Portal de Periódicos da CAPES

TET3‐mediated DNA oxidation promotes ATR‐dependent DNA damage response

2017; Springer Nature; Volume: 18; Issue: 5 Linguagem: Inglês

10.15252/embr.201643179

1469-3178

Dewei Jiang, Shu Wei, Fei Chen, Ying Zhang, Jiali Li,

Genomics and Chromatin Dynamics

Article21 March 2017free access TET3-mediated DNA oxidation promotes ATR-dependent DNA damage response Dewei Jiang Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Shu Wei Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Kunming College of Life Science, University of Chinese Academy of Sciences, Kunming, China Search for more papers by this author Fei Chen Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Ying Zhang Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Jiali Li Corresponding Author [email protected] orcid.org/0000-0002-0039-3671 Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Kunming Primate Research Center of the Chinese Academy of Sciences, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Dewei Jiang Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Shu Wei Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Kunming College of Life Science, University of Chinese Academy of Sciences, Kunming, China Search for more papers by this author Fei Chen Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Ying Zhang Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Jiali Li Corresponding Author [email protected] orcid.org/0000-0002-0039-3671 Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Kunming Primate Research Center of the Chinese Academy of Sciences, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China Search for more papers by this author Author Information Dewei Jiang1, Shu Wei1,2, Fei Chen1, Ying Zhang1 and Jiali Li *,1,3 1Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China 2Kunming College of Life Science, University of Chinese Academy of Sciences, Kunming, China 3Kunming Primate Research Center of the Chinese Academy of Sciences, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming, China *Corresponding author. Tel:/Fax: +86 871 65197043; E-mail: [email protected] EMBO Rep (2017)18:781-796https://doi.org/10.15252/embr.201643179 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract An efficient, accurate, and timely DNA damage response (DDR) is crucial for the maintenance of genome integrity. Here, we report that ten-eleven translocation dioxygenase (TET) 3-mediated conversion of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) in response to ATR-dependent DDR regulates DNA repair. ATR-dependent DDR leads to dynamic changes in 5hmC levels and TET3 enzymatic activity. We show that TET3 is an ATR kinase target that oxidizes DNA during ATR-dependent DNA damage repair. Modulation of TET3 expression and activity affects DNA damage signaling and DNA repair and consequently cell death. Our results provide novel insight into ATR-mediated DDR, in which TET3-mediated DNA demethylation is crucial for efficient DNA repair and maintenance of genome stability. Synopsis An accurate and efficient DNA damage response is crucial for the maintenance of genome integrity. This study shows that TET3-mediated DNA demethylation responds to DNA damage and promotes ATR-dependent DNA damage signaling and repair. DNA damage response (DDR) leads to changes in 5hmC levels. DDR-induced alterations in TET3 activity and 5hmC levels are ATR-dependent. TET3 is an ATR target and facilitates ATR-dependent DNA damage signaling and repair. Manipulation of TET3 activity affects DDR-induced cell death and clonogenic survival. Introduction DNA damage holds constant threats to faithfully transmit the correct genetic information to the next generation as well as to accurately transcribe the genome for cellular function and survival. Thus, efficient, accurate, and timely DDR and repair are essential to fulfill all these goals 1. Cellular genome is constantly challenged by endogenous stress such as reactive oxygen species, or those originated exogenously such as mutagens, and ultraviolet and ionic irradiations. While the impact of genomic instability is obvious in dividing cells, in non-dividing cells, such damage when left unrepaired would also lead to transcriptional abnormalities and hence result in devastating phenotypic consequences 2. To combat these threats, cells have evolved an array of DNA repair mechanisms to detect DNA damage, signal its presence, and mediate its repair 3. For terminally differentiated, long life mature cells like neurons, they further highlight the need for efficient cellular DDR program to safeguard genetic fidelity 456. However, being post-mitotic also means any DNA repair mechanisms like homologous recombination that takes part only during S-phase rendered not available. Despite that, neurons still use many other DNA repair pathways to maintain genome stability 7. DNA damage is typically classified based on the number of broken strands on the DNA double helix 8. Double-strand breaks (DSBs) represent the most biologically significant lesion. They may arise naturally in physiological processes such as V(D)J recombination, meiosis, and DNA replication but can also occur upon exogenous insults such as ionizing radiation (IR) and topoisomerase inhibitors (e.g. etoposide). Having the most significant consequence to cellular function, DSBs do not only affect cell cycle regulation but it can also result in cell death 9. In dividing cells, DSBs can repaired by two pathways—the high-fidelity homologous recombination (HR) and the less efficient non-homologous end-joining (NHEJ), whereas in non-dividing cells, only NHEJ is used 1011. On the other hand, while single-strand breaks (SSBs) are equally disastrous, the presence of intact complementary strand makes the correct restoration of the broken strand less challenging. Both dividing and non-dividing cells utilize mechanisms to repair SSBs, and these include base excision repair (BER), nucleotide excision repair (NER), and mismatch repair systems 9. Among various known sensors and regulators take part in the complex DNA damage response an repair, ataxia telangiectasia mutated (ATM) and ataxia telangiectasia and Rad3-related protein (ATR) are the most well-studied 121314151617 ones which both belong to the phosphatidylinositol 3-kinase-related kinase (PI2KK) family. Despite being structurally similar, ATM and ATR initiate different cellular response DNA damages 1418192021. By far, it has been suggested that ATM or ATR functions in complementarity faction, but recent work has shown evidence on the presence of their crosstalk during DDR and one study even suggested activation of ATR is regulated by ATM 22. The dynamic changes in DNA methylation and demethylation are essential epigenetic modifications for the maintenance genomic integrity 2324. DNA methylation is predominantly associated with transcriptional repression and is abundantly found in heterochromatin 25. In situ DNA methylation on the cytosine base to become 5-methylcytosine (5mC) is catalyzed by DNA methyltransferases (DNMT), which it is also responsible for the maintenance and de novo DNA methylation 2627. While the molecular aspects of DNA methylation have been well studied, the realization of active DNA demethylation and its epigenetic regulation is recently emphasized 2829. DNA demethylation is initiated by the oxidation of 5mC into 5-hydroxymethylcytosine (5hmC) by the TET family of proteins including TET1, TET2, and TET3 30313233. TET1 and TET2 are highly expressed in mouse embryonic stem cells, deletion of either one does not lead to impaired embryonic and postnatal development 343536. Homozygous knockout of Tet3 in mice causes death at birth 37, suggesting that TET3 is also dispensable for embryonic development. All the three TET proteins have the catalytic dioxygenase activity, yet their differential expression pattern and cell-type specificity expression indicate the existence of their specific roles in different biological processes 3839. The functional importance of 5mC and 5hmC epigenetic markers and their regulations have been identified in many biological processes 264041424344. Recent studies revealed that aberrant DNA methylation and demethylation is involved in the pathogenesis of multiple diseases including neurodegenerative disorders 45. Previous studies demonstrated a potential role of DNA methylation/demethylation in mediating DDR 464748495051. Tet-deficient cells show accumulation of damaged DNA and impaired DDR 52. Yet, the exact role and mechanism on how TET-mediated active DNA demethylation takes part in response to DNA damage and cell cycle regulation remains elusive. In this study, we demonstrate that dioxygenase TET3-mediated active DNA demethylation responds to DDR. We identify that DDR-induced TET3-mediated 5hmC production is predominantly ATR-dependent. Modulation of TET3 affects DNA damage signaling and DNA repair and consequently cell death and clonogenic survival. This supports a model in which TET3-mediated DNA oxidation in ATR-dependent DDR may function as a novel regulatory system of DNA damage signaling and genomic stability. Results DDR leads to changes in the levels of 5hmC Our recent work reported that TET1-mediated 