Structure-activity relationship study of biflavonoids on the dengue virus polymerase DENV-NS5 RdRp

Article abstract of DOI:10.1055/s-0033-1350672

Dengue virus is the worlds most prevalent human pathogenic arbovirus. There is currently no treatment or vaccine, and solutions are urgently needed. We previously demonstrated that biflavonoids from Dacrydium balansae, an endemic gymnosperm from New Caledonia, are potent inhibitors of the Dengue virus NS5 RNA-dependent RNA polymerase. Herein we describe the structure-activity relationship study of 23 compounds: biflavonoids from D. balansae (1-4) and from D. araucarioides (5-10), hexamethyl-amentoflavone (11), cupressuflavone (12), and apigenin derivatives (13-23). We conclude that 1) over the four different biflavonoid skeletons tested, amentoflavone (1) and robustaflavone (5) are the most promising ones for antidengue drug development, 2) the number and position of methyl groups on the biflavonoid moiety modulate their inhibition of Dengue virus NS5 RNA-dependent RNA polymerase, and 3) the degree of oxygenation of flavonoid monomers influences their antidengue potential. Sotetsuflavone (8), with an IC₅₀ = 0.16 μM, is the most active compound of this series and is the strongest inhibitor of the Dengue virus NS5 RNA-dependent RNA polymerase described in the literature. Georg Thieme Verlag KG Stuttgart New York.

Full text of DOI:10.1055/s-0033-1350672

Original Papers 1313
Structure-Activity Relationship Study of Biflavonoids on
the Dengue Virus Polymerase DENV‑NS5 RdRp
Authors
Paul Coulerie1, Mohammed Nour1, Alexandre Maciuk2, Cécilia Eydoux3, Jean-Claude Guillemot3, Nicolas Lebouvier1,
Edouard Hnawia1, Karine Leblanc2, Guy Lewin2, Bruno Canard3, Bruno Figadère2
1
Affiliations
Laboratoire Insulaire du Vivant et de lʼEnvironnement, Université de la Nouvelle-Calédonie, Nouméa, Nouvelle-Calédonie
Laboratoire de Pharmacognosie-Chimie des substances naturelles et chimiothérapies antiparasitaires, BioCIS, Université
2
Paris-Sud, Châtenay-Malabry, France
AFMB, Aix-Marseille Université, Campus de Luminy, Marseille, France
3
Key words
Abstract
tion of flavonoid monomers influences their anti-
dengue potential. Sotetsuflavone (8), with an
IC50 = 0.16 µM, is the most active compound of
this series and is the strongest inhibitor of the
Dengue virus NS5 RNA-dependent RNA polymer-
ase described in the literature.
"
l biflavonoids
!
"
l Dacrydium araucarioides
Dengue virus is the worldʼs most prevalent hu-
man pathogenic arbovirus. There is currently no
treatment or vaccine, and solutions are urgently
needed. We previously demonstrated that bifla-
vonoids from Dacrydium balansae, an endemic
gymnosperm from New Caledonia, are potent in-
hibitors of the Dengue virus NS5 RNA-dependent
RNA polymerase. Herein we describe the struc-
ture-activity relationship study of 23 compounds:
biflavonoids from D. balansae (1–4) and from D.
araucarioides (5–10), hexamethyl-amentoflavone
(11), cupressuflavone (12), and apigenin deriva-
tives (13–23). We conclude that 1) over the four
different biflavonoid skeletons tested, amentofla-
vone (1) and robustaflavone (5) are the most
promising ones for antidengue drug develop-
ment, 2) the number and position of methyl
groups on the biflavonoid moiety modulate their
inhibition of Dengue virus NS5 RNA-dependent
RNA polymerase, and 3) the degree of oxygena-
"
l Dacrydium balansae
"
l Podocarpaceae
"
l antiviral
"
l Dengue virus
"
l NS5 RdRp
"
l SAR study
Abbreviations
!
NC:
New Caledonia
SAR:
structure-activity relationship
Dengue virus
bovine viral diarrhea virus
RNA-dependent RNA polymerase
dimethylformamide
methyl iodide
DENV:
BVDV:
RdRp:
DMF:
MeI:
Supporting information available online at
http://www.thieme-connect.de/ejournals/toc/
plantamedica
received
revised
accepted
March 11, 2013
June 30, 2013
July 5, 2013
Introduction
from amentoflavone (1–3) with high antidengue
potential [6,7]. Herein we present an extended
study of the DENV‑NS5 RdRp inhibitory activity
of biflavonoids. Among the tested biflavonoids,
six were isolated from D. araucarioides (Podocar-
paceae), another endemic gymnosperm from
New Caledonia (5–10), one was obtained by
hemisynthesis (11), and one was commercially
purchased (12). Apigenin (13) and ten related fla-
vonoid monomers (14–23) were also tested on
the DENV‑NS5 RdRp.
!
Since a few decades DENV has become the most
widespread and prevalent human pathogenic ar-
bovirus, and no treatment or vaccine is available
yet [1]. For these reasons, the research for a vac-
cine or chemotherapy is currently a priority for
public health. The non structural protein NS5,
which is essential for virus replication [2] and
highly conserved between the four virus sero-
types of the virus [3], is an appropriate target for
the research of non cytotoxic antidengue com-
pounds [4], particularly because there is no en-
zyme with RNA-dependent RNA polymerase ac-
tivity in humans [5]. In a previous study, the bio-
guided fractionation of the leaf extract obtained
from Dacrydium balansae (Podocarpaceae) led to
the identification of three biflavonoids derived
Bibliography
DOI http://dx.doi.org/
10.1055/s-0033-1350672
Published online August 8, 2013
Planta Med 2013; 79:
1313–1318 © Georg Thieme
Verlag KG Stuttgart · New York ·
ISSN 0032‑0943
Correspondence
Mohammed Nour
Laboratoire Insulaire du Vivant
et de lʼEnvironnement
Université de la Nouvelle-
Caledonie
BP R4, 98851 Nouméa Cedex
Nouvelle-Calédonie
Phone: + 687290252
Fax: + 687254829
Results and Discussion
!
Four different biflavonoid skeletons (1, 4, 5, and
12), characterized by the bond linking the two
apigenin units, were tested. All biflavonoids iso-
nour@univ-nc.nc
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

1314 Original Papers
lated from Dacrydium spp. (1, 4, and 5) showed strong inhibition
oxygenated substitution patterns. Robustaflavone (5), character-
ized by a C3′-C6′′ linkage, is newly described as a strong inhibitor
of the DENV‑NS5 RdRp (IC50 comprised between 0.26 and
1.3 µM). Hinokiflavone (4), with an IC50 = 0.26 µM, is the most po-
tent DENV polymerase inhibitor, but it was cytotoxic on COS and
BHK cells [6]. Robustaflavone (5) also strongly inhibits the
DENV‑NS5 RdRp (CI50 = 0.33 µM) and is not cytotoxic according
to results exposed in the literature [12]. This compound was also
described as a potent HBV replication inhibitor [13]. Amentofla-
vone (1) was previously shown to be a strong and specific noncy-
totoxic inhibitor of the DENV‑NS5 RdRp [6]. Cupressuflavone (12)
was significantly less active (CI50 > 5 µM).
of the DENV‑NS5 RdRp (IC50 = 0.33 µM). Sotetsuflavone (IC50
=
0.16 µM), the 7′′-O-methylamentoflavone (8), is the strongest in-
hibitor of the DENV‑NS5 RdRp among the tested compounds. It is
also one of the strongest nonnucleotide inhibitors of the
DENV‑NS5 RdRp described in the literature so far [7,15,16]. Inhi-
bition assays of robustaflavone and amentoflavone derivatives on
infected cells should be performed to complete the evaluation of
the antidengue potential of these compounds.
Comparison of the DENV‑NS5 RdRp inhibitory activities of amen-
toflavone (1) and its methylated derivatives 2, 3, and 6–11 (see
l Table 1) shows that the position and number of methoxy
Materials and Methods
!
"
groups modulate the biological activity of the compounds. Thus,
all monomethylated compounds (2, 6, 7, and 8) displayed a lower
IC50 on the DENV polymerase (respectively, 0.75, 0.46, 0.64, and
0.16 µM) than amentoflavone (1) (IC50 = 1.3 µM). The presence of
two or three methoxy groups at C-7 and/or C-4 positions of both
apigenin units (3, 9, and 10) appears to slightly decrease the ac-
tivity (IC50 comprised between 1.0 to 3.1 µM). Methylations in
positions 7 and 7′′ on the amentoflavone skeleton seem to be
the most favorable to inhibit the DENV‑NS5 activity, since com-
pounds 6 and 8 were more active than 2 and 7. These slight mod-
ulations of the activity by the number and position of methoxy
groups could be attributed to the modification of the polarity
and of the capacity to establish hydrogen bonds with DENV poly-
merase as suggested in other studies [14]. On the contrary, meth-
ylation of both 5-OH groups clearly decreases the inhibitory ac-
tivity on the DENV-RdRp, since compound 11 displayed an IC50
equal to 50 µM probably due to steric and electronic modifica-
tions of 11 vs. other amentoflavone derivatives, related to the lack
of the well-known hydrogen bond between the 5-OH group and
carbonyl function in 11.
In order to extend our comprehension of the SAR of flavonoids on
the DENV‑NS5 RdRp, apigenin (13) and 10 related flavonoids
(14–23) were tested on the DENV polymerase. These compounds
only weakly inhibited DENV‑NS5, justifying that their inhibitory
activities are, for most of them, only described at the single con-
centration of 50 µM. Again, the number and position of methyla-
tions on the genin influenced the ability of flavonoids to inhibit
the DENV‑NS5 RdRp. However, dimethoxylation was here the
most favorable since the 7,4′-dimethylapigenin (16) was the
strongest inhibitor (92% inhibition at 50 µM) among the tested
flavonoids monomers. The results of polymerase inhibition for
other related flavonoid monomers suggest that the degree and
the position of oxygenation also influence their inhibitory activ-
ity. Thus, quercetin (21), rhamnetin (22), and kaempferol (23),
