Deoxyribonucleic acid replication in synchronized cultured mammalian cells. I. Time of synthesis of molecules of different average uanine + cytosine content

ORS12, a mammalian autonomously replicating DNA sequence, is present at the centromere of CV-1 cell chromosomes

ors12, an 812-bp-long sequence, previously isolated by extrusion of nascent DNA from replication bubbles active at the onset of S phase (G. Kaufmann, M. Zannis-Hadzopoulus, and R. G. Martin Mol. Cell. Biol. 5, 721-727, 1985), has been shown to function as an origin of DNA replication in autonomously replicating plasmids (L. Frappier and M. Zannis-Hadjopoulos Proc. Natl. Acad. Sci. USA 84, 6668-6672, 1987) and in a cell-free system (C. E. Pearson, L. Frappier, and M. Zannis-Hadzopoulos Biochim. Biophys. Acta 1090, 156-166, 1991). A portion of ors12 (nucleotides 1-168) consists of the highly reiterated alpha-satellite sequence (B. S. Rao et al. Gene 87, 233-242, 1990). We have estimated the copy number of the non-alpha-satellite portion of ors12 in CV-1 cells to be < 9 copies per haploid genome and have used it as a probe to generate a genomic map of ors12 on CV-1 DNA. In situ hybridization of CV-1 metaphase chromosomes, using a biotinylated probe of the entire ors12 sequence, positively identified the centromeres of all chromosomes. However, when the non-alpha-satellite portion of ors12 was used as a probe, it positively identified the centromeric region of only six chromosomes, namely, B4, C11, D14, D24, E25, and E27, as well as that of a marker chromosome. The results suggest that ors12 represents a centromeric putative replication origin that is present on a subset of CV-1 chromosomes and is activated at the onset of S phase.

Evidence that both G + C rich and G + C poor isochores are replicated early and late in the cell cycle

Since the G + C content of a gene is correlated to that of the isochore in which it resides, and early replicating isochores are thought to be relatively G + C rich, early replicating genes should also be rich in G + C. This hypothesis is tested on a sample of 44 mammalian genes for which replication time data and sequence information are available. Early replicating genes do not appear to be more G + C rich than late replicating genes, instead there is considerable variation in the G + C content of genes replicated during both halves of S phase. These results show that both G + C rich and poor fractions of the genome are replicated early and late in the cell cycle, and suggest that isochores are not maintained by the replication of DNA sequences in compositionally biased free nucleotide pools.

Bovine papilloma virus plasmids replicate randomly in mouse fibroblasts throughout S phase of the cell cycle

Bovine papilloma virus (BPV) replicates as a multicopy nuclear plasmid in mouse fibroblasts. Using fluorescence activated cell sorting and mitotic selection procedures, we show that the replication of BPV occurs throughout S phase of the cell cycle and that replication is confined to S phase. After one round of chromosomal DNA replication, almost one quarter of BPV plasmids have replicated more than once, while a similar number of plasmids have not replicated at all. While multiple forms of BPV exist in the cell, all forms show the same pattern of replication. These results are consistent with a model in which BPV plasmids are chosen at random for replication throughout, and only during, S phase and support the view that the completion of S phase is a specifically activated event in the cell cycle rather than simply the end of one round of chromosomal DNA replication.

Correlation between molecular clock ticking, codon usage fidelity of DNA repair, chromosome banding and chromatin compactness in germline cells

The vertebrate genome is built of long DNA regions, relatively homogeneous in GC content, which likely correspond to bands on stained chromosomes. Large differences in composition have been found among DNA regions belonging to the same genome. They are paralleled by differences in codon usage in genes differently localized. The hypothesis presented here asserts that these differences in composition are caused by different mutational bias of alpha and beta DNA polymerases, these polymerases being involved to different extents in the repair of DNA lesions in compact and relaxed chromatin, respectively, in germline cells.

