Aurones as histone deacetylase inhibitors: Identification of key features

Article abstract of DOI:10.1016/j.bmcl.2014.10.019

In this study, a total of 22 flavonoids were tested for their HDAC inhibitory activity using fluorimetric and BRET-based assays. Four aurones were found to be active in both assays and showed IC₅₀ values below 20 μM in the enzymatic assay. Molecular modelling revealed that the presence of hydroxyl groups was responsible for good compound orientation within the isoenzyme catalytic site and zinc chelation.

Full text of DOI:10.1016/j.bmcl.2014.10.019

Bioorganic & Medicinal Chemistry Letters 24 (2014) 5497–5501
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Aurones as histone deacetylase inhibitors: Identification of key
features
Vincent Zwick a, , Alkiviadis-Orfefs Chatzivasileiou b, , Nathalie Deschamps a, , Marina Roussaki b,
Claudia A. Simões-Pires a, Alessandra Nurisso a, Iza Denis c,d,e, Christophe Blanquart c,d,e,f
,
a,
Nadine Martinet g, Pierre-Alain Carrupt a, Anastasia Detsi b, , Muriel Cuendet
⇑
⇑
a School of Pharmaceutical Sciences, University of Geneva, University of Lausanne, 30 Quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland
b Laboratory of Organic Chemistry, School of Chemical Engineering, National Technical University of Athens, Zografou Campus, 15780 Athens, Greece
c INSERM, UMR892, F-44000 Nantes, France
d CNRS, UMR6299, F-44000 Nantes, France
e Université Nantes, F-44000 Nantes, France
f Réseau Epigénétique du Canceropôle Grand Ouest, France
g Institut de chimie, UMR CNRS 7272, UNSA, F-06108 Nice, France
a r t i c l e i n f o
a b s t r a c t
Article history:
In this study, a total of 22 flavonoids were tested for their HDAC inhibitory activity using fluorimetric and
BRET-based assays. Four aurones were found to be active in both assays and showed IC50 values below
20 lM in the enzymatic assay. Molecular modelling revealed that the presence of hydroxyl groups was
Received 31 August 2014
Revised 29 September 2014
Accepted 1 October 2014
Available online 13 October 2014
responsible for good compound orientation within the isoenzyme catalytic site and zinc chelation.
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
Histone deacetylase
Structure activity relationship
Docking
Aurone
Histone acetylation is one of the most important epigenetic pro-
cesses and largely influences gene expression. Increases in chroma-
tin acetylation are associated with transcriptionally active genes
whereas hypoacetylation causes transcriptionally silent genes.
The acetylation degree of histones depends on the enzymatic
activity balance between histone deacetylase (HDAC) and histone
acetylase (HAT) enzymes. In eukaryotes, 18 HDAC isoforms,
classified into 4 classes, have been identified. Classes I, II and IV
are zinc-dependant enzymes. Class III contains the sirtuins (SIRT),
which are NAD-dependent. Epigenetic perturbation due to a
dysfunction in the acetylation process have been involved in the
development of several diseases including cancer,1 cardiovascular,2
and neurodegenerative disorders.3,4 HDAC inhibition has emerged
as an interesting therapeutic strategy to restore the HDAC/HAT
balance. Some HDAC inhibitors (HDACi) are currently under
clinical studies, with others being approved by the FDA for cancer
therapy.5 The general HDACi pharmacophore consists of three
distinct structural parts: the zinc-binding group that chelates a
zinc residue within the active site, the recognition cap group
responsible for HDAC subtype selectivity and a hydrophobic linker
which links these 2 groups.6,7 The first discovered HDACi are from
natural sources. Some of them are very potent such as trichostatin
A8 and have been studied in the context of cancer therapy. Weak
HDAC ligands have also been identified in food and are mostly
associated with cancer chemoprevention. Some examples are
butyrate, organosulfur and organoselenium compounds, as well
as polyphenols.9 Among flavonoids, a common group of plant
phenolic compounds,10 some isoflavones and flavones are known
as in vitro HDAC inhibitors.11–13 Considering this, the HDAC
inhibitory activity of another flavonoid subgroup, the aurones,
was investigated in the present study.
