Pyridoxal derivatized copper(II) complexes: Evaluation of antioxidant, catecholase, and DNA cleavage activity

Article abstract of DOI:10.1016/j.ica.2017.09.063

This manuscript describes the synthesis, characterization and structural analysis of copper(II) complexes with pyridoxal-type bidentate ligands, obtained from the condensation between pyridoxal and ortho-halogenated substituted anilines (C1–C5). The complexes C1, C2 and C5 were measured by X-ray single crystal analysis which showed classic and trifurcated hydrogen bonds and interaction as a sigma-hole between the oxygen atoms of the iodine ligand derivative in the complex C5. All compounds were tested as mimetics of superoxide dismutase (SOD), through NBT photoreduction method in aqueous solution of pH 7.8, but the complex C5 showed an expressive IC₅₀ of 0.4 μM. In addition, catecholase-like activity of C1–C5 in the presence of the substrate 3,5-di-tert-butylcatechol and cleavage of plasmid DNA were also evaluated. Again, the best results were associated with the complex C5 containing iodine in the ortho position.

Full text of DOI:10.1016/j.ica.2017.09.063

Accepted Manuscript
Research paper
Pyridoxal derivatized copper(II) complexes: Evaluation of antioxidant, cate-
cholase, and DNA cleavage activity.
Mateus Brum Pereira, Liniquer Andre Fontana, Josiéli Demetrios Siqueira,
Bruna L. Auras, Marcos P. da Silva, Ademir Neves, Philipe Gabriel, Hernán
Terenzi, Bernardo Almeida Iglesias, Davi Fernando Back
PII:
S0020-1693(17)30615-1
https://doi.org/10.1016/j.ica.2017.09.063
ICA 17922
DOI:
Reference:
To appear in:
Inorganica Chimica Acta
Received Date:
Revised Date:
Accepted Date:
24 April 2017
7 July 2017
29 September 2017
Please cite this article as: M.B. Pereira, L.A. Fontana, J.D. Siqueira, B.L. Auras, M.P. da Silva, A. Neves, P. Gabriel,
H. Terenzi, B.A. Iglesias, D.F. Back, Pyridoxal derivatized copper(II) complexes: Evaluation of antioxidant,
catecholase, and DNA cleavage activity., Inorganica Chimica Acta (2017), doi: https://doi.org/10.1016/j.ica.
2
017.09.063
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Pyridoxal derivatized copper(II) complexes: Evaluation of antioxidant, catecholase, and DNA cleavage
activity.
a
a
a
b
b
b
Mateus Brum Pereira , Liniquer Andre Fontana , Josiéli Demetrios Siqueira , Bruna L. Auras , Marcos P. da Silva , Ademir Neves ,
c
c
a
a
Philipe Gabriel , Hernán Terenzi , Bernardo Almeida Iglesias and Davi Fernando Back*
a
Departamento de Química, Laboratório de Materiais Inorgânicos, Universidade Federal de Santa Maria, UFSM, 97115-900 Santa Maria - RS, Brazil.
b
Departamento de Química, Universidade Federal de Santa Catarina, Florianópolis, SC 88040-900, Brazil.
c
Centro de Biologia Molecular Estrutural, Departamento de Bioquímica, Universidade Federal de Santa Catarina, Florianópolis, SC 88040-900, Brazil.
1

ABSTRACT
This manuscript describes the synthesis, characterization and structural analysis of copper(II) complexes with pyridoxal-type
bidentate ligands, obtained from the condensation between pyridoxal and ortho-halogenated substituted anilines (C1-C5). The
complexes C1, C2 and C5 were measured by X-ray single crystal analysis which showed classic and trifurcated hydrogen bonds
and interaction as a sigma-hole between the oxygen atoms of the iodine ligand derivative in the complex C5. All compounds were
tested as mimetics of superoxide dismutase (SOD), through NBT photoreduction method in aqueous solution of pH 7.8, but the
complex C5 showed an expressive IC50 of 0.4µM.
2

In addition, catecholase-like activity of C1-C5 in the presence of the substrate 3,5-di-tert-butylcatechol and cleavage of
plasmid DNA were also evaluated. Again, the best results were associated with the complex C5 containing iodine in the ortho
position.
Keywords: Pyridoxal Schiff bases; copper(II) complexes; SOD mimetics, catecholase activity, plasmid DNA cleavage.
3

1
. INTRODUCTION
Pyridoxal acts as a coenzyme and many studies have demonstrated their performance in amino acid biosynthesis in
transamination and decarboxylation, as well as in racemization processes. In biological systems, after conversion to pyridoxal-5'-
phosphate, pyridoxine is incorporated into a variety of enzymes that catalyze these reactions [1,2]. When the pyridoxal unit is
derivatized (mainly involving condensation reactions) Schiff-base ligands are obtained. There are numerous examples in the
literature that illustrate the coordination with uranium, vanadium [3-5], iron, nickel, cobalt [6-8] and palladium [10], as well as p-block
metals such as tin [9]. However, to date, there are only a few reports about the applicability of complexes involving the pyridoxal
moiety and its antioxidant activity, specifically mimetics of superoxide dismutase (SOD) [11]. This group of enzymes can catalyze
the dismutation of the superoxide anion, converting it into oxygen and hydrogen peroxide [12]. In the human body these types of
antioxidant enzymes are divided into those that exhibit Cu/Zn or Mn metal ion centers [13-14].
There is also a lack of information in the literature about pyridoxal complexes with the potential to bind and cleave nucleic
acids. Even though interaction of DNA with complexes of many transition metal ions has been investigated [15], copper(II)
complexes are the preferred molecules for bringing about DNA cleavage. This is due to the fact that copper(II) complexes can bring
about not only oxidative DNA cleavage, but also hydrolytic, photolytic or electrolytic cleavage [16-19]. Copper(II) complexes are also
reported in the literature as potential antioxidant [20-22] antitumor, antimicrobial and anti-inflammatory molecules [23].
In this context, this study aimed to synthesize and characterize copper(II) complex derivatives from pyridoxal and ortho-
substituted haloanilines and evaluate them with regard to potential as SOD biomimetics (SOD-1), catecholase activity and
interaction with plasmid DNA.
4

