Label-Free Femtomolar Detection of Target DNA by Impedimetric DNA

Anal. Chem. 2010, 82, 1066–1072

Label-Free Femtomolar Detection of Target DNA by Impedimetric DNA Sensor Based on Poly(pyrrole-nitrilotriacetic acid) Film Jessica Baur,† Chantal Gondran,† Michael Holzinger,† Eric Defrancq,† Hubert Perrot,‡ and Serge Cosnier*,† De´partement de Chimie Mole´culaire, UMR 5250, ICMG FR 2607, CNRS, Universite´ Joseph Fourier Grenoble I, BP 53, 38041 Grenoble cedex 9, France, and Laboratoire Interfaces et Syste`mes Electrochimiques, UPR 15 CNRS, Universite´ Pierre et Marie Curie Paris VI, cpr 133, 75252 Paris cedex 05, France An ultrahigh performance impedimetric DNA sensor is presented showing detection limits in the femtomolar range. This electrochemical setup was constructed initially by electrogeneration of poly(11-pyrrol-1-yl-undecanoic acid Nr′,Nr-bis(carboxymethyl)-L-lysine amide) (poly(pyrrole-NTA)) film. The latter was then modified by the coordination of Cu2+ ions onto the chelating NTA centers followed by the immobilization of the ssHIVDNA previously modified by a polyhistidine tag by affinity binding. The immobilization of the DNA probe and hybridization with the complementary target ssHIVDNA were investigated using fluorescence microscopy and quantified with quartz crystal microbalance experiments leading to DNA probe and duplex coverage of 1.7 × 10-11 and 7.7 × 10-12 mol cm-2, respectively. The duplex formation was corroborated by amperometric measurements through the duplex labeling by a glucose oxidase. In the presence of hydroquinone as redox indicator, the DNA sensor was applied to the impedimetric detection of target DNA without a labeling step. A linear quantification of the HIV DNA target was carried out in the range 10-15 to 10-8 mol L-1. For four decades, rapid detection and monitoring in medical and food diagnostics, environmental control, and bioterrorism have paved the way for the elaboration of biosensors. The latter constitute an alternative to centralized analytical techniques as valuable, easy, inexpensive, fast, and selective tools. In particular, owing to the key role of DNA in the transfer of genetic information, considerable efforts are devoted to the development of DNA sensors. The determination of specific DNA sequences via a hybridization process was carried out by direct or indirect transduction via optical, electrochemical, or gravimetric transducers.1,2 Some of these detection methodologies require a label (enzymes, fluorophores, magnetic beads, metal complexes, organic redox marker, intercalators, or radioactive compounds) attached to the DNA target or external indicator. However, although the labeling step enhances the sensor sensitivity, * Corresponding author. E-mail: [email protected]. † Universite´ Joseph Fourier Grenoble I. ‡ Universite´ Pierre et Marie Curie Paris VI. (1) Telesa, F. R. R.; Fonseca, L. P. Talanta 2008, 77, 606–623. (2) Cosnier, S.; Mailley, P. Analyst 2008, 133, 984–991.

1066

Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

it markedly increases the time, complexity, and cost of the measurement. Among different conventional methods enabling direct detection of DNA hybridization, electrochemical impedance spectroscopy (EIS) transduction is the goal of numerous research works due to portability and low cost and power requirement. EIS is a powerful technique for the characterization of electrochemical systems3-5 and hence constitutes an efficient and sensitive tool for the characterization of biofunctionalized electrodes.6-8 This technique is especially interesting for label-free detection in bioanalytics since it bypasses the need to modify biomolecules with labels.4-11 Typically, EIS-based DNA sensors using [Fe(CN)6]3-/4- as redox indicator for the impedance signal led to detection limits for DNA hybridization between 10-10 and 5 × 10-15 mol L-1.12,13 Usually, the performance of DNA sensors based on EIS strongly depends on the physical characteristics of the immobilized DNA probe layer. In particular, DNA detection requires the precise location of a single-stranded DNA probe at a transducer surface without loss of biological activity or lack of accessibility and with appropriate orientation propitious to duplex formation with the DNA target as well as a minimum distance between the duplex and the electrode surface. However, the stable and reproducible immobilization of biological macromolecules on a surface with complete retention of their biological activity remains always a crucial problem. In this context, electrogenerated polymers attracted wide attention as interfaces between an electrode and the DNA probe due to their easy control over morphology and thickness and the reproducibility of the polymer formation with precise spatial (3) Barsoukov, E.; Macdonald, J. R. Impedance Spectroscopy: Theory, Experiment, and Applications, 2nd ed.; John Wiley & Sons, Inc: New Jersey, 2005. (4) Katz, E.; Willner, I. Electroanalysis 2003, 15, 913–947. (5) Palchetti, I.; Mascini, M. Anal. Bioanal. Chem. 2008, 391, 455–471. (6) Park, J. W.; Jung, H. S.; Lee, H. Y.; Kawai, T. Biotechnol. Bioprocess Eng. 2005, 10, 505–509. (7) Patolsky, F.; Zayats, M.; Katz, E.; Willner, I. Anal. Chem. 1999, 71, 3171– 3180. (8) Dharuman, V.; Grunwald, T.; Nebling, E.; Albers, J.; Blohm, L.; Hintsche, R. Biosens. Bioelectron. 2005, 21, 645–654. (9) Lisdat, F.; Scha¨fer, D. Anal. Bioanal. Chem. 2008, 391, 1555–1567. (10) Lucarelli, F.; Tombelli, S.; Minunni, M.; Marrazza, G.; Mascini, M. Anal. Chim. Acta 2008, 609, 139–159. (11) Sassolas, A.; Leca-Bouvier, B. D.; Blum, L. J. Chem. Rev. 2008, 108, 109– 139. (12) Zhang, K.; Ma, H.; Zhang, L.; Zhang, Y. Electroanalysis 2008, 20, 2127– 2133. (13) Kjllman, T. H. M.; Peng, H.; Soeller, C.; Travas-Sejdic, J. Anal. Chem. 2008, 80, 9460–9466. 10.1021/ac9024329  2010 American Chemical Society Published on Web 12/31/2009

resolutions.14 Moreover, the introduction of appropriated functionalities by the chemical modification of the monomer or the film can lead to DNA immobilization following various strategies involving affinity interactions,15,16 electrostatic adsorption,17,18 chemical grafting,19,20 direct electropolymerization,21,22 or entrapment process.23,24 In particular, avidin-biotin interactions are widely used for biomolecule anchoring avidin acting as a bridge between the biotinylated surface and a biotinylated biomolecule. Taking into account that DNA can be immobilized by a single attachment point, this strategy preserves the DNA accessibility for subsequent hybridization. This immobilization strategy, however, involves the presence of an intermediate avidin layer as a building block that has a detrimental effect on the sensor sensitivity for hybridization detection. Here, we report, for the first time, an innovative design of an impedimetric DNA sensor based on an electropolymerized poly(pyrrole-nitrilotriacetic acid) film (poly(pyrrole-NTA)). The latter combined with Cu2+ and histidine derivatives mimics the biological avidin-biotin interactions by replacing the bulky avidin by a copper cation. Such a polymeric affinity system was recently used for the immobilization of proteins modified by histidine groups.25,26 The possibility to efficiently modify the extremity of oligonucleotides with a peptidic moiety by using oxime ligation enabled histidine tagging of DNA for its immobilization on transducer surfaces.27 The firm anchoring of histidine tagged ssDNA from the human immunodeficiency virus (HIV) on an electrogenerated poly(pyrrole-NTA)/Cu2+ film and the subsequent hybridization process with the DNA target were thus examined using fluorescence microscopy, quartz crystal microbalance, and amperometry measurements. The performance of the resulting DNA sensor for the label-free detection of the DNA target was investigated by electrochemical impedance spectroscopy using hydroquinone as neutral redox indicator. MATERIALS AND METHODS Chemicals. Avidin (A9275), FITC-labeled avidin (A2901), Na2HPO4 (S0878), NaH2PO4 (S0751), and glucose oxidase (GOX, G2133, 179 U mg-1) were provided from Sigma. AgNO3 (21572), glucose (24379), hydroquinone (24704), and CuCl2 (14) Cosnier, S. Anal. Lett. 2007, 40, 1260–1279. (15) Ramanathan, K.; Bangar, M. A.; Yun, M.; Chen, W.; Myung, N. V.; Mulchandani, A. J. Am. Chem. Soc. 2005, 127, 496–497. (16) Dupont-Filliard, A.; Billon, M.; Livache, T.; Guillerez, S. Anal. Chim. Acta 2004, 515, 271–277. (17) Davis, F.; Nabok, A. V.; Higson, S. P. J. Biosens. Bioelectron. 2005, 20, 1531–1538. (18) Wang, J.; Jiang, M.; Mukherjee, B. Anal. Chem. 1999, 71, 4095–4099. (19) Ionescu, R. E.; Herrmann, S.; Cosnier, S.; Marks, R. S. Electrochem. Commun. 2006, 8, 1741–1748. (20) Piro, B.; Haccoun, J.; Pham, M. C.; Tran, L. D.; Rubin, A.; Perrot, H.; Gabrielli, C. J. Electroanal. Chem. 2005, 577, 155–165. (21) Livache, T.; Fouque, B.; Roget, A.; Marchand, J.; Bidan, G.; Te´oule, R.; Mathis, G. Anal. Biochem. 1998, 255, 188–194. (22) Descamps, E.; Leı¨chle´, T.; Corso, B.; Laurent, S.; Mailley, P.; Nicu, L.; Livache, T.; Bergaud, C. Adv. Mater. 2007, 19, 1816–1821. (23) Wang, J.; Jiang, M. Langmuir 2000, 16, 2269–2274. (24) Prabhakar, N.; Arora, K.; Singh, S. P.; Singh, H.; Malhotra, B. D. Anal. Biochem. 2007, 366, 71–79. (25) Haddour, N.; Cosnier, S.; Gondran, C. J. Am. Chem. Soc. 2005, 127, 5752– 5753. (26) Aravinda, C. L.; Cosnier, S.; Chen, W.; Myung, N. V.; Mulchandani, A. Biosens. Bioelectron. 2009, 24, 1451–1455. (27) Edupuganti, O. P.; Singh, Y.; Defrancq, E.; Dumy, P. Chem.sEur. J. 2004, 10, 5988–5995.

(23093) were purchased from Prolabo (VWR International) while LiClO4 (62579) came from Fluka and EZ-link Sulfo-NHSLC-biotin (21335) from Pierce (3747 N Meridian Road P.O. Box 117, Rockford, IL 61105). Acetonitrile (Rathburn, HPLC grade) and lithium perchlorate were used as received. Ultrapure water (18.2 Ω cm resistivity) was used for all aqueous solutions. All chemicals were of analytical grade and used as received. Solutions of glucose were allowed to mutarotate at room temperature for one day and were kept refrigerated. Pyrrole-nitrilotriacetic acid (pyrrole-NTA) was synthesized in the following manner according to a previously reported protocol.25 11-Pyrrol-1-yl-undecanoic acid succinimid ester (288.5 mg, 0.83 mmol) was dissolved with N,N′ bis(carboxymethyl)-L-lysine (328.3 mg, 1.25 mmol) in DMF (5 mL), TEA (300 µL), NaOH (69.2 mg, 2.15 mmol), and water (2 mL). The solution was stirred for 48 h at 55 °C. After evaporation of the solvent under vacuum atmosphere, the crude product was redissolved in water (2 mL) and the aqueous solution was filtered. Pyrrole-NTA was then obtained as oil (210.5 mg, 51% yield) after water evaporation. 1H NMR (250 MHz/CD3OD): δ (ppm) ) 6.65 (s, 2H), 6.01 (s, 2H), 3.88 (t, 2H), 3.62 (m, 5H), 3.18-3.46 (m, 4H), 2.17 (t, 2H), 1.77-1.30 (m, 24H). ESI/MS: m/z ) 496 (MH+, C25H41O7N3). Glucose oxidase (5 mg, 31 nmol) was biotinylated by reaction with sulfo-NHS-LC-biotin (63 µL 10-2 mol L-1, 630 nmol) in 1 mL phosphate buffer (pH ) 7.4, 0.01 mol L-1) containing NaCl (0.15 mol L-1) and KCl (3 × 10-3 mol L-1) for 2 h in crushed ice. Biotinylated glucose oxidase (BGOX) was purified by centrifugation (30000 MWCO concentrator with PES membrane at 7000 G) at 4 °C. The BGOX concentration was quantified using UV spectroscopy by measuring the 280 nm band intensity in PBS. Stock solutions of BGOX (0.5 mg mL-1) were prepared with phosphate buffer (pH ) 7, 0.1 mol L-1) and stored at -20 °C. All oligonucleotides were prepared on a DNA synthesizer (ABI 3400) at 1 µmol scale. The HIV probe (HIVP) bearing five histidine residues at the 5′-extremity was prepared according to a previously reported method.27 Complementary biotinylated DNA target (BHIVT) was provided by Eurogentec. Stock solutions of all oligonucleotides (0.05 mg mL-1) were prepared with phosphate buffer (0.1 mol L-1, pH ) 7) and kept frozen. The oligonucleotide sequences are the following: HIVP, HIVDNA-probe (HIVP-22 mer), 5′-NH2-(His)5-GAGACCATCAATGAGGAAGCTG-3′; BHIVT, biotinylated HIV-DNA-target (BHIVT33 mer), 5′-biotin-ATCCCATTCTGCAGCTTCCTCATTGATGGTCTC-3′; HIVT, HIV-DNA-target (HIVT-48 mer), 5′-GCTATTTGGCTACCGATCCCATTCTGCAGCTTCCTCATTGATGGTCTC-3′; NCT, HIV-DNA-noncomplementary target (NCT48 mer), 5′-GCTATTTGGCTACCGATCCCATTCTGTTATCCTTCTTCCATCAACTCT-3′. DNA Modified Electrode Preparation. Poly(pyrrole-NTA) films were generated by controlled potential electrolysis (2.5 mC cm-2) of pyrrole-NTA (4 × 10-3 mol L-1) at 0.95 V vs Ag/Ag+ in CH3CN + 0.1 mol L-1 LiClO4. The coordination of Cu2+ ions within the poly(pyrrole-NTA) film was carried out by immersing the modified electrode for 20 min in a stirred acetate buffer solution (0.1 mol L-1, pH ) 4.8) containing CuCl2 (10-2 mol L-1). The resulting electrode was carefully washed for 5 min with a 0.5 mol L-1 NaCl solution and then rinsed with Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

1067

phosphate buffer. HIVP (DNA probe) was denatured at 70 °C for 5 min, and a drop (20 µL) of the resulting solution was spread on a poly(pyrrole-NTA)/Cu2+ film electrode. After an incubation period (30 min), the electrode was rinsed with phosphate buffer and then incubated for 30 min with 20 µL of denatured HIVT (target) solution. Apparatus. Cyclic voltammetry and electropolymerization experiments were performed with a VMP3 multipotentiostat controlled by the EC-Lab software (Biologic, France). Amperometric measurements were carried out with a PRG-DEL potentiostat (Tacussel, France). Electrochemical impedance spectroscopy (EIS) was carried out with an Autolab potentiostat 100 (Eco Chemie, Utrecht, The Netherlands) using FRA software. A ZView software (Scribner Associates Inc.) was used to simulate the data obtained using an appropriate equivalent electrical circuit. An electrochemical three-electrode cell (Metrohm) was employed. Platinum or glassy carbon disk electrodes (φ ) 5 mm) were used as working electrodes and polished with 2 µm diamond paste (MECAPREX Press PM) before use. A Ag/Ag+ (AgNO3 10-2 mol L-1 in CH3CN + 0.1 mol L-1 LiClO4) electrode and a saturated calomel electrode (SCE) were used as reference electrode in acetonitrile electrolyte and aqueous solutions, respectively. A Pt wire placed in a separated compartment containing the supporting electrolyte was used as a counter electrode. All experimental impedance spectra were recorded at a rotation rate of 500 rpm at 25 °C in phosphate buffer solution (0.1 mol L-1, pH ) 7) containing hydroquinone (10-3 mol L-1) as redox probe. The frequency sweep ranges from 500 kHz to 0.1 Hz at 0.4 V vs SCE with a sine wave potential modulation of 5 mV rms. A Pt wire used as counter electrode is directly placed in the cell. Fluorescence microscopic images were recorded with an OLYMPUS BX61 microscope (exciter: D350/50, emitter: D470/ 40, beamsplitter: 400dclp; Japan). Experiments were performed on platinum microelectrode array wafers provided by the Biosensors and Biocatalysis Group of the University of Leeds (U.K.). QCM experiments were carried out with AT-cut 9-MHz quartz crystals (CQE, Troyes, France) coated with two identical Au layers (φ ) 5 mm, thickness ) 200 nm). One gold layer was modified with an electrogenerated poly(pyrrole-NTA) film by oxidative electrolysis (0.1 mC). A Plexiglas cell (Vcell ) 50 µL) connected to a micro pump (P1, Pharmacia) was used with an applied constant flow rate of 50 µL min-1. The inlet and outlet tubes were placed in the same sample for circulating flow. A labmade oscillator was used to drive the crystal at 27 MHz which corresponds to the third overtone of a quartz resonator (9 MHz). Gravimetric measurements were realized in phosphate buffer (0.05 mol L-1, pH ) 7) containing 0.5 mol L-1 of NaCl (PBS). The PBS solutions containing HIVP (0.01 mg mL-1), BHIVT (0.01 mg mL-1), and avidin (0.5 mg mL-1) were successively pumped through the cell with buffer in between. The microbalance frequency shift was monitored by a frequency counter (Fluke PM6685) as a function of time. For amperometric and fluorescent measurements, a biotinylated DNA-target (BHIVT) was used for the hybridization step instead of HIVT. Then, for amperometric experiments, the modified electrodes were successively incubated with 20 µL of 1068

Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

Figure 1. Schematic representation of the differently modified poly(pyrrole-NTA)-Cu2+-DNA films formed on platinum electrodes by electropolymerization and coordination of histidine-tagged DNA (HIVP) on the chelated copper complex (insert): (a) fluorescence detection by hybridization with a biotinylated DNA target (BHIVT) and affinity binding of FITC-labeled avidin; (b) hybridization with a DNA target (HIVT) for direct impedimetric detection; (c) hybridization with a biotinylated DNA target (BHIVT) and affinity binding of avidin (gravimetric detection) and biotinylated glucose oxidase (GOX) (amperometric measurements).

avidin (1 mg mL-1) and 20 µL of BGOX (0.5 mg mL-1) for 2 and 1 h, respectively. For fluorescence experiments, fluorescent FITC-labeled avidin (0.5 mg mL-1) was used instead of regular avidin without additional BGOX labeling process. After each incubation step, the modified electrodes were rinsed and washed with phosphate buffer for 5 min. Polypyrrole-NTA/Cu2+/HIVP modified glassy carbon electrodes were used in EIS measurements and were prepared as described before. A first Nyquist plot was recorded with the probe. Then, 20 µL portions of HIVT with steady increasing concentrations from 10-17 to 10-6 mol L-1 were deposited on a polypyrrole-NTA/Cu2+/HIVP modified glassy carbon electrode, and an impedance spectrum was recorded for each analyte concentration. Before and after HIVT incubation, the samples were washed with phosphate buffer. RESULTS AND DISCUSSION Characterization of the Specific Attachment of HIV Probe onto Poly(pyrrole-NTA)-Cu2+ Films. Electropolymerization of pyrrole-NTA (4 × 10-3 mol L-1) was carried out by controlled potential electrolysis at 0.95 V vs Ag/Ag+ in CH3CN + 0.1 mol L-1 LiClO4, and the amount of electrogenerated polymer, and hence its thickness, was controlled by integration of the oxidative charge (2.5 mC cm-2). The resulting modified electrode was then transferred to a monomer free CH3CN + 0.1 mol L-1 LiClO4 solution. As expected, the cyclic voltammogram exhibits a reversible peak system at E1/2 ) 0.47 V assigned to the well-known electroactivity of the poly(pyrrolic)

skeleton.28 The apparent surface coverage of the electropolymerized poly(pyrrole-NTA) (Γ ) 1.5 × 10-9 mol cm-2; electric yield of the polymerization process 13%) was determined from the charge recorded under the oxidation wave of the poly(pyrrole) backbone. First, we examined the possibility of fixing specifically a DNA labeled by histidine groups onto a polymer through the coordination of histidine onto the polymerized Cu2+ complex. For this purpose, the poly(pyrrole-NTA)-Cu2+ film was electrogenerated onto an array of microelectrodes in order to demonstrate the expected specific anchoring of the DNA probe (HIVP) through its spatially controlled immobilization. The modified microelectrodes were successively incubated with the DNA probe functionalized by histidine groups (HIVP), the cDNA target labeled by a biotin group (BHIVT), and finally a fluorescent avidin (FITC-labeled avidin) (Figure 1a). Figure 2A shows the fluorescence images of the microelectrode array. It appears that the fluorescence phenomenon is precisely localized onto the microelectrode surfaces (Figure 2Ad). This indicates that the fluorescent avidin was bound by affinity interactions to the biotinylated duplex resulting from the hybridization of BHIVT with the DNA probe and hence confirms the specific immobilization of HIVP onto the polymeric film. Control experiments were carried out with the incubation of the fluorescent avidin onto a bare microelectrode array or arrays previously modified by either poly(pyrrole-NTA)-Cu2+ film or poly(pyrrole-NTA)-Cu2+ film incubated with HIVP. The absence of fluorescence for these configurations indicates the absence of nonspecific adsorption of FITC-labeled avidin and hence corroborates the binding of FITC-labeled avidin via the specific attachment of HIVP onto the polymer surface. To quantify the immobilization of histidine-tagged DNA and its ability to form after immobilization a duplex with the DNA target, gravimetric experiments were performed with quartz crystal microbalance modified by the poly(pyrrole-NTA)-Cu2+ film. In aqueous solution, the quartz crystal microbalance (QCM) technique allows the measurement of minute mass changes. The modified quartz crystal was incorporated in the flow through cell, and the microbalance frequency response to buffer solutions containing HIVP, BHIVT, and avidin was monitored in continuous-flow mode as a function of time. In the presence of HIVP, a rapid decrease in frequency with time was observed followed by stabilization at a lower microbalance frequency value (Figure 2B). In addition, the circulation of PBS does not cause any strong frequency increase, indicating a firm fixation of HIVP onto the polymer. Taking into account a theoretical mass/frequency sensitivity of 360 pg Hz1- (A ) 0.2 cm2),29 this frequency shift (71 Hz) corresponds to a mass increase estimated to a density of immobilized DNA probe of 1.7 × 10-11 mol cm-2. During circulation of the cDNA strand (BHIVT), another decrease in frequency (44 Hz) was recorded corresponding to the specific immobilization of BHIVT with coverage of 7.7 × 10-12 mol cm-2. The latter leads to a hybridization ratio of 45% that confirms the presence of immobilized ssDNA onto the polymer surface and their

Figure 2. (A) Fluorescence microscopic images of microelectrode array after incubation with FITC-labeled avidin: (a) without surface modification, (b) after electrogeneration of polypyrrole-NTA/Cu2+ films, (c) same as part b after incubation with HIVP, and (d) complete DNA architecture of polymer incubated successively with HIVP and BHIVT. The scale bar is 20 µm. (B) Time course of frequency response of 27 MHz QCM (overtone 3) modified by a polypyrroleNTA/Cu2+ film: (a) to additions of (1) 10 µg mL-1 HIV-DNA probe (HIVP) solution, (2) 10 µg mL-1 HIV-DNA target (BHIVT) solution, and (3) 500 µg mL-1 avidin solution; (b) to addition of (3) 500 µg mL-1 avidin solution with flow rate 50 µL min-1. (C) (a) Steady-state current-time response of a polypyrrole-NTA/Cu2+ electrode modified by successive deposition of HIVP, BHIVT, avidin, and BGOX, to glucose (3 × 10-4 mol L-1), (b) calibration curve for glucose at the modified sensor. Applied potential: 0.7 V vs SCE in 0.1 mol L-1 stirred phosphate buffer (pH 7) at 25 °C.

(28) Ouerghi, O.; Senillou, A.; Jaffrezic-Renault, N.; Martelet, C.; Ben Ouada, H.; Cosnier, S. J. Electroanal. Chem. 2001, 501, 62–69. (29) Bizet, K.; Gabrielli, C.; Perrot, H.; Therasse, J. Biosens. Bioelectron. 1998, 13, 259–269.

availability for molecular recognition process. In addition, the circulation of avidin induced a frequency decrease (276 Hz) indicating a quantitative anchoring of the protein (Γ ) 7.4 × Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

1069

Figure 3. (A) Nyquist plot of impedance spectra obtained for the DNA sensor at different concentrations of HIVT: (a) 0 mol L-1, (b) 10-17 mol L-1, (c) 10-14 mol L-1, (d) 10-11 mol L-1, (e) 10-8 mol L-1, (f) 3.1 × 10-6 mol L-1. Impedance measurements were performed at 0.4 V vs SCE using hydroquinone 10-3 mol L-1 as redox probe in 0.1 mol L-1 phosphate buffer (pH 7), at a rotation rate of 500 rpm. Experimental (dots) and fitted data (lines) are presented. (B) Equivalent circuit for fitting the plots: Rel, electrolyte resistance; RCT, charge transfer resistance; CDL, double layer capacitance; ZWδ, finite length Warburg impedance.

10-12 mol cm-2) by avidin-biotin affinity interactions onto the immobilized BHIVT, corroborating thus the duplex formation. It should be noted that the maximum coverage corresponding to a close-packed one-molecular layer of avidin was estimated roughly to 6.4-8.3 × 10-12 mol cm-2 30. Consequently, the immobilization of avidin (7.4 × 10-12 mol cm-2) corresponds to the formation of a compact avidin monolayer indicating an efficient duplex coverage of the polypyrrole film. Cross experiments carried out with quartz crystal modified by poly(pyrroleNTA)-Cu2+ films incubated only with HIVP or BHIVT showed no frequency change in the presence of avidin. This clearly demonstrates that the previous change in the frequency response was due to the specific anchoring of avidin and hence corroborates the successive immobilization of all components as shown in Figure 1a. In parallel to gravimetric experiments, the presence of the duplex on the polymer surface was investigated by amperometric measurements. The poly(pyrrole-NTA)-Cu2+ platinum electrodes were successively incubated with HIVP, BHIVT, avidin, and then biotinylated glucose oxidase (BGOX) in order to functionalize each immobilized duplex by an enzyme (Figure 1c). The BGOX immobilized by affinity interactions catalyzes in the presence of dioxygen the oxidation of glucose with the production of H2O2. As a consequence, the amperometric performance of the modified electrodes for the determination of glucose was evaluated in 0.1 mol L-1 phosphate buffer. The electrodes were potentiostated at 0.7 V vs SCE to detect the enzymatically generated H2O2 as a function of glucose concentration. It appears that the response time after a glucose injection was extremely fast, namely 5 s (Figure 2Ca). This reflects the good permeability of the polymer film due to its hydrophilic character. 1070

Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

The amperometric current response of the BGOX-duplex electrode as a function of glucose concentrations led the conventional shape of calibration curve limited by enzyme catalysis (Figure 2Cb). The glucose sensitivity and the maximum current density are 0.5 mA mol-1 L cm-2 and 9.5 µA cm-2, respectively. It appears that the sensitivity (0.5 mA mol-1 L cm-2) was higher than those (0.15 mA mol-1 L cm-2) reported for the immobilization of one layer of biotinylated enzyme onto a poly(pyrrole-biotin) film.31 This reflects the higher hydrophilic character of the poly(pyrrole-NTA)-Cu2+ film that facilitates the H2O2 diffusion. Moreover, the maximum current density value (9.5 µA cm-2) that reflects the amount of immobilized BGOX exhibiting a maximum velocity at saturating glucose conditions is slightly higher than that (6.65 µA cm-2) reported for a BGOX monolayer immobilized on a permeable biotinylated copolymer.32 This at least indicates the formation of a compact layer of enzyme anchored on a close-packed layer of avidin, corroborating thus the gravimetric measurements. In addition, the continuous recording of the steady-state maximum current at saturating glucose conditions for 1 h showed only a weak decrease of 3% (60 nA) demonstrating thus the firm anchoring of each biological layer on the polymer surface and, in particular, the validity of the NTA-Cu2+-histidine system for the stable attachment of the DNA probe. (30) Mousty, C.; Bergamasco, J.-L.; Wessel, R.; Perrot, H.; Cosnier, S. Anal. Chem. 2001, 73, 2890–2897. (31) Cosnier, S.; Stoytcheva, M.; Senillou, A.; Perrot, H.; Furriel, R. P. M.; Leone, F. A. Anal. Chem. 1999, 71, 3692–3697. (32) Cosnier, S.; Galland, B.; Gondran, C.; Pellec, A. L. Electroanalysis 1998, 10, 808–813.

Table 1. Values of Equivalent Circuit Elements Obtained for Fitting of the Experimental Dataa

HIV-DNA sensor after deposition of HIVT 10-8 mol L-1 after deposition of NCT 10-8 mol L-1

RCT/Ω

f °/Hz

CDL/µF

R (deg)

2370 ± 80 3510 ± 110

21.0 19.7

3.2 ± 0.1 2.3 ± 0.1

77.2 78.7

2510 ± 80

21.4

3.0 ± 0.1

77.7

a RCT, charge transfer resistance; f °, relaxation frequency; CDL, double layer capacitance; R, angle.

Impedimetric Characterization of the DNA Sensor and Application to HIV Target Detection. Electrochemical impedance spectroscopy was chosen as the method for detection of HIVDNA target (Figure 1b). In contrast to the preceding amperometric configuration, the main advantage of this method lies in the direct detection of the hybridization event without the labeling step of the DNA target. Contrary to the conventional impedimetric immunosensors which generally used ferrocyanide/ferricyanide mixture as redox probe, the impedimetric HIVT detection was carried out at the poly(pyrrole-NTA)-Cu2+-HIVP electrode in presence of hydroquinone (10-3 mol L-1). This neutral redox probe can diffuse through the negatively charged poly(pyrroleNTA) whereas the commonly used ferrocyanide and ferricyanide are electrostatically rejected. The Nyquist plots of impedance spectra describe the sensor response for various concentrations of the DNA target (HIVT) at a hybridization temperature of 25 °C (Figure 3A). A general increase in impedance was observed reflecting the hybridization process. In a first approach, a basic equivalent circuit was designed for fitting the experimental impedance values (Figure 3B). The circuit includes the ohmic resistance, Rel, of the electrolyte solution, the electronic charge transfer resistance, RCT, in series with the finite length Warburg, Wδ, and in parallel with the double layer capacitance, CDL3. The high frequency semicircle of the Nyquist diagram corresponds to the charge transfer resistance in parallel of double layer capacitance. The low frequency loop is attributed to the finite length Warburg due to the diffusion of hydroquinone. The effect of the rotation rate of the DNA modified electrode on the impedance behavior was investigated. It appears that the variation of the rotation rate has no influence on the first half-circle whereas the second loop varies with the square root of the rotation rate corroborating thus its relation to the hydroquinone diffusion. The lines in Figure 3A correspond to the fit using the equivalent circuit, which demonstrates the perfect match between the experimental data (dots) and the electric model (lines). The value of Rel (400 Ω) is constant for all electrochemical impedance measurements, since the resistance of the solution is determined by the concentration of the electrolyte. The primary end point of the Warburg impedance (low frequency) is the resistance of diffusion, Rδ. Surprisingly, the value of Rδ (6570 Ω) remains constant whatever the concentration of the immobilized target (HIVT). Therefore, the diffusion layer seems to be located completely outside the film electrode. From the evolution of the impedance in Figure 3A, it appears that the semicircle increases with the increase in HIVT concentrations only at high frequencies. Table 1 summarizes the semicircle parameters recorded before and after

Figure 4. Calibration curve for the DNA sensor corresponding to the changes in the electron transfer resistances of the electrode upon detection of different HIVT concentrations: (a) ∆RCT, change in charge transfer resistance; (b) ∆Z3Hz, change of impedance modulus at 3 Hz. Experimental conditions as in Figure 3.

incubation of the DNA sensor in a HIVT solution (10-8 mol L-1) as well as those obtained for control experiments using a noncDNA target (NCT). The charge transfer resistance, RCT, increases by 1140 Ω after hybridization of the DNA receptor with the complementary target, while only a 140 Ω increase was observed for the same concentration of the noncDNA. This result reflects the high specific recognition of our setup. The relaxation frequency, f °, which characterizes the speed of the electron transfer, remains almost constant while the double layer capacitance, CDL, decreases after formation of the DNA duplex. These results demonstrate a purely geometric effect on the interface during the formation of the duplex. The electron transfer is not slowed but merely occurs on a smaller surface area since the dsDNA clutters the surface more strongly than the ssDNA. The value of the depletion angle, R, is low and almost constant indicating that the hybridization with the DNA target does not affect the surface roughness.3 Figure 4a shows the variation of the charge transfer resistance, ∆RCT, as a function of HIVT concentration. An extremely sensitive detection limit for HIVT, namely 10-15 mol L-1, was determined with the impedimetric DNA sensor. In addition, the calibration curve exhibits a wide linear variation range from 10-15 to 10-8 mol L-1 (slope ) 89.8 Ω/log unit, R2 ) 0.995). Such excellent performances may be ascribed to the permeability of the poly(pyrrole-NTA)-Cu2+film toward the diffusion of hydroquinone and the direct anchoring of a high density of DNA target onto the polymer surface. In addition, the negatively charged film should prevent the flat adsorption of HIVP facilitating thus the hybridization process with HIVT. In order to simplify the detection procedure and shorten the response time, this DNA sensor configuration was employed in combination with a conductometric detection method. The change in impedance at 3 Hz, ∆Z3Hz, was measured as a Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

1071

function of the HIVT concentration (Figure 4b). The variation of impedance follows perfectly the change of charge transfer resistance (slope ) 82.8 Ω/log unit, R2 ) 0.995). This is not surprising since the frequency of relaxation is not affected by the hybridization. The distribution frequency of the semicircle does not change either. The measured frequency is always on the same radius of the semicircle. Therefore, this architecture can also be used in conductometric sensors due to its high sensitivity and reliability. CONCLUSION An innovative DNA sensor setup for the reagentless and soft immobilization of ssDNA functionalized by histidine groups onto poly(pyrrole-nitrilotriacetic acid) films was for the first time described. The main advantages of such a sensor configuration lies in the permeable and hydrophilic character of the organic polymer film and the close proximity at the molecular level between the formed duplex and the polymer surface. The impedance transduction based on an unusual redox probe (hyd-

1072

Analytical Chemistry, Vol. 82, No. 3, February 1, 2010

roquinone) thus leads to an extremely sensitive detection limit for the hybridization event without a labeling step for the DNA target. It is thus expected that such a new procedure of ssDNA immobilization combined with EIS transduction will be useful for the development of DNA sensors. ACKNOWLEDGMENT J.B. thanks Emeline Cuzin for her valuable assistance in the impedimetric experiments. J.B. is also grateful to Marie Bernadette Villiers and Patrice Marche at “Institut Albert Bonniot” (Grenoble) for their help in enzyme modification. Arielle Lepellec should especially be acknowledged for the technical assistance. The authors would like to give thanks to the platforms “Synthesis” and “Functionalization of Surfaces and Transduction” of the scientific structure “Nanobio” for providing facilities. Received for review December 16, 2009. AC9024329

October

26,

2009.

Accepted