Synthesis and biological evaluation of deferiprone-resveratrol hybrids as antioxidants, Aβ1–42aggregation inhibitors and metal-chelating agents for Alzheimer's disease

Article abstract of DOI:10.1016/j.ejmech.2016.12.045

A series of deferiprone-resveratrol hybrids have been designed and synthesized as multitarget-directed ligands (MTDLs) through merging the chelating moiety 3-hydroxypyridin-4-one into the structure of resveratrol, a natural antioxidant agent and β-amyloid peptide (Aβ) aggregation inhibitor. The in?vitro biological evaluation revealed that most of these newly synthesized compounds exhibited good inhibitory activity against self-induced Aβ_1–42aggregation, excellent antioxidant activity and potent metal chelating capability. Compounds 3i and 4f were identified as the most promising MTDLs with triple functions, possessing micromolar IC₅₀values for Aβ_1–42aggregation inhibition, greater 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS^?+) scavenging activity than Trolox and similar pFe(III) values to that of deferiprone.

Full text of DOI:10.1016/j.ejmech.2016.12.045

Accepted Manuscript
Synthesis and biological evaluation of deferiprone-resveratrol hybrids as antioxidants,
Aβ
aggregation inhibitors and metal-chelating agents for Alzheimer's disease
1–42
Ping Xu, Minkui Zhang, Rong Sheng, Yongmin Ma
PII:
S0223-5234(16)31051-0
10.1016/j.ejmech.2016.12.045
EJMECH 9136
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 8 November 2016
Revised Date: 19 December 2016
Accepted Date: 21 December 2016
Please cite this article as: P. Xu, M. Zhang, R. Sheng, Y. Ma, Synthesis and biological evaluation of
deferiprone-resveratrol hybrids as antioxidants, Aβ
aggregation inhibitors and metal-chelating
–42
1
agents for Alzheimer's disease, European Journal of Medicinal Chemistry (2017), doi: 10.1016/
j.ejmech.2016.12.045.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Synthesis and biological evaluation of deferiprone-resveratrol hybrids as
antioxidants, Aβ1-42 aggregation inhibitors and Metal-chelating agents for
Alzheimer’s disease
Ping Xu, Minkui Zhang, Rong Sheng, Yongmin Ma

ACCEPTED MANUSCRIPT
Synthesis and biological evaluation of deferiprone-resveratrol hybrids as
antioxidants, Aβ1-42 aggregation inhibitors and metal-chelating agents for
Alzheimer’s disease
a
b
b*
a∗
Ping Xu , Minkui Zhang , Rong Sheng , Yongmin Ma
a
College of Pharmaceutical Science, Zhejiang Chinese Medical University, Hangzhou, Zhejiang, P R
China, 311402
College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang, P. R. China, 310058
b
Abstract
A series of deferiprone-resveratrol hybrids have been designed and synthesized as
multitarget-directed ligands (MTDLs) through merging the chelating moiety 3-hydroxypyridin-4-one
into the structure of resveratrol, a natural antioxidant agent and β-amyloid peptide (Aβ) aggregation
inhibitor. The in vitro biological evaluation revealed that most of these newly synthesized
compounds exhibited good inhibitory activity against self-induced Aβ1-42 aggregation, excellent
antioxidant activity and potent metal chelating capability. Compounds 3i and 4f were identified as
the most promising MTDLs with triple functions, possessing micromolar IC50 values for Aβ1-42
•+
aggregation inhibition, greater 2,2’-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS )
scavenging activity than Trolox and similar pFe(III) values to that of deferiprone.
Keywords: Resveratrol derivatives; 3-hydroxypyridin-4-one; beta amyloid aggregation; antioxidant;
metal chelator
1. Introduction
Alzheimer’s disease (AD) is a multifactorial neurodegenerative disorder, which causes a
progressive impairment in functional performance, and finally leads to death with irreversible
damage to the brain. It has become the fifth leading cause of death for the population aged over 65 in
the developed countries [1, 2]. Among all the factors related to AD, the accumulation and
aggregation of β-amyloid peptide (Aβ), the associated biometal ion dyshomeostasis and oxidative
stress have been demonstrated to be involved with the pathogenesis of AD [3, 4]. There are two
principal forms of Aβ (Aβ1-40, Aβ1-42) in the brain tissue of AD patients, Aβ1-42 is responsible for the
initial self-aggregation of Aβ, and the resulting β-amyloid oligomers and fibrils are toxic to
neurons [5-7]. In vitro studies show that the metals such as Fe(III) can bind to Aβ1-42 via His 6, Asp7,
∗
Corresponding author. Email address: yongmin.ma@zcmu.edu.cn; Tel: (86) 571 61768146

ACCEPTED MANUSCRIPT
Tyr10 and His 14 facilitating the aggregation of Aβ [5, 8]. Remarkably high concentrations of copper,
zinc and iron ions also have been found to co-localize with amyloid deposits in AD patients
compared to age-matched controls [9]. In addition, the redox active iron and copper ions can
generate reactive oxygen species (ROS), which will cause oxidative damage and may trigger the
death of neuron [10]. To combat the complex pathogenesis of AD, multitarget-directed ligands
(MTDLs), which interfere with two or more causes of AD simultaneously, may achieve better
therapeutic efficacy with complementary mechanisms of action [11]. Therefore, we envisioned that
the combination of three functions (Aβ aggregation inhibition, metal chelation and anti-oxidant
activity) into one molecule may provide a promising therapeutic strategy for AD therapy [10].
To design novel MTDLs which target AD pathological factors, a pharmacophore-combination
strategy was used with resveratrol and deferiprone as the lead compounds. Resveratrol, a native
stilbene derivative with multiple activities has been advanced into phase II clinical trial for the
treatment of AD due to its efficient Aβ aggregation inhibitory and radical scavenging activity [12-15].
Deferiprone, an orally active 3-hydroxypyridin-4-one (HPO) derivative, has been used clinically to
chelate iron in iron-overloaded thalassaemia patients for many years [16, 17]. According to the
theory of hard and soft acid and bases (HSAB), chelators containing soft donor atoms such as
nitrogen and sulfur prefer binding Fe(II) while chelators containing hard donor atoms such as oxygen
favor Fe(III). The HPO moiety favors the binding of both iron and copper in the most oxidized form
(Fe(III) and Cu(II)), which in turn prevents redox cycling [18]. Therefore, we describe here the
design, synthesis and biological evaluation of a series of novel deferiprone-resveratrol hybrids, in
which one of the benzene rings of resveratrol is replaced by the metal chelating moiety HPO (Figure
1). The designed hybrids may exert reduced Aβ aggregation and antioxidant capacity through its
stilbene-mimic scaffold and metal chelating activity via the HPO moiety. To investigate the SAR of
these new hybrids and find optimal candidates for further development, substituents on the benzene
ring and the 1-position of HPO ring were systemically modified.

ACCEPTED MANUSCRIPT
Figure 1. Rational design of deferiprone-resveratrol hybrids as MTDLs
2
．Results and discussion
2.1. Design and synthesis of resveratrol analogues
The synthetic route of the deferiprone-resveratrol hybrids 3a-3k, 4a-4k is shown in Scheme 1.
Commercial available kojic acid 5 was used as the starting material. It was treated with benzyl
chloride, thionyl chloride and triphenylphosphine subsequently to get the key intermediate 8, an
active Wittig reagent. Intermediate 8 was condensed with substituted benzaldehydes to afford 9a-9h,
followed by substitution by ammonium hydroxide or methylamine to yield 10a-10h and 11a-11h
respectively. Finally, the benzyl protecting groups was removed by hydrochloric acid to obtain the
target compounds 3a-3h and 4a-4h. In addition, intermediates 10f-10h, 11f-11h were reacted with
BBr to afford the corresponding molecules 3i-3k and 4i-4k. Prediction of BBB penetration was
3
s
calculated using Molinspiration property engine (Table 1 , supporting information). The majority of
compounds fit within the rules except 3e and 3k, in which the logBB values are close to the
minimum margin (see supporting information).

ACCEPTED MANUSCRIPT
Scheme 1.Synthesis of target compounds. Reagents and conditions: (a) BnCl, MeOH, reflux, 16h; (b) SOCl ,2h; (c)
2
PPh , CHCl , 24h; (d) NaOH; (e) NH ·H O; (f) MeNH ; (g) 6N HCl or BBr .
3
3
3
2
2
3
2
.2. Fe(III) chelating effect
Deferiprone has a higher affinity for Fe(III) than for that of either Cu(II) or Zn(II). The cumulative
stability constants of deferiprone for Fe(III), Cu(II) and Zn(II) are 37.2, 21.7 and 13.5 respectively
19]. The pFe(III) value, which is defined as the negative logarithm of free iron concentration in
[
solution, is more useful parameter than the conventional stability constant to assess the ligand
affinity for the metal under biological conditions [19]. The pFe(III) values of all the synthesized
compounds were determined by a fluorescence method with deferiprone as reference compound [20].
Most of these synthesized compounds exhibit excellent pFe(III) values ranging from 18.44 to 21.62
(Table 1). The 3, 5-dimethoxy substituted compounds 3h and 4h displayed the most potent iron
chelating activity with pFe(III) values of 21.62 and 21.18, respectively.
2.3. Anti-oxidative activity
The reduction of oxidative stress is an essential aspect of these deferiprone-resveratrol hybrids.
The ABTS (2,2’-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)) radical scavenging method [21, 22]
was employed to determine the antioxidant capacity of these compounds with Trolox as a positive

ACCEPTED MANUSCRIPT
control. Most of compounds exhibited moderate to good radical scavenging activities with IC50
values, ranging from 0.88 to 19.2 µΜ (Table 1). There are eleven compounds with superior
antioxidant activity to that of trolox (IC = 3.89 µM). As expected, the compounds which possess
5
0
one or two hydroxyl groups on the benzene ring (3i, 4i, 3j, 4j, 3k, 4k) demonstrated greater
antioxidant activities than trolox with IC values ranging from 0.88 to 2.21 µΜ. The modification of
5
0
the R moiety from hydrogen to methyl generally increases the antioxidant capacity (Table 1).
1
Table 1. Biological evaluation of compounds 3a-3k and 4a-4k.
a
Compounds R1 R2
pFe(III)
ABTS assay Inhibition of Aβ
Aβ1-42
1-42
c
b
d
IC (µM)
aggregation (%)
IC (µM)
5
0
50
3
4
3
4
3
4
3
4
3
4
3
4
3
4
3
4
3
4
3
4
3
4
a
a
b
b
c
c
d
d
e
e
f
H
H
20.99±0.43
19.77±0.23
19.51±0.036
20.80±0.17
20.24±0.05
20.12±0.40
20.06±0.13
19.24±0.056
19.54±0.19
18.44±0.16
20.12±0.26
19.00±0.16
20.22±0.16
20.58±0.18
4.03±0.53
2.11±0.88
11.5±0.85
5.72±2.71
6.29±0.57
1.31±0.07
3.41±1.15
2.52±0.45
4.79±1.28
19.2±0.28
8.12±0.49
4.02±0.34
6.74±0.57
4.17±0.51
11.4±0.58
3.74±0.43
1.73±0.14
1.37±0.01
2.21±0.15
1.83±0.42
0.88±0.09
1.19±0.03
38.25 ±3.02
21.62±3.14
11.55±0.23
14.54±1.18
53.21±1.60
52.44±1.25
49.59±2.72
44.85±0.67
n.t.
n.t.
Me
H
H
n.t.
4-Cl
n.t.
Me 4-Cl
3-F
Me 3-F
n.t.
H
18.97±2.22
19.44±0.31
29.22±3.60
n.t.
H
4-N(Me)
2
Me 4-N(Me)
2
H
4-NO
n.t.
2
2
Me 4-NO
n.t.
n.t.
H
4-OEt
Me 4-OEt
2-OMe
Me 2-OMe
3,5-di-(OMe) 21.62±0.24
Me 3,5-di-(OMe) 21.18±0.15
51.53±1.08
65.30±1.43
54.88±2.26
56.76±1.91
35.32±2.07
28.27±5.82
58.43±1.44
52.52±1.96
49.31±2.74
31.08±3.17
18.25±1.01
22.36±1.46
23.39±4.64
8.94±0.84
16.50±1.16
12.48±0.95
n.t.
f
g
g
h
h
i
H
H
n.t.
H
4-OH
Me 4-OH
2-OH
Me 2-OH
19.60±0.29
20.05±0.095
20.17±0.16
20.29±0.19
19.16±0.13
20.07±0.15
20.60 [Ref 20]
10.72±0.5
15.38±0.30
21.76±1.58
n.t.
i
j
H
j
k
k
H
3,5-di-OH
n.t.
Me 3,5-di-OH
n.t.
Deferiprone
Trolox
--
-
--
-
3.89±0.09
0.76±0.02
n.t.
n.t.
Resveratrol
Curcumin
a
64.08±1.53
56.44±0.06
11.89±2.52
18.73±0.32
e
-
-
n.t.
A fluorescence method was used. pFe(III) data expressed as the mean±SD of three independent experiments.

ACCEPTED MANUSCRIPT
b
c
Inhibition of ABTS radical by target compounds. Data expressed as IC ±SD.
5
0
The thioflavin-T fluorescence method was used. The values are expressed as the mean±SD of three independent
measurements. All values were obtained at a compound concentration of 20 µM.
d
The IC (µM) values shown are the mean ±SD of three independent experiments.
5
0
e
Not tested
2.4．Inhibition of Aβ1-42 self-induced aggregation
The ability of compounds 3a-k and 4a-k to inhibit Aβ1-42 self-induced aggregation was tested
using a thioflavin-T (ThT) based fluorometric assay [11, 23]. Resveratrol and curcumin were used as
reference compounds (Table 1). The results demonstrated that about half the compounds under
investigation exhibited similar activity (51-65%) compared to that of the resveratrol (64%) and
curcumin (56%). Three compounds 4f, 4g and 3i (IC = 8.94-12.48 µM) demonstrated the greater
5
0
Aβ1-42 self-induced aggregation inhibitory activity.
The structure-activity relationship (SAR) revealed that the modification of the substituent on
phenyl group dramatically influences the inhibitory activity, while the introduction of methyl onto
the 1-position of the pyridine ring has much weaker effect, apart from 3d/4d and 3j/4j pairs.
Compounds with 4-chloro, 4-nitro and 3,5-dihydroxy on the benzene ring exhibited poor inhibitory
activities (0-22%) for both NH and NMe pyridones, whereas compounds substituted with 3-fluoro,
4-ethoxy, 2-methoxy and 4-hydroxy exhibited good inhibitory activities (> 50%). Among them, 3i
(
IC =10.70μM) and 4f (IC =8.94μM) were found to possess the highest inhibition activities. The
5
0
50
significantly lower inhibitory activity of 3, 5-dihydroxy and 3, 5-dimethoxy derivatives compared
with the 4-ethoxy analogue is different from a previous report, where all 3,5-dihydroxy, 4-methoxy
and 3,5-dimethyoxy substituted resveratrols were found to provide good inhibitory activities [11].
This phenomenon may be due to a different Aβ1-42 conformation induced by the compounds of the
present study. To investigate this possibility, molecular docking studies were undertaken.
2.5. Molecular docking studies
To further study the interaction mode and analyze the SAR profile of these derivatives for Aβ1-42, a
molecular docking study was performed using the C-DOCKER program in Discovery Studio 2.2.5.
The 3D-sturcture of ligands was generated and optimized under a Minimization field using the

ACCEPTED MANUSCRIPT
Powell method [24]. A binding mode study of compounds 3i and 4f with Aβ1-42 (Figure 2) showed
that both 3i and 4f are located nearby the C-terminus of Aβ1-42. For 3i, the oxygen of pyridinone
moiety formed a hydrogen bond with Lys28 at a distance of 2.50 Å, and the hydroxy group of the
benzene ring interacted with Ile41 through another hydrogen bond (2.45 Å). In comparison, for 4f,
the carbonyl oxygen and the 3-hydroxy group of pyridinone moiety formed hydrogen bonds with
Lys28 at distances of 1.93 Å and 2.49 Å respectively. It is reported that Aβ1-42 units are prone to form
β-pleated “hairpin” structures which are gathered into a fiber, which is stabilized by hydrophobic
interactions and a salt bridge resulting from electrostatic interactions between the amino acid
residues Asp 23 and Lys 28. The docking results indicated that compounds 3i and 4f can form
hydrogen bonds with Lys28 to block the salt bridge formation between Lys 28 and Asp23, which
could effectively inhibit the aggregation of amyloid peptide monomers. However, it needs to mention
that the docking is only a predictive tool [25]. The actual binding orientations and interactions of
Aβ1-42 with 3i and 4f have not been verified.

