Modulation of LSD1 phosphorylation by CK2/WIP1 regulates RNF168-dependent 53BP1 recruitment in response to DNA damage

Abstract

Proper DNA damage response is essential for the maintenance of genome integrity. The E3 ligase RNF168 deficiency fully prevents both the initial recruitment and retention of 53BP1 at sites of DNA damage. In response to DNA damage, RNF168-dependent recruitment of the lysine-specific demethylase LSD1 to the site of DNA damage promotes local H3K4me2 demethylation and ubiquitination of H2A/H2AX, facilitating 53BP1 recruitment to sites of DNA damage. Alternatively, RNF168-mediated K63-linked ubiquitylation of 53BP1 is required for the initial recruitment of 53BP1 to sites of DNA damage and for its function in repair. We demonstrated here that phosphorylation and dephosphorylation of LSD1 at S131 and S137 was mediated by casein kinase 2 (CK2) and wild-type p53-induced phosphatase 1 (WIP1), respectively. LSD1, RNF168 and 53BP1 interacted with each other directly. CK2-mediated phosphorylation of LSD1 exhibited no impact on its interaction with 53BP1, but promoted its interaction with RNF168 and RNF168-dependent 53BP1 ubiquitination and subsequent recruitment to the DNA damage sites. Furthermore, overexpression of phosphorylation-defective mutants failed to restore LSD1 depletion-induced cellular sensitivity to DNA damage. Taken together, our results suggest that LSD1 phosphorylation modulated by CK2/WIP1 regulates RNF168-dependent 53BP1 recruitment directly in response to DNA damage and cellular sensitivity to DNA damaging agents.

Figure 1.

CK2 phosphorylates LSD1 at S131 and S137 both in vitro and in vivo. (A and B) Endogenous CK2α’ co-immunoprecipitated reciprocally with endogenous LSD1. Total cell lysates were extracted from 293T cells and subjected to immunoprecipitation followed by immunoblotting with antibodies as indicated. (C) LSD1 directly interacted with CK2α’. GST or GST- CK2α’ recombinant protein was employed to pull down bacterially produced HIS-LSD1. (D) CK2 phosphorylated LSD1 in vitro. GST or GST-CK2α’ recombinant protein was incubated with bacterially produced GST-LSD1 or its phosphorylation-defective mutants in the presence of ³²P-γATP in the in vitro kinase assay. (E) The phospho-specific antibodies were reactive toward LSD1 phosphorylated by CK2 in vitro. Bacterially produced GST-LSD1 was incubated with GST-CK2 and subjected to immunoblotting with antibodies as indicated. (F) Specificity of the LSD1 phospho-specific antibodies. Total cell lysates were extracted from 293T cells transiently transfected with HA-Vec or expression constructs as indicated, and the HA immunocomplexes were then subjected to immunoblotting with antibodies as indicated. (G) Inhibition of CK2 expression diminished LSD1 phosphorylation at S131. Total cell lysates were extracted from HeLa cells with mock depletion or depletion of both CK2α and CK2α’ and subjected to immunoblotting with antibodies as indicated.

Figure 2.

WIP1 dephosphorylates LSD1 at S131 and S137 both in vitro and in vivo. (A and B) Endogenous WIP1 co-immunoprecipitated reciprocally with endogenous LSD1. Total cell lysates were extracted from 293T cells and subjected to immunoprecipitation followed by immunoblotting with antibodies as indicated. (C) WIP1 directly interacted with LSD1. Bacterially produced GST-WIP1 was used to pull down bacterially produced HIS-LSD1 in vitro. (D) Overexpression of WIP1 decreased LSD1 phosphorylation at S131 and S137 in vivo. 293T cells were transiently transfected with expression constructs as indicated. Total cell lysates were harvested 48 h after transfection and subjected to immunoprecipitation with an anti-HA antibody followed by immunoblotting with antibodies as indicated. (E) WIP1 dephosphorylated LSD1 at S131 and S137 in vitro. CIP or bacterially produced GST, GST-WIP1 or GST-WIP1 (D314A) was incubated with the LSD1 immunocomplexes prepared from 293T cells in the in vitro dephosphorylation assay. *degraded GST fusions.

Figure 3.

CK2-mediated phosphorylation of LSD1 promotes its recruitment to sites of DNA damage. U2OS cells transiently expressing GFP-LSD1, GFP-LSD1(S131A) or GFP-LSD1(S137A) were irradiated with a 365-nm UV laser beam, while some GFP-LSD1 expressing cells were also pretreated with KU55933 or TBB before irradiation. Images were collected every 60 s after irradiation. Representative images are shown in (A). In each set of experiment, more than 20 GFP-positive cells were irradiated and monitored for enrichment of GFP signal along the irradiation path. Three experimental repeats were performed and statistical analysis is shown in (B).

Figure 4.

CK2-mediated phosphorylation of LSD1 promotes its association with RNF168. (A) LSD1 directly interacted with RNF168. Bacterially produced LSD1 was used to pull down bacterially produced HIS-RNF168 in vitro. (B) CK2-mediated phosphorylation of LSD1 enhanced its binding to RNF168 in vitro. Bacterially produced GST, GST-LSD1 and GST-LSD1 phosphorylated by CK2 in vitro in the absence/presence of the CK2 inhibitor TBB were employed to pull down bacterially produced HIS-RNF168. (C) Phosphorylation-defective mutants of LSD1 reduced their binding to RNF168 in vivo. 293T cells were transiently co-transfected with FLAG-RNF168 and HA-Vec/HA-LSD1/HA-LSD1(S131A)/HA-LSD1(S137A)/HA-LSD1(2A). Total cell lysates were extracted 48 h after transfection and subjected to immunoprecipitation with an anti-HA antibody followed by immunoblotting with antibodies as indicated. (D) Inhibition of CK2 activity in vivo reduced the binding of LSD1 to RNF168. 293T cells were co-transfected with FLAG-RNF168 and HA-Vec/HA-LSD1. Transfectants were treated with different concentrations of TBB 6 h before harvest. Total cell lysates were harvested 48 h after transfection and subjected to immunoprecipitation with an anti-HA antibody followed by immunoblotting with antibodies as indicated. (E) Bleomycin treatment increased LSD1 phosphorylation in the chromatin-enriched fraction. 293T cells were transiently transfected with HA-Vec or HA-LSD1. HA-LSD1 transfectants were either mock-treated or treated with bleomycin (10 μg/ml) for 1 h before harvest. Chromatin-enriched fractions (P3) were prepared 48 h after transfection and subjected to immunoprecipitation with an anti-HA antibody followed by immunoblotting with antibodies as indicated. (F) DNA damage induced an increase of interaction between RNF168 and LSD1 in the chromatin-enriched fraction. 293T cells were transiently co-transfected with FLAG-RNF168 and HA-Vec/HA-LSD1. Co-transfectants with FLAG-RNF168 and HA-LSD1 were either mock-treated or treated with bleomycin (10 μg/ml) for different time points as indicated before harvest. Chromatin-enriched fractions (P3) were prepared 48 h after transfection and subjected to immunoprecipitation with an anti-HA antibody followed by immunoblotting with antibodies as indicated.

Figure 5.

RNF168 promotes the interaction between 53BP1 and LSD1. (A) LSD1 interacted directly with 53BP1. Bacterially produced GST-53BP1-KBD was used to pull down bacterially produced HIS-LSD1 in vitro. (B) Phosphorylation-defective LSD1 mutants retained their interaction capacity with 53BP1. 293T cells were co-transfected with HA-53BP1-KBD and FLAG-Vec/FLAG-LSD1/HA-LSD1(S131A)/FLAG-LSD1(S137A)/FLAG-LSD1(2A). Total cell lysates were extracted 48 h after transfection and subject to immunoprecipitation with an anti-FLAG antibody followed by immunoblotting with antibodies as indicated. (C) Inhibition of RNF168 expression compromised the interaction between 53BP1 and LSD1. 293T cells were transfected first with siCTR or siRNF168 and then with HA-53BP1 and FLAG-Vec/FLAG-LSD1 24 h after the first transfection. Total cell lysates were harvested 48 h after the second transfection and subjected to immunoprecipitation with an anti-FLAG antibody followed by immunoblotting with antibodies as indicated. (D) Overexpression of RNF168 enhanced the interaction between 53BP1 and LSD1. 293T cells were co-transfected triply with HA-53BP1, MYC-Vec/MYC-LSD1 and FLAG-Vec/FLAG-RNF168. Total cell lysates were extracted 48 h after transfection and subjected to immunoprecipitation with an anti-MYC antibody followed by immunoblotting with antibodies as indicated.

Figure 6.

