A role of human RNase P subunits, Rpp29 and Rpp21, in homology directed-repair of double-strand breaks

DNA damage response (DDR) is needed to repair damaged DNA for genomic integrity preservation. Defective DDR causes accumulation of deleterious mutations and DNA lesions that can lead to genomic instabilities and carcinogenesis. Identifying new players in the DDR, therefore, is essential to advance the understanding of the molecular mechanisms by which cells keep their genetic material intact. Here, we show that the core protein subunits Rpp29 and Rpp21 of human RNase P complex are implicated in DDR. We demonstrate that Rpp29 and Rpp21 depletion impairs double-strand break (DSB) repair by homology-directed repair (HDR), but has no deleterious effect on the integrity of non-homologous end joining. We also demonstrate that Rpp29 and Rpp21, but not Rpp14, Rpp25 and Rpp38, are rapidly and transiently recruited to laser-microirradiated sites. Rpp29 and Rpp21 bind poly ADP-ribose moieties and are recruited to DNA damage sites in a PARP1-dependent manner. Remarkably, depletion of the catalytic H1 RNA subunit diminishes their recruitment to laser-microirradiated regions. Moreover, RNase P activity is augmented after DNA damage in a PARP1-dependent manner. Altogether, our results describe a previously unrecognized function of the RNase P subunits, Rpp29 and Rpp21, in fine-tuning HDR of DSBs.


Results
Rpp29 promotes homology-directed repair of double-strand breaks. Mock and Rpp29-depleted U2OS cells were exposed to 10 Gys of IR and examined at different time points after damage by neutral comet assay (see Materials and Methods). Results demonstrate that Rpp29 knockdown using two different specific siRNAs (Fig. 1A) disrupted DSB repair, as evident from the high percentage of DNA in the comet tail, which reflected the levels of DSBs in individual cells. Indeed, the percentage of damaged DNA in the comet tail was significantly higher in Rpp29 deficient cells at 1, 3 and 6 h after IR when compared with that seen in mock-transfected cells (Fig. 1B). This outcome indicates that Rpp29 depletion increases the accumulation of broken DNA, possibly by reducing the IR-induced DSB repair.
To gain molecular insights into the mechanism by which Rpp29 affects DNA damage repair, we looked at the phosphorylation kinetics of early DNA damage markers. We found that the phosphorylation levels of two well-characterized ATM substrates, H2AX-Ser139 and NBS1-Ser343 were significantly increased in Rpp29-depleted cells compared to control cells ( Fig. 2A,B). Similarly, RPA2 phosphorylation at Ser4 and Ser8 (pRPA) was greatly intensified in 4 h after damage induction, in Rpp29 depleted cells (Fig. 2B). Since phosphorylated RPA2 is a known indicator for ssDNA generation following DNA end resection, which promotes HDR of DSB 54,55 , these site-specific phosphorylation support the existence of persistent DNA 5′-end due to defective HDR of DSBs.
To confirm a direct role of Rpp29 in DSB repair, we engineered a traffic light reporter (TLR) system 56 integrated into the genome of U2OS cells (U2OS-TLR). This system enables simultaneous evaluation of DSB repair by HDR and NHEJ in the same cell (Fig. S1). Remarkably, Rpp29 depletion in U2OS-TLR cells using three different Real-time PCR analysis shows ~90% reduction in the steady state levels of Rpp29 mRNA in cells transfected with two different Rpp29 siRNAs, #23 and #25, and compared with those in cells transfected with control (CtRL) siRNA. P-values were calculated by two-sided Student's t-test relative to ctrl siRNA; ***p < 0.001. (B) Control and Rpp29 depleted cells were exposed to 10 Gys of ionizing radiation (IR), harvested at the indicated time points after IR, and subjected to neutral comet assay. Quantitation of the DNA percentage in comet tail reveals DSB-repair deficiency in Rpp29 depleted cells. Error bars represent standard deviation (SD) from three independent experiments (n = 80 cells). Two-way ANOVA was used to test for differences at each dose; ** and **** indicate significance at p < 0.01 and p < 0.0001, respectively.
siRNAs resulted in ~50% decrease in the number of GFP-expressing cells, when compared with cells treated with control siRNA (Fig. 2C). This decrease demonstrates that Rpp29 is required for intact HDR of DSBs.
Interestingly, the above reduction of HDR was accompanied by an increase in mCherry production (Fig. 2D), an indication that Rpp29 depletion had no deleterious effect on the integrity of NHEJ. Apparently, the activation of NHEJ pathway is a counteracting response of the cells to elevated numbers of DSBs owing to the defective HDR. As a control, treatment of U2OS-TLR cells with caffeine, which is known to inhibit PIKKs 57 , brought to a  shows the efficiency of Rpp29 knockdown in U2OS-TLR cells. Cells were transfected with control siRNA (Ctrl) or three different Rpp29 siRNAs. The y-axis represents the relative Rpp29 mRNA level normalized to that of GAPDH. P-values were calculated by two-sided Student's t-test relative to Ctrl siRNA; ***p < 0.001. (B) Rpp29 knockdown increases IR-induced levels of pRPA2 S4/S8 phosphorylation, γH2AX, and NBS1 phosphorylation levels. Control and Rpp29-depleted U2OS cells were exposed to IR (10 Gys) and protein extracts were prepared at the indicated time points and subjected to western blot analysis using antibodies directed against the indicated proteins. Total RPA and β-actin were used as internal controls. The bands intensities of γH2AX, pNBS1 or pRPA were normalized relative to the intensities of their respective β-actin or total RPA bands, respectively. The ratios are shown at the bottom of the blot. Results are representative of two independent experiments. (C) Rpp29 knockdown impairs HDR of DSBs generated by I-SceI endonuclease. A reduction of ~50% in GFP-positive cells was observed after Rpp29 depletion. Caffeine was used as a positive control. Results shown are typical of three independent experiments. Error bars represent SD. P-values were calculated by twosided Student's t-test relative to Ctrl siRNA; ** and **** indicate significance at P < 0.01 and 0.0001. Given that HDR can occur only in S/G2 cell phase, data were corrected to flow-cytomertry S/G2 values. (D) TLR results for DSBs repair by NHEJ. Rpp29 knockdown had no significant effect on the integrity of NHEJ. Data represents SD for three independent experiments. (E) Percentage of control and Rpp29-depleted cells with more than five 53BP1 foci. Data shown represent at least 300 cells. Two-way ANOVA was used to test for differences at each dose; *p < 0.05.
severe reduction in DSB repair by both HDR and NHEJ (Fig. 2C,D). Furthermore, the enhanced NHEJ in cells deficient in Rpp29 was concomitant with a moderate increase in 53BP1 foci, thereby favoring DSB repair by NHEJ (Fig. 2E).
Rpp29 and Rpp21 are rapidly and transiently recruited to laser-microirradiated DNA sites. We sought to corroborate the role of Rpp29 in DDR by monitoring its subcellular localization before and after DNA damage induction. For this purpose, U2OS-TetON cell line that conditionally expresses EGFP-Rpp29 fusion protein upon the addition of doxycycline was established (Fig. S2). The cell line, termed U2OS-TetON-EGFP-Rpp29, was subjected to laser micro-irradiation that induces a localized DNA damage within the nucleus of living cells 58,59 and protein localization was monitored. Results show that ~70% of the laser-microirradiated cells (n = 50) exhibited rapid accumulation of EGFP-Rpp29 at laser-microirradiated sites. The accumulation was detectable within 15 seconds after irradiation, thus demonstrating that Rpp29 recruitment is an early event in the DDR. Accumulation of Rpp29 peaked at ~5 minutes and was followed by gradual dispersal out of the DNA damage sites (Fig. 3A).
To further substantiate the recruitment of Rpp29 to DNA damage sites, localization of the endogenous Rpp29 was monitored by immunofluorescence analysis after cells exposure to ionizing radiation. The results revealed that endogenous Rpp29 was mobilized to IR-induced foci (IRIF) and extensively co-localized with γH2AX foci (Fig. 3B). This co-localization at IRIF is consistent with localization of exogenous Rpp29 fused to GFP at laser-microirradiated sites, and confirms that this protein is targeted to DSBs.
