Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2012 May 16;31(10):2403-15.
doi: 10.1038/emboj.2012.86. Epub 2012 Apr 10.

Phosphoproteomic Analysis Reveals That PP4 Dephosphorylates KAP-1 Impacting the DNA Damage Response

Affiliations
Free PMC article

Phosphoproteomic Analysis Reveals That PP4 Dephosphorylates KAP-1 Impacting the DNA Damage Response

Dong-Hyun Lee et al. EMBO J. .
Free PMC article

Abstract

Protein phosphatase PP4C has been implicated in the DNA damage response (DDR), but its substrates in DDR remain largely unknown. We devised a novel proteomic strategy for systematic identification of proteins dephosphorylated by PP4C and identified KRAB-domain-associated protein 1 (KAP-1) as a substrate. Ionizing radiation leads to phosphorylation of KAP-1 at S824 (via ATM) and at S473 (via CHK2). A PP4C/R3β complex interacts with KAP-1 and silencing this complex leads to persistence of phospho-S824 and phospho-S473. We identify a new role for KAP-1 in DDR by showing that phosphorylation of S473 impacts the G2/M checkpoint. Depletion of PP4R3β or expression of the phosphomimetic KAP-1 S473 mutant (S473D) leads to a prolonged G2/M checkpoint. Phosphorylation of S824 is necessary for repair of heterochromatic DNA lesions and similar to cells expressing phosphomimetic KAP-1 S824 mutant (S824D), or PP4R3β-silenced cells, display prolonged relaxation of chromatin with release of chromatin remodelling protein CHD3. Our results define a new role for PP4-mediated dephosphorylation in the DDR, including the regulation of a previously undescribed function of KAP-1 in checkpoint response.

Conflict of interest statement

The authors declare that they have no conflict of interest.

