HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 35, 27386-27392, September 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/35/27386    most recent M003126200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vispé, S. Articles by Satoh, M. S. Search for Related Content PubMed PubMed Citation Articles by Vispé, S. Articles by Satoh, M. S. Social Bookmarking                      What's this?

DNA Repair Patch-mediated Double Strand DNA Break Formation in Human Cells^*

Stéphane Vispé and Masahiko S. Satoh

From the DNA Repair Group, Health and Environment Unit, Laval University Medical Research Center, CHUQ, Faculty of Medicine, Laval University, 2705 Boulevard Laurier, Sainte-Foy, Quebec G1V 4G2, Canada

Received for publication, April 12, 2000, and in revised form, May 19, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

To investigate the mechanism of double strand DNA break formation in mammalian cells, an in vitro assay was established using closed circular DNA containing two uracils on opposite DNA strands (18 and 30 base pairs apart) and extracts prepared from human cells. In this assay, formation of double strand breaks was detected by the conversion of circular DNA to linear DNA. Approximately 4-fold more double strand DNA breaks were produced by extracts from cells deficient in DNA ligase I (46BR) relative to those produced by extracts from control cells (MRC5, derived from a clinically normal individual). In parallel with the amount of double strand DNA breaks, extracts from 46BR cells produced longer repair patches (up to 24 bases in length) than those from MRC5 cells (typically <5 bases long). When purified DNA ligase I was added to 46BR extracts to complement the DNA ligase deficiency, only a negligible difference was found between the amount of doublestrand DNA breaks or the repair patch size generated in the assay relative to MRC5 extracts. Together, our data demonstrate that double strand DNA breaks are produced through formation of DNA repair patches. We refer to this process of double strand break formation as the "DNA repair patch-mediated pathway."

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

-Rays and x-rays are cytotoxic for cells (for review, see Refs. 1 and 2). These forms of ionizing radiation produce reactive oxygen free radicals that generate a variety of DNA lesions, including oxidized DNA bases, apurinic/apyrimidinic (AP)¹ sites, and single strand DNA breaks (SSBs) (1-4). In cells, some of these initial lesions are converted to double strand DNA breaks (DSBs), which play a central role in the cytotoxicity induced by ionizing radiation (1, 2). The following mechanisms for the formation of DSBs have been proposed.

Ionizing radiation induces clusters of DNA damage, with lesions spaced 10-20 bases apart (2). The formation of two closely spaced SSBs on opposite strands can result in strand separation (1, 2) if the T_m for the intervening sequence is below 37 °C (assuming a GC content of 50% and an intervening sequence of 10 base pairs). DSBs can also occur as a result of two closely spaced oxidized bases or AP sites (5, 6). Oxidized DNA bases are removed by DNA glycosylases, resulting in the formation of AP sites (7-9). These AP sites or AP sites generated directly by hydroxy radicals are then incised by AP endonuclease, which produces SSBs (7-9). Thus, the cellular response to closely spaced AP sites or oxidized bases will result in closely spaced SSBs capable of forming DSBs as described above. In fact, it has been shown that closely spaced AP sites in duplex oligonucleotides are converted to DSBs by Escherichia coli or human AP endonuclease (5, 6). When damaged bases are located sufficiently far apart, the T_m for the intervening sequence will be too high for DSBs to form by simple strand separation.

During the repair of oxidized DNA bases or AP sites by base excision repair (BER), 1-base gaps are enzymatically created and filled by DNA polymerases (8, 10). Using a reconstituted base excision repair assay, it was found that whereas DNA polymerase  fills these gaps primarily with a single nucleotide (11), DNA polymerases  and  fill the gaps with up to 10 nucleotides in conjunction with displacement of the DNA strand from the initial gaps in the 5' to 3' direction (12-14). In vivo, repair patches up to 40 nucleotides long have been observed at damage sites (4, 15, 16). Repair of DNA breaks on opposite strands gives rise to a situation in which the DNA polymerases synthesizing new DNA to fill in the repair patches move toward each other; if the repair patches are long enough, the two polymerases will collide and fall off the DNA, resulting in a DSB. Alternatively, DNA polymerase involved in the repair of one break may reach another break on the opposite strand before completing synthesis of the repair patch, also resulting in a DSB. Thus, even when the DNA between two lesions has a T_m >37 °C, such lesions can be converted to DSBs. We refer to this pathway as the "DNA repair patch-mediated pathway" for DSB formation.

