Differential effects of toxic metal compounds on the ... - Semantic Scholar

Download PDF
1 downloads 0 Views 308KB Size Report
Comment
metals nickel, cadmium, cobalt and arsenic have been a variety of environmental mutagens including UV light, shown previously to inhibit nucleotide excision ...
Carcinogenesis vol.21 no.11 pp.2097–2104, 2000

Differential effects of toxic metal compounds on the activities of Fpg and XPA, two zinc finger proteins involved in DNA repair

Monika Asmuss1, Leon H.F.Mullenders2, Andre´ Eker3 and Andrea Hartwig1,4 1University

of Karlsruhe, Department of Food Chemistry, D-76128 Karlsruhe, Germany, 2Leiden University, MGC Department of Radiation Genetics and Chemical Mutagenesis, Leiden and 3Erasmus University Rotterdam, Department of Cell Biology and Genetics, Rotterdam, The Netherlands 4To

whom correspondence should be addressed Email: [email protected]

Even though not mutagenic, compounds of the carcinogenic metals nickel, cadmium, cobalt and arsenic have been shown previously to inhibit nucleotide excision repair and base excision repair at low, non-cytotoxic concentrations. Since some toxic metals have high affinities for –SH groups, we used the bacterial formamidopyrimidine-DNA glycosylase (Fpg protein) and the mammalian XPA protein as models to investigate whether zinc finger structures in DNA repair enzymes are particularly sensitive to carcinogenic and/or toxic metal compounds. Concentrations of ≤1 mM Ni(II), Pb(II), As(III) or Co(II) did not affect the activity of the Fpg protein significantly. In contrast, the enzyme was inhibited in a dose-dependent manner by Cd(II), Cu(II) or Hg(II), starting at concentrations of 50 µM, 5 µM and 50 nM, respectively. Simultaneous treatment with Cd(II) or Cu(II) and Zn(II) partly prevented the inhibitions, while no reversal of inhibition was observed when Zn(II) was added after Cd(II) or Cu(II). In the case of Hg(II), Zn(II) had no protective effect independent of the time of its addition; however, the enzyme activity was completely restored by glutathione. Regarding XPA, Hg(II), Pb(II) or As(III) did not diminish its binding to an UVirradiated oligonucleotide, while Cd(II), Co(II), Cu(II) and Ni(II) reduced its DNA-binding ability. Simultaneous treatment with Zn(II) prevented largely the inhibition induced by Cd(II), Co(II), and Ni(II), but only slightly in the case of Cu(II). Collectively, the results indicate that both proteins were inhibited by Cd(II) and Cu(II), XPA was additionally inactivated by Ni(II) and Co(II), and Fpg but not XPA was strongly affected by Hg(II). Even though other mechanisms of protein inactivation cannot be completely excluded, zinc finger structures may be sensitive targets for toxic metal compounds, but each zinc finger protein has unique sensitivities. Introduction Toxic metal compounds are widely distributed in the environment and are used in many industrial processes. As a result of their long persistence in biological systems and their tendency to accumulate in certain tissues, they are important Abbreviations: BSA, bovine serum albumin; DTT, dithiothreitol; Fpg, formamidopyrimidine-DNA glycosylase; NBT, nitroblue tetrazolium chloride; NER, nucleotide excision repair. © Oxford University Press

environmental hazards. From epidemiological studies and/or animal experiments it is well known that compounds of chromium, arsenic, cadmium, nickel and cobalt are carcinogenic, but—with the exception of chromate—they are only weakly mutagenic in bacterial test systems and in mammalian cells in culture. Nevertheless, these metal compounds enhance the genotoxic effects of different mutagens such as UVC radiation, X-rays, benzo[a]pyrene, cis-diamminedichloroplatinum(II) (cisplatin) or DNA alkylating agents (1) and inhibit DNA repair processes. Nucleotide excision repair (NER), which is involved in the removal of DNA damage induced by a variety of environmental mutagens including UV light, aromatic amines and polyaromatic hydrocarbons, is inhibited by Ni(II), Cd(II), Co(II) and As(III). Various steps of the repair process can be affected and various mechanisms of action have been identified. All metals reduce the frequency of incision events; in the case of Cd(II) and Ni(II), DNA– protein interactions involved in DNA damage recognition are diminished (2). In addition, Co(II) interferes with the polymerization step and As(III) with the ligation step (2). Besides NER, the repair of oxidative DNA base modifications is blocked by Ni(II) and Cd(II) in HeLa cells (3). Finally, the O6-methylguanine methyltransferase (MGMT) is inactivated by Ni(II) in HeLa cells (4). The underlying molecular mechanisms of repair inhibition by metals are still unknown, but competition of toxic metals with essential ions appears to be one important mechanism. Since metal ions are cofactors in many cellular processes and since all repair inhibitions have been observed at low, non-cytotoxic concentrations of the respective metal compounds, why are DNA repair systems sensitive towards toxic metal ions? Some toxic metal ions have high affinities for –SH groups, so zinc finger proteins could be targets of such ions. The zinc finger proteins comprise a family of proteins where zinc is complexed through four invariant cysteine and/or histidine residues forming a zinc finger domain, which is involved mostly in DNA binding, but also in protein–protein interactions (5). Even though most zinc finger structures have been described as DNA-binding motifs in transcription factors, they have also been identified in several DNA repair enzymes, including the bacterial UvrA protein involved in DNA damage recognition during NER, the bacterial formamidopyrimidine-DNA glycosylase (Fpg) protein involved in the removal of oxidative DNA base modifications and the mammalian repair proteins XPA, DNA ligase III and poly(ADP-ribose) polymerase (PARP). The effects of toxic metal ions on the activity of zinc finger transcription factors have been studied. Cu(II), Cd(II), Co(II) and Ni(II), when used to replace zinc in the bovine oestrogen receptor and in the transcription factors SP1 and IIIA, do not all have the same effect on the DNA-binding properties of the protein and protein function. For example, bovine oestrogen receptor apoprotein reconstituted with Cd(II) or Co(II) was able to bind DNA, whereas that reconstituted with Cu(II) and Ni(II) was not, even though these ions can bind to zinc finger residues (6–9). 2097

M.Asmuss et al.

