You are on page 1of 12

Journal of Molecular Liquids 359 (2022) 119230

Contents lists available at ScienceDirect

Journal of Molecular Liquids


journal homepage: www.elsevier.com/locate/molliq

Cu(II) complexes of flavonoids in solution: Impact of the Cu(II) ion on the


antioxidant and DNA-intercalating properties
Miriama Šimunková, Monika Biela, Marek Štekláč, Andrej Hlinčík, Erik Klein, Michal Malček ⇑
Institute of Physical Chemistry and Chemical Physics, Faculty of Chemical and Food Technology, Slovak University of Technology in Bratislava, Radlinského 9, SK-812 37,
Bratislava, Slovak Republic

a r t i c l e i n f o a b s t r a c t

Article history: Flavonoids are naturally occurring polyphenolic compounds showing antioxidant and metal chelating
Received 16 February 2022 properties. Chelation of the metal ion, e.g., Cu(II), may significantly affect their activity in biological sys-
Revised 7 April 2022 tems. Due to DNA-intercalating ability, metal complexes of flavonoids can act as therapeutics in the treat-
Accepted 21 April 2022
ment of various human diseases, such as cancer or neurodegenerative diseases. In this work, antioxidant
Available online 27 April 2022
and DNA-intercalating properties of Cu(II)-flavonoid complexes are investigated using spectroscopic
techniques and various computational approaches. The formation of stable Cu(II)-flavonoid complexes
Keywords:
of kaempferol, luteolin, fisetin, and apigenin in two metal:ligand ratios (1:1 and 1:2) is confirmed by
Antioxidant
Copper
UV-Vis spectroscopy. DFT calculations indicate that Cu(II) ion adopts square planar coordination in Cu
DFT (II)-flavonoid (1:2) complexes. Radical scavenging activity of flavonoids in the presence/absence of Cu
DNA intercalation (II) ion is monitored using 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) assay. Except
Flavonoid for apigenin, studied complexes show enhanced radical scavenging activity. Calculation results also imply
Molecular Docking that the presence of Cu(II) ion enhances acidity of OH groups in flavonoids. The molecular docking studies
demonstrate that the formation of Cu(II)-flavonoid complexes (1:1 and/or 1:2) improves the DNA-
intercalating ability of the parent flavonoids.
Ó 2022 Elsevier B.V. All rights reserved.

1. Introduction logically important functionality [8,9,12]. In addition, the


presence of hydroxyl groups in suitable positions also allows the
Flavonoids are polyphenolic compounds naturally occurring in chelation of metal ions, leading to a significant improvement of
plants as secondary metabolites. Their primary function is the pro- the biological activity of parent flavonoid [3,13–15]. Radical scav-
tection against UV radiation and reactive oxygen species (ROS) due enging properties of two flavonoids, kaempferol and luteolin, and
to the high reactivity of their hydroxyl groups [1]. Besides, they of their Cu(II) complexes in different metal:ligand ratios (particu-
possess antioxidant, anti-inflammatory, antibacterial and antiviral larly 1:1 and 1:2), have been recently studied by our research
functions in plants, animals and bacteria [2,3]. In addition, flavo- group using standard 2,20 -azino-bis(3-ethylbenzothiazoline-6-sul
noids show low toxicity to living cells [4], and therefore they are fonic acid) (ABTS) assay [12,16]. ABTS assay is based on the mea-
also important constituents of the human diet [5–7]. It is known surement of the ability of studied compounds to reduce ABTS+
that the daily intake of flavonoids with normal food has beneficial radical cation [17,18]. Its reduction can be monitored by UV-Vis
effects on human health and helps to treat many common diseases spectroscopy as a decrease in the absorbance of ABTS+. It has been
[6]. Due to the properties described above, flavonoids, as well as confirmed that the presence of Cu(II) ion positively affects the rad-
their metal complexes, have found an important role in the treat- ical scavenging activity of the parent flavonoid. Thus, it has been
ment of many human diseases [3,6,8,9]. The basic skeleton of flavo- proposed that the formation of the Cu(II) complexes leads to per-
noids, shown in Fig. 1, consists of two aromatic rings A and B turbations in the p-electron system of the flavonoid moiety. These
connected by heterocyclic ring C. may also influence the reactivity of hydroxyl groups, and conse-
Antioxidant properties of flavonoids primarily depend on the quently the stability of the formed phenoxy radical or phenoxide
number and positions of hydroxyl groups in their molecular struc- anion itself, which is closely connected with the antioxidant prop-
ture [2,3,5–7,10,11], especially B ring (see Fig. 1) manifests a bio- erties of the parent flavonoid.
In our previous study, we have investigated Cu(II) complexes
⇑ Corresponding author. with kaempferol, luteolin, fisetin and apigenin in a metal:ligand
E-mail address: michal.malcek@stuba.sk (M. Malček). ratio of 1:1 [19]. The molecular docking calculations have revealed

https://doi.org/10.1016/j.molliq.2022.119230
0167-7322/Ó 2022 Elsevier B.V. All rights reserved.
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

continuous variations was used [20]. The experiment was per-


formed using UV-Vis-NIR spectrometer Shimadzu UV3600 (Japan)
at room temperature (pH = 7.2) by measuring the spectra of mix-
tures containing different Cu(II) and FL fractions.
The ABTS assay was performed following procedure published
by Re and coworkers [21]. The working solution of ABTS+ radical
cation (pH = 7.2) was prepared at room temperature 24 h before
the measurement. ABTS aqueous solution (7 mM) was mixed with
aqueous solution of potassium persulfate (2.45 mM) in equivolume
ratio to ensure full oxidation of ABTS salt. The following day, 1 mL
of the pre-formed stock aqueous solution was diluted with phos-
phate buffer saline (pH = 7.2) to reach the final volume of 60 mL.
The working solution of ABTS+ was kept in the dark. The absor-
Fig. 1. Basic flavonoid structure including ring denotation (A, B, C) and carbon atom bance of the reference system (1800 lL of ABTS+ + 200 lL DMSO)
numbering. was measured at 734 nm in quartz cuvettes at room temperature
for a period of 900 s using UV-Vis-NIR spectrometer (Shimadzu).
The amount of DMSO was kept constant in all measured systems.
that these complexes are able to effectively intercalate into the
The absorbance of the sample mixture was measured as follows:
DNA structure. This finding suggests that they can be used as per-
1800 lL of ABTS+ solution and 25 lL of stock solution of particular
spective therapeutic agents in anticancer treatment [19]. The
flavonoid (kaempferol, luteolin, fisetin and apigenin, c = 110–4 M)
scheme of the studied set of flavonoids including their Cu(II) bind-
were mixed in a cuvette and filled with DMSO to reach the total
ing sites is shown in Fig. 2.
volume of 2000 lL. For the investigation of Cu(II) ion effect, the
In the presented work, the formation of Cu(II) complexes with
same procedure was repeated with the addition of copper(II) chlo-
kaempferol, luteolin, fisetin and apigenin in metal:ligand ratio of
ride stock solution to obtain molar ratio (Cu(II)-FL=1:1/1:2). The
1:2 is studied using UV-Vis spectroscopy, density functional theory
inhibition percentages were calculated according to the equation
(DFT), Bader’s quantum theory of atoms in molecules (QTAIM) and
molecular docking. Besides, the impact of the Cu(II) coordination AABTSþ  Asample
on the OA-H bond strengths of the particular flavonoid, being in %INH ¼  100% ð1Þ
AABTSþ
relation to the antioxidant properties, is monitored via three (inde-
pendent) parameters: i) delocalization index (DI), ii) bond dissoci- where AABTS+ is the absorbance of the ABTS+ reference and
ation enthalpy (BDE), and iii) proton affinity (PA) and electron Asample is the ABTS+ absorbance with sample mixture over a given
transfer enthalpy (ETE). The calculated data are compared with analytical period (900 s).
the results of the ABTS assay.
2.2. DFT calculations
2. Experimental
Optimized geometries of the studied flavonoids and their Cu(II)
2.1. Chemicals and spectroscopic measurements complexes (metal:ligand ratio 1:1) were taken from our previous
work [19]. Geometry optimizations of all Cu(II)-FL complexes
ABTS diammonium salt, potassium persulfate, phosphate buffer (metal:ligand ratio 1:2) were performed using B3LYP functional
saline pH = 7.2, dimethylsulfoxide (DMSO), copper(II) chloride and [22–25] including the D3 version of Grimme’s dispersion correc-
studied flavonoids (kaempferol, luteolin, fisetin, and apigenin) tion (GD3) [26] and 6-311G** basis set [27,28]. All calculations
were obtained from Sigma (Merck) (Germany). were carried out using the Gaussian16 program package [29] and
The interaction between the studied flavonoids (FL) and Cu(II) the default SCF procedure was employed (the energy-based crite-
ion was monitored by changes in their UV-Vis spectra. The electron rion of the SCF convergence was 10–8 Hartree) [30]. Vibrational
spectra were recorded using UV-Vis-NIR spectrophotometer analysis was employed to confirm that the optimal geometries cor-
AvaSpec-2048x14-USB2 (Avantes) in 3 ml quartz cuvettes respond to energy minima (i.e., the absence of imaginary vibra-
(HellmaÒ Analytics) at room temperature. DMSO was used as a sol- tions). Solvent effects of DMSO and methanol were described
vent for the preparation of the stock solutions of the studied flavo- using Integral Equation Formalism Polarizable Continuum solvent
noids and CuCl2, both with the final concentration of 1 mM. Prior to Model (IEF-PCM) [31–33] implemented in Gaussian16 program
measurement, 150 ll of CuCl2 stock solution was mixed with package. These two solvents were selected because of the availabil-
150/300 ll of stock solution of a particular flavonoid (Cu(II)- ity of spectroscopic data [12,19].
FL=1:1/1:2) and after 2 minutes filled with phosphate buffer saline The basis set superposition error (BSSE) corrected EBSSE
bind complex-
(pH = 7.2) to reach the final volume of 2000 ll. To determine the ation energies of the Cu(II)-FL complexes (metal:ligand ratio 1:2)
metal:ligand ratio of formed Cu(II)-FL complex, Job’s method of were calculated using the equation

