Detection of Chemically Induced DNA Damage in Layered Films by

Catalytic oxidation using 50 μM Ru(bpy)32+ (bpy = 2,2'-bipyridine) and square wave voltammetry (SWV) provided more sensitive detection of DNA damage ...
0 downloads 0 Views 200KB Size
Download PDF
Anal. Chem. 2001, 73, 4780-4786

Articles

Detection of Chemically Induced DNA Damage in Layered Films by Catalytic Square Wave Voltammetry Using Ru(Bpy)32+ Liping Zhou and James F. Rusling*

Department of Chemistry, University of Connecticut, U-60, 55 North Eagleville Road, Storrs, Connecticut 06269-3060

A sensor constructed by alternate layer-by-layer adsorption of PDDA cations and double-stranded (ds)-DNA on oxidized pyrolytic graphite electrodes was evaluated for detection of chemical damage to ds-DNA from known damage agent styrene oxide. Films made with PDDA ions of structure (PDDA/DNA)2 were ∼6 nm thick and contained 0.23 µg of ds-DNA. Catalytic oxidation using 50 µM Ru(bpy)32+ (bpy ) 2,2′-bipyridine) and square wave voltammetry (SWV) provided more sensitive detection of DNA damage than direct SWV oxidation. The catalytic peaks increased linearly with time during incubations with styrene oxide, but only minor changes were detected during incubation with nonreactive toluene. For best sensitivity, the outer layer of the film must be ds-DNA, and analysis should be done at low salt concentration. Studies of DNA and polynucleotides in solutions and films suggested that oxidation of guanine and chemically damaged adenine in partly unraveled, damaged DNA were the most likely contributors to the catalytic peak. DNA damage caused by metabolites of lipophilic pollutants and drugs formed by mammalian liver cytochrome P450 enzymes1,2 is a major toxicity pathway.3-6 Simple, inexpensive methods to detect DNA damage by these metabolites could lead to rapid in vitro toxicological screening of new chemicals. Although chromatographic or electrophoretic separation of hydrolyzed samples coupled to mass spectrometry7-10 provides (1) Schenkman, J. B., Greim, H., Eds. Cytochrome P450; Springer-Verlag: Berlin, 1993. (2) Ortiz de Montellano, P. R., Ed. Cytochrome P450; Plenum: New York, 1995. (3) Singer, B.; Grunberger, D. Molecular Biology of Mutagens and Carcinogens; Plenum: New York, 1983. (4) Pauwels, W.; Vodiceka, P.; Severi, M.; Plna, K.; Veulemans, H.; Hemminki, K. Carcinogenisis 1996, 17, 2673-2680. (5) McConnell, E. E.; Swenberg, J. A. CRC Crit. Rev. Toxicol. 1994, 24, S49S55. (6) Nestmann, E. R.; Bryant, D. W.; Carr, C. J.; Fennell, T. T.; Gorelick, N. J.; Gallagher, J. E.; Swenberg, J. A.; Williams, G. M. Regul. Toxicol. Pharmacol. 1996, 24, 9-18. (7) Cadet, J.; Weinfeld, M. Anal. Chem. 1993, 65, 675A-682A. (8) Deforce, D. L. D.; Ryniers, F. P. K.; Van den Eeckout, E. G.; Lemiere, F.; Esmans, E. L. Anal.Chem. 1996, 68, 3575-3584. (9) Schrader, W.; Linscheid, M. Arch. Toxicol. 1997, 71, 588-595. (10) Deforce, D. L. D.; Lemiere, F.; Esmans, E. L.; De Leenheer, A.; Van den Eeckout, E. G. Anal. Biochem. 1998, 258, 331-338.

4780 Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

detailed molecular information on DNA damage, toxicity screening is limited by analysis time and cost. Voltammetric oxidation of DNA is attractive for rapid, inexpensive assays. Palecek and co-workers employed adsorption of DNA on carbon electrodes and showed that single-stranded (ss)DNA is much more easily oxidized than double-stranded (ds)DNA.11-13 Adsorptive voltammetry on mercury electrodes was used to detect DNA damage from strong acid,14 methylating agents,15 and hydroxyl radicals.16,17 Damage to solubilized DNA from ionizing radiation was detected on mercury electrodes by adsorptive voltammetry18,19 in DNA solutions and by chronopotentiometry with DNA adsorbed onto carbon electrodes.20 Facile in vitro electrochemical detection of DNA damage by toxic metabolites could be envisioned with stable solid electrochemical sensors coated with DNA films. We recently reported detection of DNA damage in films containing ds-DNA and ionomers Eastman AQ and Nafion on graphite electrodes by direct oxidation using derivative square wave voltammetry (SWV).21,22 Oxidation peaks developed as the films were incubated with styrene oxide, which forms known covalent adducts with guanine and adenine in DNA with genetic consequences.4,23-26 Relative rates of DNA damage were detectable by this method.21,22 (11) Palecek, E. Electroanalysis 1996, 8, 7-14. (12) Palecek, E.; Fojta, M. Anal. Chem. 2001, 73 74A-83A. (13) Palecek, E.; Boublikova, P.; Jelen, F. Anal. Chim. Acta 1986, 187, 99-107. (14) Jelen, F.; Fojta, M.; Palecek, E. J. Electroanal. Chem. 1997, 427, 49-56. (15) Jelen, F.; Tomschik, M.; Palecek, E. J. Electroanal. Chem. 1997, 423, 141148. (16) Fojta, M.; Palecek, E. Anal. Chim. Acta 1997, 342, 1-12. (17) Fojta, M.; Stankova, V.; Palecek, E.; Koscielniak, P.; Mitas, J. Talanta 1998, 46, 155-161. (18) Sequaris, J.-M.; Valenta, P.; Nuernberg, H. W. Int. J. Radiat. Res. 1982, 42, 407-415. (19) Sequaris, J.-M.; Valenta, P. J. Electroanal. Chem. 1987, 227, 11-20. (20) Wang, J.; Rivas, G.; Ozsoz, M.; Grant, D. H.; Cai, X.; Parrado, C. Anal. Chem. 1997, 69, 1457-1460. (21) Mbindyo, J.; Zhou, L.; Zhang, Z.; Stuart, J. D.; Rusling, J. F. Anal. Chem. 2000, 72, 2059-2065. (22) Rusling, J. F.; Zhou, L.; Munge, B.; Yang, J.; Estavillo, C.; Schenkman, J. B. Faraday Discuss. 2000, 116, 77-87. (23) Bond, J. A. CRC Crit. Rev. Toxicol. 1989, 19, 227-249. (24) Latham, G. J.; Zhou, L.; Harris, C. M.; Harris, T. M.; Lloyd, R. S. J. Biol. Chem. 1993, 268, 23427-23434. (25) Latham, G. J.; Lloyd, R. S. J. Biol. Chem. 1994, 269, 28527-28530. (26) (a) Koskinen, M.; Vodicka, P.; Hemminki, K. Chem.-Biol. Interact. 2000, 124, 13-27. (b) Le, P. T. Q.; Harris, C. M.; Harris, T. M.; Stone, M. P. Chem. Res. Toxicol. 2000, 13, 63-71. 10.1021/ac0105639 CCC: $20.00

© 2001 American Chemical Society Published on Web 09/08/2001

However, disadvantages include small SWV peaks which need extraction from background. Catalytic electrochemical oxidation using transition metal complexes provides enhanced electrochemical signals for DNA.27 Thorp et al. showed that one of the most efficient catalysts is Ru(bpy)32+, which specifically oxidizes guanine bases in DNA and oligonucleotides28,29 as follows:

Ru(bpy)32+ ) Ru(bpy)33+ + e-

(1)

Ru(bpy)33+ + DNA(guanine) f Ru(bpy)32+ + DNA(guanine+) (2) DNA(guanine+) is subsequently further oxidized.30 Cycling of Ru(bpy)33+ back to Ru(bpy)32+ by the fast chemical step in eq 2 provides a catalytic current in voltammetry that is greatly enhanced over that of Ru(bpy)32+ or DNA alone. The peak current depends on the rate of this chemical step. The double-helix structure of ds-DNA shields guanine from efficient contact with Ru(bpy)33+. When the double helix is destroyed, guanine becomes more available and reacts more rapidly with the catalyst, thus providing a way to distinguish between ds- and ss-DNA.27-29 We felt that this strategy could be equally effective for detecting DNA damage by metabolites that react covalently and thus also disrupt the DNA double helix. In this paper, we examine sensors built with films of polycations and ds-DNA for their effectiveness to detect chemically induced DNA damage. The films were constructed layer by layer by alternate adsorption of layers of polycation and DNA on oxidized pyrolytic graphite (PG) electrodes.31 Square wave voltammetric oxidation catalyzed by Ru(bpy)32+ provided better sensitivity to damage of ds-DNA by styrene oxide compared to direct oxidation. EXPERIMENTAL SECTION Chemicals. Calf thymus (CT) ds-DNA (type XV), CT ss-DNA, poly(guanylic acid) (5′) (poly[G]), poly(cytidylic acid) (5′) (poly[C]), poly(adenylic acid) (5′) (poly[A]), and styrene oxide were from Sigma. Tris(2,2′-bipyridyl)dichlororuthenium(II) hexahydrate, poly(diallydimethylammonium chloride) (PDDA) and poly(sodium 4-styrenesulfonate) (PSS) were from Aldrich. Other chemicals were reagent grade. Electrochemical Experiments. A Bioanalytical Systems BAS100B/W electrochemical analyzer was used for SWV and chronocoulometry. A three-electrode thermostated cell employed a (27) Thorp, H. H. Trends Biotechnol. 1998, 16, 117-121. (28) (a) Johnston, D. H.; Glasgow, K. C.; Thorp, H. H. J. Am. Chem. Soc. 1995, 117, 8933-8938. (b) Napier, M. E.; Thorp, H. H. Langmuir 1997, 13, 63426344. (c) Farrer, B. T.; Thorp, H. H. Inorg Chem. 2000, 39, 44-49. (d) Yang, I. V.; Thorp, H. H. Inorg. Chem. 2000, 39, 4969-4976. (e) Armistead, P. M.; Thorp, H. H. Anal. Chem. 2000, 72, 3764-3770. (f) Sistare, M. F.; Codden, S. J.; Heimlich, G.; Thorp, H. H. J. Am. Chem. Soc. 2000, 122, 4742-4749. (g) Szalai, V. A.; Thorp, H. H. J. Phys. Chem. B 2000, 104, 6851-6859. (29) Onkto, A. C.; Armistead, P. M.; Kircus, S. R.; Thorp, H. H. Inorg. Chem. 1999, 38, 1842-1846. (30) Armistead, P. M.; Thorp, H. H. Anal. Chem. 2001, 73, 558-564. (31) (a) Lvov, Y. M.; Lu, Z.; Schenkman, J. B.; Zu X.; Rusling, J. F. J. Am. Chem. Soc. 1998, 120, 4073-4080. (b) Rusling, J. F. In Protein Architecture: Interfacing Molecular Assemblies and Immobilization Biotechnology; Lvov Y., Mo ¨hwald, H., Eds.; Marcel Dekker: New York, 2000; pp 337-354.

saturated calomel reference electrode (SCE), a Pt wire counter electrode, and a PG disk (Advanced Ceramics, A ) 0.2 cm2) working electrode. SWV conditions were 4 mV step height, 25 mV pulse height, and frequency 15 Hz. The electrolyte solution was 10 mM acetate buffer, pH 5.5, containing 50 mM NaCl and 50 µM Ru(bpy)32+ unless otherwise noted. The cell was thermostated at 37 °C and protected from light to avoid photodecomposition of Ru(bpy)32+. Solutions were purged with pure nitrogen for 15 min prior to each series of experiments, and a nitrogen atmosphere was maintained during data collection. Ohmic drop was compensated >95% by the BAS 100B/W system. For chronocoulometry, single steps were used from 500 to 1150 mV versus SCE at a pulse width of 30 s. A different electrode was used for each catalytic film analysis, as the oxidation modifies the nucleic acid layers in the film. Film Assembly. Basal plane PG electrodes were polished by hand with 400-grit SiC paper and then with 0.3-µm R-alumina slurries on Buehler Microcloth, washed with water, and ultrasonicated in water for 30 s. These electrodes were oxidized in 2.5% K2Cr2O7/10% HNO3 by scanning once from 1.5 to 1.7 V versus SCE to form negative carboxylate groups on the surface.32 These oxidized electrodes were dried and dipped into 2 mg mL-1 PDDA in 50 mM NaCl, washed with water, and dipped into 2 mg mL-1 CT ds-DNA in pH 7.0 Tris buffer. The polycation PDDA was chosen as a film material for its opposite charge to DNA and since it had been used previously to construct stable, layered films by this method.22,31,33 Adsorption times of 15 min were used, which provides steady-state adsorption of CT ds-DNA and PDDA on oppositely charged surfaces.31 Alternate adsorption cycles were repeated until the desired number of layers was made. Films containing ss-DNA, poly[G], poly[A], and poly[C] were assembled in a similar way. Films prepared on PG electrodes that had not been oxidized gave much smaller catalytic voltammetric signals, possibly because of smaller amounts of adsorbates in the initial layers. Analysis of Nucleic Acids in Solution. Nucleic acids were dissolved at 0.2 mg mL-1 in 10 mM acetate buffer, pH 5.5, containing 50 mM NaCl. For incubations, these solutions were stirred with saturated styrene oxide for 30 min at 37 °C. Styrene oxide was extracted with ether before SWV analysis of the aqueous phase using a bare PG electrode. Safety Note: Styrene oxide is a suspected human carcinogen and somewhat volatile. Gloves were worn, and all weighings and manipulations were done under a closed hood. All reactions were done in closed vessels. Reaction of Nucleic Acid Films with Styrene Oxide. Incubations of films in saturated styrene oxide solutions were done in a thermostated cell at 37 °C. A 120-µL sample of neat styrene oxide was added to 10 mL of acetate buffer containing 50 mM NaCl. In control experiments, 120 µL of neat toluene was added instead of styrene oxide. The electrode with DNA film or polynucleotide film was incubated in this stirred emulsion for the desired time. The electrode was then rinsed with water and transferred to the buffer containing Ru(bpy)32+ for SWV analysis. (32) Njue C. K.; Rusling, J. F. J. Am. Chem. Soc. 2000, 122, 6459-6463. (33) (a) Lvov, Y.; Ariga, K.;. Ichinose, I.; Kunitake, T. J. Am. Chem. Soc. 1995, 117, 6117-6123. (b) Lvov, Y. In Protein Architecture: Interfacing Molecular Assemblies and Immobilization Biotechnology; Lvov, Y., Mo ¨hwald, H., Eds.; Marcel Dekker: New York, 2000; pp 125-167.

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

4781

Quartz Crystal Microbalance. A quartz crystal microbalance (QCM, USI System) was used to monitor film assembly.31 QCM resonators (9 MHz, AT-cut, International Crystal Mfg. Co.) were covered by 100-nm evaporated gold electrodes (0.16 cm2). Reproducibility was (2 Hz. To simulate the oxidized PG surfaces used for voltammetry, clean Au surfaces were coated by dipping into a solution of 0.3 mM 3-mercaptopropionic acid and 0.7 mM 3-mercapto-1-propanol in ethanol.32 Layered films were prepared on this surface as above on the oxidized PG electrodes. For QCM measurements, the resonator was immersed in a given adsorbate solution for 15 min, washed, and dried in a stream of nitrogen, and the frequency change was measured at ambient temperature. The Sauerbrey equation gives the relation between adsorbed mass and frequency shift ∆F (Hz). For the 9-MHz quartz resonators, the film mass per unit area M/A (g cm-2) is33

M/A ) -∆F/(1.83 × 108)

(3)

for our resonators of A ) 0.16 ( 0.01 cm2 on one side. The nominal thickness (d) of dry films can be estimated from33

d (nm) ≈ -(0.016 ( 0.002)∆F (Hz)

(4)

RESULTS SWV Analyses of Nucleic Acids in Solution. The two purine bases guanine and adenine in DNA form chemical adducts with styrene oxide, although the majority of the reactions occur at guanine.9,26,34 It is possible that adducts formed on DNA after damage reactions, as well as unreacted guanine,27-29 could be catalytically oxidized by Ru(bpy)32+. To develop a better idea of what to expect with DNA films, we first examined catalytic oxidations of dissolved intact and damaged polynucleotides with Ru(bpy)32+. Previous work showed that the maximum rate of DNA damage by styrene oxide occurred at pH 5.5,21,22 so this pH was used for all reactions in this paper. The formation of DNA-styrene oxide adducts under these incubation conditions has been confirmed by capillary electrophoresis.21 As reported previously for other electrodes,28 Ru(bpy)32+ in aqueous solution on pyrolytic graphite electrodes gave reversible oxidation-reduction peak pairs. The midpoint (formal) potential was 1.07 V versus SCE by cyclic and square wave voltammetry. As seen in Figure 1a, SWV of 50 µM Ru(bpy)32+ gave a very small oxidation peak, which is just barely visible compared to scans in the buffer alone. After the addition of ss-DNA, this oxidation peak was greatly enhanced. For ds-DNA, a much smaller catalytic peak was observable under these conditions, but the peak was larger than that of ds-DNA alone in the buffer. When ds-DNA was incubated for 30 min with styrene oxide, a large oxidation peak at 1.07 V was found (Figure 1b). Another oxidation peak at 0.8 V was observed in styrene oxide damaged ds-DNA and also with ss-DNA (Figure 1a). This peak appeared both with and without Ru(bpy)32+ present and, thus, is not catalytic. Studies on dissolved polynucleotides were done to provide insight about origins of the catalytic signals. Figure 2a reveals a small direct oxidation peak at 1 V for poly[G], which is greatly (34) Vodicka, P.; Hemminki, K. Carcinogenesis 1988, 9, 1657-1660.

4782 Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

Figure 1. SWV of dissolved DNA (0.2 mg mL-1) in pH 5.5 acetate buffer containing 50 mM NaCl, with or without 50 µM Ru(bpy)32+ as labeled: (a) buffer, and unreacted ds-DNA and ss-DNA; (b) unreacted ds-DNA and ds-DNA after incubation at 37 °C for 30 min with styrene oxide. Points on curves in all SWV figures are for identification purposes only; SWV amplitude, 25 mV; frequency, 15 Hz; step, 4 mV.

enhanced as expected in the presence of Ru(bpy)32+. After incubation of poly[G] with styrene oxide, both the direct oxidation and catalytic Ru(bpy)32+ peaks increased. The catalytic peak was about twice the height of the direct oxidation peak. A small broad peak was found at ∼0.76 V for styrene oxide-treated poly[G] in the presence and absence of Ru(bpy)32+. For intact poly[A], only a small peak for Ru(bpy)32+ oxidation was found. After poly[A] was reacted with styrene oxide, a small increase in this peak was found (Figure 2b). However, a new peak developed at ∼0.78 V in the absence and presence of catalyst, reminiscent of the peak seen with damaged ds-DNA. Voltammograms showed no peaks between 0.4 and 1.2 V for poly[C] that had or had not been incubated with styrene oxide (Figure 2b). The Ru(bpy)32+ oxidation peak in either of these poly[C] solutions showed little evidence of catalytic enhancement. QCM Monitoring of Film Construction. Figure 3 shows a nearly linear progression of frequency shifts for films made with alternate layers of PDDA and ds-DNA, indicating regular film growth with reproducible layers. From eq 4 we estimated a

Figure 3. QCM frequency shifts for cycles of alternate PDDA/DNA adsorption on gold resonators coated with mixed monolayers of mercaptoproionic acid and mercaptopropanol. One set of experiments utilized ds-DNA (line), while the other utilized ss-DNA. Average values given for eight resonators.

Figure 2. SWV of dissolved polynucleotides (0.2 mg mL-1) in pH 5.5 acetate buffer containing 50 mM NaCl, with or without 50 µM Ru(bpy)32+ as labeled: (a) unreacted poly[G] and poly[G] after incubation at 37 °C for 30 min with styrene oxide; (b) poly[A] and poly[C] that had and had not been incubated at 37 °C for 30 min with styrene oxide. Baselines were offset for clarity.

thickness of ∼6 nm for the films of (PDDA/ds-DNA)2 that were used for voltammetric analyses. These films contain about 1.1 ( 0.3 µg cm-2 or 0.23 µg of ds-DNA per PG electrode. Films made with ss-DNA were examined for comparison (Figure 3) and showed a similar amount of adsorption to ds-DNA for the first several layers, but larger amounts for subsequent layers. Electrochemical Analysis of Damaged DNA Films. Films of the structure (PDDA/ds-DNA)2 were constructed on oxidized PG electrodes. When the ds-DNA films were reacted with styrene oxide, the SWV oxidation peaks increased with the incubation time (Figure 4). Average peak current for the ds-DNA films increased linearly with incubation time for the first 0.5 h and then decreased slightly (Figure 5). The slope of this line was 1.25 µA min-1, and the correlation coefficient was 0.985. Error bars in Figure 5 are mainly the result of electrode-to-electrode variability. When films were incubated with toluene, for which no chemical reactions with DNA have been reported, the catalytic oxidation peaks remained within electrode-to-electrode experimental error of the control experiments without toluene (Figures 4 and 5).

Figure 4. SWV of (PDDA/ds-DNA)2 films on oxidized PG in pH 5.5 acetate buffer containing 50 mM NaCl, with 50 µM Ru(bpy)32+ in buffer, and after incubations at 37 °C with styrene oxide (SO) or toluene.

Square wave voltammetric peaks after incubation with styrene oxide were broader, ∼5-fold smaller, and partly overlapped with the final current rise when PDDA rather than DNA was the outer layer in the films. This may result from a combination of electrostatic repulsion of Ru(bpy)32+ from the positively charged PDDA and decreased accessibility of the inner DNA layers to styrene oxide. Furthermore, chloride ions from the solution may be bound to the outer PDDA layer and chloride oxidation may contribute more to the background current than for an outer DNA layer. Several authors recently used chronocoulometry to differentiate between different types of DNA via the kinetics of electroAnalytical Chemistry, Vol. 73, No. 20, October 15, 2001

4783

Figure 5. Influence of incubation time with styrene oxide and toluene on the average catalytic peak current (less current for controls) for 5-15 trials per data point for (PDDA/ds-DNA)2 films. Error bars represent standard deviations.

Figure 6. Chronocoulometric charge-time (Q-t) curves of (PDDA/ ds-DNA)2 and (PDDA/ss-DNA)2 on oxidized electrodes in pH 5.5 buffer containing 50 mM NaCl and 50 µM Ru(bpy)32+. One curve shows PG/(PDDA/ds-DNA)2 electrodes after incubation with styrene oxide for 30 min. A control experiment with a (PDDA/PSS)2 film is also shown. Initial, E ) 500 mV vs SCE; final, E ) 1150 mV.

chemical catalytic schemes.30,35 Figure 6 shows charge-time curves for a single step of 30 s to a potential at which Ru(bpy)32+ is rapidly oxidized. This sets up the chemical reaction (eq 2) to complete the catalytic cycle, and the amount of charge passed is proportional to the reaction rate. (PDDA/X)2 films with X ) ssDNA, PSS, or ds-DNA were compared. Charges for all DNA films in Figure 6 were significantly larger than the X ) PSS control, for which no catalysis occurs. These experiments confirmed the results of SWV, in that ss-DNA gave the largest charge, ds-DNA incubated for 30 min with styrene oxide was intermediate, and ds-DNA gave a smaller charge. The NaCl concentration in the buffer for all experiments described thus far was 50 mM, but rates of guanine oxidation by Ru(bpy)33+ were shown to depend on ionic strength.28d,g We 4784 Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

Figure 7. SWV of (PDDA/ds-DNA)2 films showing the effect of 1 M NaCl at pH 5.5, with 50 µM Ru(bpy)32+ in buffer, and after incubations at 37 °C with styrene oxide

investigated voltammetry with 1 M NaCl in the solution in the hope of gaining some analytical advantage (Figure 7). However, compared to the low salt concentration, in 1 M NaCl the peaks for ds-DNA films incubated with styrene oxide were smaller, and the background currents were larger near the positive end of the potential window. In ss-DNA, guanine and adenine are exposed to reactions with styrene oxide and the catalyst. We did not detect significant differences in the peak heights from controls or any dependence on time for (PDDA/ss-DNA)2 films incubated with styrene oxide for up to 30 min. Catalytic peak heights for incubated films (530 min) and controls under our standard SWV conditions averaged ∼50 µA. Voltammetry of Polynucleotide Films. To gain insight into the contributions to the catalytic peaks, (PDDA/X)2 films with X ) poly[G], poly[A], and poly[C] were incubated with styrene oxide. We observed an increase in the catalytic Ru(bpy)32+ peak for poly[A] with incubation time (Figure 8a). Poly[C] gave no increase in the Ru(bpy)32+ peak upon incubation of its films with styrene oxide. Poly[G] showed small variations in peak height with incubation time, but five trials revealed no correlation in the catalytic oxidation peak height with styrene oxide incubation time (Figure 8b). Average peak heights for poly[G] films before incubation and after 30-min incubation were within (10%. DISCUSSION Results presented herein demonstrate that square wave voltammetry employing Ru(bpy)32+ as an oxidation catalyst is a viable method for measuring DNA damage from reactive metabolites in layered films of PDDA/DNA on oxidized PG electrodes (Figures 4 and 5). Chronocoulometry (Figure 6) and relative SWV peak heights suggested that films with outer ss-DNA layers are oxidized fastest by Ru(bpy)33+ and that ds-DNA films that had been incubated with styrene oxide for 30 min are oxidized much

Figure 8. SWV of (PDDA/X)2 films on oxidized PG in pH 5.5 acetate buffer containing 50 mM NaCl, with 50 µM Ru(bpy)32+ in buffer, and after incubations at 37 °C with styrene oxide. (a) X ) poly[C] or poly[A] (baselines were offset for clarity); (b) X ) poly[G].

faster than native ds-DNA films. These trends are the same as observed with all components in solution as indicated by relative SWV peak heights (Figure 1). Layer-by-layer ultrathin film assembly based on alternate adsorption of layers of oppositely charged polyions provides stable films that can withstand high salt concentrations.31,33 PDDA/DNA films stored for 2 weeks at ∼5-8 °C showed no noticeable deterioration, but they were not tested for longer storage times. Films made with heme proteins and DNA by the layer-by-layer method were usable for up to two months.31 QCM results suggest that our analytical films were ∼6 nm thick and contain roughly 0.23 µg of ds-DNA/electrode. The method provides reasonable reproducibility as shown by error bars for QCM results (Figure 3) and for peak current versus incubation time data on a large number of individual ds-DNA electrodes (Figure 5). The linear increase in peak current for styrene oxide incubation time and the lack of such an increase for nonreactive toluene suggests that the method could be used to obtain relative DNA damage rates. In direct oxidative detection of DNA damage with derivative SWV, multiple oxidation peaks ( ∼0.8 and ∼1 V) grew at different rates during styrene oxide treatment and the potentials of these peaks were somewhat variable.21 When the Ru(bpy)32+ catalyst was used, only one major oxidation peak was found at a reproduc-

ible potential close to 1.07 V versus SCE. The correlation coefficient for the linear increase in peak height with styrene oxide incubation time (0.985, Figure 5) was similar to those obtained by direct derivative SWV oxidation (0.976 for CT ds-DNA and 0.988 for salmon sperm ds-DNA).21 While different unlayered, cast ionomer/DNA films were used in the direct oxidation work, that method required incubations over several hours. The faster analysis of DNA damage rates with electrochemical catalysis reflects the enhanced signal-to-noise ratio achieved. Further, the present catalytic method employs SWV peak heights directly with no further data manipulation as needed for direct oxidation methods. Ru(bpy)32+ was reported to catalytically oxidize only guanine in nucleic acids.28 In solution, we found catalytic peaks with Ru(bpy)32+ for ss-DNA and poly[G] and also for ds-DNA and poly[A] that had been incubated with styrene oxide (Figures 1 and 2). Catalytic SWV responses for the (PDDA/X)2 films mirrored this behavior (Figures 4 and 8). Poly[C], with no known styrene oxide adducts, does not give an increased signal upon incubation with styrene oxide either in solution or in films. While poly[G] reacts with styrene oxide, this reaction caused no significant changes in the already large catalytic peak heights from oxidation of guanine (Figure 8b). On the basis of these observations, we suggest that the catalytic peaks result mainly from oxidation of exposed guanine, with a smaller contribution from adeninestyrene oxide adducts. A major influence on the catalytic peak is likely to be partial unwinding of the ds-DNA helix caused by the formation of chemical adducts during the incubation, thus allowing guanine to be exposed to the catalytic reagent. Dissolved nucleic acids also showed direct oxidation peaks at ∼0.8 V for ds-DNA, poly[G], and poly[A] that had been incubated with styrene oxide (Figures 1 and 2). These peaks may be due to direct oxidation of guanine and adenine adducts with styrene oxide.21 No such peaks were observed for damaged ds-DNA (Figures 4 and 7), poly[G], or poly[A] (Figure 8) in the (PDDA/ X)2 films that had been incubated with styrene oxide. However, small but significant current increases were observed in this potential range after incubation. It is possible that the damage occurs mainly in the outer nucleic acid layer of the films and that the subsequent direct oxidation occurs only with difficulty. In DNA/Eastman AQ ionomer films, direct oxidations by derivative SWV also showed peaks at 0.8 V.21 This difference may reflect very different molecular environments for DNA in the two types of films. In (PDDA/X)2 films, the nucleic acids are closely packed together and held into place by electrostatic forces. However, Eastman AQ ionomer films are hydrogels containing ∼90% water36 and are likely to provide an environment more like an aqueous solution, allowing considerably more molecular mobility. In summary, our results show that catalytic SWV oxidation using Ru(bpy)32+ provides a promising tool for detecting chemically induced DNA damage, as well as DNA hybridization as was already documented.27,28 An easily assembled sensor was made from stable layered films of PDDA and ds-DNA on oxidized PG electrodes. The outer layer of this film must be ds-DNA. The enhanced catalytic oxidation peak in SWV allows more sensitive and rapid detection of DNA damage than direct SWV oxidation. (35) Boon, E. M.; Ceres, D. M., Drummond, T. G.; Hill, M. G. Barton, J. K. Nature Biotechnol. 2000, 18, 1096-1100. (36) Hu, N.; Rusling, J. F. Langmuir 1997, 13, 4119-4125.

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

4785

The catalytic peaks result from oxidation of guanine and chemically damaged adenine. Comparison with previously reported data from capillary electrophoretic analysis of hydrolyzed, styrene oxide-damaged CT ds-DNA21 suggests that the present method showing clear catalytic SWV peak enhancements after 5-min incubations can detect roughly 0.1% damage or 1 damaged base/ 1000 in (PDDA/DNA)2 films. This is a 3-fold improvement over direct oxidation by SWV. 21 Relatively nontoxic chemicals (like styrene) are activated by enzymes in the human liver to genotoxic forms, e.g., styrene oxide.1-3 We recently effected such enzymatic conversions in layered films of cyt P450s and polyions.31,37 A future approach to toxicity screening could involve coupling of DNA films with (37) Zu, X.; Lu, Z.; Zhang, Z.; Schenkman, J. B.; Rusling, J. F. Langmuir 1999, 15, 7372-7377.

4786

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

enzyme activation in films, with subsequent electrochemical detection of the resulting DNA damage. ACKNOWLEDGMENT This work was supported by U.S. PHS Grant ES03154 from the National Institute of Environmental Health Sciences (NIEHS), NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NIEHS, NIH. The authors thank Mark Roy for preliminary studies.

Received for review May 18, 2001. Accepted August 6, 2001. AC0105639