RAD18 O-GlcNAcylation promotes translesion DNA synthesis and homologous recombination repair

Abstract

RAD18, an important ubiquitin E3 ligase, plays a dual role in translesion DNA synthesis (TLS) and homologous recombination (HR) repair. However, whether and how the regulatory mechanism of O-linked N-acetylglucosamine (O-GlcNAc) modification governing RAD18 and its function during these processes remains unknown. Here, we report that human RAD18, can undergo O-GlcNAcylation at Ser130/Ser164/Thr468, which is important for optimal RAD18 accumulation at DNA damage sites. Mechanistically, abrogation of RAD18 O-GlcNAcylation limits CDC7-dependent RAD18 Ser434 phosphorylation, which in turn significantly reduces damage-induced PCNA monoubiquitination, impairs Polη focus formation and enhances UV sensitivity. Moreover, the ubiquitin and RAD51C binding ability of RAD18 at DNA double-strand breaks (DSBs) is O-GlcNAcylation-dependent. O-GlcNAcylated RAD18 promotes the binding of RAD51 to damaged DNA during HR and decreases CPT hypersensitivity. Our findings demonstrate a novel role of RAD18 O-GlcNAcylation in TLS and HR regulation, establishing a new rationale to improve chemotherapeutic treatment.

Fig. 1. RAD18 binds OGT and is subject to O-GlcNAcylation predominantly at Ser130/Ser164/Thr468.

A HEK293T cells expressing Flag-RAD18 and Myc-OGT were irradiated with UV (15 J m⁻²) followed by immunoprecipitation with anti-Flag M2 beads. The cell lysates were detected with anti-Myc and anti-Flag antibodies. The experiments were carried out in triplicate. B HEK293T cells transfected with SFB-RAD18 or empty vector were incubated with Thiamet-G (TMG) and glucose. The cell lysates were denatured and immunoprecipitated with anti-S beads followed by immunoblotting with O-GlcNAc and Flag antibodies. C HEK293T cells transfected with SFB-RAD18 were treated with Thiamet-G and different concentrations of glucose. The cell lysates were immunoprecipitated and analyzed as in (B). D Schematic representation of WT- and 3A-RAD18. HEK293T cells (E) or cells incubated with glucose and TMG (F) were transfected with the indicated SFB-RAD18 constructs, followed by immunoprecipitation as in (B). G Expression of His-RAD18 (WT or 3A) and GST-OGT or GST co-transformed into E. coli Transetta (DE3) cells after IPTG (0.4 mM) induction were examined via SDS-PAGE followed by Coomassie blue staining. H His-RAD18 proteins were purified and analyzed by western blot with anti-O-GlcNAc and anti-His antibodies.

Fig. 2. O-GlcNAclylation modulates RAD18 recruitment at DNA damage sites.

A Representative image for the dynamic recruitment of WT or 3A GFP-RAD18 to laser-induced DSBs. Data were presented as mean ± SEM from 15 cells. Scale bar, 10 μm. B Quantification of the time course of WT or 3A GFP-RAD18 recruitment after laser microirradiation. C Representative images of GFP-RAD18 foci stained with DAPI after UV irradiation. Scale bars: 2 μm. The protein levels of RAD18 in RAD18+/+ and RAD18-/- U2OS cells were detected by immunoblotting. D Quantification of the percentage of RAD18-/- U2OS cells transfected with WT or 3A GFP-RAD18 constructs with more than 30 RAD18 foci after CPT, Bleomycin and UV exposure by counting at least 200 cells in each experiment. Data represent means ± SEM from three independent experiments. E Chromatin fractions of HEK293T cells expressing WT or 3A SFB-RAD18 were extracted followed by immunoblotting with Flag and H2B antibodies. The whole cell extract (WCE) was harvested and immunoblotted with Flag and β-actin antibodies.

Fig. 3. RAD18 3A mutation impairs PCNA monoubiquitination and cell survival after UV irradiation.

A HEK293T cells transfected with SFB-RAD18 were irradiated with UV (15 J m⁻²) and harvested at different time points later. The cell lysates were denatured and immunoprecipitated with anti-S beads followed by immunoblotting with O-GlcNAc and Flag antibodies. Chromatin fractions of HEK293T cells transfected with SFB-RAD18 (B) or incubated with glucose or not (C) were extracted after UV (15 J m⁻²) irradiation followed by immunoblotting with indicated antibodies. RAD18 O-GlcNAcylation promotes PCNA monoubiquitination. RAD18+/+ or RAD18-/- HEK293T (D) and U2OS (E) or OGT-depleted RAD18-/- HEK293T (F) cells were transfected with WT or 3A RAD18, and irradiated with UV (15 J m⁻²). The chromatin fractions were harvested and immunoblotted using indicated antibodies. G Clonogenic survival assays in RAD18-knockdown U2OS cells transfected with empty vector, WT or 3A Flag-RAD18 after UV irradiation. Cells were irradiated with indicated doses of UV and further incubated in medium supplemented with 0.4 mM caffeine for 7–10 days. Surviving fraction was expressed as a percentage of mock-treated cells. The representative curve is shown. Error bar: s.d., n = 3. The protein levels of RAD18 were detected by immunoblotting.

Fig. 4. O-GlcNAclylation promotes CDC7-mediated RAD18 phosphorylation at Ser434.

A WT or 3A SFB-RAD18 and GFP-Polη co-transfected HEK293T cells were treated with glucose (60 mM) and TMG (5 μM). Chromatin fraction was immunoprecipitated followed by immunoblotting with GFP and Flag antibodies after UV (15 J m⁻²) irradiation. *non-specific band. OGT knockdown HEK293T cells were transfected with WT or 3A SFB-RAD18 and GFP-Polη, followed by incubation with glucose (60 mM) and TMG (5 μM) or not. The whole cell lysates (B) and triton-insoluble fraction (C) were immunoprecipitated and immunoblotted with indicated antibodies post UV (15 J m⁻²) exposure. *non-specific band. D OGT and O-GlcNAcylation levels in (B, C). E SFB-RAD18 (WT or 3A) were transfected into RAD18-/- HEK293T cells followed by UV (15 J m⁻²) irradiation. The lysates were immunoprecipitated and analyzed via western blot by RAD18-S434p and Flag antibodies. F Flag-Polη was transfected into WT or 3A GFP-RAD18-complemented RAD18-depleted U2OS cells followed by UV (15 J m⁻²) irradiation. Quantification of the percentage of GFP-RAD18-positive cells with more than 30 Polη foci after UV (15 J m⁻²) exposure by counting at least 200 cells in each experiment. Data represent means ± SEM from three independent experiments. The protein levels of RAD18 were detected by immunoblotting. G WT or 3A SFB-RAD18 and HA-CDC7 were transfected into HEK293T cells. The lysates were immunoprecipitated using anti-HA beads and analyzed via western blot using Flag and HA antibodies. H His-RAD18 co-expressed with GST or GST-OGT was incubated with HEK293T cell lysates expressing HA-CDC7. The bound proteins were resolved by SDS-PAGE and analyzed by immunoblotting with HA antibody and staining with Ponceau S. I HEK293T cells transfected with HA-OGT and SFB-RAD18 (WT, S434A or 3A) or empty vector were lysed in a denatured condition and immunoprecipitated with anti-Flag beads followed by immunoblotting with O-GlcNAc and Flag antibodies.

Fig. 5. RAD18 O-GlcNAcylation is essential for its binding with both ubiquitin and RAD51C.

A RAD18 O-GlcNAcylation promotes HR repair. DR-GFP U2OS cells overexpressing Mcherry empty vector, WT or 3A Mcherry-Flag-RAD18 were treated with siRAD18 oligo followed by I-SceI endonuclease transfection. HR repair efficiency (GFP⁺Mcherry⁺%) was analyzed by FACS. The lower panels show immunoblots indicating the RAD18 levels in different conditions. B RAD18 O-GlcNAcylation promotes cell survival post-CPT treatment. Clonogenic survival assays in RAD18-knockdown U2OS cells transfected with empty vector, WT or 3A Flag-RAD18 after CPT treatment. Cells were treated with indicated doses of CPT and further incubated for 7–10 days. Surviving fraction was expressed as a percentage of mock-treated cells. The representative curve is shown. Error bar: s.d., n = 3. C Representative images for colony assay in (B). D 3A mutation inhibits the binding ability of RAD18 with ubiquitin. Purified GST and GST-Ubb were incubated with HEK293T cell lysates expressing SFB-RAD18 (WT or 3A) followed by immunoblotting with Flag antibody. Ponceau staining shows the amounts of GST and GST-Ubb used in the pulldown assay. E RAD18-depleted HEK293T cells were transfected with WT or 3A GFP-RAD18 and SFB-RAD51C. The cell lysates were denatured followed by immunoprecipitation with anti-GFP agarose. The immunoprecipitates were immunoblotted with Flag and GFP antibodies. F Purified His-RAD51C was incubated with OGT knockdown HEK293T cell lysates expressing SFB-RAD18 in incubation with glucose (60 mM) and TMG (5 μM) or not. The bound proteins were analyzed as in (D).

Fig. 6. RAD18 O-GlcNAcylation facilitates RAD51 focus formation at DSBs.

A WT or 3A SFB-RAD18 were transfected into siRAD18-treated U2OS cells followed by CPT treatment. Representative images of cells expressing RAD51 foci and SFB-RAD18 were shown. Scale bars: 5 μm. B The percentage of SFB-RAD18-positive cells with more than 10 RAD51 foci was determined. Data represent means ± SEM from three independent experiments by counting at least 200 cells. The protein levels of RAD18 were detected by immunoblotting. C Proposed model depicting the role of RAD18 O-GlcNAcylation in mediating TLS and HR pathways. Upon UV irradiation or CPT exposure, RAD18 undergoes O-GlcNAcylation by OGT. O-GlcNAcylated RAD18 promotes CDC7-dependent phosphorylation at Ser434 and further enhances PCNA monoubiquitination in TLS. O-GlcNAcylation also promotes the binding ability of RAD18 and ubiquitin chain at CPT-induced DSB sites, and further recruits RAD51C facilitating RAD51 loading during HR repair.