Groupe 'Réparation de l'ADN', UMR 8532 CNRS, LBPA-ENS Cachan, Institut Gustave Roussy, 39, rue Camille Desmoulins, 94805 Villejuif Cedex, France

Correspondence to: J Laval, E-mail: jlaval@igr.fr

A number of intrinsic and extrinsic mutagens induce structural damage in cellular DNA. These DNA damages are cytotoxic, miscoding or both and are believed to be at the origin of cell lethality, tissue degeneration, ageing and cancer. In order to counteract immediately the deleterious effects of such lesions, leading to genomic instability, cells have evolved a number of DNA repair mechanisms including the direct reversal of the lesion, sanitation of the dNTPs pools, mismatch repair and several DNA excision pathways including the base excision repair (BER) nucleotide excision repair (NER) and the nucleotide incision repair (NIR). These repair pathways are universally present in living cells and extremely well conserved. This review is focused on the repair of lesions induced by free radicals and ionising radiation. The BER pathway removes most of these DNA lesions, although recently it was shown that other pathways would also be efficient in the removal of oxidised bases. In the BER pathway the process is initiated by a DNA glycosylase excising the modified and mismatched base by hydrolysis of the glycosidic bond between the base and the deoxyribose of the DNA, generating a free base and an abasic site (AP-site) which in turn is repaired since it is cytotoxic and mutagenic.

Oncogene (2002) 21, 8905-8925. doi:10.1038/sj.onc.1206005

oxygen free radicals; oxidative DNA damage; DNA glycosylase; base excision; 8-oxoguanine; AP endonuclease

Introduction

Two main pathways appeared during evolution: the release of oxygen by photosynthesis and the aerobic respiration. These highly efficient metabolic systems produce useful, although dangerous molecules for the cells: free radicals (NO· ¼) and reactive oxygen species (ROS) such as O₂·^-, H₂O₂ and OH·. These molecules are constantly formed in cells by the cellular metabolism and by spontaneous chemical degradation of some biomolecules. Free radicals are implicated in a wide range of cellular processes (Bauerle et al., 1996) but an excess of such molecules could have deleterious effects on living cells leading to cell injury or cell death. Oxidative damage of main cellular components such as DNA, lipids, carbohydrates and proteins has been implemented in the ageing phenomena, ischemia (Mccord, 1987), cancer (Trus and Kensler, 1991), autoimmune diseases (Halliwell, 1982) and neural cell death (Reiter, 1995). Activation of oncogenes like c-myc (Vafa et al., 2002) and ras (Denko et al., 1994) can also induce ROS by alteration of specific metabolic pathways. Organisms are equally exposed to exogenous factors such as ionising radiation and chemical cancerogens, which can also generate ROS. Evidence has accumulated that lack of protection against free radicals and lack of repair of oxidative damage in biological macromolecules have a significant role in mutagenesis and consequently on carcinogenesis (Hoeijmakers, 2001). Thus maintenance of the cellular equilibrium between prooxidant (ROS) and antioxidant species should be tightly regulated in cells. Specialised systems have evolved, involving cellular antioxidants and DNA repair, to protect cells against ROS-induced injury. Indeed, the defence system, such as DNA repair, is highly conserved from bacteria to human.

It is generally assumed that most oxidative DNA damages - base damage, sugar damage and abasic sites - are dealt with by BER. A number of recent reviews report the main features of the BER pathway (Seeberg et al., 1995; Lindahl, 2001, Krokan et al., 1997; David and Williams, 1998; Scharer and Jiricny, 2001; Ishikawa et al., 2001; Hoeijmakers, 2001). The goal of DNA glycosylases is to locate fast and efficiently the aberrant base amongst a huge excess of normal ones. Very little is known about how these proteins achieve this goal. The comparison of the crystal structures of a number of DNA glycosylases revealed structural homologies leading to the concept of a superfamily of BER DNA glycosylases, the helix-hairpin-helix (HhH) superfamily, having similar HhH fold and a Gly/Pro-rich stretch with nearby Asp (GPD) motifs, although very little sequence similarity. This HhH motif plays an important role in the flipping out of the modified base (Thayer et al., 1995; Doherty et al., 1996). The rate of repair measured for the excision of modified bases is not always optimal and should be improved by the identification and the use of accessory proteins. The recent identification of new DNA polymerases able to replicate efficiently and accurately miscoding and modified bases have to be taken into account in the understanding of BER.

Free radical species and oxidative damage of DNA

DNA has a limited chemical stability (intrinsic or induced by exogenous agents) and is one of the most biologically important targets of ROS (Lindahl, 1993; Imlay and Linn, 1988). Maintenance of its integrity is a major goal for cells. About 100 different kinds of base and sugar damage have been identified. Free-radicals can damage nucleobases and sugar units in DNA either directly, or indirectly. Hydroxyl radicals, which are the most active species, predominantly, react with the C₈ of purines forming 7,8-dihydro-8-oxo-2'-deoxyguanosine (8-oxoG) (Devasagayam et al., 1991) and imidazol ring-opened products such as 2,6-diamino-4-hydroxy-5-formamidopyrimidine (Fapy); with C₅-C₆ double bond of pyrimidines forming glycol (Schuchmann et al., 1984; von Sonntag and Schuchmann, 1994; Cadet et al., 1999) and pyrimidines hydrates and with C8-C5' of purines forming 8,5'-cyclopurine deoxynucleosides (Dirksen et al., 1988). The abstraction of a hydrogen atom from deoxyribose at C1' and C4' generates DNA-strand breaks with 3'-phosphoglycolate ester and 3'-phosphate (Dizdaroglu et al., 1977; Giloni et al., 1981; Henner et al., 1983). Indirectly, ROS can generate reactive aldehydes, as a product of membrane lipids peroxidation, which react with DNA bases forming the exocyclic adducts 1,N⁶-ethenoadenine, 1,N²-ethenoguanine, N²,3-ethenoguanine, and 3,N⁴-ethenocytosine (Marnett, 1994; El-Ghissassi et al., 1995) and pyrimidopurinone such as M₁G (reviewed in Marnett, 1999).

Damaged bases such as 8-oxoguanine; 5-hydroxy-2'-deoxycitidine, hypoxanthine, ethenoadducts and pyrimidopurinone have miscoding properties and, if not repaired, lead to mutation upon replication (Fink et al., 1997; reviewed in Marnett, 2001). Others such as oxidised deoxyribose, formamidopyrimidine, fragmented thymine and thymine glycol cause replication block and therefore are believed to have a strong cytotoxic effect (reviewed in Laval et al., 1998).

Ionising radiation induces mutation and chromosomal aberrations in cells, which are believed to lead to cancer and loss of neural function in humans. Humans are daily exposed to low doses of radiation during air travel or from radon in homes and areas of low-level contamination. Energy from ionising radiation, such as X-rays and gamma rays, is transmitted in the water surrounding the DNA molecule in such a way that between 2 to 5 radical pairs are generated within a radius of 1 to 4 nm (Goodhead, 1994; Blaisdell et al., 2001). Clustered multiple damaged sites induced by ionising radiation, have been observed, most of them being modified bases rather than DNA strand breaks (reviewed in Sutherland et al., 2001a). The complexity of radiation-induced clustered DNA-damage depends on the ionising density of LET (Linear Energy Transfer) radiation (Cunniffe and O'Neill, 1999).

Ionising radiation was hypothesised to produce clustered damage, and clusters spanning several kilobase pairs to a few base pairs were modeled (Holley and Chatterjee, 1996; Ward, 1981; Goodhead, 1994). Irradiation of DNA in non-radioquenching solution with gamma rays was shown to induce 1 double strand break to ~0.5 Nth protein-recognised oxidised pyrimidine cluster: 1.5 Nfo protein-recognised abasic cluster: 2 Fpg protein-recognised oxidised purine clusters (Sutherland et al., 2000b). Use of the polyamines putrescine to measure abasic clusters refractory to Nfo protein cleavage reveals twice as many abasic clusters in -irradiated DNA as does Nfo, suggesting that even more clusters are produced than these values indicate (Georgakilas et al., 2002). The dose-response relations in a log-log plot are straight lines with slopes near 1, showing that clusters are formed by a single radiation hit (Sutherland et al., 2000b). All DNA lesions are not equally frequent components of clusters, with about 15% of oxidised pyrimidines, oxidised purines and abasic sites in clusters, but only about 8% of the strand breaks are in DSBs (Sutherland et al., 2000a).

Since double strand break yields are ~100 times higher in the absence of radical scavengers (Milligan et al., 2000) fewer clusters are expected in DNA irradiated in the presence of a radical scavenger such as Tris. Indeed the absolute levels of bistranded cluster levels are significantly reduced in radioscavenging solution, and, surprisingly, the ratios of specific cluster types also changed strikingly (Sutherland et al., 2001b). Cluster levels in human cells exposed to 50 kVp X-rays shows that X-rays induce 1 DSB : 0.75 Nfo-abasic cluster : 1 Fpg-oxidized purine : 0.9 Nth-oxidized pyrimidine cluster. In cells, non-DSB clusters are at least ~70% of the complex damages.

Clusters are postulated to be critical because they may be more difficult to repair than random lesions (Ward, 1995; Olive, 1998). In contrast to randomly located oxidised bases, clustered modified bases are within half a turn of the double helix, i.e. 5 nucleotides and some of them on the two strands, therefore the excision repair pathways would have difficulty in removing clustered lesions. It is anticipated that the excision/incision of oxidatively damaged bases, including AP sites, on both strands will, if not tightly regulated, either inhibit certain steps of repair or produce double strand breaks and thus be lethal for the cells (Chaudhry and Weinfeld, 1997; Mckenzie and Strauss, 2001). Studies of purified enzymes acting on oligonucleotides with defined lesions at specific relative spacing on opposing strands indicate that clusters may comprise non-repairable, highly repair-resistant and pre-mutagenic damage, and that lesion spacing and polarity are important (Chaudhry and Weinfeld, 1995, 1997; Harrison et al., 1998; David-Cordonnier et al., 2000, 2001a,b, 2002; Rasouli-Nia et al., 2001; Weinfeld et al., 2001; Budworth et al., 2002). However, the precise repair mechanisms for the clustered lesions are so far very poorly understood.

E. coli generate high levels of DSBs during repair after irradiation (Bonura et al., 1975; Blaisdell and Wallace, 2001), and the repair-generated DSBs likely result from abortive cluster repair (Blaisdell and Wallace, 2001). Although rodent cells exposed to high doses (500 Gy) also increase DSB levels (Dugle et al., 1976; Ahnstrom and Bryant, 1982), their origin and whether they are generated at low doses was not clear.

Base excision repair

Oxygen radicals generate mostly non-bulky DNA lesions, most of them are substrates for BER enzymes. Only few oxidative DNA lesions such as cyclopurine (Kuraoka et al., 2000; Brooks et al., 2000) and pyrimidopurinone (Marnett, 1999) are substrates for nucleotide excision repair (NER) enzymes. Key enzymes of the BER pathway are DNA-glycosylases (reviewed in Krokan et al., 1997). They remove damaged and mispaired bases from DNA by cleavage of the N-glycosylic bond between the abnormal base and deoxyribose, leaving an abasic site in DNA (Lindahl, 1976). Most DNA glycosylases are highly specific and can excise various types of modified bases.

There are two types of DNA glycosylases: mono- and bifunctional. The mono-functional ones cleave the N-glycosylic bond releasing the modified base and generating an AP site as a final product which in turn is recognised by an AP endonuclease (Laval, 1977). The bifunctional ones cleave the N-glycosylic bond, liberate the modified base and in a concerted manner cleave the phosphodiester bond 3' to the resulting AP site by a or - elimination mechanism (-lyase activity) generating a single-strand break with 3'-phosphate and 3'-phosphoglycolate (PGA) extremities respectively. Then the DNA backbone next to the abasic site or 3'-phosphate/PGA terminus is cleaved by an AP-endonuclease allowing a DNA polymerase to fill the gap before DNA ligase reseals the DNA (Figure 1). Experiments with E. coli mutants deficient in BER pathway demonstrated the crucial role of DNA glycosylases and AP-endonucleases in protecting cells from mutagenic and cytotoxic effects of free radicals (Michaels et al., 1992; Duwat et al., 1995; Demple et al., 1983; Cunningham et al., 1986).

In human cells BER proceeds via two alternative pathways, either 'short-patch' DNA polymerase -dependent pathway (Kubota et al., 1996) (Figure 1) which involves the replacement of a single nucleotide or 'long patch' PCNA-dependent pathway (Matsumoto et al., 1994; Frosina et al., 1996; Klungland and Lindahl, 1997) which involves the replacement of up to six nucleotides (reviewed in Wilson and Thompson, 1997). The former pathway was reconstituted in vitro with the purified human proteins uracil-DNA glycosylase (UDG), AP endonuclease 1 (APE1), DNA polymerase (pol ), the scaffold protein XRCC1 and DNA ligase (I or III) (Kubota et al., 1996). Long-patch pathway can be achieved with AP endonuclease, Pol and PCNA (Matsumoto et al., 1994). The antibodies directed against PCNA totally suppress repair patches longer than one nucleotide (Frosina et al., 1996). Reconstitution in vitro of long-patch repair showed that DNase IV/FEN 1 is required in addition to proteins implicated in short-patch pathway and that PCNA greatly stimulates the reaction (Klungland and Lindahl, 1997). Either DNA polymerase / (Pascucci et al., 1999; Matsumoto et al., 1999) or DNA polymerase (Dianov et al., 1999; Prasad et al., 2000) have been shown to perform the synthesis step in this sub-pathway. Recently, it has been demonstrated that removal of 8-oxoG could be achieved when using only hOGG1, Ape1, pol and DNA ligase I (Pascucci et al., 2002).

Base excision repair in prokaryotes (E. coli)

Bifunctional DNA glycosylases: Fpg, Nei and Nth proteins

Mechanistic studies of both classes of DNA glycosylases led to formulation of a model of unified catalytic mechanism (Dodson et al., 1994). Monofunctional DNA glycosylases cleave the glycosidic bond by a hydrolytic mechanism, activating a water molecule to attack the C1' of the damaged base, resulting in AP site as a final product. In contrast, the DNA glycosylases/AP lyases employ a nucleophilic group in the enzyme to perform attack at C1'. In all DNA glycosylases/AP lyases so far characterised, an amino group has been implicated as a nucleophile (Weiss and Grossman, 1987; Scharer et al., 1997). As a result of nucleophilic attack by amino group, a covalent imino intermediate (Schiff base) is formed between the C1' of the lesion and the enzyme. By abstracting the C2'-pro-S proton (Mazumder et al., 1991), the enzyme initiates electron rearrangement that leads to the release of the 3'-phosphate. The Schiff base is then hydrolysed, releasing the enzyme and leaving a nick with an ,-unsaturated aldehyde at the 3'-end and a phosphate at the 5'-end of the DNA. Upon completion of -elimination, some enzymes such as the Fpg protein abstract proton at the C4' and promote a -elimination (O'Connor and Laval, 1989; Bhagwat and Gerlt, 1996). The resulting structure is a single-base gap with phosphates at both 3'- and 5'-ends.

Sodium borohydride and cyanoborohydride have been used to show that the reaction involves a Schiff base intermediate (Sun et al., 1995). When the reduction reaction occurs in the glycosylase-DNA complex, the enzyme becomes irreversibly crosslinked to DNA.

The Fpg protein (formamidopyrimidine-DNA glycosylase): Among the DNA glycosylases, the E. coli Fpg protein (formamidopyrimidine-DNA glycosylase/MutM protein) has been one of the most extensively studied. The gene coding for the Fpg protein was cloned (Boiteux et al., 1987) and the physical and enzymatic properties of the protein established (Boiteux et al., 1990). It is a globular monomer of 30.2 kDa of 269 amino acids (Boiteux et al., 1990). In vitro, the Fpg protein excises a broad spectrum of modified purines (Boiteux et al., 1992), particularly 2,6-diamino - 4 - hydroxy - 5N - methylformamidopyrimidine (Fapy) and 7,8-dihydro-8-oxoguanine (8-oxoG) residues (Tchou et al., 1991). Moreover, the Fpg protein is also able to excise various pyrimidine oxidation products such as 5-hydroxycytosine and 5-hydroxyuracil (Hatahet et al., 1994), the ring fragmentation product of thymine (RT) (Jurado et al., 1998), thymine glycol and 5,6-dihydrothymine (D'Ham et al., 1999) (Table 1).

The Fpg protein is a bifunctional DNA glycosylase, endowed of an AP lyase activity that incises DNA at abasic sites by a --elimination mechanism (O'Connor and Laval, 1989; Bailly et al., 1989) and an activity excising 5'-terminal deoxyribose phosphate (dRPase) (Graves et al., 1992). In vivo, the Fpg protein has an antimutator effect preventing G/CT/A spontaneous transversion. It acts in concert with the MutY protein (Tajiri et al., 1995). The fpg mutY double mutant CC104 has an extreme mutator phenotype that can be reversed by plasmids carrying the fpg gene (Duwat et al., 1995). Interestingly, expression of the bacterial fpg gene in mammalian cells reduces the mutagenicity of -rays but has no effect on survival (Laval, 1994).

The Fpg protein in its COOH-terminus contains the zinc-finger motif (Cys-X₂-Cys-X₁₆-Cys-X₂-Cys-X₂-COOH) that is mandatory for Fpg binding to DNA and to its enzymatic activities (O'Connor et al., 1993). The active site of Fpg protein is located within the first 73 amino acid residues of the amino terminus (Tchou and Grollman, 1995). The targeted mutagenesis of conserved amino acid residues was used to elucidate the mechanism of enzymatic catalysis. It was found that conserved residues lysine 57 (K57G) (Sidorkina and Laval, 1998), lysine 155 (Rabow and Kow, 1997), glutamates 2, 5, 131 and 173 (Lavrukhin and Lloyd, 2000) and proline 2 (P2G) (Sidorkina and Laval, 2000) dramatically reduce the cleavage of oxidised bases such as 8-oxoG.

Recently, major progress has been achieved in solving the three-dimensional structure of the Fpg protein from various origins (Sugahara et al., 2000; Serre et al., 2002; Gilboa et al., 2002). These structures will be discussed below.

Two Fpg proteins have been isolated from the highly radioresistant bacteria Deinococcus radiodurans. Both excise Fapy and 8-oxoG residues and present a -lyase activity. In addition one excises thymine glycols (Bauche and Laval, 1999). The genes have been cloned and the detailed specificities established using pure proteins (Sentürker et al., 1999).

The Nth protein (Endonuclease III): The Nth protein was initially identified as an endonuclease that specifically cleaves DNA damaged by X-rays, UV light and free radicals (Radman, 1976; Demple and Linn, 1980). It is a DNA glycosylase with a broad substrate specificity, excising ring-saturated, ring-opened, and ring-fragmented pyrimidines, such as thymine glycol, 5,6-dihydrothymine, 5-hydroxy-6-hydrothymine,5,6-dihydrouracil, alloxan, 5-hydroxy-6-hydrouracil, uracil glycol (Dizdaroglu et al., 1993), 5-hydroxy-2'-deoxycytidine, 5-hydroxy-2'deoxyridine (Hatahet et al., 1994), -ureidoisobutiric acid (Mazumder et al., 1991) and also -R-hydroxy--ureidoisobutiric acid, a fragmentation product of the 5R-thymidine C5-hydrate (Jurado et al., 1998). Unexpectedly, it has been demonstrated that Nth is implicated in the repair of 8-oxoG, since it removes 8-oxoG from 8-oxoG/G mispairs and the triple mutant fpg nth nei displays increased spontaneous G/CC/G transversions as compared to single and double mutants in these three genes (Matsumoto et al., 2001). This leads to the proposal of a new model in which Nth and Nei repairs 8-oxoG in nascent and transcriptionally active DNA (Hazra et al., 2001).

The nicking activity at abasic sites of Nth is due to its AP lyase function. The phosphodiester bond cleavage occurs via -elimination (the protein does not cleave at reduced AP sites) generating 5' ends bearing 5'-phosphate and 3' ends with 2,3-unsaturated abasic residue 4-hydroxy-2-pentanal. This terminus requires further processing to be used for DNA repair synthesis (Bailly and Verly, 1987; Kim and Linn, 1988).

The gene encoding for the Nth protein, the nth gene, was cloned (Cunningham and Weiss, 1985) and protein purified to homogeneity (Asahara et al., 1989). The three dimensional structure of Nth was also established (Kuo et al., 1992). The Nth protein is a monomeric protein of 23.4 kDa (211 aa). It is an elongated protein with a cleft separating two similar size domains: a continuous domain formed by six--helices and a second domain formed by three C-terminal -helices and the N-terminal helix. The C terminal loop contains an iron-sulphur centre (4Fe-4S) that is anchored to the protein by a Cys-X6-Cys-X2-Cys-X5-Cys sequence. This cluster has a structural function in positioning basic residues for DNA-binding (Thayer et al., 1995) and it is also present in the MutY protein (Michaels et al., 1990).

E. coli nth mutants deficient in the Nth protein does not show apparent phenotype (Cunningham and Weiss, 1985). They are not sensitive to X-rays, H₂O₂ or other agents that produce ring saturation and fragmentation products of pyrimidines in DNA. However the double mutant nth nei exhibits a strong spontaneous mutator phenotype and it is hypersensitive to ionising radiation and H₂O₂ (Jiang et al., 1997).

The Nei protein (endonuclease VIII): The capacity of E. coli nth mutants defective in most of the activity to excise thymine glycol, allowed the identification of a new activity excising this oxidised base, that was named endonuclease VIII or Nei protein (Melamede et al., 1994). It has been purified from E. coli extract lacking the Nth protein. Similarly to the Nth protein, Nei removes the oxidised forms of thymine such as thymine glycol, dihydrothymine, -ureidoisobutyric acid, and urea residues when present in DNA and cleaves phosphodiester bond at AP sites (Melamede et al., 1994). The cloned nei gene encodes a 263 amino acid polypeptide with a striking similarity to the Fpg protein (Jiang et al., 1997). The catalytic site's residues are highly conserved in these two proteins (Burgess et al., 2002), however, their substrate specificities are different. Characterisation of apparently homogenous Nei protein has shown that it can equally remove the oxidised forms of cytosine such as 5-hydroxycytosine and 5-hydroxyuracil (Jiang et al., 1997). More recently, genetic evidence for Nei implication in the prevention of spontaneous GT transversions (Blaisdell et al., 1999) was confirmed since Nei was found to cleave 8-oxoG efficiently when opposite to A and G (Hazra et al., 2000). Furthermore, Dizdaroglu et al. (2001) extended the substrate specificity of the Nei protein and have demonstrated that Nei and Nth significantly differ from each other in terms of kinetic although they share common substrate specificity. The crystal structure at 1,25 Å of the Nei covalent intermediate complex with DNA largely confirms that its structure is similar to that of Fpg (Zharkov et al., 2002).

Monofunctional DNA glycosylases

AlkA protein: The main substrates for the E. coli AlkA and Tag I proteins are alkylated bases. The Tag I protein is a major DNA glycosylase removing 3-methylpurines residues (Bjelland et al., 1993; reviewed in Krokan et al., 1997). However, the AlkA protein in contrast to Tag I is also involved in repair of oxidative DNA damage. Hypoxanthine (HX) is generated in DNA by spontaneous deamination of adenine and also by the free radical nitric oxide (Wink et al., 1994). HX residues in DNA are mutagenic since they can pair with cytosine, generating AT to GC transitions after DNA replication. Hypoxantine-DNA glycosylase releases HX from DNA containing dIMP (reviewed in Laval et al., 1998). Cloning the gene coding for the HX-DNA glycosylase in E. coli, it was shown that the 3-methyladenine-DNA glycosylase (Saparbaev and Laval, 1994) coded by the alkA gene carries the HX-DNA glycosylase activity. ANPG, APDG and MAG proteins, the human, rat, and yeast functional homologues respectively of AlkA protein excise HX residues when present in DNA, the mammalian enzymes being most efficient (Saparbaev and Laval, 1994) (Table 1). The AlkA protein also catalyses the excision of ethenobases N²,3-ethenoguanine (Habraken et al., 1991) and 1,N⁶-ethenoadenine (Saparbaev et al., 1995).

The 5-formyluracil (5-foU) residue, an oxidised thymine lesion is a major DNA damage induced by ionising radiation. It induces A/T to G/C transition (Kasai et al., 1990) and is repaired in E. coli by the AlkA protein (Bjelland et al., 1994). In addition, AlkA protein excise RT residues when present in DNA with an apparent K_m=170 nM (Privezentzev et al., 2000). Thus the AlkA protein has very broad substrate specificity. The three dimensional structure of AlkA reveals a compact globular protein with a prominent hydrophobic cleft on its surface and three equal-sized domains (Yamagata et al., 1996; Labahn et al., 1996).

Uracil DNA glycosylase (UDGs) super-family: Base excision repair was first established for uracil (Lindahl, 1974). Four sub-families of uracil-DNA glycosylase (UDG) have been identified (Aravind and Koonin, 2000). The best studied family of UDGs is E. coli Ung protein. Ung is specific for uracil and it is present in a wide range of living organisms from bacteria, lower to higher eukaryotes, DNA viruses and human (reviewed in Krokan et al., 1997). So far UDG has not been implicated in the repair of oxidative DNA damage. The second family corresponds to the mismatch-specific uracil-DNA glycosylase (MUG) identified in some eukaryotes and in several prokaryotes. MUG excises thymine from G/T mismatches in DNA and it is also active on mispaired uracil in G/U pair (Barrett et al., 1998; Gallinari and Jiricny, 1996; Saparpaev and Laval, 1998). The crystal structure of Ung and Mug proteins show that they are structurally similar, despite low sequence homology (Barrett et al., 1998). Another sub-family has been characterised from thermophilic archea (Sandigursky and Franklin, 1999, 2000) and several bacteria (Sung and Mosbaugh, 2000). These double strand UDGs (dsUDG/DUG) are able to remove uracil from double strand DNA either in U/A and U/G context. The fourth sub-family, initially identified in vertebrates (Haushalter et al., 1999) is represented by the human single-strand mismatch-specific uracil-DNA glycosylase (hSMUG1). However hSMUG1 is more active on double-stranded than on single-stranded DNA, it prefers uracil opposite A and G (Nielsen et al., 2001). The 5-hydroxymethyluracil residue is a product of oxidation and subsequent deamination of 5-methylcytosine (Boorstein et al., 1992). Teebor's group identified the mammalian 5-hydroxymethyluracil DNA N-glycosylase activity as SMUG1 (Boorstein et al., 2001). A fifth sub-family has been characterised recently and is represented by pa-UDGb isolated from Pyrobaculum aerophilum (Sartori et al., 2002). pa-UDGb has broad substrate specificity. It can remove uracil and also hypoxanthine when present in DNA.

Among the members of the UDG family several enzymes are implicated in the repair of exocyclic DNA adducts which are highly mutagenic oxidative DNA lesions. The E. coli MUG protein is able to remove 3,N⁴-ethenocytosine (Saparbaev and Laval, 1998), 8-(hydroxymethyl)-3,N⁴-ethenocytosine (Hang et al., 2002) and 1,N²-ethenoguanine (Saparbaev et al., 2002b) when present in DNA.

MutY: The E. coli mutY gene encodes a DNA glycosylase of 39 kDa, which excise A opposite G and C (Michaels et al., 1990; Tsai-Wu et al., 1991). MutY also removes A opposite to 8-oxoG and 7,8-dihydro-8-oxoadenine (8-oxoA). The protein shows significant sequence homology to Nth and is also an iron-sulphur protein (Tsai-Wu et al., 1992) which contains the conserved (4Fe-4S) cluster/HhH domain. The protein was extensively characterised using various approaches including site-directed mutagenesis (reviewed by David and Williams, 1998). However, there is still controversy regarding the AP lyase activity of the MutY protein (Tsai-Wu et al., 1992; Zharkov and Grollman, 1998). Proteolytic cleavage of MutY generates two fragments of 26 kDa and 13 kDa. The 26 kDa fragment corresponds to the catalytic core domain, it retains normal DNA binding and adenine-DNA glycosylase activity but lacks its activity against A/8-oxoG. The specificity for 8-oxoG residues is due to the C-terminal domain of MutY (Chmiel et al., 2001). It is thought that MutY plays an important role in the maintenance of genome integrity following oxidative DNA damage. The presence of AP endonucleases (Xth and Nfo) greatly enhances the excision rate of A when opposite to G by MutY (Pope et al., 2002). These data suggest that the MutY-DNA complex interacts with Nfo and Xth proteins in vivo. The crystal structure of MutY catalytic core domain (cMutY) has been solved (Guan et al., 1998). cMutY is an all- protein that retains the canonical bilobal architecture of Endonuclease III. The compression of intra-strand phosphate distance by HhH and pseudo HhH domains permit the DNA bending and the nucleotide flipping. The flipped nucleotide is recognised by specified residues of the active side pocket located between the two domains. Crystal structures and mutagenesis results define that MutY cleaves the N-glycosylic bond through a hydrolytic mechanism requiring Asp138, with uncoupled, inefficient AP lyase activity due to Lys142 (Guan et al., 1998).

AP endonucleases: Xth and Nfo

Abasic sites (AP sites) occur in DNA through spontaneous depurination/depyrimidination, by the action of ROS and as secondary lesions after excision of modified bases by DNA glycosylases. AP sites have miscoding properties since the replication machinery incorporates preferentially adenine opposite to AP site (Boiteux and Laval, 1982). This observation was later named 'A rule' (Strauss, 1991). The occurrence of an enzyme recognising AP sites in E. coli was first described in the seminal work of Verly (Verly and Paquette, 1972). Two types of enzymes recognise and excise DNA at AP sites: AP endonucleases such as Nfo and Xth proteins act on this lesion by a hydrolytic mechanism (reviewed in Barzilay and Hickson, 1995) whereas enzymes carrying a -lyase activity such as Fpg, Nei and Nth proteins incise the AP site via a - and elimination mechanism respectively.

The Nfo protein (Endonuclease IV): The Nfo protein is an EDTA-resistant AP endonuclease that represents only 5% of the total AP endonucleasse activity in E. coli wild type. It is inducible by oxidative stress and is under the control of the soxRS system (Chan and Weiss, 1987; Nunoshiba et al., 1993; Tsaneva and Weiss, 1990; Demple, 1991). The gene coding in E. coli for Endonuclease IV, nfo, has been cloned (Cunningham et al., 1986), and the protein characterised (Levin et al., 1988). It is a monomer of 30 kDa having several activities: AP endonuclease, 3'-phosphatase and 3-phosphoglycoaldehyde diesterase activities (Ljungquist, 1977). It has been suggested that these 3' repair activities clean the ends, a prerequisite for DNA synthesis. E. coli nfo mutants deficient in the product of the nfo gene are extremely sensitive to the lethal effects of oxidative agents such as bleomycin and t-butyl-hydroperoxide. In addition, this mutation enhances the sensitivity of xth mutants to H₂O₂, and alkylating agents. The double mutants xth nfo are also sensitive to -radiation (Cunningham et al., 1986). We will discuss below the role of the Nfo protein and its yeast homologue Apn1 in the nucleotide incision repair pathway (NIR) (Ischenko and Saparbaev, 2002). The high-resolution structure of Nfo protein has been solved (Hosfield et al., 1999). It shows that Nfo is an protein arranged as a ₈₈-barrel. This structure, well suited for the binding to large molecules such as DNA, contains three Zn²⁺ ions at the active site which are critical for the activity. The triple Zn centre is ligated to protein by conserved residues that cluster at the centre of the crescent-shaped deep pocket. The Nfo protein detects AP site by insertion of its side chains into the DNA minor groove, it flips the target AP site and the opposite nucleotide out of the DNA base stack to produce a 90° bend in the DNA.

The Xth protein (Exonuclease III): The Xth protein, coded for by the xth gene, originally identified as a 3'5' exonuclease active on double-stranded DNA and a 3'-phosphatase, is the major AP endonuclease of E. coli: over 80% of the total AP endonuclease activity in wild type (Rogers and Weiss, 1980). This protein is also active on single stranded DNA (Shida et al., 1996). Beside its exonuclease and 3'-phosphatase activities it has 3' repair diesterase and ribonuclease H activities. The 3' termini generated by Xth are normal nucleotides with 3' hydroxyl group that are effective primers for DNA polymerases (Rogers and Weiss, 1980; Demple et al., 1986).

The xth mutants are extremely sensitive to H₂O₂, and to oxidative DNA damage generated by near-UV light. The crystal structure of the Xth protein has been solved, and it has been assumed that the base opposite to the abasic site plays an important role for the recognition of the lesion (Mol et al., 1995).

Mechanism of action and three-dimensional structure of bifunctional DNA glycosylases: the Fpg case: It was hypothesised that DNA glycosylase/AP lyases use a mechanism involving the nucleophilic attack on the C1' prime of the modified deoxynucleoside targeted for excision, thus displacing the aberrant base and forming a transient intermediate (Schiff base) with the C1'-deoxyribose moiety (Weiss and Grossman, 1987; Sun et al., 1995). The amino acid sequence of the Fpg protein begins with a Met-1 which is processed, thus Pro-2 being the N-terminal (Sidorkina and Laval, 2000). The N-terminal proline (P2) residue of the Fpg protein, a highly conserved residue, was shown to be linked with DNA containing 8-oxoG residues, suggesting that proline is the nucleophile initiating the excision of 8-oxoG (Zharkov et al., 1997). Site-directed mutagenesis demonstrated the mandatory role of the N-terminal proline residue in the 8-oxoG-DNA glycosylase activity of the Fpg protein in vitro and in vivo, as well as in its AP lyase activity upon pre-formed AP sites but less of a role in Fapy-DNA glycosylase activity (Sidorkina and Laval, 2000).

The role of the conserved lysine 57 and 155 (K57 and K155) residues upon the various catalytic activities of the Fpg protein was examined by targeted mutagenesis (Sidorkina and Laval, 1998; Rabow and Kow, 1997). The lysine 57glycine (FpgK57G) mutant protein had a dramatically reduced (55-fold) DNA glycosylase activity for the excision of 8-oxoG residues. The FpgK57G protein was poorly effective in the formation of Schiff base complex with 8-oxoG/C DNA. However, the mutant could partially restore the ability to prevent spontaneously induced G/CT/A transversions in E. coli BH990 (fpg, mutY) cells. The DNA glycosylase activity of FpgK57G using FapyGua residues as the substrate was comparable to that of the wild type enzyme. These results suggested that K57 could participate either in a direct interaction with the C₈ oxygen of the 8-oxo purines (Nash et al., 1997; Sidorkina and Laval, 1998) or in the opening of the furanose ring, in the first step of the model proposed for the mechanism of action of the Nth protein (Kow and Wallace, 1987; Weiss and Grossman, 1987). Effect of mutations of conserved glutamic and aspartic acid residues to glutamines and asparagines, have been studied (Lavrukhin and Lloyd, 2000). While the Asp to Asn mutants had no effect on the incision activity on 8-oxoG DNA, several of the substitutions at glutamates reduced Fpg activity on the 8-oxoguanosine DNA, with the E3Q and E174Q mutants being essentially devoid of activity. Unexpectedly, the AP lyase activity of all of the glutamic acid mutants was slightly reduced as compared to the wild-type enzyme. Furthermore, it has been shown that lysine 57 but not proline 2 is crucial for catalysis of oxidatively damaged pyrimidines (Saparbaev et al., 2002a).

Sugahara and co-workers determined the crystal structure of Fpg homologous protein from an extreme thermophile, Thermus thermophilus HB8 (HB8-Fpg) at 1.9 Å resolution. It reveals that the Fpg protein molecule is composed of two distinct domains connected by a flexible hinge (Sugahara et al., 2000). More recently, Castaing and colleagues resolved the structure of a non-covalent complex between the Lactococcus lactis Fpg and a 1,3-propanediol (Pr) abasic site analogue-containing DNA (Serre et al., 2002). They have shown that Fpg pushes out the Pr site from the DNA double helix, recognising the cytosine opposite the lesion and inducing a 60 degree bend of the DNA. Finally, the structure of a trapped catalytic intermediate of E. coli Fpg protein has been determined at 2.1 Å resolution (Gilboa et al., 2002). In agreement with the structure of HB8-Fpg the E. coli Fpg is a bilobal protein with a wide negatively charged DNA-binding groove. Highly conserved residues Lys-57, His-71, Asn-169 and Arg-259 are involved in binding the phosphodiester backbone of DNA, which is sharply kinked at the lesion site. Conserved residues Met-74, Arg-110 and Phe-111 are inserted into DNA helix to fill the void in DNA after nucleotide eversion. A deep hydrophobic pocket in the active site is positioned to accommodate the damaged base (Fromme and Verdine, 2002).

Base excision repair in eukaryotes

Yeast (S. cerevisiae and S. pombe)

Monofunctional DNA glycosylase: spMYH: The E. coli MutY homolog was identified in Schizosaccharomyces pombe by a homology search in genome databases. The spMYH gene encodes for a 461 amino acids protein which is highly homologous to MutY and human MYH (Lu and Fawcett, 1998). spMYH possesses the HhH motif and the [4Fe-4S] DNA binding cluster domain found in Nth/MutY. It has a DNA glycosylase and probably an AP lyase activity. SpMYH incises A opposite G (A/G), A/8-oxoG, 2-aminopurine/G and A/2-aminopurine containing DNA (Lu and Fawcett, 1998). Cells deleted for the spMYH gene (spMyH) have a mutator phenotype and are more sensitive to oxidative agents, such as hydrogen peroxide, than wild type cells (Chang and Lu, 2002a). These data show that spMYH is implicated in the avoidance of 8-oxoG induced mutations and it has an important role in the protection against oxidative stress. Interestingly, similar to human MYH (hMYH) (Parker et al., 2001), spMYH also interacts directly with PCNA (Chang and Lu, 2002b). This interaction is fundamental for the biological function of spMYH in the control of mutation avoidance since the mutation rate of spMyH expressing hMYH is reduced but it stays unchanged when spMyH cells express a mutant hMYH protein unable to interact with PCNA (Chang and Lu, 2002b).

Bifunctional DNA glycosylases S. cerevisiae yOGG1, Ntg1 and Ntg2: Saccharomyces cerevisiae 8-oxoguanine-DNA glycosylase (yOGG1) is the yeast counterpart of the E. coli Fpg protein. It was identified by functional suppression of the E. coli fpg mutY mutator phenotype (van der Kemp et al., 1996), leading to the concept that eukaryotes use the DNA glycosylase pathway to repair oxidised purines in DNA. Comparison of amino acid sequences of the yOGG1 protein and E. coli Fpg protein does not reveal any obvious homology. The yOGG1 protein has associated DNA glycosylase/AP lyase activity and cleaves abasic sites via -elimination (Girard et al., 1997). In contrast to the bacterial enzyme, the yOGG1 protein repairs 8-oxoG only when paired with pyrimidines. The yOGG1 also removes Fapy guanine-derived residues from DNA but less efficiently than 8-oxoG and -elimination was not observed (Karahalil et al., 1998). The major difference between Fpg and yOGG1 in terms of substrate specificity is that Fpg also removes adenine-derived Fapy lesions (Boiteux et al., 1992) and oxidatively damaged pyrimidines (Jurado et al., 1998; Hatahet et al., 1994). The yOGG1 is not an essential gene, as its disruption had no effect on the viability of a ogg1 haploid mutant but ogg1 cells display a mutator phenotype characterised by an augmentation of G/CT/A transversions (Thomas et al., 1997). Like fpg cells, ogg1 haploid mutants do not show particular sensitivity to H₂O₂. These data suggest that yOGG1 is the functional homologue of E. coli Fpg.

Based on the fact that the helix-hairpin-helix (HhH) DNA-binding motif was found in many DNA binding proteins (Doherty et al., 1996), searches in databases for sequences containing this motif led to identification of the NTG1 gene on chromosome I of S. cerevisiae (Eide et al., 1996). The NTG1 gene encodes a 45 kDa protein, Ntg1p, which is homologous to the E. coli endonuclease III (Nth) but lacks the (4Fe-4S) cluster DNA binding domain found in the other members of this family. Ntg1p is a DNA glycosylase/AP lyase which removes thymine glycols and unexpectedly releases formamidopyrimidine residues from DNA with a high efficiency comparable to that of the E. coli Fpg protein (Eide et al., 1996). Ntg1p also excises other lesions generated by oxidative stress such as 5,6-dihydrouracil (Augeri et al., 1997), and 5-hydroxy-purines and 8-oxoG only when paired with guanine (Sentürker et al., 1998). Targeted disruption of the NTG1 gene results in viable cells. The Ntg1 protein localises both in the nucleus and in the mitochondria and is induced by cell exposure to DNA-damaging agents (Alseth et al., 1999; You et al., 1999). Nash et al. (1996) purified and characterised a second S. cerevisiae Ogg-like protein, called Ogg2, which preferentially acts on 8-oxoG/G base pairs. Later it turned out that Ogg2 is identical to Ntg1 (Sentürker et al., 1998).

NTG2 was identified during analysis of S. cerevisiae genome as a second gene homolog to E. coli Nth in yeast. The NTG2 gene is localised on chromosome XV (Girard and Boiteux, 1997; Alseth et al., 1999) and codes for a 43.7 kDa protein, Ntg2p, which contains (4Fe-4S) cluster DNA binding domain also present in the E. coli Nth protein. The substrate specificity of Ntg2p is similar but not identical to Ntg1p; it cannot incise 8-oxoG mispaired with any of the four DNA bases (Sentürker et al., 1990). Recent results show that Ntg2p, but not Ntg1p, is implicated in the repair of certain degradation products of 8-oxoG such as 8-hydroxydeoxy-guanosine (8-OH-dG) (Kim et al., 2001). The targeted disruption of NTG2 results in viable cells but, similarly to ntg1^- cells, greatly increases the rate of spontaneous and H₂O₂-induced mutations (Alseth et al., 1999). Ntg2p is a nuclear enzyme constitutively expressed in cells. Interestingly, Ntg2p interacts with the DNA mismatch repair protein Mlh1p (Gellon et al., 2002). Ntg1p and Ntg2p are both required for repair of spontaneous and induced oxidative DNA damage.

S. cerevisiae AP endonucleases Apn1 and Apn2: Homologs of E. coli Xth and Nfo have been identified in eukaryotes. The yeast APN1 gene encodes AP endonuclease I (Apn1) homologous to E. coli Nfo protein. This 41.4 kDa protein is the major AP endonuclease in yeast, accounting for >90% (Popoff et al., 1990) of total AP endonuclease activity. Apn1 has AP endonuclease, 3'-diesterase and 3'-phosphatase activities. Like the E. coli homolog, Apn1 is a metalloenzyme excising 3'-phosphoglycoaldehyde, 3'-phosphoryl groups, and 3'-, unsaturated aldehydes (Johnson and Demple, 1988a,b). Yeast mutant lacking Apn1 (apn1) is viable but it is hypersensitive to both oxidative (H₂O₂ and t-butylhydroperoxide) and alkylating (methyl- and ethylmethane sulfonate) agents, it has 6- to 12-fold higher rate of spontaneous mutation than wild-type (Ramotar et al., 1991). This mutator phenotype is mainly characterized by a 60-fold increase in A/T to G/C transversion rate (Kunz et al., 1994). The second yeast AP endonuclease (Apn2/ETH1) was identified by blast homology search with the sequences of E. coli Xth and human Ape1 in the S. cerevisiae genome (Johnson et al., 1998). Apn2 is a 520 amino acid protein, its expression (at the mRNA level) is induced by DNA damage. Apn2 has AP endonuclease, 3'-phosphodiesterase and 3'5' exonuclease activities, it can remove 3'-phosphate and 3' phosphoglycolate termini produced by H₂O₂. Interestingly, the 3' phosphodiesterase and the 3'5' exonuclease activities of Apn2 are 30-40-fold more active than its AP endonuclease activity (Unk et al., 2001). Apn2 could be an important factor in the repair of oxidative DNA damage. Yeast lacking apn2 (apn2) alone are viable and do not show a particular phenotype. However apn1 apn2 cells are remarkably sensitive to DNA damaging agents such as H₂O₂, MMS and phleomycin D1 and they present an increase of spontaneous mutation rate (Bennett, 1999). Apn2 contains a carboxy-terminal domain which is absent in the other members of the family (Xth/Ape1). Deletion of this domain does not affect the enzymatic activity of Apn2 in vitro but the truncated protein cannot remove AP sites in vivo (Unk et al., 2000). These data strongly suggest that this domain is indispensible for protein-protein interaction and that Apn2 protein functions in vivo as a part of multiprotein complex. Interestingly, it has been shown that the human homolog of Apn2 protein, Ape2, interacts with PCNA (Tsuchimoto et al., 2001).

Mammals (human, rodent)

Bifunctional mammalian DNA glycosylases OGG1, NTH1 and NEH1 hOGG1: Several groups reported the molecular cloning of the cDNA of the human and murine 8-oxoguanine DNA glycosylase (OGG1) (reviewed in Boiteux and Radicella, 2000). The amino acid sequence of hOGG1 showed 33% identity and 54% similarity to S. cerevisiae OGG1 and conserved HhH/PVD motif which is present in endoIII/MutY/AlkA superfamily of DNA glycosylases. The human OGG1 gene, located on chromosome 3, codes for seven alternatively spliced forms of mRNA. These transcripts were classified as type 1 and 2 depending on their last exon. The main type 1 and 2 transcripts encode, respectively, a 36 kDa nuclear and a 40 kDa mitochondrial polypeptide (Nishioka et al., 1999; Takao et al., 1998). These two proteins differ by their C-terminus. Both forms possess the same catalytic activity. Purified hOGG1 protein, similarly to E. coli Fpg protein, acts as a DNA glycosylase towards duplex DNA containing 8-oxoG/C base pair, Fapy and has associated AP lyase activity (reviewed in Boiteux and Radicella, 2000) (see Table 1).

Crystal structure data at 2.1 Å resolution of the core domain of hOGG1 complexed with a 8-oxoG/C containing oligonucleotide was reported by Verdine's laboratory (Bruner et al., 2000). This structure reveals that hOGG1 recognizes in the DNA helix both 8-oxoG (using amino acids F319, Q315, G42, C253) and the cytosine opposite 8-oxoG (using amino acids N149, R154, R204, Y203). The modified base is fully extruded from the helix and inserted into an extra-helical active-site pocket on the enzyme. Interestingly, mutations of residue, which recognises the cytosine, does not alter the activity on 8-oxoG/C but significantly increases the activity toward 8-oxoG opposite other bases than cytosine. Physiologically, such mutants become both pro-mutagenic by repair of 8-oxoG/A and anti-mutagenic by repair of 8-oxoG/C. Study of search intermediates of hOGG1 protein in the presence of high molecular weight DNA by atomic force microscope shows that enzyme scans DNA at undamaged sites by inducing drastic kinks (Chen et al., 2002).

The base excision repair pathway for 8-oxoG involves the resynthesis of a single nucleotide at the lesion site as shown by in vitro repair assays with mammalian cell extracts (Dianov et al., 1999; Fortini et al., 1999) and more recently by reconstitution in vitro of this pathway by using human purified proteins (Pascucci et al., 2002). These data show that the BER short-patch pathway is predominant for the repair of major oxidative DNA damage, 8-oxoG. Activity of hOGG1 is greatly stimulated by Ape1 which binds to AP sites generated by the DNA glycosylase and enhances the enzymatic turnover (Vidal et al., 2001a). This observation suggests that Ape1 might preclude the AP lyase activity of OGG1. The finding that the dRP lyase activity of DNA polymerase is required in repair of oxidative lesions by mammalian cell extracts supports this mechanism (Allinson et al., 2001).

The nuclear hOGG1 is associated with chromatin and the nuclear matrix during interphase and with condensed-chromatin during mitosis. The fraction of hOGG1 bound to chromatin is phosphorylated on a serine residue, and the kinase responsible for this post-translational modification of hOGG1 might be protein kinase C (Dantzer et al., 2002). Homozygous ogg1-/- null mice were generated by targeted disruption (Klungland et al., 1999a). These mice are viable and, despite an increase of potentially mutagenic DNA lesions in their genome (Klungland et al., 1999a) and mitochondria (Souza-Pinto et al., 2001), they do not display any particular phenotype. The ogg1-/- null mice have an elevated spontaneous mutation rate in non- or slowly proliferating tissues with high oxygen metabolism, such as liver, but they do not develop malignancies (Klungland et al., 1999a; Osterod et al., 2001). These data are at variance with several reports which associate mutations and polymorphism of the OGG1 gene, in particular the Ser326Cys polymorphism, with increased risk of cancer (Le Marchand et al., 2002; Chevillard et al., 1998; Takezaki et al., 2002). Although the Ser326Cys polymorphism is slightly (Dherin et al., 1999), or not associated with altered OGG1 activity (Janssen et al., 2001), it might affect phosphorylation at this serine residue leading to adverse effects.

hNTH1: A human homolog of the E. coli Nth protein (endonuclease III) has been purified (Eide et al., 1996) and cloned (Aspinwall et al., 1997; Hilbert et al., 1997). This gene, called hNTH1 (human Nth homolog 1), encodes a 34.3 kDa protein that shares extensive homology with Nth including the conserved HhH-motif and the (4Fe-4S) cluster loop motif (Hilbert et al., 1997; Ikeda et al., 1998). hNTH1 protein acts as DNA glycosylase with an associated AP lyase activity, and has substrate specificity similar to the E. coli homolog towards damaged pyrimidine derivatives that result from ring saturation, ring fragmentation, ring contraction and unexpectedly ring-opened purine residues (Fapy) (Dizdaroglu et al., 1999). hNTH1 acts preferentially on 5-hydroxycytosines and AP sites when they are situated opposite to guanine (Eide et al., 2001) and removes 8-oxoG opposite G (Matsumoto et al., 2001). hNTH1 seems to be an exclusively nuclear protein and its expression is regulated during the cell cycle, showing a maximum in S-phase (Luna et al., 2000). Targeted disruption of Nth1 in mice (mNTH1) has been reported by Takao et al. (2002). Homozygous mNth1-/- mutant mice do not have any detectable phenotype defect. Sensitivity of mNth1-/- mutant embryonic cells to H₂O₂ and menadione was not changed as compared to wild-type cells. Moreover, thymine glycol induced by ionising radiation is repaired, though more slowly than in wild type. The absence of a significant phenotype could be explained by the discovery of two novel thymine-glycol DNA glycosylases in mNth1 mutant cells called TGG1 and TGG2. These ~40 kDa proteins localise in mitochondria and nucleus, respectively. Thus, these data suggest the existence of several back-up DNA glycosylases in mammals to remove thymine glycol. The redundancy of such enzymatic activity highlights the importance of BER pathway for counteracting oxidative pyrimidine damage.

Although there is no definite evidence that BER enzymes act as multiprotein complexes, several reports suggest that addition of proteins can stimulate damage recognition and lesion processing. The NER enzyme XPG stimulates the hNTH1 activity in vitro (Klungland et al., 1999b; Bessho, 1999). These data support the hypothesis that neurodegeneration in XP-patients could be due to defects in the repair of oxidative DNA damage (Reardon et al., 1997). A yeast two-hybrid screen for other interacting partners of Nth1 led to the isolation of the DNA-binding protein B (DbpB)/Y box binding protein (YB-1) (Marenstein et al., 2001). YB-1 greatly stimulates the hNTH1 activity. Altogether these data suggest that the BER pathways could be regulated by a number of non-BER proteins.

hNEH1: Human homologs of E. coli Nei/MutM were identified by database search of the genome and named hNEH1 and 2 (human Nei Homolog 1 and 2). hNEH1 was biochemically characterized (Hazra et al., 2002), it encodes a 44 kDa protein which excises Fapy, oxidised purines and 8-oxoG when present in DNA. hNEH1 shows a tissue-specific expression with an S-phase specific increase at both RNA and protein level. Tissue-specific mRNA levels of OGG1 and NEH1 are distinct. These data suggest that the activities of OGG1 and NEH1 are not redundant at the tissue level and NEH1 might be involved in the replication-coupled repair of oxidative DNA damage.

Monofunctional DNA glycosylases ANPG, hTDG, MYH

ANPG: The ANPG protein is the human counterpart of the E. coli AlkA protein. This protein releases 3-meA and 7-meG from methylated DNA (O'Connor and Laval, 1991), but it is also able to excise hypoxantine (Saparabaev and Laval, 1994) and 1,N⁶-ethenoadenine (Singer et al., 1992; Saparbaev et al., 1995) more efficiently than the AlkA protein. In contrast to the AlkA protein, the ANPG protein does not release 5-formyluracil and RT residues from DNA, however, there are activities in human cell-free extracts that can excise these lesions (Bjelland et al., 1995; Privezentzev et al., 2001). It should be noted that the human protein does not share significant amino acid sequence homology with the bacterial AlkA. Recently, it has been shown that ANPG efficiently excises 1,N²-ethenoguanine (1,N²-G) when present in DNA (Saparbaev et al., 2002b). Immunofluorescent staining for the ANPG protein in normal breast cells showed its nuclear localisation (Cerda et al., 1998). Two alternatively spliced transcripts of human ANPG have been isolated from human cells. The gene encoding for ANPG maps to chromosome 16 (Samson et al., 1991; Vickers et al., 1993). The full-length cDNA first isolated by Samson and colleagues (ANPG70, 293 AA, 1991) differs from the splice variant (ANPG60, 298 AA, Pendlebury et al., 1994) only in the sequence of the first 8 and 13 N-terminal amino acids respectively. Two truncated versions of human ANPG have also been described: ANPG40 lacks the first 63 amino acids and ANPG80 the first 73 amino acids from the N-terminus, when compared to APNG70 (O'Connor, 1993; Lau et al., 1998). It has been concluded that the non-conserved, N-terminal part apparently contributes little to their damage recognition and N-glycosylase activity (O'Connor, 1993; Roy et al., 1998). Surprisingly, it has been shown that the ANPG80, which lacks 73 amino acid residues at the N-terminus, is unable to excise 1,N²-G from the duplex oligonucleotide and that the ANPG40 has a reduced activity as compared to the ANPG70 (Saparbaev et al., 2002b). These observations indicate that the non-conserved, N-terminal part of ANPG is essential for 1,N²-G-glycosylase activity but dispensable for the release of A, hypoxanthine and N-methylpurines. The ANPG activity toward hypoxanthine-containing substrates is slightly activated in the presence of the hHR23 protein (human homolog of yeast RAD 23) implicated in NER pathway. ANPG and hHR23 act together in a complex with a greater affinity for damaged DNA than ANPG alone (Miao et al., 2000; Sidorkina and Laval, unpublished experiments 2002).

Although the core part of the ANPG70 and APDG proteins display 85% sequence identity (O'Connor and Laval, 1991), it seems that human enzyme repairs 1,N²-G somewhat more efficiently than its rat counterpart (Saparbaev et al., 2002b). In fact, a similar difference between human and mouse ANPG in the recognition of 3-methylguanine and 7-methylguanine residues has been reported (Roy et al., 1996). Interestingly, the human full-length and splice variants of ANPG have a direct repeat in N-terminal amino acid sequence, which is absent in the rat and murine homologues. These observations further support the role of the non-conserved N-terminal part of mammalian ANPG in substrate specificity. To assess the role of this repair enzyme in vivo, mice deficient in 3-meAde-DNA glycosylase (APNG/Aag null mice) have been generated (Hang et al., 1997; Engelward et al., 1997). Unexpectedly, these APNG knockout mice exhibit neither any particular sensitivity to alkylating agents nor any significant increase in the spontaneous mutation rate except for splenic T lymphocytes (Elder et al., 1998). Furthermore, Roth and Samson (2002) have shown that Aag-/- myeloid progenitor bone marrow cells are more resistant than wild-type cells to alkylating agents. This result suggests that the initiation of base excision repair could be more lethal to the cell than leaving the damaged bases unrepaired.

The crystal structures of ANPG80 (a truncated form of the ANPG protein) complexed to a DNA duplex containing pyridine and A have been established (Lau et al., 1998, 2000). ANPG80 is a single domain protein of mixed / structures, which forms a distinct structural group that does not resemble any other BER protein (Hollis et al., 2000). The enzyme bends the DNA by about 20° and intercalates into the minor groove of DNA causing the abasic pyrrolidine and A to slip into the enzyme active site. The structure of the ANPG80 nucleotide-binding pocket changes little upon flipping in the A base. The structure of the base-binding pocket of ANPG reveals the lack of specific interactions for methylated bases. This fact probably contributes to its remarkable multifunctionality, providing glycosylase activity within the same enzyme for deamination, methylation and exocyclic base damage.

hTDG: The human thymine DNA glycosylase (hTDG) was first biochemically characterised for its ability to remove T mispaired with G (Neddermann and Jiricny, 1993). A more complete characterisation shows that hTDG exhibits a broad substrate specificity (Neddermann and Jiricny, 1994): it can excise T from DNA not only when mispaired with G but also from other T-containing mispairs except T/A. Moreover, the hTDG protein more efficiently removes U from a G/U but not from a U/A mispair and 3,N⁴-ethenocytosine opposite to all four natural bases (Neddermann and Jiricny, 1994; Saparbaev and Laval, 1998; Hang et al., 1998). The hTDG gene is localised on the chromosome 12 (Sard et al., 1997), it codes for several mRNAs which were cloned (Neddermann et al., 1996). These cDNA encode a single 46 kDa polypeptide (410 amino acids) which is homologous to the E. coli MUG protein. The hTDG protein has a high affinity for AP site and stays tightly bound to it after removal of T from a G/T mispair. Ape1 activates hTDG by increasing the dissociation rate of hTDG from AP site probably through direct interaction (Waters et al., 1999). The stimulation of hTDG turnover by Ape1 is responsible for the efficient processing of 3,N⁴-ethenocytosine by hTDG (Privezentzev et al., 2001). A second factor facilitating hTDG enzymatic turnover was recently identified. Hardeland et al. (2002) have shown that SUMOylation of hTDG by SUMO-1 and SUMO-2/3 drastically reduces its affinity for AP sites, increases its enzymatic turn-over towards G/U and reduces its processing activity on G/T substrate. SUMOylation also enhances the stimulatory effect of Ape1 on hTDG.

Using the yeast one-hybrid screen, hTDG was found to interact with retinoic acid receptor (RAR) and retinoic X receptor (RXR) (Um et al., 1998) which are ligand-dependent transcription factors. Interaction between RXR, RAR/RXR and hTDG enhances the binding of the former on their responsive elements. These data suggest that hTDG has a dual function: repair enzyme and transcriptional activator. hTDG is also a transcriptional repressor since it interacts with the thyroid transcription factor 1 (TTF-1) and then strongly inhibits the expression of TTF-1 responsive genes (Missero et al., 2001). Recently, hTDG has been found to interact with the transcriptional co-activator CBP/p300 (Tini et al., 2002). The hTDG/CBP/p300 complex is competent for both BER and histone acetylation. hTDG enhances the CBP/p300 transcriptional activity and these proteins acetylate it. hTDG acetylation regulates the recruitment of Ape1 by hTDG. These data highlight the complexity of repair processes that are coupled to transcription, tightly regulated by post-translational modifications (SUMOylation and acetylation) and by protein-protein interactions.

SMUG1: The single-strand mismatch-specific uracil-DNA glycosylase (SMUG1) has been identified only in vertebrates by expression cloning (Haushalter et al., 1999). SMUG1 is a nuclear protein of 31 kDa which in fact efficiently removes uracil from double stranded DNA when paired with A and G, it also removes U from single-stranded substrates but less efficiently (Nielsen et al., 2001). Accumulation of 5-hydroxymethyluracil is associated with a cancer risk (Frenkel et al., 1998). The mammalian 5-hydroxymethyluracil DNA N-glycosylase activity is associated with SMUG1 (Boorstein et al., 2001).

Like several DNA glycosylases, SMUG1 has a low turnover rate that could be due to its inhibition by AP sites. This could explain the Ape1-dependent stimulation of the hSMUG1 activity on both single and double-stranded DNA substrates (Nielsen et al., 2001). Targeted disruption of UNG gene in mice (ung-/-) does not result in relevant increase of spontaneous mutation rates, contrary to bacteria and yeast ung^- mutants (Nielsen et al., 2000). It seems that hSMUG1 has an important role in the maintenance of genome stability since it represents the major uracil-DNA glycosylase activity in ung-/- cells (Nielsen et al., 2001).

hMYH: The human adenine-DNA glycosylase activity, called hMYH protein (human MutY homolog) was first purified from cellular extracts (Mcgoldrick et al., 1995). Later using homology search, the genomic DNA and cDNA encoding hMYH were identified (Slupska et al., 1996). Functional expression of hMYH in E. coli complements the mutator phenotype of mutY cells (Slupska et al., 1999). The hMYH gene is localised on chromosome 1. It encodes a 59 kDa monofunctional DNA glycosylase which can excise A when opposite to 8-oxoG and to a lesser extent opposite to G. hMYH, similar to OGG1, is able to excise 2-OH-A, a mutagenic oxidative adduct (Kamiya and Kasai, 1995; Ohtsubo et al., 2000). hMYH interacts with several proteins involved in the long-patch repair pathway such as PCNA, Ape1 and RPA (Parker et al., 2001). In addition, hMYH interacts with the mismatch repair proteins hMSH2/hMSH6 (Gu et al., 2002). Importantly, this interaction enhances the hMYH activity toward A/8-oxo-G. hMYH DNA glycosylase activity is equally enhanced by Ape1 on A/8-oxoG substrates. Unexpectedly this stimulation is independent of Ape1 AP endonuclease activity. Instead, Ape1 acts by stimulation of the formation of hMYH/DNA complexes (Yang et al., 2001). These data suggest that different repair pathways may cooperate through protein/protein interactions when counteracting oxidative DNA damage.

Ten different forms of mRNAs coding for hMYH were isolated. These mRNA encodes three proteins of 52, 53 and 57 kDa respectively. The 52 kDa and 53 kDa forms localise in the nucleus and the 57 kDa form in the mitochondria (Ohtsubo et al., 2000). Expression of hMYH is cell cycle-regulated with a maximum in S phase. Nuclear forms of hMYH co-localise with PCNA at DNA replication foci (Boldogh et al., 2001), suggesting a role of hMYH in replication-associated base excision repair. Recently biochemical evidence has been provided that human cell extracts perform base excision repair of 8oxo-G/A mismatches on both strands (Parlanti et al., 2002). Similarly to what is reported in yeast, following adenine excision a cytosine is preferentially inserted opposite 8oxoG. This is followed by excision repair of 8oxoG in 8oxoG/C pair. Interestingly, repair synthesis on either strand is completely inhibited by aphidicolin, suggesting that a replicative DNA polymerase is involved in the gap filling reaction (Parlanti et al., 2002). DNA polymerases / are likely to be involved in this replication-associated BER. This was confirmed in vivo using murine MYH-deficient cells (Hayashi et al., 2002). MYH-/- cells repair A/8oxoG mispairs inefficiently but expression of wild-type MYH in these cells increases the repair. Interaction with PCNA is critical, since expression of a functional MYH lacking its PCNA-binding domain had no effect on the repair efficiency of A/8oxoG in MYH-/- cells. It has been speculated that hMYH might be considered as a cancer predisposing gene in humans since several mutations in this gene are linked with somatic mutations in adenomatous polyposis coli gene (APC) (Al-Tassan et al., 2002).

AP endonucleases Ape1, Ape2: The major human AP endonuclease (HAP-1/Ape1/Apex), homologous to E. coli Xth protein (exonuclease III), was independently discovered as an AP-endonuclease (Demple et al., 1991; Robson and Hickson, 1991) and as redox-regulator of the DNA binding domain of Fos-Jun, Jun-Jun, AP-1 proteins and several other transcription factors including NF-kappa B, Myb and members of the ATF/CREB family (Xanthoudakis et al., 1992). Ape1 also activates p53 (Jayaraman et al., 1997). Ape1 is a 35.5 kDa nuclear protein (Demple et al., 1991), which shows sequence homology to the E. coli Xth protein. Targeted homozygous disruption of Apex gene (Apex-/-) results in embryonic lethality in mice (Xanthoudakis et al., 1996). The heterozygous Apex+/- mice, so far, did not reveal significant differences in phenotypes associated with oxidative stress as compared to the wild-type (Meira et al., 2001). Due to the dual role of Ape1, the early embryonic lethality could be associated either with the BER defect or with the defective regulation of several transcription factors. Ape1 can substitute for exonuclease III in E. coli and for Apn1 in S. cerevisiae (Wilson et al., 1995a). Beside its AP endonuclease activity (Wilson et al., 1995b), it exhibits other enzymatic activities: 3'5' exonuclease (Chou et al., 2000), phosphodiesterase, 3' phosphatase and RNase H (Barzilay et al., 1995; reviewed in Rothwell et al., 1997). It should be noted that these additional activities are 3-4-fold weaker than its AP endonuclease activity. In addition, Chou and Cheng (2002) have shown that Ape1 has a 3'-mismatch exonuclease activity and so it might be considered as a proof-reading enzyme.

Ape1 plays a central role in BER. It is implicated in both short-patch (Kubota et al., 1996) and long-patch repair pathway (Frosina et al., 1996). It could be also an important regulator of BER. Ape1 interacts with DNA polymerase in vivo (Bennett et al., 1997) which is a key element for BER (Sobol et al., 1996), this interaction permits the loading of pol on DNA at the AP site and the enhancement of its dRPase activity. Ape1 interacts with other BER proteins such as the scaffold protein XRCC1, which stimulates its activity (Vidal et al., 2001b), and with PCNA and the Flap endonuclease 1 (FEN1) (Dianova et al., 2001). The p21^WAF1/CIP1 is involved in the regulation of cell cycle progression, DNA replication and DNA repair (reviewed in Dotto, 2000). It can bind to PCNA inhibiting DNA replication and PCNA stimulation of long-patch BER pathway. Ape1 seems to be implicated in the compensation of the inhibitory effect of p21 on BER by stimulating and co-ordinating the long-patch BER (Tom et al., 2001). In addition, Ape1 is an activator of DNA glycosylases (Waters et al., 1999; Vidal et al., 2001a; Yang et al., 2001; Nielsen et al., 2001). Altogether these data suggest that BER pathways are co-ordinated and regulated by protein-protein interactions and that Ape1 plays an important role in these interactions. Moreover, it has been shown that Ape1 and hnRNP-L are the components of the nCARE-B2 element binding complexes in the ape1 gene promoter (Kuninger et al., 2002). This result suggests that expression of the ape1 gene is down regulated by its own product. Ape1/ref-1 could also be involved in cell response to chemotherapy since it was found that nuclear accumulation of Ape1 is inversely proportional to that of p53 in head-and-neck cancer (Koukourakis et al., 2001).

A second AP endonuclease, Ape2, homologous to the S. cerevisiae Apn2, was identified in mammals (Hadi and Wilson, 2000). The ape2 gene is localised on the X chromosome and Ape2 transcripts are ubiquitously expressed. Ape2, a 57.3 kDa protein, exhibits only a weak ability to complement E. coli and S. cerevisiae AP endonuclease deficient cells. Ape2 localises predominantly in the nucleus but also in the mitochondria (Tsuchimoto et al., 2001). In the nucleus, Ape2 co-localises with PCNA and interacts with it in vitro. These data suggest that Ape2 is implicated in both mitochondrial and nuclear BER and also that the nuclear role of Ape2 is PCNA-dependent.

Other organisms (Arabidopsis thaliana and Drosophila melanogaster)

Evidences have accumulated that oxidative DNA damage are repaired also through BER pathway in plants and insects. Homologues of the Escherichia coli Nth and Fpg proteins were characterized in Arabidopsis thaliana (Roldan-Arjona et al., 2001; Terashima et al., 1998;Garcia-Ortiz et al., 2001). Previously it was suggested that Drosophila has no BER pathway (Deutsch and Spiering, 1982). However, an homologue of the human OGG1 protein in Drosophila melanogaster has been identified and characterized (Dherin et al., 2000).

A new DNA glycosylase from A. thaliana ROS1 is a 1393 aminoacids protein, with an endonuclease III domain at the C-terminus. The recombinant ROS1 is able to incise oxidatively damaged DNA and also DNA containing 5 meC, suggesting a role in controling DNA methylation levels. This agrees with the pleiotropic developmental defects of the ros1 mutants suggesting regulatory properties of this protein. These results suggest that this protein could be a new class of DNA glycosylases, perhaps only present in plants. (Gong et al., 2002).

Alternative repair pathways

So far, several alternatives to the BER repair pathways for oxidative DNA damage have been described.

Nucleotide Incision Repair (NIR)

The BER pathway (Lindahl, 1976) requiring sequential action of two enzymes for incision of DNA (Laval, 1977) raises theoretical problems for the efficient repair of oxidative DNA damage because it generates genotoxic intermediates such as abasic (AP) sites and/or blocking 3'-termini groups that must be eliminated by additional steps before initiating DNA repair synthesis. In addition, biological evidence hints at the existence of an alternative repair pathway. E. coli and mouse mutants deficient in the DNA glycosylases that remove oxidised bases are not sensitive to reactive oxygen species (Laval et al., 1998; Blaisdell and Wallace, 2001; Klungland et al., 1999a; Takao et al., 2002). A recent finding that Nfo and Nfo-like endonucleases nick DNA on the 5' side of various oxidatively damaged bases, generating 3'-hydroxyl and 5'-phosphate termini, provides an alternative pathway to the classic BER (Ishchenko and Saparbaev, 2002) (see Table 1). It was proposed to name it the nucleotide incision repair (NIR) pathway (Figure 1). The proposed NIR pathway eliminates the genotoxic intermediates, explains the genetic data and most likely identifies the physiological and original target as the modified base and not an artificial reduced abasic site, for the long patch repair pathway described in human cells (Klungland and Lindahl, 1997). This alternative repair pathway is evolutionarily conserved from E. coli to human. The NIR pathway directly generates proper ends for DNA replication on one side and on the other side the site for elimination of the lesion by specific nucleases. This mechanistic characteristic creates an advantage as compared to BER.

Nucleotide Excision Repair

Oxidative DNA damage is a major source of genetic mutation, and it is implicated in several human disorders including Xeroderma pigmentosum (XP). Human disorder XP is characterised by defect in nucleotide excision repair, the cells from XP patients fail to remove pyrimidine dimers caused by sunlight and, as a consequence, develop multiple cancers in areas exposed to light. Despite the fact that neural tissue is shielded from sunlight-induced DNA damage, a fraction of XP patients (~18%) display progressive neurologic degeneration secondary to a loss of neurones. In fact, it was shown that in E. coli the nucleotide excision repair pathway could be involved in repair of oxidative DNA damage (Lin and Sancar, 1989; Czeczot et al., 1991). Furthermore, in human cells two major oxidative DNA lesions, 8-oxoguanine and thymine glycol, are excised from DNA in vitro by the same enzyme system responsible for removing pyrimidine dimers and other bulky DNA adducts (Reardon et al., 1997). These results support the idea that the NER pathway could be directly involved in repair of oxidative DNA damage.

Transcription-Coupled Repair (TCR)

Transcription-coupled repair (TCR) is a pathway of DNA excision repair that preferentially removes the DNA lesions present in transcribed sequences of expressed genes (Hanawalt, 2000). TCR was originally observed for DNA damage repaired by NER pathway. However, recent studies demonstrated that Cockayne syndrome (CS) cells which are deficient for TCR cannot remove 8oxo-G in transcribed sequence, despite its proficient repair in non-transcribed sequence (Le Page et al., 2000a,b). Furthermore, it has been demonstrated that the breast and ovarian cancer susceptibility genes, BRCA1 and BRCA2, may participate in the repair of the 8oxo-G residues specifically located on the transcribed DNA strand in human cells (Le Page et al., 2000c).

Translesional DNA synthesis (TLS)

In E. coli, in addition to BER pathway, the translesion DNA synthesis system can contribute to the repair of chromosomal AP sites in vivo (Otterlei et al., 2000). In the last few years, new human DNA polymerases Pol-, Pol-, and Pol- , which are able to promote replication through DNA lesions were discovered. These polymerases catalyse translesion DNA synthesis (TLS) in a distributive manner and tend to show lower replication fidelity than other polymerases when unmodified DNA is used as a template (Friedberg et al., 2002). The yeast and human Pol- replicate DNA containing 8-oxo-G efficiently and accurately by inserting a cytosine across from the lesion (Haracska et al., 2000). In support of the biochemical data, genetic studies show that synergistic increase in the rate of spontaneous mutations occurs in the absence of Pol- in the yeast ogg1 mutant. Both error-free and error-prone synthesis were observed in DNA synthesis across A residue catalysed by Pol- and Pol- (Levine et al., 2001). It has been found that Pol- preferentially misinsert G opposite to uracil thus providing a mechanism whereby mammalian cells can decrease the mutagenic potential of lesions formed via the deamination of cytosine (Vaisman and Woodgate, 2001). Altogether these observations suggest an additional role for TLS-specific DNA polymerases in the prevention of the mutagenic replication of oxidative DNA damage.

DNA Mismatch repair (MMR) pathway

DNA mismatch-repair system is involved in the repair of mispaired bases formed during replication, genetic recombination and as a result of DNA damage. Studies in S. cerevisiae indicate the involvement of the mismatch repair pathway in prevention of genotoxic effect of oxidative DNA damage. It was shown that mutations in MSH2 and MSH6 caused a synergistic increase in mutation rate in combination with mutations in OGG1, resulting in a 140- to 218-fold increase in the G/C-to-T/A transversion rate (Ni et al., 1999). Consistent with this, the MSH2-MSH6 complex binds to 8oxoG:A mispairs and 8oxo-G/C base pairs with high affinity and specificity.

Another important source of 8oxo-G in DNA is the dNTP pool. Recently it has been shown that incorporated 8oxo-dGMP contributes significantly to the mutator phenotype of MMR-deficient cells (Colussi et al., 2002). Increased expression of MTH1 in MSH2-/- cells produces a significant decrease of DNA 8oxo-G levels and is associated with a drastic reduction of the mutation rate. These findings indicate that MMR excises incorporated 8oxoGMP from newly synthesised DNA strand, thus providing cells with an additional protective strategy from oxidative stress.

Homologous recombination repair pathway

It is well established that homologous recombination (HR) system is pivotal for the repair of an important form of oxidative DNA damage induced by ionising radiation - the double-strand break (DSB) (van Gent et al., 2001). Moreover, in vivo replication forks often encounter template DNA damage that can lead to replication fork collapse generating double-strand break. It is hypothesised that DNA recombination functions as a system for reactivation of the collapsed replication fork (Cox et al., 2000). The importance of this function to the formation of lethal DNA damage is underscored by the uncontrollable fragmentation of chromosomes in Rad51-deficient vertebrate cell lines (Lim and Hasty, 1996; Sonoda et al., 1998). Another important characteristic of the HR repair pathway is that it mainly provides a non-mutagenic replicative bypass of the blocking lesion using sister chromatid or homologous chromosome as an undamaged DNA template (Greenberg et al., 2001). Recently, it has been found that defects in the homologous recombination pathway are associated with cancer-prone clinical syndromes, in particular ataxia telangiectasia, hereditary breast cancer, Bloom's syndrome and Werner's syndrome. These observations support the idea that the HR system could be an alternative to excision/incision repair pathways in counteraction of genotoxic effects of oxidative DNA damage.

Nucleotide pool-sanitising enzymes: MutT (E. coli) and MTH1 (human and mice)

Besides the modification of DNA by ROS, these species also oxidise the nucleotide pool. The modified nucleotides could be introduced in DNA and also in RNA during their synthesis leading to mutations. The E. coli mutT mutant shows a 100- to 10 000-fold increase in A/TG/C transversion rates as compared to wild-type. This effect is due to the accumulation of mutagenic 8oxo-GTP in the cellular dNTP pool which is incorporated opposite A and C during DNA synthesis, thus leading to A/TG/C transversions. The E. coli MuT protein is a 8oxo-GTPase that cleanses the cellular nucleotide pool from 8oxoGTP (Maki and Sekiguchi, 1992). In addition, MutT hydrolyses ribonucleotide 8oxo-GTP (Taddei et al., 1997), suggesting the importance of this enzyme in maintaining transcription fidelity. Mammalian cells contain MutT homolog, the MTH1 protein (MutT Homolog 1). The corresponding cDNAs were cloned in human (Furuichi et al., 1994) and murine cells (Kakuma et al., 1995), and protein was characterised. MTH1 displays only 23% identity to the E. coli counterpart but the highly conserved region essential for the 8oxo-GTPase activity contains identical residues (Fujii et al., 1999). Therefore, expression of MTH1 human and rodent origins in E. coli mutT reverts to the mutator phenotype (Furuichi et al., 1994; Kakuma et al., 1995). These results indicate that mammalian proteins may have a similar antimutagenic activity in vivo as E. coli MutT protein. Like MutT, MTH1 hydrolyses 8oxo-GTP (Mo et al., 1992) and oxidised ribonucleotides 8oxo-GTP (Hayakawa et al., 1999; Fujikawa et al., 2001). Interestingly, MTH1 possess an additional 2-hydroxyadenosine (2-OH-A) triphosphate pyrophosphorylase activity (Fujikawa et al., 1999 and 2001). Since hMYH (MutY homolog) is able to remove 2-OH-A from DNA, these data suggest that mammalian cells possess a pathway to sanitise oxidised forms of both guanine and adenine. Seven types of mRNA transcripts, coded by the MTH1 gene, have been identified in human Jurkat cells (Oda et al., 1997), which lead to the synthesis of, at least, three polypeptides of 18, 21 and 22 kDa (p18, p21 and p22) (Oda et al., 1999). MTH1 polypeptides localise mainly in the cytoplasm and in mitochondria (Kang et al., 1995; Oda et al., 1999; Nakabeppu, 2001).

The MTH1 knockout cell lines and mice were constructed by gene targeting (Tsuzuki et al., 2001). MTH1-/- cells exhibit a twofold increase of mutation rate as compared to MTH1+/+ cells which is in contrast to the increase of mutation rate in E. coli mutM cells. Therefore, it seems likely that mammalian cells have other enzyme(s) and/or pathway(s) capable of hydrolysing oxidised nucleotides. MTH1-/- mice exhibit slightly elevated tumour incidence than wild-type mice but their survival at 1.5 years is not affected. Several studies on human cancers show that mutations in hMTH1 gene, so far are not involved in these pathologies (Wu et al., 1995; Takama et al., 2000). Unexpectedly, MTH1 and translesional DNA polymerase kappa (Pol-) expressions were found to significantly increase and decrease respectively in two mammary carcinoma cell lines characterised by frequent A/TG/C transversions (Okochi et al., 2002).

Conclusion

So far, no human diseases have been linked to defects in proteins involved in the BER pathway. DNA repair genes functionally expressed in mammalian cells and now in transgenic mice having a null mutation in the gene coding for BER proteins are very important tools to ascertain the biological role of these proteins in vivo. It has been very astonishing to notice that, beside a few examples of targeted deletion of genes encoding the AP-endonuclease 1 and DNA polymerase in mice leading to embryonic lethality (Xanthoudakis et al., 1996; Sobol et al., 1996), the genotype of the other BER protein knockout mice (such as a number of DNA glycosylases) does not show any striking particularity in terms of predisposition to cancer and premature aging. Furthermore, examination of 6200 S. cerevisiae genes transcript levels after exposure to various genotoxic agents reveals that the DNA repair genes are only modestly induced (Jelinsky et al., 2000), which is in contrast to adaptive and SOS response in E. coli (Lindahl et al., 1988; Sutton et al., 2000). These observations suggest the possibility of back-up repair pathway(s) for oxidative DNA damage that have to be characterized. One could expect important breakthroughs from crosses between different strains to produce double knockouts to identify the possible back-up systems, the processes involved in the regulation and the interactions of the different pathways.

Detailed understanding of the mechanisms leading to the co-ordination of various proteins involved in the molecular reaction of BER is of paramount importance for gaining insights into the efficiency and fidelity of this key pathway for genome stability, prevention of cancer, resistance to chemotherapeutic agents, degenerative diseases and more recently in some aspects of teratogenicity.

Acknowledgements

We wish to thank Dr Eugenia Dogliotti and Dr Betsy Sutherland for constructive comments. We thank 'Electricité de France', the 'Association pour la Recherche contre le Cancer' and the Franco-Norwegian Foundation and European Community for grants.

Ahnstrom G, Bryant PE. (1982). Int. J. Radiat. Biol., 41: 671-676.

Allinson SL, Dianova II, Dianov GL. (2001). EMBO J., 20: 6919-6926.

Alseth I, Eide L, Pirovano M, Rognes T, Seeberg E, Bjoras M. (1999). Mol. Cell. Biol., 19: 3779-3787.

Al-Tassan N, Chmiel NH, Maynard J, Fleming N, Livingston AL, Williams GT, Hodges AK, Davies DR, David SS, Sampson JR, Cheadle JP. (2002). Nat. Genet., 30: 227-232. Article

Aravind L, Koonin EV. (2000). Genome Biol., 1: 1-8. Article

Asahara H, Wistort PM, Bank JF, Bakerian RH, Cunningham RP. (1989). Biochemistry, 28: 4444-4449.

Aspinwall R, Rothwell DG, Roldan-Arjona T, Anselmino C, Ward CJ, Cheadle JP, Sampson JR, Lindahl T, Harris PC, Hickson ID. (1997). Proc. Natl. Acad. Sci. USA, 94: 109-114. MEDLINE

Augeri L, Lee YM, Barton AB, Doetsch PW. (1997). Biochemistry, 36: 721-729.

Bailly V, Verly WG. (1987). Biochem. J., 242: 565-572.

Bailly V, Verly WG, O'Connor TR, Laval J. (1989). Biochemistry, 262: 581-589.

Barrett TE, Savva R, Panayotou G, Barlow T, Brown T, Jiricny J, Pearl LH. (1998). Cell, 92: 117-129. MEDLINE

Barzilay G, Hickson ID. (1995). Bioessays, 17: 713-719.

Barzilay G, Wlaker LJ, Robson CN, Hickson ID. (1995). Nucleic Acids Res., 23: 1544-1550.

Bauche C, Laval J. (1999). J. Bact., 181: 262-269.

Bauerle PA, Rupec RA, Pahl HL. (1996). Pathol. Biol., 44: 29-35.

Bennett RA, Wilson DM, Wong D, Demple B. (1997). Proc. Natl. Acad. Sci. USA, 94: 7166-7169. Article MEDLINE

Bennett RA. (1999). Mol. Cell Biol., 19: 1800-1809.

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

Bhagwat M, Gerlt JA. (1996). Biochemistry, 35: 659-665. Article MEDLINE

Bjelland S, Bjoras M, Seeberg E. (1993). Nucleic Acids Res., 21: 2045-2049. MEDLINE

Bjelland S, Eide L, Time RW, Stote R, Eftedal I, Volden G, Seeberg E. (1995). Biochemistry, 34: 14758-14764.

Bjelland S, Birkeland NK, Benneche T, Volden G, Seeberg E. (1994). J. Biol. Chem., 269: 30489-30495.

Blaisdell JO, Wallace SS. (2001). Proc. Natl. Acad. Sci. USA, 98: 7426-7430. Article MEDLINE

Blaisdell JO, Harrison L, Wallace SS. (2001). Radiat. Prot. Dosimetry, 97: 25-31.

Blaisdell JO, Hatahet Z, Wallace SS. (1999). J. Bacteriol., 181: 6396-6402. MEDLINE

Boiteux S, Radicella JP. (2000). Arch. Biochem. Biophys., 377: 1-8. Article MEDLINE

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

Boiteux S, O'Connor TR, Laval J. (1987). EMBO J., 6: 3177-3183. MEDLINE

Boiteux S, O'Connor TR, Lederer F, Gouyette A, Laval J. (1990). J. Biol. Chem., 265: 3916-3922. MEDLINE

Boiteux S. (1993). J. Photochem. Photobiol. B. Biol., 19: 87-96.

Boiteux S, Laval J. (1982). Biochemistry, 21: 6746-6751. MEDLINE

Boldogh I, Milligan D, Soog M, Bassett H, Lloyd RS, McCullough A. (2001). Nucleic Acids Res., 29: 2802-2809. MEDLINE

Bonura T, Smith KC, Kaplan HS. (1975). Proc. Natl. Acad. Sci. USA, 72: 4265-4269.

Boorstein RJ, Chiu LN, Teebor GW. (1992). Mol. Cell Biol., 12: 5536-5540.

Boorstein RJ, Cummings JrA, Marenstein DR, Chan MK, Ma Y, Neubert TA, Brown SM, Teebor GW. (2001). J. Biol. Chem., 276: 41991-41997.

Brooks PJ, Wise DS, Berry DA, Komoski JV, Smerdon MJ, Somers RL, Mackie H, Spoonde AY, Ackerman EJ, Coleman K, Tarone RE, Robbins JH. (2000). J. Biol. Chem., 275: 22355-22362. Article MEDLINE

Bruner SD, Norman DPG, Verdine GL. (2000). Nature, 403: 859-866. Article MEDLINE

Budworth H, Dianova I, Podust VN, Dianov Gl. (2002). J. Biol. Chem., 277: 21300-21305.

Burgess S, Jarugas P, Dodson ML, Lloyd SR. (2002). J. Biol. Chem., 277: 2938-2944.

Cadet J, Delatour T, Douki T, Gasparutto D, Pouget JP, Ravanat JL, Sauvaigo S. (1999). Mutat. Res., 424: 9-21.

Cerda SR, Turk PW, Thor AD, Weitzman SA. (1998). FEBS Lett., 431: 12-18.

Chan E, Weiss B. (1987). Proc. Natl. Acad. Sci. USA, 84: 3189-3193. MEDLINE

Chang DY, Lu AL. (2002a). Mol. Genet. Genomics, 266: 336-342.

Chang DY, Lu AL. (2002b). J. Biol. Chem., 277: 11853-11858.

Chaudhry MA, Weinfeld MJ. (1997). J. Biol. Chem., 272: 15650-15655.

Chaudhry MA, Weinfeld M. (1995). J. Mol. Biol., 249: 914-922.

Chen L, Haushalter KA, Lieber CM, Verdine GL. (2002). Chem. Biol., 9: 345-350.

Chevillard S, Radicella JP, Levalois C, Lebeau J, Poupon MF, Oudard S, Dutrillaux B, Boiteux S. (1998). Oncogene, 16: 3083-3086. Article MEDLINE

Chmiel NH, Golinelli MP, Francis AW, David SS. (2001). Nucleic Acids Res., 29: 553-564. Article MEDLINE

Chou KM, Cheng YC. (2002). Nature, 415: 655-659. Article MEDLINE

Chou KM, Kukhanova M, Cheng YC. (2000). J. Biol. Chem., 275: 31009-31015. Article MEDLINE

Colussi C, Parlanti E, Degan P, Aquilina G, Barnes D, Macpherson P, Karran P, Crescenzi M, Dogliotti E, Bignami M. (2002). Curr. Biol., 12: 912-918.

Cox MM, Goodman MF, Kreuzer KN, Sherratt DJ, Sandler SJ, Marians KJ. (2000). Nature, 404: 37-41. Article MEDLINE

Cunniffe S, O'Neill P. (1999). Radiat. Res., 152: 421-427.

Cunningham RP, Weiss B. (1985). Proc. Natl. Acad. Sci. USA, 82: 474-478.

Cunningham RP, Saporito S, Spitzer SG, Weiss B. (1986). J. Bacteriol., 168: 1120-1127. MEDLINE

Czeczot H, Tudek B, Lambert B, Laval J, Boiteux S. (1991). J. Bacteriol., 173: 3419-3424.

Dantzer F, Luna L, Bjoras M, Seeberg E. (2002). Nucleic Acids Res., 30: 2349-2357.

David SS, Williams SD. (1998). Chem. Rev., 98: 1221-1262. Article MEDLINE

David-Cordonnier MH, Cuniffe SM, Hickson ID, O'Neill P. (2002). Biochemistry, 41: 634-642.

David-Cordonnier MH, Laval J, O'Neill P. (2000). J. Biol. Chem., 275: 11865-11873.

David-Cordonnier MH, Boiteux S, O'Neill P. (2001a). Nucleic Acids Res., 29: 1107-1113.

David-Cordonnier MH, Boiteux S, O'Neill P. (2001b). Biochemistry, 40: 11811-11818.

Demple B, Linn S. (1980). Nature, 287: 203-208.

Demple B, Halbrook JH, Linn S. (1983). J. Bacteriol., 153: 1079-1082.

Demple B, Herman T, Chen DS. (1991). Proc. Natl. Acad. Sci. USA, 88: 11450. MEDLINE

Demple B. (1991). Annu. Rev. Genet., 25: 315-337.

Demple B, Johnson A, Fung D. (1986). Proc. Natl. Acad. Sci. USA, 83: 7731-7735.

Denko NC, Giaccia AJ, Stringer JR, Stambrook PJ. (1994). Proc. Natl. Acad. Sci. USA, 91: 5124-5128. MEDLINE

Deutsch WA, Spiering AL. (1982). J. Biol. Chem., 257: 3366-3368.

Devasagayam TP, Steenken S, Obendorf MS, Schulz WA, Sies H. (1991). Biochemistry, 30: 6283-6289.

D'Ham C, Romieu A, Jacquinod M, Gasparutto D, Cadet J. (1999). Biochemistry, 38: 3335-3344.

Dherin C, Dizdaroglu M, Doerflinger H, Boiteux S, Radicella JP. (2000). Nucleic Acids Res., 28: 4583-4592.

Dherin C, Radicella JP, Dizdaroglu M, Boiteux S. (1999). Nucleic Acids Res., 27: 4001-4007. Article MEDLINE

Dianov GL, Prasad R, Wilson SH, Bohr VA. (1999). J. Biol. Chem., 274: 13741-13743. Article MEDLINE

Dianova II, Bohr VA, Dianov GL. (2001). Biochemistry, 40: 12639-12644.

Dirksen ML, Blakely WF, Holwitt E, Dizdaroglu M. (1988). Int. J. Radiat. Biol., 54: 195-204.

Dizdaroglu M, Burgess SM, Jaruga P, Hazra TK, Rodriguez H, Lloyd RS. (2001). Biochemistry, 40: 12150-12156.

Dizdaroglu M, Karahalil B, Senturker S, Buckley TJ, Roldan-Arjona T. (1999). Biochemistry, 38: 243-246.

Dizdaroglu M, Laval J, Boiteux S. (1993). Biochemistry, 32: 12105-12111. MEDLINE

Dizdaroglu M, Schulte-Frohlinde D, von Sonntag C. (1977). Int. J. Radiat. Biol., 32: 481-483.

Dodson ML, Michaels ML, Lloyd RS. (1994). J. Biol. Chem., 269: 32709-32712. MEDLINE

Doherty AJ, Serpell LC, Ponting CP. (1996). Nucleic Acids Res., 24: 2488-2497. Article MEDLINE

Dotto GP. (2000). Biochim. Biophys. Acta. Reviews on Cancer, 1471: M43-M56.

Dugle D, Gillespie C, Chapman JD. (1976). Proc. Natl. Acad. Sci. USA, 73: 809-812.

Duwat P, de Oliveira R, Erhlich SD, Boiteux S. (1995). Microbiology, 141: 411-417.

Eide L, Luna L, Gustad EC, Henderson PT, Essigmann JM, Demple B, Seeberg E. (2001). Biochemistry, 40: 6653-6659.

Eide L, Bjoras M, Pirovano M, Alseth I, Berdal KG, Seeberg E. (1996). Proc. Natl. Acad. Sci. USA, 93: 10735-10740.

Elder RH, Jansen JG, Weeks RJ, Willington MA, Deans B, Watson AJ, Mynett KJ, Bailey JA, Cooper DP, Rafferty JA, Heeran MC, Wijhoven SWP, Van Zeeland A, Margison G. (1998). Mol. Cell Biol., 18: 5828-5837.

El-Ghissassi FE, Barbin A, Nair J, Bartsch H. (1995). Chem. Res. Toxicol., 8: 278-283.

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

Fink SP, Reddy RLJ. (1997). Proc. Natl. Acad. Sci. USA, 94: 8652-8657.

Frenkel K, Karkoszka J, Glassman T, Dubin N, Toniolo P, Taioli E, Mooney LA, Kato I. (1998). Cancer Epidemiol. Biomark. Prev., 7: 49-57.

Friedberg EC, Wagner R, Radman M. (2002). Science, 296: 1627-1630. Article MEDLINE

Fortini P, Parlanti E, Sidorkina OM, Laval J, Dogliotti E. (1999). J. Biol. Chem., 274: 15230-15236. MEDLINE

Fromme JC, Verdine GL. (2002). Nat. Struct. Biol., 9: 544-552. Article

Frosina G, Fortini P, Rossi O, Carrozzino F, Raspaglio G, Cox LS, Lane DP, Abbondandolo A, Dogliotti E. (1996). J. Biol. Chem., 271: 9573-9578. MEDLINE

Fujii Y, Shimokawa H, Sekiguchi M, Nakabeppu Y. (1999). J. Biol. Chem., 274: 38251-38259.

Fujikawa K, Kamiya H, Yakushiji H, Fujii Y, Nakabeppu Y, Kasai H. (1999). J. Biol. Chem., 274: 18201-18205.

Fujikawa K, Kamiya H, Yakushiji H, Nakabeppu Y, Kasai H. (2001). Nucleic Acids Res., 29: 449-454.

Furuichi M, Yoshida MC, Oda H, Tajiri T, Nakabeppu Y, Tsuzuki T, Sekiguchi M. (1994). Genomics, 24: 485-490.

Gallinari P, Jiricny J. (1996). Nature, 383: 735-738. MEDLINE

Garcia-Ortiz MV, Ariza RR, Roldan-Arjona T. (2001). Plant Mol. Biol., 47: 795-804.

Gellon L, Werner M, Boiteux S. (2002). J. Biol. Chem. epub ahead of print.

Georgakilas AG, Bennett PV, Sutherland BM. (2002). Nucleic Acids Res., 30: 2800-2808.

Gilboa R, Zharkov DO, Golan G, Fernandes AS, Gerchman SE, Matz E, Kycia JH, Grollman AP, Shoham G. (2002). J. Biol. Chem., 277: 19811-19816.

Giloni L, Takeshita M, Johnson F, Iden C, Grollman AP. (1981). J. Biol. Chem., 256: 8606-8615.

Girard PM, Boiteux S. (1997). Biochemie, 79: 559-566.

Girard PM, Guibourt N, Boiteux S. (1997). Nucleic Acids Res., 25: 3204-3211.

Gong Z, Morales-Ruiz T, Ariza RR, Roldan-Arjona T, David L, Zhu JK. (2002). Cell, 111: in press.

Goodhead DT. (1994). Int. J. Radiat. Biol., 65: 7-17. MEDLINE

Graves RJ, Felzenszwabb J, Laval J, O'Connor TR. (1992). J. Biol. Chem., 267: 14429-14435. MEDLINE

Greenberg RB, Alberti M, Hearst JE, Chua MA, Saffran WA. (2001). J. Biol. Chem., 276: 31551-31560.

Gu Y, Parker A, Wilson TM, Bai H, Chang DY, Lu AL. (2002). J. Biol. Chem., 277: 11135-11142.

Guan Y, Manuel RC, Arvai AS, Parikh SS, Mol CD, Miller JH, Lloyd RS, Tainer JA. (1998). Nat. Struct. Biol., 5: 1058-1064. Article MEDLINE

Habraken Y, Carter CA, Sekiguchi M, Ludlum DB. (1991). Carcinogenesis, 12: 1971-1973.

Hadi MZ, Wilson DM. (2000). Environ. Mol. Mutagen., 36: 312-324.

Halliwell B. (1982). Cell Biol. Intl. Rep., 6: 529-542.

Hanawalt PC. (2000). Nature, 405: 415-416. Article MEDLINE

Hang B, Downing G, Gulaiev AB, Singer B. (2002). Biochemistry, 41: 2158-2165.

Hang B, Medina M, Fraenkel-contrat H, Singer B. (1998). Proc. Natl. Acad. Sci. USA, 95: 13561-13566.

Hang B, Singer B, Margison GP, Elder RH. (1997). Proc. Natl. Acad. Sci. USA, 94: 12869-12874.

Haracska L, Yu SL, Johnson RE, Prakash L, Prakash S. (2000). Nat. Genet., 25: 458-461. Article MEDLINE

Hardeland U, Steinacher R, Jiricny J, Schär P. (2002). EMBO J., 21: 1456-1464.

Harrison L, Hatahet Z, Purmal AA, Wallace SS. (1998). Nucleic Acids Res., 26: 932-941.

Hatahet Z, Kow YW, Purmal AA, Cunningham RP, Wallace SS. (1994). J. Biol. Chem., 269: 18814-18820. MEDLINE

Haushalter KA, Stukenberg PT, Kirschner MW, Verdine GL. (1999). Curr. Biol., 9: 174-185. Article MEDLINE

Hayakawa H, Hofer A, Thelander L, Kitajima S, Cai Y, Oshiro S, Yakushiji H, Nakabeppu Y, Kuwano M, Sekiguchi M. (1999). Biochemistry, 38: 3610-3614.

Hayashi H, Tominaga Y, Hirano S, McKenna AE, Nakabeppu Y, Matsumoto Y. (2002). Curr. Biol., 12: 335-339.

Hazra TK, Hill JW, Izumi T, Mitra S. (2001). Proc. Nucleic Acid Res. Mol. Biol., 68: 193-205.

Hazra TK, Izumi T, Boldogh I, Imhoff B, Kow YH, Jaruga P, Dizdaroglu M, Mitra S. (2002). Proc. Natl. Acad. Sci. USA, 99: 3523-3528.

Hazra TK, Izumi T, Venkataraman R, Kow YW, Dizdaroglu M, Mitra S. (2000). J. Biol. Chem., 275: 27762-27767.

Henner WD, Rodriguez LO, Hecht SM, Haseltine WA. (1983). J. Biol. Chem., 258: 711-713. MEDLINE

Hilbert TP, Chaung W, Boorstein RJ, Cunningham RP, Teebor GW. (1997). J. Biol Chem., 272: 6733-6740. MEDLINE

Hoeijmakers JHJ. (2001). Nature, 411: 366-374. Article MEDLINE

Holley WR, Chatterjee A. (1996). Radiat. Res., 145: 188-199.

Hollis T, Lau A, Ellenberger T. (2000). Mutat Res., 460: 201-210. Article MEDLINE

Hosfield DJ, Guan Y, Haas BJ, Cunningham RP, Tainer JA. (1999). Cell, 98: 397-408. MEDLINE

Ikeda S, Biswas T, Rabindra R, Izumi T, Boldogh I, Kurosky A, Sarker AH, Seki S, Mitra S. (1998). J. Biol. Chem., 273: 21585-21593. MEDLINE

Imlay JM, Linn S. (1988). Science, 240: 1302-1308. MEDLINE

Ischenko AA, Saparbaev MK. (2002). Nature, 415: 183-187. Article

Ishikawa T, Ide F, Qin X, Zhang S, Takahashi Y, Sekiguchi M, Tanaka K, Nakatsuru Y. (2001). Mutat. Res., 477: 41-49.

Janssen K, Schlink K, Gotte W, Hippler B, Kaina B, Oesch F. (2001). Mutat Res., 486: 207-216.

Jayaraman L, Murthy KG, Zhu C, Curran T, Xanthoudakis S, Prives S. (1997). Genes Dev., 11: 558-570. MEDLINE

Jelinsky SA, Estep P, Church GM, Samson LD. (2000). Mol. Cell Biol., 20: 8157-8167. Article MEDLINE

Jiang D, Hatahet Z, Blaisdell JO, Melamede RJ, Wallace S. (1997). J. Bacteriol., 179: 3773-3782.

Johnson AW, Demple B. (1988a). J. Biol. Chem., 263: 18017-18022.

Johnson AW, Demple B. (1988b). J. Biol. Chem., 263: 18009-18016.

Johnson DE, Torres-Amos CA, Izumi T, Mitra S, Prakash S, Prakash L. (1998). Genes Dev., 12: 3137-3143.

Jurado J, Saparbaev M, Matray MJ, Greenberg MM, Laval J. (1998). Biochemistry, 37: 7757-7763.

Kakuma T, Nishida JI, Tsuzuki T, Sekiguchi M. (1995). J. Biol. Chem., 270: 25942-25948.

Kamiya H, Kasai H. (1995). Nucleic Acids Symp. Ser., 34: 233-234.

Kang D, Nishida JI, Iyama A, Nakabeppu Y, Furuichi M, Furijawa T, Sekiguchi M, Takeshige K. (1995). J. Biol. Chem., 270: 14659-14665.

Karahalil B, Girard PM, Boiteux S, Dizdaroglu M. (1998). Nucleic Acids Res., 26: 1228-1232.

Kasai H, Lida A, Yamaizumi Z, Nishimura S, Tanooka H. (1990). Mutat. Res., 243: 249-253.

Kim J, Linn S. (1988). Nucleic Acids Res., 16: 1135-1141.

Kim JE, You HJ, Choi JY, Doetsch PW, Kim JS, Chung MY. (2001). Biochem. Biophys. Res. Com., 285: 1186-1191.

Klungland A, Lindahl T. (1997). EMBO J., 16: 3341-3348. Article MEDLINE

Klungland A, Höss M, Gunz D, Constantinou A, Clarkson SG, Doetsch PW, Bolton PH, Wood RD, Lindahl T. (1999b). Mol. Cell, 3: 33-42. MEDLINE

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

Koukourakis MI, Giatromanolaki A, Kakolyris S, Sivridis E, Georgoulias V, Funtzilas G, Hickson ID, Gatter KC, Harris AL. (2001). Int. J. Radiat. Oncol. Biol. Phys., 50: 27-36.

Kow YW, Wallace SS. (1987). Biochemistry, 26: 8200-8206.

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

Kubota Y, Nash RA, Klungland A, Schär P, Barnes DE, Lindahl T. (1996). EMBO J., 15: 6662-6670. MEDLINE

Kuninger DT, Izumi T, Papaconstantinou J, Mitra S. (2002). Nucleic Acids Res., 30: 823-829.

Kunz BA, Henson ES, Roche H, Ramotar D, Nunoshiba T, Demple B. (1994). Proc. Natl. Acad. Sci. USA, 91: 8165-8169.

Kuo CF, McRee DE, Fisher CL, O'Handley SF, Cunningham RP, Tainer JA. (1992). Science, 268: 434-440.

Kuraoka I, Bender C, Romieu A, Wood RD, Lindahl T. (2000). Proc. Natl. Acad. Sci. USA, 97: 3832-3837.

Labahn J, Scharer O, Long A, Ezaz-Nikpay K, Verdine GL, Ellenberger TE. (1996). Cell, 86: 321-329. MEDLINE

Lau AY, Schrärer OD, Samson L, Verdine GL, Ellenberger T. (1998). Cell, 95: 249-258. MEDLINE

Lau AY, Wyatt MD, Glassner BJ, Samson LD, Ellenberger T. (2000). Proc. Natl. Acad. Sci. USA, 97: 13573-13578.

Laval F. (1994). Nucleic Acids Res., 22: 4943-4946.

Laval J, Jurado J, Saparbaev M, Sidorkina O. (1998). Mutat. Res., 402: 93-102. MEDLINE

Laval J. (1977). Nature, 269: 829-832. MEDLINE

Lavrukhin OV, Lloyd RS. (2000). Biochemistry, 39: 15266-15271.

Le Marchand L, Donlon T, Lum-Jones A, Seifried A, Wilkens LR. (2002). Cancer Epidemiol. Biomark. Prev., 11: 409-412.

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

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

Le Page F, Randrianarison V, Marot D, Cabannes J, Perricaudet M, Feunteun J, Sarasin A. (2000c). Cancer Res., 60: 5548-5552. MEDLINE

Levin JD, Johnson AW, Demple B. (1988). J. Biol. Chem., 263: 8066-8071.

Levine RL, Miller H, Grollman A, Ohashi E, Ohmori H, Masutani C, Hanaoka F, Moriya M. (2001). J. Biol. Chem., 276: 18717-18721.

Lim DS, Hasty P. (1996). Mol. Cell Biol., 16: 7133-7143. MEDLINE

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

Lindahl T, Sedgwick B, Sekiguchi M, Nakabeppu Y. (1988). Annu. Rev. Biochem., 57: 133-157. Article MEDLINE

Lindahl T. (1974). Proc. Natl. Acad. Sci. USA, 71: 3649-3653. MEDLINE

Lindahl T. (1976). Nature, 259: 64-66. MEDLINE

Lindahl T. (1993). Nature, 362: 709-715. MEDLINE

Lindahl T. (2001). Prog. Nucleic Acid Res. Mol. Biol., 68: 27-30.

Ljungquist S. (1977). J. Biol. Chem., 252: 2808-2814.

Lu Al, Fawcett WP. (1998). J. Biol. Chem., 273: 25098-25105.

Luna L, Bjoras M, Hoff E, Rognes T, Seeberg E. (2000). Mutat. Res., 460: 95-104.

Maki H, Sekiguchi M. (1992). Nature, 355: 273-275.

Marenstein DR, Ocampo MT, Chan MK, Altalirano A, Basu AK, Borrstein RJ, Cunningham RP, Teebor GW. (2001). J. Biol. Chem., 276: 21242-21249.

Marnett LJ. (1994). DNA adducts of alpha, beta-unsaturated aldehydes and dicarbonyl compounds Hemminki K, Dipple A, Shuker DEG, Kadlubar FF, Segerbäck D and Bartsch H (ed) Lyon, France: LARC Scientific, pp. 151-163.

Marnett LJ. (1999). Mutat Res., 424: 83-95.

Marnett LJ. (2001). Tr. Biochem. Sci., 17: 214-221.

Matsumoto Y, Kim K, Bogenhagen DF. (1994). Mol. Cell Biol., 14: 6187-6197.

Matsumoto Y, Kim K, Hurwitz J, Gary R, Levin DS, Tomkinson AE, Park MS. (1999). J. Biol. Chem., 274: 33703-33708. MEDLINE

Matsumoto Y, Zhang QM, Takao M, Yasui A, Yonei S. (2001). Nucleic Acids Res., 29: 1975-1981.

Mazumder A, Gerlt JA, Absalon MJ, Stubbe J, Cunningham RP, Withka JM, Bolton PH. (1991). Biochemistry, 30: 1116-1126.

McCord JM. (1987). Fed. Proc., 46: 2402-2406. MEDLINE

McGoldrick JP, Yeh YC, Solomon M, Essigmann JM, Lu AL. (1995). Mol. Cell Biol., 15: 989-996. MEDLINE

McKenzie JA, Strauss PR. (2001). Biochemistry, 44: 13254-13261.

Meira LB, Devaraj S, Kisby GE, Burns DK, Daniel RL, Hammer RE, Grundy S, Jialal I, Frieberg E. (2001). Cancer Res., 61: 5552-5557.

Melamede RJ, Hatahet Z, Kow YW, Ide H, Wallace SS. (1994). Biochemistry, 33: 1255-1264.

Miao F, Bouziane M, Dammann R, Masutani C, Hanaoka F, O'Connor TR. (2000). J. Biol. Chem., 275: 28433-28438. Article MEDLINE

Michaels ML, Cruz C, Grollman AP, Miller JH. (1992). Proc. Natl. Acad. Sci. USA, 89: 7022-7025.

Michaels ML, Pham L, Ngheim Y, Cruz C, Miller JH. (1990). Nucleic Acids Res., 18: 3841-3845. MEDLINE

Milligan J, Aguilera JA, Nguyen TT, Paglinawan RA, Ward JF. (2000). Int. J. Radiat. Biol., 76: 1475-1483.

Missero C, Pirro MT, Simeone S, Pischetola M, Di Lauro R. (2001). J. Biol. Chem., 276: 33569-33575.

Mo JY, Maki H, Sekiguchi M. (1992). Proc. Natl. Acad. Sci. USA, 89: 11021-11025.

Mol CD, Kuo CF, Thayer MM, Cunningham RP, Tainer JA. (1995). Nature, 374: 381-386. MEDLINE

Nakabeppu Y. (2001). Proc. Nucleic Acids Res. Mol. Biol., 68: 75-94.

Nash HM, Bruner SD, Schärer OD, Kawate T, Addona TA, Spooner E, Lane WS, Verdine GL. (1996). Curr. Biol., 6: 968-980. MEDLINE

Nash HM, Lu R, Lane WS, Verdine GL. (1997). Chem. Biol., 9: 693-702.

Neddermann P, Jiricny J. (1993). J. Biol. Chem., 268: 21218-21224. MEDLINE

Neddermann P, Jiricny J. (1994). Proc. Natl. Acad. Sci. USA, 91: 1642-1646.

Neddermann P, Gallinari P, Lettieri T, Schmid D, Truong O, Hsuan JJ, Wiebauer K, Jiricny J. (1996). J. Biol. Chem., 271: 12767-12774. Article MEDLINE

Ni TT, Marsischky GT, Kolodner RD. (1999). Mol. Cell, 4: 439-444. MEDLINE

Nielsen H, Haushalter KA, Robins P, Barnes DE, Verdine GL, Lindahl T. (2001). EMBO J., 20: 4278-4286. Article MEDLINE

Nielsen H, Rosewell I, Robins P, Skjelbred CF, Andersen S, Slupphaug G, Daly G, Krokan HE, Lindahl T, Barnes DE. (2000). Mol. Cell, 5: 1059-1065. MEDLINE

Nishioka K, Ohtsubo T, Oda H, Fujirawa T, Kang D, Sugimachi K, Nakabeppu Y. (1999). Mol. Biol. Cell, 10: 1637-1652. MEDLINE

Nunoshiba T, Hidalgo E, Li Z, Demple B. (1993). J. Bacteriol., 175: 7492-7494. MEDLINE

O'Connor TR, Laval J. (1989). Proc. Natl. Acad. Sci. USA, 86: 5222-5226.

O'Connor TR, Graves RJ, de Murcia G, Castaing B, Laval J. (1993). J. Biol. Chem., 268: 9063-9070. MEDLINE

O'Connor TR, Laval J. (1991). Biochem. Biophys. Res. Com., 176: 1170-1177.

O'Connor TR. (1993). Nucleic Acids Res., 21: 5561-5569.

Oda H, Nakabeppu Y, Furuichi M, Sekibuchi M. (1997). J. Biol. Chem., 272: 17843-17850.

Oda H, Taketomi A, Murayama R, Otih R, Nishioka K, Yakushiji H, Suzuki T, Sekiguchi M, Nakabeppu Y. (1999). Nucleic Acids Res., 27: 4335-4343.

Ohtsubo T, Nishioka K, Imaiso Y, Iwai S, Shimokawa H, Oda H, Fujiwara T, Nakabeppu Y. (2000). Nucleic Acids Res., 28: 1355-1364.

Okochi E, Ichimura S, Sugimura T, Ushijima T. (2002). Jpn. J. Cancer Res., 93: 501-506.

Olive PL. (1998). Radiat Res., 150: (Suppl) S42-S51.

Osterod M, Hollenbach S, Hengstler JG, Barnes D, Lindahl T, Epe B. (2001). Carcinogenesis, 22: 1459-1463.

Otterlei M, Kavli B, Standal R, Skjelbred C, Bharati S, Krokan HE. (2000). EMBO J., 19: 5542-5551.

Parker A, Gu Y, Mahoney W, Lee SH, Singh KK, Lu AL. (2001). J. Biol. Chem., 276: 5547-5555. MEDLINE

Parlanti E, Fortini P, Macpherson P, Laval J, Dogliotti E. (2002). Oncogene, 21: 5204-5212. Article

Pascucci B, Stucki M, Jonsson ZO, Dogliotti E, Hubscher U. (1999). J. Biol. Chem., 274: 33696-33702.

Pascucci P, Maga G, Hübscher U, Bjoras M, Seeberg E, Hickson ID, Villani G, Giordano C, Cellai L, Dogliotti E. (2002). Nucleic Acids Res., 30: 2124-2130.

Pendlebury A, Frayling IM, Santibanez Koref MF, Margison GP, Rafferty JA. (1994). Carcinogenesis, 15: 2957-2960.

Pope MA, Porello SL, David SS. (2002). J. Biol. Chem., 277: 22605-22615.

Popoff SC, Spira AI, Johnson AW, Demple B. (1990). Proc. Natl. Acad. Sci. USA, 87: 4193-4197. MEDLINE

Prasad R, Dianov GL, Bohr VA, Wilson SH. (2000). J. Biol. Chem., 275: 4460-4466. Article MEDLINE

Privezentzev CV, Saparbaev MK, Sambandam A, Greenberg MM, Laval J. (2000). Biochemistry, 39: 14263-14268.

Privezentzev CV, Saparbaev M, Laval J. (2001). Mutat. Res., 480-481: 277-284.

Rabow LE, Kow YW. (1997). Biochemistry, 36: 5084-5096.

Radman M. (1976). J. Biol. Chem., 251: 1438-1445.

Ramotar D, Popoff SC, Gralla EB, Demple B. (1991). Mol. Cell Biol., 11: 4537-4544. MEDLINE

Rasouli-Nia A, Chaudhry MA, Weinfeld M. (2001). Action of E. coli endonuclease IV with bistranded DNA lesions. In MICROS 2001 - 13th Symposium on Microdosimetry, Stresa, Italy.

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

Reiter RJ. (1995). FASEB J., 9: 526-533.

Robson CN, Hickson ID. (1991). Nucleic Acids Res., 19: 5519-5523. MEDLINE

Rogers SG, Weiss B. (1980). Gene, 11: 187-195. MEDLINE

Roldan-Arjona T, Garcia-Ortiz MV, Ruiz-Rubio M, Ariza RR. (2000). Plant. Mol. Biol., 44: 43-52.

Roth RB, Samson LD. (2002). Cancer Res., 62: 656-660.

Rothwell DG, Barzilay G, Gorman M, Morera S, Freemont P, Hickson ID. (1997). Oncol. Res., 9: 275-280.

Roy R, Biswas T, Hazra TK, Roy G, Grabowski DT, Izumi T, Srinivasan G, Mitra S. (1998). Biochemistry, 37: 580-589. MEDLINE

Roy R, Kennel SJ, Mitra S. (1996). Carcinogenesis, 17: 2177-2182.

Samson L, Derfler B, Boosalis M, Call K. (1991). Proc. Natl. Acad. Sci. USA, 88: 9127-9131.

Sandigursky M, Franklin WA. (1999). Curr. Biol., 9: 531-534. Article MEDLINE

Sandigursky M, Franklin WA. (2000). J. Biol. Chem., 275: 19146-19149. Article MEDLINE

Saparbaev M, Laval J. (1994). Proc. Natl. Acad. Sci. USA, 91: 5873-5877.

Saparbaev M, Laval J. (1998). Proc. Natl. Acad. Sci. USA, 95: 8508-8513. Article MEDLINE

Saparbaev M, Kleibl K, Laval J. (1995). Nucleic Acids Res., 23: 3750-3755.

Saparbaev M, Langouet S, Privezentzev CV, Guengerich FP, Cai H, Elder RH, Laval J. (2002b). J. Biol. Chem., 277: 26987-26993.

Saparbaev M, Sidorkina OM, Jurado J, Privezentzev CV, Greenberg MM, Laval J. (2002a). Environ. Mol. Mutagen., 39: 10-17. Article MEDLINE

Sard L, Tornielli S, Gallinari P, Minoletti F, Jiricny J, Lettieri T, Pierotti MA, Sozzi G, Radice P. (1997). Genomics, 44: 222-226.

Sartori AA, Fitz-Gibbon S, Ynag H, Miller JH, Jiricny J. (2002). EMBO J., 21: 3182-3191.

Scharer OD, Jiricny J. (2001). Bioessays, 23: 270-281. Article MEDLINE

Scharer OD, Deng L, Verdine GL. (1997). Curr. Opin. Chem. Biol., 1: 526-531.

Schuchmann MN, Steenken S, Wroblewski J, von Sonntag C. (1984). Int. J. Radiat. Biol., 46: 225-232.

Seeberg E, Eide L, Bjoras M. (1995). Trends Biochem. Sci., 20: 391-397. Article MEDLINE

Sentürker S, Auffret van der Kemp P, You HJ, Doetsch PW, Dizdaroglu M, Boiteux S. (1998). Nucleic Acids Res., 26: 5270-5276.

Sentürker S, Bauche C, Laval J, Dizdaroglu M. (1999). Biochemistry, 38: 9435-9439.

Serre L, Pereira De Jesus K, Boiteux S, Zelwer C, Castaing B. (2002). EMBO J., 21: 2854-2865.

Shida T, Noda M, Sekiguchi J. (1996). Nucleic Acids Res., 24: 4572-4576.

Sidorkina OM, Laval J. (1998). Nucleic Acids Res., 26: 5351-5357. Article MEDLINE

Sidorkina OM, Laval J. (2000). J. Biol. Chem., 275: 9924-9929. Article MEDLINE

Singer B, Antoccia A, Basu AK, Dosanjh MK, Fraenkel-Conrat H, Gallagher PE, Kusmierek JT, Qiu ZH, Rydberg B. (1992). Proc. Natl. Acad. Sci. USA, 89: 9386-9390.

Slupska MM, Baikalov C, Solomon M, Essigmann J, Lu AL. (1996). J. Bacteriol., 178: 3885-3892. MEDLINE

Slupska MM, Luther WM, Chinag JH, Yang H, Miller JH. (1999). J. Bacteriol., 181: 6210-6213. MEDLINE

Sobol RW, Horton JK, Kühn R, Gu H, Singhal RK, Prasad R, Rajewsky K, Wilson SH. (1996). Nature, 379: 183-186. MEDLINE

Sonoda E, Sasaki MS, Buerstedde JM, Bezzubova O, Shinohara A, Ogawa H, Takata M, Yamaguchi-Iwai Y, Takeda S. (1998). EMBO J., 17: 598-608. Article MEDLINE

Souza-Pinto NC, Eide L, Hogue BA, Stevsner T, Seeberg E, Klunglang A, Bohr VA. (2001). Cancer Res., 61: 5378-5381.

Strauss BS. (1991). Bioessays, 13: 79-84. MEDLINE

Sugahara M, Mikawa T, Kumasaka T, Yamamoto M, Kato R, Fukuyama K, Inoue Y, Kuramitsu S. (2000). EMBO J., 19: 3857-3869. Article MEDLINE

Sun B, Latham KA, Dodson ML, Lloyd RS. (1995). J. Biol. Chem., 270: 19501-19508. Article MEDLINE

Sung JS, Mosbaugh DW. (2000). Biochemistry, 39: 10224-10235.

Sutherland BM, Bennett PV, Sidorkina O, Laval J. (2000a). Biochemistry, 39: 8026-8031.

Sutherland BM, Bennett PV, Sidorkina O, Laval J. (2000b). Proc. Natl. Acad. Sci. USA, 97: 103-108.

Sutherland BM, Bennett PV, Saparbaev M, Sutherland JC, Laval J. (2001a). Radiat. Prot. Dosimetry, 97: 33-38.

Sutherland BM, Bennett PV, Weinert E, Sidorkina O, Laval J. (2001b). Environ. Mol. Mutagen., 38: 159-165.

Sutton MD, Smith BT, Godoy VG, Walker GC. (2000). Annu. Rev. Genet., 34: 479-497. Article MEDLINE

Taddei F, Hayakawa H, Bouton M, Cirinesi A, Matic I, Sekiguchi M, Radman M. (1997). Science, 278: 78-79.

Tajiri T, Maki H, Sekiguchi M. (1995). Mutat. Res., 336: 257-267. Article MEDLINE

Takama F, Kanuma T, Wang D, Nishida JI, Nakabeppu Y, Wake N, Mizunuma H. (2000). Int. J. Oncol., 17: 467-471.

Takao M, Aburatani H, Kobayashi K, Yasui A. (1998). Nucleic Acids Res., 26: 2917-2922. Article MEDLINE

Takao M, Kanno SI, Shiromoto T, Hasegawa R, Ide H, Ikeda S, Sarker AH, Seki S, Xing JZ, Le XC, Weinfeld M, Kobayashi K, Miyazaki JI, Muijtjens M, Heijmakers JHJ, van der Horst G, Yasui A. (2002). EMBO J., 21: 3486-3493.

Takezaki T, Gao CM, Wu JZ, Li ZY, Wang JD, Liu YT, Xu TL, Tajima K, Sugimura H. (2002). Int. J. Cancer, 99: 624-627.

Tchou J, Grollman AP. (1995). J. Biol. Chem., 270: 11671-11677. Article MEDLINE

Tchou J, Kasai H, Shibutani S, Chung MH, Laval J, Grollman AP, Nishimura S. (1991). Proc. Natl. Acad. Sci. USA, 88: 4690-4694. MEDLINE

Terashima I, Sekiguchi M, Nakabeppu Y. (1998). Mol. Gen. Genet., 259: 577-590.

Thayer MM, Ahern H, Xing D, Cunningham RP, Tainer JA. (1995). EMBO J., 14: 4108-4120. MEDLINE

Thomas D, Scott A, Barbey R, Padula M, Boiteux S. (1997). Mol. Gen. Genet., 254: 171-178. Article MEDLINE

Thomas L, Yang CH, Goldthwait DA. (1982). Biochemistry, 21: 1162-1169. MEDLINE

Tini M, Benecke A, Um SJ, Torchia J, Evans RM, Chambon P. (2002). Mol. Cell., 9: 265-277. MEDLINE

Tom S, Ranalli TA, Podust VN, Bambara RA. (2001). J. Biol. Chem., 276: 48781-48789. MEDLINE

Trus MA, Kensler TW. (1991). Free Radical Biol. Med., 10: 201-209.

Tsai-Wu JJ, Liu HF, Lu AL. (1992). Proc. Natl. Acad. Sci. USA, 89: 8779-8783.

Tsai-Wu JJ, Radicella JP, Lu AL. (1991). J. Bacteriol., 173: 1902-1910.

Tsaneva IR, Weiss B. (1990). J. Bacteriol., 172: 4197-4205.

Tsuchimoto D, Sakai Y, Sakumi K, Nishioka K, Sasaki M, Fujiwara T, Nakabeppu Y. (2001). Nucleic Acids Res., 29: 2349-2360.

Tsuzuki T, Egashira A, Igarashi H, Iwakuma T, Nakatsuru Y, Tominaga Y, Kawate H, Nakao K, Nakamura K, Ide F, Kura S, Nakabeppu Y, Katsuki M, Ishikawa T, Sekiguchi M. (2001). Proc. Natl. Acad. Sci. USA, 98: 11456-11561.

Um S, Harbers M, Benecke A, Pierrat B, Losson R, Chambon P. (1998). J. Biol. Chem., 273: 20728-20736. Article MEDLINE

Unk I, Haracska L, Jhonson RE, Prakash S, Prakash L. (2000). J. Biol. Chem., 275: 22427-22434.

Unk I, Haracska L, Prakash S, Prakash L. (2001). Mol. Cell Biol., 21: 1656-1661.

Vafa O, Wade M, Kern S, Beeche M, Pandita TK, Hampton G, Wahl GM. (2002). Mol. Cell., 9: 1031-1044. MEDLINE

Vaisman A, Woodgate R. (2001). EMBO J., 20: 6520-6529.

Van der Kemp P, Thomas D, Barbey R, de Oliveira R, Boiteux S. (1996). Proc. Natl. Acad. Sci. USA, 93: 5197-5202. Article MEDLINE

van Gent DC, Hoeijmakers JH, Kanaar R. (2001). Nat. Rev. Genet., 2: 196-206. Article MEDLINE

Verly WG, Paquette Y. (1972). Can. J. Biochem., 50: 217-224.

Vickers MA, Vyas P, Harris PC, Simmons DL, Higgs DR. (1993). Proc. Natl. Acad. Sci. USA, 90: 3437-3441. MEDLINE

Vidal AE, Boiteux S, Hickson ID, Radicella JP. (2001b). EMBO J., 20: 6530-6539. MEDLINE

Vidal AE, Hickson ID, Boiteux S, Radicella JP. (2001a). Nucleic Acids Res., 29: 1285-1292.

von Sonntag C, Schuchmann HP. (1994). Methods Enzymol., 233: 47-56.

Ward JF. (1995). Radiat. Res., 142: 362-368. MEDLINE

Ward JF. (1981). Radiat Res., 86: 185-195.

Waters TR, Gallinari P, Jiricny J, Swann PF. (1999). J. Biol. Chem., 274: 67-74. Article MEDLINE

Weinfeld M, Rasouli-Nia A, Chaudhry MA, Britten RA. (2001). Radiat Res., 156: 584-589.

Weiss B, Grossman L. (1987). Adv. Enzymol. Relat. Areas Mol. Biol., 60: 1-34.

Wilson DM, Thompson LH. (1997). Proc. Natl. Acad. Sci. USA, 94: 12754-12757. Article MEDLINE

Wilson DM, Bennett RA, Marquis JC, Ansari P, Demple B. (1995a). Nucleic Acids Res., 23: 5027-5033.

Wilson DM, Takashita M, Grollman AP, Demple B. (1995b). J. Biol. Chem., 270: 16002-16007. Article MEDLINE

Wink DA, Hanbauer I, Laval F, Cook JA, Krishna MC, Mitchell JB. (1994). Ann. N Y Acad. Sci., 738: 265-278.

Wu C, Nagasaki H, Maruyama K, Nakabeppu Y, Sekiguchi M, Yuasa Y. (1995). Biochem. Biophys. Res. Com., 214: 1239-1245.

Xanthoudakis S, Miao G, Wang F, Pan YC, Curran T. (1992). EMBO J., 11: 3323-3335. MEDLINE

Xanthoudakis S, Smeyne RJ, Wallace JD, Curran T. (1996). Proc. Natl. Acad. Sci. USA, 93: 8919-8923. Article MEDLINE

Yamagata Y, Kato M, Odawara K, Tokuno Y, Nakashima Y, Matsushima N, Yasumura K, Tomita K, Ihara K, Fujii Y, Nakabeppu Y, Sekiguchi M, Fujii S. (1996). Cell, 86: 311-319. MEDLINE

Yang H, Clendenin WM, Wong D, Demple B, Slupska MM, Chiang JH, Miller HH. (2001). Nucleic Acids Res., 29: 743-752. MEDLINE

You HJ, Swanson RL, Harrington C, Corbett AH, Jinks-Robertson S, Sentürker S, Wallace SS, Boiteux S, Dizdaroglu M, Doetsch PW. (1999). Mol. Cell Biol., 38: 11298-11306.

Zharkov DO, Grollman AP. (1998). Biochemistry, 37: 12384-12394. Article MEDLINE

Zharkov DO, Golan G, Gilboa R, Fernandes AS, Gerchman SE, Kycia JH, Rieger RA, Grollman AP, Shoham G. (2002). EMBO J., 21: 789-800. Article MEDLINE

Zharkov DO, Rieger RA, Iden CR, Grossman AP. (1997). J. Biol. Chem., 272: 5335-5341. Article MEDLINE

Zharkov DO, Rosenquist TA, Gerchman SE, Grollman AP. (2000). J. Biol. Chem., 275: 28607-28617.

Figure 1 Base excision and nucleotide incision (NIR) repair pathways for oxidative DNA damage. Beside the BER short-patch pathway outlined here which operates mainly to remove oxidised bases, a long-patch pathway described in the text operates for other lesions

Table 1 Kinetic constants for various DNA glycosylases for the excision of oxidative DNA damage