Importin KPNA2 Is Required for Proper Nuclear Localization and Multiple Functions of NBS1*

Nijmegen breakage syndrome (NBS) is a chromosomal-instability syndrome associated with cancer predisposition, radiosensitivity, microcephaly, and growth retardation. The NBS gene product, NBS1, is a component of the MRE11-RAD50-NBS1 (MRN) complex, a central player associated with double strand break (DSB) repair. In response to radiation, NBS1 is phosphorylated by ATM, and the MRN complex relocalizes to form punctate nuclear foci for DNA repair. NBS1 controls both the nuclear localization of the MRN complexes and radiation-induced focus formation. We report here that the KPNA2 (importin α1) is important for the normal nuclear localization of the MRN complex and its proper formation of the nuclear foci. KPNA2 is the only member of the importin α family that physically interacts with NBS1, and the KPNA2-mediated nucleus localization sequence (NLS) is mapped to amino acid residues 461-467 of NBS1 that is sufficient for both the interaction with KPNA2 and the proper nuclear localization. Inhibition of KPNA2 or blockage of the KPNA2 interaction with NBS1 results in a reduction of radiation-induced nuclear focus accumulation, DSB repair, and cell cycle checkpoint signaling of NBS1. Collectively, our results strongly suggest that an interaction with KPNA2 contributes to nuclear localization and multiple tumor suppression functions of the NBS1 complex.

tional complementation studies identified a single gene, NBS1, which is mutated in most patients with NBS (6,7).
The NBS1 gene is located on human chromosome 8q21 (8) and encodes a ubiquitously expressed protein of 754 amino acids termed nibrin, p95, or NBS1 (6,7,9). NBS1 is a key regulator of the MRN complex (9 -11). This complex plays important roles in the early processing of DSBs via its DNA binding and nuclease activities, participates as a double strand break sensor, and recruits ATM to broken DNA molecules (9,(12)(13)(14)(15)(16)(17)(18)(19). The MRN complex is also known to be involved in the maintenance of telomeres, which have DSB-like structures and defects here can cause telomeric fusion (20) and abnormal patterns of telomere recombination (21). It has been shown that the MRN complex forms nuclear foci after ionizing radiation (IR) (13). While such foci are also detectable in un-irradiated cells, the average number per cell and the frequency of cells with detectable foci increases in response to irradiation (18,19,22). The function of these irradiation-induced foci is unknown, but these foci may represent sites of ongoing repair or of unresolved breaks. In NBS cells, which express truncated nibrin, MRE11 and RAD50 still interact, but complexes containing these two proteins are confined to the cytoplasm and thus cannot form nuclear foci (9). This suggests that one of the major roles of NBS1 is to carry MRE11 and RAD50 into nucleus. The N-terminal portion of NBS1 contains two adjacent and potentially functional domains, a forkhead-associated (FHA) domain and a breast cancer C-terminal (BRCT) domain ( Fig. 1A) (23,24), which have been observed previously in other proteins involved in DNA damage responses or in cell-cycle checkpoint control. The C-terminal region of NBS1 binds to the MRE11-RAD50 complex and the ATM. Moreover, NBS1 contains three potential NLSs (16,17). Despite extensive study of NBS1, there is little evidence that it has any specific enzymatic activity, but rather serves mainly as a molecular chaperone, guiding MRE11-RAD50 complex and ATM to the sites of damage.
Nucleocytoplasmic transport of large complex is mediated by soluble receptors that recognize structural features (NLS and nuclear export signal) in their cargoes. In general, the transport apparatus is versatile, with individual receptors binding a variety of cargoes and vice versa (25,26). Most transport receptors belong to a large family of homologous proteins known as karyopherins or importins. They share limited sequence identity (15-25%) but adopt similar conformation. In human cells, at least twenty-two importin ␤ and six importin ␣ have been identified (25,26). All importin ␤ family members contain an N-terminal Ran-GTP-binding motif, and selectively bind nucleoporins of the nuclear pore complex (27). Nuclear translocation of target proteins through adaptor importin ␣ usually requires importin ␤, which allows passage of the complex through the nuclear pore complex. In the nucleus, the binding of Ran-GTP to importin ␤ induces its dissociation with importin ␣, thereby allowing the release of the cargo. Importin ␣ subsequently recycles to the cytoplasm via interaction with export pro- . The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Both authors contributed equally to this work. 2 To whom correspondence should be addressed. teins. Importin ␣ can also work independently by direct binding to cargoes without the help of importin ␤ (28). Although some nuclear proteins are delivered into the nucleus with the aid of the adaptor function of importin ␣, most proteins that carry cargoes directly through nuclear pore complex are members of the importin ␤ family (25,29). Furthermore, several importin ␤ can also act as adaptor to form heterodimeric importin ␤ transporters (29).
Here we have isolated an importin ␣, KPNA2, in a two-hybrid screen for the NBS1-interacting protein. We mapped the sites of interaction between the NBS1 and KPNA2 proteins in vitro using yeast two-hybrid analysis and in vivo by co-immunoprecipitation. We found that abilities of NBS1 in nuclear focus formation, DSB repair, and DSB-dependent checkpoint signaling were compromised in a KPNA2i background or in strains expressing NBS1 mutants that prevent their interaction with KPNA2. Our data fit the model that KPNA2 plays a role as an import receptor of an NBS1-containing cytoplasmic MRN complex.

MATERIALS AND METHODS
Chemicals and Antibodies-G418 was purchased from Invitrogen. All other chemicals were purchased from Sigma. The primary antibodies used were as follows: anti-FLAG M2 from Sigma; anti-Myc from Dr. L.-H. Huang (National Taiwan University Hospital); anti-KPNA2, p53, and CHK2 from Santa Cruz Biotechnology; anti-GAPDH H86504M from Biodesign; anti-pS15-p53 and pT68-CHK2 from Cell Signaling; anti-NBS1 (NB 100 -143) from Novus; and anti-NBS1 (MS-NBS1-PX1) from GeneTex. The secondary antibodies for immunofluorescence staining were Rhodamine Red TM -conjugated goat anti-rabbit and fluorescein isothiocyanate-conjugated goat anti-mouse antibodies (Molecular Probes).
Two-hybrid Screening-The two-hybrid assay was done essentially as described (32) using components generously provided by Dr. P. James and colleagues. The prey HeLa library (pGAD/X) was transformed into PJ69 -4 strain containing the NBS1 bait plasmid. Positive clones were selected on SC-Ura, -Leu, and -Ade plates and confirmed for interaction on SC-Ura, -Leu, and -His plates.
Mutagenic PCR-To conduct the reverse two-hybrid assay, the bait NBS1 plasmid was used for mutagenic PCR. NBS1 was amplified from the plasmid pGBDU-NBS1 using PCR amplification with primers that contained ϳ20 nucleotides of the vector extending outwards from the cloning sites. The 5Ј primer, pGBDU-5Ј, a 791-808 sequence, was 5Ј-TTCGATGATGAAGATACC-3Ј; the 3Ј primer, pGBDU-3Ј, a 1022-1000 sequence, was 5Ј-GATCAGAGGTTACATGGCCAAGA-3Ј. Mutagenic PCR was carried out using mixed dATP and dITP with a ratio of 9:1. The PCR conditions were as follows: 94°C for 1 min and then 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 2 min.
Reverse Two-hybrid Screen-We used the two-hybrid system to screen for KPNA2 interaction-defective NBS1 proteins. To create the necessary strain, PJ69-4A cells were transformed with the pGAD-KPNA2 plasmid (prey). The PCR products of mutagenic NBS1 were gel-purified and co-transformed with PstI-KpnI linearized pGBDU-NBS1 plasmid (bait). Because the mutagenic PCR product contains sequences homology to the gapped vector, it serves as a substrate for homologous recombination via DNA repair pathways (gapped-repair transformation (33)), creating a library of mutated NBS1 alleles. Transformants of the two-hybrid strain were grown at 30°C and selected on SC-Ura, -Leu media, which selected for the GAL activating-KPNA2 fusion plasmid (-Leu), and the library of NBS1 mutations with the GAL DNA binding plasmid (-Ura). Transformants that could grow on this selective medium have repaired the gapped pGBDU-NBS1 plasmid with the PCR-mutagenized NBS1 gene. Positive clones were replicated and selected on SC-Ura, -Leu, and -Ade plates to screen for loss of interaction.
NLS Prediction and Site-specific Mutagenesis-A program, PSORTII (psort.nibb.ac.jp), predicted that amino acid sequence 461 PSTKKRE 467 might be a potential NLS. Four point mutations, S397A, G583K, and KR465AA, were introduced into the NBS1 using QuikChange site-directed mutagenesis (Stratagene). Sequences of primers for mutagenesis are available upon request.
Cell Culture and Transfections-293T, HeLa, and NBS (provided by Dr. M. Z. Zdzienicka) cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 units/ml penicillin G, and 100 g/ml streptomycin at 37°C in a humidified 5% CO 2 atmosphere. NBS1i cells were generated as we previously described (Chiang et al. (48)). For pSUPER and pSUPER-KPNA2i transfections, cells were seeded at 70% confluence and transfected with each DNA using Effectene (Qiagen). After 24 h of transfection, cells were plated and selected in media containing 400 g/ml G418 (Invitrogen). The expression was confirmed by Western blotting with the anti-NBS1 as well as anti-KPNA2 Abs and by fluorescence microscopy.
Western Blot Analysis-Cell lysates were prepared with lysis buffer (150 mM NaCl, 1% Nonidet P-40, 1% deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 7.5, and protease inhibitors). Nuclear extracts were prepared with an NE nuclear extraction system (Pierce). Cell lysates con-taining equal amounts of protein were separated by SDS-PAGE and electroblotted to a Hybond membrane (Amersham Bioscience). The filters were probed with NBS1, KPNA2, p53, p53-p, CHK2, and CHK2-p antibodies and a polyclonal anti-GAPDH antibody as a control for protein loading, and then visualized by chemiluminescent detection (ECL, Amersham Biosciences).
Immunoprecipitation Procedure-Immunoprecipitation procedure was performed by incubating 500 g of whole cell extract prepared under cell lysis buffer (50 mM Tris, pH 7.5, 50 mM NaCl, 0.1% SDS, 1% Nonidet P-40, and 0.5% sodium deoxycholate) from each clone with the anti-NBS1, anti-KPNA2, anti-FLAG, or anti-Myc antibody for 3 h, and the immune complexes were incubated overnight with protein-G (blocked with 10% bovine serum albumin) (Pierce). The immunoprecipitate was washed three times with TNTG buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 0.1% Triton X-100, 10% glycerol), mixed with 1ϫ Laemmli dye, boiled for 10 min, and loaded to an 8% SDS gel. After transferred, the filter was blocked, probed with the primary and secondary antibodies sequentially, and developed. Data shown here are representative of two or more experiments from independent cell cultures.
Immunofluorescence Staining-The detection of nuclear foci following exposure to ionizing radiation was performed as previously described (34). Briefly, cells were grown on glass coverslips and were either mock-treated or irradiated with 15 Gy of ionizing radiation. After 8 h, cells were rinsed with phosphate-buffered saline (phosphate-buffered saline) and fixed in phosphate-buffered saline-buffered 3% paraformaldehyde and 2% sucrose at room temperature for 10 min, followed by permeabilization in Triton X buffer (0.5% Triton X-100, 20 mM Hepes-KOH, pH 7.9, 50 mM NaCl, 3 mM MgCl 2 , 300 mM sucrose) at room temperature for 10 min. For dual immunostaining, cells were blocked with 0.5% bovine serum albumin in phosphate-buffered saline and incubated at 4°C for overnight with either mouse anti-NBS1 (1:100) or rabbit anti-MRE11 (1:100) Abs. The primary antibodies were detected using Rhodamine Red-conjugated goat anti-rabbit and fluorescein isothiocyanate-conjugated goat anti-mouse antibodies (1:500). DNA was stained with Hoechst at room temperature for 1 h. Immunofluorescence was analyzed with a Zeiss Axioplan fluorescence microscope. At least 100 cells were counted for each sample.
Plasmid End-joining Assay-Plasmid end-joining assay was done as previously described (35). pGL3 plasmid (Promega) was completely linearized by restriction endonucleases HindIII or NarI and confirmed by agarose gel electrophoresis. The linearized DNA was subjected to phenol/chloroform extraction, ethanol-precipitated, and dissolved in sterilized water. DNA was then transfected into cells. The transfected cells were harvested and assayed for luciferase activity by Luciferase assay system (Promega) (36).
DSB-dependent Checkpoint Assay-DSB-dependent phosphorylation of CHK2 and p53 was determined as previously described (37). Briefly, NBS fibroblasts were transfected with pSUPER or pSUPER-KPNA2i. 48 h later, cells were transfected with pFLAG-CMV, pFLAG-NBS1, or pFLAG-NBS1-465AA. Transfected cells were harvested before or 30 min after irradiated with 9 Gy of IR. Nuclear extracts were prepared and subjected to Western blot analysis.

RESULTS
Identification of KPNA2 (importin ␣1) as an NBS1-interacting Protein Using the Yeast Two-hybrid Assay-To identify proteins that regulate NBS1 functions, we performed a yeast two-hybrid screening of a HeLa cDNA library. We constructed a plasmid vector (pGBDU-NBS1) that expressed a fusion protein with the DNA-binding domain of the Gal4 protein and NBS1. Transformation of pGBDU-NBS1 alone (the "bait") into the yeast strain PJ69 -4 did not activate reporter transcription. Of ϳ5 ϫ 10 6 clones screened, eight positive clones were obtained and sequenced. Among eight positive clones, one clone encodes TRF1 and another clone encodes a functional unknown gene. TRF1 was previously demonstrated to be an NBS1-interacting protein at telomeres (18). The remaining six independent clones encoded a single protein KPNA2 (importin ␣1), which differed only in length. The six KPNA2 clones lacked different sections of the N terminus but were capable of encoding the truncated forms of KPNA2 in-frame. The KPNA2 protein contains an N-terminal hydrophilic importin ␤ binding domain, a central hydrophobic cargo binding region composed of eight "armadillo" repeats, and a short hydrophilic C terminus (CAS binding domain). Fig.  1B indicates the schematic structures of KPNA2 and variant clones identified in our screening. These findings indicated that the armadillo repeats are maintained in all of the positive clones, suggesting the importance of this region for the interaction of NBS1. In an independent yeast two-hybrid screen using the KPNA2 as the bait and a different The IBB domain is the importin ␤ binding domain. C, two-hybrid interaction between NBS1 and KPNA2. Wild-type and mutants of NBS1 fused to the GAL4 DNA binding domain were cloned in pGBDU to test the interacting with KPNA2 in a two-hybrid assay. Yeasts were transformed with the indicated constructs (with the bait plasmid on top), and the growth of yeast transformants on a selective synthetic complete medium with or without adenine is shown. D, specific interaction between NBS1 and KPNA2 among three importin ␣ families. Interaction between NBS1 and the importin ␣ family (KPNA2, Qip1, and NPI-1) determined with a yeast two-hybrid assay. The growth of yeast transformants on selective synthetic complete medium with or without adenine is shown.
human two-hybrid library as the prey, a positive clone encoding NBS1 was also obtained (Fig. 1C). 4 These findings demonstrated that KPNA2 interacts with NBS1 in yeast.
We next asked whether the role of KPNA2 in NBS1 interaction was unique when compared with other known importin ␣ family members. In humans, there are at least six importin ␣ molecules, and these can be divided into three subfamilies, ␣1 (KPNA2), ␣3-␣4 (Qip1), and ␣5-␣6-␣7 (NPI-1) (29). Thus, we examined the binding specificity between NBS1 and the importin ␣ family members using the yeast two-hybrid assay. As shown in Fig. 1D, NBS1 interacted with KPNA2 but not with Qip1 or NPI-1, suggesting NBS1 specifically interacts with KPNA2.
To search for functional NLSs and to study the effect of the lack of NBS1-KPNA2 protein interaction on DNA repair, we performed a reverse two-hybrid screen to identify point mutations in NBS1 that disrupted the NBS1-KPNA2 interaction. We used PCR to mutagenize the full-length NBS1 (33,38). The mutagenized fragments were then assessed for their ability to interact with KPNA2 in the two-hybrid assay. Out of ϳ2 ϫ 10 4 colonies, three mutants gave reduced growth abilities on plates lacking adenine. DNA sequencing showed that those three clones had four, five, or six amino acid substitutions. One contained four amino acid substitutions N56S, A183T, S397A, and G583K (Fig.  1A). In light of the multiple mutations found in each of the mutated NBS1, we determined whether any particular mutations of NBS1 were responsible for the interaction. Desai-Mehta and coworkers (30) previously demonstrated that a region between amino acid residues 401 and 652 of NBS1 is sufficient to direct its nuclear localization. Because S397A and G583K are closer to the previously mapped sequences of NBS1 involved in directing its nuclear localization (30), we generated single and double amino acid mutations in S397A and G583K and compared their interaction with KPNA2 by the two-hybrid assay. Interestingly, mutations in the single or double amino acids did not confer loss of interaction, whereas mutations in the all four residues of NBS1 (NBS1-4m) conferred loss of interaction ( Fig. 1C and data not shown).
In parallel, we searched for potential NLSs in NBS1 using a program, PSORTII (psort.nibb.ac.jp). One of the three potential NLSs (461-467, 590 -594, and 751-754) was identified at amino acid residue 461-467 that is within the NBS1 401-652 (Fig. 1A). Additionally, the observation that a truncated version of NBS1-(1-540) could still interact with KPNA2 in a two-hybrid test (data not shown) ruled out the possibility of aa 590 -594 and 751-754 required for the KPNA2 interaction. Sitedirected mutagenesis was performed to change lysine and arginine residues at 465 and 466, respectively, to alanine. This NBS1-465AA mutant was subjected to a two-hybrid test and showed loss of interaction (Fig. 1C). This result demonstrated that the potential NLS at amino acid residue 461-467 of NBS1 is required for the NBS1-KPNA2 interaction in yeast.
NBS1-KPNA2 Binding in Vivo-To determine the interaction between NBS1 and KPNA2 in human cells, we analyzed the binding of FLAG-tagged NBS1 and Myc-tagged KPNA2. 293T cells were co-transfected with FLAG-NBS1 and KPNA2-Myc expression constructs and subjected to co-immunoprecipitation experiments with anti-Myc antibody, FLAG-NBS1 but not the empty vector, FLAG-NBS1-465AA, or FLAG-NBS1-4m controls precipitated by KPNA2-Myc from cell extracts (Fig. 2A). These results indicated the interaction between NBS1 and KPNA2. To further determine whether endogenous KPNA2 binds to NBS1, 293T cells were transfected with FLAG-NBS1 and mutant expression constructs (Fig. 2B) and subjected to immunoprecipitation with anti-FLAG antibodies. Endogenous KPNA2 was detected in anti-FLAG immunoprecipitates of only FLAG-NBS1 but not the FLAG-NBS1-465AA or FLAG-NBS1-4m mutants, suggesting that this binding is mediated through the mapped binding positions in yeast. To further test whether endogenous NBS1 and KPNA2 associate with each other, 293T cells without transfection were subjected to co-immunoprecipitation experiments (Fig. 2C). Endogenous KPNA2 was found in anti-NBS1 immunoprecipitates. Likewise, endogenous NBS1 was detected in anti-KPNA2 immunoprecipitates. These results therefore indicated that NBS1 interacts with KPNA2 in human cells.

NBS1-KPNA2 Association Is Required for the Proper IR-induced Nuclear Focus
Formation of NBS1-Based on the findings above, we hypothesized that importin KPNA2 would direct NBS1 into nucleus and execute damage repair at the DNA break sites. To determine whether KPNA2 contributes to the nucleus localization of NBS1, we first generated KPNA2 or NBS1 repression by the plasmid-mediated RNA interference technology in HeLa cells to create stable KPNA2i and NBS1i cell lines (Fig. 3A). Cells grown on glass slides were irradiated at a dose of 15 Gy and doubly stained at 8 h post-irradiation with antibodies for NBS1 and MRE11. NBS1 and MRE11 nuclear foci co-localized in irradiated cells, whereas diffuse nuclear staining for both proteins was observed in un-irradiated control cells (Fig. 3B). Cells with multiple, intense nuclear foci were visually counted, and the quantitative results are presented in Fig. 3C. Reduced nuclear intensity and focus numbers of NBS1 was observed in KPNA2i and NBS1i cells, compared with those of HeLa cells (Fig. 3, B and C). The decrease of KPNA2 expression reduced the percentage of cells with foci from 56 to 28 and the average nuclear focus numbers of NBS1 from 26.6 to 12.7 (Fig. 3C). These results suggested that KPNA2 contributes to the nuclear localization and nuclear focus formation of NBS1.
KPNA2-mediated NBS1 Localization Affects DSB Repair-To examine the potential contribution of KPNA2-mediated nuclear delivery of NBS1 in DSB repair, cells were subjected to a plasmid end-joining assay utilizing transient transfection of a linearized pGL3 plasmid harboring a luciferase reporter gene. When the reporter plasmid was linearized with restriction endonuclease HindIII, which cleaved at the linker region between promoter and coding sequence, any end-joining activity with small deletion or insertion does not affect the expression of the luciferase gene and can be considered as an overall end-joining activity. However, when the reporter plasmid was digested with NarI at the luciferase coding region, only precise end-joining can restore the luciferase activity. Repair efficiency was measured by the luciferase activities of linearized reporter constructs compared with that of the intact plasmid. pFLAG-CMV, FLAG-NBS1, or FLAG-NBS1-465AA was transfected into HeLa, KPNA2i, and NBS1i cells to test whether ablation of the NBS1-KPNA2 interaction affects DSB repair. After co-transfection with the wild-type NBS1, the frequency of DSB repair in HeLa cells was raised to 55% for the HindIII-digested plasmid, whereas co-transfection with the NBS1-465AA mutant in HeLa cells or co-transfection with wild-type NBS1 in NBS1i and KPNA2i cells caused a significant reduction in repair ability (Fig. 5). Furthermore, to completely avoid the endogenous NBS1, we transfected NBS fibroblasts with pFLAG-CMV, FLAG-NBS1, or FLAG-NBS1-465AA. Again, repair ability was significantly decreased in NBS1-465AA mutant-expressing NBS cells compared with wild-type NBS1-transfected NBS cells. Interestingly, DSB Lysates from the HeLa, KPNA2i, pSUPER vector control, and NBS1i cells were Western blotted for KPNA2 and NBS1 to verify the reduced expression of the NBS1 and KPNA2. Blots were reprobed for GAPDH to normalize lanes for protein content. B, repression of KPNA2 interferes with irradiation-induced foci of NBS1 and MRE11. HeLa, NBS1i, and KPNA2i cells were either mock treated or exposed to 15-Gy irradiation. Cells were fixed after irradiation, then co-immunostained for NBS1 and MRE11. Blue images represent Hoechst nuclear DNA staining. C, quantification of the numbers of foci-positive cells and the numbers of foci of NBS1 in each cell are shown as mean numbers and errors (ϮS.D.).  HindIII-or NarI-digested DNA and assayed for luciferase activity. Repair efficiency was calculated from the luciferase activities of linearized reporter constructs compared with that of the uncut circular plasmid. Bottom, mutations at the KPNA2-binding region interfere with ability of DSB repair of NBS1. HeLa, NBS1i, KPNA2i, and NBS cells were transiently transfected with empty plasmid, FLAG-tagged full-length NBS1, or FLAG-tagged full-length NBS1-465AA, and the ability of DSB repair was analyzed by the plasmid end-joining assay.
repair activity was slightly more impaired in NBS1-465AA mutantexpressing NBS cells than mock transfected NBS cells (Fig. 5), suggesting that NBS1-465AA partly acts in a dominant-negative manner in NBS cells. Together, these data indicated that a functional NBS1-KPNA2 interaction is required for the proper DSB repair in vivo.
DSB-dependent Checkpoint Signaling of NBS1 Is Reduced in NBS1-465AA-transfected KPNA2i Cells-The MRN complex defines one of the DSB-dependent ATM-regulated checkpoint signals (12, 22, 37, 39 -41). To further investigate the influence of the NBS1-KPNA2 interaction on signaling in this damage-response pathway, we examined the response of p53 and CHK2 phosphorylation to IR in NBS fibroblasts. In agreement with the previous observations (37), NBS1 status did not significantly affect phosphorylation of p53 on Ser-15 (Fig. 6). In contrast, while cells complemented with wild-type NBS1 efficiently activated CHK2 in response to IR, cells complemented with NBS1-465AA failed to do so (Fig. 6). KPNA2i cells exhibited a reduction in CHK2 activation in wild-type NBS1-complemented NBS cells. Notably, compared with vector-transfected cells, IR-induced CHK2 activation was more decreased in the NBS1-465AA-complemented cells, suggesting that NBS1-465AA acts in a dominant-negative manner in NBS cells. Altogether, the NBS1-KPNA2 interaction is required for the DSB-dependent checkpoint signaling of NBS1.

KPNA2-NBS1 Interaction Is Necessary for the Nuclear Localization and IR-induced Nuclear Focus Formation of the MRN Complex-To
further determine whether the direct interaction between NBS1 and KPNA2 is required for nuclear localization and IR-induced nuclear focus formation of the whole MRN complex, we examined the effect of expression of FLAG-NBS1-465AA in the NBS cells on these phenotypes by IR-treatment and staining with antibodies to FLAG and MRE11. In the NBS cells transfected with the FLAG-NBS1 plasmid, FLAG-NBS1 and MRE11 expression was in nucleus (Fig. 7A). Conversely, in cells transfected with the FLAG-NBS1-465AA construct, which lacked the KPNA2 binding site, FLAG-NBS1 and MRE11 localized to the cytoplasm (Fig. 7A). The NBS1-transfected cell line displayed increased numbers of IR-induced nuclear foci of both NBS1 and MRE11. In contrast, the number of nuclear foci-positive cells and the number of MRE11 foci per cell displayed in FLAG-NBS1-465AAtransfected cells were quantitatively similar to those displayed in mocktransfected cells (Fig. 7B). These results demonstrated that the KPNA2-NBS1 interaction is required for the nuclear localization and IR-induced nuclear focus formation of the whole MRN complex.

DISCUSSION
Two-hybrid screenings in yeast have allowed us to identify a new physical and functional interaction between NBS1 and KPNA2 importin ␣, a family member of proteins involved in the active transport of cargo proteins containing an NLS from the cytoplasm to the nucleus (25,26) (Fig. 1). KPNA2 was previously shown to mediate nuclear transport of some tumor suppressors (42,43). To cross the nuclear membrane and deliver its NLS-containing cargo, KPNA2 may bind importin-␤ through its N-terminal IBB domain (43)(44)(45). In the nucleus, these complex components, including the cargo, become dissociated, allowing the recycling of KPNA2 back to the cytoplasm (46).
We verified the interaction observed in yeast between NBS1 and KPNA2 in immunoprecipitates of endogenous NBS1 and KPNA2 from mammalian cell lysates. We next determined whether this NBS1-KPNA2 interaction was mediated via the NLS of NBS1. As the PSORT analysis of NBS1 protein sequence predicted a potential NLS localized at aa 461-467, we found that the 465AA and 4m mutations of NBS1 blocked the association with KPNA2. Therefore, we predict that aa 461-467 is a functional NLS of NBS, whereas the 4m mutation may cause a conformational change around aa 461-467, which blocks the function of this NLS.
To establish the functional role of this NBS1 NLS in vivo, the intracellular distribution of FLAG-tagged NBS1 in HeLa cells was analyzed by immunofluorescence. We showed that mutations in 465AA and 4m resulted in a cytoplasmic redistribution and a reduction of IR-induced nuclear focus formation of NBS1. This finding, together with the fact that NLS and 4m mutations disrupted the interaction with KPNA2, highlights the importance of NBS1 NLS for its binding to KPNA2 and nuclear translocation of NBS1. We observed that the majority of FLAG-NBS1-465AA has cytoplasmic fluorescence of various intensities. It is worth noting that this cytoplasmic localization was rarely observed in cells expressing FLAG-NBS1. The expression of the FLAG-NBS1-465AA causing a dramatic increase in cytoplasmic NBS1 indicates that KPNA2 is a major adaptor for the nuclear localization of NBS1. It was previously reported as data not shown that NBS1 had three potential NLSs, including 461-467 and deletion of any single NLS alone did not completely block nuclear signal of NBS1 (16,17). Here our data suggested that NLS 461-467 plays a major role for the localization of NBS1. It should be noted, specially, that additional two-hybrid screens in yeast FIGURE 6. The NBS1-KPNA2 association is required for the proper DSB-mediated checkpoint signaling of NBS1. pSUPER-or pSUPER-KPNA2-transfected NBS fibroblasts were transfected with pFLAG-CMV, pFLAG-NBS1, or pFLAG-NBS1-465AA expression constructs. Cells were harvested before (Ϫ) or 30 min after (ϩ) exposure to 9-Gy of IR. Nuclear extracts were prepared and immunoblotted for p53, p53-p, CHK2, and CHK2-p. FIGURE 7. The KPNA2-NBS1 interacting is required for the nuclear localization and IR-induced nuclear focus formation of the MRN complex. A, NBS cells were transiently transfected with the empty plasmid, FLAG-tagged full-length NBS1, or 465AA, and were exposed to 15 Gy of ionizing radiation. The immunofluorescence staining of FLAG-NBS1 and MRE11 were analyzed by immunostaining with FLAG and MRE11 antibodies. B, quantification of the numbers of MRE11 nuclear foci-positive cells and the numbers of nuclear foci of MRE11 in each cell are shown as mean numbers and errors (ϮS.D.). NOVEMBER 25, 2005 • VOLUME 280 • NUMBER 47 with different cDNA libraries did not reveal any interaction with the importin family other than KPNA2 (Fig. 1C and data not shown).

KPNA2 Contributes to Multiple Functions of NBS1
To establish the role played by KPNA2 in the DSB repair of NBS1, the wild-type and mutant NBS1 were analyzed by the plasmid end-joining assay. We showed that, in contrast to normal cells, KPNA2i cells show a considerably reduced activity in DSB repair. Inactivation of the KPNA2interacting sites in NBS1 reduced NBS1 function in DNA repair to a similar level. Furthermore, we speculated that the KPNA2-NBS1 interaction is also critical for the DNA damage-activated cell-cycle checkpoint. This is supported by the observation that inactivation of the KPNA2-interaction with NBS1 reduced the checkpoint signaling upon treatment with ionizing radiation. Therefore, this KPNA2-NBS1 interaction is functionally important.
In conclusion, NBS1 carries a functional NLS located at amino acid 461-467 that is indispensable for KPNA2 binding, and this interaction contributes to nuclear translocation and nuclear focus formation of the MRN complex. Although both the FHA and the BRCT domains of the N-terminal portion of NBS1 are required for its nuclear focus formation and phosphorylation and the C-terminal portion of NBS1 contains the domain for MRE11-interaction (23,24), the data presented here demonstrate that the amino acid sequence 461-467 in the middle region of NBS1 is involved in the association of KPNA2. This interaction contributes to multiple tumor suppression functions of the MRN complex. Interestingly, in yeast Mre11p was shown not to localize to the nucleus in the xrs2 (the yeast ortholog of NBS1) mutant. However, Mre11p fused to an NLS completely suppresses the MMS sensitivity of the xrs2 null strain (47), suggesting that Xrs2p plays its major role as a guider in translocation of the Mre11-Rad50 complex from the cytoplasm to the nucleus in yeast. The functional interaction between KPNA2 and NBS1 provides a molecular explanation of how NBS1 executes its function to carry the cytoplasmic MRN complex into the nucleus.