Multifunctional iron-chelators with protective roles against neurodegenerative diseases

Article abstract of DOI:10.1039/c3dt50406a

The multifactorial nature of Alzheimer's disease (AD), and the absence of a disease modifying drug, makes the development of new multifunctional drugs an attractive therapeutic strategy. Taking into account the hallmarks of AD patient brains, such as low levels of acetylcholine, misfolding of proteins and associated beta-amyloid (Aβ) aggregation, oxidative stress and metal dyshomeostasis, we have developed a series of compounds that merge three different approaches: metal attenuation, anti-Aβ aggregation and anti-acetylcholinesterase activity. Therefore, 3-hydroxy-4-pyridinone (3,4-HP) and benzothiazole molecular moieties were selected as starting frameworks due to their well known affinity for iron and Aβ peptides, respectively. The linkers between these two main functional groups were selected on the basis of virtual screening, so that the final molecule could further inhibit the acetylcholinesterase, responsible for the cholinergic losses. We describe herein the design and synthesis of the new hybrid compounds, followed by the assessment of solution properties, namely iron chelation and anti-oxidant capacity. The compounds were bioassayed for their capacity to inhibit AChE, as well as self- and Zn mediated-Aβ_1-42 aggregation. Finally, we assessed their effects on the viability of neuronal cells stressed with Aβ₄₂.

Full text of DOI:10.1039/c3dt50406a

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Multifunctional iron-chelators with protective roles
against neurodegenerative diseases†
Cite this: Dalton Trans., 2013, 42, 6058
Andreia Nunes,a Sérgio M. Marques,a Catarina Quintanova,a Diana F. Silva,b
Sandra M. Cardoso,b,c Sílvia Chavesa and M. Amélia Santos*a
The multifactorial nature of Alzheimer’s disease (AD), and the absence of a disease modifying drug,
makes the development of new multifunctional drugs an attractive therapeutic strategy. Taking into
account the hallmarks of AD patient brains, such as low levels of acetylcholine, misfolding of proteins
and associated beta-amyloid (Aβ) aggregation, oxidative stress and metal dyshomeostasis, we have devel-
oped a series of compounds that merge three different approaches: metal attenuation, anti-Aβ aggrega-
tion and anti-acetylcholinesterase activity. Therefore, 3-hydroxy-4-pyridinone (3,4-HP) and benzothiazole
molecular moieties were selected as starting frameworks due to their well known affinity for iron and Aβ
peptides, respectively. The linkers between these two main functional groups were selected on the basis
of virtual screening, so that the final molecule could further inhibit the acetylcholinesterase, responsible
for the cholinergic losses. We describe herein the design and synthesis of the new hybrid compounds, fol-
lowed by the assessment of solution properties, namely iron chelation and anti-oxidant capacity. The
compounds were bioassayed for their capacity to inhibit AChE, as well as self- and Zn mediated-Aβ1–42
Received 8th October 2012,
Accepted 22nd February 2013
DOI: 10.1039/c3dt50406a
www.rsc.org/dalton
aggregation. Finally, we assessed their effects on the viability of neuronal cells stressed with Aβ42.
toxicity associated with the aggregation of Aβ fibrils has been
Introduction
considered one of the major justifications for the loss of neu-
Alzheimer’s disease (AD) is a chronic, irreversible and pro- ronal cells that occur in AD patients.5
gressive disorder, which is characterized by dementia, cogni-
Although the mechanism of protein folding and aggrega-
tive impairment and memory loss and, in an advanced stage, tion in neurodegeneration remains unclear, there is plenty of
it can lead to incapacitation and finally to death.1,2 AD etiology evidence about the role of several factors in that process, such
is not completely understood, but it is recognized as a multi- as redox-active metal ions and oxidative stress.6 In fact, the
faceted disorder in which the patients brains show several fea- brain metal concentration seems to increase with age,7,8 and
tures such as low levels of acetylcholine, misfolding of this metal imbalance can be responsible for two major contri-
proteins and associated aggregation processes, oxidative stress butions to the neurodegeneration process: the redox-active
and free radical formation, metal dyshomeostasis, protein metal ions, such as Fe(III) and Cu(II), can be reduced by H2O2
phosphorylation impairment as well as mitochondrial and and promote the Fenton reaction with the formation of reac-
neuroinflammatory dysfunctions. Thus, the complexity of this tive oxygen species (ROS), which may then oxidize biomole-
pathology justifies the absence of effective disease-modifying cules and induce severe damage in neuronal cells;9 several
AD therapeutics.3,4
metal ions, such as Zn(II), can modify protein structure (mis-
The senile plaques and neurofibrillary tangles are the main folding) and lead to abnormal aggregation and formation of
hallmarks of AD and result from deposition of a beta-amyloid beta-amyloid deposits.10,11 Therefore, the critical role of the
(Aβ) peptide and τ-protein, respectively. However, the cellular metal imbalance in neurodegeneration has made the chelating
therapy an attractive strategy to be used against the develop-
ment and the progression of AD and other neurodegenerative
disorders.12–14
aCentro de Química Estrutural, Instituto Superior Técnico, Universidade Técnica de
The dysfunctional Aβ has been the pivotal pharmacological
target for AD, but up to now the only approved AD therapy is
for improving the symptomatic cognitive deficits through the
inhibition of acetylcholinesterase (AChE) (e.g. donepezil,
tacrine and rivastigmine), although only moderate benefits
have been reported.15 Quite often coadjutant drugs have also
Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal. E-mail: masantos@ist.utl.pt;
Fax: +351-218464455; Tel: +351-218419273
bCentro de Neurociências e Biologia Celular, Universidade de Coimbra,
3030 Coimbra, Portugal
cFaculdade de Medicina, Universidade de Coimbra, 3030 Coimbra, Portugal
†Electronic supplementary information (ESI)
available. See DOI:
10.1039/c3dt50406a
6058 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
been used, including antioxidants (e.g. poly-phenolic and chal- BTA) were connected with a linker, which was modeled by in
cone derivatives)16 and metal-chelators (e.g. desferrioxamine silico tools in order to best provide AChE inhibitory capacity.
and hydroxyquinolines).13,14,17
In summary, herein we describe the design of these BTA–
Hence, due to the recognized multi-factorial nature of AD, a HP polyfunctional compounds, the synthetic procedure, the
new paradigm has arisen to address this pathology, according evaluation of their solution properties (acid–base, iron com-
to which one single molecular entity can hit more than one plexation, anti-oxidant) and biological properties, such as the
target (multifunctional drugs). This leads to the development assessment of their capacity for inhibition of AChE and of self-
of hybrid compounds with at least two of the following pro- and zinc-mediated-Aβ aggregation, as well as their ability to
perties: cholinergic, antioxidant, metal attenuator, Aβ rescue neuronal cells stressed with Aβ, in the absence/presence
reductor.18,19
of iron/ascorbate.
Based on the same assumption, we have decided to develop
and study a new series of compounds that will merge different
pharmacological targets, as prospective drug candidates for
AD treatment. Specifically, in this paper we combine metal-
attenuation and anti-oxidant activity with anti-Aβ aggregation/
vectorization roles, as well as a potential cholinergic effect.
As metal-attenuators, 3-hydroxy-4-pyridinones (3,4-HP) are
proposed, due to their “privileged” properties for medicinal
metal related drugs, namely their high affinity for iron (but
low affinity for common biological electrolyte ions), anti-
oxidant properties and low toxicity (e.g. deferiprone, DFP, is a
current iron-chelating drug, see Fig. 1).20 The extrafunctionali-
zation with benzothiazole (BTA), as a thioflavin-T (ThT) deriva-
tive, is based on its known strong affinity for amyloid fibrils
and use in Aβ imaging probes.21 Recent reports have also pro-
posed a new therapeutic strategy against neurodegenerative
diseases, based on using chemical agents capable of disrupt-
ing metal–peptide interactions which are believed to induce
the Aβ aggregation. The correlation between metal ions and
amyloid plaques has been investigated, including the role of
metals in promoting Aβ aggregation and the formation of ROS,
and consequently oxidative stress.22,23 Very recently, a few
hybrid compounds with metal-chelating and amyloidophilicity
properties have been reported.14,24,25 For this new series of
potential AD hybrid drugs, the main moieties (3,4-HP and
Results and discussion
Molecular modeling
The strategy followed for the design of the new inhibitors con-
sisted of making a pre-selection of several connecting groups
(spacers) between the two main moieties, the hydroxypyridi-
none (HP) and the benzothiazole (BTA) (Fig. 1), and then per-
forming the virtual screening of these molecules against a
target enzyme (AChE). The spacer groups tested derived from a
combination of amino acid precursors that vary in length, lipo-
philic character as well as number and position of H-donating
capabilities.
Analysis of the crystallographic structures of AChE showed
that the catalytic centre of the enzyme is found in the bottom
of a deep and narrow gorge that spreads for ∼20 Å and is lined
by 14 aromatic residues with functional roles, which are con-
served between the different species. The active center is com-
posed of three main areas: the catalytic center, which is found
in the bottom of the gorge, consists of the catalytic triad:
Ser200, His440 and Glu327. Here is where the hydrolysis of the
ester bond of acetylcholine (ACh) occurs. The catalytic active
site (CAS), located in the proximities of the catalytic triad, is
composed of two aromatic residues, Trp84 and Phe330; the
Fig. 1 Deferiprone (DFP), thioflavin-T (ThT), donepezil (Dnp) and schematic representation of the new benzothiazole–hydroxypyridinone (BTA–HP) hybrid
compounds.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6059

View Article Online
Paper
Dalton Transactions
peripheral active site (PAS), which is located at the entrance of interactions at the more external part of the catalytic site,
the gorge, presents also two aromatic residues, Tyr70 and namely with PAS (Tyr70 and Trp279 residues), and a weak
Trp279.26–29 It has been proven that the PAS of the AChE can H-bond between the BTA sulfur atom and the Tyr70 OH group.
promote aggregation and deposition of the Aβ peptides. This Also some aromatic interactions between the pyridinic ring of
discovery has stimulated a great interest in a strategy of bi- HP and Phe330 can be formed, although to a lesser extent.
valent ligands that can simultaneously inhibit the hydrolysis The conformation adopted by 1b is not very different from 1a,
of ACh and the aggregation of Aβ induced by AChE.26,27
only slightly distorted within the gorge. Regarding the HP–OH
The docking calculations were performed with GOLD soft- analogues, their conformation can be quite different. Analysis
ware,30 using as a receptor an X-ray structure of the AChE, of the docking conformation obtained with 2a (analogue of 1a)
Torpedo californica variant (TcAChE), complexed with donepe- shows that the BTA group “takes” the place of the benzyl group
zil (Dnp), which was taken from the RCSB Protein Data Bank of the previous compounds at the CAS, forming π–π stacking
(PDB, entry code 1EVE).31 Dnp has some structural parallelism with the Trp84 residue, whereas the adjacent phenyl ring
with our compounds in terms of enzyme interaction features seems to highly interact with the Tyr334 phenol ring (Fig. 3b
(the BTA is expected to occupy the position of the benzyl group and Fig. S2†). The HP moiety interacts at the PAS with Trp279,
of Dnp). This fact ensures a higher proximity of the position of while its OH group is solvent exposed and its carbonyl group
the amino acid side chains in the model structure with respect may form an H-bond with the Tyr70 OH group. On the other
to our case.
hand, compound 2b (analogue of 1b) is docked in the opposite
In Fig. 2 are displayed the compounds which presented the direction. Its HP ring seems to interact at the CAS with Trp84
best results from the docking. Series 1 corresponds to the and the OH and carbonyl groups establish H-bonds with
O-benzyl protected analogues of series 2 compounds, which are Gly117 carbonyl and Tyr130 OH groups, respectively, while the
in fact the target BTA–HP hybrid multifunctional compounds. adjacent phenyl group forms van der Waals contact with the
Those compounds presented, in general, higher affinities for Phe330 aromatic ring. In this case, the BTA group is pointing
binding the AChE enzyme than the free HP analogues (series outside the cavity and forms π–π stacking with Trp279 at the
2a–2d). This fact is attributed to the strong lipophilic character PAS. According to docking, compound 2d seems to bind AChE
of the AChE catalytic gorge, which consequently interacts in a similar way to 2a, with its 5-membered ring forming
better with the more hydrophobic compounds 1. The hydroxyl hydrophobic interactions with both Phe330 and Tyr334 resi-
protection of these compounds with a benzyl group also might dues (Fig. 3c and Fig. S2†).
have a positive contribution by enhancing the blood–brain
barrier (BBB) crossing ability. To some extent, the O-benzylated
Chemistry
compounds may modulate the corresponding O-carbamate
derivatives which can be considered as prodrugs.
The synthesis of the BTA–HP compounds involved several
steps, including standard protection and activation strategies
(see Fig. 4).
The first stage of syntheses consisted of the preparation on
the two main molecular moieties, the HP and the BTA. For the
HP synthesis, the maltol (3-hydroxy-2-methyl-4-pyrone) was
first protected with a benzyl group to avoid side reactions with
the hydroxyl group in the subsequent steps. The second step
corresponded to a Michael reaction between the O-benzyl
The highest docking scores were obtained for 1a, 1b and 2c
(see Table S1 in ESI†). It was observed that these compounds
can be well inserted into the active site, maintaining several of
the aromatic interactions established by the reference inhibi-
tor, Dnp (see Fig. 3a and Fig. S1, ESI†). The O-substituent
benzyl group in 1a forms π–π stacking with the Trp84 residue
of the CAS, while the BTA–CH2–Ph-system may form aromatic
Fig. 2 General formula of the hybrid benzothiazole–hydroxypyridinone (BTA–HP) chelators (2a–2d), and corresponding benzyl derivatives (1a–1d), with the list of
the linkers selected from in silico design.
6060 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
After preparation of the two molecular fragments, the final
hybrid compounds (1 and 2) were obtained by condensation of
the BTA-amine with the HP-carboxylic compounds via for-
mation of an amide linkage. The reaction conditions
depended on the presence of the benzyl group in the final HP
moiety. In the case of the O-benzylated compounds (com-
pounds 5), there is no danger of attack in the HP group, so the
coupling reaction involved the use of a current activating agent
(EDC·HCl). As regards the unprotected compounds (com-
pounds 6), the OH group of the HP moiety is prone to be
attacked, so there is need of using a protecting group. In this
case we were able to do it in situ with 2 equivalents of ethyl-
chloroformate (ECF), which acts not only as an activating
agent of the carboxylic moiety, but also as a protecting agent
for the HP–OH group to form the respective ethyl carbonate.
The deprotection of the final product was easily accomplished
by hydrolysis with a basic solution. We have tried to prepare
compounds 2 by simply hydrogenating the benzyl-analogues,
1. However, due to the presence of a sulfide group in the BTA
moiety, the Pd/C catalyst was poisoned and the hydrogenation
could not be accomplished.
Acid–base properties
Four of the developed compounds (2a–2d) include a free HP
chelating moiety in their structure. Since HPs are well known
as strong hard metal ion chelators, these compounds are
thought to have an important role in diminishing ROS. There-
fore, 2a and 2d were chosen as models to confirm the high
iron chelating ability of these compounds and the results are
herein compared to those obtained for DFP (1,2-dimethyl-3-
hydroxy-4-pyridinone, DMHP, see Fig. 1) under the same exper-
imental conditions. In order to study the iron chelating ability,
the acid–base behavior of the compounds ought to be first
analyzed. To improve the solubility of these BTA–HP deriva-
tives, especially 2a, the solution studies were performed in a
mixed DMSO–H2O medium (50% w/w). Nevertheless, in terms
of cellular studies, the final concentration of DMSO used in
culture media did not exceed 0.05% (v/v) and no alterations
were observed in cells.
To use lower ligand concentrations, spectroscopic protona-
tion titrations were performed instead of potentiometric ones.
The stepwise protonation constants were obtained by fitting
analysis of the spectrophotometric data with the PSEQUAD
program32 and the values are included in Table 1.
Fig. 3 Docking results for some BTA–HP derivatives with AChE: (a) superimposi-
tion of 1a (cyan) with Dnp (magenta); (b) 2a (green) and 2b (yellow); (c) 2a
(green) and 2d (red). H-bonds are represented as solid black lines.
Compounds 2a and 2d were isolated in the neutral form
(HL) but both have two dissociable protons (H2L+). The proto-
protected pyrone, 3, and the suitable aminocarboxylic acids to
obtain compounds 4. These intermediates can be directly nation equilibria for this set of unprotected BTA–HP com-
coupled with the BTA derivatives (compounds 5) to afford com-
pounds 1, or they can be subjected to a prior deprotection to
pounds are shown in Scheme 1, where the first protonation
constant (K1) corresponds to the protonation of the 3-hydroxy
give compounds 6 and subsequently compounds 2. For depro- whereas the second one (K2) is attributed to the 4-hydroxy-
tection of the hydroxyl group, a standard hydrogenolysis pro-
cedure was used (1.5 bar H2 with Pd/C as a catalyst).
(N-pyridinyl) groups, as found for DFP.
Table 1 shows that the obtained log Ki values are in accord-
Regarding the benzothiazole derivatives, 5, their synthesis ance with those determined herein in the same solvent
involved the condensation of 2-aminothiophenol and different
amino acid derivatives in polyphosphoric acid to afford com-
pounds 5a–5c.
mixture for DFP. The log K2 value obtained in water for DFP is
also similar to the values achieved in the solvent mixture
although the log K1 value is lower in water (9.77) than in the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6061

View Article Online
Paper
Dalton Transactions
Fig. 4 Reagents and conditions: (a) 10% NaOH, MeOH, reflux or 0.1 M HCl–EtOH, reflux; (b) polyphosphoric acid, 220 °C; (c) EDC·HCl, DMAP, CH2Cl2, rt; (d) 10%
Pd/C, MeOH, H2, rt; (e) ECF, NMM, dry CH2Cl2; (f) 2% NaHCO3–EtOH, 65 °C. For simplification, we defined Z as the carbon chain of Y, where X and Y are the same
substituents as in Fig. 2. In compounds 1d and 2d Z is cyclic, and therefore, in this case, “ZNH” has no H-atom attached to ZN.
protonation of 2d. From this figure, it is clear that for
5.5 < pH < 8.0, 100% of the neutral HL species is present,
while for pH above 10.6 the predominant species is totally
deprotonated (L−).
Table 1 Stepwise protonation constants (log Ki) of 2a, 2d and DFP as well as
the global formation constants of their Fe(III) complexes (I = 0.1 M KCl, 50%
DMSO–H2O, T = 25.0 0.1 °C)
Compound
log K1
2a
2d
DFP
10.38(4)
3.69(8)
16.67(7)
29.02(2)
37.71(4)
19.4
10.55(1)
3.42(3)
16.00(6)
28.80(2)
38.08(4)
19.2
10.77(1)
9.77a
Iron chelating capacity
log K2
3.51(3)
3.62a
The complexation abilities of 2a and 2d towards iron(III) were
evaluated through the determination of the global stability
constants of their complexes as well as the pattern of the
species distribution curves. Once more, the spectrophoto-
metric method was adopted, instead of the potentiometric
one, because at pH 2 a high percent of FeL complex was
already formed. The global stability constants (see Table 1)
were calculated from the fitting analysis of the measured
UV-Vis spectra with the aid of the PSEQUAD program.32 Firstly,
below pH 2, under 1 : 1 M/L stoichiometric ratio conditions, a
batch titration was performed in order to determine βFeL; after-
log βFeL
16.26(5)
15.14a
29.86(2)
26.68a
40.39(3)
35.92a
20.8
log βFeL
2
log β
FeL3
pFeb
log P
−0.85
a Ref. 33 (I = 0.1 M KCl, T = 25.0 °C, in water). b pFe values at pH = 7.4
(CFe = 10−6 M, CL/CFe = 10).
50% w/w DMSO–H2O medium (10.4–10.6), which is according wards, for pH > 2, a spectrophotometric titration was per-
to the Born electrostatic solvent effect.34
formed in a 1 : 3 Fe/L molar ratio and its data analysis (while
As a representative example, Fig. S3† shows the electronic keeping constant the log βFeL value) allowed the determination
spectra and species distribution curves relative to the of log βFeL and log βFeL
.
3
2
6062 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Scheme 1 Protonation equilibria for the unprotected BTA–HP (2a–2d) compounds.
absorbance maximum at 574 nm should correspond to the
monochelated species (FeL). Furthermore, Fig. 5b shows that
the main complex formed at pH above 5.7 is the hexacoordi-
nated species (tris-chelate). Analysis of data depicted in
Table 1, in particular the pFe values at pH 7.4, evidences that
the studied compounds are strong iron chelators, similarly to
the oral clinical chelating agent DFP. The high iron(III)-chelat-
ing capacity should have an important role in the anti-oxidant
activity of the compounds, by preventing the reduction of
Fe(III) to Fe(II) and the concomitant hydroxyl radical (HO˙) for-
mation, as previously reported for DFP (redox behaviour and
HO˙ scavenging activity), which showed to be able to prevent
the redox cycling of iron.36
Anti-oxidant activity
The new compounds were screened for their anti-oxidant
activity (EC50), based on their interaction with the 2,2-diphe-
nyl-1-picrylhydrazyl (DPPH) free radical, and the correspond-
ing results are summarized in Table 2.
Analysis of the results clearly shows that, as expected, the
compounds with an O-protected HP moiety (compounds 1)
presented much lower anti-oxidant activity (all with EC50
>
1 mM) than those with a free hydroxyl group, which is sup-
posed to have an important role in inhibition of the radical
formation and also in its capture (compounds 2).36 Regarding
the unprotected compounds, 2c appears to be the least anti-
oxidant compound (713 μM), although 2a and 2b also did not
present particularly good EC50 values. Among the set of
studied compounds, 2d showed the best anti-oxidant capacity
(147 μM), very similar to that of DFP. This appears as a promis-
ing result, considering that this compound also presented
good AChE inhibitory capacity (see below). The better anti-
oxidant activity of 2d may be explained by the fact that, when
compared with the other analogues, this compound does not
include any aromatic segment in the spacer between the BTA
and HP groups, providing the lowest lipophilic character,
which can favor solubility and molecular interactions. These
features may contribute to some restriction on the molecular
ability to adopt a conformation where the opposite side (BTA
group) can make steric hindrance, which may interfere with
the interaction between the HP moiety of the compounds and
the radical that is to be extinguished.