Next, we sought to determine the effect of DNA damage on the subcellular localization of other subunits of RNase P, including its specific subunit Rpp21, which is not shared by RNase MRP 60 . Similar to Rpp29, Rpp21 also displayed rapid accumulation and dispersal kinetics at DNA damage sites (Fig. 3C). Of note, the recruitment of Rpp29 and Rpp21 to laser-microirradiated sites was independent of each other. Thus, knockdown of Rpp21 had no visible effect on Rpp29 accumulation at laser-microirradiated sites and vice versa (Fig. 3D,E). By contrast, Rpp14, Rpp25 and Rpp38, which are not components of the catalytic core of RNase P 38 , showed no perceptible recruitment to laser-microirradiated sites (Fig. S3). We concluded therefore that not all human RNase P protein subunits are recruited to DNA damage sites. The recruitment of Rpp21 to DNA damage sites prompted us to determine its effect on DSB repair using the above-described U2OS-TLR reported cells. Results show that depletion of Rpp21 using two different siRNAs (Fig. S4A) leads to ~50% decrease in GFP-positive cells compared to control siRNA-treated cells (Fig. S4B). On the other hand, Rpp21 depletion had no significant changes in the percentage of mCherry positive cells (Fig. S4C). Altogether, we concluded that, similar to Rpp29, Rpp21 predominantly facilitates HDR of DSB. Interestingly, Rpp21 and Rpp29 are amplified in many types of human cancer 61, 62 , we sought therefore to determine whether overexpression of Rpp21 and Rpp29 affects the integrity of DSB repair. Toward this end, U2OS-TLR cells were transfected with expression vectors encoding flag, flag-Rpp21 or flag-Rpp29 and the efficacy of DSB repair was determined (Fig. S5A). Results show that overexpression of neither Rpp21, nor Rpp29 has an effect on the integrity of HDR or NHEJ of DSBs (Fig. S5B,C). PARP1-depenedent recruitment of Rpp21/Rpp29 to DNA damage sites. As mentioned above, Rpp29 undergoes DNA damage-induced phosphorylation at serine 10 by PIKKs 53 . To find out if this site-specific phosphorylation is of any importance for the protein localization response to DSB, serine 10 residue was substituted with alanine, and recruitment of the resulted mutant Rpp29 protein to laser-microirradiated sites was examined. As shown in Supplementary Fig. S6, this PIKK-related mutation had no noticeable effect on Rpp29 compiling at damaged sites. Furthermore, inhibition of the ATM kinase activity by the pharmacological inhibitor KU-55933 did not influence the recruitment of Rpp29, as well as that of Rpp21, to laser-microirradiated sites (Fig. S7A,B). The efficacy of ATM inhibition was validated by visualizing CtIP at laser-microirradiated sites, in which this endonuclease marker was rather abolished (Fig. S7C), an outcome that is in agreement with a previous report 63 . Apparently, Rpp29 and Rpp21 recruitment to DNA damage sites is not regulated by ATM signaling pathway. Bioinformatics analysis revealed that Rpp29 and Rpp21 contain a consensus motif for binding of poly(ADP-ribose) (PAR) moieties (Fig. 4A). Indeed, bacterially expressed recombinant Rpp29 and Rpp21 human proteins, which were purified to near homogeneity, were able to bind PAR in a dose-dependent manner (Fig. 4B,C). Next, we sought to determine whether Rpp29 and Rpp21 undergo in vivo poly(ADP)-ribosylation (PARylation) in response to DNA damage. To do so, EGFP-Rpp29 and EGFP-Rpp21 fusions were purified using GFP-TRAP beads from undamaged and IR-damaged cells and the immunoprecipitates were immunoblotted with PAR and GFP antibodies. Results show that Rpp29 and Rpp21 were not PARylated (Fig. S8). Collectively, these observations suggest that binding of Rpp21 and Rpp29 to PAR moieties, rather than their PARylation, facilitates their mobilization to DNA damage sites. In agreement with this notion, two complementary approaches implicated PARP1 in the regulation of Rpp29 and Rpp21 recruitment to DNA breakage sites. First, depletion of U2OS cells from PARP1 by the use of siRNA (Fig. 5A) led to a remarkable decrease (~90%) in number of cells showing accumulation of Rpp21 and Rpp29 at laser-microirradiated sites, when compared with those of mock-transfected cells (Fig. 5B,C). Second, pretreating cells with PARP inhibitor Ku-0059436 abrogates the recruitment of Rpp29 and Rpp21 to DNA damage sites (Fig. 5D,E). Hence, PARP1 is critical for Rpp21 and Rpp29 recruitment to DNA damage sites.
