|Home | About | Journals | Submit | Contact Us | Français|
The catalytic subunit of the DNA-dependent protein kinase (DNA-PKcs) plays a major role in the repair of DNA double-strand breaks (DSBs) by nonhomologous end joining (NHEJ). We have previously shown that DNA-PKcs is autophosphorylated in response to ionizing radiation (IR) and that dephosphorylation by a protein phosphatase 2A (PP2A)-like protein phosphatase (PP2A, PP4, or PP6) regulates the protein kinase activity of DNA-PKcs. Here we report that DNA-PKcs interacts with the catalytic subunits of PP6 (PP6c) and PP2A (PP2Ac), as well as with the PP6 regulatory subunits PP6R1, PP6R2, and PP6R3. Consistent with a role in the DNA damage response, silencing of PP6c by small interfering RNA (siRNA) induced sensitivity to IR and delayed release from the G2/M checkpoint. Furthermore, siRNA silencing of either PP6c or PP6R1 led to sustained phosphorylation of histone H2AX on serine 139 (γ-H2AX) after IR. In contrast, silencing of PP6c did not affect the autophosphorylation of DNA-PKcs on serine 2056 or that of the ataxia-telangiectasia mutated (ATM) protein on serine 1981. We propose that a novel function of DNA-PKcs is to recruit PP6 to sites of DNA damage and that PP6 contributes to the dephosphorylation of γ-H2AX, the dissolution of IR-induced foci, and release from the G2/M checkpoint in vivo.
DNA double-strand breaks (DSBs) are the most cytotoxic form of DNA damage. In human cells there are two main pathways for the repair of DSBs, namely, nonhomologous end joining (NHEJ) and homologous recombination (HR) (reviewed in reference 26). In the initial phase of NHEJ, DSBs are detected by the Ku70/80 heterodimer, which leads to recruitment of the DNA-dependent protein kinase catalytic subunit (DNA-PKcs) and stimulation of its serine/threonine protein kinase activity. Upon autophosphorylation, DNA-PKcs undergoes a conformational change and dissociates from the DSB (25), providing other DNA repair proteins with access to the damage site (reviewed in reference 33). Another physiological substrate of DNA-PK is a histone H2A variant, H2AX. DNA-PKcs and the related protein kinase ATM (ataxia-telangiectasia mutated) both contribute to DNA damage-induced phosphorylation of H2AX on serine 139 to form γ-H2AX (51), which acts as a recruitment platform for MDC1, 53BP1, and other proteins involved in the DNA damage response and cell cycle checkpoint activation (7, 52).
While the effects of phosphorylation on the repair process have been well documented, comparatively little is known about the role of serine/threonine phosphoprotein phosphatases (PPPs) in the DNA damage response. Within the PPP family, the catalytic subunits of PP2A (PP2Ac), PP4 (PP4c), and PP6 (PP6c) are most closely related and form a subgroup referred to as the PP2A-like protein phosphatases (reviewed in reference 40). In vitro, the PP2A-like enzymes display similar sensitivities to small-molecule inhibitors such as okadaic acid and microcystin (27, 45, 53). The specificity of PP2Ac, PP4c, and PP6c function in vivo is derived from a group of regulatory subunits that, with the exception of α4/TAP42 and TIP41, are unique to each enzyme (12, 13, 27, 45, 49). PP2Ac associates with a scaffolding A-α or A-β subunit and additional B-type subunits, while four direct binding partners and several other complex partners unique to PP4c have been characterized (12). The Saccharomyces cerevisiae homologue of PP6c, known as Sit4, interacts with three related proteins: the Sit4-associated proteins SAP155, SAP185, and SAP190, each of which contains a conserved domain known as the SAPs domain (32, 50). The SAPs domain is present in three human orthologues designated PP6R1, PP6R2, and PP6R3, which are therefore considered PP6c regulatory subunits, and each has been shown to bind independently to PP6c (48). More recently, three ankyrin repeat-containing proteins (ARS-A, ARS-B, and ARS-C) were identified as PP6R1 binding partners. One of these, ARS-A, has been shown to dock all three SAPs domain proteins (50), suggesting that, like PP2Ac, PP6c forms stable heterotrimers in vivo and that together these subunits define PP6 function.
We have previously shown that inhibition of PP2A-like protein phosphatase activity by okadaic acid increases the phosphorylation status of DNA-PKcs and decreases its protein kinase activity (20), thus implicating PP2A-like phosphatases in the regulation of DNA-PK activity in vivo. More recently, both PP4 and PP2A have been shown to play roles in the DNA damage response by dephosphorylating γ-H2AX (14, 15, 28, 42). However, the potential role of PP6 in γ-H2AX dephosphorylation has not been addressed.
Here we show that DNA-PKcs interacts with PP2Ac and PP6c, as well as with the PP6c regulatory subunits, PP6R1, PP6R2, and PP6R3. Depletion of PP6c by small interfering RNA (siRNA) induces sensitivity to ionizing radiation (IR) and delayed release from the G2/M checkpoint. Furthermore, siRNA silencing of either PP6c or PP6R1 leads to sustained phosphorylation of γ-H2AX after DNA damage. Together, our studies reveal that a novel and previously unrecognized function of DNA-PKcs may be to recruit PP6 to sites of DNA damage and that PP6 regulates the phosphorylation status of γ-H2AX, the dissolution of IR-induced foci, and release from the G2/M checkpoint.
Microcystin-LR, bovine serum albumin (BSA), phenylmethylsulfonyl fluoride (PMSF), Tris base, EGTA, leupeptin, and pepstatin were purchased from Sigma-Aldrich. The DNA-PK inhibitor (NU7441) and the ATM inhibitor (KU55933) were kind gifts from Graeme Smith and Mark O'Connor (KuDOS Pharmaceuticals Inc., Cambridge, United Kingdom). Antibodies to PP6c, PP4c, PP6R1, PP6R2, PP6R3, TIP41, and TAP42/α4 were from Bethyl Laboratories. The Ku80 antibody was from Calbiochem, and antibodies to glyceraldehyde-3-phosphate dehydrogenase (GAPDH), the phosphospecific antibody to serine 2056 of DNA-PKcs, and an antibody to histone H3 were from Abcam. The antibody to PP2Ac was from Transduction Laboratories, and the DNA-PKcs antibody (DPK1) was made in-house and has been described previously (10). Phosphospecific antibodies to serine 139 of H2AX (γ-H2AX) and serine 10 of histone H3 were from Upstate Biotechnologies, and the antibody to 53BP1 was from Novus Biologicals.
