Adaptive Responses to Low-Dose/Low-Dose-Rate Rays in Normal ...

RADIATION RESEARCH

166, 849–857 (2006)

0033-7587/06 $15.00 䉷 2006 by Radiation Research Society. All rights of reproduction in any form reserved.

Adaptive Responses to Low-Dose/Low-Dose-Rate ␥ Rays in Normal Human Fibroblasts: The Role of Growth Architecture and Oxidative Metabolism Sonia M. de Toledo,a Nesrin Asaad,a Perumal Venkatachalam,a Ling Li,b Roger W. Howell,a Douglas R. Spitzb and Edouard I. Azzama,1 a

Department of Radiology, UMDNJ-New Jersey Medical School, Newark, New Jersey 07101; and b Free Radical and Radiation Biology Program, Department of Radiation Oncology, Holden Comprehensive Cancer Center, The University of Iowa, Iowa City, Iowa 52242

lying these effects are fairly well elucidated (1). In contrast, the biological effects and health risks of low-level radiation remain ambiguous (2, 3) and are currently the subject of intense debate (4–8). The frequency of human exposure to low-level ionizing radiation has been on the increase. In addition to exposures from natural sources (e.g. radon gas), the human population may be subjected to radiation during activities related to nuclear technology and deep space travel (9). Perhaps of greatest significance is the explosive growth in the use of diagnostic radiology, in which an increasing number of individuals are being exposed to radiation. It is predicted that by the year 2010, one in every three individuals residing in economically developed countries will have a computed tomography (CT) scan (average effective dose of 50 mSv or 5 rem), with the likelihood of several repeats in the patient’s lifetime (10). During CT, larger volumes of tissue are exposed to higher doses of radiation than occurs in more common X-ray procedures (0.1 mSv or 0.01 rem). As a result, there is a considerable public and scientific interest in characterizing the biological effects of ionizing radiation in the range of doses that these activities deliver, with a specific focus on elucidating the underlying molecular and biochemical mechanisms (11). Several approaches have been used to investigate the effects of low-level radiation. Although human epidemiological studies are highly relevant to characterize these effects, they are mainly observational in nature and therefore are of limited utility for elucidating mechanisms. Furthermore, considerable challenges are encountered in generating adequate sample sizes to discern dose–effect relationships and in controlling for confounding factors (e.g. individual sensitivity, latent period, lifestyle, etc.) that may affect results. Experimental animal studies offer some advantages over epidemiological investigations with regard to mechanistic studies. By taking advantage of specific genetic traits, these experiments are informative of how a particular gene or set of genes affects radiation responses. However, generalization of animal data to humans should be done with caution due to differences in physiology [e.g. thiol chemistry (12)],

de Toledo, S. M., Asaad, N., Venkatachalam, P., Li, L., Howell, R. W., Spitz, D. R. and Azzam, E. I. Adaptive Responses to Low-Dose/Low-Dose-Rate ␥ Rays in Normal Human Fibroblasts: The Role of Growth Architecture and Oxidative Metabolism. Radiat. Res. 166, 849–857 (2006). To investigate low-dose/low-dose-rate effects of low-linear energy transfer (LET) ionizing radiation, we used ␥-irradiated cells adapted to grow in a three-dimensional architecture that mimics cell growth in vivo. We determined the cellular, molecular and biochemical changes in these cells. Quiescent normal human fibroblasts were irradiated with single acute or chronic doses (1–10 cGy) of 137Cs ␥ rays. Whereas exposure to an acute dose of 10 cGy increased micronucleus formation, protraction of the dose over 48 h reduced micronucleus frequency to a level similar to or lower than what occurs spontaneously. The protracted treatment also up-regulated the cellular content of the antioxidant glutathione. These changes correlated with modulation of phospho-TP53 (serine 15), a stress marker that was regulated by doses as low as 1 cGy. The DNA damage that occurred after exposure to an acute dose of 10 cGy was protected against in two ways: (1) upregulation of cellular antioxidant enzyme activity by ectopic overexpression of MnSOD, catalase or glutathione peroxidase, and (2) inhibition of superoxide anion generation by flavincontaining oxidases. These results support a significant role for oxidative metabolism in mediating low-dose radiation effects and demonstrate that cell culture in three dimensions is ideal to investigate radiation-induced adaptive responses. Expression of connexin 43, a constitutive protein of gap junctions, and the G1 checkpoint were more sensitive to regulation by ␥ rays in cells maintained in a three-dimensional than in a two-dimensional configuration. 䉷 2006 by Radiation Research Society

INTRODUCTION

A large number of laboratory and human epidemiological studies have shown that high doses of ionizing radiation engender significant health risks. The mechanisms under1 Address for correspondence: Department of Radiology, UMDNJ— New Jersey Medical School, 185 South Orange Avenue, MSB—F451, Newark, NJ 07101; e-mail: [email protected].

849

850

de TOLEDO ET AL.

tolerance to stress, environmental conditions, etc. Unlike the human population, which is heterogeneous in nature, laboratory animals are often bred to homogeneity, and significant differences in radiosensitivity exist among species (13). In vitro tissue culture studies are another option to investigate the effects of low-level ionizing radiation. Since tissue culture was devised at the beginning of the 20th century (14), these studies have offered a range of cost-effective advantages that greatly facilitated mechanistic investigations. Tissue culture experiments permit the generation of precise and reproducible data in cells free of systemic variations that arise in animals under both normal homeostasis and the stress of an experiment (15). Cell-cell, cellmatrix and cell-soluble factor interactions can be examined in detail, and effects related to specific cell types can be rigorously distinguished. The cellular physio-chemical environment (e.g. temperature, pH, cell cycle stage and cellular physiological conditions) can be tightly controlled. However, tissue culture experiments fail to reflect integrated effects among tissues in complex biological systems. Nevertheless, they complement human epidemiology and animal experiments, and they often replace or reduce the number of live animals used in experiments. Recent advances in developing systems to culture human cells in three-dimensional (3D) architecture under in vivolike growth conditions have increased the suitability of in vitro studies to investigate radiation dose effects relevant to radioprotection of humans. Compared with the in vivo situation, some of the differences in cell behavior due to in vitro culture may be minimized when cells are grown in 3D cultures. Using normal human diploid fibroblasts grown in 3D on scaffolds called Cytomatrix娂, here we extend the investigation of responses to low-level radiation in cells grown in 2D cultures (16, 17) and examine induction of DNA damage, modulation of stress-responsive proteins, clonogenic survival and regulation of cell cycle progression in cells exposed to low dose ␥ rays delivered at various dose rates. We focus on investigating the role of oxidative metabolism in these effects. The study of endogenous oxidative damage compared to low-dose radiation-induced damage has been singled out as a significant research area to elucidate the mechanisms underlying the risk of exposure to low-dose ionizing radiation (18). MATERIALS AND METHODS

