Article pubs.acs.org/ac

Template-Independent, in Situ Grown DNA Nanotail Enabling LabelFree Femtomolar Chronocoulometric Detection of Nucleic Acids Fan Yang,†,§ Xian Yang,†,§ Yunzhao Wang,† You Qin,‡ Xiang Liu,† Xiaoqian Yan,† Ke Zou,† Yong Ning,† and Guo-Jun Zhang*,† †

School of Laboratory Medicine, Hubei University of Chinese Medicine, 1 Huangjia Lake West Road, Wuhan 430065, People’s Republic of China ‡ Cancer Center, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, 156 Wujiadun, Wuhan 430022, People’s Republic of China S Supporting Information *

ABSTRACT: A routine electrochemical DNA (E-DNA) sensor requires either an exquisite design of conformation-switchable recognition probe that is critical to facilitate electron transfer at a sensing interface, or a template-dependent DNA amplification, which often involves designing prone-to-false “sticky ends” and labeling redox tags at one end of the signal probes. Here we report an in situ grown DNA nanotail (IGT)-mediated straightforward and template-free signal amplification strategy for highly sensitive and sequence-specific DNA detection. This novel electrochemical IGT (E-IGT) DNA sensor can quantify target nucleic acids in a label-free manner because the electrochemical signals are generated by chronocoulometric interrogation of redox [Ru(NH3)6]3+ that electrostatically and quantitatively binds to the negatively charged phosphate moieties in the electrode surface-attached DNA. By introduction of terminal deoxynucleoside transferase (TdT) to this sensor design, both the sensitivity and selectivity have been significantly enhanced. This DNA sensor achieves an impressive detection limit of 20 fM for a DNA sequence with 22 nucleotides, which is lower than that of an analogous optical DNA sensor by 2 orders of magnitude. More importantly, it exhibits excellent selectivity against even a single-base mismatched sequence. In addition, this novel DNA sensor presents reliable reusability and is capable of measuring target DNA in complex matrixes, such as undiluted human serum, with minimal interference. These advantages make our E-IGT sensor a promising contender in the E-DNA sensor family for medical diagnostics.

T

can significantly enhance the sensor’s response to the target of interest by maximizing mutual advantages.16−18 Nevertheless, these methods suffer from high fabrication cost and complex photolithography procedures. Additionally, the homogeneity of the synthesized nanomaterials in size and surface functionalization was a critical issue for subsequent conjugation, and optical or electronic signal generation. Therefore, it is highly desirable to construct a simple and highly sensitive E-DNA sensor. To meet the requirements, a traditional optical molecularbeacon inspired E-DNA sensor, featuring a conformationswitchable target-responsive probe, has been created.15 But this reagentless sensor contains a typical “signal off” mode that would be hampered by false positive and limited signal increasement (≤100%). To overcome this hurdle, a series of renewed E-DNA-responsive schemes, including prehybridized metastable conformation, aptamer, split aptamer (two halves) and triplex structures, have been presented as “signal-on” sensing modes.19−23 However, these sensors often involve

he pursuit of highly sensitive, selective and label-free biosensing approaches for DNA assay has received significant attention in the past decade, due to the everincreasing demands for rapid and early diagnosis of infectious diseases and cancers.1−4 Many efforts have been devoted to this field up to now. For instance, advanced micro- and nanofabrication allows direct manufacturing of functional micro- and nanostructured devices, such as microfluidics,5,6 nanoplasmonics,7,8 nanomechanical9 and nanowire field-effect transistor (FET) sensors,10,11 for controllable and ultrasensitive DNA measurement. Nanomaterials or enzymes triggered signal amplification is another means to realize sensitive detection of DNA, either in a homogeneous solution or at the solid− liquid interface.12,13 Although a range of DNA sensing platforms have been exploited, the electrochemical DNA (EDNA) sensor has unique appeal as a promising tool for pointof-care diagnostics and many other key applications, including antiterrorism and environmental monitoring, due mainly to its intrinsic sensitive, simple, portable and inexpensive attributes.14,15 Besides, electrochemical sensing components are compatible with advanced micro- and nanofabrication technology as well as functional nanomaterials, and their combination © 2014 American Chemical Society

Received: October 5, 2014 Accepted: November 4, 2014 Published: November 4, 2014 11905

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

Figure 1. Novel IGT strategy for DNA detection. Capture probes terminated with 3′ thiol are self-assembled on the gold electrode surface via Au-s chemistry. In the presence of target DNA, the formed duplex can serve as a primer with the exposed 3′ OH end presented by target DNA bound to probes. This allows TdT to specifically recognize the primer’s 3′ OH group and thus sequentially catalyzes the addition of dNTP at 3′ terminal of the target DNA in a template-free, in situ grown manner. Prior to quantitatively chronocoulometric (CC) interrogation of IGT-mediated signal amplification of DNA detection, adequate RuHex cations are required to electrostatically bind to the negatively charged DNA complex in a stoichiometric approach. In a non-IGT DNA CC assay, a single target molecule provides equal charge with the hybridized probe molecule. In contrast, the IGT strategy can extend the target molecules linearly, leading to a much higher charge over routine method (inset).

single biotin site, whereas its bulky size would cover multiple potential biotin sites, leading to limited HRP binding. Herein, we report a novel TdT triggered sequence extension for highly sensitive DNA detection by chronocoulometric interrogation of the quantitatively bound redox [Ru(NH3)6]3+ (RuHex, a multiply charged transition metal cation) in electrode surface-confined DNA strands (Figure 1). In this work, a template-independent in situ grown DNA nanotail (IGT)-mediated signal amplification strategy is employed. A capture probe with a 3′ thiol end is self-assembled on the gold electrode surface with an optimal intermolecular nanospace. In the presence of target DNA, the rigid hybridized duplex exposes its 3′ OH group in the solution phase that serves as the initial site for the next sequential DNA extension after recognition with TdT. As a result, a long ssDNA nanotail containing several hundreds of nucleotides can be polymerized along the duplex, and thus significantly amplifies the chronocoulometric signal by electrostatically interacting with the electroactive RuHex (a redox complex that can directly reflect the amounts of DNA stands localized at the electrode surface) in a stoichiometric approach. In contrast, in the absence of target DNA, the 3′ OH terminal does not appear, which in turn eliminates the subsequent DNA elongation, leaving the intact capture probe still unhybridized at a background level. By employing this “signal-on” strategy, our novel E-IGT DNA sensor can sensitively probe femtomolar target DNA with excellent discrimination capability for singlenucleotide polymorphisms (SNP). This sensor also demonstrates favorable electrode regeneration and powerful anti-

exquisite design of a conformation-switchable recognition probe that is critical to enable the distinguishable electron transfer at the sensing interface in the presence of target of interest, and the relative low sensitivity is another barrier for further application. Apparently, in an E-DNA sensor, the detection sensitivity mainly depends on the signal variation amplitude of a hybridization event. To increase such an amplitude, a variety of strategies have been proposed to augment the signal turnovers, including supersandwich-type assay to deliver multiple redox moieties,24 nanoparticle-based biobarcode assay,25 noncovalent electrocatalytical reporter system,26,27 rolling circle amplification (RCA)28 and hybridization chain reaction (HCR) initiated strand extension.29,30 Nevertheless, these strategies often require fabricating nanoelectrodes or designing the prone-to-false “sticky ends” as templates to trigger subsequent cascade polymerization. Generally, the enzyme, like horseradish peroxidase (HRP), can also translate a single hybridization event into hundreds of thousands electrochemical signal turnovers. But its nonspecific adsorption on a carbon or gold electrode can produce a large background signal, thus degrading the sensitivity.31 Note that another enzyme, terminal deoxynucleoside transferase (TdT), can directly catalyze the incorporation of multiple dNTP or small molecule conjugated dNTP (e.g., Cy3-dNTP and biotindNTP) into a single-strand DNA (ssDNA) in the presence of its 3′ OH terminal.32−35 Thus, a string of ordered biotin array can be formed by using biotin-dNTP as the substrate of TdT.31,35 However, this strategy is incapable of reflecting the real extension power of TdT because each avidin-HRP binds a 11906

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

immediately used for DNA immobilization. The detailed experimental procedure can be found in the literature.18 Then, 4 μL of thiolated capture probes (3′ SH terminal) at appropriate concentrations were dropped on the electrode overnight at room temperature to form self-assembly monolayers (SAMs). The surface probe density can be modulated by varying the assembly concentration of capture probes (i.e., 0.2, 0.5, 1 and 2 μM). The DNA modified electrodes were subsequently treated with 2 mM MCH for 1 h to fill the unoccupied region and thus assist the SAMs maintaining a favorable interfacial orientation. After that, the functional electrodes were rinsed with Milli-Q water and dried with nitrogen for subsequent hybridization. Target DNA of varied concentrations were added to the DNA modified electrodes, allowing for probe-target binding at 37 °C for 1 h in hybridization buffer. After hybridization, the electrodes were extensively rinsed with washing buffer and dried under a stream of nitrogen. The mismatched DNA hybridization also followed the same procedure as mentioned above. In the practical sample assay, target DNA was spiked into human serum to obtain a concentration of 100 pM, and then the mixture was transferred to the DNA modified electrodes for hybridization. After removal of the unhybridized DNA sequences, the electrodes were then dried with nitrogen. TdT-Mediated Extension Reaction and Signal Amplification. After completion of hybridization, the doublestranded (ds) DNA coated electrodes were immersed into a mixed solution containing both of dATP and TdT. In the presence of the 3′ OH terminal of the captured target DNA, the enzyme TdT could specifically catalyzes the addition of dATP along the target DNA 3′ terminal. The TdT-mediated extension reaction was performed at 37 °C in a hybridization oven (UVP, HB-1000) with constant temperature and humidity. After treatment with the TdT mixture, the electrodes were rinsed with washing buffer and dried under a stream of nitrogen. Subsequently, the amplified signal was quantitatively measured with CC as previously described.18,36

interference in complex matrices that signifies great potential for diagnosis of diseases.

■

EXPERIMENTAL SECTION Materials. All oligonucleotides (Table 1) were synthesized and purified by Sangon Inc. (Shanghai, China). The Table 1. Sequences of Involved Oligonucleotidesa name capture probe target DNA one-mismatched DNA two-mismatched DNA three-mismatched DNA noncomplementary DNA (NC) a

sequence (5′-3′) AACCACACAACCTACTACCTCA-(CH2) 6-SH TGAGGTAGTAGGTTGTGTGGTT TGAGGTAGTTGGTTGTGTGGTT TGAGGTAGTTGGATGTGTGGTT TGAGGTTGTTGGATGTGTGGTT ATGCATGCATGCATGCATGCAA

The mismatched site is highlighted with italic and black underline.

concentrations of each sequence were requantified by OD260 based on their individual absorption coefficients. Capture probe is thiolated with a -(CH2)6- spacer at the 3′ end. Target DNA is a 22 base sequence that contains complementary sequences to capture probe. One-mismatch, two-mismatch and threemismatch sequences are the base mutated counterparts of target DNA, which contains a single-nucleotide, two-nucleotide and three-nucleotide mismatch to the capture probe. DNA NC is a random sequence that is noncognate to target DNA. 6-Mercapto-1-hexanol (MCH), hexaammineruthenium(III) chloride ([Ru(NH3)6]3+, RuHex), and tris(2-carboxyethyl) phosphine hydrochloride (TCEP) were purchased from Sigma-Aldrich (St. Louis, MO). Terminal deoxynucleoside transferase (TdT) and TdT buffer (5×) were purchased from Beyotime Biotechnology Co. Ltd. (Nantong, China). dATP (100 mM) was obtained from Sangon Inc. (Shanghai, China). The following buffer solutions were employed in this study: DNA immobilization buffer, 10 mM phosphate buffered saline (PBS) and 10 mM TCEP (pH 7.4); hybridization buffer, 10 mM PBS (pH 7.4) with 50 mM MgCl2. Buffers for both electrochemistry and electrode washing are 10 mM Tris−HCl solutions (pH 8.0). All solutions were prepared with Milli-Q water (18 MΩ cm resistivity) from a Millipore system. Electrochemical Measurements. All electrochemical measurements were performed with a CHI 660D electrochemical workstation (CH Instruments Inc., Austin, TX). A conventional three-electrode configuration was employed all through the experiment, which was comprised of a gold working electrode, a platinum wire auxiliary electrode, and a Ag/AgCl reference electrode. Cyclic voltammetry (CV) was carried out at a scan rate of 50 mV/s, and chronocoulometry (CC) at a pulse period of 250 ms and pulse width of 700 mV. The electrolyte buffer was thoroughly purged with nitrogen before experiments. Gold Electrode Surface-based DNA Self-Assembly and Hybridization. Gold electrodes (2 mm in diameter, CH Instruments Inc., Austin, TX) were first polished on microcloth (Buehler) with a γ micropolish deagglomerated alumina suspension (0.05 μm) for 5 min. Residual alumina powder was removed by sonicating electrodes in ethanol and water for 5 min, respectively. Then electrodes were electrochemically cleaned in 0.5 M H2SO4 to remove any remaining impurities. After being dried with nitrogen, electrodes were

■

RESULTS AND DISCUSSION IGT-Amplified Detection of DNA Hybridization. Here we implemented a redox active complex, RuHex, as the electrochemical signaling molecule in our E-IGT DNA sensor, because it could quantify the variation of surface-confined DNA strands both in density and length.37,38 CV was first used to interrogate the electrochemical behavior of RuHex localized at the electrode with DNA/MCH monolayer in a low-ionicstrength electrolyte (10 mM Tris−HCl, pH 8.0, containing 10 or 50 μM RuHex) (Figure S1, Supporting Information). Then, we observed two pairs of well-defined peaks corresponding to the reduction and oxidation of redox marker at the mixedmonolayer coated gold electrode (Figure S1, Supporting Information). Based on previous studies, one pair of peaks was ascribed to the diffusion of RuHex into the mixed monolayer, while another pair of peaks between −0.25 and −0.3 V represented the surface-confined redox process of the electrostatically trapped RuHex that was closely related to the amounts of the electrode surface-bound DNA strands.18,39 Previous studies have also revealed that the enhancement of RuHex concentration can increase the capacitive current, leading to high background current.36 Hence, in order to eliminate such diffusion-induced background signals, we first dipped the mixed monolayer functionalized electrode into the electrolyte full of redox mediators (10 μM RuHex) for several 11907

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

Figure 2. Cyclic voltammograms (A) and chronocoulometry curves (B) of C (capture probe, black line), C+T (capture probe hybridized with target molecule, blue line), C+T+A (probe-target duplex with target strand extended by TdT, orange line) in 10 mM Tris buffer (pH 8.0) after incubation with 10 μM RuHex. Scan rate = 50 mV/s. The signal of ΔQ is defined as the variation in the redox charge of RuHex before hybridization and after TdT extension (ΔQ = QC+T+A − QC). Pulse period =250 ms; pulse width = 700 mV. Intercepts at t = 0 in chronocoulometry curves represent redox charges of RuHex trapped in DNA. (C) Assembly concentrations of capture probe ranging from 0.2 μM to 2 μM are optimized to obtain rational signal amplification by nanoscale spacing of the surface-bound probes required for nanoassembly of highly ordered molecular orientation at interface, and to maintain the maximal hybridization efficiency and the minimal steric hindrance for subsequent site-specific recognition of bulk TdT. The orange solid line further exhibits the optimal probe assembly density. Error bars show the standard deviations of measurements taken from at least three independent experiments. (D) Comparison of the signals at varying time periods (i.e., 0.5, 1, 2, 3 and 4 h) of TdT-mediated reaction. The obtained charge signal in the presence of 100 nM target DNA is also compared to that in the absence of target molecules. Error bars represent standard deviations of measurements (n = 3).

We then employed CC to probe the redox process of the entrapped RuHex within diverse surface monolayers identical to that in CV. By contrast, a much obvious electrochemical signal amplification was observed from the CC curves (Figure 2B). The result is consistent with previous reports, in which CC protocol can generate more intense electrochemical response of the surface-confined RuHex than the CV scheme can.37 This is presumably due to the kinetic electroinactivity of a large portion of RuHex complex sequestered in the heterogeneous film during “dynamic” voltammetric scans, while most, if not all, RuHex molecules are electroactive in the “static” chronocoulometric measurements.36,37 Hence, to better demonstrate IGT amplified electrochemical signal, CC is more appropriate. Optimization of Capture Probe Density and Extension Time of TdT. Different from homogeneous DNA hybridization, the surface-based heterogeneous DNA pairing is severely governed by the surface density of immobilized DNA capture probe due mainly to the steric hindrance and electrostatic repulsion stemming from the negative charges of phosphate backbone. Therefore, we prepared a series of DNA probe monolayers with different surface densities by employing varied assembly concentrations. As demonstrated in Figure 2C, the signal variation (ΔQ) first increased along with the DNA probe concentration across the range of 0.2 to 1.0 μM, and then slightly decreased at 2.0 μM in the presence of 10 nM target DNA and TdT extension system (0.2 U/μL TdT and 4 mM

minutes to saturate the surface-tethered DNA sequences with RuHex via electrostatic interactions.39 Then, we made a CV characterization in the low-ionic-strength electrolyte free of RuHex complex. By employing this strategy, we have acquired a set of typical CV curves, which represented unhybridized capture probe, hybridized probe-target DNA duplex and extended complex with TdT catalysis, respectively (Figure 2A). Note that the TdT enabled a noticeable terminal extension along with the 3′ OH of the captured target DNA, which was confirmed by the enlarged redox peaks of the IGT complex. The gel electrophoresis further verified that TdT could extend the free target DNA up to approximately 500 bp that was 50 times of target DNA length (22 nt), and even longer as the ssDNA target prehybridized with capture probe, forming a rigid duplex that can easily expose its 3′ OH terminal (Figure S2, Supporting Information). Also of note, more dATP can be incorporated into the target sequence compared to other mononucleotides (i.e., dTTP, dCTP, dGTP, dNTP mixture) during TdT-mediated strand elongation.33,34,40 Thus, we here employ dATP as the only substrate for TdT. In addition, three pairs of well-defined peaks showed the reduction and oxidation of the electrostatically bound RuHex at the electrode surface, indicating a minimal influence on the IGT complex structures and the possible release of entrapped redox species arising from the low ionic strength buffer. 11908

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

Figure 3. (A) Chronocoulometry curves of electrode surface-tethered capture probe hybridized with target sequence at a series of concentrations: (a) 0 M, (b) 100 fM, (c) 1 pM, (d) 10 pM, (e) 100 pM, (f) 1 nM and (g) 10 nM in 10 mM Tris buffer (pH 8.0) after incubation with 10 μM RuHex. Note that TdT (0.2 U/μL) and dATP (2 mM) are always present in the enzyme initiated polymerization, including the control experiment that is free of target DNA (a, 0 M). (B) Logarithmic plot of charge signal (ΔQ) versus target sequence concentrations (ΔQ definition is the same as that in Figure 2). The orange line represents the plot for target detection in the presence of TdT-mediated signal amplification, whereas the green line represents the plot for target detection in the absence of signal amplification. Such an amplification renders the sensitivity improvement by 2 orders of magnitude (gray box). Inset is the linear relationship between the signal (ΔQ) and the logarithm of target DNA concentrations. Error bars show the standard deviations of measurements taken from at least three independent experiments. Error bars represent standard deviations of measurements (n = 3).

reaction efficiency gradually declined. Nevertheless, this observation is not consistent with a previous investigation, where the optimal reaction time is 1 h.31 We speculate that this discrepancy is mainly derived from the conjugated dATP, which is not a typical substrate for TdT. Furthermore, the bulk avidin-HRP may reduce the biotin exposure and may alter the configuration of the extended “tail”, which cannot reflect the real extension power of TdT. Therefore, we selected 3 h as the optimal enzymatic reaction time for subsequent incorporation of dATP. Sensitive Detection of Target DNA. We then employed the optimized E-IGT sensor to challenge with a series of complementary DNA concentrations ranging from 100 fM to 10 nM. As seen in Figure 3, the generated CC signal intensity was logarithmically related to the concentration of target DNA, spanning a response range of 5 orders of magnitude. Figure 3A represented the typical CC signal across the concentration range from 0 to 10 nM. It was found that a linear relationship between the signal variation (ΔQ) and target DNA concentration (100 fM to 1 nM) with a regression equation (ΔQ = 54.1 log C + 98.4, r = 0.991) (inset in Figure 3B) was obtained. These results were further confirmed by another sensitive electrochemical means, differential pulse voltammetry (DPV), in which a similar detectable concentration range of target DNA as well as linear relationship (Figure S4, Supporting Information) was achieved. Figure 3B also demonstrated that the lowest detectable concentration of target DNA was 100 fM in the presence of IGT, whereas it reached up to 10 pM in the absence of IGT, suggesting that the IGT nanostructure can significantly increase the sensor’s sensitivity. The detection limit of 20 fM can be estimated using 3σ/S (σ is standard deviation of the blank signal and S is the slope of the fit line shown in the inset of Figure 3B). This sensitivity exceeds that of routine fluorescent DNA quantification (usually at the level of nanomolar to picomolar) by 2 or 3 orders of magnitude.34 Also of note, this femtomolar-level sensitivity is either comparable to the sensitivity of nanoparticel-based E-DNA sensor or even better than that of the TdT-assisted HRP catalysis DNA sensor.31,36 Significantly, our scheme neither requires time-consuming conjugation of nanoparticle-DNA complex nor generates background signals arising from

dATP). It should be noted that the identification of TdT concentration and the concentration ratio between TdT and its substrate, dATP, was based on the investigation by Chilkoti group.34 We also found that the signal variation trend was in good agreement with another work that used biotin-dATP as the substrate for TdT to generate a biotin array for multiple avidin-HRP conjugations and subsequent improvement of amperometric signals.31 However, this assembly concentration was inconsistent with that reported in the gold nanoparticle (AuNP)-mediated hybridization and signal amplification.36 We speculate that two reasons are probably responsible for the concerns. First, we make a stepwise preparation of IGT complex in a relatively high salt buffer (50 mM MgCl2, 10 mM PBS) for high-efficient hybridization, whereas in the report the target DNA prehybridizes with the AuNP-loaded partially complementary sequence in a relatively low salt buffer. Second, the formed big AuNP complex with numerous negative charges is hard to approach the densely arrayed capture probes due to the overwhelming steric hindrance. In contrast, the “electroneutral” TdT requires as many duplexes (i.e., 3′ OH terminals) as possible to increase recognition sites and collision probabilities. Additionally, the surface density of capture probe of 1.0 μM assembly concentration is estimated to be approximately 3.6 × 1012 molecules per cm2 using a previously described procedure (see the Supporting Information),38,41 Therefore, the corresponding intermolecular spacing is estimated to be less than 4 nm, which is nearly half of the DNA probe length (22 nt, 7 nm), rendering a more beneficial orientation of the surface probe. We then investigated the enzyme reaction time to further improve the performance of this E-IGT sensor. As showed in Figure 2D, a gradual increase of surface charges occurred accompanied with the elongation of the enzyme reaction time spanned from 0.5 to 4 h, in the presence of 100 nM target DNA as well as 1 μM immobilization concentration of the capture probe. Of note, the increasing amplitude of CC signal slowed down over a reaction time of 3 h, and prior to that, the signal variation was relatively intense. It is probably due to the sufficient substrates of dATP that meet the requirement of high-efficient catalytic addition by TdT in the primary reaction period. With the ongoing excessive depletion of dATP, the 11909

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

nonspecific adsorption of avidin-HRP at the gold surface. To better demonstrate the advantage of our E-IGT sensor, the associated information has been compared with other similar strategies42,43 and listed in Tabel 2. It is speculated that this Table 2. Comparison between the Proposed Assay and Other Reported Methods with Diverse Signal Amplification for Varied Oligonucleotides Detection amplification strategya TdT/cy3-dNTP TdT/BiotindNTP/HRP supersandwich AuNP-DNA conjugates CSDPR/HCR RCA exonuclease III TdT/dATP

LODb

target sequence length

1 pM 1 pM

25 nt 22 nt

100 fM 10 fM

35 nt 38 nt

8 fM

29 nt

100 fM 20 pM 20 fM

21 nt 35 nt 22 nt

read-out protocolc fluorescence electrochemistry (i-t) electrochemistry (SWV) electrochemistry (CC) electrochemistry (DPV) electrochemistry (CC) electrochemistry (DPV) electrochemistry (CC)

ref 34 31

Figure 4. Selectivity of the E-IGT DNA sensor is showed by bar chart of the chronocoulometry responses of blank and other four diverse DNA sequences. The concentration of perfectly complementary (zero mismatch) strand was 100 nM. The concentration of other three sequences (i.e., One mismatch, two mismatch and three mismatch) were identical at 200 nM. Error bars show the standard deviations of measurements taken from at least three independent experiments. Error bars represent standard deviations of measurements (n = 3).

24 36 42 28 43 this work

capture probe (22 nt) and target sequence (22 nt) without any gap domain. In such a case, the hybridization dynamics is enhanced by comparison with a routine “sandwich” hybridization scheme that contains a middle unhybridized gap between the two probe strands upon target binding. It should be noted that our novel sensor can differentiate a single-base mismatch DNA only at 37 °C without relying on stringent thermal control. It is anticipated that the selectivity might be improved by using highly specific LNA or PNA as capture probes, and such an excellent selectivity synergizing with the femtomolar sensitivity will make our novel sensor well-suited for SNP assays. Regeneration and Anti-Interference of E-IGT Sensor. Beyond sensitivity and selectivity, regeneration is also an important feature for biosensors in practical applications such as disease diagnosis toward resource-poor settings. We found the IGT-based E-DNA sensor could be easily regenerated by incubating the sensor in urea (8 M) for 5 min to denaturize the hybridized duplex, and rehybridizing with the target DNA sequence. After the regeneration process was repeated for three cycles, our sensor still retained its original hybridization efficiency with a relative standard deviation of 4.0% for the measurements (Figure 5A). To further demonstrate the superiority of the E-IGT sensor, we challenged directly the detection in complex clinical samples by spiking DNA targets (100 pM) in undiluted human serum. As shown in Figure 5B, a negligible signal variation was achieved after the functionalized sensor was incubated either in buffered saline or in 100% human serum, despite its extremely complex and multicomponent nature that might degrade the performance of the sensor. This superior anti-interference of our DNA sensor is mainly contributed by the excellent base pairing ability of sequence specific hybridization between the probe and target DNA in a relatively high concentration of salts.

a

CSDPR, circular strand-displacement polymerase reaction; HCR, hybridization chain reaction; RCA, rolling circle amplification. bLOD, limit of detection. ci-t, amperometry; SWV, square wave voltammetry; CC, chronocoulometry; DPV, differential pulse voltammetry.

high sensitivity of our DNA sensor may benefit from two respects. First, a target DNA strand can be extended several hundreds of nucleotides long with the assistance of TdT, providing hundreds of phosphate sites to capture equal amounts of redox RuHex. Consequently, a single hybridization event can be turned into nearly 102 redox events, resulting in significantly improved signal intensity. Second, because a relative sensitive chronocoulometric protocol is employed in the work, the “signal on” mode renders a signal improvement over 100%. It is anticipated that the sensitivity could be further increased if other existing technologies, such as high-affinity locked or peptide nucleic acid (LNA or PNA) probes,44,45 electrical field acceleration and magnetically facilitated hybridization,46,47 are combined. Selectivity of E-IGT DNA Sensor. Due to the significance of sensor selectivity, especially in SNP assays, we here investigated the discrimination capability of this sensor by using perfectly complementary sequence, single-base, two-base and three-base mismatched sequences as well as noncognate DNA sequence as targets. As shown in Figure 4, all the mismatched sequences could be readily distinguished from the perfectly matched DNA, in which the concentration of the perfectly matched DNA was halved to that of the mismatched counterparts. It was noticed that only the perfectly matched sequence exhibited prominent signals, whereas signals corresponding to all other mismatched oligonucleotides were not significant. Such an excellent selectivity of the E-IGT sensor may benefit from (1) the employed buffer of low ionic strength (10 mM Tris buffer free of salts, pH 8.0), where the mismatched sequence could release from its pairing target in the presence of strong electrostatic repulsion between the two negatively charged strands due to the disappearance of protective electrostatic screening from high salt concentration, and (2) the complete equal-length hybridization between

■

CONCLUSION In summary, we have developed a template-free, in situ grown DNA nanotail structure-mediated method for highly sensitive and sequence-specific DNA detection of DNA via chronocoulometric measurement. This E-IGT DNA sensing strategy 11910

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

Figure 5. (A) Reusability curves for our E-ISGW DNA sensor in the presence of 100 nM target DNA. The sensor is rinsed with urea for 5 min to regenerate, which can be reused for at least three times. (B) Chronocoulometry responses for the E-IGT DNA sensor in the detection of target sequence (100 pM) in buffer and complex matrices by challenged directly in undiluted (100%) human serum. Error bars show the standard deviations of measurements taken from at least three independent experiments. Error bars represent standard deviations of measurements (n = 3).

■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21305034, 21275040, 21475034, and 81202962)

appears more straightforward than many other existing nanoparticle or nanostructure fueled amplification strategies, and competes well with them in terms of sensitivity and selectivity. Beyond that, this novel sensor has several unique merits. First, it does not require either exquisitely designed DNA template to initiate ordered DNA assembly like HCR, or multiple self-assembly procedures, like AuNP-based amplification assays. Second, TdT can directly extend DNA strands along with 3′ OH terminal of the target DNA, whereas RCA functions requiring both enzyme and circular DNA probe. Third, the small redox RuHex electrostatically binds to DNA sequence in an equal amount way, which can quantitatively reflect the extended DNA and concomitant amplified signals. Fourth, the chronocoulometric scheme can provide surfaceconfined DNA density information to finely tune the nanoscale intermolecular distance for high-efficient interfacial DNA hybridization. Moreover, our novel sensor is reusable, and can be employed directly in undiluted human serum with minimal signal degradation. By incorporation of other emerging technologies, such as nanostructured electrode-based electrocatalysis,26 electric field enhanced hybridization and single-step delivery coupled microscale electrochemical sensing array,46,48 this novel DNA sensor will be a more promising contender in the E-DNA sensor family for clinical diagnosis.

■

■

ASSOCIATED CONTENT

S Supporting Information *

CV of varied modification in RuHex (10 μM or 50 μM), gel electrophoresis, quantification of the surface-immobilized DNA probes and DPV curves of electrode surface-tethered capture probe hybridized with target sequence at a series of concentrations. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Lam, B.; Das, J.; Holmes, R. D.; Live, L.; Sage, A.; Sargent, E. H.; Kelley, S. O. Nat. Commun. 2013, 4, 2001. (2) Newman, A. M.; Bratman, S. V.; To, J.; Wynne, J. F.; Eclov, N. C. W.; Modlin, L. A.; Liu, C. L.; Neal, J. W.; Wakelee, H. A.; Merritt, R. E.; Shrager, J. B.; Loo, B. W., Jr; Alizadeh, A. A.; Diehn, M. Nat. Med. 2014, DOI: 10.1038/nm.3519. (3) Giljohann, D. A.; Mirkin, C. A. Nature 2009, 462, 461−464. (4) Hsieh, K.; Patterson, A. S.; Ferguson, B. S.; Plaxco, K. W.; Soh, H. T. Angew. Chem. 2012, 124, 4980−4984. (5) Hong, J. W.; Studer, V.; Hang, G.; Anderson, W. F.; Quake, S. R. Nat. Biotechnol. 2004, 22, 435−439. (6) Ferguson, B. S.; Buchsbaum, S. F.; Wu, T.-T.; Hsieh, K.; Xiao, Y.; Sun, R.; Soh, H. T. J. Am. Chem. Soc. 2011, 133, 9129−9135. (7) Anker, J. N.; Hall, W. P.; Lyandres, O.; Shah, N. C.; Zhao, J.; Van Duyne, R. P. Nat. Mater. 2008, 7, 442−453. (8) Tokel, O.; Inci, F.; Demirci, U. Chem. Rev. 2014, 114, 5728− 5752. (9) Lu, Y.; Peng, S.; Luo, D.; Lal, A. Nat. Commun. 2011, 2, 578. (10) Gao, A.; Lu, N.; Dai, P.; Li, T.; Pei, H.; Gao, X.; Gong, Y.; Wang, Y.; Fan, C. Nano Lett. 2011, 11, 3974−3978. (11) Yang, F.; Zhang, G.-J. Rev. Anal. Chem. 2014, 33, 95−110. (12) Lei, J.; Ju, H. Chem. Soc. Rev. 2012, 41, 2122−2134. (13) Gerasimova, Y. V.; Kolpashchikov, D. M. Chem. Soc. Rev. 2014, 43, 6405−6438. (14) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol. 2003, 21, 1192−1199. (15) Fan, C.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 9134−9137. (16) Soleymani, L.; Fang, Z.; Sargent, E. H.; Kelley, S. O. Nat. Nanotechnol. 2009, 4, 844−848. (17) Ferguson, B. S.; Buchsbaum, S. F.; Swensen, J. S.; Hsieh, K.; Lou, X.; Soh, H. T. Anal. Chem. 2009, 81, 6503−6508. (18) Zhang, J.; Song, S.; Wang, L.; Pan, D.; Fan, C. Nat. protoc. 2007, 2, 2888−2895. (19) Xiao, Y.; Lubin, A. A.; Baker, B. R.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 16677−16680. (20) Xiao, Y.; Piorek, B. D.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2005, 127, 17990−17991. (21) Zuo, X.; Song, S.; Zhang, J.; Pan, D.; Wang, L.; Fan, C. J. Am. Chem. Soc. 2007, 129, 1042−1043. (22) Zuo, X.; Xiao, Y.; Plaxco, K. W. J. Am. Chem. Soc. 2009, 131, 6944−6945.

AUTHOR INFORMATION

Corresponding Author

*G.-J. Zhang. Tel: +86-27-68890259. Fax: +86-27-68890259. E-mail: [email protected]. Author Contributions §

These authors contributed equally to this work.

Notes

The authors declare no competing financial interest. 11911

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912

Analytical Chemistry

Article

(23) Idili, A.; Amodio, A.; Vidonis, M.; Feinberg-Somerson, J.; Castronovo, M.; Ricci, F. Anal. Chem. 2014, 86, 9013−9019. (24) Xia, F.; White, R. J.; Zuo, X.; Patterson, A.; Xiao, Y.; Kang, D.; Gong, X.; Plaxco, K. W.; Heeger, A. J. J. Am. Chem. Soc. 2010, 132, 14346−14348. (25) Hu, K.; Lan, D.; Li, X.; Zhang, S. Anal. Chem. 2008, 80, 9124− 9130. (26) Lapierre, M. A.; O’Keefe, M.; Taft, B. J.; Kelley, S. O. Anal. Chem. 2003, 75, 6327−6333. (27) Gasparac, R.; Taft, B. J.; Lapierre-Devlin, M. A.; Lazareck, A. D.; Xu, J. M.; Kelley, S. O. J. Am. Chem. Soc. 2004, 126, 12270−12271. (28) Yao, B.; Liu, Y.; Tabata, M.; Zhu, H.; Miyahara, Y. Chem. Commun. 2014, 50, 9704−9706. (29) Chen, X.; Hong, C.-Y.; Lin, Y.-H.; Chen, J.-H.; Chen, G.-N.; Yang, H.-H. Anal. Chem. 2012, 84, 8277−8283. (30) Yin, B.-C.; Guan, Y.-M.; Ye, B.-C. Chem. Commun. 2012, 48, 4208−4210. (31) Wan, Y.; Xu, H.; Su, Y.; Zhu, X.; Song, S.; Fan, C. Biosens. Bioelectron. 2013, 41, 526−531. (32) Anne, A.; Bonnaudat, C.; Demaille, C.; Wang, K. J. Am. Chem. Soc. 2007, 129, 2734−2735. (33) Chow, D. C.; Chilkoti, A. Langmuir 2007, 23, 11712−11717. (34) Tjong, V.; Yu, H.; Hucknall, A.; Rangarajan, S.; Chilkoti, A. Anal. Chem. 2011, 83, 5153−5159. (35) Wan, Y.; wang, P.; Su, Y.; Zhu, X.; Yang, S.; Lu, J.; Gao, J.; Fan, C.; Huang, Q. Biosens. Bioelectron. 2014, 55, 231−236. (36) Zhang, J.; Song, S.; Zhang, L.; Wang, L.; Wu, H.; Pan, D.; Fan, C. J. Am. Chem. Soc. 2006, 128, 8575−8580. (37) Lao, R.; Song, S.; Wu, H.; Wang, L.; Zhang, Z.; He, L.; Fan, C. Anal. Chem. 2005, 77, 6475−6480. (38) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670−4677. (39) Shen, L.; Chen, Z.; Li, Y.; He, S.; Xie, S.; Xu, X.; Liang, Z.; Meng, X.; Li, Q.; Zhu, Z.; Li, M.; Le, X. C.; Shao, Y. Anal. Chem. 2008, 80, 6323−6328. (40) Khan, M. N.; Tjong, V.; Chilkoti, A.; Zharnikov, M. J. Phys. Chem. B 2013, 117, 9929−9938. (41) Shen, L.; Chen, Z.; Li, Y.; Jing, P.; Xie, S.; He, S.; He, P.; Shao, Y. Chem. Commun. 2007, 2169−2171. (42) Wang, C.; Zhou, H.; Zhu, W.; Li, H.; Jiang, J.; Shen, G.; Yu, R. Biosens. Bioelectron. 2013, 47, 324−328. (43) Xuan, F.; Luo, X.; Hsing, I. M. Anal. Chem. 2012, 84, 5216− 5220. (44) Lee, J. M.; Jung, Y. Angew. Chem., Int. Ed. 2011, 50, 12487− 12490. (45) Vasilyeva, E.; Lam, B.; Fang, Z.; Minden, M. D.; Sargent, E. H.; Kelley, S. O. Angew. Chem., Int. Ed. 2011, 50, 4137−4141. (46) Kaiser, W.; Rant, U. J. Am. Chem. Soc. 2010, 132, 7935−7945. (47) Pettersson, E.; Ahmadian, A.; Ståhl, P. L. PLoS One 2013, 8, e70504. (48) Yang, F.; Zuo, X.; Li, Z.; Deng, W.; Shi, J.; Zhang, G.; Huang, Q.; Song, S.; Fan, C. Adv. Mater. 2014, 26, 4671−4676.

11912

dx.doi.org/10.1021/ac503728s | Anal. Chem. 2014, 86, 11905−11912