Synthesis, biological evaluation and molecular modeling studies of psammaplin A and its analogs as potent histone deacetylases inhibitors and cytotoxic agents

Article abstract of DOI:10.1016/j.bmcl.2015.12.094

In this study, a concise synthetic method of psammaplin A was achieved from 3-bromo-4-hydroxybenzaldahyde and hydantoin through a four-step synthesis via Knoevenagel condensation, hydrolysis, oximation and amidation in 37% overall yield. A collection of novel psammaplin A analogs focused on the variations of substituents at the benzene ring and modifications at the oxime moiety were synthesized. Among all the synthesized compounds, 5d and 5e showed better HDAC inhibition than psammaplin A and comparable cytotoxicity against four cancer cell lines (PC-3, MCF-7, A549 and HL-60). Molecular docking and dynamics simulation revealed that (i) hydrogen atom of the oxime group interacts with Asp99 of HDAC1 through a water bridged hydrogen bond and (ii) a hydroxyl group is optimal attached on the para-position of benzene, interacting with Glu203 at the entrance to the active site tunnel.

Full text of DOI:10.1016/j.bmcl.2015.12.094

Accepted Manuscript
Synthesis, biological evaluation and molecular modeling studies ofPsammaplin
A and its analogs as potent histone deacetylases inhibitors and cytotoxic agents
Jiachen Wen, Yu Bao, Qun Niu, Jiang Liu, Jinyu Yang, Wanqiao Wang, Tao
Jiang, Yinbo Fan, Kun Li, Jian Wang, Linxiang Zhao, Dan Liu
PII:
S0960-894X(15)30407-8
DOI:
http://dx.doi.org/10.1016/j.bmcl.2015.12.094
BMCL 23455
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
27 June 2015
5 September 2015
28 December 2015
Please cite this article as: Wen, J., Bao, Y., Niu, Q., Liu, J., Yang, J., Wang, W., Jiang, T., Fan, Y., Li, K., Wang,
J., Zhao, L., Liu, D., Synthesis, biological evaluation and molecular modeling studies ofPsammaplin A and its
analogs as potent histone deacetylases inhibitors and cytotoxic agents, Bioorganic & Medicinal Chemistry Letters
(2015), doi: http://dx.doi.org/10.1016/j.bmcl.2015.12.094
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

Synthesis, biological evaluation and molecular modeling studies of Psammaplin A and its
analogs as potent histone deacetylases inhibitors and cytotoxic agents
Jiachen Wena, Yu Baob, Qun Niua, Jiang Liua, Jinyu Yanga, Wanqiao Wanga, Tao Jianga,
Yinbo Fana, Kun Lia, Jian Wanga, Linxiang Zhaoa,*, Dan Liua,*
aKey Laboratory of Structure-Based Drugs Design & Discovery of Ministry of Education,
Shenyang Pharmaceutical University, Shenyang 110016, China
bSchool of Life Sciences and Biopharmaceutics, Shenyang Pharmaceutical University,
Shenyang 110016, China
*Corresponding authors:
Tel: +86-024-23986420
Fax: +86-024-23986420
e-mail: linxiang.zhao@vip.sina.com (L. Zhao)
e-mail: sammyld@163.com (D. Liu)
1

Abstract:
In this study, a concise synthetic method of psammaplin A was achieved from
3-bromo-4-hydroxybenzaldahyde and hydantoin through
a four–step synthesis via
Knoevenagel condensation, hydrolysis, oximation and amidation in 37% overall yield. A
collection of novel psammaplin A analogs focused on the variations of substituents at the
benzene ring and modifications at the oxime moiety were synthesized. Among all the
synthesized compounds, 5d and 5e showed better HDAC inhibition than psammaplin A and
comparable cytotoxicity against four cancer cell lines (PC–3, MCF–7, A549 and HL–60).
Molecular docking and dynamics simulation revealed that (i) hydrogen atom of the oxime
group interacts with Asp99 of HDAC1 through a water bridged hydrogen bond and (ii) a
hydroxyl group is optimal attached on the para-position of benzene, interacting with Glu203
at the entrance to the active site tunnel.
Keywords: Histone deacetylases inhibitors; Psammaplin A; Antiproliferation; Structure
activity relationship; Molecular modeling
2

The acetylation and deacetylation processes of specific lysine residues on H3 or H4
histone tails play a key role in post–translational modification. Histone acetylases (HATs) and
histone deacetylases (HDACs) are two families of enzymes that catalyze such processes.
HDACs also participate in the regulation of non-histone proteins and are important in many
biological processes such as transcriptional regulation, cell-cycle progression, cell survival
and differentiation.1 Eighteen HDAC isozymes in human have been identified and could be
divided into two categories, the zinc–dependent enzymes (classes I, II, and IV) and the
NAD+–dependent enzymes (class III).2,3
Overexpression of classes I and II HDACs have been observed in many types of
cancers and correlated with poor prognosis.4-7 HDACs inhibition could cause proliferation
inhibition, apoptosis, autophagy, differentiation, susceptibility to chemotherapy and migration
inhibition of tumor cells.8 Therefore, developing HDAC inhibitors (HDACIs) as anticancer
drugs have attracted enormous attentions. A number of structurally diverse HDACIs have
been identified, many of which are or derive from natural products.9 Among them, Vorinostat ,
Romidepsin, Belinostat and Panobinostat have been approved for the treatment of cutaneous
or peripheral T-cell lymphoma (CTCL or PTCL, Figure 1) and the general pharmacophore
for HDACIs is composed of the Zn2+ binding group (ZBG), linker, and surface recognition
motif (SRM).
3

Figure 1. Structures of some documented HDACIs and its pharmacophoric characteristics.
Psammaplin A, first isolated from marine sponge in 1987, has been found to have a wide
range of bioactivities, especially for antimicrobial and antitumor activity.10 Recently,
psammaplin A was revealed to be an epigenetic modulator targeting several epigenetic
relevant targets, especially for HDACs.11 As well as Romidepsin, psammaplin A is
recognized as a nature prodrug and structure activity relationships have been enabled by
several research groups.11-15 The disulfide bond and the oxime moieties prove to be essential
for its bioactivities, whereas structural variation at the SRM is tolerated. Based on these
perspectives, a series of easy–to–prepare psammaplin A analogs bearing different substituents
at the benzene ring (5a–j) were synthesized. On the other hand, the oxime moiety, reckoned
as a metabolic issue fragment,16-18 was replaced by methyl oxime (6a–d), pyrazole (11a–c) or
isoxazole moiety (11d–g).
Restrained by the limited resource of natural derived psammaplin A, many research
groups started its total synthesis. Some documented synthetic approaches to prepare
psammaplin A were started from 4-hydroxyphenyl pyruvic acid or L-tyrosine and its
derivatives, which produced psammaplin A in relatively poor yields. In addition, different
4

substituted tyrosine derivatives suffer limited commercial sources for use to explore the
chemical space of SRM.19,20 Previously, we have reported an alternative synthetic route of
psammaplin A in good yield.21 Herein, the synthetic approach was used in the preparation of
its analogs with slight modification (Scheme 1). Briefly, benzalhydantoin 2k was obtained
from hydantoin and 3-bromo-4-hydroxybenzaldehyde 1k through Knoevenagel condensation.
After hydrolysis and oximation, the resulting α-oximic acid 4k was coupled with cystamine
dihydrochloride to yield psammaplin A through a very concise four–step synthesis in 37%
overall yield. This synthetic method was favorable for preparing diverse analogs by taking
appropriate benzaldehydes 1a–j as crude material, yielding 5a–j in good overall yields
(32%–56%). O-methyloxime analogs 6a–d were obtained through methylation of oximes (5b,
5f, 5h and 5i) with methyl iodide in 82% to 91% yields. Detailed structures are given in
Table 1.
Scheme 1. Synthesis of psammaplin A, 5a–j, 6a–d. Reagents and conditions: (a) hydantoin,
ethanolamine, EtOH, H2O, reflux, 5 h; (b) NaOH, H2O, reflux, 12 h; (c) NH2OH·HCl, NaOH,
NaHCO3, H2O, r.t., overnight; (d) cystamine dihydrochloride, EDCI, HOBt, THF, r.t., 24 h; (e)
MeI, DMF, 65°C, 2 h.
The synthesis of phenylpyrazole and phenylisoxazole analogs 11a–g is outlined in
Scheme 2. 4-Methoxyacetophenone 7a was brominated by NBS in the presence of 40%
5

sulfuric acid, yielding 3-bromo-4-methoxyacetophenone 7b as a white powder. Next, after
condensation with dimethyl oxalate, sodium ketoenolate ester 8a and 8b were produced in
good yields. Treatment of the sodium ketoenolate ester with hydroxylamine hydrochloride in
refluxing acetic acid yielded isoxazole ester 9a–b, which were then hydrolyzed to produce
5-phenyl-isoxazole-3-carboxylic acids 10a–b.22 In addition, 8a–b were readily converted into
pyrazole ester 9c–d by coupling with hydrazine hydrate, and then they were hydrolyzed into
5-phenyl-1H-pyrazole-3-carboxylic acids 10c–d. Carboxylic acids 10a–d treated with BBr3
yielded phenols 10e–g. It was unexpected that the demethylation form
5-(3-bromo-4-methoxyphenyl)isoxazole-3-carboxylic acid 10b failed to produce desired
phenol under the exact same protocol. The resulting carboxylic acids 10a–g coupled with
cystamine dihydrochloride afforded the required conjugates 11a–g in 9%–48% overall yields.
Detailed structures are given in Table 2.
Scheme 2. Synthesis of 11a–g. Reagents and conditions: (a) NBS, 40% H2SO4 in water, 60°C,
5 h; (b) dimethyl oxalate, NaOMe, Et2O, r.t., 4 h; (c) NH2OH·HCl or NH2NH2·H2O, AcOH,
6

reflux, 2 h; (d) LiOH, H2O, dioxane, r.t., 16 h; (e) BBr3, DCM, –20°C to r.t., 24 h; (f)
cystamine dihydrochloride, EDCI, HOBt, THF, r.t., 24 h.
For the enzyme-based assay, strong HDAC inhibition (IC50s at low submicromolar range)
was observed in the benzene substituents modification molecules (5a–j). The replacement of
3-bromo-4-hydroxy group from psammaplin A decreased HDAC inhibition activity (5a–c,
5f–j), while the 3-hydroxy-4-methoxy group 5d and 3-methoxy-4-hydroxy group 5e showed
comparable, even better, enzyme inhibition than psammaplin A. Unfortunately, the oxime
modification analogs 6a–d and 11a–g turned out to be less active than psammaplin A,
indicating that the free oxime moiety is responsible for enzymatic inhibition, which is in line
with the previous reports.11-15,23,24 However, an interesting trend can be observed that
phenylpyrazole analogs 11d–g exhibited generally better HDAC inhibition than those
phenylisoxazoles and O-methylated oximes. Although further exploration is required, it is
−H group could partially
apparent that the pyrazole N
replace the free oxime group acting as a
H-bond donor.
Table 1. Biological profiles of compounds 5a–j and 6a–d
IC50 (µM)a
GI50 SD (µM)b
Compd.
R
R’
HDACs
PC–3
A549
MCF–7
HL–60
Psammaplin A
3-Br-4-OH
4-OH
H
H
H
H
H
H
0.05
3.52 0.11
4.50 0.16
6.54 0.20
6.85 0.48
4.78 0.21
4.74 0.43
4.84 0.07
2.26 0.26
0.29 0.02
0.12
0.12
0.13
0.03
0.04
5.33 0.18
5.50 0.25
2.40 0.49
3.27 0.25
3.41 0.15
4.62 0.11
2.68 0.26
3.77 0.65
2.70 0.30
2.47 1.07
0.62 0.07
0.78 0.13
0.67 0.09
0.23 0.02
0.28 0.13
5a
5b
5c
5d
5e
4-OCH3
3,4-OCH2O-
3-OH-4-OCH3
3-OCH3-4-OH
7

4-CH3
4-i-Pr
H
H
0.26
0.21
0.28
0.21
0.27
3.15
7.65
8.59
4.50
6.43 0.19
2.67 0.20
4.61 0.31
5.19 0.44
6.78 0.38
26.80 0.91
> 50
8.35 0.15
4.61 0.18
6.79 0.96
5.42 0.20
4.48 0.51
38.28 0.60
> 50
4.30 0.69
3.91 1.00
4.98 0.50
3.51 0.18
3.48 0.43
16.39 1.70
> 50
0.74 0.12
0.50 0.03
0.87 0.05
0.57 0.16
0.52 0.05
5.18 0.70
11.90 0.51
8.64 0.59
6.55 0.16
5f
5g
5h
5i
2,4-di-Cl
3-Cl
H
H
3-CF3
H
5j
4-OCH3
4-CH3
2,4-diCl
3-Cl
CH3
CH3
CH3
CH3
6a
6b
6c
6d
> 50
47.09 0.56
32.02 0.75
16.53 1.03
14.35 0.58
23.13 0.08
a
Values are averages of at least two independent experiments, SD < 10%; Required pretreatment with TCEP to reduce the
disulfide bond.
b Assays were performed at least three independent experiments, data are shown as mean SD.
Table 2. Biological profiles of compounds 11a–g
IC50 (µM)a
GI50 SD (µM)b
Compd.
R
X
HDACs
8.31
PC–3
HL–60
4-OCH3
4-OH
O
O
> 50
18.33 0.22
11.62 0.54
13.31 0.42
6.45 0.52
4.11 0.21
5.52 0.67
6.14 0.55
11a
11b
11c
11d
11e
11f
3.14
31.67 0.13
38.96 0.09
26.55 0.11
19.75 0.07
16.90 0.04
17.08 0.05
3-Br-4-OCH3
4-OCH3
O
4.65
NH
NH
NH
NH
2.77
4-OH
1.81
3-Br-4-OCH3
3-Br-4-OH
1.86
0.68
11g
a
Values are averages of at least two independent experiments, SD < 10%; Required pretreatment with TCEP to reduce the
disulfide bond.
b Assays were performed at least three independent experiments, data are shown as mean SD.
For the cytotoxicity assay, compounds 5d and 5e with the highest HDAC enzyme
inhibition exhibited comparable antiproliferative activity with psammaplin A against all four
cancer cells. Besides, 5c and 5g with medium potency showed greatest cytotoxicity against
PC–3 cell line. IC50 values of 5c and 5g were 2.40 and 2.67 µM, respectively. The oxime
modification compounds 6a–d and 11a–g, consistent with the enzyme activity, were partial or
complete loss their cytotoxicity.
8

To further validate the target in cell-based model, compounds 5c–f and 6b were chosen
to investigate the acetylating level of histone H3 and α-tubulin in human HL–60 leukemia
cells. As shown in Figure 2, psammaplin A and 5c–f greatly enhanced Ac-H3 at 2 µM instead
of α-tubulin acetylation, suggesting a class I HDACs inhibition.25 However, no Ac-H3
up-regulation was observed in 6b under the assay conditions, which is in accord with the
enzyme-based assay. Vorinostat as a pan-HDACI, the up-regulation of Ac–tubulin was
already visible at 2 µM, while other compounds did not exhibited such an influence.
Figure 2. Effects on histone-H3, acetylated histone-H3, α-tubulin and acetylated α-tubulin of
Vorinostat, psammaplin A, 5c, 5d, 5e, 5f and 6b in HL–60 cells. Cells were treated with
compounds (2 µM) for 18 h.
In light of the above-mentioned bioactivity results, 5d and 5e were found to exhibit both
good HDAC inhibition activity and cytotoxicity. Thus, a docking analysis was carried out to
compare the potential binding modes of psammaplin A and its analogs with HDAC1 (PDB ID:
4BKX).26 As shown in Figure 3A, the reduced psammaplin A binds to HDAC1 and forms
key interactions with the protein in several areas. In detail, the thiol group chelates the Zn2+
9

ion, the oxime group forms a H-bond with Asp99 bridged by a water. As to the surface
recognition motif, the 3-bromo-4-hydroxy phenyl group forms few hydrophobic contacts to
His178, Tyr204 and Phe205 and the hydroxyl group forms additional hydrogen bonds with
Glu203 bridged by a water. The non-covalent interactions between the inhibitor and target
protein are quite stable through 2.5 ns dynamic simulation (supporting information). 5a–j
share a similar binding mode as psammaplin A, as exemplified by 5e shown in Figure 3B. On
the contrast, O-methylated oximes 6a–d as well as oxime cyclized analogs 11a–g fail to
interact with the key amino acid residue Asp99, leading to binding energy reduction.
Figure 3. Stereoview of the simulated docking pose of reduced psammaplin A (A) and 5e (B)
to HDAC1 (4BKX) are represented in orange and green, respectively. Important parts of the
10

enzyme for interaction are shown in magenta sticks. The zinc ion is shown as a light yellow
sphere.
In conclusion, we have developed a concise synthetic method of psammaplin A through
a four–step (37% overall yield) synthesis from 3-bromo-4-hydroxybenzaldahyde and
hydantoin via Knoevenagel condensation, hydrolysis, oximation and amidation. A collection
of novel psammaplin A derivatives focused on the variations of substituents at the benzene ring
and modifications at the oxime moiety were synthesized through this synthetic approach.
Structure-activity relationship study supported that the free oxime group and appropriate
benzene substituents were necessary for high HDACs inhibition and cytotoxicity. Among all
the synthesized compounds, 5d and 5e showed better HDAC inhibition than psammaplin A and
comparable cytotoxicity against several cancer cell lines. Molecular docking and dynamics
simulation indicated that (i) hydrogen atom of the oxime group interacts with Asp99 of HDAC1
through a water bridged hydrogen bond and (ii) a hydroxyl group is optimal attached on the
para-position of benzene, thus interacting to Glu203 at the entrance to the active site tunnel.
11

Acknowledgments
This work was supported by the National Natural Science Foundation of China (Grant No.
81102323, 81473086).
12

References and notes
1 De Ruijter, A. J.; Van Gennip, A. H.; Caron, H. N.; Kemp, S.; Van Kuilenburg, A. B.
Biochem. J. 2003, 370, 737.
2 Yoo, C. B.; Jones, P. A. Nat. Rev. Drug. Discov. 2006, 5, 37.
3 Biel, M.; Wascholowski, V.; Giannis, A. Angew. Chem. Int. Ed. Engl. 2005, 44, 3186.
4 Halkidou, K.; Gaughan, L.; Cook, S.; Leung, H. Y.; Neal, D. E.; Robson, C. N. Prostate
2004, 59, 177.
5 Zhang, Z.; Yamashita, H.; Toyama T.; Sugiura H.; Omoto Y.; Ando, Y.; Mita, K.;
Hamaguchi, M.; Hayashi, S.; Iwase, H. Clin. Cancer Res. 2004, 10, 6962.
6 Choi, J. H.; Kwon, H. J.; Yoon, B. I.; Kim, J. H.; Han, S. U.; Joo, H. J.; Kim, D. Y. Jpn. J.
Cancer Res. 2001, 92, 1300.
7 Zhu, P.; Martin, E.; Mengwasser, J.; Schlag, P.; Janssen, K. P.; Gottlicher, M. Cancer Cell
2004, 5, 455.
8 Witt, O.; Deubzer, H. E.; Milde, T.; Oehme, I. Cancer Lett. 2009, 277, 8.
9 Carafa, V.; Miceli, M.; Altucci, L.; Nebbioso, A. Expert Opin. Ther. Pat. 2013, 23, 1.
10 Piña, I. C.; Gautschi, J. T.; Wang, G. Y.; Sanders, M. L.; Schmitz, F. J.; France D.;
Cornell-Kennon, S.; Sambucetti, L. C.; Remiszewski, S. W.; Perez, L. B.; Bair, K. W.;
Crews, P. J. Org. Chem. 2003, 68, 3866.
11 Garcia, J.; Franci, G.; Pereira, R.; Benedetti, R.; Nebbioso, A.; Rodriguez-Barrios, F.;
Gronemeyer, H.; Altucci, L.; de Lera, A. R. Bioorg. Med. Chem. 2011, 19, 3637.
13

12 Baud, M. G. J.; Leiser, T.; Haus, P.; Samlal, S.; Wong, A. C.; Wood, R. J.; Petrucci, V.;
Gunaratnam, M.; Hughes, S. M.; Buluwela, L.; Turlais, F.; Neidle, S.; Meyer-Almes, F.
J.; White, A. J.; Fuchter, M. J. J. Med. Chem. 2012, 55, 1731.
13 Pereira, R.; Benedetti, R.; Perez-Rodriguez, S.; Nebbioso, A.; Garcia-Rodriguez, J.; Carafa,
V.; Stuhldreier, M.; Conte, M.; Rodriguez-Barrios, F.; Stunnenberg, H. G.; Gronemeyer,
H.; Altucci, L.; de Lera, A. R. J. Med. Chem. 2012, 55, 9467.
14 Hentschel, F.; Sasse, F.; Lindel, T. Org. Biomol. Chem. 2012, 10, 7120
15 Hong, S.; Shin, Y.; Jung, M.; Ha, M. W.; Park, Y.; Lee, Y. J.; Shin, J.; Oh, K. B.; Lee, S.
K.; Park, H. G. Eur. J. Med. Chem. 2015, 96, 218
16 Gross, P. Crit. Rev. Toxicol. 1985, 14, 87.
17 Evelo, C. T.; Spooren, A. A.; Bisschops, R. A.; Baars, L. G.; Neis, J. M. Blood Cells Mol.
Dis. 1998, 24, 280.
18 Boucher, J. L.; Delaforge, M.; Mansuy, D. Biochem. 1994, 33, 7811.
19 Hoshino, O.; Murakata, M.; Yamada, K. Bioorg. Med. Chem. Lett. 1992, 2, 1561.
20 Godert, A. M.; Angelino, N.; Woloszynska-Read, A.; Morey, S. R.; James, S. R.; Karpf, A.
R.; Sufrin, J. R. Bioorg. Med. Chem. Lett. 2006, 16, 3330.
21 Yang, Q. J.; Liu, D.; Sun, D. Y.; Yang, S.; Hu, G. D.; Wu, Z. T.; Zhao, L. X. Molecules
2010, 15, 8784.
22 Zeng, L. F.; Zhang, H. S.; Wang, Y. H.; Sanchez, T.; Zheng, Y. T.; Neamati, N.; Long, Y.
Q. Bioorg. Med. Chem. Lett. 2008, 18, 4521.
23 Baud, M. G. J.; Haus, P.; Leiser, T.; Meyer-Almes, F. J.; Fuchter, M. J. ChemMedChem
2013, 8, 149.
14

24 Baud, M. G. J.; Leiser, T.; Petrucci, V.; Gunaratnam, M.; Neidle, S.; Meyer-Almes, F. J.;
Fuchter, M. J. Beilstein J. Org. Chem. 2013, 9, 81.
25 Schafer, S.; Saunders, L.; Schlimme, S.; Valkov, V.; Wagner, J. M.; Kratz, F.; Sippl, W.;
Verdin, E.; Jung, M. ChemMedChem 2009, 4, 283.
26 Millard, C. J.; Watson, P. J.; Celardo, I.; Gordiyenko, Y.; Cowley, S. M.; Robinson, C. V.;
Fairall, L.; Schwabe, J. W. Mol. Cell 2013, 51, 57.
15

Graphical abstract
16