Overexpression of Mitochondrial Manganese Superoxide Dismutase Protects against Radiation-induced Cell Death in the Human Hepatocellular Carcinoma Cell Line HLE¹

ROS,⁴ such as superoxide, hydrogen peroxide, and hydroxyl radicals, are molecules that contain oxygen and have higher reactivity than ground-state molecular oxygen. Ionizing radiation has been shown to generate ROS in a variety of cells (1). Recent evidence suggests that ROS play an important role in cell death and signal transduction by ionizing radiation (2). When water, the most copious intracellular material, is exposed to ionizing radiation, decomposition reactions occur, which form a variety of free radicals and molecular products (3). These products can peroxidize membrane lipids and attack proteins or DNA (4).

It has been proposed that, of these radicals, superoxide is a major factor in oxygen toxicity [superoxide theory of oxygen toxicity (5, 6)]. However, it has a limited reactivity with most biological molecules, raising questions about its toxicity per se (7). To account for the toxicity of superoxide in vivo, the secondary generation of a more reactive hydroxyl radical is proposed to occur by a superoxide-assisted Fenton reaction. The production of a hydroxyl radical by this reaction requires the interaction of superoxide, hydroxy peroxide, and suitably chelated iron, all kept at low concentrations in vivo because of efficient defense systems. Whereas the rate constant for the reduction of Fe³⁺ by the superoxide is only 1 × 10⁶ M⁻¹ s⁻¹ (8), other cellular constituents, such as ascorbic acid, can reduce iron and are present in much higher concentrations than superoxide (9). Thus, the contribution of superoxide to hydroxyl radical production by the superoxide-assisted Fenton reaction may be limited in vivo, and other reactions may play important roles in superoxide toxicity.

Among other reactions, the reaction between superoxide radicals and NO to form peroxynitrite is a center of attention. Superoxide radicals can react with NO to form peroxynitrite in high yield because NO contains an unpaired electron and is paramagnetic (10). Peroxynitrite is a potent biological oxidant that has recently been implicated in diverse forms of free radical-induced tissue injury (11, 12). The reaction of peroxynitrite with membrane lipid induces a phospholipid membrane peroxidation product without requiring iron (13). A variety of aldehydes are generated as final products when lipid hydroperoxides break down. Among them, HNE is a highly toxic nine-carbon α,β-unsaturated aldehyde that can be generated by the peroxidation of ω6-unsaturated fatty acids, such as arachidonic and linoleic acid (14, 15, 16). In biological systems, HNE originates almost exclusively from phospholipid-bound arachidonic acid and may be the most reliable and sensitive marker of lipid peroxidation (14). In vitro studies have revealed that at relatively high concentrations, HNE causes rapid cell death associated with the depletion of sulfhydryl groups, disturbances in calcium homeostasis, inhibition of key metabolic enzymes, and inhibition of protein and DNA synthesis (14).

It is essential for aerobic organisms to possess enzymatic and nonenzymatic antioxidant defense systems that deal with ROS produced as a consequence of aerobic respiration. One important family of enzymes is the SOD (EC 1.15.1.1; Ref. 17). This family of enzymes is a class of metalloproteins that catalyzes the dismutation of superoxide radicals into hydrogen peroxide and molecular oxygen (5, 6). Hydrogen peroxide is further degraded to water by other antioxidant enzymes, such as glutathione peroxidase (EC 1.11.1.9) and catalase (EC 1.11.1.6). In mammalian cells, there are three types of SOD: (a) cytosolic Cu/Zn-SOD; (b) mitochondrial Mn-SOD (18); and (c) extracellular SOD (19). The biological importance of Mn-SOD is demonstrated by the following findings: (a) a lack of Mn-SOD genes in Escherichia coli and yeast makes them hypersensitive to oxidative stress (20, 21, 22); (b) homozygous mutant mice lacking Mn-SOD died within the first 10 days after birth and showed dilated cardiomyopathy, an accumulation of lipid in the liver and skeletal muscle, and metabolic acidosis (23); (c) mutant mice lacking Mn-SOD showed degenerative injury of large central nervous system neurons, particularly in the basal ganglia and brain stem, associated with damaged mitochondria and also showed progressive motor disturbances characterized by limb weakness, rapid fatigue, and circling behavior (24); (d) transfection of Mn-SOD cDNA into cultured cells rendered the cells resistant to paraquat (25)-, tumor necrosis factor (26, 27)-, doxorubicin (27)-, mitomycin C (27)-, radiation (27, 28)-, alkaline- (29)-, and cigarette smoke-induced cytotoxicity (30) and radiation-induced neoplastic transformation (31); and (e) the expression of human Mn-SOD genes in transgenic mice protected the mice against oxygen-induced pulmonary injury (32) and Adriamycin-induced cardiac toxicity (33). Thus, the expression of Mn-SOD is essential for the survival of aerobic life and the development of cellular resistance to oxygen radical-mediated toxicity.