To test the hypothesis that long repair patches can mediate DSB formation, we established an in vitro DSB formation assay using cell-free extracts and closed circular plasmid containing damage in the form of two uracils on opposite strands. Cell-free extracts were prepared from MRC5 cells (derived from a clinically normal individual) and from 46BR cells (generated from a patient with DNA ligase I deficiency) (17). 46BR cells show delayed rejoining of DNA breaks following exposure to DNA-damaging agents due to their DNA ligase I deficiency (18, 19).

Here we show that extracts from 46BR cells produced longer repair patches (up to 24 bases) relative to extracts from MRC5 cells (up to 5 bases). In parallel with the increased repair patch size, extracts from 46BR cells produced more DSBs as well. Addition of DNA ligase I to the reaction reduced the repair patch size as well as the amount of DSBs formed in the presence of 46BR extracts. Thus, our data demonstrate that the DNA repair patch-mediated pathway is involved in the formation of DSBs. In cells, this pathway may significantly increase the probability of DSB formation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Lines and Recombinant DNA Ligase I-- The simian virus 40-transformed human fibroblasts 46BR.1G1 (DNA ligase I-deficient) and MRC5 (from a clinically normal individual) were obtained from the European Collection of Cell Cultures (Salisbury, Wiltshire, United Kingdom). Lymphoblastoid GMO1953A cells (from a clinically normal individual) were from the NIGMS Human Mutant Cell Repository (Camden, NJ). DNA ligase I cDNA cloned into the E. coli expression vector pET11a was kindly provided by Dr. T. Lindahl, and the protein was purified essentially following the method of Mackenney et al. (20) using Ni²⁺-nitrilotriacetic acid resin (QIAGEN Inc.).

Construction of Substrates for DSB Assay-- The following six polyacrylamide gel electrophoresis-purified oligonucleotides were synthesized (Genosys Biotechnologies, The Woodlands, TX): KS(+), 5'-GCGCGGGGTACCGGTCCGGACATATGGGCGCCGCGCGCCCATGGGCATGCGGCCGCCT-3'; KS(+)U, 5'-GCGCGGGGTACCGGUCCGGACATATGGGCGCCGCGCGCCCATGGGCATGCGGCCGCCT-3'; KS(), 5'-CGGCCGCATGCCCATGGGCGCGCGGCGCCCATATGTCCGGACCGGTACCCCGCGC-3'; KS()U10, 5'-CGGCCGCATCCCCATGGGCGCGCGGCGCCUATATGTCCGGACCGGTACCCCGCGC-3'; KS()U18, 5'-CGGCCGCATCCCCATGGGCGCUCGGCGCCCATATGTCCGGACCGGTACCCCGCGC-3'; and KS()U30, 5'-CGGCCGCATUCCCATGGGCGCGCGGCGCCCATATGTCCGGACCGGTACCCCGCGC-3'.

KS(), KS()U10, KS()U18, and KS()U30 (1 nmol/ml) were phosphorylated using 70 units/ml T4 polynucleotide kinase (Amersham Pharmacia Biotech) with 1 mM ATP in the reaction mixture recommended by the supplier. To prepare duplex oligonucleotides, 1 nmol/ml each KS(+) and KS(), KS(+) and KS()U18, KS(+)U and KS()U10, KS(+)U and KS()U18, or KS(+)U and KS()U30 was mixed and incubated at 90 °C for 5 min, followed by 56 °C for 5 min. The resulting annealed oligonucleotides, which contained 3'-TCC protruding ends and a KpnI site on the other end (see Fig. 1A), are referred to as F0U (without uracil), F1U (one uracil), F2U10 (two uracils 10 bp apart), F2U18 (two uracils 18 bp apart), and F2U30 (two uracils 30 bp apart), respectively. These duplex oligonucleotides also contained NcoI, SfoI, and BspEI sites as illustrated in Fig. 1A.

The pBS(S) plasmid was constructed by inserting a 40-bp duplex oligonucleotide containing an SfiI site into the KpnI and EcoRI sites of pBluescript II KS⁺ (Stratagene). Then, 400 µg/ml pBS(S) was digested with 800 units/ml KpnI for 2 h at 37 °C, followed by 440 units/ml SfiI for 2 h at 50 °C. Fragments (20 bp) produced by the double digest were removed from the rest of pBS(SfiI-KpnI) by MicroSpin S-400 HR gel filtration columns (Amersham Pharmacia Biotech). 200 µg/ml pBS(SfiI-KpnI) fragment with a AGG-3' tail attached was then ligated to the 3'-TCC tail of the F0U, F1U, F2U10, F2U18, or F2U30 oligonucleotide (440 pmol/ml) using 38 Weiss units/ml T4 DNA ligase (Amersham Pharmacia Biotech) and 1 mM ATP in the reaction mixture provided by Amersham Pharmacia Biotech for 16 h at 16 °C. After phenol/chloroform extraction and ethanol precipitation of DNA, the resulting pBS(SfiI-KpnI) fragment, ligated to the annealed oligonucleotides, was digested with 800 units/ml KpnI at 37 °C for 2 h. Then, the DNA was self-circularized through the KpnI site in ligation buffer (Amersham Pharmacia Biotech) containing 1 mM ATP and 10 Weiss units/ml T4 DNA ligase for 16 h at 16 °C. The circularized DNA was concentrated by ultracentrifugation at 100,000 × g for 24 h using a Beckman SW 28 rotor. Precipitated DNA was recovered, and closed circular DNA was purified by two sequential EtBr/CsCl centrifugations. The resulting closed circular plasmids containing F0U, F1U, F2U10, F2U18, or F2U30 oligonucleotides are referred to as P0U, P1U, P2U10, P2U18, and P2U30, respectively.

Induction of Single Strand DNA Breaks at the Site of Uracils-- P2U10, P2U18, and P2U30 plasmids (640 ng) were incubated with 0.4 units of uracil-DNA glycosylase (New England Biolabs Inc.) in the reaction mixture (20 µl) provided by the supplier for 1 h at 37 °C. The resulting AP sites were hydrolyzed by 45 mM Tris-HCl (pH 8.8) at 37 °C for 24 h. DNA was then analyzed by EtBr and 1% agarose gel electrophoresis.

Preparation of Cell-free Extracts, DNA Repair Reaction, and DSB Formation Assay-- Cell-free extracts were prepared following the method of Manley et al. (21). The DNA repair reaction was carried out with 300 ng of substrate DNA, 50 µg of cell-free extract, and 2 mM NAD⁺ at 30 °C in a 50-µl reaction volume under conditions described previously (22, 23). The resulting DNA was purified by phenol/chloroform extraction followed by ethanol precipitation in the presence of 14 µg of tRNA as a carrier and 2.5 M ammonium acetate. Precipitated DNA was dissolved with 21 µl of 90 mM Tris-HCl (pH 8.0), 90 mM boric acid, and 2 mM EDTA and treated with 100 µg/ml RNase A. The DNA was then fractionated by EtBr and 1% agarose gel electrophoresis to visualize linear DNA produced by the formation of DSBs.

Analysis of Repair Patch Size-- Cell-free reactions were carried out with 300 ng of either P0U or P1U as described above in the presence of 10 µCi of [-³²P]dGTP or [-³²P]dCTP. After purification and precipitation of DNA, samples were dissolved in 10 µl of 10 mM Tris-HCl (pH 8.0) and double-digested with AflIII (200 units/ml) and NcoI (500 units/ml), SfoI (200 units/ml), BspEI (500 units/ml), or KpnI (1000 units/ml). Digested DNA fragments were fractionated by EtBr and 1% agarose gel electrophoresis. After drying the gel and exposing to x-ray film, ³²P activity incorporated into the DNA fragments was quantified by an AlphaImager (Packard Instrument Co.).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Construction of Substrate-- Closed circular DNAs containing two uracils on opposite DNA strands either 18 or 30 bp apart were constructed for use with cell-free extracts in an in vitro DSB formation assay. As shown in Fig. 1A, two oligonucleotides were annealed, and double stranded oligonucleotides containing two uracils spaced either 18 or 30 bp apart (F2U18 and F2U30, respectively) were prepared. F0U, F1U, and F2U10 (containing no uracil and one uracil and two uracils 10 bp apart, respectively) were also prepared as controls. The double-stranded oligonucleotides contained four restriction sites (NcoI (Nc), SfoI (Sf), BspEI (Bs), and KpnI (K)) to allow for analysis of repair patch size. Although double stranded oligonucleotides can be used directly in an in vitro DSB formation assay, the presence of two DNA ends that are themselves DSBs might have complicated analysis of DSB formation. In addition, repair of DNA breaks by DNA polymerases  and  does not occur efficiently on linear DNA (24). Therefore, double stranded oligonucleotides were circularized by insertion into SfiI- and KpnI-double-digested pBS(S) as summarized in Fig. 1B. AGG-3' and 3'-TCC protruding ends were attached to double-digested pBS(S) and double stranded oligonucleotides, respectively (Fig. 1, A and B). Since AGG-3' and 3'-TCC are non-palindromic sequences, intermolecular ligation of double-digested pBS(S) molecules or double stranded oligonucleotides was inhibited, allowing ligation of double-digested pBS(S) to the double stranded oligonucleotide. The ligated DNA was then treated with KpnI and self-circularized. The typical yield of substrate DNA was ~20 µg from 500 µg of pBS(S). The resulting circular DNAs contained no uracil, one uracil, two uracils 10 bp apart, two uracils 18 bp apart, or two uracils 30 bp apart and are referred to as P0U, P1U, P2U10, P2U18, and P2U30, respectively. About 5% of the resulting DNAs appeared to be closed circular dimers (Fig. 1C, Non-treated upper bands).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Schematic illustration for the construction of substrate DNA and analysis of substrate stability after induction of SSBs at uracil sites. A, two synthetic oligonucleotides were annealed, and double stranded oligonucleotides were prepared. These oligonucleotides contained NcoI (Nc), SfoI (Sf), BspEI (Bs), and KpnI (K) sites. Oligonucleotides containing no uracil, one uracil, two uracils 10 bp apart, two uracils 18 bp apart, and two uracils 30 bp apart are referred as to F0U, F1U, F2U10, F2U18, and F2U30, respectively. B, these double-stranded oligonucleotides were ligated into SfiI- and KpnI-double-digested pBS(S). After ligation through AGG-3' and 3'-TCC tails, the DNA was treated with KpnI and self-circularized. C, to test plasmid stability following induction of SSBs at the site of uracils, P2U10, P2U18, and P2U30 were treated with uracil-DNA glycosylase and weak alkaline solution to hydrolyze the resulting AP site, and formation of linear DNA was analyzed by EtBr and 1% agarose gel electrophoresis. The Tm values for the sequence between the two uracils are 30, 62, and 75 °C for P2U10, P2U18, and P2U30, respectively.

Although manipulation of DNA by this procedure did induce some damage, as detected by incorporation of dGMP (~40 fmol) into P0U (30 fmol) following incubation of the plasmid with MRC5 cell-free extracts (using DSB formation assay conditions as described under "Experimental Procedures"), the amount of damage was not significant. Assuming that a repair patch size of 1 base was generated, we estimated the number of damaged bases to be only 1-2/P0U plasmid. In addition, this degree of damage did not affect analysis of DSB formation since a negligible amount of P0U DNA was linearized when incubated with cell-free extract (data not shown).

The calculated T_m values for the intervening sequence between the two uracils in P2U18 and P2U30 are 62 and 75 °C, respectively. Thus, even if SSBs are produced at the uracil sites, these substrates are not expected to be linearized at 37 °C. In fact, after removal of uracils from P2U18 or P2U30 by uracil-DNA glycosylase followed by hydrolysis of the resulting AP sites by alkaline treatment (pH 8.8) for 24 h at 37 °C, >90% of closed circular P2U18 and P2U30 plasmids were converted to the open circular form, but only negligible amounts were linearized (Fig. 1C). On the other hand, P2U10, with a T_m of 30 °C for the intervening sequence, was linearized by this treatment.

Distribution of Repair Patch Sizes-- To determine the distribution of repair patch sizes generated by different cell extracts, a cell-free DNA repair assay was carried out with P1U, which contains one uracil. Since the counterpart of uracil is cytosine (Fig. 2A), [-³²P]dGTP was included in the reaction to label repair patches. Then, DNA was extracted from the reaction mixture and digested with AflIII (located 420 bp downstream from uracil) and one of the following restriction enzymes: NcoI (Nc), SfoI (Sf), BspEI (Bs), or KpnI (Kp) (Fig. 2A). Since NcoI is located upstream of the uracils, radioactivity incorporated into the NcoI-AflIII fragment reflects total repair activity. The digested DNA was analyzed by EtBr and 1% agarose gel electrophoresis. If a given restriction site is located within the repair patch, radioactivity should be found in the DNA fragments generated by the double digest. However, if the site is located beyond the repair patch, radioactivity found in the DNA fragment should be comparable to background levels.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Analysis of repair patch distribution. A, DNA sequences around the uracil in P1U are illustrated. Fragment size after double digestion with AflIII and NcoI (Nc), SfoI (Sf), BspEI (Bs), or KpnI (Kp) is indicated on the top. Numbers below the sequence are based on a starting position of 1 for uracil. The counterpart of uracil is cytosine. B, after incubation of P0U or P1U with extracts from either MRC5 or 46BR cells for 30 min in the presence or absence of 10 × 104 units (Weiss units) of purified DNA ligase I, DNA was double-digested and fractionated by EtBr and 1% agarose gel electrophoresis. After drying, the gel was exposed to x-ray film for autoradiography. C, 32P was quantified with an AlphaImager.

As shown in Fig. 2 (B and C,  DNA ligase I), incorporation of radioactivity was found in the NcoI-AflIII fragments derived from the P1U plasmid (B and C, P1U, Nc), whereas less [-³²P]dGMP was found incorporated into fragments extracted from P0U (B, P0U, Nc), the plasmid containing no uracil. These data suggest that incorporation of ³²P was due to repair of the uracil. When assays were carried out using extracts from MRC5 cells, significantly less ³²P activity was incorporated into P1U SfoI-AflIII fragments (Fig. 2, B and C, MRC5, P1U, Sf). Since the SfoI restriction site is located 5 bp downstream from the uracil (Fig. 2A), this result suggests that most of the repair patches generated by MRC5 cells do not exceed 5 bases in length. When [-³²P]dCTP, instead of [-³²P]dGTP, was employed (only repair patches >2 bases were labeled), the ³²P activity found in the NcoI-AflIII fragments was reduced to 20% of that found in assays with [-³²P]dGTP (data not shown). Thus, these data suggest that most of the repair patches produced by extracts from MRC5 cells are of 1 base.

When the P1U plasmid was incubated with 46BR extracts and [-³²P]dGTP, a significant amount of radioactivity was detected even in the SfoI-AflIII and BspEI-AflIII fragments (Fig. 2, B and C, 46BR, P1U, Sf and Bs). Since 46BR extracts caused higher nonspecific incorporation of [³²P]dGMP into P0U, the amount of ³²P activity incorporated into P0U fragments was subtracted from that incorporated into P1U. As shown in Fig. 2C, the amounts of [³²P]dGMP found in the SfoI-AflIII and BspEI-AflIII fragments were still significantly increased for the assay using 46BR extracts relative to MRC5 extracts. From this we conclude that most repair patches created by 46BR extracts are up to 24 bases in length, much longer than the repair patches produced by MRC5 cell extracts. When [-³²P]dCTP, instead of [-³²P]dGTP, was employed, the amount of ³²P incorporated into the NcoI-AflIII fragments was only reduced to 70% of that found in the assay with [-³²P]dGTP (data not shown). These data suggest that the proportion of single nucleotide repair occurring in assays with 46BR extracts was significantly less than that in assays with MRC5 extracts.

When 46BR extracts were tested in an assay to which purified DNA ligase I was added, ³²P activity in the SfoI-AflIII and BspEI-AflIII fragments was significantly reduced (Fig. 2, B and C, + DNA ligase I). On the other hand, addition of T4 DNA ligase had no effect on the size of repair patches (data not shown). These data demonstrate that the increased repair patch size observed with 46BR extracts is due to the deficiency in DNA ligase I. In an in vitro reconstituted BER assay, the absence of DNA ligase I caused increased repair patch size (13, 25), a result that agrees well with our observations using 46BR extracts.

DSB Formation-- Either P2U18 or P2U30 plasmids were incubated with MRC5 cell-free extracts. The resulting products were analyzed by EtBr and 1% agarose gel electrophoresis. Both open circular forms (generated by removal of uracils and incision of the resulting AP sites) and linear forms (generated by DSB formation) were generated from the closed circular plasmid (Fig. 3A, MRC5). Similar results were obtained when substrate plasmids containing AP sites instead of uracils were used (data not shown). The amount of linearized DNA became saturated above 50 µg of extracts (data not shown).

View larger version (31K): [in this window] [in a new window]   Fig. 3.   DSB formation assay. A, P1U, P2U18, or P2U30 plasmids were incubated with extracts from MRC5 or 46BR cells for various times, and the DNA was purified as described under "Experimental Procedures." The DNA was then fractionated by EtBr and 1% agarose gel electrophoresis to allow detection of linear DNA, a reflection of DSB formation. Contrast was adjusted for linear DNA, thus saturating the signals for open and closed circular DNAs. B, quantified data are shown.

To exclude the possibility that the DNA was linearized by nonspecific digestion, we analyzed P1U, which contains only one uracil. Linearization of this plasmid in the assay with MRC5 extracts would indicate that nonspecific nicking, rather than the specific repair of uracils, was responsible for linearization of the P2U18 and P2U30 plasmids. The fact that negligible amounts of linear P1U DNA were produced in the presence of extracts from MRC5 cells (Fig. 3A, MRC5) indicates that linearization of the P2U18 and P2U30 plasmids was linked to the presence of 2 damaged bases. To further confirm that DSBs were created between the two DNA uracils, linearized DNA was extracted from the agarose gel and digested with AflIII, resulting in the expected 0.6- and 2.4-kilobase pair fragments (data not shown).

Under the conditions described, ~7.0 ng of linear DNA (corresponding to 2.5% of total DNA) was formed from 300 ng of closed circular P2U18 after a 60-min incubation (Fig. 3, A and B). When the P2U30 plasmid was incubated with MRC5 extracts, linearization was reduced compared with P2U18 linearization (Fig. 3, A and B, MRC5, 7.0 ng for P2U18 versus 3.0 ng for P2U30), indicating that the probability of forming DSBs is reduced in proportion to the distance between the two damage sites. Similar results were obtained when extracts from GMO1953A cells were used (data not shown).

When the P1U plasmid was incubated with 46BR extracts in the DSB formation assay, the amount of open circular DNA produced was greater than that produced by MRC5 extracts (40% of total DNA at 60 min versus 25% for MRC5 extracts). We routinely added NAD⁺, the substrate for poly(ADP-ribose) polymerase, to the assay to promote automodification of poly(ADP-ribose) polymerase and thus inhibit binding of poly(ADP-ribose) polymerase to DNA breaks (22, 23). In the absence of NAD⁺, poly(ADP-ribose) polymerase significantly reduces the rate of DNA break rejoining by remaining on DNA breaks (22, 23). When NAD⁺ was removed, in fact, the amount of the open circular form increased to 80% of total DNA (no significant difference in the amount of open circular DNA was found for MRC5 versus 46BR extracts). These data suggest that both extracts contain a similar level of incision activity at uracil sites and that accumulation of more open circular DNA in 46BR extracts versus MRC5 extracts under standard assay conditions (i.e. in the presence of NAD⁺) is due to DNA ligase I deficiency.

Consistent with results obtained using the P1U plasmid, an increased amount of open circular DNA was formed when P2U18 was incubated with extracts from 46BR cells. In addition, ~4-fold more linear DNA (30 ng or 10% of total DNA at 60 min of incubation) was produced relative to that formed by extracts from MRC5 cells (Fig. 3, A and B). In addition, extracts from 46BR cells were capable of linearizing P2U30 (Fig. 3, A and B, 3.0 ng for MRC5 versus 14 ng for 46BR). As demonstrated (Fig. 2), extracts from 46BR cells increased the DNA repair patch size. Thus, these data suggest that the formation of long repair patches promotes increased production of DSBs.

