Human 8-Oxoguanine DNA Glycosylase Suppresses the Oxidative Stress–Induced Apoptosis Through a P53-Mediated Signaling Pathway in Human Fibroblasts

Human 8-Oxoguanine DNA Glycosylase Suppresses the Oxidative Stress–Induced Apoptosis through a p53-Mediated Signaling Pathway in Human Fibroblasts

Cha-Kyung Youn,1,2 Peter I. Song,1,5 Mi-Hwa Kim,1,2 Jin Sook Kim,6 Jin-Won Hyun,7 Sang-Joon Choi,3 Sang Pil Yoon,8 Myung Hee Chung,9 In-Youb Chang,1,4 and Ho Jin You1,2

1Korean DNA Repair Research Center, Departments of 2Pharmacology, 3Obstetrics and Gynecology, and 4Anatomy, Chosun University School of Medicine, Gwangju, South Korea; 5Department of Dermatology, University of Colorado Health Sciences Center, Denver, Colorado; 6Department of Herbal Pharmaceutical Development, Korea Institute of Oriental Medicine, Daejeon, South Korea; 7Department of Biochemistry, College of Medicine, Cheju National University, Jeju-do, South Korea; 8Department of Anatomy, College of Medicine, Seonam University, Jeollabuk-Do, South Korea; and 9Department of Pharmacology, School of Medicine, Seoul National University, Seoul, South Korea

Abstract wild-type H1299 at the same experimental condition. Human 8-oxoguanine DNA glycosylase (hOGG1) is the Moreover, the array comparative genome hybridization main defense enzyme against mutagenic effects of analyses revealed that the hOGG1-deficient GM00637 cellular 7,8-dihydro-8-oxoguanine. In this study, we showed more significant changes in the copy number of investigated the biological role of hOGG1 in DNA large regions of their chromosomes in response to H2O2 damage–related apoptosis induced by hydrogen treatment. Therefore, we suggest that although p53is a peroxide (H2O2)–derived oxidative stress. The major modulator of apoptosis, hOGG1 also plays a down-regulated expression of hOGG1 by its small pivotal role in protecting cells against the H2O2-induced interfering RNA prominently triggers the H2O2-induced apoptosis at the upstream of the p53-dependent pathway apoptosis in human fibroblasts GM00637 and human to confer a survival advantage to human fibroblasts and lung carcinoma H1299 cells via the p53-mediated human lung carcinomas through maintaining their apoptotic pathway. However, the apoptotic responses genomic stability. (Mol Cancer Res 2007;5(10):1083–98) were specifically inhibited by hOGG1 overexpression. The p53–small interfering RNA transfection into the Introduction hOGG1-deficient GM00637 markedly inhibited the Various cellular metabolites and exogenous DNA-damaging H O -induced activation of p53-downstream target 2 2 materials routinely induce cellular DNA damage, which, unless proteins such as p21, Noxa, and caspase-3/7, which promptly and properly repaired, can cause simple base or more eventually resulted in the increased cell viability. complex changes including deletions, fusions, translocations, Although the cell viability of hOGG1-knockdown or aneuploidy, resulting in cancer and a variety of genetic H1299 p53null cells was similar to that of the hOGG1 disorders (1). wild-type H1299, after the overexpression of p53the Cellular responses induced by DNA damage include the hOGG1-knockdown H1299 showed the significantly activation of several distinct biochemical pathways (2-5).First, decreased cell viability compared with that of the hOGG1 various DNA repair enzymes are activated to recognize and eliminate the damage.Second, DNA damage stimulates the specific mechanism related to cell cycle checkpoints that arrest cell cycle progression and aid in cellular survival under most circumstances.Third, apoptosis is stimulated by DNA damage Received 12/21/06; revised 6/1/07; accepted 6/6/07. Grant support: Korea Science and Engineering Foundation funded by Korea to eliminate heavily damaged or seriously deregulated cells. Government (Ministry of Science and Technology) grants M1063901001 and Although these DNA damage–induced biochemical pathways M10413010003, and Research Center for Proteinous Materials. function independently under certain circumstances, there are The costs of publication of this article were defrayed in part by the payment of page charges.This article must therefore be hereby marked advertisement in extensive interactions between these reactions.For example, the accordance with 18 U.S.C. Section 1734 solely to indicate this fact. regulatory factors in the checkpoints of cell cycle serve not only Note: P.I. Song and C-K. Youn contributed equally to this work. to delay the cell cycle but also to mediate DNA repair, both Requests for reprints: Ho Jin You, Korean DNA Repair Research Center, Bio Engineering BD 2F, Department of Pharmacology, Chosun University School directly and indirectly (3-6).However, the precise mechanisms of Medicine, 375 Seosuk-Dong, Gwangju 501-759, Republic of Korea. of assessing DNA damage both quantitatively and qualitatively Phone: 82-62-230-6337; Fax: 82-62-230-6586. E-mail: [email protected] Copyright D 2007 American Association for Cancer Research. so as to choose between mediating DNA repair and apoptosis doi:10.1158/1541-7786.MCR-06-0432 are not completely understood.

Reactive oxygen species are produced as by-products of The suppression of oxidative DNA repair activity has cellular metabolism as well as through exposure to UV, ionizing potential drawbacks; one is that the incomplete repair might radiation, and environmental carcinogens (7).The reactive result in the accumulation of mutagenic lesions in the cellular oxygen species react with DNA to produce a myriad of cytotoxic DNA.This leads to illness, death of cells, and unstoppable and mutagenic base lesions (7).Among the oxidative lesions, excessive cell division resulting in cancer and acceleration 7,8-dihydro-8-oxoguanine (8-oxoG) is the major base damage of aging process (10, 11).The human OGG1 ( hOGG1) gene produced by reactive oxygen species (3).Unlike normal is found on chromosome 3p26.2, and its allelic deletions in guanine, 8-oxoG has the propensity to mispair with adenine this region frequently occur in a variety of human cancers (9). during DNA replication and thereby gives rise to G:C to T:A It is somatically mutated in some cancer cells and is highly transversion mutations (8).Oxidatively modified bases, such as polymorphic between human population groups (12).The 8-oxoG, are mainly repaired through the base excision repair accumulation of 8-oxoG is likely to increase dramatically in pathway, the first steps being the recognition and excision of patients with various neurodegenerative diseases such as the damaged base by a specific DNA glycosylase.The major Parkinson’s disease (13), Alzheimer’s disease (14), or amyo- mammalian enzyme for removing 8-oxoG from DNA is trophic lateral sclerosis (15), which are associated with the 8-oxoguanine-DNA glycosylase (OGG1), which is a bifunc- progressive loss of cells.The hOGG1 level was also found tional enzyme with both 8-oxoG excision activity and weak AP to be lower in the orbitofrontal gyrus and the entorhinal cortex lyase strand incision activity at the abasic sites (9).Following in Alzheimer’s disease patients than in control cases.The the excision of 8-oxoG by OGG1, the resultant abasic site is accumulation of 8-oxoG increased in a majority of large further processed in sequential steps by several enzymes to motor neurons in the amyotrophic lateral sclerosis cases complete repair (2).OGG1 plays important roles in eukaryotes with decreased hOGG1 expression (14).Furthermore, several by preventing the accumulation of oxidative DNA damage, in reports have shown a correlation between the induction of both the nuclear and mitochondrial genomes, thereby suppress- apoptosis by oxidative stress and the repair of oxidative ing carcinogenesis and cell death (10). DNA lesions (16-18).Thus, mutations and deletions in the

FIGURE 1. Knockdown of hOGG1 expression by hOGG1-siRNA transfection sensitizes human fibroblast GM00637 cells to H2O2- induced apoptosis. The human fibroblast GM00637 cells were stably transfected with the control-siRNA and hOGG1-siRNA as described in Materials and Methods. The expression of hOGG1 was then analyzed by reverse transcription-PCR using hOGG1- pecific synthetic primers (A) and by Western blot using anti-hOGG1 polyclonal antibodies (B). Data were normalized with GAPDH and a-tubulin expression, respectively. C. Graph- ical plot based on A and B. Columns, mean; bars, SD. Statistically significant differences in OGG1 expression were determined by the ANOVA with P values given for both the overall (P) and the pair-wise comparisons as indicated by asterisks (*, P < 0.05). D. The OGG1 enzyme activities were analyzed by 8-oxoG nicking assay using the parental-, control-siRNA – , and two of hOGG1-siRNA – transfected cell lysates as described in Mate- rials and Methods. Purified hOGG1 and buffer alone served as positive and negative controls, respectively. Arrowhead, percent ratios of DNA cleavage products (13-mer), which are analyzed by the AxioVision Rel.4.6 Imag- ing System (Carl Zeiss International). E. Parental-, control-siRNA – , and hOGG1- siRNA – transfected cells were cultured to 70% to 80% confluence, and treated with the indicated concentration of H2O2 for 1 h. The medium was then changed to a fresh one, and the extent of cell death was assayed by trypan blue exclusion 24 h after the treatment. Cell viability was determined as the percentage of total cell number that remained unstained. Points, means from six separate experiments; bars, SD.

