Werner and Bloom helicases are involved in DNA repair in a complementary fashion

Abstract

Werner syndrome (WS) is a recessive disorder characterized by premature senescence. Bloom syndrome (BS) is a recessive disorder characterized by short stature and immunodeficiency. A common characteristic of both syndromes is genomic instability leading to tumorigenesis. WRN and BLM genes causing WS and BS, encode proteins that are closely related to the RecQ helicase. We produced WRN^−/−, BLM^−/− and WRN^−/−/BLM^−/− mutants in the chicken B-cell line DT40. WRN^−/− cells showed hypersensitivities to genotoxic agents, such as 4-nitroquinoline 1-oxide, camptothecin and methyl methanesulfonate. They also showed a threefold increase in targeted integration rate of exogenous DNAs, but not in sister chromatid exchange (SCE) frequency. BLM^−/− cells showed hypersensitivities to the genotoxic agents as well as ultraviolet (UV) light, in addition to a 10-fold increase in targeted integration rate and an 11-fold increase in SCE frequency. In WRN^−/−/BLM^−/− cells, synergistically increased hypersensitivities to the genotoxic agents were observed whereas both SCE frequencies and targeted integration rates were partially diminished compared to the single mutants. Chromosomal aberrations were also synergistically increased in WRN^−/−/BLM^−/− cells when irradiated with UV light in late S to G₂ phases. These results suggest that both WRN and BLM may be involved in DNA repair in a complementary fashion.

Introduction

Werner syndrome (WS) is a rare autosomal recessive genetic disorder causing premature aging, including short stature, juvenile cataracts, atrophy of the skin, graying and loss of hair, diabetes, arteriosclerosis, and osteoporosis, accompanied by rare cancers (Epstein et al., 1966). In vitro studies of fibroblast growth characteristics also suggest that WS may be related to normal aging since the life span of WS fibroblasts as expressed by population doubling levels is much shorter than that of normal fibroblasts (Faragher et al., 1993; Salk et al., 1985) and they have a prolonged S phase of the cell cycle. In addition to replication, WS cells are genetically unstable, as described initially by the finding of nonclonal chromosomal translocations (Salk et al., 1985) and extensive genomic deletions (Fukuchi et al., 1989).

On the other hand, Bloom syndrome (BS) is a rare autosomal recessive disorder, too. BS is characterized by proportional dwarfism, immunodeficiency, sun-sensitive facial erythema, genomic instability and the early development of a wide variety of cancers (German, 1993). Cultured fibroblasts and B-lymphoblastoid cells from BS patients are hypermutable and have chromosomal aberrations, such as an excessive number of locus-specific mutations and a high frequency of microscopically visible chromatid gaps, breaks, and rearrangements (Rosin and German, 1985). BS cells also show 10–15-fold higher levels of sister chromatid exchanges (SCEs) than normal control cells (Chaganti et al., 1974; German, 1993; Heartlein et al., 1987). The increase in SCE frequency in BS cells is further enhanced by exposure to ultraviolet (UV) light (Kurihara et al., 1987) and treatment with DNA damaging agents, such as ethyl methanesulfonate, N-ethyl-N-nitrosourea and 5-bromodeoxyuridine (BrdU) (Heartlein et al., 1987; Krepinsky et al., 1980).

WRN and BLM genes causing WS and BS, respectively, encode proteins WRN and BLM, respectively. Both proteins have seven signature motifs conserved in a wide range of DNA and RNA helicases (Ellis et al., 1995; Yu et al., 1996). Both WRN and BLM have greatest similarity to the RecQ subfamily of DNA helicases that include Escherichia coli RecQ (Nakayama et al., 1985), Saccharomyces cerevisiae Sgs1 (Gangloff et al., 1994; Watt et al., 1995), Schizosaccharomyces pombe Rqh1 (Stewart et al., 1997), human RecQL4 that is defective in some cases of Rothmund-Thomson syndrome (Kitao et al., 1999), human RecQL1 (Puranam and Blackshear, 1994; Seki et al., 1994), and human RecQL5 (Shimamoto et al., 2000). Purified recombinant WRN and BLM proteins unwind duplex DNA in a 3′-to-5′ direction and require a 3′ single strand tail. (Gray et al., 1997; Karow et al., 1997; Suzuki et al., 1997). Both enzymes interact directly with human replication protein A (RPA) that facilitates their DNA-unwinding activities (Brosh et al., 1999; 2001).