Reduced DSB Formation by 46BR Extracts in the Presence of DNA Ligase I-- As described, addition of DNA ligase I reduced the DNA repair patch size produced by extracts from 46BR cells (Fig. 2, B and C). In parallel with this reduction, addition of DNA ligase I also decreased the amount of linear DNA formed by 46BR extracts (Fig. 4, A and B). However, the equivalent amount of DNA ligase I had a less significant effect on DNA linearization by MRC5 extracts. In this assay, no circularization or dimerization of linearized DNA was observed (data not shown), suggesting that reduced formation of linear DNA was not due to rejoining of DNA ends. When T4 DNA ligase was added instead of DNA ligase I, the effect on linearization was negligible (Fig. 4, A and B). Thus, these data suggest that DSB formation in 46BR cells is linked to the deficiency in DNA ligase I and to formation of long repair patches.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Reduced DSB formation by addition of purified DNA ligase I. A, P1U and P2U18 were incubated with extracts from MRC5 or 46BR cells in the absence or presence of purified DNA ligase I (0.5, 2, or 10 × 104 Weiss units) or T4 DNA ligase (10 × 104 Weiss units) for 60 min as indicated, and DNA was purified as described under "Experimental Procedures." The DNA was then fractionated by EtBr and 1% agarose gel electrophoresis for detection of linear DNA generated by DSB formation. Contrast was adjusted for linear DNA, thus saturating signals for open and closed circular DNAs. B, quantified data are shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this report, we demonstrate a novel pathway for the production of DSBs. In mammalian cells, DSBs are often observed to develop during the post-irradiation period (26, 27), leading to speculation that cellular enzymes are involved. In this context, generation of DSBs by double incision of two closely spaced AP sites on opposite strands was demonstrated using a model substrate and purified E. coli or human AP endonuclease (5, 6). Similarly, Dianov et al. (28) showed that the excision of two uracil residues closely spaced on opposite strands by E. coli extracts led to the formation of DSBs. In addition, closely spaced SSBs and 8-hydroxyguanine sites also resulted in DSBs when the substrate DNA was treated with E. coli formamidopyrimidine-DNA glycosylase (29), an enzyme that induces DNA breaks at the site of 8-hydroxyguanine through its AP lyase activity (8, 9). These results provide a conceptual background for the formation of DSBs. However, in these cases, DSB formation is dependent on a T_m below 37 °C for dissociation of DNA between the sites of damage on opposite strands. Assuming a GC content of 50% in the intervening sequence, the T_m would be 30 °C for a 10-base stretch and 42 °C for a 14-base stretch. However, for DNA lesions spaced >14 bases apart, the intervening DNA is generally stable at 37 °C. It has been demonstrated that ionizing radiation induces closely spaced clusters of DNA damage (within a range of 10-20 bp (2) and possibly up to 80 bp in vivo (2)). These distances typically preclude DSB formation by the above mechanism. However, in this paper, we demonstrate the existence of a new pathway, the DNA repair patch-mediated pathway, which is capable of generating DSBs from two opposed sites of damage located 18 and even 30 bp apart. The existence of this pathway therefore significantly increases the likelihood of DSB formation in cells.

In Fig. 5, we present a model for DSB formation by the DNA repair patch-mediated pathway. A DNA duplex containing two SSBs on opposite strands is formed. This duplex is stable if the T_m for the intervening sequence is >37 °C. When the SSBs are repaired with short repair patches, the breaks are sealed without formation of DSBs (Fig. 5A). When long repair patches are formed, the repair patch produced from one SSB may reach all the way to the SSB on the opposite strand, resulting in the formation of a DSB (Fig. 5B). Alternatively, the repair patches generated from the two SSBs may be long enough to overlap, resulting in collision of the DNA polymerases synthesizing the repair patches and formation of a DSB (Fig. 5C). Thus, our model suggests that DNA repair with short patches presents less risk for DSB formation than repair with long patches.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Model for DNA repair patch-mediated DSB formation. DNA repair patches are formed by DNA polymerases and, in the case of long patch, by other factors (e.g. flap endonuclease 1 and proliferating cell nuclear antigen) (12, 38, 43-45). See "Discussion" for details.

DSBs are the most lethal type of DNA damage. In mammalian cells, DSBs can be rejoined either through non-homologous end joining by the DNA-dependent protein kinase catalytic subunit/Ku pathway or by homologous recombination (30-32). However, incorrect or incomplete rejoining of DSB may cause mutations or cytotoxicity. Thus, frequent formation of DSBs by virtue of a long repair patch phenotype may make cells more vulnerable to permanent mutation or even cell death. In fact, 46BR cells are hypersensitive to ionizing radiation (33). Upon exposure of 46BR cells to -rays, slightly increased amounts of DNA breaks were found immediately following exposure compared with DNA from control cells (34), possibly reflecting a reduced rate of DNA break rejoining due to the cells' DNA ligase I deficiency (17). The fact that repair of DNA damage clusters with long repair patches in these cells is more likely to give rise to DSBs may explain their hypersensitivity to ionizing radiation.

Although exposure to -rays resulted in slightly increased DNA breakage in 46BR cells immediately following exposure, Nocentini (35) found that the rate at which DNA breaks were rejoined in these cells relative to control cells was similar. Since DNA ligase I deficiency does not affect the rate of rejoining, Nocentini concluded that DNA ligase I is not involved in BER, which repairs most kinds of -ray-induced DNA damage and produces single strand DNA breaks as an intermediate (8, 10, 36). On the other hand, biochemical data strongly indicate that DNA ligase I is essential for BER (13, 25, 37-39). In this regard, we found a slightly reduced rate of DNA break rejoining at uracil sites in the presence of 46BR extracts compared with MRC5 extracts. The difference was negligible when a plasmid containing AP sites rather than uracils was used (data not shown). However, extracts from 46BR cells consistently produced increased repair patch sizes at uracil and AP sites relative to repair patch sizes in the presence of MRC5 extracts. Thus, our results suggest that DNA ligase I deficiency does not always reduce the rate of DNA break rejoining and that the major biochemical effect of the deficiency in DNA ligase I is increased DNA repair patch size. Therefore, measurements of the rate of DNA break rejoining by Nocentini may not reflect an abnormality in DNA ligase I activity.