Mol Cancer Res 2007;5(10). October 2007 Downloaded from mcr.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. hOGG1 Inhibits the Oxidative Stress–Induced Apoptosis 1085

cellular DNA, which could arise from unrepaired oxidative and Methods, to determine if the constitutive expression of DNA lesions, have been linked to the development of human OGG1 is specifically inhibited by hOGG1-small apoptosis.However, the exact mechanism by which interfering RNA (siRNA) in cultured human fibroblast oxidative stress–mediated DNA damage causes cell death GM00637 cells.Two transfectants of hOGG1-silencing is unknown. GM00637 cells blocked the normal expression of hOGG1 In this study, we investigated the effect of hOGG1 on mRNA up to 80% to 90%, compared with that of control- oxidative DNA damage–related apoptotic responses, includ- siRNA transfectants (Fig.1A and C).Western blot analysis was ing changes of apoptotic effector caspase activity, p53 carried out to determine whether the inhibition of hOGG1 phosphorylation, and the expression pattern of its upstream mRNA corresponded to that of hOGG1 protein level.As shown or downstream proteins, together with genomic DNA in Fig.1B and C, hOGG1 protein expression was 60% to 65% instability followed by the oxidative DNA damage in human inhibited by hOGG1-siRNA transfection.To determine the fibroblasts. functional significance of the inhibited hOGG1 expression, we prepared the cell lysates from parental cells and siRNA transfectants with hOGG1-siRNA and its control-siRNA, and Results examined their abilities to cleave 8-oxoG using a 21-mer Suppression of hOGG1 Gene Expression Enhances the oligonucleotide containing a single 8-oxoG at nucleotide 13 by Cytotoxic Effects of Hydrogen Peroxide in Human the glycosylase activity assay, as described in Materials and Fibroblast GM00637 Methods.As shown in Fig.1D, the hOGG1-siRNA trans- Semiquantitative reverse transcription-PCR was done using fectants showed the inefficient cleavage of the 8-oxoG:C sub- hOGG1-specific synthetic primers, as described in Materials strates compared with the parental cells or the control-siRNA

FIGURE 1 Continued.F. Apoptotic sub-G1 DNA contents were estimated by FACScan analysis after treatment with the indicated H2O2 concentration under the same experimental condition as in E. Results are representative of three independent experiments. G. DNA damages of control-siRNA – and hOGG1-siRNA – transfected cells induced by 30 Amol/L H2O2 were analyzed by single-cell gel electrophoresis (comet assay) as described in Materials and Methods. The comet tail moment of 80 cells for each experimental condition was quantified with Komet 5.5 (Andor Technology). Columns, mean; bars, SD.

transfectants, demonstrating that hOGG1-siRNA causes the the caspase-3/7 activity assay using a Promega CaspaseGlo 3/7 reduced expression of hOGG1 protein and the decrease of DNA kit, as described in Materials and Methods.The enhanced damage repair activity in human fibroblast GM00637 cells.We caspase-3/7 activity of hOGG1-siRNA transfectants in response next investigated the hOGG1 effect on cellular cytotoxicity in to H2O2 was significantly inhibited by caspase inhibitors parental cells and hOGG1-siRNA transfectants in response to (Fig.2B).The cell viability of the H 2O2-induced hOGG1- hydrogen peroxide (H2O2) oxidative stress.The cell viability of siRNA transfectants was significantly decreased in comparison hOGG1-siRNA–transfected GM00637 was significantly lower with that of their control-siRNA transfectants.In contrast, than those of control-siRNA transfectants and parental cells treatment with caspase-3/7 inhibitors causes almost complete (Fig.1E), indicating that the inhibition of hOGG1 expression recovery from the decreased cell viability of hOGG1-siRNA reduces the cell viability of fibroblasts in response to oxidative transfectants compared with that of control-siRNA–transfected stress.To determine if hOGG1 depletion has any effect on cells (Fig.2C).This indicates that caspase-3/7 activation plays an H2O2-induced apoptosis, the nuclei of GM00637 cells were important role in the development of H2O2-induced apoptosis in stained with a propidium iodide dye for flow cytometric hOGG1-depleted human fibroblast GM00637 cells. detection.The control-siRNA–transfected cells showed the apoptotic sub-G1 DNA content of 4.12% to 12.35%, when they Requirement of p53 Activation for H2O2-Induced Cell were treated with 0 to 60 Amol/L H2O2.In contrast, the hOGG1- Death in hOGG1-Deficient Cells siRNA transfection increased the proportion of apoptotic sub-G1 In response to a variety of death stimuli, p53 is stabilized DNA content from 4.97% to 24.17% at the same oxidative stress and rapidly activated in many cell types (20).One important (Fig.1F).This shows that the suppression of hOGG1 expression mechanism through which p53 promotes caspase activation is triggers the increased apoptotic responses of human fibroblasts the transactivation of proapoptotic proteins (21).To determine in response to H2O2. the contribution of a closely correlated interaction between p53 To determine whether the elevated apoptosis is arising from and hOGG1 to the oxidative stress–induced apoptosis, we DNA breaks induced by H2O2-oxidative stress, we analyzed the analyzed the changed pattern of p53 phosphorylation, the genotoxic potential of the oxidative stress on hOGG1-deficient followed regulation of p53-downstream target proteins such as GM00637 cells by use of the single-cell gel electrophoresis p21 and Noxa, the variation of apoptotic effector caspase-3/7 assay (comet assay).Comet assay is specific for detecting activity, and the trend of cellular cytotoxicity in hOGG1- various forms of DNA damage, such as single- and double- deficient and/or p53-knockdown cells and their control cells in strand breaks, in virtually any eukaryotic cell.Double-strand response to H2O2 treatment.The significantly higher level of 15 20 breaks within DNA damages can be detected by merely the H2O2-induced p53 phosphorylation at Ser and Ser , subjecting them to electrophoretic mobility at the neutral pH followed by the more activated expression of p21 and Noxa, (e.g., pH 7-8); in contrast, single-strand breaks can be detected were shown in hOGG1-deficient GM00637 cells, but not in at alkali at pH greater than pH 12.1 (19). As indicated in their control cells, in a time course–dependent manner after Fig.1G, 30 Amol/L H2O2 significantly increased both single 100 Amol/L H2O2 treatment (Fig.3A and B).These results and double DNA strand breaks of hOGG1-siRNA–transfected indicate that p53 contributes to the oxidative stress–induced cells compared with the control cells.Therefore, we suggest cellular apoptotic responses in hOGG1-deficient GM00637 that hOGG1 suppression in human fibroblasts triggers the cells.Thus, we used p53-siRNA to silence p53 expression to increased genotoxicities, including both DNA single- and test the p53-specific effect on the H2O2-induced apoptosis in double-strand breaks, through which apoptotic responses are the hOGG1-deficient GM00637.The p53 knockdown of significantly augmented. hOGG1-deficient cells by its siRNA resulted in the predom- inant decreases of the activated caspase-3/7 expression and its hOGG1-Knockdown GM00637Cells Lead to the Signif- activity in a dose-dependent manner (Fig.3C and D).We next icant Increases of Caspase-3/7Activities in Response to examined the cell cytotoxicity in response to H2O2 treatment H2O2-Oxidative Stress under the same experimental conditions (Fig.3E).The p53 The activation of caspases is one of the major processes in depletion in the hOGG1-deficient cells markedly restored their DNA damage–induced apoptosis.To understand the protective cell viability from the H2O2-induced cell death.These results effect of hOGG1 on H2O2-induced apoptosis, we examined the indicate that the p53 activation is required for H2O2-induced induced level of caspase-3/7, the key apoptotic caspases, in cellular apoptosis in hOGG1-deficient cells. hOGG1-siRNA–transfected GM00637 cells and their control- We further investigated the hOGG1 effect on the H2O2- siRNA transfectants by Western blot analysis.There was no induced cell death in human lung carcinoma H1299 p53 null significant difference in the levels of procaspase-3 and cells to determine the more exact relationship between p53 procaspase-7 between control-siRNA transfectants and activation and hOGG1 during apoptosis.We first transfected the hOGG1-siRNA–transfected GM00637 (data not shown). H1299 cells with either hOGG1-siRNA or its control GFP- Although none of the active processed products of both siRNA, together with p53-expression vector or its control- enzymes were detected in the control-siRNA transfectants up pcDNA3.1 vector. We observed that the expression of hOGG1 to 48 h after the H2O2 treatment, the cleavage of procaspase-3/7 was specifically down-regulated up to 25% to 35% in the into the active subunits was clearly observed in the hOGG1- hOGG1-siRNA–transfected H1299, and the p53 expression was siRNA–transfected GM00637 cells from 24 h after the H2O2 clearly identified in the H1299 cells transfected with p53 treatment (Fig.2A).The propensity is consistent with the expression vector (Fig.4A).When we examined the cell via- activation of their enzymatic activity, which were obtained from bility by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

