Synthesis and biological evaluation of acridine-based histone deacetylase inhibitors as multitarget agents against Alzheimer's disease

Article abstract of DOI:10.1016/j.ejmech.2020.112193

Multitarget agents simultaneously trigger molecules in functionally complementary pathways, and are therefore considered to have potential in effectively treating Alzheimer's disease (AD), which has a complex pathogenetic mechanism. In this study, the HDAC inhibitor core is incorporated into the acetylcholine esterase (ACE) inhibitor acridine–derived moiety and resulted in compounds that exhibited higher class IIa HDAC (4, 5, 7, and 9)- and class IIb HDAC6-inhibiting activity when compared to the pan-HDAC inhibitor SAHA in clinical practice. One of these compounds, 11b, displayed greater selectivity toward HDAC6 than other isoform enzymes. In contrast, the activity of compound 6a was selective toward class IIa HDAC and HDAC6. These two compounds exhibited strong activity against Aβ-aggregation as well as significantly disrupted Aβ-oligomer. Additionally, 11b and 6a strongly inhibited AChE. These experimental findings demonstrate that compounds 11b and 6a are HDAC-Aβ-aggregation-AChE inhibitors. Notably, they can enhance neurite outgrowth, but with no significant neurotoxicity. Further biological evaluation revealed the various cellular effects of multitarget compounds 11b and 6a, which have the potential to treat AD.

Full text of DOI:10.1016/j.ejmech.2020.112193

Journal Pre-proof
Synthesis and biological evaluation of acridine-based histone deacetylase inhibitors
as multitarget agents against Alzheimer's disease
Hui-Ju Tseng, Mei-Hsiang Lin, Young-Ji Shiao, Ying-Chen Yang, Jung-Chun Chu,
Chun-Yung Chen, Yi-Ying Chen, Tony Eight Lin, Chih-Jou Su, Shiow-Lin Pan, Liang-
Chieh Chen, Chen-Yu Wang, Kai-Cheng Hsu, Wei-Jan Huang
PII:
S0223-5234(20)30160-4
DOI:
https://doi.org/10.1016/j.ejmech.2020.112193
EJMECH 112193
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 9 January 2020
Revised Date: 24 February 2020
Accepted Date: 25 February 2020
Please cite this article as: H.-J. Tseng, M.-H. Lin, Y.-J. Shiao, Y.-C. Yang, J.-C. Chu, C.-Y. Chen, Y.-
Y. Chen, T.E. Lin, C.-J. Su, S.-L. Pan, L.-C. Chen, C.-Y. Wang, K.-C. Hsu, W.-J. Huang, Synthesis
and biological evaluation of acridine-based histone deacetylase inhibitors as multitarget agents against
Alzheimer's disease, European Journal of Medicinal Chemistry (2020), doi: https://doi.org/10.1016/
j.ejmech.2020.112193.
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.

Synthesis and biological evaluation of acridine-based histone deacetylase
inhibitors as multitarget agents against Alzheimer’s disease
a
b
c
d
e
Hui-Ju Tseng , Mei-Hsiang Lin , Young-Ji Shiao , Ying-Chen Yang , Jung-Chun Chu ,
e
f,g
f,g
f,g
f,g
Chun-Yung Chen , Yi-Ying Chen , Tony Eight Lin , Chih-Jou Su , Shiow-Lin Pan ,
f,h
i
a,f,g*
a,b,e∗
Liang-Chieh Chen , Chen-Yu Wang , Kai-Cheng Hsu , Wei-Jan Huang
a
Ph.D. Program in Biotechnology Research and Development, Taipei Medical University,
Taipei, Taiwan
b
School of Pharmacy, Taipei Medical University, Taipei, Taiwan
c
National Research institute of Chinese Medicine, Taipei, Taiwan
d
Department of Biotechnology and Animal Science, National Ilan University, Ilan, Taiwan
e
Graduate Institute of Pharmacognosy, Taipei Medical University, Taipei, Taiwan
f
Graduate Institute of Cancer Biology and Drug Discovery, Taipei Medical University, Taipei,
Taiwan
g
Ph.D. Program for Cancer Molecular Biology and Drug Discovery, Taipei Medical
University, Taipei, Taiwan
h
School of Life and Health Sciences and Warshel Institute for Computational Biology, The
Chinese University of Hong Kong (Shenzhen), Shenzhen, Guangdong, P. R. China
i
TaiMed Biologics Inc., Taipei, Taiwan
Abstract
Multitarget agents simultaneously trigger molecules in functionally complementary pathways, and are
therefore considered to have potential in effectively treating Alzheimer’s disease (AD), which has a
complex pathogenetic mechanism. In this study, the HDAC inhibitor core is incorporated into the
acetylcholine esterase (ACE) inhibitor acridine–derived moiety. Some resulting compounds exhibited
higher class IIa HDAC (4, 5, 7, and 9)- and class IIb HDAC6-inhibiting activity than did the reference
SAHA as a pan-HDAC inhibitor in clinical practice. One of these compounds, 11b, displayed greater
selectivity toward HDAC6 than other isoform enzymes. In contrast, the activity of compound 6a was
selective toward class IIa HDAC and HDAC6. These two compounds exhibited strong activity against
Aβ-aggregation as well as significantly disrupted Aβ-oligomer. Additionally, 11b and 6a strongly
inhibited AChE. These experimental findings demonstrate that compounds 11b and 6a are multiple
HDAC-Aβ-aggregation-AChE inhibitors. Notably, they can enhance neurite outgrowth, but with no
significant neurotoxicity. Further biological evaluation revealed the various cellular effects of
multitarget compounds 11b and 6a, which have the potential to treat AD.
Keywords: Histone deacetylase; Acetylcholine esterase; Acridine; Multitarget
*Corresponding authors
E-mail: wjhuang@tmu.edu.tw (W.-J. Huang)
E-mail: piki@tmu.edu.tw (K.-C. Hsu)
1

Introduction
Owing to the developments in healthcare, the elderly population worldwide is
increasing rapidly. Aging-related diseases including cardiovascular disease, cancer,
arthritis, dementia, cataract, osteoporosis, diabetes, hypertension and Alzheimer's
disease (AD) have become important issues. The prevalence of Alzheimer’s disease
(AD) is increasing gradually. According to statistics from the American Alzheimer’s
association, the population with AD in 2018 was estimated to be 5.7 million. In the
US in 2017, AD was the third most common cause of death of older people, behind
only heart disease and cancer. Accordingly, AD has a great impact on health, welfare
and economic development.
AD is characterized by progressive memory loss, a decline in language skills,
and other cognitive impairments. Although its etiology is yet to be fully clarified,
studies have demonstrated that AD has multiple causes, including amyloid-β (Aβ)
aggregation, tau protein hyperphosphorylation, acetylcholine deficiency, genetic
2
+
mutations, brain trauma, oxidative stress, neuroinflammation, and Ca -induced
neurotoxicity [1]. The complex pathogenetic mechanism of AD causes the problem
that currently used single-target drugs can only alleviate the symptom of cognitive
impairment, but are incapable of slowing the progression of AD. To overcome this
obstacle, a multi-target strategy or combined therapy that uses at least two drugs with
synergistic effects and involves the simultaneous triggering of different molecules in
2

the redundant pathway that is associated with AD pathogenesis can be developed. The
multi-target strategy has several advantages over combined therapy, including higher
patient compliance, a low risk of drug-drug interaction, the development of new
chemical entities and predictive pharmacokinetics [2, 3]. From 2015 to 2017, an
estimated 21% of FDA-approved new molecular entities were multi-target drugs [4].
Therefore, the use of multi-target agents is considered to be effective for treating AD
[
5, 6].
Histone deacetylases (HDACs) are a group of epigenetic enzymes that can
regulate gene expression without changing the gene sequence. HDACs remove acetyl
group from the lysine amino acid on histone, which causes the condensed nucleosome
structure to inhibit gene expression. Moreover, HDAC regulates the acetylation of
some nonhistone proteins such as tubulin, HSP90 and P53. These findings reveal that
HDAC controls a wide range of cellular events [7, 8]. HDACs comprise 18 isoform
enzymes that fall into four classes according to their sequence homology, cofactor,
and intracellular distribution. Increasing evidence indicates that these HDAC enzymes
are associated with learning, memory formation and neuroplasticity. Further,
dysregulation of histone acetylation and deacetylation is observed in AD mice models
and AD patients [9, 10]. For example, in the transgenic mice model, the
overexpression of HDAC2 reduces dendritic spine density, synaptic number and
3

synaptic plasticity, and impairs memory formation[11]. One investigation has revealed
that HDAC4 overexpression induces neurotoxicity, which is related to
neurodegeneration [12]. HDAC6 levels are substantially increased in the
hippocampus and cortex of the AD patient brain and this enzyme is implicated in tau
hyperphosphorylation [13]. Hyperphosphorylated tau protein has been demonstrated
to induce the formation of intracellular neurofibrillary tangles (NFTs), which cause
synaptic dysfunction and neuron impairment [14]. The inhibition of HDAC6
suppresses the phosphorylation of tau, stabilizing the HDAC6-tau complex [15].
Furthermore, HDAC6 inhibition downregulates Aβ aggregation [16] and alleviates
cognitive impairment in an in-vivo AD mouse model [13]. Importantly, several HDAC
inhibitors, such as 4-phenylbutrate, valproic acid, vorinstat (SAHA) [17] and MS 275
[
18], exhibit in-vitro and in-vivo activity and thus improve cognition. These findings
indicate that HDAC has a crucial role in AD and HDAC enzymes may be targeted to
treat AD.
Only four drugs - rivastigmine, galantamine, donepezil, and memantine - have
been developed to treat AD [1]. The first three are acetylcholine esterase (AChE)
inhibitors, which can prevent the degradation of acetylcholine into choline and acetic
acid and thus increase acetylcholine levels in the synapse of CNS, alleviating
cognition deficits that are associated with AD. The last one, memantine, an NMDA
4

2
+
receptor antagonist, attenuates Ca -induced neurotoxicity. These four compounds
have limited therapeutic outcomes as they alleviate some symptoms but cannot slow
the progression of AD [1]. This result is explained by the fact that AD is caused by
multiple molecular abnormalities rather than a single defect. Simultaneously
triggering molecules in functionally complementary pathways may provide
synergistic effects.
Numerous structurally diverse multifunctional agents that act on different
biological targets relevant to AD etiology have been discovered in attempts to solve
the clinical problems of current anti-AD drugs (Fig. 1). Some compounds have
advanced to clinical trials. For example, a tacrine-indole hybrid NP-61 has been
identified as a dual inhibitor of AChE and Aβ secretion, but was withdrawn in phase 1
of a clinical study [19]. Compound 1, derived from donepezil, exhibits AChE and
selective monoamine oxidase A (MAO) inhibition and functions as an antioxidant and
2
+
Cu chelator [20]. Similarly, ladostigil, which contains rivastigmine that is fused with
a rasagiline motif, exhibits enzyme inhibition against AChE and MAO and
antioxidative activity [21]. Compound 2, which comprises tacrine in combination with
a nimodipine motif, inhibits AChE as well as Aβ aggregation and secretion, and
2
+
blocks the Ca channel [22]. Ladostigil and BGC20-1259 have advanced to a clinical
study. These results demonstrate that multifunctional agents may have the potential
5

effectively to target the etiologically complex AD.
Figure 1. Identified multitarget-directed ligands against AD.
Acridine, which is a structurally diverse scaffold, is commonly used for anti-AD
drug design[23]. For example, 9-aminotetrahydroacridine (tacrine) exhibited potent
AChE inhibitory activity (Fig. 1S). Compound 12 that contains
acridine-tetrahydroacridine hybrid is demonstrated to show much higher activity than
tacrine (Fig. 1S), which suggests that acridine group contributed to anti-AChE
effect[24]. Acridine derivative 25 exhibited potent inhibitory activity against Aβ
aggregation [25] (Fig. 1S). Additionally, it displayed high blood-brain barrier (BBB)
permeability. The structure-activity relationship (SAR) study reveals that the planar
structure of acridine greatly contributed to Aβ aggregation-inhibiting activity[26].
Recently, Jiang and coworkers used a hybrid of acridine derivative and
pyridothiaoxazole scaffold of a glycogen synthase kinase-3β (GSK-3β) inhibitor to
6

develop a series of multitarget inhibitors [27]. One of these compounds, 2f, not only
exhibited potent enzyme-inhibiting activity against GSK-3β and AChE, but also
inhibited Aβ aggregation (Scheme 1), which indicates chemical conjugation at the C-9
position of acridine moiety can generate novel multitarget agents without significantly
reducing its anti-AChE effect. These findings suggest acridine derivatives can be
considered as an effective ligand for the design of multitarget-directed agents against
AChE and Aβ associated with AD pathology owing to its simple structure and high
ligand efficiency [28, 29]. Givinostat is a HDAC inhibitor currently under clinical
study[30]. It contains a phenylhydroxamic acid moiety as a core scaffold for
enzyme-inhibiting activity. In addition to givinostat, this scaffold is also found in
various HDAC inhibitors such as PCI-24781[31], tubastatin A [32], and NCC149
[
33](Fig 2S).
Based on the roles of HDAC, Aβ protein and AChE inhibitors in AD as
mentioned above, the simultaneous targeting of AChE, HDAC and Aβ may have a
synergistic anti-AD effect. We sought to develop a multiple target agent possibly as an
effective strategy against AD. In the present work, we incorporated phenylhydroxamic
acid core of givinostat into C-9 position of various acridine–derived moieties using
nitrogen as a connecting unit (Scheme 1). Novel series of acridine derivative-bearing
phenylhydroxamic acids were synthesized. The SAR of the resulting compounds that
7

inhibited various HDAC isoforms was disclosed. Of these compounds, 11b exhibited
more selectivity toward HDAC6 than other isoform enzymes (Scheme 1). In contrast,
compound 6a showed selective inhibition against class IIa HDAC and HDAC6
(Scheme 1). These acridine derivative-containing compounds were further evaluated
for inhibitory activities against Aβ aggregation as well as AChE. We found that both
compounds 11b and 6a not only exhibited strong activity against Aβ-aggregation, but
also showed strong AChE inhibition. These experimental findings identified
compounds 11b and 6a as multitarget inhibitors of HDAC, Aβ-aggregation, and
AChE. The analysis of molecular docking revealed the structural difference between
these two compounds when bound to class IIa and class IIb HDAC. Furthermore, they
can promote neurite outgrowth in mice hippocampus cells without significant
neurotoxicity. Further experiments on microsome stability and blood-brain barrier
(BBB) permeability were performed and revealed the potential of the resulting
compounds against AD.
8

1
.
Chemistry
The hydroxamates 5a-d were synthesized as described in Scheme 2. Methyl
esters 3a-c with different carbon linker chain-lengths reacted with NH OH under
2
basic conditions to give hydroxamates 4a-c, respectively. Compounds 4a-c coupled
with 9-chloroacridine 5 using 2-methoxyethanol yielded corresponding compounds
6
a-c. Using the described synthetic methodology, compound 6d was synthesized
staring from 3d.
9

Compounds 11a-h were synthesized as described in Scheme 3. Ullmann coupling of
-iodobenzoic acid 7 with different substituted aniline 8a-e provided 9a-e,
2
respectively. Friedel-Crafts acylation followed by the chlorination of 9a-e yielded
corresponding 9-chloroacridines 10a-e. Compounds 10a-h coupled with compound
4
a in the presence of 2-methoxyethanol afforded 11a-h, respectively. Details on the
synthesis, isolation and characterization of compounds 4 and 9-10 can be found in the
supplementary material. The estimated purity of compounds 6a-6d and 11a-11h is at
least 97% as determined by HPLC analysis (supplementary material).
1
0

2
.
Results and discussion
3
.1 HDAC-inhibiting activity
Compounds 6a-d were tested to determine their enzyme inhibitory activity
1
1

against a panel of HDAC isoforms, including HeLa nuclear HDACs, class IIa HDAC
4, 5, 7, 9) and class IIb HDAC (6), using SAHA as a reference (Table 1). Most HeLa
(
nuclear HDACs contained class I HDACs. Of the four compounds, compound 6a
exhibited more potent activity than SAHA against class IIa and class IIb HDACs. Of
the compounds 6a-c with different carbon chain linker between aminoacridine and
hydroxamic acid, compound 6a was the most potent, suggesting that the phenyl linker
contributed greatly to enzyme inhibitory activity. In contrast, compound 6d exhibited
low activity, suggesting meta-substituted acridine negatively contributed to activity.
These experimental results motivate the synthesis of compounds 11a-h bearing varied
acridines at the para-position of phenylhydroxamate.
Table 1
a
IC50 value (µM) of compounds 6a-d for HDACs inhibition
Class I
Class IIa
Class IIb
HDAC6
HeLa
nuclear
HDAC
Compound
HDAC4
HDAC5
HDAC7
HDAC9
6
a
0.55 ± 0.03 2.24 ± 0.06 0.85 ± 0.07 0.65 ± 0.03 0.57 ± 0.02 0.025 ± 0.00
2.07 ± 0.26 26.63 ± 5.38 1.66 ± 0.30 1.94 ± 0.03 6.42 ± 0.39 0.074 ± 0.00
0.10 ± 0.03 7.06 ± 1.10 2.58 ± 0.47 31.53 ± 2.98 28.96 ± 0.89 0.14 ± 0.00
6
b
6
c
6
d
> 40
0.034 ± 0.00 33.04 ± 6.85 10.51 ± 0.86
Data are obtained from three independent experiments.
29.38 ± 0.96 20.76 ± 3.93 16.84 ± 0.50 12.18 ± 1.15 4.69 ± 0.19
SAHA
> 40 32.26 ± 1.60 0.083 ± 0.00
a
1
2

The experimental results reveal that compounds 11a-c exhibited higher class
IIa and class IIb HDAC inhibiting activity than SAHA (Table 2). Compound 11a
exhibited greater activity against class I HDACs and retained the activity toward
other isoforms, compared to compound 6a. This experimental result indicates that
the C-2 fluoro substituent of acridine broadened the spectrum of anti-HDAC activity.
However, compounds 11b and 11c with C-2-bromo and -OCF , respectively,
3
exhibited weaker activity than compound 11a against class IIa HDACs. Compound
11d with the C-2 OBn group displayed selectivity toward HDAC4 and 5 over other
isoform enzymes. Compound 11e with the C-2, -3 and -4 methoxy groups had lower
activity against most HDAC enzymes. Similarly, compound 11f with C-2 chloro and
C-7 OMe exhibited weaker activity. Compound 11g with the tacrine moiety showed
much weaker activity toward all enzymes than compound 6a, suggesting that the
coplanar tricyclic skeleton greatly contributed to activity. Interestingly, compound
11h with quinolone exhibited HDAC6-selective inhibition.
1
3

Table 2
a
IC50 value (µM) of compounds 11a-h for HDACs inhibition
Class I
Class IIa
Class IIb
HDAC6
HeLa
nuclear
HDAC
Compound
HDAC4
HDAC5
HDAC7
HDAC9
6
a
0.55 ± 0.03 2.24 ± 0.06 0.85 ± 0.07 0.65 ± 0.03 0.57 ± 0.02 0.025 ± 0.00
0.25 ± 0.01 3.04 ± 0.73 0.69 ± 0.04 0.59 ± 0.06 0.52 ± 0.03 0.029 ± 0.00
0.16 ± 0.00 7.85 ± 0.68 7.19 ± 0.47 9.13 ± 0.30 5.51 ± 0.07 0.026 ± 0.00
0.67 ± 0.19 14.71 ± 4.31 15.83 ± 1.92 17.88 ± 1.54 31.10 ± 1.45 0.029 ± 0.00
3.50 ± 0.41 1.21 ± 0.06 0.43 ± 0.09 3.10 ± 0.30 1.25 ± 0.07 0.22 ± 0.04
0.81 ± 0.14 15.73 ± 0.08 4.23 ± 0.03 4.57 ± 0.04 4.95 ± 0.21 0.16 ± 0.02
1
1a
1b
1c
1d
1e
1f
1g
1h
SAHA
1
1
1
1
1
0.80 ± 0.06
> 40
> 40
> 40
> 40
15.18 ± 2.91 10.29 ± 0.69
17.54 ± 0.57 > 40
> 40
> 40
0.34 ± 0.00
3.55 ± 0.35
1
1
0.60 ± 0.00
3.37 ±0.50 32.67 ± 7.08 13.23 ± 1.29 0.039 ± 0.00
> 40 32.26 ± 1.60 0.083 ± 0.00
Data are obtained from three independent experiments.
0.034 ± 0.00 33.04 ± 6.85 10.51 ± 0.86
a
3
.2 Inhibition of Aβ1-42 self-aggregation
Compounds 6a and 11a-h were evaluated for the inhibition against Aβ1-42
oligomerization using the Thioflavin T (ThT) assay. Morin was used as the reference
compound. Compounds 6a and 11b-d were more potent than Morin in inhibiting
Aβ1-42 self-aggregation, as they had IC values around 3µM (Table 3). Furthermore,
50
compounds 11b-d and 11f disaggregate Aβ1-42 oligomer by more than 50% in 10 µM.
In particular, compound 11b significantly inhibited Aβ1-42 aggregation and Aβ1-42
1
4

oligomer disaggregation.
Table 3
a
Inhibitory activities of compounds 6a, and 11a-h against Aβ1-42 oligomerization
a
a
Inhibition of Aβ1−42 Aggregation Aβ1−42 of Disaggregation
Compound
R1
H
R2+R3
IC (µM)
(%)
5
0
6
a
3.0 ± 0.7
> 20
41
11a
2-F
38
70
53
57
42
53
11b
2-Br
2-OCF
1.1 ± 0.5
3.5 ± 0.9
3.0 ± 0.6
13.0 ± 0.2
34.9 ± 6.5
> 20
11c
3
11d
2-OBn
2,3,4-OMe
3-Cl
11e
11f
11g
H
H
0
0
11h
H
> 20
8.5 ± 0.6
> 20
c
b
Morin
SAHA
ND
0
a
Data are obtained from three independent experiments.
b
ND: not determined.
1
5

3
.3 AChE inhibiting activity
Compounds 6a and 11a-h were tested to evaluate their AChE-inhibiting activity
(Table 4). Four compounds 6a, 11a-b, 11e, and 11h exhibited comparable activities
with IC values in the range of 0.6-0.8 µM.
5
0
Table 4
a
IC50 value (µM) of AChE inhibitory activity of compounds 6a, and 11a-h
Compound
R1
H
R2+R3
AChE
6
a
0.72 ± 0.02
1
1a
2-F
2-Br
0.60 ± 0.03
0.88 ± 0.05
3.65 ± 0.08
2.64 ± 0.13
0.64 ± 0.01
9.25 ± 0.09
1.97 ± 0.07
11b
11c
2-OCF
3
1
1d
1e
1f
1g
1h
Acridine
2-OBn
2,3,4-OMe
3-Cl
1
1
1
H
H
1
H
0.83 ± 0.01
6.84 ± 0.06
1
6

a
Data are obtained from three independent experiments.
3
.4 Class I HDACs isoform-inhibiting activity
To confirm that compounds 6a and 11b were less potent than SAHA in
inhibiting class I HDAC, their class I HDAC1, 2 and 3 enzyme-inhibiting activities
were evaluated. The experimental results indicated that compound 6a exhibited much
lower activity against HDAC1, 2, and 3 than did SAHA. Compound 11b exhibited
less potent activity against other enzymes, except for HDAC1.
Table 5
a
IC50 value (µM) of compounds 6a, and 11b against class I HDACs 1, 2, and 3
Class I
Compound
HDAC1
HDAC2
HDAC3
6
a
0.34 ± 0.01
0.09 ± 0.01
0.025 ± 0.00
1.57 ± 0.04
0.48 ± 0.08
0.091 ±0.00
1.22 ± 0.00
0.33 ± 0.00
0.032 ± 0.00
11b
SAHA
a
Data are obtained from three independent experiment.
.5 Analysis of molecular docking of compounds 6a and 11b
To better understand the inhibition of the HDAC isozymes by the synthesized
3
compounds, a molecular docking analysis was performed. We validated the docking
protocol by docking the co-crystal ligands into their respective HDAC active sites
(Fig. 3S). The docking poses of the compounds show similar conformations to their
1
7

co-crystal confirmations. This suggests a reliable docking protocol was used.
Next we selected compounds 6a and 11b for analysis because both showed
inhibitory activity towards class IIa and IIb HDACs (Table 2). For our analysis, we
selected the docking pose that not only produced the best docking score, but also
coordinated to the zinc ion using its ZBG. The selected pose showed comparable
docking scores to multiple solutions, which suggests a favorable docking
conformation. The compounds can also be separated into three groups with respect to
traditional HDAC inhibitors – ZBG, linker and cap group. Compounds 6a and 11b
both contain a hydroxamic acid moiety that functions as the ZBG (Fig. 2). This
moiety also forms hydrogen bonds with residues H611 and G619. The linker consists
of an aniline moiety which spans a hydrophobic tunnel and its aromatic ring facilitates
hydrophobic interactions with residues such as F620 and F680. The nitrogen on the
linker forms an additional hydrogen bond with residue S568 near the periphery of the
active site. The caps of both compounds are composed of different structures.
Compound 6a contains an acridine cap, whereas compound 11b has a
2
-bromoacridine cap. Both compounds have hydrophobic contacts with residue L749.
Their similar inhibitory activities are attributable to their similar interactions.
1
8

Figure 2. Docking poses of compounds 6a and 11b in HDAC6. The docking poses of
compound 6a (A, orange) and compound 11b (B, purple) in the HDAC6 active site.
Hydrogen bonds are represented as green dash lines. HDAC6 (grey) is represented as
a stick model. Residues are labeled.
Compound 6a inhibits class IIa isozymes. The ZBG and the linker cause
HDAC7 to adopt a pose similar to that of HDAC6 (Fig. 3A). For example, the
hydroxamic acid not only coordinates the zinc ion, but also forms hydrogen bonds
with residues H669 and H670. The linker also hydrophobically interacts with residues
F679 and H709. The acridine cap of compound 6a hydrophobically interacts with
residues F737 and F738. Both residues occupy the rim of the HDAC7 active site,
which is a pocket that is favorable for the aromatic rings (Fig. 3B). The protein loops
of HDAC6 and HDAC7 occupy different positions. Furthermore, HDAC6 contains
residue S568 while HDAC7 contains residue D626. The orientation of the loop in
HDAC6 facilitates the formation of a hydrogen bond with the linker of compound 6a.
1
9

HDAC7 does not exhibit this interaction. The loss of a hydrogen bond interaction
reduces the effectiveness of inhibiting activity of compound 6a towards HDAC7.
Figure 3. Comparison of docking poses of compounds 6a in HDAC6 and HDAC7 (A)
Docking pose of compound 6a (yellow) in the HDAC7 (pink) active site. (B)
Superimposed docking poses of compound 6a in HDAC6 (orange) and HDAC7.
Hydrogen bonds are denoted as green dashed lines. HDAC6 (grey) and HDAC7 (pink)
are represented as stick models. Residues are labeled.
Compound 11b exhibits even weaker inhibitory activity toward class IIa
isozymes (Table 2). Compound 11b adopts a pose similar to that of compound 6a in
HDAC7; it forms hydrogen bonds with residues H669 and H670 and hydrophobically
interacts with residues F679 and F709 (Fig. 4A). However, the bromo moiety of the
compound 11b cap group sterically clashes with residues H709 and P809 (Fig. 4). In
contrast, the acridine cap of compound 6a does not extend into this pocket (Fig. 4B).
This fact may explain why compound 6a may exhibit stronger inhibiting activity than
2
0

compound 11b toward HDAC7. Together, the cap groups of this class of compounds
alter the inhibitory activities between HDAC isozymes.
Figure 4. The cap group affects HDAC inhibitory activity. (A)Docking pose of
compound 11b (blue) in the HDAC7 (pink) active site. The distance between the
bromo moiety and surface residues H709 and P809 is shown. (B) Docking pose of
compound 6a (yellow) in HDAC7 is superimposed onto pose of compound 11b in
HDAC7 (surface model). Hydrogen bonds are represented as green dashed lines.
Residues are labeled.
2
1

3
.6 Promotion by compounds 6a and 11b of histone H3 and α-tubulin acetylation
Histone H3 and α-tubulin are substrates of class I HDACs and HDAC6,
respectively. Compounds 6a and 11b were incubated with murine neuroblastoma
Neuro-2a cells at different concentrations using trichostatin A (TSA) as a reference
compound. Western blotting revealed that compounds 6a and 11b dose-dependently
increased the levels of acetylated histone H3 (Fig. 5). The EC values of compounds
5
0
6
a and 11b were 7.96 and 0.58 µM, respectively. These two compounds also induced
α-tubulin acetylation in a dose-dependent manner. Their EC50 was around 8 µM.
Experimental data on the promotion of histone and tubulin acetylation were consistent
with the fact that compound 11b exhibited higher activity than did compound 6a in
inhibiting HDACs 1, 2, and 3.
Figure 5. Induction by compounds 6a and 11b of histone H3 and tubulin acetylation
2
2

in mice neuroblastoma Neuro-2a cells. Western blot analysis of histone H3 and
tubulin acetylation after treatment with compound 6a and 11b were exhibited.
Neuro-2a cells were treated with various concentration of compounds 6a and 11b for
2
4 h. Protein levels after treated with compounds 6a and 11b were detected by
Western blot analysis. Quantification analysis of Western blot was conducted using
LI-COR Odyssey Fc Imaging System. The EC values of compounds 6a and 11b
5
0
were calculated using the GraphPad Prism 4 program based on a sigmoidal
dose-response equation of 10 dose concentrations using 2-fold dilution (n = 3).
3
.7 Effects of 6a and 11b on cell viability in murine neuroblastoma Neuro-2a cells
Figure 6 shows the cell viabilities of murine neuroblastoma Neuro-2a cells that
were incubated with compounds 6a and 11b. The figure indicates that both
compounds, even at concentrations as high as 10 µM, only slightly affected cell
viability, suggesting that both compounds exhibited low cytotoxicity.
Figure 6. Effects of compounds 6a and 11b on cell viability of mice neuroblastoma
Neuro-2a cells. The Cell Titer-Glo 2.0 Luminescent cell viability assay was used for
2
3

Neuro-2a cells that were treated with various concentration of compounds 6a, 11b and
reference staurosporine for 24 h. The data reported are means from two independent
experiments performed in triplicate.
3
.8 Induction by compounds 6a and 11b of neurite outgrowth in mice
hippocampus cells
To investigate the enhancement of neurite outgrowth by compounds 6a and 11b,
primary rat hippocampus neurons were incubated with these two compounds at
various concentrations of these two compounds for three days. Neuronal
differentiation was identified by immunostaining. The lengths and numbers of primary,
secondary, and tertiary neurites were obtained (supplementary material). Figure 7
shows that compounds 6a and 11b increased the number of axonal and dendritic
neurites in a dose-dependent manner compared to those of control. Meanwhile,
compound 11b increased the length of axonal neurites in a dose-dependent manner
(Fig. 4S). These results suggest that compounds 6a and 11b promote neurite
outgrowth.
2
4

Figure 7. Promotion of hippocampal neurite formation by compounds 6a and 11b
revealed by immunostaining with anti-tau antibody for axon recognition, anti-MAP2
antibody for dendrite recognition, and DAPI for nuclear recognition. The primary
hippocampus neuron was incubated with different concentrations of compounds 6a
and 11b.
3
.9 Metabolic stability of compounds 6a and 11b against human liver microsome
The pharmacokinetic profiling of compounds 6a and 11b was evaluated in terms
of metabolic stability against human liver microsome. The experimental results
revealed that compounds 6a and 11b exhibited higher T1/2 than the reference
testosterone (Table 6). Both compounds had lower CL , falling into the medium class
int
of CL classification band[34].
int
Table 6
Metabolic stability of compounds 6a, 11b, and reference compound in human liver
microsomes
-1
a
b
Compound
k (min )
T (min)
CLint
1
/2
2
5

(mL/min/mg
protein)
6
a
0.0204
0.0183
0.0270
±
±
±
0.0011
0.0011
0.0039
33.92
37.86
25.67
±
±
±
1.93
2.20
3.51
0.0407
0.0366
0.0540
±
±
±
0.0023
0.0022
0.0077
11b
Testosterone
a
T , p < 0.05 vs. testosterone. Statistical analysis was performed using one-way
1
/2
ANOVA followed by Bonferroni t-test.
b
Intrinsic clearance (CL ) was calculated based on CL = k/P, where k is the
int
int
elimination rate constant and P is the protein concentration in the incubation.
3
.10 Blood-brain barrier permeability of compounds 6a and 11b
Compounds 6a and 11b were assayed in MDR1-MDCK test to assess their BBB
permeability [35-37]. Experimental data indicated that compound 11b had moderate
-6
BBB permeability with a mean Papp value of 4.57 × 10 cm/s and an efflux ratio of
7
.42 (Table 7). However, compound 6a had a low BBB permeability with mean Papp
-6
value of 2.72 × 10 cm/s.
Table 7
a
MDR1-MDCK permeability of compounds 6a, and 11b
-6
b
c
Compound
Mean Papp (10 cm/s)
Efflux ratio
BBB category
6
a
2.72 ± 0.04
17.54 ± 1.22
7.42 ± 0.27
Low
11b
4.57 ± 0.47
Moderate
a
Test concentration 5 µM; incubation time 2 h
b
(Papp B to A) / (Papp A to B).
2
6

c
Category criteria: low permeability, Papp A to B ≥ 3.0 and ER ≥ 10, or Papp A to B
3.0; moderate permeability, Papp A to B ≥ 3.0 and 10 > ER ≥ 3; high
permeability, Papp A to B ≥ 3.0 and 3.0 > ER.
<
3
.
Conclusion
In conclusion, this study developed a novel series of acridine-containing
hydroxamates by combining the core of HDAC and AChE inhibitor. Some of these
compounds not only exhibited potent inhibiting activity against specific HDAC
isoforms, but also significantly inhibited AChE. Two of these compounds, 6a and 11b,
had excellent inhibitory activity against Aβ-aggregation and induced remarkable
Aβ-oligomer disruption. These experimental findings suggest that the compounds
were multi HDAC-Aβ-aggregation-AChE inhibitors. Compounds 6a and 11b
significantly induced histone and tubulin acetylation. Notably, they can promote
neurite outgrowth without significant neurotoxicity. These two compounds exhibited
stability against microsome in the assay of microsome stability. Additionally,
compound 11b exhibited moderate BBB permeability. Further studies of the
mechanisms of these multitarget inhibitors are ongoing.
2
7

4
.
Experimental
5
.1. General procedures
1
13
NMR spectra ( H and C NMR) were obtained using the Bruker Fourier 300
and AVIII 500 spectrometer with standard plus programs. The chemical shifts were
presented in parts per million (ppm, δ) with TMS as an internal standard. The MS data
were measured on a Finnigan Mat TSQ-7000 mass spectrometer (ESIMS and
HRESIMS). The melting point was recorded on a Fisher-Johns apparatus
(uncorrected). HPLC was performed using a C18 column (150 mm × 4.6 mm,
Ascentis) and an L-2130 pump (Hitachi, Ibaraki, Japan). Column chromatography
was performed on silica gel (70-230 mesh, Merck, Darmstadt, Germany). The TLC
analysis was performed on silica gel plates (KG60-F254, Merck). A microplate
spectrophotometer Victor 2X (Perkin Elmer Fremont, CA, United States) was used for
fluorometric analysis; an Envision 2104 Multilabel Reader (Perkin Elmer, CA, United
States) was used for luminescence analysis, and a Sunrise microplate reader (TECAN,
Männedorf, Switzerland) was used for absorbance analysis. Unless otherwise
mentioned, all chemicals and materials were used as received from commercial
suppliers without any purification. Anhydrous dichloromethane was distilled from
calcium hydride under N2.
5
.2. N-Hydroxy-4-(acridin-9-ylamino)benzamide (6a)
2
8