HeLa cells were maintained at 37°C under a humidified atmosphere of 5% CO2 in Dulbecco's modified Eagle medium (DMEM) (Invitrogen) supplemented with 5% (vol/vol) fetal bovine serum (HyClone), 50 U/ml penicillin, and 50 μg/ml streptomycin. HEK293 cells were maintained in DMEM (Invitrogen) supplemented with 10% fetal bovine serum and antibiotics as described above. U2OS human osteosarcoma cells were maintained in RPMI medium (Invitrogen) supplemented with 10% fetal bovine serum and antibiotics as described above.
Where indicated (see Fig. Fig.33 to to8),8), cells were irradiated (in medium plus serum) using a Gammacell 1000 tissue irradiator (MDS Nordion) as a 137Cs source at a dose rate of 3.53 Gy/min and were allowed to recover at 37°C under 5% CO2 for the times indicated.
Cells were washed with phosphate-buffered saline (PBS) and were harvested by trypsinization. For detergent lysis, cells were washed twice in ice-cold PBS and were lysed by incubation on ice for 30 min in ice-cold NET-N lysis buffer (0.15 M NaCl, 0.25 mM EDTA, 50 mM Tris-HCl [pH 8.0], and 1% [vol/vol] NP-40) containing a protein phosphatase inhibitor (1 μM microcystin-LR) and protease inhibitors (0.2 mM PMSF, 1 μg/ml pepstatin, 1 μg/ml aprotinin, and 1 μg/ml leupeptin), followed by sonication on ice with two 5-s bursts. Lysates were prepared by centrifugation at 10,000 × g for 10 min at 4°C. The supernatant was collected and used for immunoblotting or immunoprecipitation as indicated. To the pellet, 1 packed cell volume (PCV) of 1% sodium dodecyl sulfate (SDS) in PBS was added, and the samples were boiled for 5 min. Pellets were sonicated for 10 s and centrifuged at 10,000 × g. The supernatant was removed and used for detection of γ-H2AX by immunoblotting. Detergent lysates were stored in aliquots at −80°C. Protein concentrations were determined using the detergent-compatible dye-binding assay (Bio-Rad) with BSA as the standard.
For the generation of extracts by hypotonic lysis, cells were first harvested by trypsinization, then lysed in hypotonic buffer, and fractionated into cytoplasmic (S10) and nuclear (P10) fractions as described previously (1). Extracts were prepared in the presence of protease inhibitors (as described above) and protein phosphatase inhibitors (10 mM β-glycerol phosphate, 0.2 mM Na3VO4, and 10 mM NaF). Samples were stored in aliquots at −80°C, and protein concentrations were determined by the Bio-Rad protein assay using BSA as the standard.
Fifty micrograms of total protein from either detergent lysates or S10 or P10 extracts was resolved on SDS-polyacrylamide gel electrophoresis (PAGE) gels and transferred to nitrocellulose membranes. Membranes were blocked in 5% (wt/vol) skim milk powder in Tris-bufferd saline (TBS) buffer (20 mM Tris-HCl [pH 7.5], 500 mM NaCl) containing 0.1% (vol/vol) Tween 20 and were probed with antibodies as described. Where indicated (see Fig. Fig.3A3A and 4A and C), blots were scanned and quantitated using Image Quant software, version 5.2 (GE Healthcare).
HEK293 cells were either irradiated or left untreated as described in the figure legends. Extracts were prepared by detergent lysis as described above. Two milligrams of total protein was used per immunoprecipitation reaction. Extracts were precleared by incubation with protein A Sepharose (GE Healthcare) in PBS containing 0.25% (vol/vol) NP-40; then they were incubated overnight at 4°C with 5 μl antisera to PP6c, PP6R1, PP6R2, PP6R3, DNA-PKcs, or control preimmune sera as indicated in the figure legands. Immunoprecipitates were washed once with 1 ml of TBS containing 0.05% (vol/vol) Tween 20, twice with 1 ml of 50 mM HEPES-NaOH (pH 7.5), 40 mM NaCl, 2 mM EDTA, and 1% (vol/vol) Triton X-100, and twice with 1 ml of 50 mM HEPES-NaOH (pH 7.5), 40 mM NaCl, 2 mM EDTA, and 1% (vol/vol) Triton X-100 containing 500 mM LiCl. Beads were boiled in SDS sample buffer, run on SDS-PAGE gels, and immunoblotted for proteins as indicated.
Total-cell extracts were prepared from 4.0 × 108 HeLa cells using a modified NP-40 buffer (50 mM Tris-HCl [pH 8.0], 250 mM NaCl, 5 mM EDTA, and 0.5% [vol/vol] NP-40). Equal amounts of cell extract (80 mg) were incubated either with an affinity-purified rabbit anti-PP6c antibody (BL3704; Bethyl Laboratories) or with rabbit IgG (Bethyl Laboratories) overnight and then with 200 μl of a 50% slurry of protein A Sepharose (GE Healthcare) for 1 h. Immunoprecipitates were washed five times with modified NP-40 buffer, resolved by SDS-PAGE, and stained with Bio-Safe Coomassie stain (Bio-Rad). Discrete gel slices were dissected from the PP6c gel lane. Trypsin digestion of gel slices, mass spectrometry analysis, and database searching were performed by the Mass Spectrometry and Proteomics Core in the City of Hope National Medical Center (Duarte, CA) (http://www.cityofhope.org/research/support/Mass-Spectrometry-and-Proteomics/Pages/default.aspx) using fully automated liquid chromatography-tandem mass spectrometry (LC-MS-MS) analyses combined with protein database search capabilities.
