Signal-On Architecture for Electrochemical Aptasensors Based on

Nov 6, 2012 - In this work, we described a signal-on architecture for electrochemical aptasensors that is applicable to a wide range of aptamers. Here...
2 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Signal-On Architecture for Electrochemical Aptasensors Based on Multiple Ion Channels Li-Dong Li,* Xiao-Jiao Mu, Yi Peng, Zheng-Bo Chen, Lin Guo,* and Lei Jiang School of Chemistry & Environment, Beihang University, Beijing 100191, China S Supporting Information *

ABSTRACT: In this work, we described a signal-on architecture for electrochemical aptasensors that is applicable to a wide range of aptamers. Herein, we use thrombin as the model sensing target. The signal-on aptasensor is composed of multiple ion channels embedded within a polymeric membrane, with the antithrombin aptamers chemically modified onto the inner walls of each ion channel working as the sensing element. As the thrombin concentration increased, [Ru(NH3)6]3+, which was electrostatically absorbed onto the negatively charged phosphate backbones of aptamers beforehand, was displaced and pushed into the ion channels from the inner walls, leading to an increase in the current of redox cations at the working electrode surface. Compared with the traditional two-electrode design using a single ion channel sensing system, our ion channel sensing system is applied multiple times within an ordinary three-electrode system, providing such advantages as a high signal-to-noise ratio and suitability for a wide variety of redox species. The results indicate that multiple ion channel sensing provides improvements of orders of magnitude in signal sensitivity. In particular, this signal-on architecture avoids the problems of limited signal gain and “false positives”. Moreover, the proposed aptasensor is simple, highly selective, stable, and applicable to real samples.

T

aptasensor that is based on multiple approximately cylindrical ion channels embedded within a mechanically and chemically robust polymeric membrane. The use of multiple cylindrical ion channels overcomes the problems of low signal-to-noise ratios and rectification, allowing more accurate signals to be recorded using a three-electrode design to detect the specific redox peak current. Herein, we use thrombin as a model target molecule. In our design, the working electrode is a polymer membrane with gold nanoparticles14 sprayed on one side that has been previously embedded with multiple cylindrical ion channels. The antithrombin aptamers chemically modified onto the inner walls of each ion channel work as the sensing element. To detect target molecules, the redox-active cation [Ru(NH3)6]3+ is electrostatically absorbed onto the negatively charged phosphate backbones of DNA beforehand. Upon saturation of the DNA-modified surface with Ru(NH3)63+ (based on the adsorption isotherm), one molecule of Ru(NH3)63+ compensates the charges from three phosphate residues at physiological pH. In the presence of thrombin, [Ru(NH3)6]3+ will be increasingly displaced and pushed into the ion channels because of the positively charged nature of human thrombin (pI = 7.6). Thus, the analyte concentration can be quantitatively assessed using the increase in the redox current of [Ru(NH3)6]3+, which

ransport through ion channels has received significant attention in various fields. The controlled passage of ions through natural and artificial ion channels is a fundamental process of interest in biology, physics, and chemistry.1,2 In ion channel sensing, individual molecules pass through an ion channel, producing detectable changes in the ionic channel current. One approach to designing ion channel sensors is to decorate the ion channel walls with recognition sites for a specific analyte present in solution.3 The current of the voltage curves before and after the introduction of an analyte to the solution will change. Chemical modification of the interior ion channel surface with functional molecules may provide a highly efficient means of controlling ionic or molecular transport through the channel openings in response to ambient stimuli, such as applied light,4 pH,5 temperature,6 and specific ions.7 However, some traditional detection schemes based on a single ion channel8 use a picoammeter for measurement, which requires very strict environment conditions due to its extremely low signal-to-noise ratio. Moreover, some traditional methods produce asymmetric nanochannels, such as conical nanochannels, in polymer membranes, causing rectification.9 Thus, an ionic channel sensor with high signal-to-noise ratio and sensitive to multiple-valence ions or large molecules is highly anticipated. Aptamers, which are artificial oligonucleotide receptors originating from in vitro selection (SELEX),10 are widely recognized as highly promising tools for a variety of important applications.11−13 Here, we describe a new type of thrombin © 2012 American Chemical Society

Received: May 3, 2012 Accepted: November 6, 2012 Published: November 6, 2012 10554

dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559

Analytical Chemistry

Article

Figure 1. (a) Sketch and (b) photograph of our designed container. Steps 1−6 show the fabrication steps starting from the PET membrane and ultimately producing multiple ion channel thrombin sensors. The inner wall of each ion channel was first modified with antithrombin aptamers. Next, one side of the PET membrane was sprayed with Au nanopaticles (AuNPs) to create a working electrode. [Ru(NH3)6]3+ was then electrostatically attached to the aptamer backbones. Finally, the introduction of thrombin repels the [Ru(NH3)6]3+, leading to an enhanced redox signal.

Figure 2. SEM images of the PET membrane (a, b) at different magnified scales and (c) the membrane cross section. The figure shows that the ion channels are cylindrical with a diameter of 200 nm and a thickness of 0.7 μm.

■

MATERIALS AND METHODS Apparatus and Reagents. Cyclic voltammetric experiments were performed using a CHI660C electrochemical workstation (Shang Hai Chen Hua Instruments, China). The three-electrode system consisted of a working electrode composed of AuNPs sprayed onto one side of the PET membrane using a sputter coater (SBC-12, KYKY Technology Development Ltd., China), a platinum wire auxiliary electrode, and a Ag/AgCl (sat. KCl) reference electrode. All other reagents were of analytical reagent grade. All solutions were prepared with doubly distilled water. Phosphate buffer pH 7.0 containing 10 mmol/L [Ru(NH3)6]3+ was chosen as the supporting electrolyte for the thrombin assays. The thrombin used throughout all experiments was human thrombin (pI = 7.6). As a cross-reactivity test, bovine hemoglobin (pI = 7.1), bovine serum albumin (pI = 7.8), and lysozyme (pI = 11) were used, all of which were purchased from the Shanghai Sangon Co. The serum used for real sample analysis was human serum, which was purchased from a Beijing hospital. The thickness of the PET membrane, obtained from the GSI Co., was 0.7 μm. Ion Channel Production. In our experiment, approximately 200 nm diameter cylindrical ion channels were created by incubation with an etching solution for 30 min at 35 °C. The SEM images of the ion channels are shown in Figure 2. DNA Preparation. The DNA aptamer (5′-(NH2)-(CH2)6CCA TCT CCA CTT GGT TGG TGT GGT TGG-3′) (synthesized by the Shanghai Sangon Co., purity: >99%) was predesigned by introducing a 15-base sequence from the 3′ end that binds to thrombin with high selectivity and affinity. To immobilize the aptamer in the ion channels, an amine group was introduced to the 5′ end such that it was fastened to the ion channels via the carboxyl groups of the PET membrane by

was transported through the ion channels to the surface of the working electrode. The complete process used for the aptasensor fabrication on the PET (polyester terephthalate) membrane is shown in Figure 1. [Ru(NH3)6]3+ has been used as a redox indicator to detect many different target molecules.15,16 The mechanism of such an indicator is as follows. Initially, [Ru(NH3)6]3+ was electrostatically bound to the negatively charged phosphate backbones of DNA, which had been previously modified on the electrode surface. The target binding event induced the release of the [Ru(NH3)6]3+, thus lowering the electrochemical signal of Ru(NH3)63+ confined on the electrode surface. However, all of the previously reported aptamer-based biosensors based on the redox current change of [Ru(NH3)6]3+ suffer from the limitations associated with a “signal-off” architecture in which the presence of the target decreases the signal strength. As such, the architecture suffers from the problem of limited signal gain because, at most, only 100% of the signal can be suppressed. Moreover, for the signal-off sensor, contaminants that degrade the aptamer or its redox label can give rise to “false positives” that are difficult to distinguish from signals arising from the binding of authentic analyte. These problems would be alleviated by a “signal-on” mechanism that, instead, produces significant increases in peak current upon target recognition.17 Herein, we describe a signal-on biosensor that is fabricated in a simple fashion by the combination of the typical sensor design and the ion channel sensing system. To our best knowledge, this novel signal-on architecture for electrochemical aptamerbased sensors via multiple ion channels has never been reported previously. More importantly, this new design showed excellent sensitivity and selectivity, which may potentially be applied to the detection of many other proteins and small molecules. 10555

dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559

Analytical Chemistry

Article

Figure 3. Sensing system performance. (a) The reduction peaks of cycle voltammograms (CV) obtained after incubation as a function of thrombin concentration. (b) The dependence of the CV reduction peak current on thrombin concentration. (All electrochemical measurements were performed in 0.10 M phosphate buffer solution containing 10 mM [Ru(NH3)6]3+ at pH 7.0 at a scan rate of 100 mV/s). Inset: the CV peak current is linear with the logarithm of thrombin concentration (log C(T)) from 3 to 50 nM. The illustrated error bars represent the standard deviation of five measurements conducted using a single electrode at each concentration. The figure shows that the reduction current increases with increasing thrombin concentration, indicating a signal-on sensing mechanism.

Regeneration of Sensing Interface. After each assay, the used sensing interface can be easily regenerated by removing the residual [Ru(NH3)6]3+ and thrombin from inner ion channels by holding the membrane in the blank phosphate buffer solution (without any [Ru(NH3)6]3+ and thrombin) at +0.5 V for 30 min. The resulting electrode was ready for the next measurement.

dehydration condensation. The mixture of phosphate buffer (pH 7.0) and antithrombin aptamers was heated to 90 °C and allowed to gradually cool to room temperature. This heating and cooling step helps maintain the structural flexibility of the aptamers (for binding thrombin). To test selectivity, a randomly ordered DNA sequence was also used as a control. The DNA sequence was (5′-(NH2)-(CH2)6-CAA-TTT-GAGTGC-TTA-CCT-GAA-CCA-ACC) and was synthesized by the Shanghai Sangon Co. (purity: >99%). Aptasensor Fabrication. The PET membrane was soaked in the 1 μM ssDNA (antithrombin aptamer) solution for 18 h and then soaked in the 1 mM 4-aminothiophenol (p-ATP) solution for 6 h to fill any unoccupied gaps on the ion channel surface to prevent subsequent nonspecific adsorption. Next, one side of the modified PET membrane was sprayed with AuNPs at 50 mA for 8 min using an ion sputtering system (SBC-12, KYKY Technology Development Ltd., China). This membrane was then used as the working electrode (AFM and SEM images of the electrode surface are presented in Figure S1 and Figure S2 in the Supporting Information). The reference and counter electrodes were Ag/AgCl (sat. KCl) and platinum, respectively. The entire fabrication process is shown in Scheme S1 and S2. Next, the as-prepared PET membrane was quickly fixed to a custom-designed container and incubated with 10 mmol/L [Ru(NH3)6]3+ in 0.1 M phosphate buffer (pH 7.0) for 30 min, electrostatically binding the [Ru(NH3)6]3+ to the DNA phosphate backbones. Electrochemical measurements were performed by directly adding various concentrations of thrombin into the solution. The custom-designed container is shown in Figure 1a,b. The container volume is 2 cm3. All potentials were referred to the the reference electrode. To detect thrombin, different concentrations of thrombin were added to 0.1 M phosphate buffer solution and incubated for 0.5 h at 37 °C. Since the mechanism by which the redox probe is held is electrostatic, the ionic strength has been carefully controlled to produce the best sensitivity and signal readout. As shown in Figure S3, the signal of 1 nM thrombin could be maximized at 0.1 M phosphate buffer. Therefore, 0.1 M was chosen as the optimal concentration of buffer for all the electrochemical measurements. Upon target molecule binding, the redox marker [Ru(NH3)6]3+ was released. Next, a negative voltage scan (0 to −0.5 V) was applied, efficiently attracting all released redox cations to the working electrode surface.

■

RESULTS AND DISCUSSION Detection of Thrombin Using [Ru(NH3)6]3+ as a Redox Marker on the Electrode Surface. With increasing of the thrombin concentration, [Ru(NH3)6]3+ was increasingly displaced and released, creating a substantial CV current increase, as shown in Figure 3. In Figure 3, the [Ru(NH3)6]3+ concentration was fixed at 10 mmol/L, and the thrombin concentration was increased from 1 pmol/L to 50 nmol/L. The CV reduction peak increases gradually as the thrombin concentration increases. As single synthetic ion channels show some of the gating functions found in biological systems,18 we attribute this signal increase to two factors. First, the conformation change of the aptamers in the inner channels upon the target binding increases the inner diameter of the ion channels. The ion channel gates have been switched on such that more [Ru(NH3)6]3+ from the bulk solution enters the ion channels to approach the gold electrode. Second, upon the target binding, the [Ru(NH3)6]3+ will be released into the solution because of the positive nature of thrombin. Initially, the former factor dominates the detection process; thus, the reduction peak current increases sharply with the addition of a small amount of thrombin. Once the ion channel gates have been switched on, the inner diameter cannot increase any further. Thus, if the thrombin concentration continues to increase, the latter factor becomes increasingly dominant. Once thrombin enters, a certain amount of [Ru(NH3)6]3+ is released, explaining the strong linear relation occurring after the latter factor becomes dominant. The inset of Figure 3b shows that the signal current is linearly related to the logarithm of the thrombin concentration from 3 to 50 nM. The regression equation was I = 5568 + 430 log C(T) (unit of C, mol/L), and the regression coefficient (R) was 0.997. Of note, as the concentration of thrombin increases from 10 nM to even higher concentrations, the signal no longer increases significantly (as shown in Figure 3b). This slower tendency of current 10556

dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559

Analytical Chemistry

Article

To verify our proposed mechanism, we performed a control experiment that was identical to the actual experiment but with Fe(CN)63− as the redox probe instead of Ru(NH3)63+. As Fe(CN)63− does not bind to the aptamer, increasing the thrombin concentration will increase the number of blocked ion channels, decreasing the redox current of Fe(CN)63− (see Figure S7). Whereas our previously reported signal-off electrochemical aptasensor19 produces an 8.50% suppression for a thrombin concentration of 10 pM, this signal-on sensor produces an approximately 68.4% signal gain for the same concentration, as shown in Table 1.

increase could be attributed to the decreasing inner diameter of ion channels as more and more thrombin molecules enter the channel, leading to the less ion flux and smaller current change. Moreover, to verify the sensing mechanism, a control experiment was performed in the absence of 10 mmol/L [Ru(NH3)6]3+ in the bulk buffer solution, as shown in Figure S4, the results indicate that the signal of the redox indicator also increases with increasing thrombin concentration, which confirms that the release of the redox indicator is responsible for the signal increase. However, compared with the experiment performed in the presence of 10 mmol/L [Ru(NH3)6]3+ in the bulk solution (Figure 3), almost 10-fold decrease in current leads to the bad signal-to-noise ratio and linearity. Thus, keeping the same concentration of redox indicator inside and outside membrane is very important. Otherwise, this big concentration difference will drive the redox indicator to flow away from the electrode surface instead of moving toward it, leading to the worse signal output. In addition, the linearity over the short range of 1−10 pM was also investigated to identify the lowest detection limit (LOD) of this method. As shown in Figure S5, the regression equation was I = 211 + 5.28 C (unit of C, pM), and the regression coefficient (R) was 0.973. Thus, the LOD was 0.6 pM based on the 3σ rule. Theoretically, after the PET membrane was soaked in 1 μM antithrombin aptamer solution for 18 h, the diffusion balance was finally reached. At the aptamer concentration of 1 μM and with volume of each ion channel of 2.20 × 10−9 dm3, the numbers of aptamers bound per nanopore is 1.32 × 109. In reality, the maximum amount of thrombin that can be immobilized in each ion channel is 6.62 × 107 due to the fact that the maximum concentration of thrombin could be detected is 50 nM based on the experimental results. As thrombin includes two binding sites for its aptamer, only 10% of thrombin molecules are accessible for aptamer binding. Thus, fewer aptamers are bound per nanopore than theoretical value due to entropic barriers. On the basis of the channel density of 107/cm2 for PET membrane, the total amount of thrombin immobilized in the sensors is 4.21 × 1014. Of note, the adsorption isotherm of thrombin binding to antithrombin aptamers can be attributed to protein−aptamer interactions on the inner ion channel walls if the following assumptions are made: every binding site is equivalent and a molecule’s ability to bind is independent of whether adjacent sites are occupied. The latter is justified by the fact that DNA aptamer monolayers rarely form islands of very dense aggregates due to the electrostatic repulsion of the phosphate backbones. On the basis of the classical Langmuir model, a linearized form of the adsorption isotherm is

Table 1. Relative Average Signal Change Comparison Between Our Previous Signal-Off Sensor and the Present Signal-On Sensor thrombin concentration

average signal gain from this signal-on sensor, %

RSD (%)a (N = 5)

average signal suppression from previous signal-off sensor, %

RSD (%)a (N = 5)

10 pM 100 pM 1 nM 3 nM 5 nM 10 nM 20 nM

68.4 137 192 226 263 343 397

1.35 2.25 2.46 3.56 3.85 3.98 4.22

8.50 15.3 18.6 23.7 25.4 27.1 32.2

1.58 1.95 2.30 2.35 2.38 2.42 2.39

a

Relative standard deviation.

The signal suppression or gain is calculated as the change in the peak signal upon addition of the target. As another example, 1 nM thrombin produces an 18.6% signal drop using our previous sensor and a 192% signal increase using the new sensor. This improved signal gain increases the sensitivity by orders of magnitude. The LOD was calculated as 0.6 pM based on the 3σ rule. Such a low detection limit is comparable to that of sensors based on amplification by labeling20 or optical sensors using evanescent-wave-induced fluorescence anisotropy.21 Table 2 summarizes the electrochemical aptasensors available for thrombin. Table 2 compares our method to many other thrombin detection methods; our design provides the lowest detection limit. Table 2. LODs of Electrochemical Thrombin Aptasensors

KD c c = + ΔQ /Q (ΔQ /Q )sat (ΔQ /Q )sat

characteristics

LOD

reference

signal-on electronic aptamer sensors based on ion channels MB-taggeda thrombin aptamer sensor (signal-off)

0.6 pM 6.4 nM 3 nM 1 μM

this work 23

signal-on mechanism with MB-tagged DNA sequence two aptamers with different binding sites assembled into a sandwich assay labeled with glucose dehydrogenase in the secondary aptamer self-assembly of the DNA aptamer sensor on a microfabricated thin-film Au electrode via electrochemical impedance spectroscopy. a mixed self-assembled monolayer as a probe in the presence of the reversible redox couple, Fe(CN)63−/4−, using impedance measurements MB-intercalated thrombin aptamer sensor

where c is the thrombin concentration in solution, ΔQ/Q is the sensor signal (Q is the integrated charge of the reduction peak), (ΔQ/Q)sat is the saturated sensor signal, and KD is the dissociation constant. As shown in Figure S6, the experimental results for c/(ΔQ/Q) exhibit an almost perfect linear relationship from 1 pM to 40 nM thrombin with R = 0.9972. On the basis of the linear fit of the above equation, the KD of thrombin and the antithrombin aptamer can be calculated as 1.10 nM. This small dissociation constant indicates that the protein is strongly bound to the antithrombin aptamer.

a

10557

18 21

0.1 nM

24

2 nM

25

11 nM

21

Methylene blue-tagged. dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559

Analytical Chemistry

Article

Real Sample Analysis. To examine the applicability of this aptasensor to real samples, control experiments were performed in the standard solution and in the serum sample. Table 3

away from the electrode surface, which ensures the complete regeneration of he tsensing interface. Aptasensor Selectivity and Stability. The selectivity of the aptasensor was determined by exposing it to similar proteins (1 mM bovine hemoglobin, 1 mM bovine serum albumin, and 1 mM lysozyme). As shown in Figure 4a, the CV reduction peak for 1 nM thrombin increases dramatically relative to the background. However, the signal does not increase further when 1 mM lysozyme, 1 mM bovine serum albumin, and 1 mM bovine hemoglobin are successively added. In fact, the signals for the nonthrombin enzymes at the stated concentrations are even smaller than that of 1 nM thrombin. We infer that the enzymes have blocked the ion channels, suppressing the signal. In Figure 4b, the value at the y-axis was normalized by the CV peak current increase for thrombin versus the background, yielding relative signal changes of 6.54%, 7.81%, and 7.23% for bovine hemoglobin, bovine serum albumin, and lysozyme, respectively. Meanwhile, by changing the addition sequence of different enzymes, another selectivity test was also performed. Herein, test proteins are given before the thrombin is added. The results are shown in Figure S9. It can be seen that the reduction peaks of CVs for the different enzymes other than thrombin are only as small as the background. However, when we added only 1 nM thrombin, the CV reduction peak current increased dramatically. Moreover, the sensor selectivity was also verified using a nonspecific binding aptamer. As shown in Figure 5, the CV reduction peak current decreases rather than increases with increasing thrombin concentration, which can be attributed to an increasing number of blocked ion channels with increasing thrombin content, suppressing the signal. This excellent selectivity arises from the high selectivity of the antithrombin aptamer. The stability of the electrochemical aptasensor is also important in practical applications and was therefore investigated. The results showed that the CV signal increased by a mere 2.60% of the original value after 100 voltage cycles.

Table 3. Aptasensor Recovery for Real Sample Analysis thrombin concentration

average signal gain in standard solution, %

average signal gain in real sample, %

RSD (%)a N=5

coefficient of recovery (%)

10 pM 100 pM 1 nM 3 nM

68.4 137 192 226

70.0 126 200 230

2.30 4.11 3.23 1.92

102 92.0 104 102

a

Relative standard deviation.

shows the average signal gain for the sensor after reaction with different thrombin concentrations in the standard solution and the serum sample. Because thrombin is not present in the blood and plasma of healthy subjects when coagulation is absent,22 the same signal changes were expected for the standard solution and the real sample. As shown in Table 3, the signal gains are almost identical. The coefficients of recovery are within the range of 92.0−104%, which implies that the aptasensor is promising for analytical application in complex biological samples. Regeneration of Sensing Interface. Good reversibility is critically important for the practical applications of potential biosensors in clinical diagnoses and biological monitoring. The used sensing membrane can be readily regenerated via removing the residual [Ru(NH3)6]3+ and thrombin from inner ion channels by holding the membrane in the blank phosphate buffer solution (without any [Ru(NH3)6]3+ and thrombin) at +0.5 V for 30 min. As shown in Figure S8, both measurements were performed at 100 pM thrombin. About 99.3% of the initial signal response was retained, indicating the excellent reversibility. This good reusability should be attributed to the following two factors. First, by the electrostatic repulsion between the positive potential at the electrode surface and redox cations/positively charged thrombin, the residual [Ru(NH3)6]3+ and thrombin could be repelled and pushed out of the membrane. Second, the big concentration difference between the blank buffer solution and the solution inside the ion channels will drive [Ru(NH3)6]3+ and thrombin to move

■

CONCLUSION

In this study, we propose a universal approach to fabricating signal-on electrochemical aptasensors by combining multiple ion channels with an ordinary three-electrode system. In our design, a PET membrane with embedded ion channels and sprayed on one side with gold nanoparticles acts as the working

Figure 4. Aptasensor selectivity. (a) Reduction peaks of cycle voltammograms (CV) for the aptasensor interacting with 1 nM thrombin, 1 mM lysozyme, 1 mM bovine serum albumin (BSA), and 1 mM bovine hemoglobin (BH) successively in the same solution (0.10 M phosphate buffer solution containing 10 mM [Ru(NH3)6]3+ at pH 7.0 at a scan rate of 100 mV/s). (b) Histogram of the absolute relative response of the sensing system to the different proteins. The error bars represent the standard deviation of five measurements conducted with a single electrode at each concentration. The signal changes for the interfering proteins are very small, indicating the excellent selectivity of the sensor. 10558

dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559

Analytical Chemistry

Article

(5) Xia, F.; Guo, W.; Mao, Y. D.; Hou, X.; Xue, J. M.; Xia, H. W.; Wang, L.; Song, Y. L.; Ji, H.; Qi, O. Y.; Wang, Y. G.; Jiang, L. J. Am. Chem. Soc. 2008, 130, 8345−8350. (6) Guo, W.; Xia, H. W.; Xia, F.; Hou, X.; Cao, L. X.; Wang, L.; Xue, J. M.; Zhang, G. Z.; Song, Y. L.; Zhu, D. B.; Wang, Y. G.; Jiang, L. Chemphyschem 2010, 11, 859−864. (7) Hou, X.; Guo, W.; Xia, F.; Nie, F. Q.; Dong, H.; Tian, Y.; Wen, L. P.; Wang, L.; Cao, L. X.; Yang, Y.; Xue, J. M.; Song, Y. L.; Wang, Y. G.; Liu, D. S.; Jiang, L. J. Am. Chem. Soc. 2009, 131, 7800−7805. (8) Hou, X.; Yang, F.; Li, L.; Song, Y. L.; Jiang, L.; Zhu, D. B. J. Am. Chem. Soc. 2010, 132, 11736−11742. (9) Siwy, Z. S. Adv. Funct. Mater. 2006, 16, 735−746. (10) Hermann, T.; Patel, D. J. Science 2000, 287, 820−825. (11) Chang, H. X.; Tang, L. H.; Wang, Y.; Jiang, J. H.; Li, J. H. Anal. Chem. 2010, 82, 2341−2346. (12) Li, L. D.; Chen, Z, B.; Zhao, H. T.; Guo, L.; Mu, X. J. Sens. Actuators, B 2010, 149, 110−115. (13) Chen, Z. B.; Li, L. D.; Zhao, H. T.; Guo, L.; Mu, X. J. Talanta 2011, 83, 1501−1506. (14) Jun, I. K.; Hess, H. Adv. Mater. 2010, 22, 4823−4825. (15) Alan, K. H.; Cheng, B. X. G; Huang, Z. Y. Anal. Chem. 2007, 79, 5158−5164. (16) Shen, L.; Chen, Z.; Li, Y. H.; He, S. L.; Xie, S. B.; Xu, X. D.; Liang, Z. W.; Meng, X.; Li, Q.; Zhu, Z. W.; Li, M. X.; Le, X. C; Shao, Y. H. Anal. Chem. 2008, 80, 6323−6328. (17) Xiao, Y; Piorek, B. D.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2005, 127, 17990−17991. (18) Schiedt, B.; Healy, K.; Morrison, A. P.; Neumann, R.; Siwy, Z. Nucl. Instrum. Methods Phys. Res., Sect. B 2005, 236, 109. (19) Li, L. D.; Zhao, H. T.; Chen, Z. B.; Mu, X. J.; Guo, L. Anal Bioanal Chem. 2010, 398, 563−570. (20) Ikebukuro, K.; Kiyohara, C.; Sode, K. Anal. Lett. 2004, 37, 2901−2909. (21) Potyrailo, R. A.; Conrad, R. C.; Ellington, A. D.; Hieftje, G. M. Anal. Chem. 1998, 70, 3419−3425. (22) Shuman, M. A.; Majerus, P. W. J. Clin. Invest. 1976, 58, 1249− 1258. (23) Xiao, Y.; Lubin, A. A.; Heeger, A. J.; Plaxco, K. W. Angew. Chem. Int. Ed. 2005, 117, 5592−5595. (24) Cai, H.; Lee, T. M. H.; Hsing, I. M. Sens. Actuators, B 2006, 114, 433−437. (25) Radi, A. E.; Sánchez, J. L. A.; Baldrich, E.; O’Sullivan, C. K. Anal. Chem. 2005, 77, 6320−6323.

Figure 5. Sensing system performance. The reduction peaks of cycle voltammograms obtained after incubation at various thrombin concentrations. (The inner walls of the ion channels are modified with a randomly ordered DNA sequence). The decrease in the CV reduction peak current with increasing thrombin concentration indicates excellent DNA selectivity.

electrode. The antithrombin aptamers chemically modified onto the inner walls of each ion channel act as the sensing element, and [Ru(NH3)6]3+ is used as a redox marker. Relative to our previous signal-off aptasensor, which suffers from limited signal change, this novel signal-on architecture produces much larger signal changes for the same target molecule concentration. The signal current is linear with the logarithm of the thrombin concentration (log C(T)) from 3 to 50 nM. The improved signal gain leads to improvements in the signal sensitivity by orders of magnitude. The LOD of this method was 0.6 pM. In addition, this sensor is highly selective, stable, and applicable to real sample analysis. In summary, this approach is promising for the simple, rapid, and sensitive detection of proteins and small molecules.

■

ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (L.-D.L.); [email protected] (L.G.). Tel.: 010-82338162. Fax: 010-82338162. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China (no. 21273001, 20903008, 20973019, and 50725208), Key Project Funding of NSFC of 50831001 and 20973019, NSFC for Outstanding Young Talents (507252008 and 10825419), and National 973 program (2009CB623700).

■

REFERENCES

(1) Wen, L. P.; Hou, X.; Tian, Y.; Nie, F. Q.; Song, Y. L.; Zhai, J.; Jiang, L. Adv. Mater. 2010, 22, 1021−1024. (2) Wen, L. P.; Hou, X.; Tian, Y.; Zhai, J.; Jiang, L. Adv. Funct. Mater. 2010, 20, 2636−2642. (3) Ali, M.; Schiedt, B.; Healy, K.; Neumann, R.; Ensinger, A Nanotechnology 2008, 19, 085713−085719. (4) Wang, G. L.; Bohaty, A. K.; Zharov, I.; White, H. S. J. Am. Chem. Soc. 2006, 128, 13553−13558. 10559

dx.doi.org/10.1021/ac302351n | Anal. Chem. 2012, 84, 10554−10559