Characterization of π-stacking interactions between aromatic amino acids and quercetagetin

Journal of Molecular Structure 1128 (2017) 13e20

Contents lists available at ScienceDirect

Journal of Molecular Structure journal homepage: http://www.elsevier.com/locate/molstruc

Characterization of p-stacking interactions between aromatic amino acids and quercetagetin Farideh Badichi Akher, Ali Ebrahimi*, Najmeh Mostafavi Department of Chemistry, Computational Quantum Chemistry Laboratory, University of Sistan and Baluchestan, P.O. Box 98135-674, Zahedan, Iran

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 June 2016 Received in revised form 1 August 2016 Accepted 15 August 2016 Available online 18 August 2016

In the present study, the p-stacking interactions between quercetagetin (QUE), which is one of the most representative flavonol compounds with biological and chemical activities, and some aromatic amino acid (AA) residues has been investigated by the quantum mechanical calculations. The trend in the absolute value of stacking interaction energy jDEj with respect to AAs is HIS > PHE > TYR > TPR. The P results show that the sum of donor-acceptor interaction energy between AAs and QUE ( E2) and the P P sum of electron densities r calculated at BCPs and CCPs between the rings ( rBCPs and rCCP) can be useful descriptors for prediction of the DE values of the complexes. The OeH bond dissociation enthalpy (BDE) slightly decreases by the p-stacking interaction, which confirms the positive effect of that interaction on the antioxidant activity of QUE. A reverse trend is observed for BDE when is compared with the jDEj values. A reliable relationship is also observed between the Muliken spin density (MSD) distributions of the radical species and the most convenient OeH bond dissociations. In addition, reactivity is in good correlation with the antioxidant activity of the complexes. © 2016 Elsevier B.V. All rights reserved.

Keywords: Quercetagetin Flavonoid p-Stacking Antioxidant OH BDE

1. Introduction Flavonoids are polyphenolic compounds that are present in fruits, vegetables, coffee beans, green and black tea, red and white win, herbs and propolis [1]. They have a special importance in the medical and biological sciences due to their biological activities, e.g. antiallergic, antioxidant, antiviral, antibacterial and anticancer [2e4]. Three major subfamilies of flavonoids are flavones, flavonols and flavanones. The chemical structure of flavonoid is based on C6eC3eC6 carbon framework (subscripts 3 and 6 represent the number of carbon in each ring), consisting of two aromatic benzene rings (rings A and B) linked via an oxygen-containing pyran ring (ring C) [5,3]. There are differences in the degree of pyran ring saturation, in the placement of ring B at the C2 or C3 positions of ring C and the position and the number of hydroxyl groups [3]. Several studies have shown that flavonoids are enzyme inhibitors and are bound to active sites via noncovalent interactions [6e12]. Noncovalent interactions (hydrogen bond, van der Waals, charge transfer, p-stacking, etc.) play significant roles in drug design, supramolecular chemistry, materials science and molecular

* Corresponding author. E-mail address: [email protected] (A. Ebrahimi). http://dx.doi.org/10.1016/j.molstruc.2016.08.040 0022-2860/© 2016 Elsevier B.V. All rights reserved.

biology [13e18]. Many theoretical and empirical studies have focused on p-stacking interactions because of their biological importance [19e34], especially in the protein-ligand complexes. There is a set of small and large molecules, including aromatic rings, which make such biologically important complexes via p-stacking interactions with aromatic amino acids (AAs) like tryptophan (TRP), tyrosine (TYR), phenylalanine (PHE) and histidine (HIS) [35e39]. The crystal structures also indicate p-stacking interactions between ligands and aromatic AAs in the active sites of enzymes. The Tshaped, edge-to-face, and parallel-displaced stacking configurations are favored energetically in comparison with the sandwich configuration. It should be noted that the parallel displaced configuration is observed more frequently than the T-shaped configuration in the protein structures [26,40]. In recent decades, many computational studies have been performed to characterize the p-stacking interactions in the active sites of proteins, which are highly informative about the nature and importance of these interactions [41e45]. Intermolecular p-stacking interactions are also occurred in the crystal structures of flavonoids due to the planarity, polarity, and aromaticity [46,47]. The inhibition mechanism of cytochrome P450 2C9 (CYP2C9) has previously been investigated via a series of flavonoids [10]. The results demonstrated that all the tested flavonoids are reversible inhibitors of CYP2C9 and among them, 6-hydroxyflavone acts as

14

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

noncompetitive inhibitor and other flavonoids are competitive inhibitors. In addition, the work used computer aided docking and molecular dynamic simulation to characterize the binding sites of flavonoids in CYP2C9. Results indicated that the competitive inhibitors bind to the substrate binding site whereas the noncompetitive inhibitor binds to the allosteric site of the enzyme via a pstacking interaction with Phe100 residue and the hydrogen bonds with Leu102 and Arg105 residues. The inhibitory mechanism of the b-hydroxyl-acyl carrier protein dehydratase of Helicobacter pylori (HpFabZ) by three flavonoids (S)-sakuranetin, quercetin and apigenin has been studied by Zhang and coworkers [11]. These flavonoids were identified to be competitive inhibitors and bind to HpFabZ via two models. In one model, flavonoids act as enzyme inhibitors via noncovalent interactions and are stacked with two residues Tyr100 and Pro112. Elsewhere, Pawlotsky and coworkers [12] investigated the inhibition process of hepatitis C virus RNAdependent RNA polymerase (HCV RdRp) by six flavonoids. Quercetagetin (QUE), a type of flavonol indicated in Scheme 1, was found to be a more potent inhibitor in comparison with other flavonoids. The X-ray structure of the HCV RdRp-QUE complex shows that QUE binds to allosteric site of enzyme and alter its 3D structure. This binding is done through noncovalent interactions (a hydrogen bond with G238 and a p-stacking interaction with Phe162 residue), which are poorly described by density functionals. Moreover, p-stacking interactions may also affect the antioxidant action of flavonoids. In fact, the activity of phenolic antioxidants is not solely dependent on their structures. It may also be affected by the noncovalent interactions in the surrounding environment [2], particularly p-stacking interactions. Thus, it is important to investigate the p-stacking interaction between flavonoids and aromatic AAs and its effect on their antioxidant activity. In the present work, the p-stacking interactions of QUE and some amino acids containing five and six-membered aromatic rings have been investigated at the M06-2X/6-311þþG(d,p) level of theory. The complexes have been investigated with respect to the energy data and the results of the population analyses. Herein, the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), bond dissociation energies (BDE), and Muliken spin density distribution (MSD) values have been considered to investigate the antioxidant activity of stacked QUE. This study can provide an insight about the magnitude of p-stacking interactions between the aromatic AAs and QUE and their effect on antioxidant activity of QUE, which is useful in biological processes.

2. Computational details All calculations were carried out using the Gaussian09 program

Scheme 1. The structure of quercetagetin (3,5,6,7,30 ,40 -hexa hydroxy flavonol).

package [48]. All units were fully optimized by the hybrid meta exchange-correction functional M06-2X [49] method in conjunction with the 6-311þþG(d,p) basis set. The nucleotide units were modeled by replacing a hydrogen atom with the protein backbone of AAs (indole for TRP, phenol for TYR, benzene for PHE and imidazole for HIS). The initial structures of AAjjQUE complexes (where jj donates p-stacking interaction) were generated by aligning the centers of rings of units. Four parameters including vertical separation (R1), rotational angle (a), and horizontal displacement (R2 and R3), were considered to determine the optimal complexes (see Scheme 2). The potential energy surface scans were carried out through the single point calculations at the M06-2X/6-311þþG(d,p) level to obtain the optimized geometries of the AAjjQUE complexes. The preferred R1 value was obtained by 0.1 increments and held constant for the remaining calculations. Then, AA was rotated by 360 (12 steps of size 30 ) for a increments in right-handed sense about the axis that passes through the centers of rings. Using the optimal values for a and R1, the optimal values of R2, and R3 were determined upon the shift of AA across the face of QUE in 0.1 increments. All the above mentioned steps were performed for three rings of QUE, which specified by A, B and C letters in Scheme 1. After finding the best orientations of AAs relative to the rings A, B and C of QUE, the most stable complexes were obtained by optimization at the M06-2X/6-311þþG(d,p) level of theory. No imaginary frequency was found for any of the optimized structures. In addition, single point interaction energies were calculated by the B3LYP-D3 [50,51] dispersion corrected functional as suggested by Grimme for noncovalent contacts [52,53]. The basis set superposition error (BSSE) correction was performed by the counterpoise (CP) method. Also, to investigate the effect of p-stacking interaction on the antioxidant activity of QUE, radical structures of QUE and stacked QUE were optimized at the above mentioned level. The topological properties of electron charge density were calculated by the atoms in molecules (AIM) method [54] using the AIM2000 program package [55] on the wave functions obtained at the HF/6-311þþG(d,p) level of theory. The natural bond orbital (NBO) analysis [56] has been performed on those wave functions by the NBO3.1 software [57]. Moreover, MSDs for radical species have also been calculated and interpreted.

3. Results and discussion The interaction energies (DE) obtained for each step along the scans of variables, introduced in the previous section, are available in Supplementary material (see Table 1S). The largest DE values calculated between AAs and the rings A, B and C of QUE are reported in Table 1. The trend in the jDEj values is B < A < C in stacking

Scheme 2. Definition of variables in the AAjjQUE-C complexes.

