Ionizing radiation exposure results in up-regulation of Ku70 via a p53/ataxia-telangiectasia-mutated protein-dependent mechanism.

Genome damaging events, such as gamma-irradiation exposure, result in the induction of pathways that activate DNA repair mechanisms, halt cell cycle progression, and/or trigger apoptosis. We have investigated the effects of gamma-irradiation on cellular levels of the Ku autoantigens. Ku70 and Ku80 have been shown to form a heterodimeric complex that can bind tightly to free DNA ends and activate the protein kinase DNA-PKcs. We have found that irradiation results in an up-regulation of cellular levels of Ku70, but not Ku80, and that this enhanced level of Ku70 accumulates within the nucleus. Further, we uncovered that the postirradiation up-regulation of Ku70 utilizes a mechanism that is dependent on both p53 and damage response protein kinase ATM (ataxia-telangiectasia-mutated); however, the activation of DNA-PK does not require Ku70 up-regulation. These findings suggest that Ku70 up-regulation provides the cell with a means of assuring either proper DNA repair or an appropriate response to DNA damage independent of DNA-PKcs activation.

Ionizing radiation (IR) 1 exposure results in oxidative damage to DNA, one of the outcome of which is introduction of double strand breaks (DSBs) within the genome. Such genotoxic events activate a number of signaling pathways that serve, for example, to activate DNA repair mechanisms, halt cell cycle progression, and/or trigger advancement into apoptosis. These damage response pathways exist to facilitate genome repair and limit the accumulation of heritable genetic errors. Not surprisingly, a wide range of tumor cells has been shown to be defective in one or more of the elements that comprise these genome damage response pathways (see Ref. 1), indicating that such responses play a key role in limiting tumor cell formation.
A central figure in many damage response pathways is the tumor suppressor protein p53. p53 is a transcription factor that, when active, serves to up-regulate the cellular abundance of a cadre of molecules important in genotoxic response. Such p53-inducible proteins include p21 WAF-1/CIP-1 , bax, mdm2, gadd45, and KILLER/DR5, which perform a variety of functions in response to DNA damage such as halting cell growth and/or propelling cells into apoptosis (2) Thus, p53 performs a pivotal role in choreographing appropriate cellular response to genome damaging events. Other proteins responsible for an appropriate and timely response to IR-induced genome damage are members of a protein kinase family which share homology to the lipid kinase phosphotidylinositol-3 kinase (3,4). ATM, the protein mutated in the human genetic disorder ataxia-telangiectasia (A-T), is an important member of this kinase family. Cells from A-T patients are radiosensitive and have been shown to be deficient in a number of signaling events that normally occur following DSB formation (see Ref. 5). For example, A-T cells display faulty cell cycle checkpoint function in response to IR exposure (6 -9). Recent work has shown that ATM directly phosphorylates p53 in response to ␥-irradiation (10,11). The lack of ATM-mediated irradiation-induced phosphorylation of p53 is likely to be directly responsible for the blunted up-regulation of p21 WAF-1/CIP-1 in A-T cells resulting in their G 1 /S cell cycle checkpoint defect (12)(13)(14). In addition, A-T cells are also defective in several apparent p53-independent responses such as the activation of c-Abl (15) and the phosphorylation of the p34 subunit of replication protein A (16).
Another member of the ATM family of protein kinases which plays an important role in genome damage response is the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs). Deficiencies in DNA-PKcs activity are responsible for the lack of appropriate response to DSBs observed in the radiosensitive mammalian xrs-6 cell line (17) and scid (severe combined immunodeficiency) mouse strain (18,19). The mutant phenotypes displayed by these lines and mice indicate that loss of DNA-PKcs catalytic activity results in defective DSB rejoining and the inability to facilitate V(D)J recombination. The functional contribution of DNA-PKcs-dependent phosphorylation to DSB responses is currently unclear; however, recent studies have shown that DNA-PK, unlike ATM, is not involved in the activation of irradiation-induced cell cycle checkpoints (20,21).
