Deoxyribonucleic acid damage by neocarzinostatin chromophore: strand breaks generated by selective oxidation of C-5' of deoxyribose

Among the lesions induced in DNA by neocarzinostatin chromophore are spontaneous and alkali-dependent base release, sugar damage, and single-strand breaks with phosphate (PO4) at their 3' ends and PO4 or nucleoside 5'-aldehyde at the 5' ends. By measuring alkali-dependent thymine release and decomposition of the 5'-terminal thymidine 5'-aldehyde in drug-cut DNA, we show that the kinetics are the same for each process and that the nucleoside aldehyde is the source of about 85% of alkali-dependent thymine release. Reduction of the 5'-aldehyde ends to 5'-hydroxyls followed by incorporation of 32P from [gamma-32P]ATP by polynucleotide kinase permits their selective quantitation. Nucleoside 5'-aldehyde so measured accounts for over 80% of the drug-generated 5' ends; the remainder have PO4 termini. Since these techniques also include the contribution of alkali-labile sites in the measurement of PO4 ends, DNA sequencing was used to measure the ends directly. Using 3'-32P end-labeled DNA restriction fragments as substrates for the drug, it was found that drug attack at a T results in mainly two bands--the stronger one represents oligonucleotide with 5'-terminal nucleoside 5'-aldehyde and may account for over 90% of a particular break. Its structure was verified by its isolation from the sequencing gel, followed by various chemical and enzymatic treatments. In each case, the mobility of the product on the gel was altered in a predictable manner. In addition to spontaneous breaks, neocarzinostatin also causes alkali-labile breaks preferentially at T residues. These sites are heterogeneous in their sensitivity to alkali and are protected by reduction.     

Phosphorylation of Okazaki-like DNA fragments in mammalian cells and role of polyamines in the processing of this DNA.

In mammalian cells DNA synthesis is more complicated than in prokaryotes and less well understood. Here we incubated intact mammalian cells (polyamine auxotrophic Chinese hamster ovary cells and primary human fibroblasts) with [32P]orthophosphate and found that, besides high molecular weight DNA, a species of low molecular weight DNA, approximately 450 bp in size, became efficiently labeled. The short DNA was labeled first, and in pulse-chase experiments the labeling was transient. The isolated small DNA fragments (RNase A-treated) were phosphorylated by T4 polynucleotide kinase specific for polynucleotides with 5'-OH ends. A polynucleotide kinase phosphorylating these DNA pieces was also detected in nuclear extracts of the cells. Treatment with alkaline phosphatase removed most of the 32P label incorporated into the small DNA in vivo. Labeling with deoxyribonucleosides did not reveal these fragments. We hypothesize that the low molecular weight DNA represents Okazaki fragments and that the mammalian DNA replication machinery includes a polynucleotide kinase phosphorylating the 5'-termini of Okazaki fragments. This would imply a novel step in DNA synthesis. We also show that depriving cells of polyamines reversibly blocks synthesis of high molecular weight DNA and leads to accumulation of the short DNA pieces, suggesting a role for polyamines in joining the Okazaki fragments.     

Discontinuous DNA replication of Drosophila melanogaster is primed by octaribonucleotide primer.

To investigate the precise structure of eucaryotic primer RNA made in vivo, short DNA chains isolated from nuclei of Drosophila melanogaster embryos were analyzed. Post-labeling of 5' ends of short DNA chains with polynucleotide kinase and [gamma-32P]ATP revealed that 7% of the DNA fragments were covalently linked with mono- to octaribonucleotide primers at their 5' ends. Octaribonucleotides, the major component (ca. 30%), formed the cap structure in the reaction with vaccinia guanylyltransferase and [alpha-32P]GTP, indicating that they were the intact primer RNA with tri- (or di-) phosphate termini, and the shorter ribooligomers were degradation intermediates. The intact primers started with purine (A/G ratio, 4:1), and the starting few ribonucleotide residues were rich in A.     