FIGURE 2. Silencing hOGG1 triggers caspase-3 and caspase-7 activation in response to H2O2 in GM00637 cells. A. Cells were treated with 50 Amol/L H2O2 for 1 h. The activated (cleaved) caspase-3 and caspase-7 were detected by West- ern blot analysis using specific antibodies for caspase-3 and caspase-7 at the indicated time points after the H2O2 treatment. Data were normalized with a-tubulin expression. B. After control-siRNA – and hOGG1- siRNA – transfected GM00637 cells were treated with or without 60 Amol/L caspase inhibitor and the indicated concentration of H2O2, caspase-3/7 activities were analyzed using Promega Caspase Glo 3/7 kit as described in Materials and Methods. Data were plotted as the relative units of luciferase intensity of three separate experiments. Points, mean; bars, SD. C. Cell viability was determined by trypan blue exclusion under the same experimental condition as in B.The extent of cell death was estimated as the percentage of total cell number that remained unstained. The control-siRNA – transfected cells served as experimental controls in this study. Points, mean from five separate experiments; bars, SD.

bromide (MTT) assay at the various concentration of H2O2 oligonucleotide containing a single 8-oxoG at nucleotide 13 24 h after the transfection, the hOGG1-knockdown H1299 were significantly increased compared with those of parental exhibited the almost similar cell viability to that of its control- cells and its control vector transfectants (Fig.5B).Over- siRNA transfectants in response to H2O2 treatment, for example, expression of hOGG1 decreased the level of p53 phosphory- 73.6% versus 79.4% at 100 Amol/L H2O2 treatment, respectively lation in response to 100 Amol/L H2O2 in the overall time (Fig.4B).In contrast, after the transfection with p53 expression points (until 8 h; Fig.5C).The effect of overexpressed hOGG1 vector, the hOGG1 wild-type H1299 showed the significantly on H2O2-induced apoptotic responses in GM00637 cells increased cell viability in comparison with that of the hOGG1- was determined by estimation of their sub-G1 DNA contents knockdown H1299 at the same experimental condition, for (Fig.5D).The hOGG1-overexpressed cells showed an apoptotic example, 45.2% versus 28.6% at 100 Amol/L H2O2 treatment, sub-G1 DNA content ranging from 4.61% to 10.10% in response respectively (Fig.4B).Taken together, these results indicate that to 0 to 80 Amol/L H2O2, whereas its control vector–transfected p53 clearly modulates apoptosis in response to H2O2-oxidative cells displayed DNA content ranging from 1.21% to 18.41% stress, which is closely related with hOGG1 expression in human under the same oxidative stress.Because the oxidative stress– fibroblasts and human lung carcinomas. induced cellular apoptotic responses are specifically inhibited by hOGG1 overexpression in human fibroblast GM00637 cells, hOGG1 Overexpression Plays an Important Role to we postulate that, although p53 clearly modulates apoptosis in Protect GM00637against H 2O2-Induced Apoptosis response to H2O2-oxidative stress, hOGG1 also plays a To determine if hOGG1 overexpression suppresses H2O2- protective role in the H2O2-induced cell death. induced p53 phosphorylation and apoptosis, we first stably transfected the GM00637 cells with either hOGG1 expression Involvement of Ataxia Telangiectasia Mutated and DNA- vector or its control empty vector pcDNA3.1, as previously Protein Kinase in the H2O2-Induced p53 Phosphorylation described (22).The hOGG1 expression was highly augmented of hOGG1-Deficient GM00637Cells in two transfectants with hOGG1 expression vector (Fig.5A), Several protein kinases have been identified that detect and their 8-oxoG cleavage enzyme activities to the 21-mer genotoxic stress and initiate signaling pathway through p53

phosphorylation at the serine or threonine residues.Among Thus, we next examined H2O2-induced phosphorylation of the them, ataxia telangiectasia mutated (ATM) and DNA-protein ATM and DNA-PK in hOGG1-deficient and hOGG1-proficient kinase (DNA-PK) play an essential role in transmitting DNA cells.To determine whether these protein kinases participate in damage signals through phosphorylation of p53 (21, 23, 24). checkpoint control in response to H2O2-oxidative stress in

FIGURE 3. H 2 O 2 -induced changes in the expression levels of phosphorylated p53, p21, and Noxa, and significant protection by p53-siRNA against H2O2-induced cytotoxicity in hOGG1-silencing GM00637 cells. After control- siRNA – and hOGG1-siRNA – transfected GM00637 cells were treated with 100 Amol/L H2O2, whole-cell lysates were prepared after incubation with H2O2 for the indicated times. Western blot analysis was carried out using specific antibodies against p53 and the phosphorylated p53 such as p53- P(Ser15) and p53-P(Ser20)(A), and using specific antibodies against p21 and Noxa (B). Data were normalized to a-tubulin expression. Results are representative of three independent experiments. C. After control and hOGG1-deficient GM00637 cells were transfected with p53-siRNA, the cells were then treated with the indicated concentration of H2O2 for 24 h. Whole-cell lysates were analyzed by Western blotting using specific antibodies against p53, p21, and caspase-3/7. Data were normalized with h-actin expression. D. After hOGG1-silencing cells were transfected with control-siRNA or p53-siRNA, the cells were treated with the indicated H2O2 concentrations. Caspase- 3/7 activities were analyzed by Promega Caspase Glo 3/7 kit as described in Materials and Meth- ods. Data were plotted as the relative units of luciferase intensity from three separate experiments. Points, mean; bars, SD. E. Cell viability was determined by trypan blue exclusion under the same experimental condition as in D. The extent of cell death was estimated as the percentage of total cell number that remained unstained. Points, mean from six separate experiments; bars, SD.

hOGG1-deficient cells, we treated control-siRNA– and hOGG1-siRNA–transfected GM00637cells with 100 Amol/L H2O2 for 0.5, 2, or 8 h. As shown in Fig. 6A, the significantly increased phosphorylation of ATM and DNA-PK in response to H2O2 treatment was detected in the hOGG1-siRNA–transfected GM00637, but not in its control-siRNA transfectants. The hOGG1 defect also caused the enhanced Chk1 phosphorylation and the increased expression of Rad51 protein, which are ATM downstream signal molecules, in response to H2O2 treatment (Fig.6A). To more specifically investigate whether ATM and DNA- PK activation contributes to the oxidative stress–induced p53 phosphorylation in hOGG1-deficient GM00637 cells, we pretreated the hOGG1-siRNA–transfected cells and its control-siRNA transfectants either with caffeine, which is a potent ATM/ATM- and Rad3-related inhibitor, or wortmannin, which is a potent DNA-PK/ATM inhibitor.The H 2O2-induced phosphorylation of p53-Ser15 and p53-Ser20 in hOGG1- deficient cells was significantly reduced by caffeine pretreatment and by wortmannin pretreatment (Fig.6B).We next examined the cell viability of wortmannin-pretreated hOGG1- deficient cells by MTT assay.As indicated in Fig.6C, pretreatment with 30 mmol/L wortmannin completely restored the cell viability of hOGG1 knockdown cells from 58.8% to 105%, which was similar with that of control cells (100%). These results suggest that the H2O2-induced p53 phosphorylation of hOGG1-deficient cells was specifically enhanced by the activation of ATM and DNA-PK.

