Tankyrases Promote Homologous Recombination and Check Point Activation in Response to DSBs

Abstract

DNA lesions are sensed by a network of proteins that trigger the DNA damage response (DDR), a signaling cascade that acts to delay cell cycle progression and initiate DNA repair. The Mediator of DNA damage Checkpoint protein 1 (MDC1) is essential for spreading of the DDR signaling on chromatin surrounding Double Strand Breaks (DSBs) by acting as a scaffold for PI3K kinases and for ubiquitin ligases. MDC1 also plays a role both in Non-Homologous End Joining (NHEJ) and Homologous Recombination (HR) repair pathways. Here we identify two novel binding partners of MDC1, the poly (ADP-ribose) Polymerases (PARPs) TNKS1 and 2. We find that TNKSs are recruited to DNA lesions by MDC1 and regulate DNA end resection and BRCA1A complex stabilization at lesions leading to efficient DSB repair by HR and proper checkpoint activation.

Fig 1. MDC1 interacts with Tankyrase 1 and Tankyrase 2.

(A) Schematic representation of MDC1 and its domains. The different deletion constructs used in our study are depicted below the full-length protein. The domains marked are: Forkhead associated domain (FHA), SDT-repeats, RNF8 Binding motif (RBM), PST repeats, BRCA1 C-terminal domains (BRCTs). We also placed three regions with so far unknown functions, named”inter I-II-III”. (B) Schematic representation of the two human Tankyrase proteins with their domains; the ankyrin repeats (ANK), the sterile alpha motif (SAM) and the PARP enzymatic domain (PARP). The clones obtained in our yeast-two-hybrid screen are depicted in black below the full-length proteins. (C) Endogenous MDC1 and TNKS1 interact in vivo. HEK 293T cells were treated or not with 100 ng/ml NCS and the cell extracts subjected to immunoprecipitation using an anti-MDC1 antibody. TNKS1 is detected in the immunoprecipitate and the interaction is enhanced after DSB induction. (D) Schematic representation of the tethering system used in our study. (E) Both TNKS1 and TNKS2 interact with chromatin bound MDC1 in vivo. Full length MDC1 was tethered to chromatin in U2OS17 cells (for the system used see panel D) while cells were co-transfected with tagged versions of either TNKS1 or TNKS2. Colocalization of FN-TNKS1 or MN-TNKS2 with mCherry-lacR-MDC1 is detected by immunofluorescence staining.

Fig 2. Mapping of the Tankyrase interacting domain of MDC1.

(A) Confocal microscopy images of U2OS17 cells transfected with different mCherry-lacR-MDC1 domain constructs and flag tagged TNKS1 (FN-TNKS1). The contour of nuclei is shown. Colocalization of the two proteins can be observed in the merged images. (B) Quantification of the colocalization frequency of flag-TNKS1 (FN-TNKS1) or myc-TNKS2 (MN-TNKS2) and the different tethered MDC1 domains. Values represent mean ± SD from three independent experiments (N = 100 cells in each experiment). Statistical significance in all relevant Figs was calculated using the t-test (P<0,05 *, P<0,01 **, P<0,001 ***) (C) The PST repeats of MDC1 regulate the interaction between Tankyrases and MDC1. U2OS17 cells were transfected with different constructs expressing cherry-lacR fused versions of MDC1 (marked on the X axis) and Tankyrase (TNKS1 or TNKS2). % of cells having Tankyrase on the array was determined as for panel (B).

Fig 3. The two Tankyrase binding domains of MDC1 act cooperatively.

(A) Schematic representation of MDC1 and the two predicted TBDs with their sequences highlighted. Directed mutagenesis was conducted to replace 3 amino acids of each TBD to Alanine (shown in red capital letters). (B) Multiple alignment of known TBDs from the literature and the consensus sequence. The two predicted TBDs of MDC1 are highlighted in green. (C) Immunofluorescence staining of U2OS17 cells co-transfected with FN-TNKS1 and mCherry-lacR-MDC1 wild type or mCherry-lacR-MDC1 TBD mutant. The mutant version does not colocalize any more with TNKS. (D) Quantification of the colocalization frequency of full-length mCherry-lacR-MDC1 (wt or TBD mutants) and TNKS1/2. Disruption of the two TBDs by directed mutagenesis leads to the loss of colocalization. Values represent mean ± SD from three independent experiments (N = 100 cells in each experiment). (E) In vivo interaction of TNKS1 and the MDC1 “interII” or “BRCTs” domain detected by co-immunoprecipitation. HEK 293T cells were co-transfected with either mCherry-lacR (lanes 1,2), mCherry-lacR-interII (lanes 3–6, upper panel) or mCherry-lacR-BRCTs (lanes 3–6, bottom panel) and FN-TNKS1 (in lanes 2, 5, 6). TBD mutant versions of mCherry-lacR-interII or mCherry-lacR-BRCTs (lanes 4, 6 on upper and bottom panel, respectively) were also included in the analysis. Immunoprecipitation was performed with the M2 flag antibody and the immunoblot was developed with an anti-lacR antibody revealing an interaction between FN-TNKS1 and lacR-interII or lacR-BRCT. Flag antibody was used to detect the presence of FN-TNKS1. Arrowheads show the specific bands. (F) In vivo interaction of TNKS1 and MDC1 requires intact TBD domains. HEK 293T cells were transfected with mCherry-lacR (lanes 1, 2) or mCherry-lacR-MDC1 (lanes 3–6) and FN-TNKS1 (lanes 2, 5, 6). MDC1 was mutated on its two TBDs in lanes 4 and 6. Cell extracts were subjected to anti-flag immunoprecipitaton and analysed by Western Blotting. Only the wild type MDC1 co-precipitates with FN-TNKS (IP lane 5).

Fig 4. Tankyrase1/2 localizes to DSBs.

(A) Confocal microscopy pictures of U2OS cells co-transfected with GFP-MDC1 and FN-TNKS1 or MN-TNKS2 after laser-induced DNA damage. Cells were fixed and stained for γ-H2AX, flag or myc using the appropriate antibodies. (B) Confocal microscopy images of U2OS17 cells co-transfected with mCherry-lacR and FN-TNKS1. The cells were also transfected with an ISce-I expressing plasmid in panels e-h. TNKS localizes to pure DSBs as visualized on panels e-h. The quantification of TNKS localization frequency to DSBs is shown on the right, values obtained from three independent experiments are shown with SD (N = 100). (C) Confocal microscopy pictures of U2OS17 cells co-transfected with mCherry-lacR-MDC1 and FN-TNKS1. An ISce-I expressing plasmid was also transfected on panels m-p. Values are calculated as on panel (B). (D) Targeting of FL MDC1 to chromatin leads to DDR activation. U2OS17 cells were transfected with the indicated constructs, and the % of cells harboring γ-H2AX signal on the array was quantified. Results of three independent experiments are shown with SD (N = 100). (E) ATM mediated phosphorylation events do not affect the interaction between TNKS and the BRCTs of MDC1. U2OS17 cells were transfected with the indicated plasmids in the presence or absence of ATMi. Immunofluorescence staining was performed and the % of cells having FN-TNKS1 signal on the array quantified in three independent experiments (N = 100). Average values are shown with SD.

Fig 5. TNKSs are recruited to DSBs in a cell cycle dependent manner.

(A) Quantification of the frequency of colocalization of TNKS and BRCA1 at pure DSBs in vivo. U2OS17 cells transfected with mCherry-lacR, FN-TNKS1 and ISce-I, and were co-stained with antibodies against flag and BRCA1. Percent of cells having TNKS1 signal at the lacO array also positive for BRCA1 was determined. Results from three independent experiments are shown with SD (N = 30). (B) Quantification of the frequency of TNKS1 recruitment to pure DBSs in control or G1 blocked cells. U2OS17 cells were transfected with mCherry-lacR, FN-TNKS1 and ISce-I. 20hours before fixation, cells were treated with mimosine to block the cell cycle in G1. Percent of cells with FN-TNKS1 signal on the lacO array was determined. Results from three independent experiments are shown with SD (N = 100). (C) TNKS1 is recruited to DSBs in vivo in a subset of cells. U2OS cells were transfected with FN-TNKS1, treated with NCS and immunostained with antibodies against flag and γ-H2AX. Cells were blocked in G2 phase by the CDK1 inhibitor RO-3306 on panels e-g. Frequency of cells showing a colocalizing signal for γ-H2AX and FN-TNKS increased from 24% (asynchronous cells) to 70% (G2 cells) as shown on the graph on the right. Colocalization was analyzed by plotting the intensity of the red/green signal against the distance along the white line on the merged image. The starting point is marked by a star. (D) MDC1 is required for TNKS recruitment to pure DSBs in vivo. U2OS17 cells were transfected with control or MDC1 targeted siRNAs. 24hours after siRNA treatment, cells were co-transfected with mCherry-lacR, FN-TNKS1 or MN-TNKS2 and ISce-I. The frequency of cells having TNKS1/2 signal on the array was determined, results of three independent experiments are shown with SD (N = 100). (E) MDC1 is required for TNKS foci formation. U2OS cells were transfected with control or MDC1-directed siRNAs. 48 hours later the cells were transfected with FN-TNKS1 and 24 hours later treated with 100ng/ml NCS. 6 hours after treatment cells were stained with antibodies against γ-H2AX and FN-TNKS. Representative confocal microscopy pictures are shown.