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

with each AA. As can be seen in Table 1, the orders of the DE values are identical at the M06-2X/6-311þþG(d,p) and the B3LYP-D3/6311þþG(d,p) level. The presence of carbonyl group in the ring C of QUE can be a reason for the more stability of the AAjjQUE-C complex. The electron withdrawing character of C]O in the ring C can reduce the negative charge of p-electron cloud that decreases the repulsion between the rings and makes more favorable the electrostatic interactions. Also, two intramolecular hydrogen bonds between C4¼O and the 3-OH and 5-OH groups can reduce repulsion between the rings. The 5-OH group located in meta position relative to the 7-OH group in the ring A acts as an electron withdrawing group via the inductive effect. This effect makes AAjjQUE-A more stable than AAjjQUE-B through reduction of the negative charge of p-electron cloud. Higher jDEj values are obtained from horizontal displacement in the R2 direction, because of increase in the p-p overlap and favorite contacts between rings, which shift AAs toward the ring C. In addition, the sum of atomic ChelpG charges on the atoms of the rings A (C atoms, qA), B (C atoms, qB) and C (C and O atoms, qC) were calculated for QUE before stacking. It is interesting to note that qC (0.445) is more positive in comparison with qB (0.071) and qA (0.244) in QUE. So, the ring C can be more susceptible to pstacking interactions due to the lower negative charge of p-electron cloud that results in more favorable electrostatic interactions. On the other hand, qB is negative while qA is positive. This can be a reason of decrease in the electrostatic repulsion between stacked rings and more stability of the AAjjQUE-A complexes relative to AAjjQUE-B. Charge transfer (CT) has frequently been used to predict the intermolecular overlap and the stability of p-stacking complexes [58]. The amount of CT between AA and QUE was determined as the sum of atomic charges calculated on the QUE fragment (see Table 1). A relatively good correlation is found between the DE values and CT for the most stable AAjjQUE-X complexes (R ¼ 0.90, see Fig. 1). Thus, it can be concluded that the electrostatic interaction plays an important role in the stability of complexes. The trend in the jDEj values obtained for the stacking of AAs with the ring C (the most stable complexes) is HIS < PHE < TYR < TPR. An identical trend can also be obtained for the size of AAs, in agreement with the relative surface area, polarizability and dipole moments of the amino acids. The most stable complexes (AAjjQUE-C) were optimized after finding the preferred orientation of each AA relative to QUE by the potential energy surface scans, the energy data obtained at the

Table 1 BSSE corrected interaction energies (DE in kcal mol1) of the AAjjQUE-C complexes calculated at the M06-2X/6-311þþG(d,p) level of theory, the value of CT (10 in e) between AA and QUE, and the LUMO-HOMO energy gap (DELH in eV). Complex

AA

R2

DE

a

AAjjQUE-A

TRP TYR PHE HIS TRP TYR PHE HIS TRP TYR PHE HIS

1.4 1.4 1.5 1.4 1.1 1.2 1.3 1.5 0.8 0.8 0.8 0.9

8.8 7.2 6.7 6.1 7.5 5.1 4.4 5.6 9.5, 8.4, 7.1, 6.5,

10.3 8.6 7.9 6.8 8.5 6.1 5.1 6.0 10.8,11.7 9.1,9.4 7.8,8.1 7.0,7.3

AAjjQUE-B

AAjjQUE-C

11.50 9.20 7.70 7.60

DE

CT

DELH

11.9 4.4 3.7 0.7 16.7 11.4 9.6 13.5 18.2 0.7 5.0 5.3

e

e

21.08 21.47 21.60 21.72

a BSSE corrected interaction energies of the AAjjQUE-C complexes calculated at the B3LYP-D3/6-311þþG(d,p) level. The bold values correspond to the complexes optimized at the M06-2X/6-311þþG(d,p) level.

15

Fig. 1. Correlation between the DE and CT values.

M06-2X/6-311þþG(d,p) level are summarized in Table 1. The calculated binding energy is in the range of 7.6e11.5 kcal mol1 and trend in the stability of complexes is similar to that found in the potential energy surface scan. The optimized geometries of the AAjjQUE-C complexes and the most important geometrical parameters are gathered in Fig 2. Relative orientations of units remain approximately constant after geometry optimization of complexes, so the potential energy surface scans through the single point calculations can be suitable in order to predict the optimal complexes. Charge delocalization between the rings A and B is done via the C2eC3 bond (see Scheme 1); this role can justify the effect of pstacking interaction on the C2eC3 bond length. As can be seen in Fig. 2, the trend in the C2eC3 bond length of the AAjjQUE-C complexes is TRP > TYR > PHE > HIS. The complexes become less stable by decrease in the C2eC3 bond length, which is a result of decreasing the charge delocalization. The trend in the hydroxyl bond lengths is 5-OH > 3-OH > 7-OH > 6-OH > 30 -OH > 40 -OH in the AAjjQUE-C complexes. Higher 5-OH and 3-OH bond lengths can be related to the formation of two strong hydrogen bonds between those groups and the C4¼O carbonyl group of ring C. The 5-OH$$$O4 distance is shorter than 3OH$$$O4, so the 5-OH bond is longer than 3-OH. Therefore, there is a linear correlation between the OH bond length and the strength of intermolecular H-bond. As seen, the OH bond lengths are shorter in ring B relative to those in rings C and A. Lower tendencies for intramolecular H-bond formation are accompanied with the shorter OH bond lengths in ring B. It should be noted that after optimization, correlation between the DE values of the AAjjQUE-C complexes and CT increases from 0.91 to 0.93. The ChelpG charges implicate CT from ring B to rings A and C. Also, a good linear correlation is observed between the stabilization energy of complexes and the sum of atomic charges on P the rings A and C ( qAC ¼ qA þ qC, see Fig. 3). In order to investigate the reactivity of the AAjjQUE-C complexes, the LUMO-HOMO energy gaps (DELH ¼ ELUMO-EHOMO) were calculated. As can be seen in Table 1, the trend of the DELH values increases as TRP < TYR < PHE < HIS, which is inversely related to the stability of the complexes. DELH is an important parameter to describe the reactivity of complexes toward the free radical. The TRPjjQUE-C complex has the lowest DELH value (higher softness), which means the complex can behave as a better antioxidant in comparison with other complexes.

4. AIM and NBO analyses AIM analysis has been used to distinguish the strength of the noncovalent interactions involved in the AAjjQUE-C complexes. Topological properties of the electronic charge density of the

16

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

Fig. 2. The AAjjQUE-C complexes optimized at the M06-2X/6-311þþG(d,p) level. The C2eC3 and OH bond lengths are in Å, and CHELPG charges are in e. P P P charges on the ring B and AA ( qAA-B ¼ qAA þ qB), and qA-C is sum of the charge on the rings A and C ( qA-C ¼ qA þ qC).

P qAA-B is the sum of

Scheme 3. A typical molecular graph of the AAjjQUE-C complex. The small red, green, and yellow spheres correspond to BCPs, CCPs and RCPs, respectively. Fig. 3. Correlation between the DE values and the sum of the charges on the rings A and C.

complexes have been calculated by the AIM method on the wave functions generated at the HF/6-311þþG(d,p) level of theory. A typical molecular graph of AAjjQUE-C is presented in Scheme 3.

Bond critical points (BCPs), ring critical points (RCPs), and cage critical points (CCPs) were illustrated in this graph. The sum of P P electron densities calculated at BCPs and CCPs ( rBCPs and rCCPs) P P are shown in Table 2. The highest/lowest value of rBCP and rCCPs

Table 2 P P P The rBCP and rCCP values calculated at the BCPs observed between the rings, the rRCP values calculated at the RCPs of rings A, B and C and their summation ( rRCP), the rBCP P values corresponding to the intramolecular hydrogen bonds in QUE and their summation ( rBCP) obtained from AIM analysis on the wave function produced at the HF/6311þþG (d,p) level of theory in e/au3. P P rBCP rCCP rRCP rBCP AA P P A C B rRCP HB1 HB2 HB3 rBCP TRP TYR PHE HIS

3.863 3.775 3.432 2.468 e

1.489 0.957 0.888 0.506 e

2.107 2.101 2.096 2.092 2.079

The bold values correspond to QUE before stacking.

2.140 2.137 2.137 2.135 2.125

2.192 2.190 2.189 2.188 2.177

6.440 6.428 6.421 6.416 6.381

1.831 1.835 1.701 1.714 1.758

2.399 2.444 2.395 2.469 2.410

3.728 3.554 3.576 3.453 3.436

7.958 7.833 7.672 7.636 7.604

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

▪

Fig. 4. Correlation between the DE and rRCP values of the rings A ( ), B (

:

) and C ().

17

Fig. 7. Correlation between DE and the sum of the donor-acceptor interaction energies P of HB1, HB2 and HB3 ( E2HB).

Table 4 Bond dissociation enthalpy (BDE in kJ mol1) calculated at the M06-2X/6311þþG(d,p) level of theory for the AAjjQUE-C complexes. BDE

Fig. 5. Correlation between DE and the sum of rBCPs calculated at the intramolecular Hbonds.