Fig. 5 (a) Electronic spectra (pH 1.2–9.2) and (b) species distribution curves
with molar extinction coefficients at maximum absorption wavelengths for the
Fe3+/2d system (CL/CFe = 3.1, CFe = 4.51 × 10−5 M).
Fig. 5 (example for 2d) evidences the formation of succes-
sively stronger complexes, from bis- to hexa-coordinated
species, as indicated by the blue shift and the absorbance
increase on the metal-to-ligand charge-transfer (CT) band.
Analysis of Fig. 5a shows a clear isosbestic point at ca.
510 nm, relative to the interconversion of the bischelated
(FeL2) to the trischelated (FeL3) species. Moreover, the species
distribution curves depicted in Fig. 5b, combined with the
absorptivity at 460 and 520 nm, reveal that the maximum
absorbance at 460 nm (ε = 5173 M−1 cm−1) and 520 nm (ε =
3632 M−1 cm−1) corresponds, respectively, to the FeL3 and
Pharmacokinetic properties
FeL2 CT bands. The observed spectral parameters for the FeL3 In order to estimate the potential of the new compounds as
CT band, especially the λmax value, are analogous to those eventual drugs, a few indicators of their pharmacokinetic pro-
obtained for the CT band of the ferric tris-chelated complex files were calculated using the QikProp program.37 Parameters
with DFP in 50% w/w DMSO–H2O (460 nm, 1533 M−1 cm−1) or such as the lipo-hydrophilic character (log P), the ability to
even in water (460 nm, 5800 M−1 cm−1).35 The band with an cross the blood–brain barrier (log BB), their ability to be
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6063

View Article Online
Paper
Dalton Transactions
Table 2 Biological activities of the BTA–HP derivatives (1 and 2 series), and also the FDA-approved drugs Dnp and DFP, towards radical scavenging (DPPH), inhi-
bition of AChE and Aβ1–42 aggregation, as well as some predicted pharmacokinetic parameters
Compd
EC50 DPPHa (μM)
IC50 AChEia (μM)
Anti-Aβ aggreg.a,b (%)
clog Pc,d
log BBc,e
Caco-2 permeability (nm s−1 c
)
1a
1b
1c
1d
2a
2b
2c
2d
Dnp
DFP
>1000
>1000
>1000
>1000
397
428
713
22.3
29.2
13.8
285
>500
>500
>500
14.7
0.033 f
—
45.6
19.7
68.9
47.5
—
—
—
59.3
—
—
6.75
7.71
6.68
4.46
4.06
5.11
4.08
2.30
4.43
0.62
−1.47
−1.08
−0.87
−0.58
−1.93
−1.54
−1.28
−1.04
0.13
702
1169
1231
1289
180
332
372
301
913
1099
148
—
133
−0.28
a The standard deviation is within 10% of the values. b Inhibition of self-mediated Aβ (1–42) aggregation. The thioflavin-T fluorescence method
was used, and the measurements were carried out in the presence of an inhibitor (100 μM). c Predicted values using the QikProp program, v. 2.5,
ref. 37. d Calculated octanol–water partition coefficient. e Calculated brain–blood partition coefficient. f Ref. 38.
absorbed through the intestinal tract to the blood (Caco-2 cell Trp279. Most of the compounds with the highest inhibitory
permeability), and the verification of Lipinski’s rule of five may activities belong to the series 1 (1a, 1b, 1c), which includes the
give an idea of their druglikeness for being orally used as anti- benzyl group as well as one aromatic ring in the linker, to
AD agents.
account for hydrophobic interactions within the gorge. Inter-
As observed from Table 2, 1a–1c presented the highest pre- estingly, among the whole set 2 of unprotected compounds, 2d
dicted octanol–water log P coefficients (clog P), which are >5. presented the second highest inhibitory activity (14.7 μM), and
They also have molecular weights (MW) > 500, which makes even much higher than its benzyl analogue, 1d (285 μM),
them to have 2 Lipinski’s rule violations. This fact, together which presented the lowest inhibitory capacity of series 1. In
with the negative log BB (<−1.0 for 1a and 1b), does not lead fact, this result was not predicted by the virtual screening,
them to be likely good drug candidates, even if they show high which suggested for 2d the lowest degree of interactions with
Caco-2 permeability rates (>500 nm s−1 is considered good). AChE. A further analysis of these results (Fig. 3c and S2†) indi-
Compound 1d presents no Lipinski’s rule violation, and it has cated that the 2d BTA group might be able to form π–π stack-
the highest log BB and Caco-2 permeability values of all the ing with the Trp84 residue of CAS. Its 5-membered ring could
new compounds.
also establish van der Waals interactions within the AChE
Regarding compounds 2, all except 2b presented more gorge, and an H-bond between the HP-carbonyl and the Tyr70-
moderate log P and MW < 500. However, for 2a–2c, their OH group. The other compounds of series 2 (2a–2c) showed
log BB values are quite low, 2d having the highest value (−1.0). high IC50 values (>500 μM), thus, irrelevant inhibitory activity.
Considering the Caco-2 permeability, it seems to be moderate Therefore, some extra mechanism, unpredicted by the prelimi-
(between 180 and 372 nm s−1), and again, compound 2d dis- nary docking, must subsist to rationalize a big difference
played a relatively good value (301 nm s−1). Comparatively, the between the activities of 2d and the other unprotected BTA–HP
commercial drugs Dnp and DFP showed better pharmaco- derivatives.
kinetic parameters, namely for their log BB and Caco-2 per-
meability values.
Overall, the new compounds are weaker AChE inhibitors
than the reference inhibitor Dnp (420-fold more potent than
Altogether, these results suggested 1d and 2d as the most the best inhibitor, 1c). This feature, however, may be counter-
“drug-like” compounds of the whole set of new compounds.
balanced by eventual advantages of their extra functionalities.
The fact that the benzyl-protected derivatives (series 1) gave, in
general, better and reasonably good anti-AChE properties, may
lead to a further approach for the design of the related carba-
mate-protected BTA–HP pro-drugs.
Inhibition of Aβ42 self-induced aggregation. A number of
studies with a synthetic Aβ42 peptide have shown peptide
aggregates in vitro with amyloid cross-β-fibrils similar to the
ones found in the brains of AD patients.39,40
The in vitro Aβ aggregation can be evaluated by a method
based on the fluorescence emission by using thioflavin
T (ThT). ThT has been used for many years as a histologic
agent for the detection of amyloid protein deposits. Its
peculiar change in optical properties upon binding with the
β-pleated sheet structures of amyloid fibrils, namely with
Biological activity
AChE inhibition. The inhibitory profile of the new com-
pounds towards AChE is presented in Table 2. The best inhibi-
tory activities were found for 1a, 1b, 1c and 2d, presenting IC50
values in a low micromolar range (ca. 14–29 μM). Some of the
most active compounds are O-protected HP derivatives, which
may be attributed to an extra-interaction between the benzyl
group and aromatic residues of the enzyme active site. In fact,
the docking simulations had evidenced potential π–π stacking
interaction between the HP-O-benzyl group and the Trp84
residue of the CAS, thus anticipating a potentially good activity
for these compounds. The BTA moiety also seemed to esta-
blish aromatic interactions with the PAS residues Tyr70 and
6064 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
formation of an intense fluorescence band, not observed with concentration) is not possible, these results clearly indicate
the non-amyloidogenic conformation (unordered/α-helix) of the existence of an order inversion on their inhibitory capacity.
the soluble Aβ42 small oligomers, allows the quantification of This result may be rationalized in terms of an interplay of two
the Aβ aggregation by monitoring the fluorescence at 446 nm different effects enabled by these bifunctional compounds:
(λexc) and 490 nm (λem).41 A preliminary time scan study of the one, due to the ligand–Aβ interactions, namely involving BTA
ThT signal showed that the fluorescence signal reached a and other groups of the ligands; another one, based on the
plateau after 5 min and so the fluorimetric analysis was per- chelating effect, because the chelating group can compete with
formed at this selected time.
the peptide for the metal ion and so disrupt the metal-induced
Based on this information, the Aβ42 sample was prepared in aggregation. In this case, the first effect seems to be more rele-
phosphate buffer (pH 8.0) under favorable conditions for the vant for Aβ self-aggregation, whereas in metal-mediated aggre-
self-aggregation studies.42,43 The series of newly synthesized gation the chelating effect has also an important role. In fact,
compounds were screened to assess their inhibitory capacity of although the interaction of Zn2+ with the ligands under study
the Aβ42 self-aggregation. Aβ42 was incubated at 37 °C for 24 h is out of the scope of this paper, the Zn(II)–(3-hydroxy-4-pyridi-
in the presence of the ligands. Aggregation was monitored by none) systems have been previously reported.44 This type of
the ThT method, as described above. Table 2 reports the chelator demonstrated reasonably good affinity for the hard–
values found for the fibrillogenesis inhibition by the tested soft Zn2+ ion (for DFP, pZn = 6.3),44 which should be strong
compounds at 100 μM concentration.
enough to compete with the Aβ peptide for the metal ion, but
Analysis of the results depicted in Table 2 shows that all the without its depletion from essential metalloproteins.
compounds, except 1b, are able to inhibit Aβ42 self-aggregation
above 40% (in a concentration ratio of 1 : 5 [Aβ : I]). This set of
Cell viability
compounds evidenced a similar order of magnitude for the
Efficiency of the BTA–HP based compounds in preventing
percentage of aggregation inhibition, although 1c and 2d Aβ42-induced toxicity. To evaluate the potential therapeutic
appeared as the most potent inhibitors. This can be explained action of the BTA–HP based compounds, SH-SY5Y cells were
by their conformation which allows the compounds to interact treated with Aβ42 peptides, which constitutes a good in vitro
with the β-sheet of amyloid peptides without structural hin- cellular model for studying neurodegeneration associated with
drances, therefore inhibiting their aggregation. Since all these AD. For that, 1b–1d, 2a and 2d were tested as representative
compounds include in their structure a BTA motif, which is compounds. The cell proliferation assay makes use of a quanti-
present in the ThT structure (Fig. 1), the mechanism involved tative colorimetric method with 3-(4,5-dimethylthiazol-2-yl)-
in the inhibition of the Aβ aggregation may be rationalized by 2,5-diphenyltetrazolium bromide (MTT), and accounts for the
the strong affinity of BTA for the amyloid fibrils, and they may viability of the SH-SY5Y cells, with and without treatment (see
function either as β-sheet disruptor or β-sheet formation pre- the Experimental part).
venting agents.
Fig. 6 shows that there was a significant decrease in cell via-
bility after a 24 h treatment with 1 μM Aβ42 peptides (about
15% with respect to the control). At the concentrations used in
Zinc-induced Aβ1–42 aggregation
Due to the reported correlation between metal ions and the these tests (2.5–7 μM of the compound), only 2d changed the
amyloid plaques, it was decided to further complement the cell viability at the same end point (decreasing it about
study on activity of present compounds as Aβ anti-aggregators, 10–15%), while all the other compounds did not make appreci-
but in the presence of Zn(II). The option for this metal ion cir- able difference. Regarding the effect on the cells treated with
cumvents the fluorescence quenching effects anticipated for Aβ42 peptides, only compound 1c improved the cell viability,
paramagnetic ions Cu(II) and Fe(III). Therefore, a selection of restoring it to about 100% with respect to the control without
the compounds with the best inhibitory capacity against the Aβ42.
self-mediated Aβ aggregation (1c and 2d) was evaluated for
An analogous assay on cell viability was also carried out in
their ability to inhibit the Zn(II)-induced Aβ aggregation. The the presence of iron. Since oxidative stress is also a key feature
compounds were incubated for 24 h, at 37 °C, with Aβ peptide of AD, the efficacy of compounds 1c and 2d was tested against
and a solution of Zn2+. After incubation, a ThT solution (5 μM) such an effect mediated by iron addition. Thus, ascorbate–iron
was added and fluorescence measurements were carried out. was added, assuming that it could give rise to the reactive
As expected, the results showed that the Zn(II) could accelerate hydroxyl radical by the iron-catalyzed Haber–Weiss reaction.
the Aβ aggregation (see Fig. S4†). Since in this case the aggre- However, the results of the iron effect on the cell viability were
gation inhibition was observed for lower concentrations of the meaningless (Fig. S5†). In fact, although the addition of the
ligand (40 μM) than for the self-mediated aggregation iron–ascorbate pair could induce the iron reduction, the
(100 μM), those lower concentration conditions of the ligand absence of any strong oxidant disabled the iron reoxidation
were used. The percentage of inhibition of zinc-mediated and provoked the induced formation of the redox active
aggregation were 27.1 (1c) and 31.8 (2d), with 40 μM ligand species and oxidative stress in the cells.45 On the other hand,
concentration. Although a direct comparison between these the effect of the ligands (1c and 2d) on cell viability kept the
values and the corresponding ones obtained in the absence of same behavior independently of the presence of iron. In fact,
the zinc (68.9 for 1c; 59.3 for 2d, with 100 μM ligand in spite of the demonstrated high affinity of the chelator 2d for
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6065

View Article Online
Paper
Dalton Transactions
residues within the AChE active site and the phenylic and
cyclic groups of the compounds, as indicated by the modeling
studies. This selection of hybrid compounds evidenced also
good inhibitory capacity towards Aβ42 self-aggregation, mostly
above 40%. Among the most active compounds, a chelator 2d
(59%, 100 μM) and a “pro-chelator” 1c (68%, 100 μM), the
capacity of inhibition of the Aβ42 zinc-induced-aggregation
revealed to be superior for the chelator 2d. Overall, the anti-
Aβ42 aggregation activity of these compounds seems to result
from the contribution of two main effects, namely the BTA
interactions with the peptide and the metal chelation ability.
Finally a cellular study on the effect of the BTA–HP based com-
pounds in SH-SY5Y cells treated with Aβ42, as a model of AD
neurodegeneration, showed that compound 1c presented the
best efficiency in preventing Aβ42-induced toxicity.
Fig. 6 Effect of BTA–HP compounds on Aβ42 peptides toxicity. SH-SY5Y cells
were treated with Aβ42 (1 μM) peptides, for 24 h, in the absence or the presence
of BTA–HP compounds. Evaluation of cell viability was performed by using the
MTT reduction test. Results are expressed as the percentage of SH-SY5Y
In brief, these preliminary findings show that this set of
hybrid hydroxypyridinone–benzothiazole chelators and deriva-
tives can merge important properties, such as iron-chelation,
anti-oxidant activity, AChE inhibition and interaction with Aβ,
and some of them may be further studied in view of AD drug
development.
untreated cells, with the mean
SEM derived from 3 different experiments.
***p < 0.001; *p < 0.05, significantly different when compared with SH-SY5Y
untreated cells; #p < 0.05; significantly different when compared with Aβ42
treated SH-SY5Y cells; &p < 0.05, significantly different when compared with the
respective BTA–HP treated SH-SY5Y cells.
Experimental part
General methods
iron(III), it possesses an extremely weak affinity for iron(II), as
reported for DFP.46
Hence, based on the data obtained in the studied cellular
context, compound 1c seems to be the most promising one to
be used in further studies.
Analytical grade reagents were purchased from Sigma, Aldrich,
Fluka, Acros, and Merck, and were used as supplied. DFP and
3-hydroxy-2-methyl-4-pyrone (maltol) were from Aldrich. An
amyloid-β peptide fragment 1–42 (Aβ42) was obtained from
American Peptide Company Inc (USA). Solvents were dried
according to standard methods.47 The chemical reactions were
monitored by TLC using aluminum plates coated with silica
gel 60 F254 (Merck). Column flash chromatography sepa-
rations were performed using silica gel Merck 230–400 mesh
ASTM. The 1H and 13C NMR spectra were recorded on a Bruker
ADVANCE III 300 or 400 MHz spectrometer at room tempera-
ture. Chemical shifts (δ) are reported in ppm from the internal
reference TMS (tetramethylsilane), for organic solvents, and
DSS (3-trimethylsilyl-propionic acid-d4 sodium salt) for D2O.
The following abbreviations are used: s = singlet, d = doublet,
t = triplet, m = multiplet.
Melting points were measured with a Leica Galen III hot
stage apparatus and are uncorrected. The electrospray ioni-
zation mass spectra (ESI MS) were obtained 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. Microanalyses were performed using a
Fisons EA1108 CHNF/O instrument.
Conclusions
A set of multifunctional compounds has been developed and
studied as potential drug candidates for the multifaceted Alz-
heimer’s disease (AD). The main feature of these ligands is
based on their common framework which conjugates
3-hydroxy-4-pyridinone (HP) and benzothiazole (BTA) moieties,
which are well known for their high affinity for iron(III), and
beta-amyloid (Aβ), respectively. Furthermore, due to the multi-
factorial nature of AD, the linker between those two basic mol-
ecular units was designed to provide the whole molecule with
capability for the inhibition of acetylcholinesterase (AChE).
Following this strategy, a set of four hybrid (3-hydroxy-4-pyridi-
none)–benzothiazole (HP–BTA, 2a–2d) chelators and their HP-
O-benzyl analogues (1a–1d) were prepared and evaluated for
their physico-chemical and biological properties. For two
representative compounds (2a and 2d), the iron-complexation
study confirmed their high ability to form tris-chelated com-
plexes (FeL3) (pFe = 19.2–19.4). The chelators displayed also
significant anti-oxidant properties (in opposition to O-pro-
Molecular modelling
tected analogues), and the best activity was found for 2d. In a first approach to our drug design, a docking study was
Regarding the biological activity, reasonable AChE inhibitory performed in order to design new potential AChE inhibitors
activities (IC50 = 14–19 μM) were obtained for a set of com- based on previously known biologically-active molecular frag-
pounds (1a, 1b, 1c and 2d), which seems to be mainly rational- ments. For that, the X-ray structures of AChE were taken from
ized in terms of interactions between lipophilic amino acid the RCSB Protein Data Bank (PDB).31 In this case, the crystal
6066 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
structure with PDB code 1EVE, corresponding to the Torpedo 9.6 Hz, CH2-COOH), 2.16 (3H, s, CH3), 2.01 (2H, m,
californica variant (TcAChE) complexed with donepezil (Dnp), CH2CH2CH2). m/z (ESI MS): 302.1 (M + H)+.
was chosen.
2-(4-(3-(Benzyloxy)-2-methyl-4-oxopyridin-1(4H)-yl)phenyl)-
Solvent and auxiliary co-crystallization molecules were acetic acid (4b). 3 (2 g, 9.4 mmol) and 4-aminophenylacetic
removed from the original structure, and hydrogen atoms were acid (1.42 g, 9.4 mmol) were dissolved in a solution of aqueous
added by means of Maestro 7.5.48 The original ligand was 0.1 M HCl (50 mL) in ethanol (100 mL) and the mixture was
extracted from the complex, and the two structures were saved left under reflux overnight. After cooling, the reaction mixture
separately. The ligands were drawn using Maestro, and then was evaporated. The oil obtained was taken into a 5% NaOH
they were submitted to a conformational search (CS) of 1000 solution and washed with dichloromethane. The aqueous
cycles, using a water environment model (generalized-Born/ phase was acidified to pH ∼ 3 with 1 M HCl solution and
surface-area model) by means of Macromodel.49 The algorithm extracted with ethyl acetate. The organic phase was dried over
used was based on the Monte Carlo method, with the molecu- anhydrous sodium sulfate and evaporated till dryness. The
lar Merck force field (MMFFs) and a distance-dependent pure product was obtained as a pale solid (23% yield). 1H
dielectric constant of 1.0. The lowest-energy conformation of NMR (D2O), δ (ppm): 7.78 (1H, d, J = 10.0 Hz, 6-HPy),
each ligand was saved, and it was docked into the previously 7.46–7.44 (5H, 2d and 3t, Ph), 7.42 (2H, d, J = 11.2 Hz,
saved AChE structure, using the GOLD program, version 4.0.30 N-CHCH), 7.24 (2H, d, J = 11.2 Hz, N-CHCH), 6.67 (1H, d, J =
The region of interest was defined in order to contain the resi- 9.6 Hz, 5-HPy), 5.12 (2H, s, OCH2-Ph), 3.62 (2H, s, CH2-COOH),
dues within 15 Å from the position of the original ligand 1.81 (3H, s, CH3). m/z (ESI MS): 350.1 (M + H)+.
(Dnp). The Gold default parameters were used, and the
General procedure for the synthesis of benzothiazole
derivatives 5a–5c
ligands were submitted to 100 genetic algorithm runs. The
ASP scoring function was used in this case, since it has pre-
viously revealed, among the ones supplied by Gold (GoldScore, 2-Aminothiophenol (1 mmol) and an equivalent amount of
ChemScore, and ASP), to give the best prediction with AChE amino-carboxylic acid (e.g. 4-aminobenzoic acid, L-proline or
inhibitors.50 For each ligand, the docking solution taken into 4-aminomethylbenzoic acid) were dissolved in polyphosphoric
account was the conformation with the best score, according acid (PPA) and the reaction mixture was left under reflux (T ∼
to the fitness function used, ASP.
220 °C) for 4 h. After cooling, the reaction mixture was dis-
solved in 5% NaOH aqueous solution and extracted with
dichloromethane. The organic phase was then extracted with a
Synthesis of the compounds
Benzyloxy-2-methyl-4-pyrone (3). To a solution of 3-hydroxy- 0.1 N HCl aqueous solution and washed with dichloro-
2-methyl-4-pyrone (10 g, 79 mmol) in methanol (100 mL) with methane. The aqueous phase was basified to pH ∼ 10–12 with
the same amount of 7 M aqueous NaOH (12 mL, 79 mmol) a 5 M NaOH aqueous solution and extracted with dichloro-
was added benzyl chloride (10.1 mL, 88 mmol). The mixture methane. The organic phase was dried over anhydrous sodium
was left under reflux overnight.
sulfate and the solvent was evaporated.
After cooling, the mixture was filtered and evaporated to
4-(Benzo[d]thiazol-2-ylmethyl)aniline (5a). 2-Aminothiophe-
dryness. The resulting residual oil was dissolved in dichloro- nol (1.54 mL, 14.3 mmol) and 4-aminobenzoic acid (1.96 g,
methane and washed with a 5% NaOH aqueous solution and 14.3 mmol) give the pure hydrochloric salt as a brown solid
water. The organic phase was dried with anhydrous sodium (53% yield) after acidification with HCl-saturated methanol
sulfate and the solvent was evaporated to dryness to obtain the until pH ∼ 1 and recrystallization from acetonitrile.
