Mutagenesis Advance Access published March 23, 2012

Mutagenesis pp. 110, 2012 doi:10.1093/mutage/ges010

DNA damage induced by micro- and nanoparticlesinteraction with FPG inuences the detection of DNA oxidation in the comet assay

ller1,* J. Kain1,y, H. L. Karlsson1,2 and L. Mo

Unit for Analytical Toxicology, Department of Biosciences and Nutrition at Novum, Karolinska Institutet, SE-141 83 Huddinge, Sweden and 2Division of Molecular Toxicology, Institute of Environmental Medicine, Karolinska Institutet, SE-171 77 Stockholm, Sweden y Former name: J. Gustafsson
* To whom correspondence should be addressed: Tel: 46 8 524 810 75; Fax: 46 8 774 68 33; Email: lennart.moller@ki.se 1

Received on May 16, 2011; revised on February 10, 2012; accepted on February 14, 2012

Reliable methods for evaluation of toxicity from particles, such as manufactured nanoparticles, are needed. One promising tool is the comet assay, often used to measure DNA breaks (strand breaks and alkali-labile sites) as well as oxidatively damaged DNA, the latter by addition of specic DNA repair enzymes such as formamidopyrimidine DNA glycosylase (FPG). The aim of this study was to investigate the use of the comet assay for analysis of DNA oxidation by a range of micro- and nanoparticles in the lung cell lines A549 and BEAS-2B and to test the hypothesis that nanoparticles present in the cells during the assay performance may interact with FPG. This was done by investigating the ability of micro- and nanoparticles (stainless steel, subway particles, MnO2, Ag, CeO2, Co3O4, Fe3O4, NiO and SiO2) to induce DNA breaks, oxidatively damaged DNA (FPG sites, dominantly 8oxoguanine), intracellular production of reactive oxygen species (ROS) and non-cellular oxidation of the DNA base guanine, as well as by studying interactions of the particles and their released ions with FPG. Several particles caused DNA breaks, but low levels of FPG sites. The ability of FPG to detect DNA oxidation induced by a photosensitiser was however shown. An oxidative capacity of the particles was indicated by increased levels of intracellular ROS, and especially Ag and subway particles caused non-cellular oxidation of guanine. Incubation of FPG with the particles led to less FPG activity, particularly with nanoparticles of Ag but also with CeO2, Co3O4 and SiO2. Further investigations of these particles revealed that for Ag, the decreased activity was mainly due to released Ag ions, whereas for CeO2 and Co3O4, FPG interactions were due to the particles. We conclude that measurement of oxidatively damaged DNA in cells exposed to nanoparticles may be underestimated in the comet assay due to interactions with FPG.

Introduction The link between exposure to ambient particles in air pollution and health effects, such as cancer, cardiovascular and respiratory diseases, is well known and inammation and oxidative stress are believed to be the main routes from the

particle exposure to diseases (1,2). Particles from other sources, such as metal-based nanoparticles in consumer products, from industrial production sites and other environments such as the subway system are less studied and imply new challenges when it comes to human health risk assessment. Nanoparticles, often classied as particles with at least one dimension ,100 nm, are due to their many interesting properties more extensively used today in medical applications as well as in different industrial and consumer products. The unique properties of nanoparticles include a large surface area per unit mass, which can display a large amount of reactive surface molecules leading to high reactivity, as well as unique material properties affecting, e.g. electrical conductivity, magnetic characteristics and hardness. Such properties may not only be benecial in different contexts but can also induce toxicity (3). Due to the expanding use of nanoparticles and increased exposure risks, there is an urgent need for risk assessment and legislation. This may also concern particles from occupational settings or other milieus, e.g. the subway system, where little is known about the toxicity of the emitted particles. Many in vitro toxicity assays designed for testing of chemicals are also used for particles. Particles do, however, display several unique physicochemical properties due to their solidity, such as size, shape, agglomeration properties, elemental purity and surface area, which make them different when compared to chemicals. Some toxicity assays may therefore not be appropriate to use for particles. The unique particle properties may lead to interaction in assays and give misleading information about toxicity (4). The interference can, e.g. be due to the fact that nanoparticles have optical absorbance at the same wavelength as a coloured product in an assay or adsorb dyes or cytokines that are supposed to be measured (5). Reaction between the particle surface and the dye, endotoxin contamination of the particles and magnetic properties may also cause problems (5). In the present need for screening methods for evaluation of particle toxicity, it is therefore important to evaluate the reliability of the applied assays. One promising tool in toxicity testing of particles is the comet assay (single-cell gel electrophoresis), where DNA damage is measured in single cells. In this method, cells are lysed, the DNA is denaturated and electrophoresis is performed. During the electrophoresis, strand breaks in the DNA allow DNA to migrate out of the nucleoid, which appears as comets when evaluated microscopically. The amount of DNA in the comet tail represents DNA breaks in the form of strand breaks (SB) and alkali-labile sites (ALS). By using DNA repair enzymes, often formamidopyrimidine DNA glycosylase (FPG), more specic damage can also be measured. FPG predominantly detects oxidatively damaged DNA (FPG sites), which can be a product of oxidative stress, in the form of oxidised purines, primarily 8oxoguanine and ring-opened formamidopyrimidine bases (e.g. FapyGua) (6). Formation of 8-oxoguanine, which is mutagenic and in general assumed to be related to carcinogenesis, is 1

Downloaded from http://mutage.oxfordjournals.org/ by guest on October 14, 2013

The Author 2012. Published by Oxford University Press on behalf of the UK Environmental Mutagen Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

J. Kain et al.

considered to be a marker for oxidative stress (7). Oxidative stress and 8-oxoguanine have been shown to be increased in cancer tissue (8,9) and studies have, by using the comet assay, linked human ambient particle exposure to oxidatively damaged DNA in lymphocytes, as a measure of systemic oxidative stress (10,11). There is no standardised protocol for the comet assay, but the method has been evaluated in inter-laboratory trials conducted by the European Standards Committee on Oxidative DNA Damage (ESCODD) (12) and by the European Comet Assay Validation Group (ECVAG) (13,14). The importance of the protocol used has also been highlighted in two recently conducted studies (15,16). The comet assay seems to be a good method to use for evaluation of DNA damage induced by chemicals. However, the method is used more and more extensively in the in vitro context where cell lines are exposed to particles. A recent review identied 46 cellular in vitro studies, many of them published recently, where the comet assay was used when investigating the toxicity of nanoparticles (17). Data on possible particle interactions within the comet assay are, however, very limited. The presence of particles in the comet head (nucleoid) has been observed (1719) and the possibility of particles causing additional DNA damage during the assay has been discussed (19,20). The presence of particles close to DNA during the comet assay makes interaction of particles with FPG a possible scenario. Interaction of particles with FPG when measuring oxidatively damaged DNA with the comet assay is a potential problem that, to the best of our knowledge, has not been studied before. The aim of this study was to investigate the use of the comet assay for in vitro evaluation of oxidatively damaged DNA (FPG sites) caused by various particles and to test the hypothesis that nanoparticles present in the cells during the assay performance may interact with FPG. The particles included in the study were engineered nanoparticles: Ag, CeO2, Co3O4, Fe3O4, NiO, SiO2; one engineered microparticle: MnO2 and collected microparticles from the subway system and stainless steel industry. More specically this study i) analyses the ability of a wide range of micro- and nanoparticles to cause DNA damage in the form of DNA breaks and FPG sites, in two lung cell lines (alveolar A549 and bronchial BEAS-2B) by the comet assay, ii) investigates the ability of FPG to detect oxidatively damaged DNA induced by a non-particular agent (photocatalytic chemical agent Ro 19-8022, Ro) and light irradiation in A549 and BEAS-2B cells, iii) evaluates the oxidative potential of the particles in form of intracellular production of ROS in A549 cells and oxidation of the DNA nucleoside 2#-deoxyguanosine (dG) in a non-cellular system and iv) investigates possible interaction of particles and their released ions with FPG.

Particle characterisation Transmission electron microscopy (TEM) was used to analyse the particle size and shape. Images were captivated using a Tecnai 10 apparatus (Fei, The Netherlands) at an acceleration voltage of 100 kV and a Mega View III digital camera (Soft Imaging System, GmbH, Mu nchen, Germany). Before the analysis, the particles were suspended in MilliQ water and sonicated with a probe for 2 20 s with a 20-s break in between, in order to reduce agglomeration. Droplets of 3 ll were placed on TEM grids for 5 min and the water was removed using a lter paper before samples were analysed. Specic surface area of the particles was determined by Brunauer, Emmett, Teller analysis (BET), using a Micromeritics FlowSorb II 2300/Micromeritics Gemini V instrument at Kanthal AB, Sweden. In short, the measurement of the surface area of a particulate sample, with known mass, was estimated via the adsorption of nitrogen at cryogenic conditions. Since the size of nitrogen atoms and the amount of nitrogen atoms adsorbed on to the surface of the particles are known, an estimation of particle surface area can be measured. The size of the nanoparticles in supplemented cell medium was measured by dynamic light scattering (DLS) using a Zetasizer nano ZS (Malvern, UK). Before the measurements, the particles were suspended in supplemented Dulbeccos minimal essential medium (DMEM) medium and sonicated in the same manner as for cell exposure (see below). The particle size was calculated from the Brownian motion of the particles using the StokesEinstein equation. The method yields a hydrodynamic diameter, which is a calculated particle diameter of a sphere that has the same measured motion in the solution as the actual particle. Using the same system, the surface charge (zeta potential) of the particles was analysed. Particles were suspended in MilliQ water with 10 mM NaCl and sonicated in the same manner as for cell exposure (see below). Due to the high ionic strength, measurements could not be conducted in cell medium. The zeta potential was measured by applying an electrical eld through the solution causing particles with different charge to migrate at different speed towards the electrodes, which was measured and converted to zeta potential using Smoluchowskis theory. DLS and zeta potential measurements could not be conducted for the microparticles due to sedimentation. Cell culture Human alveolar Type II-like epithelial (A549) and bronchial epithelium (BEAS-2B, a kind gift from A. Ljungman, Department of Molecular and Clinical Medicine, University of Linko ping, Sweden) cell lines were used (both originally obtained from the American Type Culture Collection, Manassas, VA, USA). The cells were grown in DMEM supplemented with 10% heatinactivated foetal bovine serum, 100 U/ml penicillinstreptomycin and 1 mM sodium pyruvate in a humidied atmosphere at 37C and 5% CO2. Cell exposure to particles Particles were suspended to 1 mg/ml in supplemented DMEM. The suspensions were vortexed 20 s and sonicated using a probe, 2 20 s with a 20-s break in between, to minimise agglomeration of the particles. The particles were then diluted further in 37C supplemented DMEM to 20 and 40 lg/cm2 (corresponding to 40 and 80 lg/ml) for the comet assay and to 40 lg/cm2 for ROS measurements. Exposure of A549 and BEAS-2B cells (BEAS-2B, only used for comet assay) was performed in 24-well plates in a humidied atmosphere at 37C and 5% CO2, 4 h for comet assay and 1 h for measurements of intracellular ROS production; 0.16 million cells had been grown in each well for 24 h to get a 90100% conuent cell layer at the time for particle exposure. After particle exposure, cells were washed (with phosphate-buffered saline [PBS] for the comet assay and Hanks balanced salt solution (HBSS) for measurements of intracellular ROS production). The cells were removed by trypsination at 37C and mixed with supplemented DMEM before the comet assay. Control cells were exposed to supplemented DMEM medium. Cell exposure to photosensitiser (Ro 19-8022) The capacity of the FPG comet assay to detect oxidatively damaged DNA in the cells was investigated by the addition of the chemical photocatalytic agent Ro 19-8022 (F. Hoffmann-La Roche, Basel, Switzerland) followed by light irradiation of the cells. Ro together with light irradiation induces a damage prole similar to that of singlet oxygen, which predominantly produces 8oxoguanine (21). To get a 90100% conuent cell layer in six-well plates (Becton Dickinson, Franklin Lakes, NJ, USA) at exposure, 0.64 million cells of A549 and BEAS-2B were seeded in each well in supplemented DMEM at 37C and 5% CO2, 24 h before exposure. The cells were washed with PBS before Ro exposure. Ro was added to 2 ml cold PBS in the nal exposure concentrations of 0, 0.1, 0.2, 0.4 and 0.6 lM, respectively. Irradiation was performed, directly after addition of Ro to the cells, with a white light lamp (500 W halogen) for 5 min, 33 cm above the cells placed on ice. Cells were then washed, quickly trypsinated and cold supplemented DMEM was added and the cells were

Material and methods

Particles Nanoparticles (,100 nm) of Ag (American Elements, Los Angeles, CA, USA), CeO2, Co3O4, Fe3O4, NiO, SiO2 (SigmaAldrich, St Louis, MO, USA), microparticles (,10 lm) of MnO2 (SigmaAldrich), particles from the Stockholm subway system and 316L stainless steel particles (Sandvik Osprey Ltd, UK) were investigated.

FPG interaction collected into tubes on ice. All work was performed in darkness, except for the irradiation. The cells were washed once with PBS before the comet assay was performed. The comet assay DNA breaks (SB and ALS) were measured with the alkaline version of the comet assay. Escherichia coli FPG (kindly provided by Professor A. R. Collins, Department of Nutrition, School of Medicine, University of Oslo, Norway), which mainly detects purine oxidation products, was added to the assay for evaluation of oxidatively damaged DNA (FPG sites). After Ro or particle exposure, A549 and BEAS-2B cells were embedded in 0.75% agarose on microscopic slides precoated with 0.3% agarose. After drying of the gels at a cold plate in dark for a few minutes, the slides were placed in a lysis buffer with pH 10 (1% Triton X-100, 2.5 M NaCl, 10 mM Tris, 0.1 M EDTA) for 1 h to lyse the cells and separate DNA from the histones. The slides were then immersed in enzyme buffer (0.1 M KCl, 0.5 mM EDTA, 40 mM HEPES, 0.2 mg/ml BSA) 3 5 min and then placed on a cold plate. The FPG was diluted 1:3000 in enzyme buffer. Diluted enzyme or enzyme buffer only (30 ll) was added to each gel on the microscope slides and incubation was performed in a humidity chamber at 37C for 30 min. The time for incubation was based on previous studies of FPG (16). Unwinding of DNA was then performed in alkaline solution (0.3 M NaOH, 1 mM EDTA) for 40 min, on ice in darkness, followed by electrophoresis for 30 min (17 V, 1.15 V/ cm electrophoresis platform) in a black COMET-20 tank (Scie-Plas Ltd, Cambridge, UK) with cold water circulating under the platform. During the electrophoresis, the negatively charged DNA moves towards the anode and create comets. Neutralisation was performed in 0.4 M Tris for 2 5 min and in H2O for 5 min. Gels were dried over night and xed with methanol for 5 min prior to staining of the DNA with EtBr (1 lg/ml) in TAE buffer. The comets were examined with a uorescence microscope (Olympus BH2 with a 20 apochromatic objective) using the program Komet 4.0 (Kinetic Imaging Ltd). For each sample, 35 comets on two different slides (i.e. 70 comets/sample) were evaluated in each experiment. The level of FPG sites was obtained by subtracting the value of % DNA in the tail obtained without added enzyme from the value when the enzyme was present. Measurement of intracellular production of ROS using 2# , 7#-dichlorouorescin diacetate Intracellular ROS production was measured using the oxidation-sensitive uoroprobe 2# , 7#-dichlorouorescin diacetate (DCFH-DA; SigmaAldrich). DCFH-DA is a non-uorescent compound that is freely taken up by cells and hydrolysed by esterases that remove the DA group. DCFH is then oxidised to uorescent dichlorouorescein (DCF) by cellular oxidants, thereby indicating the level of intracellular ROS. A549 cells were incubated with particles, 40 lg/cm2 (80 lg/ml) for 1 h and then washed (previously described). DCFH-DA was diluted in anhydrous DMSO (SigmaAldrich) to 10 mM and stored as a stock solution in 4C. A working solution of 100 lM was prepared in 37C HBSS and was added to the cells for 30 min in 37C after the particle exposure. The DCFH-DA was then removed, and each well was washed with HBSS (37C). Cold lysis buffer (1% Triton X-100, 2.5 M NaCl, 10 mM Tris, 0.1 M EDTA, pH 10) was added to the wells, and after 1 min, the solution was collected and centrifuged at 2800 g, for 4 min at 4C; 2 100 ll of the supernatant was added to two wells in a 96-well plate (white, clear bottom) and uorescence was measured in a computerised microplate uorometer (SPECTRA Max GEMINI EM; Molecular Devices, USA) at 485 nm excitation and 530 nm emission. During the whole procedure with DCFH-DA, the plate was kept out of light to minimise fading of the uoroprobe. A mean value for the uorescence was calculated for the two wells in each experiment and the background uorescence (cells without DCFH-DA added) was subtracted. The values were expressed as times increase compared to control (unexposed cells). dG reaction and 8-oxo-2#-deoxyguanosine analysis by high-performance liquid chromatographyelectrochemical detector dG was dissolved to a concentration of 0.8 mM in saline (0.9% NaCl). Particles were suspended to 1 mg/ml in saline and sonicated 2 20 s. dG and particles were then mixed to a nal concentration of 0.4 mM dG and 50 or 100 lg/ml of particles. The reaction mix was incubated at 37C for 30 min and was mixed after 15 min by brief vortexing. The reaction mix was centrifuged at 20 000 g for 10 min at 4C in order to remove the particles, and the supernatant was immediately frozen at 80C. The formation of 8-oxo-2#-deoxyguanosine (8oxodG) was analysed using an high-performance liquid chromatography electrochemical detector (HPLC-EC) system that consisted of Scantec pump 100 (Knauer, Berlin, Germany) set at 0.75 ml/min, a 1 mm (C18) Opti-Guard column (Optimize, Portland, OR, USA), and two Delta-Pak reversed-phase columns (Waters, Milford, MA, USA). 8-oxodG was detected with an EC (Coulochem II; ESA, Chelmsford, MA, USA) and dG was detected with a UV detector (Waters) set at 290 nm. The HPLC buffer consisted of 10% v/v methanol and 20 mM sodium acetate set to pH 5.3 by acetic acid. Particle and ion interaction with FPG To evaluate the possible interaction of particles and ions with FPG, A549 cells with oxidatively damaged DNA (induced using 0.6 lM Ro light irradiation, as previously described) were evaluated with the comet assay and FPG pre-exposed to particles or ions (illustration of the procedure, Figure 6). Particles and ions located inside the cell (that are not washed away during the comet assay) are likely available for interaction with FPG in the comet assay. In the rst approach (A, in Figure 6), an estimate of the intracellular dose [15% of the total particle concentration 40 lg/cm2, based on a study on Cu nanoparticles (22)] of all particles (prepared as previously described) was mixed with FPG (1:3000) for 30 min at 37C, mimicking the enzyme incubation in the comet assay. The FPG samples were then centrifuged at 19 900 g for 4 min at 4C and the supernatants (not containing particles) were added onto the microscopic slides with Ro-exposed cells, incubated for 30 min and treated according to the comet assay protocol previously described. This procedure shows if the particles adsorb/ bind FPG, which then is centrifuged away, or if FPG gets dysfunctional after contact with the particles. The particles that caused a reduced ability of FPG to detect oxidatively damaged DNA (A, in Figure 6) were further evaluated (procedure B and C in Figure 6) to investigate the reason for the lower detection capacity of FPG. FPG was then mixed with the particles and directly used in the comet assay for detection of oxidatively damaged DNA (B, in Figure 6). The results indicate if FPG still works even though it is bound to the particle. To evaluate if the decreased enzymatic capacity was due to ions released from the particles (C, in Figure 6), the particles were incubated in enzyme buffer for 30 min in 37C, centrifuged at 19 900 g for 4 min at 4C and the supernatant with possible ions were incubated with FPG for 30 min in 37C. The ability of this FPG, pretreated with released ionic species, to detect oxidatively damaged DNA was then investigated. This procedure thus shows if the detected inhibition of FPG was due to released ions (C, in Figure 6). A control with FPG not preexposed to particles or ions was also evaluated in all the experiments. Statistics Statistical analyses were performed using PASW Statistics 18.0 (SPSS Inc., Chicago, IL, USA). Two-way ANOVA was used for comparison of the level of damage in the cell lines after Ro exposure. Pearsons correlation test was used for correlation estimates of DNA damage between A549 and BEAS-2B after particle exposure. Unpaired two-tailed Students t-test with unequal variance was used for the pairwise comparison between control and the different exposures. In all statistical analyses, a P value of ,0.05 was considered statistically signicant.

Results Particles characteristics Characteristics of the particles, size, shape, z-potential and BET area are shown in Table I. TEM pictures of the particles are presented in Figure 1. Further characteristics of the particle 316L have previously been analysed by the authors (23). Particle-induced DNA breaks and FPG sites measured with the comet assay DNA breaks. In A549 cells, several particles induced significantly increased amounts of DNA breaks compared to control (8.8% DNA in tail) at the concentration 20 lg/cm2, namely the nanoparticles CeO2 (15.6%), Co3O4 (14.6%), NiO (31.6%) as well as the micrometre particle MnO2 (16.1%) and subway particles (22.2%) (Figure 2A). At the higher concentration 40 lg/cm2, CeO2 (23.6%), NiO (29.9%) and MnO2 (18.8%) were signicantly increased when compared to control (9.2%) (Figure 2A). Subway particles showed an increase (17.1%) although not statistical signicant (P 5 0.066). The only particle type showing a signicant dose-dependent increase was CeO2 (Figure 2A). In BEAS-2B cells, the nanoparticles Ag (17.9%), Fe3O4 (12.3%), NiO (29.0%), the micrometre particle MnO2 (14.7%) and subway particles (14.4%) caused signicantly increased 3

J. Kain et al.

Table I. Particle characteristics in terms of particle size (in powder, estimated from TEM pictures and in cell medium), particle shape, z-potential and surface area Particle Particle size Powdera Micrometre particles (lm) Steel Subway MnO2 Nanoparticles (nm) Ag CeO2 Co3O4 Fe3O4 NiO SiO 10 ,25 ,50 2030 ,50 515 TEMb 3.114 0.22.6 0.30.8d 7516 425 962 2040 267 1127 DLS nd nd nd 315 225 222 ,200 167 8.7 Round Irregular Irregular Round Square Irregular Round Squaree Irregular nd nd nd 39.9 4.0 30.2 1.8 30.3 11.0 0.7 nd 37 5 70 41 42 102 nd Particle shapec Zeta potential (mV) BET area (m2/g)

nd, not detectable. a Particle size according to the manufacturer. b Particle size visually estimated from TEM pictures, longest dimension measured. c Particle shape estimated from TEM pictures. d Clusters of 35-nm particles were also present in the sample. e Square particles with round edges.

Fig. 1. TEM images of the nano- and micrometre particles for determination of primary size and shape. The bars show the size of 100 nm for all nanoparticles, 1 lm for stainless steel, 0.2 lm for subway and 0.5 lm for MnO2.

DNA breaks compared to the control (8.0%) when exposed to 20 lg/cm2 of the particles (Figure 2B). At the higher concentration 40 lg/cm2, Ag (16.2%), CeO2 (25.6%), Co3O4 (13.0%), NiO (28.5%), as well as MnO2 (26.2%) and subway particles (23.9%) caused signicantly increased levels of DNA breaks compared to control (9.2%) (Figure 2B). A statistically 4

signicant dose-related increase in DNA damage could only be seen for CeO2 (Figure 2B). FPG sites. The oxidative capacity of the different particles varied. Particles from subway and stainless steel induced

FPG interaction

signicantly increased oxidative damage to DNA in A549 cells (18.4 and 11.2% DNA in tail, respectively) at the concentration 20 lg/cm2, compared to control (7.2%). MnO2 caused an increase (12.5%), but this was not signicant (P 5 0.058) (Figure 2C). No difference in FPG sites was seen for the higher concentration, except for stainless steel where a signicantly lower amount of damage was observed for the higher concentration. These results suggest possible particle interaction with the assay (Figure 2C). In BEAS-2B cells, only Co3O4, in both 20 and 40 lg/cm2, caused a signicant increase in oxidatively damaged DNA (11.2 and 13.1%, respectively) compared to control; however, there was no signicant difference between the concentrations (Figure 2D). Correlation for A549 and BEAS-2B. There was a highly signicant correlation between DNA breaks, induced by particle exposure (40 lg/cm2), between A549 and BEAS-2B (P , 0.001, R2 5 0.85, Figure 3). Correlation between FPG sites was due to the lack of signicantly increased levels considered not relevant and therefore not shown. Intracellular ROS production using DCFH-DA DCF uorescence intensity, as a measure of the amount of intracellular ROS production in A549 cells, was signicantly increased compared to control for nanoparticles of Ag (2.4 times), Co3O4 (2.4 times), NiO (2.0 times) as well as for the microparticles of MnO2 (4.8 times) and subway particles (2.1 times) in the concentration of 40 lg/cm2. Steel particles caused an increase in uorescence intensity with 2.9 times the control, but this was not signicant due to large standard deviations (Figure 4).

Evaluation of FPG in the comet assay Ro induced oxidatively damaged DNA in a similar dosedependent manner in both A549 and BEAS-2B cells, detected with FPG in the comet assay (Figure 5). For both A549 and BEAS-2B, the lowest dose of Ro (0.1 lM) did not cause a signicant increase in damage compared to control, but all the other concentrations (0.2, 0.4 and 0.6 lM) did. When A549 and BEAS-2B were compared, there was no signicant difference in damage level (Figure 5). Particle-induced oxidation of dG, evaluated by HPLC-EC Analysis of 8-oxodG formation showed that some particles, especially subway and Ag particles, at the concentrations 50

Fig. 3. Correlation of DNA breaks (SB and ALS, % DNA in tail) in A549 and BEAS-2B after particle exposure (4 h) at the concentration 40 lg/cm2 (data from Figure 2AB). R2-value 5 0.85 (Pearson correlation test), P , 0.001.

Fig. 2. (AD) DNA breaks (SB and ALS) (A and B) and oxidatively damaged DNA (FPG sites) (C and D) in A549 and BEAS-2B cells from particle exposure; % DNA in tail of cells after 4 h of exposure to 20 lg/cm2 (grey bars) and 40 lg/cm2 (black bars) of the different particles. Each bar represents three to four independent experiments SD. Asterisks (*, ** and ***) indicate signicantly higher levels compared to controls and correspond to P , 0.05, 0.01 and 0.001, respectively. For CeO2 (A and B) and steel (C), the signicant difference between the concentrations is indicated by horizontal curly brackets.

J. Kain et al.

Fig. 4. Generation of intracellular ROS in A549 cells following particle exposure at the concentration 40 lg/cm2 for 1 h, measured as DCF uorescence. Each bar represents mean values of four to ve independent experiments SD. An asterisk (*) indicates a signicantly higher level compared to the control (P , 0.05) calculated on the measured uorescence values.

Fig. 5. Oxidatively damaged DNA (FPG sites) from the photosensitiser Ro in cultured A549 and BEAS-2B cells evaluated with FPG and the comet assay after exposure to 00.6 lM Ro and light. Each point represents the average of four independent experiment SD. Asterisks (*, **and ***) indicate signicantly higher levels compared to control and corresponds to P , 0.05, 0.01 and 0.001, respectively. The R2 values for the tted lines of the dose response were 0.991.00. There was no signicant difference between the cell lines.

MnO2, Fe3O4 and NiO detected similar levels after particle removal by centrifugation, whereas CeO2, Co3O4 and SiO2 resulted in signicantly lower levels (17.3, 28.2 and 19.3%, respectively). No FPG sites were detected by FPG preincubated with Ag particles (Figure 7A). To test whether FPG adsorbed/bound to the particles (and thus removed by centrifugation) still could be active, FPG was mixed with the particles and then directly added to the cells (without centrifugation). This showed that Ag, CeO2 and Co3O4 particles caused signicantly decreased enzymatic capacity (14.4, 22.8 and 24.6%) compared to the control (41.4%). In contrast, SiO2 particles did not cause any signicant decrease in enzymatic activity indicating that the FPG enzyme interacted with these particles but remained active (Figure 7B). To test whether the FPG interactions could be explained by released ions from the particles, FPG was incubated with the released fraction from the particles. The results showed that Ag ions caused total inhibition of FPG enzymatic activity, whereas the released ionic fraction from the other particles did not cause any inhibition of FPG in this test.

and 100 lg/ml, induced oxidation of dG. When formation of 8oxodG was normalised to the control value in each experiment, the levels of 8-oxodG caused by the subway particles were 5.6 and 12.1 times increased and for Ag 2.8 and 3.2 times increased, at the concentrations 50 and 100 lg/ml, respectively. MnO2, Co3O4 and NiO showed lower levels of 8-oxodG than the control, which may indicate a possible interference of these particles with the analysis of 8-oxodG. All particles were therefore incubated with a standard of 8-oxodG, the particles were removed by centrifugation and the remaining amount 8oxodG was analysed. This test showed that particles of MnO2, Co3O4 and NiO depleted the 8-oxodG, whereas the other particles showed no effect (data not shown). No data were therefore presented for these particles (Table II). Particle and ion interaction with FPG in the comet assay Analysis of particle interference with FPG (Figure 7A) showed that some particles caused decreased ability of FPG to detect FPG sites. When A549 cells, with oxidatively damaged DNA (induced by Ro light irradiation) were analysed with nonexposed FPG, 41.8% DNA in tail was observed. FPG preincubated with particles of stainless steel, subway particles, 6

Discussion It is well known that the risk assessment of nanoparticles lags behind the production and use in consumer products. To thoroughly complete toxicity testing for all the nanomaterials that exist today, the estimated time is between 34 and 53 years (24). This time perspective will most likely even be longer since new nanomaterials are introduced constantly. This phenomenon, that toxicity hazards of particles are tested and discovered long after the introduction to the consumer market concerns many occupational settings as well. Historically, the asbestos exposure constitutes one example (25). Due to the urgent need for toxicity testing of new particles, such as nanoparticles, accurate, reliable and easy methods are required. Preferable are in vitro tests that predict effects in humans. The increasingly used comet assay is one promising tool for measurement of DNA damage in the form of DNA breaks and oxidative damage to DNA, after particle exposure. This study investigated the use of the comet assay for analysing oxidation of DNA from a wide range of micro- and nanoparticles in human lung cell lines (A549 and BEAS-2B) and it highlights the risk of particle interaction in the applied toxicity testing method.

FPG interaction

Fig. 6. Procedure for the interaction test of particles and ions with FPG, in the comet assay. Step I: A549 cells were exposed to 0.4 lM Ro and light irradiation (5 min), which induced oxidatively damaged DNA (FPG sites, $40% DNA in tail). The damaged cells were then evaluated with the comet assay and FPG from Step II (procedures A, B and C). Step II: (A) FPG was pre-incubated with all particles for 30 min at 37C and centrifuged. The supernatant without the particles was then used to detect FPG sites of the cells from Step I in the comet assay. This procedure shows if the particles adsorb/bind FPG, which then is centrifuged away, or if FPG gets dysfunctional after contact with the particles. The particles that caused a lower activity of FPG in (A) were further evaluated in procedure (B) and (C). In (B), FPG was mixed with the particles and directly used in the comet assay for detection of oxidatively damaged DNA. The results indicate if FPG still works even though it is bound to the particle. In (C), the particles were incubated in buffer for 30 min for possible release of ions. The particles were centrifuged away and the released ions (if any) were incubated with FPG for 30 min. This procedure shows if the detected inhibition of FPG was due to released ions.

Table II. Oxidative capacity of the particles, measured by HPLC-EC, as oxidation of dG to 8-oxodG after particle incubation of the concentrations 50 and 100 lg/ml Particles Control Steel Subway MnO2 Ag CeO2 Ce3O4 Fe3O4 NiO SiO2 50 lg/ml mean 1.0 1.2 (1.01.5) 5.6 (1.311.7) ND 2.8 (1.84.2) 1.3 (1.11.5) ND 1.0 (1.01.1) ND 1.3 (1.01.7) 100 lg/ml mean 1.0 1.1 12.1 ND 3.2 1.2 ND 1.0 ND 1.3 (1.11.2) (6.522.6) (2.34.3) (1.11.2) (1.01.1) (0.91.9)

Values are based on three independent experiments and normalised to the control. Values in bold highlight the particles where the oxidation of dG was increased compared to control. Numbers in parentheses indicate the minimum maximum values. ND, no data due to interaction/binding of 8-oxodG.

When the FPG comet assay was evaluated for its capacity to detect oxidatively damaged DNA in vitro induced by the nonparticulate chemical agent photosensitiser Ro (19-8022), the present study showed that the method could detect dosedependent increase in damage in both cell lines (Figure 5). This has also been shown for A549 cells in the validation study of the comet assay by ECVAG (16), where seven out of eight laboratories could identify a doseresponse in FPG sites in Roexposed cells. A similar trial of exposed human transformed epithelial (HeLa) cells conducted by ESCODD showed, however, that only three of eight laboratories detected a doseresponse relationship in FPG sites (12), which partly may be explained by a narrower concentration range (16). The results of Ro in the present study support the general comprehension that the comet assay is a reliable method to

use for evaluation of oxidatively damaged DNA from chemicals. The photosensitiser Ro, used in this study to induce oxidative damage to DNA, induce a damage prole very similar to that of singlet oxygen, which induces oxidative purine modications, mostly in the form of 8-oxoguanine (21). FPG might in addition to 8-oxoguanine also nd small amounts of other types of DNA damage, such as FapyGua, FapyAde and ring-opened purines (6). FPG has also been found to detect alkylation of DNA (26). Formation of alkylation damage by Ro is, however, not likely but cannot be excluded. In the present study, several of the tested particles induced DNA breaks but low levels of oxidative damage to DNA when examined with the comet assay (Figure 2AD). These particles have previously shown diverse toxic and oxidative capacity by others (27,28) and a variation from low to high levels of oxidative damage to DNA was expected. Fe3O4 nanoparticles, previously shown by the authors to cause low levels of oxidatively damaged DNA (29,30), were used as low toxicity particles in this study whereas subway particles were expected to cause higher toxicity based on previous studies, using the comet assay (31) and other methods (32). Ag nanoparticles were expected to cause increased levels of oxidatively damaged DNA since, e.g. Hussain et al. (33) showed oxidative stress caused by Ag nanoparticles, but this was not seen in our study by the FPG comet assay. The oxidative capacity of several of the particles was, however, shown by their generation of intracellular ROS (Figure 4) and especially Ag nanoparticles and subway particles caused oxidation of dG to 8-oxodG when dG was exposed acellularly. The lack of oxidative damage to DNA with the comet assay, seen for many particles in this study, is in line with previous studies of microparticles and nanoparticles by the authors (29,30). Also in these studies, it is evident that many particles show a positive effect when the alkaline comet assay 7

J. Kain et al.

Fig. 7. Detection of oxidatively damaged DNA (FPG sites) in A549 cells by the comet assay, using FPG incubated with particles or ions (see Fig. 6 for procedure). The normal capacity of FPG to nd oxidatively damaged DNA is indicated by the dotted lines. (A) Particle/ion interference with FPG shows if the particles bind or inhibit the enzymatic function of FPG. (B) Loss of function of FPG due to particles (FPG particles tested without centrifugation). (C) Loss of function of FPG due to released ions from the particles. Each bar in (AC) represents a mean value of three to four independent experiments SD. Asterisks (*, ** and ***) indicate level of signicance when compared to the unexposed FPG (students t-test), corresponding to P , 0.05, 0.01 and 0.001, respectively. Lower amount of FPG sites detected (bars below the dotted line) indicate a decreased enzymatic capacity of FPG.

(measuring DNA breaks) is used, compared with the FPG version of the assay. This fact can obviously be due to low oxidative potential, but interaction within the method is also a possibility. The same trend has also been observed for certain metal ions. DNA damaging effects and lack of FPG sites have been observed for Cd(II) ions in non-cytotoxic concentrations, when the alkaline unwinding method was used (34). Again, this may be due to low oxidative capacity of the metal ions, but oxidative capacity of Cd(II) ions has been observed using other methods (35). The hypothesis that particle interaction with FPG may occur arises from the fact that particles have been observed in the comet nucleoid after particle exposure (18,19). Most likely, the particles get in contact with the nucleoid during the comet assay as previously discussed (17) and the presence of particles close to DNA during the comet assay procedure makes interaction with FPG a possible scenario. When FPG was incubated with particles in the present study, at an approximate intracellular dose [based on a study of Cu nanoparticles (22)], interference with FPG was observed for nanoparticles of Ag, CeO2, Co3O4 and SiO2. These results could be not only due to a decrease in the enzymatic activity but also to the fact that particles adsorbed the enzyme, which was then centrifuged away. A commonly used concept in nanotoxicology is the description of a protein corona, where proteins attach to the particle surface and form a corona around the nanoparticle (36). There is however a possibility that particle-bound enzyme still is active. To further investigate the enzymatic capacity of FPG when exposed to these particles, FPG was mixed with the particles and no centrifugation was performed. This experiment showed that FPG seemed to still be active when in contact with nanoparticles of SiO2, but had decreased enzymatic activity when mixed with the nanoparticles of Ag, CeO2 and Co3O4. 8

Particles (or ions from the particles) may inuence FPG to change conformation or interact with the structure, which is important for the substrate specicity of the enzyme. FPG is a zinc nger protein and it is the nger-like structure that wraps around DNA and enables the binding of the enzyme to DNA when repairing oxidatively damaged DNA, often 8oxoguanine. It has been shown that the active site of the nger structure is positively charged (37) with one Zn atom (38) anchored between four SH groups (39). This makes the active site a possible target of metal ions, which might change place with the Zn atom, necessary for the enzyme activity. Low doses of the metal ions Cd(II), Cu(II) and Hg(II) added to FPG have been observed to have a high afnity for the zinc nger, causing pronounced inhibition of the enzyme, in contrast to ions of As(III), Co(II), Ni(II) and Pb(II) (40). OConnor et al. (1993) (39) also showed that Cd(II), Cu(II) and Hg(II) ions compete for the Zn-binding site in FPG, whereas ions of Co(II), Fe(II), Mn(II) and Ni(II) did not. In contrast to this, another study showed that Ni(II) inhibited the repair capacity of FPG (34). Our study showed that the decreased activity of FPG detected after incubation of Ag nanoparticles mainly was explained by released Ag ions. FPG incubated with the released fraction from Ag nanoparticles (i.e. Ag ions) totally lost its ability to detect oxidatively damaged DNA in the comet assay. This disturbance is most likely due to the binding of Ag ions to the SH groups at the active site. Binding of Ag ions to SH groups in proteins is suggested to be the mechanism behind the well-known antibacterial effect of Ag (41). Another possible interaction of particles present in the nucleoid is the physical hindrance that these particles may cause, which can prevent the enzyme from reaching the damaged DNA. This might be an explanation to the decreased levels of FPG sites with increased concentration of steel

FPG interaction

particles (Figure 2C). Similar mechanism has been observed for some heavy metal ions, which inhibited the binding of the human DNA repair enzyme NEIL1 (42). It might be stressed that even though this study uses cell lines and not primary cells and that the amount of particles that may be found in the nucleoid is a crude estimation, it is likely that interaction of particles and/or ions released with FPG can occur. The results in this study suggest that oxidative stress caused by nanoparticles may be underestimated if the comet assay is used for analysis of oxidatively damaged DNA. In conclusion, this study highlights that particles may interact with FPG in the comet assay, leading to a decreased ability to detect oxidatively damaged DNA. Thus, this method may not be the most suitable to use for measurement of oxidatively damaged DNA from particles of all kinds, after exposure to cell lines. If the comet assay is used for this purpose, additionally, independent in vitro methods should preferably be used in parallel, which always is benecial when assessing toxicity (19). This study highlights that established, reliable genotoxicity tests for chemicals may not always be suitable for studying genotoxicity of particles.

Funding Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS 2010-968); the Swedish Research Council (Vetenskapsra det 2010-10-27); ngpannefo AFs (A reningens) Foundation for Research and Development.

Acknowledgements
Experimental help with the particle characterisation by Tao Jiang and Inger Odnevall Wallinder at the Division of Surface and Corrosion Science KTH and valuable help with the statistics and inspiring discussion about the comet assay with Clara Ersson at the Department of Biosciences and Nutrition at Karolinska Institutet is gratefully appreciated. The photosensitiser Ro 19-8022 was a generous gift from F. Hoffmann-La Roche, Basel, Switzerland. The authors are members of the Stockholm Particle Group, an operative network between three universities in Stockholm (Karolinska Institutet, The Royal Institute of Technology, and Stockholm University). The authors are partners of ECNIS (Environmental Cancer Risk, Nutrition and Individual Susceptibility), an EU network of excellence operating within the European Union sixth framework program, priority 5: Food Quality and Safety (contract no. 513943). Conict of interest statement: None declared.

References
1. Pope, C. A. and Dockery, D. W. (2006) Health effects of ne particulate air pollution: lines that connect. J. Air Waste Manag. Assoc., 56, 709742. 2. Donaldson, K., Stone, V., Borm, P. J. A. et al. (2003) Oxidative stress and calcium signaling in the adverse effects of environmental particles (PM10). Free Radic. Biol. Med., 34, 13691382. 3. Nel, A., Xia, T., Madler, L. and Li, N. (2006) Toxic potential of materials at the nanolevel. Science, 311, 622627. 4. Doak, S. H., Grifths, S. M., Manshian, B., Singh, N., Williams, P. M., Brown, A. P. and Jenkins, G. J. S. (2009) Confounding experimental considerations in nanogenotoxicology. Mutagenesis, 24, 285293. 5. Kroll, A., Pillukat, M. H., Hahn, D. and Schnekenburger, J. (2009) Current in vitro methods in nanoparticle risk assessment: Limitations and challenges. Eur. J. Pharm. Biopharm., 72, 370377. 6. Boiteux, S., Gajewski, E., Laval, J. and Dizdaroglu, M. (1992) Substratespecicity of the Escherichia-coli fpg protein (formamidopyrimidine DNA glycosylase)excision of purine lesions in DNA produced by ionizingradiation or photosensitization. Biochemistry, 31, 106110.

7. Kasai, H. (1997) Analysis of a form of oxidative DNA damage, 8-hydroxy2#-deoxyguanosine, as a marker of cellular oxidative stress during carcinogenesis. Mutat. Res., 387, 147163. 8. Matsui, A., Ikeda, T., Enomoto, K., Hosoda, K., Nakashima, H., Omae, K., Watanabe, M., Hibi, T. and Kitajima, M. (2000) Increased formation of oxidative DNA damage, 8-hydroxy-2#-deoxyguanosine, in human breast cancer tissue and its relationship to GSTP1 and COMT genotypes. Cancer Lett., 151, 8795. 9. Kondo, S., Toyokuni, S., Iwasa, Y., Tanaka, T., Onodera, H., Hiai, H. and Imamura, M. (1999) Persistent oxidative stress in human colorectal carcinoma, but not in adenoma. Free Radic. Biol. Med., 27, 401410. 10. Bra uner, E. V., Forchhammer, L., Mller, P., Simonsen, J., Glasius, M., Wa hlin, P., Raaschou-Nielsen, O. and Loft, S. (2007) Exposure to ultrane particles from ambient air and oxidative stress-induced DNA damage. Environ. Health Perspect., 115, 11771182. 11. Avogbe, P. H., Ayi-Fanou, L., Autrup, H., Loft, S., Fayomi, B., Sanni, A., Vinzents, P. and Mller, P. (2005) Ultrane particulate matter and highlevel benzene urban air pollution in relation to oxidative DNA damage. Carcinogenesis, 26, 613620. 12. Collins, A., Gedik, C., Vaughan, N. et al. (2003) Measurement of DNA oxidation in human cells by chromatographic and enzymic methods. Free Radic. Biol. Med., 34, 10891099. 13. Johansson, C., Mller, P., Forchhammer, L. et al. (2010) An ECVAG trial on assessment of oxidative damage to DNA measured by the comet assay. Mutagenesis, 25, 125132. 14. Forchhammer, L., Johansson, C., Loft, S. et al. (2010) Variation in the measurement of DNA damage by comet assay measured by the ECVAG inter-laboratory validation trial. Mutagenesis, 25, 113123. 15. Azqueta, A., Gutzkow, K. B., Brunborg, G. and Collins, A. R. (2011) Towards a more reliable comet assay: Optimising agarose concentration, unwinding time and electrophoresis conditions. Mutat. Res., 724, 4145. 16. Ersson, C. and Mo ller, L. (2011) The effects on DNA migration of altering parameters in the comet assay protocol such as agarose density, electrophoresis conditions, and durations of the enzyme or the alkaline treatments. Mutagenesis, 26, 689695. 17. Karlsson, H. L. (2010) The comet assay in nanotoxicology research. Anal. Bioanal. Chem., 398, 651666. 18. Karlsson, H. L., Nygren, J. and Mo ller, L. (2004) Genotoxicity of airborne particulate matter: the role of cell-particle interaction and of substances with adduct-forming and oxidizing capacity. Mutat. Res., 565, 110. 19. Stone, V., Johnston, H. and Schins, R. P. F. (2009) Development of in vitro systems for nanotoxicology: methodological considerations. Crit. Rev. Toxicol., 39, 613626. 20. Lin, M. H., Hsu, T. S., Yang, P. M., Tsai, M. Y., Perng, T. P. and Lin, L. Y. (2009) Comparison of organic and inorganic germanium compounds in cellular radiosensitivity and preparation of germanium nanoparticles as a radiosensitizer. Int. J. Radiat. Biol., 85, 214226. 21. Will, O., Gocke, E., Eckert, I., Schulz, I., Pum, M., Mahler, H. C. and Epe, B. (1999) Oxidative DNA damage and mutations induced by a polar photosensitizer, Ro19-8022. Mutat. Res., 435, 89101. 22. Cronholm, P., Midander, K., Karlsson, H. L., Elihn, K., Wallinder, I. O. and Moller, L. (2011) Effect of sonication and serum proteins on copper release from copper nanoparticles and the toxicity towards lung epithelial cells. Nanotoxicology, 5, 269281. 23. Hedberg, Y., Gustafsson, J., Karlsson, H. L., Mo ller, L. and Wallinder, I. O. (2010) Bioaccessibility, bioavailability and toxicity of commercially relevant iron- and chromium-based particles: in vitro studies with an inhalation perspective. Part. Fibre Toxicol., 7, 23. 24. Choi, J. Y., Ramachandran, G. and Kandlikar, M. (2009) The Impact of toxicity testing costs on nanomaterial Regulation. Environ. Sci. Technol., 43, 30303034. 25. Gee, D. and Greenberg, M. (2002) Asbestos from magic to malevolent mineral, late lesions from early warnings: the precautionary principle 1896 2000. European environment agency (EEA). Environmental Issue Report No. 22, 5263. 26. Speit, G., Schutz, P., Bonzheim, I., Trenz, K. and Hoffmann, H. (2004) Sensitivity of the FPG protein towards alkylation damage in the comet assay. Toxicol. Lett., 146, 151158. 27. Lu, S. L., Dufn, R., Poland, C., Daly, P., Murphy, F., Drost, E., MacNee, W., Stone, V. and Donaldson, K. (2009) Efcacy of simple shortterm in vitro assays for predicting the potential of metal oxide nanoparticles to cause pulmonary inammation. Environ. Health Perspect., 117, 241247. 28. Schrand, A. M., Rahman, M. F., Hussain, S. M., Schlager, J. J., Smith, D. A. and Syed, A. F. (2010) Metal-based nanoparticles and their toxicity assessment. Wiley Interdiscip Rev. Nanomed. Nanobiotechnol., 2, 544568.

J. Kain et al. 29. Karlsson, H. L., Gustafsson, J., Cronholm, P. and Mo ller, L. (2009) Sizedependent toxicity of metal oxide particlesa comparison between nanoand micrometer size. Toxicol. Lett., 188, 112118. 30. Karlsson, H. L., Cronholm, P., Gustafsson, J. and Mo ller, L. (2008) Copper oxide nanoparticles are highly toxic: a comparison between metal oxide nanoparticles and carbon nanotubes. Chem. Res. Toxicol., 21, 17261732. 31. Karlsson, H. L., Nilsson, L. and Mo ller, L. (2005) Subway particles are more genotoxic than street particles and induce oxidative stress in cultured human lung cells. Chem. Res. Toxicol., 18, 1923. 32. Karlsson, H. L., Holgersson, A. and Mo ller, L. (2008) Mechanisms related to the genotoxicity of particles in the subway and from other sources. Chem. Res. Toxicol., 21, 726731. 33. Hussain, S. M., Hess, K. L., Gearhart, J. M., Geiss, K. T. and Schlager, J. J. (2005) In vitro toxicity of nanoparticles in BRL 3A rat liver cells.. Toxicol. In Vitro, 19, 975983. 34. Dally, H. and Hartwig, A. (1997) Induction and repair inhibition of oxidative DNA damage by nickel(II) and cadmium(II) in mammalian cells. Carcinogenesis, 18, 10211026. 35. Filipic, M., Fatur, T. and Vudrag, M. (2006) Molecular mechanisms of cadmium induced mutagenicity. Hum. Exp. Toxicol., 25, 6777. 36. Lynch, I. and Dawson, K. A. (2008) Protein-nanoparticle interactions. Nano Today, 3, 4047. 37. Gilboa, R., Zharkov, D. O., Golan, G., Fernandes, A. S., Gerchman, S. E., Matz, E., Kycia, J. H., Grollman, A. P. and Shoham, G. (2002) Structure of formamidopyrimidine-DNA glycosylase covalently complexed to DNA. J. Biol. Chem., 277, 1981119816. 38. Boiteux, S., Oconnor, T. R., Lederer, F., Gouyette, A. and Laval, J. (1990) Homogeneous Escherichia coli FPG proteina DNA glycosylase which excises imidazole ring-opened purines and nicks DNA at apurinic apyrimidinic sites. J. Biol. Chem., 265, 39163922. 39. OConnor, T. R., Graves, R. J., Demurcia, G., Castaing, B. and Laval, J. (1993) FPG protein of Escherichia coli is a zinc nger protein whose cysteine residues have a structural and or functional role. J. Biol. Chem., 268, 90639070. 40. Asmuss, M., Mullenders, L. H. F., Eker, A. and Hartwig, A. (2000) Differential effects of toxic metal compounds on the activities of FPG and XPA, two zinc nger proteins involved in DNA repair. Carcinogenesis, 21, 20972104. 41. Liau, S. Y., Read, D. C., Pugh, W. J., Furr, J. R. and Russell, A. D. (1997) Interaction of silver nitrate with readily identiable groups: relationship to the antibacterial action of silver ions. Lett. Appl. Microbiol., 25, 279283. 42. Grin, I. R., Konorovsky, P. G., Nevinsky, G. A. and Zharkov, D. O. (2009) Heavy metal ions affect the activity of DNA glycosylases of the FPG family. Biochemistry (Mosc.), 74, 12531259.

10