Identification of a mutated DNA ligase IV gene in the X-ray-hypersensitive mutant SX10 of mouse FM3A cells.

The mouse carcinoma cell line SX10 is a hypersensitive mutant to x-rays and bleomycin. An earlier complementation test suggests that SX10 would belong to x-ray-cross complementing group (XRCC) 4. However, in this study, a human XRCC4 expression vector failed to complement the SX10 phenotype. Consistent with the previous report, SX10 showed the same level of DNA-dependent protein kinase activity as the wild-type SR-1. We isolated and analyzed hybrids between SX10 and human diploid fibroblast cells and found that human chromosome 13 conferred the x-ray resistance to the hybrids, suggesting that a candidate gene would be located on this chromosome. Polymerase chain reaction analysis with these hybrids and x-ray-resistant transformants obtained by introducing human chromosomes into SX10 indicated that the mutant was likely to be defective in DNA ligase IV. Sequence analysis of the DNA ligase IV gene confirmed that a defect in SX10 was attributed to a transition of G to A at nucleotide position 1413 of the gene, leading to an amino acid substitution from Trp at residue 471 to a stop codon. Revertant clones (Rev1-3) derived from SX10 showed a restored x-ray resistance; Rev1 reverted to the original nucleotide G at position 1413, whereas Rev2 and Rev3 to C. Transfection of a mouse DNA ligase IV cDNA vector into SX10 restored the resistance to both x-rays and bleomycin. SX10 showed a reduced frequency of chromosomal integration of transfected DNA, but the revertants restored the frequency found in the wild-type cells. These results suggest a possible involvement of DNA ligase IV in the integration event of foreign DNA as well as a crucial role in DNA double-strand break repair.

The mouse carcinoma cell line SX10 is a hypersensitive mutant to x-rays and bleomycin. An earlier complementation test suggests that SX10 would belong to x-raycross complementing group (XRCC) 4. However, in this study, a human XRCC4 expression vector failed to complement the SX10 phenotype. Consistent with the previous report, SX10 showed the same level of DNA-dependent protein kinase activity as the wild-type SR-1. We isolated and analyzed hybrids between SX10 and human diploid fibroblast cells and found that human chromosome 13 conferred the x-ray resistance to the hybrids, suggesting that a candidate gene would be located on this chromosome. Polymerase chain reaction analysis with these hybrids and x-ray-resistant transformants obtained by introducing human chromosomes into SX10 indicated that the mutant was likely to be defective in DNA ligase IV. Sequence analysis of the DNA ligase IV gene confirmed that a defect in SX10 was attributed to a transition of G to A at nucleotide position 1413 of the gene, leading to an amino acid substitution from Trp at residue 471 to a stop codon. Revertant clones (Rev1-3) derived from SX10 showed a restored x-ray resistance; Rev1 reverted to the original nucleotide G at position 1413, whereas Rev2 and Rev3 to C. Transfection of a mouse DNA ligase IV cDNA vector into SX10 restored the resistance to both x-rays and bleomycin. SX10 showed a reduced frequency of chromosomal integration of transfected DNA, but the revertants restored the frequency found in the wild-type cells. These results suggest a possible involvement of DNA ligase IV in the integration event of foreign DNA as well as a crucial role in DNA double-strand break repair.
DNA double-strand breaks (DSBs) 1 are caused by ionizing radiation and DNA-damaging agents (1) or occur as intermediates in certain cellular processes such as V(D)J recombina-tion (2). If unrepaired or repaired inadequately, DSBs cause cell death by loss or inactivation of an essential gene, or various chromosomal rearrangements, which lead to genetic diseases and cancer (3). Therefore, very efficient mechanisms have evolved to repair such DNA damage. In eukaryotes, there are two major pathways for DSB repair: homologous recombination that involves the exchange of genetic information between a damaged chromosome and its undamaged partner, and nonhomologous end-joining (NHEJ) that requires little or no homology at broken DNA ends (4). Although homologous recombination is a main pathway in yeast cells, NHEJ predominantly acts in mammalian cells (5).
Mammalian cell mutants hypersensitive to ionizing radiation or radiomimetic agents have been isolated and greatly contributed to understanding the mechanism of DNA repair. These mutants have been shown to be deficient in DSBR and classified into at least nine x-ray cross complementing (XRCC) groups (6). Among them, mutants in XRCC4 -7 groups exhibit an extremely high sensitivity to ionizing radiation and show severe defects in DSBR and V(D)J recombination (7,8). With the hamster mutant XR-1 cell line, the XRCC4 gene has been cloned by rescue of its deficiency in DSBR and V(D)J recombination (9). This gene encodes a protein that interacts with DNA ligase IV and stimulates the enzyme activity (9 -11). The XRCC5 and XRCC6 genes encode Ku86 and Ku70, respectively, which are subunits of Ku protein (12)(13)(14). The Ku86-defective hamster cell line xrs-6 exhibits defects in both DSBR and V(D)J recombination processes (12,15). Ku protein binds to broken DNA ends and recruits DNA-PKcs encoded by XRCC7 (16 -20) to form a large protein complex. This complex recruits DNA ligase IV bound to XRCC4 and undergoes the end-joining reaction. Also, the complex reveals an activated protein kinase activity, which phosphorylates certain unidentified substrate(s) involved in the regulation of cell cycle checkpoints and controls the processing of the cell cycle during the process of DSBR. The mouse SX9 (21) and hamster V3 (19,20,22) mutant cell lines and severe-combined immune-deficient (scid) mice (20,23,24) all exhibit defects in DNA-PKcs and are defective in both DSBR and V(D)J recombination. The NHEJ pathway is conserved in yeast cells. Except for DNA-PKcs, four homologues corresponding to Ku70, Ku80, XRCC4, and DNA ligase IV also exist in Saccharomyces cerevisiae and function to repair DSBs in a similar manner (7).
In 1986, Sato et al. (25) isolated the x-ray-sensitive mutant SX10 by mutagenizing SR-1 cells of the mouse mammary carcinoma FM3A cell line with a potent mutagen N-methyl-NЈnitro-N-nitrosoguanidine. SX10 possesses a greatly reduced rejoining capacity of DSBs compared with the wild-type cells (26) and have been suggested to fall into the same complementation group as that of x-ray-hypersensitive mouse lymphoma M10 cells (25). Although SX10 has been characterized in several ways (25)(26)(27), a gene responsible for the defective phenotypes has not yet been identified. Here we show that SX10 is a DNA ligase IV-defective mutant and that the defect results from a nonsense mutation in the ligase gene. Also we show that SX10 has a lower activity to integrate transfected DNA into the genomic sites than wild-type cells, suggesting the involvement of NHEJ mechanism in the integration event.

EXPERIMENTAL PROCEDURES
Cells and Culture Methods-Mouse mammary carcinoma FM3A cell lines, wild-type SR-1, mutant SX10, and its 6-thioguanine-and ouabain-resistant derivative SX10TOR (25) were used; cells were grown in suspension in ES medium (Nissui Seiyaku Co., Tokyo) (28) supplemented with 4% bovine calf serum (HyClone). The mouse leukemia cell lines L5178Y and its x-ray-hypersensitive mutant M10 (29) were used; cells were grown in suspension in ES medium supplemented with 10% bovine calf serum. HeLa and normal human diploid fibroblast TIG-3 cells (obtained from the Tokyo Institute of Gerontology, Tokyo) were also used; cells were grown as monolayers in Dulbecco's modified Eagle's medium (Nissui Seiyaku Co.) supplemented with 10% calf serum or 10% fetal bovine serum (HyClone), respectively, and were subcultured by dispersing with 0.1% trypsin solution. All cultures were incubated at 37°C in an atmosphere of 5% CO 2 in air. Logarithmically growing cells were used for all experiments.
Survival and Growth Inhibition Assays-Cells were exposed to xrays (MBR-1520, Hitachi Medico Tokyo) at a dose rate of 1.1 Gy/min with a 0.5-mm Al/0.2-mm Cu filter at room temperature. Then the cells were diluted appropriately with growth medium and plated at 100 -10 4 cells/dish into 60-mm bacterial plastic dishes containing 5 ml of 0.15% agarose-containing growth medium and cultured for 14 -18 days. For growth inhibition assays, x-ray-treated or untreated cells were plated at 10 4 cells/well into 24-well plates (NUNC), cultured for 3-5 days, and counted for cell number. To determine sensitivity to bleomycin (Wako Pure Chem. Ind., Osaka), cells were plated at 100 -10 4 cells/dish into 60-mm dishes containing 5 ml of 0.15% agarose-containing growth medium with different drug concentrations and cultured for 14 -18 days. Resulting colonies were counted and plating efficiencies were calculated.