GST-PP6c (NM_002721), GST-PP6R1 (NM_014931), GST-PP6R2 (NM_014678), and GST-PP6R3 (NM_018312) were cloned into the glutathione S-transferase (GST) fusion expression vector pGEX-6P-1 (GE Healthcare), expressed, and purified under the conditions recommended by the manufacturer.
SMARTpool siRNA oligonucleotides were purchased from Dharmacon (Lafayette, CO). HeLa cells were plated in antibiotic-free medium for 24 h prior to transfection. The target siRNA or a scrambled siRNA control (100 nM each) was transfected using Oligofectamine (Invitrogen) according to the manufacturer's instructions. Twenty-four hours after transfection, fresh medium was added, and cells were left for a further 48 h, at which point cells were either left untreated or irradiated and harvested as described above.
Phospho-H3 assays were carried out as described previously (58). Briefly, U2OS cells were transfected with an siRNA to PP6c as described above. Seventy-two hours after transfection, cells were irradiated with 3 Gy IR and were allowed to recover for the times indicated at 37°C under 5% CO2. Cells were then fixed with 0.9% (wt/vol) NaCl-95% (vol/vol) ethanol, resuspended in PBS containing 0.25% (vol/vol) Triton X-100, incubated on ice for 15 min, and then incubated in PBS containing 1% BSA and 75 μg/ml phospho-H3 antibody (Upstate, Billerica, MA) for 3 h. Samples were then incubated for 30 min at room temperature with a fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit antibody (Jackson ImmunoResearch, West Grove, PA) (diluted 1:30 with PBS containing 1% BSA) and were analyzed by flow cytometry using a FACScan flow cytometer (Becton Dickinson, Franklin Lakes, NJ); results were plotted using Modfit by the University of Calgary Flow Cytometry Facility.
Wild-type (wt) DNA-PKcs and the Ku 70/80 heterodimer (Ku) were purified from HeLa cells as described previously (23). The ABCDE/ala mutant (A6-DNA-PKcs) of DNA-PKcs was purified as described previously (36). For autophosphorylation reactions, purified wt DNA-PKcs or A6-DNA-PKcs (5 μg) and the purified Ku70/80 heterodimer (1.6 μg) were incubated in the presence of 50 mM HEPES-NaOH (pH 7.5), 50 mM KCl, 10 mM MgCl2, 0.25 mM ATP or the nonhydrolyzable analogue AMP-PNP, and sonicated calf thymus DNA (at 10 μg/ml) for 60 min at 30°C and were then used in GST pulldown assays.
For GST pulldown assays, 5 μg of purified GST, purified GST-PP6c, GST-PP6R1, GST-PP6R2, or GST-PP6R3 was bound to glutathione-Sepharose beads (GE Healthcare) and incubated either with purified DNA-PKcs alone (see Fig. Fig.1E)1E) or with autophosphorylated or mock-autophosphorylated DNA-PK (prepared by incubation of DNA-PKcs and the Ku heterodimer with DNA and ATP as described above) (see Fig. 2C to E) at 4°C for 2 h with end-over-end rotation. Beads were harvested by centrifugation at 1,000 × g for 1 min and were washed once with 1 ml TBS containing 0.05% (vol/vol) Tween 20, three times with 1 ml 50 mM HEPES-NaOH (pH 7.5), 40 mM NaCl, 2 mM EDTA, and 1% (vol/vol) Triton X-100, and three times with 1 ml 50 mM HEPES-NaOH (pH 7.5), 40 mM NaCl, 2 mM EDTA, 1% (vol/vol) Triton X-100, and 500 mM LiCl. Samples were immunoblotted for DNA-PKcs using the DPK1 antibody.
HeLa cells were exposed to a dose of 20 Gy IR, and neutral comet assays were carried out using the Comet Assay system (Trevigen, Gaithersburg, MD) according to the manufacturer's instructions. Fluorescence images were captured using a Leica DMIRE2 microscope equipped with a digital charge-coupled device (CCD) camera (Hamamatsu, Photonics, K.K.), and the comet tail length (100 cells per condition) was measured using Open Lab software.
HeLa cells were transfected with a siRNA to PP6c or PP6R1 or with a scrambled control as described above. Seventy-two hours after transfection, cells were exposed to various doses of IR and were seeded in triplicate in medium containing 5% (vol/vol) fetal bovine serum. After 14 days, colonies were fixed with methanol, stained with crystal violet, and counted. Results are plotted as the fraction surviving, with the standard error of the mean.
HeLa cells were transfected either with a siRNA to PP6c or PP6R1 or with a scrambled control as described above and were grown on poly-l-lysine-coated coverslips. Cells either were left untreated or were treated with 2 Gy IR and left to recover for the times indicated. Cells were fixed in 3.7% (wt/vol) formaldehyde for 10 min, permeabilized in PBS containing 0.5% (vol/vol) Triton X-100 for 10 min, and then blocked in 1% BSA in PBS for 30 min. Fixed, permeabilized cells were incubated with an anti-γ-H2AX antibody at a 1:1,000 dilution and an anti-53BP1 antibody at a 1:2,000 dilution (in PBS containing 1% BSA) for 1.5 h, washed in 0.05% Tween 20 in PBS, incubated for 30 min with an Alexa Fluor 488-conjugated goat anti-mouse secondary antibody and an Alexa Fluor 594-conjugated goat anti-rabbit secondary antibody (Molecular Probes, Eugene, OR), each at a 1:500 dilution (in PBS containing 1% BSA), and then washed in PBS. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) (1 μg/ml in PBS) for 10 min. Coverslips were mounted in Vectashield (Vector Laboratories Inc., Burlingame, CA). Fluorescence images were captured using a Leica DMIRE2 microscope as described above.
We have previously shown that the phosphorylation status and activity of DNA-PKcs is regulated by a PP2A-like protein phosphatase (20, 21). To further explore which phosphatase or phosphatases might regulate DNA-PK function in vivo, DNA-PKcs was immunoprecipitated from human HEK293 cells, and immunoprecipitates were probed with antibodies to the catalytic subunits of the PP2A-like protein phosphatases, PP2Ac, PP4c, and PP6c. PP2Ac and PP6c were found to immunoprecipitate with DNA-PKcs, whereas PP4 did not (Fig. (Fig.1A).1A). Independently, DNA-PKcs and the known PP6 regulatory subunits PP6R1, PP6R2, and PP6R3 were identified in immunoprecipitates of PP6c using mass spectrometry (Fig. (Fig.1B).1B). The PP6 regulatory subunits PP6R1, PP6R2, and PP6R3 were also detected in immunoprecipitates of DNA-PKcs (Fig. (Fig.1C),1C), and the interaction between DNA-PKcs and PP6 was confirmed in reciprocal immunoprecipitation assays using antibodies to PP6c, PP6R1, PP6R2, or PP6R3 (Fig. (Fig.1D1D).
To determine whether or not the interaction between DNA-PKcs and PP6 was direct, we carried out GST-pulldown assays. PP6c, PP6R1, PP6R2, and PP6R3 were expressed in bacteria as GST fusion proteins, purified (see Fig. S1 in the supplemental material), and incubated with purified DNA-PKcs (in the absence of Ku). Protein complexes were isolated by the addition of glutathione-Sepharose beads, and after a wash, beads were probed for DNA-PKcs by immunoblotting. DNA-PKcs was found to interact directly with PP6c as well as with each of the PP6 regulatory subunits, PP6R1, PP6R2, and PP6R3 (Fig. (Fig.1E1E).
Given that DNA-PKcs interacts with PP6 catalytic and regulatory subunits, we asked whether PP6 was a substrate for DNA-PK in vitro. Purified recombinant PP6c (Fig. (Fig.2A),2A), PP6R1, PP6R2, or PP6R3 (Fig. (Fig.2B)2B) was incubated with purified DNA-PKcs, the Ku heterodimer, and DNA either in the absence or in the presence of 32P-labeled ATP, and phosphorylation was determined by SDS-PAGE and autoradiography. No phosphorylation of PP6c was detected under conditions in which the known DNA-PK substrate (PHAS-I) was highly phosphorylated (Fig. (Fig.2A),2A), indicating that PP6c is not an in vitro substrate of DNA-PKcs. In contrast, PP6R1, PP6R3, and, to a lesser extent, PP6R2 were phosphorylated by DNA-PK in vitro (Fig. (Fig.2B).2B). These data suggested to us that DNA-PK-mediated phosphorylation of PP6 regulatory subunits might regulate DNA-PKcs-PP6 interactions.
We have shown previously that DNA-PKcs is highly autophosphorylated in vitro and that autophosphorylation leads to disruption of the interaction between DNA-PKcs and Ku (10, 19, 25, 37). We therefore asked whether autophosphorylation might affect the interaction between DNA-PKcs and the PP6 subunits. For these experiments, purified DNA-PKcs and the Ku heterodimer were incubated with DNA and purified GST-PP6c, either alone or in the presence of either ATP or the nonhydrolyzable ATP analogue AMP-PNP (Fig. (Fig.2C).2C). Where indicated, the highly selective DNA-PKcs inhibitor NU7441 (60) was added to reaction mixtures either before the addition of ATP (Fig. (Fig.2C,2C, lane 4) or after incubation with ATP (Fig. (Fig.2C,2C, lane 5). GST-PP6c was then pulled down by the addition of glutathione-Sepharose beads, and after a wash, DNA-PKcs was detected by SDS-PAGE and immunoblotting. The presence of ATP in the reaction mixture (i.e., autophosphorylation-permissive conditions) resulted in abrogation of the PP6c-DNA-PKcs interaction (Fig. (Fig.2C,2C, lane 3). In contrast, when the protein kinase activity of DNA-PK was inhibited by NU7441 prior to the addition of ATP (Fig. (Fig.2C,2C, lane 4), or when ATP was replaced with the nonhydrolyzable analogue AMP-PNP (Fig. (Fig.2C,2C, lane 6), the interaction between DNA-PKcs and PP6c was maintain. When NU7441 was added to DNA-PK reaction mixtures after preincubation with ATP, the interaction between DNA-PKcs and PP6c was again disrupted (Fig. (Fig.2C,2C, lane 5). Together, these data indicate that autophosphorylation of DNA-PKcs results in disruption of the DNA-PKcs-PP6c interaction in vitro.
We previously identified multiple in vitro phosphorylation sites in DNA-PKcs, including a cluster between amino acids 2609 and 2647 (also called the ABCDE cluster) (21). This cluster of sites is phosphorylated in vivo in response to DNA damage, and autophosphorylation of these sites is required for release of DNA-PKcs from DSBs in vivo (8, 19, 25, 54; reviewed in reference 33). We therefore asked whether phosphorylation of the ABCDE cluster was required for autophosphorylation-induced release of PP6c from DNA-PKcs. DNA-PKcs containing alanine mutations at the ABCDE sites (A6-DNA-PKcs) was incubated with Ku, DNA, ATP, and GST-tagged PP6c; then GST-PP6c was immobilized on beads, and interaction with DNA-PKcs was determined as for Fig. Fig.2C.2C. A6-DNA-PKcs was found to dissociate from PP6c after autophosphorylation, indicating that the ABCDE sites are not involved in the phosphorylation-induced dissociation of DNA-PKcs from PP6c (Fig. (Fig.2D).2D). Thus, autophosphorylation of DNA-PKcs at sites located outside the ABCDE cluster is involved in the ATP-induced dissociation of PP6c from DNA-PKcs.
We next asked whether phosphorylation affected the interaction between DNA-PKcs and the PP6 regulatory subunits PP6R1, PP6R2, and PP6R3. As described for Fig. 2C and D, DNA-PKcs and Ku were incubated with DNA and GST-tagged PP6R1, PP6R2, and PP6R3 either in the absence or in the presence of ATP. GST-tagged regulatory subunits were then bound to glutathione-Sepharose beads, washed, and probed for the presence of DNA-PKcs as described above. Preincubation of DNA-PKcs and Ku with either PP6R2 or PP6R3 under autophosphorylation-permissive conditions disrupted the interaction with DNA-PKcs (Fig. (Fig.2E,2E, lanes 5 and 7, upper panel), indicating that phosphorylation of DNA-PKcs, PP6R2, or PP6R3 disrupts these interactions. In contrast, although preincubation of PP6R1 with DNA-PKcs, Ku, and ATP results in phosphorylation of PP6R1 (Fig. (Fig.2B,2B, lane 4), the interaction between DNA-PKcs and PP6R1 was maintained (Fig. (Fig.2E,2E, lane 3). As found for PP6c, similar results were observed for A6-DNA-PKcs, indicating that these phosphorylation sites are not involved in the disruption of DNA-PKcs-PP6R2 and DNA-PKcs-PP6R3 complexes (Fig. (Fig.2E,2E, lower panel).
Since DNA-PKcs is a nuclear protein that is involved in, and required for, the cellular response to IR, we asked whether PP6c and its regulatory subunits also localized to the nucleus. Cytoplasmic (S10) and nuclear (P10) extracts were prepared as described in Materials and Methods and were probed for PP6c, PP6R1, PP6R2, and PP6R3 by immunoblotting. Extracts were also probed for GAPDH as a cytoplasmic marker, as well as for DNA-PKcs and Ku80 as loading controls. Consistent with its nuclear function, most of the DNA-PKcs was found in the nuclear (P10) fraction (Fig. (Fig.3A).3A). Although PP6c, PP6R1, PP6R2, and PP6R3 were primarily cytoplasmic (i.e., located in the S10 fraction), small amounts were consistently detected in the nuclear (P10) fraction (Fig. (Fig.3A,3A, lane 4). We also observed a small (1.6- to 2-fold) but reproducible increase in the amount of nuclear PP6c and nuclear PP6R1 in cells following irradiation (Fig. (Fig.3A,3A, lanes 4 to 6, and Fig. Fig.3B).3B). In contrast, IR-induced changes in the levels of nuclear PP6R2 and PP6R3 were less than 1.2-fold (data not shown) and were not considered significant.
We next determined whether the association of DNA-PKcs with PP6 was affected by IR. DNA-PKcs was immunoprecipitated from unirradiated or irradiated cells, and the presence of associated PP6 subunits was detected by immunoblotting. DNA-PKcs was found to coimmunoprecipitate with PP6c, PP6R1, PP6R2, or PP6R3 in both unirradiated and irradiated cells (Fig. (Fig.3C),3C), indicating that the interaction of DNA-PKcs with the PP6 complex is constitutive and is not regulated by exposure to IR. Therefore, the modest increase in nuclear PP6 levels observed after IR (Fig. (Fig.3A)3A) is unlikely to be due to an IR-induced increase in the association of PP6 with DNA-PKcs (to be addressed further in the Discussion).
Given that DNA-PKcs interacts with PP6 and is phosphorylated in response to DNA damage in vivo (8, 11, 19, 36), we next asked whether PP6 might play a role in the dephosphorylation of DNA-PKcs after DNA damage. To address this question, PP6c was first depleted in HeLa cells using siRNA. Levels of PP6c were reduced by approximately 90% compared to the level with a control siRNA (see Fig. S2A in the supplemental material). Seventy-two hours posttransfection, cells were irradiated with 10 Gy, and phosphorylation of DNA-PKcs on serine 2056, a known DNA damage-induced autophosphorylation site (11, 19), was determined by immunoblotting. Depletion of PP6c had no effect on the IR-induced phosphorylation of DNA-PKcs on serine 2056 (see Fig. S3 in the supplemental material), suggesting that PP6 does not regulate IR-induced DNA-PKcs phosphorylation in vivo.
We next asked whether depletion of PP6c affected the phosphorylation of other proteins involved in the DNA damage response. The related protein kinase ATM is rapidly autophosphorylated on serine 1981 in response to DNA damage (4), and autophosphorylation on serine 1981 is considered a marker of ATM activation (4, 30). However, silencing of PP6c did not affect either the kinetics of ATM 1981 phosphorylation or that of the phosphorylation of the ATM targets SMC1 and Chk2 (see Fig. S3 in the supplemental material), suggesting that these proteins are not targets of PP6 in vivo.
Histone H2AX is a variant of nucleosomal histone H2A that is rapidly phosphorylated by ATM and DNA-PK on serine 139 in response to IR and other DNA-damaging agents, creating a form referred to as γ-H2AX (42, 51). Several recent studies have implicated the PP2A-like protein phosphatases PP2A and PP4 in the regulation of γ-H2AX phosphorylation in vivo (14, 15, 28, 42). We therefore asked whether PP6 also plays a role in γ-H2AX dephosphorylation after DNA damage. PP6c was silenced using siRNA (see Fig. S2A in the supplemental material); 72 h after transfection, cells were irradiated (10 Gy), and γ-H2AX phosphorylation was determined by immunoblotting to phosphoserine 139. Significantly, siRNA depletion of PP6c caused sustained phosphorylation of γ-H2AX 2 to 8 h after IR, in contrast to cells transfected with a control siRNA (Fig. 4A and B).
Since other PP2A-like protein phosphatases have been implicated in the dephosphorylation of γ-H2AX (14, 15, 28, 42), it was important to determine whether silencing of PP6c affected the levels of PP2Ac and PP4c. Extracts from PP6c-depleted cells were therefore probed for the catalytic and regulatory subunits of other PP2A-like phosphatases as well as for the PP2A-like phosphatase regulatory proteins TIP41 and α4/TAP42 (29). Knockdown of PP6c had no effect on the levels of any of the proteins tested (see Fig. S4 in the supplemental material).