2
. EXPERIMENTAL
2
.1. General Instrumentation
All manipulations were conducted by use of standard argon atmosphere. CHN% elemental analyses were performed at a
Shimadzu EA112 microanalysis instrument. Compounds were analyzed using a mass spectrometer with electrospray ionization
ESI-MS) in the positive mode using an Amazon X Ion Trap (Bruker Daltonics, Billerica, MA). Mass spectra were recorded with the
methanolic solutions of around 500 ppb concentration with a flow of 180 µL/min and capillary of 3500 V. High resolution ESI-MS
HRMS-ESI) was performed on a micrOTOF QII mass spectrometer (Bruker Daltonics, Billerica, MA). IR spectra were recorded on
(
(
-1
a Tensor 27 Bruker spectrometer with KBr pellets in the 4000-400 cm region. UV-vis spectra were recorded on a UV-2600
Shimadzu spectrometer. Cyclic voltammograms were recorded with an EcoChemie PGSTAT 32N system at room temperature and
under argon atmosphere, in dry acetonitrile solution. Electrochemical grade tetrabutylammonium hexafluorophosphate (TBAPF )
6
was used as supporting electrolyte (0.1 mol/L acetonitrile solution). These CV experiments were carried out by employing a
standard three-component system: a glassy carbon working electrode; a platinum wire auxiliary electrode and a platinum wire
pseudo-reference electrode. To monitor the reference electrode, the ferrocenium/ferrocene couple was used as an internal
reference [24].
5

2
.2. X-Ray crystallography
Data were collected on a Bruker D8 Venture Photon 100 diffractometer equipped with an Incoatec IµS high brilliance Mo Kα X-
ray tube with two-dimensional Montel micro-focusing optics. The structure was solved by direct methods using SHELXS [25].
Subsequent Fourier-difference map analyses yielded the positions of the non-hydrogen atoms. Refinements were carried out with
2
the SHELXL package [25]. All refinements were made by full-matrix least-squares on F with anisotropic displacement parameters
for all non–hydrogen atoms. Hydrogen atoms were included in the refinement in calculated positions but the atoms (of hydrogens)
that are commenting performing special bond were located in the Fourier map. Drawings were done using DIAMOND for Windows
[
26]. Crystal data and more details of the data collection and refinements of the complexes C1, C2 and C5 are presented in Table
1
.
2
.3. DNA Cleavage assays
The plasmid pBSK-II (Stratagene, USA) was used in the cleavage tests. This plasmid was produced in Escherichia coli DH5-
α, following the protocol recommended by Ausubel and co-workers [27] and described in detail by Oliveira and collaborators [28].
DNA was extracted and purified following the protocol for extraction and purification of plasmid DNA (HiSpeed Plasmid Maxi Kit-
Qiagen). The extracted plasmid DNA concentration was quantified using UV spectrophotometry and its integrity verified by agarose
gel electrophoresis [27,28].
6

A typical cleavage reaction (20 µL) contained 330 ng of DNA pBSK II, buffered with 10 mM HEPES (pH 7.0) and treated with
complexes at various concentrations. The reaction time was 16 hours at 37 ºC. During the reaction, the samples were kept
sheltered from light. To interrupt the cleavage reactions, 5 µL of sample buffer (0.25 M EDTA, pH 8.0, 50% glycerol and 0.01%
bromophenol blue) were added. The samples were then kept refrigerated (4ºC) until subjected to 1% agarose gel electrophoresis,
containing ethidium bromide (0.3 µg/mL) for 100 minutes at 90 V with TBE 0.5X running buffer (44.5 mM Tris, 44.5 mM boric acid
and 1 mM EDTA at pH 8.0). The gels were documented using a DigiDoc-It (UVP, USA) system and the fraction of each plasmid
DNA form quantified by densitometry, using KODAK Molecular Imaging Software 5.0 (Carestream Health, USA). Since ethidium
bromide intercalates to a lesser extent to the supercoiled form of the plasmid DNA, a correction factor of 1.47 was used when
measuring this topological form of plasmid DNA [29].
2
.4. ROS scavenger’s effect
DNA cleavage can occur by, at least, two distinct mechanisms: hydrolytic or oxidative [30]. To evaluate the nature of such
cleavage mechanisms, experiments were conducted in the presence of reactive oxygen species scavengers, such as 0.5 mM KI, a
1
peroxide generation inhibitor; and 0.5 mM NaN
3
, a singlet oxygen scavenger species ( O
2
).
2
.5. DNA groove binders: effect on DNA cleavage
7

To determine the interaction specificity for the minor or major DNA grooves during the cleavage process, two small molecules
known for binding with each groove were alternately added 30 minutes before the addition of complex, being them netropsin, a
classical minor groove binder [31], and methyl green, a major groove binder [32].
2
.6. DNA cleavage kinetics assays
Cleavage kinetics experiment is a key point to compare the efficiency of different catalyzers. The assays were performed using
a final volume of 120 µL, to which 2 µg of plasmid DNA in 10 mM HEPES (pH 7.0) were added, followed by 30 µL of copper(II)
complex (250 µM). Aliquots of 20 µL were taken at different times (0, 2, 4, 8 and 24 hours) and subjected to agarose gel
electrophoresis. Cleavage kinetic constants (kobs) were estimated for each specific concentration, taking these reactions as pseudo
first-order. The value of kobs was obtained directly from the angular coefficient of the linear regressions attained from the plot of the
natural logarithm of the quantity of DNA versus reaction time.
2
.7. Thermal denaturation with CT-DNA
8

