J Nat Med

DOI 10.1007/s11418-017-1097-2

ORIGINAL PAPER

Semisynthesis and biological evaluation of prenylated resveratrol

derivatives as multi-targeted agents for Alzheimer’s disease
Thanchanok Puksasook1 • Shinya Kimura2 • Sarin Tadtong3 •
Jutamas Jiaranaikulwanitch4 • Jaturong Pratuangdejkul5 •
Worawan Kitphati6 • Khanit Suwanborirux7 • Naoki Saito2 •
Veena Nukoolkarn1

Received: 27 January 2017 / Accepted: 26 May 2017

 The Japanese Society of Pharmacognosy and Springer Japan 2017

Abstract A series of prenylated resveratrol derivatives Moreover, this compound showed no neurotoxicity along
were designed, semisynthesized and biologically evaluated with a greater ability to inhibit oxidative stress on P19-
for inhibition of b-secretase (BACE1) and amyloid-b (Ab) derived neuronal cells (50.59% cell viability at 1 nM). The
aggregation as well as free radical scavenging and neuro- neuritogenic activity presented more branching numbers
protective and neuritogenic activities, as potential novel (9.33) and longer neurites (109.74 lm) than the control,
multifunctional agents against Alzheimer’s disease (AD). and was comparable to the quercetin positive control.
The results showed that compound 4b exhibited good anti- Taken together, these results suggest compound 4b had the
Ab aggregation (IC50 = 4.78 lM) and antioxidant activity greatest multifunctional activities and might be a very
(IC50 = 41.22 lM) and moderate anti-BACE1 inhibitory promising lead compound for the further development of
activity (23.70% at 50 lM), and could be a lead compound. drugs for AD.

Electronic supplementary material The online version of this

article (doi:10.1007/s11418-017-1097-2) contains supplementary
material, which is available to authorized users.

& Veena Nukoolkarn

[email protected];
[email protected]
1
Department of Pharmacognosy, Faculty of Pharmacy,
Mahidol University, Bangkok 10400, Thailand
2
Graduate School of Pharmaceutical Sciences, Meiji
Pharmaceutical University, Tokyo 204-8588, Japan
3
Department of Pharmacognosy, Faculty of Pharmacy,
Srinakharinwirot University, Nakhon Nayok 26120, Thailand
4
Department of Pharmaceutical Sciences, Faculty of
Pharmacy, Chiang Mai University, Chiang Mai 50200,
Thailand
5
Department of Microbiology, Faculty of Pharmacy, Mahidol
University, Bangkok 10400, Thailand
6
Department of Physiology, Faculty of Pharmacy, Mahidol
University, Bangkok 10400, Thailand
7
Department of Pharmacognosy and Pharmaceutical Botany,
Faculty of Pharmaceutical Sciences, Center for Bioactive
Natural Products from Marine Organisms and Endophytic
Fungi (BNPME), Chulalongkorn University, Bangkok 10330,
Thailand

123
 J Nat Med

Graphical Abstract
Semisynthesis and biological evaluation of prenylated resveratrol derivatives as
multi-targeted agents for Alzheimer’s disease

Thanchanok Puksasook, Shinya Kimura, Sarin Tadtong,

Jutamas Jiaranaikulwanitch, Jaturong Pratuangdejkul,
Worawan Kitphati, Khanit Suwanborirux, Naoki Saito and Veena Nukoolkarn

Keywords Alzheimer’s disease (AD)  Prenylated strategies which consist of (i) developing acetyl-
resveratrol derivatives  b-Secretase (BACE1)  Amyloid-b cholinesterase inhibitors, and (ii) designing ACh receptor
(Ab) aggregation  Neuroprotective  Neuritogencity agonists to balance the level of ACh. The amyloid
hypothesis, which has two tactics, has been employed to
relieve Ab aggregation [6]. The first tactic is to design b-
Introduction secretase (BACE1) or c-secretase inhibitors with non-nat-
ural amino acids or transition metal complexes to block the
Alzheimer’s disease (AD) is the most common form of formation of amyloid plaques [7]. The second tactic is the
dementia in the elderly population and has become the use of antibodies or catalytic antibodies to bind and/or
fourth leading cause of death in developed countries. degrade Ab [8]. In addition, suppressing the abnormal post-
Currently, AD is increasing in people[65 years of age and translational modifications of the tau protein has been
affects [37 million people worldwide. The size of the performed using various methods, such as designing
affected population is expected to increase to 42.3 million antagonists to inhibit serine/threonine kinases or agonists
in 2020 and 81.1 million in 2040 if no efficient treatment is to promote the dephosphorylation of the hyper-phospho-
discovered. It has become one of the costliest diseases, rylated tau protein [9, 10].
bringing a heavy social and financial burden to both society However, no effective cure is applied in clinics except
and families [1, 2]. A highly complex and rapidly pro- for symptom-relieving strategies. The current clinical
gressive disease, AD manifests in the cholinergic branch of therapy for AD patients are mainly based on the cholin-
the central nervous system (CNS). The key histopatho- ergic hypothesis, which suggests that a decline in ACh
logical hallmarks of AD are (i) the generation of extra- levels in specific brain regions leads to cognitive and
cellular insoluble senile plaques composed of b-amyloid memory deficits [11], and so focuses on sustaining or
(Ab) peptide, (ii) intracellular neurofibrillary tangles recovering cholinergic functions. Supporting this notion,
composed of hyperphosphorylated tau proteins, and (iii) four cholinesterase inhibitors—tacrine, rivastigmine,
decreased levels of the neurotransmitter acetylcholine galanthamine and donepezil—have been approved for
(ACh) [3]. During the past two decades, extensive research clinical use [4]. Due to the complex nature of AD, these
on reducing or clearing the related AD pathological phe- drugs can only reverse the symptoms for a short period of
nomena has been attempted using different strategies. The time without preventing progressive neurodegeneration.
cholinergic hypothesis represents one of the classical Thus, a more appropriate approach, termed the multi-target
hypotheses associated with AD [4, 5]. There are two major directed ligands (MTDLs) strategy, has been proposed to

123
J Nat Med

confront this disease [12, 13]. This strategy aims for a neurodegenerative injuries and disorders. Thus, drugs that
single compound that can simultaneously modulate dif- specifically show all of these effects could be useful for
ferent targets involved in the neurodegenerative AD cas- either the prevention or the treatment of AD [30]. The
cade. Searching for such candidate molecules that act on MTDLs proposed compounds were designed using
multiple sites of the pathologic cascade has become a new resveratrol as a core structure due to its antioxidant [31],
strategy in the design of new drugs for AD. anti-Ab aggregation [32, 33] and neuroprotective activi-
Several active compounds originate from herbal plants, ties [34]. The resveratrol scaffold (Group C) was com-
suggesting that herbs could potentially serve as sources of bined with prenyl or geranyl moiety, the natural active
novel multi-targeted drugs for AD [14–17]. Resveratrol is moiety of BACE1 inhibitors (Group A) and neurito-
one natural compound with a stilbene structure that is genicity agents (Group B). A recent study reported that b-
widely presented in foods and medicinal plants and well secretase inhibition activity of the natural flavone from the
known as exerting a variety of promising biological stem bark of Morus ihou, named kuwanon C, was 20-fold
activities. Recently, several studies have shown that more effective than its parent compound, noratocarpetin.
resveratrol can counteract Ab toxicity with its antioxidant This enhanced activity was related to the presence of the
properties in cellular models, and can inhibit Ab aggre- prenyl side chain [35]. In addition, 5-geranyloxy-8-
gation in animal models [18, 19]. Moreover, resveratrol is methoxypsoralen was shown to be more effective than
also effective in prevention of neurodegenerative diseases unsubstituted analogs with the geranyl side chain [36].
and its phase II and III clinical trials for AD patients are Moreover, prenylated xanthone derivatives which were
ongoing at the present [20]. Resveratrol has shown sig- extracted from the wood of Garcinia xanthochymus pro-
nificant potential in promising activities related to AD, moted neuritogenicity in PC12D cells [37]. However,
such as a-secretase promotion [21], and anti-inflammatory artepillin C, a major component of Brazilian propolis,
and neuroprotective activity [22]. Indeed, over the past exhibited a higher neuritogenic level than its derivative
few years, evidence from high-quality studies concluded without alkyl moieties on the aromatic ring. It was
that resveratrol was found to be a safe and well-tolerated approximately seven-fold greater than that induced by
substance in populations at risk for AD and other neu- nerve growth factor alone in PC12m3 cells [38]. There-
rodegenerative diseases [23, 24]. These results indicate fore, the alkyl side chain was selected as an active moiety.
that resveratrol may be suitable as a starting compound in In the present study, two series of resveratrol derivatives
the design of a multifunctional drug for the treatment of were prepared. The C-alkyl resveratrol derivatives were
AD. Therefore, we are interested in the discovery and designed to increase the hydrophobic or p–p interactions
development of a new series of resveratrol derivatives that of MTDLs for the potential treatment of AD. In addition,
could be employed as effective multi-target-directed the O-alkyl resveratrol derivatives were also semisynthe-
agents for AD as shown in Fig. 1. We focused our sized to examine the importance of the –OH group on the
attention on multifunctional activities including inhibition resveratrol scaffold for multifunctional activities of AD.
of Ab aggregation and b-secretase, and antioxidant, neu-
roprotective and neuritogenic activities for AD, due to
increasing evidence in support of these effects amelio- Experimental
rating AD symptoms in AD model animals and AD
patients. According to available reports, strategies Chemistry
involving the inhibition of b-secretase (BACE1) and Ab
1
aggregation in an APP transgenic mouse model improved H-NMR and 13C-NMR spectra were recorded in CDCl3 or
memory deficits and prevented cholinergic dysfunction dimethyl sulfoxide (DMSO)-d6 using a Bruker Avance
[25, 26]. The free radical and oxidative stress theory of DPX-300 FT-NMR spectrometer. Mass spectra (MS) were
aging also suggests that oxidative damage is an important recorded on either a Thermo Finnigan or LCMS Bruker
player in neuronal degeneration. There is evidence to MicroTof. Infrared (IR) spectra were recorded on a Nicolet
support the fact that antioxidants could be able to atten- FTIR 550. Melting points were recorded on electrothermal
uate the AD syndrome, and prevent progression of the IA9000 digital melting-point apparatus. Resveratrol was
disease [27]. In addition, neuroprotection refers to mech- extracted from the stem and root of Polygonum cuspidatum
anisms within the nervous system which relatively con- and purchased from Xi’an Huarui Bio-Engineering Co.
serve the neuronal structure and/or function [28, 29]. Ltd. All the chemical reactions were monitored by thin-
Neuritogenic activity is one of the focuses of the study layer chromatography on silica gel GF254 plates from E.
looking at the preventive and therapeutic effects on AD Merck (Germany). Where appropriate, crude products were
due to the regeneration of neurons. This represents a purified by column chromatography using silica gel
promising strategy for drugs targeted against (230–400 mesh) purchased from E. Merck (Germany).

123
 J Nat Med

Fig. 1 Designed strategy for alkylated resveratrol derivatives

Preparation of compounds 2a–d

Methylated resveratrol derivatives were synthesized by (E)-3,5,40 -Trimethoxystilbene (2a): colorless crystal;
alkylation using a mixture of resveratrol (500 mg, m.p. 57–59 C; IR (KBr)/cm: 2990, 2931, 2832, 1588,
2.19 mmol), anhydrous potassium carbonate (605.3 mg, 1509, 1458, 1424, 1345, 1326, 1315, 1248, 1208, 1192,
2.0 eq, 4.38 mmol) and methyl iodide (0.41 mL, 3.0 eq, 1152, 1070, 1030, 960, 839, 830, 772, 682. 1H-NMR
6.57 mmol) in dry acetone (10 mL). The reaction mix- (CDCl3) d:7.46 (2H, d, J = 8.8 Hz, H-20 and H-60 ), 7.06
ture was refluxed and stirred for 3 h under a nitrogen (1H, d, J = 16.3 Hz, H-a), 6.92 (1H, d, J = 16.3 Hz,
atmosphere. After complete reaction, the mixture was H-b), 6.91 (2H, d, J = 8.8 Hz, H-30 and H-50 ), 6.66 (2H, d,
diluted with water and extracted with ethyl acetate J = 2.2 Hz, H-2 and H-6), 6.39 (1H, t, J = 2.2 Hz, H-4),
(15 mL) three times. The combined extracts were 3.83 (9H, s, 3x-OCH3). 13C-NMR (CDCl3) d: 161.1 (s),
washed with brine and dried with anhydrous sodium 159.5 (s), 139.8 (s), 130.1 (s), 128.9 (d), 127.9 (d), 126.7
carbonate (Na2CO3). After removal of the solvent, the (d), 114.3 (d), 104.5 (d), 104.4 (d), 99.7 (d), 55.5 (q).
crude products were purified by silica gel column HRESIMS m/z: 271.1331 ([M ? H]?, calcd for C17H19O3:
chromatography with MeOH/CH2Cl2 (0.5:9.5) to obtain 271.1334).
2a (21.4 mg, 4%), 2b (84.7 mg, 15%), 2c (109.4 mg, (E)-3,40 -Dimethoxy-5-hydroxystilbene (2b): colorless
21%) and 2d (139.0 mg, 26%) [39, 40]. crystal; m.p. 115–117 C; IR (KBr)/cm: 3357, 2929, 2837,

123
J Nat Med

1594, 1513, 1456, 1303, 1265, 1179, 1157, 1013, 992, 967, concentration of the crude products, column chromato-
830. 1H-NMR (CDCl3) d: 7.44 (2H, d, J = 8.7 Hz, H-20 graphic purification with MeOH/CH2Cl2 (0.5:9.5) gave 3a
and H-60 ), 7.02 (1H, d, J = 16.2 Hz, H-a), 6.90 (2H, d, (23.1 mg, 3%), 3b (86.4 mg, 14%), 3c (134.8 mg, 24%)
J = 8.7 Hz, H-30 and H-50 ), 6.87 (1H, d, J = 16.2 Hz, and 3d (55.6 mg, 10%).
H-b), 6.62 (1H, t, J = 1.9 Hz, H-6), 6.58 (1H, t, (E)-3,5,40 -Triethoxystilbene (3a) [41]: colorless crystal;
J = 1.9 Hz, H-2), 6.32 (1H, t, J = 1.9 Hz, H-4), 3.83 (3H, m.p. 75–77 C; IR (KBr)/cm: 2979, 2928, 1592, 1511,
s, –OCH3), 3.82 (3H, s, –OCH3). 13C-NMR (CDCl3) d: 1444, 1391, 1297, 1250, 1172, 1115, 1057, 961. 1H-NMR
161.2 (s), 159.6 (s), 157.0 (s), 140.3 (s), 129.1 (s), 128.2 (CDCl3) d: 7.43 (2H, d, J = 8.7 Hz, H-20 and H-60 ), 7.03
(d), 127.9 (d), 126.4 (d), 114.3 (d), 105.8 (d), 104.8 (d), (1H, d, J = 16.3 Hz, H-a), 6.89 (2H, d, J = 8.7 Hz, H-30
100.8 (d), 55.5 (q). HRESIMS m/z: 257.1173 ([M ? H]?, and H-50 ), 6.89 (1H, d, J = 16.2 Hz, H-b), 6.64 (2H, d,
calcd for C16H17O3: 257.1178). J = 2.19 Hz, H-2 and H-6), 6.37 (1H, t, J = 2.19 Hz,
(E)-3-Methoxy-40 ,5-hydroxystilbene (2c): colorless H-4), 4.05 (6H, q, J = 6.96 Hz, 3x-CH2), 1.43 (9H, t,
crystal; m.p. 117–118 C; IR (KBr) cm-1: 3384, 2926, J = 6.96 Hz, 3x-CH3). 13C-NMR (CDCl3) d: 160.4 (s),
2852, 1591, 1511, 1457, 1344, 1301, 1250, 1150, 1058, 158.9 (s), 139.8 (s), 130.0 (s), 128.8 (d), 127.9 (d), 126.7
1031, 961, 833, 681. 1H-NMR (DMSO-d6) d: 9.57 (1H, brs, (d), 114.8 (d), 105.0 (d), 100.6 (d), 63.6 (t), 15.0 (q).
–OH), 9.43 (1H, brs, –OH), 7.41 (2H, d, J = 8.6 Hz, H-20 HRESIMS m/z: 313.1811 ([M ? H]?, calcd for C20H25O3:
and H-60 ), 7.04 (1H, d, J = 16.4 Hz, H-a), 6.87 (1H, d, 313.1804).
J = 16.4 Hz, H-b), 6.76 (2H, d, J = 8.6 Hz, H-30 and (E)-3,40 -Diethoxy-5-hydroxystilbene (3b): colorless
H-50 ), 6.22 (1H, t, J = 2.1 Hz, H-4), 6.57 (1H, brs, H-2), crystal; m.p. 92–94 C; IR (KBr) cm-1: 3385, 2928, 2855,
6.53 (1H, brs, H-6), 3.72 (3H, s, –OCH3); 13C-NMR 1697, 1601, 1511, 1455, 1249, 1171, 1115, 1052, 838. 1H-
(DMSO-d6) d: 160.7 (s), 158.6 (s), 157.3 (s), 139.5 (s), NMR (CDCl3) d: 7.42 (2H, d, J = 8.6 Hz, H-20 and H-60 ),
128.5 (s), 128.0 (d), 125.4 (d), 115.6 (d), 105.8 (d), 102.6 7.00 (1H, d, J = 16.3 Hz, H-a), 6.88 (2H, d, J = 8.6 Hz,
(d), 100.4 (d), 55.0 (q). HRESIMS m/z: 243.1026 H-30 and H-50 ), 6.85 (1H, d, J = 16.3 Hz, H-b), 6.62 (1H,
([M ? H]?, calcd for C15H15O3: 243.1021). t, J = 1.74 Hz, H-2), 6.56 (1H, t, J = 1.74, 1.44 Hz, H-6),
(E)-40 -Methoxy-3,5-dihydroxystilbene (2d): colorless 6.30 (1H, t, J = 1.74, H-4). 4.05 (4H, dq, J = 4.5 Hz, 2x-
crystal; m.p. 176–178 C; IR (KBr)/cm: 3384, 2924, 2854, CH2), 1.42 (6H, td, J = 2.64, 6.96 Hz, 2x-CH3) 13C-NMR
1735, 1599, 1512, 1461, 1258, 1031, 801. 1H-NMR (CDCl3) d: 160.6 (s), 158.9 (s), 156.9 (s), 140.2 (s), 129.9
(DMSO-d6) d: 9.23 (2H, s, 2x-OH), 7.51 (2H, d, (s or d), 128.0 (d), 126.3 (d), 114.8 (s), 105.7 (d), 105.5 (d),
J = 8.8 Hz, H-20 and H-60 ), 6.99 (1H, d, J = 16.4 Hz, 101.2 (d), 63.7 (t), 15.0 (q). HRESIMS m/z: 285.1487
H-a), 6.93 (2H, d, J = 8.8 Hz, H-30 and H-50 ), 6.89 (1H, d, ([M ? H]?, calcd for C18H21O3: 285.1491).
J = 16.4 Hz, H-b), 6.41 (2H, d, J = 2.1 Hz, H-2 and H-6), (E)-3-Ethoxy-40 ,5-hydroxystilbene (3c): colorless crys-
6.14 (1H, t, J = 2.1 Hz, H-4). 3.76 (3H, s, –OCH3) 13C- tal; m.p. 89–91 C; IR (KBr)/cm: 3370, 2980, 1591, 1512,
NMR (DMSO-d6) d: 158.9 (s), 158.6 (s), 139.1 (s), 129.7 1446, 1342, 1236, 1168, 1054, 1014, 962, 837. 1H-NMR
(s), 127.8 (d), 127.5 (d), 126.7 (d), 114.2 (d), 104.5 (d), (DMSO-d6) d: 9.58 (1H, s, –OH), 9.39 (1H, s, –OH), 7.40
102.0 (d), 55.2 (q). HRESIMS m/z: 243.1028 ([M ? H]?, (2H, d, J = 8.6 Hz, H-20 and H-60 ), 7.03 (1H, d,
calcd for C15H15O3: 243.1021). J = 16.4 Hz, H-a), 6.86 (1H, d, J = 16.4 Hz, H-b), 6.76
(2H, d, J = 8.6 Hz, H-30 and H-50 ), 6.55 (1H, s, H-2), 6.50
Preparation of compounds 3a–d (1H, s, H-6), 6.19 (1H, t, J = 2.1 Hz, H-4), 3.96 (2H, q,
J = 7.0 Hz, –CH2), 1.29 (3H, t, J = 7.0 Hz, –CH3). 13C-

Compounds 3a–d were prepared according to the method NMR (DMSO-d6) d: 159.9 (s), 158.6 (s), 157.3 (s), 139.4
of compounds 2a–d, except for using 1-bromoethane as a (s), 128.4 (s), 128.0 (d), 127.9 (d), 125.4 (d), 115.5 (d),
reagent (0.49 mL, 2.0 eq, 6.57 mmol). After filtration and

123
 J Nat Med

105.7 (d), 103.1 (d), 100.8 (d), 62.8 (t), 14.8 (q). HRESIMS H-a), 6.80 (1H, d, J = 16.1 Hz, H-b), 6.77 (2H, d,
m/z: 257.1182 ([M ? H]?, calcd for C16H17O3: 257.1178). J = 8.6 Hz, H-30 and H-50 ), 6.47 (1H, d, J = 2.1 Hz, H-6),
(E)-40 -Ethoxy-3,5-dihydroxystilbene (3d): colorless 6.23 (1H, d, J = 2.1 Hz, H-4), 5.03 (1H, t, J = 6.6 Hz,
crystal; m.p. 169–171 C; IR (KBr)/cm: 3348, 2981, 2931, H-200 ), 3.29 (2H, d, J = 6.5 Hz, H-100 ), 1.74 (3H, s, H-500 ),
1591, 1512, 1477, 1445, 1351, 1252, 1160, 1116, 1042, 1.60 (3H, s, H-400 ). 13C-NMR (DMSO-d6) d: 157.6 (s),
1008, 964, 836. 1H-NMR (DMSO-d6) d: 9.22 (2H, s, 2x- 156.2 (s), 156.1 (s), 137.9 (s), 129.5 (s), 129.4 (d or s),
OH), 7.47 (2H, d, J = 8.7 Hz, H-20 and H-60 ), 6.98 (1H, d, 128.2 (d), 124.9 (d), 123.9 (d), 117.1 (s), 116.0 (d), 103.2
J = 16.4 Hz, H-a), 6.90 (2H, d, J = 8.7 Hz, H-30 and (d), 102.2 (d), 26.0 (s), 24.3 (t), 18.3 (s). HRESIMS m/z:
H-50 ), 6.88 (1H, d, J = 16.4 Hz, H-b), 6.40 (2H, d, 295.1349 ([M-H]-, calcd for C19H19O3: 295.1334).
J = 2.1 Hz, H-2 and H-6), 6.12 (1H, t, J = 2.1 Hz, H-4), (E)-3,5,40 -Trihydroxy-4-prenylstilbene (4b) [40, 43]:
4.03 (2H, q, J = 7.0 Hz, –CH2), 1.34 (3H, t, J = 7.0 Hz, – pale yellow powder; m.p. 157–159 C; IR (KBr) cm-1:
CH3). 13C-NMR (DMSO-d6) d: 158.5 (s), 158.2 (s), 139.1 3547, 3404, 1607, 1582, 1512, 1425, 1238, 1043, 966, 833.
1
(s), 129.5 (s), 127.8 (d), 127.5 (d), 126.6 (d), 114.6 (d), H-NMR (CD3OD) d: 7.30 (2H, d, J = 8.4 Hz, H-20 and
104.4 (d), 101.9 (d), 63.1 (t), 14.7 (q). HRESIMS m/z: H-60 ), 6.88 (1H, d, J = 16.4 Hz, H-a), 6.74 (1H, d,
257.1178 ([M ? H]?, calcd for C16H17O3: 257.1178). J = 8.4 Hz, H-30 and H-50 ), 6.72 (1H, d, J = 16.4 Hz,
H-b), 6.44 (2H, s, H-2 and H-6), 5.22 (1H, m, H-200 ), 3.26
(2H, d, J = 7.2 Hz, H-100 ), 1.74 (3H, s, H-500 ), 1.63 (3H, s,
H-400 ). 13C-NMR (CD3OD) d: 158.1 (s), 157.2 (s), 137.6
Preparation of compounds 4a–c

4a 4b 4c 4d

Dry lithium carbonate (325.1 mg, 2.0 eq, 4.92 mmol) was (s), 131.1 (s), 130.6 (s), 128.6 (d), 128.2 (d), 127.2 (d),
added to a solution of resveratrol (500 mg, 2.19 mmol) in 124.6 (d), 116.4 (d), 115.8 (d), 105.6 (d), 26.0 (q), 23.3 (t),
dry dimethylformamide (DMF, 30 mL). 3,3-Dimethyl 17.9 (q). HRESIMS m/z: 295.1333 ([M-H]-, calcd for
allylbromide (0.76 mL, 3.0 eq, 3.68 mmol) was added C19H19O3: 295.1334).
dropwise. The reaction mixture was refluxed and stirred (E)-3,5,40 -Trihydroxy-2,6-prenylstilbene (4c) [44]: pale
under a nitrogen atmosphere. After 3 h, the reaction mix- yellow oil; IR (CH2Cl2) cm-1: 3386, 2968, 2921, 1592,
ture was filtered. The filtrate was diluted with water and 1512, 1445, 1374, 1258, 1169, 1084, 1038, 832. 1H-NMR
extracted with ethyl acetate (3 9 30 mL). The combined (DMSO-d6) d: 9.50 (s, –OH), 8.89 (s, 2x-OH), 7.28 (2H, d,
organic extracts were dried (Na2SO4), and the solvent was J = 8.6 Hz, H-20 and H-60 ), 6.84 (1H, d, J = 16.6 Hz,
removed under reduced pressure. The crude material was H-a), 6.75 (2H, d, J = 8.6 Hz, H-30 and H-50 ), 6.29 (1H, d,
purified by silica gel column chromatography eluting with J = 16.6 Hz, H-b), 6.31 (1H, s, H-4), 5.06 (2H, t,
gradient CH2Cl2/hexane (1:1 to 20:1) to afford 4a J = 6.8 Hz, H-200 and H-2¢¢¢), 3.16 (4H, d, J = 6.5 Hz,
(402.3 mg, 62%), 4b (16.4 mg, 3%) and 4c (39.7 mg, 5%). H-100 and H-1¢¢¢), 1.59 (3H, s, H-500 ), 1.53 (3H, s, H-400 ).
13
(E)-3,5,40 -Trihydroxy-2-prenylstilbene (4a) [42]: color- C-NMR (DMSO-d6) d: 157.0 (s), 153.3 (s), 138.7 (s),
less needle crystal; m.p. 156-158 C; IR (KBr) cm-1: 3332, 132.7 (d), 128.6 (s), 127.4 (d), 124.9 (d), 123.9 (d), 116.5
2972, 2926, 1603, 1513, 1447, 1345, 1304, 1243, 1170, (d), 115.5 (d), 101.3 (d), 25.8 (q), 25.6 (q), 17.9 (d).
1138, 1009, 961, 836. 1H-NMR (DMSO-d6) d: 9.56 (1H, s, HRESIMS m/z: 363.1980 ([M-H] -, calcd for C24H27O3:
–OH), 9.14 (1H, s, –OH), 8.98 (1H, s, –OH), 7.35 (2H, d, 363.1960).
J = 8.6 Hz, H-20 and H-60 ), 7.07 (1H, d, J = 16.1 Hz,

123
J Nat Med

Preparation of compounds 5a–c

5a 5b 5c

Compounds 5a–c were prepared according to the method (DMSO-d6) d: 9.49 (1H, s, –OH), 8.89 (2H, s, 2xOH), 7.25
of compounds 4a–c, except for using geranyl bromide (2H, d, J = 8.0 Hz, H-20 and H-60 ), 6.83 (1H, d,
(1.26 mL, 3.0 eq, 3.68 mmol) as a reagent. After filtration J = 16.6 Hz, H-a), 6.74 (2H, d, J = 8.0 Hz, H-30 and
and concentration, the crude products, purified by silica gel H-50 ), 6.29 (1H, d, J = 16.6 Hz, H-b), 6.33 (1H, s, H-4),
column chromatography eluting with gradient CH2Cl2/ 5.05 (4H, m, H-200 , H-2¢¢¢, H-600 and H-6¢¢¢), 3.18 (4H, d,
hexane (1:1 to 20:1) furnished 5a (510.5 mg, 64%), 5b J = 5.8 Hz, H-100 and H-1¢¢¢), 1.97 (4H, brt, J = 6.1 Hz,
(26.7 mg, 3%), and 5c (58.0 mg, 5%). H-500 and H-5¢¢¢), 1.91 (4H, brd, J = 6.8 Hz, H-400 and
(E)-3,5,40 -Trihydroxy-2-geranylstilbene (5a) [45]: light H-4¢¢¢), 1.59 (6H, s, H-900 and H-9¢¢¢), 1.53 (12H, brs, H-800 ,
brown oil; IR (CH2Cl2) cm-1: 3385, 2923, 2853, 1599, H-8¢¢¢, H-1000 and H-10¢¢¢). 13C-NMR (DMSO-d6) d: 157.0
1514, 1453, 1357, 1263, 1172, 835. 1H-NMR (DMSO-d6) (s), 153.3 (s), 138.8 (s), 138.9 (s), 132.6 (s), 132.5 (s),
d: 9.57 (1H, s, –OH), 9.16 (1H, s, –OH), 9.01 (1H, s, –OH), 132.2 (d), 127.2 (d), 124.9 (d), 124.2 (s or d), 116.5 (s),
7.34 (2H, d, J = 8.4 Hz, H-20 and H-60 ), 7.05 (1H, d, 115.4 (d), 101.2 (d), 39.5 (t), 26.2 (t), 25.6 (q), 25.5 (t),
J = 16.1 Hz, H-a), 6.80 (1H, d, J = 16.1 Hz, H-b), 6.77 17.5 (q), 15.9 (q). HRESIMS m/z: 501.3367 ([M ? H]?,
(2H, d, J = 8.4 Hz, H-30 and H-50 ), 6.50 (1H, d, calcd for C34H45O3: 501.3369).
J = 1.8 Hz, H-6), 6.26 (1H, d, J = 1.77 Hz, H-4), 5.02
(2H, m, H-200 and H-600 ), 3.31 (2H, d, J = 6.1 Hz, H-100 ), Preparation of (E)-3,5,40 -
1.94 (2H, m, H-400 and H-500 ), 1.74 (3H, s, H-900 ), 1.54 (3H, tri(methoxymethoxy)stilbene (6)
s, H-800 ), 1.47 (3H, s, H-1000 ). 13C-NMR (DMSO-d6) d:
157.2 (s), 155.9 (s), 155.8 (s), 137.8 (s), 137.5 (s), 130.7 Resveratrol (1) (92.2 mg, 0.4 mmol) was dissolved in
(s), 129.0 (d), 127.8 (d), 124.6 (d), 124.2 (d or s), 116.8 (s), DMF (20 mL) at 0 C. Then, NaH (96.0 mg, 6.0 eq,
115.7 (s or t), 102.9 (d), 101.8 (d), 39.5 (t), 26.3 (t), 23.8 2.4 mmol) and methoxymethyl chloride (290.0 mg, 9.0 eq,
(t), 16.1 (q), 25.5 (q), 17.6 (q). HRESIMS m/z: 365.2113 3.6 mmol) were added, respectively. The reaction mixture
([M ? H]?, calcd for C24H29O3: 365.2117). was stirred at 0 C for 2 h. After complete reaction, water
(E)-3,5,40 -Trihydroxy-4-geranylstilbene (5b) [42]: light (10 mL) was added to the reaction mixture at 0 C and
brown oil; IR (CH2Cl2) cm-1: 3383, 2965, 2926, 1677, 1605, extracted with EtOAc (3 9 10 mL). The extract was
1513, 1438, 1377, 1236, 1171, 1043, 963, 837. 1H-NMR washed with saturated NH4Cl solution (10 mL) and H2O
(DMSO-d6) d: 9.55 (1H, s, –OH), 9.11 (2H, s, 2x-OH), 7.38 (10 mL), dried (Na2SO4), and evaporated under reduced
(2H, d, J = 8.1 Hz, H-20 and H-60 ), 6.78 (2H, s, H-a and H-b), pressure. After removal of the solvent, the products were
6.75 (2H, d, J = 8.1 Hz, H-30 and H-50 ), 6.44 (2H, s, H-2 and purified by flash column chromatography using hexane/
H-6), 5.17 (1H, brs, H-200 ), 5.05 (1H, brs, H-600 ), 3.17 (2H, d, ethyl acetate (20:1) to obtain 6 (134.5 mg, 93%) as col-
J = 6.2 Hz, H-100 ), 1.98 (2H, brs, H-500 ), 1.90 (2H, d, orless prism [46, 47], m.p. 67–68 C; IR (KBr) cm-1:
J = 6.7 Hz, H-400 ) 1.70 (3H, s, H-900 ), 1.60 (3H, s, H-800 ), 1.53 3022, 3017, 2959, 1591, 1508, 1290, 1238, 1152, 1082,
(3H, s, H-1000 ). 13C-NMR (DMSO-d6) d: 157.1 (s), 156.1 (s), 1036, 1007, 962, 924, 841. 1H-NMR (CDCl3) d: 7.44 (2H,
135.6 (s), 133.1 (d), 130.7 (s), 128.2 (d), 127.8 (d), 126.9 (d), d, J = 8.7 Hz, H-20 and H-60 ), 7.04 (1H, d, J = 16.4 Hz,
124.3 (d), 123.3 (d), 115.6 (d), 113.9 (s), 104.3 (s or d), 39.5 (t), H-a), 7.03 (2H, d, J = 8.7 Hz, H-30 and H-50 ), 6.91 (1H, d,
25.5 (t or q), 22.1 (t), 17.6 (q), 16.0 (q). HRESIMS m/z: J = 16.4 Hz, H-b), 6.85 (2H, d, J = 2.1 Hz, H-2 and H-6),
365.2114 ([M ? H]?, calcd for C24H29O3: 365.2117). 6.64 (1H, t, J = 2.1 Hz, H-4), 5.20 (6H, s, 3x-CH2) 3.50
(E)-3,5,40 -Trihydroxy-2,6-geranylstilbene (5c): light (6H, s, 2x-OCH3) 3.49 (3H, s, –OCH3). 13C-NMR (CDCl3)
brown oil; IR (CH2Cl2) cm-1: 3376, 2968, 2923, 1606, d: 158.6 (s), 157.1 (s), 139.9 (s), 131.1 (s), 128.9 (d), 127.9
1513, 1445, 1377, 1261, 1170, 1104, 830. 1H-NMR (d), 126.7 (d), 116.4 (d), 107.8 (d), 104.0 (d), 94.6 (t), 94.5

123
 J Nat Med

(t), 56.2 (q), 56.2 (q). HREIMS m/z: 360.1574 ([M]?, calcd Biological evaluation
for C20H24O6: 360.1573).
Determination of the inhibitory effect on b-secretase
Preparation of (E)-3,5,40 -tri(methoxymethoxy)-4- (BACE1)
prenylstilbene (7)
Enzyme inhibition assay was carried out using a BACE1
(E)-3,5,40 -Tri(methoxymethoxy)stilbene (6) (1.80 g, fluorescence resonance energy transfer assay, according to
5 mmol) was dissolved in tetrahydrofuran (THF, the manufacturer’s instruction [49]. The assay procedure
100 mL) and placed in an ice-bath. Then, 1.61 M n-BuLi was conducted as follows—BACE1 100 lg was diluted
in hexane (3.4 mL, 1.1 eq, 5.5 mmol) was added to a with 50 mM sodium acetate buffer (pH 4.5) with 10%
solution and stirred at 0 C for 1 h 30 min. Prenyl bro- glycerol to obtain stock solution of the enzyme (0.5 U/mL).
mide (1.6 mL, 2.5 eq, 12.5 mmol) was added dropwise The working solution of the enzyme (0.01 U/mL) was
to the reaction mixture at 0 C, which was stirred at prepared by diluting the stock solution of the enzyme with
0 C for 2 h. After complete reaction, the reaction 100 mM sodium acetate buffer (pH 4.5). The BACE1
mixture was diluted with saturated NH4Cl solution substrate IV stock solution (1250 lM) in DMSO was
(30 mL) and water (70 mL) and then extracted with freshly prepared before use and also diluted with 100 mM
EtOAc (3 9 100 mL). The combined extracts were sodium acetate buffer (pH 4.5) to provide the working
washed with brine (100 mL), dried (Na2SO4), and solution of substrate (125 lM). The mixture of the solution
evaporated under reduced pressure. After removal of the of BACE1 enzyme (30 lL), 125 lM of substrate solution
solvent, the products were purified by silica gel column (20 lL) and sample dissolved in 5% DMSO were intro-
chromatography using hexane/ethyl acetate (19:1) to duced to the 96-well plate and gently mixed. Sodium
give compounds 7 (1.14 g, 53%) as an oil [48], IR acetate buffer was then added to adjust the final volume of
(CHCl3) cm-1: 3009, 2961, 1599, 1574, 1510, 1236, 100 lL. The reaction mixture was incubated at 37 C for
1200, 1153, 1109, 1080, 1049, 1007, 922. 1H-NMR 30 min in the dark. After complete reaction, the fluores-
(CDCl3) d: 7.44 (2H, d, J = 8.4 Hz, H-20 and H-60 ), 7.02 cence was monitored at 380 nm (excitation wavelength)
(2H, d, J = 8.4 Hz, H-30 and H-50 ), 6.95 (1H, d, and 510 nm (emission wavelength). b-secretase inhibitor
J = 16.3 Hz, H-a), 6.93 (2H, s, H-2 and H-6), 6.92 (1H, IV (Calbiochem) was used as a reference inhibitor. The
d, J = 16.3 Hz, H-b), 5.25 (4H, s, 2x-CH2), 5.22 (1H, percent inhibition was calculated by the following
m, H-200 ), 5.20 (2H, s, –CH2), 3.51 (3H, s, –OCH3), 3.50 equation:
(6H, s, 2x-OCH3), 3.39 (2H, d, J = 7.2 Hz, H-100 ), 1.80  
ðC  C0 Þ  ðS  S0 Þ
(3H, s, H-500 ), 1.67 (3H, s, H-400 ). 13C-NMR (CDCl3) d: b - secretase inhibitionð%Þ ¼
ðC  C0 Þ
157.0 (s), 156.0 (s), 136.7 (s), 131.4 (s), 131.2 (s), 127.9
 100
(d), 127.8 (d), 127.3 (d), 122.9 (d), 119.7 (d), 116.5 (d),
106.1 (d), 94.6 (t), 94.5 (t), 56.2 (q), 56.1 (q), 25.9 (q), where C was the fluorescence of the control (mixture of
22.9 (t), 17.9 (q). HREIMS m/z: 428.2201 ([M]?, calcd enzyme, buffer, and substrate) after 30 min of incubation,
for C20H24O6: 428.2199). C0 was the fluorescence of the control at zero time, S was
the fluorescence of the tested samples (mixture of enzyme,
Preparation of (E)-3,5,40 -trihydroxy-4- sample solution and substrate) after incubation, and S0 was
prenylstilbene (4b) by optimized condition the fluorescence of the tested samples at zero time. The
assays were run in triplicate. Samples having inhibitory
A solution of (E)-3,5,40 -tri(methoxymethoxy)-4-prenylstil- activity [50% were further evaluated for IC50 and data
bene (7) (279.7 mg, 0.65 mmol) in EtOH (47 mL) was were analyzed via GrahPad Prism4 in nonlinear regression
added to concentrated HCl (47 lL) at room temperature. curve fit.
The reaction mixture was refluxed at 110 C for 1 h. After
complete reaction, the reaction mixture was diluted with Determination of the inhibitory effect on Ab
saturated NaHCO3 solution (20 mL) and extracted with aggregation
EtOAc (3 9 20 mL). Removal of solvent at reduced
pressure left an oily residue, which was then purified by The thioflavin-T fluorescence method was used as descri-
silica gel column chromatography using hexane/ethyl bed by Jiaranaikuwanitch et al. [50]. Experiments were
acetate (4:1) to obtain 4b as pale yellow powder performed by diluting the Ab(1–42) stock solution (Anaspec;
(156.4 mg, 80%). 1384 lM in 1% NH4OH) with 50 mM Tris buffer (pH 7.4)
to give 25 lM of working solution. All compounds were

123
J Nat Med

prepared in DMSO at a concentration of 1000 lM. Then, 1 nitrogen radical and had a UV–vis absorption maximum at
lL of each compound and 9 lL of 25 lM Ab(1–42) peptide 517 nm. DPPH free radical scavenging activity was mea-
were added to the black 96-well plates. The final concen- sured by reducing absorbance at 517 nm, which resulted
tration of each compound was 100 lM and was prepared in from reduction of DPPH free radicals (DPPH•) to the
independent triplicates. The reaction was incubated in the hydrazine form (DPPH-H). To test free radical scavenging
dark at 37 C for 48 h. After incubation, 200 lL of 5 lM effects, samples were adjusted with methanol solution to
thioflavin-T in 50 mM Tris buffer (pH 7.4) was added to final concentrations of 20–200 lM. Methanolic DPPH
each well. The fluorescence was then measured on a (100 lM) was used in the reaction mixture. Serial dilutions
SpectraMaz Gemini EMTM Reader (Thermo Scientific) of the test sample (100 lL) were combined with a freshly
with excitation and emission wavelengths at 446 and prepared DPPH solution (100 lL, 100 lM) in a 96-well
490 nm, respectively. The fluorescence intensities were microtitre plate. MeOH was used as a negative control. In
compared and the percent inhibition of the inhibitor was addition, ascorbic acid was used as a positive control. After
calculated by the following formula: 100-IFi/IFc*100, incubation at 37 C for 30 min in the dark, the absorbance
where IFi and IFc were the fluorescence intensities obtained was measure at 517 nm. Mean values were obtained from
for Ab1–42 in the presence and in the absence of inhibitors, triplicate experiments. Percent inhibition was calculated
respectively. The results of test compounds were calculated using the equation:
for percent inhibition and IC50 values using GrahPad DPPH radical scarvenging activity ð%Þ 
Prism4 in nonlinear regression curve fit. The assays were ðAblank  ðBsample  Cblank of sample ÞÞ
run in triplicate. ¼  100
Ablank
Molecular docking study where A was the absorbance of DPPH radical-methanol
solution, B was the absorbance of sample (DPPH and
All calculations and analysis were carried out using compounds), and C was the absorbance of sample (com-
Autodock 4.2 and PyMOL program. The X-ray crystal pounds) alone.
structure of BACE1 and Ab(1–42) (PDB code 1IYT) used in Percent inhibition was plotted against concentration, and
the docking study were obtained from the Protein Data the equation for the line is used to obtain the IC50 value and
Bank. Heteroatoms and water molecules in the PDB files compared to the positive controls.
were removed at the beginning, and all hydrogen atoms
were subsequently added to the proteins. AUTOGRID Cell culture and neuronal differentiation of P19 cells
program was used to perform pre-calculated atomic affinity
grid maps for each atom type in the ligand, plus an elec- Murine P19 embryonic carcinoma cells were obtained from
trostatics map and a separate desolvation map presented in the American Type Culture Collection (USA) and cultured
the substrate molecule. Flexible ligand docking was per- as described by McBurney and MacPheson [52, 53]. The
formed for the compounds. Each docked system was per- P19 cells were grown in a tissue flask containing P19GM
formed by 100 runs of the AUTODOCK search by the (alpha minimal essential (a-MEM) supplemented with
Lamarckian genetic algorithm. The population size was 7.5% newborn calf serum, 2.5% fetal bovine serum (FBS)
150 and the maximum number of energy evaluations was and 1% antibiotic-antimycotic solution) and then incubated
increased to 5,000,000 per run and the maximum number in a 5% CO2 atmosphere at 37 C. Cells in monolayer
of generations was 27,000. Other than the above-mentioned cultures were maintained in exponential growth by sub-
parameters, the remaining parameters were accepted as culturing every 2 days. To induce differentiation, P19 cells
default. A cluster analysis was performed on the docking were seeded at 2 9 106 cells/mL onto 100-mm non-adhe-
results using a root mean square tolerance of 2.0 and the sive bacteriological-grade dishes in 10 mL P19IM (a-
lowest energy conformation of the highest populated MEM containing 5% FBS, 1% antibiotic-antimycotic
cluster was selected for analysis. Graphic manipulations solution) and 0.5 lM all trans-retinoic acid (RA). Large
and visualizations were performed by PyMOL software aggregates of cells were formed in suspension. After
[49, 50]. 4 days of RA treatment, cell aggregates were dissociated
into single cells by mechanical force, resuspended and
Antioxidant activity assay transferred to poly-L-lysine pre-coated 96-well plates
(plates were previously coated with 50 lg/mL poly-L-
The 1,1-diphenyl-2-picryl hydrazyl (DPPH) radical was lysine dissolved in a phosphate buffer saline (PBS) over-
used to evaluate free radical scavenging activity according night and sterilized under UV light for 30 min) at a cell
to a modified version of Brand-Williams method [51]. The density of 7 9 104 cells/mL in P19SM (a-MEM containing
DPPH radical was a commercially available organic 10% FBS, 1% antibiotic-antimycotic solution). After

123
 J Nat Med

incubation for 24 h, P19SM was changed to P19SM sup- (n = 3), with the medium (a-MEM supplemented with
plemented with 10 lM cytosine-1-b-D-arabinoside (Ara- 10 lM Ara-C, 10% FBS and 1% antibiotic-antimycotic
C). Ara-C was added on the first day after plating to inhibit solution) as a control representing 100% cell viability.
the proliferation of non-neuronal cells. The cultured med-
ium was replenished every 2–3 days. The differentiated Neuritogenicity assay
P19-derived neurons were used after day 14 of the differ-
entiation process [53]. Neuritogenicity on neuronal P19 cells was evaluated by
measuring the length and number of neurites in indi-
Neuronal viability assay vidual cells as described by Tadtong et al. [54–56]. The
assay was carried out with P19-derived neurons cultured
The viability of P19-derived neurons was determined by on a poly-L-lysine-precoated 6-well plate. The a-MEM
XTT reduction assay [54–56]. After 14 days of differenti- supplemented with 10% FBS, 10 lM Ara-C, and 1%
ation process, P19SM in the presence of 10 lM Ara-C was antibiotic-antimycotic solution was removed after
removed and replaced with DMSO solutions of samples 14 days of differentiation process and replaced with
diluted with P19 SM containing 10 lM Ara-C to obtain sample solution in DMSO, diluted with a-MEM sup-
0.001, 0.01, 0.1,1, and 10 lM concentrations. 0.5% DMSO plemented with 10% FBS, 10 lM Ara-C, and 1%
was added to the cultures as a solvent control and P19SM antibiotic-antimycotic solution. Subsequently, the plate
supplemented with 10 lM Ara-C was added to the control was incubated at 37 C in a humidified atmosphere at
wells. The cells were incubated for 18 h at 37 C. Then, 5% CO2 for 24 h. The P19-derived neurons were pho-
150 lL of the medium was removed and replaced with 50 tographed by an inverted microscope using phase-con-
lL of XTT solution (1 mg/mL XTT in a-MEM with trast objectives and measured for the length and number
25 lM of phenazine methosulfate). After incubation for of neurites as well as compared to the control (vehicle
4 h at 37 C, 100 lL of PBS pH 7.4 was added to each well without samples). Quercetin (1 nM) was used as a pos-
and the absorbance of each well was measured at 450 nm itive control. Neurite processes with length equal to or
with a microplate reader. Data were expressed as greater than the diameter of the neuron cell body were
mean ± SEM (n = 3); the medium as a control repre- scored as a neurite-bearing cell. The neurite-bearing cells
sented 100% cell viability. The lowest concentration that were determined and expressed as the mean ± SEM for
enhanced survival of cultured neurons more than the con- three independent experiments (n = 3). Each experiment
trol was further investigated for neuritogenic and neuro- measured the average length and number of neurites of
protective activities against oxidative stress induced by 30 neurons.
serum deprivation.
Statistical analysis
Neuroprotective assay
Statistical analysis was performed using GraphPad Prism 4
Neuroprotective assay was performed on P19-derived for multifunctional activities including anti-Ab aggrega-
neurons cultured in a 96-well plate using the serum tion, anti-b-secretase and antioxidant. The mean values
deprivation method [54–56]. The differentiated P19- were analyzed using one-way analysis of variance (one-
derived neurons were cultured in P19SM plus 10 lM Ara- way ANOVA) and Tukey to compare the statistical sig-
C (a-MEM supplemented with 10% FBS, 10 lM Ara-C nificance between either the resveratrol starting compound
and 1% antibiotic-antimycotic solution). Then, the medium or the positive control and resveratrol derivatives of each
was removed and the DMSO solution of resveratrol biological activity. Significant differences between means
derivatives diluted with a-MEM supplemented with 10 lM at a level of 0.05 (p \ 0.05) were considered significant.
Ara-C, and 1% antibiotic-antimycotic solution without For neuroprotective activity, the average viability of the
FBS was added to give a final concentration of test samples neurons was statistical analyzed using OriginPro 8.5.1.
that enhanced survival of cultured neurons more than the Student’s t test was carried out to identify significant dif-
control. The cells cultured in medium (a-MEM supple- ferences between either the solvent of the control oxidative
mented with 10 lM Ara-C, and 1% antibiotic-antimycotic stress conditions or the resveratrol starting compound and
solution) without FBS were used to create conditions of selected resveratrol derivatives. The average length and
oxidative stress. DMSO was added to the cultures at 0.5% branching numbers of the neurites were statistically sig-
as a solvent control. The cells were incubated in a 5% CO2 nificant between either the control or positive control and
humidified atmosphere at 37 C for 18 h. Cell viability was selected resveratrol derivatives for neuritogenic activity.
determined by XTT reduction assay compared to the sol- Differences were considered significant when the p value
vent control. Data were shown as the mean ± SEM was \0.05.

123
J Nat Med

Results and discussion had lower inhibition activities for Ab(1–42) aggregation
(IC50 [ 100 lM) than resveratrol, with the exception of
Two series of O-alkyl resveratrol derivatives (2a–d, 3a–d) compound 3c. These results indicated that the –OH groups
and C-alkyl resveratrol derivatives (4a–c, 5a–c) were on the aromatic ring are likely to be important in the sta-
synthesized by alkylation to connect the alkyl side chain to bility of the Ab(1–42) from self-aggregation. To investigate
resveratrol. The reaction mixture of the resveratrol and the the possible binding modes of the compounds with
alkyl halide reagents were refluxed for 3 h in the presence Ab(1–42), their interaction with the Ab(1–42) structure (PDB
of DMF or acetone under a nitrogen atmosphere to afford code: 1IYT) was modeled using the Autodock 4.2 and
alkyl products at low to moderate yields (2–62%), as PyMOL software, as shown in Fig. 2. The –OH group on
shown in Scheme 1. As alkyl halide reacts non-selectively the A ring of 5b formed a hydrogen bond with the back-
with the hydroxyl (–OH) groups or aromatic ring, three or bone –NH group of Gln15 with a distance 1.7 Å. Moreover,
four side products were obtained for each series. The the geranyl side chain of 5b displayed hydrophobic and
structures of all synthesized compounds were elucidated by dipole–dipole interactions with Val12, Gln15, Lys16,
IR, Mass, 1H-NMR and 13C-NMR spectroscopy. Phe19 and Phe20 residues (Fig. 2A), which are one of the
The multipotential profiles of the alkylated resveratrol central hydrophobic regions in the aggregation process
derivatives (2a–d, 3a–d, 4a–c, 5a–c) were evaluated for [57]. However, curcumin formed a hydrogen bonding with
their anti-Ab aggregation, BACE1 inhibition, and antioxi- Gln 15 (1.9 Å), the same amino acid residues as 5b and
dant, neuroprotective and neuritogenicity activities. They showed hydrophobic interactions. These results indicated
were first evaluated for inhibition of Ab(1–42) aggregation that the key residue Gln15 and hydrophobic interaction
using a thioflavin T assay [50]. The results showed that C- played a significant role in determining the inhibitory
alkyl resveratrol derivatives (4a–c, 5a–c) were 2- to activity for Ab(1–42) aggregation. The formation of hydro-
20-fold more potent than the resveratrol starting com- gen bonds or hydrophobic interactions of the core struc-
pound, with IC50 values ranging from 1.06-10.00 lM, as tures possibly blocked Ab from self-aggregation because
shown in Table 1. The best results were obtained from Gln15 plays an important role in intermolecular staggering
derivatives bearing a geranyl group at C-4 on the resorcinol of the b-strands of the Ab(1–42) aggregation [58]. The –OH
ring. Compound 5b (IC50 = 1.06 lM) was shown to be group of the resorcinol moiety on A ring of 5a and on B
comparable to that of curcumin (the positive control; IC50 ring of compound 5c provided H-bond interactions with
value of 0.77 lM), followed by 4b (IC50 = 4.78 lM) Asp 23 instead of Gln 15 at a distance of 2.0 and 2.2 Å,
bearing a prenyl group at the same position with 5b. respectively. Moreover, the geranyl groups of these two
However, most of the O-alkylated resveratrol derivatives compounds fitted less well into the hydrophobic region

Scheme 1 Synthesis of O-alkyl resveratrol derivatives (2a–d, 3a–d) and C-alkyl resveratrol derivatives (4a–c, 5a–c)

123
 J Nat Med

Table 1 Multifunctional activity of alkyl resveratrol derivatives

Compounds Anti-Ab aggregation Anti-b-secretase Antioxidant
(DPPH)
% Inhibition (at 100 lM) ICa50 (lM) % Inhibition ICb50 (lM) % Inhibition ICc50 (lM)
(at 50 lM) (at 100 lM)

Resveratrol 77.91 ± 1.37 27.50 ± 1.00 24.98 ± 1.07 [50 82.80 ± 0.24 66.90 ± 0.89
2a 7.39 ± 0.37 [100 10.55 ± 1.54 [50 1.33 ± 0.27 [100
2b 47.37 ± 1.85 [100 18.09 ± 0.78 [50 18.23 ± 0.99 [100
2c 71.01 ± 1.54 44.40 ± 1.01* 37.20 ± 1.17 [50 45.24 ± 0.67 [100
2d 49.32 ± 1.93 [100 8.54 ± 1.92 [50 17.95 ± 0.98 [100
3a 6.19 ± 0.36 [100 12.41 ± 1.86 [50 10.57 ± 0.45 [100
3b 45.73 ± 0.90 [100 20.93 ± 0.99 [50 19.19 ± 0.34 [100
3c 64.12 ± 1.45 22.00 ± 1.42 35.69 ± 1.80 [50 39.07 ± 0.53 [100
3d 49.32 ± 0.55 [100 10.47 ± 0.63 [50 18.06 ± 0.17 [100
4a 85.27 ± 1.30 7.28 ± 1.16* 39.15 ± 1.11 [50 82.19 ± 0.72 31.19 ± 0.82*
4b 87.98 ± 1.50 4.78 ± 0.83* 23.70 ± 0.41 [50 81.00 ± 0.10 41.22 ± 0.57*
4c 84.38 ± 1.79 10.00 ± 0.48* 73.61 ± 1.79 31.39 ± 1.36* 85.17 ± 0.11 28.70 ± 1.41*
5a 85.19 ± 1.53 6.02 ± 0.23* 62.11 ± 1.53 36.70 ± 1.76* 88.04 ± 0.22 21.09 ± 0.79*
5b 88.22 ± 1.39 1.06 ± 0.12* 48.21 ± 1.71 [50 88.19 ± 0.69 25.03 ± 0.79*
5c 91.90 ± 1.25 7.29 ± 0.61* 89.34 ± 1.92 24.99 ± 0.65* 92.25 ± 0.13 20.71 ± 1.12*
Vitamin C – – – – 95.78 ± 0.20 21.63 ± 0.18
Inh IV 96.51 ± 1.33 0.015 ± 0.4e – –
Curcumin 86.28 ± 0.50 0.77 ± 0.10d – – – –
* p \ 0.05 compared with resveratrol starting compound
** p \ 0.05 compared with positive control
a
Data are shown as IC50, inhibitor concentration (mean ± SD) of three experiments required for 50% inactivation of Ab(1-42) self-induced
aggregation
b
Data are shown as IC50, inhibitor concentration (mean ± SD) of three experiments required for 50% inactivation of BACE1
c
Data are shown as IC50, the test compounds concentration that inhibits 50% of free radicals (mean ± SD)
d
Ref. [50]
e
Inh IV is b-secretase inhibitor IV (Calbiochem) as positve control [49].

compared to 5b, which likely accounts for the decreased 8.9–89.3% range at a concentration of 50 lM (Table 1).
prevention of Ab(1–42) formation (Fig. 2b). In addition, 5c The most potent BACE1 inhibitors were (in order) 5c, 4c
was found to be slightly less effective than 5a because of and 5a (IC50 values of 24.99, 31.39 and 36.70 lM,
weak hydrophobic interaction of two long geranyl groups respectively). The binding modes and the structure activity
at C-2 and C-6 of 5c (Fig. 3b). Moreover, 4b containing a relationship of these compounds were studied in silico
prenyl group at C-4 was less effective than 5b containing a using Autodock 4.2 and PyMOL software, which revealed
geranyl group at the same position. This may be due to the three key points. First, the substitution at the C2 and/or C6
shorter alkyl side chain leading to less hydrophobic inter- position (compounds 5c, 5a and 4c) increased BACE1
actions (Fig. 2c). Therefore, these results confirmed that inhibitory activity by fitting into the hydrophobic S3 and
the hydrophobic region possibly prevented Ab(1–42) for- S30 pockets (Fig. 3a) [59]. These compounds had
mation. In the same manner, 4a and 4c exhibited a lower hydrophobic interactions with amino acid residues Gly11,
inhibition of Ab(1–42) aggregation than 5a and 5c, indi- Gln12, Ile110, Val69, Pro70, and Tyr71. Moreover, the –
cating that the length of the alkyl moiety on the resorcinol OH group on the A ring of 5c and 4c was involved in
ring (geranyl vs prenyl) is important for enhancing the hydrogen bonding interactions with the active site Asp32
inhibition of Ab(1–42) aggregation. (2.1 and 2.0 Å distance, respectively), whereas the –OH
The inhibition of BACE1 was measured by a spec- group on the B ring of 5a was observed to bind to Asn 233
trofluorometric method [49, 50]. All the alkylated resver- instead of Asp 32 via two hydrogen bonding interactions at
atrol derivatives showed BACE1 inhibition in the 3.1 Å distance (Fig. 3B). Second, the long chain alkyl

123
J Nat Med

Moreover, all O-alkyl resveratrol derivatives were found to

be less active BACE1 inhibitors than C-alkyl resveratrol
derivatives. These results support that the –OH group is
essential for the BACE1 inhibitory activity.
The reduction of oxidative stress is one the crucial aspects
in designing agents for AD treatment. Consequentially, the
antioxidant activity of these alkyl resveratrol derivatives was
determined using the 2,2-diphenyl-1-picrylhydrazyl (DPPH)
radical scavenging assay [60], and the results are shown in
Table 1. Most of the C-alkyl resveratrol derivatives showed
stronger activity than O-alkyl resveratrol derivatives at
concentrations of 100 lM. Thus, these results indicated that
the free –OH group is seemingly essential for the antioxidant
activity of the alkyl resveratrol derivatives, presumably
because the presence of the free –OH group in the aryl
resorcinol moiety can donate hydrogen atoms and stabilize
electrons by conjugation [61]. Compound 5c was the most
potent antioxidant with an IC50 value of 20.71 lM, followed
by compounds 5a (IC50 = 21.09 lM) and 5b
(IC50 = 25.03 lM), respectively. The IC50 of three ger-
anylated resveratrol derivatives were compared to that of
vitamin C used as a positive control (IC50 = 21.63 lM).
However, prenylated resveratrol derivatives showed slightly
less active than geranylated resveratrol derivatives. The
slightly higher antioxidant activity of all geranylated
resveratrol derivatives (5a–c), compared to the prenylated
resveratrol derivatives (4a–c), may be due to the longer alkyl
groups playing a role in determining the antioxidant activity.
Based on the results for antioxidant activity, C-alkyl
resveratrol derivatives showed effective radical scavenging
activity and also exhibited moderate b-secretase and good
Ab aggregation inhibitory activities. Therefore, they were
selected to determine their influence on neuronal viability
using the P19-derived neuron cell line, which is a
pluripotent stem cell line differentiated into neurons by
retinoic acid and the XTT reduction assay [54, 62], with the
results shown in Fig. 4. The prenylated resveratrol
derivatives promoted a high viability of the cultured neu-
Fig. 2 Possible binding mode of compounds 5a (orange stick), 5b
(green stick), 5c (blue stick), 4b (red stick) with Ab(1–42) (PDB code:
rons, while the geranylated resveratrol derivatives showed
1IYT). a The binding modes of compound 5b compared to curcumin a stronger neurotoxicity at 1 nM (% viability
as positive control (purple stick) with hydrogen-bonding interaction 51.25 ± 13.12, 70.07 ± 36.33 and 34.17 ± 29.98 for 5a,
(black dashed lines) and hydrophobic interaction (yellow residue 5b and 5c, respectively). Interestingly, only 4b showed
lines). b The binding modes of compound 5b compared to compounds
5a and 5c with hydrogen-bonding interaction (black dashed lines) and
[100% neuron viability at 1 nM to 10 lM concentrations,
hydrophobic interaction (yellow residue lines). c The binding modes whereas 4a and 4c at the highest concentration (10 lM)
of compound 5b compared to compound 4b with hydrogen-bonding tended towards neurotoxicity. Overall results suggested
(black dashed lines) and hydrophobic interaction (yellow residue that the prenyl substituent at the C-4 position might play an
lines)
important role in neuronal viability. From these results,
compound 4b was selected for further determination of its
(geranyl) substitution had a higher inhibitory activity than neuroprotective and neuritogenic activities.
the short alkyl (prenyl) chain, due to its greater access to The neuroprotective ability of compound 4b at 1 nM
the S3 and S30 pockets, as shown in 5c to 4c. Finally, the – and 10 lM was evaluated in a serum deprivation model.
OH group at the C-40 position was important in the binding Serum is a complex heterogeneous mixture that is required
with BACE1 via hydrogen bonding, as seen in 5a. for the proliferation of cells in culture. The lack of serum

123
 J Nat Med

Fig. 3 Possible binding mode of compound 4c (pink stick), 5a consist of eight subsites from S4 to S40 [59] and b the binding modes
(orange stick) and 5c (blue stick) with BACE1. a Surface represen- of compounds with hydrogen-bonding interactions (black dashed
tations of compounds interacting with the active site of BACE1 lines)

Fig. 4 Neurotoxicity of 4a–c

and 5a–c on cultured P19-
derived neurons incubated for
18 h. Percent cell viability for
tested compounds versus
control cells was calculated by
absorbance at 450 nm. Each
point represents the average of
three independent
experiments ± SEM (bar) and
each experiment was run in
triplicate

induces oxidative stress conditions that result in cell virtue of their ability to stimulate the outgrowth of neurites
apoptosis [55]. Interestingly, at both 1 nM and 10 lM, 4b from the neuronal and local circuits in the damaged CNS
significantly protected the cultured neurons against serum [63, 64]. Measurement of the length of outgrowth per cell
deprivation (assumed ROS toxicity from serum-deprivation is the most commonly used method to assess the effect of a
induced oxidative stress) at 50.59 ± 3.98 and compound to the growth of neurites. Compound 4b at
53.19 ± 12.48% viability, respectively, as shown in Fig. 5, 1 nM and 10 lM concentrations was evaluated for neuri-
compared to the solvent control group at a 2.95 ± 2.37% togenicity on cultured P19-derived neurons in a poly-L-
viability. In addition, the active compound 4b was more lysine-precoated six-well plate for 18 h. The morphology
effective than the resveratrol starting compound of 30 neuronal cells from three independent experiments
(37.41 ± 4.40% viability) and comparable to that of the (totally counted for 90 neuronal cells) was examined from
quercetin positive control (58.04 ± 9.20% viability). phase-contrast light micrographs. It was evident in all
Neuritogenic substances have the potential for thera- ninety P19-derived neuronal cells that neurite outgrowth
peutic efficacy in the treatment of neuronal injuries by was induced by 4b compared to the untreated (control) and

123
J Nat Med

of new promising drugs for AD. For this reason, 4b was

selected to optimize conditions in order to increase the
percent yield. The semisynthesis was improved in the
following three steps, methoxymethyl (MOM) protection
of three hydroxyl groups, regioselective prenylation and
removal of MOM group as shown in Scheme 2 [35]. First,
(E)-3,5,4́- tri(methoxymethoxy)stilbene (6) was prepared
with NaH and methoxymethyl chloride (MOMCl) in DMF
from resveratrol at high yields (93%). The desired inter-
mediate (7) was obtained by alkylation with n-BuLi in THF
followed by slow addition of prenyl bromide in order to
selectively substitute the prenyl groups on the aromatic
ring and the yield was found to be moderate (53%). Finally,
removal of MOM groups of 7 with concentrated HCl in
ethanol afforded the prenyl target compound (4b) at high
Fig. 5 The effect of 4b in neuronally differentiated P19 cells against yields (80%). Moreover, this active compound (4b) was
oxidative stress induced by serum deprivation. The histogram shows found to be present at a higher level than 4b in the previous
the percentage of cell viability relative to vehicle-treated control
cultures. Each bar represents the mean ± SEM from three individual condition (3% yield) as shown in Scheme 1.
measurements (*p \ 0.05 compared with solvent control of serum-
deprivation condition (a-MEM? 0.5% DMSO) and **p \ 0.05
compared with resveratrol starting compound) Conclusion
quercetin-treated (positive control) cells. Compound 4b In conclusion, two series of O-alkyl (2a–d, and 3a–d) and
both at 1 nM and 10 lM significantly increased the neurite C-alkyl (4a–c, and 5a–c) resveratrol derivatives were
length (109.74 ± 47.18 and 107.90 ± 26.30 lm, respec- designed, semisynthesized and evaluated for the inhibition
tively) approximately two-fold compared to the untreated of Ab aggregation and b-secretase (BACE1), and antioxi-
condition (60.44 ± 21.77 lm) and comparable to that of dant, neuroprotective, and neuritogenicity activities, for the
the quercetin positive control (104.33 ± 41.70 lm). potential treatment of AD. Among the two series of
Moreover, the number of neurites were significantly greater resveratrol derivatives, the C-alkyl resveratrol derivatives
in neuron cells treated with compound 4b at both 1 nM and showed higher rates of multifunctional activities than O-
10 lM concentrations (9.33 ± 5.69 and 11.76 ± 4.10, alkyl resveratrol derivatives. The results confirm that the
respectively) than the control by approximately five-fold substitution of C-alkyl moieties and the –OH groups on the
(2.12 ± 0.81). Interestingly, only compound 4b at 10 lM resveratrol scaffold are essential for good activities of AD.
showed more branching numbers of neurites than the Most of the geranylated resveratrol derivatives were found
quercetin positive control (8.02 ± 4.49), as shown in to possess multifunctional activity with three actions.
Fig. 6. Furthermore, the neurons treated with compound 4b Compound 5c exhibited the most potent BACE1 inhibitory
at the very low concentration of 1 nM remarkably activity and radical scavenging activity as well as inhibi-
enhanced the number and length of neurites compared to tion of Ab aggregation, while 5a showed the most potent
the 10 lM concentration. These results indicated that the antioxidant activity and also moderated anti-BACE1
active compound 4b might be considered as a potentially activity as well as anti-Ab aggregation. Compound 5b was
safe lead compound for the treatment and prevention of the most potent inhibitor of Ab aggregation with good
AD. antioxidant and moderate inhibition of BACE1. However,
All alkylated resveratrol derivatives were screened for all three of these geranylated resveratrol derivatives
their multifunctional activities, including anti-b-secretase, exhibited neurotoxicity on the P19-derived neurons
anti-Ab aggregation, and antioxidant, neuroprotective and whereas a prenylated resveratrol derivative 4b showed no
neuritogenic activities for potential AD treatment. The neurotoxicity on the P19-derived neuronal cells with good
geranylated resveratrol derivatives (5a–c) were found to multifunctionality, including antioxidant, anti-Ab aggre-
possess multifunctional activity, as shown in Table 1. gation and anti-BACE1 activities. Moreover, the neuro-
However, all of them also exhibited neurotoxic effects to protective effect of 4b was better than that of resveratrol
P19-derived neurons. Only 4b showed good anti-Ab and was comparable to that of quercetin as a positive
aggregation, and antioxidant, neuroprotective and neurito- control. Furthermore, the neuritogenic activity of 4b
genic activities without any toxicity to the P19-derived caused more branching numbers and longer neurites than
neurons, which may be useful for the further development the control, and was comparable to the positive control.

123
 J Nat Med

Fig. 6 Phase-contrast micrographs of the neuritogenicity of P19- quercetin (positive control), d 1 nM compound 4b, and e 10 lM
derived neurons after 18 h of incubation and treated with a no compound 4b (Bar = 50 lm, at 1009)
compound (control), b 0.5% DMSO (solvent control), c 1 nM

Scheme 2 Semisynthesis of active prenylated resveratrol derivative (4b)

123
J Nat Med

Collectively, these results suggested that compound 4b 17. Li SY, Wang XB, Kong LY (2014) Design, synthesis and bio-
might be a promising multi-target lead candidate for the logical evaluation of imine resveratrol derivatives as multi-tar-
geted agents against Alzheimer’s disease. Eur J Med Chem
further development of new drugs for AD. 71:36–45
18. Baur JA, Sinclair DA (2006) Therapeutic potential of resveratrol:
Acknowledgements This project was supported by Mahidol the in vivo evidence. Nat Rev Drug Discov 5:493–506
University and the National Research Council of Thailand (NRCT). 19. Galli RL, Bielinski DF, Szprengiel A, Shukitt-Hale B, Joseph JA
This work was supported in part by a grant from the Dementia Drug (2006) Blueberry supplemented diet reverses age-related decline in
Resource Development Center Project (DRC), the Ministry of Edu- hippocampal HSP70 neuroprotection. Neurobiol Aging 27:344–350
cation, Culture, Sports, Science and Technology (MEXT), Japan 20. Smoliga JM, Baur JA, Hausenblas HA (2011) Resveratrol and
(S1511016). We also express their gratitude for research funding from health—a comprehensive review of human clinical trials. Mol
the Meiji Pharmaceutical University Asia/Africa Center for Drug Nutr Food Res 55:1129–1141
Discovery (MPU-AACDD). 21. Choi B, Kim S, Jang BG, Kim MJ (2016) Piceatannol, a natural
analogue of resveratrol, effectively reduces beta-amyloid levels
via activation of alpha-secretase and matrix metalloproteinase-9.
References J Funct Foods 23:124–134
22. Yazir Y, Utkan T, Gacar N, Aricioglu F (2015) Resveratrol exerts
anti-inflammatory and neuroprotective effects to prevent memory
1. Alzheimer’s association (2016) Alzheimer’s disease facts and deficits in rats exposed to chronic unpredictable mild stress.
figures. Alzheimers Dement 12:1–80 Physiol Behav 138:297–304
2. Okamura H, Ishii S, Ishii T, Eboshida A (2013) Prevalence of 23. Turner RS, Thomas RG, Craft S, Van Dyck CH, Mintzer J,
dementia in Japan: a systematic review. Dement Geriatr Cogn Reynolds BA, Aisen PS (2015) A randomized, double-blind,
Disord 36:111–118 placebo-controlled trial of resveratrol for Alzheimer disease.
3. Klafki HW, Staufenbiel M, Kornhuber J, Wiltfang J (2006) Neurology 85:1383–1391
Therapeutic approaches to Alzheimer’s disease. Brain 24. Tellone E, Galtieri A, Russo A, Giardina B, Ficarra S (2015)
129:2840–2855 Resveratrol: a focus on several neurodegenerative diseases. Oxid
4. Terry JRAV, Buccafusco JJ (2003) The cholinergic hypothesis of Med Cell Longev doi:10.1155/2015/392169
age and Alzheimer’s disease-related cognitive deficits: recent 25. Ohno M, Sametsky EA, Younkin LH, Oakley H, Younkin SG,
challenges and their implications for novel drug development. Citron M, Disterhoft JF (2004) BACE1 deficiency rescues
J Pharm Exp Ther 306:821–827 memory deficits and cholinergic dysfunction in a mouse model of
5. Giacobini E (2004) Cholinesterase inhibitors: new roles and Alzheimer’s disease. Neuron 41:27–33
therapeutic alternatives. Pharmocol Res 50:433–440 26. DaSilva KA, Shaw JE, McLaurin J (2010) Amyloid-b fibrillo-
6. Selkoe DJ, Hardy J (2016) The amyloid hypothesis of Alzhei- genesis: structural insight and therapeutic intervention. Exp
mer’s disease at 25 years. EMBO Mol Med 8:595–608 Neurol 223:311–321
7. Bihel F, Das C, Bowman MJ, Wolfe MS (2004) Discovery of a 27. Nunomura A, Castellani RJ, Zhu X, Moreira PI, Perry G, Smith
subnanomolar helical D-tridecapeptide inhibitor of b-secretase. MA (2006) Involvement of oxidative stress in Alzheimer disease.
J Med Chem 47:3931–3933 J Neuropathol Exp Neurol 65:631–641
8. Nishiyama Y, Taguchi H, Hara M, Planque SA, Mitsuda Y, Paul 28. Mandel S, Youdim MB (2004) Catechin polyphenols: neurode-
S (2014) Metal-dependent amyloid b-degrading catalytic anti- generation and neuroprotection in neurodegenerative diseases.
body construct. J Biotechnol 180:17–22 Free Radic Biol Med 37:304–317
9. Maccioni RB, Farias G, Morales I, Navarrete L (2010) The 29. Singh NA, Mandal AKA, Khan ZA (2016) Potential neuroprotec-
revitalized tau hypothesis on Alzheimer’s disease. Arch Med Res tive properties of epigallocatechin-3-gallate (EGCG). Nutr J 15:60
41:226–231 30. More SV, Koppula S, Kim IS, Kumar H, Kim BW, Choi DK
10. Boutajangout A, Sigurdsson EM, Krishnamurthy PK (2011) Tau (2012) The role of bioactive compounds on the promotion of
as a therapeutic target for Alzheimer’s Disease. Curr Alzheimer neurite outgrowth. Molecules 17:6728–6753
Res 8:666–677 31. Zhao HF, Li N, Wang Q, Cheng XJ, Li XM, Liu TT (2015)
11. Anand P, Singh B (2013) A review on cholinesterase inhibitors Resveratrol decreases the insoluble Ab-42 Level in hippocamous
for Alzheimer’s disease. Arch Pharm Res 36:375–399 and protects the integrity of the blood-brain barrier in AD rats.
12. Cavalli A, Bolognesi ML, Minarini A, Rosini M, Tumiatti V, Neuroscience 310:641–649
Recanatini M, Melchiorre C (2008) Multi-target-directed ligands 32. Bastianetto S, Zheng WH, Quirion RJ (2000) Neuroprotective
to combat Neurodegenerative diseases. J Med Chem 51:347–372 abilities of resveratrol and other red wine constituents against
13. Leon R, Garcia AG, Contelles JM (2013) Recent advances in the nitric oxide-related toxicity in cultured hippocampal neurons. Br
multitarget-directed ligands approach for the treatment of Alz- J Pharmacol 131:711–720
heimer’s disease. Med Res Rev 33:139–189 33. Vella F, Ferry G, Delagrange P, Boutin JA (2005) NRH:quinone
14. Li RS, Wang XB, Hu XJ, Kong LY (2013) Design, synthesis and reductase 2: an enzyme of surprises and mysteries. Biochem
evaluation of flavonoid derivatives as potential multifunctional Pharmacol 71:1–12
acetylcholinesterase inhibitors against Alzheimer’s disease. 34. Bastianetto S, Menard C, Quirion R (2015) Neuroprotective
Bioorg Med Chem Lett 23:2636–2641 action of resveratrol. Biochim Biophys Acta 1852:1195–1201
15. Xu ZC, Wang XB, Yu WY, Xie SS, Li SY, Kong LL (2014) 35. Cho JK, Ryu YB, Curtis-Long MJ, Kim JY, Kim D, Lee WS,
Design, synthesis and biological evaluation of benzylisoquinoline Park KH (2011) Inhibition and structural reliability of prenylated
derivatives as multifunctional agents against Alzheimer’s disease. flavones from the stem bark of Morus lhou on b-secretase
Bioorg Med Chem Lett 24:2368–2373 (BACE-1). Bioorg Med Chem Lett 21:2945–2948
16. Lu C, Guo Y, Yan J, Luo Z, Luo HB, Yan M, Huang L, Li X 36. Marumoto S, Miyazawa M (2012) Structure–activity relation-
(2013) Design, synthesis, and evaluation of multitarget-directed ships for naturally occurring coumarins as b-secretase inhibitor.
resveratrol derivatives for the treatment of Alzheimer’s disease. Bioorg Med Chem 20:784–788
J Med Chem 56:5843–5859

123
 J Nat Med

37. Chanmahasathien W, Li Y, Satake M, Oshima Y, Ruangrungsi N, 51. Brand-Williams W, Cuvelier ME, Berset C (1995) Use of a free
Ohizumi Y (2003) Prenylated xanthones with NGF-potentiating radical method to evaluate antioxidant activity. Lebensm Wiss
activity from Garcinia xanthochymus. Phytochemistry Technol 28:25–30
64:981–986 52. McBurney MW (1993) P19 embryonal carcinoma cells. Intl J
38. Kano Y, Horie N, Doi S, Aramaki F, Maeda H, Hiragami F, Dev Biol 37:135–140
Kawamura K, Motoda H, Koike Y, Akiyama J, Eguchi S, 53. MacPherson PA, McBurney MW (1995) P19 embryonal carci-
Hashimoto K (2008) Artepillin C derived from propolis induces noma cells: a source of cultured neurons amenable to genetic
neurite outgrowth in PC12m3 cells via ERK and p38 MAPK manipulation. Methods 7:238–252
pathways. Neurochem Res 33:1795–1803 54. Tadtong S, Athikomkulchai S, Sareedenchai V (2012) Neurito-
39. Chao J, Li H, Cheng KW, Yu MS, Chang RC, Wang M (2010) genic activity of Thai plant extracts. J Health Res 26:293–296
Protective effects of pinostilbene, a resveratrol methylated 55. Tangsaengvit N, Kitphati W, Tadtong S, Bunyapraphatsara N,
derivative, against 6-hydroxydopamine-induced neurotoxicity in Nukoolkarn V (2013) Neurite outgrowth and neuroprotective
SH-SY5Y cells. J Nutr Biochem 21:482–489 effects of quercetin from Caesalpinia mimosoides Lamk. on
40. Orsini F, Verotta L, Lecchi M, Restano R, Curia G, Redaelli E, oultured P19-derived neurons. Evid Based Complement Alternat
Wanke E (2004) Resveratrol derivatives and their role as potas- Med doi:10.1155/2013/838051
sium channels modulators. J Nat Prod 67:421–426 56. Tadtong S, Kanlayavattanakul M, Lourith N (2013) Neuritogenic
41. Ruan BF, Huang XF, Ding H, Xu C, Ge HM, Zhu HL, Tan RX and neuroprotective activities of fruit residues. Nat Prod Com-
(2006) Synthesis and cytotoxic evaluation of a series of resver- mun 8:1583–1586
atrol derivatives. Chem Biodivers 3:975–981 57. Hamada Y, Miyamoto N, Kiso Y (2015) Novel b-amyloid
42. Kumano T, Richard SB, Noel JP, Nishiyama M, Kuzuyama T aggregation inhibitors possessing a turn mimic. Bioorg Med
(2008) Chemoenzymatic syntheses of prenylated aromatic small Chem Lett 25:1572–1576
molecules using Streptomyces prenyltransferases with relaxed 58. Ahmed M, Davis J, Aucoin D, Sato T, Ahuja S, Aimoto S, Elliott
substrate specificities. Bioorg Med Chem 16:8117–8126 JI, Nostrand WV, Smith SO (2010) Structural conversion of
43. Park BH, Lee HJ, Lee YR (2011) Total synthesis of chiricanine neurotoxic amyloid-beta(1–42) oligomers to fibrils. Nat Struct
A, arahypin-1, trans-arachidin-2, trans-arachidin-3, and arahy- Mol Biol 17:561–567
pin-5 from peanut seeds. J Nat Prod 74:644–649 59. Shimizu H, Tosaki A, Kaneko K, Hisano T, Sakurai T, Nukina N
44. Verotta L, Orsini F, Gerhauser C, Klimo K (2009) Biologically- (2008) Crystal structure of an active form of BACE1, an enzyme
activity stilbene derivatives and compositions thereof. PCT Int responsible for amyloid b protein production. Mol Cell Biol
Appl WO 2009012910 A1 (patent number) 28:3663–3671
45. Kumano T, Tomita T, Nishiyama M, Kuzuyama T (2010) 60. Kim SH, Kumar CN, Kim HJ, Kim DH, Cho J, Jin C, Lee YS
Functional characterization of the promiscuous prenyltransferase (2009) Glucose-containing flavones—their synthesis and antiox-
responsible for furaquinocin biosynthesis identification of a idant and neuroprotective activities. Bioorg Med Chem Lett
physiological polyketide substrate and its prenylated reaction 19:6009–6013
products. J Biol Chem 285:39663–39671 61. Hossain SU, Bhattacharya S (2007) Synthesis of O-prenylated
46. Rodrı́guez RA, Lahoz IR, Faza ON, Cid MM, Lopez CS (2012) and O geranylated derivatives of 5-benzylidene2,4-thiazolidine-
Theoretical and experimental exploration of the photochemistry diones and evaluation of their free radical scavenging activity as
of resveratrol: beyond the simple double bond isomerization. Org well as effect on some phase II antioxidant/detoxifying enzymes.
Biomol Chem 10:9175–9182 Bioorg Med Chem Lett 17:1149–1154
47. Mattarei A, Azzolini M, Carraro M, Sassi N, Zoratti M, Paradisi 62. Tadtong S, Meksuriyen D, Tanasupawat S, Isobe M, Suwan-
C, Biasutto L (2013) Acetal derivatives as prodrugs of resvera- borirux K (2007) Geldanamycin derivatives and neuroprotective
trol. Mol Pharm 10:2781–2792 effect on cultured P19-derived neurons. Bioorg Med Chem Lett
48. Hartung AM, Beutler JA, Navarro HA, Wiemer DF, Neighbors 17:2939–2943
JD (2014) Stilbenes as K-selective, non-nitrogenous opioid 63. More SV, Koppula S, Kim IS, Kumar H, Kim BW, Choi DK
receptor antagonists. J Nat Prod 77:311–319 (2012) The role of bioactive compounds on the promotion of
49. Jiaranaikulwanitch J, Boonyarat C, Fokin VV, Vajragupta O neurite outgrowth. Molecules 7:6728–6753
(2010) Triazolyl tryptoline derivatives as b-secretase inhibitors. 64. Li P, Matsunaga K, Yumakuni T, Ohizumi Y (2002) Picrosides I
Bioorg Med Chem Lett 20:6572–6576 and II, selective enhancers of the mitogen-activated protein
50. Jiaranaikulwanitch J, Govitrapong P, Fokin VV, Vajragupta O kinase-dependent signaling pathway in the action of neuritogenic
(2012) From BACE1 inhibitor to multifunctionality of tryptoline substances on PC12D cells. Life Sci 71:1821–1835
and tryptamine triazole derivatives for Alzheimer’s disease.
Molecules 17:8312–8333

123