1
pure product as a pale yellow oil (74% yield). H NMR (CDCl3),
1H-NMR (D2O), δ (ppm): 7.98 (1H, d, J = 9 Hz, NC-CH), 7.93
δ (ppm): 7.59 (1H, d, J = 5.7 Hz, 6-HPy), 7.39–7.33 (5H, 1d and (1H, d, J = 9 Hz, SC-CH), 7.56 (2H, t, J = 9 Hz, NC-CHCH and
2t, Ph), 6.36 (1H, d, J = 5.4 Hz, 5-HPy), 5.16 (2H, s, CH2), 2.09 SC-CHCH), 7.50 (2H, d, J = 9 Hz, CH-CHNH2), 7.42 (2H, d, J =
(3H, s, CH3). m/z (ESI MS): 217.1 (M + H)+.
9 Hz, CHCH-NH2), 4.57 (2H, s, CH2). m/z (ESI MS): 241.1
3-Benzyloxy-1-(3′-carboxylpropyl)-2-methyl-4-pyridinone (4a). (M + H)+.
3 (8.27 g, 38 mmol) and amino butyric acid (3.91 g, 9.4 mmol)
(4-(Benzo[d]thiazol-2-yl)phenyl)methanamine (5b). 4-Amino-
were dissolved in methanol (130 mL). The pH was adjusted to methyl benzoic acid (1.02 g, 6 mmol) and 2-aminothiophenol
13 with an aqueous solution of 10% NaOH and the mixture (0.7 mL, 6 mmol) give the pure product as a pale greenish-blue
was left under reflux for 5 h. After cooling, the reaction solid (53% yield). 1H-NMR (CDCl3), δ (ppm): 8.08–8.05 (3H, m,
mixture was evaporated. The oil obtained was dissolved in NC-CH and Ph), 7.90 (1H, d, J = 7.6 Hz, SC-CH), 7.52–7.36 (4H,
ethyl acetate and extracted with water. The aqueous phase was m, NC-CHCH, SC-CHCH and Ph), 3.96 (2H, s, CH2). m/z (ESI
acidified till pH ∼ 4 with a 2 M HCl solution and extracted MS): 241.1 (M + H)+.
with dichloromethane. The organic phase was dried over anhy-
2-(Pyrrolidin-2-yl)benzo[d]thiazole (5c). 2-Aminothiophenol
drous sodium sulfate and evaporated till dryness. The pure (1.54 mL, 14.3 mmol) and L-proline (2.92 g, 14.3 mmol) afford
product was obtained as a beige solid (48% yield). 1H NMR the pure hydrochloric salt as an orange solid (71% yield) after
(CDCl3), δ (ppm): 7.52 (1H, d, J = 5.7 Hz, 6-HPy), 7.38–7.30 acidification with HCl-saturated methanol until pH ∼ 1 and
(5H, 2d and 3t, Ph), 6.58 (1H, d, J = 5.7 Hz, 5-HPy), 5.15 (2H, s, recrystallization from acetonitrile. 1H-NMR (D2O), δ (ppm):
OCH2Ph), 3.95 (2H, t, J = 10.2 Hz, CH2N), 2.37 (2H, t, J = 8.09 (2H, d, J = 9 Hz, NC-CH and SC-CH), 7.65–7.57 (2H, 2t,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6067

View Article Online
Paper
Dalton Transactions
NC-CHCH and SC-CHCH), 5.28 (1H, t, J = 7.5 Hz, C-CHNH), C35H29N3O3S·1.5H2O: C 70.29, H 5.38, N 7.03, S 5.36%; found:
3.62 (2H, m, NHCH2), 2.73 (1H, m, C-CHCH2), 2.42 (1H, m, C 70.34, H 5.55, N 6.67, S 5.38%.
C-CHCH2), 2.26 (2H, m, CHNH-CH2). m/z (ESI MS): 205.1
N-(4-(Benzo[d]thiazol-2-yl)benzyl)-2-(4-(3-(benzyloxy)-2-methyl-
4-oxopyridin-1(4H)-yl)phenyl)acetamide (1c). 4b (51 mg,
0.146 mmol) and 5b (33 mg, 0.138 mmol) afford the pure
product as a pale green solid (70% yield) with recrystallization
(M + H)+.
General procedure for the synthesis of compounds 1a–1d
5a, 5b or 5c (1 mmol), 4a or 4b (2 mmol), 1-ethyl-3-(3-dimethyl- from ethyl ether. M.p. 93–95 °C.
aminopropyl) carbodiimide hydrochloride (EDC·HCl, 4 mmol)
1H NMR (MeOD), δ (ppm): 8.03 (4H, m, Ph), 7.67 (1H, d, J =
and a catalytic amount of 4-dimethylaminopyridine (DMAP, 7.4 Hz, 6-HPy), 7.53 (3H, m, SC-CH, NC-CH and NHC-CH),
5% mol) were dissolved in dichloromethane and the mixture 7.45 (5H, m, Ph) 7.35 (3H, m, NC-CHCH, SC-CHCH and Ph),
was left stirring at room temperature overnight. The reaction 7.29 (2H, m, Ph), 6.55 (1H, d, J = 7.4 Hz, 5-HPy), 5.15 (2H, s,
mixture was washed with 5% NaOH solution, 1 M HCl solution OCH2Ph), 4.49 (2H, s, CH2-NH), 3.70 (2H, s, CH2CONH), 1.88
and water. The organic phase was dried over anhydrous (3H, s, CH3). 13C NMR (MeOD), δ (ppm): 210.0 (CvO), 201.0
sodium sulfate and evaporated to dryness.
(CvO), 194.4 (SCvN), 191.4 (C-O), 175.1 (C-CH3), 173.2 (C
N-(4-(Benzo[d]thiazol-2-ylmethyl)phenyl)-4-(3-(benzyloxy)- arom-HP), 143.7 (C arom-BTA), 141.8 (C arom-BTA), 141.6
2-methyl-4-oxopyridin-1(4H)-yl)butanamide (1a). 5a (50 mg, (6-CPy), 139.3 (C arom-BTA), 138.4 (2 × C arom-BTA), 131.9 (2 ×
0.181 mmol) and 4a (61 mg, 0.203 mmol) afford the pure C arom-HP), 130.3 (2 × C arom-HP), 129.4 (2 × C arom-HP),
product as a beige solid (69% yield) with recrystallization from 129.3 (C arom-HP), 128.7 (2 × C arom-HP), 127.9 (C arom-BTA),
ethyl acetate. M.p. 73–75 °C.
127.8 (C arom-HP), 126.7 (C arom-BTA), 123.8 (C arom-BTA),
1H NMR (CDCl3), δ (ppm): 8.40 (1H, d, J = 7.5 Hz, 6-HPy), 123.0 (2 × C arom-BTA), 116.9 (5-Cpy), 74.70 (OC-Ph), 30.67
7.98 (1H, d, J = 8.3 Hz, NC-CH), 7.79 (1H, d, J = 7.8 Hz, SC-CH), (CH2), 24.23 (CH2), 14.92 (CH3). m/z (ESI MS): 572.1 (M + H)+.
7.55 (2H, d, J = 7.6 Hz, NH-CH), 7.45 (2H, t, J = 7.8 Hz, Analysis calc. for C35H29N3O3S·0.3Et2O: C 70.08, H 5.22,
NCCH-CH and SCCH-CH), 7.36–7.32 (5H, 2d and 3t, Ph), 7.21 N 6.76, S 5.15%; found: C 70.15, H 4.85, N 6.45, S 4.88%.
(2H, d, J = 7.6 Hz, NHCH-CH), 6.33 (1H, d, J = 7.5 Hz, 5-HPy),
1-(4-(2-(Benzo[d]thiazol-2-yl)pyrrolidin-1-yl)-4-oxobutyl)-
5.17 (2H, s, OCH2Ph), 4.40 (2H, s, CH2-CN), 3.91 (2H, t, J = 3-(benzyloxy)-2-methylpyridin-4(1H)-one (1d). 4a (1 g,
6.9 Hz, N-CH2CH2), 2.35 (2H, t, J = 6.7 Hz, CO-CH2), 2.14 (3H, 3.32 mmol) and 5c (80 mg, 3.32 mmol) afford the pure
s, CH3), 2.02 (2H, m, CH2CH2CH2). 13C NMR (CDCl3), δ (ppm): product as a hygroscopic orange solid (80% yield) with recrys-
1
173.2 (CvO), 171.6 (CvO), 170.3 (SCvN), 153.3 (C arom-BTA), tallization from ethyl acetate. H NMR (MeOD), δ (ppm): 7.93
146.3 (C-O), 142.0 (6-CPy), 139.1 (C arom-HP), 138.2 (C arom- (2H, m, NC-CH and SC-CH), 7.69 (1H, d, J = 7.4 Hz, 6-HPy),
BTA), 137.3 (C arom-BTA), 135.7 (C arom-BTA), 132.5 (C arom- 7.54–7.31 (7H, m, NC-CHCH, SC-CHCH and Ph), 6.48 (1H, d,
HP), 129.7 (2 × C arom-BTA), 128.9 (2 × C arom-HP), 128.5 (2 × J = 7.4 Hz, 5-HPy), 5.47 (1H, m, N-CH-CvN), 5.07 (2H, s,
C arom-HP), 128.3 (C arom-BTA), 126.1 (C arom-BTA), 124.9 OCH2Ph), 4.02 (2H, t, J = 7.7 Hz, N-CH2CH2CH2), 3.68 (2H, m,
(C-CH3), 122.8 (5-Cpy), 121.7 (C arom-BTA), 120.4 (2 × C arom- N-CH2CH2), 2.50 (2H, t, J = 7.7 Hz, CH2-CO), 2.41 (1H, m,
HP), 116.9 (C arom-BTA), 73.20 (OCH2-Ph), 53.15 (CH2), 40.15 NCH-CH2), 2.23 (1H, m, NCH-CH2), 2.18 (3H, s, CH3), 1.96
(CH2), 32.38 (CH2), 25.63 (CH2), 12.56 (CH3). m/z (ESI MS): (2H, m, N-CH2CH2). 13C NMR (MeOD), δ (ppm): 176.3 (CvO),
524.3 (M + H)+. Analysis calc. for C31H29N3O3S·H2O: C 68.91, H 174.8 (CvO), 173.1 (SCvN), 154.2 (C-N), 145.2 (C-O), 141.1
5.76, N 7.78, S 5.93%; found: C 68.90, H 5.63, N 7.72, S 5.97%.
(6-CPy), 138.5 (C arom-HP), 135.8 (C-S), 130.3 (2 × C arom-HP),
N-(4-(Benzo[d]thiazol-2-ylmethyl)phenyl)-2-(4-(3-(benzyloxy)- 129.4 (2 × C arom-HP), 129.3 (C arom-HP), 127.4 (C arom-BTA),
2-methyl-4-oxopyridin-1(4H)-yl)phenyl)acetamide (1b). 5a 126.4 (C arom-BTA), 123.4 (C arom-BTA), 123.3 (C arom-BTA),
(25 mg, 0.1 mmol) and 4b (17.7 mg, 0.1 mmol) give the pure 123.0 (C-CH3), 117.4 (5-Cpy), 74.48 (OC-Ph), 60.60 (CH-N),
product as a pale yellow solid (73% yield) with recrystallization 54.50 (CH2), 47.89 (CH2), 33.48 (CH2), 31.34 (CH2), 26.36
1
from ethyl acetate. M.p. 89–91 °C. H NMR (MeOD), δ (ppm): (CH2), 25.25 (CH2), 12.80 (CH3). m/z (ESI MS): 488.3 (M + H)+.