Aurones, (Z)-2-benzylidenebenzofuran-3-(2H)-ones are respon-
sible for the bright yellow color of some popular ornamental flow-
ers, and are biosynthesized from chalcones by the key enzyme
aureusidin synthase.14 Probably due to their scarce occurrence in
nature, aurones have not been extensively studied for their
biological activities although this is rapidly changing. Natural and
synthetic aurones have been shown so far to possess a broad
spectrum of bioactivity including anticancer,15,16 antioxidant,17
⇑
Corresponding authors. Tel.: +30 2107724126 (A.D.), +41 223793386 (M.C.).
E-mail addresses: adetsi@chemeng.ntua.gr (A. Detsi), muriel.cuendet@unige.ch
(M. Cuendet).
These authors equally contributed to the work.
http://dx.doi.org/10.1016/j.bmcl.2014.10.019
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

5498
V. Zwick et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5497–5501
antiparasitic18,19 and neuroprotective.20 Some aurone analogues have
been recently identified by Manjulatha et al. as SIRT1 inhibitors.21 In
the frame of our research involving the design, synthesis and activity
evaluation of aurones, a series of 20 natural and synthetic analogues
were tested for their zinc-dependent HDAC inhibitory activity.
The synthesis of the aurones under study was implemented via
the methodology that was previously applied17,18 and consisted of
the oxidative cyclization of the corresponding chalcones using
mercury(II) acetate as the oxidative agent. The required chalcones
were prepared through the Claisen–Schmidt condensation reaction
between appropriately substituted 20-hydroxy-acetophenones and
benzaldehydes carried out using 20% aqueous KOH in ethanol
(Scheme 1).17,18
The synthesis and structure characterization of aurones 17–19,
21–23, and 25–29 were reported in detail in previously published
work.17,18 Aurones 20, 24, 30, 31 and 32 are new compounds which
were synthesized for the purpose of this study. The hydroxylated
aurones 33–35 were prepared by deprotection of the correspond-
ing methoxylated aurones 17, 23 and 24, respectively, using BBr3
in CH2Cl2 as the deprotecting agent, whereas the natural product
aureusidin (36) was synthesized by removing the methoxy-
methyl-protecting group from aurone 29, as described previously
(Scheme 2). Experimental details and analytical data for the new
compounds are given in Supplementary data.
HDAC inhibitory activity at 100 lM by using a HeLa nuclear extract
in a fluorimetric assay (Table S1). The IC50 of compounds 23, 29, 33,
34, 35 and 36 (Table 1) exhibiting an inhibition P50% was
determined. Among these, only the hydroxylated aurones 33–36
presented an IC50 lower than 20 lM and were further tested with
various HDAC isoforms (HDAC1, HDAC2 and HDAC6). Phenolic
groups on ring B appear to be critical for the activity as the
alkoxylated analogues 17, 23, 24 and 29, respectively, showed no
or significantly lower activity. Aurones 34 and 35 possessing a
catechol substitution pattern were the most active compounds
on each HDAC isoform tested, showing no isoform selectivity,
whereas aurone 33 which is lacking the catechol system on ring
B was less active. The natural product aureusidin (36), which pos-
sesses two additional hydroxyl groups at the 4- and 6-positions of
ring A, exhibited lower activity than aurone 34, therefore implying
that substitution on ring B is more important for the activity.