Cell Fusion and Hybrid Selection-Cell fusion was carried out as described previously (25). Human diploid fibroblast TIG-3 cells were used as a partner of cell fusion. Briefly, SX10TOR and TIG-3 cells were harvested and washed with serum-free ES medium; both cells (each 1 ϫ 10 6 ) were mixed in a glass centrifuge tube and centrifuged. To the cell pellet was gently added 0.3 ml of a 50% PEG1500 solution (Roche Molecular Biochemicals). After 1 min at room temperature, serum-free ES medium was gradually added and suspended. The mixture was spun down at low centrifugation, and the cell pellet was dispersed into growth medium, plated into 100-mm bacterial dishes, and incubated. On the next day, hybrid selection started in growth medium containing HAT/ouabain (75 M hypoxanthine, 0.5 M amethopterin, 20 M thymidine, and 1 mM ouabain). Two weeks later, the resulting colonies were picked, transferred to normal medium, and subcultured.
Chromosome-mediated Gene Transfer (CMGT)-This method was carried out as described previously (30). Briefly, metaphase chromosome preparations from human HeLa cells were passed through a needle and precipitated with calcium phosphate, and the precipitate was added to SX10 cells. On the next day, the cells were harvested, plated, and grown in 0.15% agarose-containing growth medium with 500 ng/ml bleomycin. The drug-resistant colonies were isolated, subcultured in growth medium without bleomycin, and assayed for sensitivity to x-rays. The presence of human Alu sequences was screened by PCR using Alu-PCR probe 451 as described (31). Also, the human DNA ligase IV sequence was examined by PCR by using the following primers h21F (5Ј-GATGTATTGATGGTTAATAATAAAAAGCTAGGG-3Ј) and hR1 (5Ј-AAAGCTAGCTTTAAATCAAATACTGGTTTTCTTC-3Ј). The PCR was performed in 20 l of a reaction mixture composed of 1ϫ PCR buffer (20 mM Tris-HCl, pH 8.4, 50 mM KCl), 1.5 mM MgCl 2 , 200 M each of four deoxynucleoside triphosphates, 0.2 l of the above primers, 40 ng of template genomic DNA, and 0.12 l of Taq polymerase (5 units/l, Life Technologies, Inc.). The reaction conditions were: 94°C for 3 min, 94°C for 40 s, 56°C for 30 s, and 72°C for 1 min for 35 cycles and 72°C for 10 min.
Identification of a Mutation-Total cellular DNA and RNA were prepared as described previously (32). To examine a mutation in the mouse DNA ligase IV gene, the following primers were used for both genomic PCR and RT-PCR: mF1 (5Ј-CAAACCGGAGGATTTGTTGTC-GCT-3Ј), mF2 (5Ј-CCCATTTATTCACAATGCGTTCGGGA-3Ј), mR1 (5Ј-CATCATCACTGTCTGCTGTTGTCTGACA-3Ј), mR2 (5Ј-TAGGTCAG-CCTTTGGATAACCATCTAA-3Ј), mR3 (5Ј-AGATGGCCTGTCACCAG-GAGGTGGTG-3Ј), and mR4 (5Ј-GGGTTGTAGGCCATCATCTCACCG-3Ј). The PCR was carried out as described above. The reaction conditions were: 94°C for 3 min, 94°C for 40 s, 56°C for 30 s, and 72°C for 1 min for 35 cycles and 72°C for 10 min. Amplified fragments were cloned into pGEM-T Easy Vector (Promega) and directly sequenced using a DNA sequencer model 4000 (Li-COR) with SequiTherm Long-Read cycle sequencing kits LC (Epicenter Technologies) as described (32). Otherwise, the fragments were digested with the restriction enzyme DdeI (New England BioLabs), electrophoresed on a 8% polyacrylamide gel, stained with 0.5 g/ml ethidium bromide, and photographed. Southern Blot Analysis-Southern blotting was performed as described previously (33). Briefly, 10 g of genomic DNA purified from cells were digested with DdeI, electrophoresed in a 1.0% agarose gel, and transferred to a nylon membrane (Hybond-N ϩ , Amersham Pharmacia Biotech). Then the membrane was probed with a 32 P-labeled 487-bp fragment of the mouse DNA ligase IV sequence (see Fig. 3A). The probe had been prepared by amplifying with mF2 and mR2 primers and SR-1 cell DNA as a template, and by digesting the 1099-bp product with DdeI.