Replication pattern of human repeated DNA sequences

Either aphidicolin- or thymidine-synchronized human HL-60 cells were used to study the replication pattern of a family of human repetitive DNA sequences, the Eco RI 340 bp family (alpha RI-DNA), and of the ladders of fragments generated in total human DNA after digestion with XbaI and HaeIII (alpha satellite sequences). DNAs replicated in early, middle-early, middle-late and late S periods were labelled with BUdR or with [3H]thymidine. The efficiency of the cell synchronization procedure was confirmed by the transition from a high-GC to a high-AT average base composition of the DNA synthesized going from early to late S periods. By hybridizing EcoRI 340 bp repetitive fragments to BUdR-DNAs it was found that this family of sequences is replicated throughout the entire S period. Comparing fluorograph densitometric scans of [3H]DNAs to the scans of ethidium bromide patterns of total HL-60 DNA digested with XbaI and HaeIII, it was observed that DNA synthesized in different S periods is characterized by approximately the same ladder of fragments, while the intensity of each band may vary through the S phase; in particular, the XbaI 2.4 kb fragment becomes undetectable in late S.

Temporal replication of an interspersed repeated sequence of mouse DNA

The temporal replication profile of an interspersed repeated DNA sequence (variously named MIF-1, Bam and L1Md) of mouse was determined by isotope analysis of a resolvable restriction fragment differentially labeled in pre- and post-synchrony cultures. While the temporal replication profile of the fragment was similar to that of total nuclear DNA, an average time lag of about 20 min was evident for this interspersed repeated family (called Bam in this paper). In addition, the sequence organisation of Bam homologues were examined for the separable early- and late-replication domains of the hamster genome. The data suggest that late-replicating domains of the rodent genome are slightly enriched in Bam homologous sequences. Furthermore, this repeated sequence family has different sequence organisations in the separable replication domains of hamster.

Occurrence of short-sized oligo(A) fragments during course of cell cycle and ageing

Affinity chromatography of nucleic acids precipitated by N-cetyl-N,N,N-trimethyl ammonium bromide on poly(U)-Sepharose has proved to be a suitable method for a nearly quantitative isolation of oligo(A) sequences down to a chain length of 4 nucleotide units. Analysis of short oligo(A) fragments in synchronized L5178y mouse lymphoma cells after labeling with [3H]Ado revealed that the percentage of A2-6 sequences on the total radioactivity amounted in S-phase cells to 1.6%, while the value obtained for the stationary L-cell system was 8.0%. The alterations of occurrence and chain length distribution of short oligo(A) fragments during ageing were studied in two age groups of female quails: mature (250-320 days old) and senescent animals (3-3.5 yr old). It was found that the amount of low molecular weight oligo(A) fragments gradually decreases during ageing of the animals; the amount in the mature animal group was significantly higher (6-fold) than in the old animal group. The decreased amounts of oligo(A) during S phase and ageing could in part be due to posttranslational modification of enzymes involved in poly(A) metabolism. It could be demonstrated that both homogeneous poly(A) anabolic poly(A) polymerase and homogeneous poly(A) catabolic endoribonuclease IV are phosphorylated by nuclear protein kinase NI.

G-band-like structures and centromeric asymmetry in the BrdU containing mouse chromosomes

Mouse cell cultured in the presence of BrdU or BrdC for one replication cycle were stained in a 4Na-EDTA Giemsa solution which stains BrdU-containing chromatin preferentially (Takayama and Tachibana, 1980). With this treatment clear bands (B-bands) were revealed along the length of the chromosomes. The B-banding patterns were identical with the G-banding patterns of this species except for the centromeric region in which lateral asymmetry of Giemsa staining was seen. The concomitant occurrence of the lateral asymmetry with the B-banding supports the assumption that the B-bands visualized by the present technique reflect the BrdU-rich chromatin regions differentially localized along the chromosomes. Most of the chromosomes constituting the mouse karyotype showed their own characteristics appearance of the asymmetry, but in some of them the asymmetry was not clear and the Y did not show any specific, centromeric staining. The marked coincidence of the B- and G-banding patterns seems to provide evidence of the involvement of AT-rich chromatin in the induction of positive G-bands. The present technique also seems quite useful to analyze chromosomes of some species in which ordinary G-banding techniques have been known to bring about only unsatisfactory results.