Fig. 2. Scheme of the studied flavonoids: a) kaempferol, b) luteolin, c) fisetin, d) apigenin, and their potential Cu(II) binding sites (Cu1–Cu3).

2
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

 
EBSSE BSSE
bind ¼ EFLCuFL  EFL þ EBSSE BSSE
Cu þ EFL ð2Þ 3. Results and discussion

where EFL-Cu-FL is the energy of Cu(II)-FL (1:2) complex, EBSSE


FL and 3.1. UV-Vis spectroscopy
EBSSE
Cu are the BSSE corrected energies of particular flavonoid moieties
and Cu(II), respectively, obtained via the counterpoise scheme of Since flavonoids are known to be poorly soluble in water,
Boys and Bernardi [34]. DMSO, applicable also for biochemical application, was used. Cu
The QTAIM analysis [35] implemented in the AIMAll package (II) ion was allowed to interact with studied flavonoids in the form
[36] was performed for every relevant system under study using of CuCl2 aqueous solution two minutes prior the measurement. The
the Gaussian16 checkpoint files. The QTAIM method was employed spectra presented in the Fig. 3 are typical for polyphenolic mole-
to estimate character and strength of the CuAO and OAH bonds cules characterized by p?p* electronic transitions [15,48,49].
using properties in the bond critical points (BCP) and delocaliza- The interaction between the studied flavonoids and Cu(II) ion is
tion indexes (DI). Visualization of the optimized structures was documented by marked spectral shifts pointing on the formation
performed using Molekel software suite [37]. of Cu(II)-FL complexes in the case of all four studied systems
The thermochemistry of the OAH bonds cleavage, i.e., bond dis- (kaempferol, luteolin, fisetin and apigenin). These changes are a
sociation enthalpy (BDE), proton affinity (PA) and electron transfer consequence of disturbed electronic structure of flavonoid moiety
enthalpy (ETE), were determined via B3LYP-GD3/6-311++G** cal- upon Cu(II) chelation. The presented UV-Vis spectra (Fig. 3) of Cu
culated total enthalpies (H) using the following equations (II)-FL complexes are similar to the ones in methanol reported in
our previous work [19].
BDE = H(FLO ) + H(H ) —H(FLOH) ð3Þ A Job’s method was applied to study the stereochemistry of the
Cu(II)-FL interaction and the determination of the metal:ligand
ratio. The Job’s plots derived from UV-Vis measurements (see
PA = H(FLO— ) + H(Hþ ) —H(FLOH) ð4Þ Fig. S1) confirmed an approximate metal:ligand ratio of 1:2 for
Cu(II)-kaempferol and Cu(II)-luteolin complexes. On the other
hand, in the case of Cu(II)-fisetin and Cu(II)-apigenin complexes
ETE = H(FLO ) + H(e— ) —H(FLO— ) ð5Þ preferred metal:ligand ratio of 1:1 was found, see Fig. S1. Never-
where FL represents the moiety linked to OH group in flavonoid or theless, the formation of a minor amount of the Cu(II)-fisetin
Cu(II)-FL complex. The enthalpies of proton H(H+) and electron H and/or Cu(II)-apigenin complexes in metal:ligand ratio of 1:2 in
(e–), taken from the work of Rimarčík et al. [38] were determined solution cannot be excluded [13]. This is also supported by the
as a sum of their gas-phase enthalpy and the corresponding solva- observation of different types of spectral shifts in UV-Vis spectra
tion enthalpy in DMSO (or methanol). The gas-phase enthalpy of of Cu(II)-fisetin and/or Cu(II)-apigenin complexes.
proton is 6.197 kJ mol–1 and the gas-phase enthalpy of electron is
3.145 kJ mol–1 [39]. Note that the application of basis set containing 3.2. ABTS assay
diffuse functions (6-311++G**) is essential for the correct treatment
of anions [40]. The radical scavenging capacity of studied compounds in the
absence/presence of Cu(II) ions, were investigated using the stan-
2.3. Molecular docking dard ABTS assay [18], see Fig. 4. The main purpose of this experi-
ment was to shed light on the effect of Cu(II) on the radical
Semi-flexible docking of the Cu(II)-FL complex (metal:ligand scavenging activity of flavonoids after the destabilization of their
ratio 1:2) with the highest Boltzmann population (BP) ratio was p-electron system upon Cu(II) coordination.
performed for each flavonoid using AutoDock4.2.6 suite [41,42]. As it can be derived from the results of the ABTS assay (Fig. 4
QTAIM charges and Mulliken charge distribution were utilized to and Table 1), the presence of Cu(II) ion enhances the radical scav-
preserve the net positive charge of the complexes. Calf thymus enging activity of kaempferol, luteolin and fisetin. In the case of
DNA (CT-DNA) dodecamer d(CGCGAATTCGCG) (PDB ID: 1bna) kaempferol and fisetin, the highest time decay of ABTS+ absor-
[43] was downloaded from RSCB database [44] to serve as the tar- bance is found for the Cu(II)-FL (1:2) mixture, while in the case
get structure. CT-DNA is frequently used in experimental DNA of luteolin, the highest decay of ABTS+ absorbance is found for
intercalation studies [45,46] owing to its relative ease of extraction the Cu(II)-luteolin (1:1) complex. In contrast, apigenin alone shows
from thymus gland of calf, and its use allows mutual comparison of better radical scavenging ability than its Cu(II) complexes, see
results. Fig. 4d. A similar decrease in the radical scavenging activity of api-
To ensure the comparability of the results with our previous genin in the presence of Cu(II) ion has been recently reported by Xu
work [19], the computational details specifying the docking proto- et al. [13].
cols were kept largely the same. Grid box of 75  75  126 points Considering the results of radical scavenging properties moni-
with the resolution of 0.325 Å was used to calculate the potential tored by UV-Vis measurements, we have focused our further atten-
maps. The docking calculations were performed using the Lamar- tion on the DFT, QTAIM and molecular docking studies of Cu(II)-FL
ckian genetic algorithm, with the number of docking runs set to complexes in metal:ligand ratio of 1:2. Cu(II)-FL complexes in
100. The maximum number of energy evaluations was set to 30 metal:ligand ratio of 1:1 have been studied in detail in our previ-
106, while each generation contained 300 individuals and the ous work [19].
maximum number of generations was restrained to 27 000. Local
search parameters were kept at their default values. The resulting
poses were sampled with 2.0 Å tolerance. The putative hydrogen 3.3. Conformation analysis of Cu(II)-FL (1:2) complexes
bond patterns were visualized, both three- and two-
dimensionally, using AutoDockTools [41,42] and LigPlot+ [47], Molecular structure of flavonoids, i.e., the presence of hydroxyl
respectively. The free energy of binding (DG) was further used to and keto functional groups, offers several metal chelation sites, see
express binding constants at room temperature (T = 298.15 K), Fig. 2. The optimized geometries of the potential conformations of
using the equation the Cu(II)-FL complexes with the metal:ligand ratio 1:2 are com-
piled in Figs. S2–S5. In all presented conformations, the Cu(II)
K b ¼ exp ðDG=RT Þ ð5Þ adopts a square planar coordination environment typical for cupric
3
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Fig. 3. UV-Vis spectra of FLs: a) kaempferol (KAM), b) luteolin (LUT), c) fisetin (FIS), d) apigenin (API) dissolved in DMSO (red line) before and after the mixing with CuCl2 using
molar ratios Cu(II)-FL = 1:1 (green line) and 1:2 (blue line), in phosphate buffer pH = 7.2. Spectra were measured 2 min after the mixing of the CuCl2 solution with flavonoids
[Cu(II)] = [FL] = 0.1 mM.

ion [50]. Calculated relative energies (in kJ mol–1), BP ratios and Last but not least, as it can be seen in Table 2, there are more con-
BSSE corrected complexation energies (EBSSE
bind ,
in kJ mol ) of the –1 formations with nonnegligible Boltzmann populations in the cases
studied systems are presented in Table 2. As it can be seen in of Cu(II)-kaempferol and Cu(II)-fisetin, in particular: Cu1-Cu2-cis-
Table 2, the trans conformation via Cu1 site is the energetically pre- kaempferol (ca 17%) and Cu2-cis-fisetin (ca 11%). Even though the
ferred mode of the Cu(II)-complex formation in the case of kaemp- existence of these two conformations in solution cannot be
ferol, luteolin and apigenin, with BP ratios of 77%, 100% and 100%, excluded, our further attention will be focused only on the energet-
respectively. In the case of fisetin, which lacks Cu1 site, the most ically most stable conformations (see Table 2 and Fig. 5).
preferred mode of binding is the trans conformation via Cu2 site
(BP ratios are 88% in DMSO and 89% in methanol). These results
3.4. QTAIM analysis of bond critical points
are further supported by the most negative values of the calculated
complexation energies (EBSSE
bind ), ranging from –1024 to –1057 As already mentioned above, Cu(II) ion in the studied com-
kJ mol–1, see bold values in Table 2. The more negative EBSSE bind value plexes adopts the distorted square planar coordination environ-
indicates the energetically more favorable complex formation ment. Besides a binding distance, being a natural measure of
and consequently the stronger binding to Cu(II) ion. When consid- bond strength, the Bader’s QTAIM analysis offers several parame-
ering only the most stable conformations, the strength of Cu(II) ters which are related to the nature of chemical bond. Under the
binding decreases in the order: luteolin  apigenin > kaempferol Bader’s formalism, the BCP is a saddle point of electron density
> fisetin. Interestingly, slightly stronger Cu(II) binding is found in between the two atoms forming a chemical bond [35]. Conse-
the case of flavonoids lacking 3 OH group (luteolin and apigenin). quently, the charge density in BCP (qBCP) and its Laplacian (DqBCP)
The data shown in Table 2 correspond well with the found ener- are direct measures of the bond strength. Further useful parameter,
getic preference of the Cu(II) binding sites reported for Cu(II)-FL DI is a measure of the number of electrons shared between the
complexes in metal:ligand ratio of 1:1 [19]. In this work, we have atoms forming a chemical bond [51]. Calculated CuAO bond dis-
reported that the binding sites of the rings A and C (Cu1 and Cu2) tances, BCP characteristics and DI values for the most stable Cu
are more favored when compared to those of the ring B (Cu3) [19]. (II)-FL complexes are presented in Table 3. A comparison to rele-
Optimized geometries of the energetically preferred Cu(II)-FL com- vant Cu(II)-FL systems in metal:ligand ratio of 1:1 [19] is also
plexes are shown in Fig. 5, including their BP ratios. given.
4
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Fig. 4. Time decay of absorbance of the reference ABTS+ radical cation (black dotted line) measured at 734 nm at pH = 7.2 after addition of particular flavonoid (red line) a)
kaempferol, b) luteolin, c) fisetin, d) apigenin and Cu(II)-FL mixture in two molar ratios (1:2 - blue dashed line; 1:1 - green dash-dotted line) (for details regarding the
concentrations of the working solutions and experimental setup, see Section 2.2.).

Table 1 delocalized over the four coordinating oxygens. For completeness,


The percentage of ABTS+ scavenging capacity of the studied flavonoids and their Cu the spin density distributions of the studied complexes are shown
(II) complexes in two molar ratios (1:1 and 1:2).
in Fig. S6.
System FL Cu(II)-FL (1:1) Cu(II)-FL (1:2) Further investigated functionalities are the OAH bonds of the
Kaempferol 47 53 55 hydroxyl groups, which are associated with the antioxidant activ-
Luteolin 36 43 41 ity of polyphenolic compounds [2,10–12,16]. As mentioned above,
Fisetin 54 57 61 DI value, corresponding to the number of electrons shared between
Apigenin 96 77 83
the two atoms, represents a direct measure of bond strength [51].
Calculated DI values of OAH bonds of studied flavonoids and their
Cu(II) complexes in different metal:ligand ratios (1:1 and 1:2) are
The results summarized in Table 3 are very similar to each shown in Table 4. As it can be seen from Table 4, the calculated
other. The Cu(II) is coordinated via two equally strong bonds with DI values of the OAH bonds of studied flavonoids and their Cu(II)
oxygens of keto groups (DI values around 0.5, distances around complexes have well defined trends: the OAH bond strength
1.9 Å), and with two slightly weaker bonds with oxygens of decreases after the coordination of one FL moiety to the Cu(II)
hydroxyl groups (DI values around 0.35, distances around 2.0 Å) ion, but the coordination of the second FL molecule leads to an
in all studied complexes. Naturally, the CuAO bonds of Cu(II)-FL increase of DI values. The only exception is the 5 OAH bond, where
(1:1) complexes are found to be stronger than those of Cu(II)-FL the trend is opposite: the DI values increase upon the coordination
(1:2) complexes when considering the BCP parameters and DI of Cu(II) ion, which can be explained by the orientation of the 5 OH
values (Table 3). The calculated QTAIM charge and spin densities group in the optimized structures. While in the case of FL moiety,
at the Cu atom in all studied Cu(II)-FL (1:2) complexes are the 5 OH group is oriented towards the neighboring C4=O keto
approximately 1.3e and 0.7e, respectively, which indicates that group (forming intramolecular H-bond), in the case of Cu(II)-FL
most of the unpaired electron remains localized at Cu center. the OAH bond is oriented in the opposite direction due to the pres-
The rest of the unpaired electron, being approximately 0.3e, is ence of the Cu(II) ion, see Fig. 5.

5
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Table 2 fer of electron, is characterized by ETE [38,40,52–57]. Calculated


Calculated B3LYP-GD3/6-311G**/PCM=DMSO relative energies (in kJ mol–1), BP ratios BDE, PA and ETE values of the OAH bonds of studied flavonoids
and BSSE corrected binding energies (in kJ mol–1) of the studied Cu(II)-FL (1:2)
complexes. Values obtained in methanol solvent model (PCM=methanol) are given in
and their Cu(II) complexes in different metal:ligand ratios (1:1
parentheses. For Cu atom binding sites labeling, see Fig. 2. The energetically preferred and 1:2) are shown in Tables 5 and 6.
structures are set in bold. Results compiled in Table 5 imply that the calculated BDE val-
Flavonoid rel. energy / kJ mol–1 BP ratio / %
ues have well-defined trends in all studied systems for both sol-
EBSSE
bind / kJ mol
–1

Cu binding site vent models (DMSO and methanol). In particular, the BDEs of the
kaempferol
3’, 4’, and 7 OH groups increase in the order: FL < Cu(II)-FL (1:1)
Cu1, cis 16.6 (16.9) 0.1 (0.1) –1020.8 (–1033.6) < Cu(II)-FL (1:2), while the BDEs of the 3 OH and 5 OH groups fol-
Cu1, trans 0.0 (0.0) 76.3 (77.4) –1031.8 (–1034.3) low the trend: FL > Cu(II)-FL (1:2) > Cu(II)-FL (1:1). Therefore, it can
Cu2, cis 11.2 (11.7) 0.8 (0.7) –994.3 (–998.1) be concluded that the presence of Cu(II) ion in the vicinity of C4=O
Cu2, trans 6.8 (7.1) 4.8 (4.5) –997.4 (–999.8)
keto group positively affects the ability of neighboring OH groups
Cu1-Cu2, ‘‘cis” 3.7 (3.8) 17.2 (16.7) –1011.3 (–1008.3)
Cu1-Cu2, ‘‘trans” 11.4 (11.8) 0.8 (0.7) –1005.3 (–1020.5) to undergo homolytic cleavage. In addition, the presence of Cu(II)
luteolin ion is able to alter thermodynamically preferred OH group (see
Cu1, cis 21.8 (22.3) 0.0 (0.0) –1044.4 (–1049.4) bold values in Table 5). In the case of luteolin, fisetin and apigenin,
Cu1, trans 0.0 (0.0) 100 (100) –1057.3 (–1062.0) the lowest BDE value is always ascribed to the B ring (i.e., 3’ OH and
Cu3, cis 127.0 (130.6) 0.0 (0.0) –853.6 (–859.2)
4’ OH group, respectively), while in the case of their Cu(II)-FL (1:2)
Cu3, trans 127.4 (131.1) 0.0 (0.0) –854.3 (–860.1)
Cu1-Cu3, ‘‘cis” 61.1 (62.8) 0.0 (0.0) –963.8 (–965.8) complexes, the lowest BDE values are ascribed to either 3 OH or 5
Cu1-Cu3, ‘‘trans” 60.6 (62.4) 0.0 (0.0) –962.5 (–964.2) OH group. The only exception is kaempferol, where the lowest BDE
fisetin value is found for the 3 OH group of kaempferol itself as well as for
Cu2, cis 4.8 (5.2) 12.4 (10.9) –1020.0 (–1024.0)
its Cu(II) complexes. Last but not least, the impact of solvent effects
Cu2, trans 0.0 (0.0) 87.6 (89.1) –1024.2 (–1027.1)
Cu3, cis 108.8 (111.0) 0.0 (0.0) –856.2 (–855.3) on the calculated BDE values can be considered negligible [38,54–
Cu3, trans 109.1 (111.3) 0.0 (0.0) –864.8 (–863.4) 56].
Cu2-Cu3, ‘‘cis” 50.8 (52.0) 0.0 (0.0) –952.5 (–953.5) Abovementioned data imply that the coordination of Cu(II) ion
Cu2-Cu3, ‘‘trans” 50.6 (51.8) 0.0 (0.0) –952.0 (–953.2) via OH groups of the B ring (Cu3 site, see Fig. 2) by luteolin or fise-
apigenin
tin may lead to a decrease in BDE values of these OH groups (3’ OH
Cu1, cis 21.7 (22.3) 0.0 (0.0) –1038.9 (–1048.6)
Cu1, trans 0.0 (0.0) 100 (100) –1051.9 (–1062.1) and 4’ OH). However, such mode of Cu(II) binding is not energeti-
cally favorable, see Table 2.
Similar as BDEs, the calculated PA values have well defined
With the exception of the 5 OAH bond, it can be concluded that trends for both studied solvents. Note that, the solvent effects play
the formation of Cu(II) complex leads to a weakening of OAH a non-negligible role when evaluating PAs of phenolic compounds
bonds of the parent FL which is reflected in the decrease in their [38,54–56]. The calculated PA values decrease mainly in the order:
DI values, see Table 4. In the case of Cu(II)-luteolin complexes, FL > Cu(II)-FL (1:2) > Cu(II)-FL (1:1), see Table 6. Extremely low PA
the decrease in DI(OAH) values follows the order: luteolin > Cu values in the case of the Cu(II)-FL (1:1) complexes indicate the
(II)-luteolin (1:2) > Cu(II)-luteolin (1:1). This is in perfect agree- existence of these species in the forms of phenoxide anions in
ment with an increase in the ABTS+ scavenging activity: 36% lute- the solution (especially in DMSO). Moreover, as in the case of BDEs,
olin < 41% Cu(II)-luteolin (1:2) < 43% Cu(II)-luteolin (1:1). In the the formation of Cu(II)-FL complex alters thermodynamically pre-
case of Cu(II)-kaempferol and Cu(II)-fisetin complexes, the calcu- ferred OH groups. In the case of FL moieties, the lowest PA values
lated trends in DI values are only in partial agreement with the are ascribed either to OH groups of B ring or to 7 OH group. On the
reported ABTS assays. Still, the weakening of OAH bonds as well other hand, their Cu(II) complexes have the lowest PA values for 5
as enhanced radical scavenging activity upon the formation of Cu or 3 OH groups in both studied metal:ligand ratios. Overall, the
(II) complexes are confirmed. DI(OAH) values of apigenin and its data presented in Table 6 suggest that the heterolytic cleavage of
Cu(II) complexes follow the same trends as other systems under the OAH bonds in studied flavonoids is more preferred than the
study. However, these trends do not agree with the found decrease homolytic one when the Cu(II) ion is present in solution.
in ABTS+ scavenging activity of Cu(II)-apigenin complexes when For completeness, the calculated ETE values are shown in
compared to free apigenin, see Table 1. Table 6 as well. Calculated ETE values show negligible differences
for the two studied environments. This suggests that these two sol-
vents do not have a notable effect on ETE. The values in Table 6
3.5. Thermodynamics of OAH bond cleavage indicate that these enthalpies are relatively high. In general, it
can be observed that low PA values are supplemented by high
Besides the DI value, three thermochemical parameters are con- ETE values and vice versa [40,52,53]. The calculated ETE values
sidered to estimate the antioxidant properties of the studied Cu(II)- decrease in the following order: Cu(II)-FL (1:1) > Cu(II)-FL (1:2) >
FL complexes. BDE, PA and ETE represent the reaction enthalpies of FL, which corresponds to the above-mentioned trend of PA values.
homolytic and heterolytic bond cleavage followed by successive
electron transfer, respectively, see Equations (3-5). As it is already 3.6. Molecular Docking
mentioned in the Introduction section, the B ring of flavonoids is an
important functionality in the radical scavenging process, operat- Computational approaches describing biological effects are usu-
ing via hydrogen-atom or sequential proton loss — electron trans- ally focused on examining relations between computable proper-
fer to radical species [12]. Hydrogen atom transfer (HAT) ties and biological activity of the compound. However, it is
represents a single-step radical (homolytic) cleavage of the OAH equally important to consider the extent to which the selected
bond, which is governed by the OAH BDE [38,40,52–57]. Heteroly- compound binds to the receptor. Molecular docking can be
tic cleavage of the OAH bond, i.e., formation of the phenoxide employed for this purpose, providing energetical and structural
anion, is the first step of the sequential proton-loss electron- information about the binding process and the formation of
transfer (SPLET) mechanism which also plays an important role ligand-receptor complex [58–61]. Even though there is an incon-
in metal chelation [5]. Its first step is governed by the PA sistency in free binding energies (docking scores) acquired with
[38,40,52–57]. The second step of SPLET mechanism, i.e., the trans- different scoring functions implemented by various software pack-
6
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

a) BP 76.3% (77.4%) b) BP 100% (100%)

c) BP 87.6% (89.1%) d) BP 100% (100%)


Fig. 5. B3LYP-GD3/6-311G**/PCM=DMSO optimized geometries of the most stable Cu(II)-FL complexes (metal:ligand ratio 1:2): a) kaempferol, b) luteolin, c) fisetin, d)
apigenin. Values obtained in methanol (PCM=methanol) are given in parentheses.

ages [58], molecular docking allows users to compare binding being able to fully intercalate into DNA grooves. Although Cu(II)-
affinity of compounds calculated with the same docking protocol. luteolin (1:2) and Cu(II)-fisetin (1:2) complexes show mild affinity
Hence, the free binding energies and putative hydrogen bond pat- towards the major DNA groove, their binding poses prevent any
terns of Cu(II)-FL complexes (1:2) with CT-DNA calculated in this interactions with DNA bases. Cu(II)-kaempferol (1:2), Cu(II)-
work are compared with results of Cu(II)-FL complexes (1:1) apigenin (1:2) complexes are merely located in proximity of one
obtained in similar manner [19]. Studied complexes were split into strand of the DNA helix, see Fig. S7. Two-dimensional depictions
two sets, based on the charge distribution used to calculate the of putative binding patterns of studied compounds with QTAIM
ligand partial atomic charges, i.e., Mulliken and QTAIM. charge distribution with CT-DNA can be found in Fig. S8.
It has been revealed that the charge distribution has a drastic This observation is in agreement with the docking poses of Cu
impact on the outcome of the docking protocol, i.e., the free energy (II)-FL complexes (1:1) [19] with this charge distribution and it is
of binding and the predicted pose. The docking calculations with likely the result of high charge concentration on atoms, strength-
QTAIM ligand partial atomic charges resulted in neither compound ening electrostatic interactions with negatively charged phosphate

7
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Table 3
Calculated B3LYP/6-311G**/PCM=DMSO CuAO bond lengths d and corresponding QTAIM BCP characteristics (charge density qBCP, Laplacian DqBCP and delocalization index DI) of
the studied Cu(II)-FL (1:2) systems. For Cu atoms labeling see Fig. 2. A comparison to reference Cu(II)-FL (1:1) complexes is given.

