Conversion of topoisomerase I cleavage complexes on the leading strand of ribosomal DNA into 5'-phosphorylated DNA double-strand breaks by replication runoff - PubMed (original) (raw)

Conversion of topoisomerase I cleavage complexes on the leading strand of ribosomal DNA into 5'-phosphorylated DNA double-strand breaks by replication runoff

D Strumberg et al. Mol Cell Biol. 2000 Jun.

Abstract

Topoisomerase I cleavage complexes can be induced by a variety of DNA damages and by the anticancer drug camptothecin. We have developed a ligation-mediated PCR (LM-PCR) assay to analyze replication-mediated DNA double-strand breaks induced by topoisomerase I cleavage complexes in human colon carcinoma HT29 cells at the nucleotide level. We found that conversion of topoisomerase I cleavage complexes into replication-mediated DNA double-strand breaks was only detectable on the leading strand for DNA synthesis, which suggests an asymmetry in the way that topoisomerase I cleavage complexes are metabolized on the two arms of a replication fork. Extension by Taq DNA polymerase was not required for ligation to the LM-PCR primer, indicating that the 3' DNA ends are extended by DNA polymerase in vivo closely to the 5' ends of the topoisomerase I cleavage complexes. These findings suggest that the replication-mediated DNA double-strand breaks generated at topoisomerase I cleavage sites are produced by replication runoff. We also found that the 5' ends of these DNA double-strand breaks are phosphorylated in vivo, which suggests that a DNA 5' kinase activity acts on the double-strand ends generated by replication runoff. The replication-mediated DNA double-strand breaks were rapidly reversible after cessation of the topoisomerase I cleavage complexes, suggesting the existence of efficient repair pathways for removal of topoisomerase I-DNA covalent adducts in ribosomal DNA.

PubMed Disclaimer

Figures

FIG. 1

FIG. 1

Map of the human rRNA gene repeat. The human rRNA gene forms a 44-kb repeat unit, with the four segments defined by _Eco_RI (E) sites. Boxes show the positions of sequences coding for the 18S, 5.8S, and 28S rRNA genes. Spacer, nontranscribed region. The horizontal arrow indicates the transcribed region. Numbers for genomic positions are according to GenBank (accession no. U13369). 1 to 3656, 5′ external spacer; 3657 to 5527, 18S segment; 5528 to 6622, internal spacer I; 6623 to 6779, 5.8S segment; 6780 to 7943, internal spacer II; 7944 to 12969, 28S segment; 12970 to 13314, 3′ external spacer. Replication starts bidirectionally in the nontranscribed intergenic spacer (18, 28, 74). Unidirectional replication fork barriers are located at the 3′ end of the transcribed region (23, 28).

FIG. 2

FIG. 2

Diagram of the LM-PCR protocol to detect top1-induced DNA single-strand breaks and replication-mediated DNA double-strand breaks. Top1 is shown as a shaded oval with covalent linkage to the 3′ end of a DNA single-strand break. In the assay for top1-induced DNA single-strand breaks, top1-induced DNA single-strand breaks (i.e., top1 cleavage complexes) were detected (upper left) as described previously (41) by annealing primer 1 (P1) to denatured genomic DNA. After primer extension and in vitro phosphorylation of the 5′-OH termini with T4 polynucleotide kinase, ligation to the double-stranded linker was performed. Thereafter, rRNA gene-specific DNA fragments were amplified with Taq DNA polymerase using the linker-primer and a nested, gene-specific PCR primer. After 26 cycles of PCR, a third primer (5′ end labeled with 32P; star) was used for two primer extension cycles before the samples were separated in 7% denaturing polyacrylamide gels. In the assay for replication-mediated DNA double-strand breaks, collision between a replication fork and a top1 cleavage complex is proposed to lead to replication runoff, with generation of a DNA double-strand break (upper right). Because of in vivo 5′-end phosphorylation of replication-mediated DNA double-strand breaks, ligation to the linker could be performed without prior T4 polynucleotide kinase reaction. The following reaction steps were the same as for the detection of top1-induced DNA single-strand breaks. Note that the single-strand break assay detects both single- and double-strand breaks.

FIG. 3

FIG. 3

Replication-mediated DNA double-strand breaks on the leading strand of the 18S human rRNA gene are prevented by the DNA synthesis inhibitor aphidicolin or treatment at 0°C. DNA single-strand breaks (S) and double-strand breaks (D) were determined in untreated HT29 cells (lanes 1 and 2) or after 1 h of treatment with CPT alone (lanes 3 and 4) or in combination with aphidicolin (Aph; 10 μM, 5-min pretreatment and 1-h cotreatment with CPT; lanes 5 and 6) or after treatment with CPT for 1 h on ice (lanes 7 and 8). Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369).

FIG. 4

FIG. 4

Replication-mediated DNA double-strand breaks are 5′-phosphorylated and coincide with top1-induced single-strand breaks. HT29 cells were exposed to CPT for 4 h. Experiments were performed with the RA primers (Table 1). Lane 1, complete reaction (see Fig. 2); lanes 2 to 5, 5′ phosphorylation of genomic DNA with T4 polynucleotide kinase omitted; lane 3, 5′ dephosphorylation with shrimp alkaline phosphatase (USB); lanes 4 and 5, incubation with Taq DNA polymerase and dNTPs omitted; lane 5, linker ligation omitted. Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369).

FIG. 5

FIG. 5

Repair of replication-mediated DNA double-strand breaks (panel D) induced by top1 cleavage complexes (panel S) on the leading strand of the 18S human rRNA gene. HT29 cells were treated with CPT for either 10 min or 4 h. Top1 cleavage complexes (i.e., DNA single-strand breaks [SSB], panel S) and replication-mediated DNA double-strand breaks (DSB, panel D) were studied at the indicated times (in hours after CPT removal). Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369). Lanes C, control DNA from untreated cells.

FIG. 6

FIG. 6

Replication-mediated DNA double-strand breaks are not detectable on the lagging strand of the 18S human rRNA gene. HT29 cells were treated with CPT for 4 h. DNA was extracted immediately after drug treatment at time 0 (lanes 3, 9, 12, and 16) or at various times after CPT removal. Time in drug-free medium (in hours) is indicated above lanes 4 to 7, 13, and 14. Lanes C, control (untreated) cells; lanes S, detection of top1 cleavage complexes (i.e., DNA single-strand breaks); lanes D, detection of replication-mediated DNA double-strand breaks. The scheme at the bottom illustrates the positions of the primers used with the landmarks of the rRNA gene. Data were obtained with primer set RC (panel A) and primer set RD (panel B). Lanes G+A, Maxam-Gilbert sequencing reactions. Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369).

FIG. 7

FIG. 7

Top1 cleavage complexes occur on both strands of the human 28S rRNA gene, whereas replication-mediated DNA double-strand breaks are detectable only on the leading strand. The scheme at the bottom illustrates the positions of the primers used with the landmarks of the rRNA gene. Experiments were performed with primer set RF for the leading DNA strand (panel A) and primer set RG for the lagging strand (Table 1) (panel B). HT29 cells were treated with CPT for 4 h. DNA was extracted immediately after drug treatment at time 0 (lanes 2, 8, 14, and 18) or at the indicated times after CPT removal. Time in drug-free medium is indicated above lanes 3 to 6, 9 to 12, 15, and 16. Lanes C, control (untreated) cells; lanes S, detection of DNA single-strand breaks; lanes D, detection of double-strand breaks; lanes G+A, Maxam-Gilbert sequencing reactions. Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369).

FIG. 8

FIG. 8

In vivo 5′ phosphorylation is not detectable on the lagging strand for DNA replication, suggesting that 5′-end phosphorylation is only detectable at replication-mediated DNA double-strand breaks. HT29 cells were exposed to CPT for 4 h. Experiments were performed with primers RC for the lagging DNA strand. Lane 1, complete reaction as described before (41), using primer 1; lane 2, incubation with T4 polynucleotide kinase omitted; lane 3, annealing and extension of primer 1 omitted; lane 4, linker ligation omitted. Numbers correspond to genomic positions of the DNA lesions (GenBank accession no. U13369).

FIG. 9

FIG. 9

Proposed interactions of DNA replication forks with CPT-stabilized top1 cleavage complexes and hypothetical repair pathways. Two covalent top1 cleavage complexes (shaded ovals) are shown, one on each side of a growing replication bubble (top). Parental DNA strands are represented as thick lines. Leading-strand synthesis is shown as thin arrows, and Okazaki fragments are shown as broken-line arrows. The differential effect of replication fork collision into top1 cleavage complexes on the leading and lagging strands is shown in the middle panel. Our results suggest that replication-mediated DNA double-strand breaks are formed by replication fork runoff on the leading strand with phosphorylation (P) of the 5′ end of the DNA template strand. By contrast, replication-mediated DNA double-strand breaks are not detectable on the lagging strand, which suggests that the replication fork is arrested upstream from the top1 cleavage complex without bypass, that the replication complex forces the dissociation of the top1 cleavage complex, or that Okasaki fragment synthesis bypasses the top1-mediated single-strand break interruption (see Discussion). In any case, no replication runoff would occur on the replicating lagging strand. (Bottom) Hypothetical excision repair of top1 cleavage complexes on the leading strand (see text for details and references). On the lagging strand, top1 might religate the DNA template strand directly upon drug removal. Replication-mediated DNA double-strand breaks are potential targets for homologous and illegitimate recombination.

Similar articles

Cited by

References

    1. Anderson R D, Berger N A. Mutagenicity and carcinogenicity of topoisomerase-interactive drugs. Mutat Res. 1994;309:109–142. - PubMed
    1. Avemann K, Knippers R, Koller T, Sogo J M. Camptothecin, a specific inhibitor of type I DNA topoisomerase, induces DNA breakage at replication forks. Mol Cell Biol. 1988;8:3026–3034. - PMC - PubMed
    1. Baker S D, Wadkins R M, Stewart C F, Beck W T, Danks M K. Cell cycle analysis of amount and distribution of nuclear DNA topoisomerase I as determined by fluorescence digital imaging microscopy. Cytometry. 1995;19:134–145. - PubMed
    1. Bendixen C, Thomsen B, Alsner J, Westergaard O. Camptothecin-stabilized topoisomerase I-DNA adducts cause premature termination of transcription. Biochemistry. 1990;29:5613–5619. - PubMed
    1. Bennett C B, Westmoreland T J, Snipe J R, Resnick M A. A double-strand break within a yeast artificial chromosome (YAC) containing human DNA can result in YAC loss, deletion, or cell lethality. Mol Cell Biol. 1996;16:4414–4425. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources