¹Department of Molecular and Structural Biology, University of Aarhus, DK-8000 Aarhus C, Denmark

²Laboratory of Molecular Gerontology, National Institute on Aging, National Institutes of Health, Baltimore, Maryland, MD 21224, USA

Correspondence to: V A Bohr, Laboratory of Molecular Gerontology, National Institute on Aging, NIH 5600 Nathan Shock Drive, MD 21224, USA; E-mail: Vbohr@nih.gov

^aCurrent address: Laboratoire de Neurobiologie Moleculaire, Centre National de la Recherche Scientifique, Unite de Recherche Associee 2182, Institut Pasteur, 28 Rue du Dr. Roux, 75724 Paris Cedex 15, France

^bCurrent address: MRC Radiation and Genome Stability Unit, Medical Research Council, Harwell, Oxfordshire, OX11 ORD, UK

Cockayne syndrome (CS) is an autosomal recessive human disease characterized by UV-sensitivity as well as neurological and developmental abnormalities. Two complementation groups have been established, designated CS-A and CS-B. Traditionally, CSA and CSB have been ascribed a function in the transcription-coupled repair (TCR) pathway of nucleotide excision repair (NER) that efficiently removes bulky lesions from the transcribed strand of RNA polymerase II transcribed genes. To assess the role of the CSB protein in the repair of the highly mutagenic base lesion 7,8-dihydro-8-oxoguanine (8-oxoG), we have investigated the removal of this lesion using an in vitro incision approach with cell extracts as well as an in vivo approach with a modified protocol of the gene-specific repair assay, which allows the measurement of base lesion repair in intragenomic sequences. Our results demonstrate that the integrity of the CSB protein is pivotal for processes leading to incision at the site of 8-oxoG and that the global genome repair (GGR) of this lesion requires a functional CSB gene product in vivo.

Oncogene (2002) 21, 3571-3578. DOI: 10.1038/sj/onc/1205443

oxidative DNA damage; repair; DNA instability; Cockayne syndrome

Introduction

Cockayne syndrome (CS) is a rare autosomal recessive genetic human disease with diverse clinical symptoms including severe impairment of physical development, progressive neurological degeneration, accelerated aging and a hypersensitivity to sunlight (Nance and Berry, 1992; Woods, 1998). Two complementation groups of CS have been described, designated CS-A and CS-B (Lehmann, 1982), and the corresponding genes have been cloned (Henning et al., 1995; Troelstra et al., 1992). Although the precise roles of CSA and CSB are yet undetermined, it is clear that they have a key function in promoting the repair of DNA lesions, which pose an obstacle to the progressing RNA polymerase II during gene transcription. It has been shown that CS cells are deficient in the nucleotide excision repair (NER) pathway that removes ultraviolet light (UV) induced lesions from the transcribed strand of active genes (van Hoffen et al., 1993; Venema et al., 1990). Defective transcription-coupled repair (TCR) in CS cells is not only observed after UV irradiation, but also after oxidative stress (Cooper et al., 1997; Le Page et al., 2000b). The clinical features of CS are difficult to reconcile with a repair deficiency limited to TCR of UV-induced or bulky lesions. Consequently, it has been speculated that subtle defects in the repair of oxidative DNA lesions may contribute to the complex phenotype of afflicted individuals (Cooper et al., 1997; Dianov et al., 1997, 1999; Le Page et al., 2000b; Nouspikel et al., 1997).

Recent experiments from our laboratory as well as others have established that CS-B cells have a reduced capacity to repair certain oxidative lesions. Using an in vitro approach with whole cell extracts (WCE) and oligonucleotide-based duplex substrates containing site-specific lesions, we demonstrated earlier that human CS-B cells have a reduced capacity to make incisions at the base lesion 8-oxoG compared to extracts from normal cells (Dianov et al., 1999). In a plasmid-based shuttle vector system, it has also been shown that CS cells exhibit reduced repair of 8-oxoG, which leads to a higher mutation frequency (Le Page et al., 2000b).

All living organisms are constantly being exposed to endogenous and environmental reactive oxygen species (ROS), which are highly genotoxic by their potential to induce DNA damage. 8-oxoG is a base lesion caused primarily by oxidative stress from endogenous metabolic processes and it is repaired mainly by the base excision repair (BER) pathway (Dianov et al., 1998; and for reviews Bohr and Dianov, 1999; Le Page et al., 1999). A possible increase in the load of unrepaired oxidative DNA lesions in CS cells will eventually lead to replicational and transcriptional deficiencies that may potentially contribute to the detrimental features displayed in CS patients (for review see Nance and Berry, 1992; Rolig and McKinnon, 2000).

The 8-oxoG lesions are thought to cause GCTA transversion mutations if left unrepaired during replication (Cheng et al., 1992), and the ubiquitous expression throughout evolution of a specific enzyme that recognizes and removes 8-oxoG bases from DNA emphasizes the importance of preventing the mutagenic potential of this lesion. In mammalian cells, the oxoguanine DNA glycosylase 1 (OGG1) protein seems to be the major glycosylase that efficiently removes 8-oxoG bases from nontranscribed DNA (Klungland et al., 1999; Le Page et al., 2000a) whereas the repair proteins involved in repairing 8-oxoG positioned in transcribed sequences remain more elusive. However, the existence of more than one OGG enzyme and the presence of other glycosylases could contribute significantly to the excision of 8-oxoG lesions in these sequences (Bessho et al., 1993; Hazra et al., 1998).

Sequence determination of the CSB gene product revealed that it belongs to the SWI/SNF family of proteins, which all contain seven putative helicase/ATPase motifs (Gorbalenya et al., 1989; Troelstra et al., 1992). However, in conventional strand displacement assays CSB failed to exhibit any helicase activity (Selby and Sancar, 1997b; Tantin et al., 1997). Targeting two different ATPase motifs by site-directed mutagenesis, we and the Hoeijmakers laboratory have demonstrated that the ATPase domain of CSB is essential for its function in TCR-dependent and -independent pathways (Brosh et al., 1999; Citterio et al., 1998). Besides the ATPase domain, the CSB gene also encodes an N-terminal region with an extremely high content of acidic amino acids (Troelstra et al., 1992). We previously demonstrated that this region does not have any significant impact on TCR and non-TCR pathways after UV induced DNA damage (Brosh et al., 1999; Sunesen et al., 2000). However, the contribution of the ATPase domain or the acidic portion of the protein in processing oxidative DNA damage was not addressed in those studies.

To gain insight into the molecular-genetic significance of the CSB gene product in the repair of oxidative DNA damage, we have directly examined the function of this protein in in vitro incision assays as well as in in vivo gene-specific repair assays. Previously established UV61 cell lines (the hamster homologue of CS-B) harboring mutations in the acidic region or the ATPase domain II (Brosh et al., 1999; Sunesen et al., 2000) have been used in this study.

Assessing gene-specific repair experiments, we have exploited the polar photosensitizer RO19-8022 (RO), which predominantly induces 8-oxoG base modifications, upon light exposure to hamster cells (Will et al., 1999). The E. coli formamidopyrimidine-DNA glycosylase (Fpg), which primarily recognizes and removes 8-oxoG (Boiteux et al., 1992), was previously implemented to measure the induction and removal of lesions in mitochondrial sequences (Anson et al., 1998). Here, we have used the Fpg enzyme to measure 8-oxoG levels in intragenomic sequences of nuclear DNA of stably isogenic UV61 cell lines, stably transfected with site-specific CSB alleles.

Results

We have previously shown that extracts prepared from human CS-B cells harbor normal levels of thymine glycol (Tg) and uracil incision activity, whereas the level of 8-oxoG incision is reduced (Dianov et al., 1999). We also demonstrated that the deficiency could be complemented by transfection of the wild type CSB gene (Dianov et al., 1999). To determine the repair status of the Chinese hamster ovary (CHO) cell line UV61 (the hamster homologue of CS-B), the repair incision activity was measured on single lesion constructs. WCE were prepared from UV61 cells transfected with wild type and mutant CSB alleles (Figure 1) in order to evaluate the relative importance of the ATPase domain and the acidic region of the CSB protein in the repair of oxidative lesions.

Reduced 8-oxoG incision activity of the ATPase mutant of CSB

To determine the OGG1 enzyme activity, we conducted a series of incision assays with extracts from the UV61 cell lines. Mammalian OGG1 removes both the 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine and 8-oxoG from DNA and cleaves the DNA strand 5' of the lesion by its associated AP lyase activity, introducing a single-stranded nick (Bjoras et al., 1997; Boiteux and Radicella, 1999; Krokan et al., 1997; Roldan-Arjona et al., 1997). The incision assays were performed on oligonucleotides containing either no lesion, a thymine glycol, a uracil, or an 8-oxoG lesion at a defined site (Figure 2a). We did not observe any difference in the uracil incision activity (Figure 2b, lanes 1 to 4) or in the thymine glycol incision activity (data not shown). However, the 8-oxoG incision level was about 60% lower in UV61/pc3.1 cells compared to UV61/pc3.1-CSBwt (compare Figure 2b, lanes 11 and 12, and 2c). These data are consistent with our previous observations with human cell extracts (Dianov et al., 1999), which suggests that deficient incision of 8-oxoG is a general feature of CS-B cell lines. Interestingly, when extracts were prepared from a UV61 cell line transfected with the CSBE646Q allele, which contains a point mutation in motif II of the ATPase domain, the 8-oxoG incision activity was at an intermediate level being significantly different from both UV61/pc3.1 (P<0.05, t-test) and UV/pc3.1-CSBwt (P<0.05) (Figure 2b, lanes 9 and 12, and 2c). In contrast to the CSBE646Q allele, CSBAc conferred full complementation of the low incision level of UV61/pc3.1 WCE (Figure 2b, lanes 10 and 2c). These experiments suggest that the ATPase domain of CSB encodes a significant but not essential function of the CSB protein in repair of 8-oxoG lesions, whereas the acidic region of CSB is completely dispensable.

RO induction of Fpg-sensitive sites in DHFR

Since no transcriptional processes are taking place in the assay conditions for the in vitro incision assay, these assays are generally thought to reflect global genome repair (GGR) (Mu et al., 1996; Shivji et al., 1994). Having shown in vitro that WCEs prepared from rodent as well as from human CS-B cell lines harbor a lower level of 8-oxoG incision activity, we next sought to assay whether the repair of oxidative lesions positioned in transcriptionally active genes are affected by CSB mutations. Recent data from plasmid-based repair experiments have suggested that 8-oxoG lesions are removed from DNA by a transcription-coupled repair mechanism (Le Page et al., 2000b). To address this question directly in intragenomic nuclear DNA, we set up gene-specific experiments to determine the removal of Fpg-sensitive sites (FSS) after oxidative DNA damage induction in the UV61 transfectant cell lines.

In order to induce oxidative DNA damage, the UV61 transfectant cell lines were exposed to the photosensitizer RO19-8022 (RO) plus light. When exposed to RO plus light (see Materials and methods), relatively high levels of FSS were generated in the housekeeping DHFR gene (initial lesion frequency 0.59±0.16 FSS per restriction fragment) (Figure 3, lanes 7 and 8).

No significant difference in initial lesion frequency was observed between the employed cell lines (data not shown).

Consistent with previously reported results (Will et al., 1999), no lesions could be detected by enzymatic digestion of genomic DNA from non-treated cells (Figure 3, lanes 1 and 2) or when either RO treatment (Figure 3, lanes 3 and 4) or light exposure (Figure 3, lanes 5 and 6) was omitted.

Removal of Fpg-sensitive sites in DHFR is dependent on a functional CSB protein

Using this experimental approach, we analysed the repair kinetics of FSS in the actively transcribed DHFR gene of the UV61 transfectant cell lines. Interestingly, 42 and 56% of all the FSS were repaired after 6 and 8 h, respectively, in the UV61/pc3.1-CSBwt cell line (Figure 4), whereas only 14 and 19% respectively were repaired at the same time points in the UV61/pc3.1 cell line. Thus, correlating with the observed in vitro incision data, the CSBwt allele confers a highly significant 2-3-fold stimulation of the repair capabilities exhibited by UV61.

Also corresponding with the observed in vitro incision data, we found the repair capability of the UV61/pc3.1-CSBE646Q cell line to be at an intermediate level (33% after 6 h and 38% after 8 h, Figure 4) and the UV61/pc3.1-CSBAc cell line to be comparable to UV61/pc3.1CSBwt (38% after 6 h and 50% after 8 h, Figure 4). These findings suggest that the acidic domain is dispensable for CSB in repair of oxidative damage to nuclear DNA of an actively transcribed gene. A mutation in the highly conserved ATPase domain in CSB, on the other hand, seems to be partially detrimental to the repair of Fpg-sensitive sites. This is the first direct demonstration that CSB is involved in the repair of oxidative lesions in actively transcribed intra-genomic sequences in rodent cells.

Collectively, our results demonstrate that the integrity of the CSB protein is fundamental for the removal of the highly mutagenic 8-oxoG base lesion from the overall genome.

RO survival

Using the classical clonogenic survival assay, we wanted to determine whether the observed results were obtained due to different survival rates of the UV61 transfectant cell lines after RO treatment. As shown in Figure 5, we do not find any significant difference in the survival rate: UV61/pc3.1 was as RO resistant as UV61 transfected with either the CSBwt or the CSBE646Q allele. These data also suggest that the reduced 8-oxoG repair capacity of UV61/pc3.1 is not a critical parameter for cell survival after RO induced DNA damage.

Discussion

The BER pathway normally processes the unavoidable but potentially detrimental DNA alterations induced by endogenous ROS. The repair pathway is initiated by various different DNA glycosylases, which individually recognize and excise a relatively narrow subset of aberrant DNA bases (for recent reviews, see Krokan et al., 2000; Memisoglu and Samson, 2000). In the present study, we have demonstrated that the in vitro incision activity towards one of the most important base lesions, 8-oxoG, is greatly compromised when WCE from the hamster CS-B homologue cell line UV61 is used. This deficiency could be rescued by transfection of the wild type CSB gene. Whereas the acidic region is dispensable for the function of CSB in the repair steps leading to incision, a CSB allele carrying a point mutation in the highly conserved ATPase motif II did not fully complement the deficient 8-oxoG incision activity of UV61 WCE. However, some significant residual activity remained and it may be speculated that the CSB-E646Q mutant may stabilize some protein-protein interactions important for the incision reaction. Also, the ATPase activity of the CSB-E646Q mutant have not been characterized and we can not exclude the possibility that some residual ATP hydrolysis may account for the residual 8-oxoG incision activity from WCE of the UV61/pc3.1-CSBE646Q cell line.

In previous experiments with human CS-B cell lines, we found similar results: an approximate twofold increase in incision level of 8-oxoG containing oligonucleotide substrates when extracts of CS1AN (CS-B) cells complemented with the CSBwt gene were used (Dianov et al., 1999). We show here that the in vitro incision deficiency extends to rodent cells as well, which suggests that the deficiency is a general defect of CS-B cell lines in the processing of oxidatively damaged DNA. We speculate that the CSB protein is directly involved in the steps of 8-oxoG repair that leads to the incision. We are currently investigating the hypothesis that CSB plays an upstream role in the BER incision process of hOGG1. This could, for example, be via a function as a transcription factor and CSB has been shown to play a role in transcription (Balajee et al., 1997; Dianov et al., 1997; Selby and Sancar, 1997a; Tantin, 1998; van Gool et al., 1997). Alternatively, CSB may directly interact with proteins (such as OGG1) involved in the BER incision complex.

Recent reports have ascribed a TCR pathway for the removal of oxidative lesions from DNA. Cooper et al. (1997) found that Tgs are repaired in TCR fashion from the actively transcribed MT1A gene, and that CS-B cells are deficient in this form of TCR (Cooper et al., 1997). Furthermore, recently Le Page et al. (2000b) exploited a plasmid-based shuttle vector system to show that besides Tg, 8-oxoG is also strand-specifically removed from DNA, and that the removal is defective in CS cells, including CS-B, XP-B/CS, XP-D/CS, and XP-G/CS (Le Page et al., 2000b). Surprisingly, the defect was limited to the transcribed strand of the active gene examined.

In this study, we have examined the repair of FSS in intragenomic sequences of hamster UV61 cell lines. After induction of lesions with RO plus light treatment, the repair kinetics of the DHFR gene was significantly compromised in UV61 cells compared to UV61 transfected with the wild type CSB allele, which is in accordance with the results of Le Page et al. (2000b). However, together with our in vitro incision data, our results indicate that the deficit arises as a consequence of deficient GGR of 8-oxoG, whereas Le Page et al. (2000b) found that only the TCR pathway for the removal of 8-oxoG was affected in CS cells (Le Page et al., 2000b). Consistent with the findings described in this report, we and others have previously demonstrated that repair of FSS is not changed significantly whether being positioned in transcribed or nontranscribed sequences (Grishko et al., 1997; Taffe et al., 1996; Thorslund et al., 2002), which suggests that the major repair mechanism of this lesion is not TCR dependent. On the other hand, from the experiment described in this report, we are not able to exclude the possibility of an analogous TCR mechanism of 8-oxoG that operates in parallel with the GGR system. However, very recent 8-oxoG-repair experiments from our group do not suggest this to be the case, since the Fpg-sensitive sites were not removed preferentially from the transcribed strand of DHFR in hamster CHOWT cells (Thorslund et al., 2002).

CS cells are hypersensitive to UV (Mayne and Lehmann, 1982). It has been speculated that the reduced survival rate may be caused by blockage of RNA polymerase II progression by UV induced photoproducts in the transcribed strand (Ljungman and Zhang, 1996). The situation is more controversial for oxidative lesions. For one particular oxidative base lesion, Tg, it has been demonstrated that it blocks ongoing transcription when positioned in the template strand (Bessho, 1999; Htun and Johnston, 1992), and is hence repaired by a TCR mechanism (Cooper et al., 1997). However, Tg is unusual in that (1) it causes more helix distortion than other forms of oxidative DNA damage (Kung and Bolton, 1997) and (2) it has been shown that the human NER system that normally processes the bulky UV induced photoproducts may be able to serve as a backup mechanism for the repair of Tg (Kung and Bolton, 1997; Lin and Sancar, 1989; Reardon et al., 1997). Whether the 8-oxoG base lesion, which only induces minor alteration in the structure of the double helix, blocks the progressing RNA polymerase like Tg is still not clear. Results on plasmid DNA have been obtained, which suggest that 8-oxoG actually inhibits the RNA polymerase progression through the site of the lesion (Le Page et al., 2000b). Whether 8-oxoG or an 8-oxoG binding activity poses such an obstacle to RNA polymerase II in genomic sequences is still elusive. On the contrary, it has been demonstrated that 8-oxoG can be bypassed by DNA polymerases (Cheng et al., 1992) and E. coli RNA polymerases (Viswanathan and Doetsch, 1998), which does not favor a model of TCR that invokes arrest of RNA polymerase II by an 8-oxoG adduct.

We also determined the repair kinetics of UV61/pc3.1-CSBE646Q and UV61/pc3.1-CSBAc after RO plus light treatment. Whereas the CSB allele harboring the acidic deletion complemented the repair deficiency of UV61 to a level equal to the CSB wild type allele, the ATPase mutant failed to reach the same level of repair after 8 h. These data suggest that the ATPase activity of the CSB protein has some, albeit minor, function in the GGR pathway of oxidative lesions. On the contrary, we have now characterized the acidic deletion mutant after both UV radiation (Sunesen et al., 2000) and oxidative DNA damage induction, without being able to detect deficiencies in repair or survival. Whether this distinctive region, which encodes approximately 60% acidic amino acids in a stretch of 39 amino acids, has its major impact in more direct transcription related functions of CSB has to be further investigated.

Knockout mice deficient in 3-methyladenine-DNA-glycosylase or OGG1 accumulate base lesions in their genomes but are still as viable as normal mice (Engelward et al., 1997; Klungland et al., 1999). Also, we fail to detect an increased sensitivity in viability to RO-induced DNA damage of the UV61 cell line despite the repair deficiencies. These observations suggest that deficient oxidative DNA damage processing does not directly affect the survival rate as a consequence of a genotoxic dose. Rather, the cytotoxicity of the UV61 cell lines upon treatment with the photosensitizer RO probably results from damage to other cellular constituents, such as membranes or proteins.

In conclusion, these studies have demonstrated that CSB is a pivotal constituent of GGR of 8-oxoG and that the integrity of the ATPase domain has a significant role, whereas the acidic region seems to be dispensable for repair related functions of CSB in NER as well as BER. Since CS patients suffer from dramatic neurodegeneration, it has been difficult to reconcile CSB with defective TCR of UV lesions only. Certainly, as described above, recent evidence suggest that defective repair of oxidative lesions may be a key element in CS etiology. Future studies of CSB and its interactions should provide further insight into the role of CSB in repair of oxidative lesions.

Materials and methods

Cell lines and culture conditions

The UV61 cell lines used in this study, their selection, and their UV-repair characteristics were previously described (Brosh et al., 1999; Sunesen et al., 2000). The UV61 cell line, derived from AA8, belongs to rodent complementation group 6 and is homologous to human CS-B cells. UV61 cells transfected with the mammalian expression vector pcDNA3.1 (abbreviated as pc3.1, Invitrogen, San Diego, CA, USA) or pc3.1 containing the wild type human CSB gene are designated UV61/pc3.1 and UV61/pc3.1-CSBwt, respectively. UV61 cells transfected with pc3.1-CSB containing a 39 amino acid deletion of the acidic region are designated UV61/pc3.1-CSBAc, whereas UV61 transfected with a point mutation in motif II in the ATPase domain of the CSB protein (E646Q) is designated UV61/pc3.1-CSBE646Q. CSB expression and UV-sensitivity have been previously characterized for the entire cell lines used in this study (Brosh et al., 1999; Sunesen et al., 2000). All the isogenic cell lines were routinely grown in Ham's F-10 and DMEM (1 : 1) media containing 400 g/ml geneticin (Gibco BRL, Paisley, UK).

Oligonucleotides

The oligonucleotides used in this study were purchased from DNA Technology (Aarhus, DK). The following oligonucleotides were employed: uracil; ATATACCGCGGUCGGCCGATCAAGCTTAT, no lesion; ATATACCGCGGCCGGCCGATCAAGCTTAT, 8-oxoG; ATATACCGCG[8-oxoG]CCGGCCGATCAAGCTTAT, Tg; CCAGCGCACGACGCATgGCACGACGACCGGG. All oligonucleotides were purified on a 20% polyacrylamide gel, ³²P-5'-end labeled and annealed with the complementary strand as previously described (Dianov et al., 1999). To generate an oligonucleotide containing a single thymine glycol lesion, 2 g of an oligonucleotide harboring a unique thymine residue was incubated in 50 l of 100 mM osmium tetroxide containing 2% pyridine for 30 min at room temperature. The oligonucleotide was subsequently purified by gel filtration using Sephadex-G25 (Amersham-Pharmacia).

Incision assay

Whole cell extracts were prepared by the method of Manley et al., 1983). Incision reactions (25 l) contained approximately 0.5 nmol of oligonucleotide duplex, 45 mM HEPES-KOH, pH 7.8, 70 mM KCl, 1 mM DTT, 2 mM EDTA, 100 ng of random single-stranded competitor DNA, and 50 g of whole cell extract. The reaction mixture including the uracil containing duplex was incubated at 37°C for 1 h while the thymine glycol and 8-oxoG oligonucleotides were incubated with extracts for 3 h at 37°C. Repair incision reactions were terminated by the addition of 2 l 0.5 M EDTA. After the reaction, samples were treated with proteinase K, phenol-chloroform extracted, 20 l of formamide-dye (0.1% xylene cyanol, 0.1% bromophenol blue dissolved in 100% deionized formamide) was added, incubated for 5 min at 90°C, and electrophoresed on a 20% polyacrylamide gel containing 7 M urea, 89 mM Tris-borate (pH 8.0) and 2 mM EDTA. The incision reaction products were visualized by Phosphorlmager analysis and quantitated using the ImageQuant software.

Induction of oxidative DNA base damage in cell culture

Exponential phase cells were incubated in RO19-8022 (RO) (250 M, OD₄₂₅ 0.4) in phosphate buffered saline with glucose (PBS-G) (140 mM NaCl, 3 mM KCl, 8 mM Na₂HPO₄, 1 mM KH₂PO₄, 1 mM CaCl₂, 0.5 mM MgCl₂, 0.1% Glucose, pH 7.4) at 37°C for 45 min. The RO treated cells were exposed to visible light from a 1000 W halogen lamp at 15 cm distance on ice for 5 min. Cells were washed twice in PBS before repair incubation in normal growth medium.

Gene-specific repair

The handling of cells after treatment with RO and the subsequent isolation of genomic DNA was essentially performed according to previously published techniques (Anson et al., 1998; Bohr and Okumoto, 1988). DNA was resuspended in TE and the concentration was measured spectrophotometrically at 260 nm.

The first repair incubations were performed with bromodeoxyuridine and fluorodeoxyuridine in the media to allow subsequent determination of the contribution of replication via CsCl gradient separation. However, no significant replication occurred within the 8 h repair incubation and this step was therefore subsequently omitted.

The E. coli Fpg protein excises primarily 8-oxoG and faPy lesions and cleaves the DNA backbone 3' of the base; thus, incisions in DNA at Fpg sensitive sites generates single-stranded breaks, which can readily be detected in the gene specific repair assay as described previously (Anson and Bohr, 1999; Bohr et al., 1985). Fpg treated and untreated DNA samples (10 g) were electrophoresed under alkaline conditions on a 0.5% agarose gel. DNA was transferred to a Hybond N+membrane (Amersham Pharmacia Biotech) by Posiblot (Stratagene) using standard protocols. The membrane was treated in prehybridization buffer (0.342 M Na₂HPO₄+0.088 M NaH₂PO₄, pH 7.2, 7% SDS, and 2 mM EDTA) for a minimum of 2 h. Double stranded DNA probes were prepared with a random-primed labeling kit (Amersham Pharmacia Biotech) using a 3.4 kb dihydrofolate reductase (DHFR) gene template. Membranes were hybridized with the probe at 68°C overnight and non-hybridizing probe was removed in stringent washes. The blots were visualized by Personal Imager PX and quantitated using the QuantityOne software (Bio-Rad, Hercules, CA, USA).

RO clonogenic survival assays

UV61 transfectant cell lines were trypsinized and 300 cells were seeded per 10-cm dish and allowed to grow for 16 h. For RO treatment, the cells were either non-treated or treated with the indicated doses of RO plus 5 min of visible light as described in the previous section. The cells were grown for 7 days, washed with PBS, fixed with methanol, and stained with methylene blue. The stained cells were washed and blue colonies were counted to determine the clonogenic survival of cells.

Acknowledgements

We appreciate the comments by T Thorslund and the technical assistance by UB Henriksen. RO19-8022 was a gift from Hoffman la Roche AG (Basel, Switzerland) and S Boiteux kindly provided the Fpg enzyme. TS was supported by The Danish Research Council and MS was supported by EU grant (QLK6-CT-1999-02002). The Novo Nordic Foundation, Eva and Henry Frænkels Foundation, and the foundation of 17.12.1981 supported the project.

Anson RM, Bohr VA. (1999). Methods Mol. Biol., 113: 257-279. MEDLINE

Anson RM, Croteau DL, Stierum RH, Filburn C, Parsell R, Bohr VA. (1998). Nucleic Acids Res., 26: 662-668. MEDLINE

Balajee AS, May A, Dianov GL, Friedberg EC, Bohr VA. (1997). Proc. Natl. Acad. Sci. USA, 94: 4306-4311. MEDLINE

Bessho T. (1999). Nucleic Acids Res., 27: 979-983. MEDLINE

Bessho T, Roy R, Yamamoto K, Kasai K, Nishimura S, Tano K, Mitra S. (1993). Proc. Natl. Acad. Sci. USA, 90: 8901-8904. MEDLINE

Bjoras M, Luna L, Johnsen B, Hoff E, Haug T, Rognes T, Seeberg E. (1997). EMBO J., 16: 6314-6322. Article MEDLINE

Bohr VA, Dianov GL. (1999). Biochimie, 81: 155-160. Article MEDLINE

Bohr VA, Okumoto DS. (1988). Friedberg EC and Hanawalt PC. (eds) New York: Marcel Dekker, pp 347-366.

Bohr VA, Smith CA, Okumoto DS, Hanawalt PC. (1985). Cell, 40: 359-369. MEDLINE

Boiteux S, Gajewski E, Laval J, Dizdaroglu M. (1992). Biochemistry, 31: 106-110. MEDLINE

Boiteux S, Radicella JP. (1999). Biochimie, 81: 59-67. Article MEDLINE

Brosh JrRM, Balajee AS, Selzer RR, Sunesen M, De Santis LP, Bohr VA. (1999). Mol. Biol. Cell, 10: 3583-3594. MEDLINE

Cheng KC, Cahill DS, Kasai H, Nishimura S, Loeb LA. (1992). J. Biol. Chem., 267: 166-172. MEDLINE

Citterio E, Rademakers S, van der Horst GT, van Gool AJ, Hoeijmakers JH, Vermeulen W. (1998). J. Biol. Chem., 273: 11844-11851. MEDLINE

Cooper PK, Nouspikel T, Clarkson SG, Leadon SA. (1997). Science, 275: 990-993. Article MEDLINE

Dianov G, Bischoff C, Piotrowski J, Bohr VA. (1998). J. Biol. Chem., 273: 33811-33816. MEDLINE

Dianov G, Bischoff C, Sunesen M, Bohr VA. (1999). Nucleic Acids Res., 27: 1365-1368. MEDLINE

Dianov GL, Houle JF, Iyer N, Bohr VA, Friedberg EC. (1997). Nucleic Acids Res., 25: 3636-3642. MEDLINE

Engelward BP, Weeda G, Wyatt MD, Broekhof JL, de Wit J, Donker I, Allan JM, Gold B, Hoeijmakers JH, Samson LD. (1997). Proc. Natl. Acad. Sci. USA, 94: 13087-13092. MEDLINE

Gorbalenya AE, Koonin EV, Donchenko AP, Blinov VM. (1989). Nucleic Acids Res., 17: 4713-4730. MEDLINE

Grishko VI, Driggers WJ, LeDoux SP, Wilson GL. (1997). Mutat. Res., 384: 73-80. MEDLINE

Hazra TK, Izumi T, Maidt L, Floyd RA, Mitra S. (1998). Nucleic Acids Res., 26: 5116-5122. MEDLINE

Henning KA, Li L, Iyer N, McDaniel LD, Reagan MS, Legerski R, Schultz RA, Stefanini M, Lehmann AR, Mayne LV, Friedberg EC. (1995). Cell, 82: 555-564. MEDLINE

Htun H, Johnston BH. (1992). Methods Enzymol., 212: 272-294. MEDLINE

Klungland A, Rosewell I, Hollenbach S, Larsen E, Daly G, Epe B, Seeberg E, Lindahl T, Barnes DE. (1999). Proc. Natl. Acad. Sci. USA, 96: 13300-13305. Article MEDLINE

Krokan HE, Nilsen H, Skorpen F, Otterlei M, Slupphaug G. (2000). FEBS Lett., 476: 73-77. Article MEDLINE

Krokan HE, Standal R, Slupphaug G. (1997). Biochem. J., 325: 1-16. MEDLINE

Kung HC, Bolton PH. (1997). J. Biol. Chem., 272: 9227-9236. MEDLINE

Le Page F, Gentil A, Sarasin A. (1999). Biochimie, 81: 147-153. MEDLINE

Le Page F, Klungland A, Barnes DE, Sarasin A, Boiteux S. (2000a). Proc. Natl. Acad. Sci. USA, 97: 8397-8402.

Le Page F, Kwoh EE, Avrutskaya A, Gentil A, Leadon SA, Sarasin A, Cooper PK. (2000b). Cell, 101: 159-171. MEDLINE

Lehmann AR. (1982). Mutat Res., 106: 347-356. MEDLINE

Lin JJ, Sancar A. (1989). Biochemistry, 28: 7979-7984. MEDLINE

Ljungman M, Zhang F. (1996). Oncogene, 13: 823-831. MEDLINE

Manley JL, Fire A, Samuels M, Sharp PA. (1983). Methods Enzymol., 101: 568-582. MEDLINE

Mayne LV, Lehmann AR. (1982). Cancer Res., 42: 1473-1478. MEDLINE

Memisoglu A, Samson L. (2000). Mutat Res., 451: 39-51. Article MEDLINE

Mu D, Hsu DS, Sancar A. (1996). J. Biol. Chem., 271: 8285-8294. MEDLINE

Nance MA, Berry SA. (1992). Am. J. Med. Genet., 42: 68-84. MEDLINE

Nouspikel T, Lalle P, Leadon SA, Cooper PK, Clarkson SG. (1997). Proc. Natl. Acad. Sci. USA, 94: 3116-3121. Article MEDLINE

Reardon JT, Bessho T, Kung HC, Bolton PH, Sancar A. (1997). Proc. Natl. Acad. Sci. USA, 94: 9463-9468. MEDLINE

Roldan-Arjona T, Wei YF, Carter KC, Klungland A, Anselmino C, Wang RP, Augustus M, Lindahl T. (1997). Proc. Natl. Acad. Sci. USA, 94: 8016-8020. Article MEDLINE

Rolig RL, McKinnon PJ. (2000). Trends Neurosci., 23: 417-424. MEDLINE

Selby CP, Sancar A. (1997a). Proc. Natl. Acad. Sci. USA, 94: 11205-11209. Article MEDLINE

Selby CP, Sancar A. (1997b). J. Biol. Chem., 272: 1885-1890. MEDLINE

Shivji MK, Eker AP, Wood RD. (1994). J. Biol. Chem., 269: 22749-22757. MEDLINE

Sunesen M, Selzer RR, Brosh JrRM, Balajee AS, Stevnsner T, Bohr VA. (2000). Nucleic Acids Res., 28: 3151-3159. MEDLINE

Taffe BG, Larminat F, Laval J, Croteau DL, Anson RM, Bohr VA. (1996). Mutat Res., 364: 183-192. MEDLINE

Tantin D. (1998). J. Biol. Chem., 273: 27794-27799. MEDLINE

Tantin D, Kansal A, Carey M. (1997). Mol. Cell Biol., 17: 6803-6814. MEDLINE

Thorslund T, Sunesen M, Bohr VA, Stevnsner T. (2002). DNA Repair In Press.

Troelstra C, van Gool A, de Wit J, Vermeulen W, Bootsma D, Hoeijmakers JH. (1992). Cell, 71: 939-953. MEDLINE

van Gool AJ, Citterio E, Rademakers S, van Os R, Vermeulen W, Constantinou A, Egly JM, Bootsma D, Hoeijmakers JH. (1997). EMBO J., 16: 5955-5965. Article MEDLINE

van Hoffen A, Natarajan AT, Mayne LV, van Zeeland AA, Mullenders LH, Venema J. (1993). Nucleic Acids Res., 21: 5890-5895. MEDLINE

Venema J, Mullenders LH, Natarajan AT, van Zeeland AA, Mayne LV. (1990). Proc. Natl. Acad. Sci. USA, 87: 4707-4711. MEDLINE

Viswanathan A, Doetsch PW. (1998). J. Biol. Chem., 273: 21276-21281. MEDLINE

Will O, Gocke E, Eckert I, Schulz I, Pflaum M, Mahler HC, Epe B. (1999). Mutat Res., 435: 89-101. MEDLINE

Woods CG. (1998). Arch. Dis. Child, 78: 178-184. MEDLINE

Figure 1 A schematic presentation of the CSB protein containing the conserved helicase/ATPase motifs and the N-terminal acidic region. Using site-directed mutagenesis a conserved glutamate in the ATPase motif II is substituted with a glutamine (CSBE646Q). The acidic region of the CSB protein is removed by a deletion of 39 consecutive amino acids of which approximately 60% are acidic residues (CSBAc). UV61/pc3.1-CSBwt, UV61/pc3.1-CSBE646Q, and UV61/pc3.1-CSBAc denote isogenic UV61 transfectant cells harboring the wild type, ATPase mutant, and acidic region mutant, respectively. Black boxes indicate putative nuclear localization sequences

Figure 2 Glycosylase activities in extracts from UV61 transfectant cell lines. (a) Schematic presentation of the in vitro incision assay on oligonucleotide duplexes containing a single well-defined uracil, thymine glycol or 8-oxoG DNA lesion. (b) Fifty g of WCE prepared from the indicated cell lines were incubated 3 h with the single lesion oligonucleotide duplex substrate containing either no lesion or 8-oxoG or for 30 min with the uracil-containing oligonucleotide duplexes. (c) Levels of 8-oxoG incision were quantitated and the mean±s.d. from three independent experiments is expressed relative to the incision level of UV61/pc3.1-CSBwt

Figure 3 Formation of Fpg-sensitive sites in the DHFR gene. Cells were either treated (+) or non-treated (-) with RO plus cells were either exposed (+) or non-exposed (-) to visible light and genomic DNA was isolated as described in Materials and methods. DNA was subsequently digested with KpnI and either treated (+) or non-treated (-) with Fpg. DNA was analysed by Southern blot analysis using double-stranded probes for the DHFR gene. Hybridization signals were detected and analysed by Personal Imager PX and QuantityOne analysis

Figure 4 Repair of Fpg-sensitive sites in the actively transcribed DHFR gene. Genomic DNA (10 g) was isolated from cells at 0, 4, 6 or 8 h after RO treatment as described in Materials and methods. DNA was subsequently digested with KpnI and either treated (+) or non-treated (-) with Fpg. DNA was analysed by Southern blot analysis using double-stranded probes for the DHFR gene. Blots were visualized using a Molecular Imager FX reader and quantitated using the QuantityOne software. Repair data is depicted as the mean±s.d. from three independent experiments

Figure 5 Clonogenic survival following RO plus light exposure. Isogenic clonal populations of UV61/pc3.1-CSBE646Q and UV61/pc3.1-CSBwt were exposed to the indicated concentration of RO plus exposure to visible light for 5 min as described in Materials and methods. Survival data are expressed as the fraction of RO plus light treated cells able to form colonies compared to non-treated cells and represents the mean±s.d. from three independent experiments