5hmC responds to ATM-dependent DDR 45. To explore whether TET dioxygenases-mediated active DNA demethylation is also involved in ATR-dependent DNA damage response (DDR), HT22 cells were exposed to ultraviolet light (UV, 254 nm, 20 J/m2) or treated with topoisomerase I inhibitor, camptothecin (CPT) to induce DNA damage. After 3 h of treatment, cells were fixed and stained with 5hmC and γ-H2AX antibodies. We found that both UV- and CPT-induced DDR led to significantly increased intensities of 5hmC (Fig 1A and B). Furthermore, we analyzed 5hmC-specific immunostaining data and revealed a significant increase in 5hmC intensity in DDR, with a 2.25 ± 0.13-fold increase upon UV exposure (P < 0.01), and a 2.36 ± 0.12-fold increase after CPT treatment (P < 0.05). Indeed, increased 5hmC showed partially co-localized with γ-H2AX nuclear foci (Fig 1A). Similar phenomenon on DDR-induced changes in intensities of 5hmC was also observed in other cell types such as MEFs, N2a, and HEK293T cells (Fig EV1A–C). Figure 1. DDR induces changes in 5hmC production A. HT22 cells were stained with 5hmC (green), γ-H2AX (red), and DAPI (blue) at 3 h post-UV exposure (254 nm, 20 J/m2) or after camptothecin (CPT, 10 μM) treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White arrows indicate representative cells. B. Relative intensities of 5hmC immunostaining illustrated in panel (A) were quantified by use of ImageJ software. Data are presented as mean ± SEM (n = 36–88 cells per treatment). Each bar represents the average of three independent experiments (*P < 0.05, unpaired t-test). C, D. 5hmC-specific dot-blot intensities of genomic DNA isolated from HT22 cells at post-UV exposure or after CPT treatment at indicated time. Total genomic DNA was stained using methylene blue as the quality of loading control in the dot blot assays (lower panel). E, F. Relative intensities of 5hmC-specific dot signals illustrated in panels (C and D) were quantified by use of ImageJ software. The 5hmC contents were quantified and calculated in comparison with total genomic DNA sample (300 ng/each). Data are presented as mean ± SEM (n = 3–5 repetitions of the experiment). Download figure Download PowerPoint Click here to expand this figure. Figure EV1. DDR leads to changes in conversion of 5mC to 5hmC A. Wild-type MEFs were stained with 5hmC (green) and γ-H2AX (red) at 3 h post-UV exposure (254 nm, 20 J/m2) or camptothecin (CPT, 10 μM) treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White squares indicate representative cells. B. N2a cells were stained with 5hmC (green), γ-H2AX (red) at 3 h post-UV or CPT treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White circles indicate representative cells. C. HEK293T cells were stained with stained with 5hmC (green) and γ-H2AX (red) at 3 h post-UV or CPT treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White circles indicate representative cells. D, E. Changes in global 5hmC contents at post-UV exposure or after CPT treatment at indicated time were examined with Quest 5hmC™ DNA ELISA Kit. To quantitate the 5hmC percentage in each sample, a standard curve was generated using 100 ng control samples with 0, 0.03, 0.12, 0.23, and 0.55% 5hmC, respectively. Data are presented as mean ± SEM. Each bar represents the average of three independent experiments. Download figure Download PowerPoint To validate DDR-induced increase in 5hmC levels, we determined the total global 5hmC level at pre- and post-DDR using a dot-blot assay in HT22 cells. As expected, we observed that the amount of 5hmC was significantly increased after UV exposure or treatment with CPT (Fig 1C–F), with a 2.10 ± 0.16-fold increase of the percentage of 5hmC content in genomic DNA at 3 h post-UV exposure, and a 1.5 ± 0.14-fold increase after CPT treatment. Next, to investigate whether changes in 5hmC levels in response to DDR is a dynamic one during DNA damage and repair, we examined the levels of 5hmC in the DNA damage response at different time points. Interestingly, while dynamic changes in 5hmC levels showed differences between UV exposure and CPT treatment, higher levels of 5hmC occurred at early DDR period gradually decreased toward later stages (Fig 1C–F). Alternatively, DDR-induced changes in total level of global 5hmC in N2a cells were examined with Quest 5hmC™ DNA ELISA Kit (Zymo Research). Quantitative analysis revealed a significant increase in the total amount of global 5hmC at 3 h post-UV exposure or CPT treatment, with a 1.65 ± 0.45-fold increase upon UV exposure (P < 0.05) (Fig EV1D), and a 1.51 ± 0.38-fold increase after CPT treatment (P < 0.01) (Fig EV1E). Similarly, distinctly dynamic changes in the total amount of global 5hmC were showed at 6, 12 and 24 h post-UV exposure and CPT treatment (Fig EV1D and E). To investigate whether DDR-induced 5hmC production resulted from conversion of 5mC, we examined changes in intensities of 5mC in several cell types including HT22, N2a, and HEK293. Notably, DDR resulted in significant reduction in 5mC intensities (Fig EV2A–C). Similar reductions in 5mC levels were found in cells after UV exposure or treatment with CPT (Fig EV2D and E). Click here to expand this figure. Figure EV2. DDR leads to a decrease in 5mC levels HT22 cells were stained with 5mC (green), γ-H2AX (red) at 3 h post-UV or CPT treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White arrows indicate representative cells. N2a cells were stained with 5mC (green), γ-H2AX (red) at 3 h post-UV or CPT treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White circles indicate representative cells. HEK293T cells were stained with stained with 5mC (green) and γ-H2AX (red) at 3 h post-UV or CPT treatment for 3 h. The total number of cells was counted by DAPI-stained nuclei. White circles indicate representative cells. 5mC-specific dot-blot intensities of genomic DNA isolated from HT22 cells at post-UV exposure or after CPT treatment at indicated time. Total genomic DNA was stained using methylene blue as the quality of loading control in the dot blot assays (lower panel). Relative intensities of 5mC-specific dot signals illustrated in panel (D) were quantified by use of ImageJ software. The 5mC contents were quantified and calculated in comparison with total genomic DNA sample (250 ng/each) (*P < 0.05, unpaired t-test; mean ± SEM; n = 3 repetitions of the experiment). Download figure Download PowerPoint TET dioxygenases differentially respond to ATR-dependent DDR While the three TET dioxygenases mediate active DNA demethylation in similar fashion, their differential tempo-spatial expression patterns in different cell types suggest they have potentially distinct roles in DDR. Therefore, to investigate which TET dioxygenase is at preference to respond to UV- and CPT-induced DDR, and predominantly responsible for changes observed in 5hmC levels, we performed studies in N2a cells. Strikingly, both UV and CPT led to a significant increase in the protein levels of TET3 in contrast to small fluctuations observed in TET1 and TET2 (Fig EV3A and B). Phospho-ATR (S428-ATR), cleaved PARP-1, and γ-H2AX were used at the same time to access DNA damage and repair. To assess whether DDR-induced changes in protein levels of TET3 is a transcriptional response, total RNA was extracted from UV-exposed or CPT-treated N2a cells and subjected to quantitative real-time PCR (qRT–PCR) analysis. Consistent with our previous findings 45, no significant changes in mRNAs levels of three family Tet genes in response to DNA damage were observed, indicating the change in TET protein level is not a gene transcription event (Fig EV3C). Click here to expand this figure. Figure EV3. DDR leads to differential changes in protein levels of the TET family dioxygenases Protein extracts from N2a cells at post-UV exposure or after CPT treatment at indicated time were immunoblotted with antibodies against TET1, TET2, TET3, S428-ATR, ATR, PARP-1, γ-H2AX. β-Actin was used as a loading control. Relative intensities of immunoblot signals illustrated in panel (A) were quantified by use of ImageJ software. Error bars denote standard deviations. *P < 0.05 (by unpaired t-test, n = 3 repetitions of the experiment). The mRNA levels of Tet1, Tet2, and Tet3 showed little change in N2a cells with UV exposure or CPT treatment. Total RNA was extracted from N2a cells at post-UV exposure or after CPT treatment at indicated time, and quantitative RT–PCR was performed. Data are presented as mean ± SEM. Each bar represents the average of three independent experiments. Protein extracts from human control and Seckel syndrome (ATR-deficient, AtrΔ/Δ) fibroblasts after CPT treatment at indicated times were immunoblotted with antibodies against TET1, TET2, TET3, S428-ATR, ATR, PARP-1, γ-H2AX. β-Actin was used as a loading control. Relative intensities of immunoblot signals illustrated in panel (D) were quantified by use of ImageJ software. n = 3 repetitions of the experiment. Source data are available online for this figure. Download figure Download PowerPoint Although both ATM and ATR mediate DNA damage response, it has been well documented that ATR plays a more dominant role in mediating UV- and CPT-induced DDR 2021. To further confirm our findings, we examined