Construction of an XRCC4 Expression Vector-Based on the reported sequence of human XRCC4 cDNA (9), we amplified the fragment of the entire open reading frame and inserted it into a pcDNA3 vector (Invitrogen). The resulting vector, designated phXRCC4, was purified by 2 cycles of CsCl density centrifugation and used for transfection.
Transfection by Electroporation-To isolate stable transfectants with the above phXRCC4 plasmid and the mouse DNA ligase IV expression vector pCLig4 (a gift from Frederic W. Alt), transfection was carried out with the plasmids linearized by digestion with restriction enzymes BglII and PvuI (Amersham Pharmacia Biotech), respectively, which cut the vector backbone once. SX10 was transfected by electroporation as described previously and incubated in growth medium for 24 h (33). The cells were then harvested, replated, and cultured in 0.15% agarosecontaining growth medium with 0.8 mg/ml active Geneticin (G418, Life Technologies, Inc.) for about 2 weeks. The resulting colonies were picked with micropipets, grown to mass culture in the presence of the drug, and used for further analysis.
To examine the ability of SR-1 and SX10 cells to integrate foreign DNA into their chromosomes, quantitative transfection assays were carried out with plasmid pSV2neo linearized with EcoRI before use (33). Briefly, 2 g of the plasmid DNA was transfected into 4 ϫ 10 6 cells by electroporation, and the cells were diluted appropriately in growth medium and cultured for 24 h. Then, the cells were collected, diluted, and replated at 1-5 ϫ 10 5 and 10 2 cells/60-mm dish into 0.15% agarosecontaining growth medium with and without active G418 (0.9 mg/ml), respectively. They were cultured for 14 -18 days at 37°C. The number of resulting colonies was scored, and the integration frequency was calculated by dividing the number of G418 r colonies with surviving cells.

RESULTS
SX10 Is Likely to Differ from XRCC1-9 in the Complementation Group -Sato et al. (25) reported that SX10 would belong to the same complementation group as that of the M10 mouse leukemia line, which is hypersensitive to x-rays (26). Because M10 had been suggested to belong to the XRCC4 group (34), we constructed a human XRCC4 expression vector, phXRCC4, transfected it into wild-type SX10 and M10, and obtained G418-resistant clones. We assayed their sensitivity to x-rays using two representative clones, SX10/hXRCC4 and M10/ hXRCC4. Fig. 1 clearly shows that SX10/hXRCC4 was as sensitive to x-ray irradiation as SX10 but that M10/hXRCC4 restored the resistance of L5178Y cells, the parental cells of M10, indicating that the XRCC4 vector was able to complement M10 but not SX10. We therefore conclude that SX10 is not part of the XRCC4 group. Although we measured DNA-PK activity in whole cell extracts by pull-down assays (35), no difference was found between SR-1 and SX10 cells (data not shown), indicating that SX10 would fall into a complementation group distinct from those of XRCC5-7. We measured the level of sister chro-matid exchange in SR-1 and SX10 and observed no significant difference between the two cell types (data not shown). Because the elevated sister chromatid exchange level is characteristic for a ␥-ray-sensitive Chinese hamster ovary mutant EM-9 classified into XRCC1 (36), our result suggests that SX10 does not fall into the XRCC1 group. Like the XRCC4 -7 mutant lines reported, SX10 revealed much greater sensitivity to ionizing irradiation than XRCC2 (37), XRCC3 (38), XRCC8 (39), and XRCC9 (40); therefore, we infer that SX10 would belong to a complementation group different from these groups.
Human Chromosome 13 Complements the X-ray Sensitivity of SX10 -To determine a human chromosome that could complement the defect in SX10, we established three hybrid lines between SX10 and normal human diploid fibroblast TIG-3 cells by selection in growth medium containing HAT/ouabain after cell fusion as described under "Experimental Procedures." We first isolated one of the hybrid clones, CC3, which exhibited a high resistance to x-rays, and subcloned CC3-2J and CC3-2P after serial passages. We examined their sensitivity to x-rays by growth inhibition assays. As shown in Fig. 2A, CC3-2J was almost as resistant as wild-type SR-1 whereas CC3-2P showed a striking sensitivity like SX10. Moreover, subclone CC3-2J-13D derived from the CC3-2J line lost the resistance to x-rays, indicating that a human chromosome conferring the x-ray resistance segregated from the CC3-2J cells. Then, with these hybrid lines, PCR analysis was performed using primer sets (MapPaires, Research Genetics) specific for each of human chromosomes. As a result, we found that only a 205-bp fragment amplified by using a primer set (D13S325) specific for human chromosome 13 was correlated with the x-ray resist-ance of the hybrid lines (Fig. 2B). These results imply that the restored resistance was concordant with human chromosome 13, thus suggesting a candidate gene residing in chromosome 13.
