Catechol-Based Ligands as Potential Metal Chelators Inhibiting Redox Activity in Alzheimer's Disease

Article abstract of DOI:10.1002/ejic.201700385

The loss of homeostasis of redox metal ions, namely, copper and iron, has been considered to be an important contributing factor in neuronal death observed in the brain of patients affected by Alzheimer's disease (AD). Specific ligands able to regulate the homeostasis of copper and, to a lesser extent, iron ions in the brain have therefore been considered to be potential drugs for the treatment of AD. Herein, the synthesis and metal coordination properties of a series of ligands based on a bis(2,3-dihydroxyalkylbenzamide) scaffold, which are potentially able to regulate the homeostasis of redox metal ions, are reported. These catechol-based ligands exhibit high specific affinities for Cu^II and Fe^III, and a lower affinity for Zn^II, which indicates that they may interact with redox metal ions without disturbing the structural and dynamic role of zinc chelation in many proteins. These ligands inhibit the reduction of dioxygen induced by Cu^II–Aβ_1–16, and this suggests that they may be able to efficiently reduce the oxidative stress induced by metal-loaded amyloids in the AD brain.

Full text of DOI:10.1002/ejic.201700385

A Journal of
Accepted Article
Title: Catechol-based ligands as potential metal chelators inhibiting
redox activity in Alzheimer's disease
Authors: Anne Robert
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201700385
Link to VoR: http://dx.doi.org/10.1002/ejic.201700385

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
Catechol-based ligands as potential metal chelators inhibiting redox activity
in Alzheimer's disease
Michel Nguyen,[a] Bernard Meunier,[a,b]* Anne Robert,[a]*
[a] Laboratoire de Chimie de Coordination du CNRS, 205 route de Narbonne, BP 44099,
31077 Toulouse cedex 4, France,
[b] School of Chemical Engineering and Light Industry, Guangdong University of Technology
(GDUT), Higher Education Mega Center, 100 Waihuan Xi road, Panyu District, Guangzhou,
510006, P. R. China.
Corresponding authors: Bernard Meunier (bmeunier@lcc-toulouse.fr) and Anne Robert
(anne.robert@lcc-toulouse.fr). https://www.lcc-toulouse.fr/article432.html
___________________________________________________________________________
Key words
Alzheimer’s disease; Catechol, Copper; Iron; Polydentate ligands.
___________________________________________________________________________
Abstract
The loss of homeostasis of redox metal ions, namely copper and iron, has been considered as
an important contributing factor in neuronal death observed in the brain of patients affected
by Alzheimer’s disease (AD). Specific ligands able to regulate the homeostasis of copper and,
in a lesser extent, of iron ions in the brain have therefore been considered as potential drugs
for the treatment of AD. We report here the synthesis and metal coordination properties of a
series of ligands based on a bis(2,3-dihydroxyalkylbenzamide) scaffold, which are potentially
able to regulate the homeostasis of redox metal ions. These catechol-based ligands exhibit
high specific affinities for Cu(II) and Fe(III), and a lower affinity for Zn(II), indicating that
they may interact with redox metal ions, without disturbing the structural and dynamic role of
zinc chelation in many proteins. These ligands inhibit the reduction of dioxygen induced by
CuII–Aβ1-16, suggesting that they might be able to efficiently reduce the oxidative stress
induced by metal-loaded amyloids in AD brain.
1
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
Introduction
The loss of homeostasis of redox metal ions, namely copper and iron, has been considered as
important in neuronal death observed in brains of patients affected by Alzheimer’s disease
(AD). Specific ligands able to regulate the homeostasis of copper and, in a lesser extent, iron
ions in the brain have therefore been considered as potential drugs for the treatment of AD.[1-5]
In addition, the disease is associated with degeneration of central cholinergic tracts. There are
substantial evidences that the nerve growth factor (NGF) has a APP-mediated role in
promoting survival and growth of central cholinergic neurons both during development and in
regeneration after injury.[6] NGF signaling directly controls the APP phosphorylation, BACE
cleavage and, finally, amyloid-β accumulation (APP: amyloid precursor protein, BACE: beta-
(a)
site amyloid precursor protein cleaving enzyme).[7]
Administration of NGF has therefore been considered to rescue
R
O
(CH2)n
HO
impaired neuronal response due to Alzheimer's disease.[8]
Efforts have been made to find small molecules that have
neurotrophic properties similar to those of NGF and/or can
enhance the activity of endogenous neurotrophic factors.
Among them, gentisides (Figure 1a) are naturally occurring
based on a long chain 2,3-dihydroxyalkylbenzoate scaffold.
These molecules, recently isolated from the traditional Chinese
medicine Gentiana rigescens Franch., exhibit a significant
neuritogenic activity against pheochromocytoma PC12
cells.[9] For this reason, this motif is of potential interest for
the therapy of neurodegenerative diseases, and series of
OH
O
R =
(b)
HO
Gentisides, n = 14-18
X = O or S
CH3
X
(CH2)n
OH
O
Synthetic 2,3-dihydroxybenzoates
and 2,3-dihydroxythiobenzoates
Figure 1. Structures of (a)
gentisides, and (b) synthetic 2,3-
dihydroxyalkylbenzoate- and 2,3-
dihydroxyalkylthiobenzoate ligands.
OH
HO
4
3
synthetic
2,3-dihydroxybenzoates
and
2,3-
5
2
OH
HO
6
1
H
N
H
N
dihydroxythiobenzoates were synthesized. Their activity as
neurite outgrowth inducers, compared to that of NGF, was found
promising in vitro.[10,11]
R
O
O
R =
Ligand
(CH2)7
(CH2)2–O–(CH2)2
(CH2)2–O–(CH2)2–O–(CH2)2
1
2
3
As an attempt to combine these two properties, regulation of
Figure 2. Structure of the
bis(2,3-dihydroxyalkyl-
benzamides) ligands 1-3.
homeostasis of Cu(II) and Fe(III) which are the two main redox
metals in vivo, and promotion of neurite outgrowth, we designed
2
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
specific ligands based on a bis(2,3-dihydroxyalkylbenzamide) scaffold (Figure 2). These
ligands should have a higher affinity for copper and iron ions to avoid any perturbation of the
interactions of zinc ions that have a key structural role in many biomolecules (allosteric
interactions, zinc-finger proteins, ...).
Results and Discussion
Synthesis and characterization of bis(2,3-dihydroxyalkylbenzamide) ligands 1-3
In the series of 2,3-dihydroxybenzoate and 2,3-dihydroxythiobenzoate (Figure 1b), their
neurogenic activity does not depend on the structural diversity at the end of the alkyl chain,
nor is decisive the benzoate or thiobenzoate functions. Contrary, the chain length is important
for activity. In the benzoate series, the higher activity was found for derivatives with alkyl
chains containing 8 to 16 carbon atoms (i.e. 13-21 bonds between the two catechol
rings).[10,11] So, in order to get ligands able (i) to chelate with high affinity the biologically
relevant redox metal ions Cu2+ and Fe3+, and (ii) to promote a NGF like activity, we designed
and synthesized a series of tetradentates bis-2,3-dihydroxyalkylbenzamide derivatives having
10-13 bonds between
(i)
(ii)
OMe
OMe
OH
HO
BnO
HO
the two catechol rings
(Figure 2). Ligands 1
and 3 were obtained by
OH
O
OBn
O
OH
O
5
4
(iii)
(v)
(iv)
1
H
N
H
N
OH
3
coupling
of
1,7-
BnO
R
OBn
BnO
OBn
O
O
OBn
OBn
O
R =
(CH2)7
(CH2)2–O–(CH2)2–O(CH2)2
6
diaminoheptane or 2,2’-
(ethylenedioxy)bis(ethyl
amine), respectively,
7
8
Scheme 1. Synthesis of ligands 1 and 3 from 2,3-dihydroxybenzoic acid. (i) BF3 in
CH3OH, reflux, 5 h; (ii) K2CO3/NaI, BnBr in DMF, RT, overnight; (iii) KOH in H2O,
90 °C, 2 h; (iv) CDI, 1,7-diaminoheptane (for 1) or 2,2’-
(ethylenedioxy)bis(ethylamine) (for 3) in CH2Cl2, RT, overnight; (v) H2/Pd-C/EtOH.
Pd-C. in C2H5OH.
with
benzyloxybenzoic acid
6, in the presence of
carbonyldiimidazole (CDI)
2,3-
H
N
H
N
(i)
OH
HO
R
OH
HO
OH
O
O
OH
OH
O
2
(Scheme 1).[12] Deprotection of
the tetrabenzylated coupling
products 7 and 8 was achieved
R = (CH2)2–O–(CH2)2
Scheme 2. Synthesis of ligand 2 from 2,3-dihydroxybenzoic acid.
(i) HOBt, DCC in THF, 2,2’-oxybis(ethylamine), RT, overnight.
3
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
by catalytic hydrogenation, providing ligands 1 and 3, with overall yields of 47% and 36%,
respectively, from the starting 2,3-dihydroxybenzoic acid.
Ligand 2 was obtained by direct coupling of two equivalents of 2,3-dihydroxybenzoic acid
with 2,2’-oxybis(ethylamine), in the presence of dicyclohexylcarbodiimide, with a moderate
yield (11%, Scheme 2).
Complexation of ligands 1-3 by Cu(II)
Complexation of ligands 1-3 by copper(II) was achieved by addition of aliquots of CuCl2 until
to 20 mole equivalents with respect to
231
0,2
0,8
the ligand, and was monitored by UV-
0,15
0,7
visible spectrometry as previously
0,1
0,6
reported.[13] The same behavior was
0,05
0,5
0
observed for ligands 1-3. As a matter of
0,4
0
0,5
1
1,5
2
2,5
Cu2+/1 molar ratio
example, results obtained with ligand 1
are depicted in Figure 3. The
absorbances at 315-317 nm of free
ligands (316 nm for 1, Figure 3, dashed
trace) disappeared in favor of a band at
341-344 nm (342 nm for Cu-1)
assigned to the MLCT transition of the
complexes. Spectral modifications were
immediately observed from 0.25 to 1 mole
equivalent of Cu2+, as the result of a fast
complexation. Detection of well defined
isosbestic points at 214-217 nm and 317-318
nm is in agreement with the formation of
one type of Cu(II) complex with each ligand.
Saturation was obtained with one molar
equivalent of Cu2+, indicating that the
stoichiometry of the copper complexes
formed is 1:1 (Figure 3, Insert).
0,3
0,2
0,1
0
342
316
252
210
260
310
360
Wavelength, nm
Figure 3. Titration of ligand 1 (15 µM) from 0 (dashed trace)
to 1.75 mole equivalent of CuCl2, in Tris-buffered saline pH
7.4/methanol, 1/1, v/v. Insert: Absorbance at 342 nm against
the Cu2+/1 mole ratio.
0,18
0,16
343 nm
0,14
0,12
314 nm
0,1
0,08
0,06
0,04
0,02
0
374 nm
260 280 300 320 340 360 380 400 420
Wavelength, nm
Figure 4. Determination of log Kapp of CuII-1 by using
UV-visible spectrometry in Tris-buffered saline pH
7.4/methanol, 1/1, v/v. Dotted line: ligand 1, full line:
Cu2+/1 = 1/1, dashed line: Cu2+/1/EDTA = 1/1/1.
4
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
The affinity constants of Cu(II) for ligands 1-3 were measured by UV-visible spectrometry,
using EDTA as competitive chelator, as previously reported.[14] In a solution containing
equimolar amounts of ligand L, CuCl2, and EDTA (15 µM each), in Tris buffered saline pH
7.4/methanol, 1/1, v/v. Log Kapp [CuII-L] was calculated from log Kapp [CuII-EDTA] in these
conditions (15.9)[15] and the proportion of CuII-L evaluated at equilibrium. Measurements
were made at 374 nm because neither free ligand L, nor EDTA or Cu-EDTA absorb at this
wavelength. The proportion of ligand coordinated to Cu(II) was (56.5 ± 4)% for ligand 1
(Figure 4), (77.6 ± 0.5)% for ligand 2, and (36.6 ± 2)% for ligand 3 (not shown). So,
calculated log Kapp [CuII-L] were 16.2 ± 0.1 for L=1, 17.0 ± 0.1 for L=2, and 15.4 ± 0.1 for
L=3. These values are about 5-6 orders of magnitude higher than the reported affinities of
copper for β-amyloids (log Kapp=10-11),[16] and close to the affinities of other tetradentate
copper specific ligands.[14]
Complexation of ligands 1-3 by Fe(III)
Complexation of ligands 1-3 by Fe(III) was monitored by addition of aliquots of FeCl3 in a
solution of ligand, in the same way as reported above for Cu(II). Upon metallation, the
absorbance of the free ligand at 316 nm underwent a bathochromic and hyperchromic shift.
The results obtained with ligands 1 and 2 were the same, and the results obtained with ligand
1 are depicted in Figure 5. The maximum wavelength was 330 nm after addition of 2 mole
equivalents of iron salt (Figure 5a), and 337 nm after addition of 20 mole equivalents of Fe3+
(b)
0,16
(a)
0,16
0,14
0,12
0,10
0,08
0,06
0,04
0,02
0,00
330
Fe3+/1 mol ratio =
1.0
0,15
0,14
0,13
0,12
0,11
0,10
0,09
0,08
0.7
316
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
270
290
310
330
350
370
390
Fe3+/1 molar ratio
Wavelength, nm
Figure 5. (a) Titration of ligand 1 (15 µM) from 0 (dashed trace) to 1.2 mole equivalent of FeCl3, in Tris-
buffered saline pH 7.4/methanol, 1/1, v/v. (b) Absorbance at 333 nm against the Fe3+/1 mole ratio.
5
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
(data not shown). The plot of absorbance at 333 nm against the Fe3+/1 molar ratio exhibit a
slight but significant change of slope for the Fe3+/1 molar ratio = 0.65-0.7 (Figure 5b). This
result evidenced the rapid formation of a complex with a 3/2 L/Fe stoichiometry (L = 1 or 2),
suggesting three catechol residues involved in the octahedral coordination sphere of each Fe
ion. Upon further addition of FeCl3, several other iron complexes may exist, as evidenced by
the plot of absorbance at 333 nm against the Fe3+/1 molar ratio until Fe3+/1 = 6 (Figure S1).
Upon metallation of ligand 3 by Fe3+, the plot of absorbance at 323 nm against the Fe3+/3
molar ratio is a broken line (Figure S2) which reveals that several FeIII –(1)m complexes exist
n
or coexist, even at low Fe3+/3 ratio. In addition, in the presence of Fe3+/L mol ratios higher
than 1/1, the precipitation of Fe(II) oxydes/hydroxydes is likely to occur. This complex
situation was not investigated further.
We did not measure the affinity constant of Fe(III) for ligands 1 and 2 with the method
used above for Cu(II). In fact, the equations used to calculate log Kapp by this method are
accurate only for metal complexes with stoichiometry L/M = 1/1, and that is not the case for
the Fe(III) complexes of 1 and 2. However, the cumulative constants log β3 of FeIII(catechol)3
complexes have already been reported to be as high as 40.[17]
Complexation of ligands 1-3 by Zn(II)
Titration of ligands 1-3 (15 µM) by Zn2+ was achieved in a similar way, and monitored by
UV-visible. In all the three cases, the absorbances of the free ligands at 228-232 nm and 341-
344 nm increased upon addition of ZnCl2 (0-20 mole equivalents), with well-defined
isosbestic points at 216, 249-250, 254-259, 283 and 317 nm, for the ligands 1-3. Twenty mole
equivalents of Zn2+ were required to saturate ligand 2, indicating that free ligand 2 and its
Zn(II) complex were in equilibrium in the solution (Figure 6). The plot of absorbance at 334
nm against the Zn2+/2 molar ratio indicated that half metallation occurred when 1 mole
equivalent of Zn2+ has been introduced (Figure 6, Insert). So, calculated from this titration,
log Kapp affinity [Zn(II)-2] was 5.1.
6
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
0,18
1
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
0,16
0,14
0,12
0,1
216
½
metallation
250
254
0
1
2
337
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
Zn2+/2 molar ratio
316
283
317
310
210
260
360
Wavelength, nm
Figure 6. (a) Titration of ligand 2 (15 µM) from 0 (dashed trace) to 20 mole equivalent of ZnCl2, in Tris-
buffered saline pH 7.4/methanol, 1/1, v/v. (b) Absorbance at 334 nm against the Zn2+/2 mole ratio.
Ligands 1 and 3 were not saturated upon addition of 20 mole equivalents of Zn2+.
Consequently, the plots of absorbance against the Zn2+/L molar ratio indicated that more than
4.9 and 2.8 mole equivalents of Zn2+ were necessary to achieve half metallation of 1 and 3,
respectively (Figures S3). Then, the values of log Kapp [Zn(II)-L] were below 4.2 and 4.5, for
L = 1 and 3, respectively. So, the affinity of these ligands for Zn(II) is at least 11 orders of
magnitude below that of Cu(II) and Fe(III) at physiological pH. In addition, these values are
below the affinity of many proteins for Zn(II) [log Kapp (Zn-HSA) = 6-8],[18] suggesting that
the bis-catecholates should not interact with zinc proteins.
Redox activity of copper complexes of 1 and 2
We measured the consumption of ascorbate by Cu–Aβ1-16 in the presence of bis(2,3-
dihydroxyalkylbenzamide) ligands 1 and 2 as an indirect evaluation of their ability to reduce
dioxygen and, consequently, to induce an oxidative stress.[19,20] The short amyloid 1-16 is
usually considered as a reliable model for metal coordination by amyloids, because it contains
the coordination sites of metal ions (Cu, Fe, and Zn).[21] The assay was carried out in saline
Hepes buffer containing 1.05 vol% of DMSO, and the ascorbate consumption was measured
by monitoring UV absorbance at 265 nm. Results were the same for 1 and 2, and results
obtained with ligand 1 are depicted in Figure 7. The residual autoxidation of ascorbate was
3% in 5 min and 19% in 30 min (trace a). As expected, the ascorbate was extensively
7
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
oxidized in the presence of CuCl2 (trace
b), or Cu–Aβ1-16 (trace c) as a result of
the ability of Cu2+/Cu+ or CuII/I–Aβ1-16 to
1,6
1,4
1,2
1
Ascorbate consumption =
(e)
(a)
Cu–1, 1/1.1
< 3%
11%
19%
(f)
Ascorbate
Cu-Aβ + 1
catalytically
reduce
dioxygen.
0,8
0,6
0,4
0,2
0
Interestingly, in the presence of 1.1 mole
equivalent of ligand 1 with respect to Cu,
the ascorbate consumption was negligible
(< 3% in 30 min, trace e), thus confirming
that the copper complex of 1 was unable
to both reduce dioxygen and release
catalytically active copper ions.
Moreover, addition of 1 (1.1 mole
equivalent) to Cu–Aβ1-16 resulted in
(c)
Cu-Aβ, 1/1.1
(d)
Cu–1, 1/0.7
(b)
CuCl2
98%
0
5
10
15
Time, min
20
25
30
Figure 7. UV-visible (265 nm) kinetic spectra of ascorbate
consumption under air in the presence of (a) no additive,
(b) CuCl2, (c) CuII–Aβ1-16, 1/1, (d) CuII–1, 1/0.7, (e) CuII–1,
1/1.1, (f) CuII–Aβ1-16, 1/1.1, + 1 (1.1 mole equiv). Spectra
(c)-(f) were obtained after substraction of the substrate
absorbance at 265 nm. [Ascorbate] = 100 µM, [Cu] = 10
µM, [Aβ1-16] = 11 µM, [1] = 7 or 11 µM.
efficient inhibition of the ascorbate consumption (11% in 30 min, trace f), indicating that
ligand 1 was able to remove copper from Aβ1-16, to yield the redox inactive CuII–1 complex.
This behavior is consistent with a lower affinity of Cu(II) for Aβ (log Kapp [CuII–Aβ] = 10-
11)[22] compared to ligand 1, and suggests that 1 may be able to reduce oxidative stress in AD
brain. In addition, it can be noted that 0.7 mole equivalent of 1 with respect to Cu–Aβ1-16 was
not efficient to inhibit the ascorbate consumption (trace d). This confirms the 1/1 M/L
stoichiometry of the Cu–1 complex.
Conclusion
To regulate the homeostasis of redox metal ions in AD brains, without touching zinc which
has a structural role in many biomolecules, we proposed bis(2,3-dihydroxyalkylbenzamide)
ligands. The catechol residues of these ligands are able to selectively chelate redox active
metal ions compared to Zn(II), their affinity for Zn(II) being at least 11 orders of magnitude
below that of Cu(II) and Fe(III) at physiological pH [log Kapp ZnII-1 = 5, log Kapp CuII-1 = 16,
and log β3 expected for FeIII-1 around 40]. So, these ligands should be able to chelate Cu(II)
and Fe(III) without disturbing zinc-containing proteins. Moreover, these ligands were found
to inhibit the reduction of molecular oxygen induced by CuII–Aβ1-16, suggesting that they may
have a benefit activity against oxidative stress induced by metal loaded amyloids in AD brain.
8
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
In addition, the Cu(II) complexes of these catecholate ligands were unable to oxidize
ascorbate by their own. This should be due to the an
R
R
electron transfer from the catecholate ligand to the metal
center, giving rise to a reduced metal-semiquinonate
complex (Scheme 3).[23] This equilibrium, which is in fact
an intramolecular oxido-reduction, may render these
complexes unable to reduce an extra species such as
dioxygen. In the case of copper, the reduced metal-
semiquinonate form is favored by other electron
O
O
O
O
Mn+
M(n-1)+
Oxidized metal-
catecholate
Reduced metal-
semiquinonate
Scheme 3.
Possible equilibrium
between the η2-catecholate and
semiquinonate forms of iron and
copper complexes.
donating ligands bound to copper, which is obviously the case in the bis(catecholate)-copper
complexes of 1-3.[24,25] This feature should also account for the antioxidant properties of
natural polyphenols, a series considered for therapeutic potential against AD.[26] In addition, it
was recently reported that dysfunction of dopamine, with its 4-substituted catechol structure,
may be involved in AD, and that dopamine agonists may have beneficial effects against the
cognitive decline.[27]
In addition, the calculated logP value of ligand 1 was 2.3-2.4, a conducive condition
for blood-brain barrier penetration (for several classes of CNS active substances, BBB
penetration was found optimal when logP values were in the range 1.5-2.7, with the mean
value of 2.1).[28]
Experimental Section
Materials and instruments
All solvents and commercially available reagents were purchased from Sigma-Aldrich, Fluka,
Acros, or TCI, except amyloid-β1-16 (purchased from Bachem). They were used without
further purification. Reactions were monitored by thin layer chromatography (TLC), using
silica gel-coated aluminium plates with fluorescent indicator 254 nm (Fluka). Preparative
chromatography was performed on silica gel column (Fluka 63-200 µm) or using a
CombiFlash Companion (Teledyne Isco) with RediSep silica flash columns (35-70 µm). Aβ1-
16 was dissolved in DMSO and the concentration was measured by UV-visible spectroscopy
[ε276 nm (Tyr10) = 1410 M-1 cm-1].[22a]
9
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
UV-visible spectra were recorded on Agilent 8453 or Specord 205 (Analytik Jena)
1
13
spectrophotometers. H- and C-NMR spectra spectra were recorded on a Brucker Avance
400 spectrometer using tetramethylsilane (TMS) as the external standard. Mass spectra were
obtained on ThermoFisher (DCI) or GCT Premier spectrometer (DCI, CH4, HRMS), or Xevo
G2 QTOF (Waters, ESI+, HRMS).
LogP values of the ligands were calculated using ChemDraw Prime 16.0.0.82.
Syntheses
Compound 4. A solution of BF3 1.5 M in methanol (15 mL) was added to a solution of 2,3-
dihydroxybenzoic acid (5 g, 32 mmol) in anhydrous methanol (100 mL) and the mixture was
heated under argon, at reflux for 5 h. After evaporation of MeOH under reduced pressure,
cold water (200 mL) was added and the pH was adjusted to 6 with a 4 M aqueous NaOH
solution. The aqueous phase was extracted with CH2Cl2 (3 x 250 mL). The organic phases
were dried over Na2SO4, filtered, and the solvent was removed under reduced pressure to give
1 as a white solid (4.29 g, 79%). 1H NMR (400.0 MHz, CDCl3): δ = 3.98 (s, 3H, O–CH3),
5.66 (s, 1H, C3–OH), 6.82 (t, 3J = 8.0 Hz, 1H, H5), 7.13 (d, 3J = 8.0 Hz, 1H, H6), 7.39 (dd, 3J
4
13
= 8.0 Hz, J = 1.6 Hz, 1H, H4), 10.91 (s, 1H, C2–OH). C NMR (100.6 MHz, CDCl3) δ =
52.4 (O–CH3), 112.4 (C1), 119.2 (C6), 119.9 (C4), 120.6 (C5), 145.0 (C3), 148.8 (C2), 170.8
(CO). MS (DCI, NH3): m/z = 169 [M+H]+.
Compound 5. Compound 4 (6 g, 35.7 mmol), anhydrous K2CO3 (19.7 g, 142.7 mmol), and
NaI (0.32 g, 2.14 mmol) were dissolved in dry DMF (50 mL) under argon. To this solution,
benzyl bromide (8.6 mL, 71 mmol) in DMF (10 mL) was slowly added. The reaction mixture
was stirred overnight and filtered. The filtrate was diluted with Et2O (250 mL), washed with
2% NaOH aqueous solution, and then water. The ether layer was dried over Na2SO4, filtered
and evaporated under reduced pressure to give a yellow oil. The product was precipitated
from ethyl acetate/hexane as a white solid (10.3 g, 83%). 1H NMR (400.0 MHz, CDCl3): δ =
3.87 (s, 3H, O–CH3), 5.14 (s, 2H, C3–O–CH2–C6H5), 5.17 (s, 2H, C2–O–CH2–C6H5), 7.10 (t,
J = 8.0 Hz, 1H, H5), 7.17 (dd, 3J = 8.0 Hz, 4J = 1.6 Hz, 1H, H6), 7.29-7.52 [m, 11H, H4 + (2
13
x C6H5)]. C NMR (100.6 MHz, CDCl3) δ = 52.1 (O–CH3), 71.3 (C6H5–CH2–OC3), 75.7
(C6H5–CH2–OC2), 118.1 (C6), 122.9 (C4), 124.0 (C5), 126.9 (C1), 127.5 (2 x o-Cbenzyl–C3),
10
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
127.93 (p-Cbenzyl–C3), 128.1 (p-Cbenzyl–C2), 128.3 (2 x o-Cbenzyl–C2), 128.6 (2 x m-Cbenzyl–C2
+ 2 x m-Cbenzyl–C3), 136.6 (ipso-Cbenzyl–C3), 137.4 (ipso-Cbenzyl–C2), 148.3 (C3), 152.8 (C2),
166.8 (CO). HRMS (DCI, CH4) m/z calculated for [C22H20O4+H]+ = 348.1362, found
348.1362.
Compound 6. A solution of 40% KOH in water (30 mL) was added to a solution of 5 in THF
(3 mL). After stirring for 2 h at 90°C, the mixture was acidified by aqueous 10% HCl,
extracted with ethyl acetate, dried over Na2SO4, filtered and evaporated under reduced
pressure to give 6 as a white solid (1 g, 95%). 1H NMR (400.0 MHz, CDCl3): δ = 5.22 (s, 2H,
C3–O–CH2–C6H5), 5.29 (s, 2H, C2–O–CH2–C6H5), 7.22 (t, 3J = 8.0 Hz, 1H, H5), 7.28 (dd, 3J
= 8.0 Hz, 4J = 1.6 Hz, 1H, H6), 7.32-7.55 (m, 10H, 2 x C6H5), 7.77 (dd, 3J = 8.0 Hz, 4J = 1.6
Hz, 1H, H4). 13C NMR (100.6 MHz, CDCl3) δ = 71.6 (C3–O–CH2–C6H5), 77.2 (C2–O–CH2–
C6H5), 119.0 (C6), 123.0 (C1), 124.5 (C4), 125.1 (C5), 127.8 (2 x o-Cbenzyl–C3), 128.6 (p-
Cbenzyl–C3), 128.8 (2 x o-Cbenzyl–C2), 128.9 (2 x m-Cbenzyl–C3), 129.3 (2 x m-Cbenzyl–C2),
129.3 (p-Cbenzyl–C2), 134.6 (ipso-Cbenzyl–C3), 135.8 (ipso-Cbenzyl–C2), 147.1 (C3), 151.3 (C2),
165.0 (CO). MS (DCI, NH3) m/z = 335 [M+H]+.
Compound 7. Compound 7 was prepared by the method described by Joshua et al.[12]
Compound 6 (1.3 g, 3.89 mmol) and carbonyldiimidazole (CDI, 0.63 g, 3.89 mmol) were
dissolved in dry methylene chloride (8 mL) under argon, at room temperature. The reaction
mixture was stirred for 1 h, and 1,7-diaminoheptane was added (295 µL, 1.94 mmol). After
stirring at room temperature for 24 h, the solvent was removed under reduced pressure, and
the residue was taken up in ethyl acetate (50 mL), washed with 2% NaOH, water, and brine,
dried over Na2SO4, and filtered. The solvent was evaporated to provide the yellow oil, which
was purified by column chromatography on silica gel (10 % hexane in ethyl acetate).
1
Evaporation of the eluate to dryness provided 7 as a white solid (1.29 g, 87%). H NMR
(400.0 MHz, CDCl3): δ = 1.15 (br, 6H, 3 x CH2), 1.24-1.38 (m, 4H, 2 x NH–CH2–CH2),
3.21-3.32 (m, 4H, 2 x NH–CH2), 5.09 (s, 4H, C3–O–CH2–C6H5), 5.17 (s, 4H, C2–O–CH2–
C6H5), 7.13-7.20 (m, 4H, 2 x H5 and 2 x H6), 7.30-7.52 (m, 20H, 4 x C6H5), 7.77 (2 x d, 3J =
8.0 Hz, 2H, 2 x H4), 7.93 (br t, 3J = 5.4 Hz, 2H, 2 x NH). 13C NMR (100.6 MHz, CDCl3) δ =
26.9 (2 x CH2), 29.0 (1 x CH2), 29.2 (2 x CH2), 39.7 (2 x NH–CH2), 71.3 (2 x C3–O–CH2–
C6H5), 76.4 (2 x C2–O–CH2–C6H5), 116.8 (2 x C6), 123.4 (2 x C4), 124.4 (2 x C5), 127.5 (2
x C1), 127.7 (4 x o-Cbenzyl–C3), 128.3 (2 x p-Cbenzyl–C3), 128.6 (4 x o-Cbenzyl–C2), 128.67 (4 x
11
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
m-Cbenzyl–C3 + 2 x p-Cbenzyl–C2), 128.72 (4 x m-Cbenzyl–C2), 136.39 (2 x ipso-Cbenzyl–C3),
136.44 (2 x ipso-Cbenzyl–C2), 146.7 (2 x C3), 151.7 (2 x C2), 164.9 (2 x CO). HRMS (ESI+)
m/z calculated for [C49H50N2O6+H]+ = 763.3747, found 763.3763.
Compound 1. Compound 7 (370 mg, 0.485 mmol) and 10% Pd-C (20% w/w) were suspended
in degassed ethanol (20 mL) and the mixture was purged with hydrogen and vigorously
stirred for 90 min under hydrogen pressure (3 bars). The Pd-C was filtered through celite and
the solvent was removed under vacuum. The obtained brown residue was taken up in acetone
(6 mL) and the solution was poured in diethyl ether mixture (120 mL). The precipitate was
discarded by filtration through a polyester syringe filter (Chromafil® GF/PET-45/25). After
1
evaporation of the solvent, the product 1 was obtained as a white solid (170 mg, 87%). H
NMR (400.0 MHz, CDCl3): δ = 1.36-1.48 (m, 6H, 3 x CH2), 1.65 (m, 4H, 2 x NH–CH2–
CH2), 3.47 (m, 4H, 2 x NH–CH2), 5.86 (s, 2H, 2 x HO–C3), 6.37 (br, 2H, 2 x NH), 6.77 (t, 3J
= 8.0 Hz, 2H, 2 x H5), 6.90 (dd, 3J = 8.0 Hz, 4J = 1.2 Hz, 2H, 2 x H6), 7.06 (dd, 3J = 8.0 Hz,
4J = 1.2 Hz, 2H, 2 x H4), 12.77 (br, 2H, 2 x HO–C2). 13C NMR (100.6 MHz, CDCl3) δ = 26.7
(2 x CH2), 28.8 (1 x CH2), 29.3 (2 x CH2), 39.6 (2 x NH–CH2), 114.0 (2 x C1), 115.7 (2 x C6),
118.0 (2 x C4), 118.6 (2 x C5), 146.0 (2 x C3), 149.1 (2 x C2), 170.0 (2 x CO). HRMS (DCI,
CH4) m/z calculated for [C21H26N2O6+H]+ = 403.1869, found 403,1871.
Compound 2. 2,3-Dihydroxybenzoic acid (0.5 g, 3.24 mmol) and 1-hydroxybenzotriazole
(HOBt, 438 mg, 3.24 mmol) were dissolved in dry THF (10 mL) under argon. A solution of
N,N’-dicyclohexylcarbodiimide (DCC, 802 mg, 3.89 mmol) in dry THF (8 mL) was added
dropwise. After stirring at room temperature for 15 min, 2,2’-oxybis(ethylamine) (176 µL,
1.62 mmol) was added and the mixture was stirred overnight. Dicyclohexylurea was removed
by filtration and the solvent was evaporated. The residue was solubilised with ethyl acetate
and washed with a saturated aqueous solution of NaHCO3, water, dried over Na2SO4, and
filtered. The solvent was removed under vacuum and the product was purified by column
chromatography on silica gel (CH2Cl2/CH3OH, 95/5, v/v). Evaporation of the eluate to
1
dryness provided 2 as a beige hygroscopic powder (70 mg, 11%). H NMR (400.0 MHz,
CDCl3): δ = 3.60-3.80 (m, 8H, 2 x NH–CH2–CH2–O), 5.78 (s, 2H, 2 x HO–C3), 6.69 (br s,
3
3
4
2H, 2 x NH), 6.72 (tr, J = 8.0 Hz, 2H, 2 x H5), 6.88 (dd, J = 8.0 Hz, J = 1.2 Hz, 2H, 2 x
3
4
13
H6), 7.06 (dd, J = 8.0 Hz, J = 1.2 Hz, 2H, 2 x H4), 12.61 (br, 2H, 2 x HO–C2). C NMR
(100.6 MHz, CDCl3) δ = 39.3 (2 x NH–CH2), 69.5 (2 x CH2–O), 113.8 (2 x C1), 116.0 (2 x
12
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
C6), 118.2 (2 x C4), 118.7 (2 x C5), 145.9 (2 x C3), 149.1 (2 x C2), 170.1 (2 x CO). MS (DCI,
NH3): m/z = 377 [M+H]+. Elemental analysis for C18H20N2O7•0.12 CH2Cl2 (apparent MW =
386.5): calcd C 56.30, H 5.28, N 7.25; found C 56.34, H 5.27, N 6.82.
Compound 8. Compound 8 was prepared by the method described as reported above for
compound 7, but using 2,2’-(ethylenedioxy)bis(ethylamine) (285 µL, 1.94 mmol) instead of
1,7-diaminoheptane. [12] The obtained yellow oil was purified by column chromatography on
silica gel (ethyl acetate). Evaporation of the eluate to dryness provided 8 as a white solid (1.26
1
g, 83%). H NMR (400.0 MHz, CDCl3): δ = 3.34 (s, 4H, 2 x NH–CH2), 3.43 and 3.50 (2 x m,
2 x 4H, 2 x NH–CH2–CH2–O and O–CH2–CH2–O), 5.06 (s, 4H, 2 x C3–O–CH2–C6H5), 5.14
(s, 4H, 2 x C2–O–CH2–C6H5), 7.14 (m, 4H, 2 x H5 and 2 x H6), 7.25-7.49 (m, 20H, 4 x
C6H5), 7.71 (2 x d, 3J = 8.0 Hz, 2H, 2 x H4), 8.19 (br tr, 3J = 5.2 Hz, 2H, 2 x NH). 13C NMR
(100.6 MHz, CDCl3) δ = 39.5 (2 x NH–CH2), 69.5 (2 x NH–CH2–CH2–O), 70.0 (O–CH2–
CH2–O), 71.2 (2 x C3–O–CH2–C6H5), 76.1 (2 x C2–O–CH2–C6H5), 116.8 (2 x C6), 123.1 (2
x C4), 124.4 (2 x C5), 127.6 (2 x C1), 127.7 (4 x o-Cbenzyl–C3), 128.2 (2 x p-Cbenzyl–C3),
128.5 (4 x o-Cbenzyl–C2 + 2 x p-Cbenzyl–C2), 128.6 (4 x m-Cbenzyl–C3), 128.8 (4 x m-Cbenzyl
–
C2), 136.37 (2 x ipso-Cbenzyl–C3), 136.41 (2 x ipso-Cbenzyl–C2), 146.6 (2 x C3), 151.8 (2 x C2),
165.2 (2 x CO). HRMS (DCI, CH4) m/z calculated for [C48H48N2O8+H]+ = 781.3489, found
781.3494.
Compound 3. The deprotection of compound 8 was achieved by catalytic hydrogenation in a
similar way as reported above for compound 7. Compound 3 was obtained as a white powder
1
(70%). H NMR (400.0 MHz, CDCl3): δ = 3.60-3.75 (m, 12H, CH2), 5.88 (br, 2H, 2 x HO–
C3), 6.73 (tr, 3J = 8.0 Hz, 2H, 2 x H5), 6.84 (br tr, 2H, 2 x NH), 6.89 (dd, 3J = 8.0 Hz, 4J = 1.2
Hz, 2H, 2 x H6), 7.04 (dd, 3J = 8.0 Hz, 4J = 1.2 Hz, 2H, 2 x H4), 12.68 (br s, 2H, 2 x HO–C2).
13C NMR (100.6 MHz, CDCl3) δ = 39.3 (2 x NH–CH2), 69.4 (2 x NH–CH2–CH2–O), 70.2
(O–CH2–CH2–O), 113.8 (2 x C1), 116.0 (2 x C6), 118.2 (2 x C4), 118.6 (2 x C5), 145.9 (2 x
C3), 149.1 (2 x C2), 170.0 (2 x CO). HRMS (ESI+) m/z calculated for [C20H24N2O8+H]+ =
421.1611, found 421.1605. Elemental analysis for C20H24N2O8•0.01 EtOH•0.27 CH2Cl2•0.36
hexane (apparent MW = 474.8): calcd C 56.79, H 6.29, N 5.90; found C 56.75, H 6.15, N
5.88.
Metallation of 1-3 by Cu(II), Zn(II), and Fe(III)
13
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
To a solution of 15 µM of 1, 2 or 3 in methanol/20 mM Tris•HCl pH 7.4, 150 mM NaCl
(1/1, v/v), aliquots of freshly prepared aqueous solutions of CuCl2, ZnCl2 or FeCl3 were added
(1 to 20 mol equiv). Stock solutions of ligands were 300 µM in methanol. Stock solution of
metal salts were 3.75 mM or 37.5 mM in H2O, for 0.25 to 3.5 mol equiv, or 5-20 mol equiv,
respectively, of metal ion with respect to ligand. Total variation of volume in the cuvette was
below 2% upon addition of 20 mol equiv of metal. UV-visible spectrum was recorded after
each addition of metal salt. Absorbance was plotted against the number of metal ion molar
equivalents.
Affinity of ligands 1, 2, and 3 for Cu(II).
The following mother solutions were prepared: (A) ligand L, 300 µM in CH3OH; (B)
EDTA, 300 µM in 20 mM Tris·HCl pH 7.4, 150 mM NaCl; (C) CuCl2, 15 mM in H2O. In the
UV-vis cuvette, 100 µL of (A), 900 µL of CH3OH, 100 µL of (B), 900 µL of 20 mM Tris·HCl
pH 7.4, 150 mM NaCl, and 2 µL of (C) were added in this order. Final concentrations of
ligand L, EDTA, and Cu2+ were 15 µM. Spectrum was recorded from 250 to 600 nm at 20 °C,
and compared to the spectra of L [2 µL of H2O instead of 2 µL of (C)], and of Cu2+-L [100 µL
of 20 mM Tris·HCl pH 7.4, 150 mM NaCl, instead of 100 µL of (B)]. All ligands and
complexes studied followed the Beer-Lambert law in the experimental conditions used. Log
Kapp [CuII-L] was calculated according to reference 14. Experiments were performed in
triplicate.
Affinity of ligands 1, 2 and 3 for Zn(II).
Log Kapp [ZnII-L] were calculated from the above mentioned titration of L by ZnCl2.
Since L and ZnII-L were in equilibrium in the UV-visible cuvette, log Kapp [Zn-L] can be
calculated from the number of mole equivalents of Zn2+ (n) required to achieve half-
metallation of L. Details of the calculation are below.
14
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Concentrations
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
t = 0
1.5 x 10-5 n x 1.5 x 10-5
[L]
+
[M]
[L-M]
0
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
at equilibrium
(half-reaction)
1.5 x 10-5
2
(n x 1.5 x 10-5)– 1.5 x 10-5 1.5 x 10-5
2
2
= 0.75 x 10-5
= (n – ½) x 1.5 x 10-5
= 0.75 x 10-5
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
At equilibrium, Kapp [Zn-L] = [L-M] / [L] [M] = 105 / 1.5 x (n – ½); where
n = number of mole equivalents of Zn2+ required for half-metallation of L.
⇒ log Kapp [Zn-L] = 4.82 – log(n – 0.5).
Oxidation of ascorbate monitored by UV-visible. Stock solutions were prepared as follows:
ligand L (3 mM in DMSO), Aβ1-16 (1.53 mM in DMSO) or CuCl2 (3 mM in H2O) and sodium
ascorbate (30 mM in H2O). For each experiment, a fresh solution of ascorbate was prepared.
The copper complexes CuII–L or CuII–Aβ1-16 were prepared by mixing adequate volumes of
the metal salt and ligand stock solutions, and incubated for 30 min. Where a ligand 1-3 was
added to CuII–Aβ1-16, the mixture was incubated for a further 30 min. Final concentrations in
the cuvette were: [CuII = 10 µM, [Aβ1-16] = 11 µM, [L] = 7 or 11 µM. The copper complexes
(or CuCl2, final concentration 10 µM, as a positive control) were then added in the UV
cuvette containing sodium ascorbate (final concentration: 100 µM) in Hepes buffer 50 mM,
pH 7.4, under continuous stirring, at 25 °C (final volume = 1.5 mL). The DMSO
concentration was below 1 vol%. Monitoring of ascorbate absorbance (λmax = 265 nm, ε =
14500 M-1 x cm-1)[20] was immediately started and continued over 30 min.
Acknowledgements
This work was supported by the CNRS and partially by GDUT (BM).
15
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
References
[1] A. Robert, Y. Liu, M. Nguyen, B. Meunier, Acc. Chem. Res. 2015, 48, 1332-1339, and
references therein.
[2] K. J. Barnham, A. I. Bush, Chem. Soc. Rev. 2014, 43, 6727-6749.
[3] M. A. Telpoukhovskaia, C. Orvig, Chem. Soc. Rev. 2013, 42, 1836-1846.
[4] L. R. Perez, K. J. Franz, Dalton Trans. 2010, 39, 2177-2187.
[5] H. Tamano, A. Takeda, Metallomics 2015, 7, 1205-1212.
[6] E. A. Milward, R. Papadopoulos, S. J. Fuller, R. D. Moir, D. Small, K. Beyreuther, C.
L. Masters, Neuron 1992, 9, 129-137.
[7] V. Triaca, V. Sposato, G. Bolasco, M. T. Ciotti, P. Pelicci, A. C. Bruni, C. Cupidi, R.
Maletta, M. Feligiono, R. Nisticò, N. Canu, P. Calissano, P. Aging Cell. 2016, 15, 661-
672.
[8] R. De Rosa, A. A. Garcia, C. Braschi, S. Capsoni, L. Maffei, N. Berardi, A. Cattaneo,
Proc. Natl. Acad. Sci. USA 2005, 102, 3811-3816.
[9] L. Gao, J. Li, J. Qi, Bioorg. Med. Chem. 2010, 18, 2131-2134.
[10] Y. Luo, K. Sun, L. Li, L. Gao, G. Wang, Y. Qu, L. Xiang, L. Chen, H. Hu, J. Qi,
ChemMedChem 2011, 6, 1986-1989.
[11] G. Wang, L. Bian, H. Zhang, Y. Wang, L. Gao, K. Sun, L. Xiang, J. Qi, Chem. Pharm.
Bull. 2016, 64, 161-170.
[12] A. V. Joshua, J. R. Scott, Tetrahedron Lett. 1984, 25, 5725-5728.
[13] C. Deraeve, A. Maraval, L. Vendier, V. Faugeroux, M. Pitié, B. Meunier, Eur. J. Inorg.
Chem. 2008, 5622-5631.
[14] M. Nguyen, A. Robert, A. Sournia-Saquet, L. Vendier, B. Meunier, Chem. Eur. J. 2014,
20, 6771-6785.
[15] Stability constants of metal-ion complexes. A. E. Martell, Ed., Chem. Soc. Special
Publication, 2nd Edition (1971).
[16] P. Faller, C. Hureau, Chem. Eur. J. 2012, 18, 15910-15920.
[17] a) A. Avdeef, S. R. Sofen, T. L. Bregante, K. N. Raymond, J. Am. Chem. Soc. 1978,
100, 5362-5370; b) T. Kabawata, V. Schepkin, N. Haramaki, R. S. Phadke, L. Packer,
Biochem. Pharmacol. 1996, 51, 1569-1577.
16
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
[18] C. A. Blindauer, I. Harvey, K. E. Bunyan, A. J. Stewart, D. Sleep, D. J. Harrison, S.
Berezenko, P. J. Sadler, J. Biol. Chem. 2009, 284, 23116-23124.
[19] M. Nguyen, C. Bijani, N. Martins, B. Meunier, A. Robert, Chem. Eur. J. 2015, 21,
17085-17090.
[20] G. R. Buettner, B. A. Jurkiewitz, Free Rad. Biol. Med. 1993, 14, 49-55.
[21] P. Faller, C. Hureau, Dalton Trans. 2009, 1080-1094.
[22] a) B. Aliès, E. Renaglia, M. Rózga, W. Bal, P. Faller, C. Hureau, Anal. Chem. 2013, 85,
1501-1508; b) C. J. Sarell, C. D. Syme, S. E. J. Rigby, J. H. Viles, Biochemistry 2009,
48, 4388-4402.
[23] T. R. Demmin, M. D. Swerdloff, M. M. Rogić, J. Am. Chem. Soc. 1981, 103, 5795-
5804.
[24] I. A. Koval, P. Gamez, C. Belle, K. Selmeczi, J. Reedijk, Chem. Soc. Rev. 2006, 35,
814-840.
[25] R. M. Buchanan, C. Wilson-Blumenberg, C. Trapp, S. K. Larsen, D. L. Greene, C. G.
Pierpont, Inorg. Chem. 1986, 25, 3070-3076.
[26] D.-Y. Choi, Y.-J. Lee, J. T. Hong, H.-J. Lee, Brain Res. Bull. 2012, 87, 144-153.
[27] A. Martorana, G. Koch, Frontiers in Aging Neurosc. 2014, 6, Art.252.
[28] C. Hansch, A. J. Leo, Substituent constant for correlation analysis in chemistry and
biology, New York: Wiley (1979).
17
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
Captions for Figures and Schemes
Figure 1. Structures of (a) gentisides, and (b) synthetic (2,3-dihydroxyalkylbenzoates)
ligands.
Figure 2. Structure of the bis(2,3-dihydroxyalkylbenzamides) ligands 1-3.
Figure 3. Titration of ligand 1 (15 µM) from 0 (dashed line) to 1.75 mole equivalent of
CuCl2, in Tris-buffered saline pH 7.4/methanol, 1/1, v/v. Insert: Absorbance at
342 nm against the Cu2+/1 mole ratio.
Figure 4. Determination of log Kapp of CuII-1 by using UV-visible spectrometry in Tris-
buffered saline pH 7.4/methanol, 1/1, v/v. Dotted line: ligand 1, full line: Cu2+/1
= 1/1, dashed line: Cu2+/1/EDTA = 1/1/1.
Figure 5. (a) Titration of ligand 1 (15 µM) from 0 (dashed trace) to 1.2 mole equivalent of
FeCl3, in Tris-buffered saline pH 7.4/methanol, 1/1, v/v. (b) Absorbance at 333
nm against the Fe3+/1 mole ratio.
Figure 6. (a) Titration of ligand 2 (15 µM) from 0 (dashed trace) to 20 mole equivalent of
ZnCl2, in Tris-buffered saline pH 7.4/methanol, 1/1, v/v. (b) Absorbance at 334
nm against the Zn2+/2 mole ratio.
Figure 7. UV-visible (265 nm) kinetic spectra of ascorbate consumption under air in the
presence of (a) no additive, (b) CuCl2, (c) CuII–Aβ1-16, 1/1, (d) CuII–1, 1/0.7, (e)
CuII–1, 1/1.1, (f) CuII–Aβ1-16, 1/1, + 1 (1.1 mole equiv). Spectra (c)-(f) were
obtained after substraction of the substrate absorbance at 265 nm. [Ascorbate] =
100 µM, [Cu] = 10 µM, [Aβ1-16] = 11 µM.
Scheme 1. Synthesis of ligands 1 and 3 from 2,3-dihydroxybenzoic acid. (i) BF3 in CH3OH,
reflux, 5 h; (ii) K2CO3/NaI, BnBr in DMF, RT, overnight; (iii) KOH in H2O,
90 °C,
2
h; (iv) CDI, 1,7-diaminoheptane (for 1) or 2,2’-
(ethylenedioxy)bis(ethylamine) (for 3) in CH2Cl2, RT, overnight; (v) 10% Pd-C in
C2H5OH.
Scheme 2. Synthesis of ligand 2 from 2,3-dihydroxybenzoic acid. (i) HOBt, DCC in THF,
2,2’-oxobis(ethylamine), RT, overnight.
Scheme 3. Possible equilibrium between the η2-catecholate and semiquinonate forms of iron
and copper complexes.
18
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
Figure S1. Absorbance at 333 nm against the Fe3+/1 mole ratio, from 0 to 6.
Figure S2. Absorbance at 323 nm against the Fe3+/3 mole ratio, from 0 to 5.
Figure S3. Titration of ligands 1 and 3 (15 µM) from 0 to 20 mole equivalent of ZnCl2, in
Tris-buffered saline pH 7.4/methanol, 1/1, v/v. Absorbance at 329 nm (a) and 338
nm (b) for L = 1 and 3, respectively, against the Zn2+/L mole ratio.
19
This article is protected by copyright. All rights reserved.

10.1002/ejic.201700385
European Journal of Inorganic Chemistry
Revised manuscript ejic.201700385 for the European Journal of Inorganic Chemistry, May 12th, 2017
Synopsis for the Table of Contents
Key topic. Bis(2,3-dihydroxyalkylbenzamide as specific redox metal chelators.
Bis(2,3-dihydroxyalkylbenzamide) ligands exhibit high specific affinities for Cu(II) and
Fe(III), and efficiently inhibit the reduction of dioxygen induced by CuII–Aβ1-16. These data
suggest that they should be able to act as regulators of redox metal homeostasis in the brain of
patients with Alzheimer’s disease, without interaction with zinc ions.
Graphic for the Table of Contents
Bis(hydroxybenzamide) ligands:
Regulate metal homeostasis & Promote neurite repair
Bis(catechols)
Specific chelators
of Cu and Fe
HO
OH
HO
OH
H
N
Y
Gentisides
R
O
O
Promote neurite
outgrowth
20
This article is protected by copyright. All rights reserved.