Full-Length BARD1 Mediates Aurora B Degradation, Cancer-Associated BARD1B Scaffolds Aurora B and BRCA2

Published OnlineFirst January 27, 2009; DOI: 10.1158/0008-5472.CAN-08-2134

Research Article

Distinct Roles of BARD1 Isoforms in Mitosis: Full-Length BARD1 Mediates Aurora B Degradation, Cancer-Associated BARD1B Scaffolds Aurora B and BRCA2

Stephan Ryser,1,2 Eva Dizin,1,2 Charles Edward Jefford,1,2 Be´ne´dicte Delaval,3 Sarantis Gagos,4 Agni Christodoulidou,4 Karl-Heinz Krause,2 Daniel Birnbaum,3 and Irmgard Irminger-Finger1,2

1Molecular Gynecology and Obstetrics Laboratory, Department of Gynecology and Obstetrics and 2Department of Medical Genetics and Laboratory Medicine, University Hospitals Geneva, Geneva, Switzerland; 3Laboratoire d’Oncologie Mole´culaire, Centre de Recherche en Cance´rologiede Marseille, UMR891 Inserm-Institut Paoli-Calmettes, Marseille, France; and 4Laboratory of Genetics, Foundation of Biomedical Research, Academy of Athens, Athens, Greece

Abstract NH2 terminal RING finger domains (4), and either Bard1 or Brca1 The BRCA1-associated ring domain protein 1 (BARD1) deletion induces breast cancer in tissue-specific conditional interacts with BRCA1 via its RING finger domain. The knockout mice (5). BARD1 is the major protein-binding partner BARD1-BRCA1 complex participates in DNA repair, cell cycle of BRCA1, with whom it forms a stable heterodimer and acts in control, genomic stability, and mitotic spindle formation tumor suppressor functions but has also BRCA1-independent through its E3 ubiquitin ligase activity. Cancer cells express functions in apoptosis (6, 7). The BARD1-BRCA1 heterodimer is an severalBARD1 protein isoforms, includingthe RING finger– E3 ubiquitin ligase implicated in DNA repair and homologous deficient variant BARD1B. Here, we show that BARD1 has recombination (8, 9), centrosome duplication (10), and mitotic BRCA1-dependent and BRCA1-independent functions in spindle assembly (11), which are essential functions for maintain- mitosis. BARD1, but not BRCA1, localizes to the midbody at ing genomic stability (10, 12). telophase and cytokinesis, where it colocalizes with Aurora B. During S-phase, BRCA1, BRCA2, and BARD1 are partially The 97-kDa full-length (FL) BARD1 coimmunoprecipates with colocalized to distinct nuclear dots (13, 14). During mitosis, BRCA1 BRCA1, but the 82-kDa BARD1B coimmunoprecipitates with localizes to spindle poles (15), and BRCA2 localizes to the midbody Aurora B and BRCA2. We used selective small interfering RNAs during telophase and cytokinesis, wherein it is involved in to distinguish the functions of FL BARD1 and BARD1B. contractile ring and midbody formation and completion of Depletion of FL BARD1 had only minor effects on cell growth cytokinesis (16). Depletion of BRCA2 impedes abscission of the and did not abolish midbody localization of BARD1 staining, midbody and cell separation (16), providing an explanation for the but resulted in massive up-regulation of Aurora B. In contrast, genetic instability observed in cancers associated with mutations suppression of FL BARD1 and BARD1B led to growth arrest in BRCA2. Midbody formation and abscission are also controlled by and correlated with various mitotic defects and disappearance the microtubule-binding protein transforming acidic coiled coil– of midbody localization of BARD1 staining. Our data suggest a containing protein 1 (TACC1) and the mitotic kinase Aurora B (17). novelfunction of FL BARD1 in Aurora B ubiquitination and TACC1 was reported to interact with BARD1 in Caenorhabditis degradation, opposing a proproliferative function of BARD1B elegans (18). BARD1 expression is up-regulated during mitosis (19), in scaffolding Aurora B and BRCA2. Thus, loss of FL BARD1 presumably due to phosphorylation by cyclin-dependent kinase/ and up-regulation of Aurora B, as observed in cancer cells, can cyclin complexes (20), but its subcellular localization during be explained by an imbalance of FL BARD1 and BARD1B. mitosis has not been investigated. [Cancer Res 2009;69(3):1125–34] The mitotic functions of BRCA1 and BARD1 (11) could explain why normal proliferating cells are not viable without BRCA1 or BARD1. However, mouse trophoblast giant cells, which Introduction undergo endomitosis, a nuclear division not followed by cell Breast and ovarian cancers with mutations in BRCA1 and BRCA2 division, are not affected in the Bard1 knockout mouse (3), show a severe genomic instability phenotype (1). Genomic suggesting that the lethal phenotype of BARD1 depletion is linked instability and a premalignant phenotype are also features of to a function at exit of mitosis, similar to functions attributed BRCA1-associated ring domain protein 1 (BARD1)–deficient cells to BRCA2 (16). (2), causing early embryonic death of BARD1 knockout mice (3). A genetic link between BRCA2 and BARD1 was found with the BARD1 and BRCA1 form a stable heterodimer via their respective BRCA2 mutation 999del5, which, when combined with the BARD1 variant Cys557Ser, results in 100% probability for developing breast and/or ovarian cancer for carriers of the BRCA2 999del5/BARD1 Cys557Ser double mutation (21). Thus, BRCA2 and BARD1 might Note: Supplementary data for this article are available at Cancer Research Online act in a common pathway. (http://cancerres.aacrjournals.org/). Whereas mutations in BARD1 are rare in cancers (7), aberrant S. Ryser and E. Dizin contributed equally. Requests for reprints: Irmgard Irminger-Finger, University Hospitals Geneva (HUG), up-regulation of isoforms is associated with poor prognosis in Blvd de la Cluse, Geneva, CH-1211, Geneva, Switzerland. Phone: 41-22-382-4327; Fax: breast and ovarian cancer (22, 23). We found aberrantly expressed, 41-22-382145; E-mail: [email protected]. I2009 American Association for Cancer Research. differentially spliced BARD1 isoforms, which lack the BRCA1- doi:10.1158/0008-5472.CAN-08-2134 interacting RING finger, in breast, ovarian, and endometrial cancer www.aacrjournals.org 1125 Cancer Res 2009; 69: (3). February 1, 2009

Cancer Research cells (23, 24) and human cytotrophoblasts (25). Repression of these on a logarithmic scale after determining the cell number along the isoforms in cancer cells that lack full-length (FL) BARD1 led to exponential phase and calculating the mean population-doubling time. growth arrest (23), which suggests that BRCA1-independent Immunofluorescence microscopy. Cells were fixed with 2% parafor- proproliferative, presumably mitotic, functions are retained in maldehyde for 15 min at room temperature (RT) or with methanol for 6 min at À20jC and rinsed in acetone for 30 s. Paraformaldehyde-fixed cells were cancer-associated BARD1 isoforms. permeabilized in 1% Triton/PBS for 15 min at RT and then blocked in 1% We therefore designed this study to understand the different serum/PBS for 30 min. Coverslides were incubated with appropriate roles of FL BARD1 and cancer-associated isoforms in mitosis. We antibodies for 1 h at RT in 1% FCS/PBS, washed, and stained with 4¶,6- show that FL BARD1 and the RING finger–deficient p82 isoform diamidino-2-phenylindole for 3 min. Coverslips were mounted using BARD1h have different functions and protein-protein interaction fluorogard solution and analyzed under a Nikon epifluorescence micro- properties. FL BARD1 interacts with BRCA1 and is involved in scope, and images were captured with a 3.3-megapixel CCD camera. Images Aurora B ubiquitination and degradation during mitosis. In were processed with Metamorph software (Visitron). Primary antibodies contrast, BARD1h stabilizes Aurora B and forms a complex with used were BRCA1 (Ab-1; Calbiochem), BARD1 (H-300; Santa Cruz), BARD1 BRCA2 and Aurora B, proteins known to be involved in midbody (BL518; Bethyl Laboratories), Aurora B (AIM-1; BD Biosciences). PVC was formation (16, 17). These findings establish an unexpected link reported previously (2). Antibody p25 was produced in rabbits against between BARD1 and Aurora B and BRCA2, which are aberrantly peptide sequence MVAVPGTVAPRC encoded in alternative open reading frame (ORF) of exon 1. Antibody specificity was probed in inversed ELISAs overexpressed or mutated in cancer. We suggest a molecular with different peptides and on human cancer cell lines (data not shown). pathway that explains the function of FL BARD1 as tumor Immunoprecipitation and Western blot. Protein lysates used for suppressor together with BRCA1 and loss of FL BARD1, but up- immunoprecipitation were prepared from HeLa cells, unsynchronized and regulated expression of RING finger–deficient isoforms with synchronized in G1-S or G2-M. At 24 h after plating, cells were arrested in S proproliferative functions, in cancer. phase using 2 mmol/L thymidine for 18 h, released from the arrest for 9 h,

arrested a second time using thymidine (2 mmol/L) for 18 h (G1-S), and released from the arrest for 5 h (G2-M). Materials and Methods Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer [50 mmol/L Tris (pH 8), 150 mmol/L NaCl, 2 mmol/L EDTA, 0.5% NP40, Plasmid constructs. BRCA1 and BARD1 small interfering RNA (siRNA) 10% glycerol] supplemented with protease inhibitors (complete EDTA-free, constructs were generated by annealing complementary oligonucleotides Roche Applied Science). Protein concentrations were determined by the and inserting them into the pSuperScript vector as BglII/HindIII fragment: Bradford procedure (Bio-Rad). Lysates containing 400 Ag of protein were human BRCA1-siRNA (si-56), human BARD1-siRNA 34 (si-34), and human precleared by stirring with 100 AL of protein G-Sepharose for 1 h at 4jC. BARD1-siRNA 78 (si-78) forward and reverse oligonucleotides have been After centrifugation, immunoprecipitation was performed with the described previously (23, 26); hBARD1-siRNA 423 (si-423) forward, 5¶-G ATC precleared lysate and 1 Ag of antibody for 3 h at 4jC. Thirty microliters CCC GTG CTC AGC AAG ACT CAT ATT CAA GAG ATA TGA GTC TTG of protein G-Sepharose were added and incubated for 30 min at 4jC. After CTG AGC ACT TTT TGG AAA-3¶, and reverse, 5¶-A GCT TTT CCA AAA AGT centrifugation, beads were washed twice with lysis buffer, and proteins GCT CAG CAA GAC TCA TAT CTC TTG AAT ATG AGT CTT GCT GAG were eluted by heating at 95jC for 5 min in SDS loading buffer and CAC GGG-3¶. The pSuper fragments ClaI/EcoRI containing siRNA sequences 100 mmol/L DTT. were then subcloned in the lentiviral vector pLVTHM (27). pLVTHM Proteins were subjected to SDS-PAGE and blotted onto polyvinylidene and pLV-tTRKRAB plasmids were kindly provided by D. Trono. pLVTHM- difluoride membranes (Immobilon-P, Millipore) using standard methods. SIGN (si-SIGN), kindly provided by V. Piguet, is used as a negative The following primary antibodies were used: BRCA1 (Ab-1; Calbiochem), control (28). BARD1 (H-300; Santa Cruz), BARD1 (BL518; Bethyl Laboratories), Aurora B Human FL BARD1 and splice variant BARD1h cDNAs were amplified (AIM-1; BD Biosciences), BRCA2 (Ab-1; Calbiochem), TACC1 (Upstate from human primary fibroblasts and HeLa cells, respectively. Flag epitope Biotechnology Euromedex), g-tubulin (C11; Santa Cruz), h-tubulin M2 and c-myc tags were inserted at the 5¶ and 3¶ ends of the cDNA, and the (D10; Santa Cruz), cdc-2 (POH-1; Cell Signaling), P-cdc2-Tyr 15 (Cell resulting construct recombined into the 2K7bsd lentiviral vector (29). Signaling), actin (C-2; Santa Cruz), cyclin A (BF-683; Santa Cruz), anti-HA Si-78–resistant FL BARD1 (FLm78) construct was generated by site-directed (HA.11; Covance). Antibody used against Aurora A was obtained from C. mutagenesis using the FL BARD1 construct as template and the following Prigent (31). Horseradish peroxidase–conjugated secondary antibodies primers Mut78-F 5¶-AAG TGT ATG CTC GGA ATA CTC AAT GGA TG-3¶ and (Amersham Biosciences) were used for detection of immunoreactive Mut78-R 5¶-AGC ATC CAT TGA GTA TTC CGA GCA TAC AC-3¶. proteins by enhanced chemiluminescence (Amersham Biosciences). The expression plasmid HA-tagged ubiquitin was kindly provided by Protein stability assay. To determine the effect of FL BARD1 and G. Courtois and M. Huber. spliced variant BARD1-h on the stability of Aurora B, cells growing in six- Cell culture and transduction with lentivirus. The 293T and HeLa cell well plates were transfected with 1 Ag of 2K7 vector (control) or 2K7 vector lines were cultured in DMEM supplemented with 10% fetal calf medium. All expressing human FL BARD1 or spliced variant BARD1h and 3 ALof recombinant lentiviruses were produced and purified according to standard Fugene-6 (Roche Applied Biosciences). Forty-four hours after transfection, protocols (30). 293T cells were cotransfected with the lentivirus vector, the 100 Ag/mL of cycloheximide (Sigma) was added to cells to inhibit protein packaging vector psPAX2, and the envelope vector pMD2G by calcium synthesis, and cells were harvested in RIPA buffer at various time intervals, phosphate precipitation. After 16 h, medium was changed, and recombinant as indicated in the figure. lentiviruses were harvested 24 h later. For lentivirus transduction, HeLa In vivo ubiquitination assay. At 30% confluence, cells growing in cells were plated on six-well plates (104 cells per well) for 24 h and then 10-cm-diameter dish were cotransfected with 1 Ag of vector encoding incubated with the medium containing recombinant lentivirus vectors for HA-tagged ubiquitin and 2 Ag of 2K7 vector (control) or 2K7 vector 48 h. To control the expression of the siRNA, the cells were cotransduced expressing human FL BARD1 or spliced variant BARD1h and 9 ALof with lentivirus derived from the vectors pLVTHM and pLV-tTRKRAB (27). Fugene-6. Thirty-six hours after transfection, cells were treated with Coexpressed tTR-KRAB locks onto the tet-on promoter. Upon doxycyclin 10 Amol/L of the proteasome inhibitor MG132 (Calbiochem) for 12 h. After exposure, KRAB is derepressed and allows transcription of siRNAs. cell lysis, immunoprecipitations were performed with 800 Ag of total cellular Generation of growth curves. HeLa cells were trypsinized, centrifuged, proteins and 2 Ag of anti-HA antibody. The immunocomplexes were pulled and resuspended in 5 mL of medium. The cells were counted using a down, as described above, and analyzed by SDS-PAGE on 10% gels and hemocytometer, then diluted to have a 20 to 30% confluency in the Petri probed for the presence of polyubiquitinated-Aurora B using anti-Aurora B dish. Triplicate plates were counted every 48 to 72. The results were plotted antibody.

Cancer Res 2009; 69: (3). February 1, 2009 1126 www.aacrjournals.org

Distinct Roles of BARD1 Isoforms in Mitosis

Figure 1. Cell cycle–regulated localization of BARD1. A, BARD1 translocation during mitosis from centrosomes to the midbody. Double staining with anti-BARD1 H300 and anti–h-tubulin antibodies are shown in HeLa cells at different cell cycle stages. BARD1 localization is observed mostly in the nucleus during S phase. At metaphase, BARD1 is mostly localized to the spindle pole and colocalizes with tubulin (yellow, short arrow). In anaphase, colocalization of BARD1 and tubulin (yellow, short arrow) and abundant BARD1 staining at the spindle midzone (long arrow) can be observed. In telophase, abundant BARD1 staining and colocalization with tubulin (short arrow)is observed at the midbody (long arrow). Colocalization of BARD1 and DNA can be observed (pink) in metaphase and most prominently in anaphase. B, distinct localization of BARD1 and BRCA1 during telophase and cytokinesis. Double labeling for BRCA1 and BARD1 (using antibody BL518) was performed in HeLa cells. BRCA1 is localized to nuclear dots during telophase and especially during cytokinesis. In contrast, BARD1 shows highest concentration at the presumed contractile ring during telophase and is localized to the midbody (arrow) during cytokinesis. C, costaining for BARD1 (using BARD1 H300) and Aurora B shows transition of BARD1 staining from the spindle poles in metaphase to the metaphase plate during anaphase and colocalization with Aurora B at the midplate (short arrow) from anaphase to cytokinesis. Aurora B staining is concentrated at the metaphase plate, colocalizing with DNA (merge; light blue, long arrow), during anaphase and regresses to distinct dots at the midbody during cytokinesis (long arrow). During cytokinesis, BARD1 and Aurora B costaining gradually dissociates (long arrow; blow-up). Aurora B is more proximal to the abscission point than BARD1.

Results BARD1 and BRCA1 showed colocalization during metaphase and BARD1, but not BRCA1, colocalizes with Aurora B at the anaphase but localized separately during telophase and cytokinesis midbody during cytokinesis. To investigate the function of (Fig. 1B). In early telophase, BARD1 staining was observed around BARD1 during mitosis, we studied its intracellular localization at the chromatin but increased at the site of cleavage furrow different stages of mitosis by immunofluorescence microscopy. ingression. BRCA1 staining was around the chromatin but less Double labeling with anti-BARD1 and anti–h-tubulin antibodies staining was observed in regions of the presumed contractile ring. revealed that BARD1 staining was partially localized along mitotic During telophase and cytokinesis, BARD1 staining was localized at spindle microtubules in metaphase and early anaphase cells the midbody but no BRCA1 staining was observed at the midbody (Fig. 1A). This localization was also observed in several other cell (Fig. 1B). lines, namely human HeLa, PC-3, and mouse TAC-2 cells (data not The observation of BARD1 staining at spindle poles during shown), with BARD1 H300 antibody, directed against the NH2 metaphase and at the midbody during cytokinesis suggested that terminal half of the BARD1 protein. During telophase and BARD1 interacts with BRCA1, as reported (10, 11), at early stages of cytokinesis, colocalization of BARD1 and microtubules was also mitosis, but with proteins other than BRCA1 during telophase and observed, specifically at the midbody, whereas staining during cytokinesis. anaphase and early telophase was not restricted to spindle The specific subcellular localization of BARD1 during mitosis microtubules. was reminiscent of the localization of regulators of cell division, To investigate whether BARD1 localized to the midbody known to translocate from the midplate to the midbody as part of independently of BRCA1, we performed costaining experiments the chromosomal passenger complex (CPC; ref. 32). One such with anti-BARD1 and anti-BRCA1 antibodies in mitotic cells. protein, the mitotic kinase Aurora B, is of particular interest www.aacrjournals.org 1127 Cancer Res 2009; 69: (3). February 1, 2009

Cancer Research because it is a member of CPC (33–35), is required for midbody Importantly, BRCA2 coimmunoprecipitated with BARD1 but formation (17), and is up-regulated in breast cancer (36). Double- exclusively in mitotic cell extracts. These data suggest that BARD1, labeling for BARD1 and Aurora B was performed to determine a TACC1, Aurora B, and BRCA2 act in a common pathway at the exit possible colocalization. Aurora B was only weakly expressed during of mitosis. S phase and localized to distinct spots in the nucleus in HeLa cells. To confirm BARD1 interaction with TACC1, Aurora B, and During metaphase, Aurora B staining was increased and located at BRCA2, we performed coimmunoprecipitations with anti-TACC1, the midplate. In anaphase, Aurora B staining formed a tight line at anti-BRCA2, and anti-Aurora B antibodies (Fig. 2B). As a positive the midplate and colocalized with BARD1 (Fig. 1C). During immunoprecipitation control, we used anti-BRCA1 antibodies. telophase and cytokinesis, when Aurora B is gradually degraded BRCA1 immunoprecipitation, when probed on Western blots with to form two distinct spots at the midbody adjacent to the the anti-BARD1 antibody BL518 directed against an epitope microtubule abscission point, BARD1 staining was partially overlapping with Aurora B staining. Further inspection showed that BARD1 and Aurora B colocalize already in metaphase and gradually dissociate during cytokinesis (data not shown). To investigate whether colocalization with Aurora B is also a property of BRCA1, we performed double labeling experiments with anti-BRCA1 and anti-Aurora B antibodies. In anaphase, BRCA1 staining was observed at the midplate, but only weakly colocalized with Aurora B compared with BARD1 colocalization with Aurora B. In telophase, no BRCA1 staining at the midbody and no colocalization with Aurora B were observed (Supplementary Fig. S1). These data show BARD1 association with BRCA1 at the spindle poles during metaphase but costaining for BARD1 and Aurora B during anaphase, telophase, and cytokinesis. This suggests that BARD1, but not BRCA1, colocalizes with Aurora B at the midbody. BARD1 interacts with Aurora B, TACC1, and BRCA2 during mitosis. Because BARD1 colocalizes with Aurora B during mitosis, we investigated whether BARD1 is part of an Aurora B protein complex during mitosis. Aurora B forms a complex with TACC1, which, like Aurora B, localizes at the metaphase plate during anaphase and the midbody during cytokinesis (17). TACC1 was also identified as binding partner of BARD1 in C. elegans cells (18). Like Aurora B, TACC1 is up-regulated and acts as an oncogene in breast and ovarian cancer (37). To show that BARD1, TACC1, and Aurora B physically interact during mitosis, we performed coimmunoprecipitation experiments using anti-BARD1 H300 antibodies, which recognize the NH2 terminal 300 amino acids of the BARD1 protein. Immunoprecip- itation with anti-BARD1 antibodies confirmed TACC1 interaction with BARD1 (Fig. 2A). We then probed anti-BARD1 immunoprecipitations with antibodies against Aurora B and Aurora A, the mitotic kinase that interacts with and phosphorylates BRCA1 (33–35). Both Aurora A and Aurora B levels were increased in G2-M. However, coimmunoprecipitation with BARD1 was only observed for Aurora B and Figure 2. BARD1 interaction with BRCA1, TACC1, BRCA2, and Aurora B. Immunoprecipitations were performed with HeLa cells, either nonsynchronized only in G -M cell extracts (Fig. 2A). Aurora A levels were very low, 2 (NS) or enriched in G1-S or G2-M. A, Western blots of total cell extracts and a binding to BARD1 cannot be excluded. Together, these data (total), BARD1 immunoprecipitations using BARD1 H300 (IP BARD1), and control immunoprecipitation (omission of BARD1 antibody), were separated indicate that BARD1 forms a complex with TACC1 and Aurora B on 6% gels and probed with anti-Aurora A, Aurora B, TACC1, BRCA1, and during mitosis. BRCA2 antibodies. An additional band (*), presumably phosphorylated TACC1, TACC1 and Aurora B are proteins required for midbody formation is detected and increased in G2-M of total lysates. B, coimmunoprecipitation assays were performed with anti-BRCA1, TACC1, BRCA2, and Aurora B (17), a function also involving BRCA2 (16). Because BRCA2 also antibodies. Western blots were probed with anti-BARD1 BL518 (recognizing localizes to the midbody (16) and a genetic interaction of BARD1 and epitope in exon 4). Total cell extracts show 97-kDa FL BARD1 and phosphorylated (*) and 82-kDa BARD1h. FL BARD1 is increased in G2-M; BRCA2 suggests that they act in a common pathway (21), we tested BARD1h is not. FL BARD1 is coprecipitated with BRCA1 and TACC1 whether BARD1 interacts with BRCA2. We probed BARD1 (IP TACC1). No coprecipitation of BARD1 is found in the control assays with coimmunoprecipitations with BRCA1, as positive control, and beads only. IP BRCA2 coprecipitates BARD1h. IP Aurora B also coprecipitates p82/BARD1h, but only in G2-M cell extracts. C, expression of cell cycle BRCA2 antibodies. BARD1 and BRCA1 coimmunoprecipitated in regulators. Cell extracts were separated on 8% gels and were probed with G1-S and G2-M cell extracts, and binding was not increased in mitotic anti–phosphorylated CDC2 (p-cdc2), anti-CDC2, anti–cyclin A, and anti-BARD1. Detection of phosphorylated CDC2 in G1-S and increase of cyclin A in G2-M extracts, although total BRCA1 and BARD1 levels increased in G2-M showed synchronization of cells. FL BARD1, but not BARD1h, is up-regulated cells (Fig. 2A). and phosphorylated (*) during mitosis.

Cancer Res 2009; 69: (3). February 1, 2009 1128 www.aacrjournals.org

Distinct Roles of BARD1 Isoforms in Mitosis

Figure 3. Specific repression and localization of FL BARD1 and BARD1h. A, schematic diagram of BARD1 exon structure with approximate localization of conserved protein motifs [RING finger domain (RING), Ankyrin repeats, BRCT domains], and exon structure of BARD1 isoforms are presented. BL518 marks approximate position of reactive epitope of anti-BARD1 antibody BL518. Approximate positions of siRNA target sequences (si-34, si-423, si-78) are indicated on the BARD1 structure. Target exons are colored for si-34 (purple) and si-78 (red). For si-423, yellow arrow indicates position in exon 4 on isoform BARD1h. Capacity of BRCA1 interaction is indicated. Isoform BARD1h uses an alternative ORF in exon 1 and translates into a protein of f82 kDa (23, 25). B, time course of FL BARD1 and p82/BARD1h repression after induction of siRNA expression by doxycyclin (Dox) in HeLa cells. Effect of different siRNAs, targeting BARD1 (si-34, si-423, and si-78), BRCA1 (si-56), or unrelated gene SIGN (si-SIGN), was monitored on Western blots probed with antibodies recognizing BARD1 (BL518), BRCA1, or actin, as loading control. BARD1 antibody BL518 recognizes two bands of 97 and 82 kDa (also recognized by BARD1 H300; Supplementary Fig. S2), FL BARD1 and BARD1h. si-SIGN does not repress either band; si-34 and si56 repress only FL BARD1, si-423 FL BARD1, and BARD1h, but the latter is more resistant; si-78 efficiently represses both. BRCA1 expression is reduced with all siRNAs except si-SIGN. C, intracellular localization of FL BARD1 during cytokinesis. HeLa cells were cotransduced with inducible BARD1 siRNA si-78 (doxycyclin-dependent expression) and si-78–resistant form of FL BARD1, FLm78, lentiviruses. Immunofluorescence with anti-BARD1 BL518 shows nuclear and cytoplasmic staining in nontransduced HeLa cells (control), but increased nuclear staining is observed during S phase in cells expressing exogenous BARD1-FLm78. In cytokinesis, BARD1 staining is around chromatin and more particularly at the midbody. In cells depleted of FL BARD1 and BARD1h by si-78 repression and complemented with FLm78 (right), BARD1 staining is absent from the midbody during cytokinesis. Arrows point at midbodies. Expression of siRNA is monitored with GFP expression. Blow-up of GFP expression is shown to visualize midbody. D, complementation of si-78 repression with wild-type FL BARD1 (FLwt) or degradation-resistant mutated form (FLm78) in HeLa cells. Expression of si-78 was induced with doxycyclin (Dox,+/À). Western blot of cell extracts was probed with anti-BARD1 antibody BL518. FL BARD1 and p82/BARD1h are expressed in noninduced cells; FL BARD1, but not BARD1h, is only observed in induced cells expressing exogenous BARD1-FLm78.

encoded on exon 4, showed 97-kDa FL BARD1 interaction with as BARD1h, but not of p97 FL BARD1. BARD1h lacks the region BRCA1. Negative immunoprecipitation control showed no precip- comprising the BRCA1-interacting RING domain (23, 25), hence itation. Anti-TACC1 antibodies also coprecipitated FL BARD1 and cannot interact with BRCA1 in anti-BRCA1 immunoprecipitation. coprecipitated slightly better in extracts of mitotic cells. However, Thus, BARD1h binds to BRCA2 and Aurora B, and the interaction is immunoprecipitations with anti-BRCA2 and anti-Aurora B showed strongest with Aurora B and restricted to mitotic cells. These coprecipitation of a 82-kDa protein, which we previously identified results also show that FL BARD1 does not bind to BRCA2 and www.aacrjournals.org 1129 Cancer Res 2009; 69: (3). February 1, 2009

Cancer Research

Aurora B, suggesting that the NH2 terminal RING finger does not BARD1, and its repression was observed later than repression of FL promote but rather hinders BARD1 interaction with BRCA2 and BARD1. BARD1h protein levels exceeded those of FL BARD1, which Aurora B. was also observed in different other cell lines (ref. 23; Supplemen- To investigate cell cycle stage-specific expression, cell extracts tary Fig. S2). used for immunoprecipitations were prepared from cells, either BARD1B, but not FL BARD1, localizes to the midbody. nonsynchronized or synchronized, and enriched for G1-S or G2-M Because BARD1, but not BRCA1, staining was observed at the cell stages. In parallel to immunoprecipitations, cell extracts were midbody during telophase and cytokinesis, we wondered whether monitored for expression of mitotic marker proteins, such as the observed staining was due to the presence of FL BARD1 or phosphorylated CDC2 and cyclin A (Fig. 2C). BARD1h isoform. To distinguish localization and possible functions Distinct repression of FL BARD1 or FL BARD1B. To of FL BARD1 and BARD1h, we performed immunofluorescence distinguish localization and function of FL BARD1 or the RING staining in cells that are deficient of isoforms but express FL finger–deficient isoform BARD1h, we performed repression experi- BARD1, using the anti-BARD1 BL518 antibody, which recognizes ments with siRNAs targeting different regions or exons of BARD1 both FL BARD1 and BARD1h, for detection (Fig. 3C). in HeLa cells (Fig. 3A). To determine the phenotype of BARD1 To generate these cells, BARD1-FLm78 (FL BARD1 expression depletion, we generated stable inducible cell lines by cotransducing clone carrying a silent mutation at the si-78 target region) was siRNAs in lentiviral vectors with the tTRKRAB repressor operated constitutively expressed in si-78 expressing cells. Western blots through the Tet-on system (27, 30). probed with anti-BARD1 BL518 confirmed the identity of FL The efficiency of BARD1 and BRCA1 siRNAs was analyzed BARD1 and showed that levels of exogenous BARD1-FLm78 were on Western blots probed with anti-BARD1 BL518 (Fig. 3B)or not affected by si-78–induced repression (Fig. 3D). H300 antibodies (Supplementary Fig. S2). Expression of si-34, Using anti-BARD1 BL518 in S-phase cells, BARD1 staining was directed against exon 2 of BARD1, led to repression of 97-kDa FL found in the nucleus and cytoplasm. In cells that expressed BARD1, but not BARD1h (Fig. 3B). FL BARD1 levels were also exogenous BARD1-FLm78, nuclear staining was increased, indicat- decreased after repression of BRCA1 (si-56) due to mutual ing that FL BARD1 localized to the nucleus, whereas cytoplasmic stabilization of the two proteins. BARD1h was repressed by staining might reflect BARD1h localization (Fig. 3C). This si-423, directed against exon 4 and si-78, directed against exon 9 localization was reported previously for NH2 terminal truncated but not by si-34, consistent with the structure of this isoform forms of BARD1 (19). In anaphase, BARD1 staining was observed at (Fig. 3A). The BARD1h isoform seemed more stable than FL the spindle poles and the midplate, and in telophase, an intense

Figure 4. BARD1h isoform is essential for cell growth. A, images of inverse-phase microscopy and GFP expression of HeLa cells transduced with Tet-ON inducible siRNAs (si-34, si-423, si-78) transcribed from the same plasmid as the Krab repressor and GFP are shown before siRNA induction (si-78) and 5 d after siRNA induction (+Dox) as bright field and GFP fluorescence for siRNA expression control. B, growth curves of cells expressing Tet-ON inducible BARD1 si-34, si-423, and si-78; BRCA1 si-56; and control si-SIGN. si-78 and si-423 show no or reduced growth; si-34 and si-56 have similar reduced growth rate compared with control cells si-SIGN. C, growth curves of vector control cells and BARD1 si-78 with or without doxycyclin (Fdox) induction are shown: HeLa, empty vector (pLVHTM), vector expressing GFP and tTRKRAB repressor (K-GFP; Fdox), and BARD1 si-78 cells (si-78; Fdox). Induction of BARD1 si-78 expression leads to complete growth arrest.

Cancer Res 2009; 69: (3). February 1, 2009 1130 www.aacrjournals.org

Distinct Roles of BARD1 Isoforms in Mitosis

Figure 5. FL BARD1 and BARD1h modulate Aurora B degradation. A, correlation of BARD1 and BARD1h expression with Aurora B protein level. Time courses of siRNA induction are shown for si-34, si-423, and si-78. Cells were induced with doxycyclin for 10d, and cell extracts were probed on Western blots with anti-BARD1 (BL518), anti-Aurora B, and anti-tubulin for loading control. Histogram shows quantification of Aurora B protein levels using image analysis software Image J. Gels and quantifications are representative experiments repeated at least twice. B, monitoring Aurora B degradation after cycloheximide (CHX) treatment. Protein extracts from HeLa cells, transiently transfected with empty vector (Control), FL BARD1, or BARD1h expression constructs, were analyzed on Western blots probed with anti-Aurora B antibodies. Protein extracts were prepared from cells incubated with CHX for 0, 5, or 8h.Bottom, quantification of Aurora B repression in different cell types. C, FL BARD1, but not BARD1h, overexpression induces Aurora B ubiquitination in vivo. Protein extracts from HeLa cells, transiently cotransfected with empty vector (Control), FL BARD1, or BARD1h expression constructs, and HA-tagged ubiquitin expression clone and treated with proteasome inhibitor MG132, were immunoprecipitated with anti-HA antibodies. Western blots were probed with anti-Aurora B antibodies. Input cell lysates were tested for BARD1 and Aurora B expression (bottom).

staining was observed at the midbody. BARD1-FLm78 expression in BRCA1 (si-56). However, expression of siRNA si-78, targeting both cells depleted of FL BARD1 and BARD1h by si-78 expression did FL BARD1 and BARD1h, led to growth arrest. When compared not change the pattern of BARD1 staining during S phase, with various control cells, growth rates of si-78 expressing cells metaphase, or anaphase (data not shown). However, staining to were zero after 2 to 3 days of doxycyclin addition (Fig. 4C). si-423 the midbody in telophase was missing. These data indicate that expression, targeting exon 4 common to FL BARD1 and BARD1h, the BARD1h isoform, but not FL BARD1, is localized to the resulted in growth arrest but only after 5 to 6 days (Fig. 4B). This midbody during telophase and cytokinesis. difference in growth arrest induced by si-78 or si-423 expression We further confirmed that BARD1h, and not FL BARD1, localizes was correlated with the depletion of BARD1h, but not FL BARD1, to the midbody by using antibodies that selectively recognize the proteins (Fig. 3B). These data indicate that BARD1h, but not FL two isoforms. Antibody BARD1 H300 (directed against the 300 NH2 BARD1 or BRCA1, retain the minimal protein functions required terminal amino acids) shows staining at midbody in cytokinesis, for cell proliferation. but antibody BARD1 PVC (directed against exon 3, which is deleted The depletion of BARD1 induced various mitotic defects and in BARD1h) does not show staining at the midbody (Supplemen- genetic instability. We monitored the effects of BARD1 depletion tary Fig. S3A and B). by si-34 or si-78 and of BRCA1 depletion by si-56 in HeLa cells by We also generated an antibody, p25, directed against an epitope time-lapse video imaging, immunofluorescence microscopy, and encoded in the alternative ORF of exon 1, hence unique to BARD1h cytogenetic analysis (Supplementary Video S1–S4; Supplementary (Supplementary Fig. S3A). This p25 antibody specifically recognizes Table S1; Supplementary Figs. S4–S7). BARD1-depleted cells had BARD1h by immunoprecipitation and Western blot (Supplemen- difficulties in forming bipolar mitotic spindles and in abscission tary Fig. S3C). Importantly, p25 shows staining at the midbody in of the midbody. Although cell cycle times were only slightly cytokinesis (Supplementary Fig. S3B). elevated for cells expressing si-34 or si-56, cell cycle times of si-78 Thus, immunofluorescence staining of cells depleted of expressing cells could not be determined, because only few cells FL BARD1 and BARD1h and complementation with mutated FL completed mitosis. Presumably, due to these defects, high levels BARD1, use of FL BARD1-specific antibody PVC, and use of of chromosomal instability were observed, consistent with BARD1h-specific p25 show that BARD1h, but not FL BARD1, is previous observations (2, 3). localized to the midbody. BARD1 controls protein expression levels of Aurora B. The BARD1B is required for cell growth. To analyze the phenotype phenotype of cytokinesis failure of BARD1-depleted cells is similar of BARD1 depletion, we first monitored cell growth over a period of to the phenotype resulting from BRCA2 depletion (16) and, 10 days. Repression of BARD1 with different siRNAs led to distinct together with the observed interaction of BRCA2 with BARD1h results (Fig. 4A and B). siRNA si-34, which affects only FL BARD1 (Fig. 2), supports the hypothesis that BARD1h, Aurora B, and (Fig. 3A), had little effect on cell growth, similar to repression of BRCA2 act in a common pathway in cytokinesis. www.aacrjournals.org 1131 Cancer Res 2009; 69: (3). February 1, 2009

Cancer Research

To determine whether Aurora B expression depends on BARD1, on microtubule bridges connecting sister cells were larger than in we analyzed Aurora B and BARD1 expression in HeLa cells that control cells, and their localization was not restricted to the were either untreated, depleted of BARD1 by si-RNA, or overex- abscission point. Aurora B was accumulated in the cytoplasm and pressed FL BARD1. not restricted to the midbody. We monitored protein levels of Aurora B after BARD1 repression Our results indicate that (a) FL BARD1 overexpression results in or overexpression of FL BARD1 on Western blots. FL BARD1, polyubiquitination of Aurora B and degradation and (b) FL BARD1 BARD1h, and Aurora B expression was analyzed at various time depletion leads to increased Aurora B accumulation, suggesting points after siRNA induction on Western blots (Fig. 5A). Anti– that FL BARD1 plays an important role in Aurora B degradation h-tubulin staining was performed as loading control. Interestingly, and turnover. However, BARD1h counteracts Aurora B degrada- Aurora B levels increased with time of induction of si-34 and tion. Whereas BARD1h interaction with Aurora B is strong, FL paralleled the depletion of FL BARD1 but not of BARD1h. In si-423 BARD1 interaction with Aurora B is not observed and might be cells, wherein a rapid depletion of FL BARD1 but slow decrease of transient. One possible explanation for these observations might be BARD1h was observed, Aurora B levels first increased and then that BARD1h protects Aurora B from degradation by forming a declined in parallel with the repression of BARD1h. Consistently, in BARD1h-Aurora B-BRCA2 complex at the midbody. si-78 cells, wherein FL BARD1 and BARD1h were rapidly degraded, A role for BARD1B in BRCA2-Aurora B interaction and no significant change of Aurora B expression is observed. The midbody formation. FL BARD1 and BARD1h belong to different overexpression of FL BARD1, on the contrary, was associated with protein complexes and have distinct intracellular localization reduced Aurora B levels (Fig. 5A). during mitosis. Both might be required for the regulated gradual Seemingly, FL BARD1 and BARD1h have opposing effects on degradation of Aurora B during mitosis, which is disturbed in many Aurora B stability. To determine how they affected Aurora B cancer cells. stability, cells were transfected with vectors expressing either FL Because a common pathway for TACC1 and Aurora B in BARD1 or BARD1h or empty vector as control and then treated cytokinesis has been described (17) and a function of BRCA2 in with the protein translation inhibitor cycloheximide. We monitored cytokinesis was also reported (16), we asked whether BRCA2 Aurora B degradation during a time course of 8 hours. In cells interacts with Aurora B. Immunoprecipitation with anti-BRCA2 overexpressing FL BARD1, Aurora B was degraded more efficiently antibodies showed coprecipitation of Aurora B (Fig. 6A). Therefore, than in control cells, whereas in cells overexpressing BARD1h, Aurora B and BRCA2 form a complex in vivo. Immunoprecipitation Aurora B remained nearly stable (Fig. 5B). of BRCA2 and TACC1 did not show interaction (data not shown). We then asked whether the observed Aurora B degradation When the same experiment was carried out with cells depleted was associated with Aurora B ubiquitination and whether of FL BARD1 after siRNA induction, Aurora B expression levels ubiquitination of Aurora B was related to FL BARD1. Cells were were increased and the proportion of Aurora B interacting with cotranfected with expression plasmids encoding either FL BARD1 BRCA2 was also increased. These experiments suggest that or BARD1h, and HA-tagged ubiquitin expression vector, and BARD1h promotes the formation of a BRCA2-Aurora B complex. treated with the proteasome inhibitor MG132 to allow accumula- These data suggest that BARD1h acts with BRCA2 and Aurora B in tion of ubiquitinated proteins. Immunoprecipitation of HA-tagged midbody formation and abscission (Fig. 6B). ubiquitinated Aurora B was then visualized by Western blotting with anti-Aurora B antibodies. The ubiquitinated form of Aurora B was only observed in cells overexpressing FL BARD1 and not in Discussion cells overexpressing BARD1h or control cells (Fig. 5C). The size of BARD1 has multiple functions in association with BRCA1 (7), ubiquitinated Aurora B was consistent with previously found including mitotic spindle formation (11). We show novel BARD1 polyubiquitinated forms of Aurora B (38). These experiments functions during the late phases of mitosis, which involve show that FL BARD1, but not BARD1h, is involved in Aurora B interactions with TACC1, Aurora B, and BRCA2. Our data support degradation and suggest that FL BARD1 acts on Aurora B the view that, during mitosis, FL BARD1 and the BARD1h isoform, degradation via ubiquitination of Aurora B. BARD1h, on the which is expressed in many cancer cells (23), sequentially interact contrary, is not associated with Aurora B ubiquitination and its with BRCA1, TACC1, Aurora B, and BRCA2. BARD1-BRCA1 overexpression inhibits Aurora B degradation. heterodimers are formed during metaphase and early anaphase To investigate Aurora B expression in cells as a function of at the centrosome, consistent with previous reports of BRCA1 BARD1 expression levels, we also performed immunofluorescence localization to the spindle poles (11, 15) and its centrosome-related staining in wild-type HeLa cells and HeLa cells with exogenous functions (39), whereas BARD1h interacts with Aurora B and overexpression or si-34 repression of FL BARD1. In untreated HeLa BRCA2 during anaphase and cytokinesis. cells, Aurora B staining was confined to the nucleus, colocalized BRCA1-independent and BRCA2-related functions of BARD1 with chromatin during S phase, prophase, and metaphase, and during the late stages of mitosis. During the late stages of aligned at the midplate during anaphase. Small dots of Aurora B mitosis, in particular cytokinesis, BARD1h acts independently of staining were observed at the midbody during telophase and BRCA1. The phenotypes induced by selective repression of FL cytokinesis (Supplementary Fig. S8). In cells overexpressing FL BARD1 (si-34) or BRCA1 (si-56) were less pronounced than the BARD1, Aurora B staining was decreased at all stages compared phenotypes induced by si-78 and si-423, two siRNAs that target FL with control cells and, most importantly, during anaphase. In BARD1 and the BARD1h isoform, which lacks the BRCA1 contrast, Aurora B staining in cells depleted of FL BARD1 by si-34 interaction domain. BARD1h is also more abundant than FL expression was increased. This was most pronounced in cells BARD1 in many cancer cell lines (ref. 23; Supplementary Fig. S2C). arrested in cytokinesis. Typically, si-34 repressed cells were During anaphase, telophase, and cytokinesis, BARD1, but not connected with elongated midbodies, as was also observed by BRCA1, staining is observed at the midbody (Fig. 1). Staining at the anti–h-tubulin staining (Supplementary Fig. S6). The Aurora B dots midbody was not reconstituted by expression of exogenous FL

Cancer Res 2009; 69: (3). February 1, 2009 1132 www.aacrjournals.org

Distinct Roles of BARD1 Isoforms in Mitosis

Figure 6. Aurora B, BRCA2, and BARD1h complex formation. A, immunoprecipitations of BRCA2 in HeLa cells (control) or FL BARD1-depleted (si-34) cells were tested for coprecipitation of Aurora B on Western blots. IgGs were probed as control for antibody cross-reactivity. Input shows Western blot of FL BARD1 and BARD1h expression in control HeLa and si-34 expressing HeLa cells. B, schematic diagram of BARD1 and BARD1h interactions during mitosis. FL BARD1 and BRCA1 form a heterodimer at metaphase. FL BARD1 affects Aurora B protein levels, presumably its degradation. During anaphase, BARD1 colocalizes with Aurora B and presumably with diffusely localized BRCA2. TACC1 also interacts and colocalizes with Aurora B during anaphase and cytokinesis and translocates together with Aurora B to the midbody (17). During telophase and cytokinesis, residual Aurora B dots are localized directly at the abscission point, overlapping with BARD1 and more distantly with BRCA2 (16). Interaction studies suggest that BARD1 interacts with TACC1, Aurora B, and BRCA2, but TACC1 interacts only with Aurora B and not with BRCA2. BARD1h acts as scaffold for Aurora B and BRCA2 and enables their respective roles in midbody formation, e.g., myosin II organization (16).

BARD1 in cells depleted of all BARD1 isoforms (Fig. 3C), was not midbody during telophase and cytokinesis (Fig. 1C). Aurora B observed with an antibody specific for FL BARD1, but was observed degradation seems to be under tight spatial and temporal control. with an antibody directed against the epitope encoded by an Aurora B degradation, involving ubiquitination by APC, has been alternative ORF translated in BARD1h. These data suggest that reported (42). Although the mechanism of inhibition of Aurora B BARD1h, and not FL BARD1, is implicated in functions at degradation in FL BARD1 depleted cells is not clear at the moment, late stages of mitosis. Midbody localization was also reported it is clear that the expression levels of FL BARD1 and BARD1h are for BRCA2 (16). There is strong evidence that both proteins act in critical (Fig. 5). a common pathway: (a) the phenotype of BARD1 depletion is Our findings suggest further that BARD1h acts together with similar to BRCA2 depletion (Supplementary Fig. S4), (b) BRCA2 Aurora B in midbody formation and abscission and provide an coprecipitates with BARD1h (Fig. 2), and (c) strong genetic explanation for the proproliferative action of BARD1 isoforms in interaction of BRCA2 and BARD1 is observed in double mutation cancer cells (23). carriers (21). BARD1B links Aurora B and BRCA2 functions in cytokinesis. Opposing roles of BARD1 and BARD1B in Aurora B The BARD1 deficiency phenotype is similar to the phenotype degradation. The CPC protein Aurora B acts in a common induced by BRCA2 depletion and the Capan-1 mutation (16), pathway with TACC1 (17) in centrosome formation and cytokinesis namely cells with elongated midbodies (Supplementary Fig. S6). We and cancer progression (37, 40). TACC1 also interacts with show that BRCA2 interacts with Aurora B and BARD1h promotes BARD1, as was shown in C. elegans (18). We show here that this interaction, suggesting that the BARD1h-Aurora B-BRCA2 human BARD1 interacts with TACC1 and Aurora B. Furthermore, complex is required for completion of cytokinesis and deregulation BARD1 expression levels affect Aurora B stability and function: of any of the components induces aneuploidy. depletion of FL BARD1 results in Aurora B accumulation (Fig. 5) BRCA2 depletion affects myosin II organization at the cleavage and hyperstabilization of the midbody, reminiscent to Aurora B furrow (16). Indeed, BARD1 protein concentrations (BARD1 up-regulation in cancer (41). BARD1 depletion also results overexpression and repression) influence ingression of the in chromosome missegregation and aneuploidy (Supplementary cleavage furrow and abscission,5 presumably by controlling Figs. S5–S7), as is observed in cancers with Aurora B up-regulation. Aurora B stability and promoting Aurora B-BRCA2 interaction. Thus, mitotic and cytogenetic defects, induced by loss of FL BARD1 Because the role of Aurora B in cleavage furrow formation is well repression, could result from Aurora B up-regulation. established (43, 44), our data link BRCA2 and Aurora B functions, Interestingly, although FL BARD1 repression affects Aurora B and we hypothesize that Aurora B and BRCA2 act on myosin II ubiquitination and accumulation, FL BARD1 and Aurora B binding organization and BARD1h might be involved in that process was not observed in coimmunoprecipitations, which might be due (Fig. 6B). to a transient interaction between Aurora B and the BARD1-BRCA1 complex. In contrast to FL BARD1, BARD1h showed strong interaction with Aurora B in coimmunoprecipitations, which suggests that it is BARD1h that colocalized with Aurora B at the 5 S. Ryser and I. Irminger-Finger, unpublished. www.aacrjournals.org 1133 Cancer Res 2009; 69: (3). February 1, 2009

Cancer Research

Our observations imply that (a) FL BARD1 versus BARD1h Disclosure of Potential Conflicts of Interest expression levels are important for degradation of Aurora B during No potential conflicts of interest were disclosed. mitosis, because FL BARD1 promotes and BARD1h hampers degradation (Fig. 5) and (b) BARD1h is promoting BRCA2-Aurora B interaction, which is likely to be crucial for midbody formation and Acknowledgments abscission (Fig. 6). Received 6/4/2008; revised 10/29/2008; accepted 11/9/2008. In conclusion, we have shown that BARD1 and its cancer- Grant support: Swiss National Science Foundation grant 3100-068222 (I. Irminger- h Finger), Inserm Ligue Nationale Contre le Cancer and Cance´ropoˆle grants associated isoform BARD1 act as functional link between early (D. Birnbaum), General Secretariat of Research of the Greek Ministry of Development and late phases of mitosis and a few key regulators, BRCA1, Aurora grant 05NON-EU-449 (S. Gagos), and Fondation Pour La Recherche and Cancer and Solidarite´(I. Irminger-Finger and S. Ryser). B, and BRCA2. Thus, the function of FL BARD1, the tumor The costs of publication of this article were defrayed in part by the payment of page suppressor, in Aurora B degradation and the opposing function of charges. This article must therefore be hereby marked advertisement in accordance BARD1h in scaffolding Aurora B and BRCA2 links the hitherto with 18 U.S.C. Section 1734 solely to indicate this fact. We thank D. Suter for sharing precious constructs before publication and A. unrelated cancer-promoting pathways, in which Aurora B and Caillon, L. Kcjancic Curty, S. Arnaudeau, and M. Hiourea for excellent technical BRCA2 are involved. assistance.

References 15. Lotti LV, Ottini L, D’Amico C, et al. Subcellular 31. Cremet JY, Descamps S, Verite F, Martin A, Prigent C. localization of the BRCA1 gene product in mitotic cells. Preparation and characterization of a human aurora-A 1. Grigorova M, Staines JM, Ozdag H, Caldas C, Edwards Genes Chromosomes Cancer 2002;35:193–203. kinase monoclonal antibody. Mol Cell Biochem 2003; PA. Possible causes of chromosome instability: compar- 16. Daniels MJ, Wang Y, Lee M, Venkitaraman AR. 243:123–31. ison of chromosomal abnormalities in cancer cell lines Abnormal cytokinesis in cells deficient in the breast 32. Ruchaud S, Carmena M, Earnshaw WC. Chromo- with mutations in BRCA1, BRCA2, CHK2 and BUB1. cancer susceptibility protein BRCA2. Science 2004;306: somal passengers: conducting cell division. Nat Rev Mol Cytogenet Genome Res 2004;104:333–40. 876–9. Cell Biol 2007;8:798–812. 2. Irminger-Finger I, Soriano JV, Vaudan G, Montesano 17. Delaval B, Ferrand A, Conte N, et al. Aurora 33. Ducat D, Zheng Y. Aurora kinases in spindle R, Sappino AP. In vitro repression of Brca1-associated B-TACC1 protein complex in cytokinesis. Oncogene 2004; assembly and chromosome segregation. Exp Cell Res RING domain gene, Bard1, induces phenotypic 23:4516–22. 2004;301:60–7. changes in mammary epithelial cells. J Cell Biol 18. Boulton SJ, Martin JS, Polanowska J, Hill DE, Gartner 34. Adams RR, Eckley DM, Vagnarelli P, et al. Human 1998;143:1329–39. A, Vidal M. BRCA1/BARD1 orthologs required for DNA INCENP colocalizes with the Aurora-B/AIRK2 kinase on 3. McCarthy EE, Celebi JT, Baer R, Ludwig T. Loss of repair in Caenorhabditis elegans. Curr Biol 2004;14:33–9. chromosomes and is overexpressed in tumour cells. Bard1, the heterodimeric partner of the Brca1 tumor 19. Jefford CE, Feki A, Harb J, Krause KH, Irminger- Chromosoma 2001;110:65–74. suppressor, results in early embryonic lethality and Finger I. Nuclear-cytoplasmic translocation of BARD1 35. Adams RR, Carmena M, Earnshaw WC. Chromo- chromosomal instability. Mol Cell Biol 2003;23:5056–63. is linked to its apoptotic activity. Oncogene 2004;23: somal passengers and the (aurora) ABCs of mitosis. 4. Wu LC, Wang ZW, Tsan JT, et al. Identification of a 3509–20. Trends Cell Biol 2001;11:49–54. RING protein that can interact in vivo with the BRCA1 20. Hayami R, Sato K, Wu W, et al. Down-regulation of 36. Smith SL, Bowers NL, Betticher DC, et al. Over- gene product. Nat Genet 1996;14:430–40. BRCA1-1 ubiquitin ligase by CDK2. Cancer Res 2005;65: expression of aurora B kinase (AURKB) in primary non- 5. Shakya R, Szabolcs M, McCarthy E, et al. The basal-like 6–10. small cell lung carcinoma is frequent, generally driven mammary carcinomas induced by Brca1 or Bard1 21. Stacey SN, Sulem P, Johannsson OT, et al. The BARD1 from one allele, and correlates with the level of genetic inactivation implicate the BRCA1/BARD1 heterodimer Cys557Ser variant and breast cancer risk in Iceland. instability. Br J Cancer 2005;93:719–29. in tumor suppression. Proc Natl Acad Sci U S A 2008; PLoS Med 2006;3:e217. 37. Cully M, Shiu J, PiekorzRP, Muller WJ, Done SJ, Mak 105:7040–5. 22. WuJY,VlastosAT,PelteMF,etal.Aberrant TW. Transforming acidic coiled coil 1 promotes 6. Irminger-Finger I, Leung WC, Li J, et al. Identification expression of BARD1 in breast and ovarian cancers transformation and mammary tumorigenesis. Cancer of BARD1 as mediator between proapoptotic stress and with poor prognosis. Int J Cancer 2006;118:1215–26. Res 2005;65:10363–70. p53-dependent apoptosis. Mol Cell 2001;8:1255–66. 23. Li L, Ryser S, Dizin E, et al. Oncogenic BARD1 38. Nguyen HG, Chinnappan D, Urano T, Ravid K. 7. Irminger-Finger I, Jefford CE. Is there more to BARD1 isoforms expressed in gynecological cancers. Cancer Res Mechanism of Aurora-B degradation and its dependen- than BRCA1? Nat Rev Cancer 2006;6:382–91. 2007;67:11876–85. cy on intact KEN and A-boxes: identification of an 8. Stark JM, Pierce AJ, Oh J, Pastink A, Jasin M. Genetic 24. Feki A, Jefford CE, Berardi P, et al. BARD1 induces aneuploidy-promoting property. Mol Cell Biol 2005;25: steps of mammalian homologous repair with distinct apoptosis by catalysing phosphorylation of p53 by DNA- 4977–92. mutagenic consequences. Mol Cell Biol 2004;24:9305–16. damage response kinase. Oncogene 2005;24:3726–36. 39. Chiba N, Parvin JD. The BRCA1 and BARD1 9. Westermark UK, Reyngold M, Olshen AB, Baer R, Jasin 25. Li L, Cohen M, Wu J, et al. Identification of BARD1 association with the RNA polymerase II holoenzyme. M, Moynahan ME. BARD1 participates with BRCA1 in splice-isoforms involved in human trophoblast invasion. Cancer Res 2002;62:4222–8. homology-directed repair of chromosome breaks. Mol Int J Biochem Cell Biol 2007;39:1659–72. 40. Gergely F, Karlsson C, Still I, Cowell J, Kilmartin J, Cell Biol 2003;23:7926–36. 26. Fabbro M, Savage K, Hobson K, et al. BRCA1–1 Raff JW. The TACC domain identifies a family of 10. Starita LM, Machida Y, Sankaran S, et al. BRCA1- complexes are required for p53Ser-15 phosphorylation centrosomal proteins that can interact with micro- dependent ubiquitination of g-tubulin regulates centro- and a G1-S arrest following ionizing radiation-induced tubules. Proc Natl Acad Sci U S A 2000;97:14352–7. some number. Mol Cell Biol 2004;24:8457–66. DNA damage. J Biol Chem 2004;279:31251–8. 41. Li JJ, Li SA. Mitotic kinases: the key to duplication, 11. Joukov V, Groen AC, Prokhorova T, et al. The BRCA1/ 27. Wiznerowicz M, Trono D. Conditional suppression of segregation, and cytokinesis errors, chromosomal insta- BARD1 heterodimer modulates ran-dependent mitotic cellular genes: lentivirus vector-mediated drug-inducible bility, and oncogenesis. Pharmacol Ther 2006;111:974–84. spindle assembly. Cell 2006;127:539–52. RNA interference. J Virol 2003;77:8957–61. 42. Stewart S, Fang G. Destruction box-dependent 12. Hsu LC, Doan TP, White RL. Identification of a 28. Arrighi JF, Pion M, Wiznerowicz M, et al. Lentivirus- degradation of aurora B is mediated by the anaphase- g-tubulin-binding domain in BRCA1. Cancer Res 2001; mediated RNA interference of DC-SIGN expression promoting complex/cyclosome and Cdh1. Cancer Res 61:7713–8. inhibits human immunodeficiency virus transmission 2005;65:8730–5. 13. Scully R, Chen J, Ochs RL, et al. Dynamic changes of from dendritic cells to T cells. J Virol 2004;78:10848–55. 43. Vader G, Medema RH, Lens SM. The chromosomal BRCA1 subnuclear location and phosphorylation state 29. Suter DM, Dubois-Dauphin M, Krause KH. Genetic passenger complex: guiding Aurora-B through mitosis. are initiated by DNA damage. Cell 1997;90:425–35. engineering of embryonic stem cells. Swiss Med Wkly J Cell Biol 2006;173:833–7. 14. Jin Y, Xu XL, Yang MC, et al. Cell cycle-dependent 2006;136:413–5. 44. Dulyaninova NG, Bresnick AR. The long myosin colocalization of BARD1 and BRCA1 proteins in discrete 30. Zufferey R, Nagy D, Mandel RJ, Naldini L, Trono D. light chain kinase is differentially phosphorylated nuclear domains. Proc Natl Acad Sci U S A 1997;94: Multiply attenuated lentiviral vector achieves efficient during interphase and mitosis. Exp Cell Res 2004;299: 12075–80. gene delivery in vivo. Nat Biotechnol 1997;15:871–5. 303–14.

Cancer Res 2009; 69: (3). February 1, 2009 1134 www.aacrjournals.org

Distinct Roles of BARD1 Isoforms in Mitosis: Full-Length BARD1 Mediates Aurora B Degradation, Cancer-Associated BARD1β Scaffolds Aurora B and BRCA2

Stephan Ryser, Eva Dizin, Charles Edward Jefford, et al.

Cancer Res 2009;69:1125-1134. Published OnlineFirst January 27, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-08-2134

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2009/01/26/0008-5472.CAN-08-2134.DC1

Cited articles This article cites 42 articles, 20 of which you can access for free at: http://cancerres.aacrjournals.org/content/69/3/1125.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/69/3/1125.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/69/3/1125. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.