The DNA-PK holoenzyme consists of three subunits: a ϳ450-kDa catalytic subunit (DNA-PKcs) and a heterodimeric complex composed of the proteins Ku70 (70 kDa) and Ku80 (86 kDa). The Ku proteins were originally identified as autoantigens in patients with autoimmune disorders (22) and were found to bind tightly to linear duplex DNA (23). Following the generation of DSBs by either exogenous forces (e.g. ␥-irradia-tion) or intrinsic mechanisms (e.g., during V(D)J recombination) the Ku70-Ku80 complex binds to free DNA ends and subsequently recruits and activates DNA-PKcs at the site of DSBs (24).
In this manuscript we report that cellular levels of Ku70, but not Ku80, increase multifold within the nucleus of irradiated cells. Up-regulation of Ku70 following IR requires both functional p53 and ATM. Furthermore, we report that up-regulation of Ku70 is not required for appropriate activation of DNA-PK activity following genome damage. These results indicate that, in addition to its characterized function in activating DNA-PK, Ku70 plays yet another role(s) in DNA damage response.

EXPERIMENTAL PROCEDURES
Cell Lines and Culture Conditions-Normal diploid human lung fibroblast line MRC-5 was obtained from the American Type Tissue Collection. Cells were cultured in Eagle's minimum essential medium supplemented with 10% fetal calf serum. Human lung carcinoma line H1299 and its derivative temperature-sensitive line H1299-Val-138 (25) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C as were fibroblasts cultured from c-ablϪ/Ϫ mice. Where indicated, H1299-Val-138 cells were shifted to an incubator maintained at 32°C for 6 h prior to irradiation. Normal (GM0536A and B-310) and A-T lymphoblast (GM01525E and IARC12/ AT3) lines were grown in RPMI 1640 containing 20% heat-inactivated fetal calf serum. GM0536A and GM01525E lines were obtained from NIGMS Human Genetic Mutant Cell Repository (Camden, NJ). All cell lines were grown in a humidified 5% CO 2 atmosphere at 37°C. Where indicated, cells were exposed to 10 Gy of IR from a 137 Cs source (Gammacell 1000, Atomic Energy of Canada Ltd.; dose rate, 318 rad/min). Following irradiation, cells were returned to the incubator and harvested at the indicated time points.
Antibodies-The Ku70 mouse monoclonal antibody Ab-5, Ku80 monoclonal antibody Ab-2, and DNA-PKcs monoclonal antibody Ab-2 were purchased from Lab Vision Corporation (Fremont, CA). The anti-p53 antibody DO-1 was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-␣-tubulin monoclonal antibody DM1A was a generous gift from Dr. D. W. Cleveland (Ludwig Institute for Cancer Research, La Jolla, CA).
Electrophoresis and Immunoblotting-At the indicated time points, cells were harvested with trypsin, washed three times with cold (4°C) PBS, and lysed by the addition of SDS-solubilization buffer (125 mM Tris-HCl, pH 7.5, 5 mM EDTA, and 1% SDS). Lysates were then boiled for 5 min, sonicated briefly, centrifuged (5 min/14,000 rpm), and the supernatants were removed and stored at Ϫ80°C. Protein concentrations were determined using the BCA protein assay (Pierce). Prior to electrophoresis, appropriate volumes of cell lysates were diluted in 3 ϫ SDS-sample buffer (150 mM Tris-HCl, pH 6.8, 10% ␤-mercaptoethanol, 20% glycerol, 3% SDS, 0.01% bromphenol blue, 0.01% pyronin-Y) and boiled for 2 min. SDS-polyacrylamide gel electrophoresis, electrotransfer of the gels to nitrocellulose sheets, and antibody labeling were conducted as outlined (26). Immunoreactive bands were visualized using Supersignal CR-HRP (Pierce) chemiluminescent substrate and recorded on Kodak XAR film. Quantitation of immunoblot signals was performed using an Agfa SnapScan flatbed scanner and the NIH Image (version 1.62) software package using the integrated density function. In all cases, protein loadings and antibody dilutions were adjusted to assure that immunoblotting was conducted at subsaturating conditions.
