Ionizing Radiation-Induced Cell Death Is Partly Caused by Increase of Mitochondrial Reactive Oxygen Species in Normal Human Fibroblast Cells Author(s): Shinko Kobashigawa, Genro Kashino, Keiji Suzuki, Shunichi Yamashita, and Hiromu Mori Source: Radiation Research, 183(4):455-464. Published By: Radiation Research Society DOI: http://dx.doi.org/10.1667/RR13772.1 URL: http://www.bioone.org/doi/full/10.1667/RR13772.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

RADIATION RESEARCH

183, 455–464 (2015)

0033-7587/15 $15.00 Ó2015 by Radiation Research Society. All rights of reproduction in any form reserved. DOI: 10.1667/RR13772.1

Ionizing Radiation-Induced Cell Death Is Partly Caused by Increase of Mitochondrial Reactive Oxygen Species in Normal Human Fibroblast Cells Shinko Kobashigawa,a,1 Genro Kashino,b Keiji Suzuki,c Shunichi Yamashitac and Hiromu Moria a Department of Radiology, School of Medicine, Oita University, 1-1 Idaigaoka, Hasama-machi, Yufu City, Oita, Japan; b Advanced Molecular Imaging Center, School of Medicine, Oita University, Oita, Japan; and c Department of Radiation Medical Sciences, Atomic Bomb Disease Institute, Nagasaki University, Nagasaki, Japan

reported that the delivery of alpha radiation from polonium at a dose in excess of 250 Gy to the cytoplasm had no effect on cell proliferation. By contrast, the mean lethal dose to the nucleus was less than 1.5 Gy (1). These experiments indicate that the nucleus is at least 100 times more radiosensitive than the cytoplasm, therefore, damage to nuclear DNA has been thought to be the main cause of lethality in cells after irradiation. However, several groups have reported that cells overexpressing mitochondrial manganese superoxide dismutase (MnSOD) were more radioresistant than control cells (2–8). MnSOD is an antioxidant enzyme specifically localized in mitochondria that converts superoxide to hydrogen peroxide. Fisher et al. reported that MnSOD-overexpressing irradiated cells exhibited increased loss of phosphorylated histone H2AX protein. Furthermore, mitochondrion-targeted catalase overexpression increased the survival of irradiated cells (9), and Atkinson et al. have also shown that a mitochondriontargeted inhibitor of cytochrome c peroxidase mitigated radiation-induced death (10). These results support the hypothesis that mitochondrial reactive oxygen species (ROS) can regulate radioresistance. It is not well understood whether mitochondrial ROS induced by radiation exposure or ROS produced by mitochondrial metabolism influence radiosensitivity. The effects of radiation on cells have been classified as direct effects and indirect effects. Direct effects are the damage caused by radiation itself. Indirect effects are the damage caused by free radicals that are produced by the displacement of electrons from water by radiation. These indirect effects are considered to be responsible for about 70% of the biological effects of low-LET radiation, so they require more detailed consideration. Previously, we reported that gamma radiation delayed the induction of an increase in mitochondrial ROS that depended on Drp1 localization to mitochondria (11). Drp1 is a mitochondrial fission protein, and excessive Drp1 localization to mitochondria causes mitochondrial fragmentation. The significance of this phenomenon is still unclear, as is the question of whether it causes cell death. Other groups have also shown that radiation can induced an increase of mitochondrial ROS

Kobashigawa, S., Kashino, G., Suzuki, K., Yamashita, S. and Mori, H. Ionizing Radiation-Induced Cell Death is Partly Caused by Increase of Mitochondrial Reactive Oxygen Species in Normal Human Fibroblast Cells. Radiat. Res. 183, 455–464 (2015).

Radiation-induced cell death is thought to be caused by nuclear DNA damage that cannot be repaired. However, in this study we found that a delayed increase of mitochondrial reactive oxygen species (ROS) is responsible for some of the radiation-induced cell death in normal human fibroblast cells. We have previously reported that there is a delayed increase of mitochondrial O2–, measured using MitoSOXe Red reagent, due to gamma irradiation. This is dependent on Drp1 localization to mitochondria. Here, we show that knockdown of Drp1 expression reduces the level of DNA double-strand breaks (DSBs) remaining 3 days after 6 Gy irradiation. Furthermore, cells with knockdown of Drp1 expression are more resistant to gamma radiation. We then tested whether the delayed increase of ROS causes DNA damage. The antioxidant, 2-glucopyranoside ascorbic acid (AA-2G), was applied before or after irradiation to inhibit ROS production during irradiation or to inhibit delayed ROS production from mitochondria. Interestingly, 1 h after exposure, the AA-2G treatment reduced the level of DSBs remaining 3 days after 6 Gy irradiation. In addition, irradiated AA-2G-treated cells were more resistant to radiation than the untreated cells. These results indicate that delayed mitochondrial ROS production may cause some of the cell death after irradiation. Ó 2015 by Radiation Research Society

INTRODUCTION

Radiosensitivity has been well described in terms of the radiation-induced damage to nuclear DNA. It has been Editor’s note. The online version of this article (DOI: 10.1667/ RR13772.1) contains supplementary information that is available to all authorized users. 1 Address for correspondence: Oita University, School of Medicine, Department of Radiology, 1-1 Idaigaoka, Hasamamachi, Yufu City, Oita 879-5593, Japan; e-mail: [email protected]. 455

456

KOBASHIGAWA ET AL.

production (12–14). In addition, Ogura et al. showed that treatment with radical scavengers a few minutes after irradiation effectively suppressed radiation-induced apoptosis in leukemia cells (15). Furthermore, Koyama et al. and Kinashi et al. showed that treatment with ascorbic acid, a radical scavenger, suppressed radiation-induced mutation 20 h after irradiation (16, 17). These findings suggest that a secondary increase in ROS appears to have significant effects after irradiation, and clarification of these effects is one of the most important issues to determine the biological effects of radiation. We hypothesized that radiosensitivity is influenced not only by radiation-induced ROS but also delayed ROS caused by mitochondrial metabolism after irradiation. In the current study, we investigated whether delayed ROS produced from mitochondria-induced cell death and whether treatment with 2-glucopyranoside ascorbic acid (AA-2G), an antioxidant, 1 h after irradiation inhibited this cell death. Our results showed that the cells treated with AA-2G before or after irradiation (6 Gy) were 2 times more radioresistant than untreated cells. In addition, treatment with AA-2G 1 h after irradiation suppressed the number of DNA double-strand breaks (DSBs) that remained 3 days after irradiation. These results indicate that ROS caused by mitochondrial metabolism can indeed influence radiosensitivity. MATERIALS AND METHODS Cell Culture and Gamma-Ray Irradiation Normal human foreskin fibroblast (BJ-hTERT) cells, severe combined immunodeficiency (SD01) mouse cells and CB09 cells were cultured in Eagle’s minimum essential medium with supplements containing 10% fetal bovine serum in a 5% CO2 incubator at 378C, as described previously (18). AA-2G was purchased from Wako Pure Chemical Industries Ltd., Osaka, Japan. For the knockdown of Drp1 expression, lentivirus carrying smallhairpin RNA (sh-RNA) for Drp1: 1. clone no. TRCN0000001098, ATTAGGATTACTGATGAACCG; 2. clone no. TRCN0000010594, TAACAAATTCTAGCACCACCG (Open Biosystems, GE Healthcare, Lafayette, CO) was used. sh-RNA for EGFP was used as a control sh-RNA. BJ-hTERT cells were infected with the lentivirus, then incubated for 3 days before use in the experiment. Drp1 knockdown cells were used within 2 weeks after transfection, since longer cultivation results in mitochondrial membrane depolarization. The cells were irradiated with various doses of gamma rays from 137Cs. For survival assays, appropriate numbers of cells were plated onto 100 mm dishes and grown for 14 days in a 5% CO2 incubator at 378C to produce 100 surviving colonies per plate. Cell Cycle Analysis Cells plated onto 100 mm dishes with approximately 50% confluence were trypsinized and washed once with PBS–. Cells were resuspended in 3 ml PBS– and fixed by adding 7 ml ethanol. The cells were then washed once with PBS– and treated with 0.5% RNase (Sigma-Aldricht LLC, St. Louis, MO) in PBS– for 30 min. After treatment, the cells were added to 1 mg/ml propidium iodide (Molecular Probest, Life Technologies, Grand Island, NY) for a final concentration of 50 lg/ml. The fluorescence intensity of propidium

iodide was measured by BD FACSVersee (Becton Dickinson and Co., Franklin Lakes, NJ). Measurement of Intracellular Oxidative Stress (APF and MitoSOX Test) For detection of ROS, we used 2-[6-(4 0 -amino)phenoxy-3Hxanthen-3-on-9-yl] benzoic acid (APF) and MitoSOX Red reagents. APF selectively and dose-dependently yielded a strongly fluorescent compound, fluorescein, upon reaction with reactive nitrogen  and oxygen species such as ONOO–, OH and OCl, but not other ROS (19). MitoSOX is a fluoroprobe, a derivative of hydroethidine (HE), which was introduced for selective detection of O2– in the mitochondria of live cells. The positive charge on the phosphonium group in MitoSOX Red selectively targets this cell-permeant HE derivative to mitochondria, where it accumulates as a function of mitochondrial membrane potential and exhibits fluorescence upon oxidation and subsequent binding to mitochondrial DNA (20). The cells were washed with PBS þ (containing Mg and Ca) twice, then treated with 5 lM 2-[6-(4 0 -amino)phenoxy-3H-xanthen-3-on-9-yl] benzoic acid (APF) in PBS þ (containing Ca and Mg) solution for 60 min, or 1 lM MitoSOX Red (Molecular Probes, Life Technologies) for 20 min at 378C in a 5% CO2 incubator. After treatment, the cells were trypsinized, suspended in PBS– solution at 4 3 104 cells/ml, and fluorescent intensity was measured using an F-2700 fluorescence spectrophotometer (Hitachi Ltd., Tokyo, Japan). The APF excitation wavelength was 490 nm and the emission wavelength was 515 nm. The MitoSOX excitation wavelength was 510 nm and the emission wavelength was 580 nm. The cells were treated with APF or MitoSOX 30 min before and during irradiation, for the detection of oxidative stress immediately after irradiation. Measurement of Foci Number of Histone H2AX Phosphorylated at Ser139 (phospho-H2AX) and 53BP1 For immunofluorescence staining of phospho-H2AX and 53BP1, the cells plated onto a cover glass were washed once with PBS– and fixed with 4% formaldehyde in PBS– for 10 min. The cells were then permeabilized with 0.5% Triton X-100 in PBS– for 5 min at room temperature. After permeabilization, primary antibodies for phospho-H2AX (mouse; clone no. JBW301, Upstatet, Millipore, Billerica, MA) and 53BP1 (Bethyl Laboratories Inc., Montgomery, TX) were applied for 2 h in a 378C humidified CO2 incubator. The antibodies were dissolved in TBS-DT (20 mM Tris-HCl, 137 mM NaCl, 0.1% Tween 20, 125 g/ml ampicillin and 5% skim milk). The cells were then washed with PBS three times. Secondary antibody conjugated to Alexa Fluort 488 or 594 (Life Technologies) was applied for 1 h in an incubator. After washing with PBS five times, cover slips were mounted onto slide glasses with 10% glycerol in PBS. Digital images of the primary antibodies were acquired using fluorescence microscopy (DP72, Olympus, Japan). Measurement of Foci Number of Drp1 For detecting Drp1 attached with mitochondria, the cells plated onto a cover glass were washed once with CSK buffer and permeabilized with cold 0.5 % Triton X-100 in CSK buffer for 2 min on ice before fixation. During permeabilization, detached Drp1 in the cytosol was then washed out from the cells. The cells were then fixed with 4% formaldehyde in PBS– for 20 min. After fixation, the cells were permeabilized again with 0.5% NP-40 in PBS– for 5 min at room temperature. After permeabilization, the cells were washed with PBS– five times. Drp1 antibody (BD Biosciences, San Jose, CA) dissolved in TBS-DT was applied for 2 h in a 378C humidified CO2 incubator. The cells were then washed with PBS three times. Secondary antibody conjugated to Alexa 488 (Life Technologies) was applied for

THE EFFECT OF DELAYED ROS INDUCED BY RADIATION ON NORMAL HUMAN CELLS

457

FIG. 1. Knockdown of Drp1 expression results in radioresistance. Panel A: Western blot analysis for shEGFP-, sh-Drp1(1)- and sh-Drp1(2)-expressing cells. The control represents unirradiated cells, and the days 1–3 represent the time after 6 Gy irradiation. Panel B: Radiation survival curves for sh-EGFP, sh-Drp1(1) and shDrp1(2) cells gamma irradiated (0–6 Gy). Appropriated numbers of cells were plated onto 100 mm dishes to produce on average 100 surviving colonies, after 14 days of growth after irradiation. Colonies containing more than 30 cells were counted and plotted as the log of the survival fraction of cells versus radiation dose. Data represent mean 6 SE of three independent experiments. The data are significantly different between the shEGFP cells, and the sh-Drp1(1) and sh-Drp1(2) cells after 6 Gy irradiation (t test, P , 0.01). Panel C: Cell cycle analysis of sh-EGFP, sh-Drp1(1) and sh-Drp1(2) cells by propidium iodide staining with or without irradiation (6 Gy). Cells (20,000) were counted in each experiment at 3 days after 6 Gy irradiation by flow cytometer. Data represent mean 6 SE of three independent experiments. The percentages of S-phase cells were significantly different between sh-EGFP and sh-Drp1(2) at 3 days after 6 Gy irradiation (1.81% vs. 2.56%) (t test, P , 0.05). **P , 0.01. (The percentage of S-phase cells in sh-Drp1(1) cells was 1.99% at 3 days after 6 Gy irradiation.) 1 h in an incubator. After washing with PBS five times, cover slips were mounted onto slide glasses with 10% glycerol in PBS. Digital images of the Drp1 were acquired using fluorescence microscopy (Olympus). The numbers of Drp1 foci were counted with particle analysis of Image J software (National Institutes of Health, Bethesda, MD). Western Blot Analysis Western blot analysis was performed as previously described (21). The anti-Drp1 monoclonal antibody (clone 8, Becton Dickinson and Co., Franklin Lakes, NJ) and anti-tubulin (Cell Signaling Technologyt, Danvers, MA) were used.

RESULTS

Knockdown of Drp1 Causes Radioresistance

We previously showed that gamma irradiation with 6 Gy caused a delayed increase in mitochondrial ROS in normal human fibroblast cells (BJ-hTERT). This phenomenon was dependent on mitochondrial fragmentation caused by Drp1 localization to mitochondria after irradiation (11). In the current study, we examined the number of surviving cells after gamma irradiation in BJ-hTERT cells with knockdown of Drp1 expression (sh-Drp1) (Fig. 1A). As shown in Fig. 1B, both types of 6 Gy irradiated sh-Drp1(1) and sh-

458

KOBASHIGAWA ET AL.

FIG. 2. Knockdown of Drp1 expression reduces the number of 53BP1 foci remaining 3 days after 6 Gy irradiation. Panel A: The average number of DNA DSBs per cell, counted using 53BP1 foci. BJ-hTERT cells cultured on cover slips were not irradiated (0 Gy) or gamma irradiated (6 Gy), then fixed at the indicated time points, followed by immunofluorescence staining for 53BP1. Panel A’s top figure shows unirradiated (0 Gy) cells, the middle figure shows cells 6 h after 6 Gy irradiation and the bottom figure shows cells 3 days after 6 Gy irradiation. More than 200 cells were analyzed for each case. The average number of foci per nucleus was calculated and is indicated above the bars in the graph. Data represent mean 6 SE of three independent experiments. The data are significantly different between sh-EGFP and sh-Drp1(1) cells 3 days after 6 Gy irradiation (t test, P , 0.01). Panel B: The percentages of cells without DSBs detected in terms of 53BP1 foci. None of the cells were 53BP1 foci-negative at 6 h after 6 Gy irradiation. More than 200 cells were analyzed for each case. Data represent mean 6 SE of three independent experiments. The data are significantly different between sh-EGFP cells and sh-Drp1(1) cells 3 days after 6 Gy irradiation (t test, P , 0.01). **P , 0.01.

Drp1(2) cells were more resistant than cells expressing control sh-RNA (sh-EGFP). The cell cycle analysis revealed very little difference between the sh-EGFP cells, and the shDrp1(1) and sh-Drp1(2) cells (Fig. 1C). Fewer DSBs in Cells with Knockdown of Drp1 Expression than in Control Cells

It is well known that amplification of the DNA damage signal is correlated to cell cycle arrest and colony formation after irradiation (22, 23). Therefore, we checked the number of DNA DSBs remaining after irradiation by staining 53BP1 protein, which localizes and accumulates at DSB sites. The size of 53BP1 foci was the same in shDrp1(1) cells and sh-EGFP cells after 6 Gy irradiation. There was also no difference in the numbers of 53BP1 foci between sh-Drp1(1) cells and sh-EGFP cells 6 h after irradiation (Fig. 2A, middle panel). These findings indicate that the number of initially induced DSBs and the amplification of the DNA damage signal were the same in these cells. Interestingly, the number of 53BP1 foci

remaining 3 days after irradiation differed almost twofold between sh-Drp1 cells and sh-EGFP cells, i.e., the shDrp1(1) cells had 1.8 foci per cell and sh-EGFP cells had 4.0 foci per cell (Fig. 2A, bottom panel). Since the DNA damage signal is correlated to cell cycle arrest, we counted the cells without 53BP1 foci. As expected, the percentages of cells without 53BP1 foci differed twofold between shEGFP and sh-Drp1 (Fig. 2B). Therefore, we speculated that the differences in survival and the number of DSBs remaining after irradiation between sh-EGFP and sh-Drp1 cells were caused by a delayed increase of ROS from mitochondria. AA-2G Treatment Can Suppress ROS Produced Both Immediately and After Irradiation

We then examined whether the effects of the delayed increase of ROS could be altered by treatment with 2glucopyranoside ascorbic acid as an antioxidant. AA-2G has glucose bound to the hydroxyl group of the second carbon (C2) of ascorbic acid. This glucose protects the

THE EFFECT OF DELAYED ROS INDUCED BY RADIATION ON NORMAL HUMAN CELLS

459

ascorbic acid from high temperatures, pH, metal ions and other mechanisms of degradation, so AA-2G can function for a longer period than ascorbic acid (24). Tamari et al. reported that ascorbic acid offset growth defects observed in SOD2-depleted cells and also lowered mitochondrial superoxide to physiological levels in both SOD1- and SOD2-depleted cells (25). AA-2G was used as a nontoxic antioxidant that mimics the function of cytoplasmic and mitochondrial SODs. As shown in Fig. 3A, treatment with 2.5 mM AA-2G alone increased ROS in unirradiated cells. However, treatment with 2.5 mM AA-2G could significantly suppress the initial radiation-induced ROS (Fig. 3A). In addition, 2.5 mM AA-2G suppressed the delayed increase of intracellular ROS and mitochondrial O2– (Fig. 3B, C). However, there was no significant difference between 2.5 mM and 4 mM AA-2G treatments for the suppression of mitochondrial O2– 3 days after 6 Gy irradiation (Fig. 3C). Delayed ROS Cause the Retention of DSBs and Cell Death

To distinguish the initial radiation-induced ROS and the delayed increase of ROS from mitochondria, we applied four different AA-2G treatments (Fig. 4A): 1. no treatment; 2. pre-treatment: cells were treated with 2.5 mM AA-2G from 24 h before irradiation to 1 h after irradiation; 3. posttreatment: cells were treated with 2.5 mM AA-2G from 1 h after irradiation; and 4. pre þpost-treatment: cells were treated with 2.5 mM AA-2G from 24 h before irradiation (Fig. 4A). As shown in Fig. 4B, AA-2G-treated cells were significantly resistant to radiation compared with untreated cells. Surprisingly, the levels of resistance to radiation upon pre-treatment and post-treatment with AA-2G were the same, but pre þpost-treatment with 2.5 mM AA-2G resulted in an additive effect of the pre- and post-treatment with AA2G (Fig. 4B). Next, we examined whether delayed ROS influenced the level of DSBs remaining after 6 Gy irradiation with AA-2G treatment. As shown in Fig. 4C, there was no difference in the numbers of 53BP1 foci between untreated cells and AA-2G-treated cells (pre-, postand pre þpost-treatment cells) 6 h after irradiation. In