Array Comparative Genome Hybridization Characteriza- tion of the hOGG1-Deficient Fibroblasts after H2O2 Treatment To identify the genetic copy number changes associated with the knockdown of hOGG1, an array comparative genome hybridization (array CGH) was done using genomic DNAs derived from hOGG1 knockdown cells and its control cells, which had been treated with 10 Amol/L H2O2 for 4 weeks (Fig.7A) and 8 weeks (Fig.7B).The array CGH is a powerful molecular cytogenetic method, which enables genome-wide screening for copy number losses and gains of the chromosomal parts by single hybridization (25). Currently, array-based comparative genomic hybridization (array CGH) using bacterial artificial chromosome arrays spanning the entire human genome is used to examine genomic alterations (26).The advancements in genome-wide bacterial artificial chromosome array CGH technology have FIGURE 4. The role of p53 in H2O2-mediated cell death of hOGG1- deficient H1299 lung carcinoma p53 null cells. A. After H1299 cells were contributed to the understanding of the underlying genetic transfected with control- and/or hOGG1-siRNA together with pcDNA3.1 or changes that correspond to disease progression.The extent of p53-pcDNA3.1 expression vector, whole-cell lysates were prepared, and Western blot analysis was done using anti-hOGG1 polyclonal antibodies numerical changes and their location in DNA contents of the and anti-p53 polyclonal antibodies. Data were normalized with h-actin H2O2-induced hOGG1-deficient cells and its control cells expression. Columns, mean based on the results from Western blot were examined by the array CGH.The hOGG1-deficient cells analysis; bars, SD. Statistically significant differences in the expression of hOGG1 and p53 were determined by the ANOVA with P values given for showed more significant changes in the copy number of large both the overall (P) and the pair-wise comparisons as indicated by regions of their chromosomes, compared with their control asterisks (*, P < 0.05). B. The cells in A were cultured to 70% to 80% cells in response to H O treatment (Fig.7).We observed the confluence and treated with the indicated concentration of H2O2 for 1 h. 2 2 The medium was then changed to a fresh one, and the extent of cell death representative chromosomal losses around 1q44 (gene loci of was estimated with MTT assay kit as described in Materials and Methods. LOC339529, ZNF238, and LOC440742) and 7q36.3 (gene Cell viability was determined as the percentage of total cell number. loci of VIPR2 and LOC442366), and the clear chromosomal Points, mean from three separate experiments; bars, SD. The level of cell viability at 100 Amol/L H2O2 treatment was separately described together gains at 17q11.2 (gene loci of MYO18A), 18q23 (gene locus with the graphical presentation.

FIGURE 5. The overexpression of hOGG1 inhibits apoptotic cell death in GM00637 cells. A. After cells were transfected with either empty vector or hOGG1 expression vector, whole- cell lysates from parental, control-pcDNA3.1 – transfected GM00637 cells and two hOGG1 transfectants were then separated by 12% SDS- PAGE. The expression of hOGG1 was analyzed by Western blotting using anti- hOGG1 polyclonal antibodies. Data were normalized with a-tubulin expression. B. The OGG1 enzyme activities were analyzed by an 8-oxoG nicking assay using the parental-, empty vector transfected – , and two of hOGG1 expression vector– transfected cell lysates, as described in Materials and Methods. Purified hOGG1 was used as positive control. Arrow- head, percent ratios of DNA cleavage products (13-mer), which are analyzed by the AxioVision Rel.4.6 Imaging Sys- tem (Carl Zeiss International). C. After control and hOGG1- overexpressed GM00637 cells were treated with 100 Amol/L H2O2 for the indicated times, the level of phosphorylated p53 was examined by Western blot analysis using specific antibodies against phosphorylated p53 such as p53-P(Ser15) and p53- P(Ser20). Data were normalized to a-tubulin expression. Data are representative of three independent experiments. D. The cells transfected with control or hOGG1 expression vector were cultured to 70% to 80% confluences. After the cells were treated with the indicated concentration of H2O2 for 1 h, apoptotic sub-G1 DNA contents were estimated by FACScan analysis.

of FLJ25715 and CTDP1), and Xq13.1 (gene loci of PJA1 of all chromosomes by the array CGH.These results suggest and RP13-153N15.1), respectively, based on the precise that the suppression of hOGG1 repair activity can lead to an measurement of the level of gains (log2 ratio >0.225, green alteration of the gene copy number in response to H2O2- arrows) and losses (log2 ratio less than À0.225, red arrows) oxidative stress.

Discussion cell death, mutagenesis, aging, cancer, diabetes mellitus, Reactive oxygen species such as superoxide, hydroxyl atherosclerosis, neurologic degeneration, schizophrenia, and radicals, and H2O2 oxidize macromolecules in living cells, autoimmune disorders like arthritis (13-15, 27-29).Among leading to various types of cellular dysfunctions, for example, many modified DNA macromolecules generated by the

FIGURE 6. H2O2-induced p53 phosphorylation is specifically related with the activation of DNA-PK and ATM in hOGG1-deficient GM00637 cells. A. After transfection with control- and hOGG1-siRNA, the cells were treated with 100 Amol/L H2O2 for the indicated times. The level of normal and phosphorylated DNA-PK catalytic subunit (DNA-PKcs), ATM, and Chk1 together with the expression of Rad51 was examined by Western blot analysis using each specific antibodies as described in Materials and Methods. B and C. After hOGG1-silencing cells were pretreated either with 6 mmol/L caffeine and/or 30 Amol/L wortmannin for 30 min, the cells were then incubated with 100 Amol/L H2O2 for the indicated times. The level of phosphorylated p53 together with normal p53 was examined by Western blot analysis using specific antibodies against all related proteins. Data were normalized toh-actin expression (B). Cell viability was estimated with MTT assay kit as described in Materials and Methods (C). Data are representative of three independent experiments.

FIGURE 7. Array CGH of the control-siRNA – and hOGG1-siRNA – transfected GM00637 cells. The cells were stably transfected with control- and/or hOGG1-siRNA and then treated with 10 Amol/L H2O2 for 4 wk (A) and 8 wk (B). The genomic DNA was extracted and digested with DpnII, followed by random prime labeling with Cy3 for the test DNA derived from hOGG1-siRNA – transfected cells and Cy5 for the reference DNA from control-siRNA – transfected cells. Whole-genome array CGH profiles showed log2-transformed hybridization ratios of the test DNA versus the control-reference DNA. Representative chromosomal gains (log2 ratio >0.25, green arrows) and losses (log2 ratio less than À0.25, red arrows) are also indicated along with their chromosomal locations and candidate genes.

oxidative stresses, 8-oxoG is the most abundant product that led to the loss of 80% to 90% endogenous hOGG1 expression seems to play a major role in mutagenesis and carcinogenesis and the increased cellular sensitivity to H2O2 (Fig.1).We also (3).Hence, understanding the mechanisms for processing observed a significant increase of sub-G1 population in the oxidative DNA damage is important in elucidating the hOGG1-silencing cells but not in the control-siRNA–trans- oxidative stress–mediated diseases.Recently, it was reported fected cells (Fig.1F).The elevated apoptosis in hOGG1- that the frequency of apoptosis and necrosis is related to the depleted GM00637 cells may be caused by the increased nonrepair oxidative DNA adducts.For example, a decrease in genotoxicities, including both DNA single- and double-strand hOGG1 confers cellular hypersensitivity to H2O2 (30), and breaks (Fig.1G).These results indicate that hOGG1, as a major hOGG1-deficient human leukemia cells (KG-1) could be forced DNA repair enzyme for 8-oxoG, plays a protective role to to undergo apoptosis by 8-hydroxyguanosine (31).On the other confer a survival advantage in response to H2O2-oxidative hand, the expression of hOGG1 suppresses the oxidative stress in human fibroblasts. stress–derived DNA damages, followed by the enhanced cell Apoptosis is associated with the activation of caspases by survival (32-34).However, the exact mechanism by which a cytosolic multiprotein complex formed upon the release of oxidative stress–mediated DNA damage causes cell death is cytochrome c from permeabilized mitochondria, which results unclear. in the loss of the mitochondrial membrane potential or In this study, we examined the role of hOGG1 in the activation of a proapoptotic protein such as Bax (35). cytotoxicity of human fibroblast GM00637 cells in response The released cytochrome c from mitochondria induces the to H2O2-oxidative stress.It was achieved by manipulating the proteolytic activation of procaspase-9 and the downstream capability of the cells to remove the 8-oxoG from free radical activation of caspases-3 and caspase-7, which are key damages and by assessing the effects of this manipulation on mediators of apoptosis (36, 37).Likewise, during the H 2O2- the cellular survival after H2O2 treatment.The RNA interfer- induced apoptosis, procapase-3 and procaspase-7 were more ence–mediated knockdown of hOGG1 in the GM00637 cells intensively cleaved into detectable active forms of caspase-3

(14 and 22 kDa) and caspase-7 (26 kDa), respectively, in the contribute to the increased sensitivity of the hOGG1-deficient hOGG1-knockdown cells than in their control cells (Fig.2A). cells to oxidative stress.Further studies need to identify these The caspase-3/7 inhibitor specifically suppressed the apoptosis pathways and determine their roles in oxidative stress–induced of hOGG1-knockdown cells as shown in Fig.2C.Therefore, cellular responses in hOGG1-proficient and hOGG1-deficient we postulate in this study that unrepaired oxidative DNA cells. lesion like 8-oxoG in hOGG1-depleted cells may also play an Chatterjee et al.(39) and Achanta et al.(40) indicated that important role in the H2O2-induced apoptotic process through p53 is a major regulator of hOGG1 expression and hOGG1 the mitochondrial apoptotic pathway. activity, because p53 knockout resulted in a severe reduction in p53 is a tumor-suppressor protein that transmits the signals hOGG1 mRNA and hOGG1 protein levels, and addition of p53 arising from various forms of cellular stress to the genes as well protein increased the ability to incise 8-oxoG from a synthetic as the factors that induce cell cycle arrest and apoptosis.The oligonucleotide, suggesting involvement of p53 in damage p53 is not only induced by a variety of apoptotic stimuli but the recognition and enhancement of glycosylase activity.Although overexpression of p53 has also been shown to induce apoptosis a transcriptional/translational regulation of hOGG1 by p53 is in a variety of cell types (20, 21).Several p53 target important to their relationships in response to oxidative stress, it mitochondrial proteins, such as Bax, Noxa, and PUMA, affect is consistent with our present data at a viewpoint of feedback the mitochondrial membrane potential, which is an important down-regulated p53 activation by hOGG1.That is, p53 is able determinant of mitochondrial apoptotic signaling (21).How- to regulate the upstream of hOGG1; conversely, activated ever, p53 can also directly engage in the major cellular hOGG1 may also play a feedback down-regulation of p53 apoptotic pathways, promoting both death receptor signaling activity in response to oxidative stress. and mitochondrial perturbations (21).In this study, we show the Therefore, taken together, we suggest that although p53 is a contribution of a closely correlated interaction between p53 and major modulator of apoptosis, hOGG1 also plays a protective hOGG1 to the oxidative stress–induced apoptosis, because the role in the H2O2-induced apoptosis at the upstream of the p53- significantly higher level of the H2O2-induced p53 phosphor- dependent pathway to confer a survival advantage to human ylation at Ser15 and Ser20, followed by the more activated fibroblasts and human lung carcinomas.We postulate the expression of p53-downstream target proteins such as p21 and hOGG1 effect on protecting cells against the oxidative stress– Noxa, were shown in hOGG1-deficient GM00637 cells, but not induced apoptosis through p53-dependent apoptotic signaling in their control cells (Fig.3A and B).Moreover, the p53 pathway as shown in Fig.8: knockdown of hOGG1-deficient GM00637 resulted in the decreased activation of caspase-3/7, and markedly restored the (a) In hOGG1 wild-type, oxidative damage triggers the cell viability of hOGG1-deficient cells from the H O -induced 2 2 activation of hOGG1-related DNA repair to remove oxidative cell death (Fig.3C-E).We further confirmed the protective DNA adducts like 8-oxoG, followed by the decrease in effect of hOGG1 on the H O -induced cell death in human lung 2 2 apoptosis through blocking of p53 phosphorylation. carcinoma H1299 p53 knockout cells.The cell viability of (b) In hOGG1 knockdown by siRNA, the oxidative DNA hOGG1-knockdown H1299 was similar to that of the hOGG1 wild-type H1299, which is 73.6% versus 79.4% at 100 Amol/L H2O2 treatment, respectively (Fig.4).After the overexpression of p53, the hOGG1-knockdown H1299 showed the significantly decreased cell viability (28.6%) compared with that of the hOGG1 wild-type H1299 (45.2%) at the same experimental condition.The propensity for the cell viability in response to H2O2-oxidative stress in this study is coincided with the previous report (38), in which Chen et al.showed a significantly increased viable cells in the p53 null H1299— but not in the p53 wild-type H460 cells—in response to trichloroethylene, a potent H2O2 accumulant, time and dose dependently.In contrast, Chatterjee et al.(39, 40) recently showed that the combined suppression of hOGG1 and p53 reduced the cell viability of human mammary gland epithelia, MCF12A, when compared with suppression of each gene individually. The discrepancy of cell viability between human fibroblasts, lung carcinomas, and mammary gland epithelia in the case of both knockdowns of p53 and hOGG1 may be derived from different cellular characteristics in the fields of sensitivity against specific oxidative stress, changing period of cell cycle in response to the stress, origin of cell lines, and so on.In addition, many other cellular processes, for example, mitogen- FIGURE 8. Schematic diagram of p53-mediated apoptotic pathway in activated protein kinase activation, are also affected by the hOGG1-deficient and hOGG1-proficient cells in response to H2O2 hOGG1 defect after H2O2 treatments (34), which may exposure.

adducts are increased by oxidative damages because of the breaks, was significantly increased in hOGG1-depleted cells inhibited hOGG1 activation, followed by increased apopto- compared with control cells, at neutral and alkali pH ranges, sis through the activation of p53. respectively (Fig.1G).These results indicate that hOGG1 (c) In p53 knockdown by siRNA, hOGG1 repair activity deficiency may be a major factor of the increased DNA is comparatively ineffective in oxidative stress–induced strand breaks caused by H2O2-oxidative stress.Generally, the apoptosis.Because of the intrinsic blocking of p53 acti- oxidative stress–induced DNA damages in hOGG1-depleted vation, cell viability in p53 knockdown cells is prominently cells increase oxidized purine bases, such as 8-oxoG and high in response to any oxidative stress in comparison with Fapy-G DNA adducts, which are repaired by base excision that in the p53 wild-type cells. repair.The base excision repair requires a DNA glycosylase and an AP endonuclease, which generate gaps in ssDNA. Such interruption, if encountered by a replication fork, can p53 is phosphorylated at multiple sites by several different cause double-strand breaks (44, 45).It is also possible that protein kinases, including ATM and DNA-PK, enhancing its unrepaired oxidized purines in hOGG1-deficient cells can tetramerization, stability, and activity (20).The activation of lead to topoisomerase I–oxidized DNA cleavage complexes ATM kinases, which are important mediators of signals (46), which may activate ATM and DNA-PK.Therefore, as originating from DNA damage, is caused by ionizing depicted in Fig.8, the activation of ATM and DNA-PK irradiation and genotoxic drugs (24).DNA-PK has recently protein may represent an important signal of H2O2-oxidative been identified as a crucial molecule activating apoptotic stress, and the hOGG1 defect may enhance p53 phosphor- machinery in response to DNA damage, particularly double- ylation through the activated ATM/DNA-PK signal transduc- strand breaks (23).DNA-PK complex with p53 has been tion, resulting in the increased apoptotic cell death.However, shown to act as a sensor of abnormal DNA structure, in the exact mechanism for the H2O2-oxidative stress–induced which DNA-PK selectively regulates a p53-dependent apoptotic responses is not clear yet. apoptotic response (41).Therefore, activation of these protein There are many previous reports for the role of OGG1 in kinases in H2O2-treated hOGG1-deficient cells would shed various human tumor suppression, for example, the frequent light on the activation of its upstream signaling molecule.To loss of the OGG1 chromosomal locus in human lung and renal evaluate this possibility, we determined ATM and DNA-PK cancers, and the significantly lower OGG1 activity in lung phosphorylation in response to H2O2 in hOGG1-deficient and cancer patients (9).The inactivation or knockdown of mouse hOGG1-proficient cells.The results obtained from Western OGG1 gene also led to the accumulation of 8-oxoG, followed blot experiments revealed that the phosphorylation of ATM by an increase in the mutation frequency in liver, kidney, and protein at Ser981 and DNA-PK protein at Tyr2609 was skin (47, 48).Thus, we investigated the genomic profile of significantly higher in hOGG1-deficient cells in response to hOGG1-depleted cells in chronic exposure of H2O2 to detect H2O2 compared with that of hOGG1-proficient cells (Fig.6). the gains or the losses of specific oncogenes and tumor- Additionally, the phosphorylation of Chk1 in response to suppressor genes.The analysis of the overall genomic integrity H2O2, which has been shown to act downstream of ATM as by the high-resolution array CGH revealed a significantly signal transducers in DNA damage signaling (24), was also higher level of alteration of particular gene copy number in the significantly increased in hOGG1-deficient cells; however, the H2O2-treated hOGG1-deficient cells than in the control cells H2O2-mediated phosphorylation of Chk1 does not signifi- (Fig.7).The bacterial artificial chromosome array profiles of cantly occur in hOGG1-proficient cells.Moreover, H 2O2 the H2O2-treated hOGG1 knockdown cells showed chromo- treatment increased the level of Rad51 protein, which has somal losses of LOC339529, ZNF238, LOC440742, VIPR2, been reported to be up-regulated by posttranslational and LOC442366, and chromosomal gains of MYO18A, modifications mediated by ATM (42) in hOGG1-deficient FLJ25715, CTDP1, PJA1, and RP13-153N15.1, indicating cells but not in hOGG1-proficient cells. that hOGG1 plays an important role in maintaining the To identify the regulatory pathways of p53 activation in genomic stability against oxidative stress.We are carrying hOGG1-deficient GM00637 cells, we pretreated the cells with out the more detailed characterization of the genes, in vitro and caffeine, a potent inhibitor of ATM, and wortmannin, a potent in vivo, these days. inhibitor of DNA-PK.Caffeine significantly abolished the In summary, we show in this study that the down-regulation 15 20 H2O2-induced phosphorylation of p53-Ser and p53-Ser , of hOGG1 expression prominently triggers the H2O2-induced and wortmannin almost completely attenuated the p53-Ser20 apoptosis in human fibroblast GM00637 cells and human lung phosphorylation in hOGG1-deficient cells (Fig.6B).Because carcinoma H1299 cells via the p53-mediated apoptotic pathway. ATM and DNA-PK are normally activated by double-strand The hOGG1-mediated protection against the oxidative stress– breaks or DNA replication arrest (23, 24, 41), we postulate induced apoptosis is achieved through the down-regulation of that H2O2 treatment may increase DNA double-strand breaks caspase-3/7 and p53 phosphorylation.Although the precise in hOGG1-deficient cells.However, clustered damages such mechanism of the hOGG1-inhibited p53-dependent apoptosis as double-strand breaks are only found after treatment with a remains to be determined, it may be related to the shift in the very high concentration of H2O2 at 0jCto4jC (43).Thus, balance of damage/repair process toward the accumulation of we examined more specifically the possibility of clustered oxidative DNA lesions such as 8-oxoG.Chromosomal instabil- damages of hOGG1-deficient GM00637 cells in response to ity may also play a critical role in tumor evolution of hOGG1- H2O2-oxidative stress by comet assay.We observed that DNA deficient cells through the increased rate of genetic selection of strand breaks, including both single- and double-strand cancer genes with altered DNA copy number.

The future challenge is to define the target molecules of the Semiquantitative Reverse Transcription-PCR oxidative stress–induced p53 activation and their biological RNA extraction was carried out using the RNA-STAT-60 significance for the activated signaling pathways.Studies according to the manufacturer’s instructions (TEL-TEST, focusing on these biochemical steps would extend our Inc.). Two micrograms of the total RNA were reverse- understanding of the oxidative DNA damage signaling cascades transcribed using a M-MLV cDNA synthesis system (Prom- stimulated by oxidative stress during normal metabolic and ega), and the reverse-transcribed DNA was subjected to PCR. pathologic processes. The profile of replication cycles was denaturation at 94jCfor 50 s, annealing at 58jC for 50 s, and polymerization at 72jC Materials and Methods for 1 min.In each reaction, the expression of glyceraldehyde- Maintenance of Cell Lines 3-phosphate dehydrogenase (GAPDH) served as the internal Human fibroblast GM00637 cells (Coriell Institute for control.The primers used for PCR are as follows: hOGG1 Medical Research, Camden, NJ) were maintained in Earle’s forward, 5¶-CTGCCTTCTGGACAATCTTT-3¶; hOGG1 re- MEM supplemented with 10% fetal bovine serum, 100 units/mL verse, 5¶-TAGCCCGCCCTGTTCTTC-3¶ designed to amplify penicillin, and 100 mg/mL streptomycin (Invitrogen).Human a 551-bp region; GAPDH forward, 5¶-CCATGGAGAAGG- lung carcinoma H1299 (p53-null) cells were purchased from CTGGGG-3¶; and GAPDH reverse 5¶-CAAAGTTGTCATG- American Type Culture Collection and grown in RPMI 1640 GATGACC-3¶ designed to amplify a 194-bp region (total supplemented with 10% fetal bovine serum, 100 units/mL number of cycles, 26).The PCR products were resolved on penicillin, and 100 mg/mL streptomycin (Invitrogen).The cells 1% agarose gels, stained with ethidium bromide, and then were maintained in a humidified incubator in an atmosphere photographed. containing 5% CO2 at 37jC. Small Interfering RNA Western Blotting Three target sites within human OGG1 genes were chosen j Cultured cells were washed with PBS and lysed at 0 C from the human OGG1 mRNA sequence (Genbank accession for 30 min in a lysis buffer [20 mmol/L HEPES (pH 7.4), no.AF003595), which was extracted from the National Center 2 mmol/L EGTA, 50 mmol/L glycerol phosphate, 1% Triton for Biotechnology Information Entrez nucleotide database. X-100, 10% glycerol, 1 mmol/L DTT, 1 mmol/L phenyl- After selection, each target site was searched with National methylsulfonyl fluoride, 10 mg/mL leupeptin, 10 mg/mL Center for Biotechnology Information BLAST to confirm the aprotinin, 1 mmol/L Na3VO4, and 5 mmol/L NaF].After specificity only to the human OGG1.The sequences of the cellular protein contents were determined using a dye-binding 21-nucleotide sense and antisense RNA are as follows: A microassay (Bio-Rad), 20 g of the protein per lane were hOGG1-siRNA, 5¶-GUACUUCCAGCUAGAUGUUUU-3¶ electrophoresed on 10% SDS-polyacrylamide gels after boiling (sense) and 5¶-AACAUCUAGCUGGAAGUACUU-3¶ (anti- for 5 min in a Laemmli sample buffer.The proteins were blotted sense) for the hOGG1 gene (nucleotides 292–312); LacZ onto Hybond enhanced chemiluminescence membranes (Amer- siRNA, 5¶-CGUACGCGGAAUACUUCGAUU-3¶ (sense), sham Biosciences).After electroblotting, the membranes were 5¶-AAUCGAAGUAUUCCGCGUACGUU-3¶ (antisense) for blocked with TBS containing Tween 20 [10 mmol/L Tris-HCl the LacZ gene.These siRNAs were prepared using a (pH 7.4), 150 mmol/L NaCl, 0.1% Tween 20] and 5% milk. transcription-based method with a Silencer siRNA construction The membranes were then incubated with appropriate primary kit (Ambion) according to the manufacturer’s instructions. antibodies in a blocking buffer for 4 h.We followed the LacZ siRNA was used as the negative control.Cells were manufacturer’s protocol for dilution of all primary antibodies. transiently transfected with siRNA duplexes using Oligofect- The membranes were then washed, incubated with the amine (Invitrogen) and stably transfected with the constructed appropriate secondary antibodies (1:4,000) in a blocking buffer siRNA expression plasmid based on pSilence hygro vector for 2 h, and washed again.The blotted proteins were detected (Ambion), which includes a human U6 promoter and a using an enhanced chemiluminescence detection system hygromycin resistance gene, using LipofectAMINE (Invitro- (iNtRON Biotech). gen).If required, we selected several resistant colonies against 100 Ag/mL hygromycin in the medium after transfection. Antibodies Antibodies used in this study are anti-human OGG1 polyclonal antibody (Santa Cruz Biotechnology); anti-p53 Single-Cell Gel Electrophoresis (Comet Assay) polyclonal antibody, anti–p53-P(Ser15) polyclonal antibody, Single-cell gel electrophoresis assay was done essentially as anti–p53-P(Ser20) polyclonal antibody (Cell Signaling Tech- described by Sing et al.(49).After treatment, aliquots of nology); anti-p21 monoclonal antibody (mAb; BD PharMin- the cell suspension (50–100 AL, 0.5 Â 105 –1.0 Â 105 cells) gen); anti-Noxa mAb (Calbiochem); anti–h-actin mAb (BD were transferred to 1.5-mL Eppendorf tubes and centrifuged at PharMingen); anti–a-tubulin mAb (BD PharMingen); anti– 200 Â g for 5 min.The supernatant was discarded and the DNA-PK catalytic subunit mAb (Santa Cruz Biotechnology); pellet was mixed with 75 AL low-melting agarose (0.7% in anti–DNA-PK catalytic subunit-P (Tyr2609) (Abcam); anti- PBS) and distributed onto conventional microscope slides.The ATM mAb (Santa Cruz Biotechnology); anti–ATM-P (Ser1981) slides were precoated with normal melting agarose (0.5% in mAb (Cell Signaling Technology); anti–Chk1-P (Ser345) PBS) and dried at room temperature.After the agarose had polyclonal antibody, anti-Chk1 polyclonal antibody (Cell solidified (4jC for 10 min), a second layer of normal melting Signaling Technology); anti-Rad51 mAb (Oncogene). agarose was applied similarly to the first.The slides were then