In Vitro Kinase Assays-Kinase reactions were performed using an adaptation of the DNA-PKcs immunoprecipitation-based protocol of Kurimasa et al. (27). Briefly, cells were irradiated with 10 Gy of ␥-radiation and collected at indicated time points. The cells were subsequently washed with ice-cold PBS and lysed in 1 ϫ lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.5% Nonidet P-40, 1 mM NAF, 1 mM dithiothreitol, 1 mM sodium vanadate, and protease inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pepstatin-A)). After adjusting the lysate concentration for equal protein content, DNA-PKcs antibody was added to the lysate at a concentration of 5.0 g/mg lysate and incubated for 2 h on ice. After this step, 15 l of a 50% slurry (in PBS) of protein A/G-Sepharose beads (Amersham Pharmacia Biotech) were added, and the incubation was continued for another 1 h on an end-over-end rotator at 4°C. The immune complexes were washed three times with lysis buffer containing 500 mM NaCl and twice with kinase buffer (25 mM HEPES, pH 7.5, 500 mM KCl, 0.5 mM EDTA, 5 mM dithiothreitol, 2.5 mM phenylmethylsulfonyl fluoride). The beads were then suspended in a minimal volume of kinase buffer and used in kinase reactions.
Kinase reactions were carried out in a final reaction volume of 35 l. To the slurry of beads containing immunoprecipitated DNA-PKcs, 1 g of bacterially synthesized, purified recombinant GST-p53 fusion protein, 5 M cold ATP, 30 Ci of [ 32 P-␥]ATP (7,000 Ci/mmol, ICN) and 500 ng of sonicated salmon sperm DNA were added. This reaction was incubated at room temperature for 30 min and terminated by adding an equal volume of 3 ϫ SDS-sample buffer followed by immersion in a boiling water bath. The reaction products were resolved on 8% acrylamide gels and electrically transferred onto Immobilon-P sheets (Millipore) using a Bio-Rad semidry transfer apparatus. Quantitation was done using a Bio-Rad laser densitometer (GS-710 calibrated imaging densitometer).
Subcellular Fractionation-Subcellular fractionation was carried out as described previously (26). Approximately 2 ϫ 10 7 irradiated (10 Gy) and nonirradiated MRC-5 fibroblasts were harvested by scraping and rinsed twice in-cold PBS. The cells were then swollen in ice-cold hypotonic lysis buffer (20 mM HEPES, pH 7.1, 5 mM KCl, 1 mM MgCl 2 , 10 mM N-ethylmaleimide, and protease inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 5.0 g/ml pepstain-A, 2.0 g/ml chymostain, 5.0 g/ml leupeptin, 5.0 g/ml aprotinin, 5.0 g/ml antipain)) for 10 min on ice. The cells were then lysed by 20 strokes in a Dounce homogenizer with a tight pestle and the nuclei cleared by centrifugation (400 ϫ g, 10 min). Following this step, the supernatant (cytosolic fraction) was transferred to a Centricon microconcentrator (Amicon), concentrated approximately 5-fold, and subsequently stored at Ϫ80°C. The nuclei were washed three times in lysis buffer and stored at Ϫ80°C prior to analysis.
Immunofluorescence Microscopy-Indirect immunofluorescence microscopy was conducted as outlined previously (26). MRC-5 fibroblasts were cultured on sterilized glass coverslips to ϳ75% confluence prior to irradiation. Postirradiation the cells were fixed by immersion in PBS containing 3.0% formaldehyde for 30 min at room temperature, after which the cells were permeabilized by immersion in PBS containing 0.1% Triton X-100. Following this step the coverslips were rinsed in PBS and blocked by immersion in PBS containing 5% bovine serum albumin. Immunofluorescent staining was done using anti-Ku70 followed by fluorescein isothiocyanate-labeled anti-mouse secondary antibody (Kirkegaard and Perry Laboratories, Gaithersburg, MD). Cells were counterstained with Hoescht 33258 (0.1 g/ml final concentration) to reveal DNA. Images were captured on 35-mm film (Kodak T-Max 400) using a Nikon microscope equipped with epifluorescence optics.

Cellular Levels of Ku70
Rise Following IR Exposure-In our efforts to define signal transduction pathways that are activated following genotoxic events we have investigated the fluctuation of the Ku70 and Ku80 autoantigens following IR exposure. The Ku70-Ku80 complex is responsible for the activation of DNA-PKcs in part because of its intrinsic property of binding to free DNA ends with high affinity. Hence, it is conceivable that the cellular levels of Ku70 and/or Ku80 could be influenced positively in response to the introduction of DSBs into the genome. An initial report on this subject (28) indicated that in normal peripheral blood lymphocytes and a leukemia tumor cell line exposure to ␥-irradiation resulted in an apparent 3-4fold rise in Ku70.