which inhibit more than 90% of the enzyme activity at 50 µM,
were significantly more active than apigenin (13), pinocembrin
(18), naringenin (19), and galangin (20). Compounds 21, 22, and
23 belong to the 3-flavonol subtype, such as galangin (20). They
are unsubstituted on the lateral ring and present a weaker activ-
ity than 20. These results suggest that the simultaneous presence
of oxygenated substituents on the three rings of the flavone in-
creases the inhibitory activity on the DENV‑NS5 RdRp.
Plant material and crude extracts
Leaves of Dacrydium araucarioides were harvested in Prony in
April 2009 (22°16′02′′S; 166°50′07′′E), in the South province of
New Caledonia (Province Sud authorization N°10919–2009). A
voucher specimen (reference: Cou 11), identified by Dr. Jérome
Munzinger (IRD), was deposited at the herbarium of the IRD cen-
ter of Nouméa.
General experimental procedures
Cupressuflavone (12), apigenin (13), pinocembrin (18), naringe-
nin (19), galangin (20), quercetin (21), rhamnetin (22), and
kaempferol (23) were purchased from Extrasynthèse. The struc-
ture of all isolated compounds from D. araucarioides was deter-
mined by NMR and HRMS data analysis and by comparison with
data from the literature [8–10]. Spectral data are available in the
Supporting Information. 1D and 2D NMR spectra were recorded
on a Bruker AM-300 and Bruker AM-400 instruments, at room
temperature. APCI and ESIMS spectra were recorded on an Agi-
lent MSD G1946D spectrometer and HRESIMS spectra on a
Bruker MicroTOF Q‑II spectrometer. Copies of the original spectra
are available from the corresponding author. Analytical HPLC‑UV
were performed on an Agilent HP1100 instrument equipped
with a Waters Sunfire C18 column (4.6 × 150 mm, 5 µm), at a flow
rate of 1 mL/min. Preparative HPLC was performed on a Waters
Deltaprep equipped with a sunfire RP-18 column (19 × 250 mm,
10 µm; Waters) with a flow rate of 17 mL/min. The purity of all
tested compounds (> 95%) was verified by HPLC‑UV/MS. All
"
chemical structures are given in l Table 1.
Compound isolation
A crude extract was obtained from maceration of the 1200 g
dried leaf powder from D. araucarioides in methanol under
agitation and sonication at room temperature (3 × 4 L, 6 h each).
The methanolic extract was dried and dissolved in 1 L of MeOH
80% to be successively extracted by C6H12 (3 × 1 L) and CH2Cl2
(3 × 1 L). The methylene chloride fraction was reduced to 1 L and
washed with water (2 × 1 L) to obtain the total biflavonoid extract
(20 g). This extract was then subjected to silica gel chromatogra-
phy (63–200 µm; 5 × 60 cm; Merck) eluted with a step gradient of
CH2Cl2/MeOH (90:10; 75:25 and 0:100) to obtain 10 fractions
(F1–F10). F3 (300 mg) was then fractionated by Sephadex LH20
chromatography (2.5 × 40 cm; Sigma-Aldrich) with CH2Cl2/
MeOH 3:1 to obtain 5 fractions (F3.1–F3.5). F3.3 (40 mg) was fi-
nally separated by preparative HPLC on an RP-18 column eluted
with an isocratic mix of TFA(0.5%)/MeOH 22:78 to obtain 9
(14.9 mg) and 10 (3.7 mg). F5 (1.8 g) was fractionated by silica
gel chromatography (63–200 µm; 2.5 × 30 cm; Merck) with a step
gradient of CH2Cl2/MeOH (92:8; 75:25 and 0:100) to obtain 8
The SAR study of biflavonoids and related flavonoid monomers
on the DENV‑NS5 RdRp confirms that the biflavonoid skeleton is
a promising template for the design of an antidengue compound.
Their antidengue potential is modulated by the biflavonoid skel-
eton and by the number and position of methylation groups. We
also correlated the modulation of antiviral potential with the
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

Original Papers 1315
Table 1 Inhibitory activity of biflavonoids and flavonoid monomers against the DENV‑NS5 RdRp. Percentage inhibition of Dengue 2 NS5 RNA-dependent RNA
polymerase (RdRp) by isolated compounds from D. balansae at 50 µg/mL and IC50 of DENV‑NS5 RdRp (in µM). Positive control was measured with the 3′dGTP.
Compounds
Structure
Inhibitory potency on the DENV‑NS5 RdRp
at 50 µM (%)
IC50 (µM)
ꢀ1
100
1.3 0.1
ꢀ2
100
0.75 0.03
ꢀ3
100
3.12 0.09
ꢀ4
100
0.26 0.01
ꢀ5
100
0.33 0.01
ꢀ6
ꢀ7
ꢀ8
100
100
100
0.46 0.01
0.64 0.05
0.16 0.02
continued
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

1316 Original Papers
Table 1 Continued
Compounds
Structure
Inhibitory potency on the DENV‑NS5 RdRp
at 50 µM (%)
IC50 (µM)
ꢀ9
98
1.6 0.1
10
11
12
100
1.0 0.1
52
4
–
80
4
> 5
13
14
15
16
17
> 50
–
31
5
6
2
–
48
–
92
13.8 1.9
21 12
–
continued
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

Original Papers 1317
Table 1 Continued
Compounds
Structure
Inhibitory potency on the DENV‑NS5 RdRp
at 50 µM (%)
IC50 (µM)
18
35
9
–
19
33
4
–
20
21
67
3
0
–
100
3.1 0.4
22
91
8
0
2.2 0.2
23
99
5.5 1.9
3′dGTP
–
100
0.02
fractions (F5.1–F5.8). F5.2 (45 mg) was then separated by prepar-
ative HPLC with an isocratic elution mix of TFA(0.5%)/MeOH
30:70 to obtain 4 (22 mg) and 6 (3.8 mg). F5.6 (28 mg) was also
separated by preparative HPLC with an isocratic elution mix of
TFA(0.5%)/MeOH 30:70 to obtain 7 (3.1 mg) and 8 (0.7 mg).
500 mg of F7 (2.0 g) were separated on a Sephadex LH20 column
eluted with CH2Cl2/MeOH 3:1 to obtain 5 fractions (F7.1-F7.5).
F7.3 (70 mg) was then fractionated by preparative HPLC using an
isocratic mix of TFA(0.5%)/MeOH 30:70 to obtain 1 (36 mg).
Compounds 2 and 3 were previously isolated from D. balansae
[6] and were not detected in D. araucarioides.
from 2 hours for monomethylated compounds to 3 days for per-
methylation. Apigenin derivatives were then separated on a silica
gel column using an isocratic mix of CHCl3/EtOAc/CH3COOH
60:35:5.
Dengue NS5 polymerase assay
DENV‑NS5 RdRp activity was assayed by monitoring the incorpo-
ration of radiolabeled guanosine (Amersham-Bioscience) into a
homopolymeric cytosine RNA template obtained from Amer-
sham-Pharmacia as previously described [6]. Positive and nega-
tive controls consisted of a reaction mix supplemented by 10 µM
[3H]-GTP (> 98%, 5.1 Ci/mmol, from Amersham-Bioscience) and
5% DMSO for the positive control, and 20 mM EDTA or 10 µM 3′
dGTP (> 90%; Trilink Biotech, N-3002–5) for the negative con-
trols.
Preparation of O-methyl flavonoids
O-methylated flavonoids were prepared following a protocol
adapted from the literature [11]. Amentoflavone (1) was placed
in a solution with DMF, a large excess of MeI and K2CO3 for three
days under agitation, to obtain hexametylamentoflavone (11).
Acacetin (14), genkwanin (15), 7,4′-dimethylapigenin (16), and
trimethylapigenin (17) were prepared with the same procedure
using apigenin as a precursor, and the time of the reaction ranged
Supporting information
All spectroscopic data of biflavonoids isolated from Dacrydium
spp. are available as Supporting Information.
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

1318 Original Papers
6 Coulerie P, Eydoux C, Hnawia E, Stuhl L, Maciuk A, Lebouvier N, Canard B,
Figadère B, Guillemot JC, Nour M. Biflavonoids of Dacrydium balansae
with potent inhibitory activity on Dengue 2 NS5 polymerase. Planta
Med 2012; 78: 672–677
7 Noble CG, Shi PY. Structural biology of dengue virus enzymes: towards
rational design of therapeutics. Antiviral Res 2012; 96: 115–126
8 Hodges R. 5,5′-dihydroxy-7,4′,7′′,4′′′-tetramethoxy-8,3′′-biflavone from
Dacrydium cupressinum Lamb. Aust J Chem 1965; 18: 1491–1492
9 Markham RK, Sheppard C, Geiger H. 13C NMR studies of some naturally
occurring amentoflavone and hinokiflavone biflavonoids. Phytochem-
istry 1987; 26: 3335–3337
Acknowledgments
!
This work was supported by the Government of New Caledonia
and grants from the Fond Pacifique. We thank Jerome Munzinger
and the herbarium of IRD in New Caledonia for the authentica-
tion of botanical samples. We are also grateful to Jean Popovici
for the comments and corrections of this manuscript.
10 Ilyas N, Ilyas M, Rahman W, Okigawa M, Kawano N. Biflavones from the
leaves of Araucaria excelsa. Phytochemistry 1978; 17: 987–990
11 Kim J, Park K, Lee C, Chong Y. Synthesis of a complete series of O-methyl
analogues of Naringenin and Apigenin. Bull Korean Chem Soc 2007;
28: 2527–2530
Conflict of Interest
!
The authors state that there is no conflict of interests.
References
12 Lin YM, Chen FC, Lee KH. Hinokiflavone, a cytotoxic principle from Rhus
succedanea and the cytotoxicity of the related biflavonoids. Planta Med
1989; 55: 166–168
1 WHO. Dengue guidelines for diagnosis, treatment, prevention and con-
trol: new edition. Geneva: WHO Press; 2009
13 Lin YM, Zembower DE, Flavin MT, Schure RM, Anderson HM, Korba BE,
Chen FC. Robustaflavone, a naturally occurring biflavanoid, is a potent
non-nucleoside inhibitor of hepatitis B virus replication in vitro. Bioorg
Med Chem Lett 1997; 7: 2325–2328
14 Molfetta FA, Honório KM, Alves CN, Da Silva ABF. A study on the anti-HIV
activity of biflavonoid compounds by using quantum chemical and
chemometric methods. J Mol Struct 2004; 674: 191–197
15 Yin Z, Chen YL, Schul W, Wang QY, Gu F, Duraiswamy J, Kondreddi RR,
Niyomrattanakit P, Lakshminarayana SB, Goh A, Xu HY, Liu W, Liu B,
Lim JYH, Ng CY, Qing M, Lim CC, Yip A, Wang G, Chan WL, Tan HP, Lin K,
Zhang B, Zou G, Bernard KA, Garrett C, Beltz K, Dong M, Weaver M, He H,
Pichota A, Dartois V, Keller TH, Shi PY. An adenosine nucleoside inhib-
itor of dengue virus. Proc Natl Acad Sci U S A 2009; 106: 20435–20439
16 Niyomrattanakit P, Chen YL, Dong H, Yin Z, Qing M, Glickman JF, Lin K,
Mueller D, Voshol H, Lim JYH, Nilar S, Keller TH, Shi PY. Inhibition of
Dengue virus polymerase by blocking of the RNA tunnel. J Virol 2010;
84: 5678–5686
2 Massé N, Davidson A, Ferron F, Alvarez K, Jacobs M, Romette JL, Canard B,
Guillemot JC. Dengue virus replicons: production of an interserotypic
chimera and cell lines from different species, and establishment of a
cell-based fluorescent assay to screen inhibitors, validated by the eval-
uation of ribavirinʼs activity. Antiviral Res 2010; 86: 296–305
3 Bollati M, Alvarez K, Assenberg R, Baronti C, Canard B, Cook S, Coutard B,
Decroly E, De Lamballerie X, Gould EA, Grard G, Grimes JM, Hilgenfeld R,
Jansson AM, Malet H, Mancini EJ, