In this work we report the development of an electrochemical DNA biosensor with high sensitivity for mercury ion detection. A new matrix based on gold nanoparticles (AuNPs)-glutathione (GSH)/cysteine was investigated. The interaction between DNA oligonucleotides and Hg2+ ions followed by the formation of Thymine–Hg2+–Thymine (T–Hg2+–T) structures was quantified using different electrochemical methods. It has been shown that the electrochemical impedance spectroscopy (EIS) measurements and the differential pulse voltammetry (DPV) confirmed the specific interaction between the oligonucleotide receptor layer and the Hg2+ ions. Besides, the developed sensor exhibited high sensitivity towards mercury among some examined metal ions such as Pb2+, Cu2+ and Cd2+. As a result, a high electrochemical response and low detection limit of 50 pM were estimated in the case of Hg2+ ions. The developed DNA biosensor was applied successfully to the determination of Hg2+ions in wastewater samples.

INTRODUCTION

The development of electrochemical sensors for the recognition of heavy metal ions is very important due to their fundamental role in biological, environmental and chemical processes (Prodi et al. 2000). Therefore, many sensor types have been used for environmental analysis, industrial quality control and clinical diagnostics (Andreescu & Sadik 2004). Among various types of sensors, electrochemical receptor layers, such as enzymes, nucleic acids, and antibodies, came to a special prominence. Recently, the interactions of heavy metal ions, in particular mercury, with nucleic acids have been studied, due to the possible toxicity and cancerogenicity of these ions (Anastassopoulou 2003). Furthermore, the mercury ion is a highly toxic environmental pollutant and has posed a serious threat to human health (Bolger & Schwetz 2002). Mercury can also affect many different areas of the brain and their associated functions, resulting in symptoms such as tremors, vision problems, deafness and loss of muscle coordination, sensation and memory (Stern 2005). In addition to the brain, inorganic mercury can damage the heart, kidney, stomach, and intestines (Zheng et al. 2003; Mutter et al. 2005; Wojcik et al. 2006). Due to their serious harm, the binding of metal ions to nucleic acids is used for the construction of biosensors for the detection of heavy metal ions (Liu et al. 2009). Recently, it has been reported that there is specific and strong coordination between Hg2+and the two DNA thymine bases (T) to form a mediated base pair (T–Hg2+–T) (Ono & Togashi 2004). Classical methods for mercury detection, such as atomic absorption spectroscopy (Li et al. 2006), colorimetric (Lin et al. 2010) and fluorescence (Chiang et al. 2008) are widely used. However, electrochemical methods have received particular attention due to their high sensitivity and selectivity, such as differential pulse stripping analysis (Yantasee et al. 2003; Wang et al. 2007), electrochemical impedance spectroscopy (EIS) (Lin et al. 2011) and square wave voltammetry (Jiang et al. 2015). In this study, we describe an electrochemical biosensor with high sensitivity and selectivity for Hg2+ detection based on DNA oligonucleotides. This probe report based on a DNA-AuNPs-glutathione/cysteine/Au modified electrode to capture mercury (II) ions. First, the interaction between thymine and mercury ions was evaluated by the EIS. Then, the electrochemical reduction of the surface provides a readout signal for the quantitative detection of Hg2+. We also demonstrate that the sensitivity of this Hg2+ sensor could be significantly improved with the DNA oligonucleotides immobilized by using gold nanoparticles (AuNPs).

EXPERIMENTAL

Reagents

Cysteine (95%), glutaraldehyde (Glu, 25%), glutathione (GSH reduced form), hexacyanoferrate (II/III), gold colloid solution (20 nm), N-hydroxysuccinimide (NHS), phosphate buffered saline (PBS), trishydroxymethylaminomethane and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) were purchased from Sigma Aldrich (St. Louis, USA). CuCl2.2H2O, 3CdSO4.8H2O, Pb(NO3) and HgCl2 were purchased from Fluka-Chemika (Buchs, Switzerland). The DNA oligonucleotide was purchased from Bioneer Oligo Synthesis Report (Bioneer, BPS Bioscience, Biotools – Tunisia). The sequence is 5′-NH2-(CH2)6-ATTTGTTCATGCCT-3′. All electrolytic solutions were prepared using ultra-pure water.

Apparatus

The electrochemical measurements were performed using an Autolab (PGSTAT 302 N, Eco Chemie). The measurements were made using a conventional three electrode electrochemical system consisting of a gold electrode, a platinum wire auxiliary electrode and Ag/AgCl/KCl reference electrode. The geometrical area of the working electrode was 0.031 cm2. Impedance measurements were performed in the frequency range from 0.1 to 100,000 Hz. All electrochemical measurements were performed in a Faraday cage at room temperature (25 °C) to avoid any stray light or electrical perturbation from external sources.

Working electrode pretreatment and SAM formation

To obtain a clean, activated and reproducible electrochemical surface, the gold electrode was mechanically polished with alumina slurry followed by rinsing with distilled water and sonication in acetone for 2 min. After mechanical cleaning, the gold electrode was treated using a chemical process by immersion in a piranha solution (H2SO4/H2O2, 1:3 v/v) for 1 min and rinsed with ultrapure water. Afterwards, the clean gold electrode was immersed into 5 mM cysteine in 0.1 M PBS solution at pH 7.4 for 2 h. Then, the modified electrode was rinsed with ultrapure water to remove non-covalent attached cysteine molecules.

Immobilization of the probe DNA on the Au/cysteine modified electrode

The Au-cysteine modified electrode was activated in an atmosphere saturated with glutaraldehyde vapor for 1 h. The DNA probe was immobilized on the Au/cysteine modified electrode via two immobilization strategies:

  1. In the first case, the DNA senor was fabricated via an adsorption process. The Au/Cysteine electrode was immersed in DNA probe solution for 1 h. Then, the modified electrode surface was washed with PBS solution for desorbing the non-attached DNA probes. Figure 1(a) shows the procedure of the preparation of the DNA/cysteine/Au modified electrode.

  2. In the second case, the Au/Cysteine electrode was immersed for 1 h into a mixture of glutathione-AuNPs (1:1) solution previously stored for 16 h in a refrigerator. Then, the modified electrode was activated for 1 h in 1:1 (v/v) EDC/NHS mixture (10 mM EDC and 10 mM NHS, pH 5). Finally, the modified electrode was immersed in the DNA probe solution for 1 h and then washed with PBS solution to remove the unbound DNA. Figure 1(b) shows the preparation steps of the DNA biosensor based on the Au nanoparticles.

Figure 1

The construction steps and design mechanism of the electrochemical sensor based on (a) Au/cysteine/DNA and (b) Au/cysteine/glutathione-Au NPs/DNA electrodes.

Figure 1

The construction steps and design mechanism of the electrochemical sensor based on (a) Au/cysteine/DNA and (b) Au/cysteine/glutathione-Au NPs/DNA electrodes.

RESULTS AND DISCUSSION

Electrochemical characterization of the modified electrode

The immobilization of DNA on Au/cysteine using two approaches was characterized by cyclic voltammetry and impedance spectroscopy measurements in 0.1 M PBS solution containing 5 mM of the [Fe(CN)6]3-/4- as redox agent at the scan rate of 50 mVs−1. Figure 2(a) shows the CVs behavior of the bare and the modified gold electrode. The self-assembly of the cysteine monolayer on the electrode surface induces an increase of the peak current, explaining the diffusion controlled process for the [Fe(CN)6]3-/4- redox agent. The significant change in current after DNA immobilization explains that the DNA probe has been successfully immobilized on the modified electrode surface. Thus, the decrease in current can be assigned to the electrostatic repulsion interaction between [Fe(CN)6]3-/4-and the negative charge phosphate backbone of the DNA probe. Furthermore, the incorporation of the gold nanoparticles in the sensing matrix enhanced the electrochemical performances of the DNA sensor by increasing the effective electrode surface area. This result improves the rate of electron transfer reaction, which was evidenced by an increase in the voltammetric responses of the [Fe(CN)6]3-/4- agent in comparison with the probe/cysteine modified electrode (Du et al. 2009). In addition, the EIS method was used to explain and gain more information about the modified solid-liquid interface. Figure 2(b) shows the Nyquist plots, which represent the control of the charge transfer during the modification of the electrode surface. The lower semicircle diameter was obtained in the case of the DNA/GSH-Au NPS/cysteine modified electrode in comparison with the DNA/cysteine/Au modified electrode. This result demonstrates that the charge transfer resistance (Rct) decreases in the presence of the gold nanoparticles, which enhances the specific surface and incidentally improves the electron transfer mechanism.

Figure 2

(a) Cyclic voltammograms of Au bare electrode, Au/cysteine, Au/cysteine/DNA, Au/cysteine/GSH-Au NPs/DNA modified electrodes (scan rate: 50 mV/s) and (b) Nyquist plots for bare gold electrode, Au/cysteine, Au/cysteine/DNA and Au/cysteine/GSH-Au NPs/DNA modified electrode in 5 mM of ferri/ferrocyanide redox couple.

Figure 2

(a) Cyclic voltammograms of Au bare electrode, Au/cysteine, Au/cysteine/DNA, Au/cysteine/GSH-Au NPs/DNA modified electrodes (scan rate: 50 mV/s) and (b) Nyquist plots for bare gold electrode, Au/cysteine, Au/cysteine/DNA and Au/cysteine/GSH-Au NPs/DNA modified electrode in 5 mM of ferri/ferrocyanide redox couple.

Electrochemical sensing of Hg2+ions

The electrochemical detection of Hg2+ ions was performed by incubating the DNA modified electrode in different concentrations of Hg2+ ions dissolved in Tris-HCl (50 mmol.L–1) for 1 h. The pH of the mercury solutions was adjusted to 6 in order to inhibit the formation of HgO species. After the immersion step, the modified electrode was carefully washed with Tris-HCl and the electrochemical measurements were conducted. In the presence of Hg2+ ions, the specific coordination between Hg2+ and thymine bases can change parallel DNA from linear to hairpin structures. This structure enhances the steric and coulombic interaction between adjacent DNA sequences (Liu et al. 2008). The electrochemical response of the modified electrode towards Hg2+ ions was evaluated using the differential pulse voltammetry (DPV) method. Thus, the reduction of surface confined Hg2+ ions was investigated. Figure 3 shows the voltammograms corresponding to the DNA/cysteine modified electrode upon incubating with various concentrations of Hg2+ from 1 nM to 0.1 μM. As can be seen in Figure 3, a reduction peak at 0.55 V (vs Ag/AgCl) was obtained, which can be attributed to the reduction potential of the surface confined Hg2+ (Zhu et al. 2009). Consequently, we observe that the peak reduction currents are intensified with an increase of the Hg2+ concentration.

Figure 3

DPV spectra of cysteine/DNA modified electrode for different concentrations of Hg2+.

Figure 3

DPV spectra of cysteine/DNA modified electrode for different concentrations of Hg2+.

Electrochemical sensing for Mercury (II) ions using gold nanoparticles

The electroanalytical performances of the T–Hg2+–T based on biosensor were evaluated for different Hg2+ion concentrations using the DNA/GSH-Au NPS/cysteine modified electrode. Figure 4(a) represents the DPV voltammograms obtained before and after incubation of the DNA biosensor in Hg2+ solution in which the concentration varies from 50 pM to 1 μM. By plotting the concentration of Hg2+ versus the peak current, the analytical calibration curves are shown in Figure 4(b). The DNA/GSH-Au NPS/cysteine modified electrode response exhibit a good linear relationship with the Hg2+ concentration in the range of 50 pM to 0.1 μM, with a sensitivity of 0.143 μM/pHg and a detection limit around 50 pM. On the other hand, the DNA/cysteine modified electrode demonstrates a linear relationship in the range of 1 nM to 0.1 μM and a sensitivity of 0.053 μA/pHg. This comparison indicates that the modified electrode using the gold nanoparticles presents a higher sensitivity and a lower detection limit towards Hg2+ ions.

Figure 4

(a) Differential pulse voltammograms of cysteine/GSH-Au NPs/DNA modified electrode for different concentration of Hg2+. (b) Calibration curves for (a) cysteine/DNA and (b) cysteine/GSH-Au NPs/DNA modified electrode.

Figure 4

(a) Differential pulse voltammograms of cysteine/GSH-Au NPs/DNA modified electrode for different concentration of Hg2+. (b) Calibration curves for (a) cysteine/DNA and (b) cysteine/GSH-Au NPs/DNA modified electrode.

The value of 50 pM is below the maximum contaminant level for inorganic mercury in drinking water set by US EPA (0.002 mgL–1, ca. 10 nmol L–1) (US Environmental Protection Agency 2009).

In order to confirm the interaction between DNA thymine and mercury ion in the presence of gold nanoparticles, electrochemical impedance measurements were investigated. As shown in Figure 5(a), the charge transfer resistance (Rct), which corresponds to the diameter of the semicircle in Nyquist plot, increases gradually when the mercury ions' concentration increases. Higher concentrations of Hg2+ ions result in higher charge transfer resistance (Rct) of the negative charge redox indicator, [Fe(CN)6]3-/4-. This result can be attributed to the thymine–Hg2+–thymine complexes formation, which inhibits the redox activity of the electroactive couple and limits access to the electrode surface. In addition, the impedance values are fitted to a standard Randle's equivalent circuit as shown in Figure 5(b). The components in the equivalent circuit included the solution resistance Rs, the charge transfer resistance Rct, the constant phase element related to double layer capacitance (CPE) and the Warburg impedance (W). As summarized in Table 1, the charge transfer resistance values (Rct) of DNA-modified electrodes before and after incubation in Hg2+ solution was obtained from the numeric simulation of the impedance plots. As a result, the addition of 1 μM of Hg2+ to the sensor leads to an apparent increase in the charge transfer resistance from 2 to 11.6 KΩ. Moreover, the Warburg impedance and the double layer capacitance decreased slightly with the increase of mercury concentration. By plotting the charge transfer resistance versus the logarithm value of Hg2+ concentration, a calibration curve was obtained (Figure 5(b)). The Rct values exhibit a linear relationship with the Hg2+ concentration in the range of 50 pM to 1 μM, with a sensitivity of 2.609 kΩ/pHg. From the Rct values, the apparent electrode coverage (θ) of the mercury sensor can be approximately calculated according to Equation (1) (Troughton et al. 1982): 
formula
1
where Rct and Rct′ are the charge transfer resistance of the bare and the AuNPs -GSH/cysteine modified electrodes, respectively. It can also be seen that the coverage surface increases considerably with the Hg2+ concentration. This result confirms the wide linear biosensor response range.
Table 1

Impedance fitted parameters of DNA/AuNPs-glutathione/cysteine modified electrode after immersion in different concentrations of Hg2+, obtained from the analysis of impedance data with the Randles circuit in Figure 5(b)

[Hg2+](M) Rs (Ω) CPE (μF) Rtc (kΩ) W(μF) Ѳ(%) 
 136.3 1.876 0.875 1.033 625 
5 × 10–11 140.87 1.826 0.861 2.023 620 0.489 
10–10 26.64 1.634 0.891 2.816 563 0.633 
5 × 10–10 40.61 1.28 0.891 4.754 410 0.782 
10–9 35.75 1.174 0.891 6.832 314 0.848 
10–8 38.06 1.098 0.889 8.483 206 0.878 
10–7 36.62 1.038 0.887 10.475 136 0.901 
10–6 35.62 0.952 0.89 11.657 116 0.911 
[Hg2+](M) Rs (Ω) CPE (μF) Rtc (kΩ) W(μF) Ѳ(%) 
 136.3 1.876 0.875 1.033 625 
5 × 10–11 140.87 1.826 0.861 2.023 620 0.489 
10–10 26.64 1.634 0.891 2.816 563 0.633 
5 × 10–10 40.61 1.28 0.891 4.754 410 0.782 
10–9 35.75 1.174 0.891 6.832 314 0.848 
10–8 38.06 1.098 0.889 8.483 206 0.878 
10–7 36.62 1.038 0.887 10.475 136 0.901 
10–6 35.62 0.952 0.89 11.657 116 0.911 
Figure 5

(a) Nyquist spectra of the cysteine/GSH-AuNPs/DNA modified electrode for different Hg2+ concentrations. (b) Calibration curves for cysteine/GSH-AuNPs/DNA modified electrode and the corresponding equivalent circuit.

Figure 5

(a) Nyquist spectra of the cysteine/GSH-AuNPs/DNA modified electrode for different Hg2+ concentrations. (b) Calibration curves for cysteine/GSH-AuNPs/DNA modified electrode and the corresponding equivalent circuit.

Selectivity of the electrochemical DNA biosensors for Hg2+ detection

The sensitivity of the developed DNA biosensors towards other heavy metal ions such as Cu2+, Pb2+ and Cd2+ was investigated. As shown in Figure 6(a), the higher charge transfer resistance variation was obtained for the mercury ions, whereas a slight variation of Rct was observed for the other tested metal ions. This result indicates that the electrochemical DNA biosensor has high sensitivity and more selectivity towards Hg2+ ions. It can be attributed to the specific coordination between thymine bases and Hg2+ ions. Moreover, Figure 6(b) illustrates the electrochemical response for Hg2+ ions with and without a mixture of three different interfering metal ions, such as Cu2+, Cd2+and Pb2+, at a concentration of 1 μM. As shown in Figure 6(b), the current response for the Hg2+ ions, peak decreased when the modified electrode was immersed in the mixed ions solution. This decrease can be attributed to the ion interference effect. In a previous work it was demonstrated that the Pb2+ ion exhibits a significant affinity towards guanine by the formation of a G-quadruplex structure (Jarczewska et al. 2015). For this reason, the interference of Pb2+ ions cannot be excluded, as they interact specifically with guanine.

Figure 6

(a) Nyquist plot of the cysteine/GSH-AuNPs/DNA senor for 0.1 μM Hg2+ against other metal cations (Cu2+, Pb2+ and Cd2+ (0.1 μM)). (b) DPV corresponding to the detection of 0 M, 0.1 μM of Hg2+ in metal ion mixture (Cu2+,Pb2+ and Cd2+ (1 μM)) and 0.1 μM of Hg2+.

Figure 6

(a) Nyquist plot of the cysteine/GSH-AuNPs/DNA senor for 0.1 μM Hg2+ against other metal cations (Cu2+, Pb2+ and Cd2+ (0.1 μM)). (b) DPV corresponding to the detection of 0 M, 0.1 μM of Hg2+ in metal ion mixture (Cu2+,Pb2+ and Cd2+ (1 μM)) and 0.1 μM of Hg2+.

Determination of Hg2+ in field samples

The analytical performance of the developed electrochemical DNA modified electrode was applied for the determination of the mercury ion in wastewater samples. The tested water samples were obtained from the Tunisian public industrial company ONAS, which is interested in the management of the sanitation sector. The wastewater samples were analyzed with voltammetric and impedance metric DNA sensors, and were compared to the results obtained using the spectrophotometric method. The results presented in Table 2 indicate the applicability of the voltammetric and impedancemetric methods to quantify mercury ions in wastewater samples. This result proves the usefulness of DNA/AuNPs-glutathione/cysteine modified electrodes for the detection of mercury in aqueous samples. Furthermore, in the present system the sensor probe can be regenerated in 100 mM of ascorbic acid solution for 1 hour to reduce Hg2+ into Hg+, and might exhibit a weak coordination with thymine as described by Liu et al. (2009). The mercury sensor showed good reproducibility and efficacious regeneration (relative standard deviation below 5.0%).

Table 2

Application of the DNA-Sensor to Hg2+ ions' determination in wastewater samples. The Hg2+ amount was given in mg.L–1

Wastewater Impedancemetric method Voltammetric method Spectrophotometric method 
Sample 1 0.00459 0.00468 0.00462 
Sample 2 0.00876 0.00880 0.00870 
Sample 3 0.0120 0.0119 0.0122 
Wastewater Impedancemetric method Voltammetric method Spectrophotometric method 
Sample 1 0.00459 0.00468 0.00462 
Sample 2 0.00876 0.00880 0.00870 
Sample 3 0.0120 0.0119 0.0122 

CONCLUSIONS

In this work an electrochemical DNA biosensor has been successfully fabricated based on thymine-Hg2+-thymine complexation for trace mercury ion quantification. Incidentally, the developed biosensor, based on an AuNPs-glutathione/cysteine-DNA matrix, was simple, sensitive and rapid. The Hg2+ ions were detected and identified at trace level quantities with a detection limit of 50 pM. A linear relationship of the biosensor response to the analyte concentration was obtained in the dynamic range from 50 pM to 0.1 μM. The obtained results revealed that this sensor exhibited a high sensitivity for mercury among other tested metal ions. The applicability of the DNA sensor for the determination of low concentration levels of Hg(II) ions in wastewater samples was successfully tested with a high reproducibility.

REFERENCES

REFERENCES
Anastassopoulou
J.
2003
Metal–DNA interactions
.
J. Mol. Struct.
651–653
,
19
26
.
Bolger
P. M.
Schwetz
B. A.
2002
Mercury and health
.
N. Engl. J. Med.
347
,
1735
1736
.
Jarczewska
M.
Kierzkowska
E.
Ziółkowski
R.
Górski
Ł.
Malinowska
E.
2015
Electrochemical oligonucleotide –based biosensor for the determination of lead ion
.
Bioelectrochemistry
101
,
35
41
.
Mutter
J.
Naumann
J.
Schneider
R.
Walach
H.
Haley
B.
2005
Mercury and autism: accelerating evidence?
Neuro Endocrinol. Lett.
26
,
439
446
.
Prodi
L.
Bolletta
F.
Montalti
M.
Zaccheroni
N.
2000
Luminescent chemosensors for transition metal ions
.
Coord. Chem. Rev.
205
,
59
83
.
US Environmental Protection Agency
2009
National Primary Drinking Water Regulations. EPA 816-F-09
.
Wojcik
D. P.
Godfrey
M. E.
Christie
D.
Haley
B. E.
2006
Mercury toxicity presenting as chronic fatigue, memory impairment and depression: diagnosis, treatment, susceptibility, and outcomes in a New Zealand general practice setting (1994–2006)
.
Neuro. Endocrinol. Lett.
27
,
415
423
.
Zheng
W.
Aschner
M.
Ghersi-Egea
J. F.
2003
Brain barrier systems: a new frontier in metal neurotoxicological research
.
Toxicol. Appl. Pharmacol.
192
,
1
11
.