Lig4 and rad54 are required for repair of DNA double-strand breaks induced by P-element excision in Drosophila

Site-specific double-strand breaks (DSBs) were generated in the white gene located on the X chromosome of Drosophila by excision of the w(hd) P-element. To investigate the role of nonhomologous end joining (NHEJ) and homologous recombination (HR) in the repair of these breaks, the w(hd) P-element was mobilized in flies carrying mutant alleles of either lig4 or rad54. The survival of both lig4- and rad54-deficient males was reduced to 25% in comparison to the wild type, indicating that both NHEJ and HR are involved in the repair P-induced gaps in males. Survival of lig4-deficient females was not affected at all, implying that HR using the homologous chromosome as a template can partially compensate for the impaired NHEJ pathway. In rad54 mutant females survival was reduced to 70% after w(hd) excision. PCR analysis indicated that the undamaged homologous chromosome may compensate for the potential loss of the broken chromosome in rad54 mutant females after excision. Molecular analysis of the repair junctions revealed microhomology (2-8 bp)-dependent DSB repair in most products. In the absence of Lig4, the 8-bp target site duplication is used more frequently for repair. Our data indicate the presence of efficient alternative end-joining mechanisms, which partly depend on the presence of microhomology but do not require Lig4.

Genetic interactions between BLM and DNA ligase IV in human cells

BLM has been implicated in DNA double-strand break (DSB) repair, but its precise role remains obscure. To explore this, we generated BLM(-/-) and BLM(-/-)LIG4(-/-) cells from the human pre-B cell line Nalm-6. BLM(-/-) cells exhibited retarded growth, increased mutation rates, and hypersensitivity to agents that block replication fork progression. Interestingly, these phenotypes were significantly suppressed by deletion of LIG4, suggesting that nonhomologous end-joining (NHEJ) is unfavorable for integrity and survival of cells lacking BLM. We propose that the absence of BLM leads to accumulation of replication-associated, one-ended DSBs, which are deleterious to cells and lead to genomic instability when repaired by NHEJ. In addition, the NHEJ pathway per se was marginally affected by BLM deficiency, as evidenced by x-ray sensitivity and I-SceI-based DSB repair assays. More intriguingly, however, these experiments revealed the presence of an alternative, DNA ligase IV-independent end-joining pathway, which was significantly affected by the loss of BLM. Collectively, our results provide the first evidence for genetic interactions between BLM and NHEJ in human cells.

TheDrosophila melanogasterDNALigase IVGene Plays a Crucial Role in the Repair of Radiation-Induced DNA Double-Strand Breaks and Acts Synergistically WithRad54

DNA Ligase IV has a crucial role in double-strand break (DSB) repair through nonhomologous end joining (NHEJ). Most notably, its inactivation leads to embryonic lethality in mammals. To elucidate the role of DNA Ligase IV (Lig4) in DSB repair in a multicellular lower eukaryote, we generated viable Lig4-deficient Drosophila strains by P-element-mediated mutagenesis. Embryos and larvae of mutant lines are hypersensitive to ionizing radiation but hardly so to methyl methanesulfonate (MMS) or the crosslinking agent cis-diamminedichloroplatinum (cisDDP). To determine the relative contribution of NHEJ and homologous recombination (HR) in Drosophila, Lig4; Rad54 double-mutant flies were generated. Survival studies demonstrated that both HR and NHEJ have a major role in DSB repair. The synergistic increase in sensitivity seen in the double mutant, in comparison with both single mutants, indicates that both pathways partially overlap. However, during the very first hours after fertilization NHEJ has a minor role in DSB repair after exposure to ionizing radiation. Throughout the first stages of embryogenesis of the fly, HR is the predominant pathway in DSB repair. At late stages of development NHEJ also becomes less important. The residual survival of double mutants after irradiation strongly suggests the existence of a third pathway for the repair of DSBs in Drosophila.

