Replicon origins in Chinese hamster cell DNA. II. Reproducibility

Eukaryotic DNA replication

One of the fundamental characteristics of life is the ability of an entity to reproduce itself, which stems from the ability of the DNA molecule to replicate itself. The initiation step of DNA replication, where control over the timing and frequency of replication is exerted, is poorly understood in eukaryotes in general, and in mammalian cells in particular. The cis-acting DNA element defining the position and providing control over initiation is the replication origin. The activation of replication origins seems to be dependent on the presence of both a particular sequence and of structural determinants. In the past few years, the development of new methods for identification and mapping of origins of DNA replication has allowed some understanding of the fundamental elements that control the replication process. This review summarizes some of the major findings of this century, regarding the mechanism of DNA replication, emphasizing what is known about the replication of mammalian DNA. J. Cell. Biochem. Suppls. 32/33:1-14, 1999.

The spatial position and replication timing of chromosomal domains are both established in early G1 phase

Mammalian chromosomal domains replicate at defined, developmentally regulated times during S phase. The positions of these domains in Chinese hamster nuclei were established within 1 hr after nuclear envelope formation and maintained thereafter. When G1 phase nuclei were incubated in Xenopus egg extracts, domains were replicated in the proper temporal order with nuclei isolated after spatial repositioning, but not with nuclei isolated prior to repositioning. Mcm2 was bound both to early- and late-replicating chromatin domains prior to this transition whereas specification of the dihydrofolate reductase replication origin took place several hours thereafter. These results identify an early G1 phase point at which replication timing is determined and demonstrate a provocative temporal coincidence between the establishment of nuclear position and replication timing.

Origins of replication and the nuclear matrix: the DHFR domain as a paradigm

The eukaryotic genome appears to be organized in a loopwise fashion by periodic attachment to the nuclear matrix. The proposal that a chromatin loop corresponds to a functional domain has stirred interest in the properties of the DNA sequences at the bases of these loops, the matrix-attached regions (MARs). Evidence has been presented suggesting that certain MARs act as boundary elements isolating domains from their chromosomal context. MARs have also been found in the vicinity of promoters and enhancers and they could act by displacing these cis-regulatory elements into the proper nuclear subcompartment. Attachment to the matrix might also play a role in DNA replication. A large body of evidence indicates that replication occurs on the nuclear matrix. This implies that any DNA sequence will be attached to the matrix at a certain time during the cell cycle. This transient mode of attachment contrasts with the proposed permanent attachment of origins of DNA replication with the nuclear matrix. While some data exist that support this suggestion, the current lack of understanding of the mammalian replication origin precludes definitive conclusions regarding the role of MARs in the initiation process.

Searching for replication origins in mammalian DNA

The attempts at identifying precise replication origins (ori) in mammalian DNA have been pursued mainly through physico-chemical and biochemical approaches, in view of the essential failure of the search for autonomously replicating sequences in cultured cells. These approaches involve the mapping of short stretches of nascent DNA, the identification of the regions where either leading or lagging strands switch polarity, or the localization of replication intermediates by two-dimensional gel electrophoresis. Due to the complexity of animal cell genomes, most of these studies have been performed on amplified domains and with the use of synchronization procedures. The results obtained have been controversial. In order to avoid the use of experimental procedures potentially affecting the physiological mechanism of DNA replication, we have developed a method for the localization of ori in single-copy loci in exponentially growing cells. This method entails the absolute quantification of the abundance of selected DNA fragments along a genomic region within samples of newly synthesized DNA by competitive polymerase chain reaction (PCR); the latter is immune to all the uncontrollable variables which severely affect the reproducibility of conventional PCR. The application of this method to SV40 ori-driven plasmid replication precisely identifies the known ori localization. Using the same approach, we have mapped an ori for bi-directional DNA replication in a 13.7-kb locus of human chromosome 19 encoding lamin B2.

Mammalian origins of replication

It has been almost twenty-five years since Huberman and Riggs first showed that there are multiple bidirectional origins of replication scattered at approximately 100 kb intervals along mammalian chromosomal fibers. Since that time, every conceivable physical property unique to replicating DNA has been taken advantage of to determine whether origins of replication are defined sequence elements, as they are in microorganisms. The most thoroughly studied mammalian locus to date is the dihydrofolate reductase domain of Chinese hamster cells, which will be used as a model to discuss the various methods of investigation. While several laboratories agree on the rough location of the 'initiation locus' in this large chromosomal domain, different experimental approaches paint different pictures of the mechanism by which initiation occurs. However, a variety of new techniques and synchronizing agents promises to clarify the picture for this particular locus, and to provide the means for identifying and isolating other origins of replication for comparison.

DNA cruciforms and the nuclear supporting structure

Cruciforms have been suggested as potential recognition structures at or near origins of DNA replication in eukaryotic cells. Monoclonal antibodies with structural specificity for DNA cruciforms have been produced (Frappier et al. J. Mol. Biol. 193, 751, 1987). The effect of these antibodies, when introduced into permeabilized cells, was to increase overall DNA synthesis and relative copy number of genes (Zannis-Hadjopoulos et al. EMBO J. 7, 1837, 1988); this was interpreted to be a consequence of antibody stabilization of the cruciforms located at or near replication origins resulting in multiple initiations of DNA replication at a single site. Fluorescent labeling of nuclei with anti-cruciform antibodies produces a nonuniform pattern of fluorescence in cells arrested at the G1/S boundary which then changes with progression through S-phase (Ward et al. Exp. Cell Res. 188, 235, 1990). In order to determine the relationship of cruciform distribution in DNA with the nuclear matrix/chromosomal scaffold, we assessed the susceptibility of DNA containing cruciforms to digestion with DNase I. The majority of the cruciforms detectable at G1/S and throughout the nucleus are readily digested by DNase, suggesting that cruciform structures may not be intimately associated with matrix proteins. The fraction that is resistant to DNase I appears associated with nuclear membrane and the nucleolus. No cruciforms could be detected in metaphase chromosomes; cruciforms either are not present or are inaccessible--buried in the scaffold. The absence of cruciforms from metaphase chromosomes would be consistent with the viewpoint that the cruciform in vivo is a transient structure dependent upon and interacting with proteins essential for replication or transcription.

The dynamic distribution and quantification of DNA cruciforms in eukaryotic nuclei

Cruciforms have been suggested as potential recognition structures at or near origins of DNA replication in eukaryotic cells. Monoclonal antibodies specific for cruciforms have been produced. The antibody binds to structural determinants at the base of the cruciform stem, the "elbow." Labeling of nuclei with anti-cruciform antibodies produces a nonuniform pattern of fluorescence in cells arrested at the G1/S boundary. This pattern of fluorescence changes when these cells are released from synchrony. Using fluorescence flow cytometry to quantify the number of DNA cruciform structures in cells throughout the cell cycle, we observed two major populations of nuclei with different numbers of cruciforms; the modal number of cruciforms in these populations was 0.6 x 10(5) and 3 x 10(5) cruciforms per nucleus. Synchronized cells (doubly arrested by serum starvation and aphidicolin) displayed a biphasic distribution of the number of cruciforms over the first 6 h after release from synchrony with maxima at 0 and 4 h after release.

Evidence for contamination of origin DNA isolated after an in vivo treatment of mammalian cells with 4,5',8-trimethylpsoralen

We have studied the effect of in vivo treatment with trioxsalen on DNA replication in mammalian cells. In vitro cultured bovine liver cells were exposed to two or four cycles of treatment with 45 microM trioxsalen followed by irradiation with long-wave ultraviolet light. Thymidine incorporation was reduced by about 95% during the first hour after a double treatment. A large proportion of the label was released in alkaline sucrose gradients as a low molecular weight fraction (average length about 500 nucleotides) which was supposed to consist of replication origins containing DNA fragments. From the relative quantities of this DNA obtained at different times of the S phase we concluded that it contains a considerable but not precisely determinable proportion of non-origin DNA. We also find that the fraction is contaminated by a large excess of non-replicating bulk DNA.

The relationship between chromosomal origins of replication and the nuclear matrix during the cell cycle

A cytological investigation into the dynamic behaviour of the origins of replication with respect to the nuclear matrix has been carried out on Xenopus laevis cultured cells. In order to preferentially label origins or 'non-origin' regions along DNA fibres, 5-fluoro-2'-deoxyuridine (FUdR)-treated cells were pulsed with [3H]deoxyadenosine in early or late S phase. Samples were then allowed to proceed through the cell cycle for increasing times. The DNA loops were induced in situ to completely uncoil around the nuclear matrix. The autoradiographic analysis shows that, under the experimental conditions used, 'non-origin' regions behave as expected from previous studies, i.e., they associate with the nuclear matrix only when they become part of a replication fork, whereas active origins of replication remain associated with the matrix throughout the cell cycle.

Newly-initiated DNA isolated from Physarum in early S phase consists of nascent-nascent duplexes

We have studied the nature of newly initiated DNA released during DNA isolation at the beginning of S phase of Physarum polycephalum. The released DNA was separated from the bulk DNA by sedimentation through sucrose gradients. Gentle shearing strongly enhanced the release of newly initiated DNA. The additionally released material had a larger average molecular weight. Buoyant density analysis after labelling with bromodeoxyuridine (BrdU) revealed that the released DNA consisted of nascent-nascent duplexes for more than 90%. This indicates that the release of newly initiated DNA occurs by branch migration. We conclude that shearing enhances branch migration by destabilization of the double helix.

Organization of DNA replication in Physarum polycephalum. Attachment of origins of replicons and replication forks to the nuclear matrix

We have investigated the attachment of the DNA to the nuclear matrix during the division cycle of the plasmodial slime mold Physarum polycephalum. The DNA of plasmodia was pulse labelled at different times during the S phase and the label distribution was studied by graded DNase digestion of the matrix-DNA complexes prepared from nuclei isolated by extraction with 2 M NaCl. Pulse labelled DNA was preferentially recovered from the matrix bound residual DNA at any time of the S phase. Label incorporated at the onset of the S phase remained preferentially associated with the matrix during the G2 phase and the subsequent S phase. The occurrence of the pulse label in the matrix associated DNA regions was transiently elevated at the onset of the subsequent S phase. Label incorporated at the end of the S phase was located at DNA regions which, in the G2 phase, were preferentially released from the matrix by DNase treatment. From the results and previously reported data on the distribution of attachment sites it can be concluded that origins of replicons or DNA sites very close to them are attached to the matrix during the entire nuclear cycle. The data further indicate that initiations of DNA replication occur at the same origins in successive S phases. Replicating DNA is bound to the matrix, in addition, by the replication fork or a region close to it. This binding is loosened after completion of the replication.

Replication of ribosomal DNA in Xenopus laevis

The study of the localization of the replication origins of rDNA in Xenopus laevis has been approached by two different methods. 1. The DNA of X. laevis larvae was fractionated by CsCl gradient centrifugation in bulk and ribosomal DNA and examined in the electron microscope. In bulk DNA, clusters of microbubbles, which are related with the origins of replication, appear to be spaced along the DNA molecules at intervals comparable with the size of the 'average' replicon of X. laevis. In ribosomal DNA, the distance between adjacent clusters is much shorter and corresponds to the size of the rDNA repeating unit. When ribosomal DNA was submitted to digestion with restriction enzymes (Eco RI and HindIII) the microbubbles are observed in the non-transcribed spacer-containing fragment. 2. Cultured cells of X. laevis were synchronized by mitotic selection and incubated with 5-fluoro-2-deoxyuridine for a time longer than the G1 phase. This treatment synchronizes the replicons and allows them to start replicating very slowly. It was thus possible to obtain a preferential labelling of the regions containing the origins. The analysis by gel electrophoresis of the Eco Ri-digested rDNA showed that the radioactivity was preferentially incorporated in the fragments which contain the non-transcribed spacer. The results of these two approaches indicate that the rRNA gene cluster consists of multiple units of replication, possibly one per gene unit. Furthermore they show that the origins of replication are localized into the non-transcribed spacer.

Single-stranded molecules in DNA preparations from cultured mammalian cells at different moments of cell cycle

Long single-stranded DNA molecules have been observed at electron microscope in DNA preparations from synchronized Chinese hamster cells. The amount of single strandedness in parental DNA increases following a prolonged block of DNA synthesis by hydroxyurea as judged by the results obtained using an improved hydroxyapatite chromatography (Hanania et al., 1975). As far as newly replicated DNA is concerned, an increase of the single strand amount has been observed in DNA preparations from cells actively synthesizing DNA.

Variations in the length of S-phase related to the time cells are blocked at the G1-S interface

The inhibition of DNA synthesis with hydroxyurea or 5-fluorode-oxyuridine decreases the duration of S-phase of synchronously growing Chinese hamster cultures. - The observed drug effects are discussed in relation to an alteration of programmed DNA replication.

Lipid-F1 nucleohistone interactions

High concentrations of phospholipids determine destabilization of F1 histone-DNA complex at the weight ratios, histone:DNA, 0.8:1 and 1:1, but low concentrations cause only negligible destabilization. Cholesterol at high weight ratios has little effect on nucleohistone stability. Only linolenic acid of the fatty acids used reproduces similar changes in the thermal stability of F1 histone-DNA complex as phospholipids. The type of interaction of phospholipids with the F1 histone-DNA complex is analyzed, and the involvement of phospholipids in DNA replication in vivo is discussed.

Regulation of DNA replication on subchromosomal units of mammalian cells

The regulation of DNA replication at a subchromosomal level in mammalian cells has been investigated. DNA fiber autoradiographs were prepared from mouse L-929 cells pulse labeled with (3H)thymidine. Initiation events and subsequent chain growth occurring over short stretches (up to three replication units in length) of chromosomal DNA were analyzed. The results show that adjacent units usually initiate replication synchronously and that this synchrony is related to the proximity of initiation sites. In addition, adjacent units are of similar size and the rates of replication fork progression within units and on adjacent units are similar. The rate of fork progression increases with increasing replication unit size. Finally, no evidence for fixed termination sites for the units has been found. These observations suggest that despite large variations in size of replication units, timing of initiation events, and rates of fork progression found in chromosomal DNA as a whole, these processes are closely regulated within subchromosomal clusters of active replication units.