Purified recombinant XLF is phosphorylated by DNA-PK in vitro. Full-length His-XLF was purified over NTA-agarose and incubated with purified DNA-PKcs and Ku as described in Section 2. (A) Coomassie stained SDS PAGE gel showing purified His-tagged XLF. Positions of molecular weight markers (in kDa) are shown on the left. (B) His-XLF was incubated with ³²P-γ-ATP without (−) and/or with (+) purified DNA-PK as indicated. Reactions were stopped by the addition of SDS and samples were analyzed by SDS PAGE followed by autoradiography. GST-XRCC4 was included as a positive control. The positions on the autoradiogram of autophosphorylated DNA-PKcs and phosphorylated Ku70 (indicated by the *), XRCC4 and XLF are shown. (C) His-XLF was incubated with DNA-PK, non-radioactive ATP and wortmannin (5 μM) as indicated, then analyzed by SDS PAGE followed by western blot using an antibody to phosphorylated SQ/TQ sites (Cell Signaling Inc). The position of phosphorylated DNA-PKcs and His-XLF are shown. (D) Serine/threonine to alanine mutations were generated in full-length GST-tagged XLF at serines 245 and 251 as indicated. The GST-tagged protein was cleaved using PreScission Protease (GE Healthcare), and the purified proteins were incubated without or with DNA-PK as indicated under standard assay conditions as described in Section 2 and analyzed by SDS PAGE and autoradiography. Upper panel, region of the Coomassie stained SDS PAGE gel corresponding to XLF; lower panel, region of the autoradiogram corresponding to XLF. Phosphorylation of the mutant proteins as a percentage of phosphorylation of wt-XLF is shown in the box. (E) Truncation mutations were generated in GST-XLF and the purified proteins, consisting of amino acids 1 –239 and 240 –299 as indicated, were phosphorylated by DNA-PK as described above. Upper panel, Coomassie stained SDS PAGE gel, lower panel, autoradiogram.

DNA-PK and ATM are required for in vivo phosphorylation of XLF-FLAG on serine 245 and serine 251, respectively. (A) Cells were transfected with wt-XLF-FLAG (lanes 1–4) or XLF-FLAG with either serine 245 or 251 mutated to alanine [XLF(S245A) or XLF(S251A), respectively], (lanes 5 and 6). Cells were unirradiated (lane 1), or irradiated as indicated, then, after 2 h, cell extracts were immunoprecipitated with anti-FLAG agarose beads and probed with either phosphospecific antibodies to serine 245 or 251 or an antibody to FLAG as indicated. (B) Cells were transfected with either wt-XLF-FLAG (lanes 1–4) or phosphorylation mutant XLF-FLAG as indicated(lanes 5 and 6), irradiated at 5Gy, then analyzed at the times indicated by immunoprecipitation and western blot using antibodies to phosphoserine 251 or FLAG as indicated. (C) Cells were transfected with wt-XLF-FLAG (lanes 1–7) and incubated with either the DNA-PK inhibitor NU7441 (lanes 3, 4 and 7) or the ATM inhibitor KU55933 (lanes 5–7) as indicated, or an equivalent volume of DMSO. The sample in lane 1 was not irradiated. Samples in lanes 2–7 were irradiated with 5Gy and two hours after irradiation, XLF was immunoprecipitated using anti-FLAG agarose beads and analyzed using phosphospecific antibodies to phosphoserine 245, phosphoserine 251 or FLAG as indicated.

The DNA-binding domain of XLF is located in the C-terminal 60 amino acids. Full-length GST-XLF was incubated with (A) 100, (B) 200 or (C) 300bp dsDNA and protein–DNA complexes were analyzed by EMSA as described in Section 2. The left hand figure in each panel shows lanes containing probe alone (lane 1), 0.012 (lane 2), 0.06 (lane 3), or 0.3 pmoles (lane 4) of purified Ku heterodimer. Lanes 6 and 8 contained 48 pmoles of either BSA or purified GST protein as indicated. The right hand panels contained 0 (lane 1), 6 (lanes 2, 6 and 10), 12 (lanes 3, 7 and 11), 24 (lanes 4, 8 and 12) or 48 (lanes 5, 9 and 13) pmoles of either full-length GST- XLF, GST-XLF_1–239 or GST-XLF_240–299, as indicated.

Live cell imaging of wild-type and phospho-mutant XLF. (A) Accumulation of EGFP-tagged wild-type and phosphomutant XLF at laser-induced DSBs. EGFP-XLF and EGFP-XLF containing mutations at either four or six of the phosphorylation sites (4A mutant and 6A mutant, respectively) were expressed in U2OS cells. DSBs were induced in the defined area of the nucleus by a pulse of laser microbeam (shown by arrows), and accumulation of EGFP-XLF was monitored by time-lapse imaging. (B) FRAP analysis of EGFP-XLF at DSBs. Following induction of DSBs by laser micro-irradiation, cells were incubated to allow maximum accumulation of EGFP-XLF at DSBs. The damaged area in the nucleus was photobleached, and the recovery of the fluorescence signal was then monitored by time-lapse imaging. Mean values of relative fluorescence intensities at each time point were calculated from 10 independent measurements. (C) Two hour time-course of EGFP-XLF accumulation at DSBs. Mean values of the relative fluorescence at each time point were calculated from 10 independent measurements.

XLF-FLAG is phosphorylated in a DNA-PK and ATM-dependent manner in vivo. (A) HeLa cells were transfected with vector alone (lane 1) or XLF-FLAG as indicated (lanes 2–6). Samples in lanes 3 –6 were irradiated (10 Gy) after addition of ³²P-inorganic phosphate. Where indicated, cells were incubated for 30 min with wortmannin (5 μM), the DNA-PK inhibitor NU7441 (10 μM), or the ATM inhibitor KU55933 (10 μM) in phosphate-free media as described in Section 2. Cells were then incubated with ³² P-inorganic phosphate and proteins were immunoprecipitated as described in Section 2. The upper panel shows the resulting autoradiogram and the lower panel the western blot for FLAG. Phosphorylation in each sample as a percentage of the phosphorylation in lane 2 is shown in the boxed region. (B) Cells were transfected with vector alone (lane 1), wt-XLF-FLAG (lane 2) or XLF-FLAG containing serine to alanine mutations at amino acids 132, 203, 245 and/or 251 (XLF-4A in figure) as indicated (lanes 3–7). Cells were labeled with ³²P-inorganic phosphate and immunoprecipitated as described in panel (A). The upper panel shows the resulting autoradiogram and the lower panel the western blot for FLAG. Quantitation was as described in panel (A).

Endogenous XLF is phosphorylated on serine 245 and 251 in irradiated human cells. XLF was partially purified from either unirradiated or irradiated human cells in the presence of protein phosphatase inhibitors as described in Supplementary materials and methods. XLF-containing fractions from either unirradiated cells (lanes 1 and 2) or irradiated cells (lanes 3 and 4) was analyzed by SDS PAGE and probed with phosphospecific antibodies to either serine 245 or serine 251 or total XLF as indicated. Where indicated (lanes 2 and 4, indicated by +), samples were incubated with λ phosphatase prior to SDS PAGE.

Phosphorylated C-terminal XLF retains its ability to interact with DNA. (A) Coomassie blue stained SDS gel showing purified GST-XLF_240–299 alone (control, c), after incubation with DNA-PKcs, Ku and DNA in the absence of ATP (mock, m) or the presence of ATP (phosphorylated, p). Phosphorylated protein was indicated by the supershifted bands (lane 3). (B) Non-phosphorylated, mock-phosphorylated or phosphorylated GST-XLF_240–299 was analyzed in EMSA assays for interaction with 300 bp dsDNA as described in Fig. 5. Lane 1 contained probe alone, lanes 2, 6 and 10 contained 6 pmoles of protein, lanes 3, 7 and 11 contained 12 pmoles of protein, lanes 4, 8 and 12 contained 24 pmoles and lanes 5, 9 and 13, contained 48 pmoles of protein. (C) Samples containing control (c), mock phosphorylated (m) and phosphorylated (p) GST-XLF_240–299 were prepared exactly as described above but 0.7 μg phosphospecific antibody to serine 251 (lanes 2-4) or a non-specific (ns) IgG (lane 5) was added after addition of the DNA probe. Samples were incubated and analyzed by EMSA as in panel B.

Phosphorylation mutant 6A-XLF complements DSB repair defects in XLF-deficient 2BN cells. (A) 2BN cells were transiently transfected with pEGFP-wt-XLF (solid line, closed squares) or pEGFP-6A-XLF (dashed lines, open circles) and cells were probed for γ-H2AX as described in Section 2. γ-H2AX foci were counted in GFP-labeled cells and compared to those in non-GFP labeled cells (solid line, open squares). The results are plotted as average number of foci per cell. (B) 2BN cells were transiently transfected with vector alone, pEGFP-wt-XLF, or pEGFP-6A-XLF. Cells were irradiated (1 Gy) 24 h after transfection, then seeded in triplicate and colonies were counted after crystal violet staining 14 days later. XLF protein levels were constant for at least 6 days post-transfection (Supplementary Fig. 5). Surviving fraction was calculated from number of colonies counted divided by numbers of cells seeded × plating efficiency. Results are representative of three separate experiments.