Fabrication of a Deoxyribonuclease Sensor Based on the Electrical Characteristics of DNA Molecules.

In this study, we developed a sensing device that can detect deoxyribonuclease (DNase) based on the electrical properties of deoxyribonucleic acid (DNA). We estimated the equivalent circuit between the electrodes with immobilized DNA and investigated whether the characteristics of the electrodes change before and after the DNase reaction. This method detects DNase by simply evaluating the electrical properties of DNA without using a fluorescent reagent. Therefore, inexpensive and highly accurate measurements can be performed with simple operations. However, detection sensitivity must be increased for practical feasibility. Hence, we investigated whether DNA immobilization is restricted by changing the shape of the electrode to a triangle with sharp edges, which may improve the sensitivity of DNase. Additionally, we attempted to detect DNase from an extremely small amount of sample solution using a microchannel. The device was able to quantitatively analyze DNase I activity with a detection limit of 5.5 × 10-5 unit/μL. The results demonstrate the effectiveness of the proposed sensing device for various medical applications.

Introduction

Deoxyribonucleic acid (DNA) exhibits excellent molecular recognition and self-organizing functions, which make it a viable option for constructing various nanostructures.[1,2] For instance, DNA origami is a nanostructure created by weaving long single-stranded DNA (ssDNA) and short ssDNA via self-assembly.[3−6] Additionally, DNA has an electrical conductivity that relies on the base sequence and environmental conditions.[7−11] It can also bind to specific biomolecules spontaneously.[12,13] These characteristics enable DNA to serve as an effective sensing material in the medical field.

Deoxyribonuclease (DNase) is an enzyme that nonspecifically hydrolyzes the phosphodiester bonds of DNA. It exists in the cells and tissues of living organisms; in body fluids, including blood and urine; on the human skin surface; in the atmosphere; and in tap water.[14,15] The DNase concentration in blood is being increasingly investigated as a diagnostic marker owing to its active involvement in specific diseases, such as myocardial infarction[16,17] and systemic lupus erythematosus.[18,19] In other words, the development of DNase measurement sensors can facilitate the early diagnosis of these diseases. Additionally, studies on DNA, such as genetic manipulation experiments, have demonstrated that DNase contamination causes major problems. Therefore, a small sensor for DNase detection can significantly improve the reliability of DNA-based experiments. One of the conventional methods for the detection of DNase is the single radial-enzyme-diffusion method,[17,20] which uses a fluorescent reagent and employs fluorescence resonance energy transfer.[21] Although these methods are highly sensitive, they present certain disadvantages, such as a long duration from preparation to detection and a relatively high cost of the fluorogenic oligonucleotide. Therefore, an inexpensive and highly accurate DNase measurement sensor with simpler operations is required. DNase can be electrochemically detected using ferrocenyloligonucleotide (FcODN)-immobilized electrodes. The FcODN is immobilized on the electrode through Au-S linkages.[22] DNase I was detected by measuring the oxidation current with a detection limit of 1 × 10–4 unit/μL. A highly effective sensing method was recently developed to detect endonuclease activity using copper nanoparticles as the electrochemical reporters. This sensor had a wide linear range of 10–3 to 10–1 unit/mL and a limit of detection of 10–3 unit/mL.[23]

In this study, we developed a highly accurate sensing device that can detect DNase in small volumes using the electrical properties of DNA. As DNA molecules typically form a complex random coil in solutions, the molecules must be stretched into a linear shape and immobilized on a substrate for sensing applications.[24,25] Therefore, we used an electrostatic orientation to immobilize and stretch the DNA between the two microelectrodes. The behavior of DNA can be manipulated using the electric-field strength and applied frequency based on the electrostatic orientation.[10,26,27] Furthermore, DNA can be stretched and immobilized at arbitrary positions depending on the shape and arrangement of the electrodes. In a previous study, we investigated the electrical properties of lambda phage DNA (λDNA) molecules, which were manipulated by electrostatic orientation, based on their current–voltage characteristics and complex impedance plots.[28] Additionally, we developed a DNase-detecting device using multiple λDNA nanowires and performed several experiments using rod-shaped parallel electrodes.[29] The device comprised a PDMS reservoir with a volume of 400 μL and two aluminum electrodes. This method enabled a simple and highly reproducible detection of DNase without DNA pretreatment or electrode surface modification. Furthermore, this DNase detection was simpler and faster than that presented by conventional fluorometric detection methods. Furthermore, the conductivity of DNA and the properties of the triangular electrodes were determined through the experimental analysis of DNase detection using the triangular electrodes.[30] In this study, we developed a simple method based on a microfluidic channel that can be precisely handled with extremely small volumes. The microchannel allows sample small volumes to be filled between electrodes without requiring micropipettes or other tools to control the volume of the sample. Therefore, the dead volume of our device is extremely small, which avoids wastage of the sample and facilitates the conduction of experiments with minimal sample volume. We also estimated the equivalent circuit between the triangular electrodes with multiple λDNA molecules using the electrochemical impedance spectroscopy[31−33] and investigated whether the properties between the electrodes change before and after the DNase treatment. As triangular electrodes facilitate the immobilization of DNA molecules in a specific location, they are expected to exhibit more noticeable impedance changes than those exhibited by parallel electrodes. The experimental results indicate that the detection sensitivity of DNase can be improved, owing to the limited immobilization site of DNA molecules.

Materials and Methods

Device for Electrostatic Orientation

Figure a depicts a schematic of the device used to detect DNase. This device comprises an aluminum thin-film electrode fabricated on a glass substrate and a microchannel (50 μm in depth, 500 μm in width, and 10 mm in length)[28] composed of polydimethylsiloxane (PDMS, KE-106, Shin-Etsu Chemical Co., Ltd.). Initially, an aluminum film (thickness of 200 nm) was evaporated on a 30 mm square glass substrate using a vacuum evaporation system. Subsequently, a photoresist was applied, and ultraviolet exposure and development were performed. The electrode pattern was then formed via wet etching. To limit the locations for immobilizing the DNA molecules, the edges of the electrodes were sharpened. In this experiment, the electrode gap was set to 14 μm, which was slightly shorter than the length of DNA (16 μm).

Figure 1

Schematic of the microfluidic device used for deoxyribonucleic acid (DNA) stretching and immobilization. (a) Bird’s-eye view of the device before mounting a polydimethylsiloxane (PDMS) microchannel on a glass plate with two electrodes. The microchannel is 50 μm deep, 500 μm wide, and 10 mm long. (b) Top view of the electrodes and microchannel. Two triangular electrodes were placed 14 μm apart such that the electrode gap was slightly shorter than the length of λDNA, which was 16 μm.

DNA Immobilization and Electrical Characterization

The DNA specimens used in this study were the λDNA extracted from Escherichia coli and procured from Takara Bio Inc. λDNA is a double-stranded DNA helix that comprises 48,502 base pairs (length 16 μm). A DNA solution (1 nmol/L) was introduced into the microchannel with a volume of less than 10 μL, and an alternating current (AC) voltage was applied between the triangular electrodes to immobilize the λDNA molecules. The formation of the AC electric field immobilized the λDNA molecules on the sharp edges of the electrode,[10] and their electrical properties were evaluated (Figure b). An impedance analyzer (HIOKI, IM3570) was used for the evaluation. The frequency was varied between 500 Hz and 5 MHz, and the impedance, Z, and phase, θ, between the electrodes were measured at different frequencies in steps. The impedance was divided into real and imaginary parts based on these measured values, and a complex impedance plot of the DNA immobilized between the electrodes was generated. Subsequently, the equivalent circuit between the electrodes was estimated using the shape of the obtained complex impedance plot. The impedance of λDNA was determined by calculating the appropriate values for each element of the equivalent circuit.

Optimization of Flow Rate

Typically, the DNA molecules may peel off from the electrodes when a solution is introduced through a microchannel at an extremely high speed, regardless of the presence or absence of DNase. Therefore, we attempted to optimize the flow rate of the solution using the following procedure.

Initially, a λDNA solution was introduced into a PDMS microchannel, and an AC voltage of 1 MHz and 20 Vp-p were applied between the two electrodes using a function generator (TEXIO FGX-295). The AC voltage was applied for 15 min to stretch and immobilize the λDNA molecules between the electrodes.

Ultrapure water was then introduced into the microchannel using an electric syringe pump (Fusion Touch 200, CX07200) to remove the scattered λDNA molecules, which were not immobilized from the microchannel. Subsequently, the AC voltage application was terminated.

An impedance analyzer (HIOKI, IM3570) was used to measure the impedance, Z, between the two electrodes and the phase difference, θ, between the voltage and current at each frequency. The measurements were performed at frequencies ranging from 500 Hz to 5 MHz, with an applied potential of 100 mV.

Ultrapure water was introduced again into the microchannel at the same flow rate as in Step (2), and Z and θ were measured six times for every 10 min in 1 h from the beginning of immobilization.

The impedance value of the immobilized λDNA molecules was estimated by creating a complex impedance plot with the Z and θ values measured from the initiation of DNA immobilization up to 60 min after the washing with ultrapure water. The frequency of the applied voltage was used as a parameter.

The impedance change of DNA was evaluated corresponding to each flow rate, and the decrease rate of conductance was graphically presented to verify the exfoliation of DNA caused by the solution operation.

DNase Detection

When DNase was introduced between the electrodes after immobilizing DNA, λDNA molecules were cleaved by an enzymatic reaction (Figure ). This reaction changes the impedance between the electrodes, and the extent of change depends on the concentration of DNase. In this study, we attempted to detect DNase based on the electrical characterization between the electrodes before and after the enzymatic reaction using the following procedure. The concentration of the DNase solution was adjusted using an appropriate buffer because DNase I requires Ca2+ and Mg2+ to hydrolyze double-stranded DNA.

Figure 2

Principle of deoxyribonuclease (DNase) detection using immobilized lambda phage DNA (λDNA) molecules. (a) The λDNA molecules are stretched and immobilized between two electrodes via electrostatic orientations. (b) After DNase treatment, λDNA molecules are cleaved by the enzymatic reaction and removed from the electrodes.

A λDNA solution was introduced into a PDMS microchannel, and an AC voltage was applied for 30 min to stretch and immobilize the λDNA molecules between triangular electrodes with 12 gaps.

The DNA molecules were stretched and immobilized between the triangular electrodes with 12 gaps using an electrostatic orientation.

The impedance, Z, and phase difference, θ, were measured, and a complex impedance plot was generated. After estimating the equivalent circuit, the resistance component of λDNA molecules before the introduction of DNase (Rbefore) was calculated.

A solution of DNase (Recombinant DNase I, Takara Bio Inc.) with a concentration in the range of 10–5 to 10–1 unit/μL was introduced into the microchannel at an optimized flow rate. After the microchannel was filled with the DNase solution, the device was placed in an incubator maintained at 37 °C for 30 min.

After removing the device from the incubator, the microchannel was washed with a solution of ultrapure water for 1 h.

The Z and θ values between the electrodes were then measured once again. The resistance value of the equivalent circuit after the DNase treatment (Rafter) was determined, and the resistance values before and after the DNase treatment were compared.

Results and Discussion

In a previous study, we evaluated the electrical properties before and after DNA immobilization. We observed that the complex impedance plot before DNA immobilization consisted of a circular arc, while that after DNA immobilization consisted of two semicircles, with a semicircle in the high-frequency range and a circular arc in the low-frequency range.[28] In this experiment, we focused on the diameter of the semicircle that appeared in the high-frequency range (the real part of the impedance, i.e., the resistance component). This was used to determine any changes in the number of DNA immobilized between the electrodes. Figure depicts the impedance change between the electrodes obtained when DNA was stretched and immobilized between the 12 sharp electrodes and ultrapure water was supplied continuously at a flow rate of 1.0 μL/min. This indicates that the size of the semicircle that appears in the high-frequency range increases as a function of the liquid transfer time. In the real part, the resistance value was 760 kΩ immediately after DNA immobilization, 890 kΩ 30 min after feeding, and 1.7 MΩ after 1 h. The increase in the resistance value from 30 min to 1 h indicates that the immobilized DNA molecules were unable to maintain their state and gradually peeled off between the electrodes at a flow rate of 1.0 μL/min.

Figure 3

Example of a complex impedance plot between two electrodes from a validation experiment of flow rate (flow rate = 1.0 μL/min). The red circles represent the complex impedance plot obtained from the immobilized λDNA molecules before supplying the solution. The blue squares represent the complex impedance plot obtained after 30 min of pumping. The black squares represent the complex impedance plot obtained after 60 min of pumping. The diameter of the semicircle observed in the high-frequency range increased with each pumping, indicating the detachment of λDNA from the electrode.

Furthermore, the flow rate was changed from 0.3 to 5.0 μL/min, resistance component at the start of liquid feeding (R0) changed after n min of liquid feeding (R), and increase ratio was determined to be R/R0. Figure depicts the relationship between the solution-sending time and impedance-increase ratio for each flow rate. When the flow rate was 5.0 μL/min, a large increase was observed in the resistance value of approximately two times 10 min after the onset of pumping, indicating that the DNA molecules were detached. Additionally, the resistance value increased slightly when the liquid feeding operation was performed for 1 h even at a flow rate of 0.5 μL/min. Conversely, the resistance remained unchanged at a flow rate of 0.3 μL/min. This analysis indicated that 0.3 μL/min was the optimal flow rate for operating the proposed device.

Figure 4

Relationship between the time the solution was sent and the increase in the impedance ratio. The increase ratio was calculated as R/R0 using the resistance component before pumping, R0, and the resistance after n minutes of pumping, R. The impedance increased by a factor of two within 10 min of pumping at flow rates of 5 and 10 μL/min. The impedance increased by a factor of two when the pumping operation was performed for over 60 min even at a flow rate of 1.0 μL/min. The increase ratio remained at one for the flow rate of 0.3 μL/min, even after 60 min of pumping, indicating that the impedance remained unchanged.

Detection of DNase

Following the immobilization of the DNA molecules between 12 sharp electrodes, a DNase solution was introduced at a flow rate of 0.3 μL/min. Typically, when the immobilized DNA molecules were cleaved by the enzymatic reaction, the impedance between the electrodes changes according to the number of cleavages. Therefore, we attempted to quantify the DNase concentration by analyzing the characteristics of the electrodes before and after the enzymatic reaction.

To determine the stretch/immobilization of DNA between the electrodes, we observed the λDNA molecules that were preliminarily labeled with YOYO-1. Figure a presents a fluorescence image of the λDNA stretched and immobilized between the electrodes in the microchannel. The stretched λDNA was cleaved by the enzymatic reaction after the DNase treatment, and a fluorescent image of the curled molecules was obtained at the tip of the electrode, as shown in Figure b. Figure c illustrates the complex impedance plot of the immobilized λDNA molecules before and after the DNase treatment. The red circles represent the complex impedance obtained before the DNase treatment (after λDNA immobilization) within the frequency range of 500 Hz to 5 MHz. A small semicircle was observed in the high-frequency range, and the circular shape of the plot indicates that the immobilized λDNA molecules comprise resistive and capacitive components. Therefore, the equivalent circuit includes a parallel resistance–capacitance circuit, as shown in Figure d. We then used the constant phase element (CPE), ZCPE, which is a circuit element, to represent non-ideal capacitive interfacial behaviors.[34−36] The CPE impedance, ZCPE, can be given aswhere j denotes an imaginary unit, ω denotes the angular frequency (rad/s), X represents the CPE coefficient (Fs), and p represents the CPE exponent that takes a value between 0 and 1. The resistive component, Rbefore, which corresponds to the diameter of the semicircle at higher frequencies, presumably represents the resistance of the λDNA molecules. The blue square markers in Figure c represent the complex impedances obtained after the DNase treatment. When the DNase solution was introduced at a concentration of 10–4 unit/μL, the complex impedance changed, and the semicircle increased slightly. Comparing the estimated resistance values of λDNA, the value was determined to be 400 kΩ after λDNA immobilization and changed to approximately 1.6 MΩ after introducing the DNase solution. The dashed and dotted black lines in Figure c represent the fitted curves obtained using the R-CPE circuit model and the parameter values listed in Table . The value increases by four times when expressed as an increase ratio (Rafter/Rbefore). This increase can be attributed to the cleavage of the λDNA molecules immobilized between the electrodes by DNase, since the liquid transfer operation does not present a significant effect. In other words, DNase can be detected by measuring the impedance using a microfluidic device.

Figure 5

Fluorescence image of λDNA stretched and immobilized between two electrodes (a) before and (b) after DNase treatment. The λDNA was labeled with the fluorescence dye, YOYO-1, prior to alignment. (c) Complex impedance plots and (d, e) equivalent circuits of the DNase treatment. Red circles represent the complex impedance plot obtained from the immobilized λDNA molecules before the DNase treatment. The equivalent circuit after λDNA immobilization is assumed to be a series connection of two parallel R-CPE circuits. The resistance value of λDNA (Rbefore) was estimated to be 400 kΩ. Blue squares represent the complex impedance plot obtained after the DNase treatment with 10–4 unit/μL, which increased the resistance value of DNA (Rafter) to 1.6 MΩ.

Table I

Parameter Values of the R-Constant Phase Element (CPE) Circuit Model that Were Used for Curve-Fittinga

label	R1 (kΩ)	X1 (Fsp-1)	p1	R2 (MΩ)	X2 (Fsp-1)	p2
before treatment	400	1.0 × 10–11	0.9	70	4.0 × 10–9	0.75
after treatment	1600	3.5 × 10–12	0.98	80	1.5 × 10–9	0.8

X represents the CPE coefficient, and p represents the CPE exponent (0 > p > 1). When p = 0 and p = 1, the corresponding CPEs are equivalent to the ideal resistance and ideal capacitance, respectively.

We were able to successfully obtain the increased impedance ratio, which was dependent on the DNase concentration (Figure a), when DNase solutions of various concentrations were introduced. Their definite correlations in the range of 10–5 to 10–1 units/μL of DNase concentration demonstrated that the device can be used to quantitatively analyze DNase. The limit of detection for DNase I was expected to be 5.5 × 10–5 unit/μL based on the correlations obtained for DNase I concentrations in the range of 10–5 to 10–4 unit/μL. It was demonstrated that detection in the 10–5 to 10–4 unit/μL range is possible for samples with very small volumes (<10 μL). Furthermore, the increased ratio of the resistance components of the parallel and triangular electrodes were compared to determine the detection sensitivity based on the electrode shape. The results indicate that the triangular electrode generates a higher rate of increase in all DNase concentrations and improves the measurement sensitivity. This can be attributed to the smaller number of immobilized λDNA molecules observed when the triangular electrode was used (Figure b). The corresponding impedance change associated with the DNA cleavage was larger. In future experiments, the impedance change before and after serum introduction must be analyzed using serum from venous blood separated by centrifugation based on previously reported demonstration experiments.[37] Furthermore, the electrical properties of the double-stranded M13mp18 DNA (which contains 7249 base pairs) have already been analyzed and can be simulated using an equivalent circuit model similar to that of λDNA.[38] Comparison experiments must be performed between M13mp18 DNA and λDNA to determine the detection limits of the DNA probes.

Figure 6

(a) Relationship between the deoxyribonuclease (DNase) concentration and increased impedance ratio. Red circles represent the relationship between the DNase concentration and increase ratio obtained from the triangular electrodes. The open circle on the ordinate represents the increase ratio in the absence of DNase I. The blue triangle represents the increase ratio obtained from the parallel electrodes. (b) Comparative images of λDNA stretched and immobilized between the triangular electrodes and the parallel electrodes. The parallel electrodes could not detect impedance changes at DNase concentrations smaller than the 10–4 unit/μL level. When the triangular electrodes were used, an increased ratio by a factor of two times approximately was obtained even at the 10–5 unit/μL level. The standard deviation for the increased impedance ratio of the λDNA molecules was less than 20%, indicating that the immobilized λDNA molecules can be used for DNase detection.

Conclusions

In this study, we fabricated a microdevice for detecting DNase by immobilizing λDNA molecules between two triangular electrodes. When the concentration of DNase was 10–4 unit/μL, an impedance-increase rate of approximately four times was obtained after the DNase treatment. Furthermore, a definite correlation between the DNase concentration and increased impedance ratio was obtained. This result can be applied to the quantitative analyses of DNase. Moreover, the difference in the increased impedance ratio was investigated based on the shape of the electrode. We observed that the detection sensitivity of DNase improved when the immobilization site of DNA was restricted. In future, the proposed device can be used as a simple DNase sensing kit prior to genetic research experiments or can be mounted on ultrapure water production systems for genetic research.