NBS1 I171V variant underlies individual differences in chromosomal radiosensitivity within human populations

Genetic information is protected against a variety of genotoxins including ionizing radiation (IR) through the DNA double-strand break (DSB) repair machinery. Genome-wide association studies and clinical sequencing of cancer patients have suggested that a number of variants in the DNA DSB repair genes might underlie individual differences in chromosomal radiosensitivity within human populations. However, the number of established variants that directly affect radiosensitivity is still limited. In this study, we performed whole-exome sequencing of 29 Japanese ovarian cancer patients and detected the NBS1 I171V variant, which is estimated to exist at a rate of approximately 0.15% in healthy human populations, in one patient. To clarify whether this variant indeed contributes to chromosomal radiosensitivity, we generated NBS1 I171V variant homozygous knock-in HCT116 cells and mice using the CRISPR/Cas9 system. Radiation-induced micronucleus formation and chromosomal aberration frequency were significantly increased in both HCT116 cells and mouse embryonic fibroblasts (MEFs) with knock-in of the NBS1 I171V variant compared with the levels in wild-type cells. These results suggested that the NBS1 I171V variant might be a genetic factor underlying individual differences in chromosomal radiosensitivity.

genetic variants of interest. However, chromosomal radiosensitivity might be influenced by many confounding factors, such as age, smoking, environment, and other genetic background 7 . It is therefore necessary to evaluate the genetic variants underlying individual differences in chromosomal radiosensitivity in a cell line with a uniform genetic background.
We previously reported genome editing technology-mediated candidate variant knock-in in human cultured cells with a uniform genetic background to confirm that heterozygous ATM mutations, which cause a rare autosomal-recessive disease, ataxia-telangiectasia [A-T; Online Mendelian Inheritance in Man (OMIM): 067585], indeed underlie chromosomal radiosensitivity 8 . Since heterozygous carriers of most hyper-radiosensitive recessive disorders such as A-T exist at a rate of ~ 1% in human populations 9 , it was suggested that the heterozygous mutations might also determine individual differences in chromosomal radiosensitivity within human populations.
The association between increased chromosomal radiosensitivity and sporadic breast cancer risk has been reported 6 , and a number of DNA DSB repair genes including BRCA1, BRCA2, ATM, CHEK2, PALB2, TP53, Rad51C, Rad51D, and NBS1 have been identified as being causative of hereditary breast and ovarian cancer (HBOC) 10 . On the basis of this background, multi-gene panel testing of the DNA repair genes is clinically used as a tool for screening breast or ovarian cancer patients. In this study, we searched for variants in DNA DSB repair genes using whole-exome sequencing of 29 Japanese ovarian cancer patients. The heterozygous NBS1 c.511A>G, p.Ile171Val variant was identified as a unique candidate underlying chromosomal radiosensitivity. It was reported that a Japanese girl with a homozygous variant of NBS1 I171V developed idiopathic aplastic anemia, but did not show the typical features of hyper-radiosensitive and cancer-prone Nijmegen breakage syndrome [OMIM: 602667.0007] 11 . Meta-analysis based on 60 publications with ~ 40,000 cancer cases and ~ 65,000 controls demonstrated that the NBS1 I171V variant is associated with a significant increase in overall cancer risk 12 , while several epidemiological studies did not show any contribution of this variant to carcinogenesis [13][14][15][16] . Given these conflicting interpretations of pathogenicity, we decided to analyze the exact functional impact of this variant on chromosomal radiosensitivity in a uniform genetic background. Here, we generated NBS1 variant knock-in human cultured cells and MEFs using genome editing technology. Semiautomated CBMN and chromosome aberration analyses in both genome-edited human and mouse cells could quantify the effect of this variant on chromosomal radiosensitivity. These findings could provide a unique insight into the genetic basis underlying the heterogeneity of chromosomal radiosensitivity within human populations.

