ARTICLE Corrected: Publisher correction DOI: 10.1038/s41467-017-01051-z OPEN ID3 regulates the MDC1-mediated DNA damage response in order to maintain genome stability 1,2 1,3 1 1 1 1,2 Jung-Hee Lee , Seon-Joo Park , Gurusamy Hariharasudhan , Min-Ji Kim , Sung Mi Jung , Seo-Yeon Jeong , 4 5 5 6 6 1,7 In-Youb Chang , Cheolhee Kim , Eunae Kim , Jihyeon Yu , Sangsu Bae & Ho Jin You MDC1 plays a critical role in the DNA damage response (DDR) by interacting directly with several factors including γ-H2AX. However, the mechanism by which MDC1 is recruited to damaged sites remains elusive. Here, we show that MDC1 interacts with a helix–loop–helix (HLH)-containing protein called inhibitor of DNA-binding 3 (ID3). In response to double- strand breaks (DSBs) in the genome, ATM phosphorylates ID3 at serine 65 within the HLH motif, and this modiﬁcation allows a direct interaction with MDC1. Moreover, depletion of ID3 results in impaired formation of ionizing radiation (IR)-induced MDC1 foci, suppression of γ- H2AX-bound MDC1, impaired DSB repair, cellular hypersensitivity to IR, and genomic instability. Disruption of the MDC1–ID3 interaction prevents accumulation of MDC1 at sites of DSBs and suppresses DSB repair. Thus, our study uncovers an ID3-dependent mechanism of recruitment of MDC1 to DNA damage sites and suggests that the ID3–MDC1 interaction is crucial for DDR. Laboratory of Genomic Instability and Cancer Therapeutics, Cancer Mutation Research Center, Chosun University School of medicine, Gwangju 501-759, Republic of Korea. Department of Cellular and Molecular Medicine, Chosun University School of medicine, Gwangju 501-759, Republic of Korea. 3 4 Department of Premedical Sciences, Chosun University School of medicine, Gwangju 501-759, Republic of Korea. Department of Anatomy, Chosun University School of medicine, Gwangju 501-759, Republic of Korea. College of Pharmacy, Chosun University, 375 Seosuk-dong, Gwangju 501-759, Republic 6 7 of Korea. Department of Chemistry, Hanyang University, Seoul 04763, Republic of Korea. Department of Pharmacology, Chosun University School of medicine, Gwangju 501-759, Republic of Korea. Jung-Hee Lee and Seon-Joo Park contributed equally to this work. Correspondence and requests for materials should be addressed to J.-H.L. (email: email@example.com) or to H.J.Y. (email: firstname.lastname@example.org) NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 1 | | | ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z he integrity of genomic DNA is challenged by genotoxic underlying mechanism of action are only partially understood. In insults that originate from either normal cellular metabo- order to better characterize the regulatory network relevant to Tlism or external sources. To ensure proper maintenance of MDC1 and to gain further insight into the molecular mechanism genomic integrity, eukaryotes have evolved a DNA damage of action of MDC1 in the DDR, a yeast two-hybrid screen was response (DDR) system that senses damage and transduces this performed using a HeLa cDNA plasmid library with the C- information within the cell in order to orchestrate DNA repair, terminal fragment (amino acid 1882–2082) of human MDC1 as 1 7 cell-cycle checkpoints, chromatin remodeling and apoptosis . The the bait. Out of the 2.6 × 10 transformants that were screened, 43 functional importance of DDR in maintaining genomic integrity independent positive clones were isolated. When cDNA from is highlighted by the fact that it is conserved among eukaryotes. each of the positive clones was sequenced, one was identiﬁed as Mutations that disrupt the activity of DDR components con- 53BP1, a protein known to associate with the tandem BRCA1-C tribute directly to tumorigenesis ; therefore, it is important to terminus (tBRCT) domain of MDC1. The other clones, which understand these complex mechanisms at the molecular level to had no previously identiﬁed association with MDC1, encoded further our understanding of cancer progression and treatment. ID3 (NM_002167), PIAS1 (NM_016166), UBE2I (NM_194259), DNA double-strand breaks (DSBs), which are generated through KPNA2 (NM_0022266), ZNF114 (NM_153608), KIFC1 ionizing radiation (IR) and through various DNA-damaging che- (NM_002263), CASP8AP2 (NM_001137667), C2orf44 micals, are the most dangerous DNA lesions, because if they are not (NM_025203), SRSF11 (NM_001190987), TINP1 (NM_014886), efﬁciently and accurately repaired, they can result in mutations, GPRC5C (NM_022036), and WWC1 (NM_015238). Among genomic rearrangements, and cell death, which can lead to these, DNA-binding protein inhibitor ID3 was particularly 1, 2 cancer . The ability of cells to detect and properly repair DSBs is notable because this HLH-containing protein has been shown to therefore essential for maintaining genome stability and preventing activate a DNA repair process and, correspondingly, when ID3 3 19 cancer . Central to the DSB checkpoint response is ATM protein is inactivated, excess DNA damage accumulates . kinase, which, when activated by DSBs, initiates a signaling cascade To verify that an interaction between MDC1 and ID3 occurs in that starts with phosphorylation of the histone variant H2AX (γ- human cells, we used co-immunoprecipitation assays followed by H2AX) at DSB sites, and is followed by recruitment of upstream western blotting to assess protein–protein interactions. As shown 1, 4, 5 factors including MDC1 . MDC1 functions as an assembly in Fig. 1a, b, endogenous MDC1 and ID3 co-immunoprecipitated platform to help localize and maintain signaling and repair factors reciprocally and, although the association occurred in non- at and around DSB sites . In this role, MDC1 ampliﬁes DNA irradiated cells, it was enhanced in response to DNA damage. To damage signals by binding to phosphorylated H2AX and subse- examine this further, we expressed HA-tagged MDC1 and GFP- quently binding and retaining additional DDR factors at sites of tagged ID3 ectopically in HEK293T cells and attempted to DNA damage. The accumulation of these DDR factors at DSB sites immunoprecipitate MDC1 using anti-HA antibody. As shown in is generally believed to facilitate DNA damage repair and check- Fig. 1c, GFP-ID3 was observed in the anti-HA immunoprecipi- point control. Thus, MDC1 has been recognized as the “master tate. Reciprocally, HA-MDC1 was immunoprecipitated together regulator” that modulates a speciﬁc chromatin microenvironment with GFP-ID3 using anti-GFP antibody (Fig. 1d) and these required to maintain genomic stability. interactions were more pronounced following exposure to IR. MDC1-knockout (KO) mice show chromosomal instability, This protein–protein interaction did not require the presence of defects in DSB repair, radiosensitivity, and cancer DNA, as neither ethidium bromide nor DNase affected the co- 7, 8 predisposition . Furthermore, downregulation of MDC1 is immunoprecipitation (Supplementary Fig. 1a). Furthermore, ID3 associated with multiple cellular phenotypes including hyper- appears to be the only member of the ID family of proteins with sensitivity of cells to DSBs, improper activation of the G2/M and which MDC1 interacts, because we were unable to detect any intra-S checkpoints, aberrant activation of DNA damage-induced interaction of MDC1 with ID1 or ID2 (Supplementary Fig. 1b). apoptosis, and inefﬁcient phosphorylation of DDR regulatory To conﬁrm that the tBRCT domain of MDC1 is the location at proteins . It has been suggested that, in addition to its central role which ID3 binds, we examined the interaction between these two 10, 11 in the DDR, MDC1 directly mediates HR and non- proteins using a series of internal deletion mutations of MDC1 homologous end joining (NHEJ) , activation of the decatena- (Supplementary Fig. 2a). We found that when expressed in tion checkpoint , regulation of the DNA replication check- HEK293T cells, all MDC1 mutants co-immunoprecipitated with 14 15 16 point , mitosis , and spindle assembly checkpoint . GFP-ID3 except for a mutant lacking tBRCT domain (amino Clearly, MDC1 is quickly recruited to DNA damage sites, acids 1893–2082) (Supplementary Fig. 2b), indicating that this allowing multiple protein–protein interactions that are crucial for domain is indeed required for binding to ID3. To further analyze proper DDR processes. However, the precise mechanisms by direct physical associations between ID3 and MDC1, a GST which MDC1 is recruited to protect cells from the deleterious fusion to the tBRCT domain of MDC1 was made and used in a effects of DNA damage are not fully understood. The current pull-down assay with IR-treated HeLa cell lysates. Immunoblot- study was initiated with the goal of better understanding how ting of the pull-down samples revealed that the tBRCT domain of MDC1 is recruited to DNA damages sites and how the role of MDC1 interacts with ID3 (Fig. 1e); this polypeptide also pulled MDC1 in DDR is regulated in response to DNA damage. Since a down the exogenously expressed GFP-tagged ID3 (Fig. 1f), tandem BRCA1 C-terminal (tBRCT) domain of MDC1 is conﬁrming that ID3 directly binds to this region of MDC1 essential for recruitment of MDC1 to DNA damage sites ,we in vitro. screen for tBRCT domain of MDC1-associated proteins and identify a helix–loop–helix (HLH) domain-containing protein called inhibitor of DNA-binding 3 (ID3), which we propose ATM-dependent and MDC1-dependent phosphorylation of interacts directly with MDC1 and is a key factor in the interaction ID3. To examine the potential role of ID3 in regulating DDR, we of MDC1 with γ-H2AX, recruiting MDC1 to DSB sites and asked whether IR exposure induced phosphorylation of ID3. regulating DDR function of MDC1. ATM/ATR-like kinases convey signals upstream of the DDR pathway by preferentially phosphorylating substrate proteins on Results either a serine or threonine residue that is typically followed by a MDC1 interacts with ID3. Although the role of mammalian glutamine residue in the so-called serine–glutamine (SQ) or 20, 21 MDC1 in the DDR is well documented, its regulation and threonine–glutamine (TQ) motifs . We identiﬁed two 2 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE putative ATM/ATR phosphorylation motifs in the ID3 amino in this regulatory pathway. In IR-treated HeLa cells, phosphor- acid sequence, threonine 62 (Thr62) and serine 65 (Ser65), both ylation at Ser65 was strongly inhibited by KU55933 but was not located in the conserved HLH domain (Fig. 2a). A sequence affected by NU7026 (Fig. 2d). Because MDC1 associates with ID3, alignment of the human ID family of proteins, including ID1, we looked for a role for MDC1 speciﬁcally in the phosphorylation ID2, and ID3, revealed that ID3 is the only member that possesses of