BRCA1 Interacts with Poly(A)-binding Protein

BRCA1 has been implicated in a number of cellular processes, including transcription regulation, DNA damage repair, cell cycle control, and apoptosis. We identified poly(A)-binding protein 1 (PABP) as a novel BRCA1-interacting protein in a yeast two-hybrid screen and confirmed the interaction by in vitro assays and coimmunoprecipitation in mammalian cells. Endogenous interaction between BRCA1 and PABP was also observed. This interaction was abolished by BRCA1 cancer-associated mutations, suggesting that it may be physiologically relevant. Deletion mapping demonstrated that the RNA recognition motifs 1–4 region of PABP is required to mediate the interaction with BRCA1. To understand the biological function of the BRCA1-PABP complex, we sought to determine whether BRCA1 is a modulator of translation. We showed here that inhibition of endogenous BRCA1 using a small interfering RNA-based approach decreased protein synthesis. Conversely, overexpression of BRCA1 activated translation. Using a RNA transfection approach, we clearly showed that BRCA1 modulates translation, independently of any transcriptional activity. The data presented here suggest that BRCA1 modulates protein synthesis via its interaction with PABP, providing a novel mechanism by which BRCA1 may exert its tumor suppressor function.

The breast cancer susceptibility gene breast cancer 1 (BRCA1) acts as a tumor suppressor gene (1). Germ-line mutations of BRCA1 are found in about 40% of patients with inherited breast cancer and up to 90% of families with breast and ovarian cancer (2). The decreased BRCA1 expression found in approximately one-third of sporadic breast tumors may also play a role in these cancers (3,4). Several lines of approach tend to define the biochemical function of BRCA1 protein. BRCA1 has thus far been involved in regulation of cell cycle check-points, apoptosis, DNA damage repair, transcription, and ubiquitination (5). Although the implication of BRCA1 in these cellular processes has been demonstrated, the exact mechanism of BRCA1 function remains unclear.
The BRCA1 protein contains a RING finger at its N-terminal region (1), two nuclear export signals near its N terminus (6,7), two nuclear localization signals in the central portion of the protein (8), and a tandem of two BRCT domains at its C-terminal region (9 -11). The majority of cancer-associated BRCA1 mutations affect the BRCT tandem, resulting in truncated products lacking one or two BRCT domains (2). These findings together with the observation that deletion of the Brca1 BRCT domains is responsible for tumor development in mice (12) demonstrate that BRCT domains play a central role in the BRCA1 tumor suppressor function.
In this study we further explored new functions of BRCA1 through its BRCT tandem using the yeast two-hybrid approach. We show that BRCA1 interacts with poly(A)-binding protein 1 (PABP), 5 a highly conserved protein involved in mRNA stabilization and translation (19 -21) and recently described as a canonical translation initiation factor (22). We first demonstrated that BRCA1 binds to PABP in vitro as well as in vivo. Notably, this interaction is disrupted by germ line BRCA1 mutations that affect the BRCT repeats. We then provide evidence showing that BRCA1 can stimulate translation, supporting a model in which control of translation would be mediated in part via the BRCA1-PABP interaction.

EXPERIMENTAL PROCEDURES
Chemicals-The primary antibodies used in this study were as follows. Mouse monoclonal antibody to human BRCA1 (OP92) was purchased from Oncogene Research Products, mouse monoclonal antibody to actin was from ICN Biochemicals, mouse monoclonal antibody to Myc (anti-Myc) from Roche Applied Science, and the mouse monoclonal antibody against the synthetic V5 epitope was from Invitrogen. Mouse monoclonal antibodies (as ascites fluids) against human PABP (clone 10E10) were gifts from Gideon Dreyfuss (University of Pennsylvania) (23). Polyclonal antibody to P300 (N-15) was from Santa Cruz Biotechnology. Secondary antibodies used were peroxidase-conjugated anti-mouse immunoglobulin (Amersham Biosciences). Cycloheximide (CHX) was purchased from Sigma.
Mammalian Expression Constructs-The pCDNA3␤ plasmid expressing human BRCA1 full-length protein and the N-terminal hemagglutinin-tagged pCDNA3␤ plasmid expressing the truncated mutant Y1853X were previously described (24).
Renilla luciferase cDNA sequence was PCR-amplified from the pRL-CMV vector (Promega) using the primers 5Ј-CGGGATCCTACTTCGAAAGTTTATGATCCAG-3Ј and 5Ј-TCCCCCGGGTTGTTCATTTTTGAGAACTCGC-3Ј, digested with BamHI and SmaI (PCR added restriction sites), and cloned into the multiple cloning region of the monocistronic plasmid derived from pGEM-2 (Promega), between the T7 promoter and poly(A) stretch. The structure of the resulting DNA construct, named PGEM-Renil, was verified by restriction enzyme digestion and sequencing. The luciferase sequence is under control of the T7 RNA polymerase promoter for in vitro transcription.

Yeast Two-hybrid Screen
The L40 strain (Invitrogen) was cotransformed with pGBT9-BRCT and a Gal4 transactivation domain-tagged 11-day-old mouse embryo cDNA library (Clontech). Yeasts were plated on a Leu/Trp/His amino acid-depleted DOB medium containing 1 mM 3-aminotriazol. Transformants were tested for ␤-galactosidase activity using a yeast colony-filter assay. Positive (blue) colonies were isolated, and the ␤-galactosidase assay was repeated. Plasmids of these colonies were recovered and used to transform an Electromax DH10B bacterial strain. The prey plasmids amplified in bacteria were retransformed with pGBT9 or pGBT9-BRCT into L40 strain and analyzed by DNA sequencing.

Cell Culture and Plasmid Transfection
HBL100 human epithelial mammary cells were maintained in Dulbecco's modified Eagle's medium containing 1 g/liter glucose supplemented with 10% fetal calf serum, 100 g/ml streptomycin, and 100 units/ml penicillin. Bosc human embryonic kidney cells were maintained in Dulbecco's modified Eagle's medium containing 4.5 g/liter glucose (or 1 g/liter glucose for metabolic labeling experiments) supplemented with 10% fetal calf serum, 100 g/ml streptomycin, and 100 units/ml of penicillin.
Bosc cells were plated at 4 ϫ 10 6 cells per 10-cm-diameter dish 24 h before transfection. Cells were transfected with 4 g of BRCA1 expressing vector (pCDNA3␤) and 20 l of ExGen 500 (Euromedex) following the supplier's procedure. Fortyeight hours after transfection cells were processed for immunoblotting or immunoprecipitation. For metabolic labeling experiments Bosc cells were plated at 4 ϫ 10 5 cells per plate in a 6-well flask 24 h before transfection. Cells were transfected with 1 g of pCDNA3.1 control plasmid or pCDNA3␤ plasmid expressing BRCA1 and 10 l of ExGen 500 (Euromedex). Where pCDNA-PAIP1 plasmid was used, cells were transfected with 2 g of plasmid and 20 l of ExGen. Forty-eight hours after transfection cells were processed for metabolic labeling.
Lysates containing 2 mg of protein were precleared by stirring with 100 l of protein G-Sepharose for 1 h at 4°C. After centrifugation immunoprecipitation was performed with the precleared lysate and 2 g of antibody for 3 h at 4°C. Thirty microliters of protein G-Sepharose were added and incubated for 30 min at 4°C. After centrifugation, beads were washed 3 times with lysis buffer A, and proteins were eluted by heating at 95°C for 5 min in SDS-loading buffer and 100 mM dithiothreitol.
Proteins were subjected to SDS-PAGE and blotted onto polyvinylidene difluoride membranes (Immobilon-P, Millipore). Membranes were blocked in Tris-buffered saline solution containing 0.05% Tween 20 and 5% nonfat milk and incubated with primary antibodies. Horseradish peroxidase-conjugated secondary antibodies (Amersham Biosciences) were used for detection of immunoreactive proteins by chemiluminescence (ECL, Amersham Biosciences).

GST Pulldown Assay
Cells were lysed in lysis buffer A supplemented with protease inhibitor (Complete EDTA-free, Roche Applied Science). One mg of protein was incubated for 1 h with 100 l of 50% glutathione-Sepharose beads solution to eliminate nonspecific interactions. After a centrifugation of 2 min at 5000 rpm, supernatants were incubated with 5 g of GST or GST-BRCT proteins and 50 l glutathione-Sepharose beads for 3 h at 4°C. Protein complexes were washed three times with buffer A. Protein complexes were released by heating at 95°C in SDS-loading buffer and 100 mM dithiothreitol. Proteins were analyzed by immunoblotting assay as described above. The presence of GST fusion proteins was examined by Coomassie Blue staining. Where indicated, lysates were treated with RNase I (0.1 mg/ml) for 1 h at room temperature.
When the GST pulldown assay was performed on in vitro translated proteins, 2 g of GST or GST-BRCT proteins were incubated with 10 l of [ 35 S]methionine-labeled protein previously synthesized in a reticulocyte lysate-coupled transcription/translation system (Promega). Binding assays were carried out as above.

Far Western Assay
Induced bacterial extracts containing GST, GST-BRCT, and GST-BRCA1-N (containing the N-terminal region of BRCA1) were subjected to SDS-PAGE and transferred to polyvinylidene difluoride membrane. Membranes were blocked in Tris-buffered saline solution containing 0.05% Tween 20 and 5% nonfat milk and incubated for 2 h with 40 l of [ 35 S]methionine-labeled in vitro translated protein in 3 ml of Tris-buffered saline solution containing 0.05% Tween 20 and 2% nonfat milk. Membranes were washed four times, dried, and radiographed.
For transfection, HBL100 cells were plated at 4 ϫ 10 5 cells per plate in a 6-well flask 24 h before transfection. Cells were transfected with 0.125 nmol of siRNA and 2 l of Lipofectamine 2000 (Invitrogen) using the protocol of the supplier. 72 h after transfection cells were processed for immunoblotting or metabolic labeling.

Metabolic Labeling
After transfection of the cells with siRNAs or plasmids, cells were incubated for 1 h with 3 ml of methionine-depleted medium. Then cells were pulse-labeled in 1 ml of methionine-depleted medium supplemented with 20 Ci of [ 35 S] methionine for 1 h. Cells were washed with phosphate-buffered saline and harvested, and cytoplasmic extracts were prepared as follows. Cells were first treated in lysis buffer B (10 mM Hepes-KOH pH 8, 1.5 mM MgCl 2 , 10 mM KCl) supplemented with protease inhibitor (Complete EDTA free, Roche Applied Science) for 15 min and subsequently incubated with buffer B containing 0.5% Nonidet P-40 for 5 min. Cells were briefly centrifuged, and protein concentration was determined by Bradford procedure (Bio-Rad). 10 -20 g of protein were precipitated with 10% trichloroacetic acid. Precipitates were washed with 5% trichloroacetic acid, and radioactivity was determined by scintillation counting. Values were normalized for sample protein contents. When newly synthesized proteins were subjected to SDS-PAGE, cells were lysed in buffer A and analyzed (50 g of protein) using 6% SDS-PAGE, and labeled proteins were visualized by autoradiography. For assessing total protein translation after treatment with CHX, cells were pretreated for 2 h with CHX (10 g/ml) in Dulbecco's modified Eagle's medium, then starved, labeled (treatment with CHX continued during the methionine starvation and the radioactive pulse), and harvested as described above.

In Vitro Transcription
For in vitro transcription DNA was linearized at the EcoRI site. Synthesis of capped transcripts was carried out with the bacteriophage T7 RNA polymerase as previously described (27). GTP concentration was reduced to 0.32 mM, and m7GpppG RNA capping analog (Invitrogen) was added at a concentration of 1.3 mM. The reaction was carried at 37°C for 2 h. The resulting mRNAs were treated with DNase, purified from nonincorporated nucleotides by gel filtration chromatography with RNeasy kit (Qiagen) and precipitated by ethanol. The integrity of the RNAs was checked by electrophoresis on 1% agarose gels, and their concentrations were measured by spectrophotometry. Radiolabeled mRNAs were synthesized as described above in the presence of 10 Ci of [ 32 P]UTP. For in vitro synthesis of unadenylated luciferase mRNAs, the plasmid DNA was linearized at the SmaI restriction site.

Transfection of RNA and in Vivo Translation Assay
Bosc cells were plated at 2 ϫ 10 5 cells per plate in a 12-well flask 24 h before transfection. Cells were transfected with 1 g of pCDNA3.1 control plasmid or pCDNA3␤ plasmid expressing BRCA1, Y1853X, or PAIP1 and 5 l of ExGen 500 (Euromedex). Twenty-four hours after transfection, cells were transfected with 1 g of in vitro synthesized RNA and 4 l of DMRIE-C (Invitrogen). Cells were exposed to the RNA/DM-RIE-C mixture for 4 h, the medium was changed to fresh, serum-containing Dulbecco's modified Eagle's medium, and incubation was continued for an additional 6 h. Cell lysis and the analysis of translational products from transfected mRNA template were done using the Renilla luciferase assay (Promega). Values were normalized for samples protein contents.

Assay for mRNA Stability
Cells were transfected with 1 g of radiolabeled RNAs as described above. 6 h post-transfection, RNAs were isolated with TRIzol reagent (Invitrogen), and 5 g of total RNAs were loaded on a guanidine thiocyanate-denaturing agarose gel as described (28) and submitted to autoradiography.

Statistical Analysis
Results are expressed as the mean Ϯ S.D. Statistical analysis was performed with a two-tailed paired Student's t test.

Identification of PABP as a BRCA1 BRCT Tandem Interacting
Protein-To identify proteins that interact with the tandem of two BRCT domains of BRCA1, we used the two-hybrid system. The mouse BRCA1 residues 1583-1812 were fused to the Gal4 DNA binding domain to generate the bait construct pGBT9-BRCT. The pGBT9-BRCT demonstrated a minimal transactivation activity, which was suppressed by 3-aminotriazol when used at a concentration of 1 mM. A screen of 10 6 transformants of an 11-day-old mouse embryo cDNA library (29) yielded 273 His ϩ colonies. 115 ␤-galactosidase-positive clones were isolated. Sequence analysis of the interacting clones showed that one of them encoded amino acids 205-406 of Pabp.
The protein sequence of the two BRCT domains of mouse Brca1 share 75 and 58% identity with those of human BRCA1, respectively (10), and the protein sequence of mouse Pabp is highly conserved with that of human PABP. Therefore, we examined whether the interaction between BRCA1 and PABP was conserved in human cells.
To test whether human full-length PABP directly binds to the BRCT domains of human BRCA1, a Far Western analysis was performed. The fusion proteins GST, GST-BRCT, and GST-BRCA1-N (containing the N-terminal region of BRCA1) were probed with [ 35 S]methionine-labeled PABP. A specific hybridization was obtained with GST-BRCT only (Fig. 1A). As another approach, a GST pulldown assay was performed on Bosc cell lysates. The specific retention of PABP was observed with GST-BRCT but not with the control GST (Fig. 1B). Therefore, BRCT repeats of BRCA1 interact with full-length PABP in vitro. Because PABP associates with mRNA, it was possible that BRCA1 binds to PABP via RNA. Treatment of the cell lysates with RNase I before GST pulldown shows that the association is not mediated by RNA (Fig. 1B).
To address the functional importance of the BRCT-PABP interaction, we investigated the effect of four germ line BRCA1 mutations located in BRCT domains on in vitro binding with PABP (Fig. 1C). GST pulldown assays were performed on Bosc cell lysates. As shown in Fig. 1D, the two single amino acid substitutions (A1708E and P1749R) and the two truncating mutations (R1835X and Y1853X) abolished the interaction with PABP. These results demonstrate that tumorigenic lesions of the BRCT domains prevent the association of BRCA1 with PABP, suggesting that the BRCA1-PABP interaction may be of physiologic relevance.
Interaction of BRCA1 with PABP in Mammalian Cells-To determine whether BRCA1 could interact with PABP in human cells, lysates from Bosc cells transiently expressing BRCA1 were immunoprecipitated with anti-PABP antibody and further analyzed by immunoblotting using anti-BRCA1 antibody. As shown in Fig. 2A, BRCA1 coprecipitated with PABP. The negative control, an unrelated mouse monoclonal antibody anti-V5, was unable to coprecipitate BRCA1. Anti-PABP immunoprecipitation of lysates from Bosc cells transiently expressing the BRCA1 Y1853X mutant failed to coprecipitate BRCA1 (Fig.  2B). Western blot analysis confirmed that BRCA1 and the Y1853X mutant were expressed at equivalent levels after transfection (Figs. 2, A and B). This reaffirms the physiologic rele-vance of the interaction of BRCA1 with PABP via its BRCT tandem.
To further examine whether endogenous BRCA1 and PABP form a complex, we performed coimmunoprecipitation analysis in untransfected epithelial mammary HBL100 cells. Immunoprecipitation of PABP resulted in the coprecipitation of BRCA1, whereas no bands were observed in negative control (Fig. 2C). In reciprocal experiments, anti-BRCA1 antibody coprecipitated PABP, but not the negative control (Fig. 2D). Therefore, BRCA1 and PABP form an endogenous complex in vivo.
To determine whether BRCA1 and PABP interact in the nucleus or the cytoplasm, Bosc cells overexpressing BRCA1 were fractionated into nuclear and cytoplasmic fractions. The fractionation of these cells was checked with antibodies raised against nuclear P300 and cytoplasmic actin. An immunoprecipitation assay using anti-PABP antibody resulted in the coprecipitation of BRCA1 in the cytoplasmic fraction, whereas no band was observed in the nuclear fraction (Fig. 2E). An immunoprecipitation assay using an unrelated monoclonal antibody anti-V5 was used as control. Therefore, BRCA1 interacts with PABP in the cytoplasm of cells.
PABP Associates with BRCA1 through Its RRM-To map the interaction domain of BRCA1 on PABP, GST pulldown assays were performed on [ 35 S]methionine-labeled fragments of PABP. As shown in Fig. 3, the three N-terminal fragments RRM1 and -2, RRM3 and -4, and RRM1-4 clearly bound to the BRCT domains, whereas the C-terminal fragments C1, C2, and C1ϩC2 of PABP failed to bind. GST did not bind to any of these fragments, and GST-BRCT strongly bound to full-length PABP. It is noteworthy that the intensity of the signals obtained in Fig. 3C (RRM1-4, RRM1 and -2, and RRM3 and -4) are rather proportional to that of the radiolabeled products shown in Fig. 3B, suggesting that the BRCT domains interact with PABP through RRM1 and -2 and RRM3 and -4 with a nearly comparable efficiency.
Therefore, BRCA1 interacts through its BRCT domains with the N-terminal region of PABP encompassing the four RRM motifs. This finding suggests that BRCA1 could participate in regulation of mRNA translation. Indeed, previous studies indicated that the RRM motifs of PABP interact with 3Ј mRNA poly(A) tail (30), eukaryotic translation initiation factor 4G (eIF4G) (31), PAIP1 (32), and PAIP2 (33) and, therefore, are crucial for the translational activity of PABP.
BRCA1 Affects Translation-To investigate the role of BRCA1 in modulating translation, we used a siRNA approach on HBL100 cells. Inhibition of endogenous BRCA1 expression after transfection with two BRCA1-specific siRNA oligonucleotides (Si-BRCA1-1 and Si-BRCA1-2) was analyzed by Western blotting. RNA Interference strongly reduced the expression of BRCA1 protein compared with HBL100 cells transfected with the control siRNA (Si-control), with no effect on the amount of PABP. Actin was used as an internal control (Fig. 4A). To assess the potential effects of RNA Interference -mediated BRCA1 silencing on cellular translation, HBL100 cells transfected with siRNAs were incubated with [ 35 S]methionine. The [ 35 S]methionine incorporation into proteins of HBL100 cells transfected with Si-BRCA1-1 and Si-BRCA1-2) FIGURE 2. BRCA1 and PABP associate in vivo. A, Bosc cells were transfected with an expression plasmid encoding BRCA1 and subjected to immunoprecipitation (IP) with anti-PABP or with the unrelated monoclonal antibody anti-V5. The presence of endogenous PABP and ectopically expressed BRCA1 was determined by immunoblotting using PABP and OP92 antibodies, respectively. WB, Western blot. Input represents 5% of the total cellular lysate. B, Bosc cells were transfected with an expression plasmid encoding the Y1853X mutant and subjected to immunoprecipitation as in A. Input represents 5% of the total cellular lysate. C, HBL100 cells were subjected to immunoprecipitation as in A. The presence of endogenous PABP and BRCA1 proteins was determined using PABP and OP92 antibodies, respectively. Input represents 5% of the total cellular lysate. D, HBL100 cells were subjected to immunoprecipitation with anti-BRCA1 (OP92) or with an unrelated monoclonal antibody anti-V5 (V5). The presence of endogenous PABP and BRCA1 proteins was determined using PABP and OP92 antibodies, respectively. Input represents 5% of the total cellular lysate. E, subcellular localization of the BRCA1-PABP interaction. Nuclear (N) and cytosolic (C) fractions were prepared from Bosc cells overexpressing BRCA1. Equivalent aliquots of each fraction were subjected to an immunoprecipitation assay using anti-PABP antibody or an unrelated monoclonal antibody (V5). The presence of endogenous PABP and ectopically expressed BRCA1 was determined by immunoblotting using PABP and OP92 antibodies, respectively. P300 served as marker for nuclear distribution, and actin served as marker for cytoplasmic distribution.
To further ascertain the role of BRCA1 in cellular translation, Bosc cells were transfected with an expression plasmid encoding BRCA1 and were exposed to [ 35 S]methionine. In these cells expression of BRCA1 protein was largely increased without affecting the amount of PABP (Fig. 4C). In the cells that overexpressed BRCA1, the [ 35 S]methionine incorporation into proteins was significantly increased to 120 Ϯ 3% (mean Ϯ S.D.; n ϭ 3) of the control cells (Fig. 4D). To verify that the increase in translation was not due to a preferential translation of the transcriptional product from the plasmid expressing BRCA1, Bosc cells transfected with plasmid encoding BRCA1 (BRCA1) or empty plasmid (control) were lysed, and pulse-labeled proteins were subjected to 6% SDS-PAGE. As shown in Fig. 4E, no marked band of 220 kDa that could represent BRCA1 was present among the newly synthesized proteins visualized by autoradiography.
We further investigated the effect of the Y1853X mutant lacking the ability to bind to PABP (Figs. 1D and 2B) on the cellular translation. In the cells that overexpressed the Y1853X mutant (Fig. 4C), the [ 35 S]methionine incorporation into proteins was not significantly increased (105 Ϯ 5% (mean Ϯ S.D.; n ϭ 3)), suggesting that the BRCT of BRCA1 is required to achieve the stimulatory effect (Fig. 4D).
As above, treatment with CHX resulted in a strong inhibition of translation rate compared with control cells. We further verified that [ 35 S]methionine incorporation was also increased in Bosc cells that overexpressed PAIP1, a protein previously reported to stimulate translation (32). In these cells overexpression of PAIP1 did not affect the amount of PABP (Fig. 4C). As expected, the [ 35 S]methionine incorporation into proteins was increased (Fig. 4D). Interestingly, increased translation measured in cells overexpressing BRCA1 was similar to that observed in cells overexpressing PAIP1. Taken together, these data show that overexpression of BRCA1 efficiently increased protein synthesis and to the same extent as BRCA1 silencing repressed it, thus confirming the implication of BRCA1 in translation.

BRCA1 Modulates Translation through Its Interaction with PABP-
To exclude the possibility that the effect on translation observed may reflect some transcriptional activity of BRCA1, we used an RNA transfection approach. The Renilla luciferase cDNA was inserted in the PGEM-Renil construct, and capped RNAs were generated in vitro. Transfection of these RNAs in Bosc cells overexpressing BRCA1, the Y1853X mutant, or PAIP1 created conditions in which only modulation of luciferase translation was evaluated, independently of its transcription.
To investigate whether BRCA1 participates in translation, we cotransfected Bosc cells with vectors expressing either BRCA1 or the Y1853X mutant lacking the ability to bind to PABP together with capped and polyadenylated luciferase reporter RNA. In these cells, the Y1853X mutant was expressed at comparable levels to those of wild-type BRCA1 protein, with no effect on the amount of PABP (Fig. 5A, left panel). The effect of BRCA1 or Y1853X on the luciferase protein synthesis was determined by measuring luciferase activity. Transfection with BRCA1 caused about a 140% increase in luciferase activity compared with that with vector alone, whereas transfection with the Y1853X mutant expressing plasmid caused weak stimulation (116%), suggesting that the BRCT of BRCA1 is required to achieve the full stimulatory effect (Fig. 5B). The stimulation of translation obtained with PAIP1 was comparable with that previously described (Figs. 5, A, right panel, and B) (32).
Therefore, overexpression of BRCA1 stimulates translation independently of transcriptional activity. In addition, the BRCA1 BRCT domains are required to achieve the stimulatory effect.
To verify that BRCA1 had no effect on mRNAs stability, mRNA coding luciferase was monitored over time. As shown in Fig. 5C, luciferase mRNA stability was not modified by the presence of BRCA1 or Y1853X over the course of the experiment. This strongly suggests that the increase in luciferase activity is due to differences in the translation efficiency and not due to the stability of the luciferase mRNA.
To further ensure that translational stimulation involved PABP, Bosc cells were cotransfected with capped and unadenylated luciferase reporter RNA together with vector expressing BRCA1. The effect of BRCA1 on the luciferase protein synthesis was determined as above. Transfection with BRCA1 caused no significant increase in luciferase activity compared with that with vector alone (Fig. 5D). Collectively, these data show that BRCA1 modulates translation and not mRNA stability through its interaction with PABP.

DISCUSSION
This study describes for the first time an interaction between BRCA1 and a protein of translation, namely PABP. Initiation of mRNA translation, the rate-limiting step in protein synthesis, is well regulated by initiation factors including eIF4E, eIF4G, and eIF4A (34). Recent studies have shown that PABP also should be considered as a canonical translation initiation factor (22).
In this report we showed that the RRM motifs were required for the interaction of PABP with BRCA1. The RRMs are highly conserved motifs. They support most of the important biochemical functions of PABP. RNA binding studies using yeast (35) and Xenopus (36,37) PABPs demonstrated that RRM1 and -2 binds specifically to poly(A) RNA. Second, RRM1 and -2 suffices for protein synthesis in vitro (38). Finally, RRMs support interactions with eIF4G (31), PAIP1 that stimulates translation (32) and PAIP2 that represses translation (33). Thus, RRM motifs play important roles for translation regulation. Accordingly, our finding that BRCA1 interacts with PABP at the RRM motifs suggests that BRCA1 participates in regulation of mRNA translation. It is interesting to speculate that BRCA1 might modulate the interaction of PABP with eIF4G, PAIP1, or PAIP2, as BRCA1 shares the PABP binding site with these proteins. Further studies will be needed to delineate the entire protein complex containing BRCA1 and its molecular mechanism.
Additional PABP-interacting proteins have recently been identified. DAZL proteins, known to play an essential role in gametogenesis, interact with the C-terminal region of PABP and, therefore, stimulate translation of specific mRNAs (39). Human ataxin-2 (ATX2), whose cellular function is unknown, interacts with PABP via PABC and may have a potential role in RNA metabolism (40). The antiproliferative protein Tob is involved in the translational suppression of IL-2 through its interaction with PABC of PABP (41). Thus, in light of the discovery of these unexpected new regulators of translation, we might speculate that other regulators remain to be identified. Here, BRCA1 represents a novel regulatory protein of translation through binding to PABP.
A number of studies have suggested that translation is linked to cancer. eIF4E is the most strongly implicated in malignancy. eIF4E is overexpressed in a variety of tumors and malignant cell lines including carcinomas of head, neck, lung, breast, colon, and in non-Hodgkin's lymphoma (42). In addition, the function of eIF4E as a bona fide oncogene was established through in vivo studies utilizing transgenic mice (43). Overexpression of eIF4E in immortalized human mammary epithelial cells caused their transformation, as judged by their ability to form foci on a monolayer of cells and grow in soft agar (44). Other members of the eIF4F initiation complex seem to be implicated in cancer. Overexpression of the scaffold protein eIF4G in rodent fibroblast cell lines leads to transformation (45), and amplification of the eIF4G gene has been detected in squamous cell lung carcinoma (46). Elevated levels of the eIF4A helicase are detected in human melanoma and hepatoma cells (47)(48)(49). However, few data to date implicate PABP in tumorigenesis. In the early stages of cancer, an increase of cellular PABP level was observed, suggesting a link between growth control and PABP level (50). Here, the novel interaction of PABP with the tumor suppressor BRCA1 protein provides new elements to implicate PABP in malignancy.
Unlike the well established involvement of defective cellular processes like DNA repair and genome stability in predisposition toward malignant transformation, dysregulation of protein synthesis has only recently been recognized as a major step in tumorigenesis. In the present study we found that BRCA1 overexpression results in a moderate but global increase of protein synthesis. This was further demonstrated by using an RNA transfection approach showing that a BRCA1 mutation located in the BRCT repeats affected its translational activity. Conversely, the knockdown of endogenous BRCA1 by using small interfering RNAs resulted in an inhibition of total cellular FIGURE 4. BRCA1 affects protein translation. A and B, HBL100 cells were transfected with control siRNA (Si-control) or siRNAs directed against BRCA1 (Si-BRCA1-1 and Si-BRCA1-2). Seventy-two hours after transfection, cells were treated as follows. A, BRCA1 and PABP proteins were detected from cell lysates by immunoblotting analysis using OP92 and anti-PABP antibodies, respectively. Actin protein was used as internal control. WB, Western blot. B, cells were pulse-labeled with [ 35 S]methionine for 1 h, and radioactivity was determined from protein extracts. CHX, non-transfected cells were treated with CHX for 2 h before labeling and during pulse labeling. Values were normalized for sample protein contents. The [ 35 S]methionine incorporation measured in cells transfected with Si-control was set at 100%. Values represent the mean Ϯ S.D. of experiments performed in triplicate. ***, p Ͻ 0.001 compared with Si-control. C and D, Bosc cells were transfected with expression plasmids encoding BRCA1, the Y1853X mutant, PAIP1, or empty plasmid (Control). Forty-eight hours after transfection, cells were treated as follows. In C, BRCA1 and the Y1853X mutant were detected by immunoblotting analysis using OP92. PAIP1 and PABP proteins were detected using anti-Myc and anti-PABP antibodies, respectively. Actin protein was used as internal control. In D, cells were pulse-labeled with [ 35 S]methionine for 1 h, and radioactivity was determined from protein extracts. CHX, nontransfected cells were treated with CHX for 2 h before labeling and during pulse labeling. Values were normalized for sample protein contents. The [ 35 S]methionine incorporation measured in cells transfected with the control plasmid was set at 100%. Values represent the mean Ϯ S.D. of experiments performed in triplicate. ***, p Ͻ 0.001 compared with control plasmid. The threshold for significance was set at p Ͻ 0.05. No significant difference was found between the control and the Y1853X mutant. E, Bosc cells transfected with expression plasmid encoding BRCA1 or empty plasmid (Control) were pulse-labeled with [ 35 S]methionine for 1 h. Cell lysates (50 g of protein) were subjected to 6% SDS-PAGE, and newly synthesized proteins were visualized by autoradiography. translation by the same order of magnitude. Because BRCA1 Y1853X mutant does not bind PABP, our data clearly show that the interaction of BRCA1 with PABP is required for the translational stimulatory activity of BRCA1 to be effective. Our data using the unadenylated luciferase mRNA transfection approach further reinforced the notion that the activity of BRCA1 involves PABP. The bulk of our experiments demon-strate that BRCA1 modulates translation without affecting mRNA stability through its interaction with PABP. Interestingly, in our experimental assay these effects were very similar to those obtained upon overexpression of PAIP1, previously described as a PABP-interacting protein with the ability to enhance translation (32). Because PAIP1 shares part of the PABP RRMs binding site with BRCA1, it is interesting to speculate that BRCA1 and PAIP1 might present an additive effect on translation.
Although our data suggest that BRCA1 is clearly not an essential factor for translation, they nevertheless indicate that the interaction between PABP and BRCA1 has a physiological relevance by exerting a modulating role on total protein synthesis. More importantly, because these results were obtained by measurement of the metabolic activity of the cell, it remains to be determined whether BRCA1 affects translation of all cytoplasmic cellular mRNAs or whether it specifically targets a specific subset of mRNAs that is involved in cell surveillance or other classes of mRNAs that are tightly regulated at the translational level. In addition, it can also be speculated that the effect of BRCA1 on translation may vary greatly under the physiological status of the cell, e.g. upon induction of the stress response, UV-induced DNA damages, or malignant transformation. Further studies will be needed to answer these questions.
Although BRCA1 was initially believed to be predominantly nuclear (51,52), it was subsequently reported as shuttling between nucleus and cytoplasm (6). Thus, unlike nuclear BRCA1, which has been extensively studied (53), the function of the cytoplasmic form of BRCA1 remains largely unknown. Some studies have suggested an increased cytoplasmic localization of cellular BRCA1 in breast cell lines and in breast tumors (8, 54 -56). In response to radiation-induced DNA damage, a fraction of endogenous BRCA1 was exported from the nucleus to the cytoplasm in MCF7 breast cancer cells (57,58). Finally, we and others have previously shown that ectopically expressed BRCA1 is located predominantly in the cytoplasm (26, 59). . BRCA1 overexpression up-regulates protein synthesis. A-C, capped and polyadenylated luciferase reporter mRNA was transfected into Bosc cells overexpressing BRCA1, the Y1853X mutant, or PAIP1. Bosc cells transfected with empty plasmid served as control. Six hours after transfection, cells were lysed and treated as follow. A, BRCA1 and the Y1853X mutant were detected by immunoblotting analysis using OP92 antibody. PAIP1 and PABP proteins were detected using anti-Myc and anti-PABP antibodies, respectively. Actin protein was used as an internal control. WB, Western blot. B, luciferase protein synthesis was determined by measuring luciferase activity. Values were normalized for sample protein contents. The luciferase activity measured in cells transfected with the control plasmid was set at 100%. Values represent the mean Ϯ S.D. of experiments performed in triplicate. ***, p Ͻ 0.001 compared with control plasmid. C, luciferase mRNA stability was analyzed by agarose gel electrophoresis. The stability of the luciferase RNAs was verified by extracting the radiolabeled RNAs 6 h post-transfection and running them on a denaturing agarose gel as indicated on the figure. As a size control, 3 ng of in vitro transcribed luciferase RNA that has not been transfected into the cells was run in parallel (separate lane). D, capped and unadenylated luciferase reporter mRNA was transfected into Bosc cells overexpressing BRCA1. Bosc cells transfected with empty plasmid served as control. Six hours after transfection, cells were lysed, and luciferase protein synthesis was determined by measuring luciferase activity as above. The threshold for significance was set at p Ͻ 0.05. No significant difference was found between the control and BRCA1. Luciferase mRNA stability was analyzed by denaturing agarose gel electrophoresis as described under "Experimental Procedures." Here, we showed that BRCA1 interacts with PABP in the cytoplasm of cells overexpressing BRCA1. Therefore, it is plausible that the increased protein translation in BRCA1-overexpressing cells mimics a physiological role of BRCA1 in stress conditions such as DNA damage or during malignant transformation. This reinforces the notion that BRCA1 may translate a subset of RNAs specific from certain cellular status. Beside the cap-dependent mechanism of translation, several internal ribosome entry sites containing mRNAs can utilize an alternative mode of translation initiation predominantly during stress and apoptosis (60). In addition, a recent study showed that PABP could stimulate the activity of cellular internal ribosome entry sites (61). It is, therefore, plausible that BRCA1 may act through this specific translational context. BRCA1 acts as a scaffold protein implicated in multiple cellular functions, such as transcription, DNA repair, ubiquitination (5), and recently lipogenesis (62). In this regard, it is noteworthy that the implication of BRCA1 in these functions is greatly linked to the functions of the BRCA1 interacting proteins identified to date (53,62). In our study, the potential of BRCA1 to interact with PABP and to modulate global cellular translation provides a new mechanism by which BRCA1 may exert its tumor suppressor functions.