Cell Growth in 3D Cultures Cells in passage 10 or 11 were adapted to grow on scaffolds called Cytomatrix娂 (Cell Science Therapeutics, Woburn, MA). This highly porous and biocompatible structure is made entirely of carbon material; its open geometry and organization into an array of continuously interconnected dodecahedrons with no dead space make it ideal for cell growth in three dimensions (19). About 90% of the Cytomatrix娂 consists of pores that are ⬃500 ␮m in diameter, with channels (the conduits connecting adjacent pores) 100–150 ␮m wide. This inert, nondegradable material with a high surface area to volume ratio is not coated with metal; hence complications with radiation dosimetry due to metal-induced absorption and scattering of incident radiation are eliminated. Cytomatrices immersed in PBS were sterilized by autoclaving. After cooling to room temperature, they were incubated overnight at room temperature in centrifuge tubes containing FNC coating solution (AthenaES, Baltimore, MD) consisting of fibronectin and collagen to facilitate cell attachment. Air pockets in the scaffold pores were eliminated by brief centrifugation (500 rpm for 30 s). Cells (2 ⫻ 105) were seeded on FNCcoated matrices placed in the wells (one matrix/well containing 1 ml growth medium) of 24-well plates (Falcon). The growth medium was replaced on a daily basis for 7 days. To alleviate complications in interpretation of results due to differential response of cells at different phases of the growth cycle (20), cells were fed on day 8 with medium containing 0.5% FCS to allow them to accumulate in G0/G1 phase. Irradiation was started 24 h after the last feeding when ⬃95% of the cells were in G0/ G1 as determined by [3H]thymidine uptake. For experiments with cells grown in 2D cultures, about 1.2 ⫻ 105 cells/ dish were seeded in 30-mm polystyrene dishes. The cells were subsequently fed on days 5 and 7 with fresh growth medium. On day 8, they were fed medium containing 0.5% FCS, and experiments were started 24 h later. At that time, 95–98% of the cells were also in G0/G1 phase as determined by [3H]thymidine uptake. Irradiation Cells were exposed to low-dose/low-dose-rate ␥ rays in a custom-designed Cs irradiator containing a 0.666 TBq (18 Ci) source and fitted with a beam shaper and a computer-controlled mercury attenuator system (21). Exposure occurred at 37⬚C in a humidified atmosphere of 5% CO2 in air. The irradiator delivers dose rates ranging from 30 to less than 0.01 cGy/h (21). For acute irradiation, cell cultures were exposed to ␥ rays from a 137Cs source at 6 cGy/min (for low doses) or 3.3 Gy/min (for high doses) (J. L. Shepherd Mark I, San Fernando, CA). Control cells were handled in parallel with cells destined for irradiation but were sham-irradiated. 137

Clonogenic Survival Cell survival was determined by the colony formation assay. Sham-treated and irradiated cultures were trypsinized; cells were diluted in culture medium and seeded in 100-mm dishes at numbers estimated to give about 150 to 200 clonogenic cells per dish. Three to four replicates were done for each exposure point. After an incubation period of 2 weeks, the cells were rinsed with PBS, fixed in ethanol and stained with crystal violet, and macroscopic colonies were counted. Survival values were corrected for the plating efficiency of sham-irradiated cells, and data from representative experiments (repeated two to five times) are shown in the Results.

Cells Apparently normal AG1522 and GM1603 human skin fibroblasts and GM3396 and GM3397 fibroblasts derived from individuals that are heterozygous for the ataxia telangiectasia mutated (ATM) gene were obtained from the Genetic Cell Repository at the Coriell Institute for Medical Research (Camden, NJ). Stock cultures were routinely maintained in 2D architecture at 37⬚C in a humidified incubator in an atmosphere of 5% CO2 in air. Cells were grown in Eagle’s minimal essential medium (CellGro) supplemented with 12.5% heat-inactivated fetal calf serum (FCS) (Hyclone), 100 U/ml penicillin, and 100 ␮g/ml streptomycin (CellGro).

Autoradiographic Measurement of Labeling Indices To characterize the radiation-induced G1 checkpoint, control and irradiated cultures were subcultured in growth medium containing 37 kBq/ ml [3H]thymidine (PerkinElmer LAS, Inc., 37 MBq/ml, specific activity 0.74 TBq/mmol), seeded in 30-mm dishes and incubated at 37⬚C. At regular intervals, duplicate dishes were rinsed with PBS and fixed with ethanol, and Kodak NTB2 nuclear emulsion was applied. After a 3-week exposure at 4⬚C, the dishes were developed with D-19 developer (Kodak), fixed with Kodak GBX fixer, and stained with crystal violet. To determine the labeling indices, a minimum of 1000 cells were scored on each dish.

851

RESPONSES OF HUMAN CELLS TO LOW-DOSE RADIATION

Western Blot Analysis Sham-treated and irradiated cultures were washed in PBS and lysed in chilled RIPA buffer [50 mM Tris-Cl (pH 7.5), 150 mM NaCl, 50 mM NaF, 5 mM EDTA, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS] supplemented with protease inhibitor cocktail (Sigma) and sodium orthovanadate (1 mM). Anti- phospho-TP53 (serine 15) (no. 9284) and anticonnexin 43 (GJA1) (no. 3512) from Cell Signaling, anti-p34cdc2 (CDC2) (no. Sc-54) from Santa Cruz Biotechnology, and anti-␣-tubulin (TUBA) (no. CP06) from EMD Biosciences were used in the analyses. Reaction with anti-␣-tubulin or non-specific (NS) antibody reaction was used to verify whether equal amounts of sample were fractionated. Secondary antibodies conjugated with horseradish peroxidase and the enhanced chemiluminescence system from Amersham Biosciences were used for protein detection. Micronucleus Assay Radiation-induced DNA damage and its repair were assessed by measuring the frequency of micronucleus formation by the cytokinesis-block technique (22). After treatments, cell populations were subcultured, and approximately 3 ⫻ 104 cells were seeded in chamber flaskettes (Nunc) in the presence of 2 ␮g/ml cytochalasin B (Sigma). After 72 h incubation, the cells were rinsed in PBS, fixed in cold methanol, stained with Hoechst 33342 (1 ␮g/ml PBS), and viewed with a fluorescence microscope. At least 1000 cells/experiment were examined, and only micronuclei in binucleated cells were considered for analysis. At the concentration used, cytochalasin B was not toxic to AG1522 cells. Vectors and Cell Transfections Replication-defective recombinant adenovirus type 5 with the E1 region substituted with the human genes encoding MnSOD (AdSOD2), glutathione peroxidase (AdGPX) or catalase (AdCAT) and the control gene E. coli ␤-galactosidase (23) were obtained from ViraQuest (North Liberty, IA). The infectious units of the adenovirus were typically at 1 ⫻ 1010 PFU/ml. At the time of infection, the growth medium was replaced with serum-free fresh medium, adenovirus was added to a multiplicity of infection (MOI) of 100, and cells were incubated for 24 h. They were then fed fresh medium and were used for experiments 24 h later.