Advanced sequence alignment analysis showed a putative exonuclease domain near the N-terminus of WRN (Mushegian et al., 1997). This domain contains three conserved motifs that resemble the conserved motifs in the proofreading exonuclease domain of E. coli DNA polymerase I and in E. coli RNaseD. An exonuclease activity of purified WRN protein has been described, although there is some disagreement on its directionality and dependency on the helicase activity (Huang et al., 1998; Kamath-Loeb et al., 1998; Shen et al., 1998; Suzuki et al., 1999).

Hypersensitivities of WS cells to 4-nitroquinoline 1-oxide (4-NQO) and camptothecin (CPT) have been reported (Gebhart et al., 1988; Ogburn et al., 1997; Poot et al., 1999). However, WS cells do not exhibit a hypersensitivity to other DNA-damaging agents such as most alkylating agents, and X-rays, bleomycin, or H₂O₂ that produce reactive oxygen species (Fujiwara et al., 1977; Gebhart et al., 1988; Higashikawa and Fujiwara, 1978; Okada et al., 1998), as well as to UV irradiation (Krepinsky et al., 1979). On the other hand, BS cells show hypersensitivities to N-ethyl-N-nitrosourea, ethyl methanesulfonate, methyl methanesulfonate (MMS) and 4-NQO, as well as UV irradiation (Krepinsky et al., 1979; Kurihara et al., 1987; Shiraishi et al., 1985). These observations suggest that both WRN and BLM may be directly or indirectly involved in DNA repair and may prefer certain types of DNA damage, respectively.

A possible role in recombination is implicated for WRN and BLM as their homologs in yeast, Sgs1 and Rqh1+, and in E. coli RecQ, negatively regulate recombination (Gangloff et al., 1994; Hanada et al., 1997; Stewart et al., 1997; Watt et al., 1995). Notably human WRN and BLM can suppress increased homologous and illegitimate recombination in the sgs1 mutant (Yamagata et al., 1998). WRN and BLM have abilities to recognize Holliday junctions and to efficiently accelerate its ATP-dependent branch migration in vitro, suggesting that WRN and BLM may suppress homologous recombination (HR) by disrupting recombinogenic molecules that arise at sites of stalled replication forks in vivo (Constantinou et al., 2000; Karow et al., 2000). We (Imamura et al., 2001) and Wang et al. (2000) developed BLM^−/− chicken B cell DT40 lines, and showed that the BLM^−/− DT40 cells have a higher sensitivity to several genotoxic agents, increases in the levels of SCE and targeted integration of exogenous DNAs. We also showed that the chicken BLM is involved in early S phase-specific surveillance of damaged adducts of DNA (Imamura et al., 2001). An experiment system using DT40 cell lines enabled us to target several genes in a single cell line for analysing complementary functions of these genes. In this study, we produced WRN^−/− and WRN^−/−/BLM^−/− cells from DT40 in addition to BLM^−/− DT40 cells and studied their phenotypes to further investigate the roles of WRN and BLM on DNA repair. Our study demonstrated that both WRN and BLM are involved in DNA repair in a complementary fashion.

Results

WRN targeting constructs and production of WRN-deficient DT40 clones

A full-length cDNA encoding GdWRN was isolated from a cDNA library prepared from the chicken DT40 cell line. The helicase domain of chicken WRN shares a 75% identity with the helicase domain of human WRN at the amino acid level, and a 38, 42, 29 and 36% identity with the helicase domains of human BLM, RecQL1, RecQL4 and RecQL5, respectively (Figure 1). Further analysis of the sequence of the N-terminal region shows that chicken WRN shares a 73% identity with the exonuclease domain of human WRN at the amino acid level providing evidence that chicken WRN must be an ortholog of human WRN.