ACCEPTED MANUSCRIPT
Figure 2. Docking studies of 3i and 4f with Aβ1−42 (PDB code 1IYT). (A, B) Binding mode of 3i (green sticks) with
Aβ1−42 monomer. The possible hydrogen bonds between 3i and residues Lys28, Ile41 are indicated by yellow lines.
(C, D) Binding mode of 4f (yellow sticks) with Aβ1−42 monomer. The possible hydrogen bonds between 4f and
residues Lys28 are indicated by yellow lines.
2.6. The choose of drug candidates
Based on the above biological evaluation results, compounds 3i (IC50 (ABTS)=1.73 µΜ, IC50(Aβ1-42)
=10.7 µΜ) and 4f (IC50 (ABTS) =4.02 µΜ, IC50 (Aβ1-42)=8.94 µΜ)) exhibited significant abilities of
antioxidation and self-induced Aβ1-42 aggregation inhibition as well as the strong ability to bind
iron(III). Therefore, these two compounds have been chosen as drug candidates for further
evaluation.
2.7. Stoichiometry property
To determine the stoichiometry of the deferiprone-resveratrol hybrid complexes of Fe(III), the
molar ratio method was used [26]. Taking 3i as an example, the chelating effect of 3i on FeNTA (Fe :
nitrilotriacetic acid at molar ratio 1:2) in 3-morpholinopropanesulfonic acid buffer (MOPS, pH 7.4)
was tested by UV-vis spectrometry with wavelength ranging from 250 to 550 nm. UV-vis spectra of
compound 3i (60 µM) at increasing Fe(III) concentrations (0-20 µM) are presented in Figure 3(A).
The addition of FeNTA does not lead to a red or blue shift of wavelength but does change the
maximum absorbance intensity. A decrease of absorbance at 350nm and an increase of absorbance at
4
3
20nm were observed. Plots of absorbance versus mole ratios of FeNTA-3i are presented in Figure
(B). In theory, the stoichiometry of Fe(III) to the bidentate ligand is predicted to be at ratio of 1:3
under ambient conditions. In practice, UV spectrometry of FeNTA-3i complex demonstrated that the
maximal absorbance intensity is reached at an iron:ligand ratio of 1:4, revealing that the ligand
exceeds the sensitivity predicted by their chelating stoichiometry [27]. Similar curves were obtained
for other analogues.

ACCEPTED MANUSCRIPT
Figure 3. The Fe(III)- dependent UV/vis spectra of 3i and absorbance plot of 3i-Fe(III). (A): UV spectra of
compound 3i (60µM) with FeNTA in different molar ratio; (B): Determination of the stoichiometry of FeNTA and
3
i by molar ratio method. Plots of absorbance versus molar ratios of FeNTA-3i were presented at 348 nm.
2.8. ThT and TEM assay of inhibition of metal-induced Aβ1-42 aggregation
To investigate the ability of deferiprone-resveratrol hybrids to inhibit Fe(III)- and Cu(II)-induced
Aβ1-42 aggregation, Thioflavin T (ThT) fluorescence and Transmission Electron Microscope (TEM)
experiments were undertaken. For the ThT florescence assay, considering the chelating ratio of these
HPOs to Fe(III) and Cu(II) at 3:1 and 2:1 respectively, the ratios of Aβ-Fe(III)-chelator and
Aβ-Cu(II)-chelator were set up at 1:1:3 and 1:1:2 respectively. Aβ1-42 was treated with Cu(II) or
Fe(III) (for 24h) and then incubated with resveratrol, 3i or 4f for 24h at 37°C.
For the ThT florescence assay, the fluorescence of Aβ1-42 when treated with metal ions alone was
set at 100% (Figure 4). In the presence of 3i and 4f, the ThT fluorescence of Aβ1-42 with Fe(III)

ACCEPTED MANUSCRIPT
markedly decreased to 62% and 64% respectively. A similar phenomenon was observed for that of
the Aβ1-42 samples with copper (remaining 65% and 57% respectively). This data indicates the ability
of these compounds to inhibit Fe(III)- and Cu(II)-induced Aβ1-42 aggregation in comparison with
resveratrol (remaining 87% with iron and 86% with copper).
Similarly, the TEM assay of inhibition of Fe(III)- or Cu(II)-induced Aβ1-42 aggregation by the two
compounds also indicated a similar effect (Figures 5 and 6). In the presence of 3i and 4f, only a few
Aβ fibrils were observed. The above results suggest that compounds 3i and 4f possess good ability to
inhibit iron- and copper-induced Aβ aggregation.
Figure 4. ThT binding analysis of Fe(III)/Cu(II)-induced Aβ1-42 aggregation inhibited by resveratrol, 3i and 4f. The
fluorescence of Aβ1-42 treating with Fe(III) alone was set at 100%
7°C for 24h ([Aβ] = 25 µM, [compound] = 75 µM [FeNTA] = 25 µM) or [Aβ] = 25 µM, [compound] = 50 µM
Cu(II)] = 25 µM). Statistical significance was analyzed by ANOVO: ( ) P<0.01, versus control.
. The assay was measured in PBS (pH 6.6) at
3
[
**

ACCEPTED MANUSCRIPT
Figure 5. TEM images (the inhibition of Fe(III)-induced Aβ1-42 aggregation in the presence of resveratrol, 3i and 4f.
(
PBS, pH=6.6, [Aβ] = 25 µM, [compound] = 75 µM [Fe(III)] = 25 µM). (A) Aβ1-42, 0h; (B) Aβ1-42, 24h; (C)
Aβ1-42+FeNTA, 24h; (D) Aβ1-42+FeNTA +Resveratrol, 24h; (E) Aβ1-42+FeNTA +3i, 24h; (F) Aβ1-42+FeNTA +4f,
4h.
2
Figure 6. TEM images of Cu(II)-induced Aβ1-42 aggregation in the presence of resveratrol, 3i and 4f (PBS, pH=6.6,
Aβ] = 25 µM, [compound] = 50 µM [Cu(II)] = 25 µM). (A) Aβ1-42, 0h; (B) Aβ1-42+Cu(II), 24h; (C) Aβ1-42+Cu(II)
[
+
Resveratrol, 24h; (D) Aβ1-42+Cu(II) +3i, 24h; (E) Aβ1-42+Cu(II) +4f, 24h.
2.9．ThT and TEM assay of disaggregation of self- and Fe(III)- or Cu(II)-induced Aβ1-42 aggregation.
The ability of compounds 3i and 4f to disaggregate iron- or copper-induced Aβ aggregation was
undertaken with fibrils which were generated by incubating fresh Aβ with or without metal ions for
4h at 37 °C. Compounds 3i, 4f or resveratrol were then added to the sample and incubated for a
2

ACCEPTED MANUSCRIPT
further 24h at 37°C. As shown in Figure 7, the fluorescence of Aβ1-42 treated without metal ions was
set at 100%. The ThT fluorescence assay revealed that the two metal ions led to a reduced ThT
fluorescence in comparison with Aβ alone. This phenomenon is similar to the previous report [28]
and probably results from the emission being quenched by the paramagnetic metal ion-induced
formation of nonfibrilliar Aβ aggregates [29]. These results demonstrate that both compound 3i and
4f have an appreciable ability to resolve Aβ fibrils in comparison with resveratrol (remaining 58%
for iron and 35% for copper). The disaggregation effects on Fe(III)-induced fibrils (remaining 25%
for 3i and 16% for 4f respectively) were greater than those of Cu(II)-induced fibrils (remaining 69%
and 47 respectively).
Similar results were obtained for the TEM assay of the disaggregation of Fe(III)- and
Cu(II)-induced Aβ1-42 aggregation. In comparison with the well-defined Aβ fibrils (Figures 8-10, B),
fewer Aβ fibrils were observed when resveratrol, 3i or 4f was added (Figures 8-10; C-E), indicating
compound 3i and 4f could disassemble metal-induced Aβ fibrils as efficiently as resveratrol.
Figure 7. ThT binding analysis of Aβ disaggregation in the presence of Fe(III)/Cu(II) by resveratrol, 3i and 4f. The
fluorescence of Aβ1-42 treating without metal ions alone was set at 100%. The assay was measured in PBS (pH 7.4)
at 37°C for up to 48h ([Aβ] = 25 µM, [compound] = 75 µM, [Fe(III)] = 25µM or [Aβ] = 25 µM, [compound] = 50
µM, [Cu(II)] = 25 µM). Statistical significance was analyzed by ANOVO: ( ) P<0.01, versus control.
*
*

ACCEPTED MANUSCRIPT
Figure 8. TEM images of the disaggregation of self-induced Aβ1-42 aggregation (PBS, pH=7.4, [Aβ] = 25 µM,
[Compound] = 50 µM). (A) Aβ1-42, 0h; (B) Aβ1-42, 24h; (C) Aβ1-42+ Resveratrol, 24h; (D) Aβ1-42 +3i, 24h; (E)
Aβ1-42 +4f, 24h.
Figure 9.Schematic of the disaggregation experiments. TEM images of the disaggregation of Fe(III)-induced Aβ1-42
aggregation (PBS, pH=6.6, [Aβ] = 25 µM, [compound] = 75 µM, [Fe(III)] = 25 µM). (A) Aβ1-42, 0h; (B)
Aβ1-42+FeNTA; (C) Aβ1-42+FeNTA +Resveratrol; (D) Aβ1-42+FeNTA +3i; (E) Aβ1-42+FeNTA +4f.