In this study, we examined the possible role of mitochondrial Mn-SOD in radiation-induced cell death. We report here that increased expression of mitochondrial Mn-SOD suppresses radiation-induced mitochondrial ROS generation and intracellular HNE production and also prevents nuclear condensation, DNA fragmentation, and cell death. Our results demonstrate that the production of mitochondrial ROS and intracellular HNE is an important mechanism by which radiation causes cellular lethality, suggesting that the removal of superoxide radicals by Mn-SOD in the mitochondria is a critical step in preventing radiation-induced cell death.

A human hepatocellular carcinoma cell line (HLE; Ref. 34) was purchased from the Health Science Research Resources Bank of Japan Health Sciences Foundation (Osaka, Japan). pCR3.1-Uni plasmid (Invitrogen, Carlsbad, CA) containing a sense human Mn-SOD cDNA insert was a kind gift of Dr. Makoto Akashi (National Institute of Radiological Sciences, Chiba, Japan). A sequence analysis of the Mn-SOD gene in the construct showed that the sequence was identical to that of accession number Y00472, except that C (nucleotide 113) was changed to T, and C (nucleotide 529) was changed to G, which make alanine to valine and glutamine to glutamic acid, respectively. The HLE cell line was transfected using the GenePORTER transfection procedure (Gene Therapy Systems, San Diego, CA) according to the manufacturer’s instructions. Briefly, cells were plated 24 h before transfection at 60% confluence in a 60-mm dish. The cells were stably transfected with 6 μg of the pCR3.1-Uni plasmids containing a sense human Mn-SOD cDNA insert and linearized by ScaI in serum-free Dulbecco’s Modified Eagle Medium (Life Technologies, Inc., Grand Island, NY). The controls were transfected with pCR3.1-Uni plasmids without human Mn-SOD cDNA insert and linearized by ScaI. Stable clones of both Mn-SOD and control plasmid transfectants were selected with Geneticin (Life Technologies, Inc.) at a final concentration of 500 μg/ml. Selected cellular clones that expressed Mn-SOD (Mn-SOD clones 6, 7, 10, and 13) or selectable marker alone (NEO-clones 1 and 2) and parental cells (HLE) were used in all experiments. Selected clones were routinely maintained in Dulbecco’s Modified Eagle Medium containing 10% fetal bovine serum (JRH Biosciences, Lenexa, KS) and 500 μg/ml Geneticin at 37°C in humidified air containing 5% CO₂. Geneticin was removed at least 24 h before the experiments reported below were performed.

Each dish was irradiated to a dose of 15 Gy at room temperature using a Pantak 320S X-ray unit (Pantak Inc., East Haven, CT). The machine was operated at 200 kVp and 20 mA with a filter of 0.5 mm of Cu and 0.5 mm of Al. The dose rate was 1.0 Gy/min at a focus surface distance of 50 cm.

A nondenatured gel assay for SOD activity was performed according to the method described previously (35), with slight modifications. Cells were sonicated in 50 mm potassium phosphate buffer (pH 7.8). A total of 50 μg protein/lane was electrophoresed through a nondissociating riboflavin gel consisting of 5% stacking gel (pH 6.8) and 12% running gel (pH 8.8) at 4°C. To visualize the SOD activity, gels were first incubated in 2.43 mm nitroblue tetrazolium (Wako Pure Chemical Industries, Ltd., Osaka, Japan) in deionized water for 20 min and then incubated in 0.028 mm riboflavin (Wako Pure Chemical Industries, Ltd.) and 280 mm N,N,N′,N′-tetramethylethylenediamine (Sigma Chemical Co., St. Louis, MO) in 50 mm potassium phosphate buffer (pH 7.8) for 15 min in the dark. Gels were then washed in deionized water and illuminated under fluorescent light until clear zones of SOD activity were evident. The images were obtained as TIFF files by a CCD camera in connection with Power Macintosh G3 (Apple Computer, Inc., Cupertino, CA). The bands of Mn-SOD were quantified by NIH Image 1.59, which is available on the internet via a file-transfer protocol.⁵ Mn-SOD activity of the parental cell was normalized to 1, and the relative Mn-SOD activities of other cells were calculated. The mean of the integrated density obtained from the three independent files was used as a representative value for the experiment.

To determine the growth characteristics of the cells, they were plated in 60-mm tissue culture dishes at 5 × 10⁴ cells/dish and cultured for 5 days. They were then trypsinized, and the number of cells was counted using a Coulter Z1 particle counter (Beckman Coulter, Fullerton, CA) daily. All experiments were repeated three times. The cell population doubling time was calculated as 0.693 t/ln (Nt/N₀) [t is time (in hours), Nt is the cell number at time t, and N₀ is the initial cell number].

Survival after irradiation was analyzed by in vitro colony formation. A stock culture near confluence was trypsinized, and the number of cells was counted by a Coulter Z1 particle counter (Beckman Coulter). One thousand cells were plated in plastic dishes. Twenty h after plating, the cells were irradiated and then returned to the incubator. After incubation at 37°C with 5% CO₂ for 13 days, the cells were stained with crystal violet, and colonies containing more than 50 cells were counted using a dissecting microscope. The surviving fraction was calculated by dividing the mean colony count at each radiation dose by the mean colony count of the unirradiated control group. Survival curves were plotted as the log of the surviving fraction of cells versus the radiation dose. All experiments were repeated three times.

Cells grown on glass-bottomed (35-mm) dishes (MatTek Corp., Ashland, MA) were stained with Hoechst 33342 fluorescent dye (Molecular Probes, Eugene, OR). Seventy-two h after irradiation, the cells were fixed for 30 min in a solution containing 4% formaldehyde in PBS and then incubated in PBS with 1 μg/ml dye for 30 min. The cells were washed twice with PBS and then washed twice with water. Fluorescence was visualized using an IX90 inverted microscope with an UPlanAPO ×20 objective lens (Olympus Optical Co., Tokyo, Japan). The dye was excited at 340 nm, and emission was filtered with a 510 nm barrier filter. Photographs of microscope fields were taken using a C5810-01 color chilled 3CCD camera (Hamamatsu Photonics K. K., Hamamatsu, Japan). More than 300 cells/culture dish were counted, and counts were made in three separate cultures per irradiation. Analyses were performed without knowledge of the irradiation history of the culture dishes. The percentage of apoptotic cells (apoptotic index) in each culture dish was determined.

