Introduction

Electrochemical techniques in recent years raise increasing interest in detecting DNA damage mainly because of their simplicity, fast response, and low cost. A variety of different electrochemical strategies have been employed to detect DNA lesions, based on either the direct electrochemical response of guanine bases in DNA or the redox signal of electroactive probes that have specific interactions with DNA [1]. Ultraviolet (UV) radiation from the sun, especially that characterized by shorter wavelengths and high frequencies, is associated with skin cancer and immunosuppression and caused damage to DNA due to the fact that DNA bases directly absorb UV photons [2]. This effect is determined by the formation of DNA lesions, which leads to the creation of covalent bonds between the pyrimidine bases and the formation of cyclobutane pyrimidine dimers [3]. Induction of DNA damage by solar UV irradiation can result in (i) misincorporation of bases during the replication process, (ii) hydrolytic damage, which results in deamination of bases, depurination, and depyrimidination, (iii) oxidative damage, caused by direct interaction of ionizing irradiations with the DNA molecules as well as mediated by UV irradiation-induced free radicals or reactive oxygen species (ROS), and (iv) formation of alkylating agents that may result in modified bases, all these depending on the type of UV irradiation [4].

In particular, the formation of ROS in living organisms is a well-known process, and their levels are regulated by enzymes or antioxidants. However, the excessive accumulation of ROS can possess a potential risk for human and animal health. The balance disturbance between ROS production and antioxidant activity is called oxidative stress, which is accompanied by the degradation of all cellular components (proteins, lipids, and DNA), leading to different types of diseases or genetic disorders. Regarding DNA damage as the result of an interaction between generated ROS radicals and a DNA molecule, various changes of the primary DNA structure can occur, ranging from the exchange/loss of DNA bases to the formation of strand breaks and cross-links [5].

For characterization of the DNA structure, indicators that bind to the DNA non-covalently (via intercalation of planar aromatic molecules between base pairs in the duplex DNA [6]) are of special interest [7, 8]. Methylene blue is an organic dye that belongs to the phenothiazine family, intercalates into dsDNA, and can be efficiently reduced through DNA-mediated charge transport [9, 10]. Several strategies have been employed to study the interaction between DNA and methylene blue [11, 12]. The detection of hybridization was accomplished by using the specific interaction of methylene blue with guanine [13]. Lin et al. [14] developed a dsDNA/graphene oxide-chitosan/AuNPs/GCE sensor for the detection of DNA damage induced by endocrine-disrupting compounds with the use of methylene blue as a redox indicator. García-González et al. [15] used methylene blue as an electroactive label covalently attached to ssDNA for the preparation of bioassay. Methylene blue was also used by De Grozals et al. [16] for DNA labeling.

Amalgam electrodes are environmentally friendly alternatives to mercury electrodes suitable for batch analysis or flow injection analysis. They are being considered as non-toxic alternative analytical tools to traditional mercury electrodes, with a broad negative potential window offering high hydrogen overpotential, good mechanical stability, and sufficient sensitivity in the determination of various electrochemically active organic compounds [17]. Several types of amalgam electrodes have been introduced in DNA analysis and DNA biosensor development, including detecting DNA hybridization [18] and DNA-chemical agent interactions [19]. Polished and mercury film-modified silver solid amalgam electrodes were used for the measurements of intrinsic redox and tensammetric voltammetric signals of single-stranded and double-stranded, linear, or supercoiled DNA, synthetic polynucleotides, and free adenine base [20] and for the voltammetric determination of adenine and guanine purine bases [21]. Mercury meniscus-modified silver solid amalgam electrodes (m-AgSAEs) were used by Danhel et al. [22, 23] for electrochemical studies of dsDNA modified by multi-redox labels. Only one publication from 2004 [24] was devoted to the detection of DNA damage caused by a cleaving agent at a silver solid amalgam electrode in connection with alternating current voltammetry. Here, the qualitative differences between the voltammetric behaviour of covalently closed supercoiled DNA and linear DNA allowed sensitive detection of strand breaks in DNA structure. Due to the difficulty of the whole procedure (when the copper cleaving complex was prepared by mixing CuSO4 and 1,10-phenanthroline), the researcher did not pay proper attention to the detection of DNA damage on amalgam electrodes.

Therefore, this study should provide a relatively fundamental attempt to move a bit further in this field, so the amalgam electrodes, on which the DNA is spontaneously immobilized very well, can compete with carbon electrodes, where the immobilization step is not always quite simple without previous modification.

In this study, methylene blue and hexaammineruthenium chloride (RuHex) as the redox-active molecules provide a highly sensitive way to detect UV-induced DNA damage at the m-AgSAE, which surface can be easily renewed, either mechanically (old meniscus is wiped off and after short polishing, a new one is formed) or electrochemically by a series of suitable potential cleaning pulses [25]. Two different types of dsDNA (salmon sperm and calf thymus) and two different wavelengths (254 nm and 365 nm) are used for the irradiation of the biosensor surface. A complex DNA biosensor response is based on difference pulse voltammetric (DPV) signal of the intercalative indicator methylene blue, cyclic voltammetric (CV) response of the RuHex redox indicator in solution, and modern electrochemical impedance spectroscopy (EIS) detection.

Materials and methods

Reagents

Low molecular salmon sperm dsDNA (ssp DNA) and calf thymus dsDNA (ct DNA) of final concentration 0.1 mg/mL obtained from Sigma-Aldrich, USA, were prepared by dilution with the 0.1 M phosphate buffer (PB) pH 7.0 and stored in the fridge at 4 °C. The stock solution of methylene blue (MB, c = 1 mM) was prepared by dissolving an accurately weighed amount of dye in deionised water and stored in the fridge at 4 °C. Working solutions of MB (c = 0.01 mM) were prepared freshly before every use. A 1 mM RuHex was used as a redox probe; a 0.1 M PB of pH 7.0 was prepared in the usual way and used as a supporting electrolyte; a 0.2 M potassium chloride solution (KCl) was used for the activation of the working electrode. Deionised water produced by a Milli-Q Plus system (Millipore, USA) was used to prepare all solutions. All reagents (Lach-Ner, Czech Republic) were of p.a. purity grade and stored in glass vessels at ambient temperature unless stated otherwise.

Apparatus

Voltammetric and impedimetric measurements were performed with the potentiostat Autolab III/FRA2 (Methrom Autolab, Switzerland) driven by a NOVA 1.11 software. All measurements were carried out using a three-electrode system with a mercury meniscus–modified silver solid amalgam working electrode (m-AgSAE), silver/silver chloride reference electrode (Ag/AgCl/sat. KCl) and platinum auxiliary electrode, all contained in an electrochemical cell with a volume of 10 mL at ambient temperature. UV lamp (P-LAB, Czech Republic) with working wavelengths 254 nm (UV-C) and 365 nm (UV-A) and radiation power of 15 W (from 1 cm light source distance) was used for the UV irradiation. The pH measurements were carried out with a digital pH meter with a combined glass electrode.

Procedures

Before starting measurements, oxygen was removed from the measured solution by purging with nitrogen for 60 s. Before entering a voltammetric cell, nitrogen passed through a bubbler containing deionised water.

The m-AgSAE was prepared by immersing the surface of polished silver solid amalgam electrode (polished with alumina powder, 1.1 µm) into liquid mercury for 15 s while stirring and then rinsed with deionised water. After creating a new meniscus, each day before starting measurements and after electrode passivation, the working electrode was activated electrochemically in the 0.2 M KCl at − 2200 mV for 300 s. Then, an electrochemical regeneration of the working electrode took place in order to reduce the influence of electrode passivation. Regeneration was based on 150 times switching the electrode potential between an initial (E1reg) and a final (E2reg) regeneration potential for 0.05 s.

The electrochemical DNA-based biosensor was prepared by dipping the m-AgSAE surface into 0.1 mg/mL ssp DNA or ct DNA solution in PB for 1 min under stirring. dsDNA was thus physically adsorbed on the surface of the activated m-AgSAE. The thus-prepared biosensor was washed with deionised water and then used. The surface of the DNA biosensor was exposed to UV irradiation with wavelength 254 nm or 365 nm for various periods of time from the 1 cm light source distance. After the exposure, the biosensor was rinsed with deionised water and the electrochemical measurements were performed.

DPV of MB was performed using the modulation amplitude of 50 mV, the modulation time of 100 ms, the interval time of 150 ms, the step potential of − 3 mV, and the scan rate of 20 mV/s. The peak heights recorded using DPV were evaluated from the straight lines connecting the minima before and after the peak. The CV of RuHex was performed using the step potential of − 3 mV and the scan rate of 20 mV/s. For the EIS of both, MB and RuHex, the polarization potential of − 200 mV, the frequency range from 0.1 to 5000 Hz (51 frequency steps), and the amplitude of 10 mV were used. EIS data were obtained from Nyquist plots as an imaginary impedance component against the real impedance component (at 5 kΩ) at each excitation frequency.

The DPV cathodic peak current of MB and the CV cathodic peak current of RuHex as well as the EIS imaginary impedance component against the real impedance component (of both MB and RuHex) are expressed as the relative values compared to the values obtained at the biosensors before application of the UV irradiation and are calculated using the ratio:

$${S}_{rel}=\frac{{S}_{1}}{{S}_{0}}\times 100 \%$$
(1)

where S0 and S1 are signals recorded before and after the UV-induced DNA damage at the biosensor.

All experiments were repeated three times (n = 3). The parameters of all mathematical and statistical quantities (for significant level α = 0.05) were calculated using an Origin Pro 9.0 software (OriginLab, USA).

Results and discussion

Methylene blue and DNA interaction study

The electrochemical behaviour of DNA-MB interactions was investigated by means of DPV at the m-AgSAE. DP voltammogram of MB (c = 0.01 mM) (Fig. 1a) shows a cathodic reduction peak at around − 0.15 V after 1 min of incubation at the bare m-AgSAE, while no significant adsorption of MB at the bare m-AgSAE was observed. However, with the concentration and incubation time increase, the adsorption of MB onto the surface of the m-AgSAE became more evident (not shown), which was undesirable for further measurements. At the sspDNA/m-AgSAE biosensor (Fig. 1a), the response of 0.01 mM MB solution was obtained after 1 min incubation which results in a slight shift of MB peak to lower potential values and a significant increase of this peak which can be described as the intercalation of the indicator into dsDNA [13]. The concentration and time of incubation of DNA had no significant effect. The concentration of dsDNA to form a layer (0.1 mg/mL), the time of the dsDNA spontaneous immobilization on the surface of the electrode (1 min), the concentration of MB (0.01 mM), and the time of the accumulation of MB into the dsDNA layer (1 min) were previously optimized for further biosensor development and applications [26]. The biosensor prepared as described above can be further used as a tool for DNA damage evaluation. The same trend was observed for ct DNA.

Fig. 1
figure 1

Electrochemical behaviour of MB at the m-AgSAE and the DNA/m-AgSAE biosensor: DP voltammograms (a) and Nyquist plots (b) of MB (cMB = 0.01 mM) measured at the m-AgSAE and at the sspDNA/m-AgSAE biosensor (cDNA = 0.1 mg/mL) after 1 min contact for the accumulation

Full size image

The behaviour of DNA-MB interaction was also investigated by means of EIS, where the charge transfer resistance of the sspDNA/m-AgSAE in the presence of MB in PB pH 7.0 was probed at − 200 mV (Fig. 1b). In the Randles equivalent circuit model used for fitting the EIS spectra, Rs simulates the ohmic (contact) resistance; Rct represents the charge transfer resistance of the electrode–electrolyte interface; the element C represents the capacitance of double electric layer which is charged simultaneously with the occurring electrochemical reaction. The experimental data are described by an equivalent electric circuit characteristic of the processes that include the adsorption of an intermediate compound; the double layer capacity and the adsorption capacity are replaced by the respective constant phase elements (CPE). This element contains many surface defects such as local charge inhomogeneities (from MB) or adsorbed species (from DNA). Consequently, in the Nyquist plot, the semicircle becomes flattened and exhibits the ideal profile of a purely diffusion-controlled reaction [27]. The impedance of this type of working electrode appeared almost as a linear line, which also indicated excellent conductivity, good electrocatalytic activity, and seamless electron transport on the electrode–electrolyte interface [28]. The resistance of the electron transfer is significantly lower at sspDNA/m-AgSAE compared to the unmodified electrode (Fig. 1b). This can be in correspondence with the fact that the positively charged MB is intercalated into the base pairs of negatively charged dsDNA while DNA film serves here as a bridge for enhanced charge transport which leads to increasing the conductivity and decreasing the resistance at the electrode surface [1]. Again, the same trend was observed for ct DNA experiments.

The voltammograms displayed in Fig. 2 illustrate the shift of the MB peak potential (around − 0.1 V) toward more negative values, and also early increase and subsequent gradual decrease of the MB peak current values. Along with the increase of ssp DNA concentration, we can observe an increase of the new peak current values (around + 0.05 V) and a slight shift of the peak potential. The change of the electrochemical behaviour of both DNA and MB molecules can indicate a formation of a mutual complex. Such redox behaviour of MB in the presence of dsDNA is a well-defined mechanism that has an intercalative mode [29]. Similar behaviour was observed for ct DNA (not shown), when peak current increase and subsequent gradual decrease while accompanied with the slight shift of the peak potential values, however, this time, without the formation of a new peak.

Fig. 2
figure 2

Interaction study of MB and DNA at the m-AgSAE: DP voltammograms of MB (cMB = 0.01 mM) in the absence (black line) and presence of ssp DNA (cDNA(1–9) = 1, 2, 5, 10, 20, 50, 100, 200, and 500 μg/L) (a) and the dependence of the Ip1 (grey) and Ip2 (black) on the log(cDNA) (b). Measured at the m-AgSAE after 1 min contact for the accumulation

Full size image

These data showed that the observed electrochemistry is most efficient at DNA-modified electrodes when the redox probe is bound to the DNA attached at the electrode surface. The previous electrode surface passivation studies showed that in the case of intercalating MB, the DNA base pairs attached at the electrode surface provide a bridge for charge transport. As a consequence, surface passivation has no significant effect on the electrochemistry of MB [30]. Any electrochemical processes are possible to observe due to the fact that MB was employed as a reporter of DNA-mediated processes, as the surface is inaccessible to the molecule [9].

Damage to DNA after UV irradiation of the biosensor in the presence of methylene blue

Two independent electrochemical techniques, i.e., DPV and EIS, were used to detect and evaluate changes in the DNA layer induced by UV irradiation (with wavelengths 254 nm and 365 nm) at the DNA/m-AgSAE biosensor surface using the redox response of MB. The detection strategies before and after the exposure of the biosensor to UV irradiation are schematically represented in Fig. 3. Firstly, the m-AgSAE surface was modified with dsDNA and the prepared biosensor was immersed into the solution of 0.01 mM MB. Upon intercalation of MB into the dsDNA structure, the cathodic peak of MB was registered (Fig. 3a). In order to detect the DNA damage, a newly prepared biosensor was exposed to daylight as well as to the UV irradiation of different wavelengths (254 and 365 nm) and then immersed in the MB solution (Fig. 3b). The formation of strand breaks and the release of dsDNA fragments from the electrode surface were detected indirectly via monitoring the decrease of the MB cathodic peak upon intercalation. In addition, these structural changes can be investigated using EIS which controls the charge transfer resistance before and after the breakage of the DNA chains. To detect the DNA damage by UV irradiation, exposure of the DNA biosensor to daylight was used as a control test. The relative change of MB cathodic peak current (Fig. 4a) and the relative change of the Nyquist plot values (Fig. 4b) indicated damage to ssp DNA as a progressive release of the degraded DNA fragments from the biosensor surface (decrease of the Ip,rel and Z"rel values). This degradation became deeper with the exposure time to UV irradiation and was evidently the highest for the most harmful UV-C irradiation with a wavelength of 254 nm. Results showed the slight DNA degradation independent of the time of the irradiation for the wavelength 365 nm and also demonstrated the stability of the prepared biosensor at daylight.

Fig. 3
figure 3

Scheme for DNA damage detection at the dsDNA/m-AgSAE surface using MB: monitoring of MB redox response upon its intercalation to the dsDNA structure before (a) and after (b) the exposure to the irradiation using pulse voltammetric and impedimetric approaches

Full size image
Fig. 4
figure 4

Indirect electrochemical analysis of the DNA damage using response of 0.01 mM MB: the relative change of DPV cathodic peak current values Ip,rel (left) and of the Nyquist plot values Z"rel (right) at the sspDNA/m-AgSAE biosensor (a, b) and the ctDNA/m-AgSAE biosensor (c, d) after the exposure to UV irradiation (254 nm (dark grey) and 365 nm (light grey)) and to daylight (black) for a given time, evaluated using Eq. (1)

Full size image

On the other hand, the behaviour of ct DNA was somehow different. Although the deepest degradation of DNA was recorded again for UV-C irradiation of 254 nm, here the time dependence showed the opposite trend. The relative change of MB cathodic peak current (Fig. 4c) and the relative change of the Nyquist plot (Fig. 4d) were represented as an early decrease and subsequently gradual increase of the Ip,rel and Z"rel values. This can be probably due to the different structure of the DNA chain (10,000 base pairs, compared to ssp DNA with 200 base pairs), where the damage to DNA does not immediately represent the strand breaks and release of the DNA fragments from the electrode surface. In difference, the DNA layer exhibits damage by the opening of the helix structure and, therefore, to the intercalation with the MB present in the solution. For the wavelength 365 nm and for daylight were these values stable up to 10 min. The structure of ct DNA was found to be more stable under these conditions than the structure of ssp DNA; thus, ct DNA needs more time to be damaged because of its more difficult structure and length than ssp DNA.

Damage to DNA after UV irradiation of the biosensor in the presence of hexaammineruthenium chloride

Another approach how to detect damage to DNA at the m-AgSAE can be a use of a non-intercalating redox indicator such as RuHex. It binds in the groove of DNA duplex, is reduced at approximately the same potential as MB, and associates with DNA through electrostatic and hydrogen-bonding interactions [30, 31]. Although RuHex is a well-known redox probe to determine the impedances and the electron transfer through self-assembled monolayers, its use in biosensor applications is still not fully developed for all types of electrode materials (it is mostly used for gold surfaces) [32].

Here, the CV and EIS of the m-AgSAE and the dsDNA/m-AgSAE biosensor took place in order to characterize the sensor before and after the modification step. CV curves of the probe manifested a pair of well-developed redox peaks. As Fig. 5a shows, the signal of the redox probe at CV measurements drifted to higher currents after the immobilization of dsDNA at the m-AgSAE surface. The EIS plot shown in Fig. 5b consists of two regions. One is a semi-circular part in the high-frequency range and the other is a straight line in the low-frequency range. The semi-circular part corresponds to the charge transfer resistance (Rct) between the electrode surface and the electrolyte solution. Also, its intercept at high frequencies on the real axis corresponds to the electrolyte resistance (Rs). The straight line corresponds to the diffusion-controlled process at low frequencies, representing the Warburg impedance (Zw). After modification with the dsDNA layer, it almost became a straight line indicating an improvement in electron transfer because of increasing conductivity at the electrode surface. The enhanced charge transfer occurred since positively charged RuHex complexes are attracted by the immobilized negatively charged dsDNA [33]. The results obtained were in good agreement with CV measurements.

Fig. 5
figure 5

Electrochemical behaviour of RuHex at the m-AgSAE and the DNA/m-AgSAE biosensor: CV (a) and Nyquist plots (b) of RuHex (cRuHex = 1 mM) measured at the m-AgSAE and at the sspDNA/m-AgSAE biosensor (cDNA = 0.1 mg/mL) after 1 min contact for the accumulation

Full size image

Well, CV and EIS took place in order to evaluate UV-induced DNA damage after the irradiation of the biosensor surface for various periods of time from the 1 cm light source distance. Again, the DNA/m-AgSAE biosensor was exposed to UV irradiation of two different wavelengths (254 nm for UV-C and 365 nm for UV-A) and to daylight. A decrease of the CV cathodic peak current (Fig. 6a) in correspondence with a decrease in the Nyquist plot values (Fig. 6b) was the evidence of DNA damage under UV irradiation. The degradation of the biolayer became again deeper with the time of exposure to UV irradiation, which is most evident in the UV-C irradiation. The results again proved the stability of the biosensor response at daylight.

Fig. 6
figure 6

Indirect electrochemical analysis of the DNA damage using response of 1 mM RuHex: the relative changes of CV cathodic peak current values Ip,rel (left) and of the Nyquist plot values Z"rel (right) at the sspDNA/m-AgSAE biosensor (a, b) and the ctDNA/m-AgSAE biosensor (c, d) after the exposure to UV irradiation (254 nm (dark grey) and 365 nm (light grey)) and to daylight (black) for a given time, evaluated using Eq. (1)

Full size image

The same approach was applied to the biosensor with ct DNA attached to the electrode surface. Similarly, a decrease of the CV cathodic peak current was in correspondence with a decrease in the Nyquist plot values, but the sensitivity of the biosensor response was in all cases (for both wavelengths and daylight) comparable. The relative change of CV cathodic peak current values (Ip,rel) (Fig. 6c) and the relative change of the Nyquist plot values (Z"rel) (Fig. 6d) obtained in these experiments are compared. This suggests that ssp DNA attached to the electrode surface, structurally not as complex as ct DNA, is more sensitive to the changes after the short-term UV irradiation.

In comparison with the use of intercalating MB, this concept was less sensitive for the detection of the structural changes of surface-attached DNA, which makes the introducing MB-based detection technique more convenient. It is possible that in the case of RuHex, the DNA film serves to concentrate the molecule near the electrode surface; passivation of the surface could limit direct contact and, therefore, turns off the RuHex electrochemistry [30].

The approaches described in this study are, therefore, expected to provide quite substantial advances in the field of silver amalgam electrodes. Several strategies have been employed to study the detection of DNA damage caused by various physical and chemical agents [34,35,36,37,38], and many of them required chemically modified DNA sequences, e.g. thiol or amine modifications (which rapidly increases the cost), highly specialized equipment, and, in general, high adsorption times. Some of the works dealing with the evaluation of UV-induced DNA damage are summarized and compared in Table 1. In the case of silver amalgam electrodes, neither further modification of the electrode surface nor higher adsorption times are necessary, as the DNA probe is immobilized very well spontaneously compared to commonly used carbon-based or gold electrodes.

Table 1 Comparison of detection techniques, electrode materials and probes for the evaluation of UV-induced DNA damage using DNA-based biosensors
Full size table

Conclusions

Electrochemical detection using a simple electrochemical DNA-based biosensor with mercury meniscus–modified silver solid amalgam working electrode is reported to evaluate UV-induced damage to a surface-attached double-stranded DNA. Three independent electrochemical techniques, (i) differential pulse voltammetric response of DNA intercalator methylene blue, (ii) cyclic voltammetry of hexaammineruthenium chloride redox indicator, and (iii) electrochemical impedance spectroscopy in the presence of methylene blue or hexaammineruthenium chloride, were applied to characterize time changes of DNA structure when it is influenced by UV irradiation of different wavelengths. An amalgam working electrode, as an environmentally friendly alternative to traditional mercury electrodes, is reported here for the first time to detect damage to DNA. These approaches could provide a promising way for a simple and sensitive indication of DNA structural changes caused by various types of physical (drugs, pesticides, nanoparticles) and chemical agents (UV light) from the environment. For example, UV-C is a well-characterized genotoxic agent that is used to study DNA damage and repair [46]. Recently, it has become relevant in various medical applications, including acting as an anticancer agent [47] and as a disinfectant for localized treatments of multidrug-resistant infections [48, 49]. On the other hand, the monitoring of DNA damage represents an important aspect of cancer treatment. However, until recently, the available methods were limited by poor sensitivity and/or long turn-around time. Thus, the emergence of a rapid, sensitive method to quantify the DNA damage that is apparent shortly after treatment has great potential to monitor not only the response of the tumour but also the effects of the treatment on “at-risk” normal tissues, to thus assess the therapeutic ratio of used chemotherapeutics.