FIG. 1. AG1522 normal human fibroblasts easily adapt to grow in a three-dimensional architecture on the Cytomatrix娂 carbon scaffold. Panel A: Architecture of the Cytomatrix娂 carbon scaffold. Panel B: AG1522 cells growing on the periphery of the scaffold. Panel C: AG1522 cells growing within the pores of the scaffold.

tometric method of Lawrence and Burk (28) using cumene hydroperoxide as the substrate. One unit of enzymatic activity, expressed as mU/mg protein, is defined as the amount of protein that oxidizes 1 mmol of NADPH/min. Diphenyleneiodonium (DPI) Treatment The flavoprotein inhibitor DPI (Sigma) was dissolved in DMSO and added to quiescent cultures at a concentration of 0.1 ␮M 24 h prior to irradiation. Sham-treated cultures received the same volume of diluent (0.01%, v/v) and were handled in parallel with the test cultures. Experiments (data not shown) confirmed that DPI inhibits NAD(P)H-oxidase activity in AG1522 cells and results in significant reduction in the level of DCFH-DA (2⬘,7⬘-dichlorodihydrofluorescence diacetate) reactive oxidants. Statistical Analysis Chi-square analysis was used to determine the significance of differences in the magnitude of measured end points before and after irradiation. Experiments were repeated two to five times, and standard errors of the means are indicated on the figures when they are greater than the size of the data point symbols.

Measurements of Antioxidant Enzyme Activity Cell cultures were rinsed with ice-cold PBS and harvested by scraping at 4⬚C. Pelleted cells (⬃107 cells) were rinsed in PBS, repelleted and suspended in 50 mM potassium phosphate buffer containing 1.34 mM diethylenetriaminepenta-acetic acid (pH 7.8), and homogenized at 4⬚C. The activity of SOD in whole cell homogenates was determined as described (24) with modification. Briefly, this assay is a competitive inhibition assay using xanthine-xanthine oxidase-generated superoxide to reduce nitroblue-tetrazolium (NBT) at a constant rate (absorbance of 0.02/ min), which is monitored spectrophotometrically at 560 nm. Increasing amounts of purified SOD protein or cellular homogenates containing SOD activity progressively inhibit the rate of reduction of NBT. The amount of protein that attenuates reduction of NBT to 50% of maximal inhibition is defined as 1 unit of SOD activity (6–9 ng ⫽ 1 unit of activity for purified CuZnSOD from Oxis International; Portland, OR). The assay also uses the inhibition of CuZnSOD by 5 mM cyanide to differentiate between CuZnSOD and MnSOD activities. This concentration of cyanide inhibits the activity associated with 150–250 ng of purified CuZnSOD (24). CuZnSOD activity is determined by subtracting MnSOD activity from total SOD activity. Protein concentration was determined by the method of Lowry (25), and enzymatic activity is expressed in units/mg protein. Catalase activity was determined using the spectrophotometric method of Beers and Sizer (26) and expressed as kU/mg protein as described by Aebi (27). This method measures the exponential disappearance of H2O2 (10 mM) at 240 nm in the presence of cellular homogenates. Total glutathione peroxidase activity was measured by the spectropho-

RESULTS

Normal Human Diploid Fibroblasts Easily Adapt to Grow in 3D Architecture To characterize responses to low-dose ionizing radiation in cells growing in an in vivo-like environment, optimum conditions (described in the Materials and Methods) for the 3D culture of AG1522 normal human diploid fibroblasts were established. The light microscopy data in Fig. 1 show 3D cultures consisting of healthy AG1522 cells growing on biocompatible carbon scaffolds called Cytomatrix娂. Several other cell strains derived from normal and radiosensitive individuals (e.g. GM1603, GM3396, GM3397, etc.; data not shown) also adapted to grow on the Cytomatrix娂, indicating suitability of this 3D culture system to examine the cellular response to radiation. Cells dissociated from the Cytomatrix娂 had a cloning efficiency similar to that of cells derived from monolayer cultures growing in 2D configuration (0.45 ⫾ 0.01 for AG1522 cells cultured on the Cytomatrix娂 compared to 0.48 ⫾ 0.03 for AG1522 cells cultured in 2D architecture).

852

de TOLEDO ET AL.

FIG. 2. Western blot analyses of connexin 43 (cx43) in control and ␥irradiated AG1522 cells cultured in 2D or 3D configuration. Quiescent G0/G1 cells were exposed to 0, 5 or 10 cGy (6 cGy/min) and harvested for analyses 3 h later. Expression of the native, phosphorylated (P) and hyperphosphorylated (PP) forms of connexin 43 were detected by the antibody used.

Sensitivity of the Response to Low-Dose Ionizing Radiation in Cells Cultured in 3D: Modulation of Connexin 43 (GJA1) Expression Homeostatic maintenance of cells in tissues depends on a complex network of communication modalities that allow coordinated interactions among themselves and with their environment. Cells are linked to each other by specialized junctions and respond to factors in the extracellular matrix. Among the many junctions that synchronize the activities of cells in multicellular tissues is the gap junction. This widespread and specialized plasma membrane structure is composed of connexin proteins (29). Expression of connexin 43, a constitutive protein of gap-junction channels found in skin cells, was similar in control cells from 3D or 2D cultures (Fig. 2). Interestingly, its modulation by lowdose radiation was more sensitive in cells maintained in 3D than in 2D cultures. Relative to control cells, a 1.5- and 2-fold increase (by scanning densitometry) in the native and phosphorylated forms of the protein, respectively, occurred in cells acutely exposed to 5 cGy (6 cGy/min) in 3D cultures; no increase was observed in cells maintained in 2D cultures (Fig. 2). At 10 cGy, levels of the three forms

