Donepezil-based hybrids as multifunctional anti-Alzheimer's disease chelating agents: Effect of positional isomerization

Article abstract of DOI:10.1016/j.jinorgbio.2020.111039

The intricate and multifactorial nature of Alzheimer's disease (AD) requires the development of compounds able to hit different pathophysiological targets, such as cholinergic dysfunction, deposits of amyloid beta (Aβ) peptide and metal dyshomeostasis. In order to continue the search for new anti-AD drugs, a design strategy was once more followed based on repositioning donepezil (DNP) drug, by ortho-attaching a benzylpiperidine mimetic of DNP moiety to a hydroxyphenyl-benzimidazole (BIM) chelating unit (compound 1). Herein, compound 1 and a positional isomer 2 are compared in terms of their potential multiple properties: both present good acetylcholinesterase (AChE) inhibition (low μmolar range) and are moderate/good inhibitors of Aβ self- and Cu-mediated aggregation, the inhibition process being mainly due to ligand intercalation between the β-sheets of the fibrils; compound 1 has a higher chelating capacity towards Cu²⁺ and Zn²⁺ (pCu = 14.3, pZn = 6.4, pH 7.4, C_L/C_M = 10, C_M = 10^?6 M) than 2 (pCu = 10.7, pZn = 6.3), attributed to its ability to establish a tridentate (N,O,O) coordination to the metal ion. Both compounds are eligible as drug candidates for oral administration but compound 1 shows improved neuroprotective role by completely preventing Aβ-induced cell toxicity.

Full text of DOI:10.1016/j.jinorgbio.2020.111039

Journal Pre-proof
Donepezil-based hybrids as multifunctional anti-Alzheimer's
disease chelating agents: Effect of positional isomerization
Marina Costa, Romane Josselin, Diana F. Silva, Sandra M.
Cardoso, Nóra V. May, Sílvia Chaves, M. Amélia Santos
PII:
S0162-0134(20)30067-2
DOI:
https://doi.org/10.1016/j.jinorgbio.2020.111039
JIB 111039
Reference:
To appear in:
Journal of Inorganic Biochemistry
Received date:
Revised date:
Accepted date:
17 October 2019
12 February 2020
13 February 2020
Please cite this article as: M. Costa, R. Josselin, D.F. Silva, et al., Donepezil-based
hybrids as multifunctional anti-Alzheimer's disease chelating agents: Effect of positional
isomerization, Journal of Inorganic Biochemistry (2020), https://doi.org/10.1016/
j.jinorgbio.2020.111039
This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. 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.
©
2020 Published by Elsevier.

Journal Pre-proof
1
Donepezil-based hybrids as multifunctional anti-Alzheimer´s disease
chelating agents: effect of positional isomerization
a)
a)
b)
b),c)
Marina Costa, Romane Josselin, Diana F. Silva, Sandra M. Cardoso,
Nóra
d)
a)*
a)*
V. May, Sílvia Chaves, M. Amélia Santos
a)
Centro de Quimica Estrutural, Instituto Superior Tecnico, Universidade de Lisboa, Av,
Rovisco Pais 1, 1049-001, Lisboa, Portugal
CNC – Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra,
b)
Portugal
c)
Institute of Molecular and Cell Biology, Faculty of Medicine, University of Coimbra,
Coimbra, Portugal
d)
Research Centre for Natural Sciences, 1117 Magyar tudósok körútja 2 Budapest,
Hungary
Keywords: Hydroxyphenyl-benzimidazole; Donepezil; Anti-neurodegeneratives;
Alzheimer´s disease; Multifunctional drugs; Metal chelation.
*
Corresponding authors:
(
Sílvia Chaves) E-mail: silvia.chaves@tecnico.ulisboa.pt; Phone: +351-218419269
M. Amélia Santos) E-mail: masantos@tecnico.ulisboa.pt; Phone: +351-218419273
(

Journal Pre-proof
2
Abstract
The intricate and multifactorial nature of Alzheimer´s disease (AD) requires the
development of compounds able to hit different pathophysiological targets, such as
cholinergic dysfunction, deposits of amyloid beta (Aβ) peptide and metal
dyshomeostasis. In order to continue the search for new anti-AD drugs, a design
strategy was once more followed based on repositioning donepezil (DNP) drug, by
ortho-attaching a benzylpiperidine mimetic of DNP moiety to a hydroxyphenyl-
benzimidazole (BIM) chelating unit (compound 1). Herein, compound 1 and a
positional isomer 2 are compared in terms of their potential multiple properties: both
present good acetylcholinesterase (AChE) inhibition (low μmolar range) and are
moderate/good inhibitors of Aβ self- and Cu-mediated aggregation, the inhibition
process being mainly due to ligand intercalation between the β-sheets of the fibrils;
2
+
2+
compound 1 has a higher chelating capacity towards Cu and Zn (pCu = 14.3, pZn =
−
6
6
.4, pH 7.4, C /C = 10, C = 10 M) than 2 (pCu = 10.7, pZn = 6.3), attributed to its
L M M
ability to establish a tridentate (N,O,O) coordination to the metal ion. Both compounds
are eligible as drug candidates for oral administration but compound 1 shows improved
neuroprotective role by completely preventing Aβ-induced cell toxicity.

Journal Pre-proof
3
1
. Introduction
Alzheimer´s disease (AD) is one of the most common and devastating age-dependent
neurodegenerative disorders and its prevalence is expected to enormously increase in
the coming years because of the population growth and ageing [1]. Hence, there has
been a worldwide intensive research aimed to understand and find treatment for AD, but
no cure has been found so far [2,3]. Currently approved AD therapeutics are only able to
attenuate the dementia symptoms by compensating the cholinergic deficit, as the
acetylcholinesterase (AChE) inhibitors (tacrine, donepezil, rivastigmine, galantamine)
or the glutamate N-methyl D-aspartate (NMDA) receptor antagonist (memantine) [4].
The recognized multifactorial nature and consequent complexity of AD are believed to
account for the absence of disease-modifying drugs. In fact, the formation of misfolded
protein aggregates, namely β-amyloid (Aβ) plaques (a prominent hallmark of AD),
dyshomeostasis of metal ions and oxidative stress are among the many factors that are
implicated in the trigger and progression of AD [5]. Furthermore, although Aβ
imbalance has been considered to be a causative agent in AD, high levels of metals (e.g.
Fe, Cu, Zn) have been found in the Aβ plaques of AD brains and, in particular Cu and
Zn bind to Aβ peptides, promoting their aggregation, while redox active metal ions (e.g.
Cu, Fe) promote overproduction of reactive oxygen species (ROS) that can damage
biomolecules [6-8]. Therefore, to address the multiple pathological factors and their
potential interconnections in AD pathogenesis, in the last decade there has been a
change of paradigm from single-target drugs towards multi-target drugs as chemical
entities (the so called multi-target ligands) able to address multiple biological targets
involved in the disease, aimed to get possible therapeutic answers [9-11].
In recent years, several research groups, including ours, have developed hybrid
compounds, via a rational structure-based methodology which integrates a moiety with
capacity for AChE inhibition with a framework able to interact with Aβ and also possess
metal cheating capacity, so that to associate the symptom treatment with disease-
modifying roles, namely to inhibit the Aβ self- and metal-induced aggregation and to
attenuate the metal-promoted ROS formation [12-16]. Herein we report the synthesis,
physico-chemical and biological evaluation as well as molecular modeling of a new
donepezil-hydroxyphenylbenzimidazole (DNP-BIM) chelating hybrid (1) (Fig 1). This
compound results from a structural modification of a recently reported positional isomer