ACCEPTED MANUSCRIPT
Figure 10. TEM images of the disaggregation of Cu(II)-induced Aβ1-42 aggregation (PBS, pH=6.6, [Aβ] = 25 µM,
[compound] = 50 µM, [Cu(II)] = 25 µM). (A) Aβ1-42, 0h; (B) Aβ1-42+Cu(II); (C) Aβ1-42++Cu(II) +Resveratrol; (D)
Aβ1-42++Cu(II) +3i; (E) Aβ1-42++Cu(II) +4f.
3. Conclusion
In conclusion, a series of deferiprone-resveratrol hybrids have been synthesized and evaluated as
multifunctional anti-Alzheimer agents. According to the ABTS assay, the majority of the synthesized
analogues displayed superior anti-oxidant properties than that of trolox. The ThT fluorometric assay
demonstrates that a proportion of the synthesized hybrids displayed higher effective inhibitory
potencies against Aβ1-42 aggregation than either resveratrol or curcumin. Structure–activity
relationship analysis for all synthesized analogues was undertaken and the analysis suggested that the
introduction of an electron-donating group such as hydroxyl or alkoxy at the 2 or 4 position of the
benzene ring is crucial to the inhibitory potencies of the compounds against Aβ1-42 aggregation.
Further investigations of compound 3i and 4f by Docking experiments confirmed that they can
efficiently interact with C-terminus of Aβ1-42 by hydrogen bonds. The TEM assay revealed that 3i and
4f can effectively inhibit self or metal (iron, copper) induced β-sheet aggregation and fibril formation,
which is consistent with the results of ThT binding assay. All the above data suggests that 3i and 4f
are both potentially effective agents for the treatment of AD and further studies are currently in
progress.
4
4
. Experimental section
.1. General procedures.

ACCEPTED MANUSCRIPT
All the chemical reagents used in the synthesis were of research grade and were obtained from
commercial sources. All the reactions were monitored by thin-layer chromatography (TLC). All
target products were characterized by using NMR and High resolution mass spectra (HRMS). NMR
spectra were acquired on a Bruker 600 spectrometer with TMS as the internal standard in DMSO or
CDCl . The chemical shifts are reported in parts per million (ppm), the coupling constants (J) are
3
expressed in hertz (HZ).
General procedure of chemical intermediates for deferiprone-resveratrol hybrids are presented in
supporting information.
4
4
1
.2. Chemical Synthesis
.2.1. 5-Hydroxy-2-styryl-1H-pyridin-4-one (3a)
0a (1.06g, 3.5mmol) was dissolved in hydrochloric acid (6N, 15ml). The mixture was refluxed
overnight and filtered and recrystallized from methanol to yield a white solid (0.5g, yield; 72%).
1
H NMR (600 MHz, DMSO) δ 8.10 (s, 1H, Py-H), 7.73 (d, J = 16.6 Hz, 1H, CH=CH), 7.64 (d, J =
7.4 Hz, 2H, Ar-H), 7.55 (s, 1H, Py-H), 7.47 (t, J = 7.5 Hz, 2H, Ar-H), 7.42 (t, J = 7.3 Hz, 1H, Ar-H),
1
3
7
.35 (d, J=16.6 Hz, 1H, CH=CH) (note: protons of OH and NH are not detected). C NMR (151
MHz, DMSO) δ 161.2, 144.8, 143.7, 136.8, 135.5, 130.2, 129.6, 127.9, 126.7, 119.43, 110.0. HRMS
+
(
ESI): Calcd for (M+H) (C H NO ) 214.0863, found 214.0872.
1
3
11
2
Analogous procedures starting with 10a-10h, 11a-11h with hydrochloric acid (6N) gave 3b-3h,
4a-4h.
4.2.2. 2-[2-(4-Chlorophenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3b)
1
(
Yield: 98%). H NMR (600 MHz, DMSO) δ 8.10 (s, 1H, Py-H), 7.73 (d, J = 16.5 Hz, 1H, CH=CH),
7.66 (d, J = 8.5 Hz, 2H, Ar-H), 7.53 (m, 3H, Ar-H, Py-H), 7.36 (d, J = 16.6 Hz, 1H, CH=CH) (note:
1
3
protons of OH and NH are not detected). C NMR (151 MHz, DMSO) δ 161.3, 144.9, 143.4, 135.4,
+
1
2
4
34.5, 134.4, 129.6, 129.5, 126.8, 120.3, 110.2. HRMS (ESI): Calcd for (M+H) (C H NO Cl)
1
3
10
2
48.0473, found 248.0487.
.2.3. 2-[2-(3-Fluorophenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3c)
1
(
Yield: 99%). H NMR (600 MHz, DMSO) δ 8.12 (s, 1H, Py-H), 7.74 (d, J = 16.5 Hz, 1H, CH=CH),
7.53 (s, 1H, Py-H), 7.51 (t, J = 7.3 Hz, 2H, Ar-H), 7.46 (d, J = 7.7 Hz, 1H, Ar-H), 7.42 (d, J = 16.6
1
3
Hz, 1H, CH=CH), 7.26 (t, J = 5.2 Hz, 1H, Ar-H) (note: protons of OH and NH are not detected). C
NMR (151 MHz, DMSO) δ 161.2, 145.0, 143.1, 137.9, 135.4, 131.6, 126.9, 124.4, 121.1, 117.2,

ACCEPTED MANUSCRIPT
+
1
17.0, 113.9, 110.4. HRMS (ESI): Calcd for (M+H) (C H NO F) 291.9968, found 291.9991.
1
3
10
2
4.2.4. 2-[2-(4-Dimethylaminophenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3d)
1
(
Yield: 96%). H NMR (600 MHz, DMSO) δ = 8.02 (s, 1H, Py-H), 7.62 (d, J = 16.2 Hz, 1H,
CH=CH), 7.53 (d, J = 7.2 Hz, 2H, Ar-H), 7.47 (s, 1H, Py-H), 7.08 (d, J = 16.2 Hz, 1H, CH=CH),
.97 (d, J = 7.2 Hz, 2H, Ar-H), 3.01 (s, 6H, N(CH ) ) (note: protons of OH and NH are not detected).
6
3
2
1
3
C NMR (126 MHz, DMSO) δ 161.1, 144.4, 144.2, 136.6, 136.5, 130.6, 129.4, 129.3, 126.2, 116.3,
+
1
09.3, 42.5. HRMS (ESI): Calcd for (M+H) (C H N O ) 257.1285, found 257.1309.
1
5
16
2
2
4.2.5. 5-Hydroxy-2-[2-(4-nitrophenyl)-vinyl]-1H-pyridin-4-one (3e)
1
(
Yield: 95%). H NMR (600 MHz, DMSO) δ 8.32 (d, J = 8.2 Hz, 2H, Ar-H), 8.13 (s, 1H, Py-H), 7.89
d, J = 8.2 Hz, 2H, Ar-H), 7.84 (d, J = 16.6 Hz, 1H, CH=CH), 7.54 (s, 1H, Py-H), 7.53 (d, J = 16.6
(
1
3
Hz, 1H, CH=CH) (note: protons of OH and NH are not detected). C NMR (151 MHz, DMSO) δ
161.3, 147.9, 145.3, 142.7, 142.1, 134.1, 128.8, 127.3, 124.8, 124.0, 110.9. HRMS (ESI): Calcd for
+
(
M+H) (C H N O ) 259.0713, found 259.0735.
1
3
10
2
4
4.2.6. 2-[2-(4-Ethoxyphenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3f)
1
(
Yield: 94%). H NMR (600 MHz, DMSO) δ 8.05 (s, 1H, Py-H), 7.66 (d, J = 16.5 Hz, 1H, CH=CH),
7
.58 (d, J = 8.8 Hz, 2H, Ar-H), 7.49 (s, 1H, Py-H), 7.17 (d, J = 16.5 Hz, 1H, CH=CH), 7.01 (d, J =
.8 Hz, 2H, Ar-H), 4.08 (q, J = 7.0 Hz, 2H, CH ), 1.34 (t, J = 7.0 Hz, 3H, CH ) (note: protons of OH
8
2
3
1
3
and NH are not detected). C NMR (151 MHz, DMSO) δ = 161.3, 160.3, 144.5, 144.2, 136.6, 129.6,
+
1
27.9, 126.3, 116.8, 115.4, 109.4, 63.8, 15.0. HRMS (ESI): Calcd for (M+H) (C H NO ) 258.1125,
1
5
15
3
found 258.1121.
4.2.7. 5-Hydroxy-2-[2-(2-methoxyphenyl)-vinyl]-1H-pyridin-4-one (3g)
1
(
Yield: 82%). H NMR (600 MHz, DMSO) δ 8.10 (s, 1H, Py-H), 7.76 (d, J = 16.6 Hz, 1H, CH=CH),
7.63 (d, J = 7.6 Hz, 1H, Ar-H), 7.58 (m, 1H, Ar-H), 7.40 (d, J = 16.6 Hz, 1H, CH=CH ), 7.40 (s, 1H,
Py-H), 7.12 (d, J=6.8 Hz, 1H, Ar-H), 7.04 (m, 1H, Ar-H), 3.90 (s, 3H, CH ) (note: protons of OH and
3
1
3
NH are not detected). C NMR (151 MHz, DMSO) δ 161.2, 158.0, 144.7, 132.0, 131.7, 128.7, 128.6,
+
1
2
4
26.5, 123.8, 121.3, 120.0, 112.3, 109.3, 56.2. HRMS (ESI): Calcd for (M+H) (C H NO )
1
4
13
3
44.0968, found 244.0982.
.2.8. 2-[2-(3,5-Dimethoxyphenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3h)
1
(
Yield: 39.6%). H NMR (600 MHz, DMSO) δ 8.10 (s, 1H Py-H), 7.64 (d, J = 16.5 Hz, 1H,
CH=CH), 7.51 (s, 1H, Py-H), 7.36 (d, J = 16.5 Hz, 1H, CH=CH), 6.79 (d, J = 2.2 Hz, 2H, Ar-H),

ACCEPTED MANUSCRIPT
6
.55 (t, J = 2.2 Hz, 1H, Ar-H), 3.80 (s, 6H, 2CH ) (note: protons of OH and NH are not detected).
3
1
3
C NMR (151 MHz, DMSO) δ 161.3, 161.2, 144.9, 143.5, 137.5, 136.6, 126.7, 120.2, 110.1, 105.8,
+
1
02.1, 55.8. HRMS (ESI): Calcd for (M+H) (C H NO ) 274.1074, found 274.1086.
1
5
15
4
4.2.9. 5-Hydroxy-1-methyl-2-styryl-1H-pyridin-4-one (4a)
1
(
Yield: 76%). H NMR (600 MHz, DMSO) δ 8.29 (s, 1H, Py-H), 7.78 (m, 2H, CH=CH, Ar-H), 7.68
(
s, 1H, Py-H), 7.44 (m, 5H, CH=CH, Ar-H), 4.11 (s, 3H, CH ) (note: proton of OH is not detected).
3
1
3
C NMR (151 MHz, DMSO) δ 160.1, 146.1, 144.8, 139.3, 135.6, 132.7, 130.3, 129.4, 128.5, 118.5,
+
1
10.4, 44.5. HRMS (ESI): Calcd for (M+H) (C H NO ) 228.1019, found 228.1037.
1
4
13
2
4.2.10. 2-[2-(4-Chlorophenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4b)
1
(
Yield: 66.7%). H NMR (600 MHz, DMSO) δ 8.26 (s, 1H, Py-H), 7.82 (d, J = 8.6 Hz, 2H, Ar-H),
7.63 (s, 1H, Py-H), 7.54 (d, J = 8.5 Hz, 2H, Ar-H), 7.46 (d, J = 1.0 Hz, 2H, CH=CH), 4.10 (s, 3H,
1
3
CH ) (note: proton of OH is not detected). C NMR (151 MHz, DMSO) δ 160.3, 145.9, 144.9, 137.8,
3
+
1
2
4
34.8, 134.6, 132.7, 130.1, 129.4, 119.4, 110.5, 44.5. HRMS (ESI): Calcd for (M+H) (C H NO Cl)
1
4
12
2
62.0629, found 262.0636.
.2.11. 2-[2-(3-Fluorophenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4c)
1
(
Yield: 64.6%). H NMR (600 MHz, DMSO) δ 8.31 (s, 1H, Py-H), 7.73 (d, J = 10.3 Hz, 1H,
CH=CH), 7.68 (s, 1H, Py-H), 7.59 (d, J = 8.1 Hz, 1H, Ar-H), 7.50 (m, 3H, 1×CH=CH, 2×Ar-H),
1
3
7
.26 (dd, J = 8.1, 2.4 Hz, 1H, Ar-H), 4.12 (s, 3H, Me) (note: proton of OH is not detected). C NMR
(151 MHz, DMSO) δ 160.1, 145.7, 145.0, 137.9, 137.7, 132.9, 131.4, 125.2, 120.1, 117.1, 116.9,
+
1
14.3, 110.6, 44.6. HRMS (ESI): Calcd for (M+H) (C H NO F) 306.0124, found 306.0147.
1
4
12
2
4.2.12. 2-[2-(4-Dimethylaminophenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4d)
1
(
Yield: 87%). H NMR (600 MHz, DMSO) δ 8.31 (s, 1H, Py-H), 7.02 (d, J = 6.2 Hz, 2H, Ar-H), 7.01
(s, 1H, Py-H), 6.96 (d, J = 12.1 Hz, 1H, CH=CH), 6.67 (d, J = 6.2 Hz, 2H, Ar-H), 6.30 (d, J = 12.1
1
3
Hz, 1H, CH=CH), 3.92 (s, 3H, NMe), 2.92 (s, 6H, N(CH ) ) (note: proton of OH is not detected). C
3
2
NMR (151 MHz, DMSO) δ 159.9, 146.8, 144.8, 144.2, 132.9, 132.2, 130.8, 130.1, 130.0, 113.5,
+
1
09.4, 44.4, 44.2. HRMS (ESI): Calcd for (M+H) (C H N O ) 271.1441, found 271.1456.
1
6
18
2
2
4.2.13. 5-Hydroxy-1-methyl-2-[2-(4-nitrophenyl)-vinyl]-1H-pyridin-4-one (4e)
1
(
Yield: 95%). H NMR (600 MHz, DMSO) δ 8.31 (d, J = 8.8 Hz, 2H Ar-H), 8.27 (s, 1H, Py-H), 8.05
(d, J = 8.8 Hz, 2H, Ar-H), 7.66 (d, J = 16.1 Hz, 1H, CH=CH), 7.64 (s, 1H, Py-H), 7.61 (d, J = 16.1
1
3
Hz, 1H, CH=CH), 4.13 (s, 3H, NMe) (note: proton of OH is not detected). C NMR (151 MHz,

ACCEPTED MANUSCRIPT
DMSO) δ 160.4, 148.1, 145.4, 145.2, 142.2, 136.6, 132.8, 129.4, 124.5, 122.9, 111.1, 44.5. HRMS
+
(
ESI): Calcd for (M+H) (C H N O ) 273.0870, found 273.0876.
1
4
12
2
4
4.2.14. 2-[2-(4-Ethoxyphenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4f)
1
(
Yield: 92%). H NMR (600 MHz, DMSO) δ 8.27 (s, 1H, Py-H), 7.73 (d, J = 8.8 Hz, 2H, Ar-H), 7.67
(s, 1H, Py-H), 7.44 (d, J = 16.0 Hz, 1H, CH=CH), 7.25 (d, J = 16.0 Hz, 1H, CH=CH), 7.00 (d, J =
8
.8 Hz, 2H, Ar-H), 4.09 (m, 5H, NCH , CH ), 1.35 (t, J = 7.0 Hz, 3H, CH ) (note: proton of OH is
3 2 3
1
3
not detected). C NMR (151 MHz, DMSO) δ = 160.5, 160.0, 146.6, 144.4, 139.1, 132.4, 130.2,
+
1
28.1, 115.7, 115.2, 109.8, 63.8, 44.4, 15.0. HRMS (ESI): Calcd for (M+H) (C H NO ) 272.1281,
1
6
17
3
found 272.1282.
4.2.15. 5-Hydroxy-2-[2-(2-methoxyphenyl)-vinyl]-1-methyl-1H-pyridin-4-one (4g)
1
(
Yield: 97%). H NMR (600 MHz, DMSO) δ 8.29 (s, 1H, Py-H), 7.72 (d, J = 6.9 Hz, 1H, Ar-H), 7.66
(d, J = 12.5 Hz, 1H, CH=CH), 7.51 (m, 1H, Ar-H), 7.42 (m, 1H, Ar-H), 7.25 (m, 2H, CH=CH, Py-H),
1
3
7
.04 (m, 1H, Ar-H), 4.09 (s, 3H, CH ), 3.87 (s, 3H, NCH ) (note: proton of OH is not detected). C
3 3
NMR (151 MHz, DMSO) δ 160.2, 157.9, 146.4, 144.7, 133.8, 132.6, 128.9, 126.4, 123.9, 121.2,
+
1
2
4
18.8, 112.2, 110.0, 56.2 , 44.4. HRMS (ESI): Calcd for (M+H) (C H NO ) 258.1125, found
1
5
15
3
58.1143.
.2.16. 2-[2-(3,5-Dimethoxyphenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4h)
1
(
Yield: 95.7%). H NMR (600 MHz, DMSO) δ 8.27 (s, 1H, Py-H), 7.63 (s, 1H, Py-H), 7.43 (d, J =
16.1 Hz, 1H, CH=CH), 7.36 (d, J = 16.1 Hz, 1H, CH=CH), 6.96 (d, J = 2.2 Hz,, 2H, Ar-H), 6.57 (t, J
=
2.2 Hz, 1H, Ar-H), 4.12 (s, 3H, NCH ), 3.81 (s, 6H, 2CH ) (note: proton of OH is not detected).
3
3
1
3
C NMR (126 MHz, DMSO) δ 161.2, 159.4, 145.9, 141.7, 136.8, 128.8, 128.7, 126.1, 121.4, 110.6,
+
1
03.5, 55.9, 44.4. HRMS (ESI): Calcd for (M+H) (C H NO ) 288.1230, found 288.1244.
1
6
17
4
4.2.17. 5-Hydroxy-2-[2-(4-hydroxyphenyl)-vinyl]-1H-pyridin-4-one (3i)
Boron tribromide (1.5g, 6eq) was added dropwise at -20℃ under nitrogen to a solution of 10f (0.35g,
mmol) in dichlomethane (20ml), then the resulting mixture was slowly warmed to room
1
temperature and stirred overnight. Methanol (6ml) was slowly added, and the mixture stirred for
another 0.5h. The resulting mixture was evaporated and recrystallized from methanol, filtered to get
yellow solid (0.29g, yield: 99%)
1
H NMR (600 MHz, DMSO) δ 7.99 (s, 1H, Py-H), 7.55 (d, J = 16.4 Hz, 1H, CH=CH), 7.50 (d, J =
8.7 Hz, 2H, Ar-H), 7.40 (s, 1H, Py-H), 7.03 (d, J = 16.4 Hz, 1H, CH=CH), 6.86 (d, J = 8.7 Hz, 2H,

ACCEPTED MANUSCRIPT
1
3
Ar-H) (note: protons of OH and NH are not detected). C NMR (151 MHz, DMSO) δ 161.3, 159.8,
+
144.8, 144.2, 137.3, 129.8, 126.6, 126.4, 116.4, 115.7, 109.2. HRMS (ESI): Calcd for (M+H)
(
C H NO ) 230.0812, found 230.0833.
1
3
11
3
Analogous procedures starting with 10g-h, 11f-h with boron tribromide gave 3j-k, 4i-k.
4.2.18. 5-Hydroxy-2-[2-(2-hydroxyphenyl)-vinyl]-1H-pyridin-4-one (3j)
1
(
Yield: 98%). H NMR (600 MHz, DMSO) δ 7.98 (s, 1H, Py-H), 7.73 (d, J = 16.6 Hz, 1H, CH=CH),
7.56 (m, 1H, Ar-H), 7.42 (s, 1H, Py-H), 7.33 (d, J = 16.6 Hz, 1H, CH=CH), 7.24 (m, 1H, Ar-H), 6.95
1
3
(
m, 1H, Ar-H), 6.89 (m, 1H, Ar-H) (note: protons of OH and NH are not detected). C NMR (151
MHz, DMSO) δ 161.3, 158.8, 144.8, 144.4, 133.0, 131.5, 128.9, 126.8, 122.1, 120.0, 118.9, 116.7,
+
1
09.3. HRMS (ESI): Calcd for (M+H) (C H NO ) 230.0812, found 230.0815.
1
3
11
3
4.2.19. 2-[2-(3,5-Dihydroxyphenyl)-vinyl]-5-hydroxy-1H-pyridin-4-one (3k)
1
(
Yield: 94%). H NMR (600 MHz, DMSO) δ 8.05 (s, 1H, Py-H), 7.51 (d, J = 16.4 Hz, 1H, CH=CH),
7
2
1
.50 (s, 1H, Ar-H), 7.12 (d, J = 16.4 Hz, 1H, CH=CH), 6.52 (d, J = 2.1 Hz, 2H, Ar-H), 6.32 (t, J =
1
3
.1 Hz, 1H, Py-H) (note: protons of OH and NH are not detected). C NMR (151 MHz, DMSO) δ =
61.3, 159.2, 144.5, 144.0, 137.6, 137.0, 126.7, 118.7, 109.8, 106.1, 104.8. HRMS (ESI): Calcd for
+
(
M+H) (C H NO ) 246.0761, found 246.07810.
1
3
11
4
4.2.20. 5-Hydroxy-2-[2-(4-hydroxyphenyl)-vinyl]-1-methyl-1H-pyridin-4-one (4i)
1
(
Yield: 98.7%). H NMR (600 MHz, DMSO) δ 8.14 (s, 1H, Py-H), 7.63 (d, J = 8.6 Hz, 2H, Ar-H),
7.47 (s, 1H, Py-H), 7.40 (d, J = 15.9 Hz, 1H, CH=CH), 7.18 (d, J = 15.9 Hz, 1H, CH=CH), 6.84 (d, J
1
3
=
8.6 Hz, 2H, Ar-H), 4.08 (s, 3H, CH ) (note: proton of OH is not detected). C NMR (151 MHz,
3
DMSO) δ 159.9, 147.1, 139.7, 132.3, 131.2, 130.4, 126.8, 116.2, 116.0, 114.6, 109.6, 44.4. HRMS
+
(
ESI): Calcd for (M+H) (C H NO ) 244.0968, found 244.0977
1
4
13
3
4.2.21. 5-Hydroxy-2-[2-(2-hydroxyphenyl)-vinyl]-1-methyl-1H-pyridin-4-one (4j)
1
(
Yield: 94.3%). H NMR (600 MHz, DMSO) δ 8.17 (s, 1H, Py-H), 7.73 (m, 1H, Ar-H), 7.58 (d, J =
16.1 Hz, 1H, CH=CH), 7.45 (s, 1H, Py-H), 7.40 (d, J = 16.1 Hz, 1H, CH=CH), 7.25 (m, 1H, Ar-H)
7
.14 (m, 1H, Ar-H), 6.90 (m, 1H, Ar-H), 4.08 (s, 3H, CH ) (note: protons of OH are not detected).
3
1
3
C NMR (151 MHz, DMSO) δ = 160.0, 156.7, 147.0, 144.4, 134.9, 132.8, 131.7, 128.8, 122.3,
+
1
2
4
19.9, 117.6, 116.6, 109.8, 44.4. HRMS (ESI): Calcd for (M+H) (C H NO ) 244.0968, found
1
4
13
3
44.0955.
.2.22. 2-[2-(3,5-Dihydroxyphenyl)-vinyl]-5-hydroxy-1-methyl-1H-pyridin-4-one (4k)

ACCEPTED MANUSCRIPT
1
(
Yield: 65%). H NMR (600 MHz, DMSO) δ 9.45 (s, 1H, OH), 8.17 (s, 1H, Py-H), 7.48 (s, 1H,
Py-H), 7.30 (d, J = 15.9 Hz, 1H, CH=CH), 7.21 (d, J = 15.9 Hz, 1H, CH=CH), 6.60 (d, J = 1.9 Hz,
2
H, Ar-H), 6.32 (d, J = 1.9 Hz, 1H, Ar-H), 4.09 (s, 3H, NCH ) (note: protons of OH are not detected).
3
1
3
C NMR (151 MHz, DMSO) δ 160.0, 159.1, 146.5, 144.5, 140.1, 137.3, 132.5, 117.9, 110.4, 106.6,
+
1
4
4
04.8, 44.4. HRMS (ESI): Calcd for (M+H) (C H NO ) 260.0917, found 260.0939.
1
4
13
4
.3 Biological evaluation
.3.1. Antioxidant activity assay
The
antioxidant
activity
was
determined
by
the
ABTS
(2,2’-azino-bis
(3-ethylbenzthiazoline-6-sulfonic acid)) radical scavenging method. ABTS was dissolved in purified
water to a 7 mM concentration, which was then reacted with 2.45 mM potassium persulfate (final
concentration) and kept it in the dark at room temperature for 16h to obtain ABTS radical cation
+
+
(
ABTS ). The ABTS solution was filtered by PFDV (0.2mm) and diluted with phosphate buffer
(0.2M, pH7.4) to 100 µΜ. 50 µL of 3a-k, 4a-k and trolox at different concentrations (80-0.004µΜ,
three times dilution) were added to 150 µL of 100 µΜ ABTS solution in 96-well plates and each
sample was prepared in triplicate, which was then incubated in dark at room temperature for 10
minutes. Then the optical absorbance of ABTS at 415 nm was measured.
4.3.2 Inhibition and disaggregation of self-induced Aβ1-42 aggregation
Inhibition and disaggregation of self-induced Aβ1-42 aggregation was measured using a
ThT-binding assay. Aβ1-42 was firstly dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) by brief
sonication for 10min, and kept overnight at room temperature. Then the HFIP was removed by
evaporation under vacuum. HFIP pretreated Aβ1-42 samples were re-dissolved in DMSO and a 10
mM phosphate buffer (pH 7.4) in order to have a stable stock solution ([Aβ1-42] = 50 µΜ). For
inhibition of self-induced Aβ1-42 aggregation, a mixture of the peptide (20 µL, 25 µΜ, final
concentration) with or without the inhibitor (20 µL, 20 µΜ, final concentration) was incubated in the
o
dark at 37 C for 24h. For disaggregation of self-induced Aβ1-42aggregation, the peptide was firstly
o
incubated in the dark at 37 C for 24h，then the inhibitor was added and the mixture was incubated
under the same conditions for another 24h. After incubation, 20 µL of the sample was extracted into
a black 96-well plate and diluted to a final volume of 200 µL with 50 µΜ gly-NaOH buffer (pH 8.5)
containing 5 µL ThT. Fluorescence was measured on Infinite 2000 multi-mode plate reader with
excitation and emission wavelengths at 448 nm and 485 nm, respectively. The percentage of

ACCEPTED MANUSCRIPT
inhibition or disaggregation of self-induced Aβ1-42 aggregation was calculated by using the following
equation: (1-IF /IF ) × 100% in which IF and IF were the fluorescence intensities gained for
i
c
i
c
absorbance in the presence and absence of inhibitors which had subtracted the absorbance of the
background of ThT.
4.3.3 Inhibition and disaggregation of metal-induced Aβ1-42 aggregation
The effects of inhibitors on metal-induced Aβ1-42aggregation were also determined by using ThT.
Most of the conditions were the same as in the self-induced Aβ1-42 aggregation test. The stock Aβ1-42
solution was diluted with 20 µΜ HEPES (pH 6.6) containing 150 µΜ NaCl. Stock solution of CuCl2,
FeNTA (10mM) was prepared using distilled water. For the inhibition assay, a mixture of the peptide
(
20 µL, 25 µΜ, final concentration), CuCl or FeNTA (20 µL, 25 µΜ, final concentration) and the
2
inhibitor (20µL, 50µΜ for iron sample; 75µΜ for copper sample, final concentration), were
o
incubated in the dark at 37 C for 24h. For disaggregation of self-induced Aβ aggregation, the
1
-42
o
peptide and metal ions was firstly incubated in the dark at 37 C for 24h，then the inhibitor was added
and the mixture was incubated under the same conditions for another 24h. After incubation, 20µL of
the sample was extracted into a black 96-well plate and diluted to a final volume of 200µL with
50µΜ gly-NaOH buffer (pH 8.5) containing 5 µM ThT. The absorbance was assayed and calculated
as described for the inhibition and disaggregation of self-induced Aβ1-42 aggregation assay.
4.3.4. TEM Assay
To further study the activity of compound 3i and 4f for the inhibition and disaggregation of
self-induced or metal-induced Aβ1-42 aggregation, transmission electron microscopy (TEM) was
used[30, 31]. Aliquots (5 µL) of the samples were prepared in the same manner as used in the ThT
assay and placed on a carbon-coated copper/rhodium grid for 2 minutes. Filter paper was used to
drain off the excess staining solution. Each grid was stained with uranyl acetate (1%, 5 µL) for 1
minute. Then the excess staining solution was removed by filter paper. The specimen was then
transferred for imaging by transmission electron microscopy (JEOL JEM-1230, Japan).
4.3.5 Metal chelating assay
For the molar ratio assay, chelating studies were made in DMSO and MOPS
(3-morpholinopropanesulfonic acid) buffer (pH 7.4) using UV-vis spectrophotometer (SHIMADZU
UV-2550). The absorption spectra of compound 3i, alone or in the presence of different
concentration of FeNTA was recorded at room temperature in a 1 cm quartz cell.

ACCEPTED MANUSCRIPT
For the determination of pFe(III) values, the metal chelation assay was performed in DMSO and
MOPS buffer using a spectrofluorophotometer (SHIMADZU, RF-5301PC) with emission mode of
excitation wavelength (435nm) and emission wavelength from 440nm to 550nm. CP645 and CP20
were synthesized as previously described [32-35]. MOPS buffer was treated with Chelex 100 before
use. Firstly, fluorescence of CP645 (6µΜ, 4ml, total volume) was quenched by the addition of
FeNTA (2µΜ, 4ml, total volume) and then recovered by addition of 3a-3k, 4a-4k (60µΜ, 4ml),
respectively. After 48h, the fluorescence of all the samples was detected. The maximum fluorescence
intensities of CP645 (6µΜ) and CP645-Fe complex (6µΜ: 2µΜ) were set at 100% and 0%
respectively, based on this, the relative fluorescence intensity of the compounds mixed with
CP645-Fe could be calculated by using the equation: fluorescence (%) =11.743 pFe(III)-169 [20].
Quoted values are the mean of 3 independent determinations.
4.3.6. Docking study
The docking studies of compounds 3i and 4f with Aβ1-42 monomer model (PDB ID: 1IYT)[36]
were performed by using the C-DOCKER program in Discovery Studio 2.2.5. The 3D-structures of
ligands were generated and optimized under a minimization field using the Powell method. The
receptor was prepared by checking the atom types, adding hydrogens, and assigning Kollman
charges. The binding pocket was defined to cover all the amino acid residues within a 30 Å radius
sphere. Docking experiments were carried out taking into account the flexibility of both ligand and
receptor. The structural optimization was performed for 100,000 generations using a TRIPOS
forcefield and genetic algorithm. The selected best scoring complexes were kept for further analyses
based on the reference of CDOCKER ENERGY and the conformation with the lowest energy was
selected. Docked models of the ligand-receptor complex were visualized using PyMOL.
Acknowledgments
We thank Zhejiang provincial Natural Science Foundation of China (LY17H300005) and Zhejiang
Chinese Medical University (2015ZG15) for final support of this study. We also thank Robert Hider
from King’s College London for helpful discussion of the manuscript.
Reference
[
1] A. Alzheimer's, 2014 Alzheimer's disease facts and figures, Alzheimer's & dementia, 10 (2014)
e47-92.

ACCEPTED MANUSCRIPT
[
2] R. Jakob-Roetne, H. Jacobsen, Alzheimer's disease: from pathology to therapeutic approaches,
Angew. Chem, 48 (2009) 3030-3059.
3] C. Reitz, Alzheimer's disease and the amyloid cascade hypothesis: a critical review, Int. J.
Alzheimer's Dis, 2012 (2012) 369808.
4] A.S. Darvesh, R.T. Carroll, A. Bishayee, W.J. Geldenhuys, C.J. Van der Schyf, Oxidative stress
and Alzheimer's disease: dietary polyphenols as potential therapeutic agents, Expert Rev. Neurother,
0 (2010) 729-745.
5] R.C. Hider, Y. Ma, F. Molina-Holgado, A. Gaeta, S. Roy, Iron chelation as a potential therapy for
neurodegenerative disease, Biochem. Soc. Trans, 36 (2008) 1304-1308.
6] J.L. Tomic, A. Pensalfini, E. Head, C.G. Glabe, Soluble fibrillar oligomer levels are elevated in
Alzheimer's disease brain and correlate with cognitive dysfunction, Neurobiol. Dis, 35 (2009)
52-358.
7] G. Pepeu, M.G. Giovannini, Cholinesterase inhibitors and beyond, Curr. Alzheimer Res, 6 (2009)
6-96.
8] R.C. Hider, S. Roy, Y.M. Ma, X. Le Kong, J. Preston, The potential application of iron chelators
for the treatment of neurodegenerative diseases, Metallomics, 3 (2011) 239-249.
9] P. Wang, Z.Y. Wang, Metal ions influx is a double edged sword for the pathogenesis of
Alzheimer’s disease. Ageing Res. Rev, (2016) doi:10.1016/j.arr.2016.10.003.
10] A. Gaeta, R.C. Hider, The crucial role of metal ions in neurodegeneration: the basis for a
promising therapeutic strategy, Br. J. Pharmacol, 146 (2005) 1041-1059.
11] C. Lu, Y. Guo, J. Yan, Z. Luo, H.B. Luo, M. Yan, L. Huang, X. Li, Design, synthesis, and
[
[
1
[
[
3
[
8
[
[
[
[
evaluation of multitarget-directed resveratrol derivatives for the treatment of Alzheimer's disease, J.
Med. Chem, 56 (2013) 5843-5859.
[
12] K. Ono, M.M. Condron, L. Ho, J. Wang, W. Zhao, G.M. Pasinetti, D.B. Teplow, Effects of grape
seed-derived polyphenols on amyloid beta-protein self-assembly and cytotoxicity, Biol. Chem, 283
(2008) 32176-32187.
[
13] L. Ho, L.H. Chen, J. Wang, W. Zhao, S.T. Talcott, K. Ono, D. Teplow, N. Humala, A. Cheng,
S.S. Percival, M. Ferruzzi, E. Janle, D.L. Dickstein, G.M. Pasinetti, Heterogeneity in red wine
polyphenolic contents differentially influences Alzheimer's disease-type neuropathology and
cognitive deterioration, J. Alzheimer's Dis : JAD, 16 (2009) 59-72.
[
14] K. Ono, H. Naiki, M. Yamada, The development of preventives and therapeutics for Alzheimer's
disease that inhibit the formation of beta-amyloid fibrils (fAbeta), as well as destabilize preformed
fAbeta, Curr. Pharm. Des, 12 (2006) 4357-4375.
[
15] J. Wang, L. Ho, W. Zhao, K. Ono, C. Rosensweig, L. Chen, N. Humala, D.B. Teplow, G.M.
Pasinetti, Grape-derived polyphenolics prevent Abeta oligomerization and attenuate cognitive
deterioration in a mouse model of Alzheimer's disease, J. Neurosci, 28 (2008) 6388-6392.
[
16] T. Zhou, Y. Ma, X. Kong, R.C. Hider, Design of iron chelators with therapeutic application,
Dalton Trans, 41 (2012) 6371-6389.
17] A. Maggio, G. D'Amico, A. Morabito, M. Capra, C. Ciaccio, P. Cianciulli, F. Di Gregorio, G.
[
Garozzo, R. Malizia, C. Magnano, A. Mangiagli, G. Quarta, M. Rizzo, D.G. D'Ascola, A. Rizzo, M.
Midiri, Deferiprone versus deferoxamine in patients with thalassemia major: a randomized clinical
trial, Blood Cells, Mol., Dis, 28 (2002) 196-208.
[
18] A. Gaeta, F. Molina-Holgado, X.L. Kong, S. Salvage, S. Fakih, P.T. Francis, R.J. Williams, R.C.
Hider, Synthesis, physical-chemical characterisation and biological evaluation of novel

ACCEPTED MANUSCRIPT
2-amido-3-hydroxypyridin-4(1H)-ones: Iron chelators with the potential for treating Alzheimer's
disease, Bioorg. Med. Chem, 19 (2011) 1285-1297.
[
(
[
19] Z.D. Liu, R.C. Hider, Design of clinically useful iron(III)-selective chelators, Med. Res. Rev, 22
2002) 26-64.
20] Y. Ma, Y. Xie, R.C. Hider, A novel fluorescence method for determination of pFe3+, The
Analyst, 138 (2013) 96-99.
21] R. Sheng, L. Tang, L. Jiang, L.J. Hong, Y. Shi, N.M. Zhou, Y.Z. Hu, Novel
-Phenyl-3-hydroxy-4-pyridinone Derivatives as Multifunctional Agents for the Therapy of
Alzheimer's Disease, Acs Chem Neurosci, 7 (2016) 69-81.
22] H. Schugar, D.E. Green, M.L. Bowen, L.E. Scott, T. Storr, K. Bohmerle, F. Thomas, D.D. Allen,
[
1
[
P.R. Lockman, M. Merkel, K.H. Thompson, C. Orvig, Combating Alzheimer's disease with
multifunctional molecules designed for metal passivation, Angew. Chem, 46 (2007) 1716-1718.
[
23] A.K. Sharma, S.T. Pavlova, J. Kim, J. Kim, L.M. Mirica, The effect of Cu(2+) and Zn(2+) on
the Abeta42 peptide aggregation and cellular toxicity, Metallomics, 5 (2013) 1529-1536.
24] R. Tycko, Progress towards a molecular-level structural understanding of amyloid fibrils, Curr.
Opin. Struct. Biol, 14 (2004) 96-103.
25] F. Prado-Prado, I. Garcia, Review of theoretical studies for prediction of neurodegenerative
inhibitors, Mini-Rev. Med. Chem, 12 (2012) 452-466.
26] M.L. Bolognesi, A. Cavalli, L. Valgimigli, M. Bartolini, M. Rosini, V. Andrisano, M. Recanatini,
[
[
[
C. Melchiorre, Multi-target-directed drug design strategy: from a dual binding site
acetylcholinesterase inhibitor to a trifunctional compound against Alzheimer's disease, J. Med. Chem,
5
[
0 (2007) 6446-6449.
27] Y.M. Ma, R.C. Hider, The selective quantification of iron by hexadentate fluorescent probes,
Bioorg. Med. Chem, 17 (2009) 8093-8101.
28] A.K. Sharma, J. Kim, J.T. Prior, N.J. Hawco, N.P. Rath, J. Kim, L.M. Mirica, Small Bifunctional
[
Chelators That Do Not Disaggregate Amyloid beta Fibrils Exhibit Reduced Cellular Toxicity, Inorg
Chem, 53 (2014) 11367-11376.
[
29] L. Jiang, M.K. Zhang, L. Tang, Q.J. Weng, Y.H. Shen, Y.Z. Hu, R. Sheng, Identification of
-subsituted benzothiazole derivatives as triple-functional agents with potential for AD therapy, Rsc
Adv, 6 (2016) 17318-17327.
30] R. Sheng, L. Tang, L. Jiang, L. Hong, Y. Shi, N. Zhou, Y. Hu, Novel
-Phenyl-3-hydroxy-4-pyridinone Derivatives as Multifunctional Agents for the Therapy of
Alzheimer's Disease, ACS Chem. Neurosci, 7 (2016) 69-81.
31] L. Huang, C. Lu, Y. Sun, F. Mao, Z. Luo, T. Su, H. Jiang, W. Shan, X. Li, Multitarget-directed
2
[
1
[
benzylideneindanone derivatives: anti-beta-amyloid (Abeta) aggregation, antioxidant, metal
chelation, and monoamine oxidase B (MAO-B) inhibition properties against Alzheimer's disease, J.
Med. Chem, 55 (2012) 8483-8492.
[
32] Y. Ma, S. Roy, X. Kong, Y. Chen, D. Liu, R.C. Hider, Design and synthesis of fluorinated iron
chelators for metabolic study and brain uptake, J. Med. Chem, 55 (2012) 2185-2195.
33] Y. Ma, W. Luo, P.J. Quinn, Z. Liu, R.C. Hider, Design, synthesis, physicochemical properties,
and evaluation of novel iron chelators with fluorescent sensors, J. Med. Chem, 47 (2004) 6349-6362.
34] S. Piyamongkol, Y.M. Ma, X.L. Kong, Z.D. Liu, M.D. Aytemir, D. van der Helm, R.C. Hider,
Amido-3-hydroxypyridin-4-ones as iron(III) ligands, Chemistry, 16 (2010) 6374-6381.
35] Z.D. Liu, R. Kayyali, R.C. Hider, J.B. Porter, A.E. Theobald, Design, synthesis, and evaluation
[
[
[

ACCEPTED MANUSCRIPT
of novel 2-substituted 3-hydroxypyridin-4-ones: structure-activity investigation of metalloenzyme
inhibition by iron chelators, J. Med. Chem, 45 (2002) 631-639.
[
36] O. Crescenzi, S. Tomaselli, R. Guerrini, S. Salvadori, A.M. D'Ursi, P.A. Temussi, D. Picone,
Solution structure of the Alzheimer amyloid beta-peptide (1-42) in an apolar
microenvironment-Similarity with a virus fusion domain, Eur. J. Biochem, 269 (2002) 5642-5648.

ACCEPTED MANUSCRIPT
Highlights
1
2
3
. 22 3-Hydroxypyridin-4-one-resveratrol hybrids were designed.
. 3i and 4f are good antioxidants, metal-chelating agents and Aβ anti aggregation.
. SAR between compounds and the Aβ1-42 inhibition activity was outlined.