Development of curcumin-based amyloid β aggregation inhibitors for Alzheimer's disease using the SAR matrix approach

Article abstract of DOI:10.1016/j.bmc.2021.116357

Amyloid β (Aβ) aggregation inhibitor activity cliff involving a curcumin structure was predicted using the SAR Matrix method on the basis of 697 known Aβ inhibitors from ChEMBL (data set 2487). Among the compounds predicted, compound B was found to posses

Full text of DOI:10.1016/j.bmc.2021.116357

Bioorg. Med. Chem. 46 (2021) 116357
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Development of curcumin-based amyloid β aggregation inhibitors for
Alzheimer’s disease using the SAR matrix approach
,
Rohmad Yudi Utomoa, Yasunobu Asawaa, Satoshi Okadaa b, Hyun Seung Banc,
,b,*
Atsushi Yoshimorid, Jürgen Bajorathe, Hiroyuki Nakamuraa
a School of Life Science and Technology, Tokyo Institute of Technology, 4259 Nagatsuta‑cho, Midori‑ku, Yokohama 226‑8501, Japan
b Laboratory for Chemistry and Life Science, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta‑cho, Midori‑ku, Yokohama 226‑8503,
Japan
c Korea Research Institute of Bioscience and Biotechnology, Daejeon 34141, South Korea
d Institute for Theoretical Medicine, Inc., 26‑1, Muraoka‑Higashi 2‑chome, Fujisawa, Kanagawa 251‑0012, Japan
e
¨
Department of Life Science Informatics, B‑IT, LIMES Program Unit Chemical Biology and Medicinal Chemistry, Rheinische Friedrich-Wilhelms-Universitat Friedrich-
Hirzebruch-Allee 6, 53115 Bonn, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords:
Amyloid β (Aβ) aggregation inhibitor activity cliff involving a curcumin structure was predicted using the SAR
Matrix method on the basis of 697 known Aβ inhibitors from ChEMBL (data set 2487). Among the compounds
predicted, compound B was found to possess approximately 100 times higher inhibitory activity toward Aβ
aggregation than curcumin. TEM images indicate that compound B induced the shortening of Aβ fibrils and
increased the generation of Aβ oligomers in a concentration dependent manner. Furthermore, compound K, in
which the methyl ester of compound B was replaced by the tert-butyl ester, possessed low cytotoxicity on N2A
cells and attenuated Aβ-induced cytotoxicity, indicating that compound K would have an ability for preventing
neurotoxicity caused by Aβ aggregation.
SAR matrix
Curcumin
Amyloid β
Alzheimer’s disease
Neurotoxicity
1. Introduction
such antibodies, the immunogenic effect is a critical issue for therapy7,8
.
On the other hand, RS0406, a low-molecular-weight β-sheet breaker, has
entered a phase II clinical trial, but no significant clinical improvement
has been demonstrated. Therefore, more potent and efficacious Aβ in-
Since the Global leaders aim to find an effective way of treating
Alzheimer’s Disease (AD) by 20251, many researchers elucidate various
therapeutic strategies following recent findings concerning pathophys-
iological markers. The initial strategy using acetylcholine-based therapy
has thus far exerted the most progressive impact with four FDA-
approved drugs2. However, these compounds only relieve symptoms
with moderate clinical efficacy3. Other strategies targeting the mis-
folded Tau protein are challenged by lack of efficacy in clinical trials and
drug administration problems4,5. Learning from these clinical observa-
tions, recent research strategies focus on targeting amyloid β (Aβ) ag-
gregation, following a well-known hypothesis of therapeutic AD
intervention6.
hibitors are still required9,10
.
Curcumin, a natural polyphenol, was proposed to be a promising
candidate with considerable apparent efficacy in transgenic models of
AD; however, in phase II clinical trials, non-negligible side effects such
as gastrointestinal symptoms were observed11. Evaluation of curcumin
derivatives has indicated that the two benzene rings, Michael acceptor
framework, and length of linkers contribute to the Aβ inhibitory activ-
ity12. In addition, molecular modeling studies on curcumin in complex
with Aβ have suggested contributions of the ketone groups and methoxy
substituents on its benzene rings to biological activity13,14. So far, over
200 curcumin derivatives have been synthesized to further optimize the
curcumin structure as a potential Aβ inhibitor, and some of these com-
pounds have shown a promising apparent inhibition of the Aβ fibril
Aβ-targeted strategies, which account for 54% of AD clinical trials,
however, encounter several obstacles. A series of trials have been con-
ducted with monoclonal antibodies, such as Solanezumab and Aduca-
numab, which bind directly to Aβ. Despite its promising efficacies of
formation15,16
.
* Corresponding author at: Laboratory for Chemistry and Life Science, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta‑cho,
Midori‑ku, Yokohama 226‑8503, Japan.
E-mail address: hiro@res.titech.ac.jp (H. Nakamura).
https://doi.org/10.1016/j.bmc.2021.116357
Received 7 July 2021; Received in revised form 30 July 2021; Accepted 1 August 2021
Available online 8 August 2021
0968-0896/© 2021 Elsevier Ltd. All rights reserved.

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
We note, however, that curcumins are known as pan assay interfer-
ence compounds (PAINS)17 that may give raise to assay artifacts and
false-positive activity readouts. Artificial activities of curcumin are
intensely debated18 and special care must be taken to further assess
apparent activities of curcuminoids. Nonetheless, in light of previous
findings, we have further investigated curcumin derivatives as potential
inhibitors of fibril formation. Several lines of evidence are provided in
support of true activity of curcumin derivatives we have developed on
the basis of curcuminoids available in the ChEMBL database, as reported
in the following19,20. Specifically, given the uncertainties associated
with curcumin activities, we have attempted to systematically explore
structural relationships between curcumin derivatives and associated
structure–activity relationships. We have reasoned that this analysis
scheme would be less vulnerable to anticipated incidents of false-
positives reported in the literature than focusing on individual curcu-
min derivatives.
models17. In general, the activity is gradually changing with com-
pound modifications in the presence of SAR continuity. However, in the
presence of SAR discontinuity, the activity can drastically change with
small chemical modifications, leading to the formation of an activity
cliff (i.e., a pair of structural analogues with a large difference in po-
tency)22,23. Local Free-Wilson type QSAR predictions are generally
confined to compound subsets forming continuous SARs, but SARMs
also enable the detection of activity cliffs in regions of SAR
discontinuity.
2. Results and discussion
To construct SARMs for our analysis, 697 known Aβ inhibitors with
available IC50 value were retrieved from ChEMBL (data set 2487). After
the existing curcumin derivatives or analogues were fragmented by
systematic cleavage of exocyclic bonds using the SARM methodology,
SARMs were obtained including virtual curcumin derivatives, and the
activity of several of these derivatives was predicted. (Figure 1).
Analyzing the compound distribution across SARMs using the Molecular
Grid Map (MGM)24, we focused on a putative activity cliff associated
with virtual analogues A–C in Figure 1, for which the activity was pre-
dicted. An activity cliff is formed by any of these compounds and the
known weakly active analogs at adjacent positions in the MGM. All
compounds in this activity cliff region possessed distinct benzene sub-
stitution patterns with combination of several functional groups such as
hydroxyl, methoxy, methyl methoxyacetate, methyl hydroxyacetate,
and trifluoromethyl groups. In addition to compounds A–C, we also
synthesized compound D as a new curcumin derivative not originating
In this paper, we introduced the success application of SAR Matrix
(SARM) to generate novel curcumin derivatives with improved biolog-
ical activity. We have applied SARM approach21, which systematically
extracts analogue series from compound data sets and organizes series
with structurally related cores in individual matrices akin of R-group
tables. Through color coding of matrix cells representing compounds,
activity information can be included in the analysis. We have recently
shown that the SARM method is useful for systematic analysis of SAR
data sets. For structurally related analogue series, SARM produces vir-
tual analogues of known active compounds that represent not yet
explored combinations of existing core structures and R-groups whose
potential activity can be predicted by SARM-based local Free-Wilson
Fig. 1. Activity cliff prediction from SAR Matrix (SARM) method.
2

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
from the SARM and compounds E and F as the reported curcumin de-
rivatives with Aβ-inhibitory activity.