Amplified dihydrofolate reductase genes are located in chromosome regions containing DNA that replicates during the first half of S-phase

To obtain a better understanding of the relationship between metaphase chromosome banding patterns and genome organization, attention was focused on regions of metaphase chromosomes that were found to contain the genes for a specific cellular enzyme, dihydrofolate reductase (DHFR). These studies involved the use of highly methotrexate-resistant mouse lymphoblastoid cells (L5178YR), which contain approximately 300 times the number of DHFR genes present in parental cells (L5178YS). Karyotypic analysis revealed the presence of two very large, nonhomologous, marker chromosomes that were absent in the parental line. In situ hybridization of 3H-labeled cloned DHFR cDNA to metaphase chromosomes of L5178YR cells was used to localize the DHFR genes to a very large Giemsa (G)-negative region on each of the two large marker chromosomes. Regional patterns of DNA replication in metaphase chromosomes were studied by autoradiographic visualization of [3H]thymidine incorporation and by fluorescent microscopic visualization of bromodeoxyuridine incorporation. Because the amplified DHFR genes were present within two prominent cytogenetic regions on two easily identifiable chromosomes, it was possible to observe the following. The amplified DHFR genes were located in chromosome regions that replicated at the same time during the first half of a 9-h-S-phase. DNA replication began simultaneously and terminated simultaneously at many locations throughout each amplified region. We conclude that transcriptionally active DHFR genes are located within large G-negative regions of metaphase chromosomes and that the DNA within these regions replicates during the first half of S-phase.

Satellite versus total DNA replication in relation to endopolyploidy of decidual cells in the mouse

In rodents, decidualization produces large endopolyploid cells. Amongst the various endocycles which have been demonstrated in animals and plants, different modes of DNA replication have been characterized: either total reproduction of all DNA types, or else, underreplication or amplified synthesis affecting specific parts of the genome. A double labelling method was used to determine to which of these categories the case of decidual cells belongs. A mixture of purified DNA from hormonally-stimulated control endometrium labelled by 3H-thymidine and from decidua labelled by 14C-thymidine was ultra-centrifuged to equilibrium in a Cs2 SO4-Ag gradient. Optical density at 260 nm and 14C/3H ratio were evaluated in serial fractions along the gradient. Since the 14C/3H ratio did not significantly vary along the gradient, it may be concluded that in the case of decidual cells, endopolyploidy corresponds to uniform replication of all nuclear DNA.

Replication of alpha and beta globin DNA sequences occurs during early S phase in murine erythroleukemia cells

Murine erythroleukemia cells (MELC) can be induced to express the characteristics of erythroid differentiation by a variety of agents. Previous studies indicate that an action of inducer, occurring during early S phase, may be critical to the expression of differentiated characteristics such as initiation of accumulation of newly synthesized alpha and beta globin mRNAs. In this investigation, the time of replication of globin genes in MELC was studied. DNA was isolated from synchronous populations of cells obtained by centrifugal elutriation. Newly replicated DNA sequences were prepared from synchronized cells cultured for 1 1/2 hr with 5-bromodeoxyuridine; bromodeoxyuridine-containing DNA was isolated by CsCl gradient centrifugation. By employing cloned probes for hybridization to newly synthesized DNA, it was found that alpha and beta globin gene sequences are replicated early in S phase, while ribosomal RNA gene sequences are replicated to about the same extent in early, middle, and late S phases.

Arabinosyl nucleosides. XXXII. Mechanism of inhibition of L5178y mouse lymphoma cells by 9-beta-D-arabinofuranosylguanine

The mechanism of the growth-inhibitory action of 9-beta-D-arabinofuranosylguanine (ara-G) on mouse lymphoma cells (L5178y) at the level of nucleic acid synthesis was studied. Ara-G inhibits the cell proliferation at an ED50 concentration of 12.3 microM. The cells die because of unbalanced growth. At cytostatic concentrations ara-G inhibits the incorporation rate of dThd into DNA to a much higher degree compared to those of the precursors into RNA or proteins. Ara-G has no effect on the induction of DNA polymerase alpha occurring at the beginning of the S-phase; this compound exerts its inhibitory activity during the S-phase. Ara-G is intracellularly phosphorylated to ara-GTP. It was demonstrated that ara-G is incorporated into DNA but not into RNA; one molecule of ara-G is incorporated per 7050 molecules of deoxyguanosine. The ara-GMP moieties, incorporated into DNA of L5178y cells are not removed during a culture period of 7.7 doubling steps; compared to the controls no reduction of the viability and growth rate was observed.

Cytogenetics of the parthenogenetic grasshopper Warramaba (formerly Moraba) virgo and its bisexual relatives. VI. DNA replication patterns of the chromosomes

The distribution of late-replicating segments along the chromosomes of five clones of W. virgo is described. Some, but not all of these segments correspond to C-bands. In general, the "autoradiographic profiles" (histograms of linear grain density along the length of chromosomes labeled with tritiated thymidine in late S-phase) show strong resemblances throughout the five clones. However, some significant differences exist, and these are particularly marked in the case of the Boulder clone, which is anomalous in many other respects. --A similar study has also been carried out on the two bisexual species of Warramaba ("P169" and "P196") that gave rise, by hybridization more than half a million years ago, to the parthenogenetic W. virgo. In the case of P169, the autoradiographic profiles of the three large chromosomes (X + A, B + 5, CD) which it has contributed to the W. virgo karyotype are extremely similar to those of the corresponding chromosomes in the virgo clones we have studied. In the case of P196 there is likewise, in most instances, a close resemblance of the autoradiographic profiles of the AB, X1 and CD chromosomes to those of the same chromosomes in the virgo clones, but that of the X1 shows no particular resemblance to the anomalous profile of the X1 in the Boulder clone, in which the X1 has undergone a structural reorganisation. The autoradiographic profile of the P196 CD chromosome does, however, show a much closer resemblance to that of the corresponding chromosome in the Boulder clone than to those of the CD196 in the other four virgo clones studied. These investigations confirm the considerable evolutionary stability of DNA replication patterns.

Initiation of DNA replication in chromosomes of Chinese hamster ovary cells

The initiation of DNA replication and the subsequent chain elongation were studied using Chinese hamster ovary cells synchronized at the beginning of S phase. The cells were synchronized by a combination of mitotic selection and treatment with 5-fluorodeoxyuridine (FdU). The use of this drug at a concentration of 10(-5) M was found to effectively prevent the leakage of cells into S phase. Reversal of the FdU block by supplying thymidine resulted in the synchronous onset of initiation at multiple sites in each cell. The length of the nascent chains, as determined by autoradiography and velocity sedimentation in alkaline gradients, increased linearly with time during the first twenty minutes of S phase after release. We applied these procedures to study the effects of the length of an FdU block on the number of functional origins per cell, the rate of chain growth, and the rate of DNA synthesis per cell following reversal of the block. Although no change was noted in the rate of DNA synthesis in cells held at the beginning of S phase from 10.5 to 24 h after division, the rate of chain growth decreased from 0.94 to 0.28 microns per min. This decrease indicated that the number of functional origins increased markedly with length of FdU block. The calculated number of utilized origins per cell increased from 1,900 to 5,700. We also presented arguments that 1,900 origins per cell represents the approximate number of origins utilized by any cell held at the beginning of S phase for less than 10.5 h after division.

Adaptation of thymidine utilization to changing rates of DNA synthesis in the cell cycle

In synchronous cultures of P-815 murine mastocytoma and of Chinese hamster ovary (CHO) cells, the relative contribution of exogenous thymidine to DNA synthesis was studied by comparing rates of (3H)thymidine incorporation with the rate of DNA synthesis as derived from incorporation of (3H)thymidine (10(-5) M) in the presence of amethopterin. In synchronous P-815 cultures, time-dependent variations of DNA synthesis rates were in close agreement with those of (3H)thymidine incorporation rates at concentrations of the precursor ranging from 5 x 10(-8) to 10(-5) M. Similarly, in synchronous CHO cell cultures prepared by two different methods, time-dependent changes in DNA synthesis rate were almost identical with those of the rate of incorporation of (3H)thymidine supplied at 5 x 10(-8) M. Thus, at a given thymidine concentration in the medium, the proportion of thymine residues in DNA that were derived from exogenous thymidine remained nearly constant, even though rates of cellular DNA synthesis underwent pronounced changes. This indicates that in the synchronous culture systems used, utilization of exogenous thymidine is efficiently adapted to changing rates of DNA synthesis.

