Chromosome breakage elicits transient silencing of ribosomal RNA synthesis, but the mechanisms involved remained elusive. Here we discover an in trans signalling mechanism that triggers pan-nuclear silencing of rRNA transcription in response to DNA damage. This is associated with transient recruitment of the Nijmegen breakage syndrome protein 1 (NBS1), a central regulator of DNA damage responses, into the nucleoli. We further identify TCOF1 (also known as Treacle), a nucleolar factor implicated in ribosome biogenesis and mutated in Treacher Collins syndrome, as an interaction partner of NBS1, and demonstrate that NBS1 translocation and accumulation in the nucleoli is Treacle dependent. Finally, we provide evidence that Treacle-mediated NBS1 recruitment into the nucleoli regulates rRNA silencing in trans in the presence of distant chromosome breaks.
The inability to properly respond to DNA damage leads to various disorders and enhanced rates of tumour development in mammals. To avoid genome instability, cells respond to DNA damage by activating a complex network of signalling pathways, collectively termed the DNA damage response (DDR), which is designed for detecting, signalling and repairing of aberrant DNA structures1.
A central player in the DDR is the ATM protein, a large serine/threonine kinase that is activated by DNA double-strand breaks (DSBs) and is mutated in the genome instability syndrome Ataxia telangiectasia2. Several proteins are implicated in efficient recruitment of ATM to sites of DSBs and its subsequent activation: NBS1, for example, which is a subunit of the conserved MRE11–RAD50–NBS1 (MRN) complex, directly interacts with ATM, and the MRN complex is involved in the efficient activation of ATM in response to DSBs (ref. 3). NBS1 is mutated in Nijmegen breakage syndrome (NBS), a rare autosomal recessive congenital disorder that is associated with short stature, cranio-facial abnormalities including microcephaly, mild mental retardation, immunodeficiency and a strong predisposition to lymphoid malignancy4.
Other proteins such as the phosphorylated form of the histone variant H2AX (termed γH2AX) and the adaptor MDC1 help to establish a specialized chromatin environment in the regions flanking DSBs, a process believed to enhance and maintain the ATM-dependent DNA damage signalling pathways5. Recruitment and accumulation of the MRN complex at sites of DSBs requires the direct interaction between a unique divalent Forkhead-associated (FHA)-tandem BRCA1 carboxy-terminal (BRCT) phosphopeptide-binding architecture located at the amino terminus of NBS1, with a constitutively phosphorylated acidic repeat region in MDC1 (SDT repeat region; refs 6, 7, 8, 9, 10, 11).
Induction of DSBs triggers transient inhibition of cell cycle progression (so-called cell cycle checkpoints), especially during S-phase and at the G2/M border. Whether the transcriptional machinery is similarly affected by the presence of DSBs remains a long-standing question. Recent evidence suggests that ATM-dependent signalling cascades can also suppress transcription. For example, an ATM-dependent transcriptional silencing mechanism prevented RNA polymerase II (Pol II) elongation-dependent chromatin decondensation at regions distal to DSBs (ref. 12). In addition, ATM-dependent signalling was shown to inhibit local RNA polymerase I (Pol I) transcription in the nucleoli in response to chromosome breaks13. In both cases, transcriptional silencing seemed to be limited to chromatin regions that are closely located to the sites of breaks and a pan-nuclear transcriptional silencing mechanism in response to DNA damage has so far not been described.
Here we describe an ATM-dependent in trans signalling mechanism that triggers pan-nuclear silencing of Pol I transcription in response to DNA breaks. Furthermore, we present evidence that in trans inhibition of rRNA synthesis is regulated by rapid and transient DNA damage-induced nucleolar translocation and accumulation of NBS1, a process that is regulated by ATM and is dependent on direct interaction between NBS1 and the nucleolar factor Treacle. In summary, our findings reveal an unexpected trans-compartmental communication between sites of DNA damage and Pol I-mediated transcription in the nucleoli.
In trans inhibition of Pol I transcription in response to DNA damage is associated with NBS1 recruitment into the nucleoli
It was previously shown that laser micro-irradiation of an individual nucleolus leads to transient silencing of rRNA synthesis within this nucleolus, whereas transcription in the other nucleoli of the same cell was unaffected13.
To test whether induction of DSBs outside the nucleolar compartment had the capacity to trigger rRNA silencing in trans we established an imaging-based assay that combined laser micro-irradiation with 5-EU RNA labelling followed by Click-iT chemistry. Surprisingly, we found that irradiation of any sub-nuclear region had a pronounced effect on rRNA transcription in all of the nucleoli of the irradiated cells irrespective of whether or not they were directly hit by the laser beam (Fig. 1a and Supplementary Fig. 1a). This suggests the existence of an as yet unidentified in trans signalling mechanism that transduces the signal generated by the presence of chromosome breaks into the nucleoli. Further support for such an in trans signalling mechanism came from the observation that low doses of ionizing radiation gave a very similar degree of rRNA inhibition as compared to high-dose treatment (Supplementary Fig. 2a).
As Pol I silencing in cis after exposure of cells to ionizing radiation was shown to be dependent on ATM signalling and on the presence of the two DDR mediator–adaptor proteins MDC1 and NBS1 (ref. 13), we investigated whether these factors were also involved in this trans-compartmental signalling mechanism. In our hands MDC1 depletion had only a minor effect but we found that rRNA repression was dependent on ATM signalling and on NBS1, consistent with previous findings13 (Supplementary Fig. 3).
In support of a functional implication of NBS1 in rRNA silencing in response to DNA damage, we observed a focal pattern of GFP-tagged NBS1 in the nucleus that increased after laser micro-irradiation (Fig. 1b). These distinct NBS1–GFP foci co-localized with nucleolar Fibrillarin, suggesting bona fide nucleolar relocalization (Supplementary Fig. 1b). Interestingly, GFP-tagged MRE11 did not show this inter-compartmental dynamics in response to DNA damage (Supplementary Fig. 1d). We observed the same focal pattern of endogenous NBS1 in the nucleoli 10 min after 5 Gy of ionizing radiation, although the endogenous nucleolar NBS1 foci were smaller than the NBS1–GFP-decorated structures. Again, no signal associated with endogenous MRE11 could be detected, thus suggesting an MRE11-independent function for NBS1 (Fig. 1c). Consistently, we also found that rRNA silencing in response to DNA damage was NBS1 dependent but MRE11 independent (Supplementary Fig. 2b).
We then combined laser micro-irradiation with real-time imaging (reviewed in ref. 14) and observed a rapid accumulation of NBS1 around the DSB-containing subnuclear tracks and with a small temporal delay, the focal NBS1–GFP pattern appearing in the nucleoli (Supplementary Fig. 1c). Whereas NBS1–GFP stably associated at sites of laser damage for the entire duration of the experiment (70 min; see Fig. 1d) the nucleolar NBS1–GFP signal reached a maximum at around 15 min post irradiation and gradually decreased, reaching background levels at around 40 min. The kinetics of NBS1 recruitment into the nucleoli perfectly matched the kinetics of rRNA silencing in response to DNA damage (compare Fig. 1d, e), thus suggesting that these two events are functionally linked. In summary, these data reveal that DNA damage-induced in trans silencing of Pol I transcription is associated with relocalization of a fraction of NBS1 into the nucleoli.
NBS1 is enriched at sites of rRNA transcription in response to DNA damage
Immunofluorescence analysis in laser micro-irradiated cells showed that NBS1–GFP nucleolar foci co-localized with Pol I in the dense fibrillar components of the nucleoli where active transcription of rDNA takes place (Fig. 2a). To directly show NBS1 enrichment within transcribed regions of the nucleoli we used chromatin immunoprecipitation (ChIP) followed by quantitative real-time PCR in NBS1–GFP expressing cell lines. The primer pairs for the quantitative real-time PCR included a promoter region 5′ of the transcription start site (H0), a segment inside the coding region of the primary transcript (H1) and a region within the intergenic spacer (H18) of rDNA. A primer pair from the GAPDH gene served as a negative control (Fig. 2b). ChIP data showed that, as for Pol I, NBS1–GFP is present at the rDNA promoter (H0) and more abundant within the coding region (H1; Fig. 2c, d). Moreover, Pol I localization at the transcriptional start site and within the coding region was not markedly affected after ionizing radiation treatment (Fig. 2c), whereas NBS1–GFP association with transcribed regions of the rDNA increased after ionizing radiation treatment, thus suggesting a DNA damage-induced recruitment of NBS1 into actively transcribed regions within the nucleoli (Fig. 2d).
The NBS1 FHA domain mediates nucleolar accumulation of NBS1
Consistent with Pol I silencing in response to DNA damage being ATM dependent13 (Supplementary Fig. 3a), we observed a significant reduction of the percentage of laser micro-irradiated cells with nucleolar NBS1 on depletion of ATM by RNA interference (Fig. 3a). NBS1 itself is an ATM target but a mutant derivative of NBS1 lacking the three ATM phosphorylation sites (3A mutant) was still efficiently forming nucleolar foci on laser micro-irradiation, indicating that NBS1 phosphorylation by ATM is not required for its nucleolar translocation (Fig. 3b).
NBS1 recruitment into γH2AX-marked chromatin regions at sites of DSBs is mediated by MDC1. However, MDC1 depletion by short interfering RNA (siRNA) did not provoke a reduction in nucleolar NBS1–GFP accumulation, even though NBS1 recruitment to sites of DSBs was significantly reduced under these conditions (Fig. 3c). Interestingly, though, a mutation in the N-terminal FHA-BRCT region of NBS1 that abrogates its ability to interact with phosphorylated proteins (R28A ref. 10) led to a complete loss of nucleolar residence and DNA damage-induced nucleolar enrichment of NBS1–GFP (Fig. 3d). These data strongly suggest that an as yet unidentified phosphorylation-dependent interaction partner of NBS1 that is distinct from MDC1 mediates its nucleolar recruitment in response to DNA damage.
Treacle is a NBS1 FHA-BRCT-interacting protein
To identify NBS1 FHA-BRCT-interacting proteins that may mediate trans-compartmental enrichment of NBS1 in the nucleoli we performed an unbiased proteomic screen (see Methods for details). The first N-terminal 382 amino acids of human NBS1 were used as bait and were C-terminally modified with two affinity tags (Strep, HA; see Fig. 4a). One construct was wild type, the other carried the R28A point mutation in the FHA domain, which completely abrogated nucleolar residence of NBS1 (see above). Protein complexes were isolated from untreated cells and from cells treated with 10 Gy of ionizing radiation by double-affinity purification. Retained proteins were then identified by mass spectrometry.
As expected, unique NBS1 bait peptides were identified in all four purifications (Fig. 4b). MDC1 was co-purified only with wild-type NBS1 FHA-BRCT and the interaction was not altered on ionizing radiation treatment in agreement with published results6,7,15,16. Interestingly, several nucleolar proteins co-purified with the NBS1 FHA-BRCT region. Most pronounced according to the number of unique peptides was TCOF1 (Fig. 4b). The TCOF1 gene encodes a low-complexity nucleolar phosphoprotein often referred to as Treacle. TCOF1 is mutated in Treacher Collins syndrome (TCS), one of the most severe autosomal dominant congenital disorders of craniofacial development. Functionally, TCOF1-Treacle has been implicated in the spatiotemporal control of ribosome biosynthesis, most notably in rRNA transcription and pre-ribosomal RNA processing (reviewed in ref. 17).
We first confirmed the interaction between NBS1 and Treacle biochemically by GST pulldown experiments with bacterially purified NBS1 N-terminal fragments. Wild-type NBS1–GST (amino acids 1–382) but not an NBS1 FHA-BRCT double mutant (DM: R28A, K160M) pulled down detectable amounts of Treacle from the extract, indicating that interaction of the NBS1 FHA-BRCT region with Treacle requires intact FHA and BRCT phosphopeptide-binding functions (Fig. 4c).
Co-immunoprecipitation experiments with overexpressed tagged proteins showed that full-length Strep-tagged Treacle could retrieve endogenous NBS1 from cell extracts (Fig. 4d). Moreover, a wild-type Flag–HA-tagged N-terminal NBS1 fragment efficiently co-immunoprecipitated with endogenous Treacle, whereas the FHA mutated fragment (R28A) and the double-mutant fragment (DM: R28A, K160M) did not. The BRCT mutated fragment (K160M) showed residual binding activity, albeit significantly weaker than the wild type (Fig. 4e). This indicates that the NBS1 FHA domain is the predominant interaction partner of Treacle in vitro and that the NBS1 BRCT domains also partially contribute to the interaction.
Endogenous Treacle also co-immunoprecipitated with endogenous NBS1 (Fig. 4f) and, as already shown in the mass spectrometry analysis, Treacle association with NBS1 was not dependent on DNA damage (Fig. 4e, f).
NBS1 binds to conserved phosphorylated SDT-like motifs in Treacle
Treacle consists of three structurally distinct regions: an N-terminal region, a central region consisting of 10 consecutive acidic and serine-rich sequence stretches and a C-terminal region containing nuclear and nucleolar localization signals18 (Fig. 5a). Treacle interacts with CK2 (ref. 19) and its primary sequence contains numerous potential CK2 phosphorylation sites. We found that bacterially purified Treacle fragments are heavily phosphorylated by CK2 in vitro, especially within the repeat region (Supplementary Fig. 4). To map the NBS1 interaction site we systematically deleted the individual domains of Treacle to assess their impact on the interaction with NBS1 (Fig. 5b). These experiments suggested that the main NBS1 interaction site is located within the N-terminal 225 amino acids of Treacle with a potential minor contribution of the C-terminal region.
A purified GST fragment of this region was efficiently phosphorylated by CK2 in vitro (Supplementary Fig. 4; T-2) and this fragment pulled down the MRN complex from HeLa nuclear extracts only when previously phosphorylated by CK2 (Fig. 5c). Notably, this N-terminal region of Treacle features three motifs that are highly reminiscent of the previously described MDC1 SDT motifs6,7,8,9 and form CK2 consensus sites (Fig. 5d). The first motif consists of a serine residue followed by a glutamic acid and threonine, flanked by three consecutive glutamic acid residues (SETE amino acids 171–174). Further downstream, two additional sequence stretches show similarity to the MDC1 SDT motifs: a serine residue followed by two acidic amino acids and by a threonine residue (SEDT, amino acid 200–203 and SDET, amino acid 207–210, respectively).
To fine-map the NBS1 interaction site in Treacle we assessed the binding affinity of phosphopeptides derived from this region by isothermal titration calorimetry (ITC; Fig. 5e). The SDET motif around Thr 210 showed the highest binding affinity with a Kd around 1 μM.
Phosphorylated Thr 210 in Treacle is the main interaction site for NBS1
We next mutated Thr 210 to alanine in the context of full-length HA-tagged Treacle and analysed its binding to Flag-tagged N-terminal NBS1 in 293T cell extracts by Flag immunoprecipitation (Fig. 6a). The T210A mutant was completely deficient for NBS1 interaction and we did not observe a further decrease by additional serine/threonine substitutions. Mutation of the SETE motif led to a significant reduction in NBS1 binding but did not abrogate it (Fig. 6b), thus suggesting that sequence around Thr 210 is the predominant NBS1-interaction site in Treacle, with a minor contribution of the region around Thr 173.
Next we sought to confirm that Thr 210 is phosphorylated in vivo. Proteins extracted from GFP–Treacle-expressing 293T cells were sequentially digested with trypsin and Asp-N, followed by an enrichment of CK2 phosphopeptides through an affinity resin (Fig. 6c; see Methods for detail). This approach led to the unambiguous identification of a short peptide sequence harbouring Thr 210 of Treacle (Fig. 6d), and the mass of this peptide reveals Thr 210 as being phosphorylated (Fig. 6e). Collectively, these data strongly indicate that Thr 210 is indeed phosphorylated by CK2 in vivo, and that this modification constitutes the main interaction site for NBS1.
Treacle mediates NBS1 nucleolar translocation in response to DNA damage and a small fraction of it is transiently recruited to sites of DNA damage
Treacle has originally been described as a nucleolar protein that interacts with upstream binding factor (UBF), an RNA Pol I transcription factor. Indeed, we confirmed nucleolar localization of Treacle and co-localization with UBF by immunofluorescence (Supplementary Fig. 5a). Treacle staining is most pronounced in the dense fibrillar components of the nucleolus where the rDNA transcription takes place, and Treacle and Fibrillarin foci co-localize19 (Fig. 7a).
Interestingly, on induction of DSBs by laser micro-irradiation, NBS1–GFP nucleolar foci co-localized with Treacle, indicating that they occupied the same fibrillar structures after DNA damage (Fig. 7b). ChIP experiments in stable GFP–Treacle-expressing cell lines confirmed that, as for Pol I, Treacle mainly occupies the promoter and coding regions of the rRNA gene arrays (Supplementary Fig. 5b).
Significantly, depletion of Treacle by siRNA led to a complete loss of NBS1–GFP nucleolar localization and accumulation in response to DNA damage, thus revealing that Treacle acts as a mediator for DNA damage-induced recruitment of NBS1 in the nucleoli (Fig. 7c).
On the basis of sequence similarity of Treacle with Nopp140, a trafficking nucleolar phosphoprotein that shuttles between the nucleolus, the cytoplasm and Cajal bodies20, it was speculated that Treacle might also be trafficking to other cellular compartments19. Indeed, fluorescence recovery after photobleaching (FRAP) experiments in cells expressing a GFP-tagged version of the full-length Treacle complementary DNA revealed that GFP–Treacle is a mobile protein that shuttles between the nucleoplasm and the nucleoli (Supplementary Fig. 5c). We next sought to determine whether nucleoplasmic Treacle was recruited to sites of DNA damage. Laser micro-irradiation followed by fluorescence microscopy revealed weak accumulation of GFP–Treacle and endogenous Treacle within the irradiated subnuclear volume (Fig. 7d, e).
To more robustly demonstrate that a subset of Treacle is transiently recruited to sites of DNA damage we exploited a transcript variant of Treacle (so-called isoform c). This variant uses an alternative terminal exon and lacks ten 3′ exons. The resulting isoform has a significantly shorter and distinct C terminus as compared with the canonical transcript variant and lacks several putative nuclear and nucleolar localization signals (Supplementary Fig. 6a). This Treacle variant exclusively localized to the nucleoplasm and was efficiently recruited to microlaser-induced DNA breaks, probably owing to the lack of competition with the strong nucleolar binding (Fig. 7f). We next sought to determine the requirements for the mobilization of Treacle to DNA lesions. We first examined whether NBS1 is needed for Treacle recruitment to sites of DSBs. Surprisingly GFP–Treacle-c was still efficiently recruited to sites of laser-induced DNA damage in NBS1-depleted cells (Supplementary Fig. 6b). However, chemical inhibition of PARP activity significantly reduced Treacle association with DSB sites whereas an ATM inhibitor did not yield such an effect (Fig. 7g). In addition, PARP inhibition also led to a small but reproducible decrease in the number of cells with NBS1–GFP-positive nucleoli (Supplementary Fig. 6c), indicating that Treacle recruitment to sites of DNA damage might be involved in mediating efficient NBS1 nucleolar accumulation.
Together, these data strongly indicate that Treacle mediates DNA damage-induced recruitment of NBS1 in the nucleoli. Moreover, these data also show that the nucleoplasmic pool of Treacle is transiently recruited to sites of DNA damage in a manner that is partly dependent on PARP activity but is independent of ATM activity.
Treacle-dependent NBS1 recruitment into the nucleoli mediates rRNA silencing in response to DNA damage
Having established that in response to DNA damage, Treacle mediates NBS1 recruitment into the nucleoli, we next sought to determine whether Treacle is involved in the trans-signalling process that leads to downregulation of rRNA transcription (see above). Treacle itself is an essential factor for efficient rRNA transcription and its inactivation by mutation or its downregulation by RNA interference leads to reduced rates of rRNA synthesis21,22. Therefore, we carefully compared the reduction in rRNA transcription in response to DNA damage with the reduction in rRNA synthesis on Treacle depletion. This analysis revealed that rRNA synthesis drops by about 40% in response to DNA damage and on depletion of Treacle, respectively. However, there was no further reduction of rRNA transcription in response to DNA damage in the absence of Treacle (Fig. 8a). This lack of an additive effect was not due to an exhaustion of rRNA silencing, because rRNA transcription can be reduced further by treatment with actinomycin D (Fig. 8b). Therefore, these data indicate that the inhibition of rRNA transcription that occurs after Treacle depletion and exposure to DNA damage is conveyed through the same pathway.
However, these data do not provide direct evidence that NBS1 accumulation in the nucleoli is negatively regulating rRNA synthesis in response to DNA damage. Thus, to strengthen this conclusion, we fused a nucleolar localization signal (NOLS) to our NBS1–GFP construct and expressed it in HeLa cells. Under these conditions, recombinant NBS1–GFP–NOLS accumulated in the nucleoli in a subset of cells where it formed a similar focal pattern as observed after DNA damage. Significantly, EU labelling in those cells revealed that accumulation of NBS1 in the nucleoli correlates with low levels of rRNA transcription even in the absence of DNA damage (Fig. 8c). These data support the conclusion that Treacle-mediated NBS1 accumulation in the nucleoli regulates rRNA silencing in response to DNA damage.
In this study we identified a signalling pathway that globally inhibits rRNA transcription in response to DSBs through an in trans mechanism and we present compelling evidence that this Pol I silencing program is associated with rapid and transient recruitment of NBS1 into the nucleoli; a process that depends on the direct interaction between the NBS1 N-terminal FHA-BRCT region and the nucleolar phosphoprotein Treacle. This signalling mode is distinct from the previously reported DSB-induced transcriptional silencing that occurred in the vicinity of the DNA lesions and did not require long-distance signal transduction mechanisms12,13,23. Specifically, we show that DNA damage induction in randomly chosen nuclear compartments is accompanied by a global rRNA silencing response, which demands trans-compartmental communication between sites of DSBs and the transcription machinery within the nucleoli. Mechanistically, we provide evidence that this trans-compartmental communication is regulated by a genome surveillance complex, consisting of the DDR adaptor protein NBS1 and a nucleolar regulator Treacle (Fig. 8d).
Our findings indicate that the communication between the repair reactions locally at damaged chromosomes and other vital chromatin transactions elsewhere in the nucleus are more widespread than previously thought, and the example of the NBS1–Treacle dynamics might pave the way to dissect the physiological relevance of such intra-nuclear communication. Our previous work and this study show that nuclear dynamics of NBS1 is coordinated by two distinct ‘adaptor’ proteins, MDC1 and Treacle, respectively. Intriguingly, NBS1 interacts with these adaptors by the same mechanism based on direct interaction of its N-terminal FHA-BRCT region with repeated acidic CK2 phosphorylation sites present both in MDC1 and Treacle. Moreover, the thermodynamic parameters of the association between the NBS1 N-terminal FHA-BRCT region with phosphopeptides derived from the respective interaction sites in MDC1 and Treacle are virtually identical. Thus, it is likely that any given interaction of NBS1 molecules with either MDC1 or Treacle is mutually exclusive and that the relative proportion of MDC1–NBS1 versus Treacle–NBS1 complexes depends only on the relative concentration of MDC1 and Treacle in the nucleoplasm. As the steady-state levels of Treacle in the nucleoplasm are very low, we can assume that the bulk of NBS1 associates with MDC1. Nevertheless, as we report in this study, a small fraction of a constitutive NBS1–Treacle complex exists, and can be mobilized after DNA damage.
Mechanistically, the local increase in NBS1 concentration in the nucleoli on induction of DNA damage is not yet clear. It is possible that post-translational modifications of Treacle in response to DNA damage may lead to a conformational change that may favour Treacle–NBS1 association or facilitate release of the NBS1–Treacle complex from DSBs. Consistent with this idea, Treacle has been identified as an ATM target24 and we could confirm these results (Supplementary Fig. 7).
An interesting question is how Treacle-dependent NBS1 mobilization in the nucleoli mediates rRNA silencing in response to DNA damage. One simple mechanism may be that NBS1 mediates its effect on rRNA transcription indirectly through Treacle. When in a complex with NBS1, Treacle may simply not be able anymore to promote efficient rRNA transcription.
Both TCS and NBS are associated with craniofacial dysmorphism but only NBS is associated with microcephaly and mental retardation4. This raises the question of whether the defect in craniofacial development in TCS and that in NBS are the consequence of a common functional role of the two affected genes, TCOF1 and NBS1, respectively. However, whereas the dysmorphic head features of NBS patients derive from a primary developmental defect in the formation of the brain, the cranio-facial abnormalities in TCS patients are probably the result of increased apoptosis in the neuronal crest stem cell population early during development25. In addition, the recent finding that TCS is also caused by mutations in Pol I and Pol III subunits strongly supports the hypothesis that it is primarily a ribosomopathy26.
Finally, several benefits that cells could derive from a global rRNA silencing program triggered by the presence of DSBs come to mind. One intriguing possibility is provided by recent observations in yeast, where high rRNA transcription rates are associated with DNA repair defects and genome instability27. These observations suggest that in response to DNA damage, transient silencing of transcriptional activity within the rDNA repeats may be important to maintain genome integrity. □
U2OS, HeLa and 293T cells were grown in DME containing 10% fetal bovine serum (Invitrogen), 100 U penicillin and 100 μg ml−1 streptomycin. Where indicated, the culture medium was supplied with 50 ng ml−1 actinomycin D, 10 μM ATM inhibitor (KU55933) or Veliparib (ABT888) for 1 h. U2OS derivative cell lines expressing GFP–Treacle protein in a doxycycline-responsive manner were isolated by co-transfecting U2OS cells with pcDNA6/TR (Invitrogen) and pcDNA4-TO-GFP–Treacle constructs and selecting stably transfected cells with 400 μg ml−1 Zeocin and 5 μg ml−1 Blasticidin S (Invitrogen). U2OS cells stably expressing GFP–MRE11 and U2OS GFP–Treacle-c were selected with 400 μg ml−1 G418. The U2OS-NBS1–GFP2 cell line was previously described16.
Plasmids and transfections.
Transient transfections of HeLa and U2OS cells were done using Lipofectamine LTX with plus reagent (Invitrogen) according to the manufacturer’s specifications. Transient transfection of 293T cells was done using calcium phosphate. Full-length Treacle was synthesized in pcDNA4–TO–strep–HA (Genscript) and GFP was inserted to generate pcDNA4–TO–strep–HA–GFP–Treacle. NBS1–2GFP, NBS1-R28A–2GFP and NBS1-3A–2GFP (Ser 278, Ser 343 and Ser 397) were described previously16,28. NBS1–GFP–NOLS and GFP–NOLS were generated by digestion of NBS1–2GFP with BsrGI, NotI to remove one GFP and insert the NOLS (QDLWQWRKSL) sequence C terminally.
LpEXPR-IBA103-Nterm-NBS1-NLS-FLAG-HA-Strep-WT, R28A, K160M or DM (R28A and K160M) was generated by insertion of N-term-NBS1 (1–380)-NLS into pEXPR-IBA103-oligo. GFP–Treacle-c construct was generated by insertion of the TCOF1 isoform C into pAcGFP-C1. Treacle GST-fragments were generated by PCR amplification and subcloning into pGEX-4T3. Strep–HA–GFP–Treacle fragments were generated by PCR amplification and religation. Strep–HA-tagged NBS1 for co-purification and mass spectrometry was generated by insertion of an N-terminal NBS1 WT/R28A fragment (1–380) into pTGSH with a C-terminal strep–HA fusion.
RNA interference and site-directed mutagenesis.
The siRNA oligonucleotides were obtained from Microsynth AG. For annotations and sequences see Supplementary Table 1. siRNA oligonucleotides targeting MRE11 were from Ambion (MRE11 siRNA-59: AMBION 4427038, siRNA ID#:8959; MRE11 siRNA-60, AMBION 4427038, siRNA ID#:8960). The siRNA transfections were performed with 50 or 100 nM siRNA duplexes using Lipofectamine RNAiMAX (Invitrogen). Samples were collected 72 h after initiation of transfection unless stated otherwise. For mutagenesis primer sequences see Supplementary Table 1. Short hairpin RNA (shRNA) was generated by insertion of an oligonucleotide (Supplementary Table 1) into the pSUPERIOR.puro vector (Oligoengine). Cells were transfected with Lipofectamine LTX with plus reagent (Invitrogen) according to the manufacturer’s specifications and collected 4–5 days post transfection.
Generation of DNA damage.
X-ray irradiation was done with a YXLON.SMART 160E – 1.5 device (150 kV, 6 mA; YXLON International A/S) delivering 11.8 mGy s−1. Soft X-rays were largely filtered out with a 3 mm aluminum filter.
In situ detection of nascent rRNA was done with the Click-iT RNA Alexa Fluor 594, 488 Imaging Kit (Invitrogen, Molecular Probes). Briefly, cells were incubated for 20 min in the presence of 5-EU starting at the time of irradiation. Samples were fixed in 4% paraformaldehyde at room temperature for 12 min and permeabilized in 0.25% Triton X-100 for 5 min at room temperature. Samples were then processed according to the manufacturer’s recommendation. For high-content ScanR analysis fibrillarin staining was used to create a nucleolar mask for quantification of rRNA transcription. In general, the mean intensity of the 5-EU signal was measured within the nucleolar mask. In the case of ActD-treated cells the total intensity of the 5-EU signal was measured in the nucleus.
Microscopy and image analysis.
Quantitative image analysis for measurement of fluorescence intensities was done as described previously29. The images were obtained with a ×20 0.75 NA (UPLSAPO20x) dry objective, a quadruple-band filter set for DAPI, FITC, Cy3 and Cy5 fluorescent dyes, a MT20 Illumination system and a digital monochrome Hamamatsu C9100 EM-CCD (electron-multiplying charge-coupled device) camera. Camera resolution is 200 nM × 200 nM per pixel (binning 1, ×40). Image analysis was performed with Olympus propriety ScanR automated image and data analysis software using standard algorithms for detection of nuclei and sub-objects within nuclei. Typically, 49 images (corresponding to 2,000–4,000 sub-objects) were acquired under non-saturating conditions for each data point allowing robust measurements of experimental parameters such as intensities.
Laser micro-irradiation experiments were performed as described previously28 with a PALM micro beam equipped with a 355 nm ultraviolet-A laser (Zeiss). Confocal images were obtained on a LSM780 (Zeiss) using a Plan Apo ×40 1.4 Oil DICII objective. All images were taken under non-saturated conditions and images to be compared were acquired with the same settings. Non-confocal images were acquired on a Leica DMI 6000B equipped with a Leica DFC365FX camera using the HCX PL APO ×63 1.4 oil. Scoring of cells with NBS1–GFP2-positive nucleoli and GFP–Treacle or GFP–Treacle-c laser stripes was done manually. Cells would be scored as positive or negative on the basis of visual presence of GFP signal within nucleoli/laser tracks. Intensity plots for co-localization were generated with the Plot Profile function in ImageJ. For co-localization one region of interest (ROI) was applied to individual channels of the same image and intensities were measured. Time-lapse microscopy to follow accumulation of Nbs1–GFP2 within laser tracks and in the nucleoli was done using a Axiocam MRm (Zeiss) and Axio Vision software. Images were obtained with a LD Plan Neofluar ×40/0.6 corrM27 dry objective. Quantitative analysis of time-lapse images was done in ImageJ where time-lapse images were converted into stacks. Stacks were aligned using the StackReg plugin. After alignment a single ROI was applied to the stack and the average intensity or standard deviation was measured for the NBS1–GFP2 within laser tracks or in the nucleoli respectively.
FRAP experiments were performed on an UltraView Vox live-cell imaging system (Perkin Elmer) mounted on a Nikon Ti-E inverted microscope and integrated with a Photokinesis unit for photo-bleaching experiments. Cells were grown in glass-bottom Labtek dishes (Nunc, 155361) and kept under physiologic conditions for the duration of the experiment (37 °C incubator, custom-made CO2-independent medium, covered by a layer of mineral oil to prevent evaporation). Images were acquired using a Nikon ×60/1.4 APO oil objective, filter sets for GFP, and a Hamamatsu C9100-50 EM-CCD camera using propriety Volocity software (Perkin Elmer). GFP–Treacle protein was bleached by a 488 nm laser (100% output, 4 iterations). The bleached region was designed to bleach one nucleolus. Fluorescence recovery in the bleached region was monitored by acquiring 10 images before bleaching, 10 frames s−1 for the first 3 s followed by 30 frames min−1 for 15 min. An unbleached region was measured to allow background subtraction. The average of signal intensities before bleaching was set to 1 and the background level to 0 and intermediate values were transformed accordingly.
RNA was isolated with RNeasy, Qiagen. cDNA synthesis was based on 1 μg RNA using M-MLV Reverse Transcriptase (Promega, M170A) according to the manufacturer’s recommendation with random hexamer primers. Real-time PCR was performed using 10 ng cDNA with Roche SYBR Green I Master (Cat. 04707516001). Samples were analysed on the LightCycler480 and quantified with LightCycler480 quantification software. For primer sequences see Supplementary Table 1.
Cells were fixed with 1% formaldehyde for 10 min and the reaction was stopped by addition of 1.25 M glycine. Cells were washed in cold PBS and suspended in buffer A (100 mM Tris-HCl pH 8, 10 mM dithiothreitol), followed by 15 min incubation on ice and 15 min incubation at 37 °C. Cells were pelleted and resuspended in Buffer B (10 mM EDTA, 10 mM EGTA, 10 mM HEPES at pH 8, 0.25% Triton X-100). Again, cells were pelleted and resuspended in Buffer C (10 mM EDTA, 0.5 mM EGTA, 10 mM HEPES pH 8, 200 mM NaCl). Finally, cells were pelleted and resuspended in lysis Buffer D (1% SDS, 10 mM EDTA, 50 mM Tris-HCl at pH 8.1). After sonication to yield DNA fragments of 0.5–1 kilobases, 100 μg chromatin was diluted tenfold with IP buffer (16.7 mM Tris-HCl at pH 8.1, 167 mM NaCl, 1.2 mM EDTA, 0.01% SDS, 1.1% Triton X-100). Proteins of interest were immunoprecipitated with salmon sperm DNA/Protein A/G agarose (Millipore, 16-157/16-201) overnight with the respective antibodies. Beads were washed extensively with washing buffer 1A (10 mM Tris-HCl at pH8, 150 mM NaCl, 0.1% SDS, 1% Triton X-100), washing buffer 1B (10 mM Tris-HCl pH 8, 500 mM NaCl, 0.1% SDS, 1% Triton X-100) and washing buffer 2 (0.25 M LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl at pH 8) before elution in 100 μl elution buffer (1% SDS, 100 mM NaHCO3). Proteinase K was added to each sample and reversion of crosslinks was done by heating for 1 h at 37 °C followed by overnight incubation at 65 °C. QIAquick PCR purification kit (Qiagen) was used to purify eluted DNA. Seven per cent of the purified DNA was amplified by 40 cycles using a LightCycler480 (Roche) with the SYBR Green detection system. A standard curve was included for each primer pair to compare the abundance at different regions of the rDNA repeat.
Total cell lysates were prepared in Laemmli sample buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue and 10% glycerol). Alternatively, cell extracts were prepared in the following buffer (50 mM Tris, pH 7.5, 120 mM NaCl, 0.5% NP-40, and 1 mM EDTA). Extracts were supplemented with 2 mM MgCl2 and benzonase and incubated for 30 min at 4 °C. Lysates were cleared by centrifugation at 14,000g for 15 min and protein concentration was measured using Bradford protein assay. Immunoprecipitations were done using Anti-Flag M2 affinity gel (Sigma, A2220), Monoclonal anti-HA agarose (Sigma, A2095), Dynabeads M-280 Streptavidin (Invitrogen, 112.05D) or Treacle antibody, in combination with Protein A agarose.
Other antibodies used in this study were: mouse γH2AX (Millipore, clone JBW302, catalogue number 05-636, 1:500), rabbit RPA194 (Santa Cruz, sc28714, 1:1,000), rabbit NBS1 (Novus Biologicals, NB100-143, 1:500), mouse NBS1 (Genetex, Clone 1D7, GTX70224, 1:500), Treacle antibody (Sigma Life Science, HPA038237, western blotting: 1:500, immunofluorescence 1:100), rabbit HA (Abcam, ab9110, 1:2,000), rabbit Flag (Sigma, F7425, 1:1,000), mouse fibrillarin (Abcam, 38F3, ab4566, 1:200), rabbit SMC1 (Abcam, ab9262, 1:500), rabbit Bid pS61 (Bethyl, A300-527A, 1:500), mouse Mcm7 (Santa Cruz, DCS141, sc65469, 1:500), mouse UBF (Santa Cruz f-9, sc-13125, 1:300), mouse GFP (Roche, 11814460001, 1:1,000), mouse tubulin (Sigma, DM1A, T6199, 1:2,000), mouse Mre11 (Abcam, 12D7, ab214, 1:500), sheep MDC1 3835 (gift from S. Jackson, 1:2,000).
CK2 phosphorylation assay and pulldown were done as previously described6. Briefly, 100 ng of recombinant CK2a (Milipore) was added to 1 μg of purified GST-fusion protein in CK2 kinase buffer (20 mM MOPS, pH 7.2, 25 mM β-glycerophosphate, 5 mM EGTA, 1 mM sodium orthovanadate, 37.5 mM MgCl2, 1 μM ATP and 10 μCi [γ-32P]ATP) and incubated for 10 min at 30 °C. In the case of GST-pulldown after in vitro phosphorylation, [γ-32P]ATP was excluded. Kinase reactions were inactivated by boiling in SDS sample buffer and were run on SDS polyacrylamide gels.
Isothermal titration calorimetry.
The affinity and thermodynamic parameters for NBS1 interactions with synthetic phosphopeptides were determined as previously described10. Briefly, isothermal titration calorimetry was done with a VP-ITC instrument (MicroCal). NBS1 samples were dialysed extensively into 50 mM HEPES (pH 7.5), 150 mM NaCl, 2 mM β-mercaptoethanol. Before use, all peptides were desalted and buffer exchanged using NAP-5 purification columns (GE Healthcare) into the relevant buffer. Peptides (0.5–1.5 mM) were titrated into 0.05 mM NBS1. Data were analysed with Origin 7.0 software. The peptide sequences were as follows: SETE: TLVpSEpTEEE; SDET: SSSpSDEpTDVE; SDSE: SEEpSDpSEEE.
Double-affinity purification and mass spectrometry analysis for identification of Treacle was done by Dualsystems Biotech AG (CaptiVate Shotgun proteomics service). For enrichment and mass spectrometric analysis of CK2-specific phosphopeptides, 293T cells were transfected with GFP–TCOF1 and collected in high-salt RIPA buffer (50 mM Tris pH 7.5, 400 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 0.1% Na-deoxycholate, 2 mM Na-orthovanadate, 5 mM NaF, 5 mM glycero-2-phosphate, protease inhibitors (Roche)). Proteins from cleared lysate were precipitated in acetone, dissolved in urea, reduced with dithiothreitol, alkylated with chloroacetamide30, and digested using Lys-C and modified sequencing-grade trypsin (Sigma) followed by digestion with sequencing grade Asp-N (Promega). Protease digestion was terminated by addition of trifluoroacetic acid and peptides were purified using reversed-phase Sep-Pak C18 cartridges (Waters). CK2-specific phosphopeptides were enriched using the PTMScan Phospho-CK2 Substrate Motif (S*DXE or T*DXE) Kit (Cell Signaling). Mass spectrometric experiments were performed on a nanoscale UHPLC system (EASY-nLC1000 from Proxeon Biosystems) connected to an Orbitrap Q-Exactive equipped with a nanoelectrospray source (Thermo Fisher Scientific). Each peptide fraction was auto-sampled and separated on a 15 cm analytical column (75 μm ID) in-house packed with 1.9 μm C18 beads using a 2 h gradient ranging from 5% to 40% acetonitrile in 0.5% formic acid at a flow rate of 200 nl min−1. The Q Exactive was operated in data-dependent acquisition mode and all samples were analysed using a previously described ‘sensitive’ acquisition method31. All raw data analysis was performed with MaxQuant software suite32 version 126.96.36.199 supported by the Andromeda search engine33.
We thank S. Jackson (The Gurdon Institute, Cambridge, UK) for reagents, R. Santoro for technical support in relation to rRNA transcription, C. Dinant for critical discussion of the FRAP data and members of the Department of Gynecology, the Novo Nordisk Foundation Center for Protein Research and the National Institute of Medical Research for helpful discussions. This work was supported by grants from the Swiss National Foundation (3100A0-111818 and 31003A-144284), Promedica Foundation, Lundbeckfonden (R93-A8863), Novo Nordisk Foundation and by the Kanton of Zürich. S.J.S. is supported by the Medical Research Council UK (U117584228).
About this article
Molecular Neurobiology (2018)