In order to confirm the HDACi properties of the compounds,
histone H3 acetylation was measured in living cells by using a bio-
luminescent resonance energy transfer technology (BRET)-based
assay.22 An increase in BRET signal reflects an increase in histone
H3 acetylation. Experimental details are given in Supplementary
data. All BRET experiments were validated by using SAHA as a
positive control (Fig. S1A and B). Compounds 23, 30, 33, 34, 35
and 36 increased BRET signal in a dose dependent manner
(Fig. 1) demonstrating their HDACi activity in living cells. A dose
dependent toxicity was observed for all of these compounds
The 20 synthesized aurones as well as chalcones 2 and 3,
precursors of aurones 18 and 19, respectively, were tested for their
OH
R2
O
CHO
R4
R5
CH3
+
R1
R3
R6
R4
(a)
R3
O
OH
O
R5
R6
R2
4
3
(b)
5
2
O
R4
2'
A
1'
R1
R3
R1
6
7
R5
6'
R2
B
3'
5'
4'
R6
1 16
-
17 R1=R2=R3=R4=R5=H, R6=OCH3
18
R1=R2=R3=R4=R5=H, R6=CH3
19 R1=R2=R3=R4=R5=H, R6=Cl
20 R1=R2=R3=R4=R5=H, R6=COOH
R1=R2=R3=R5=R6=H, R4=OCH3
21
22 R1=R2=R3=R4=R6=H, R5=OCH3
23 R1=R2=R3=R4=H, R5=R6=OCH3
R1=R2=R3=R6=H, R4=R5=OCH3
24
25 R1=R3=OCH3, R2=R4=R5=R6=H
26 R1=R3=R6=OCH3, R2=R4=R5=H
R1=R3=R4=OCH3, R2=R5=R6=H
27
28 R1=R3=OCH3, R4=Cl, R2=R5=R6=H
29 R1=R3= R5= R6=OCH2OCH3, R2=R4=H
R1=R3=R4=H, R2=Br, R5=R6=OCH3
30
31 R1=R3=R6=H, R2=Cl, R4=R5=OCH3
32 R1=R3=R4=R5=H, R2=Cl, R6=COOH
Scheme 1. General synthetic route to chalcones 1–16 and aurones 17–32. Reagents and conditions: (a) 20% aq KOH, EtOH, rt. (b) Hg(OAc)2, pyridine, 110 °C, 1 h.

V. Zwick et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5497–5501
5499
R1
R1
O
O
R2
R3
R2
(a) or (b)
R4
R4
R6
O
R3
O
R5
R5
R6
R1=R2=R3=R4=R5=H, R6=OCH3
17
R1=R2=R3=R4=R5=H, R6=OH
33
23 R1=R2=R3=R4=H, R5=R6=OCH3
24 R1=R2=R3=R6=H, R4=R5=OCH3
34 R1=R2=R3=R4=H, R5=R6=OH
35 R1=R2=R3=R6=H, R4=R5=OH
R1=R3= R5= R6=OCH2OCH3, R2=R4=H
29
R1=R3= R5= R6=OH, R2=R4=H
36
Scheme 2. Procedure for the synthesis of hydroxylated aurones 33–36. Reagents and conditions: Route (a) (for aurones 33–35): BBr3, CH2Cl2, 0 °C; Route (b) (for aurone 36):
10% aq HCl, MeOH, reflux 15 min.
Table 1
HDAC inhibition (IC50 in
l
Ma) of selected aurones in a nuclear extract and towards various isoforms
HeLa nuclear extract
Compounds
Class I
Class IIb
HDAC6
HDAC1
HDAC2
O
O
71.0 11.6
NDb
ND
ND
OCH3
OCH3
23
CH3OCH2O
O
O
40.0 6.0
ND
ND
ND
CH3OCH2O
OCH2OCH3
OCH2OCH3
29
O
O
12.6 3.4
129.2 25.6
100.5 13.0
135.9 21.9
33
OH
O
O
8.2 1.2
11.4 0.7
5.1 0.10
26.9 6.9
OH
34
OH
O
OH
18.6 2.5
21.0 4.1
16.4 0.3
26.9 8.5
O
OH
35
OH
O
O
18.5 2.3
38.0 7.7
38.9 6.2
48.5 5.7
HO
OH
O
36
OH
O
OH
N
H
0.0075 0.0017
0.0074 0.0027
0.0282 0.0009
0.0082 0.0024
N
Trichostatin Ac
a
b
c
Results are the means SD of three independent experiments.
ND: not determined.
Positive control.

