Enzyme-Modulated Cleavage of dsDNA for Supramolecular Design of

Fei Yan and Omowunmi A. Sadik*. Department of Chemistry, State University of New York at Binghamton, P.O. Box 6016, Binghamton, New York 13902-6016...
0 downloads 0 Views 147KB Size
Download PDF
Anal. Chem. 2001, 73, 5272-5280

Enzyme-Modulated Cleavage of dsDNA for Supramolecular Design of Biosensors Fei Yan and Omowunmi A. Sadik*

Department of Chemistry, State University of New York at Binghamton, P.O. Box 6016, Binghamton, New York 13902-6016

Supramolecular docking and immobilization of biotinylated dsDNA onto a self-assembled monolayer of avidin have been measured using impedance spectroscopy and quartz crystal microbalance technique. The formation of the serial assembly was first achieved by linearizing circular plasmid dsDNA using BamH I endonuclease enzyme. This was followed by a bisulfite-catalyzed transamination reaction in order to biotinylate the dsDNA. The reaction is single-strand specific, and it specifically targets unpaired cytosine bases generated during the enzyme cleavage. The biotinylated dsDNA was then used as a ligand at a gold electrode containing avidin. The process was monitored by ac impedance spectroscopy that was used to probe the changes in interfacial electron-transfer resistance upon binding and a microgravimetric quartz crystal microbalance that reflected in situ mass changes on the dsDNA-functionalized substrates. Our results demonstrated that this approach could be employed for the determination of small-molecular-weight organics such as cisplatin, daunomycin, bisphenol A, chlorinated phenols, and ethidium bromide. A detection limit in the magnitude of ca. 10 nM was achieved. This immobilization technique provides a generic approach for dsDNAbased sensor development and for the monitoring of DNA-analyte interactions. The recent development of DNA biosensors has attracted substantial research efforts directed toward clinical diagnostics as well as forensic and biomedical applications.1 The highly specific, interchain binding in nucleic acids has been used extensively with selected target molecules for hybridization and bimolecular interactions.2 This property can also be exploited for biosensor application by using DNA or RNA molecules as the capturing agents to generate analyte-dependent signals. In chemical analysis, the application of nucleic acids as ligands is still a new area of investigation with only a few specific examples to date, primarily in clinical diagnostic.3 DNA and RNA have been * To whom correspondence should be addressed. Phone: 607-777-4132. Fax: 607-777-4478. E-mail: [email protected]. (1) (a) Landergren, U.; Kraiser, C. T.; Hodd, L. Science 1988, 242, 229. (b) Hashimoto, K.; Ito, K.; Ishimori, Y. Anal. Chem. 1994, 66, 3820-3833. (c) Yang, M.; McGovern, M. E.; Thompson, M. Anal. Chim. Acta 1997, 346, 259-275. (d) Wilson, E. K. Chem. Eng. News 1998, 76, 47-49. (e) Downs, M. E. A. Biochem. Soc. Trans. 1991, 19, 39. (f) McGown, L. B.; Joseph, M. J.; Pitner, J. B.; Vonk, G. P.; Linn, C. P. Anal. Chem. 1995, 67, 663A668A. (2) Christopoulos, T. K. Anal. Chem. 1999, 71, 425R-438R.

5272 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001

identified as the primary targets of many small molecules, especially in matrixes of clinical importance, for example, cisplatin,4 daunomycin,5 and ethidium bromide.6 In some cases, the toxic actions of many carcinogenic and mutagenic pollutants have been attributed to their interaction with DNA.7 However, few studies have been devoted to the design of environmental DNA biosensors.8 The design of biosensors requires the successful immobilization of the biological reagents. A number of approaches for immobilizing dsDNA layers on electrodes have been reported,9 yet the quest for a molecularly organized but reproducible immobilization (of dsDNA) continues to pose a challenge. In a previous study, we demonstrated that a multilayer bis-biotinylated dsDNA could be immobilized onto metal substrates after endonuclease cleavage and that the resulting substrate could be used to study the bimolecular interaction of DNA with small molecules.10 In this report, we will show that the supramolecular assembly of dsDNA on solid substrates could provide a generic format for detecting small-molecular-weight organics. Hence, we have selected analytes that are of interest to clinical and environmental investigations. These include polychlorinated biphenyls (PCBs), chlorophenols, bisphenol A, cisplatin, daunomycin, ethid(3) Wang, J. Nucleic Acids Res. 2000, 28, 3011-3016. (4) (a) Lippard, S. J. Science 1982, 218, 1075. (b) Sherman, S. E.; Gibson, D.; Wang, A. H.-J.; Lippard, S. J. Science 1985, 230, 412. (c) Eastman, A. Biochemistry 1986, 25, 3912. (d) Takahara, P. M.; Frederick, C. A.; Lippard, S. J. J. Am. Chem. Soc. 1996, 118, 12309-12321. (5) (a) Chairs, J. B.; Dattagupta, N.; Crothers, D. M. Biochemistry 1982, 21, 3933-3940. (b) Liu, L. F. Annu. Rev. Biochem. 1989, 58, 351-375. (c) Frederick, C. A.; Williams, L. D.; Ughetto, G.; van der Marel, G. A.; van Boom, J. H.; Rich, A.; Wang, A. H.-J. Biochemistry 1990, 29, 2538-2549. (d) Taaties, D. J.; Gaudiano, G.; Resing, K.; Koch, T. H. J. Med. Chem. 1996, 39, 5135-4138. (6) (a) Jones, R. L.; Lanier, A. C.; Keel, R. A.; Wilson, W. D. Nucleic Acids Res. 1980, 8, 1613. (b) Saenger, W. Principles of Nucleic Acid Structure; Springer-Verlag: New York, 1984. (c) Meyer-Almes, F. J.; Porschke, D. Biochemistry 1993, 32, 4246-4253. (7) Lambert, B.; Le Pecq, J. In DNA-Ligand Interactions: From Drugs to Proteins; Guschlbauer, W., Saenger, W., Eds.; Plenum Press: New York, 1986, 141-156. (8) (a) Pandey, P. C.; Weetall, H. H. Anal. Chem. 1995, 67, 787. (b) Wang, J.; Rivas, G.; Luo, D.; Cai, X.; Valera, F. S.; Dontha, N. Anal. Chem. 1996, 68, 4365-4369. (9) (a) Ghosh, S. S.; Musso, G. F. Nucleic Acids Res. 1987, 15, 5353. (b) Moser, I.; Schalkhammer, T.; Pittner, F.; Urbant, G. Biosens. Bioelectron. 1997, 12, 729-737. (c) Higashi, N.; Takahashi, M.; Niwa, M. Langmuir 1999, 15, 111-115. (d) 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, 6781-6784. (e) Pang, D.; Zhang, M.; Wang, Z.; Qi, Y.; Cheng, J.; Liu, Z. J. Electroanal. Chem. 1996, 403, 183-188. (f) Bamdad, C. Biophys. J. 1998, 75, 1997-2003. (g) Yang, M.; Yau, H. C. M.; Chan, H. L. Langmuir 1998, 14, 6121-6129. (10) Yan, F.; Sadik, O. A. J. Am. Chem. Soc. 2001, in press. 10.1021/ac015516v CCC: $20.00

© 2001 American Chemical Society Published on Web 10/04/2001

ium bromide, and formaldehyde. dsDNA was immobilized after linearization with endonuclease cleavage, and the characterization was carried out using circular dichroism, gel electrophoresis, UV/ vis spectroscopy, and HPLC protocols. Faradaic impedance spectroscopy and mass measurements clearly showed that the successful immobilization and interactions of dsDNA were accomplished at nanomolar levels of the small organic molecules tested. Therefore, we indicate that this method is capable of offering a very general approach to the design of dsDNA-based sensors. MATERIALS AND METHODS Avidin and biocytin hydrazide was obtained from Pierce Chemicals (Rockford, IL), and PCBs (Aroclor standards 1254, 1242, and 1221) were purchased from Ultra Scientific, North Kingstown, RI. The 2961bp pBluescript II SK (-) phagemid dsDNA vector was purchased from Stratagene Inc. (La Jolla, CA). All other chemicals were obtained from either Sigma or Aldrich (Milwaukee, WI). These include bisphenol A, chlorinated phenols (o-chlorophenol, 2,4,5-trichlorophenol, and 2,4,6-trichlorophenol), ethidium bromide, daunomycin, and cisplatin. All chemicals were used as received from the vendors without further purification. Electrochemical measurements were carried out using a 2-mL, 3-electrode electrochemical cell with an EG&G lock-in amplifier (model 5210) combined with an EG&G 263A potentiostat. Data acquisition was conducted using EG&G M270 and M398 software for cyclic voltammetry and impedance measurement, respectively. The impedance measurements were carried out in the presence of 0.1 M phosphate buffer (pH 7.0) or a 10 mM K3[Fe(CN)6]/ K4[Fe(CN)6] (1:1) mixture as a redox probe at the formal potential of the system (E0 ) 200 mV) using an alternating voltage of 10 mV. In addition, the frequency range during impedance measurements was from 100 to 10 kHz (otherwise, as indicated). The impedance spectra were plotted using complex plane diagrams (or Nyquist plots). Experimental impedance data were exported from the EG&G M398 software using the Boukamp-mode procedure, and these were transferred to a software package using Equivalent Circuit (EQUIVCRT.PAS version 4.61). The circuit utilizes nonlinear least squares fit (NLLSF). Electrochemical experiments were conducted using a well-type, minicell fitted with a gold-coated quartz crystal electrode as the working electrode (0.196 cm2). Potentials were measured relative to an aqueous, saturated Ag/AgCl double junction (as reference electrode) and a platinum wire (auxiliary electrode). All measurements were carried out at ambient temperature (22 ( 2 °C). An open-circuit system was used for the QCM measurements in which only the gold-coated, quartz crystal, working electrode was connected in the absence of auxiliary and reference electrodes. The Au-coated quartz crystals (AT-cut, 9 MHz) of 0.2-cm2 geometric area/face were obtained from EG&G Instruments (Princeton Applied Research). The resonant frequency was determined using a QCA 917 quartz crystal analyzer (Seiko EG&G). RESULTS AND DISCUSSIONS To generate a specific chemical modification of the nucleic acids, we used a method involving the nucleophilic attack of the cytosine, as described by Shapiro et al.11 The reaction consisted of a bisulfite-catalyzed transamination of the cytosine and cytidine,

Scheme 1. Reaction of Biocytin Hydrazide with Cytosine (Cytidine) Residues of dsDNA

which gave rise to N4-substituted molecules, as shown in Scheme 1. This transamination reaction was reported to be single-strandspecific12 and had been widely used in analytical biochemistry to prepare non-isotopic probes for nucleic acid hybridization assays. Presently, it appears that this reaction has not yet been reported for DNA immobilization on solid substrates or for biosensor purposes. Our experiments involved the overnight digestion of a 2961bp pBluescript II SK (-) phagemid dsDNA vector using endonuclease enzyme BamH I. The endonuclease enzyme possessed only one cleavage site for the phagemid dsDNA vector. The specific recognition sequence of 5′-GVG-A-T-C-C-3′ obtained with BamH I was used to cleave the dsDNA, and this resulted in welldefined fragments having free cytosines, as shown in Scheme 2.13 It was possible to verify the restriction cleavage of the dsDNA using HPLC, electrophoresis in 1.0% agarose gels, and circular dichroism spectroscopy.14 Our circular dichroism experiments showed evidence of such a single strand-specific reaction.10 Scheme 3 shows a schematic representation of the formation of a biotinylated monolayer on gold-coated quartz crystal electrodes. The stability of the self-assembled avidin monolayer was improved by using an underpotentially deposited silver adlayer.16 The serial docking of the biotinylated plasmid dsDNA onto avidin layers was shown to have increased the sensitivity of the cisplatin measurement.10 The enzymatic hydrolysis of the surface-functionalized dsDNA that was bound to the first avidin layer was investigated using 10 mM Tris-HCl buffer (pH 7.5) with 10 mM MgCl2 and 1 mM dithioerythritol (DTE). The understudied plasmid dsDNA has one cleavage site for restriction endonuclease Kpn I, which can specifically hydrolyze a double-stranded 5′-GGTAC/C-3′ se(11) Shapiro, R.; Weisgras, J. M. Biochem. Biophys. Res. Commun. 1970, 40, 839-843. (12) (a) Goddard, J. B.; Schulman, L. H. J. Biol. Chem. 1972, 247, 3864-3867. (b) Shapiro, R.; Braveman, B.; Louis, J. B.; Servis, R. E. J. Biol. Chem. 1973, 248, 4060-4064. (c) Peden, K. W. C.; Nathans, D. Proc. Natl. Acad. Sci. U.S.A. 1982, 79, 7214-7217. (d) Reisfeld, A.; Rothenberg, J. M.; Bayer, E. A.; Wilchek, M. Biochem. Biophys. Res. Commun. 1987, 142, 519-526. (13) Roberts, R. J.; Wilson, G. A.; Young, F. E. Nature 1977, 265, 82-84. (14) Maniatis, T.; Fritsch, E. F.; Sambrook, J. In Molecular Cloning: A Laboratory Manual; Cold Spring Harbor Laboratory: New York, 1982. (15) Avignolo, C.; Valente, P.; Cai, S.; Roner, R.; Fulle, A.; Pizzorno, G.; Bignone, F. A. Biochem. Biophys. Res. Commun. 1990, 170, 243-250. (16) (a) Jennings, G. K.; Laininis, P. E. J. Am. Chem. Soc. 1997, 119, 52085214. (b) Jennings, G. K.; Laininis, P. E. Langmuir 1996, 12, 6173. (c) Burgess, J. D.; Hawkridge, F. M. Langmuir 1997, 12, 3781-3786.

Analytical Chemistry, Vol. 73, No. 21, November 1, 2001

5273

Scheme 2. Cleavage of dsDNA by a Restriction Enzyme Endonuclease (e.g., BamH I)

Scheme 3. Schematic Representation of the Immobilization of Double-Stranded DNA onto the Electrodeposited Avidin Monolayer

Figure 1. In situ frequency monitoring during the hydrolysis of surface-bound plasmid dsDNA upon the addition of endonuclease Kpn I. A frequency increase (mass decrease) was seen immediately after the introduction of Kpn I. The frequency change was ca. 40 Hz after 30 min. Kpn I cutting site locates 759 position base in the 2961bp pBluescript II SK (-) phagemid dsDNA vector.

quence.17 Figure 1 shows that the frequency of the QCM increased instantly after the injection of Kpn I, and this was accompanied by a frequency increase (or mass decrease on the surface). The (17) Tomassini, J. et al. Nucleic Acid Res. 1978, 5, 4055.

5274

Analytical Chemistry, Vol. 73, No. 21, November 1, 2001

frequency change was ca. 40 Hz after 30 min. The Kpn I cutting site was located at the position base of the 2961bp pBluescript II SK (-) phagemid dsDNA vector. This might indicate that the oriented plasmid dsDNA retained its bioactivity after binding to the avidin, which is one of the desirable characteristics of such a sensing approach based on the surface-functionalized dsDNA. Impedance Spectroscopy of Bimolecular Interactions of Small Molecule-DNA Interface. Electrochemical impedance spectroscopy (EIS) enabled us to observe the magnitude of the interfacial capacity and the pattern of interconnectivity of ideal circuit elements at a high resolution.18 A sinusoidal signal of small amplitude was applied to the electrochemical cell, and the current response was measured under potentiostatic control. The impedance was calculated as the ratio of the system voltage phasor, U (jω), and the current phasor, I (jω), which were generated by the frequency response analyzer during the experiment (eq 1), where j ) x-1, ω is expressed in rad s-1, ω )2πf, and f is the excitation frequency (Hz).

Z(jω) ) U(jω)/I(jω) ) Zre(ω) + jZim(ω)

(1)

Impedance spectroscopy has been used mainly to investigate (18) Macdonald, J. R., Ed. Impedance Spectroscopy: Emphasizing Solid Materials and Systems; John Wiley & Sons: New York, 1987.

Figure 2. Nyquist diagram (Zim vs Zre) in the presence of 0.1 M phosphate buffer (pH 7.0) at (A) bare Au-coated quartz electrode, (B) avidinmodified Au electrode at 0 V and -0.2 V, (C) biotinylated plasmid dsDNA docking of avidin-modified Au electrode at 0 V and -0.2 V. Frequency range, 100 mHz-10 kHz.

nonbiological systems, such as the protective coating covering the metal substrate.19 This technique is becoming popular for novel applications, such as in the modeling of biological systems through their impedance response using networks of electrical components.20 The technique is also being used to monitor the interfacial properties of layered electrodes upon formation of oligonucleotide complexes and precipitation of enzymatic product, including the capacitance and the electron-transfer resistance.21 Previously, we used EIS to study antibody-antigen interactions on conducting polymer-modified electrodes.22 The advantages of EIS lie in its ability to provide a nondestructive signal during biomolecular interaction. EIS Measurements with and without Redox-Active Probes. In these particular experiments, EIS experiments were first carried out in the presence of phosphate buffer solution (i.e., in the (19) (a) Xiao, H.; Mansfeld, F. J. Electrochem. Soc. 1994, 141, 2332-2337. (b) Van Westing, E. P. M.; Ferrari, G. M.; de Wit, J. H. W. Electrochim. Acta 1994, 39, 899-910. (20) Vandernoot, T. J.; Levinkind, M. Electro. Magnetobiol. 1994, 13, 211223. (21) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; Wiley: New York, 1980. (22) (a) Sargent, A.; Loi, T.; Gal, S.; Sadik, O. A. J. Electroanal. Chem. 1999, 470, 144-156. (b) Sargent, A.; Sadik, O. A. Electrochim. Acta 1999, 44, 4667-4675.

absence of any redox probes). We observed that there were no well-defined, electron-transfer trends generated. This was evidenced from the scattered distribution of the frequency response over the range of 100 mHz to 100 kHz. Figure 2 shows the result obtained in the absence of the redox-active probe in which no noticeable trend in impedance was observed and no electron transfer was measured. As a result of this trend, we decided to consider the use of a redox probe to examine the impedance signals. In the presence of the redox-active probe and with the aid of RC circuits, we could depict the electron-transfer measurement system using simple electrical circuitry. Scheme 4 shows the general equivalent circuit of the ac impedance spectroscopy performed on metal electrodes. The circuit was based on the simplest electrode processes shown in the literature21,23 that are capable of explaining more complex situations. The two components of the electronic circuit, that is, Rs and Zw, represent the bulk properties of the electrolyte solution and diffusion features of the redox probe in solution (i.e., Fe(CN)63-/4-). The doublelayer capacitance Cd consists of the capacitance of a bare electrode (CAu) and a variable capacitance originating from the electrode (23) Stoynov, Z. B.; Grafov, B. M.; Savova-Stoynov, B. S.; Elkin, V. V. Electrochemical Impedance; Nauka: Moscow, 1991.

Analytical Chemistry, Vol. 73, No. 21, November 1, 2001

5275

Scheme 4. Equivalent Electrical Circuit for a Simple Charge-Transfer Reaction at a Planar Surfacea

a C , double-layer capacitance; R , charge-transfer resistance; d et Zw, Warburg impedance; and Rs, solution resistance between surface and reference electrode.

surface modifier (Cmod). Therefore, the double-layer capacitance can be expressed as

1/Cd ) 1/CAu + 1/Cmod

(2)

Any electrode modifier of an insulating feature will decrease the double-layer capacitance when compared to the bare metal electrode. Ret controls the electron-transfer kinetics of the redox probe at the electrode interface. Thus, the insulating modifier on the electrode is expected to retard interfacial electron-transfer kinetics, and this will increase the electron-transfer resistance accordingly. Ret is given by eq 3, where RAu and Rmod represent the electron-transfer resistance of the bare gold electrode and the variable electron-transfer resistance introduced by the modifier, respectively.

Ret ) RAu + Rmod

(3)

The typical shape of an ac impedance spectrum (presented in the form of a Nyquist plot) includes a semicircle followed by a straight line. In general, two frequency regions can be distinguished in the presence of electroactive species. The semicircle portion observed at higher frequencies corresponds to the electron-transfer-limited process, and the straight line represents the diffusion-limited electron-transfer process. The intercepts of the semicircle with the Zre axis at high and low frequencies correspond to Rs and Rs + Ret, respectively. Thus, the diameter of the semicircle is equal to Ret. An increase in the diameter of the semicircle can be used to correlate the serial assembly of the insulating layers to the electrodes.10 For example, in Figure 3, for the 1A-modified gold-quartz electrode, Ret ) 907.6 Ω, whereas Ret increased to 1125.8 Ω upon association of the complex between avidin and biotinylated plasmid dsDNA. These results are consistent with the fact that the negative charge associated with the phosphate groups of the plasmid increases after the multistep organization of the assembly. In addition to this simple electrostatic interaction, the observed change in Ret could be attributed to multiple fundamental microscopic processes, such as the flow of charged atoms or atomic agglomerates in the electrolytes, charge-transfer reactions resulting from the diffusion features of the redox probe in solution (i.e., Fe(CN)63-/4-), and specific adsorption. The flow may be further impeded by band structure anomalies at any grain boundaries present especially due to additional phases by the avidin-biotin layers. 5276

Analytical Chemistry, Vol. 73, No. 21, November 1, 2001

Figure 3. Nyquist diagram in the presence of 10 mM Fe(CN)63-/4at (a) a bare gold electrode, (b) 1A-, (c) 1A1B-, (d) 2A1B-, (e) 2A2B-, and (f) 3A2B-modified gold electrode (from left to right).

Scheme 5. Two Equivalent Circuits that Can Have Exactly Identical Frequency Dispersions

Monitoring the Serial Deposition of dsDNA Layer Using EIS. Usually, impedance data can be interpreted in two ways. One way is to reconcile the data within the context of microscopic equations of motion driven by a harmonic voltage excitation. The other is to use an equivalent circuit model based on the appearance of the impedance spectrum and the physical model of the system.24 From the EQUIVCRT.PAS program, the circuit description code (CDC) allows us to use a variety of equivalent circuits to be analyzed or simulated. Although only two types of circuit types can be represented by the CDC (i.e., a series and a parallel), the various circuits that are different in appearance can still be modeled to give identical frequency dispersions (Scheme 5).25 In addition, Scheme 5 circuits can have exactly identical frequency dispersions, thus allowing us to use CDCs such as RQ (RQ) to simulate the impedance data obtained. Figure 4 is the fit-quality plots of the data of one experiment. In the impedance mode, an almost perfect match can be observed between the measured and the simulated data, while a residual error of