Table 3 The donor-acceptor interaction energies (in kcal mol1) obtained from the NBO analysis on the wave functions calculated at the HF/6-311þþG(d,p) level for the AAjjQUE-C complexes. aP 2 bP 2 AA E2AA/QUE E2QUE/AA E2HB1 E2HB2 E2HB3 E E HB TRP 3.71 4.91 8.62 TYR 3.10 3.36 6.46 PHE 2.78 3.19 5.97 HIS 2.22 3.67 5.89 P P a P 2 E ¼ E2AA/QUE þ E2QUE/AA. b P 2 E ¼ E2HB1 þ E2HB2 þ E2HB3.

1.60 1.73 1.16 1.17

3.41 3.58 3.39 3.79

16.47 14.83 15.05 13.95

21.48 20.14 19.60 18.91

corresponds to the TRPjjQUE-C/HISjjQUE-C complex. The stabilities P P of the complexes increase by increasing the rBCP and rCCP P P values. It demonstrates that the rBCP and rCCP values are useful

Fig. 6. Correlation between DE and the sum of the donor-acceptor interaction energies P observed between AA and QUE ( E2).

AA

30 -OH

40 -OH

3-OH

5-OH

6-OH

7-OH

TRP TYR PHE HIS

372.48 371.71 372.93 374.37

325.53 326.64 326.69 327.21

353.13 355.80 354.83 349.18

369.50 377.20 378.28 372.94

328.41 334.42 334.69 333.11

411.13 381.98 472.34 471.99

to describe interactions of the aromatic rings. The electron charge density calculated at the RCP (rRCP) are also summarized in Table 2. As can be seen, the p stacking interaction increases the rRCP values of the rings A, B and C of QUE. The DE values correlate with the rRCP values of QUE, such that the stability of complexes increases by increasing the rRCP values of the rings A, B and C (see Fig. 4). The AIM analysis reveals the presence of three intramolecular hydrogen bonds 5-OH$$$O4 (HB1), 3-OH$$$O4 (HB2) and 3-O$$$60 H (HB3) in the AAjjQUE-C complexes. The rBCP values of the intramolecular hydrogen bonds are presented in Table 2. The rBCP values increase as HB3 < HB2 < HB1 in each complex. The p-stacking interaction can affect the strength of intramolecular hydrogen bond. A reasonable correlation is observed between the DE values and the sum of rBCPs calculated at the intramolecular hydrogen bonds (see Fig. 5). The donor-acceptor interaction energy (E2) can also be used as a measure of the strength of the noncovalent interactions involved in the AAjjQUE-C complexes. The data obtained from the NBO analysis are shown in Table 3. The sum of the E2 values for interactions P P P between AA and QUE ( E2 ¼ E2AA/QUE þ E2QUE/AA) can be used as a descriptor for prediction of the DE values of the

Fig. 8. Correlation between BDE and HOMO-LUMO energy gap (DELH).

18

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

Fig. 9. MSD on QUE in the AAjjQUE-C complexes.

P 2 complexes. The highest/lowest E value corresponds to the TRPjjQUE-C/HISjjQUE-C complex. As can be seen in Fig. 6, a P reasonable correlation is observed between the DE and E2 values. 2 In addition, the E values of the nO4/s*5OH, nO4/s*3OH and nO3/s*60 CH interactions can be considered as measures of the

strength of the intramolecular hydrogen bonds. Table 3 shows that the E2 values increase as HB3 < HB2 < HB1 in each complex, which is in agreement with the data obtained from the AIM analysis. Good linear correlations are observed between the E2 and rBCP values of the intramolecular hydrogen bonds (R ¼ 0.98 for HB1 and HB2, and

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

R ¼ 1.00 for HB3). Also, the sum of the E2 values of HB1, HB2 and P HB3 ( E2HB ¼ E2HB1 þ E2HB2 þ E2HB3) can be used as a descriptor to predict the stability of the complexes (see Fig. 7).

19

radicals which is in agreement with the highest BDE value corresponding to the related complexes. 6. Conclusions

5. Bond dissociation energy (BDE) The OH BDEs of QUE in the AAjjQUE-C complexes are important parameters to evaluate the antioxidant activity [59]. The lower BDE values indicate higher hydrogen donation ability from the OH groups [5]. The OH BDEs of QUE have been calculated by the following equation BDE ¼ H(AAjjQUE-O) þ H(H)  H(AAjjQUE-OH) where AAjjQUE-OH indicates the parent complex and AAjjQUE-O is the corresponding radical. The enthalpy value of H was calculated to be 0.495834 hartree at the M06-2X/6-311þþG(d,p) level. From Table 4, it is obvious that the lowest values of BDE in all complexes are attributed to the 40 -OH and 6-OH groups on the rings B and A, respectively in the complexes. H atom abstraction of the 40 OH and 6-OH groups leads to the formation of a relatively strong intramolecular hydrogen bond (OH/O) with the OH group located at adjacent carbon (30 -OH$$$O40 and 7-OH$$$O6) that may be the reason of more stability of the AAjjQUE-O40  and AAjjQUE-O6 radicals. On the other hand, the results indicate that the H atom abstraction from the 40 -OH group (ring B) is easier than 6-OH (ring A). Charge delocalization is happened for the 40 -OH radical via the C2eC3 bond, while such a delocalization is not observed for the 6OH radical [3]. The higher BDE values in the AAjjQUE-O3 and AAjjQUE-O5 radicals can be attributed to the intramolecular hydrogen bonds of 3-OH and 5-OH with the C4¼O group. The 5OH$$$O4 H-bond is stronger than 3-OH$$$O4, so the BDE value of 5-OH is higher than 3-OH. The highest value of BDE is observed for the 7-OH group of the complexes. The existence of the 5-OH group in QUE can increase the 7-OH BDE in comparison with other OH groups. The 5-OH group located in the meta position of the 7-OH group behaves as an electron withdrawing group and enhances the BDE value of 7OH. The order of BDE for 40 -OH is found to be TRP < TYR < PHE < HIS in the complexes; this is in agreement with the DELH values of the complexes. There is a good relationship between reactivity of complexes and antioxidant activity of stacked QUE (see Fig. 8). The p-stacking interaction slightly decreases the 40 -OH BDE and increases the antioxidant activity of QUE. MSD can be used as a realistic parameter to describe the reactivity [60]. The antioxidant ability of QUE in the AAjjQUE-C complexes can also be visualized via the MSD analysis of corresponding radicals. Therefore, in order to discuss the reactivity of different OH groups of QUE and consequently the variations in the BDE values, MSD analysis has been performed on various QUE radical forms of the complexes. A higher MSD in a radical leads to a lower BDE and makes easier radical formation [61]. As can be seen in Fig. 9, the highest MSD corresponds to the AAjjQUE-O40  radicals. This can be attributed to the C2eC3 bond which allows the MSD distribution on two rings B and C. This result can also be explained by high spin density on the C0 1 atom at para position of the 40 -OH group in the AAjjQUE-O40  radicals, and spin delocalization via the C2eC3 bond by resonance. There is not such a delocalization for the AAjjQUEO6 radicals which is in agreement with the BDE data. In addition, higher MSDs for the AAjjQUE-O3 radicals in comparison with the AAjjQUE-O5 radicals confirm the lower BDE value for 3-OH. The atomic MSD is delocalized on two rings B and C in the AAjjQUE-O3 radicals, while it just is delocalized on the ring A in AAjjQUE-O5. On the other hand, the lowest MSD corresponds to the AAjjQUE-O7

In the present study, the quantum mechanical calculations were used to characterize the stacking interactions utilized by QUE. To the best of our knowledge, this is the first study that considers the staking interactions between QUE as a flavonol and four aromatic AAs, and therefore the effect of that interaction on the antioxidant activity of QUE. This information can be useful since the magnitudes of descried p-stacking interactions are difficult to obtain directly from experiments where it is hard to separate other effects. A broader understanding of the effects of charge on both the preferred geometries and the magnitude of the binding energies has been revealed in the AAjjQUE-X complexes. The results indicate that the ring C in QUE is more suitable than the rings A and B for the p-stacking interaction. The sum of atomic ChelpG charges on the atoms of the ring C is more positive than those of A and B, which results in more favorable electrostatic interactions and therefore stronger p-stacking interactions. A relatively good correlation is found between the DE values and CT for the most stable AAjjQUE-X complexes. On the other hand, the interaction energies of the most stable complexes (AAjjQUE-C) with respect to AAs decrease as TRP > TYR > PHE > HIS, which this trend is dictated by the relative surface area, polarizability and dipole moments of AAs. The stability of complexes is accompanied with decrease in the C2eC3 bond length, which is a result of decreasing the charge delocalization. Reactivity of the AAjjQUE-C complexes estimated using the LUMOHOMO energy gap decreases as TRP > TYR > PHE > HIS. According to the AIM analysis, the complexes become more P P stable by increasing the rBCP and rCCP values obtained between P P rings. So, it demonstrates that the rBCP and rCCP values are useful to describe interactions of the aromatic rings. On the other hand, The NBO analysis shows that the sum of donor-acceptor interaction energies between AA and QUE P P P ( E2 ¼ E2AA/QUE þ E2QUE/AA) increases by increasing the DE values of the complexes. The lowest values of BDE corresponds to the 40 -OH and 6-OH groups on the rings B and A, respectively, in the complexes. The pstacking interaction slightly decreases the 40 -OH BDE such that the trend becomes HIS > PHE > TYR > TRP and confirms the positive effect of p-stacking in increasing the antioxidant reactivity of QUE. On the other hand, the most MSD in the radical species corresponds to AAjjQUE-O40 , which is in agreement with the BDE data. Acknowledgment We thank the University of Sistan and Baluchestan for financial supports and Computational Quantum Chemistry Laboratory for computational facilities. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.molstruc.2016.08.040. References [1] M. Leopoldini, F. Rondinelli, N. Russo, M. Toscano, J. Agric. Food Chem. 58 (2010) 8862e8871. nek, E. Klein, Phys. Chem. Chem. Phys. 15 (2013) [2] J.L. yel, J. Rimar cík, A. Vaga 10895e10903. [3] W.A. Peer, A.S. Murphy, in: E. Grotewold (Ed.), The Science of Flavonoids, Springer, New York, NY, USA, 2006. [4] P.M. Mitrasinovic, J. Chem, Inf. Model. J. Chem. Inf. Model 55 (2015) 421e433.