ID3. We compared normal and MDC1-depleted HeLa cells TQ (including Thr62) and SQ (including Ser65) motifs in the and observed that IR-induced ID3 phosphorylation was attenu- Helix2 domain, consistent with our earlier observation that ID3 is ated when MDC1 was expressed at lower levels (Fig. 2e). the only of the three family members relevant to this pathway. To Together, these results suggest that ATM and MDC1 act together test whether DNA damage induces phosphorylation of ID3, we in mediating phosphorylation of ID3 Ser65 in response to DNA immunoprecipitated ID3 from HeLa cells and examined phos- damage in vivo. phorylation levels before and after cells had been exposed to IR. Using antibodies that bind to phosphorylated serine or threonine, we showed that in IR-induced cells, serine was phosphorylated, Recruitment of phosphorylated ID3 to sites of DNA damage. but threonine was not (Fig. 2b). Western blot analysis, using The ﬁnding that ID3 is phosphorylated at residue Ser65 after lysates prepared from both control and irradiated cells and exposure to IR suggests that this form of ID3 may localize to site polyclonal rabbit anit-pSer65–ID3 antibody, conﬁrmed that IR of DNA damage. To identify the cellular location of phosphory- treatment enhanced the phosphorylation of endogenous ID3 at lated ID3, we used immunoﬂuorescence staining with anti- Ser65 (Fig. 2c). Recognition of pSer65–ID3 by this antibody was pSer65–ID3 antibody. As shown in Fig. 3a, in IR-treated HeLa blocked by the presence of phosphorylated peptides but not by cells, foci of pSer65–ID3 were detected, and these foci co-localized non-phosphorylated peptides (Supplementary Fig. 3a) and the with γ-H2AX, an intracellular protein that associates with recognition was inhibited by ID3 siRNA (Supplementary Fig. 3b), damaged DNA. The staining of pSer65–ID3 was speciﬁc, because ﬁndings that establish the speciﬁcity of this antibody. the number of foci decreased in the presence of a competing Next, we explored upstream signaling molecules in order to phosphorylated peptide but not in the presence of a non- identify those that may contribute to DNA damage-induced phosphorylated peptide (Supplementary Fig. 3c). Because both phosphorylation of ID3. ATM kinase and DNA-dependent ATM and MDC1 affect the phosphorylation of ID3, we looked protein kinase (DNA-PK) are largely responsible for initiating for a role for either of these proteins on pSer65–ID3 foci for- and maintaining DNA damage signals after exposure to IR .We mation. In the presence of the ATM inhibitor KU55933, a few IR- therefore used KU55933, an inhibitor of ATM kinase, and induced pSer65–ID3 foci were observed, whereas in the presence NU7026, an ID-PK to assess the importance of these two proteins of NU7026, the DNA-PK inhibitor that had no effect in a a b MDC1 IgG Input ID3 IgG Input IR −+−+ −+ IR −+ −+ −+ ID3 MDC1 IP: MDC1 IP: ID3 17 MDC1 ID3 c d HA IgG input GFP IgG Input IR −+−+ −+ IR −+−+ −+ GFP HA IP: HA IP: GFP HA GFP e f IB: anti-GFP IB: anti-ID3 (ID3) IP: GST IP: GST IB: anti-GST 35 IB: anti-GST (MDC1) (MDC1) Fig. 1 Protein–protein interactions between MDC1 and ID3. a HeLa cells, with or without exposure to IR, were collected after 3 h, and whole-cell lysates were subjected to immunoprecipitation using an anti-MDC1 antibody followed by western blotting using the antibodies indicated to the right of the blot. b HeLa cells were prepared as in a, and lysates were subjected to immunoprecipitation using an anti-ID3 antibody followed by western blotting using the antibodies indicated to the right of the blot. c HA-tagged MDC1 and GFP-tagged ID3 were co-transfected into HEK293T cells and exposed to IR. After 3 h, whole-cell lysates were subjected to immunoprecipitation using an anti-HA antibody followed by western blotting using the antibodies indicated to the right of the blot. d HEK293T cells were prepared as in c, and lysates were subjected to immunoprecipitation using an anti-GFP antibody followed by western blotting using the antibodies indicated to the right of the blot. e, f A GST-tagged fragment of MDC1 or a GST bead alone was incubated with total cell lysates from HeLa cells exposed to IR (e) or from GFP-ID3 transfected HEK293T cells exposed to IR (f). GST pull-downs were immunoblotted with antibodies as indicated. Uncropped blots of this ﬁgure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 3 | | | GST–MDC1 -1893–2082 GST Input GST GST–MDC1 -1893–2082 Input ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z previous experiment, the number of foci was comparable to the observed effects are attributed to the ID3 protein (Fig. 4d, e). control, as expected (Fig. 3b). Likewise, in MDC1-depleted cells MDC1 is required for the retention of additional DDR proteins exposed to IR, accumulation of pSer65–ID3 at DSBs was strongly and is thus considered to be an upstream regulator of this 8, 22, 23 reduced (Fig. 3c, d), supporting the prediction that both MDC1 process . Therefore, we predicted that depletion of ID3 and ATM are required for efﬁcient localization of ID3 under might also affect the localization of several DDR factors to these conditions. nuclear DSB sites. Indeed, accumulation of NBS1, BRCA1, 53BP1, RNF8, and RNF168 at DSBs in response to IR exposure was consistently impaired in both HeLa and U2OS cells depleted ID3 is required for the recruitment of MDC1 to DSBs. The of ID3, whereas γ-H2AX, a protein that acts upstream in the above results, pointing to a biochemical interaction between ID3 DDR pathway, still formed foci (Fig. 4f, g). and MDC1, prompted the prediction that ID3 is required for To further conﬁrm the critical role for ID3 in recruitment of recruitment of MDC1 to DSBs. To test this hypothesis, we MDC1 to nuclear foci after DNA damage, we generated ID3-KO exposed control and ID3 knockdown HeLa and U2OS cells human U2OS cells using the CRISPR/Cas9 genome-editing (Fig. 4a) to IR, and used immunoﬂuorescence staining to detect system (Supplementary Fig. 5a–c). Noticeably, IR-induced MDC1 at various time points. As shown in Fig. 4b, c, in control MDC1 foci were signiﬁcantly reduced in ID3 KO cells than in siRNA-transfected cells, MDC1 foci formed rapidly in response to control cells (Supplementary Fig. 6a). Moreover, ID3 KO IR treatment and the percentage of cells containing a notable impaired the recruitment of NBS1, BRCA1, 53BP1, RNF8, and number of foci increased steadily for up to 1 h. In contrast, cells RNF168, but not γ-H2AX, at DSBs after IR exposure (Supple- transfected with ID3-targeted siRNA had signiﬁcantly fewer mentary Fig. 6b, c). Together, these data suggest that ID3 is MDC1 foci and the percentage of cells with foci remained low required for localization of MDC1 at DSBs thereby facilitating throughout the entire time course. We also observed that HeLa recruitment of downstream factors. cells with a stable knockdown of ID3, created using two different shRNAs, had dramatically less recruitment of MDC1 to DNA damage sites (Supplementary Fig. 4a, b). Similar results were pSer65–ID3 is essential for IR-induced MDC1 foci formation. obtained when DNA damage was induced in HeLa cells using the Phosphorylation of SQ or TQ motifs leads to altered interactions radiomimetic neocarzinostatin (Supplementary Fig. 4c). Intro- between proteins in the DNA repair complex and between other duction of shRNA-resistant ID3 into cells depleted of endogenous checkpoint proteins . Furthermore, proteins with BRCT ID3 restored formation of MDC1 foci, conﬁrming that the domains, like MDC1, have been shown to recognize phospho- ab IR 0 1 3 h Helix1 Loop Helix2 ID HLH p-serine ID1 NKKVSKMEILQH IP: ID3 ID2 NRKVSKVEILQH p-threonine ID3 GTQLSQVEILQR ID3 Homo sapiens (a.a. 61–72) GTQLSQVEILQR Mus musculus (a.a. 61–72) GTQLSQVEILQR β-actin ID3 Rattus norvegicus (a.a. 61–72) GTQLSQVEILQR Input ID3 Bos taurus (a.a. 61–72) GTQLSQVEILQR c d DMSO Ku55933 Nu7026 IR 0 10 20 40 60 180 min IR01 1 3013 0 3h 20 20 p-ID3 p-ID3 ID3 ID3 β-actin β-actin siRNA Control MDC1 IR 01 3 01 3h MDC1 ID3-p ID3 β-actin Fig. 2 ATM- and MDC1-mediated phosphorylation of ID3 at Ser65 in response to IR. a An outline of the HLH domain and amino acid sequences including the TQ/SQ sites for human ID1, ID2, and ID3 is shown at the top of the panel. The bottom of the panel includes an alignment of ID3 HLH from four different species, demonstrating the highly conserved TQ and SQ motifs (highlighted). b HeLa cells, with or without exposure to IR, were collected at the indicated times and whole-cell lysates were subjected to immunoprecipitation using an anti-ID3 antibody followed by western blotting using anti-p-serine, anti-p- threonine, and anti-ID3 antibodies, as indicated to the right of the blot. c HeLa cells, with or without exposure to 2 Gy of IR for the indicated times, were lysed and analyzed by western blotting using anti-pSer65–ID3 and anti-ID3 antibodies. d HeLa cells were pretreated with DMSO, ATM inhibitor KU55933 (10 μM), or DNA-PK inhibitor NU7026 (5 μM) for 1 h, and then treated with or without exposure to IR for the indicated times. Cell lysates were analyzed by western blotting using anti-pSer65–ID3 and anti-ID3 antibodies. e HeLa cells transfected with either control siRNA or MDC1-speciﬁc siRNA, were treated with or without exposure to IR for the indicated times. Cell lysates were analyzed by western blotting using anti-MDC1, anti-pSer65–ID3, and anti-ID3 antibodies. Uncropped blots of this Figure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14 4 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE ** p-ID3 γ-H2AX DAPI Merge IR 0 min 40 ** 30 min 60 min 0 30 60 Minutes post IR (2 Gy) treatment DMSO ns IR 0 0.5 1 3 h Ku55933 ns 30 ** Nu7026 ** p-ID3 ns DAPI ** p-ID3 DAPI 0 0.5 1 3 Hours post IR (2 Gy) treatment p-ID3 c siRNA Ctrl MDC1 MDC1 DAPI ID3 β-actin IR UT 0 10 20 40 60 180 min Control siRNA p-ID3 MDC1 siRNA DAPI ** p-ID3 ** ** ** ** DAPI UT 0 10 20 40 60 180 Minutes post IR (2 Gy) treatment Fig. 3 Phosphorylated ID3 is recruited to sites of DNA damage. a Endogenous phospho-ID3 co-localizes with γ-H2AX after DNA damage. HeLa cells were exposed to 2 Gy of IR and ﬁxed at the indicated time points. Immunohistochemistry was performed using antibodies against pSer65–ID3 and γ-H2AX. Co- localization is visible as yellow staining in the column labeled “merge”. Nuclei were stained with DAPI. The histogram in the right panel shows the percentage of pSer65–ID3 foci that co-localized with γ-H2AX foci. At least 100 cells were analyzed for each treatment (n = 3). Scale bars: 10 μm. b ATM, but not DNA-PK, mediates phospho-ID3 foci formation. HeLa cells were pretreated with DMSO, KU55933 (10 μM), or NU7026 (5 μM), exposed to 2 Gy of IR, and ﬁxed at the indicated