Almost nothing is known about the interaction of toxic metal ions with zinc finger DNA repair enzymes and the functional implications. O’Connor et al. (10) reported the displacement of radioactive zinc in the Fpg protein by 100 µM of Hg(II), Cu(II) or Cd(II), but no functional analysis has been conducted. Therefore, in the present study we used the bacterial Fpg protein and the mammalian XPA protein as models for zinc finger repair proteins and examined their activity in the presence of Cd(II), Ni(II), Co(II), Cu(II), Pb(II), Hg(II) or As(III). Fpg is a glycosylase initiating base excision repair in Escherichia coli. It recognizes and removes 7,8-dihydro-8oxoguanine (8-oxoguanine), the imidazole ring-opened purines 2,6-diamino-4-hydroxy-5-formamidopyrimidine (Fapy-Gua) and 4,6-diamino-5-formamidopyrimidine (Fapy-Ade) as well as small amounts of 7,8-dihydro-8-oxoadenine (8-oxoadenine). Among these DNA base modifications, 8-oxoguanine has the highest affinity and, due to its mutagenic potential, is also believed to be the biologically most relevant substrate. Fpg combines the function of a glycosylase, an AP-lyase and a 5⬘-terminal deoxyribose phosphate-excising activity, thus converting the DNA base damage into single-strand breaks (11,12). DNA binding is mediated by a single zinc finger domain in the C-terminal region, where zinc is complexed by four cysteines (Cys244, Cys247, Cys264 and Cys267). Substitution of any cysteine in the zinc finger destroys DNA-binding capacity and enzyme function as a whole (10). XPA consists of 273 amino acids and plays a central role in the first steps of NER. Loss of XPA function leads to xeroderma pigmentosum type A, a severe disorder characterized by UV-hypersensitivity and enhanced cancer risk. The protein contains specific binding sites for other NER proteins such as ERCC1, TFIIH and RPA, and has been proposed to coordinate these factors in the pre-incision complex of NER (13). XPA binds specifically to damaged DNA, including lesions induced by UVC, benzo[a]pyrene or cis-platinum (14–16); its binding affinity is enhanced by the RPA protein (17). Whether the binding of XPA or the heterodimer XPA/RPA to damaged DNA is preceded by binding of XPC/ HR23B is controversial (18,19). XPA contains a single zinc finger motif (20) which is part of the minimal binding domain (MBD) necessary for DNA binding. In the zinc-binding core (Asp101–Lys137), zinc is complexed by Cys105, Cys108, Cys126 and Cys129 (21). Substitution of any of these cysteines leads to a severe reduction in NER activity (22).

(37%) were from Merck (Darmstadt, Germany). The Fpg protein was kindly given by Dr Serge Boiteux, France. PM2 DNA We used PM2 DNA as a substrate for the Fpg protein. PM2 is a bacteriophage with a circular DNA of 10 kb. PM2 DNA was prepared according to the method of Salditt et al. (23) with a few alterations. Briefly, PM2 bacteriophage was amplified in Alteromonas espejiana by adding 1010 phages to each 1 l of bacterial suspension (OD600 nm ⫽ 0.6). After full lysis of the bacteria, PM2 phages were precipitated for 艌18 h with polyethylene glycol (43 g/l suspension) and sodium dextran sulfate (2.36 g/l suspension) at 4°C on ice. The sediment was resolved and washed in 2 vols NTC buffer (1 M NaCl, 16 mM Tris– HCl, 10 mM CaCl2, pH 7.5). To remove bacterial cell remains, the sediment was homogenized twice with 30 ml NTC using a Potter homogenisator and centrifuged for 20 min at 12 000⫻g. Subsequently, the supernatant containing the PM2 phages was centrifuged for 3 h at 30 000⫻g. The pellets were homogenized again in 20 ml NTC buffer before the addition of 0.39 g CsCl2 per ml, resulting in a solution density of 1.27–1.29 g/ml. After centrifugation for 艌12 h at 35 000⫻g, PM2 phages appeared as a dense, opal-coloured band in the upper third of the gradient. The phage fraction was removed using a syringe, loaded on to an agarose column and eluted with NTC buffer. The DNA was isolated by phenol–chloroform extraction in the presence of 1% sodium dodecyl sulfate in the first step. After precipitation with ethanol/ 125 mM sodium acetate, PM2 DNA was dissolved and stored in buffer containing 100 mM NaCl, 20 mM Tris–HCl, 1 mM EDTA, pH 7.5. DNA prepared by this procedure retained ~90% supercoiled molecules. Linear DNA fragments were not detected. Induction of DNA damage PM2 DNA (20 µg/ml) dissolved in enzyme buffer (50 mM sodium phosphate buffer, 100 mM NaCl, pH 7.5) was oxidatively damaged by addition of the photoreactive thiazin dye methylene blue (final concentration 10 µg/ml) and subsequently irradiated with visible light (40 s, 60 W, 50 cm distance, resulting in 216 J/m2). Afterwards the PM2 DNA was precipitated with ethanol/125 mM sodium acetate for 艌30 min and centrifuged for 4 min at 12 000 r.p.m. (8000⫻g). The supernatant was removed carefully and the DNA pellet resuspended in enzyme buffer. Damage induction and all subsequent steps were carried out in the dark to prevent additional DNA damage. Detection and quantification of enzyme activity Oxidatively damaged PM2 DNA (10 µl; 200 ng per sample) and Fpg protein (final concentration 1 µg/ml, 30 µl per sample) were incubated for 30 min at 37°C. The reaction was terminated by adding 7 µl stop solution (0.25% bromophenol blue, 0.25% xylene cyanol, 15% Ficoll 400). When investigating effects of metal treatment, the Fpg protein was preincubated with various concentrations of Cd(II), Hg(II), Ni(II), Cu(II), Co(II), Pb(II) or As(III) for the times indicated in the legends to the respective figures. Supercoiled and closed circular forms of the PM2 molecules were separated by electrophoresis in a 1% agarose gel in buffer (890 mM Trizma base, 890 mM boric acid, 10 mM EDTA) for 3 h at 90 V and stained with ethidium bromide. The density of the bands was measured using a Herolab gel detection system (EASY win 32). For calculation of break frequencies, a Poisson distribution was assumed and a correction factor of 1.4 was applied to compensate for the relative lower fluorescence of the supercoiled form (24): N ⫽ – ln [(1.4I)/(1.4I ⫹ II)]

Materials and methods

where N is the number of strand breaks/molecule PM2, I is the percentage of supercoiled PM2 DNA and II is the percentage of open circular PM2 DNA. The overall number of strand breaks per 10 000 base pairs represents the sum of single-strand breaks and incisions generated by the repair enzyme.