RNA-linked nascent DNA pieces in phage T7-infected Escherchia coli. II. Primary structure of the RNA portion.

Short DNA chains were purified from phage T7 infected E. coli cells and 5' ends were labeled with 32P. By an alkali-treatment, pNp's rich in pAp and pCp were liberated from the T7 short DNA chains. After digestion of the [5'-32P] short DNA with the 3' to 5' exonuclease of T4 DNA polymerase, [5'-32P] mono- to pentaribonucleotides tipped with a deoxyribonucleotide residue at their 3' ends were isolated. 5' terminal ribonucleotides were; exclusively AMP in the penta- and the tetraribonucleotides, mostly CMP in the triribonucleotide and mainly CMP and AMP in di- and monoribonucleotides. The 5' terminal dinucleotide of the penta- and the tetraribonucleotides was pApC. The nucleotide sequence of the tetraribonucleotide was mainly pApCpCpN and some pApCpApN, where N was mainly A and C. These results indicate that oligoribonucleotides shorter than trinucleotide may result from in vivo degradation of the tetra- and pentaribonucleotides. A possibility that the tetra- and pentaribonucleotides with a 5' triphosphate terminus are the intact primers for the discontinuous T7 DNA replication is discussed.     

Analysis of the 5'-termini of nascent DNA chains synthesized in permeable cells of Bacillus subtilis

The nascent DNA synthesized by permeable cells of Bacillus subtilis in the presence of 5'-mercurideoxycytidine triphosphate and 2',3'-dideoxyATP has been isolated and characterized. The newly synthesized DNA was isolated free from other cellular nucleic acids by affinity chromatography on thiol-substituted agarose. The number average chain length of the nascent DNA synthesized in one minute at 25 degrees C was 33 nucleotide residues, due to the chain-terminating action of 2',3'-dideoxyATP. Several lines of evidence indicated that at least 90% of the DNA thus isolated carried a terminally phosphorylated RNA moiety at its 5'-end: (1) the nascent DNA was resistant to exonucleolytic degradation by spleen phosphodiesterase unless first hydrolyzed by strong alkali or ribonuclease; (2) the 5'-termini of nascent DNA could not be phosphorylated by polynucleotide kinase unless first treated with alkaline phosphatase or subjected to hydrolysis by strong alkali or ribonuclease; (3) alkaline hydrolysis of nascent DNA labeled with 32P at the 5'-end released unlabeled DNA with a free 5'-terminus and 32P-labeled ribonucleoside 3',5'-bisphosphates; (4) ribonuclease degradation of similarly labeled material produced an unlabeled DNA-containing polynucleotide fraction and 32P-labeled ribo-oligonucleotides; (5) chromatography on dihydroxyboryl cellulose showed that the RNA moiety lacked a 3'-terminal cis-diol grouping (even after treatment with alkaline phosphatase) unless first subjected to the 3'-exonucleolytic action of bacteriophage T4 DNA polymerase. The sequence of the ribonucleotide chains was elucidated by end-group labeling with polynucleotide kinase and digestion with various ribonucleases. The ribonucleotide moiety was primarily three and four residues in length with the predominant sequence (pp)pApG(pC)1-2pDNA. The possibility that it represents a primer for discontinuous DNA synthesis is discussed.     

Analysis of 5'-ends of short DNA fragments excreted by phytohemagglutinin stimulated lymphocytes

Human peripheral blood lymphocytes were stimulated with phytohemagglutinin and the excreted DNA was isolated from the medium after four days of incubation of cells. The excreted DNA was labeled at the 5'-end with [gamma-32P]ATP and polynucleotide kinase. Analysis of the end-labeled material revealed a size distribution with a chain length of 6 - 60 nucleotides. These short DNA fragments did not contain ribo-nucleotides at their 5'-termini. P1 nuclease digestion did not release specific deoxyribonucleoside monophosphates from the 5'-end of the excreted DNA fragments. These results point to the non-specific degradation of DNA excreted by stimulated lymphocytes.     