Calf-Thymus DNA denaturation can be detected by the raising on light absorption (hyperchromic effect) [33]. The needed
temperature to denature half of the DNA molecules is called denaturation mean temperature (T ). When a molecule interacts with
DNA the T can be altered. Based on such concepts, the T of CT-DNA was calculated in the absence or presence of the most
m
m
m
active complex. The assays were performed in a UV-Visible Ultrospec 2100 spectrophotometer (Amersham Biosciences, USA). CT-
DNA solution (50 µM) untreated or treated with complex (5 µM) remained 10 minutes at the initial start temperature (40 °C). The
temperature was then raised by 1 ºC / minute until 90 ºC. The T
non-linear regression [19,34,35].
m
was calculated using the mean point of the Boltzmann sigmoidal
2
.8. Superoxide dismutase activity assays
The superoxide dismutase activity (SOD activity) of complexes C1, C2, C3, C4 and C5 was determined by measuring the
inhibition of nitrotetrazolium blue chloride (NBT) photoreduction assays, according to a previously described procedure [11,36-42].
-3
The reaction mixture consisted of 2.4 mL of a sodium phosphate buffer solution (pH = 7.8) containing 9.53 x 10 mol/L of
-
5
methionine and 3.8 x 10 mol/L of NBT. Temperature was set to 25 °C. A 50 µL aliquot of a dilute solution containing the complexes
-
6
in DMF was added to the solutions, followed by addition of riboflavin (final concentration = 3.2 x 10 mol/L). All experiments were
conducted in triplicate, included the blank samples (only solvent, no added complex). The solutions were illuminated with a
fluorescent lamp at constant light intensity. NBT reduction was monitored spectrophotometrically at λ = 560 nm as a function of the
illumination period (t). NBT reduction rate in the absence and in the presence of complexes were determined for a range of complex
concentrations.
9

Inhibition percentage was calculated according to {(∆Abs/t)without complex – (∆Abs/t)with complex} x 100/(∆Abs/t)without complex. The IC50
values represent the concentration of the SOD mimic that induces a 50% inhibition of NBT reduction. The catalytic rate constants
4
-1 -1
(
kinetic constants of McCord and Fridovich) were calculated as kMcCF = kNBT [NBT]/IC50, where kNBT (pH = 7.8) = 5.94 x 10 M s
[
43].
2
.9. Catecholase activity
The catalytic activity of the copper(II) complexes was evaluated by the oxidation reaction using 3,5-di-tert-butylcatechol as
substrate (3,5-DTBC). Kinetic experiments were performed in duplicate under substrate excess conditions by monitoring
spectrophotometrically on a UV-vis spectrometer Varian Cary 50 Bio coupled to a thermostated bath, the change in absorbance that
-1
-1
occurred at 400 nm (ε = 1.645 M cm ) due to formation of 3,5-di-tert-butilquinone (3,5-DTBQ). The reactions were monitored from
to 5% of substrate conversion to product and the data were treated by the method of initial rates, which were obtained directly
2
from the substrate concentration versus time.
Optical glass cuvettes were used (4 mL) and optical path of 1.0 cm, closed with a teflon cap, to which was added 50 µL of
-
2
aqueous solution ([T]final = 3.33 x 10 mol/L) buffers: (MES pH = 4.0 to 6.5 and TRIS pH = 7.0 to 10.0), 1100-1380 µL of methanol
-
5
saturated with oxygen and 50 µL of a solution of compound in DMSO ([C]final= 2.67 x 10 mol/L). The reaction was started by
-4
-3
adding 50 to 300 µL of a methanolic solution of the substrate ([S] = 5.88 x 10 to 7.06 x 10 mol/L) and monitored for 10 minutes. In
all experiments, the final volume of the reaction mixture in the cuvette was 1.5 mL, respectively.
1
0

Corrections spontaneous oxidation of the substrate 3,5-DTBC were performed under identical conditions without the addition
of the complex. The initial rates were obtained from the slope of the absorbance versus time curve during the first minutes of the
reaction by the method of initial rates. The formation of hydrogen peroxide in the oxidation reactions of 3,5-DTBC catalyzed by
complexes C1 to C5 was detected by a modification of the iodometric method [44].
-
5
A reaction mixture was prepared similar to kinetic experiments: complex concentration [C]Final = 2.67 x 10 mol/L (pH = 6.5)
-3
and substrate concentration [S]final = 3.33 x 10 mol/L. After one hour of reaction, an equal volume of water was added and the
-3
quinone was extracted with dichloromethane. The aqueous layer was acidified with sulfuric acid ([Acid] = 5 x 10 mol/L) at pH ≅ 2,
to stop the oxidation reaction, and 1.0 mL of aqueous solution of potassium iodide ([iodide] = 0.3 mol/L) were added.
In the presence of hydrogen peroxide is the following reaction:
+
-
H
2
O
2
2 2
+ 2H + 2I → I + 2H O
and iodide ion excess occurs the formation of triiodide ion:
-
-
3
I
2(aq.) + I → I
-
This reaction is usually slow, but in acid solution becomes virtually instantaneous. The formation of triiodide ion (I ) can be
3
-1
-1
monitored spectrophotometrically due to the emergence of a characteristic band at 353 nm (ε = 26.000 M cm ).
2
.10. Ligand synthesis – General procedure (Scheme 1)
1
1

In a round bottom flask, pyridoxal hydrochloride (0.201 g; 1.0 mmol) was added to methanol (15 mL) and the resulting
suspension was subjected to magnetic stirring at room temperature until complete dissolution of the pyridoxal molecule. Then, the
corresponding aniline (0.120 g of 2-fluoroaniline; 0.130 g of 2-chloroaniline, 0.170 g of 2-bromoaniline or 0.220 g of 2-iodoaniline;
1
.0 mmol for each) dissolved in methanol (5 mL) was added. The resulting mixture was kept under magnetic stirring at reflux
temperature for 2.0 h. After this period, the flask contents were cooled to room temperature and the solvent evaporated. The
resulting solid was washed with small portions of ice water and diethyl ether and dried in a desiccator with CaCl . For the synthesis
of the ligand containing aniline, the same conditions were used, but without potassium hydroxide ( 0.093 g of aniline was added to
2
1
5 mL of methanol and 1 mmol of pyridoxal hydrochloride; 0.201 g).
Insert Scheme1 about here
5
2
3
-(hydroxymethyl)-2-methyl-4-(phenylimino)methyl)pyridin-3-ol (L1): Properties - light yellow crystals, yield 73%. Melting point:
-1
23 °C. Anal.Calc. for [C14
H
14
N
2
O ]HCl: C, 60.38; H, 5.43; N, 10.03. Found: C, 60.33; H, 5.42; N, 10.05%. IR (KBr pellets, cm ):
2
218 [w, ν(O-H)]; 1605 [m, ν(C=N)]; 1375 [m, ν(C–N)], 1165 [s, ν(C-O)phenol], 1060 [s, ν(C-O)alcohols] and 655 [s, ν(O-H)alcohols] (w =
1
weak; m = medium; s = strong). H NMR (ppm, CDCl
and 9.18 (s, 1H, CHimine).
3
): δ 2.5 (s, 3H, CH
3
); 4.84 (s, 2H, CH ); 7.32-7.46 (m, 5H, Ar); 7.87 (s, 1H, Py)
2
1
2

4
-(2-fluorophenylimino)methyl)-5-(hydroxymethyl)-2-methylpyridin-3-ol (L2): Properties - yellow crystals, yield 71%. Melting
-1
point: 206 °C. Anal.Calc. for [C14
H
14FN
2
O ] C, 56.60; H, 4.75; N, 9.45. Found: C, 56.67; H, 4.76; N, 9.44%. IR (KBr pellets, cm ):
2
3
124 [w, ν(O-H)]; 1612 [m, ν(C=N)]; 1388 [m, ν(C–N)], 1296 [s, ν(C-O)phenol]; 1090 [m, ν(C-F)]; 1026 [s, ν(C-O)alcohols] and 630 [s,
1
ν(O-H)alcohols] (w = weak; m = medium; s = strong). H NMR (ppm, DMSO-d
6
): δ 2.45 (s, 3H, CH
3
); 4.82 (s, 2H, CH ); 7.04-7.68 (m,
2
4
H, Ar); 8.03 (s, 1H, Py) and 9.30 (s, 1H, CHimine).
4
-(2-chlorophenylimino)methyl)-5-(hydroxymethyl)-2-methylpyridin-3-ol (L3): Properties - orange crystals, yield 83%. Melting
-1
point: 202 °C. Anal.Calc. for [C14
H
13ClN
2
O ] C, 53.70; H, 4.49; N, 8.89. Found: C, 53.69; H, 4.51; N, 8.94%. IR (KBr pellets, cm ):
2
3
113 [w, ν(O-H)]; 1614 [m, ν(C=N)]; 1387 [m, ν(C–N)], 1263 [s, ν(C-O)phenol], 1026 [s, ν(C-O)alcohols], 855 [m, ν(C-Cl)] and 625 [s,
1
ν(O-H)alcohols] (w = weak; m = medium; s = strong). H NMR (ppm, DMSO-d
6
): δ 2.46 (s, 3H, CH
3
); 4.81 (s, 2H, CH ); 7.41-7.69 (m,
2
4
H, Ar); 8.03 (s, 1H, Py) and 9.25 (s, 1H, CHimine).
4
-(2-bromophenylimino)methyl)-5-(hydroxymethyl)-2-methylpyridin-3-ol (L4): Properties - orange crystals, yield 85%. Melting
-1
point: 207 °C. Anal.Calc. for [C14
H
13BrN
2
O ] C, 47.02; H, 3.95; N, 7.83. Found: C, 46.93; H, 3.90; N, 7.79%. IR (KBr pellets, cm ):
2
3
209 [w, ν(O-H)]; 1627 [m, ν(C=N)]; 1387 [m, ν(C–N)], 1257 [s, ν(C-O)phenol], 1033 [s, ν(C-O)alcohols], 694 [m, ν(C-Br)alcohols] and 620
1
[
s, ν(O-H)alcohols] (w = weak; m = medium; s = strong). H NMR (ppm, DMSO-d
6
): δ 2.46 (s, 3H, CH
3
); 4.80 (s, 2H, CH ); 7.32-7.81
2
(
m, 4H, Ar); 8.03 (s, 1H, Py) and 9.21 (s, 1H, CHimine).
1
3

4
-(2-iodophenylimino)methyl)-5-(hydroxymethyl)-2-methylpyridin-3-ol (L5): Properties - orange crystals, yield 81%. Melting
-
1
point: 214 °C. Anal.Calc. for [C14
H
13lN
2
O ] C, 41.56; H, 3.49; N, 6.92. Found: C, 41.50; H, 3.48; N, 6.99%. IR (KBr pellets, cm ):
2
3
070 [w, ν(O-H)]; 1604 [m, ν(C=N)]; 1381 [s, ν(C–N)], 1265 [s, ν(C-O)phenol], 1026 [s, ν(C-O)alcohols], 615 [s, ν(O-H)alcohols] and 525
1
[
m, ν(C-I)alcohols] (w = weak; m = medium; s = strong). H NMR (ppm, DMSO-d
6
): δ 2.46 (s, 3H, CH
3
); 4.79 (s, 2H, CH ); 7.14-8.00
2
(
m, 4H, Ar); 8.03 (s, 1H, Py) and 9.12 (s, 1H, CHimine).
2
.12. Copper(II) complexes synthesis – General procedure (Scheme 2)
The synthesis of the complex C1 was carried out in situ. Pyridoxal hydrochloride (0.203 g; 0.1 mmol), CuCl
2
.2H
2
O (0.085 g;
0
.05 mmol) and aniline (0.093 g; 0.1 mmol) were dissolved in methanol (10 mL). The solution was heated in an oil-bath at 50 ºC and
stirred for 1 h. After 2 days, green crystals were obtained by slow evaporation of the solvent.
In a round bottom flask, the corresponding ligand (L2, L3, L4 and L5; 0.1 mmol for each) was added to methanol (10 mL) and
the resulting mixture was subjected to magnetic stirring until complete dissolution of the ligand, followed by addition of triethylamine
(
30 µL). In the next step, copper(II) perchlorate (0.187 g; 0.05 mmol) dissolved in methanol (5 mL) was added and the resulting
solution was subjected to magnetic stirring at 60 °C for 1.5 h. After this period, the solution was cooled to room temperature and
slow evaporation of the solvent resulted in crystals of complexes C2-C5.
Insert Scheme 2 about here
1
4