Cu(II)-FL (1:2) d(Cu-O) / Å qBCP(Cu-O)/ bohr-3 DqBCP(Cu-O)/ bohr-5 DI(Cu-O)/ –


kaempferol
Cu1, trans 1.88, 1.88 0.099, 0.099 0.578, 0.582 0.492, 0.494
BP = 76.3% 1.99, 1.99 0.073, 0.072 0.421, 0.431 0.345, 0.345
luteolin
Cu1, trans 1.87, 1.87 0.102, 0.102 0.595, 0.592 0.508, 0.507
BP = 100% 1.98, 1.99 0.074, 0.073 0.437, 0.432 0.345, 0.346
fisetin
Cu2, trans 1.89, 1.90 0.099, 0.099 0.545, 0.545 0.506, 0.505
BP = 87.6% 2.02, 2.02 0.069, 0.069 0.382, 0.383 0.344, 0.344
apigenin
Cu1, trans 1.87, 1.87 0.102, 0.102 0.594, 0.592 0.508, 0.507
BP = 100% 1.98, 1.98 0.074, 0.073 0.437, 0.433 0.346, 0.347
Reference Cu(II)-FL (1:1)a
kaempferol-Cu1a 1.90a, 2.13a 0.095a, 0.056a 0.586a, 0.275a 0.491a, 0.269a
luteolin-Cu1a 1.83a, 1.96a 0.116a, 0.080a 0.639a, 0.475a 0.658a, 0.390a
fisetin-Cu2a 1.88a, 2.04a 0.103a, 0.067a 0.555a, 0.367a 0.586a, 0.339a
apigenin-Cu1a 1.82a, 1.94a 0.121a, 0.084a 0.659a, 0.507a 0.726a, 0.413a
a
B3LYP/6-311G**/PCM=DMSO values from Šimunková et al.[19]

Table 4
Calculated B3LYP-GD3/6-311G**/PCM=DMSO delocalization indexes (DI) of OAH bonds in FLs and Cu(II)-FL complexes (metal:ligand ratio 1:1 and 1:2). Comparison to ABTS+
scavenging activity (Table 1) is given.

Kaempferol Bond FL Cu(II)-FL (1:1) Cu(II)-FL (1:2)


ABTS+ decay / % 47 53 55
DI value 4’ O-H 0.637 0.606 0.630
3 O-H 0.578 0.533 0.576
5 O-H 0.510 0.577 0.554
7 O-H 0.631 0.612 0.617
Luteolin
ABTS+ decay / % 36 43 41
DI value 3’ O-H 0.621 0.610 0.617
4’ O-H 0.605 0.590 0.599
5 O-H 0.492 0.547 0.553
7 O-H 0.632 0.612 0.617
Fisetin
ABTS+ decay / % 54 57 61
DI value 3’ O-H 0.624 0.605 0.615
4’ O-H 0.608 0.586 0.597
3 O-H 0.573 0.553 0.560
7 O-H 0.631 0.617 0.620
Apigenin
ABTS+ decay / % 96 77 83
DI value 4’ O-H 0.634 0.624 0.628
5 O-H 0.493 0.541 0.553
7 O-H 0.632 0.611 0.617

Table 5
Calculated B3LYP-GD3/6-311++G**/PCM=DMSO BDE of OAH bonds in FLs and Cu(II)-FL complexes (metal:ligand ratio 1:1 and 1:2). Values obtained in methanol solvent
(PCM=methanol) are given in parentheses. The lowest value for each molecule is set in bold.

Kaempferol Bond FL Cu(II)-FL (1:1) Cu(II)-FL (1:2)


–1
BDE / kJ mol 4’ O-H 336 (336) 353 (354) 395 (396)
3 O-H 324 (324) 306 (306) 351 (352)
5 O-H 378 (379) 356 (356) 379 (379)
7 O-H 357 (358) 393 (394) 441 (442)
Luteolin
BDE / kJ mol–1 3’ O-H 325 (325) 335 (335) 419 (419)
4’ O-H 336 (336) 337 (338) 416 (417)
5 O-H 393 (394) 378 (378) 403 (403)
7 O-H 364 (364) 422 (423) 488 (488)
Fisetin
BDE / kJ mol–1 3’ O-H 321 (321) 338 (338) 396 (395)
4’ O-H 326 (326) 337 (338) 392 (392)
3 O-H 327 (327) 270 (270) 327 (326)
7 O-H 355 (355) 398 (398) 448 (448)
Apigenin
BDE / kJ mol–1 4’ O-H 348 (348) 368 (368) 454 (454)
5 O-H 393 (394) 370 (370) 404 (404)
7 O-H 365 (365) 416 (416) 494 (494)

8
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Table 6
Calculated B3LYP-GD3/6-311++G**/PCM=DMSO PAs and ETEs of OAH bonds in FLs
and Cu(II)-FL complexes (metal:ligand ratio 1:1 and 1:2). Values obtained in
methanol (PCM=methanol) are given in parentheses. The lowest value for each
molecule is set in bold.

Kaempferol Bond FL Cu(II)-FL (1:1) Cu(II)-FL (1:2)


PA / kJ mol–1 4’ O-H 89 (167) –31 (43) 98 (95)
3 O-H 103 (182) –40 (34) 95 (92)
5 O-H 118 (197) –56 (17) 22 (19)
7 O-H 81 (159) –12 (61) 117 (115)
ETE / kJ mol–1 4’ O-H 370 (367) 507 (509) 497 (499)
3 O-H 344 (340) 469 (470) 455 (459)
5 O-H 383 (379) 534 (536) 557 (558)
7 O-H 400 (396) 528 (530) 524 (525)
Luteolin
PA / kJ mol–1 3’ O-H 81 (160) –63 (10) 104 (97)
4’ O-H 89 (168) –51 (23) 139 (109)
5 O-H 132 (211) –74 (0) 30 (27)
7 O-H 83 (162) –17 (57) 117 (115)
ETE / kJ mol–1 3’ O-H 366 (363) 521 (523) 515 (520)
4’ O-H 369 (366) 512 (513) 477 (506)
5 O-H 384 (380) 575 (576) 573 (574)
7 O-H 404 (400) 562 (564) 570 (570)
Fisetin
PA / kJ mol–1 3’ O-H 87 (165) –65 (8) 78 (75)
4’ O-H 97 (176) –52 (21) 91 (87)
3 O-H 113 (192) –103 (–30) 21 (18)
7 O-H 81 (160) –28 (45) 123 (121)
ETE / kJ mol–1 3’ O-H 357 (353) 526 (528) 517 (519)
4’ O-H 352 (348) 512 (514) 501 (502)
3 O-H 337 (333) 496 (498) 505 (506)
7 O-H 397 (393) 548 (551) 525 (525)
Apigenin
PA / kJ mol–1 4’ O-H 84 (162) –56 (18) 60 (135)
5 O-H 132 (211) –91 (–18) –48 (26)
7 O-H 84 (162) –34 (40) 40 (115)
ETE / kJ mol–1 4’ O-H 387 (383) 546 (548) 517 (517)
5 O-H 384 (380) 584 (586) 575 (576)
7 O-H 404 (400) 572 (574) 577 (577)

