A Highly Sensitive and Signal-On Electrochemical Aptasensor Based on Restriction Endonuclease Induced Background Elimination

The oligonucleotides were designed according to 'mfold' program (http://mfold.rit.albany.edu/?q=mfold/DNA-Folding-Form) and synthesized by Sangon Biotechnology Co., Ltd. (Shanghai, China) with the sequences: 5^'-SH-(CH₂)₆-GTCGACAAGGATAAATCCTTCAATGAAGTGGGTCTGTCGAC-(CH₂)₆-NH₂-3^'. The underline part referred to recognition site of Sal I endonuclease with a GTCGAC palindrome structure. The italic portion was the sequence of original aptamer for cocaine²⁴ The 5^' and 3^' end of the oligonucleotide were modified with thiol group and amino group, respectively.

Cocaine, caffeine, morphine and theophylline were all obtained from National Institutes for Food and Drug Control (Beijing, China). Ferrocene monocarboxylic acid and 6-Mercaptohecanol (MCH) were bought from Acros organics. N-(3dimethylaminopropyl)-N^'-ethylcarbodiimide (EDC) and N-hydroxysuccinimide (NHS) were received from Sigma. Tris was obtained from Dingguo Biotechnology Co., Ltd. (Changsha, China). All other chemicals were of analytical-reagent grade and used as received. Triply distilled water (resistance > 18 MΩ·cm) was used throughout the experiments.

Endonuclease set for Sal I was bought from Takara Biotechnology Co., Ltd. (Dalian, China), including Sal I endonuclease and 10 × Sal I endonuclease digestion buffer. The digestion buffer contained 50 mM Tris-HCl (pH 7.8), 100 mM NaCl, 10 mM MgCl₂, and 1 mM DTT. Other two buffers were also used as the stock and dilution buffer which contained 50 mM Tris-HCl (pH 7.8), 300 mM NaCl and 1 mM MgCl₂.

Electrochemical measurements were all carried out on CHI 660E electrochemical workstation (Shanghai Chenhua Instruments, China) in a self-made measuring cell at ambient temperature. Conventional three-electrode system was used: gold electrode (polycrystalline gold rod with 2 mm diameter) as the working electrode, KCl saturated calomel electrode (SCE) as the reference electrode and platinum foil as the auxiliary electrode. All potentials were referenced to the SCE reference electrode.

The 3^' terminal of aptamer probe was labeled with ferrocene prepared according to previous work²³ with a minor modification. In brief, 1 mg of ferrocene monocarboxylic acid was mixed thoroughly with 1 mL of fresh prepared stock buffer containing 0.1 M EDC and 0.1 M NHS, followed by adding 500 μL of 1.0 μM aptamer solution into the mixture and stirring at room temperature for 2 h. Subsequently, the resulting ferrocene modified aptamer probe solution was stored in refrigerator at 4°C until use.

A 2 mm diameter gold electrode was polished with 0.3 and 0.05 μm alumina slurries sequentially on microcloth pads until mirror smoothness, followed by rinsing successively in ultrasonic bath with distilled water, absolute alcohol and distilled water for 5 min respectively. Then, the gold electrode was immersed in piranha solution (a mixture of H₂SO₄ and 30% H₂O₂ with a ratio of 7:3 by volume) for 10 min and rinsed with distilled water. The cleaned electrode was subsequently electrochemically treated in 0.1 M H₂SO₄ by cycling the potential between −0.3 and +1.5 V until a reproducible cyclic voltammogram was obtained. After rinsing with distilled water, 20 μL of ferrocene labeled aptamer probe was pipetted on the treated gold electrode and incubated for 2 h in a water-saturated atmosphere. The resulting electrode was rinsed with dilution buffer to remove the unbound probes. Then 20 μL of 1 mM 6-mercaptohexanol solution was dripped onto the aptamer probe immobilized electrode and maintained for 15 min to block the unreacted sites. After rinsing with distilled water, the sensing interface was finally fabricated and ready for cocaine detection.

Cyclic voltammetry and AC impedance were carried out for characterization of the sensor fabrication process. Cyclic voltammetry was performed at a scan rate of 100 mV/s in 5 mL of 10 mM PBS (pH 7.4) containing 5 mM [Fe(CN)₆] ³⁻/[Fe(CN)₆] ⁴⁻ and 0.1 M KCl. The impedance measurements were performed at the formal potential of the system (E° = 240 mV) using alternating voltage of 5 mV. The frequency range was from 1 Hz to 100 kHz.

For target detection, 20 μL of cocaine solution with a certain concentration (prepared with dilution buffer) was dropped on the prepared sensing interface and allowed to incubate at 37°C for 2 h. After rinsing with dilution buffer, 20 μL of endonuclease solution containing 0.9 U/μL Sal I was placed on the electrode surface. The enzyme digestion reaction was allowed to maintain at 37°C for another 2 h. Finally, the resulting electrode was fully rinsed in dilution buffer solution under stirring for 1 min to remove the physically adsorbed molecules and ready for electrochemical measurements. AC voltammogram measurements in 5 mL of 1.0 M NaClO₄ solution from 0.0 to 0.6 V (versus SCE) at a specific frequency were carried out to evaluate the response characteristics of proposed aptasensor. The peak current observed at 0.25 V was used to assess the analytical performances of the aptasensor.

In the present contribution, we proposed a restriction exdonuclease digestion based electrochemical aptasensor to solve the problem of background current and to improve the analytical performances, using cocaine as the model analyte. Figure 1 shows the design of aptamer probe and the detection principle of the electrochemical aptasensor. The aptamer probe contains a 31 bases of origin aptamer sequence (as shown in italic portion in Experimental section) for cocaine with 3 bases extension at the 5^' end and 7 bases extension at the 3^' end, respectively. In the absence of cocaine molecules, the aptamer probe adopts a hairpin structure as shown in the left panel of Figure 1A by predicting with the 'mfold' software. A double-stranded palindrome structure with the sequence of GTCGAC is formed in the stem, which can be recognized specifically by Sal I restriction endonuclease and cutoff between guanine (G) and thymine (T). On the contrary, when binding specifically with cocaine molecules, the aptamer probes undergo structure switching to form tee valve-like conformation, which resulting in the deformation of palindrome structure and the disappearance of recognition site for Sal I as shown in the right panel of Figure 1A. Making use of this unique design, a background current eliminated electrochemical aptasensor was developed for cocaine detection as depicted in Figure 1B. The aptamer probe was labeled with thiol group at the 5^' terminal and electrochemical active ferrocene at the 3^' terminal. In the absence of target molecules, the palindrome structure was specifically recognized and digested by Sal I and the ferrocene-involved cleavage products were moved from the electrode surface. As a result, almost no peak current was observed as shown in the upper panel of Figure 1B, indicating the thorough elimination of background current for blank sample. Introduction of cocaine triggered the structure switching of the aptamer probe and the deformed conformation cannot be cutoff by the endonuclease. Accordingly, a high peak current could be detected since the electrochemical active ferrocene tag was still located close to the electrode surface as seen in the lower panel of Figure 1B. The current intensity directly related to the concentration of cocaine molecules. By this means, not only a background current eliminated electrochemical cocaine aptasensor with high signal-to-noise ratio was developed but also a signal-on signaling scheme was proposed. Moreover, only one oligonucleotide sequence was used to fabricate the proposed aptasensor, endowing the sensing system with unique characteristics such as simple fabrication, timesaving operation, and easy control.

In order to validate the feasibility of the presented electrochemical aptsensing strategy, a control experiment was carried out and the results were shown in Figure 2. Target sample with a high concentration of saturated cocaine caused a striking peak current of about 150 nA (solid line) which is almost equal to the current intensity of freshly prepared sensing interface. This phenomenon confirmed the assumption that the aptamer probe deformed and the recognition site for Sal I disappeared when binding with target molecules. Meanwhile, the typical current signal of ferrocene indicated that the ferrocene moiety was insusceptible to enzyme digestion reaction and was still kept in close vicinity of electrode surface due to the target binding induced hairpin folding of aptamer probe. Oppositely, no peak current of ferrocene was obtained for blank sample (dotted line) by ac voltammogram measurements, indicating that the palindrome structure for Sal I was successfully formed and completely cutoff via enzyme digestion. As a result, a signal-on signaling mechanism with eliminated background peak current and high signal-to-noise ratio was distinctly achieved.

Zoom In Zoom Out Reset image size

Impedance spectroscopy was performed in order to monitor the sensor fabrication procedure and further check the validity of the proposed sensing strategy. Different stages of the same electrode were characterized by Faradic impedance measurements as shown in Figure 3. A very small impedance was found for the bare gold electrode (line a), accounting for excellent electrochemical conductivity of the treated electrode. Line b represented the impedance spectra of aptamer probe immobilized electrode. Obviously, self-assembly of the aptamer probes induced significant increase of electrochemical impedance. This should be attributed to the fact that the hairpin-structured aptamer probe was rigid and easy to form well-ordered monolayer on gold electrode and the negatively charged phosphate backbone hindered the electron transfer. Treatment with 6-mercaptohexanol replaced the weak adsorbed molecules and blocked the residual active sites of the gold electrode surface. As a result, a well ordered mixed self-assembled monolayer of aptamer probe and thiol molecule was achieved, which induced a additional increase of impedance (line c) since the diffusion of ferricyanide toward the electrode surface was hindered. When the aptamer probes bond with cocaine target molecules, the electrochemical impedance increased further as shown in line d. This might attributed to the fact that the hairpin-structured aptamer probe could provide sufficient interstitial space, which facilitated the binding event with small cocaine molecules and resulted in additional prevention of electroactive probe from reaching the electrode surface. However, enzyme digestion by restriction endonuclease Sal I induced an obvious decrease of electrochemical impedance as shown in line e. The measured experimental data directly verified the sensor designing principle from two points. On one hand, the unreacted hairpin aptamer probes could be specifically recognized and digested by restriction endonuclease. The negative-charged digest productions were removed from the electrode, resulting in the enhancement of electrons transfer ability and the decreases of impedance. On the other hand, part of the aptamer probes had switched structure to specifically bind with cocaine target molecules, in which the palindrome structure was deformed and cannot be cutoff by the restriction endonuclease. Therefore, the observed electrochemical impedance after digestion was a little higher than line d since the reacted aptamer probes were still anchored on the electrode surface. Obviously, these results further confirm the success of designing the proposed sensing strategy.

Zoom In Zoom Out Reset image size

Aptamer probes used in the presented sensing strategy exhibited two major functions. On one hand, the aptamer probes acted as capture probe by self-assembly of 5^' labeled thiol groups. On the other hand, the aptamer probes served as signaling probe via 3^' modified electrochemical active ferrocene. Therefore, the assembly time of aptamer probe should be investigated since it directly influences the stability and surface density of the aptamer probe. As shown in Figure 4A, the current response of electrode for different assembly periods of aptamer probe were investigated. Ten minutes of incubation time caused only a relatively low current since the aptamer probe was probably unstable under such short assembly time. As the assembly time prolonged, the peak current increased accordingly and tended to constant at about 120 min, indicating the saturated self-assembly of aptamer probes. Therefore, the incubation time of 120 min was adopted as the optimum assembly time for aptamer probes in subsequent experiments.

Zoom In Zoom Out Reset image size

The Sal I restriction endonuclease with a palindrome structured recognition site of GTCGAC is often used in Double Digests. In this study, the Sal I endonuclease was utilized for designing electrochemical aptasensor for the first time. The concentration of Sal I endonuclease should be optimized because it directly influenced the cleavage degree of the palindromic sequence. As shown in Figure 4B, a peak current of 32 nA was detected when a low concentration of 0.1 U/μL endonuclease was adopted, which was only about 20% of the original peak current for fresh prepared sensing interface. This result meant that about 80% of the signaling aptamer probes were cutoff and removed from the electrode surface, which directly demonstrated the significant digestion efficiency of the Sal I restriction endonclease. As the endonuclease concentration gradually increased, the peak current of the sensing interface declined sharply and finally turned into zero when a 0.9 U/μL of endonuclease concentration was used. The results showed that the aptamer probes self-assembled onto the electrode surface without target small molecules could be completely and efficiently cutoff by Sal I at a relatively high concentration. Therefore, the concentration of 0.9 U/μL for Sal I restriction endonuclease was selected in the following experiments.

The cleavage degree of the recognition site also related to digestion time of the Sal I restriction endonuclease. Figure 4C showed the influence of different digestion periods on the current response of the presented aptasensor. The peak current was 53 nA (about 65% loos compared with the peak current of the original sensing interface) when the enzyme digestion proceeded for only 10 min using 0.9 U/μL of Sal I restriction endonuclease, indicating that most of aptamer probes were cutoff within several minutes due to the extremely high efficiency and specificity of Sal I. With the increment of digestion time, peak current of signaling aptamer probe modified electrode decreased sharply and zero current was achieved when a period of 120 min was involved, indicating that the aptamer probes were completely digested by Sal I restriction endonuclease. To achieve a high detection capability, a digestion period of 120 min was used in the experiments.

Detecting and quantifying small target species such as drugs, hormones and the corresponding derivatives are of fundamental importance in clinical assay, forensic test and environmental analysis.²⁶ The presented work utilized Sal I restriction endonuclease for the first time to develop a background current eliminated signal-on aptasensing strategy for cocaine detection. In order to validate the utility of the proposed aptasensing method, different concentrations of target molecules were measured and the AC voltammograms were recorded as shown in Figure 5. The current intensity at 0.25 V was used to obtain the calibration curve and evaluate the analytical performances of the aptasensor. As the cocaine concentration increased, the peak current of the proposed biosensor enhanced accordingly. A representative calibration curve between the peak current intensity and logarithm of cocaine concentration was obtained with a wide linear relationship ranging from 0.8 nM to 500 nM as shown in inset of Figure 5. The calibration equation was I = 21.46 Log C + 4.97 with a correlation coefficient of 0.9935 (n = 5, R.S.D = 8.9%), where I and C represented the peak current (nA) and cocaine concentration (nM), respectively. The detection limit was estimated at 0.3 nM, at which target cocaine molecules could trigger an observable peak current slightly higher than the blank sample. Compared with previous aptasensors for cocaine analysis, the detection limit of the proposed aptasensing strategy is far lower than reported optical methods, such as strand-displacement polymerization and fluorescence resonance energy transfer based fluorescent aptasensing strategy (200 nM),²⁷ minor groove binder based energy transfer (200 nM),²⁸ DNA-Ag nanoclusters fluorescent probe based turn-on aptamer sensor (100 nM),²⁹ strand displacement amplification based label-free fluorescent aptasensor (2 nM),¹¹ and so on. Meanwhile, the achieved detection limit is also superior to most of reported electrochemical techniques, such as methylene blue labeled signal-on aptasensing strategy (10 μM),²⁴ klenow fragment polymerase reaction based electrochemical aptasensor (200 μM),³ quantum dot encoded electrochemical strategy (50 nM),³⁰ electrochemical aptasensor using quantum dots as immobilized substrate (30 nM),³¹ catalytic redox-recycling amplification based method (1 nM),³² and so on. The achievement of unexpected analytical performance might attribute to the fact that background current is essentially eliminated by restriction endonuclease mediated digestion and meanwhile a signal-on transducing mechanism is developed. Moreover, since only one oligonucleotide probe is utilized to fabricate the sensor and no extra amplification means is needed, the presented aptasensing approach is much more simple, controllable and inexpensive compared with the above-mentioned reported sensing strategies.

Zoom In Zoom Out Reset image size

A successful biosensor should not send out false positive or false negative signals in practical application, therefore high selectivity is of significant importance.³³ In the proposed work, the specificity of the biosensor was mainly determined by two key points: specific recognition ability of the aptamer for cocaine molecules and high fidelity of the Sal I restriction endonuclease for palindrome structured recognition site. As shown in Figure 6, other three drug molecules namely morphine, caffeine and theophylline were measured to inspect the selectivity of the electrochemical aptameric biosensor. The intensity of peak currents (obtained by searching manually using the software of CHI 660E electrochemical workstation since the peak currents were apparently too low to observe) upon the interferential small molecules were not more than 3% of the current intensity triggered by 500 nM target cocaine even though their concentrations were 1000-fold higher than that of cocaine. Moreover, the peak current intensity for the minimum concentration (0.8 nM) of cocaine in the linear range was also much higher than that of interferents with concentrations of 625000-fold higher, indicating a high detection selectivity was achieved in the presented aptameric sensing system.

Zoom In Zoom Out Reset image size

Good reproducibility is the important and indispensable element of a promising biosensor. Three samples of different concentrations in the linear detection range were detected to evaluate the intra- and inter- assay reproducibility of the proposed aptasensor. The intra-assay reproducibility was obtained from five repetitive measurements for each sample at the same electrode, while the inter-assay reproducibility was acquired from detections of the same sample at five different electrodes. The maximum value of the relative standard deviations was 9.8% for intra-assay and 11.5% for inter-assay. In consideration of the change of the electrode positions and/or the difference of the surface areas from electrode to electrode, the proposed sensing system could offer an acceptable reproducibility for cocaine detection.

In order to evaluate the applicability and reliability of the presented sensing system, the recovery experiments in 10 fold dilution of human serum samples with different concentrations of cocaine in the linear range were performed. All the measurements were implemented for three times, and the results were recorded in Table I. Recovery in the range of 96–105% with a average relative standard deviation of 9.2% was achieved, indicating that the proposed aptasensor was appropriate for cocaine detection.