Ionophoric polyphenols selectively bind Cu2+, display potent antioxidant and anti-amyloidogenic properties, and are non-toxic toward Tetrahymena thermophila

Article abstract of DOI:10.1016/j.bmc.2016.06.012

Alzheimer's disease (AD) is the most common form of dementia affecting more than 28 million people in the world. Only symptomatic treatments are currently available. Anticipated tri-fold increase of AD incidence in the next 50 years has established the ne

Full text of DOI:10.1016/j.bmc.2016.06.012

Accepted Manuscript
2
+
Ionophoric polyphenols selectively bind Cu , display potent antioxidant and
anti-amyloidogenic properties, and are non-toxic toward Tetrahymena thermo-
phila
Alberto Martínez, Ralph Alcendor, Tanzeen Rahman, Magdalena Podgorny,
Ismaila Sanogo, Rebecca McCurdy
PII:
S0968-0896(16)30435-7
DOI:
http://dx.doi.org/10.1016/j.bmc.2016.06.012
BMC 13068
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
28 March 2016
30 May 2016
3 June 2016
Please cite this article as: Martínez, A., Alcendor, R., Rahman, T., Podgorny, M., Sanogo, I., McCurdy, R.,
2
Ionophoric polyphenols selectively bind Cu , display potent antioxidant and anti-amyloidogenic properties, and
+
are non-toxic toward Tetrahymena thermophila, Bioorganic & Medicinal Chemistry (2016), doi: http://dx.doi.org/
1
0.1016/j.bmc.2016.06.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

2+
Ionophoric polyphenols selectively bind Cu , display potent
antioxidant and anti-amyloidogenic properties, and are non-toxic
toward Tetrahymena thermophila
a,
b
a
a
Alberto Martínez *, Ralph Alcendor , Tanzeen Rahman , Magdalena Podgorny , Ismaila
a
a
Sanogo and Rebecca McCurdy
a
Department of Chemistry, New York City College of Technology, The City University of New
York (CUNY), Brooklyn, New York, 11201, USA.
b
Department of Biology, New York City College of Technology, The City University of New
York (CUNY), Brooklyn, New York, 11201, USA.
*
Corresponding author: Dr. Alberto Martínez
Tel: 1-718-260-5855
E-mail: amartinez@citytech.cuny.edu
In memory of Roberto A. Sánchez-Delgado

ABSTRACT
Alzheimer’s disease (AD) is the most common form of dementia affecting more than 28 million
people in the world. Only symptomatic treatments are currently available. Anticipated tri-fold
increase of AD incidence in the next 50 years has established the need to explore new possible
treatments. Accumulation of extracellular amyloid- (A) plaques, intracellular tangles in the
brain, and formation of reactive oxygen species (ROS) are the major hallmarks of the disease.
2
+
The active role of some metal ions, especially Cu , in promoting both A aggregation and
reactive oxygen species formation has rendered ionophoric drugs as a promising treatment
strategy. In this work, a series of 5 disease-modifying and multi-target ionophoric polyphenols
(
1-5), inspired on the structure of natural resveratrol, have been synthesized and characterized.
2
+
All compounds bind Cu selectively over other biologically relevant metal ions. They form 2:1
2
+
(
compound:Cu ) complexes with association constants logK
a
12-14 depending on the molecular
design. Our results indicate that compounds 1-5 possess excellent antioxidant properties: they
2
+
inhibit the Cu -catalyzed reactive oxygen species production between 47% and 100%, and they
scavenge DPPH (1,1-diphenyl-2-picryl-hydrazyl) and AAPH (2,2’-azobis(2-amindino-propane)
dihydrochloride) free radicals in general better than clioquinol, resveratrol and ascorbic acid. In
2
+
addition, compounds 1-5 interact with A peptides and inhibit both the Cu -catalyzed
aggregation and the self-assembly of A(1-40) up to a ~92% extent. Interestingly, 1-5 are also
able to disaggregate up to ~91% of pre-formed A(1-40) aggregates. Furthermore, cytotoxic
studies show remarkably low toxicity of 1-5 toward Tetrahymena thermophila with LD50 values
higher than 150 M, comparable to non-toxic natural resveratrol.

KEYWORDS: Alzheimer’s disease, polyphenol, ionophoric compound, amyloid-beta
aggregation, multi-target compound, reactive oxygen species, antioxidant, Tetrahymena
1
. Introduction
Alzheimer’s disease (AD) is the most common and incurable form of dementia, affecting more
1
,2
than 28 million people worldwide. It is most often diagnosed in people over 65 years old,
although early-onset AD can also occur. These numbers are expected to increase rapidly due to
the aging population, the rise in life expectancy in both the developed and developing countries,
3
,4
and the lack of any effective treatment. While the etiology and pathological mechanisms of
AD are still under debate, the accumulation of extracellular amyloid- (A) plaques, intracellular
tangles in the brain, and formation of reactive oxygen species (ROS) are the major hallmarks of
5
-10
the disease that lead to an uncontrolled neuronal death. Fragments A(1-40) and A(1-42) are
9
,10
the typical components of A plaques.
The production of A is a normal physiological
1
1,12
process occurring in the brain and cerebrospinal fluid throughout life.
However, according to
7
,8
the amyloid cascade hypothesis, an increased Aproduction and/or accumulation leads first to
the formation of A oligomers, then to protofibrils and ultimately to fibrils that are the principal
constituent of the plaques.
Therapeutic targets for the treatment of AD have been classified into two groups: 1) symptomatic
therapies, and 2) disease-modifying targets. Current approved treatments (Figure 1) are old,
ineffective and based on symptomatic therapies, i.e. drugs that only attenuate the symptoms but
do not interdict the underlying causal pathobiology. They involve four acethylcholinesterase
inhibitors (tacrine, rivastigmine, galantamine and donepezil) and an NMDA receptor antagonist
(
memantine). The rationale for the use of these drugs is based on the cholinergic hypothesis, one

1
3
of the earliest and most extensively researched theories regarding AD. However, the benefits
from their use are very limited and none of them clearly shows to delay or halt the progression of
AD. In addition, the use of anticholinergic drugs has been recently linked to higher risk of
1
4
dementia and, thus, the search for novel chemotherapies for Alzheimer’s disease clearly
requires urgent attention.
Tacrine
Rivastigmine
Galantamine
Donepezil
Clioquinol
Memantine
Desferroxiamine
Resveratrol
PBT2
Figure 1. Five approved AD treatments (tacrine, rivastigmine, galantamine, donepezil and
memantine), examples of chelating agents that reached clinical trials in AD (desferroxiamine,
clioquinol and PBT2), and polyphenol resveratrol.
Metal ions are thought to be involved in AD in different ways, and consensus arises in the case
2
+
2+
of Cu . A wealth of evidence for this comes from two facts: 1) Cu has been detected in
1
2,15
2+
elevated concentrations in A plaques,
and 2) Cu binding to A is aberrant and does not

take place under healthy conditions. Essential copper (Cu, referring to its different oxidation
states) is normally present in proteins related to electron and oxygen transport, although it can
also be found in synaptic vesicles as “free” ions (loosely bound to low affinity ions, for instance
1
6
glutamate). Importantly, there is evidence indicating that Cu homeostasis is altered in AD
1
7-20
brains, resulting in abnormal concentrations of its metal ions in the synaptic cleft.
The
binding of aberrant concentrations of “free” Cu to Ais then able to promote aggregation into
2
+
toxic and neurodegenerative forms of the protein. In addition, redox active metal ions like Cu
are crucial for the production of reactive oxygen species (ROS) and oxidative stress that
2
1
2+
ultimately lead to neuronal death. Proof has been provided that the complex with A (Cu -A)
2
2,24
might be directly involved in the production of ROS.
This mounting evidence supporting the
2
+
role of metal ions, specifically Cu , in the pathophysiology of AD has rendered metal
ionophoric therapy as a promising treatment strategy. In vitro studies have shown that production
2
+
of ROS via interaction of Cu with Ais significantly attenuated by the presence of
diethylenetriaminepentaacetic acid (DTPA), triethylenetetraamine (TETA) and desferroxiamine
2
5
(
DFO), known as potent metal chelators. Furthermore, the latter, DFO, clioquinol (CQ) and
second generation PBT2 (Figure 1) have shown indication of slowing neurological decline in
2
5-29
early stage clinical trials.
Therefore, the potential shown by this family of compounds has
opened a promising line of research and, although some groups have lately reported their work
2
+
2+ 30-34
on small molecules able to chelate Cu and Zn ,
research on this issue clearly requires
further attention.
Promising strategies in the treatment of AD also involve the use of natural products capable of
crossing the blood brain barrier (BBB) to display neuroprotective properties by either
detoxification of ROS, anti-amyloidogenic potential or interaction with relevant proteins closely

3
5
36-39
related to AD. Prominent examples are resveratrol, curcumin or flavonoids.
Among them,
resveratrol (3,5,4’-trans-trihydroxystilbene, Figure 1), mostly found in berries and grapes, has
currently completed phase II clinical trials in 120 patients with possible AD diagnosis due to its
4
0
ability to modify important AD biomarkers. Several health-promoting properties have been
reported for resveratrol due to its intrinsic antioxidant capacity. Interestingly, this natural
compound is able to penetrate the blood-brain barrier (BBB) and exert strong neuroprotective
4
1
effects, even at low concentrations. It has proven to fight the neuronal dysfunction caused in
4
2,43
AD by several proposed mechanisms.
Very importantly, resveratrol also exerts its
neuroprotective activity by acting as an anti-amylodiogenic and fibril-destabilizing
4
4,45
polyphenol,
and by detoxifying metal-Ainduced ROS, although it is not sufficient to fully
block metal-Atoxicity, likely because of its inability to effectively compete with toxic forms of
4
6
A peptides for the binding of AD-related metal ions.
Due to the complexity of AD and the identification of many potential targets, the design of
disease-modifying multi-target compounds has recently attracted the attention of researchers in
4
7,48
the pursuit of improved drugs.
In line with this, herein we present the synthesis of a family of
multi-functional polyphenols (compounds 1-5 in Schemes 1 and 2), inspired on the structure of
2
+
natural resveratrol, that additionally incorporate a Cu -selective metal ionophoric motif in their
2
+
structure for enhanced activity. We also report: 1) their Cu binding properties and ion
2
+
selectivity, 2) their ability to both inhibit the Cu -A(1-40) induced ROS formation and
scavenge DPPH (1,1-diphenyl-2-picryl-hydrazyl) and AAPH (2,2’-azobis(2-amindino-propane)
dihydrochloride) free radicals, 3) their potential to inhibit the aggregation of A(1-40) into
pathological forms of the peptide, and 4) their cytotoxicity toward healthy eukaryotic cells of
Tetrahymena thermophila.

2
2
. Results and discussion
.1. Drug design and drug likeliness
Ionophoric polyphenols 1-5 (Schemes 1 and 2) were designed inspired on clinically tested
resveratrol, because of its proven anti-AD properties and lack of toxicity. Importantly, the
2
+
resveratrol-like structures were provided with a Cu selective ionophoric motif for increased
anti-AD potential resulting from the combination. Thus, compounds 1-5 were expected to
maintain and/or enhance all intrinsic antioxidant and anti-amyloidogenic properties of natural
2
+
polyphenols, and display the added benefit of selective Cu chelation (Scheme 1). In general,
two major issues must be tackled when designing anti-AD ionophoric agents: 1) finding the right
balance between activity, selectivity and metal ion affinity, and 2) achieving blood brain barrier
(
BBB) permeability. On one hand, ligands displaying a high affinity and/or low selectivity for
metal ions might present undesired side effects. In this context, the molecular design
incorporates a low-medium affinity and selective chelating motif (bidentate NH-OH or N-OH)
into the structure of a polyphenol-like compound (see section 2.3 below for expanded discussion
and results on metal ion selectivity). On the other hand, failure to cross the BBB is a major
4
9
impediment for the design of effective drugs in anti-AD therapy. BBB permeability of
molecules can be predicted from Lipinski's rules, a set of criteria that evaluate molecular
5
0,51
properties important for drug pharmacokinetics.
These rules are based on the observation that
most of the orally active drugs satisfy certain standards regarding molecular weight, lipophilicity
(
(
cLogP), number of hydrogen bond donors (HBD) and acceptors (HBA), and polar surface area
PSA). As observed in Table 1, the compounds herein presented satisfy the drug-like criteria
comparably to both clinically tested resveratrol and CQ. Importantly for us, the calculation of

logBB for predicted BBB penetration ability (see equation on footnote of Table 1) using
Lipinski’s rules confirms probable brain distribution for all compounds.
Scheme 1. Design of ionophoric polyphenols combining the properties of natural polyphenols
and ionophoric agents. Chelating motif in grey shade.
Table 1. Physicochemical properties of the ionophoric polyphenol compounds 1-5, CQ and
resveratrol.
Lipinski’s
1
2
3
4
5
Resveratrol
CQ
Rules
≤450
≤5.0
≤10
MW
cLogP
HBA
215.25 215.25 263.30 231.25 279.30
228.25
2.833
3
305.5
3.729
2
1.755
3
1.755
3
3.59
3
1.158
4
2.924
4
HBD
PSA
3
3
2
4
3
3
1
≤5
52.49
52.49
-0.30
52.82
-0.11
72.72
-0.77
73.05
-0.51
60.69
-0.27
32.59
0.21
≤90
LogBB -0.30
*
*
Calculated according to logBB = -0.0148xPSA + 0.152xclogP + 0.130. Drugs with logBB below -1 are expected to be poorly
50,51
distributed in the brain.
MW, molecular weight; cLogP, calculated logarithm of the octanol-water partition coefficient; HBA, hydrogen bond acceptor;
HBD, hydrogen bond donor; PSA, polar surface area.

One last consideration in the molecular design is concerned with the nature of the nitrogen
functional group. In line with this, imino compounds 3 and 5 were expected to display different
electron delocalization properties than amino 1, 2 and 4 when scavenging free radicals. In
addition, we anticipated that reduced compounds 1, 2 and 4 would present better solubility and,
to some extent, different metal ion affinity or selectivity. Thus, the synthesis of both types was of
interest in order to establish structure-activity relationships.
2
.2. Chemistry
Target compounds 1, 2 and 4 were synthesized by means of a typical two-step reductive
5
2,53
amination, similarly to procedures previously described in the literature.
In the first step, the
amino group of 2-aminophenol was condensed with the aldehyde group of 4-
hydroxybenzaldehyde, 3-hydroxybenzaldehyde or 3,4-dihydroxybenzaldehyde, respectively, in
refluxing dry methanol through a Schiff base reaction. In the second step, the resulting imine was
reduced at 0-2ºC by addition of NaBH to produce the desired product in good yields.
4
Compounds 3 and 5 were synthesized by a Schiff base condensation of 2-hydroxy-1-
naphtaldehyde with 2-aminophenol or 4-aminoresorcinol, respectively, in refluxing dry
methanol, and isolated as stable imino compounds (Scheme 2).
Regarding stereochemical considerations, compounds 3 and 5 were obtained in their E
1
1
conformation for the C-N double bond, as confirmed by H- H NOESY experiments showing a
diagnostic signal between the proton in the CH of the imino group and the hydrogen in ortho of
the phenol group (See compound 3 as example in Supplementary Information, Figure S1).
Additionally, energy optimization calculations revealed a difference of 10.8 and 14.6 kcal/mol
between Z and E configurations (Z displaying higher energy than E) for 3 and 5, respectively.

Imino intermediates of 1, 2 and 4 were also isolated, but NMR showed decomposition in
presence of water or D
further testing.
2
O and, therefore, only the reduced amino derivatives were used for
1
2
5
4
3
Scheme 2. Synthesis of compounds 1-5. a, b and d) i. Reflux, 2h, MeOH; ii. NaBH
ºC; c and e) Reflux, 5h, MeOH.
4
, MeOH, 0-
2
2
.3. Metal ion binding properties and selectivity
Although metal chelation represents a promising strategy for the treatment of AD, long term use
of ionophoric agents might affect the homeostasis of numerous biometals and the normal
2
9,31
physiological functions of biomolecules that require of those metal ions.
For this reason, it is
of critical importance that ligands targeting toxic concentrations of specific ions present both an
ion-selective binding rather than systemic, and a moderate-low ion affinity. This would allow to

2
+
compete with A for metal binding (A forms 1:1 complexes with Cu with LogK
a
values in the
), but not with metalloenzymes, therefore,
resulting in a combination that could promote restoration of ionic balance rather than total
1
7,54
range 8-10, depending on experimental conditions
5
5
2+
removal of metal ions. Given the active role of Cu in promoting A aggregation and ROS
production, compounds 1-5 were studied spectrophotometrically and compared to CQ and
2
+
resveratrol to determine 1) their ability to selectively chelate Cu versus other biologically
2
+
2+
2+
2+
2+
2+
relevant metal ions, such as Zn , Fe , Ca , Mn and Mg , and 2) their Cu binding
2
+
properties, including stoichiometry and association constants of the compound-Cu adducts. CQ
2
+
(
used here as a control) is known to form 2:1 complexes with Cu with low association
5
6,57
57,58
constants in the range of LogK
a
10-16,
depending on buffer conditions;
these values are
5
9
lower than those observed for other 8-hydroxyquinolines. Because of its low metal binding
affinity, literature argues that CQ has the potential not to interfere with intracellular metal-
5
7
binding enzymes and, therefore, CQ is claimed to be able to remove metal ions from A and
5
5
redistribute them within the brain regions where they can promote synaptic function. However,
severe myelo-optic acute neuropathy on patients treated with CQ has been associated to lack of
2
+ 29
tissue and/or ion selectivity and, most especially, to the ability to chelate Zn . Our results for
CQ metal binding agree, in general, with these latter findings; CQ displayed an average LogK
a
of
2
+
2+
2+
2+
2+
2+
1
5.58, but exhibited a poor selectivity for Cu with respect to Zn , Fe , Ca , Mn and Mg
(
Figure S2 in Supplementary Information). On the other hand, natural resveratrol showed a
2
+
2+
2+
2+
2+
2+
selective binding for Cu and Fe over Zn , Ca , Mn or Mg (Figure S2 in Supplementary
2
+
Information), but displayed a much lower association constant for Cu (logK = 6.70, on
a
average) likely insufficient to compete with A for metal binding and, therefore, to prevent AD
Cu toxicity.

The interaction of 1-5 with metal ions was spectrophotometrically followed by changes in the
absorption spectra upon interaction with metal ions. For metal ion selectivity, buffered solutions
2
+
2+
2+
of compounds 1-5 were prepared and treated with the metal ion under study (Zn , Fe , Ca ,
2
+
2+
Mn and Mg ). After 5 minutes incubation, the optical response was observed, followed by
2
+
addition of Cu and consequent analysis of further spectral changes (see Experimental Section
for more details). Results for compounds 1-5 (only 1 and 3 are shown in Figure 2 as
representative examples of amino and imino compounds; see the behavior of compounds 2, 4, 5,
CQ and resveratrol in Supplementary Information, Figure S2) show that these ionophoric
2
+
polyphenols bind Cu with a remarkably better selectivity than clinically tested CQ; this feature
could have important implications in preventing side effects. Binding stoichiometry and apparent
2
+
association constants for Cu binding were calculated from spectrophotometric titrations of 1-5
2
+
upon stepwise additions of a Cu solution (see Figure S3 in Supplementary Information). These
2
+
experiments revealed that all the ionophoric polyphenols form 2:1 complexes with Cu
2
+
(
compound: Cu ; binding through the chelating motif NH-OH for amino compounds, or N-OH
for imino compounds is proposed. See Scheme 1), similarly to what has been observed for CQ.
Interestingly, compounds 1-5 displayed low values for LogK (see Table 2), as proposed in the
a
molecular design (Section 2.1), adequate to prevent competition with cuproenzymes. When
compared within the series of ionophoric polyphenols, amino compounds 1, 2 and 4 showed
lower association constants and slightly worse metal selectivity than imino derivatives 3 and 5.
2
+
The experiments described in the following sections were aimed at correlating Cu binding
properties with the compounds’ ability to inhibit ROS production and A(1-40) aggregation.

A
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
M2+ + Cu2+
M2+
Normalized
Abs374nm
Fe
Zn
Ca
Mn
Mg
B
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Normalized
Abs465nm
Fe
Zn
Ca
Mn
Mg
Figure 2. Relative changes in optical response (absorbance intensity) when compounds 1 (A)
and 3 (B) interact with metal ions (see Figure S2 in Supplementary Information for results on
compounds 2, 4, 5, CQ and resveratrol). Black bars represent normalized affinity for metal ions
2
+
2+
2+
2+
2+
2+
M (M= Zn , Fe , Ca , Mn or Mg ), while blue bars indicate affinity for Cu in presence of
M.

2
+
Table 2. Apparent association constants for Cu , and binding affinity to A(1-40) peptide
(
calculated from tyrosine fluorescence quenching in titration curves; Figure S7 in Supplementary
Information) of compounds 1-5, CQ and resveratrol.
2
+
Apparent Association Constant for Cu
Compound
5
-1
Binding Affinity for A(1-40) (x 10 M )
(
LogK )
a
CQ
15.58 ± 0.34
6.70 ± 0.25
12.82 ± 0.40
13.07 ± 0.40
14.01 ± 0.19
12.62 ± 0.12
14.29 ± 0.35
3.69± 0.59
1.65± 0.17
3.27 ± 0.65
1.55± 0.15
1.91± 0.19
3.55± 0.52
2.46± 0.13
Resv
1
2
3
4
5
2
+
2
.4. Antioxidant ability: Inhibition of Cu -A(1-40) hydroxyl radical production and
scavenging of DPPH and AAPH radicals
Evidence indicates that one of the main causes of neuronal cell loss in AD is due to excessive
.
.-
free radical generation. ROS, such as OH and O
2
, which are produced under oxidative stress,
seem to play a role in the neurotoxicity of A, and research has shown that at least trace amounts
6
0-62
of metals are necessary for the formation of such toxic species.
A causative link between the
2
+
2+
Cu -A complex and oxidative stress is supported by a body of evidence: 1) sequestering Cu
from A induces a significant decrease in A toxicity, 2) only Cu-containing A aggregates are
toxic to primary neurons, and 3) Cu-containing fibrils are more toxic to PC12 cells than other

1
7
2+
.-
forms of the peptide. The complex Cu -A is known to catalyze the generation of O
2
, H
2
O
2
.
and OH through standard Fenton reactions in presence of molecular oxygen and a reducing
6
3
agent. This becomes relevant given that the brain environment is well supplied in oxygen and
several reducing agents (ascorbate, L-dopa, dopamine, cholesterol, etc) are easily available at
relatively high concentrations (for instance, brain concentration of ascorbate ranges between 200

M and 10 mM).
2
+
In this work we have measured the ability of compounds 1-5 to inhibit the Cu -A(1-
.
4
0)catalyzed formation of hydroxyl radical (OH ), one of the most toxic radicals in biological
systems. The effect of resveratrol and CQ was also analyzed for comparison purposes. Hydroxyl
2
+
radical production was promoted through Fenton reaction of Cu -A(1-40) with dissolved O in
2
presence of ascorbate and 2 eq. of the compounds under study (see Experimental Section for
more details). Furthermore, time course plots (Figure S4 in Supplementary Information) were
used to calculate the apparent rate of ROS formation by determining the slope of the linear
segment of the graphs. All results are displayed in Figure 3 and Table 3, and show that, in


general, compounds 1-5 both inhibit and slow down the Cu -A(1-40) catalyzed generation of
hydroxyl radical better than parent resveratrol. Importantly, compounds 3 and 5 completely
2
+
suppress the Cu -A(1-40) catalyzed ROS formation, better than clinically tested resveratrol
2
+
and CQ, most likely due to their higher Cu binding affinity. Results for CQ agree with
6
3
previously reported experiments.

1
20
00
1
80
60
40
20
0
Cu-Ab Cu-Resv Cu-CQ Cu-1
Cu-2
Cu-3
Cu-4
Cu-5
.
2+
Figure 3. Percent OH generation by the adduct Cu -A1-40) in presence of 2 eq. of
2
+
compounds 1-5, CQ and resveratrol. ROS generation by the Cu -A(1-40) adduct was taken as
2
+
1
4
00% positive control. See time course production of hydroxyl radical by the adduct Cu -A(1-
0) in presence of compounds 1-5 in Supplementary Information, Figure S4.
In addition, the total antioxidant capacity of compounds 1-5 was also determined in absence of
2
+
64
Cu ions by the DPPH assay and a modification of the ORAC (oxygen radical absorbance
6
5,66
capacity) assay.
The DPPH assay is based on the reduction of the stable free radical 1,1-
diphenyl-2-picryl-hydrazyl (DPPH). Because of its unpaired electron, DPPH gives a strong
absorption maxima at 517 nm (purple color), as the odd electron of the radical becomes paired
off in presence of a radical scavenger, the absorption intensity decreases and the discoloration is
6
7
stoichiometric with the number of electrons captured. The DPPH assay is commonly used to
65
establish the antioxidant activity of phenolic compounds as well as natural plant extracts. On
the other hand, assays based on the scavenging of 2,2’-azobis(2-amindino-propane)
dihydrochloride (AAPH), like the ORAC assay, are said to be more relevant because they use a

biologically significant radical source. This method determines the ability of antioxidants to
protect fluorescein (FS) from damage by free radicals. It consists of measuring the decrease in
fluorescence of FS when it suffers oxidative damage caused by a source of peroxyl radical
.
(
ROO ) such as AAPH. This is important given that the presence of oxygen radicals in AD brains
is a common feature.
2
+
Table 3. Antioxidant properties of compounds 1-5. Ability to inhibit Cu -A1-40) catalyzed
hydroxyl radical production and to scavenge DPPH and AAPH free radicals.
2
+
%
inhibition of Cu -
Rate of radical
formation
IC50 for DPPH
radical** (r.a.)
IC50 for AAPH
radical** (r.a.)
Compound
A(1-40) catalyzed
ROS production
0
(a.u./min)*
13.0 ± 0.2
2
+
Cu -A(1-40)
--
--
CQ
87.7 ± 0.9
9.0 ± 3.3
--
1.2 ± 0.4
11.2 ± 0.4
--
40.6 ± 6.3 (0.4)
17.8 ± 0.8 (1.0)
12.8 ± 0.8 (1.4)
6.27 ± 0.59 (2.8)
5.87 ± 0.48 (3.0)
38.5 ± 5.8 (0.5)
3.05 ± 0.10 (5.8)
13.5 ± 0.3 (1.3)
9.72 ± 0.42 (0.2)
1.70 ± 0.04 (1.0)
62.2 ± 0.2 (0.02)
1.38 ± 0.13 (1.2)
1.24 ± 0.15 (1.4)
11.5 ± 0.5 (0.1)
0.61 ± 0.08 (2.8)
3.15 ± 0.08 (0.5)
Resv
Ascorbic Acid
1
2
3
4
5
47.3 ± 1.0
51.1 ± 1.3
100.0
6.8 ± 0.6
5.4 ± 0.8
0.00
59.9 ± 2.4
98.5 ± 0.2
4.0 ± 0.2
0.03
*
Calculated as the slope of the lineal part of the time course plot fluorescence intensity vs. time. Figure S4 in Supplementary
Information.
*All values in M concentration.
r.a. (relative activity): IC50 resveratrol/IC50 compound.
*

6
4
The DPPH assay was performed as previously published by others. IC50 values for the
scavenging of DPPH radicals were calculated and compared to those of resveratrol, CQ and
ascorbic acid used as controls. All results are shown in Table 3 and Figure S5 in Supplementary
Information (dose vs. response curves), and reveal that amino compounds 1, 2 and 4 are between
2
.8 and 5.8 better at scavenging DPPH radicals than resveratrol, and between 2 and 4.2 times
more active than ascorbic acid. On the other hand, compound 5 proved to have a better activity
than resveratrol and comparable to that of ascorbic acid, while compound 3 and CQ showed poor
DPPH scavenging properties. Our results for ascorbic acid agree with previously reported values
6
4
measured under the same conditions, whereas values for resveratrol have been reported at
6
5,66
different concentrations or buffer solutions and cannot be directly compared.
experimental conditions were chosen to ensure spectrophotometric accuracy at 517 nm
absorbance between 0.1-1.0).
To measure the ability of compounds 1-5 to scavenge peroxyl radicals, we used a modification of
Our
(
6
5,66
the ORAC assay.
The protective effect on FS of the ionophoric polyphenols was then
compared to that of resveratrol, CQ and ascorbic acid used as controls. Results, shown in Table 3
and Figure S6 in Supplementary Information (dose vs. response curves), generally follow a
similar trend to that observed in the DPPH assay. Amino compounds 1, 2 and 4 display a better
ability to reduce the peroxyl radical AAPH in absence of metal ions. Importantly, compound 4 is
up to 2.8 times better scavenger than resveratrol under the studied conditions. In general,
compounds 3, 5 and CQ are less efficient than resveratrol, while ascorbic acid proved to have a
low peroxyl radical protecting ability when compared to the polyphenols, as previously reported
6
6,69
by others.

Results obtained for compounds 3 and 5 in all antioxidant tests herein performed seem to
indicate that the activity observed in the hydroxyl radical inhibition experiment (Figure 3) was
2
+
2+
mainly due to their better ability to sequester Cu ions as a consequence of a higher Cu
2
+
binding affinity, whereas amino compounds 1, 2 and 4, despite having a lower Cu affinity and
therefore being worse inhibitors of fenton-produced radicals, they are better antioxidant agents in
absence of metal ions. We hypothesize then that this behavior could be partly due to radical
scavenging by hydrogen transfer from the amino group. In particular, all antioxidant tests taken
together show that compound 4 displays the best overall antioxidant activity of our ionophoric
2
+
polyphenols, inhibiting to a 60% extent the Cu -A(1-40) catalyzed production of hydroxyl
radicals, while being able to reduce and scavenge DPPH and AAPH free radicals better than both
resveratrol and ascorbic acid. These observations might be of importance in the rational design
of future generations of anti-AD agents with antioxidant properties.
2
+
2
.5. Inhibition of Cu -induced A(1-40)aggregation
The amyloid cascade hypothesis suggests that increased production in A peptides leads first to
the formation of A amorphous oligomers that evolve then to protofibrils and ultimately to
9
,10
2+
fibrils, the main constituents of the amyloid plaques.
There is evidence indicating that Cu
promotes A aggregation into neurodegenerative forms of the protein in a pH dependent fashion.
2
+
It has been reported that Cu -catalyzed aggregation is more significant at slightly acidic pH
70
(
6.6-6.8) than at 7.4, and this becomes relevant given that altered proton homeostasis, known as
7
1,72
2+
cerebral acidosis, is hypothesized in AD.
Thus, the ability of 1-5 to inhibit the Cu -induced
A(1-40) aggregation was tested and compared with that of CQ and resveratrol at pH=6.60.
2
+
Solutions of A(1-40) were incubated up to 18 days with 1 eq. of Cu and 2 eq. of compounds

1
-5. The amount of aggregates was then measured after 3, 7 and 18 days by means of the
thioflavin-T (ThT) and turbidity assays. ThT exhibits a moderately strong fluorescence emission
at 485 nm upon interaction with mature A fibrils and, therefore, the ThT assay is used as a key
diagnostic for the formation of this kind of aggregates. On the other hand, the turbidity assay
measures light scattering due to the presence of protein aggregates in solution and, therefore, it
can determine the total amount of aggregates, including both the initial amorphous oligomeric
form and mature fibril aggregates that form A plaques. However, in agreement with what other
7
3
researchers have suggested, we observed that amorphous aggregates exhibited remarkably
higher turbidity than mature fibrils and, therefore, in this work turbidity results were taken
mostly as evidence of misfolded structures of A(1-40) in the oligomeric form. Whether the
7
4
oligomeric or the fibril is the most toxic form of A aggregates still remains controversial.
Figure 4 shows results when ThT fluorescence and turbidity were monitored up to 18 days.

A
350
300
250
200
150
100
3 days incubation
days incubation
7
50
0
Ab Ab-Cu Cu-CQCu-Resv Cu-1 Cu-2 Cu-3 Cu-4 Cu-5
0.025
0.020
0.015
0.010
0.005
0.000
B
3 days incubation
7
1
days incubation
8 days incubation
Abs550 nm
Ab
Ab-Cu Cu-CQ Cu-Resv Cu-1
Cu-2
Cu-3
Cu-4
Cu-5
2
+
Figure 4. Inhibition of Cu -induced A(1-40) aggregation by compounds 1-5, CQ and
resveratrol studied by ThT (A) and turbidity (B) assays. No significant changes were observed in
ThT fluorescence after 18 days of incubation, and, therefore, that data is not shown for clarity.
All values of fluorescence and percent inhibition of fibril aggregation can be found in Table S1
in Supplementary Information.

Several interesting observations can be made when turbidity and ThT assays results are taken
7
5-78
together: 1) Previously reported results by ThT, AFM and electron microscopy
indicate that
2
+
the presence of Cu exacerbates the formation of amorphous aggregates (when compared to A
self-assembly) and inhibits the fibrilization of A. The increase in turbidity and decrease of ThT
2
+
fluorescence observed when A(1-40) was treated with Cu in our experiments (see Figure 4)
agree with previous findings and validate our observation of higher turbidity for amorphous
2
+
forms of A(1-40)aggregates. The Cu ion effect has been attributed to its preferential binding
to high affinity protein sites of the N-terminal side of the A peptide that are crucial for -sheet
formation and consequent stabilization of non-amyloidogenic conformations. A slight decrease
2
+
of turbidity combined with an increase in fluorescence is observed for control Cu -A(1-
4
0)from 3 to 18 days incubation, which suggests that amorphous aggregates might be evolving
to mature A(1-40) fibrils with time. 2) ThT and turbidity assays for resveratrol and CQ show
significantly distinct behavior for the two controls. The natural polyphenol attenuates the
appearance of mature A(1-40) fibrils to an
ꢀ 85% extent (see Table S1 in Supplementary
Information for all values of fluorescence and percent inhibition) by promoting the aggregation
7
9
into amorphous forms, a phenomenon that has been previously observed for resveratrol and
7
3
several other natural polyphenols like morin, quercitin, ferulic acid or gossypol. It is
hypothesized that fibrillar structures possess structural features in which aromatic residues of the
A peptide engage in stacking interactions that are absent in the oligomeric forms. It is then
suggested that resveratrol and other polyphenols can interact with aromatic chains of A, which
thus explains the fact that such polyphenols seem to selectively destabilize the A fibril
7
9 and refs. therein
population.
In particular, it has been found that resveratrol remodels the A
7
9
aggregation to form non-toxic high-molecular weight oligomeric aggregates. The formation of

such aggregates would agree with the increase of turbidity observed in our experiments. On the
2
+
other hand, CQ directs and enhances the Cu -induced aggregation process into assembly to
2
+
mature A(1-40) fibrils (~195-250% fibrils with respect to the Cu -A(1-40) control are
detected in our experiments; see Table S1 in Supplementary Information), a phenomenon
previously reported and associated to two main factors: Firstly, metal removal from the A(1-40)
protein by CQ seems to restore the aggregational properties of the native peptide. Secondly, the
interaction of CQ with hydrophobic domains of A(1-40) could stimulate protein conversion into
8
0
mature fibrils (in section 2.6 the ability of CQ to interact with A(1-40) has been tested). 3)
2
+
Compound 3 prevents the formation of Cu -induced total aggregates by inhibiting the
misfolding of A(1-40) into its amorphous form up to a ~50% extent, better than resveratrol and
slightly better than CQ, and the fibrillary assembly up to a ~67% extent. 4) ThT assay reveals
that all compounds inhibit the formation of mature A(1-40) fibrils between a ~67% and a
~92% extent, comparably to resveratrol and much better than CQ; and 5) compounds 1, 2, 4 and
5
destabilize the A(1-40) fibril formation by initially promoting amorphous aggregates. As
previously stated, literature on polyphenols, as well as on small molecules and peptides, suggests
that this kind of compounds might redirect the aggregation cascade instead of inhibiting it
8
1-83
completely.
Importantly, the high similarity with the results observed for resveratrol allows
us to hypothesize that our ionophoric polyphenols 1-5 might be exerting a comparable effect to
7
9
that of the natural polyphenol and remodeling the aggregation towards non-toxic high
molecular weigth oligomeric forms of A, although this statement cannot be fully confirmed
with the data at hand. In addition, our results show that prolonged exposure to compounds 1-5
(
up to 18 days were studied in this work) induces a decrease in the amount of amorphous
aggregates without increasing the fibril population. The differential turbidity pattern observed for

compound 3 with respect to the rest of ionophoric polyphenols could be partly explained by its
2
+
2+
higher Cu affinity and thus better ability to counteract Cu effect, although lack of the same
pattern for compound 5 cannot be explained with the data at hand. Comparison of 1-5 with both
controls CQ and resveratrol leads to postulate that our ionophoric polyphenols prevent A(1-40)
aggregation by mechanisms different from both, and that the observed anti-amyloidogenic effect
2
+
is probably consequence of Cu removal from A(1-40) as well as fibril destabilizing
2
+
interactions of 1-5 (or their respective Cu complexes) with the peptide. It is worth highlighting
that these results confirm the initial postulate of molecular design; compounds 1-5 were
conceived hypothesizing that the combination of both ionophoric and polyphenol moieties in one
single compound would result in a cooperative effect able to enhance the activity of the parent
drugs. Our results also show that no significant difference is observed within the three amino
compounds (1, 2 and 4), suggesting that the position and number of hydroxy groups in these
compounds do not play an important role in the observed activity. A final remark concerns with
2
+
increase of A(1-40) fibril concentration with time. It is theorized that Cu -induced A
oligomeric aggregates are not a final product and that those oligomeric forms can eventually
2
+
evolve to mature fibrils. 47% increase in Cu -induced A(1-40) fibrils formation between 3 and
days combined with a decrease of turbidity in the same time period seem to support this fact.
Importantly, this conversion is also inhibited by compounds 1-5.
7
2
.6. Binding affinity to A(1-40) and inhibition of self-induced A(1-40)aggregation
Molecular self-assembly is an ubiquitous process in biological systems and crucial to the
function of cells. Lipids assemble to form membranes, individual strands of nucleic acids do it to
form the double helix of DNA and the assembly of proteins results in quaternary structures with

specific functions. However, molecular self-assembly processes of inappropriately folded
peptides into amyloid fibers are responsible for the formation of insoluble protein aggregates that
have been associated with several serious human ailments, most especially, neurodegenerative
9
,10
2+
2+
diseases.
Because of the low Cu binding affinity and distinct Cu -induced aggregation
2
+
profile to CQ we have postulated mechanism(s) additional to ion chelation to explain the Cu -
induced aggregation inhibition of compounds 1-5. These observations, therefore, provided the
rationale to further investigate the ability of our ionophoric polyphenols to interact with A(1-
4
0) and inhibit its self-induced aggregation in absence of metal ions.
The interaction of 1-5, CQ and resveratrol with A(1-40) was studied by fluorimetric titration,
monitoring the quenching of A(1-40) tyrosine fluorescence upon consecutive additions of
compound solutions. Results of binding affinity for 1-5, collected in Table 2, are comparable to
8
4
those described for other small ionophoric compounds and similar to both controls CQ and
resveratrol despite their different A(1-40) aggregation inhibition profile. These observations
taken together suggest that the compound-protein interactions although of same magnitude must
be of different nature.

A
250
3
7
days incubation
days incubation
200
150
100
50
0
Ab
CQ
Resv
1
2
3
4
5
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
B
3 days incubation
7
days incubation
1
8 days incubation
Abs550 nm
Ab
CQ
Resv
1
2
3
4
5
Figure 5. Inhibition of self-induced A(1-40) aggregation by compounds 1-5, CQ and
resveratrol studied by ThT (A) and turbidity (B) assays. No significant changes were observed in
ThT fluorescence after 18 days of incubation, and, therefore, that data is not shown for clarity.
All values of fluorescence and percent inhibition of fibril aggregation can be found in Table S2
in Supplementary Information.

In a different set of experiments, the ability of 1-5, resveratrol and CQ to inhibit self-induced
A(1-40) aggregation was studied by the ThT and turbidity assays. A solution of A(1-40) was
incubated for 3-18 days in absence or presence of 2 eq. of compounds 1-5, resveratrol or CQ at
2
+
pH=6.60 for comparison purposes with the Cu -induced experiment. Results in Figure 5 show
that compounds 1-5 inhibit between a ~77% and a ~96% extent the A(1-40) self-assembly into
fibril aggregates in the time frame studied (see Table S2 in Supplementary Information for all
values of fluorescence and percent inhibition). In general, results for compounds 1-5 in absence
2
+
2+
of Cu display consistent trends to those in presence of Cu , confirming that the activity of the
ionophoric polyphenols must be explained beyond metal chelation. In our interpretation of
turbidity and ThT results, and similarly to what we previously discussed, resveratrol and the
ionophoric polyphenols seem to destabilize fibril formation by initially re-directing aggregation
7
9
to non-toxic high molecular weight amorphous populations that end up re-dissolving upon
long-term exposure to 1-5, which translates into an overall decrease of the aggregate population.
2
+
Importantly, the fact that compound 3 does not show a distinctive behavior in absence of Cu
2
+
ions could confirm that the pattern observed in Cu -containing experiments might be partly due
to its stronger metal ion affinity. Interestingly, CQ again stimulates the conversion to misfolded
structures of A(1-40) peptide in both amorphous and fibril forms of the protein, as reported by
8
0,85
others,
and consistently proves to have poor anti-amyloidogenic properties. On the other
2
+
hand, resveratrol follows a parallel pattern to that observed in presence of Cu , suggesting that
its effect on A(1-40) aggregation is metal ion independent, in good agreement with its low
metal binding affinity. The effect of resveratrol, curcumin and some other polyphenols on A
3
7,73,79
aggregation has been previously reported
and generally agree with our findings.

2
.7. Disaggregation of pre-formed A(1-40) aggregates
7
9
It has been described that resveratrol, curcumin and some flavonoids seem to exert their anti-
amyloidogenic activity by preferably binding amyloid fibrils rather than primary monomeric A
structures and, therefore, the ability of curcumin -and other drugs like CQ- to promote A
7
3,86,87
disaggregation of pre-formed aggregates at different levels has been studied.
In line with
this, we have observed that compounds 1-5 suppress the formation of A(1-40) fibrils but they
require long times of incubation to produce an inhibitory effect on amorphous aggregates,
suggesting that these potential drugs might also interact preferably with some type of pre-formed
A aggregates or with certain structural patterns present only in mature fibrils. In this context,
thus, it was relevant to test the ability of 1-5 to disaggregate mature A(1-40) forms of the
2
+
peptide after aggregation had been initiated both in presence and absence of Cu ions.
2
+
First, A(1-40) peptide was incubated for 7 days with 1 eq. of Cu to allow aggregate formation.
After that time, amorphous and fibril A(1-40) aggregates were measured by turbidity and ThT
assays, followed by addition of 2 eq. of compounds 1-5 and incubation for 12 more days. Results
2
+
are depicted in Figure 6. Cu -A(1-40) control behaved as previously observed in section 2.5.
Fluorescence ThT assay confirms that addition of CQ also promotes fibril aggregation under
these experimental conditions. On the other hand, results on turbidity for CQ do not significantly
differ from the control, which in turn translates in an overall increase of the aggregate
population. These results for CQ are consistent with all our observations although they contrast
8
5
2+
with some reported studies. Interestingly, resveratrol affects fibrilization of pre-formed Cu -
A(1-40) aggregates only after long periods of incubation, while not changing the turbidity
profile with respect to the control. The ionophoric polyphenols display different trends.

2
+
Compounds 1 and 2 show a decrease in turbidity similar to that of the Cu -A(1-40) control,
while stopping the aggregation into mature fibrils, resulting in a lower overall amount of
aggregates. Best results are obtained for compound 3, which actually shows ability to
2
+
disaggregate up to a ~57% extent Cu -A(1-40) fibrils and lowers up to a ~22% the total
turbidity of A(1-40) aggregates after 12 days incubation, while compounds 4 and 5 seem to be
redirecting aggregation from mature fibrils to amorphous aggregates (see Table S3 in
Supplementary Information for all values of fluorescence and percent inhibition).
Disaggregation experiments of self-assembled A(1-40) aggregates were carried out following
2
+
2+
the same experimental conditions as for the Cu -induced experiment, but lacking Cu ions.
Results are shown in Figure 7 and reveal that fibrilization is remarkably reduced between ~73%
and ~91% extent after 4 or 12 days of treatment with compounds 1-5 (see Table S4 in
Supplementary Information for all values of fluorescence and percent inhibition). Turbidity
experiments are consistent with our previous observations. Resveratrol and compounds 1-5 seem
to remodel mature A fibrils into high molecular weight oligomeric amorphous aggregates, as
7
9
previously reported, which also disaggregate when exposed to the ionophoric polyphenols 1-5
for long periods of time (i.e. ~46% turbidity reduction when treated with compound 2, ~25%
when treated with compound 3 for 12 days) without raising the amount of mature fibrils.
Therefore, ThT and turbidity assays taken together indicate a remarkable overall decrease of pre-
formed A(1-40) aggregates upon treatment with ionophoric polyphenols 1-5. Comparison with
clinically tested controls CQ and resveratrol show a significant better anti-amyloidogenic profile
for our family of polyphenols. Lastly, we speculate that differences observed in presence and
2
+
absence of Cu ions might be a consequence of competition between, at least, three types of