LSD1 phosphorylation promotes the interaction between 53BP1 and RNF168 in chromatin. (A) LSD1 interacted directly with 53BP1 and CK2-mediated phosphorylation promoted this interaction. Bacterially produced GST-53BP1-KBD was used to pull down bacterially produced HIS-RNF168 in the absence of HIS-LSD1 or in the presence of HIS-LSD1, HIS-LSD1 phosphorylated by CK2α’ in vitro in the absence/presence of the CK2 inhibitor TBB. (B) Inhibition of LSD1 expression compromised the interaction between 53BP1 and RNF168 in the chromatin-enriched fraction. 293T cells were transfected first with siCTR or siLSD1 and then with FLAG-RNF168 24 h after the first transfection. Cells were treated with bleomycin (10 μg/ml) for 1 h before harvest. Chromatin-enriched fractions (P3) were extracted 48 h after the second transfection and subjected to immunoprecipitation followed by immunoblotting with antibodies as indicated. (C) Phosphorylation-defective LSD1 mutants reduced the interaction between 53BP1 and RNF168. 293T cells were transfected first with siCTR or siLSD1 and then co-transfected with FLAG-RNF168 and HA-Vec/HA-LSD1/HA-LSD1(S131A)/HA-LSD1(S137A)/HA-LSD1(2A) 24 h after the first transfection. Chromatin-enriched fractions (P3) were extracted 48 h after the second transfection and subjected to immunoprecipitation with an anti-53BP1 antibody followed by immunoblotting with antibodies as indicated. (D) Inhibition of LSD1 expression compromised RNF168-mediated 53BP1 ubiquitination. 293T cells were transfected first with siCTR or siLSD1 and then co-transfected with HA-Vec/HA-UB(K63) and MYC-Vec/MYC-RNF168 24 h after the first transfection. Total cell lysates were extracted 48 h after the second transfection and subjected to immunoprecipitation with an anti-53BP1 antibody followed by immunoblotting with antibodies as indicated.

Figure 7.

LSD1 phosphorylation promotes 53BP1 recruitment to the DNA damage sites. (A and B) Inhibition of CK2 activity compromised bleomycin-induced 53BP1 recruitment. U2OS cells were either untreated or treated with bleomycin (10 μg/ml) for 1 h without or with pretreatment with TBB (100 μM) for 6 h. Cells were fixed and proceeded for immunofluorescence (IF) staining with an anti-53BP1 antibody. Representative images are shown in (A) and the quantitation data of 53BP1 foci per cell is in (B). (C and D) Overexpression of WIP1 compromised 53BP1 recruitment. U2OS cells were transfected with FALG-Vec, FLAG-WIP1 or FLAG-WIP1 (D314A). Transfectants were either untreated or treated with bleomycin (10 μg/ml) for 1 h before IF. IF was performed 48 h after transfection with antibodies as indicated. Representative images are shown in (C) and the quantitation data of > = 10 53BP1 foci per cell is in (D). (E and F) Defective recruitment of 53BP1 to the DNA damage sites induced by inhibition of LSD1 expression was rescued by re-expression of the wild-type LSD1, but not the phosphorylation-defective LSD1 mutants. U2OS cells were engineered to stably express HA-Vec, HA-LSD1, HA-LSD1(S131A), HA-LSD1(S137A) or HA-LSD1(2A) by retroviral infection. Infectants were transfected with siCTR or siLSD1. Transfectants were synchronized by double thymidine blocks 6 h after transfection and released for 6 h. Cells were then either untreated or treated with bleomycin (10 μg/ml) for 1 h. IF was performed with antibodies as indicated. Representative images are shown in Supplementary Figure S5, while the quantitation data of 53BP1 foci per cell is in (E), and immunoblotting data is in (F).

Figure 8.

LSD1 phosphorylation promotes cell proliferation and survival in response to DNA damage. (A and B) Inhibition of LSD1 expression sensitized cells to bleomycin treatment. HEK293 cells were stably infected with lentiviral shCTR, shLSD1–1, shLSD1–2 or shLSD1–3. Total cell lysates were extracted and subjected to immunoblotting with antibodies as indicated in (A). Infectants were treated with bleomycin with increasing concentrations for 72 h. Relative cell proliferation was determined by MTT assay (B). (C–E) Phosphorylation-defective LSD1 mutants failed to correct LSD1 depletion-induced cellular sensitization to bleomycin treatment. The shLSD1–1-mediated stable knockdown HEK293 cells described in (A) were engineered to stably express HA-Vec, HA-LSD1res, HA-LSD1(S131A)res, HA-LSD1(S137A)res or HA-LSD1(2A)res by retroviral infection. Infectants were treated with bleomycin with increasing concentrations for 72 h. LSD1 protein levels were shown in (C), relative cell proliferation was determined by MTT assay (D) and relative cell survival was measured by clonogenic assay (E). *P < 0.05.

Figure 9.

A working model showing that CK2-mediated phosphorylation of LSD1 at S131 and S137 promotes RNF168-dependent recruitment of 53BP1 to the sites of DNA damage. A solid arrow depicts a positive regulation of the particular process or promotes the interaction between the two proteins, while a dotted arrow indicates a multiple-step process, and a solid line represents a constitutive or unregulated interaction between the two proteins. Details of this model are described in the Discussion section.