As mentioned above, both DNA-PKcs and ATM contribute to DNA damage-induced phosphorylation of γ-H2AX (42, 51). In addition, PP2A-type protein phosphatases have been linked to the regulation of ATM (22) and DNA-PK kinase activity (20). Although we did not see an effect of PP6c silencing on the autophosphorylation of either DNA-PK or ATM (see Fig. S3 in the supplemental material), it remained possible that PP6c was acting on γ-H2AX indirectly by promoting the activity of ATM or DNA-PK. To eliminate this possibility, we examined the turnover of γ-H2AX after inhibition of ATM and/or DNA-PK using the small-molecule inhibitor KU55933 or NU7441, respectively. We first examined the relative contributions of ATM and DNA-PKcs to γ-H2AX in HeLa cells. Consistent with earlier reports (42, 51), inhibition of either ATM or DNA-PK reduced IR-induced phosphorylation of H2AX, while inhibition of both together effectively eliminated γ-H2AX formation (see Fig. S5A in the supplemental material). We next examined the effects of ATM/DNA-PK inhibition on H2AX phosphorylation in cells depleted of PP6c. PP6c was depleted by siRNA; then cells were irradiated to induce γ-H2AX phosphorylation. Sixty minutes post-IR, cells were incubated with both KU55933 and NU7441 to inhibit further phosphorylation of H2AX, and the decay of γ-H2AX phosphorylation was monitored by immunoblotting. Depletion of PP6c resulted in slower kinetics of γ-H2AX dephosphorylation (see Fig. S5B in the supplemental material), indicating that PP6c contributes to the dephosphorylation of γ-H2AX in vivo, independently of any effects on ATM and DNA-PK activity.
To further explore the effects of PP6 on H2AX phosphorylation, we transiently depleted PP6R1, PP6R2, or PP6R3 using siRNA (see Fig. S2B to D in the supplemental material). Silencing of PP6R1, PP6R2, or PP6R3 was specific for the target protein and had no effect on the levels of other PP6 subunits or PP2A-like phosphatases (see Fig. S6 to S8 in the supplemental material). siRNA-depleted cells were irradiated, and H2AX phosphorylation was examined by immunoblotting. Silencing of PP6R1 resulted in an increase in H2AX serine 139 phosphorylation, in particular at early times after IR, as well as in sustained phosphorylation of H2AX at 4 and 8 h post-IR (Fig. 4C and D; see also Fig. S6 in the supplemental material). In contrast, silencing of PP6R2 or PP6R3 had no effect on H2AX phosphorylation (see Fig. S7 and S8 in the supplemental material). Thus, PP6c and PP6R1, but not PP6R2 and PP6R3, regulate γ-H2AX phosphorylation in response to IR in vivo. Given that silencing of PP6R2 and PP6R3 had no effect on the kinetics of γ-H2AX dephosphorylation, for subsequent experiments we concentrated on PP6c and PP6R1 only.
We next examined the effects of PP6c and PP6R1 depletion on γ-H2AX and 53BP1 focus formation using immunofluorescence. As found by immunoblotting, silencing of PP6c resulted in persistence of γ-H2AX and 53BP1 foci (Fig. (Fig.5).5). Surprisingly, however, and in contrast to the results obtained by immunoblotting, silencing of PP6R1 did not result in the persistence of either γH2AX or 53BP1 foci (Fig. (Fig.5).5). We speculate that these results suggest that PP6c complexes composed of PP6c with different regulatory subunits may dephosphorylate different populations of H2AX in the nucleus (to be addressed further in the Discussion).
Since cells with DSB repair defects also display sustained γ-H2AX phosphorylation (46, 59), we considered it possible that sustained phosphorylation of γ-H2AX in PP6c- or PP6R1-depleted cells might result from a direct defect in DSB repair. We therefore determined whether silencing of PP6c or PP6R1 induced DNA damage by using neutral comet assays. The kinetics of DSB formation and repair were found to be similar in cells in which either PP6c or PP6R1 had been depleted (Fig. (Fig.6),6), suggesting that the sustained phosphorylation of γ-H2AX is due to a defect in PP6-mediated dephosphorylation and is not due to an inability to repair DNA DSBs.
To explore further a potential role for PP6 in the DNA damage response, we asked whether silencing of PP6c or PP6R1 conferred sensitivity to IR. PP6c or PP6R1 was silenced as described above (see Fig. S2 in the supplemental material), and 72 h after transfection, cells were irradiated with 0.5, 2, or 4 Gy. Radiation sensitivity was determined using the colony formation assay. Silencing of PP6c resulted in decreased survival after 2 Gy IR (approximately 25% survival, compared with 50% for cells transfected with a scrambled siRNA). After 4 Gy IR, fewer than 3% of PP6c-depleted cells formed colonies, compared to more than 18% of control cells (Fig. (Fig.7).7). In contrast, depletion of PP6R1 did not enhance radiation sensitivity over that of control cells (Fig. (Fig.7).7). These results further support a role for PP6c in the DNA damage response but again point to subtle differences between the roles of PP6c and PP6R1.
siRNA silencing of PP4, which also contributes to the dephosphorylation of γ-H2AX in vivo, causes delayed release from the G2/M checkpoint after IR (42). We therefore determined whether, like that of PP4, silencing of PP6c also affected the G2/M checkpoint. PP6c was depleted in U2OS cells; then cells were irradiated with 3 Gy IR, and the G2/M checkpoint function was determined by monitoring histone H3 phosphorylation on serine 10. Silencing of PP6c had no effect on the initiation of the G2/M checkpoint; cells treated either with a siRNA to PP6c or with a scrambled control arrested 1 h after 3 Gy IR. In contrast, knockdown of PP6c resulted in a significant delay in release from G2, as indicated by reduced numbers of cells entering mitosis 24 h post-IR (Fig. (Fig.8).8). Together, our results reveal that PP6 is required for the dephosphorylation of γ-H2AX, the dissolution of 53BP1 foci, and release from the G2/M checkpoint.
Here we show that DNA-PKcs coprecipitates with the catalytic subunit of PP6 (PP6c) as well as with the PP6 regulatory subunits, PP6R1, PP6R2, and PP6R3. Moreover, the interaction between DNA-PKcs and each of the PP6 subunits (PP6c, PP6R1, PP6R2, and PP6R3) is direct and occurs independently. These results confirm and extend those of Larner and colleagues, who recently reported that DNA-PKcs interacts with PP6c and PP6R1 (38). We also show that DNA-PKcs immunoprecipitates with the catalytic subunit of PP2A (PP2Ac). Interestingly, PP2A was recently shown to interact with Ku and to regulate NHEJ through direct dephosphorylation of DNA-PKcs (56), providing a possible explanation for our earlier observations that dephosphorylation of DNA-PKcs by a PP2A-like phosphatase activates its protein kinase activity (20).
Consistent with a role in the DNA damage response, we show that although the majority of PP6c, PP6R1, PP6R2, and PP6R3 is cytoplasmic, a fraction of each protein is located in the nuclear fraction. Indeed, although PP6c, PP6R1, PP6R2, and PP6R3 are considered to be mainly cytosolic proteins (48), other studies have also indicated that PP6 is present in the nucleus (47). We observed modest increases in the levels of total PP6c and PP6R1 in the nuclear fraction after IR; however, these increases did not appear to coincide with increased association of PP6 subunits with DNA-PKcs in immunoprecipitation assays. Interestingly, Larner and colleagues recently reported that IR induces a small increase in the nuclear accumulation of PP6R1 and that this is dependent on DNA-PKcs (38). We speculate that since DNA-PKcs and PP6 are abundant proteins in the cell, small changes in their association might be below the limits of detection by immunoprecipitation.
We and others have previously shown that DNA-PKcs is autophosphorylated in vivo in response to DNA damage and that DNA-PKcs autophosphorylation is important for its function in NHEJ in vivo (8, 19, 25, 54; reviewed in reference 33). Since DNA-PKcs interacts directly with PP6, we initially hypothesized that PP6 might act to dephosphorylate and thus activate DNA-PKcs; however, when PP6c was silenced using siRNA, IR-induced phosphorylation of DNA-PKcs on Ser-2056 was not affected. Similarly, Larner and colleagues reported that depletion of either PP6c or PP6R1 has no effect on DNA-PKcs phosphorylation at either Ser-2056 or Thr-2609 (38). It therefore seems unlikely that PP6 directly regulates DNA-PKcs phosphorylation in response to DNA damage; however, we cannot exclude the possibility that PP6 regulates the phosphorylation of additional DNA-PKcs sites in vivo.
DNA damage results in transient recruitment of DNA-PKcs to the DSB, where it undergoes rapid autophosphorylation and then release, thus facilitating pathway progression (25, 54; reviewed in reference 33). Our results suggest that phosphorylation by DNA-PK may also regulate interactions of the DNA-PK-PP6 complex. In vitro, autophosphorylation of DNA-PKcs resulted in disruption of the interaction between DNA-PKcs and PP6c, PP6R2, or PP6R3, but not PP6R1. Dissociation of DNA-PKcs from the PP6 complex involved DNA-PKcs autophosphorylation events outside the ABCDE cluster, at sites that remain to be identified. The phosphorylation-induced dissociation of DNA-PKcs and PP6 subunits was not reflected in immunoprecipitation assays, however, perhaps because the amount of DNA-PKcs actually associated with DSBs in vivo would be a very small fraction of the total DNA-PKcs population and therefore would be difficult to measure by immunoprecipitation.
We also show that PP6R1, PP6R3, and, to a lesser extent, PP6R2 are phosphorylated by DNA-PKcs in vitro. Interestingly, proteomics screens have shown that PP6R1, PP6R2, and PP6R3 are phosphorylated in vivo (5, 6, 17, 18, 34, 43). Indeed, of the five in vivo phosphorylation sites identified in PP6R3, two (Thr-631 and Ser-634) were identified as DNA damage-induced phospho-SQ/TQ sites and are therefore potential targets for DNA-PK or the related protein kinases ATM and ATR in vivo (34). Together, these studies suggest that DNA damage-induced phosphorylation of PP6 subunits may regulate PP6 function in vivo.
We also examined whether PP6 might regulate the phosphorylation of other proteins involved in the DNA damage response. Silencing of PP6c had no effect on the phosphorylation of ATM on Ser-1981, of SMC1 on Ser-957, or of Chk2 on Thr-68; however, silencing of either PP6c or PP6R1 caused increased and sustained phosphorylation of total γ-H2AX by immunoblot analysis. We also analyzed γ-H2AX and 53BP1 focus formation following either PP6c or PP6R1 silencing. Interestingly, silencing of PP6c led to a significant increase in the number of both γ-H2AX and 53BP1 foci remaining 8 h after IR, whereas in PP6R1-silenced cells, the number of H2AX foci was the same as that in control cells. Although several interpretations of these results are possible, one possible explanation is that PP6c may target different populations of γ-H2AX depending on the regulatory subunit with which it is associated. As we have shown, DNA-PKcs can interact directly with either PP6c, PP6R1, PP6R2, or PP6R3, and PP6c can itself interact directly and independently with PP6R1, PP6R2, or PP6R3 (48); thus, DNA-PKcs may recruit multiple combinations of PP6 complexes to the break site, and these complexes may have different properties in the cell (Fig. (Fig.9).9). Indeed, in budding yeast, the regulatory subunits SAP1, SAP2, and SAP3 have each been shown to confer different properties on Sit4, the yeast homologue of PP6c (41). We speculate that one possible function of the different PP6 complexes may be to target γ-H2AX at foci that are located at different distances from the DSB site. Alternatively, different complexes may dephosphorylate H2AX located in areas enriched for either euchromatin or heterochromatin (24). It is also possible that PP6 complexes may dephosphorylate γ-H2AX that is present in the nucleoplasm rather than associated with chromatin (Fig. (Fig.9).9). Indeed, this might provide a possible explanation for the different effects of PP6R1 depletion on γ-H2AX dephosphorylation in focus assays compared to immunoblot assays (Fig. (Fig.44 and and5).5). Different PP6 complexes may also be involved in dephosphorylating γ-H2AX induced by different forms of DNA damage, as has previously been suggested for PP4 and PP2A (15, 42). We also note that although this study focused on γ-H2AX dephosphorylation, it is possible that PP6 regulates the phosphorylation of other proteins involved in the DNA damage response.