-
1
Complex C1: Red/green crystals. Yield: 70% Melting point: 212 °C (dec.). Anal.Calc: C28
32
H N
4
O
6
Cl Cu (650,98 g.mol ): C, 51.38;
2
-
1
H, 4.99; N, 8.52. Found: C, 51.34, H, 4.92; N, 8.55%. IR (KBr pellets, cm ): 3526 [m, ν(O–H)alcohols]; 1599 [s, ν(C=N)]; 1415 [m,
ν(C–O)phenol.]; 1312 [m, ν(O–H)alcohol] and 1012 [s, ν(C–O) alcohols].
-
1
Complex C2: Red crystals. Yield: 56% Melting point: 253-254 °C (dec.). Anal.Calc: C28
H
24
F
2
N
4
O
4
Cu (582,06 g mol ): C, 57.80; H,
-1
4
.19; N, 9.62. Found: C, 57.78, H, 4.16; N, 9.63%. IR (KBr pellets, cm ): 3353 [m, ν(O–H)alcohols]; 1605 [s, ν(C=N)]; 1421 [m, ν(C–
N)]; 1187 [m, ν(C–O)phenol]; 1044 [m, ν(C-F)] and 1021 [s, ν(C–O) alcohols].
-
1
Complex C3: Red crystals. Yield: 90% Melting point: 208-209 °C (dec.). Anal.Calc: C28
H
2
24Cl N
4
O Cu (614,96 g mol ) C, 54.71; H,
4
-
1
3
.91; N, 9.12. Found: C, 54.69, H, 3.93; N, 9.11%. IR (KBr pellets, cm ): 3153 [m, ν(O–H)alcohols]; 1603 [s, ν(C=N)]; 1419 [m, ν(C–
N) ]; 1187 [m, ν(C–O)phenol]; 1026 [s, ν(C–O) alcohols] and 767 [m, ν(C-Cl)].
.
-
1
Complex C4: Red crystals. Yield: 96% Melting point: 255-256 °C (dec.). Anal.Calc: C28
H
24Br
2
N
4
O
4
Cu (703,86 g mol ) C, 47.80; H,
.46; N, 7.99. Found: C, 47.78, H, 3.44; N, 7.96%. IR (KBr pellets, cm ): 2916 [m, ν(O–H)alcohols]; 1606 [s, ν(C=N)]; 1420 [m, ν(C–
N) ]; 1187 [m, ν(C–O)phenol]; 1023 [s, ν(C–O) alcohols] and 735 [m, ν(C-F)].
-
1
3
.
1
5

-
1
Complex C5: Red crystals. Yield: 93% Melting point: 255-256 °C (dec.). Anal.Calc: C28
H
24
I
2
N
4
O
4
Cu (797,86 g mol ). C, 42.04; H,
.06; N, 6.97. Found: C, 42.15, H, 3.03; N, 7.02%. IR (KBr pellets, cm ): 3153 [m, ν(O–H)alcohols]; 1593 [s, ν(C=N)]; 1422 [m, ν(C–
N) ]; 1193 [m, ν(C–O)phenol] and 563 [m, ν(C–I)].
-
1
3
.
Insert Table 1 about here
3
. RESULTS AND DISCUSSIONS
3
.1. Crystal structures of C1, C2 and C5
In the molecular structure of complex C1 (Figure 1), the copper(II) cation is bonded to two phenolic oxygen atoms, as well as
to the two nitrogen atoms present in the imine function, thus giving a total of four bonds around the metallic center, with square
planar coordination geometry. This is shown by the angles formed between the O(1)-Cu-O(1)# and N(2)-Cu-N(2)# Å; O(1)-Cu-N(2)
Å atoms, which are respectively, 180(10)º and 89.53(7)º. These values are consistent with the literature [45]. The complex C1 is
similar to that described by Marinovich et al. [46]. Symmetry operations used to generate equivalent atoms: (#) -x + 2, y, z + 2.
The length of the bond involving the copper atom and the phenolic oxygen atoms of the ligand is 1.9134(15) Å, for Cu-O1. The
Cu-N2 imine bond presented a distance of 2.0062(18) Å. As expected, these values are within the range of Cu-O and Cu-N bond
1
6

lengths reported in the literature (Cu–O(phenol) 1.87-1.88 Å and Cu–N(imine) 1.99 Å) [47,48]. Other relevant bond lengths and angles
for the structural analysis of C1 are shown in Table S1 (see in the supplementary information section).
Insert Figure 1 about here
In the C2 complex, the coordination of the ligand (L2) is carried out through the iminic nitrogens, as well as from the phenolic
oxygens of the two pyridoxal molecules (Figure 2). These bond distances are in accordance with literature values [49,50,51]. Other
relevant bond lengths and angles for the structural analysis of C2 are shown in Table S1 (see in the supplementary information
section).
Apparently, the fluorine atom does not produce a change in the metal-halogen interaction thus, a second molecule of the
ligand is coordinated to the metal center generating the expected geometry.
Insert Figure 2 about here
Figure 1S (supplementary information section) allows the visualization of intermolecular O2#-H2#····N1 secondary interactions in
the bc plane, and angles and distances confirm that these intermolecular bonds support the growth of a supramolecular assembly
1
7

