Chem. Res. Toxicol. 2000, 13, 541-549
541
DNA Damage Induced in Cells by γ and UVA Radiation As Measured by HPLC/GC-MS and HPLC-EC and Comet Assay J.-P. Pouget,† T. Douki,† M.-J. Richard,‡ and J. Cadet*,† De´ partement de Recherche Fondamentale sur la Matie` re Condense´ e, SCIB/Laboratoire “Le´ sions des Acides Nucle´ iques”, CEA/Grenoble, F-38054 Grenoble Cedex 9, France, and LBSO/LCR7 n°8M, Faculte´ de Pharmacie, Universite´ Joseph Fourier, F-38700 La Tronche, France Received February 7, 2000
The aim of the work was to measure DNA damage induced within tumoral human monocytes by γ rays, UVA radiation, and exogenous photosensitizers. The accurate HPLC-EC assay was used to determine the level of 8-oxodGuo. The formation of FapyGua and FapyAde was monitored by HPLC/GC-MS analyses after formic acid hydrolysis at room temperature. For this purpose, cells were exposed to relatively high doses of γ rays and UVA radiation. The extent of formation of FapyGua in the DNA of cells exposed to γ rays was estimated to be more than 2-fold higher than that of 8-oxodGuo, i.e., about 0.027 lesion per 106 bases per Gy. The yield of FapyAde was estimated to be 1 order of magnitude lower. The latter results were used to calibrate the alkaline comet assay associated with DNA N-glycosylases. The latter approach allowed the determination of the background level (0.11-0.16 Fpg-sensitive site/106 bases) and the yields of strand breaks and DNA base damage upon low irradiation doses. Insights into the mechanism of radiation-induced DNA damage were gained from these measurements. A major involvement of 1O2 with respect to hydroxyl radicals and type I photosensitization was thus observed within cells exposed to UVA radiation.
Introduction Ionizing and solar radiations are known to be involved in mutagenic and carcinogenic processes (1, 2). The way by which the electromagnetic radiations interact with DNA depends on the energy of the incident photon. Around 70% of the interaction of γ rays with DNA proceed through radiolysis of water (3). This leads to the formation of reactive oxygen species such as •OH, which can further react with DNA. The 30% remaining consist of a direct interaction of the ionizing radiation with DNA. Hydroxyl radical is the principal damaging species involved in the formation of DNA strand breaks, modified bases, abasic sites, sugar alterations, and DNA-protein cross-links (4, 5). Among the spectrum of base damage, 8-oxo-7,8-dihydroguanine (8-oxoGua)1 is of main relevance (6). The latter lesion is also expected to be the major degradation product upon exposure of cellular DNA to UVA (320-400 nm) and visible (400-800 nm) radia* To whom correspondence should be addressed: Laboratoire “Le´sions des Acides Nucle´iques”, DRFMC/SCIB, UMR-CNRS 5046, CEA/ Grenoble, F-38054 Grenoble Cedex 9, France. Phone: (33)-4-76-8849-87. Fax: (33)-4-76-88-50-90. E-mail:
[email protected]. † CEA/Grenoble. ‡ Universite ´ Joseph Fourier. 1 Abbreviations: 8-oxoGua, 8-oxo-7,8-dihydroguanine; 8-oxodGuo, 8-oxo-7,8-dihydro-2′-deoxyguanosine; dGuo, 2′-deoxyguanosine; dGMP, 2′-deoxyguanosine 5′-monophosphate; FapyGua, 2,6-diamino-4-hydroxy-5-formamidopyrimidine; FapyAde, 4,6-diamino-5-formamidopyrimidine; Fpg, E. coli formamidopyrimidine DNA N-glycosylase; endo III, E. coli endonuclease III; GC-MS, gas chromatography-mass spectrometry; HPLC-EC, high-performance liquid chromatographyelectrochemical detection; HPLC-MS/MS, high-performance liquid chromatography-tandem mass spectrometry; DSB, double strand break; SSB, single strand break; ALS, alkali-labile site; FCS, fetal calf serum; PBS, phosphate-buffered saline; BSA, bovine serum albumin; OD, optical density; AO, acridine orange; RB, rose bengal; BSTFA, N-bis(trimethylsilyl)trifluoroacetamide.
tions (7-9). The molecular effects of UVA radiation can be explained by the presence within the cells of photosensitizers which absorb energy. Among them, porphyrins and flavins absorb between 334 and 434 nm (10). Excited photosensitizers can then directly react with DNA by electron abstraction (type I reaction) and/or via the production of singlet oxygen (type II) (11). Type I reaction can give rise to superoxide radical anion (O2•-) as a single product of the reaction of the radical anion of the photosensitizer with oxygen. The rather unreactive O2•- species can be converted into the hydroxyl radical • OH via a Fenton reaction. Singlet oxygen 1O2 (type II reaction) is involved in the formation of 8-oxoGua (for a comprehensive review, see ref 12). Therefore, the relative proportion of strand breaks versus 8-oxoGua can be used as an indicator of the contribution of both type I and type II photosensitizing effects. The main aim of this study was to measure DNA damage induced in cultured human cells by γ rays, UVA radiation, or exogenous photosensitizers such as rose bengal or acridine orange. Exogenous photosensitizers were chosen for their ability to mostly produce 1O2 upon excitation by visible light. Direct measurement of 8-oxoGua, FapyGua, and FapyAde was performed in the DNA of a human cultured THP-1 monocyte cell line using HPLC-EC and HPLC/ GC-MS methods. Much attention was devoted to the prevention of drawbacks associated with the application of the latter assays as outlined in recent publications (13-15). Indeed, the large variability in the levels of 8-oxoGua within cellular DNA has been reported in the past decade (16). The measurement of 8-oxoGua is important because this lesion which is known to be
10.1021/tx000020e CCC: $19.00 © 2000 American Chemical Society Published on Web 06/06/2000
542
Chem. Res. Toxicol., Vol. 13, No. 7, 2000
mutagenic is commonly used as a marker of oxidative stress (17-20). Measurement of Fapy derivatives by GCMS was achieved using an optimized assay involving a quantitative formic acid hydrolysis at room temperature (21, 22). Then, the levels of modified purines were compared with data obtained using a modified version of the comet assay. Under its standard version, the latter method allows the measurement of single and double strand breaks (SSB and DSB, respectively) together with alkali-labile sites (ALS). When associated with DNA repair enzymes such as Escherichia coli formamidopyrimidine DNA N-glycosylase (Fpg) and E. coli endonuclease III (endo III), single-cell gel electrophoresis analysis provides insights into the measurement of several modified purines and pyrimidines, respectively (23-26). Modified bases are thus converted into strand breaks which can be further assessed by the comet assay. The DNA repair enzyme Fpg has also been associated with other techniques, including nick translation (27) and alkaline elution (28-30). 8-OxoGua together with 2,6-diamino-4hydroxy-5-formamidopyrimidine (FapyGua) and 4,6-diamino-5-formamidopyrimidine (FapyAde) were detected using the Fpg protein (31-33). Endo III was used in a similar way to assess the level of degradation products of pyrimidine bases, which may include, among others, 5-hydroxy-5-methylhydantoin and thymine glycols (34). The measurement of 8-oxoGua, FapyGua, and FapyAde by chromatographic methods allowed the calibration of the modified comet assay for its subsequent application to the determination of the level of modified purine bases. It was thus possible to determine the steady-state level together with the formation rate of the targeted DNA lesions. In addition, attempts were made to assess mechanisms involved in the DNA damaging properties of the selected oxidizing agents.
Materials and Methods Chemicals. Agarose and low-melting point agarose were obtained from Promega (Madison, WI). Tris and Na2EDTA were purchased from Interbiotech (Interchim, Montluc¸ on, France). Nuclease P1, acid phosphatase, RNase A, RNase T1, DMSO, Triton X-100, bovine serum albumin (BSA), sodium sarcosinate, MgCl2, NaI, ethidium bromide, deferoxamine mesylate, and SDS were from Sigma Co. (St. Louis, MO). Proteinase was obtained from Qiagen Genomic (Hilden, Germany), whereas RPMI-1640, fetal calf serum (FCS), and Dulbecco’s PBS were from Life Technologies (Paisley, U.K.). Slides and cover glasses were from Touzart & Matignon (Vitry-sur-Seine, France) and Marienfeld, respectively. Fpg and endo III E. coli proteins were kindly provided by S. Boiteux (CEA, Fontenay-aux-Roses, France). Nap 25 columns were purchased from Pharmacia Biotech (Uppsala, Sweden). FapyGua and [2,6-diamino,5-formamido-15N3]FapyGua ([15N3]FapyGua) were synthesized as described in ref 35. [4,6diamino-15N2,5-formamido-13C]FapyAde ([15N2,13C]FapyAde) was a gift from H. E. Poulsen (University of Copenhagen, Copenhagen, Denmark). FapyAde was purchased from Sigma Co. Cell Line and Culture. The THP-1 leukemia acute monocytic human cell line from ATCC was isolated from the peripheral blood of a young donor. It was grown in RPMI-1640 medium supplemented with 10% decomplemented FCS that contained 2 mM glutamin in a 5% CO2 atmosphere at 37 °C. UVA and γ Irradiations. UVA irradiation was carried out using a high-pressure Tecimex apparatus (Dixwell, St. Symphorien d’Ozon, France). The spectral distribution of the lamp exhibited a maximum at 373 nm. The dose rate was determined to be 45 mW/cm2 using a radiometer (Dixwell). γ-Ιrradiation was performed using a 60Co source at a dose rate of either 50 or 20 Gy/min for the HPLC/GC-MS and HPLC-EC measurements,
Pouget et al. respectively. Prior to UVA or γ irradiation, cells were pelleted at 240g for 4 min. Then, they were washed once in PBS buffer, and resuspended in PBS to give a cell suspension with a density of either 2 × 106 or 20 × 106 cells/mL for the comet assay and the chromatographic analysis, respectively. All irradiations were carried out at room temperature. Nonirradiated cells were kept in the dark at 4 °C. Immediately after irradiation, cells used for the comet assay were mixed with 1.2% low-melting point agarose, and the resulting mixture was laid on the slides. Cells used for the HPLC/GC-MS and HPLC-EC assays were pelleted at 240g for 4 min. Then, resulting pellets were resuspended in lysis buffer A prior to DNA extraction. Photosensitization Reactions. Acridine orange and rose bengal were used at concentrations of 5 10-3 OD (λmax ) 492 nm, ) 28 100 L cm-1 mol-1, 0.18 µM) and 10-2 OD (λmax ) 548 nm, ) 82 000 L cm-1 mol-1, 0.12 µM), respectively. The experiments were also repeated for both photosensitizers at a concentration of 1 OD (36 and 12 µM for acridine orange and rose bengal, respectively). Prior to light exposure, cells were pelleted at 240g for 4 min to remove culture medium. Then, they were washed once in PBS buffer and resuspended in PBS to give a cell suspension with a density of 2 × 106 or 20 × 106 cells/mL for the comet assay and the chromatographic analysis, respectively. Cells were then incubated for 5 min in the presence of either rose bengal or acridine orange and washed twice with PBS to remove the remaining photosensitizer. Irradiations were carried out at room temperature using a commercial 100 W visible lamp (Philips, Eindhoven, Netherlands) at a distance of 18 cm. As control experiments, cells were either kept in the dark or exposed to visible light alone without the photosensitizers. Immediately after irradiation, cells used for the comet assay, the HPLC/GC-MS method, or the HPLC-EC analysis were treated as previously described (vide supra). DNA Extraction. The chaotropic method reported by Helbock et al. (14) was used for DNA extraction. Typically, immediately after irradiation, cells were pelleted at 240g for 4 min. The resulting pellets (15 × 106 cells/sample) were resuspended twice in 2 mL of lysis buffer A (320 mM sucrose, 5 mM MgCl2, 10 mM Tris-HCl, 0.1 mM deferoxamine, and 1% Triton X-100) and centrifuged at 1500g for 10 min. The pellets were then resuspended in 0.6 mL of lysis buffer B (5 mM Na2EDTA, 10 mM Tris-HCl, and 0.15 mM deferoxamine). Subsequently, 10% SDS (35 µL) was added, and the mixture was vortexed before the addition of 30 units of ribonuclease A (1 mg/mL) and 8 units of ribonuclease T1. The resulting samples were incubated at 50 °C for 15 min. Then, 30 µL of protease (20 mg/mL) was added, and the samples were incubated for 1 h at 37 °C. DNA was precipitated by the addition of 0.8 mL of NaI solution (20 mM Na2EDTA, 7.6 M NaI, 40 mM Tris-HCl, and 0.3 mM deferoxamine) and 2 mL of 100% propan-2-ol. DNA was centrifuged at 5000g for 10 min and washed with 1 mL of 40% propan-2-ol and 1 mL of 70% ethanol. DNA was then dissolved in 100 µL of 0.1 mM deferoxamine. HPLC-EC-UV Measurements. The HPLC system consisted of an L 6200 Merck pump (Darmstadt, Germany) connected to a SIL 9A automatic injector (Shimadzu, Kyoto, Japan). The isocratic mobile phase was made of 50 mM KH2PO4 (pH 4.5) that contained 8% methanol. The flow rate was 1 mL/min. Separation of the nucleosides was performed using a C18 reversed-phase Uptisphere (5 µm, 4.6 mm × 250 mm) column from Interchim maintained at 30 °C. The retention times of 8-oxodGuo and dGuo were 18.6 and 13 min, respectively. A Coulochem II, model 5200A, electrochemical detector (ESA, Chelmsford, MA) was used for the measurement of 8-oxodGuo. The oxidation potentials of the electrodes of the analytical cell (model 5011, ESA) which were connected to the column were set at 200 and 450 mV for E1 and E2, respectively. In addition, the potential of the guard cell placed prior to the inlet of the injector was set at 500 mV. Elution of unmodified nucleosides was monitored using an UV detector (model 2151, LKB Bromma, Uppsala, Sweden) set at 280 nm. After DNA extraction, enzymatic hydrolysis was performed at 37 °C for 90 min using
Comet Assay and HPLC/GC-MS and HPLC-EC Measurements 10 units of nuclease P1 and 0.5 unit of acidic phosphatase dissolved in 10 µL of nuclease P1 buffer [300 mM sodium acetate and 1 mM ZnSO4 (pH 5.3)]. Thereafter, 50 µL of chloroform was added to precipitate proteins. The aqueous layer was collected and injected onto the HPLC-EC-UV system. For each sample, the amount of DNA injected onto the column was estimated using the UV signal of dGuo after appropriate calibration. HPLC/GC-MS Measurements. For GC-MS measurements, 35 × 106 cells (350 µg of DNA) were required to be able to detect Fapy derivatives in cellular DNA. Then, DNA extraction was performed as described above, up to the enzymatic hydrolytic step. Nucleoside 5′-monophosphates were obtained by addition of 20 units of nuclease P1 dissolved in 20 µL of buffer [300 mM sodium acetate (pH 5.3) and 1 mM ZnSO4] to the extracted DNA. The two isotopically labeled internal standards consisting of [15N3]FapyGua (100 pmol) and [15N2,13C]FapyAde (100 pmol) were added at this stage. An aliquot fraction (100 µL) of the samples was dried under vacuum prior to the acidic hydrolysis step. The remainder of the solution (20 µL) was used to measure the amount of DNA in each sample by HPLC-UV detection. For this purpose, the HPLC-EC-UV system described above (EC signal turned off) was used to measure dGMP and thus to estimate the amount of DNA in the samples after appropriate calibration. The mobile phase was a mixture of 50 mM KH2PO4 and methanol in a 99:1 volume ratio with the flow rate set at 1 mL/min. Under these conditions, the retention time of dGMP was 18 min. Acidic hydrolysis of the enzymatically released Fapy nucleotides was achieved with 100 µL of formic acid (88%) maintained at room temperature for 10 min. Then, the resulting samples were resuspended twice with 50 µL of distilled water and dried under vacuum. Subsequently, the dry residue was resuspended in a mobile phase that consisted of 25 mM ammonium formate. The HPLC prepurification system consisted of a 2150 LKB pump (Pharmacia LKB Biotechnology) connected to a SIL-9A autosampler (Shimadzu). The separations were performed using a C18 reversed-phase Uptisphere (5 µm, 4.6 mm × 250 mm) column from Interchim. The mobile phase consisted of the following gradient. The mobile phase was a 25 mM ammonium formate solution (100%) for the first 10 min. Then, methanol was added to the mobile phase so the proportion reached 20% within a 15 min period. The mixture was maintained for 5 min, and then, the percentage of methanol was decreased to 0%. The flow rate was 1 mL/min. Five fractions were collected, between 2 and 7 min after injection of 100 µL of the sample, using a FC 204 fraction collector (Gilson, Villiers-le-Bel, France). The retention times of FapyGua and FapyAde were 3.5 and 4.5 min, respectively (fractions 2 and 3, respectively). The collected fractions were then lyophilized to remove ammonium formate before the derivatization of the samples. For this purpose, samples were dissolved in a mixture of 25 µL of BSTFA and 25 µL of acetonitrile. Then, the silylation of the modified bases was achieved by heating the resulting solution at 120 °C for 20 min. GC-MS analysis was performed on an HP 5890 series II gas chromatograph (Hewlett-Packard, Les Ulis, France). The capillary column (0.25 mm, 15 m) used for the separation was coated with a 0.1 µm film of methylsiloxane substituted with 5% phenylsiloxane (HP5-MS, Hewlett-Packard). The injection of the sample (2 µL) was performed in the splitless mode with the temperature of the injection port set at 250 °C. The carrier gas was helium used at a velocity of 47 cm/s. The temperature of the interface/transfer line was set at 280 °C. The column temperature was initially maintained at 110 °C and raised to 280 °C within 11.3 min of injection. Detection of the positive ions was achieved with an HP 5972 mass detector operating in the electron impact ionization mode. Chromatograms were recorded in the single-ion monitoring mode. The characteristic ions of FapyGua and [15N3]FapyGua were detected at m/z 457 and 460, respectively, corresponding to the addition of four trimethylsilyl groups to the FapyGua molecule. The ions monitored for FapyAde and [15N2,13C]FapyAde were detected at
Chem. Res. Toxicol., Vol. 13, No. 7, 2000 543
m/z 369 and 372, respectively, corresponding to the addition of three trimethylsilyl groups to the FapyAde molecule. The respective retention times of silylated FapyGua and FapyAde were 7 and 5.5 min. Alkaline Single-Cell Gel Electrophoresis. Typically, 150 µL of 1% standard agarose was laid on a frosted slide and covered by a coverslip before being kept at room temperature for 5 min. Immediately after the irradiation (UVA, γ, or photosensitizer and visible light), 10 µL of the cell suspension was embedded in 100 µL of 1.2% low-melting point agarose. Cover glasses were removed, and the mixture was laid on the slides. Slides were covered by a coverslip and kept at 4 °C to allow the gel to solidify. The cover glasses were removed, and the slides were immersed in a lysis buffer [2.5 M NaCl, 100 mM Na2EDTA, 10 mM Tris (pH 10), 1% sodium sarcosinate, 1% Triton X-100, and 10% DMSO] maintained at 4 °C for 1 h. To convert radiation-induced base damage into strand breaks, two E. coli DNA N-glycosylases such as Fpg and endo III were used separately. For this purpose, slides were rinsed three times with Tris-HCl (0.4 M, pH 7.4) to remove the remaining lysis buffer and twice with Fpg buffer [20 mM Tris, 0.2 mg/mL bovine serum albumin, 0.5 mM Na2EDTA, and 0.1 M KCl (pH 8)]. Then, 100 µL of either Fpg (1 µg/mL) or endo III protein (1 µg/ mL) was laid on the slides. Cover slips were added, and the slides were then incubated at 37 °C for 45 min before being immersed in the electrophoresis buffer (300 mM NaOH and 1 mM Na2EDTA) for 20 min. Electrophoresis was performed at a voltage of 25 V and a current of 300 mA for 45 min. Then, the slides were twice rinsed with Tris-HCl (0.4 M, pH 7.4) before staining the nuclei with 60 µL of ethidium bromide (0.5 mg/ mL) per slide. Analysis of the slides was performed using an epifluorescence microscope (Carl Zeiss, Microscope Division, OberKochen, Germany) equipped with a short arc mercury HBO lamp (50 W, 516-560 nm) from Zeiss. The computer image analysis software Komet 3.1 (Kinetic Imaging, Liverpool, U.K.) was used to measure the tail moment of 50 cells per slide (expressed in micrometers). The tail moment is defined as the product of the tail length by the ratio of the intensity between the head and tail. For each irradiation dose, the average tail moment was determined using five different slides. Fapy Derivative Stability under the Conditions of the Alkaline Comet Assay. An aqueous solution of isolated calf thymus DNA was exposed to γ radiation (100 Gy). Then, the pH was increased from 7 to 13 by addition of 300 mM NaOH. The solution was maintained at 4 °C for increasing periods of time. Then, the solution was neutralized by the addition of 300 mM HCl. Subsequently, Nap 25 size exclusion columns were used to eliminate the salts and to separate any released Fapy derivatives from DNA. After evaporation to dryness, the fraction that contained DNA was then dissolved into 100 µL of 0.1 mM deferoxamine. Subsequently, enzymatic hydrolysis was performed to allow the measurement of Fapy purines by HPLC/ GC-MS (vide supra).
Results Measurement of 8-OxodGuo by HPLC-EC-UV. (1) γ and UVA Irradiation. The measurement of 8-oxodGuo in the DNA of cells exposed to γ rays was performed using the accurate HPLC-EC assay. For this purpose, DNA was extracted using the chaotropic NaI method (14). A linear formation (0.011 8-oxodGuo per 106 bases per Gy) as a function of the applied dose (0-80 Gy) was observed (Figure 1). The background level was estimated to be 0.2 8-oxodGuo/106 bases. This is 9-fold lower than the background measured using the conventional NaCl DNA extraction procedure (26). Interestingly, the relatively low value of the steady-state level of 8-oxodGuo allows the measurement of a significant increase in the formation of the latter lesion at a dose of
544
Chem. Res. Toxicol., Vol. 13, No. 7, 2000
Pouget et al.
Table 1. Yields (expressed in the number of lesions per 106 bases) of 8-OxodGuo, FapyGua, Strand Breaks, Sensitive Sites to Fpg or Endo III As Measured by HPLC-EC, GC-MS, and the Modified Comet Assaya
8-oxodGuo (HPLC-EC) FapyGua (HPLC/GC-MS) sites sensitive to Fpg sites sensitive to endo III SSB + DSB + ALS
γ rays (yield/Gy)
UVA (yield/kJ-1 m-2)
0.01 OD of rose bengal (yield/min)
0.005 OD of acridine orange (yield/min)
1 OD of rose bengal (yield/min)
1 OD of acridine orange (yield/min)
0.011 ( 0.001
0.00098 ( 0.0002
0.054 ( 0.006
0.097 ( 0.013
1.88 ( 0.04
3.93 ( 0.21
0.027 ( 0.001
ND1b
ND1
ND1
0.66 ( 0.08
0.9 ( 0.2
0.048 ( 0.025
0.0019 ( 0.0007
0.05 ( 0.02
0.10 ( 0.03
ND2c (saturation)
ND2 (saturation)
0.053 ( 0.035
0.0003 ( 0.0002