FIG. 3. Increase of ROS was suppressed by 2.5 mM AA-2G treatment. Panel A: Intracellular ROS is detected using APF reagent just after 6 Gy irradiation. APF reagent was applied 30 min before irradiation and the intensity was then detected immediately after irradiation. Relative APF fluorescence was normalized by the intensity of the unirradiated cells without AA-2G treatment (0 Gy, 0 mM). Data represent mean 6 SE of three independent experiments. The data are significantly different between the 6 Gy irradiated cells receiving no AA-2G treatment (0 mM) and those treated with AA-2G (t test, P , 0.01). AA-2G was applied 24 h before 6 Gy irradiation. Panel B: Intracellular ROS detected using APF reagent 3 days after 6 Gy irradiation. Relative APF fluorescence was normalized by the intensity of the unirradiated cells without AA-2G treatment (0 Gy, 0 mM). Data

represent mean 6 SE of three independent experiments. The data are significantly different between the 6 Gy irradiated cells receiving no AA-2G treatment (0 mM) and those treated with AA-2G (t test, P , 0.01). Data are also significantly different between the 0 and 6 Gy irradiated cells on day 3 with or without AA-2G treatment (t test, P , 0.01). AA-2G was applied 24 h before irradiation, and the medium was changed with or without AA-2G 1 h after irradiation. Panel C: Mitochondrial O2– detected using MitoSOX reagent 3 days after 6 Gy irradiation. Relative MitoSOX fluorescence was normalized by the intensity of the unirradiated 0 mM cells. Data represent mean 6 SE of three independent experiments. The data are significantly different between the 6 Gy irradiated cells receiving no AA-2G treatment (0 mM) and those treated with 2.5 mM or 4 mM AA-2G (t test, P , 0.01). AA-2G was applied 24 h before irradiation, and the medium was changed with or without AA-2G 1 h after irradiation. *P , 0.05; **P , 0.01.

460

KOBASHIGAWA ET AL.

FIG. 4. Treatment with 2.5 mM AA-2G results in radioresistance. Panel A: AA-2G treatment scheme. AA-2G was applied at 2.5 mM as described above. Untreated cells (no treatment) received no AA-2G and were unirradiated. Pre-treatment cells received AA-2G 24 h before irradiation to 1 h after irradiation. Post-treatment cells received AA-2G 1 h after irradiation. Pre þpost-treatment cells received AA-2G 24 h before irradiation. The medium was changed 24 h before irradiation and 1 h after irradiation. Panel B: Radiation survival curves for cells with or without 2.5 mM AA-2G and gamma irradiated (0–6 Gy). Cells were plated onto 100 mm dishes to

THE EFFECT OF DELAYED ROS INDUCED BY RADIATION ON NORMAL HUMAN CELLS

addition, there was no difference in the size of 53BP1 foci between cells with or without AA-2G treatment. This indicates that the number of initially induced DSBs and the amplification of the DNA damage signal were the same in these cells. However, the number of DSBs detected using 53BP1 foci was significantly less in the cells treated with AA-2G than in the untreated cells 3 days after 6 Gy irradiation (Fig. 4C). Also, the numbers of phospho-H2AX foci were reduced in cells post-treatment and pre þposttreatment of AA-2G compared to cells not treated 3 days after 6 Gy irradiation (Fig. 4D). These findings strongly suggest that it is the delayed increase of ROS that induces DNA damage in BJ-hTERT cells. In addition, cells without 53BP1 foci were significantly increased in the AA-2G postand pre þpost-treatment groups compared with the untreated group (Fig. 4E). Delayed ROS is Downstream of Drp1 Recruitment to Mitochondria

We next examined whether the delayed ROS is essential for Drp1 recruitment to mitochondria. To detect Drp1 attached with mitochondria, we permeabilized cells before fixation so that floating Drp1 in the cytosol could be excluded. As shown in Fig. 4F, Drp1 attached with mitochondria were increased 3 days after 6 Gy irradiation. However, there was no difference in the numbers of Drp1 foci between untreated and AA-2G-treated cells after irradiation (Fig. 4F). We previously reported that knockdown of Drp1 expression suppressed delayed ROS after 6 Gy irradiation (11). According to these data, we concluded that delayed ROS is downstream of Drp1 recruitment to mitochondria.

461

Delayed ROS Causes DSBs De Novo after 6 Gy Irradiation

It is unclear whether the remaining DNA damage due to delayed production of ROS accumulation is from new damage or from initial DNA damage that has been left unrepaired. To investigate, we used severe combined immunodeficiency (SD01) mouse cells, which have a dysfunctional non-homologous recombination repair pathway, because of its mutation in DNA-PKcs. For wild-type control cells, CB09 mouse cells were used. If remaining DNA damage is the result of initial DNA damage left unrepaired, AA-2G post-treatment should effectively decrease the number of 53BP1 foci in CB09 cells only and not in SD01 cells. However, in Fig. 5A we show that this is not the case and that mitochondrial O2– was increased 3 days after 6 Gy irradiation in both the CB09 and SD01 cells. Next we counted the number of 53BP1 foci after 6 Gy with or without 2.5 mM AA-2G treatment after irradiation to test if it would suppress delayed production of ROS. As shown in Fig. 5B, post-treatment with AA-2G significantly decreased the number of 53BP1 foci remaining 3 days after 6 Gy irradiation in both the CB09 and SD01 cells (Fig. 5B). Finally, cells post-treated with AA-2G were also found to be slightly more resistant than the untreated CB09 and SD01 cells (Fig. 5C). DISCUSSION

In the current study, we showed the levels of radiation resistance in sh-Drp1 cells and AA-2G-treated cells (Figs. 1 and 4B). Even the cells treated with AA-2G 1 h after irradiation showed radioresistance. To clarify whether this phenomenon involves a response due to the antioxidant effect of AA-2G, we treated cells with 2.5 mM glucose

produce 100 surviving colonies and grown for 14 days after irradiation. Colonies containing more than 30 cells were counted and plotted as the log of the survival fraction of cells versus radiation dose. Data represent mean 6 SE of three independent experiments. The data are significantly different between the untreated cells and those receiving AA-2G treatment (pre-, post, pre þpost-treatment) (t test, P , 0.01). Panel C: The average number of DNA DSBs per cell, counted using 53BP1 foci. BJ-hTERT cells with or without 2.5 mM AA-2G cultured on cover slips were not irradiated (0 Gy) or gamma irradiated (6 Gy). Cells were then fixed at the indicated time points, followed by immunofluorescence staining for 53BP1. More than 200 cells were analyzed for each case. The average number of foci per nucleus was calculated and is indicated above the bars in the graph. Data represent mean 6 SE of three independent experiments. The data are significantly different between the cells that did not receive 2.5 mM AA-2G (untreated) and cells treated with AA-2G (pre-, post- and pre þposttreatment of AA-2G) 3 days after 6 Gy irradiation (t test; *P , 0.05; **P , 0.01). Panel D: The average number of phospho-H2AX foci per cell. BJ-hTERT cells with or without 2.5 mM AA-2G cultured on cover slips were not irradiated (0 Gy) or gamma irradiated (6 Gy). They were then fixed at 3 days after 6 Gy irradiation and immunofluorescence stained for phosphorylated-H2AX. More than 200 cells were analyzed for each case. The average number of foci per nucleus was calculated and is indicated above the bars in the graph. Data represent mean 6 SE of three independent experiments. The data are significantly different between the cells that did not receive 2.5 mM AA-2G (untreated) and cells treated with AA-2G post- and pre þpost-treatment 3 days after 6 Gy irradiation (t test; **P , 0.01). Panel E: The percentages of cells without DSBs detected using 53BP1 foci. More than 200 cells were analyzed for each case. Data represent mean 6 SE of three independent experiments. The data are significantly different between untreated cells and post-treatment/pre þpost-treatment cells 3 days after 6 Gy irradiation (t test, P , 0.05). Panel F: The average number of Drp1 foci per cell. BJ-hTERT cells with or without 2.5 mM AA-2G cultured on cover slips were not irradiated (0 Gy) or gamma irradiated (6 Gy, day 3). Cells were then fixed at 3 days after 6 Gy irradiation, followed by immunofluorescence staining for Drp1. More than 200 cells were analyzed for each case. The average number of foci per nucleus was calculated and is indicated above the bars in the graph. Data represent mean 6 SE of three independent experiments.

462

KOBASHIGAWA ET AL.

FIG. 5. Delayed ROS production results in new DNA damage after irradiation. Panel A: Mitochondrial O2– detected using MitoSOX reagent 3 days after 6 Gy irradiation in CB09 and SD01 cells. Data represent mean 6 SE of three independent experiments. The data are significantly different between unirradiated (0 Gy) cells and irradiated cells (6 Gy, day 3) in both the CB09 and SD01 cells (t test, P , 0.01). Panel B: The average number of DNA DSBs per cell, counted using 53BP1 foci. CB09 cells and SD01 cells with or without 2.5 mM AA-2G post-treatment (treatment 1 h after irradiation) were cultured on cover slips and fixed 3 days after 6 Gy irradiation, followed by immunofluorescence staining for 53BP1. More than 100 cells were analyzed for each case. The average number of foci per nucleus was calculated and is indicated above the bars in the graph. Data represent mean 6 SE of three independent experiments. The data are significantly different between the cells that did not receive 2.5 mM AA-2G (untreated) and cells post-treated with AA-2G (treatment 1 h after irradiation) at 3 days after 6 Gy irradiation (t test; *P , 0.05; **P , 0.01). Panel C: Radiation survival curves for cells with or without 2.5 mM AA-2G post-treatment, gamma irradiated (0–6 Gy). One hour after irradiation, cells were plated onto 100 mm dishes with or without 2.5 mM AA-2G to produce 100 surviving colonies and grown for 10 days after irradiation. Colonies containing more than 30 cells were counted and plotted as the log of the survival fraction of cells versus radiation dose. Data represent mean 6 SE of three independent experiments. The data are significantly different between CB09 cells that received no AA-2G treatment and CB09 cells that received 2.5 mM AA-2G after 4 Gy irradiation (t test, P , 0.05).