Ordered replication of DNA sequences: synthesis of mouse satellite and adjacent main band sequences

The replication of mouse satellite DNA was delayed when synchronized 3T3 cells were exposed to low concentrations of hydroxyurea during S phase, It appears that the onset of satellite replication is not a time dependent event, but instead requires that a certain amount of main band DNA be synthesized first. Using hydroxyapatite chromatography and S1 nuclease digestion, a procedure was developed to quantitate the synthesis of both satellite and neighboring main band sequences. The replication kinetics of satellite determined by this method agree with previous estimates. Main band sequences adjacent to satellite appear to replicate in concert with satellite DNA. The results are discussed and related to the limitations of the techniques utilized.

Organization, replication and modification of the human genome: temporal order of synthesis and methylation of two classes of HeLa nDNA separated in Ag+--Cs2-SO4 gradients

During the HeLa S-phase, DNA was methylated, at 1-hr intervals in isolated nuclei and fractionated in Ag+-Cs2SO4 gradients providing a heavy GC-rich peak and a main light AT-rich peak. Both size and specific methylation of these peaks changed during the nDNA duplicative phase. Replication of the heavy GC-rich nDNA fraction, which contains genes for ribosomal RNA, occurred in early S; in contrast, replication of the main AT-rich nDNA fraction was maximal in late S. Concomitantly, specific methylation of the GC-rich nDNA was maximal in the first part of S, while that of the AT-rich nDNA was maximal in the second part of S. This suggested that genes are replicated and methylated with order during the S-phase.

Organization, replication and modification of the human genome: differential methylation of two classes of HeLa nuclear DNA separated on Ag+--Cs2SO4 gradients

HeLa nuclear DNA sediments as a single peak, in neutral CsCl, while it is separated in a heavier and a lighter components, in alkaline Ag+--Cs2SO4. The heavy fraction, on the average, represents about 20% of the total DNA. CsCl analytical ultracentrifugation shows that heavy DNA bands at 1.715 g/cm3 and contains 53% GC (10% of the total GC), whereas light DNA bands at 1.703 g/cm3 and contains 40% GC (32% of the total GC). Coherently, Tm values in 0.1 x SSC are 82.5 degrees C, for heavy DNA, and 72.5 degrees C, for light DNA. After treatment with [3H-methyl-S-adenosyl-L-methionine in isolated nuclei, the concentration of labelled 5-methylcytosine was found to be highest in the more dense regions of the heavy peak and in the less dense regions of the light peak. Exposure to ultrasound modifies the quantitative relationship of the two peaks and improves the separation of supermethylated AT- and GC-rich DNAs. Four possible triplets as sites for DNA-methylase recognition are discussed.

The effect of the DNA-suppressing factor (DSF) on host DNA synthesis in synchronized cell cultures

Purified host DNA-suppressing factor (DSF) produced into culture fluid of HeLa C-9 cells infected with measles virus inhibited cellular DNA synthesis in HeLa cells. When purified DSF was added into cultures of synchronous HeLa cells at the early G1-phase, cellular DNA synthesis was irreversibly inhibited. However, DSF did not affect the stability of native double-stranded DNA nor the chain-elongation of single-stranded DNA in cells of the S-phase.

Alterations of activities of ribonucleases and polyadenylate polymerase in synchronized mouse L cells

The activities of the three known catabolic and the one anabolic polyadenylate enzymes have been determined in synchronized L5178y cells: endoribonuclease, exoribonuclease, 5'-nucleotidase and poly(A) polymerase (Mg2+-dependent). These four enzymes were found primarily in the nuclear fraction. The activity of poly(A) polymerase remains essentially constant during the transition from G1 to S phase. However, the poly(A) catabolic enzyme activities increase parallel with DNA synthesis; the endoribonuclease activity increases 4-fold during G1 to S phase, the exoribonuclease and the nucleotidase activities increasing 30-fold and 16-fold. During the S phase the poly(A)-degrading enzymes are far more active than the poly(A)-synthesizing activity of poly(A) polymerase. We conclude that in L5178y cells the poly(A)-degrading enzymes probably function in regulation of the post-transcriptional net-polyadenylation of heterogeneous nuclear RNA during the phase of DNA synthesis.