Structural and functional adaptations to extreme temperatures in psychrophilic, mesophilic, and thermophilic DNA ligases

Psychrophiles, host of permanently cold habitats, display metabolic fluxes comparable to those exhibited by mesophilic organisms at moderate temperatures. These organisms have evolved by producing, among other peculiarities, cold-active enzymes that have the properties to cope with the reduction of chemical reaction rates induced by low temperatures. The emerging picture suggests that these enzymes display a high catalytic efficiency at low temperatures through an improved flexibility of the structural components involved in the catalytic cycle, whereas other protein regions, if not implicated in catalysis, may be even more rigid than their mesophilic counterparts. In return, the increased flexibility leads to a decreased stability of psychrophilic enzymes. In order to gain further advances in the analysis of the activity/flexibility/stability concept, psychrophilic, mesophilic, and thermophilic DNA ligases have been compared by three-dimensional-modeling studies, as well as regards their activity, surface hydrophobicity, structural permeability, conformational stabilities, and irreversible thermal unfolding. These data show that the cold-adapted DNA ligase is characterized by an increased activity at low and moderate temperatures, an overall destabilization of the molecular edifice, especially at the active site, and a high conformational flexibility. The opposite trend is observed in the mesophilic and thermophilic counterparts, the latter being characterized by a reduced low temperature activity, high stability and reduced flexibility. These results strongly suggest a complex relationship between activity, flexibility and stability. In addition, they also indicate that in cold-adapted enzymes, the driving force for denaturation is a large entropy change.

The viral potassium channel Kcv: structural and functional features

The chlorella virus PBCV-1 was the first virus found to encode a functional potassium channel protein (Kcv). Kcv is small (94 aa) and basically consists of the M1-P-M2 (membrane-pore-membrane) module typical of the pore regions of all known potassium channels. Kcv forms functional channels in three heterologous systems. This brief review discusses the gating, permeability and modulation properties of Kcv and compares them to the properties of bacterial and mammalian K+ channels.

DNA ligase IV suppresses medulloblastoma formation

Substantial neural defects are often present in mice with targeted inactivation of DNA repair factors such as DNA ligase IV (Lig4). Whereas Lig4(-/-) mice undergo widespread neural apoptosis and die during development, p53 deficiency rescues this death. We found that all Lig4(-/-)p53(-/-) mice developed medulloblastoma, but did not develop other tumors of the nervous system. Lig4(-/-)p53(-/-) medulloblastoma occurred as early as 21 days of age, originated in the external granule layer of the developing cerebellum, and was synaptophysin immunoreactive. These data reveal a pronounced susceptibility of the cerebellum to the effects of chronic DNA damage and provide a direct link between genotoxic stress and medulloblastoma formation.

An evaluation of the role of LIG4 in genomic instability and adaptive mutagenesis in Candida albicans

The non-homologous end-joining (NHEJ) pathway of DNA recombination is important for genomic stability in animal cells, since the absence of Ku70, Ku80, Lig4 or Xrcc4 results in non-reciprocal translocation and chromosome fragmentation. The role of LIG4 in the genomic instability of Candida albicans has been analyzed. We have found that both cell transformation and 5'-fluoroorotic acid selection steps used to obtain several lig4 mutants (LIG4/lig4 Ura(+); LIG4/lig4 Ura(-); lig4/lig4 Ura(+); lig4/lig4 Ura(-); and revertant lig4/LIG4 Ura(+)) resulted in significant alterations in chromosome R (ChrR). However, this effect is not specific for LIG4, since disruption of SHE9, a gene unrelated to recombination, also caused alterations in the mobility of ChrR. On the other hand, we could not detect reciprocal or non-reciprocal translocations between non-homologous chromosomes in several lig4 mutants. Furthermore, propagation of these mutants in rich medium did not cause other alterations in the mobility of ChrR. Adaptive mutagenesis of C. albicans, determined by the appearance of L-sorbose-utilizing mutants on L-sorbose plates, was also independent of the presence of Lig4 and occurred by monosomy of Chr5. Accordingly, the NHEJ pathway does not appear to be involved in the adaptive mutagenesis mediated by alterations in chromosome copy number.

DNA ligase I null mouse cells show normal DNA repair activity but altered DNA replication and reduced genome stability

DNA ligase I is the key ligase for DNA replication in mammalian cells and has also been reported to be involved in a number of recombination and repair processes. Our previous finding that Lig1 knockout mouse embryos developed normally to mid-term before succumbing to a specific haematopoietic defect was difficult to reconcile with a report that DNA ligase I is essential for the viability of cultured mammalian cells. To address this issue, we generated a second Lig1 targeted allele and found that the phenotypes of our two Lig1 mutant mouse lines are identical. Widely different levels of Lig1 fusion transcripts were detected from the two targeted alleles, but we could not detect any DNA ligase I protein, and we believe both are effective Lig1 null alleles. Using foetal liver cells to repopulate the haematopoietic system of lethally irradiated adult mice, we demonstrate that the haematopoietic defect in DNA-ligase-I-deficient embryos is a quantitative deficiency relating to reduced proliferation rather than a qualitative block in any haematopoietic lineage. DNA ligase I null fibroblasts from Lig1 mutant embryos showed an accumulation of DNA replication intermediates and increased genome instability. In the absence of a demonstrable deficiency in DNA repair we postulate that, unusually, genome instability may result directly from the DNA replication defect. Lig1null mouse cells performed better in the survival and replication assays than a human LIG1 point mutant, and we suggest that the complete absence of DNA ligase I may make it easier for another ligase to compensate for DNA ligase I deficiency.

Different types of V(D)J recombination and end-joining defects in DNA double-strand break repair mutant mammalian cells

The end-joining pathway of DNA double-strand break (DSB) repair is necessary for proper V(D)J recombination and repair of DSB caused by ionizing radiation. This DNA repair pathway can either use short stretches of (micro)homology near the DNA ends or use no homology at all (direct end-joining). We designed assays to determine the relative efficiencies of these (sub)pathways of DNA end-joining. In one version, a DNA substrate is linearized in such a way that joining on a particular microhomology creates a novel restriction enzyme recognition site. In the other one, the DSB is made by the RAG1 and RAG2 proteins. After PCR amplification of the junctions, the different end-joining modes can be discriminated by restriction enzyme digestion. We show that inactivation of the 'classic' end-joining factors (Ku80, DNA-PK(CS), ligase IV and XRCC4) results in a dramatic increase of microhomology-directed joining of the linear substrate, but very little decrease in overall joining efficiency. V(D)J recombination, on the other hand, is severely impaired, but also shows a dramatic shift towards microhomology use. Interestingly, two interstrand cross-linker-sensitive cell lines showed decreased microhomology-directed end-joining, but without an effect on V(D)J recombination. These results suggest that direct end-joining and microhomology-directed end-joining constitute genetically distinct DSB repair pathways.

Molecular mechanism of PCNA-dependent base excision repair

In higher eukaryotes, base excision repair can proceed by two alternative pathways: a DNA polymerase beta-dependent pathway and a proliferating cell nuclear antigen (PCNA)-dependent pathway. Recently, we have reconstituted the PCNA-dependent AP site repair reaction with six purified human proteins: AP endonuclease, replication factor C (RFC), PCNA, flap endonuclease 1 (FEN1), DNA polymerase delta (pol delta), and DNA ligase I. In this reconstituted system, the number of nucleotides replaced during the repair reaction (patch size) was predominantly two nucleotides. PCNA can directly interact with RFC, pol delta, FEN1 and DNA ligase I. These interactions are partly through a consensus motif, QXX(I/L/M)XX(F/H)(F/Y), found in each of the four proteins. PCNA functions as a molecular adaptor for recruiting these factors to the site of DNA repair. Two DNA-N-glycosylases among those so far cloned from human, UNG2 and MYH, are found to have the same PCNA-binding motif. Major substrates of these enzymes, a uracil opposite an adenine for UNG2 and an adenine opposite an 8-oxoguanine for MYH, are formed during DNA replication. Therefore, UNG2 and MYH may serve for replication-coupled base excision repair through the direct interaction with PCNA in the replication machinery.

Mammalian DNA beta-polymerase in base excision repair of alkylation damage

DNA beta-polymerase (beta-pol) carries out two critical enzymatic reactions in mammalian single-nucleotide base excision repair (BER): DNA synthesis to fill the repair patch and lyase removal of the 5'-deoxyribose phosphate (dRP) group following cleavage of the abasic site by apurinic/apyrimidinic (AP) endonuclease (1). The requirement for beta-pol in single-nucleotide BER is exemplified in mouse fibroblasts with a null mutation in the beta-pol gene. These cells are hypersensitive to monofunctional DNA methylating agents such as methyl methane-sulfonate (MMS) (2). This hypersensitivity is associated with an abundance of chromosomal damage and induction of apoptosis and necrotic cell death (3). We have found that beta-pol null cells are defective in repair of MMS-induced DNA lesions, consistent with a cellular BER deficiency as a causative agent in the observed hypersensitivity. Further, the N-terminal 8-kDa domain of beta-pol, which contains the dRP lyase activity in the wild-type enzyme, is sufficient to reverse the methylating agent hypersensitivity in beta-pol null cells. These results indicate that lyase removal of the dRP group is a pivotal step in BER in vivo. Finally, we examined MMS-induced genomic DNA mutagenesis in two isogenic mouse cell lines designed for study of the role of BER. MMS exposure strongly increases mutant frequency in beta-pol null cells, but not in wild-type cells. With MMS treatment, beta-pol null cells have a higher frequency of all six base-pair substitutions, suggesting that BER plays a role in protecting the cell against methylation-induced mutations.