To confirm that Ku70 levels are altered in response to IR exposure we subjected normal human diploid lung fibroblasts (MRC-5 cells) to analysis following ␥-irradiation. 2 h following 10 Gy of IR we measured an approximate 3-fold rise in Ku70 (mean ϭ 2.7 Ϯ 0.5 S.D., n ϭ 5) over the levels present in nonirradiated cells (Fig. 1a). This quantitative increase in Ku70 was observed 30 min after irradiation and persisted during the analyzed 2-h time frame. We have examined MRC-5 cells 6 h postirradiation and observed no further significant Ku70 accumulation than what was noted 2 h after IR exposure (data not shown). We noted a slight increase in cellular levels of the Ku80 protein after IR exposure (Fig. 1b); however, this increase was significantly blunt compared with the observed rise in Ku70 during the same time period. As expected, we noted a 5-7-fold rise in p53 following IR exposure during this period (Fig. 1c). An immunoblot run in parallel and probed with anti-tubulin confirmed equivalent protein loadings (Fig. 1d).
p53 and ATM Are Required for Postirradiation Up-regulation of Ku70 Levels-p53 is responsible for the up-regulation of a number of proteins in response to genotoxic events (2). To determine if Ku70 up-regulation observed in normal human fibroblasts is a p53-dependent event, we measured Ku70 accumulation in irradiated H1299 cells, a p53-null human lung carcinoma cell line (29). Irradiation of these cells with 10 Gy of IR resulted in no significant increase in cellular levels of Ku70 at 2 h (1.1-fold rise Ϯ 0.15, n ϭ 7) (Fig. 2a).
To confirm that loss of functional p53 in H1299 cells was the underlying cause of loss of irradiation-dependent Ku70 increase we employed a derivative cell line that expresses a conditionally functional form of p53. This cell line, termed H1299-Val-138, is engineered to express ectopically a human p53 molecule that contains an alanine to valine mutation at codon 138 (25). This change renders the molecule temperaturesensitive, similar to an equivalent alanine to valine mutation at codon 135 in murine p53 (30). When H1299-Val-138 cells were cultured for 6 h at the permissive temperature (32°C) and then irradiated with 10 Gy of IR these cells showed, 2 h after irradiation, an approximate 2-fold up-regulation of cellular Ku70 levels (1.9 Ϯ 0.5, n ϭ 8) (Fig. 2b). Conversely, when irradiated H1299-Val-138 cells were maintained at the nonpermissive temperature (37°C) there was no observed increase in Ku70 levels at the same time point (1.0 Ϯ 0.4 h, n ϭ 8) (Fig. 2c). Taken together, these findings validate that the postirradiation rise in Ku70 is dependent on p53.
ATM has been shown to phosphorylate p53 in response to IR exposure (10,11) and is required for the p53-dependent rise in several proteins following DSB formation (see Refs. 5 and 31).
To determine if p53-dependent up-regulation of Ku70 levels requires ATM function we subjected normal human and A-T patient lymphoblasts to 10 Gy of ␥-irradiation. (Lymphoblasts are Epstein-Barr virus-immortalized peripheral blood lymphocytes and have been shown to possess intact p53-dependent responses (7,13).) Consistent with the observations of Kumaravel et al. (28) we observed an approximate 2-fold increase in Ku70 in normal lymphoblasts 2 h after IR exposure (2.1 Ϯ 0.6, n ϭ 5) (Fig. 3a). However, we observed no postirradiation increase in Ku70 in either the A-T lymphoblast lines IARC12/ AT3 (1.0 Ϯ 0.26, n ϭ 5) or GM01525E (1.0 Ϯ 0.39, n ϭ 8) (Fig.  3, b and c) 2 h after IR exposure. Further, two additional A-T lymphoblast lines displayed parallel results (data not shown). These observations indicate that in addition to p53, ATM is required for the up-regulation of Ku70 after irradiation.