Journal Pre-proof
4
DNP hybrid (2) [17], for the sake of increasing ligand denticity and thus improving the
metal chelating capacity as well as the control/modulation of interconnected multiple
pathological features of AD, namely those that could arise from the synergy of its metal
chelation and Aβ interaction properties.
O
N
O
O
Donepezil
1
2
Fig. 1. DNP and DNP-BIM hybrids (1 and 2), chelators with different attachment points
between the benzylpiperidine and the BIM moieties.
2
. Experimental
2
2
.1. Chemistry
.1.1. General methods and materials
Analytical grade reagents were purchased and used as supplied. Solvents were dried
according to standard methods [18]. The chemical reactions were monitored by TLC
using alumina plates coated with silica gel 60 F254 (Merck). Column chromatography
separations were performed on silica gel Merck 230-400 mesh (Geduran Si 60). The
melting points (m.p.) were measured with a Leica Galen III hot stage apparatus and are
1
13
uncorrected. The H- and C-NMR spectra were recorded on Bruker AVANCE III
spectrometers at 300 and 400 MHz, respectively. Chemical shifts (δ) are reported in
ppm from the standard internal reference tetramethylsilane (TMS). The following
abbreviations are used: s = singlet, d = doublet, t = triplet, m = multiplet. Mass spectra

Journal Pre-proof
5
(ESI-MS) were performed on a 500 MS LC Ion Trap (Varian Inc., Palo Alto, CA, USA)
mass spectrometer equipped with an ESI ion source, operated in the positive or negative
ion mode. High resolution mass spectrum (HRMS) was obtained on a Bruker Impact II
quadrupole mass spectrometer (Bruker Daltoniks). The electronic spectra were recorded
with a Perkin Elmer Lambda 35 spectrophotometer, using thermostated 1-cm path
length cells and the fluorescence measurements were performed with a Spectramax
Gemini EM (Molecular Devices) fluorimeter.
2
.1.2. Synthesis of compound 1
2
.1.2.1. 2-((1-Benzylpiperidin-4-yl)methyl)isoindoline-1,3-dione (1a). Phthalic acid
anhydride (1 eq) and 4-(aminomethyl)piperidine (1 eq) were heated at 160 °C for 4 h.
The resulting dark brown solid was mixed with K CO (6.6 eq), triethylamine (1.4 eq)
2
3
and benzyl chloride (2.3 eq) and refluxed in acetonitrile for 3 h. Then, the mixture was
cooled to r.t., added to water and then extracted with ethyl acetate (three times). The
organic layers were collected, dried over sodium sulphate anhydrous and concentrated
under reduced pressure. The crude was purified through chromatography column
(
eluent: ACN/H O/NH OH 92:8:0.1), affording a white solid ( = 29%); m.p. = 130 °C.
2
4
1
H NMR (300 MHz, MeOD-d ), δ (ppm): 1.37 (m, 2H, Pip), 1.67 (d, 2H, J = 12.0 Hz,
4
Pip), 1.80 (m, 1H, Pip), 2.02 (td, 2H, J = 12.0 Hz, Pip), 2.91 (d, 2H, J = 12.0 Hz, Pip),
3
7
.51 (s, 2H, NCH Ph), 3.58 (d, 2H, J = 12.0 Hz, NCH CH), 7.34-7.29 (m, 5H, Ph),
2 2
+
.89-7.79 (m, 4H, Phth); m/z (ESI-MS): 335.00 (M+H) .
2
.1.2.2.
(1-Benzylpiperidin-4-yl)methanamine
(1b).
2-((1-Benzylpiperidin-4-
yl)methyl)isoindoline-1,3-dione (1 eq) and hydrazine monohydrate (6 eq) were
dissolved in absolute ethanol and warmed at 60 °C for 3 h. The reaction mixture was
then filtered and concentrated under reduced pressure. The residue was dissolved in
dichloromethane, washed with brine (three times), dried over anhydrous sodium
sulphate and concentrated under reduced pressure, obtaining a pure yellow-orange oil (
1
=
30.5%). H NMR (300 MHz, MeOD-d ), δ (ppm): 1.01-1.21 (m, 2H, Pip), 1.64 (d,
4
2
2
H, J = 12.0 Hz, Pip), 1.86 (t, 1H, J = 12.0 Hz, Pip), 2.40 (d, 2H, J = 12.0 Hz, Pip),
.78 (d, 2H, J = 12.0 Hz, NCH CH) 2.91 (d, 2H, J = 12.0 Hz, Pip), 3.51 (s, 2H,
2
NCH Ph), 7.34-7.19 (m, 5H, Ph).
2

Journal Pre-proof
6
2
.1.2.3. 2-(2-Hydroxyphenyl)-1H-benzo[d]imidazole-4-carboxylic acid (1c). To a
solution of salicylaldehyde (1 eq) and Na S O (1.2 eq) in the green solvent
2
2
5
(glycerol/choline chloride (2:1)) was added the 2,3-diaminobenzoic acid (1 eq). The
mixture was heated at 50 °C for 30 min. Then, the mixture was cooled at r.t., added to
water and filtrated. The afforded solid was subsequently washed with water and
dichloromethane and then it was purified by recrystallisation from EtOAc and MeOH.
1
The title compound was obtained as a purple solid ( = 25.2%), m.p. > 300°C. H NMR
(
300 MHz, MeOD-d ), δ (ppm): 7.08 (d, 2H, J = 8.0 Hz, BIM); 7.42 (m, 2H, J = 8.0 Hz,
4
+
BIM), 7.96 (t, 2H, J = 8.0 Hz, BIM), 8.25 (d, 1H, BIM); m/z (ESI-MS): 253 (M-1) ,
+
2
55.00 (M+1) .
2
.1.2.4. N-((1-benzylpiperidin-4-yl)methyl)-2-(2-hydroxyphenyl)-1H-benzo[d]imidazole-
-carboxamide (1). To a solution of the 2-(2-hydroxyphenyl)-1H-benzo[d]imidazole-4-
4
carboxylic acid (1 eq, 1c) and (1-benzylpiperidin-4-yl)methanamine (1 eq, 2b), in dry
DMF, was added N-hydroxysuccinimide (1 eq) and DCC (1 eq), and the reaction
mixture was left stirring for 24 h at r.t.. Then the precipitate obtained (urea) was filtered
off and the liquid phase was diluted with ethyl acetate and washed with brine (3 times),
dried over anhydrous sodium sulfate, concentrated under reduced pressure and purified
by recrystallization in EtOAc, affording a yellow solid ( = 24.2%); m.p. = 192−193
1
°
C. H RMN (300 MHz, MeOD-d ), δ (ppm): 1.75-1.95 (m, 5H, Pip), 2.17 (t, 2H, J =
4
1
2 Hz, Pip), 3.02 (d, 2H, J = 12 Hz, Pip), 3.49 (d, 2H, J = 6 Hz, CH NCO), 3.60 (s, 2H,
2
CH Ph), 7.00-7.08 (m, 2H, Ph), 7.28-7.42 (m, 7H, Ph), 7.79 (d, 1H, J = 9 Hz, BIM),
2
1
3
7
.89 (d, 1H, J = 8 Hz, BIM), 8.16 (d, 1H, J = 9 Hz, BIM). C RMN (100 MHz, MeOD-
d4), δ (ppm): 29.25, 33.35, 44.31, 52.77, 62.44, 116.65, 119.45, 119.54, 122.13, 127.18,
1
1
4
27.59, 127.84, 129.50, 129.59, 131.84, 131.91, 135.91, 136.56, 139.11, 139.33,
+
51.40, 157.44, 161.38; HRMS (ESI), m/z: calculated for C H N O [M+H] ,
2
7
29
4
2
41.2293; found 441.2298.
2
.2.
Molecular modeling for ligand-AChE docking
To perform the ligand-enzyme docking study, the X-ray crystallographic structure of
Torpedo californica-AChE (TcAChE) complexed with an inhibitor (donepezil, DNP)
was taken from RCSB Protein Data Bank (PDB entry 1 EVE) [19], in order to be used
as enzyme model. This structure was chosen because of the similarity between the

Journal Pre-proof
7
original inhibitor (DNP) and ligand 1, namely relatively to the benzylpiperidine moiety.
The original structure was treated using Maestro v. 9.3 [20], by removing the original
ligand, solvent, and co-crystallization molecules, and then adding the hydrogen atoms.
The ligand structure was built using Maestro, and then, using Ghemical v. 2.0 [21], it
was submitted to random conformational search (RCS) of 1000 cycles, and 2500
optimization steps using Tripos 5.2 force field [22]. The minimized ligand was docked
into the AChE structure with GOLD software v. 5.1 [23], and the zone of interest was
defined as the residues within 10 Å from the original position of the ligand in the crystal
structure. The ‘allow early termination’ option was deactivated and the remaining
default parameters of Gold were used. The ligand was subjected to 100 genetic
algorithm steps using ASP as fitness function. Figure S1 encloses the result of re-
docking the original ligand (DNP) inside the TcAChE active site, under the simulation
conditions used in the present study, evidencing an excellent superposition and so
giving support to the used simulation methodology.
2
.3. Prediction of pharmacokinetic properties
To analyze the drug like properties of compound 1, and of DNP (for comparison
reasons), the prediction of the main pharmacokinetic parameters, was performed in
silico. This procedure determined parameters such as the lipo-hydrophilic character
(clog P), blood–brain barrier partition coefficient (log BB), the ability to be absorbed
through the intestinal tract (Caco-2 cell permeability) and the CNS activity.
The ligands were built and minimized as previously mentioned for the docking studies.
The structures were submitted to the calculation of these relevant pharmacokinetic
properties and descriptors using QikProp v. 2.5 [24]. These predictions are for orally
delivered drugs and assume non-active transport.
2
.4. Metal chelation studies
2
.4.1. Potentiometric and UV-vis spectrophotometric titrations
Solutions. The aqueous copper (CuCl , 0.015 M) and zinc (ZnCl , 0.0156 M) stock
2
2
solutions for potentiometric and spectrophotometric equilibrium studies were prepared
from 1000 ppm standards (Titrisol) and their metal content was evaluated by atomic
absorption. The titrant solution (0.1 M KOH) was obtained from a carbonate-free
commercial concentrate (Titrisol) in 50% (w/w) DMSO/H O medium and standardized
2
by potentiometric titration with potassium hydrogen phthalate. The titrant solution was

Journal Pre-proof
8
discarded whenever the percentage of carbonate (Gran’s method) [25] was about 0.5%
of the total amount of base.
Measurements. Compound 1 was titrated in 50% (w/w) DMSO/H O medium, at T =
2
2
5.0  0.1 °C and ionic strength (I) 0.1 M KCl, by using 0.1 M KOH as titrant. Both
glass and Ag/AgCl reference electrodes were previously conditioned in DMSO/H O
2
mixtures with increasing amount of DMSO and the response of the glass electrode was
controlled by strong acid – strong base (HCl/KOH) calibrations with the determination
of the Nernst parameters by Gran’s method [25]. The measurements were performed in
-4
a final volume of 30.00 mL, with the ligand concentrations (C ) 6.7  10
M
L
-5
(
potentiometry) and 4.0  10 M (spectrophotometry), under different C /C ratios: 0:1
M
L
(
L), 1:1 and 1:2 (M = Cu, Zn). The spectrophotometric measurements were carried out
at pH ca 2.5-11.5, in the 300–420 nm wavelength range. All titrations were performed in
duplicate and, under the stated experimental conditions, the pK value (13.6) was
w
determined and subsequently used in the computations. The stepwise protonation
constant of the ligand, K = H L/H LH, and the overall metal–complex stability
i
i
i-1
m
h
l
constants, M
= M H L /M H L , were calculated by fitting the
H L
l
m
h l
m
h
potentiometric and spectrophotometric data with, respectively, Hyperquad 2008 [26]
and PSEQUAD programs [27]. The metal hydrolysis constants were determined under
the defined experimental conditions (I = 0.1 M KCl, 50% (w/w) DMSO/H O, T = 25.0
2
°

0.1 C) and the following values of stability constants were included in the fitting of
2
+
2+
experimental data towards the equilibrium models related to the Cu /L and Zn /L
systems: log = -9.99; log ZnH-2= -14.79, log ZnH-3 = -23.48. Species distribution

Cu H
2
2
curves were plotted with the Hyss program [26], and this program was also used for the
−
6
determination of pM values (pM = −log[M] at pH 7.4, C /C = 10, C = 10 M) [28].
L
M
M
Conditional dissociation constants (K´ ) for the copper complexes at pH 7.4 were
d
+
+ 2
determined as 1/K´CuL and K´CuL = KCuL/ ( = 1+ [H ]/K + [H ] /(K K ) +
L
L
1
1 2
+
3
[
H ] /(K K K )).
1 2 3
2
.4.2. ESI-MS spectra
-4
-4
The pH of the Cu(II)/1 (C = 2.25  10 M) and Zn(II)/1 (C = 2.25  10 M) systems
L
L
in 50% (w/w) DMSO/H O medium was previously ascertained to 4.28 (C /C = 1) and
2
L
Cu
6
.42 (C /C = 2), as well as to 7.01 and 8.39 (C /C = 2), by using KOH solution, and
L Cu L Zn
the respective mass spectra were obtained from a LCQ Fleet mass spectrometer

