CDYL1 fosters double-strand break-induced transcription silencing and promotes homology-directed repair

CDYL1 fosters double-strand break-induced transcription silencing and promotes homology-directed... Abstract Cells have evolved DNA damage response (DDR) to repair DNA lesions and thus preserving genomic stability and impeding carcinogenesis. DNA damage induction is accompanied by transient transcription repression. Here, we describe a previously unrecognized role of chromodomain Y-like (CDYL1) protein in fortifying double-strand break (DSB)-induced transcription repression and repair. We showed that CDYL1 is rapidly recruited to damaged euchromatic regions in a poly (ADP-ribose) polymerase 1 (PARP1)-dependent, but ataxia telangiectasia mutated (ATM)-independent, manner. While the C-terminal region, containing the enoyl-CoA hydratase like (ECH) domain, of CDYL1 binds to poly (ADP-ribose) (PAR) moieties and mediates CDYL1 accumulation at DNA damage sites, the chromodomain and histone H3 trimethylated on lysine 9 (H3K9me3) mark are dispensable for its recruitment. Furthermore, CDYL1 promotes the recruitment of enhancer of zeste homolog 2 (EZH2), stimulates local increase of the repressive methyl mark H3K27me3, and promotes transcription silencing at DSB sites. In addition, following DNA damage induction, CDYL1 depletion causes persistent G2/M arrest and alters H2AX and replication protein A (RPA2) phosphorylation. Remarkably, the ‘traffic-light reporter’ system revealed that CDYL1 mainly promotes homology-directed repair (HDR) of DSBs in vivo. Consequently, CDYL1-knockout cells display synthetic lethality with the chemotherapeutic agent, cisplatin. Altogether, our findings identify CDYL1 as a new component of the DDR and suggest that the HDR-defective ‘BRCAness’ phenotype of CDYL1-deficient cells could be exploited for eradicating cancer cells harboring CDYL1 mutations. CDYL1, PARP1, double-strand breaks, homology-directed repair, non-homologous end joining, H3K27me3, EZH2 Introduction Our genome is constantly attacked by exogenous and endogenous mutagens that cause various types of DNA damage. To cope with the massive amount of DNA lesions, cells have evolved diverse mechanisms, collectively named DNA-damage response (DDR), to sense and repair DNA lesions (Tubbs and Nussenzweig, 2017). Sensing and repairing DNA lesions are mediated by rapid and highly orchestrated changes in chromatin structure and dynamics, followed by accumulation of DNA damage responsive proteins to DNA breakage sites (Jackson and Bartek, 2009; Adam and Polo, 2014). Several reports showed that DNA damage induction is accompanied by transient transcription pause to eliminate production of abnormal transcripts and to avoid deleterious collisions between transcription and repair machineries (Kruhlak et al., 2007; Chou et al., 2010; Shanbhag et al., 2010; Svejstrup, 2010; Pankotai et al., 2012; Adam and Polo, 2014; Kakarougkas et al., 2014; Ui et al., 2015; Wickramasinghe and Venkitaraman, 2016; Awwad et al., 2017; Polo, 2017). Double-strand breaks (DSBs) are considered the most cytotoxic form of DNA damage and vertebrates utilize two distinct DSB repair pathways. The first is homology-directed repair (HDR); an error-free process that functions only in late S and G2 phases of the cell cycle, where an intact chromatid is available to serve as a template for repair. In this pathway the ends of the DSB are recognized by the MRE11/RAD50/NBS1 (MRN) complex that catalyzes the formation of 3’ single stranded DNA (ssDNA) overhangs. The ssDNA is swiftly coated by the replication protein A (RPA), which is then displaced by Rad51 protein to initiate DNA strand invasion into the sister chromatid. This process requires the activity of several DNA repair proteins including BRCA1 and BRCA2 (Jasin and Rothstein, 2013; Hustedt and Durocher, 2016). The second repair pathway is non-homologous end joining (NHEJ); an error-prone process that functions throughout the cell cycle (Hustedt and Durocher, 2016). In NHEJ, Ku70-Ku80 heterodimer and 53BP1 mediator protein bind the DSB ends to inhibit end resection (Lieber, 2010). Subsequently, several factors including DNA-dependent protein kinase catalytic subunit, polynucleotide kinase and XRCC4-DNA ligase IV complex are recruited to promote end joining of the broken ends. Quantitative phospho-proteomic screens revealed that CDYL1 protein undergoes phosphorylation in response to DNA damage (Matsuoka et al., 2007; Bennetzen et al., 2010; Elia et al., 2015). These data prompted us to investigate the alleged role of CDYL1 in DDR. Human CDYL1 gene has three splicing variants: CDYL1a, CDYL1b, and CDYL1c. CDYL1b is the most abundant variant and consists of an N-terminal chromodomain (CD), a central hinge region and a C-terminal enoyl-CoA hydratase-like (ECH) domain. CDYL1 directly interacts with the catalytic subunit of polycomb repressive complex 2 (PRC2), enhancer of zeste homolog 2 (EZH2), and facilitates the establishment and propagation of H3K27me3 mark (Zhang et al., 2011). In addition, it was recently reported that CDYL1 plays an important role in the maintenance of repressive histone marks during DNA replication (Liu et al., 2017b). Moreover, it was shown that CDYL1 chromodomain bridges between the repressor element-1 silencing transcription factor (REST) and G9a methyltransferase (catalyzes the dimethylation of H3K9) to repress transcription (Mulligan et al., 2008). In line with this, CDYL1b contains a functional chromodomain that can bind H3K9me2/3 (Franz et al., 2009) and H3K27me3 (Vermeulen et al., 2010) and is involved in X-chromosome inactivation (Escamilla-Del-Arenal et al., 2013). The ECH domain of CDYL1b is essential for its multimerization and mediates its interaction with histone deactylases HDAC1 and HDAC2 to promote transcription repression (Caron et al., 2003). Furthermore, it was recently shown that the ECH domain of CDYL1 negatively regulates the transcriptionally active histone mark, lysine crotonylation (Wu et al., 2009a; Liu et al., 2017a). Notably, CDYL1 is a cancer-associated gene with a notable tumor suppressor activity. CDYL1 depletion increases the expression of several known oncogenes (e.g. TrkC) and promotes cellular transformation as evident by enhanced growth of human mammary epithelial cells (HMEC) in semi-solid media (Mulligan et al., 2008). Furthermore, the International Cancer Genome Consortium databases (http://icgc.org/) show that CDYL1 is mutated in 7.8% of diverse human cancers (n = 20343 cases). On the other hand, it was also shown that CDYL1 is required for repressing the CDH1 (E-cadherin) gene, whose loss is an essential event in epithelial–mesenchymal transition and is crucial for cancer metastasis, indicating that CDYL1 could facilitate tumorigenesis (Caron et al., 2003). Altogether, these data suggest that CDYL1 may possess oncogenic and tumor suppressor activities and can play distinct roles in specific biological processes when present in different protein complexes. Here, we present several lines of evidence implicating CDYL1 in DDR (hereafter, we refer to the endogenous protein as CDYL1 and the overexpressed form as CDYL1b). We showed that the ECH domain of CDYL1b and PARP1 activity promote CDYL1b recruitment to DNA damage sites. Remarkably, CDYL1 promotes EZH2 recruitment, fosters trimethylation of H3K27 at laser-microirradiated regions and underpins DSB-induced transcription silencing. Moreover, we demonstrated that CDYL1 exclusively facilitates HDR of DSBs. Lastly, we observed that CDYL1-depleted cells exhibit hypersensitivity to the widely used chemotherapeutic drug, cisplatin, and thus providing the basis for a new targeted therapy of CDYL1-deficient cancer cells. Results CDYL1 is rapidly recruited to DNA damage sites Given that CDYL1 is phosphorylated in response to DNA damage (Matsuoka et al., 2007; Bennetzen et al., 2010; Elia et al., 2015), we assumed that it might participate in DDR. To test this assumption, we monitored the subcellular localization of the most abundant variant, CDYL1b, following DNA damage induction. First, we tested the localization of EGFP-CDYL1b fusion at DNA breakage sites induced by laser microirradiation (Khoury-Haddad et al., 2014; Abu-Zhayia et al., 2017; Awwad et al., 2017) in four different cell types: Osteosarcoma human cells (U2OS); human breast adenocarcinoma cells (MCF7); non-tumorigenic breast epithelial cells (MCF10A) and mouse embryonic fibroblasts (MEF). Results showed rapid accumulation of EGFP-CDYL1b at laser-microirradiated regions in all tested cell lines (Figure 1A–D). Notably, the rapid recruitment kinetics of EGFP-CDYL1b is comparable to the mediator of DNA damage checkpoint 1 (MDC1) protein, which becomes detectable within 20 sec after damage induction, suggesting it is an early event in DDR (Supplementary Figure S1). Figure 1 View largeDownload slide CDYL1 is rapidly recruited to DNA damage sites. EGFP-CDYL1b shows swift accumulation at DNA breakage sites induced by laser microirradiation in U2OS (A), MCF7 (B), MCF10A (C), and MEF (D). Graphs in the right display fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites. Error bars indicate standard deviation (SD) from three independent experiments, and each measurement represents >10 different cells. White arrows mark the microirradiated regions. (E−G) U2OS (E), MCF7 (F), and MEF (G) cells were exposed to laser microirradiation and 5 min later were fixed and co-stained for γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). Results are typical of two independent experiments (n > 25). Scale bar, 2 μm. (H) Biochemical fractionation demonstrated that IR leads to CDYL1 accumulation at the chromatin-bound fraction. U2OS cells were exposed to 10 Gys of IR followed by biochemical fractionation at 5 and 30 min after IR. Protein samples were immunoblotted using the indicated antibodies. (I) Endogenous CDYL1 is enriched at I-SceI-induced DSBs. ChIP was performed 2 h after tamoxifen treatment using IgG, CDYL1, and γH2AX (as a control for damage induction) antibodies. Data are presented as fold enrichment around the DSB site relative to the negative control (IgG) and mean ± SD from three independent experiments. Statistical analysis was performed by Student’s t-test, **P < 0.01. Figure 1 View largeDownload slide CDYL1 is rapidly recruited to DNA damage sites. EGFP-CDYL1b shows swift accumulation at DNA breakage sites induced by laser microirradiation in U2OS (A), MCF7 (B), MCF10A (C), and MEF (D). Graphs in the right display fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites. Error bars indicate standard deviation (SD) from three independent experiments, and each measurement represents >10 different cells. White arrows mark the microirradiated regions. (E−G) U2OS (E), MCF7 (F), and MEF (G) cells were exposed to laser microirradiation and 5 min later were fixed and co-stained for γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). Results are typical of two independent experiments (n > 25). Scale bar, 2 μm. (H) Biochemical fractionation demonstrated that IR leads to CDYL1 accumulation at the chromatin-bound fraction. U2OS cells were exposed to 10 Gys of IR followed by biochemical fractionation at 5 and 30 min after IR. Protein samples were immunoblotted using the indicated antibodies. (I) Endogenous CDYL1 is enriched at I-SceI-induced DSBs. ChIP was performed 2 h after tamoxifen treatment using IgG, CDYL1, and γH2AX (as a control for damage induction) antibodies. Data are presented as fold enrichment around the DSB site relative to the negative control (IgG) and mean ± SD from three independent experiments. Statistical analysis was performed by Student’s t-test, **P < 0.01. To rule out a possibility that the observed accumulation is due to the fusion between CDYL1b and EGFP or CDYL1b overexpression, we tracked the subcellular localization of endogenous CDYL1 protein upon DNA damage induction. Toward this end, we devised three distinct but complementary approaches: First, U2OS, MCF7, and MEF cells were subjected to laser microirradiation and co-stained for γH2AX and CDYL1. The suitability of CDYL1 antibody for immunofluorescence analysis was confirmed (Supplementary Figure S2). Results showed that endogenous CDYL1 protein accumulates at DNA breakage sites marked by γH2AX (Figure 1E–G). Interestingly, once triggered, CDYL1 accumulation lasts for at least 1 h (Supplementary Figure S3). Second, biochemical fractionation of untreated and ionizing radiation (IR)-treated cells showed that the endogenous CDYL1 protein is enriched at damaged chromatin both at 5 and 30 min after IR (Figure 1H). Third, we monitored the localization of CDYL1 at DSB sites. Toward this end, we induced DSBs using I-SceI endonuclease in U2OS-TRE-I-SceI-19 cells (Awwad et al., 2017). Chromatin immunoprecipitation (ChIP) analysis showed significant enrichment of CDYL1 protein at DSB sites (Figure 1I). Notably, CDYL1 does not accumulate at UV-induced DNA damage sites marked by cyclobutane pyrimidine dimers (CPD) (Supplementary Figure S4). Altogether, these findings provide a firm evidence that CDYL1 is rapidly recruited to DSB sites in tumorigenic and non-tumorigenic cells. Also, CDYL1b recruitment is not cell type-specific and is conserved between human and mouse. CDYL1b exhibits prominent accumulation at damaged euchromatin We showed that CDYL1b foci colocalize with three heterchromatic markers: Hoechst-bright spots, H3K9me3, and HP1α (Figure 2A–C). These results are in line with previous findings showing CDYL1 enrichment at heterochromatic regions (Fischle et al., 2008; Franz et al., 2009). We took advantage of CDYL1b informative localization at heterochromatic regions to determine whether it is recruited to DNA damage sites induced within heterochromatic and euchromatic regions. Toward this end, MEF cells expressing EGFP-CDYL1b were subjected to laser microirradiation at heterochromatic and euchromatic regions and the intensity of EGFP signal was measured overtime. Results showed very slight increase in the fluorescence intensity of EGFP-CDYL1b signal at laser-microirradiated heterochromatin, when compared to damaged euchromatic regions that exhibit ~3.5-fold increase in the intensity of EGFP-CDYL1b (Figure 2D and E). To further substantiate this finding, we irradiated euchromatic and heterochromatic regions in the same cell. Again, results nicely showed prominent accumulation of EGFP-CDYL1b at euchromatic regions (Figure 2F). Altogether, our data strongly suggest that CDYL1b shows prominent accumulation at DNA breakage sites induced within euchromatic regions. Figure 2 View largeDownload slide CDYL1b accumulates prominently at damaged euchromatic regions. (A−C) CDYL1b is enriched at heterochromatic regions. MEF cells expressing EGFP-CDYL1b were stained with Hoechst (blue) (A), H3K9me3 (B), and HP1α (C) antibodies. Cells are representative of two independent experiments (n > 100). (D−F) Representative MEF cells showing EGFP-CDYL1b accumulation at laser-microirradiated sites induced in euchromatin (D) and heterochromatin (E). Graphs in the right show fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites marked by white arrows. (F) MEF cells showing EGFP-CDYL1b accumulation in both euchromatic and heterochromatic regions within the same cell. Graph shows absolute fluorescence intensity (%) of EGFP-CDYL1b at damaged heterochromatic and euchromatic sites. Cells are representative of three biological repeats (n > 10). Error bars indicate SD. Scale bar, 2 μm. Figure 2 View largeDownload slide CDYL1b accumulates prominently at damaged euchromatic regions. (A−C) CDYL1b is enriched at heterochromatic regions. MEF cells expressing EGFP-CDYL1b were stained with Hoechst (blue) (A), H3K9me3 (B), and HP1α (C) antibodies. Cells are representative of two independent experiments (n > 100). (D−F) Representative MEF cells showing EGFP-CDYL1b accumulation at laser-microirradiated sites induced in euchromatin (D) and heterochromatin (E). Graphs in the right show fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites marked by white arrows. (F) MEF cells showing EGFP-CDYL1b accumulation in both euchromatic and heterochromatic regions within the same cell. Graph shows absolute fluorescence intensity (%) of EGFP-CDYL1b at damaged heterochromatic and euchromatic sites. Cells are representative of three biological repeats (n > 10). Error bars indicate SD. Scale bar, 2 μm. The carboxyl terminus of CDYL1b mediates its recruitment to DNA damage sites We sought to map the region of CDYL1b that regulates its accumulation at laser-microirradiated sites. Toward this end, deletion mutants of CDYL1b that lack either the chromodomain (CDYL1bdel(CD)), the hinge region (CDYL1bdel(Hinge)) or the C-terminal region including the ECH domain (CDYL1bdel(ECH)) were tagged with EGFP (Figure 3A) and tested for recruitment to laser-microirradiated sites. While the recruitment kinetic of EGFP-CDYL1bdel(Hinge) is comparable to the wild-type protein (Figure 3B and C), EGFP-CDYL1bdel(ECH) exhibits defective accumulation at DNA damage sites (Figure 3D). These observations suggest that the ECH domain facilitates CDYL1b recruitment to DNA damage sites. In contrast, EGFP-CDYL1bdel(CD) mutant shows faster and superior accumulation when compared to the wild-type protein (Figure 3E). We concluded therefore that CDYL1b chromodomain counteracts its recruitment to DNA damage sites. On this basis, we assumed that the binding of CDYL1b to H3K9me3 mark via its chromodomain is dispensable for its recruitment. To validate this assumption, we tested the recruitment of MonoRed-CDYL1b to laser-microirradiated sites in cells overexpressing EGFP-KDM4D protein. KDM4D overexpression causes a severe reduction in H3K9me2/3 (Khoury-Haddad et al., 2014) (Supplementary Figure S5). Results showed that both KDM4D and CDYL1b are recruited to DNA damage sites (Figure 3F). Altogether, we concluded that the compiling of CDYL1b at DNA breakage sites is independent of its binding to H3K9me2/3. Figure 3 View largeDownload slide The C-terminus of CDYL1b regulates its recruitment to DNA damage sites. (A) Western blot shows the expression of the indicated deletion mutants and the wild-type (WT) EGFP-CDYL1b fusion in U2OS cells. (B−E) Representative images of three independent experiments (n > 15) showing the recruitment of CDYL1b WT and deletion mutants to laser-microirradiated sites marked by white arrows. Graph in the right shows quantitative measurements of the fluorescence intensity of the indicated EGFP-CDYL1b fusions at the irradiated sites. Error bars indicate SD. (F) KDM4D overexpression shows no detectable effect on CDYL1b recruitment to laser-microirradiated sites. Laser microirradiation was applied on U2OS cells expressing EGFP-KDM4D (green) and MonoRed-CDYL1b (red). Graph displays the increase in the red and green fluorescence signals at DNA damage sites. Error bars indicate SD. Scale bar, 2 μm. Figure 3 View largeDownload slide The C-terminus of CDYL1b regulates its recruitment to DNA damage sites. (A) Western blot shows the expression of the indicated deletion mutants and the wild-type (WT) EGFP-CDYL1b fusion in U2OS cells. (B−E) Representative images of three independent experiments (n > 15) showing the recruitment of CDYL1b WT and deletion mutants to laser-microirradiated sites marked by white arrows. Graph in the right shows quantitative measurements of the fluorescence intensity of the indicated EGFP-CDYL1b fusions at the irradiated sites. Error bars indicate SD. (F) KDM4D overexpression shows no detectable effect on CDYL1b recruitment to laser-microirradiated sites. Laser microirradiation was applied on U2OS cells expressing EGFP-KDM4D (green) and MonoRed-CDYL1b (red). Graph displays the increase in the red and green fluorescence signals at DNA damage sites. Error bars indicate SD. Scale bar, 2 μm. PARP1, but not ATM, regulates CDYL1b recruitment to DNA damage sites Given the established role of ATM activity in promoting the recruitment of various DNA repair proteins to DNA damage sites (Shiloh and Ziv, 2013), we examined whether