INT6 is required for cell survival and G2/M checkpoint following γ-irradiation. (A) HeLa cells were transfected with control siRNAs, ATM siRNAs or two siRNAs targeting distinct regions of the INT6 mRNA. Fourty hours later, cells were exposed with increasing doses of γ-radiation and allowed to grow for 72 h before cell proliferation was measured by a MTT-based assay. Shown are the averages of triplicate samples. Error bars correspond to standard errors. (B) HeLa cells transfected for 60 h with siRNAs control or targeting INT6 were untreated or γ-irradiated (6 Gy) and harvested 2h later. Cells were labeled with an antibody to phospho-histone H3 and propidium iodide. Percentages of mitotic cells (boxed area) were determined by FACS.

INT6 knock-down impairs activation of CHK2 and CHK1. (A) HeLa cells transfected with siRNAs control or targeting INT6 were untreated or γ-irradiated (6 Gy) and collected after 0.5 h or 2 h. Immunoblots were performed using antibodies to CHK2 and CHK1 unphosphorylated or phosphorylated on Thr68 and Ser345, respectively. RNAi efficacy was controlled by detection of INT6. (B) Hela cells were transfected with the parental vector (pCEP-FGFP), the vector containing the INT6 shRNA (pCEP-SUPI6-FGFP), or the one containing the INT6 shRNA plus an shRNA-resistant INT6 cDNA (pCEP-SUPI6-I6M5). Plasmid description is in Supplemental Materials. Cells were treated with 200 ng/ml NCS for 30 min and collected at indicated time points. Immunoblots were performed with phospho-specific antibodies to CHK2 and CHK1. Efficiency of INT6-depletion and re-expression and equal protein loading were controlled by detection of INT6 and β-actin, respectively.

Silencing INT6 impairs sustained accumulation of ATM to DNA damage sites. (A) HeLa cells transfected with siRNAs were exposed, or not, to γ-irradiation (10 Gy) and collected after 1 h. Immunoblots were performed using antibodies to unphosphorylated or Ser1981-phosphorylated ATM. RNAi efficacy and protein loading were controlled as in Fig. 2B. (B) HeLa cells transfected with control or INT6-targeting siRNAs were untreated or γ-irradiated (10 Gy). Cells were immunostained 1 h later with an antibody to Ser1981-phosphorylated ATM. Representative confocal images are shown. Scale bar, 10μM. (C) Accumulation of GFP-ATM at laser-induced DSBs was monitored by time-lapse imaging in U2OS cells transfected with control or INT6-targeting siRNAs. White arrows on left images indicate micro-irradiated areas. Scale bar, 10 μM. (D) Kinetics of GFP-ATM relative accumulation at laser-induced DSBs in control and INT6-depleted cells. Mean value of fluorescence intensities for each time point was calculated from at least 30 independent measurements. Error bars represent standard deviations. (E) Biochemical fractionation of HeLa cells transfected with control and INT6-specific siRNAs. Cells were treated with NCS (200 ng/ml, 15 min) and fractionated 3 h later as described in Supplemental Methods. Aliquots of each fraction corresponding to an equal cell number were used for immunoblots using antibodies to ATM, either total or autophosphorylated. Successful fractionation was verified through detection of RRM2 in the cytoplasm and a modified form of histone H3 on chromatin. RNAi efficiency was checked by detecting INT6.

INT6 is recruited to DSBs induced by NCS treatment or micro-irradiation. (A) HeLa cells transfected with control or INT6-targeting siRNAs were treated, or not, with NCS (200 ng/ml, 15 min), and immunostained 3 h later with antibodies against INT6 and γ-H2AX. Representative confocal images are shown. Co-localized pixels were determined using the Co-localization Highlighter plug-in for ImageJ software. They appear as white dots on right panels. Scale bar, 10 μM. (B and C) U2OS cells were untreated (B) or pre-treated with 5 μM KU55933 (C), micro-irradiated, fixed at the indicated time points, immunostained with antibodies to INT6 and γ-H2AX, and counterstained with DAPI. The merged red and green channels show co-localization in yellow. Bar, 10 μM. (D) Recruitment of INT6 was assessed in ATM-deficient cells (AT5) processed as above.

INT6 interacts physically with ATM. (A) Association of ectopically-expressed ATM and INT6. 293T cells were transiently transfected with vectors expressing FLAG-ATM and INT6. Cell extracts were blotted directly (lanes 1–3) or after immunoprecipitation with pre-immune serum (lane 7) or an antibody to INT6 (lanes 4–6). An antibody against FLAG was used to detect FLAG-ATM. (B) Reverse pull-down. Extracts as above were immunoprecipitated using an antibody to FLAG (lanes 4–6) or a control antibody (lane 7) and immunoblotted with an antibody against INT6. (C) Association of the endogenous proteins. HeLa cells were non-treated (lane 1) or treated with NCS (200 ng/ml, 30 min) and collected after 1 h (lane 2). Whole cell lysates were immunoprecipitated with pre-immune serum (lanes 3, 5) or an antibody to INT6 (lanes 4, 6) and immunoblotted using an antibody against ATM. Recovery of INT6 is shown on bottom panel. Uncropped images of blots showing the co-precipitation of ATM and INT6 are in Fig. S7. (D) Mapping of interacting regions between ATM and INT6. A schematic representation of ATM is shown at top of the panel. Structural domains within ATM include the FRAP/ATM/TRRAP (FAT) domain, the kinase domain (PI3K), and the FAT c-terminal (FAT-C) domain. Various GST-ATM fusion proteins that cover the full ATM protein sequence were incubated with lysates from 293T cells untreated or γ-irradiated (6 Gy) and bound proteins were analyzed by immunoblotting with an antibody to INT6. Purification of GST-ATM proteins on glutathione agarose beads was controlled on a gel stained with coomassie blue.

INT6 activity in DDR does not rely on a translational effect. (A) HeLa cells were treated, or not, with NCS (200 ng/ml, 30 min) and whole cell extracts were prepared 1 h later. Lysates were immunoprecipitated with preimmune serum (lanes 1, 3) or an antibody to INT6 (lanes 2, 4). Co-immunoprecipitated proteins were analyzed by immunoblotting using antibodies against total eIF3 (only the part of the blot corresponding to the largest eIF3 subunits is shown) or against EIF3L and EIF3D. Recovery of INT6 is shown on bottom panel. (B) Cytoplasmic extracts were prepared from HeLa cells treated as in (A) and were immunoprecipitated with a control antibody (lanes 1, 3) or an antibody to EIF3B (lanes 2, 4). Co-immunoprecipitated proteins were immunoblotted with an antibody to INT6. The same membrane was reprobed with an antibody against EIF3B to verify its pull-down. (C) Hela cells were treated with NCS (200 ng/ml, 15 min) in the presence or absence of 50 μg/ml cycloheximide (CHX). After washing out of NCS, CHX was maintained for 2 h and cells were immunostained using antibodies to Ser1981-phosphorylated ATM and γ-H2AX. Representative confocal images are shown. Scale bar, 10 μM. White squares on merge images delineate regions shown in right panels. These are composite images obtained using the Co-localization Highlighter plug-in for ImageJ. Co-localized pixels appear as white dots. (D) UV-absorbance profiles of cytoplasmic extracts from HeLa cells through a 10–50% sucrose gradient. Cells were transfected with siRNAs control or targeting INT6 for 70 h, treated with NCS (200 ng/ml, 1 h), and collected after 1 h. Positions of 40S and 60S ribosomal subunits, 80S monosomes and polysomes are shown. (E) Transcripts encoding DDR proteins were measured using the NanoString nCounter system from total and polysomal RNAs isolated from cells transfected as in (D). Results are expressed as the mean fold-change of three independent experiments (INT6 knock-down versus control) and error bars correspond to standard deviation. Detailed results and procedures are in Tables S1, S2 and Supplemental Methods.

INT6 is required for an efficient DDR. (A–C) HeLa cells transfected with siRNAs were treated with NCS (200 ng/ml, 15 min) and immunostained 3 h post-treatment with antibodies to γ-H2AX and Ser343-phosphorylated NBS1 (panel A), conjugated ubiquitin (FK2 antibody, panel B), and BRCA1 (panel C). ~30 cells were acquired for each cell population and experiments were repeated at least twice. Representative confocal images shown were processed as in Fig. 6C. (D) Fluorescence intensities of γ-H2AX, phospho-NBS1, ubiquitin and BRCA1 labeling were measured within the entire nucleus (Total) and within γ-H2AX foci (Foci). Ten nuclei from control cells and from INT6-deficient cells were analyzed for each experiment. Shown are the mean values of fluorescence intensities. Error bars represent standard errors of the mean.