5500
V. Zwick et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5497–5501
Figure 1. Validation of HDACi activity of compounds in living cells. MeT-5A cells were transfected with phRluc-C1-BrD and pEYFP-C1 histone H3 and treated for 16 h with
increasing doses of the various compounds. BRET signals were measured as described in Experimental section. Results are the means SEM of three independent
experiments.
Figure 2. Compound 34 in complex with HDAC2 and HDAC6. (A) Compound 34 (green sticks) in complex with HDAC2. (B) Compound 34 (green sticks) in complex with
HDAC6.
(Fig. S1C). It was not possible to confirm the HDACi activity of com-
pound 29 due to strong toxicity (Table S1 and Fig. S2). Compound
30, which is different from compound 23 by the presence of a bro-
mine at the 5-position of ring A, showed an HDACi activity only in
the cell-based assay. This could be explained by a cellular process
leading to the active molecule or impacting its stability and then its
intracellular concentration (Fig. 1). Each compound was also tested
for the modulation of sirtuin 2 by thermal shift assay and for
changes in DNA methylation based on specific PCR for RARb. None
of the compound showed any activity (data not shown).
In order to rationalize the results coming from the enzymatic
assays, the four most active aurones (Table 1) were docked into
the catalytic pocket of a representative isoform from class I
(HDAC2)23 and class II (HDAC6). The molecular docking protocol
was first validated in both enzymes through the re-docking of
co-crystallized SAHA within the HDAC2 protein structure.24
Docking calculations revealed the capability of the docking tool
to retrieve the correct binding mode within HDAC2. Poses with
Root Mean Square Deviation (RMSD) values lower that 1 Å with
respect to the ligand of reference were obtained (Fig. S3A). The
same inhibitor was then docked into the catalytic pocket of
HDAC6.25 A unique binding mode, in line with data coming from
the literature,26 was retrieved (Fig. S3B). The docking of 33 in both
HDAC2 and HDAC6 isoforms suggested that this compound,
through the single para-hydroxy substitution, can occupy similar
positions in the catalytic pocket through a monodentate chelation
of the zinc ion. Other hydrophobic interactions stabilized the com-
plexes (further details provided in Supplementary information,
Fig. S4).
Regarding 34 and 35, molecular docking results showed a
conserved hydrophobic interaction network within the HDAC
proteins but a different hydrophilic one. The aurone 34–HDAC2

V. Zwick et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5497–5501
5501
complex presented a monodentate zinc chelation via the 40-OH
group and polar interactions between the 30-OH and the His146
side chain (Fig. 2A). Through the 40- and 30-OH groups, 34
interacted with the zinc ion and Gly619 of the HDAC6 catalytic
pocket, respectively (Fig. 2B). Aurone 35 interacted via 20-OH with
the catalytic zinc ion of HDAC2 whereas the 30-OH formed a
hydrogen bond with the Gln265 side chain (Fig. S5A). In HDAC6,
a monodentate zinc chelation via the 30-OH group and polar con-
tacts between the 20-OH group and Asp742 were observed
(Fig. S5B). Aureusidin (36) was the most substituted compound
among the active ones (Table 1). In HDAC2, a convergent docking
pose highlighted a hydrophilic network between 30-OH and the
catalytic metal, between 40-OH and the His146 side chain, and
between 6-OH and the Leu276 backbone (Fig. S6A). Whereas the
docking algorithm identified only one possible position of 36 in
the HDAC2 pocket, two statistically represented binding modes
were observed in HDAC6. The first one was in line with the
position adopted by compounds 33–35 (Fig. S6B). The second
one, on the contrary, displayed a flip of the ligand within the
pocket, with the hydroxy groups of ring A pointing toward the
zinc binding domain (Fig. S6C). The four most active aurones
and la Ligue Nationale Contre le Cancer (ID Grant). A.N. is also grate-
ful to the program ‘Boursière d’Excellence’ of the University of
Geneva.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
10.019.
References and notes
1. Khan, O.; La Thangue, N. B. Immunol. Cell Biol. 2012, 90, 85.
2. Pons, D.; de Vries, F. R.; van den Elsen, P. J.; Heijmans, B. T.; Quax, P. H. A.;
Jukema, J. W. Eur. Heart J. 2009, 30, 266.
3. Fischer, A.; Sananbenesi, F.; Mungenast, A.; Tsai, L. H. Trends Pharmacol. Sci.
2010, 31, 605.
4. Simões-Pires, C. A.; Zwick, V.; Nurisso, A.; Schenker, E.; Carrupt, P.-A.; Cuendet,
M. Mol. Neurodegen. 2013, 8, 7.
5. Wagner, J. M.; Hackanson, B.; Luebbert, M.; Jung, M. Clin. Epigenet. 2010, 1, 117.
6. Suzuki, T. Chem. Pharm. Bull. 2009, 57, 897.
7. Bieliauskas, A. V.; Pflum, M. K. H. Chem. Soc. Rev. 2008, 37, 1402.
8. Khan, N.; Jeffers, M.; Kumar, S.; Hackett, C.; Boldog, F.; Khramtsov, N.; Qian, X.;
Mills, E.; Berghs, S. C.; Carey, N.; Finn, P. W.; Collins, L. S.; Tumber, A.; Ritchie, J.
W.; Jensen, P. B.; Lichenstein, H. S.; Sehested, M. Biochem. J. 2008, 409,
581.
are
hydroxylated
(Z)-2-benzylidenebenzofuran-3-(2H)-one
compounds. Hydroxylation seems to play a key role on HDAC mod-
ulation via zinc chelation. Moreover, docking results of the inactive
compounds (data not shown) confirmed their importance: methyl-
ation or apolar substitutions were responsible for the missing
interactions with the catalytic zinc ion. From the enzymatic data
(Table 1), a slight preferential inhibition of the aurone derivatives
toward class I HDACs can be observed. As elucidated in the past,
HDAC2 and HDAC6 catalytic sites present differences in terms of
residues, shape and size (Fig. S7).25,26 Thus, it can be hypothesized
that this feature may be responsible for a hydrophobic sandwich
effect, stabilizing the linker portion and the ring A of the com-
pounds in the HDAC2 pocket. In fact, by docking, a higher number
of hydrophobic contacts within HDAC2 active site were detected.
Based on the wide range of aurone bioactivity, a series of 20
analogues was tested for their HDAC inhibitory activity. Among
9. Ho, R.; Nievergelt, A.; Simões-Pires, C. A.; Cuendet, M. Stud. Nat. Prod. Chem.
2012, 38, 247.
10. Yao, L. H.; Jiang, Y. M.; Shi, J.; Tomas-Barberan, F. A.; Datta, N.; Singanusong, R.;
Chen, S. S. Plant Foods Hum. Nutr. (Dordrecht, Neth.) 2004, 59, 113.
11. Bontempo, P.; Mita, L.; Miceli, M.; Doto, A.; Nebbioso, A.; De Bellis, F.; Conte,
M.; Minichiello, A.; Manzo, F.; Carafa, V.; Basile, A.; Rigano, D.; Sorbo, S.;
Castaldo Cobianchi, R.; Schiavone, E. M.; Ferrara, F.; De Simone, M.; Vietri, M.;
Cioffi, M.; Sica, V.; Bresciani, F.; de Lera, A. R.; Altucci, L.; Molinari, A. M. Int. J.
Biochem. Cell Biol. 2007, 39, 1902.
12. Son, I. H.; Chung, I.-M.; Lee, S. I.; Yang, H. D.; Moon, H. I. Bioorg. Med. Chem. Lett.
2007, 17, 4753.
13. Pandey, M.; Kaur, P.; Shukla, S.; Abbas, A.; Fu, P.; Gupta, S. Mol. Carcinog. 2012,
51, 952.
14. Ono, E.; Fukuchi-Mizutani, M.; Nakamura, N.; Fukui, Y.; Yonekura-Sakakibara,
K.; Yamaguchi, M.; Nakayama, T.; Tanaka, T.; Kusumi, T.; Tanaka, Y. Proc. Natl.