whether CPT-induced changes in the protein levels of TET enzymes is ATR-dependent. Human Seckel syndrome (ATR-deficient, AtrΔ/Δ) fibroblasts and those obtained from non-diseased control were treated with CPT for various time points. In contrast to the mild increase observed for TET3 protein levels in ATR-deficient cells, CPT treatment led to a robust increase of such in non-diseased control cells. Again, both TET1 and TET2 showed little responses in response to CPT challenge in both non-diseased control and ATR-deficient fibroblasts (Fig EV3D and E), implying that TET3 is mainly responsible for ATR-dependent DDR-induced changes in 5hmC levels. ATR but not ATM is required for UV- and CPT-induced changes in 5hmC levels With our data showing TET3 induction is an ATR-dependent event, we next want to analyze whether ATM also plays a role such regulation, or solely an ATR-specific response. With the same human ATR-deficient and non-diseased control fibroblasts pre-treated with either inhibitor against ATR (VE821) or ATM (KU55933) for an hour followed by DNA damage induction by CPT or etoposide for another 3 h later, cells were fixed and both 5hmC and γ-H2AX were analyzed. While significant inductions in intensity of 5hmC in non-diseased control fibroblast following either CPT or etoposide were observed, CPT failed to induce in 5hmC in ATR-deficient fibroblasts but treatment with etoposide resulted in strong 5hmC signals (Fig 2A–D). Similar concept was proven with pharmacological intervention. Treatment of VE821 in control fibroblast prevented CPT-induced increased 5hmC intensity, whereas KU55933 treatment prevented etoposide-induced increased intensity of 5hmC in both control and ATR-deficient fibroblasts. To assess whether TET3 enzymatic activity responds to ATR-dependent DDR, we examined the protein levels of TET3 and γ-H2AX in both control and ATR-deficient fibroblasts. As expected, CPT increased the protein level of TET3 in control cells but not in ATR-deficient fibroblasts, and pre-treatment of VE821 prevented the induction in control cells (Fig 2E). We subsequently analyzed the amount of global 5hmC was in both control and ATR-deficient fibroblasts with the dot blot assay. While 5hmC level was significantly increased in control cells after CPT or etoposide treatment (Fig 2F–I), non-significant changes were observed in CPT-treated ATR-deficient fibroblasts (Fig 2F and H). Rather, significant induction in 5hmC levels was observed when ATR-deficient fibroblasts were treated with etoposide (Fig 2G and I). Conversely, VE821 treatment impaired CPT-induced inductions of 5hmC level in control cells, and KU55933 treatment prevented etoposide-induced increased levels of 5hmC in both control and ATR-deficient fibroblasts. To exclude the possibilities that pharmacological inhibitions of ATR and ATM directly affect Tet genes expression, protein lysates and total mRNA of N2a cells pre-treated with either VE821 or KU55933 for 3 h followed by CPT treatment were harvested. Consistent with earlier findings, while VE821 prevented CPT-mediated induction in TET3 protein level, both mRNA and protein levels of Tet1, Tet2, and Tet3 genes remained stable in all treatment groups (Appendix Fig S1A–D). Figure 2. CPT-induced changes in 5hmC levels are ATR-dependent A, B. Human control and ATR-deficient fibroblasts were pre-treated with VE821 (ATR inhibitor) and KU55933 (ATM inhibitor) for 1 h followed by CPT (10 μM) or etoposide (10 μM) treatment. Six hours later, fibroblasts were fixed and stained with 5hmC and γ-H2AX antibodies. C, D. Relative intensities of 5hmC immunostaining illustrated in panel (A and B) were quantified by use of ImageJ software. Data are presented as mean ± SEM (n = 62–88 cells per treatment). Each bar represents the average of three independent experiments (*P < 0.05, unpaired t-test). E. Protein extracts from human control and ATR-deficient fibroblasts pre-treated with or without VE821 for 1 h followed by CPT (10 μM) treatment were immunoblotted with antibodies against TET3, γ-H2AX, ATR. β-Actin was used as a loading control (notes: NT, no treatment, C+P, CPT plus VE821). F, G. 5hmC-specific dot-blot intensities of genomic DNA isolated from human control and ATR-deficient fibroblasts with VE821 and KU55933 pre-treatment for 1 h followed by CPT or etoposide treatment for 6 h. H, I. Relative intensity of 5hmC-specific dot signals illustrated in panels (F and G) was quantified by use of ImageJ software. The 5hmC contents were quantified and calculated in comparison with total genomic DNA sample (300 ng/each) (*P < 0.05, unpaired t-test; mean ± SEM; n = 3–5 repetitions of the experiment). Source data are available online for this figure. Source Data for Figure 2 [embr201643179-sup-0006-SDataFig2.pdf] Download figure Download PowerPoint Next, we assessed whether ATM is also involved in CPT-mediated inductions in 5hmC level. With human control and A-T fibroblasts pre-treated with VE821 followed by etoposide or CPT treatment, respectively, cells were fixed and 5hmC intensity was analyzed (Fig EV4A). In control fibroblast, 5hmC intensity was substantially induced by upon etoposide or CPT treatment. In contrast, etoposide failed to do so in A-T cells but treatment with CPT, however, resulted in strong 5hmC signal. By contrast, VE821 prevented CPT-mediated induction of 5hmC in both control and A-T cells. We further measured the amount of global 5hmC in control and A-T fibroblasts. After etoposide or CPT treatment, 5hmC was significantly induced in control fibroblasts (Fig EV4B and C). However, in A-T cells, etoposide failed to induce 5hmC but significant inductions were observed after CPT treatment (Fig EV4B and C). Together, these findings indicate that UV- and CPT-mediated inductions in 5hmC levels are mainly dependent on ATR but not on ATM. Click here to expand this figure. Figure EV4. ATM is dispensable for DDR-induced changes in 5hmC levels Human control and A-T fibroblasts were pre-treated with VE821 (ATR inhibitor) for 1 h followed by etoposide (10 μM) or CPT (10 μM) treatment. Three hours later, fibroblasts were fixed and stained with 5hmC antibody. 5hmC-specific dot-blot intensities of genomic DNA isolated from control and A-T fibroblasts with VE821 (ATR inhibitor) pre-treatment for 1 h followed by etoposide (10 μM) or CPT (10 μM) for 3 h. Total genomic DNA was stained using methylene blue as the quality of loading control in the dot blot assays (lower panel). Quantification of three repetitions of the experiment illustrated in (B). Amounts of 5hmC are shown as a percentage of total genomic DNA (*P < 0.05, unpaired t-test; mean ± SEM). Download figure Download PowerPoint TET3 mediates CPT-mediated inductions in 5hmC To gain an in-depth insight on which specific TET enzyme is responsible for CPT-mediated inductions in 5hmC, we performed knockdown experiments with shRNA against Tet-1, Tet-2, and Tet-3. We created MEF and N2a cell which stably express shmTet1, shmTet2, shmTet3, or scramble shRNA and validated the knockdown efficiencies (Appendix Fig S2A–C). The mechanisms underlying CPT-mediated induction of 5hmC were then analyzed in these cells with various knockdowns. To our surprise, no significant reduction in basal levels of 5hmC was observed in knockdowns of Tet1, Tet2, or Tet3. However, upon CPT treatment, Tet3 knockdown is the only one showed impaired induction in 5hmC (Fig 3A and B). We quantitatively analyzed 5-hmC-specific immunostaining data and revealed a significant reduction (−1.35 ± 0.16-fold, P < 0.05) in 5hmC intensity in shmTet3-expressing cells. In contrast, no significant reductions in 5hmC intensities were observed in both shmTet1- and shmTet2-expressing cells in response to DNA damage. Lastly, we extracted the genomic DNA and again the total amount of global 5hmC was analyzed. As expected, both shmTet1 and shmTet2 failed to prevent CPT-mediated induction in 5hmC (Fig 3C and D), whereas shmTet3 significantly reduced the global 5hmC. Together, these observations indicate that TET3 is the key isoform to mediate CPT-mediated induction of 5hmC. Figure 3. CPT-induced changes in 5hmC levels are mainly mediated by TET3 Scramble shRNA-, shmTet1-, shmTet2-, and shmTet3-expressing N2a cell lines were treated with CPT (10 μM) for 6 h and then fixed and stained with 5hmC antibody. The white circles indicate positive cells. Relative intensities of 5hmC immunostaining illustrated in panel (A) were quantified by use of ImageJ software. Data are presented as mean ± SEM (n = 39–68 cells per group). Each bar represents the average of three independent experiments (*P < 0.05, unpaired t-test). 5hmC-specific dot-blot intensities of genomic DNA isolated from scramble shRNA-, shmTet1-, shmTet2-, and shmTet3-expressing N2a cell lines, respectively, after CPT (10 μM) treatment for 6 h. Total genomic DNA was stained using methylene blue as the quality of loading control in the dot blot assays (lower panel). Relative intensities of 5hmC-specific dot signals illustrated in panel (C) were quantified by use of ImageJ software. The 5hmC contents were quantified and calculated in comparison with total genomic DNA sample (300 ng/each) (*P < 0.01, unpaired t-test; mean ± SEM; n = 3 repetition of the experiment;