7.94 (2H, t, J = 9.3 Hz, NC-CHCH and SC-CHCH), 7.71 (1H, d, Analysis calc. for C28H29N3O3S·0.1HCl·0.8H2O: C 66.66, H 6.13,
J = 7.5 Hz, 6-HPy), 7.59 (4H, m, Ph), 7.52 (2H, m, Ph), N 8.33, S 6.36%; found: C 66.68, H 6.13, N 8.48, S 5.96%.
7.47–7.31 (9H, m, NHC-CH, NC-CHCH, NC-CH, Ph), 6.59 (1H,
General procedure for the synthesis of 6a and 6b
d, J = 7.5 Hz, 5-HPy), 5.17 (2H, s, OCH2Ph), 4.45 (2H, s,
CH2CvN), 3.81 (2H, s, CH2CONH), 1.92 (3H, s, CH3). 13C NMR In a hydrogenation flask, 4a or 4b (1 mmol) was dissolved in
(MeOD), δ (ppm): 175.2 (CvO), 173.9 (CvO), 171.3 (SCvN), methanol and 10% Pd/C was added (10% w/w). The reaction
146.6 (C-O), 145.4 (C-CH3), 141.8 (C arom-HP), 141.3 (6-CPy), suspension was left for 2 h under H2 atmosphere at 2 bar
139.1 (C arom-BTA), 138.5 (C arom-BTA), 134.5 (C arom-BTA), pressure. The reaction mixture was filtered and the solvent was
132.0 (2 × C arom-BTA), 130.7 (2 × C arom-HP), 130.3 (2 × C evaporated to dryness, affording the pure product.
arom-HP), 129.4 (2 × C arom-HP), 129.3 (C arom-HP), 127.9
4-(3-Hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)butanoic acid
(2 × C arom-HP), 127.4 (C arom-BTA), 126.3 (C arom-HP), 123.2 (6a). 4a (200 mg, 0.664 mmol) afforded the pure product as a
1
(C arom-BTA), 122.9 (C arom-BTA), 121.6 (2 × C arom-BTA), pale brown solid (95% yield). H NMR (MeOD), δ (ppm): 7.62
117.0 (5-Cpy), 74.64 (OC-Ph), 43.95 (CH2), 40.31 (CH2), 14.89 (1H, d, J = 7.2 Hz, 6-HPy), 6.41 (1H, d, J = 7.2 Hz, 5-HPy), 4.10
(CH3). m/z (ESI MS): 594.0 (M + Na)+. Analysis calc. for (2H, t, J = 7.65 Hz, CH2N), 2.47 (3H, s, CH3), 2.40 (2H, t, J =
6068 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
6.9 Hz, CH2-COOH), 2.02 (2H, m, CH2CH2CH2). m/z (ESI MS): 7.14 (2H, d, J = 5.7 Hz, NC-CH), 6.45 (1H, d, J = 7.2 Hz, 5-HPy),
212.1 (M + H)+.
4.40 (2H, s, CH2CONH), 3.80 (2H, s, CH2CN), 2.11 (3H, s, CH3).
2-(4-(3-Hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)phenyl)acetic 13C NMR (MeOD), δ (ppm): 171.7 (CvO), 171.5 (CvO), 146.9
acid (6b). 4b (200 mg, 0.573 mmol) afforded the pure product (SCvN), 142.1 (C-O), 139.5 (6-CPy), 139.2 (C-N), 139.0 (C-S),
1
as a light brown solid (93% yield). H NMR (MeOD), δ (ppm): 137.9 (C arom-BTA), 136.7 (C arom-BTA), 134.5 (C-CH3), 133.1
7.59 (1H, d, J = 7.2 Hz, 6-HPy), 7.53 (2H, d, J = 8.1 Hz, Ph), 7.35 (2 × C arom-HP), 132.1 (2 × C arom-BTA), 130.8 (2 × C arom-
(2H, d, J = 8.1 Hz, Ph), 6.48 (1H, d, J = 7.2 Hz, 5-HPy), 3.73 (2H, HP), 128.1 (2 × C arom-BTA), 127.5 (C arom-BTA), 126.4
s, CH2), 2.13 (3H, s, CH3). m/z (ESI MS): 259.2 (M + H)+.
(C arom-BTA), 123.2 (C arom-BTA), 122.9 (C arom-BTA), 121.7
(2 × C arom-HP), 112.7 (5-Cpy), 44.05 (CH2), 40.39 (CH2), 13.71
(CH3). m/z (ESI MS): 482.3 (M + H)+. Analysis calc. for
General procedure for the synthesis of compounds 2a–2d
A solution of 6a or 6b (1 mmol) in dry dichloromethane under C28H23N3O3S·0.2HCl: C 65.63, H 4.67, N 8.13, S 6.21%; found:
a nitrogen atmosphere was cooled over an ice–salt–water bath. C 65.68, H 4.52, N 7.46, S 6.68%.
Ethylchloroformate (ECF, 2.1 mmol) and N-methylmorpholine
N-(4-(Benzo[d]thiazol-2-yl)benzyl)-2-(4-(3-hydroxy-2-methyl-4-
(NMM, 4 mmol) were added and the reaction mixture was left oxopyridin-1(4H)-yl)phenyl)acetamide (2c). From 6b (103 mg,
reacting for 40 min. Then, 5a, 5b or 5c (1.1 mmol) was added 0.397 mmol) and 5b (105 mg, 0.437 mmol) the general
and the reaction mixture was left at room temperature for 3 h. method afforded the pure product as a light yellow solid (28%
The reaction mixture was filtered and the solvent was evapo- yield). M.p. 96–98 °C.
rated. The oil obtained was dissolved in ethyl acetate and
1H NMR (CDCl3), δ (ppm): 8.06 (1H, d, J = 7.5 Hz, 6-HPy),
washed with water and a solution of 0.05 M HCl. The organic 7.92 (1H, d, J = 7.8 Hz, NC-CH), 7.67 (1H, d, J = 7.8 Hz, SC-CH),
phase was dried under anhydrous sodium sulfate and evapo- 7.43 (2H, d, J = 7.2 Hz, NHC-CH), 7.40 (2H, t, J = 9.3 Hz,
rated to dryness.
NC-CHCH and SC-CHCH), 7.37 (2H, d, J = 7.2 Hz,
The solid was taken in an aqueous solution of 2% NHC-CHCH), 7.34 (1H, d, J = 7.2 Hz, NC-CHCH), 7.16 (2H, d,
NaHCO3–EtOH (1 : 1) and left at ∼65 °C for 4 h. The reaction J = 9.0 Hz, NC-CH), 6.56 (1H, d, J = 7.5 Hz, 5-HPy), 4.54 (2H, d,
mixture was evaporated and the oil obtained was dissolved in J = 5.7 Hz, CH2-NH), 3.70 (2H, s, CH2CONH), 1.89 (3H, s, CH3).
ethyl acetate (for all compounds, except for 2d that is dissolved 13C NMR (MeOD), δ (ppm): 184.7 (SCvN), 143.7 (C-O), 139.4
in dichloromethane) and washed with water. The organic (C-CH3), 134.6 (6-CPy), 131.8 (2 × C arom-BTA), 129.4 (C arom-
phase was dried under anhydrous sodium sulfate and evapo- BTA), 128.7 (2 × C arom-HP), 128.0 (2 × C arom-BTA), 127.8
rated till dryness. The solid obtained was recrystallized with (C arom-BTA), 126.7 (C arom-BTA), 123.8 (C arom-HP), 122.9
diethyl ether to afford the pure product.
(2 × C arom-HP), 112.6 (5-Cpy), 44.03 (CH2), 43.30 (CH2), 13.81
N-(4-(Benzo[d]thiazol-2-ylmethyl)phenyl)-4-(3-hydroxy-2-methyl- (CH3). m/z (ESI MS): 482.0 (M + H)+. Analysis calc. for
4-oxopyridin-1(4H)-yl)butanamide (2a). From 6a (80 mg, C28H23N3O3S·1.4EtOH: C 64.77, H 5.27, N 7.72, S 5.89%;
0.379 mmol) and 5a (81 mg, 0.293 mmol), the general method found: C 64.85, H 4.99, N 7.26, S 5.74%.
afforded the pure product as a beige solid (29% yield). M.p.
1-(4-(2-(Benzo[d]thiazol-2-yl)pyrrolidin-1-yl)-4-oxobutyl)-3-hydroxy-
91–93 °C. 1H NMR (MeOD), δ (ppm): 7.94 (2H, t, J = 9.2 Hz, 2-methylpyridin-4(1H)-one (2d). From 6a (234 mg, 1.11 mmol)
NCCH-CH and SCCH-CH), 7.71 (1H, d, J = 6.8 Hz, 6-HPy), and 5c (293 mg, 1.22 mmol), the general method afforded the
7.57–7.49 (3H, 2d and 1t, J = 8.3 Hz, NC-CH and Ph), 7.42–7.34 pure product as a beige solid (15% yield). M.p. 83–85 °C.
(3H, 1t and 2d, SC-CH and Ph), 6.49 (1H, d, J = 6.8 Hz, 5-HPy),
1H NMR (MeOD), δ (ppm): 7.94 (2H, t, J = 9.4 Hz, NC-CH
4.43 (2H, s, CH2-CN), 4.19 (2H, t, J = 7.4 Hz N-CH2CH2), 2.49 and SC-CH), 7.63 (1H, d, J = 7.2 Hz, 6-HPy), 7.52–7.42 (2H, m,
(2H, m, CO-CH2), 2.16 (2H, t, J = 7.4 Hz, CH2CH2CH2), 2.03 NCCH-CH and SCCH-CH), 6.41 (1H, d, J = 7.2 Hz, 5-HPy), 5.49
(3H, s, CH3). 13C NMR (DMSO), δ (ppm): 171.4 (CvO), 170.1 (1H, m, N-CH-CvN), 4.11 (2H, t, J = 7.6 Hz, N-CH2CH2CH2),
(CvO), 145.2 (SCvN), 138.1 (C-O), 137.6 (6-CPy), 135.0 3.84 (2H, m, N-CH2CH2), 3.67 (2H, m), 2.56 (2H, t, J = 7.6 Hz,
(C-CH3), 135.8 (C-S and C-N), 129.5 (2 × C arom), 126.1 (C CO-CH2), 2.20 (3H, s, CH3), 2.13 (2H, m, NCH-CH2), 1.96 (2H,
arom), 124.9 (C arom), 122.3 (C arom), 122.1 (C arom), 119.5 m, N-CH2CH2). 13C NMR (MeOD), δ (ppm): 176.3 (CvO), 173.2
(2 × C arom), 110.6 (5-Cpy), 56.09 (CH2), 52.53 (CH2), 32.45 (CvO), 170.6 (SCvN), 154.1 (C-N), 147.3 (C-OH), 138.8 (6-
(CH2), 25.86 (CH2), 11.39 (CH3). m/z (ESI MS): 434.1 (M + H)+. CPy), 135.8 (C-S), 132.9 (C arom-BTA), 127.4 (C arom-BTA),
Analysis calc. for C24H23N3O3S·H2O: C 63.98, H 5.57, N 9.33, 126.4 (C arom-BTA), 123.3 (C arom-BTA), 123.0 (C-CH3), 112.7
S 7.12%; found: C 64.02, H 5.56, N 9.15, S 6.91%.
