Ferulic acid-carbazole hybrid compounds: Combination of cholinesterase inhibition, antioxidant and neuroprotection as multifunctional anti-Alzheimer agents

Article abstract of DOI:10.1016/j.bmc.2016.01.010

In order to search for novel multifunctional anti-Alzheimer agents, a series of ferulic acid-carbazole hybrid compounds were designed and synthesized. Ellman's assay revealed that the hybrid compounds showed moderate to potent inhibitory activity against the cholinesterases. Particularly, the AChE inhibition potency of compound 5k (IC₅₀ 1.9 μM) was even 5-fold higher than that of galantamine. In addition, the target compounds showed pronounced antioxidant ability and neuroprotective property, especially against the ROS-induced toxicity. Notably, the neuroprotective effect of 5k was obviously superior to that of the mixture of ferulic acid and carbazole, indicating the therapeutic effect of the hybrid compound is better than the combination administration of the corresponding mixture.

Full text of DOI:10.1016/j.bmc.2016.01.010

Bioorganic & Medicinal Chemistry xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Ferulic acid–carbazole hybrid compounds: Combination of
cholinesterase inhibition, antioxidant and neuroprotection as
multifunctional anti-Alzheimer agents
c,
Lei Fang a,b, Mohao Chen a, Zhikun Liu a,c, Xubin Fang a, Shaohua Gou a, , Li Chen
⇑
⇑
a Jiangsu Province Hi-Tech Key Laboratory for Bio-medical Research and School of Chemistry and Chemical Engineering, Southeast University, Nanjing 211189, China
b State Key Laboratory of Natural Medicines, China Pharmaceutical University, Nanjing 210009, China
c Department of Natural Medicinal Chemistry, China Pharmaceutical University, Nanjing 210009, China
a r t i c l e i n f o
a b s t r a c t
Article history:
In order to search for novel multifunctional anti-Alzheimer agents, a series of ferulic acid–carbazole
hybrid compounds were designed and synthesized. Ellman’s assay revealed that the hybrid compounds
showed moderate to potent inhibitory activity against the cholinesterases. Particularly, the AChE
Received 5 November 2015
Revised 5 January 2016
Accepted 7 January 2016
Available online xxxx
inhibition potency of compound 5k (IC50 1.9 lM) was even 5-fold higher than that of galantamine. In
addition, the target compounds showed pronounced antioxidant ability and neuroprotective property,
especially against the ROS-induced toxicity. Notably, the neuroprotective effect of 5k was obviously
superior to that of the mixture of ferulic acid and carbazole, indicating the therapeutic effect of the hybrid
compound is better than the combination administration of the corresponding mixture.
Ó 2016 Elsevier Ltd. All rights reserved.
Keywords:
Hybrid compounds
Antioxidant
Anti-Alzheimer
AChE inhibitor
1. Introduction
Facing the complex etiology of AD, developing novel agents
with multiple pharmacological effects has become a promising
Alzheimer’s disease (AD) is a neurodegenerative and life-threat-
ening disease characteristic of a progressive impairment of cogni-
tive functions, behavioral disturbances and a decreasing ability to
perform basic activities of daily living. Though the disease has been
identified for more than 100 years, it is still incurable due to its
complex pathogenesis. Substantial evidences have revealed that
AD is a multifactorial syndrome derived from a complex array of
neurochemical factors, involving the deficiency of synaptic
acetylcholine and other related neurotransmitters, the formation
of neurotoxic beta-amyloid (Ab) peptide, oxidative stress, the
inflammation of neurons, and so on.1,2 Nowadays, only three
acetylcholinesterase inhibitors (AChEIs), including galantamine,
strategy in the search for new anti-AD agents. In this context,
hybrid molecules, which consist of two or more pharmacophores
in one molecule and could simultaneously target different
pathogenic factors of AD, have attracted more and more attention.
Carbazole derivatives have been reported to be able to inhibit the
aggregation of Ab which is one of the neuropathology characteris-
tics of AD and plays a key role in triggering a neurotoxic cascade.3
Before long, we also found that 2,8-disubstituted carbazole deriva-
tives (e.g., 1a), which could be regarded as the D-ring opened ana-
logs of galantamine, could inhibit cholinesterase (ChE) and protect
neurons from the toxicity induced by Ab oligomers.4 These unique
characters of carbazole derivatives make them ideal lead structures
for developing multifunctional anti-AD agents. Reactive oxygen
species (ROS) are now regarded as another major etiological factor
of AD since it is confirmed that ROS relate to formation of both
amyloid plaques and neurofibrillary tangles, which are two major
pathological hallmarks of AD.5 Natural antioxidant ferulic acid
(FA) could effectively scavenge ROS and exert protective effect on
neurons against the ROS-induced toxicity in vitro, suggesting it
may be useful for the treatment of AD.6 Thereby, in order to take
advantage of these two scaffolds, we have designed and synthe-
sized a series of FA–carbazole hybrid compounds which contain
the carbazole moiety connected to the carboxylic acid group of
FA via an amide bond (Fig. 1). We hypothesized that the hybrid
rivastigmine and donepenzil, and one N-methyl-D-aspartic acid
receptor antagonist memantine are clinically available for the
treatment of AD. The oldest AChEI tacrine was withdrawn from
the market due to its serious hepatotoxicity. Notably, though
AChEIs can enhance the level of synaptic acetylcholine (ACh) and
consequently improve the cholinergic function in the central ner-
vous system, they cannot halt the progression of AD and thereby
only a symptom-ameliorating effect can be obtained.
⇑
Corresponding authors. Tel./fax: +86 25 83272381
E-mail addresses: sgou@seu.edu.cn (S. Gou), chenliduo12@gmail.com (L. Chen).
http://dx.doi.org/10.1016/j.bmc.2016.01.010
0968-0896/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

2
L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
carbazole derivatives could inhibit cholinesterases, we wondered
COOH
R1
whether the hybrid compounds retained such activity. Thus, the
hybrid compounds were screened for the cholinesterase inhibitory
activity in vitro by Ellman’s assay. The results (Table 1) turned out
that most of the synthesized compounds (except 5d, 5e, 5h–j)
showed moderate to potent inhibitory effect on the AChE, with
HO
OCH3
Ferulic acid
O
NH
N
R2
H
HO
N
IC50 values ranging from 1.9 to 88.2 lM. Notably, the potency of
R2
OCH3
Hybrid compounds
compounds 1k and 5k was even higher than that of galantamine.
Considering AChEs from different enzyme sources show significant
structural difference, we further tested the inhibition effect of 1f,
1k, 5f, and 5k on human AChE. The results turned out that all of
the four compounds showed potent inhibitory activity with IC50
R1
N
H
Carbazoles
values from 5.1 to 22.0 lM is at the same level of the potency
Figure 1. Drug design rationale of the target compounds.
against AChE from Electric Eel. Analyzing the structure–activity
relationship we found that the substituents R1 and R2 had impor-
tant influence on the activity. When R2 was ethyl (e.g., 5g) or iso-
propyl (e.g., 5f) which had a relative small size, the target
compounds showed good activity. When the size of R2 increased,
the activity obviously reduced, indicating small alkyl substituents
were optimal for the activity. Besides, the electric effect of R1 also
played an important role in the activity. Replacing the electron-
donating methoxyl group with electron-accepting chlorine atom
significantly improved the inhibitory activity, indicating the elec-
tron-withdrawing substituents could make a contribution to the
activity. As for the butyrylcholinesterase (BChE), the activity levels
of all compounds against BChE were generally higher than those
compounds could simultaneously possess the ChE inhibition activ-
ity from the carbazole scaffold and the antioxidant and neuropro-
tective activity from the FA scaffold, and as a consequence, novel
multifunctional anti-AD agents could be obtained.
2. Results and discussion
2.1. Chemistry
The synthesis of FA–carbazole hybrids 5a–k was outlined in
Schemes 1 and 2. Firstly, the carbazole derivatives 1a–k were suc-
cessfully prepared using our previously reported method
(Scheme 1).4 Then, the phenol hydroxyl functional group of FA
was protected by acylation with ethyl carbonochloridate to give
the ester 2, which was further reacted with oxalyl chloride to form
the acyl chloride intermediate 3. Thereafter, compounds 4a–k
were synthesized by treating 3 with 1a–k in the presence of pyri-
dine, respectively. Finally, the protection group was removed by
the treatment with aminoethanol in 95% ethanol aqueous solution
to yield the target compounds 5a–k (Scheme 2).
against AChE with IC50 values varying from 1.9 to 25.9 lM. Besides,
the potency difference was also observed between the hAChE and
hBChE groups. The difference of the activity against these two
isoenzymes may origin from the structural difference of the isoen-
zymes. It was reported that at the midgorge level several aromatic
residues of AChE are replaced by smaller aliphatic ones in BChE,
which leads to the larger void along the BChE gorge with respect
to AChE.7
2.3. Molecular modeling and ADME prediction
2.2. Cholinesterase inhibition
In order to investigate the interaction mode of the synthesized
compounds with the target enzymes, we further performed molec-
ular modeling study. It was revealed that the large active gorge of
BChE allowed the ligands with large steric hindrance (e.g., 5i) to
enter the pocket and interact with the corresponding residues,
whereas in the case of AChE the large steric hindrance of the
The deficiency of synaptic ACh and other related neurotrans-
mitters is one of most important etiological factors of AD. In order
to compensate ACh in the brain of AD patients, AChEIs have been
developed and acting as the mainstay for the symptomatic treat-
ment of AD. Since our former study revealed that 2,8-disubstituted
H3C
H3C
Br
B(OH)2
i
ii
+
R1
R1
NO2
CH 3
N
H
R1
NO2
BrH2C
H
N
H3C
N
R2
iii
iv
v
R1
R1
R1
N
N
O
S
O
O
S
O
O
S
O
Ph
Ph
Ph
H
R1
R2
R1
R2
N
R2
1g
1h
1i
H
H
H
H
Cl
CH2CH3
1a
OCH3
CH(CH3)2
CH2CH3
1b OCH3
vi
R1
1c
1d
OCH3
OCH3
N
H
1j
CH2Ph
1k
CH2CH3
1e OCH3
1f
CH2Ph
1a-k
H
CH(CH3)2
Scheme 1. The synthetic procedure of compounds 1a–k. Regents and condition: (i) Na2CO3, Pd(PPh3)4, DME, reflux, 20 h; (ii) PPh3, DCB, reflux, 3 h; (iii) PhSO2Cl, NaH, THF,
0 °C/rt, overnight; (iv) NBS, AIBN, CCl4, reflux, 3 h; (v) alkylamine, KI, K2CO3, anhydrous acetone, rt, overnight; (vi) 2 N NaOH, ethanol, reflux, overnight.
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
3
COCl
COOH
COOH
O
O
i
ii
O
O
O
O
HO
OCH3
OCH3
R1
OCH3
2
3
R1
O
O
NH
NH
iv
iii
O
O
N
N
R2
R2
O
HO
OCH3
4a-k
OCH3
5a-k
Scheme 2. The synthetic procedure of the target compounds 5a–k. Regents and condition: (i) EtOCOCl, NaOH, H2O, rt, 0.5 h; (ii) oxalyl chloride, CH2Cl2, DMF, À5 °C, 1 h; (iii)
1a–k, pyridine, CH2Cl2, 0 °C,1 h; (iv) NH2CH2CH2OH, 95% EtOH, rt.
Table 1
Inhibitory effect of 1b, 1f, 1k, 5a–k and galantamine (abbreviated as Gal.) on AChE and BChE (IC50 values)
R1
O
NH
N
R2
HO
OCH3
5a-k
Compd
R1
R2
IC50 SEMa
(lM)
AChEb
hAChEe
BChEb
hBChEe
Gal.
1a
1f
1k
5a
5b
—
—
8.5 1.5
28.1 6.2
10.1 7.7c
8.9 2.0
1.9 0.2
13.9 2.6
19.8 4.2
AOCH3
AH
ACl
AOCH3
AOCH3
ACH(CH3)2
ACH(CH3)2
ACH2CH3
ACH(CH3)2
ACH2CH3
50.8 12.1c
21.6 7.2
2.1 0.6
18.0 2.1
5.1 0.8
8.1 1.1
7.9 0.5
58.2 11.2
61.6 13.8
5c
AOCH3
88.2 17.1
25.9 9.3
5d
AOCH3
>100
20.5 3.2
5e
5f
5g
AOCH3
AH
AH
ACH2Ph
ACH(CH3)2
ACH2CH3
>100
11.1 2.4
18.0 1.9
n.d.d
12.7 1.3
13.8 2.8
22.0 4.0
9.0 2.1
5h
AH
>100
31.6 11.4
5i
AH
>100
22.0 2.5
5j
AH
ACH2Ph
>100
n.d.d
5k
ACl
ACH2CH3
1.9 0.8
6.9 0.9
3.1 0.4
2.8 0.4
a
b
c
Data are the mean values of at least three determinations.
AChE from Electric Eel and BChE from equine serum were used.
Values are cited from Ref. 4.
n.d. Means not determined.
Human AChE and BChE were used.
d
e
cyclohexyl group of 5i totally blocked the interactions
(Fig. 2A and B). In contrast, compound 5k whose R2 is a small ethyl
group could effectively enter the active pocket of AChE and interact
candidate should be not higher than 5. The high Clogp value indi-
cates a good lipophilicity of the target compound. Though the high
lipophilicity decreases the aqueous solubility of 5k as compared
with tacrine, it guarantees a good ability of the target compound
to penetrate BBB, which was also confirmed by the prediction.
with the key residue Trp84 through a p–p interaction (Fig. 2C). Its
interaction mode is similar to that of galantamine.8
Given the target compounds were designed for the treatment of
the central nervous system disease, the physicochemical proper-
ties such as Clogp and blood–brain barrier (BBB) penetration abil-
ity were essential for the drug-likeness of the target compounds.
Thereby, ADMET of 5k was predicted using ADMET predictor 7.0
(Table 2). It was found that the Clogp of 5k is 4.39. This value is
higher than that of tacrine but still within the range given by
Lipinski’s Rule of Five which suggests the optimal logp value of drug
2.4. Free radical scavenging activity
The antioxidant effect is believed to be responsible for the anti-
AD property of FA. Thereby, using the UV spectroscopy method,9
the capacity of different concentrations (1, 2, 10, 50, 100 lM) of
the target compounds to eliminate 2,2-diphenyl-1-picrylhydrazyl
free radical (DPPH) as well as galvinoxyl radicals was determined
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

4
L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
Figure 2. (A) The interaction of 5i and BChE (Protein Data Bank code 1P0P); (B) the interaction of 5i and AChE (Protein Data Bank code 3I6M); (C) the interaction of 5k and
AChE (Protein Data Bank code 3I6M). The pictures were generated from MOE.
Table 2
in vitro. The free radical scavenging activity (FRSA) of the mixture
of 1k and FA (molar ratio: 1:1), which can be considered as corre-
The ADMET prediction results of 5k and tacrine
sponding bioactive components of the hybrid compound, was also
tested for the comparison of the activity of the hybrid compound
and the combination administration of the corresponding mixture.
The results were shown in Tables 3 and 4. It was found that all of
the tested compounds could effectively scavenge the free radicals.
Similar to the action manner of FA, the FRSA of the target com-
pounds against DPPH (Table 3) was generally higher than that
against the galvinoxyl radicals (Table 4). As for DPPH, even at the
Properties
5k
Tacrine
Clogp
4.39
2.79
10.06
pKa
11.91, 9.02
1.14 Â 10À3
270.51
High
6.16
hE, Hp
Solubilitya
MDCKb
4.71 Â 10À1
353.51
High
BBB penetrationc
ADMET riskd
Toxicitye
5.45
ra, SG, Hp, Mu
a
b
c
Water solubility (mg/mL).
lowest concentration (i.e., 1
compounds reached 20–40%; in contrast, the galvinoxyl FRSA% of
the same compounds at was only in the range of
lM), the DPPH FRSA% of the target
MDCK permeability (cm/s  107).
Likelyhood of blood–brain barrier penetration.
d
1 lM
A score in the 0–24 range indicating the number of potential ADMET problems a
compound might have. The higher the number is, the higher risk of a compound is.
hE = hERG, SG = SGOT and SGPT evaluation, Hp = hepatotoxicity, ra = acute rat
toxicity, Mu = ames positive.
1.1–25.9%. Our former study revealed that the different potency
against these two radicals may probably be attributed to the fact
that DPPH and galvinoxyl radicals belong to two different types
of radicals: DPPH is a kind of typical nitrogen radicals while
e
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
5
Table 3
In vitro DPPH free radical scavenging activity (FRSA) of the target compounds
A
5k
5k + H2O2 (200 µM)
Compd
DPPH FRSA%a
10
1.5
1.0
0.5
0.0
mixture + H2O2 (200 µM)
100
l
m
50
l
m
l
m
2
lm
1 lm
**
5a
5b
5c
5d
5f
5g
5h
5i
5k
98.0 2.1
94.9 6.0
90.4 7.9
88.6 10.1
90.9 2.8
86.3 7.2
90.3 2.7
89.2 7.2
91.9 8.0
93.2 2.0
92.2 7.0
72.9 5.1
86.4 8.3
86.8 5.9
82.0 8.6
84.5 2.9
82.8 8.0
85.7 5.2
72.0 7.6
77.3 2.3
71.4 1.7
72.0 9.2
68.2 5.0
71.3 2.1
67.4 9.1
50.9 3.9
68.9 3.5
55.0 4.1
66.6 7.9
63.1 7.2
66.0 5.1
60.1 3.0
63.1 5.5
55.0 7.7
59.2 5.5
54.0 7.7
47.9 1.8
56.2 6.0
40.1 2.2
54.5 4.2
52.4 8.7
49.1 7.7
44.7 4.3
44.4 2.1
45.1 4.7
33.1 4.6
34.8 3.3
38.6 1.6
44.1 5.1
24.7 1.6
25.9 2.0
30.3 2.9
41.8 4.2
31.6 2.1
30.3 6.0
**
**
*
#
)
ol
r
M
M
M
M
M
t
μ
μ
μ
μ
μ
n
o
1
0
1k + FA
FA
10
50
C
0.
1.0
20
(
2O2
H
a
Data are the mean values SEM (n = 3).
B
5k
5k + Aβ(5 μM)
mixture + Aβ(5 μM)
galvinoxyl radicals belong to reactive oxygen species. From a
chemical point of view, DPPH radicals are more reactive than galvi-
noxyl radicals, and more easily captured by phenolic hydroxy
groups which are responsible for the antioxidant activity of FA.4
As for the mixture of 1k and FA, it was found that the performance
of the mixture was similar to FA, but showed a relative low FRSA as
compared with 5k. As far as the R1 and R2 groups are concerned,
they seem have little effect on the antioxidant activity of the target
compounds. When the substituents were changed, no obvious
change of the FRSA was observed. This finding suggested that the
antioxidant effect of the target compounds may originate from
the FA moiety.
1.5
1.0
0.5
0.0
**
**
**
#
l
)
o
r
M
M
M
M
M
t
μ
μ
μ
μ
μ
n
o
1
5
(
10
50
C
0.
1.0
β
A
Figure 3. The neuroprotective effect of 5k and its corresponding mixture (1k + FA,
molar ratio: 1:1) against the H2O2 (A) or Ab42 (B) induced toxicity. #p <0.01 versus
2.5. Neuroprotective effect
control, *p <0.05 versus Ab (5 M), **p <0.01 versus Ab (5
l lM), n = 3.
Both FA and the carbazole derivatives have been reported to
possess neuroprotective effects.3,4 Given the newly synthesized
compounds are hybrid compounds from FA and carbazoles, it
may be interesting to investigate whether these hybrid compounds
could protect neurons from the exogenous toxins. Thus, using H2O2
or Ab42 as the toxins, the neuroprotective effect of compound 5k
was determined in vitro by MTT assay.4,10 For the comparison,
the protective effect of the mixture of 1k and FA (molar ratio:
1:1) was also measured. The results were shown in Figure 3. When
PC12 cells were incubated with different concentrations of com-
manner, indicating compound 5k could effectively protect the neu-
rons from the exogenous toxin. It was noticed that the protective
effect of compound 5k against the H2O2-induced toxicity was
higher than that against Ab42-induced toxicity. H2O2 is well
known for its oxidative damage while the toxicity caused by
Ab42 is more complex, which includes the generation of abnor-
mally high concentration of reactive oxygen species, activating
the release of damaging cytokines (e.g., interleukin-1, inter-
pound 5k (0.1, 1, 10, 50
as compared with the vehicle group. In contrast, when the cells
were treated with H2O2 (200 M) or Ab42 (5 M), a significant tox-
lM) for 24 h, no cytotoxicity was observed
leukin-6, TNF-a
), and causing mitochondrial dysfunctions.11 The
protective effect of 5k against Ab42-induced toxicity may originate
from the carbazole part as our former study revealed this scaffold
could inhibit the aggregation of Ab and exert neuroprotective
effect. In contrast, the high protective potency against the H2O2-
induced toxicity suggested that the antioxidant property, which
was mainly contributed to the FA moiety, also played important
roles in the neuroprotective effect of the target compound.
As far as the mixture of 1k and FA (molar ratio: 1:1) was con-
cerned, only a slight protective effect against the Ab42-induced
toxicity was observed. In the H2O2 assay, the mixture showed no
obvious difference from the vehicle group. These results strongly
support that the combination of equimolar parts of FA and car-
bazole derivative in one hybrid molecule is clearly superior to a
simple mixture of both parts.
l
l
icity to PC12 cells was observed as the viability of the cells dropped
from 1.0 to 0.42 and 0.54, respectively. Interestingly, when PC12
cells were co-treated with compound 5k (0.1, 1, 10, 50 lM), the
induced toxicity was significantly alleviated in a dose-dependent
Table 4
In vitro galvinoxyl free radical scavenging activity (FRSA) of the target compounds
Compd
Galvinoxyl FRSA%a
10
100
lm
20
lm
l
m
2
lm
1 lm
5a
5b
5c
5d
5f
5g
5h
5i
5k
87.2 9.0
91.9 6.7
87.4 11.7
88.6 6.3
90.8 9.7
86.3 7.0
90.3 7.6
87.2 8.0
83.2 6.7
83.2 7.0
79.0 7.6
51.7 7.5
78.2 3.9
66.8 8.1
72.0 7.7
87.5 8.4
82.8 7.2
85.7 3.9
42.0 3.7
57.0 5.5
51.4 5.2
52.2 7.0
23.4 4.2
40.3 4.8
37.4 6.2
50.9 5.9
68.0 7.3
55.0 3.2
66.6 6.7
23.1 4.0
38.1 5.0
30.1 2.7
33.7 1.7
8.0 3.1
1.1 0.7
29.2 3.1
24.0 5.5
40.9 5.0
55.2 7.0
40.1 5.7
54.5 5.4
12.4 2.7
22.4 1.7
14.7 3.0
15.8 2.0
11.1 1.2
14.8 2.7
25.6 1.5
24.1 2.7
24.7 2.0
25.9 1.4
3.3 0.7
8.3 1.1
11.0 1.7
6.6 0.3
3. Conclusion
In summary, a series of FA–carbazole hybrid compounds were
designed and synthesized with the hope to achieve a synergic action
from the antioxidant FA moiety and the ChE inhibitory carbazole
moiety. In vitro ChE assay revealed that most of the hybrid com-
pounds showed moderate inhibitory activity against AChE as well
as BChE. Particularly, the potency of the inhibition activity of
1k + FA
FA
a
Data are the mean values SEM (n = 3).
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

6
L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
compound 5k was even higher than that of the positive control
galantamine. The SAR analysis demonstrated that the electron-with-
drawing substituent on the carbazole template and the small size of
the alkyl substituent on the side chain were beneficial for the ChE
inhibition activity. In addition, the target compounds showed pro-
nounced antioxidant ability and neuroprotective property, espe-
cially against the H2O2-induced toxicity. Altogether, the results
suggest that the FA–carbazole hybrid compounds, in particular com-
pound 5k, can be considered as potential therapeutic agents for AD.
CDCl3): d 1.26(d, J = 6.9 Hz, 6H), 3.57 (s, 3H), 3.91 (s, 3H), 4.15–
4.17 (m, 1H), 5.22 (s, 2H), 6.46 (d, J = 15.2, 1H), 6.60–6.62
(m, 1H), 6.70 (d, J = 7.6 Hz, 1H), 6.78 (d, J = 7.4 Hz, 1H), 6.93 (d,
J = 7.4 Hz, 1H), 6.99 (s, 1H), 7.13 (s, 1H), 7.30–7.33 (m, 2H), 7.70
(d, J = 15.2 Hz, 1H), 7.98 (d, J = 7.6 Hz, 1H), 8.37 (s, 1H). HRMS
(ESI) m/z Calcd for
C
27H27N2O4 [MÀH]À 443.19711 found
443.19786.
4.1.4.2. N-Ethyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-meth-
oxy-9H-carbazol-4-yl)methyl)acryl-amide (5b). White solid.
Yield 90%. mp 102–104 °C. IR (KBr, cmÀ1): 3271 (OH); 1642
(C@O); 1609, 1581, 1513 (Ar, C@C). 1HNMR (300 MHz, CDCl3): d
1.05 (t, J = 6.8 Hz, 3H), 3.59 (s, 3H), 3.93 (s, 3H), 4.30 (q,
J = 6.8 Hz, 2H), 5.83 (s, 2H), 6.59 (d, J = 14.1 Hz, 1H), 6.64 (s, 1H),
6.71–6.80 (m, 2H), 6.89–7.02 (m, 3H), 7.33–7.39 (m, 2H), 7.67 (d,
J = 14.1 Hz, 1H), 7.99 (d, J = 7.2 Hz, 1H), 8.64 (s, 1H). HRMS (ESI)
m/z Calcd for C26H25N2O4 [MÀH]À 429.18146 found 429.18129.
4. Experimental protocols
4.1. Chemistry
4.1.1. Materials and instruments
Melting points were determined using a capillary apparatus
(RDCSY-I). All of the compounds synthesized were purified by col-
umn chromatography on silica gel 60 (200–300 mesh). IR spectra
were measured on KBr pellets with a Nicolet IR200 FTIR spectrom-
4.1.4.3. N-Cyclopentyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-
eter, and 1H NMR spectra were recorded with
a Bruker
methoxy-9H-carbazol-4-yl)methyl)acryl-amide (5c).
White
solid. Yield 90%. mp 113–115 °C. IR (KBr, cmÀ1): 3239 (OH); 1639
(C@O); 1608, 1579, 1512 (Ar, C@C). 1HNMR (300 MHz, CDCl3): d
1.56–1.77 (m, 8H), 3.44 (s, 3H), 3.82 (s, 3H), 4.70 (m, 1H), 5.29 (s,
2H), 6.54–6.58 (m, 2H), 6.67 (m, 2H), 6.96–7.10 (m, 5H), 7.77 (d,
J = 15.3 Hz, 1H), 8.01 (d, J = 7.5 Hz, 1H), 9.43 (s, 1H). HRMS (ESI)
m/z Calcd for C29H29N2O4 [MÀH]À 469.21256 found 469.21298.
300/500 MHz spectrometer. Mass spectra were measured on a
Bruker Esquire ESIMS instrument. The purity of all tested
compounds was characterized by high resolution mass spectrum
(Angilent technologies LC/MSD TOF). Individual compounds with
the purity of >95% were used for other experiments.
4.1.2. 3-(4-Ethoxycarbonyloxy)-3-methoxyphenyl)acrylic acid (2)
4.1.4.4.
N-Cyclohexyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-
(5d). Yellow
Compound
2 was synthesized using a previous reported
method.1 Generally, to a solution of ferulic acid (5 mmol) in
20 mL 1 M NaOH aqueous solution, ethyl chloridocarbonate
(6 mmol) was added dropwise at 5 °C. The mixture was allowed
to react for 6 h with stirring at 5 °C. The pH of the mixture was
adjusted to 2 with 3 M HCl solution, and white precipitate was
formed. Filter, wash with water, and then recrystallize the crude
product with acetone to obtain white crystal. Yield 61%, mp
157–158 °C (lit.1 mp 157–160 °C).
methoxy-9H-carbazol-4-yl)methyl)acryl-amide
solid. Yield 86%. mp 96–98 °C. IR (KBr, cmÀ1): 3403 (NH); 3256
(OH); 1639 (C@O); 1609, 1579, 1513 (Ar, C@C). 1HNMR
(500 MHz, CDCl3): d 1.40–1.49 (m, 3H), 1.60–1.63 (m, 2H), 1.74
(m, 3H), 1.91 (m, 2H), 3.59 (s, 3H), 3.91 (s, 3H), 4.77–4.78 (m,
1H), 5.18 (s, 2H), 5.73 (s, 1H), 6.47 (d, J = 15.2 Hz, 1H), 6.61 (s,
1H), 6.70 (d, J = 8.1 Hz, 1H), 6.79 (d, J = 8.0 Hz, 1H), 6.93 (d,
J = 8.4 Hz, 1H), 7.00 (s, 1H), 7.14 (d, J = 5.6 Hz, 1H), 7.13 (d,
J = 5.6 Hz, 1H), 7.31–7.32 (m, 1H), 7.69 (d, J = 15.2 Hz, 1H), 7.98
(d, J = 8.6 Hz, 1H), 8.36 (s, 1H). HRMS (ESI) m/z Calcd for
4.1.3. 4-(3-Chloro-3-oxoprop-1-enyl)-2-methoxylphenyl ethyl
carbonate (3)
C
30H31N2O4 [MÀH]À 483.22841 found 483.22846.
To a solution of compound 2 (1 mmol) in 15 mL anhydrous
CH2Cl2 and 2 drops of DMF, a solution of oxalyl chloride (5 mmol)
in 15 mL anhydrous CH2Cl2 was added dropwise in 0.5 h at 5 °C
under N2 atmosphere. Then the mixture was heated to reflux for
3 h to obtain green–yellow solution and concentrated to afford
yellow needle solid which was used for the next reaction without
purification.
4.1.4.5. N-Benzyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-meth-
oxy-9H-carbazol-4-yl)methyl)acrylamide (5e). Yellow solid.
Yield 90%. mp 103–106 °C. IR (KBr, cmÀ1): 3398 (NH); 3268 (OH);
1640 (C@O); 1608, 1580, 1512 (Ar, C@C). 1HNMR (500 MHz,
CDCl3): d 3.65 (s, 3H), 3.86 (s, 3H), 4.89 (s, 2H), 5.21 (s, 2H), 5.8
(s, 1H), 6.65 (d, J = 15.1 Hz, 1H), 6.62–7.18 (m, 7H), 7.27–7.33 (m,
6H), 7.72 (d, J = 15.1 Hz, 1H), 7.85 (d, J = 8.6 Hz, 1H), 8.39 (s,1H).
HRMS (ESI) m/z Calcd for C31H28N2O4 [MÀH]À 491.19711 found
491.19730.
4.1.4. General procedure for the preparation of compounds 5a–k
To a solution of compounds 1a–k (0.8 mmol) in 30 mL anhy-
drous CH2Cl2 and pyridine (1.2 mmol), the solution of flesh com-
4.1.4.6.
N-Isopropyl-3-(4-hydroxy-3-methoxyphenyl)-N-((9H-
pound
3 (1 mmol) in 30 mL anhydrous CH2Cl2 was added
carbazol-4-yl)methyl)acrylamide (5f). White solid. Yield 90%.
mp 98–100 °C. IR (KBr, cmÀ1): 3268 (OH); 1641 (C@O); 1605,
1580, 1512 (Ar, C@C). 1HNMR (500 MHz, CDCl3): d 1.25 (d,
J = 6.9 Hz, 6H), 3.56 (s, 3H), 5.15 (m, 1H), 5.22 (s, 2H), 5.72 (s,
1H), 6.46 (d, J = 15.1 Hz, 1H), 6.60 (s, 1H), 6.72 (d, J = 7.5 Hz, 1H).
6.78 (d, J = 7.5 Hz, 1H), 6.80 (m, 1H), 7.38 (m, 1H), 7.41 (d,
J = 8.0 Hz, 1H), 7.49 (t, J = 7.5 Hz, 8.0 Hz, 2H), 7.54 (d, J = 7.7 Hz,
1H), 7.70 (d, J = 15.1 Hz, 1H), 8.14 (d, J = 7.9 Hz, 1H), 8.44 (s, 1H).
HRMS (ESI) m/z Calcd for C26H25N2O3 [MÀH]À 413.18654 found
413.18676.
dropwise in 0.5 h at 5 °C. The mixture was stirred for another 3 h
at room temperature, concentrated in vacuo and the crude product
was purified by column chromatography (CH2Cl2/EtOAc = 5:1) to
give the compounds 4a–k. Then compounds 4a–k was dissolved
in a mixture of 50 mL of 95% ethanol and 200 lL of aminoethanol
and stirred for 12 h at 25 °C. The mixture was concentrated in vacuo,
and the residue was diluted with 30 mL water, extracted with
CH2Cl2 (20 mL Â 3). The organic layer was dried over anhydrous
Na2SO4, filtered, and concentrated to afford compounds 5a–k.
4.1.4.1.
N-Isopropyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-
4.1.4.7.
N-Ethyl-3-(4-hydroxy-3-methoxyphenyl)-N-((9H-car-
methoxy-9H-carbazol-4-yl)methyl)acrylamide (5a). Pale yel-
low solid. mp 98–99 °C. Yield 88%. IR (KBr, cmÀ1): 3207 (OH);
1640 (C@O); 1608, 1579, 1512 (Ar, C@C). 1HNMR (500 MHz,
bazol-4-yl)methyl)acrylamide (5g). Yellow solid. Yield 82%. mp
110–112 °C. IR (KBr, cmÀ1): 3405 (NH); 3270 (OH); 1642 (C@O);
1606, 1581, 1512 (Ar, C@C). 1HNMR (500 MHz, CDCl3): d 1.09(t,
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
7
J = 6.8 Hz, 3H), 3.58 (s, 3H), 4.39 (q, J = 6.8 Hz, 2H), 5.31 (s, 2H), 6.52
(d, J = 15.2 Hz, 1H), 6.63 (d, J = 1.5 Hz, 1H), 6.72 (d, J = 8.2 Hz, 1H),
6.80 (dd, J = 8.2 Hz, 1.5 Hz, 1H), 7.32 (m, 2H), 7.41 (d, J = 8.0 Hz,
1H), 7.48 (t, J = 7.45 Hz, 8.0 Hz, 2H), 7.53 (d, J = 8.1 Hz, 1H), 7.62
(d, J = 15.2 Hz, 1H), 8.10 (d, J = 8.0 Hz, 1H), 8.33 (s, 1H). HRMS
10À4 M (10% EtOH in the stock solution did not influence enzyme
activity). In order to obtain an inhibition curve, at least five differ-
ent concentrations (normally 10À4–10À9 M) of the test compound
were measured at 25 °C at 412 nm, each concentration in triplicate.
For buffer preparation, 1.36 g of potassium dihydrogen phosphate
(10 mmol) were dissolved in 100 mL of water and adjusted with
KOH to pH = 8.0 0.1. Enzyme solutions were prepared to give
2.5 units mLÀ1 in 1.4 mL aliquots. Furthermore, 0.01 M DTNB solu-
tion, 0.075 M ATC and BTC solutions, respectively, were used. A
(ESI) m/z Calcd for
C
25H23N2O3 [MÀH]À 399.17089 found
399.17066.
4.1.4.8. N-Cyclopentyl-3-(4-hydroxy-3-methoxyphenyl)-N-((9H-
carbazol-4-yl)methyl)acrylamide (5h). Yellow solid. Yield
cuvette containing 3.0 mL of phosphate buffer, 100
respective enzyme, and 100 L of the test compound solution
was allowed to stand for 5 min, then 100 L of DTNB were added,
and the reaction was started by addition of 20 L of the substrate
lL of the
91%. mp 115–117 °C. IR (KBr, cmÀ1): 3394 (NH); 3240 (OH);
1640 (C@O); 1605, 1580, 1512 (Ar, C@C). 1HNMR (500 MHz,
CDCl3): d 1.58–1.66 (m, 8H), 3.54 (s, 3H), 4.70 (m, 1H), 5.22 (s,
2H), 5.72 (s, 1H), 6.46 (d, J = 14.9 Hz, 1H), 6.59 (d, J = 1.5 Hz, 1H),
6.70 (d, J = 8.2 Hz, 1H), 6.77 (dd, J = 8.2 Hz, 1.5 Hz, 1H), 7.30 (m,
2H), 7.42 (d, J = 8.0 Hz, 1H), 7.47 (t, J = 7.45 Hz, 8.0 Hz, 2H), 7.51
(d, J = 8.1 Hz, 1H), 7.68 (d, J = 14.9 Hz, 1H), 8.12 (d, J = 8.0 Hz, 1H),
8.45 (s, 1H). HRMS (ESI) m/z Calcd for C28H27N2O3 [MÀH]À
439.20219 found 439.20253.
l
l
l
solution (ATC/BTC). The solution was mixed immediately, and
exactly 2 min after substrate addition the absorption was mea-
sured. For the reference value, 100
compound solution. For determining the blank value, additionally
100 L of water replaced the enzyme solution. The inhibition curve
lL of water replaced the test
l
was obtained by plotting the percentage enzyme activity (100% for
the reference) versus logarithm of test compound concentration.
4.1.4.9. N-Cyclohexyl-3-(4-hydroxy-3-methoxyphenyl)-N-((9H-
carbazol-4-yl)methyl)acrylamide (5i).
Yellow solid. Yield
4.2.2. Antioxidant assay in vitro
92%. mp 112–114 °C. IR (KBr, cmÀ1): 3403 (NH); 3263 (OH);
1641 (C@O); 1606, 1579, 1512 (Ar, C@C). 1HNMR (500 MHz,
CDCl3): d 1.41–1.48 (m, 2H), 1.60–1.63 (m, 3H), 1.74 (m, 3H),
1.91–1.93 (m,2H), 3.57 (s, 3H), 4.78 (m, 1H), 5.25 (s, 2H), 5.68 (s,
1H), 6.47 (d, J = 15.2 Hz, 1H), 6.61 (s, 1H), 6.70 (d, J = 8.2 Hz, 1H),
6.79 (d, J = 7.9 Hz, 1H), 7.18 (s, 1H), 7.33 (t, J = 7.3 Hz, 7.0 Hz, 1H),
7.37 (m, 2H), 7.47 (t, 7.1 Hz, 7.9 Hz, 1H), 7.54 (d, J = 7.9 Hz, 1H),
7.70 (d, J = 15.2 Hz, 1H), 8.13 (d, J = 7.9 Hz, 1H), 8.42 (s, 1H). HRMS
The experiments of FA and its hybrids to trap galvinoxyl radi-
cals and DPPH radicals were performed following a previous
report.4 DPPH radical (0.1 mM) and galvinoxyl radical (2
lM) were
dissolved in ethanol to record the absorbance (Abs0) at 517 and
428 nm, respectively. The concentration ranges of the ethanol solu-
tion of the tested compounds are from 1 to 100 lM. The absor-
bance (Abst) of the mixtures became stable after the tested
compounds were added to DPPH radical for 4 h and to galvinoxyl
radical for 19 h. The percentages of DPPH radical and galvinoxyl
radical scavenged by the tested compounds were calculated by
(1 À Abst/Abs0) Â 100.
(ESI) m/z Calcd for
C
29H29N2O3 [MÀH]À 453.21784 found
453.21748.
4.1.4.10. N-Benzyl-3-(4-hydroxy-3-methoxyphenyl)-N-((9H-car-
bazol-4-yl)methyl)acrylamide (5j).
Yellow solid. Yield 88%.
4.2.3. Neuroprotective activity in vitro
mp 90–92 °C. IR (KBr, cmÀ1): 3253 (NH, OH); 1641 (C@O); 1582,
1511 (Ar, C@C). 1HNMR (300 MHz, CDCl3): d 3.58 (s, 3H), 4.99 (s,
2H), 5.34 (s, 2H), 5.48 (s, 1H), 6.68–6.90 (m, 5H), 7.03–7.16 (m,
3H), 7.36–7.51 (m, 6H), 7.85–7.95 (m, 3H), 9.17 (s, 1H). HRMS
PC12 cells, obtained from the Cell Bank of the Chinese Academy
of Sciences (Shanghai, China), were cultured in a humidified, 5%
CO2 atmosphere at 37 °C, and maintained in monolayer culture
in F-12 medium supplemented with 10% fetal bovine serum
(FBS), 100 mg/mL of streptomycin and 100 mg/mL of penicillin.
The cell viability was determined by MTT assay (MTT: 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide). For the
preparation of Ab42 solution, Ab42 (Chinapeptides Co., Ltd) was
initially dissolved to 1 mM in hexafluoroisopropanol (Sigma) and
separated into aliquots in sterile microcentrifuge tubes. Hexafluo-
roisopropanol was removed under vacuum in a Speed Vac, and the
peptide film was stored desiccated at À20 °C. F-12 culture medium
(ESI) m/z Calcd for
461.18621.
C
30H25N2O3 [MÀH]À 461.18654 found
4.1.4.11. N-Ethyl-3-(4-hydroxy-3-methoxyphenyl)-N-((7-chloro-
9H-carbazol-4-yl)methyl)acrylamide (5k). Yellow solid.
