The nuclear basket of nuclear pore complexes (NPCs) is composed of three nucleoporins: Nup153, Nup50 and Tpr. Nup153 has a role in DNA double-strand break (DSB) repair by promoting nuclear import of 53BP1 (also known as TP53BP1), a mediator of the DNA damage response. Here, we provide evidence that loss of Nup153 compromises 53BP1 sumoylation, a prerequisite for efficient accumulation of 53BP1 at DSBs. Depletion of Nup153 resulted in reduced SUMO1 modification of 53BP1 and the displacement of the SUMO protease SENP1 from NPCs. Artificial tethering of SENP1 to NPCs restored non-homologous end joining (NHEJ) in the absence of Nup153 and re-established 53BP1 sumoylation. Furthermore, Nup50 and Tpr, the two other nuclear basket nucleoporins, also contribute to proper DSB repair, in a manner distinct from Nup153. Similar to the role of Nup153, Tpr is implicated in NHEJ and homologous recombination (HR), whereas loss of Nup50 only affects NHEJ. Despite the requirement of all three nucleoporins for accurate NHEJ, only Nup153 is needed for proper nuclear import of 53BP1 and SENP1-dependent sumoylation of 53BP1. Our data support the role of Nup153 as an important regulator of 53BP1 activity and efficient NHEJ.
Nuclear pore complexes (NPCs) are gates for all molecular trafficking between the cytoplasm and the nucleus of interphase eukaryotic cells. NPCs consist of multiple copies of ∼30 different nucleoporins (Nups) (Cronshaw et al., 2002; Ori et al., 2013; Rout et al., 2000; for reviews, see Dickmanns et al., 2015; Kabachinski and Schwartz, 2015) resulting in an overall complex of ∼110 MDa in vertebrates (Ori et al., 2013; Reichelt et al., 1990). On the nuclear side of the NPCs, eight filaments emanate to form the nuclear basket, which is primarily composed of the nucleoporins Nup153, Nup50 and Tpr (Beck et al., 2004, 2007; Duheron et al., 2014; Fahrenkrog and Aebi, 2003; Frenkiel-Krispin et al., 2010; Lim et al., 2008; Maimon et al., 2012; Ori et al., 2013). All three basket nucleoporins are important players in nucleocytoplasmic transport (Dickmanns et al., 2015) and are evolutionarily conserved, with functional homologues in yeast, Drosophila, Xenopus and other species (Bogerd et al., 1994; De Souza et al., 2009; Dilworth et al., 2001, 2005; Kosova et al., 2000; Strambio-de-Castillia et al., 1999).
Recent years have provided increasing evidence that nucleoporin function exceeds nucleocytoplasmic transport and that they directly or indirectly control aspects of DNA-based processes, such as DNA transcription, replication and repair (Gay and Foiani, 2015; Géli and Lisby, 2015; Ibarra and Hetzer, 2015). In this context, it has been shown that Nup153 mediates the nuclear import of the DNA repair mediator 53BP1 (also known as TP53BP1), regulating the non-homologous end joining (NHEJ) pathway of DNA double-strand break (DSB) repair (Cobb et al., 2016; Lemaitre et al., 2012; Moudry et al., 2011). Nup50 appears to be required for the nuclear import of XRCC1, which plays roles in single-strand break and base excision repair (Kirby et al., 2015). So far nothing is known about a putative role for Tpr in the DNA damage response (DDR), but yeast cells lacking Mlp1 and Mlp2, the yeast homologue of Tpr (Strambio-de-Castillia et al., 1999), accumulate DNA damage and are highly sensitive to damaging drugs (Hediger et al., 2002; Palancade et al., 2007; Zhao et al., 2004). The role of Mlp1 and Mlp2 in DDR is not well characterised, but appears to be linked to their anchor function of the yeast SUMO-protease Ulp1 (Palancade et al., 2007; Zhao et al., 2004). Small ubiquitin-like modifiers (SUMOs) are proteins that are covalently conjugated to a large set of target proteins, thereby altering their activity (Flotho and Melchior, 2013; Mukhopadhyay and Dasso, 2007). Ulp1 belongs to a family of cysteine proteases that specifically target SUMOs to reverse their action (Mukhopadhyay and Dasso, 2007; Yeh, 2009). The closest mammalian homologous of Ulp1 are the two SUMO-specific proteases 1 and 2 (SENP1 and SENP2). SENP1 and SENP2 are nuclear proteins that harbour nuclear localisation signals (NLSs) as well as nuclear export signals (NESs) (Bailey and O'Hare, 2004; Gong et al., 2000; Itahana et al., 2006; Kim et al., 2005; Yeh, 2009). They associate with NPCs and are concentrated at the nuclear basket, where they dynamically and primarily interact with Nup153 (Chow et al., 2012; Hang and Dasso, 2002; Zhang et al., 2002). No interaction of SENP2 with Tpr has been found (Zhang et al., 2002), but SENP2 expression was reduced in HeLa cells depleted for Tpr (David-Watine, 2011). Delocalisation of Ulp1 from NPCs causes DNA damage and cell cycle defects, phenotypes attributed to inappropriate desumoylation of nucleoplasmic target proteins that are normally protected from Ulp1 (Nie and Boddy, 2015; Palancade et al., 2007; Zhang et al., 2002). Similarly, it has been shown that a mutant form of SENP2 that is unable to interact with Nup153 more effectively promotes desumoylation (Hang and Dasso, 2002), indicating that NPCs in fact play important roles in spatially restricting the activity of SUMO protease.
Here, we provide evidence that displacement of SENP1 from NPCs in cells depleted for Nup153 affects sumoylation of 53BP1 and reduces the efficiency of DSB repair by NHEJ.
Depletion of nuclear basket nucleoporins alters the DDR
It has previously been shown that the depletion of Nup153 by use of small interfering RNAs (siRNAs) sensitised cells to DNA DSB-inducing agents (Lemaitre et al., 2012; Moudry et al., 2011). To address the question of whether or not Nup50 and Tpr, the two partners of Nup153 at the NPC nuclear basket, are also involved in DSB repair, we analysed the effect of their depletion for clonogenic survival of U2OS cells after exposure to genotoxic stress. As shown in Fig. 1A and Table 1, U2OS cells depleted for Nup153 and Tpr, respectively, exhibited a slight increased sensitivity to the radiomimetic drug neocarzinostatin (NCS) and a trend towards decreased viability as compared to Nup50-depleted cells and control cells transfected with scrambled non-targeting (NT) siRNAs. This decreased viability was statistically relevant for Nup153-depleted cells when treated with 2 µg/ml NCS (Fig. 1B). The efficiency of the silencing of the three nucleoporins was verified by immunofluorescence microscopy and western blot analyses (Fig. S1A,B). Next, we sought to directly test which pathway for DSB repair was affected by the respective depletion of Nup153 and Tpr. To this end, we utilised cell lines that were co-transfected with siRNAs and GFP reporter constructs to monitor the efficiency of NHEJ (Seluanov et al., 2004) and homologous recombination (HR) (Pierce et al., 1999), respectively. To monitor NHEJ efficiency, U2OS cells were depleted for the particular nucleoporin, transfected with a linearised pEGFP-Pem1-Ad2 plasmid and analysed by flow cytometry (see Materials and Methods). As shown in Fig. 1C, depletion of either of the three nuclear basket nucleoporins decreased NHEJ efficiency as compared to that seen in control cells (normalised to 1), similar to what is observed in 53BP1-depleted cells. To monitor HR, the recombination substrate DR-GFP was utilised, in which a gene conversion event results in the expression of an intact GFP protein, and assessed by flow cytometry (see Materials and Methods). Similar to what is seen upon depletion of Rad51 in U2OS cells, cells lacking Nup153 and Tpr, respectively, were less competent for HR as compared to Nup50-depleted and control cells (Fig. 1D). Taken together, these data indicate that in particular Nup153 and Tpr are implicated in DNA DSB repair.
Localisation of 53BP1 is only affected by depletion of Nup153
Previously it has been demonstrated that Nup153 is involved in the nuclear import of 53BP1 (Moudry et al., 2011). We therefore next asked whether the depletion of Nup50 and Tpr also affects 53BP1 nuclear import. By performing immunofluorescence analyses, we found, however, that only down-regulation of Nup153 reduced the nuclear accumulation of 53BP1 (Fig. 2A,B), while the overall expression levels of 53BP1 remained unchanged (Fig. 2D).
Having seen that cells lacking Nup153 or Tpr are also less competent at HR, we wondered whether BRCA1 nuclear localisation would be altered by the depletion of these two nucleoporins. In control cells, BRCA1 was mainly present in the nucleus, but a small proportion was also detected in the cytoplasm (Fig. 2A). This distribution of BRCA1 did not change upon the respective down-regulation of Nup153, Nup50 and Tpr in U2OS cells (Fig. 2A,C) as was also the case for BRCA1 expression levels (Fig. 2E). Taken together, our data indicate that general nuclear import defects are unlikely to be the exclusive cause for impaired DSB repair in cells lacking nuclear basket components.
Nup153 and Tpr are required for the dissolution of repair foci
We next analysed the behaviour of 53BP1 and BRCA1 in U2OS cells lacking Nup153, Nup50 and Tpr, respectively, after the induction of DSBs by NCS (50 ng/ml for 15 min). We monitored the formation of 53BP1 and BRCA1 foci at DSBs and their dissolution over time. As shown in Fig. 3A, the particular downregulation of the nuclear basket nucleoporins did not impair 53BP1 and BRCA1 foci formation after NCS treatment, although more 53BP1 was detectable in the cytoplasm of Nup153-depleted cells. In control cells, the number of BRCA1 (Fig. 3B) and 53BP1 (Fig. 3C) foci increased during the first 2 h after NCS treatment and then progressively decreased. The decrease in foci number was accompanied by an increase in the volume of the remaining foci (Fig. 3D,E). We revealed a similar evolution for BRCA1 foci in Nup50-depleted cells, whereas Nup153- and Tpr-depleted cells showed a reduced number of BRCA1 foci with no increment in their volume (Fig. 3B,D), potentially due to impaired recruitment of BRCA1 to the break sites. 53BP1 foci were formed within the first 2 h after DNA damage induction in control and nucleoporin-depleted cells (Fig. 3C), which, in contrast to in control cells, remained constant in number (Fig. 3B) and volume (Fig. 3E) in the depleted cells, which might be due to inefficient replenishment of 53BP1 to the damaged sites resulting in an inefficient dissolution of the 53BP1 repair foci.
Sumoylation of 53BP1 is impaired upon Nup153 depletion
Previous work revealed that efficient accumulation of 53BP1 at sites of DSBs requires the E3 SUMO ligase PIAS4 and 53BP1 modification by SUMO1 (Galanty et al., 2009). We hypothesised that sumoylation of 53BP1 might be hampered in the absence of the nuclear basket components. To address this hypothesis, we treated U2OS cells stably expressing GFP–SUMO1, GFP–SUMO2 and GFP–SUMO3, respectively (Galanty et al., 2009), with NCS and carried out colocalisation experiments analysing the particular SUMO forms and 53BP1. We confirmed that, upon DSB induction, 53BP1 colocalised with SUMO1, but not SUMO2 and SUMO3 (Fig. 4A). Similar results were obtained with NCS-treated HeLa cells stably expressing YFP–SUMO1 (Fig. S1C). We next transiently transfected the GFP–SUMO cell lines with an N-terminally HA-tagged 53BP1 (HA–53BP1). Western blot analysis revealed that, in U2OS cells treated with NCS, HA–53BP1 is sumoylated by GFP–SUMO1, but not by GFP–SUMO2 and GFP–SUMO3 (Fig. 4B). Co-expression of the SUMO-isopeptidase SENP1 with HA–53BP1 in the GFP–SUMO1-expressing U2OS cells prevented the appearance of the modified HA–53BP1 upon NCS-mediated DSB induction, confirming that 53BP1 is indeed modified by SUMO1 (Fig. 4C). Upon the respective depletion of Nup153, Nup50 and Tpr, we found that the association of GFP–SUMO1 with 53BP1 was reduced in U2OS cells depleted for Nup153, but increased in the absence of Tpr (Fig. 4D–G), indicating that altered sumoylation of 53BP1 might indeed account for the defects in NHEJ in these cells. Downregulation of Nup50, in contrast, had no effect on the association of GFP–SUMO1 with 53BP1 (Fig. 4D–G). The role of Tpr for protein sumoylation appears to be more general, whereas Nup153 appears to be specifically required for SUMO1 modification of 53BP1; cells depleted for Tpr exhibited a significant increase in global SUMO1 and SUMO2 modification of proteins, whereas global SUMO profiles remained unchanged in the respective absence of Nup153 and Nup50 (Fig. S2). Sumoylation takes place in a cascade of three enzymatic steps, of which the third step comprises the actual transfer of SUMO to its target protein by an E3-SUMO ligase. The E3 SUMO ligase for 53BP1 is PIAS4, and it is in fact required for 53BP1 recruitment to DSB sites (Galanty et al., 2009). We therefore tested whether PIAS4 localisation and/or expression is altered in the absence of Nup153, but observed no difference in PIAS4 nuclear localisation and expression levels upon depletion of Nup153, consistent with an apparently proper 53BP1 recruitment to DSBs (Fig. S3). Likewise, the nuclear import and expression levels of PIAS1, another E3 SUMO ligase known to be critical for proper DSB repair (Galanty et al., 2009; Ishiai et al., 2004; Zlatanou and Stewart, 2010), remained unaffected by Nup153 depletion (Fig. S3). Moreover, overexpressed mCherry-tagged versions of PIAS1, PIAS2, PIAS3 and PIAS4 localised to the nucleus in NCS-treated and Nup153-depleted cells in a manner that was indistinguishable from that seen in control cells (Fig. S4).
SENP1 is partially displaced in the respective absence of Nup153 and Tpr
Two SUMO proteases, SENP1 and SENP2, are known to localise to the nuclear periphery, and Nup153 is involved in their recruitment (Chow et al., 2012, 2014; Cubenas-Potts et al., 2013; Hang and Dasso, 2002; Zhang et al., 2002). Delocalisation of Ulp1, the yeast homologue of SENP1 and SENP2, from NPCs caused increased DNA damage sensitivity and cell cycle defects (Lewis et al., 2007; Palancade et al., 2007; Zhao et al., 2004). We considered it possible that 53BP1 sumoylation is impaired in Nup153-depleted cells due to a displacement of SENP1 and/or SENP2 from NPCs. We therefore analysed the effect of Nup153 depletion on the localisation of SENP1 and SENP2, respectively. Detection of SENP1 by immunofluorescence microscopy revealed that SENP1 was enriched at the nuclear envelope (NE) in control and Nup50-depleted cells, whereas it was reduced at the NE and partially displaced to the cytoplasm in Nup153- as well as Tpr-depleted cells (Fig. 5A,B). Consistent with the redistribution of SENP1 in cells lacking Nup153 or Tpr, the relative fluorescence intensity at the NE decreased by ∼40% in Nup153-depleted cells and 30% in Tpr-depleted cells (Fig. 5C). Western blot analysis of U2OS lysates revealed that SENP1 levels were lowered in cells lacking Nup153 (Fig. 5D; see also Fig. S5A for control of SENP1 depletion by siRNAs directed against SENP1), as previously described (Chow et al., 2014). In addition, in Tpr-depleted, but not in control or Nup50-depleted cells, relative levels of SENP1 were reduced (Fig. 5E). In contrast to SENP1, SENP2 expression remained unaffected by depletion of the nuclear basket nucleoporins (Fig. S5B; see also Fig. S5A for control of SENP2 depletion by siRNAs directed against SENP2). Owing to the fact that antibodies were largely inadequate for immunofluorescence, potential changes in SENP2 localisation could not be monitored (data not shown).
Tethering of SENP1 to the NPC protects cells against deficiencies in DNA damage response
In order to confirm that the loss of SENP1 from the NE contributes to the defects in DSB repair observed in cells lacking Nup153 or Tpr, we next tethered SENP1 artificially to the NPC. Specifically, we fused the N-terminal domain of Nup153, which contains its NPC-targeting domain (Ball and Ullman, 2005), to SENP1 (Nup153–SENP1; Fig. 6A). This fusion protein is resistant to siRNAs directed against the C-terminal region of Nup153 (Fig. 6B). Nup153–SENP1 localised to NPCs in transiently transfected U2OS cells treated beforehand with non-targeting siRNAs as well as siRNAs against Nup153 and Tpr (Fig. 6C).
To characterise the effect on DSB repair of artificially tethering SENP1 to NPCs, we next monitored the efficiency of NHEJ and HR using the reporter constructs outlined above. U2OS cells were first treated with the experimental siRNAs for 72 h, and subsequently co-transfected with Nup153–SENP1 and the respective reporter constructs for NHEJ and HR. After a further 24 h, NHEJ and HR efficiency were each analysed by flow cytometry. As shown in Fig. 6D, artificial tethering of SENP1 to NPCs could rescue NHEJ efficiency in cells lacking Nup153, but not in cells lacking Tpr (see also Fig. 1B). NHEJ efficiency of Nup153-depleted cells with tethered SENP1 was similar to that of cells treated with non-targeting siRNAs expressing only endogenous SENP1. Tethering SENP1 to NPCs in the control cells further enhanced the efficiency of NHEJ (defined as value 1). In contrast to NHEJ, artificial tethering of SENP1 to NPCs could only partially rescue HR deficiency in Nup153-depleted cells as compared to control cells expressing endogenous SENP1 or tethered SENP1, while again HR deficiency could not be rescued in Tpr-depleted cells (Figs 6E and 1C).
To confirm that the displacement of SENP1 and the defects in NHEJ in Nup153-depleted cells are in fact due to impaired sumoylation of 53BP1, we next tested whether artificial tethering of SENP1 to NPCs also rescues the reduced association of GFP–SUMO1 with 53BP1 as seen in Nup153-depleted cells (Fig. 4D), although it is somewhat counter-intuitive that a SUMO protease would promote SUMO modification of 53BP1. As shown in Fig. 6F, alongside the quantification in Fig. 6G, Nup153-depleted cells with tethered SENP1 showed an accumulation of GFP–SUMO1 at 53BP1 foci in U2OS cells treated with NCS that was comparable to that in control cells, in contrast to Nup153-depleted cells and SENP1-depleted cells. SENP1 depletion, in contrast to Nup153 depletion, did not affect nuclear import of 53BP1 (and BRCA1, Fig. S5C,D). As seen for Nup153, depletion of SENP1 does not compromise the localisation and expression of the two E3 SUMO ligases PIAS1 and PIAS4 (Fig. S3) nor the localisation of over-expressed mCherry-tagged versions of PIAS1–PIAS4 (Fig. S4). Taken together, our data indicate that defects in NHEJ, but not in HR, in Nup153-depleted cells are, at least in part, due to displacement of SENP1 from NPCs. Tpr acts in a manner that is unrelated to SENP1 localisation at NPCs, suggesting that Nup153 and Tpr have different and independent roles in DNA DSB repair.
Silencing Nup153 compromises SUMO1 modification of 53BP1
The results of this study broaden the spectrum of the function of Nup153 in the maintenance of genome integrity. Previous studies have shown that Nup153 is important for the nuclear import of 53BP1 and efficient NHEJ (Lemaitre et al., 2012; Moudry et al., 2011). By carrying out a more systematic analysis of the function of Nup153 in the DDR, we demonstrate here that Nup153 is similarly required for the sumoylation of 53BP1. We observed no impaired recruitment of 53BP1 to DSBs, but decreased growth of repair foci and their delayed dissolution (Fig. 3). Efficient accumulation of 53BP1 at repair sites requires its modification by SUMO1 (Galanty et al., 2009), and SUMO pathway components are often associated with nuclear pores (for reviews, see Palancade and Doye, 2008; Rodriguez, 2014). While 53BP1 and SUMO1 largely colocalised at repair foci in control cells, this was diminished in Nup153-depleted U2OS cells, but not in Tpr- or Nup50-depleted cells (Fig. 4). This reduced sumoylation of 53BP1 was not provoked by impaired nuclear import of PIAS4, the E3 SUMO ligase promoting SUMO1 modification of 53BP1 (Galanty et al., 2009), but rather by a displacement of the SUMO protease SENP1 from NPCs in the absence of Nup153 (Fig. 5; Fig. S3). Artificial tethering of SENP1 to NPCs restored NHEJ efficiency and 53BP1 sumoylation after silencing of Nup153 (Fig. 6). It is somewhat counter-intuitive that a SUMO protease promotes SUMO modification of 53BP1, but similar observations have been made in fission yeast, where the deletion of Nup132, the Schizosaccharomyces pombe homologue of human Nup133, caused delocalisation of Ulp1 from NPCs, leading to DNA damage and cell cycle defects (Nie and Boddy, 2015). These defects were thought to arise from inappropriate desumoylation of nucleoplasmic targets that are normally spatially protected from Ulp1. Delocalised Ulp1 allowed an accumulation of SUMO chains on Pli1, a PIAS family SUMO E3 ligase and the S. pombe homologue of PIAS1, and its targeting for proteosomal degradation, which resulted in profound SUMO pathway and cell cycle defects (Nie and Boddy, 2015). It will be interesting to see whether a similar crosstalk between sumoylation and ubiquitylation of a PIAS family or another SUMO E3 ligase contributes to the defects in DDR seen in Nup153-depleted cells. PIAS1 and PIAS4, the two most obvious candidates in the context of 53BP1 and DSB repair, appear, however, unaffected in cells lacking Nup153 (Figs S3 and S4); we observed no changes in their intracellular localisation or stability. Nevertheless, the notion that the balance between sumoylation and desumoylation of SUMO E3 ligases contributes to effective DSB repair is further supported by our observation that depletion of Tpr led to defects in NHEJ, to displacement of SENP1 from NPCs and to an increase in 53BP1 sumoylation (Figs 1–4). This increase in 53BP1 sumoylation in the absence of Tpr likely arises because of the impact of Tpr on the level and function of SENP2 (David-Watine, 2011; Xu et al., 2007). Which other components of the SUMO pathway, and potentially the ubiquitin pathway, are out of balance in Nup153- and Tpr-depleted cells is an interesting question to address in the future. It will be furthermore interesting to see whether our observations made here hold true in non-immortalised cells and for endogenous 53BP1. However, endogenous sumoylation of proteins is challenging to detect due to the dynamic nature of the modification and its low steady-state levels (see, for example, Becker et al., 2013).
Nup153 and HR-mediated repair
We observed here that silencing Nup153 sensitised U2OS cells to treatment with the radiomimetic drug NCS, which coincided with a decrease in cell viability as well as in NHEJ and HR (Fig. 1). In contrast to a previous study (Lemaitre et al., 2012), we noticed no shift from NHEJ to HR in Nup153-depleted cells. This discrepancy might arise from the different cell lines that were used; U2OS cells in our study and the human fibroblast line GC92 in the previous study (Lemaitre et al., 2012; Mamouni et al., 2014; Rass et al., 2009). Our data are, however, consistent with the fact that 53BP1 promotes the fidelity of homology-directed DNA repair (Ochs et al., 2016). Silencing or exhausting 53BP1 results in a switch from the usage of RAD51 as recombinase in HR to usage of RAD52, which translates into a shift from error-free repair to the mutagenic single-strand annealing pathway (Ochs et al., 2016). One prerequisite for the loading of RAD51 on damaged DNA is, among others, the recruitment of BCRA1, a key factor in HR (Daley and Sung, 2014). We observed a reduced number of BRCA1 foci with no increase in their volume after silencing of Nup153 (Fig. 3A), which would be consistent with defects in RAD51 loading to damage sites. For future studies, it will therefore be interesting to deepen the analysis of the impact of Nup153 on RAD51-dependent HR.
Other nuclear basket nucleoporins and DDR
Our study further revealed that the two other components of the NPC nuclear basket, Tpr and Nup50, both somewhat contribute to effective DSB repair, but that their respective depletion did not inhibit the nuclear import of 53BP1 (Fig. 3A). We found that rather than a decrease, there was in fact an increase in overall level of sumoylation in Tpr-depleted cells, in contrast to a previous study (David-Watine, 2011), which might be attributed to the different cell lines used (HeLa versus U2OS). Tpr-depleted cells also showed an increase in sumoylation of 53BP1 and a partial displacement of SENP1 from NPCs, but defects in NHEJ in the absence of Tpr could not be restored by artificially tethering SENP1 to NPCs (Fig. 6). The role of Tpr and Nup50 hence needs to be further elucidated, but might be linked to repair at telomeres, as seen in yeast; Nup60 in yeast, the homologue of Nup50, is involved in anchoring Ulp1 to NPCs, similar to Mlp1 and Mlp2, the two yeast homologues of Tpr (Palancade et al., 2007). There, the displacement of Ulp1 from NPCs affected sumoylation of yKu70, a key player in NHEJ. In mlp1/mlp2 mutants, it was shown that yKu70 function is impaired in DNA repair at telomeres (Hediger et al., 2002). It will therefore be interesting to see whether Tpr and Nup50 are also involved in NHEJ at telomeres in human cells. Tpr is furthermore phosphorylated upon DNA damage by the ATM/ATR kinases and involved in the proper activation of G2/M and intra-S check point (Matsuoka et al., 2007). If and how this might be linked to our observations here remains a subject for future studies.
MATERIALS AND METHODS
All experiments were carried out at room temperature (RT) unless otherwise stated.
Cell lines and transfections
All experiments were conducted in U2OS cells (a kind gift from Dr Harald Herrmann, German Cancer Research Center, Heidelberg, Germany; tested for contamination, but not authenticated) grown in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (FBS) plus penicillin and streptomycin. Cells were transfected with plasmids by using the Turbofect transfection reagent (Thermo Scientific Fermentas, St. Leon-Rot, Germany) and with siRNAs by using Lipofectamine RNAiMAX (Life Technologies Invitrogen, Ghent, Belgium) following the instructions of the manufacturer. Smart-pool small interfering RNAs were obtained from Dharmacon (GE Healthcare Europe, Diegem, Belgium) as follows: Nup50 (L-012369-01), Nup153 (L-005283-00), Tpr (L-010548-00), 53BP1 (L-003548-00), Rad51 (L-003530-00), SENP1 (L-006357-00-0005), SENP2 (L-006033-00-0005) and non-targeting siRNAs (D-001810-10).
Plasmids and stable cell lines
U2OS cell lines (not tested and not authenticated) expressing SUMO1–GFP, SUMO2–GFP and SUMO3-GFP were a gift from Dr Steve Jackson (University of Cambridge, Cambridge, UK); HeLa cells expressing YFP–SUMO1 were a gift from Dr Angus Lamond (University of Dundee, Dundee, UK), pDRGFP and pCBASCeI were from Addgene [Addgene plasmids 26475 (Pierce et al., 1999) and 26477 (Richardson et al., 1998), deposited by Maria Jasin], pEGFP-Pem1-Ad2 and pDsRed2-N1 were kindly provided by Dr Vera Gorbunova (University of Rochester, NY), pEGFP-SENP1 by Dr Mary Dasso (NIH, Bethesda, USA), pmCherry-PIAS constructs were received from Dr Simon E. Fisher (Radboud University, Nijmegen, The Netherlands) and pCMH6K53BP1 (HA-53BP1) was a gift from Dr Aidan Doherty (University of Sussex, Brighton, UK).
For the generation of pcDNA3.1-FLAG-Nup153-HA-SENP1, SENP1 was PCR amplified and subcloned from pEGFP-SENP1 into KpnI- and NotI-cut pcDNA3.1-myc-His (reading frame B; Invitrogen, Carlsbad, CA), thereby integrating a HA tag, with the forward primer (5′-GGGGTACCATGTACCCATACGATGTTCCAGATTACGCTGATGATATTGCTGATAGG-3′) and a STOP codon with the reverse primer (5′-ATAAGAATGCGGCCGCTCACAAGAGTTTTCGGTG-3′) to result in pcDNA3.1-HA-SENP1. Nup153 nucleotides 1–1968 (amino acids 1–656) were amplified by a PCR and subcloned from pEGFP-Nup153 (Fahrenkrog et al., 2002) into HindIII- and KpnI-cut pcDNA3.1-HA-SENP1. A FLAG tag was thereby integrated into the forward primer (5′-CCCAAGCTTATGGACTACAAAGACGATGACGACAAGGCCTCGGGAGCCGGA-3′), the reverse primer was (5′-GGGGTACCTAAACTCTCCCCAAACCC-3′). In-frame insertion was confirmed by DNA sequencing.
The following antibodies were used: polyclonal rabbit anti-53BP1 [Santa Cruz Biotechnology sc-22760; lot I-1813; immunofluorescence (IF) 1:500; western blotting (WB) 1:1000], rabbit anti-α-tubulin (Abcam ab18251; lot GR52624-1; WB 1:2000), rabbit ant-importin-β (kind gift of Ralph Kehlenbach, University of Göttingen, Germany; WB 1:1000), rabbit anti-SENP1 (Abcam ab108981; lot YG102603C; IF 1:1500; WB 1:1000), rabbit anti-Nup50 (Abgent AP1913b; lot SH05120G; IF 1:1000; WB 1:250), rabbit anti-SENP2 (kind gift of Michael Matunis, Johns Hopkins Bloomberg School of Public Health, MD; IF 1:50; WB 1:500), rabbit anti-SENP2 (Santa Cruz Biotechnology, sc-67075; IF: 1:50; lot H-1813), as well as monoclonal mouse anti-BRCA1 (Santa Cruz Biotechnology sc-6954; lot D1713; IF 1:1000; WB 1:1000), mouse anti-HA (hybridoma supernatant; WB 1:50), mouse anti-Nup153 (SA1; hybridoma supernatant; IF 1:800; WB 1:200), mouse monoclonal anti-Tpr (Abnova H00007175-M01; lot 10131-1A8; IF 1:6000; WB 1:2000), mouse anti-FLAG (Sigma-Aldrich F-3165; lot 087K6002V; IF: 1:200), and rat anti-GFP (Chromotek, clone 3H9; lot 130624; WB 1:1000).
Clonogenic survival assay
U2OS were treated with siRNAs for 72 h, trypsinised, stained with Trypan Blue solution according to the manufacturer's instructions (Sigma-Aldrich, Diegem, Belgium) and counted using a haemocytometer. 1000 cells were seeded per well of a 6-well plate and grown for 12–14 h. Next, cells were treated with neocarzinostatin (NCS; Sigma-Aldrich, Diegem, Belgium) with doses ranging from 0 through 10 µg/ml for 15 min, washed extensively in PBS and released into fresh medium. Colonies were allowed to grow for 10 days, the wells were washed in PBS, fixed in 4% paraformaldehyde for 15 min, washed in PBS, stained with 0.5% Crystal Violet solution and colonies were counted manually. The surviving fractions were calculated and normalised to untreated control.
NHEJ fidelity was determined according to Seluanov et al. (2004). In brief, the pEGFP-Pem1-Ad2 reporter plasmid was linearised with I-SceI (New England Biolabs) and purified using a Qiagen gel extraction kit; complete digestion was confirmed by gel electrophoresis. U2OS cells were treated with siRNAs against the nuclear basket nucleoporins Nup153, Nup50 and Tpr, respectively, as well as scrambled siRNAs, for 72 h and next co-transfected with the linearised pEGFP-Pem1-Ad2 (to monitor NHEJ) and pDsRed2-N1 (to monitor transfection efficiency) (Seluanov et al., 2004). The cells were harvested 24 h after transfection, resuspended in 0.5 ml of PBS, and assayed for the expression of EGFP and DsRed by flow cytometry (FACS Calibur, BD Biosciences) according to the guidelines in Seluanov et al. (2004).
HR efficiency was determined using U2OS cells stably expressing the integrated HR reporter DR-GFP (see above). Cells were transfected with siRNAs against the nuclear basket nucleoporins Nup153, Nup50, and Tpr, respectively, as well as scrambled siRNAs, for 40 h and next co-transfected with pCBA-I-SceI (to monitor HR) and pDsRed2-N1 (to monitor transfection efficiency). The cells were harvested at 48 h after plasmid transfection, resuspended in 0.5 ml of PBS, and assayed for the expression of EGFP and DsRed by flow cytometry according to Pierce et al. (1999).
Cells were grown on coverslips, fixed in 4% paraformaldehyde for 15 min and permeabilised in Tris-buffered saline (TBS) containing 0.3% Triton X-100 for 5 min. After blocking for 60 min in TBS containing 5% BSA, cells were incubated with primary antibodies overnight at 4°C. Following washing steps, cells were incubated with fluorochrome-coupled secondary antibody for 45 min at RT and mounted in Mowiol 4-88 (Sigma) containing 1 µg/ml DAPI. Images were acquired with a Zeiss LSM 710 laser-scanning confocal microscope and the Zen system software. Images were processed using ImageJ and Adobe Photoshop.
Image J quantification
Quantification of western blots and immunofluorescence images were carried out following the same principles. In brief, areas of interest and background areas were delimited and their ‘mean grey values’ were measured using ImageJ. Values of background areas were subtracted from values of areas of interest and respective ratios were calculated.
Volumes and numbers of 53BP1 and BRCA1 foci were determined on confocal Z-stack pictures, using the ‘3D object counter’ ImageJ plugin (https://imagej.nih.gov/ij/plugins/track/objects.html) according to the developer's guidelines (http://imagejdocu.tudor.lu/lib/exe/fetch.php?media=plugin:analysis:3d_object_counter:3d-oc.pdf; Bolte and Cordelieres, 2006).
U2OS cells were lysed in lysis buffer [50 mM Tris-HCl, pH 7.8, 150 mM NaCl, 1% Nonidet-P40 and protease inhibitor cocktail tablets (Roche, Basel Switzerland)]. In case of the GFP–SUMO-expressing U2OS lines, the lysis buffer additionally contained 20 mM N-ethylmaleimide (Sigma-Aldrich) to inhibit de-sumoylation. 20 µg of protein were loaded and separated by SDS-PAGE (5% or 7% gels). The proteins were transferred onto a PVDF membrane (Immobilon-P, Millipore) and the membranes were blocked with TBS containing 0.1% Tween 20 and 5% non-fat dry milk for 1 h. The membranes were then incubated for 1 h in blocking solution containing a primary antibody followed by washing 3× in TBS containing 0.1% Tween 20 and 5% non-fat dry milk. The membranes were next incubated with alkaline phosphatase-labelled secondary antibodies for 1 h, washed 3× in TBS and developed using Lightning CDP Star Chemiluminescence reagent (Applied Biosystem) and an X-ray film. Films were scanned and processed using ImageJ.
The authors thank Drs. Mary Dasso (NIH, Bethesda, USA), Simon E. Fisher (Radboud University, Nijmegen, The Netherlands), Harald Herrmann (German Cancer Research Center, Heidelberg), Maria Jasin (Memorial Sloan Kettering Cancer Center, New York, USA), Ralph Kehlenbach (University of Göttingen, Germany), Angus Lamond (University of Dundee, Scotland, UK), and Mike Matunis (Johns Hopkins University, Baltimore, USA) for sharing reagents. Ralph Kehlenbach is further acknowledged for critically reading the manuscript.
Conceptualization: V.D., N.N., B.F.; Validation: V.D., N.N., S.P., V.M., B.F.; Formal analysis: V.D., N.N., B.F.; Investigation: V.D., N.N., S.P., V.M.; Writing - original draft: V.D., B.F.; Writing - review & editing: V.D., N.N., B.F.; Supervision: V.D., B.F.; Project administration: V.D., B.F.; Funding acquisition: B.F.
This work was supported by grants from the Fonds De La Recherche Scientifique - FNRS, Belgium (grant numbers 1.5091. 12 and T.0237.13, Télévie grants 7.4519.12F and 7.6534.16F), the Fédération Wallonie-Bruxelles (ARC 4.110.F.000092F), the Fonds Brachet, the Fonds Alice et David van Buuren, and by the Université Libre de Bruxelles, Belgium.
The authors declare no competing or financial interests.