Structure–activity relations of rosmarinic acid derivatives for the amyloid β aggregation inhibition and antioxidant properties

Article abstract of DOI:10.1016/j.ejmech.2017.07.026

Amyloid-β aggregation inhibitors are expected to be therapeutic or prophylactic agents for Alzheimer's disease. Rosmarinic acid, which is one of the main aggregation inhibitors derived from Lamiaceae, was employed as a lead compound and its 25 derivatives were synthesized. In this study, the structure–activity relations of rosmarinic acid derivatives for the amyloid-β aggregation inhibitory effect (MSHTS assay), antioxidant properties, and xanthine oxidase inhibition were evaluated. Among the tested compounds, compounds 16d and 19 were found to the most potent amyloid aggregation inhibitors. The SAR revealed that the necessity of the presence of the phenolic hydroxyl on one side of the molecule as well as the lipophilicity of the entire molecule. The importance of these structural properties was also supported by docking simulations.

Full text of DOI:10.1016/j.ejmech.2017.07.026

Accepted Manuscript
Structure–activity relations of rosmarinic acid derivatives for the amyloid β
aggregation inhibition and antioxidant properties
Riho Taguchi, Koki Hatayama, Tomohito Takahashi, Takafumi Hayashi, Yuki Sato,
Daisuke Sato, Kiminori Ohta, Hiroto Nakano, Chigusa Seki, Yasuyuki Endo, Kiyotaka
Tokuraku, Koji Uwai
PII:
S0223-5234(17)30548-2
10.1016/j.ejmech.2017.07.026
EJMECH 9589
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 3 March 2017
Revised Date: 20 June 2017
Accepted Date: 16 July 2017
Please cite this article as: R. Taguchi, K. Hatayama, T. Takahashi, T. Hayashi, Y. Sato, D. Sato, K.
Ohta, H. Nakano, C. Seki, Y. Endo, K. Tokuraku, K. Uwai, Structure–activity relations of rosmarinic acid
derivatives for the amyloid β aggregation inhibition and antioxidant properties, European Journal of
Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.07.026.
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
Structure–activity relations of rosmarinic acid derivatives for the amyloid β aggregation inhibition
and antioxidant properties.
Riho Taguchia, Koki Hatayamaa, Tomohito Takahashia, Takafumi Hayashib, Yuki Satob, Daisuke
Satob, Kiminori Ohtab, Hiroto Nakanoa, Chigusa Sekia, Yasuyuki Endob, Kiyotaka Tokurakua, and
Koji Uwaia*
a) Division of Sustainable and Environmental Engineering, Graduate School of Engineering,
Muroran Institute of Technology, 27-1 Mizumoto, Muroran 050-8585, Japan
b) Tohoku Medical and Pharmaceutical University, 4-4-1 Komatsushima, Aoba-ku, Sendai
981-8558, Japan
(Corresponding author)
Dr. Koji Uwai
Division of Sustainable and Environmental Engineering, Graduate School of Engineering, Muroran
Institute of Technology, 27-1 Mizumoto, Muroran 050-8585, Japan
e-mail: uwai@mmm.muroran-it.ac.jp
Graphical Abstract
Highlights
• 25 rosmarinic acid derivatives with anti-AD properties have been synthesized
• In vitro anti-Aβ aggregation, antioxidant properties, XOD inhibition were evaluated
• Structure–activity relations were discussed
• Compounds 16d and 19 are the most potent amyloid aggregation inhibitors

ACCEPTED MANUSCRIPT
Abstract
Amyloid-β aggregation inhibitors are expected to be therapeutic or prophylactic agents for
Alzheimer’s disease. Rosmarinic acid, which is one of the main aggregation inhibitors derived from
Lamiaceae, was employed as a lead compound and its 25 derivatives were synthesized. In this study,
the structure–activity relations of rosmarinic acid derivatives for the amyloid-β aggregation
inhibitory effect (MSHTS assay), antioxidant properties, and xanthine oxidase inhibition were
evaluated. Among the tested compounds, compounds 16d and 19 were found to the most potent
amyloid aggregation inhibitors. The SAR revealed that the necessity of the presence of the phenolic
hydroxyl on one side of the molecule as well as the lipophilicity of the entire molecule. The
importance of these structural properties was also supported by docking simulations.
Key words: Alzheimer’s disease, structure–activity relationship, inhibitor, rosmarinic acid,
amyloid-beta, aggregation
1. Introduction
Alzheimer’s disease (AD) is one of the most common neurodegenerative diseases in late adult
life and accounts for 70 percent of dementia. The onset of AD leads to the progressive loss of
learning ability and mental, behavioral and functional decline [1]. The number of the patients with
AD is estimated to reach 150 million by 2050 due to the acceleration of demographic aging [2].
Choline esterase inhibitors (donepezil, galanthamine, rivastigmine) and an N-methyl-D-aspartic
acid receptor antagonist (memantine) are now used for the medical treatment of AD. However,
these treatments are used only for delaying the progression of AD [3]. Thus, the development of a
fundamental AD treatment is becoming increasingly urgent.
In the AD patient’s brain, the amyloid β (Aβ) protein is cut out of the amyloid precursor protein
by the action of β-secretase and γ–secretase, is oligomerized, and becomes fibrotic by aggregation,
finally forming senile plaques. This insoluble Aβ fiber is thought to induce neuronal death and
cognitive dysfunction as a result of the deposition of senile plaques formed from the Aβ fiber in the
vascular wall in the brain (the Aβ cascade hypothesis) [4]. Thus, Aβ aggregation inhibitors are
expected to be therapeutic or prophylactic agents for AD.
Previously, a series of bisphenols [5], alkaloids [6], polyphenols [7], and peptides with
C-terminal motif of Aβ [8] have been reported as the target compound Aβ aggregation inhibitors.
On the other hand, based on the demonstration of the correlation between the Aβ aggregation
inhibitory activity and antioxidant activity [9], it appears that the study of the Aβ aggregation
inhibitors based on the polyphenols can make a significant contribution to the development of AD
treatment. For example, the relation between the structure of curcumin, an active polyphenol, and
the Aβ aggregation inhibitory activity has been well-studied [10,11].

ACCEPTED MANUSCRIPT
Recently, we established a novel method for the evaluation of the Aβ aggregation inhibitory
activity called the “microliter scale high - throughput screening (MSHTS) system” [12], and
showed that the results obtained by this method were comparable to those obtained using
Thioflavin-T (ThT) assay, which is the standard method for the evaluation of the Aβ aggregation
inhibitory activity. Using this new method, we found that of 52 tested spices, a member of
Lamiaceae showed the strongest activity, and rosmarinic acid (1) isolated from Satureja hortensis
(one of the species of Lamiaceae family) was a main Aβ aggregation inhibitor. Additionally, it was
reported that rosmarinic acid led to an improvement in the cognitive function in Tg2576 mice [13].
Rosmarinic acid (1) is also reported to show a high antioxidant activity [14]. Aβ-induced
neuronal toxicity is regulated by the reactive oxygen species (ROS) generation that assists
subsequent Aβ aggregation and speeds up the AD progression [15, 16] and antioxidants with
catechol moiety showed to inhibit Aβ aggregation [17]. Thus, the relations between antioxidant
activity and Aβ aggregation in the viewpoint of the compound structure is also of interest.
Herein, we studied the structure–activity relations between the structures of rosmarinic acid
derivatives and the Aβ aggregation inhibitory activity and also performed simulations of the
docking of these derivatives with the Aβ monomer. Moreover, their potency as scavengers of the
superoxide radical, produced by xanthine oxidase (XOD) (enzymatic method) and by the diphenyl
picrylhydrazyl (DPPH) assay (chemical method) were also studied.
2. Results and Discussion
2.1 Chemistry
Rosmarinic acid (1) has five key structural groups: a carboxy group, two phenol hydroxy groups, an
unsaturated C–C bond, an alkoxy group, and an ester moiety (Fig. 1).

ACCEPTED MANUSCRIPT
(Single column fitting image)
Fig. 1. Key points in the structure of rosmarinic acid (1)
Initially, we directed our attention to the carboxy group and prepared two analogs of rosmarinic
acid, rosmarinic amide (2)[18,19] and caffeic acid 3,4-dihydroxyphenethyl ester (12c)[20,21,22]
(Scheme 1). Rosmarinic amide (2) was synthesized using Won’s protocol [18,19]. However, since
the yield obtained by this approach was quite poor, we attempted to synthesize it via an acid
chloride prepared with oxalyl chloride and NH3 [23]. We attempted to prepare the caffeic acid
3,4-dihydroxyphenethyl ester (12c) by the direct Mitsunobu reaction of 3,4-dihydroxyphenethyl
alcohol with caffeic acid (3); however, the target yield was only 4.8%. Examination of the structure
of 3,4-dihydroxyphenethyl alcohol shows that it possesses several alcohol groups that can function
as a nucleophile in basic conditions. Additionally, the electrophilicity of the carboxy moiety of
caffeic acid (3) was decreased because of the conjugation of the phenolic hydroxyls. Therefore, we
planned to protect the phenolic hydroxyls in caffeic acid (3) and 3,4-dihydroxyphenethyl alcohol.
The phenolic hydroxyls of 3,4-dihydroxyphenethyl alcohol were protected with
tert-butyldimethylsilyl (TBS) groups, and the caffeic acid (3) phenolic hydroxyls were protected
with the 2-nitrobenzensulfonyl (Ns) groups. NsO-groups exhibit an electron-withdrawing nature,
and therefore the electrophilicity of the conjugated carboxy moiety was expected to be increased. In
particular, caffeic acid (3) was esterified by Fischer’s ester synthesis and the phenolic hydroxyls
were protected by the Ns groups, with acid hydrolysis of the ester giving carboxylic acid (6a). The
coupling partner, the protected 3,4-dihydroxyphenethyl alcohol (10a), was synthesized from 3,
4-dihydroxyphenylacetic acid (7). Allyl esterification of carboxylic acid (7) followed by
TBS-protection of phenolic hydroxyls and reduction of the ester with lithium aluminum hydride
gave alcohol (10a). Condensation of the protected carboxylic acid (6a) and alcohol (10a) followed
by the deprotection of Ns and TBS groups gave caffeic acid 3,4-dihydroxyphenethyl ester (12c).

ACCEPTED MANUSCRIPT
(2-column-fitting image)
Scheme 1. a) oxalyl chloride, NH3, THF, rt, 19.5 h; b) EDCI, DMAP, MeOH; c) NsCl, Et3N,
CH3CN; d) AcOH, 0.5 M HCl; e) EDCI, DMAP, Allyl alcohol; f) TBSCl, imidazole, DMF; g)
LiAlH4, THF; h) DEAD, TPP, THF, 0 °C → rt; i) PhSH, Cs2CO3, CH3CN, 0 °C → rt; j) TBAF,
THF, 0 °C → rt; k) H2, Pd/C; l) DIAD, TPP, THF, 0 °C → rt

ACCEPTED MANUSCRIPT
Similarly, to investigate the effect of phenolic hydroxyls and unsaturated C–C bond, a series of
carboxylic acids (6a, b) and alcohols (10a-c) were coupled, followed by deprotection to give esters
(11a, 12a-d). Unsaturated C–C bond in compound 11a, 12a, and 12c were reduced by catalytic
hydrogenation to give compounds 13a-c.
Caffeic acid alkyl esters (16a-g, i, k) were also prepared by the condensation reaction of caffeic
acid (3) and corresponding alkyl alcohols (15a-i) to study the effect of the aromatic ring on one side
of the molecule (Scheme 2). Additionally, to confirm the effect of the ester group, ketone (19)[24]
and amide (21a, b) were synthesized.
(2-column-fitting image)
Scheme 2. a) DIAD, PPh3, THF, r.t.; b) PhSH, Cs2CO3, CH3CN, 0 °C → rt; c) TBAF, THF, 0 °C
→ rt; d) pyrolidine, AcOH, THF, reflux; e) iPr2NEt, DMAP, EDCI, DMF, rt
2.2 Aβ(1-42) aggregation inhibitory activity
Elucidation of the essential structural factor for the inhibition of the Aβ(1-42) aggregation
inhibitory activity would lead to the development of more effective inhibitors and would obtain
vital information for the elucidation of the inhibitory action mechanism. Therefore, we directed our
attention to the five structural features of rosmarinic acid (1) that show the strong activity: carboxy
group, phenolic hydroxyls, unsaturated C–C bond, alkoxy group, and ester moiety. We synthesized
various derivatives and evaluated their Aβ(1-42) aggregation inhibitory activities (Table 1).

ACCEPTED MANUSCRIPT
Table 1. Aβ(1−42) aggregation inhibitory activity by MSHTS system and CLogP values of
rosmarinic acid derivatives
Compounds
EC50 [ M]
CLogP
Compounds
16a
EC50 [ M]
CLogP
20.3
1.10
23.2
1.98
1
2
21.9
216
0.27
0.97
1.20
4.56
3.30
2.96
2.03
3.30
4.38
3.11
1.85
4.03
24.7
12.0
2.58
5.13
10.4
9.52
80.1
54.8
2.82
69.0
11.4
2.26
2.79
3.32
3.85
4.38
5.43
1.33
1.86
3.53
1.37
2.66
16b
16c
16d
16e
16f
3
4
67.7
1040
7.95
39.5
20.6
8.41
4536
5.93
14.9
274
11a
12a
12b
16g
16i
16k
19
12c
12d
13a
13b
21a
21b
13c
13d
First, we studied the effect of carboxy group on the activity. Compared to the Aβ(1-42) aggregation
inhibitory activity of rosmarinic acid (1, EC50 = 20.3 µM), amide (2) and compound 12c, which
eliminate the carboxylic acid, showed almost equivalent activities of 21.9 µM and 20.6 µM,
respectively. This suggested that the carboxy group did not affect the activity.
Next, the effect of phenolic hydroxyls was evaluated. Compared to the EC50 value of compound
12c (20.6 µM), the EC50 values of compound 12a, for which the phenolic hydroxyls on the
carboxylic side were eliminated, and compound 12d, for which the phenolic hydroxyls on the

ACCEPTED MANUSCRIPT
alkoxy moiety were eliminated, were 7.95 and 8.41 µM, respectively. Thus, the activity was
enhanced by the elimination of phenolic hydroxyls. However, interestingly, compound 11a, for
which the phenolic hydroxyls on both sides were deleted, exhibited the activity of 1,040 µM. These
results suggested that lateral phenolic hydroxyls on odd side were important for the activity.
Similarly, compound 12b, for which the phenolic hydroxyls were masked with methoxyl functional
groups, showed significantly lower activity (EC50 = 39.5 µM) than compound 12c. In line with the
results described above, in contrast to the activity of compound 13a (EC50 = 4,536 µM), for which
the phenolic hydroxyls on both sides were eliminated, compound 13d, for which hydroxyl groups
on the alkoxy side were masked with methoxy groups, showed a higher activity (EC50 = 274 µM).
Moreover, compound 13b, for which the hydroxyls were present on the alkoxy side, showed
significantly higher activity (EC50 = 5.93 µM) than compound 13d. These results suggested that not
only the existence of polar functional groups but also the functional groups that could make
hydrogen bonds or were hydrophilic affected the activity.
To evaluate the effect of the aromatic rings on the activity, we tested the compounds for which one
phenyl ring was substituted with an alkyl chain. Compound 16 (EC50 = 5.13 µM) for which the
alkyl chain was as long as that of compound 12d showed almost the same activity as the parent
compound. This result suggested that the existence of the aromatic ring was not important for the
activity; rather, the presence of a hydrophobic group in the alkoxy moiety was vital.
Next, the effect of the alcohol substitution for catechol was investigated. The activity of compound
16k (EC50 = 54.8 µM) with alcohol in alkoxy moiety was significantly impaired when compared to
compound 12c with catechol. This result suggested that catechol was important for the activity,
seemingly contradicting the interpretation of the results for 16e and 12d. Nevertheless, this
discrepancy may be resolved by considering the fact that it is easier for compound 12c with the
catechol moiety to make a hydrophilic bond via π–π stacking than for compound 16k with the
Aβ(1-42) peptide.
We then investigated the effect of the alkyl chain length on the alkoxy moiety. The activity of
caffeic acid alkyl esters was enhanced in accordance with the elongation of alkyl chain in the 1 to 5
range. The activity was maximum at the C5-ester (16d, EC50 = 2.58 µM) and then decreased
gradually. Thus, the length of the alkyl side chain (length of the liphophilic moiety) was important
for the activity.
The effect of the unsaturated C–C bond in the linker that connected the two catechols was
investigated. Comparison of compounds 11a, 12a, and c with the α,β-unsaturated ester and
compounds 13a, b, and c with the saturated ester showed that the activities of the corresponding
compounds were almost equivalent. This suggested that the unsaturated bond in the linker was not
important for the activity.

ACCEPTED MANUSCRIPT
The mode of connection linking the two catechols was then investigated. The activity of compounds
21a (69.0 µM) and 21b (11.4 µM) that connected the carboxyl and alkoxy group via the amide were
significantly impaired compared to the corresponding ester compounds 12c and 16d. On the other
hand, the activity of ketone 19 (2.82 µM) was almost equivalent to that of compound 16d. These
results suggested that the resonance effect of amide nitrogen may fix the conformation and affect
the activity.
2.3 Kinetics of Aβ(1-42) aggregation in the existence of inhibitors
The aggregation of Aβ(1-42) was followed by an increase in the standard deviation (SD) value of
fluorescence intensities owing to the variability of QDAβ(1-42) in coaggregates. The representative
aggregation curve on Fig. 2 demonstrates a good fit between the Boltzmann equation used for
fitting and experimental data that is confirmed by plotting the residuals versus time in the inset.
Parallel measurements at 37°C with 30 µM Aβ(1-42) showed good reproducibility of the maximal
SD of fluorescence 6.77 and the aggregation growth rate (k = 0.533 min-1) in the control (without
inhibitor). The lag time of the reaction was 4.44 min. By contrast, by the addition of 20.3 µM
(EC50) of rosmarinic acid (1) or 5.13 µM of C6-ester (16e), the maximal SD of fluorescence was
3.40 and 3.68 and the aggregation growth rate (k = 0.508 and 0.190 min−1), without any change in
the lag-period of the reaction, respectively. These results demonstrated that the existence of
rosmarinic acid (1) did not affect the aggregation rate but reduced the total amount of aggregates
and extended the lag time (5.58 min). On the contrary, the existence of C6-ester (16e) reduced the
aggregation rate, the total amount of aggregates and lag-period (1.10 min). Thus, it was suggested
that the mechanism of inhibition of rosmarinic acid (1) and C6-ester (16e) was different.

ACCEPTED MANUSCRIPT
8
6
4
2
0
0
5
10
15
20
Time [h]
(1 column fitting image)
Fig. 2 Time-lapse analyses of process of Aβ(1-42) aggregation with/without 20.3 µM (EC50) of
rosmarinic acid (1) or 5.13 µM of C6-ester (16e) obtained by MSHTS system. –•–: control, ---•---:
rosmarinic acid (1), ⋯ • ⋯: C6-ester (16e)
2.4 CLogP values of rosmarinic acid derivatives
For exhibiting central nervous system (CNS) activity, compounds need to pass through the blood–
brain barrier (BBB). One of the factors relevant to success of CNS drugs is lipophilicity. In the
common logarithm of octanol/water partition coefficient (cLogP) that showed the degree of
calculated lipophilicity from chemical structure, CNS drugs showed significantly higher value than
that of peripheral drugs. The average cLogP value of existing CNS drugs was 3.43 [25]. ClogP
values of the compounds were shown in Table 1. On comparing the EC50 values and cLogP values,
the active compounds (12a, 12d, 13b, 16d, 16e, 19) tended to show suitable cLogP values (3.30,
3.30, 3.11, 3.32, 3.85, 3.53, respectively). Thus, these compounds were expected to show the
Aβ(1-42) aggregation inhibitory activity also in vivo.
2.5 Docking simulation between rosmarinic acid derivatives and Aβ(1-42) monomer
Clarifying the binding site of rosmarinic acid derivatives on Aβ(1-42), we would be able to presume
important amino acid residues in Aβ(1-42) for aggregation and to design more effective inhibitors.
Thus, docking simulation between Aβ(1-42) monomer (PDB: 1Z0Q)[26] and 25 of rosmarinic
derivatives were performed using AutoDock [27,28].

ACCEPTED MANUSCRIPT
The results were shown in S-1. Concerning the lowest energy cluster in 50 calculated clusters for
each derivatives, the derivatives could be classified in three groups by the activity and binding site:
1) derivatives with EC50 ≦10 µM tended to bind to amide NH and carboxylate in Ala42 with their
phenolic hydroxyls via hydrogen bond (Fig. 3A); 2) derivatives with 10 µM < EC50 < 100 µM
tended to bind to amide carbonyl oxygen in Ala30 with their phenolic hydroxyl via hydrogen bond
(Fig. 3B); 3) derivatives with EC50 ≧100 µM tended to bind to side chain in His14 with their ester
functional group via hydrogen bond (Fig. 3C). Concerning this result that higher effective
derivatives tended to bind Ala42 with the report that Aβ(1-42) might aggregate faster than Aβ(1-40)
[29], it was suggested that Ala42 played an important role in their aggregation process.

ACCEPTED MANUSCRIPT
(2 column fitting image)

ACCEPTED MANUSCRIPT
Fig. 3. Docking model between Aβ(1-42) monomer and rosmarinic acid derivatives calculated by
AutoDock.
2.6 XOD inhibitory effect
Among the 25 tested rosmarinic acid derivatives (Table 2), 10 displayed XOD inhibitory potencies.
Compounds 2, 12c, 12d, 14, 16b, 16c, 16d, 16j, 16k, 19 showed better potencies. A detailed
evaluation of the chemical structures of the active rosmarinic acid derivatives showed a general
requirement for two or three hydrophilic moieties for better potency with compounds 2, 12c, 16i, k
being the potent inhibitors in the tested compounds. Additionally, the activity of series of caffeic
acid alkyl esters were depending on the length of alkyl side chain. Most potent derivatives were
compound 16c and 16d (butyl and pentyl ester).
Table 2. XOD inhibitory and DPPH radical-scavenging activity
XOD inhibition (%)
DPPH radical scavenging
activity (%)
compound
1
1 µM
10 µM
< 10
100 µM
27
100 µM
47
500 µM
66
48
63
64
38
< 10
< 10
< 10
< 10
50
82
< 10
65
39
49
2
3
48
8f
< 10
22
< 10
40
< 10
56
11a
12a
12b
12c
12d
13a
13b
13c
13d
16a
16b
16c
16d
16e
16f
43
43
59
26
80
37
46
< 10
< 10
< 10
< 10
< 10
< 10
< 10
<10
11
62
35
51
< 10
< 10
21
< 10
38
< 10
54
37
49
< 10
22
< 10
46
< 10
62
61
46
63
> 90
> 90
< 10
25
46
62
46
62
10
45
61
11
46
63

ACCEPTED MANUSCRIPT
< 10
< 10
< 10
12
40
84
51
76
30
33
45
41
42
48
47
40
61
54
55
62
64
53
16g
16i
16k
19
< 10
< 10
> 90
21a
21b
< 10
allopurinol
edaravone
50-53
71-75
2.7 Antioxidant ability (DPPH radical scavenging assay)
The antioxidant activities of rosmarinic acid derivatives were based on their interaction with the
2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical (Table 2) [30]. As expected, the results
suggested that compounds without catechol type aromatic ring did not exhibit significant
antioxidant activities, as it is shown by the low inhibition percentages at very high concentrations
(<10% at 500 µM concentration). Thus, the significant radical scavenging activity of the tested
compounds can be explained by the presence of catechol phenolic hydroxyls that are important in
the scavenging action of ROS species. However, we could not find a specific structural feature of
caffeic acid-type compounds is present in all the compounds, and caffeic acid itself has significant
antioxidant activity.
3. Conclusions
In summary, rosmarinic acid derivatives were synthesized and their Aβ(1-42) aggregation
inhibitory activity was investigated. After detailed consideration of structure–activity relations, it
was suggested that the structural requirement of rosmarinic acid derivatives to express the activity
was phenolic hydroxyls in unilateral side and the length of the alkoxy moiety.
Additionally, normally ester functional group was labile intravitally to hydrolases to metabolite to
carboxylic acid and alcohol. Thus, the fact that compound 19 showed the equivalent activity to
compound 16d gave us hope to develop in vivo experiment.
Docking simulation of Aβ(1-42) monomer and rosmarinic acid derivatives by AutoDock suggested
that potent Aβ aggregation inhibitor tend to dock the Ala42, Ile31, and Val39. This result suggested
the importance of C-terminal of Aβ(1-42) for the aggregation. XOD inhibitory effect was affected
by the compounds lipophilicity. DPPH antioxidation activity was found in the derivatives with
catechol type moiety.
The present study sheds new light on the properties of this interesting group of compounds. Further
studies are now underway in our laboratory, including a search for more effective analogs and

ACCEPTED MANUSCRIPT
studies exploring the mechanism of action of these compounds.
4. Experimental Part
4.1 MSHTS system
The EC50 values of compounds were determined by a modified MSHTS system [12]. More
specifically, various concentrations of rosmarinic acid derivatives, 30 nM QD - labeled Aβ(1-42)
(QDAβ(1-42)), and 30 µM Aβ(1-42) in PBS containing 5% EtOH and 3% DMSO were incubated
in a 1536 - well plate at 37˚C for 24 h, and the formed QDAβ(1-42)–Aβ(1-42) coaggregates in each
well were observed by an inverted fluorescence microscope equipped with a color CCD camera.
SD values of fluorescence intensities of 10,000 pixels (100 × 100 pixels: 167 × 167 mm) in the
central region of each well were measured by the ImageJ software. The SD values, which were
approximately proportional to the amount of aggregates, were plotted against the concentrations of
added rosmarinic acid derivatives to establish an inhibition curve. EC50 values were estimated from
the inhibition curve using the EC50 shift by the Prism global fitting software.
4.2 Data analysis and kinetic formation
The kinetics of Aβ(1-42) aggregation could be described as sigmoid curves, and aggregation
parameters were determined by fitting the plot of the SD values of fluorescence intensities versus
time in the Boltzmann equation.
ꢂꢃ
ꢇꢂꢃ
ꢄꢈꢉ
ꢄꢅꢆ
ꢀꢁ = ꢊꢇꢋꢌꢍ
,
ꢎꢏ
ꢐ
ꢓ ꢕ
ꢇꢐ ꢔ
ꢑꢒ
where SDmin is the initial SD level, SDmax is SD of maximal aggregation level, t50 is the time when
fluorescence reaches half maximal, and k is the rate constant of aggregate formation. The lag time is
calculated as
ꢖꢗꢘ = ꢙꢚꢛ − 2/ꢜ
4.2 Molecular Docking [27,28]
AutoDock, PytonMolecularViewer, and Cygwin were used for the docking simulation of
interactions between Aβ(1-42) and inhibitors. ChemBioDraw was used for the preparation of
chemical structure data of inhibitors for the docking simulation. ChemBio3D was used for the
calculation of molecular dynamics and the most stable conformation of inhibitors. The results of
docking simulation were analyzed by using Discovery Studio. PDB: 1Z0Q [26] was used for the
docking simulation with Aβ(1-42) aggregation inhibitors. The compounds used for the docking
simulation were 25 rosmarinic acid derivatives.
4.3 XOD inhibition test

ACCEPTED MANUSCRIPT
35 µL of phosphate buffer (100 mM, pH=7.6), 50 µL of a solution of allopurinol and rosmarinic
acid derivatives in DMSO, 30 µL of a solution of XOD (from butter milk, Oriental Yeast Co., Ltd.)
in phosphate buffer were added in 96 well micro plate. Then, 60 µL of a solution of 500 µM
xanthine in phosphate buffer was added to the solution at rt. After 8 min, 1N HClaq was added and
absorbance at 290 nm was measured. The inhibition rate (%) was calculated according to the
following equation:
ꢥꢠꢦꢏꢧꢨꢁ ꢟꢞℎꢟꢠꢟꢙꢟꢡꢞ ꢗꢩꢙꢟꢪꢟꢙꢫꢓ − ꢥꢠꢦꢏꢦꢗꢬꢭꢖꢣꢓ
ꢝꢞℎꢟꢠꢟꢙꢟꢡꢞ ꢢꢗꢙꢣ ꢏ%ꢓ = ꢤ1 −
ꢯ × 100
ꢏ
ꢓ
ꢥꢠꢦꢏꢧꢨꢁ ꢗꢩꢙꢟꢪꢟꢙꢫꢓ − ꢥꢠꢦꢏꢮꢖꢗꢞꢜꢓ
4.4 1,1-Diphenyl-2-picrylhydrazyl (DPPH) radical scavenging assay
40 µL of 1 mM DPPH solution and 150 µL of EtOH were added to 96 well plate. Then, 10 µL of 2
mM or 10 mM of edaravone, or rosmarinic acid derivatives in DMSO was added at rt. And then, the
absorbance at 517 nm was continuously measured for 60 min, and radical trapping capacity was
calculated with the data at 20 min following the equation:
ꢥꢠꢦꢏ20 ꢬꢟꢞꢓ − ꢥꢠꢦꢏꢦꢗꢬꢭꢖꢣꢓ
ꢁꢰꢰꢱ ꢢꢗꢲꢟꢩꢗꢖ ꢙꢢꢗꢭꢭꢟꢞꢘ ꢩꢗꢭꢗꢩꢟꢙꢫ ꢏ%ꢓ = ꢳꢴ
ꢵꢶ × 100
ꢥꢠꢦꢏꢁꢰꢰꢱꢓ − ꢥꢠꢦꢏꢠꢖꢗꢞꢜꢓ
4.5 Typical reaction procedure for the construction of rosmarinic acid derivatives (12c)
4.5.1 Methyl caffeate (4)
A mixture of caffeic acid (309.2 mg, 1.72 mmol), EDCI (385.3 mg, 2.01 mmol) and DMAP (22.1
mg, 0.181 mmol) in methyl alcohol (400 µL) was stirred at 60°C for 13 h. The resulting mixture
was quenched with 2N HCl and extracted with AcOEt. The combined organic phases were washed
with sat. NaHCO3, dried over anhydrous magnesium sulfate, and evaporated under reduced
pressure. The residue was purified by column chromatography on silica gel (MeOH–CHCl3) (1:5
v/v) to give a title compound (287.5 mg, 86%) as a white solid. Mp: 161°C–164°C [31], IR ṽ: 3489,
3303, 2955, 2925, 1679, 1601, 1519 cm−1. 1H-NMR (500 MHz, acetone-d6) δ: 7.51 (d, 1H, J = 16.1
Hz), 7.13 (d, 1H, J = 2.0 Hz), 7.02 (dd, 1H, J = 2.0, 8.0 Hz), 6.84 (d, 1H, J = 8.0 Hz), 6.26 (d, 1H, J
13
= 16.1 Hz), 3.69 (s, 3H). C-NMR (125 MHz, acetone-d6) δ: 167.0, 147.9, 145.5, 144.9, 126.8,
121.7, 115.6, 114.5, 114.4. High-resolution MS calcd. for C10H10O4 (M+): 194.0579. Found:
194.0585. All spectral data agreed with previously reported data [32].
4.5.2 Methyl 3,4-bis[[(2-nitrophenyl)sulfonyl]oxy] cinnamate (5)
To a solution of 4 (194.6 mg, 1.00 mmol) in CH3CN (2.2 mL) were added triethylamine (0.63 mL,
4.50 mmol) and 2-nitrobenzenesulfonyl chloride (487.1 mg, 2.30 mmol) at 0°C. The mixture was
stirred at 0°C for 0.5 h. The reaction mixture was quenched with 2M HCl and extracted with

ACCEPTED MANUSCRIPT
CH2Cl2. The combined organic phases were washed with brine, dried over anhydrous magnesium
sulfate and evaporated under reduced pressure. The residue was purified by column
chromatography on silica gel (AcOEt–n-hexane) (3:2 v/v) to give a title compound (562.9 mg, 99%)
as white solid. Mp: 118°C–123°C, IR ṽ: 3099, 1714, 1542, 1390, 1191, 850, 814, 693 cm−1.
1H-NMR (500 MHz, CDCl3) δ: 8.03 (dd, 2H, J = 1.3, 7.9 Hz), 7.89–7.84 (m, 2H), 7.81–7.74 (m,
4H), 7.58 (d, 11H, J = 16.0 Hz), 7.48 (dd, 1H, J = 2.0, 8.5 Hz), 7.45 (d, 1H, J = 2.0 Hz), 7.36 (d, 1H,
13
J = 8.5 Hz), 6.39 (d, 1H, J = 16.0 Hz), 3.82 (s, 3H). C-NMR (125 MHz, CDCl3) δ: 166.4, 148.2,
148.1, 141.6, 141.4, 141.1, 136.0, 135.1, 132.5, 131.6, 128.2, 128.1, 127.8, 124.9, 123.4, 120.7,
51.8.
4.5.3 3,4-Bis[(2-nitrophenyl)sulfonyl]oxycinnamic acid (6)
To a solution of 5 (113.3 mg, 0.20 mmol) in AcOH (3.1 mL) was added 0.5M HCl aq (3.1 mL) at
0°C. The mixture was stirred at 85°C for 37 h. The reaction mixture was added H2O and extracted
with AcOEt. The combined organic phases were washed with brine, dried over anhydrous
magnesium sulfate and evaporated under reduced pressure. The residue was purified by column
chromatography on silica gel (MeOH–CHCl3) (1:20 v/v) to give a title compound (110.4 mg, 99%)
as white solid. Mp: 182°C–186°C, IR ṽ: 3104, 1687, 1544, 1378, 1355, 1180, 1124, 851, 821, 721
cm-1. 1H NMR (500 MHz, acetone-d6): δ 8.12–8.04 (m, 6H), 7.97–7.92 (m, 2H), 7.79 (dd, 1H, J =
2.1, 8.6 Hz), 7.67 (d, 1H, J = 2.1 Hz), 7.64 (d, 1H, J = 16.0 Hz), 7.41 (d, 1H, J = 8.6 Hz), 6.55 (d,
13
1H, J = 16.0 Hz). C-NMR (125 MHz, acetone-d6) δ: 167.1, 148.49, 148.77, 142.3, 142.0, 141.8,
137.3, 136.1, 133.44, 133.40, 132.1, 128.8, 128.0, 125.8, 125.5, 125.3, 124.3, 121.9, 78.7.
High-resolution MS calcd for C21H14N2O12S2 (M+): 549.9988. Found: 549.9969.
4.5.4 Allyl (3,4-dihydroxyphenyl)acetate (8)
To a mixture of (3,4-dihydroxyphenyl)acetic acid (7, 1.01 g, 6.02 mmol), EDCI (1.38 g, 7.21 mmol)
and DMAP (73.4 mg, 0.60 mmol) was added allyl alcohol (4.5 mL) at room temperature, then
heated at 60°C for 2 h. The resulting mixture was quenched with 2N HCl and extracted with AcOEt.
The combined organic phases were washed with sat. NaHCO3, dried over anhydrous magnesium
sulfate and evaporated under reduced pressure. The residue was purified by column
chromatography on silica gel (acetone–n-hexane) (1:5 v/v) to give a title compound (1.11 g, 89%)
as yellow oil. IR ṽ: 3389, 1706, 1606, 1281, 988, 961, 795 cm−1. 1H NMR (500 MHz, acetone-d6): δ
6.77 (d, 1H, J = 2.2 Hz), 6.73 (d, 1H, J = 8.2 Hz), 6.59 (dd, 1H, J = 2.2, 8.2 Hz), 5.93–5.85 (m, 1H),
5.24 (dq, 1H, J = 1.6, 17.3 Hz), 5.13 (dq, 1H, J = 1.6, 10.6 Hz), 4.52 (dt, 2H, J = 1.6, 5.5 Hz), 3.47
(s, 2H). 13C NMR (125 MHz, acetone-d6): δ 171.8, 145.6, 144.8, 133.5, 126.7, 121.5, 117.7, 117.1,
115.9, 65.4, 40.8. High-resolution MS calcd for C11H12O4 (M+): 208.0736. Found: 208.0727.

ACCEPTED MANUSCRIPT
4.5.5 Allyl [3,4-bis(tert-butyldimethylsilyl)oxyphenyl]benezeneacetate (9)
To a solution of compound 8 (1.06 g, 5.11 mmol), TBSCl (1.86 g, 12.3 mmol) and imidazole (1.75
g, 25.7 mmol) in DMF (2.3 mL) was stirred at 0°C for 0.5 h. The reaction mixture was added sat.
NH4Cl and extracted with AcOEt. The combined organic phases were washed with sat. NaHCO3
and brine, dried over anhydrous magnesium sulfate, and evaporated under reduced pressure. The
residue was purified by column chromatography on silica gel (AcOEt–n-hexane) (1:20 v/v) to give a
title compound (2.16 g, 97%) as colorless oil. IR
MHz, CDCl3): δ 6.79 (d, 1H, J = 2.0 Hz), 6.76 (d, 1H, J = 8.0 Hz), 6.71 (dd, 1H, J = 2.0, 8.0 Hz),
ṽ
: 2930, 1739, 1228, 905, 836 cm−1. 1H NMR (500
5.91
–
5.84 (m, 1H), 5.24 (dq, 1H, J = 1.6, 17.2 Hz), 5.18 (dq, 1H, J = 1.3, 10.1 Hz), 4.58 (dt, 2H, J =
13
1.4, 5.7 Hz), 3.51 (s, 2H), 0.99 (s, 9H), 0.98 (s, 9H), 0.193 (s, 6H), 0.188 (s, 6H). C NMR (125
MHz, CDCl3): δ 171.5, 146.8, 146.1, 132.2, 127.0, 122.3, 122.2, 121.0, 118.2, 65.4, 40.8, 26.0, 18.5,
−4.0. High-resolution MS calcd for C23H40O4Si2 (M+): 436.2466. Found: 436.2463.
4.5.6 2-[3,4-bis(tert-butyldimethylsilyl)oxyphenyl]ethanol (10a)
To a suspension of LiAlH4 (278.9 mg, 4.90 mmol) in THF (2.00 mL) at 0°C were added compound
9 (2.13 g, 4.88 mmol) in THF (17.5 mL). The reaction mixture was stirred at room temperature for
0.5 h. The mixture was filtered through a Celite pad and evaporated under reduced pressure. The
residue was purified by column chromatography on silica gel (AcOEt–n-hexane) (1:4 v/v) to give a
title compound (1.83 g, 98%) as colorless oil. IR ṽ: 3326, 2929, 1463, 980, 834 cm−1. 1H NMR (500
MHz, CDCl3): δ 6.69 (d, 1H, J = 8.3 Hz), 6.62 (d, 1H, J = 2.0 Hz), 6.58 (dd, 1H, J = 2.0, 8.3 Hz),
3.70 (t, 2H, J = 6.5 Hz), 2.66 (t, 2H, J = 6.5 Hz), 0.92 (s, 9H), 0.91 (s, 9H), 0.121 (s, 6H), 0.117 (s,
13
6H). C NMR (125 MHz, CDCl3): δ 146.7, 145.4, 131.3, 121.84, 121.79, 120.1, 63.7, 38.5, 25.8,
18.4, −4.1. High-resolution MS calcd for C20H38O3Si2 (M+): 382.2360. Found: 382.2370.
4.5.7
2-[3,4-bis(tert-butyldimethylsilyl)oxyphenyl]ethyl
3,4-bis[[(2-nitrophenyl)sulfonyl]oxy]cinnamate (11d)
To a cooled (0°C) solution of 6a (278.5 mg, 0.51 mmol) and 10a (118.7 mg, 0.31 mmol) in dry
THF (1.6 mL) were added PPh3 (205.5 mg, 0.78 mmol) and DEAD (0.12 mL, 0.78 mmol). The
reaction mixture was stirred at room temperature for 21 h. Then, the reaction was worked up by
removal of the solvent, redissolved on AcOEt, and extracted with sat. NaHCO3. The organic phase
was washed with brine, dried over anhydrous magnesium sulfate, and evaporated under reduced
pressure. The residue was purified by column chromatography on silica gel (AcOEt–CHCl3) (1:30
v/v) to give a title compound (266.1 mg, 94%) as yellow-green solid. Mp: 57°C–61°C, IR ṽ: 2930,
1
2858, 1714, 1544, 1471, 1391, 1193, 1128, 903, 827, 781 cm−1. H NMR (500 MHz, CDCl3): δ

ACCEPTED MANUSCRIPT
8.04–8.01 (m, 2H), 7.88–7.84 (m, 2H), 7.80–7.73 (m, 4H), 7.55 (d, 1H, J = 16.1 Hz), 7.47–7.45 (m,
2H), 7.35 (d, 1H, J = 8.3 Hz), 6.77 (d, 1H, J = 8.2 Hz), 6.71 (d, 1H, J = 2.3 Hz), 6.68 (dd, 1H, J =
2.3, 8.2 Hz), 6.37 (d, 1H, J = 16.1 Hz), 4.36 (t, 2H, J = 7.0 Hz), 2.88 (t, 2H, J = 7.0 Hz), 0.981 (s,
9H), 0.979 (s, 9H), 0.192 (s, 6H), 0.189 (s, 6H). 13C NMR (125 MHz, CDCl3): δ 165.8, 148.2, 148.1,
146.5, 245.4, 141.6, 141.3, 141.1, 135.9, 135.1, 132.4, 131.7, 130.6, 128.2, 128.2, 127.8, 124.9,
1247, 123.4, 121.7, 121.6, 121.0, 120.965.5, 34.3, 25.8, 18.3, −4.2.
4.5.8 2-(3,4-Dihydroxyphenyl)ethyl caffeate (12c)
2-[3,4-bis(tert-butyldimethylsilyloxy)phenyl]ethyl caffeate
To a suspension of Cs2CO3 (1.19 g, 3.55 mmol) in CH3CN (1.90 mL) were added thiophenol (0.18
mL, 1.79 mmol) and 11d (132.2 mg, 0.14 mmol) at 0°C. The mixture was stirred at room
temperature for 1.5 h. The reaction was quenched with sat. NH4Cl and extracted with AcOEt. The
organic phase was dried over anhydrous magnesium sulfate and evaporated under reduced pressure.
The residue was purified by column chromatography on silica gel (MeOH–CHCl3) (1:100 v/v) to
give a title compound (64.6 mg, 82%) as yellow-green solid. Mp: 130°C–134°C, IR ṽ: 3463, 2929,
1
1670, 1278, 980, 906, 836, 782 cm−1. H NMR (500 MHz, CDCl3): δ 7.56 (d, 1H, J = 16.1 Hz),
7.08 (s, 1H), 7.60 (dd, 1H, J = 2.0, 8.0 Hz), 6.87 (d, 1H, J = 8.0 Hz), 6.76 (d, 1H, J = 8.0 Hz), 6.72
(d, 1H, J = 2.0 Hz), 6.68 (dd, 1H, J = 2.0, 8.0 Hz), 6.25 (d, 1H, J = 16.1 Hz), 4.34 (t, 2H, J = 7.0
13
Hz), 2.89 (t, 1H, J = 7.0 Hz), 0.98 (s, 18H), 0.19 (s, 12H). C NMR (125 MHz, CDCl3): δ 167.7,
146.6, 146.3, 145.5, 150.0, 143.7, 130.8, 127.5, 122.4, 121.8, 121.7, 121.0, 115.6, 115.5, 114.4,
65.4, 34.5, 25.9, 18.4, −4.1. High-resolution MS calcd for C29H44O6Si2 (M+): 544.2677. Found:
544.2688.
To a solution of above compound (60.4 mg, 0.11 mmol) in THF (3.6 mL) were added TBAF (1.1 M
in THF) (0.40 mL, 0.44 mmol) at 0°C. The mixture was stirred at room temperature for 2.5 h. Then,
the reaction was worked up by removal of the solvent, redissolved on AcOEt, and extracted with
H2O. The organic phase was washed with brine, dried over anhydrous magnesium sulfate, and
evaporated under reduced pressure. The residue was purified by column chromatography on silica
gel (MeOH–CHCl3) (1:20 v/v) to give a title compound (29.6 mg, 84%) as white solid. Mp: 118–
120 ºC [28], IR ṽ: 3283, 1680, 1394, 1168, 853, 810 cm−1. 1H NMR (500 MHz, MeOH-d4): δ 7.51
(d, 1H, J = 15.7 Hz), 7.02 (d, 1H, J = 2.0 Hz), 6.92 (dd, 1H, J = 2.0, 8.0 Hz), 6.76 (d, 1H, J = 8.0
Hz), 6.70–6.68 (m, 2H), 6.56 (dd, 1H, J = 2.0, 8.0 Hz), 6.23 (d, 1H, J = 15.7 Hz), 4.27 (t, 2H, J =
13
7.2 Hz), 2.82 (t, 2H, J = 7.2 Hz). C NMR (125 MHz, MeOH-d4): δ 169.3, 149.5, 146.9, 146.8,
146.2, 145.0, 144.9, 130.8, 127.7, 123.0, 121.2, 117.0, 116.5, 116.4, 115.1, 66.5, 35.6.
High-resolution MS calcd for C17H16O4 (M+): 316.0947. Found: 316.0962. All spectral data agreed
with previously reported data [33].

ACCEPTED MANUSCRIPT
4.6 Calculation of cLogP value of tested compounds
CLogP values of synthesized compounds were calculated using ChemBio3D Ultra 12.0
(PerkinElmer).
Ackowledgements
This work is supported by the MuIT Grant for Selective Research (K.U.) and JSPS KAKENHI
Grant Numbers 16K01909 (K.U.) and 16H03288 (K. T. and K.U.).
References
[1] C. R. Jack Jr., D. S. Knopman, W. J. Jagust, L. M. Shaw, P. S. Aisen, M. W. Weiner, R. C.
Petersen, J. Q. Trojanowski, Hypothetical model of dynamic biomarkers of the Alzheimer's
pathological cascade, Lancet Neurol. 9 (2010) 119-128.
[2] H. Hanyu, Lifestyle-related desease and dementia, Dementia Japan 30 (2016) 317-318.
[3] S. Gauthier, Alzheimer's disease: current and future therapeutic perspectives, Prog.
Neuropsychopharmacol. Biol. Psychiatry 25 (2001) 73-89.
[4] J. A. Hardy, G. A. Higgins, Alzheimer's disease: the amyloid cascade hypothesis, Science 256
(1992) 184-185.
[5] Y. Zhou, C. Jiang, Y. Zhang, Z. Liang, W. Liu, L. Wang, C. Luo, T. Zhong, Y. Sun, L. Zhao, X.
Xie, H. Jiang, N. Zhou, D. Liu, H. Liu, Structural optimization and biological evaluation of
substituted bisphenol a derivatives as β-amyloid peptide aggregation inhibitors, J. Med. Chem. 53
(2010) 5449-5466.
[6] A. F. McKoy, J. Chen, T. Schupbach, M. H. Hecht, Structure-activity relationships for a series
of compounds that inhibit aggregation of the Alzheimer's peptide, Aβ42, Chem. Biol. Drug. Design.
84 (2014) 505-512.
[7] M. Necula, R. Kayed, S. Milton, C. G. Glabe, Small molecule inhibitors of aggregation indicate
that amyloid β oligomerization and fibrillization pathways are independent and distinct, J. Biol.
Chem. 287 (2007) 10311-10324
.
[8] T. Arai, A. Ohno, K. Mori, T. Kakizawa, H. Kuwata, T. Ozawa, M. Shibanuma, S. Hara, S.
Ishida, M. Kurihara, N. Miyata, H. Nakagawa, K. Fukuhara, Design, synthesis, and evaluation of
Trolox-conjugated amyloid-β C-terminal peptides for therapeutic intervention in an in vitro model
of Alzheimer’s disease, Bioorg. Med. Chem. 24 (2016) 4138-4143.
[9] K. Murakami, K. Irie, H. Ohigashi, H. Hara, H.; Nagao, M.; Shimizu, T.; Shirasawa, Formation
and stabilization model of the 42-mer Abeta radical: implications for the long-lasting oxidative
stress in Alzheimer's disease, J. Am. Chem. Soc. 127 (2005) 15168-15174.

ACCEPTED MANUSCRIPT
[10] D. Yanagisawa, H. Taguchi, S. Morikawa, T. Kato, K. Hirao, N. Shirai, I. Tooyama, Novel
curcumin derivatives as potent inhibitors of amyloid
(2015) 357-368.
aggregation, Biochem. Biophys. Rep. 4
[11] H. Endo, Y. Nikaido, M. Nakadate, S. Ise, H. Konno, Structure activity relationship study of
curcumin analogues toward the amyloid-beta aggregation inhibitor, Bioorg. Med. Chem. Lett. 24
(2014) 5621-5626.
[12] Y. Ishigaki, H. Tanaka, T. Ogara, K. Uwai, K. Tokuraku, A microliter-scale high-throughput
screening system with quantum-dot nanoprobes for amyloid-β aggregation inhibitors, PLoS. ONE.
8 (2013) e72992.
[13] T. Hamaguchi, K. Ono, A. Murase, M. Yamada, Phenolic compounds prevent Alzheimer's
pathology through different effects on the amyloid-beta aggregation pathway, Am. J. Pathol. 175
(2009) 2557-2565.
[14] J. H. Chen, C.-T. Ho, Antioxidant activities of caffeic acid and its related hydroxycinnamic
acid compounds, J. Agric. Food. Chem. 45 (1997) 23742378.
[15] A. Kern, C. Behl, The unsolved relationship of brain aging and late-onsetAlzheimer disease.
Biochim. Biophys. Acta 1790 (2009) 1124–1132.
[16] S. J. Yang, W. J. Lee, E. A. Kim, K. D. Nam, H. G. Hahn, S. Y. Choi, S. W. Cho, Effects of
N-adamantyl-4-methylthiazol-2-amine on hyperglycemia, hyperlipidemia and oxidative stress in
streptozotocin-induced diabetic rats. Eur. J. Pharmacol. 736 (2014) 26–34.
[17] X.-L. Bu, P. P. N. Rao, Y.-J. Wang, Anti-amyloid aggregation activity of natural compounds:
implications for Alzheimer’s drug discovery, Mol. Neurobiol. 53 (2016) 3565-3575.
[18] J. Won, Y.-G. Hur, S.-J. Kim, S.-H. Park, D. Park, Pharmaceutical composition comprising
hydroxylphenyl derivatives of rosmarinic acid for anticancer, WO 2005072723 A1 2005-08-11.
[19] J. Won, K. Lee, S. Park, S.-J. Kim, S.-Y. Yun, M.-A. Kang, T.-G. Hur, J. Youn, Y. Yun, D.
Park, D., Derivatives of Hydroxyphenyl, A Method for Preparing Thereof and Their Pharmaceutical
Composition, WO 2003089405 A1 2003-10-30.
[20] Y. Liu, X. Li, D.-C. Xiong, B. Yu, X. Pu, X.-S. Ye, Synthetic Phenylethanoid Glycoside
Derivatives as Potent Neuroprotective Agents, Eur. J. Med. Chem. 95 (2015) 313-312.
[21] Z. Zhang, B. Xiao, Q. Chen, X.-Y. Lian, Synthesis and Biological Evaluation of Caffeic Acid
3,4-Dihydroxyphenethyl Ester, J. Nat. Prod. 73 (2010) 252-254.
[22] S. Okombi, D. Rival, A. Boumendjel, A.-M. Mariotte, E. J. L. Perrier, Use of
p-Hydroxycinnamic
acid
derivatives
in
cosmetic
or
dermatological
compositions, GB 2431876 A 2007-05-09.
[23] Z. N. Kayani, R. A. Lewis, S. Naseem, Development of novel chiral dopants to be used in
ferroelectric liquid crystal system, J. Mol. Liq. 180 (2013) 74-88.

ACCEPTED MANUSCRIPT
[24] Y. Zhu, A. Mohammadi, J. Ralph, Facile synthesis of 4-hydroxycinnamaldehydes, Bioenerg.
Res. 5 (2012) 407-411.
[25] M. S. Alavijeh, M. Chishty, M. Z. Qaiser, A. M. Palmer, Drug metabolism and
pharmacokinetics, the blood-brain barrier, and central nervous system drug discovery, NeuroRx. 2
(2005) 554-571.
[26] The RCSB PDB. http://www.rcsb.org/pdb/home/home.do
[27] Auto Dock. http://autodock.scripps.edu/
[28] S. M. D. Rizvi, S. Shakil, M. Haneef, A simple click by click protocol to perform docking:
Autodock 4.2 made easy for non-bioinformaticians, EXCLI J., 12 (2013) 831-857.
[29] G. K. Gouras, J. Tsai, J. Naslund, B. Vincent, M. Edgar, F. Checler, J. P. Greenfield, V.
Haroutunian, J. D. Buxbaum, H. Xu, P. Greengard, N. R. Relkin, N. R., Intraneuronal Abeta42
accumulation in human brain, Am. J. Pathol. 156 (2000) 15-20.
[30] B. Tepe, D. Daferera, A. Sokmen, M. Sokmen, M. Polissiou, Antimicrobial and antioxidant
activities of the essential oil and various extracts of Salvia tomentosa Miller (Lamiaceae), Food
Chem. 90 (2005) 333-340.
[31] K. Ohara, Y. Ichimura, S. Nagaoka, Kinetic study of singlet-oxygen quenching by cffeic acid
and related phenols, Bull Chem Soc Jpn, 82, (2009) 689-691.
[32] K. Uwai, Y. Osanai, T. Imaizumi, S. Kanno, M. Takeshita, M. Ishikawa, Inhibitory effect of
the alkyl side chain of caffeic acid analogues on lipopolysaccharide-induced nitric oxide production
in RAW264.7 macrophages, Bioorg. Med. Chem. 16 (2008) 7795-7803.
[33] L. Chapado, P. J. Linares-Palomino, S. Salido, J. Altarejos, J. A. Rosado, G. M. Salido, G. M.,
Synthesis and evaluation of the platelet antiaggregant properties of phenolic antioxidants
structurally related to rosmarinic acid, Bioorg Chem, 38 (2010) 108-114.