(5-Cpy), 60.58 (CH-N), 54.39 (CH2), 47.89 (CH2), 33.49 (CH2),
N-(4-(Benzo[d]thiazol-2-ylmethyl)phenyl)-2-(4-(3-hydroxy- 31.42 (CH2), 26.54 (CH2), 25.22 (CH2), 11.88 (CH3). m/z (ESI
2-methyl-4-oxopyridin-1(4H)-yl)phenyl)acetamide (2b). From MS): 398.1 (M + H)+. Analysis calc. for C21H23N3O3S·0.2H2O·
6b (60 mg, 0.232 mmol) and 5a (70 mg, 0.255 mmol), the 0.2EtOH: C 62.13, H 5.94, N 10.27, S 7.84%; found: C 62.13,
general method afforded the pure product as a beige solid H 5.94, N 10.37, S 7.72%.
(20% yield). M.p. 75–77 °C. 1H NMR (CDCl3), δ (ppm): 7.97
(1H, d, J = 6.3 Hz, 6-HPy), 7.78 (1H, d, J = 8.0 Hz, NC-CH), 7.67
Spectroscopic studies
Solutions. The aqueous FeCl3 stock solution (1.77 × 10−2 M),
(1H, d, J = 8.0 Hz, SC-CH), 7.51 (2H, d, J = 7.2 Hz, NHC-CH),
7.45 (2H, t, J = 8.0 Hz, NC-CHCH and SC-CHCH), 7.34 (2H, d, prepared in acid medium, was obtained from Merck and stan-
J = 7.2 Hz, NHC-CHCH), 7.23 (1H, d, J = 5.7 Hz, NC-CHCH), dardized by atomic absorption. Its exact HCl content was
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6069

View Article Online
Paper
Dalton Transactions
determined by titration with 0.1 M HCl (Titrisol) for values of 0.925 mL of the sample were mixed in a cell and left to incu-
pH ≥ 2. The titrant solution (0.1 M KOH) was prepared from a bate for 15 min. Subsequently, 75 μL of AChI were added. The
carbonate-free commercial concentrate (Titrisol), in a 50% w/w initial rate of the enzymatic reaction was followed by reading
DMSO–H2O medium, standardized by potentiometric titration the solution absorbance at 405 nm during the first 5 min of
with potassium hydrogen phthalate and discarded when the the reaction. For that, a Perkin-Elmer Lambda 35 UV-Vis spec-
percentage of carbonate (Gran’s method)51 was about 0.5% of trophotometer was used. Samples were prepared diluting the
the total amount of base. An HCl solution (0.099 M) in 50% inhibitors in methanol, in a range of different concentrations.
DMSO–H2O was used for the electrode calibration and for the A control reaction was carried out using no inhibitor, and it
batch titrations at pH ≤ 2.
was considered 100% activity. The percentage of inhibition
Measurements. Spectrophotometric titrations (250–370 nm) (%I) was calculated by eqn (1), in which
of 2a (4.0 × 10−5 M) and 2d (8.0 × 10−5 M) were performed in
50% w/w DMSO–H2O solution at ionic strength (I) 0.1 M KCl
ꢀ
ꢁ
vI
v0
%I ¼ 100 ꢀ
ꢁ 100
ð1Þ
and the working temperature was maintained at 25.0 0.1 °C.
The iron complexation studies were also done by spectrophoto-
metric titration (300–700 nm) at a 1 : 3 Fe–L stoichiometry
(CL = 1–1.5 × 10−4 M, 10% excess of ligand). Solutions of com-
plexes with pH ≥ 2 were prepared with a total volume solution
of 20 mL and for pH ≤ 2 (1.2–2), using a batch titration
(5 points), in which the amounts of acid and KCl to be added
were calculated for the total volume of the solution used
(4 mL). All titrations were done in duplicate or triplicate and
for comparison purposes the protonation and global iron
stability constants in 50% w/w DMSO–H2O were also calcu-
lated for DFP (CL = 3 × 10−4 M) from spectrophotometric data.
Under the experimental conditions used, the value determined
for pKw was 14.11.
vI is the initial rate in the presence of the inhibitor and v0 is
the initial rate of the control reaction.
The inhibition curves were obtained by plotting the relation
percentage of enzymatic inhibition vs. inhibitor concentration
and a calibration curve was drawn from which the linear
regression parameters were obtained.
Anti-oxidant activity
The anti-oxidant activity was measured by the 2,2-diphenyl-1-
picrylhydrazyl radical (DPPH) method, as described by Tepe
et al.53 To a 2.5 mL solution of DPPH, four samples of com-
pound solution with different volumes were added. The
volume required for the 3.5 mL total volume was attained with
methanol. The mixtures were incubated for 30 min at room
temperature. The absorbance was measured at 517 nm against
the corresponding blank. The control solution corresponds to
the sample with only DPPH and methanol. The anti-oxidant
activity was calculated using eqn (2), in which AA
Calculation of equilibrium constants. The stepwise protona-
tion constants of the compounds, Ki = [HiL]/[Hi−1L][H], and
the overall iron-complex stability constants, βFe
=
mHbL1
[FemHhLl]/[Fe]m[H]h[L]l, were calculated by fitting analysis of
the spectroscopic data (PSEQUAD program32) in the presence
and in the absence of Fe3+. The Fe3+ hydrolytic species under
the defined experimental conditions (I = 0.1 M KCl, 50% w/w
DMSO–H2O, T = 25.0 0.1 °C), herein determined by potentio-
metry with the use of the HYPERQUAD program,52 were
Asample ꢀ ADPPH
%AA ¼
ꢁ 100
ð2Þ
ADPPH
is the anti-oxidant activity, Asample is the absorbance of the
sample compound against the blank and ADPPH is the absor-
bance of the DPPH solution (control solution) against the
blank. The compound concentrations provided 50% of anti-
oxidant activity (EC50) by plotting the anti-oxidant activity
against the compound concentration.
included in the equilibrium model (log βFeH = −4.07, log
-1
βFeH = −11.42) and the species distribution curves were
-3
plotted with the HYSS program.52
Acetylcholinesterase inhibition
The determination of the AChE inhibitory capacity of the syn-
thesized compounds was performed by an adaptation of a
method previously described.50 The buffer used for the enzy-
Inhibition of self-mediated Aβ (1–42) aggregation
matic tests was 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfo- An amyloid β-peptide (1–42) (Aβ1–42) was purchased from Bio-
nic acid (HEPES) 50 mM at pH 8. The Ellman reagent, 5,5′- peptide Co. Inc. as a lyophilized powder and stored at −20 °C.
dithiobis-(2-nitrobenzoic acid) (DTNB, 3 mM), was dissolved in Following a reported protocol,42 samples were treated with
a solution of buffer previously prepared (HEPES 50 mM, pH 8, 1,1,1,1,1,1-hexafluoropropan-2-ol (HFIP) to avoid self-aggrega-
NaCl 50 mM, MgCl2 20 mM). The substrate of the enzyme tion and reserved. HFIP pre-treated Aβ42 samples were resolu-
used in the tests was acetylthiocholine iodide (AChI, 15 mM of bilized with a CH3CN–Na2CO3 (300 μM)–NaOH (250 μM)
stock solution). The AChE (0.05 U mL−1) extracted from Electro- (48.3 : 48.3 : 4.3, v/v/v) solvent mixture in order to have a stable
phorus electricus was purchased from Sigma-Aldrich and stock solution. This Aβ42 alkaline solution (500 μM) was
diluted in the buffer solution (HEPES 50 mM, pH 8). A stock diluted in phosphate buffer (0.215 M, pH 8.0) to obtain a
solution of the inhibitor was prepared in methanol.
20 μM solution. Compounds under study were firstly dissolved
374 μL of HEPES (50 mM, pH 8), 476 μL of DTNB, 10, 20, in methanol (1 mg mL−1), due to their hydrophobic nature,
30, 40 or 50 μL of a stock solution of the inhibitor, 25 μL of and then further diluted in phosphate buffer to the final con-
AChE solution and the amount of methanol necessary to have centration of 100 μM.
6070 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
To study the Aβ42 aggregation inhibition, a known spectro- cells the medium was refreshed, and 1 μM of Aβ42 peptide
fluorimetric method, based on the fluorescence emission of was added from a 1 mM stock solution of 276.9 μM, prepared
thioflavin T (ThT), was followed.42,43 Firstly, Aβ42 (10 μL) in sterile water, for 24 h. Low-molecular weight oligomers of
samples and the tested compounds (10 μL) were diluted with Aβ42 were preferentially detected in the fresh preparations
the phosphate buffer to a 20 μM final concentration of (Aβ) used.54
and 100 μM compound concentration, and then they were
The tested compounds (1b–1d, 2a and 2d) were dissolved in
incubated for 24 h at 37 °C, without stirring. As for the DMSO at a concentration of 10 mM and aliquots were stored at
control, a sample of the peptide was incubated under identical −20 °C. These compounds were added to the medium at 5 μM
conditions but without the inhibitor. After incubation, the (1c), 2.5 μM (2a, 2d), or 7 μM (1b, 1d) final concentrations.
samples were added to a black, flat and clear bottom 96-well Before the addition of Aβ42, the cells were pre-incubated for
plate (BD Falcon) with 180 μL of 1.5 μM ThT (Sigma CAS 2390- 2 h with the BTA–HP compounds.
54-7) in 50 mM glycine–NaOH (pH 8.5). Blank samples were
In addition, the pair ascorbate (0.8 mM)–iron (2.5 μM) was
prepared for each concentration in a similar way, devoid of added to the cells, alone or with Aβ42 peptide, in the presence
peptides. After 5 min incubation with the dye, the fluorescence or absence of compounds 1c and 2d.
of ThT was measured using a Spectramax Gemini EM (Molecu-
The final concentration of DMSO in culture media did not
lar Devices) at the wavelength 490 nm (emission) with exci- exceed 0.05% (v/v) and no alterations in cells were observed.
tation at 446 nm. The percentage inhibition of the self- For all conditions tested, control experiments were performed
induced aggregation due to the presence of the test compound in which the compounds tested, Aβ42 or ascorbate/iron were
was calculated using eqn (3), where IFi and IF0 correspond to not added.
the fluorescence intensities in the presence and absence
MTT cell proliferation assay
ꢀ
ꢁ
IFi
IF0
%I ¼ 100 ꢀ
ꢁ 100
ð3Þ
Cell reduction ability as a surrogate of cell viability was
measured by using a quantitative colorimetric assay with
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
(MTT) according to the Mosmann method.55 In viable cells,
cellular dehydrogenases metabolize 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−1) was
added to each well. The plate was then incubated at 37 °C for
2 h. At the end of the incubation period the formazan precipi-
tates were solubilized with 0.5 mL of acidic isopropanol
(0.04 M HCl–isopropanol). The absorbance was measured at
570 nm using a Spectramax Plus 384 spectrophotometer (Mol-
ecular Devices). Cell reduction ability was expressed relatively
to control values obtained for untreated cells.
of the test compound, respectively, minus the fluorescence
intensities due to the respective blanks. The reported values
were obtained as the mean SEM from two different exper-
iments, each one made in duplicate.
Inhibition of metal-induced Aβ1–42 aggregation
The preparation of the Aβ1–42 solutions was similar to that
described above for the self-induced method. To study the
effects of 1c and 2d on metal-induced Aβ1–42 aggregation a
reported protocol was followed.22 Solutions of Zn2+ were pre-
pared from dilution of a 0.016 M ZnCl2 standard solution to
final concentration of 200 μM, using phosphate buffer (0.2 M
at pH 8.0). Solutions of the compounds 1c and 2d were pre-
pared in methanol for storage and diluted with phosphate
buffer before use.
Abbreviations
Aβ1–42 (20 μM) was incubated with Zn2+ (20 μM) in phos-
phate buffer with or without the ligand (40 μM). The incu-
bation was performed at 37 ° C for 24 h. After the incubation,
the samples were diluted with 170 μL of glycine–NaOH buffer
(50 mM, pH 8.5) containing thioflavin-T (5 μM). Fluorescence
was measured at 450 nm (λexc) and 485 (λem) using a Spectra-
max Gemini EM (Molecular Devices).
AD
Aβ
ACh
AChE
AChI
ROS
Alzheimer’s disease
Beta-amyloid
Acetylcholine
Acetylcholinesterase
Acetylcholine iodide
Reactive oxygen species
3,4-HP 3-Hydroxy-4-pyridinones
HP
BTA
Hydroxypyridinone
Benzothiazole
Cell line and treatment
Human neuroblastoma SH-SY5Y cell lines (ATCCCRL-2266) DFP
were grown in 75 cm2 tissue flasks in DMEM and Ham’s CAS
F12 medium with 10% supplemental non-dialyzed fetal bovine PAS
serum and 100 IU mL−1 penicillin and 50 μg mL−1 streptomy- Dnp
Deferiprone
Catalytic active site
Peripheral active site
Donepezil
cin. Cells were maintained at 37 °C in a humidified incubator DPPH 2,2-Diphenyl-1-picrylhydrazyl radical
under an atmosphere of 95% air and 5% CO2.
Aβ42
Amyloid-β peptide fragment 1–42
(3-(4,5-Dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide)
For the MTT assay, the cells were plated in 24-well plates MTT
at a density of 0.1 × 106 cells per well. 24 h after seeding the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6071

View Article Online
Paper
Dalton Transactions
24 A. Dedeoglu, K. Cormier, S. Payton, K. A. Tseitlin,
J. N. Kremsky, L. Lai, X. Li, R. D. Moir, R. E. Tanzi,
A. I. Bush, N. W. Kowall, J. T. Rogers and X. Huang, Exp.
Gerontol., 2004, 39, 1641.
Acknowledgements
The authors thank the Portuguese NMR Network (IST-UTL
Center) for providing access to the NMR facilities as well as the
Portuguese Fundação para a Ciência e Tecnologia (FCT) for 25 L. E. Scott, M. Telpoukhovskaia, C. Rodríguez-Rodríguez,
financial support with the project PEst-OE/QUI/UI0100/2011
and the post-Doc grant SFRH/BPD/29874/2006 (S.M. Marques).
We also wish to thank Prof. Adriano Martinelli, from the Uni-
versity of Pisa for allowing us to use QikProp software, in order
M. Merkel, M. L. Bowen, B. D. G. Page, D. E. Green,
T. Storr, F. Thomas, D. D. Allen, P. R. Lockman,
B. O. Patrick, M. J. Adam and C. Orvig, Chem. Sci., 2011, 2,
642.
to obtain the estimation of the pharmacokinetic properties of 26 S. Gupta, A. Fallarero, P. Järvinen, D. Karlsson,
the compounds.
M. S. Johnson, P. M. Vuorela and C. G. Mohan, Bioorg.
Med. Chem. Lett., 2011, 21, 151105.
27 W. Luo, Y. P. Li, Y. He, S. L. Huang, J. H. Tan, T. M. Ou,
D. Li, L. Q. Gu and Z. S. Huang, Bioorg. Med. Chem., 2011,
19, 763.
References
1 C. L. Masters, R. Cappai, K. J. Barnham and 28 L. Huang, Z. Luo, F. He, J. Lu and X. Li, Bioorg. Med. Chem.,
V. L. Villemagne, J. Neurochem., 2006, 97, 1700.
2 M. Citron, Nat. Rev. Drug Discovery, 2010, 9, 387.
3 R. Jakob-Roetne and H. Jacobsen, Angew. Chem., Int. Ed.,
2009, 48, 3030.
2010, 18, 4475.
29 M. L. Bolognesi, A. Cavalli, L. Valgimigli, M. Bartolini,
M. Rosini, V. Andrisano, M. Recanatini and C. Melchiorre,
J. Med. Chem., 2007, 50, 6446.
4 H. W. Querfurth and F. M. La Ferla, N. Engl. J. Med., 2010, 30 G. Jones, P. Willett, R. C. Glen, A. R. Leach and R. Taylor,
362, 329.
J. Mol. Biol., 1997, 267, 727.
5 A. Abbott, Nature, 2008, 456, 161.
31 H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland,
T. N. Bhat, H. Weissig, I. N. Shindyalov and P. E. Bourne,
Nucleic Acids Res., 2000, 28, 235; http://www.pdb.org/pdb/
home/home.do (accessed on February–May 2011).
32 L. Zékány, I. Nagypál and G. Peintler, PSEQUAD 501,
2001.
33 E. T. Clarke and A. E. Martell, Inorg. Chim. Acta, 1992, 196,
185.
34 M. Born, Z. Phys., 1920, 1, 45.
35 E. T. Clarke and A. E. Martell, Inorg. Chim. Acta, 1992, 191,
57–63.
36 S. Chaves, S. Canário, M. Carrasco, L. Mira and
M. A. Santos, J. Inorg. Biochem., 2012, 114, 38.
37 QikProp, Version 2.5, Schrödinger, LLC, New York, NY,
2005.
6 M. Stefani and C. M. Dobson, J. Mol. Med., 2003, 81, 678.
7 K. J. Barnham, C. L. Masters and A. I. Bush, Nat. Rev. Drug
Discovery, 2004, 3, 205.
8 A. I. Bush, Curr. Opin. Chem. Biol., 2000, 4, 184.
9 F. Zhou and G. Millhauser, Coord. Chem. Rev., 2012, 256,
2285.
10 H. Kozlowski, Coord. Chem. Rev., 2012, 256, 2129.
11 A. I. Bush and R. E. Tanzi, Neurotherapeutics, 2008, 5, 421.
12 A. I. Bush, Trends Neurosci., 2003, 26, 207.
13 A. I. Mot, A. G. Wedd, L. Sinclair, D. R. Brown, S. J. Collins
and M. W. Brazier, Expert Rev. Neurother., 2011, 11, 1717.
14 C. Rodríguez-Rodríguez, M. Telpoukhovskaia and C. Orvig,
Coord. Chem. Rev., 2012, 256, 2308.
15 J. Kao and G. Grossberg, Cholinesterase inhibitors, in Alzhei-
mer’s Disease, ed. I.-F. Lau and M. A. Brodney, Springer- 38 M. F. Snape, A. Misra, T. K. Murray, R. J. De Souza,
Verlag, Berlin, 2008, pp. 26–51.
16 U. Cornelli, Neurodegener. Dis., 2010, 7, 193.
J. L. Williams, A. J. Cross and A. R. Green, Neuropharma-
cology, 1999, 38, 181.
17 Y. Biran, C. L. Masters, K. J. Barnham, A. I. Bush and 39 J. C. Rochet, Curr. Opin. Struct. Biol., 2000, 8, 10.
P. A. Adlard, J. Cell. Mol. Med., 2009, 13, 61.
18 A. Cavali, M. L. Bolognesi, A. Minarini, M. Rosini,
V. Tumiatti, M. Recanatini and C. Melchiorre, J. Med.
Chem., 2008, 51, 347.
40 D. M. Walsh, E. Thulin, A. M. Minogue, N. Gustavavsson,
E. Pang, D. B. Teplow and S. Linse, FEBS J., 2009, 276,
1266.
41 H. Levine, Protein Sci., 1993, 2, 404.
19 H. Zheng, M. B. H. Youdim and M. Fridkin, J. Med. Chem., 42 M. Bartolini, C. Bertucci, M. L. Bolognesi, A. Cavalli,
2009, 52, 4095.
C. Melchiorre and V. Andrisano, ChemBioChem, 2007, 12,
2152.
43 X. Chao, X. He, Y. Yang, X. Zhou, M. Jin, S. Liu, Z. Cheng,
P. Liu, Y. Wang, J. Yu, Y. Tan, Y. Huang, J. Qin and
S. Rapposelli, Bioorg. Med. Chem. Lett., 2012, 22, 6498.
20 M. A. Santos, S. M. Marques and S. Chaves, Coord. Chem.
Rev., 2012, 256, 240.
21 K. Serdons, K. Van Laere, P. Janssen, H. F. Kung, G. Bormans
and A. Verbruggen, J. Med. Chem., 2009, 52, 7090.
22 L. Huang, C. Lu, Y. Sun, F. Mao, Z. Luo, T. Su, H. Jiang, 44 S. Chaves, R. Jelic, C. Mendonça, M. Carrasco,
W. Shan and X. Li, J. Med. Chem., 2012, 55, 8483.
23 A. K. Sharma, S. T. Pavlova, J. Kim, D. Finkelstein,
Y. Yoshikawa, H. Sakurai and M. A. Santos, Metallomics,
2010, 2, 220.
N. J. Hawco, N. P. Rath, J. Kim and L. M. Mirica, J. Am. 45 L. D. Devanur, H. Neubert and R. C. Hider, J. Pharm. Sci.,
Chem. Soc., 2012, 134, 6625.
2008, 97, 1454.
6072 | Dalton Trans., 2013, 42, 6058–6073
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
46 M. Merkofer, A. Domazou, T. Nauser and W. H. Koppenol,
M. L. Serralheiro and M. A. Santos, J. Enzyme Inhib. Med.
Chem., 2011, 26, 485.
Eur. J. Inorg. Chem., 2006, 671.
47 W. L. F. Armarego and D. D. Perrin, Purification of Labora- 51 F. J. C. Rossotti and H. Rossotti, J. Chem. Educ., 1965, 42,
tory Chemicals, Butterworth-Heinemann, Oxford, 4th edn,
1999.
48 Maestro, Version 7.5, Schrödinger Inc., Portland, OR, 2005.
49 Macromodel, Version 8.5, Schrödinger Inc., Portland, OR,
1999.
375.
52 P. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 1739.
53 B. Tepe, O. Eminagaoglu, A. Akpulat and E. Aydin, Food
Chem., 2007, 100, 985.
54 M. L. Michaelis, R. T. Dobrowsky and G. Li, J. Mol. Neuro-
sci., 2002, 19, 289.
50 J. Sebestík, S. M. Marques, P. L. Falé, S. Santos,
D. M. Arduíno, S. M. Cardoso, C. R. Oliveira, 55 T. Mosmann, J. Immunol. Methods, 1983, 65, 55.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6058–6073 | 6073