of complex C2 along the crystallographic b axis [4,52]. The comparison of hydrogen secondary bonds for compound C2 is shown in
Table S2 (supplementary information section).
In addition to these intermolecular interactions, it is also noteworthy that along the a axis there is an expressive number of
further trifurcated hydrogen bonding, as shown in Figure 2S (supplementary information section) [53-58]. These secondary bonds are
of an electrostatic nature, but sometimes their nature can undergo variations. According to Desiraju and Steiner [57], for
supramolecular organizations the directionality of the connection depends on the donor and the receiver.
In the case of complex C5 (Figure 3), the ligand is also in bidentate mode, with the two phenolate oxygens and two iminic
nitrogens in the equatorial positions. The bond distance values from the metal center to the imine nitrogen and the oxygen atoms of
phenolate are in accordance with the literature [59]. The most relevant values for the bond lengths and angles of complex C5 and a
comparison to C1 and C2 complexes are listed in Table S2 (see in the supplementary information section).
Insert Figure 3 about here
Although the secondary bonds of the iodine atom are below the sums of the Van der Waals rays (3.38 Å) [60], there is no
evidence of the existence of bonds/interactions between Cu···I (Figure 3S in the supplementary information section). However,
considering the secondary bonds of iodine and oxygen O2, the following packaging arrangement occurs: the bonding distance of
the I···O2 is 2.979(1) Å and the angle is 170.53(2)°, these values being compatible with strong iodine···oxygen interactions [61].
3
.2. UV-vis analysis of copper complexes C1-C5
1
8

In general, for the free ligands L1-L5 two bands were observed in the UV-vis spectra, which can be assigned to the π→π*
intraligand transition at 298 nm (imine moiety) and the n→π* transition in the 345-349 nm range.
The electronic spectra of the mononuclear copper(II)-complexes C1-C5, measured in DMF solutions, reveal two strong
absorption bands (Table 3S in the supplementary information section) (see Figure 4S in the supplementary information section) in the
2
90-500 nm range. While the band observed at 298 nm, can be assigned to the π→π* intraligand transition, the band in the 400-410
nm range is best described as being the result of a ligand-to-metal charge-transfer (LMCT) transition from pyridoxal to the Cu(II). In
any case, d-d transitions were not observed in the 500-800 nm electronic spectra range under these experimental conditions due to
the low solubility and absorption of these transitions. However, such transitions were clearly observed when solid state diffuse
reflectance spectra of the complexes C1-C5 were obtained (Table 3S in the supplementary information section).
3
.3. Electrochemical properties of copper(II) complexes C1-C5
Electrochemical analysis of the free ligands L1-L5 was carried out through cyclic voltammetry in dry DMF solution and cyclic
voltammograms (CV’s) revealed the presence of one irreversible process in the anodic region, with oxidation peaks at Epa = -0.08 V
1
9

(
R = H), Epa = +0.17 V (R = F), Epa = +0.10 V (R = Cl), Epa = +0.14 V (R = Br) and Epa = +0.11 V (R= I), respectively, which can be
attributed to ligand oxidation. In general, at the cathodic region (0.0 to -1.60 V), the ortho-substituted halogenated ligands presented
a reduction peak in the -1.40 to -1.60 V potential range, which we tentatively attribute to the reduction of the imine group.
In the case of the copper(II) complexes, the cyclic voltammograms of C1-C5 are presented in Figure 4. The CV´s of C1-C5 in
dry DMF solution display quasi-reversible redox responses in the -1.50 to -1.75 V potential range, which can be attributed to the
II
I
one-electron Cu /Cu redox pair. These redox potentials lie in the range found for mononuclear Cu(II) complexes containing SALEN-
type Schiff base ligands [62]. From the E1/2 values listed in Table 2, it seems that the ortho-substituted haloanilines only slightly
affect the N-donor capacity of the imine nitrogen coordinated to the copper(II) center. During the anodic potential scan, all the
complexes show two irreversible oxidative responses in the -0.28 to +0.11 V range, which are most probably related to the
sequential oxidation of the ligand molecules when coordinated to the Cu(II) center (Figure 4). The oxidation observed for the
ligands L1-L5 in the -0.08 – 0.17 V range and the presence of two oxidation processes for C1-C5 strongly corroborates this
II
assignment. In fact, the oxidation of the second ligand bound to Cu is expected to occur at a higher potential due to the loss of
electronic density of the whole molecule when the first ligand becomes oxidized. All the potentials were referenced to the
ferrocene/ferrocenium redox pair and potential values are listed in Table 2.
Insert Figure 4 about here
Insert Table 2 about here
3
.4. DNA cleavage assays
2
0

All complexes were able to cleave DNA in a concentration-dependent mode, i.e. as the complex concentration in the reaction
medium rises the DNA cleavage capacity also increases.
In this study, complex C2 (R = F) presented the lowest cleavage capacity. When the DNA sample was treated with 10 µM of
this Cu(II) complex, a single cleavage in the supercoiled DNA was observed generating a ~15% of FII (Figure 5B). The progressive
raise in complex C2 concentration raised the DNA cleavage reaching ~40% of FII with 250 µM of complex. The complexes with
chlorine (complex C3) and bromine halogen (complex C4) presented a cleavage potential similar to the unsubstituted complex C1
(
Figure 5A, 4C and 5D). The DNA treated with 250 µM of these complexes presents a ~60% of FII. In comparison to these
complexes, the one containing iodine (complex C5) demonstrated the higher activity, since at 250 µM concentration; ~80% of the
DNA was present as FII (Figure 5E). At the concentration range of 100-250 µM it could be seen the formation of linear DNA (FIII)
with ~10% and ~20% respectively, which indicates a double strand rupture of the DNA helix structure, being the last concentration
all treated plasmid DNA was cleaved with ~80% of FII and ~20% of FIII. Although the reactivity order may be better clarified by
kinetic tests, currently it can be suggested that: C2 << C4 ≈ C3 << C5.
These results may indicate that there is a relationship between electronegativity and atomic radius of each halogen and the
cleavage DNA potential, being the fluorine, having the highest electronegativity and smallest atomic radius, the one with lower
cleavage potential. The iodine, having the lowest electronegativity and larger atomic radius, presented higher cleavage potential.
Insert Figure 5 about here
2
1

3
.4.1. DNA groove binders: effect on cleavage
It was aimed to determine if the interaction between complex and DNA occurred due to dependency for minor or major DNA
grooves. Tests were realized using minor and major groove blockers, respectively netropsin and methyl green. The results are
presented in Figure 6.
Complex C1 (control) and the copper(II) complexes C3 and C4 have shown a clear inhibition in the presence of the minor
groove binder, netropsin, resulting in ~25% inhibition for complex C1, ~57% for the complex C4, and ~39% for the complex C3,
respectively. This suggests that such Cu(II) complexes access the DNA by a minor groove interaction (Figure 6). The ortho-I-
substituted complex C5 has not shown an alteration of DNA cleavage in the presence of both groove binders. The ortho-F-
substituted complex C2 has not presented a cleavage alteration in the presence of the minor groove binder netropsin, however, in
the presence of a major groove blocker, methyl green, the cleavage potential increased in ~40% (Figure 6). This may be explained
by the methyl green interacting with the DNA and facilitating the access of the complex to the structure, raising the cleavage
potential, effect that is already known and described for others groove binders such as distamicyn [62].
Insert Figure 6 about here
3
.4.2. ROS scavengers effect on cleavage
2
2

Since the description by Sigman and co-workers [63] of the metallic complexes capacity to cleave DNA generating reactive
oxygen species, studies were conducted to investigate the potential cleavage mechanism for each metal new complex described.
Reactive oxygen species (ROS) can be formed by the reduction of molecular oxygen, which can result in damage to the DNA by
oxidative cleavage [64]. Thus the DNA cleavage potential in the presence of two ROS scavenger species were tested, being
potassium iodine (KI) as scavenger of superoxide radical and sodium azide (NaN ) as singlet oxygen inhibitor.
3
Results presented in Figure 7 indicate that the cleavage mechanism of such copper(II) complexes should not be dependent of
this ROS, since no cleavage inhibition is seen in the presence of the scavengers, excepting in the case of the complex containing
bromine atom (complex C4). This complex presented a slight inhibition in the presence of the scavenger species, but still not
enough to confirm dependency [65]. The present results also indicate that the complexes cleave DNA through a hydrolytic
mechanism, although, it was reported that other complexes are able to oxidize directly the DNA molecule (the DNA cleavage
potential of the complexes was not affected in anaerobiosis (Figure 5S (supplementary information section), thus the reaction
seems to be independent of oxygen), without the generation of a reactive species [66-69].
Insert Figure 7 about here
3
.4.3. DNA cleavage kinetics
The kinetic profile of DNA cleavage promoted by the copper(II) complexes was determined by disappearance of the
supercoiled plasmid DNA form (FI) as a function of time (Figure 8). Considering the data as a pseudo-first-order reaction kinetic
2
3

[
70] where there is an excess of catalyzer relative to substrate, the cleavage DNA observed constant (kobs) and half-life time (t1/2
)
are presented in the inset of Figure 8.
Kinetic tests were performed at a single concentration, being 250 µM for all complexes, except for the complex C5 (R = I), for
which were used a lower concentration (50 µM), since it is the most active. The test indicates significant difference among the
observed constants from different derivatives, the complex with fluorine atom (complex C2) presenting a lower kobs and
consequently higher half-life time for the cleavage of the DNA form, corroborating with the results found at the concentration
dependence test (Figure 8). The complexes containing chlorine and bromine atoms (complexes C3 and C4) presents difference in
the kobs constants, diverging from the concentration tests, which showed a similar cleavage potential. At the kinetic test, the complex
C3 presents a higher kobs in comparison to Br-substituted complex C4 (Figure 8). The complex containing iodine atom in the ortho
position (complex C5), even in at 5 times lower concentration, presented a lower half-life time relative to complex C3 (R = Cl), which
presented the higher kobs at the 250 µM concentration (Figure 8A-E).
Thus, once more the complex C5 containing the iodine halogen atom is more active, and considering the data obtained by the
kinetic assay it is possible to express the following cleavage potential order: C2<< C4 << C3 <<C5.
Insert Figure 8 about here
3
.5. Thermal denaturation
2
4

The thermal denaturation assays were conducted with the complex that contains iodine atom (complex C5), the most active,
and also with a control complex without halogen (Cu-aniline – complex C1), and another control only with DNA and solvent
(
DMSO). The experiment was performed with the concentration of 5 µM complex.
The found DNA thermal denaturation temperature (T ) without complexes was 58,6ºC, while in the presence of complex C5
the T raised to 60.0ºC. The increase in the T was not sufficient to confirm that the complex increased the DNA denaturing
temperature as result of interaction between them (Figure S6 in the supplementary information section).
m
m
m
3
.6. Evaluation of superoxide dismutase mimetic activity (SOD)
We have evaluated the mimetic activity of superoxide dismutase of ligands and coordination compounds by analyzing
inhibition of the reaction between the superoxide ion and NBT and consequent formazan production. Thus, compounds that have
antioxidant activity are able to react more easily with superoxide radicals.
The reaction was monitored spectrophotometrically at 560 nm and the mean absorbance of triplicate values was subtracted
from the absorbance values obtained for the sample considered control. Thus, the result expressed as superoxide ion production
was inhibited at each concentration of the antioxidant. This value is then expressed as a percentage of inhibition of photoreduction
reaction of NBT. The straight line equation was used to quantify the IC50 value, the concentration which inhibits 50% of the reaction,
in order to relate the various concentrations of the compound tested with their respective percentages of inhibition of NBT
photoreduction.
2
5