Replication kinetics of three DNA sequence families in synchronized mouse cells

Balb/C3T3 cells entered the quiescent Go state following serum deprivation. On addition of fresh serum, more than 95% of the culture resumed growth, but with asynchronous kinetics. If hydroxyurea were added just before the first cells reached S phase, at at least 90% of the cells accumulated at the Gl/S border over the next ten hours. When the block was removed, the culture moved synchronously into S phase. As the cells traversed S, the replication kinetics of three classes of rapidly renaturing DNA were analyzed. Main band highly repeated DNA and foldback DNA replicated continuously, In contrast, satellite DNA replication did not commence until three hours into S, whereupon its rate of synthesis increased ver rapidly, reaching a maximum with the next two hours. These results are discussed in the light of earlier work utilizing other methods of cell synchronization.

Mechanisms of chromosome banding. IX. Are variations in DNA base composition adequate to account for quinacrine, Hoechst 33258 and daunomycin banding?

Prior studies on subfractions of mouse and Kangaroo rat DNA have suggested that variations in base concentration within a given genome may not be great enough to account for Q-banding. To examine this with another species, calf DNA was subfractionated by CsCl ultracentrifugation into GC-rich satellites and the main band DNA was further fractionated into AT-rich, intermediate and GC-rich portions. The effect of varying concentrations of these DNAs on quinacrine and Hoechst 33258 fluorescence was examined. Although with both compounds there was less fluorescence in the presence of the GC-rich satellites than main band fractions, these results per se did not answer the question of whether the variation in base composition alone was adequate to account for chromosome banding. To answer this the fluorescence observed in the presence of DNA of a given base composition was related to the fluorescence observed in the presence of DNA of 40% GC content (F/F40). This allowed the derivation of a term B which indicated the relative change in fluorescence per 1% change in base composition of DNA. To determine the percent change in fluorescence observed in Q-banding, the photoelectric recordings of Caspersson et al. (1971) were used. From these data we conclude: 1. Quinacrine is twice as sensitive to changes in base composition as Hoechst 33258. 2. Variation in the base content of DNA along the base content of DNA along the chromosome is sufficient to account for most Q-banding, except possibly for some of the extremes of quinacrine fluorescence. This was further examined with daunomycin. Even though daunomycin gives good fluorescent banding, DNAs varying in base composition from 100 to 40% GC content all resulted in the same relative fluorescence of 0.03. However, in the presence of poly (dA-dT) the relative fluorescence was 0.85, indicating a great sensitivity to very AT-rich DNA. This suggests that with daunomycin and possibly other fluorochromes, stretches of very AT-rich DNA may be more important in fluorescent banding than simple variation in mean base composition.

Is the centromeric heterochromatin of Mus musculus late replicating?

Quantitative analysis of DNA replication in embryonic cultures of C57BL/6J mice was carried out, using autoradiography after tritiated thymidine incorporation. The centromeric regions are late-replicating.

Percent satellite DNA as a function of tissue and age of mice

A selective loss of satellite DNA was found to occur to different extents as a function of tissue and age of mice using several common DNA extraction and purification procedures. This result emphasizes a serious problem that may be encountered in comparative studies of DNA structure and composition if selective loss of specific DNA sequences occurs. We have developed a DNA extraction and purification procedure that is simple and reliable and gives a high percent DNA yield, which substantially reduces the selective loss of heterochromatin DNA sequences. The method features a centrifugation step of a proteolytic digest of chromatin in 2.4 M CsCl. Percent DNA yield of 82-98% are routinely obtained with no apparent loss of satellite DNA sequences from different tissues or ages of mice. Utilizing this method, percent satellite DNA was found to remain essentially constant at 11 +/- 1% for spleen, kidney, and brain tissues obtained from mice of 10-780 days of age. However, for liver, percent satellite DNA remained at about 7-8% from 10 to 300 days of age and then increased to about 12-13% from 300 to 600 days of age. During this latter time interval (300-600 days), an increase of DNA per nucleus of about 3-fold occurred, due to the formation of tetra- and octaploid cell types. A steady loss in the total number of nuclei per gram of liver as a function of age was also found. These two opposing effects resulted in a nearly constant amount of DNA per gram and per organ for liver throughout the lifespan of the mouse.