Results
Identification of NBS1 I171V variant in a Japanese ovarian cancer patient and generation of knock-in HCT116 cell clones using genome editing technology. To screen genetic variants underlying chromosomal radiosensitivity, we performed whole-exome sequencing of genomic DNA from the peripheral blood cells of 29 Japanese ovarian cancer patients. Average coverage for the exons was more than 100 ×. Since most of the mutations detected in HBOC patients are located in the DNA repair genes, we first extracted genetic variants in the top 10 HBOC genes, namely, BRCA1, BRCA2, ATM, CDH1, CHEK2, PALB2, TP53, Rad51C, Rad51D, and NBS1 10 . We also narrowed down the candidate variants on the basis of filtering criteria consisting of the ClinVar database evaluation, genomic position, function, and zygosity. As expected, of these patients, five cases harbored heterozygous mutations of either BRCA1 or BRCA2 (Table 1). In addition, heterozygous ATM missense variants (rs551411717 and rs587782298) in two ovarian cancer patients and a heterozygous TP53 missense variant (rs201382018) in two patients, which might be causative mutations of ataxia-telangiectasia (A-T, OMIM: 607585) and Li-Fraumeni syndrome (OMIM: 191170), respectively, were detected (Table 2). We also identified heterozygous variant c.511A>G, p.Ile171Val, in the NBS1 gene (NM: 002485.5, OMIM: 602667.0007), which encodes a component of the MRE11-RAD50-NBS1 (MRN) complex sensing DNA DSB sites for appropriate repair 17 , in one patient ( Table 2, Fig. 1a,b). The NBS1 c.511A>G, p.Ile171Val variant was predicted to be "disease-causing" and "probably damaging" by MutationTaster and Polyphen2, respectively. In addition, it was previously reported that a Japanese homozygote of this variant showed aplastic anemia rather than the typical features of Nijmegen breakage syndrome, such as severe microcephaly, immunodeficiency, and cancer predisposition 11 . These findings suggested that the NBS1 c.511A>G, p.Ile171Val variant might be involved in individual differences in chromosomal radiosensitivity within human populations.
To demonstrate that the NBS1 c.511A>G, p.Ile171Val variant indeed underlies chromosomal radiosensitivity, we attempted to generate cultured human cells with knock-in of this variant along with a uniform genetic www.nature.com/scientificreports/ background using the CRISPR/Cas9 system. We constructed a plasmid vector expressing both Cas9 protein and single guide RNA (sgRNA) for cutting NBS1 exon 5 including c.511A>G, p.Ile171Val (Fig. 1c). The plasmid vector (px459 provided by Addgene) contained the Cas9 gene and a puromycin resistance gene separated by a 2A peptide sequence, expressing the discrete proteins from a single open reading frame. As a targeting donor, we chemically synthesized single-strand oligonucleotides (ssODNs) harboring the NBS1 c.511A>G, p.Ile171Val variant and putative silent CRISPR/Cas9 blocking mutations in the PAM and sgRNA sequences (Fig. 1c). These silent mutations also functioned as an ScaI site for checking ssODN knock-in easily (Fig. 1c). We also designed ssODNs carrying the silent mutations alone in order to evaluate their effect on chromosomal radiosensitivity (Fig. 1c). Since the human colon cancer cell line HCT116 has two copies of the NBS1 allele and relatively high efficacy of ssODN knock-in 18,19 , we transfected both the Cas9-2A-puromycin resistance gene plasmid vector and the ssODN targeting donors into HCT116 cells. After transient puromycin selection for 48 h post-transfection and subsequent culture for 2 weeks, the drug-resistant colonies were isolated, and their genotypes were analyzed by ScaI digestion and direct sequencing of the PCR amplicon of the target locus. Finally, two clones of the biallelic NBS1 c.511A>G, p.Ile171Val knock-in cells (NBS1 I171V/I171V cells) and one clone of the biallelic wild-type NBS1 with silent mutation knock-in cells (NBS1 +/+ cells) were generated ( Fig. 1d-f). We also established two clones of the NBS1 null HCT116 cells (NBS1 −/− cells) with c.514_514insG, p.Val172fs and c.660_6678del, p.Gln220fs mutations. Western blotting analysis revealed that the NBS1 −/− clones had no signal of NBS1 protein, while the NBS1 I171V/I171V clones showed almost the same amounts of NBS1 protein as the NBS1 +/+ parental HCT116 cells and NBS1 +/+ cell clone, suggesting that the NBS1 c.511A>G, p.Ile171Val variant is not involved in the stability of NBS1 protein (Fig. 2a). It was reported that over extended culturing of HCT116 cells caused mutations in the MRE11 promoter thereby causing loss of expression of MRE11 20 . These generated clones showed almost the same amounts of MRE11 protein and mRNA level (Fig. 2a,b). Thus, CRISPR/Cas9 system-mediated knock-in technology in the HCT116 cell line enabled the generation of an experimental system for comparing the biological effects among the NBS1 variants with a uniform genetic background. www.nature.com/scientificreports/ Chromosomal radiosensitivity is more enhanced in NBS1 I171V/I171V -edited HCT116 cell clones than in NBS1 +/+ HCT116 cells. To investigate whether the NBS1 c.511A>G, p.Ile171Val variant affects the cellular lethality after γ-ray irradiation, we performed the colony survival assay of the set of NBS1 genome-edited cell clones. Consistent with previous reports 21, 22 , the NBS1 −/− clone showed much higher lethal radiosensitivity than the others. The NBS1 I171V/I171V cell clone did not show significant cellular lethality after less than 2 Gy of γ-ray irradiation in comparison with the NBS1 +/+ cell line (Fig. 2c, S3a-c, Table S1), while their lethal radiosensitivity was clearly segregated from those of the NBS1 +/+ cell line in the higher dose exposure. These results implied that the NBS1 c.511A>G, p.Ile171Val variant might contribute to cellular radiosensitivity. Next, to quantify effect of the NBS1 c.511A>G, p.Ile171Val variant on radiosensitivity at the chromosomal level, radiation-induced micronuclei (MN) in the binucleated (BN) cells were measured ( Fig. 2d-f). We evaluated the average ratio of MN to BN cells scored on more than 1000 BN images acquired from Metafer system as a semiautomatic approach 8 , in three independent conditions along with the standard error for each point and create representative dose-response calibration curves. The NBS1 −/− cell clones showed extreme IR-induced micronucleus formation, while the NBS1 I171V/I171V cell clones demonstrated a highly radiosensitive phenotype in comparison with the NBS1 +/+ parental HCT116 cell line and NBS1 +/+ cell clone with silent mutations (Fig. 2g, S3d-f, Table S2). To compare the chromosomal radiosensitivity among these cell clones in a more quantitative manner, we used a linear-quadratic model (MN frequency = c + βD + αD 2 , D: exposure dose) to analyze the dose-response curves of the ratio of MN/BN cells among the NBS1-edited HCT116 cell clones. α, β, and c coefficients were scored by the chromosomal aberration calculation (Cabas) software version 2.0 (http:// www. pu. kielce. pl/ ibiol/ cabas) 23 . Since a linear-quadratic model is converted to (MN frequency-c)/D = αD + β, the sum of α and β coefficients at D = 1 Gy accurately represents the radiosensitivity of cells to γ-ray irradiation 24 . The mean data scores of the sum of α and β (D = 1 Gy) from the NBS1-edited HCT116 cell clones were obtained ( Table 3), indicating that biallelic NBS1 null mutations and the c.511A>G, p.Ile171Val variant contributed to approximately 2.0-and 1.3-fold increases of chromosomal radiosensitivity, respectively, in the HCT116 cell genetic background.
To confirm the radiosensitivity of the NBS1 c.511A>G, p.Ile171Val variant at the chromosomal level more directly, we stained the chromosomes with Giemsa dye in the γ-ray-irradiated NBS1 edited cell lines, and then measured the chromosomal aberrations including chromatid and chromosome gaps and breaks 25 . Consistent with the results of the semiautomatic CBMN assay, the NBS1 −/− cells showed the highest ratio of radiationinduced chromosomal aberrations, while the NBS1 I171V/I171V cells demonstrated an incidence of them intermediate between those of the NBS1 −/− cells and the NBS1 +/+ cell clones ( Fig. 2h- Table S3). These findings revealed that the NBS1 c.511A>G, p.Ile171Val variant has potency for enhancing chromosomal radiosensitivity.
Chromosomal radiosensitivity increases in a manner dependent on the copy number of Nbs1 I171V variant. We attempted to generate heterozygous NBS1 I171V/+ HCT116 cell clones using the genome editing method mentioned above. However, the candidate cell clones obtained were all genetically mosaics consisting of NBS1 +/+ and NBS1 I171V/ I171V cells, and no heterozygous NBS1 I171V/+ single-cell clones were isolated.
Using genome editing technology in mouse fertilized embryos, it is possible to establish heterozygous cell clones for the ssODN knock-in allele. We thus attempted to establish Nbs1 I171V knock-in mice by co-electroporation of CRISPR/Cas9 RNP, which constitutes recombinant Cas9 protein, chemically synthesized crRNA and tracRNA, and ssODN containing the variant, into mouse fertilized eggs (Fig S1a). Two mouse lines were generated from distinct electroporated embryos, in which the Nbs1 p.Ile171Val variant was introduced into the genome. Sanger sequencing confirmed correct knock-in of the variant in both lines. Since the two mouse lines did not exhibit different phenotypes, we mainly addressed data obtained from one line.
No obvious phenotypes were observed in heterozygous mutant mice up to 1 year of age. When they were inbred, wild-type (Nbs1 +/+ ), heterozygous (Nbs1 I171V/+ ), and homozygous (Nbs1 I171V/I171V ) mice were born at the expected Mendelian ratio. Nbs1 I171V/I171V mice developed and grew normally in the laboratory environment and showed normal reproductive ability, and did not show any hematological profile such as aplastic anemia, which was previously reported in a human homozygote of the NBS1 c.511A>G, p.Ile171Val variant 11 (Fig. S2).
To compare chromosomal radiosensitivity among the Nbs1 +/+ , Nbs1 I171V/+ , and Nbs1 I171V/I171V mice, we intercrossed the Nbs1 I171V/+ mice to generate MEFs with each Nbs1 genotype (Fig S1b-d). Western blotting analysis revealed that both Nbs1 I171V/+ MEFs and Nbs1 I171V/I171V MEFs showed almost the same amount of Nbs1 protein as the Nbs1 +/+ MEFs (Fig. 3a). Next, to quantify the effect of the copy number of the Nbs1 I171V allele on chromosomal radiosensitivity, we measured the MN frequencies post-IR in these MEF lines using the semiautomatic CBMN assay (Fig. 3b-d). The Nbs1 I171V/I171V MEFs showed the highest rate of IR-induced micronucleus formation, and the Nbs1 I171V/+ MEFs demonstrated a moderately radiosensitive phenotype in comparison with the Nbs1 +/+ clones (Fig. 3e, S4a-c, Table S4). We also used Cabas software to evaluate the dose-response curves of the Nbs1 mutant MEFs ( Table 4). The mean data scores (D = 1 Gy) suggested that one copy of the Nbs1 I171V allele contributed to an approximately 1.1-fold increase and two copies an approximately 1.7-fold increase of radiosensitivity compared with that of the biallelic wild-type Nbs1. Cytogenetic analysis also demonstrated that IR-induced chromosomal aberrations increased in a manner dependent on the copy number of the Nbs1 I171V allele (Fig. 3f-i, S4d-f, Table S5). Taken together, these findings suggest that the NBS1 c.511A>G, p.Ile171Val variant might be a genetic factor underlying individual differences in chromosomal radiosensitivity within human populations.