Journal Pre-proof
9
operated in the ESI positive and negative ion mode (Thermo Scientific). The optimized
parameters were as following: ion spray voltage, +4.5/-2.5 kV; capillary voltage, 16/-28
V; tube lens offset, -63 V; sheath gas (N ), 80 arbitrary units; auxiliary gas, 5 arbitrary
2
units; capillary temperature, 250 ᵒC. The spectra were recorded in the range 100–1000
Da. Spectra typically correspond to the average of 20–35 scans.
2
.4.3. EPR spectra
CW-EPR spectra were recorded with a BRUKER EleXsys E500 spectrometer
microwave frequency 9.85 GHz, microwave power 10 mW, modulation amplitude 5 G,
modulation frequency 100 kHz). Two samples were prepared for each compound (1 and
), one at ligand excess and another at around equimolar concentration in 50% (w/w)
(
2
2
+
-3
DMSO/H O solution. The pH of the Cu /1 (C = 1.48  10 M) system was
2
L
2
+
ascertained to 6.75 (C /C = 1) and 6.91 (C /C = 1.5), while that of the Cu /2 (C =
L
Cu
L
Cu
L
-3
2
.23  10 M) system was to 5.46 (C /C = 1) and 5.72 (C /C = 2), by using KOH
L
Cu
L
Cu
solution. Room temperature spectra were measured in capillary at 298 K. Anisotropic
EPR spectra were recorded in quartz EPR tubes immersed in a Dewar containing liquid
nitrogen (77 K); 0.025 mL MeOH was added to the 0.2 mL sample to avoid water
crystallization upon freezing. All recorded EPR spectra, and anisotropic EPR spectra at
7
7 K, were simulated by the EPR program [29]. To describe the isotropic spectra the
Cu
N
parameters g , A copper hyperfine (I = 3/2) and a nitrogen (I = 1) superhyperfine
0
0
Cu
0
N
2
+
couplings have been taking into account. The four Cu linewidths were treated with
Cu
individual linewidth parameters. The g-tensor (g g , g ,), copper hyperfine tensor (A ,
x,
y
z
x
Cu
Cu
z
N
N
N
z
A , A ) and nitrogen hyperfine tensor (a ,a ,a , for wich x y and z denotes the
y
x
y
directions of the g-tensor) parameters were fitted. For the description of the linewidth
the orientation dependent  and  parameters were used to set up each component
spectra. Since a natural copper(II) chloride was used for the measurements, both the
6
3
isotropic and anisotropic spectra were calculated as the sum of the spectra of Cu and
6
5
Cu weighted by their natural abundances. The hyperfine and super-hyperfine coupling
-4
constants and the relaxation parameters were obtained in field units (Gauss = 10 T).
2
.4.4. Molecular modeling of the copper complexes
In the absence of X-ray structures, a molecular modelling study was performed aimed at
providing some further insight into the 1:1 and 1:2 (M/L) copper complex structures of
compounds 1 and 2. In the case of the 1:1 complexes, one and two water molecules

Journal Pre-proof
10
were considered in the coordination shell, respectively, for complexes with compounds
and 2. These studies were performed with full geometry optimization of the copper
1
complexes by quantum mechanical calculations based on DFT methods [30], included
in the Gaussian 03 program software [31] with the B3PW91 hybrid functional [32].
Preliminary molecular simulation was obtained by using 3-21G basis set [33] and
afterwards 6-31G** [34]. No symmetry constrains were enforced during geometry
optimization with the two basis sets. Identical calculations were also performed for one
water molecule. Concerning the 1:1 copper complexes, the electronic energies (Eb1)
obtained at the PBE0/b1 level of theory were converted to free energy (G ) at 298.15 K
b1
and 1 atm by using zero point energy and thermal energy corrections based on structural
and vibration frequency data calculated at the same level.
2
2
.5. Biological studies
.5.1. Acetylcholinesterase inhibition
The enzymatic activity of AChE was determined using an adaptation of Ellman’s
method, as already described [35,36]. Firstly, a stock solution of AChE was prepared by
dissolving enzyme 500 U in 10 mL of tris(hydroxymethyl)aminomethane (TRIS) buffer
(50 mM, pH 8). Then, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)
buffer was used for further dilution of the enzyme solution, therefore obtaining the final
working solution of AChE. The assay solution consisted of 374 μL of HEPES buffer (50
mM and pH 8.0), a variable volume (10–50 μL) of compound 1 stock solution (1
mg/mL of MeOH), 476 μL of 3 mM bis(3-carboxy-4-nitrophenyl) disulfide (DTNB), 25
μL of AChE stock solution and the necessary amount of methanol to attain the same
volume of sample mixture in a 1 mL cuvette. After mixing, samples were left to
incubate for 15 min and then 75 μL of 16 mM acetylthiocholine iodide (AChI) solution
was added and immediately the reaction was monitored at 405 nm for 5 min. Firstly,
enzyme activity was determined by measuring the rate of enzymatic reaction in solution
by recording absorbance (with or without enzyme) at 405 nm. To calculate the percent
inhibition of AChE, the enzyme activity in the presence of increasing concentrations of
compound 1, as well as for a blank reaction (methanol without compound), were
measured at 405 nm.
The obtained results were expressed in percentage of inhibition through the Eq. (1):
%
I = 100−(v / v x 100)
(1)
1
0

Journal Pre-proof
11
Where v is the initial reaction rate in the presence of the inhibitor and v is the initial
1
0
rate of the control reaction. The curves of inhibition were obtained using a plot of
percentage of enzymatic inhibition versus inhibitor concentration, and a calibration
curve was obtained from which the linear regression parameters were obtained.
2
+
2
.5.2. Inhibition of self- and Cu -mediated Aβ aggregation
Amongst the methods used to measure the aggregation of Aβ1-42, the assay based on
Thioflavin T (ThT) is the standard procedure [37,38]. This test is performed with Aβ1-42
peptide solution, previously prepared by dissolving it in 1,1,1,3,3,3–hexafluoro-2-
propanol (HFIP), an organic solvent useful to solubilize and monomerize the β–sheet
protein aggregates and reserve them in the fridge. Films were re-dissolved in an
Eppendorf tube with a fresh solution of a mixture of CH CN/Na CO /NaOH, to have a
3
2
3
stable stock solution, and, afterwards, added to an optimized fibrillation phosphate
buffer. The samples were prepared using MeOH (1 mg/mL) as a solvent with the
specific procedure [16], followed by incubation in a water bath for 24 h at 37 °C with
gentle shaking. After incubation, the samples were added to a 96-well plate (BD Falcon)
with 180 µL of 5 µM ThT in 50 mM glycine-NaOH (pH 8.5) buffer. Blank samples
were prepared for each concentration in a similar way, devoid of peptide. After 5-min
incubation with the dye, the ThT fluorescence was measured at 446 nm (excitation) and
4
85 nm (emission).
The obtained results were expressed in percentage of inhibition of Aβ1-42 aggregation
through the Eq. (2): (2)
%
I = 100−(F / F × 100).
1
0
F1 is the difference between the fluorescence of a solution containing Aβ (or Aβ and
copper) plus the inhibitor and the respective blank solution without Aβ; F is the
0
difference between the fluorescence of a solution containing Aβ (or Aβ and copper) and
the respective blank solution whithout Aβ. The reported values were obtained as the
mean±SEM of three different experiments.
2
.5.3. Cell viability and neuroprotection
SH-SY5Y human neuroblastoma cell line (ATCC-CRL-2266) is grown in Dulbecco’s
modified Eagle’s medium (DMEM) and was obtained from Gibco-Invitrogen (Life
Technologies Ltd, UK) with 10% heat inactivated fetal calf serum, containing 50 U/mL