ATM regulates CDYL1b recruitment. Pharmacological inhibition of ATM has no discernable effect on CDYL1b recruitment to laser-microirradiated regions (Supplementary Figure S6A and B). To test the potency of ATM inhibition, we monitored CtIP recruitment to laser-microirradiated regions during S and G2 phase of the cell cycle. To do so, cells expressing MonoRed-CtIP and EGFP, which mark S/G2 and M phases, respectively, fused to the N-terminal domain of Geminin (Sakaue-Sawano et al., 2008) were subjected to laser microirradiation. Consistent with previous observations (You et al., 2009), ATM inhibition abrogates CtIP recruitment to DNA damage sites (Supplementary Figure S6C and D). Altogether, we concluded that CDYL1b accumulates at DNA damage sites irrespective of ATM activity. In agreement with this, the ATM-dependent phosphorylation of CDYL1 on Serine 160 is not required for its accumulation at damage sites. As shown in Supplementary Figure S7, the recruitment kinetics of EGFP-CDYL1bS160A phospho-mutant is comparable to the wild-type protein. We and others have shown that the recruitment of many DDR responsive proteins to DNA damage sites is dependent on PARP1 activity (Sousa et al., 2012; Tallis et al., 2014; Khoury-Haddad et al., 2015; Abu-Zhayia et al., 2017; Awwad et al., 2017). On this basis, we sought to determine whether PARP1 regulates EGFP-CDYL1b recruitment to damage sites. Here, we provide two lines of evidence demonstrating that PARP1 activity regulates EGFP-CDYL1b accumulation at laser-microirradiated sites. First, pharmacological inhibition of PARP1/2 abrogates EGFP-CDYL1b and the endogenous CDYL1 accumulation at laser-microirradiated sites (Figure 4A–D). Second, siRNA depletion of PARP1 (Figure 4E) leads to a severe reduction in the percentage of cells that show EGFP-CDYL1b recruitment to DNA damage sites (Figure 4F). Altogether, these results provide firm evidence that PARP1 activity is critical for CDYL1 recruitment to sites of damage. Figure 4 View largeDownload slide PARP1 activity regulates CDYL1 recruitment to DNA damage sites. (A−D) Pharmacological inhibition of PARP1/2 abolishes CDYL1 recruitment to DNA damage sites. Time-lapse images of U2OS cell expressing EGFP-CDYL1b that were treated with PARP inhibitor (A) and DMSO (B) prior to laser microirradiation. Graph (right) shows increase in the fluorescence intensity of EGFP-CDYL1b at damage sites. Data are representative of three biological repeats (n > 30). Error bars indicate SD. (C and D) U2OS cells were treated with PARP inhibitor (C) or DMSO (D), subjected to laser microirradiation and 5 min later immunostained with γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). This experiment was repeated twice (n > 20). (E) Western blot shows siRNA knockdown of PARP1. U2OS cells were transfected with PARP1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (F) PARP1 siRNA impairs EGFP-CDYL1b recruitment to DNA damage sites. Control and PARP1-deficient U2OS cells expressing EGFP-CDYL1b were subjected to laser microirradiation (indicated by red arrows). Scale bar, 2 μm. (G) Sequence alignment showed that CDYL1 protein contains a putative PAR-binding motif. (H) Purification of EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) from HEK293T cells using GFP-TRAP beads. Eluted proteins were separated and stained with Coomassie. Mr indicates protein marker. (I) PAR-binding assay with EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) recombinant proteins. EGFP-NLS and EGFP-Rpp21 were used as negative and a positive controls, respectively. IB, immunoblot. 32P, phosphate radiolabeled PAR. Figure 4 View largeDownload slide PARP1 activity regulates CDYL1 recruitment to DNA damage sites. (A−D) Pharmacological inhibition of PARP1/2 abolishes CDYL1 recruitment to DNA damage sites. Time-lapse images of U2OS cell expressing EGFP-CDYL1b that were treated with PARP inhibitor (A) and DMSO (B) prior to laser microirradiation. Graph (right) shows increase in the fluorescence intensity of EGFP-CDYL1b at damage sites. Data are representative of three biological repeats (n > 30). Error bars indicate SD. (C and D) U2OS cells were treated with PARP inhibitor (C) or DMSO (D), subjected to laser microirradiation and 5 min later immunostained with γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). This experiment was repeated twice (n > 20). (E) Western blot shows siRNA knockdown of PARP1. U2OS cells were transfected with PARP1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (F) PARP1 siRNA impairs EGFP-CDYL1b recruitment to DNA damage sites. Control and PARP1-deficient U2OS cells expressing EGFP-CDYL1b were subjected to laser microirradiation (indicated by red arrows). Scale bar, 2 μm. (G) Sequence alignment showed that CDYL1 protein contains a putative PAR-binding motif. (H) Purification of EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) from HEK293T cells using GFP-TRAP beads. Eluted proteins were separated and stained with Coomassie. Mr indicates protein marker. (I) PAR-binding assay with EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) recombinant proteins. EGFP-NLS and EGFP-Rpp21 were used as negative and a positive controls, respectively. IB, immunoblot. 32P, phosphate radiolabeled PAR. To shed further mechanistic insights into how PARP1 regulates CDYL1 recruitment, we checked CDYL1 poly (ADP-ribosylation) before and after ionizing radiation (IR). To do so, endogenous CDYL1 protein was immunoprecipitated from undamaged and IR-damaged cells and the immunoprecipitates were immunoblotted with poly (ADP-ribose) (PAR) and CDYL1 antibodies. Results demonstrated that CDYL1 is neither PARylated before, nor after IR induction (Supplementary Figure S8). Collectively, these observations suggest that CDYL1 recruitment to DNA damage sites is not mediated by poly (ADP-ribosylation). Interestingly, bioinformatic analysis revealed that CDYL1 contains a putative PAR-binding motif encompassing amino acids 535−542 at the very end of its C-terminal region (Figure 4G). This finding prompted us to test whether CDYL1 binds to PAR moieties. Toward this end, we performed radioactive in vitro PAR-binding assay using recombinant EGFP-CDYL1bWT, EGFP-CDYL1bdel(CD), and EGFP-CDYL1bdel(ECH) (Figure 4H). Results showed that CDYL1b directly binds to PAR moieties independent of the chromodomain, but dependent on the carboxyl terminus containing the ECH domain (Figure 4I). CDYL1 promotes transcription repression at double-strand break sites Given that CDYL1 is recruited to DNA damage sites (Figure 1) and is known to function as a transcriptional repressor (Caron et al., 2003; Mulligan et al., 2008; Escamilla-Del-Arenal et al., 2013), we predicted that it is implicated in DSB-induced transcription silencing. To test this prediction, we used U2OS-TRE-I-SceI-19 reporter cell line (Awwad et al., 2017) to measure the effect of CDYL1 depletion on the transcription of MS2 gene before and after DSB induction. This single-cell assay allows monitoring nascent transcription of the MS2 gene following DSB induction upstream its promoter region using I-SceI endonuclease (Ui et al., 2015). To activate transcription of MS2 gene, cells were transfected with pCherry-tTA-ER plasmid. This plasmid expresses a cytoplasmic Cherry-tTA-ER chimera, which migrates into the nucleus upon tamoxifen (Tam) addition, where it binds to the TRE repeats and induces transcription of MS2 gene. To visualize nascent transcription of MS2, cells were co-transfected with pYFP-MS2 plasmid, which expresses YFP-MS2 protein that binds to the MS2 stem loops. To study the effect of DSB on the transcription of MS2 gene, U2OS-TRE-I-SceI-19 cells were transfected with pCMV-NLS-I-SceI plasmid expressing I-SceI endonuclease that generates DSB upstream the MS2 gene. DSB induction can be validated by immunostaining with γH2AX antibody (Ui et al., 2015). To test whether CDYL1 is implicated in transcription repression after DSB induction, U2OS-TRE-I-SceI-19 cells were transfected with control and CDYL1 siRNAs (Figure 5A). Seventy-two hours post-transfection, nascent MS2 transcripts were visualized in the absence of DSBs (by co-transfecting the cells with pCherry-tTA-ER and pYFP-MS2 plasmids) and in the presence of DSBs (by co-transfecting the cells with pCherry-tTA-ER, pYFP-MS2, and pCMV-NLS-I-SceI plasmids). Consistent with previous findings (Rafalska-Metcalf et al., 2010; Ui et al., 2015; Awwad et al., 2017), MS2 gene is transcribed in cells expressing Cherry-tTA-ER fusion as evident by the co-localization of YFP and mCherry foci (Figure 5B, top panel), and DSB induction represses MS2 expression (Figure 5B, bottom panel). Remarkably, CDYL1 depletion using two different siRNA sequences leads to MS2 expression in the presence of DSBs in ~75% of the cells, whereas has no noticeable effect on MS2 expression in the absence of DSBs (Figure 5C and D). Quantification of CDYL1 effect on MS2 expression before and after DSB induction is shown in Figure 5E. Notably, CDYL1 depletion has no detectable effect on the intensity of γH2AX at the I-SceI-induced DSB site (Figure 5F). Collectively, these observations strongly suggest that CDYL1 underpins DSB-induced transcription silencing. Figure 5 View largeDownload slide CDYL1 promotes transcription repression at double-strand break sites. (A) Western blot shows the efficacy of CDYL1 depletion in U2OS-TRE-I-SceI-19 cells using two different siRNA sequences. β-actin is used as a loading control. (B−D) CDYL1 facilitates transcription silencing at DSB sites. (B) Representative images of control U2OS-TRE-I-SceI-19 cells that were transfected with plasmids expressing the transcription activator cherry-tTA-ER and YFP-MS2 fusion and then treated with tamoxifen to induce MS2 expression (top panel). To generate DSB, the cells were co-transfected with a third plasmid expressing I-SceI endonuclease (bottom panel). (C and D) Cells were treated as in B, except that the cells were transfected with two different siRNA sequences against CDYL1. Scale bar, 2 μm. (E) Graph displays the percentage of γH2AX-positive cells that show co-localization of YFP-MS2 and Cherry-tTA-ER. Error bars represent SD from four independent experiments (n > 60). P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001 (F) Quantitative measurements show that CDYL1 depletion has no detectable effect on the fluorescence intensity of γH2AX foci at DSB sites. Red horizontal bars indicate the mean fluorescence intensity of γH2AX in cells. Figure 5 View largeDownload slide CDYL1 promotes transcription repression at double-strand break sites. (A) Western blot shows the efficacy of CDYL1 depletion in U2OS-TRE-I-SceI-19 cells using two different siRNA sequences. β-actin is used as a loading control. (B−D) CDYL1 facilitates transcription silencing at DSB sites. (B) Representative images of control U2OS-TRE-I-SceI-19 cells that were transfected with plasmids expressing the transcription activator cherry-tTA-ER and YFP-MS2 fusion and then treated with tamoxifen to induce MS2 expression (top panel). To generate DSB, the cells were co-transfected with a third plasmid expressing I-SceI endonuclease (bottom panel). (C and D) Cells were treated as in B, except that the cells were transfected with two different siRNA sequences against CDYL1. Scale bar, 2 μm. (E) Graph displays the percentage of γH2AX-positive cells that show co-localization of YFP-MS2 and Cherry-tTA-ER. Error bars represent SD from four independent experiments (n > 60). P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001 (F) Quantitative measurements show that CDYL1 depletion has no detectable effect on the fluorescence intensity of γH2AX foci at DSB sites. Red horizontal bars indicate the mean fluorescence intensity of γH2AX in cells. CDYL1 promotes EZH2 accumulation and H3K27me3 induction at DNA breakage sites To gain molecular insights into the mechanism by which CDYL1 facilitates transcription repression at DSB sites, we monitored the localization of several repressive factors that were previously shown to be involved in DNA damage-induced transcription silencing in control and CDYL1-deficient cells. Since CDYL1 directly interacts with HDAC2 (Mulligan et al., 2008), which is recruited to DNA damage sites (Miller et al., 2010), we anticipated that CDYL1 might regulate HDAC2 recruitment to damaged chromatin. As shown in Supplementary Figure S9A and B, CDYL1 depletion has no detectable effect on HDAC2 accumulation at DNA damage sites. Afterward, we monitored the recruitment of CHD4, which is known to accumulate at DNA damage sites and promote the recruitment of repressive factors to damaged chromatin (Chou et al., 2010; Larsen et al., 2010; Polo et al., 2010; Xia et al., 2017). Similar to HDAC2, CHD4 is recruited to DNA damage sites independent of CDYL1 (Supplementary Figure S9C and D). Then, we looked at the repressive epigenetic mark H2A-K119 monoubiquitination, which is known to be induced locally at DNA damage sites to promote transcription silencing (Bergink et al., 2006; Marteijn et al., 2009; Zhu et al., 2009; Wu et al., 2009b; Ui et al., 2015). Results showed that monoubiquitination of H2A-K119 at DNA damage sites was also not affected by CDYL1 depletion (Supplementary Figure S9E and F). Since EZH2 is recruited to DNA breakage sites (Chou et al., 2010; Campbell et al., 2013) and interacts with CDYL1 (Zhang et al., 2011), we sought to determine whether CDYL1 regulates EZH2 recruitment to damage sites. To do so, we induced DSB using I-SceI in control and CDYL1-depleted U2OS-TRE-I-SceI-19 cells (Figure 6A), followed by ChIP analysis using EZH2 and γH2AX antibody as a positive control for DSB induction. Results showed that CDYL1 depletion impairs EZH2 accumulation at DSB sites (Figure 6B). Altogether, these observations demonstrated that CDYL1 underpins EZH2 recruitment to DSB sites. The fact that EZH2 catalyzes the methylation of H3K27me3, a repressive mark known to be induced at DNA damage sites (Chou et al., 2010), prompted us to test whether the level of H3K27me3 at damage sites is affected by CDYL1. Results showed that while CDYL1 depletion has no detectable effect on global H3K27me3 levels (Figure 6C), it diminishes the local induction of H3K27me3 at laser-microirradiated sites at 5, 60, and 120 min after damage (Figure 6D and E). Collectively, our data suggest that CDYL1 fortifies DSB-induced transcription silencing by promoting EZH2 compiling and H3K27 trimethylation at laser-microirradiated sites. Figure 6 View largeDownload slide CDYL1 promotes the accumulation of EZH2 and H3K27me3 at DNA damage sites. (A) Western blot shows siRNA knockdown of CDYL1. U2OS-TRE-I-SceI-19 cells were transfected with CDYL1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (B) CDYL1 knockdown abrogates EZH2 recruitment to DSB sites. ChIP was performed as described in Figure 1I using IgG, EZH2, and γH2AX antibodies. (C) Western blot shows that CDYL1b depletion has no detectable effect on the global levels of H3K27me3 mark. (D) CDYL1 depletion impairs the increase in H3K27me3 at laser-microirradiated sites. U2OS cells transfected with control or CDYL1 siRNA were subjected to laser microirradiation and co-immunostained with γH2AX (green) and H3K27me3 (red) antibodies at the indicated time points. DNA is stained with Hoechst (blue). Scale bar, 2 μm. (E) Graph displays H3K27me3 fluorescence intensity normalized to background in mock and CDYL1-depleted U2OS cells at the indicated time points after laser microirradiation. Error bars represent SD from three independent experiments (n > 60). P-values were calculated by two-sided Student’st-test relative to Ctrl siRNA. **P < 0.01. Figure 6 View largeDownload slide CDYL1 promotes the accumulation of EZH2 and H3K27me3 at DNA damage sites. (A) Western blot shows siRNA knockdown of CDYL1. U2OS-TRE-I-SceI-19 cells were transfected with CDYL1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (B) CDYL1 knockdown abrogates EZH2 recruitment to DSB sites. ChIP was performed as described in Figure 1I using IgG, EZH2, and γH2AX antibodies. (C) Western blot shows that CDYL1b depletion has no detectable effect on the global levels of H3K27me3 mark. (D) CDYL1 depletion impairs the increase in H3K27me3 at laser-microirradiated sites. U2OS cells transfected with control or CDYL1 siRNA were subjected to laser microirradiation and co-immunostained with γH2AX (green) and H3K27me3 (red) antibodies at the indicated time points. DNA is stained with Hoechst (blue). Scale bar, 2 μm. (E) Graph displays H3K27me3 fluorescence intensity normalized to background in mock and CDYL1-depleted U2OS cells at the indicated time points after laser microirradiation. Error bars represent SD from three independent experiments (n > 60). P-values were calculated by two-sided Student’st-test relative to Ctrl siRNA. **P < 0.01. CDYL1 depletion triggers DNA damage-induced G2/M arrest and alters the phosphorylation of DNA damage markers Mammalian cells respond to DNA-damaging agents by activating cell cycle checkpoints that lead to a temporary pause at a specific stage of the cell cycle to allow DNA damage repair (Guleria and Chandna, 2016). Flow cytometric analysis revealed that CDYL1-deficient cells (Figure 7A) accumulate in G2/M and exhibit persistent cell cycle arrest at 12 and 24 h after IR, likely due to defective DNA damage repair (Figure 7B). To clarify whether CDYL1 is implicated in DSB repair, we monitored γH2AX levels (earliest known marker of DSBs) in control and CDYL1-depleted cells. Results showed that CDYL1-deficient cells exhibit elevated levels of γH2AX at different time points after IR (Figure 7C). This finding indicates that CDYL1-depleted cells show elevated levels of damaged DNA, possibly due to inefficient repair of IR-induced DSBs. To further investigate the role of CDYL1 in DSB repair, we tested the levels of RPA2 phosphorylation at Ser4/Ser8, a marker for ssDNA generation following DNA-end resection that promotes HDR of DSB, after CPT treatment that causes DSBs by inhibiting topoisomerase I and promoting collision of replication forks. Results demonstrated defective phosphorylation of RPA2 in CDYL1-deficient cells when compared to control cells (Figure 7D), suggesting that CDYL1 may regulate DNA-end resection, a prerequisite step for HDR of DSBs. Figure 7 View largeDownload slide CDYL1 depletion causes G2/M arrest, alters the activation of DNA damage markers, and disrupts the integrity of homology-directed repair of DSBs. (A) Constitutive depletion of CDYL1 using specific shRNA sequence. U2OS cells were stably infected with lentiviral vectors expressing either scramble or CDYL1 shRNA and subjected to western blot analysis using the indicated antibodies. (B) CDYL1 knockdown leads to persistent G2/M arrest after IR. Control and CDYL1-deficient cells were exposed to IR (3 Gys) and samples were collected at the indicated time points for cell cycle analysis by flow cytometry. (C) Western blot analysis shows that CDYL1-deficient U2OS cells have elevated levels of γH2AX in response to IR. (D) Western blot shows that CDYL1 knockdown decreases CPT-induced levels of pRPA2 S4/S8 phosphorylation. Total RPA and β-actin were used as internal controls. The numbers below the blot indicate the relative intensity of pRPA bands, which was normalized to the intensity of total RPA band. (E) Western blot shows siRNA knockdown of CDYL1 in U2OS-TLR cells. (F) U2OS-TLR cells show that CDYL1 depletion impairs HDR of DSBs generated by I-SceI endonuclease. A reduction of ~70%−80% in GFP-positive cells was observed after CDYL1b depletion. Caffeine was used as a positive control. Given that HDR can occur only in S/G2 cell phase, data were corrected to flow-cytometry S/G2 values. (G) TLR results for DSBs repair by NHEJ. CDYL1 knockdown has no significant effect on the integrity of NHEJ. Data represent SD of three independent experiments. All the results are typical of three independent experiments. Error bars represent SD. P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001. Figure 7 View largeDownload slide CDYL1 depletion causes G2/M arrest, alters the activation of DNA damage markers, and disrupts the integrity of homology-directed repair of DSBs. (A) Constitutive depletion of CDYL1 using specific shRNA sequence. U2OS cells were stably infected with lentiviral vectors expressing either scramble or CDYL1 shRNA and subjected to western blot analysis using the indicated antibodies. (B) CDYL1 knockdown leads to persistent G2/M arrest after IR. Control and CDYL1-deficient cells were exposed to IR (3 Gys) and samples were collected at the indicated time points for cell cycle analysis by flow cytometry. (C) Western blot analysis shows that CDYL1-deficient U2OS cells have elevated levels of γH2AX in response to IR. (D) Western blot shows that CDYL1 knockdown decreases CPT-induced levels of pRPA2 S4/S8 phosphorylation. Total RPA and β-actin were used as internal controls. The numbers below the blot indicate the relative intensity of pRPA bands, which was normalized to the intensity of total RPA band. (E) Western blot shows siRNA knockdown of CDYL1 in U2OS-TLR cells. (F) U2OS-TLR cells show that CDYL1 depletion impairs HDR of DSBs generated by I-SceI endonuclease. A reduction of ~70%−80% in GFP-positive cells was observed after CDYL1b depletion. Caffeine was used as a positive control. Given that HDR can occur only in S/G2 cell phase, data were corrected to flow-cytometry S/G2 values. (G) TLR results for DSBs repair by NHEJ. CDYL1 knockdown has no significant effect on the integrity of NHEJ. Data represent SD of three independent experiments. All the results are typical of three independent experiments. Error bars represent SD. P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001. CDYL1 exclusively promotes homology-directed repair of double-strand breaks To confirm a direct role of CDYL1 in DSB repair, we engineered a traffic light reporter (TLR) system (Certo et al., 2011) integrated into the genome of U2OS cells (Abu-Zhayia et al., 2017). The observed reporter cell line, hereafter called U2OS-TLR, has an I-SceI recognition site integrated within a GFP-reporter cassette and out-of-frame mCherry gene. While DSB repair by homologous recombination results in reconstitution of a functional GFP gene, manifested in green-colored cells, error-prone repair of DSBs by NHEJ introduces a frame-shift within the out-of-frame mCherry gene, leading to the expression of mCherry (Abu-Zhayia et al., 2017). Remarkably, CDYL1 depletion in U2OS-TLR cells using two different siRNAs (Figure 7E) resulted in 70%−80% decrease in the number of GFP-positive cells, when compared to cells treated with control siRNA (Figure 7F). This decrease demonstrated that CDYL1 is required for intact HDR of DSBs. Notably, the reduction of HDR in cells was accompanied by mild increase in mCherry-positive cells (Figure 7G), an indication that CDYL1 depletion had no deleterious effect on the integrity of NHEJ. Altogether, we concluded that CDYL1 exclusively promotes HDR but not NHEJ of DSBs. CDYL1 depletion sensitizes cells to the chemotherapeutic drug cisplatin Interstrand cross-linking agents, such as cisplatin, cause DNA adducts that can be repaired by various mechanisms including HDR (Dronkert and Kanaar, 2001). Therefore, HDR-deficient cells exhibit hypersensitivity to cisplatin (D’Andrea and Grompe, 2003). For example, it was shown that knocking down BRCA1 and BRCA2 sensitizes cells to cisplatin (Bartz et al., 2006). Since CDYL1 depletion disrupted the integrity of HDR (Figure 7F), we sought to examine the sensitivity of CDYL1-depleted cells to cisplatin. Toward this end, we first knocked out CDYL1 gene in non-malignant mammary epithelial cells, MCF10A, using CRISPR-Cas9 methodology. MCF10A cells were co-transfected with vectors expressing EGFP-Cas9 and specific guide-RNAs targeted against exon 4 of CDYL1. CDYL1 knockout was confirmed by DNA sequencing (Figure 8A), western blot (Figure 8B) and immunofluorescence analysis (Figure 8C). Next, control and CDYL1-knockout MCF10A cells were treated with increasing concentrations of cisplatin and subjected to short-term growth delay assay. Results showed that CDYL1-depleted cells are pronouncedly more sensitive to cisplatin when compared to control cells (Figure 8D). To validate that the hypersensitivity of CDYL1-depleted cells to cisplatin is caused by CDYL1 depletion, CDYL1-knockout MCF10A cells were infected with lentiviral vector expressing CDYL1 protein (Figure 8E) and tested for cisplatin sensitivity. As seen in Figure 8D, introducing CDYL1 suppresses the hypersensitivity to cisplatin, confirming that CDYL1 depletion indeed sensitizes cells to cisplatin. Altogether, we concluded that CDYL1-deficient cells exhibit synthetic lethality with cisplatin. Figure 8 View largeDownload slide CDYL1-knockout MCF10A cells display synthetic lethality with cisplatin. (A) DNA sequence alignment shows the deletion and the mismatches (in red) in exon 4 of CDYL1 gene at the 3′ of the gRNA sequence. (B and C) Control and CDYL1-knockout MCF10A cells were subjected to western blot (B) and immunofluorescence analysis (C). (D) CellTiter-Glo viability assay in wild-type, CDYL1-deficient cells, and CDYL1-deficient cells infected with lentiviral vector expressing CDYL1b. Cells were treated with increasing concentrations of cisplatin for 96 h. Data are representative of three independent experiments. Error bars indicate SD. Two-way ANOVA was used to test for differences at each dose. **P < 0.01. (E) Lentiviral vector expressing CDYL1b was used to restore the expression of CDYL1b in CDYL1-deficient MCF10A cells. Western blot analysis shows that the restored protein level of CDYL1 is comparable to the endogenous level in control MCF10A cells. Figure 8 View largeDownload slide CDYL1-knockout MCF10A cells display synthetic lethality with cisplatin. (A) DNA sequence alignment shows the deletion and the mismatches (in red) in exon 4 of CDYL1 gene at the 3′ of the gRNA sequence. (B and C) Control and CDYL1-knockout MCF10A cells were subjected to western blot (B) and immunofluorescence analysis (C). (D) CellTiter-Glo viability assay in wild-type, CDYL1-deficient cells, and CDYL1-deficient cells infected with lentiviral vector expressing CDYL1b. Cells were treated with increasing concentrations of cisplatin for 96 h. Data are representative of three independent experiments. Error bars indicate SD. Two-way ANOVA was used to test for differences at each dose. **P < 0.01. (E) Lentiviral vector expressing CDYL1b was used to restore the expression of CDYL1b in CDYL1-deficient MCF10A cells. Western blot analysis shows that the restored protein level of CDYL1 is comparable to the endogenous level in control MCF10A cells. Discussion Here, we describe a novel role of CDYL1 in DDR. We unprecedentedly demonstrated that CDYL1 is recruited to DSB sites in a PARP1-dependent manner, and that the C-terminal region, containing the ECH domain, binds to PAR moieties in vitro (Figure 4). We speculate therefore that CDYL1 accumulation at damage sites is mediated via binding to PAR moieties, which are known to be locally induced at DNA damage sites and form a docking platform for recruiting DNA repair proteins (Gibson and Kraus, 2012; Khoury-Haddad et al., 2014, 2015; Tallis et al., 2014; Abu-Zhayia et al., 2017; Awwad et al., 2017). CDYL1 undergoes phosphorylation in response to DNA damage on consensus site recognized by ATM and ATR (Matsuoka et al., 2007). However, we found that this phosphorylation is dispensable for its accumulation at DNA damage sites (Supplementary Figure S7). In line with this, CDYL1 is recruited irrespective of ATM activity (Supplementary Figure S6). On this basis, we anticipate that the DNA damage-induced phosphorylation may regulate CDYL1 activity at DNA breakage sites or modulate its interactome after damage. Previous report showed that CDYL1 foci colocalize with the heterchromatic region mark, H3K9me3 (Franz et al., 2009). We confirmed this finding and demonstrated that CDYL1 colocalizes also with bright DAPI foci and HP1α protein (Figure 2A−C). Altogether, these observations provide compelling evidence that CDYL1 foci represent heterochromatic regions. It should be noted, however, that laser microirradiation of heterochromatic regions may hit also neighboring euchromatic regions. Nevertheless, only negligible increase was observed in the fluorescence intensity of EGFP-CDYL1b at laser-microirradiated heterochromatic regions (Figure 2D and F). The lack of CDYL1b accumulation at damaged heterochromatin might be due to the fact that CDYL1b is enriched at heterchromatic regions regardless of DNA damage and therefore there is no need for further accumulation after damage induction. We and others recently showed that PARP1 activity promotes transcription repression at DNA damage sites by recruiting the repressive complexes NuRD and PcG (Chou et al., 2010) and the negative transcription elongation factor, NELF-E (Awwad et al., 2017; Polo, 2017). Here, we identified CDYL1 as an additional PARP1-regulated protein that contributes to DSB-induced transcription silencing at damage sites. Future studies will be required to clarify whether the recruitment of those repressive factors depends on each other and whether they have overlapping functions or work in different pathways or in different chromatin environment to achieve transcription silencing at DSB sites. Interestingly, since CDYL1 is not recruited to UV-induced damage sites (Supplementary Figure S4), we propose therefore that CDYL1 may not contribute to UV-induced transcription repression. Our data favor a model that CDYL1 enhances the recruitment of the endogenous H3K27 methyltransferase, EZH2, to DSB sites and subsequently leads to an increase in H3K27me3 methylation at DNA breakage sites. Similar findings showing CDYL1-dependent increase in EZH2 and H3K27me3 levels at I-SceI-induced DSBs were recently reported (Liu et al., 2017b). Notably, EZH2 accumulation at DSB sites was only observed by ChIP methodology. Several attempts to visualize EGFP-EZH2 enrichment at laser-microirradiated sites were proved unsuccessful (data not shown). Possible reasons for this variation may arise from the overexpression or the fusion of EZH2 with EGFP tag. Using a sophisticated reporter cell line, we showed that CDYL1 exclusively facilitates HDR of DSBs (Figure 7F). The fact that CDYL1 is recruited to DNA damage sites favors a scenario where CDYL1 directly regulates DSB repair. However, we cannot rule out a possibility that the defective HDR in CDYL1-depleted cells could also result from alterations in the expression of CDYL1 target genes (Zhang et al., 2011). Future studies will be required to shed mechanistic insight into how CDYL1 regulates HDR of DSBs. Interestingly, the reduction in HDR was not accompanied by an increase in NHEJ suggesting that the excess DSBs are not repaired by the classical NHEJ. We predict therefore that the resulted DSBs could be fixed by alternative repair pathways such as microhomology-mediated end-joining (MMEJ) and synthesis-dependent (SD)-MMEJ (Kostyrko and Mermod, 2016) or by RNA-templated DSB repair (Shen et al., 2011). Inhibiting DNA repair factors has become an attractive therapeutic concept in cancer therapy (Kelley et al., 2014). Our observations demonstrated that CDYL1 depletion increases the sensitivity of breast epithelial cells to cisplatin (Figure 8), which is one of the most commonly used chemotherapeutic drugs. On this basis, we propose that cisplatin treatment might be particularly effective to preferentially eradicate cancer cells harboring mutations in CDYL1 gene. Strikingly, CDYL1 mutations have been found in 13% of 1787 breast cancer cases and in 19.5% of 1287 hepatocellular cancer cases (http://icgc.org/). Future work will be required to measure the efficacy of cisplatin inhibition on the growth of CDYL1-deficient tumor cells in vivo. Material and methods Plasmids pEGFP-C1-CDYL1b, CDYL1bdel(CD), CDYL1bdel(Hinge), CDYL1bdel(ECH), CDYL1-S160A, MonoRed-CDYL1b, pLKO.1-TRC-CDYL1-shRNA, and MonoRed-CtIP were constructed as described in Supplementary Table S1. A complete list of all primers and their sequences is described in Supplementary Table S2. pSpCAS9 (BB)-2A-GFP (PX458; #48183) and PsPgRNA (#47108) vectors were purchased from Addgene. All constructs used in this study were verified by nucleotide sequencing or restriction digestion. mAG-hGeminin (1/110) plasmid expressing EGFP-Geminin was a kind gift from Dr Atsushi Miyawaki (RIKEN Brain Science Institute, Japan). pCSC-SP-PW-(GENE)-IRES/GFP lentiviral plasmid and packaging plasmids were a kind gift from Dr Hava Gil-Henn (Bar-Ilan University, Israel). Cell lines All cell lines used in this study were cultured in media supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 100 units/ml penicillin, and 100 μg/ml penicillin/streptomycin. U2OS, MEF, HEK293T, and U2OS-TRE-I-SceI-19 cell lines were grown in Dulbecco’s modified Eagle’s medium (Gibco). MCF-7 cell line was cultured in RPMI-1640 media (Gibco). MCF10A cell line was cultured in DMEM/F12 media (Invitrogen) supplemented with 5% horse serum (Invitrogen), 20 ml EGF, 0.5 mg/ml hydrocortisone, 100 ng/ml cholera toxin, 10 μg/ml insulin, 100 μg/ml penicillin/streptomycin, and 2 mM L-glutamine. Stable U2OS-TLR cells were grown in media containing 0.6 μg/ml puromycin. Transfections and drug treatments Cell transfections with DNA plasmids and siRNAs (Supplementary Table S3) were performed using PolyJet transfection reagent (BioConsult) and Lipofectamine2000 reagent (Invitrogen), respectively, following the manufacturer’s instructions. Where mentioned, cells were treated with 4 mM Caffeine (Sigma; C0750) for 72 h, 1 μM PARP inhibitor (Ku-0059436) for 1 h, 5 μM ATM inhibitor (KU-55933) for 2 h, 1 μM Camptothecin (CPT) (C9911; Sigma) for 2 h. Cisplatin (Selleck S1166) was added at the indicated concentrations. Where indicated, cells were exposed to ionizing radiation (IR) using the X-ray machine (Faxitron, CellRad). Western blot Western blot analysis was performed as previously described (Khoury-Haddad et al., 2014). Briefly, protein extracts were prepared using Hot-lysis buffer (1% SDS, 5 mM EDTA, 50 mM Tris, pH 7.5) and protease inhibitor mixture (Calbiochem). Samples were separated on SDS-PAGE gel and membranes were immunoblotted with the relevant antibodies. A complete list of antibodies and their dilutions is described in Supplementary Table S4. The immunoblots were developed using Quantum ECL detection kit (K-12042-D20, Advansta). The intensity of the immunoblot bands was determined using ImageJ software. Immunoprecipitation Control and CDYL1-depleted U2OS cells were left untreated or exposed to 10 Gys of IR and kept for 5 min to recover. Immunoprecipitation was performed as previously described (Escamilla-Del-Arenal et al., 2013) with the following modifications. Cells were fractionated using Buffer A (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl2, 10% glycerol, 0.34 M sucrose, 1 mM DTT, 0.1% Triton, PMSF, and protease inhibitor mixture) for 5 min on ice. Nuclear fraction was purified by centrifugation and incubated with lysis buffer (50 mM HEPEs, pH 7.4, 150 mM NaCl, 0.5% NP-40, 10 mM EDTA, 1 mM DTT) and Benzonase (Novagen) for 1 h on ice, followed by 30 min centrifugation. Supernatants were subjected to overnight immunoprecipitation (IP) using 1 μg of CDYL1 antibody and protein A magnetic beads (GenScript). The immune-complexes were washed, and subjected to western blot analysis. Immunofluorescence Cells were grown on coverslips for 24 h and subjected to immunofluorescence as previously described (Khoury-Haddad et al., 2014). Cells were immunostained with the appropriate antibodies (Supplementary Table S4). Slides were visualized using the inverted Zeiss LSM-700 confocal microscope with 40× oil EC Plan Neofluar objective. Laser microirradiation Cells were subjected to laser microirradiation as previously described (Khoury-Haddad et al., 2014). Briefly, cells were plated on flourodish (Ibidi; Cat#81158) and pre-sensitized with 1 μg/μl Hoechst 3334 dye for 10 min at 37°C. Laser microirradiation was executed using an LSM-700 inverted confocal microscope equipped with CO2 module and 37°C heating chamber. DNA damage was induced by micro-irradiating a single region in the nucleus with 15 iterations of 405 nm laser beam. Time-laps images were acquired using 488 nm laser. Signal intensity at damaged sites was measured using Zen 2009 software. In Figures 1E−G, 4C, D, 6D and Supplementary Figures S3 and S9, cells were plated on gridded plates (Ibidi; Cat#80826). Prior to 405 nm laser-microirradaition, cells were pre-sensitized with Hoechst for 5 min. After damage induction cells were pre-extracted with CSK buffer (10 mM HEPES-KOH, pH 7.9, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 1 mM EGTA, and 0.5% v/v Triton X-100) for 5 min on ice, fixed with 4% PFA for 10 min at room temperature and stained with the indicated antibodies. Generation of lentiviral particles and cell transduction In order to knockdown CDYL1 expression in U2OS and U2OS-TRE-I-SceI-19 cells, shRNA sequence was used as previously described (Awwad et al., 2017). First, scramble short hairpin oligonucleotides and short hairpin oligonucleotide directed against CDYL1 were annealed and inserted into pLKO.1-TRC lentiviral vector digested with EcoRI and AgeI. The generated lentiviral vectors were verified by nucleotide sequencing. Viral particles containing the shRNA construct were generated by transfecting HEK293T cells together with plasmids encoding the lentiviral proteins Gag, Pol, and VSV-G. Viral particles expressing CDYL1b protein were generated by co-transfecting HEK293T cells with pCSC-SP-PW-CDYL1b-IRES/GFP plasmid together with three plasmids expressing pRSV-REV, pMDL, and VSVG. Media containing the viral particles were collected 48 h post-transfection and filtered with 0.45 μm filters. Then, the viral particles were used to infect the indicated cell lines. At 72 h post-infection, cells were selected with 1 μg/ml puromycin for 1 week. Cell cycle analysis by flow cytometry Flow cytometric analysis was performed as previously described (Khoury-Haddad et al., 2014). Briefly, cells were fixed with ice-cold 75% ethanol. DNA was stained with 100 mg/ml propidium iodide (Sigma-Aldrich) in phosphate buffer solution (PBS) containing 0.1% Triton-X-100 and 0.5 mg/ml DNase free RNase A (Sigma-Aldrich). Samples were analyzed using flow cytometry of 10000 events on a BD LSR-II flow cytometer (Becton Dickinson). Data were analyzed with FCS express software. Traffic light reporter assay TLR assay was performed as previously described (Certo et al., 2011; Schmidt et al., 2015; Abu-Zhayia et al., 2017). In brief, U2OS-TLR cells were transfected with control or CDYL1 siRNAs. Ten hours later, cells were co-transfected with plasmids expressing I-SceI nuclease fused to infrared fluorescent protein (IFP) and donor plasmid expressing GFP donor sequence fused to blue fluorescent protein (BFP). Seventy-two hours after siRNA transfection, cells were harvested, and GFP and mCherry signals (reflecting HDR and NHEJ, respectively) were measured by four-color fluorescent flow-cytometry using a BD LSRFortessa™ cell analyzer (BD Biosciences). Minimum 10000 double-positive (IFP and BFP) cells were scored for each condition from three independent experiments. Results of siRNA-transfected cells were normalized to control siRNA-transfected cells. HDR values for each condition were normalized to the percentage of cells at S and G2 phase monitored by propidium iodide-based standard flow-cytometry. Chromatin fractionation Biochemical fractionation was performed as previously described (Khoury-Haddad et al., 2014) with the following modifications. U2OS cells were left untreated or exposed to 10 Gys of IR followed by 5 min and 30 min recovery. Cells were lysed with Buffer A for 5 min at 4°C. Cell lysates were centrifuged at 1500× g for 5 min at 4°C and the supernatant was removed. Then, the pellet was incubated with Buffer B (3 mM EDTA, 0.2 mM EGTA, 1 mM DTT, PMSF, and protease inhibitor mixture) for 10 min on ice followed by centrifugation at 1700× g for 5 min at 4°C. To prepare chromatin-bound fraction, pellet was resuspended with Hot-lysis buffer, boiled for 15 min, and sonicated with two 15-sec pulses of 35% amplitude. treated with Benzonase (Novagen) for 30 min at room temperature, centrifuged at maximum speed for 20 min at 12°C, and the supernatant was recovered. Chromatin-bound fractions were subjected to western blot analysis and immunoblotted with the indicated antibodies. Short-term growth delay assay Cells were seeded in 96-well plates at a density of 3000 per well (Thermo) and left for 1 h at room temperature to adhere before being returned to the incubator. At 24 h after seeding, cisplatin was added at the indicated concentrations. Cell viability was measured 96 h after drug treatment using the CellTiter-Glo Kit (Promega), following the manufacturer’s protocol. Visualizing MS2 expression before and after DSB in U2OS-TRE-I-SceI-19 The effect of DSB on the transcription of MS2 gene was monitored as previously described (Awwad et al., 2017). Briefly, U2OS-TRE-I-SceI-19 cells were transfected with pCherry-tTA-ER plasmid, which expresses a cytoplasmic Cherry-tTA-ER chimera, and treated with 1 μM tamoxifen to drive its migration into the nucleus and induce transcription of MS2 gene. To visualize nascent transcripts of MS2, cells were co-transfected with pYFP-MS2 plasmid, which expresses YFP-MS2 protein that binds to the MS2 stem loops. To generate DSB, U2OS-TRE-I-SceI-19 cells were co-transfected with pCMV-NLS-I-SceI. Generation of CDYL1-knockout MCF10A cell line using CRISPR/Cas9 methodology MCF10A cells were co-transfected with pSpCAS9 (BB)-2A-GFP vector expressing GFP-Cas9 and PsPgRNA-CDYL1gRNA vector containing a specific gRNA to introduce DSB within CDYL gene. At 24 h after transfection, GFP-positive cells were sorted using BD LSRFortessa™ cell analyzer (BD Biosciences) and plated in 96-well plates at a dilution of one cell per well. Clones were first screened by western blot analysis. CDYL1-knockout clones were further validated by sequencing and immunofluorescence. Chromatin immunoprecipitation ChIP experiment was carried out as previously described (Ui et al., 2015; Awwad et al., 2017). Briefly, U2OS/TRE/I-SceI-19 cells were plated in 150-mm dishes and transfected with pCherry-rTA-ER plasmid only or together with a plasmid expressing I-SceI endonuclease, pCMV-NLS-I-SceI. At 24 h following transfection, cells were treated with 1 μM tamoxifen for 2 h. Cells were then crosslinked with 1% PFA for 10 min at room temperature, and cross-linking was stopped with 0.125 M Glycine for 5 min. After cell lysis, DNA was sheared to the size of 300–500 bp using a Vibra cell sonicator (15 sec ON, 30 sec OFF, 35% duty, 20 cycles). Five percent of each supernatant was used as input control and processed with the cross-linking reversal step. The rest of the supernatant was subjected to overnight immunoprecipitation (IP) using either 1 μg of CDYL1, γH2AX antibody (Millipore 05-636) or EZH2 and protein A magnetic beads (GenScript). Following reverse cross-linking; the precipitated DNA was purified using the PureLinkTM PCR Micro Kit. Quantification of the immunoprecipitated DNA was carried out by Step-One-Plus real-time PCR using Fast SYBR Green Master mix (Applied Biosystems) and the primers around the transcription start sites (TSS), GACGTAAACGGCCACAAGTT and GAACTTCAGGGTCAGCTTGC (80 bp downstream of TSS). Fold induction of binding surrounding the break site was calculated using the CT cycles in which untreated and treated IP samples values were normalized to the no-antibody control (IgG). PAR-binding assay Recombinant proteins including EGFP-CDYL1WT, CDYL1bdel(CD), and CDYL1bdel(ECH) were overexpressed in HEK293T cells and immunoprecipitated using GFP-TRAP beads following the manufacturer’s guidelines (Chromotek). The immunoprecipitated proteins were tested for their ability to bind to PAR moieties using the PAR-binding assay, as previously specified (Khoury-Haddad et al., 2015). Briefly, 1–5 pmol of proteins were blotted onto a nitrocellulose membrane and blocked with TBST buffer supplemented with 5% milk. Radioactively labeled PAR moieties were produced by auto-modified PARP1 prepared by in vitro PARylation reaction. This reaction was performed at room temperature for 20 min in a reaction buffer (50 mM Tris-HCl, pH 8, 25 mM MgCl2, 50 mM NaCl) supplemented with radiolabelled NAD+ (Perkin Elmer), activated DNA, and PARP1 enzyme (Trevigen). PAR moieties were detached from PARP1 using proteinase K and blotted membrane was incubated for 2 h with the radiolabelled PAR diluted in TBST buffer. Membranes were then washed with TBST, subjected to autoradiography and western blotting using GFP antibody. Localized UV damage Induction of localized UV irradiation was done as described in Mone et al. (2001). Briefly, cells were grown on coverslips, washed once with 1× PBS, covered by isopore polycarbonate membrane filter (Millipore), and then cells were irradiated with 150 J/m2 UV light. The filter was then removed and cells were fixed or incubated for additional recovery time. Cells were co-stained for CDYL1 and cyclobutane pyrimidine dimers (CPD). Slides were visualized using the inverted Zeiss LSM 700 confocal microscope with 40× oil EC Plan Neofluar objective. Statistical analysis Statistical analyses were performed using the demo version of GRAPHPAD prism software version. Addendum While this manuscript was in preparation, similar observations implicating CDYL1 in DSB repair were reported (Liu et al., 2017b). Supplementary material Supplementary material is available at Journal of Molecular Cell Biology online. Acknowledgements We thank Noga Gutmann-Raviv, Rami Aqeilan, Arnon Henn, Oded Kleifeld, and Yossi Shiloh for critical discussion of the manuscript. We thank Jacob Hanna (Weizmann Institute), Sarah Selig (Technion), Amir Orian (Technion), Hava Henn (Bar-Ilan University), Yehuda Assaraf (Technion), and Aaron Ciechanover (Technion) for providing plasmids and antibodies. Funding This work was supported by grants from the Israel Science Foundation (ISF, Grant no. 2021242), the Israel Cancer Association (Grant no. 2019404), the Binational Science Foundation (Grant no. 2023065), the Israel Cancer Research Fund (ICRF, Grant no. 2021762), and Volkswagen Foundation (Grant no. 2020594). E.R.A.-Z. and S.W.A. are supported by the Council for Higher Education 19 fellowship for outstanding minority M.Sc. and Ph.D. students, respectively. N.A. is supported by the Neubauer Family Foundation. Conflict of interest: none declared. Author contributions: E.R.A.-Z. performed the experiments described in this study (except the one indicated below), wrote the experimental procedures, and helped in proofreading the manuscript. S.W.A. performed the experiments described in Figures 4H, 4I, 8D and Supplementary Figures S4, S6C, S6D, S9 and helped in proofreading the manuscript. B.M.B.-O. helped in performing the experiments described in Figure 8E, constructed several plasmids used in this study, and helped in proofreading the manuscript. H.K.-H. performed the experiment described in Figure 3A and Supplementary Figure S5 and helped in proofreading the manuscript. N.A. conceived the study, planed the experiments, and wrote the manuscript. References Abu-Zhayia, E.R., Khoury-Haddad, H., Guttmann-Raviv, N., et al.  . ( 2017). A role of human RNase P subunits, Rpp29 and Rpp21, in homology directed-repair of double-strand breaks. Sci. Rep.  7, 1002. Google Scholar CrossRef Search ADS PubMed  Adam, S., and Polo, S.E. ( 2014). Blurring the line between the DNA damage response and transcription: the importance of chromatin dynamics. Exp. Cell Res.  329, 148– 153. Google Scholar CrossRef Search ADS PubMed  Awwad, S.W., Abu-Zhayia, E.R., Guttmann-Raviv, N., et al.  . ( 2017). NELF-E is recruited to DNA double-strand break sites to promote transcriptional repression and repair. EMBO Rep.  18, 745– 764. Google Scholar CrossRef Search ADS PubMed  Bartz, S.R., Zhang, Z., Burchard, J., et al.  . ( 2006). Small interfering RNA screens reveal enhanced cisplatin cytotoxicity in tumor cells having both BRCA network and TP53 disruptions. Mol. Cell. Biol.  26, 9377– 9386. Google Scholar CrossRef Search ADS PubMed  Bennetzen, M.V., Larsen, D.H., Bunkenborg, J., et al.  . ( 2010). Site-specific phosphorylation dynamics of the nuclear proteome during the DNA damage response. Mol. Cell. Proteomics  9, 1314– 1323. Google Scholar CrossRef Search ADS PubMed  Bergink, S., Salomons, F.A., Hoogstraten, D., et al.  . ( 2006). DNA damage triggers nucleotide excision repair-dependent monoubiquitylation of histone H2A. Genes Dev.  20, 1343– 1352. Google Scholar CrossRef Search ADS PubMed  Campbell, S., Ismail, I.H., Young, L.C., et al.  . ( 2013). Polycomb repressive complex 2 contributes to DNA double-strand break repair. Cell Cycle  12, 2675– 2683. Google Scholar CrossRef Search ADS PubMed  Caron, C., Pivot-Pajot, C., van Grunsven, L.A., et al.  . ( 2003). Cdyl: a new transcriptional co-repressor. EMBO Rep.  4, 877– 882. Google Scholar CrossRef Search ADS PubMed  Certo, M.T., Ryu, B.Y., Annis, J.E., et al.  . ( 2011). Tracking genome engineering outcome at individual DNA breakpoints. Nat. Methods  8, 671– 676. Google Scholar CrossRef Search ADS PubMed  Chou, D.M., Adamson, B., Dephoure, N.E., et al.  . ( 2010). A chromatin localization screen reveals poly (ADP ribose)-regulated recruitment of the repressive polycomb and NuRD complexes to sites of DNA damage. Proc. Natl Acad. Sci. USA  107, 18475– 18480. Google Scholar CrossRef Search ADS   D’Andrea, A.D., and Grompe, M. ( 2003). The Fanconi anaemia/BRCA pathway. Nat. Rev. Cancer  3, 23– 34. Google Scholar CrossRef Search ADS PubMed  Dronkert, M.L., and Kanaar, R. ( 2001). Repair of DNA interstrand cross-links. Mutat. Res.  486, 217– 247. Google Scholar CrossRef Search ADS PubMed  Elia, A.E., Boardman, A.P., Wang, D.C., et al.  . ( 2015). Quantitative proteomic atlas of ubiquitination and acetylation in the DNA damage response. Mol. Cell  59, 867– 881. Google Scholar CrossRef Search ADS PubMed  Escamilla-Del-Arenal, M., da Rocha, S.T., Spruijt, C.G., et al.  . ( 2013). Cdyl, a new partner of the inactive X chromosome and potential reader of H3K27me3 and H3K9me2. Mol. Cell. Biol.  33, 5005– 5020. Google Scholar CrossRef Search ADS PubMed  Fischle, W., Franz, H., Jacobs, S.A., et al.  . ( 2008). Specificity of the chromodomain Y chromosome family of chromodomains for lysine-methylated ARK(S/T) motifs. J. Biol. Chem.  283, 19626– 19635. Google Scholar CrossRef Search ADS PubMed  Franz, H., Mosch, K., Soeroes, S., et al.  . ( 2009). Multimerization and H3K9me3 binding are required for CDYL1b heterochromatin association. J. Biol. Chem.  284, 35049– 35059. Google Scholar CrossRef Search ADS PubMed  Gibson, B.A., and Kraus, W.L. ( 2012). New insights into the molecular and cellular functions of poly(ADP-ribose) and PARPs. Nat. Rev. Mol. Cell Biol.  13, 411– 424. Google Scholar CrossRef Search ADS PubMed  Guleria, A., and Chandna, S. ( 2016). ATM kinase: much more than a DNA damage responsive protein. DNA Repair  39, 1– 20. Google Scholar CrossRef Search ADS PubMed  Hustedt, N., and Durocher, D. ( 2016). The control of DNA repair by the cell cycle. Nat. Cell Biol.  19, 1– 9. Google Scholar CrossRef Search ADS PubMed  Jackson, S.P., and Bartek, J. ( 2009). The DNA-damage response in human biology and disease. Nature  461, 1071– 1078. Google Scholar CrossRef Search ADS PubMed  Jasin, M., and Rothstein, R. ( 2013). Repair of strand breaks by homologous recombination. Cold Spring Harb. Perspect. Biol.  5, a012740. Google Scholar CrossRef Search ADS PubMed  Kakarougkas, A., Ismail, A., Chambers, A.L., et al.  . ( 2014). Requirement for PBAF in transcriptional repression and repair at DNA breaks in actively transcribed regions of chromatin. Mol. Cell  55, 723– 732. Google Scholar CrossRef Search ADS PubMed  Kelley, M.R., Logsdon, D., and Fishel, M.L. ( 2014). Targeting DNA repair pathways for cancer treatment: what’s new? Future Oncol.  10, 1215– 1237. Google Scholar CrossRef Search ADS PubMed  Khoury-Haddad, H., Guttmann-Raviv, N., Ipenberg, I., et al.  . ( 2014). PARP1-dependent recruitment of KDM4D histone demethylase to DNA damage sites promotes double-strand break repair. Proc. Natl Acad. Sci. USA  111, E728– E737. Google Scholar CrossRef Search ADS   Khoury-Haddad, H., Nadar-Ponniah, P.T., Awwad, S., et al.  . ( 2015). The emerging role of lysine demethylases in DNA damage response: dissecting the recruitment mode of KDM4D/JMJD2D to DNA damage sites. Cell Cycle  14, 950– 958. Google Scholar CrossRef Search ADS PubMed  Kostyrko, K., and Mermod, N. ( 2016). Assays for DNA double-strand break repair by microhomology-based end-joining repair mechanisms. Nucleic Acids Res.  44, e56. Google Scholar CrossRef Search ADS PubMed  Kruhlak, M., Crouch, E.E., Orlov, M., et al.  . ( 2007). The ATM repair pathway inhibits RNA polymerase I transcription in response to chromosome breaks. Nature  447, 730– 734. Google Scholar CrossRef Search ADS PubMed  Larsen, D.H., Poinsignon, C., Gudjonsson, T., et al.  . ( 2010). The chromatin-remodeling factor CHD4 coordinates signaling and repair after DNA damage. J. Cell Biol.  190, 731– 740. Google Scholar CrossRef Search ADS PubMed  Lieber, M.R. ( 2010). The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu. Rev. Biochem.  79, 181– 211. Google Scholar CrossRef Search ADS PubMed  Liu, S., Yu, H., Liu, Y., et al.  . ( 2017a). Chromodomain protein CDYL acts as a Crotonyl-CoA hydratase to regulate histone crotonylation and spermatogenesis. Mol. Cell  67, 853– 866.e5. Google Scholar CrossRef Search ADS PubMed  Liu, Y., Liu, S., Yuan, S., et al.  . ( 2017b). Chromodomain protein CDYL is required for transmission/restoration of repressive histone marks. J. Mol. Cell Biol.  9, 178– 194. Google Scholar CrossRef Search ADS PubMed  Marteijn, J.A., Bekker-Jensen, S., Mailand, N., et al.  . ( 2009). Nucleotide excision repair-induced H2A ubiquitination is dependent on MDC1 and RNF8 and reveals a universal DNA damage response. J. Cell Biol.  186, 835– 847. Google Scholar CrossRef Search ADS PubMed  Matsuoka, S., Ballif, B.A., Smogorzewska, A., et al.  . ( 2007). ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science  316, 1160– 1166. Google Scholar CrossRef Search ADS PubMed  Miller, K.M., Tjeertes, J.V., Coates, J., et al.  . ( 2010). Human HDAC1 and HDAC2 function in the DNA-damage response to promote DNA nonhomologous end-joining. Nat. Struct. Mol. Biol.  17, 1144– 1151. Google Scholar CrossRef Search ADS PubMed  Mone, M.J., Volker, M., Nikaido, O., et al.  . ( 2001). Local UV-induced DNA damage in cell nuclei results in local transcription inhibition. EMBO Rep.  2, 1013– 1017. Google Scholar CrossRef Search ADS PubMed  Mulligan, P., Westbrook, T.F., Ottinger, M., et al.  . ( 2008). CDYL bridges REST and histone methyltransferases for gene repression and suppression of cellular transformation. Mol. Cell  32, 718– 726. Google Scholar CrossRef Search ADS PubMed  Pankotai, T., Bonhomme, C., Chen, D., et al.  . ( 2012). DNAPKcs-dependent arrest of RNA polymerase II transcription in the presence of DNA breaks. Nat. Struct. Mol. Biol.  19, 276– 282. Google Scholar CrossRef Search ADS PubMed  Polo, S.E. ( 2017). Switching genes to silent mode near DNA double-strand breaks. EMBO Rep.  18, 659– 660. Google Scholar CrossRef Search ADS PubMed  Polo, S.E., Kaidi, A., Baskcomb, L., et al.  . ( 2010). Regulation of DNA-damage responses and cell-cycle progression by the chromatin remodelling factor CHD4. EMBO J.  29, 3130– 3139. Google Scholar CrossRef Search ADS PubMed  Rafalska-Metcalf, I.U., Powers, S.L., Joo, L.M., et al.  . ( 2010). Single cell analysis of transcriptional activation dynamics. PLoS One  5, e10272. Google Scholar CrossRef Search ADS PubMed  Sakaue-Sawano, A., Kurokawa, H., Morimura, T., et al.  . ( 2008). Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell  132, 487– 498. Google Scholar CrossRef Search ADS PubMed  Schmidt, C.K., Galanty, Y., Sczaniecka-Clift, M., et al.  . ( 2015). Systematic E2 screening reveals a UBE2D-RNF138-CtIP axis promoting DNA repair. Nat. Cell Biol.  17, 1458– 1470. Google Scholar CrossRef Search ADS PubMed  Shanbhag, N.M., Rafalska-Metcalf, I.U., Balane-Bolivar, C., et al.  . ( 2010). ATM-dependent chromatin changes silence transcription in cis to DNA double-strand breaks. Cell  141, 970– 981. Google Scholar CrossRef Search ADS PubMed  Shen, Y., Nandi, P., Taylor, M.B., et al.  . ( 2011). RNA-driven genetic changes in bacteria and in human cells. Mutat. Res.  717, 91– 98. Google Scholar CrossRef Search ADS PubMed  Shiloh, Y., and Ziv, Y. ( 2013). The ATM protein kinase: regulating the cellular response to genotoxic stress, and more. Nat. Rev. Mol. Cell Biol.  14, 197– 210. Google Scholar CrossRef Search ADS   Sousa, F.G., Matuo, R., Soares, D.G., et al.  . ( 2012). PARPs and the DNA damage response. Carcinogenesis  33, 1433– 1440. Google Scholar CrossRef Search ADS PubMed  Svejstrup, J.Q. ( 2010). The interface between transcription and mechanisms maintaining genome integrity. Trends Biochem. Sci.  35, 333– 338. Google Scholar CrossRef Search ADS PubMed  Tallis, M., Morra, R., Barkauskaite, E., et al.  . ( 2014). Poly(ADP-ribosyl)ation in regulation of chromatin structure and the DNA damage response. Chromosoma  123, 79– 90. Google Scholar CrossRef Search ADS PubMed  Tubbs, A., and Nussenzweig, A. ( 2017). Endogenous DNA damage as a source of genomic instability in cancer. Cell  168, 644– 656. Google Scholar CrossRef Search ADS PubMed  Ui, A., Nagaura, Y., and Yasui, A. ( 2015). Transcriptional elongation factor ENL phosphorylated by ATM recruits polycomb and switches off transcription for DSB repair. Mol. Cell  58, 468– 482. Google Scholar CrossRef Search ADS PubMed  Vermeulen, M., Eberl, H.C., Matarese, F., et al.  . ( 2010). Quantitative interaction proteomics and genome-wide profiling of epigenetic histone marks and their readers. Cell  142, 967– 980. Google Scholar CrossRef Search ADS PubMed  Wickramasinghe, V.O., and Venkitaraman, A.R. ( 2016). RNA processing and genome stability: cause and consequence. Mol. Cell  61, 496– 505. Google Scholar CrossRef Search ADS PubMed  Wu, H., Min, J., Antoshenko, T., et al.  . ( 2009a). Crystal structures of human CDY proteins reveal a crotonase-like fold. Proteins  76, 1054– 1061. Google Scholar CrossRef Search ADS PubMed  Wu, J., Huen, M.S., Lu, L.Y., et al.  . ( 2009b). Histone ubiquitination associates with BRCA1-dependent DNA damage response. Mol. Cell. Biol.  29, 849– 860. Google Scholar CrossRef Search ADS PubMed  Xia, L., Huang, W., Bellani, M., et al.  . ( 2017). CHD4 has oncogenic functions in initiating and maintaining epigenetic suppression of multiple tumor suppressor genes. Cancer Cell  31, 653– 668 e657. Google Scholar CrossRef Search ADS PubMed  You, Z., Shi, L.Z., Zhu, Q., et al.  . ( 2009). CtIP links DNA double-strand break sensing to resection. Mol. Cell  36, 954– 969. Google Scholar CrossRef Search ADS PubMed  Zhang, Y., Yang, X., Gui, B., et al.  . ( 2011). Corepressor protein CDYL functions as a molecular bridge between polycomb repressor complex 2 and repressive chromatin mark trimethylated histone lysine 27. J. Biol. Chem.  286, 42414– 42425. Google Scholar CrossRef Search ADS PubMed  Zhu, Q., Wani, G., Arab, H.H., et al.  . ( 2009). Chromatin restoration following nucleotide excision repair involves the incorporation of ubiquitinated H2A at damaged genomic sites. DNA Repair  8, 262– 273. Google Scholar CrossRef Search ADS PubMed  © The Author(s) (2017). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Molecular Cell Biology Oxford University Press

CDYL1 fosters double-strand break-induced transcription silencing and promotes homology-directed repair

Loading next page...
 
/lp/ou_press/cdyl1-fosters-double-strand-break-induced-transcription-silencing-and-0DtVRr3vCT
Publisher
Oxford University Press
Copyright
© The Author(s) (2017). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved.
ISSN
1674-2788
eISSN
1759-4685
D.O.I.
10.1093/jmcb/mjx050
Publisher site
See Article on Publisher Site

Abstract

Abstract Cells have evolved DNA damage response (DDR) to repair DNA lesions and thus preserving genomic stability and impeding carcinogenesis. DNA damage induction is accompanied by transient transcription repression. Here, we describe a previously unrecognized role of chromodomain Y-like (CDYL1) protein in fortifying double-strand break (DSB)-induced transcription repression and repair. We showed that CDYL1 is rapidly recruited to damaged euchromatic regions in a poly (ADP-ribose) polymerase 1 (PARP1)-dependent, but ataxia telangiectasia mutated (ATM)-independent, manner. While the C-terminal region, containing the enoyl-CoA hydratase like (ECH) domain, of CDYL1 binds to poly (ADP-ribose) (PAR) moieties and mediates CDYL1 accumulation at DNA damage sites, the chromodomain and histone H3 trimethylated on lysine 9 (H3K9me3) mark are dispensable for its recruitment. Furthermore, CDYL1 promotes the recruitment of enhancer of zeste homolog 2 (EZH2), stimulates local increase of the repressive methyl mark H3K27me3, and promotes transcription silencing at DSB sites. In addition, following DNA damage induction, CDYL1 depletion causes persistent G2/M arrest and alters H2AX and replication protein A (RPA2) phosphorylation. Remarkably, the ‘traffic-light reporter’ system revealed that CDYL1 mainly promotes homology-directed repair (HDR) of DSBs in vivo. Consequently, CDYL1-knockout cells display synthetic lethality with the chemotherapeutic agent, cisplatin. Altogether, our findings identify CDYL1 as a new component of the DDR and suggest that the HDR-defective ‘BRCAness’ phenotype of CDYL1-deficient cells could be exploited for eradicating cancer cells harboring CDYL1 mutations. CDYL1, PARP1, double-strand breaks, homology-directed repair, non-homologous end joining, H3K27me3, EZH2 Introduction Our genome is constantly attacked by exogenous and endogenous mutagens that cause various types of DNA damage. To cope with the massive amount of DNA lesions, cells have evolved diverse mechanisms, collectively named DNA-damage response (DDR), to sense and repair DNA lesions (Tubbs and Nussenzweig, 2017). Sensing and repairing DNA lesions are mediated by rapid and highly orchestrated changes in chromatin structure and dynamics, followed by accumulation of DNA damage responsive proteins to DNA breakage sites (Jackson and Bartek, 2009; Adam and Polo, 2014). Several reports showed that DNA damage induction is accompanied by transient transcription pause to eliminate production of abnormal transcripts and to avoid deleterious collisions between transcription and repair machineries (Kruhlak et al., 2007; Chou et al., 2010; Shanbhag et al., 2010; Svejstrup, 2010; Pankotai et al., 2012; Adam and Polo, 2014; Kakarougkas et al., 2014; Ui et al., 2015; Wickramasinghe and Venkitaraman, 2016; Awwad et al., 2017; Polo, 2017). Double-strand breaks (DSBs) are considered the most cytotoxic form of DNA damage and vertebrates utilize two distinct DSB repair pathways. The first is homology-directed repair (HDR); an error-free process that functions only in late S and G2 phases of the cell cycle, where an intact chromatid is available to serve as a template for repair. In this pathway the ends of the DSB are recognized by the MRE11/RAD50/NBS1 (MRN) complex that catalyzes the formation of 3’ single stranded DNA (ssDNA) overhangs. The ssDNA is swiftly coated by the replication protein A (RPA), which is then displaced by Rad51 protein to initiate DNA strand invasion into the sister chromatid. This process requires the activity of several DNA repair proteins including BRCA1 and BRCA2 (Jasin and Rothstein, 2013; Hustedt and Durocher, 2016). The second repair pathway is non-homologous end joining (NHEJ); an error-prone process that functions throughout the cell cycle (Hustedt and Durocher, 2016). In NHEJ, Ku70-Ku80 heterodimer and 53BP1 mediator protein bind the DSB ends to inhibit end resection (Lieber, 2010). Subsequently, several factors including DNA-dependent protein kinase catalytic subunit, polynucleotide kinase and XRCC4-DNA ligase IV complex are recruited to promote end joining of the broken ends. Quantitative phospho-proteomic screens revealed that CDYL1 protein undergoes phosphorylation in response to DNA damage (Matsuoka et al., 2007; Bennetzen et al., 2010; Elia et al., 2015). These data prompted us to investigate the alleged role of CDYL1 in DDR. Human CDYL1 gene has three splicing variants: CDYL1a, CDYL1b, and CDYL1c. CDYL1b is the most abundant variant and consists of an N-terminal chromodomain (CD), a central hinge region and a C-terminal enoyl-CoA hydratase-like (ECH) domain. CDYL1 directly interacts with the catalytic subunit of polycomb repressive complex 2 (PRC2), enhancer of zeste homolog 2 (EZH2), and facilitates the establishment and propagation of H3K27me3 mark (Zhang et al., 2011). In addition, it was recently reported that CDYL1 plays an important role in the maintenance of repressive histone marks during DNA replication (Liu et al., 2017b). Moreover, it was shown that CDYL1 chromodomain bridges between the repressor element-1 silencing transcription factor (REST) and G9a methyltransferase (catalyzes the dimethylation of H3K9) to repress transcription (Mulligan et al., 2008). In line with this, CDYL1b contains a functional chromodomain that can bind H3K9me2/3 (Franz et al., 2009) and H3K27me3 (Vermeulen et al., 2010) and is involved in X-chromosome inactivation (Escamilla-Del-Arenal et al., 2013). The ECH domain of CDYL1b is essential for its multimerization and mediates its interaction with histone deactylases HDAC1 and HDAC2 to promote transcription repression (Caron et al., 2003). Furthermore, it was recently shown that the ECH domain of CDYL1 negatively regulates the transcriptionally active histone mark, lysine crotonylation (Wu et al., 2009a; Liu et al., 2017a). Notably, CDYL1 is a cancer-associated gene with a notable tumor suppressor activity. CDYL1 depletion increases the expression of several known oncogenes (e.g. TrkC) and promotes cellular transformation as evident by enhanced growth of human mammary epithelial cells (HMEC) in semi-solid media (Mulligan et al., 2008). Furthermore, the International Cancer Genome Consortium databases (http://icgc.org/) show that CDYL1 is mutated in 7.8% of diverse human cancers (n = 20343 cases). On the other hand, it was also shown that CDYL1 is required for repressing the CDH1 (E-cadherin) gene, whose loss is an essential event in epithelial–mesenchymal transition and is crucial for cancer metastasis, indicating that CDYL1 could facilitate tumorigenesis (Caron et al., 2003). Altogether, these data suggest that CDYL1 may possess oncogenic and tumor suppressor activities and can play distinct roles in specific biological processes when present in different protein complexes. Here, we present several lines of evidence implicating CDYL1 in DDR (hereafter, we refer to the endogenous protein as CDYL1 and the overexpressed form as CDYL1b). We showed that the ECH domain of CDYL1b and PARP1 activity promote CDYL1b recruitment to DNA damage sites. Remarkably, CDYL1 promotes EZH2 recruitment, fosters trimethylation of H3K27 at laser-microirradiated regions and underpins DSB-induced transcription silencing. Moreover, we demonstrated that CDYL1 exclusively facilitates HDR of DSBs. Lastly, we observed that CDYL1-depleted cells exhibit hypersensitivity to the widely used chemotherapeutic drug, cisplatin, and thus providing the basis for a new targeted therapy of CDYL1-deficient cancer cells. Results CDYL1 is rapidly recruited to DNA damage sites Given that CDYL1 is phosphorylated in response to DNA damage (Matsuoka et al., 2007; Bennetzen et al., 2010; Elia et al., 2015), we assumed that it might participate in DDR. To test this assumption, we monitored the subcellular localization of the most abundant variant, CDYL1b, following DNA damage induction. First, we tested the localization of EGFP-CDYL1b fusion at DNA breakage sites induced by laser microirradiation (Khoury-Haddad et al., 2014; Abu-Zhayia et al., 2017; Awwad et al., 2017) in four different cell types: Osteosarcoma human cells (U2OS); human breast adenocarcinoma cells (MCF7); non-tumorigenic breast epithelial cells (MCF10A) and mouse embryonic fibroblasts (MEF). Results showed rapid accumulation of EGFP-CDYL1b at laser-microirradiated regions in all tested cell lines (Figure 1A–D). Notably, the rapid recruitment kinetics of EGFP-CDYL1b is comparable to the mediator of DNA damage checkpoint 1 (MDC1) protein, which becomes detectable within 20 sec after damage induction, suggesting it is an early event in DDR (Supplementary Figure S1). Figure 1 View largeDownload slide CDYL1 is rapidly recruited to DNA damage sites. EGFP-CDYL1b shows swift accumulation at DNA breakage sites induced by laser microirradiation in U2OS (A), MCF7 (B), MCF10A (C), and MEF (D). Graphs in the right display fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites. Error bars indicate standard deviation (SD) from three independent experiments, and each measurement represents >10 different cells. White arrows mark the microirradiated regions. (E−G) U2OS (E), MCF7 (F), and MEF (G) cells were exposed to laser microirradiation and 5 min later were fixed and co-stained for γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). Results are typical of two independent experiments (n > 25). Scale bar, 2 μm. (H) Biochemical fractionation demonstrated that IR leads to CDYL1 accumulation at the chromatin-bound fraction. U2OS cells were exposed to 10 Gys of IR followed by biochemical fractionation at 5 and 30 min after IR. Protein samples were immunoblotted using the indicated antibodies. (I) Endogenous CDYL1 is enriched at I-SceI-induced DSBs. ChIP was performed 2 h after tamoxifen treatment using IgG, CDYL1, and γH2AX (as a control for damage induction) antibodies. Data are presented as fold enrichment around the DSB site relative to the negative control (IgG) and mean ± SD from three independent experiments. Statistical analysis was performed by Student’s t-test, **P < 0.01. Figure 1 View largeDownload slide CDYL1 is rapidly recruited to DNA damage sites. EGFP-CDYL1b shows swift accumulation at DNA breakage sites induced by laser microirradiation in U2OS (A), MCF7 (B), MCF10A (C), and MEF (D). Graphs in the right display fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites. Error bars indicate standard deviation (SD) from three independent experiments, and each measurement represents >10 different cells. White arrows mark the microirradiated regions. (E−G) U2OS (E), MCF7 (F), and MEF (G) cells were exposed to laser microirradiation and 5 min later were fixed and co-stained for γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). Results are typical of two independent experiments (n > 25). Scale bar, 2 μm. (H) Biochemical fractionation demonstrated that IR leads to CDYL1 accumulation at the chromatin-bound fraction. U2OS cells were exposed to 10 Gys of IR followed by biochemical fractionation at 5 and 30 min after IR. Protein samples were immunoblotted using the indicated antibodies. (I) Endogenous CDYL1 is enriched at I-SceI-induced DSBs. ChIP was performed 2 h after tamoxifen treatment using IgG, CDYL1, and γH2AX (as a control for damage induction) antibodies. Data are presented as fold enrichment around the DSB site relative to the negative control (IgG) and mean ± SD from three independent experiments. Statistical analysis was performed by Student’s t-test, **P < 0.01. To rule out a possibility that the observed accumulation is due to the fusion between CDYL1b and EGFP or CDYL1b overexpression, we tracked the subcellular localization of endogenous CDYL1 protein upon DNA damage induction. Toward this end, we devised three distinct but complementary approaches: First, U2OS, MCF7, and MEF cells were subjected to laser microirradiation and co-stained for γH2AX and CDYL1. The suitability of CDYL1 antibody for immunofluorescence analysis was confirmed (Supplementary Figure S2). Results showed that endogenous CDYL1 protein accumulates at DNA breakage sites marked by γH2AX (Figure 1E–G). Interestingly, once triggered, CDYL1 accumulation lasts for at least 1 h (Supplementary Figure S3). Second, biochemical fractionation of untreated and ionizing radiation (IR)-treated cells showed that the endogenous CDYL1 protein is enriched at damaged chromatin both at 5 and 30 min after IR (Figure 1H). Third, we monitored the localization of CDYL1 at DSB sites. Toward this end, we induced DSBs using I-SceI endonuclease in U2OS-TRE-I-SceI-19 cells (Awwad et al., 2017). Chromatin immunoprecipitation (ChIP) analysis showed significant enrichment of CDYL1 protein at DSB sites (Figure 1I). Notably, CDYL1 does not accumulate at UV-induced DNA damage sites marked by cyclobutane pyrimidine dimers (CPD) (Supplementary Figure S4). Altogether, these findings provide a firm evidence that CDYL1 is rapidly recruited to DSB sites in tumorigenic and non-tumorigenic cells. Also, CDYL1b recruitment is not cell type-specific and is conserved between human and mouse. CDYL1b exhibits prominent accumulation at damaged euchromatin We showed that CDYL1b foci colocalize with three heterchromatic markers: Hoechst-bright spots, H3K9me3, and HP1α (Figure 2A–C). These results are in line with previous findings showing CDYL1 enrichment at heterochromatic regions (Fischle et al., 2008; Franz et al., 2009). We took advantage of CDYL1b informative localization at heterochromatic regions to determine whether it is recruited to DNA damage sites induced within heterochromatic and euchromatic regions. Toward this end, MEF cells expressing EGFP-CDYL1b were subjected to laser microirradiation at heterochromatic and euchromatic regions and the intensity of EGFP signal was measured overtime. Results showed very slight increase in the fluorescence intensity of EGFP-CDYL1b signal at laser-microirradiated heterochromatin, when compared to damaged euchromatic regions that exhibit ~3.5-fold increase in the intensity of EGFP-CDYL1b (Figure 2D and E). To further substantiate this finding, we irradiated euchromatic and heterochromatic regions in the same cell. Again, results nicely showed prominent accumulation of EGFP-CDYL1b at euchromatic regions (Figure 2F). Altogether, our data strongly suggest that CDYL1b shows prominent accumulation at DNA breakage sites induced within euchromatic regions. Figure 2 View largeDownload slide CDYL1b accumulates prominently at damaged euchromatic regions. (A−C) CDYL1b is enriched at heterochromatic regions. MEF cells expressing EGFP-CDYL1b were stained with Hoechst (blue) (A), H3K9me3 (B), and HP1α (C) antibodies. Cells are representative of two independent experiments (n > 100). (D−F) Representative MEF cells showing EGFP-CDYL1b accumulation at laser-microirradiated sites induced in euchromatin (D) and heterochromatin (E). Graphs in the right show fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites marked by white arrows. (F) MEF cells showing EGFP-CDYL1b accumulation in both euchromatic and heterochromatic regions within the same cell. Graph shows absolute fluorescence intensity (%) of EGFP-CDYL1b at damaged heterochromatic and euchromatic sites. Cells are representative of three biological repeats (n > 10). Error bars indicate SD. Scale bar, 2 μm. Figure 2 View largeDownload slide CDYL1b accumulates prominently at damaged euchromatic regions. (A−C) CDYL1b is enriched at heterochromatic regions. MEF cells expressing EGFP-CDYL1b were stained with Hoechst (blue) (A), H3K9me3 (B), and HP1α (C) antibodies. Cells are representative of two independent experiments (n > 100). (D−F) Representative MEF cells showing EGFP-CDYL1b accumulation at laser-microirradiated sites induced in euchromatin (D) and heterochromatin (E). Graphs in the right show fold increase in the relative fluorescence intensity of EGFP-CDYL1b at damaged sites marked by white arrows. (F) MEF cells showing EGFP-CDYL1b accumulation in both euchromatic and heterochromatic regions within the same cell. Graph shows absolute fluorescence intensity (%) of EGFP-CDYL1b at damaged heterochromatic and euchromatic sites. Cells are representative of three biological repeats (n > 10). Error bars indicate SD. Scale bar, 2 μm. The carboxyl terminus of CDYL1b mediates its recruitment to DNA damage sites We sought to map the region of CDYL1b that regulates its accumulation at laser-microirradiated sites. Toward this end, deletion mutants of CDYL1b that lack either the chromodomain (CDYL1bdel(CD)), the hinge region (CDYL1bdel(Hinge)) or the C-terminal region including the ECH domain (CDYL1bdel(ECH)) were tagged with EGFP (Figure 3A) and tested for recruitment to laser-microirradiated sites. While the recruitment kinetic of EGFP-CDYL1bdel(Hinge) is comparable to the wild-type protein (Figure 3B and C), EGFP-CDYL1bdel(ECH) exhibits defective accumulation at DNA damage sites (Figure 3D). These observations suggest that the ECH domain facilitates CDYL1b recruitment to DNA damage sites. In contrast, EGFP-CDYL1bdel(CD) mutant shows faster and superior accumulation when compared to the wild-type protein (Figure 3E). We concluded therefore that CDYL1b chromodomain counteracts its recruitment to DNA damage sites. On this basis, we assumed that the binding of CDYL1b to H3K9me3 mark via its chromodomain is dispensable for its recruitment. To validate this assumption, we tested the recruitment of MonoRed-CDYL1b to laser-microirradiated sites in cells overexpressing EGFP-KDM4D protein. KDM4D overexpression causes a severe reduction in H3K9me2/3 (Khoury-Haddad et al., 2014) (Supplementary Figure S5). Results showed that both KDM4D and CDYL1b are recruited to DNA damage sites (Figure 3F). Altogether, we concluded that the compiling of CDYL1b at DNA breakage sites is independent of its binding to H3K9me2/3. Figure 3 View largeDownload slide The C-terminus of CDYL1b regulates its recruitment to DNA damage sites. (A) Western blot shows the expression of the indicated deletion mutants and the wild-type (WT) EGFP-CDYL1b fusion in U2OS cells. (B−E) Representative images of three independent experiments (n > 15) showing the recruitment of CDYL1b WT and deletion mutants to laser-microirradiated sites marked by white arrows. Graph in the right shows quantitative measurements of the fluorescence intensity of the indicated EGFP-CDYL1b fusions at the irradiated sites. Error bars indicate SD. (F) KDM4D overexpression shows no detectable effect on CDYL1b recruitment to laser-microirradiated sites. Laser microirradiation was applied on U2OS cells expressing EGFP-KDM4D (green) and MonoRed-CDYL1b (red). Graph displays the increase in the red and green fluorescence signals at DNA damage sites. Error bars indicate SD. Scale bar, 2 μm. Figure 3 View largeDownload slide The C-terminus of CDYL1b regulates its recruitment to DNA damage sites. (A) Western blot shows the expression of the indicated deletion mutants and the wild-type (WT) EGFP-CDYL1b fusion in U2OS cells. (B−E) Representative images of three independent experiments (n > 15) showing the recruitment of CDYL1b WT and deletion mutants to laser-microirradiated sites marked by white arrows. Graph in the right shows quantitative measurements of the fluorescence intensity of the indicated EGFP-CDYL1b fusions at the irradiated sites. Error bars indicate SD. (F) KDM4D overexpression shows no detectable effect on CDYL1b recruitment to laser-microirradiated sites. Laser microirradiation was applied on U2OS cells expressing EGFP-KDM4D (green) and MonoRed-CDYL1b (red). Graph displays the increase in the red and green fluorescence signals at DNA damage sites. Error bars indicate SD. Scale bar, 2 μm. PARP1, but not ATM, regulates CDYL1b recruitment to DNA damage sites Given the established role of ATM activity in promoting the recruitment of various DNA repair proteins to DNA damage sites (Shiloh and Ziv, 2013), we examined whether ATM regulates CDYL1b recruitment. Pharmacological inhibition of ATM has no discernable effect on CDYL1b recruitment to laser-microirradiated regions (Supplementary Figure S6A and B). To test the potency of ATM inhibition, we monitored CtIP recruitment to laser-microirradiated regions during S and G2 phase of the cell cycle. To do so, cells expressing MonoRed-CtIP and EGFP, which mark S/G2 and M phases, respectively, fused to the N-terminal domain of Geminin (Sakaue-Sawano et al., 2008) were subjected to laser microirradiation. Consistent with previous observations (You et al., 2009), ATM inhibition abrogates CtIP recruitment to DNA damage sites (Supplementary Figure S6C and D). Altogether, we concluded that CDYL1b accumulates at DNA damage sites irrespective of ATM activity. In agreement with this, the ATM-dependent phosphorylation of CDYL1 on Serine 160 is not required for its accumulation at damage sites. As shown in Supplementary Figure S7, the recruitment kinetics of EGFP-CDYL1bS160A phospho-mutant is comparable to the wild-type protein. We and others have shown that the recruitment of many DDR responsive proteins to DNA damage sites is dependent on PARP1 activity (Sousa et al., 2012; Tallis et al., 2014; Khoury-Haddad et al., 2015; Abu-Zhayia et al., 2017; Awwad et al., 2017). On this basis, we sought to determine whether PARP1 regulates EGFP-CDYL1b recruitment to damage sites. Here, we provide two lines of evidence demonstrating that PARP1 activity regulates EGFP-CDYL1b accumulation at laser-microirradiated sites. First, pharmacological inhibition of PARP1/2 abrogates EGFP-CDYL1b and the endogenous CDYL1 accumulation at laser-microirradiated sites (Figure 4A–D). Second, siRNA depletion of PARP1 (Figure 4E) leads to a severe reduction in the percentage of cells that show EGFP-CDYL1b recruitment to DNA damage sites (Figure 4F). Altogether, these results provide firm evidence that PARP1 activity is critical for CDYL1 recruitment to sites of damage. Figure 4 View largeDownload slide PARP1 activity regulates CDYL1 recruitment to DNA damage sites. (A−D) Pharmacological inhibition of PARP1/2 abolishes CDYL1 recruitment to DNA damage sites. Time-lapse images of U2OS cell expressing EGFP-CDYL1b that were treated with PARP inhibitor (A) and DMSO (B) prior to laser microirradiation. Graph (right) shows increase in the fluorescence intensity of EGFP-CDYL1b at damage sites. Data are representative of three biological repeats (n > 30). Error bars indicate SD. (C and D) U2OS cells were treated with PARP inhibitor (C) or DMSO (D), subjected to laser microirradiation and 5 min later immunostained with γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). This experiment was repeated twice (n > 20). (E) Western blot shows siRNA knockdown of PARP1. U2OS cells were transfected with PARP1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (F) PARP1 siRNA impairs EGFP-CDYL1b recruitment to DNA damage sites. Control and PARP1-deficient U2OS cells expressing EGFP-CDYL1b were subjected to laser microirradiation (indicated by red arrows). Scale bar, 2 μm. (G) Sequence alignment showed that CDYL1 protein contains a putative PAR-binding motif. (H) Purification of EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) from HEK293T cells using GFP-TRAP beads. Eluted proteins were separated and stained with Coomassie. Mr indicates protein marker. (I) PAR-binding assay with EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) recombinant proteins. EGFP-NLS and EGFP-Rpp21 were used as negative and a positive controls, respectively. IB, immunoblot. 32P, phosphate radiolabeled PAR. Figure 4 View largeDownload slide PARP1 activity regulates CDYL1 recruitment to DNA damage sites. (A−D) Pharmacological inhibition of PARP1/2 abolishes CDYL1 recruitment to DNA damage sites. Time-lapse images of U2OS cell expressing EGFP-CDYL1b that were treated with PARP inhibitor (A) and DMSO (B) prior to laser microirradiation. Graph (right) shows increase in the fluorescence intensity of EGFP-CDYL1b at damage sites. Data are representative of three biological repeats (n > 30). Error bars indicate SD. (C and D) U2OS cells were treated with PARP inhibitor (C) or DMSO (D), subjected to laser microirradiation and 5 min later immunostained with γH2AX (green) and CDYL1 (red). DNA is stained with DAPI (blue). This experiment was repeated twice (n > 20). (E) Western blot shows siRNA knockdown of PARP1. U2OS cells were transfected with PARP1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (F) PARP1 siRNA impairs EGFP-CDYL1b recruitment to DNA damage sites. Control and PARP1-deficient U2OS cells expressing EGFP-CDYL1b were subjected to laser microirradiation (indicated by red arrows). Scale bar, 2 μm. (G) Sequence alignment showed that CDYL1 protein contains a putative PAR-binding motif. (H) Purification of EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) from HEK293T cells using GFP-TRAP beads. Eluted proteins were separated and stained with Coomassie. Mr indicates protein marker. (I) PAR-binding assay with EGFP-CDYL1WT, EGFP-CDYL1bdel(ECH), and EGFP-CDYL1bdel(CD) recombinant proteins. EGFP-NLS and EGFP-Rpp21 were used as negative and a positive controls, respectively. IB, immunoblot. 32P, phosphate radiolabeled PAR. To shed further mechanistic insights into how PARP1 regulates CDYL1 recruitment, we checked CDYL1 poly (ADP-ribosylation) before and after ionizing radiation (IR). To do so, endogenous CDYL1 protein was immunoprecipitated from undamaged and IR-damaged cells and the immunoprecipitates were immunoblotted with poly (ADP-ribose) (PAR) and CDYL1 antibodies. Results demonstrated that CDYL1 is neither PARylated before, nor after IR induction (Supplementary Figure S8). Collectively, these observations suggest that CDYL1 recruitment to DNA damage sites is not mediated by poly (ADP-ribosylation). Interestingly, bioinformatic analysis revealed that CDYL1 contains a putative PAR-binding motif encompassing amino acids 535−542 at the very end of its C-terminal region (Figure 4G). This finding prompted us to test whether CDYL1 binds to PAR moieties. Toward this end, we performed radioactive in vitro PAR-binding assay using recombinant EGFP-CDYL1bWT, EGFP-CDYL1bdel(CD), and EGFP-CDYL1bdel(ECH) (Figure 4H). Results showed that CDYL1b directly binds to PAR moieties independent of the chromodomain, but dependent on the carboxyl terminus containing the ECH domain (Figure 4I). CDYL1 promotes transcription repression at double-strand break sites Given that CDYL1 is recruited to DNA damage sites (Figure 1) and is known to function as a transcriptional repressor (Caron et al., 2003; Mulligan et al., 2008; Escamilla-Del-Arenal et al., 2013), we predicted that it is implicated in DSB-induced transcription silencing. To test this prediction, we used U2OS-TRE-I-SceI-19 reporter cell line (Awwad et al., 2017) to measure the effect of CDYL1 depletion on the transcription of MS2 gene before and after DSB induction. This single-cell assay allows monitoring nascent transcription of the MS2 gene following DSB induction upstream its promoter region using I-SceI endonuclease (Ui et al., 2015). To activate transcription of MS2 gene, cells were transfected with pCherry-tTA-ER plasmid. This plasmid expresses a cytoplasmic Cherry-tTA-ER chimera, which migrates into the nucleus upon tamoxifen (Tam) addition, where it binds to the TRE repeats and induces transcription of MS2 gene. To visualize nascent transcription of MS2, cells were co-transfected with pYFP-MS2 plasmid, which expresses YFP-MS2 protein that binds to the MS2 stem loops. To study the effect of DSB on the transcription of MS2 gene, U2OS-TRE-I-SceI-19 cells were transfected with pCMV-NLS-I-SceI plasmid expressing I-SceI endonuclease that generates DSB upstream the MS2 gene. DSB induction can be validated by immunostaining with γH2AX antibody (Ui et al., 2015). To test whether CDYL1 is implicated in transcription repression after DSB induction, U2OS-TRE-I-SceI-19 cells were transfected with control and CDYL1 siRNAs (Figure 5A). Seventy-two hours post-transfection, nascent MS2 transcripts were visualized in the absence of DSBs (by co-transfecting the cells with pCherry-tTA-ER and pYFP-MS2 plasmids) and in the presence of DSBs (by co-transfecting the cells with pCherry-tTA-ER, pYFP-MS2, and pCMV-NLS-I-SceI plasmids). Consistent with previous findings (Rafalska-Metcalf et al., 2010; Ui et al., 2015; Awwad et al., 2017), MS2 gene is transcribed in cells expressing Cherry-tTA-ER fusion as evident by the co-localization of YFP and mCherry foci (Figure 5B, top panel), and DSB induction represses MS2 expression (Figure 5B, bottom panel). Remarkably, CDYL1 depletion using two different siRNA sequences leads to MS2 expression in the presence of DSBs in ~75% of the cells, whereas has no noticeable effect on MS2 expression in the absence of DSBs (Figure 5C and D). Quantification of CDYL1 effect on MS2 expression before and after DSB induction is shown in Figure 5E. Notably, CDYL1 depletion has no detectable effect on the intensity of γH2AX at the I-SceI-induced DSB site (Figure 5F). Collectively, these observations strongly suggest that CDYL1 underpins DSB-induced transcription silencing. Figure 5 View largeDownload slide CDYL1 promotes transcription repression at double-strand break sites. (A) Western blot shows the efficacy of CDYL1 depletion in U2OS-TRE-I-SceI-19 cells using two different siRNA sequences. β-actin is used as a loading control. (B−D) CDYL1 facilitates transcription silencing at DSB sites. (B) Representative images of control U2OS-TRE-I-SceI-19 cells that were transfected with plasmids expressing the transcription activator cherry-tTA-ER and YFP-MS2 fusion and then treated with tamoxifen to induce MS2 expression (top panel). To generate DSB, the cells were co-transfected with a third plasmid expressing I-SceI endonuclease (bottom panel). (C and D) Cells were treated as in B, except that the cells were transfected with two different siRNA sequences against CDYL1. Scale bar, 2 μm. (E) Graph displays the percentage of γH2AX-positive cells that show co-localization of YFP-MS2 and Cherry-tTA-ER. Error bars represent SD from four independent experiments (n > 60). P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001 (F) Quantitative measurements show that CDYL1 depletion has no detectable effect on the fluorescence intensity of γH2AX foci at DSB sites. Red horizontal bars indicate the mean fluorescence intensity of γH2AX in cells. Figure 5 View largeDownload slide CDYL1 promotes transcription repression at double-strand break sites. (A) Western blot shows the efficacy of CDYL1 depletion in U2OS-TRE-I-SceI-19 cells using two different siRNA sequences. β-actin is used as a loading control. (B−D) CDYL1 facilitates transcription silencing at DSB sites. (B) Representative images of control U2OS-TRE-I-SceI-19 cells that were transfected with plasmids expressing the transcription activator cherry-tTA-ER and YFP-MS2 fusion and then treated with tamoxifen to induce MS2 expression (top panel). To generate DSB, the cells were co-transfected with a third plasmid expressing I-SceI endonuclease (bottom panel). (C and D) Cells were treated as in B, except that the cells were transfected with two different siRNA sequences against CDYL1. Scale bar, 2 μm. (E) Graph displays the percentage of γH2AX-positive cells that show co-localization of YFP-MS2 and Cherry-tTA-ER. Error bars represent SD from four independent experiments (n > 60). P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001 (F) Quantitative measurements show that CDYL1 depletion has no detectable effect on the fluorescence intensity of γH2AX foci at DSB sites. Red horizontal bars indicate the mean fluorescence intensity of γH2AX in cells. CDYL1 promotes EZH2 accumulation and H3K27me3 induction at DNA breakage sites To gain molecular insights into the mechanism by which CDYL1 facilitates transcription repression at DSB sites, we monitored the localization of several repressive factors that were previously shown to be involved in DNA damage-induced transcription silencing in control and CDYL1-deficient cells. Since CDYL1 directly interacts with HDAC2 (Mulligan et al., 2008), which is recruited to DNA damage sites (Miller et al., 2010), we anticipated that CDYL1 might regulate HDAC2 recruitment to damaged chromatin. As shown in Supplementary Figure S9A and B, CDYL1 depletion has no detectable effect on HDAC2 accumulation at DNA damage sites. Afterward, we monitored the recruitment of CHD4, which is known to accumulate at DNA damage sites and promote the recruitment of repressive factors to damaged chromatin (Chou et al., 2010; Larsen et al., 2010; Polo et al., 2010; Xia et al., 2017). Similar to HDAC2, CHD4 is recruited to DNA damage sites independent of CDYL1 (Supplementary Figure S9C and D). Then, we looked at the repressive epigenetic mark H2A-K119 monoubiquitination, which is known to be induced locally at DNA damage sites to promote transcription silencing (Bergink et al., 2006; Marteijn et al., 2009; Zhu et al., 2009; Wu et al., 2009b; Ui et al., 2015). Results showed that monoubiquitination of H2A-K119 at DNA damage sites was also not affected by CDYL1 depletion (Supplementary Figure S9E and F). Since EZH2 is recruited to DNA breakage sites (Chou et al., 2010; Campbell et al., 2013) and interacts with CDYL1 (Zhang et al., 2011), we sought to determine whether CDYL1 regulates EZH2 recruitment to damage sites. To do so, we induced DSB using I-SceI in control and CDYL1-depleted U2OS-TRE-I-SceI-19 cells (Figure 6A), followed by ChIP analysis using EZH2 and γH2AX antibody as a positive control for DSB induction. Results showed that CDYL1 depletion impairs EZH2 accumulation at DSB sites (Figure 6B). Altogether, these observations demonstrated that CDYL1 underpins EZH2 recruitment to DSB sites. The fact that EZH2 catalyzes the methylation of H3K27me3, a repressive mark known to be induced at DNA damage sites (Chou et al., 2010), prompted us to test whether the level of H3K27me3 at damage sites is affected by CDYL1. Results showed that while CDYL1 depletion has no detectable effect on global H3K27me3 levels (Figure 6C), it diminishes the local induction of H3K27me3 at laser-microirradiated sites at 5, 60, and 120 min after damage (Figure 6D and E). Collectively, our data suggest that CDYL1 fortifies DSB-induced transcription silencing by promoting EZH2 compiling and H3K27 trimethylation at laser-microirradiated sites. Figure 6 View largeDownload slide CDYL1 promotes the accumulation of EZH2 and H3K27me3 at DNA damage sites. (A) Western blot shows siRNA knockdown of CDYL1. U2OS-TRE-I-SceI-19 cells were transfected with CDYL1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (B) CDYL1 knockdown abrogates EZH2 recruitment to DSB sites. ChIP was performed as described in Figure 1I using IgG, EZH2, and γH2AX antibodies. (C) Western blot shows that CDYL1b depletion has no detectable effect on the global levels of H3K27me3 mark. (D) CDYL1 depletion impairs the increase in H3K27me3 at laser-microirradiated sites. U2OS cells transfected with control or CDYL1 siRNA were subjected to laser microirradiation and co-immunostained with γH2AX (green) and H3K27me3 (red) antibodies at the indicated time points. DNA is stained with Hoechst (blue). Scale bar, 2 μm. (E) Graph displays H3K27me3 fluorescence intensity normalized to background in mock and CDYL1-depleted U2OS cells at the indicated time points after laser microirradiation. Error bars represent SD from three independent experiments (n > 60). P-values were calculated by two-sided Student’st-test relative to Ctrl siRNA. **P < 0.01. Figure 6 View largeDownload slide CDYL1 promotes the accumulation of EZH2 and H3K27me3 at DNA damage sites. (A) Western blot shows siRNA knockdown of CDYL1. U2OS-TRE-I-SceI-19 cells were transfected with CDYL1 siRNA and 72 h afterward, protein samples were immunoblotted with the indicated antibodies. (B) CDYL1 knockdown abrogates EZH2 recruitment to DSB sites. ChIP was performed as described in Figure 1I using IgG, EZH2, and γH2AX antibodies. (C) Western blot shows that CDYL1b depletion has no detectable effect on the global levels of H3K27me3 mark. (D) CDYL1 depletion impairs the increase in H3K27me3 at laser-microirradiated sites. U2OS cells transfected with control or CDYL1 siRNA were subjected to laser microirradiation and co-immunostained with γH2AX (green) and H3K27me3 (red) antibodies at the indicated time points. DNA is stained with Hoechst (blue). Scale bar, 2 μm. (E) Graph displays H3K27me3 fluorescence intensity normalized to background in mock and CDYL1-depleted U2OS cells at the indicated time points after laser microirradiation. Error bars represent SD from three independent experiments (n > 60). P-values were calculated by two-sided Student’st-test relative to Ctrl siRNA. **P < 0.01. CDYL1 depletion triggers DNA damage-induced G2/M arrest and alters the phosphorylation of DNA damage markers Mammalian cells respond to DNA-damaging agents by activating cell cycle checkpoints that lead to a temporary pause at a specific stage of the cell cycle to allow DNA damage repair (Guleria and Chandna, 2016). Flow cytometric analysis revealed that CDYL1-deficient cells (Figure 7A) accumulate in G2/M and exhibit persistent cell cycle arrest at 12 and 24 h after IR, likely due to defective DNA damage repair (Figure 7B). To clarify whether CDYL1 is implicated in DSB repair, we monitored γH2AX levels (earliest known marker of DSBs) in control and CDYL1-depleted cells. Results showed that CDYL1-deficient cells exhibit elevated levels of γH2AX at different time points after IR (Figure 7C). This finding indicates that CDYL1-depleted cells show elevated levels of damaged DNA, possibly due to inefficient repair of IR-induced DSBs. To further investigate the role of CDYL1 in DSB repair, we tested the levels of RPA2 phosphorylation at Ser4/Ser8, a marker for ssDNA generation following DNA-end resection that promotes HDR of DSB, after CPT treatment that causes DSBs by inhibiting topoisomerase I and promoting collision of replication forks. Results demonstrated defective phosphorylation of RPA2 in CDYL1-deficient cells when compared to control cells (Figure 7D), suggesting that CDYL1 may regulate DNA-end resection, a prerequisite step for HDR of DSBs. Figure 7 View largeDownload slide CDYL1 depletion causes G2/M arrest, alters the activation of DNA damage markers, and disrupts the integrity of homology-directed repair of DSBs. (A) Constitutive depletion of CDYL1 using specific shRNA sequence. U2OS cells were stably infected with lentiviral vectors expressing either scramble or CDYL1 shRNA and subjected to western blot analysis using the indicated antibodies. (B) CDYL1 knockdown leads to persistent G2/M arrest after IR. Control and CDYL1-deficient cells were exposed to IR (3 Gys) and samples were collected at the indicated time points for cell cycle analysis by flow cytometry. (C) Western blot analysis shows that CDYL1-deficient U2OS cells have elevated levels of γH2AX in response to IR. (D) Western blot shows that CDYL1 knockdown decreases CPT-induced levels of pRPA2 S4/S8 phosphorylation. Total RPA and β-actin were used as internal controls. The numbers below the blot indicate the relative intensity of pRPA bands, which was normalized to the intensity of total RPA band. (E) Western blot shows siRNA knockdown of CDYL1 in U2OS-TLR cells. (F) U2OS-TLR cells show that CDYL1 depletion impairs HDR of DSBs generated by I-SceI endonuclease. A reduction of ~70%−80% in GFP-positive cells was observed after CDYL1b depletion. Caffeine was used as a positive control. Given that HDR can occur only in S/G2 cell phase, data were corrected to flow-cytometry S/G2 values. (G) TLR results for DSBs repair by NHEJ. CDYL1 knockdown has no significant effect on the integrity of NHEJ. Data represent SD of three independent experiments. All the results are typical of three independent experiments. Error bars represent SD. P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001. Figure 7 View largeDownload slide CDYL1 depletion causes G2/M arrest, alters the activation of DNA damage markers, and disrupts the integrity of homology-directed repair of DSBs. (A) Constitutive depletion of CDYL1 using specific shRNA sequence. U2OS cells were stably infected with lentiviral vectors expressing either scramble or CDYL1 shRNA and subjected to western blot analysis using the indicated antibodies. (B) CDYL1 knockdown leads to persistent G2/M arrest after IR. Control and CDYL1-deficient cells were exposed to IR (3 Gys) and samples were collected at the indicated time points for cell cycle analysis by flow cytometry. (C) Western blot analysis shows that CDYL1-deficient U2OS cells have elevated levels of γH2AX in response to IR. (D) Western blot shows that CDYL1 knockdown decreases CPT-induced levels of pRPA2 S4/S8 phosphorylation. Total RPA and β-actin were used as internal controls. The numbers below the blot indicate the relative intensity of pRPA bands, which was normalized to the intensity of total RPA band. (E) Western blot shows siRNA knockdown of CDYL1 in U2OS-TLR cells. (F) U2OS-TLR cells show that CDYL1 depletion impairs HDR of DSBs generated by I-SceI endonuclease. A reduction of ~70%−80% in GFP-positive cells was observed after CDYL1b depletion. Caffeine was used as a positive control. Given that HDR can occur only in S/G2 cell phase, data were corrected to flow-cytometry S/G2 values. (G) TLR results for DSBs repair by NHEJ. CDYL1 knockdown has no significant effect on the integrity of NHEJ. Data represent SD of three independent experiments. All the results are typical of three independent experiments. Error bars represent SD. P-values were calculated by two-sided Student’s t-test relative to Ctrl siRNA. ***P < 0.001. CDYL1 exclusively promotes homology-directed repair of double-strand breaks To confirm a direct role of CDYL1 in DSB repair, we engineered a traffic light reporter (TLR) system (Certo et al., 2011) integrated into the genome of U2OS cells (Abu-Zhayia et al., 2017). The observed reporter cell line, hereafter called U2OS-TLR, has an I-SceI recognition site integrated within a GFP-reporter cassette and out-of-frame mCherry gene. While DSB repair by homologous recombination results in reconstitution of a functional GFP gene, manifested in green-colored cells, error-prone repair of DSBs by NHEJ introduces a frame-shift within the out-of-frame mCherry gene, leading to the expression of mCherry (Abu-Zhayia et al., 2017). Remarkably, CDYL1 depletion in U2OS-TLR cells using two different siRNAs (Figure 7E) resulted in 70%−80% decrease in the number of GFP-positive cells, when compared to cells treated with control siRNA (Figure 7F). This decrease demonstrated that CDYL1 is required for intact HDR of DSBs. Notably, the reduction of HDR in cells was accompanied by mild increase in mCherry-positive cells (Figure 7G), an indication that CDYL1 depletion had no deleterious effect on the integrity of NHEJ. Altogether, we concluded that CDYL1 exclusively promotes HDR but not NHEJ of DSBs. CDYL1 depletion sensitizes cells to the chemotherapeutic drug cisplatin Interstrand cross-linking agents, such as cisplatin, cause DNA adducts that can be repaired by various mechanisms including HDR (Dronkert and Kanaar, 2001). Therefore, HDR-deficient cells exhibit hypersensitivity to cisplatin (D’Andrea and Grompe, 2003). For example, it was shown that knocking down BRCA1 and BRCA2 sensitizes cells to cisplatin (Bartz et al., 2006). Since CDYL1 depletion disrupted the integrity of HDR (Figure 7F), we sought to examine the sensitivity of CDYL1-depleted cells to cisplatin. Toward this end, we first knocked out CDYL1 gene in non-malignant mammary epithelial cells, MCF10A, using CRISPR-Cas9 methodology. MCF10A cells were co-transfected with vectors expressing EGFP-Cas9 and specific guide-RNAs targeted against exon 4 of CDYL1. CDYL1 knockout was confirmed by DNA sequencing (Figure 8A), western blot (Figure 8B) and immunofluorescence analysis (Figure 8C). Next, control and CDYL1-knockout MCF10A cells were treated with increasing concentrations of cisplatin and subjected to short-term growth delay assay. Results showed that CDYL1-depleted cells are pronouncedly more sensitive to cisplatin when compared to control cells (Figure 8D). To validate that the hypersensitivity of CDYL1-depleted cells to cisplatin is caused by CDYL1 depletion, CDYL1-knockout MCF10A cells were infected with lentiviral vector expressing CDYL1 protein (Figure 8E) and tested for cisplatin sensitivity. As seen in Figure 8D, introducing CDYL1 suppresses the hypersensitivity to cisplatin, confirming that CDYL1 depletion indeed sensitizes cells to cisplatin. Altogether, we concluded that CDYL1-deficient cells exhibit synthetic lethality with cisplatin. Figure 8 View largeDownload slide CDYL1-knockout MCF10A cells display synthetic lethality with cisplatin. (A) DNA sequence alignment shows the deletion and the mismatches (in red) in exon 4 of CDYL1 gene at the 3′ of the gRNA sequence. (B and C) Control and CDYL1-knockout MCF10A cells were subjected to western blot (B) and immunofluorescence analysis (C). (D) CellTiter-Glo viability assay in wild-type, CDYL1-deficient cells, and CDYL1-deficient cells infected with lentiviral vector expressing CDYL1b. Cells were treated with increasing concentrations of cisplatin for 96 h. Data are representative of three independent experiments. Error bars indicate SD. Two-way ANOVA was used to test for differences at each dose. **P < 0.01. (E) Lentiviral vector expressing CDYL1b was used to restore the expression of CDYL1b in CDYL1-deficient MCF10A cells. Western blot analysis shows that the restored protein level of CDYL1 is comparable to the endogenous level in control MCF10A cells. Figure 8 View largeDownload slide CDYL1-knockout MCF10A cells display synthetic lethality with cisplatin. (A) DNA sequence alignment shows the deletion and the mismatches (in red) in exon 4 of CDYL1 gene at the 3′ of the gRNA sequence. (B and C) Control and CDYL1-knockout MCF10A cells were subjected to western blot (B) and immunofluorescence analysis (C). (D) CellTiter-Glo viability assay in wild-type, CDYL1-deficient cells, and CDYL1-deficient cells infected with lentiviral vector expressing CDYL1b. Cells were treated with increasing concentrations of cisplatin for 96 h. Data are representative of three independent experiments. Error bars indicate SD. Two-way ANOVA was used to test for differences at each dose. **P < 0.01. (E) Lentiviral vector expressing CDYL1b was used to restore the expression of CDYL1b in CDYL1-deficient MCF10A cells. Western blot analysis shows that the restored protein level of CDYL1 is comparable to the endogenous level in control MCF10A cells. Discussion Here, we describe a novel role of CDYL1 in DDR. We unprecedentedly demonstrated that CDYL1 is recruited to DSB sites in a PARP1-dependent manner, and that the C-terminal region, containing the ECH domain, binds to PAR moieties in vitro (Figure 4). We speculate therefore that CDYL1 accumulation at damage sites is mediated via binding to PAR moieties, which are known to be locally induced at DNA damage sites and form a docking platform for recruiting DNA repair proteins (Gibson and Kraus, 2012; Khoury-Haddad et al., 2014, 2015; Tallis et al., 2014; Abu-Zhayia et al., 2017; Awwad et al., 2017). CDYL1 undergoes phosphorylation in response to DNA damage on consensus site recognized by ATM and ATR (Matsuoka et al., 2007). However, we found that this phosphorylation is dispensable for its accumulation at DNA damage sites (Supplementary Figure S7). In line with this, CDYL1 is recruited irrespective of ATM activity (Supplementary Figure S6). On this basis, we anticipate that the DNA damage-induced phosphorylation may regulate CDYL1 activity at DNA breakage sites or modulate its interactome after damage. Previous report showed that CDYL1 foci colocalize with the heterchromatic region mark, H3K9me3 (Franz et al., 2009). We confirmed this finding and demonstrated that CDYL1 colocalizes also with bright DAPI foci and HP1α protein (Figure 2A−C). Altogether, these observations provide compelling evidence that CDYL1 foci represent heterochromatic regions. It should be noted, however, that laser microirradiation of heterochromatic regions may hit also neighboring euchromatic regions. Nevertheless, only negligible increase was observed in the fluorescence intensity of EGFP-CDYL1b at laser-microirradiated heterochromatic regions (Figure 2D and F). The lack of CDYL1b accumulation at damaged heterochromatin might be due to the fact that CDYL1b is enriched at heterchromatic regions regardless of DNA damage and therefore there is no need for further accumulation after damage induction. We and others recently showed that PARP1 activity promotes transcription repression at DNA damage sites by recruiting the repressive complexes NuRD and PcG (Chou et al., 2010) and the negative transcription elongation factor, NELF-E (Awwad et al., 2017; Polo, 2017). Here, we identified CDYL1 as an additional PARP1-regulated protein that contributes to DSB-induced transcription silencing at damage sites. Future studies will be required to clarify whether the recruitment of those repressive factors depends on each other and whether they have overlapping functions or work in different pathways or in different chromatin environment to achieve transcription silencing at DSB sites. Interestingly, since CDYL1 is not recruited to UV-induced damage sites (Supplementary Figure S4), we propose therefore that CDYL1 may not contribute to UV-induced transcription repression. Our data favor a model that CDYL1 enhances the recruitment of the endogenous H3K27 methyltransferase, EZH2, to DSB sites and subsequently leads to an increase in H3K27me3 methylation at DNA breakage sites. Similar findings showing CDYL1-dependent increase in EZH2 and H3K27me3 levels at I-SceI-induced DSBs were recently reported (Liu et al., 2017b). Notably, EZH2 accumulation at DSB sites was only observed by ChIP methodology. Several attempts to visualize EGFP-EZH2 enrichment at laser-microirradiated sites were proved unsuccessful (data not shown). Possible reasons for this variation may arise from the overexpression or the fusion of EZH2 with EGFP tag. Using a sophisticated reporter cell line, we showed that CDYL1 exclusively facilitates HDR of DSBs (Figure 7F). The fact that CDYL1 is recruited to DNA damage sites favors a scenario where CDYL1 directly regulates DSB repair. However, we cannot rule out a possibility that the defective HDR in CDYL1-depleted cells could also result from alterations in the expression of CDYL1 target genes (Zhang et al., 2011). Future studies will be required to shed mechanistic insight into how CDYL1 regulates HDR of DSBs. Interestingly, the reduction in HDR was not accompanied by an increase in NHEJ suggesting that the excess DSBs are not repaired by the classical NHEJ. We predict therefore that the resulted DSBs could be fixed by alternative repair pathways such as microhomology-mediated end-joining (MMEJ) and synthesis-dependent (SD)-MMEJ (Kostyrko and Mermod, 2016) or by RNA-templated DSB repair (Shen et al., 2011). Inhibiting DNA repair factors has become an attractive therapeutic concept in cancer therapy (Kelley et al., 2014). Our observations demonstrated that CDYL1 depletion increases the sensitivity of breast epithelial cells to cisplatin (Figure 8), which is one of the most commonly used chemotherapeutic drugs. On this basis, we propose that cisplatin treatment might be particularly effective to preferentially eradicate cancer cells harboring mutations in CDYL1 gene. Strikingly, CDYL1 mutations have been found in 13% of 1787 breast cancer cases and in 19.5% of 1287 hepatocellular cancer cases (http://icgc.org/). Future work will be required to measure the efficacy of cisplatin inhibition on the growth of CDYL1-deficient tumor cells in vivo. Material and methods Plasmids pEGFP-C1-CDYL1b, CDYL1bdel(CD), CDYL1bdel(Hinge), CDYL1bdel(ECH), CDYL1-S160A, MonoRed-CDYL1b, pLKO.1-TRC-CDYL1-shRNA, and MonoRed-CtIP were constructed as described in Supplementary Table S1. A complete list of all primers and their sequences is described in Supplementary Table S2. pSpCAS9 (BB)-2A-GFP (PX458; #48183) and PsPgRNA (#47108) vectors were purchased from Addgene. All constructs used in this study were verified by nucleotide sequencing or restriction digestion. mAG-hGeminin (1/110) plasmid expressing EGFP-Geminin was a kind gift from Dr Atsushi Miyawaki (RIKEN Brain Science Institute, Japan). pCSC-SP-PW-(GENE)-IRES/GFP lentiviral plasmid and packaging plasmids were a kind gift from Dr Hava Gil-Henn (Bar-Ilan University, Israel). Cell lines All cell lines used in this study were cultured in media supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 100 units/ml penicillin, and 100 μg/ml penicillin/streptomycin. U2OS, MEF, HEK293T, and U2OS-TRE-I-SceI-19 cell lines were grown in Dulbecco’s modified Eagle’s medium (Gibco). MCF-7 cell line was cultured in RPMI-1640 media (Gibco). MCF10A cell line was cultured in DMEM/F12 media (Invitrogen) supplemented with 5% horse serum (Invitrogen), 20 ml EGF, 0.5 mg/ml hydrocortisone, 100 ng/ml cholera toxin, 10 μg/ml insulin, 100 μg/ml penicillin/streptomycin, and 2 mM L-glutamine. Stable U2OS-TLR cells were grown in media containing 0.6 μg/ml puromycin. Transfections and drug treatments Cell transfections with DNA plasmids and siRNAs (Supplementary Table S3) were performed using PolyJet transfection reagent (BioConsult) and Lipofectamine2000 reagent (Invitrogen), respectively, following the manufacturer’s instructions. Where mentioned, cells were treated with 4 mM Caffeine (Sigma; C0750) for 72 h, 1 μM PARP inhibitor (Ku-0059436) for 1 h, 5 μM ATM inhibitor (KU-55933) for 2 h, 1 μM Camptothecin (CPT) (C9911; Sigma) for 2 h. Cisplatin (Selleck S1166) was added at the indicated concentrations. Where indicated, cells were exposed to ionizing radiation (IR) using the X-ray machine (Faxitron, CellRad). Western blot Western blot analysis was performed as previously described (Khoury-Haddad et al., 2014). Briefly, protein extracts were prepared using Hot-lysis buffer (1% SDS, 5 mM EDTA, 50 mM Tris, pH 7.5) and protease inhibitor mixture (Calbiochem). Samples were separated on SDS-PAGE gel and membranes were immunoblotted with the relevant antibodies. A complete list of antibodies and their dilutions is described in Supplementary Table S4. The immunoblots were developed using Quantum ECL detection kit (K-12042-D20, Advansta). The intensity of the immunoblot bands was determined using ImageJ software. Immunoprecipitation Control and CDYL1-depleted U2OS cells were left untreated or exposed to 10 Gys of IR and kept for 5 min to recover. Immunoprecipitation was performed as previously described (Escamilla-Del-Arenal et al., 2013) with the following modifications. Cells were fractionated using Buffer A (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl2, 10% glycerol, 0.34 M sucrose, 1 mM DTT, 0.1% Triton, PMSF, and protease inhibitor mixture) for 5 min on ice. Nuclear fraction was purified by centrifugation and incubated with lysis buffer (50 mM HEPEs, pH 7.4, 150 mM NaCl, 0.5% NP-40, 10 mM EDTA, 1 mM DTT) and Benzonase (Novagen) for 1 h on ice, followed by 30 min centrifugation. Supernatants were subjected to overnight immunoprecipitation (IP) using 1 μg of CDYL1 antibody and protein A magnetic beads (GenScript). The immune-complexes were washed, and subjected to western blot analysis. Immunofluorescence Cells were grown on coverslips for 24 h and subjected to immunofluorescence as previously described (Khoury-Haddad et al., 2014). Cells were immunostained with the appropriate antibodies (Supplementary Table S4). Slides were visualized using the inverted Zeiss LSM-700 confocal microscope with 40× oil EC Plan Neofluar objective. Laser microirradiation Cells were subjected to laser microirradiation as previously described (Khoury-Haddad et al., 2014). Briefly, cells were plated on flourodish (Ibidi; Cat#81158) and pre-sensitized with 1 μg/μl Hoechst 3334 dye for 10 min at 37°C. Laser microirradiation was executed using an LSM-700 inverted confocal microscope equipped with CO2 module and 37°C heating chamber. DNA damage was induced by micro-irradiating a single region in the nucleus with 15 iterations of 405 nm laser beam. Time-laps images were acquired using 488 nm laser. Signal intensity at damaged sites was measured using Zen 2009 software. In Figures 1E−G, 4C, D, 6D and Supplementary Figures S3 and S9, cells were plated on gridded plates (Ibidi; Cat#80826). Prior to 405 nm laser-microirradaition, cells were pre-sensitized with Hoechst for 5 min. After damage induction cells were pre-extracted with CSK buffer (10 mM HEPES-KOH, pH 7.9, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 1 mM EGTA, and 0.5% v/v Triton X-100) for 5 min on ice, fixed with 4% PFA for 10 min at room temperature and stained with the indicated antibodies. Generation of lentiviral particles and cell transduction In order to knockdown CDYL1 expression in U2OS and U2OS-TRE-I-SceI-19 cells, shRNA sequence was used as previously described (Awwad et al., 2017). First, scramble short hairpin oligonucleotides and short hairpin oligonucleotide directed against CDYL1 were annealed and inserted into pLKO.1-TRC lentiviral vector digested with EcoRI and AgeI. The generated lentiviral vectors were verified by nucleotide sequencing. Viral particles containing the shRNA construct were generated by transfecting HEK293T cells together with plasmids encoding the lentiviral proteins Gag, Pol, and VSV-G. Viral particles expressing CDYL1b protein were generated by co-transfecting HEK293T cells with pCSC-SP-PW-CDYL1b-IRES/GFP plasmid together with three plasmids expressing pRSV-REV, pMDL, and VSVG. Media containing the viral particles were collected 48 h post-transfection and filtered with 0.45 μm filters. Then, the viral particles were used to infect the indicated cell lines. At 72 h post-infection, cells were selected with 1 μg/ml puromycin for 1 week. Cell cycle analysis by flow cytometry Flow cytometric analysis was performed as previously described (Khoury-Haddad et al., 2014). Briefly, cells were fixed with ice-cold 75% ethanol. DNA was stained with 100 mg/ml propidium iodide (Sigma-Aldrich) in phosphate buffer solution (PBS) containing 0.1% Triton-X-100 and 0.5 mg/ml DNase free RNase A (Sigma-Aldrich). Samples were analyzed using flow cytometry of 10000 events on a BD LSR-II flow cytometer (Becton Dickinson). Data were analyzed with FCS express software. Traffic light reporter assay TLR assay was performed as previously described (Certo et al., 2011; Schmidt et al., 2015; Abu-Zhayia et al., 2017). In brief, U2OS-TLR cells were transfected with control or CDYL1 siRNAs. Ten hours later, cells were co-transfected with plasmids expressing I-SceI nuclease fused to infrared fluorescent protein (IFP) and donor plasmid expressing GFP donor sequence fused to blue fluorescent protein (BFP). Seventy-two hours after siRNA transfection, cells were harvested, and GFP and mCherry signals (reflecting HDR and NHEJ, respectively) were measured by four-color fluorescent flow-cytometry using a BD LSRFortessa™ cell analyzer (BD Biosciences). Minimum 10000 double-positive (IFP and BFP) cells were scored for each condition from three independent experiments. Results of siRNA-transfected cells were normalized to control siRNA-transfected cells. HDR values for each condition were normalized to the percentage of cells at S and G2 phase monitored by propidium iodide-based standard flow-cytometry. Chromatin fractionation Biochemical fractionation was performed as previously described (Khoury-Haddad et al., 2014) with the following modifications. U2OS cells were left untreated or exposed to 10 Gys of IR followed by 5 min and 30 min recovery. Cells were lysed with Buffer A for 5 min at 4°C. Cell lysates were centrifuged at 1500× g for 5 min at 4°C and the supernatant was removed. Then, the pellet was incubated with Buffer B (3 mM EDTA, 0.2 mM EGTA, 1 mM DTT, PMSF, and protease inhibitor mixture) for 10 min on ice followed by centrifugation at 1700× g for 5 min at 4°C. To prepare chromatin-bound fraction, pellet was resuspended with Hot-lysis buffer, boiled for 15 min, and sonicated with two 15-sec pulses of 35% amplitude. treated with Benzonase (Novagen) for 30 min at room temperature, centrifuged at maximum speed for 20 min at 12°C, and the supernatant was recovered. Chromatin-bound fractions were subjected to western blot analysis and immunoblotted with the indicated antibodies. Short-term growth delay assay Cells were seeded in 96-well plates at a density of 3000 per well (Thermo) and left for 1 h at room temperature to adhere before being returned to the incubator. At 24 h after seeding, cisplatin was added at the indicated concentrations. Cell viability was measured 96 h after drug treatment using the CellTiter-Glo Kit (Promega), following the manufacturer’s protocol. Visualizing MS2 expression before and after DSB in U2OS-TRE-I-SceI-19 The effect of DSB on the transcription of MS2 gene was monitored as previously described (Awwad et al., 2017). Briefly, U2OS-TRE-I-SceI-19 cells were transfected with pCherry-tTA-ER plasmid, which expresses a cytoplasmic Cherry-tTA-ER chimera, and treated with 1 μM tamoxifen to drive its migration into the nucleus and induce transcription of MS2 gene. To visualize nascent transcripts of MS2, cells were co-transfected with pYFP-MS2 plasmid, which expresses YFP-MS2 protein that binds to the MS2 stem loops. To generate DSB, U2OS-TRE-I-SceI-19 cells were co-transfected with pCMV-NLS-I-SceI. Generation of CDYL1-knockout MCF10A cell line using CRISPR/Cas9 methodology MCF10A cells were co-transfected with pSpCAS9 (BB)-2A-GFP vector expressing GFP-Cas9 and PsPgRNA-CDYL1gRNA vector containing a specific gRNA to introduce DSB within CDYL gene. At 24 h after transfection, GFP-positive cells were sorted using BD LSRFortessa™ cell analyzer (BD Biosciences) and plated in 96-well plates at a dilution of one cell per well. Clones were first screened by western blot analysis. CDYL1-knockout clones were further validated by sequencing and immunofluorescence. Chromatin immunoprecipitation ChIP experiment was carried out as previously described (Ui et al., 2015; Awwad et al., 2017). Briefly, U2OS/TRE/I-SceI-19 cells were plated in 150-mm dishes and transfected with pCherry-rTA-ER plasmid only or together with a plasmid expressing I-SceI endonuclease, pCMV-NLS-I-SceI. At 24 h following transfection, cells were treated with 1 μM tamoxifen for 2 h. Cells were then crosslinked with 1% PFA for 10 min at room temperature, and cross-linking was stopped with 0.125 M Glycine for 5 min. After cell lysis, DNA was sheared to the size of 300–500 bp using a Vibra cell sonicator (15 sec ON, 30 sec OFF, 35% duty, 20 cycles). Five percent of each supernatant was used as input control and processed with the cross-linking reversal step. The rest of the supernatant was subjected to overnight immunoprecipitation (IP) using either 1 μg of CDYL1, γH2AX antibody (Millipore 05-636) or EZH2 and protein A magnetic beads (GenScript). Following reverse cross-linking; the precipitated DNA was purified using the PureLinkTM PCR Micro Kit. Quantification of the immunoprecipitated DNA was carried out by Step-One-Plus real-time PCR using Fast SYBR Green Master mix (Applied Biosystems) and the primers around the transcription start sites (TSS), GACGTAAACGGCCACAAGTT and GAACTTCAGGGTCAGCTTGC (80 bp downstream of TSS). Fold induction of binding surrounding the break site was calculated using the CT cycles in which untreated and treated IP samples values were normalized to the no-antibody control (IgG). PAR-binding assay Recombinant proteins including EGFP-CDYL1WT, CDYL1bdel(CD), and CDYL1bdel(ECH) were overexpressed in HEK293T cells and immunoprecipitated using GFP-TRAP beads following the manufacturer’s guidelines (Chromotek). The immunoprecipitated proteins were tested for their ability to bind to PAR moieties using the PAR-binding assay, as previously specified (Khoury-Haddad et al., 2015). Briefly, 1–5 pmol of proteins were blotted onto a nitrocellulose membrane and blocked with TBST buffer supplemented with 5% milk. Radioactively labeled PAR moieties were produced by auto-modified PARP1 prepared by in vitro PARylation reaction. This reaction was performed at room temperature for 20 min in a reaction buffer (50 mM Tris-HCl, pH 8, 25 mM MgCl2, 50 mM NaCl) supplemented with radiolabelled NAD+ (Perkin Elmer), activated DNA, and PARP1 enzyme (Trevigen). PAR moieties were detached from PARP1 using proteinase K and blotted membrane was incubated for 2 h with the radiolabelled PAR diluted in TBST buffer. Membranes were then washed with TBST, subjected to autoradiography and western blotting using GFP antibody. Localized UV damage Induction of localized UV irradiation was done as described in Mone et al. (2001). Briefly, cells were grown on coverslips, washed once with 1× PBS, covered by isopore polycarbonate membrane filter (Millipore), and then cells were irradiated with 150 J/m2 UV light. The filter was then removed and cells were fixed or incubated for additional recovery time. Cells were co-stained for CDYL1 and cyclobutane pyrimidine dimers (CPD). Slides were visualized using the inverted Zeiss LSM 700 confocal microscope with 40× oil EC Plan Neofluar objective. Statistical analysis Statistical analyses were performed using the demo version of GRAPHPAD prism software version. Addendum While this manuscript was in preparation, similar observations implicating CDYL1 in DSB repair were reported (Liu et al., 2017b). Supplementary material Supplementary material is available at Journal of Molecular Cell Biology online. Acknowledgements We thank Noga Gutmann-Raviv, Rami Aqeilan, Arnon Henn, Oded Kleifeld, and Yossi Shiloh for critical discussion of the manuscript. We thank Jacob Hanna (Weizmann Institute), Sarah Selig (Technion), Amir Orian (Technion), Hava Henn (Bar-Ilan University), Yehuda Assaraf (Technion), and Aaron Ciechanover (Technion) for providing plasmids and antibodies. Funding This work was supported by grants from the Israel Science Foundation (ISF, Grant no. 2021242), the Israel Cancer Association (Grant no. 2019404), the Binational Science Foundation (Grant no. 2023065), the Israel Cancer Research Fund (ICRF, Grant no. 2021762), and Volkswagen Foundation (Grant no. 2020594). E.R.A.-Z. and S.W.A. are supported by the Council for Higher Education 19 fellowship for outstanding minority M.Sc. and Ph.D. students, respectively. N.A. is supported by the Neubauer Family Foundation. Conflict of interest: none declared. Author contributions: E.R.A.-Z. performed the experiments described in this study (except the one indicated below), wrote the experimental procedures, and helped in proofreading the manuscript. S.W.A. performed the experiments described in Figures 4H, 4I, 8D and Supplementary Figures S4, S6C, S6D, S9 and helped in proofreading the manuscript. B.M.B.-O. helped in performing the experiments described in Figure 8E, constructed several plasmids used in this study, and helped in proofreading the manuscript. H.K.-H. performed the experiment described in Figure 3A and Supplementary Figure S5 and helped in proofreading the manuscript. N.A. conceived the study, planed the experiments, and wrote the manuscript. References Abu-Zhayia, E.R., Khoury-Haddad, H., Guttmann-Raviv, N., et al.  . ( 2017). A role of human RNase P subunits, Rpp29 and Rpp21, in homology directed-repair of double-strand breaks. Sci. Rep.  7, 1002. Google Scholar CrossRef Search ADS PubMed  Adam, S., and Polo, S.E. ( 2014). Blurring the line between the DNA damage response and transcription: the importance of chromatin dynamics. Exp. Cell Res.  329, 148– 153. Google Scholar CrossRef Search ADS PubMed  Awwad, S.W., Abu-Zhayia, E.R., Guttmann-Raviv, N., et al.  . ( 2017). NELF-E is recruited to DNA double-strand break sites to promote transcriptional repression and repair. EMBO Rep.  18, 745– 764. Google Scholar CrossRef Search ADS PubMed  Bartz, S.R., Zhang, Z., Burchard, J., et al.  . ( 2006). Small interfering RNA screens reveal enhanced cisplatin cytotoxicity in tumor cells having both BRCA network and TP53 disruptions. Mol. Cell. Biol.  26, 9377– 9386. Google Scholar CrossRef Search ADS PubMed  Bennetzen, M.V., Larsen, D.H., Bunkenborg, J., et al.  . ( 2010). Site-specific phosphorylation dynamics of the nuclear proteome during the DNA damage response. Mol. Cell. Proteomics  9, 1314– 1323. Google Scholar CrossRef Search ADS PubMed  Bergink, S., Salomons, F.A., Hoogstraten, D., et al.  . ( 2006). DNA damage triggers nucleotide excision repair-dependent monoubiquitylation of histone H2A. Genes Dev.  20, 1343– 1352. Google Scholar CrossRef Search ADS PubMed  Campbell, S., Ismail, I.H., Young, L.C., et al.  . ( 2013). Polycomb repressive complex 2 contributes to DNA double-strand break repair. Cell Cycle  12, 2675– 2683. Google Scholar CrossRef Search ADS PubMed  Caron, C., Pivot-Pajot, C., van Grunsven, L.A., et al.  . ( 2003). Cdyl: a new transcriptional co-repressor. EMBO Rep.  4, 877– 882. Google Scholar CrossRef Search ADS PubMed  Certo, M.T., Ryu, B.Y., Annis, J.E., et al.  . ( 2011). Tracking genome engineering outcome at individual DNA breakpoints. Nat. Methods  8, 671– 676. Google Scholar CrossRef Search ADS PubMed  Chou, D.M., Adamson, B., Dephoure, N.E., et al.  . ( 2010). A chromatin localization screen reveals poly (ADP ribose)-regulated recruitment of the repressive polycomb and NuRD complexes to sites of DNA damage. Proc. Natl Acad. Sci. USA  107, 18475– 18480. Google Scholar CrossRef Search ADS   D’Andrea, A.D., and Grompe, M. ( 2003). The Fanconi anaemia/BRCA pathway. Nat. Rev. Cancer  3, 23– 34. Google Scholar CrossRef Search ADS PubMed  Dronkert, M.L., and Kanaar, R. ( 2001). Repair of DNA interstrand cross-links. Mutat. Res.  486, 217– 247. Google Scholar CrossRef Search ADS PubMed  Elia, A.E., Boardman, A.P., Wang, D.C., et al.  . ( 2015). Quantitative proteomic atlas of ubiquitination and acetylation in the DNA damage response. Mol. Cell  59, 867– 881. Google Scholar CrossRef Search ADS PubMed  Escamilla-Del-Arenal, M., da Rocha, S.T., Spruijt, C.G., et al.  . ( 2013). Cdyl, a new partner of the inactive X chromosome and potential reader of H3K27me3 and H3K9me2. Mol. Cell. Biol.  33, 5005– 5020. Google Scholar CrossRef Search ADS PubMed  Fischle, W., Franz, H., Jacobs, S.A., et al.  . ( 2008). Specificity of the chromodomain Y chromosome family of chromodomains for lysine-methylated ARK(S/T) motifs. J. Biol. Chem.  283, 19626– 19635. Google Scholar CrossRef Search ADS PubMed  Franz, H., Mosch, K., Soeroes, S., et al.  . ( 2009). Multimerization and H3K9me3 binding are required for CDYL1b heterochromatin association. J. Biol. Chem.  284, 35049– 35059. Google Scholar CrossRef Search ADS PubMed  Gibson, B.A., and Kraus, W.L. ( 2012). New insights into the molecular and cellular functions of poly(ADP-ribose) and PARPs. Nat. Rev. Mol. Cell Biol.  13, 411– 424. Google Scholar CrossRef Search ADS PubMed  Guleria, A., and Chandna, S. ( 2016). ATM kinase: much more than a DNA damage responsive protein. DNA Repair  39, 1– 20. Google Scholar CrossRef Search ADS PubMed  Hustedt, N., and Durocher, D. ( 2016). The control of DNA repair by the cell cycle. Nat. Cell Biol.  19, 1– 9. Google Scholar CrossRef Search ADS PubMed  Jackson, S.P., and Bartek, J. ( 2009). The DNA-damage response in human biology and disease. Nature  461, 1071– 1078. Google Scholar CrossRef Search ADS PubMed  Jasin, M., and Rothstein, R. ( 2013). Repair of strand breaks by homologous recombination. Cold Spring Harb. Perspect. Biol.  5, a012740. Google Scholar CrossRef Search ADS PubMed  Kakarougkas, A., Ismail, A., Chambers, A.L., et al.  . ( 2014). Requirement for PBAF in transcriptional repression and repair at DNA breaks in actively transcribed regions of chromatin. Mol. Cell  55, 723– 732. Google Scholar CrossRef Search ADS PubMed  Kelley, M.R., Logsdon, D., and Fishel, M.L. ( 2014). Targeting DNA repair pathways for cancer treatment: what’s new? Future Oncol.  10, 1215– 1237. Google Scholar CrossRef Search ADS PubMed  Khoury-Haddad, H., Guttmann-Raviv, N., Ipenberg, I., et al.  . ( 2014). PARP1-dependent recruitment of KDM4D histone demethylase to DNA damage sites promotes double-strand break repair. Proc. Natl Acad. Sci. USA  111, E728– E737. Google Scholar CrossRef Search ADS   Khoury-Haddad, H., Nadar-Ponniah, P.T., Awwad, S., et al.  . ( 2015). The emerging role of lysine demethylases in DNA damage response: dissecting the recruitment mode of KDM4D/JMJD2D to DNA damage sites. Cell Cycle  14, 950– 958. Google Scholar CrossRef Search ADS PubMed  Kostyrko, K., and Mermod, N. ( 2016). Assays for DNA double-strand break repair by microhomology-based end-joining repair mechanisms. Nucleic Acids Res.  44, e56. Google Scholar CrossRef Search ADS PubMed  Kruhlak, M., Crouch, E.E., Orlov, M., et al.  . ( 2007). The ATM repair pathway inhibits RNA polymerase I transcription in response to chromosome breaks. Nature  447, 730– 734. Google Scholar CrossRef Search ADS PubMed  Larsen, D.H., Poinsignon, C., Gudjonsson, T., et al.  . ( 2010). The chromatin-remodeling factor CHD4 coordinates signaling and repair after DNA damage. J. Cell Biol.  190, 731– 740. Google Scholar CrossRef Search ADS PubMed  Lieber, M.R. ( 2010). The mechanism of double-strand DNA break repair by the nonhomologous DNA end-joining pathway. Annu. Rev. Biochem.  79, 181– 211. Google Scholar CrossRef Search ADS PubMed  Liu, S., Yu, H., Liu, Y., et al.  . ( 2017a). Chromodomain protein CDYL acts as a Crotonyl-CoA hydratase to regulate histone crotonylation and spermatogenesis. Mol. Cell  67, 853– 866.e5. Google Scholar CrossRef Search ADS PubMed  Liu, Y., Liu, S., Yuan, S., et al.  . ( 2017b). Chromodomain protein CDYL is required for transmission/restoration of repressive histone marks. J. Mol. Cell Biol.  9, 178– 194. Google Scholar CrossRef Search ADS PubMed  Marteijn, J.A., Bekker-Jensen, S., Mailand, N., et al.  . ( 2009). Nucleotide excision repair-induced H2A ubiquitination is dependent on MDC1 and RNF8 and reveals a universal DNA damage response. J. Cell Biol.  186, 835– 847. Google Scholar CrossRef Search ADS PubMed  Matsuoka, S., Ballif, B.A., Smogorzewska, A., et al.  . ( 2007). ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science  316, 1160– 1166. Google Scholar CrossRef Search ADS PubMed  Miller, K.M., Tjeertes, J.V., Coates, J., et al.  . ( 2010). Human HDAC1 and HDAC2 function in the DNA-damage response to promote DNA nonhomologous end-joining. Nat. Struct. Mol. Biol.  17, 1144– 1151. Google Scholar CrossRef Search ADS PubMed  Mone, M.J., Volker, M., Nikaido, O., et al.  . ( 2001). Local UV-induced DNA damage in cell nuclei results in local transcription inhibition. EMBO Rep.  2, 1013– 1017. Google Scholar CrossRef Search ADS PubMed  Mulligan, P., Westbrook, T.F., Ottinger, M., et al.  . ( 2008). CDYL bridges REST and histone methyltransferases for gene repression and suppression of cellular transformation. Mol. Cell  32, 718– 726. Google Scholar CrossRef Search ADS PubMed  Pankotai, T., Bonhomme, C., Chen, D., et al.  . ( 2012). DNAPKcs-dependent arrest of RNA polymerase II transcription in the presence of DNA breaks. Nat. Struct. Mol. Biol.  19, 276– 282. Google Scholar CrossRef Search ADS PubMed  Polo, S.E. ( 2017). Switching genes to silent mode near DNA double-strand breaks. EMBO Rep.  18, 659– 660. Google Scholar CrossRef Search ADS PubMed  Polo, S.E., Kaidi, A., Baskcomb, L., et al.  . ( 2010). Regulation of DNA-damage responses and cell-cycle progression by the chromatin remodelling factor CHD4. EMBO J.  29, 3130– 3139. Google Scholar CrossRef Search ADS PubMed  Rafalska-Metcalf, I.U., Powers, S.L., Joo, L.M., et al.  . ( 2010). Single cell analysis of transcriptional activation dynamics. PLoS One  5, e10272. Google Scholar CrossRef Search ADS PubMed  Sakaue-Sawano, A., Kurokawa, H., Morimura, T., et al.  . ( 2008). Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell  132, 487– 498. Google Scholar CrossRef Search ADS PubMed  Schmidt, C.K., Galanty, Y., Sczaniecka-Clift, M., et al.  . ( 2015). Systematic E2 screening reveals a UBE2D-RNF138-CtIP axis promoting DNA repair. Nat. Cell Biol.  17, 1458– 1470. Google Scholar CrossRef Search ADS PubMed  Shanbhag, N.M., Rafalska-Metcalf, I.U., Balane-Bolivar, C., et al.  . ( 2010). ATM-dependent chromatin changes silence transcription in cis to DNA double-strand breaks. Cell  141, 970– 981. Google Scholar CrossRef Search ADS PubMed  Shen, Y., Nandi, P., Taylor, M.B., et al.  . ( 2011). RNA-driven genetic changes in bacteria and in human cells. Mutat. Res.  717, 91– 98. Google Scholar CrossRef Search ADS PubMed  Shiloh, Y., and Ziv, Y. ( 2013). The ATM protein kinase: regulating the cellular response to genotoxic stress, and more. Nat. Rev. Mol. Cell Biol.  14, 197– 210. Google Scholar CrossRef Search ADS   Sousa, F.G., Matuo, R., Soares, D.G., et al.  . ( 2012). PARPs and the DNA damage response. Carcinogenesis  33, 1433– 1440. Google Scholar CrossRef Search ADS PubMed  Svejstrup, J.Q. ( 2010). The interface between transcription and mechanisms maintaining genome integrity. Trends Biochem. Sci.  35, 333– 338. Google Scholar CrossRef Search ADS PubMed  Tallis, M., Morra, R., Barkauskaite, E., et al.  . ( 2014). Poly(ADP-ribosyl)ation in regulation of chromatin structure and the DNA damage response. Chromosoma  123, 79– 90. Google Scholar CrossRef Search ADS PubMed  Tubbs, A., and Nussenzweig, A. ( 2017). Endogenous DNA damage as a source of genomic instability in cancer. Cell  168, 644– 656. Google Scholar CrossRef Search ADS PubMed  Ui, A., Nagaura, Y., and Yasui, A. ( 2015). Transcriptional elongation factor ENL phosphorylated by ATM recruits polycomb and switches off transcription for DSB repair. Mol. Cell  58, 468– 482. Google Scholar CrossRef Search ADS PubMed  Vermeulen, M., Eberl, H.C., Matarese, F., et al.  . ( 2010). Quantitative interaction proteomics and genome-wide profiling of epigenetic histone marks and their readers. Cell  142, 967– 980. Google Scholar CrossRef Search ADS PubMed  Wickramasinghe, V.O., and Venkitaraman, A.R. ( 2016). RNA processing and genome stability: cause and consequence. Mol. Cell  61, 496– 505. Google Scholar CrossRef Search ADS PubMed  Wu, H., Min, J., Antoshenko, T., et al.  . ( 2009a). Crystal structures of human CDY proteins reveal a crotonase-like fold. Proteins  76, 1054– 1061. Google Scholar CrossRef Search ADS PubMed  Wu, J., Huen, M.S., Lu, L.Y., et al.  . ( 2009b). Histone ubiquitination associates with BRCA1-dependent DNA damage response. Mol. Cell. Biol.  29, 849– 860. Google Scholar CrossRef Search ADS PubMed  Xia, L., Huang, W., Bellani, M., et al.  . ( 2017). CHD4 has oncogenic functions in initiating and maintaining epigenetic suppression of multiple tumor suppressor genes. Cancer Cell  31, 653– 668 e657. Google Scholar CrossRef Search ADS PubMed  You, Z., Shi, L.Z., Zhu, Q., et al.  . ( 2009). CtIP links DNA double-strand break sensing to resection. Mol. Cell  36, 954– 969. Google Scholar CrossRef Search ADS PubMed  Zhang, Y., Yang, X., Gui, B., et al.  . ( 2011). Corepressor protein CDYL functions as a molecular bridge between polycomb repressor complex 2 and repressive chromatin mark trimethylated histone lysine 27. J. Biol. Chem.  286, 42414– 42425. Google Scholar CrossRef Search ADS PubMed  Zhu, Q., Wani, G., Arab, H.H., et al.  . ( 2009). Chromatin restoration following nucleotide excision repair involves the incorporation of ubiquitinated H2A at damaged genomic sites. DNA Repair  8, 262– 273. Google Scholar CrossRef Search ADS PubMed  © The Author(s) (2017). Published by Oxford University Press on behalf of Journal of Molecular Cell Biology, IBCB, SIBS, CAS. All rights reserved.

Journal

Journal of Molecular Cell BiologyOxford University Press

Published: Nov 21, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

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

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

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.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off