Attomole DNA Electrochemical Sensor for the Detection of Escherichia

Mar 3, 2009 - Víctor Ruiz-Valdepeñas Montiel , Eloy Povedano , Eva Vargas , Rebeca ... Phil De Luna , Sahar S. Mahshid , Jagotamoy Das , Binquan Lua...
58 downloads 0 Views 1MB Size
Download PDF
Anal. Chem. 2009, 81, 2470–2476

Attomole DNA Electrochemical Sensor for the Detection of Escherichia coli O157 Wei-Ching Liao and Ja-an Annie Ho* BioAnalytical Laboratory, Department of Chemistry, National Tsing Hua University, Hsinchu 30013 Taiwan Enterohemorrhagic Escherichia coli O157, a verocytotoxin (VT1/2)-producing pathogen, can be deadly because it can induce acute or chronic renal failure. To speed up the clinical diagnosis of related syndromes caused by E. coli O157, there is an urgent need for rapid, simple, and reliable analytical tools for its quantitation. In this study, we developed a novel electrochemical competitive genosensor, featuring gold-electrodeposited screen-printed electrodes (nanoAu/SPE) modified with a self-assembled monolayer of thiol-capped single-stranded DNA (capture probe), for the detection of the rfbE gene, which is specific to E. coli O157. This assay functions based on competition between the target gene (complementary to the capture probe DNA) and reporter DNA-tagged, hexaammineruthenium(III) chloride-encapsulated liposomes. The current signal of the released liposomal Ru(NH3)63+ was measured using square wave voltammetry, yielding a sigmoidally shaped dose-response curve whose linear portion was over the range from 1 to 106 fmol. This liposomal competitive assay provides an amplification route for the detection of the rfbE gene at ultratrace levels; indeed, we could detect as little as 0.75 amol of the target rfbE DNA (equivalent to the amount present in 5 µL of a 0.15 pM solution). The rapid, specific detection of nucleic acid sequences has important applications in a wide range of fields, from clinical medicine to experimental biology and research diagnostics.1,2 A variety of optical,3-5 acoustic,6,7 and electronic8 “gene detection” approaches have been reported in the past few years. Because of its high sensitivity, low cost, rapid response, small dimensions (portability), low manpower requirements, and compatibility with microfabrication technology, it is likely that biosensors based on electrochemical transduction will ultimately be preferable for DNA * Corresponding author. Fax: +886-3-571-1082. E-mail: [email protected]. (1) Wang, J. Chem.sEur. J. 1999, 5, 1681–1685. (2) (a) Patolsky, F.; Lichtenstein, A.; Willner, I. J. Am. Chem. Soc. 2001, 123, 5194–5205. (b) Patolsky, F.; Lichtenstein, A.; Willner, I. Nat. Biotechnol. 2001, 19, 253–257. (c) Patolsky, F.; Katz, E.; Willner, I. Angew. Chem., Int. Ed. 2002, 41, 3398–3402. (d) Patolsky, F.; Lichtenstein, A.; Willner, I. Chem.sEur. J. 2003, 9, 1137–1145. (3) Xu, X. H.; Bard, A. J. J. Am. Chem. Soc. 1995, 117, 2627–2631. (4) Cao, Y. W. C.; Jin, R. C.; Mirkin, C. A. Science 2002, 297, 1536–1540. (5) Peterson, A. W.; Wolf, L. K.; Georgiadis, R. M. J. Am. Chem. Soc. 2002, 124, 14601–14607. (6) Cooper, M. A.; Dultsev, F. N.; Minson, T.; Ostanin, V. P.; Abell, C.; Klenerman, D. Nat. Biotechnol. 2001, 19, 833–837. (7) Hook, F.; Ray, A.; Norden, B.; Kasemo, B. Langmuir 2001, 17, 8305–8312. (8) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol. 2003, 21, 1192–1199.

2470

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

detection. Many electrochemical approaches for the detection of nucleic acid have been demonstrated, including the direct reduction or oxidation of DNA,9-12 the mediated catalytic oxidation of guanine in DNA strands by polypyridyl complexes of Ru(II) and Os(II),13,14 the voltammetric monitoring of redox-active indicators15-18 that interact preferentially with double-stranded DNA, and the labeling of DNA with redox-active reporter molecules,19,20 enzymes,21,22 or metal nanoparticles23-25 to amplify the electrochemical hybridization signal. In general, DNA-specific redox indicator detection strategies involve either a three-component “sandwich” assay,26-28 where the redox label is attached to a synthetic sequence specifically designed to bind an overhang portion of the probe-target complex or a competitive assay,29,30 in which competition occurs between the target DNA (the sample) and the labeled DNA for a limited number of capture probe DNA strands. Bringing the probe and labeled sequences together through hybridization enables indirect determination of target DNA and eliminates the need to modify the target strand. Enterohemorrhagic Escherichia coli (E. coli) O157 serotypes were first isolated in 1986 from the feces of a Chinese patient (9) Palecek, E. Anal. Biochem. 1988, 170, 421–431. (10) Singhal, P.; Kuhr, W. G. Anal. Chem. 1997, 69, 4828–4832. (11) Jelen, F.; Yosypchuk, B.; Kourilova, A.; Novotny, L.; Palecek, E. Anal. Chem. 2002, 74, 4788–4793. (12) Wang, J.; Kawde, A. B. Analyst 2002, 127, 383–386. (13) Yang, I. V.; Thorp, H. H. Anal. Chem. 2001, 73, 5316–5322. (14) Yang, I. V.; Ropp, P. A.; Thorp, H. H. Anal. Chem. 2002, 74, 347–354. (15) Hashimoto, K.; Ito, K.; Ishimori, Y. Anal. Chem. 1994, 66, 3830–3833. (16) Kelley, S. O.; Barton, J. K.; Jackson, N. M.; Hill, M. G. Bioconjugate Chem. 1997, 8, 31–37. (17) Maruyama, K.; Mishima, Y.; Minagawa, K.; Motonaka, J. Anal. Chem. 2002, 74, 3698–3703. (18) del Pozo, M. V.; Alonso, C.; Pariente, F.; Lorenzo, E. Anal. Chem. 2005, 77, 2550–2557. (19) Palecek, E.; Fojta, M.; Jelen, F. Bioelectrochemistry 2002, 56, 85–90. (20) Xiao, Y.; Lubin, A. A.; Baker, B. R.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 16677–16680. (21) Dominguez, E.; Rincon, O.; Narvaez, A. Anal. Chem. 2004, 76, 3132–3138. (22) Kavanagh, P.; Leech, D. Anal. Chem. 2006, 78, 2710–2716. (23) Wang, J.; Xu, D.; Kawde, A. N.; Polsky, R. Anal. Chem. 2001, 73, 5576– 5581. (24) Hansen, J. A.; Mukhopadhyay, R.; Hansen, J. O.; Gothelf, K. V. J. Am. Chem. Soc. 2006, 128, 3860–3861. (25) Liu, G.; Lin, Y. J. Am. Chem. Soc. 2007, 129, 10394–10401. (26) Umek, R. M.; Lin, S. W.; Vielmetter, J.; Terbrueggen, R. H.; Irvine, B.; Yu, C. J.; Kayyem, J. F.; Yowanto, H.; Blackburn, G. F.; Farkas, D. H.; Chen, Y. P. J. Mol. Diagn. 2001, 3, 74–84. (27) Gibbs, J. M.; Park, S. J.; Anderson, D. R.; Watson, K. J.; Mirkin, C. A.; Nguyen, S. T. J. Am. Chem. Soc. 2005, 127, 1170–1178. (28) Fojta, M.; Kostecka, P.; Trefulka, M.; Havran, L.; Palecek, E. Anal. Chem. 2007, 79, 1022–1029. (29) Pa¨ke, O.; Kirbs, A.; Lisdat, F. Biosens. Bioelectron. 2007, 22, 2656–2662. (30) Bonanni, A.; Esplandiu, M. J.; del Valle, M. Electrochim. Acta 2008, 53, 4022–4029. 10.1021/ac8020517 CCC: $40.75  2009 American Chemical Society Published on Web 03/03/2009

suffering from hemorrhagic colitis.31 It is one of the most dangerous of the pathogens that cause hemorrhagic colitis and severe hemolytic uremic syndrome, which may result in death via acute or chronic renal failure.32 E. coli O157 is commonly found in ground beef, unpasteurized or raw milk, cold sandwiches, vegetables, apple cider, and drinking water; it can be transmitted through contaminated foods and drinks or spread by person-toperson contact.33 Outbreaks associated with the contamination of E. coli O157 have occurred in many countries, including the United States, Japan, England, Canada, and Sweden.34-36 The U.S. Centers for Disease Control have estimated that E. coli O157 bacteria cause thousands of cases of serious illness in the United States each year. Given the magnitude and severity of outbreaks caused by E. coli O 157 infection, there is an urgent need to develop sensitive, specific, rapid, and powerful new tools to combat food-borne disease and bioterror threats and also to speed up the clinical diagnosis, surveillance, and monitoring of the presence of such pathogens in foodstuffs. The rfbE gene, which encodes an enzyme necessary for O-antigen biosynthesis, is highly conserved in E. coli O157 serotypes.37 Thus, in this study we chose a synthetic DNA strand homologous to part of the target gene rfbE.36 This gene has been used previously in conjunction with polymerase chain reactionbased techniques to diagnose the presence of E. coli O157 specifically.36,38-44 In this present study, we developed a nucleic acid-based electrochemical detection method using a competition and liposomal amplification strategy. Liposomes, spherical vesicles composed of a phospholipid bilayer surrounding an aqueous cavity, were originally developed as models to study the properties of cell membranes.45 Because of their ability to accommodate various water-soluble agents in their aqueous cavities, liposomes are used widely in diagnostic and drug delivery applications. The liposomes used in detection assay systems are mainly immunoliposomes presenting antibodies on their surfaces or are tagged with DNA or analytes of interest.46,47 In this study we prepared (31) Xu, J. G.; Liu, Q. Y.; Jing, H. Q.; Pang, B.; Yang, J. C.; Zhao, G. F.; Li, H. W. Microbiol. Immunol. 2003, 47, 45–49. (32) Karmali, M. A. Clin. Microbiol. Rev. 1989, 2, 15–38. (33) Griffin, P. M.; Tauxe, R. V. Epidemiol. Rev. 1991, 13, 60–98. (34) Willshaw, G. A.; Smith, H. R.; Cheasty, T.; O’Brien, S. J. Int. J. Food Microbiol. 2001, 66, 39–46. (35) Welinder-Olsson, C.; Stenqvist, K.; Badenfors, M.; Brandberg, Å.; Floren, K.; Holm, M.; Holmberg, L.; Kjellin, E.; Marild, S.; Studahl, A.; Kaljser, B. Epidemiol. Infect. 2004, 132, 43–49. (36) Yu, G.; Niu, J.; Shen, M.; Shao, H.; Chen, L. Clin. Chim. Acta 2006, 366, 281–286. (37) Maurer, J. J.; Schmidt, D.; Petrosko, P.; Sanchez, S.; Bolton, L.; Lee, M. D. Appl. Environ. Microb. 1999, 65, 2954–2960. (38) Paton, A. W.; Paton, J. C. J. Clin. Microbiol. 1998, 36, 598–602. (39) Fortin, N. Y.; Mulchandani, A.; Chen, W. Anal. Biochem. 2001, 289, 281– 288. (40) Chizhikov, V.; Rasooly, A.; Chumakov, K.; Levy, D. D. Appl. Environ. Microbiol. 2001, 67, 3258–3263. (41) Morin, N. J.; Gong, Z.; Li, X.-F. Clin. Chem. 2004, 50, 2037–2044. (42) Liu, Y.; Gong, Z.; Morin, N.; Pui, O.; Cheung, M.; Zhang, H.; Li, X.-F. Anal. Chim. Acta 2006, 578, 75–81. (43) Bertrand, R.; Roig, B. Water Res. 2007, 41, 1280–1286. (44) Liu, Y.; Gilchrist, A.; Zhang, J.; Li, X.-F. Appl. Environ. Microbiol. 2008, 74, 1502–1507. (45) Papahadjopoulos, D.; Portis, A.; Pangborn, W. Ann. N.Y. Acad. Sci. 1978, 308, 50–66. (46) Esch, M. B.; Baeumner, A. J.; Durst, R. A. Anal. Chem. 2001, 73, 3162– 3167. (47) Ahn-Yoon, S.; DeCory, T. R.; Baeumner, A. J.; Durst, R. A. Anal. Chem. 2003, 75, 2256–2261.

Table 1. Sequences of ss-DNA Targets Used in This Study name

sequence (5′ f 3′)

Probe capture reporter

5′-HS-(CH2)6-ATGTACAGCTAATCCTTGGCC-3′ 5′-HS-(CH2)6-GGCCAAGGATTAGCTGTACAT-3′

Hybridization Target target ssrAa

5′-GGCCAAGGATTAGCTGTACAT-3′ 5′-TCGAACTATCCCTGTCGAAT-3′

a

Sequence designated for the ssrA gene of L. monocytogenes.

reporter gene-tagged liposomes encapsulating an electroactive redox marker, [Ru(NH3)6]Cl3, and investigated their application to the development of a DNA sensor for the detection of the rfbE gene. MATERIALS AND METHODS Reagents and Materials. All chemicals and organic solvents were of reagent grade or better. Potassium chloride, sodium chloride, sodium citrate, formamide, Ficoll type 400, ethylenediaminetetraacetic acid tetrasodium salt hydrate (EDTA), and hexaammineruthenium(III) chloride (RuHex) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Hydrogen tetrachloroaurate(III) was purchased from Alfa Aesar (Ward Hill, MA). Tris (base), sucrose, and potassium phosphate monobasic (KH2PO4) were obtained from J. T. Baker (Phillipsburg, NJ). Potassium phosphate dibasic (K2HPO4) was purchased from Riedel-de Hae¨n (Seelze, Germany). Triton X-100 was obtained from Amershan Pharmacia Biotech (Amershan Place, U.K.). Dipalmitoylphosphatidylcholine (DPPC), dipalmitoylphosphatidylglycerol (DPPG), 1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine-N-[4-(p-maleimidomethyl)cyclohexanecarboxamide] (PE-MCC) were obtained from Avanti Polar Lipids (Alabaster, AL). The thiol-capped single-stranded capture and reporter oligonucleotides were synthesized by MDBio, Inc. (Taipei, Taiwan); sulfhydryl groups were introduced at the 5′ end via a C6-spacer, forming the sequences 5′-HS-C6-ATGTACAGCTAATCCTTGGCC-3′ and 5′-HS-C6-GGCCAAGGATTAGCTGTACAT-3′, respectively. All other single-stranded DNA (ss-DNA) targets were purchased from MDBio, Inc. (Taipei, Taiwan); their sequences are listed in Table 1. All solutions were prepared with deionized water having a resistivity not less than 18 MΩ cm (Milli-Q, Bedford, MA). Apparatus. Cyclic voltammetry (CV), square wave voltammetry (SWV), and chronocoulometry (CC) measurements were performed using a CHI 633 electrochemical analyzer/workstation (CH Instruments, Inc., Austin, TX). Disposable electrochemical screen-printed electrodes (SPE), comprised of a carbon working electrode, carbon counter electrode, and silver pseudoreference electrode, were purchased from Zensor R&D (Taichung, Taiwan). The effective diameter and zeta potential of the liposomes were measured using a Brookhaven 90Plus nanoparticle size analyzer and zeta potential analyzer (Brookhaven Instruments Co., Holtsville, NY). X-ray photoelectron spectroscopy (XPS) analyses were conducted using an ULVAC-PHI Quantera SXM XPS spectrometer (Chigasaki, Japan). Surface plasmon resonance (SPR) experiments were performed on a Biacore T100 system (Uppsala, Sweden). SPR Assay. The Au surface of a Biacore sensor chip (SIA Kit Au) was rinsed with 1 M NaCl containing 50 mM NaOH. The Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2471

Scheme 1. Flow Diagram Displaying the Concept Behind the Competitive Assay-Based Performance of the Developed Genosensor

SPR analysis was initiated by injecting the running buffer (10 mM Tris-HCl, 1 mM EDTA, and 1 M NaCl; pH 7.4) through the system at a rate of 30 µL min-1 until the baseline became stable. One channel was unmodified to provide an additional reference surface; the other channel on the chip was modified with capture probe DNA (30 µL of 1 µM DNA) in 1 M potassium phosphate buffer (0.5 M KH2PO4 and 0.5 M K2HPO4; pH 7) at a rate of 30 µL min-1. Subsequently, 5 mM Tris-HCl buffer containing10 mM NaCl (pH 7.4) was introduced to wash out the free, unbound DNA probe. The kinetic binding study was then performed by analyzing various concentrations of target DNA (0, 5.63, 31.3, 62.5, 125, 250, and 500 nM), which we obtained through a serial 2-fold dilution of a stock 500 nM solution. NaCl (1 M) containing 50 mM NaOH was used as the regeneration buffer to dissociate the capturing DNA probes from their complementary target DNA. Finally, the SPR data were evaluated using Biacore T100 evaluation software (v. 1.1.1) to calculate the values of the kinetic parameters ka and kd. Fabrication of Electrode. Prior to immobilizing the capture probe ss-DNA, the working electrode of the SPE was preconditioned electrochemically by cycling the potential repeatedly between -0.6 and +0.6 at 0.5 V s-1 in 20 mM Tris-HCl buffer (pH 7.4). A Au nanostructured platform was formed on the working electrode of the SPE in a single step through controlled electrodeposition as described in Ho et al.48 As shown in Scheme 1, the SPE was placed in a solution of 10 mM HAuCl4 containing 0.1 M KCl, and then controlled electrodeposition was performed at -0.66 V for 10 s. The SPE was dried in air after rinsing with distilled deionized (d.d.) water. Subsequently, a droplet (6 µL) of a 1 µM solution of the thiolated capture DNA in 1 M potassium phosphate buffer (pH 7.0) was placed onto the working electrode and left to react overnight at ambient temperature. Finally, the thiol-capped ssDNA self-assembled SPE (ss-DNA/SPE) was rinsed in 5 mM Tris-HCl buffer (pH 7.4) containing 10 mM NaCl. (48) Ho, J. A. A.; Chiu, J. K.; Hsu, W. L.; Hong, J. C.; Lin, C. C.; Hwang, K. C.; Hwu, J. R. R. J. Nanosci. Nanotechnol. 2008, 8, 1–6.

2472

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

Preparation of Reporter Gene-Tagged Liposomal Biolabels. Reporter DNA-tagged liposomes were prepared from a lipid mixture using the reversed-phase evaporation method, which has been described in detail previously.2,49 Briefly, a lipid mixture consisted of DPPC, cholesterol, DPPG, and PE-MCC (10:10:1: 0.25 molar ratio) was dissolved in a mixture of chloroform, isopropyl ether, and methanol (6:6:1 volume ratio, 4 mL) and then 150 mM aqueous Ru(NH3)6Cl3 (1 mL) was added. After sonication of the mixture for 3 min, the organic solvent was evaporated under reduced pressure, leaving a milky-white jelly of liposomes. Another aliquot of Ru(NH3)6Cl3 was added to this residue and then the mixture was sonicated for 3 min and vortexed at 45 °C. The liposome sizes were regulated via extrusion 20 times through 1- and 0.4-µm polycarbonate filters; they were then purified through gel filtration to remove free, unencapsulated Ru(NH3)6Cl3. The collected liposomes fractions were incubated overnight (at 4 °C on a shaker) with an appropriate amount of the 5′-thiol capped reporter probe. Finally, the reaction mixture was passed through a Sephadex G-75 column to separate the DNA-tagged liposomes from the free reporter probe. This liposome suspension was stored at 4 °C until required for use. Characterization of Reporter Gene-Tagged Liposomal Biolabels. The liposomes were characterized using a particle analyzer and a zeta potential analyzer to determine their sizes and electrokinetic potentials. In addition, they were assayed for their phospholipid contents using Bartlett assays,50 which were performed as follows. Samples of liposomes (10 or 20 µL) were dehydrated at 155 °C for 10 min, and then d.d. water (1 mL) was added. Each sample was digested to inorganic phosphates with 10 N H2SO4 (0.5 mL) for 3 h at 155 °C. H2O2 (30%, 100 µL) was added to each sample, and then the mixtures were returned to the oven for 1.5 h. The tubes were cooled to ambient temperature prior to, and vortexed vigorously following, each addition. Finally, 0.22% ammonium molybdate (4.6 mL) and the (49) Rule, G. S.; Montagna, R. A.; Durst, R. A. Clin. Chem. 1996, 42, 1206– 1209. (50) Bartlett, G. R. J. Biol. Chem. 1959, 234, 466–468.

Fiske-Subbarow reagent (prepared by mixing sodium bisulfite [15% (w/v), 40 mL], sodium sulfite (0.2 g), and 1-amino-4naptholsulfonic acid (0.1 g) at ambient temperature for 1 h and then filtering out the undissolved solids; 0.2 mL) were added. The tubes were heated in a boiling-water bath for 7 min and then quickly cooled in an ice-water bath. The absorbance at 830 nm was recorded. Standards prepared from DPPE stock [2.23 mg mL-1 in CHCl3/MeOH (8:2)] were subjected to the same procedure concurrently. The phospholipid content of the liposomes was determined from a calibration curve prepared from the standards analyzed in each run. The total lipid concentration was calculated by multiplying the phospholipid concentration by the initial ratio of total lipid to phospholipid. Assay Performance. The assay procedure (shown in Scheme 1) was initiated by activating the capture ss-DNA self-assembled SPE surface by pipetting the hybridization buffer51 (60% formamide, 6× SSC (saline-sodium citrate buffer), 0.15 M sucrose, 0.8% Ficoll type 400, and 0.01% Triton X-100) onto the ss-DNA/SPE and incubating for 20 min. The subsequent hybridization was performed by directly applying the target sequence and liposome mixture at an appropriate dilution onto the working electrode and incubating the system for 40 min at room temperature with continuous shaking. A fixed volume (1 µL) of the target sequence solution was introduced onto the sensor in each assay; the ratio of the hybridization buffer and the liposome solution was adjusted accordingly to obtain a total volume of 5 µL. Next, the electrode was rinsed with a solution containing 10% formamide, 3× SSC, 0.2 M sucrose, 0.2% Ficoll type 400, and 0.01% Triton X-100 to remove the nonbound target gene and reporter DNA-tagged liposomes. The SPEs were then dried for 20 min under vacuum at ambient temperature prior to electrochemical analysis. Electrochemical measurements were performed in 20 mM Tris-HCl (pH 7.4). The reduction signal of Ru(NH3)63+ (RuHex) was measured using SWV by scanning from 0 to -0.6 V with an amplitude of 25 mV and a step potential of 4 mV at 15 Hz. RESULTS AND DISCUSSION Analysis of DNA Kinetic Binding Using SPR Spectroscopy. To confirm the binding affinity between the synthetic capture probe ss-DNA and its complementary target ss-DNA, we used the kinetic function of an SPR spectrometer to determine the binding parameters. Thus, we injected solutions of the target ss-DNA from low concentration to high concentration to interact with the immobilized capture ss-DNA on the sensor chip. We fitted the resulting data (Figure 1) to a simple bimolecular model, obtaining values of ka and kd of 5.76 (±0.09) × 104 M-1 s-1 (RSD ) 0.02) and 6.75 (±0.30) × 10-5 s-1 (RSD ) 0.04), respectively. The equilibrium dissociation constant KD, which we calculated based on the ratio kd/ka, was 1.17 (±0.07) × 10-9 M (RSD ) 0.06). Characterization of Liposomal Biolabels. In this study, we used DNA-tagged liposomal biolabels as signal amplifiers. A homogeneous liposome population (with respect to their size) is essential for any diagnostic application requiring acceptable reproducibility from an assay. Thus, we extruded the liposome preparations to increase their size homogeneity. The average (51) Kwakye, S.; Goral, V. N.; Baeumner, A. J. Biosens. Bioelectron. 2006, 21, 2217–2223.

Figure 1. Kinetic analysis (Biacore T100 system) of the interaction of the target ss-DNA with the immobilized capture ss-DNA. From bottom to top, the injected concentrations of target ss-DNA were 0, 15.63, 31.25, 62.5, 125, 250, and 500 nM.

diameter and zeta potential of the resulting liposomes were 212.4 nm and -18.26 mV, respectively, suggesting that the average volume of a single liposome was ∼6.3 × 10-19 L, with an entrapped volume (assuming a bilayer thickness of 4 nm) of ∼5.0 × 10-19 L. Assuming that the RuHex concentration inside the liposomes was equal to that (150 mM) in the original solution and comparing the current signal of the lysed liposomes with that of standard RuHex solution, we calculated that there were ∼7.5 × 1016 liposomes L-1 and that each liposome contained ∼4.5 × 104 molecules of RuHex. The phospholipid concentration determined using Bartlett’s phosphorus assay was ∼0.25 g L-1. On the basis of the average size and phospholipid concentration in the liposome preparation, we calculated the liposome concentration to be ∼1.5 × 1016 liposomes L-1, which was accurate to within the same order of magnitude as the concentration calculated according to the RuHex information. Characterization of Bioelectrodes. Figure 2A displays the CV behavior of Fe(CN)63-/4- on the bare gold-nanostructured SPE (nanoAu/SPE) and the capture probe DNA-modified electrode (DNA/SPE) in 100 mM PBS (containing 0.15 M NaCl, pH 7.0). We observed a pair of well-defined peaks corresponding to the reduction and oxidation of Fe(CN)63-/4on the bare nanoAu/SPE surface; the redox reaction of Fe(CN)63-/4- was markedly less reversible on the DNA/SPE surface, as evidenced by increased peak splitting (∆Ep became larger). This phenomenon was due to repulsive electrostatic interactions between the negative DNA layer and the anionic Fe(CN)63-/4-, which impeded the anions from reaching the electrode surface.52 These observations suggested that the capture probes were successfully immobilized on the nanoAu/ SPE. Furthermore, we used X-ray photoelectron spectroscopy (XPS) to verify the presence of the capture probe DNA on the surface of the nanoAu/SPE. The intensity of the S 2p peak from the capture probe-modified electrode was higher than that from the bare nanoAu/SPE, once again confirming that the capture probes were immobilized on the electrode. (52) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670–4677.

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2473

Figure 2. (A) Cyclic voltammetric traces of Fe(CN)63-/4- on the bare Au-nanostructured SPE (solid curve) and the capture DNA-modified electrode (dashed curve). (B) CC response curves for the capture DNA-modified electrode in the absence (solid curve) and presence (dashed curve) of 50 µM Ru(NH3)63+.

We used the chronocoulometric (CC) analysis procedure described by Steel et al.,52 employing 50 µM Ru(NH3)63+ containing 10 mM Tris-HCl (pH 7.4) at a pulse period of 250 ms and a pulse width of 300 mV, to determine the density of the selfassembled capture DNA on the electrode surface. We measured the redox charges (Figure 2B) of Ru(NH3)63+, which was associated electrostatically with the anionic DNA backbone,52 to estimate the surface density of the capture probe DNA to be 1.14 (±0.07) × 1013 molecules cm-2 with addition of 1.0 µM capture probe (coverage rate was calculated to be 35.8%). This value (for details of the calculation, see the Supporting Information) agrees reasonably well with the findings reported by Steel et al.52 Optimization of Assay System. We investigated the optimal amounts of the capture probe ss-DNA and liposomal biolabels used for the electrochemical bioassay. First, we applied an excess of the DNA-tagged liposomes to the sensing surface to hybridize with various amounts of the immobilized capture probe DNA (0.1, 1.0, and 10 µM). As indicated in Table 2B, we compared the signal obtained from the control group (without addition of target ssDNA) with that from the experimental group (with addition of 2 × 105 fmol of the target ss-DNA). When more capture probes 2474

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

Figure 3. (A) Square-wave voltammetric traces acquired when using the competitive genosensor to detect different amounts of the target rfbE gene. (B) Dose-response curve for the rfbE target gene. Each data point represents an average ( 1 standard deviation of three replicates.

Figure 4. Square-wave voltammetric traces obtained from the competitive genosensor hybridized with 2.5 × 105 fmol of (a) the target rfbE gene and (b) ssrA gene. (c) Control assay performed without the addition of the target rfbE gene.

were immobilized on the SPE, we obtained higher current signals from the control assays. The lowest current signal resulted when using the 0.1 µM capture probe solution, presumably because fewer hybridization sites were provided

Table 2. (A) Cartoon Representations of the Analyses of (I) the Control Group in the Absence of Target DNA and (II) the Experimental Group with Addition of 2 × 105 fmol of Target DNA, (B) Signal-Ratio Differences for Various Amounts of Capture DNA on the Modified Electrodes, and (C) Relationships between the Dilution of the DNA-Tagged Liposome and the Corresponding Electrochemical Signals Obtained from the Experimental and Control Groupsa

a The signal ratio percentage was calculated in terms of the current signal obtained from the experimental group divided by the current obtained from the control group.

by the limited number of immobilized capture probes. By comparing the two sets of data obtained when using the 1.0 and 10 µM solutions, we observed a slightly higher current signal from the immobilization of the 10 µM capture probe solution on the SPE surface. Nevertheless, we detected no significant difference in the signal ratio (i.e., the ratio of the currents obtained from the experimental and control groups). From previous experience developing competitive immunoassays, we suspected that an excess of binding sites (in this case, hybridization sites) might lead to a poor detection limit; thus, we chose 1.0 µM as the optimal capture DNA concentration for the immobilization process. Table 2C summarizes the effect of the dilution factor of the added liposome on the performance of the electrochemical genosensor toward the detection of the target rfbE gene. Samples of the liposome solution (1 µL, containing ∼7.5 × 1010 liposomes) were diluted by 2-, 5-, and 10-fold. We found that the optimal concentration of the added liposome preparation was obtained after 5-fold dilution (i.e., 1 µL of the liposome solution diluted to 5 µL in the appropriate buffer); this sample contained ∼7.5 × 1010 liposomes and encapsulated ∼5.6 × 10-9 mol of the RuHex marker. Determination of rfbE Target. The following experiments were undertaken to develop a competitive binding assay that facilitates the detection of an unknown amount of nonlabeled

target ss-DNA with the help of a known amount of a liposomelabeled ss-DNA competitor. Competitive binding was observed after incubating the sensing surface with various mixtures containing liposome-labeled and nonlabeled target ss-DNA. We investigated the target rfbE gene at contents ranging from 5 × 10-2 to 106 fmol, performing triplicate analyses at each concentration. The current signals of the released liposomal Ru(NH3)63+, obtained using SWV, provided a sigmoidally shaped doseresponse curve whose linear portion was over the range from 1 to 106 fmol (Figure 3B). An increase in the concentration of the target ss-DNA during hybridization resulted in fewer liposomelabeled ss-DNA strands binding to the immobilized capture DNA, which led to a decrease in the peak current. We suspect that the minor shift in the peak potential in Figure 3A might be due to minor surface variations arising during the electrode modification process. Nevertheless, this phenomenon did not have a significant influence on the electrochemical behavior. The current signal decreased in a dose-dependent manner with respect to the amount of target present in the range from 1 to 106 fmol, a wide dynamic range at least 6 orders of magnitude, and provided a limit of detection (LOD) of 0.75 amol per assay, defined by subtracting 3 times the standard deviation of the control (free of target DNA) from its average value. The limit of quantification of the proposed genosensor is 3.26 fmol per assay, calculated by Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

2475

subtracting 10 times the standard deviation of the control (free of target DNA) from its average value. To test the sensor for its ability to withstand nonspecific hybridization, we substituted a noncomplementary ssrA sequence for the target ss-DNA and investigated its interaction with the capture probe-modified electrode. Figure 4 reveals that better hybridization was observed for the complementary target gene, leading to a decrease in the reduction current. No significant change in the current signal occurred in the presence of the noncomplementary ssrA sequence, suggesting that no hybridization occurred between the capture probes and ssrA. Thus, this gene sensor exhibits specificity for the recognition of its target rfbE. CONCLUSION We have developed a simple, yet sensitive, DNA-based detection platform employing liposomal amplification. The controlled electrodeposition of Au nanoparticles provided a nanopatterned electrode that we used to conveniently prepare a genosensor exhibiting advanced electrochemical functions and providing an attractive surface for the immobilization of thiolated capture DNA probes. Using SPEs as sensing substrates offers several advantages: simplicity, rapidity, cost-effectiveness, and disposability. We selected synthetic DNA molecules homologous to part of the rfbE gene as the target sequence for the detection of E. coli O157. Our newly developed assay for the rfbE gene, with its LOD of 0.75 amol of DNA (equivalent to 5 µL of a 0.15 pM solution) and exceptional dynamic range of 6 orders of magnitude, is much more

2476

Analytical Chemistry, Vol. 81, No. 7, April 1, 2009

sensitive than those reported previously; it is more than 1300 times more sensitive than the sandwich strip assay system reported by Baeumner53 for the detection of E. coli clpB gene (LOD, 1 fmol per assay) and it is 650 times more sensitive than the nanoparticle amplification-based quartz crystal microbalance DNA sensor (LOD, 0.5 fmol, equivalent to 500 µL of a 10-12 M solution) reported by Mao54 for the detection of the E. coli eaeA gene. We believe that our significant improvement in LOD arose because of (i) the nanostructured pattern modified on the SPE increasing the efficiency of electron transfer and (ii) the signal amplification provided by of RuHex-encapsulated liposomes. Our electrochemical genosensor is capable of detecting the rfbE gene within 15 min (excluding the 40 min hybridization time). The low detection limit, high assay speed, cost-effectiveness, and portability enable this proposed platform technology to be extended to the monitoring and surveillance of various food pathogenic bacteria. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review September 28, 2008. Accepted February 10, 2009. AC8020517 (53) Baeumner, A. J.; Pretz, J.; Fang, S. Anal. Chem. 2004, 76, 888–894. (54) Mao, X.; Yang, L.; Su, X.-L.; Li, Y. Biosens. Bioelectron. 2006, 21, 1178– 1185.