Figures

Figure 1
Figure 1
PP4C influences the phosphorylation status of multiple DDR proteins. (A) Schematic for phosphoproteomics-based identification of putative PP4C substrates. (B) Identification of regulated phosphorylation in response to PP4C depletion. The geometric mean of iTRAQ spectral peak height ratios for replicate samples was plotted as a function of the sum of the geometric means of all iTRAQ spectral peak heights. Maximum approximate conditional likelihood (MACL) was used to determine an intensity-based variance function from which the 95% acceptance region was calculated (grey curve and Supplementary Table 1). Several DDR proteins, including KAP1, CHD4 and TP53BP1, exhibited hyperphosphorylation in response to deplection of PP4C. (C) Validation of the PP4 targets. (Left panel) HeLa S3 cells transfected with PP4C or scrambled siRNAs were exposed to IR, lysed after 2 h and immunoprecipitated (IP) using a pan-phosphoSer antibody and probed for DDR proteins as indicated. The relative band intensities are provided below each immunoblot. (Right panel) Proteins identified as putative PP4 substrates were confirmed using Phos–tag. HeLa S3 cells were transfected with PP4C siRNA and irradiated with 10-Gy IR. Lysates were subjected to SDS–PAGE containing 20 μM Phos–tag and immunobloted with indicated antibodies. Lysates treated with λ protein phosphatase (λPP) served as control for the Phos–tag-induced mobility shift.
Figure 2
Figure 2
KAP-1 phosphorylation on S824 is regulated by a PP4C–R3β complex. (A) In unperturbed cells, phosphorylation of KAP-1 on S824 is elevated in PP4-depleted cells. HeLa cells transfected with siRNAs against indicated PP4 subunits were harvested after 72 h and immunoblotting was performed with the indicated antibodies. (B) HeLa cells transfected with siRNAs against PP4C or PP4R3β were irradiated and harvested at the indicated times and p-KAP-1 was assessed by immunoblot using phospho-KAP-1 antibody (phosphoSerine 824). The kinetics of pS824-KAP-1 formation was monitored after irradiation by loading 20 μg of cell lysate per lane. (C) PP4R3β depletion attenuates pS824-KAP-1 turnover after IR. Primary human fibroblasts were transfected with control or PP4R3β siRNAs. After 72 h, cells were irradiated, fixed at the indicated times and immunostained for pS824-KAP1 (red), γH2AX (green) and DAPI (blue). Right Panel: The average pS824-KAP-1 signal intensity per nucleus was quantified using ImageJ software. Data represent average and s.d. of three independent experiments. (D) PP4C/PP4R3β interacts with KAP-1. HeLa S3 cells stably expressing empty vector (FH-Ctrl), FH-tagged KAP-1 (top and bottom panel), and PP4C or PP4R3β (middle panel) were subjected to immunoprecipitation using anti-FLAG beads at indicated time points before or after IR. PP4R3β or control siRNAs were transfected in cells 72 h prior to immunoprecipitation (right panel, bottom). The immunoprecipitate was probed with antibodies against endogenous KAP-1, PP4C and PP4R3β as indicated. Figure source data can be found with the Supplementary data.
Figure 3
Figure 3
Hyperphosphorylation of S824-KAP-1 in PP4-silenced cells is not due to ectopic activation of ATM. (A) pS824-KAP-1 is lost rapidly in the absence of ongoing ATM activity. (Upper panel) 1BR3 primary fibroblasts were irradiated with 3-Gy IR and, 0.5 h later, were incubated with either DMSO or 10 μM of ATMi and harvested at the indicated times. Cells were fixed and immunostained for pS824-KAP-1 (red), γH2AX (green) and DAPI (blue). (Lowerpanel) GM02188 lymphoblastoid cells were irradiated with 10-Gy IR and, 20 min later, were incubated with DMSO or ATMi. Cells were harvested at the indicated times, processed and immunoblotted for pS824-KAP-1 and total KAP-1 (top panel). Immunoblot signal was quantified and plotted (lower panel). The half-lives of signal decay were calculated. (B) PP4C/PP4R3β depletion attenuates pS824-KAP-1 turnover after IR independent of ATM activity. HeLa cells were transfected with siRNAs. After 72 h, cells were irradiated with 10-Gy IR. At 0.5 h post IR, DMSO or ATMi was added. Cells were harvested at the indicated times and immunoblotting was performed and band intensities quantified using the Odyssey Infrared Imaging System. The intensity of pS824-KAP-1 was normalized relative to total KAP-1 and represented graphically. (C) PP4R3β depletion attenuates pS824-KAP-1 turnover after IR independent of ATM activity. Primary human fibroblasts were transfected with siRNAs. After 72 h, cells were irradiated with 3-Gy IR. At 0.5 h post IR, DMSO or ATMi was added. Cells were fixed at the indicated times and immunostained for pS824-KAP1 (red), γH2AX (green) and DAPI (not shown). The data were quantified as in Figure 2C. (D) PP4R3β regulates pS824-KAP-1 at late-repairing IRIF. Primary human fibroblasts were transfected with control or PP4R3β siRNA. After 72 h, cells were irradiated with 8-Gy IR. After 24 h, ATMi was added (t=0 min) and cells were fixed at the indicated times up to 120 min. Fixed cells were then immunostained for pS824-KAP-1 (red), γH2AX (green) and DAPI (blue). The average pS824KAP-1 signal intensity (from (A)) overlapping with γH2AX foci (i.e., ‘at IRIF') was quantified using ImageJ software. Data represent average and s.d. of three independent experiments. The approximate half-life (t1/2) of pS824-KAP-1 signal following the addition of ATMi is indicated. Data represent average and s.d. of three independent experiments. Figure source data can be found with the Supplementary data.
Figure 4
Figure 4