Neocarzinostatin induction of DNA repair synthesis in HeLa cells and isolated nuclei.

The antitumor antibiotic neocarzinostatin that causes DNA strand breaks in vivo and in vitro is shown to induce DNA repair synthesis in HeLa S3 cells. In the repair assay, the parental DNA was prelabeled with 32P and a density label (bromodeoxyuridine) was introduced into the new synthesized DNA. Quantitation of the repair synthesis as measured by the incorporation of [3H]thymidine into the light parental DNA at varying doses of the drug indicate that there is a significant repair response at low levels of the drug (0.2--0.5 microgram/ml) which cause DNA strand breakage and inhibition of DNA synthesis. In isolated HeLa nuclei neocarzinostatin stimulates the incorporation of dTMP many-fold. This enhancement of dTMP incorporation, which requires the presence of a sulfhydryl agent, is a consequence of the drug-induced DNA strand breakage and is in the parental DNA. These results suggest that an intact cell membrane is not required for DNA strand breakage and its subsequent repair.     

RNA-linked nascent DNA pieces in phage T7-infected Escherichia coli. III. Detection of intact primer RNA

RNA-linked DNA fragments of T7-infected Escherichiacoli were labeled with [(32)P]orthophosphate invivo. The RNA segments of the labeled fragments were isolated by degrading the DNA portion with the 3'--> 5' exonuclease intrinsic to bacteriophage T4 DNA polymerase and fractionated according to net charge by a DEAE-Sephadex A-25 column chromatography in the presence of 7 M urea. Tri-, tetra- and pentanucleotides were obtained which have ATP residues at their 5' ends. Most of the pentanucleotides had a single deoxynucleotide at the 3' end but a minor portion was totally an oligoribonucleotide. In the light of prior results, the former is a cooligomer of an intact tetraribonucleotide primer and a monodeoxynucleotide and the latter is an intact pentaribonucleotide primer. Tri- and tetraribonucleotides with ATP at the 5' ends had no deoxynucleotide at the 3' ends, therefore it is not clear if intact triribonucleotide primers are present. The 5'-terminal dinucleotides of the tetra- and pentanucleotides were mostly pppApC and a trace amount of pppApA was present.Images     

Structure of the RNA portion of the RNA-linked DNA pieces in bacteriophage T4-infected Escherichia coli cells.

A method for the isolation of the RNA portion of RNA-linked DNA fragments has been developed. The method capitalizes on the selective degradation of DNA by the 3′ to 5′ exonuclease associated with bacteriophage T4 DNA polymerase. After hydrolysis of the DNA portion, the RNA of RNA-linked DNA is recovered mostly as RNA tipped with a deoxyribomononucleotide and a small fraction as pure RNA. On the other hand, the 5′ ends of RNA-free DNA are recovered mostly as dinucleotides and a small fraction as mononucleotides.Using this method, we have isolated the primer RNA for T4 phage DNA synthesis. Nascent short DNA pieces were isolated from T4 phage-infected Escherichia coli cells and the 5′ ends of the pieces were dephosphorylated and then phosphorylated with polynucleotide kinase and [γ-³²P]ATP. After selective degradation of the DNA portions, [5′-³²P]oligoribonucleotides (up to pentanucleotide) were obtained with covalently bound deoxymononucleotides at their 3′ ends. More than 40% of the oligoribonucleotides isolated were pentanucleotides with pApC at the 5′-terminal dinucleotide. The 5′-terminal nucleotide of the tetraribonucleotides was AMP, but that of the shorter chains was not unique. The pentanucleotide could represent the intact primer RNA for T4 phage DNA synthesis.     

Discontinuous synthesis of both strands at the growing fork during polyoma DNA replication in vitro.

