ROS-responsive and multifunctional anti-Alzheimer prodrugs: Tacrine-ibuprofen hybrids via a phenyl boronate linker

Article abstract of DOI:10.1016/j.ejmech.2020.112997

Current drugs available in clinic for Alzheimer's disease (AD) treatment can only alleviate disease symptoms without clearly curing or delaying the process of AD. And some AD drugs failed in Phase III clinical trials are only focused on targeting amyloid-

Full text of DOI:10.1016/j.ejmech.2020.112997

Journal Pre-proof
ROS-Responsive and Multifunctional Anti-Alzheimer Prodrugs: Tacrine-Ibuprofen
Hybrids via a Phenyl Boronate Linker
Zhikun Liu, Bin Zhang, Shengjin Xia, Lei Fang, Shaohua Gou
PII:
S0223-5234(20)30969-7
DOI:
https://doi.org/10.1016/j.ejmech.2020.112997
EJMECH 112997
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 24 September 2020
Revised Date: 26 October 2020
Accepted Date: 3 November 2020
Please cite this article as: Z. Liu, B. Zhang, S. Xia, L. Fang, S. Gou, ROS-Responsive and
Multifunctional Anti-Alzheimer Prodrugs: Tacrine-Ibuprofen Hybrids via a Phenyl Boronate Linker,
European Journal of Medicinal Chemistry, https://doi.org/10.1016/j.ejmech.2020.112997.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Masson SAS.

Graphical abstract
ROS-Responsive and Multifunctional Anti-Alzheimer Prodrugs:
Tacrine-Ibuprofen Hybrids via a Phenyl Boronate Linker
a,b,1
a,b,1
a,b
a,b *
a,b *
Zhikun Liu , Bin Zhang , Shengjin Xia , Lei Fang , Shaohua Gou
A series of ROS-responsive and multi-functional anti-Alzheimer compounds
were designed and synthesized. The representative compound 22 possessed
pronounced cholinesterase-inhibiting, anti-inflammatory and neuron-protective effects
in vitro, and perfectly ameliorated the cognitive damage of AD in vivo. The study
offers a promising strategy for developing new anti-Alzheimer drugs.

ROS-Responsive and Multifunctional Anti-Alzheimer Prodrugs:
Tacrine-Ibuprofen Hybrids via a Phenyl Boronate Linker
a,b,1
a,b,1
a,b
a,b *
a,b *
Zhikun Liu
, Bin Zhang ,Shengjin Xia , Lei Fang , Shaohua Gou
a
Jiangsu Province Hi-Tech Key Laboratory for Biomedical Research, Southeast
University, Nanjing, 211189, China
b
Pharmaceutical Research Center and School of Chemistry and Chemical
Engineering, Southeast University, Nanjing 211189, China
ABSTRACT Current drugs available in clinic for Alzheimer’s disease (AD)
treatment can only alleviate disease symptoms without clearly curing or delaying the
process of AD. And some AD drugs failed in Phase III clinical trials are only focused
on targeting amyloid-β (Aβ). Therefore, an alternative strategy in AD drug design is
meaningful to be involved in the multiple pathogenic factors which can affect each
other at multiple levels. Herein, we report a series of ROS-responsive prodrugs based
on multi-target-directed ligands (MTDLs) approach, which can specifically release
tacrine derivatives and ibuprofen under oxidation of ROS and show
acetylcholinesterase (AChE)-inhibiting, neuron-protective and anti-inflammatory
effects in extracellular or intracellular assays. Related biological study illustrated that
compound 22 was able to permeate blood-brain-barrier (BBB) showing little
hepatotoxicity in comparison to tacrine. Besides, 22 hinted a therapeutic clue in
AD-treatment by regulating proinflammatory factors (IL-1β and TNF-α) and
apoptosis related proteins (Bax, Bcl-2 and cleaved caspase-3). Further spatial memory
assays in Aβ-induced AD model showed that 22 enhanced the ability of learning and
memory. Our study proves that the strategy of ROS-responsive prodrugs has promise
for AD treatments in future and offers a way for AD drug development.
1
These authors contributed equally.
*
Corresponding authors
E-mail: lei.fang@seu.edu.cn (L. Fang); 2219265800@qq.com (S. Gou)

Keywords: ROS stimulation, Prodrug, Multifunctional agent, Alzheimer
1
. INTRODUCTION
Alzheimer's disease (AD), one of the most common forms of dementia, is an
age-related neurodegenerative disease characterized by progressive cognitive decline,
[
1,2]
behavior disorders and degenerative brain disease
. It is estimated that about 50
million people have been affected by AD in 2018 and the number is expected to reach
1
2
52 million costing more than $2.0 trillion for treatment in the United States along by
[3,4]
050
. Without doubt, the sustaining and high-speed growth of AD patients will
bring about huge social and economic burden, revealing that treatment of AD is of
great significance.
Much attention has been paid to AD, but no efficient therapeutic treatment is
developed at present because of complex etiology of this neurodegenerative disease
[
5,6]
and its multifactorial progression . Several hypotheses based on AD’s pathogenic
factors have been proposed, including the production and aggregation of amyloid-β
(Aβ), accumulation of hyperphosphorylated τ-proteins, oxidative stress, decrease of
[
7,8]
synaptic cholinergic level and inflammation response
. Upon these hypotheses,
only a few of acetylcholinesterase inhibitors (AChEIs shown in Figure 1) including
tacrine (withdrawn from the market for hepatotoxicity), donepezil, rivastigmine,
galantamine and the N-methyl-D-aspartate receptor (NMDAR) antagonist
(memantine), were obtained for AD treatment. However, these drugs can only
alleviate the progress of AD by improving memory and cognitive function to a certain
degree instead of definitively curing or stopping this disease. Therefore, development
of novel approach to treat AD is much necessary and required. In view of the complex
pathogenesis, present research and development on AD drugs are mainly focused on
multitarget-directed ligands (MTDLs) approach by efficiently ameliorating memory
[
9-13]
and modifying cognitive decline
.
Indeed, a number of drug candidates have undergone clinical trials and most of
them aim to reduce the production of toxic aggregated Aβ or promote the
[
14-17]
depolymerization or clearance of aggregation Aβ proteins to protect neurons
.
However, with further understanding of mechanisms about AD, oxidative stress is

found to happen earlier than other pathogenesis and then affect each other at multiple
[
18-20]
levels
. When dopamine (DA) is oxidized by oxygen (O ) or upon the formation
2
of metal ion-binding Aβ protein, a large number of reactive oxygen species (ROS)
•
−
(
e.g. superoxide anion radical (O ) and hydrogen peroxide (H O ) generate and
2 2 2
subsequently induce more pathogenesis such as oxidative damage to biomolecules
(lipids, proteins, and DNA), mitochondrial dysfunction and final disorders of neuronal
[
21-23]
homeostasis in the brain
. Moreover, prominent activation of inflammatory
process (inflammatory cytokines, IL-1β, TNF-α) induced by Aβ may drive AD
pathology or become ongoing disease process, though neuroinflammation exerts both
[
24-26]
toxic and protective effects in AD process
. Alternatively, in order to reduce the
toxic side effects induced by anti-inflammatory drugs and increase neuron-protective
activities, much attention has also been paid on microenvironment-responsive prodrug
[
25, 27a]
strategies
.
In recent years, quite a number of H O -specific probes or H O -responsive
2
2
2
2
prodrugs for antitumor study have been designed and synthesized using arylboronic
[
27].
acid or ester as the protecting group
As a ROS-responsive unit, phenylboronate
group can be self-immolative upon exposure to H O , which has been widely used in
2
2
[
28-30]
both oxidation-responsive polymers and prodrugs
. Tacrine, the first AChEI
approved for AD treatment, has been withdrawn from clinical use, but we believe that
it is still of great significance for further investigation on tacrine derivatives to reduce
[
6, 31,32]
hepatotoxicity
. Inkielewicz-Stepniak and co-workers have explained the great
therapeutic potential of tacrine in Alzheimer’s disease via multifunctional ligands
[32c]
modification
. In addition, nonsteroidal anti-inflammatory drugs (NSAIDs) like
ibuprofen or aspirin were reported to reduce inflammatory markers and reverse spatial
memory deficits in APPsw transgenic mice or improve memory and learning, and
[
25,33,34]
decrease stress-related behaviors in FAD5X/Ppara-null mice
. Therefore, the
great therapeutic potential effects of tacrine and ibuprofen motivate us to develop a
novel agent to simultaneously tackle the cognitive impairment caused by Alzheimer's
disease. Much different from ‘one molecule, one target’ paradigm, MTDLs approach
usually offers more than one functional biological activities, which is suitable for

complex pathogenesis of AD. Besides, MTDLs approach can also offer an ample
variety of derivatives in which key structure or linker length can modulate their
biological properties. Moreover, considering the overproduction of ROS in brain of
AD and specific characteristics of local inflammation, we have constructed
self-immolative dendritic prodrugs consisting of two ligands, ibuprofen and tacrine
derivatives, based on the MTDLs with a concept mentioned above.
As illustrated in Figure 2, we pursue to develop ROS-responsive prodrugs based
on AD’s pathological characteristics, including oxidative stress decreased synaptic
cholinergic level and inflammation response. With this aim in mind, we expect these
prodrugs release those two ligands at specific pathological oxidation environment of
AD upon the stimulation of ROS. In order to achieve our designing goals, quite a
number of biological assays have been performed including intracellular/extracellular
tests and therapeutic effects in AD models in vivo. Herein we present such a
promising strategy for the design of AD drugs as well as significant results we have
obtained.
2
. RESULTS AND DISCUSSION
2
.1. Chemistry
Preparation of tacrine derivatives is shown in Scheme S1, in which intermediates
[
31,32a]
8
-11 were synthesized according to the previous reports
. Synthetic way of our
target compounds 19-22 is displayed in Scheme 1. Compound 13 was obtained
through modifying with two protection groups of TBDMS on 12, using imidazole as
an absorb acid agent in anhydrous N,N-dimethylformamide (DMF). Then, the
nucleophilic
substitution
-bromomethylphenylboronic acid pinacol ester produced compound 14. Intermediate
5 was obtained from deprotection reaction of 14 under the condition of catalytic
reaction
between
compound
13
and
4
1
amount of PTSA in methanol at room temperature. Single TBDMS-covered 16 was
separated with silica gel chromatography. Compound 18 was made from deprotection
of TBDMS-covered 17 followed by condensation reaction with ibuprofen.
Compounds 19-22 were obtained by modifying two ligands of 18 and tacrine

derivatives 8-11, respectively, in the presence of N,N'-carbonyldiimidazole (CDI) in
DMF.
2
.2. Cytotoxicity assay
Cytotoxicity of the target compounds (19-22) against SH-SY5Y cells was
measured by MTT method. As shown in Figure 3A, no obvious cytotoxicity was
induced by 19-22 in the concentration ranged from 7.81 to 31.25 μM. Interestingly, 19
(with a two-methylene linker) and 21 (with a four-methylene linker) exhibit moderate
cytotoxicity to SH-SY5Y cells at 31.25 μM, whereas both of them have negligible
cytotoxicity at 7.81 μM and 15.62 μM. Moreover, 20 (with a three-methylene linker)
and 22 (with a six-methylene linker) have better cellular adaptability, especially for
compound 22. Judging from the results mentioned above, 22 exhibited negligible
cytotoxicity to SH-SY5Y cells even at higher concentration of 31.25 μM as
comparison to that of control group (Figure 3A), whereas 19-21 induced little to high
cytotoxicity to some degree under this concentration. Therefore, we decided to select
2
2 to investigate the following biological activities in alleviating Alzheimer’s disease.
As reducing the damage from oxidative stress in cells is believed to be feasible to
[34b-34d]
protect the nerve cells
, we subsequently investigated neuroprotective effects of
compounds 19-22 against H O . Briefly, SH-SY5Y cells were pre-incubated with
2
2
o
1
9-22 followed by the addition of H O . After another 24 h co-incubation at 37 C,
2 2
cell viability of SH-SY5Y cells was obtained. The results in Figure 3B indicate that
these compounds at the concentration of 5 μM show moderate to high neuroprotective
activity against H O and the values are all more than 50 % but significantly increase
2
2
to more than 80 % when the concentration was elevated to 20 µM (exception for 21),
exerting stronger neuroprotective activities than that of H O -treated group (cell
2
2
viability was about 20 %). However, ibuprofen, a component of the target compound,
does not show apparent detoxification against H O . On the other hand, the combined
2
2
group of ibuprofen and 11, two components of 22, seems not to rescue the cells
against cytotoxicity induced by H O even at high concentration of 20 µM. Taken all
2
2
the results together, 22 is the most potential compound in protecting neurons. Thus,
compound 22 is selected for further biological experiments.

Encouraged by this, we further investigate the H O -scavenging capability of our
2
2
target compounds. The consuming H O after 24 h co-incubation with 22 was
2
2
quantified via QuantiChrom™ Peroxide Assay Kit. As illustrated in Figure S1A, the
amount of scavenging H O is significantly elevated in a liner relationship as the
2
2
increasing dose of 22. Trolox, a positive control, also shows strong H O -cosuming
2
2
capacity (Figure S1A). In contrast, 11 and ibuprofen only show negligible
H O -eliminating capability as compared with 22 and Trolox. All the illustrated
2
2
results mentioned above revealed that compound 22 can protect neurons from
apoptosis against H O and be beneficial for decrease of oxidation stress of
2
2
Alzheimer.
2
.3. The release of ibuprofen and tacrine derivative 11
The target compound, especially compound 22, induces no cytotoxicity to
SH-SY5Y cells (Figure 3A) but seems to be more potential in protecting neurons
from apoptosis against the ROS (Figure 3B). Motivated by this, we evaluated the
stability of compound 22 under the oxidation of H O . A mixed solution of 22 and
2
2
o
H O at 37 C was used to detect the stability of 22 by reverse phase high
2
2
performance liquid chromatography (RP-HPLC). For the purpose of matching the
peaks generating from 22, solutions of ibuprofen, 11 and 12 at certain concentrations
were prepared. As shown in Figure S1, the peak attributed to 22 in RP-HPLC
appeared at about 12.381 min (Figure S1F). However, when treated with H O , the
2
2
peak of 22 gradually decreased with the generation of other three peaks matching the
position of compound 11, compound 12 and ibuprofen, respectively. This
phenomenon indicated that 22 acted as a ROS-responsive prodrug. Moreover, the
reaction rate did not accelerate a lot during the first 5 hours until the addition of
another portion of H O , hinting that amount of H O was scavenged with
2
2
2
2
self-immolation of 22 (Figure S1F).
The results in Figure S1F indicated that 22 degrades into the fragments of 11, 12
and ibuprofen gradually within 10 h upon stimulation with H O . Therefore, we
2
2
speculated that the envisaged collapse mechanism of 22 under oxidative condition
was as follows (Scheme S3): Firstly, phenylboronic ester reacted with H O to
2
2

generate phenyl oxide structure. Subsequently, the electron rearrangement around the
aromatic structure and finally caused rapidly release of the two ligands. In order to
1
substantiate this speculation, the process was also monitored with H-NMR
characterization. As shown in Figure S2, initially, there existed no peaks from
4
.10-4.60 ppm where were methylene protons of benzyl alcohols. However, when
o
mixed with H O (6.0 equiv.) at 37 C, methylene protons of 22 (a, b) gradually
2
2
decreased along with another increased peaks at 4.50 and 4.36 ppm (a′, b′, Figure S2).
These two peaks (a′, b′) belong to the methylene protons of phenylboronate and cresol
units, indicating the collapse of 22 under the oxidative condition of H O .
2
2
Interestingly, the pinacol unmasking reaction was not detected may due to the fast
1
degradation of this intermediate. Moreover, the significant changes in H-NMR peaks
of e′-f′ and c also proved the hydrolysis of the benzyl ester bonds and produce of
fragments 11 and ibuprofen under the oxidative condition of H O . In sum,
2
2
considering the results in Figure S1F, we can conclude that our target compound 22
served as a ROS-responsive prodrug and obtained great potential in treatment of
Alzheimer’s disease.
2
.4. Blood-brain barrier (BBB) permeation and cellular uptake and assays
[
35a]
Brain permeation abilities are initially required for development of AD drugs
.
Therefore, in order to investigate the ability of 22 to penetrate the BBB we
constructed a widely adopted BBB model by growing a dense layer of bEnd.3 cells on
[
35b]
the upper chamber of a transwell
. Briefly, bEnd.3 cells are cultured and allowed
to adhere in monolayer in DMEM (10% FBS) and transendothelial electrical
resistance (TEER) values of cells are monitored until the TEER values reach about
2
1
80 Ω
•
cm and used for BBB assays. The schematic of BBB model is given in Figure
S3A while change of TEER values during the monolayer culture is showed in Figure
S3B. As illustrated in Figure S3C, TEER values are decreased in bEnd.3 cells
monomers upon the addition of 22 while little change in TEER value occurs in
DMSO-treated group. The decreased TEER at the first 15 min is induced by solvent,
[
36a]
which was reported previously
. Further immunofluorescence staining indicates
that 22 induces discontinued intervals or even full absence to tight-junction related

protein (ZO-1) between bEnd.3 monomer cells (Figure S3D), which reveals that 22
regulates tight junction molecules to assist 22 to permeate BBB for delivering drug to
brain. Furthermore, we also tested the content of 22 in the medium and intracellular
abilities of lower chamber via BBB model by HPLC. As illustrated in Figure S4B, 22
easily passes through monomer cells in significant dose-dependent manner and can be
easily ingested by SH-SY5Y cells in lower chamber (Figure S4C), indicating better
BBB permeation abilities and intracellular properties. In sum, all of the results
indicate that 22 efficiently permeates BBB via regulation expression of ZO-1 between
bEnd.3 cells and easily entered SH-SY5Y cells.
For further research of visual experiment in BBB permeation assays and
intracellular uptake assays, compound 25 with a fluorescent coumarin moiety at the
[
36, 37]
position of 11 was constructed according to previous reports
(Scheme S2).
Intensely florescent images in SH-SY5Y cells in the BBB model indicate that 25 can
efficiently cross through the monolayer barrier and be easily ingested by SH-SY5Y
cells (Figure S4D). In order to investigate the uptake by SH-SY5Y and BV-2 cells, 25
were co-incubated with SY-SY5Y or BV-2 cells, respectively, and cellular uptake
results were obtained from laser confocal microscope. The intensely florescent color
in Figure 4 indicates that 25 has entered either SY-SY5Y or BV-2 cells efficiently.
Considering the great BBB permeation properties and efficient intracellular uptake of
2
2 and 25 in SHSY5Y cells we conclude that 22 is in great potential for treatment of
Alzheimer’s disease.
2
.5. The AChE inhibition study
The results in Figure S1F illustrated that our target compounds act as
ROS-responsive prodrugs upon stimulation with H O . Therefore, we forecast that
2
2
AChE inhibition activity of our target compounds may vary greatly before and after
treatment with H O . In order to confirm our hypothesis, AChE (from electric eel,
2
2
eeAChE) inhibition assays were performed using tacrine and untreated-compounds
[
34, 38]
1
9-22 as controls with Ellman’s methods
. The decreased enzymatic activity
results induced by 19-22 or H O -sitimulated 19-22 are shown in Figure 5. It is noted
2
2
in Figure 5A that compounds 19-22 have moderate or no obvious AChE inhibitory

activity even at 10 μM. Besides, the inhibitory rate of AChE increased from 28.25%
19) to 47.25% (22), probably due to the extension of carbon linkers that enhanced the
(
interaction between AChE and the tacrine group. However, after co-incubation with
H O , the AChE inhibitory effects of the target compounds were enhanced apparently
2
2
with the increasing alkyl chains. As shown in Figure 5B, H O -treated 19-22 exhibit
2
2
AChE inhibitory activity with IC values in range of 39.16 - 608.1 nM, of which the
5
0
activity of compound 22 containing a six methylene linker tacrine group is the
strongest with a value comparable to that of tacrine (IC value was 21.63 nM) and 11
5
0
(
IC value was 54.58 nM). It is noticed that the length of the alkyl chain has a close
50
relationship with the AChE activity of the resulting compounds. Motivated by AChE
inhibitory activity of 22, we chose the compound 22 for the following biological
study.
2
.6. Molecular docking
The AChE inhibition study proved that our target compounds 19-22 showed little
inhibitory effects on AChE, whereas they turned out to be efficient AChE imbibition
agents upon exposure to H O . Generally, AChE was reported as dumbbell-shape with
2
2
two active sites: the catalytic active site (CAS) at the bottom and the peripheral active
[
31]
site (PAS) at the lip . To investigate the difference before and after treatment with
H O , we performed the molecular docking studies of compounds 11 and 22. The
2
2
crystal structure of torpedo californica AChE (2CKM) was obtained from RSCB
protein data bank, and the structure of 22 was constructed in Chem 3D 14.0 with
energy minimized in Sybyl-X 2.0. The results obtained from Sybyl-X 2.0
surflex-dock model were shown in Figure 6. The binding model illustrated that the
tacrine fragment of 11 bound to the bottom of AChE, is surrounded with Asp 72, Val
7
1, Tyr 70, Trp 84, Glu 199, Met 83, Gly 118 and Pro 86, exhibiting a potent π-π
interaction between the heterocycle moiety and Trp 84; while the aliphatic alkyl chain
is surrounded by Ser 200, Trp 84, Gly 441, His 440 and Try 442 (Figure 6A and 6B).
The formation of H-bond between aliphatic alkyl chain of 11 and residues Tyr 442
and Ser 200 may contribute to the inhibitory effect on AChE as well. However, the
binding model in Figure 6C and 6D suggested that 22 only occupies the mouth of the

gorge and was only surrounded with residues Gly 335, Phe 331, His 440, Asp 72, Val
1, Try 70, Asp 276 and Trp 279 without significant intermolecular interactions with
7
residues around the PAS. Taken all these docking results together, the apparent
difference of the binding models between 22 or 11 and AChE may explain why 22
shows little AChE inhibition activity excepting for stimulating with H O .
2
2
2
.7. Inflammatory cytokines assay by qRT-PCR and LPS-induced ROS accumulation
assays
Proinflammatory cytokines (such as TNF-α and IL-β) secreted by microglia play
[39-41]
an important role in the pathogenesis and development of cognitive impairment
.
Therefore, BV-2 cells were co-incubated with 22 for 6 h followed with treatment of
LPS for another 24 h. The total RNA was isolated, purified, quantified and reversely
[
25,39a]
transcripted to cDNA according to the previous report by q-PCR
. The results in
Figure 7 showed that, under stimulation of LPS at concentration of 1 μg/mL and 3
μg/mL, expression of IL-1β was significantly increased to 2.22- and 3.68-fold, while
that of TNF-α was significantly increased to 2.51- and 3.86-fold when compared to
those of the control. This clearly indicates that LPS is capable of inducing
inflammation in BV-2 cells. However, when co-incubated with 22 (5 μM, 10 μM) and
LPS (3 μg/mL) for another 24 h, the expression of IL-1β in BV-2 cells decreased to
1
.36- and 0.86-fold in a concentration-dependent manner while the expression of
TNF-α decreased to 1.60- and 1.20-fold as comparison to control group (Figure 7).
Moreover, the expression level of both IL-1β and TNF-α was decreased to the same
level as that of untreated group, revealing that compound 22 acquired the
anti-inflammatory activity by regulating relative proinflammatory cytokines (TNF-α
and IL-β). Similarly, ibuprofen also showed the same anti-inflammatory effects by
maintaining the normal level of the expression of TNF-α and IL-1β.
The results in Figure 7 indicated that 22 modulates LPS-induced IL-1β and
TNF-α to the normal levels in LPS-stimulated BV-2 cells as comparison with that of
untreated cells, which may raise a question that whether the anti-inflammation effects
come from 22 itself or from the released ibuprofen from 22 under the stimulation of
LPS. No doubtfully, LPS elevated the intracellular ROS accumulation, which

[
39b]
thereafter modulates the expression of neurotoxic factors in microglia
. Besides,
disruption of the production or accumulation of intracellular ROS provides a method
[
39c]
to investigate the proinflammatory signals in cells
. Therefore, in order to
investigate this mechanism of anti-inflammation of 22, intracellular ROS
accumulation in BV-2 cells were studied. As shown in Figure S5, the intracellular
ROS levels in BV-2 cells treated with LPS (3 μg/mL) were significantly elevated
(Figure S5A2-S5C2), labeling as more cells were stained in strong green-florescence
as comparison with untreated-cells (Figure S5A1-S5C1). However, when
co-incubated with 22 (10 µM) and LPS (3 μg/mL) for another 18 h, ROS levels were
apparently decreased, illustrating that 22 regulated oxidative stress induced by LPS
(Figure S5A3-S5C3). Interestingly, the remarkable ROS accumulation in the cells
treated with combined group of ibuprofen (10 µM) and 11 (10 µM) indicated that both
ibuprofen and 11 did not detoxify against LPS by eliminating the ROS accumulation
(Figure S5A4-S5C4), though the cells were pretreated with the two components of 22
for 6 h. Moreover, BV-cells of 6 h pretreatment of LPS (3 μg/mL) and another 18 h
co-incubation with 22 (10 µM) were also in normal ROS levels condition when
compared with that of untreated cells (Figure S5A5- S5C5), revealing that 22 also
showed neuroprotective activity in a model of oxidative stress in BV-2 cells. Taken
the anti-inflammatory of 22 in LPS-treated cells in consideration, we believed that 22
rescued cells against cytotoxicity induced by LPS not only depending on the collapse
of 22 in oxidative condition but also on its ROS-eliminating capability.
2
.8. Apoptosis detection
On the basis of the promising neurons protection activity of 22 against H O , we
2
2
subsequently performed apoptosis assays by FACS to investigate whether the
neuroprotection was attributed to apoptosis inhibition. Firstly, the apoptosis of
SH-SY5Y cells induced by 22 was studied. As shown in Figure 8A-8D, 22 did not
exert obvious cytotoxicity against the cells after 24 h co-incubation with a cell
viability value over 90%, consistent with the result observed in Figure 3. In order to
evaluate the neuroprotective effect of 22 against H O , SH-SY5Y cells were incubated
2
2
with 22 (5 μM, 10 μM) for 6 h followed by another 24 h treatment of H O (100 μM).
2
2

Similarly, cells exposed to equal concentration H O were used as positive control.
2
2
The apoptotic results are given in Figure 8E-8H. Notably, the percentage of living
cells decreased from 93.60% (DMSO-treated group) to 41.7% (H O -treated group).
2
2
In comparison with H O , SH-SY5Y cells treated with 22 exhibited high percentage
2
2
of living cells from 73.70% of 22 (5 μM) to 86.60% of 22 (10 μM). Significantly, the
potency of 22 in protecting neurons from H O induced cytotoxicity presented in a
2
2
dose-dependent manner. In sum, these findings verify the results as obtained in
Figure 3 and the neuroprotective activity of compound 22.
2
.9. Hepatotoxicity study
Tacrine was the first AChEI approved for AD treatment, but it has been
withdrawn from the clinical use due to its hepatotoxicity. However, this drug
[
31,
molecule still deserves to be investigated for non-hepatotoxic tacrine derivatives
4
0-44]
.
It is learned that the hepatotoxicity induced by tacrine mainly resulted in
elevating the levels liver transaminase, decreasing albumin concentration and
[
31, 42-44]
inducing ROS in hepatocyte
. In order to investigate whether our target
compound can decrease the hepatotoxicity, compound 22 was used to for ROS
accumulation tests with tacrine as positive control. As shown in Figure 9, in
comparison to the control group, tacrine caused significant hepatotoxicity as
illustrated with more HepG2 cells in green florescence. Interestingly, less amount of
HepG2 cells in green florescence were detected when cells were treated with 22 (25
or 50 μM) for 24 h. Generally, the higher the accumulation of ROS was present, the
[
45,46]
more serious damage was induced to the cells
. Taken all of the results mentioned
above, compound 22 caused lower hepatotoxicity at a concentration of 25 or 50 μM
than tacrine which induced serious hepatotoxicity at 50 μM.
2
.10. Cell morphology assays under H O
2 2
In order to investigate the change of cell morphology, calcein AM/PI double
staining probes were used. SH-SY5Y cells were pre-incubated with 22 (10 μM) for 6
o
h at 37 C, using DMSO as control group. Subsequently, the cells in 22-treated group
and untreated group were incubated with H O (100 μM) for another 24 h and stained
2
2
with calcein AM/PI double staining probes. As shown in Figure 10, the cells without

any treatment are in spindle outline and mostly stained in green florescence (Figure
0A1), while little cells are stained in red florescence (Figure 10B1). However, after
treated with H O most cells become oval or irregular nucleus in apoptosis or death
1
2
2
state and are stained in red florescence by the PI probe (Figure 10A2-10C2).
Moreover, after pre-treatment with 22 (Figure 10B3), less cells with red florescence
are found as compared with H O -treated group (Figure 10A2-10C2), illustrating that
2
2
2
2 shows perfect neuroprotective activities against the apoptosis induced by H O .
2 2
Besides, the neuroprotective effects also confirmed the better anti-apoptosis activity
of 22 as evaluated in Figure 8.
2
.11. Western blotting assays
Given the great potential neuroprotective activities of 22 revealed above, we
subsequently explore the mechanism of its neuroprotective effects against cytotoxicity
induced by H O via western blotting technique. SH-SY5Y cells were pre-incubated
2
2
with DMSO, 22(5 μM, 10 μM), ibuprofen (10 μM) and ferulic acid (10 μM) for 6 h
and subsequently treated with H O (100 μM) for another 24 h. The total cytosolic
2
2
protein of SH-SY5Y cells was collected and subjected to western blotting assay with
β-actin as internal reference. As shown in Figure 11, levels of relative apoptosis
related proteins of Bcl-2, Bax and cleaved-caspase 3 in H O -treated group are
2
2
significantly changed in comparison to those of untreated-group, which clearly
revealed that H O can induce apoptosis through mitochondrial apoptosis pathway.
2
2
However, when cells co-incubated with 22 and H O , the levels of Bax and
2
2
cleaved-caspase 3 were decreased while Bcl-2 was up-regulated in contrast to those of
H O -treated group, indicating that 22 exerted neural protective effects by regulating
2
2
the expression of apoptotic protein and anti-apoptotic protein. Moreover, this
neuroprotective ability of 22 was in a concentration-dependent manner in regulating
related apoptotic proteins. As expected, the positive control (ferulic acid-treated cells)
also exerted the same neuroprotective effects by regulating the expression of relative
proteins.
2
.12. Morris Water Maze (MWM) assay

Lastly, we carried out morris water maze (MWM) test to investigate the
therapeutic effect of 22 on the learning and memory of AD model rats induced by
intrahippocampal aggregated Aβ25-35 (2 μL, 5 μg/μL). After another a week recovery
from the surgery, the rats were treated daily intraperitoneal injection of tacrine (1.0
µmol/100g b.wt., 0.2 mg/100g, qd), 22 (1.0 µmol/100g b.wt., 0.9 mg/100g, qd), using
equal amount of normal saline for control group. As shown in Figure 12, the escape
latency of rats in different groups had no significant difference of approximately 65 s.
During the following four days, rats of tacrine-treated group and 22-treated group
displayed a significant decrease in escape latency, especially for short searching
distance of tacrine-treated and 22-treated groups on the day five (Figure 12B).
Moreover, the shorter searching path of rats in 22-treated group revealed that 22
gained better memory improvement than tacrine. Not surprisingly, the escape latency
of the model group changed little in comparison to that of the first day. These results
apparently proved that 22 can ameliorate cognitive deficit induced by Aβ polymers.
Furthermore, the results in Figure 12C indicated that rats in the control group
swam aimlessly in the tank, whereas rats in tacrine-treated and 22-treated groups
preferentially swam in the northeast of the target quadrant. Similarly, rats in
2
2-treated group took more time (2.2 seconds) in the targeting platform than that of
the control or tacrine-treated rats (1.3 seconds), demonstrating that 22 has potential in
ameliorating cognitive deficit and improving spatial memory when compared with
tacrine-treated group. More excitingly, the more times (4.2) of 22-treated rats crossed
the platform than those of the control (0.4) and tacrine-treated rats (2.4) disclosed that
2
2 showed significantly therapeutic effect in ameliorating cognitive deficit and
improving spatial memory of AD model rats.
3
. CONCLUSION
In conclusion, a series of ROS-responsive prodrugs have been designed and
synthesized based on the clinic pathological characteristics of ROS overproduction,
inflammation and decrease of choline transmitters in brain. The prodrugs were
constructed and catered to our design with MTDLs strategies as a potential and novel
symptomatic- and disease-improving approach in AD treatments. The cytotoxicity

tests showed that the resulting compounds themselves had little cytotoxic activity
toward SH-SY5Y cells, but exerted excellent neuron-protective effects in the presence
of H O . Besides, little inhibitory activity on AChE was detected even at 10 μM,
2
2
whereas H O -treated 19-22 showed AChE inhibitory activity with IC values from
2
2
50
3
9.16 to 608.1 nM. Further examination indicated that upon treatment with H O , 22
2 2
can release 11 and ibuprofen, and show strong neuroprotective by inhibiting
proinflammatory cytokines (TNF-α and IL-1β) in LPS-treated microglial cells (BV-2)
and regulating apoptosis related proteins (Bax, Bcl-2 and cleaved caspase-3). In
addition, 22 showed neglected hepatotoxicity in HepG2 cells by inducing less
intracellular ROS than tacrine, demonstrating that this compound has better
biocompatibility. More importantly, compound 22 exhibited significantly therapeutic
effect in ameliorating cognitive deficit and improving spatial memory of Aβ-induced
AD model rats. Taken together, our research on these ROS-responsive prodrugs has
promise for AD treatments in future and offers a way for developing new AD drugs.
4
. EXPERIMENTAL SECTION
4
.1. Materials and Instruments
All the reagents and solvents were reagent grade with a minimum purity of 98%
purchased from Aladdin, Energy Chemical or Sinopharm Chemical Reagent Co., Ltd
and were used without further purification. All reactions were carried out under
1
nitrogen in anhydrous conditions, unless otherwise indicated. The H-NMR and
1
3
C-NMR were recorded with Bruker 600 MHz spectrometer in DMSO-d or CDCl3
6
with TMS as the internal standard. Chemical shifts were obtained in parts per million
(
ppm) and spin multiplicities in NMR spectra were abbreviated as: s (singlet), d
doublet), t (triplet), m (multiplet) or br (broad). All of the intermediates and target
(
products were purified by recrystallization or silica gel chromatography. Silica gel
200-300 mesh) was purchased from Qingdao Haiyang Chemical Co. Ltd. and was
(
used without further treatment. Reactions were monitored by UV detector using thin
player chromatography (TLC) on silica gel plates (Yantai Jiangyou Silica Gel
Development Co., Ltd.). High-resolution mass spectrometry (HR-MS) were recorded
by an Agilent 6224 ESI/TOF MS instrument. All target compounds with purity of

9
5% or above were used in the following biological studies. The purity testes were
made with RP-HPLC (waters, e2695 system) equipped with ODS column (250×4.6
mm, 5 μm) and UV/Vis detector (2489 system, using wavelength of 254 nm), running
4
5:55 acetonitrile: water (V/V) (0.05% trifluoroacetic acid in water) at a flow rate of 1
mL/min. ROS tracker (KGAF018), mitochondrial membrane potential test kit
KGA604), and Annexin-V/PI double staining kit as well as the cells used were
(
purchased from KeyGEN BioTECH Corp. Primary antibodies of Bax, Bcl-2 and
cleaved capase-3 were obtained from Imgenex, USA. Secondary antibodies labeled
with horseradish peroxidase with a dilution of 1:2000 were from Santa Cruz
Biotechnologies. Primary antibody of anti-ZO-1 (ab216880) and secondary goat
anti-rabbit IgG H&L (ab150077) were from Abcam. QuantiChrom™ Peroxide Assay
Kit was purchased from BioAssay Systems, ROS tracker (DCFH-DA, S0033S,
Beyotime).
4
.2. Synthesis of Intermediates
,6-bis(((tert-butyldimethylsilyl)oxy)methyl)-4-methylphenol (13). To a solution
2
of compound 12 (4.0 g, 23.7 mmol) and imidazole (3.7 g, 54.5 mmol) in DMF (30
mL), TBDMSCl (7.9 g, 52.3 mmol) was added dropwise at 0 ℃ and kept stirring for 2
h at the same temperature. After the completion of the reaction, EA (300 mL) was
added to dissolve the residue. The resulting solution was washed with sat. saline
solution three times. The organic layer was dried with anhydrous Na SO and
2
4
concentrated in vacuum to give crude oil, which was isolated on silica gel
1
chromatography to give compound 13. (colorless oil, 8.4 g). Yield 90.0%. H NMR
(
600 MHz, CDCl ) δ 8.03 (s, 1H, Ar-OH), 6.91 (s, 2H, Ar-H), 4.83 (s, 4H, Ar-CH ),
3
2
2
.26 (s, 3H, Ar-H), 0.95 (s, 18H, C(CH ) ), 0.13 (s, 12H, Si(CH ) ).
3
3
3 2
(((5-methyl-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)-1,3-p
henylene)bis(methylene))bis(oxy))bis(tert-butyldimethylsilane) (14). A solution of
compound 13 (4.2 g, 10.6 mmol) in DMF (15 mL) was cooled to 0 °C, then K CO
2
3
(2.8 g, 20.2 mmol) was added and stirred for 10 min at the same temperature before
the addition of 2-(4-bromobenzyl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (3.0 g).
After the completion of the reaction monitored by TLC, EA (300 mL) was added to

dissolve the residue that was washed with sat. saline solution three times. The organic
layer was dried with anhydrous Na SO and concentrated in vacuum to give crude
2
4
colorless oil, which was purified by silica gel chromatography to obtain compound 14.
1
(
colorless oil, 4.8 g). Yield 75.1%. H NMR (600 MHz, CDCl ) δ 7.85 (d, J = 8.0 Hz,
3
2
4
H, Ar-H), 7.43 (d, J = 8.0 Hz, 2H, Ar-H), 7.18 (s, 2H, Ar-H), 4.90 (s, 2H, Ar-CH2),
.71 (s, 4H, Ar-CH ), 2.35 (s, 3H, Ar-CH ), 1.37 (s, 12H, OC(CH ) ), 0.93 (s, 18H,
2
3
3 2
-
C(CH ) ), 0.08 (s, 12H, -Si(CH ) ).
3 3 3 2
(
5-methyl-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)-1,3-ph
enylene)dimethanol (15). To a solution of compound 14 (1.1 g, 1.86 mmol) in MeOH
15 mL) was added PTSA (35.4 mg). The mixture was stirred at room temperature for
0 min, then the solvent was removed under reduced pressure after the completion of
(
3
the reaction. EA (150 mL) was added to dissolve the residue that was washed with sat.
saline solution three times. The organic layer was dried with anhydrous Na SO and
2
4
concentrated in vacuum. The residue was isolated by silica gel chromatography to
1
give compound 15 (colorless oil, 640.0 mg). Yield 89.5%. H NMR (600 MHz,
CDCl ) δ 7.82 (d, J = 7.9 Hz, 2H, Ar-H), 7.40 (d, J = 7.9 Hz, 2H, Ar-H), 7.13 (s, 2H,
3
Ar-H), 4.90 (s, 2H, Ar-CH ), 4.62 (s, 4H, Ar-CH ), 2.30 (s, 3H, Ar-CH ), 1.35 (s, 12H,
2
2
3
OC(CH ) )).
3
2
(
3-(((tert-butyldimethylsilyl)oxy)methyl)-5-methyl-2-((4-(4,4,5,5-tetramethyl-1,3,
-dioxaborolan-2-yl)benzyl)oxy)phenyl)methanol (16). To a solution of compound 15
1.5 g, 3.9 mmol) and imidazole (398.2 g, 5.9 mmol) in DMF (15 mL) was added
2
(
TBDMSCl (764.0 g, 5.1 mmol) dropwise at 0 °C. The mixture was stirred overnight
at room temperature. After the completion of the reaction monitored by TLC, EA
(150 mL) was added to dissolve the residue that was washed with sat. saline solution
three times. The organic layer was dried with anhydrous Na SO and concentrated in
2
4
vacuum to give crude oil, which was isolated on silica gel chromatography to obtain
1
compound 16. (colorless oil, 964.0 mg). Yield 49.6%. H NMR (600 MHz, CDCl ) δ
3
7
.97 (s, 1H, ), 7.85 (d, J = 7.9 Hz, 2H, Ar-H), 7.43 (d, J = 7.9 Hz, 2H, Ar-H), 7.22 (d,
J = 1.5 Hz, 1H, Ar-H), 7.10 (d, J = 1.7 Hz, 1H, Ar-H), 4.93 (s, 2H, Ar-CH ), 4.72 (s,
2

2
0
H, Ar-CH ), 4.63 (s, 2H, Ar-CH ), 2.33 (s, 3H, Ar-CH ), 1.36 (s, 12H, -OC(CH ) ),
2 2 3 3 2
.93 (s, 9H, -C(CH ) ), 0.09 (s, 6H, -Si(CH ) ).
3
3
3 2
3-(((tert-butyldimethylsilyl)oxy)methyl)-5-methyl-2-((4-(4,4,5,5-tetramethyl-1,3,2
-dioxaborolan-2-yl)benzyl)oxy)benzyl 2-(4-isobutylphenyl)propanoate (17).
A
solution of ibuprofen (1.48 g, 7.2 mmol) and DCC (1.9 g, 9.4 mmol) in anhydrous
DCM (25 mL) was stirred at room temperature for 20 min. Then, compound 16 (3.2 g,
6
.6 mmol) was added dropwise at the same temperature and stirred overnight. After
the completion of the reaction, the deposit was filtered out and the organic layer was
concentrated under reduced pressure. The residue was further isolated by silica gel
chromatography with the eluent of PE and EA (30:1-25:1) to obtain compound 17.
1
(
colorless oil, 3.4 g). Yield 75.6%. H NMR (600 MHz, CDCl ) δ 7.81 (d, J = 7.7 Hz,
3
2
H, Ar-H), 7.34 (d, J = 7.7 Hz, 2H, Ar-H), 7.23 (s, 1H, Ar-H), 7.17 (d, J = 8.0 Hz, 2H,
Ar-H), 7.04 (d, J = 8.0 Hz, 2H, Ar-H), 6.93 (s, 1H, Ar-H), 5.17-5.12 (m, 2H,
Ar-CH -O), 4.75 (s, 2H, Ar-CH -O), 4.65 (s, 2H, Ar-CH -O), 3.72 (q, J = 7.1 Hz, 1H,
2
2
2
Ar-CH-CO), 2.40 (d, J = 7.2 Hz, 2H, Ar-CH -CH), 2.27 (s, 3H, Ar-CH ), 1.83-1.77
2
3
(
m, 1H, -CH(CH ) ), 1.49 (d, J = 7.2 Hz, 3H, CH(CH ) ), 1.37 (s, 12H, -OC(CH ) ),
3 2 3 2 3 2
0
.92 (s, 9H, -C(CH ) ), 0.88 (s, 3H, -CH(CH ) ), 0.87 (s, 3H, -CH(CH ) ), 0.06 (s, 6H,
3 3 3 2 3 2
-
Si(CH ) ).
3 2
3
-(hydroxymethyl)-5-methyl-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)
benzyl)oxy)benzyl 2-(4-isobutylphenyl)propanoate (18). To a solution of compound
7 (3.7 g, 5.5 mmol) in MeOH (25 mL) PTSA (104.0 mg) was added. The mixture
1
was stirred at room temperature for 30 min, the solvent was removed under reduced
pressure and EA (150 mL) was added. The resulting solution was washed three times
with sat. saline solution. The organic layer was dried with anhydrous Na SO and
2
4
concentrated in vacuum. The residue was isolated by silica gel chromatography to
1
give compound 18. (colorless oil, 1.9 g). Yield 63.0%. H NMR (600 MHz, CDCl )
3
δ 8.24 (d, J = 9.1 Hz, 2H, Ar-H), 7.81 (d, J = 7.9 Hz, 2H, Ar-H), 7.36 (d, J = 7.9 Hz,
2
1
5
H, Ar-H), 7.26 (d, J = 2.3 Hz, 1H, Ar-H), 7.25 (d, J = 2.3 Hz, 1H, Ar-H), 7.20 (d, J =
.8 Hz, 1H, Ar-H), 7.06 (d, J = 1.8 Hz, 1H, Ar-H), 5.23 (s, 2H, Ar-CH -O), 5.23 –
2
.14 (m, 2H, Ar-CH -O), 4.84 (s, 2H, Ar-CH -O), 3.74 (q, J = 7.1 Hz, 1H,
2
2

Ar-CH-CO), 2.38 (d, J = 7.2 Hz, 2H, Ar-CH -CH), 2.27 (s, 3H, Ar-CH ), 1.83 – 1.75
2
3
(
m, 1H, -CH(CH ) ), 1.50 (d, J = 7.2 Hz, 3H, -CH(CH )), 1.36 (s, 12H, -OC(CH ) ),
3 2 3 3 2
0
4
.87 (s, 3H, -CH(CH ) ), 0.86 (s, 3H, -CH(CH ) ).
3 2 3 2
.3. General preparations of compounds 19-22
To a solution of compound 18 (300 mg, 0.52 mmol) in DMF (5 mL) CDI (92.8
mg, 0.57 mmol) was added. The mixture was stirred at room temperature for 2 h, then
compound 8/9/10/11 (0.68 mmol) was added. After the resulting solution was kept
stirring for another 2 h, DCM (150 mL) was added. The organic phase was washed
three times with sat. saline solution, and then dried over anhydrous Na SO and
2
4
concentrated in vacuum. The residue was isolated by silica gel chromatography with
the eluent of DCM and MeOH (40:1-10:1) to generate target compounds 19-22.
5-methyl-3-((((2-((1,2,3,4-tetrahydroacridin-9-yl)amino)ethyl)carbamoyl)oxy)me
thyl)-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)benzyl2-(4-isobut
1
ylphenyl)propanoate (19). (pale yellow solid, 150 mg). Yield 34.3 %. H NMR (600
MHz, CDCl ) δ 14.40 (s, 1H, -NH-), 8.23 (d, J = 8.5 Hz, 1H, Ar-H), 8.11 (d, J = 8.5
3
Hz, 1H, Ar-H), 7.76 (d, J = 7.9 Hz, 2H, Ar-H), 7.35 (d, J = 7.9 Hz, 2H, Ar-H), 7.20 –
7
6
.15 (m, 2H, Ar-H), 7.14 (d, J = 8.0 Hz, 2H, Ar-H), 7.02 (d, J = 8.0 Hz, 3H, Ar-H),
.94 (s, 1H, Ar-H), 6.64 (s, 1H, -NH-), 5.15 (s, 2H, Ar-CH -O), 5.15 – 5.08 (m, 2H,
2
Ar-CH -O), 4.80 (s, 2H, Ar-CH -O), 4.07 (s, 2H, NHCH -), 3.71-3.67 (m, 3H,
2
2
2
Ar-CH-CO, NHCH -), 3.15 (s, 2H, -CH -), 2.45 (s, 2H, -CH -), 2.38 (d, J = 7.2 Hz,
2
2
2
2
H, Ar-CH -CH), 2.19 (s, 3H, Ar-CH ), 1.81-1.73 (m, 5H, -CH(CH ) , -CH -), 1.46
2 3 3 2 2
(
d, J = 7.2 Hz, 3H, -CH(CH )), 1.33 (s, 12H, -OC(CH ) ), 0.86 (s, 3H, -CH(CH ) ),
3 3 2 3 2
1
3
0
1
.85 (s, 3H, -CH(CH ) ); C NMR (150 MHz, CDCl ) δ 174.44, 158.65, 155.54,
3 2 3
53.10, 150.65, 140.62, 139.88, 138.28, 137.41, 134.96(C×2，Ar-H，overlapped
signals), 134.20, 131.72, 130.87, 130.61, 129.51, 129.47, 129.40, 129.34(C×2，Ar-H，
overlapped signals ), 127.19(C×2，Ar-H，overlapped signals), 126.89(C×2, Ar-H，
overlapped signals), 124.69, 124.40, 120.04, 115.42, 110.91, 83.87(C×2 ，
CH -C-(CH ) , overlapped signals), 62.21, 61.51, 50.66, 45.12, 44.97, 40.71, 30.13,
3
3 2

2
1
9.70, 28.33, 24.88(C×4, -CH ,overlapped signals ), 23.87, 22.39, 21.87, 20.79, 20.60,
3
+
8.35, 14.11；HR-MS(ESI) calcd for [M+H] = 840.4759, found 840.4827.
5-methyl-3-((((3-((1,2,3,4-tetrahydroacridin-9-yl)amino)propyl)carbamoyl)oxy)
methyl)-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)benzyl2-(4-iso
1
butylphenyl)propanoate (20). (pale yellow solid, 134.4 mg). Yield 30.3 %. H NMR
(
600 MHz, CDCl3) δ 14.69 (s, 1H, -NH-), 8.45 (d, J = 8.5 Hz, 1H, Ar-H), 8.16 (d, J =
.5 Hz, 1H, Ar-H), 7.77 (d, J = 7.8 Hz, 2H, Ar-H), 7.61 (t, J = 7.8 Hz, 1H, Ar-H), 7.38
t, J = 7.8 Hz, 1H, Ar-H), 7.34 (d, J = 7.7 Hz, 2H, Ar-H), 7.32 (s, 1H, Ar-H), 7.15 (d,
J = 7.9 Hz, 2H, Ar-H), 7.03 (d, J = 7.9 Hz, 2H, Ar-H), 6.94 (s, 1H, Ar-H), 5.78 (s, 1H,
NH-), 5.16 – 5.09 (m, 4H, Ar-CH -O), 4.79 (s, 2H, Ar-CH -O), 3.88 (s, 2H,
8
(
-
2
2
NHCH -), 3.72-3.68 (m, 1H, Ar-CH-CO), 3.37-3.34 (m, 2H, -NHCH -), 3.27-3.24 (m,
2
2
2
H, -CH -), 2.66 (s, 2H, -CH -), 2.38 (d, J = 7.2 Hz, 2H, Ar-CH -CH), 2.20 (s, 3H,
2 2 2
Ar-CH ), 1.84-1.76 (m, 7H, -CH(CH ) , -CH -), 1.47 (d, J = 7.1 Hz, 3H, -CH(CH )),
3
3 2
2
3
13
1
.32 (s, 12H, -OC(CH ) ), 0.86 (s, 3H, -CH(CH ) ), 0.85 (s, 3H, -CH(CH ) ); C
3 2 3 2 3 2
NMR (150 MHz, CDCl ) δ 174.46, 157.75, 155.80, 153.29, 151.17, 140.62, 139.97,
3
1
38.61, 137.45, 134.97(C×2，Ar-H, overlapped signals), 134.08, 132.14, 130.80,
30.79, 129.61, 129.42, 129.38, 129.35(C×2，Ar-H, overlapped signals), 127.21(C×2，
1
Ar-H, overlapped signals), 126.89(C×2，Ar-H, overlapped signals), 125.21, 123.93,
1
4
2
20.79, 115.76, 111.29, 83.86(C×2, CH -C-(CH ) , overlapped signals), 62.13, 61.54,
3 3 2
5.15, 44.98, 43.70, 37.28, 31.24, 30.13, 28.46, 24.87(C×4, -CH , overlapped signals),
3
4.17, 22.39, 22.38, 22.04, 20.79, 20.66, 18.37, 14.13; HR-MS(ESI) calcd for [M
+
+
H] = 854.4916, found 854.5190.
5-methyl-3-((((4-((1,2,3,4-tetrahydroacridin-9-yl)amino)butyl)carbamoyl)oxy)me
thyl)-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)benzyl2-(4-isobut
1
ylphenyl)propanoate (21). (pale yellow solid, 141.0 mg). Yield 31.2 %. H NMR (600
MHz, DMSO) δ 13.89 (s, 1H, -NH-), 8.37 (d, J = 8.5 Hz, 1H, Ar-H), 7.94 (d, J = 8.5
Hz, 1H, Ar-H), 7.83 (t, J = 7.7 Hz, 1H, Ar-H), 7.73 (s, 1H, Ar-H), 7.66 (d, J = 7.8 Hz,
2
5
6
H, Ar-H), 7.54 (t, J = 7.7 Hz, 1H, Ar-H), 7.35 (d, J = 7.8 Hz, 2H, Ar-H), 7.28 (t, J =
.6 Hz, 1H, Ar-H), 7.14 (d, J = 8.1 Hz, 2H, Ar-H), 7.04 (d, J = 8.0 Hz, 2H, Ar-H),
.92 (s, 1H, -NH-), 5.09 (d, J = 12.6 Hz, 1H, Ar-CH -O), 5.04 (d, J = 12.5 Hz, 1H,
2

Ar-CH -O), 4.97 (s, 2H, Ar-CH -O), 4.76 (s, 2H, Ar-CH -O), 3.84-3.82 (m, 2H,
2
2
2
NHCH -), 3.79 – 3.75 (m, 1H, Ar-CH-CO), 3.01-2.98 (m, 4H, -NHCH -, -CH -), 2.64
2
2
2
(s, 2H, -CH -), 2.35 (d, J = 7.1 Hz, 2H, Ar-CH -CH), 2.16 (s, 3H, Ar-CH ), 1.81-1.80
2 2 3
(
m, 4H, -CH ), 1.77-1.68 (m, 3H, -CH -), 1.47 – 1.43 (m, 2H, -CH -), 1.36 (d, J = 7.1
2
2
2
Hz, 3H, -CH(CH )), 1.30 (s, 12H, -OC(CH ) ), 0.81 (s, 3H, -CH(CH ) ), 0.81 (s, 3H,
3
3 2
3 2
1
3
-
CH(CH ) ). C NMR (150 MHz, CDCl ) δ 174.48, 156.62, 155.50, 153.22, 151.17,
3 2 3
1
1
40.62, 139.92, 138.67, 137.43, 134.97(C×2，Ar-H, overlapped signals), 134.06,
32.20, 130.79, 130.68, 129.77, 129.34(C×3，Ar-H, overlapped signals), 127.20(C×3，
Ar-H, overlapped signals ), 127.00(C×2，Ar-H, overlapped signals), 125.12, 124.30,
1
6
2
20.71, 115.72, 110.84, 83.86(C×2, CH -C-(CH ) , overlapped signals), 61.70,
3 3 2
1.55(C×2, Ar-CH -O, overlapped signals), 45.14, 44.97, 40.20, 31.58, 30.13, 28.37,
2
7.96, 27.04, 24.87(C×4, -CH , overlapped signals), 22.38(C×2,-CH -, overlapped
3
2
+
signals), 21.89, 20.79, 20.59, 18.35, 14.14; HR-MS(ESI) calcd for [M +H] =
68.5072, found 868.5058.
-methyl-3-((((6-((1,2,3,4-tetrahydroacridin-9-yl)amino)hexyl)carbamoyl)oxy)m
8
5
ethyl)-2-((4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)benzyl2-(4-isobu
1
tylphenyl)propanoate (22). (pale yellow solid, 116.4 mg). Yield 25.0 %. H NMR
(
600 MHz, CDCl ) δ 15.10 (s, 1H, -NH-), 8.53 (d, J = 8.5 Hz, 1H, Ar-H), 8.18 (d, J =
3
8
7
.5 Hz, 1H, Ar-H), 7.78 (d, J = 7.7 Hz, 2H, Ar-H), 7.66 (t, J = 7.7 Hz, 1H, Ar-H),
.41 (t, J = 7.7 Hz, 1H, Ar-H), 7.33 (d, J = 7.7 Hz, 2H, Ar-H), 7.14 (d, J = 7.9 Hz, 2H,
Ar-H), 7.12 (s, 1H, Ar-H), 7.03 (d, J = 7.9 Hz, 2H, Ar-H), 6.94 (s, 1H, Ar-H), 5.99 (s,
1
2
3
2
1
H, -NH-), 5.16 – 5.05 (m, 4H, Ar-CH -O), 4.89 (t, J = 5.7 Hz, 1H, NH-CH ), 4.77 (s,
2 2
H, Ar-CH -O), 3.90-3.87 (m, 2H, -NH-CH -), 3.70 (q, J = 7.1 Hz, 1H, Ar-CH-CO),
2
2
.29 (t, J = 6.1 Hz, 2H), 3.19-3.15 (m, 2H, -CH -), 2.58 (t, J = 5.8 Hz, 2H, -CH -),
2
2
.38 (d, J = 7.2 Hz, 2H, Ar-CH -CH), 2.21 (s, 3H, Ar-CH ), 1.80-1.80 (m, 2H, -CH -),
2
3
2
.84 – 1.77 (m, 6H, -CH -), 1.47-1.46 (m, J = 7.2 Hz, 7H, -CH -), 1.34 (s, 12H,
2
2
13
-
OC(CH ) ), 0.86 (s, 3H, -CH(CH ) ), 0.85 (s, 3H, -CH(CH ) ); C NMR (150 MHz,
3 2 3 2 3 2
CDCl ) δ 174.44, 156.48, 155.46, 153.38, 151.30, 140.60, 139.94, 138.85, 137.47,
3
1
34.98(C×2，Ar-H), 134.05, 132.24, 130.97, 130.81, 129.79(C×2，Ar-H), 129.40,

1
1
29.34(C×2，Ar-H), 127.21(C×2，Ar-H), 126.98(C×2，Ar-H), 125.12, 124.27, 120.95,
15.78, 110.74, 83.85(C×2, CH -C-(CH ) , overlapped signals), 61.73, 61.58(C×2,
3
3 2
Ar-CH -O-, overlapped signals), 45.15, 44.99, 40.59, 30.91, 30.12, 29.81, 28.38,
2
2
1
6.03, 26.00, 24.88(C×4,-CH ,), 23.80, 22.40, 22.38, 21.93, 20.78, 20.65, 18.37,
3
+
4.13; HR-MS(ESI) calcd for [M +H] = 896.5585, found 896.5562.
6-((2-oxo-2H-chromen-7-yl)oxy)hexanoic acid (24). To a solution of compound
2
3 (1.0 g, 6.17 mmol) in DMF (10 mL), K CO (3.4 g, 24.8 mmol), 6-bromohexanoic
2
3
acid (2.4 g, 12.34 mmol) and catalytic amount of KI were added. The resulting
o
solution was stirred at 55 C for 6 h, then the solvent was removed under reduced
pressure. The residue was isolated on silica gel with the eluent of DCM and MeOH
1
(
80:1-50:1) to produce compound 24. Pale yellow solid (1.3 g). Yield 76.5%. H
NMR (600 MHz, DMSO) δ 12.00 (s, 1H, -COOH), 7.98 (d, J = 9.5 Hz, 1H, Ar-H),
.61 (d, J = 8.6 Hz, 1H, Ar-H), 6.97 (d, J = 2.4 Hz, 1H, Ar-H), 6.93 (dd, J = 8.6, 2.4
Hz, 1H, Ar-H), 6.27 (d, J = 9.5 Hz, 1H, Ar-H), 4.06 (t, J = 6.5 Hz, 2H, Ar-O-CH -),
7
2
2
.23 (t, J = 7.4 Hz, 2H, -CH -COOH), 1.76 – 1.71 (m, 2H, -CH -), 1.6-1.54 (m, 2H,
2 2
-
CH -), 1.45-1.39 (m, 2H, -CH -).
2 2
3-(((2-(4-isobutylphenyl)propanoyl)oxy)methyl)-5-methyl-2-((4-(4,4,5,5-tetramet
hyl-1,3,2-dioxaborolan-2-yl)benzyl)oxy)benzyl6-((2-oxo-2H-chromen-7-yl)oxy)hexan
oate (25). DCC (174.5 g, 0.84 mmol) was added to a solution of compound 24 (180.0
mg, 0.65 mmol) in DCM (15 mL). The mixing solution was stirred for 10 min until
white solids precipitated. Then, compound 17 (372.1 mg, 0.65 mmol) and catalytic
amount of DMAP was added to the solution which was stirred for another 2 h. After
removal of the precipitate, the product was isolated with silica gel chromatography
with the eluent of EA and PE (10:1-5:1) as compound 25 (150 mg, white solid). Yield
1
2
7.8 %. H NMR (600 MHz, CDCl ) δ 7.80 (d, J = 8.0 Hz, 2H, Ar-H), 7.62 (d, J = 3.9
3
Hz, 1H, Ar-H), 7.35 – 7.34 (m, 3H, Ar-H), 7.16 (d, J = 8.0 Hz, 2H, Ar-H), 7.13 (d, J
1.9 Hz, 1H, Ar-H), 7.04 (d, J = 8.0 Hz, 2H, Ar-H), 6.98 (d, J = 1.9 Hz, 1H, Ar-H),
.81-6.79 (m, 1H, Ar-H), 6.78 (d, J = 2.3 Hz, 1H, Ar-H), 6.24 (d, J = 9.4 Hz, 1H,
Ar-H), 5.17 – 5.11 (m, 2H, Ar-CH -O), 5.11 (s, 2H, Ar-CH -O), 4.80 (s, 2H,
=
6
2
2
Ar-CH -O), 3.97 (t, J = 6.4 Hz, 2H, -O-CH - ), 3.71 (q, J = 7.1 Hz, 1H, Ar-CH(CH )),
2
2
3

2
.40 (d, J = 7.2 Hz, 2H, Ar-CH -CH(CH ) ), 2.34 (t, J = 7.5 Hz, 2H, OOCCH ), 2.24
2 3 2 2
(
s, 3H, Ar-CH -), 1.83-1.76 (m, 3H, -CH -, -CH(CH ) ), 1.71-1.66 (m, 2H, -CH -),
3 2 3 2 2
1
.51 – 1.46 (m, 5H, -CH -, Ar-CH(CH )), 1.35 (s, 12H, -OC(CH ) ), 0.87 (d, J = 0.9
2 3 3 2
1
3
Hz, 3H, -CH(CH ) ), 0.86 (d, J = 0.9 Hz, 3H, -CH(CH ) ). C NMR (150 MHz,
3
2
3 2
CDCl ) δ 174.45, 173.26, 162.31, 161.32, 155.91, 153.72, 143.48, 140.61, 139.95,
3
1
37.52, 135.03(C×2, Ar-H, overlapped signals), 134.11, 131.29, 131.16, 129.53,
29.36(C×2, Ar-H, overlapped signals), 129.27, 128.74(C×2, Ar-H, overlapped
1
signals), 127.24(C×2, Ar-H, overlapped signals), 126.88(C×2, Ar-H, overlapped
signals), 112.98, 112.94, 112.45, 101.37, 83.86(C×2, CH3-C(CH ) , overlapped
3
2
signals), 76.99, 68.28, 61.63, 61.51, 45.19, 45.01, 34.13, 30.15, 28.67, 25.57,
2
4.90(C×4, -CH , overlapped signals), 24.58, 22.42, 22.40, 20.78, 18.39.
3
4
.4. Cell Culture
SH-SY5Y, BV-2, and b.End.3 cells were maintained in RPMI 1640 or
Dulbecco’s Modified Eagle Medium (DMEM, Life Technologies) supplemented with
0 % fetal bovine serum (FBS, Gibco), penicillin (100 U/mL) and streptomycin (100
μg/mL) solution. All cells were grown at 37 °C in 5% CO atmosphere and 95% air.
1
2
4
.5. Cytotoxicity assay
The cytotoxicity was measured with MTT methods. SH-SY5Y cells were seeded
5
into 96-well plates at a density of 1×10 cells/mL and allowed to adhere for overnight.
Then compounds of 19-22 and tacrine at different concentrations in DMSO were
added to the well and co-incubated for 24 h, 48 h or 72 h at an atmosphere of 5 %
CO and 95 % air. Subsequently, MTT (5 mg/mL) was added for an extra 4 h of
2
incubation. The medium was replaced with DMSO to dissolve formazan precipitate
and the absorbance of OD was recorded with enzyme-labeling instrument at a
wavelength of 570 nm. The results of cell viabilities were analyzed with SPSS
software for three independent experiments. Neuroprotective assays of 19-22 were
carried out followed the procedures mentioned above. Briefly, SH-SY5Y cells were
pre-incubated with 19-22 (5, 10, 20 μM), ibuprofen (5, 10, 20 μM) and combined
group (ibuprofen and 11, 5, 10, 20 μM) for 6 h followed with another 24h treatment of

H O . The results obtained through MTT assays were shown as mean ± SD of three
2
2
independent experiments.
4
.6. Release of ibuprofen and tacrine derivative 11
A solution of 22 (40 μmmol/L) was prepared in purified water and acetonitrile
(
50 % : 50 %, V/V), then the solution of 22 and H O (3.0 equiv.) were co-incubated
2 2
o
at 37 C in dark. The release process was recorded and analyzed with reversed phase
high performance liquid chromatography (RP-HPLC, waters, e2695 system; 2489
UV/Vis detector). If the process slowed down, extra portion of H O (3.0 equiv.) was
2
2
added until the completion of release. Similarly, other solutions of ibuprofen (40
μmmol/L), 11 (10 μmmol/L), and 12 (40 μmmol/L) were prepared in purified water
and acetonitrile (50%:50%, V/V). All of the tests were detected on ODS column
(250×4.6 mm, 5 μm) at different time points with an eluent of acetonitrile/water
(10:90-100:0, 16 min; 100:0, 20 min; 50:50, V/V) (0.05 % trifluoroacetic acid in
water). The volume of sample injection was 10 μL and the flow rate was 1 mL/min
o
with column oven of 35 C. All solutions prefiltered with 0.45 µm filter were detected
using UV detector under 254 nm.
4
.7. In vitro H O -eliminating capability
2 2
Various concentrations of trolox (1,5,10, 20, 30 µM), 22 (1,5,10, 20, 30 µM),
ibuprofen (1,5,10, 20, 30 µM) and 11 (1,5,10, 20, 30 µM) were incubated in 4 mL
o
PBS (7.4) containing 200 µM H O at 37 C for 24 h. The remaining H O in various
2
2
2
2
reaction solutions were detected via QuantiChrom Peroxide Assay Kit (DIOX-250,
BioAssay Systems) while the scavenged H O was calculated by graph pad prism 8.0.
2
2
The relationship between consuming H O and concentrations of various compounds
2
2
were graphed via origin pro 2016.
4
.8. Blood-brain barrier (BBB) permeation assays and cellular uptake
As for the blood-brain barrier permeation assays, b.End.3 cells at a density of
5
2
2
×10 cells/cm were seeded into the upper chamber of transwell while SH-SY5Y
5
2
cells (1×10 cells/cm ) were seeded into the lower chamber of transwell. All of the
cells were cultured in monolayer. Trans-Endothelial electrical resistance (TEER)
values of cells in upper chamber were monitored with voltammeter (Millipore) and

2
used for BBB assays until the TEER values reached more than 180 Ω cm . Thereafter,
•
2
2 and 25 (10 μM) were added to the upper chamber, respectively and incubated for 2
o
h at 37 C. The concentration of 22 in the lower culture medium was analyzed with
HPLC. After another 6 h co-incubation, SH-SY5Y cells were collected, lysed with
RIPA lysis buffer and cytoplasm was collected for HPLC detection. The concentration
of 22 in SH-SY5Y cells calculated with Graph Pad Prism 8.0 and showed as one of
three independent experiment. TEER= (R -R )*A, R and R represent the total
t
b
t
b
resistance and background resistance, respectively, and A is the transwell area (0.33
2
cm ). As for 25, the cells in lower chamber were washed twice with cold PBS and
gathered for images with laser confocal microscope.
As for uptake of 25, SH-SY5Y or BV-2 cells were seeded onto six-well plates
5
o
(
1×10 cells/mL) and allowed to adhere at 37 C for overnight. After 80 % confluence,
o
the cells were treated with compound 25 (10 μM) for 12 h at 37 C. Subsequently, the
cells were washed twice with PBS, recorded and analyzed with laser scanning
confocal microscope. (Ex. 405 nm, Em. 488 nm).
4
.9. Immunofluorescence assays of ZO-1
5
BEnd.3 (5×10 cells/mL) cells were seeded into glass bottom cell culture dishes
till the formation of monomer cells. Then the cells were treated with DMSO or 22 for
5 min. After that, the cells were washed with cold PBS, fixed in paraformaldehyde
4%) and permeabilized with 0.5% Triton X-100 in PBS. Then, the cells were labeled
4
(
with primary antibodies (ab216880, abcam) followed with Alexa-Fluoro488-labeled
goat anti-rabbit secondary antibody. The nuclei were stained with DAPI, while images
were recorded and analyzed via laser confocal microscope. The results were shown as
one of three independent experiments.
4
.10. LPS-induced ROS accumulation assays
BV-2 cells were seeded onto six-well plates (1×10 cells/mL) and allowed to
adhere overnight. After 80 % confluence, the cells were pretreated with compound 22
10 μM), ibuprofen+11 (10 µM + 10 µM) for 6 h followed with another 18 h
5
(
co-treatment with LPS (3 μg/mL). Similarly, untreated cells (DMSO) and LPS (3
μg/mL)-treated cells were used as controls. BV-2 cells pretreated with LPS (3 μg/mL)

for 6 h followed with another 18 h co-incubation with 22 (10 µM) was also used as
o
control. After 24 h co-treatment at 37 C, the cells were trypsinized, collected for
DCFH-DA staining at dark for 20 min and imaged via laser confocal microscope with
the excitation wavelength and emission wavelength of 488 nm and 530 nm. Results
were displayed as one of three independent experiments.
4
.11. AChE inhibition study
The AChE inhibition was investigated with colorimetric method according to the
[32,34]
research previously reported by Ellman et al
. The details of materials were as
follows: AChE (EC3.1.1.7, Type VI−S, Electric Eel); (5,5'-Dithiobis-(2-nitrobenzoic
acid)) DTNB (pH=8.0 Gluco Biotechnology Co., Ltd, Nanjing); acetylthiocholine
chloride (ATC, Sigma Aldrich); tacrine (Sigma Aldrich). Tacrine and target
-9
-5
compounds were dissolved in DMSO and diluted to 10 -10 M with 0.1 M
NaH PO /Na HPO buffer (pH 7.4). As for the measurement of AChE assay of target
2
4
2
4
compounds under H O . Compounds 19-22 was dissolved in DMSO and diluted with
2
2
PBS buffer to 40 mM followed with the addition of H O (6.0 equiv.) and
2
2
o
co-incubated for 10 h at 37 C in dark. The release solution was diluted with PBS
buffer to a final concentration range and used for the following assays.
For measurement, the AChE was diluted to 2.0 units/mL in 2 mL aliquots. After
5
min co-incubation of contained PBS (1 mL), AChE (10 μL), test compound solution
o
(
10 μL) and DTNB (0.01 M, 30 μL) at 25 C, the reaction was started with
o
acetylthiocholine chloride (0.01 M, 30 μL) and incubated for 1 min at 37 C. The
absorbance of various solution was recorded with enzyme-labeled instrument at 410
nm for three independent experiments. Similarly, pure water was tested as control.
The IC values were calculated with Graph Pad Prism 6.0.
5
0
4
.12. Molecular docking studies
The moleular docking were performed with Sybyl-X 2.0 (Sybyl-X 2.0 was
available from state key laboratory for the chemistry and molecular engineering of
medicinal resources of guangxi normal university). The structures used in this
stdudies were constructed with chem 3D pro 14.0 followed by ennergy minimized
with tripos force field, using optimization gradient of 0.001 kcal/mol·Å and a

maximum optimal step size of 1000. The crystal structure of torpedo californica
AChE (PDB ID: 2CKM) was obtained from the RCSB protein data bank and
modified with Sybyl-X 2.0. For preparing receptor protein, the water molecules were
removed and the missing hydrogen atoms were added. The docking results were
obtained in Sybyl-X 2.0 Surflex-Dock model. Besides, the docking parameters were
kept as default values.
4
.13. The detection of inflammatory factors in BV-2 cells by real-time quantitative
PCR (qRT-PCR)
The effects of 22 on expression of inflammatory factors induced by H O were
2
2
detected by real-time quantitative PCR (qRT-PCR). BV-2 cells were seeded onto
5
o
six-well plates (1×10 cells/mL) and allowed to adhere at 37 C for overnight. When
cells were 80% confluence, the cells were pre-treated with or without 22 (5 μM, 10
μM) for 6 h and treated with H O (100 μM) for another 24 h at the same condition.
2
2
The total RNA isolated with TRIzol reagent (Invitrogen Co., Carlsbad, CA), purified
with chloroform and isopropanol and quantified with the NanoDrop 2000c
spectrophotometer (ThermoFisher Scientific, Madrid, Spain) at 260/280 wavelength.
After reverse transcription cDNA by a High Capacity RNA-to-cDNA kit (TOYOBO,
Japan), real-time quantitative PCR was performed with 10 μL of final reaction volume
including SYBR green I master mix. The method of amplification was carried out
[
47]
according to the previous report
and the results were analyzed with ΔΔCt method
[
48]
.
The primer sequences were used as follows: IL-1β R-AGC TCT CCA CCT CAA
TGG AC; TNF-α R-CTG TAG CCC ACG TCG TAG; GAPDH R-CAC TGA GGA
CCA GGT TGT CT
4
.14. Apoptosis detection
The neuroprotective effects of compound 96 were studied using apoptosis assay.
5
SH-SY5Y cells were seeded onto six-well plates (1×10 cells/mL) and allowed to
o
adhere at 37 C for overnight. The cells were pretreated with control (DMSO),
o
compound 96 (5 μM and 10 μM) for 6 h at 37 C. Then the cells were treated H O
2
2
(100 μM) and incubated for another 24 h at the same condition. Then both the

adherent and floating cells were collected, centrifuged (2000 rpm for 5 min) and
washed twice with cold PBS. Lastly, the cells were resuspended with 500 μL binding
buffer, stained with Annexin V-FITC (5 μL) and propidium iodide (5 μL) for 15 min
at room temperature and analyzed with flow cytometry (Cell Quest; BD Biosciences).
4
.15. Hepatotoxicity Studies
In order to investigate whether our target copunds obtained decreased
hepatotoxicity in coparison to tacrine. ROS accumulation assays induced by 22 or
5
tacrine were measured. Briefly, Hep G2 were seeded onto six-well plates (1×10
o
cells/mL) and allowed to adhere at 37 C for overnight. Subsequently, the cells were
treated with control (DMSO), compound 22 (25 μM and 50 μM) and tacrine (50 μM)
o
at 37 C for another 24 h. Thereafter, the cells were washed twice with cold PBS,
stained with ROS tracker followed by washing with PBS. The accumulation of ROS
was recorded and analyzed with laser confocal microscope (Ex. 490 nm/Em. 535 nm).
4
.16. Cell morphology assay
The cell morphology assay was detected with cytotoxicity assay kit (KGAF001,
5
KeyGEN BioTECH). SH-SY5Y cells were seeded onto six-well plates (1×10
o
cells/mL) and allowed to adhere at 37 C for overnight. Then the cells were treated
o
with control (DMSO), compound 22 (5 μM and 10 μM) for 6 h at 37 C. Subsequently,
the cells were treated with H O (100 μM) and incubated for another 24 h at the same
2
2
condition. Lastly, the cells were stained with calcein AM/PI probe prepared according
to the illustration of the manufacture for 30 min in dark at room temperature. The
cells were washed twice with cold PBS and analyzed with laser confocal microscope.
4
.17. Western blotting assay
In order to investigate the expression of related protein in the mitochondrial
apoptosis pathway, the western blotting study was performed. SH-SY5Y cells were
5
seeded into six-well plates at a density of 1×10 cells/mL and allowed to adhere at 37
o
C for overnight. Thereafter, the cells were pre-incubated with different conditions,
including 22 (5 μM and 10 μM), control (DMSO) for 6 h. Then the cells were exposed
to H O (100 μM) (one group of control was not treated) and incubated for another 24
2
2