time points. Immunostaining experiments were performed using an anti-pSer65–ID3 antibody. The histogram in the right panel shows the number of phospho-ID3 foci per cells (n = 3). Scale bars: 10 μm. c HeLa cells were transiently transfected with either control siRNA or MDC1-speciﬁc siRNA, and the levels of endogenous MDC1 and ID3 were analyzed by western blotting. d HeLa cells transfected with either control siRNA or MDC1-speciﬁc siRNA were exposed to 2 Gy of IR and ﬁxed at the indicated time points. Immunoﬂuorescence was performed using antibodies against pSer65–ID3. The histogram in the right panel shows the number of phospho-ID3 foci per cell (n = 3). Scale bars: 10 μm. Data are presented as means ± s.d. P values are based on Student’s two-tailed t-test: **P < 0.01; ns, not signiﬁcant 25–27 Ser/Thr motifs . Therefore, we predicted that phosphoryla- control, we exchanged Thr62 to alanine (T62A). As shown in tion of Ser65 within the HLH domain of ID3 is important for Fig. 5c, the presence of the S65A mutation impaired interactions modulating the interactions between MDC1 and ID3. To test this, with MDC1, but the T62A mutation did not have a notable affect, we generated GFP-tagged constructs that included the following suggesting that phosphorylation of ID3 at Ser65 is required for its versions of ID3: the HLH domain (amino acids 25–81), a mutant interaction with MDC1. We next mutated Ser65 to aspartic acid lacking the HLH (ΔHLH; deletion of residues 25–81), a mutant (S65E), which mimics the phosphorylated state. This mutant lacking the carboxyl terminus (ΔC; deletion of residues 82–119), bound to MDC1, as predicted (Fig. 5d). Interactions between ID3 and a mutant lacking the amino terminus (ΔN; deletion of resi- and the tBRCT domain of MDC1 were abolished after treatment dues 1–24) (Fig. 5a). ID3 from each of these constructs was co- with phosphatase, supporting the conclusion that the interaction expressed with full-length HA-tagged MDC1 in HEK293T cells. is phosphorylation dependent (Supplementary Fig. 7). As shown in Fig. 5b, when the HLH domain was present, ID3 The BRCT domain of MDC1 is required for the formation of interacted with MDC1, but if the HLH domain was deleted, the foci at sites of DNA damage . Because ID3 interacts with this interaction was impaired. same domain, and because depletion of ID3 decreased the We then generated a point mutation within the HLH domain formation of MDC1 foci, we hypothesized that a direct of ID3, replacing Ser65 with alanine (S65A). As a negative interaction between ID3 and MDC1 is required for MDC1 to NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 5 | | | Nu7026 Ku55933 DMSO p-ID3 foci no. per cells p-ID3 foci MDC1 siRNA Control siRNA co-localization (%) p-ID3 foci no. per cell ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ab IR UT 0 10 20 40 60 180 min HeLa HeLa U2OS MDC1 siRNA Ctrl ID3 Ctrl ID3 17 17 ID3 Control siRNA DAPI ID2 17 17 17 17 ID1 MDC1 MDC1 245 245 ID3 siRNA 100 100 NBS1 DAPI BRCA1 245 245 Control siRNA 53BP1 245 ID3 siRNA 17 17 H2AX 50 RNF8 63 63 RNF168 ** ** β-actin ** ** ** UT 0 10 20 40 60 180 Mins post IR (2 Gy) treatment IR UT 0 10 20 40 60 180 min U2OS ID3 shRNA+ HA-MDC1 MDC1 ID3-GFP Mock WT Control siRNA DAPI d ID3 shRNA+ HA-MDC1 GFP-ID3 Mock WT MDC1 ID3 siRNA GFP DAPI Control siRNA HA-MDC1 ID3 siRNA 245 ** β-actin 30 20 ** ** ** ** ** ID3-GFP Mock WT UT 0 10 20 40 60 180 ID3 shRNA+ HA-MDC1 Minutes post IR (2 Gy) treatment fg HeLa U2OS MDC1 γ-H2AX NBS1 BRCA1 53BP1 RNF8 RNF168 MDC1 γ-H2AX NBS1 BRCA1 53BP1 RNF8 RNF168 Control siRNA Control siRNA ID3 siRNA ID3 siRNA 100 100 Control siRNA Control siRNA ID3 siRNA ID3 siRNA 80 80 ** 60 40 40 ** ** ** ** ** ** ** 20 20 ** ** ** ** MDC1 γ-H2AX NBS1 BRCA1 53BP1 RNF8 RNF168 MDC1 γ-H2AX NBS1 BRCA1 53BP1 RNF8 RNF168 1 h post IR (2 Gy) treatment 1 h post IR (2 Gy) treatment be recruited to sites of DNA damage. We tested this using HeLa recruitment of HA-MDC1 to DSB sites in each cell type was then cells that had been transfected with ID3 shRNA in order to measured. As shown in Fig. 5f, there was signiﬁcantly more deplete the endogenous pool of protein, and then reconstituted recruitment of HA–MDC1 to DSBs in cells expressing HLH ID3 with either an shRNA-resistant HLH domain (HLH ID3) or an when compared to cells expressing ID3-ΔHLH, indicating that HLH deletion mutant (ID3-ΔHLH) (Fig. 5e). The level of the HLH domain of ID3 is indeed involved in recruitment of 6 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | % of cell with MDC1 >10 foci % cells with protein > 5~10 foci % of cells with MDC1 >10 foci % cells with protein > 5~10 foci % of cell with MDC1 > 10 foci DAPI GFP-ID3 HA-MDC1 NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE MDC1. Moreover, immunoﬂuorescence analysis suggested that Fig. 8b). Thus, Arg1933 in the tBRCT domain, which is known when ID3 knockdown cells are reconstituted with either wild-type to be involved in binding to the tail of γ-H2AX , was much more HLH domain of ID3 or the T62A HLH mutant, recruitment of ﬂexible when in a complex with pHLH than it is in tBRCT alone. MDC1 to DSBs occurred normally. But when reconstituted with Consequently, it is likely that the weakness in the salt bridge of the S65A HLH mutant, recruitment was signiﬁcantly decreased the tBRCT–pHLH complex may allow for better capture of the γ- (Fig. 5g, h). On the other hand, reconstitution of ID3 knockdown H2AX pentapeptide. Of note, the tBRCT–γ-H2AX–pHLH trimer cells with the S65E HLH mutant, which mimics the phosphory- contains a substantial salt bridge network with aggregation of lated state, restored MDC1 recruitment to normal levels. These hydrophobic side chains, including Phe1979, Ala2078, and results point to a requirement for a phospho-dependent Phe2079 of tBRCT and Val67 of pHLH (Fig. 6d), indicating that interaction between MDC1 and ID3 in order for proper the stability of the salt bridge increased when γ-H2AX docked localization of MDC1 to sites of DNA damage. with the tBRCT–pHLH complex. Together, these results suggest a molecular mechanism by which the interaction of MDC1–tBRCT with ID3-pHLH increases binding afﬁnity between ID3 controls the interaction between γ-H2AX and MDC1. The MDC1–tBRCT and γ-H2AX. domains of MDC1 identiﬁed in this study as important for interactions with ID3 (Fig. 1) overlap with those previously identiﬁed as important for interactions with γ-H2AX . There- ID3 promotes DSB repair through its interaction with MDC1. fore, we explored whether the lack of ID3 or H2AX would affect To deﬁne a possible role for ID3 in DDR, we ﬁrst investigated interactions between MDC1 and γ-H2AX. We found that in ID3 whether cells lacking ID3 would be more sensitive to DNA knockdown cells, but not H2AX knockdown cells, the association damage. Cell viability, as measured by the clonogenic survival between MDC1 and γ-H2AX in response to IR irradiation was assay, was measured for HeLa cells with and without depletion of abolished (Fig. 6a, b), indicating that the interaction between ID3 through introduction of each of two different shRNAs. MDC1 and γ-H2AX occurs downstream of the ID3–MDC1 Under normal conditions, the ID3 knockdown cells were interaction. equivalent to wild-type cells. However, the cells depleted in ID3 To more closely examine the structural differences in MDC1 had a signiﬁcant decrease in survival in response to IR (Fig. 7a). when interacting with either ID3 or ID3 and γ-H2AX together, Moreover, the survival fractions of colonies following IR were we used molecular dynamic (MD) simulations to compare the much lower in ID3 KO cells than WT cells (Supplementary interactions. We observed that the most stable conformation of Fig. 6d). the MDC1–tBRCT domain, as derived from normal MD We next investigated whether ID3 affects DSBs repair by simulations , is similar to that determined through X-ray measuring γ-H2AX staining and comet tail moments. We found diffraction data, and the salt bridge network including Arg1933, that knockdown of ID3 in HeLa cells had signiﬁcantly more Glu2063, and Thr2067 involved in the binding of γ-H2AX is also residual DSBs than control cells, as evidenced by the increase in 17, 29 in agreement (Supplementary Fig. 8a) . Interestingly, the signal intensity of γ-H2AX staining (Fig. 7b) and by the increase distal C-terminal region of the tBRCT domain of MDC1 contains in comet tail moments (Fig. 7c). Notably, depletion of both ID3 two positively charged lysine residues at positions 2071 (K2071) and MDC1 did not further increase comet tail length from what and 2075 (K2075) that are fully ﬂexible, indicating that they may was observed with ID3 depletion alone (Fig. 7d, e), indicating that be relevant for binding to phospho-ID3. To explore this further, ID3 and MDC1 act in the same pathway. We also conﬁrmed that K2071 and K2075 were sequentially replaced with the neutral the impaired DSB repair was due to an insufﬁcient amount of amino acid methionine in order to create both single-point and ID3, because ID3 knockdown cells could be rescued through the double-point mutants. The K2071M mutant bound to ID3- introduction of shRNA-resistant ID3 (Fig. 7f, g). These results phosphorylated HLH (pHLH) equally as well as wild-type MDC1 suggest that ID3 promotes DSB repair. (Fig. 6c). On the other hand, interactions between the K2076M Because we observed a role for ID3 in DSB repair, we predicted mutant and ID3-pHLH were signiﬁcantly decreased. Intriguingly, that a possible reason for the interaction between ID3 and MDC1 when both lysine residues were mutated, interactions were almost might also be to function in DNA repair. To test this, we stably completely abolished. These results suggest a prominent role for transfected HeLa cells with ID3 shRNA in order to deplete K2075 in interactions with phospho-ID3, and perhaps a endogenous ID3, reconstituted these cells with shRNA-resistant supportive role for K2071. ID3-HLH or ID3-ΔHLH (Fig. 7h), and then measured the We then performed a docking simulation to more closely amount of DSB repair after IR exposure using the neutral comet analyze the interactions between the tBRCT domain of MDC1 assay. As shown in Fig. 7i, comet tails for cells depleted in ID3 or and the pHLH domain of ID3. Our simulation indicated the expressing ID3-ΔHLH were signiﬁcantly longer than for those formation of a new hydrophobic core including Val67 of ID3 as a expressing ID3-HLH, suggesting that efﬁcient DNA repair ligand and Phe1979, Ala2078, and Phe2079 of tBRCT as a required the presence of the HLH domain. Further, we found receptor. The Cβ−methyl group of Thr2067 in tBRCT was located that reconstitution of ID3-depleted cells with the S65E HLH close to this cooperative hydrophobic