SX10 Appears to Be Defective in DNA Ligase IV-So far, the genes responsible for DNA repair residing in human chromosome 13 are DNA ligase IV (41), BRCA2 (42), and XPG (43). Because SX10 is highly sensitive to x-rays but insensitive to UV (25), we reasoned that the DNA ligase gene could be a responsible gene. Therefore, we established three x-ray-resistant transformants (Tor1-3) by the CMGT with human chromosomes prepared from HeLa cells, as described under "Experimental Procedures." The transformants all showed restored resistance to x-ray treatment, although Tor2 was less resistant than Tor1 and Tor3, as determined by a growth inhibition assay (Fig. 2C). To examine the existence of the human DNA ligase IV gene in the transformants, genomic PCR using h21F and h21R primers specific for the gene sequence was carried out. Like human diploid fibroblast TIG-3 cells, a 1645-bp fragment was amplified in the three transformants, whereas no fragment was amplified in either mouse SR-1 or SX10 cells (Fig. 2D). These results strongly suggest that SX10 is complemented by DNA ligase IV located in human chromosome 13.
A Point Mutation Is Present at the DNA Ligase IV Gene in SX10 -To verify directly the defect in SX10, we determined the nucleotide sequences of the DNA ligase IV gene in SR-1 and SX10. Because the mouse DNA ligase IV gene consists of only one exon (44), we were able to amplify a genomic sequence, including the entire open reading frame, using several pairs of primers prepared based on the mouse DNA ligase IV sequence and genomic DNA as a template (Fig. 3). We identified a point mutation from G to A at nucleotide position 1413 of the ligase gene in SX10 cells, resulting in an amino acid substitution from Trp at residue 471 to a stop codon (TGA). No other change was detected in the sequence in the mutant cells.
During the CMGT experiments mentioned above, we isolated three additional clones that were found on PCR amplification with either human Alu-PCR primers or the h21F and h21R primers for the human DNA ligase IV gene. We examined whether these clones, designated as Rev1-3, were sensitive to x-rays. Fig. 4 shows that they all were resistant to the radiation (data not shown for Rev3 cells). So we determined their sequences of the DNA ligase IV gene. As shown in Fig. 3A (upper  panel), the mutated nucleotide A in SX10 was changed to the original G in Rev1, and to C in both Rev2 and Rev3. Therefore, we conclude that these cells resulted from reversion of the mutated gene in SX10.
SX10 Is Hemizygous for the DNA Ligase IV Gene-We asked whether the mutation found in SX10 cells was homozygous or hemizygous. Because the transition of G to A at position 1413 results in the appearance of the recognition site (CTGAG) for DdeI, this mutation will generate 245-and 84-bp fragments from the 329-bp fragment observed in SR-1 (see Fig. 3, lower panel). We amplified a region (including the mutation site) of DNA ligase IV cDNA from SR-1, SX10, and the two revertants and a pair of mF2 and mR3, and analyzed them on a polyacrylamide gel (Fig. 5A). Without DdeI digestion, cDNAs from all the cell lines gave a 559-bp fragment. When digested with DdeI, the wild-type cDNA was split into 329-and 209-bp frag-ments, whereas the SX10 cDNA was into 245-, 209-, and 84-bp fragments (Fig. 5A). The two revertants Rev1 and Rev2 revealed the pattern of the wild-type cells, reflecting a nucleotide change at the mutated site (Fig. 3, upper panel). The 21-bp fragment was too small to be visible on the gel for any cell line. These results suggest that SX10 has only a single mutated allele of DNA ligase IV. To further confirm this, we analyzed genomic DNA digested with DdeI by Southern blotting. Fig. 5B shows that, different from SR-1, Rev1, and Rev2, only SX10 displayed a smaller 245-bp band that came from the mutated sequence. These results demonstrate that SX10 is hemizygous  5. Hemizygosity of SX10 for the DNA ligase IV gene. A, mutation detected by RT-PCR analysis. The sequence of the DNA ligase IV gene was amplified by RT-PCR with total RNA from wild-type SR-1, SX10, and two revertants and the primers mF2 and mR3 (see Fig. 3,  lower panel). The transition of G to A at nucleotide position 1413 results in the occurrence of sequence CTCAG, a recognition site of DdeI, so that digestion of the amplified 559-bp fragment with DdeI generates 329and 209-bp fragments in SR-1, Rev1, and Rev2, but 245-, 209-, and 84-bp fragments in SX10. B, Southern blot analysis. Genomic DNA from SR-1, SX10, and the two revertants were digested with DdeI and analyzed by Southern blotting using a probe shown in Fig. 3A (lower  panel).