Base excision repair of mitochondrial DNA damage in mammalian cells

This review of the work from our laboratory describes initial studies in which base excision repair in mtDNA was first seen. It considers the results of experiments in which the substrates for mtDNA repair were identified. The discussion then focuses on studies during which the sequence context for mtDNA damage and repair were explored. Next, it addresses factors that have been identified that influence mtDNA repair. Finally, it summarizes the results of studies that evaluated cell-specific differences in the repair of mtDNA and explored some of the biological consequences of these differences.

Multiple DNA glycosylases for repair of 8-oxoguanine and their potential in vivo functions

8-Oxoguanine (8-oxoG) is a critical mutagenic lesion because of its propensity to mispair with A during DNA replication. All organisms, from bacteria to mammals, express at least two types of 8-oxoguanine-DNA glycosylase (OGG) for repair of 8-oxoG. The major enzyme class (OGG1), first identified in Escherichia coli as MutM (Fpg), and later in yeast and humans, excises 8-oxoG when paired with C, T, and G but rarely with A. In contrast, a distinct and less abundant OGG, OGG2, prefers 8-oxoG when paired with G and A as a substrate, and has been characterized in yeast and human cells. Recently, OGG2 activity was detected in E. coli which was subsequently identified to be Nei (Endo VIII). In view of the ubiquity of OGG2, we have proposed a model named "bipartite antimutagenic processing of 8-oxoguanine" and is an extension of the original "GO model." The GO model explains the presence of OGG1 (MutM) that excises 8-oxoG from nonreplicated DNA. If 8-oxoG mispairs with A during replication, MutY excises A and provides an opportunity for insertion of C opposite 8-oxoG during subsequent repair replication. Our model postulates that whereas OGG1 (MutM) is responsible for global repair of 8-oxoG in the nonreplicating genome, OGG2 (Nei) repairs 8-oxoG in nascent or transcriptionally active DNA. Interestingly, we observed that MutY and MutM reciprocally inhibited each other's catalytic activity but observed no mutual interference between Nei and MutY. This suggests that the recognition sites on the same substrate for Nei and MutY are nonoverlapping. Human OGG1 is distinct from other oxidized base-specific DNA glycosylases because of its extremely low turnover, weak AP lyase activity, and nonproductive affinity for the abasic (AP) site, its first reaction product. OGG1 is activated nearly 5-fold in the presence of AP-endonuclease (APE) as a result of its displacement by the latter. These results support the "handoff" mechanism of BER in which the enzymatic steps are coordinated as a result of displacement of the DNA glycosylase by APE, the next enzyme in the pathway. The physiological significance of multiple OGGs and their in vivo reaction mechanisms remain to be elucidated by further studies.

Completion of base excision repair by mammalian DNA ligases

Three mammalian genes encoding DNA ligases--LIG1, LIG3, and LIG4--have been identified. Genetic, biochemical, and cell biology studies indicate that the products of each of these genes play a unique role in mammalian DNA metabolism. Interestingly, cell lines deficient in either DNA ligase I (46BR.1G1) or DNA ligase III (EM9) are sensitive to simple alkylating agents. One interpretation of these observations is that DNA ligases I and III participate in functionally distinct base excision repair (BER) subpathways. In support of this idea, extracts from both DNA ligase-deficient cell lines are defective in catalyzing BER in vitro and both DNA ligases interact with other BER proteins. DNA ligase I interacts directly with proliferating cell nuclear antigen (PCNA) and DNA polymerase beta (Pol beta), linking this enzyme with both short-patch and long-patch BER. In somatic cells, DNA ligase III alpha forms a stable complex with the DNA repair protein Xrcc1. Although Xrcc1 has no catalytic activity, it also interacts with Pol beta and poly(ADP-ribose) polymerase (PARP), linking DNA ligase III alpha with BER and single-strand break repair, respectively. Biochemical studies suggest that the majority of short-patch base excision repair events are completed by the DNA ligase III alpha/Xrcc1 complex. Although there is compelling evidence for the participation of PARP in the repair of DNA single-strand breaks, the role of PARP in BER has not been established.