Journal Pre-proof
12
penicillin and 50 μg/mL streptomycin, under a humidified atmosphere of 5% CO at 37
2
6
ºC. Cells were plated at 0.12 × 10 cells/mL for cell viability assay. The tested
compounds were dissolved in DMSO at a concentration of 50 mM and aliquots were
stored at -20 °C. A dose-response screening (from 1 μM up to 10 μM) was performed in
order to choose the highest non-toxic concentration. As a result the 5 μM final
concentrations was selected for compound 1. The final concentration of DMSO in
culture media did not exceed 0.05% (v/v) and no alterations on cells were observed. The
compound was added to the cell media 1 h before the incubation with Aβ1-42 or L-
ascorbic acid/ferrous sulfate. Aβ1-42 or L-ascorbic acid/ferrous sulfate where incubated
alone or with the compound for an additional 24 h. Aβ1-42 was prepared as 276.9 μM
stock in sterile water and added to the medium at 2.5 μM final concentration. Ferrous
sulfate was freshly prepared as 0.36 M stock in water and added to the medium at 500
μM final concentration. L-Ascorbic Acid was freshly prepared as 80 mM stock in water
and added to the medium at 5 mM final concentration. Aβ1-42 was purchased from
Bachem (Torrance, CA, USA) and ferrous sulfate and L-ascorbic acid from Sigma
Chemical Co (St. Louis, MO, USA).
Cell viability was determined by the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) reduction test. In viable cells, the enzyme succinate
dehydrogenase metabolizes MTT into a formazan that absorbs light at 570 nm.
Following the cell treatment protocol, the medium was aspirated and 0.5 mL MTT (0.5
mg/mL) was added to each well. The plate was then incubated at 37 °C for 1 h 30 min,
protected from light. At the end of the incubation period, the formazan precipitates were
solubilized with 0.5 ml of acidic isopropanol (0.04 M HCl/isopropanol). The absorbance
was measured at 570 nm [39]. Cell reduction ability was expressed as a percentage of
untreated control cells.
All data were expressed as mean ±S.E.M. of at least three independent experiments
performed in triplicates. Statistical analyses were performed using GraphPad Prism 6
(GraphPad Software, San Diego, CA, USA). Differences between two datasets were
evaluated by two-tailed unpaired Student’s t test and a P value < 0.05 was considered
statistically significant.
3
. Results and Discussion
3
.1. Synthesis of compound 1

Journal Pre-proof
13
The synthesis of compound 1 involved the set of chemical reactions represented in
Scheme 1.
Scheme 1. Reagents and conditions: i) phthalic anhydride, 160 °C, 4 h; ii) benzyl
bromide, K CO , triethylamine, acetonitrile, 50 °C, 3 h; (iii) N H xH O, absolute
2
3
2
4
2
ethanol, reflux, 3 h, then RT, overnight; iv) dimethylacetamide, Na S O , 100 °C, 12 h;
2
2
5
v) DCC, NHS , dry DMF, RT, 24 h.
The synthetic work concerned the preparation of two main frameworks, a benzyl-
piperidine-methanamine (1b) and a carboxylic derivative of benzimidazole (1c), as well
as their attachment through an amidic linkage. The preparation of 1b involved a 3-step
reaction. Firstly, the primary amine group of the commercially available
piperidinemethanamine was protected, as a phthalimide group, by reaction with phthalic
anhydride, followed by the piperidine N-benzylation, by reaction with benzyl bromide
in basic conditions, affording 1a. That amine group was subsequently released, from
phthalimide group, by reaction with hydrazine hydrate in absolute ethanol, giving the
free amine derivative (1b). In parallel, the hydroxyphenylbenzimidazole carboxylic acid
(1c) was prepared by Mannich reaction between 3,4-diamino benzoic acid and
salicylaldehyde and cyclization in the presence of the reducing agent sodium
metabissulfite [16]. Finally, the intermediates 1b and 1c were coupled to each other via
amide bond formation, affording the novel hybrid 1. This reaction was carried out in dry
DMF with N-hydroxysuccinimide (NHS) and dicyclohexylcarbodiimide (DCC) as
activators, under N atmosphere for two days.
2
3
.2. Molecular Design and modeling studies of ligand-AChE interactions

Journal Pre-proof
14
DNP is undoubtedly a contemporary choice as anti-AD drug and therefore it was herein
used as an inspiration to design the new hybrid 1. In this way, N-benzylpiperidine was
enclosed in the structure of the hybrid and further connected to a hydroxyphenyl-
benzimidazole (BIM) unit. The N-benzylpiperidine portion mimics the binding
interaction of DNP within the catalytic anionic site (CAS) of AChE [12] while the BIM
moiety is expected to mimic the indanone segment of DNP in its interaction with the
peripheral anionic site (PAS) of the enzyme, besides providing additional properties to
the hybrids, such as anti-Aβ aggregation and metal chelating activity [16,17,40,41].
4
2
It is well known that DNP establishes a strong interaction within the active site of
human acetylcholine esterase, hAChE, and also of the electric ray homologue (Torpedo
californica), TcAChE. Notwithstanding some differences between the respective
inhibitor-enzyme complexes, both enzymes are fairly conservative in terms of the main
aminoacid residues that coat the active site gorge [42] and so the herein described
modeling study was performed with TcAChE. The binding of DNP to TcAChE includes
three types of interactions: in the bottom of the gorge, the benzyl group establishes a
parallel π-π stacking with the indole group of Trp84; in the middle of the gorge, the
charged nitrogen atom can establish a cation-π binding interaction with the phenyl
group of Phe330; at the entrance of the gorge, the indanone ring interacts through π-π
stacking with the peripheral anionic residue Trp279.

Journal Pre-proof
15
Fig. 2. Superposition of compound 1 (above) and 2 [17] (below) and the original ligand
DNP), from PDB code 1EVE, inside the TcAChE active site: DNP (green), compound
(yellow) and compound 2 (brown).
(
1
The docking simulations with compounds 1 and 2 show some differences in their
accommodation inside the gorge of the enzyme, namely concerning the interaction of
the benzylpiperidine moiety with the aminoacid residues of CAS: in the case of
compound 2 there is an excellent superimposition with the original ligand (DNP), while
compound 1, thought it can establish identical binding through the piperidine nitrogen
atom, it slightly deviates from the position of the original ligand (see Fig. 2). On the
other way, the BIM moieties of both compounds seem to be involved quite similarly in
the interactions with PAS, allowing a π-π interaction with the indole group from Trp279,
although probably with a less favorable orientation than DNP. This fact is eventually
due to the larger size of the BIM unit when compared with that of the indanone group of
DNP. Therefore, as a conclusion, the modeling studies seem to predict that the ortho
compound 1 is a less good AChE inhibitor than its positional isomer (compound 2).
3
.3. Acid-base properties
The chelating capacity of compound 1, containing a BIM unit with the carboxamide
substituent in position adjacent (1,2) to the point of attachment between the two rings
2
+
2+
(
benzene and imidazole), was studied by pH-potentiometric (M = Cu , Zn ) as well as
2+
ultraviolet-visible (UV-vis), spectrophotometric (M = Cu ) titrations for the different
M/L systems: 0:1, 1:1 and 1:2. A 50% (w/w) DMSO/H O medium was adopted to
2
perform these studies, due to some solubility problems of the compound. Nevertheless,

Journal Pre-proof
16
in cell studies, the amount of ligand used is lower ( 7 M) and so the final
concentration of DMSO employed in culture media would be even inferior ( 1%),
therefore not expected to involve modifications in biological tissues.
In order to establish the metal chelating model, first the protonation constants of 1 were
determined by both above referred experimental techniques (see Fig. 3) and the values
obtained are contained in Table 1.
1
1
1
2
1
0
9
8
7
6
5
4
3
a)
pH
L
Cu/L 1:1
Cu/L 1:2
Zn/L 1:1
Zn/L 1:2
-
2
-1
0
1
a
0.8
b)
0.6
A 0.4
0.2
0
300
320
340
360
(nm)
380
400


Journal Pre-proof
17
25000
20000
15000
10000
c)
H3L
L
HL
5
000
0
H2L
3
00
320
340
360
(nm)
380
400

-4
Fig. 3. a) Potenciometric titration curves of compound 1 (C = 6.7  10 M) in 50%
L
-5
(
w/w) DMSO/H O; b) Absorption spectra of 1 (2.56  pH  11.53, C = 4  10 M) in
2
L
5
0% (w/w) DMSO/H O; c) respective individual calculated spectra by PSEQUAD [27].
2
Table 1. Stepwise protonation constants of compound 1 and compound 2, for
a
comparison purpose, as well as global formation constants for their Cu(II) and Zn(II))
º
b
complexes (T = 25.0  0.1 C, I = 0.1 M KCl, 50% (w/w) DMSO/water) and pM
values.
Compound
M H L
h l
log Ki
log
log
m
pCu pZn


CumH Ll
Zn mH L
h
h
l
c
(011)
(021)
(031)
(111)
(101)
8.98(2)
d
c
9
.00(1)
7.78(4)
d
c
7
.11(3)
2.86(5)
d
2
.38(8)
c
c
18.96(8)
1
13.81(7)
d
8.864(6)
11.71(7)
c
c
-
14.3 6.4
d
c
1
3.08(2)
c
(
1-11)
(122)
102)
1-12)
4.44(4)
-
13.9
d
d
5
.12(2)
c
c
1
37.27(5)
-
26.51(9)
c
(
22.18(9)
-
d
c
1
9.18(9)
c
(
14.64(6)
3.77(8)
(
(
(
011)
021)
031)
8.671(1)
7.39(1)
3.21(1)

Journal Pre-proof
18
(
(
(
111)
122)
102)
16.43(4)
13.13(3)
25.24(8)
9.94(8)
-
10.7 6.3
32.13(4)
17.91(4)
8.20(6)
e
2
(1-12)
a
m
h
l b
M
= M H L /M H L ; pM = −log[M] at pH 7.4 (C /C = 10, C =
H L
h l
m
h
l
L
M
M
m
−
6
c
d
e
1
0 M) [28]; pH-potentiometric data; UV-vis spectrophotometric data; ref. [40].
Compound 1 was isolated in its neutral mono-protonated form (HL, a = 0 in Fig. 3a))
after the synthetic procedure, although the fully protonated species corresponds to
2
+
H L . The three protonation constants contained in Table 1 for compound 1 are in
3
accordance, as expected, to those previously obtained for compound 2 and correspond
to the protonation of the phenolic oxygen of the BIM moiety (log K ), then the N-
1
piperidine nitrogen atom (log K ) and finally the imidazole nitrogen N(3) of the BIM
2
portion (log K ) [40]. The values obtained by both experimental techniques
3
(
potentiometry and UV-vis spectrophotometry) are quite similar for log K values but
1
present some discrepancies for log K (i =2, 3), probably due to the fact that the UV-vis
i
spectrophotometric titrations have higher experimental errors also since they are not
performed in automated manner as the potentiometric ones.
Taking into consideration the outset in Fig. 3b), enclosing the calculated spectra by
PSEQUAD of the different protonated species for compound 1, it is possible to see that:
2
+
the fully protonated species (H L ) has three absorption maxima at 306, 334 and 343
3
nm; the absorption at 332 nm correspond to both di- and mono-protonated species
+
-
(
H L and HL); the mono-protonated HL and the deprotonated L species have a large
2
absorption band with maximum at ca 373 nm. Moreover, at pH 7.4 and considering CL
-5
=
10 M, as an attempt to mimic the physiological conditions, the predominant species
+
is: for compound 1, H L (70%) with a contribution of 29% of the neutral HL species;
2
+
for compound 2, the neutral HL (50%) and 48% of H L . Even though the lipo-
2
hydrophilic character is not only determined by the molecular charge but also by the
ability to establish solute-solvent interactions, the existence of the neutral HL species in
significant concentrations (ca 29-50%) explains the need of using a 50% (w/w)
DMSO/water medium in the solution studies. In fact, in particular compound 2 showed
to be insoluble for inferior % of DMSO content in the DMSO/H O mixtures.
2
3
3
.4. Metal chelation
.4.1. Solution equilibrium studies

Journal Pre-proof
19
The existence of a BIM moiety in the structure of potential anti-neurodegenerative
drugs was already proven to endow the compounds with capacity for hitting several AD
targets, such as biometal chelation, eventual dual AChE interaction, as well as self- and
Cu-mediated Aβ aggregation [16,17,40]. Concerning metal chelation, the BIM unity is a
bidentate moiety and its (N,O) coordination mode implicates a good potential
coordination towards hard-soft metal ions. The study of the metal chelating properties
of compound 1, a positional isomer of the already studied compound 2, containing the
BIM unit ortho-attached to the benzylpiperazinic moiety, will also show the possibility
of formation of a (N,O,O) metal coordination type involving both the BIM moiety and
the adjacent carbonyl oxygen atom.
The chelating ability of compound 1 was studied by using the same experimental
techniques (potentiometry, UV-vis spectrophotometry) and experimental medium used
for the titration of the ligand alone, 50% (w/w) DMSO/H O. In the following
2
calculations for the 1:1 and 1:2 M/L systems (M = Cu, Zn), the log K values previously
i
obtained by each experimental method will be used in the complexation studies
performed by the same methodology. Fig. 3a) evidences a change in the deprotonation
profile of the ligand titration curve due to the presence of the metal ion since the curves
for the M(II)/L systems lye well below that of the ligand for a  -2 (M = Cu) or a  -1
-
(
M = Zn), thus supporting the formation of metal complexes with the deprotonated (L )
and mono-protonated (HL) forms of the ligand with comparative stability order Cu 
Zn. The potentiometric curves seem to be in accordance with a (N,O) coordination
mode, involving the BIM moiety of compound 1, as already found for other BIM
derivatives [40,43]. Table 1 contains the equilibrium models obtained from the fitting
analysis of the potentiometric curves (M = Cu, Zn) and the UV-vis spectral data (M =
Cu).

Journal Pre-proof
20
a)
0.6
0.4
0.2
0
A
300
320
340
360
(nm)
380
400

b)
30000
20000
10000
0
CuL2
CuH L
-
1
CuHL
CuL
3
00
320
340
360
(nm)
380
400

2
+
Fig. 4. a) Spectrophotometric absorption spectra of the 1:2 Cu / 1 system (2.58 < pH <
-5
1
1.51, C = 4.0  10 M) in 50% (w/w) DMSO/water and b) respective individual
L
calculated spectra by PSEQUAD [27].
2
+
In the obtained metal complex models, MHL and MH L (not found from 1:2 Cu /1
2
2
spectral data) correspond to species with the ligand containing one protonated nitrogen
atom at the piperidine moiety, while ML and ML correspond to complexes involving
2
the completely deprotonated form of the ligand; MH L and MH L (not found from 1:2
-1
-1 2
2
+
Cu /1 spectral data) are mixed ligand-hydroxo metal complexes. The models obtained
2
+
for Cu /1 system from potentiometric and spectrophotometric data (see Table 1) are
not identical but quite similar.
-1
-1
The outset of Fig. 4 evidences that the calculated spectra (max (, M cm ) at 317 nm
3
4
(
9.3  10 ) and 380 nm (1.5  10 )) for both CuHL and CuL complexes are equal,

Journal Pre-proof
21
resulting from the fact that the coordination core is exactly the same and so the
piperidinic nitrogen does not contribute to the copper coordination as already found for
compound 2 [40]; the intensity of the calculated spectrum rises for the CuL species,
2
-1
-1
4
that displays max (, M cm ) values corresponding to a shoulder at 361 nm (3.2  10 )
4
and a band at 374 (3.4  10 ) nm. As expected, to the CuH L species corresponds also a
-1
more intense spectrum with a blue shift of the bands relatively to CuHL/CuL (372 nm
4
(
1.5  10 ), corresponding to the formation of a complex that has an additional
coordination site occupied by an apical water molecule.
Fig. 5 shows the species distribution curves for the 1:2 Cu(II)/1 and Zn(II)/1 systems by
using the potentiometric determined model at the used experimental conditions.
100
80
60
40
20
0
100
80
60
40
20
0
CuH2L2
Zn
ZnHL
CuH L
1 2
-
ZnH L
1 2
-
CuHL
ZnH2L2
CuL2
Cu
ZnH-2
7 . 0
ZnH-3
8 . 0
4 . 0
5 . 0
6 . 0
pH
2
4
6
8
pH
b)
a)
Fig. 5. Species distribution curves for the 1:2 (M/L; L = 1) systems: a) M = Cu; b) M =
-4
Zn (C = 6.7 10 M).
L
Comparison of the metal chelating capacity of compound 1 with that of the already
studied analogous compound 2 (see Table 1) reveals that the pM values (at pH 7.4,
−6
C /C = 10, C = 10 M) [28] are analogous for zinc but higher for copper, which may
L
M
M
point towards the possibility of compound 1 having a different copper coordination
core, eventually including the adjacent carbonyl oxygen atom, with correspondingly
higher stability of the copper complex with 1. In fact, previous studies on the copper
chelation ability of 8-hydroxyquinoline (HQ, pCu = 13.2) in 60% (w/w) DMSO/H O
2
[
44], a commonly known stong chelator, revealed that compound 1 presented a slightly
2
+
higher Cu chelating capacity than HQ while compound 2 shows a small reduction.
To further examine the proposed complexation models, ESI-MS data was also obtained
from solutions of M/1 (M = Cu, Zn), in 50% (w/w) DMSO/H O and 1:1 and 1:2 M/L
2

Journal Pre-proof
22
stoichiometric conditions, at specific pH values. The obtained spectra confirmed the
existence of 1:1 and 1:2 metal complexes (see Table 2 and Fig. 6). The ESI-MS
2
+
analysis of a solution of the 1:1 Cu /1 system, at pH 4.28 and in the negative mode,
showed a group of peaks at m/z 571/573, suggesting a characteristic isotopic
-
distribution of a [Cu(L-H)Cl ] species. However, due to the lower intensity of the
2
-
signal, a clear fit with a calculated isotopic distribution for [(C H N O )CuCl ] was
27
27
4
2
2
not achieved.
2
+
2+
Table 2. ESI-MS ions for the Cu and Zn complexes of compound 1 in 50% (w/w)
-4
DMSO/H O (1:2 (M/L), C = 2.25  10 M).
2
L
M = Cu
M = Zn
(m/z)
Species
pH
(m/z)
-
[
M(L-H)Cl2]
7.01
6.42
8.39
-
573.2/575.2/577.1
-
+
[
M(L-H)L]
M(L-H)L]
942.2/944.2
-
+
[
943.2/945.2
L represents the neutral mono-protonated form of compound 1
a)

Journal Pre-proof
23
b)
Fig. 6. a) Expanded ESI(+) mass spectrum for the Cu(II)/1 system (C /C = 2, C =
L
Cu
L
-4
2
.25  10 M, pH=6.42). The insert shows the theoretical isotopic pattern for
+
[
(C H N O )Cu] ; b) Expanded ESI(-) mass spectrum for the Zn(II)/1 system (C /C
54 55 8 4 L Zn
-4
=
2, C = 2.25  10 M, pH=7.01). The insert shows the theoretical isotopic pattern for
L
-
[
(C H N O )ZnCl ] .
27 27 4 2 2
Finally, to analyze the copper coordination core of the complexes with compounds 1
and 2, EPR spectra of these copper systems were obtained, at room temperature and in
frozen solution together with the calculated EPR spectra of the component species (see
Fig. 7). The solution EPR spectra for the studied systems could be mainly described by
free copper, CuLH and CuL H , as well as the ratio of the species agreed well with the
2
2
distribution curves at the selected pH working values. For the systems with compound 1
1:1) in RT solution and 2 in frozen solution, only the CuLH complex or the free copper
(
and CuLH could be detected, respectively. In those cases, the 1:2 (Cu/L) complex was
not formed. Nevertheless, for compound 1, the frozen solution sample recorded at 1:1
ratio showed a dimer species (Cu L H ), besides CuLH, which frequently forms for
2
2
2
aromatic flat ligands by the off-centered parallel stacking interaction of two mono-
complexes arranged above each other. The frozen solution spectra of the CuLH and
CuL H complexes show well resolved nitrogen splitting, and could be simulated,
2
2
respectively, with one nitrogen and two nitrogens, in accordance with the involvement
of one or to BIM chelating cores. The obtained EPR data are collected in Table 3.

Journal Pre-proof
24
The comparison of the EPR data for the CuLH complexes of compounds 1 and 2 allows
concluding that the isotropic and anisotropic g values reflect higher ligand field for
compound 1 than for compound 2, since the g values are lower for compound 1. This
result supports the already suggested (N,O,O) tridentate coordination mode for
compound 1, with the involvement of the adjacent carbonyl oxygen atom. For the 1:2
copper complex of compound 1 it is probable that the second ligand coordinates
equatorially with its imidazole nitrogen and with a lower coordination angle with the
phenolate oxygen atom, which seems to be in accordance with the geometries found by
molecular modeling studies (see section 3.4.2.). This coordination change agrees with
the facts that g value decreases only 0.01, comparing with the same value of the 1:1
z
complex, and the A value also declines evidencing an axial coordination with small
z
rhombic distortion.
a)

Journal Pre-proof
25
b)
2
+
Fig. 7. Measured (black) and calculated (grey) EPR spectra obtained for: a) Cu /1 1:1
2
+
(
pH = 6.75) and 1:1.5 (pH = 6.91); b) Cu /2 1:1 (pH = 5.46) and 1:2 (pH = 5.72)
systems at room temperature (left) and in frozen solution (right).
2
+
Table 3 - Isotropic and anisotropic EPR parameters obtained for the Cu
a
complexes with compounds 1 and 2
Compound/Sp g0 A0
a
g
gx
gy
gz Ax Ay Az
/G /G /G
a
a
a
z
0
0,cal
x
y
N
b
c
N
N
N
ecies
/G
/
G
/G /G /G
1
CuLH
2.1 54. 14. 2.13 2.0 2.0 2.2 15. 16. 161 14. 17. 4.8
58 57 87 .4
2.1 66. 14. 2.12 2.0 2.0 2.2 20. 3.9 149 14. 15. 9.9
48 59 75 .0
2
3
9
5
4
8
2
2
3
CuL H
2
2
2
0
0
5
7
7
5
0
10. 11. 10.
1
9
3
c
Cu L H
2
2.0
2
2
6
6
2
Cu
2.1 31.
2.19 2.0
2.4 6.9 11

Journal Pre-proof
26
9
4
6
6
86
2.1 59. 13. 2.14 2.0
60
17
2.3 15.
24
0
CuLH
145 14. 14. 10.
5
6
6
6
8
2
1
1
0
CuL H
2.1 61. 10.
2
2
1
1
3
8
a
The experimental errors were ±0.002 for g and g and ±0.001 for g and g ; ± 2 G for A and
A and ±1 G for A and A Calculated by the equation g
x
y
z
0
c
x
b
= (g +g +g )/3. The dimer species
x y z
y
z
0
0.calc
was calculated as a doublet. The peak distances resulted in the dipolar coupling of D = 287 G,
from which the copper-copper distance of 4.14 Å can be calculated.
So, the spectral data collected, namely EPR information, and the callculated equilibrium
models obtained for compound 1 point towards a chelating core composed by the N(3)
imidazole and O-phenol atoms of the ligands, excluding the hypothesis of metal
coordination through the distant nitrogen atom of the piperidine [40], but eventually
including the adjacent carbonyl oxygen atom in the case of the copper complexes.
3
.4.2. Molecular modeling of the copper complexes
In the absence of X-ray structures, a molecular modelling study was accomplished
aimed at providing some further insight into the 1:1 and 1:2 (M/L) copper complex
structures with compounds 1 and 2. In that way, a tridentate (N,O,O) coordination mode
was considered for compound 1, involving one coordinated water molecule, while a
bidentate (N,O) coordination type was adopted for compound 2 and two coordinated
water molecules were included in the chelating core. All the molecular modelling
2
+
studies were carried out with full geometry optimization of the Cu complexes by
quantum mechanical calculations based on DFT methods [30] included in the Gaussian
0
3 program software [31] with the B3PW91 functional [31]. Geometry optimizations
were obtained by using 3-21G basis set [33] for all elements, first, and afterwards re-
optimized using 6-31G** [33] for the main group elements and the Stuttgart–Dresden
pseudopotential and associated basis set for Cu (SDD [45]). No symmetry constrains
were enforced during geometry optimizations. The calculations were performed as well
for one water molecule by using the B3PW91 functional, first with the 3-21G basis set
and subsequently re-optimized via 6-31G** basis set.

Journal Pre-proof
27
The energy-minimized structures of the 1:1 copper complexes present a distorted square
planar four-coordinate configuration for 1 while for 2 a distorted tetrahedral geometry is
indicated (see Fig. 8).
Fig. 8. DFT-minimized structures of 1:1 copper complexes with compound a) 1 and b)
2
. Coloring of atoms: Cu (light blue), N (blue), O (red), C (grey) and H (white).
In the obtained modeling simulations, some identical coordination bond distances can
be found between the two structures for Cu-H O (2.02−2.05 Å) and Cu-O-phenol
2
(1.86−1.88 Å), while the Cu-N distance increases somehow from the copper complex

Journal Pre-proof
28
with 1 (1.88 Å) to that with 2 (1.91 Å), probably due to the stronger tridentate
coordination established in the case of complex with 1, that also involves the adjacent
O-carbonyl atom (distance Cu-O-carbonyl 1.96 Å). In the case of the complex with
compound 1, the distortion in relation to a regular square planar geometry is well
°
demonstrated by the coordination bond angles (83.6−93.0 ) and the almost planar
°
geometry as showed by the bond angles of O-phenol-Cu-O-carbonyl (173.1 ) and N-
°
Cu-water (174.8 ). On the other hand, the copper coordination with compound 2 also
exhibits a strong distortion from a tetrahedral geometry, with coordination bond angles
°
ranging from 82.4 to 99.7 .
Since these two complexes do not have exactly the same number of atoms, comparison
of complex stability was based on the determination of the free energy difference (G =
2
5.6 kcal/mol) between Cu-1 complex plus one water molecule and Cu-2 molecule.
Therefore, the calculations show that the copper complex with compound 1 is more
stable than that with compound 2, which is according to the previously presented
solution studies.
Fig. 9 shows that the energy-minimized structures of the 1:2 copper complexes present
a distorted four-coordinate tetrahedral geometry for the complex with compound 2,
while for the respective complex with isomer 1 the geometry is intermediate between
square planar and tetrahedral.

Journal Pre-proof
29
O
N
N
Cu
O
a)
Fig. 9. DFT-minimized structures of 1:2 (M:L) copper complexes with the ligands, a) 1
and b) 2. Coloring of atoms: Cu (light blue), N (blue), O (red), C (grey), H (white).
Hydrogen atoms linked to carbon atoms are hidden for clarity purposes.