THE EFFECT OF DELAYED ROS INDUCED BY RADIATION ON NORMAL HUMAN CELLS

instead of AA-2G. Radioresistance was not observed in cells treated with glucose (data not shown). To test the involvement of O2– in these observations, we treated cells with 2.5 mM N-acetylcysteine (NAC), which is an OH and H2O2 radical scavenger but not O2– scavenger. As shown in the supplementary data (http://dx.doi.org/10.1667/ RR13772.1.S1), radioresistance did not improve in cells treated with 2.5 mM NAC. These findings indicate that the increase of delayed mitochondrial O2– is partly responsible for radiation-induced cell death in 6 Gy irradiated cells. In the current study we demonstrated that AA-2G treatment did not effect Drp1 localization to mitochondria (Fig. 4F). In addition, we previously demonstrated knockdown of Drp1 expression suppressed delayed ROS (11). Thus we concluded that delayed ROS is downstream of Drp1 activation after irradiation. There was a significant difference in the level of DSB signals remaining between sh-Drp1 and sh-EGFP cells (Fig. 2). AA-2G-treated cells also showed significantly fewer 53BP1 foci than control cells (Fig. 4C). In addition, the cells post-treated with AA-2G showed an increase in the percentage of cells without 53BP1 foci compared with the untreated cells (Fig. 4D), suggesting that the delayed increase of mitochondrial ROS influences DNA damage. The data thus suggest two possibilities: 1. Delayed increase of mitochondrial ROS causes DNA damage; or 2. Delayed increase of mitochondrial ROS inhibits some of the DSB repair. In the current study we demonstrated post-treatment with AA-2G reduced the number of 53BP1 foci remaining 3 days after 6 Gy irradiation in SD01 cells. SD01 cells have a dysfunctional non-homologous end joining repair pathway, and initial DNA damage will remain unrepaired even if treated with AA-2G. Thus, we concluded that it is the delayed production of ROS that causes DNA damage de novo. Since amplification of DNA damage signal is correlated with cell cycle arrest and colony formation (23, 26), the number of cells without 53BP1 foci, as concluded from our data, indicates that delayed increase of ROS caused the difference in the number of remaining DSBs. These remaining DSBs were responsible for some of the radiation-induced cell death. The target of radiation for inducing mitochondrial dysfunction is still unclear. Zhou et al. reported that cytoplasmic irradiation with alpha particles induced mutations in nucleic DNA (27). In addition, cells irradiated with four alpha particles through the cytoplasm induced about 10% cell death. Zhang et al. reported that irradiation of cytoplasm resulted in reduced mitochondrial respiratory chain function and Drp1-dependent mitochondrial fission (28). These findings support the idea that mitochondria are a target of radiation that can induce mitochondrial dysfunction. One possibility is that the exposure of mitochondria itself induces dysfunction of the electron transport chain and the loss of mitochondrial membrane potential. Another possibility is that a nucleic DNA damage signal induces mitochondrial dysfunction. Yamamori et al. reported that

463

cells in the G2/M phase had higher mitochondrial content and cellular oxidative stress level than cells in the G1 or S phase, regardless of whether the cells were irradiated. This suggested that radiation-induced G2/M arrest contributed to the increase in the mitochondrial ROS level by causing the accumulation of cells in the G2/M phase (13). In summary, we demonstrated that radiation can cause delayed increase of ROS, which leads to new delayed DNA damage that is partially responsible for radiation-induced cell death. The delayed increase of ROS is not only observed in BJ-hTERT cells and CB09 cells, but also in human embryo cells and Syrian-golden hamster embryo (SHE) cells (data not shown). Further study is needed to understand how delayed ROS is produced and the biological significance of delayed ROS production after irradiation. SUPPLEMENTARY INFORMATION

Fig. S1. Increase of ROS was suppressed by 2.5 mM NAC treatment. Fig. S2. Intracellular ROS detected using APF reagent 3 days after irradiation (6 Gy). Fig. S3. Mitochondrial O2– detected using MitoSOX reagent 3 days after irradiation (6 Gy). Fig. S4. A scheme of 2.5 mM NAC treatment. And radiation survival curves for cells with or without 2.5 mM NAC treatment plus gamma irradiated (0–6 Gy). ACKNOWLEDGMENTS This study was supported by the Global Center of Excellence (GCOE) Program and Grants-in-Aid for Scientific Research, Japan. Received: April 21, 2014; accepted: January 13, 2015; published online: March 25, 2015

REFERENCES 1. Munro TR. Irradiation of selected parts of single cells. Ann N Y Acad Sci 1961; 95:920–31. 2. Epperly MW, Chaillet JR, Kalash R, Shaffer B, Goff J, Franicola D, et al. Conditional radioresistance of Tet-inducible manganese superoxide dismutase bone marrow stromal cell lines. Radiat Res 2013; 180:189–204. 3. Epperly MW, Epperly LD, Niu Y, Wang H, Zhang X, Franicola D, et al. Overexpression of the MnSOD transgene product protects cryopreserved bone marrow hematopoietic progenitor cells from ionizing radiation. Radiat Res 2007; 168:560–6. 4. Fisher CJ, Goswami PC. Mitochondria-targeted antioxidant enzyme activity regulates radioresistance in human pancreatic cancer cells. Cancer Biol Ther 2008; 7:1271–9. 5. Hirose K, Longo DL, Oppenheim JJ, Matsushima K. Overexpression of mitochondrial manganese superoxide dismutase promotes the survival of tumor cells exposed to interleukin-1, tumor necrosis factor, selected anticancer drugs, and ionizing radiation. FASEB J 1993; 7:361–8. 6. Hosoki A, Yonekura S-I, Zhao Q-L, Wei Z-L, Takasaki I, Tabuchi Y, et al. Mitochondria-targeted superoxide dismutase (SOD2) regulates radiation resistance and radiation stress response in HeLa cells. J Radiat Res 2012; 53:58–71. 7. Hu H, Luo ML, Du XL, Feng YB, Zhang Y, Shen XM, et al. Up-

464

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

KOBASHIGAWA ET AL.

regulated manganese superoxide dismutase expression increases apoptosis resistance in human esophageal squamous cell carcinomas. Chinese Med J-Peking 2007; 120:2092–8. Zhang X, Epperly MW, Kay MA, Chen ZY, Dixon T, Franicola D, et al. Radioprotection in vitro and in vivo by minicircle plasmid carrying the human manganese superoxide dismutase transgene. Hum Gene Ther 2008; 19:820–6. Fisher CJ, Goswami PC. Mitochondria-targeted antioxidant enzyme activity regulates radioresistance in human pancreatic cancer cells. Cancer Biol Ther 2008; 7:1271–9. Atkinson J, Kapralov AA, Yanamala N, Tyurina YY, Amoscato AA, Pearce L, et al. A mitochondria-targeted inhibitor of cytochrome c peroxidase mitigates radiation-induced death. Nat Commun 2011; 2:497. Kobashigawa S, Suzuki K, Yamashita S. Ionizing radiation accelerates Drp1-dependent mitochondrial fission, which involves delayed mitochondrial reactive oxygen species production in normal human fibroblast-like cells. Biochem Biophys Res Commun 2011; 414:795–800. Saenko Y, Cieslar-Pobuda A, Skonieczna M, Rzeszowska-Wolny J. Changes of reactive oxygen and nitrogen species and mitochondrial functioning in human K562 and HL60 cells exposed to ionizing radiation. Radiat Res 2013; 180:360–6. Yamamori T, Yasui H, Yamazumi M, Wada Y, Nakamura Y, Nakamura H, et al. Ionizing radiation induces mitochondrial reactive oxygen species production accompanied by upregulation of mitochondrial electron transport chain function and mitochondrial content under control of the cell cycle checkpoint. Free Radic Biol Med 2012; 53:260–70. Yoshida T, Goto S, Kawakatsu M, Urata Y, Li TS. Mitochondrial dysfunction, a probable cause of persistent oxidative stress after exposure to ionizing radiation. Free Radic Res 2012; 46:147–53. Ogura A, Oowada S, Kon Y, Hirayama A, Yasui H, Meike S, et al. Redox regulation in radiation-induced cytochrome c release from mitochondria of human lung carcinoma A549 cells. Cancer Lett 2009; 277:64–71. Kinashi Y, Sakurai Y, Masunaga S, Suzuki M, Nagata K, Ono K. Ascorbic acid reduced mutagenicity at the HPRT locus in CHO cells against thermal neutron radiation. Appl Radiat Isot 2004; 61:929–32. Koyama S, Kodama S, Suzuki K, Matsumoto T, Miyazaki T, Watanabe M. Radiation-induced long-lived radicals which cause mutation and transformation. Mutat Res 1998; 421:45–54.

18. Watanabe M, Suzuki M, Suzuki K, Nakano K, Watanabe K. Effect of multiple irradiation with low doses of gamma-rays on morphological transformation and growth ability of human embryo cells in vitro. Int J Radiat Biol 1992; 62:711–8. 19. Setsukinai K, Urano Y, Kakinuma K, Majima HJ, Nagano T. Development of novel fluorescence probes that can reliably detect reactive oxygen species and distinguish specific species. J Biol Chem 2003; 278:3170–5. 20. Mukhopadhyay P, Rajesh M, Hasko G, Hawkins BJ, Madesh M, Pacher P. Simultaneous detection of apoptosis and mitochondrial superoxide production in live cells by flow cytometry and confocal microscopy. Nat Protoc 2007; 2:2295–301. 21. Yamauchi M, Suzuki K, Kodama S, Watanabe M. Abnormal stability of wild-type p53 protein in a human lung carcinoma cell line. Biochem Biophys Res Commun 2005; 330:483–8. 22. Oka Y, Yamauchi M, Suzuki M, Yamashita S, Suzuki K. Persistence and dynamics of DNA damage signal amplification determined by microcolony formation and live-cell imaging. J Radiat Res 2011; 52:766–74. 23. Yamauchi M, Oka Y, Yamamoto M, Niimura K, Uchida M, Kodama S, et al. Growth of persistent foci of DNA damage checkpoint factors is essential for amplification of G1 checkpoint signaling. DNA Repair 2008; 7:405–17. 24. Huang WY, Lee PC, Huang LK, Lu LP, Liao WC. Stability studies of ascorbic acid 2-glucoside in cosmetic lotion using surface response methodology. Bioorg Med Chem Lett 2013; 23:1583–7. 25. Tamari Y, Nawata H, Inoue E, Yoshimura A, Yoshii H, Kashino G, et al. Protective roles of ascorbic acid in oxidative stress induced by depletion of superoxide dismutase in vertebrate cells. Free Radic Res 2013; 47:1–7. 26. Oka Y, Yamauchi M, Suzuki M, Yamashita S, Suzuki K. Persistence and dynamics of DNA damage signal amplification determined by microcolony formation and live-cell imaging. J Radiat Res 2011; 52:766–74. 27. Zhou H, Suzuki M, Randers-Pehrson G, Vannais D, Chen G, Trosko JE, et al. Radiation risk to low fluences of alpha particles may be greater than we thought. Proc Natl Acad Sci U S A 2001; 98:14410–5. 28. Zhang B, Davidson MM, Zhou H, Wang C, Walker WF, Hei TK. Cytoplasmic irradiation results in mitochondrial dysfunction and DRP1-dependent mitochondrial fission. Cancer Res 2013; 73:6700–10.