The results demonstrate that some coordination compounds exhibited potential for inhibition of the superoxide ion while the
ligands did not demonstrate such a function.
When the halogen of the ortho-substituted was evaluated, it was found that the C2 complex containing the 2-fluoroaniline
ligand showed unsatisfactory disproportionation of superoxide ion. This fact can be explained by the high electronegativity of the
fluorine atoms that lead to a greater polarization of charges in the complex molecule, forming a negative pole that repels the
superoxide radical, making it difficult to approach to the metal center, which consequently promote the reaction between the
superoxide and NBT molecules. On the other hand, the complex containing iodine exhibits better IC50 values compared to the other
complexes, this can be explained (associated) to the lower value of the reduction potential of Cu (II)-Cu(I) species (Table 2), thus,
the dismutation of the superoxide ion promoted by copper occurs easily.
Values of the catalytic rate constants for superoxide disproportionation shown in table 3 clearly indicate that these complexes
can be used as superoxide scavengers. Comparison of kMcCF of copper complexes and bibliographic references with similar
complexes, with C5 being the most suitable for reacting with superoxide.
Insert Table 3 about here
3
.6. Catecholase activity assay
Several studies in the literature report on the catalytic activity of Cu (II) mononuclear complexes tested on the oxidation of the
substrate model 3,5-di-tert-butylcatechol (3,5-DTBC). This substrate is widely used an oxidative model due to its low redox potential
2
6

that facilitates oxidation in quinone (3,5-DTBQ), furthermore, the presence of bulky substituents promotes secondary reactions of
oxidation, such as ring opening [75 76,77].
The data obtained from kinetic studies of oxidation of 3,5-DTBC promoted by complexes of C1 to C5, were treated by
monitoring the appearance of the intense absorption band at 400 nm region, as a result of the formation of the corresponding
quinone (3,5-DTBQ; ε = 1.900 L/mol.cm). A kinetic study carried out for C1-C5, under anaerobic conditions, shows that only one
equivalent of the substrate is oxidized to 3,5-DTBQ (stoichiometry substrate:complex 1:1). Such results suggest that two
mononuclear Cu(II) centers of the catalyst should be reduced to Cu(I) in a simultaneous two-electron transfer process. The
interaction of the 3,5- DTBC substrate with two mononuclear C1-C5 units is assumed to generate the Michaelis intermediate. When
oxygen (O ) was bubbled into the solution immediately the catalytic activity is regenerated indicating that the catalyst is still active
2
and must participate directly in the catalytic cycle acting as a thermodynamic driving force by reoxidizing any Cu(I) species back to
the active Cu(II) complex [77,44,78].
In order to determine the influence of pH on the catalytic activity of C1-C5, reactions were carried out in the pH range 4.0 to
1
0.0. The dependence of the reaction rate (v ) on pH for all complexes is given in Figure 9(9A). Interestingly, for all the complexes,
0
bell shape profiles (but not well defined) were observed and a pH optimum of 6.4 was determined. In addition, a kinetic pK
a
of about
II
-
5
.5 can be extracted for C1 to C5. Drop in activity at pH ≥ 6.5 most probably occurs due to the presence of Cu coordinated OH
groups which reduces the association of the substrate 3,5-DTBC to the catalyst. It is reasonable to assume that in solution, under
-
II
the present kinetic experimental conditions a H
2
O/OH ligand occupies the fifth Cu coordination site [77].
Insert Figure 9 about here
2
7

After defining the ideal pH for the oxidation of 3,5-DTBC by complexes C1-C5, dependence studies of the reaction rate on the
substrate concentration were performed. For comparison, all curves were obtained at pH = 6.4. The graph of initial rates of the
reaction (V ) versus concentration of 3,5-DTBC show saturation profiles (Figure 9B), which suggests the formation of an
o
intermediate substrate/complex and thus can be treated by the Michaelis-Menten model developed for enzyme kinetics. Kinetic
parameters were obtained from nonlinear fits of the curves and the results are presented in Table 4.
Insert Table 4 about here
In general, the kinetic parameters found for complexes C1 to C5 are comparable to those previously reported for mono- di-
and poly-nuclear copper(II) complexes [17,18,19,44]. From the data shown in Table 4 it’s possible to observe that the catalytic
activity (kcat) of these complexes is of the same magnitude but with complex C5 being slightly more active. In fact, these
experimental results are in full agreement with the cyclic voltammetric data, which clearly show that E1/2 of C5 is anodic shifted by
~
150 mV when compared to C1-C4, due to the iodine substituent at the aniline ring. On the other hand, complex C1 shows the
lower K value within the series of complexes, a result which is most probably related to some steric effect of the halogen
M
substituents in C4-C5. At this stage it is important to emphasize that under the kinetic experimental conditions, it is assumed that a
dimer is most probably the catalytically active species in the oxidation of 3,5-DTBC. Indeed, several factors need to be considering
in assessing the difference in catalytic activity of these complexes. For the complexes shown in the present work, it seems that
electrochemical properties and steric match play the major role in reactivity of complexes C1-C5.
2
8

Finally, the peroxide detection was performed according to the iodometric method proposed by Meyer and co-workers [79],
which detected the presence of H accumulation compared with the experiment without the presence of the catalyst. This implies
that the mechanism of catalysis performed by complexes C1 to C5, is different from that proposed for the native enzyme [80] and is
likely a 1:1 stoichiometric process 3,5-DTBC + O → 3,5-DTBQ + H without the formation of water in the catalytic cycle.
2
O
2
2
O
2 2
6
. FINAL REMARKS
Schiff base-type ligands derived from pyridoxal and halogen-substituted anilines showed great potential to coordinate
copper(II) ions. The X-ray diffraction analysis revealed that the complexes presented square planar geometry with N donor
O
2 2
atoms. The structures of C2 and C5 complexes present crystalline networks with hydrogen bonding and substantially distinct
interactions, although they have the same spatial group and coordination environment.
All copper complexes showed antioxidant activity as mimetic for SOD, however, the best result was found with the complex
containing ortho iodine which is the strongest oxidant within the series of complexes. On the other hand, the complex containing
fluorine in the ortho position presented inexpressible IC50% showing a clear trend related to electronic conditions generated by the
substituents DNA and cathecolase.
Finally, cathecolase activity of the complexes in the oxidation of 3,5-di-tert-butylcatechol reveals that complex C5, containing
the iodine substituent at the aniline ring and the highest E1/2 value, also shows the highest catalytic activity. A similar trend was
found for DNA cleavage, leading us to conclude that these complexes are able to directly oxidize the DNA molecule without the
generation of ROS.
2
9