The mechanism of switching among multiple BER pathways

To preserve genomic beta DNA from common endogenous and exogenous base and sugar damage, cells are provided with multiple base excision repair (BER) pathways: the DNA polymerase (Pol) beta-dependent single nucleotide BER and the long-patch (2-10 nt) BER that requires PCNA. It is a challenge to identify the factors that govern the mechanism of switching among these pathways. One of these factors is the type of DNA damage induced in DNA. By using different model lesions we have shown that base damages (like hypoxanthine and 1, N6-ethenoadenine) excised by monofunctional DNA glycosylases are repaired via both single-nucleotide and long-patch BER, while lesions repaired by a bifunctional DNA glycosylase (like 7,8-dihydro-8-oxoguanine) are repaired mainly by single-nucleotide BER. The presence of a genuine 5' nucleotide, as in the case of cleavage by a bifunctional DNA glycosylase-beta lyase, would then minimize the strand displacement events. Another key factor in the selection of the BER branch is the relative level of cellular polymerases. While wild-type embryonic mouse fibroblast cell lines repair abasic sites predominantly via single-nucleotide replacement reactions (80% of the repair events), cells homozygous for a deletion in the Pol beta gene repair these lesions exclusively via long-patch BER. Following treatment with methylmethane sulfonate, these mutant cells accumulate DNA single-strand breaks in their genome in keeping with the fact that repair induced by monofunctional alkylating agents goes predominantly via single-nucleotide BER. Since the long-patch BER is strongly stimulated by PCNA, the cellular content of this cell-cycle regulated factor is also extremely effective in driving the repair reaction to either BER branch. These findings raise the interesting possibility that different BER pathways might be acting as a function of the cell cycle stage.

DNA damage recognition and repair pathway coordination revealed by the structural biochemistry of DNA repair enzymes

Cells have evolved distinct mechanisms for both preventing and removing mutagenic and lethal DNA damage. Structural and biochemical characterization of key enzymes that function in DNA repair pathways are illuminating the biological and chemical mechanisms that govern initial lesion detection, recognition, and excision repair of damaged DNA. These results are beginning to reveal a higher level of DNA repair coordination that ensures the faithful repair of damaged DNA. Enzyme-induced DNA distortions allow for the specific recognition of distinct extrahelical lesions, as well as tight binding to cleaved products, which has implications for the ordered transfer of unstable DNA repair intermediates between enzymes during base excision repair.

Yeast base excision repair: interconnections and networks

The removal of oxidative base damage from the genome of Saccharomyces cerevisiae is thought to occur primarily via the base excision repair (BER) pathway in a process initiated by several DNA N-glycosylase/AP lyases. We have found that yeast strains containing simultaneous multiple disruptions of BER genes are not hypersensitive to killing by oxidizing agents, but exhibit a spontaneous hyperrecombinogenic (hyper-rec) and mutator phenotype. The hyper-rec and mutator phenotypes are further enhanced by elimination of the nucleotide excision repair (NER) pathway. Furthermore, elimination of either the lesion bypass (REV3-dependent) or recombination (RAD52-dependent) pathway results in a further, specific enhancement of the hyper-rec or mutator phenotypes, respectively. Sensitivity (cell killing) to oxidizing agents is not observed unless multiple pathways are eliminated simultaneously. These data suggest that the BER, NER, recombination, and lesion bypass pathways have overlapping specificities in the removal of, or tolerance to, exogenous or spontaneous oxidative DNA damage in S. cerevisiae. Our results also suggest a physiological role for the AP lyase activity of certain BER N-glycosylases in vivo.

