Preparation and study of new poly-8-hydroxyquinoline chelators for an anti-alzheimer strategy

Article abstract of DOI:10.1002/chem.200701024

Fourteen different ligands have been synthesized with two covalently linked 8-hydroxyquinoline motifs that favor metal complexation. These bis-chelators include different bridges at the C2 positions and different substituents to modulate their physicochemical properties. They can form metal complexes in a ratio of one ligand per metal ion with Cu^II and Zn^II, two metal ions involved in the formation of amyloid aggregates of the toxic Aβ-peptides in the Alzheimer disease. The apparent affinity of all bis-8-hydroxyquinoline ligands for Cu^II and Zn^II are similar with logK_CuII, ≈16 and logK_ZnII≈13 and are 10000 times more efficient than for the corresponding 8-hydroxyquinoline monomers. Their strong chelating capacities allow them to inhibit more efficiently than the corresponding monomers the precipitation of Aβ-peptides induced by Cu^II and Zn^II and also to inhibit the toxic formation of H₂O₂ due to copper complexes of Aβ. The best results were obtained with a one-atom linker between the two quinoline units. X-ray analyses of single-crystals of Cu ^II, Zn^II or Ni^II complexes of 2,2′-(2,2-propanediyl)-bis(8-hydroxyquinoline), including a oneatom linker, showed that all heteroatoms of the bis-8-hydroxyquinoline ligand chelate the same metal ion in a distorted square-planar geometry. The Cu^II and Zn^II complexes include a fifth axial ligand and are pentacoordinated.

Full text of DOI:10.1002/chem.200701024

DOI: 10.1002/chem.200701024
Preparation and Study of New Poly-8-Hydroxyquinoline Chelators for an
anti-Alzheimer Strategy
CØline Deraeve,[a] Christophe Boldron,[a] Alexandrine Maraval,[a] HonorØ Mazarguil,[b]
Heinz Gornitzka,[c] Laure Vendier,[a] Marguerite PitiØ,*[a] and Bernard Meunier[a, d]
Abstract: Fourteen different ligands
have been synthesized with two cova-
lently linked 8-hydroxyquinoline motifs
that favor metal complexation. These
bis-chelators include different bridges
at the C2 positions and different sub-
stituents to modulate their physico-
chemical properties. They can form
metal complexes in a ratio of one
ligand per metal ion with CuII and ZnII,
two metal ions involved in the forma-
tion of amyloid aggregates of the toxic
Ab-peptides in the Alzheimer disease.
The apparent affinity of all bis-8-hy-
droxyquinoline ligands for CuII and
ZnII are similar with logKCuII ꢀ16 and
logKZnII ꢀ13 and are 10000 times more
efficient than for the corresponding 8-
hydroxyquinoline monomers. Their
strong chelating capacities allow them
to inhibit more efficiently than the cor-
responding monomers the precipitation
of Ab-peptides induced by CuII and
ZnII and also to inhibit the toxic forma-
tion of H2O2 due to copper complexes
of Ab. The best results were obtained
with a one-atom linker between the
two quinoline units. X-ray analyses of
single-crystals of CuII, ZnII or NiII com-
plexes of 2,2’-(2,2-propanediyl)-bis(8-
hydroxyquinoline), including
a one-
atom linker, showed that all heteroa-
toms of the bis-8-hydroxyquinoline
ligand chelate the same metal ion in a
distorted square-planar geometry. The
CuII and ZnII complexes include a fifth
axial ligand and are pentacoordinated.
Keywords: amyloids
· chelates ·
copper · peptides · zinc
Introduction
related peptides. They are toxic to neurons and their capaci-
ty to strongly chelate metal ions, particularly CuII and ZnII
(and also iron in a lesser extent), has been associated to this
toxicity. It has also been suggested that Cu–Ab complexes
can generate H2O2.[1,2] The accumulation of redox active
metal ions in these amyloid plaques is probably involved in
the oxidative stress inducing neuronal lesions leading to irre-
versible brain damage. Both CuII and ZnII also affect inter-
actions between Ab peptides and lipid membranes.
Among the different Ab peptides, Ab1–42 has the highest
affinity to metal ions and is fostered by Alzheimer disease-
promoting mutations and risks factors. Ab1–42 is probably the
most toxic amyloid peptide, but also the most difficult to
study due to its capacity to generate self-organized aggre-
gates.
The Alzheimer disease is associated with the aggregation of
b-type amyloid peptides (Ab) in the brain, leading to the
formation of amyloid plaques.[1] These 39- to 43-mer pep-
tides result from the cleavage of the amyloid precursor pro-
tein (APP). Ab1–40 and Ab1–42 are the two major pathology-
[a]Dr. C. Deraeve, Dr. C. Boldron, Dr. A. Maraval, Dr. L. Vendier,
Dr. M. PitiØ, Dr. B. Meunier
Laboratoire de Chimie de Coordination du CNRS
205 route de Narbonne, 31077 Toulouse cedex 4 (France)
Fax : (+33)561-55-30-03
E-mail: pitie@lcc-toulouse.fr
Different metal-chelating agents are able to dissolve Ab-
deposits by removing metal ions from the amyloid aggre-
gates.[3] The use of a hydrophobic metal chelator like clio-
quinol showed some clinical improvements in a Phase II
trial, indicating that chelators are not only useful to study
the disease but might also have some therapeutic poten-
tial.[4]
[b]Dr. H. Mazarguil
Institut de Pharmacologie et Biologie Structurale du CNRS
205 route de Narbonne, 31077 Toulouse cedex 4 (France)
[c]Pr. H. Gornitzka
Laboratoire d’HØtØrochimie Fondamentale et
AppliquØe (UMR 5069-CNRS), UniversitØ Paul Sabatier
118 route de Narbonne, 31062 Toulouse cedex 9 (France)
[d]Dr. B. Meunier
We have recently shown that ligand 1, with a covalent
linkage between two 8-hydroxyquinoline units was particu-
Current adress:
PALUMED, BP 28262, 31682 Lab›ge cedex (France)
682
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

FULL PAPER
larly efficient to inhibit the CuII- or ZnII-induced aggrega-
tion of the b-amyloid peptide Ab1–42 when compared with
clioquinol which is a monomeric 8-hydroxyquinoline deriva-
tive (Figure 1).[5] Our idea was to favor metal complexation
Figure 1. Structures of compound 1, 8-hydroxyquinaldine and clioquinol.
by having two 8-hydroxyquinoline motifs around the metal
ion to obtain in vivo a classical metal/ligand ratio for CuII
and ZnII for such type of ligand, having in mind that this
ratio has a very poor probability to be observed in vivo
when the two chelating entities are not covalently linked.
Ligand 1 can also inhibit the production of H2O2 induced by
the copper–Ab1–42 complexes and involved in the toxicity of
the peptide. These results suggest that this ligand is a good
candidate to study the role of metal ions in Alzheimer dis-
ease and might be considered as a prototype of a drug can-
didate.
We report here the preparation of different bis-8-hydro-
quinoline in order to study the influence of the linker length
or of the nature of substituents with respect to their behav-
ior with amyloid plaques. These parameters can modulate
their physicochemical properties and, therefore, can influ-
ence their bioavailability, their stability and their metaboli-
sation in vivo. All ligands are without asymmetric centers
and are symmetric for synthetic purposes. Their activity was
compared with clioquinol but also to 8-hydroxyquinaldine
(Figure 1) that can be both considered as parent monomers
of this bis-8-hydroxyquinoline series. They showed a good
anti-amyloid activity to dissolve metal-aggregated Ab1–42
and to inhibit H2O2 production that correlates with their
high ability to chelate the metal ions involved in the Alz-
heimer disease. Since the experiments were performed on
the most amyloidogenic Ab1–42 peptide, it is reasonable to
propose that these ligands can be active against all the other
Ab-peptides encountered in amyloid aggregates.
Figure 2. ORTEP drawing of complex 2.
Table 1. Selected bonds lengths [Š]and angles [ 8]for 2, 27, 28 and 29.
2[a]
27
28
29
ꢁ
O1 C1
1.327(6)
1.326(6)
1.860(3)
1.854(3)
1.859(4)
1.840(4)
1.330(3)
1.333(3)
1.9511(18)
1.9643(19)
1.969(2)
1.301(7)
1.322(7)
1.942(5)
1.941(4)
1.982(5)
1.970(5)
1.342(3)
1.334(3)
2.0982(19)
1.9706(19)
1.983(2)
1.960(2)
ꢁ
O2 C20
M1ꢁO1
M1ꢁO2
ꢁ
M1 N1
ꢁ
M1 N2
1.956(2)
2.2641(18)
ꢁ
M1 O2(1
ꢁ
M1 O3
2.283(5)
ꢁ
M1 Cl1
2.3267(9)
80.88(8)
135.99(9)
90.75(8)
88.46(9)
157.65(9)
83.24(9)
O1-M1-N1
O1-M1-N2
O1-M1-O2
N1-M1-N2
O2-M1-N1
O2-M1-N2
O1-M1-O2(1
O2-M1-O2(1
N1-M1-O2(1
N2-M1-O2(1
O1-M1-O3
O2-M1-O3
N1-M1-O3
N2-M1-O3
O1-M1-Cl1
O2-M1-Cl1
N1-M1-Cl1
N2-M1-Cl1
C9-C10-C13
C11-C10-C12
88.04(17)
177.28(17)
88.92(16)
94.34(18)
176.93(17)
88.69(17)
85.50(9)
160.10(8)
95.03(8)
91.70(10)
171.96(9)
85.07(9)
94.12(7)
86.23(8)
101.74(8)
105.73(8)
84.5(2)
162.0(2)
94.93(18)
90.7(2)
165.7(2)
85.38(19)
98.30(19)
92.09(19)
102.2(2)
99.72(19)
94.09(6)
100.70(7)
100.55(7)
129.89(7)
111.1(2)
106.8(2)
118.5(4)
110.0(4)
113.0(2)
107.8(2)
116.0(5)
106.7(6)
Results and Discussion
[a]Two molecules of complex appeared in the asymmetric crystal unit
cell. Both showed similar values. Therefore, for clarity, values were only
given here for one of both complexes.
Ligands design and synthesis: Different substitutions on the
CH2 linker of 1 were performed (Scheme 1). Methylation
yielded 3 as previously described by Yamamoto et al.[6]
During the synthesis, single-crystals of [2,2-isopropylidenedi-
8-quinolinolato]nickel(II) complex 2 suitable for X-ray anal-
ysis were obtained (Figure 2 and Table 1). A symmetric and
monomeric NiII complex was observed. The geometry is
square planar around the metal center with the two depro-
tonated 8-hydroxyquinoline chelating entities in cis-position,
as expected from the design of this ligand series.[5]
Fluorinated ligand 4 was obtained by electrophilic fluori-
nation of the linker of 1 with Selectfluor.[7] When compared
with methylation, fluorination does not induce steric con-
straints because H and F substituents are isosteric. Ligand 1
was also quantitatively oxidized in 5 in aqueous alkaline
conditions.
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
683

M. PitiØ et al.
line residues in poly-8-hydroxyquinoline against metal com-
plexes of Ab.
Bis-8-hydroxyquinoline ligands with other two-atom link-
ers were also prepared. A reductive amination of 8-hydroxy-
quinoline-2-carboxaldehyde by 2-amino-8-hydroxyquinoline
produced 12 (with a CH2NH linker, Scheme 3). A peptidic
Scheme 1. Modifications of the CH2 linker of compound 1.
a) Ni(CH3CO2)2, CH3OH, 608C; b) CH3I, NaOH, CH3OH, reflux;
c) HCl, ethanol or EDTA, H2O; d) Selectfluor, CH3CN, RT; e) NaHCO3,
H2O, THF, RT.
The length of the linker between the 8-hydroxyquinoline
entities was also increased, from one to two and four atoms.
Compound 7, with an ethylene linker, was previously pre-
pared by Albrecht et al.[8] and by Kitamura et al.[9] but with
very poor yields: 5.6 and 13% from 8-hydroxyquinaldine,
respectively. We optimized the synthesis (Scheme 2). A first
Scheme 3. Synthesis
b) NaBH(CH3CO2)3, RT.
of
compound
12.
a) Cl(CH2)2Cl,
RT;
coupling reaction between 2-amino-8-hydroxyquinoline and
8-hydroxyquinoline-2-carboxylic acid allowed to obtain 13
(with
a
C(O)NH
linker,
Scheme 4).
Different substituents were
also introduced in the bis-8-hy-
droxyquinoline ligands. In the
presence of N-chlorosuccini-
mide in acidic conditions, a se-
lective chlorination on C5 car-
bons of 1, 3 and 7 allowed to
obtain 14, 15 and 16, respective-
ly, in high yields (Scheme 5).[10]
Then, a successive reaction with
N-iodosuccinimide in the same
conditions led to 17 (a dimeric
analogue of clioquinol) from
16.
Ligands 25 and 26 with meth-
yloxy substituents on C7
carbon, in the ortho-position of
the phenolic function, were also
prepared. This type of substitu-
ents provide ligands with a par-
Scheme 2. Synthesis of compounds 7, 9, 11 and 26. a) CH3I, K2CO3, acetone, RT; b) 1 equiv LDA, THF,
ꢁ958C then Br(CH2)2Br; c) 48% HBr, reflux; d) 2 equiv LDA, THF, ꢁ958C then Br(CH2)2Br; e) 3.7 equiv
LDA, THF, 08C then CuCl2.
improvement consisted in performing a homo-coupling be-
tween two 8-methyloxyquinaldine in the presence of LDA
(lithium diisopropylamine) and 1,2-dibromoethane at
ꢁ958C to produce 6 (with the ethylene linker) in 38%
yield. The product was contaminated by compound 8 (with
a butylene linker) that was easily removed by recrystalliza-
Scheme 4. Synthesis of compound 13. a) BOP, HOBT, Et3N, CH2Cl2, RT;
tion. Then acid hydrolysis to remove the 8-methyloxy pro-
tective group yielded 7 (or 9 with the butylene linker). In
fact the protection of the 8-hydroxyl function was found
later as unnecessary and 7 was directly obtained in a better
yield (52%) by the same homo-coupling protocol from 8-hy-
droxyquinaldine. Interestingly, modifications of the reaction
conditions (replacement of 1,2-dibromoethane by CuCl2 as
oxidizing agent and increase of the reaction temperature to
08C) gave an additional product from 8-methyloxyquinal-
dine: trimer 11 (Scheme 2). Compound 11 allowed us to
document the effect of a higher number of 8-hydroxyquino-
b) H2O, RT.
Scheme 5. Halogenation of bis-8-hydroxyquinoline ligands. a) N-chloro-
succinimide, H2SO4, RT; b) N-iodosuccinimide, H2SO4, CH3OH, RT.
684
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
tial homology of structure with curcumin, which is a dimeric
molecule active in dissolving Ab-deposits.[11] Selective bro-
mination, in alkaline conditions, on the C7 position of 8-hy-
droxyquinoline monomer generated the expected bromine-
containing compounds (Scheme 6).[12] Then, a nucleophilic
Scheme 6. Synthesis of 7-methyloxy-8-hydroxyquinoline derivatives. a) N-
bromosuccinimide, tBuNH2, toluene; b) NaOCH3, CuCl2, DMF, reflux.
Figure 3. ORTEP drawing of compound 24.
substitution of the bromine atom by CH3ONa led to the 7-
methyloxy-8-hydroxyquinoline precursors 19 and 21.[13] They
were used to prepare 25 (with
a methylene linker,
Scheme 7) and 26 (including an ethylene linker, Scheme 2)
shifted to lower energy with concomitant appearance of an
absorption band in the visible region (lmax near 374–420 nm)
probably due to a MLCT transition. At higher M/L ratios,
no further change in UV-visible absorption was observed.
The observation of isobestic points was in accordance with
the formation of only one type of metal complex. These re-
sults correlated with the ability of the bis-8-hydroxyquino-
line ligands to form only one type of CuII or ZnII complex
with a 1:1 metal/ligand ratio. This M/L ratio was confirmed
by mass spectrometry. A typical example, obtained for the
CuCl2 titration with 3, is given on Figure 4.
Scheme 7. Synthesis of compound 25. a) Benzylchloride, K2CO3, CH3CN,
reflux; b) m-CPBA, CH2Cl2, RT; c) CH2(CO2CH3)2, acetic anhydride,
RT; d) 37% HCl, reflux.
with methods used for the syntheses of 1[6] and 7, respective-
ly. Interestingly, X-ray analysis of single-crystals of inter-
mediate 24 confirmed the position of the substituents on the
ligand (Figure 3).
Figure 4. UV-visible spectra of ligand 3 (L, 15mm) in CH3OH/Tris saline
buffer pH 7.4 1:1, in the presence of different ratios of CuCl2. Black
arrows show the variation of absorbance when the Cu/L ratio increases.
Grey arrows show isobestic points. The variation of absorbance at l=
383 nm show the stabilization of the complexation when Cu/L = 1.
Copper(II) and zinc(II) chelation: The stoichiometry of the
complexes with these different ligands and CuII or ZnII—the
two major metal ions (M) involved in the formation of amy-
loid plaques—was determined by metal titration. Experi-
ments were performed in buffered saline conditions (at
pH 7.4) in order to mimic the medium used during experi-
ments involving the amyloid-b peptide.
For the majority of the bis-8-hydroxyquinoline ligands (L)
studied here for a M/L ratio increased from 0 to 1, the p!
p* transition centered near 248–262 nm for the free ligand
Interestingly, major differences were observed in the case
of 13 (including an amide linker) that formed multiple com-
plexes with CuCl2. This ligand also required more than one
equivalent of ZnII per ligand to obtain the stabilization of
the titration although only one species was formed. The 13-
ZnII complex, with a M/L ratio of one, was observed by
mass spectrometry. One can propose that the affinity con-
stant of 13 for ZnII was weak in the conditions used. There-
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
685

M. PitiØ et al.
fore an excess of ZnII was required for 100% of complexa-
tion.
pH 7.4; logKLZnII = 9.1, logKL ZnII = 8.8 and logKappLZnII =
2
6.4 at pH 7.4),[14b] the parent compound for the bis-8-hydroxy-
quinoline series, indicates that the bis-8-hydroxyquinoline
No reliable data were obtained during UV/Vis analyses
involving CuII or ZnII ions and the ligands 1, 14 and 25 con-
taining a CH2 linker. The spectra changed with time showing
a poor stability of these complexes under the conditions
used. However, the ability of 1 and 14 to form a 1:1 metal/
ligand complexes was confirmed by mass spectrometry.
The affinity of bis-8-hydroxyquinoline ligands (excepted
1, 13, 14 and 25) for CuII or ZnII was estimated, in the buf-
fered saline conditions used for UV titrations, via competi-
tive chelation experiments with EDTA. This well-known
ligand was selected after several trials with different chela-
tors covering a large scale of affinity constants for CuII or
ZnII. EDTA appeared to be the most convenient competitor
for all ligands. The logarithm of the apparent affinity con-
stants of EDTA are 15.9 for CuII and 13.7 for ZnII in the
used conditions at pH 7.4.[14,15] The percentage of bis-8-hy-
droxyquinoline complex obtained in the presence of one
equivalent of EDTA and one equivalent of metal ion was
deduced from UV/Vis spectra and this value was used to es-
timate the apparent affinity constants (Table 2), as described
in the Experimental Section.
ACHTREUNG
strategy is particularly efficient to obtain ligands with high
affinity constants for CuII and ZnII.
Crystal structures of copper(II) complexes of ligand 3:
Ligand 3 was metalated with one equivalent of different
cupric salts [Cu(OAc)2, CuSO4 or CuCl2]. Different single-
G
crystals were obtained and were analyzed by X-ray diffrac-
tion. All complexes are five-coordinated, neutral, with a
more or less distorted square-pyramidal configuration and a
coordination sphere with all the heteroatoms of the ligand.
Oxygen and nitrogen atoms of the ligand occupy the equa-
torial plane. However, major differences are observed due
to the recrystallization conditions or to the copper salt used
for metalation. They provide information concerning the
structures that can adopt such cupric complexes of bis-8-hy-
droxyquinoline bridged on the C2 position of the quinoline
entity. To our knowledge, they have never been previously
described from analysis of single-crystals.
When ligand 3 was metalated in the presence of Cu-
A
All studied bis-8-hydroxyquinoline ligands have similar
apparent affinity with logKCuII ꢀ16 and logKZnII ꢀ13 in the
experimental conditions used. Interestingly, a comparison
with the apparent affinity constants of 8-hydroxyquinaldine
dimeric complex 27 was obtained (Figure 5 and Table 1)
with both copper centers showing the same distorted
square-pyramidal geometry. One of the two oxygens of each
ligand is m-coordinated, with a participation to the equatori-
al plane of one copper center and a second engagement in
(logKLCuII = 11.0, logKL CuII = 7.7 and logKappLCuII = 9.3 at
2
Table 2. Affinity constants of bis-8-hydroxyquinoline ligands for CuII and ZnII (logKaff), partition coefficients (LogD7.4) and inhibition of the precipitation
of Ab and of the production of H2O2.
logKaff for
logD7.4
Soluble Ab (%)[a]
nmol H2O2 (% recovered H2O2)[b]
LMII
Compound
Size of Z
Z
R1 R2
CuII
ZnII
No metal CuII
ZnII
No Ab
Cu/Ab=2
Cu/Ab=0.75
no ligand
1
14
25
3
15
4
5
7
16
17
26
12
13
9
–
1
1
1
1
1
1
1
2
2
2
2
2
2
4
–
–
–
–
H
H
–
nd
nd
–
nd
nd
nd
–
55Æ4
18Æ2 29Æ3 2.93Æ0.18 2.77Æ0.16
1.08
0.76
0.61
nd
CH2
CH2
CH2
C(CH3)2
C(CH3)2 Cl
CF2
C=O
(CH2)2
(CH2)2
(CH2)2
(CH2)2
CH2NH
C(O)NH
(CH2)4
–
H
Cl
H
H
3.3Æ0.1 63Æ4
60Æ5 63
45Æ5 57
48Æ3 50
47Æ3 59
0.54
0.56
nd
0.75Æ0.11
0.80Æ0.23
0.70
3.5Æ0.1 66
CH3O nd
nd
53
H
H
H
H
H
H
I
15.9
13.0
12.5
13.9
14.4
13.9
13.0
14.2
13.3
12.6
nd
4.4Æ0.1 56
4.9Æ0.1 55
3.8Æ0.1 60
3.9Æ0.1 56
3.5Æ0.1 59
4.2Æ0.1 59
5.6Æ0.5 36
3.5Æ0.1 63Æ3
0.84 (100) 0.71Æ0.02 (100) 0.55
15.5
15.7
16.6
16.1
15.5
16.2
30
49
0.12
0.39
0.82
0.75
0.66
0.66
0.37Æ0.09 (97) 0.35
0.55Æ0.08 (97) nd
0.86Æ0.11 (92) nd
H
H
H
Cl
Cl
H
H
H
H
–
51Æ1 50
58Æ3 54
35
36
19
60
58
47
0.89Æ0.07
0.89Æ0.09
0.82Æ0.08
0.69
0.68
0.64
0.82
nd
nd
0.45
nd
nd
0.79
0.57
CH3O 16.0
38Æ4 42Æ4 nd
H
H
H
–
–
–
15.7
nd
nd
45
43 nd nd
nd
3.0Æ0.1 66
3.8Æ0.1 52
3.7Æ0.1 48
2.4Æ0.1 69
3.8Æ0.1 63
46Æ3 62Æ5 0.47
0.47
nd
nd
0.77
15.7
nd
13.5
nd
42
48
52
43
45
62
nd
11(trimer)
8-hydroxyquinaldine
clioquinol
nd
–
–
–
–
11[c]
–
9.1[c]
–
0.72
0.54
55Æ3 37
0.78Æ0.08
[a]A b1–42 (5 mm) was incubated for 1 h at 378C in 20 mm Tris·HCl buffer (pH 7.4), 150 mm NaCl with or without 20 mm metal salt (CuCl2 or ZnCl2) then
for another 1 h in the presence of 200 mm ligand followed by centrifugation. Quantities of soluble and precipitated Ab1–42 were determined in the super-
natant and the pellet, respectively, with MicroBCA. [b]Experiments were performed in the presence of A b1–42 (0.2 or 0.53 mm), 0.4 mm CuCl2, 0.4 mm
ligand (0.8 mm in the case of 8-hydroxyquinaldine and clioquinol), 10 mm ascorbate and air. Generated H2O2 (in nanomoles) was quantified with Am-
plexRed Assay. The recovered percentage of 1.5 nmol of H2O2 added in the medium was given into brackets. [c]From reference [14b]; nd: not deter-
mined.
686
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
monomeric complex 29 were obtained (Figure 7 and
Table 1). The geometry of the copper center of 29 can be
described as a highly distorted square-planar pyramid with a
Figure 5. ORTEP drawing of complex 27.
the axial position of the second copper center. Since the
complex is neutral the phenolic substituents of 3 are depro-
tonated to generate 27.
When ligand 3 was dissolved in DMF to be metalated
with an aqueous solution of CuSO4, single-crystals of mono-
meric complex 28 were obtained (Figure 6 and Table 1).
Figure 7. a) ORTEP drawing of complex 29. b) Stabilization of two com-
plexes by hydrogen bonds.
chloride ion as axial ligand. Only one of the two phenolic
positions of 3 is deprotonated to form 29, it was named O2
on the structure. On the other quinoline unit of the ligand, a
hydrogen atom was located on O1 by Fourier differences
ꢁ
map, with an O1 H bond of 0.93 Š. However, the angles in
the crystal structure deviate from ideal values for a square
planar pyramid and the geometry can be also described as a
highly distorted trigonal bipyramidal arrangement around
the metal ion. In this last proposition, the equatorial posi-
tions are occupied by the chloride ion, the phenolic atom of
one of the both quinoline units of the ligand and by the ni-
trogen of the other quinoline unit. The other nitrogen and
oxygen of the ligand are in axial position and this last
oxygen is deprotonated in a phenolate form. Importantly,
two copper complexes are stabilized by two intermolecular
hydrogen bonds between the phenol hydrogen of one com-
plex and the phenolate of the other one at 1.69 Š. The
O···O separation was 2.439(3) Š and the angle O-H···O was
135.78.
Figure 6. ORTEP drawing of complex 28.
Interestingly, the two pK values of 8-hydroxyquinoline
and 2-methyl-8-hydroxyquinoline monomer have been re-
ported to be around 9.8 (close to the 9.6 value of phenol)
and 11.7 for pK1 and 5.0 (close to the 5.2 value of pyridine)
and 4.6 for pK2, respectively.[14b] It is reasonable to assume
that bis-8-hydroxyquinoline ligands have similar values.
