|Home | About | Journals | Submit | Contact Us | Français|
Phosphorylation of the DNA-dependent protein kinase catalytic subunit (DNA-PKcs) upon ionizing radiation (IR) is essential for cellular radioresistance and nonhomologous-end-joining-mediated DNA double-strand break repair. In addition to IR induction, we have previously shown that DNA-PKcs phosphorylation is increased upon camptothecin treatment, which induces replication stress and replication-associated double-strand breaks. To clarify the involvement of DNA-PKcs in this process, we analyzed DNA-PKcs phosphorylation in response to UV irradiation, which causes replication stress and activates ATR (ATM-Rad3-related)/ATM (ataxia-telangiectasia mutated) kinases in a replication-dependent manner. Upon UV irradiation, we observed a rapid DNA-PKcs phosphorylation at T2609 and T2647, but not at S2056, distinct from that induced by IR. UV-induced DNA-PKcs phosphorylation occurs specifically only in replicating cells and is dependent on ATR kinase. Inhibition of ATR activity via caffeine, a dominant-negative kinase-dead mutant, or RNA interference led to the attenuation of UV-induced DNA-PKcs phosphorylation. Furthermore, DNA-PKcs associates with ATR in vivo and is phosphorylated by ATR in vitro, suggesting that DNA-PKcs could be the direct downstream target of ATR. Taken together, these results strongly suggest that DNA-PKcs is required for the cellular response to replication stress and might play an important role in the repair of stalled replication forks.
DNA-dependent protein kinase (DNA-PK), composed of a Ku70/80 heterodimer and a catalytic subunit (DNA-PKcs), is the key component of nonhomologous-end joining (NHEJ), the predominant DNA double-strand break (DSB) repair pathway in mammalian cells. The intrinsic kinase activity of DNA-PKcs is essential for radioresistance and NHEJ-mediated DSB repair (18), most likely through phosphorylation of NHEJ components, including DNA-PKcs itself. DNA-PKcs is rapidly autophosphorylated in vitro upon activation and is phosphorylated in vivo after IR. Many in vitro and in vivo phosphorylation sites of DNA-PKcs have been identified thus far, including the T2609 cluster (7, 11, 32), S2056 (8), and the recently identified C-terminal phosphorylation sites (21). The majority of these phosphorylation sites are the (S/T)Q motifs (serine or threonine followed by a glutamine residue) common in many DNA damage repair proteins and are the cognate substrates of phosphoinositide kinase-related protein kinases (PIKKs), including DNA-PKcs and ATM (ataxia-telangiectasia mutated) and ATR (ATM-Rad3-related) kinases (17, 35). Similar to its kinase activity, DNA-PKcs phosphorylation is also required for NHEJ-mediated DSB repair. Mutations at the T2609 cluster or S2056 severely compromise the ability of DNA-PKcs to restore the radioresistance and DSB repair defects in DNA-PKcs-deficient cells (3, 7, 8, 10).
We demonstrated that among the many DNA-PKcs phosphorylation sites identified, phosphorylation at T2609, S2056, and two additional sites within the T2609 cluster, T2638 and T2647, can be detected in culture cells after IR, and the in vivo phosphorylation is involved with DNA-PKcs itself as well as ATM kinase (7-9). IR-induced S2056 phosphorylation is mediated by a DNA-PKcs autophosphorylation event and detected in cells expressing wild-type DNA-PKcs but diminished in cells expressing kinase-dead mutant DNA-PKcs (8). On the other hand, IR-induced phosphorylation within the T2609 cluster (T2609 and T2647) remained detectible in cells expressing kinase-dead mutant DNA-PKcs but was much reduced in ataxia-telangiectasia cells, suggesting that ATM is likely the main kinase responsible for IR-induced DNA-PKcs phosphorylation at the T2609 cluster (8, 9).
