Quantitation, Cationic Binding Activity, and Electron-Transfer Kinet

Burnaby, British Columbia V5A 1S6, Canada. To examine DNA-modified surfaces, we have developed a simple, convenient, and reliable procedure based on ...
0 downloads 0 Views 143KB Size
Download PDF
Anal. Chem. 2003, 75, 3902-3907

Voltammetric Procedure for Examining DNA-Modified Surfaces: Quantitation, Cationic Binding Activity, and Electron-Transfer Kinetics Hua-Zhong Yu,*,† Chuan-Yun Luo,† Carlo G. Sankar,‡ and Dipankar Sen‡

Department of Chemistry and Department of Molecular Biology and Biochemistry, Simon Fraser University, Burnaby, British Columbia V5A 1S6, Canada

To examine DNA-modified surfaces, we have developed a simple, convenient, and reliable procedure based on the voltammetric response of multiply charged transition metal cations (such as [Ru(NH3)6]3+) bound electrostatically to the DNA probes. At micromolar concentrations of the redox molecules in the electrolyte, the reduction and oxidation waves resulting from the immobilized cations on DNA-modified electrodes are well defined, stable, and reproducible. The surface densities of both single- and double-stranded oligonucleotides were accurately determined by integration of the peak for reduction of [Ru(NH3)6]3+ to [Ru(NH3)6]2+. In addition, the binding constant and electron-transfer rate constant of [Ru(NH3)6]3+ on DNA-modified electrodes were evaluated with the help of classical models. The present research provides not only an applicable and simple protocol for the quantitation of DNA probes on chips but also a versatile and powerful tool for the investigation of the binding activity and electron-transfer kinetics of cationic analytes on DNA-modified surfaces. DNA microarrays are typically monolithic, flat surfaces that bear multiple, high-density single-stranded oligonucleotide sequences. These probes are used to detect complementary DNA fragments based on traditional imaging technologies, most often fluorescence or radioisotopic detection.1 Recent progress in lithography and self-assembly techniques has contributed significantly to the improvement of both the complexity (different probes per unit area) of the chip and the surface density of probe molecules. Tremendous efforts have been made to search for ideal substrates, optimal synthetic routes, and sensitive detection methods for DNA chips. Glass slides, polymer substrates, carbon, gold, and silicon have been employed for immobilization of DNA.1a In particular, thiolate DNA monolayers on gold are of great interest because of their simple preparation and versatility for * To whom correspondence should be addressed. Fax: (604) 291-3765. E-mail: [email protected]. † Department of Chemistry. ‡ Department of Molecular Biology and Biochemistry. (1) For recent reviews of DNA chips, see: (a) Pirrung, M. C. Angew. Chem., Int. Ed. 2002, 41, 1276. (b) Blohm, D. H.; Guiseppi-Elie, A. Curr. Opin. Biotechnol. 2001, 12, 41. (c) Wang, J. Nucleic Acids Res. 2000, 28, 3011. (d) Freeman, W. M.; Robertson, D. J.; Vrana, K. E. Biotechniques 2000, 29, 1042. (e) Niemeyer, C. M.; Blohm, D. H. Angew. Chem., Int. Ed. 1999, 38, 2865.

3902 Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

molecular modification. As model systems, they are ideal for the study of long-range electron transfer through the DNA double helix, as well as for the correlation of DNA structures and their biomolecular recognition properties.2 An area for continued investigation in DNA-chip technology is quality control for the attachment, structure, and heterogeneous hybridization/denaturation efficiency of DNA probes on the substrate. Many different techniques have been applied, including 32P-radiolabeling, X-ray photoelectron spectroscopy, surface plasmon resonance spectroscopy, and infrared spectroscopy, as well as fluorescence-based methods.2a,3 Electrochemical approaches have also been explored based on the synthesis of electroactive oligodeoxynucleotides,2c-d,4 intercalation of redox-active labels,2b,3b chronocoulometric response of electrostatically bound species,5 and amplified electrocatalytical signals.6 However, these methods are often not fast or simple enough to be used as part of routine laboratory operations by nonspecialists. In this paper, we report a simple and straightforward procedure based on the voltammetric response of multiply charged transition metal molecules bound electrostatically to DNA-modified surfaces. The key feature of this approach is the easy distinction of the surface waves from the signal of diffused species when using micromolar concentrations of the redox molecules. This research stems from the continued (2) For examples of studies of DNA-modified surfaces, see: (a) Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 8916. (b) Kelley, S. O.; Jackson, N. M.; Hill, M. G.; Barton, J. K. Angew. Chem., Int. Ed. 1999, 38, 941. (c) Yu, C. J.; Wan, Y. J.; Yowanto, H.; Li, J.; Tao, C. L.; James, M. D.; Tan, C. L.; Blackburn, G. F.; Meade, T. J. J. Am. Chem. Soc. 2001, 123, 11155. (d) Kertesz, V.; Whittemore, N. A.; Chambers, J. Q.; McKinney, M. S.; Baker, D. C. J. Electroanal. Chem. 2000, 493, 28. (e) Hartwich, G.; Caruana, D. J.; de Lumley-Woodyear, T.; Wu, Y. B.; Campbell, C. N.; Heller, A. J. Am. Chem. Soc. 1999, 121, 10803. (f) Pike, A. R.; Lie, L. H.; Eagling, R. A.; Ryder, L. C.; Patole, S. N.; Connolly, B. A.; Horrocks, B. R.; Houlton, A. Angew. Chem., Int. Ed. 2002, 41, 615. (3) For examples, see: (a) Peterlinz, K. A.; Georgiadis, R. M.; Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 3401. (b) Kelley, S. O.; Barton, J. K. Bioconjugate Chem. 1997, 8, 31. (c) Brewer, S. H.; Anthireya, S. J.; Lappi, S. E.; Drapcho, D. L.; Franzen, S. Langmuir 2002, 18, 4460. (d) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A., III; Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535. (4) Whittemore, N. A.; Mullenix, A. N.; Inamati, G. B.; Manoharan, M.; Cook, P. D.; Tuinman, A. A.; Baker, D. C.; Chambers, J. Q. Bioconjugate Chem. 1999, 10, 261. (5) (a) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670. (b) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Bioconjugate Chem. 1999, 10, 419. (6) (a) Patolsky, F.; Weizmann, Y.; Willner, I. J. Am. Chem. Soc. 2002, 124, 770. (b) Patolsky, F.; Katz, E.; Willner, I. Angew. Chem., Int. Ed. 2002, 41, 3398. 10.1021/ac034318w CCC: $25.00

© 2003 American Chemical Society Published on Web 06/12/2003

efforts by many research groups to study both fundamental and practical aspects of the interaction between redox-active cations and DNA7-12 and from our ultimate goal of developing varieties of DNA-based sensing devices allowing direct electrical and electrochemical measurements.13 In retrospect, voltammetric measurements have long been used to characterize chemically modified electrodes and have been particularly successful in the investigation of the quantitative aspects of the partition of redox ions into polymer thin films.14 RESULTS AND DISCUSSION Preliminary Studies. As depicted in Figure 1, exposure of DNA-modified electrodes to an aqueous solution of a redox cation, such as ruthenium hexaammine, [Ru(NH3)6]3+, at a low ionic strength leads to an ion exchange equilibrium between [Ru(NH3)6]3+ and the native charge compensation ions (presumably Na+) associated with the anionic DNA backbone. The reaction assumes that penetration of the DNA film by [Ru(NH3)6]3+ results in ejection of Na+ on a 1:1 (charge) basis. The surface concentration, ΓRu (mol/cm2), of the [Ru(NH3)6]3+ ions incorporated into the DNA film depends on the charge, size, and concentration C (M) in the bathing solution, as well as the number of bases and the surface density, ΓDNA (molecules/cm2), of oligonucleotides immobilized on the electrode surfaces. The cyclic voltammograms that result from partitioning [Ru(NH3)6]3+ into DNA monolayers are shown in Figure 1 for a series of [Ru(NH3)6]3+ concentrations (in a 10 mM Tris buffer at pH 7.4). Gold electrodes modified with single-stranded thiolate DNA (Au-S-C6-TCGATCTGACGTCAGCTAAA, ssDNA) were prepared by denaturation of the double helices (dsDNA) to remove the complementary strand.15 In each case, the cathodic peak that represents the one-electron reduction of [Ru(NH3)6]3+ to [Ru(NH3)6]2+ systematically increases by equilibration of the DNA monolayer with the higher concentration of [Ru(NH3)6]3+ in the (7) (a) Carter, M. T.; Bard, A. J. J. Am. Chem. Soc. 1987, 109, 7528. (b) Carter, M. T.; Rodriguez, M.; Bard, A. J. J. Am. Chem. Soc. 1989, 111, 8901. (8) Millan, K. M.; Mikkelsen, S. R. Anal. Chem. 1993, 65, 2317. (9) Wang, J.; Cai, X.; Rivas, G.; Shiraishi, H.; Faria, P. A. M.; Donta, N. Anal. Chem. 1996, 68, 2629. (10) Holmlin, R. E.; Stemp, E. D. A.; Barton, J. K. J. Am. Chem. Soc. 1996, 118, 5236. (11) (a) Pang, D. W.; Abrun ˜a, H. D. Anal. Chem. 1998, 70, 3162. (b) Pang, D. W.; Abrun ˜a, H. D. Anal. Chem. 2000, 72, 4700. (12) (a) Horrocks, B. R.; Mirkin, M. V. Anal. Chem. 1998, 70, 4653. (b) Aslanoglu, M.; Isaac, C. J.; Houlton, A.; Horrocks, B. R. Analyst 2000, 125, 1791. (13) Fahlman, R. P.; Sen, D. J. Am. Chem. Soc. 2002, 124, 4610 and references therein. (14) (a) Oyama, N.; Anson, F. C. Anal. Chem. 1980, 52, 1192. (b) Peerce, P. J.; Bard, A. J. J. Electroanal. Chem. 1980, 112, 97. (c) Schneider, J. R.; Murray, R. W. Anal. Chem. 1982, 54, 1508. (d) Guadalupe, A. R.; Abrun ˜a, H. D. Anal. Chem. 1985, 57, 142. (15) Gold surfaces bearing high-density single-stranded oligonucleotide sequences can be prepared by two different methods. One is to immobilize single-stranded thiolate DNA directly from its dilute solution via selfassembly and then treat it with a dilute alkanethiol solution for surface passivation (refs 2a, 3a, and 5a). The key is the precise control of the time for deposition of thiolate ssDNA and exposure to alkanethiol in order to achieve both the high density (for higher sensitivity) and near-unity hybridization efficiency. An equilibrated adsorption would yield very dense ss-DNA monolayers on the surface (∼10 × 1012 molecules/cm2), which would exhibit a considerably lower hybridization ratio as result of spatial hindrance.5a The other method, as demonstrated in this work (see also ref 3b), is to prepare a dsDNA-modified surface first and then remove the complementary strand, which naturally produces an ssDNA-modified surface with optimal surface coverage for hybridization.

Figure 1. Pictorial representation and voltammetric response of multiply charged transition metal complexes bound electrostatically to DNA-modified surfaces: cyclic voltammograms of gold electrodes modified with dsDNA (A) and ssDNA (B) in 10 mM Tris buffer (pH 7.4) in the presence of [Ru(NH3)6]3+ at different concentrations as listed. The scan rate was 50 mV/s.

solution. Both cathodic and anodic peaks are well defined and symmetric, with negligible peak separations at a scan rate of 50 mV/s. It is noticeable that the anodic peaks are slightly smaller than the corresponding cathodic waves. In addition, the responses from electrodes modified with dsDNA monolayers are much stronger than from those modified with ssDNA. Repeated potential cycles did not change the features shown in Figure 1, indicating that the DNA probes remain intact. At potential scan rates in the range of 10-500 mV/s, the peak currents of cyclic voltammograms shown in Figure 1 are linearly dependent on the scan rate. In principle, for adsorbed electroactive species, the peak current of their surface waves should be proportional to the scan rate, υ, in contrast to the υ1/2 dependence observed for redox peaks of diffusing species.16 The charges under the peaks (and thus the measured quantity of partitioned [Ru(NH3)6]3+) are independent of the potential scan rate. These are typical features of Nernstian equilibration of the electroactive [Ru(16) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2001.

Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

3903

Figure 2. Electrochemical response of surface-confined and solution-diffused multiply charged redox cations: cyclic voltammograms of [Ru(NH3)6]3+ at different concentrations in 10 mM Tris buffer (pH 7.4) at 50 mV/s on dsDNA-modified (solid lines) and bare (dashed lines) gold electrodes.

(NH3)6]3+ species, in case they are electrostatically trapped in the DNA films with potential applied to the gold electrode. To further confirm the negligible disturbance by the redox reaction of solution species that diffuse to the electrode surfaces, we compared the cyclic voltammograms of bare gold electrodes and those modified with DNA monolayers in the presence of [Ru(NH3)6]3+ at relatively high concentrations (3.5-12.0 µM). Figure 2 shows that the latter surface waves are much more distinct, and the capacitive currents are more compressed relative to the diffusing responses. The influence from diffused species can be even further suppressed by the application of high-potential scan rates. Quantitation of DNA Probes. The above experimental data demonstrate that the voltammetric ΓRu value is a direct measure of the total amount of [Ru(NH3)6]3+ on the DNA-modified electrodes. It can be easily calculated from

Q ) nFAΓRu

(1)

where Q is the charge obtainable by integration of the redox peaks in the cyclic voltammograms (Figures 1 and 2),17 n is the number of electrons in the reaction, and A is the area of the working electrode. The adsorption isotherms of [Ru(NH3)6]3+ on DNA-modified surfaces were obtained by determination of the charges of surfaceconfined redox species at different solution concentrations. As shown in Figure 3A, the saturated charge values of DNA-modified electrodes become much lower after denaturation of the doublestranded to single-stranded oligonucleotides. The saturated surface (17) To be more accurate, at these relatively high concentrations (Figure 2), the area of the left half of the peak (with negligible contribution from the diffusing peaks) was measured and multiplied by 2.

3904

Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

Figure 3. Ionic binding activities of [Ru(NH3)6]3+ on DNA-modified surfaces. (A) Adsorption isotherms of [Ru(NH3)6]3+ on gold electrodes modified with dsDNA (b) and ssDNA (O). The surface charge due to adsorbed [Ru(NH3)6]3+ was determined by integration of the cathodic current (Figure 1). The solid lines are only to guide the eyes. (B) Linearized adsorption isotherms of [Ru(NH3)6]3+ on gold electrodes modified with dsDNA (b) and ssDNA (O) based on the Langmuir model. The lines are the best linear fits to the experimental data (method of least-squares) from which the binding constants were determined.

concentrations of the adsorbed redox marker, ΓRu, were (10.1 ( 1.2) × 10-11 and (5.9 ( 0.4) × 10-11 mol/cm2 for gold electrodes modified with dsDNA and ssDNA, respectively. Under the condition of saturation, the measured value can be directly converted to the surface density of DNA using the relationship

ΓDNA ) ΓRu(z/m)NA

(2)

where m is the number of nucleotides in the DNA, z is the charge of the redox molecules, and NA is Avogadro’s number. Evaluated from the ΓRu values, the surface densities, ΓDNA, for electrodes modified with dsDNA ((4.6 ( 0.6) × 1012 molecule/cm2) and ssDNA ((5.3 ( 0.4) × 1012 molecule/cm2),18 are close to the theoretical estimation (5.2 × 1012 molecule/cm2).2e This is the calculated maximal surface density assuming that the DNA monolayers are compact and the double helices “stand” on the surface with a small tilt angle relative to the normal. Of particular interest, the voltammetrically determined ΓDNA value for electrodes

modified with ssDNA is consistent with the reported optimal coverage for hybridization as the electrodes were prepared by direct adsorption of thiolate ssDNA and subsequent exposure to a dilute 6-mercapto-1-hexanol (MCH) solution.2a,3a,5a,15 The validity of electrochemical methods for the quantitation of DNA probes relies on the following assumptions: (1) the redox cations bind to the DNA strictly through electrostatic interactions; (2) the amount of trapped redox cations can be accurately determined; and (3) the charge compensation of the DNA phosphate functionalities is provided solely by the redox cations. We are confident that our experimental conditions fulfill these assumptions. This is supported by the strong binding of [Ru(NH3)6]3+ to the DNA-modified electrodes as evidenced by the low concentrations (0.1-12 µM) usable in the measurements (also see the Cationic Binding Activity section for a quantitative analysis). Based on the electrostatic interaction between metal cations and DNA on electrodes, a chronocoulometric protocol has been proposed by Tarlov and co-workers for the quantitation of oligonucleotide probes on chips.5 However, the attractive aspects of the voltammetric quantitation described in this article are the simplicity and ease of interpretation of the results compared to other electrochemical methods. Cyclic voltammetry has been a popular technique for initial electrochemical studies of new systems and has proven useful for gathering information about complicated electrode reactions. The main advantages of this procedure over the chronocoulometric scheme are the easier application of corrections for background contributions to the charge and the more direct route to electrode-transfer rate constants (as described below). Considering the different approaches to the preparation of ssDNA-modified electrodes with optimal surface density for hybridization,15 cyclic voltammetry was also used to monitor the surface hybridization. In this case, the modified gold electrodes were prepared by direct adsorption of the thiolate ssDNA for a specific period and subsequent exposure to the target complementary strand to prepare double helices on the surface. Preliminary results confirmed that the voltammetric procedure could be used to successfully monitor the immobilization of ssDNA probes, MCH passivation, and hybridization reactions on electrode surfaces. Cationic Binding Activity. Both adsorption isotherms of [Ru(NH3)6]3+ shown in Figure 3A exhibit identical general features, despite the fact that the maximum charge obtained from electrodes modified with ssDNA is significantly less than that from those modified with dsDNA. To further quantify the cationic binding activities on DNA-modified surfaces, we determined the binding constants of [Ru(NH3)6]3+ based on the Langmuir model.5b This classical model assumes that every binding site is equivalent and that the ability of a molecule to bind is independent of occupation of nearby sites. A linearized form (Figure 3B) of the Langmuir isotherm in terms of the redox cation concentration, (18) We noticed that the surface density of DNA probes on electrodes modified with dsDNA ((4.6 ( 0.6) × 1012 molecule/cm2) was slightly lower than that of ssDNA ((5.3 ( 0.4) × 1012 molecule/cm2), within the experimental uncertainties. This could be due to incomplete heterogeneous denaturation of the duplex or to imperfect hybridization when the duplex was prepared in solution before self-assembly on the electrode surfaces. These are challenging topics in DNA chip technology for which further investigations are in progress in our laboratory.

C, the accumulated charge at the electrode surface, Q, and the saturated charge, Qsat, is given by eq 3:

1 C C + ) Q Qsat KQsat

(3)

The calculated binding constants, K, (2.0 ( 0.5) × 106 M-1 for the dsDNA-modified and (1.5 ( 0.5) × 106 M-1 for the ssDNAmodified surfaces, respectively, are consistent with those reported previously.5b The larger binding constant obtained for the former can be understood in terms of its higher negative charge density (see Figure 1). It should be noted that the voltammetric procedure provides not only an accurate approach to evaluate the cationic binding activities but also a convenient way to study ion exchange selectivities of DNA-modified electrodes for different redox cations. For this purpose, competitive measurements can be carried out in electrolyte solutions containing paired redox cations (experiments in progress in our laboratory).14c Electron-Transfer Kinetics. Another important feature of the present approach is the study of electron-transfer kinetics at DNAmodified electrode surfaces. In particular, the electron-transfer rate constants of [Ru(NH3)6]3+ at both dsDNA- and ssDNAmodified gold electrodes can be readily determined based on the Laviron classical model,19

Epc ) E°′ Epa ) E°′ -

RnFυ RT ln RnF RTk

(

(

)

(4)

)

(1 - R)nFυ RT ln RTk (1 - R)nF

(5)

where Epc and Epa are the cathodic and anodic peak potentials, respectively, E°′ is the formal potential, n is the number of electrons transferred, and R is the transfer coefficient for the cathodic process. R, T, and F have their usual meanings. The simplest analytical procedure consists of recording cyclic voltammograms at different scan rates (υ) and subsequent plotting of the relative peak potentials (E°′ - Ep) versus lnυ. The electrontransfer rate constant, k, and the transfer coefficient, R, can be derived from an analysis of the linear part of the plot at the higher scan rates. Figure 4 shows the Laviron plots for [Ru(NH3)6]3+. The k and R were found to be 2.2 ( 0.2 s-1 and 0.48 ( 0.05, respectively, for the electrodes modified with dsDNA, and 2.8 ( 0.4 s-1 and 0.55 ( 0.09, respectively, for those modified with ssDNA. It is not surprising that the electron-transfer rate constants of these one-dimensionally aligned 20-mer single- and doublestranded oligonucleotide monolayers are 2 orders of magnitude smaller than those reported for electrodes coated with bulk disordered DNA molecules.11b Previous scanning probe microscopy studies confirmed that the thiolate DNA helices form an ordered and closely packed monolayer on gold with a tilt angle of 45° with respect to the surface normal, and the orientation of DNA helices is potential dependent.20 In addition, the electrontransfer rate constant for the ssDNA-modified electrodes is slightly larger than that for dsDNA-modified electrodes, which emphasizes (19) Laviron, E. J. Electroanal. Chem. 1979, 101, 19.

Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

3905

It should be noted that the advantages of the voltammetric method for the study of DNA-modified surfaces go beyond simplicity and accuracy, in comparison with other techniques.3,5 As demonstrated experimentally, voltammetric responses are sensitive to chemical reactions (e.g., ligand binding) coupled to the electron-transfer steps; therefore, this method is also capable of supplying kinetic data of the adsorbed species, which are relatively difficult to get from chronocoulometric measurements.23 More importantly, this method avoids the complexities inherent in the synthesis of electroactive oligonucletides2c,d,4 and in the stoichiometry of the intercalated redox complexes.2b,3b The present approach deserves further investigation regarding the binding activity of molecular analytes on DNA-modified surfaces and the structure-dominated electron transport in the DNA double helix. These are fundamental and essential perspectives in the development of DNA-based sensing technology. CONCLUSIONS In summary, we have explored the application of a simple voltammetric procedure for the examination of DNA-modified surfaces. The surface density of DNA probes can be accurately determined by integration of the well-defined redox peaks resulting from electrostatically trapped, multiply charged transition metal complexes (e.g., [Ru(NH3)6]3+). Furthermore, binding constants for the metal complexes with DNA on surfaces can be derived from the adsorption isotherms, while the electron-transfer kinetics can be evaluated based on a classical model. This useful and straightforward protocol provides a versatile and powerful tool for the investigation of the immobilization, structure, and reaction of DNA probes on conductive substrates. Figure 4. Laviron plots for 3.5 µM [Ru(NH3)6]3+ in 10 mM pH 7.4 Tris-HCl buffer at (A) dsDNA- and (B) ssDNA-modified gold electrodes.

the importance of molecular orientation (the DNA monolayers become more disordered following the denaturation from duplexes into single strands). With respect to the ongoing discussion about DNA-mediated charge transport,21 a satisfactory explanation for the rather small difference is not available at the present time. One possibility is that not the DNA helices but the [Ru(NH3)6]3+ cations aligned along the oligonucleotides are transporting electrons. However, it has been shown that the interaction with oligonucleotides will change the voltammetric response of metal cations in electrolyte solutions.7-12 Particularly, the dependence of electron-transfer rate constants on the base stacking of oligonucleotides has been examined by cyclic voltammetry of Ru(bpy)32+ (bpy ) 2,2′-bipyridine) in the presence of oligonucleotides.22 To clarify the detailed electron-transfer mechanism, further investigation of different base-stacking designs of oligonucleotide probes appears warranted. (20) (a) Kelley, S. O.; Barton, J. K.; Jackson, N. M.; McPherson, L. D.; Potter, A. B.; Spain, E. M.; Allen, M. J.; Hill, M. G. Langmuir 1998, 14, 6781. (b) Sam, M.; Boon, E. M.; Barton, J. K.; Hill, M. G.; Spain, E. M. Langmuir 2001, 17, 5727. (21) For recent reviews of charge transport in DNA, see: (a) Giese, B.; Spichty, M. ChemPhysChem 2000, 1, 195. (b) Schuster, G. B. Acc. Chem. Res. 2000, 33, 253. (22) Sistare, M. F.; Codden, S. J.; Heimlich, G.; Thorp, H. H. J. Am. Chem. Soc. 2000, 122, 4742.

3906 Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

EXPERIMENTAL SECTION The DNA oligomers were synthesized at Core DNA Services, Inc. (Calgary, AB, Canada). For the purpose of self-assembly of the oligomers on gold, a 5′-thiolate modification to one of the strands was needed, for which the 5′-Thiol-Modifier C6 S-S from Glen Research (Sterling, VA) was used. For the thiolate oligomer, a DMT-on Synthesis serves the dual purpose of protecting the disulfide and providing an appropriate hydrophobic group for HPLC purification. The sequences are DMTO-C6-S-S-C6-5′TCGATCTGACGTCAGCTAAA3′ for the thiolate and 5′TTTAGCTG ACGTCAGATCGA3′ for the complementary strand. These DNA sequences were designed to include approximately equal numbers of purine and pyrimidine bases. Gold substrates were purchased from Evaporated Metal Films Inc. (Ithaca, NY); [Ru(NH3)6]Cl3 (98%) was obtained from Sigma-Aldrich (Milwaukee, WI). Deionized water (>18.3 MΩ‚cm) was from a Barnstead EasyPure UV/ UF compact water system (Dubuque, IA). Thiolate oligomers were deprotected with saturated aqueous ammonia for 12 h at 55 °C. The resulting oligomer was purified (23) In the determination of adsorbed reactants by chronocoulometry, reactants arrive at the diffusion-limited rate: Anson, F. C.; Osteryoung, R. A. J. Chem. Educ. 1983, 60, 293. Heterogeneous kinetic data can be obtained by using a step potential that is insufficiently extreme to enforce diffusion-controlled electrolysis throughout the experimental time domain and analyzed by appropriate theoretical considerations: Christie, J. H.; Lauer, G.; Osteryoung, R. A. J. Electroanal. Chem. 1964, 7, 60. It is not easy to measure kinetic parameters of the adsorbed reactant in the presence of diffusing reactants at concentrations high enough to contribute significantly to the measured charges.

by reversed-phase HPLC on a C18 Vydac column (218TP54). The purified sample was then treated with 100 mM dithiothreitol at pH 8.5 for 30 min in order to reduce the disulfide and desalted on a Pharmacia Nap-5 Column (G-25 Sephadex). The 10 µM solution of the duplex was hybridized in deoxygenated 10 mM Tris buffer at pH 7.4 by heating to 90 °C followed by slow cooling to room temperature. Gold substrates were cleaned by immersion in a 3:1 mixture of concentrated H2SO4 and 30% H2O2 for 5 min at ∼90 °C and then rinsed with copious amounts of water. Double-stranded DNAmodified electrodes were prepared by spreading a drop of 10 µM duplex solution on the cleaned gold surface for 18-48 h at ambient conditions. Modified electrodes were rinsed with Tris buffer and dried under N2 before characterization. Single-stranded DNA electrodes were prepared by denaturation of double-stranded DNA samples using 8 M urea incubation at 90 °C for 2 min and rinsing several times with 90 °C hot water. It has been confirmed that the desorption of thiols on gold electrodes is negligible under similar conditions, although the structure of the monolayers could be distorted.24 Cyclic voltammetry was performed using a µAutolab II potentiostat/galvanostat (Eco Chemie B.V., Utrecht, Netherlands). The (24) For an example, see: Badia, A.; Lennox, R. B.; Reven, L. Acc. Chem. Res. 2000, 33, 475.

working electrode (DNA-modified gold substrates) was pressed against an opening in the cell bottom using an O-ring seal with an exposed area of 0.69 cm2. The electrolyte solution was degassed with argon for 15 min. All potentials are reported with respect to a Ag|AgCl|3M NaCl reference electrode, and the counter electrode was a Pt wire. ACKNOWLEDGMENT This research was financially supported by the Natural Science and Engineering Research Council of Canada (NSERC) and Simon Fraser University. The authors are grateful to Dr. Fred C. Anson (Caltech) for valuable discussions and informative comments, and to Miss Li Su for her help with some of the experiments. SUPPORTING INFORMATION AVAILABLE Plots of the peak currents for DNA-modified gold electrodes versus scan rate, ion-exchange voltammetry, and voltammetric quantitation of surface hybridization. This material is available free of charge via the Internet at http://pubs.acs.org. NOTE ADDED AFTER ASAP An error on the Figure 4 axis was corrected and the paper re-posted on June 20, 2003. Received for review March 29, 2003. Accepted April 30, 2003. AC034318W

Analytical Chemistry, Vol. 75, No. 15, August 1, 2003

3907