of connexin 43 were increased in cells cultured in either configuration, but the increase was greater in cells cultured in 3D (twofold increase in the native isoform in cells maintained in 3D compared to 1.5-fold increase in cells maintained in 2D) (Fig. 2). While a rise in connexin 43 expression was consistently observed in cells from 3D cultures exposed to low-dose ␥ rays, a significant increase was not always observed in cells from 2D cultures. Increased expression of connexin 43 is associated with greater communication between contiguous cells (30); its up-regulation in irradiated cells cultured in 3D suggests enhanced capacity for intercellular communication. In contrast, phosphorylation of connexin proteins has been shown to both facilitate and restrict gap-junction intercellular communication (31). Radiation-Induced Cell Killing/G1 Delay and Relationship to Growth Architecture We and others have demonstrated the existence of lowdose, low-LET ionizing radiation-induced protective effects (adaptive responses) in cells maintained in 2D (16, 32, 33). In studies designed to extend this investigation to normal human fibroblasts maintained in 3D, we have characterized and compared clonogenic survival and progression through G1 phase of the growth cycle in control and ␥-irradiated AG1522 cells maintained in either configuration. Survival data indicated that the killing effect of ␥ rays is independent of growth architecture (Fig. 3A). Dose–response curves showed that radiosensitivity of quiescent cells maintained in 3D was comparable to that of concurrently exposed cells cultured in 2D. A dose of about 4 Gy reduced survival to 10% (D10) in cultures grown in either configuration. The similar radiosensitivity of cells maintained in 2D or 3D (Fig. 3A) correlated with the similarity of the duration of the radiation-induced G1 delay. Quiescent cell cultures synchronized in G0/G1 and maintained in 2D or 3D config-

FIG. 3. Clonogenic survival and the G1 checkpoint in control and irradiated AG1522 cells cultured in 2D or 3D configuration. Panel A: Quiescent cells were exposed to increasing doses of ␥ rays at a rate of 3.3 Gy/min and assayed for survival immediately after exposure. (䉭) Cells cultured in 2D; (䢇) cells cultured in 3D. Panel B: Representative data from four experiments showing no difference in magnitude of the G1 delay between cells cultured in 2D or 3D configuration. Control and irradiated cultures synchronized in G0/G1 were subcultured and allowed to progress in the cell cycle in growth medium containing 37 kBq/ml [3H]thymidine. The fraction of labeled cells was determined by autoradiography as a function of time after release from the G0/G1 phase. Panel C: Representative data from two experiments showing marked differences in the magnitude of the radiation-induced G1 checkpoint between cells cultured in 2D or 3D. The cumulative labeling indexes were determined as in panel B.


uration were exposed to 0 or 4 Gy. Since cellular interactions with components of the extracellular matrix control the radiation-induced G1 arrest (34), cell cultures grown in 2D were maintained on dishes coated with the same fibronectin/collagen mixture as the Cytomatrix娂 used for 3D growth. Immediately after exposure, control and irradiated cell populations were subcultured and allowed to proliferate in 2D configuration in medium containing [3H]thymidine (37 kBq/ml). Movement into S phase was monitored autoradiographically by measuring the cumulative labeling indices at multiple times up to 60 h after subculture. Control cells from 2D or 3D cultures progressed into S phase at a comparable rate (Fig. 3B). As expected, a significant transient delay in G1 occurred in cells from cultures exposed to 4 Gy; at 50% of the maximum labeling index, radiation caused a delay (⬃5 h) in entry into S phase of cells cultured in either 2D or 3D configuration. The effect of growth architecture on the radiation-induced G1 checkpoint was examined in six independent experiments. The above data on cell cycle progression (Fig. 3B) are representative of four experiments. In the remaining experiments, however, a clear difference in the duration of the radiation-induced G1 checkpoint was seen when cells were cultured in 2D or 3D configuration. Exposure of quiescent cells maintained in 3D to 4 Gy followed by immediate subculture to lower density resulted in a longer G1 delay than irradiated cells maintained in 2D. The delays calculated at 50% of the maximum labeling index were 7 h in 3D cells compared to 5 h in 2D cells (Fig. 3C). The reason for these discrepancies is unclear; it may be related to effects of morphological changes and/or metabolic alterations related to a possible decrease in partial oxygen tension in cells maintained in 3D. Exposure to Low-Dose/Low-Dose-Rate ␥ Rays Decreases the Frequency of DNA Damage to a Level below the Spontaneous Rate We have previously shown that exposure to low-dose/ low-dose-rate ␥ rays reduces the frequency of neoplastic transformation in C3H 10T½ mouse embryo fibroblasts to a level below the spontaneous rate (33). To examine whether similar protective responses exist in normal human fibroblasts maintained in 3D, AG1522 cell cultures grown on Cytomatrix娂 were exposed to 10 cGy delivered at various dose rates. Immediately after exposure, cells were subcultured to lower density and assayed for micronucleus formation. Micronuclei arise predominantly from un-rejoined DNA double-strand breaks (35) and have been strongly implicated in the process of cancer development in humans (36). Relative to control, exposure to 10 cGy delivered acutely or over 2 h resulted in a significant (P ⬍ 0.01) increase in micronucleus formation (Fig. 4A). The magnitude of the increase was reduced when the 10-cGy dose was delivered over 24 h (P ⬍ 0.05). Importantly, when the dose was protracted over 48 h, the micronucleus frequency

853

in exposed cells was often (five out of eight experiments) reduced to a level below the spontaneous rate. The level of significance of the decrease, as calculated by ␹2 analyses, varied among experiments (P ⬍ 0.2 to P ⬍ 0.05). The above data on micronucleus formation (Fig. 4A) correlated with changes in phosphorylation of serine 15 in TP53 (Fig. 4B), a protein that is activated and stabilized in response to a wide range of cellular stresses. Similar to micronucleus frequency, a significant increase (fourfold) in TP53 phosphorylation occurred in cells exposed to a dose of 10 cGy delivered in 2 h. When the dose was protracted over 48 h, phospho-TP53 (serine 15) level was 50% lower than in control cells (Fig. 4B). These data are consistent with expression of p34cdc2 (CDC2), a growth-related protein that is regulated by ionizing radiation in a TP53-dependent mechanism (37, 38). As expected, decrease in phosho-TP53 correlated with increased (threefold) p34cdc2 expression in cells exposed to a 10-cGy dose protracted over 48 h. These data (Fig. 4B) indicate that TP53 signaling is highly sensitive to low-dose/low-dose-rate exposure and that phospho-TP53 (serine 15) is a suitable marker to detect DNA damage soon after such exposure, being modulated by doses as low as 1 cGy (5 cGy/h). Oxidative Metabolism Modulates the Cellular Response to Low-Dose Ionizing Radiation To investigate the contribution of metabolic oxidation/ reduction (redox) reactions to induction of DNA damage in cells exposed to low-dose radiation delivered at a high dose rate (Fig. 4A), we measured micronucleus formation in control and ␥-irradiated (10 cGy, 6 cGy/min) AG1522 cells in which antioxidant enzyme activity had been up-regulated by ectopic overexpression of either MnSOD, glutathione peroxidase (GPX) or catalase (CAT). Transduction of AG1522 cell cultures with adenovirus constructs for either of the above enzymes resulted in increased activity in MnSOD (threefold), GPX (threefold) and catalase (eightfold), respectively (Table 1). Similar to control nontransduced cells (Fig. 4A), a significant increase (P ⬍ 0.001) in micronucleus formation occurred in irradiated AG1522 fibroblasts transduced with an empty adenovirus vector (Fig. 5A). In contrast, when cells with increased activity of MnSOD, GPX or CAT were exposed to the same dose and dose rate (10 cGy, 6 cGy/min), the micronucleus frequency was significantly attenuated and was similar to that in the respective nonirradiated control cells. The slight differences in the extent of attenuation by the different antioxidants may be related to biochemical/signaling events, which result from overexpression of each specific enzyme. Together, these data indicate that up-regulation of antioxidants protects against chromosomal damage by small acute doses of low-LET ionizing radiation. They suggest that the decrease in micronucleus formation to levels similar to or lower than the spontaneous rate in cells exposed to low-dose radiation delivered at a very low dose rate (Fig. 4A) may be due to

854

de TOLEDO ET AL.

