DNA and nanophotonics: original methodological approach

1 Introduction

Bionanophotonics is a recently emerged, but already well-defined, truly interdisciplinary field of science and technology aimed at establishing and using the peculiar properties of light and nanoscale light-matter interaction with an emphasis on life science applications [1–9].

The deoxyribonucleic acid (DNA) molecule is well known as the blueprint of life, amazingly rich in information content and very robust. However, its unique structural features and powerful recognition capabilities can also be of interest for assembling artificial structures for a variety of applications in nanophotonics. A DNA helix is, itself, a nanoobject that can be manipulated in various ways, but it can also be treated as a versatile molecular scaffold for building nanoscale devices from the bottom up [1]. DNA’s unique properties are intensively studied from different points of view: as a molecular wire, as a drug delivery system, as a ladder for ordered arrangements of various nanostructures, as a spacer to control distances between nanoobjects, etc.

Recently there is tendency in modern medicine to use such nanoscale interactions, as so-called fluorescence resonance energy transfer (FRET) in the donor-acceptor pair of dye molecules, which intensely absorb light in the visible regions of spectrum and have significant quantum yield [10–15]. Microscopic FRET is also used for cytological diagnosis of tumors [16–19].

On the other hand, there are nearly no works where quantitative estimation of FRET applied to DNA are given. For instance, energy transfer (ET) effectiveness, quantitative estimation of DNA double helix conditions, accessibility of nuclear DNA to intercalator, etc. have not been yet estimated. In this regard, it is to date the modeling of DNA defects in solution and estimation of the quality of double helix that has been investigated. It is interesting to study stress impacts on the DNA by transitive metal (TM) ions, laser irradiation, and heating.

The aim of the present work is to look at the DNA double helix’s application to: 1) double proton transfer in GC and AT pairs; 2) mobile adsorption of H₃O⁺ ions; 3) inner mobility including intercalation; 4) formation of densely packed regular structure suitable for light re-emission for spectroscopic and thermodynamic study of DNA catalytic properties in the following processes: a) redox; b) nanoscale resonance radiationless electron excitation energy transfer; c) formation of interstrand crosslinks; d) performing of photodynamic effects; e) photoinduced conformation transfers in silver nanoparticles and making structures that have plasmon characteristics.

1.1 Inner mobility of DNA

Comprehensive data on the inner mobility of DNA are presented in [20]. Table 1 presents some characteristics of the inner mobility of DNA and RNA. By comparing these movements with their thermodynamic characteristics, a certain correlation can be observed. We can predict that for the excitation of a small amplitude of atom mobility, application of thermal energy is sufficient, but B→A and B→Z transition excitation requires a change in its surroundings (humidity, ionic strength, alcohols, transition metal ions, etc.). Thus, for modeling DNA structural changes, stimulated by the interactions with various ligands, particularly, metal ions (M²⁺), it is necessary to estimate the correlation between the energy of interaction and the lifetime of such complexes, on the one hand, and dynamic characteristics of DNA, on the other.

Table 1

Characteristics of inner movements in DNA.

Type of movement	Time (s)	Excitation energy (kcal/mol)
1. Various small-amplitude oscillating movements of atoms (about 0.1 Å) inside the components of DNA	10-14–10-13 (3500–300 cm-1)	RT(T=300 K)=0.6(210 cm-1)
2. Limited movements of phosphates, sugars, and nucleobases relative to the equilibrium position, torsional, and flexural oscillations of the double helix	10-10–10-8
3. Large-amplitude movements of phosphates, sugars, and bases occurring in connection with the transition of the double helix from one form to another	10-7–10-5	5–6 for B→A transitions and 21 for B→Z transitions [21]
4. Change of free energy ΔG needed for opening of central pairs in double chain RNA at 298 K, kcal/mol bases [22]
	∼3×10-6a	7.5
	∼10-8a	4.0

^aThe time of pair opening is evaluated by expression (3).

1.3 Intercalation

One more interesting characteristic of DNA is intercalation. Intercalator binding coefficients with natural DNA and model nucleotides is between the 10⁴ and 10⁶ sequence. Such a type of molecule intercalator binding with the DNA double helix takes place, as a rule, in two steps. The first step is adsorption of intercalators on periphery areas of the surface; at the second step, the intercalators penetrate into the DNA molecule, which is accompanied by unfolding and elongation of the helix and, certainly, by increasing the helix rigidity. The first step is characterized by weak binding. Binding constants given above are typical for intercalation processes. Besides, the intercalation process is characterized by the principal of excluding the nearest binding places. In the case of proflavine, only 44% of possible binding places are busy [29]. It means that every second place is vacant. But there are some data that allow doubt in the statement. In particular, our group has managed to obtain fluorescence spectra of the DNA ternary complex DNA-AO-EB at a specific concentration of three intercalator molecules to 5 base pairs (bp) of DNA (see Figure 1). It should be underlined that unlike soft ions such as Ag(I), ions of Ni(II), Co(II), Zn(II), Mn(II), Cu(II) choose stable nucleotide dimmers. Intercalators when binding with DNA prefer the same dimmers. For instance, overlapping is much stronger in the sequence pyrimidine – 3′, 5′ – purine than at the intercalation of sequence purine – 3′, 5′ – pyrimidine. It means that, in the first case, the energy gain is 29–54 kJ/mol more than in the second case [31].

Figure 1

Fluorescence spectra of ternary complex DNA-AO-EB. [AO]-3×10^-5 mol/l, [EB]-0 mol/l (1); [AO]-2×10^-5 mol/l, [EB]-1.0×10^-5 mol/l (2); [AO]-1.5×10^-5 mol/l, [EB]-1.5×10^-5 mol/l (3); [AO]-1×10^-5 mol/l, [EB]=2×10^-5 mol/l (4). [DNA]-10^-4 mol/l (P), [NaCl]-0.01 mol/l, pH7 [30].

2 DNA as catalyst in redox reactions

2.1 Oxidation

Figure 2 shows the absorption spectra for AgNPs and AgNPs-DNA complexes. Figure 3 presents first derivatives of absorption spectra for AgNPs and AgNPs in the complex with the DNA.

Figure 2

Absorption spectra of AgNPs and AgNPs-DNA complexes. [AgNPs] – 0.72×10^-4 mol/l (Ag⁰), [DNA] – 1.6×10^-4 mol/l (P), [NaNO₃] – 10^-2 mol/l.

Figure 3

First derivative of absorption spectra of AgNPs and AgNPs-DNA. The blue shift is evident.

The analysis of Figures 2 and 3 shows that at DNA interaction with AgNPs, a short-wave shift of AgNPs absorption band (6 nm) takes place. Besides, a 20% hypochromic effect can be observed.

It should be pointed out that the values of the hypsochromic shift and hypochromic effects depend to a certain extent on the lifetime of AgNPs. A distinct trend is observed in the decrease of these effects connected with time. The short-wave shift points out that at the interaction with DNA, there is a kind of loosening of the interaction between the silver atoms inside the AgNPs. The decrease in the intensity of the absorption band is due to the partial corrosion of AgNPs in the presence of DNA. At DNA interaction with AgNPs and H₃O⁺ in water solutions, AgNPs-DNA and H₃O⁺-DNA complexes are formed. H₃O⁺ ions being in a mobile adsorbed state on the surface of DNA [32, 36] can form complexes with AgNPs. Nanoparticles, in turn, moisturize the DNA surface. We have not found any data on O₂ and NO₃^- molecule adsorption on the DNA surface, though high concentrations of NO₃^- (0.3 mol/l) can oxidize AgNPs (see Figures 4 and 5). Photo-irradiation in water also oxidizes AgNPs [37].

Figure 4

Absorption spectra of AgNPs after various ion strength applications. [AgNPs] – 0.7×10^-4 mol/l (Ag⁰), [NaNO₃] – 10^-2 mol/l (C₁); [NaNO₃] – 3×10^-1 mol/l (C₂).

Figure 5

Absorption spectra of AgNPs and DNA after various ion strength applications. [AgNPs] – 0.7×10^-4 mol/l (Ag⁰), [DNA] – 1.8×10^-4 mol/l (P), [NaNO₃] – 10^-2 mol/l (C₁); [NaNO₃] – 3×10^-1 mol/l (C₂).

As early as 1980 [38], one of the authors of the present work proposed a definition of various ion and molecule affinity to electrons in water solutions as reciprocal to the activation energy ΔGa of a hydrated electron (e^-_aq) with organic and inorganic substances in the water solution rate constant which is evaluated as k_e-/aq=10¹¹exp(-ΔGa/RT) [39]. Taking as an example the study of DNA interaction with transition metal ions by absorption spectrometry and equilibrium dialyses, we have shown that as a rule, the bigger is K_e-/aq value for the metal ions, the more intensive is their interaction with DNA [39].

Table 2 shows rate constants k_e-/aq and 1/ΔGa for the reaction with Ag⁺, H₃O⁺, NO₃^- ions and O₂, H₂O, and DNA molecules, where ΔGa is the hydrated electron activation energy according to k_e-/aq=10¹¹exp(-ΔGa/RT). Besides, in the last column of Table 2, the concentration of the substances under investigation is given.

Table 2

Values for reaction rate constants of hydrated with Ag⁺, H₃O⁺, NO₃^- ions, and O₂, H₂O and DNA molecules.

Ions and molecules	pH	ke/aq (mol-1 s-1)	ΔGa-1 (mol/kcal)	Concentration in solution (mol/l)
Ag+	7	3.2×1010	1.468	1.1×10-4
AgNPs	6	–	–	0.74×10-4 (Ag0)
H3O+	4-5	2.36×1010	1.157	0.3×10-4
NO3-	7	1.1×1010	0.757	10-2
H2 0	8.4	1.6×101	0.074	–
Na+	7	<105	0.121	0.8×10-4
O2	7	1.88×1010	1	1.33×10-6
DNA	7	∼108	0.242	2·10-4 (P)

It can be explained by the ability of DNA to adsorb H₃O⁺ (pK∼4) and metal ions of the first transition row M²⁺ (pK=4-6) [40]. It should be noted that in the study of the interaction of transition metal ions with DNA, non-buffer solutions are usually used. As a rule, the pH of the solutions is 5.5–6. At these pH values, the concentration of H₃O⁺ in the solution is satisfactory for the attacks at DNA because H₃O⁺ adsorption on DNA is of a mobile character [36]. Thus, the surface of DNA can adsorb both AgNPs and H₃O⁺ and serves as a catalyst in the AgNP oxidation reaction. So, we can state with definite certainty that in AgNPs-H₃O⁺ complexes, which exist on the DNA surface, electron transfer from the silver atom to H₃O⁺ can occur because the value of electro-negativity of hydrogen atom is bigger than that of the silver atom and much more than that of the H₃O⁺ ion. We can express it by the following equation:

2AgNPs(n)+2H+→2AgNPs(n-1)+(2AgNPs+DNA)+H2↑

It is necessary to underline that AgNPs used in the experiment have the size of 1–2 nm, and consequently, their concentration is at least two orders less than the one for atoms, which constitute the particles. Assuming that all nanoparticles of the solution are adsorbed on the DNA, and they interact from the side of the double helix major groove, and their average size is 1.5 nm, we can conclude that the average distance between AgNPs adsorbed on the DNA is nearly 60 nm (one particle for about 170 bp). As soon as H₃O⁺ reduction starts (as a result of AgNPs oxidation), the water in the solution dissociates to OH^- and H⁺ supplying new H₃O⁺ until a new dynamic balance of the system is achieved. Ag⁺ ions, which are created during the reaction, have a high stability constant with DNA (pK≥10.8 [36]). As early as 1969, Wilhelm and Daune [41] showed that Ag⁺ ions make two kinds of intra-spherical complexes with G-C DNA pairs: chelate N₇G-O₆G and intrastrand linear complex between N₁G and N₃C, the so-called crosslink. The authors [41] believe that in making the complex of the second type, H₃O⁺ is released from the DNA guanine into the solution. It is an additional mechanism of H₃O⁺ formation in solution. As formation of the second type of complex depends on the DNA dynamic characteristics, i.e., on the frequency of unfolding the base pairs, the possibility of the process will be much higher in case when DNA is denatured. We could observe it in our experiment.

2.2 Reduction

Figure 6 demonstrates the absorption spectra of AgNPs and of the following complexes DNA-AgNO₃-AgNPs, DNA-AgNO₃-AgNPs-AA. In contrast to the spectra shown in Figure 7, the absorption spectra of quaternary complex DNA-AgNO₃-AgNPs-AA have greater intensity and different places for absorption maximums. It can be seen that ascorbic acid (AA) reduced silver ions in ternary complex DNA-AgNO₃-AgNPs. Thus, AgNPs activate the process of quick reduction of Ag⁺ ions to silver atoms. Analyzing the absorption spectra given in Figure 6, we can draw the following conclusions: 1) Ag⁺ ions interfere in H₃O⁺ ion mobility and prevent oxidation of AgNPs and 2) AgNPs, in turn, activate the process of Ag⁺ ion reduction in the presence of AA. The AgNP absorption spectra shift to the red side (6–7 nm), and significant increase in the absorption spectra intensity undoubtedly point out the increase in AgNPs size in the complex with DNA. It should be underlined that, without DNA, nothing happens to AgNPs in water solution.

Figure 6

Absorption spectra of AgNPs and the following complexes DNA-AgNO₃-AgNPs and DNA-AgNO₃-AgNPs-AA. [AgNPs] – 0.7×10^-4 mol/l (Ag⁰), [DNA] – 1.8×10^-4 mol/l (P), [AgNO₃] – 0.7×10^-4 mol/l, [AA] – 1.4×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 7

Spectra for silver ion reduction in complexes. DNA-AgNO₃-AA. [DNA] – 5.9×10^-4 mol/l (P), [AgNO₃] – 1.2×10^-4 mol/l, [AA] – 2.4×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 7 presents the AA effect on the absorption spectra of the DNA complex with silver ions in the visible region. By 160 min, we can clearly see the appearance of a notable spectrum from the reduction of silver atoms. In 24 h, the absorption spectra specific for silver atoms can be observed. It is seen that the absorption spectra have a complex form, they are reminiscent of the absorption spectra of substances having specific resonance interactions. Resonance interactions are specific for structures having identical chromophore and hard structure [42], e.g., for the absorption spectra of poly-peptide in the α-helix state [43]. Interaction of silver ions, which is characterized by inter-cross type of links with DNA, reduces dynamic characteristics of double helix and makes it harder. The other reason for the complexity of the spectra is the reduction of silver ions and their further condensation in clusters and nanoparticles with the size of more than 50 nm [44].

3 Photo-induced DNA-dependent conformational changes in AgNPs

Figure 8 shows the AgNP absorption spectra before and after their heating followed by their cooling. Colloidal silver suspension was placed in a hermetic test tube, and it was incubated in a vessel with boiling water for 15 min. Then, the test tube with the sample was cooled in an ice bath to room temperature. It can be seen from Figure 8 that nanoparticles exposed to heating demonstrate noticeable weakening of absorption intensity, and what is very important is the widening of absorption spectra from 140 nm to 215 nm.

Figure 8

Absorption spectra of AgNPs before (1) and after (2) heating (15 min incubation at T=373 K) [AgNPs] – 0.72×10^–4 mol/l (Ag⁰), [NaNO₃] – 10^-2 mol/l.

Hypsochromic shift and widening of the AgNP absorption spectra caused by their interaction with DNA and heating demonstrates, on the one hand, loosening of AgNPs, i.e., penetration of solvent in nanoparticles, the atoms of which show weaker polarizability than water [45, 46].

The company NanoComposix [44, 47] gives sample absorption spectra for spherical AgNPs with particle sizes of 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 nm at the same mass concentration 0.02 mg/ml. The given data show that with the increase in the size of the particles, widening of the red shift of the absorption band can be observed. It is notable that despite the growth of the particle size, i.e., decrease in their total number in the solution, the intensity of the absorption bands for nanoparticles with the sizes from 10 to 40 nm is not practically changed, only a small shift of absorption band maximums can be observed. The above explicitly points out that chromophore units are silver atoms and not nanoparticles. Thus, we can draw a conclusion that silver atoms in AgNPs are sufficiently isolated and bound together by dispersion interaction (induced dipole-induced dipole). As these interactions are performed in the surrounding water, they should be considerably amplified at the expense of the so-called hydrophobic effect [48], which means compaction and then minimization of the surface (decrease of the system entropy).

We should especially point out that in a nanoparticle, which consists of one kind of atoms, along with the mentioned dispersion interaction, the so-called resonance interaction should take place [42]. Such types of interactions are typical for molecular crystals, and they usually lead to exciton splitting of the principal absorption band. Inevitable condition for exciton splitting is the presence of a system consisting of identical groups and having a hard structure [43]. The absence of splitting can mean that AgNPs under investigation (Figure 2) have a liquid structure resembling a drop, which under definite conditions such as temperature, photo-irradiation, variations in dielectric constant of the environment should be characterized by conformational transitions. So, once again, analyzing the absorption spectra presented in Figure 2, we can make the conclusion that hypsochromic shift shows not only loosening of AgNPs, but moreover, the transfer of the drop into a spherical segment with the wetting angle θ<π/2 (see Figure 9). It means that the DNA surface is moistened by silver nanoparticles. First of all, it is connected with great affinity of soft ions (Cu⁺, Ag⁺, Hg⁺⁺, etc.) and metal ions M⁰ with the DNA double helix [36]. Photo-diffusion of AgNPs (see Figure 10) on the DNA double helix resembles flesh-desorption phenomena, i.e., fast heating of AgNPs by photons, then desorption of silver ions with their following adsorption by DNA double helix, including creation of crosslinks between silver atoms and DNA chains.

Figure 9

Model of DNA moisturizing by silver nanoparticles.

Figure 10

Absorption spectra of AgNPs-DNA before and after irradiation (5-min interval). [AgNPs] – 1.94×10^-4 mol/l (Ag⁰), [DNA] – 1.6×10^-4 mol/l (P), [NaNO₃] – 10^-2 mol/l.

The kinetics of AgNP photoirradiation has also been studied. Figures 10 and 11 show superposed absorption spectra of free AgNPs and AgNPs in complex with DNA before and after irradiation using a water filter.

The analysis of the spectra in Figures 10 and 11 demonstrates that only the irradiated complexes AgNPs-DNA have distinctly expressed isosbestic point. The test with the free AgNPs shows that, as a result of photoirradiation, desorption of silver atoms and their oxidation to Ag⁺ ions takes place. The presence of isosbestic points in the absorption spectra of irradiated AgNPs-DNA complexes proves that the system has no less than two states, i.e., AgNPs-DNA complexes have several forms of existence joint by structural photodiffusive transition from one form, e.g., spherical one, to an extended long, and probably one-dimensional form along the DNA double helix. The analysis of the spectra really shows with good correlation (≤5%) that the space under the spectra is preserved, which means that there are no changes in the chromophore electron structure. Besides, half the width of the absorption spectra Δλ_1/2 is changed from 140 nm to 360 mn. Red shift and widening of the AgNP absorption band points to the typical for molecular system increase of electron conjugation (linear and cyclic conjugated systems [49]).

Figure 11

Absorption spectra of AgNPs before and after irradiation (5-min interval). [AgNPs] – 1.94×10-⁴ mol/l (Ag⁰), [NaNO₃] – 10^-2 mol/l.

We have also carried out the evaluation of energy needed for the heating of AgNPs (1–2 nm). Under the condition when atomic (specific) heat capacity of liquid silver is equal to 30.5 J/(g-atom×grad); single photon energy (λ=430 nm) is 46.2×10^-20 J, it is possible to heat a single silver nanoparticle (size 1 nm) consisting of 30 silver atoms up to 610 K; in the case when a silver nanoparticle has a size of 2 nm, i.e., consists of 240 atoms, it can be heated up to 340 K. It means that photo-irradiation (λ=430 nm) has the absolute ability to cause the photodiffusion of AgNPs, especially of those with size about 1 nm.

To check the concept, we have carried out thermodynamic kinetic analysis of absorption spectra of AgNPs-DNA complexes (see Figure 10).

Let us consider the changes in absorption spectra for photo-irradiated AgNPs-DNA complexes given in Figure 10versus the duration of irradiation in M_t/M_e and t^1/2 (see Figure 12). M_e=A_t=0-A_t=6600 is the number of silver atoms in the nanoparticles at the beginning (absorption A_t when λ=430 nm at t=0), M_t=A_t=0-A_t is the molar quantity of silver atoms desorbed by the time moment t (difference between absorption A_t=0-A_t at λ=430 nm). As can be seen, the curves in Figure 12 have an S-shape for photodesorption kinetics of silver atoms both from the surface of free AgNPs and AgNPs-DNA complexes. The S-shape appearance of the curves denotes that photo-induced desorption of atoms is a complex and multiphase process [37]; it means diffusion of silver atoms from the inner part of a nanoparticle to its surface, conformational changes in the particles especially in the ones that are adsorbed on the DNA surface. Next, we are going to consider the results given in Figures 10 and 11 in ln[M_e/(M_e-M_t) and t coordinates, which are given in Figure 13.

Figure 12

Kinetic curve of photo-desorption in M_t/M_e and t^1/2 coordinates for AgNPs bound with DNA.

Figure 13

Kinetic curve of photodesorption in ln[M_e/(M_e-M_t)] and t coordinates for AgNPs bound with DNA.

The analysis of the curve shows that only the initial stage of the given curves of the desorption kinetics obey the linear law of Langmuir first-order equation

The constant of desorption rate for silver atoms from the surface of AgNPs has been evaluated from the slopes of the curves, and the data is, for AgNPs bound with DNA, k_d≅9×10^-5s^-1. The values allow us to estimate the activation energy E_d for the desorption reaction using the equation

where υ₀ is the pre-exponential factor assumed as υ₀≈10¹⁰s^-1 (reciprocal quantity to silver atom oscillation time in nanoparticles). In this case, we have acquired the value for E_d≅80 kJ/mol Ag⁰ for AgNPs bound with DNA at T=300 K As E_d=E_a+Q_a, where E_a is the adsorption activation energy, and Q_a is the adsorption heat of nanoparticles, so Q_a≅80 kJ/M Ag⁰ under the condition that formation of nanoparticles is not an activated process. The value of heat is specific for cluster nanostructures [50].

Silver atom photodesorption from the surface of nanoparticles, included in the DNA complex, has a desorption activation energy of 80 kJ/mol Ag⁰, and consequently, we can assume that photo-induced diffusion of AgNPs on DNA double helix takes place along with the activation of silver atom desorption energy, which equals to 80 kJ/ mol Ag⁰, and thus, we can state that the energy of Ag⁰ interaction with the DNA double helix is not <80 kJ/mol Ag⁰. In accordance with the Gibbs equation

and taking into account the evaluated energy 80 kJ/mol Ag⁰, we assume that the stability constant of the complex is not <10¹⁴, and consequently, the lifetime of the complexes is equal to 10⁴ s. The lifetimes of 1 s order and more are characteristic for interstrand interactions with the participation of transition metal ions, so-called crosslinks [22]. As early as 1969, Wilhelm and Daune [23] showed that Ag⁺ ions form crosslinks between DNA chains, thus, releasing protons bound with N1 guanine and N3 thymine into the solution. We have estimated stability constants of Cu⁺ and Ag⁺ ions with DNA, which are equal to pK=10.8 for Ag⁺ and pK=14.9 for Cu⁺. Thereafter the change of free energy for Ag⁺ is 63 kJ/mol and 86 kJ/mol for Cu⁺ and life times are 0.63 s for Ag⁺ and 8.6×10³ s for Cu⁺ [4].

4 Nonradiative energy transfer between intercalator molecules and defects in the DNA duplex

Interaction of intercalator molecules with DNA, particularly, acridine orange (AO) and ethidium bromide (EB) (Figure 14), depends on ionic strength of the solution, DNA nucleotide content, its sequence [51], and double helix structure. Besides, such an interaction depends on transition metal ions (TM) [32], which cause ejection of intercalators, though intercalator molecules and TM ions have different binding sites on DNA. We should point out that AO and EB molecules as well as TM ions [Mn(II), Co(II), Ni(II), Cu(II), and Zn(II)] at the interaction with the DNA double helix have similar values of stability constants, and their pK are in the 4–6 interval [51]. At the interaction with DNA, TM ions provoke point defects such as double proton transfer in GC pairs, depurinization, create interstrand crosslinks [27], which are the reason for intercalator ejection from DNA [51]. In its turn, the above influences the efficiency of nonradiative transfer of electronic excitation energy from donor (D) to acceptor (A).

Figure 14

(A) Structure of acridine orange and (B) structure of ethidium bromide.

The base of the Förster mechanism of nonradiative transfer of electronic excitation energy is the so-called inductive-resonance transfer of energy from D to A, where dipole-dipole interactions dominates. According to Förster, in our case, the rate of energy transfer k_ET is in direct ratio to the donor emission quantum yield (q_oD), overlap integral donor emission spectra, and extinction coefficient of acceptor molecule, and inversely to the solvent refraction index in the fourth degree, distance between excited D* and A in the sixth degree, and τ_D is the donor emission decay time [52].

On the other hand, at inductive-resonance transfer of energy through thin dielectric layers with alternative thickness d and permanent concentration of D and A, the following correlation will be achieved [53, 54].

where s=4, 6 and 2 are the energy transfers, consequently, for electric dipole-electric dipole, electric quadruple-electric dipole and magnetic dipole-electric dipole, d₀ is the critical thickness of the layer. Correlation (7) in the case of energy transfer from AO to EB intercalated in the DNA double helix can be rewritten as follows:

where q_oD is the quantum efficiency of donor fluorescence when the distance between energy donor and acceptor R→∞, q_D at a given R, q_oA is the quantum efficiency of sensibilized acceptor fluorescence at R→0 and q_A at a given R. The value e_ET=1-q_AO/q_oAO was estimated as electron excitation energy transfer efficiency.

So, having a definite concentration of intercalated pair D-A in DNA and changing the concentration of the DNA double helix, the efficiency of energy transfer in the particular D-A pair can be sufficiently changed. Thus, the efficiency of energy transfer is proportional to the DNA double helix site concentration. In Section 3, we show the validity of such an approach for DNA-AO-EB ternary complexes.

Figure 15 shows the fluorescence spectra of binary and ternary AO-DNA and AO-ED-DNA complexes where concentration of DNA changes the distance between donor AO and acceptor EB. AO and EB concentrations were constant and equal to 0.14×10^-4 mol/l. DNA concentration varied from 0.5×10^-4 to 10×10^-4 mol/l per bp. Figure 16 shows the ignition curves of AO fluorescence depending on the distance between AO and EB intercalated in DNA. The distance between donor and acceptor is given in nm and in bp units. The ignition curves of AO fluorescence are built in correspondence to Eq. (8) for the s index equal to 4, 6, or 2. An important characteristic of the energy transfer process is Foerster distance R_0.. At this distance, half of the donor molecules decay by energy transfer, and the other half decays by the usual radiative and nonradiative rates [55].

Figure 15

Fluorescence spectra of double and ternary DNA-AO and DNA-AO-EB complexes where concentration of DNA changes the distance between donor AO and acceptor EB. (a) DNA-AO [DNA] – 2.8×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l. (b) DNA-AO-EB [DNA] – 10×10^-4 mol/l (P), [EB] – 0.14×10^-4 mol/l, (c) DNA-AO-EB [DNA] – 8×10^-4 mol/l (P); (d) DNA-AO-EB [DNA] – 6×10^-4 mol/l (P); (e) DNA-AO-EB [DNA] – 4×10^-4 mol/l (P); (f) DNA-AO-EB [DNA] – 2.8×10^-4 mol/l (P); (g) DNA-AO-EB [DNA] – 1.4×10^-4 mol/l (P); (h) DNA-AO-EB [DNA] – 10^-4 mol/l (P).

Figure 16

Ignition curves of AO fluorescence depending on the distance between AO and EB intercalated in DNA.

This expression allows to calculate the Förster distance from the spectral properties (J(λ)) of the donor and the acceptor and the donor quantum yield (q_oD), i.e., in terms of the experimentally known values taking into account the environment refractive index (n) and orientation factor (κ²) of fluorescent molecules. In our case, J(λ)=2.72×10¹⁴ (mol/l)^-1 (cm)^-1 (nm)⁴, q_oD=0.75 [56], κ²=2/3. For the index of refraction, we choose n=1.6 [57]. Thus, the Förster distance is calculated as R₀=3.5±0.3 nm, which is in good agreement with the experimental value R₀=3.9±0.3 nm.

Table 3 presents the calculated values for R₀ for different n and κ². As a refraction index, some authors use n=1.33 (water) [58], others assumed n to be 1.40, which is valid for biomolecules in aqueous solution [55] or n=1.60 in the case of DNA [57]. In all cases, any dispersion effects are ignored. The term κ² is a factor describing the relative orientation of donor and acceptor transition dipoles in space. κ² is usually assumed to be equal to 2/3, which is appropriate for dynamic random average orientation of the donor and acceptor. When donor and acceptor are oriented in parallel planes, then κ²=1/2 (for details, see [55]). Evidently, the orientation factor κ² and refractive index n for different media do not affect significantly the R₀ value. In Figure 16, we can see that the theoretical curve describing electron excitation transfer for the case electric dipole-electric dipole (S=4) corresponds to experimental data, which is the best. Table 4 gives the values for efficiency e_ET=(1-q_D/q_OD) for AO intercalated in DNA depending on the distance between AO and EB given in bp units. The same table allows to estimate the distance between donor and acceptor in correspondence with the values of e_ET evaluated from fluorescence spectra of FRET.

It is well known and it was shown in our work [27] that transition metal ions at the interaction with DNA cause or participate in different conformational changes, e.g., Cu(II) ions initiate DNA transition from B to C form [59]; with the help of TM ions and ethanol, B-Z transition can be initiated [27]. The effect of Ag(I) ions on the DNA structure is known [41]. As early as 1996, we discovered the ejection of AO and EB intercalators from DNA. Besides, the rise in temperature in the solution causes melting of the DNA double helix. In this connection, it was interesting to investigate the electron excitation transfer in D-A pairs intercalated in DNA under the effect of different stress factors with the aim of finding intact sites of DNA double helix.

5 Nanoscale donor-acceptor fluorescent probing of the DNA double helix and stress factors: models and mechanisms

The paragraph is devoted to the effects of Cu(II), Cu(I), and Ag(I) ions, silver nanoparticles, temperature, and laser irradiation on electron excitation energy transfer, efficiency between intercalated in DNA fluorescence molecules of acridine orange (donor) and ethidium bromide (acceptor). Besides, the paragraph deals with the models and mechanisms of DNA defect formation.

Figures 17–20 demonstrate the influence of Cu(II), Cu(I), Ag(I), AgNPs, and laser irradiation (λ=457 nm) on electron excitation energy transfer efficiency from AO to EB intercalated in DNA, which is shown in the rise of effectiveness of FRET. The estimated data are given in Table 4. There are also the data for the distance between AO and EB in bp units, as well as the relative concentrations of DNA sites applicable for intercalation. Figure 17–20 also show that these ions quench AO fluorescence – the phenomena was studied by our group in [36, 51], and it is connected with both dynamic quenching and the quenching caused by nonradiative transfer of excitation energy. On the one hand, Cu(II), Cu(I), Ag(I), AgNPs, and laser irradiation (λ=457 nm) quench AO fluorescence and, on the other hand, increase FRET intensity. It obviously points out that there are different reasons for the phenomenon, in particular, FRET intensity is connected with the quality of double helix, i.e., stability constant of AO and EB with DNA.

Figure 17

Influence of Cu(II) and laser irradiation (λ=457 nm) on electron excitation energy transfer effectiveness from AO to EB intercalated in DNA. (a) DNA-AO; (b) DNA-AO-EB; (c) DNA-AO-Cu(II); (d) DNA-AO-Cu(II) 10 min irradiation; (e) DNA-AO-Cu(II)-EB; (f) DNA-AO-Cu(II) 10-min irradiation+EB. [DNA] – 9.6×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [CuCl₂] – 0.14×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 18

Influence of Cu(I) ([ascorbic acid]/[Cu²⁺] 2:1) and laser irradiation (λ=457 nm) on electron excitation energy transfer effectiveness from AO to EB intercalated in DNA. (a) DNA-AO; (b) DNA-AO-EB; (c) DNA-AO-Cu(I); (d) DNA-AO-Cu(I) 10-min irradiation; (e) DNA-AO-Cu(I)-EB; (f) DNA-AO-Cu(II) 10-min irradiation+EB. [DNA] – 9.6×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [CuCl₂] – 0.14×10^-4 mol/l, [AA] – 0.24×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 19

Quenching of fluorescence by Ag⁺ ion in DNA-AO-EB complex. (a) DNA-AO, (b) DNA-AO-EB-Ag⁺ (C₁), (c) DNA-AO-EB-Ag⁺ (C₂), (d) DNA-AO-EB-Ag⁺ (C₃). [DNA] – 2.8×10^-4 mol/l (P), [AO] – 0.14·10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [Ag⁺] – 0 (C₁), [Ag⁺] – 6.0×10^-6 mol/l (C₂), [Ag⁺] – 30.0×10^-6 mol/l (C₃), [NaNO₃] – 10^-2 mol/l. λ=460 nm.

Figure 20

Quenching of fluorescence by AgNPs in DNA-AO-EB complex. (a) DNA-AO, (b) DNA-AO-EB-AgNPs (C₁), (c) DNA-AO-EB-AgNPs (C₂), (d) DNA-AO-EB-AgNPs (C₃). [DNA] – 2.8×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [AgNPs] – 0 (C₁), [AgNPs] – 6.0×10^-6 mol/l (C₂), [AgNPs] – 18.0×10^-6 mol/l (C₃), [NaNO₃] – 10^-2 mol/l. λ=460 nm.

Besides, the temperature effect on FRET stability was investigated. Figure 21 shows the heating effect on DNA solution located in a hermetic test tube in a thermostat. Two milliliters of DNA solution was heated at various temperatures T=323, 333, 343, 353, and 363 K, and then, it was taken out and put into an icy bath (T=273 K). At 293 K, an intercalator pair AO-EB was added to the solution, and the fluorescence spectra were registered.

Figure 21

Heating effect on DNA solution located in a hermetic test tube in thermostat at various temperatures t=323, 333, 343, 353 and 363K. (a) DNA-AO, (b) DNA-AO-EB, (c) DNA-AO-EB 323K, (d) DNA-AO-EB 333K, (e) DNA-AO-EB 343K, (f) DNA-AO-EB 353K, (g) DNA-AO-EB 363K . [DNA] – 2.8×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 22 shows the fluorescence spectra for ternary complexes AO-DNA-EB. The complexes were prepared as follows: DNA solutions put in hermetic test tubes were kept in vitro in thermal bath at the temperature 373 K for different periods of time (5, 10, and 20 min). Then, they were quickly cooled in an icy bath (T=273 K). After the procedure, the intercalator pair AO-EB was added to the solution, and fluorescence spectra were registered. From the spectra given in Figures 21 and 22, and with the application of Table 4, the effectiveness of FRET, the data for the distance between AO and EB in bp units, as well as the relative concentrations of DNA sites applicable for intercalation were estimated. The results are given in Table 5.

Figure 22

Fluorescence spectra for ternary complexes AO-DNA-EB. The complexes were prepared as follows: DNA solutions put in hermetic test tubes were kept in vitro in thermal bath at a temperature of 100°C for different periods of time (5, 10, and 20 min). (a) DNA-AO, (b) DNA-AO-EB (0 min), (c) DNA-AO-EB (5 min), (d) DNA-AO-EB (10 min), (e) DNA-AO-EB (20 min). [DNA] – 2.8×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 23 presents changes in the values of the AO molecule (λ=502 nm) absorption spectra maximums in water solutions and in binary and ternary complexes with DNA-AO and DNA-AO-Cu(II), Cu(I), and Ag(I). Analyzing the data, we can see that Cu (II) and Cu (I) ions decrease the destructive effect of irradiation, while Ag(I) ions increase it.

Figure 23

Laser irradiation effect on AO molecules in binary and ternary complexes AO-DNA, AO-DNA-Cu(II), AO-DNA-Cu(I), DNA-AO-Ag, and AO. ▸—AO-DNA-Cu(I); •—AO-DNA-Cu(II), ▹—AO-DNA, ■—DNA-AO-Ag(I) □—AO. [DNA] – 0.7×10^-3 mol/l (P), [AO] – 0.7×10^-4 mol/l, [EB] – 0.7·10^-4 mol/l, [AA] – 1.4×10^-4 mol/l, [Ag⁺] – 0.7×10^-4 mol/l, [CuCl₂] – 0.7×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

Figure 24 shows the change of energy transfer efficiency from 13 bp to 7 bp from AO to ethidium bromide intercalated in DNA after 20 min of irradiation of the binary DNA-AO complex (see Table 5).

Figure 24

Influence of laser irradiation (λ=457 nm) on electron excitation energy transfer effectiveness from AO to EB intercalated in DNA. (A) AO-DNA; (B) AO-DNA-EB; (C) AO-DNA irradiation 25 min; (D) AO-DNA irradiation 25 min+EB. [DNA] – 7×10^-4 mol/l (P), [AO] – 0.14×10^-4 mol/l, [EB] – 0.14×10^-4 mol/l, [NaNO₃] – 10^-2 mol/l.

The following paragraphs will show new models and mechanisms of defect formation in DNA caused by stress factors.

5.1 Double proton transfer in GC pairs in DNA

The influence of H⁺ ions on UV spectra of guanine and cytosine has been known for a long time [60, 61] as well as its explanation by keto-enolic and amino-imino transformations in guanine and cytosine, correspondingly. It should be noted that the mentioned transfers of proton ions between proton-donor and proton-acceptor groups N1 and O6 in guanine and N4 and N3 in cytosine are definitely performed with the help of the hydrated H₂O molecules; they form a kind of cyclic structures with the above groups in guanine and cytosine. Because of the self-congruent transition of protons from N1 of guanine and N4 of cytosine to a molecule of hydrated H₂O, reorientation of H₂O molecules takes place in both cases.

Thus, we can state that the specific energy of H bonds with two energy states and self-congruent transfer of protons in cyclic structures is a necessary and sufficient condition for keto-enol and amino-imino tautomeric transformations of polar organic molecules, particularly guanine, cytosine, and adenine in solution. Water molecules act as a mediator in the process of forming the joint electron-proton complementary complex between the proton-acceptor/donor groups of the solute molecules and those of the solvent. So, here there is no need for water molecules as mediators in the DNA duplex because a pair of complementary bases always forms a cyclic structure.

Double proton transfer (DPT) in DNA is demonstrated spectroscopically as bathochromic shifts, weak hypochromic effects, and small widening of the absorption band [27, 36]. We shall discuss the mechanism in detail, using GC pairs as an example for the following reasons:

1. the effect of H⁺ interactions with guanine and cytosine [41, 62], as well as the interaction between transition metal ions and guanine are easily observed in the UV spectra;

2. spontaneous mutations of a genome occur more often at GC pairs than at AT pairs [63–65];

3. guanine and cytosine are often populated with rare enol and imino forms [66–68];

4. GC pairs are far less resistant to tunneling transitions compared to AT pairs [69–71];

5. in the DNA duplex, the site of preferable binding of H⁺ and metal ions is the endocyclic N7 of guanine located in the major groove [31, 32].

In general, the keto-enol and amino-imine tautomeric transformations in GC pairs are conditioned by the electric charges on G-N1 and C-N3 of endocyclic nitrogens, which play the principal role in H bonding of the base pairs. Figure 25 illustrates double-proton transfer in GC pair of DNA.

Figure 25

Double proton transfer (DPT) in a G-C part of DNA.

5.2 Interstrand crosslinks in the DNA double helix: ion sorption as a multistage adsorption process

Soft ions, in particular, Cu(I), Ag(I), Pt(II), Hg(II) ions are able to form the so-called interstrand crosslinks in DNA [60]. Let us consider the process on the example of Ag(I). First, silver ions are adsorbed on the DNA major groove (Guanine-N7 (G-N7) or chelate complex G-N7 and G-O6). At small silver ion concentration on DNA does not cause ejection of AO and EB [36]. On the other hand, silver ions at an interaction with DNA induce double proton transfer in the GC pair (see Section 5.1). Chelate complex with silver ions makes it easy to unfold the DNA double helix with the wrong Watson-Crick GC pair. In this process, guanine’s atom O6 is in the enol form, the nitrogen atom G-N1 is in the pyridine state and cytosine-N3 (C-N3) is in the pyrole state. After unfolding of the double helix in neutral water solution the C-N3 atom cannot keep the enol state for a long time, and it should transfer into its usual pyridine state. At the same time, silver ions can, with definite possibility, attack nitrogen atoms G-N1 still existing in the pyridine state. During the following folding of the double helix, inter-crosslink formation between G*-N1 and C-N3 takes place. In this way, the process of inter-crosslink formation can be considered as such a simple process as 1. Silver ion adsorption on DNA (G-N7) and double proton transfer of GC pair with the lifetime τ₁, 2. The unfolding of the double helix causes the formation of G*-N1- Ag⁺ binding, C-HN3 transfer to C-N3 and formation of link between G*-N₁-Ag+-C-N₃. The total time of the process is τ₂, 3. DNA folding with the formation of a stereoscopically distorted double helix with inter-crosslinks (τ₃). So, in the case of the DNA compound, the absorption process of inter-crosslink formation can be reduced to a multistage adsorption process consisting of several simple adsorption processes named above with the total time of τ₁+τ₂+τ₃. Scheme 1 illustrates it.

Scheme 1

The interstrand crosslink in DNA induced by silver ions. 1. Regular Watson-Crick pair. 2. Wrong Watson-Crick pair. 3. Open pair. 4. In the open wrong pair, silver ions start to transfer from N7 to N1. 5. In the open pair, the cyclic water molecules system makes proton transfer from N3 to N4 in cytosine. 6. The interstrand crosslink.

5.4 Phosphordiester linkage hydrolysis in DNA: photodynamic effect and the Jablonski diagram

The hydrolysis reaction requires the presence of H⁺ and OH^- ions, the concentration of which in water at pH=7 (natural pH for living cell) is 10^-7mol/l, i.e., the electric dissociation constant is 10^-14. Naturally, at such concentrations, the probability to hydrolysis is very small. In vivo phosphordiester link hydrolysis in DNA and RNA is carried out by nuclease ferments (DNase and RNase), and in vitro photoenergy dissipation in solution acts as a catalyst, first of all, in phosphordiester and glycoside link hydrolysis reactions in Figure 26. It cannot be excluded that in discoloration and degradation of aromatic compounds, the hydrolysis reaction is also important. A lifetime (τ) of exited AO singlet state is ∼10^-9 s (τ_AO=1.5×10^-9, τ_AO-DNA=5.2×10^-9 s [72]), the possibility that the oxygen molecule ([O₂]=1.3×10^-6 mol/l at T=298 K, see Table 2) can collide with the AO molecule is too small. At the same time, the triplet excited state is unusual for AO molecule at room temperature [73, 74].

Figure 26

Segment of polynucleotide DNA change 1- glycoside link, 2- and 3- phosphordiester links.

So, in the case of the DNA complex with AO at photodynamic effect in solutions, the principal oxidant is not an oxygen molecule but H⁺ ions. It is also connected to the fact, which we have shown, that for H⁺ ions, the mobile adsorption state of DNA is typical. It should be noted that at the photodynamic effect, the H⁺ local concentration depends on electrolytic dissociation of the water molecule. So, at photon absorption by chromophore, a part of energy dissipates into heat as a result of conversion, which we consider as energy transfer from electron vibration levels of organic molecules (in our case AO) to water molecule vibration levels (see IRET in Figure 27), and it should increase the constant of electrolytic dissociation. For instance, a 373 K temperature rise causes increase in water dissociation constant by three orders (Arrhenius equation). It is interesting how many H⁺ and OH^- ions one photon (λ=457 nm) can generate if it totally dissipates into heat. As electrolytic dissociation reaction activation energy requires 81.6 kJ/mol, the energy of the absorbed photon is enough for three water molecules to undergo electrolytic dissociation. In the place where the AO exited molecule undergoes total energy dissipation, we can get high local concentration of H⁺ and OH^- ions in immediate proximity from DNA and, thus, the possibility to hydrolysis both phosphordiester and glycoside is significant. In particular, phosphor-di-ester links a) P-O₍₃′₎ and b) O₍₅′₎-P can be presented as two schemes.

Figure 27

Scheme of electron-vibration-rotation levels of intramolecule transitions in compound organic molecules and singlet-singlet energy transfer between donor and acceptor.

6 Conclusions

Using spectrophotometry and thermodynamic approaches, we have shown that 1) on interaction with DNA, silver nanoparticles with the size 1–2 nm (AgNPs) are adsorbed, and only partial corrosion of nanoparticles at the level of Ag⁺ ions is observed; 2) AA reduces silver ions in ternary complex DNA-AgNO₃-AgNPs. Thus, AgNPs activate the process of quick reduction of Ag⁺ ions to silver atoms; 3) at photo-irradiation (λ=436 nm or full spectrum of visible band) desorption of silver atoms from the surface of AgNPs takes place. The atoms are first adsorbed on the surface of DNA and then penetrate inside the double helix (crosslinks between complementary DNA base pairs) making a prolate stretched structure (AgNPs absorption spectrum width is changed from 140 nm to 360 nm at half-height); 4) the kinetic study of photodesorption makes it possible to determine the desorption rate constant k_d and adsorption heat Q_a that are equal to k_d≅9×10^-5 s^-1; Q_a≥80 kJ/mol Ag⁰ for AgNPs bound with DNA; 5) AgNPs represent liquid drops, which moisturize the DNA surface at interaction. At photo-irradiation of the AgNPs-DNA complex, DNA-dependent conformational transition takes place due to fast and intensive heating.

The nano-scale method of laser-induced fluorescence resonance energy transfer (FRET) to donor-acceptor intercalator pair for quantitative and qualitative study of stability quality, the DNA double helix in solution in real time is offered. The approach is based on the example of the acridine orange molecule (donor) and ethidium bromide (acceptor) intercalated in DNA.

It is shown that ions Cu(II), Cu(I), Ag(I), and AgNPs, laser irradiation of AO, and the effect of heating decrease the concentration of undamaged areas of DNA double helix, i.e., the sites able to intercalate dye molecules such as AO and EB.

FRET radii were experimentally estimated in the background electrolyte solution (0.01 mol/l NaNO₃) and proved to be 3.9±0.3 nm, and the data are in satisfactory agreement with the theoretically calculated value R₀=3.5±0.3 nm.

The FRET method allows to estimate the concentration of double helix areas with high quality stability applicable for intercalation in DNA after it was subjected to a stress effect. It gives the opportunity to compare various types of DNAs with 1) different origins; 2) various damage degrees; 3) and in various functional states.

Alternative models and mechanisms of photodynamic effect on DNA in solutions are proposed. They are based on photo energy degradation in solutions. The energy activates electrolytic dissociation of water molecules on H₃O⁺ and OH^- and acts as a catalyst for hydrolysis reactions of phosphor-di-ester and glycoside linkages.

7 Experimental part