Acad. Sci. U.S.A. 2006, 103, 11075.
15. Sim, H. M.; Loh, K. Y.; Yeo, W. K.; Lee, C. Y.; Go, M. L. ChemMedChem 2011, 6,
713.
16. Cheng, H.; Zhang, L.; Liu, Y.; Chen, S.; Cheng, H.; Lu, X.; Zheng, Z.; Zhou, G.-C.
Eur. J. Med. Chem. 2010, 45, 5950.
17. Detsi, A.; Majdalani, M.; Kontogiorgis, C. A.; Hadjipavlou-Litina, D.; Kefalas, P.
Bioorg. Med. Chem. 2009, 17, 8073.
these compounds, four exhibited an IC50 lower than 20 lM and
were considered active. Unfortunately, they did not show any
selectivity with the isoforms tested. The activity of the compounds
was confirmed in living cells by using a BRET-based assay. The
presence of hydroxyl groups on ring B was identified to play an
important role for the activity. These results could lead to the
development of more selective and potent HDAC inhibitors.
18. Roussaki, M.; Lima, S. C.; Kypreou, A. M.; Kefalas, P.; Cordeiro da Silva, A.; Detsi,
A. Int. J. Med. Chem. 2012, 196921.
19. Kayser, O.; Chen, M.; Kharazmi, A.; Kiderlen, A. F. Z. Naturforsch., C: J. Biosci.
2002, 57, 717.
20. Shrestha, S.; Natarajan, S.; Park, J. H.; Lee, D. Y.; Cho, J. G.; Kim, G. S.; Jeon, Y. J.;
Yeon, S. W.; Yang, D. C.; Baek, N. I. Bioorg. Med. Chem. Lett. 2013, 23, 5150.
21. Manjulatha, K.; Srinivas, S.; Mulakayala, N.; Rambabu, D.; Prabhakar, M.;
Arunasree, K. M.; Alvala, M.; Basaveswara Rao, M. V.; Pal, M. Bioor. Med. Chem.
Lett. 2012, 22, 6160.
22. Blanquart, C.; Francois, M.; Charrier, C.; Bertrand, P.; Gregoire, M. Curr. Cancer
Drug Targets 2011, 11, 919.
23. Di Micco, S.; Chini, M. G.; Terracciano, S.; Bruno, I.; Riccio, R.; Bifulco, G. Bioorg.
Med. Chem. 2013, 21, 3795.
24. Lauffer, B. E. L.; Mintzer, R.; Fong, R.; Mukund, S.; Tam, C.; Zilberleyb, I.; Flicke,
B.; Ritscher, A.; Fedorowicz, G.; Vallero, R.; Ortwine, D. F.; Gunzner, J.;
Modrusan, Z.; Neumann, L.; Koth, C. M.; Lupardus, P. J.; Kaminker, J. S.; Heise, C.
E.; Steiner, P. J. Biol. Chem. 2013, 288, 26926.
25. Butler, K. V.; Kalin, J.; Brochier, C.; Vistoli, G.; Langley, B.; Kozikowski, A. P.
J. Am. Chem. Soc. 2010, 132, 10842.
Acknowledgments
The authors thank Dr. Daumantas Matulis from the Biotechnol-
ogy Institute of Vilnius, Lituania for the thermal shift assays and
Dr. Pierre-Marie Allard from the University of Geneva, Switzerland
for the MS measurements. This work is part of COST Actions
TD0905: ‘Epigenetics: Bench to Bedside’ and CM1106: ‘Chemical
Approaches to Targeting Drug Resistance in Cancer Stem Cells’.
For the synthetic part of this work, funding by the Empirikion
Foundation is gratefully acknowledged. The authors thank INSERM,
CNRS and Grants from la Ligue Inter-régionale Contre le Cancer
(Comités Départementaux du Grand Ouest: CD85 and CD17)
26. Estiu, G.; Greenberg, E.; Harrison, C. B.; Kwiatkowski, N. P.; Mazitschek, R.;
Bradner, J. E.; Wiest, O. J. Med. Chem. 2008, 51, 2898.