Fig 6. Tankyrase1/2 plays a role in RAD51 recruitment to DSBs and efficient HR.

(A) Cells depleted of TNKS1/2 show defective foci formation of RAD51. U2OS cells were treated with NCS and fixed 6 hours later. Frequency of RAD51 foci positive cells was determined in three independent experiments (N = 100). Results are shown as relative frequencies compared to control. (B) Representative confocal microscopy images corresponding to panel (A). (C) Quantification of RAD51 on pure DSBs in U2OS17 cells reveals a drop in siTNKS conditions. U2OS17 cells were transfected with the indicated siRNAs, and then co-transfected with mCherry-lacR and ISce-I. Cells were fixed and stained using an antibody against RAD51. Colocalization frequency of the lacR and the RAD51 signal was determined. Relative frequencies compared to the control are results of three independent experiments with SD. (D) Forced binding of TNKS1 to a DSB increases RAD51 recruitment efficiency. U2OS17 cells were transfected with GFP-lacR or GFP-lacR-TNKS1 and ISce-I. The percentage of cells having RAD51 signal on the array was determined after immunofluorescence staining. Results from three independent experiments are shown with SD (N = 100). (E) MDC1-mediated TNKS recruitment is necessary for efficient RAD51 binding to DSBs in vivo. U2OS17 cells were transfected with mCherry-lacR-MDC1 wt or its TBD mutant version and ISce-I. Percent of cells harboring RAD51 signal on the array was determined. Results from three independent experiments are shown with SD (N = 100). (F) HR efficiency is decreasing in TNKS1/2 depleted HRind cells measured in a stable cellular system [49]. Results of three independent experiments are shown with SD. N>10⁵ cells were observed by FACS for GFP expression. Western blot analysis of rescue-efficiency is shown on the right. “res” stands for siRNA rersistant version of the construct, “mut” stands for mutation in the PARP activity.

Fig 7. TNKSs recruit the CtIP-BRCA1 complex to chromatin.

(A) TNKS depletion leads to defects in CtIP loading and RPA phosphorylation. U2OS19 cells were transfected with the indicated siRNAs and pure DSBs were generated by transfecting ISce-I together with mCherry-lacR. Colocalization frequency of RPA-P or CtIP with the lacO array was established in 100 cells in three independent experiments. Values represent mean ± SD. (B) TNKS is required for the recruitment of the BRCA1 to pure DSBs in vivo. U2OS17 cells were transfected with the indicated siRNAs, and then co-transfected with mCherry-lacR and ISce-I. Immunofluorescence staining with an anti-BRCA1 antibody was performed and the colocalization frequency of the lacR and BRCA1 signal quantified. Relative values compared to the control are the results of three independent experiments with SD (N = 100 cells in each condition). (C) TNKS depleted cells are deficient for BRCA1 foci formation. U2OS cells were transfected with siRNAs as shown and treated with NCS 48 hours later. Cells were fixed 6 hours after treatment, immunofluorescence staining detecting BRCA1 was performed. The frequency of foci-positive cells was determined in three independent experiments (N = 100). Representative confocal microscopy pictures for BRCA1 staining are shown on the bottom. (D) TNKS tethering to the chromatin results in BRCA1 recruitment in a partially PARP activity dependent manner. U2OS17 cells were transfected with the indicated plasmids and immunofluorescence staining was performed to establish colocalization frequencies between GFP-lacR-TNKS1 and BRCA1. Representative confocal microscopy pictures are shown on the bottom. (E) MDC1-mediated TNKS recruitment is necessary for efficient binding of BRCA1 to DSBs in vivo. U2OS17 cells were transfected with mCherry-lacR-MDC1 wt or its TBD mutant version and ISce-I. Percent of cells harboring BRCA1 signal on the array was determined. Results from three independent experiments are shown as relative frequencies compared to control with SD (N = 100).

Fig 8. TNKS loads the BRCA1A complex on chromatin.

(A) TNKS tethering to the lacO array leads to MERIT40 and RAP80 binding in a PARP activity independent manner even in the absence of DSBs. U2OS17 cells were transfected with GFP-lacR, GFP-lacR-TNKS1 or the PARP activity mutant version: GFP-lacR-TNKS1mut. ISce-I was co-transfected to induce pure DSBs at the lacO array. Immunofluorescence staining was performed to visualize the localization pattern of MERIT40 or RAP80 in the cells. The values represent the results of three independent experiments with SD (N = 100). On the right representative confocal microscopy pictures are shown. (B) The recruitment of MERIT40 and RAP80 is affected in TNKS1/2 knock down cells. U2OS17 cells were transfected with the indicated siRNAs. After depletion, cells were transfected with mCherry-lacR and ISce-I where indicated. Immunofluorescence staining was performed to visualize Merit40 or RAP80 in the cells. Relative values compared to the control are the results of three independent experiments with SD (N = 100 cells in each condition). (C) TNKS depleted cells are deficient for RAP80 foci formation. U2OS cells were transfected with siRNAs as shown and treated with NCS 48 hours later. The frequency of foci-positive cells was determined in three independent experiments (N = 100) and is shown as relative to the control (D) MDC1-mediated TNKS recruitment is necessary for efficient binding of the BRCA1 complex to DSBs in vivo. U2OS17 cells were transfected with mCherry-lacR-MDC1 wt or its TBD mutant version and ISce-I. Percent of cells harboring MERIT40 or RAP80 signal on the array was determined after immunofuorescence staining. Results from three independent experiments are shown with SD (N = 100) as relative frequencies compared to control. (E) MERIT40 stabilizes BRCA1 on the TNKS-bound chromatin. Control or MERIT40 depleted U2OS17 cells were transfected with GFP-lacR, GFP-lacR-TNKS1, GFP-lacR-TNKS1mut and ISce-I. Percent of cells harboring BRCA1 signal on the array was determined. Results from three independent experiments are shown with SD (N = 100).

Fig 9. TNKSs stabilize the BRCA1A complex at DSBs and activate the G2/M checkpoint.

(A) RNF8 mediated ubiquitination is dispensable for BRCA1A complex’s loading onto the chromatin. U2OS17 cells were transfected with siSCR control or siRNF8, and GFP-lacR, GFP-lacR-TNKS1 or its PARP activity mutant version. The frequency of MERIT40, RAP80 and BRCA1 signal on the lacO array was analyzed after immunofluorescence staining with the corresponding antibodies. Results of three independent experiments are shown with SD (N = 100). (B) RAP80 is recruited to TNKS1 bound chromatin in an ubiquitin-binding independent manner. U2OS17 cells were transfected with GFP-lacR or GFP-lacR-TNKS1 together with mCherry-RAP80 (wt or UIM mutant). Pure DSBs were inflicted with the cotransfection of ISce-I. Colocalization frequency of the GFP and mCherry signal is shown from three independent experiments (N = 100) with SD. (C) TNKS depleted cells show RAP80 foci-forming deficiency at later timepoints after DSB induction. U2OS cells were transfected with the indicated siRNAs, treated with NCS and fixed at the indicated timepoints. Results from three independent experiments are shown with SD (N = 100). (D) Knockdown of TNKS1/2 leads to G2/M checkpoint escape and elevated number of mitotic cells after DSB induction compared to the control. U2OS cells were transfected with the indicated siRNAs. 48 hours later cells were treated with Phleomycin and released for 6 hours. Mitotic cells were detected by an anti- histone H3 S10P fluorescence staining and their frequency determined. Values are represented on a Whisker box plot as relative to the control from three independent experiments. (E) Schematic action of TNKSs in DSB repair is shown.