Figure 1

(a) Schematic representation of the homology between chicken WRN (GdWRN) and human (Hs) WRN. The darkened areas indicate the locations of the helicase domains shared by the RecQ helicase family. The hatched areas are putative nuclease domains homologous to bacterial RNase D and DNA polymerase I. The light grayish areas are acidic regions contained in most of RecQ helicase family members. The numbers in brackets indicate each area of GdWRN homologous to the areas of HsWRN calculated by the GeneWorks (IntelliGenetics). (b) Amino acid alignments in the helicase domains of GdWRN and HsWRN. Thick lines shown above the sequence indicate helicase motifs. The darkened areas represent identical amino acid species. For convenience of alignment, spaces are inserted

Next, we isolated genomic clones of the WRN locus using long-range PCR amplification with primers designed from the chicken WRN cDNA. The genomic clones were partially sequenced to determine the positions of exons and introns. To produce WRN deletion constructs, approximately 0.8 kb of genomic sequence of the WRN locus was replaced at the helicase motif regions with either the histidinol- or blasticidin-resistance gene (His^R or Bsr^R) (Figure 2a). We expected the targeted integration of these constructs to delete amino acids 615–654. To isolate heterozygous WRN^+/− mutant clones, the WRN-His^R construct was transfected into wild-type DT40 cells. After Southern blot analysis of BamH I/ScaI-digested genomic DNA, drug-resistant clones that had a 3.8-kb band were selected (Figure 2a,b). One WRN^+/− mutant clone was then transfected with the WRN-Bsr^R construct to isolate homozygous WRN^−/− mutant clones. The disruption of the WRN gene was assessed by northern blot analysis and RT–PCR (Figure 2c,d).

Figure 2

Generation of WRN^−/− clones. (a) Schematic representation of the partial restriction map of the chicken WRN locus, the two gene disruption constructs and the configuration of the targeted loci. Black boxes are the positions of exons. Relevant restriction enzyme sites are shown as follows: B, BamHI; Sc, ScaI. (b) Southern blot analysis of wild-type (+/+), heterozygous mutant (+/−) and homozygous mutant (−/−) clones. Genomic DNA digested with BamHI and ScaI was hybridized with the probe DNA shown in (a). (c) Northern blot analysis of total RNA of the indicated genotype after hybridization with a chicken WRN cDNA probe and chicken β-actin probe. An asterisk indicates shortened transcription products resulted from the genome disruption. (d) RT–PCR analysis of purified total RNA from wild-type (+/+) and homozygous mutant (−/−) cells. cDNA was synthesized and was used as a template for amplification with primers that were designed based on the sequence of a coding region containing the area lost by targeting

BLM^−/− clones were produced by replacing approximately 1 kb of genomic sequence in the series of helicase motifs of the BLM locus with either the His^R or Bsr^R as described by Imamura et al. (2001). To produce WRN^−/−/BLM^−/− clones, the two BLM targeting constructs containing a selection marker of either neomycin or puromycin were sequentially transfected into a WRN^−/− clone. The disruption of the BLM gene was also verified using Southern blot analysis (data not shown).

Proliferative properties of mutant cells

The proliferative properties of WRN^−/−, BLM^−/− and WRN^−/−/BLM^−/− mutant clones were monitored by growth curves and cell cycle analyses. WRN^−/− and BLM^−/− cells proliferated at slightly lower rates than wild-type cells, while WRN^−/−/BLM^−/− cells proliferated at a considerably lower rate than either single mutant cells (Figure 3). The approximate doubling times were 9, 10, 12 and 13 h for wild-type cells, WRN^−/−, BLM^−/− and WRN^−/−/BLM^−/− mutant cells, respectively. Although our previous study showed a slight retardation of cell cycle from G₁ to S phase in asynchronously cultured BLM^−/− cells (Imamura et al., 2001), neither the cell cycle profile of WRN^−/− cells nor that of WRN^−/−/BLM^−/− cells analysed by flow cytometry showed obvious differences from the cell cycle profile of wild-type cells in this study (data not shown).

Figure 3

Proliferative characteristics of wild-type, BLM^−/−, WRN^−/− and BLM^−/−/WRN^−/− cells. Growth curves corresponding to the indicated cell cultures in the absence of histidiol, blasticidin, neomycin and puromycin. Shown data are the mean values for three experiments

Increases in sensitivities of mutant cells to genotoxic agents

To examine the sensitivities to various genotoxic agents in a colony survival assay, the mutant cells were grown in medium containing various concentrations of DNA-damaging chemicals or were grown after UV irradiation. Our previous study (Imamura et al., 2001) with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay showed that BLM^−/− cells were hypersensitive to DNA genotoxic agents, such as methyl methanesulfonate (MMS), etoposide and 4-NQO and with UV irradiation. In addition to these genotoxic agents, BLM^−/− cells also showed a hypersensitivity to CPT when compared with wild-type cells in the present study (Figure 4).

Figure 4

Dose-dependent growth inhibition curves showing the sensitivities of asynchronous cells to various genotoxic agents. The y and x axes mean rates of the surviving cells and the doses of each genotoxic agents, respectively. The genotypes of cells are at the right bottom panel. Each value represents the mean of survival rates from three independent experiments. Error bars show the standard deviation of the mean

WRN^−/− cells showed markedly increased sensitivities to 4-NQO, MMS and CPT, and a slightly increased sensitivity to etoposide, compared with wild-type cells. However, these cells showed no marked difference in sensitivity to UV irradiation in the range of 1 to 10 J/m², which markedly affects on the growth of BLM^−/− cells compared with wild-type cells (Figure 4).

WRN^−/−/BLM^−/− cells were more sensitive to 4-NQO, MMS, CPT, etoposide and UV irradiation than either WRN^−/− or BLM^−/− cells (Figure 4). These results suggest that separate DNA repair pathways involving WRN and BLM exist and that both pathways might partially complement each other in repairing DNA damages in DT40 cells.

Chromosomal aberrations associated with DNA damage during progression from late S to G₂ phase in WRN^−/−/BLM^−/− cells

DT40 cells have a stable karyotype with a modal chromosome number of 80 and show no obvious abnormalities except for a trisomy of chromosome 2 and one additional microchromosome (Sonoda et al., 1998). Chromosomal aberrations were hardly detectable in asynchronously grown wild-type, BLM^−/−, and WRN^−/− cells, whereas they were more frequent in WRN^−/−/BLM^−/− cells (Table 1). When cells were irradiated with 1 J/m² of UV light, wild-type cells at the late S to G₂ phase were expected to enter mitosis within 0–6 h after UV irradiation. As reported by our group (Imamura et al., 2001), chromatid-type gaps and breaks were frequent in these cells (Table 1), reflecting the formation of double-stranded DNA (dsDNA) breaks after DNA replication.

Table 1 Frequencies of chromosomal aberrations in BLM-, WRN- and WRN/BLM-deficient DT40 cells after irradiation with UV light

BLM^−/− cells that had been in the G₁ to early S phase on UV irradiation and entered into mitosis 6–12 h after irradiation also showed increased frequencies of chromatid-type gaps and breaks in addition to a marked increase in frequency of chromatid exchanges, i.e., chromosomal quadriradials (Table 1), suggesting that Holliday junctions persisting during the S phase lead to the induction of dsDNA breaks in BLM-deficient cells (Imamura et al., 2001).

In contrast to BLM^−/− cells, WRN^−/− cells showed no obvious change in frequency of chromosomal aberrations after UV irradiation (Table 1), being consistent with the observation that WRN^−/− cells have almost the same susceptibility to UV irradiation as wild-type cells (Figure 4). However, WRN^−/−/BLM^−/− cells that were in the late S to G₂ phase on UV irradiation and entered into mitosis 0–6 h after irradiation showed a marked increase in frequency of chromatid-type breaks and gaps with no increase in frequencies of chromosome-type breaks and gaps and chromatid exchanges. This result implies a possible involvement of WRN in the repair of dsDNA breaks produced at the sites of stalled replication forks.

Targeted integration of exogenous genomic DNA in mutant cells

We examined targeted integration frequencies at the hypoxanthine-guanine phosphoribosyl transferase gene (HPRT) locus by comparing wild-type cells with mutant cells. As we had previously measured the targeting integration frequencies at the β-actin and Ovalbumin loci in BLM^−/− cells (Imamura et al., 2001), the disruption of BLM markedly increased the frequency of targeted integration of the exogenous HPRT without a marked increase in transfection efficiency (Table 2). WRN^−/− cells showed a 2–3-fold increase in the frequency of targeted integration compared with wild-type cells without a marked increase in transfection efficiency (Table 2), suggesting that WRN may suppress spontaneous HR. However, additional disruption of WRN in BLM^−/− cells caused a sixfold increase, but lower, in the frequency of targeted integration in contrast to an approximately 10-fold increase in BLM^−/− cells, suggesting that WRN may accelerate the HR that arises from BLM-deficiency.

Table 2 Targeted integration frequencies in WRN^−/−/BLM^−/− cells

Increased SCE frequency in BLM^−/− cells was reduced by disruption of the WRN gene

BLM^−/− DT40 cells show a 10-fold higher number of SCEs than wild-type cells (Imamura et al., 2001; Wang et al., 2000), suggesting that the defect in chicken BLM is sufficient to cause the high SCE phenotype characteristic of BS cells (Chaganti et al., 1974; German, 1993; Heartlein et al., 1987). A considerable number of the SCEs observed in chicken BLM^−/− cells are formed depending on the Rad54 function, i.e., they are formed by HR (Wang et al., 2000). In this study, wild-type cells had as few as 1.8 SCEs/cell (Figure 5). By contrast, BLM^−/− cells had a high 20.7 SCEs/cell on average as reported by Wang et al. (2000) and Imamura et al. (2001). In contrast to BLM^−/− cells, WRN^−/− cells showed no marked increase in SCE frequency (2.3±1.4 SCEs/cell) compared with wild-type cells. However, additional disruption of WRN in BLM^−/− cells caused a slight reduction of the SCE frequency (14.6±3.9 SCEs/cell), suggesting that WRN may be partially involved in HR increased in BLM^−/− cells.

Figure 5

Distribution of population of wild-type, BLM^−/−, WRN^−/− and BLM^−/−/WRN^−/− cells with indicated frequency of SCEs. Cells were labeled with BrdU during two cell cycles. Spontaneous SCE in the macrosomes of 200 metaphase cells were counted. Histograms show the frequency of cells with the indicated number of SCEs in each cell

Discussion

Complementary involvement of WRN and BLM in DNA repair

As DT40 cell lines, like most transformed chicken cell lines, do not express p53 (Takao et al., 1999), they may proliferate well throughout the cell cycle and are killed by a genotoxic agent without a checkpoint control downstream of p53. The present study demonstrates that BLM^−/− DT40 cells are very similar in their phenotype to human BS cells. BLM^−/− DT40 cells showed a hypersensitivity to several genotoxic agents, such as 4-NQO, MMS, CPT, etoposide, to the same extent as UV irradiation. CPT is known to stabilize topoisomerase I resulting in formation of dsDNA breaks during the S phase (Tsao et al., 1993). Etoposide stabilizes topoisomerase II resulting in formation of a cleavable DNA complex (Zhang et al., 1992). The other chemical genotoxins cause specific types of DNA damages. All of them contribute to the formation of certain types of mutagenic adducts in DNA (Galiegue-Zouitina et al., 1984; Smith and Grisham, 1983). UV irradiation also forms DNA adducts such as cyclobutane pyrimidine dimers (Tornaletti and Pfeifer, 1996). Usually, DNA adducts are repaired by the nucleotide-excision repair (NER) pathway (Wood, 1989). However, large numbers of DNA adducts may be overlooked by the NER pathway and may subsequently produce strand discontinuities during replication, leading to chromatid-type gaps and breaks (Sonoda et al., 1999; Yamaguchi-Iwai et al., 1999). Such strand discontinuities can be repaired by HR with the sister chromatid after replication, as is the case for recombinational repair in yeast cells (Kadyk and Hartwell, 1993). BLM has an ability to recognize DNA adducts or DNA structures associated with stalled replication forks, such as Holliday junctions, and to disrupt these DNA secondary structures leading to recombination repair (Imamura et al., 2001). In the presence of BLM, HR after replication might be prevented by rapid disruption of aberrant DNA secondary structures.

Transformed B-lymphoblastoid cells from WS patients are hypersensitive to 4-NQO and CPT (Gebhart et al., 1988; Ogburn et al., 1997; Poot et al., 1999). WS cells are also hypersensitive to etoposide (Elli et al., 1996). However, transformed B-lymphoblastoid cells from WS patients do not exhibit a hypersensitivity to other DNA-damaging agents, such as UV irradiation, most other alkylating agents, bleomycin, H₂O₂ or X-ray irradiation (Fujiwara et al., 1977; Gebhart et al., 1988; Higashikawa and Fujiwara, 1978; Okada et al., 1998). In the present study, WRN^−/− DT40 cells showed marked hypersensitivity to MMS, etoposide, 4-NQO and CPT, but not to UV irradiation. These phenotypes characterized in WRN^−/− DT40 cells are similar to the phenotypes in transformed B-lymphoblastoid cells from WS patients. WRN has been shown to bind to and/or functionally interact with RPA (Brosh et al., 1999; Constantinou et al., 2000), proliferating cell nuclear antigen (PCNA), DNA topoisomerase I (Lebel et al., 1999), Ku 86/70 (Cooper et al., 2000; Orren et al., 2001), and DNA polymerase δ (Kamath-Loeb et al., 2000). Each of these interacting proteins is involved in DNA manipulations including those that resolve alternative DNA structures or repair DNA damages. The synergistic increases in sensitivities of WRN^−/−/BLM^−/− DT40 cells to genotoxic agents suggest the complementary involvement of WRN and BLM in repair of damaged DNAs. Moreover, the hypersensitivity of BLM^−/− cells to UV irradiation is enhanced by the concomitant disruption of WRN despite the fact that WRN^−/− cells show no increase in sensitivity to UV irradiation. These findings suggest that WRN may have a role on DNA transitions in a manner different from BLM. This explanation might be supported by characteristics unique to WRN where WRN exonuclease preferentially hydrolyzes alternative DNA that contains bubbles, extra-helical loops, 3-way or 4-way junctions (Machwe et al., 2000; Shen and Loeb, 2000). In addition, WRN as well as BLM can resolve aberrant DNA structures such as G-quadruplex and G-triplex DNAs (Brosh et al., 2001; Sun et al., 1998).

The synergistic increase in sensitivity of WRN^−/−/BLM^−/− DT40 cells to UV irradiation is also supported by the data from karyotypic analyses. As previously reported (Imamura et al., 2001), the fact that UV irradiation in the G₁ to early S phase causes chromosomal aberrations such as chromatid-type gaps and breaks in BLM^−/− cells suggests BLM may have a role in resolving DNA secondary structures occurring at stalled forks in the early S phase. In the present study, chromatid breaks were frequently observed in WRN^−/− cells, as well as in wild-type cells, when irradiated with UV light during the S to G₂ phases. In WRN^−/−/BLM^−/− cells, UV irradiation in the late S to G₂ phase synergistically enhanced the increases in the number of chromatid-type gaps and breaks without any remarkable change in the number of chromosome-type breaks and gaps. These results imply that WRN may, in the late S to G₂ phase, contribute to repair of dsDNA breaks which occur due to the lack of rapid resolution of aberrant DNA secondary structures by BLM. Therefore, the sensitivity of DT40 cells to UV irradiation and to other DNA damaging agents might be increased synergistically when both BLM and WRN were disrupted. In addition, the marked increase in frequency of spontaneous chromosomal aberrations leading to cell death may explain the decreased rate of proliferation of WRN^−/−/BLM^−/− cells without any change in cell cycle profile.

Possible involvement of WRN in homologous recombination

In BS cells, the increases in frequencies of SCEs and interchange between homologous chromosomes are observed (Chaganti et al., 1974; German, 1993). Like the yeast sgs1 disruptants that show an increased frequency of spontaneous recombination (Watt et al., 1996; Onoda et al., 2001), BLM^−/− DT40 cells have a hyper-recombination phenotype resulting in increased frequencies in both SCEs and targeted genome integration (Imamura et al., 2001). Sonoda et al. (1999) demonstrated that HR between sister chromatids is the primary mechanism for SCE in DT40 cells. The level of SCE in BLM^−/− DT40 cells is considerably reduced in the absence of Rad54 (Wang et al., 2000). These findings suggest that a large number of the SCEs in BLM^−/− cells occur via HR.

WRN^−/− DT40 cells also showed an increase in incidence of spontaneous recombination indicated by increased frequency of targeted genome integration. However, disruption of WRN partially diminished the SCE frequency that increased in BLM^−/− DT40 cells despite the fact that the SCE frequency did not change in WRN^−/− DT40 cells compared with wild-type cells, suggesting that WRN might be partially involved in accelerating the post-replicational HR repair occurring in BLM^−/− DT40 cells. The above speculation is compatible with the recent report that the close proximity of a pair of sister chromatids allows efficient HR-mediated repair of dsDNA breaks during late S to G₂ phases (Sonoda et al., 2001). A possible role of WRN on HR repair contrasts with the function of BLM that may prevent HR by disrupting recombinogenic molecules arising at sites of stalled replication forks which lead to dsDNA breaks (Karow et al., 2000). Suppression of spontaneous HR by WRN, that is indicated by the increase in targeted genome integration upon the disruption of WRN, may be explained by an ability of WRN to unwind a DNA triple helix, as well as BLM, which is a joint molecule formed by aberrant recombination between exogenous DNA and its homologous genomic DNA (Brosh et al., 2001; Constantinou et al., 2000). Because human WRN and BLM can suppress the increased homologous and illegitimate recombination in the yeast sgs1 mutant (Yamagata et al., 1998), both WRN and BLM have been considered as suppressors of HR. We believe this is the first report of a possible function of WRN to accelerate HR in response to DNA damage in vertebrate cells.

In summary, our results with chicken DT40 cells suggest (1) BLM accelerates branch migration of Holliday junctions encountering stalled replication forks and may occasionally overlook some stalled forks in which dsDNA breaks occurred and (2) WRN may accelerate post-replicational recombination repair of dsDNA breaks occurring at stalled forks. Thus, WRN and BLM may contribute to maintain genomic stability in a complementary fashion, resulting in the prevention of tumorigenesis.

Materials and methods

Cloning the chicken homolog of WRN gene

Degenerate primers corresponding to the amino acid residues in motifs I (TGGGKSLC) and V (ATIAFGMG) of human WRN were used to clone a cDNA fragment coding for a portion of the helicase domain of chicken WRN (GdWRN). The polymerase chain reaction (PCR) product was subcloned into a pGEM-T vector (Promega, Madison, WI, USA). The cloned cDNA fragment was used to screen a cDNA library of avian leukosis virus-induced B cell line DT40 (Buerstedde et al., 1990) to obtain full-length chicken WRN cDNA clones (GenBank/DDBJ/EMBL, Accession No. AB035866), which were verified by DNA sequencing.

Plasmid constructs

The 10-kb genomic chicken WRN locus was amplified from DT40 genomic DNA by long-range PCR using primers designed for the chicken WRN cDNA sequence. The positions of the exons and introns were located by sequencing. The chicken WRN disruption constructs WRN-his and WRN-bsr were made by replacing approximately 0.8 kb of BamHI/BamHI-linked genomic sequence with histidinol- or blasticidin-selection marker cassettes His^R or Bsr^R under the control of the β-actin promoter (Bezzubova et al., 1997). The WRN disruption constructs were made linear before transfection to DT40 cells by electroporation. The chicken BLM targeting constructs were made as described by Imamura et al. (2001) using His^R, Bsr^R, puromycin- and neomycin-selection markers.

Gene targeting and cell culture

DT40 cells were maintained in RPMI#1640 medium supplemented with penicillin, streptomycin, 10% fetal bovine serum and 1% chicken serum (Sigma, St. Louis, MO, USA) at 37°C. For DNA transfection, 10⁷ cells were suspended in 0.5 ml of phosphate-buffered saline (PBS) containing 30 μg of linearized plasmid and were electroporated using a GENE Pulser apparatus (BioRad, Hercules, CA, USA) at 550 V and 25 μF. Following electroporation, the cells were transferred to 20 ml of fresh medium and were incubated for 24 h. The cells were then resuspended in 90 ml of medium containing the appropriate drugs and divided into four 96-well microtiter plates. After 7–10 days, drug-resistant colonies were transferred to 24-well plates. Gene disruption was confirmed by Southern and Northern blot analysis in addition to reverse transcriptase (RT)–PCR.

Northern blot analysis

10⁷ cells were washed once with PBS and the total RNA was extracted using TRIzol Reagent (Gibco-BRL, Grand Island, NY, USA). RNA (20 μg/lane) was separated in a 1.2% formaldehyde gel, was transferred to a nylon membrane, and was hybridized with a ³²P-labeled chicken WRN or BLM cDNA fragment.

Measurement of sensitivity of cells to genotoxic agents

To determine the sensitivity to genotoxic agents, 50 to 5×10⁵ cells were plated in triplicate onto 6-well clusters with 5 ml/well of 1.5% (w/v) methylcellulose (Aldrich, Milwaukee, WI, USA) plates containing D-MEM/F-12 (Gibco-BRL) with 15% fetal calf serum and 1.5% chicken serum in the presence or absence of various concentrations of genotoxic agents. Colonies resistant to the genotoxic compound were counted 7–10 days after inoculation. To measure the UV light sensitivity, cells were irradiated with various doses of UV light (λ=254 nm) before plating. The percentage survival was calculated relative to the numbers of colonies from untreated cells.

Karyotype analysis

The karyotype was analysed as described by Sonoda et al. (1998). Briefly, cells were treated for 3 h with a medium containing 0.1 μg/ml colcemid (Gibco-BRL) before harvesting. The harvested cells were incubated in 1 ml of 0.9% sodium citrate for 15 min at room temperature and were fixed with 5 ml of a freshly prepared 3 : 1 mixture of methanol and acetic acid. The cell suspension was dropped onto a glass slide, which had been wetted with 50% ethanol, and was immediately flame-dried. The cells on the slides were stained with 3% Giemsa solution at pH 6.4 for 10 min.

Measurements of targeted integration frequencies

To analyse the targeted integrations at the HPRT locus, the disruption DNA cassette containing the hygromycin-resistance gene within the locus (Fukagawa et al., 1999) was transfected into cells and Southern blot analysis was used for selecting clones resistant to hygromycin.

Measurement of SCE levels

Cells were cultured together with 10 μM BrdU for 18–28 h (two cell cycles) at 37°C and were incubated with 0.1 μg/ml colcemid for the last 3 h. The harvested cells were treated with 75 mM KCl for 15–30 min at room temperature and were subsequently fixed with a freshly prepared 3 : 1 mixture of methanol and acetic acid for at least 30 min. The cells were fixed onto glass slides wetted with 50% ethanol, and were dried at 40–42°C. The dried slides were incubated with 10 μg/ml of Hoechst 33258 in PBS (pH 6.8) for 20 min, followed by rinsing with MacIlvaine solution (Sonoda et al., 1998). The slides were irradiated with black light (λ=352 nm) for 60 min, and were incubated in 2×SSC solution at 62°C for 1 h before staining with 3% Giemsa solution (pH 6.8) and were then subjected to microscopic observation.