H1 RNA knockdown interferes with the recruitment of Rpp21 and Rpp29 to DNA damage sites. As mentioned above, RNAs are implicated in recruiting DDR proteins to DNA damage sites. Since Rpp21 and Rpp29 can bind the catalytic H1 RNA of human RNase P 33, 38, 64 , we sought to assess whether RNA is required for these two proteins to be recruited to DNA damage sites. Thus U2OS-TetON cells expressing EGFP-Rpp21 or EGFP-Rpp29 were treated with RNase A 65 and then were exposed to laser-microirradiation. RNase A treatment caused ~3.3-fold decrease in the number of cells (n = 30) that showed accumulation of Rpp29 and Rpp21 at DNA damaged sites (Fig. 6A effect on the recruitment of PARP1 to laser-microirradiated sites (Fig. 6C), thereby ruling out the possibility of a general indirect effect on the subcellular distribution of DDR proteins. Interestingly, inhibition of RNA Pol II activity by α-amanitin showed no detectable effect on the recruitment of EGFP-Rpp21 or EGFP-Rpp29 to DNA laser-microirradiated sites. Altogether, we concluded that RNA molecules, rather than active transcription, are critical for EGFP-Rpp21 or EGFP-Rpp29 accumulation at DNA damage sites (Fig. S9).
To gain further insights into the identity of the RNA molecules that may regulate Rpp21 and Rpp29 recruitment to DNA damage sites, we checked the involvement of H1 RNA mainly because it associates with Rpp21 and Rpp29 32, 37-39 . Toward this end, H1 RNA was knocked down in U2OS-TetON cells by shRNA using lentivirus plasmid (pLKO.1-TRC). This shRNA was directed against a single-stranded RNA region spanning positions 164-184 of H1 RNA. Knock down of this relatively stable 340-nt RNA was determined by real-time qRT-PCR analysis. Remarkably, a reduction of ~65% in the steady state levels of H1 RNA (Fig. 7A) was accompanied by ~50% decrease in the number of cells with Rpp21 and Rpp29 localized at laser-microirradiated sites (Fig. 7B,C). Considering the knockdown efficiency of this stable RNA, these results indicate that H1 RNA is critical for accumulation of Rpp29 and Rpp21 at DSB sites, possibly through RNA-protein interactions.

RNase P activity increases after DNA damage in a PARP1-dependent manner. The fact that H1
RNA is critical for Rpp29 and Rpp21 accumulation at DNA damage sites (Fig. 7) prompted us to check for RNase P activity in response to DNA damage. We found that the activity of this catalytic RNP was elevated in whole extracts (S20) prepared from U2OS cells that were harvested immediately after IR, as compared to that seen in extracts of untreated cells (Fig. 8A, 0 min; orange vs black graph; Fig. S10, A, lanes 3-6 versus 7-10). However, basal enzymatic activity was recovered in 30 min post irradiation (Fig. 8A). This swift and transient change in RNase P activity is consistent with the rapid recruitment of Rpp29 and Rpp21 to DNA damage sites and its dependence on H1 RNA. Moreover, PARP1 inhibition resulted in a marked decrease in RNase P activity in maturation of tRNA in whole extracts (Fig. 8B, orange vs black graphs; Fig. S10B, 11-14 vs S9A, lanes 11-14; 3′ tRNA band) and in coupled transcription/processing system of human tRNA Arg (UCU) gene (Fig. 8C, lane 7 vs 3, and 8D) in extracts derived from irradiated U2OS cells, but not from untreated cells. The decrease was transient, as it was seen in 30 min post irradiation. Since the steady state levels of H1 RNA, as well as those of Rpp20 and Rpp40, remained unchanged in irradiated cells (Fig. S11), the observed PARP-dependent inhibition of RNase P activity and protein subunit recruitment to DSBs provide mechanistic basis for the involvement of a novel variant of nuclear RNase P RNP in DSB repair.

Discussion
RNase P is classically defined as an RNA enzyme that carries out the excision of the 5′ leader of precursor tRNA. The catalytic asset of human RNase P is related to the H1 RNA subunit. Rpp29 and Rpp21 serve as cofactors for H1 RNA with which these two subunits form an active RNP complex 34,38,66,67 . In this study, we have shown that Rpp29, Rpp21 and H1 RNA are involved in DDR. Functional analyses confirm that depletion of Rpp29 leads to accumulation of broken DNA, as determined by comet assay. Since other subunits, such as Rpp14, Rpp25 and Rpp38, do not similarly respond to the induction of DSBs, and the recruitment of Rpp21 and Rpp29 to DSB sites is independent of each other, but both subunits rely on H1 RNA, we speculate that a non-canonical form of RNase P responds to these lethal DNA lesions. In this regard, our data raise two possibilities: H1 RNA catalytic activity might be required for Rpp21 and Rpp29 recruitment to DSB sites. Alternatively, H1 RNA may serve as a scaffold for recruiting Rpp21 and Rpp29 subunits to DNA damage sites. In support of this, we know today about several ncRNAs such as: TSIX 68 , Air 69 , Tug1 70 , NRON 71 , Kcnq1ot1 72 and HOTAIR 73 which are able to interact with RNA binding proteins to regulate their activity and subcellular localization. The dependence of recruitment and processing activity of RNase P on PARP1 indicates that this RNP participates in the major PARP1-mediated repair pathway. One challenging question is how does PARP1 regulate Rpp21 and Rpp29 recruitment to DNA breakage sites? Our data show that Rpp21 and Rpp29 possess PAR-binding motifs, bind PAR moieties in vitro (Fig. 4), but do not undergo ADP-ribosylation (Fig. S8). These observations altogether favor a model by which PARP1-mediated ADP-ribosylation of histones and non-histone proteins at sites of damage provide a platform for recruiting Rpp21 and Rpp29. Additional question: How is PARP1 involved in the DNA damage-induced increase of RNase P activity? While we cannot rule out a possibility that PARP1 may regulate RNase P activity by PARylating protein subunits other than Rpp21 and Rpp29, we assume that Rpp21/ Rpp29 binding to PAR moieties may increase RNase P catalytic activity. In agreement with this assumption, several reports show that proteins activity is altered following their binding to PAR moieties 74 .
Our discovery that human RNase P has a role in DDR is supported by three indirect reports. First, Rpp29 undergoes IR-induced phosphorylation 53 . Second, Rpp29 interacts with histone H3.3 and represses its incorporation into chromatin 75 . H3.3 deposition is implicated in DDR, as previous studies reported on active deposition of H3.3 variant at UV-C damage sites 76 and at laser microirradiation-induced DSB repair by NHEJ 77 . Third, genetic studies in Drosophila melanogaster with a mutated RNase P due to null mutations in the RPP30 gene, which codes for the core protein subunit Rpp30, reveal that sterile female display replication stress in atrophied ovaries 78 . The atrophied ovaries contain high steady state levels of precursor tRNAs with unprocessed 5′ leader sequences, but normal levels of mature tRNA. Though the molecular mechanism underlying replication stress remains unknown, it is proposed that structural and functional defects in RNase P lead to the activation of a number of key DNA damage checkpoint proteins, including p53, Claspin, and Chk2 78 .
Our results reveal that the role of human RNase P RNP in DSB repair is confined to HDR and not NHEJ. In general, Rpp29 and Rpp21 may promote HDR of DSBs through several pathways: (i) by serving as a scaffold for recruiting HDR proteins to DSB sites. (ii) By recruiting non-coding RNA molecules to DSB sites. In support of this, Rpp29 and Rpp21 are known to directly bind RNA 32,[37][38][39] . In addition, a growing number of evidence showed that RNA transcripts serve as a template for HDR of DSBs 79-81 . (iii) By processing noncoding RNA molecules in response to DNA damage. In this regard, tRNA-derived fragments have been shown to act as tumor suppressors through a posttranscriptional mechanism in human cells 82 . The tRNA-derived fragments are processed, by yet unknown enzyme(s), upon exposure to hypoxic stress 82 . Notably, previous studies reveal some other damage-induced RNAs processed to smaller RNAs by DICER and DROSHA or by DICER-like proteins 5, 10, 83 . On the other hand, CU1276, a tRNA derived 22-nt RNA that modulates DDR 84 is generated in a DROSHA-and DICER-independent manner, suggesting that other RNA-processing machineries are implicated in tRNA processing [85][86][87][88] . Based on these studies, and based on our findings showing increased RNase P activity upon DNA damage, we hypothesize that human RNase P might be involved in DSB repair by generating stress-induced tRNA fragments. Future studies need to be conducted to elucidate whether indeed RNase P enzymatic activity is required for HDR of DSB.
In summary, our data identified a variant of human RNase P as a new component of DNA repairome that underpins HDR of DSBs (Fig. 8E). The remarkable evolutionary conservation of this RNP may shed new light on common primordial mechanisms by which eukaryotes preserve their genomes. mock and PARP1-siRNA transfected cells. Each measurement is representative of at least 10 cells. Error bars indicate SD. Column graphs (right) display the percentage of PARP1-depleted cells exhibiting accumulation of EGFP-Rpp29 and EGFP-Rpp21 at damage sites, as compared with mock-transfected cells. Error bars represent the SEM from two independent experiments. P-values were calculated by two-sided Student's t-test relative to mock; ***p < 0.001. (D,E) Representative time-lapse images (left) show localization of EGFP-Rpp29 (D) and EGFP-Rpp21 (E) fusions to laser-microirradiated sites (marked by red arrow) in U2OS cells treated for 1 h with either DMSO or 5 μM of the PARP1/2 inhibitor (Ku-0059436)(PARPi). Results shown are typical of two independent experiments and represent at least 30 different cells. Line graphs (middle) show fold increase in the relative fluorescence intensity of EGFP-Rpp29 and EGFP-Rpp21 at laser-microirradiated sites in untreated and PARPi-treated cells. Column graphs (right) display the percentage of PARPi-treated cells with accumulation of EGFP-Rpp29 and EGFP-Rpp21 at damage sites, as compared with untreated cells. Error bars express the SEM from two independent experiments. P-values were calculated by two-sided Student's t-test relative to DMSO; ***p < 0.001. ing EGFP-Rpp21 and EGFP-Rpp29 were established as previously described 58 . Briefly, EGFP-Rpp21 and EGFP-Rpp29 genes were sub-cloned into pTRE2/Puro vector (Clontech) and resulted constructs were introduced into U2OS-TetON cells (Clontech). Puromycin-resistant clones were then selected and tested for doxycycline  Supplementary Fig. S10, were quantified and ratios of product/substrate were plotted. (C) Processing of a nascent precursor tRNA Arg (UCU) is inhibited by PARP1 inhibitor. U2OS cells were treated as in Supplementary  Fig. S10, and dialyzed S20 extracts were assayed for processing of nascent precursor tRNA Arg transcribed from a cloned gene for the indicated times (in min), as previously described 50 . Labeled RNAs were resolved in an 8% polyacrylamide sequencing gel. The positions of the primary transcript, 93 nt in length, and transcript processed at 5′ end by RNase P, 88 nt in size, are shown. Extracts of U2OS cells are not as efficient as HeLa cells in tRNA gene transcription and splicing to mature tRNA (not shown), thus producing weak labeled RNA signals. (D) The 93-and 88-nt transcript bands seen in C were quantitated and the ratios of processed to unprocessed tRNA Arg were plotted. P-values were calculated by two-sided Student's t-test relative to scr shRNA; *, **, and *** indicates significance of p < 0.05, 0.01 and 0.001, respectively. (E) A hypothetical model shows that local ADP-ribosylation at DSB site and H1 RNA molecule underpin the recruitment of Rpp29 and Rpp21 to promote HDR of DSBs.
(Dox)-induced expression of the fusion proteins by fluorescent microscopy. Cell clones that showed expression of the fusion proteins in the presence of Dox (Sigma, D9891) were selected for further characterization.

Generation of U2OS-TLR cells. Viral particles containing the pCVL Traffic Light Reporter 1.1 (Sce target)
Ef1a Puro 56 were generated by transfecting HEK293T cells together with plasmids encoding the lentiviral proteins Gag, Pol and VSV-G. Media containing the viral particles were collected in 48 h post transfection, filtered and applied on U2OS cells. Infected cells with integrated reporters were selected in the presence of 0.6 μg/ml of puromycin for one week. Puromycin-resistant cells were sorted using FACSAria Cell Sorter (BD Biosciences) and mCherry negative cells were collected and used for TLR assays.
Western blot. Hot-lysis buffer was used to prepare protein extracts. Samples were separated on SDS-PAGE gels and membranes were immunoblotted with relevant antibodies (Supplementary Table 1). The immunoblots were developed using Quantum ECL detection kit (K-12042-D20, Advansta). The intensity of the immunoblot bands was performed using ImageJ software.
GFP Trap pull-down. GFP only, GFP-Rpp21, and GFP-Rpp29 fusions were purified from U2OS cells using GFP-TRAP methodology as previously described 58 . Briefly, U2OS cells were transfected with pEGFP-C1, pEGFP-Rpp21 or pEGFP-Rpp29 expression vectors and whole-cell extracts were prepared using NP40 lysis buffer and subjected to pull-down using GFP-Trap beads (Chromotek). Next, the immuno-complexes were washed, resolved by SDS-PAGE and blotted with the indicated antibodies.
Immunofluorescence. Immunofluorescence analysis was done as previously described 58 . Cells were seeded on coverslips at least 24 h prior to the experiment, immunostained with the indicated antibodies (Table S1) and visualized using the inverted Zeiss LSM 700 confocal microscope with 40× oil EC Plan Neofluar objective. For 53BP1 foci quantification experiment, U2OS cells were seeded in 96-well plates 24 h prior experiment. Cells were exposed to 3 Gys of IR, fixed at the indicated time points and immunostained with 53BP1 antibody. High-content screening microscope (IN Cell Analyzer 2000; GE Healthcare) was used for automatic acquisition of at least 300 cells at each time point. The number of 53BP1 foci was calculated using the IN Cell Analyzer Workstation.
RNA isolation, reverse transcription, and quantitative real-time PCR. RNA was extracted from U2OS cells transfected with siRNA/shRNA using TRizol reagent, according to the manufacturer's instructions (Ambion). Aliquots of RNA (1 µg) were used for cDNA synthesis using the qScript cDNA Synthesis Kit (Quanta) using random primers. The steady state levels of mRNAs encoding for Rpp29 and Rpp21, as well as H1 RNA, were measured by real-time PCR in a Step-One-Plus Real time PCR System (Applied Biosystems) using Fast SYBR Green Master mix (Applied Biosystems) and specific primers for Rpp29, Rpp21, H1 RNA and GAPDH (Table S2). Each PCR reaction was repeated 3 times, data analysis and quantification were performed using the Step-One software V2.2 supplied by Applied Biosystems. RNA levels were calculated using ΔΔCt method from results that were normalized to GAPDH gene expression.
Laser microirradiation. Cells were subjected to laser microirradiation as previously described 58 . Briefly, cells were plated on a 35-mm imaging dish with glass bottom (Ibidi; Cat#81158) and pre-sensitized with 10 µM of Hoechst 3334 dye for 15 min at 37 °C. Laser microirradiation was executed using a LSM-700 inverted confocal microscope equipped with CO 2 module and 37 °C heating chamber. DNA damage was induced by microirradiating of a single region in the nucleus with 10 iterations of 405-nm laser beam. Time-lapse images were then acquired and intensity of fluorescence signals at laser-microirradiated sites were measured using a Zen 2009 software (Carl Zeiss).
PAR-binding assay. Recombinant Rpp29 and Rpp21 human proteins were overexpressed in and purified from Escherichia coli BL21 strains, as previously described 38 . The soluble affinity purified His-tagged Rpp21 and Rpp29 proteins were tested for their ability to bind PAR moieties using the PAR-binding assay, as formerly specified 59 . Briefly, 1-5 pmol of recombinant Rpp29 and Rpp21 proteins were blotted onto a nitrocellulose membrane and blocked with TBST buffer supplemented with 5% milk. Radioactively labeled PAR moieties were produced by auto-modified PARP1 prepared by in vitro PARylation reaction. This reaction was carried out at room temperature for 20 min in a reaction buffer (50 mM Tris-HCl, pH 8, 25 mM MgCl 2 , 50 mM NaCl) supplemented with radiolabelled NAD + (Perkin Elmer), activated DNA, and PARP1 enzyme (Trevigen). PAR moieties were detached from PARP1 using proteinase K and blotted membrane was incubated for 2 h with the radiolabelled PAR diluted in TBST buffer. Membranes were then washed with TBST, subjected to autoradiography and Western blotting using α-His and αH3 antibodies. TLR assay. TLR assay was performed as previously described 56,89 . Briefly, U2OS-TLR cells were subjected to two sequential transfections with a time interval of 10 h. Cells were first transfected with siRNAs and then with two plasmids: pRRL-sEF1a HA.NLS.Sce(opt).T2A.IFP, expressing I-SceI fused to infrared fluorescent protein (IFP), and pRRL SFFV d20GFP.T2A.mTagBFP donor plasmid, expressing GFP donor sequence fused to blue fluorescent protein (BFP) 56 . Cells were harvested after 72 h and signals of GFP + (reflecting HDR) and mCherry + (reflecting NHEJ) were measured by four-color fluorescent flow-cytometry using a BD LSRFortessa ™ cell analyser (BD Biosciences). A minimum of 10,000 double-positive (IFP + and BFP + ) cells were scored for each condition from three independent experiments. Results of treated cells were normalized to control cells. HDR values for each condition were normalized for percentage of cells at S and G2 phase monitored by propidium iodide-based standard flow-cytometry.
Scientific RepoRts | 7: 1002 | DOI:10.1038/s41598-017-01185-6 Neutral comet assays. Comet assays were carried out with the Single Cell Gel Electrophoresis Assay-kit (Trevigen). U2OS cells were transfected with control or Rpp29 siRNAs. After 72 h of transfection, cells were exposed to 10 Gys of ionizing radiation from an X-ray machine (Faxitron, CellRad), followed by recovery for 1, 3 and 6 h. Approximately, 3 × 10 3 cells were combined with molten LM Agarose at a ratio of 1:10 and then were applied onto comet slide for 30 min at 4 °C in the dark to solidify. Cells were lysed by incubation with lysis solution for 1 h at 4 °C and slides were run for 30 min at 13 volts. Slides were incubated for 30 min at RT with DNA precipitation solution and fixed with 70% ethanol for 30 min at RT. Samples were dried by incubation for 30 min at RT and stained with SYBR-green. Comet images were acquired using Nikon Eclipsed E400 Epi-fluorescence microscope and percentage of DNA in the comet tail was measured using commercially available COMET SCORETM (TriTeK Corporation) software.
Transcription and processing of tRNA. Whole cell extracts (S20; 25 µl) were assayed for coupled transcription and processing of tRNA Arg , as previously described [48][49][50] . Transcription reactions (55-µl in volume) contained transcription buffer, rNTP and 10 µCi of [α-32 P]UTP. In parallel, extracts were assayed for RNase P activity in processing of the 5′ leader of an internally 32 P-labeled S. pombe precursor tRNA Ser (pSupS1) or E. coli precursor tRNA Tyr in 1 × MRP/TNET buffer containing 0.3 units of rRNasin, 150 ng of cold precursor tRNA and 10 mM DTT, as previously described [48][49][50] . Labeled RNAs were separated in denaturing 8% polyacrylamide/7 M urea gels that were then dried and exposed to autoradiography. Quantitation of the values of the RNA bands was done by the use of the EZQuant-Gel software for densitometric quantitation.
Statistical analysis. Statistical analyses were performed using the demo version of GRAPHPAD prism software version.