In addition to IR induction, DNA-PKcs phosphorylation could be induced upon treatment with DNA replication-inhibiting agents, including camptothecin, UV light (UV), and hydroxyurea (8). The possible involvement of DNA-PKcs in replication stress responsiveness is also supported by evidence that DNA-PKcs is required for RPA2 (the p34 subunit of replication protein A [RPA]) hyperphosphorylation after DNA damage (4, 6). RPA is a multisubunit single-stranded-DNA-binding protein that is essential for normal DNA replication as well as DNA damage repair. The N terminus of RPA2 becomes hyperphosphorylated upon DNA damage. Consequently, hyperphosphorylation of RPA2 leads to down-regulation in DNA replication but not DNA repair, suggesting that RPA2 hyperphosphorylation could function as a molecular switch to direct RPA activity from DNA replication to DNA damage repair (2). Furthermore, DNA-PKcs is required for cellular resistance (24) and response to UV irradiation (27), which is known to cause replication stress. It was reported that UV-induced replication arrest is normal in DNA-PKcs-proficient M059K cells but attenuated in DNA-PKcs-deficient M059J cells, implying that DNA-PKcs is required for UV-induced replication arrest (27).
In light of this evidence, we hypothesized that DNA-PKcs could play a significant role in the cellular response to replication stress. To test this hypothesis, we investigated whether UV irradiation-generated replication stress could activate DNA-PKcs and induce its phosphorylation. Here we report that, upon UV irradiation, DNA-PKcs is rapidly phosphorylated at both T2609 and T2647 within the T2609 phosphorylation cluster but not at S2056, which is distinct from the response to IR induction. UV-induced DNA-PKcs phosphorylation is also dependent on ATR kinase, the key signaling molecule in response to stalled replication forks and the S-phase checkpoint (1, 14, 30). Taken together, these results strongly support the assertion that DNA-PKcs is involved in DNA repair or signaling in response to replication stress.
All cell lines, including human cervical adenocarcinoma HeLa cells, normal human skin fibroblasts (HSFs), Chinese hamster ovary (CHO) wild-type AA8 and DNA-PKcs-defective V3 cells, and human osteosarcoma U2OS cells expressing inducible wild-type ATR (GW33) or dominant-negative kinase-dead mutant ATR (GK41) were grown in alpha-minimum essential medium supplemented with 10% fetal calf serum and penicillin-streptomycin and were maintained in a humidified atmosphere with 5% CO2. The expression of wild-type and kinase-dead mutant ATR was induced by 1 μg/ml of doxycycline for 2 days, as described previously (25).
For UV irradiation, exponentially growing cells on culture dishes were washed once with phosphate-buffered saline (PBS) and then subjected to UV-C (254 nm) at a rate of 0.5 to 1 J/m2/s to achieve the cumulative desired doses. The culture dishes were replenished with fresh culture medium immediately after irradiation. To generate localized UV exposure, a UV-blocking 3-μm Isopore polycarbonate filter (Millipore) was applied to overlie culture cells prior to UV irradiation (16). IR was carried out in a Mark-II cesium-137 irradiator (JL Shepherd & Associates) at a rate of 57.8 mGy/s to achieve the cumulative desired doses. For wortmannin and caffeine treatments, various concentrations of each were added to cell cultures 1 h prior to UV irradiation.
HSFs were subjected to serum starvation (0.2% fetal calf serum) for 48 h followed by 4 h release in regular medium (10% fetal calf serum) to synchronize in G1 phase. For S-phase synchronization, serum-starved HSFs were released in regular medium for 8 h, treated with aphidicolin (2 μg/ml) overnight, and then released in regular medium for 3 h to synchronize the cell cycle in mid-S phase. For flow cytometry analysis, cells were harvested and fixed in 70% ethanol. Prior to propidium iodide (PI) staining, cells were washed twice with PBS and resuspended in ice-cold hypotonic PI solution (0.1 mg/ml RNase A, 0.1% Triton X-100, 20 mg/ml PI solution, 1 mg/ml trisodium citrate) at a concentration of 1 × 106 cells per ml. The cell suspension was tumbled at 4°C for 30 min in darkness. DNA content was measured by fluorescence-activated cell sorting (FACS) scan, and cell cycle compartments were analyzed by CellQuest software (BD Biosciences).
For immunofluorescence (IF) staining, cells were fixed in 4% paraformaldehyde for 10 min or cold methanol for 20 min, permeabilized in 0.5% Triton X-100 for 10 min, and blocked in 5% normal goat serum or bovine serum albumin for 1 h at room temperature. The cells were incubated with primary antibodies for 1 h, washed three times in PBS, and then incubated with rhodamine red- and Alexa-488-conjugated secondary antibodies for 30 min (Molecular Probes). Cells were then washed three times in PBS and mounted in Vectashield mounting medium with 4,6-diamidino-2-phenylindole (Vector Laboratories). BrdU labeling was carried out with a 5-bromo-2-deoxyuridine labeling and detection kit (Roche). Fluorescence images were captured by using an Olympus BX51 epifluorescence microscope equipped with a MicroPublisher digital charge-coupled device camera and QCapture software (QImaging).
HeLa nuclear extract was prepared as previously described (7). After preadsorption with protein G-agarose (Roche), nuclear extract was incubated with antibody-conjugated resin at 4°C overnight with gentle rotation. The precipitated protein complex was washed five times with lysis buffer (20 mM HEPES [pH 7.5], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, and protease and phosphatase inhibitors) and was subjected to Western blot analysis or an in vitro kinase assay. The ATR in vitro kinase assay was carried out as previously described (34) with [γ-32P]ATP or unlabeled ATP. The recombinant glutathione S-transferase (GST) fusion proteins GST-Xrcc4 (Xrcc4 amino acids 251 to 334) (19), GST-p53 (full length), and GST-PKcs (DNA-PKcs amino acids 2261 to 2700), covering the T2609 cluster in the wild-type sequence (if not indicated) or with alanine substitutions at the T2609 cluster (6A mutant) were expressed and purified from Escherichia coli. The kinase reaction was terminated with sodium dodecyl sulfate (SDS) loading buffer. Substrates were separated by SDS-polyacrylamide gel electrophoresis and were scanned on a Storm PhosphorImager (GE Healthcare) or analyzed by Western blotting.
Phosphospecific anti-DNA-PKcs antibodies (anti-pS2056, anti-pT2609, and anti-pT2647) were described previously (7-9). Anti-DNA-PKcs mouse monoclonal antibody (MAb) (NeoMarkers), anti-γH2AX MAb (Upstate), anti-cyclobutane-pyrimidine dimer (CPD) MAb (MBL International), and anti-ATR (Bethyl Laboratories), anti-CHK1, and anti-CHK2 antibodies (Cell Signaling) are commercially available.
Small inhibitory RNA (siRNA) oligonucleotides designed against ATR (CGAGACUUCUGCGGAUUGCdTdT) or ATM (UGGUGCUAUUUACGGAGCUdTdT) were synthesized (Dharmacon Research Inc.). Transfection of RNA oligonucleotides was carried out with Oligofectamine (Invitrogen) as described previously (12). Seventy-two hours after initial transfection, the transfected cells were subjected to UV irradiation and were harvested for further analysis.
UV sensitivity for each cell line was measured by the colony-forming abilities of irradiated cell populations. Three hundred cells were seeded on 60-mm culture dishes for 4 h prior to UV irradiation (2.5, 5, 7.5, and 10 J/m2). Seven to ten days after UV irradiation, cells were fixed and stained with crystal violet. Colonies containing more than 50 cells were scored, and the mean value for triplicate culture dishes was determined. Cell survival was normalized to plating efficiency of untreated controls for each cell type.
To further investigate the possible role of DNA-PKcs in cellular response to replication stress, we examined whether UV irradiation could induce DNA-PKcs phosphorylation similar to that observed after IR (7-9). Exponentially growing monolayers of HeLa cells were mock treated or subjected to 10 J/m2 of UV irradiation and were harvested at 2 h after UV exposure. As a positive control, HeLa cells were subjected to 10 Gy of IR and were harvested at 1 h after IR. The irradiated and unirradiated samples were analyzed for DNA-PKcs phosphorylation at S2056, T2609, and T2647 by Western blotting. As shown in Fig. Fig.1A,1A, DNA-PKcs phosphorylation at all three sites was low in unirradiated cells and increased significantly after IR. The increase of DNA-PKcs phosphorylation at T2609 and T2647 was also detected after UV irradiation; however, DNA-PKcs phosphorylation at S2056 was not detectible at 2 h after UV irradiation. It is clear that DNA-PKcs phosphorylation can be induced by UV irradiation but in a pattern distinct from that of IR-induced DNA-PKcs phosphorylation (7-9).
Although T2609 and T2647 are both phosphorylated after IR and UV irradiation and are in close proximity to each other within the T2609 cluster of phosphorylation sites, their responses to DNA damage are not identical. The increase of T2647 phosphorylation after UV irradiation is comparable to that induced by IR under the experimental conditions, whereas the increase of T2609 phosphorylation after UV irradiation is much less than that induced by IR (Fig. (Fig.1A).1A). Further kinetics analysis revealed that the increase of T2609 and T2647 phosphorylation was clearly detected at 1 h after UV irradiation (Fig. (Fig.1B).1B). UV-induced T2647 phosphorylation reaches the peak level earlier (at 2 h), while UV-induced T2609 phosphorylation increases progressively and reaches the peak at 16 h after UV irradiation (Fig. (Fig.1B).1B). Taken together, these results suggest that T2647 is the predominant phosphorylation site of DNA-PKcs in response to UV irradiation.
We previously reported that DNA-PKcs phosphorylation in response to IR occurs at concentrated nuclear foci and colocalizes with histone H2AX S139 phosphorylation (γH2AX), presumably at the sites of DNA double-strand breaks (7, 8). We expected that UV-induced DNA-PKcs phosphorylation would also coincide with UV-induced DNA lesions. To test this hypothesis, we used a UV-blocking Isopore polycarbonate filter to localize cells' UV exposure (16) and analyzed whether DNA-PKcs phosphorylation occurs specifically in UV-exposed areas. The UV-blocking polycarbonate filter overlay the monolayer of HeLa cells. The filter-covered cells were subjected to 30 J/m2 of UV irradiation and were fixed at 1 h after UV irradiation for subsequent immunofluorescence staining. The localized UV exposure was confirmed by the positive staining of UV-induced CPD within the nuclei (22). UV-induced DNA-PKcs phosphorylation at T2609 and T2647 coincided precisely with CPD staining (Fig. (Fig.2A),2A), indicating that phosphorylated DNA-PKcs was localized at DNA lesions generated after UV irradiation. In addition, UV-induced T2609 and T2647 phosphorylation colocalized with UV-induced γH2AX at both localized UV lesions (Fig. (Fig.2B)2B) and UV-induced γH2AX foci (Fig. (Fig.2C)2C) (37).
The induction of γH2AX after UV irradiation occurs primarily in S-phase cells in response to UV-induced replication stress (37). To examine whether replication stress in fact induces DNA-PKcs phosphorylation after UV irradiation, we analyzed UV-induced DNA-PKcs phosphorylation in asynchronous cells as well as G1-phase- or S-phase-synchronized cells after UV irradiation. Normal HSFs were serum starved and released to synchronize cell cycles at either G1 or S phase (Fig. (Fig.3A)3A) and were subjected to UV irradiation. Two hours after UV irradiation, the cells were harvested for Western blot analysis of DNA-PKcs phosphorylation at T2647. The result showed that S-phase-synchronized HSFs exhibit a significant increase in T2647 phosphorylation after UV irradiation compared to asynchronous HSFs, whereas G1-phase-synchronized HSFs exhibit only a minimal increase of T2647 phosphorylation after UV irradiation (Fig. (Fig.3B3B).