for the DNA ligase IV gene.
Transfection of the Mouse DNA Ligase IV Expression Vector Is Able to Complement the Defect in SX10 -To prove complementation with exogenous DNA ligase IV, we transfected the PvuI-linearized, mouse DNA ligase IV expression vector pCLig4 into SX10 cells and isolated several G418 r clones. Two representative clones, designated as Tec1 and Tec2, were assayed for their radiosensitivity. It is evident that both transfectants restored a normal level of resistance to treatment with either x-rays (Fig. 6A) or bleomycin (Fig. 6B). We assayed the growth rates of SR-1, SX10, revertants, and transfectants by culturing at a density of 10 4 cells/well per ml in 24-well plates and counting the cell number every day. SX10 grew with a longer doubling time of 23 h compared with that of 14 h found in SR-1. Rev1 and Rev2 grew with a doubling time of 19 and 20 h, respectively; Tec1 and Tec2 cells revealed the doubling times of 21 and 20 h, respectively. These results indicate that the lower growth rate in SX10 was slightly recovered in the revertants and transfectants.
SX10 Exhibits Reduced Chromosomal Integration of Transfected DNA-Transfection of foreign DNA into mammalian cells results in the integration of the DNA into random chromosomal sites (45). To assess whether the defect in DNA ligase IV affected the integration event, we carried out quantitative transfection assays with EcoRI-linearized pSV2neo and compared integration frequencies in SR-1, SX10, and two revertants. In Fig. 7, the basal level of the integration frequency in wild-type SR-1 was 5.6 ϫ 10 Ϫ4 per surviving cell on average (three determinations). The frequency in SX10 was reduced to 5% of the frequency found in SR-1. The two revertants Rev1 and Rev2 considerably resumed the integration activity, although Rev1 exhibited slightly lower frequencies whereas Rev2 had higher frequencies. This difference between the wild-type and SX10 cells could come from a reduced uptake of the DNA into the nucleus following transfection. To test this possibility, we carried out transient assays by transfecting expression plasmids pSV2Luc or pSV2CAT linearized at the backbone and determining the expression of luciferase or CAT activities, respectively, 24 h after transfection; no significant differences were found in the expression levels per cell between SR-1 and SX10 (data not shown). This rules out a possible reduction of DNA uptake into the nucleus in the mutant. DISCUSSION In this study, we have presented evidence showing that the mouse SX10 cell line that is hypersensitive to x-rays has a mutated DNA ligase IV gene.
The mutation identified is a transition of G to A at nucleotide position 1413 of the DNA ligase IV gene, resulting in the generation of a stop codon (Fig. 3A). The mutation site in SX10 cells is located at the C-terminal portion of the catalytic domain. The mouse and human DNA ligase IV genes have no intron, consisting of one open reading frame (44,46). DNA ligase IV is composed of a catalytic domain in the N-terminal region and two domains containing BRCT motifs in the Cterminal region; it binds to the XRCC4 protein in the region between the two motifs (47). We have not directly examined the catalytic activity of the mutated gene product. Nevertheless, the product must have no activity, because it is a truncated form that has lost both parts of the catalytic domain and the BRCT domains along with the XRCC4-binding site. Recently, the radiosensitive 180BR cell line derived from a leukemia patient has been shown to possess a mutation at position 833 of the DNA ligase IV cDNA, leading to an Arg to His substitution (48). This mutation decreases the ability of the ligase IV to form an enzyme-adenylate complex.