ATP-dependent DNA ligases

SUMMARY

By catalyzing the joining of breaks in the phosphodiester backbone of duplex DNA, DNA ligases play a vital role in the diverse processes of DNA replication, recombination and repair. Three related classes of ATP-dependent DNA ligase are readily apparent in eukaryotic cells. Enzymes of each class comprise catalytic and non-catalytic domains together with additional domains of varying function. DNA ligase I is required for the ligation of Okazaki fragments during lagging-strand DNA synthesis, as well as for several DNA-repair pathways; these functions are mediated, at least in part, by interactions between DNA ligase I and the sliding-clamp protein PCNA. DNA ligase III, which is unique to vertebrates, functions both in the nucleus and in mitochondria. Two distinct isoforms of this enzyme, differing in their carboxy-terminal sequences, are produced by alternative splicing: DNA ligase IIIalpha has a carboxy-terminal BRCT domain that interacts with the mammalian DNA-repair factor XrccI, but both alpha and beta isoforms have an amino-terminal zinc-finger motif that appears to play a role in the recognition of DNA secondary structures that resemble intermediates in DNA metabolism. DNA ligase IV is required for DNA non-homologous end joining pathways, including recombination of the V(D)J immunoglobulin gene segments in cells of the mammalian immune system. DNA ligase IV forms a tight complex with Xrcc4 through an interaction motif located between a pair of carboxy-terminal BRCT domains in the ligase. Recent structural studies have shed light on the catalytic function of DNA ligases, as well as illuminating protein-protein interactions involving DNA ligases IIIalpha and IV.

Impaired nonhomologous end-joining provokes soft tissue sarcomas harboring chromosomal translocations, amplifications, and deletions

Although nonhomologous end-joining (NHEJ) deficiency has been shown to accelerate lymphoma formation in mice, its role in suppressing tumors in cells that do not undergo V(D)J recombination is unclear. Utilizing a tumor-prone mouse strain (ink4a/arf(-/-)), we examined the impact of haploinsufficiency of a NHEJ component, DNA ligase IV (Lig4), on murine tumorigenesis. We demonstrate that lig4 heterozygosity promotes the development of soft-tissue sarcomas that possess clonal amplifications, deletions, and translocations. That these genomic alterations are relevant in tumorigenesis is supported by the finding of frequent mdm2 amplification, a known oncogene in human sarcoma. Together, these findings support the view that loss of a single lig4 allele results in NHEJ activity being sufficiently reduced to engender chromosomal aberrations that drive non-lymphoid tumorigenesis.

Non-homologous end-joining proteins are required for Agrobacterium T-DNA integration

Agrobacterium tumefaciens causes crown gall disease in dicotyledonous plants by introducing a segment of DNA (T-DNA), derived from its tumour-inducing (Ti) plasmid, into plant cells at infection sites. Besides these natural hosts, Agrobacterium can deliver the T-DNA also to monocotyledonous plants, yeasts and fungi. The T-DNA integrates randomly into one of the chromosomes of the eukaryotic host by an unknown process. Here, we have used the yeast Saccharomyces cerevisiae as a T-DNA recipient to demonstrate that the non-homologous end-joining (NHEJ) proteins Yku70, Rad50, Mre11, Xrs2, Lig4 and Sir4 are required for the integration of T-DNA into the host genome. We discovered a minor pathway for T-DNA integration at the telomeric regions, which is still operational in the absence of Rad50, Mre11 or Xrs2, but not in the absence of Yku70. T-DNA integration at the telomeric regions in the rad50, mre11 and xrs2 mutants was accompanied by gross chromosomal rearrangements.

DNA ligase IV-deficient cells are more resistant to ionizing radiation in the absence of Ku70: Implications for DNA double-strand break repair

Vertebrate cells have evolved two major pathways for repairing DNA double-strand breaks (DSBs), homologous recombination (HR) and nonhomologous DNA end-joining (NHEJ). To investigate the role of DNA ligase IV (Lig4) in DSB repair, we knocked out the Lig4 gene (LIG4) in the DT40 chicken B-lymphocyte cell line. The LIG4(-/-) cells showed a marked sensitivity to X-rays, bleomycin, and VP-16 and were more x-ray-sensitive in G(1) than late S or G(2)/M, suggesting a critical role of Lig4 in DSB repair by NHEJ. In support of this notion, HR was not impaired in LIG4(-/-) cells. LIG4(-/-) cells were more x-ray-sensitive when compared with KU70(-/-) DT40 cells, particularly at high doses. Strikingly, however, the x-ray sensitivity of KU70(-/-)/LIG4(-/-) double-mutant cells was essentially the same as that of KU70(-/-) cells, showing that Lig4 deficiency has no effect in the absence of Ku. These results indicate that Lig4 is exclusively required for the Ku-dependent NHEJ pathway of DSB repair and that other DNA ligases (I and III) do not substitute for this function. Our data may explain the observed severe phenotype of Lig4-deficient mice as compared with Ku-deficient mice.