immersed in lysis solution [2.5 mol/L NaCl, 100 mmol/L Cytotoxicity Assay Na2EDTA, 10 mmol/L Tris-HCl (pH 10), containing freshly The extent of cell death was determined by trypan blue added 1% Triton X-100 and 10% DMSO] for 1 h at 4jC and exclusion.The cell viability was measured as a percentage of then placed into a horizontal electrophoresis apparatus filled the total cell number that remained unstained.If required, the with freshly made buffer [1 mmol/L Na2EDTA, 300 mmol/L cell cytotoxicity was also assessed using a MTT-based NaOH (pH >13 for alkaline conditions or pH 7-8 for neutral colorimetric assay kit (Roche Applied Science), according to conditions)].After 20 min of preincubation (unwinding of the manufacturer’s instructions. DNA), electrophoresis was run for 30 min at a fixed voltage of 25 V and 300 mA, which was adjusted by raising or lowering Propidium Iodide Staining the level of the electrophoresis buffer in the tank.At the end of The floating and trypsin-detached GM00637 cells were the electrophoresis, slides were washed with the 70% ethanol, collected and washed once with ice-cold PBS, followed by A A stained with 50 L ethidium bromide (20 g/mL), and kept in fixing in 70% cold ethanol.The cells were then stained in PBS j a moist chamber in the dark at 4 C until analysis.All of the and propidium iodide (50 Ag/mL), RNase A (50 Ag/mL), and above-reported steps were carried out under red light to prevent 0.05% Triton X-100 for 45 min. The DNA content of the additional DNA damage. GM00673 cells was analyzed by fluorescent-activated cell sorting (FACSort, Becton Dickinson).At least 10,000 events Comet Detection and Statistical Analysis were analyzed, and the percentage of cells in sub-G1 population Cells were analyzed 24 h after staining at Â400 magnifi- was calculated.Aggregates of cell debris at the origin of cation using a fluorescence microscope (Zeiss, R.G.) equipped histogram were excluded from the sub-G1 cells. with a 50-W mercury lamp.The microscope images revealed circular shapes, indicating undamaged DNA, and comet-like Caspase-3/7Activity Assay shapes, indicating the DNA had migrated out from the head to The caspase-3/7 activity was detected by a Caspase-Glo form a tail (damaged DNA).The extension of each comet was 3/7 Assay system (Promega) after preincubating the analyzed using a computerized image analysis system (Komet GM00637 cells (1 Â 106/60-mm plate) with or without 5.5, Andor Technology) that provided a ‘‘tail moment,’’ which 60 Amol/L caspase inhibitor I (Calbiochem) for 12 h, followed is defined as the product of DNA in the tail and the mean by treating with various H2O2 concentrations (0-70 Amol/L) distance of its migration in the tail and considered to be the for another 12 h.The background luminescence associated variable most directly related to DNA damage.Calculation of with the cell culture and assay reagent (blank reaction) was the extent of DNA damage, which was not homogeneous, was subtracted from the experimental values.The activity of based on analysis of 80 to 100 randomly selected comets from caspase-3/7 was presented as the means of triplets for the each slide. given cells.

8-OxoG Glycosylase Activity Assay Bacterial Artificial Chromosome Array-CGH The cells at the exponential phase were centrifuged at Human fibroblast genomic DNAs derived from si-hygro– 800 Â g for 5 min.The cell pellets (10 6 cells per each assay) and hOGG1-siRNA–transfected GM00637 were prepared by were then suspended in 2 volumes of a homogenization using the PUREGENE DNA Isolation kit (Gentra D5500A). buffer [50 mmol/L Tris-HCl, 50 mmol/L KCl, 1 mmol/L For each CGH hybridization, we digested 2 Ag of genomic EDTA, 5% glycerol, and 0.05% 2-mercaptoethanol (pH 7.5)] DNA from the reference and the corresponding experimental and homogenized.After the supernatants were obtained by sample with DpnII.All digests were done for a minimum of centrifugation, followed by dialyzing against the homogeni- 2hat37jC and then verified by agarose gel analysis.The zation buffer, they were used for the endonuclease-nicking samples were then filtered by using the QIAQuick PCR assay as cell extracts.8-OxoG–containing 21-mer with clean-up kit (Qiagen).Labeling reactions were done with 0.8 the sequence 5¶-CAGCCAATCAGTXCACCATTC-3¶ (X = Ag of purified DNA and a Bioprime labeling kit (Invitrogen) 8-oxoG) along with its complementary strand was chemically according to the manufacturer’s instructions in a volume of synthesized (The Midland Certified Reagent Co.). The 50 AL with a modified deoxynucleotide triphosphate pool synthetic oligonucleotide was 3¶-end–labeled using terminal containing 120 Amol/L each of dATP, dGTP, dTTP, 60 Amol/L transferase and [a-32P]ddATP (Amersham Biosciences, 3,000 dCTP, and 60 Amol/L Cy5-dCTP (for the reference from si- Ci/mmol).The end-labeled oligomer was annealed with its hygro–transfected GM00637) or Cy3-dCTP (for the exper- complementary oligonucleotide, and the resulting duplex imental sample from hOGG1-siRNA–transfected GM00637). DNA was used as the assay substrate.The duplex substrate Experimental and reference targets for each hybridization DNA (20 pmol) was incubated with the cell extracts (10 mg were pooled and mixed with 50 Ag of human Cot-1 DNA of protein) at 37jC for 1 h in 1 mL of reaction mixture (Invitrogen).The target mixture was ethanol precipitated [50 mmol/L Tris-HCl, 50 mmol/L KCl and 1 mmol/L EDTA and resuspended to a final volume of 40 AL hybridization (pH 7.5)]. The reaction was quenched by heating at 90jC buffer.Before hybridization to the array, the hybridization for 3 min, followed by electrophoresis on 20% denaturing mixtures were denatured at 70jC for 15 min and applied to (7 mol/L urea) polyacrylamide gels (DNA sequencing gel). the array CGH chip slide, MacArray Karyo 4000 (Macrogen The gels were wrapped in saran wrap and exposed to Kodak Korea).Hybridization was carried out for 48 to 72 h at film for visualization. 37jC in a rotating chamber (MAUI, BioMicro Systems, Inc.).