DAFs (Daiichi Pure Chemicals, Tokyo, Japan) are fluorescent indicators for NO (36). They do not react with NO itself, but with NO+ equivalents, such as nitric anhydride (N₂O₃), which are produced by autooxidation of NO. DAF-FM DA, which was a kind gift from Daiichi Pure Chemicals, is a newly synthesized DAF that permeates well into cells and is quickly converted into water-soluble DAF-FM by esterases in the cytosol, where the dye can remain for a long time. Under aerobic conditions, DAF-FM traps NO to yield highly fluorescent triazofluoresceins by nitrosation and dehydration. Triazofluoresceins are not formed in the presence of NO. Glass-bottomed (35-mm) dishes (MatTek Corp.) with monolayers were prepared for staining with DAF-FM DA. One h after irradiation, the cell culture medium was replaced with modified HBSS containing 10.0 mm HEPES, 1.0 mm MgCl₂, 2 mm CaCl₂, and 2.7 mm glucose adjusted to pH 7.3 ± 0.05. Then, the cells were loaded with 10 μm DAF-FM DA by incubation for 30 min at 37°C. Bioimages of DAF-FM DA were acquired using a CSU-10 confocal laser scanning unit (Yokogawa Electric Co., Tokyo, Japan) coupled to an IX90 inverted microscope with UPlanAPO ×20 objective lens (Olympus Optical Co.) and a C5810-01 color chilled 3CCD camera (Hamamatsu Photonics K. K.). The DAF-FM DA was excited at 488 nm, and the emission was filtered using a 515 nm barrier filter. The intensity of the laser beam, the exposure time of the 3CCD camera, and the gain of the amplifier were held at 500 μW, 5 s, and 18 decibels, respectively, to allow quantitative comparisons of the relative fluorescence intensity of the cells between groups. Cells were chosen for analysis on a random basis and scanned twice. The values for the average fluorescence intensity/cell were obtained using IPLab Spectrum version 3.0 (Scanalystics Inc., Fairfax, VA) software with some modification of the program by the author (H. J. M.). The fluorescence intensity (which was acquired by confocal laser microscopy and analyzed by computer) after 15 Gy of irradiation divided by the intensity of sham-irradiated cells, i.e., the relative fluorescence intensity, is calculated as the relative fluorescence intensity that indicates the “increment” of the intensity induced by irradiation in each cell. Note that relative fluorescence intensity is not the ratio of the fluorescence intensity to the control plasmid-transfected cells or parental cells.

DHR (Molecular Probes) is an oxidation-sensitive lipophilic dye that enters a cell and fluoresces when oxidized by mitochondrial ROS to the positively charged rhodamine 123 derivatives. The relative level of mitochondrial ROS loaded with DHR was qualified by confocal laser microscope image using the same procedures described for DAF measurements, except that the final concentration of the dye used in the study was 10 μg/ml, and the exposure time of the 3CCD camera was 1.6 s.

Glass-bottomed (35-mm) dishes (MatTek Corp.) with monolayers were prepared for immuno-fluorescence staining with monoclonal antibody directed against proteins modified with the major membrane lipid peroxidation product, HNE. This monoclonal antibody (NOF Corp., Tokyo, Japan), which was specific for HNE-modified proteins, was raised by immunizing mice with a HNE-modified keyhole limpet hemocyanin conjugate (37). The antibody was tested for cross-reactivity toward glutaraldehyde, formaldehyde, 1-hexanal, 2-hexanal, 4-hydroxy-2-hexanal, and 2-nonenal. ELISAs with these potential competitors were performed. The results indicated that the anti-HNE antibody is highly specific to HNE-derived modifications to protein. Two h after irradiation, cells were fixed with 4% formaldehyde/PBS at room temperature for 30 min and rinsed twice with PBS, and membranes were permeabilized by incubation in 95% ethanol with 5% acetic acid for 10 min. After washing twice with PBS, cells were incubated for 4 min in a blocking serum (0.1% BSA in PBS) and incubated for 1 h in anti-HNE mouse monoclonal antibody at a dilution factor of 200. The cells were rinsed twice with 0.1% BSA in PBS and reincubated with Alexa Fluor 488 goat antimouse IgG (H+L) conjugate (Molecular Probes) for 1 h at room temperature. Image acquisition and analysis were similar to those of DAF-FM DA, except that the exposure time of the 3CCD camera was 5 s.

A statistical analysis was performed by an ANOVA, followed by Fisher’s post hoc tests. P < 0.05 was considered to be statistically significant. Data were presented as the mean ± SE. Calculations were performed with a statistical package, StatView 5.0J (SAS Institute Inc., Cary, NC), on a Power Macintosh G3 (Apple Computer, Inc.).

The production of Mn-SOD activity in these transfectants was investigated in cell lysates (Fig. 1). The Mn-SOD activity of the parental cell line was normalized to 1, and the relative Mn-SOD activities of the other cells were calculated. The activity of Mn-SOD in HLE and NEO clone 1 and 2 cells was 1.0.97 ± 0.02 and 1.02 ± 0.13, respectively. The Mn-SOD activity in Mn-SOD clones 6, 7, 10, and 13 was clearly detectable, and the relative activities were 3.18 ± 0.33, 5.99 ± 1.10, 3.53 ± 0.36, and 3.18 ± 0.10, respectively, i.e., the human Mn-SOD activity in the Mn-SOD-transfected cells was greater than that in the control cells. These results confirmed that (a) human Mn-SOD was not expressed much in the hepatocellular carcinoma cell line HLE, (b) but in the Mn-SOD-transfected cells, it remained stably expressed.

Table 1 shows the doubling time of the seven cell lines. The doubling times of the cells were not very different, except for that of Mn-SOD clone 7, which presented the strongest activity of Mn-SOD.

Cell morphology was detected by phase-contrast microscopy. The control cell typically showed an epithelial-like shape, whereas the Mn-SOD-transfected cells showed a long dendritic process (data not shown).

To determine the effect of radiation on cell survival, we performed an in vitro colony formation assay. Fig. 2 presents the survival curves of all cells used in the experiment irradiated with different doses of X-rays. The SF2 values are shown in Table 2. These two results show that compared with the parental cells and control plasmid transfectants, Mn-SOD-transfected clones were more resistant to irradiation.

To determine the effect of radiation on apoptotic cell death, we performed a microscopic assessment of nuclear chromatin condensation and a fragmentation assay using Hoechst 33342 staining. The percentage of apoptotic cells (apoptotic index) in each dish cultured for 72 h after irradiation was determined. The apoptotic index after 15 Gy of irradiation, which was divided by the apoptotic index of sham (0 Gy)-irradiated cells, i.e., the relative apoptotic index, was calculated (see Table 3). The result shows that the relative apoptotic index was suppressed in the Mn-SOD-transfected cells compared with the HLE and NEO clones 1 and 2. This fact indicates that Mn-SOD suppresses radiation-induced apoptosis.

To determine the effect of Mn-SOD on radiation-induced intracellular NO generation, DAF-FM DA, a dye sensitive to a change in the intracellular NO, was used. The dye was loaded 1 h after irradiation, and the images were acquired after 30 min of incubation. The fluorescence intensity (which was acquired by confocal laser microscopy and analyzed by computer) after 15 Gy of irradiation divided by the intensity on sham (0 Gy)-irradiated cells, i.e., the relative fluorescence intensity, is shown in Table 4. This result shows that NO was almost equally increased at 15 Gy in all cells used in the experiment, indicating that Mn-SOD overexpression does not influence radiation-induced NO generation in the cell.

To determine the effect of Mn-SOD on radiation-induced mitochondrial ROS generation, a dye sensitive to a change in the mitochondrial ROS was used. For an analysis of the levels of mitochondrial ROS, we used the same analytic technique used for NO. The dye was loaded 1 h after irradiation, and the images were acquired after 30 min of incubation. The relative fluorescence intensity is shown in Table 5. This result shows that the relative fluorescence intensity of DHR was depressed in the Mn-SOD-transfected cells compared with the HLE and NEO clones 1 and 2 cells, indicating that Mn-SOD suppresses radiation-induced ROS generation in the mitochondria.

To determine whether changes in mitochondrial ROS generation are accompanied by an increase in the lipid peroxidation product, the levels of HNE were evaluated by immunohistochemical staining. The relative HNE staining intensity, which was obtained 2 h after irradiation, is shown in Table 6. This result shows that the relative HNE staining intensity was suppressed in the Mn-SOD-transfected cells compared with that in HLE and NEO clones 1 and 2, indicating that Mn-SOD suppresses the levels of the radiation-induced generation of intracellular HNE.

To better understand the relationship between mitochondrial ROS, intracellular lipid peroxidation, and cell death, we used a scattergram to plot (a) the relative DHR staining intensity against the relative HNE staining intensity, (b) the relative HNE staining intensity against the relative apoptotic index, and (c) the relative HNE staining intensity against SF2, and then we analyzed the linear regression. Fig. 3,a shows a linear regression analysis of the relative DHR staining intensity versus the relative HNE staining intensity (r = 0.922; P = 0.0013). Fig. 3,b shows a linear regression analysis of the relative HNE staining intensity versus the relative apoptotic index (r = 0.822; P = 0.0199). These results show a strong positive correlation between the mitochondrial ROS, the intracellular lipid peroxidation products, and apoptosis. Fig. 3 c illustrates a linear regression analysis of the relative HNE staining intensity versus SF2 (r = −0.696; P = 0.0853). Intracellular lipid peroxidation products have a strong correlation with apoptosis but a smaller correlation with the surviving fraction assessed by a colony formation assay.

In this study, the protective role of Mn-SOD against radiation injury was analyzed by using DHR fluorescence as a measure of ROS in mitochondria, immunohistochemical staining of HNE (a marker for lipid peroxidation), in vitro colony formation to examine cell viability, and Hoechst 33342 staining to detect apoptosis. Our results indicate that Mn-SOD seems to be an enzyme effective in protecting cells against ROS induced by ionizing radiation. Furthermore, the production of intracellular NO after irradiation was not suppressed by Mn-SOD.

In almost all aerobic cells, oxygen metabolism generates ROS, such as superoxide, hydroxyl radicals, and hydrogen peroxide. These ROS can peroxidize membrane lipids of a cell and organelles and can also attack DNA or protein (4). Among the various intracellular targets for ROS-mediated injury, mitochondria are thought to be particularly prone to ROS-induced damage (38). Because in most cells mitochondria consumes >95% of the cell’s oxygen, the NADH-coenzyme Q reductase complex and the QH2 cytochrome c segment (cytochrome bc1 segment) of the mitochondrial electron transport chain are believed to be a principal source of endogeneous ROS generation. The electron transport chain yields electrons that produce a univalent reduction of oxygen to generate superoxide radicals (39). Mitochondrial DNA is highly sensitive to mutation by these endogenous ROS because mitochondrial DNA has no introns (making it more likely that a random mutation will strike a coding DNA sequence), no protective histones, and no effective DNA repair system (40). Mutations in any of the genes coding for cytochrome oxidase, cytochrome bc1, NADH dehydrogenase, or ATPase complexes may lead to an imperfect function of these enzymes. As a matter of course, under physiological conditions, only a small amount of oxygen consumed in mitochondria for respiration is changed into superoxide radicals (41). ROS production in mitochondria by ionizing radiation was apparently increased in control plasmid transfectants and parental cells as contrasted with Mn-SOD transfectants. There is a possibility that ROS generated by radiation in the mitochondria injure mitochondrial DNA and, as a result, lead to cell death. It has also been shown that ROS-generating alkylating toxins can cause mitochondrial DNA damage (42). Because SOD is specific for the elimination of superoxide radicals (6), our results also suggest that superoxide is the primary radical that leads to the increased amount of ROS detected.

Evidence has accumulated that mitochondria are major participants in apoptosis and that the ROS produced in mitochondria contribute to cell death by acting as apoptotic signaling molecules (43). Activators of apoptosis, such as caspase-2, caspase-9, cytochrome c (an activator of caspases), and apoptosis-inducing factor, all exist in mitochondria (44, 45). The release of cytochrome c (46, 47) and of apoptosis-inducing factor (44) from mitochondria is an irreversible implementation in the apoptotic process. Mitochondria also contain members of the Bcl-2 family, which regulates apoptosis (48). Overexpression of Bcl-2 prevented tumor necrosis factor-induced apoptosis and mitochondrial transmembrane potential, which is essential for mitochondrial function (49).

Recent studies suggest that the apoptosis of neuronal cells induced by oxidative stress is mediated by HNE, the major alkenal formed from oxidative degradation of membrane lipids (50); HNE can also directly mediate apoptotic and differentiating effects in K562 cells (51). Our study shows that the mitochondrial ROS, levels of HNE, and the apoptotic index are correlated with each other. There is a possibility that mitochondrial ROS form HNE and that this might help to release cytochrome c from the mitochondria and induce apoptosis. Mn-SOD might be able to prevent this course by reducing the production of mitochondrial ROS and HNE. Fig. 4 shows a schematic of the proposed hypothesis regarding how mitochondrial ROS and lipid peroxidation products accelerate cell death and the inhibition of cell death by Mn-SOD.

We note that in HLE and NEO clone 1 and 2 cells, the increase in the total intracellular HNE levels was correlated with increased mitochondrial ROS. Furthermore, in Mn-SOD-transfected cells, there was no measurable increase in the levels of both mitochondrial ROS and total intracellular HNE, suggesting that the removal of superoxide radicals in the mitochondria reduced the cellular oxidative stress at a distance from the mitochondria. A similar previous study (28) revealed that overexpression of human Mn-SOD protected against radiation-induced cell death, as assessed by a trypan blue dye exclusion test. The dead cell with a damaged cell membrane cannot exclude the trypan blue dye. This fact shows that although Mn-SOD exists in the mitochondria, it can protect the cell membrane. Why would the elimination of superoxide radicals in the mitochondria affect the levels of total cytosolic lipid peroxidation products? It is believed that in addition to its important role in respiration, mitochondrial respiration may also play a role in supporting the cellular redox status by decreasing cytosolic superoxide radicals (52). The cytosolic superoxide radical scavenging of mitochondria enhances the spontaneous dismutation of superoxide, which diffuses into the mitochondrial intermembrane space. The mitochondrial intermembrane space has a localized proton-rich condition that can protonate superoxide radicals to form hydroperoxyl radicals, which can then spread into mitochondrial matrices and are scavenged by Mn-SOD. Thus, superoxide consumption in the mitochondria creates a gradient for superoxide radicals, which favors spreading from the cytosolic space to the mitochondrial space (52). For this reason, the removal of superoxide radicals from the mitochondria affects the levels of total cytosolic superoxide radicals and, as a result, the levels of lipid peroxidation products.

This is the first report of intracellular NO formation using a NO-detecting dye, DAF-FM DA, after irradiation. NO was produced homogeneously in the cell by irradiation according to an observation by confocal laser microscopy. In addition, the data indicated that Mn-SOD expression does not affect NO production. Similar previous research work was reported by Fricker et al. (53), who described the influence of SOD on NO production. The addition of SOD could not suppress the NO production of the RAW 264 macrophage, which was stimulated to produce NO by lipopolysaccharide and IFN-γ.

Superoxide can react with NO to form peroxynitrite and, in this way, can act as a mediator of ROS-mediated toxicity. The reaction of peroxynitrite with membrane lipid induces a membrane peroxidation product (13). The dye DHR was used to quantify the relative levels of mitochondrial ROS (54). DHR localizes to the mitochondria and fluoresces when oxidized to a positively charged rhodamine 123 derivative. Peroxynitrite is particularly effective in oxidizing DHR (55). For this reason, some investigators use this dye to measure the relative levels of peroxynitrite (12). Therefore, in our study, it is apparent that peroxynitrite has shown enhanced production in the mitochondria of control cells after irradiation. If cytosolic NO and cytosolic superoxide radicals form peroxynitrite, and this peroxynitrite induces membrane peroxidation, mitochondrial Mn-SOD cannot affect the total intracellular HNE. Therefore, there is a possibility that peroxynitrite was produced in the mitochondria and may diffuse to the cytosol and peroxidize the membrane lipids.

We thank Dr. Makoto Akashi for MnSOD vectors and Dr. Daret K. St. Clair for helpful discussion.