core (Supplementary mutant, but not the S65A HLH mutant, restored DSB repair Fig. 4 ID3 depletion reduces MDC1 foci formation. a HeLa and U2OS cells were transiently transfected with either control siRNA or ID3-speciﬁc siRNA. western blotting using antibodies to the indicated proteins shows the expression levels of each. b, c Control or ID3-depleted HeLa (b) and U2OS (c) cells were exposed to 2 Gy of IR and ﬁxed at the indicated time points. Immunostaining experiments were performed using an anti-MDC1 antibody. Nuclei were stained with DAPI. Representative images (upper panel) and quantiﬁcation (lower panel) of MDC1 foci in control and ID3-depleted cells (n = 3). Scale bars: 10 μm. d, e A stable knockdown of ID3 in HeLa cells co-expressing shRNA-resistant GFP-tagged wild-type (WT) ID3 and HA-MDC1 was generated. Levels of exogenous ID3 and MDC1 were conﬁrmed by immunoblotting using the indicated antibodies (d). Images depict representative nuclei showing MDC1 foci at 1 h after IR treatment (e). Scale bars: 10 μm. The histogram in the lower panel is a quantiﬁcation of the average number of cells containing MDC1 foci 1 h after exposure to IR (n = 3). f, g Control or ID3-depleted HeLa (f) and U2OS (g) cells were exposed to 2 Gy of IR. Images depict representative nuclei showing MDC1, γ-H2AX, NBS1, BRCA1, 53BP1, RNF8, and RNF168 foci at 1 h after IR treatment. The lower panel shows the number of IR-induced foci in control and ID3-depleted cells (n = 3). Scale bars: 10 μm. Uncropped blots of this Figure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14. Data are presented as means ± s.d. P values are based on Student’s two-tailed t-test: **P < 0.01 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 7 | | | ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z + HA-MDC1 GFP-ID3 a 48 IB: anti-GFP MDC1 binding 1 25 81 119 IP: anti-HA (ID3) ++ ID3 WT HLH (MDC1) ++++ ID3 HLH +/– ID3 ΔHLH IB: anti-HA ++ ID3 ΔC (MDC1) 245 +++ ID3 ΔN IB: anti-GFP (ID3) (Whole-cell lysates) c d e ID3 shRNA + HA-MDC1 + HA-MDC1 + HA-MDC1 GFP-ID3 GFP-ID3 WT S65A S65E GFP-ID3 WT T62A S65A 48 GFP-ID3 GFP-ID3 GFP IP: HA IP: HA HA-MDC1 HA-MDC1 HA-MDC1 180 245 Input 48 48 Input GFP-ID3 GFP-ID3 β-actin f g h ID3 shRNA + HA-MDC1 ID3 shRNA + HA-MDC1 GFP- WT T62A S65A S65E ID3-GFP Mock HLH ΔHLH ID3 HLH ID3 shRNA + HA-MDC1 GFP-ID3 HLH GFP-ID3 HA-MDC1 α-tubulin ns ** ** ** ns GFP-ID3 HLH ID3-GFP Mock HLH ΔHLH ID3 shRNA + HA-MDC1 ID3 shRNA + HA-MDC1 Fig. 5 An interaction between ID3 and MDC1 is required for the recruitment of MDC1 at DSBs. a A schematic representation of wild-type ID3 (WT) and deletion constructs is shown. Domain boundaries, including the HLH region (25–81), are indicated using numbering for the human ID3 protein sequence. The relative afﬁnity of MDC1 for the ID3 deletion mutants is indicated on the right side of the ﬁgure. b The HLH domain of ID3 is required for binding to MDC1. HEK293T cells were co-transfected with HA-tagged MDC1 and GFP-tagged deletion mutants of ID3, as indicated. After 48 h, cells were exposed to IR, and 3 h later, cell lysates were subjected to immunoprecipitation (IP) and immunoblotting (IB) using the indicated antibodies. * indicated heavy chain. c HEK293T cells were co-transfected with HA-MDC1 and one of three different versions of GFP-ID3: WT, T62A, or S65A. Cell lysates were subjected to IP using an anti-HA antibody and IB using antibodies indicated to the right of the image. d HEK293T cells were co-transfected with HA-MDC1 and one of three different versions of GFP-ID3: WT, S65A, or S65E. Cell lysates were subjected to IP using an anti-HA antibody and IB using antibodies indicated to the right of the image. e, f HeLa cells with a stable ID3 knockdown expressing HA-WT MDC1 were reconstituted with the indicated ID3 constructs or with a mock control. The exogenous ID3 and MDC1 were analyzed by IB using the indicated antibodies (e). Immunostaining for HA-MDC1 and GFP-ID3 was carried out 1 h after exposure to IR (f). Scale bars: 10 μm. The histogram in the lower panel quantiﬁes the number of IR-induced foci (n = 3). g, h Similar to e, f but with the indicated point mutants in the HLH of ID3 expressed in stable ID3 knockdown HeLa cells (n = 3). Scale bars: 10 μm. Uncropped blots of this Figure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14. Data are presented as means ± s.d. P values are based on Student’s two-tailed t-test: **P < 0.01. ns, not signiﬁcant 30–32 (Fig. 7j, k), suggesting that phosphorylated S65 of ID3 is damage . Thus, we looked for a role for ID3 in these cell-cycle necessary. These ﬁndings established that the characteristics of checkpoints as well and found that an ID3 knockdown did not ID3 that are important for interactions with MDC1 are also noticeably affect the G2/M checkpoints in response to IR necessary to promote DSB repair. (Supplementary Fig. 9). ID3 is involved in the transcriptional Another intracellular role for MDC1 is to regulate intra-S- repression of p27, a key regulator of cell-cycle progression, and 33, 34 phase and G2/M cell-cycle checkpoints in response to DNA thus induced the G1/S transition . Indeed, HeLa cells 8 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | % of cell with HA- MDC1 > 10 foci DAPI GFP-ID3 MDC1 WT T62A S65A S65E HA- DAPI GFP-ID3 MDC1 HA-MDC1 foci no. per cells WT WT HLH ΔHLH T62A Mock ΔC S65A HLH ΔN ΔHLH S65E NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE a b siRNA Control ID3 siRNA Control H2AX MDC1 MDC1 IP:MDC1 γ-H2AX 245 IP : MDC1 ID3 ID3 MDC1 MDC1 Input γ-H2AX 17 Input γ-H2AX 17 17 ID3 ID3 c d +HA-HLH ID3 pS65 K2075 K2071 V67 A2078 GFP–tBRCT F1979 GFP–BRCT F2079 T2067 IP : HA E2063 R1933 HA-ID3 T1898 HA-ID3 17 Y142 Input GFP–BRCT 48 pS139 K1936 Fig. 6 ID3 promotes binding of MDC1 to γ-H2AX. a Control and ID3-depleted HeLa cells were treated with and without exposure to IR. 3 h later, whole-cell lysates were subjected to immunoprecipitation (IP) using an anti-MDC1 antibody, and western blotting using anti-MDC1, anti-γ-H2AX and anti-ID3 antibodies as indicated to the right of the blot. b Control and H2AX-depleted HeLa cells were treated with and without exposure to IR. After 3 h, whole-cell lysates were subjected to immunoprecipitation using an anti-MDC1 antibody, and western blotting using anti-MDC1, anti-γ-H2AX, and anti-ID3 antibodies as indicated to the right of the blot. c GFP-tagged versions of each of the tBRCT point mutants were expressed in HEK293T cells together with the HLH domain of ID3. About 3 h after irradiation, cell lysates were subjected to immunoprecipitation and western blotting using antibodies as indicated to the right of the blot. d A model of the three-dimensional protein structures for the proposed trimer of MDC1–tBRCT, γ-H2AX, and ID3-pHLH is shown. The blue ribbon represents phosphorylated ID3-HLH, the gray ribbon represents the tBRCT domain of MDC1, the orange ribbon represents the region of tBRCT where ID3 binds, and the pink ribbon represents the γ-H2AX pentapeptide. Relevant amino acid residues are numbered, and notable hydrogen bonding is indicated by a black dashed line. Uncropped blots of this Figure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14 transfected with ID3 siRNA were less proliferative compared to and MRC-5 cells (Supplementary Fig. 13), indicating that the control siRNA-transfected cells (Supplementary Fig. 10). This DSB repair deﬁciency observed when ID3 is depleted results in a may be explained by the fact that IR-induced G2/M checkpoints failure to maintain chromosome integrity. were not affected in ID3-deﬁcient cells. Discussion ID3 promotes DSB repair and genomic stability. DSBs can be ID3 belongs to a family of proteins that consists of four members, repaired through one of two pathways, HR or NHEJ. Because we ID1–ID4, which are ubiquitously expressed in many different had established a role for ID3 in promoting DNA repair, the next tissues . Most of the ID proteins contain a basic DNA recog- step was to look more closely at speciﬁc pathways using well- nition region that allows binding to either an E-box (CANNTG) 35, 36 established reporter assays for both HR and NHEJ .We or an N-box (CACNAG) in a promoter region; however, ID3 is found that ID3 depletion led to decreased HR repair (Fig. 8a, b). lacking this region and instead, functions through dimerization NHEJ repair was also attenuated by knockdown of ID3 (Fig. 8c, with other transcriptional regulators . Consequently, ID3 inhi- d). Knockdown of ID3 did not cause an additional reduction of bits the activity of several transcription factors by directly binding HR (Supplementary Fig. 11a) and NHEJ repair (Supplementary and preventing their association with regulatory sequences in 39–41 Fig. 11b) in MDC1-depleted cells, suggesting that ID3 regulates target genes . ID3 plays an important role in controlling cell HR and NHEJ repair through MDC1. cycle, proliferation, differentiation, and apoptosis. ID3 has been Next, we measured the number of metaphase spreads contain- also implicated as relevant to the pathology of vascular and 42–44 ing chromosomal breaks that result from exposure to IR. As metabolic disease . These biological effects are mainly shown in Fig. 8e, cells in which ID3 had been depleted had an attributed to the aforementioned inhibition of transcription fac- increased number of chromosomal breaks following IR treatment tors, speciﬁcally basic-HLH (bHLH) transcription factors. To our when compared to control cells. We then performed array- knowledge, the data presented here demonstrates for the ﬁrst comparative genome hybridization (array CGH) , comparing time that ID3 binds directly to proteins that are not transcription the proﬁles of ID3-depleted and wild-type human ﬁbroblast factors, and exerts its effects in a transcription-independent GM00637 and human embryonic lung ﬁbroblasts MRC-5 cells. manner. We showed that ID3 directly binds to MDC1 and this We found that chromosomal abnormalities were detected as interaction requires a tBRCT domain in MDC1 as well as an HLH chromosomal gains (dots over +0.5) and losses (dots below −0.5) domain in ID3. The BRCT domain of MDC1 is fairly conserved that were widely distributed throughout the entire genome of between human and murine proteins and it binds selectively to ID3-depleted cells GM00637 (Fig. 8f and Supplementary Fig. 12) peptides containing a phosphorylated serine residue. Importantly, NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 9 | | | WT K2071M K2075M K2071MK2075M 16 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ab c IR (h) 0 0.5 16 γ-H2AX ** Control shRNA ** ID3-1 shRNA IR (min) UT 0 15 30 60 ID3-2 shRNA DAPI ** ** ** γ-H2AX ** DAPI 2 510 IR treatment (Gy) Control siRNA Control siRNA ID3 siRNA ID3 siRNA MDC1 siRNA −− + + ** ID3 siRNA −+ − + ** ** ID3 50 MDC1 25 α-tubulin 48 0 0 0 0.5 16 h UT 0 15 30 60 min e f g h MDC1 siRNA −− + + ID3 shRNA ID3 siRNA −+ − + GFP-ID3 Mock WT ID3 shRNA ID3 shRNA GFP-ID3 Mock WT GFP-ID3 ns 35 GFP GFP ** 25 HA-MDC1 MDC1 ** β-actin β-actin 0 0 GFP-ID3 Mock WT MDC1 siRNA −−++ ID3 shRNA ID3 siRNA −+−+ ij k ID3 shRNA ID3 shRNA ID3 shRNA GFP-ID3 Mock HLH ΔHLH GFP-IHLH-ID3 Mock HLH S65A S65E GFP-IHLH-ID3 GFP ns MDC1 ** ** ** β-actin GFP-ID3 Mock HLH S65A S65E GFP-ID3 Mock HLH ΔHLH ID3 shRNA ID3 shRNA Fig. 7 ID3-depleted cells are sensitive to IR and are defective in DSB repair. a HeLa cells with a stable ID3 knockdown were exposed to the indicated doses of IR and assessed for colony forming ability (n = 3). The cell viability of untreated cells is deﬁned as 100%. b, c γ-H2AX foci (b) and comet moments (c)at indicated time points after exposure to IR in both control and ID3-depleted HeLa cells. Representative images (upper panel) and quantiﬁcation (lower panel) of unrepaired DSBs are shown (n = 3). d HeLa cells were transfected with indicated siRNA combinations, and the expression levels of ID3 and MDC1 were assessed by western blotting using antibodies indicated. e HeLa cells were transfected with indicated siRNA combinations. 