International Journal of Biological Macromolecules 67 (2014) 228–237

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac

Probing the binding mode of psoralen to calf thymus DNA Xiaoyue Zhou, Guowen Zhang ∗ , Langhong Wang State Key Laboratory of Food Science and Technology, Nanchang University, No. 235, Nanjing East Road, Nanchang 330047, China

a r t i c l e

i n f o

Article history: Received 20 January 2014 Received in revised form 13 March 2014 Accepted 14 March 2014 Available online 28 March 2014 Keywords: Psoralen Calf thymus DNA Intercalation Molecular modeling Multivariate curve resolution–alternating least squares (MCR–ALS)

a b s t r a c t The binding properties between psoralen (PSO) and calf thymus DNA (ctDNA) were predicted by molecular docking, and then determined with the use of UV–vis absorption, fluorescence, circular dichroism (CD) and Fourier transform infrared (FT-IR) spectroscopy, coupled with DNA melting and viscosity measurements. The data matrix obtained from UV–vis spectra was resolved by multivariate curve resolution–alternating least squares (MCR–ALS) approach. The pure spectra and the equilibrium concentration profiles for PSO, ctDNA and PSO–ctDNA complex extracted from the highly overlapping composite response were obtained simultaneously to evaluate the PSO–ctDNA interaction. The intercalation mode of PSO binding to ctDNA was supported by the results from the melting studies, viscosity measurements, iodide quenching and fluorescence polarization experiments, competitive binding investigations and CD analysis. The molecular docking prediction showed that the specific binding most likely occurred between PSO and adenine bases of ctDNA. FT-IR spectra studies further confirmed that PSO preferentially bound to adenine bases, and this binding decreased right-handed helicity of ctDNA and enhanced the degree of base stacking with the preservation of native B-conformation. The calculated thermodynamic parameters indicated that hydrogen bonds and van der Waals forces played a major role in the binding process. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Deoxyribonucleic acid (DNA) is an important genetic substance in the organism and a significant component of cell. It not only carries and expresses the hereditary information but also decides the type and function of cells. It plays a decisive role in growth, breeding, heredity, variation and transformation and a series of life phenomena [1]. DNA is a common intracellular target for antiviral, antitumor, anticancer and antibiotic drugs, therefore to explore the binding mechanisms of drugs with DNA for guiding the rational design and construction of new and more efficient drugs targeted to DNA is of great importance [2]. Drugs can interact with DNA through the following three non-covalent modes: (i) intercalation between the base pairs; (ii) interactions with DNA groove; and (iii) electrostatic attractions with the anionic sugar–phosphate backbone of DNA [3]. Intercalative binding and groove binding are related to the grooves in DNA double helix, while the electrostatic binding takes place out of the groove. Psoralen (PSO, structure shown in Fig. 1) is known as a furocoumarin, which is naturally occurring or synthetic tricyclic aromatic compound deriving from the condensation of a coumarine

∗ Corresponding author. Tel.: +86 79188305234; fax: +86 79188304347. E-mail address: [email protected] (G. Zhang). http://dx.doi.org/10.1016/j.ijbiomac.2014.03.038 0141-8130/© 2014 Elsevier B.V. All rights reserved.

nucleus with a furan ring [4]. It is the main active ingredient extracted from the fruits of Psoralea corylifolia L. [5]. PSO exhibits many therapeutic effects, for example, promoting MCF-7 cell proliferation significantly as ER␣ agonists [6], showing multidrug resistance effect on leukemia cell [7], acting through the activation of BMP signaling to promote osteoblast differentiation [5], etc. Also, it was reported that derivates of PSO could cause a linkage to DNA which lead to some changes in DNA properties [8,9]. In recent years, some studies have shown that PSO exerts activity in preventing proliferation of bladder carcinoma, mucoepidermoid carcinoma and mammary cancer cells in vitro [10]. Owing to a variety of biological and pharmacological activity of PSO, it is necessary to clarify the interaction mechanism between PSO and DNA. Various techniques have been adopted to investigate the interaction of small molecules with DNA, including fluorescence spectroscopy, electrochemistry, UV–vis absorption [11], circular dichroism (CD), Fourier transform infrared (FT-IR), nuclear magnetic resonance (NMR) [12], and atomic force microscopy (AFM) [13]. Though spectroscopic techniques possess the advantages of reproducibility, selectivity and convenience among these methods, it is difficult to obtain more information about the composition in complex system of more than two components due to the overlap of their response signals. Therefore, a chemometrics method, multivariate curve resolution–alternating least squares (MCR–ALS), has been used to analyze the overlapping response signals in recent years because it

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

13

O

O

O1

Fig. 1. Molecular structure of psoralen (PSO).

can extract simultaneously the concentration information and the pure spectra of all species involved in interaction [14–16]. In this work, the molecular docking was employed to predict the probable binding site and binding mode of PSO with ctDNA, and then multispectroscopic methods including UV–vis absorption, fluorescence, CD and FT-IR spectroscopy along with DNA melting and viscosity measurements were used to determine the interaction between PSO and ctDNA at physiological buffer (pH 7.4). Furthermore, the MCR–ALS method was applied to decompose the expanded UV–vis absorption spectral data collected from the PSO–ctDNA mixtures, and then the corresponding pure spectra of each component and their concentration profiles were simultaneously extracted from composite responses to evaluate the interaction process of PSO with ctDNA. 2. Experimental 2.1. Chemicals PSO (analytical grade) was obtained from the National Institute for the Control of Pharmaceutical Biological Products (Beijing, China). The stock solution (4.94 × 10−3 mol L−1 ) of PSO was prepared in absolute ethanol. The ctDNA was purchased from Sigma Chemical Co., and it was dissolved in 0.10 mol L−1 NaCl solution to get ctDNA solution. As the ratio of absorbance at 260 and 280 nm (A260 /A280 ) was 1.86, the ctDNA solution was deemed to be free from protein sufficiently [17]. Also, the concentration of ctDNA solution was determined to be 3.61 × 10−3 mol L−1 by UV absorption at 260 nm using the molar absorption coefficient ε260 = 6600 L mol−1 cm−1 [18]. All other reagents and solvents were of analytical reagent grade, and ultrapure water was used throughout the experiment. All the solutions used in the experiments were kept in cold storage at 0–4 ◦ C. 2.2. Procedures 2.2.1. Molecular docking studies The molecular modeling investigation was aimed at predicting the binding mode and binding site of PSO with ctDNA. Docking studies were carried out with the use of Autodock (version 4.2) software. The structure of 24 bp long DNA used for docking was acquired from the Protein Data Bank with identifier 453D [19]. Then the structure of macromolecule was optimized for docking by adding polar hydrogen atoms and Gasteiger charges. The 3D structure of the ligand PSO was generated in Sybyl × 1.1 (Tripos Inc., St. Louis, USA) and its conformation was energy-minimized using MMFF94 force field. Rotatable bonds in the ligands were assigned with AutoDock Tools and docking carried out by the AutoDock 4.2 Lamarckian Genetic Algorithm (LGA) [20]. DNA was enclosed in a grid possessing 0.375 A˚ spacing and PSO molecule was allowed to move within the whole region. The output from AutoDock was rendered with PyMol. 2.2.2. UV–vis absorption measurements UV–vis absorption spectra measurements were conducted to investigate the interaction of PSO with ctDNA over a wavelength range of 215–390 nm, and two different experiments were performed in pH 7.4 Tris–HCl buffer at room temperature.

229

For experiment 1: the concentration of PSO constant was kept at 1.98 × 10−5 mol L−1 , and different amounts of ctDNA (0–1.62 × 10−4 mol L−1 in increment of 6.01 × 10−6 mol L−1 , total 27 solutions) were added successively to the solution. For experiment 2: the concentration of ctDNA was kept at 7.22 × 10−5 mol L−1 , and PSO was added into the solution in increment of 8.23 × 10−7 mol L−1 (total 27 solutions) with a final concentration of 2.22 × 10−5 mol L−1 . All the solution samples were allowed to stand for 3 min to equilibrate after addition, and then the UV–vis absorption spectra were collected every 1 nm on a Shimadzu UV-2450 spectrophotometer (Shimadzu, Japan). Thus, two data matrices DPSO (27 × 176) and DctDNA (27 × 176), were obtained from these measurements, and column-wise expanded data matrix for the two experiments were constructed. 2.2.3. Fluorescence measurements A quantitative analysis of the interaction between PSO and ctDNA was carried out by fluorimetric titration with a Hitachi spectrofluorometer model F-7000 (Hitachi, Japan) equipped with a 150 W xenon lamp. The diluent PSO solution (1.21 × 10−5 mol L−1 ) was pipetted to a 1.0 cm quartz cuvette, and then successive titrated by the 3.61 × 10−3 mol L−1 ctDNA solution to give a final concentration of 1.20 × 10−4 mol L−1 . After equilibrating for 3 min, the fluorescence spectra of the solution were recorded at four temperatures (292, 298, 304 and 310 K) in the wavelength range of 370–550 nm with an exciting wavelength at 280 nm. Both the excitation and emission bandwidths were set at 5.0 nm. The background of fluorescence was corrected by subtracting appropriate blanks of the Tris–HCl buffer. The fluorescence data were corrected for absorption of excitation light and emitted light to eliminate the re-absorption and inner filter effect caused by UV–vis absorption according to the relationship [21]. Fc = Fm e(A1 +A2 )/2

(1)

where Fc and Fm represent the corrected and measured fluorescence, respectively. A1 and A2 are the absorbance of the ctDNA solutions at excitation and emission wavelengths, respectively. Iodide quenching effects were compared according to the quenching constants calculated from fluorescence data of titrating KI to PSO and PSO–ctDNA complex solutions. Fluorescence polarization was measured by keeping the concentration of PSO at 1.21 × 10−5 mol L−1 while varying ctDNA concentrations from 0 to 1.20 × 10−4 mol L−1 . After setting a pair of polarizers, the samples were performed at the corresponding excitation and emission wavelengths: 280 and 463 nm, respectively. 2.2.4. DNA melting studies DNA melting experiments were conducted by determining the absorption of ctDNA at 260 nm in the absence and presence of PSO at temperatures varying from 20 to 100 ◦ C with an interval of 4 ◦ C. The transition midpoints of the curves of fss = (A − A0 )/(Af − A0 ) versus temperature (T) were regarded as the melting temperatures (Tm ) of ctDNA and PSO–ctDNA complex, where A0 and Af represent the absorbance intensities at 20 and 100 ◦ C, respectively, and A is the absorbance intensity of the corresponding temperature [22]. 2.2.5. Viscosity measurements Viscometric titrations were performed on an Ubbelohde viscometer (˚ 0.7–0.8 mm, Shanghai Qianfeng Rubber and Glass Co., Shanghai, China), which was kept at 25 ± 0.1 ◦ C by a constant temperature bath. Each flow time of the solutions was measured with a digital stopwatch for five times after being added appropriate amounts of PSO to give a certain r (r = [PSO]/[ctDNA]) value while the ctDNA concentration was kept in constant. The average times

230

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

(t) of five replicated measurements were used to evaluate the viscosity of the samples according to  = (t − t0 )/t0 , t and t0 of which are average flow times of ctDNA containing solutions and corrected buffer solution, respectively. The data was presented as (/0 )1/3 versus the [PSO]/[ctDNA] ratio, where 0 and  represent the viscosity of pure ctDNA and that of ctDNA with different amount of PSO, respectively [23]. 2.2.6. Competitive binding investigations using methylene blue (MB) and ethidium bromide (EB) as probes The competitive experiments were carried out using the intercalators MB and EB as fluorescence probes. First, a fixed concentration of the ctDNA–MB complex solution was titrated by successive additions of 4.94 × 10−3 mol L−1 PSO stock solution. After standing 3 min to equilibrate, the fluorescence spectra were recorded over the wavelength range of 650–800 nm with an excitation wavelength at 630 nm. In addition, the ctDNA solution in the absence and presence of different amount of PSO was progressively titrated with EB. The fluorescence spectra were measured in the wavelength range of 550–750 nm after being excited with the wavelength at 525 nm. All the competitive experiments were conducted at 298 K. 2.2.7. CD spectra studies CD measurements of ctDNA and the PSO–ctDNA complex were made on a Bio-Logic MOS 450 CD spectrometer (Bio-Logic, France) at wavelengths between 220 and 320 nm under steady nitrogen flush. The CD spectra of ctDNA in the absent and present of PSO were recorded in pH 7.4 Tris–HCl buffer at room temperature and corrected for the signal of buffer solution. 2.2.8. FT-IR spectroscopic measurements The infrared spectra were measured by a FT-IR spectrometer (Thermo Nicolet-5700, USA) equipped with a germanium attenuated total reflection (ATR) accessory, a DTGS KBr detector and a KBr beam splitter after 2 h incubation of PSO with the ctDNA solution and recorded via the ATR method over the spectral range 1800–800 cm−1 with a resolution of 4 cm−1 and 64 scans. To remove the spectral features of water to get a flat baseline around 2200 cm−1 , the H2 O solution at pH 7.0 ± 0.2 was chosen as a reference to complete water subtraction. The band at 968 cm−1 in difference spectra [(ctDNA solution + PSO solution) − PSO solution], which is due to deoxyribose C C stretching vibrations, exhibited no spectral changes upon PSO–ctDNA complex and was used as an internal reference. The plots of the relative intensity (R) of several peaks of DNA caused by in-plane vibrations of base pairs and the stretching vibrations of PO2 versus the PSO concentrations were obtained after peak normalization using R = Ii /I968 , where Ii is the intensity of absorption peak at i cm−1 for pure ctDNA and its complex with different concentration of PSO, and I968 is the intensity of the 968 cm−1 peak (internal reference band). 2.2.9. Chemometrics method: Multivariate curve resolution–alternating least squares (MCR–ALS) Multivariate curve resolution (MCR) is used for purpose of mathematically decomposing an instrumental response such as spectrum of mixture into the pure contributions of each constituent existed in the system [15]. Data of spectra monitored during a chemical process can be arranged in a matrix D (NR × NC ), where NR is the number of spectra recorded throughout the process and NC is the instrumental responses measured at each wavelength. The MCR decomposition of matrix D is carried out according to the multivariate extension of the Beer’s law: D = CS T + E

(2) ST

(NX × NC ) where the columns in the C (NR × NX ) and the rows in are data matrices containing concentration profiles and pure

component spectra present in the experiment, respectively. E (NR × NC ) is the residual matrix with data variance which is unexplained by the product CST . The number of species, NX , is directly related with the number of main components in matrix D. The number N is estimated by rank analysis with the use of singular value decomposition (SVD) [24]. The rank of the matrix calculated by the method is assumed to be the number of chemical species in the system. The concentration profiles were obtained by initial estimating for the concentration matrix via evolving factor analysis (EFA). As described above, two kinds of experiments were performed under different conditions: one with the constant concentration of PSO (cPSO ) and the other with the constant concentration of ctDNA (cctDNA ). The data arrangement evaluated by the extended MCR–ALS analysis can be expressed as follows:



DPSO DctDNA





=

C PSO C ctDNA





T

×S +

E PSO E ctDNA



(3)

where [DPSO ; DctDNA ] is the column-wise augmented matrix obtained by formatting one individual matrix above the other. The data matrices DPSO and DctDNA correspond to the two different experiments 1 and 2, respectively. The new column-wise augmented data matrix [CPSO ; CctDNA ] is the corresponding augmented column-wise concentration matrix, and ST is the data matrix of the recovered pure spectra for each species. The last term is the residual variance. With the purpose of resolving the complex spectra to obtain the pure spectra and the concentration profiles of all molecular components, each column of this amplified spectral data matrices were combined and submitted to MCR–ALS analysis. 3. Results and discussion 3.1. Molecular docking studies As the number of biological structures in data banks rapidly increases, molecular docking is becoming an important approach to predict, evaluate or even to elucidate the interaction between potential ligands and macromolecular targets. DNA has been a familiar molecular target for a wide range of anticancer and antitumor drugs [25]. In this study, a total of 100 runs were performed at the end of each autodock execution, which resulted in the generation of a list of clusters and their energies to infer the most probable binding site of PSO to ctDNA. Cluster profile in Fig. 2A was performed using a rmsd tolerance ´˚ Multimember conformational clusters obtained from 100 of 2.0 A. docking runs were in total of 29. Each cluster presents the occurrence number of different binding domains, and the binding energy corresponds to the lowest energy among all of the binding gestures in each domain. The ligand is most likely to bind to the domain possessed 17 binding gestures with highest occurrences (the red bar in Fig. 2A). Considered from energetic perspective, the most energetically favorable binding gesture should be the one with binding energy of −4.72 kcal mol−1 . As the ligand has no rotatable atom, the torsional energy was 0 kcal mol−1 , and its binding energy was equal to the intermolecular energy, thus the predict free energy (Gbinding = intermolecular energy + torsional energy) was worked out to be −4.72 kcal mol−1 . Fig. 2B shows the probable binding mode of the ligand PSO to ctDNA. It was obvious that the whole planar molecular intercalated into the intermediate space of the two adjacent base pairs (DA6–DT19 and DT7–DA18). Furthermore, there were two hydrogen bonds between PSO and adenine bases (Fig. 2B, green line). As shown in Fig. 2C, hydrogen bond 1 was formed between oxygen atom O13 of PSO and hydrogen atom H61 associated with N6 (DA6 of A chain), and hydrogen bond 2 was constructed between oxygen

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

231

(A) 1.5

Absorbance

1.2

246nm259nm

0.9 28 0.6

1

0.3

0.0 215

250

285

320

355

390

355

390

Wavelength/nm

(B) 1.0

0.8

Absorbance

246nm259nm 28

0.6

1

0.4

0.2

0.0 215

250

285 320 Wavelength/nm

Fig. 3. Spectra obtained from different experiments. (A) Experiment 1: absorption data matrix DPSO , the concentration of PSO was 1.98 × 10−5 mol L−1 , and ctDNA was added at different concentrations (0–1.62 × 10−4 mol L−1 in increment of 6.01 × 10−6 mol L−1 ) for curves 1–28, respectively; (B) experiment 2: absorption data matrix DctDNA , the concentration of ctDNA was kept at 7.22 × 10−5 mol L−1 and PSO was added into the solution in increment of 8.23 × 10−7 mol L−1 with a final concentration of 2.22 × 10−5 mol L−1 for curves 1–28, respectively.

Fig. 2. (A) Cluster analyses of the AutoDock docking runs of PSO with DNA; (B) most probable docking mode of PSO interaction with DNA (possible hydrogen bonds shown as green). (C) Possible hydrogen bonds (dashed line) between PSO and bases. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

atom O1 of PSO and hydrogen atom H62 related to N6 (DA18 of B chain). At the same time, hydrogen bond forming of the 100 different simulated models was statistical analyzed to find that the gestures with hydrogen bond accounted for 85%, including 63% with adenines. These results portended that the hydrogen bonds may play an important role in the intercalative binding of PSO to ctDNA, and adenine bases were the preferential binding site. 3.2. Absorption spectra of interaction between PSO and ctDNA The UV spectrometry is one of the most common methods to detect the interaction of small molecules with DNA. Fig. 3A shows the UV spectra of PSO in the absence and presence of

different concentrations of ctDNA which corresponds to experiment 1. PSO exhibited two absorption bands at around 246 and 259 nm, respectively. The absorption intensity of PSO at 246 increased and presented a 2 nm red shift with the increasing amounts of ctDNA. The UV absorption spectrum of ctDNA has a peak at about 259 nm (Fig. 3B, curve 1), and the peak ascended with the increasing addition of PSO. It can be found that it is difficult to judge the formation of a complex and monitor the progress of PSO interacting with ctDNA in the system with the serious overlapping absorption spectra of the mixture. Hence, MCR–ALS approach was introduced to resolve the complicated absorption spectra and extract the equilibrium concentration profiles and pure spectra of the reacting species from the spectral data. 3.3. Application of MCR–ALS to the two-way absorption spectra The two different UV–vis absorption profile matrices were combined into an expanded data matrix, which was submitted to be resolved by the MCR–ALS method in MATLAB 6.5 software. To ignore variation below a particular threshold to reduce the data with the assurance that the main relationships have been preserved, the number of components could be determined by the

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

(A)

Absorbance

0.6

PSO—ctDNA

0.4

PSO

0.2 DNA

0.0 215

250

285

355

390

(B) 2.0 1.6 PSO

1.2 0.8 PSO- ctDNA

0.4 ctDNA

0.0 0

2

4

6

8

[ctDNA]:[PSO]

3.4. Analysis of fluorescence quenching of PSO by ctDNA

(4)

where F0 and F represent the fluorescence intensities of PSO in the absence and presence of ctDNA, respectively. KSV is the Stern–Volmer dynamic quenching constant which was determined by linear regression of a plot of F0 /F against [Q]. Kq denotes the quenching rate constant of the bimolecular and the value of the maximum scattering collision quenching constant is 2.0 × 1010 L mol−1 s−1 [31].  0 refers to the average lifetime of the fluorophore in the absence of quencher with a constant value of 10−8 s [32], and [Q] represents the concentration of ctDNA. As shown in Fig. 5B, the good linearity of the curves of F0 /F versus [Q] at four different temperatures (292, 298, 304, and

(C) 7.5 -5

Relative Concentration (10 )

Fluorescence quenching studies can be used to detect the interactions for determining the accessibility of a fluorophore to an added quenching agent and get some binding information of small molecule substances to macromolecules on the molecular level, such as the binding mechanism, binding mode, binding constant, intermolecular distances, etc. [29]. PSO exhibited a strong fluorescence emission peak at 463 nm after being excited with a wavelength of 280 nm (Fig. 5A). It was evident that the fluorescence intensity of PSO was remarkably quenched by increasing the concentration of ctDNA, but the maximum emission wavelength did not apparently shift. This phenomenon indicated that ctDNA interacted with PSO and quenched the intrinsic fluorescence of PSO. In general, fluorescence quenching processes were divided into dynamic quenching and static quenching. In order to ascertain the quenching process between PSO and ctDNA, the fluorescence quenching data were analyzed using the Stern–Volmer equation [30]: F0 = 1 + KSV [Q] = 1 + Kq 0 [Q] F

320

Wavelength/nm

-5

SVD method [26]. The first four Eigen values were evaluated to be 112.20, 19.95, 14.45 and 1.205, which indicated that there were only three main factors in the system, that is, PSO, ctDNA and PSO–ctDNA complex. Then, the pure spectra and concentrations for the three postulated species, which were not easily got by conventional methods, were extracted by the MCR–ALS analysis from the seriously overlapped spectra. The constraints applied were nonnegativity for both the concentration and spectral profiles. The pure spectra of PSO and ctDNA were applied as known spectra as equality constraint for spectral profiles. In addition, as the concentrations of PSO and/or ctDNA were known in the experiments, they were included as a closure constraint for the concentration profiles [26,27]. The extracted spectra (solid line, Fig. 4A) of the free PSO and ctDNA were in accord with their measured spectra (dashed line, Fig. 4A), suggesting that the concentration profiles of the three reacting species were correctly resolved [28]. In addition, the spectrum of the PSO–ctDNA complex was successful depicted. The EFA method can be used to provide the initial indication of the changes in the concentration profiles of the components in the mixture. The concentration profiles got from MCR–ALS at constant concentration of PSO or ctDNA (corresponding to experiment 1 and 2, respectively) are presented in Fig. 4B and C, respectively. From Fig. 4B, it was found that the concentration of PSO–ctDNA complex increased gradually accompanied by the decrease of PSO concentration after adding different amounts of ctDNA. Similarly, Fig. 4C displays an increasing tendency of the concentration of the PSO–ctDNA complex and a decreasing tendency of the ctDNA concentration with increasing amounts of PSO. These results were powerful evidences to prove that PSO could interact with ctDNA and form the PSO–ctDNA complex.

Relative Concentration (10 )

232

6.0 ctDNA

4.5 3.0 PSO

1.5

PSO— ctDNA

0.0 0.00

0.07

0.14

0.21

0.28

0.35

[PSO]:[ctDNA] Fig. 4. Results of the simultaneous analysis of the UV–vis absorption data of the two experiments. (A) Recovered UV–vis absorption spectra for PSO, ctDNA and PSO–ctDNA complex; solid line: resolved spectra from MCR–ALS; dashed line: measured spectra in the experiments; (B) equilibrium concentration profiles for experiment 1 corresponding to C PSO ; (C) equilibrium concentration profiles for experiment 2 corresponding to C ctDNA .

310 K) suggested that either static or dynamic quenching existed in PSO–ctDNA interaction. The values of KSV at different temperatures are listed in Table 1. The results showed that the values of KSV decreased with the increasing temperature, indicating that the quenching mechanism of the PSO–ctDNA binding reaction probably arose from compound formation rather than by dynamic collision [33], which was in accordance with the result of MCR–ALS analysis.

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

was smaller than that of some typical intercalators, ethidium bromide (6.58 × 104 L mol−1 ), acridine orange (2.69 × 104 L mol−1 ) and methylene blue (2.13 × 104 L mol−1 ) [35], but accorded with some intercalative compounds such as triadimenol (9.06 × 103 L mol−1 ) [36], clodinafop-propargyl (5.66 × 103 L mol−1 ) [37] and naringenin (4.19 × 103 L mol−1 ) [38]. This may be due to the different molecular planarity of the intercalators and the differences in the surrounding environment of inserting sites.

(A)

Fluorescence Intensity

520

1 390

11 260

3.5. Determination of thermodynamic parameters The four classes of noncovalent interactions forces which can work in the binding of small molecules to biomolecules are hydrogen bonds, van der Waals forces, hydrophobic interactions, and electrostatic forces [39]. The thermodynamic parameters of binding reaction, enthalpy change (H◦ ) and entropy change (S◦ ) are the major evidence for identifying the binding force. H◦ can be taken as a constant if there is no distinct change in temperature, and then its value and that of S◦ can be calculated from the van’t Hoff equation [40] as follows:

130

0 370

415

460

505

550

Wavelength/nm (B) 1.5

log Ka = −

292K 298K 304K 310K

1.4

F0/F

233

1.3

1.0 6

-5

9

-1

12

(7)

where R is the gas constant and the temperatures used in this experiment were 292, 298, 304 and 310 K. The slope and intercept of linear plot of log Ka versus 1/T allows the determination of H◦ and S◦ . Then the value of free energy change (G◦ ) was obtained from Eq. (7). The thermodynamic parameters for the interaction of PSO with ctDNA are shown in Table 1. The values of H◦ and S◦ were −31.47 kJ mol−1 and −29.77 J mol−1 K−1 , respectively, indicating that hydrogen bonds and van der Waals forces played a major role in the binding of PSO to ctDNA [41,42]. The process of PSO binding to ctDNA is spontaneous due to a negative value of G◦ . The value (−5.32 kcal mol−1 ) of G◦ in 310 K was close to the predictive free energy in docking (−4.72 kcal mol−1 ), and the little difference of the results could be attributed to the lack of desolvation energy in the simulated vacuum environment of Autodock. This is another proof that can affirm the reliability of the molecular docking result.

1.1

3

(6)

G◦ = H ◦ − TS ◦

1.2

0

H ◦ S ◦ + 2.303RT 2.303R

15

[ctDNA]/(10 mol L ) Fig. 5. (A) Fluorescence spectra of PSO in the absence and presence of ctDNA (pH 7.4, T = 298 K, ex = 280 nm, em = 463 nm). c(PSO) = 1.21 × 10−5 mol L−1 , and c(ctDNA) = 0, 0.12, 0.24, 0.36, 0.48,0.60, 0.72, 0.84, 0.96, 1.08 and 1.20 × 10−4 mol L−1 corresponding to the curves from 1 to 11, respectively; (B) the Stern–Volmer plots for the fluorescence quenching of PSO by ctDNA at four different temperatures.

The binding constants were calculated by the modified Stern–Volmer equation [34]:

3.6. Iodide quenching experiments

F0 1 1 1 + = F0 − F fa Ka [Q] fa

The interaction mode of PSO to ctDNA could be obtained through iodide quenching experiments. Intercalative bound molecule should be protected from being quenched by anionic quencher because a highly negatively charged quencher will be repelled by the negatively charged phosphate backbone of DNA. On the contrary, the free complexes or groove binding molecules should be quenched readily by anionic quenchers [43]. The quenching data were analyzed according to the Stern–Volmer equation, the calculated KSV of bound PSO (1.47 × 102 L mol−1 ) quenched by iodide was lower than that of free PSO (3.01 × 102 L mol−1 ) (shown in Fig. 6A). This difference of iodide quenching effect between the PSO in the absence and presence of ctDNA could make a preliminary proof

(5)

where Ka , fa denote the modified Stern–Volmer association constant for the accessible fluorophores and the fraction of accessible fluorescence, respectively. The values of Ka at different temperatures were determined by a linear regression of F0 /(F0 − F) versus 1/[Q]. It can be seen from Table 1 that the decreasing trend of Ka with increasing temperature was consistent with KSV ’s dependence on temperature as discussed above, this result was a characteristic that coincides with the type of static quenching [21]. The Ka value of the PSO–ctDNA interaction was 9.74 × 103 L mol−1 at 298 K, which

Table 1 The quenching constants (KSV ), association constants (Ka ), and relative thermodynamic parameters for the interaction of PSO with ctDNA at different temperatures. T (K)

KSV (L mol−1 )

Ra

Ka (L mol−1 )

Rb

292 298 304 310

2.66 × 103 1.99 × 103 1.62 × 103 1.20 × 103

0.9994 0.9990 0.9978 0.9972

11.46 × 103 9.74 × 103 6.80 × 103 5.67 × 103

0.9944 0.9962 0.9955 0.9963

a b

Correlation coefficient for the KSV values. Correlation coefficient for the Ka values.

H◦ (kJ mol−1 ) −31.47

S◦ (J mol−1 ) −29.77

G◦ (kJ mol−1 ) −22.77 −22.59 −22.42 −22.24

234

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

(A)

2.0 PSO PSO— ctDNA

F0/ F

1.8 1.6 1.4 1.2 1.0 0.0

0.5

1.0

1.5

2.0

-3

2.5

3.0

-1

[KI]/(10 mol L ) (B) 0.28

Polarization

0.26 0.24 0.22 0.20 0.18 0

3

6

9 -5

12

15

-1

[ctDNA]/(10 mol L ) Fig. 6. (A) The Stern–Volmer plots for the quenching of PSO by KI in the absence and presence of ctDNA at pH 7.4 and room temperature. c(PSO) = 1.21 × 10−5 mol L−1 , and c(ctDNA) = 3.61 × 10−5 moL L−1 . (B) Effect of the ctDNA concentration on fluorescence polarization of PSO at pH 7.4 and room temperature (ex = 280 nm, em = 463 nm). c(PSO) = 1.21 × 10−5 mol L−1 .

Fig. 7. (A) Melting curves of ctDNA in the absence and presence of PSO at pH 7.4. c(PSO) = 8.89 × 10−6 mol L−1 , and c(ctDNA) = 4.88 × 10−5 mol L−1 . (B) Effect of increasing amounts of PSO on the relative viscosity of ctDNA at pH 7.4. c(ctDNA) = 2.39 × 10−5 mol L−1 .

for the prediction by molecular docking that PSO interacted with ctDNA via intercalation.

the addition of ctDNA caused an increase in the fluorescence polarization of PSO, suggesting that PSO intercalated into the double helix of ctDNA.

3.7. Fluorescence polarization analysis

3.8. DNA melting studies

Fluorescence polarization is an effective parameter for investigating dynamic characteristics of PSO in different microenvironments. Due to the rapid tumbling motion in aqueous media, small molecules polarize weakly. However, if it intercalates to the helix of DNA, its rapid rotational motion would be restricted and embodied in the increase of fluorescence polarization. On the other hand, binding to the phosphate backbone or to the DNA grooves does not result in a significant enhancement of the fluorescence polarization [44]. Fluorescence polarization can be obtained from the following equation:

The melting temperature Tm , which is defined as the temperature at which half of the total base pairs are unwound, is determined from the midpoint of DNA melting curves. Generally, the melting temperature increases about 5–8 ◦ C when small molecule bind to DNA by intercalation, whereas the non-intercalation binding causes no visible enhancement in Tm , since intercalation of the molecule into DNA base pairs causes stabilization of base stacking and hence raises the melting temperature of the double-stranded DNA [45,46]. Fig. 7A shows the melting curves of ctDNA and the PSO–ctDNA complex, the Tm values of ctDNA in the absence and presence of PSO were 67.9 and 73.5 ◦ C, respectively, which indicated the interaction mode of PSO with ctDNA was an intercalation.

P=

IVV − GIVH IVV + GIVH

(8)

where IVV and IVH denote the emission intensities measured through vertically and horizontally oriented polarizer in the excitation and emission beam, respectively. G represents the instrument grating correction factor, G = IHV /IHH , where IHV is the vertical polarization intensity parallel to excitation and IHH is the horizontal polarization intensity parallel to excitation. As seen from Fig. 6B,

3.9. Viscosity measurements Since the viscosity is sensitive to length changes of the double helix of DNA, viscosity measurements are regarded as the most unambiguous and decisive test for providing reliable evidence of DNA binding mode [47]. Because of the increase in separation

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

1:0 1:0.5 1:1 1:2

2.10 × 10 0.80 × 105 0.54 × 105 0.33 × 105

0.062 0.060 0.053 0.042

of base pairs at intercalation sites, a classic intercalation could increase the viscosity of DNA solutions dramatically by the resulting increase in the overall DNA contour length. Nevertheless, ligand that bind in DNA grooves via partial and/or non-classic intercalation generally cause either a less obvious or no change at all in DNA viscosity [48]. Fig. 7B represents that the relative viscosity of ctDNA increased greatly upon the addition of PSO, which was similar to that caused by the typical intercalator ethidium bromide (EB) in the same range of [M]/[ctDNA] ratios. It may be due to lengthening of the DNA helix as base pairs separated to accommodate the aromatic chromophore of the bound molecule. This behavior further demonstrated that PSO bound to ctDNA through an intercalative mode. 3.10. Competitive binding of MB and EB with PSO for ctDNA MB was used as a fluorescence probe to investigate competitive fluorescence interactions of PSO with ctDNA. It was reported that the mode of MB interacting with DNA is different in various condition. When the [DNA]/[MB] ratio is low (10) [DNA]/[MB] ratio [49]. In the ratio of 10.54, if PSO intercalates into the helix of ctDNA, it would compete with MB for the interaction sites in ctDNA and dissociate MB from the ctDNA–MB complex, which leads to an increase in fluorescence intensity of the ctDNA–MB. The emission spectra of the ctDNA–MB complex in the absence and presence of PSO are shown in Fig. 8A. With the addition of PSO to the complex solution, the fluorescence intensity of ctDNA–MB increased approximately 47%. Besides, PSO could neither quench the free MB fluorescence nor produce new peaks. These results indicated that MB was replaced by PSO and released from the hydrophobic environment inside the ctDNA helix into the aqueous medium for the interaction between PSO and ctDNA, which is another indicative of the intercalative binding of PSO to ctDNA. It is well known that EB is a classical intercalator to DNA, which plays very important role in the research of interaction between small molecules and DNA. Thus, in this study, the fluorescence titration of EB with ctDNA in the presence of various amount of PSO was performed to provide more information about the intercalation. The fluorescence data were analyzed by the Scatchard equation r/Cf = KEB (n − r), where r is the molar ratio of bound EB to ctDNA (r = Cb /[ctDNA]), Cf is the molar concentration of free EB, n is the number of binding sites per nucleotide and KEB denotes the binding constant of EB with ctDNA [50,51]. The Scatchard plots are shown in Fig. 8B, and the values of KEB and n were obtained from the linear regression of r/Cf versus r (Table 2). It clearly showed that the number of binding sites per nucleotide n for EB decreased with the increasing amount of PSO, suggesting that the binding of PSO to ctDNA was so strong that bound PSO could not be easily displaced by added EB. What’s more, the KEB values gradually decreased when the molar ratio of ctDNA to PSO increased, which indicated that the binding between EB and ctDNA was weakened [51]. The experimental phenomena illustrated that the binding mode of PSO to ctDNA was intercalation, and the PSO–ctDNA binding decreased the binding ability of EB to ctDNA. The results from competitive

150

Fluorescence Intensity

n

5

120 9

90 1

60 30 0 650

680

710

740

770

800

Wavelength/nm (B) 12

-1

KEB (L mol−1 )

9

5

[ctDNA]/[PSO]

(A)

r/Cf / (10 mol L )

Table 2 Binding parameters of EB to ctDNA in the presence of various amount of PSO.

235

6

[ctDNA]/[PSO] 1:0 1:0.5 1:1 1:2

3

0 0.000

0.015

0.030

r

0.045

0.060

Fig. 8. (A) Fluorescence spectra of ctDNA–MB system in the presence of PSO at different concentrations. c(ctDNA) = 1.37 × 10−4 mol L−1 , c(MB) = 1.30 × 10−5 mol L−1 , and c(PSO) = 0, 0.16, 0.33, 0.49, 0.66, 0.82, 0.99, 1.15 and 1.32 × 10−4 mol L−1 corresponding to the curves from 1 to 9, respectively. (B) Scatchard plots for the binding of EB with ctDNA in the absence and presence of increasing the concentration of PSO. c(ctDNA) = 3.61 × 10−5 mol L−1 , c(EB) = 0, 0.33, 0.67, 1.00, 1.33, 1.67, 2.00, 2.33, 2.67, 3.00, 3.33 × 10−6 mol L−1 .

binding studies with two different intercalative probes have provided powerful evidence for the intercalation of PSO to ctDNA. 3.11. CD spectra studies CD is a useful technique to access whether DNA undergoes conformational changes when the environmental conditions change. In particular, B–DNA form shows two conservative CD bands in the UV region: the negative band at 245 nm due to the right-handed helicity and the positive band at 275 nm due to base stacking [52]. The CD spectra of ctDNA in the presence of increasing concentrations of PSO were recorded in Fig. 9. It was apparent that the intensity of the negative band decreased and the positive band increased with the increase in molar ratios of PSO to ctDNA. As intercalation induces DNA unwinding, a common assumption for rationalizing intercalation-induced DNA unwinding is that it enables the sugar–phosphodiester backbone to span the bound intercalator and still maintain the link between the two flanking base pairs [23]. Hence, the decrease in negative band correlated with helix unwinding caused by intercalative action of PSO. Another report has pointed that increase in the molar ellipticity of positive band is attributed to an intercalative mode, since the stacking interaction

236

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

of the intercalator with the base pairs leads to an enhancement of the positive band [53]. The CD spectra indicated that PSO intercalating into ctDNA helix decreased ctDNA right–handed helicity and enhanced the degree of base stacking [54], but the B conformation was not changed [55].

3 3

CD(medg)

1

1

PSO

3.12. FT-IR analysis

-1

3 1

-3 220

240

260 280 Wavelength/nm

300

320

Fig. 9. CD spectra of ctDNA in the presence of increasing amounts of PSO at pH 7.4 and room temperature. c(ctDNA) = 3.61 × 10−4 mol L−1 . The molar ratios of PSO/ctDNA were (1) 0, (2) 1/30, and (3) 1/5.

895

834

895

968

834

968

1088 1088

1223

1493

diff.1/25

895

834 834

895

968 968

1088

834

895

968

1223

diff.1/100

diff.1/200

1088

1223 1223

1496 1496

1088

diff.1/50

1494

1712 1662 1612 1661 1613

1712 1712

1660 1616

Absorbance

1223

1493

1611

1712 1662 1612

1712 1662

(A)

FT-IR spectroscopy is often used to characterize the specific binding nature of small molecules–DNA interaction and to monitor the effects of micromolecules on DNA structure. Fig. 10A presents the FT-IR spectra of free ctDNA and difference spectra of PSO–ctDNA complex in the region of 1800–800 cm−1 at different [PSO]/[ctDNA] ratios. The vibrational bands of DNA at 1712, 1660, 1616 and 1496 cm−1 correspond to guanine (G), thymine (T), adenine (A) and cytosine (C) nitrogenous bases, and bands at 1223 and 1088 cm−1 denote phosphate asymmetric and symmetric vibrations, respectively [35]. After PSO adding to ctDNA solution, the guanine band had no shift, the thymine band shifted from 1660 to 1662 cm−1 , the adenine band at 1616 shifted toward lower wavenumber 1611 cm−1 , and the cytosine band at 1496 appeared at 1493 cm−1 (Fig. 10A). All these shifting may be related with binding of PSO to the thymine, adenine and cytosine bases of ctDNA [56]. As shown in Fig. 10B, when the PSO concentration was low (r = 1/200, 1/100), intensities of the adenine band increased while that of the other three bases bands declined. Associated with molecular docking studies, the reason why adenine was different from the rest could be speculated to be the occurrence of specific binding between PSO and adenines. At high PSO concentration (r = 1/25), the intensities of the four bases bands increased to different extent, which could be attributed to some degree of helix destabilization [57]. Phosphate stretching (both asymmetric and symmetric) showed only intensity variations (increase or decrease) and no major shifting in wavenumber was observed, suggesting that there was no interaction of PSO with phosphate group of ctDNA backbone [56]. No change in the marker bands for B – form DNA at 895 and 834 cm−1 , which meant the maintenance of ctDNA in B – conformation [58].

Free ctDN A

1800

1600

1400

1200

1000

800

4. Conclusions

-1

Wavenu mber s ( cm ) (B)

Intensity/968

1.8

1712 1660 1616 1496 1223 1088

1.4

1.0

0.6

0.2 0

1/200

1/100

1/50

1/25

PSO/ctDNA molar ratios Fig. 10. (A) FT-IR spectra and difference FT-IR spectra [(ctDNA solution + PSO solution) − PSO solution] of the free ctDNA and PSO–ctDNA complex at different molar ratios in the region of 1800–800 cm−1 in aqueous solution. (B) Intensity ratio variations for several DNA in-plane vibrations as a function of PSO concentration.

In this paper, the interaction between PSO and ctDNA was forecast by molecular modeling software in a visualized way and further investigated by various spectroscopic techniques with the aid of chemometrics method. The expanded two-way spectroscopic data matrix obtained from two different UV–vis spectral titration experiments was resolved by MCR–ALS algorithm to get the concentration profiles and the pure spectra for the three reaction components (PSO, ctDNA and PSO–ctDNA complex) to monitor the progress of PSO interaction with ctDNA. The results of fluorescence titration suggested that the fluorescence of PSO could be markedly quenched by ctDNA and the quenching mechanism was considered as a static quenching procedure resulting from the formation of the PSO–ctDNA complex. The binding of PSO to ctDNA was able to increase ctDNA viscosity and melting temperature, displace the bound MB from the MB–ctDNA complex, decrease the binding ability of EB to ctDNA, induce changes in the CD spectra of ctDNA. Moreover, the iodide quenching effect was decreased and the fluorescence polarization of PSO was enhanced. All these experimental results demonstrated that PSO interacted with ctDNA by intercalative mode. Analysis of FT-IR spectra revealed that PSO was more prone to bind to adenine bases of ctDNA, which was in good accordance with the docking prediction. The negative values of H◦ and S◦ announced that the binding process was primarily driven by hydrogen bonds and van der Waals forces. It is noteworthy that the

X. Zhou et al. / International Journal of Biological Macromolecules 67 (2014) 228–237

application of molecular modeling to give a foresight for interaction of the small molecule drugs with DNA is faithful to some extent. The quick and reliable experimental means could provide more help in understanding the binding mechanism of PSO with ctDNA and the pharmacological effects of PSO as well as designing the structure of new and efficient drug molecules. Acknowledgments We are grateful for financial support provided by the National Natural Science Foundation of China (nos. 21167013 and 31060210), the Research Program of State Key Laboratory of Food Science and Technology of Nanchang University (SKLF–ZZB– 201305, SKLF–ZZA–201302, and SKLF–KF–201203), the Program of Jiangxi Provincial Department of Science and Technology (20112BBF60010), and the Natural Science Foundation of Jiangxi Province (20114BAB204019). References [1] N. Shahabadi, L. Nemati, DNA Cell Biol. 31 (2012) 883–890. [2] J.F. Li, C. Dong, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 71 (2009) 1938–1943. [3] N. Shahabadi, S.M. Fili, F. Kheirdoosh, J. Photochem. Photobiol., B: Biol. 128 (2013) 20–26. [4] P. Deepa, P. Kolandaivel, K. Senthilkumar, Mater. Sci. Eng., C 32 (2012) 423–431. [5] D.Z. Tang, F. Yang, Z. Yang, J. Huang, Q. Shi, D. Chen, Y.J. Wang, Biochem. Biophys. Res. Commun. 405 (2011) 256–261. [6] D. Xin, H. Wang, J. Yang, Y.F. Su, G.W. Fan, Y.F. Wang, Y. Zhu, X.M. Gao, Phytomedicine 17 (2010) 126–131. [7] E. Wang, C.N. Casciano, R.P. Clement, W.W. Johnson, Pharm. Res. 18 (2001) 432–438. [8] B.L. Lee, A. Murakami, K.R. Blake, S.B. Lin, P.S. Miller, Biochemistry 27 (1988) 3197–3203. [9] M.S. Rocha, N.B. Viana, O.N. Mesquita, J. Chem. Phys. 121 (2004) 9679–9683. [10] V. Carneiro Leite, R. Ferreira Santos, L. Chen Chen, L. Andreu Guillo, J. Photochem. Photobiol., B: Biol. 76 (2004) 49–53. [11] Q.S. Wang, L. Yang, T.T. Fang, S. Wu, P. Liu, X.M. Min, X. Li, Appl. Surf. Sci. 257 (2011) 9747–9751. [12] F. Ahmadi, B. Jafari, M. Rahimi-Nasrabadi, S. Ghasemi, K. Ghanbari, Toxicol. In Vitro 27 (2013) 641–650. [13] Z.G. Liu, S.N. Tan, Y.G. Zu, Y.J. Fu, R.H. Meng, Z.M. Xing, Micron 41 (2010) 833–839. [14] Y.N. Ni, M. Wei, S. Kokot, Int. J. Biol. Macromol. 49 (2011) 622–628. [15] M. Garrido, F.X. Rius, M.S. Larrechi, Anal. Bioanal. Chem. 390 (2008) 2059–2066. [16] Y. Zhang, G.W. Zhang, Y. Li, Y.T. Hu, J. Agric. Food Chem. 61 (2013) 2638–2647. [17] X.Y. Xu, D.D. Wang, X.J. Sun, S.Y. Zeng, L.W. Li, D.Z. Sun, Thermochim. Acta 493 (2009) 30–36. [18] S. Kashanian, M.M. Khodaei, P. Pakravan, DNA Cell Biol. 29 (2010) 639–646. [19] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig, I.N. Shindyalov, P.E. Bourne, Nucleic Acids Res. 28 (2000) 235–242. [20] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, R.K. Belew, A.J. Olson, J. Comput. Chem. 19 (1998) 1639–1662. [21] Y.D. Ma, G.W. Zhang, J.H. Pan, J. Agric. Food Chem. 60 (2012) 10867–10875. [22] P. Zhao, J.W. Huang, W.J. Mei, J. He, L.N. Ji, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 75 (2010) 1108–1114.

237

[23] E. Grueso, G. López-Pérez, M. Castellano, R. Prado-Gotor, J. Inorg. Biochem. 106 (2012) 1–9. [24] H. Gampp, M. Maeder, C.J. Meyer, A.D. Zuberbühler, Talanta 32 (1985) 1133–1139. [25] C.G. Ricci, P.A. Netz, J. Chem. Inf. Model. 49 (2009) 1925–1935. [26] B.M. Kalejahi, M. Bahram, A. Naseri, S. Bahari, M. Hasani, J. Iran. Chem. Soc. 11 (2014) 241–248. [27] Q. Zhang, Y. Ni, S. Kokot, J. Agric. Food Chem. 61 (2013) 7730–7738. [28] Y.N. Ni, Y.X. Wang, S. Kokot, Sens. Actuators, B: Chem. 156 (2011) 290–297. [29] S.Y. Bi, L.L. Yan, Y. Wang, B. Pang, T.J. Wang, J. Lumin. 132 (2012) 2355–2360. [30] Y.L. Zhang, X. Zhang, X.C. Fei, S.L. Wang, H.W. Gao, J. Hazard. Mater. 182 (2010) 877–885. [31] J.R. Lakowicz, G. Weber, Biochemistry 12 (1973) 4161–4170. [32] M. Xu, Z.R. Ma, L. Huang, F.J. Chen, Z.Z. Zeng, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 78 (2011) 503–511. [33] P. Mussardo, E. Corda, V. González-Ruiz, J. Rajesh, S. Girotti, M.A. Martín, A.I. Olives, Anal. Bioanal. Chem. 400 (2011) 321–327. [34] Z.G. Liu, Q.S. Xiang, L.H. Du, G. Song, Y.T. Wang, X.B. Liu, Food Chem. 141 (2013) 289–296. [35] S. Nafisi, A.A. Saboury, N. Keramat, J.F. Neault, H.A. Tajmir-Riahi, J. Mol. Struct. 827 (2007) 35–43. [36] Y.P. Zhang, G.W. Zhang, P. Fu, Y.D. Ma, J. Zhou, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 96 (2012) 1012–1019. [37] S. Kashanian, S. Askari, F. Ahmadi, K. Omidfar, S. Ghobadi, F.A. Tarighat, DNA Cell Biol. 27 (2008) 581–586. [38] A.H. Hegde, S.N. Prashanth, J. Seetharamappa, J. Pharm. Biomed. Anal. 63 (2012) 40–46. [39] N. Shahabadi, S. Kashanian, F. Darabi, DNA Cell Biol. 28 (2009) 589–596. [40] Y.M. Temerk, M.S. Ibrahim, M. Kotb, W. Schuhmann, Anal. Bioanal. Chem. 405 (2013) 3839–3846. [41] S. Kashanian1, J. Ezzati Nazhad Dolatabadi, DNA Cell Biol. 28 (2009) 535–540. [42] F.L. Cui, G.Q. Hui, X.Y. Jiang, G.S. Zhang, Int. J. Biol. Macromol. 50 (2012) 1121–1126. [43] S. Kashanian, M.M. Khodaei, F. Kheirdoosh, J. Photochem. Photobiol., B: Biol. 120 (2013) 104–110. [44] Z.C. Xu, G. Bai, C. Dong, Bioorg. Med. Chem. 13 (2005) 5694–5699. [45] Y.J. Liu, X.Y. Wei, F.H. Wu, W.J. Mei, L.X. He, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 70 (2008) 171–176. [46] C.Y. Qiao, S.Y. Bi, Y. Sun, D.Q. Song, H.Q. Zhang, W.H. Zhou, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 70 (2008) 136–143. [47] B. Macías, M.V. Villa, R. Lapresa, G. Alzuet, J. Hernández-Gil, F. Sanz, J. Inorg. Biochem. 115 (2012) 64–71. [48] J. Liu, T.X. Zhang, T.B. Lu, L.H. Qu, H. Zhou, Q.L. Zhang, L.N. Jig, J. Inorg. Biochem. 91 (2002) 269–276. [49] R.W. Armstrong, T. Kurucsev, U.P. Strauss, J. Am. Chem. Soc. 92 (1970) 3174–3181. [50] D. Xu, X. Wang, L. Ding, Carbohydr. Polym. 83 (2011) 1257–1262. [51] W. Chen, N.J. Turro, D.A. Tomalia, Langmuir 16 (2000) 15–19. [52] F. Ma, X.F. Ge, H.Y. Huang, C. Yang, L. Han, J.H. Zhou, X.D. Yang, Spectrochim. Acta, A: Mol. Biomol. Spectrosc. 109 (2013) 158–163. [53] B.K. Sahoo, K.S. Ghosh, R. Bera, S. Dasgupta, Chem. Phys. 351 (2008) 163–169. [54] P. Uma Maheswari, M. Palaniandavar, J. Inorg. Biochem. 98 (2004) 219–230. [55] B. Ranjbar, P. Gill, Chem. Biol. Drug Des. 74 (2009) 101–120. [56] D.K. Jangir, G. Tyagi, R. Mehrotra, S. Kundu, J. Mol. Struct. 969 (2010) 126– 129. [57] S.T. Saito, G. Silva, C. Pungartnik, M. Brendel, J. Photochem. Photobiol., B: Biol. 111 (2012) 59–63. [58] D.K. Jangir, G. Tyagi, R. Mehrotra, S. Kundu, J. Photochem. Photobiol., B: Biol. 105 (2011) 143–148.