DNA-PK Activation Occurs Normally in the Absence of IRinduced Ku70 Accumulation-Previous work has shown that
Ku70 and Ku80 play an essential role in activating DNA-PKcs (24). Hence, we examined if the observed rapid up-regulation of Ku70 levels was required for timely activation of DNA-PKcs after IR exposure. To address this issue we irradiated H1299-Val-138 cells cultured at both the permissive and nonpermissive temperature with 10 Gy of IR and subsequently measured DNA-PKcs activity by employing an in vitro kinase assay using recombinant GST-p53 as a substrate (p53 has been shown to be an in vitro substrate for DNA-PKcs catalytic activity (32)). At the permissive temperature (active p53) we observed a 4-fold rise in DNA-PK-associated p53 phosphorylation after ␥-irradiation (Fig. 4a). Equivalent results were observed in extracts of irradiated H1299-Val-138 cells cultured at the nonpermissive temperature (Fig. 4b). Immunoblots with anti-DNA-PKcs confirmed the presence of equivalent amounts of immunoprecipitated DNA-PKcs in each of the kinase assays. Further, we observed normal activation of DNA-PK activity in irradiated A-T lymphoblastoid cells (data not shown). These findings clearly indicate that the irradiation-induced rise in Ku70 is not required for DNA-PK activation.  2. p53 is required for the postirradiation rise in Ku70 levels. a, cultures of the p53-null human lung carcinoma cell line H1299 were subjected to 10 Gy of IR, and extracts were prepared at the indicated times. Extracts were subjected to immunoblot analysis with anti-Ku70 (upper panel) or anti ␣-tubulin (lower panel). b, cultures of H1299-Val-138 (V138) cells which express a temperature-sensitive form of human p53 were cultured at the permissive temperature (32°C) for 6 h and then irradiated (10 Gy). Extracts were formed and analyzed as in a. c, H1299-Val-138 cells cultured at the nonpermissive temperature (37°C) were irradiated (10 Gy), and extracts were formed and analyzed as indicated in a. The relative Ku70 immunoblot signal intensity is indicated.

FIG. 3. A-T cells do not display an irradiation-induced rise in
Ku70 levels. a, normal lymphoblast line B-310 cells were exposed to 10 Gy of IR and subjected to immunoblot analysis with anti-Ku70 (upper panel) and anti-␣-tubulin (lower panel) at the indicated time points. b, A-T lymphoblast line IARC12/AT3 cells were exposed to 10 Gy of IR and analyzed as in a. c, A-T lymphoblast line GM01525E cells were exposed to 10 Gy of IR and analyzed as in a. The relative Ku70 immunoblot signal intensity is indicated.
we found Ku70 localized exclusively within the nucleus in both nonirradiated cells and irradiated cells (Fig. 5a). Negligible cytoplasmic staining was observed.
To confirm these findings independently, we subjected irradiated and nonirradiated MRC-5 cells to subcellular fractionation (Fig. 5b). Analysis of nuclear fractions by immunoblotting showed a significant elevation of Ku70 within the nuclear fraction of irradiated cells compared with nonirradiated controls, consistent with the immunofluorescence studies. Ku70 was not detected in the cytoplasmic fractions obtained from either irradiated or nonirradiated fibroblasts. ␣-Tubulin (a cytoplasmic protein) was detected predominantly in the cytoplasmic fractions, indicating that lysis was virtually complete. We are unsure why our results differ from the observations of Kumaravel et al. (28); however, others have noted a difference in Ku70 localization in lymphoid and fibroblast type cells (33). Nevertheless, our data indicate that this protein accumulates exclusively within the nucleus in irradiated fibroblasts. DISCUSSION The evidence presented in this study indicates that ionizing radiation results in increased cellular levels of Ku70, a finding that is in agreement with a previously published report (28). Using lymphoid cells, Kumaravel and co-workers noted a 3-4fold increase in Ku70 2 h postirradiation (earlier time points were not assayed), and this elevated level persisted for 72 h. We have extended this work by demonstrating that in both human fibroblasts and lymphoblasts, Ku70 undergoes a similar (2-3fold) rise rapidly (within 30 min) after irradiation and that this phenomenon requires both functional ATM and p53. Additionally, we have found that Ku70 levels rise in irradiated normal and c-ablϪ/Ϫ mouse fibroblasts, 2 indicating that the up-regulation of Ku70 is independent of c-Abl activity and IR-induced Ku70 up-regulation is a feature common to many, if not all, normal cell types.