Since cells that contain defects in NHEJ components such as DNA ligase IV (46) display elevated levels of γ-H2AX foci after IR due to increased numbers of unresolved DSBs, we tested whether knockdown of PP6c or PP6R1 induced a DNA repair defect. Using the neutral comet assay, we found that the repair of breaks in the presence of a siRNA to PP6c or PP6R1 was as efficient as in the siRNA control cells, suggesting that PP6 depletion does not result in impaired repair of DSBs. However, we caution that limitations of the comet assay, for example, its inability to detect low levels of DNA damage (discussed in reference 35), could have precluded detection of some forms of DNA damage, and we cannot exclude the possibility that knockdown of PP6 produces more-subtle effects on DSB repair that were not apparent in this assay. Further work will be required to explore this possibility.
To further explore the role of PP6 in the DNA damage response, we examined sensitivity to IR in cells that had been depleted of PP6c. The fraction surviving after 2 Gy irradiation (SF2) was approximately 0.25, compared with 0.5 for cells transfected with a scrambled siRNA, while after 4 Gy, the fraction surviving was 0.03, compared to 0.18 for control cells. The degree of radiosensitivity is thus less than that reported for NHEJ-deficient human cell lines such as the DNA-PKcs/ATM-deficient cell line M059J (9, 31), for which the SF2 is 0.02 (3), or for DNA ligase IV- or XLF-deficient human fibroblasts, which also have an SF2 of less than 0.03 (estimated from reference 16). However, the radiosensitivity of human cells depleted of PP6c by siRNA, shown here, is similar to that reported for siRNA depletion of either XRCC4 or XLF in human cells, for which an SF2 of approximately 0.3 has been reported (calculated from reference 2). Thus, the observed radiation sensitivity of PP6c-silenced human cells is similar to that observed for human cells in which other NHEJ factors have been silenced. In contrast, silencing of PP6R1 did not result in a level of radiation sensitivity higher than that of cells transfected with a scrambled control siRNA. As discussed above, our results suggest that PP6c and PP6R1 have subtly different roles in the DNA damage response. In addition, we note that whereas PP6c silencing would result in almost complete loss of PP6 function, knockdown of PP6R1 would eliminate only the specific roles of the PP6c-PP6R1 complex. Moreover, it is possible that some of the functions of the PP6c-PP6R1 complex affected by knockdown of PP6R1 could be compensated for by other PP6-targeting subunits, such as PP6R2 and PP6R3.
Recently, PP4 was shown to dephosphorylate γ-H2AX and to be required for release from G2 after IR (42). We therefore examined the effects of PP6c depletion on cell cycle checkpoint function. Significantly, we found that depletion of PP6c also resulted in delayed release from G2 after IR. Taken together, our studies suggest that DNA-PKcs recruits PP6 complexes to DSBs, where PP6 contributes to the dephosphorylation of γ-H2AX, the dissolution of foci, and release from the G2/M checkpoint (Fig. (Fig.9).9). However, it remains possible that the effects of PP6 depletion may not be a consequence of loss of the interaction between DNA-PKcs and PP6 but rather could be due to the disruption of other functions of PP6. Further work will be required to distinguish between these possibilities.
In summary, our studies reveal a new role for PP6 in γ-H2AX dephosphorylation after IR. Since the other members of the PP2A-like protein phosphatase family, PP2A and PP4, also regulate DNA damage-induced phosphorylation of γ-H2AX (14, 15, 28, 42), we speculate that the different PP2A-like protein phosphatases might target different populations of γ-H2AX, perhaps at different distances from the break site or within different nuclear subfractions, such as euchromatin versus heterochromatin (24). Alternatively, different PP2A-like phosphatases may be involved in γ-H2AX dephosphorylation depending on the degree or type of DNA insult or the stage of the cell cycle at which the damage was incurred. Thus, the three PP2A-like phosphatases may have distinct roles in γ-H2AX dephosphorylation.
Finally, we note that DNA-PKcs and/or the DNA-PK complex has been reported to interact with several protein phosphatases, including PP6 (this study and reference 38), PP2A (this study and reference 56), PP5 (57), and PP1c (39). Together, these studies raise the interesting possibility that DNA-PKcs is involved in the recruitment of multiple protein phosphatases to sites of DNA damage. It is well established that protein phosphatase regulatory subunits play a critical role in targeting protein phosphatases to their substrates, and HEAT repeats have been shown to be important in the interaction of the scaffolding subunit of PP2A (PR65) with both the catalytic and regulatory B-subunits (55). Since the N-terminal region of DNA-PKcs is also composed of a series of HEAT repeats (44), it is tempting to speculate that the HEAT repeats of DNA-PKcs may also function as phosphatase docking or recruitment sites. Thus, we speculate that DNA-PKcs might act as a potential regulatory subunit for multiple protein phosphatases involved in the DNA damage response.
We thank S. Fang for excellent technical assistance, C. Williamson and L. Williamson for help with the checkpoint assays and statistical analysis, respectively, B. Mahaney for help with figures, the University of Calgary Flow Cytometry Facility for FACS analysis, E. W. McIntush (Bethyl Laboratories) for providing the PP6 antibodies for testing, G. Smith and M. O'Connor (KuDOS Pharmaceuticals, Inc.) for NU7441 and KU55933, and E. Kurz, J. Cobb, and members of the Lees-Miller and Xu laboratories for helpful discussions and suggestions.
This work was supported by grant MOP-13639 from the Canadian Institutes of Health Research (CIHR) to S.P.L.-M., grant CCI-85678 from the CIHR China-Canada Joint Health Research Initiative to S.P.L.-M. and X.X., and funds from the Natural Science Foundation of China (30570371, 90608014, and 30711120570), the Program for New Century Excellent Talents in University (NCET-06-0187), the Beijing Natural Science Foundation Program and Scientific Research Key Program of the Beijing Municipal Commission of Education (KZ200810028014), and the Funding Project for Academic Human Resources Development in Institutions of Higher Learning under the Jurisdiction of Beijing Municipality (PHR[IHLB]) to X.X. S.P.L.-M. is a Scientist of the Alberta Heritage Foundation for Medical Research and holds the Engineered Air Chair in Cancer Research at the University of Calgary.
We declare that we have no conflict of interest.
Published ahead of print on 11 January 2010.
†Supplemental material for this article may be found at http://mcb.asm.org/.