Comparison of the Nonspecific Binding of DNA-Conjugated Gold

Nov 25, 2008 - The nonspecific binding of DNA-conjugated gold nanoparticles (AuNPs) to solid surfaces is more difficult to control than that of DNA mo...
0 downloads 4 Views 2MB Size
Download PDF
Langmuir 2009, 25, 235-241

235

Comparison of the Nonspecific Binding of DNA-Conjugated Gold Nanoparticles between Polymeric and Monomeric Self-Assembled Monolayers Jagotamoy Das,† Chan-Hwa Huh,† Kiyeon Kwon,† Sangjin Park,‡ Sangyong Jon,‡ Kyuwon Kim,§ and Haesik Yang*,† Department of Chemistry and Chemistry Institute for Functional Materials, Pusan National UniVersity, Busan 609-735, Korea, Department of Life Science, Gwangju Institute of Science and Technology, Gwangju 500-712, Korea, and Department of Chemistry, UniVersity of Incheon, Incheon 402-749, Korea ReceiVed August 5, 2008. ReVised Manuscript ReceiVed October 13, 2008 The nonspecific binding of DNA-conjugated gold nanoparticles (AuNPs) to solid surfaces is more difficult to control than that of DNA molecules due to the more attractive interactions from the large number of DNA molecules per AuNP. This paper reports that the polymeric self-assembled monolayers (SAMs) formed on indium-tin oxide (ITO) electrodes significantly inhibit the nonspecific binding of DNA-conjugated AuNPs. The random copolymers used to prepare the polymeric SAMs consist of three functional parts: an ITO-reactive silane group, a DNA-blocking poly(ethylene glycol) (PEG) group, and an amine-reactive N-acryloxysuccinimide group. In order to compare the polymeric SAMs with various monomeric SAMs, the relative nonspecific binding of the DNA-conjugated AuNPs to the ITO electrodes modified with (3-aminopropyl)triethoxysilane (APTES), 3-aminopropylphosphonic acid, 3-phosphonopropionic acid, or 11-phosphonoundecanoic acid is examined by measuring the electrocatalytic anodic current of hydrazine caused by the nonspecifically absorbed AuNPs and by counting the AuNPs adsorbed onto modified ITO electrodes. Carboxylicacid-terminated and amine-terminated monomeric SAMs cause high levels of nonspecific binding of DNA-conjugated AuNPs. The monomeric SAM modified with the carboxylic-acid-terminated poly(amidoamine) dendrimer shows low levels of nonspecific binding (2.0% nonspecific binding relative to APTES) due to the high surface density of the negative charge. The simply prepared polymeric SAM produces the lowest level of nonspecific binding (0.8% nonspecific binding relative to APTES), resulting from the combined effect of (i) DNA-blocking PEG and carboxylic acid groups and (ii) dense polymeric SAMs. Therefore, thin and dense polymeric SAMs may be effective in electrochemical detection and easy DNA immobilization along with low levels of nonspecific binding.

Introduction Although the nonspecific binding of DNA to solid surfaces is inevitable in heterogeneous phase bioassays, its minimization is essential for sensitive DNA detection.1-3 Highly negatively charged DNA causes strong interactions with positively charged and even neutral surfaces. The modification of solid surfaces with a highly negatively charged species such as poly(acrylic acid),4 carboxylated dextran,5 and carboxylated dendrimer6 is a common strategy for reducing the nonspecific binding of DNA. * To whom correspondence should be addressed. E-mail: hyang@ pusan.ac.kr. Telephone: 82-51-510-3681. Fax: 82-51-516-7421. † Pusan National University. ‡ Gwangju Institute of Science and Technology. § University of Incheon. (1) Sassolas, A.; Leca-Bouvier, B. D.; Blum, L. J. Chem. ReV. 2008, 108, 109–139. (2) Rosi, N. L.; Mirkin, C. A. Chem. ReV. 2005, 105, 1547–1562. (3) Palee`ek, E.; Fojta, M. Talanta 2007, 74, 276–290. (4) Krieg, A.; Ruckstuhl, T.; Seeger, S. Anal. Biochem. 2006, 349, 181–185. (5) Yao, X.; Li, X.; Toledo, F.; Zurita-Lopez, C.; Cutova, M.; Momand, J.; Zhou, F. Anal. Biochem. 2006, 354, 220–228. (6) Benters, R.; Niemeyer, C. M.; Drutschmann, D.; Blohm, D.; Wo¨hrle, D. Nucleic Acids Res. 2002, 30, e10. (7) Schlapak, R.; Armitage, D.; Saucedo-Zeni, N.; Hohage, M.; Howorka, S. Langmuir 2007, 23, 10244–10253. (8) Schlapak, R.; Pammer, P.; Armitage, D.; Zhu, R.; Hinterdorfer, P.; Vaupel, M.; Fru¨hwirth, T.; Howorka, S. Langmuir 2006, 22, 277–285. (9) Plutowski, U.; Jester, S. S.; Lenhert, S.; Kappes, M. M.; Richert, C. AdV. Mater. 2007, 19, 1951–1956. (10) Ameringer, T.; Hinz, M.; Mourran, C.; Seliger, H.; Groll, J.; Moeller, M. Biomacromolecules 2005, 6, 1819–1823. (11) Kyo, M.; Yamamoto, T.; Motohashi, H.; Kamiya, T.; Kuroita, T.; Tanaka, T.; Engel, J. D.; Kawakami, B.; Yamamoto, M. Genes Cells 2004, 9, 153–164. (12) De Paul, S. M.; Falconnet, D.; Pasche, S.; Textor, M.; Abel, A. P.; Kauffmann, E.; Liedtke, R.; Ehrat, M. Anal. Chem. 2005, 77, 5831–5838.

Another common strategy is to modify the solid surfaces with hydrophilic poly(ethylene glycol) (PEG).7-14 Indium-tin oxide (ITO) electrodes are widely used for electrochemical detection on account of their wide potential window, low background current, and good optical transparency.15-20 The strong adsorption of phosphate ions onto ITO electrodes21 causes high levels of nonspecific binding of DNA molecules to ITO electrodes, particularly single-stranded DNA (ssDNA).22 A complete coverage on ITO electrodes with a blocking film containing PEG and/or negatively charged group may be a good way of minimizing such nonspecific binding. However, if the blocking film is thick, the film acts as an insulating layer, which limits electrochemical detection. Therefore, a thin but dense (i.e., fully surface-covering) blocking film is essential (13) Cha, T.-W.; Boiadjiev, V.; Lozano, J.; Yang, H.; Zhu, X.-Y Anal. Biochem. 2002, 311, 27–32. (14) Kannan, B.; Kulkarni, R. P.; Majumdar, A. Nano Lett. 2004, 4, 1521– 1524. (15) Stotter, J.; Show, Y.; Wang, S.; Swain, G. Chem. Mater. 2005, 17, 4880– 4888. (16) Asanov, A. N.; Wilson, W. W.; Oldham, P. B. Anal. Chem. 1998, 70, 1156–1163. (17) Zudans, I.; Paddock, J. R.; Kuramitz, H.; Maghasi, A. T.; Wansapura, C. M.; Conklin, S. D.; Kaval, N.; Shtoyko, T.; Monk, D. J.; Bryan, S. A.; Hubler, T. L.; Richardson, J. N.; Seliskar, C. J.; Heineman, W. R. J. Electroanal. Chem. 2004, 565, 311–320. (18) Das, J.; Aziz, M. A.; Yang, H. J. Am. Chem. Soc. 2006, 128, 16022– 16023. (19) Aziz, M. A.; Park, S.; Jon, S.; Yang, H. Chem. Commun. 2007, 2610– 2612. (20) Das, J.; Jo, K.; Lee, J. W.; Yang, H. Anal. Chem. 2007, 79, 2790–2796. (21) Popovich, N. D.; Yen, B. K.; Wong, S.-S. Langmuir 2003, 19, 1324– 1329. (22) Armistead, P. M.; Thorp, H. H. Anal. Chem. 2000, 72, 3764–3770.

10.1021/la802531d CCC: $40.75  2009 American Chemical Society Published on Web 11/25/2008

236 Langmuir, Vol. 25, No. 1, 2009

for good electrochemical detection and low levels of nonspecific binding. DNA-conjugated gold nanoparticles (AuNPs) are used as probes in DNA sensors and microarrays on account of their unique optical, electrical, and electrochemical properties.2,23-26 However, the nonspecific binding of DNA-conjugated AuNPs to solid surfaces is much more severe than that of DNA.5,14,27-31 For example, the level of nonspecific binding of DNA-conjugated AuNPs to self-assembled monolayers (SAMs) of 1,11-undecanedicarboxylic acid on ITO electrodes is high,32 whereas that of DNA to SAMs of 1,12-dodecanedicarboxylic acid is quite low.33 Furthermore, the nonspecific binding of DNA-conjugated AuNPs occurs onto Au electrodes modified with dense SAMs,27,28 even containing PEG.30 Multiple attractive interactions between a solid surface and many DNA molecules per DNA-conjugated AuNP may substantially increase the levels of nonspecific binding. Monomeric SAMs formed on silicon oxide or metal oxide surfaces are widely used for low levels of nonspecific binding and easy immobilization of biomolecules.34,35 However, in many cases, these monomeric SAMs are not sufficiently dense to inhibit nonspecific binding. On the other hand, it has been reported that polymeric SAMs of PEG-silane random copolymers on silicon oxide surfaces are quite effective in reducing the levels of nonspecific binding of proteins.36,37 Moreover, the simply prepared polymeric SAMs are ultrathin (∼1 nm) but cover the surface well.36,37 Recently, it was reported that polymeric SAMs of a PEG-silane random copolymer containing an additional amine-reactive group (referred to as poly(TMSMA-r-PEGMAr-NAS)) allow the specific immobilization of proteins as well as the low nonspecific binding of proteins.38,39 This paper reports that polymeric SAMs containing PEG, silane, and carboxylic acid (the hydrolyzed form of polymeric SAMs of poly(TMSMA-r-PEGMA-r-NAS)) formed on ITO electrodes are highly resistant to the nonspecific binding of DNAconjugated AuNPs. The nonspecific binding of DNA-conjugated AuNPs to silane and phosphonate SAMs is also presented for comparison. The relative nonspecific binding is assessed by measuring the electrocatalytic anodic current of hydrazine caused by the nonspecifically absorbed AuNPs and by counting the AuNPs adsorbed on modified ITO electrodes.

Experimental Section Chemicals. The random copolymer poly(TMSMA-r-PEGMAr-NAS) was synthesized by radical polymerization as reported (23) Katz, E.; Willner, I. Angew. Chem., Int. Ed. 2004, 43, 6042–6108. (24) Daniel, M.-C.; Astruc, D. Chem. ReV. 2004, 104, 293–346. (25) Niemeyer, C. M. Angew. Chem., Int. Ed. 2001, 40, 4128–4158. (26) Castan˜eda, M. T.; Alegret, S.; Merkoc¸i, A. Electroanalysis 2007, 19, 743–753. (27) Demers, L. M.; Park, S.-J.; Taton, T. A.; Li, Z.; Mirkin, C. A. Angew. Chem., Int. Ed. 2001, 40, 3071–3073. (28) Zhang, H.; Li, Z.; Mirkin, C. A. AdV. Mater. 2002, 14, 1472–1474. (29) Han, S.; Lin, J.; Satjapipat, M.; Baca, A. J.; Zhou, F. Chem. Commun. 2001, 609–610. (30) He, L.; Musick, M. D.; Nicewarner, S. R.; Salinas, F. G.; Benkovic, S. J.; Natan, M. J.; Keating, C. D. J. Am. Chem. Soc. 2000, 122, 9071–9077. (31) Festag, G.; Steinbru¨ck, A.; Wolff, A.; Csaki, A.; Mo¨ller, R.; Fritzsche, W. J. Fluoresc. 2005, 15, 161–170. (32) Cerruti, M. G.; Sauthier, M.; Leonard, D.; Liu, D.; Duscher, G.; Feldheim, D. L.; Franzen, S. Anal. Chem. 2006, 78, 3282–3288. (33) Napier, M. E.; Thorp, H. H. Langmuir 1997, 13, 6342–6344. (34) Onclin, S.; Ravoo, B. J.; Reinhoudt, D. N. Angew. Chem., Int. Ed. 2005, 44, 6282–6304. (35) Ulman, A. Chem. ReV. 1996, 96, 1533–1554. (36) Jon, S.; Seong, J.; Khademhosseini, A.; Tran, T.-N. T.; Liabinis, P. E.; Langer, R. Langmuir 2003, 19, 9989–9993. (37) Park, S.; Chi, Y. S.; Choi, I. S.; Seong, J.; Jon, S. J. Nanosci. Nanotechnol. 2006, 6, 1–5. (38) Park, S.; Lee, K.-B.; Choi, I. S.; Langer, R.; Jon, S. Langmuir 2007, 23, 10902–10905. (39) Kim, E. J.; Shin, H.-Y.; Park, S.; Sung, D.; Jon, S.; Sampathkumar, S.-G.; Yarema, K. J.; Choi, S.-Y.; Kim, K. Chem. Commun. 2008, 3543–3545.

Das et al. elsewhere.38 The initial feed ratio of the three monomers, poly(ethylene glycol) methyl ether methacrylate, 3-(trimethoxysilyl)propyl methacrylate, and N-acryloxysuccinimide, was 1:1:1. The DNA obtained from Genotech (Daejeon, Korea) had the following sequences: DNA conjugated to AuNP, 5′-GCA ATA TTA ATG AAG-A20-3′-C3-SH and ssDNA, NH2-5′-C9-AAA GAA GCC AGC TCA A-3′. The DNA-conjugated AuNPs were prepared as reported elsewhere.40 All other chemicals were purchased from Aldrich, Fluka, or Sigma and used as received. Modification of the ITO Electrodes. ITO electrodes were pretreated as reported previously.20 The pretreated ITO electrodes were immersed in either of the following solutions for 36 h at room temperature to allow the formation of phosphonate or silane monolayers: an aqueous solution containing 0.1 mM 3-phosphonopropionic acid (PPA), an aqueous solution containing 0.1 mM 3-aminopropylphosphonic acid (APPA), a methanolic solution containing 0.1 mM 11-phosphonoundecanoic acid (PUA), or a methanolic solution containing 3% (3-amidopropyl)triethoxysilane (APTES). The PPA- and APPA-modified electrodes (PPA/ITO and APPA/ITO electrodes) were washed with distilled water, and the PUA- and APTES-modified ITO electrodes (PUA/ITO and APTES/ ITO electrodes) were washed with methanol and distilled water. The preparation of the ITO electrodes modified with amine-terminated generation-4 poly(amidoamine) dendrimer (dendrimer/ITO electrodes) was begun by activating the carboxylic acid group of the PPA/ITO electrodes by immersing the electrodes into a mixed solution of 50 mM 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride and 25 mM N-hydroxysuccinimide for 2 h. A methanolic solution containing 100 µM generation-4 poly(amidoamine) dendrimer was then dropped onto the carboxylic-acid-activated ITO electrodes. The electrodes were incubated in a refrigerator (ca. 7 °C) for 2 h and then washed with 0.1 M phosphate buffer solution (pH 7.4) containing 0.1% sodium dodecyl sulfate to remove the nonspecifically adsorbed dendrimer. The ITO electrodes modified with the amine-terminated SAMs (i.e., APTES/ITO, APPA/ITO, dendrimer/ITO electrodes) were immersed into a N,N-dimethylformamide solution containing 2 g/mL glutaric anhydride (GA) overnight at room temperature. Subsequently, the electrodes were washed with methanol and water.6 For the immobilization of amine-terminated ssDNA, the ITO electrodes modified with the carboxylic-acid-terminated SAMs were dipped into a mixed solution containing 50 mM 1-ethyl-3-[3(dimethylamino)propyl]carbodiimide hydrochloride and 25 mM N-hydroxysuccinimide for 2 h to activate the exposed carboxylic groups. Subsequently, a 0.1 M phosphate buffer solution (pH 7.4) containing 1 µM amine-terminated ssDNA was dropped onto the activated ITO electrode for 2 h at room temperature. The electrodes were washed with a 0.1 M phosphate buffer solution (pH 7.4) containing 0.1% sodium dodecyl sulfate. The polymeric SAMs were formed by dipping the pretreated ITO electrodes into a dichloromethane solution containing 1% random copolymer poly(TMSMA-r-PEGMA-r-NAS) for 2 h at room temperature followed by washing with dichloromethane.38 The polymeric SAM-modified ITO (pSAM/ITO) electrodes were cured at 100 °C for 20 min. The polymeric SAMs were modified with ssDNAs by immersing the pSAM/ITO electrodes into a phosphate buffer solution containing 1 µM amine-terminated ssDNA for 2 h, which allowed the immobilization of amine-terminated ssDNA. This was followed by washing with a 0.1 M phosphate buffer solution (pH 7.4) containing 0.1% sodium dodecyl sulfate. The pSAM/ITO or ssDNA/pSAM/ITO electrodes were immersed into a 0.1 M borate buffer (pH 11) for 1 h to hydrolyze the unreacted N-acryloxysuccinimide group of polymeric SAMs. Nonspecific Binding of DNA-Conjugated AuNPs and Their Activation. A solution containing 0.7 nM DNA-conjugated AuNPs was dropped onto the modified ITO electrodes for 2 h, followed by washing with a gentle flow of a 0.1 M phosphate buffer solution (pH 7.4) containing 0.1% sodium dodecyl sulfate from a washing bottle (40) Selvaraju, T.; Das, J.; Jo, K.; Kwon, K.; Huh, C.-H.; Kim, T. K.; Yang, H. Langmuir 2008, 24, 9883–9888.

Nonspecific Binding of DNA-Conjugated AuNPs

Langmuir, Vol. 25, No. 1, 2009 237

Figure 1. Schematic diagram of the modified ITO electrodes: (a) APTES/ITO, (b) GA/APTES/ITO, (c) ssDNA/GA/APTES/ITO, (d) APPA/ITO, (e) GA/APPA/ITO, (f) ssDNA/GA/APPA/ITO, (g) PPA/ITO, (h) ssDNA/PPA/ITO, (i) PUA/ITO, (j) ssDNA/PUA/ITO, (k) GA/dendrimer/PPA/ITO, (l) ssDNA/GA/dendrimer/PPA/ITO, (m) pSAM/ITO, and (n) ssDNA/pSAM/ITO electrodes. (o) Chemical structure of poly(TMSMA-r-PEGMAr-NAS). ITO, APTES, GA, ssDNA, APPA, PPA, PUA, and pSAM represent indium-tin oxide, (3-aminopropyl)triethoxysilane, glutaric anhydride, single-stranded DNA, 3-aminopropylphosphonic acid, 3-phosphonopropionic acid, 11-phosphonoundecanoic acid, and polymeric self-assembled monolayer, respectively.

for ∼10 s. The AuNPs were activated by dipping the electrodes into a Tris buffer solution containing 10 mM NaBH4 for 15 min, followed by washing with distilled water.41 Electrochemical Measurements and Scanning Electron Microscope (SEM) Images. The electrochemical experiments were carried out using a CHI708C instrument (CH instruments, Inc.). The electrochemical cell consisted of a modified ITO working electrode, a Pt wire counter electrode, and a Ag/AgCl reference electrode. The cell was filled with a 0.1 M phosphate buffer solution (pH 8) containing 2 mM hydrazine. A fresh hydrazine solution was prepared each day. The SEM images were obtained using a Hitachi S-4200 microscope at 15 kV after coating the surfaces with platinum. The images were magnified by 100 000 times.

Results and Discussion Many different types of SAMs, such as silane, phosphonate, and carboxylate monolayers, can be formed on ITO electrodes. Silane (e.g., APTES) monolayers are used most commonly to modify silicone oxide and metal oxide surfaces. On the other hand, phosphonate (e.g., APPA, PPA, and PUA) and carboxylate monolayers are used to modify metal oxide surfaces only. It is known that the surface coverage of carboxylate monolayers on ITO electrodes is lower than that of phosphonate monolayers and that phosphonates bind more strongly to ITO surfaces than carboxylates.33,42,43 Therefore, only silane and phosphonate (41) Das, J.;Yang, H.; submitted. (42) Pawsey, S.; McCormick, M.; Paul, S. D.; Graf, R.; Lee, Y. S.; Reven, L.; Spiess, H. W. J. Am. Chem. Soc. 2003, 125, 4174–4184.

monolayers were selected to evaluate the nonspecific binding of DNA-conjugated AuNPs. Figure 1 shows a schematic diagram of all the modified ITO electrodes used to test the nonspecific binding of DNA-conjugated AuNPs. It is expected that DNA has higher levels of nonspecific binding on amine-terminated surfaces than on carboxylic-acidterminated surfaces. Despite the high levels of nonspecific binding to amine-terminated surfaces, this study first examined the nonspecific binding of DNA-conjugated AuNPs to two amineterminated ITO electrodes (APTES/ITO (Figure 1a) and APPA/ ITO (Figure 1d) electrodes) to determine the level of nonspecific binding of DNA-conjugated AuNPs to amine-terminated surfaces. The relative nonspecific binding of DNA-conjugated AuNPs to the modified ITO electrodes was evaluated by comparing the cyclic voltammograms (Figure 2) or SEM images (Figure 3). The change in the electrocatalytic properties of an electrode caused by the nonspecific binding of DNA-conjugated AuNPs to the electrode was determined from the cyclic voltammograms. The AuNPs show better electrocatalytic properties for hydrazine electrooxidation than the ITO electrodes.41 Therefore, when AuNPs are bound nonspecifically to a modified ITO electrode, the electrocatalytic properties of the electrode are expected to be significantly improved. However, there was no noticeable improvement in the electrocatalytic properties when the DNAconjugated AuNPs were bound nonspecifically (data not shown). (43) Vercelli, B.; Zotti, G.; Schiavon, G.; Zecchin, S.; Berlin, A. Langmuir 2003, 19, 9351–9356.

238 Langmuir, Vol. 25, No. 1, 2009

Das et al.

Figure 2. Cyclic voltammograms of the modified ITO electrodes obtained in a 0.1 M phosphate buffer solution containing 2 mM hydrazine at a scan rate of 50 mV/s. Before cyclic voltammetry, a solution containing 0.7 nM DNA-conjugated AuNPs was dropped onto and spread over the modified ITO electrodes for 2 h, and NaBH4 treatment was then performed by immersing the electrodes for 15 min in a Tris buffer (pH 9) solution containing 10 mM NaBH4. The background of the pSAM/ITO electrode in (f) was obtained without treating the pSAM/ITO electrode with a solution of DNA-conjugated AuNPs.

The electrocatalytic activity of DNA-conjugated AuNPs for hydrazine electrooxidation is very low. The slow electron-transfer kinetics at DNA-conjugated AuNPs requires some overpotential to electrooxidize hydrazine. When the distance between AuNPs and an electrode is short, high anodic current of hydrazine can be observed within an electrochemical window of ITO electrodes. On the other hand, when the distance between AuNPs and an electrode is longer than a certain level, anodic current of hydrazine is not observed within an electrochemical window because of high overpotential for hydrazine electrooxidation that is due to both slow electron transfer kinetics and slow electron tunneling. The long DNA layer of the DNA-conjugated AuNPs impedes electron tunneling between the AuNPs and ITO electrodes.41 It was reported previously that the electrocatalytic properties of AuNPs are enhanced significantly after NaBH4 treatment.41,44 If the electrocatalytic activity of AuNPs is significantly enhanced, the overpotential due to slow electron transfer kinetics is substantially reduced. In this case, we can see anodic current of hydrazine within an electrochemcial window, although the current is not large because the overpotential due to slow electron tunneling is still high. Hence, the DNA-conjugated AuNPs were (44) Das, J.; Patra, S.; Yang, H. Chem. Commun. 2008, 4451–4453.

activated by NaBH4 treatment before obtaining the cyclic voltammograms for hydrazine electrooxidation. If there is no nonspecific binding of DNA-conjugated AuNPs to the modified ITO electrodes, the two cyclic voltammograms would be similar regardless of the electrodes being incubated in a solution containing DNA-conjugated AuNPs. SEM images were used to both visualize and quantify the nonspecifically bound DNA-conjugated AuNPs. AuNPs of 20 nm size were used to easily identify the AuNPs from the SEM images. The relative nonspecific binding was calculated by counting the number of AuNPs per unit area. The relative percentage of nonspecific binding to the APTES/ITO electrodes is defined as 100%. Table 1 shows the relative percentage of nonspecific binding to the other modified ITO electrodes. On the APTES/ITO electrodes, the peak current for hydrazine electrooxidation was relatively high (Figure 2a). The current was slightly higher than that on the APPA/ITO electrodes (Figure 2b). SEM images of the two electrodes (Figure S1a and c in the Supporting Information) show high levels of nonspecific binding of DNA-conjugated AuNPs to the amine-terminated surfaces. Table 1 shows that the relative percentage of nonspecific binding to the APPA/ITO electrodes was 98%. This suggests that there

Nonspecific Binding of DNA-Conjugated AuNPs

Langmuir, Vol. 25, No. 1, 2009 239

Figure 3. SEM images of the modified ITO electrodes that were obtained after incubation in a solution containing DNA-conjugated AuNPs. The electrodes are the (a) ssDNA/GA/APTES/ITO, (b) ssDNA/GA/APPA/ITO, (c) ssDNA/PPA/ITO, (d) ssDNA/PUA/ITO, (e) ssDNA/GA/dendrimer/ PPA/ITO, and (f) ssDNA/pSAM/ITO electrodes.

is a high level of nonspecific binding to the APPA/ITO electrodes, which is similar to that to the APTES/ITO electrodes. The change in relative nonspecific binding by the conversion of an amine group to a carboxylic acid group was assessed. A dicarboxylic-acid-generating compound, that is, GA, was used to convert amine-terminated surfaces into carboxylic-acidterminated surfaces (Figure 1b, e, and k).6 The peak currents on the GA/APTES/ITO and GA/APPA/ITO electrodes were lower than those on the APTES/ITO and APPA/ITO electrodes, respectively. However, the peak currents were still high. The SEM images of the two electrodes (Figure S1b and d in the Supporting Information) also show high levels of nonspecific binding of the DNA-conjugated AuNPs to the GA/APTES/ITO and GA/APPA/ITO electrodes. Table 1 shows that the relative percentage of nonspecific binding was 53 and 60%, respectively.

This suggests that the conversion of an amine group to a carboxylic acid group does not cause a significant decrease in nonspecific binding, even though the electrode surface is highly negatively charged. The nonspecific binding to the two ITO electrodes modified with carboxylic-acid-terminated monolayers (PPA/ITO (Figure 1i) and PUA/ITO (Figure 1j) electrodes) was also measured. Although the PPA/ITO and PUA/ITO electrodes are negatively charged, the anodic current of hydrazine caused by the nonspecific binding of DNA-conjugated AuNPs was quite high (Figure 2c and d). SEM images (Figure S1e and f in the Supporting Information) show high levels of nonspecific binding to these carboxylic-acid-terminated electrodes. Table 1 shows that the relative percentage of nonspecific binding was 94 and 90%, respectively. These results clearly show that the carboxylic-

240 Langmuir, Vol. 25, No. 1, 2009

Das et al.

Table 1. Relative Percentage of Nonspecific Binding of DNA-Conjugated AuNPs to Modified ITO Electrodes type of modified ITO electrode

relative percentage of nonspecific binding (%)

APTES/ITO GA/APTES/ITO ssDNA/GA/APETS/ITO APPA/ITO GA/APPA/ITO ssDNA/GA/APPA/ITO PPA/ITO ssDNA/PPA/ITO PUA/ITO ssDNA/PUA/ITO GA/dendrimer/PPA/ITO ssDNA/GA/dendrimer/PPA/ITO pSAM/ITO ssDNA/pSAM/ITO

100 53 40 98 60 41 94 65 90 55 2.0 2.0 0.8 0.8

acid-terminated electrodes (PPA/ITO and PUA/ITO electrodes) are not significantly more resistant to the nonspecific binding of DNA-conjugated AuNPs than the amine-terminated electrodes (APTES/ITO and APPA/ITO electrodes). Dendrimer-modified surfaces show high immobilization efficiency for DNA because of the high surface density of functional groups.6,45,46 The nonspecific binding to GA-treated generation-4 poly(amidoamine) dendrimer surfaces (Figure 1k) was also investigated. The anodic current of hydrazine on the GA/ dendrimer/PPA/ITO electrodes (Figure 2e) was significantly lower than that of the other carboxylic-acid-terminated electrodes (GA/APTES/ITO, GA/APPA/ITO, PPA/ITO, and PUA/ITO electrodes) (Figure 2a-d). The SEM image (Figure S1g in the Supporting Information) clearly shows lower levels of nonspecific binding of DNA-conjugated AuNPs to the GA/dendrimer/PPA/ ITO electrodes. Table 1 shows that the relative percentage of nonspecific binding was only 2.0%. One dendrimer contains many peripheral amine groups. If most of the amine groups are converted to carboxylic acid groups, the GA/dendrimer/ITO electrodes would become more highly negatively charged than the other carboxylic-acid-terminated electrodes (GA/APTES/ ITO, GA/APPA/ITO, PPA/ITO, and PUA/ITO electrodes). It appears that the high surface density of negative charge makes the electrode surface highly resistant to the nonspecific binding of DNA-conjugated AuNPs. The nonspecific binding to the ITO electrodes modified with polymeric SAMs containing PEG and carboxylic acid (pSAM/ ITO electrodes) (Figure 1m) was examined. The anodic current of hydrazine on the pSAM/ITO electrodes was quite low but was slightly higher than the background current obtained in the absence of nonspecific binding (Figure 2f). The current was even lower than that on the GA/dendrimer/ITO electrodes (Figure 2e). Only two nonspecifically bound AuNPs were observed in the SEM image (Figure S1h in the Supporting Information), indicating very low nonspecific binding. Table 1 shows only 0.8% nonspecific binding to the pSAM/ITO electrodes. The nonspecific binding of DNA-conjugated AuNPs is mainly due to (i) the interaction between DNAs and functional groups on the ITO surface and (ii) the interaction between DNAs and bare ITO surfaces. The use of the functional groups such as PEG and carboxylic acid could allow weak interaction between DNAs and functional groups.4-14 The minimization of the strong interaction between phosphate ions and the bare ITO surface21 (45) Mark, S. S.; Sandhyarani, N.; Zhu, C.; Campagnolo, C.; Batt, C. A. Langmuir 2004, 20, 6808–6817. (46) Benters, R.; Niemeyer, C. M.; Wo¨hrle, D. ChemBioChem 2001, 2, 686– 694.

could allow low levels of nonspecific binding of DNA-conjugated AuNPs to ITO electrodes. The random copolymer used in this study consisted of three parts: ITO surface-reactive silane group, DNA-blocking PEG group, and amine-reactive N-acryloxysuccinimide group (Figure 1o). The carboxylic acid group acts as a DNA-blocking group when it is generated after the hydrolysis of an unreacted N-acryloxysuccinimide group. Accordingly, the presence of both PEG and carboxylic acid groups on the pSAM/ ITO electrodes could lead to very weak interaction between DNAconjugated AuNPs and these functional groups. The polymeric SAMs cover the surface well.36,37 Therefore, a complete coverage of the polymeric SAMs on ITO electrodes could minimize the interaction between phosphate groups of DNA-conjugated AuNPs and the bare ITO surface. Moreover, a high surface density of PEG and carboxylic acid groups on ITO electrodes could also minimize all other types of interaction between DNA-conjugated AuNPs and ITO electrodes. If surface coverage of the polymeric SAMs was not sufficiently high, then high levels of nonspecific binding could occur. Although GA/APTES/ITO, GA/APPA/ ITO, PPA/ITO, and PUA/ITO electrodes contain many carboxylic acid groups, an insufficient surface coverage of monomeric SAMs may cause high levels of nonspecific binding of DNA-conjugated AuNPs. Consequently, the combined effect of (i) DNA-blocking PEG and carboxylic acid groups and (ii) dense polymeric SAMs allows low interaction between DNA-conjugated AuNPs and ITO electrodes, resulting in very low levels of nonspecific binding of DNA-conjugated AuNPs. In addition, the relative nonspecific binding of DNA-conjugated AuNPs was obtained after immobilizing the ssDNAs (Figure 1c, f, j, l, and n). The anodic current of hydrazine on the ssDNA/ GA/APTES/ITO, ssDNA/GA/APPA/ITO, ssDNA/PPA/ITO, and ssDNA/PUA/ITO electrodes was still high but lower than that on the GA/APTES/ITO, GA/APPA/ITO, PPA/ITO, and PUA/ ITO electrodes, respectively (Figure 2a-d). SEM images (Figure 3a-d) and Table 1 show high levels of nonspecific binding even to the ssDNA-modified electrodes. A high surface density of negatively charged ssDNAs was achieved after covalently attaching amine-terminated ssDNA to the carboxylic-acidterminated surfaces. It is possible that some amine-terminated ssDNAs are adsorbed nonspecifically onto the ITO electrodes modified with monomeric SAMs. Therefore, the additional nonspecific binding of DNA to the ssDNA-modified ITO electrodes would be quite low. However, the nonspecific binding of DNA-conjugated AuNPs to the ssDNA-modified ITO electrodes was quite high. This clearly shows that the level of nonspecific binding of DNA-conjugated AuNPs is much higher than that of the ssDNA. The anodic current of the ssDNA/GA/dendrimer/PPA/ITO and ssDNA/pSAM/ITO electrodes was similar to that of the GA/dendrimer/PPA/ITO and pSAM/ITO electrodes, respectively (Figure 2e and f). SEM images (Figure 3e and f) and the relative percentage of nonspecific binding (2.0 and 0.8%) suggest that there is no further decrease in nonspecific binding after modification with ssDNA. It was reported that washing with a formamide solution significantly decreases the level of nonspecific binding of DNAconjugated AuNPs to DNA-conjugated magnetic beads.40,47 However, the cyclic voltammogram of the PPA/ITO electrodes obtained before the formamide treatment was similar to that after the formamide treatment (Figure 2c). This suggests that a formamide treatment does not inhibit the nonspecific binding of DNA-conjugated AuNPs to the PPA/ITO electrodes. It was also found that the use of bovine serum albumin (i.e., blocking agent)31 (47) Hill, H. D.; Mirkin, C. A. Nat. Protoc. 2006, 1, 324–336.

Nonspecific Binding of DNA-Conjugated AuNPs

and sodium dodecyl sulfate (i.e., detergent),48 which are commonly used to reduce the level of nonspecific binding, has no significant effect on the nonspecific binding of DNAconjugated AuNPs. It was reported that filling the defect sites of monomeric SAMs with n-dodecyl phosphate,49 sodium phosphate, sodium pyrophosphate, or sodium tripolyphosphate21 reduces the level of nonspecific binding of DNA. However, treatment with these phosphates did not inhibit the nonspecific binding of DNA-conjugated AuNPs. This shows that the nonspecific binding of DNA-conjugated AuNPs is quite high and cannot be avoided easily by a treatment with blocking agents or detergents. Although the GA/dendrimer/PPA/ITO electrodes are quite resistant to nonspecific binding, their preparation is quite complex. Multiple preparation steps are also required in the case of previously reported dextran-modified surfaces5 and PEG-modified surfaces based on monomeric SAMs.7-14 On the other hand, the polymeric SAMs in this study were prepared in a simple onestep process, and the surfaces were highly resistant to the nonspecific binding of DNA-conjugated AuNPs. Furthermore, the immobilization of amine-terminated ssDNA was readily (48) Rose, K.; Mason, J. O.; Lathe, R. BioTechniques 2002, 33, 54–58. (49) Gao, Z.; Yang, Z. Anal. Chem. 2006, 78, 1470–1477.

Langmuir, Vol. 25, No. 1, 2009 241

achieved because the polymeric SAMs contain an amine-reactive group.

Conclusions It was demonstrated that the nonspecific binding of DNAconjugated AuNPs to ITO electrodes can be reduced significantly using polymeric SAMs containing PEG, silane, and carboxylic acid without a treatment with blocking agents or detergents. On the other hand, the levels of nonspecific binding of DNAconjugated AuNPs to the ITO electrodes modified with carboxylic-acid-terminated or amine-terminated monomeric SAMs were quite high. Ultrathin, simply prepared, and well surfacecovering polymeric SAMs are suitable for electrochemical detection and easy DNA immobilization as well as for low nonspecific binding. Polymeric SAMs could be applied easily to glass and other metal oxide substrates to minimize the nonspecific binding of DNA-conjugated AuNPs, which are used widely as labels in DNA sensors and microarrays. Acknowledgment. This work was supported for two years by Pusan National University Research Grant. Supporting Information Available: Additional SEM images. This material is available free of charge via the Internet at http://pubs.acs.org. LA802531D