Our observation that the up-regulation of Ku70 requires both functional ATM and p53 leads us to propose the following signal cascade: DSB Ͼ ATM Ͼ p53 Ͼ Ku70. The involvement of p53 in postirradiation Ku70 up-regulation suggests that this phenomenon is caused by increased p53-mediated transcription of the Ku70 gene. However, preliminary evidence failed to detect a rise in Ku70 mRNA 2 h after ␥-irradiation as judged by Northern blot analysis (data not shown). Thus, at this time, we are unsure of the nature of the molecular mechanism that gives rise to the observed increase in Ku70 levels.
Although we documented the multifold up-regulation of Ku70 after irradiation, we noted no similar rise in its heterodimeric binding partner Ku80, a finding that is in agreement with other reports (34). This observation suggests that the postirradiation rise in Ku70 does not result in higher cellular concentrations of the Ku70-Ku80 heterodimer. Interestingly, a splice variant of the Ku80 protein termed KARP-1 (35) has been also shown to be up-regulated following IR exposure in an ATM/p53-dependent manner (34). Like the Ku70-Ku80 complex, several lines of evidence indicate that KARP-1 is involved in modulating DNA-PK activity (35). Thus, it is tempting to speculate that KARP-1 and Ku70 may form a complex after DNA damage because such events lead to increased abundance of both molecules.
Diminution of DNA-PK activity has been shown to be responsible for the scid defect in mice (18,19). One of the phenotypic hallmarks of scid cells is extreme radiosensitivity, indicating the requirement for DNA-PK activity in responding appropriately to genome damage. Similarly, cells harboring mutated Ku80 or Ku70 alleles are also radiosensitive (36 -39), reinforcing the concept that Ku function is required for DNA-PKcs activation.
Observations made on recently developed Ku70Ϫ/Ϫ or Ku80Ϫ/Ϫ mice have begun to shed light on possibility that the Ku proteins perform functions that are independent of their role in DNA-PKcs activation. Ku80- (37,38) and Ku70-deficient mice (39), in addition to displaying radiosensitivity, are of significantly reduced size compared with wild type or heterozygous littermates. On the other hand, DNA-PKcs-deficient mice do not display such growth retardation (40), indicating that Ku proteins play a role in development which is independent of DNA-PKcs. Furthermore, scid mice and DNA-PKcs-deficient mice display blocked V(D)J coding end joining but show overall normal signal joint formation (40 -42), whereas Ku80Ϫ/Ϫ mice display both defective coding and signal joint formation (37,38). Paradoxically, whereas DNA-PKcsϪ/Ϫ and Ku80Ϫ/Ϫ mice display both abnormal T-and B-lymphocyte development, Ku70Ϫ/Ϫ mice showed abnormal B-lymphocyte but normal T-lymphocyte development and maturation (39) 4. Lack of Ku70 up-regulation has no effect on the postirradiation activation of DNA-PKcs. a, H1299-Val-138 cells were cultured at the nonpermissive temperature (32°C) for 6 h prior to exposure to ␥-irradiation (10 Gy). At the indicated time points after irradiation, extracts from the cells were formed, DNA-PKcs was immunoprecipitated, and associated kinase activity was determined by in vitro phosphorylation of purified recombinant human p53-GST fusion protein (upper panel) as outlined under "Experimental Procedures." To ensure that equivalent amounts of DNA-PKcs were subjected to kinase assays, the immunoprecipitated material was subjected to immunoblot analysis with DNA-PKcs antibody (lower panel). b, H1299-Val-138 cells were cultured at the nonpermissive temperature (37°C), exposed to ␥-irradiation (10 Gy), and analyzed as in a. The relative signal intensity of the radiolabeled p53-GST is shown.