SX10 was found to be hemizygous for the DNA ligase IV gene. We could find only one allele in both revertants Rev1 and Rev2 (Fig. 5, A and B), which were obtained in the CMGT experiments. We had examined the reversion of SX10 by selecting over 10 7 cells in bleomycin-containing plates but failed to obtain any revertants. Introduction of foreign DNA or chromosomes into cells seems to induce a mutation or rearrangements in the host genome. 2 We have not examined whether the parental SR-1 cells are hemizygous for the DNA ligase IV gene. However, it may be possible that the hemizygosity has been established prior to mutant selection following mutagenesis 2 D. Ayusawa, and H. Koyama, unpublished data. with N-methyl-NЈ-nitro-N-nitrosoguanidine, because the karyotype of SX10 considerably differs from that of SR-1 (data not shown).
Hybrids between SX10 and XRCC4-deficient M10 were unable to complement each other for their radiosensitivity, suggesting that the two mutant lines would belong to the same complementation group (25). However, in this study, we found that this was not the case. The expression vector of the human XRCC4 failed to rescue the SX10 defect (Fig. 1). The reason for this difference is not known, but the complementing chromosome in the hybrid cells might have segregated during the early passage after the hybrid selection, or it is also possible that, in the hybrid cells, the mutated XRCC4 protein would interfere with the normal XRCC4 that specifically binds to DNA ligase IV and acts in the NHEJ process.
Transfection of SX10 with the mouse DNA ligase IV-expressing vector pCLig4 (44) restored the x-ray and bleomycin resistance seen in wild-type cells (Fig. 6). This result clearly demonstrates the deficiency of DNA ligase IV in SX10 cells. The extent of the recovery in the transfectants was lesser than that found in revertant Rev1 (data not shown). In addition, Rev2 slightly resumed the resistance compared with Rev1 (data not shown). This would be attributed to a reversion of the mutated nucleotide A to C in Rev2 in place of G, a wild-type nucleotide observed in Rev1. This incomplete recovery of the sensitivity to x-rays in the revertants and transfectants may suggest that SX10 has a genetic or epigenetic defect(s) other than the nonsense mutation in the DNA ligase IV gene.
The growth rate of SX10 was not completely recovered in either revertants or transfectants. The doubling time of SX10 was 23 h, whereas that of the parental SR-1 was 14 h. Rev1, Rev2, Tec1, and Tec2 grew with reduced doubling times of 19 -21 h, thus showing a 20 -40% recovery of growth rates compared with SX10 cells. This partial recovery might also result from genetic or epigenetic defect(s) other than the defect in DNA ligase IV. Recently, DNA ligase IV-deficient mice have been reported to be embryonic lethal due to massive neuronal cell death at the developmental stage (44,49,50); however, fibroblast cells cultured from the mutant embryos are viable and grow but reveal a reduced growth rate and a lowered saturation density (44). Furthermore, we have established DNA ligase IV-negative mutants from chicken B-cell lymphoma DT40 cells by gene targeting; the mutants exhibit a small but significant reduction of growth rate. 3 Taken together, these observations may suggest that DNA ligase IV could partially be involved in DNA replication or repair-coupled DNA replication. Although in mammalian cells DNA ligase I is thought to function in DNA replication, it is not ruled out that other DNA ligases, including ligase IV, may also contribute to DNA replication (51). However, human pre-B cell mutants deficient in DNA ligase IV do not exhibit growth retardation (46); this may imply that the mutants have a level of DNA ligase I activity enough to fully support the replication.
Finally, the present study shows that the chromosomal integration of transfected, linear DNA into SX10 cells was greatly reduced as compared with wild-type SR-1 cells (Fig. 7) and this defect was restored to normal levels in the two revertants. DNA ligase IV has been implicated in DSB repair through NHEJ (51). This end-joining mechanism has long been postulated to be involved in the chromosomal integration of transfected DNA; that is, the integration event could occur by joining of both ends of the input DNA and broken chromosomal DNA (45,52), although there is no direct evidence to support this view. Our finding indicates that DNA ligase IV is involved in an integration process, and, in other words, suggests that the integration event occurs through an NHEJ process in support of the above view (45,52). Recently, we have obtained the data supporting the contribution of NHEJ to the integration process by an extensive structural analysis of integration junctions. 4 However, this observation is inconsistent with the data by Merrihew et al. (53), which do not agree with the end-joining mechanism for the chromosomal integration. We are now obtaining further evidence for the involvement of an end-joining mechanism in random integration.