, including sequences that normally regulate [alpha]-globin gene expression. The pattern of replication timing in most of the remaining GC-rich isochore is early but has lost the characteristic peaks (at ~205 000 and ~270 000) seen in normal chromosomes (Fig. 5B and Table 4). In both cell lines, the most distal (~50 kb) region, in which the [alpha] genes are now juxtaposed to the telomere, replicates later than the corresponding sequence in the normal allele. Again, these findings indicate that the proximal segment of the 325 kb region does not contain any sequences that are essential for replication of this region. However, early replicating sequences moved close to the telomere replicate later than normal in the cell cycle.
The influence of the [alpha]-globin regulatory element on the pattern of replication
It appears that the 325 kb region replicates somewhat earlier in MEL cells than in lymphoblastoid cells (Fig. 3): at present, it is not clear if this is an erythroid-specific phenomenon or a particular feature of MEL cells. In the [beta]-globin cluster (11p15.5), a chromosomal segment far upstream of the [beta] genes containing the [beta]-locus control region ([beta]-LCR) influences transcription (47), the timing of replication (34) and the choice of origins (35) in an erythroid-specific manner. A natural deletion of this upstream region abolishes [beta]-globin expression and has been shown to alter the pattern of replication throughout a >150 kb chromosomal region, including the [beta] cluster, in erythroid (MEL×11 hybrid) cells (34). The [alpha] cluster is regulated similarly by a remote element (33) located 40 kb from the [alpha]-like globin genes (HS -40 in Fig. 1) and, again, natural deletions of this region abolish globin gene expression in vivo (48) and in MEL×16 hybrids (49; reviewed in ref. 30). Here we have examined the influence of HS -40 on replication in erythroid cells.
We first examined the pattern of replication in an MEL×16 hybrid containing a chromosome (JJ523) with an ~35 kb deletion spanning HS -40 derived from a previously described individual (IJ) with [alpha]-thalassaemia (29,50). All remaining sequences appear to replicate early in a similar way to the normal chromosome (Table 5 and Fig. 5). To refine this observation, we also analysed replication in a chromosome (C40) from which an ~1.1 kb segment containing HS -40 was deleted and replaced by a Neo selectable marker using homologous recombination (51). To ensure that the Neo marker did not influence the replication characteristics of the isochore per se, we confirmed the observations (Table 5 and Fig. 5) in a derivative of C40 (C10) in which the Neo marker was removed by FLP recombinase (51). It is clear from these data that, unlike parallel observations on the [beta] cluster (52), removal of the [alpha]-globin regulatory element has no effect on replication within or around the globin locus.
Table 5. The percentage doublets for three MEL×16 hybrids (JJ523, C40 and C10) in which various segments including HS -40 are deleted
| Cosmid | % Doublets | ||
| JJ523 | C40 | C10 | |
| cSam D4: mouse | 31 | 27 (95) | 24 |
| cNFG9 | 47 | 49 (100) | |
| cRA36 | 53 | 51 (100) | |
| cNFG7 | Deleted | 59 (95) | |
| cGG4 | Deleted | 65 (98) | |
| cGG2 | Deleted | 65 (100) | |
| cRN24 | 67 (100) | 79 | |
| cGG1 | 69 | 64 (95) | 80 |
| cos12 | 56 | 56 (85) | 81 |
| cRJ14 | 62 | 86 (100) | |
| c310H5 | 66 | 75 (86) | |
| c314G4 | 70 | 71 (90) | |
| c419C1 | 76 | 74 (100) | 78 |
| P48: mouse | 60 | 53 (96) | 65 (94) |
Replication of segments of 16p13.3 in transgenic mice
Although the pattern of early replication in this region appears to be maintained by a genetic or epigenetic programme, no deletion appears to disturb it in a specific way consistent with the presence of more than one origin of replication in this 325 kb segment. It was therefore of interest to see whether small segments of this region contain functional origins and the sequences to specify early replication. To address this, we examined the pattern of replication in two previously described lines of transgenic mice: one (GG1/GG2) contains 70 kb from this region (53) and the other ([alpha]PAC100) contains 150 kb (30). Both transgenes include the entire human [alpha]-globin locus which is expressed, albeit at suboptimal levels, in a tissue- and developmental stage-specific manner in these mice.
