Nucleoporin NUP153 guards genome integrity by promoting nuclear import of 53BP1

NUP153 is involved in nuclear import of 53BP1

By knocking down 1346 genes involved in the ubiquitin-proteasome system and genes encoding zinc-finger proteins in human U2OS and HeLa cell lines, respectively, we aimed at identifying novel factors required for proper assembly of 53BP1 into radiation-induced foci. The lists of all targeted genes and the respective scores obtained in the screen are presented in Supplementary Tables 1 and 2, and the lists of 10 top-scoring genes (‘hits') in U2OS and HeLa cells are provided in Tables 1 and 2, respectively. The fact that the known DSB regulators MDC1, RNF8 and RNF168 scored prominently in either cell type support the biological relevance of both our overall screening strategy and the additional top-scoring hits (Tables 1 and 2). Although their precise roles within the DDR machinery remains to be elucidated, it is noteworthy that among the top-scoring hits are proteins implicated in signal transduction and regulatory pathways operating through protein modifications, such as ubiquitylation (ANAPC10 and ANAPC4 components of the APC/C ubiquitin ligase, UBA1 ubiquitin activating enzyme and ASB18 – a candidate substrate-recognition component of the SCF-like ECS ubiquitin ligase complex), phosphorylation (MAP3K3 kinase involved in modulation of several key transcription factors) and acetylation (ATXN7, a component of the SAGA acetyl transferase complex).

One of the strongest hits that scored in both cellular backgrounds was NUP153 (Tables 1 and 2), whose downregulation reduced nuclear accumulation of 53BP1 and impaired formation of 53BP1 foci in cells exposed to IR (Figure 1a). Data mining of a previous genome-wide screen analyzing spontaneous DDR foci²⁰ identified only two additional genes among the 26 tested NPC components (NUP93 and SEC13L1) whose knockdown resulted in partially cytoplasmic 53BP1 in a fraction of cells (Supplementary Table 3). In the presented screen of U2OS cells, the impact of NUP153 knockdown on IR-induced 53BP1 focus formation was the most robust, almost as pronounced as the knockdown of MDC1 and RNF168, two known regulators of 53BP1 retention at damaged chromatin that have no known relationship to the nuclear pore and were used here as positive controls (Figure 1b).²⁰

The suppression of 53BP1 nuclear accumulation and focus formation was validated by two independent siRNAs to NUP153 (Figures 1c and d), reproduced in HeLa cells (Supplementary Figure S1A) and observed up to 8 h after IR (Supplementary Figure S1B). Biochemical fractionation confirmed that although total levels of 53BP1 remained unchanged in NUP153-depleted cells, 53BP1 partially redistributed to the cytoplasm after NUP153 knockdown (Figure 1e). To look more closely at 53BP1 localization in NUP153-depleted cells, we acquired images on confocal microscope (Figure 1c). In the efficiently NUP153-depleted cells, 53BP1 localized in the cytoplasm and also around the nucleus, presumably in nuclear pores. The latter localization pattern likely also explains residual amounts of 53BP1 in the nuclear fraction after NUP153 knockdown (Figure 1e), whereas the appearance of a sizable fraction of 53BP1 in the cytoplasmic fraction (Figure 1e) is consistent with the subcellular distribution observed by fluorescence microscopy. The partial distribution between the cytosolic and nuclear fractions seen by immunoblots may partly also reflect the fact that individual cells within the siRNA-treated population showed diverse extent of NUP153 depletion, with consequent variable impact on 53BP1 subcellular localization. In otherwise unperturbed cells, NUP153 knockdown selectively affected nuclear localization of 53BP1, with only modest effects on BRCA1 in subset of cells (Supplementary Figure S2A) and no effects on DDR factors, such as MDC1, NBS1, Rad51 and TopBP1. Finally, 53BP1 nuclear accumulation and formation of spontaneous and IR-induced 53BP1 foci were rescued by expression of siRNA-resistant GFP-NUP153^*, confirming the specificity of the observed phenotype (Figure 1f).

Immunofluorescence analysis revealed that the IR-induced nuclear focus formation of BRCA1, MDC1, γH2AX and RNF168 was not affected by knockdown of NUP153, suggesting that loss of NUP153 specifically had an impact on the localization and foci formation of 53BP1 (Figure 2a). This result is consistent with a report showing that NUP153 knockdown does not impair global transport function of NPC.²² Time-lapse microscopy using a stable U2OS cell line expressing GFP-53BP1¹⁰ uncovered that NUP153 knockdown impaired reentry of 53BP1 into the nucleus at the time of mitotic exit and reassembly of the nuclear envelope (Figure 2b). Although a fraction of 53BP1 eventually entered the nucleus, the bulk of 53BP1 remained in the cytoplasm for most of the cell cycle (our unpublished observation). Together, these data suggest that NUP153 is required for nuclear import of 53BP1.

Knockdown of NUP153 delays DSB repair and impairs cell survival

To test the significance of NUP153 knockdown on DNA repair, we exposed siRNA-treated cells to a moderate dose of IR (2 Gy) and followed the MDC1 nuclear DSB foci in time. The number of nuclei with more than five MDC1 foci detected early (1 h) after IR was very similar in control, NUP153- and RNF168-deficient cells (93%, 91% and 98%, respectively). However, although in control cells MDC1 progressively dissociated from DSBs (consistent with ongoing DNA repair), knockdown of NUP153 and RNF168 (the upstream regulator of 53BP1 used here as a positive control) caused persistence of the MDC1 foci even at the time points when the control cells returned to the pre-damage values (Figures 3a and b; and see Supplementary Figure S1C for siRNA efficiency). We detected only modest changes in the cell cycle profile in NUP153-silenced cells (Supplementary Figure S2B), indicating that indirect effects of cell cycle position were unlikely to account for the observed effects on DNA repair. Consistent with the impaired resolution of repair foci, we found that NUP153 knockdown impaired the long-term survival of U2OS cells exposed to IR (Figure 3c).

Depletion of NUP153 enhances DNA damage caused by replication stress

In addition to the delayed DSB repair and decreased survival after IR, depletion of NUP153 reduced the intensity of 53BP1 nuclear bodies in otherwise unstressed G1 cells (Supplementary Figure S1D). As these nuclear compartments have been associated with shielding DNA lesions arising from unresolved replication intermediates,^{9, 10} we hypothesized that cytoplasmic sequestration of 53BP1 may also undermine cellular responses to replication stress. Indeed, NUP153-depleted cells in which 53BP1 was retained in the cytoplasm showed increased fraction of nuclei with elevated number of γH2AX foci after treatment with low doses of aphidicolin, an inhibitor of DNA polymerases α and δ (Figures 4a and b), consistent with the notion that 53BP1 participates in protecting cells against replication stress. We note that NUP153 depletion reduced, but not completely eliminated, accumulation of 53BP1 in G1 nuclear bodies (Supplementary Figure S1D); it is therefore possible that additional factors whose nuclear import is regulated by NUP153 cooperate with 53BP1 to protect genomic loci affected by replication stress.

The C-terminus of NUP153 is required for 53BP1 nuclear import

To determine which domains of NUP153 are necessary for nuclear import of 53BP1, we performed a functional complementation experiment using full-length and previously described NUP153 mutants²² in which the C-terminal part containing the FG repeats, central zinc finger domain or both were deleted (Figure 5a). As reported, the N-terminal part of NUP153 is required for incorporation into NPC.²³ Among the tested NUP153 variants, only the fragment containing the intact N and C termini (N+C) was able to fully rescue 53BP1 nuclear accumulation in cells depleted of endogenous NUP153 (Figure 5b), indicating that the internal zinc-finger domain of NUP153 is not required for nuclear import of 53BP1. Instead, these experiments suggested that 53BP1 nuclear import could be mediated via the C-terminus of NUP153, whose FG repeats are known to interact with importin-β.^{11, 12}

Disruption of importin-β transport pathway blocks nuclear import of 53BP1

To test whether importin-β is involved in 53BP1 nuclear import, we took advantage of previous reports showing that overexpression of the N-terminal and C-terminal part of NUP153 inhibits transportin-1 and importin-β-mediated transport, respectively.²⁴ We expressed these fragments in U2OS cells and assayed their impact on 53BP1 localization. Although the N-terminal fragment of NUP153 did not affect 53BP1 localization, cells overexpressing the C-terminal fragment showed a predominantly cytoplasmic 53BP1 (Figure 6a). Consistent with published results, we observed interaction of the NUP153 C-terminal fragment with importin-β (Supplementary Figure S3A) and also an interaction of endogenous NUP153 with importin-β protein (Supplementary Figure S3B). Thus, overexpression of the C-terminal part of NUP153 appears to generate a dominant negative effect on nuclear import of 53BP1, indicating that accumulation of 53BP1 in the nucleus may rely on the importin-β pathway.

Next, we asked to which extent the nuclear import of 53BP1 is dependent on importin-β and transportin-1, respectively. siRNA-mediated knockdown of importin-β was relatively toxic, likely because importin-β is essential for nuclear import of a large number of cargoes;²⁵ nevertheless, its depletion resulted in nuclear exclusion of 53BP1 in a fraction of cells (Figure 6b) within the time frame of the experiment. In contrast, knockdown of transportin-1 had no effect on 53BP1 localization (Figure 6b). Nuclear import of proteins is also dependent on Ran GTPase.²⁶ Indeed, knockdown of Ran resulted in a detectable cytoplasmic 53BP1 in the majority of cells, analogous to knockdown of importin-β (Figure 6b). Efficiency of transportin-1 and importin-β knockdown, respectively, was confirmed by immunoblotting (Supplementary Figure S3C).

Finally, we focused on potential physical interactions among the key proteins implicated by the above results in the nuclear import of 53BP1. Indeed, we were able to co-immunoprecipitate endogenous 53BP1 with GFP-NUP153, but not with a control unrelated NUP GFP-NUP35 (Figure 6c), and, vice versa, GFP-NUP153, but not GFP-NUP35, with endogenous 53BP1 (Figure 6d), suggesting specificity of the observed interaction. Moreover, we found an interaction between endogenous 53BP1 and endogenous importin-β (Figure 6e). The latter interaction was detectable regardless of whether NUP153 was knocked down in the analyzed cells or not (Figure 6e), a result that is consistent with the current models of importin-β forming its complexes with cargo proteins (here 53BP1) in the cytosol, prior to the interaction of such importin-β/cargo complex with the NPC.²⁷ These results further support the role of the NUP153/importin-β pathway in nuclear import of 53BP1.

Screen for 53BP1 regulators

A custom-designed siRNA library (Ambion, Austin, TX, USA; silencer select siRNAs) comprising of 1346 targets involved in ubiquitin–proteasome system (see Supplementary Table 4 for used siRNA sequences), SUMO regulators and zinc-finger proteins were used as described.²⁰ Briefly, U2OS or HeLa cells were seeded on the siRNA arrays, incubated for 3 days, irradiated with two Gy, immunostained 1 h later with an antibody to 53BP1 and analyzed for 53BP1 nuclear foci. Each gene was targeted by two independent siRNAs and only the cases where both siRNAs showed the same phenotype were considered as hits.

Image acquisition, image analysis and Z-score calculation

Image acquisition of the siRNA arrays was performed on a wide-field fluorescence microscope (Zeiss Axiovert 200, Zeiss, Jena, Germany) equipped with a motorized-stage, × 20 air objective (PlanApochromat 20 × /0.8, Zeiss), CCD camera (Coolsnap HQ, Roper Scientific, Ottobrunn, Germany) operated by Metamorph software (version 6.2r6; Molecular Devices, Sunnyvale, CA, USA). The number of 53BP1 foci per nucleus was determined by automated routine using free software ImageJ (http://rsb.info.nih.gov/ij/). The nuclei were detected using the Hoechst signal, the average number of 53BP1 foci per nucleus was determined for each frame, and the values normalized to the average number of foci in nuclei exposed to the non-targeting controls within the given siRNA microarray. Impact of each siRNA on 53BP1 foci formation was determined according to its Z-score: Z=(X−μ)/σ (where X is the raw score, σ is the S.D. of the population and μ is the mean of the population).

Cell culture

Human U2OS and HeLa cell lines were cultured in Dulbecco's Modified Eagle's medium supplemented with 10% foetal bovine serum (Invitrogen, Carlsbad, CA, USA) and penicillin/streptomycin (Sigma-Aldrich, St.Louis, MO, USA) in a humidified atmosphere of 5% CO2 at 37 °C.

Plasmids and antibodies

The plasmid for GFP-NUP153 (pEGFP-NUP153) was described previously.¹⁷ To generate siRNA-insensitive GFP-NUP153^* mutant, silent mutations were introduced into the siNUP153 no. 5 target sequence in the NUP153 coding region using the QuikChange II Site-Directed Mutagenesis Kit (Stratagene/Agilent Technologies, Santa Clara, CA, USA). Plasmid transfections were performed using FuGENE 6 (Roche, Mannheim, Germany) according to the manufacturer's instructions. Antibodies used in this study included mouse monoclonal antibodies to γH2AX (Millipore, Billerica, MA, USA), BRCA1, GFP (Santa Cruz Biotechnology, Santa Cruz, CA, USA), MDC1 (Sigma-Aldrich), NBS1 (GeneTex, Irvine, CA, USA) and NUP153 (Abcam, Cambridge, UK); rabbit polyclonal antibodies to 53BP1, RAD51, TopBP1 (Santa Cruz Biotechnology), 53BP1 (Millipore), MDC1 (Abcam) and RNF168;²⁸ and goat polyclonal antibody to γ-tubulin (Santa Cruz Biotechnology). Note that some but not all antibodies to 53BP1 recognize a triplet of bands on western blots (see Figure 1e) that likely represent modified versions of the protein. Importantly, the cellular staining signal obtained with these antibodies selectively disappeared after siRNA-mediated knockdown of 53BP1 (data not shown), also supporting the notion that the three western blot bands represent modified 53BP1.

RNA interference

All transient siRNA transfections were performed using Lipofectamine RNAiMAX (Invitrogen) and oligonucleotides with siRNA duplexes purchased from Ambion. siRNA used: negative control siRNA no. 1 (s229084), siNUP153 no. 1,²² siNUP153 no. 4 (s19374), siNUP153 no. 5 (s19375), siNUP153 no. 6 (s19376), siRNF168 (ID no. 126171), siIMP-β (s7919), siTNPO1 (s7933) and siRan (ID no. 120462).

Immunofluorescence and immunoblotting

Cells cultured on glass coverslips were fixed in 4% formaldehyde for 15 min at RT, permeabilized 5 min with 0.2% (v/v) Triton X-100 in PBS and incubated with primary antibodies for 60 min at RT. Following washing step, the coverslips were incubated with goat anti-rabbit or goat anti-mouse Alexa Fluor 488 or Alexa Fluor 568 (Invitrogen) secondary antibodies for 60 min at RT and mounted using Vectashield mounting reagent with DAPI (Vector Laboratories, Burlingame, CA, USA). Image acquisition was done using a confocal microscope LSM510 (Zeiss) as described previously.^{10, 20} The analysis of cell lysates by gel electrophoresis under denaturing conditions, followed by immunoblotting was performed as described.⁴⁵ Briefly, whole cell extracts were separated by 6 or 10% SDS-PAGE and transferred to nitrocellulose membranes (GE Healthcare, Little Chalfont, UK). The membranes were blocked with 5% (w/v) dry milk in 0.1% (v/v) Tween-20 in PBS and probed with the primary antibodies listed above, followed by HRP-labeled secondary antibodies (Vector Laboratories and Santa Cruz Biotechnology) and visualized using ECL detection reagents (GE Healthcare).

Clonogenic survival assay

Sensitivity to IR was determined by transfecting U2OS cells with control and NUP153-targeting siRNA for 72 h followed by irradiation (2 and 4 Gy) and plating in increasing cell densities on a six-well plate. Colonies were allowed to grow for 10 days, fixed in 70% ethanol and stained with 1% crystal violet in ethanol. Colonies of >50 cells were counted, and the surviving fraction was calculated and normalized to untreated control.

Cellular fractionation

Cytoplasmic proteins were extracted from U2OS cells by incubating in buffer A (1 mM HEPES, 10 mM KCl, 1.5 mM MgCl₂, 0.5 mM DTT, 0.1% Triton X-100) at 4°C for 10 min and spinning down at 5000 × g for 2 min at 4°C. Nuclear fraction was obtained by extraction of remaining pellet in SDS sample buffer.

Immunoprecipitation

Cells were subjected to mild sonication in ice-cold extraction buffer (PBS pH 7.4, 1% Triton X-100, 2 mM EDTA, 1 mM EGTA, 0.5 mM DTT, 1 mM NaF, 25 mM β-glycerolphosphate, 1 mM Na₃VO₄, protease inhibitors). Cell extracts were clarified by 10 min centrifugation 20 000 × g at 4°C. Antibodies were bound to protein A/G resin (Thermo Fischer Scientific, Rockford, IL, USA) and incubated with normalized cell extracts at 4°C for 3 h. Immunoprecipitates were extensively washed and eluted by SDS sample buffer.