Yield 88%. mp 90–92 °C. IR (KBr, cmÀ1): 3255 (OH); 1655 (C@O);
1580, 1515 (Ar, C@C). 1HNMR (300 MHz, CDCl3):
1.09 (t,
J = 6.9 Hz, 3H), 3.59 (s, 3H), 3.91 (s, 3H), 4.06 (q, J = 6.9 Hz, 2H),
5.22 (s, 2H), 6.49 (d, J = 15.2, 1H), 6.60–6.62 (m, 1H), 6.70 (d,
J = 7.6 Hz, 1H), 6.78 (d, J = 7.4 Hz, 1H), 6.93 (d, J = 7.4 Hz, 1H),
6.99 (s, 1H), 7.13 (s, 1H), 7.56–7.59 (m, 2H), 7.78 (d, J = 15.2 Hz,
1H), 7.99 (d, J = 7.6 Hz, 1H), 8.47 (s, 1H). HRMS (ESI) m/z Calcd
for C25H22ClN2O3 [MÀH]À 433.13191 found 433.13166.
d
was added to bring the peptide to a final concentration of 100 lM
and incubated at 4 °C for 24 h. The tested compounds were dis-
solved in DMSO and diluted to the required concentration with
culture medium (DMSO final concentration <0.5%). The suspension
of 2000 cells/well was plated in 96-well culture plates with culture
medium and was incubated for 24 h at 37 °C, in a 5% CO2 incubator.
4.2. Biological studies
Then 24
ranging between 0.1
solution (final concentrations ranging between 0.1
containing Ab42 (final concentration 5 M) or H2O2 (final concen-
tration 200 M) was added. Cells were incubated at 37 °C for 24 h.
respectively. After that, the cells were treated with 10 mL MTT dye
solution (5 mg/mL) for 4 h cultivation. The media with MTT solu-
tion were removed with 100 mL of DMSO solution. The absorbance
of formazane solution was measured at 540 nm with an automatic
microplate ELISA reader. Wells without cells were used as blanks
and were subtracted as background from each sample. Results
were expressed as a percentage of control. Three independent tri-
als were analyzed and the results were expressed as mean SEM.
l
L of tested compound solution (final concentrations
M and 50 M), or 24 L of tested compound
M and 50 M)
l
l
l
4.2.1. Cholinesterase inhibition assay in vitro
l
l
The cholinesterase inhibitory activity of the target compounds
was measured using Ellman’s assay.2 AChE (E.C.3.1.1.7, Type VI-S,
from Electric Eel) and BChE (E.C.3.1.1.8, from equine serum) were
purchased from Sigma–Aldrich (Steinheim, Germany). DTNB
(Ellman’s reagent), ATC and BTC iodides were obtained from Fluka
(Buchs, Switzerland). The assay was performed as described in the
following procedure: stock solutions of the test compounds were
l
l
prepared in ethanol, 100 lL of which gave a final concentration
of 10À3 M when diluted to the final volume of 3.32 mL. The highest
concentration of the test compounds applied in the assay was
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010

8
L. Fang et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
4.2.4. Molecular modeling and ADME prediction
Research Funds for the Central Universities (No. 3207045413).
Dr. Fang is thankful to the support of the Open Project Program
of State Key Laboratory of Natural Medicines, China Pharmaceuti-
cal University.
Homology modeling was carried out using the MOE (Molecular
Operating Environment) software (Chemical Computing Group
Inc). Initial minimization was performed within the homology
modeling function of MOE. The model from MOE was minimized
with a few thousand cycles of minimization using the ABNR
(adopted-basis Newton–Raphson) method. Ligands were modeled
by positioning them in the active site in accordance with the pub-
lished crystal structures (PDB code: 1P0P and 3I6M). The entire
complex was then subjected to alternate cycles of minimization
and dynamics. Each dynamics run was short, about 3 ps. The intent
was to get a satisfactory structure for the complex that was
consistent with the published crystal structure. For the ADME
and Toxicity prediction, the compound was firstly sketched using
discovery studio 3.0 and then saved as sd format. It was then
imported into ADMET predictor 7.0 (Simulation plus, USA) for
ADME and toxicity prediction. Calculation was performed under
pH = 7.4. Other parameters were set as default. The results were
exported into an sd file for further reading.
References and notes
1. Cacabelos, R. Methods Mol. Biol. 2008, 448, 213.
2. Fang, L.; Gou, S.; Fang, X.; Cheng, L.; Fleck, C. Mini Rev. Med. Chem. 2013, 13, 870.
3. Yang, W.; Wong, Y.; Ng, O. T.; Bai, L. P.; Kwong, D. W.; Ke, Y.; Jiang, Z. H.; Li, H.
W.; Yung, K. K.; Wong, M. S. Angew. Chem., Int. Ed. 2012, 51, 1804.
4. Fang, X.; Fang, L.; Gou, S.; Lupp, A.; Lenhardt, I.; Sun, Y.; Huang, Z.; Chen, Y.;
Zhang, Y.; Fleck, C. Eur. J. Med. Chem. 2014, 76, 376.
5. Markesbery, W. R.; Carney, J. M. Brain Pathol. 1999, 9, 133.
6. Picone, P.; Bondi, M. L.; Montana, G.; Bruno, A.; Pitarresi, G.; Giammona, G.; Di
Carlo, M. Free Radical Res. 2009, 43, 1133.
7. Saxena, A.; Redman, A. M. G.; Jiang, X.; Lockridge, O.; Doctor, B. P. Biochemistry
1997, 36, 14642.
8. Greenblatta, H. M.; Krygera, G.; Lewis, T.; Silman, I.; Sussman, J. L. FEBS Lett.
1999, 463, 321.
9. Feng, J. Y.; Liu, Z. Q. J. Agric. Food Chem. 2009, 57, 11041.
10. Shi, Y. F.; Zhang, H. Y.; Wang, W.; Fu, Y.; Xia, Y.; Tang, X. C.; Bai, D. L.; He, X. C.
Acta Pharmacol. Sin. 2009, 30, 1195.
11. Jakob-Roetne, R.; Jacobsen, H. Angew. Chem., Int. Ed. 2009, 48, 3030.
Acknowledgments
This work is supported by Natural Science Foundation of
Jiangsu Province, China (No. BK20151402) and the Fundamental
Please cite this article in press as: Fang, L.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.01.010