The FASEB Journal • Research Communication

Protective role of RAD50 on chromatin bridges during abnormal cytokinesis Bianca Schröder-Heurich,*,1 Britta Wieland,* Martin F. Lavin,†,‡ Detlev Schindler,§ and Thilo Dörk* *Gynaecology Research Unit, Hannover Medical School, Hannover, Germany; †Queensland Institute of Medical Research, Brisbane, Queensland, Australia; ‡University of Queensland Centre for Clinical Research, Herston, Brisbane, Queensland, Australia; and §Institute of Human Genetics, University of Würzburg, Germany Faithful chromosome segregation is required for preserving genomic integrity. Failure of this process may entail chromatin bridges preventing normal cytokinesis. To test whether RAD50, a protein normally involved in DNA double-strand break repair, is involved in abnormal cytokinesis and formation of chromatin bridges, we used immunocytochemical and protein interaction assays. RAD50 localizes to chromatin bridges during aberrant cytokinesis and subsequent stages of the cell cycle, either decorating the entire bridge or focally accumulating at the midbody zone. Ionizing radiation led to an ⬃4-fold increase in the rate of chromatin bridges in an ataxia telangiectatica mutated (ATM)-dependent manner in human RAD50-proficient fibroblasts but not in RAD50-deficient cells. Cells with a RAD50-positive chromatin bridge were able to continue cell cycling and to progress through S phase (44%), whereas RAD50 knockdown caused a deficiency in chromatin bridges as well as an ⬃4-fold prolonged duration of mitosis. RAD50 colocalized and directly interacted with Aurora B kinase and phospho-histone H3, and Aurora B kinase inhibition led to a deficiency in RAD50-positive bridges. Based on these observations, we propose that RAD50 is a crucial factor for the stabilization and shielding of chromatin bridges. Our study provides evidence for a hitherto unknown role of RAD50 in abnormal cytokinesis.—Schröder-Heurich, B., Wieland, B., Lavin, M. F., Schindler, D., Dörk, T. Protective role of RAD50 on chromatin bridges during abnormal cytokinesis. FASEB J. 28, 1331–1341 (2014). www.fasebj.org ABSTRACT

Key Words: Aurora B 䡠 DNA damage 䡠 MRN complex 䡠 anaphase bridges 䡠 mitosis

Abbrevations: ATM, ataxia telangiectatica mutated; ATMi, ATM inhbitor; DSB, double-strand break; DAPI, 4,6-diamidino-2-phenylindole; DMSO, dimethylsulfoxide; H3, histone 3; HRR, homology-directed recombinational repair; MRN, MRE11-RAD50-NBN; NBS, Nijmegen breakage syndrome; NGS, normal goat serum; NRS, normal rabbit serum; PBS, phosphate-buffered saline; PFA, paraformaldehyde; PLA, proximity ligation assay; SMC, structural maintenance of chromosome; UFB, ultrafine bridge 0892-6638/14/0028-1331 © FASEB

DNA double-strand lesions constitute the most serious form of DNA damage. Double-strand breaks (DSBs) lead to activation of DNA DSB repair pathways via the ataxia telangiectatica mutated (ATM) kinase and the MRE11-RAD50-NBN (MRN) complex. Repair of DNA DSBs occurs primarily by either homologous recombination or nonhomologous end joining (1–2). Defective repair can result in persisting DSBs, leading to chromosomal fragmentation, or in faulty recombination, leading to multiradial chromosomes or translocations (3– 5); unrepaired DSBs can be lethal for a cell. Sequelae of unrepaired DSBs can also entail chromatin bridges that link sister chromatids during cell division. Chromatin bridges are revealed by conventional DNA dyes and contain missegregated DNA (6 –7). A different class of DNA bridges that interlink sister chromatids during cell division are ultrafine bridges (UFBs) connecting the centromeres of sister chromatids (8). They are not detectable by staining with classical DNA dyes and have been shown to be coated by the Bloom protein BLM (8). At their extremities, FANCD2 and FANCI, two key Fanconi anemia (FA) proteins, are located in early mitosis, while FANCM is recruited to the UFBs at a later stage (9). RAD50 is the core protein of the MRN complex and is a member of the structural maintenance of chromosome (SMC) protein family (1–2, 10). RAD50 plays a role in tethering broken DNA ends by holding them in proximity (11). This function is highly conserved in evolution (12). Deficiency of Rad50 causes embryonic lethality in mice (13–14), whereas a human hyphomorphic mutation gives rise to a Nijmegen breakage syndrome (NBS)-like syndrome, RAD50 deficiency [Online Mendelian Inheritance in Man (OMIM) no. 613078; http://www.omim.org]. At the cellular level, this disorder is characterized by chromosomal instabil1 Correspondence: Hannover Medical School, Gynaecology Research Unit (OE 6411), Carl-Neuberg-Str. 1, D-30625 Hannover, Germany. E-mail: schroeder-heurich. [email protected] doi: 10.1096/fj.13-236984 This article includes supplemental data. Please visit http:// www.fasebj.org to obtain this information.

1331

ity, radiosensitivity, cell cycle abnormalities, and defective DNA damage response (15–16). One possible consequence of chromosomal instability can be the formation of chromatin bridges during mitosis (17), which reflects incompletely segregated chromosomes between 2 dividing daughter cells (18). Such segregation defects can result from dysfunctional telomeres (19) in addition to unrepaired DNA DSBs (4). Chromatin bridges can be fragmented during cytokinesis, but it is still unclear how they can be resolved or whether they are stabilized during cytokinetic progression (8, 20). Chromatin bridges appear to induce a delay in abscission, and this cell division control, dependent on Aurora B kinase, protects against tetraploidization. Aurora B kinase activity is sustained during abnormal cytokinesis, which associates with an apparent stabilization of the chromatin bridge through retention of actin patches (20 –21). However, the mechanisms to regulate the stabilization of chromatin bridges still remain to be resolved. In the present study we report on the identification of RAD50 as a component of chromatin bridges in different murine and human cell types. We show that RAD50 is required for the formation of these bridges and that it colocalizes with cytoskeletal and cytokinetic proteins during abscission and interphase of the cell cycle. We also demonstrate that RAD50 directly interacts with Aurora B and that RAD50 localization on chromatin bridges depends on Aurora B kinase activity. Our results provide evidence that RAD50 acts as an essential stabilization factor during abnormal cytokinesis.

MATERIALS AND METHODS Cell culture Murine immortalized NIH3T3 fibroblasts were obtained from the laboratory of Penelope Jeggo (Genome and Stability Centre, University of Sussex, Brighton, UK) and were cultured in DMEM (Life Technologies, Carlsbad, CA, USA), 10% fetal calf serum (Biochrom, Berlin, Germany), 500 U/ml penicillin (PAA Laboratories, Pasching, Austria), 0.5 mg/ml streptomycin (PAA), and 2 mM l-glutamine (PAA). Human large T antigen-immortalized ADD-T (wild-type) and F239-T (RAD50-deficient) fibroblasts and human hTERT-immortalized BJ5-ta (wild-type) fibroblasts, obtained from American Type Culture Collection (ATCC; Manassas, VA, USA), were maintained in DMEM with 15% fetal calf serum, 500 U/ml penicillin, 0.5 mg/ml streptomycin, and 2 mM l-glutamine. Human HCC1395 and HCC1937 breast cancer cell lines were obtained from ATCC and cultured in RPMI1640 (Life Technologies), supplemented as above. MCF10A breast epithelial cells (wild-type) were obtained from ATCC and were cultured in MEBM (Lonza, Basel, Switzerland), supplemented with MEGM Single Quots (Lonza) according to the manufacturer’s instructions. Wild-type lymphoblastoid cell line HA325 (EBV-immortalized) was cultured in RPMI with 20% fetal calf serum, 500 U/ml penicillin, 0.5 mg/ml streptomycin, and 2 mM l-glutamine. All cells were maintained at 37°C in a humidified atmosphere supplemented with 5% CO2. For treatment with ionizing radiation, cells were irradiated at different doses (6 Gy; 12 Gy) using a Mevatron MD-2 1332

Vol. 28

March 2014

accelerator (Siemens, Munich, Germany). The ATM inhibitor Ku-55933 (ATMi; KuDos Pharmaceuticals, Cambridge, UK) was used at a final concentration of 10 ␮M and was added to the culture medium 1 h prior irradiation. Because Ku-55933 was dissolved in dimethylsulfoxide (DMSO; SigmaAldrich, Steinheim, Germany), the same volume of DMSO was added to the control. Aurora B kinase was inhibited by using barasertib (AZD1152-HQPA). To test the effectiveness of the inhibitor, NIH3T3 cells were grown on cover glasses in 6-well plates (BD Biosciences, San Jose, CA, USA) and treated with different concentrations of barasertib, and the amount of phosphorylated histone 3 (pH3) S10-positive cells was evaluated after 24 h through immunocytochemistry by using a specific antibody. For chromatin bridge evaluation, the cells were irradiated with 6 Gy, incubated for 24 h at 37°C and subsequently were treated with different concentrations of barasertib for another 24 h (37°C) and evaluated by immunocytochemistry. Because barasertib was dissolved in DMSO, the same volume of DMSO was added to the control. siRNA-mediated knockdown of RAD50 NIH3T3 cells were transfected with 180 pmol siRNA targeting mouse RAD50 (Sigma-Aldrich) or an equivalent volume of phosphate-buffered saline (PBS) as control per 0.8 ⫻ 10e5 cells using Metafectene PRO (Biontex Laboratories, Martinsried, Germany) according to the manufacturer’s instructions. After incubation for 72 h, cells were irradiated with 12 Gy, incubated for 24 h, and processed for immunofluorescence analysis. Antibodies For immunofluorescence, the following primary antibodies were used: rabbit anti-pH3 (Ser10; 1:200, 3377; Cell Signaling, Danvers, MA, USA), mouse anti-pH3 (Ser10; 1:200, 9706; Cell Signaling), rabbit anti-RAD50 (1:200, 07-1781; Millipore, Billerica; MA, USA), rabbit anti-RAD50 (1:200, AB3754; Millipore), mouse anti-MRE11 (1:100, GTX70212; GeneTex, Irvine, CA, USA), rabbit anti-MRE11 (1:200, 4895; Cell Signaling), rabbit anti-NBN (1:200, NB100-143; Novus Biologicals, Littleton, CO, USA), rabbit anti-MDC1 (1:100, ab11169; Abcam, Cambridge, UK), rabbit anti-phospho-histone H2AX (S139; 1:200, 2212-1, Epitomics, Burlingame, CA, USA), mouse anti-phospho-histone H2AX (S139, 1:200, 05-636; Millipore), rabbit anti-53BP1 (1:200, 4937; Cell Signaling), mouse anti-Aurora B (1:50, ab3609; Abcam), rabbit antiSMC1 (1:200, A300-055A-3; Bethyl, Montgomery, AL, USA), goat anti-BLM (1:150, sc-7790; Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse anti-␣-tubulin-FITC (1:500; Sigma-Aldrich), and rabbit anti-␣-tubulin (1:500, 2871; Epitomics, Burlingame, CA, USA). The following secondary antibodies were used: rabbit anti-goat Alexa 488 (1:200; Invitrogen/Life Technologies), goat anti-mouse Alexa 546 (1:200; Invitrogen/Life Technologies), goat anti-rabbit Alexa 546 (1:200; Invitrogen/Life Technologies), and goat anti-rabbit Alexa 488 (1:200; Invitrogen/Life Technologies). For coimmunoprecipitation, mouse anti-Aurora B (7 ␮g, ab3609; Abcam) was used. Immunocytochemistry Subconfluent cells grown on cover glasses in 6-well plates (BD Biosciences) were fixed with 3% (w/v) paraformaldehyde (PFA; Sigma-Aldrich) and 2% (w/v) sucrose (Roth, Karlsruhe, Germany) for 10 min. Cells were permeabilized with 0.2% (v/v) TritonX-100 (Sigma-Aldrich) in 1⫻ PBS for 3 min and

The FASEB Journal 䡠 www.fasebj.org

SCHRÖDER-HEURICH ET AL.

rinsed 3⫻ with PBS. Cells were incubated for 1.5-2 h at room temperature with the primary antibody [diluted in 2% w/v normal goat serum (NGS; Dianova, Hamburg, Germany) or normal rabbit serum (NRS; Dianova) in 1⫻ PBS], washed 3⫻ with PBS, and incubated with the secondary antibody (diluted in 2% w/v NGS or NRS in 1⫻ PBS) for 1 h at room temperatur in the dark. Cells were washed with PBS, incubated with 4=, 6-diamidino-2-phenylindole (DAPI; Invitrogen/Life Technologies; 1:50000 in PBS) for 10 min and mounted using Prolong Gold (Invitrogen/Life Technologies). Images were routinely taken by using a Leica DMI 6000B Microscope (Leica, Heidelberg, Germany), and image acquisition was carried out using Corel Photo PaintX4 software (Corel Corp., Ottawa, ON, Canada) and evaluated using GraphPad PrismX4 software (GraphPad, San Diego, CA, USA). For confocal microscopy, immunofluorescence images were taken by using a Leica Inverted-2 confocal microscope (TCS SP2; Leica) with oil-immersion objectives HCX PL APO CS40 (⫻40) or HCX PL APO BL (⫻63) as z stacks with a sequential scan between frames. For scanning, a multiline argon laser (458, 476, 488, and 514 nm), an HeNe laser (543 nm), and an HBO lamp were used. Images were represented as average projections. Acquired images were adjusted using Corel Photo PaintX4. In nuclease treatment experiments, cells grown on cover glasses were irradiated or left untreated. After 24 h, cells were permeabilized with 0.2% (v/v) TritonX-100 in PBS as before and treated with DNaseI (10 U; New England Biolabs, Ipswich, MA, USA) for 10 min at 37°C or with RNaseH (50 U; New England Biolabs) for 20 min at 37°C before immunofluorescence staining. For labeling S-phase cells in some further experiments, EdU detection was performed before antibody staining with the Click-iT EdU Alexa Fluor 555 Imaging Kit (Invitrogen/Life Technologies,) according to the manufacturer’s instructions. Time-lapse microscopy Live cell imaging to study cell cycle progression was performed using a Leica DMI 6000B microscope equipped with Incubator BL (Leica) for heating (37°C) and CO2 supply (5%). Cells were cultured in 6-well plates and imaging was directly started after irradiation or 72 h after transfection of siRNA. Images were taken by using phase-contrast optics with a ⫻20 objective (L40 ⫻ PH2) every 8 min for a total imaging time of 48 h. The acquired images were analyzed using Leica Application Suite 1.9.0, Corel Photo PaintX4, and GraphPad PrismX4. Proximity ligation assay (PLA) PLA (Olink Bioscience, Uppsala, Sweden) was used to analyze protein-protein interactions. Briefly, cells grown in 6 well plates were fixed with 3% (w/v) PFA, 2% (w/v) sucrose (Roth) for 10 min, permeabilized with 0.2% Triton X-100, washed 3⫻ with PBS, and incubated with the indicated primary antibodies in 2% NGS for 2 h. PLA detection was performed according to the manufacturer’s instructions. As a positive control, mouse anti-MRE11 (GTX70212; GeneTex, Irvine, CA, USA) and rabbit anti-RAD50 (07–1781; Millipore) antibody were used. F239-T was included as a negative control. Images were taken by using a Leica DMI 6000B microscope, and image acquisition was carried out using Corel Photo PaintX4. Coimmunoprecipitation For coimmunoprecipitation in human fibroblasts (BJ5-ta), protein A and protein G magnetic beads (Invitrogen/Life RAD50 ON CHROMATIN BRIDGES

Technologies) were washed 3 times with 1⫻ PBS, 0.02% Tween-20 (Sigma-Aldrich) and were incubated with the specific antibody for 2 h at 4°C. Beads were washed 3⫻ with 1⫻ PBS and 0.02% Tween-20 and incubated with 1 mg of protein lysate for 2 h at 4°C. After being washed 3⫻ with 1⫻ PBS, the beads were processed for SDS-PAGE and immunoblotting. Coimmunoprecipitates from the human lymphoblastoid cell line HA325 were performed with protein G-Sepharose beads (Sigma-Aldrich). After spinning of the beads (1 min, 16,100 rcf), the supernatant was discarded, and the beads were washed 3⫻ with PBS and 0.02% Tween-20. Beads were incubated with the specific antibody for 1 h at 4°C, washed (PBS and 0.02% Tween-20), and incubated with 1 mg of protein lysate for 2 h at 4°C. After being washed with PBS and 0.02% Tween-20, the beads were sedimented by centrifugation (1 min, 16,100 rcf) and processed for SDS-PAGE and immunoblotting. Lysate preparation and Immunoblotting Briefly, cells were lysed in cell extraction buffer (50 mM Tris, pH 7.4; Merck, Darmstadt, Germany), 150 mM NaCl (Merck), 2 mM ethylene glycol tetraacetic acid (EGTA; Sigma-Aldrich), 2 mM ethylene diamine tetraacetic acid (EDTA; SigmaAldrich), 25 mM NaF (Sigma-Aldrich), 0.1 mM Na3VO4 (Sigma-Aldrich), 0.1 mM phenylmethanesulfonylfluoride (PMSF; Sigma-Aldrich), 2 mg/ml leupeptin (Serva Feinbiochemika, Heidelberg, Germany), 2 mg/ml aprotinin (Serva Feinbiochemika), 0.2% Triton X-100, and 0.3% Nonidet P-40 (Sigma-Aldrich) for 30 min on ice and centrifuged at 16,100 rcf for 15 min. Protein extracts were separated through SDS-PAGE and immunoblotting. The following primary antibodies were used: mouse anti-RAD50 (1:500, ab89; Abcam) and mouse anti-␤-actin (1:3000, A5541; Sigma-Aldrich). Antimouse IgG horseradish peroxidase-labeled secondary antibody (1:5000, NA9310; GE Healthcare, Little Chalfont, UK) and ECL (Thermo Scientific/Pierce, Rockford, IL, USA) were used for visualization of immunoreactive bands.

RESULTS We have previously shown that cells from a patient with RAD50 deficiency exhibit high levels of spontaneous chromosomal instability (15). For analyzing possible chromosome segregation defects in RAD50-deficient cells, we used a large T-antigen-transformed cell line from human F239 fibroblasts, originally derived from this patient who had an NBS-like syndrome. We confirmed by immunoblotting that there is a residual level of RAD50 protein in these cells, ⬍5% of that in wild-type cells (Fig. 1A). We irradiated these cells with either 6 or 12 Gy and initially analyzed them for the occurrence of chromatin bridges after 24 h. F239-T (RAD50-deficient) cells displayed a higher basal level of 3.7% of cells with a chromatin bridge compared to 2% in RAD50-wild-type cells (ADD-T). While exposure of wild-type cells to radiation led to an ⬃4-fold increase in chromatin bridges, there was no additional increase in RAD50-deficient cells (Fig. 1B). An assessment of the mitotic index of both cell lines did not reveal noticeable differences between untreated and irradiated cells after 24 h (Supplemental Fig. 1A), so that this finding could not be attributed to aberrant cell cycle arrest after irradiation. These results indi1333

Figure 1. RAD50 localization on chromatin bridges. A) Reduced protein levels of RAD50 in RAD50-deficient (F239-T, right lane) compared with wild-type fibroblasts (ADD-T, left lane). B, E) Quantitative assessment of DAPI⫹ and RAD50⫹ bridges. B) Human wild-type fibroblasts (ADD-T) but not RAD50deficient fibroblasts (F239-T) showed a significant increase of cells with DAPI⫹ chromatin bridges 24 h after irradiation (6 Gy; 12 Gy). ***P ⬍ 0.001. C, D) RAD50 was either forming a short bridge (C) or decorating the whole bridge in a dot wise pattern (D). Right panels present enlarged view of boxed areas in left panels. Scale bars ⫽ 20 ␮m. E) Fraction of RAD50⫹ bridges 24 h after irradiation was increased ⬃3-fold in wild-type fibroblasts (ADD-T) but remained unaltered in RAD50-deficient fibroblasts (F239-T). UNT, untreated. **P ⬍ 0.01.

cated that RAD50 protein is important for the formation of chromatin bridges. To further validate the possible connection between chromatin bridge formation and RAD50, we investigated the subcellular localization of RAD50 in cells connected through a chromatin bridge. Immunocytochemistry using conventional DNA staining (DAPI) and RAD50 staining showed a strikingly consistent localization of RAD50 on chromatin bridges (Fig. 1C, D). More than ⬃86% of chromatin bridges showed a localization of RAD50, whereas in ⬃14% of bridged wild-type cells, the RAD50 localization on bridges was not detectable. We defined those bridges that were positive for both DAPI staining and RAD50 localization as RAD50-positive (RAD50⫹) chromatin bridges, and bridges that were positive for DAPI staining alone as DAPI-positive (DAPI⫹) chromatin bridges. We observed 2 different patterns of RAD50 distribution coincident with DAPIstained chromatin bridges (Fig. 1C, D). In some instances, RAD50 staining was found to be densely and focally pinched in the central part of the chromatin bridge (Fig. 1C). In other cells, RAD50 staining showed up like a dotwise pattern, decorating the whole bridge (Fig. 1D). When we determined the frequency of cells with RAD50⫹ bridges induced by radiation, the data were similar to those for chromatin bridges with controls showing a marked induction whereas 1334

Vol. 28

March 2014

RAD50-deficient cells showed no significant increase (Fig. 1E). We tested further fibroblast and epithelial cell lines and observed RAD50⫹ bridge formation in normal murine fibroblasts (NIH3T3; Fig. 2A) as well as in human epithelial breast cancer cell lines (HCC1937 and HCC1395; Fig. 2B) and in wild-type human breast epithelial cells (MCF10A; Fig. 2B). To determine whether RAD50⫹ bridges are dependent on the presence of chromatin, we treated cells with DNaseI for a short incubation time (10 U, 10 min), which resulted in a significant reduction to ⬃47% of RAD50⫹ bridge cells (Fig. 2A). By contrast, treatment with RNAseH showed no decrease in RAD50⫹ bridges (Supplemental Fig. 1C). These data corroborated that RAD50 marks intercellular bridges in diverse cell types in dependence of chromatin. Because ATM is crucial for the intracellular radiation response, we investigated the extent to which formation of chromatin bridges depends on ATM kinase activity. NIH3T3 cells were treated with the ATM inhibitor (ATMi) Ku-55933 before irradiation with 12 Gy and examined for chromatin and RAD50⫹ bridge formation at 24 h after irradiation. Cells with DAPI-stained chromatin bridges after ATMi treatment were ⬃2-fold reduced (49%) compared with irradiated cells without ATMi treatment (Fig. 2C). More specifically, RAD50

The FASEB Journal 䡠 www.fasebj.org

SCHRÖDER-HEURICH ET AL.

Figure 2. RAD50 and ATM are involved in chromatin bridge formation after irradiation. A) DNaseI treatment before immunofluoresce nce staining resulted in an ⬃2-fold decrease of cells with a RAD50⫹ bridge; n ⫽ 3. Data are presented as means ⫾ se. *P ⬍ 0.05; 2-tailed Student’s t test. B) Radiation-induced increase of RAD50⫹ bridges in BRCA1-deficient human breast cancer epithelial cells (HCC1937, HCC1395) and wild-type mammary epithelial cells (MCF10A) at 30 min, 24 h, or 48 h after irradiation with 6 Gy. Experiments were performed in quadruplicates (n⬎300 cells/case). C) Inhibition of ATM kinase with KU-55933 was associated with a significant decrease of cells with RAD50⫹ or DAPI⫹ chromatin bridges. Experiments were performed in triplicates (n⬎300 cells/case). **P ⬍ 0.01; 2-tailed Student’s t test. D) Transfection control of RAD50 siRNA knockdown in NIH3T3 cells by immunoblotting. E) An ⬃7-fold increase of cells with a RAD50⫹ bridge after irradiation (12 Gy, 24 h) of NIH3T3 cells. Knockdown of RAD50 led to a significant reduction of cells with a RAD50⫹ chromatin bridge. Experiments were performed in triplicates (n⬎300 cells/case). Data are presented as means ⫾ se. *P ⬍ 0.05; 2-tailed Student’s t test.

staining after ATM inhibition revealed a reduction by 63% of cells with a RAD50⫹ bridge. To confirm the RAD50 dependency of chromatin bridge formation, we performed siRNA knockdown of RAD50 by transfection (Fig. 2D), which led to a 3-fold decrease in cells with a RAD50⫹ chromatin bridge in comparison to mocktransfected irradiated cells (Fig. 2E), indicating that RAD50 is crucial for bridge formation. We next tested whether cell lines deficient in other proteins of the DNA DSB repair pathway are similarly impaired in chromatin bridge formation. Analysis of 2 breast cancer epithelial cell lines that were deficient in p53 and BRCA1 showed a 2- to 3-fold increase of RAD50⫹ bridges after 24 h and a 3- to 6-fold increase after 48 h (Fig. 2B), indicating that neither p53 nor BRCA1 is essential for bridge formation. Previous studies have detected phospho-Ser139 H2A.X, referred to as ␥H2AX, on chromatin bridges, suggesting that they contain unrepaired DSBs (22). To further evaluate whether RAD50 marks DSBs on chromatin bridges as part of the MRN complex, we tested by immunostaining the localization of its interacting partners NBN and MRE11, as well as the DSB markers H2A.X and 53BP1, and 3 additional proteins of the ATM pathway, SMC1, MDC1, and BLM. None of the RAD50-interacting proteins MRE11, NBN, MDC1, or SMC1 were specifically detected on chromatin bridges in these experiments, although MRE11 and NBN appeared to be distributed within cytoplasmic compartments including intercellular bridges (Supplemental RAD50 ON CHROMATIN BRIDGES

Fig. S3A, B). Similarly, BLM did not focally stain chromatin bridges (Supplemental Fig. S3C). In some instances, ␥H2AX and 53BP1 were found localized at the edges of the chromatin bridge (Supplemental Fig. S4A). Additional double-immunostaining experiments targeting ␥H2AX and RAD50 revealed that these proteins were not colocalizing on chromatin bridges, even when both were present (Supplemental Fig. 4B), suggesting that RAD50 is not merely a marker of DNA DSBs. Because chromosome segregation requires the cytoskeleton and mitotic kinases, we next asked whether the RAD50 pattern on chromatin bridges was related to cytoskeletal and cytokinetic proteins. Immunostaining against the cytoskeletal protein ␣-tubulin showed colocalization with RAD50 at the dividing plane during abscission in cells connected through a chromatin bridge (Fig. 3A). The RAD50 signal appeared to increase from early to late abscission in cytokinetic cells and was sustained even at the midbody remnant and in interphase cells where RAD50 decorated the whole bridges (Fig. 3A). A mitotic kinase that had previously shown sustained activity at the midbody in cells with trapped chromatin is Aurora B (21). To further address whether RAD50 interacts with Aurora B in abnormal cytokinesis, we analyzed ADD-T fibroblasts by immunostaining against Aurora B. RAD50 colocalized with Aurora B on chromatin bridges during abscission and at the midbody remnant (Fig. 3B). In interphase cells, RAD50 immunostaining was distributed over the whole 1335

Figure 3. RAD50 colocalization with ␣-tubulin and Aurora B at different stages of abnormal cytokinesis and direct interaction with Aurora B in situ. A) Representative immunofluorescence images showing the localization of RAD50 (red) and ␣-tubulin (green) in different stages of cytokinesis in NIH3T3 cells with trapped chromatin. Colocalization of RAD50 and ␣-tubulin was detectable during early abscission, late abscission, at the midbody remnant, and in interphase cells (open arrowheads). B) Immunofluorescence double staining with RAD50 (green) and Aurora B (red) showing a lack of RAD50 signal at the midbody zone in normal cytokinetic cells (ADD-T), whereas RAD50 and Aurora B colocalized during abnormal cytokinesis and at the midbody remnant. RAD50 decorated the entire chromatin bridge in interphase cells (open arrowheads). Right panels show enlargement of the areas indicated by open arrowheads in left panels. DNA is counterstained with DAPI (blue). Scale bar ⫽ 20 ␮m. C) PLA performed on ADD-T fibroblasts. Cells were fixed and PLA stained for RAD50 and Aurora B, detecting abundant PLA dots in interphase cells. RAD50 and Aurora B resided in proximity in untreated ADD-T cells (top panel) and 24 h after irradiation (6 Gy; bottom panel). Scale bars ⫽ 20 ␮m. D) Focal localization as a single PLA dot in the center of a chromatin bridge (left panel, white arrow) and enlarged view (right panel). All images are from the same biological sample. Scale bar ⫽ 20 ␮m. E) Coimmunoprecipitation of Aurora B and RAD50. Cell extracts of human fibroblasts (BJ5-ta) or lymphoblastoid cells (HA325) were immunoprecipitated (IP) with Aurora B antibody, separated on SDS-PAGE, and immunoblotted against specific Aurora B and RAD50 antibodies. F, G) Dependence of RAD50 localization on Aurora B kinase activity. Murine NIH3T3 cells were irradiated with 6 Gy, incubated for 24 h at 37°C, treated with different doses (5, 10, 20, and 50 nM) of Aurora B kinase inhibitor barasertib (AZD1152-HQPA), and incubated for a further 24 h. Cells were fixed, immunostained against RAD50, and stained for DAPI. Inhibition of Aurora B kinase was associated with a significant decrease of cells with a RAD50 localizaton on chromatin bridges (F) though not significantly with DAPI⫹ chromatin bridges (G). Experiments were performed in triplicates (n⬎200 cells/case). UNT, untreated; DMSO, DMSO-treated control cells. Data are presented as means ⫾ se. *P ⬍ 0.05, **P ⬍ 0.01; 2-tailed Student‘s t test.

chromatin bridge in the absence of Aurora B. On the other hand, normally dividing cells that were stained positive for Aurora B at the midbody zone were lacking a RAD50 signal. These results indicated that RAD50 and Aurora B may specifically synergize on chromatin bridges in abnormal cytokinesis. We used the PLA to investigate direct protein-protein interactions of RAD50 and Aurora B, in giving a 1336

Vol. 28

March 2014

positive PLA signal when both proteins reside in proximity (⬃40 nm, according to the manufacturer). PLA was performed on ADD-T (wild-type) and F239-T (RAD50-deficient) cells using a rabbit anti-RAD50, and either mouse anti-MRE11 or mouse anti-Aurora B antibody, respectively. As shown in Fig. 3C, we found abundant positive PLA signals suggesting proximity and interaction of RAD50 and Aurora B in both untreated

The FASEB Journal 䡠 www.fasebj.org

SCHRÖDER-HEURICH ET AL.

and irradiated interphase cells. In a few cells, single dots indicating RAD50-Aurora B interaction could be detected at the center of a chromatin bridge (n⫽5; Fig. 3D). For control of specificity, any positive PLA signal for either MRE11 or Aurora B was missing in F239-T cells, as anticipated for the case of RAD50 deficiency. The interaction of RAD50 and Aurora B was further confirmed by coimmunoprecipitation of both proteins in 2 different cell lines (Fig. 3E). Because Aurora B kinase is crucial for mitotic processes and has been found to have a sustained kinase activity at chromatin bridges (20, 21), we investigated whether RAD50 localization on chromatin bridges is dependent on the kinase activity of Aurora B. The effectiveness of Aurora B inhibition was confirmed by determining pH3 S10 phosphorylation in NIH3T3 fibroblasts (Supplemental Fig. S2). Cells treated with barasertib at different concentrations resulted in decreasing and significantly reduced numbers of RAD50⫹ bridges (Fig. 3F), whereas the total numbers of chromatin bridges was not significantly affected (Fig. 3G). These data indicate a dependency on Aurora B kinase activity for RAD50 localization on preformed chromatin bridges. Chromatin bridges can either get resolved or can lead to binucleated or tetranucleated cells (21). We

therefore asked to what extent cells with a RAD50 signal on the chromatin bridge were able to keep on with cell cycle progression. To specifically investigate actively replicating cells in S phase, we grew cells in EdU-containing medium for a short incubation time (20 min) before fixation. EdU and RAD50 costaining uncovered several bridged cells, of which at least 1 had progressed through S phase and stained positive for RAD50 (Fig. 4A). Almost half (21/48, 44%) of RAD50⫹ bridged cell pairs were in S phase. Additional staining against phospho (Ser10) histone H3 revealed that some 3% of all cells with a bridge were marked as mitotic (Supplemental Fig. 1B). Interestingly, phospho (Ser10) histone H3 was detected on chromatin bridges even in interphase cells, and double immunostaining with mouse anti-pH3 S10 and rabbit anti-RAD50 antibodies confirmed that these two proteins colocalized on chromatin bridges (Fig. 4B). Investigation of protein-protein interactions via PLA showed abundant positive PLA signals for pH3/RAD50 in interphase and at the center of chromatin-bridged cells (Fig. 4C). To investigate how the presence of RAD50 affects mitotic cell division, we performed time-lapse microscopy to analyze mitotic cell division time in untreated, irradiated, or RAD50 siRNA-knockdown NIH3T3 cells, as well as in RAD50-deficient and RAD50-proficient human fi-

Figure 4. Recycling of RAD50⫹ bridged cells and interaction of RAD50 with pH3 in situ. A) Representative immunofluorescence staining with RAD50 and EdU in NIH3T3 cells. Cells with a chromatin bridge were quantified as positive for S phase with ⱖ1 EdU positive, and 48 pairs of chromatin-bridged cells (n⫽96 cells) were evaluated. B) Double immunostaining in ADD-T fibroblasts with the mitotic marker pH3 S10 and RAD50, showing colocalization of both proteins on a bridge. C) PLA performed on ADD-T fibroblasts after cells were fixed and PLA stained for RAD50 and pH3 S10, detecting abundant PLA dots in interphase cells and a focal localization at the center of chromatin-bridged cells. Scale bars ⫽ 20 ␮m. RAD50 ON CHROMATIN BRIDGES

1337

broblasts (Fig. 5). Cell division was followed from onset of prometaphase until complete segregation of the 2 dividing daughter cells (Fig. 5A). Untreated NIH3T3 cells had a median division time of 104 min (n⫽22), whereas irradiated cells showed a median division time of 174 min (n⫽32). Sixty-five percent of irradiated cells underwent a prolonged cell division during which the daughter cells remained connected through a cytoplasmic bridge, whereas some 22% of the cells failed to divide (Fig. 5B). NIH3T3 fibroblasts depleted of RAD50 by siRNA transfection showed a median division time of 380 min (n⫽23; Fig. 5B). In line with these results, cell division time in RAD50-mutant human fibroblasts (F239-T) also showed a markedly prolonged median division time of 344 min (n⫽15) in comparison to RAD50-proficient human fibroblasts (ADD-T; median division time 272 min, n⫽19; Fig. 5C), indicating that RAD50 strongly impacts on the timely progression through mitosis and cytokinesis.

DISCUSSION Chromatin bridges arise from sister chromatids that fail to completely segregate and are a rare mitotic event but consistently observed in mammalian cells with cytokinetic failure. Previous investigations of chromatin

bridges in tumor cells have shown that bridged cells can undergo different fates including apoptosis, fusion to form multinucleated cells and breakage or resolution of the bridge (22). However, the molecular mechanisms behind chromatin bridge stabilization or resolvement are not well established to date. RAD50 is a highly conserved, essential protein that, as part of the MRN complex, participates in faithful DNA damage signaling and repair and thus ensures the maintenance of chromosome stability (23–25). In the present study, we show that RAD50 is also required for the formation and maintenance of chromatin bridges after ionizing radiation in mammalian cells. Previous studies have found that cells defective in homology-directed recombinational repair (HRR) as well as ATM-deficient cells display an increased level of chromatin bridges after irradiation (4, 22) suggesting that chromatin bridge formation is due to unrepaired DNA DSBs. Unexpectedly, our studies of RAD50 indicated a decrease of bridge formation after RAD50 knockdown in murine NIH3T3 cells and no bridge induction in RAD50deficient human fibroblasts after irradiation. On the other hand, breast cancer cells defective in homologydirected recombinational repair (BRCA1⫺/⫺; TP53⫺/⫺) were clearly proficient in chromatin bridge formation,

Figure 5. Time-lapse microscopy showed a prolonged cell division time in RAD50-knockdown, RAD50-deficient, and irradiated cells. Irradiated (12 Gy) or RAD50-depleted NIH3T3 cells both underwent a prolonged cell division time. A) Cell division time was evaluated from onset of pro/metaphase (a, b) until abscission of the 2 dividing daughter cells at the midbody (h, arrowhead). After onset of ana-/telophase (c), irradiated dividing cells were frequently connected through a cytoplasmic bridge (d, e, f, g; arrows). Insets bordered in red panels show enlargements of the areas circled in red. B) Quantification of division time in untreated (n⫽22) and irradiated (12 Gy; n⫽32) NIH3T3 cells and siRNA RAD50-knockdown cells (n⫽23). Median division times were 104 min for untreated cells, 174 min after irradiation with 12 Gy, and 380 min after RAD50 depletion by transfection. C) Median cell division time in RAD50-deficient fibroblasts (F239-T; n⫽15) was 344 min, while RAD50-proficient fibroblasts (ADD-T) showed a median divison time of 272 min (n⫽19). 1338

Vol. 28

March 2014

The FASEB Journal 䡠 www.fasebj.org

SCHRÖDER-HEURICH ET AL.

arguing for a specific role of RAD50 independent of HRR. Disrupting ATM kinase function by chemical inhibition revealed a decrease in bridges in our model system, indicating that ATM is required for either the formation or maintenance of chromatin bridges, perhaps via an effect on RAD50 which is a phosphorylation target of ATM (16). Collectively, the results indicated a hitherto unknown role of RAD50 for the presence of chromatin bridges. Strikingly, RAD50 protein itself was localized on chromatin bridges and stained those in different cell types, among those large T-transformed human fibroblasts as well as in spontaneously immortalized murine fibroblasts and breast epithelial and breast cancer cells. With the use of RAD50 immunocytochemistry, these bridges could be observed in unperturbed cells but their frequency significantly increased after irradiation. It is unlikely that RAD50⫹ bridges simply represent DNA damage, as we did not observe a specific staining of the bridges for MRE11 or NBN, its binding partners within the MRN complex, or for MDC1 which also localizes to DNA breaks. By contrast, RAD50 colocalized with Aurora B and ␣-tubulin in cytokinetic cells with trapped chromatin. The direct interaction of Aurora B with RAD50, as detected by the PLA and coimmunoprecipitation, may play a role in abnormal cytokinesis after DNA damage has occurred. In a previous study by Steigemann et al. (21), a delay in abscission by sustained Aurora B kinase activity was linked to the prevention of tetraploidization. In budding yeast, the inhibition of premature abscission ensures that trapped chromatin is cleared from the midbody zone (26). It is thus possible that the interaction between Aurora B and pH3 with RAD50 augments the resolution of chromatin bridges. Cells with persistent bridges appeared to further proceed through the cell cycle and were active in S phase in the presence of chromatin at the midbody zone. It is possible that RAD50 acts as a shielding protein to ensure the maintenance of bridges. Moreover, the mitotic period in RAD50-depleted cells and in RAD50-deficient cells was prolongated, indicating a role for RAD50 in mitotic and cytokinetic progression. This adds the RAD50 protein to other known components of DNA repair pathways, such as the Fanconi anemia proteins, that protect cells from cytokinesis failure (9, 27). Based on the previous results, there may be at least two explanations, not mutually exclusive, for RAD50 function on chromatin bridges. First, RAD50 gets recruited to DNA damage sites as part of the DNA damage repair machinery and remains chromatinbound while tethering broken DNA ends. When repair is not sufficient or in time, it stays as a stabilization factor on chromatin bridges to help cells in continuing cell cycle progression and complete repair at later stages. During that time, RAD50 may ensure the proximity of sister chromatids to allow for error-free repair. While sister chromatids are usually tethered through the cohesin complex, the onset of anaphase involves cohesin dissociation by separase (28). A similar course RAD50 ON CHROMATIN BRIDGES

of events has also been proposed to take place for de novo loaded cohesin in postreplicative DNA repair to promote accessibility to repair factors and restore DNA integrity (29). The protective role of RAD50 may be part of a more extensive shielding of damaged sites throughout later cell cycle stages, which is supported by our observation of a sustained pH3 signal and colocalization with RAD50 on bridges, indicating that chromatin bridges failed to complete a normal mitotic dephosphorylation program. If RAD50 marks damage sites, its presence on chromatin bridges could be indicative of DNA DSBs arising from heterochromatic regions. This could explain the lack of colocalization between ␥H2AX and RAD50 on the same bridge, since heterochromatin sites are repaired with slower kinetics and ␥H2AX foci rarely detectable at such sites (30). While this is a conceivable scenario, chromatin retention of RAD50 at sites of DNA damage in the context of the MRN complex is usually mediated by phospho-dependent interactions between NBN and MDC1 (31), so that the absence of detectable MRE11 or MDC1 foci in conjunction with RAD50 on bridges may argue against a bridge-associated function in classical DSB repair. However, RAD50 can act without MRE11 in chromatin bound or free states (32). An alternative explanation would thus be that RAD50 is specifically located to chromatin bridges in assisting the resolution of the bridges at cytokinetic progression to stabilize the intercellular canals and this function might be in concert with Aurora B and phosphorylated histone H3. RAD50 protomers are capable of longdistance communications through their coiled-coil domains that end in a conserved hook-shaped domain (33–34) by which they could form intermolecular complexes or interlocking loops via cysteine-mediated zinc ion coordination to tether 2 DNA molecules (1, 35–36). The coiled-coil domains are estimated as large as 50 nm (37), providing RAD50 with the molecular capability to bridge remarkably large distances up to 100 nm, which may aid its role during cytokinesis. The Aurora B-mediated phosphorylation of histone H3 on Ser10 is required for proper chromosome condensation and segregation (38) and is associated with recruitment of members of the SMC proteins onto mitotic chromosomes (39 – 41). It is possible that the structurally related RAD50 takes part in this process. In line with this hypothesis, RAD50-depleted NIH3T3 fibroblasts exhibited a pronounced delay in the mitotic duration between prometaphase onset and cytokinetic abscission. A similar observation has been made recently in Xenopus eggs where MRE11 knockdown, probably affecting RAD50 levels, or inhibition with the MRN inhibitor mirin caused a significant metaphase delay (42). Collectively, these data suggest a previously unanticipated role of RAD50 in the regulation of mitosis. Based on our results, we suggest that RAD50 might be important for both requirements, ensuring the integrity as well as shielding and resolution of chromatin bridges. We found RAD50 in 2 different patterns on chromatin bridges during different cytokinetic stages 1339

and even in aneuploid interphase cells, and only the centrally localized pattern was associated with Aurora B colocalization. In summary, we have shown that RAD50 localizes to DNA on chromatin bridges and that it is necessary for the presence of these bridges. RAD50 was focally pinched on bridges in cytokinetic cells, where it colocalized with Aurora B kinase, and stained bridges in a dotwise pattern in interphase cells. Cells with a RAD50⫹ bridge were able to proceed with cell cycle progression, suggesting a protective role for RAD50 on chromatin bridges. Altogether, these data document a novel role for RAD50 in abnormal cytokinesis, which might contribute to its essential functions in tumor suppression and development.

10.

11.

12.

13.

14.

The authors thank Julia Menzel, Martin Werner, Jörg Frühauf, Johann Hinrich Karstens, and Hans Christiansen (Department of Radiation Oncology), Peter Hillemanns (Department of Obstetrics and Gynaecology), and Wolfgang Posselt (Confocal Microscopy Center) for support at Hannover Medical School. Further, the authors thank Mark O’Driscoll and Penelope Jeggo for initial support and for providing NIH3T3 cells. Author contributions: B.S.H., M.F.L., D.S., and T.D. participated in the conception and design of the study; B.S.H. and B.W. performed data acquisition; B.S.H. performed data analyses; B.S.H. and T.D. drafted the manuscript; all authors read and approved the final manuscript. The authors declare no conflicts of interest.

2.

3.

4. 5.

6. 7.

8.

9.

1340

De Jager, M., Dronkert, M. L., Modesti, M., Beerens, C. E., Kanaar, R., and van Gent, D. C. (2001) DNA-binding and strand-annealing activities of human Mre11: implications for its roles in DNA double-strand break repair pathways. Nucleic Acids Res. 29, 1317–1325 Kinoshita, E., van der Linden, E., Sanchez, H., and Wyman, C. (2009) RAD50, an SMC family member with multiple roles in DNA break repair: how does ATP affect function? Chromosome Res. 17, 277–288 Pfeiffer, P., Goedecke, W., Kuhfittig-Kulle, S., and Obe, G. (2004) Pathways of DNA double-strand break repair and their impact on the prevention and formation of chromosomal aberrations. Cytogenet. Genome Res. 104, 7–13 Acilan, C., Potter, D. M., and Saunders, W. S. (2007) DNA repair pathways involved in anaphase bridge formation. Genes Chromosomes Cancer 46, 522–531 Gisselsson, D., Jonson, T., Petersen, A., Strombeck, B., Dal Cin, P., Hoglund, M., Mitelman, F., Mertens, F., and Mandahl, N. (2001) Telomere dysfunction triggers extensive DNA fragmentation and evolution of complex chromosome abnormalities in human malignant tumors. Proc. Natl. Acad. Sci. U. S. A. 98, 12683–12688 Olive, P. L. (1998) The role of DNA single- and double-strand breaks in cell killing by ionizing radiation. Radiat. Res. 150, S42–51 Tamulevicius, P., Wang, M., and Iliakis, G. (2007) Homologydirected repair is required for the development of radioresistance during S phase: interplay between double-strand break repair and checkpoint response. Radiat. Res. 167, 1–11 Chan, K. L., North, P. S., and Hickson, I. D. (2007) BLM is required for faithful chromosome segregation and its localization defines a class of ultrafine anaphase bridges. EMBO J. 26, 3397–3409 Vinciguerra, P., Godinho, S. A., Parmar, K., Pellman, D., and D’Andrea, A. D. (2010) Cytokinesis failure occurs in Fanconi anemia pathway-deficient murine and human bone marrow hematopoietic cells. J. Clin. Invest. 120, 3834 –3842

Vol. 28

March 2014

16.

17. 18.

REFERENCES 1.

15.

19.

20. 21.

22.

23.

24.

25. 26.

27. 28.

Tittel-Elmer, M., Lengronne, A., Davidson, M. B., Bacal, J., Francois, P., Hohl, M., Petrini, J. H., Pasero, P., and Cobb, J. A. (2012) Cohesin association to replication sites depends on rad50 and promotes fork restart. Mol. Cell. 48, 98 –108 Hopfner, K. P., Craig, L., Moncalian, G., Zinkel, R. A., Usui, T., Owen, B. A., Karcher, A., Henderson, B., Bodmer, J. L., McMurray, C. T., Carney, J. P., Petrini, J. H., and Tainer, J. A. (2002) The Rad50 zinc-hook is a structure joining Mre11 complexes in DNA recombination and repair. Nature 418, 562–566 Dolganov, G. M., Maser, R. S., Novikov, A., Tosto, L., Chong, S., Bressan, D. A., and Petrini, J. H. (1996) Human Rad50 is physically associated with human Mre11: identification of a conserved multiprotein complex implicated in recombinational DNA repair. Mol. Cell. Biol. 16, 4832–4841 Luo, G., Yao, M. S., Bender, C. F., Mills, M., Bladl, A. R., Bradley, A., and Petrini, J. H. (1999) Disruption of mRad50 causes embryonic stem cell lethality, abnormal embryonic development, and sensitivity to ionizing radiation. Proc. Natl. Acad. Sci. U. S. A. 96, 7376 –7381 Zhu, J., Petersen, S., Tessarollo, L., and Nussenzweig, A. (2001) Targeted disruption of the Nijmegen breakage syndrome gene NBS1 leads to early embryonic lethality in mice. Curr. Biol. 11, 105–109 Waltes, R., Kalb, R., Gatei, M., Kijas, A. W., Stumm, M., Sobeck, A., Wieland, B., Varon, R., Lerenthal, Y., Lavin, M. F., Schindler, D., and Dork, T. (2009) Human RAD50 deficiency in a Nijmegen breakage syndrome-like disorder. Am. J. Hum. Genet. 84, 605–616 Gatei, M., Jakob, B., Chen, P., Kijas, A. W., Becherel, O. J., Gueven, N., Birrell, G., Lee, J. H., Paull, T. T., Lerenthal, Y., Fazry, S., Taucher-Scholz, G., Kalb, R., Schindler, D., Waltes, R., Dork, T., and Lavin, M. F. (2011) ATM protein-dependent phosphorylation of Rad50 protein regulates DNA repair and cell cycle control. J. Biol. Chem. 286, 31542–31556 Hoffelder, D. R., Luo, L., Burke, N. A., Watkins, S. C., Gollin, S. M., and Saunders, W. S. (2004) Resolution of anaphase bridges in cancer cells. Chromosoma 112, 389 –397 Laulier, C., Cheng, A., and Stark, J. M. (2011) The relative efficiency of homology-directed repair has distinct effects on proper anaphase chromosome separation. Nucleic Acids Res. 39, 5935–5944 Stewenius, Y., Gorunova, L., Jonson, T., Larsson, N., Hoglund, M., Mandahl, N., Mertens, F., Mitelman, F., and Gisselsson, D. (2005) Structural and numerical chromosome changes in colon cancer develop through telomere-mediated anaphase bridges, not through mitotic multipolarity. Proc. Natl. Acad. Sci. U. S. A. 102, 5541–5546 Chen, C. T., and Doxsey, S. (2009) A last-minute rescue of trapped chromatin. Cell 136, 397–399 Steigemann, P., Wurzenberger, C., Schmitz, M. H., Held, M., Guizetti, J., Maar, S., and Gerlich, D. W. (2009) Aurora B-mediated abscission checkpoint protects against tetraploidization. Cell 136, 473–484 Huang, H., Fletcher, L., Beeharry, N., Daniel, R., Kao, G., Yen, T. J., and Muschel, R. J. (2008) Abnormal cytokinesis after X-irradiation in tumor cells that override the G2 DNA damage checkpoint. Cancer Res. 68, 3724 –3732 Hari, F. J., Spycher, C., Jungmichel, S., Pavic, L., and Stucki, M. (2010) A divalent FHA/BRCT-binding mechanism couples the MRE11-RAD50-NBS1 complex to damaged chromatin. EMBO Rep. 11, 387–392 Williams, G. J., Lees-Miller, S. P., and Tainer, J. A. (2010) Mre11-Rad50-Nbs1 conformations and the control of sensing, signaling, and effector responses at DNA double-strand breaks. DNA Repair (Amst.) 9, 1299 –1306 Stracker, T. H., and Petrini, J. H. (2011) The MRE11 complex: starting from the ends. Nat. Rev. Mol. Cell Biol. 12, 90 –103 Norden, C., Mendoza, M., Dobbelaere, J., Kotwaliwale, C. V., Biggins, S., and Barral, Y. (2006) The NoCut pathway links completion of cytokinesis to spindle midzone function to prevent chromosome breakage. Cell 125, 85–98 Daniels, M. J., Wang, Y., Lee, M., and Venkitaraman, A. R. (2004) Abnormal cytokinesis in cells deficient in the breast cancer susceptibility protein BRCA2. Science 306, 876 –879 Uhlmann, F., Lottspeich, F., and Nasmyth, K. (1999) Sisterchromatid separation at anaphase onset is promoted by cleavage of the cohesin subunit Scc1. Nature 400, 37–42

The FASEB Journal 䡠 www.fasebj.org

SCHRÖDER-HEURICH ET AL.

29.

30.

31.

32.

33.

34. 35.

McAleenan, A., Clemente-Blanco, A., Cordon-Preciado, V., Sen, N., Esteras, M., Jarmuz, A., and Aragon, L. (2013) Post-replicative repair involves separase-dependent removal of the kleisin subunit of cohesin. Nature 493, 250 –254 Goodarzi, A. A., Noon, A. T., and Jeggo, P. A. (2009) The impact of heterochromatin on DSB repair. Biochem. Soc. Trans. 37, 569 –576 Chapman, J. R., and Jackson, S. P. (2008) Phospho-dependent interactions between NBS1 and MDC1 mediate chromatin retention of the MRN complex at sites of DNA damage. EMBO Rep. 9, 795–801 Hopfner, K. P., Karcher, A., Shin, D. S., Craig, L., Arthur, L. M., Carney, J. P., and Tainer, J. A. (2000) Structural biology of Rad50 ATPase: ATP-driven conformational control in DNA double-strand break repair and the ABC-ATPase superfamily. Cell 101, 789 –800 Moreno-Herrero, F., de Jager, M., Dekker, N. H., Kanaar, R., Wyman, C., and Dekker, C. (2005) Mesoscale conformational changes in the DNA-repair complex Rad50/Mre11/Nbs1 upon binding DNA. Nature 437, 440 –443 Hirano, T. (2006) At the heart of the chromosome: SMC proteins in action. Nat. Rev. Mol. Cell Biol. 7, 311–322 Anderson, D. E., Trujillo, K. M., Sung, P., and Erickson, H. P. (2001) Structure of the Rad50 x Mre11 DNA repair complex from Saccharomyces cerevisiae by electron microscopy. J. Biol. Chem. 276, 37027–37033

RAD50 ON CHROMATIN BRIDGES

36. 37.

38. 39. 40. 41.

42.

Lichten, M. (2005) Rad50 connects by hook or by crook. Nat. Struct. Mol. Biol. 12, 392–393 Van der Linden, E., Sanchez, H., Kinoshita, E., Kanaar, R., and Wyman, C. (2009) RAD50 and NBS1 form a stable complex functional in DNA binding and tethering. Nucleic Acids Res. 37, 1580 –1588 Wei, Y., Yu, L., Bowen, J., Gorovsky, M. A., and Allis, C. D. (1999) Phosphorylation of histone H3 is required for proper chromosome condensation and segregation. Cell 97, 99 –109 Losada, A., and Hirano, T. (2001) Shaping the metaphase chromosome: coordination of cohesion and condensation. Bioessays 23, 924 –935 Harvey, S. H., Krien, M. J., and O’Connell, M. J. (2002) Structural maintenance of chromosomes (SMC) proteins, a family of conserved ATPases. Genome Biol. 3, REVIEWS3003 Schmiesing, J. A., Gregson, H. C., Zhou, S., and Yokomori, K. (2000) A human condensin complex containing hCAP-ChCAP-E and CNAP1, a homolog of Xenopus XCAP-D2, colocalizes with phosphorylated histone H3 during the early stage of mitotic chromosome condensation. Mol. Cell. Biol. 20, 6996 – 7006 Rozier, L., Guo, Y., Peterson, S., Sato, M., Baer, R., Gautier, J., and Mao, Y. (2013) The MRN-CtIP pathway is required for metaphase chromosome alignment. Mol. Cell. 49, 1097–1107 Received for publication July 3, 2013. Accepted for publication November 26, 2013.

1341