Electrochemical Coding of Single-Nucleotide Polymorphisms By

The size of the surface-modified Au nanoparticle was found to be 8.46 ± 1.53 nm by using atomic force microscopy. If there is ... Analytical Chemistr...
0 downloads 0 Views 197KB Size
Download PDF
Anal. Chem. 2004, 76, 1877-1884

Electrochemical Coding of Single-Nucleotide Polymorphisms By Monobase-Modified Gold Nanoparticles Kagan Kerman,†,‡ Masato Saito,† Yasutaka Morita,† Yuzuru Takamura,† Mehmet Ozsoz,‡ and Eiichi Tamiya*,†

School of Materials Science, Japan Advanced Institute of Science and Technology, 1-1 Asahidai, Tatsunokuchi, Ishikawa 923-1292, Japan, and Department of Analytical Chemistry, Faculty of Pharmacy, Ege University, Bornova, Izmir, 35100 Turkey

Rapidly increasing information about the human genome requires a fast and simple method for the detection of single-nucleotide polymorphisms (SNPs). To date, the conventional SNP detection technologies have been unable to identify all possible SNPs and needed further development in cost, speed, and sensitivity. Here we describe a novel method to discriminate and code all possible combinations. SNPs were coded by monitoring the changes in the electrochemical signal of the monobase-modified colloidal gold (Au) nanoparticles. First, a chitosan layer was formed on the alkanethiol self-assembled monolayer-modified Au nanoparticle. The monobases were then attached onto the chitosan-coated Au nanoparticles through their 5′ phosphate group via the formation of a phosphoramidate bond with the free amino groups of chitosan. The size of the surface-modified Au nanoparticle was found to be 8.46 ( 1.53 nm by using atomic force microscopy. If there is a SNP in DNA and the mismatched bases are complementary to the monobase, Au nanoparticles accumulate on the electrode surface in the presence of DNA polymerase I (Klenow fragment), thus resulting in a significant change in the Au oxide wave. In this report, monobase-modified Au nanoparticles show not only the presence of a SNP, but also identify which bases are involved within the pair. Especially, the identification of a transversion SNP, which contains a couple of the same pyrimidine or purine bases, is greatly simplified. A model study was performed by using a synthetic 21-base DNA probe related to tumor necrosis factor (TNF-r) along with its all possible mutant combinations. This versatile nanoparticle-based electrochemical protocol is a promising candidate for coding all mutational changes. Current efforts in DNA-based research are focused on making use of the extensive library of the Human Genome Project to reveal the secrets of biological events. The biggest challenge in these efforts is certainly the detection of a single nucleotide * To whom correspondence should be addressed. E-mail: [email protected]. † Japan Advanced Institute of Science and Technology. ‡ Ege University. 10.1021/ac0351872 CCC: $27.50 Published on Web 03/06/2004

© 2004 American Chemical Society

polymorphism (SNP). A base change in the somatic cells may lead to an inherited or a noninherited genetic disease. The transmission of the genetic code depends entirely on the specific pairings of adenine (A) with thymine (T), and cytosine (C) with guanine (G), as first described by Watson and Crick five decades ago.1 The base-pair mismatch types can be grouped into transition SNPs, which pair a purine with the wrong pyrimidine; and transversion SNPs, which pair either two purines or two pyrimidines. The main drawback in SNP identification protocols is that many combinations can be formed even with the alteration of a single base. As a matter of fact, eight possible SNPs can be foreseen; such as A-C, A-A, A-G, C-C, C-T, T-T, T-G, and G-G. In the past decade, several important protocols for detecting SNPs have been described; however, these protocols cannot be applied to detect all possible SNPs listed above. Cleaving the region of the DNA, which contained SNP with a single-strandspecific DNase, such as S1 nuclease2 or T4 endonucleoase VII,3 was a direct approach; however, many different combinations for enzyme and reaction conditions made this protocol a difficult task. Additionally, A-A and T-T transversion SNPs were not cleaved in any mismatched strands.3 Protein recognition of SNPs has been an attractive research field, because no enzyme cleavage is required. A protein, which had the ability to recognize and bind to a SNP was isolated from E. coli, and referred to as mutS.4 Once the mutS bound to the SNP, the detection of this complex could be carried out by using several different analytical methods.5 The main disadvantage of this protocol was that not all mismatches could be detected. For instance, C-C mismatches could not be recognized by mutS.4,5 Bi et al.6 has recently reported a protein chip based on a genetically modified mutS. The genetically modified mutS could successfully detect SNPs in PCR amplified samples. (1) Watson, J. D.; Crick, F. H. C. Nature 1953, 171, 737-738. (2) Dodgson, J. B.; Wells, R. D. Biochemistry 1977, 16, 2374-2379. (3) Youil, R.; Kemper, B. W.; Cotton, R. G. H. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 87-91. (4) Lishanski, A.; Ostrander, E. A.; Rine, J. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 2674-2678. (5) Smith, J.; Modrich, P. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 4374-4379. (6) Bi, L.-J.; Zhou, Y. F.; Zhang, J.-Y.; Zhang, Z.-P.; Xie, B.; Zhang, C.-G. Anal. Chem. 2003, 75, 4113-4119.