20

F.B. Akher et al. / Journal of Molecular Structure 1128 (2017) 13e20

[5] D. Ami, D.D. Ami, D. Belo, V. Rastija, B. Lui, N. Trinajsti, Curr. Med. Chem. 14 (2007) 827e845. [6] Y. Araia, S. Endoa, N. Miyagia, N. Abeb, T. Miurac, T. Nishinakac, T. Teradac, M. Oyamab, H. Godad, O. El-Kabbanie, A. Haraf, T. Matsunagaa, A. Ikaria, Fitoterapia 101 (2015) 51e56. [7] K.E. Tsitsanoua, J.M. Hayesa, M. Keramiotia, M. Mamaisa, N.G. Oikonomakosa, A. Katoc, D.D. Leonidasd, S.E. Zographosa, Food. Chem. Toxicol. 61 (2013) 14e27. [8] S.A. Hudson, H. Ecroyd, F.C. Dehle, I.F. Musgrave, J.A. Carver, J. Mol. Bio. 392 (2009) 689e700. [9] N. Ferreiraa, A.P. Henriquesa, M.R. Almeida, Biochem. Biophys. Rep. 3 (2015) 123e133. [10] D. Si, Y. Wang, Y.H. Zhou, Y. Guo, J. Wang, H. Zhou, Z.S. Li, J.P. Fawcett, Drug. Metab. Dispos. 37 (2009) 629e634. [11] L. Zhang, Y. Kong, D. Wu, H. Zhang, J. Wu, J. Chen, J. Ding, L. Hu, H. Jiang, X. Shen, Protein. Sci. 17 (2008) 1971e1978. [12] A.A. Belkacem, J.F. Guichou, R. Brillet, N. Ahnou, E. Hernandez, C. Pallier, J.M. Pawlotsky, Nucleic. Acids. Res. 42 (2014) 9399e9409. [13] S.E. Wheeler, A.J. McNeil, P. Müller, T.M. Swager, K.N. Houk, J. Am. Chem. Soc. 132 (2010) 3304e3311. [14] M.L. Waters, Curr. Opin. Chem. Biol. 6 (2002) 736e741. [15] E.A. Meyer, R.K. Castellano, F. Diederich, Angew. Chem. Int. Ed. Engl. 42 (2003) 1210e1250. [16] P. Hobza, Phys. Chem. Chem. Phys. 10 (2008) 2581e2583. [17] C.A. Hunter, K.R. Lawson, J. Perkins, C.J. Urch, J. Chem. Soc. Perkin Trans. 2 (2001) 651e669. [18] C.A. Hunter, J.K.M. Sanders, J. Am. Chem. Soc. 112 (1990) 5525e5534. [19] E.C. Lee, D. Kim, P. Jurecka, P. Tarakeshwar, P. Hobza, K.S. Kim, J. Phys. Chem. A 111 (2007) 3446e3457. [20] D. Quinonero, A. Frontera, P.M. Deya, I. Alkorta, J. Elguero, Chem. Phys. Lett. 460 (2008) 406e410. [21] D. Escudero, A. Frontera, D. Quinonero, P.M. Deya, J. Phys. Chem. A 112 (2008) 6017e6022. [22] M. Zaccheddu, C. Filippi, F. Buda, J. Phys. Chem. A 112 (2008) 1627e1632. [23] T. Janowski, P. Pulay, Chem. Phys. Lett. 447 (2007) 27e32. [24] M.J. Rashkin, M.L. Waters, J. Am. Chem. Soc. 124 (2002) 1860e1861. [25] M.O. Sinnokrot, E.F. Valeev, C.D. Sherrill, J. Am. Chem. Soc. 124 (2002) 10887e10893. [26] S. Tsuzuki, K. Honda, T. Uchimaru, M. Mikami, K. Tanabe, J. Am. Chem. Soc. 124 (2002) 104e112. [27] E.C. Lee, B.H. Hong, J.Y. Lee, J.C. Kim, D. Kim, Y. Kim, P. Tarakeshwar, K.S. Kim, J. Am. Chem. Soc. 127 (2005) 4530e4537. [28] B.K. Mi shra, N. Sathyamurthy, J. Phys. Chem. A 109 (2005) 6e8. [29] S. Beg, K. Waggoner, Y. Ahmad, M. Watt, M. Lewis, Chem. Phys. Lett. 455 (2008) 98e102. [30] M.O. Sinnokrot, C.D. Sherrill, J. Phys. Chem. A 107 (2003) 8377e8379. [31] M. Busker, Y.N. Svartsov, T. Haber, K. Kleinermanns, Chem. Phys. Lett. 467

(2009) 255e259. [32] W. Versees, S. Loverix, A. Vandemeulebroucke, J. Mol. Biol. 338 (2004) 1e6. [33] A. Ebrahimi, M. Habibi-Khorassani, A.R. Gholipour, H.R. Masoodi, Theor. Chem. Acc. 124 (2009) 115e122. [34] A. Ebrahimi, M. Habibi, R.S. Neyband, A.R. Gholipour, Phys. Cem. Chem. Phys. 11 (2009) 11424e11431. [35] D.D. Boehr, A.R. Farley, G.D. Wright, J.R. Cox, Chem. Biol. 9 (2002) 1209e1217. [36] C. Bissantz, B. Kuhn, M. Stahl, J. Med. Chem. 53 (2010) 5061e5084. [37] M. Brandl, M.S. Weiss, A. Jabs, J. Sühnel, R. Hilgenfeld, J. Mol. Biol. 307 (2001) 357e377. [38] S.K. Burley, G.A. Petsko, SCI 229 (1985) 23e28. [39] S.K. Burley, G.A. Petsko, Adv. Protein Chem. 39 (1988) 125e189. [40] G.B. McGaughey, M. Gagne, A.K. Rappe, J. Biol. Chem. 273 (1998) 15458e15463. [41] F.B. Akher, A. Ebrahimi, J. Mol. Graph. Model 61 (2015) 115e122. [42] Y. Zhao, D.G. Truhlar, J. Chem. Theory Comput. 3 (2007) 289e300. [43] M.J.G. Moa, M. Mandado, R.A. Mosquera, J. Phys. Chem. A 111 (2007) 1998e2001. [44] D. Mikulski, K. Eder, M. Molski, Comput. Theor. Chem. 1046 (2014) 118e125. [45] A. Ebrahimi, P. Karimi, F.B. Akher, R. Behazin, N. Mostafavi, Mol. Phys. 112 (2014) 1047e1056. [46] J. Souza, R.H. de Almeida Santos, M.M.C. Ferreira, F.A. Molfettaa, A.J. Camargo, rio, A.B.F. da Silva, Eur. J. Med. Chem. 38 (2003) 929e938. K.M. Hono [47] J. Sridhar, J. Liu, M. Foroozesh, C.L.K. Stevens, Molecules 17 (2013) 9283e9305. [48] M.J. Frisch, et al., Gaussian 09, Revision A.02, Gaussian Inc., Wallingford, CT, 2009. [49] Y. Zhao, D.G. Truhlar, Theor. Chem. Acc. 120 (2008) 215e241. [50] S. Grimme, J. Comput. Chem. 25 (2004) 1463e1473. [51] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 132 (2010) 154104. [52] W. Hujo, S. Grimme, J. Chem. Theory Comput. 9 (2012) 308e315. [53] L.A. Burns, A. Vazquez-Mayagoitia, B.G. Sumpter, C.D. Sherrill, J. Chem. Phys. 134 (2011) 84107e84114. [54] R.F.W. Bader, Atoms in Molecules: a Quantum Theory, Oxford University Press, Oxford, 1990. [55] K.F. Biegler, J. Schonbohm, D. Bayles, J. Comput. Chem. 22 (2001) 545e559. [56] A.E. Reed, L.A. Curtiss, F. Weinhold, donoreacceptor viewpoint, Chem. Rev. 88 (1988) 899e926. [57] E.D. Glendening, A.E. Reed, J.E. Carpenter, F. Weinhold, NBO Version 3.1. Theoretical Chemistry Institute, University of Wisconsin, Madison, 1990. [58] P. Xu, F. Zahariev, M.S. Gordon, J. Chem. Theory Comput. 10 (2014) 1576e1587. [59] K. Sadasivam, R. Kumaresan, Spectrochim. Acta Part A 79 (2011), pp. 282e2930. [60] R.J. Fehir, J.K. McCusker, J. Phys. Chem. A 113 (2009) 9249e9260. [61] S. Fiorucci, J. Golebiowski, D. Cabrol-Bass, S. Antonczak, J. Agric. Food Chem. 55 (2007) 903e911.