moieties of sugar-phosphate backbone. This is further illustrated


by high contribution of intermolecular electrostatic interactions
towards free energies of binding, see Table S1.
Contrastingly, docking of studied complexes with Mulliken
charge distribution suggest their ability to intercalate into the
minor DNA grooves, see Fig. 6. The two-dimensional depictions
of putative binding patterns of studied compounds with Mulliken
charge distribution with CT-DNA can be found in Fig. S9. The pre-
dicted binding poses of all four complexes show that both FL moi-
eties are found within the minor groove, as opposed to interacting
with just one FL moiety and the second one being located outside
of the groove, i.e., the binding poses change with the introduction Fig. 6. Docking poses of Cu(II)-FL complexes (metal:ligand ratio 1:2) with Mulliken
charge distribution bound to CT-DNA: a) kaempferol, b) luteolin, c) fisetin, d)
of the second FL substituent. Kaempferol and apigenin Cu(II) com-
apigenin. Deoxy cytosine (DC) is yellow, deoxy guanine (DG) is blue, deoxy adenine
plexes have almost identical binding modes, with the former being (DA) is red and deoxy thymine (DT) is orange.
additionally stabilized by the interaction between the ligand
hydrogen at O3 and OP1 (DG10, see Fig. S9) group. This hydroxy lated free energies of binding of Cu(II) complexes with Schiff base
moiety is absent in apigenin, resulting in slightly higher free ligands to CT-DNA are within the range from –5.7 to –7.0 kcal mol–
1
energy of binding. Although the Cu(II)-fisetin complex binds to (i.e., from –23.8 to –29.3 kJ mol–1) [49,62].
the same region, its binding pose is a reflection in a mirror plane The values of free binding energies, contributions of van der
passing through the copper ion and separating both flavonoid Waals/hydrogen and electrostatic interactions, as well as more
units. Furthermore, the angles between B and A/C ring planes detailed description of the predicted bonding pattern for Mulliken
appear to be greater than in the cases of kaempferol and apigenin charge distribution can be found in Table 7.
complexes, see Fig. S9. This is likely a result of an additional The free energies of binding of ligand-DNA complexes with
hydroxy group in the B ring (3’ OH and 4’ OH), which does not QTAIM charge distribution on ligand atoms have the following
lie on the axis joining rings B and C. Therefore, the rotation of B ring trend with respect to metal:ligand ratio: FL > Cu(II)-FL (1:2) > Cu
along this axis offers more means of complex stabilization, and the (II)-FL (1:1), see Fig. 7a. The first decrease in docking score can
Cu(II)-fisetin complex exhibits the highest number of predicted be explained by presence of Cu(II), which increases net charge on
hydrogen bonds. In contrast, luteolin binds to different CT-DNA the ligand, leading to stronger intermolecular electrostatic interac-
regions interacting with oxygens in the sugar-phosphate backbone tions. The increase from 1:1 to 1:2 ratio is likely a result of spatial
while having the most negative free binding energy of DG = –43.60 constraints introduced by the second FL moiety leading to greater
kJ mol–1 (Kb = 4.34107). For comparison, recently reported calcu- distance between interacting atomic charges.

9
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Table 7
Energies related to binding process of Cu(II)-FL complexes (metal:ligand ratio 1:2) with DNA (kJ mol–1), binding constants (Kb) and hydrogen bond details (bond length (Å) and
bond angle (°)) of compound-DNA complexes for Mulliken charge distribution. Atom numbers refer to coordinating carbon while nucleobases to which the atoms belong are in
parentheses (1—12 corresponds to chain A, 13—24 corresponds to chain B).

Cu(II)-FL (1:2) DGa Eb1 Ec2 Kb Hydrogen Bonding


Compound CT-DNA Bond length Bond angle
Kaempferol –41.67 –34.52 -16.36 2.00107 3 OH OP1(DG10) 1.903 136.944
O7 NH22(DG16) 1.940 123.191
7 OH O4’(DA17) 2.069 123.931
Luteolin –43.60 –34.61 –16.19 4.34107 7 OH O4’(DA6) 2.160 129.826
9’ OH O4’(DC9) 2.111 162.919
3’ OH OP1(DC23) 1.762 176.112
Fisetin –41.88 –34.64 –16.40 2.17107 7 OH O2(DT7) 2.003 174.839
9’ OH O4’(DC11) 1.800 162.322
O10’ NH22(DG16) 1.917 153.127
10’ OH O4’(DA17) 1.951 159.789
4’ OH OP1(DT20) 2.105 158.504
Apigenin –39.50 –31.51 –14.85 8.31106 O7 NH22(DG16) 1.795 120.271
7 OH O4’(DA17) 1.880 134.791
a
DG is the estimated free energy of binding calculated using AutoDock scoring function.
b
E1 is the sum of intermolecular van der Waals energy, hydrogen bonding energy and desolvation free energy.
c
E2 is the intermolecular electrostatic energy.

Fig. 7. Free binding energies of FLs and Cu(II)-FL complexes (metal:ligand ratio 1:1 and 1:2) with CT-DNA: a) QTAIM ligand partial atomic charges, b) Mulliken ligand partial
atomic charges. DG values of FL and Cu(II)-FL (1:1) complexes are taken from Šimunková et al. [19].

On the other hand, the complexes with Mulliken charge distri- such small magnitude that its observation is beyond limits of
bution on ligand atoms have no conclusive trend with respect to molecular docking approaches. Therefore, a more rigorous
metal:ligand ratio and are different for each flavonoid, see approach is required for corroboration of the experimental results
Fig. 7b. Kaempferol and luteolin show slight decrease in the free presented in Šimunková et al. [12] and Jomová et al. [16].
energies of binding with two coordinating flavonoid units, while
fisetin exhibits minor increase. It should be noted that these dis- 4. Conclusion
tinctions are minute and well within the scoring function error
[41]. The only notable change in the free energies of binding is that The antioxidant and DNA-intercalating activity of the Cu(II)-FL
of apigenin, whose affinity towards DNA decreases upon coordina- complexes in solution have been studied using UV-Vis spec-
tion of copper by the second FL moiety. troscopy, ABTS assay, DFT calculations and molecular docking pro-
It is apparent that ligand charges have dramatic effects on dock- tocol. Obtained UV-Vis spectra confirmed the formation of Cu(II)-
ing accuracy [60]. Unlike the Mulliken charge distribution, Cu(II)- FL complexes in different metal:ligand ratios (1:1 and 1:2). Accord-
FL complexes (1:2) with QTAIM charges have not been predicted ing to DFT calculations, in the case of Cu(II)-FL (1:2) complexes, the
to intercalate into DNA. This observation agrees with our previous Cu(II) ion is coordinated by four oxygens from two FL moieties ori-
calculations of complexes with 1:1 metal:ligand ratio [19]. Inspec- ented in trans conformation. Cu(II) ion adopts the square planar
tion of the free energies of binding (binding constants) of com- coordination environment in all studied (1:2) complexes and the
plexes with Mulliken charges reveals that changes linked to strongest binding is found in the case of interaction with luteolin
metal:ligand ratio are within the calculation error [41]. Therefore, (calculated complexation energy is –1057 kJ mol–1).
it can be surmised that the impact of the second flavonoid moiety The ABTS assay showed that the presence of Cu(II) ion enhances
in Cu(II)-FL complex on its binding affinity towards CT-DNA is of the radical scavenging activity of kaempferol, luteolin and fisetin.
10
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