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004 1877

The different electrophoretic behavior of mismatched DNA sequences from the perfect complementary ones was monitored by using special gel systems.7 Denaturing gradient gel electrophoresis8 was applied for the identification of SNPs, but this method also had some drawbacks. The nature of the mobility shifts between the perfect and mismatched sequences is yet to be described in detail. A SNP should not be at the far end of a DNA sequence for denaturing gradient gel electrophoresis to detect it. Ganguly et al.9 attached radiolabeled water-soluble carbodiimides to mismatched T and G, and SNP could be detected by monitoring the radioisotope label with polyacrylamide gel electrophoresis. The limitations of this approach, its dependence on only T and G bases and the use of radioactive agents, made it undesirable for routine analysis. In the past decade, several electrochemical10-16 and optical17-20 DNA biosensor schemes have been introduced to detect SNPs. The detection scheme has always been to detect the presence of a SNP in a given target DNA sequence. In this report, the identification of the bases within the SNP is revealed for the first time by monobase-modified gold (Au) nanoparticles. Au nanoparticles have been widely used in the design of DNA biosensors owing to their unique optical and electrochemical properties.21-26 Taton et al.27 labeled oligonucleotide targets with Au nanoparticles and monitored the melting profiles of the hybrids. In connection with the Au nanoparticle-catalyzed silver (Ag) reduction, SNPs were selectively detected on an optical scanometric array.28 Park et al.29 hybridized Au-tagged probes with their target oligonucleotides immobilized in a nanogap between two electrodes. A (7) Kourkine, I. V.; Hestekin, C. N.; Buchholz, B. A.; Barron, A. E. Anal. Chem. 2002, 74, 2565-2572. (8) Khrapko, K.; Hanekamp, J. S.; Thilly, W. G.; Belenkii, A.; Foret, F.; Karger, B. L. Nucleic Acids Res. 1994, 22, 364-369. (9) Ganguly, A.; Prockop, D. J. Nucleic Acids Res. 1990, 18, 3933-3939. (10) Millan, K. M.; Saraulo, A.; Mikkelsen, S. R. Anal. Chem. 1994, 66, 29432948. (11) Palecek, E.; Jelen, F. Crit. Rev. Anal. Chem. 2002, 32, 261-270. (12) Marrazza, G.; Chiti, G.; Mascini, M.; Anichini, M. Clin. Chem. 2000, 46, 31-37. (13) Patolsky, F.; Lichtenstein, A.; Willner, I. Nat. Biotechnol. 2001, 19, 253257. (14) Boon, E. M.; Ceres, D. M.; Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol. 2000, 18, 1096-1100. (15) Ozsoz, M.; Erdem, A.; Kerman, K.; Ozkan, D.; Tugrul, B.; Topcuoglu, N.; Ekren, H.; Taylan, M. Anal. Chem. 2003, 75, 2181-2187. (16) Ozkan, D.; Erdem, A.; Kara, P.; Kerman, K.; Meric, B.; Hassmann, J.; Ozsoz, M. Anal. Chem. 2002, 74, 5931-5936. (17) Zhong, X.-B.; Reynolds, R.; Kidd, J. R.; Kidd, K. K.; Jenison, R.; Marlar, R. A.; Ward, D. C. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 11559-11564. (18) Mariotti, E.; Minunni, M.; Mascini, M. Anal. Chim. Acta 2002, 453, 165172. (19) Wang, X.; Krull, U. J. Anal. Chim. Acta 2002, 470, 57-70. (20) Gerion, D.; Chen, C.. F.; Kannan, B.; Fu, A.; Parak, W. J.; Chen, D. J.; Majumdar, A.; Alivisatos, A. P. Anal. Chem. 2003, 75, 4766-4672. (21) Shipway, A. N.; Katz, E.; Willner, I. ChemPhysChem 2000, 1, 18-52. (22) Niemeyer, C. M. Angew. Chem., Int. Ed. 2001, 40, 4128-4158. (23) Hamad-Schifferli, K.; Schwartz, J. J.; Santos, A. T.; Zhang S.; Jacobson, J. M.; Nature 2002, 415, 152-155. (24) Sauthier, M. L.; Carroll, R. L.; Gorman, C. B.; Franzen, S. Langmuir 2002, 18, 1825-1830. (25) Mirkin. C. A.; Taton, T. A. Nature 2000, 405, 626-627. (26) Taton, T. A.; Lu, G.; Mirkin, C. A. J. Am. Chem. Soc. 2001, 123, 51645165. (27) Taton, T. A.; Mirkin, C. A.; Letsinger, R. L. Science 2000, 289, 1757-1760. (28) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078-1081. (29) Park, S.-J.; Taton, T. A.; Mirkin, C. A. Science 2002, 295, 1503-1506.

1878

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

novel Au nanoparticle-based protein assay has also recently been reported by Mirkin and co-workers.30 Target protein was captured between magnetic microparticles and DNA-coated Au nanoparticles. After separation using a magnetic field, DNA was released from the Au nanoparticles and quantified. Depending on the size of the Au nanoparticle, numerous DNA strands could be released into the solution for each protein molecule captured. Wang et al.31 described a magnetically induced electrochemical detection method for DNA hybridization by binding streptavidin-coated Au nanoparticles to biotinylated hybrid anchored on a magnetic bead. Wang et al.32 also utilized PbS, ZnS, and CdS quantum dots for simultaneous detection of multiple synthetic DNA targets on mercury film electrodes. Braslavsky et al.33 recently reported a new SNP coding technology by using fluorescently labeled monobases, which supported our hypothesis that monobase-modified labels could give information at the single molecule level. DNA polymerase enzyme in connection with Cy-3- and Cy-5-modified monobases and anchored DNA templates enabled optical identification of SNPs. DNA polymerase and a mismatched species of labeled monobase were incubated with the 75-base target DNA for 5 min. After imaging of the surface, the positions of the fluorescent molecules were compared with the positions of the DNA molecules that were detected beforehand. A second reaction was then performed with the fluorescently labeled complementary monobase. When the images were superimposed, it was found that Cy3modified uracyl (U) bound to the complementary A, whereas Cy3modified C did not bind to A. Willner and co-workers34,35 assembled thiolated oligonucleotides, complementary to the target DNA as far as one base before the SNP site, on a Au electrode or a Au-quartz piezoelectric crystal. After hybridizing the target DNA, normal or mutant, with the sensing oligonucleotide, the nonhybridized free base in the SNP site was reacted with a biotinylated nucleotide in the presence of DNA polymerase I (Klenow fragment). The labeled nucleotide was coupled only to the double-stranded hybrid that included the SNP site. Subsequent binding of avidin-alkaline phosphatase to the biotinylated nucleotide on the SNP site and the biocatalyzed precipitation of an insoluble product on the transducer provided the amplified detection of the SNPs. However, Willner and coworkers34,35 did not employ their detection system to the identification of the unknown nonhybridized free bases in the mutant site. In this paper, we have utilized a similar detection system by using monobase-modified gold nanoparticles as the signaling source for the electrochemical identification of these mismatched bases. Here we describe a novel electrochemical method for detection of SNPs by hybridization of mismatched bases with complementary base-modified Au nanoparticles in the presence of DNA polymerase I (Klenow fragment). The procedure provides the basis for the detection of both transversion and transition SNPs (30) Nam, J.-M.; Thaxton, C. S.; Mirkin, C. A. Science 2003, 301, 1884-1886. (31) Wang, J.; Xu, D.; Polsky, R. J. Am. Chem. Soc. 2002, 124, 4208-4209. (32) Wang, J.; Liu, G.; Merkoci, A. J. Am. Chem. Soc. 2003, 125, 3214-3215. (33) Braslavsky, I.; Hebert, B.; Kartalov, E.; Quake, S. R. Proc. Natl. Acad. Sci. 2003, 100, 3960-3964. (34) Willner, I.; Patolsky, F.; Weizmann, Y.; Willner, B. Talanta 2002, 56, 847856. (35) Patolsky, F.; Lichtenstein, A.; Willner, I. Nat. Biotechnol. 2001, 19, 253257.

by using the electrochemistry of Au nanoparticles. A synthetic 21-base oligonucleotide (probe) related to tumor necrosis factor (TNF-R) gene36 was used as a model in this report. TNF-R is a pleiotropic cytokine with proinflammatory and immunoregulatory effects.37 A SNP from G to A at position -308 within the TNF-R promoter was reported to affect the capacity to secrete TNF-R in response to endogenous (e.g., cytokines) or exogenous stimuli (e.g., lipopolysaccharides).38 In this report, several SNP combinations were artificially prepared to challenge the selectivity and sensitivity of the method. Although the optimization studies in this report were performed by using the TNF-R gene, this versatile method can readily be used to code SNPs in amplified DNA fragments related to other genetic disorders. EXPERIMENTAL SECTION Apparatus. Square-wave voltammetry (SWV) was performed with a 660A electrochemical workstation (CH Instruments Inc., Austin, TX). A Hitachi U-3010 spectrophotometer in connection with UV Solutions software (Tokyo, Japan) was used with quartz cuvettes of 1 mL and 10-mm path length. The detection was carried out in a 2.0-mL electrochemical cell containing a carbon paste electrode (CPE, 3.0-mm i.d.), a reference electrode (Ag/ AgCl, Bioanalytical Systems, West Lafayette, IN), and a platinum wire as the auxiliary electrode. Atomic force microscopy (AFM) was performed in air with a commercial unit (SPA400-SPI3800, Seiko Instruments Inc., Chiba, Japan) equipped with a calibrated 20-µm x-y-scan and 10-µm z-scan range PZT-scanner. A silicon nitride tip (SI-DF40, spring constant ) 42 N/m, Seiko Instruments Inc.) was used, and images were taken in a dynamic force mode (DFM mode) at an optimal force. Mica substrates for AFM measurements were purchased from Furuuchi Chemical Co. (Tokyo, Japan). Reagents. Colloidal gold nanoparticles (3.5-6.5 nm, 0.75 A520 units/mL), strepavidin-coated superparamagnetic iron oxide particles (∼1 µm), double-stranded DNA (dsDNA), and singlestranded DNA (ssDNA) from calf thymus were purchased from Sigma. The synthetic oligonucleotides were purchased from Fasmac (Kagawa, Japan) and had the following sequences: Probe: 5′-biotin-CAA GAC CAC CAC TTC GAA ACC-3′ Complementary: 5′-GGT TTC GAA GTG GTG GTC TTG-3′ Guanine mismatch: 5′-GGT TTC GAA GGG GTG GTC TTG-3′ Cytosine mismatch: 5′-GGT TTC GAA GCG GTG GTC TTG-3′ Adenine mismatch: 5′-GGT TTC GAA GAG GTG GTC TTG-3′ Noncomplementary: 5′-GGGGC ACGTT TATCC GTCCC TCCTA GTGGC GTGCCCC-3′ The oligonucleotide stock solutions (100 mg/L) were prepared with 10 mM Tris-HCl and 1 mM EDTA (pH 8.0, TE) and were kept frozen. More dilute solutions of the probe were prepared using 0.50 M acetate buffer containing 20 mM NaCl (pH 4.8, ABS). More dilute solutions of the targets were prepared using 20 mM Tris-HCl buffer solution containing 20 mM NaCl (pH 7.0, TBS). (36) Schlu ¨ ter, B.; Erren, M.; Schotte, H.; Junker, R.; Rust, S.; Assmann, G. Clin. Chim. Acta 2002, 320, 135-138. (37) Rink, L.; Kirchner, H. Int. Arch. Allergy Immunol. 1996, 111, 199-209. (38) Abraham, L. J.; Kroeger, K. M. J. Leukocyte Biol. 1999, 66, 562-566.

All stock and buffer solutions were prepared with ultrapure water from a Millipore Milli-Q system (Bedford, MA). The concentrations of all nucleic acids were determined by following a spectrophotometric method.39 Biotin and chitosan were obtained from Wako Pure Chemical Industries Ltd., (Tokyo, Japan). Adenosine 5′-monophosphate, cytidine 5′-monophosphate, guanosine 5′monophosphate, and thymidine 5′-monophosphate were purchased from ICN Biomedicals Inc. (Irvine, CA). 3-Mercaptopropionic acid (MPA), 3-mercaptopropanol (MP), N-(3-dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride (EDC), and N-hydroxysuccinimide (NHS) were purchased from Aldrich. Cloned Klenow fragment (large-fragment Escherichia coli DNA polymerase I, 200 U) was purchased from Takara Bio Inc. (Shiga, Japan). Other chemicals were of analytical reagent grade. All experiments were conducted at room temperature (23 ( 0.5 °C). Procedure. Biotinylated Probe Modification onto StrepavidinCoated Beads. The bead-labeled probes were prepared using a modified procedure reported by Wang et al.31,32 except for the washing buffer of the beads contained 0.2% Tween 20 and 0.1% SDS after centrifugation for 3 min at 12 000 rpm. Hybridization in Solution. After the probe-modified beads were resuspended in 100 µL of 10 mM phosphate buffer solution (PBS, pH 7.0) with 0.10 M NaCl (hybridization solution), they were pipetted into 15 µg/mL target-containing hybridization solution. The hybridization proceeded for 20 min (unless otherwise reported) at room temperature under gentle mixing. After hybridization, a magnetic field was applied by placing an external magnet beneath the vial to separate the hybrid-modified beads from the solution. The separated beads were resuspended in blank hybridization solution and then separated by magnetic field in a similar fashion three times. Both a magnetic separation and washing of the beads from the solution greatly suppressed the effects of nonspecifically binding species. Preparation of Biotin-Modified CPE. The desired amounts of graphite powder and mineral oil (Acheson 38, Fisher) (30/70% (w/w) graphite/oil) were mixed thoroughly, then a portion of the carbon paste was mixed with biotin so that the final quantity was 5% (w/w). The mixture was tightly packed into a Teflon electrode holder and polished to a smooth finish. The biotin-modified CPE was then dipped into the hybridization solution that contained hybrid-modified strepavidin-coated beads. The immobilization of beads onto CPE continued for 1 h with gentle mixing. Preparation of Monobase-Modified Au Nanoparticles. A selfassembled monolayer (SAM) of monobases on Au nanoparticles was prepared in a manner similar to previous reports,40-42 except for the incubation of the MPA-coated Au nanoparticles in 0.02 M MP for 1 h to act as a spacer thiol. Chitosan of different molecular weights was dissolved in a 5 mM ABS (pH 4.50). The solution was then adjusted to a final concentration of 0.05% chitosan in ABS. A 100-µL portion of Au nanoparticle solution was added to 100 µL of the chitosan solution and kept for 1 h for the immobilization of the primary amines of chitosan onto the (39) Essentials of Molecular Biology; Malacinski, G. M., Freifelder, D., Eds.; Jones and Bartlett Publishers: Boston, 1993. (40) Gooding, J. J.; Praig, V. G.; Hall, E. A. H. Anal. Chem. 1998, 70, 23962402. (41) Yang, W.; Hibbert, D. B.; Gooding, J. J. J. Electroanal. Chem. 2001, 516, 10-16. (42) Kerman, K.; Ozkan, D.; Kara, P.; Meric, B.; Gooding, J. J.; Ozsoz, M. Anal. Chim. Acta 2002, 462, 39-47.

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

1879

carboxyl-terminated SAM after the covalent activation by 2 mM EDC and 5 mM NHS in 50 mM phosphate buffer solution (PBS, pH 7.40). The monobase solution contained 200 µg of monobase/ mL in TBS. A 100-µL portion of monobase solution was added to 100 µL of the Au/chitosan solution and kept for 1 h. The monobases were attached onto the chitosan-coated Au nanoparticles through their 5′ phosphate group via the formation of a phosphoramidate bond with the free amino groups of the immobilized chitosan.43-45 The solution of Au/chitosan/monobasemodified nanoparticles was used for the hybridization without further alterations. Determination of the Loading of the Monobases on ChitosanModified Au Nanoparticles. The procedure was performed in a fashion similar to that reported by Patolsky et al.46 Briefly, the absorption spectrum of the chitosan-modified Au nanoparticles of known concentration (0.30 OD at 520 nm) was recorded prior to monobase immobilization. The absorption spectrum of the monobase-modified Au nanoparticles was normalized to the same OD value at 520 nm for the bare Au nanoparticles. Since DNA is not absorbing at 520 nm, the subtraction of the monobase-modified Au nanoparticles spectrum from the bare Au nanoparticles’ spectrum showed the absorbance of the associated DNA. The absorbance difference at 260 nm provided the quantification of DNA. With the help of the known Au nanoparticle concentration, the monobase coverage on the Au nanoparticle could be calculated. AFM Imaging of Monobase-Modified Au Nanoparticles. To confirm the electrostatic binding event between chitosan and the negatively charged mica surface, mica substrates without any modification were utilized. Au nanoparticles of 0.30 OD from solution were deposited onto the mica. After 2 min, the mica disk was rinsed 2 times with 100 µL of ultrapure water and dried by a gentle stream of nitrogen gas. All AFM operations were done in dynamic force mode (DFM mode) at an optimal force in the moisture control box at room temperature and 30-40% of humidity. Hybridization between Mutant Site and Monobase-Modified Au Nanoparticles. The hybrid-immobilized CPEs were dipped into the desired amount of Au nanoparticle-tagged monobase solution containing 20 U/µL DNA polymerase I (Klenow fragment) in 20 mM TBS (pH 7.8) with 60 mM KCl and 10 mM MgCl2 at 37 °C for 1 h in a 2.0-mL vial. The nonspecifically bound Au nanoparticles were removed by rinsing the electrode with PBS containing 0.30 M NaNO3, as described by Lee et al.47 Quantitation of Surface Coverage of Hybrid-Immobilized Beads on CPE. The surface coverage of the hybrid-immobilized beads was determined by following the voltammetric oxidation signal of guanine in DNA, as reported by Wang et al.48 Briefly, the surface density of double-stranded oligonucleotides was determined by monitoring the oxidation signal of guanine at ∼1.0 V in 0.50 M ABS by using the biotin-modified CPE in connection with square(43) Borchard, G. Adv. Drug Delivery Rev. 2001, 52, 145-150. (44) Corsi, K.; Chellat, F.; Yahia, L’H.; Fernandes, J. C. Biomaterials 2003, 24, 1255-1264. (45) Kara, P.; Kerman, K.; Ozkan, D.; Meric, B.; Erdem, A.; Nielsen, P. E.; Ozsoz, M. Electroanalysis 2002, 14, 1685-1690. (46) Patolsky, F.; Gill, R.; Weizman, Y.; Mokari, T.; Banin, U.; Willner, I. J. Am. Chem. Soc. 2003, 125, 13918-13919. (47) Lee, T. M.-H.; Li, L.-L.; Hsing, I.-M. Langmuir 2003, 19, 4338-4343. (48) Wang, J.; Palecek, E.; Nielsen, P. E.; Rivas, G.; Cai, X.; Shiraishi, H.; Dontha, N. Luo, D.; Farias, P. A. M. J. Am. Chem. Soc. 1996, 118, 7667-7670.

1880

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

wave voltammetry with an amplitude of 25 mV and a step potential of 4 mV at 15 Hz. Voltammetric Transduction. The oxidation signal of Au was measured by using square-wave voltammetry (SWV) in blank TBS by scanning from +0.50 to +1.40 V with the same parameters described above. The raw voltammograms were treated by using the least-squares smoothing (level 21) and linear baseline correction of the electrochemical workstation software. RESULTS AND DISCUSSION An ideal SNP assay should be able to discriminate and code all eight possible SNPs. In this report, we are reporting the achievement of this goal by using specific monobase-modified Au nanoparticles that can hybridize with the mismatched bases and reveal exactly which SNP is present in the sample. This study shows that unknown SNPs can be electrochemically coded by using monobase-modified Au nanoparticles. In the first step, the biotinylated probe is attached onto the strepavidin-coated magnetic beads. The beads are then transferred into a hybridization solution containing the target DNA. After hybridization, a magnetic field is used to separate the hybrid-modified beads from the mixture. After washing the beads with blank PBS followed by magnetic separation three times, a biotin-modified CPE is immersed into the solution, and the hybrid-modified beads are immobilized onto the CPE surface. The collection of the beads from the solution by using strong streptavidin/biotin affinity enables further suppression of the nonspecific binding of the remaining unwanted constitutents onto the electrode surface. Then, in a batch experiment, four hybrid-modified CPEs, which are prepared in the same fashion, are exposed to four different Au nanoparticle solutions, each modified with a different base. Each hybrid should be tested by using each one of the four bases in order to be able to evaluate the voltammograms and, thus, determine the mismatched bases, as illustrated in Scheme 1. The modification of Au nanoparticles with chitosan and monobase layers is illustrated in Scheme 2. Chitosan is covalently attached onto the alkanethiol SAM via EDC/NHS chemistry. Monobases are also attached to the chitosan surface via a phosphoramidate bond. Characterization of the chitosan modification on Au nanoparticles was performed by using AFM, and is displayed in Figure 1A and B. Bare Au nanoparticles without any modification could not be bound onto the negatively charged mica surface. Thus, only 5 ( 2 nanoparticles/µm2 could be observed on the mica surface (Figure 1A). On the contrary, positively charged chitosan polymer modification enabled electrostatic attachment of 73 ( 3 nanoparticles/µm2 (Figure 1B). An ∼15fold increase in the attachment of Au nanoparticles onto the mica surface confirmed the chitosan modification on the surface of the nanoparticles. Line profiles of the nanoparticles revealed that the modification of the bare 5 nm Au nanoparticles (inset Figure 1A) with chitosan increased the size of the nanoparticles up to 8.46 ( 1.53 nm (inset Figure 1B). During this observation, the heights of the nanoparticles were taken into account, because the lateral dimensions were distorted due to the AFM tip dimensions, as recently reported by Patolsky et al.46 If the mutation site has bases complementary to the introduced Au nanoparticles, the hybridization of these bases with the aid of the DNA polymerase I results in an enhancement in the accumulation of Au nanoparticles on the electrode surface, thus also

Scheme 1. Schematic Representation of the Principle for the Electrochemical Identification of SNP by Using Monobase-Modified Au Nanoparticlesa

a Which bases are involved in an unknown SNP can be identified by comparing the voltammetric signals obtained from the four different monobase-modified Au nanoparticles.

in the Au oxidation signal. Although the binding constant of the monobase on the Au nanoparticle and the nonhybridized base should be low, it was assumed that one particle could bind to many mutant sites, and thus, a synergistic association was observed. Monitoring these changes in the electrochemical signal at around 1.25 V, which was previously reported by Gonzales-Garcia et al.49 and Ozsoz et al.,15 enables the coding of mismatched bases in the hybrid. Figure 2 shows the electrochemical signals obtained from a hybrid, which involved a transition A-C mismatch. A hybrid containing a A-C mismatch was prepared by hybridization between the single-stranded probe and the C-mismatch containing the target DNA. The exposure of the electrode to the T-modified Au nanoparticles (Figure 2a) resulted in a high signal, since T could hybridize with the mismatched A bases on the hybrid. As expected, G-modified Au nanoparticles (Figure 2b) showed a high signal after hybridization with the mismatched C bases on the hybrid. On the contrary, hybridization with C- (Figure 2c) and A-modified (Figure 2d) Au nanoparticles resulted in similarly low signals. These low signals are attributed to the nonspecific attachment of Au nanoparticles on the hybrid, which could not be fully avoided by the washing step (see Experimental Section).

Data for the coding of the A-C mismatch by using four different monobase-modified Au nanoparticles are given in vertical bar charts in Figure 3, where each bar shows the mean and the standard deviation of the signal values obtained with five different replicates of the electrode (n ) 5). Another transition mismatch, which was A-G pairing, was also identified as shown in Figure 4 by monitoring the changes in the electrochemical oxidation signal of Au nanoparticles. High signals were observed after hybridization with T- and C-modified Au nanoparticles, whereas low signals were obtained after hybridization with noncomplementary A- and G-modified Au nanoparticles. The detection of a transversion mismatch is greatly simplified by using this protocol. A mismatched pair composed of the same bases provides more targets for the complementary base-modified nanoparticles to attach, thus leading to a greater enhancement in the signal. For example, A-A pairing resulted in a significantly enhanced signal when the electrode was exposed to T-modified Au nanoparticles, as shown in Figure 5, whereas the hybridization with C-, A-, and G-modified Au nanoparticles resulted in low (49) Gonzales-Garcia, M. B.; Costa-Garcia, A. Bioelectrochem. Bioenerg. 1995, 38, 389-395.

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

1881

Scheme 2. Illustration of the Monobase-Modified Gold Nanoparticlea

a The gold nanoparticle is coated with a self-assembled monolayer of alkanethiol (SAM), then chitosan biopolymer was covalently bound to the SAM. The monobases were attached onto the chitosan biopolymer via their phosphate groups, leaving the bases free to hybridize with the mismatched base in the SNP site.

Figure 1. Atomic force microscopic images of 0.30 OD (A) bare Au nanoparticles, and (B) monobase-modified chitosan-coated Au nanoparticles on a nonmodified mica surface. Images were taken in dynamic force mode (DFM mode) at an optimal force.

signals. The high difference between the signals enables us to determine a transversion mismatch in a very simple and rapid scheme. 1882

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

Control experiments were performed by exposing the probemodified beads to 15 µg/mL of noncomplementary target DNA, 10 µg/mL dsDNA, and also ssDNA in separate or in mixture

Figure 2. Electrochemical signals reveal the identity of the mismatched bases. Electrochemical Au oxidation signal obtained from A-C mutation hybrid at 29.75 ( 2.34 pmol/cm2 after interaction with T-modified Au nanoparticles (a), G-modified Au nanoparticles (b), C-modified Au nanoparticles (c), and A-modified Au nanoparticles (d) of 0.30 OD each.

Figure 5. Detection of a transversion SNP. Histograms for A-A SNP at 29.17 ( 1.38 pmol/cm2 by using monobase-modified Au nanoparticles of 0.30 OD each (n ) 5).

Figure 6. Calibration plot for the dependence of Au oxidation signals on the Au nanoparticle concentration after hybridization with target DNA, which would form A-C SNP at 29.75 ( 2.34 pmol/cm2 (n ) 5).

Figure 3. Detection of a transition SNP. Histograms for A-C SNP at 29.75 ( 2.34 pmol/cm2 by using monobase-modified Au nanoparticles of 0.30 OD each with the mean and the standard deviation of the signal values obtained with five different replicates of the electrode (n ) 5).

Figure 4. Detection of a transition SNP. Histograms for A-G SNP at 28.45 ( 1.67 pmol/cm2 by using monobase-modified Au nanoparticles of 0.30 OD each (n ) 5).

setups. In all cases, the nonspecific binding of these DNA molecules to the probe was suppressed by the magnetic separation of the hybrids, as described in the Experimental Section. After the magnetic separation step, the nonspecifically bound DNA was removed from the probe, and the beads were collected from the solution with a biotin-modified electrode. This result indicated that the nonspecific binding effects could be greatly suppressed, which was crucial for detecting an unknown SNP in this study. The perfect complementary hybrid was also challenged with the four monobase-modified Au nanoparticle solutions (not shown). A low signal was obtained after the exposure of the perfect complementary hybrid to each monobase-modified Au nanoparticle solutions. Observation of these signals even with perfect complementary hybrid electrodes showed that Au nanoparticles could also nonspecifically stack themselves into the double helix, but still did not interfere with the SNP coding.

The surface coverage of the chitosan-modified Au nanoparticles with monobases was determined by following a spectroscopic method. The average of five independent measurements of samples indicated that the monobase surface coverage on Au nanoparticles was 33 ( 1.42 pmol/cm2. Despite slight nanoparticle diameter variation in the samples, the area-normalized surface coverages were similar for the four different monobases on the Au nanoparticles. The strong adsorptive accumulation of hybrid DNA-immobilized beads onto biotin-modified CPE was further exploited. The immobilization process on the CPE transducer was monitored via the intrinsic anodic signal of DNA (not shown). Conditions for attaining full surface coverage were assessed by measuring the dependence of the guanine current peak height upon the DNA solution concentration or the adsorption time. Using 90 µg/mL of beads in connection with 6 µg/mL probe and 15 µg/mL target in ABS, surface saturation was observed for adsorption periods longer than 6 min. The stability of the immobilized DNA layer was examined by monitoring its signal dependence upon the immersion time in a stirred blank ABS solution. No diminution of the guanine response was observed over a 30-min period, indicating a stable DNA layer.48 The integration of the peak current height at varying concentrations of A-C SNP containing hybrid oligonucletides indicated the full surface coverage value as 29.75 ( 2.34 pmol/cm2. Such a high density of hybrid molecules provided more binding sites for the Au nanoparticles; thus, this synergistic association enhanced the voltammetric response. The remarkable signal amplification associated with the Au nanoparticle concentration is displayed in Figure 6. Au nanoparticle concentration was monitored by following the absorbance changes at 520 nm before and after hybridization of the monobasemodified nanoparticles to the high density of mutant sites on the electrode surface. Such an effective hybridization step permitted convenient quantitation of trace (nanomolar) levels of Au. The favorable signal-to-noise ratio (S/N) characteristics of these data Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

1883

indicate a detection limit of 0.02 OD (S/N ) 3). Further signal amplification and lowering of the detection limits are expected in connection with longer deposition times.48 When the Au concentration was increased from 0.05 to 0.30 OD, the Au signal increased linearly, reaching saturation above 0.30 OD. Thus, it was concluded that a chitosan-modified Au nanoparticle solution of 0.30 OD would provide the surface saturation. According to the reference method proposed by Millan and Mikkelsen,50 calibration data obtained at the hybrid immobilized and bare CPE (not shown) were used to estimate the partition coefficient of monobase-modified Au nanoparticles in the microenvironment near the CPE surface, as in the following equation,

Aubound ibound - ifree ) Aufree ifree

where Aubound and Aufree are the concentrations of free and bound Au nanoparticle complexes, respectively; ibound is the voltammetric peak current obtained at the hybrid-modified CPE; and ifree the current obtained from the bare CPE. The validity of this equation depended on the following assumptions that were reported by Millan and Mikkelsen:50 (a) the equilibration of the free and bound forms of Au nanoparticles occurred rapidly on the voltammetric time scale, (b) the diffusion coefficient of Au nanoparticles was the same in hybrid DNA as in the bulk solution, and (c) the (50) Millan, K. M.; Mikkelsen, S. R. Anal. Chem. 1993, 65, 2317-2323.

1884

Analytical Chemistry, Vol. 76, No. 7, April 1, 2004

preconcentration of Au nanoparticles into hybrid DNA at CPE surface did not significantly change the optimum Au nanoparticle complex concentration. The estimation of the partition coefficient of Au nanoparticles was made as 0.94 by using the voltammetric peak currents obtained under the same conditions at hybrid-modified and bare CPEs at various Au nanoparticle concentrations. Such a high partition coefficient indicated that monobase-modified Au nanoparticles could efficiently attach onto the high-density surface-immobilized mutant sites with the aid of DNA polymerase I. CONCLUSIONS Electrochemical coding for SNPs by using monobase-modified Au nanoparticles is a versatile method which has the potential to be implemented onto a microarray. The demonstration here that the transition and transverse SNPs can easily be identified in a 21-base probe, even in the presence of interfering DNA molecules, leads us to propose that longer PCR samples can also be studied in connection with magnetic separation. Further implementations of this method into a microarray, which will enable us to monitor mismatched bases in several PCR samples simultaneously, are in progress in our laboratory. ACKNOWLEDGMENT K.K. acknowledges the Monbukagakusho scholarship for research students from the Japan Ministry of Education, Culture, Sports, Science and Technology (MEXT). Received for review October 7, 2003. Accepted January 30, 2004. AC0351872