1 h after exposure to IR, the cells were assessed using the comet assay. Representative images (upper panel) and quantiﬁcation (lower panel) of comet tail moments are shown (n = 3). f A western blot of GFP-tagged wild-type (WT) ID3 or GFP alone (mock construct) expressed in ID3 knockdown HeLa cells is shown. g Comet moments of complemented HeLa cells are shown as in (e). ID3-depleted cells were complemented with a GFP-WT ID3 construct or with GFP alone. h A western blot of the indicated GFP-ID3 constructs reconstituted into ID3 knockdown HeLa cells is shown. i Comet moments of complemented HeLa cells are shown as in (e). ID3-depleted cells were reconstituted with either HLH ID3 or an HLH deletion mutant (ΔHLH) of ID3 (n = 3). j A western blot demonstrating comparable expression of the indicated GFP-tagged point mutants in the ID3 HLH domain expressed in ID3 knockdown HeLa cells is shown. k Similar to i but with the indicated point mutants in the ID3 HLH domain expressed in ID3 knockdown HeLa cells (n = 3). Scale bars: 40 μm(b) and 200 μm(c, e, g, i, k). Uncropped blots of this Figure are shown in Supplementary Fig. 14. Data are presented as means ± s.d. P values are based on Student’s two-tailed t-test: **P < 0.01. ns, not signiﬁcant 10 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | Relative clonal survival (%) Comet tail moment (%) Comet tail moment (%) Relative γ-H2AX staining (%) Mock HLH S65A S65E ID3 siRNA Control siRNA Comet tail moment (%) Comet tail movement (%) ID3 Control siRNA siRNA Comet tail moment (%) Mock HLH ΔHLH 53BP1 NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE a single phosphorylation of Ser65 in the HLH of ID3, is both effect of various mutant ID3 proteins, we showed that a phos- necessary and sufﬁcient for binding to MDC1. In addition, phorylated Ser65 in the HLH domain was required for proper positively charged amino acid lysine at position 2075, which is localization of MDC1 to damaged DNA sites. Furthermore, we fully ﬂexible at the distal C-terminal region of tBRCT domain, demonstrated that when cellular levels of ID3 were low, there plays an important role in association of MDC1–tBRCT with were fewer DSB repair events, HR and NHEJ activities were ID3–pHLH. ID1, ID2, and ID3 have a high degree of sequence suppressed, and cells are more sensitive to IR. Notably, ID3- similarity and have widely overlapping biological functions in deﬁcient cells transfected with either wild-type ID3 or the HLH many tissues . However, ID3 is the only one of these three domain of ID3, but not ΔHLH ID3 or S65A ID3, fully rescued proteins that contains Ser65 motifs in the HLH domain, which we both IR-induced MDC1 foci and repair of DSBs. Thus, we con- propose is the reason why ID1 and ID2 did not interact with clude that ID3 contributes to the ability of MDC1 to regulate MDC1. This suggests that ID3 may have a unique role in DNA DNA damage repair by directly interacting with MDC1. repair that is distinct from the other ID proteins. Our data also highlight the necessary role that ID3 plays in The direct and highly speciﬁc interaction between ID3 and allowing binding between MDC1 and γ-H2AX, early DNA MDC1 suggests that they are functionally linked. Consistent with damage sensor protein. MDC1 mainly acts as an adapter protein this idea, we have shown that ID3 becomes phosphorylated at that recruits DDR proteins to sites of DNA damage . After DNA Ser65 in response to IR, and that the localization of ID3 to DSBs damage, ATM is recruited to DSBs where it phosphorylates in response to DNA damage resembles that of MDC1. Moreover, H2AX , and then γ-H2AX serves as a docking platform for our ﬁndings suggest that functional ID3 is required for MDC1 to MDC1, which binds to γ-H2AX via its tBRCT domain . Abro- localize appropriately after IR treatment. When we compared the gation of the MDC1-γ-H2AX interaction disrupts IR-induced ac b d siRNA siRNA I-SceI endonuclease I-SceI endonuclease ID3 restriction site ID3 restriction site puro GFP- SceGFP iGFP MDC1 MDC1 + I-Scel + I-Scel 245 ATAA β-actin DSB GFP- 48 β-actin TATT NHEJ 1.2 HR wtGFP iGFP wtGFP 0.9 0.6 ** ** 0.3 3 Control ID3 Control ID3 siRNA siRNA siRNA siRNA e g *** DNA damage Control siRNA ID3 siRNA <ID3 presence > < ID3 absence > ATM ATM ATM p MDC1 MDC1 ID3 p ATM MDC1 MDC1 ID3 Control siRNA ID3 siRNA γ-H2AX γ-H2AX p 53BP1 BRCA1 ATM RNF168 2.00 RNF168 NBS1 RNF8 RNF8 MDC1 NBS1 1.50 BRCA1 ID3 1.00 0.50 γ-H2AX γ-H2AX –0.50 Intact DSB repair Impaired DSB repair –1.00 –1.50 –2.00 Genomic stability Genomic instability Genomic position Fig. 8 Efﬁciency of DSB repair in ID3-depleted cells. a A diagram of the ﬂuorescence-based assay for measuring levels of DSB repair via homologous recombination (HR) is shown. b The efﬁciency of HR repair was measured by FACS analysis in DR-GFP-U2OS cells transfected with either control or ID3 siRNA (lower panel) (n = 3). The levels of endogenous MDC1 and ID3 were analyzed by western blotting (upper panel). c A diagram of the assay for measuring non-homologous end joining (NHEJ) repair using an EJ5-GFP reporter is shown. d The efﬁciency of NHEJ repair was measured by FACS analysis in HeLa cells that contained EJ5-GFP and had been transfected with either control or ID3 siRNA (lower panel) (n = 3). The levels of endogenous MDC1 and ID3 were analyzed by western blotting (upper panel). e Chromosome spreads from control or ID3-depleted HeLa cells exposed to IR are shown. Images (left panel) and a plot (right panel) of the average frequencies of chromosomal breaks (n = 50). Scale bars: 10 μm. f Array CGH proﬁles of clones derived from control or ID3-depleted GM00637 cells are shown. g A schematic representation of the role of ID3 in regulating the DDR functions of MDC1 is shown. The presence (left pathway) and absence (right pathway) of ID3 are compared. Uncropped blots of this Figure accompanied by the location of molecular weight markers are shown in Supplementary Fig. 14. Data are presented as means ± s.d. P values are based on Student’s two-tailed t-test: **P < 0.01, ***P < 0.001 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 11 | | | MDC1 Log2 (ratio) No. of chromosomal breaks GFP expressing cells (%) Control ID3 GFP expressing cells (%) Control ID3 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z MDC1 foci formation and renders cells radiosensitive . In this Antibodies. Polyclonal MDC1 antibody (R2) was raised in rabbit against GST fusion protein containing BRCT domain of MDC1 (residues 1882–2089 aa). study, we observed the interaction between MDC1 and γ-H2AX Antibody against S65 phosphorylated ID3 was generated to phospho-peptide of was abolished by depletion of ID3. Our MD simulations suggest ID3 encompassing amino acids 61–72 (GTQLpSQVEILQR) using polyclonal that interaction of MDC1–tBRCT with ID3-pHLH increases antibody production service (Abfrontier). The antibodies used for immunoblotting ﬂexibility of Arg1933 of MDC1–tBRCT and therefore may give a analysis, immunoprecipitation assay and immunostaining assay are provided in Supplementary Table 1. better chance for catch the γ-H2AX. ID3 does not have obvious sequence homology with known DDR protein domains or rele- vant enzymatic active sites. Instead, ID3 may act as a scaffold for ID3 siRNA and generation of stable ID3 knockdown cell lines. For the the MDC1-γ-H2AX complex by maintaining the interactions, knockdown of ID3, MDC1, or ATM expression, cells were transiently transfected with siRNA using lipofectamine RNAiMAX (Invitrogen) according to the manu- stability, and DSB targeting of these proteins. Consequently, facturer’s instructions. The siRNA target sequences were as follows: ON- depletion of ID3 diminishes both the protein–protein interactions TARGETplus SMARTpool ID3 siRNA (ID3 CDS, NM_002167, GE Dharmacon), and the DSB-targeted activities of MDC1. 