In discontinuous polyoma DNA replication, the synthesis of Okazaki fragments is primed by RNA. During viral DNA synthesis in nuclei isolated from infected cells, 40% of the nascent short DNA fragments had the polarity of the leading strand which, in theory, could have been synthesized by a continuous mechanism. To rule out that the leading strand fragments were generated by degradation of nascent DNA, they were further characterized. DNA fragments from a segment of the genome which replication forks pass in only one direction were strand separated. The sizes of the fragments from both strands were similar, suggesting that one strand was not specifically degraded. Most important, however, the majority of the Okazaki fragments of both strands were linked to RNA at their 5' ends. For identification, the RNA was labeled at the 5' ends by [beta-32P]GTP, internally by [3H]CTP, [3H]GTP, and [3H]UTP, or at the 3' ends by 32P transfer from adjacent [32P]dTMP residues. All three kinds of labeling indicated that an equal proportion of DNA fragments from the two strands was linked to RNA primers.     

The relationship between DNA strand-scission and DNA synthesis inhibition in HeLa cells treated with neocarzinostatin.

Neocarzinostatin inhibits DNA synthesis in HeLa S3 cells and induces the rapid limited breakage of cellular DNA. The fragmentation of cellular DNA appears to precede the inhibition of DNA synthesis. Cells treated with drug at 37 degrees C for 10 min and then washed free of drug show similar levels of inhibition of DNA synthesis or cell growth, or of strand-scission of DNA as when cells were not washed. If cells are preincubated with neocarzinostatin at 0 degrees C before washing, the subsequent incubation of 37 degrees C results in no inhibition of DNA synthesis or cell growth, or cutting of DNA. Isolated nuclei or cell lysates derived from neocarzinostatin-treated HeLa S3 cells are inhibited in DNA synthesis but this can be overcome in cell lysates by adding activated DNA. A cytoplasmic fraction from drug-treated cells can stimulate DNA synthesis by nuclei isolated from untreated cells, whereas nuclei from drug-treated cells are not stimulated by the cytoplasmic fraction from untreated cells. By contrast, neocarzinostatin does not inhibit DNA synthesis when incubated with isolated nuclei, but it can be shown that under these conditions the DNA is already degraded and is not further fragmented by the drug. These data suggest that the drug's ability to induce breakage of cellular DNA in HeLa S3 cells is an essential aspect of its inhibition of DNA replication and may be responsible for the cytotoxic and growth-inhibiting actions of neocarzinostatin.     

Capped and conserved terminal structures in human rotavirus genome double-stranded RNA segments.

Both 3'- and 5'-terminal structures of human rotavirus genome double-stranded RNA segments were determined. RNAs were labeled at the 3'-termini with [32P]pCp by incubation with RNA ligase and at the 5'-termini with [32P]phosphate by polynucleotide kinase or, in the case of 5' caps, with 3H by chemical modification with [3H]NaBH4. Examination of radiolabeled termini released by digestion with several base-specific RNases revealed that rotavirus RNA segments are base paired end-to-end and contain the same terminal structures: (formula; see text)     

Analysis of products formed during bleomycin-mediated DNA degradation

By the use of DNA, copolymers of defined nucleotide composition, and a synthetic dodecanucleotide having putative bleomycin cleavage sites in proximity to the 5'- and 3'-termini, the products formed concomitant with DNA strand scission have been isolated and subjected to structural identification and quantitation via direct comparison with authentic synthetic samples. The products of DNA strand scission by Fe(II)-bleomycin include oligonucleotides having each of the four possible nucleoside 3'-(phosphoro-2'-O-glycolates) at their 3'-termini, as well as the four possible base propenals. At least for 3-(adenin-9'-yl)propenal and 3-(thymin-1'-yl)propenal, the products formed were exclusively of the trans configuration.     

Replication of the linear mitochondrial DNA of Tetrahymena pyriformis.

1. Electron micrographs of the linear mtDNA from Tetrahymena pyriformis strain GL show linear molecules with a duplex 'eye' of variable size in the middle. This indicates that replication of this DNA starts near the middle of the molecule and proceeds bidirectionally to the ends, as previously shown for the mtDNA of strain ST (Arnberg, A.C., Van Bruggen, E.F.J., Clegg, R.A., Upholt, W.B. and Borst, P. (1974) Biochim. Biophys. Acta 361, 266-276). The mtDNAs of these two strains have little base sequence homology beyond the ribosomal RNA cistron (Goldbach, R.W., Bollen-De Boer, J.E., Van Bruggen, E.F.J. and Borst, P. (1978) Biochim. Biophys. Acta 521, 187-197). 2. Electron micrographs of mtDNA from strain ST, spread under non-denaturing conditions, contain only molecules with fully duplex ends. mtDNA spread under conditions of early denaturation contains duplex loops on one end (40% of all molecules) or both ends (37%). The loops are stable to partial denaturation and vary in size from 0.15 to approximately 1.0 micron, most loops measuring 0.25--0.40 micron. No loops are formed with single-stranded DNA under analogous conditions and we conclude from this result that loop formation is based on the presence of straight, rather than inverted, duplications near the ends. 3. When full-length 3H-labelled mtDNA from strain ST, 32P-labelled at the 5'-termini with T4 polynucleotide kinase, was sedimented in alkaline sucrose gradients, greater than 70% of the 3H and less than 30% of the 32P cosedimented with full-length molecules; the remaining 32P sedimented heterogeneously and predominantly with the DNA less than 10% the size of intact single strands. Brief incubations of full-length mtDNA with DNA polymerase I from Escherichia coli and labelled dNTPs at 15 degrees C did not lead to preferential labelling of terminal EcoRI fragments of the DNA. From these results we infer that the DNA contains nicks or gaps near the termini and that these are not bordered by free 3'-OH groups. 4. A model is presented in which straight sequence repetitions at the termini of Tetrahymena pyriformis mtDNA are involved in the later stages of replication. This model can also account for the pronounced terminal heterogeneity previously observed in this DNA.     

A method for the enzymatic synthesis and purification of [alpha-32P] nucleoside triphosphates.

A simplified method is described for the enzymatic synthesis and purification of [alpha-32P]ribo- and deoxyribonucleoside triphosphates. The products are obtained at greater than 97% radiochemical purity with yields of 50--70% (relative to 32Pi) by a two-step elution from DEAE-Sephadex. All reactions are done in one vessel as there is no need for intermediate product purifications. This method is therefore suitable for the synthesis of these radioactive compounds on a relatively large scale. The sequential steps of the method involve first the synthesis of [gamma-32P]ATP and the subsequent phosphorylation of nucleoside 3' monophosphate with T4 polynucleotide kinase to yield nucleoside 3', [5'-32P]diphosphate. Hexokinase is used after the T4 reaction to remove any remaining [gamma-32P]ATP. Nucleoside 3',[5'-32P]diphosphate is treated with nuclease P-1 to produce the nucleoside [5'-32P]monophosphate which is phosphorylated to the [alpha-32P]nucleoside triphosphate with pyruvate kinase and nucleoside monophosphate kinase. Adenosine triphosphate used as the phosphate donor for [alpha-32P]deoxynucleoside triphosphate syntheses is readily removed in a second purification step involving affinity chromatography on boronate-polyacrylamide. [alpha-32P]Ribonucleoside triphosphates can be similarly purified when deoxyadenosine triphosphate is used as the phosphate donor.     