PP4-mediated dephosphorylation of KAP-1 on S824 impacts its role in the DNA damage response. (A) PP4-mediated regulation of KAP-1 impacts the expression of Gadd45α and p21. HeLa cells were transfected with PP4R3β siRNA, or endogenous KAP-1 was replaced with WT, mutant forms A (KAP1-S824A) or D (KAP1-S824D). After IR, cells were harvested and RNA purified at indicated times and quantitative real-time PCR (qRT–PCR) was performed. Data represent average and s.d. of four independent experiments. (B, C) PP4R3β depletion and KAP1 phosphomimetic mutant (S824D) prolong damage-induced chromatin relaxation. HeLa cells were transfected with control or PP4R3β siRNA, or endogenous KAP-1 was replaced with WT or KAP1-S824D mutant (D). Cells were irradiated and harvested at indicated times. Nuclei were purified, treated with micrococcal nuclease; DNA isolated and analyzed as described in experimental procedures. The relative intensity of each nucleosomic form (mono, bi, tri or poly) is expressed as the percentage of the total signal (for a given lane). Bar graphs represent average and s.d. of three independent experiments. The KAP-1 replacement has been shown in the immunoblot with endogenous and Myc-tagged KAP-1 indicated. (D) The ‘return' of CHD3 to sites of slow-repairing DSBs following ATMi addition is significantly attenuated in PP4R3β-depleted cells. Primary human fibroblasts were transfected with either control or PP4R3β siRNA. After 72 h, cells were irradiated with 8-Gy IR. At 24 h post IR, cells were treated with ATMi and harvested 10, 30, 60 and 120 min later. To examine the retention of CHD3 at sites of late-repairing damage, cells were pre-extracted with PBS containing 0.1% (v/v) Triton × 100 for 30 s before being fixed and immunostained for CHD3, γH2AX and DAPI (left panel). The signal intensity of CHD3 at regions of γH2AX foci (as determined by computer analysis) was measured (∼200 foci per sample). The data represent the mean and s.d. of multiple experiments. siRNA-mediated knock-down efficiency was independently verified for each experiment and was >80% (of cells with good knock-down) (right panel). It shows representative images at 60 min time point after ATMi addition from quantified data. In the zoomed-in and three-dimensional plotted images, note the relative difference in red–green overlap (yellow signal), indicative of changes in CHD3 abundance at sites of ongoing DSB repair, following the addition of ATMi for 1 h.
Figure 5
Figure 5
IR-induced phosphorylation of KAP1 on S473 is regulated by a combination of ATM/CHK2 and PP4C/R3β. (A) Alignment of region surrounding S473 of KAP1 across different organisms. Sequence alignment was performed with ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2) and consensus sequence is shown in color-shaded boxes (Upper panel). Phosphorylation of KAP-1 on S473 after IR is CHK2-mediated. HeLa cells after transfection with CHK1 or CHK2 siRNAs were irradiated after 3 days and harvested at indicated times (lower panel). Immunoblotting was performed with the indicated antibodies. (B) HeLa cells transfected with siRNAs for PP4C or PP4R3β were irradiated and harvested at the indicated times and pS473-KAP-1 was assessed by immunoblot using phospho-KAP-1 antibody (phosphoSerine 473). (C) PP4R3β depletion attenuates pS473-KAP-1 turnover after IR independent of Chk2 or ATM activity. HeLa cells transfected with siRNAs were irradiated after 3 days. At 0.5 h post IR, Chk2i (left panel) and ATMi (right panel) were added. Cells were harvested at the indicated times and immunoblotting was performed with p-KAP-1 antibody (S473). Figure source data can be found with the Supplementary data.
Figure 6
Figure 6
PP4 directly dephosphorylates KAP-1 and impacts expression of stress-reponse genes and G2/M checkpoint. (A) PP4 dephosphorylates KAP1 in vitro. Wild-type PP4C and mutant PP4C (D82A) were purified using the baculoviral system and were serially diluted in the phosphatase reaction. λ phosphatase served as positive control for the reaction. PP4C dephosphorylates phospho-KAP1 on both S824 and S473 in a dose-dependent manner. Phosphatase reactions were probed with indicated antibodies. (B) PP4-mediated regulation of KAP-1 impacts the expression of Gadd45α, p21 and p53AIP1. In HeLa cells, endogenous KAP-1 was replaced with WT, mutant forms A (S824A or/and S473A) and D (S824D or/and S473D). After IR, cells were harvested and RNA purified at indicated times and quantitative real-time PCR (qRT–PCR) was performed. Data represent average and s.d. of four independent experiments. (C) Hyperphosphorylation of KAP1 on S473 impacts G2/M checkpoint. U2OS cells after depleting PP4R3β or replacing endogenous KAP-1 with wild-type, phosphonull mutants, or phosphomimetic mutants were exposed to irradiation (5 Gy) and then released in medium for 3 h and fixed. Mitotic cells were stained with anti-phospho-S10-Histone H3 (p-H3) antibody and quantified by flow cytometry. Data were analyzed with FlowJo software. The results from three independent experiments are graphically represented (S473A, P<0.029; S473D, P<0.011; PP4R3β, P<0.0321; ATMi, P<0.041). Figure source data can be found with the Supplementary data.
Figure 7
Figure 7
KAP-1 phosphonull mutants reverse the phenotype induced by silencing PP4R3β. (A) In cells expressing phosphonull KAP-1 mutants, PP4 depletion does not affect the expression of Gadd45α, p21. In HeLa cells, endogenous KAP-1 was replaced with WT, mutant forms A (S824A or/and S473A), and either control siRNA or PP4R3β siRNA was transfected. After IR, cells were harvested and RNA purified at indicated times and quantitative real-time PCR (qRT–PCR) was performed. Data represent average and s.d. of four independent experiments. (B) KAP-1 S473A mutant, not S824A, reverses G2/M checkpoint phenotype induced by PP4R3β depletion. The G2/M checkpoint assay was performed as described in Figure 6D after silencing PP4R3β with U2OS cells expressing wild-type or phosphonull KAP-1 mutants. The results from three independent experiments are graphically represented (WT/PP4R3β siRNA, P<0.041; S824A/PP4R3β siRNA, P<0.011; S473A/Ctrl siRNA, P<0.019; S473A/PP4R3β siRNA, P<0.009). (C) Working model. In response to IR, ATM directly phosphorylates KAP-1 at S824 and indirectly induces the phosphorylation at S473 via CHK2. Phosphorylation of both S824 and S473 regulates chromatin structure and the expression of KAP-1 target genes, whereas only S473 is involved in the checkpoint response. A PP4 complex dephosphorylates KAP-1 at S824 and at S473 impacting the role of KAP-1 in DDR.

Comment in

Similar articles

See all similar articles

Cited by 48 articles

See all "Cited by" articles

Publication types

Feedback