Transgenic lines with a low copy number were selected for analysis to allow unequivocal scoring from a single integration site and to avoid the potential confounding problems associated with multiple tandemly repeated copies. Two low copy number transgenic lines were examined: line [alpha]53 (53) is homozygous for the 70 kb GG1/GG2 construct (Figs 1 and 5) integrated at a single site in band E of mouse chromosome 12, and [alpha]PAC100 (30; unpublished data) contains two tandemly repeated copies of the 150 kb [alpha]PAC construct (Figs 1 and 5), integrated at a single site within bands A3-A5 of mouse chromosome 4. Replication was studied in erythroblasts from the spleens of phenylhydrazine-treated mice.
The results (Table 6) indicate that in both transgenic lines, replication occurs early in S-phase, at a time similar to that of the endogenous mouse [alpha] cluster. Line [alpha]53 was analysed on two occasions: despite differences in the proportions of erythroid cells present, the pattern of replication was similar in both experiments, suggesting that the transgene replicates at similar times in erythroid and non-erythroid cells. It is possible that both constructs (GG1/GG2 and [alpha]PAC100) fortuitously integrated near to a mouse early activating origin and that this is responsible for the pattern of early replication. However, the [alpha]PAC construct is large (150 kb) and is present as two copies (300 kb). Signals from both copies co-localize in interphase nuclei, and the proportion of doublets observed with probes corresponding to the insert was similar to the mouse early replicating control. If this region were replicated from an origin lying outside of the transgenic fragment, it would take at least 3 h to duplicate this region and it would appear to be later replicating than the early controls by the hybridization assay. Therefore, it seems most likely that replication initiates at one or more regions within the 150 kb transgene.
Table 6. Replication timing of transgenes in primary cells extracted from the spleen of phenylhydrazine-treated transgenic mice
| Cosmid | % Doublets | ||
| [alpha]53 A | [alpha]53 B | [alpha]PAC | |
| P48: mousea | 57 | 53 | 43 |
| cGG1 | 60 (100) | 58 | 41, 43 |
| cRN24 | ND | ND | 41 |
| cSam D4: mouseb | 24 | 28 | 19 |
| % erythroid | 60 | 82 | 83 |
ND, probe not represented in this construct.
aEndogenous mouse [alpha] cluster probe (early replicating).
bcSam D4 is the endogenous mouse late replicating control.
The data from these transgenic mice therefore suggest that both constructs contain functional origins or zones of initiation of DNA replication and any associated remote sequences required to establish and maintain the pattern of early replication.
DISCUSSION
To date, there have been few studies to establish the temporal order of replication throughout an extensive segment of the mammalian genome and relate the findings to specific aspects of chromosome structure and function. Here we have shown that the most telomeric region (~20 kb) of the short arm of human chromosome 16 consistently replicates late in S-phase whereas the adjacent GC-rich isochore replicates early. The observed pattern of replication would be most easily explained by the presence of one or more origins in the 325 kb segment, but replication analysis of chromosomes with large, overlapping deletions of this region (CL, IC and TAT) shows that no single element is essential to maintain the normal programme of replication.
The relationship between the pattern of replication and chromatin structure
As previously proposed (19), it seems possible that in mammalian chromosomes chromatin structure plays an important role in defining the choice of origins and possibly when they are activated in S-phase. At present, there is no entirely satisfactory way to define chromatin structure in vivo but, based on the presence of DNase I-hypersensitive sites, unmethylated CpG islands and the distribution of widely expressed genes, we previously suggested that this region contains long tracts of apparently open chromatin surrounded by segments of closed heterochromatin including the telomere and subtelomeric regions (27). Human telomeric regions, like yeast telomeres (54), contain repeated sequences that may be assembled into heterochromatin. This would be consistent with the pattern of late replication observed at the natural 16p telomere and the relatively late replication of chromosomal regions that have been juxtaposed next to new telomeres synthesized to heal broken chromosomes (IC and TAT). Similar chromatin-mediated position effects have been demonstrated in the well-characterized telomeres of Saccharomyces cerevisiae (55). The only previous study of human telomeres, using filter hybridization to DNA from FACS-sorted cells (56), suggested that they replicate at all stages of the cell cycle, possibly by passive extension of replication from the adjacent DNA. This would not explain the pattern observed here since in that case we would have expected the normal, and in particular the abnormal, 16p telomere to replicate early in S-phase.
Other regions may also influence the pattern of replication and chromatid separation because of their chromatin structure. An Alu-rich, densely methylated, DNase I-insensitive segment of chromatin lying downstream of the [alpha] genes (~180 000-218 000) replicated somewhat later than its flanking regions and often remained closely paired throughout S-phase. When juxtaposed next to a structurally normal [alpha] gene by a natural deletion, this region silences expression by exerting a position effect (30), suggesting that, like the telomeric region, it assumes a repressive chromatin structure in vivo.
The remaining segments of the chromosome containing widely expressed, tissue-specific and developmentally regulated genes represent `open' chromatin (27-29). Presumably since many cis-acting transcriptional elements are always accessible in these regions, cis-acting elements regulating replication might be similarly available. Our findings are consistent with this and previous observations suggesting that transcriptionally active regions replicate early in S-phase, although the relationship between the two is not clear. It has been suggested that disruption of chromatin during replication offers a `window of opportunity' in the cell cycle for transcription factors to gain access to their cognate binding sites (e.g. see ref. 57); the timing of replication may define the repertoire of factors available and thus contribute to the programming of gene expression (58). Alternatively, some transcription factors may also activate replication, providing a direct link between the two processes (59). Although the patterns of replication and transcription are tightly coupled for some tissue-specific and developmentally regulated genes, here we show that this is not always the case. Our findings are consistent with preliminary elutriation data suggesting that the human and mouse [alpha]-like globin genes, which are expressed in a strictly tissue- and developmental-stage specific manner, replicate early in both erythroid and non-erythroid cells (60-62). These observations provide a clear example where early replication may be necessary but is not sufficient for gene expression.
Thus the pattern and timing of replication throughout this telomeric region correlate well with the underlying chromatin structure; open chromatin replicating early in the cell cycle and heterochromatic regions being duplicated late in S-phase.
Where does replication initiate in the GC-rich isochore?
The combined data from transgenic experiments, the analysis of normal chromosomes and the findings in chromosomes from which large segments of DNA have been deleted suggest that the 325 kb region contains sequences lying within the region of open chromatin that act as efficient origins of replication in vivo. Which sequences might subserve this function? To date, only ~30 mammalian origins have been identified, but a comparison of six well-characterized origins allowed the identification of a loose consensus (63) which is found in the only mammalian replicator identified so far (15). Allowing one-base mismatches, we found 11 potential sites in the 325 kb region including four within the subtelomeric region which may correspond to origins close to the X and Y subtelomeric elements of S.cerevisiae. Although our data would be consistent with these sites acting as origins, there are no additional data such as co-localization with DNase I-hypersensitive sites, CpG islands or scaffold attachment regions (SARs) to strengthen the observation.
It has been suggested that replication complexes and origins may be fixed to the nuclear scaffold or matrix (64-66). Progress in answering this question has been hampered by the wide range of experimental conditions used to analyse interactions with the matrix (reviewed in ref. 67). Previous analysis of 150 kb around the [alpha]-globin cluster using LIS-extracted nuclei revealed no classical SARs (68), although others have shown that all telomeres are attached to the nuclear matrix via their terminal (TTAGGG)n repeats (69). It is clear that not all well-characterized mammalian replication origins co-localize with SARs and/or matrix attachment regions (70). More consistent results have been obtained recently using an assay in which nuclei are prepared in isotonic conditions, in which it appears that all gene-rich regions interact throughout the cell cycle with matrix-bound proteins that regulate transcription, repair and replication (reviewed in ref. 67).
It is interesting that, of all the origins studied so far, about half are associated with the promoters of genes where transcription complexes assemble, supporting the view that some transcription factors may play a direct role in regulating DNA replication. Recently, using a nascent strand assay, Delgado et al. (71) found that CpG islands, which are associated with the promoters of 50% of all genes, constitute a significant fraction of endogenous, early firing origins of replication. The 325 kb region studied here contains 14 CpG islands, 13 of which are unmethylated in all cells, and each is associated with a constitutive DNase I-hypersensitive site (27,28). It may be that CpG islands in this region play a role in initiating both transcription and DNA replication.
A paradigm for replication in GC-rich and -poor regions of the genome
The analysis of this telomeric region may provide a paradigm for how similar regions of the human genome are replicated, and yet it provides a contrast with others. We previously have pointed out the differences between the chromosomal contexts associated with the human [alpha]- and [beta]-globin clusters (summarized in Table 7) and suggested how this might explain some of the differences in the way in which the two clusters are regulated in vivo. The globin gene clusters may thus provide key examples of how nuclear processes differ between genes located in broadly defined subcompartments of the genome. Whereas the [beta] cluster is contained in a region of closed chromatin that opens in a tissue-specific manner, the [alpha] cluster appears to lie in a region of constitutively open chromatin. These differences may be reflected in many different aspects of chromatin-restricted processes including transcription, repair and recombination (Table 7). This concept is strengthened by the interesting observation that mutations of a transcriptional regulator (ATRX) thought to modify gene expression by a chromatin-mediated effect perturb [alpha] but not [beta] gene expression (72). Here we have shown that such differences also extend to the pattern of replication in and around the globin genes. In the [beta] cluster, replication is initiated normally from an origin containing a well-defined replicator (15), which fires early in erythroid cells but late in non-erythroid cells, correlating with changes in chromatin accessibility (34). When transferred experimentally to a permissive chromatin environment, the [beta]-globin origin fires independently of the [beta]-LCR, suggesting that the [beta] cluster's normal pattern of replication may be determined by chromatin-mediated access to a specific sequence (15). By contrast, although deletions of the [alpha]-globin regulatory element (HS -40) severely reduce [alpha]-globin expression, the surrounding region remains as transcriptionally active, open chromatin (29) and, as we have shown here, the pattern of early replication is unaffected. Therefore, in GC-rich, Alu-rich isochores containing a high density of genes, the synchronous pattern of early replication may be determined by the activation of the many potential origins [possibly CpG islands (71)] made accessible to the replication machinery by the constitutively open chromosomal environment.
Table 7. Differences in the structure and function of the [alpha]- and [beta]-globin gene clusters
| [alpha] Cluster | [beta] Cluster | References | |
| Location | 16p13.3 telomeric | 11p15.5 interstitial | 81,82 |
| GC content | 54% | 39.5% | 27 |
| CpG islands | Common | None | 27 |
| Gene density | High | Low | 27 |
| Alu family repeats | 25% | ~5% | 27 |
| LINE repeats | Rare | Present | 27 |
| Chromatin | Open | Closed->open | 29,52,83,84 |
| Matrix attachment sites | None detected | Common | 68 |
| Effect of hATRX mutations | Present | Absent | 72 |
| Regulatory element | Enhancer | LCR | 33,47 |
| Predominant mutations | Deletions | Point mutations | 30 |
| Evolution of intergenic regions | Rapid | Slow | 85 |
| Expression in transient assays | Enhancer independent | Enhancer dependent | 86,87 |
| Expression in hybrids | Early | Late | 88 |
Conclusion
To date, eukaryotic replication has been best characterized in S.cerevisiae. Although there are clear differences in the mechanisms of replication between yeast and human, the effect of chromosomal deletions, telomere truncations and replication in transgenes described here highlight several important parallels between the two systems (for examples, compare with refs 44,54,73,74). In both organisms, chromosomal context exerts an important effect on the temporal pattern of replication in S-phase. Furthermore, within an accessible region, the choice of regions, although not random, may be flexible within and between individual cells.
MATERIALS AND METHODS
Cell culture and preparation of nuclei
EBV-transformed lymphoblastoid and interspecific somatic cell hybrid cell lines (as described in the text) were grown in RPMI 1640 medium (Sigma, Poole, UK) supplemented with 50 U/ml penicillin (Gibco BRL, Paisley, UK), 50 µg/ml streptomycin (Gibco BRL), 2 mM L-glutamine (Gibco BRL) and 15% (v/v) fetal calf serum (Sigma). Selective pressure for hybrid cells was achieved by the addition of 1 mM adenine (Sigma), 1 mM methotrexate (Lederle Laboratories, Gosport, UK) and 30 µM thymidine (Sigma). Ouabain (0.5 µM; Sigma) was also included in the first 14 days of culture to prevent the background growth of EBV-transformed lymphocytes in preference to mouse-human hybrid cells. Cells were fed every 2-3 days to maintain asynchrony.
For each preparation of nuclei, ~2 × 107 cells were grown to mid-log phase. To identify S-phase cells, BrdU (Sigma) was added to the culture to give a final concentration of 0.1 mM, 90 min prior to harvesting. Cells were then collected by centrifugation, washed once with ice-cold phosphate-buffered saline (PBS) and swollen in hypotonic solution (0.07 M KCl), containing 0.1 µg/ml colcemid (Gibco BRL) for 10-20 min. Cells were then gently resuspended in freshly made fixative (3:1 methanol:glacial acetic acid), and placed at 20°C for 30 min. Three further changes of fixative were performed prior to making a batch of ~30 slides. Using a pasteur pipette, a single drop of suspension was released onto a clean glass slide (Superfrost; BDH, Poole, UK) from ~10 cm above, and laid flat to dry. Slides were examined using phase contrast microscopy (Olympus BH2, ×10 objective, ×12.5 eyepiece) to check for correct cell density, absence of cytoplasm, and large interphase nuclei, mid-grey in colour. Slides were aged by placing in a sealed, opaque slide box containing silica gel desiccant for between 7 days and 3 months. Slides were then hybridized sequentially, each with an individual probe, in as short a time period as possible to reduce artefacts due to ageing. Any slides older than 3 months were discarded.
Prior to FISH, slides prepared from a single fixed preparation were assayed to evaluate the percentage of nuclei in S-phase (BrdU-positive nuclei). Only preparations with >30% S-phase were used. Slides were denatured in 0.07 M NaOH/ethanol (2:5) for 90 s, then blocked in PBS/1% dried skimmed milk for 10 min. Anti-BrdU-fluorescein isothiocyanate (FITC) (Boehringer Mannheim, Lewes, UK) was diluted 1:10 in 4× SSC/5% dried skimmed milk, and 30 µl applied to each slide under a coverslip. Slides were incubated in a moist chamber at 37°C for 1 h, then rinsed in three changes of 4× SSC/0.1% Tween-20, 2 min each, at 37°C. Slides were mounted in Vectashield antifade (Vector, Peterborough, UK) containing 1 µg/ml 4[prime],6-diamidino-2-phenylindole (DAPI) counterstain and coverslips sealed with rubber solution. Slides were viewed under a DAPI filter to assess the total number of nuclei. The same field of view was then scanned through the FITC filter to identify which of these were in S-phase. Approximately 100 nuclei were scored per slide.
Nick translation labelling of probes
A 2 µg aliquot of cosmid or plasmid DNA was added to 4 µl of 10× nick translation buffer [0.5 M Tris, pH 7.5 (BDH), 0.1 M MgSO4 (BDH), 1 mM dithiothreitol (BDH), 0.5 mg/ml bovine serum albumin fraction V (Sigma)], 4 µl of each of the following: 2 mM dATP, 2 mM dGTP, 2 mM dCTP (all Pharmacia, St Albans, UK), diluted 1:30 with sterile distilled H2O, 1 mM DIG-11-UTP (Boehringer Mannheim), 2 µl of 0.5 mM dTTP (Pharmacia), 1 µl of DNA polymerase (10 U/µl; Gibco BRL) and 2 µl of DNase I (7.5 U/µl; Pharmacia), mixed and incubated at 16°C for 1.5 h. Following incubation, 60 µl of TE pH 7.5 was added and unincorporated nucleotides removed by centrifugation of the solution through a G50 Sephadex (Sigma) column, pre-equilibrated with TE pH 7.5. The sources of probes used are set out in the legend to Figure 1 and Table 1 footnotes, and others are described in the text.
FISH
For each slide, 20-50 ng of labelled probe was co-precipitated with 0-5 µg of human Cot-1 DNA, 0-5 µg of mouse Cot-1 DNA (both Gibco BRL) and 5 µg of salmon sperm DNA (Sigma) as carrier, with 2 vol of ethanol. DNA was dried under vacuum (Savant DNA Speed Vac 120) and resuspended in 10 µl of hybridization buffer (50% formamide; Fluka, Watford, UK), 2× SSC, 1% Tween-20 (BDH), 10% dextran sulfate at 37°C for 10 min. DNA probe and competitor were denatured at 70°C for 5 min, and allowed to pre-anneal at 37°C for 15-30 min before applying to the slide. Slides were incubated with 100 µl of 100 µg/ml RNase A (Boehringer Mannheim) under a coverslip in a moist chamber at 37°C for 1 h. Slides were then rinsed quickly in 2× SSC, dehydrated through a 70, 90, 100% alcohol series, 2 min each, and air-dried. To denature the chromosomal DNA, slides were placed in 70% formamide/2× SSC at 70°C, for between 30 s and 2 min, then transferred quickly to ice-cold 70% ethanol for 2 min, then through 90 and 100% ethanol for 2 min each and air-dried. Denatured probe was applied to each slide under a coverslip, sealed with rubber solution, and slides incubated at 37°C overnight. The following day, slides were washed in four changes of 50% formamide/2× SSC, followed by four more changes of 2× SSC, each change for 3 min at 45°C. Stringent washing was carried out using three changes of 0.1× SSC, 4 min each, at 60°C, before transferring to wash solution (4× SSC/0.1% Tween-20). Following a 10 min incubation in blocking solution (4× SSC/5% dried skimmed milk), slides were treated with two antibody layers. For each layer, slides were incubated in a moist chamber at 37°C for 20 min. Antibodies were diluted in blocking solution and 30 µl applied per slide, under a coverslip. Layer 1, 1:10 anti-BdU (Boehringer Mannheim)/1:50 sheep anti-DIG-FITC conjugate (Boehringer Mannheim); layer 2, 1:100 rabbit anti-sheep-FITC conjugate (Vector)/1:100 anti-mouse-Texas red (Vector). Between layers 1 and 2, and after layer 2, slides were washed in four changes of 4× SSC/0.1% Tween-20, 3 min each, at 37°C. Finally, slides were mounted in antifade Vectashield (Vector)/1 µg/ml DAPI counterstain (Sigma) and coverslips sealed with rubber solution. Slides were stored at 4°C, in the dark until ready to view.
Slide scoring
Hybridized slides were examined using a fluorescence microscope (Olympus BX60) equipped with a Pinkel filter wheel containing a DAPI, Texas red, FITC, dual (FITC and Texas red) and triple filters. Images were captured using a Photometrics cooled CCD camera and analysed using MacProbe version 3.3 software (Perceptive Scientific International, Chester, UK). Scoring was carried out using the ×100 oil eyepiece.
Slides of overall poor quality were not scored. Poor quality was due to: (i) a weak hybridization signal; (ii) high non-specific background within the nuclei, which made the true signal difficult to identify; or (iii) high non-specific background over the whole of the slide. Slides were also discarded if identifying nuclei in S-phase (BrdU-positive nuclei) was difficult due to a weak fluorescent signal. Nuclei with no signal or too many signals (due to either background or aneuploidy) were not scored; neither were nuclei with abnormal morphology.
Each slide of adequate quality was scored `blind', the scorer being unaware of with which cosmid the slide had been hybridized. Slides were scanned under the dual filter to identify S-phase nuclei (Texas red) with cosmid signal (FITC). Cosmid signal was then confirmed further as either a singlet or doublet under the single FITC filter. A total of 200 chromosomes were scored per slide, with each slide taking ~1 h to score. As an internal control for each slide, hybridization efficiency was assessed by scoring at least 10 metaphase spreads, as each chromosome should display two signals, one per sister chromatid. In diploid cell lines, viewing non-S-phase nuclei was also used as an indication of hybridization efficiency as hybridization patterns should show a lack of asynchrony: G1 nuclei a 1:1 pattern and G2 nuclei a 2:2 pattern.
Timing of replication was measured relative to cosmids or plasmids which are specific to regions which are known to be early or late replicating (Table 1).
ACKNOWLEDGEMENTS
We are grateful to Drs L. Kearney, V. Buckle and W. Bickmore for helpful advice in establishing the FISH protocol. We are also grateful to Dr W.G. Wood, Ms J. Sharpe, Ms S. Butler and Ms J. Sloane-Stanley for providing material from transgenic mice and cell lines for analysis. W.G. Wood also provided helpful support and suggestions throughout. We thank many current laboratory members for helpful comments on the manuscript.
REFERENCES
*To whom correspondence should be addressed. Tel: +44 1865 222 393; Fax: +44 1865 222 500; Email: drhiggs{at}worf.molbiol.ox.ac.uk
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Garrick, M. De Gobbi, V. Samara, M. Rugless, M. Holland, H. Ayyub, K. Lower, J. Sloane-Stanley, N. Gray, C. Koch, et al. The role of the polycomb complex in silencing {alpha}-globin gene expression in nonerythroid cells Blood, November 1, 2008; 112(9): 3889 - 3899. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Brown, J. Green, R. P. das Neves, H. A.C. Wallace, A. J.H. Smith, J. Hughes, N. Gray, S. Taylor, W. G. Wood, D. R. Higgs, et al. Association between active genes occurs at nuclear speckles and is modulated by chromatin environment J. Cell Biol., September 22, 2008; 182(6): 1083 - 1097. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Brown, J. Leach, J. E. Reittie, A. Atzberger, J. Lee-Prudhoe, W. G. Wood, D. R. Higgs, F. J. Iborra, and V. J. Buckle Coregulated human globin genes are frequently in spatial proximity when active J. Cell Biol., January 17, 2006; 172(2): 177 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. R. E. Williams, V. Azuara, P. Perry, S. Sauer, M. Dvorkina, H. Jorgensen, J. Roix, P. McQueen, T. Misteli, M. Merkenschlager, et al. Neural induction promotes large-scale chromatin reorganisation of the Mash1 locus J. Cell Sci., January 1, 2006; 119(1): 132 - 140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Hughes, J.-F. Cheng, N. Ventress, S. Prabhakar, K. Clark, E. Anguita, M. De Gobbi, P. de Jong, E. Rubin, and D. R. Higgs Annotation of cis-regulatory elements by identification, subclassification, and functional assessment of multispecies conserved sequences PNAS, July 12, 2005; 102(28): 9830 - 9835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Cohen, T. Dagan, L. Stone, and D. Graur GC Composition of the Human Genome: In Search of Isochores Mol. Biol. Evol., May 1, 2005; 22(5): 1260 - 1272. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. P. Steensma, R. J. Gibbons, and D. R. Higgs Acquired {alpha}-thalassemia in association with myelodysplastic syndrome and other hematologic malignancies Blood, January 15, 2005; 105(2): 443 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Wang and J. Zhu The hTERT Gene Is Embedded in a Nuclease-resistant Chromatin Domain J. Biol. Chem., December 31, 2004; 279(53): 55401 - 55410. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. White, O. Emanuelsson, D. Scalzo, T. Royce, S. Kosak, E. J. Oakeley, S. Weissman, M. Gerstein, M. Groudine, M. Snyder, et al. DNA replication-timing analysis of human chromosome 22 at high resolution and different developmental states PNAS, December 21, 2004; 101(51): 17771 - 17776. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-A. Deville, R. Ouazana, F. Morle, and A. Bernet Disruption of the Mechanism of Long Range Activation within the Human {alpha}-Globin Complex J. Biol. Chem., May 21, 2004; 279(21): 21793 - 21801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Gomez and N. Brockdorff Heterochromatin on the inactive X chromosome delays replication timing without affecting origin usage PNAS, May 4, 2004; 101(18): 6923 - 6928. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. P. Steensma, V. Viprakasit, A. Hendrick, D. K. Goff, J. Leach, R. J. Gibbons, and D. R. Higgs Deletion of the {alpha}-globin gene cluster as a cause of acquired {alpha}-thalassemia in myelodysplastic syndrome Blood, February 15, 2004; 103(4): 1518 - 1520. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Viprakasit, R. J. Gibbons, B. C. Broughton, J. L. Tolmie, D. Brown, P. Lunt, R. M. Winter, S. Marinoni, M. Stefanini, L. Brueton, et al. Mutations in the general transcription factor TFIIH result in {beta}-thalassaemia in individuals with trichothiodystrophy Hum. Mol. Genet., November 1, 2001; 10(24): 2797 - 2802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Li, J. Chen, M. Izumi, M. C. Butler, S. M. Keezer, and D. M. Gilbert The replication timing program of the Chinese hamster {beta}-globin locus is established coincident with its repositioning near peripheral heterochromatin in early G1 phase J. Cell Biol., July 23, 2001; 154(2): 283 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Daniels, J. F. Peden, C. Lloyd, S. W. Horsley, K. Clark, C. Tufarelli, L. Kearney, V. J. Buckle, N. A. Doggett, J. Flint, et al. Sequence, structure and pathology of the fully annotated terminal 2 Mb of the short arm of human chromosome 16 Hum. Mol. Genet., February 1, 2001; 10(4): 339 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Flint, C. Tufarelli, J. Peden, K. Clark, R. J. Daniels, R. Hardison, W. Miller, S. Philipsen, K. C. Tan-Un, T. McMorrow, et al. Comparative genome analysis delimits a chromosomal domain and identifies key regulatory elements in the {{alpha}} globin cluster Hum. Mol. Genet., February 1, 2001; 10(4): 371 - 382. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Cimbora, D. Schübeler, A. Reik, J. Hamilton, C. Francastel, E. M. Epner, and M. Groudine Long-Distance Control of Origin Choice and Replication Timing in the Human beta -Globin Locus Are Independent of the Locus Control Region Mol. Cell. Biol., August 1, 2000; 20(15): 5581 - 5591. [Abstract] [Full Text]
|
|
|
|
|
|
V. M. Barbour, C. Tufarelli, J. A. Sharpe, Z. E. Smith, H. Ayyub, C. A. Heinlein, J. Sloane-Stanley, K. Indrak, W. G. Wood, and D. R. Higgs alpha -Thalassemia resulting from a negative chromosomal position effect Blood, August 1, 2000; 96(3): 800 - 807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. L. McDowell, R. J. Gibbons, H. Sutherland, D. M. O'Rourke, W. A. Bickmore, A. Pombo, H. Turley, K. Gatter, D. J. Picketts, V. J. Buckle, et al. Localization of a putative transcriptional regulator (ATRX) at pericentromeric heterochromatin and the short arms of acrocentric chromosomes PNAS, November 23, 1999; 96(24): 13983 - 13988. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Ofir, A. C. C. Wong, H. E. McDermid, K. L. Skorecki, and S. Selig Position effect of human telomeric repeats on replication timing PNAS, September 28, 1999; 96(20): 11434 - 11439. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Esperet, S. Sabatier, M.-A. Deville, R. Ouazana, E. E. Bouhassira, J. Godet, F. Morle, and A. Bernet Non-erythroid Genes Inserted on Either Side of Human HS-40 Impair the Activation of Its Natural alpha -Globin Gene Targets without Being Themselves Preferentially Activated J. Biol. Chem., August 11, 2000; 275(33): 25831 - 25839. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Anglana and M. Debatisse Dual Control of Replication Timing. STOCHASTIC ONSET BUT PROGRAMMED COMPLETION OF MAMMALIAN CHROMOSOME DUPLICATION J. Biol. Chem., September 21, 2001; 276(39): 36639 - 36646. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Anguita, C. A. Johnson, W. G. Wood, B. M. Turner, and D. R. Higgs Identification of a conserved erythroid specific domain of histone acetylation across the alpha -globin gene cluster PNAS, October 9, 2001; 98(21): 12114 - 12119. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||