Such an increase in radical scavenging activity of these FLs upon Appendix A. Supplementary material
the Cu(II) coordination is further confirmed by the weakening of
their OAH bonds, monitored via decreased DI(OAH) values. On Supplementary data to this article can be found online at
the other hand, the coordination of Cu(II) decreases the radical https://doi.org/10.1016/j.molliq.2022.119230.
scavenging activity of apigenin. Moreover, the presence of Cu(II)
alters the thermodynamic preference of OAH bonds of FLs to References
undergo homolytic and/or heterolytic cleavage, as it is documented
by the changes in the calculated BDE and PA values. Naturally, the [1] C. Brunetti, A. Fini, F. Sebastiani, et al., Modulation of Phytohormone Signaling :
OH groups in the vicinity of the coordinated Cu(II) are the most A Primary Function of Flavonoids in Plant – Environment Interactions, Front
Plant Sci 9 (2018) 1042, https://doi.org/10.3389/fpls.2018.01042.
affected ones. Extremely low calculated PA values point out that
[2] S. Kumar, A.K. Pandey, Chemistry and Biological Activities of Flavonoids : An
the formed Cu(II)-FL complexes easily undergo deprotonation in Overview, Sci World J 162750 (2013).
solution. [3] M. Grazul, E. Budzisz, Biological activity of metal ions complexes of chromones,
coumarins and flavones, Coord Chem Rev 253 (2009) 2588–2598, https://doi.
In addition, molecular docking studies confirmed the ability of
org/10.1016/j.ccr.2009.06.015.
Cu(II)-FL complexes to interact with DNA molecule via intercala- [4] G. Galati, P.J. O’Brien, Potential toxicity of flavonoids and other dietary
tion. Interestingly, the calculated free binding energies of the inter- phenolics: Significance for their chemopreventive and anticancer properties,
action between Cu(II)-FL complexes and DNA are comparable for Free Radic Biol Med 37 (2004) 287–303, https://doi.org/10.1016/j.
freeradbiomed.2004.04.034.
both 1:1 and 1:2 metal:ligand ratios (all DG values are around – [5] I. Gülçin, Antioxidant activity of food constituents: An overview, Arch Toxicol
40 kJ mol–1, when Mulliken charge distribution is considered). Fur- 86 (2012) 345–391, https://doi.org/10.1007/s00204-011-0774-2.
thermore, it has been observed that Cu(II)-FL complexes with par- [6] B.H. Havsteen, The biochemistry and medical significance of the flavonoids,
Pharmacol Ther 96 (2002) 67–202, https://doi.org/10.1016/S0163-7258(02)
tial atomic charges calculated using QTAIM approach show no 00298-X.
ability to intercalate into the DNA grooves. This contrasts with _ Gulcin, Antioxidants and antioxidant methods: an updated overview, Arch
[7] I.
complexes characterized by Mulliken charges, which are all found Toxicol 94 (3) (2020) 651–715.
[8] K. Jomová, L. Hudecova, P. Lauro, M. Simunkova, S.H. Alwasel, I.M. Alhazza, M.
intercalated in the minor DNA groove. Possible explanation for this Valko, A switch between antioxidant and prooxidant properties of the phenolic
behavior comes from high QTAIM charge concentration on atoms compounds myricetin, morin, 30 , 40 -dihydroxyflavone, taxifolin and 4-
favoring electrostatic interactions with negatively charged phos- hydroxy-coumarin in the presence of copper (II) ions: a spectroscopic,
absorption titration and DNA damage study, Molecules 24 (23) (2019) 4335.
phate moieties of sugar-phosphate backbone. Contrastingly, [9] E. Rodríguez-Arce, M. Saldías, Antioxidant properties of flavonoid metal
charges calculated using Mulliken population analysis do not exhi- complexes and their potential inclusion in the development of novel
bit such charge extremes (very high or low partial charges on cer- strategies for the treatment against neurodegenerative diseases, Biomed
Pharmacother 143 (2021), https://doi.org/10.1016/j.biopha.2021.112236
tain atoms) and are stabilized by intermolecular van der Waals
112236.
interactions inside the DNA groove. [10] K.E. Heim, A.R. Tagliaferro, D.J. Bobilya, Flavonoid antioxidants: chemistry,
The presented results demonstrate that the binding of Cu(II) ion metabolism and structure-activity relationships, J. Nutr. Biochem. 13 (10)
by flavonoids significantly enhances their antioxidant as well as (2002) 572–584.
[11] S.V. Jovanovic, S. Steenken, M. Tosic, B. Marjanovic, M.G. Simic, Flavonoids as
DNA-intercalating properties. This suggests potential application antioxidants, J Am Chem Soc 116 (11) (1994) 4846–4851.
of these compounds as effective and perspective therapeutics in [12] M. Simunkova, Z. Barbierikova, K. Jomova, et al., Antioxidant vs. Prooxidant
the treatment of neurodegenerative disorders or as DNA- Properties of the Flavonoid, Kaempferol, in the Presence of Cu (II) Ions : A ROS-
Scavenging Activity, Fenton Reaction and DNA Damage Study, Int J Mol Sci 22
damaging agents e.g., in cancer therapy. (1619) (2021), https://doi.org/10.3390/ijms22041619.
[13] Y.i. Xu, J. Yang, Y. Lu, L.-L. Qian, Z.-Y. Yang, R.-M. Han, J.-P. Zhang, L.H. Skibsted,
Copper(II) Coordination and Translocation in Luteolin and Effect on Radical
CRediT authorship contribution statement Scavenging, J Phys Chem B 124 (2) (2020) 380–388.
[14] C.E. Lekka, J. Ren, S. Meng, E. Kaxiras, Structural, Electronic, and Optical
Properties of Representative Cu - Flavonoid Complexes, J Phys Chem B 113
Miriama Šimunková: Conceptualization, Methodology, Writing (2009) 6478–6483, https://doi.org/10.1021/jp807948z.
– original draft. Monika Biela: Investigation, Writing – original [15] J.M. Dimitrić Marković, Z.S. Marković, T.P. Brdarić, V.M. Pavelkić, M.B. Jadranin,
Iron complexes of dietary flavonoids: Combined spectroscopic and
draft. Marek Štekláč: Conceptualization, Methodology, Writing – mechanistic study of their free radical scavenging activity, Food Chem 129
original draft. Andrej Hlinčík: Investigation. Erik Klein: Investiga- (4) (2011) 1567–1577.
tion, Writing – original draft. Michal Malček: Conceptualization, [16] K. Jomova, L. Hudecova, P. Lauro, M. Simunková, Z. Barbierikova, M. Malcek, S.
H. Alwasel, I.M. Alhazza, C.J. Rhodes, M. Valko, The effect of luteolin on DNA
Methodology, Supervision, Writing – original draft.
damage mediated by a copper catalyzed Fenton reaction, J. Inorg. Biochem.
226 (2022).
[17] N.J. Miller, C. Rice-Evans, M.J. Davies, V. Gopinathan, A. Milner, A Novel
Declaration of Competing Interest Method for Measuring Antioxidant Capacity and its Application to Monitoring
the Antioxidant Status in Premature Neonates, Clin Sci 84 (4) (1993) 407–412.
[18] N. Echegaray, M. Pateiro, P.E.S. Munekata, J.M. Lorenzo, Z. Chabani, M.A. Farag,
The authors declare that they have no known competing finan- R. Domínguez, Measurement of antioxidant capacity of meat and meat
cial interests or personal relationships that could have appeared products: Methods and applications, Molecules 26 (13) (2021) 3880.
to influence the work reported in this paper. [19] M. Šimunková, M. Štekláč, M. Malček, Spectroscopic, computational and
molecular docking study of Cu(ii) complexes with flavonoids: From cupric ion
binding to DNA intercalation, New J Chem 45 (2021) 10810–10821, https://
doi.org/10.1039/d1nj01960k.
Acknowledgement [20] J.S. Renny, L.L. Tomasevich, E.H. Tallmadge, D.B. Collum, Method of continuous
variations: Applications of job plots to the study of molecular associations in
This work received financial support from Slovak Grant Agen- organometallic chemistry, Angew Chemie - Int Ed 52 (2013) 11998–12013,
https://doi.org/10.1002/anie.201304157.
cies APVV (contract No. APVV-17-0513, APVV-19-0087 and [21] Re R, Pellegrini N, Proteggente A, et al (1999) Development and
APVV-20-0213) and VEGA (contracts No. 1/0139/20, 1/0482/20, characterisation of carbon nanotube-reinforced polyurethane foams. Free
1/0504/20, 1/0461/21 and 1/0078/21). MM is also grateful to the Radic Biol Med 26:1231–1237. https://doi.org/doi.org/10.1016/S0891-5849
(98)00315-3
HPC center at the Slovak University of Technology in Bratislava,
[22] A.D. Becke, Density-functional exchange-energy approximation with correct
which is a part of the Slovak Infrastructure of High Performance asymptotic behavior, Phys Rev A 38 (6) (1988) 3098–3100.
Computing (SIVVP project, ITMS code 26230120002, funded by [23] C. Lee, W. Yang, R.G. Parr, Development of the Colic-Salvetti Correlation-
the European region development funds) for the computational Energy Formula into a Functional of the Electron D, Phys Rev B 37 (1988) 785–
789.
time and resources made available. To all financing sources the [24] A.D. Becke, Density-functional thermochemistry.III. The role of exact
authors are greatly indebted. exchange, J Chem Phys 98 (7) (1993) 5648–5652.

11
M. Šimunková, M. Biela, M. Štekláč et al. Journal of Molecular Liquids 359 (2022) 119230

[25] S.H. Vosko, L. Wilk, M. Nusair, Accurate spin-dependent electron liquid [46] J.E.N. Dolatabadi, S. Kashanian, A review on DNA interaction with synthetic
correlation energies for local spin density calculations: a critical analysis, Can J phenolic food additives, Food Res Int 43 (2010) 1223–1230, https://doi.org/
Phys 58 (1980) 1200–1211, https://doi.org/10.1139/p80-159. 10.1016/j.foodres.2010.03.026.
[26] S. Grimme, Semiempirical GGA-Type Density Functional Constructed with a [47] A.C. Wallace, R.A. Laskowski, J.M. Thornton, LIGPLOT: a program to generate
Long-Range Dispersion Correction, J Comput Chem 27 (2006) 1787–1799, schematic diagrams of protein-ligand interactions The LIGPLOT program
https://doi.org/10.1002/jcc. automatically generates schematic 2-D representations of protein-ligand
[27] R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, Self-consistent molecular orbital complexes from standard Protein Data Bank file input, Protein Eng 8 (1995)
methods. XX. A basis set for correlated wave functions, J Chem Phys 72 (1) 127–134, https://doi.org/10.1093/protein/8.2.127.
(1980) 650–654. [48] E.H. Anouar, J. Gierschner, J.L. Duroux, P. Trouillas, UV/Visible spectra of
[28] A.D. McLean, G.S. Chandler, Contracted Gaussian basis sets for molecular natural polyphenols: A time-dependent density functional theory study, Food
calculations. I. Second row atoms, Z=11–18, J Chem Phys 72 (10) (1980) 5639– Chem 131 (2012) 79–89, https://doi.org/10.1016/j.foodchem.2011.08.034.
5648. [49] S. Gurusamy, K. Krishnaveni, M. Sankarganesh, R.N. Asha, A. Mathavan,
[29] Frisch MJ, Trucks GW, Schlegel HB, et al (2016) Gaussian 16, Revision C.01 Synthesis, characterization, DNA interaction, BSA/HSA binding activities of VO
[30] K.N. Kudin, G.E. Scuseria, E. Cancès, A black-box self-consistent field (IV), Cu (II) and Zn (II) Schiff base complexes and its molecular docking with
convergence algorithm: One step closer, J Chem Phys 116 (2002) 8255– biomolecules, Journal of Molecular Liquids 345 (2022), https://doi.org/
8261, https://doi.org/10.1063/1.1470195. 10.1016/j.molliq.2021.117045.
[31] S. Miertus, E. Scrocco, J. Tomasi, Electrostatic Interaction of a Solute with a [50] E.W. Dahl, N.K. Szymczak, Hydrogen Bonds Dictate the Coordination Geometry
Continuum. A Direct Utilization of ab Initio Molecular Potentials for the of Copper: Characterization of a Square-Planar Copper(I) Complex, Angew
Prevision of Solvent Effects, Chem Phys 55 (1981) 117–129. Chemie 128 (2016) 3153–3157, https://doi.org/10.1002/ange.201511527.
[32] J. Tomasi, B. Mennucci, R. Cammi, Quantum Mechanical Continuum Solvation [51] R.F.W. Bader, M.E. Stephens, Spatial localization of the electronic pair and
Models, Chem Rev 105 (2005) 2999–3093, https://doi.org/10.1021/cr9904009. number distributions in molecules, J Am Chem Soc 97 (1975) 7391–7399,
[33] J. Tomasi, B. Mennucci, E. Cancès, The IEF version of the PCM solvation https://doi.org/10.1021/ja00859a001.
method: an overview of a new method addressed to study molecular solutes at [52] A. Vagánek, J. Rimarčík, K. Dropková, J. Lengyel, E. Klein, Reaction enthalpies of
the QM ab initio level, J Mol Struct 464 (1-3) (1999) 211–226. OH bonds splitting-off in flavonoids: The role of non-polar and polar solvent,
[34] S.F. Boys, F. Bernardi, The calculation of small molecular interactions by the Comput Theor Chem 1050 (2014) 31–38.
differences of separate total energies. Some procedures with reduced errors, [53] M. Biela, A. Kleinová, E. Klein, Thermodynamics of radical scavenging effect of
Mol Phys 19 (1970) 553–566, https://doi.org/10.1080/00268977000101561. deprotonated isoflavones in aqueous solution, J Mol Liq 345 (2022), https://
[35] R.F.W. Bader, Atoms in Molecules: A Quantum Theory, Oxford University Press, doi.org/10.1016/j.molliq.2021.117861 117861.
Oxford, 1990. [54] Y.Z. Zheng, G. Deng, D.F. Chen, et al., The influence of C2[dbnd]C3 double bond
[36] Keith TA AIMAll, version 14.04.17; TK Gristmill Software: Overland Park, KS, on the antiradical activity of flavonoid: Different mechanisms analysis,
2014 (aim.tkgristmill.com) Phytochemistry 157 (2019) 1–7, https://doi.org/10.1016/
[37] S. Portmann, H.P. Luthi, MOLEKEL: An interactive molecular graphics tool, j.phytochem.2018.10.015.
Chimia (Aarau) 54 (2000) 766–770. [55] Y.Z. Zheng, G. Deng, R. Guo, et al., The influence of the H5  O[dbnd]C4
[38] J. Rimarčík, V. Lukeš, E. Klein, M. Ilčin, Study of the solvent effect on the intramolecular hydrogen-bond (IHB) on the antioxidative activity of flavonoid,
enthalpies of homolytic and heterolytic N-H bond cleavage in p- Phytochemistry 160 (2019) 19–24, https://doi.org/10.1016/
phenylenediamine and tetracyano-p-phenylenediamine, J Mol Struct j.phytochem.2019.01.011.
THEOCHEM 952 (2010) 25–30, https://doi.org/10.1016/j. [56] H. Boulebd, Comparative study of the radical scavenging behavior of ascorbic
theochem.2010.04.002. acid, BHT, BHA and Trolox: Experimental and theoretical study, J Mol Struct
[39] J.E. Bartmess, Thermodynamics of the Electron and the Proton, J Phys Chem 98 1201 (2020), https://doi.org/10.1016/j.molstruc.2019.127210 127210.
(1994) 6420–6424, https://doi.org/10.1021/j100076a029. [57] A. Galano, G. Mazzone, R. Alvarez-Diduk, T. Marino, J.R. Alvarez-Idaboy, N.
[40] M. Biela, J. Rimarčík, E. Senajová, A. Kleinová, E. Klein, Antioxidant action of Russo, Food Antioxidants: Chemical Insights at the Molecular Level, Annu Rev
deprotonated flavonoids: Thermodynamics of sequential proton-loss electron- Food Sci Technol 7 (1) (2016) 335–352.
transfer, Phytochemistry 180 (2020), https://doi.org/10.1016/ [58] T. Cheng, X. Li, Y. Li, Z. Liu, R. Wang, Comparative assessment of scoring
j.phytochem.2020.112528. functions on a diverse test set, J Chem Inf Model 49 (4) (2009) 1079–1093.
[41] G.M. Morris, R. Huey, W. Lindstrom, et al., AutoDock4 and AutoDockTools4: [59] Y. Li, L. Han, Z. Liu, R. Wang, Comparative assessment of scoring functions on
Automated Docking with Selective Receptor Flexibility, J Comput Chem 30 an updated benchmark: 2. Evaluation methods and general results, J Chem Inf
(2009) 2785–2791, https://doi.org/10.1002/jcc. Model 54 (2014) 1717–1736, https://doi.org/10.1021/ci500081m.
[42] M.F. Sanner, Python: a programming language for software integration and [60] M.A. Husain, H.M. Ishqi, S.U. Rehman, T. Sarwar, S. Afrin, Y. Rahman, M. Tabish,
development, J Mol Graph Model 17 (1999) 57–61. Elucidating the interaction of sulindac with calf thymus DNA: Biophysical and:
[43] H.R. Drew, R.M. Wing, T. Takano, C. Broka, S. Tanaka, K. Itakura, R.E. Dickerson, In silico molecular modelling approach, New J Chem 41 (24) (2017) 14924–
Structure of a B-DNA dodecamer: conformation and dynamics, Proc Natl Acad 14935.
Sci U S A 78 (4) (1981) 2179–2183. [61] R.N. Asha, M. Sankarganesh, N. Bhuvanesh, B.R.D. Nayagam, Synthesis,
[44] H.M. Berman, T. Battistuz, T.N. Bhat, W.F. Bluhm, P.E. Bourne, K. Burkhardt, Z. structural, spectral, antidiabetic, DNA interactions and molecular docking
Feng, G.L. Gilliland, L. Iype, S. Jain, P. Fagan, J. Marvin, D. Padilla, V. investigations of a piperidine derivative, J Mol Struct 1250 (2022), https://doi.
Ravichandran, B. Schneider, N. Thanki, H. Weissig, J.D. Westbrook, C. org/10.1016/j.molstruc.2021.131692 131692.
Zardecki, The protein data bank, Acta Crystallogr Sect D Biol Crystallogr 58 [62] G.S. Senthilkumar, M. Sankarganesh, J. Dhaveethu Raja, P.R. Adwin Jose, A.
(6) (2002) 899–907. Sakthivel, T. Christopher Jeyakumar, R. Nandini Asha, Water soluble Cu(II) and
[45] G. Psomas, Mononuclear metal complexes with ciprofloxacin: Synthesis, Zn(II) complexes of bidentate-morpholine based ligand: synthesis, spectral,
characterization and DNA-binding properties, J Inorg Biochem 102 (2008) DFT calculation, biological activities and molecular docking studies, J Biomol
1798–1811, https://doi.org/10.1016/j.jinorgbio.2008.05.012. Struct Dyn 40 (3) (2022) 1074–1083.

12

You might also like