Therefore, their cupric complexes could be deprotonated on
both phenolic moieties of the ligand during experiments in
the buffered medium (at pH 7.4) that was used in the other
These experimental conditions are similar to those used by
Di Vaira et al. to obtain single-crystals of (clioqui-
nol)2CuII.[17] Oxygen of a molecule of water crystallization
solvent (O3) occupies the axial position of the copper com-
plex.
When ligand 3 was dissolved in CH3OH and then metalat-
ed with an aqueous solution of CuCl2 (the same salt used
for the determination of affinity constants), mono-crystals of
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
687

M. PitiØ et al.
experiments described in this manuscript. It can be noted
that crystal structures of cupric complexes of 8-hydroxyqui-
noline[18] or clioquinol[17] have been obtained with two quin-
oline entities, chelating, as phenolate, a copper ion in a
more or less distorted square-planar geometry. Copper was
four-coordinated. Therefore, such square-planar geometry
could exist in buffered medium for cupric complexes of bis-
8-hydroxyquinoline. This proposition is reasonable because:
i) all the heteroatoms of the ligand are more or less in an
equatorial position in the copper complexes 27, 28 and 29;
ii) the phenolic part of these complexes appeared easily de-
protonable since they are partially or totally deprotonated
in the complexes although no special treatment (like an ad-
dition of base) was performed to favor this deprotonation;
and iii) this type of square planar geometry was also ob-
served in the case of 2, the nickel(II) complex of ligand 3.
However, in such organization of bis-8-hydroxyquinoline
copper complexes involving two phenolate residues, five-or
six-coordinated copper complexes with one or two solvent
molecule (such as water) could be also present in the buf-
fered solution since this type of structure has been also pre-
viously observed for CuII complexes involving two 8-hydroxy-
quinolines[19] or two 8-hydroxyquinaldines.[20]
observed for CuII complexes of the same ligand. Complex 30
is five-coordinated, distorted square pyramidal and shows a
structure where all the heteroatoms of a same ligand mole-
cule coordinate the same metal ion. Oxygen and nitrogen
atoms of the ligand 3 occupy the equatorial plane. The axial
position is occupied by the oxygen of a molecule of DMSO
used as solvent of crystallization.
Analysis of the hydrophobicity of the ligands: The partition
of the ligands between 1-octanol and a physiological aque-
ous phase was also measured (Table 2). 8-Hydroxyquinal-
dine appeared to be poorly hydrophobic (logD7.4 =2.4 sug-
gesting that this molecule is 250 times more soluble in 1-oc-
tanol than in the buffer at physiological pH 7.4 value). The
halogenation increased highly the solubility in 1-octanol
since clioquinol had a log D7.4 value of 3.8 (the hydrophobic-
ity is known to favor the passage of a molecule through the
blood-brain barrier).[21]
The logD7.4 =3.3 value for 1 was higher than that of 8-hy-
droxyquinaldine, its monomeric analogue. This can be con-
sidered as an advantage for molecules that should go to the
brain. The logD7.4 increased also with the addition of fluoro,
keto or methyl substituents on the linker of 1 (logD7.4 =3.3,
3.8, 3.9 and 4.4 for 1, 4, 5 and 3, respectively). This logD7.4
value also increased with the length of the aliphatic linker
(logD7.4 =3.3, 3.5 and 3.8 for 1, 7 and 9, respectively).
Chlorination of the aromatic heterocycles allowed also to in-
crease this parameter (logD7.4 =3.3 and 3.5, 4.4 and 4.9, 3.5
and 4.2 when 1, 3 and 7 are compared to their 5-dichloro de-
rivatives 14, 15 and 16, respectively). Iodination had the
same effect (compound 17). Other substituents such as
methyloxy on the C7 carbons did not change the logD7.4
value (that was the same for 7 and 26).
Crystal structure of the (ligand 3)–zinc(II) complex: For
comparison, we tried to grow single-crystals of the ZnII com-
plexes of ligand 3. Unfortunately they crystallized as partic-
ularly thin needles and it was difficult to obtain single-crys-
tals suitable for X-ray analysis. We obtained, however,
single-crystals of complex 30 (Figure 8) when ligand 3 was
Therefore the modifications of bis-8-hydroxyquinoline
could modulate the biodistribution of the molecules. Impor-
tantly, excepted for 17, the modifications allowed to obtain
molecules with molecular weight <500, which is another pa-
rameter known to be favorable to a good bioavailability.[21]
Inhibition of the precipitation of Ab1–42 induced by CuII or
ZnII: The ability of poly-8-hydroxyquinoline ligands to in-
hibit the precipitation of Ab1–42 induced by CuII or ZnII was
compared in the conditions previously used to analyze
ligand 1 (Table 2).[5] The peptide was incubated for one hour
in the presence of four equivalents of CuCl2 or ZnCl2 fol-
lowed by the addition of 40 equivalents of the ligand. After
another hour of incubation, the quantity of soluble Ab1–42
after a centrifugation step was determined by microBCA
assay.
Figure 8. ORTEP drawing of complex 30.
Without addition of metal ion or ligand, only 55% of
Ab1–42 was soluble. The addition of CuCl2 strongly reduced
the solubility of Ab-peptide (only 18% of Ab1–42 remained
soluble), more than with ZnCl2 (29% of soluble Ab1–42).
Compound 1 appeared as the most efficient ligand to in-
hibit the peptide precipitation but other 8-hydroxyquinoline
derivatives were active. Important remarks can be deduced
from a comparison of the results obtained.
metalated with one equivalent of ZnSO4 then recrystallized
from a H2O/DMSO mixture. The crystal was very weakly
diffracting, so that a large proportion of essentially “unob-
served” reflections were used in the refinement, with an ele-
vated Rint value (>20%). Nevertheless, our model is good
enough for the discussion of the main features of this com-
plex. ZnII complex 30 shows a structure similar to the ones
688
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
The incorporation of a substituent on the quinoline resi-
dues induced a decrease of the inhibition of the precipita-
tion of Ab1–42 (compare 1 with 14 and 25; 3 with 15; 7 with
26; 16 with 17, respectively). Compound 17 with chloride
and iodide substituents was rather inefficient.
Compounds including a one-atom linker (1, 14, 25, 3, 15,
4 and 5) are generally more active than ligands including
larger linkers (7, 16, 17, 26, 12, 13 and 9). This result was
particularly clear in the inhibition of the precipitation of
Ab1–42 induced by CuCl2 (47–60% of solubility of Ab1–42 for
experiments involving one-atom linker ligand to compare to
only 35–46% for experiments involving two-atom linkers in
the case of R1 =R2 =H).
The nature of the linker modulates also the activity. In
the one-atom linker series, the best results were obtained
with a CH2 linker but bis-8-hydroxyquinoline ligands includ-
ing CF2 (4) or CO (5) gave good results since they were also
able to restore the solubility of Ab1–42 observed in the ab-
sence of added metal ion.
Figure 9. Variation of the percentage of soluble Ab1–42 peptide in the
presence of different stoichiometries of CuCl2 and 8-hydroxyquinoline
derivatives. Dark grey: Ab1–42 (5 mm) was incubated in the presence of
4 equiv of CuCl2 then 40 equiv of 8-hydroxyquinoline derivative then
centrifuged. Clear grey: Ab1–42 (5 mm) was incubated in the presence of
2.5 equiv of CuCl2 then 2.5 equiv of bis-8-hydroxyquinoline derivative
(5 equiv in the case of 8-hydroxyquinaldine and clioquinol) then centri-
fuged. Percentage of peptide in the supernatant (soluble Ab) and the
precipitate (aggregated Ab) were quantified with MicroBCA Assay.
The incorporation of another 8-hydroxyquinoline residue
did not increase the activity of poly-8-hydroxyquinoline for
the activity studied here (compare dimers 1 or 7 with the
trimer 11).
CuCl2) in the presence of ascorbate were catalytically able
to produce H2O2.
In the absence of Ab, all ligands inhibited the H2O2 for-
mation mediated by CuCl2 in the presence of ascorbate
(2.93 nmol H2O2).
A clear interest of the use of 1 compared with clioquinol
was previously observed when Ab1–42 was precipitated with
2.5 equiv of CuCl2.[5] This ratio of CuII per peptide is the
minimal quantity of CuII inducing the maximum of Ab1–42
precipitation. One equivalent of 1 per CuII was sufficient to
obtain the same peptide solubility observed without metal
ion. Clioquinol was less efficient, although two equivalents
of clioquinol per CuII were used to respect the metal/ligand
ratio of two 8-hydroxyquinoline residues per metal ion in
the case of 1 and clioquinol.
The same experiment was performed for the most active
ligands of the amyloid dissolution experiments: 1, 3, 4 and 5
(with one-atom linker) and 13 (the best compound in the
two-atom linker series) (Figure 9). Importantly, all bis-8-hy-
droxyquinoline gave statistically the same results for Ab1–42
solubility when compared to results obtained in conditions
involving an excess of metal ion and ligand. For these li-
gands, a 1:1 ratio for ligand/Cu was sufficient whereas
mono-8-hydroxyquinoline derivatives (8-hydroxyquinaldine
or clioquinol) failed to induce a large inhibition of Ab1–42
precipitation.
All chelators were also able to inhibit H2O2 formation
mediated by Ab1–42·Cun complexes in the presence of ascor-
bate. For a Cu/Ab1–42 ratio = 2, the ligands (in the propor-
tion of one equivalent per copper ion) were able to reduce
the production of H2O2 to 15–30% of the quantity produced
by Ab1–42·Cu2 in the absence of chelator (only 0.37–
0.89 nmol H2O2 were produced compare with 2.77 nmol
without ligand added). For a Cu/Ab1–42 ratio=0.75, the
copper complex of amyloid peptide produced less H2O2
(1.08 nmol) and again one equivalent of chelator per copper
reduced significantly the formation of H2O2.
The addition of a known quantity of H2O2 in the reaction
medium containing 3, 15 or 4 (three one-atom linker ligands
that showed interesting performances in the experiments
just above) allowed us to verify that their copper complexes
did not degrade H2O2 (97–100% of the added H2O2 was still
present at the end of the analysis). We have previously
shown that Ab-Cu did not degrade H2O2 in these condi-
tions.[22] This control allowed us to confirm that the observed
decrease of H2O2 formation was in fact due to product in-
hibition (and not to its putative degradation) by the addition
of 3, 15 or 4. These ligands probably removed copper from
Ab to generate redox-inactive copper complexes (3·Cu,
15·Cu or 4·Cu) in the presently used experimental condi-
tions. Importantly, this proposition is not valuable for all the
different ligands evaluated here since for experiments per-
formed in the presence of 5, a weak but real degradation of
H2O2 was observed (only 92% of added H2O2 was recov-
ered).
Therefore, the use of bis-8-hydroxyquinoline ligands
having a one-atom linker is favorable to maintain the solu-
bility of Ab-peptides.
Inhibition of the synthesis of H2O2 induced by copper com-
plexes of Ab1–42: Ab-Cu, in the presence of a reductant,
exerts a toxicity correlated with the generation of H2O2.[2]
The ability of the bis-8-hydroxyquinoline chelators to inhibit
the formation of H2O2 by CuCl2 in the absence of Ab1–42 or
for a Cu/Ab1–42 ratio = 2 or 0.75 was compared, see Table 2.
In the conditions used, Ab1–42, CuCl2 or copper complexes
of Ab1–42 did not generate a significant quantity of H2O2 in
the absence of ascorbate but copper complexes (Ab·Cun or
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
689

M. PitiØ et al.
Conclusion
Experimental Section
General: 2,2’-Methanediyl-bis(8-hydroxyquinolinium) dichloride dihy-
drate (1),[6] 8-methyloxyquinaldine[9] and 7-bromo-8-hydroxyquinoline
(18)[12] were synthesized as previously described. DMF and CH3CN were
dried over 4 Š molecular sieves. CH2Cl2 was dried over basic alumina.
THF was distilled over benzophenone and sodium. Other commercially
available reagents and solvents were purchased from standard chemical
suppliers and used without further purification. NMR spectra were re-
corded on Bruker spectrometers at 200, 250, 300 or 500 MHz. UV/Vis
spectra were recorded on a Hewlett-Packard 8452 A diode array spectro-
photometer or a Perkin-Elmer Lambda 35 spectrophotometer. Chroma-
tographic purifications were performed on silica gel.
Fourteen symmetric bis-8-hydroxyquinoline ligands were
prepared and they are efficient chelators of metal ions such
as CuII and ZnII. These metal chelators are able to inhibit
the precipitation of a 42-mer Ab-peptide enhanced by CuII
and ZnII ions and also to inhibit the H2O2 production associ-
ated to an oxidative stress induced by Ab. The hydrophobic-
ity of bis-8-hydroxyquinoline ligands seems to be adapted
for the passage of the blood-brain barrier to reach the toxic
amyloid aggregates involved in Alzheimer disease. Varia-
tions of substituents on these bis-8-hydroxyquinolines have
limited effects on the metal affinity or on the ability to in-
hibit H2O2 production but significant structure/activity rela-
tionship can be deduced from ligand-induced dissolution of
metal-aggregated Ab1–42. The best results were obtained for
a one-atom linker between both 8-hydroxyquinoline units.
Compound 1 (with a CH2 linker) was particularly active but
its chelation to CuII or ZnII was difficult to analyze in phys-
iological conditions. Substitutions of the methylene hydro-
gens of 1 by two methyl groups (ligand 3), fluoro atoms
(ligand 4) or a keto function (ligand 5) induced only a limit-
ed decrease of their activity and these chelators are less sus-
ceptible to an oxidative metabolism in vivo. Introduction of
substituents such as chloride, iodide or methyloxy on 8-hy-
droxyquinoline heterocycle had a negative effect in experi-
ments involving Ab-peptide solubility.
X-ray analysis: Data collection were collected at low temperature (180 or
193 K) on a Bruker-CCD (for 2), an Xcalibur Oxford Diffraction diffrac-
tometer (for 24, 28 and 29) or an IPDS STOE diffractometer (for 27 and
30) using a graphite-monochromated MoKa radiation (l=0.71073 Š) and
equipped with an Oxford Cryosystems (Cryostream Cooler Device). The
structures have been solved by direct methods using SHELXS-97[23] (for
2) or SIR92[24] (for 24, 27–30) and refined by means of least-squares pro-
cedures on F2 with the aid of the program SHELXL97[25] include in the
software package WinGX version 1.63.[26] The Atomic Scattering Factors
were taken from International tables for X-Ray Crystallography.[27] Most
of the hydrogens atoms were geometrically placed and refined by using a
riding model, few of them were located by Fourier differences map and
refined by using a riding model. All non-hydrogens atoms were aniso-
tropically refined, and in the last cycles of refinement a weighting
Scheme was used, where weights are calculated from the following for-
mula: w=1/[s2(Fo2)+(aP)2 +bP]where P=(Fo2 + 2Fc2)/3. Drawings of
molecule are performed with the program ORTEP32 with 30% probabil-
ity displacement ellipsoids for non-hydrogen atoms (excepted for 2 that
was drawn with 50% probability).[28] The crystal data for the determined
structures are given in Table 3.
Better in vitro results were obtained against the amyloid
peptide with these bis-quinoline ligands compared with the
monomeric derivatives, especially when a 1:1 Cu/ligand ratio
was utilized.
CCDC 652781 (29), 652782 (24), 652783 (28), 652784 (27), 652785 (30),
and 652786 (2) contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif
Table 3. Details of crystallographic measurements and refinements.
2
24
27
28
29
formula
Mw
T [K]193 (2)
C
435.13
22.5H22N2NiO3.5
C39H34N2O8
658.68
C42H32Cu2N4O4 4(CH4O)
911.98
180
C21H18CuN2O3
409.91
C21H17ClCuN2O2
428.36
180
110
180
crystal system
space group
triclinic
P1
10.2593(9)
10.8311(10)
15.2192(14)
monoclinic
P1211
triclinic
P1
15.9572(15)
9.1716(8)
23.850(2)
–
orthorhombic
Pbca
triclinic
P1
¯
¯
¯
a [Š]12.456(4)
b [Š]12.650(4)
c [Š]13.984(5)
a [8]112.260(5)
b [8]105.956(5)
g [8]90.115(6)
V [Š3]1946.7(11)
Z
8.4208(12)
10.5632(16)
11.4970(17)
89.769(12)
87.858(12)
81.443(12)
1010.6(3)
2
9.4456(10)
9.9688(12)
11.4489(8)
64.633(9)
82.430(7)
63.962(11)
873.12(19)
8
–
97.622(7)
90
–
–
1676.2(3)
4
1.305
0.092
908
3490.5 (5)
1
2
1calcd [gcmꢁ3]1.485
m [mmꢁ1]1.026
F(000)
1.499
1.114
1.560
1.276
1.629
1.423
692
474
1688
438
crystal size [mm]0.1” 0.4”0.8
q range [8]5.10–23.26
collected/independent reflns
reflns with [F2 >4s(F2)]5521
Rint
0.49”0.17”0.15
2.70–32.12
12267/5521
0.15”0.12”0.05
2.89–32.04
17105/5799
0.125”0.1”0.05
2.13–25.99
0.2”0.1”0.02
2.46–32.16
25850/3402
10909/6440
9591/5588
3358
6440
3402
5588
0.0614
533
0.977
0.0593
446
0.892
0.0478/0.0980
0.366/ꢁ0.426
0.0869
277
0.661
0.1251
246
0.966
0.0625/0.1612
0.650/ꢁ1.294
0.0386
246
0.959
0.0367/0.0880
0.589/ꢁ0.689
parameters
GOF on F2
R1[F2 >4s(F2)]/wR2
D1max/D1min [eŠ3]0.821/
0.0532/0.1081
0.0466/0.0522
0.771/ꢁ0.587
ꢁ0.466
690
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
[2,2’-(2,2-Propanediyl)-bis[8-quinolinolato]nickel(II) (2): The synthesis
was performed according to ref. [6]. More precise 1H NMR data than
those previously published are given, other unpublished characterizations
are added. 1H NMR (250 MHz, CDCl3): d=8.15 (d, J=9.0 Hz, 2H), 7.45
(brd, J=8.5 Hz, 2H), 7.33 (brt, J=8.0 Hz, 2H), 6.91 (brd, J=7.5 Hz,
2H), 6.81 (brd, J=6.5 Hz, 2H), 1.93 ppm (s, 6H); 13C NMR (63 MHz,
CDCl3): d=166.4, 159.1, 144.3, 138.7, 130.7, 127.7, 119.3, 113.6, 110.4,
50.7, 32.4 ppm; MS (DCI, NH3): m/z: 387 [M+H]+, 404 [M+NH4]+;
HRMS: m/z: calcd for C21H17N2O2Ni: 387.0644, found: 387.0646. Recrys-
tallization from CH3OH provided green single-crystals suitable for analy-
sis by X-ray diffraction.
129.5, 121.8, 117.9, 111.2 ppm; UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4,
NaCl 150 mm 1:1]: lmax (e)=254 (43300), 273 (26100), 310 (9600, sh),
375 nm (2300mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 317 [M+H]+; elemental
analysis calcd (%) for C19H12N2O3·0.5H2O: C 70.15, H 4.03, N 8.61;
found: C 70.11, H 3.83, N 8.94.
2,2’-(1,2-Ethanediyl)-bis(8-methyloxyquinoline) (6):
A solution of 8-
methyloxyquinaldine (5.32 g, 30.75 mmol) in dry THF (47 mL) under N2
was cooled to ꢁ958C in a CH3OH/liquid N2 bath. Lithium diisopropyl-
amine·THF (20.5 mL, 30.7 mmol, 1.5m in cyclohexane) in dry THF
(63 mL) was then added over 2 h. The solution was stirred 2 h more at
ꢁ958C before the addition of 1,2-dibromoethane (5.30 mL, 61.5 mmol).
The mixture was allowed to warm to room temperature and stirred over-
night. Water (32 mL) was then added, resulting in the formation of a
white precipitate removed by filtration. The filtrate was concentrated
under vacuum and purified by chromatography (CHCl3, 0–90% AcOEt).
The resulting product and the precipitate were recrystallized from hot
CH3OH to give 6 as a white solid (2.04 g, 38%). 1H NMR (250 MHz,
CDCl3): d=8.04 (d, J=8.5 Hz, 2H), 7.40 (d, J=8.5 Hz, 2H), 7.39 (t, J=
7.5 Hz, 2H), 7.35 (dd, J=1.5, 8.0 Hz, 2H), 7.06 (dd, J=1.5, 7.5 Hz, 2H),
4.10 (s, 6H), 3.61 ppm (s, 4H); 13C NMR (63 MHz, CDCl3): d=160.8,
155.0, 139.8, 136.3, 128.0, 125.9, 122.1, 119.5, 107.8, 56.1, 39.2 ppm; MS
(DCI, NH3): m/z: 345 [M+H]+; elemental analysis calcd (%) for
C22H20N2O2: C 76.72, H 5.85, N 8.13; found: C 76.58, H 6.04, N 7.99.
2,2’-(2,2-Propanediyl)-bis(8-hydroxyquinoline) (3): Complex
2 (3.32 g,
8.60 mmol) was suspended in ethanol (80 mL) and 37% HCl (16 mL)
was added to give a green solution. Hot H2O (320 mL) was added por-
tion-wise resulting in yellow crystals. After cooling, the crystals were col-
lected, washed with aqueous 1m HCl, air-dried overnight at room tem-
perature and dissolved in hot ethanol (32 mL). A hot aqueous solution of
AcONa·3H2O (1.60 g in 64 mL) was then added portion-wise. After cool-
ing, the precipitate was collected by centrifugation, washed with H2O and
extracted with CH2Cl2 to give 3 as a white product after evaporation of
the solvent (1.81 g). The supernatant of the previous steps contained re-
maining nickel complex. After removal of the solvent, it was demetalled
with EDTA (2”10 g in 500 mL H2O) under stirring for 1 h in CH2Cl2
(1 L). The organic layer was collected, washed with saturated aqueous
NaHCO3 and concentrated under vacuum. The mixture was purified by
chromatography (CH2Cl2, 0–1% CH3OH) to give 828 mg of 3 (total
mass=2.64 g, 93%). 1H NMR (250 MHz, CDCl3): d=8.30 (brs, 2H),
8.02 (d, J=9.0 Hz, 2H), 7.45 (t, J=8.0 Hz, 2H), 7.30 (dd, J=1.2, 7.5 Hz,
2H), 7.22 (d, J=9.0 Hz, 2H), 7.21 (dd, J=1.5, 7.5 Hz, 2H), 2.0 ppm (s,
6H); 13C NMR (63 MHz, CDCl3): d=164.7, 152.1, 136.8, 136.5, 127.5,
126.8, 121.2, 117.5, 110.3, 49.3, 28.0 ppm; UV/Vis [CH3OH/Tris·HCl
20 mm pH 7.4, NaCl 150 mm 1:1]: lmax (e)=203 (77700), 251 (84600),
308 nm (6900 mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 331 [M+H]+; elemental
analysis calcd (%) for C21H18N2O2: C 76.34, H 5.49, N 8.48; found: C
75.80, H 5.30, N 8.38.
2,2’-(1,2-Ethanediyl)-bis(8-hydroxyquinoline) (7): MethodA : To a solu-
tion of 8-hydroxyquinaldine (1.00 g, 6.29 mmol) in THF (10 mL), under
argon and at ꢁ958C, lithium diisopropylamine·THF (8.4 mL, 12.6 mmol,
1.5m in cyclohexane) in dry THF (12.5 mL) was added over 45 min. The
solution was stirred for 1.5 h then 1,2-dibromoethane (1.08 mL,
12.58 mmol) was added. The mixture was allowed to warm to room tem-
perature and stirred for 15 h. The product was precipitated with H2O
(12 mL), collected by filtration and washed with H2O. Then it was dis-
solved in CH2Cl2 (150 mL) where a few drops of AcOH were added to
dissolve the product that was washed with saturated aqueous NaHCO3
(2”100 mL) then with H2O to give 7 (0.51 g, 1.62 mmol, 52%) after
drying.
MethodB : A solution of 6 (2.83 g, 8.23 mmol) in 48% HBr (150 mL)
was heated under reflux for 24 h. After cooling to room temperature, the
mixture was neutralized with aqueous 3m NaOH, resulting in the forma-
tion of a pale green precipitate. The product was extracted with CH2Cl2
and washed with H2O and brine. The organic layer was dried (Na2SO4)
and the solvent was evaporated under vacuum to afford 7 as a pale green
powder (2.53 g, 97%).
1H NMR (250 MHz, CDCl3): d=9.35 (brs, 2H), 8.20 (d, J=8.5 Hz, 2H),
7.53 (d, J=8.5 Hz, 2H), 7.37 (t, J=6.5 Hz, 2H), 7.32 (dd, J=2.0, 6.5 Hz,
2H), 7.05 (dd, J=2.0, 6.5 Hz, 2H), 3.57 ppm (s, 4H); 13C NMR (75 MHz,
[D6]DMSO): d=160.2, 153.0, 138.1, 136.6, 127.6, 127.0, 122.8, 118.0,
111.4, 37.2 ppm; UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm
1:1]: lmax (e)=204 (71800), 248 (74200), 305 nm (5200mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 317 [M+H]+; elemental analysis calcd (%) for
C20H16N2O2·0.1NaBr: C 73.53, H 4.93, N 8.58; found: C 73.81, H 4.73, N
8.50.
2,2’-(Difluromethanediyl)-bis(8-hydroxyquinoline) (4): Compound 1 was
suspended in CH2Cl2 and washed with sodium acetate buffer (0.1m, pH
7.0) to obtain its neutral form 1’ (quantitative yield). Then, to an orange
suspension of 1’ (110 mg, 0.36 mmol) in dry CH3CN (30 mL) was added,
under N2, 1-chloromethyl-4-fluoro-1,4-diazoniabicyclo(2.2.2)octane bis-
A
small portions, resulting in the subsequent products dissolution to a
yellow solution. The mixture was stirred for 1.5 h at room temperature.
After solvent removal, the product was dissolved in CH2Cl2 and washed
two times with H2O. The organic layer was concentrated and dried under
vacuum to give
4 as a yellow powder (123 mg, quantitative yield).
1H NMR (250 MHz, CDCl3): d=8.34 (d, J=8.5 Hz, 2H), 8.00 (d, J=
8.5 Hz, 2H), 7.78 (s, 2H), 7.52 (t, J=8.0 Hz, 2H), 7.39 (dd, J=1.0,
8.0 Hz, 2H), 7.19 ppm (dd, J=1.0, 7.5 Hz, 2H); 13C NMR (126 MHz,
CDCl3): d=150.3 (t, 2J(C,F)=30.1 Hz, Cq), 152.3, 137.6, 137.2, 129.3,
128.5, 118.7 (t, 3J(C,F)=3.3 Hz, CH), 117.9, 117.0 (t, 1J(C,F)=245.6 Hz,
Cq), 111.0 ppm; 19F NMR (188 MHz, CDCl3, ref: CF3CO2H): d=
ꢁ22.2 ppm (s); UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm
1:1]: lmax (e)=241 (50000), 251 (55600), 314 nm (4700mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 339 [M+H]+, 356 [M+NH4]+; elemental analysis calcd
(%) for C19H12F2N2O2·0.3H2O: C 66.39, H 3.69, N 8.15; found: C 66.25,
H 3.62, N 8.27.
2,2’-(1,4-Butanediyl)-bis(8-methyloxyquinoline) (8): It was obtained
during the synthesis of 6 as major by-product in the supernatant of the
CH3OH recrystallization. A second crystallization in CH3OH allowed to
obtain a pure methanolic solution of the desired product which was
evaporated and dried under vacuum to give 8 as a white powder (0.68 g,
6%). 1H NMR (250 MHz, CDCl3): d=8.01 (d, J=8.5 Hz, 2H), 7.40 (d,
J=8.5 Hz, 2H), 7.39 (t, J=7.5 Hz, 2H), 7.35 (dd, J=1.5, 7.5 Hz, 2H),
7.06 (dd, J=1.5, 7.5 Hz, 2H), 4.01 (s, 6H), 3.10 (m, 4H), 1.95 ppm (m,
4H); 13C NMR (75 MHz, CDCl3): d=161.8, 155.0, 139.7, 136.2, 127.9,
125.7, 121.8, 119.4, 107.7, 56.1, 39.4, 30.0 ppm; MS (DCI, NH3): m/z: 373
[M+H]+; elemental analysis calcd (%) for C24H24N2O2·0.5H2O: C 75.57,
H 6.61, N 7.34; found: C 75.41, H 6.62, N 7.16.
Bis(8-hydroxy-2-quinolinyl)methanone (5): Compound
1
(60 mg,
0.16 mmol) was stirred in THF/saturated aqueous NaHCO3 1:1 (8 mL
total) for 48 h. CH2Cl2 (10 mL) and H2O (10 mL) were added. The organ-
ic phase was collected and the aqueous phase extracted further with
CH2Cl2 (2”10 mL). The organic phase was washed with H2O (10 mL)
and the solvent was evaporated under vacuum. The product was purified
by filtration over silica gel with CH2Cl2/CH3OH 97:3 to give 5 as an
orange powder (42 mg, 81%). 1H NMR (250 MHz, CDCl3): d= 8.36 (d,
J=8.5 Hz, 2H), 8.19 (d, J=8.5 Hz, 2H), 8.04 (brs, 2H), 7.59 (m, 2H),
7.43 (dd, J=1.0, 8.5 Hz, 2H), 7.22 ppm (dd, J=1.0, 7.5 Hz, 2H);
13C NMR (126 MHz, CDCl3): d=192.2, 153.1, 151.4, 137.1, 137.0, 130.4,
2,2’-(1,4-Butanediyl)-bis(8-hydroxyquinoline) (9): Compound 8 (50 mg,
0.13 mmol) was heated under reflux in 48% HBr (2.7 mL) for 24 h. After
cooling down to 08C, the reaction mixture was alkalinized to pH 8 with
aqueous 3m NaOH (5 mL, 15 mmol) and saturated aqueous NaHCO3
(2 mL). After adding H2O until the volume reached 15 mL, the aqueous
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
691

M. PitiØ et al.
phase was extracted with CH2Cl2 (3”10 mL). The organic phase was
washed with H2O (10 mL) and the solvent was evaporated under
vacuum. The product was purified by flash chromatography (CH2Cl2/
CH3OH/AcOH 96:3:1). The product was then dissolved in CH2Cl2
(10 mL) and washed with sodium acetate buffer (10 mL, 0.1m, pH 7).
Evaporation of the solvent gave 9 as a white powder (34 mg, 77%).
1H NMR (250 MHz, CDCl3): d=8.04 (d, J=8.5 Hz, 2H), 7.38 (t, J=
7.5 Hz, 2H), 7.31–7.25 (m, 4H), 7.15 (d, J=7.5 Hz, 2H), 3.02 (m, 4H),
1.93 ppm (m, 4H); 13C NMR (75 MHz, [D6]DMSO): d=160.8, 153.2,
138.3, 136.7, 127.6, 126.9, 122.5, 117.9, 111.4, 38.3, 29.4 ppm; UV/Vis
[CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax (e)=202
(81300), 244 (90800), 303 nm (6300mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 345
[M+H]+; elemental analysis calcd (%) for C22H20N2O2·0.25H2O: C 75.73,
H 5.92, N 8.35; found: C 75.39, H 5.69, N 8.63.
lmax (e)=253 (52100), 267 (42900), 345 nm (3600mꢁ1 cmꢁ1, sh); MS
(DCI, NH3): m/z: 318 [M+H]+; HRMS: m/z: calcd for C19H16N3O2:
318.1243, found: 318.1238.
8-Hydroxy-N-(8-hydroxy-2-quinolinyl)-2-quinolinecarboxamide (13): To a
suspension of 8-hydroxyquinoline-2-carboxylic acid (50 mg, 0.26 mmol) in
CH2Cl2 (5 mL) were added 1-hydroxybenzotriazole monohydrate (71 mg,
0.53 mmol),
(benzotriazole-1-yloxy)tris(dimethylamino)phosphonium
hexafluorophosphate (177 mg, 0.40 mmol) and Et3N (0.037 mL,
0.26 mmol). After stirring for 0.5 h at room temperature, 2-amino-8-hy-
droxyquinoline (85 mg, 0.53 mmol) and Et3N (0.037 mL, 0.26 mmol then,
after 5 min, 0.117 mL, 0.794 mmol) were added. After stirring for 18 h,
H2O (10 mL) and CH2Cl2 (5 mL) were added and the organic phase was
collected. The product was purified by chromatography (CH2Cl2/CH3OH
99.5:0.5). The product was then dissolved in CH2Cl2 (30 mL) and washed
with saturated aqueous NaHCO3. The solvent was evaporated to give 13
as a white powder (28 mg, 32%). 1H NMR (250 MHz, [D6]DMSO): d=
11.89 (s, 1H), 10.89 (s, 1H), 9.55 (s, 1H), 8.60 (d, J=8.5 Hz, 1H), 8.41 (s,
2H), 8.32 (d, J=8.5 Hz, 1H), 7.63 (dd, J=7.0 and 8.0 Hz, 1H), 7.53 (dd,
J=1.0, 8.0 Hz, 1H), 7.42 (dd, J=2.0, 8.0 Hz, 1H), 7.37 (dd, J=7.0,
8.0 Hz, 1H), 7.23 (dd, J=1.0, 7.0 Hz, 1H), 7.13 ppm (dd, J=2.0, 7.0 Hz,
1H); 13C NMR (75 MHz, [D6]DMSO): d= 163.8, 154.7, 152.7, 150.0,
147.3, 138.6, 138.6, 137.6, 137.2, 130.50, 130.46, 127.4, 126.5, 119.6, 118.3,
118.0, 116.6, 112.8, 112.5 ppm; UV/Vis [DMSO/Tris·HCl 20 mm pH 7.4,
NaCl 150 mm 8:2]: lmax (e)=259 (49100), 332 nm (11200mꢁ1 cmꢁ1, sh);
MS (DCI, NH3): m/z: 332 [M+H]+; HRMS: m/z: calcd for C19H14N3O3:
332.1035, found: 332.1011.
2,2’,2’’-(1,2,3-Propanetriyl)-tris(8-methyloxyquinoline) (10): To a solution
of 8-methyloxyquinaldine (1.41 g, 8.14 mmol) in dry THF (20 mL), under
argon and cooled on an ice-bath, was added lithium diisopropylami-
ne·THF (20 mL, 30.52 mmol, 1.5m) over 1 min. The solution was stirred
for 1 h at 48C before the addition of dry CuCl2 (1.32 g, 9.81 mmol) then
the mixture was stirred for 36 h at room temperature under argon. Water
(100 mL) was then added, and the crude product was extracted with
CHCl3, washed with brine and the solvent was evaporated before a pu-
rification step by chromatography (CHCl3, 20–100% AcOEt). The result-
ing product was dissolved in CH2Cl2 and washed with an aqueous solu-
tion of EDTA to remove copper traces and purified by chromatography
again in the previous conditions to give 10 as a white powder (150 mg,
11%). 1H NMR (250 MHz, CDCl3): d=7.81 (d, J=8.5 Hz, 1H), 7.75 (d,
J=8.5 Hz, 2H), 7.37–7.18 (m, 9H), 7.00 (dd, J=1.3, 7.5 Hz, 1H), 6.95
(dd, J=1.3, 7.5 Hz, 1H), 4.49 (dd, J=8.0, 7.0 Hz, 1H), 4.05 (s, 3H), 4.01
(s, 6H), 3.78 (dd, J=14.0, 8.0 Hz, 2H), 3.62 ppm (dd, J=14.0, 7.0 Hz,
2H); 13C NMR (63 MHz, CDCl3): d=162.7, 159.8, 155.4, 155.1, 140.0,
139.7, 135.8, 135.6, 128.1, 127.8, 125.7, 125.6, 122.8, 122.0, 119.5, 119.3,
108.2, 107.6, 56.4, 56.0, 48.4, 44.3 ppm; MS (DCI, NH3): m/z: 516
[M+H]+; HRMS: m/z: calcd for C33H30N3O3: 516.2287, found: 516.2300.
General protocol for chlorination of C5 of bis-8-hydroxyquinoline deriva-
tives: To a solution 0.1m of ligand in 97% H2SO4 at 08C was added N-
chlorosuccinimide (2 equiv) in small portions. The mixture was stirred for
15 min at 08C and 4 h at room temperature. It was then poured onto ice
to give a suspension that was neutralized with aqueous 6m NaOH. The
mixture was centrifuged, the supernatant was removed and the precipi-
tate was suspended in H2O and extracted with CH2Cl2. The organic layer
was washed with H2O and the solvent was evaporated under vacuum to
give the product.
2,2’,2’’-(1,2,3-Propanetriyl)-tris(8-hydroxyquinoline) (11): Compound 10
(30 mg, 0.06 mmol) was refluxed in 48% HBr (1.2 mL) for 36 h. The acid
was evaporated under vacuum and the residue was dissolved in CH2Cl2
(10 mL) and washed with sodium acetate buffer (10 mL, 0.1m, pH 7.0).
The organic phase was collected and the aqueous phase extracted further
with CH2Cl2 (2”10 mL). The pooled organic phases were washed with
H2O (10 mL) and the solvent was evaporated under vacuum to give 11 as
an orange powder (20 mg, 73%). 1H NMR (250 MHz, CDCl3): d=7.94
(d, J=8.0 Hz, 1H), 7.91 (d, J=8.0 Hz, 2H), 7.42–7.10 (m, 12H), 4.47 (dd,
J=8.0, 6.5 Hz, 1H), 3.71 (dd, J=14.5, 8.0 Hz, 2H), 3.57 ppm (dd, J=
14.5, 6.5 Hz, 2H); 13C NMR (63 MHz, [D6]DMSO): d=162.8, 159.3,
152.9, 152.8, 138.03, 138.01, 136.3, 136.0, 127.5, 127.4, 127.0, 123.6, 123.3,
118.0, 117.9, 111.2, 111.0, 45.9, 43.6 ppm; UV/Vis [CH3OH/Tris·HCl
20 mm pH 7.4, NaCl 150 mm 1:1]: lmax (e)=204 (106000), 248 (113000),
302 nm (9200mꢁ1 cmꢁ1); MS (CDI, NH3): m/z: 474 [M+H]+; elemental
analysis calcd (%) for C30H23N3O3·0.5H2O: C 74.67, H 5.01, N 8.71;
found: C 74.54, H 5.09, N 9.19.
2,2’-Methanediyl-bis(5-chloro-8-hydroxyquinoline) (14): Pale orange
powder (440 mg, 98%). 1H NMR (250 MHz, CDCl3): d=8.45 (d, J=
8.5 Hz, 2H), 8.12 (brs, 2H), 7.56 (d, J=8.5 Hz, 2H), 7.48 (d, J=8.0 Hz,
2H), 7.10 (d, J=8.0 Hz, 2H), 4.74 ppm (s, 2H); 13C NMR (63 MHz,
CDCl3): d=157.4, 151.0, 138.1, 134.2, 127.2, 125.0, 123.3, 120.4, 110.3,
47.5 ppm; UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]:
lmax (e)=258 (59900), 320 (7000), 460 (2000), 489 (2900), 518 nm
(2100mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 371 [M+H]+; elemental analysis
calcd (%) for C19H12Cl2N2O2·0.1Na2SO4: C 59.21, H 3.14, N 7.27; found:
C 59.27, H 2.58, N 7.05.
2,2’-(2,2-Propanediyl)-bis(5-chloro-8-hydroxyquinoline)
(15):
White
powder (55 mg, 93%). 1H NMR (250 MHz, CDCl3): d=8.37 (d, J=
9.0 Hz, 2H), 8.19 (brs, 2H), 7.50 (d, J=8.0 Hz, 2H), 7.32 (d, J=9.0 Hz,
2H), 7.13 (d, J=8.0 Hz, 2H), 2.0 ppm (s, 6H); 13C NMR (63 MHz,
CDCl3): d=165.2, 151.1, 137.3, 134.0, 127.2, 124.7, 121.9, 120.5, 110.2,
49.4, 27.9 ppm; UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm
1:1]: lmax (e)=258 (72500), 315 nm (8300mꢁ1 cmꢁ1); MS (DCI, NH3):
m/z: 399 [M+H]+; elemental analysis calcd (%) for C21H16Cl2N2O2: C
63.17, H 4.04, N 7.02; found: C 63.02, H 3.76, N 6.87.
2-{[8-Hydroxy-2-quinolinyl)amino]methyl}-8-quinolinol (12): A solution
of 2-amino-8-hydroxyquinoline (100 mg, 0.62 mmol) and 8-hydroxyquino-
line-2-carboxaldehyde (130 mg, 0.75 mmol) in 1,2-dichloroethane (8 mL)
was stirred during 1 h at room temperature then AcO3BHNa (291 mg,
1.29 mmol) was added. After 1.5 h, CH2Cl2 (50 mL) and saturated aque-
ous NaHCO3 (50 mL) were added and the product was extracted with
CH2Cl2 (2”120 mL), dried (Na2SO4), then the solvent was evaporated.
The product was dissolved in CH2Cl2 (75 mL) and one volume of hexane
was added. After filtration and reduction of the solvent volume, the
product was purified by chromatography (CH2Cl2, 0.5–1% CH3OH) to
give 12 as a clear brown powder (32 mg, 16%). 1H NMR (250 MHz,
[D6]DMSO): d=8.29 (d, J=8.5 Hz, 1H), 8.01 (t, J=4.5 Hz, 1H), 7.93 (d,
J=9.0 Hz, 1H), 7.57 (d, J=8.5 Hz, 1H), 7.39 (m, 2H), 7.08 (m, 4H), 6.91
(dd, J=1.5, 8.0 Hz, 1H), 5.08 ppm (d, J=4.5 Hz, 2H); 13C NMR
(75 MHz, [D6]DMSO): d=157.3, 156.0, 153.2, 150.8, 137.5, 137.5, 137.0,
137.0, 128.0, 127.4, 123.4, 122.2, 121.1, 118.04, 118.02, 114.0, 111.7, 111.0,
46.8 ppm; UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]:
2,2’-(1,2-Ethanediyl)-bis(5-chloro-8-hydroxyquinoline) (16): Pale green
solid (258 mg, 71%). 1H NMR (250 MHz, [D6]DMSO): d=9.73 (brs,
2H), 8.37 (d, J=8.5 Hz, 2H), 7.70 (d, J=8.5 Hz, 2H), 7.52 (d, J=8.0 Hz,
2H), 7.06 (d, J=8.0 Hz, 2H), 3.64 ppm (s, 4H); 13C NMR (100 MHz,
[D6]DMSO): d=161.5, 153.0, 139.2, 133.5, 127.4, 125.4, 124.4, 119.6,
112.1, 37.1 ppm. 1H/13C NMR correlations spots between C5 and H4 al-
lowed to attribute the halogen position; UV/Vis [dioxane/Tris·HCl 20 mm
pH 7.4, NaCl 150 mm 1:1]: lmax (e)=254 (75000), 313 nm (7200mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 385 [M+H]+; elemental analysis calcd (%) for
C20H14Cl2N2O2·0.1Na2SO4: C 60.14, H 3.53, N 7.01; found: C 59.92, H
3.07, N 6.65.
2,2’-(1,2-Ethanediyl)-bis(5-chloro-7-iodo-8-hydroxyquinoline) (17): To a
solution of 16 (300 mg, 0.78 mmol) in CH3OH (12 mL) under N2 was
added 97% H2SO4 (250 mL) to give a yellow suspension cooled on an ice
692
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
bath. N-Iodosuccinimide (418 mg, 1.86 mmol) was then slowly added.
The mixture was stirred for 15 min at 08C and at room temperature for
5 d. It was then poured onto ice and decolorized by addition of Na2S2O5
(450 mg). The suspension was neutralized with aqueous ammonia and
centrifuged. The supernatant was removed and the precipitate was dis-
solved in CH2Cl2 and washed with Tris·HCl buffer (0.1m, pH 7). The or-
ganic layer was evaporated under vacuum to afford 17 as a yellow
4.2 Hz, 1H), 8.06 (dd, J=1.5, 8.5 Hz, 1H), 7.59–7.51 (m, 3H), 7.36–7.24
(m, 5H), 5.40 (s, 2H), 3.92 ppm (s, 3H); 13C NMR (63 MHz, CDCl3): d=
151.3, 149.5, 142.9, 141.2, 137.4, 135.3, 127.9, 127.4, 127.1, 123.6, 122.9,
118.5, 114.8, 75.2, 56.1 ppm; MS (DCI, NH3): m/z: 266 [M+H]+; elemen-
tal analysis calcd (%) for C17H15NO2·0.05 CHCl3: C 75.49, H 5.59, N 5.16;
found: C 75.74, H 5.40, N 5.33.
8-(Benzyloxy)-7-methyloxyquinolin-N-oxide (23): To dry CH2Cl2 (11 mL)
at 48C under argon were successively added 22 (0.30 g, 1.13 mmol) and
m-chloroperbenzoic acid 77% wt (0.38 g, 1.69 mmol). After stirring for
48 h at room temperature, CH2Cl2 (40 mL) was added. The reaction mix-
ture was washed with saturated aqueous NaHCO3 (2”50 mL) and the
solvent was evaporated. The product was purified by chromatography
CH2Cl2/CH3OH 95:5 affording 23 as a brown oil (0.18 g, 58%). 1H NMR
(250 MHz, CDCl3): d=8.43 (dd, J=1.0, 6.0 Hz, 1H), 7.68–7.58 (m, 4H),
7.41–7.30 (m, 4H), 7.10 (dd, J=6.0, 8.5 Hz, 1H), 5.23 (s, 2H), 3.96 ppm
(s, 3H); 13C NMR (63 MHz, CDCl3): d=138.4, 128.9, 128.7, 128.2, 127.8,
127.6, 125.7, 124.9, 123.4, 119.5, 118.8, 116.8, 107.7, 75.2, 57.1 ppm; MS
(DCI, NH3): m/z: 282 [M+H]+; elemental analysis calcd (%) for
C17H15NO3: C 72.58, H 5.37, N 4.98; found: C 72.74, H 4.91, N 4.87.
1
powder (350 mg, 70%). H NMR (250 MHz, [D6]DMSO): d=8.36 (d, J=
8.5 Hz, 2H), 7.90 (s, 2H), 7.70 (d, J=8.5 Hz, 2H), 3.68 ppm (s, 4H);
13C NMR (63 MHz, [D6]DMSO): d=162.1, 153.1, 137.3, 134.4, 133.5,
124.6, 124.5, 120.2, 78.4, 37.1 ppm; UV/Vis [DMSO/Tris·HCl 20 mm pH
7.4, NaCl 150 mm 8:2]: lmax (e)=262 (69500), 319 nm (7900mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 637 [M+H]+; elemental analysis calcd (%) for
C20H12Cl2I2N2O2: C 37.71, H 1.90, N 4.40; found: C 38.15, H 2.04, N 4.15.
7-Bromo-2-methyl-8-hydroxyquinoline (20): Tert-butylamine (3.23 mL)
was stirred under argon in toluene (100 mL) during 2 h at room tempera-
ture with 4 Š molecular sieves (10 g). After cooling to ꢁ708C, N-bromo-
succinimide (5.48 g, 30.78 mmol) and 8-hydroxyquinaldine (5.00 g,
30.78 mmol) were added and the temperature was slowly increased to
208C (4 h). After filtration and washing with diethyl ether, the filtrate
was washed with H2O (3”50 mL), dried (Na2SO4) then the solvent was
evaporated. The obtained solid was refluxed in hexane (100 mL) then fil-
trated and dried to give 20 as a white powder (4.85 g, 67%). 1H NMR
(250 MHz, CDCl3): d=8.00 (d, J=8.5 Hz, 1H), 7.52 (d, J=9.0 Hz, 1H),
7.31 (d, J=8.5 Hz, 1H),7.16 (d, J=9.0 Hz, 1H), 2.72 ppm (s, 3H);
13C NMR (63 MHz, CDCl3): d=157.9, 149.2, 137.7, 136.2, 130.1, 125.4,
122.9, 118.2, 103.9, 24.8 ppm; MS (DCI, NH3): m/z: 238 [M+H]+; ele-
mental analysis calcd (%) for C10H8NOBr: C 50.45, H 3.39, N 5.88,
found: C 50.02, H 3.37, N 6.12.
2,2’-(Dimethylmalonatediyl)-bis(8-(benzyloxy)-7-methyloxyquinoline)
(24): To a solution of 23 (0.72 g, 2.56 mmol) in dry CH2Cl2 (7 mL) were
added dimethylmalonate (0.307 mL, 2.69 mmol) and acetic anhydride
(1.18 mL) and the mixture was stirred for 24 d protected from the light
and the humidity (CaCl2 trap). CH3OH (20 mL) was added and, after
15 h at ꢁ208C, a precipitate was removed by centrifugation. The solvent
of the supernatant was evaporated and the product was dissolved in
CH3OH (0.5 mL) and precipitated with H2O (3 mL). Supernatant was re-
moved and pellet was dissolved in CH2Cl2 (20 mL) and washed with H2O
(2”20 mL). After evaporation of the solvent, the product was purified by
preparative thin layer chromatography eluted with CH2Cl2/CH3OH 99:1.
Residual impurities were then removed by crystallization from CH3OH
General method for synthesis of 7-methyloxy-8-hydroxyquinoline deriva-
tives: To a solution of 7-bromo-8-hydroxyquinoline derivative (70 mm in
DMF) was added 10 equiv of CH3ONa (30% wt in CH3OH) and the
mixture was stirred for 10 min under argon. CuCl2.2H2O (0.3 equiv) was
added and the mixture was heated under reflux for 20 h. After cooling to
room temperature, H2O and EDTA (0.3 equiv) were added and the stir-
ring was carried on for 1 h. The solution was slightly acidified with
AcOH and then carefully basified with saturated NaHCO3. The aqueous
phase was extracted with CH2Cl2. The organic layer was dried (Na2SO4)
and the solvent was evaporated. The residue was purified by chromatog-
raphy using a gradient from CH2Cl2/CH3OH/AcOH 94:4:2 to CH2Cl2/
CH3OH 90:10. The fractions containing the product were combined and
washed with saturated aqueous NaHCO3 (3”100 mL). The organic phase
was dried (Na2SO4) and the solvent was evaporated under reduced pres-
sure affording desired product.
1
to give 24 as white needles (12.3 mg, 1.5%). H NMR (250 MHz, CDCl3):
d=7.93 (d, J=8.5 Hz, 2H), 7.63 (d, J=8.5 Hz, 2H), 7.59 (m, 4H), 7.47
(d, J=9.0 Hz, 2H), 7.40–7.28 (m, 8H), 5.41 (s, 4H), 3.95 (s, 6H),
3.84 ppm (s, 6H); 13C NMR (63 MHz, CDCl3): d=169.3, 157.5, 152.2,
142.2, 142.1, 138.2, 135.9, 128.4, 128.2, 127.7, 123.3, 123.0, 121.1, 116.0,
75.8, 75.2, 57.2, 53.0 ppm; MS (DCI, NH3): m/z: 659 [M+H]+; HRMS:
m/z: calcd for C39H35N2O8: 659.2393; found: 659.2399. Crystallization
from CH3OH provided single-crystals suitable for analysis by X-ray dif-
fraction as white needles.
2,2’-Methanediyl-bis(7-methyloxy-8-hydroxyquinoline) dihydrochloride
(25): Compound 24 (40 mg, 0.06 mmol) was refluxed in 37% aqueous
HCl (3 mL) for 1.5 h then the solvent was evaporated. The product was
dissolved in CH3OH then precipitated by pouring in diethyl ether
(3 mL). The precipitate was centrifuged and washed with diethyl ether
before to be dried under vacuum to give 24 as yellow crystals (28 mg,
7-Methyloxy-8-hydroxyquinoline (19): Grey powder (0.65 g, 40%).
1H NMR (250 MHz, CDCl3): d=8.77 (dd, J=1.5, 4.0 Hz, 1H), 8.11 (dd,
J=1.5, 8.5 Hz, 1H), 7.36 (m, 2H), 7.31 (dd, J=4.0, 8.5 Hz, 1H),
4.06 ppm (s, 3H); 13C NMR (63 MHz, CDCl3): d=148.6, 144.0, 139.7,
138.6, 136.0, 123.5, 119.7, 117.7, 116.4, 57.3 ppm; MS (DCI, NH3): m/z:
176 [M+H]+; elemental analysis calcd (%) for C10H9NO2·0.1H2O: C
67.86, H 5.24, N 7.91; found: C 67.82, H 5.11, N 7.95.
1
98%). H NMR (250 MHz, CD3OD): d=8.97 (d, J=8.0 Hz, 2H), 7.86 (s,
4H), 7.68 (d, J=8.0 Hz, 2H), 4.16 ppm (s, 6H); 13C NMR (63 MHz,
CDCl3): d=152.5, 150.4, 147.6, 135.1, 129.8, 123.9, 120.4, 119.8, 117.2,
56.4 ppm; UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]:
lmax (e)=203 (47800), 246 (34100, sh), 257 (49800), 334 nm
(3500mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 363 [M+H]+.
7-Methyloxy-2-methyl-8-hydroxyquinoline (21): White powder (480 mg,
60%). 1H NMR (250 MHz, CDCl3): d=7.95 (d, J=8.5 Hz, 1H), 7.26 (s,
2H), 7.15 (d, J=8.5 Hz, 1H), 4.03 (s, 3H), 2.69 ppm (s, 3H); 13C NMR
(63 MHz, CDCl3): d=157.6, 143.9, 139.3, 138.0, 136.0, 121.7, 120.6, 117.3,
115.2, 57.2, 25.1 ppm; MS (DCI, NH3): m/z: 190 [M+H]+; elemental
analysis calcd (%) for C11H11NO2: C 69.83, H 5.86, N 7.40, found: C
69.40, H 5.85, N 7.47.
2,2’-(1,2-Ethanediyl)-bis(7-methyloxy-8-quinolinol) (26):
A
solution
under argon of 21 (0.50 g, 2.64 mmol) in THF (10 mL) was stirred 0.5 h
with 4 Š molecular sieves (2.0 g) at room temperature. After cooling to
ꢁ958C, lithium diisopropylamine·THF (3.70 mL, 5.57 mmol, 1.5m in cy-
clohexane) was added over 10 min. The mixture was allowed to warm to
ꢁ508C, then cooled again to ꢁ908C and 1,2-dibromoethane (0.50 mL,
5.82 mmol) was added and the mixture was allowed to warm to room
temperature before adding H2O (5 mL) and AcOH (0.50 mL). After 1 h
of stirring, molecular sieves were removed by filtration. Saturated aque-
ous NaHCO3 (10 mL), H2O (10 mL) and diethyl ether (40 mL) were
added and the organic phase was collected and dried (Na2SO4). Solvent
was evaporated and the solid obtained was recrystallized from CH3OH
8-(Benzyloxy)-7-methyloxyquinoline (22): Compound 19 (1.10 g,
6.29 mmol) and K2CO3 (1.30 g, 9.43 mmol) in dry CH3CN (30 mL) were
stirred for 10 min under argon. Benzyl chloride (0.87 mL, 7.54 mmol) was
added and the mixture was heated under reflux overnight. After cooling
to room temperature, the precipitate was filtered, washed with CH2Cl2
and the filtrate was dried under reduced pressure. The product was dis-
solved in CH2Cl2 (50 mL) and washed with aqueous 2m NaOH (3”
50 mL) then H2O (50 mL). The organic layer was dried (Na2SO4) and the
solvent was evaporated under reduced pressure affording 22 as a brown
oil (1.28 g, 77%). 1H NMR (250 MHz, CDCl3): d=8.94 (dd, J=1.5,
(10 mL) to give 26 as
a
white powder (306 mg, 62%). 1H NMR
(250 MHz, CDCl3): d=8.00 (d, J=8.5 Hz, 2H), 7.28 (s, 4H), 7.24 (d, J=
8.5 Hz, 2H), 4.04 (s, 6H), 3.57 ppm (s, 4H); 13C NMR (63 MHz, CDCl3):
d=160.1, 144.0, 139.3, 137.9, 136.3, 121.9, 120.4, 117.5, 115.5, 57.2,
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
693

M. PitiØ et al.
37.1 ppm; UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]:
lmax (e)=204 (83000), 208 (53200, sh), 252 (97200), 338 nm
(6600mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 377 [M+H]+; elemental analysis
calcd (%) for C22H20N2O4·0.2H2O: C 69.53, H 5.41, N 7.37, found: C
69.40, H 5.30, N 7.47.
4·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=205 (64000), 255 (47600), 276 (34400), 415 nm (2900mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 400 [LCuIIꢁH]+, 417 [LCuIINH4ꢁ2H]+.
5·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=245 (41100), 306 (27900), 348 (12100, sh), 485 nm (2500mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 378 [LCuIIꢁH]+, 395 [LCuIINH4ꢁ2H]+.
Single-crystals of copper(II) or zinc(II) complexes of ligand 3
7·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (57700), 259 (68000), 268 (46200, sh), 392 nm (4900mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 378 [LCuIIꢁH]+, 395 [LCuIINH4ꢁ2H]+.
Complex 27: To a solution of 3 (10.0 mg, 0.03 mmol) in ethanol (0.4 mL)
was added Cu(OAc)2·H2O (6.05 mg, 0.03 mmol) dissolved in H2O
(0.4 mL). The mixture was stirred for 0.5 h at 608C then precipitated at
48C. Supernatant was removed after a centrifugation step and the pellet
was washed twice with H2O before to be dried at 1108C. Product was
then recrystallized in CH3OH to give 27 as green needles.
9·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=204 (56000), 260 (71800), 374 nm (4800mꢁ1 cmꢁ1); MS (DCI, NH3):
m/z: 406 [LCuIIꢁH]+, 423 [LCuIINH4ꢁ2H]+.
12·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=267 (45600), 286 (26800, sh), 320 (4700, sh), 389 nm (2700mꢁ1 cmꢁ1);
MS (ESMS): m/z: 379 [LCuIIꢁH]+, 415 [LCuCl]+.
Complex 28: To a solution of 3 (11.7 mg, 0.035 mmol) in DMF (0.4 mL)
was added CuSO4·5H2O (8.85 mg, 0.035 mmol) dissolved in H2O
(0.4 mL) to give a green solution. Slow evaporation of the solvent at
room temperature gave 28 as green needles.
14·CuII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=256 (38300), 278 (35900), 310 (10300), 342 (7300), 398 nm
(4000mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 434 [LCuIIꢁH]+, 451
[LCuIINH4ꢁ2H]+.
Complex 29: To
a solution of 3 (10.6 mg, 0.032 mmol) in CH3OH
(0.4 mL) was added CuCl2·2H2O (5.53 mg, 0.032 mmol) dissolved in H2O
(0.4 mL). A green precipitate was obtained. It was dissolved by heating
then CH3OH (4 mL) was added and slow evaporation of the solvent gave
29 as green-yellow crystals.
15·CuII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=261 (59500), 274 (46900, sh), 347 (4700), 409 nm (5800mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 462 [LCuIIꢁH]+, 479 [LCuIINH4ꢁ2H]+; elemental
analysis calcd (%) for C21H14N2O2Cl2Cu: calcd C 54.74, H 3.06, N 6.08;
found: C 54.50, H 2.84, N 5.92.
Complex 30: To a solution of 3 (51.7 mg, 0.156 mmol) in THF (2.0 mL)
was added ZnSO4·7H2O (45.08 mg, 0.156 mmol) dissolved in H2O
(2.0 mL). THF (2.0 mL) was added and the solution was slowly evaporat-
ed at room temperature until the formation of a biphasic system. Super-
natant was removed and the volume of the lower phase was reduced. A
minimum volume of DMSO was added to obtain a homogenous solution
at 60–708C. The mixture was stand at room temperature until the forma-
tion of single-crystals of complex 30 as white needles. The overall quality
of the data was poor due to the weak diffracting crystal, which leads to a
high Rint value (11826/3679 collected/independents reflections, Rint
0.3374, R1[F2 >4s(F2)]=0.0726, wR2=0.1582). We have limited the dis-
cussion of the main features of the complex. Complex 30 crystallizes in
the orthorhombic space group Pca21 with the cell parameters: a=
19.690(3), b=10.036(2), c=10.220(2) Š.
16·CuII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=264 (64700), 276 (46700), 413 nm (5300mꢁ1 cmꢁ1); MS (DCI, NH3):
m/z: 448 [LCuIIꢁH]+, 465 [LCuIINH4ꢁ2H]+.
17·CuII: UV/Vis [DMSO/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 8:2]: lmax
(e)=278 (56000), 340 (55000), 392 (6200), 420 nm (5000mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 700 [LCuIIꢁH]+, 717 [LCuIINH4ꢁ2H]+.
26·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=212 (65300), 263 (65300), 275 (47300, sh), 319 (4500), 421 nm
(5900mꢁ1 cmꢁ1); MS (ESMS): m/z: 438 [LCuIIꢁH]+, 475 [LCuCl]+.
1·ZnII: MS (DCI, NH3): m/z: 365 [LZnIIꢁH]+, 382 [LZnIINH4ꢁ2H]+.
3·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (78700), 258 (74100), 268 (46700), 378 nm (4600mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 393 [LZnIIꢁH]+, 410 [LZnIINH4ꢁ2H]+.
Determination of the metal to ligand stoichiometry: UV-visible absorp-
tion spectra and spectrophotometric titrations were performed in the
presence of 20 mm Tris·HCl pH 7.4 buffer, 150 mm NaCl since it was the
solvent used during Ab dissolution assays. Organic solvent (CH3OH, di-
oxane or DMSO) was added in order to have a good solubility of ligands
and metal complexes at the concentrations used in these experiments. To
a 15 mm solution of ligand were added aliquots of concentrated solutions
of CuCl2 or ZnCl2 in order to induce negligible volume variations. Upon
these additions changes in UV-visible absorption spectra were immedi-
ately observed and were stable between two additions as a result of a fast
complexation process. The metal to ligand stoichiometries were also con-
firmed from elemental analysis of significant examples and are included
here.
4·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=205 (43300), 258 (50500), 274 (30900), 404 nm (3000mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 401 [LZnIIꢁH]+.
5·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=238 (31100), 298 (26200), 346 (10500, sh), 473 nm (300mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 379 [LZnIIꢁH]+, 396 [LZnIINH4ꢁ2H]+.
7·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (55500), 258 (63600), 268 (43400), 374 nm (4000mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 379 [LZnIIꢁH]+, 396 [LZnIINH4ꢁ2H]+.
9·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (63600), 260 (65600), 368 nm (4400mꢁ1 cmꢁ1); MS (DCI, NH3):
m/z: 407 [LZnIIꢁH]+, 424 [LZnIINH4ꢁ2H]+.
Preparation of metal complexes for mass spectrometry characterization:
Solutions of ligands in CH3OH or dioxane were metalated in the pres-
ence of 1 equiv of metal ion during 1 h at room temperature before to be
evaporation of the solvent. CuCl2 (for 1, 4, 5, 9, 12, 13, 25 and 26),
Cu(OAc)2 (for 3, 7, 14, 15, 16 and 17), Zn(OAc)2 (for 1, 3, 7, 14, 15, and
16) or ZnCl2 (for 4, 5, 9, 12, 13, 17, 25 and 26) were used. Control UV/
Vis spectrophotometric analyses of the different complexes are the same
than those obtained during titration experiments.
1·CuII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=256 (41900), 272 (32200, sh), 304 (8800), 338 (5100), 376 nm
(3100mꢁ1 cmꢁ1); MS (DCI, NH3): m/z: 364 [LCuIIꢁH]+, 381
[LCuIINH4ꢁ2H]+; elemental analysis calcd (%) for C19H12N2O2Cu: calcd
C 52.73, H 2.33, N 6.47; found: C 52.52, H 1.64, N 6.48.
12·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=267 (41400), 287 (26900, sh), 320 (5600, sh), 372 nm
(3100mꢁ1 cmꢁ1); MS (ESMS): m/z: 380 [LZnIIꢁH]+, 415 [LZnCl]+.
14·ZnII: MS (DCI, NH3): m/z: 435 [LZnIIꢁH]+, 452 [LZnIINH4ꢁ2H]+.
15·ZnII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=265 (69200), 275 (45700), 346 (5800), 405 nm (6100mꢁ1 cmꢁ1); MS
(DCI, NH3): m/z: 463 [LZnIIꢁH]+, 482 [LZnIINH4ꢁ2H]+.
16·ZnII: UV/Vis [dioxane/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=264 (57800), 272 (43600, sh), 390 nm (5400mꢁ1 cmꢁ1); MS (DCI,
NH3): m/z: 449 [LZnIIꢁH]+, 466 [LZnIINH4ꢁ2H]+.
3·CuII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (70800), 254 (58900), 268 (42400, sh), 383 nm (4300mꢁ1 cmꢁ1);
MS (DCI, NH3): m/z: 392 [LCuIIꢁH]+, 409 [LCuIINH4ꢁ2H]+; elemental
analysis calcd (%) for C21H16N2O2Cu·0.3C2H4O2: calcd C 63.33, H 4.16,
N 6.84; found: C 63.30, H 3.50, N 6.82.
17·ZnII: UV/Vis [DMSO/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 8:2]: lmax
(e)=276 (54200), 340 (5800), 352 (7700), 397 nm (5100mꢁ1 cmꢁ1); MS
(ESMS): m/z: 735 [LZnCl]ꢁ.
25·ZnII: MS (ESMS): m/z: 425 [LZnIIꢁH]+.
694
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696

anti-Alzheimer Drugs
FULL PAPER
26·ZnII: UV/Vis [CH3OH/Tris·HCl 20 mm pH 7.4, NaCl 150 mm 1:1]: lmax
(e)=202 (66900), 264 (92500), 276 (47100, sh), 318 (5500), 398 nm
(6700mꢁ1 cmꢁ1); MS (ESMS): m/z: 439 [LZnIIꢁH]+.
cubated 5 min under stirring. The amount of H2O2 produced was quanti-
fied using the AmplexRed H2O2 Assay Kit (Molecular Probes) from an
analytical grade H2O2 standard curve. Some experiments received also
known quantities of H2O2 before the addition of AmplexRed.
Estimation of the metal to ligand affinity constants: Solutions (15 mm) of
studied ligand (Ls), EDTA (competitor chelator Lc) and metal ion (M) in
1:1:1 ratio were analyzed spectrophotometrically at 208C to determine
the concentration of Ls, MLs, Lc or MLc species at the equilibrium. Ex-
periments were performed in 20 mm Tris·HCl pH 7.4, 150 mm NaCl
buffer where an organic solvent (CH3OH, dioxane or DMSO) was added
in order to have a good solubility of ligands and metal complexes at the
concentrations used. Controls showed that all ligands and complexes
studied were in accordance with the Beer–Lambert law in the experimen-
tal conditions used. In these experimental conditions, and since all li-
gands studied Ls form MLs species, complexation reactions can be sum-
Acknowledgement
CNRS and PALUMED are gratefully acknowledged for their financial
support. Dr F. CoslØdan (PALUMED) is acknowledged for fruitful dis-
cussions.
Ks
Kc
ƒ
ƒ!
ƒ!
marized as: Ls + M
ML with K = [MLs]/[Ls][M] and Lc + M
s s
ƒ
MLc with Kc = [MLc]/[Lc][M]
[1]a) M. Mattson, Nature 2004, 430, 631–639; b) M. Citron, Nat. Rev.
Neurosci. 2004, 5, 677–685; c) E. Gaggelli, H. Kozlowski, D. Valen-
sin, G. Valensin, Chem. Rev. 2006, 106, 1995–2044.
[2]a) C. Opazo, X. Huang, R. Cherny, R. Moir, A. Roher, A. White, R.
Cappai, C. Masters, R. Tanzi, N. Inestrosa, A. Bush, J. Biol. Chem.
2002, 277, 40302–40308; b) G. Ciccotosto, D. Tew, C. Curtain, D.
Smith, D. Carrington, C. Masters, A. Bush, R. Cherny, R. Cappai,
K. Barnham, J. Biol. Chem. 2004, 279, 42528–42534.
Kc is the Kapp value for the competitive chelator at pH 7.4 where the ex-
periments were performed. It was determined as in reference 15. It can
be deduced that at the equilibrium: Ks = Kc[MLs][Lc]/[MLc][Ls].
The initial concentration was C. All the metal is chelated during the ex-
periments and [MLs] + [MLc] = C = [MLs] + [Ls] = [MLc] + [Lc] . At
the equilibrium [MLs] = x% of C, then [Ls] = (1ꢁx)% of C, [MLc] =
(1ꢁx)% of C and [Lc] = x% of C. Therefore, the apparent affinity con-
stant of Ls for the metal ion M, at pH 7.4, is: Ks = (Kc x2)/(1ꢁx)2.
logD7.4 determination: Ligand (2.0 mg) was dissolved in 1-octanol
(2.00 mL) then Tris·HCl buffer (2.00 mL, 20 mm, pH 7.4), 150 mm NaCl
was added. After a 3 min vortex at room temperature followed by centri-
fugation for 5 min at 5500 g, the concentration of ligand in each layer
was determined by UV/Vis spectrophotometry. Samples from 1-octanol
layer were repartitioned until consistent partition coefficient values were
obtained. The measurement was carried out in triplicate.
[3]a) R. Cherny, J. Legg, C. McLean, D. Fairlie, X. Huang, C. Atwood,
K. Beyreuther, R. Tanzi, C. Masters, A. Bush, J. Biol. Chem. 1999,
274, 23223–23228; b) R. Cherny, K. Barnham, T. Lynch, I. Volitakis,
Q. Li, C. McLean, G. Multhaup, K. Beyreuther, R. Tanzi, C. Mas-
ters, A. Bush, J. Struct. Biol. 2000, 130, 209–216; c) J. Lee, J. Fried-
man, I. Angel, A. Kozak, J. Koh, Neurobiol. Aging 2004, 25, 1315–
1321; d) A. Dedeoglu, K. Cormier, S. Payton, K. Tseitlin, J. Krem-
sky, L. Lai, X. Li, R. Moir, R. Tanzi, A. Bush, N. Kowall, J. Rogers,
X. Huang, Exp. Gerontol. 2004, 39, 1641–1649; e) Z. Cui, P. Lock-
man, C. Atwood, C. Hsu, A. Gupte, D. Allen, R. Mumper, Eur. J.
Pharm. Biopharm. 2005, 59, 263–272; f) C. Boldron, I. van der Au-
wera, C. Deraeve, H. Gornitzka, S. Wera, M. PitiØ, F. van Leuven,
B. Meunier, ChemBioChem 2005, 6, 1976–1980.
[4]a) R. Cherny, C. Atwood, M. Xilinas, D. Gray, W. Jones, C. McLean,
K. Barnham, I. Volitakis, F. Fraser, Y. Kim, X. Huang, L. Goldstein,
R. Moir, J. Lim, K. Beyreuther, H. Zheng, R. Tanzi, C. Masters, A.
Bush, Neuron 2001, 30, 665–676; b) C. Ritchie, A. Bush, A. Mackin-
non, S. Macfarlane, M. Mastwyk, L. McGregor, L. Kiers, R. Cherny,
Q. Li, A. Tammer, D. Carrington, C. Mavros, I. Volitakis, M. Xilinas,
D. Ames, S. Davis, K. Beyreuther, R. Tanzi, C. Masters, Arch.
Neurol. 2003, 60, 1685–1691.
[5]C. Deraeve, M. PitiØ, B. Meunier, New J. Chem. 2007, 31, 193–195.
[6]Y. Yamamoto, A. Miura, A. Kawamata, M. Miura, S. Takei, Bull.
Chem. Soc. Jpn. 1978, 51, 3489–3495.
[7]R. E. Banks, N. J. Lawrence, A. L. Popplewell, J. Chem. Soc. Chem.
Commun. 1994, 343–344.
[8]M. Albrecht, O. Blau, K. Witt, E. Wegelius, M. Nissinen, K. Rissa-
nen, R. Frçhlich, Synthesis 1999, 10, 1819–1829.
[9]C. Kitamura, N. Maeda, N. Kamada, M. Ouchi, A. Yoneda, J. Chem.
Soc. Perkin Trans. 1 2000, 781–785.
Reagents for the experiments involving Ab1–42 amyloid peptide: Before
use, MilliQ H2O (Millipore) and all buffers were treated with Chelex-100
resin (Biorad) (5.0 mgmLꢁ1) and filtered through a 0.2 mm filter (What-
man) to remove any metal ion or particulate matter. Ab1–42 amyloid pep-
tide was synthesized, purified (up to 95% purity) and characterized by
HPLC analysis and MALDI-TOF mass spectrometry by CNRS-IPBS
(Toulouse, France). It was dissolved and quantified as previously de-
scribed.[5,22] Ligands were used as chloride salts. Excepted for 1 and 25
that were synthesized as chloride salts, they were generated by addition
of 1 equiv of HCl per nitrogenous functions of ligand dissolved in
DMSO. After evaporation of the solvent, the chloride salts were dis-
solved at the desired concentration in DMSO and stored at ꢁ208C. All
experiments involving the peptide were performed in duplicate when the
discrepancy between experimental values was less than 5%. In the other
cases and for major results triplicates were done and standard deviations
were calculated. Stirrings were performed with a Thermomixer.
Inhibition of Ab precipitation: Final concentrations were given. Ab1–42
(500 mL, 5 mm) was incubated in Tris·HCl buffer (20 mm, pH 7.4), 150 mm
NaCl with CuCl2 or ZnCl2 (20 mm) for 1 h at 378C under stirring at
1400 rpm. Tested ligand (200 mm) dissolved in DMSO (50 mL) was added
and the samples were incubated for one additional hour at 378C under
stirring at 1400 rpm. Samples were then centrifuged for 20 min at 5500 g
and 500 mL of supernatant was removed. The test tube containing residu-
al supernatant and pellet received 450 mL of experimental buffer/DMSO
91:9. Then both supernatant and pellet were analyzed for protein concen-
tration by MicroBCA Assay (Pierce). This double quantification allowed
to monitor the initial quantity of peptide and to confirm the validity of
the measurements. In the case of 1, 5, 12, 13, 14, 25, 26 and 8-hydroxyqui-
naldine that showed an absorption at 562 nm in the dosage conditions,
blanks were performed in the presence of the tested ligand and the stud-
ied metal ion. Some experiments were performed with other ratios of
CuII (2.5 equiv) and ligand (2.5 or 5 equiv) per Ab1–42.
[10]H. Gershon, M. W. Mc Neil, J. Heterocycl. Chem. 1972, 9, 659–667.
[11]a) F. Yang, G. P. Lim, A. N. Begum, O. J. Ubeda, M. R. Simmons,
S. S. Ambegaokar, P. Chen, R. Kayed, C. G. Glabe, S. A. Frautschy,
G. M. Cole, J. Biol. Chem. 2005, 280, 5892–5901; b) Y. Porat, A.
Abramowitz, E. Gazit, Chem. Biol. Drug. Des. 2006, 67, 27–37.
[12]a) D. E. Pearson, R. D. Wysong, C. V. Breder, J. Org. Chem. 1967,
32, 2358–2360; b) G. E. Collis, A. K. Burrell, K. D. John, P. G. Plieg-
er, Acta Crystallogr. Sect. C 2003, 59, o443-o444.
[13]a) M. Numazawa, Y. Ogura, J. Chem. Soc. Chem. Commun. 1983,
533–534; b) D. Nobel, J. Chem. Soc. Chem. Commun. 1993, 419–
420; c) D. Planchenault, R. Dhal, J. Robin, Tetrahedron 1995, 51,
1395–1404.
Hydrogen peroxide assay: Final concentrations were given. Ab1–42 (0.2 or
0.53 mm) and CuCl2 (0.4 mm) were incubated for 1 h at 378C under stirring
at 1400 rpm in 375 mL of sodium phosphate buffer (20 mm, pH 7.4). Then
2 mL of tested ligand in DMSO (0 to 0.8 mm) were added. After one more
hour of incubation, 2 mL of sodium ascorbate (10 mm) were added and in-
[14]a) The National Institute of Standards and Technology Standard
Reference Data base 46’’ NIST (Critically selected Stability Con-
stants of Metal Complexes Database, version 4.0, US Departement
of Commerce); b) L. G. Sillen, A. E. Martell, Stability constants of
Chem. Eur. J. 2008, 14, 682 – 696
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
695

M. PitiØ et al.
metal-ion complexes, The Chemical Society London Publication,
1971.
[22]C. Deraeve, M. PitiØ, B. Meunier, J. Inorg. Biochem. 2006, 100,
2117–2126.
[15]a) A. Ringblom, Complexation in Analytical Chemistry, Interscience,
New York, 1963; b) G. Schwarznbach, H. Flaschka, Complexometric
Titrations, Meuthuen, New York, 1969.
[16]H. Gershon, M. W. McNeil, S. G. Schulman, J. W. Parkes, Anal.
Chim. Acta 1972, 62, 43–47.
[17]M. Di Vaira, C. Bazzicalupi, P. Orili, L. Messori, B. Bruni, P. Zatta,
Inorg. Chem. 2004, 43, 3795–3797.
[18]F. Kanamaru, K. Ogawa, I. Nitta, Bull. Chem. Soc. Jpn. 1963, 36,
422–427.
[19]N. Okabe, H. Saishu, Acta Crystallogr. Sect. E Acta Cryst. 2001, 57,
m714–m716.
[20]a) M. Shoda, H. Gershon, D. Bray, D. D. Clarke, Monatsh. Chem.
1998, 129, 843–853; b) F. Jian, Y. Wang, L. Lu, X. Yang, X. Wang, S.
Chantrapromma, H. K. Fun, I. Abdul Razak, Acta Crystallogr. Sect.
C 2001, 57, 714–716.
[23]G. M. Sheldrick, Acta Crystallogr. Sect. A 1990, 46, 467–473.
[24]SIR92—A program for crystal structure solution: A. Altomare, G.
Cascarano, C. Giacovazzo and A. Guagliardi, J. Appl. Crystallogr.
1993, 26, 343–350.
[25]SHELX97 [Includes SHELXS97, SHELXL97, CIFTAB—] Programs
for Crystal Structure Analysis (Release 97-2). G. M. Sheldrick, Insti-
tut fꢁr Anorganische Chemie der Universitꢂt, Tammannstrasse 4,
3400 Gçttingen (Germany), 1998.
[26]WINGX—1.63 Integrated System of Windows Programs for the So-
lution, Refinement and Analysis of Single Crystal X-Ray Diffraction
Data: L. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[27]International tables for X-ray crystallography, Vol. IV, Kynoch press,
Birmingham (Great Britain), 1974.
[28]ORTEP32 for Window: L. J. Farrugia, J. Appl. Crystallogr. 1997, 30,
565.
[21]H. van de Waterbeemd, E. Gifford, Nat. Rev. Drug Discovery 2003,
2, 192–204.
Received: July 4, 2007
Published online: October 30, 2007
696
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 682 – 696