Materials 5-Bromo-4-chloro-3-indolyl-phosphate (X-phosphate), nitroblue tetrazolium chloride (NBT), anti-digoxigenin–Fab fragments and the blocking reagent were from Boehringer (Mannheim, Germany). Ethidium bromide, sodium chloride, xylene cyanol FF, cupric(II) sulfate, sodium(meta) arsenite, cadmium(II) chloride and HEPES were purchased from Fluka Chemie (Buchs, Germany). Ficoll 400 and dextran sulfate came from Pharmacia (Uppsala, Sweden). Maleic acid, Tween 20, dithiothreitol (DTT), TEMED, glycerol (99%) and acrylamide–bisacrylamide solution (37.5:1; 40%) were obtained from Serva (Heidelberg, Germany) and Trizma base, polyethylene glycol 6000, potassium chloride and bovine serum albumin (BSA) from Sigma– Aldrich (Deisenhofen, Germany). Agarose type II and medium EEO were from Sigma Chemical Co. (St Louis, MO). All other chemicals, including bromphenol blue, absolute ethanol, Na2HPO4, NaH2PO4·xH2O, methylene blue, magnesium(II) chloride, sodium acetate, boric acid, EDTA, mercury(II) chloride, nickel(II) chloride, cobalt(II) chloride, lead(II) acetate, zinc(II) chloride, bromphenol blue–sodium salt, ammonium peroxodisulfate (APS) and HCl

Isolation and DNA binding activity of XPA Recombinant mouse XPA protein was purified as described previously (18,25). The DNA-binding activity of the XPA protein was determined by gel mobility shift experiments using a digoxygenin end-labelled synthetic oligonucleotide (70 bp; MWG Biotech, Ebersberg, Germany) with the following sequence together with its complementary strand: 5⬘-ATATGTGCACATGGCGCACGTATGTATCTATAGTCTGCCATCACGCCAGTCAATCGCTGTGGTATATGCA-3⬘. XPA (500 ng) was pretreated with metal compounds where indicated in a gel shift buffer (final concentration: 25 mM HEPES–KOH, 10% glycerol, 30 mM KCl, 4 mM MgCl2, 1 mM EDTA, 45 µg/ml BSA, 0.9 mM DTT, pH 8.3) for 15 min at room temperature. Afterwards, 240 fmol of the digoxigeninlabelled oligonucleotide either unirradiated or irradiated with 18 kJ/m2 UVC (254 nm germicidal lamp; Bioblock Scientific, VL-6.C) were added for 30 min at room temperature in the dark. The binding mixture was loaded on a 5% polyacrylamide gel (37.5 acrylamide:1.0 bisacrylamide; 45 mM Tris–HCl, 45 mM boric acid, 1 mM EDTA, pH 8.0) and electrophoresis was conducted

2098

Effects of toxic metal compounds on Fpg and XPA

Fig. 2. Effect of Ni(II), Pb(II), Co(II) and As(III) on the activity of the Fpg protein. The protein was incubated for 30 min at 37°C with the respective metal compounds at the indicated concentrations. The activity of the Fpg protein was determined on methylene blue-damaged PM2 DNA irradiated with visible light for 40 s. Shown are mean values of at least three determinations ⫾ SD.

Fig. 1. Induction of Fpg-sensitive sites and DNA strand breaks in PM2 DNA by methylene blue plus visible light. PM2 DNA was irradiated with visible light (60 W, at a distance of 50 cm) for the indicated times in the presence of 30 µM methylene blue. Subsequently, enzyme buffer with or without Fpg protein was added for 30 min at 37°C. (A) Supercoiled (sc) and closed circular (cc) PM2 DNA as detected by agarose gel electrophoresis. (B) Quantification of lesion frequencies as described in Materials and methods. Shown are mean values of at least three determinations ⫾ SD. at 90 V for 1.25 h. Southern blotting was done in a semi-dry electroblotting apparatus using a positively charged nylon membrane (Hybond-N⫹; Amersham, Braunschweig, Germany), followed by fixation for 1.5 h at 90°C. Digoxigenin-labelled oligonucleotide was detected colorimetrically by alkaline phosphatase conjugated to an anti-digoxigenin antibody using NBT and X-phosphate as substrates. The bands were quantified with a Herolab detection system.

Results Effect of metal compounds on the bacterial Fpg protein A test system was first established to measure the activity of the Fpg protein in the absence and presence of toxic metal compounds. We isolated and applied DNA from bacteriophage PM2, which consists of 10 kb and which was ~90% preserved in its supercoiled form after isolation. Induction of DNA strand breaks, either directly or by incisions at sites of oxidative DNA base modifications by the Fpg protein, converts the supercoiled PM2 molecule into the open circular form; the two forms were separated electrophoretically and the frequency of DNA strand breaks was determined as described in Materials and methods. The background level of DNA strand breaks after PM2 preparation did not exceed 0.15 breaks/10 000 bp (mean 0.12 ⫾ 0.02 breaks/10 000 bp). Values for Fpg-sensitive sites in undamaged DNA were ⬍0.07 breaks/10 000 bp (mean 0.05 ⫾ 0.016 breaks/10 000 bp). To test the activity of the Fpg protein on an oxidatively damaged template, PM2 DNA was modified with methylene blue plus visible light, which has been shown to yield 8-oxoguanine (26) and small amounts of Fapy-Gua (12), presumably due to the generation of singlet oxygen. The result is shown in Figure 1. Treatment of the DNA with methylene blue in the absence of visible light only slightly increased the frequency of these two types of DNA

lesions. However, when the modified DNA was irradiated with visible light, the frequency of Fpg-sensitive modifications increased in a dose-dependent manner. The low frequency of DNA strand breaks as compared with the high number of Fpgsensitive sites confirms results obtained by other authors; it resembles the damage distribution of both types of lesions induced by singlet oxygen as opposed to OH•– radicals (27,28). When PM2 DNA was exposed to Cd(II), Ni(II), Cu(II), Co(II), Hg(II), Pb(II) or As(III), only Cu(II) increased the basic rate of strand breaks in PM2 at concentrations of ⬎10 µM. None of the other metal compounds showed an effect at concentrations of 艋1 mM (data not shown). In subsequent experiments, we used methylene blue-modified DNA irradiated for 40 s as substrate and tested the enzyme activity after 30 min preincubation with different concentrations of Cd(II), Ni(II), Cu(II), Co(II), Hg(II), Pb(II) or As(III). As shown in Figure 2, Ni(II), Pb(II), As(III) or Co(II) at 艋1 mM did not affect the activity of Fpg protein significantly. With respect to arsenic, its retention in the trivalent form appears to be crucial (29); thus, the experiment has been repeated in the presence of 1 or 5 mM β-mercaptoethanol. Again, no inhibition was observed (data not shown). In contrast, Cd(II), Cu(II) and Hg(II) inhibited the enzyme nearly completely. The lowest concentration of Cd(II) that was inhibitory was 50 µM, and at 1 mM there was only 3% of the activity in the control. Cu(II) decreased the enzyme activity to 40% of control level at 5 µM and almost complete inhibition was observed at 10 µM. Of all the metals tested, Hg(II) had the strongest inhibitory effect on the Fpg protein; complete inhibition of activity was seen at 50 nM Hg(II) (Figure 3). To find out whether this inhibition was due to displacement of zinc in the zinc finger structure, we tested whether simultaneous or subsequent incubation with Zn(II) would prevent the enzyme from being inhibited. Zn(II) alone did not modulate the enzyme activity at concentrations up to 1 mM (data not shown). With respect to the effect of Cd(II), after simultaneous incubation with equimolar concentrations of Cd(II) and Zn(II) the inhibition was less pronounced, giving ~60% of the control enzyme activity, with no further increase seen at 500 µM Zn(II). However, when the Fpg protein was incubated with Cd(II) before the addition of Zn(II), no reversal of inactivation 2099

M.Asmuss et al.

Fig. 3. Effect of Cd(II), Cu(II) and Hg(II) on the activity of Fpg protein on methylene blue-damaged PM2 DNA. The protein was incubated with the respective metal compounds for 30 min at 37°C, at the indicated concentrations. Shown are mean values of at least three determinations ⫾ SD.

was observed (Figure 4A). Regarding Cu(II), the simultaneous incubation with an equimolar concentration of Zn(II) had no effect, but the enzyme activity was restored to ~40% at 100 and 500 µM Zn(II). When the enzyme was incubated with Cu(II) before adding Zn(II), no substantial reversal of inhibition took place (Figure 4B). In contrast to the results obtained with Cd(II) and Cu(II), inhibition by Hg(II) was not affected by simultaneous treatment with Zn(II) (data not shown). To elucidate further the involvement of –SH groups in enzyme inactivation by the metal ions, we investigated whether increasing concentrations of reduced glutathione during the last 15 min of metal treatment would alter the degree of enzyme inhibition by Cd(II), Cu(II) or Hg(II). As demonstrated in Figure 5, the reversal of enzyme inhibition was most pronounced in the case of Hg(II), where an equimolar concentration of glutathione restored the Fpg activity to 70%, with a further increase at higher concentrations. With respect to Cd(II), the Fpg activity increased to ~50% when adding an equimolar concentration of glutathione; a 10-fold higher glutathione concentration led to a complete reversal of enzyme inhibition. In contrast, even a 1000-fold excess of glutathione had no effect on the enzyme inhibition by Cu(II). Effect of metal compounds on the mammalian XPA protein The activity of the XPA protein was investigated by its ability to bind to an undamaged or UVC-damaged oligonucleotide by gel mobility shift analysis. In the absence of toxic metal compounds, XPA bound to both oligonucleotides. Only with small amounts of XPA (艋300 ng), a moderately preferential binding to damaged DNA was observed. At higher concentrations, no difference was seen between the damaged and undamaged oligonucleotide, which is in agreement with results obtained by Wakasugi and Sancar (19). Due to the sharper band obtained with 500 ng XPA, this amount was used for the subsequent experiments, and both specific and non-specific binding of XPA in the presence of different toxic metal ions was analysed. However, no differences were observed between both oligonucleotides for any metal compound, so in the following figures only the results obtained with the UVCdamaged oligonucleotide are shown. As demonstrated in Figure 6, preincubation of XPA with either Hg(II), Pb(II) or As(III) had no effect on the binding 2100

Fig. 4. Effect of Zn(II) on the inhibitory effects of Cd(II) (A) or Cu(II) (B) towards the Fpg protein. Shown are enzyme activities on methylene bluedamaged PM2 DNA after 30 min simultaneous incubation of Fpg with Cd(II) or Cu(II) and Zn(II) at 37°C or after 15 min incubation with Cd(II) or Cu(II), whereafter Zn(II) was added and incubated for a further 15 min at 37°C. Shown are mean values of at least three determinations ⫾ SD.

Fig. 5. Effect of glutathione on Fpg inhibition by Cd(II), Cu(II) or Hg(II). Enzyme activity was determined on methylene blue-damaged PM2 DNA after 30 min incubation with the respective metal compounds at 37°C with different concentrations of glutathione added during the last 15 min where indicated. Shown are mean values of at least three determinations ⫾ SD.

Effects of toxic metal compounds on Fpg and XPA

Fig. 6. Effect of As(III), Pb(II) or Hg(II) on the DNA-binding activity of XPA to a UVC-irradiated oligonucleotide. XPA protein was incubated with the respective metal compounds for 15 min at room temperature and its binding activity was analysed by gel mobility shift assay as described in Materials and methods. Shown are mean values of at least three determinations ⫾ SD.

behaviour of XPA. To ensure that As(III) is present in its trivalent form and not complexed by DTT present in the gel shift buffer, the experiment was also performed with βmercaptoethanol in the reaction mixture; again, no inhibitory effect was observed (data not shown). In contrast, Cd(II), Co(II), Cu(II) and Ni(II) diminished the binding activity of XPA to the UVC-damaged oligonucleotide (Figure 7). The strongest effect was seen with Co(II): the binding started to decrease at 50 µM and was completely inhibited at 200 µM. In the case of Cd(II), the shape of the band started to fade at 100 µM (Figure 7A); pronounced, almost complete inhibition was seen at 200 µM. At 200 µM Cu(II) and Ni(II), 40% and 90% residual binding activities were observed, respectively; at 500 µM, complete inhibition took place (Figure 7B). The effect of Zn(II) on the observed metal-induced inhibition was investigated next. Since an excess of Zn(II) may be inhibitory itself, the effect of Zn(II) on the binding behaviour of XPA was analysed first. Up to 2 mM Zn(II) did not alter DNA binding; at higher concentrations, the binding decreased slightly (data not shown). Thus, in the following experiments, only Zn(II) concentrations of 艋2 mM were applied. Simultaneous treatment of XPA with Cd(II), Co(II) or Ni(II) with Zn(II) inhibited XPA with differing efficiencies. With respect to Cd(II), an equimolar concentration of Zn(II) was sufficient to achieve 90% binding activity. Regarding Co(II) and Ni(II), a 5-fold or 3.3-fold excess, respectively, was needed for partial XPA binding; at 2 mM Zn(II), DNA binding of XPA was completely restored (Figure 8). In the case of Cu(II), no significant change in XPA inhibition was seen after simultaneous treatment with Cu(II) and Zn(II) as compared with Cu(II) alone; only at 2 mM Zn(II) plus 200 µM Cu(II) was a slight increase in XPA binding activity, from 40% to 59%, observed (data not shown). Discussion The results presented in this study illustrate for the first time that zinc finger DNA repair proteins may be inactivated by toxic metals like Cd(II), Hg(II), Co(II), Cu(II) and Ni(II), but the effects depend strongly on the actual repair factor

Fig. 7. Effect of Cd(II), Co(II), Cu(II) and Ni(II) on the DNA binding activity of XPA to a UVC-irradiated oligonucleotide. XPA protein was incubated with the respective metal compounds for 15 min at room temperature and its binding activity was analysed by gel mobility shift assay as described in Materials and methods. (A) Alterations in XPA-induced gel mobility shift after preincubation with Cd(II); (B) quantification of XPA binding derived after preincubation with Cd(II), Co(II), Cu(II) and Ni(II). *UVC-irradiated oligonucleotide. Shown are mean values of at least three determinations ⫾ SD.

investigated. Cd(II), Cu(II) and Hg(II), but not Pb(II), Co(II), Ni(II) or As(III), inhibited the catalytic activity of the bacterial Fpg protein. The metal ions varied in terms of the concentrations required for inhibition and their characteristics in competition experiments. The strongest inhibition was observed in the presence of Hg(II), followed by Cu(II) and Cd(II). With respect to Cd(II), the enzyme activity was inhibited at 艌50 µM. This inhibition was partly prevented when simultaneously incubating with equimolar concentrations of Zn(II), suggesting that competition between these elements is important. However, when the enzyme was preincubated with Cd(II), the inhibition could not be reversed by addition of a 5-fold excess of Zn(II), indicating that Cd(II) is complexed with a higher affinity than Zn(II). Partial reversibility was also observed when adding equimolar concentrations of glutathione and complete reactivation occurred at a 10-fold molar excess of glutathione, suggesting that Cd(II) is complexed to glutathione with similar strength as in the zinc finger structure. In the case of Cu(II), pronounced inhibition was observed at 艌5 µM, which was 2101

M.Asmuss et al.

Fig. 8. Effect of Zn(II) on the diminished DNA binding of XPA in the presence of Cd(II), Co(II) or Ni(II). XPA protein was incubated simultaneously with the respective toxic metal compounds and Zn(II) for 15 min at room temperature and its binding activity was analysed by gel mobility shift assay as described in Materials and methods. Shown are mean values of at least three determinations ⫾ SD.

only partly prevented when simultaneously treating with an 20-fold excess of Zn(II) and essentially irreversible even at a 100-fold excess of Zn(II) once the enzyme had been preincubated with Cu(II). In contrast to the results obtained with Cd(II), the copper-induced inhibition could not be reversed by the addition of glutathione, indicating a high affinity for the zinc finger structure. Displacement of Zn(II) by Cu(II) in the Fpg protein was also shown by O’Connor et al. (10); it cannot be excluded, however, that other mechanisms account for the copper-induced enzyme inhibition. Copper catalyses the formation of reactive oxygen species and thus the oxidation of cellular macromolecules. Even though, in our experiments, direct DNA damage by Cu(II) was only observed at concentrations of ⬎10 µM, Cu(II) bound to specific sites on proteins may have greater prooxidative activity (30), which could in turn lead to the oxidation of amino acids inside or outside the zinc finger (31). Hg(II) inactivated the enzyme at concentrations of 艌10 nM, even in the presence of a 1000-fold excess of Zn(II). Yet, substantial reactivation was seen at equimolar concentrations of reduced glutathione, pointing towards the involvement of –SH groups in enzyme inhibition. Several explanations could account for this observation. First, the main target for Hg(II) may not be Zn(II) in the zinc finger structure but rather –SH groups in amino acids outside the zinc finger structure. However, if there were other –SH groups outside the zinc finger readily accessible to metal ions, Pb(II) would be expected to inhibit the enzyme as well. Taking into account the displacement of zinc by mercury shown before (10), one other possible explanation consists in an extremely high affinity of Hg(II) for the zinc finger structure where Zn(II) is not able to compete. The observation that glutathione can reverse the inhibition may be explained by the fact that it has extremely high affinities for Hg(II) as well. Thus, the apparent dissociation constants were found to be in the order of 10–42 for Hg(GS)2 and 10–44 for Hg(RS)2 in complex with cysteine (32). Altogether, with respect to the three inhibitory metal compounds, our data agree with the displacement of 65Zn(II) observed previously (10) and the zinc finger appears to be the predominant target within the Fpg protein for these metal ions. With regard to the XPA protein, a different pattern of 2102

inhibition by toxic metals was observed. While 艋1 mM of Hg(II), Pb(II) or As(III) did not impair DNA binding, DNA– protein interactions were diminished by Ni(II), Cd(II), Co(II) and Cu(II). In the case of Ni(II), the result agrees with the diminished binding activity of XPA to a cisplatin-damaged oligonucleotide reported previously (33). Cd(II) and Cu(II) inhibited both XPA and Fpg, while As(III) and Pb(II) were not inhibitory to either repair protein. Hg(II) inactivated Fpg most strongly, but did not reduce XPA DNA binding. In contrast, Ni(II) and Co(II) inactivated XPA but did not affect Fpg. For all metals except Cu(II), simultaneous incubation with Zn(II) largely prevented the XPA inactivation, suggesting the displacement of zinc as an underlying mechanism of inhibition. Interestingly, in two very recent studies, the XPA MBD has been constructed with Cd2⫹ or Co2⫹ instead of Zn2⫹ (34,35). Structural investigations by different spectroscopic methods revealed a tetrahedral coordination of all three metal ions with no major distortion of the XPA MBD. In the case of Cd2⫹, however, an increased Cd–S bond length was observed (2.54 Å as opposed to 2.34 Å for Zn–S). Even though the authors considered the changes too small to disrupt DNA–protein interactions, our experiments show a diminished XPA–DNA binding by both Cd2⫹ and Co2⫹, supporting the importance of functional analyses of the protein in question. The different zinc concentrations required to prevent inhibition by the respective toxic metal compounds may again be explained by different affinities of the zinc finger structures for the diverse metal ions. With respect to Cu(II), the weak protective effect of zinc(II) is similar to the results obtained with the Fpg protein discussed above. How do these data compare with those from metal competition studies derived from zinc finger transcription factors? There are several different types of zinc finger, characterized by the nature and spacing of their zinc-chelating residues. They include the classical zinc fingers Cys2His2 found in TFIIIA or SP1 and the Cys4 type present for example in the oestrogen receptor. Even though clearly different, both types contain a β-hairpin and an α-helix folded around a zinc ion; DNA contact is mediated by the α-helix with the major groove of the DNA (5). When investigating a synthetic 26 amino acid peptide based on the consensus sequence of 131 zinc finger Cys2His2 domains [CP-1(Cys-Cys-His-His)], Krizek et al. (36) observed clearly preferential binding of Zn2⫹ over first-row and second-row transition metals including Cd2⫹, as would be predicted from ligand field stabilization in a tetrahedral environment and from hard–soft acid–base effects. However, when the two histidines were changed to cysteines [CP1(Cys-Cys-Cys-Cys)], the affinity for Cd2⫹ was increased dramatically by five orders of magnitude, resulting in a preferential binding of Cd2⫹ over Zn2⫹. Yet, metal exchange reactions in naturally occurring zinc finger proteins appear to be more complex and difficult to predict. With respect to the Cys2His2 type, Cd(II) has a lower affinity for apoTFIIIA than Zn(II) (37); nonetheless, it prevented DNA binding of TFIIIA at concentrations of 艌0.1 µM (9). Both zinc finger repair proteins applied in the present study belong to the Cys4 type; in agreement with the model studies reported above, both were inhibited by Cd(II), but—depending on the enzyme under investigation—also by other metals. With respect to Cd(II) and Cu(II), the inhibition resembles results reported by Predki and Sarkar (7), who observed that both metals had higher affinities than Zn(II) did for the bovine oestrogen receptor, which is also a Cys4-type zinc finger. However, in contrast to the results

Effects of toxic metal compounds on Fpg and XPA

obtained in the present study with the XPA protein, Co(II) and Ni(II) had lower affinities in the bovine oestrogen receptor zinc finger. Thus, these and other results show that even though zinc finger structures are common motifs in many DNAbinding proteins, each member of this family appears to have its own structural features and sensitivities towards toxic metal compounds. With respect to the inhibition of DNA repair by toxic and/ or carcinogenic metal compounds, our results indicate that zinc finger structures may be sensitive targets for metals like Cd(II), Co(II), Ni(II), Cu(II) or Hg(II), even though the latter metal did not affect XPA. In cultured mammalian cells, the incision frequency during NER is reduced by low concentrations of Ni(II), Cd(II) or Co(II) (2); studies to investigate the effects of Hg(II) and Cu(II) on this step are in progress. The diminished DNA damage recognition by nuclear proteins derived from cadmium-treated cells can be reversed by the addition of Zn(II) (38). With respect to arsenic compounds, the results are still puzzling. Several groups (39–41) have demonstrated that As(III) inhibits nucleotide excision repair at non-cytotoxic concentrations in the low micromolar range, but no arsenic-sensitive protein involved in the processing of DNA damage has been detected yet. Thus, neither DNA ligase I nor DNA ligase III nor DNA polymerase β was inhibited at concentrations below several millimolar (42). Similarly, in the present study, neither Fpg nor XPA was affected by As(III), even though arsenic is known to have a high affinity for vicinal –SH groups. Further studies are needed to elucidate the mechanism of arsenic-induced repair inhibition. Concerning the potential relevance of the observed effects, the effective concentrations have to be considered. For Co(II) and Ni(II), the inhibitory concentrations are similar to those necessary for repair inhibition in cultured cells and in the subtoxic range (43,44). For Cd(II), the 50 µM concentration required for Fpg inhibition and even 200 µM for XPA inactivation are considerably higher when compared with effects on DNA damage recognition by nuclear proteins of cadmium-treated cells (38), indicating that other repair proteins may be more sensitive to cadmium(II). Nevertheless, it has to be kept in mind that, although cadmium concentrations in blood and urine are in the nanomolar range in humans, high amounts of cadmium are stored in liver and kidney, reaching concentrations of several hundred micromolar even in occupationally unexposed individuals (45). Cadmium is mainly bound to metallothionein in these organs, but preliminary results from our laboratory indicate that metallothinein-bound cadmium is still available for inhibitory effects (S.Hoffmann and A.Hartwig, unpublished). Taken together, our results add further evidence to the hypothesis that there is no single mechanism accounting for the repair inhibitions observed by the diverse metal compounds; however, the displacement of Zn(II) in zinc finger structures may be one important mode of action. Acknowledgements The authors would like to thank Dr Serge Boiteux (Commisariat Energie Atomique, Fontanay aux Roses, France) for kindly providing the Fpg protein. This work was supported by the Deutsche Forschungsgemeinschaft, grant no. Ha 2372/1-2.

References 1. Hartwig,A. (1995) Current aspects in metal genotoxicity. BioMetals, 8, 3–11.

2. Hartwig,A. (1998) Carcinogenicity of metal compounds: possible role of DNA repair inhibition. Toxicol. Lett., 102/103, 235–239. 3. 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, 1021–1026. 4. Iwitzki,F., Schlepegrell,R., Eichhorn,U., Kaina,B., Beyersmann,D. and Hartwig,A. (1998) Nickel(II) inhibits the repair of O6-methylguanine in mammalian cells. Arch. Toxicol., 72, 681–689. 5. Mackay,J.P. and Crossley,M. (1998) Zinc fingers are sticking together. Trends Biochem. Sci., 23, 1–4. 6. Thiesen,H.-J. and Bach,C. (1991) Transition metals modulate DNA–protein interactions of SP1 zinc finger domains with its cognate target size. Biochem. Biophys. Res. Commun., 176, 551–557. 7. Predki,P.F. and Sarkar,B. (1992) Effect of replacement of ‘zinc finger’ zinc on estrogen receptor DNA interactions. J. Biol. Chem., 267, 5842–5846. 8. Nagaoka,M., Kuwahara,J. and Sugiura,Y. (1993) Alteration of DNA binding specificity by nickel(II) substitution in three zinc(II) fingers of transcription factor SP1. Biochem. Biophys. Res. Commun., 194, 1515–1520. 9. Hanas,J.S. and Gunn,C.G. (1996) Inhibition of transcription factor IIIA– DNA interactions by xenobiotic metal ions. Nucleic Acids Res., 24, 924–930. 10. O’Connor,T.R., Graves,R.J., de Murcia,G., Castaing,B. and Laval,J. (1993) Fpg protein of Escherichia coli is a zinc finger protein whose cysteine residues have a structural and/or functional role. J. Biol. Chem., 268, 9063–9070. 11. Tchou,J., Kasai,H., Shibutani,S., Chung,M.-H., Laval,J., Grollman,A.P. and Nishimura,S. (1991) 8-Oxoguanine (8-hydroxyguanine) DNA glycosylase and its substrate specificity. Proc. Natl Acad. Sci. USA, 88, 4690–4694. 12. Boiteux,S., Gajewski,E., Laval,J. and Dizdaroglu,M. (1992) Substrate specificity of the Escherichia coli Fpg protein (formamidopyrimidineDNA glycosylase): excision of purine lesions in DNA produced by ionizing radiation or photosensitization. Biochemistry, 31, 106–110. 13. Cleaver,J.E. and States,J.C. (1997) The DNA damage recognition problem in human and other eucaryotic cells: the XPA damage binding protein. Biochem. J., 328, 1–12. 14. Robins,P., Jones,C.J., Biggerstaff,M., Lindahl,T. and Wood,R.D. (1991), Complementation of DNA repair in xeroderma pigmentosum group A cell extracts by a protein with affinity for damaged DNA. EMBO J., 10, 3913–3921. 15. Jones,C.J. and Wood,R.D. (1993) Preferential binding of the xeroderma pigmentosum group A complementing protein to damaged DNA. Biochemistry, 32, 12096–12104. 16. Asahina,H., Kuraoka,I., Shirakawa,M., Morita,E.H., Miura,N., Miyamoto,I., Ohtsuka,E., Okada,Y. and Tanaka,K. (1994) The XPA protein is a zinc metalloprotein with an ability to recognize various kinds of DNA damage. Mutat. Res., 315, 229–237. 17. Lee,S.H., Kim,D.K. and Drissi,R. (1995) Human xeroderma pigmentosum group A protein interacts with human replication protein A and inhibits DNA replication. J. Biol. Chem., 270, 21800–21805. 18. Sugasawa,K., Ng,J., Masutani,C., Iwai,S., van der Spek,P.J., Eker,A., Hanaoka,F. and Hoeijmakers,J. (1998) Xeroderma pigmentosum group C protein complex is the initiator of global genome nucleotide excision repair. Mol. Cell, 2, 223–232. 19. Wakasugi,M. and Sancar,A. (1999) Order of assembly of human DNA repair excision nuclease. J. Biol. Chem., 274, 18759–18768. 20. Buchko,G.W., Iakoucheva,L.M., Kennedy,M.A., Ackerman,E.J. and Hess,N.J. (1999) Extended X-ray absorption fine structure evidence for a single metal binding domain in Xenopus laevis nucleotide excision repair protein XPA. Biochem. Biophys. Res. Commun., 254, 109–113. 21. Buchko,G.W., Ni,S., Thrall,B.D. and Kennedy,M.A. (1998) Structural features of the minimal DNA binding domain (M98–F219) of human nucleotide excision repair protein XPA. Nucleic Acids Res., 26, 2779–2788. 22. Miyamoto,I., Miura,N., Niwa,H., Miyazaki,J. and Tanaka,K. (1992) Mutational analysis of the structure and function of the xeroderma pigmentosum group A complementing protein. J. Biol. Chem., 267, 12182–12187. 23. Salditt,M., Braunstein,S.N., Camerini-Otero,R.D. and Franklin,R.M. (1972) Structure and synthesis of a lipid-containing bacteriophage-improved techniques for the purification of bacteriophage PM2. Virology, 48, 259–262. 24. Mueller,E., Boiteux,S., Cunningham,R.P. and Epe,B. (1990) Enzymatic recognition of DNA modifications induced by singlet oxygen and photosensitizers. Nucleic Acids Res., 18, 5969–5973. 25. Eker,A.P.M., Vermeulen,W., Miura,N., Tanaka,K., Jaspers,N.G.J., Hoeijmakers,J.H.J. and Bootsma,D. (1992) Xeroderma pigmentosum group A correcting protein from calf thymus. Mutat. Res., 274, 211–224. 26. Floyd,R.A., West,M.S., Eneff,K.L. and Schneider,J.E. (1989) Methylene

2103

M.Asmuss et al. blue plus light mediates 8-hydroxyguanine formation in DNA. Arch. Biochem. Biophys., 273, 106–111. 27. Schneider,J.E., Price,S., Maidt,M.L., Gutteridge,J.M.C. and Floyd,R.A. (1990) Methylene blue plus light mediates 8-hydroxy-2⬘-deoxyguanosine formation in DNA preferentially over strand breakage. Nucleic Acids Res., 18, 631–635. 28. Epe,B., Pflaum,M. and Boiteux,S. (1993) DNA damage induced by photosensitizers in cellular and cell-free systems. Mutat. Res., 299, 135–145. 29. McKinney,J.D. (1992) Metabolism and disposition of inorganic arsenic in laboratory animals and humans. Environ. Geochem. Health, 14, 43–48. 30. Mukhopadhyay,C.K., Mazumder,B., Lindley,P.F. and Fox,P.L. (1997) Identification of the prooxidant site of human ceruloplasmin: a model for oxidative damage by copper bound to protein surfaces. Proc. Natl Acad. Sci. USA, 94, 11546–11551. 31. Stadtman,E.R. (1993) Oxidation of free amino acids and amino acid residues in proteins by radiolysis and by metal-catalyzed reactions. Annu. Rev. Biochem., 62, 797–821. 32. Stricks,W. and Kolthoff,I.M. (1953) Reactions between mercuric mercury and cysteine and glutathione: apparent dissociation constants, heats and entropies of formation of various forms of meruric mercapto-cysteine and -glutathione. J. Am. Chem. Soc., 75, 5673–5681. 33. Krueger,I., Mullenders,L.H.F. and Hartwig,A. (1999) Nickel(II) increases the sensitivity of V79 Chinese hamster cells towards cisplatin and transplatin by interference with distinct steps of DNA repair. Carcinogenesis, 20, 1177–1184. 34. Buchko,G.W., Hess,N.J. and Kennedy,M.A. (2000) Human nucleotide excision repair protein XPA: summary of the EXAFS studies on the Zn(II), Co(II) and Cd(II) associated minimal DNA-binding domain. Protein Peptide Lett., 7, 49–56. 35. Buchko,G.W., Hess,N.J. and Kennedy,M.A. (2000) Cadmium mutagenicity and human nucleotide excision repair protein XPA: CD, EXAFS and 1H/

2104

15N-NMR spectroscopic studies on the zinc(II) and cadmium(II)-associated minimal DNA-binding domain. Carcinogenesis, 21, 1051–1057. 36. Krizek,B.A, Merkle,D.L. and Berg,J.M. (1993) Ligand variation and metal ion binding specificity in zinc finger peptides. Inorg. Chem., 32, 937–940. 37. Makowski,G.S. and Sunderman,F.W. Jr (1992) The interactions of zinc, nickel, and cadmium with Xenopus transcription factor IIIA, assessed by equilibrium dialysis. J. Inorg. Biochem., 48, 107–119. 38. Hartmann,M. and Hartwig,A. (1998) Disturbance of DNA damage recognition after UV-irradiation by nickel(II) and cadmium(II) in mammalian cells. Carcinogenesis, 19, 617–621. 39. Okui,T. and Fujiwara,Y. (1986) Inhibition of human excision DNA repair by inorganic arsenic and the co-mutagenic effect in V79 Chinese hamster cells. Mutat. Res., 172, 69–76. 40. Lee-Chen,S.F., Yu,C.T. and Jan,K.Y. (1992) Effect of arsenite on the repair of UV-irradiated Chinese hamster ovary cells. Mutagenesis, 7, 51–55. 41. Hartwig,A., Gro¨ blinghoff,U.D., Beyersmann,D., Natarajan,A.T., Filon,R. and Mullenders,L.H.F. (1997) Interaction of arsenic(III) with nucleotide excision repair in UV-irradiated human fibroblasts. Carcinogenesis, 18, 399–405. 42. Hu,Y., Su,L. and Snow,E.T. (1998) Arsenic toxicity is enzyme specific and its affects on ligation are not caused by the direct inhibition of DNA repair enzymes. Mutat. Res., 408, 203–218. 43. Hartwig,A., Mullenders,L.H.F., Schlepegrell,R., Kasten,U. and Beyersmann,D. (1994) Nickel(II) interferes with the incision step in nucleotide excision repair. Cancer Res., 54, 4045–4051. 44. Kasten,U., Mullenders,L.H.F. and Hartwig,A. (1997) Cobalt(II) inhibits the incision and the polymerization step of nucleotide excision repair in human fibroblasts. Mutat. Res., 383, 81–89. 45. Barregard,L., Svalander,C., Schu¨ tz,A., Westberg,G., Sa¨ llsten,G., Blohme,I., Mo¨ lne,J., Attman,P.O. and Haglind,P. (1999) Cadmium, mercury and lead in kidney cortex of the general Swedish population: a study of biopsies from living kidney donors. Environ. Health Perspect., 107, 8657–871.

Received March 31, 2000; revised August 7, 2000; accepted August 15, 2000