The correlation between replication stress and UV-induced DNA-PKcs phosphorylation was further examined via bromodeoxyuridine (BrdU) cell labeling and immunofluorescence staining analysis. Exponentially growing HeLa cells were pulse-labeled with BrdU in culture medium for 10 min followed by UV irradiation. Two hours after UV irradiation, the cells were fixed and immunostained for BrdU incorporation and UV-induced DNA-PKcs phosphorylation. Immunofluorescence analysis revealed that UV-induced T2647 and T2609 phosphorylation occurs specifically in BrdU-incorporating or replicating cells but not in BrdU-negative cells (Fig. 3C and D), thus demonstrating that UV-induced DNA-PKcs phosphorylation is indeed associated with replication stress.
In response to replication stress, ATR is the major kinase activated, and it plays an important role in the S-phase checkpoint to prevent cell cycle progression with unrepaired DNA damage (1, 14, 30). It is likely that ATR kinase, but not ATM kinase or DNA-PKcs itself, is responsible for DNA-PKcs phosphorylation at the T2609 cluster upon replication stress. To examine the involvement of all three kinases, we analyzed the sensitivity of UV-induced DNA-PKcs phosphorylation to wortmannin, which preferentially inhibits the kinase activities of ATM and DNA-PKcs (29). We previously reported that IR-induced DNA-PKcs phosphorylation (mediated by DNA-PKcs and ATM) is sensitive to low concentrations of wortmannin treatment (7-9). Contrary to that induced by IR, DNA-PKcs phosphorylation at T2609 and T2647 induced by UV was not affected by increasing concentrations of wortmannin (up to 100 μM [Fig. [Fig.4A4A ]), and neither was ATR-dependent CHK1 phosphorylation at serine 345 (S345). In contrast, ATM-dependent CHK2 phosphorylation at threonine 68 (T68) was much reduced under the influence of wortmannin (Fig. (Fig.4A).4A). These results suggest that the kinase activities of ATM and DNA-PKcs are dispensable for UV-induced DNA-PKcs phosphorylation.
Phosphorylation in DNA-PKcs-deficient CHO V3 cells complemented with wild-type DNA-PKcs (V3-wt) or kinase-dead mutant DNA-PKcs (V3-kd) (18) was also analyzed. We reported previously that IR-induced DNA-PKcs phosphorylation at S2056 was robust in V3-wt cells but was diminished in V3-kd cells, suggesting that DNA-PKcs autophosphorylation is responsible for IR-induced S2056 phosphorylation (8). In contrast, UV-induced DNA-PKcs phosphorylation at T2609 and T2647 was detected in both V3-wt and V3-kd cell lines without significant differences (Fig. (Fig.4B),4B), confirming that the kinase activity of DNA-PKcs is not required for UV-induced DNA-PKcs phosphorylation. This leads to the next query: whether ATR is the kinase responsible for DNA-PKcs phosphorylation.
To further investigate the role of ATR, we analyzed whether caffeine, which inhibits ATR kinase (15, 28), could attenuate UV-induced DNA-PKcs phosphorylation. As shown in Fig. Fig.4C,4C, in the presence of 2 mM caffeine, UV-induced DNA-PKcs phosphorylation at T2609 and T2647 was significantly reduced, whereas the steady-state protein level of DNA-PKcs remained constant. In addition, ATR-dependent CHK1 phosphorylation at S345 was completely blocked by 2 mM caffeine treatment (Fig. (Fig.4C).4C). Taken together, these results consistently demonstrate that ATR is the kinase that mediates DNA-PKcs phosphorylation at the T2609 cluster after UV irradiation.
The ATR dependency of DNA-PKcs phosphorylation was further tested in human osteosarcoma U2OS cells expressing either wild-type ATR (ATR-wt, GW33 cell line) or dominant-negative kinase-dead mutant ATR (ATR-kd, GK41 cell line) through the tetracycline-inducible system (25). The expression of ATR-kd resulted in dominant-negative effects, including inhibition of ATR downstream signaling and disruption of the cell cycle checkpoints (25, 26). Similarly, ATR-kd expression completely blocked UV-induced DNA-PKcs phosphorylation at T2647 in immunofluorescence analysis (Fig. (Fig.5B),5B), whereas ATR-wt expression did not affect UV-induced T2647 phosphorylation (Fig. (Fig.5A).5A). UV-induced DNA-PKcs phosphorylation at T2609, like that at T2647, was also blocked by the expression of ATR-kd (Fig. (Fig.5C5C).
ATR-kd inhibition of UV-induced DNA-PKcs phosphorylation was also determined by Western blot analysis. As shown in Fig. Fig.6A,6A, we observed similar increases of DNA-PKcs phosphorylation at T2647 after IR and after UV irradiation in ATR-wt-expressing U2OS cells. In contrast, UV-induced T2647 phosphorylation was completely blocked in U2OS cells expressing ATR-kd, whereas IR-induced T2647 phosphorylation was not affected in the same cell line (Fig. (Fig.6A),6A), indicating that ATR-kd inhibition of DNA-PKcs phosphorylation is specific and restricted to replication stress.
In addition to the dominant-negative ATR-kd inhibition, ATR-dependent DNA-PKcs phosphorylation was analyzed in HeLa cells by the RNA interference (RNAi) knockdown approach (12). Monolayers of HeLa cells were transfected with double-stranded siRNA oligonucleotides against ATR or ATM. Three days after initial transfection, when siRNA inhibition reached its peak, the transfected HeLa cells were UV irradiated and analyzed by Western blotting for UV-induced DNA-PKcs phosphorylation. Transient transfection of siRNA oligonucleotide against ATR or ATM was effective in reducing the steady-state protein level of both kinases (Fig. (Fig.6B).6B). UV-induced DNA-PKcs phosphorylation at both T2609 and T2647 was significantly reduced in ATR knockdown cells but not in ATM knockdown or mock-transfected cells (Fig. (Fig.6B6B).
Our results consistently point out that ATR is the kinase that mediates DNA-PKcs phosphorylation after UV irradiation. If ATR plays a direct role in phosphorylating DNA-PKcs, we expect that DNA-PKcs could complex or interact with ATR in vivo and that antibodies against either one of them would be able to coimmunoprecipitate both proteins. HeLa nuclear extracts were prepared from unirradiated or UV-irradiated cells and were subjected to immunoprecipitation using control immunoglobulin G (IgG) or specific antibodies against either DNA-PKcs or ATR. In reciprocal coimmunoprecipitation assays (Fig. 7A and B), we were able to bring down both DNA-PKcs and ATR using either anti-DNA-PKcs or anti-ATR antibodies. The coimmunoprecipitation of DNA-PKcs and ATR was not affected by the presence of ethidium bromide, suggesting that their interaction is not mediated through DNA (Fig. S2 in the supplemental material). The interaction between DNA-PKcs and ATR was also detected in both unirradiated and UV-irradiated HeLa nuclear extract without significant differences, indicating that their interaction is not regulated by UV irradiation.
To examine whether ATR could phosphorylate DNA-PKcs at the T2609 cluster directly in vitro, ATR kinase was immunoprecipitated from HeLa nuclear extract and was subjected to in vitro kinase reaction with recombinant GST fusion proteins GST-Xrcc4, GST-p53, and GST-DNA-PKcs (GST-PKcs) covering the entire T2609 cluster region (Fig. (Fig.7C).7C). As shown in Fig. Fig.7C,7C, ATR kinase immunoprecipitated from HeLa nuclear extract was able to phosphorylate both GST-p53 and GST-PKcs but not GST-Xrcc4, a good substrate for the DNA-PKcs in vitro kinase reaction (19). Furthermore, ATR-mediated phosphorylation of GST-PKcs required the T2609 cluster of phosphorylation sites (Fig. (Fig.7D).7D). Alanine substitutions at all six phosphorylation sites within the T2609 cluster (6A mutant) greatly reduced the phosphorylation of GST-PKcs in ATR in vitro kinase reaction (Fig. (Fig.7D).7D). Finally, in vitro ATR-mediated phosphorylation at the T2609 cluster was further confirmed by Western blot analysis using phosphospecific antibodies against T2647 (Fig. (Fig.7E),7E), suggesting that DNA-PKcs could be the direct target of ATR kinase.
DNA-PKcs phosphorylation upon IR is essential not only for DSB repair but also for cellular resistance and survival after IR (7, 8, 10). We speculated that DNA-PKcs phosphorylation after UV irradiation is also required for cellular resistance against UV, as cell lines deficient in DNA-PKcs were UV sensitive (24). To determine whether DNA-PKcs phosphorylation is required for UV resistance, we analyzed the UV sensitivity of CHO V3 cells complemented with human wild-type DNA-PKcs (V3-wt) or mutant DNA-PKcs harboring alanine substitutions at the T2609 cluster of phosphorylation sites (V3-6A) (Fig. (Fig.8A).8A). Alanine substitutions at the T2609 cluster do not affect the kinase activity of DNA-PKcs, as shown previously (10) and confirmed here (see Fig. S1 in the supplemental material). In clonogenic survival analysis, complementation of wild-type DNA-PKcs restored UV resistance in V3 cells, as the survival of V3-wt cells was comparable to that of wild-type CHO AA8 cells (Fig. (Fig.8B).8B). In contrast, V3 complemented with 6A mutant DNA-PKcs remained as UV sensitive as the parental V3 cells, indicating that DNA-PKcs phosphorylation at the T2609 cluster is required for cellular resistance to UV irradiation. In addition to clonogenic survival, we observed that V3 and V3-6A cells exhibit increased and prolonged kinetics of γH2AX relative to V3-wt cells after UV irradiation (Fig. (Fig.8C),8C), suggesting that UV-induced DNA-PKcs phosphorylation is also required for DNA damage repair after UV.
In the present study, we report that DNA-PKcs phosphorylation is induced by replication stress caused by UV irradiation. DNA-PK has been implicated in UV damage response. Cell lines lacking functional components of DNA-PK exhibit elevated sensitivity to UV irradiation compared to their parental cell lines (24). However, nucleotide excision repair (NER), the major repair pathway for removing UV-induced DNA lesions from the genome (23, 33), was not affected in DNA-PK-deficient cells, suggesting that DNA-PK is not directly involved in the repair of UV lesions through the NER pathway (24). Instead, DNA-PK is likely involved in cellular responses to replication stress, which also induces histone H2AX phosphorylation and activation of ATR/ATM kinases after UV irradiation (37, 38). Our results demonstrate that UV-induced DNA-PKcs phosphorylation is indeed associated with replication stress and occurs specifically in replicating cells (Fig. (Fig.3).3). Consistent with this assertion, UV-induced replication arrest is defective in DNA-PKcs-deficient M059J cells but not in DNA-PKcs-proficient M059K cells (27). Our findings that DNA-PKcs is phosphorylated by ATR upon UV irradiation provides further evidence and confirms the direct link of DNA-PKcs in the replication stress response and, perhaps, in ATR-mediated cell cycle checkpoints (1, 14, 30).
In response to replication stress, ATR is the major kinase activated, and it plays a key role in the S-phase checkpoint as well as in resolving stalled replication forks (1, 14, 30). Our results demonstrate that UV-induced DNA-PKcs phosphorylation is indeed mediated by ATR. Inhibition of ATR activity via caffeine treatment, a dominant-negative kinase-dead mutant, and RNAi knockdown all resulted in the reduction of UV-induced DNA-PKcs phosphorylation at the T2609 cluster. Furthermore, ATR is associated with DNA-PKcs and can phosphorylate DNA-PKcs at the T2609 cluster directly in vitro (Fig. (Fig.7),7), indicating that DNA-PKcs could be the direct downstream target of ATR. This is consistent with our recent findings that DNA-PKcs phosphorylation at the T2609 cluster after IR is mediated mainly by ATM but not by DNA-PKcs itself (9). DNA-PKcs, ATM, and ATR are members of the PIKK family of protein kinases and play important roles in the maintenance of genome integrity in mammalian cells (31). All three kinases recognize and phosphorylate the common (S/T)Q motif (17, 35) and share overlapping downstream targets like H2AX and p53. It has also been suggested that there may be coordination and cross-talks among them in response to DNA damage. Our findings that DNA-PKcs phosphorylation at the T2609 cluster is targeted by ATM after IR (9) and by ATR after UV irradiation in the present study provide the direct evidence for such cross-talks and regulations between DNA-PKcs and ATM/ATR kinases. The cross-talks between these kinases may be important for coordinating all the signaling events to ensure proper DNA damage repair, cell cycle checkpoints, apoptosis, and ultimately the integrity of the genome (31).
Although DNA-PKcs phosphorylation at the T2609 cluster is elicited after IR and UV, DNA-PKcs phosphorylation at S2056 is detected only after IR, not after UV, under the experimental conditions (Fig. (Fig.1A).1A). We previously showed that DNA-PKcs phosphorylation at S2056 in response to IR is an autophosphorylation event (8), possibly in trans through two DNA-PKcs molecules in close proximity after synapsis of two DNA ends (39). The lack of S2056 phosphorylation after UV irradiation indicates that there are probably no DSBs generated immediately after UV irradiation. In contrast, treatment with camptothecin, which inhibits topoisomerase I and induces replication-associated DSBs, elicits both S2056 and T2609 phosphorylation (8). Nevertheless, DSB formation occurs after UV irradiation through conversion of UV-induced CPD lesions during DNA replication, and DSBs increase significantly at a later time point after UV irradiation (13). Under similar conditions (8 h after initial UV irradiation), we observed an increase of DNA-PKcs phosphorylation at S2056 (data not shown). The same study concluded that conversion of UV-induced CPD lesions into DSBs during DNA replication may be the principal cause of UV-mediated cytotoxicity (13). This notion is consistent with our results that V3-wt cells are UV resistant while V3-6A cells remain as UV sensitive as the parental V3 cells (Fig. (Fig.8B),8B), as DNA-PKcs phosphorylation at the T2609 cluster is required for DSB repair (3, 7, 8, 10). Furthermore, UV resistance of V3-wt cells also correlates with the reduced level of γH2AX kinetics (Fig. (Fig.8C),8C), suggesting that DNA-PKcs and its phosphorylation at the T2609 cluster are required for DNA damage repair after UV irradiation.
In our model, we hypothesize that there is no DSB formation immediately after UV irradiation. If this model is correct, how is DNA-PKcs recruited to UV lesions or replication stalling sites? And how is DNA-PKcs phosphorylated by ATR? DNA-PKcs was reported to interact directly with RPA1 (p70), the large subunit of RPA (36). It is possible that DNA-PKcs is recruited to replication stalling sites through direct interaction with single-strand-DNA-RPA complexes, similar to the recruitment of the ATR/ATRIP complex (40). Alternatively, DNA-PKcs could be recruited through interaction with other repair molecules like MDC1 (20). Consequently, DNA-PKcs could phosphorylate RPA2 (5) and other molecules to facilitate resolution of the stalled replication forks and, perhaps, the ATR-mediated S-phase checkpoint.
In summary, our studies show that DNA-PKcs phosphorylation is induced after UV irradiation but in a pattern distinct from that induced by IR: DNA-PKcs is rapidly phosphorylated at T2609 and T2647 but not at S2056 after UV irradiation. Our results demonstrate that UV-induced DNA-PKcs phosphorylation occurs specifically in replicating cells and is associated with replication stress. Furthermore, UV-induced DNA-PKcs phosphorylation is dependent on ATR but not ATM or DNA-PKcs. Finally, UV-induced DNA-PKcs phosphorylation is required for both UV resistance and DNA damage repair after UV irradiation.
We thank Paul Nghiem for providing ATR-inducible U2OS cell lines, Junhua Wang for flow cytometry analysis, Elizabeth Miller for editing the manuscript, and David J. Chen and Aroumougame Asaithamby for critical reading of the manuscript. Particularly, we thank David J. Chen for his strong support of this work.
This work was supported by an institutional start-up fund granted to B.P.C.C. and NIH grant CA50519.
Published ahead of print on 14 August 2006.
†Supplemental material for this article may be found at http://mcb.asm.org/.