Table 1
Inhibition of Aβ aggregation using ThT assay[a]
Compound
IC50 (µM) [b]
Synthesis of curcumin derivatives A–F is shown in Scheme 1. Alde-
hydes 1a–b were chosen as starting materials and reacted with methyl
bromoacetate under basic conditions. The resulting aldehyde 2a un-
derwent the condensation reaction with acetylacetone to give the cor-
responding conjugated ketone 3a in 61% yield. The microwave (MW)
irradiated condition is essential for this condensation reaction and the
desired product 3a was obtained only in 5% yield without MW irradi-
ation. Further condensation reaction was performed between compound
3a with compound 1b to obtain asymmetric curcumin derivative A. For
obtaining other asymmetric curcumins, first compound 1a was reacted
with acetyl acetone to give compound 4a. Further condensation reaction
with various aldehydes, such as compounds 1b, 2a-b, and 3-trifluoro-
methyl benzaldehyde, produced asymmetric curcumin derivatives E,
F, B, and D. Compound C was obtained from the condensation reaction
between compound 2c (obtained by hydrolysis of compound 2b) with
compound 4a.
A
0.022 ± 0.013
0.007 ± 0.001
0.698 ± 0.010
0.013 ± 0.002
0.052 ± 0.006
0.029 ± 0.005
0.693 ± 0.013
0.022 ± 0.043
B
C
D
E
F
Curcumin
Tacrine
[a]
The ThT fluorescence assay was carried out by
incubating 20 µM of Aβ with serial concentration of
curcumin derivatives in PBS pH 7.4 at 37 ◦C. [b] The
drug concentration required to inhibit Aβ aggregation
by 50% (IC50) was determined from semi-logarithmic
dose–response plots, and results represent the mean
± s.d. of triplicate samples.
Since the predicted curcumin analogues and their derivatives were
successfully synthesized, we next examined the in vitro effect of the
synthesized compounds on the inhibitory activity of Aβ aggregation. In
this study, we choose Aβ42 considering as the most neurotoxic Aβ iso-
form24. The aggregation of Aβ42 occurred immediately under the bio-
logical conditions (pH 7.4), thus could be detected by thioflavin T (ThT)
which is known to be interacted on hydrophobic site of Aβ4225. In this
assay, an Aβ42 inhibitor was defined by the ability of the compound to
competitively interact at the binding site of ThT, resulting in the
decreasing of fluorescence intensity during detection. In the case of
curcuminoids, results from this assay must be considered with special
care18. Therefore, detected activities were further investigated, as
described below. The results we have obtained using the ThT assay are
summarized in Table 1. Curcumin inhibited the aggregation in a dose-
dependent manner with the IC50 Aβ42 value of 0.693 ± 0.013 µM.
Further screening using the ThT fluorescence assay revealed that several
compounds containing hydroxyl group and methoxycarbonylmethyloxy
group on different positions on their benzene ring such as compounds A,
B, and F displayed stronger inhibitory activity than curcumin with IC50
values of 0.022 ± 0.004, 0.007 ± 0.001, and 0.029 ± 0.005, respec-
tively. In addition, compounds D and E, which do not contain a
methoxycarbonylmethyloxy group but contain a hydroxyl group, also
exhibited comparable inhibitory activity to compound F, indicating a
crucial role for the trifluoromethyl and hydroxyl groups. Although the
presence of the carboxymethyloxy group and hydroxyl group on
different benzene rings decreased the inhibitory activity, the activity
was still comparable to curcumin as shown by compound C. We also
confirmed the ability of SARM to predict the Aβ inhibitory activity of
novel curcumin derivatives. On the basis of the data we obtained, SARM
analysis was continued in search of other promising derivatives.
The results observed from ThT fluorescence assay revealed that
several curcumin derivatives possessed higher Aβ aggregation inhibitory
activity than curcumin. Therefore, we updated the SARM analysis by
including the experimental IC50 values of all compounds to find other
more potent curcumin derivatives. The revised SARM analysis revealed
several other curcumin derivatives, including novel compounds G, H, I,
and J, possessing more potent inhibitory activity than curcumin and
comparable with compound B (Figure S1).
Synthetic routes for the preparation of compounds G, H, I, and J,
which were derived from the updated-SARM analysis are also shown in
Scheme 1. The compounds G and H were prepared by reacting com-
pound 3a with 3-trifluoromethyl benzaldehyde and compound 2b,
respectively. For compounds I and J were obtained from condensation
reaction between compound 4b with 3-trifluoromethyl benzaldehyde or
compound 2b. Furthermore, we also synthesized compound K, a de-
rivative of compound B, in which the methyl ester was replaced by the
tert-butyl ester. The tert-butyl group was first introduced on compound
1b, and the resulting compound 5 was treated with 4a to afford com-
pound K.
To further assess SARM predictions, compounds G, H, I, and J were
tested in the ThT assay (Table 2). The dose-dependent ThT assay showed
these curcumin derivatives were more potent than curcumin: their IC50
values were 0.018–0.109 µM whereas the IC50 of curcumin was 0.693
µM. Interestingly compounds
D
and
I
which have
a
3-
Scheme 1. Synthetic scheme of curcumin derivatives A–K. Reaction conditions: (a) DMF, K2CO3, 80 ◦C, 2 h; (b) DMF, B(OH)3, morpholine, MW 150 ◦C, 15 min; (c)
NaOH, H2O, 100 ◦C, 30 min.
3

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
observed the formation of Aβ fibrils and oligomers, indicated by long-
filament surrounded by spherical-shape species31. The addition of
compound B induced shortening of Aβ fibrils and increased the gener-
ation of Aβ oligomers in a concentration dependent manner, resulting in
the inhibition of the fibril formation from oligomers. These findings
provided strong support for the validity of the detected activity of the
new curcumin derivatives.
Table 2
Inhibition of Aβ aggregation using ThT assay.
compound
SARM prediction
IC50 (µM)
G
H
I
8.8
8.5
8.8
8.2
–
0.046 ± 0.012
0.055 ± 0.009
0.018 ± 0.001
0.109 ± 0.037
0.055 ± 0.006
J
K
3. Conclusion
trifluoromethylphenyl group in one of two aromatic rings on the cur-
cumin skeleton were more potent than tacrine and comparable with
compound B. These data indicated the importance of the trifluoromethyl
group for the amyloid inhibition activity.
In conclusion, we identified compound B as a potent Aβ inhibitor
from a curcumin-based activity cliff region originating from SARM
analysis and revealed by MGM. Compound B possessed low cytotoxicity
on N2A cells and attenuate Aβ-induced cytotoxicity. Negative stained
TEM images indicated that compound B induced the shortening of Aβ
fibrils and increased the generation of Aβ oligomers in a concentration
dependent manner. Furthermore, compound K in which the methyl ester
of compound B was replaced by the tert-butyl ester possessed lower
cytotoxicity on N2A cells than compound B and attenuated Aβ-induced
cytotoxicity, indicating that compound K has the ability to prevent
neurotoxicity caused by Aβ aggregation. Taken together, the results
provide several lines of evidence for the relevance of the activities of
curcumin derivatives observed in our study, although curcumin has
PAINS character and is prone to eliciting assay artifacts.
Aggregation of Aβ monomer to oligomers or fibrils has been reported
to cause neurotoxicity in Alzheimer’s disease26. Therefore, we next
evaluated whether the synthesized curcumin derivatives prevented the
neurotoxicity by inhibiting Aβ aggregation. We chose compounds B and
D as the most potent inhibitors and compound K as a derivative of
compound B for comparison. We used mouse neuroblastoma (N2A) and
human glioblastoma (U87MG) cells, that have been commonly used for
the Aβ cytotoxicity study of Alzheimer’s disease27,28. We first examined
cell viability of selected curcumin derivatives toward N2A and U87MG
cells using 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium
bromide (MTT) assay. The results are summarized in Table 3. Curcumin
induced a cytotoxic effect on N2A and U87MG cells with IC50 values of
22.6 ± 1.48 and 18.4 ± 3.59, which were similar to those previously
reported29. In contrast, the curcumin derivatives B, D, and K were less
cytotoxic toward both cell lines than curcumin. Especially, compound K
did not show significant cytotoxicity at 100 µM concentration. We next
examined effects of the curcumin derivatives B and K on Aβ-stimulated
cytotoxicity in N2A cells. As expected, Aβ caused a dose-dependent
cytotoxic effect on N2A cells as reported30 (Figure 2A). Next, the ef-
fects of compounds B and K on the viability of N2A cells stimulated by
Aβ were evaluated (Figure 2B). The cell viability was reduced to ca. 55%
by Aβ (10 µM). In the presence of compound B at 1 µM concentration,
the cell viability was reduced to ca. 70% with or without Aβ, indicating
that compound B relieved the Aβ-stimulated cytotoxicity. Indeed, these
reduced cell viabilities were caused by the cytotoxicity of compound B.
Surprisingly, compound K was found not only to possess low cytotox-
icity but also a significant protective ability against the Aβ-stimulated
cytotoxicity, and >90% of cell viability was observed in the Aβ-stimu-
lated cells treated with compound K. These results indicated that com-
pound K would was able to prevent neurotoxicity caused by Aβ
aggregation.
4. Experimental section
SAR Matrix software beta version was provided under the licensed
from Dr. Atsushi Yoshimori from Institute for Theoretical Medicine, Inc,
Japan. All the solvents used were in analytical standard grade. The NMR
spectra were measured on a Bruker biospin AVANCE II (400 MHz for 1H
and 100 MHz for 13C) or a Bruker biospin AVANCE III (500 MHz for 1H,
125 MHz for 13C, and 470 MHz for 19F). Chemical shifts (δ) was reported
in ppm relative to internal tetramethylsilane. The HRMS data were
recorded on Bruker ESI-TOF-MS micrOTOF II instrument with sodium
formate as the calibration standard. Transmission electron microscopy
was performed using Hitachi H-8100 operated at 200 kV. Vanillin was
purchased from Wako (Japan). Acetylacetone and methyl bromoacetate
were purchased from Sigma (USA). Morpholine was purchased from
Wako (Japan). The 4-hydroxy-3-(trifluoromethyl) benzaldehyde was
purchased from Fluorochem (UK). The 3-(trifluoromethyl) benzalde-
hyde was purchased from Combi-Blocks (USA). The 4-fluoro-3-(trifluor-
omethyl) benzaldehyde and tert-butyl bromoacetate were purchased
from Tokyo Chemical Industry (Japan). Tacrine hydrochloride was
purchased from Cayman Chemical (USA). Microwave for synthesis was
conducted on Biotage® Initiator + instrument. Column chromatography
was performed on silica gel Chromatorex (Japan). Purity analysis of
compound A-K was determined by HPLC analysis using Inertsil ODS-3 5
µm (4.6 × 75 mm; GL Science) with a linear gradient of 0.1% formic acid
in water/0.1% formic acid in MeCN detected by UV lamp for 25 min.
The amyloid β (Aβ42) peptide was purchased from Peptide Institute
(Japan). The thioflavin T for fluorescence detection of Aβ42 was pur-
chased from Sigma (USA). The fluorescence intensity was measured on
Tecan Infinite 200.
The ThT assay revealed the effectiveness of the curcumin derivatives
synthesized in the current study to inhibit Aβ aggregation. To further
examine the morphological change after the treatment of Aβ monomer
with compound B, which would confirm the activity detected in the
assays discussed above, we performed negative stained TEM imaging
under similar conditions of the ThT assay. As shown in Figure 3, we
Table 3
Cell viability of selected compounds toward N2A and U87MG cells using MTT
assay[a]
compound
IC50 (µM)[b]
4.1. Synthesis of compounds
N2A
U87MG
4.1.1. General procedure to Synthesis compound 2a and 2b
B
38.7 ± 8.47
65.7 ± 7.44
>100
90.0 ± 5.93
88.7 ± 2.78
>100
Compound 1a or 1b (1 mmol) and potassium carbonate (1 mmol)
were dissolved in DMF 10 mL. Methyl bromoacetate (1.1 mmol) was
added then the mixture was refluxed for 2 h at 80 ◦C. The reaction was
quenched by HCl 0.1 N then extracted by ethyl acetate. The organic
layer was dried using MgSO4 then, purified by column chromatography
(silica gel, hexane:ethyl acetate 2:1). Methyl 2-(4-formyl-2-methox-
yphenoxy) acetate (2a). Yield: 96%. 1H NMR (400 MHz, CDCl3): δ
(ppm) 3.86 (–OCH3, s, 3H), 4.01 (–OCH3, s, 3H), 4.84 (–CH2, s, 2H), 6.92
D
K
Curcumin
22.6 ± 1.48
18.4 ± 3.59
[a]
The MTT assay was carried out after treating the cells with serial concen-
tration of curcumin derivatives for 24 h. [b] The drug concentration required to
inhibit the cell growth by 50% (IC50) was determined from semilogarithmic
dose–response plots, and results represent the mean ± s.d. of three independent
experiments in triplicate.
4

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
Fig. 2. Curcumin derivatives attenuate Aβ-induced cytotoxicity. (A) Cytotoxic effect of Aβ on N2A cells. (B) Effect of compounds B and K on the cell viability of Aβ-
stimulated N2A cells.
Fig. 3. Negative stained TEM images of Aβ incubated at 37 ◦C in the absence and presence of Compound B. Fibril (→), oligomer (–>).
(=CH, J = 8.1, d, 1H), 7.46 (=CH, J = 5.5, 1.8 Hz, dd, 1H), 7.49 (=CH, J
= 1.7, d, 1H), 9.92 (–CHO, s, 1H). 13C NMR (100 MHz, CDCl3): δ (ppm)
52.89, 56.52, 66.24, 110.34, 112.78, 126.62, 131.59, 150.37, 152.89,
169.01, 190.38. LRMS-ESI (m/z): calcd for C11H12O5 224.07; found
225.09 [M + H]+. Methyl 2-(4-formyl-2-(trifluoromethyl)phenoxy)
acetate (2b). Yield: 94%. 1H NMR (400 MHz, CDCl3): δ (ppm) 3.82
(–OCH3, s, 3H), 4.01 (–OCH3, s, 3H), 4.76 (–CH2, s, 2H), 6.88 (=CH, J =
8.8, d, 1H), 8.05 (=CH, J = 8.4, d, 1H), 8.18 (=CH, s, 1H), 9.96 (–OH, s,
1H). 13C NMR (100 MHz, CDCl3): δ (ppm) 52.15, 65.21, 112.86, 119.55
(JC-F = 30.7 Hz, d), 122.69 (JC-F = 271.4 Hz, d), 128.81 (JC-F = 5.2 Hz,
d), 129.59, 134.92, 159.97, 159.98, 167.66, 189.54. LRMS-ESI (m/z):
calcd for C11H19F3O4 262.05; found 285.09 [M + Na]+.
4.1.3. General procedure to Synthesis compound 3a
Compound 3a was synthesized using microwave-assisted reaction.
Acetylacetone (20 mmol) and boric acid (20 mmol) were dissolved in
DMF. Compound 2a (2 mmol) was added followed by morpholine (0.6
mmol). The mixture was irradiated in microwave at 150 ◦C for 10 min.
The reaction was stopped by the addition of HCl 0.1 N. Crude product
was extracted by ethyl acetate then the organic phase was dried using
MgSO4. Purification to obtain the desired compound was conducted
using column chromatography (silica gel, hexane:ethyl acetate 2:1).
Methyl 2-{4-[(1E)-3,5-dioxohex-1-en-1-yl]-2-methoxyphenoxy}ac-
etate (3a). Yield: 61%. 1H NMR (400 MHz, CDCl3): δ (ppm) 3.85
(–OCH3, s, 3H), 3.97 (–OCH3, s, 3H), 4.78 (–CH2, s, 2H), 5.69 (–CH2, s,
2H), 6.39 (=CH, J = 15.6 Hz, d, 1H), 6.83 (=CH, J = 8.8 Hz, d, 1H), 7.10
(=CH, J = 6.8 Hz, d, 1H), 7.12 (=CH, J = 6.8 Hz, d, 1H), 7.57 (=CH, J =
16.0 Hz, d, 1H).). 13C NMR (125 MHz, CD3CN): δ (ppm) 27.00, 52.66,
56.42, 66.25, 101.72, 111.66, 114.10, 118.32, 122.18, 122.87, 130.09,
149.94, 150.45, 169.98, 178.59, 198.89. LRMS-ESI (m/z): calcd for
4.1.2. General procedure to Synthesis compound 2c
Compound 2b (0.26 mmol) was dissolved in 0.5 mL methanol, then
2 mL of NaOH 1 M was added. The mixture was stirred for 1 h on 100 ◦C.
The reaction was quenched by the addition of HCl 0.1 M then was
extracted using ethyl acetate. The desired compound was purified by
column chromatography (silica gel, hexane:ethyl acetate 2:1). 2-(4-
formyl-2-(trifluoromethyl)phenoxy)acetic acid (2c). Yield: 41%. 1H
NMR (400 MHz, CDCl3): 4.94 (=CH, J = 8.6 Hz, d, 1H) 7.08 (=CH, J =
8.6 Hz, d, 1H), 8.09 (=CH, J = 8.7 Hz, d, 1H), 8.21 (=CH, J = 1.6, d, 1H),
9.99 (–CHO, s, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 65.43,
68.99, 113.89, 117.19 (JC-F = 30.9 Hz, q), 123.30 (JC-F = 270.9 Hz, q),
123.33, 128.09 (JC-F = 5.7 Hz, q), 128.84 (JC-F = 5.6 Hz, q), 129.31,
135.41, 159.15, 166.04, 169.25, 191.02. LRMS-ESI (m/z): calcd for
C
16H18O6 306.11; found 307.02 [M + H]+.
4.1.4. General procedure to Synthesis compound a and G
Compound A and G was synthesized using microwave-assisted re-
action. Compound 3a (0.26 mmol) and boric acid (0.26 mmol) was
suspended in DMF. Compound Compound 1b or compound 3-trifluoro-
methyl benzaldehyde (0.26 mmol) was added followed by morpholine
(0.1 mmol). The mixture was irradiated in microwave at 150 ◦C for 10
min. HCl 0.1 N was added to quench the reaction. Crude product was
extracted by ethyl acetate then dried using MgSO4. The desired com-
pound was purified by column chromatography (silica gel, hexane:ethyl
C
10H7F3O4 248.03; found 247.12 [Mꢀ H]-.
5

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
acetate 2:1). Methyl 2-(4-((1E,6E)-7-(4-hydroxy-3-(trifluoromethyl)
phenyl)-3,5-dioxohepta-1,6-dien-1-yl)-2-methoxyphenoxy)acetate
(A). Yield: 40%. 1H NMR (400 MHz, CD3CN): δ (ppm) 3.76 (–OCH3, s,
3H), 3.91 (–OCH3, s, 3H), 4.76 (–CH2, s, 2H), 5.47 (–CH2, s, 2H), 5.47
(=CH, s, 1H), 5.97 (=CH, s, 1H), 6.72 (=CH, J = 16.0 Hz, d, 1H), 6.76
(=CH, J = 16.0 Hz, d, 1H), 6.91 (=CH, J = 8.4 Hz, d, 1H), 7.09 (=CH, J
= 8.4 Hz, d, 1H), 7.18 (=CH, J = 8.4 Hz, d, 1H), 7.30 (=CH, J = 1.9 Hz,
d, 1H), 7.62 (=CH, J = 16.0 Hz, d, 1H), 7.64 (=CH, J = 16.0 Hz, d, 1H),
7.76 (=CH, J = 8.4, 2.0 Hz, dd, 1H), 7.86 (=CH, J = 1.6 Hz, d, 1H), 8.34
(=CH, s, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 52.16, 55.99,
58.75, 65.28, 101.67, 111.32, 112.35 (JC-F = 270.3 Hz, q), 116.36 (JC-F
= 32.1 Hz, q), 117.95, 122.90, 125.95, 127.62, (JC-F = 4.4 Hz, q),
128.28, 128.50 (JC-F = 3.8 Hz, q), 128.74, 133.66, 135.82, 139.35,
140.63, 149.37, 157.94, 169.29, 183.31, 183.65. 19F NMR (470 MHz,
DMSO‑d6): δ 61.2. HRMS-ESI (m/z): calcd for C24H21F3O7 478.1239;
found 477.1241 [Mꢀ H]-.HPLC purity 94.3%, retention time 5.74 min.
chromatography (silica gel, hexane:ethyl acetate 2:1). (1E,6E)-1-(4-
hydroxy-3-methoxyphenyl)-7-[3-(trifluoromethyl)phenyl]hepta-
1,6-diene-3,5-dione (B). Yield: 30%. 1H NMR (400 MHz, DMSO‑d6):
3.72 (–OCH3, s, 3H), 3.85 (–OCH3, s, 3H), 5.07 (–CH2, s, 2H), 6.12 (=CH,
s, 1H), 6.81 (=CH, J = 15.8 Hz, d, 1H), 6.83 (=CH, J = 8.1 Hz, d, 1H),
6.93 (=CH, J = 15.9 Hz, d, 1H), 7.18 (=CH, J = 8.1 Hz, d, 1H), 7.2%
(=CH, J = 8.8 Hz, d, 1H), 7.34 (=CH, s, 1H), 7.58 (=CH, J = 15.8, d,
1H), 7.63 (=CH, J = 15.9, d, 1H), 7.98 (=CH, J = 8.6, d, 1H), 8.01 (=CH,
s, 1H).). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 52.19, 55.88, 65.29,
101.64, 111.59, 114.36, 115.94, 118.01 (JC-F = 30.9 Hz, q), 121.29,
123.54, 123.54 (JC-F = 270.4 Hz, q), 124.19, 126.45, 127.09 (JC-F = 4.7
Hz, q), 128.17, 133.81, 137.99, 141.68, 148.23, 149.78, 156.70, 168.55,
181.73, 184.89. 19F NMR (470 MHz, DMSO‑d6): δ 60.9. HRMS-ESI (m/
z): calcd for C24H21F3O7 478.1239; found 477.1235 [Mꢀ H]-. HPLC
purity 96.1%, retention time 5.88 min. 2-(4-((1E,6E)-7-(4-hydroxy-3-
methoxyphenyl)-3,5-dioxohepta-1,6-dien-1-yl)-2-(trifluoromethyl)
phenoxy)acetic acid (C). Yield: 15%. 1H NMR (400 MHz, CD3CN): 3.98
(–OCH3, s, 3H), 4.84 (–CH2, s, 2H), 5.98 (=CH, s, 1H), 6.72 (=CH, J =
16.0 Hz, d, 1H), 6.78 (=CH, J = 16.0 Hz, d, 1H), 6.89 (=CH, J = 8.0 Hz,
d, 1H), 7.10 (=CH, J = 8.8 Hz, d, 1H), 7.18 (=CH, J = 8.2, 1.9 Hz, dd,
1H), 7.29 (=CH, J = 1.8 Hz, d, 1H), 7.65 (=CH, J = 15.6 Hz, d, 2H), 7.84
(=CH, J = 8.6 Hz, d, 1H), 7.94 (=CH, J = 1.9 Hz, d, 1H). 13C NMR (125
MHz, DMSO‑d6): δ (ppm) 55.92, 66.81, 101.58, 111.59, 114.52, 115.99,
Methyl
2-(4-((1E,6E)-3,5-dioxo-7-(3-(trifluoromethyl)phenyl)
hepta-1,6-dien-1-yl)-2-methoxyphenoxy)acetate (G). Yield: 35%. 1H
NMR (400 MHz, CDCl3): δ (ppm) 3.83 (–OCH3, s, 3H), 3.97 (–OCH3, s,
3H), 4.77 (–CH2, s, 2H), 5.85 (–CH2, s, 1H), 6.55 (=CH, J = 15.8 Hz, d,
1H), 6.69 (=CH, J = 15.8 Hz, d, 1H), 6.83 (=CH, J = 8.7 Hz, d, 1H), 7.13
(=CH, s, 1H), 7.15 (=CH, J = 7.2, 1.8 Hz, dd, 1H), 7.55 (=CH, J = 8.0
Hz, t, 1H), 7.65 (=CH, J = 15.8 Hz, d, 1H), 7.69 (=CH, J = 15.8 Hz, d,
1H), 7.73 (=CH, J = 7.8 Hz, d, 1H), 7.82 (=CH, s, 1H). 13C-NM‘R (125
MHz, DMSO‑d6): δ (ppm) 52.01, 55.87, 65.14, 102.08, 111.27, 113.25,
122.71, 122.95, 124.19 (JC-F = 270.9 Hz, q), 124.82, 126.40, 128.49,
130.02 (JC-F = 32.8 Hz, q), 130.16, 132.05, 136.11, 138.01, 141.33,
149.27, 149.39, 169.11, 181.21, 185.28. 19F NMR (470 MHz,
DMSO‑d6): δ 61.2. HRMS-ESI (m/z): calcd for C24H21F3O6 462.1290;
found 461.1280 [Mꢀ H]-. HPLC purity 90.7%, retention time 6.76 min.
117.66 (JC-F = 30.2 Hz, q), 121.29, 122.67, 123.58, 123.74 (JC-F
=
270.7 Hz, q), 126.40, 126.62, 126.74, 127.04 (JC-F = 4.7 Hz, q), 127.22,
127.49 (JC-F = 5.2 Hz, q), 133.61, 138.37, 141.62, 141.98, 148.28,
157.84, 170.25, 182.01, 184.68, 198.26. 19F NMR (470 MHz,
DMSO‑d6): δ 60.8. HRMS-ESI (m/z): calcd for C23H19F3O7 464.1083;
found 463.1087 [Mꢀ H]-. HPLC purity 96.2%, retention time 5.69 min.
methyl 2-(4-((1E,6E)-7-(4-hydroxy-3-methoxyphenyl)-3,5-dioxo-
hepta-1,6-dien-1-yl)-2-methoxyphenoxy)acetate (F). Yield: 35%. 1H
NMR (400 MHz, CDCl3): δ (ppm) 3.82 (–OCH3, s, 3H), 3.94 (–OCH3, s,
3H), 3.94 (–OCH3, s, 3H), 4.75 (–CH2, s, 2H), 5.82 (–CH2, s, 1H), 6.47
(=CH, J = 15.8 Hz, d, 1H), 6.52 (=CH, J = 15.8 Hz, d, 1H), 6.82 (=CH, J
= 8.8 Hz, d, 1H), 6.94 (=CH, J = 8.2 Hz, d, 1H), 7.04–7.13 (=CH, m,
4.1.5. General procedure to Synthesis compound 4a and 4b
Compound 4a and 4b was synthesized using microwave-assisted
reaction. Acetylacetone (20 mol) and boric acid (20 mmol) were sus-
pended in DMF. Compound 1a or 1b (2 mmol) was added then followed
by morpholine (0.6 mmol). The mixture was irradiated in microwave at
150 ◦C for 10 min. The reaction mixture was quenched by HCl 0.1 N and
extracted using ethyl acetate. The organic phase was dried using MgSO4.
The crude product was purified by column chromatography (silica gel,
hexane:ethyl acetate 2:1) to obtain the desired compound and curcumin.
(4-hydroxy-3-methoxyphenyl) hex-5-ene-2,4-dione (4a). Yield:
65%. 1H NMR (400 MHz, CDCl3): δ (ppm) 2.20 (–CH3, s, 3H), 3.98
(–OCH3, s, 3H), 5.68 (=CH, s, 1H), 5.87 (=CH, s, 1H), 6.37 (=CH, J =
15.8, d, 1H), 6.97 (=CH, J = 8.2, d, 1H), 7.06 (=CH, J = 1.8 Hz, d, 1H),
4H), 7.58 (=CH, J = 15.8 Hz, d, 1H), 7.62 (=CH, J = 15.8 Hz, d, 1H). 13
C
NMR (100 MHz, CDCl3): δ (ppm) 52.29, 55.96, 66.15, 101.28, 109.75,
110.87, 113.77, 114.89, 121.76, 122.74, 127.63, 129.65, 139.89,
140.79, 146.86, 147.98, 149.03, 149.75, 169.03, 182.69, 183.78.
HRMS-ESI (m/z): calcd for C24H24O8 440.1471; found 439.1471
[Mꢀ H]-. HPLC purity 94.9%, retention time 5.15 min. methyl 2-(4-
((1E,6E)-7-(4-hydroxy-3-(trifluoromethyl)phenyl)-3,5-dioxohepta-
1,6-dien-1-yl)-2-(trifluoromethyl)phenoxy)acetate (J). Yield: 43%.
1H NMR (400 MHz, DMSO‑d6): δ (ppm) 3.72 (–CH3, s, 3H), 5.07 (–CH2,
s, 2H), 6.14 (–CH2, s, 1H), 6.89 (=CH, J = 16.0 Hz, d, 1H), 6.98 (=CH, J
= 16.0 Hz, d, 1H), 7.10 (=CH, J = 8.4 Hz, d, 1H), 7.27 (=CH, J = 8.8 Hz,
d, 1H), 7.62–7.68 (=CH, m, 2H), 7.87 (=CH, J = 8.8 Hz, d, 1H), 7.90
(=CH, s, 1H), 7.99 (=CH, J = 8.7 Hz, d, 1H), 8.02 (=CH, s, 1H). 13C NMR
(125 MHz, DMSO‑d6): δ (ppm) 52.20, 65.29, 101.86, 114.41, 116.26 (JC-
F = 30.1 Hz, q), 117.79, 117.96 (JC-F = 30.5 Hz, q), 122.83, 123.51 (JC-F
= 271.3 Hz, q), 123.92 (JC-F = 272.5 Hz, q), 124.16, 125.83, 127.18,
127.61 (JC-F = 4.1 Hz, q), 128.07, 133.70, 133.96, 138.48, 139.65,
156.78, 157.83, 168.55, 182.63, 183.99. 19F NMR (470 MHz,
DMSO‑d6): δ 60.9, 61.2. HRMS-ESI (m/z): calcd for C24H18F6O6
516.1008; found 515.1008 [Mꢀ H]-. HPLC purity 93.4%, retention time
6.33 min.
7.13 (=CH, J = 8.2, 1.8 Hz, dd, 1H), 7.58 (=CH, J = 15.8 Hz, d, 1H). 13
C
NMR (125 MHz, CDCl3): δ (ppm) 26.59, 55.86, 100.69, 111.39, 115.88,
119.87, 123.09, 126.55, 140.47, 148.19, 149.41, 178.48, 196.85. LRMS-
ESI (m/z): calcd for C13H14O4 234.09; found 235.13 [M + H]+. (E)-6-(4-
hydroxy-3-(trifluoromethyl)phenyl)hex-5-ene-2,4-dione
(4b).
Yield: 23%. 1H NMR (400 MHz, CDCl3): δ (ppm) 2.20 (–CH3, s, 3H), 5.67
(–CH2, s, 1H), 6.41 (=CH, J = 16.0 Hz, d, 1H), 7.01 (=CH, J = 8.4 Hz, d,
1H), 7.57 (=CH, J = 16.0 Hz, d, 1H), 7.61 (=CH, J = 8.4, 1.6 Hz, dd,
1H), 7.70 (=CH, s, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 26.79,
101.24, 101.32, 116.53 (JC-F = 30.1 Hz, q), 117.94, 121.48, 124.15 (JC-F
= 270.9 Hz, q), 125.99, 127.33, 128.24 (JC-F = 5.2 Hz, q), 133.37,
138.59, 157.93, 177.59, 197.78. LRMS-ESI (m/z): calcd for C13H11F3O3
272.07; found 271.03 [Mꢀ H]+.
4.1.7. General procedure to Synthesis compound d and I
4.1.6. General procedure to Synthesis compound B, C, F, and J
Compound D was synthesized using microwave-assisted reaction.
Compound 4a (0.26 mmol) and boric acid (0.26 mmol) was suspended
in DMF. Compound 3-trifluoromethyl benzaldehyde (0.26 mmol) was
added followed by morpholine (0.1 mmol). The mixture was irradiated
in microwave at 150 ◦C for 10 min. The HCl 0.1 N was added to quench
the reaction. Crude product was extracted by ethyl acetate then dried
using MgSO4. The desired compound was purified by column chroma-
tography (silica gel, hexane:ethyl acetate 2:1). (1E,6E)-1-(4-hydroxy-3-
Compound B, C, F, and J was synthesized using microwave-assisted
reaction. Compound 4a or 4b (0.26 mmol) and boric acid (0.26 mmol)
was suspended in DMF. Compound 2a, 2b, and 2c (0.26 mmol) was
added followed by morpholine (0.1 mmol). The mixture was irradiated
in microwave at 150 ◦C for 10 min. The HCl 0.1 N was added to quench
the reaction. Crude product was extracted by ethyl acetate then dried
using MgSO4. The desired compound was purified by column
6

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
methoxyphenyl)-7-[3-(trifluoromethyl) phenyl]hepta-1,6-diene-
3,5-dione (D). Yield: 50%. 1H NMR (400 MHz, CDCl3): 3.98 (–OCH3, s,
3H), 5.88 (–CH2, s, 2H), 6.53 (=CH, J = 15.8 Hz, d, 1H), 6.68 (=CH, J =
15.9 Hz, d, 1H), 6.97 (=CH, J = 8.2, d, 1H), 7.09 (=CH, J = 1.5 Hz, d,
1H), 7.17 (=CH, J = 8.2, 1.6 Hz, dd, 1H), 7.55 (=CH, J = 7.8 Hz, t, 1H),
7.65 (=CH, J = 15.8 Hz, d, 1H), 7.68 (=CH, J = 15.9 Hz, d, 1H), 7.73
(=CH, J = 7.8 Hz, d, 1H), 7.82 (=CH, s, 1H). 13C NMR (125 MHz,
DMSO‑d6): δ (ppm) 55.89, 102.12, 116.02, 121.38, 123.69, 124.26 (JC-F
= 271.1 Hz, q), 125.78 (JC-F = 4.1 Hz, q), 126.29, 126.44 (JC-F = 4.8 Hz,
q), 129.58 (JC-F = 32.1 Hz, q), 129.98, 130.12, 130.27, 131.94, 133.36,
136.25, 137.69, 142.20, 148.31, 150.01, 180.54, 185.96. 19F NMR (470
MHz, DMSO‑d6): δ 61.4. HRMS-ESI (m/z): calcd for C21H17F3O4
390.1079; found 389.1069 [Mꢀ H]-. HPLC purity 98.8%, retention time
6.78 min. (1E,6E)-1-(4-hydroxy-3-(trifluoromethyl)phenyl)-7-(3-
(trifluoromethyl)phenyl)hepta-1,6-diene-3,5-dione (I) Yield: 28%.
1H NMR (400 MHz, CD3CN): δ (ppm) 6.05 (–CH2, s, 2H), 6.81 (=CH, J =
16.0 Hz, d, 1H), 6.96 (=CH, J = 16.0 Hz, d, 1H), 7.09 (=CH, J = 8.8 Hz,
d, 1H), 7.65–7.74 (=CH, m, 3H), 7.78 (=CH, J = 8.4, 2.0 Hz, dd, 1H),
7.88 (=CH, J = 2.1 Hz d, 1H), 7.92 (=CH, J = 7.7 Hz, d, 1H), 7.99 (=CH,
s, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 48.83, 102.33, 116.39
(JC-F = 30.1 Hz, q), 117.89, 122.83, 123.99 (JC-F = 271.3 Hz, q), 124.26
7.67 (=CH, J = 15.8 Hz, d, 1H), 7.85 (=CH, J = 8.8, 1.9 Hz, dd, 1H),
7.96 (=CH, J = 1.9 Hz, d, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm)
52.04, 52.20, 55.89, 65.16, 65.28, 101.71, 111.25, 113.28, 143.39,
117.93 (JC-F = 30.4 Hz, q), 122.42, 122.69, 123.51 (JC-F = 271.0 Hz, q),
124.19, 127.14 (JC-F = 4.9 Hz, q), 128.07, 128.59, 133.91, 138.34,
140.86, 149.29, 156.75, 168.54, 169.15, 182.45, 184.22. 19F NMR (470
MHz, DMSO‑d6): δ 60.9. HRMS-ESI (m/z): calcd for C27H25F3O9
550.1451; found 549.1449 [Mꢀ H]-. HPLC purity 99.8%, retention time
6.52 min.
4.1.10. General procedure to Synthesis compound 5
Compound 1b (1 mmol) and potassium carbonate (1 mmol) were
dissolved in DMF 10 mL. Tert-butyl bromoacetate (1.1 mmol) was added
then was refluxed for 2 h at 80 ◦C. The reaction was quenched by HCl
0.1 N then extracted by ethyl acetate. The organic layer was dried using
MgSO4 then, purified by column chromatography (silica gel, hexane:
ethyl acetate 2:1). Synthesis of (E)-6-(3-(trifluoromethyl)phenyl)
hex-5-ene-2,4-dione. Yield: 97%. 1H NMR (400 MHz, CDCl3): δ (ppm)
1.48 (3x-CH3, s, 9H), 4.75 (–CH2, s, 2H), 7.02 (=CH, J = 8.6 Hz, d, 1H),
8.05 (=CH, J = 8.6, 2.0 Hz, dd, 1H), 8.16 (=CH, J = 1.8 Hz, d, 1H), 9.95
(–CHO, s, 1H). 13C NMR (125 MHz, CDCl3): δ (ppm) 27.46, 65.62, 82.69,
112.74, 119.34 (JC-F = 31.6 Hz, q), 122.71 (JC-F = 271.1 Hz, q), 128.61
(JC-F = 5.0 Hz, q), 134.79, 160.10, 166. 06, 189.36. LRMS-ESI (m/z):
calcd for C14H15F3O4 304.09; found 303.75 [Mꢀ H]+.
(JC-F = 270.7 Hz, q), 124.87, 125.72, 126.42, 127.72, 130.14 (JC-F
=
31.6 Hz, q), 130.17, 132.07, 133.72, 136.17, 138.14, 140.16, 158.12,
181.41, 185.11. 19F NMR (470 MHz, DMSO‑d6): δ 61.32, 61.37. HRMS-
ESI (m/z): calcd for C21H14F6O3 428.0847; found 427.0849 [Mꢀ H]-.
HPLC purity 93.7%, retention time 7.28 min.
4.1.11. General procedure to Synthesis compound K
Compound K was synthesized using microwave-assisted reaction.
Compound 4a (0.26 mmol) and boric acid (0.26 mmol) was suspended
in DMF. Compound 5 (0.26 mmol) was added followed by morpholine
(0.1 mmol). The mixture was irradiated by microwave at 150 ◦C for 10
min. The HCl 0.1 N was added to quench the reaction. Crude product
was extracted by ethyl acetate then dried using MgSO4. The desired
compound was purified by column chromatography (silica gel, hexane:
ethyl acetate 2:1. tert-butyl 2-(4-((1E,6E)-7-(4-hydroxy-3-methox-
yphenyl)-3,5-dioxohepta-1,6-dien-1-yl)-2-(trifluoromethyl)phe-
noxy)acetate (K). Yield: 28%. 1H NMR (400 MHz, CDCl3): δ (ppm) 1.51
(–CH3, s, 9H), 3.99 (–CH3, s, 3H), 4.70 (–CH2, s, 2H), 5.86 (–CH2, S, 2H),
6.53 (=CH, J = 16.0 Hz, d, 1H), 6.57 (=CH, J = 16.0 Hz, d, 1H) 6.92
(=CH, J = 8.6 Hz, d, 1H), 6.99 (=CH, J = 8.0 Hz, d, 1H), 7.11 (=CH, s,
1H), 7.19 (=CH, J = 7.9 Hz, d, 1H), 7.61–7.69 (=CH, m, 3H), 7.85
(=CH, s, 1H). 13C NMR (125 MHz, DMSO‑d6): δ (ppm) 27.82, 55.87,
65.61, 81.98, 101.58, 111.56, 114.18, 115.82, 117.89 (JC-F = 30.5 Hz,
q), 121.26, 123.53 (JC-F = 271.3 Hz, q), 123.56, 124.09, 126.39, 127.11
(JC-F = 5.5 Hz, q), 127.96, 133.74, 141.66, 148.18, 149.71, 167.06,
181.77, 184.84. 19F NMR (470 MHz, DMSO‑d6): δ 60.9. HRMS-ESI (m/
z): calcd for C27H27F3O7 520.1709; found 519.1713 [Mꢀ H]-. HPLC
purity 98.4%, retention time 6.79 min.
4.1.8. General procedure to Synthesis compound E
Compound E was synthesized using microwave-assisted reaction.
Compound 4a (0.26 mmol) and boric acid (0.26 mmol) was suspended
in DMF. Compound 1b (0.26 mmol) was added followed by morpholine
(0.1 mmol). The mixture was irradiated in microwave at 150 ◦C for 10
min. The HCl 0.1 N was added to quench the reaction. Crude product
was extracted by ethyl acetate then dried using MgSO4. The desired
compound was purified by column chromatography (silica gel, hexane:
ethyl acetate 2:1). (1E,6E)-1-(4-hydroxy-3-(trifluoromethyl)phenyl)-
7-(4-hydroxy-3-methoxyphenyl)hepta-1,6-diene-3,5-dione
(E).
Yield: 28%. Yield: 28%. 1H NMR (400 MHz, CD3CN): δ (ppm) 3.94
(–OCH3, s, 3H), 5.96 (–CH2, s, 1H), 6.70 (=CH, J = 15.6 Hz, d, 1H), 6.74
(=CH, J = 15.6 Hz, d, 1H), 6.89 (=CH, J = 8.0 Hz, d, 1H), 7.08 (=CH, J
= 8.4 Hz, d, 1H), 7.17 (=CH, J = 8.4, 1.8 Hz, dd, 1H), 7.28 (=CH, J =
1.8 Hz, d, 1H), 7.61 (=CH, J = 15.6 Hz, d, 2H), 7.75 (=CH, J = 8.6, 1.8
Hz, dd, 1H), 7.86 (=CH, J = 1.6 Hz, d, 1H). 13C NMR (125 MHz,
DMSO‑d6): δ (ppm) 55.85, 101.32, 115.86, 116.19 (JC-F = 30.0 Hz, q),
117.75, 121.22, 121.98 (JC-F = 271.3 Hz, q), 122.88, 123.46, 124.99,
125.90, 126.42, 127.41 (JC-F = 4.7 Hz, q), 133.54, 138.84, 141.35,
148.16, 149.63, 157.66, 182.42, 184.22. 19F NMR (470 MHz,
DMSO‑d6): δ 61.2. HRMS-ESI (m/z): calcd for C21H17F3O5 406.1028;
found 405.1022 [Mꢀ H]-. HPLC purity 94.1%, retention time 5.63 min.
4.2. ThT fluorescence assay
4.1.9. General procedure to Synthesis compound H
Procedure on solubilizing the Aβ was conducted according the
known literature by slightly modification32. To prepare the Aβ monomer
stock, in amount 0.5 mg of the lyophilized Aβ (Peptide Institute) was
dissolved in NaOH 2 mM by gently mixing without vortexing to obtain
500 µM as final concentration. The solution was centrifuged at 13,200
rpm, 4 ◦C, and 10 min. The supernatant was collected and stored in
ꢀ 80 ◦C. The ability of curcumin derivatives to inhibit Aβ42 Aggregation
was determined using a thioflavin T (ThT)-based fluorescence assay.
This assay was conducted in 96-well plate full black non-binding with
frequent 15 s of linear shaking every 5 min and constant heating at 37 ◦C
for three hours. The Thioflavin T (ThT) stock at concentration 200 µM
was freshly prepared in tris glycine 10 mM pH 8.5. For assay system, the
ThT was diluted by PBS 7.4 to reach concentration 20 µM in each well.
Curcumin stock 10 mM was diluted in each well to obtain 10-fold
dilution (1, 0.1, 0.01, 0.001, 0.0001, and 0.00001 µM). The Aβ stock
was transferred to each well to reach final concentration 20 µM. The
Compound H was synthesized using microwave-assisted reaction.
Compound 3a (0.26 mmol) and boric acid (0.26 mmol) was suspended
in DMF. Compound 2b (0.26 mmol) was added followed by morpholine
(0.1 mmol). The mixture was irradiated in microwave at 150 ◦C for 10
min. The HCl 0.1 N was added to quench the reaction. Crude product
was extracted by ethyl acetate then dried using MgSO4. The desired
compound was purified by column chromatography (silica gel, hexane:
ethyl acetate 2:1). Methyl 2-(2-methoxy-4-((1E,6E)-7-(4-(2-methoxy-
2-oxoethoxy)-3-(trifluoromethyl)phenyl)-3,5-dioxohepta-1,6-dien-
1-yl)phenoxy)acetate (H). Yield: 38%. 1H NMR (400 MHz, CD3CN): δ
(ppm) 3.77 (–OCH3, s, 3H), 3.79 (–OCH3, s, 3H), 3.91 (–OCH3, s, 3H),
4.76 (–CH2, s, 2H), 4.89 (–CH2, s, 2H), 6.00 (–CH2, s, 1H), 6.77 (=CH, J
= 15.8 Hz, d, 1H), 6.81 (=CH, J = 15.8 Hz, d, 1H), 6.90 (=CH, J = 8.4
Hz, d, 1H), 7.12 (=CH, J = 8.7 Hz, d, 1H), 7.21 (=CH, J = 8.4, 1.9 Hz,
dd, 1H), 7.32 (=CH, J = 1.9 Hz, d, 1H), 7.62 (=CH, J = 15.8 Hz, d, 1H),
7

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
fluorescence intensity was immediately measured with an excitation of
430 nm and emission 480 nm using microplate reader (Tecan Infinite
F200, Tecan, Switzerland). The data of fluorescence intensity was cor-
rected toward blank containing PBS and ThT while the IC50 values ob-
tained by plotting the intensity at half Vmax of each group toward
concentration.
References
1
Vradenburg G. A pivotal moment in Alzheimer’s disease and dementia: how global
unity of purpose and action can beat the disease by 2025. Expert Rev Neurother. 2015;
15(1):73–82.
2
Cummings J, Aisen PS, DuBois B, et al. Drug development in Alzheimer’s disease: the
path to 2025. Alz Res Therapy. 2016;8(1). https://doi.org/10.1186/s13195-016-
0207-9.
3
4
Schneider LS, Sano M. Current Alzheimer’s disease clinical trials: Methods and
placebo outcomes. Alzheimers Dement. 2009;5(5):388–397.
Gauthier S, Feldman HH, Schneider LS, et al. Efficacy and safety of tau-aggregation
inhibitor therapy in patients with mild or moderate Alzheimer’s disease: a
randomised, controlled, double-blind, parallel-arm, phase 3 trial. Lancet. 2016;388
(10062):2873–2884.
4.3. MTT assay
The N2A and U87 MG cells were suspended in Dulbecco’s Modified
Eagle Medium (DMEM) medium supplemented with 10% fetal bovine
serum. Briefly, N2A cells (8x103 cells/well) and U87 MG (8x103 cells/
well) was grown on 96-well plate overnight at 37 ◦C CO2 5%. For the
single treatment on N2A and U87MG using curcumin derivatives, cells
were treated by serial concentration of curcumin, compound B, D, and K
(1, 5, 10, 25, 50, 100, and 200 µM). To screen the toxicity of Aβ, N2A
cells were treated by Aβ at concentration of 5, 10 and 20 µM. Evaluation
of protective effect was conducted by treating the N2A cells with Aβ 10
µM with or without curcumin derivatives 1 µM. An MTT reduction assay
was conducted as described previously33. Briefly, MTT powder was
dissolved in PBS pH 7.4 to obtain 5 mg/mL concentration stock, then
was diluted into 0.5 mg/mL in DMEM medium. After removal of me-
dium on 96-well plate containing-treated cell, each well was added by
100 µL MTT 0.5 mg/mL and incubated at 37 ◦C CO2 5%. After 3 h in-
5
6
7
8
9
Lee HE, Lim D, Lee JY, Lim SM, Pae AN. Recent tau-targeted clinical strategies for the
treatment of Alzheimer’s disease. Future Med Chem. 2019;11(15):1845–1848.
Scheltens P, Blennow K, Breteler MMB, et al. Alzheimer’s disease. Lancet. 2016;388
(10043):505–517.
Honig LS, Vellas B, Woodward M, et al. Trial of Solanezumab for Mild Dementia Due
to Alzheimer’s Disease. N Engl J Med. 2018;378(4):321–330.
Schneider L. A resurrection of aducanumab for Alzheimer’s disease. Lancet Neurol.
2020;19(2):111–112.
O’Hare E, Scopes DIC, Kim E-M, et al. Orally bioavailable small molecule drug
protects memory in Alzheimer’s disease models. Neurobiol Aging. 2013;34(4):
1116–1125.
10 Nakagami Y, Nishimura S, Murasugi T, et al. A novel β-sheet breaker, RS-0406,
reverses amyloid β-induced cytotoxicity and impairment of long-term potentiation in
vitro. Br J Pharmacol. 2002;137(5):676-682.
11 Ringman JM, Frautschy SA, Teng E, et al. Oral curcumin for Alzheimer’s disease:
tolerability and efficacy in a 24-week randomized, double blind, placebo-controlled
study. Alz Res Therapy. 2012;4(5):43. https://doi.org/10.1186/alzrt146.
12 Reinke AA, Gestwicki JE. Structure–activity Relationships of Amyloid Beta-
aggregation Inhibitors Based on Curcumin: Influence of Linker Length and
Flexibility. Chem Biol Drug Des. 2007;70(3):206–215.
cubation, 100 μL DMSO was added following the absorbance measure-
ment at wavelength of 550 nm using microplate reader (Tecan Infinite
F200, Tecan, Switzerland). Calculation of % cells viability was measured
by dividing the absorbance of untreated cells with the absorbance of
curcumin-treated cells.
13 Ono K, Hasegawa K, Naiki H, Yamada M. Curcumin has potent anti-amyloidogenic
effects for Alzheimer’s β-amyloid fibrils in vitro. J Neurosci Res. 2004;75(6):742–750.
14 Rao PPN, Mohamed T, Teckwani K, Tin G. Curcumin Binding to Beta Amyloid: A
Computational Study. Chem Biol Drug Des. 2015;86(4):813–820.
15 Bairwa K, Grover J, Kania M, Jachak SM. Recent developments in chemistry and
biology of curcumin analogues. RSC Adv. 2014;4(27):13946. https://doi.org/
10.1039/c4ra00227j.
4.4. Negative-Staining TEM imaging
16 Chen M, Du Z-Y, Zheng Xi, Li D-L, Zhou R-P, Zhang K. Use of curcumin in diagnosis,
prevention, and treatment of Alzheimer’s disease. Neural Regen Res. 2018;13(4):742.
https://doi.org/10.4103/1673-5374.230303.
The negative-staining TEM images were obtained according to the
following procedure. Elastic carbon grids (ELS-C10, STEM, Japan) was
made hydrophilic using an ion coater (IB-2, Eiko, Japan) with 3 mA of
plasma current for 40 s before applying sample solution. 5 µM of Aβ only
or in combination with compound B was incubated on PBS pH 7.4 at
37 ◦C for 30 min. Briefly, 5 µL of the Aβ sample was applied to a hy-
drophilic grid and incubated for one minute at RT. After gently dried
with filter paper, the grid sample was washed with Milli-Q water and
dried again with filter paper three times. Finally, the grid was incubated
with 5 µL of 1% Nano-W negative staining solution (NY, USA) for one
minute followed by complete drying using filter paper. The negative
stained sample was observed using TEM H-8100 (Hitachi) operated at
200 kV.
17 Baell JB, Holloway GA. New Substructure Filters for Removal of Pan Assay
Interference Compounds (PAINS) from Screening Libraries and for Their Exclusion in
Bioassays. J Med Chem. 2010;53(7):2719–2740.
18 Nelson KM, Dahlin JL, Bisson J, Graham J, Pauli GF, Walters MA. The Essential
Medicinal Chemistry of Curcumin. J Med Chem. 2017;60(5):1620–1637.
19 Chainoglou E, Hadjipavlou-Litina D. Curcumin in Health and Diseases: Alzheimer’s
Disease and Curcumin Analogues, Derivatives, and Hybrids. Int J Mol Sci. 2020;21
(6):1975.
20 Shabbir U, Rubab M, Tyagi A, Oh D-H. Curcumin and Its Derivatives as Theranostic
Agents in Alzheimer’s Disease: The Implication of Nanotechnology. Int J Mol Sci.
2021;22(1):196.
21 Asawa Y, Yoshimori A, Bajorath J, Nakamura H. Prediction of an MMP-1 inhibitor
activity cliff using the SAR matrix approach and its experimental validation. Sci Rep.
2020;10(1):14710.
22 Wassermann AM, Haebel P, Weskamp N, Bajorath J. SAR Matrices: Automated
Extraction of Information-Rich SAR Tables from Large Compound Data Sets. J Chem
Inf Model. 2012;52(7):1769–1776.
Declaration of Competing Interest
23 Gupta-Ostermann D, Shanmugasundaram V, Bajorath J. Neighborhood-Based
Prediction of Novel Active Compounds from SAR Matrices. J Chem Inf Model. 2014;
54(3):801–809.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
24 Pike CJ, Burdick D, Walencewicz AJ, Glabe CG, Cotman CW. Neurodegeneration
induced by beta-amyloid peptides in vitro: the role of peptide assembly state.
J Neurosci. 1993;13(4):1676–1687.
25 Sulatskaya AI, Sulatsky MI, Antifeeva IA, Kuznetsova IM, Turoverov KK. Structural
Analogue of Thioflavin T, DMASEBT, as a Tool for Amyloid Fibrils Study. Anal Chem.
2019;91(4):3131–3140.
Acknowledgments
26 Um JW, Nygaard HB, Heiss JK, et al. Alzheimer amyloid-β oligomer bound to
postsynaptic prion protein activates Fyn to impair neurons. Nat Neurosci. 2012;15(9):
1227–1235.
We thank Prof. Kenjiro Ono (Showa University Medical School),
Prof. Mayumi Tsuji (Showa University Pharmacological Research Cen-
ter), and Prof. Motomu Kanai (Graduate School of Pharmaceutical Sci-
ences, The University of Tokyo) for their valuable suggestions and Prof.
Yoshitaka Kitamoto (Tokyo Institute of Technology) for the technical
support with TEM measurements. H. S. Ban thanks Five-star Alliance
overseas researcher invitation program in “NJRC Mater. & Dev.”.
27 Bettayeb K, Oumata N, Zhang Y, et al. Small-molecule inducers of Aβ-42 peptide
production share a common mechanism of action. FASEB J. 2012;26(12):5115–5123.
28 Li Y, Cheng D, Cheng R, et al. Mechanisms of U87 Astrocytoma Cell Uptake and
Trafficking of Monomeric versus Protofibril Alzheimer’s Disease Amyloid-β Proteins.
Ferreira ST, ed. PLoS ONE. 2014;9(6):e99939.
29 Sidhar H, Giri RK. Induction of Bex genes by curcumin is associated with apoptosis
and activation of p53 in N2a neuroblastoma cells. Sci Rep. 2017;7(1):41420.
30 Yang T, Zhu Z, Yin E, et al. Alleviation of symptoms of Alzheimer’s disease by
diminishing Aβ neurotoxicity and neuroinflammation. Chem Sci. 2019;10(43):
10149–10158.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmc.2021.116357.
8

R. Yudi Utomo et al.
Bioorganic & Medicinal Chemistry 46 (2021) 116357
31 Xiao Y, Matsuda I, Inoue M, Sasahara T, Hoshi M, Ishii Y. NMR-based site-resolved
profiling of β-amyloid misfolding reveals structural transitions from pathologically
relevant spherical oligomer to fibril. J Biol Chem. 2020;295(2):458–467.
32 Mohamed T, Shakeri A, Tin G, Rao PPN. Structure-Activity Relationship Studies of
Isomeric 2,4-Diaminoquinazolines on β-Amyloid Aggregation Kinetics. ACS Med
Chem Lett. 2016;7(5):502–507.
33 Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to
proliferation and cytotoxicity assays. J Immunol Methods. 1983;65(1-2):55–63.
9