Characteristics of the nascent and non-nascent small DNA molecules found in Escherichia coli

We have purified nascent DNA molecules from Escherichia coli pulse-labeled with 5-bromo[6-3H]deoxyuridine by repeated chromatography on nitrocellulose and isopycnic centrifugation in CsCl. The nascent molecules were labeled with 32P either at their 5' ends using polynucleotide kinase or at their 3' ends using terminal transferase. Compared to the non-nascent DNA of normal density, the nascent dense DNA contained a higher proportion of molecules terminated at their 5' ends with ribonucleotides. Exposure of the dense DNA to alkali generated 5' OH termini quantitatively equivalent to the number of molecules bearing 5' ribonucleotides. Experiments designed (1) to detect structures at the 5' ends of phosphatase-treated nascent DNA molecules that caused them to be resistant to hydrolysis by spleen exonuclease or (2) to detect polypeptides that were associated covalently with small DNA molecules and could be iodinated with the Bolton-Hunter reagent did not yield positive results. We conclude that many, if not all, of the intermediates in E. coli DNA replication are initiated with one or more ribonucleotides. The nascent molecules are outnumbered by small non-nascent DNA molecules in the cell, many of which appear to become slightly longer when cells are pulsed with thymidine. Many of the non-nascent DNA molecules behave as if they were self-complementary or crosslinked.     

The 3''-terminal nucleotide sequences of adenovirus types 2 and 5 DNA.

Short nucleotide sequences at the 3'-termini of adenovirus types 2 and 5 DNA have been determined using T4 DNA polymerase as described by P. T. Englund (1972). The terminal sequences of both serotypes appear to be completely identical. Both molecular ends of type 2 as well as of type 5 DNA terminate with the sequence ...pCpC...pGpApTpG3', consistent with the presence of an inverted terminal repetition in adenovirus DNA.     

Chromatin 3''-phosphatase/5''-OH kinase cannot transfer phosphate from 3'' to 5'' across a strand nick in DNA.

Rat liver chromatin contains a 3'-phosphatase/5'-OH kinase which may be involved in the repair of DNA strand breaks limited by 3'-phosphate/5'-OH ends. In order to determine whether the phosphate group can be transferred directly from the 3' to the 5' position, a polynucleotide duplex was synthesized between poly (dA) and oligo (dT) segments which had 3'-[32P]phosphate and 5'-OH ends. The oligo (dT) segments were separated by simple nicks as shown by the ability of T4 DNA ligase to seal the nick after the 3'-phosphate was removed by a phosphatase and the 5' end was phosphorylated with a kinase. The chromatin 3'-phosphatase/5'-OH kinase was unable to transfer phosphate directly from the 3' to the 5' end of the oligo (dT) segments in the original duplex; ATP was needed to phosphorylate the 5'-OH end. It is concluded that the chromatin 3'-phosphatase/5'-OH kinase is unable to convert a 3'-phosphate/5'-OH nick which cannot be repaired by DNA ligase directly into a 3'-OH/5'-phosphate nick which can be repaired by DNA ligase; the chromatin enzyme rather acts in two steps: hydrolysis of the 3'-phosphate followed by ATP-mediated phosphorylation of the 5'-OH end.     

DNA sequence selectivity of guanine-N7 alkylation by nitrogen mustards is preserved in intact cells.

Nitrogen mustard alkylating agents react with isolated DNA in a sequence selective manner, and the substituent attached to the drug reactive group can impose a distinct sequence preference. It is not clear however to what extent the observed DNA sequence preferences are preserved in intact cells. The highly reiterated sequence of human alpha DNA has been used to determine the sites of guanine-N7 alkylation following treatment of cells with three nitrogen mustards, mechlorethamine, uracil mustard and quinacrine mustard, known to react in isolated DNA with distinctly different sequence preferences. Alpha DNA from drug treated cells was extracted, purified, end-labeled, and a 296 base pair, singly end-labelled, fragment isolated. Following the quantitative conversion of alkylation sites to strand breaks the fragments were separated on DNA sequencing gels. Clear differences were observed between the alkylation patterns of the three compounds, and the selectivities were qualitatively similar to those predicted and observed in the same sequence alkylated in vitro. In particular the unique preferences of uracil and quinacrine mustards for 5'-PyGC-3' and 5'-GT/GPu-3' sequences, respectively, were preserved in intact cells suggesting that the pattern of sequence dependent reactivity is not grossly affected by the nuclear milieu.