FIG. 4. DNA damage and modulation of TP53Ser15 and p34cdc2 after exposure of AG1522 cells maintained in 3D to low-dose ␥ rays delivered over four different periods. Panel A: Frequency of micronucleus formation in cells assayed immediately after exposure to 10 cGy. Panel B: Expression of TP53Ser15 and p34cdc2 in cells harvested for analyses immediately after ␥-ray exposure of 0, 1, 5 or 10 cGy delivered at three different dose rates.

increased antioxidant activity. Our data confirm this concept by indicating a slight but consistent increase in glutathione levels to 1.4 ⫾ 0.2 times the control level in cells exposed to 10 cGy protracted over 48 h. Involvement of oxidative metabolism in low-dose radiation effects is further supported by data in Fig. 5B. Incubation of cells in 0.1 ␮M DPI, an inhibitor of superoxide generating NAD(P)H-oxidase and other flavin-containing enzymes, significantly attenuated micronucleus formation in cells exposed to 10 cGy delivered acutely (6 cGy/min). DISCUSSION

Recent advances in cellular and molecular biology are providing sensitive assays to characterize effects induced in cells exposed to low doses/low fluences of ionizing radiation (3, 39, 40). Elucidation of the molecular/biochemical events associated with these effects has been considered of particular significance to our understanding of lowdose radiation risk (16). This study further contributes to these advances and characterizes a novel 3D tissue culture system, which mimics aspects of in vivo growth, to investigate radiation effects. A variety of skin fibroblasts derived from apparently normal individuals (Fig. 1) or carriers of ATM (not shown), a gene that predisposes to radiation senTABLE 1 Antioxidant Levels in AG1522 Cells Transduced with Replication-Defective Recombinant Adenovirus Type 5 with the E1 Region Substituted with the Human Genes Encoding MnSOD, Glutathione Peroxidase or Catalase SOD (U/mg)

Control Ad-empty Ad-catalase Ad-GPX Ad-MnSOD

Total

MnSOD

41.7 41.7 55.6 26.3 71.4

16.7 20.8 45.5 23.3 55.6

GPX Catalase CuZnSOD (mU/mg) (m␬U/mg) 25 20.9 10.1 3.0 15.8

15.6 19.1 25.8 46.6 14.3

5.1 6.6 43.8 5.8 9.2

sitivity, easily adapted to grow in 3D on a carbon skeleton that provided them with an in vivo-like microenvironment. Cloning efficiencies of control and irradiated cells (Fig. 3A) originating from cultures maintained in 3D or 2D were similar. In contrast, expression of connexin 43 (GJA1), a constitutive protein of junction channels that link contiguous cells, was more sensitive to regulation by low-dose ␥ rays in normal human cells maintained in 3D rather than 2D architecture (Fig. 2). These differences may result from physical forces related to growth architecture and may also be due to metabolic alterations, perhaps due to variation in partial oxygen tension and/or other factors. Normal human fibroblasts cultured in 3D (Fig. 4B) or 2D (not shown) revealed that phosphorylation of serine 15 in TP53 is a highly sensitive marker to monitor stressful effects induced by low-dose ␥ rays. Up-regulation of phospho-TP53 (ser15) was detected within 10 min after exposure of AG1522 cells to an acute dose of 1 cGy (Fig. 4B). As expected, protracting delivery of the radiation dose (1 to 10 cGy) attenuated this increase. Modulation of TP53 mirrored changes in micronucleus frequency (Fig. 4A). While a significant (P ⬍ 0.01) increase in micronuclei occurred in cells exposed to an acute dose of 10 cGy, protraction of the same dose over 24 h attenuated this increase (Fig. 4A). Strikingly, when delivery of 10 cGy was further protracted over 48 h, micronucleus frequency and phosphoTP53 ser15 levels in irradiated cells were often lower than or equal to the spontaneous levels (Fig. 4A and B). These effects in human cells are consistent with data in mouse embryo fibroblasts and HeLa/normal human hybrid cells exposed to low-dose radiation, in which the frequency of neoplastic transformation decreased to a level below the background rate (33, 41). Currently, for the purposes of radiation protection, the deleterious effects of ionizing radiation are assumed to have a linear dose response with no threshold. Furthermore, it is assumed that protracted (chronic) exposures require about twice the dose to cause the same effect as an acute exposure (42). The effects of sequential doses are assumed to be


855

FIG. 5. Oxidative metabolism modulates the cellular response to low-dose ionizing radiation. Panel A: Frequency of micronucleus formation in AG1522 normal human diploid fibroblasts where the activity of the antioxidant enzymes MnSOD, GPX or catalase was regulated prior to exposure to a single 10-cGy ␥-ray dose (6 cGy/min). Panel B: AG1522 cells were incubated in the presence or absence of 0.1 ␮M DPI for 24 h, exposed to 10 cGy (6 cGy/min), and assayed for micronucleus formation.

additive. One consequence of this hypothesis is the assumption that exposure to any dose of radiation, however small, increases the risk of detrimental health effects. Our data on in vitro low-dose-rate exposures (Fig. 4) and those of others (17, 43) show that exposure to low-dose/low-LET radiation may not universally damage exposed cells. In contrast, they suggest that such exposure up-regulates protective mechanisms, which attenuate damage caused by normal endogenous metabolism and/or the damage caused by the radiation delivered at low dose rates or by subsequent challenge high-dose-rate exposures. Changes in chromatin conformation (44), apoptosis (45), intercellular communication (46, 47), up-regulation of DNA repair, and/or antioxidation (48–51) may mediate such adaptive processes. Our experiments indicate that up-regulation of antioxidant enzyme activity (Table 1) by ectopic expression of either MnSOD, catalase or glutathione peroxidase attenuates DNA damage induced by an acute dose of 10 cGy to a level similar to background (Fig. 5A). Comparable effects occurred when superoxide production by flavin-containing oxidases [presumably NAD(P)H-oxidase] was inhibited (Fig. 5B). These latter data strongly support the role of oxidative metabolism in mediating low-dose effects (52– 54). Protection against radiation-induced DNA damage by either up-regulating antioxidant activity (Fig. 5A) or inhibiting superoxide production by flavin oxidases (Fig. 5B) suggests a critical role for flavin-containing oxidases in the cellular response to stress. One of these oxidases, NAD(P)H-oxidase, was shown in vascular cells to be activated by reactive oxygen species to induce oxidant production and cellular injury (55). It is attractive to speculate that excess residual pro-oxidants, produced soon after radiation exposure, activate NAD(P)H-oxidase to produce more oxidants than required for normal physiological functions, thus prolonging the period of cellular oxidative stress. A disruption of the balance between oxidant production and antioxidant capacity alters the cellular redox environment, resulting in important biological consequences to the cell

(54). Genetic alterations in the progeny of irradiated cells have been associated with oxidative stress (56, 57), and inhibition of flavin oxidases attenuated the transmission of stressful effects from irradiated to nonirradiated cells (58, 59). An important finding of our studies is the increase in glutathione content in AG1522 cells exposed to a 10-cGy dose protracted over 48 h. These findings are consistent with those of Kojima et al., who particularly found that elevation of intracellular glutathione by low-dose ␥ rays (60) is associated with increase in ␥-glutamlcysteine synthetase mRNA by 3 to 6 h after exposure (61). Glutathione is a fundamental antioxidant that protects cells against oxidant-induced injury (62). Furthermore, it modulates redoxregulated signal transduction and has a role in cellular proliferation (63). Its increase in normal human cells exposed to low-dose/low-dose-rate ␥ rays may be a factor in reducing DNA damage in these cells to levels below the spontaneous rate (Fig. 4A). Such an effect would be in contrast to observations in HeLa ⫻ human hybrid cells in which suppression of glutathione synthesis by incubation with buthionine sulfoximine had no effect on suppression of neoplastic transformation by low-dose ␥ rays (64). The environmental milieu in which cells are maintained has been shown to significantly affect experimental outcome (34, 65). In addition to enhanced sensitivity in regulation of connexin 43 by low-dose ␥ rays in cells cultured in 3D, our studies have shown that the magnitude of the radiation-induced G1 checkpoint may also depend on growth architecture (Fig. 3C). Greater delays in G1 often occurred when cells were exposed to ␥ rays while maintained in 3D. The molecular and biochemical basis for this effect has not been identified. The radiation-induced G1 checkpoint is thought to be a regulator of cellular sensitivity (66); it presumably ensures that DNA damage is repaired prior to DNA replication (67, 68). We have previously found that greater cell cycle delays contribute to expression of radiation-induced adaptive responses in cells grown in

856

de TOLEDO ET AL.

2D (69). Thus the occurrence of yet more prominent growth delays after in vivo-like exposure to ionizing radiation may result in adaptive responses that are greater in magnitude. Collectively, these studies highlight the usefulness of 3D growth systems to investigate the cellular response to ionizing radiation; in conjunction with in vivo studies (70, 71), they should increase our understanding of mechanisms underlying low-dose effects. In vitro, the effects of low-dose/ low-LET ionizing radiation are often subtle; however, overwhelming evidence indicates that they cannot be predicted from effects observed at high doses (72, 73). Importantly, this study provides further evidence that the rate at which ␥-ray energy is deposited in the cell is critical in triggering the system responses that determine the outcome of low dose exposure.

14. 15. 16.

17.

18.

19.

ACKNOWLEDGMENTS 20. We thank Drs. Debkumar Pain, Badri Pandey and Donna Gordon for helpful discussions. Research grants FG02-02ER63447 (EA and DRS) and FG02-05ER64050 (DRS and EA) from the U.S. Department of Energy supported this investigation.

21.

Received: March 25, 2006; accepted: August 2, 2006 22.

REFERENCES 1. J. B. Little, Radiation carcinogenesis. Carcinogenesis 21, 394–404 (2000). 2. NCRP, Evaluation of the Linear-Nonthreshold Dose-Response Model for Ionizing Radiation. Report No. 136, National Council on Radiation Protection and Measurements, Bethesda, MD, 2001. 3. W. M. Bonner, Low-dose radiation: Thresholds, bystander effects, and adaptive responses. Proc. Natl. Acad. Sci. USA 100, 4973–4975 (2003). 4. R. E. Mitchel, J. S. Jackson, R. A. McCann and D. R. Boreham, The adaptive response modifies latency for radiation-induced myeloid leukemia in CBA/H mice. Radiat. Res. 152, 273–279 (1999). 5. J. L. Redpath, Radiation-induced neoplastic transformation in vitro: Evidence for a protective effect at low doses of low LET radiation. Cancer Metastasis Rev. 23, 333–339 (2004). 6. D. L. Preston, D. A. Pierce, Y. Shimizu, E. Ron and K. Mabuchi, Dose response and temporal patterns of radiation-associated solid cancer risks. Health Phys. 85, 43–46 (2003). 7. R. J. Preston, The LNT model is the best we can do—today. J. Radiol. Prot. 23, 263–268 (2003). 8. D. J. Brenner, R. Doll, D. T. Goodhead, E. J. Hall, C. E. Land, J. B. Little, J. H. Lubin, D. L. Preston, R. J. Preston and M. Zaider, Cancer risks attributable to low doses of ionizing radiation: Assessing what we really know. Proc. Natl. Acad. Sci. USA 100, 13761–13766 (2003). 9. E. J. Hall, Radiation and Life, 2nd ed. Pergamon Press, New York, 1984. 10. S. R. Baker, Musings at the beginning of the hyper-CT era. Abdominal Imaging 28, 110–114 (2003). 11. E. I. Azzam and J. B. Little, The radiation-induced bystander effect: Evidence and significance. Hum. Exp. Toxicol. 23, 61–65 (2004). 12. J. P. Messina and D. A. Lawrence, Effects of 2-mercaptoethanol and buthionine sulfoximine on cystine metabolism by and proliferation of mitogen-stimulated human and mouse lymphocytes. Int. J. Immunopharmacol. 14, 1221–1234 (1992). 13. B. Ponnaiya, M. N. Cornforth and R. L. Ullrich, Radiation-induced

23.

24.

25.

26.

27. 28.

29. 30.

31.

32.

33.