The arrays were then disassembled in 50% formamide, 2Â SSC 8-hydroxyguanine in mitochondria of MCF-7 and MDA-MB-468 human breast at 46jC for 15 min, followed by washing with 2Â SSC, cancer cell lines.Cancer Res 2002;62:1349 – 55. 0.1% SDS at 46jC for 30 min.Slides were dried and 18.Hyun JW, Jung YC, Kim HS, et al.8-hydroxydeoxyguanosine causes death of human leukemia cells deficient in 8-oxoguanine glycosylase 1 activity by scanned with a GenePix4000B microarray scanner (Axon inducing apoptosis.Mol Cancer Res 2003;1:290 – 9. Instruments-Molecular Devices).Microarray images were 19.Collins AR.The comet assay for DNA damage and repair: principles, analyzed by using Macviewer software (Macrogen).Default applications, and limitations.Mol Biotechnol 2004;26:249 – 61. settings were used.Whole-genome array CGH profiles 20.Lavin MF, Gueven N.The complexity of p53 stabilization and activation. Cell Death Differ 2006;13:941 – 50. showed log2-transformed hybridization ratios of test DNA versus control-reference DNA as described in the figure 21.Haupt S, Berger M, Goldberg Z, Haupt Y.Apoptosis—the p53 network. J Cell Sci 2003;116:4077 – 85. legends. 22.Lee MR, Kim SH, Cho HJ, et al.Transcription factors NF-YA regulate the induction of human OGG1 following DNA-alkylating agent methylmethane sulfonate (MMS) treatment.J Biol Chem 2004;279:9857 – 66. Statistical Analysis 23.Collis SJ, DeWeese TL, Jeggo PA, Parker AR.The life and death of DNA- Results are expressed as mean F SD.For statistical analysis, PK.Oncogene 2005;24:949 – 61. ANOVA with P values were done for both the overall (P) and 24.Shiloh Y.ATM and ATR: networking cellular responses to DNA damage. the pair-wise comparison as indicated by asterisks.Values of Curr Opin Genet Dev 2001;11:71 – 7. P < 0.05 were considered to be significant. 25.Kallioniemi A, Kallioniemi OP, Sudar D, et al.Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors.Science 1992; 258:818 – 21. References 26.Wolf M, Mousses S, Hautaniemi S, et al.High-resolution analysis of gene 1.Eyfjord JE, Bodvarsdottir SK.Genomic instability and cancer: networks copy number alterations in human prostate cancer using CGH on cDNA involved in response to DNA damage.Mutat Res 2005;592:18 – 28. microarrays: impact of copy number on gene expression.Neoplasia 2004;6: 2.Ide H, Kotera M.Human DNA glycosylases involved in the repair of 240 – 7. oxidatively damaged DNA.Biol Pharm Bull 2004;27:480 – 5. 27.Klaunig JE, Kamendulis LM.Therole of oxidative stress in carcinogenesis. 3.Fortini P, Pascucci B, Parlanti E, D’Errico M, Simonelli V, Dogliotti E. Annu Rev Pharmacol Toxicol 2004;44:239 – 67. 8-Oxoguanine DNA damage: at the crossroad of alternative repair pathways. 28.Macip S, Igarashi M, Berggren P, Yu J, Lee SW, Aaronson SA.Influenceof Mutat Res 2003;531:127 – 39. induced reactive oxygen species in p53-mediated cell fate decisions.Mol Cell 4.Sancar A, Lindsey-Boltz LA, Unsal-Kacmaz K, Linn S.Molecular Biol 2003;23:8576 – 85. mechanisms of mammalian DNA repair and the DNA damage checkpoints. 29.Macip S, Kosoy A, Lee SW, O’Connell MJ, Aaronson SA.Oxidative stress Annu Rev Biochem 2004;73:39 – 85. induces a prolonged but reversible arrest in p53-null cancer cells, involving a 5.Norbury CJ, Hickson ID.Cellular responses to DNA damage.Annu Rev Chk1-dependent G2 checkpoint.Oncogene 2006;25:6037 – 47. Pharmacol Toxicol 2001;41:367 – 401. 30.Yang N, Chaudhry MA, Wallace SS.Base excision repair by hNTH1 and 6.Mitra S, Boldogh I, Izumi T, Hazra TK.Complexities of the DNA base hOGG1: a two edged sword in the processing of DNA damage in g-irradiated excision repair pathway for repair of oxidative DNA damage.Environ Mol human cells.DNA Repair (Amst) 2006;5:43 – 51. Mutagen 2001;38:180 – 90. 31.Hyun JW, Choi JY, Zeng HH, et al.Leukemic cell line, KG-1 has a functional 7.Barnes DE, Lindahl T.Repairand genetic consequences of endogenous DNA loss of hOGG1 enzyme due to a point mutation and 8-hydroxydeoxyguanosine base damage in mammalian cells.Annu Rev Genet 2004;38:445 – 76. can kill KG-1.Oncogene 2000;19:4476 – 9. 8.Shibutani S, Takeshita M, Grollman AP.Insertion of specific bases during 32.Dobson AW, Xu Y, Kelley MR, LeDoux SP, Wilson GL.Enhanced DNA synthesis past the oxidation-damaged base 8-oxodG.Nature 1991;349: mitochondrial DNA repair and cellular survival after oxidative stress by targeting 431 – 4. the human 8-oxoguanine glycosylase repair enzyme to mitochondria.J Biol Chem 2000;275:37518 – 23. 9.Boiteux S, Radicella JP.The human OGG1 gene: structure, functions, and its implication in the process of carcinogenesis.Arch Biochem Biophys 2000; 33.Wu M, He YH, Kobune M, Xu Y, Kelley MR, Martin WJ II. 377:1 – 8. Protection of human lung cells against hyperoxia using the DNA base 10.Nakabeppu Y, Tsuchimoto D, Furuichi M, Sakumi K.The defense excision repair genes hOgg1 and Fpg.Am J Respir Crit Care Med 2002; mechanisms in mammalian cells against oxidative damage in nucleic acids and 166:192 – 9. their involvement in the suppression of mutagenesis and cell death.Free Radic 34.Kannan S, Pang H, Foster DC, Rao Z, Wu M.Human 8-oxoguanine DNA Res 2004;38:423 – 9. glycosylase increases resistance to hyperoxic cytotoxicity in lung epithelial 11.Bohr VA.Repair of oxidative DNA damage in nuclear and mitochondrial cells and involvement with altered MAPK activity.Cell Death Differ 2006;13: DNA, and some changes with aging in mammalian cells.Free Radic Biol Med 311 – 23. 2002;32:804 – 12. 35.Garrido C, Galluzzi L, Brunet M, Puig PE, Didelot C, Kroemer G. 12.Fan CY, Liu KL, Huang HY, et al.Frequent allelic imbalance and loss of Mechanisms of cytochrome c release from mitochondria.Cell Death Differ 2006; protein expression of the DNA repair gene hOGG1 in head and neck squamous 13:1423 – 33. cell carcinoma.Lab Invest 2001;81:1429 – 38. 36.Norbury CJ, Zhivotovsky B.DNA damage-induced apoptosis.Oncogene 13.Cardozo-Pelaez F, Cox DP, Bolin C.Lackof the DNA repair enzyme OGG1 2004;23:2797 – 808. sensitizes dopamine neurons to manganese toxicity during development.Gene 37.Lakhani SA, Masud A, Kuida K, et al.Caspases 3 and 7: key mediators of Expr 2005;12:315 – 23. mitochondrial events of apoptosis.Science 2006;311:847 – 51. 14.Iida T, Furuta A, Nishioka K, Nakabeppu Y, Iwaki T.Expression of 38.Chen SJ, Wang JL, Chen JH, Huang RN. Possible involvement of 8-oxoguanine DNA glycosylase is reduced and associated with neurofibril- glutathione and p53 in trichloroethylene- and perchlorethylene-induced lipid lary tangles in Alzheimer’s disease brain.Acta Neuropathol (Berl) 2002; peroxidation and apoptosis in human lung cancer cells.Free Radic Biol Med 103:20 – 5. 2002;33:464 – 72. 15.Kikuchi H, Furuta A, Nishioka K, Suzuki SO, Nakabeppu Y, Iwaki T. 39.Chatterjee A, Mambo E, Osada M, Upadhyay S, Sidransky D.Theeffect of Impairment of mitochondrial DNA repair enzymes against accumulation of p53-RNAi and p53 knockout on human 8-oxoguanine DNA glycosylase (hOgg1) 8-oxo-guanine in the spinal motor neurons of amyotrophic lateral sclerosis.Acta activity.FASEB J 2006;20:112 – 4. Neuropathol (Berl) 2002;103:408 – 14. 40.Achanta G, Huang P.Role of p53 in sensing oxidative DNA damage in 16.Harrison JF, Hollensworth SB, Spitz DR, Copeland WC, Wilson GL, response to reactive oxygen species-generating agents.Cancer Res 2004;64: LeDoux SP.Oxidative stress-induced apoptosis in neurons correlates with 6233 – 9. mitochondrial DNA base excision repair pathway imbalance.Nucleic Acids Res 41.Achanta G, Pelicano H, Feng L, Plunkett W, Huang P.Interaction of p53 and 2005;33:4660 – 71. DNA-PK in response to nucleoside analogues: potential role as a sensor complex 17.Mambo E, Nyaga SG, Bohr VA, Evans MK. Defective repair of for DNA damage.Cancer Res 2001;61:8723 – 9.

42.Daboussi F, Dumay A, Delacote F, Lopez BS.DNA double-strand break topoisomerase I-mediated DNA damage and cell death.J Biol Chem 2004;279: repair signalling: the case of RAD51 post-translational regulation.Cell Signal 14587 – 94. 2002;14:969 – 75. 47.Klungland A, Rosewell I, Hollenbach S, et al.Accumulation of premutagenic 43.Dahm-Daphi J, Sass C, Alberti W.Comparison of biological effects of DNA DNA lesions in mice defective in removal of oxidative base damage.Proc Natl damage induced by ionizing radiation and hydrogen peroxide in CHO cells.Int J Acad Sci U S A 1999;96:13300 – 5. Radiat Biol 2000;76:67 – 75. 48.Kunisada M, Sakumi K, Tominaga Y, et al.8-Oxoguanine formation induced 44.Lu AL, Li X, Gu Y, Wright PM, Chang DY.Repair of oxidative DNA by chronic UVB exposure makes Ogg1 knockout mice susceptible to skin damage: mechanisms and functions.Cell Biochem Biophys 2001;35:141 – 70. carcinogenesis.Cancer Res 2005;65:6006 – 10. 45.Kuzminov A.Single-strand interruptions in replicating chromosomes cause 49.Singh NP, McCoy MT, Tice RR, Schneider EL.A simple technique for double-strand breaks.Proc Natl Acad Sci U S A 2001;98:8241 – 6. quantitation of low levels of DNA damage in individual cells.Exp Cell Res 1988; 46.Daroui P, Desai SD, Li TK, Liu AA, Liu LF.Hydrogen peroxide induces 175:184 – 91.

Mol Cancer Res 2007;5(10). October 2007 Downloaded from mcr.aacrjournals.org on October 1, 2021. © 2007 American Association for Cancer Research. Human 8-Oxoguanine DNA Glycosylase Suppresses the Oxidative Stress−Induced Apoptosis through a p53-Mediated Signaling Pathway in Human Fibroblasts

Cha-Kyung Youn, Peter I. Song, Mi-Hwa Kim, et al.

Mol Cancer Res 2007;5:1083-1098.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/5/10/1083

Cited articles This article cites 49 articles, 14 of which you can access for free at: http://mcr.aacrjournals.org/content/5/10/1083.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/5/10/1083.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/5/10/1083. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.