5ʹ-GCACUCAGCUUAGCCAGGU-3ʹ,5ʹ-GAACGCAGUCUGGCCAUCG-3ʹ,5ʹ- ID proteins are frequently overexpressed in many cancer cells, GGGAACUGGUACCCGGAGU-3ʹ,5ʹ-GGAAGGUGACUUUCUGUAA-3ʹ; and disease severity and poor prognosis is associated with a high MDC1 siRNA (MDC1 cDNA 58-76), 5ʹ-UCCAGUGAAUCCUUGAGGUdTdT-3ʹ; 41, 46, 47 ATM siRNA (ATM cDNA 1266-1284), 5ʹ-GAUACCAGAUCCUUGGAGAdTdT- level of these proteins . As a result, ID proteins are now 3ʹ; Negative control siRNA (Bioneer), 5ʹ-CCUACGCCACCAAUUUCGUdTdT-3ʹ. considered important targets for potential anti-cancer drugs as a For generation of stable ID3-depleted cell lines, oligonucleotides encoding the 41, 48 way to counteract tumor progression . However, a KO of ID3 target sequence for ID3-1- forward, 5ʹ-GATCCCCACTGCTACTCC resulted in the accumulation of DNA damage in colon cancer CGCCTGTTCAAGAGACAGGCGGGAGTAGCAGTGGTTTTTTG GAAA-3ʹ; reverse, 5ʹ-AGC TTTTCCAAAAAACCACTGCTACTCCCGCCTGTCTCTT- initiating cells , and conversely, ectopic expression of ID3 acti- 18 GAACAGGCGGGAGTAGCAGTGG G-3ʹ and ID3-2- forward, 5ʹ- vated DNA repair processes in pancreatic β cells , thereby GATCCTCCTACAGCGCGTCATCGATTCAAGAGATCGATGA supporting a role for ID3 in maintaining genomic stability. Our CGCGCTGTAGGATTTTTTGGAAA-3ʹ; reverse, 5ʹ- studies revealed that ID3 depletion led to signiﬁcant increases in AGCTTTTCCAAAAAATCCTACAGCGCG TCATCGATCTCTTGAAT CGATGACGCGCTGTAGGAG-3ʹ were annealed and inserted into psilencer2.1- DNA damage accumulation and chromosomal aberrations. The U6-neo vector (Ambion). Cells were transfected with pSilencer2.1-U6-neo control development of gross chromosomal abnormalities, which is a shRNA or pSilencer2.1-U6-neo ID3 shRNA using lipofectamine 2000 (Invitrogen) hallmark of cancer, in ID3-deﬁcient non-transformed cells again −1 and cultured in selection medium containing 500 μgml neomycin for 2–3 weeks. supports the role of ID3 in maintaining chromosome stability. After selection, stably ID3 knockdown clones were conﬁrmed by western blot The association of ID3 with MDC1 and the effect on DSB repair analysis of ID3. that we have described in this study support our conclusion that ID3 functions to maintain genomic stability, very likely by Plasmid constructs and transfection. The plasmids encoding wild-type MDC1 modulating a DDR function of MDC1. Thus, although the and various truncated MDC1 (Δ55–124, Δ200–420, Δ421–624, Δ625–1129, inhibition of ID3 may suppress cancer cell proliferation and Δ1130–1661, Δ1893–2082) were obtained from Zhenkun Lou . For GST pull- down assay, the BRCT fragment of MDC1 was produced by PCR using wild-type tumor progression, the impaired ability of MDC1 to repair DNA MDC1 (pcDNA HA-MDC1) as template and then subcloned in the EcoRI/XhoI damage under these conditions might lead to a signiﬁcant accu- site of pGEX4T-1 vector. To generate the full-length, N-terminal, C-terminal, mulation of chromosomal abnormalities and might subsequently HLH, ΔHLH constructs of ID3, each region of ID3 was ampliﬁed from human promote tumorigenesis and cancer progression. This hypothesis HeLa cDNA, and the PCR products were cloned into pEGFP-N3 or pcDNA HA (Fig. 2a; all amino acid positions were based on the sequence of accession is supported by a recent ﬁnding that ID3-KO mice develop T-cell NP_002158). To prepare the serial deletion constructs of tBRCT (1882–2089, lymphomas . 1882–2062, 1882–2042, 1882–2022, 1882–1992 (BRCT1), 1993–2089 (BRCT2), In summary, the interactions between ID3 and the DNA 2028–2089, 2063–2089), each fragment was PCR-ampliﬁed using pcDNA HA- damage mediator protein MDC1 suggest that ID3 participates in MDC1 as template, and the PCR products were inserted into the EcoRI and XhoI the DNA damage signaling pathway to maintain genomic stabi- sites of pEGFP-N3 vector. Point mutations of ID3 and tBRCT were generated by site-directed mutagenesis (Invitrogen). To construct the expression vector encod- lity. The model depicted in Fig. 8g provides an integrated view of ing shRNA-resistant GFP-tagged ID3, we performed mutagenesis using GEN- how MDC1 recruitment is linked to the upstream DNA damage EART Site-Directed Mutagenesis System (Invitrogen). The shRNA targeting signaling processes, and how ID3 may function at sites of DSBs. sequence, 5ʹ-CCACTGCTACTCCCGCCTG-3ʹ, was replaced by 5ʹ- Our data establish a mechanistic basis for the ID3–MDC1 CCACTGCTATTCCCGACTT-3ʹ. All sequences were conﬁrmed by automated DNA sequencing. Cells were transfected with the appropriate plasmids using cooperation and suggest that, in response to IR, ID3 is phos- lipofectAMINE 2000 (Invitrogen) according to the manufacturer’s instructions. phorylated by ATM and MDC1, speciﬁcally at Ser65 within the HLH domain, which then binds directly to MDC1–tBRCT domain, promoting both the interaction of MDC1 with γ-H2AX Immunoprecipitation assay and western blot analysis. Cells were lysed in ice- and the accumulation of MDC1 at DSB sites, leading to the cold RIPA buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM dithiothreitol, 1 recruitment of additional DDR factors at DSB sites. Thus, our −1 −1 mM phenylmethanesulfonyl ﬂuoride, 10 μgml leupeptin and 10 μgml apro- results illuminate a novel interaction between ID3 and MDC1 tinin). Equal amounts of cell or tissue extracts were separated by 6–12% that is crucial for DSB repair and genome stability. SDS–PAGE followed by electrotransfer onto a polyvinylidene diﬂuoride membrane (PALL life sciences). The membranes were blocked for 1 h with TBS-t (10 mM Tris-HCl (pH 7.4), 150 mM NaCl and 0.1% Tween-20) containing 5% non-fat milk and then incubated with indicated primary antibodies overnight at 4 °C. The blots Methods were washed four times for 15 min with TBS-t and then incubated for 1 h with Cell culture and treatment. The human cervix carcinoma HeLa cells, human peroxidase-conjugated secondary antibodies (1:5000, Jackson ImmunoResearch osteosarcoma U2OS cells and human embryonic kidney HEK293T cells were Inc). The blots were washed four more times with TBS-t and developed using an cultured in Dulbecco’s modiﬁed Eagle’s medium supplemented with 10% heat- enhanced chemiluminescence detection system (ECL; iNtRON Biotechnology, inactivated fetal bovine serum (FBS: Lonza), 100 units per ml penicillin and 100 μg Korea). The amount of MDC1 protein was quantiﬁed using Scion Image software −1 ml streptomycin (Invitrogen). The human ﬁbroblast GM00637 and human (Scion Corp.) For the immunoprecipitation assay, lysates were precleared with embryonic lung ﬁbroblasts MRC-5 cells were cultured in Earle’s MEM containing protein A-Sepharose beads (GE Healthcare) prior to adding the antibody. If DNase 10% FBS, penicillin, and streptomycin. Cells were maintained in a humidiﬁed −1 I or Ethidium Bromide was used, the lysates were either treated with 100 μgml incubator with an atmosphere of 5% CO at 37 °C. All cell lines were from the −1 DNase I (Invitrogen) for 20 min at 37 °C or 50 μgml ethidium bromide (Sigma) American Type Culture Collection (ATCC). To induce DNA DSB, exponentially on ice for 30 min. Next, after removing the protein A-Sepharose by centrifugation, growing cells were irradiated at 2 or 10 Gy from Cs source (Gammacell 3000 the supernatant was incubated at 4 °C overnight with appropriate antibodies Elan irradiator, Best Theratronics) and allowed to recover at 37 °C incubator for (Supplementary Table 1). After the addition fresh protein A-Sepharose bead, the various times. Inhibitors of ATM (Ku55933) and DNA-PKcs (Nu7426) were from incubation was continued for an additional 1 h, and then beads were washed ﬁve TOCRIS bioscience. times with RIPA buffer. Immunoprecipitated proteins were denatured in SDS 12 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE sample buffer, boiled for 5 min, and analyzed by western blotting using the cytometry. For each analysis, 10,000 cells were processed and each experiment was appropriate antibodies (Supplementary Table 1). repeated three times. Chromosomal aberration analysis. For chromosomal aberration analysis, indi- Immunoﬂuorescence microscopy. To visualize DNA damage foci, cells cultured cated transfected-U2OS cells were treated with 1 Gy of γ-ray for 24 h. To arrest on coverslips coated with poly-L-lysine (Sigma) were irradiated at 2 Gy and −1 cells in metaphase, 300 ng ml colcemid (Sigma) was added 4 h before cell col- allowed to recover at 37 °C for adequate times. Cells were washed twice with PBS lection. Colcemid depolymerizes microtubules and inhibits the formation of and ﬁxed with 4% paraformaldehyde for 10 min and ice-cold 98% methanol for 5 mitotic spindle. Cell were collected in 15 ml tubes, gently resuspended in 40% of min, followed by permeabilization with 0.3% Triton X-100 for 10 min at room culture media for 10 min at 37 °C, and then ﬁxed in equivalent volume of a freshly temperature. After permeabilization, coverslips were washed three times with PBS prepared ﬁxative solution (3:1 mixture of methanol/acetic acids, Carnoy’s solution). and then were blocked with 5% BSA in PBS for 1 h. Cells were single or double After removal of supernatant, pellets were resuspended in ﬁxative solution, drop- immunostained with primary antibodies against the indicated proteins (Supple- ped onto a cleaned glass slide and air-dried overnight. The slide was mounted in mentary Table 1) overnight at 4 °C. Cells were washed with PBS and then stained Vectashield with DAPI (Vector Laboratories). Metaphase images were captured with appropriate Alexa Fluor 488- (green, Molecular Probe), Alexa Fluor 594- (red, using confocal microscope (Zeiss LSM 510 Meta; Carl Zeiss) and analyzed with Molecular Probe) conjugated secondary antibodies (Supplementary Table 1). After Zeiss microscope image software ZEN (Carl Zeiss). washing, the coverslips were mounted onto slides using Vectashield mounting medium with 4,6 diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA). Fluorescence images were taken using a confocal microscope (Zeiss LSM 510 CGH array and data analysis. Human ﬁbroblast GM00637 cells were stably Meta; Carl Zeiss) and analyzed with Zeiss microscope image software ZEN (Carl transfected with control or ID3 shRNA, and genomic DNA was isolated using Zeiss). The foci number per cells was counted at least 100 cells. The error bars AccuPrep® Genomic DNA Extraction kit (Bioneer) according to the manu- represent the standard error from three independent experiments. facturer’s instructions. Array CGH analysis was performed using the NimbleGen Human CGH 12 × 135 K whole-genome tiling v3.1 Array (Agilent Technologies). Human genomic DNA (1 μg) from ID3-depleted cells and reference DNA samples GST pull-down assay. Bacterially expressed fusion proteins containing GST and from control cells were independently labeled with ﬂuorescent dyes (Cy3/Cy5), co- the BRCT domains of MDC1 (residues 1893–2082) or GST alone immobilized hybridized at 65 °C for 24 h, and then subjected to the array. The hybridized array onto Glutathion Sepharose 4B beads (GE Heathcare) and incubated with lysates was scanned using NimbleGen’s MS200 scanner (NimbleGen systems Inc.) with 2 prepared from HeLa cells or cells transiently transfected with GFP-ID3 expression μm resolution. Log2-ratio values of the probe signal intensities were calculated and vector for 3 h at 4 °C. Cells were lysed in TEN100 buffer (20 mM Tris-HCl (pH plotted vs. genomic position using Roche NimbleGen’s NimbleScan v2.5 software. 7.4), 0.1 mM EDTA, 100 mM NaCl, 1 mM dithiothreitol, 1 mM phenylmethane- −1 −1 Data are displayed and analyzed in Roche NimbleGen SignalMap software and sulfonyl ﬂuoride, 10 μgml leupeptin, and 10 μgml aprotinin). If λ-phosphatase CGH-explorer v2.55. treatment was required, TEN100 lysates were incubated with 400 units of λ- phosphatase (New England BioLabs) at 30 °C for 30 min. The GST beads were washed ﬁve times with NTEN buffer (0.5% NP-40. 20 mM Tris-HCl (pH 7.4), 1 G2/M checkpoint assay. HeLa cells were transiently transfected with control or mM EDTA, and 300 mM NaCl), and bound proteins were separated by ID3 siRNA. After 48 h of transfection, cells were exposed to IR for 3 h and then −1 SDS–PAGE and analyzed by western blotting using the appropriate antibodies. placed in 100 ng ml nocodazole-containing media for 3 h, and the cells were collected and washed with PBS and then ﬁxed with 1% formaldehyde for 10 min at 37 °C. The cells were chilled on ice for 1 min and then permeabilized with 90% Clonogenic cell survival assay. After treatment with irradiation, 5 × 10 cells were methanol at −20 °C overnight. The ﬁxed cells were washed with PBS and blocked immediately seeded on 60 mm dish in triplicate and grown for 2–3 weeks at 37 °C with incubation buffer (0.5% BSA in PBS) for 10 min. The cells were stained with to allow colonies to form. Colonies were stained with 2% methylene blue/50% anti-phospho-histone H3(S10)-Alexa Fluor 647-conjugated antibody (Cell Signal- ethanol and were counted. The fraction of surviving cells was calculated as the ratio ing Technology, 9716) at a 1:10 dilution in incubation buffer for 1 h in darkness at for the plating efﬁciency of treated cells over untreated cells. Cell survival results room temperature, and the cells were then washed and resuspended in PBS con- are reported as the mean value ± s.d. for three independent experiments. −1 taining 50 g ml propidium iodide. At least 10,000 cells were analyzed by ﬂuorescence-activated cell sorting (FACSort, Becton Dickinson, San Jose, CA). The acquired data were analyzed using the CellQuest Pro software (Becton Dickinson). Single-cell gel-electrophoresis (Comet assay). DSB repair was visualized by neutral single-cell agarose-gel electrophoresis. Brieﬂy, indicated cells were collected (~ 10 cells per pellet), mixed with low-melting agarose, and layered onto agarose- BrdU incorporation assay. To determine the replication rate, control and ID3- coated glass slides. The slides were maintained in the dark for all of the remaining depleted HeLa cells were seeded in a 48-well plate. After 24 h, 10 μM BrdU was steps. Slides were submerged in lysis solution (Cat.#4250-050-01,TREVIGEN® added to the cells, and the culture was then incubated for 2 h at 37 °C. After Instructions, Gaithersburg, MD, USA) for 1 h and incubated for 30 min in neutral ﬁxation of the cells, immune complexes were formed using peroxidase-coupled electrophoresis solution (100 mM Tris, 300 mM Sodium Acetate at pH 9.0). After BrdU-antibodies. Colored products were measured in a microplate reader at 405 −1 incubation slides were electrophoresed (~ 30 min at 1 V cm tank length), and nm with a reference wavelength at ~490 nm. The relative DNA synthesis was then gently immerse slide in DNA Precipitation Solution (1.5 M NH Ac) for 30 calculated as the absorbance of ID3-depleted cells from the absorbance of control min at room temperature. After air-dried, comet slide stained with SYBR green. cells. The data are presented as the mean ± s.d. from triplicate experiments. Average comet tail moment was scored for 40–50 cells per slide using a compu- terized image analysis system (Komet 5.5; Andor Technology, South Windsor, CT, Generation of U2OS-ID3-knockout cells. The ID3 CRISPR (Guide sequence USA). insert: 5ʹ-CGAGGCGGTGTGCTGCCTGTCGG)-3ʹ construct targeting exon 1 of the human ID3 gene were designed using the Cas9-Designer (http://www.rgenome. HR assay. To measure the HR repair, stable cell lines expressing DR-GFP reports net/cas-designer) . SpCas9 expression plasmids (500 ng) and sgRNA plasmids were generated by transfection using lipofectamine 2000. Clones were selected in (500 ng) were transfected to 2 × 10 U2OS cells using 4D-nucleofector (Lonza) SE −1 the medium containing 500 μgml neomycin for 2 weeks and screened for sig- kit and program CM-104 in 20 µl Nucleovette Strips. Genomic DNA was isolated niﬁcant induction of GFP-positive cells following infection with I-SceI expressing with the Nucleospin Tissue Kit (MACHEREY-NAGEL) 72 h post-transfection. adenovirus. U2OS-DR-GFP cells were transfected with control or ID3 siRNA using Target sites were ampliﬁed with adapter primers using Phusion polymerase (New lipofectamine RNAiMAX, and then infected with I-SceI-carrying adenovirus at an England Biolabs). The resulting deep sequencing libraries were subjected to paired- estimated MOI of 10. After 72 h, GFP-positive cells were measured by end sequencing with the MiSeq system (Illumina). After MiSeq, paired-end reads ﬂuorescence-activated cell sorting (FACSCalibur, BD Biosciences). The acquired were joined by the Fastq-join. Fastq-joined ﬁles were analyzed using Cas-Analyzer data was analyzed using CellQuest Pro software (BD Biosciences). The data are (http://www.rgenome.net/cas-analyzer/) to obtain a mutation frequency in edited presented as the mean ± s.d. value in three independent experiments. cells. For single-cell colony expansion, transfected cells were diluted and sorted with a density of 1 cell per well into 96-well-plates. After 7–10 days incubation, only single-cell colonies were screened visually under microscope. When colonies NHEJ assay. The NHEJ assay was measured in HeLa EJ5-GFP cells, using methods reached 70% conﬂuence, the cells were transferred to 24-well-plates. Genomic previous described . EJ5-GFP contains a promoter that is separated from a GFP DNA was isolated from each single colony. Sequences of ID3 target region were coding region by puromycin resistance gene, which is ﬂanked by two I-SceI sites analyzed by a targeted deep sequencing in each single-cell colony and further that are in the same orientation. When the I-SceI-induced DSBs is repaired by conﬁrmed by Sanger sequencing. NHEJ in HeLa EJ5-GFP cells, the puro gene is removed, and the promoter is rejoined to the rest of the GFP expression cassette, leading GFP expression. HeLa EJ5-GFP cells were kindly provided by Dr. Kee at the University of South Florida. Model system preparation. The three dimensional (3D) structure of both ID3 Similar to the above HR assay, HeLa EJ5-GFP cells were transfected with control or protein (PDB code: 2LFH) as ligand and MDC1/NFBD1-γ-H2AX complex (PDB ID3 siRNA using lipofectamine RNAiMAX, and then infected with I-SceI-carrying code: 2AZM) as receptor are revealed by X-ray experiments. The MDC1–tBRCT adenovirus at an estimated MOI of 10. After 3 days, the percentage of GFP-positive domain of MDC1/NFBD1-γ-H2AX complex (residues 1891–2083) is deﬁned to cells which had repaired the DSBs generated by I-SceI was determined by ﬂow bind the γ-H2AX tail (residues 138–142) in the phospho-peptide recognition. For NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 13 | | | ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z the docking simulation, we need the isolated three systems; MDC1–tBRCT domain 8. Lou, Z. et al. MDC1 maintains genomic stability by participating in the such as the receptor, γ-H2AX and ID3 such as the ligand. To prepare the ligand, ampliﬁcation of ATM-dependent DNA damage signals. Mol. Cell 21, 187–200 the initial 25 residues of disordered N-terminal of ID3 protein are discarded. We (2006). used a total 44 amino acids (residues 25–68) to forms the HLH domain of ID3 9. Coster, G. & Goldberg, M. The cellular response to DNA damage: a focus on protein and the residue number is different with that of the experimental source MDC1 and its interacting proteins. Nucleus 1, 166–178 (2010). (NCBI Refseq: NM_002167.3). We assigned the new residue number, corre- 10. Zhang, J., Ma, Z., Treszezamsky, A. & Powell, S. N. MDC1 interacts with Rad51 and sponded to the experimental data. The PDB residues 25–68 are agreement with facilitates homologous recombination. Nat. Struct. Mol. Biol. 12,902–909 (2005). NCBI sequences 40–83. To relax the neutralized systems in a cubic box of TIP3P 11. Xie, A. et al. Distinct roles of chromatin-associated proteins MDC1 and 53BP1 water model, pre-equilibration is performed under NPT environment before the in mammalian double-strand break repair. Mol. Cell 28, 1045–1057 (2007). NVT process. Fully NVT equilibrium is performed until all properties of interest 12. Dimitrova, N. & de Lange, T. MDC1 accelerates nonhomologous end-joining of have stabilized. All simulation temperature is at 277 K (4 °C) like our experiments. dysfunctional telomeres. Genes Dev. 20, 3238–3243 (2006). The MD simulations are performed with amber99SB-nmr-ILDN force ﬁeld with a 13. Luo, K., Yuan, J., Chen, J. & Lou, Z. Topoisomerase IIalpha controls the 53, 54 phospho-Serine . decatenation checkpoint. Nat. Cell Biol. 11, 204–210 (2009). 14. Wang, J., Gong, Z. & Chen, J. MDC1 collaborates with TopBP1 in DNA replication checkpoint control. J. Cell Biol. 193, 267–273 (2011). Docking and molecular dynamics calculation. The docking protocol are used to explore the vast conformational space of ligands in a short time but the major 15. Townsend, K. et al. Mediator of DNA damage checkpoint 1 (MDC1) regulates drawback is the poor ﬂexibility of protein, which is not permitted to adjust its mitotic progression. J. Biol. Chem. 284, 33939–33948 (2009). conformation upon ligand binding. To predict more reliable protein-ligand com- 16. Eliezer, Y., Argaman, L., Kornowski, M., Roniger, M. & Goldberg, M. Interplay plexes, MD simulation can deal with both ligand and protein in a ﬂexible way, between the DNA damage proteins MDC1 and ATM in the regulation of the allowing for an induced ﬁt of the receptor-binding site around the newly intro- spindle assembly checkpoint. J. Biol. Chem. 289, 8182–8193 (2014). duced ligand. To complement the computational methods, we perform their 17. Stucki, M. et al. MDC1 directly binds phosphorylated histone H2AX to regulate combined approach to provide reliable starting complex structures from docking cellular responses to DNA double-strand breaks. Cell 123, 1213–1226 (2005). calculations and to incorporate protein ﬂexibility and analyze complex stability in 18. Lee, S. H., Hao, E., Levine, F. & Itkin-Ansari, P. Id3 upregulates BrdU MD simulation. In order to ﬁnd the reliable complex in the 3D space, HEX protein incorporation associated with a DNA damage response, not replication, in 55 56 57 docking server , and GROMACS software patched with PLUMED 1.3 were human pancreatic beta-cells. Islets 3, 358–366 (2011). used for docking and MD studies. From our biology experimental result, we reports 19. O’Brien, C. A. et al. ID1 and ID3 regulate the self-renewal capacity of human that Lys2075 of the MDC1 BRCT domain binds phosphorylated serine (pSer65) of colon cancer-initiating cells through p21. Cancer Cell 21, 777–792 (2012). ID3 protein to enhance the binding afﬁnity of the γ-H2AX pentapeptide and the 20. Shiloh, Y. ATM and related protein kinases: safeguarding genome integrity. newly special-binding site of the MDC1 BRCT domain is residues 2063–2089. To Nat. Rev. Cancer 3, 155–168 (2003). understand the complex conformation in the atomic level, the MDC1–ID3 com- 21. Matsuoka, S. et al. ATM and ATR substrate analysis reveals extensive protein plex is built to restraint the distance between Lys2075 of the receptor and pSer65 of networks responsive to DNA damage. Science 316, 1160–1166 (2007). the ligand and the angular search range about its own coordinate origin is set up in 22. Sobhian, B. et al. RAP80 targets BRCA1 to speciﬁc ubiquitin structures at DNA HEX docking server. Repeatedly docking calculation is performed by various damage sites. Science 316, 1198–1202 (2007). angular docking parameters to search the possible docking model. The type of 23. Bekker-Jensen, S. & Mailand, N. The ubiquitin- and SUMO-dependent signaling docking calculation is performed “shape-only”. After getting the highest-scoring response to DNA double-strand breaks. FEBS Lett. 585, 2914–2919 (2011). orientations by the docking calculation, the possible docking complex is solvated in + − 24. Traven, A. & Heierhorst, J. SQ/TQ cluster domains: concentrated ATM/ATR the water molecules and is neutralized with Na or Cl . Various docking models kinase phosphorylation site regions in DNA-damage-response proteins. are performed using MD simulation to discriminate among conformations of Bioessays 27, 397–407 (2005). different stability. During MD simulation, all complex models have a restraint 25. Durocher, D., Henckel, J., Fersht, A. R. & Jackson, S. P. The FHA domain is a condition between the speciﬁc binding residues as Lys2075 of the MDC1 and modular phosphopeptide recognition motif. Mol. Cell 4, 387–394 (1999). pSer65 of ID3. The restraint distance is about 4.0 Å and the restraint force is about −1 26. Yu, X., Chini, C. C., He, M., Mer, G. & Chen, J. The BRCT domain is a 100 KJ mol . The trimer of tBRCT-γ-H2AX-pHLH complex is calculated as same phospho-protein binding domain. Science 302, 639–642 (2003). condition. Compared with the structural stability by the potential energy of system, 27. Manke, I. A., Lowery, D. M., Nguyen, A. & Yaffe, M. B. BRCT repeats as we can predict one of possible docking systems and show the 3D structural model phosphopeptide-binding modules involved in protein targeting. Science 302, to explain the interaction with the receptor and the ligand. 636–639 (2003). 28. Dodson, G. G., Lane, D. P. & Verma, C. S. Molecular simulations of protein Statistical analysis. Data in all experiments are represented as mean ± s.d. Two- dynamics: new windows on mechanisms in biology. EMBO Rep. 9, 144–150 tailed Student’s t-test was conducted using GraphPad Prism software (GraphPad (2008). software, Inc.) and Excel (Microsoft). P values < 0.01 were considered statistically 29. Campbell, S. J., Edwards, R. A. & Glover, J. N. Comparison of the structures signiﬁcant (**). and peptide binding speciﬁcities of the BRCT domains of MDC1 and BRCA1. Structure 18, 167–176 (2010). Data availability. The data that support the ﬁndings of this study are available 30. Lou, Z., Minter-Dykhouse, K., Wu, X. & Chen, J. MDC1 is coupled to activated from the corresponding authors upon reasonable request. CHK2 in mammalian DNA damage response pathways. Nature 421, 957–961 (2003). 31. Goldberg, M. et al. MDC1 is required for the intra-S-phase DNA damage Received: 8 December 2016 Accepted: 14 August 2017 checkpoint. Nature 421, 952–956 (2003). 32. Stewart, G. S., Wang, B., Bignell, C. R., Taylor, A. M. & Elledge, S. J. MDC1 is a mediator of the mammalian DNA damage checkpoint. Nature 421, 961–966 (2003). 33. Chassot, A. A. et al. Id3 is a novel regulator of p27kip1 mRNA in early G1 phase and is required for cell-cycle progression. Oncogene 26, 5772–5783 (2007). References 34. Williams, S. A. et al. USP1 deubiquitinates ID proteins to preserve a 1. Ciccia, A. & Elledge, S. J. The DNA damage response: making it safe to play mesenchymal stem cell program in osteosarcoma. Cell 146, 918–930 (2011). with knives. Mol. Cell. 40, 179–204 (2010). 35. Pierce, A. J., Johnson, R. D., Thompson, L. H. & Jasin, M. XRCC3 promotes 2. Jackson, S. P. & Bartek, J. The DNA-damage response in human biology and homology-directed repair of DNA damage in mammalian cells. Genes Dev. 13, disease. Nature 461, 1071–1078 (2009). 2633–2638 (1999). 3. Kastan, M. B. & Bartek, J. Cell-cycle checkpoints and cancer. Nature 432, 36. Bennardo, N., Cheng, A., Huang, N. & Stark, J. M. Alternative-NHEJ is a 316–323 (2004). mechanistically distinct pathway of mammalian chromosome break repair. 4. Polo, S. E. & Jackson, S. P. Dynamics of DNA damage response proteins at PLoS Genet. 4, e1000110 (2008). DNA breaks: a focus on protein modiﬁcations. Genes Dev. 25, 409–433 (2011). 37. Wolf, M. et al. High-resolution analysis of gene copy number alterations in 5. Lukas, J., Lukas, C. & Bartek, J. More than just a focus: the chromatin response human prostate cancer using CGH on cDNA microarrays: impact of copy to DNA damage and its role in genome integrity maintenance. Nat. Cell Biol. number on gene expression. Neoplasia 6, 240–247 (2004). 13, 1161–1169 (2011). 38. Yokota, Y. Id and development. Oncogene 20, 8290–8298 (2001). 6. Huen, M. S. & Chen, J. Assembly of checkpoint and repair machineries at DNA 39. Massari, M. E. & Murre, C. Helix-loop-helix proteins: regulators of damage sites. Trends Biochem. Sci. 35, 101–108 (2010). transcription in eucaryotic organisms. Mol. Cell Biol. 20, 429–440 (2000). 7. Minter-Dykhouse, K., Ward, I., Huen, M. S., Chen, J. & Lou, Z. Distinct versus 40. Mueller, C., Baudler, S., Welzel, H., Bohm, M. & Nickenig, G. Identiﬁcation of a overlapping functions of MDC1 and 53BP1 in DNA damage response and novel redox-sensitive gene, Id3, which mediates angiotensin II-induced cell tumorigenesis. J. Cell Biol. 181, 727–735 (2008). growth. Circulation 105, 2423–2428 (2002). 14 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications | | | NATURE COMMUNICATIONS | DOI: 10.1038/s41467-017-01051-z ARTICLE 41. Perk, J., Iavarone, A. & Benezra, R. Id family of helix-loop-helix proteins in Acknowledgements cancer. Nat. Rev. Cancer 5, 603–614 (2005). We thank Zhenkun Lou for providing MDC1 deletion constructs and helpful comments 42. Doran, A. C. et al. Id3 is a novel atheroprotective factor containing a on the manuscript. This work is supported by the National Research Foundation of functionally signiﬁcant single-nucleotide polymorphism associated with Korea (NRF), funded by the Ministry of Science, ICT, and Future Planning (NRF- intima-media thickness in humans. Circ. Res. 106, 1303–1311 (2010). 2015R1A5A2009070, NRF- 2017R1A2B2007557 and NRF-2017R1A2B2008064). 43. Lipinski, M. J. et al. Loss of Id3 increases VCAM-1 expression, macrophage −/− accumulation, and atherogenesis in Ldlr mice. Arterioscler. Thromb. Vasc. Author contributions Biol. 32, 2855–2861 (2012). J.-H.L., S.-J.P. and H.J.Y. designed the experiments; J.-H.L., S.-J.P., M.-J.K., S.M.J. and 44. Cutchins, A. et al. Inhibitor of differentiation-3 mediates high fat diet-induced S.-Y.J. performed the experiments; G.H. and I.-Y.C. provided help for the experiments; visceral fat expansion. Arterioscler. Thromb. Vasc. Biol. 32, 317–324 (2012). C.K. and E.K. performed MD simulations; J.Y. and S.B. generated ID3 KO U2OS cells; 45. Strauss, C. & Goldberg, M. Recruitment of proteins to DNA double-strand J.-H.L., S.-J.P. and H.J.Y. analyzed data; J.-H.L., S.-J.P., E.K., S.B. and H.J.Y. wrote the breaks: MDC1 directly recruits RAP80. Cell Cycle 10, 2850–2857 (2011). manuscript. 46. Ruzinova, M. B. & Benezra, R. Id proteins in development, cell cycle and cancer. Trends Cell Biol. 13, 410–418 (2003). 47. Asirvatham, A. J., Carey, J. P. & Chaudhary, J. ID1-, ID2-, and ID3-regulated Additional information gene expression in E2A positive or negative prostate cancer cells. Prostate 67, Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 1411–1420 (2007). 017-01051-z. 48. Iavarone, A. & Lasorella, A. ID proteins as targets in cancer and tools in neurobiology. Trends Mol. Med. 12, 588–594 (2006). Competing interests: The authors declare no competing ﬁnancial interests. 49. Li, J. et al. Mutation of inhibitory helix-loop-helix protein Id3 causes gammadelta T-cell lymphoma in mice. Blood 116, 5615–5621 (2010). Reprints and permission information is available online at http://npg.nature.com/ 50. Wu, L., Luo, K., Lou, Z. & Chen, J. MDC1 regulates intra-S-phase checkpoint reprintsandpermissions/ by targeting NBS1 to DNA double-strand breaks. Proc. Natl Acad. Sci. USA 105, 11200–11205 (2008). Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in 51. Park, J., Lim, K., Kim, J. S. & Bae, S. Cas-analyzer: an online tool for assessing published maps and institutional afﬁliations. genome editing results using NGS data. Bioinformatics 33, 286–288 (2017). 52. Wong, M. V., Jiang, S., Palasingam, P. & Kolatkar, P. R. A divalent ion is crucial in the structure and dominant-negative function of ID proteins, a class of helix- loop-helix transcription regulators. PLoS ONE 7, e48591 (2012). Open Access This article is licensed under a Creative Commons 53. Li, D. W. & Bruschweiler, R. NMR-based protein potentials. Angew. Chem. Int. Attribution 4.0 International License, which permits use, sharing, Ed. Engl. 49, 6778–6780 (2010). adaptation, distribution and reproduction in any medium or format, as long as you give 54. Craft, J. W. Jr. & Legge, G. B. An AMBER/DYANA/MOLMOL phosphorylated appropriate credit to the original author(s) and the source, provide a link to the Creative amino acid library set and incorporation into NMR structure calculations. J. Commons license, and indicate if changes were made. The images or other third party Biomol. NMR 33,15–24 (2005). material in this article are included in the article’s Creative Commons license, unless 55. Macindoe, G., Mavridis, L., Venkatraman, V., Devignes, M. D. & Ritchie, D. W. indicated otherwise in a credit line to the material. If material is not included in the HexServer: an FFT-based protein docking server powered by graphics article’s Creative Commons license and your intended use is not permitted by statutory processors. Nucleic Acids Res. 38, 445–449 (2010). regulation or exceeds the permitted use, you will need to obtain permission directly from 56. Pronk, S. et al. GROMACS 4.5: a high-throughput and highly parallel open the copyright holder. To view a copy of this license, visit http://creativecommons.org/ source molecular simulation toolkit. Bioinformatics 29, 845–854 (2013). licenses/by/4.0/. 57. Bocahut, A., Bernad, S., Sebban, P. & Sacquin-Mora, S. Relating the diffusion of small ligands in human neuroglobin to its structural and mechanical properties. J. Phys. Chem. B. 113, 16257–16267 (2009). © The Author(s) 2017 NATURE COMMUNICATIONS 8: 903 DOI: 10.1038/s41467-017-01051-z www.nature.com/naturecommunications 15 | | |
Nature Communications – Springer Journals
Published: Oct 12, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera