A novel series of l-2-benzyloxycarbonylamino-8-(2-pyridyl)- disulfidyloctanoic acid derivatives as histone deacetylase inhibitors: Design, synthesis and molecular modeling study

Article abstract of DOI:10.1016/j.ejmech.2012.03.009

Histone deacetylases inhibitors (HDACIs) have become an attractive class of anticancer agents. In order to find some novel potent HDACIs, we designed and synthesized a series of l-2-benzyloxycarbonylamino-8-(2-pyridyl)- disulfidyloctanoic acid derivatives. All compounds exhibited potent HDAC-inhibitory activity, and two of them had similar potency to TSA. The introduction of 2-amino-4-phenylthiazole or 9-methyleneoxy-fluorenyl group at the surface recognize domain of these HDACIs could greatly increase their HDAC-inhibitory activity. Molecular modeling studies indicated that coordination of the zinc ion by these inhibitors, and formation of hydrogen bond and hydrophobic interaction between inhibitors and HDACs were essential for the HDAC-inhibitory activities of these inhibitors. Asp181, Asp269, Leu276 and Tyr308 in the active site of HDAC2 gave favorable contributions for binding with all compounds.

Full text of DOI:10.1016/j.ejmech.2012.03.009

European Journal of Medicinal Chemistry 52 (2012) 111e122
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
A novel series of
L
-2-benzyloxycarbonylamino-8-(2-pyridyl)-disulfidyloctanoic
acid derivatives as histone deacetylase inhibitors: Design, synthesis and molecular
modeling study
*
Dawei Huang, Xiaohui Li , Yingdong Wei, Zhilong Xiu
School of Life Science and Biotechnology, Dalian University of Technology, No.2 Linggong Road, Dalian 116024, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Histone deacetylases inhibitors (HDACIs) have become an attractive class of anticancer agents. In order to
Received 26 October 2011
Received in revised form
5 March 2012
Accepted 6 March 2012
Available online 15 March 2012
find some novel potent HDACIs, we designed and synthesized a series of L-2-benzyloxycarbonylamino-8-
(2-pyridyl)-disulfidyloctanoic acid derivatives. All compounds exhibited potent HDAC-inhibitory activity,
and two of them had similar potency to TSA. The introduction of 2-amino-4-phenylthiazole or 9-
methyleneoxy-fluorenyl group at the surface recognize domain of these HDACIs could greatly increase
their HDAC-inhibitory activity. Molecular modeling studies indicated that coordination of the zinc ion by
these inhibitors, and formation of hydrogen bond and hydrophobic interaction between inhibitors and
HDACs were essential for the HDAC-inhibitory activities of these inhibitors. Asp181, Asp269, Leu276 and
Tyr308 in the active site of HDAC2 gave favorable contributions for binding with all compounds.
Ó 2012 Elsevier Masson SAS. All rights reserved.
Keywords:
Histone deacetylase
Histone deacetylase inhibitor
Docking
Molecular dynamics
1. Introduction
HDACs, HDAC1 and HDAC2 are highly conserved, with about 82%
sequence identity between the two proteins. The amino acid resi-
The reversible acetylation of the
3
dues in the active pockets of both enzymes are identical.
Evidence has shown that treatment of tumor cells with HDAC
inhibitors (HDACIs) can arrest cell growth and lead to differentia-
tion and apoptosis [8e10]. Thus research focusing on turning
HDACIs into a promising class of anticancer agents is growing [11].
SAHA [12] and FK228 [13] (Fig. 1) have been approved by the US
FDA for the treatment of cutaneous T-cell lymphoma (CTCL). A
number of other HDACIs, such as valproic acid [14] and MS-275
[15,16] (Fig. 1), are in different phases of clinical trails for the
treatment of different cancers.
The classic pharmacophore of HDACIs consists of three distinct
domains: metal binding, linker and surface recognition domains
[17]. The metal binding domain coordinates the zinc ion in the
bottom of the active pocket of HDACs, and it appears to be
a precondition for the possession of HDAC-inhibitory activity. The
metal binding domain usually comprises a group with metal
chelating ability, such as the hydroxamic group [12,18] and thiol
group [13,19]. The linker occupies the hydrophobic channel, so
hydrophobicity is important for the linker domain. The surface
recognition domain is essential for recognizing and binding to the
rim of active pocket of enzymes.
dues at the N-terminal domain of histone catalyzed by histone
acetyl transferases (HATs) and histone deacetylases (HDACs) is one
of the most important chromatin epigenetic modifications [1].
Deacetylation of histone would lead to chromatin condensation,
and transcriptional repression is associated with high levels of
histone deacetylation. HDACs therefore play a pivotal role in the
regulation of gene expression, cell growth and proliferation [2].
Overexpression of HDACs has been linked to the development of
cancers in human [3]. Thus, HDAC has become an important target
enzyme for anticancer therapies. There are eighteen HDACs in
human and these enzymes are subdivided into four classes based
on their homologies to yeast HDACs, subcellular localizations and
enzymatic activities. Class I includes HDAC1, 2, 3 and HDAC8; Class
II has six members, HDAC4-7, HDAC9 and HDAC10; Class III HDACs,
also known as Sirtuins, include Sirt1-7, which are NADþ-dependent
enzymes; Class IV which comprises HDAC11, exhibits properties of
both class I and class II HDACs [4]. Class I, II and IV are Zn2þ
dependent enzymes, and these HDACs are the targets often used for
designing anticancer drugs. The structural information of Class I
HDACs was obtained from the crystal structures of HDAC-like
protein (HDLP) [5], HDAC8 [6] and HDAC2 [7]. Among class I
In this study, a series of
L-2-benzyloxycarbonylamino-8-(2-
pyridyl)-disulfidyloctanoic acid derivatives (Fig. 2) were synthe-
sized as HDACIs. These compounds were designed based on the
structural features of HDACs. The metal binding domain was
* Corresponding author. Tel.: þ86 411 84706326; fax: þ86 411 84706369.
E-mail address: lxhxh@dlut.edu.cn (X. Li).
0223-5234/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2012.03.009

112
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
calculated by molecular mechanics generalized Born surface area
(MM-GBSA) method. This study applied molecular modeling to
the designing of HDACIs to reveal the structure-activity relation-
ship (SAR) of
L-2-benzyloxycarbonylamino-8-(2-pyridyl)-disul-
fidyloctanoic acid based HDACIs.
2. Results and discussion
2.1. Docking studies
In order to estimate the rationality of designed compounds and
confirm that all of them could bind to the active pocket of HDACs
and thus have HDAC-inhibitory activities, all compounds with thiol
groups as metal biding domains were docked at the active site of
HDAC2 before they were synthesized. As shown in Fig. 3, all
compounds were successfully docked into the active pocket. The
linker occupied the channel, the surface recognition domain
interacted with the rim of the active pocket and the thiol group was
at the bottom of the channel. Hydrogen bond and hydrophobic
interaction in the binding between the compounds and HDACs
enzyme were also examined since these two forces are important
in the binding between inhibitor and enzyme. As shown in Fig. 4,
when compound 1 is bound to HDAC2, the atoms of the surface
recognition domain could hydrogen bond to the amino acid resi-
dues in the active site of the enzyme, and the thiol group could
interact with the zinc ion that lies in the bottom of the active
pocket. Other parts of the compounds made contact with some of
the surrounding amino acid residues, such as Phe, Leu, Gly and His
via hydrophobic interactions. The results of docking confirmed that
these compounds could bind to HDACs. It indicated that the
structural differences in the surface recognition domain among
these compounds could affect their binding to HDACs. The
1-naphthylamine, cyclohexylamino and 2-amino-4-phenylthiazole
groups were introduced to compounds 1, 2 and 3, respectively, in
order to compare the effect of polycyclic aromatic hydrocarbon,
cycloalkane and heterocyclic compound on the interactions
between inhibitors and HDACs. The N-terminal benzyloxycarbonyl
groups tended to contact the benzene rings of Phe210 or Tyr209 via
Fig. 1. Structures of some HDACIs.
designed as a thiol group that is protected as a disulfide hybrid [20],
but could be reduced to a thiol group by intracellular -glutathione.
The linker domain of these compounds was designed as six
methylenes. In the surface recognition domain, some different
groups, especial large aromatic groups were introduced at the
L
carboxyl group of
L-2-benzyloxycarbonylamino-8-(2-pyridyl)-
disulfidyloctanoic acid. Dockings were performed for the binding of
the compounds to HDAC2 before they were synthesized to make
sure that all compounds could bind to the active pocket of HDACs.
After the HDAC-inhibitory and anti-proliferative activities against
tumor cells were assessed, molecular dynamics (MD) simulations
were conducted to further investigate the interactions between
this series of inhibitors and HDACs. Binding free energy was
computed by molecular mechanics PoissoneBoltzmann surface
area (MM-PBSA) method. Contributions of amino acid residues
in the active site of HDAC2 to binding free energy were
pep interactions, whereas the C-terminal groups bound to the
hydrophobic groove formed by His33, Pro34, Phe155 and Leu276,
so these rings in the C-terminus had little influence on their
bindings to HDAC2. For compounds 4 and 5, the different substi-
tution sites of biphenyl group had some effect on their interactions
with HDAC2. The 2-aminobiphenyl group could contact both
Tyr209 and Phe210, but the 4-aminobiphenyl group only interacted
with Tyr209. Both benzyloxycarbonyl groups of these two
compounds bound to the hydrophobic groove. The C-terminus of
compound 6 was an ester, and there was a methylene between the
ester bond and benzene ring, so the biphenyl group contacted the
hydrophobic groove, and the N-terminal benzyloxycarbonyl group
of compound 6 interacted with Tyr209. The N-terminal benzylox-
ycarbonyl group in compounds 7 and 8 both interacted with
Phe210. The C-terminus of compound 7 was 9-aminofluorene, and
it bound to the hydrophobic groove. The C-terminus of compound 8
was an ester and there was one more methylene group than
compound 7, and it tended to bind to the groove formed by Tyr209
and Leu276. The 4-nitrobenzyl group of compound 9 contacted
Phe210, and the 3-nitrobenzyl group of compound 10 interacted
with Leu276 and Arg275. In compound 10, the nitro group was at
the m-site, so it could interact with the guanidino group of Arg275
in a good direction. The o- or m- substitution of naphthalene
evidently affected the binding of compounds 11 and 12 to HDAC2. It
showed that the 2-naphthylmethyl group of compound 11 bound to
the groove formed by Tyr209 and Leu276, but the 1-
naphthylmethyl group of compound 12 interacted with Phe210.
Fig. 2. Structures of designed compounds.

D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
113
Fig. 3. Top views of the surfaces of the active site region of HDAC2 with compounds from docking.
When the 1-naphthylmethyl group was replaced with
a
9-
The target compounds were synthesized by the procedure
outlined in Scheme 1. All synthesized compounds were confirmed
by HRMS, 1H NMR and 13C NMR.
anthracenylmethyl group in compound 13, it seemed that the 9-
anthracenylmethyl group tended to contact Tyr209. The results
indicated that there were several sites at the rim of active pocket
that could accommodate the surface recognition domains of this
series of inhibitors. The structural differences in the C-terminus of
these compounds could affect their binding with HDAC2.
2.3. Biology
The in vitro HDAC-inhibitory activity of all compounds was
determined with HDAC Fluorescent Assay/Drug Discovery Kits (AK-
511 and AK500, Biomol), and the anti-proliferation activity of all
compounds against MCF-7 (human breast cancer cell line) cells was
determined by MTT assay. TSA was used as a positive control. All
compounds inhibited the activities of HDAC1, HDAC2 and exerted
2.2. Chemistry
Non-natural amino acid
L-2-amino-8-bromooctanoic acid
(L-Ab8) was synthesized and its overall yield was 18.2%.
Fig. 4. Hydrogen bonds and hydrophobic contacts between HDAC2 and compound 1.

114
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
Scheme 1. Synthetic route of designed compounds.
anti-proliferation activity against MFC-7 cells at low concentrations
(Table 1). Compounds 3 and 8 were the most potent compounds,
being comparable to TSA. The rest of the compounds exhibited
similar HDAC-inhibitory activity, which was about 10 folds less
compared to compounds 3 and 8. They also showed similar
anti-proliferative activity against cancer cells. This suggested
that introduction of different groups at the C-terminus of
conducted for the enzymeeinhibitor complexes using HDAC2 as
a representative enzyme. The energetic and structural properties of
the complexes were analyzed to explore their dynamic stability.
The structures remained stable during the last 500 ps of the
simulations for all systems, so all the complexes had converged and
reached equilibrium. Atom coordinates in the last 500 ps were
chosen to analyze the structure in detail.
L
-2-benzyloxycarbonylamino-8-(2-pyridyl)-disulfidyloctanoic acid,
Root-mean-squared-fluctuation (RMSF) of Ca, C, N atoms for
all complexes were analyzed (Fig. 5). It indicated that all
complexes had similar RMSF. Though some amino acid residues
in some complexes fluctuated much more than that in other
complexes, the conformational changes of the amino acid resi-
dues in the active site, especially the residues at the bottom of
the channel of HDAC2 (His145, His146, Asp181, His183 and
Asp269) were all very small (Fig. 6). It suggested that in HDAC2,
some amino acids could shift far away from their normal posi-
tions when the enzyme is bound to an inhibitor, but for the
amino acids that are in the active site, binding with an inhibitor
could make them more rigid.
HDAC2 is a zinc-dependent enzyme, and so the ability of
a compound to chelate the zinc ion is a precondition for the inhi-
bition of against the enzyme. The zinc ion coordination states in MD
simulations showed that the zinc ion in each complex coordinated
five or six atoms, resulting in a pentahedral or hexahedral geometry
(Table 2). The carboxylate oxygens of Asp181 and Asp269, and both
nitrogen and oxygen atoms of His183 could coordinate the zinc ion.
One coordination site was occupied by the sulfur atom of the
inhibitor, so coordination of the zinc ion by a sulfur atom was
essential for this series of compounds to exert their HDAC-
inhibitory activity.
which resulted in a change of surface recognition domain could
have a substantial effect on HDAC binding. The introduction of 2-
amino-4-phenylthiazole (compound 3) or 9-methyleneoxy-fluo-
renyl (compound 8) group at the surface recognition domain could
largely increase their HDAC-inhibitory activity. The change
between other compounds seemed to have little effect on their
HDAC-inhibitory activities.
2.4. MD simulations
To study the interaction between HDAC and each of the
synthesized compounds in aqueous solution, MD simulations were
Table 1
HDAC-inhibitory and anti-proliferative activities against tumor cells data of
synthesized compounds.
Compd.
IC50 (mM)
HDAC1
HDACs
Cell line (MCF-7)
TSA
1
2
3
4
5
6
7
8
0.0023
0.0652
0.0669
0.0051
0.0712
0.0709
0.0641
0.0682
0.0076
0.0644
0.0667
0.0652
0.0647
0.0592
0.0011
0.0382
0.0441
0.0039
0.0452
0.0466
0.0407
0.0481
0.0052
0.0455
0.0485
0.0403
0.0396
0.0436
1.7
15.3
19.6
3.2
20.8
18.1
15.6
17.2
3.5
16.9
15.2
14.8
15.3
14.0
Hydrogen bond formed between a ligand and a protein is very
important for stabilizing the ligandeprotein complex. In the
current MD simulations, the hydrogen bonds between each of the
compounds and HDAC2 were examined (Table 3). The result
showed that the sulfur atom of each compound could form strong
hydrogen bond with Tyr308 of HDAC2, and this gave rise to the
potency of these compounds with respect to HDAC inhibition. For
compounds 3 and 8, there was another strong hydrogen bond
between the compounds and HDAC2. In compound 3, the sulfur
9
10
11
12
13

D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
115
Fig. 5. RMSF for all amino acid residues of HDAC2 during the last 500 ps in MD simulations.

116
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
Fig. 6. RMSF for residues in the active site of HDAC2 during the last 500 ps in MD simulations.

D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
117
Table 2
Average distances between zinc ion and surrounding atoms in HDAC2-compound complexes.
Compd.
Distance (Å)
Asp181:OD1
Asp181:OD2
His183:ND1
His183:O
Asp269:OD1
Asp269:OD2
Compd.:S
1
2
3
4
5
6
7
8
2.60
2.72
2.63
2.61
2.71
2.70
2.62
2.70
2.68
2.66
2.63
2.66
2.59
2.68
2.59
2.68
2.74
2.61
2.61
2.65
2.61
2.57
2.58
2.72
2.71
2.71
2.89
2.81
5.68
2.88
2.83
2.80
2.92
2.81
2.88
2.87
2.85
2.86
2.86
6.15
2.90
4.75
7.25
2.89
2.91
5.18
2.82
4.90
4.87
2.84
3.54
6.65
2.64
2.64
2.69
2.66
2.63
4.21
2.68
4.05
2.61
3.12
3.73
3.41
2.88
2.90
4.18
2.79
2.81
4.14
2.64
2.94
2.65
3.49
2.62
2.62
2.71
2.65
2.79
2.79
2.84
2.82
2.79
2.77
2.81
2.80
2.79
2.77
2.82
2.78
2.79
9
10
11
12
13
atom of the metal binding domain also hydrogen bonded with
His183. The introduction of a thiazole group at the surface recog-
nition domain may have a large effect on the interaction between
the surface recognition domain and the amino acid residues in the
rim of the active pocket, and this could further affect the interaction
between the metal binding domain and surrounding amino acid
residues of HDAC2. In compound 8, the longer linker of the surface
recognition domain could help the carbonyl oxygen to form
hydrogen bond with Phe210 of HDAC2. The additional strong
hydrogen bonds formed by compound 3 and 8 gave these two
compounds higher affinity for HDACs compared to the other
compounds.
The electrostatic interactions between water molecules and
inhibitors were stronger than that between HDAC2 and inhibitors,
and therefore the total electrostatic interactions ( Gele) were
D
nonpol
sol
unfavorable for binding. Hydrophobic interaction
(
D
G
)
contributed favorably, and it was more than half of the total binding
energy ( Gb) for some compounds. This was expected since the
D
linker and surface recognition domains of the inhibitors are
hydrophobic, and the active pocket of HDAC2 has some hydro-
phobic amino acid residues, such as Phe, Tyr and Leu, which can
generate strong hydrophobic interaction with these inhibitors. The
binding energies of compounds 3 and 8 were much lower
compared to the binding energies of the rest of the compounds, and
this was in accordance with their higher HDAC-inhibitory activity.
Contributions of the amino acid residues in the active site of
HDAC2 to the binding free energy were calculated using MM-GBSA
method (Table 5). The results revealed that Asp181, Asp269 Leu276
and Tyr308 were all favorable for binding with all compounds, so
these amino acid residues were essential for stabilizing the HDAC2-
inhibitor complexes.
2.5. Free energy calculation
Free energy associated with the binding of inhibitor to HDAC2
was analyzed by the MM-PBSA method integrated in AMBER9. As
shown in Table 4, intermolecular van der Waals (
trostatics (
Eienlet ) interactions provided major contributions to the
binding. non-polar salvation terms (
Gnsoolnpol) also contributed
Gesolel) opposed binding.
D
Eivndtw) and elec-
D
D
favorably, whereas polar solvation terms (
D
3. Conclusions
We successfully synthesized a series of HDACIs with differences
in the surface recognition domains, and all compounds showed
potent HDAC-inhibitory abilities and anti-proliferative activities
against cancer cells. The introduction of 2-amino-4-phenylthiazole
or 9-methyleneoxy-fluorenyl group at the surface recognition
domain could largely increase their HDAC-inhibitory activity.
Docking studies showed that changes in the structures of these
compounds could affect the mode of their binding with HDACs. MD
simulations revealed that zinc coordination, formation of hydrogen
bond and the presence of hydrophobic effect contributed to the
strong inhibitory activity of these compounds toward HDACs.
Asp181, Asp269, Leu276 and Tyr308 of HDAC2 contributed favor-
ably to the binding between enzyme and each of the compounds, so
these amino acids should have an important effect on deacetylation
ability of HDAC2. The results obtained from molecular modeling
were in accordance with the HDAC-inhibitory activity profiles, so
therefore could be considered as a reliable method available for the
designing and screening of new HDACIs that have similar structures
with this series compounds. The information obtained from this
study would provide new insight into the mechanism of the
interaction between inhibitors and HDACs, which could assist with
the rational design of some novel and potent HDACIs for medical
application.
Table 3
Hydrogen bonds between the compounds and HDAC2 active site obtained from MD
simulations.
Compd. Donor
AcceptorH
Acceptor
MD simulation
Occupied
(%)
Distance
(Å)
1
2
3
Compd. 1:S
Tyr308:HH
Tyr308:HH
Tyr308:HH
Tyr308:OH
Tyr308:OH
Tyr308:OH
100.00
100.00
100.00
85.40
100.00
100.00
100.00
100.00
100.00
68.20
2.96
2.89
2.94
3.19
2.95
2.91
2.90
2.98
2.88
3.08
2.96
2.94
2.92
3.20
Compd. 2:S
Compd. 3:S
Compd. 3:S
Compd. 4:S
Compd. 5:S
Compd. 6:S
Compd. 7:S
Compd. 8:S
His183:HE2 His183:NE2
4
5
6
7
8
Tyr306:HH
Tyr308:HH
Tyr308:HH
Tyr308:HH
Tyr308:HH
Tyr306:OH
Tyr308:OH
Tyr308:OH
Tyr308:OH
Tyr308:OH
Phe210:N
Tyr308:OH
Tyr308:OH
Tyr308:OH
Phe210:N
Compd. 8:O2 Phe210:H
Compd. 9:S Tyr308:HH
Compd. 10:S Tyr308:HH
Compd. 11:S Tyr308:HH
9
10
11
100.00
100.00
100.00
15.60
Compd.
11:O2
Phe210:H
12
13
Compd. 12:S Tyr308:HH
Compd. 13:S Tyr308:HH
Tyr308:OH
Tyr308:OH
100.00
100.00
33.60
2.89
2.94
2.94
Compd.
13:O2
His183:HE2 His183:NE2

118
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
Table 4
Binding free energies of all compounds to HDAC2 from MD simulations.
nonpol
sol
ele
int
vdw
int
ele
sol
Compd.
D
E
D
E
DEint
D
G
D
G
D
Gsol
DGele
DGb
1
2
3
4
5
6
7
8
ꢂ68.80
ꢂ63.59
ꢂ69.62
ꢂ70.36
ꢂ68.87
ꢂ64.18
ꢂ69.86
ꢂ68.19
ꢂ77.63
ꢂ75.53
ꢂ64.03
ꢂ63.37
ꢂ68.97
ꢂ30.07
ꢂ30.03
ꢂ34.41
ꢂ30.28
ꢂ28.79
ꢂ28.25
ꢂ28.95
ꢂ34.9
ꢂ98.87
ꢂ93.62
ꢂ5.79
ꢂ5.07
ꢂ6.61
ꢂ6.75
ꢂ6.60
ꢂ5.66
ꢂ5.48
ꢂ5.45
ꢂ6.06
ꢂ5.81
ꢂ5.38
ꢂ5.23
ꢂ6.09
95.11
88.8
89.31
83.73
88.23
92.24
89.9
84.47
90.83
86.97
97.61
92.25
91.34
88.47
92.35
26.31
25.21
25.21
28.62
27.63
25.94
26.45
24.23
26.04
22.53
32.69
30.33
29.47
ꢂ9.55
ꢂ9.89
ꢂ15.80
ꢂ8.40
ꢂ7.76
ꢂ7.96
ꢂ7.98
ꢂ16.12
ꢂ7.91
ꢂ9.30
ꢂ9.49
ꢂ9.56
ꢂ11.26
ꢂ104.03
ꢂ100.64
ꢂ97.66
94.83
98.98
96.50
90.12
96.31
92.42
103.67
98.06
96.72
93.7
ꢂ92.43
ꢂ98.81
ꢂ103.09
ꢂ105.52
ꢂ101.55
ꢂ100.83
ꢂ98.03
9
ꢂ27.89
ꢂ26.02
ꢂ36.80
ꢂ34.66
ꢂ34.64
10
11
12
13
ꢂ103.61
98.44
4. Experiments and computational methodology
was monitored by TLC (CHCl3:CH3OH:CH3COOH ¼ 90:10:2) and
HPLC. After the reaction had completed, the acetone was evaporated
and the residue was washed with aether. Citric acid was added until
the pH reached 3.0. The residue was then dissolved in ethyl acetate
and thenwashed in sodium bicarbonate solution (4%) and brine, and
4.1. Synthesis
4.1.1. Instruments and materials
Thin layer chromatography (TLC) and/or high performance
liquid chromatography (HPLC) were used to check all intermediate
products and target compounds. TLC was performed on aluminum-
backed silica gel plates (Merck DC-Alufolien Kieselgel 60 F254) and
the spots were visualized by ultraviolet light or charring. Analytical
HPLC was performed on Jasco PU-1586 using a Chromolith
Performance RP-18e column (4.6 ꢀ 100 mm, Merck). The mobile
phases used were A: H2O with 10% CH3CN and 0.1% trifluoroacetic
acid (TFA); B: CH3CN with 0.1% TFA. Elution of the desired
compounds was carried out using a solvent gradient of A to B over
15 min at a flow rate of 2 mL/min. Eluent was detected at 220 nm.
High-resolution mass spectrometry (HRMS) were measured on
a UPLC/Q-Tof Mass Spectrometer instrument. Nuclear magnetic
resonance (NMR) spectra were recorded on a Varian INOVA
400 MHz spectrometer.
dried over anhydrous MgSO4. Z-
acetate was evaporated (1.04 g, 93.2%). To a solution of Z-
L
-Ab8 was obtained as oil after ethyl
L
-Ab8
(1.04 g) in tBueOH (14 mL), Boc anhydride (1.22 g) and
4-dimethylamiopryidine (DMAP) (34 mg) were added. After the
reaction had completed as monitored by TLC (CHCl3:CH3OH ¼ 9:1)
and HPLC, the tBueOH was evaporated and the residue was purified
by silica gel chromatography using a mixture of hexane and ethyl
acetate (8:1) to yield Z-
L
-Ab8-OtBu (1.08 g, 90.2%). The side chain of
Z-L
-Ab8-OtBu was changed from a bromo group to a thioacetyl
group by treatment with potassium thioacetate (433 mg) in N,N-
dimethylformamide (DMF) for 5 h. The thioester was extracted as
before to yield Z-L
-Am8(SAc)-OtBu (1.0 g, 93.3%). The compound was
then dissolved in DMF, and 2, 20-dipyridyldisulfide (624 mg) was
added, followed by 40% solution of CH3NH2/CH3OH (1.27 mL) under
argon atmosphere and stirred for 6 h. After evaporation of DMF, the
residue was extracted and purified by silica gel chromatography
using mixture of dichloromethane (DCM) and methanol (149:1) to
4.1.2. Synthesis of L-Ab8
L-Ab8 was synthesized according to the procedure reported in
ref. [21] using diethyl acetamidomalonate and 1, 6-dibromohexan
as starting material.
yield Z-L
-Am8(S2Py)-OtBu as powder (743 mg, 64.2%). The
compound was dissolved in TFA (1.5 mL) at 0 ꢁC and kept for 2 h. TFA
in the sample was evaporated and Z-Am8(S2Py)-OH was obtained as
a yellow oil (661 mg,100%). Z-Am8(S2Py)-OH (217 mg) was coupled
with 1-naphthylamine (72 mg) in cooled DMF, and N-hydrox-
ybenzotrizole (HOBt$H2O) (92 mg), 2-(1H-benzotriazole-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) (228 mg),
4.1.3. Synthesis of compound 1
To a solution of L-Ab8 (714 mg) in H2O (3 mL), Na2CO3 (382 mg),
acetone (3 mL) and N-(benzyloxycarbonyloxy) succinimide (Z-OSu)
(897 mg) were added and stirred for 5 h. The progress of the reaction
and N,N-diisopropylethylamine (DIEA) (87
mL) were added and
Table 5
Contributions of the amino acid residues in the active site of HDAC2 to the binding free energy (kcal/mol).
Compd.
Amino acid residues
His145
His146
Gly154
Phe155
Asp181
His183
Asp269
Leu276
Tyr308
1
2
3
4
5
6
7
8
1.22
3.68
2.92
1.02
1.06
1.71
4.11
1.20
2.22
3.50
1.08
1.33
0.63
9.02
1.52
4.57
4.20
6.17
3.22
5.40
7.88
6.10
5.44
ꢂ0.07
ꢂ0.07
8.35
1.29
0.14
0.35
0.87
0.33
0.50
0.60
ꢂ0.88
ꢂ0.14
0.89
6.02
6.55
5.67
7.25
6.14
6.49
5.20
6.37
6.73
6.54
6.79
6.43
5.86
ꢂ69.01
ꢂ64.20
ꢂ64.45
ꢂ67.72
ꢂ63.65
ꢂ64.26
ꢂ67.81
ꢂ63.04
ꢂ66.23
ꢂ66.11
ꢂ61.57
ꢂ63.67
ꢂ68.48
0.45
3.16
7.33
2.16
4.13
3.27
2.59
3.43
4.02
3.32
3.15
4.35
ꢂ0.53
ꢂ63.11
ꢂ59.51
ꢂ61.51
ꢂ62.78
ꢂ59.89
ꢂ59.83
ꢂ64.79
ꢂ59.69
ꢂ62.77
ꢂ62.25
ꢂ55.91
ꢂ59.25
ꢂ61.28
ꢂ20.29
ꢂ22.08
ꢂ21.70
ꢂ20.96
ꢂ21.70
ꢂ21.12
ꢂ20.27
ꢂ22.91
ꢂ20.00
ꢂ20.27
ꢂ23.31
ꢂ20.87
ꢂ22.29
ꢂ19.28
ꢂ19.57
ꢂ19.73
ꢂ18.38
ꢂ20.40
ꢂ20.52
ꢂ19.72
ꢂ19.92
ꢂ19.31
ꢂ20.76
ꢂ19.71
ꢂ19.74
ꢂ19.88
9
10
11
12
13
0.13
ꢂ0.10
1.04

D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
119
stirred overnight. After the reaction had completed as monitored by
TLC (CHCl3:CH3OH ¼ 9:1) and HPLC, the DMF was evaporated and
the residue was dissolved in ethyl acetate and washed with citric
acid solution (10%), followed by sodium bicarbonate solution (4%)
and brine, and then dried over anhydrous MgSO4. The remaining
ethyl acetate was evaporated and the residue was purified by silica
gel chromatography using a mixture of DCM and methanol (149:1)
to yield compound 1 (256 mg, 91.5%; Overall yield: 46.1%) as a whitþe
solid. M.p. 111e113 ꢁC. HPLC purity: 98.2%. HRMS: [M þ Na]
582.1866 for C31H33N3O3S2 (calcd 582.1861); 1H NMR (CDCl3,
129.23, 129.12, 128.59, 128.48, 128.32, 128.12, 128.07, 124.67,
121.34, 120.56, 119.65, 67.20, 55.80, 38.77, 32.41, 29.73, 28.73,
28.70, 28.15, 25.16.
4.1.7. Synthesis of compound 5
This compound was synthesized according to the procedure
reported for compound 1 using 4-aminobiphenyl instead of 1-
naphthylamine. Compound 5 yield 45.7% as a white solid. M.p
114e116 ꢁC. HPLC purity: 98.1%. HRMS: [M þ Na]þ 608.2012 for
C33H35N3O3S2 (calcd 608.2018); 1H NMR (CDCl3, 400 MHz,
d, ppm)
400 MHz,
d
, ppm) 8.44 (1H, d, J ¼ 4.1 Hz), 7.83 (3H, m), 7.69 (1H, d,
8.45 (1H, d, J ¼ 2.5 Hz), 8.30 (1H, s), 7.71 (1H, d, J ¼ 7.7 Hz), 7.63 (1H,
t, J ¼ 7.7 Hz), 7.54 (5H, m), 7.42 (2H, t, J ¼ 7.0 Hz), 7.35 (5H, m), 7.26
(1H, s), 7.07 (1H, t, J ¼ 4.4 Hz), 5.41 (1H, d, J ¼ 5.8 Hz), 5.14 (2H, t,
J ¼ 11.7 Hz), 4.28 (1H, s), 2.76 (2H, t, J ¼ 6.4 Hz), 1.95 (1H, t,
J ¼ 8.0 Hz), 7.64 (2H, m), 7.44 (2H, t, J ¼ 4.6 Hz), 7.38 (1H, t, J ¼ 7.8 Hz),
7.28 (5H, m), 7.04 (1H, t, J ¼ 5.7 Hz), 5.68 (1H, d, J ¼ 7.7 Hz), 5.10 (2H, t,
J ¼ 12.9 Hz), 4.49 (1H, s), 2.75 (2H, t, J ¼ 7.1 Hz),1.97 (1H, t, J ¼ 6.7 Hz),
1.74 (1H, t, J ¼ 7.7 Hz), 1.66 (2H, t, J ¼ 6.8 Hz), 1.36 (7H, m); 13C NMR
J ¼ 7.6 Hz), 1.64 (4H, m), 1.38 (6H, m); 13C NMR (CDCl3, 400 MHz,
d,
(CDCl3, 400 MHz,
d, ppm) 170.78, 160.59, 156.80, 153.33, 149.58,
ppm) 170.00, 160.61, 156.66, 149.61, 140.46, 137.40, 137.05, 136.83,
135.97, 128.80, 128.64, 128.37, 128.14, 127.62, 127.17, 126.87, 120.61,
120.28, 119.73, 67.42, 55.79, 38.67, 31.99, 28.65, 28.07, 25.37.
137.05, 135.97, 134.07, 132.03, 128.63, 128.59, 128.30, 128.04, 127.20,
126.39, 126.02, 125.96, 125.61, 120.93, 120.90, 120.59, 119.67, 67.37,
55.75, 38.77, 32.13, 28.84, 28.73, 28.20, 25.54.
4.1.8. Synthesis of compound 6
4.1.4. Synthesis of compound 2
Z-Am8(S2Py)-OH was obtained as described above. Z-
Am8(S2Py)-OH (217 mg) was coupled with 4-biphenylmethanol
(111 mg) in DMF by the aid of DMAP (6.1 mg) and dicyclohex-
ylcarbodiimide (DCC) (124 mg) and stirred for 8 h. After the reaction
had completed as monitored by TLC (CHCl3:CH3OH ¼ 9:1) and HPLC,
the DMF was evaporated and the residue was dissolved in ethyl
acetate and washed with citric acid solution (10%), followed by
sodium bicarbonate solution (4%) and brine, and then dried over
anhydrous MgSO4. The remaining ethyl acetate was evaporated and
the residuewas purified by silica gel chromatography using a mixture
of DCM and methanol (149:1) to yield compound 6 (272 mg, 90.7%;
Overall yield: 46.0) as a colorless oil. HPLC purity: 97.9%. HRMS:
[M þ Na]þ 623.2001 for C34H36N2O4S2 (calcd 623.2014); 1H NMR
This compound was synthesized according to the procedure
reported for compound 1 using cyclohexylamine instead of 1-
naphthylamine. Compound 2 yield 41.1% as a wþhite solid. M.p
78e80 ꢁC. HPLC purity: 97.8%. HRMS: [M þ Na] 538.2165 for
1
C27H37N3O3S2 (calcd 538.2174); H NMR (CDCl3, 400 MHz,
d, ppm)
8.46 (1H, d, J ¼ 4.5 Hz), 7.71 (1H, d, J ¼ 8.0 Hz), 7.65 (1H, t, J ¼ 7.1 Hz),
7.35 (4H, m), 7.26 (1H, s), 7.08 (1H, t, J ¼ 5.8 Hz), 5.77 (1H, d,
J ¼ 8.6 Hz), 5.29 (1H, d, J ¼ 6.8 Hz), 5.10 (1H, s), 4.03 (1H, d,
J ¼ 7.2 Hz), 3.74 (1H, s), 2.77 (1H, t, J ¼ 7.2 Hz), 1.87 (1H, t, J ¼ 7.9 Hz),
1.78 (1H, t, J ¼ 7.5 Hz), 1.67 (3H, m), 1.61 (5H, m) 1.30 (9H, m), 1.13
(3H, m); 13C NMR (CDCl3, 400 MHz,
d, ppm) 170.47, 160.60, 156.13,
149.61, 137.01, 136.25, 128.58, 128.25, 128.11, 120.57, 119.64, 67.04,
55.09, 48.32, 38.76, 33.06, 32.93, 32.72, 29.74, 28.76, 28.70, 28.16,
25.47, 25.24, 24.77.
(CDCl3, 400 MHz,
d, ppm) 8.45 (1H, d, J ¼ 4.4 Hz), 7.68 (1H, d,
J ¼ 8.1 Hz), 7.58 (5H, m), 7.42 (4H, q, J ¼ 7.8 Hz), 7.35 (6H, m), 7.06 (1H,
t, J ¼ 6.5 Hz), 5.11 (1H, s), 2.71 (1H, t, J ¼ 7.2 Hz),1.61 (5H, m),1.26 (9H,
4.1.5. Synthesis of compound 3
m), 0.98 (2H, d, J ¼ 6.7 Hz); 1H NMR (CDCl3, 400 MHz,
d C
, ppm); 13
This compound was synthesized according to the procedure
reported for compound 1 using 2-amino-4-phenylthiazole hydro-
bromide monohydrate instead of 1-naphthylamine. Compound 3
yield 44.0% as a light yellow solid. M.p 76e78 ꢁC. HPLC purity:
NMR (CDCl3, 400 MHz, d, ppm) 172.43, 167.71, 160.60, 155.87, 149.61,
141.48, 140.51, 136.97, 136.24, 134.28, 132.39, 130.95, 128.87, 128.57,
128.24, 128.16, 127.56, 127.38, 127.13, 120.53, 119.59, 71.83, 67.04,
66.91, 53.90, 38.82, 32.62, 29.74, 28.71, 28.18, 27.75, 24.93, 19.20.
98.5%. HRMS: [M
þ
Na]þ 615.1528 for C30H32N4O3S3 (calcd
, ppm) 8.46 (1H, d,
615.1534); 1H NMR (CDCl3, 400 MHz,
d
4.1.9. Synthesis of compound 7
J ¼ 4.7 Hz), 7.79 (1H, d, J ¼ 1.3 Hz), 7.76 (1H, s), 7.70 (1H, d,
J ¼ 8.2 Hz), 7.63 (1H, t, J ¼ 7.9 Hz), 7.36 (7H, m), 7.26 (1H, s), 7.14 (1H,
s), 7.07 (1H, t, J ¼ 5.8 Hz), 5.47 (1H, d, J ¼ 7.7 Hz), 5.16 (1H, q,
J ¼ 12.1 Hz), 4.52 (1H, d, J ¼ 7.0 Hz), 2.75 (2H, t, J ¼ 7.2 Hz), 1.92 (1H,
s),1.66 (3H, m),1.34 (7H, m) 1.00 (1H, d, J ¼ 6.7 Hz); 13C NMR (CDCl3,
This compound was synthesized according to the procedure
reported for compound 1 using 9-aminofluorene hydrochloride
instead of 1-naphthylamine. Compound 7 yield 40.8% as a light
yellow solid. M.p 130e132 ꢁC. HPLC purity: 98.6%. HRMS:
[M þ Na]þ 620.2023 for C34H35N3O3S2 (calcd 620.2018); 1H NMR
400 MHz,
d
, ppm) 170.28, 167.72, 160.59, 157.80, 156.49, 149.56,
(CDCl3, 400 MHz, d, ppm) 8.42 (1H, s), 7.64 (4H, m), 7.48 (2H, s), 7.37
137.07, 135.79, 133.57, 132.38, 130.95, 128.83, 128.62, 128.40, 128.30,
126.15,120.61,119.76,107.96, 71.83, 67.72, 55.13, 38.69, 32.06, 29.74,
28.67, 28.06, 27.75, 25.27, 19.20.
(3H, t, J ¼ 7.3 Hz), 7.30 (6H, m), 7.05 (1H, s), 6.51 (1H, s), 6.17 (1H, d,
J ¼ 8.0 Hz), 5.43 (1H, s), 4.92 (1H, s), 4.23 (1H, s), 2.77 (1H, t,
J ¼ 6.3 Hz),1.87 (1H, t, J ¼ 5.0 Hz), 1.66 (3H, s), 1.35 (8H, m); 13C NMR
(CDCl3, 400 MHz, d, ppm) 172.55, 160.57, 156.05, 149.59, 144.06,
4.1.6. Synthesis of compound 4
143.86, 140.61, 136.98, 136.07, 128.79, 128.54, 128.24, 128.08, 127.85,
125.07, 125.00, 120.55, 120.05, 119.65, 67.03, 55.14, 54.70, 38.76,
32.91, 29.73, 28.77, 28.70, 28.19, 25.27.
This compound was synthesized according to the procedure
reported for compound 1 using 2-aminobiphenyl instead of 1-
naphthylamine. Compound 4 yield 47.2% as a white solid. M.p
112e114 ꢁC. HPLC purity: 98.8%. HRMS: [M þ Na]þ 608.2026 for
4.1.10. Synthesis of compound 8
C33H35N3O3S2 (calcd 608.2018); 1H NMR (CDCl3, 400 MHz,
d
, ppm)
This compound was synthesized according to the procedure
reported for compound 6 using 9-fluorenemethanol instead of 4-
biphenylmethanol. Compound 8 yield 44.5% as a light yellow oil.
HPLC purity: 98.3%. HRMS: [M þ Na]þ 635.2026 for C35H36N2O4S2
8.44 (1H, d, J ¼ 3.1 Hz), 8.29 (1H, d, J ¼ 7.9 Hz), 7.72 (2H, t,
J ¼ 7.6 Hz), 7.63 (1H, t, J ¼ 7.0 Hz), 7.34 (9H, m), 7.26 (2H, d,
J ¼ 7.1 Hz), 7.21 (1H, d, J ¼ 7.3 Hz), 7.06 (1H, t, J ¼ 5.2 Hz), 5.15 (1H,
s), 5.07 (1H, q, J ¼ 11.7 Hz), 4.08 (1H, s), 2.75 (1H, t, J ¼ 6.8 Hz), 1.61
(calcd 635.2014); 1H NMR (CDCl3, 400 MHz,
d, ppm) 8.47 (1H, d,
(4H, m), 1.25 (9H, m); 13C NMR (CDCl3, 400 MHz,
d
, ppm) 169.64,
J ¼ 4.6 Hz), 7.75 (2H, m), 7.58 (2H, m), 7.41 (1H, m), 7.35 (2H, t,
160.57, 155.96, 149.61, 137.85, 136.99, 136.02, 134.18, 132.60, 130.09,
J ¼ 8.8 Hz), 7.26 (8H, m), 7.07 (1H, t, J ¼ 6.9 Hz), 5.12 (1H, s), 4.51 (1H,

120
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
d, J ¼ 6.8 Hz), 4.20 (1H, t, J ¼ 5.1 Hz), 2.78 (1H, t, J ¼ 6.9 Hz), 1.67 (2H,
t, J ¼ 8.2 Hz), 1.55 (10H, m), 1.26 (3H, s); 13C NMR (CDCl3, 400 MHz,
4-biphenylmethanol. Compound 12 yield 41.9% as a light yellow oil.
HPLC purity: 98.7%. HRMS: [M þ Na]þ 597.1863 for C32H34N2O4S2
d
, ppm) 172.28, 160.48, 155.96, 150.25, 149.59, 147.80, 142.55,
(calcd 597.1858); 1H NMR (CDCl3, 400 MHz,
d, ppm) 8.46 (1H, d,
138.53, 137.05, 136.15, 128.58, 128.47, 128.29, 128.13, 127.58, 123.88,
120.98,120.60,119.64, 67.12, 65.50, 53.91, 38.72, 36.62, 32.34, 29.72,
28.92, 28.66, 28.61, 28.35, 28.11, 25.11, 25.00.
J ¼ 4.3 Hz), 8.00 (1H, d, J ¼ 8.0 Hz), 7.87 (2H, t, 7.4 Hz), 7.69 (1H, d,
J ¼ 8.1 Hz), 7.62 (1H, t, J ¼ 7.2 Hz), 7.52 (2H, m), 7.44 (1H, t,
J ¼ 7.3 Hz), 7.34 (4H, m), 7.26 (1H, s), 7.11 (1H, t, J ¼ 4.8 Hz), 7.06 (1H,
t, J ¼ 6.5 Hz), 5.68 (1H, s), 5.58 (1H, s), 5.28 (1H, d, J ¼ 8.2 Hz), 5.01
(1H, s), 4.41 (1H, d, J ¼ 6.1 Hz), 2.70 (1H, t, J ¼ 7.4 Hz), 1.64 (6H, m),
4.1.11. Synthesis of compound 9
Z-Am8(S2Py)-OH was obtained as described above. Z-
Am8(S2Py)-OH (217 mg) was coupled with 4-nitrobenzyl bromide
1.26 (2H, s), 1.15 (3H, m), 1.06 (1H, m); 13C NMR (CDCl3, 400 MHz,
d,
ppm) 172.41, 160.63, 155.82, 149.60, 137.43, 136.97, 136.23, 133.75,
131.61, 130.80, 129.66, 128.80, 128.55, 128.21, 128.13, 128.02, 126.70,
126.07, 125.26, 123.48, 120.53, 119.60, 67.01, 65.56, 53.91, 38.83,
32.53, 30.95, 28.64, 28.58, 28.08, 24.77.
(130 mg) in cooled DMF, triethylamine (84 mL) was added and the
solution was stirred overnight. After the reaction had completed as
monitored by TLC (CHCl3:CH3OH ¼ 9:1) and HPLC, the DMF was
evaporated and the residue was dissolved in ethyl acetate and
washed with citric acid solution (10%), followed by sodium bicar-
bonate solution (4%) and brine, and then dried over anhydrous
MgSO4. The remaining ethyl acetate was evaporated and the residue
was purified by silica gel chromatography using a mixture of DCM
and methanol (149:1) to yield compound 9 (273 mg, 95.8%; Overall
yield: 48.3%) as a light yellow oil. HPLC purity: 97.7%. HRMS:
[M þ Na]þ 592.1560 for C28H31N3O6S2 (calcd 592.1552); 1H NMR
4.1.15. Synthesis of compound 13
This compound was synthesized according to the procedure
reported for compound 6 using 9-anthracenemethanol instead of
4-biphenylmethanol. Compound 13 yield 43.8% as a light yellow oil.
HPLC purity: 98.1%. HRMS: [M þ Na]þ 647.2030 for C36H36N2O4S2
(calcd 647.2014); 1H NMR (CDCl3, 400 MHz,
d, ppm) 8.51 (1H, s),
8.45 (1H, d, J ¼ 4.5 Hz), 8.31 (1H, d, J ¼ 8.8 Hz), 8.02 (1H, d,
J ¼ 8.4 Hz), 7.66 (1H, d, J ¼ 8.1 Hz), 7.58 (2H, m), 7.49 (2H, t,
J ¼ 7.8 Hz), 7.32 (5H, m), 7.26 (1H, s), 7.05 (1H, t, J ¼ 5.4 Hz), 6.33
(1H, d, J ¼ 12.6 Hz), 6.11 (1H, d, J ¼ 12.5 Hz), 5.25 (1H, d, J ¼ 7.8 Hz),
5.07 (2H, t, J ¼ 12.1 Hz), 4.39 (1H, d, J ¼ 6.1 Hz), 2.63 (1H, t,
J ¼ 7.2 Hz), 1.63 (4H, m), 1.45 (1H, m), 1.26 (2H, s), 1.04 (6H, m); 13C
(CDCl3, 400 MHz,
d
, ppm) 8.46 (1H, d, J ¼ 4.7 Hz), 8.19 (2H, d,
J ¼ 8.3 Hz), 7.70 (1H, d, J ¼ 8.0 Hz), 7.64 (1H, t, J ¼ 6.6 Hz), 7.50 (2H, d,
J ¼ 8.3 Hz), 7.34 (5H, m), 7.07 (1H, t, J ¼ 5.1 Hz), 5.26 (2H, s), 5.11 (2H,
s), 4.42 (1H, d, J ¼ 5.2 Hz), 2.76 (1H, t, J ¼ 7.2 Hz),1.83 (1H, t, J ¼ 7.3 Hz),
1.66 (3H, m),1.32 (8H, m); 13C NMR (CDCl3, 400 MHz,
d, ppm) 172.28,
160.48, 155.96, 149.59, 147.80, 142.55, 137.05, 136.15, 128.58, 128.47,
128.29, 128.13, 123.88, 120.60, 119.64, 67.12, 65.50, 53.91, 38.72,
36.62, 32.34, 28.92, 28.66, 28.61, 28.35, 28.11, 25.11, 25.00.
NMR (CDCl3, 400 MHz, d, ppm) 172.70, 160.65, 155.80, 149.59,
136.96, 136.23, 131.35, 131.03, 129.52, 129.19, 128.54, 128.19, 128.12,
126.82, 125.58, 125.20, 123.81, 120.51, 119.57, 66.98, 59.86, 53.91,
38.81, 32.51, 29.73, 28.55, 28.49, 28.01, 24.72, 19.19.
4.1.12. Synthesis of compound 10
This compound was synthesized according to the procedure
reported for compound 9 using 3-nitrobenzyl bromide instead of 4-
nitrobenzyl bromide. Compound 10 yield 47.0% as a light yellow oil.
HPLC purity: 98.5%. HRMS: [M þ Na]þ 592.1564 for C28H31N3O6S2
4.2. HDAC-inhibitory assay
HDAC-inhibitory activities of all compounds were assayed with
the Biomol AK511 (containing human HDAC1) and AK500 (con-
taining HeLa nuclear extract, rich in HDAC1 and HDAC2) kits to
determine IC50 values against HDAC1 and HDAC2, and the assay
was performed according to the manufacturer’s instructions in
triplicate and repeated three times. The disulfide bond of each
compound was reduced by dithiothreitol (DTT) to give a thiol group
that acts as metal binding domain before the assay.
(calcd 592.1552); 1H NMR (CDCl3, 400 MHz,
d, ppm) 8.46 (1H, d,
J ¼ 4.5 Hz), 8.21 (2H, d, J ¼ 10.2 Hz), 7.70 (3H, m), 7.55 (1H, t,
J ¼ 7.7 Hz), 7.35 (5H, m), 7.07 (1H, t, J ¼ 6.6 Hz), 5.26 (2H, d,
J ¼ 3.7 Hz), 5.11 (2H, s), 4.43 (1H, d, J ¼ 5.4 Hz), 2.75 (1H, t,
J ¼ 7.2 Hz), 1.84 (1H, t, J ¼ 7.9 Hz), 1.67 (4H, m), 1.29 (7H, m); 13C
NMR (CDCl3, 400 MHz,
d, ppm) 172.28, 160.54, 155.92, 149.61,
148.38, 137.44, 137.01, 136.15, 134.00, 129.77, 128.57, 128.27, 128.16,
123.42, 122.93, 120.56, 119.62, 67.13, 65.57, 53.88, 38.74, 32.41,
28.67, 28.62, 28.12, 25.06.
4.3. Cell growth inhibitory assay
Growth inhibition was determined using a standard MTT assay.
MCF-7 was cultured in DMEM medium supplemented with 10%
fatal bovine serum. Cell cultures were incubated at 37 ꢁC with 5%
CO2 in air. Cells were diluted to 5e9 ꢀ 104 cells/mL with medium
and plated in a 96-well plate. After overnight incubation, aliquot of
10-time serial dilution of each compound to be evaluated was
4.1.13. Synthesis of compound 11
This compound was synthesized according to the procedure
reported for compound 9 using 2-(bromomethyl)naphthalene
instead of 4-nitrobenzyl bromide. Compound 11 yield 48.1% as
a light yellow oil. HPLC purity: 98.0%. HRMS: [M þ Na]þ 597.1853
for C32H34N2O4S2 (calcd 597.1858); 1H NMR (CDCl3, 400 MHz,
d
,
added, and the plate was incubated for 48 h. Then 200 mL MTT
ppm) 8.45 (1H, d, J ¼ 4.5 Hz), 8.01 (1H, s), 7.70 (3H, t, 3.9 Hz), 7.68
(1H, d, J ¼ 8.0 Hz), 7.62 (1H, t, J ¼ 6.7 Hz), 7.50 (2H, m), 7.44 (1H, d,
J ¼ 8.3 Hz), 7.34 (4H, m), 7.12 (1H, t, J ¼ 5.8 Hz), 7.06 (1H, t,
J ¼ 6.4 Hz), 5.36 (1H, d, J ¼ 12.2 Hz), 5.30 (1H, t, J ¼ 9.8), 5.11 (1H, s),
4.44 (1H, d, J ¼ 5.7 Hz), 2.95 (1H, s), 2.88 (1H, s), 2.69 (1H, t,
J ¼ 7.3 Hz), 1.81 (1H, d, J ¼ 7.8), 1.58 (3H, m), 1.26 (7H, m); 13C NMR
solution diluted in serum free medium (0.5 mg/mL) was added to
each well and the cells were incubated for an additional 4 h fol-
lowed by addition of 200 mL DMSO to dissolve the dark blue crystal
(formazan). The plate was then read at 570 nm using a spectro-
photometer. The IC50 values were determined in triplicate and the
experiment was repeated at least three times.
(CDCl3, 400 MHz, d, ppm) 172.42, 162.56, 160.62, 158.98, 155.88,
149.60, 137.43, 136.97, 136.24, 133.20, 133.13, 132.73, 128.56, 128.22,
128.14, 128.03, 127.75, 126.44, 125.89, 121.13, 120.53, 119.59, 67.29,
67.03, 53.91, 38.81, 36.52, 32.59, 29.74, 28.66, 28.13, 24.89.
4.4. Docking studies
AutoDock4.0 program [22,23] was used in the docking studies.
Initial X-ray structure of HDAC2 was obtained from the Protein Data
Bank (PDB) (PDB ID: 3MAX). The disulfide bonds in the metal biding
domains of inhibitors were changed to thiol groups in docking,
because the thiol group was the active group that interacted with
4.1.14. Synthesis of compound 12
This compound was synthesized according to the procedure
reported for compound 6 using 1-naphthalenemethanol instead of

D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
121
enzymes. The inhibitors were put into the active pocket of HDAC2,
and the grid maps were centered on the inhibitors and comprised
70 ꢀ 70 ꢀ 70 points with 0.375 Å spacing. For all inhibitors, all the
single bonds except the amide bonds were treated as active
torsional bonds. Lamarckian genetic algorithm was used with
equilibrated after 1.5 ns. Hence, atom coordinates for the last 500 ps
were used to analyze the structure in detail.
4.6. Binding free energy calculations
a
maximum number of 25,000,000 energy evaluations and
For MM-PBSA methodology, snapshots were taken at 5 ps
intervals from the corresponding 500 ps MD trajectories. Explicit
water molecules were removed from the snapshots. The energy
components were calculated using very large cutoff (999 Å). The
a maximum number of 5000 generations for each run. The other
parameters were all set as default in AutoDock4.0. For all calcula-
tions, 150 docking runs were performed. Results differed by less
than 2.0 Å in positional root-mean-square deviation (RMSD) were
clustered together. The best docking results were chosen for MD
simulations. The LigPlot program [24] was employed to analyze the
docking results that focused on hydrogen bonding and hydrophobic
interactions.
binding free energy,
Gb Emm
D
Gbind
S (1) where
,
was estimated as follows:
D
¼
D
þ
D
Gsol ꢂ T
D
D
Emm is the molecular
ele
int
vdw
ele
mechanics energy given by
DEmm
¼
D
E
þ
D
E
(2) where E
D
int
int
vdw
and
D
E
represent the HDAC2-inhibitor electrostatic and van der
int
Waals interactions, respectively. The solvation free energy (DGsol)
was estimated as the sum of electrostatic solvation free energy
4.5. MD simulations
D
( E
)
and non-polar solvation free energy
(
D
Gnsoolnpol):
ele
sol
nonpol
sol
HDAC2-compound complexes were used for further MD simu-
lations to understand how the compounds bind to HDAC2 in
explicit aqueous solution. The initial structures for the simulations
were obtained from docked conformations. The AMBER9 [25]
package was used for all molecular mechanics and dynamics
calculations. The standard AMBER ff99 force field [26] parameters
were assigned to enzyme and water atoms, and the general AMBER
force field (GAFF) [27] were used for the inhibitors. Zinc ion was
modeled using the Stote non-bonded model (q ¼ þ2eꢂ, r ¼ 1.7 Å,
ele
sol
ele
D
Gsol
¼
D
G
þ
D
G
(3) where
D
G
was calculated using the
sol
PoissoneBoltzmann (PB) model and
as function of the solvent-accessible surface area [36]:
D
Gnsoolnpol was determined with
a
nonpol
sol
D
G
¼ A þ b where A is the solvent-accessible surface area
g
estimated with the LCPO method [37] and the parameters
g
¼ 0.0072 kcal/(mol Å2) and b ¼ 0.0 kcal/mol. The change in solute
entropy upon complexation, TDSconf, was not estimated in this work.
It seemed that entropy did not contribute much to the relative
binding free energies of the ligands with similar to the same protein,
and in some studies there have been good agreements between the
experimental and calculated relative binding energies [38].
3
¼ 0.67 kcal/mol) [28]. The local hydrogen bonding network
around the histidine residues was checked. His183 of HDAC2 was
assigned as HIE (histidine with hydrogen on its epsivlon nitrogen),
and other histidine residues as HID (histidine with hydrogen on its
delta nitrogen). The force field parameters of all inhibitors were
prepared with the Antechamber module [29] of AMBER9 package.
Atomic charges of the inhibitors were derived with the AM1-BCC
charge method [30]. Hydrogen atoms were added to the crystal-
lographic HDAC2 with the AMBER Leap module. Sodium counter-
ions were added to neutralize the system. The complex was then
soaked in a truncated octahedron box of TIP3P [31] water molecules
with a margin of 10 Å along each dimension. The particle mesh
Ewald (PME) method [32] was applied to treat long-range elec-
trostatic interactions. The cutoff distance for the long-range elec-
trostatic and the van der Waals energy terms were set at 12.0 Å. All
covalent bonds to hydrogen atoms were constrained using the
SHAKE algorithm [33].
Energy minimization was achieved in three steps. In the first
step, movement was allowed only for water molecules and coun-
terions. In the second step, the inhibitor and the amino acid residues
of HDAC2 were all allowed to move, while water molecules, together
with counterions were constrained. In the final step, all atoms were
permitted to move freely. In each step, energy minimization was
executed by the steepest descent method for the first 5000 steps and
by the conjugated gradient method for the subsequent 2500 steps.
Periodic boundary conditions were used. The time steps were 2 fs
during the production dynamics. The temperature was maintained
by rescaling the velocities using the Berendsen weak-coupling
algorithm [34], with a time constant of 2 ps for the heat bath.
After the system was stepwise warmed-up from the initial
0 Ke300 K using the NVT ensemble in 120 ps, molecular dynamics
were performed at a constant temperature of 300 K. Each simulation
was performed for 2 ns under periodic boundary conditions with
NPT ensemble at a constant pressure of 1 atm.
4.7. Binding free energy decomposition
MM-GBSA method was used for the binding energy calculation.
This decomposition was used for molecular mechanics and solva-
tion energies but not for entropies. Snapshots were taken at 5 ps
time intervals from the last 500 ps MD trajectories. Explicit water
molecules were removed from the snapshots. Energy decomposi-
tion was performed for gasphase energies (Egas) and desolvation
free energies (Esol). The gasphase energies include internal energy
(Eint), van der Waals energy (Evdw) and electrostatic energy (Eele).
The polar contribution (DGgb) of desolvation was computed using
a modified GB model developed by Onufriev et al. [39] The LCPO
method [37] was used to calculate the solvent-accessible surface
area (SASA) for the estimation of the non-polar solvation free
energy (DGnp) [36]. The total relative binding free energy of a given
residue can be obtained by summing the contribution of each atom
of this residue. The separate contribution of backbones and side-
chains can be organized from the relevant atoms.
Acknowledgments
This work was supported by The National High Technology
Research and Development Program of China (863 program),
2007AA021604. We thank Dr. Alan K Chang for his help with the
revision of the manuscript.
References
[1] M. Grunstein, Histone acetylation in chromatin structure and transcription,
Nature 389 (1997) 349e352.
The convergence of energies, temperatures, and pressures of the
systems, and the atomic RMSD of HDAC2 and the inhibitor were
used to verify the stability of the systems. Trajectories were
analyzed using the PTRAJ modules [35]. In the present MD simu-
lations, the overall structure of all complexes appeared to be
[2] C. Hildmann, D. Riester, A. Schwienhorst, Histone deacetylasesean important
class of cellular regulators with a variety of functions, Applied Microbiology
and Biotechnology 75 (2007) 487e497.
[3] D.C. Drummond, C.O. Noble, D.B. Kirpotin, Z. Guo, G.K. Scott, C.C. Benz, Clinical
development of histone deacetylase inhibitors as anticancer agents, Annual
Review of Pharmacology and Toxicology 45 (2005) 495e528.

122
D. Huang et al. / European Journal of Medicinal Chemistry 52 (2012) 111e122
[4] A.J.M. De Ruijter, A.H. Van Gennip, H.N. Caron, S. Kemp, A.B.P. Van Kuilenburg,
[21] L.A. Watanabe, B. Jose, T. Kato, N. Nishino, M. Yoshida, Synthesis of L-alpha-
amino-omega-bromoalkanoic acid for side chain modification, Tetrahedron
Letters 45 (2004) 491e494.
Histone deacetylases (HDACs): characterization of the classical HDAC family,
Biochemical Journal 370 (2003) 737e749.
[5] M.S. Finnin, J.R. Donigian, A. Cohen, V.M. Richon, R.A. Rifkind, P.A. Marks,
R. Breslow, N.P. Pavletich, Structures of a histone deacetylase homologue
bound to the TSA and SAHA inhibitors, Nature 401 (1999) 188e193.
[6] J.R. Somoza, R.J. Skene, B.A. Katz, C. Mol, J.D. Ho, A.J. Jennings, C. Luong,
A. Arvai, J.J. Buggy, E. Chi, J. Tang, B.C. Sang, E. Verner, R. Wynands, E.M. Leahy,
D.R. Dougan, G. Snell, M. Navre, M.W. Knuth, R.V. Swanson, D.E. McRee,
L.W. Tari, Structural snapshots of human HDAC8 provide insights into the
class I histone deacetylases, Structure 12 (2004) 1325e1334.
[7] A.R. Gangloff, J.C. Bressi, A.J. Jennings, R. Skene, Y.Q. Wu, R. Melkus, R. De Jong,
S. O’Connell, C.E. Grimshaw, M. Navre, Exploration of the HDAC2 foot pocket:
synthesis and SAR of substituted N-(2-aminophenyl)benzamides, Bioorganic &
Medicinal Chemistry Letters 20 (2010) 3142e3145.
[8] T. Jenuwein, C.D. Allis, Translating the histone code, Science 293 (2001)
1074e1080.
[9] P.A. Marks, R.A. Rifkind, V.M. Richon, R. Breslow, T. Miller, W.K. Kelly, Histone
deacetylases and cancer: causes and therapies, Nature Reviews Cancer 1
(2001) 194e202.
[22] D.S. Goodsell, R. Huey, G.M. Morris, A.J. Olson, A semiempirical free energy
force field with charge-based desolvation, Journal of Computational Chem-
istry 28 (2007) 1145e1152.
[23] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, R.K. Belew,
A.J. Olson, Automated docking using a Lamarckian genetic algorithm and an
empirical binding free energy function, Journal of Computational Chemistry
19 (1998) 1639e1662.
[24] A.C. Wallace, R.A. Laskowski, J.M. Thornton, Ligplot - a program to generate
schematic diagrams of protein ligand interactions, Protein Engineering 8
(1995) 127e134.
[25] D.A. Case, T.A. Darden, T.E. Cheatham, C.L. Simmerling, J. Wang, R.E. Duke,
R. Luo, K.M. Merz, D.A. Pearlman, M. Crowley, R.C. Walker, W. Zhang, B. Wang,
S. Hayik, A. Roitberg, G. Seabra, K.F. Wong, F. Paesani, X. Wu, S. Brozell, V. Tsui,
H. Gohlke, L. Yang, C. Tan, J. Mongan, V. Hornak, G. Cui, P. Beroza,
D.H. Mathews, C. Schafmeister, W.S. Ross, P.A. Kollman, AMBER Version 9,
University of California, San Francisco, CA, 2006.
[26] J.M. Wang, P. Cieplak, P.A. Kollman, The ability of the AMBER ff99, Journal of
Computational Chemistry 21 (2000) 1049e1074.
[27] J.M. Wang, R.M. Wolf, J.W. Caldwell, P.A. Kollman, D.A. Case, Development and
testing of a general amber force field, Journal of Computational Chemistry 25
(2004) 1157e1174.
[28] R.H. Stote, M. Karplus, Zinc-binding in proteins and solution - a simple but
accurate nonbonded representation, Proteins-Structure Function and Genetics
23 (1995) 12e31.
[29] J.M. Wang, W. Wang, P.A. Kollman, Antechamber: an accessory software
package for molecular mechanical calculations, Abstracts of Papers of the
American Chemical Society 222 (2001) U403.
[30] A. Jakalian, B.L. Bush, D.B. Jack, C.I. Bayly, Fast, efficient generation of high-
quality atomic charges. AM1-BCC model: I. Method, Journal of Computa-
tional Chemistry 21 (2000) 132e146.
[31] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Klein,
Comparison of simple potential functions for simulating liquid water, Journal
of Chemical Physics 79 (1983) 926e935.
[32] T. Darden, D. York, L. Pedersen, Particle mesh Ewald e an NLog(N) method for
Ewaldsumsinlargesystems, JournalofChemicalPhysics 98(1993)10089e10092.
[33] J.P. Ryckaert, G. Ciccotti, H.J.C. Berendsen, Numerical-integration of Cartesian
equations of motion of a system with constraints e molecular-dynamics of N-
alkanes, Journal of Computational Physics 23 (1977) 327e341.
[34] H.J.C. Berendsen, J.P.M. Postma, W.F. Vangunsteren, A. Dinola, J.R. Haak,
Molecular-Dynamics with coupling to an external bath, Journal of Chemical
Physics 81 (1984) 3684e3690.
[10] S. Minucci, P.G. Pelicci, Histone deacetylase inhibitors and the promise of
epigenetic (and more) treatments for cancer, Nature Reviews Cancer 6 (2006)
38e51.
[11] H.E. Carraway, S.D. Gore, Addition of histone deacetylase inhibitors in
combination therapy, Journal of Clinical Oncology 25 (2007) 1955e1956.
[12] V.M. Richon, S. Emiliani, E. Verdin, Y. Webb, R. Breslow, R.A. Rifkind,
P.A. Marks, A class of hybrid polar inducers of transformed cell differentiation
inhibits histone deacetylases, Proceedings of the National Academy of
Sciences of the United States of America 95 (1998) 3003e3007.
[13] H. Ueda, T. Manda, S. Matsumoto, S. Mukumoto, F. Nishigaki, I. Kawamura, K.
Shimomura, Fr901228, a Novel Antitumor Bicyclic Depsivpeptide Produced by
Chromobacterium-Violaceum No-968 .3. Antitumor Activities on
Experimental-Tumors in Mice, Journal of Antibiotics, 47 (1994) 315e323.
[14] A. Kuendgen, S. Knipp, F. Fox, C. Strupp, B. Hildebrandt, C. Steidl, U. Germing,
R. Haas, N. Gattermann, Results of a phase 2 study of valproic acid alone or in
combination with all-trans retinoic acid in 75 patients with myelodysplastic
syndrome and relapsed or refractory acute myeloid leukemia, Annals of
Hematology 84 (Suppl. 1) (2005) 61e66.
[15] T. Suzuki, T. Ando, K. Tsuchiya, N. Fukazawa, A. Saito, Y. Mariko, T. Yamashita,
O. Nakanishi, Synthesis and histone deacetylase inhibitory activity of new
benzamide derivatives, Journal of Medicinal Chemistry 42 (1999) 3001e3003.
[16] A. Saito, T. Yamashita, Y. Mariko, Y. Nosaka, K. Tsuchiya, T. Ando, T. Suzuki,
T. Tsuruo, O. Nakanishi, A synthetic inhibitor of histone deacetylase, MS-27-
275, with marked in vivo antitumor activity against human tumors,
Proceedings of the National Academy of Sciences of the United States of
America 96 (1999) 4592e4597.
[35] B. Honig, A. Nicholls, Classical electrostatics in biology and chemistry, Science
268 (1995) 1144e1149.
[17] T.A. Miller, D.J. Witter, S. Belvedere, Histone deacetylase inhibitors, Journal of
Medicinal Chemistry 46 (2003) 5097e5116.
[18] M. Yoshida, M. Kijima, M. Akita, T. Beppu, Potent and specific-inhibition of
mammalian histone deacetylase both invivo and invitro by trichostatin-A,
Journal of Biological Chemistry 265 (1990) 17174e17179.
[19] T. Suzuki, Y. Nagano, A. Kouketsu, A. Matsuura, S. Maruyama, M. Kurotaki,
H. Nakagawa, N. Miyata, Novel inhibitors of human histone deacetylases:
design, synthesis, enzyme inhibition, and cancer cell growth inhibition of
SAHA-based non-hydroxamates, Journal of Medicinal Chemistry 48 (2005)
1019e1032.
[36] D. Sitkoff, K.A. Sharp, B. Honig, Accurate calculation of hydration free-energies
using macroscopic solvent models, Journal of Physical Chemistry 98 (1994)
1978e1988.
[37] J. Weiser, P.S. Shenkin, W.C. Still, Optimization of Gaussian surface calcula-
tions and extension to solvent-accessible surface areas, Journal of Computa-
tional Chemistry 20 (1999) 688e703.
[38] Z.G. Zhou, M. Madrid, J.D. Evanseck, J.D. Madura, Effect of a bound non-nucleoside
RT inhibitor on the dynamics of wild-type and mutant HIV-1 reverse transcrip-
tase, Journal of the American Chemical Society 127 (2005) 17253e17260.
[39] A. Onufriev, D. Bashford, D.A. Case, Modification of the generalized Born
model suitable for macromolecules, Journal of Physical Chemistry B 104
(2000) 3712e3720.
[20] K.T. Hirashima Y, Nishino N, Maeda S, Nishino TG, Yoshida M., In Proceedings
of 29th European Peptide Symposium (Peptides), (2006) 2.