ZNF281 is recruited on DNA breaks to facilitate DNA repair by non-homologous end joining

Efficient repair of DNA double-strand breaks (DSBs) is of critical importance for cell survival. Although non-homologous end joining (NHEJ) is the most used DSBs repair pathway in the cells, how NHEJ factors are sequentially recruited to damaged chromatin remains unclear. Here, we identify a novel role for the zinc-finger protein ZNF281 in participating in the ordered recruitment of the NHEJ repair factor XRCC4 at damage sites. ZNF281 is recruited to DNA lesions within seconds after DNA damage through a mechanism dependent on its DNA binding domain and, at least in part, on poly-ADP ribose polymerase (PARP) activity. ZNF281 binds XRCC4 through its zinc-finger domain and facilitates its recruitment to damaged sites. Consequently, depletion of ZNF281 impairs the efficiency of the NHEJ repair pathway and decreases cell viability upon DNA damage. Survival analyses from datasets of commonly occurring human cancers show that higher levels of ZNF281 correlate with poor prognosis of patients treated with DNA-damaging therapies. Thus, our results define a late ZNF281-dependent regulatory step of NHEJ complex assembly at DNA lesions and suggest additional possibilities for cancer patients’ stratification and for the development of personalised therapeutic strategies.

blot analysis (right panel) showing the correct activation of the DNA damage signalling cascade in response to 4-OHT-mediated I-PpoI activation. Red arrows showing 53BP1 foci around nucleoli indicate the induction of DNA damage in the rDNA. Scale bars, 10 µm. (c) Representative WB of U2OS cells exposed to 5 Gy of IR and harvested at the indicated time points. GAPDH was used as a loading control. (d) Representative WB of U2OS and two ZNF281 -/clones exposed to 5 Gy of IR and harvested 8 hours later.
GAPDH was used as loading control. (e) Schematic depicting the HR reporter system (top). The HR cassette carries two mutated copies of the GFP gene, both of which are inactive. Upon I-SceI digestion, the functional GFP gene is restored by gene conversion through HR and is quantified using FACS analysis. Scrambled siRNA-transfected U2OS harbouring integrated HR reporter cassette were used as control to measure the relative repair efficiency of BRCA1-and ZNF281-depleted cells (left). Graphs present means ± SD; n=3; * p<0.05 (two-tailed Student's t-test), n.s. not significant. Western blot showing the knock-down efficiency of BRCA1 and ZNF281 (right); GAPDH was used as a loading control. (f) WB of sub-cellular fractions of U2OS cells exposed to 10 Gy of IR and then harvested at the indicated time points of recovery. γ-H2AX was used as control for DNA damage induction, while the amount of total H2AX shows the separation of the chromatin-bound fraction from the soluble fraction. (g) WB of DIvA cells (relative to Fig. 1f) demonstrates ZNF281 knockdown efficiency and pATM S 1981 activation in response to 4-OHT treatment; Ku70 was used as a loading control. Figure S3. (a, b and c) The recruitment of ZNF281 Wt to UV laser-induced DNA damage sites was compared to S785A (a), S807A (b) and S785A+S807A (c) phospho-mutants in U2OS cells. The

SUPPLEMENTARY MATERIALS AND METHODS
Clonogenic assay. HEK293T cells treated with the indicated siRNAs for 48h or U2OS parental and KO clones were irradiated with different doses of IR. Immediately after irradiation cells were trypsinized, counted and replated in triplicate at low density. Colonies were allowed to form for 12-16 days and then fixed in 3:1 methanol:acetic acid before staining with 0.1% (w/v) crystal violet in PBS. Colonies with at least 50 cells were automatically scored with the ImageJ software and the survival fraction was determined from the plating efficiency of the irradiated cells relative to the plating efficiency of the 0 Gy controls, as described in 4 .  Table S2. Uncropped scans of all western blots are shown in Supplementary Fig. S6 and S7.

Immunofluorescence (IF).
For g-H2AX IF, cells were seeded on glass coverslips, treated as indicated and fixed with 4% paraformaldehyde for 10 min at room temperature (RT). For 53BP1 IF, cells were pre-extracted for 4 min with CSK buffer (100 mM NaCl, 300 mM sucrose, 3mM MgCl2, 10 mM PIPES pH 6.8, 10mM β-glycerol phosphate, 50mM NaF, 1 mM EDTA, 1 mM EGTA, 5 mM Na orthovanadate, 0.5% Triton X-100) and then fixed with 4% paraformaldehyde on ice for 20 min. In both cases, after the fixation step, cells were permeabilized for 5 min in 0.25% Triton X-100 and blocked in 10% goat serum for 1h. Slides were then incubated with the indicated primary antibodies (Supplementary Table S2) 1h at RT and afterwards with secondary antibody (Alexa Fluor 488, invitrogen) along with Hoechst to counterstain nuclei. Images were acquired using a Zeiss LSM510 confocal microscope and analysis performed using the Volocity software (PerkinElmer). All images of a given experiment set were analysed under the same parameters.
Translocation assay. U2OS-AsiSI cells were induced with 4-OHT as indicated, and genomic DNA was extracted using the PureLink kit (Invitrogen). Illegitimate rejoining frequencies between DSBs induced in the same chromosome (MIS12::TRIM37) or in different chromosomes (TRIM37:: RBMXL1) were assessed by qPCR (Applied Biosystems® 7500 fast) or with endpoint PCR using Q5 High-Fidelity DNA Polymerase (New England Biolabs) according to the manufacturer's instructions. Translocation frequencies were calculated using the 2 -DDCt method, normalising on two control regions far from any AsiSI cut sites, as previously described 3 .
Primers used for translocation assay are listed in the Supplementary Table S1. 48-72h and analysed by flow cytometry. The DSBs repair efficiency was calculated as the ratio between the percentage of GFP (successfully repaired DSBs) and mCherry (transfection efficiency) positive cells.

Laser micro-irradiation and live cell imaging. U2OS cells expressing the indicated fluorescent-fusion proteins
were seeded into Lab-Tek chambered coverglass wells (#155383, Thermo Scientific) and pre-sensitized with 20 μM Hoechst 33342 (Thermo Scientific) for 10 min at 37 °C prior the UVA laser micro-irradiation. DNA damage was induced focally using the 3i 'Ablate' UV 355 nm pulsed laser system in a circle of radius 1 µm. Image acquisition was performed using a spinning disk inverted microscope system (3i), while cells were maintained at 37 °C and 5% CO2 in a stage top environmental control system (okolab) for the whole duration of the experiment. The 488 nm (for EGFP-constructs) or the 561 nm (for mCherry-constructs) channels were imaged every 5 s to monitor the fluorescence intensity. Background subtraction and photobleaching correction were applied using the 3i Slidebook software and, for each cell, relative fluorescence intensity was calculated by normalising the intensity of pre-damaged cells (t=0s) to a value of 1.

Site-directed DNA damage induction and Chromatin immunoprecipitation (ChIP). U2OS cells were
infected with a lentivirus expressing the dd-HA-ER-I-PpoI 5 endonuclease for 24h in presence of 8 µg/ml of polybrene. To enrich the I-PpoI-containing population, cells were subjected to a 24h round of puromycin selection (2.5 µg/ml). Once the puromycin was washed out, cells were allowed to recover for additional 24h. To