46BR cells are sensitive to alkylating agents in addition to ionizing radiation (33), although these agents do not form damage clusters (2). In addition, alkylating agents must cause significantly more DNA lesions than those produced by ionizing radiation to result in equivalent cytotoxicity. For example, a 63% killing dose of ionizing radiation induces ~1 × 10³ damages (SSBs) per cell (2), whereas a similar level of killing by alkylating agents (such as N-methyl-N-nitrosourea) requires ~8 × 10⁵ damages (7-methylguanine) per cell (2). The formation of a large number of damage sites within cellular DNA apparently increases the probability of forming closely spaced DNA lesions on opposite strands, even in the absence of damage clusters like those observed upon exposure to ionizing radiation. In fact, it has been demonstrated that alkylating agents cause DSB formation (40). It is possible that the long repair patch-mediated pathway is involved in DSB formation downstream of exposure to both alkylating agents and ionizing radiation.

In this work, we used 46BR cells as a model to establish the relationship between long repair patches, DSB formation, and DNA ligase I deficiency. However, even in the presence of normal levels of DNA ligase I activity, BER is known to produce repair patches of various lengths. Repair patch size depends to some extent on the DNA polymerase used to fill the gaps. For example, DNA polymerase  together with the DNA ligase III-XRCC1 complex (10, 11) generally create a short repair patch, typically only 1 base long (11, 41), whereas DNA polymerases  and  produce longer repair patches (up to 10 bases in vitro) (12-14). Based on our model, repair of DNA damage by DNA polymerase  would be associated with a low risk of DSB formation (Fig. 5). Although DNA polymerase  lacks 3' to 5' proofreading activity and thus has an error rate significantly higher than that of other DNA polymerases (42), the fact that this polymerase gives rise to repair patches only 1 base in length should confer an advantage on cells with regard to DSB formation following DNA damage.

	ACKNOWLEDGEMENTS

We thank Cristina Ward for editing and scientific comment and Sachiko Sato for discussion.

	FOOTNOTES

^* This work was supported in part by the National Cancer Inst. of Canada for the Terry Fox Run (to M. S. S.). The Canada Foundation for Innovation and the Quebec Government provided infrastructure support.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by a scholarship from the Medical Research Council of Canada. To whom correspondence should be addressed. Tel.: 418-656-4141 (ext. 7340); Fax: 418-654-2159; E-mail: Masahiko.Sato@crchul.ulaval.ca.

Published, JBC Papers in Press, May 25, 2000, DOI 10.1074/jbc.M003126200

	ABBREVIATIONS

The abbreviations used are: AP, apurinic/apyrimidinic; SSB, singlestrand break; DSB, double strand break; T_m, melting temperature; BER, base excision repair; bp, base pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				E. L. Y. Ho, M. Parent, and M. S. Satoh Induction of Base Damages Representing a High Risk Site for Double-strand DNA Break Formation in Genomic DNA by Exposure of Cells to DNA Damaging Agents J. Biol. Chem., July 27, 2007; 282(30): 21913 - 21923. [Abstract] [Full Text] [PDF]

				E. L. Y. Ho and M. S. Satoh Repair of single-strand DNA interruptions by redundant pathways and its implication in cellular sensitivity to DNA-damaging agents Nucleic Acids Res., December 1, 2003; 31(23): 7032 - 7040. [Abstract] [Full Text] [PDF]

				S. Vispe, E. L. Y. Ho, T. M. C. Yung, and M. S. Satoh Double-strand DNA Break Formation Mediated by Flap Endonuclease-1 J. Biol. Chem., September 12, 2003; 278(37): 35279 - 35285. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/35/27386    most recent M003126200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Vispé, S. Articles by Satoh, M. S. Search for Related Content PubMed PubMed Citation Articles by Vispé, S. Articles by Satoh, M. S. Social Bookmarking                      What's this?