Discussion
NBS1 encodes the NBS1 protein, which consists of 754 amino acids and plays a central role in DNA DSB repair as a member of the MRN protein complex, which recognizes DNA DSB sites and orchestrates the DNA damage response 26  www.nature.com/scientificreports/ domains (a forkhead-associated domain and two BRCA1 C-terminal domains) and is recruited to DNA DSB sites. The NBS1 I171V variant, which was identified in this study, is located in the BRCT1 domain (Fig. 1a) 17 . It was reported that the overexpression of the NBS1 I171V protein in HeLa cells interfered with phosphorylated MDC1 and inhibited IR-induced formation of foci of NBS1 27 . Although overexpression analyses of the mutant protein are quick and informative, it is more precise to evaluate its effect on chromosomal radiosensitivity under the endogenous and physiological expression level. Consistent with previous overexpression studies, the endogenous promoter driven NBS1 I171V protein in HCT116 cells and MEFs also interfered chromosomal stability and cellular survival post IR irradiation. These results suggested that the NBS1 I171V variant might be involved in individual radiosensitivity within human populations. However, the epidemiological cancer risk of this variant remains controversial. Therefore, we attempted to evaluate the precise effect of this variant on chromosomal radiosensitivity in a uniform genetic background.
In the presence of conflicting forward genetics data, it is difficult to evaluate whether the variants of interest are indeed involved in chromosomal radiosensitivity. The number of similar situations is increasing with the recent clinical sequencing using deep-sequencers for genetic diagnosis 28,29 . There are several criteria for evaluating the pathological significance of variants. We propose that a reverse genetics approach could be useful because this directly investigates the causality. In this study, we used the CRISPR/Cas9 system to introduce the NBS1 I171V variant into human cultured HCT116 cells and mouse fertilized eggs. The biallelic NBS1 I171V variant in human and mouse genomes significantly enhanced IR-induced micronucleus formation and chromosomal aberrations, clarifying that this variant underlies chromosomal instability after IR irradiation.
In general, simple knock-out using the CRISPR/Cas9 system can be applied in any proliferative human cultured cell lines. Previously, we generated ATM heterozygous knock-out hTERT-RPE1 cells derived from normal retinal tissues to demonstrate that the ATM heterozygous mutations affect cellular radiosensitivity 8 . The activity of homologous directed recombination (HDR) repair in human cultured cells from normal tissues is generally insufficient for ssODN-mediated knock-in. Therefore, here we used HCT116 cells with sufficient HDR activity and a diploid karyotype from colon cancer tissue. As expected, we developed two independent HCT116 cell clones with the biallelic NBS1 I171V variant. However, NBS1 I171V/+ cells were not obtained because the ssODNnon-targeted allele contained indel mutations in the CRISPR/Cas9 system-recognition site; however, mosaic colonies of NBS1 +/+ cells and NBS1 I171V/ I171V cells were formed. To control the copy number of the NBS1 I171V variant in the genome and to confirm that the biological effects of the variant on radiosensitivity are conserved among species, we generated mice heterozygous for the Nbs1 I171V variant. Intercrossing of their heterozygous mice enabled MEFs with three copy numbers of the variant to be obtained. Notably, IR-induced chromosomal instabilities in the MEFs increased in a manner dependent on the copy number of the Nbs1 I171V variant. Unlike a human NBS1 I171V homozygote, the homozygous mice did not show aplastic anemia. Although heterozygotes of NBS1 I171V are estimated to exist at a rate of approximately 0.15% in human populations (Table 2), a second homozygote has not been reported, implying that the existence of other gene mutations underlying aplastic anemia in the reported homozygote cannot be completely ruled out. Importantly, NBS1 171V did not increase the spontaneous frequencies of MN in HCT116 cells and MEFs in the spite of NBS1 I171V carriers have significantly increased risk for multiple cancers even they are not radiated. Since NBS1 participates is not only DNA DSB repairs but also other DNA damage responses including Rad18/Polη-dependent translation DNA synthesis 30 and RNF20-meditaed chromatin remodeling 31 , NBS1 I171V might influence the genomic integrity against a variety of genotoxins to enhance the spontaneous carcinogenesis risk the NBS1 I171V carriers. Further investigations are needed to confirm whether the heterozygous and homozygous Nbs1 I171V exhibit a predisposition for cancer with or without IR irradiation, and to elucidate how NBS1 I171V protein interferes the DNA damage response after IR irradiations. Taking the above findings together, it is safe to conclude that the NBS1 I171V variant is a genetic factor underlying individual differences of radiation-induced chromosomal instability at the cellular level.
A series of radiation biological and epidemiological studies has suggested that many variants in the DNA repair genes might contribute to the existence of a portion of radiosensitive populations 32 . Besides the NBS1 I171V variant, here we also extracted several suspicious variants in the ATM and TP53 genes from Japanese ovarian cancer patients. Whether they are merely polymorphisms or indeed determinants of IR-induced chromosomal instability and carcinogenesis is a key question for precision medicine in the fields of radiology and Full-length gel is presented in Supplementary Fig. 5. GAPDH antibody was used as a loading control. The intensity of NBS1 and MRE11 bands was normalized to that of GAPDH and shown as a percentage, regarding the score of NBS1 +/+ parent cell clones as 100%. (b) RT-PCR data showing mRNA expression levels of MRE11 in NBS1 I171V edited HCT116 cell clones. GAPDH was used as a control. (mean ± SEM; no significant change in each t-test parameter; n = 4) (c) Survival fractions for NBS1 I171V edited HCT116 cell clones for 11 days after irradiation (mean ± SD based on averages from triplicate samples; t-test; n = 3).  www.nature.com/scientificreports/ oncology. Indeed, in breast cancer patients with variants such as in BRCA1, BRCA2, and ATM, it is recommended that the contralateral breast dose in mammography and radiotherapy be minimized in order to prevent secondary carcinogenesis 33 . Since chromothripsis, which is massive and clustered genomic rearrangements, occurs in micronuclei to drive cancer evolution 34,35 , the frequent radiation-induced micronucleus formation in the carriers with NBS1 I171V and other DNA repair gene variants might enhance the risk of secondary carcinogenesis. Thus, knowledge of the genetic basis underlying individual differences in chromosomal radiosensitivity might allow us to provide safer approaches in the context of CT imaging and radiotherapy. Although the reverse genetics approach used in this study is useful for precision medicine, the throughput should be improved. Recent advances of CRISPR/Cas systems provide the base-editor technology based on several cytidine deaminases and their chemical evolved products fused to catalytically dead Cas9 (dCas9), thereby converting the base pair at the specific site directly 36 . Prime editing technology based on dCas9 fused with an engineered reverse transcriptase is another potential approach of base pair conversion 37 . If the base-editor and prime editing technologies are applied to the human haploid cell line HAP1 derived from chronic myeloid leukemia 38 , the higher-throughput validation of single-nucleotide variants by reverse genetics might provide convincing genetic markers for personal radiological protection and accurate genetic diagnosis.

Materials and methods
Whole-exome analysis. This study was reviewed and approved by the Ethics Committee of Hiroshima University Hospital, and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from the participants. Genomic DNA was extracted from the peripheral lymphocytes of ovarian cancer patients and exonic DNA was captured using Agilent SureSelect Human All Exon V5 kit. Sequencing was performed with 150-bp paired-end reads on a HiSeq2500 (Illumina). We used BWA (http:// bio-bwa. sourc eforge. net/) for alignment and mapping, Samtools (http:// samto ols. sourc eforge. net/) and Picard (http:// broad insti tute. github. io/ picard/) for SAM/BAM handling, GATK (http:// www. broad insti tute. org/ gatk/) and Samtools for variant calls, and Annovar (http:// annov ar. openb ioinf ormat ics. org/) for annotation, as described previously. Functional predictions associated with amino acid changes were performed using PolyPhen-2 (http:// genet ics. bwh. harva rd. edu/ pph2) and MutationTaster (http:// www. mutat ionta ster. org/ index. html). All reported genomic coordinates were in the Genome Reference Consortium Human build 37 (GRCh37/hg19) assembly. PCR amplification followed by Sanger sequencing with an Applied Biosystems 3130 sequencer (ThermoFisher) was used to validate mutations identified by whole-exome sequencing.
Cell cultures. HCT116 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and gentamycin at 37 °C in humidified air with 5% CO 2 . Primary MEFs were cultured in DMEM supplemented with 20% FBS and penicillin/streptomycin at 37 °C in humidified air with 5% CO 2 .

Animals and generation of MEFs. This study was reviewed and approved by the Ethics Committee for
Experimental Animals of Hiroshima University, and all animals were treated in accordance with the guidelines of the Institutional Animal Care and Use Committee. The E13.5 embryos were thoroughly minced with razor or scalpel blades. The cells spun down at 500×g for 5 min were incubated in DMEM supplemented with 20% FBS and penicillin/streptomycin at 37 °C in humidified air with 5% CO 2 . After 24 h, a resuspension of the 2.5% trypsin/EDTA (Gibco)-treated cells was filtrated through a 40 µm cell strainer and subsequently centrifuged at 500×g for 5 min. The cellular pellet was resuspended in DMEM supplemented with 20% FBS and penicillin/ streptomycin and cultured in 60-mm culture dishes.  Table 3. Radiosensitivity coefficients (α, β, c, and α + β) in NBS1 I171V variant knock-in HCT116 cells. α, β, and c coefficients were extracted from dose-response calibration curves in the CBMN assay using Cabas software. Relative capacity to repair DNA after acute γ-irradiation was assessed at a dose of 1 Gy. www.nature.com/scientificreports/ Heterozygous knock-in mice were then intercrossed to produce homozygous mice. All animal methods were performed in accordance with the ARRIVE guidelines.

Generation of
Western blotting analysis. Cells were resuspended in lysis buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 2 mM EDTA, 0.1% SDS, 1% NP-40, 0.1% sodium deoxycholate, protease inhibitors, and 10 mM NaF) and incubated on ice for 20 min. Samples were centrifuged at 8000×g for 20 min and the supernatants were collected. Protein concentration in the supernatant was determined by Bradford protein assay. Cell lysates were then boiled with sample buffer (60 mM Tris-HCl pH 6.8, 1% SDS, 5% β-mercaptoethanol) for 5 min. Proteins were separated via SDS-PAGE and transferred to a PVDF membrane. The membrane was blocked with 5% skim milk in TBS-T (0.05% Tween-20) for 30 min. Primary antibodies were diluted in blocking buffer and the membrane was incubated with diluted antibodies overnight at 4 °C. HPR-conjugated anti-rabbit and anti-mouse IgG (#NA934V and #NA931V, respectively; GE Healthcare) were used as secondary antibodies. Images were acquired using ImageQuant LAS4000mini system (GE Healthcare). Quantitation of the levels of NBS1, MRE11 and GAPDH was performed by ImageQuant TL 7.0 software version 8.1 (GE Healthcare). The primary antibodies used were anti-NBS1 mouse monoclonal antibody (GTX70222, GeneTex and MAB15731, R&D), anti-MRE11 rabbit polyclonal antibody (ab33125, Abcam) and anti-GAPDH mouse monoclonal antibody (sc-32233, Santa Cruz Biotechnology).

RT-PCR.
RT-PCR analysis was carried out as described previously 39 . Briefly, HCT116 genome editing cells in 96-well plates were cultured in DMEM supplemented with 10% FBS for 24 h. The cells were washed with ice-cold PBS, lysed, and then, cDNAs were synthesized with reverse transcriptase from their extracted total RNA using Cells to CT Kit (ThermoFisher), in accordance with the manufacturer's protocol. Expression was measured by TaqMan PCR analysis using MRE11 and GAPDH TaqMan probes (ThermoFisher) on a CFX Connect™ Real-Time PCR Detection System (BioRad); the levels of MRE11 expression were normalized to those of GAPDH. The primers and TaqMan probes were MRE11 (Hs00967437_m1) and GAPDH (Hs99999905_g1). www.nature.com/scientificreports/ Gamma irradiation. Cells were irradiated from 1 to 4 Gy ( 137 Cs gamma ray source, 148 TBq, Gammacell 40 Exactor; Best Theratronics). The dose rate used was around 1 Gy/min.

Semiautomatic cytochalasin-block micronucleus (CBMN) assay.
Cells were seeded in six-well plates, incubated for 4 h, and then irradiated. After 44 h of incubation, 3 μg/ml cytochalasin B (WAKO) was added to block cytokinesis for 24 h. Cells were harvested and centrifuged. The pellet was treated with hypotonic solution (75 mM KCl) for 15 min at 37 °C and subsequently rinsed in Carnoy solution (methanol: acetic acid, 3:1) three times for fixation. The slides were prepared using HANABI Metaphase Spreader (ADSTEC, Japan) and stained with 1 μg/ml 4′,6-diamidino-2-phenylindole (DAPI). These cells were scanned at 10 × magnification with a Metafer 4 Scanning System comprising a Carl Zeiss Axioplan Imager Z1 connected to Metafer 4 software version 3.11.4 (MetaSystems GmbH, Altlussheim Germany). Captured images were analyzed with the Metafer 4_MN program (MetaSystems). Criteria for selecting BN cells and MN were as previously described 8 .
All scanned images of BN cells and MN were re-evaluated visually to exclude false-positive and -negative images of BN cells and for MN to be scored in at least 1000 BN cells from each condition.
Survival assay. An appropriate number of cells were irradiated and then seeded in three 100-mm dishes per condition. After 11 days of incubation, cell colonies were fixed with 100% methanol at room temperature for 2 min and stained with 10% Giemsa (Wako) for 2 min. Survival curves were obtained by counting the number of cell colonies.
Chromosomal aberration analysis. Cells were seeded in six-well plates, incubated for 4 h (HCT116 cells) or 6 h (MEFs), and then irradiated. After 24 h of incubation, colcemid (0.1 mg/ml, Gibco) was added to induce arrest in metaphase for 1 h (HCT116 cells) or 6 h (MEFs). The cells were then treated with a hypotonic solution (0.075 M KCl) for 15 min at 37 °C and subsequently fixed with Carnoy solution. Slides were prepared using HANABI Metaphase Spreader (ADSTEC, Japan) and then stained with 10% Giemsa solution (Wako) for 5 min. Chromosome images were scanned at 10 × and 63 × magnifications with Metafer 4 software (MetaSystems, Germany) connected to a motorized ZEISS AxioImager M1 (Zeiss, Germany). Chromosome gaps and breakages were scored visually by fewer than 25 metaphases per slide of well-spread metaphase chromosomes. Fifty metaphase chromosomes from each condition were examined for spontaneous chromosomal aberrations.

Statistical analysis.
The experiments were performed independently three times, and the data are shown as mean ± s.e. Differences between groups were evaluated for statistical significance using Student's t-test. Values of p < 0.05 were considered to be statistically significant.