FIG. 5. Ku70 accumulates within the nucleus of irradiated human fibroblasts. a, MRC-5 fibroblasts were seeded on presterilized coverslips and either unexposed (No IR) or exposed to 10 Gy of IR and processed for immunofluorescence microscopy 2 h (ϩIR-2 h) or 18 h (ϩIR-18 h) postirradiation. The cells were labeled with anti-Ku70 and counterstained with Hoescht 33258 to stain DNA fluorescently. Note the colocalization of Ku70 with nuclei in all cells and the general absence of Ku70 staining in the cytoplasm. b, approximately 2 ϫ 10 7 irradiated MRC-5 cells (10 Gy, 2 h postirradiation) and an equal number of nonirradiated cells were subjected to subcellular fractionation. 10% of the isolated nuclear (lanes 1 and 3) and cytoplasmic (lanes 2 and 4) fractions from nonirradiated (lanes 1 and 2) and irradiated cells (lanes 3 and 4) were subjected to immunoblot analysis with anti-Ku70 (upper panel). Additionally, 2.5% of these fractions were subjected to immunoblot analysis with anti-tubulin (lower panel) to confirm complete cell lysis. and Ku80 but not Ku70. Taken together, these observations argue that both Ku80 and Ku70 perform functions independent of the activation of DNA-PKcs and that Ku70 and Ku80 may possess some separable functions.
Several lines of evidence indicate that the Ku proteins play an essential role in DNA repair. For example, Ku80-and Ku70deficient mice show severe impairment in their ability to repair DSBs (38,39). Additionally, Saccharomyces cerevisiae mutants that do not express either the yeast homologs of Ku80 or Ku70 show diminished capacity for the repair of DSBs (43,44). (It bears noting that although S. cerevisiae does express a number of ATM-related kinases (i.e. Mec1p, Tel1p, Tor1p, and Tor2p), a DNA-PKcs homolog has not been identified in this organism to date.) In addition to the aforementioned property of binding to free DNA ends, in vitro studies have shown that the Ku heterodimer can translocate internally on DNA fragments (45), transfer between DNA fragments possessing cohesive ends (46), and join DNA fragments (47). Further, Ku70-Ku80 facilitates ligation of DNA molecules possessing small (Ͻ4 base pairs), blunt, or noncohesive termini by eukaryotic ligases (48), and Ku has been shown to protect free DNA ends from nuclease digestion (49). Such observations have fueled a current view that Ku proteins perform an important role(s) in nonhomologous end joining, the primary mode of DSB repair in mammalian cells (50).
Although not grossly deficient in DNA repair, several studies indicate that A-T cells display faulty DSB repair (51)(52)(53). Given our findings that Ku70 levels are not up-regulated in irradiated A-T cells, we propose that the lower relative abundance of Ku70 in A-T cells following irradiation may contribute to the observed deficiencies in DSB repair in these cells. This potential role for the IR-induced increase in cellular levels of Ku70 is in agreement with our finding that this protein accumulates within the nucleus of irradiated cells. Hence, by up-regulating nuclear Ku70 levels following irradiation the cell may have evolved a mechanism that ensures a sufficient supply of Ku70 to fully facilitate DSB repair following genome damage.
It is becoming apparent that ␥-irradiation results in global changes in transcriptional activity via both p53-dependent and -independent mechanisms (54). Ku has been shown to be a sequence-specific DNA-binding protein (55), and several genes have been identified which contain regions that bind Ku (56 -59). Although little direct evidence indicates that the binding of Ku to these or any other genes has a direct effect on modulating transcriptional activity, these findings do make this a possibility worthy of consideration. Alternatively, Ku70 has been shown to interact with the nuclear tyrosine kinase c-Abl (60) and thus may potentially influence the phosphorylation of the carboxyl-terminal domain of RNA polymerase II by c-Abl. Phosphorylation of the RNA polymerase II COOH-terminal domain has recently been shown to have a strong effect on RNA polymerase II-mediated transcription (61).
In conclusion, we have found that ␥-irradiation results in an up-regulation of Ku70 leading to enhanced levels of Ku70 within the nucleus. We have also provided evidence that the up-regulation of Ku70 utilizes a mechanism that is dependent on both ATM and p53 but that Ku70 up-regulation is not a required event for DNA-PK activation. These findings support the notion that the damage-induced up-regulation of Ku70 provides the cell with a means of assuring either proper DNA repair or appropriate response to DNA damage.