34.

chromosomal instability in BALB/c and C57BL/6 mice: The difference is as clear as black and white. Radiat. Res. 147, 121–125 (1997). A. Carrel, On the permanent life of tissues outside the organism. J. Exp. Med. 15, 516–528 (1912). R. I. Freshney, Culture of Animal Cells, 3rd ed. Wiley-Liss, New York, 1983. National Research Council, Committee on the Biological Effects of Ionizing Radiation, Health Risks from Exposure to Low Levels of Ionizing Radiation (BEIR VII). National Academy Press, Washington, DC, 2005. S. Wolff, The adaptive response in radiobiology: Evolving insights and implications. Environ. Health Perspect. 106 (Suppl. 1), 277–283 (1998). L. E. Feinendegen, V. P. Bond, C. A. Sondhaus and K. I. Altman, Cellular signal adaptation with damage control at low doses versus the predominance of DNA damage at high doses. C. R. Acad. Sci. III 322, 245–251 (1999). M. C. Poznansky, R. H. Evans, R. B. Foxall, I. T. Olszak, A. H. Piascik, K. E. Hartman, C. Brander, T. H. Meyer, M. J. Pykett and D. T. Scadden, Efficient generation of human T cells from a tissueengineered thymic organoid. Nat. Biotechnol. 18, 729–734 (2000). R. Terasima and L. J. Tolmach, X-ray sensitivity and DNA synthesis in synchronous populations of HeLa cells. Science 140, 490–492 (1963). R. W. Howell, S. M. Goddu and D. V. Rao, Design and performance characteristics of an experimental Cs-137 irradiator to simulate internal radionuclide dose rate patterns. J. Nucl. Med. 38, 727–731 (1997). M. Fenech and A. A. Morley, Measurement of micronuclei in lymphocytes. Mutat. Res. 147, 29–36 (1985). E. W. Lam, R. Zwacka, E. A. Seftor, D. R. Nieva, B. L. Davidson, J. F. Engelhardt, M. J. Hendrix and L. W. Oberley, Effects of antioxidant enzyme overexpression on the invasive phenotype of hamster cheek pouch carcinoma cells. Free Radic. Biol. Med. 27, 572–579 (1999). D. R. Spitz and L. W. Oberley, An assay for superoxide dismutase activity in mammalian tissue homogenates. Anal. Biochem. 179, 8– 18 (1989). O. H. Lowry, N. J. Rosebrough, A. L. Farr and R. J. Randall, Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265– 275 (1951). R. F. Beers and I. W. Sizer, A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. J. Biol. Chem. 195, 133–140 (1952). H. Aebi, Catalase in vitro. Methods Enzymol. 105, 121–126 (1984). R. A. Lawrence and R. F. Burk, Glutathione peroxidase activity in selenium-deficient rat liver. Biochem. Biophys. Res. Commun. 71, 952–958 (1976). A. L. Harris, Emerging issues of connexin channels: Biophysics fills the gap. Q. Rev. Biophys. 34, 325–472 (2001). B. J. Darrow, V. G. Fast, A. G. Kleber, E. C. Beyer and J. E. Saffitz, Functional and structural assessment of intercellular communication. Increased conduction velocity and enhanced connexin expression in dibutyryl cAMP-treated cultured cardiac myocytes. Circ. Res. 79, 174–183 (1996). J. C. Saez, A. D. Martinez, M. C. Branes and H. E. Gonzalez, Regulation of gap junctions by protein phosphorylation. Braz. J. Med. Biol. Res. 31, 593–600 (1998). E. I. Azzam, S. M. de Toledo, G. P. Raaphorst and R. E. Mitchel, Radiation-induced radioresistance in a normal human skin fibroblast cell line. In Low Dose Irradiation and Biological Defense Mechanisms (T. Sugahara, L. A. Sagan and T. Aoyama, Eds.), pp. 291–294. Elsevier, Amsterdam, 1992. E. I. Azzam, S. M. de Toledo, G. P. Raaphorst and R. E. Mitchel, Low-dose ionizing radiation decreases the frequency of neoplastic transformation to a level below the spontaneous rate in C3H 10T1/2 cells. Radiat. Res. 146, 369–373 (1996). D. M. Gadbois, E. M. Bradbury and B. E. Lehnert, Control of ra-


35.

36. 37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52. 53.

diation-induced G1 arrest by cell-substratum interactions. Cancer Res. 57, 1151–1156 (1997). J. A. Heddle and A. V. Carrano, The DNA content of micronuclei induced in mouse bone marrow by gamma-irradiation: Evidence that micronuclei arise from acentric chromosomal fragments. Mutat. Res. 44, 63–69 (1977). P. Karran, DNA double strand break repair in mammalian cells. Curr. Opin. Genet. Dev. 10, 144–150 (2000). E. I. Azzam, S. M. de Toledo, M. J. Pykett, H. Nagasawa and J. B. Little, CDC2 is down-regulated by ionizing radiation in a p53-dependent manner. Cell Growth Differ. 8, 1161–1169 (1997). S. M. de Toledo, E. I. Azzam, P. Keng, S. Laffrenier and J. B. Little, Regulation by ionizing radiation of CDC2, Cyclin A, Cyclin B, Thymidine Kinase, Topoisomerase II␣, and RAD51 expression in normal human diploid fibroblasts is dependent on p53/p21Waf1. Cell Growth Differ. 9, 887–896 (1998). K. Rothkamm and M. Lobrich, Evidence for a lack of DNA doublestrand break repair in human cells exposed to very low X-ray doses. Proc. Natl. Acad. Sci. USA 100, 5057–5062 (2003). C. N. Coleman, W. F. Blakely, J. R. Fike, T. J. MacVittie, N. F. Metting, J. B. Mitchell, J. E. Moulder, R. J. Preston, T. M. Seed and R. S. Wong, Molecular and cellular biology of moderate-dose (1–10 Gy) radiation and potential mechanisms of radiation protection: Report of a workshop at Bethesda, Maryland, December 17–18, 2001. Radiat. Res. 159, 812–834 (2003). J. L. Redpath and R. J. Antoniono, Induction of an adaptive response against spontaneous neoplastic transformation in vitro by low-dose gamma radiation. Radiat. Res. 149, 517–520 (1998). United Nations Scientific Committee on the Effects of Atomic Radiation, Sources, Effects, and Risks of Ionizing Radiation. 1988 Report to the General Assembly, with Annexes. United Nations, New York, 1988. J. L. Redpath, Radiation-induced neoplastic transformation in vitro: Evidence for a protective effect at low doses of low LET radiation. Cancer Metastasis Rev. 23, 333–339 (2004). I. Belyaev, I. M. Spivak, A. Kolman and M. Harms-Ringdahl, Relationship between radiation induced adaptive response in human fibroblasts and changes in chromatin conformation. Mutat. Res. 358, 223–230 (1996). B. R. Scott, A biological-based model that links genomic instability, bystander effects, and adaptive response. Mutat. Res. 568, 129–143 (2004). S. J. Ko, X. Y. Liao, S. Molloi, E. Elmore and J. L. Redpath, Neoplastic transformation in vitro after exposure to low doses of mammographic-energy X rays: Quantitative and mechanistic aspects. Radiat. Res. 162, 646–654 (2004). K. Ishii and M. Watanabe, Participation of gap-junctional cell communication on the adaptive response in human cells induced by low dose of X-rays. Int. J. Radiat. Biol. 69, 291–299 (1996). E. I. Azzam, S. M. de Toledo, G. P. Raaphorst and R. E. Mitchel, Re´ponse adaptative au rayonnement ionisant des fibroblastes de peau humaine. Augmentation de la vitesse de re´paration de l’ADN et variation de l’expression des ge`nes. J. Chim. Phys. 91, 931–936 (1994). O. Rigaud and E. Moustacchi, Radioadaptation for gene mutation and the possible molecular mechanisms of the adaptive response. Mutat. Res. 358, 127–134 (1996). A. Bravard, C. Luccioni, E. Moustacchi and O. Rigaud, Contribution of antioxidant enzymes to the adaptive response to ionizing radiation of human lymphoblasts. Int. J. Radiat. Biol. 75, 639–645 (1999). G. Guo, Y. Yan-Sanders, B. D. Lyn-Cook, T. Wang, D. Tamae, J. Ogi, A. Khaletskiy, Z. Li, C. Weydert, J. A. Longmate and J. J. Li, Manganese superoxide dismutase-mediated gene expression in radiation-induced adaptive responses. Mol. Cell. Biol. 23, 2362–2378 (2003). L. E. Feinendegen, Reactive oxygen species in cell responses to toxic agents. Hum. Exp. Toxicol. 21, 85–90 (2002). A. Petkau, Role of superoxide dismutase in modification of radiation injury. Br. J. Cancer (Suppl.) 8, 87–95 (1987).

857

54. D. R. Spitz, E. I. Azzam, J. J. Li and D. Gius, Metabolic oxidation/ reduction reactions and cellular responses to ionizing radiation: A unifying concept in stress response biology. Cancer Metastasis Rev. 23, 311–322 (2004). 55. W. G. Li, F. J. Miller, Jr., H. J. Zhang, D. R. Spitz, L. W. Oberley and N. L. Weintraub, H2O2-induced O2-production by a non-phagocytic NAD(P)H oxidase causes oxidant injury. J. Biol. Chem. 276, 29251–29256 (2001). 56. S. M. Clutton, K. M. S. Townsend, C. Walker, J. D. Ansell and E. G. Wright, Radiation-induced genomic instability and persisting oxidative stress in primary bone marrow cultures. Carcinogenesis 17, 1633–1639 (1996). 57. A. R. Snyder and W. F. Morgan, Differential induction and activation of NF-␬B transcription complexes in radiation-induced chromosomally unstable cell lines. Environ. Mol. Mutagen. 45, 177–187 (2005). 58. E. I. Azzam, S. M. de Toledo, D. R. Spitz and J. B. Little, Oxidative metabolism modulates signal transduction and micronucleus formation in bystander cells from ␣-particle-irradiated normal human fibroblast cultures. Cancer Res. 62, 5436–5442 (2002). 59. P. K. Narayanan, E. H. Goodwin and B. E. Lehnert, ␣ particles initiate biological production of superoxide anions and hydrogen peroxide in human cells. Cancer Res. 57, 3963–3971 (1997). 60. S. Kojima, H. Ishida, M. Takahashi and K. Yamaoka, Elevation of glutathione induced by low-dose gamma rays and its involvement in increased natural killer activity. Radiat. Res. 157, 275–280 (2002). 61. K. Teshima, A. Yamamoto, K. Yamaoka, Y. Honda, S. Honda, T. Sasaki and S. Kojima, Involvement of calcium ion in elevation of mRNA for ␥-glutamylcysteine synthetase (␥-GCS) induced by lowdose gamma-rays. Int. J. Radiat. Biol. 76, 1631–1639 (2000). 62. J. E. Biaglow, M. E. Varnes, E. R. Epp, E. P. Clark, S. W. Tuttle and K. D. Held, Role of glutathione and other thiols in cellular response to radiation and drugs. Drug Metab. Rev. 20, 1–12 (1989). 63. R. H. Burdon, Control of cell proliferation by reactive oxygen species. Biochem. Soc. Trans. 24, 1028–1032 (1996). 64. M. C. Pant, X. Y. Liao, Q. Lu, S. Molloi, E. Elmore and J. L. Redpath, Mechanisms of suppression of neoplastic transformation in vitro by low doses of low LET radiation. Carcinogenesis 24, 1961– 1965 (2003). 65. T. Waldman, Y. Zhang, L. Dillehay, J. Yu, K. Kinzler, B. Vogelstein and J. Williams, Cell cycle arrest versus cell death in cancer therapy. Nat. Med. 3, 1034–1036 (1997). 66. L. Hartwell, T. Weinert, L. Kadyk and B. Garvik, Cell cycle checkpoints, genomic integrity, and cancer. Cold Spring Harb. Symp. Quant. Biol. 59, 259–263 (1994). 67. S. J. Elledge, Cell cycle checkpoints: Preventing an identity crisis. Science 274, 1664–1672 (1996). 68. J. B. Little, Changing views of cellular radiosensitivity. Radiat. Res. 140, 299–311 (1994). 69. R. E. Mitchel, E. I. Azzam and S. M. de Toledo, Adaptation to ionizing radiation in mammalian cells. In Stress-Inducible Processes in Higher Eukaryotic Cells (T. M. Koval, Ed.), pp. 221–243. Plenum Press, New York, 1997. 70. E. Cardis, M. Vrijheid, M. Blettner, E. Gilbert, M. Hakama, C. Hill, G. Howe, J. Kaldor, C. R. Muirhead and K. Veress, Risk of cancer after low doses of ionising radiation: Retrospective cohort study in 15 countries. Br. Med. J. 331, 77 (2005). 71. L. Y. Krestinina, D. L. Preston, E. V. Ostroumova, M. O. Degteva, E. Ron, O. V. Vyushkova, N. V. Startsev, M. M. Kossenko and A. V. Akleyev, Protracted radiation exposure and cancer mortality in the Techa River Cohort. Radiat. Res. 164, 602–611 (2005). 72. A. M. Hooker, M. Bhat, T. K. Day, J. M. Lane, S. J. Swinburne, A. A. Morley and P. J. Sykes, The linear no-threshold model does not hold for low-dose ionizing radiation. Radiat. Res. 162, 447–452 (2004). 73. B. N. Pandey, D. A. Gordon, D. Pain and E. I. Azzam, Normal human fibroblasts exposed to high or low dose ionizing radiation: Differential effects on mitochondrial protein import and membrane potential. Antioxid. Redox Signal. 8, 1253–1261 (2006).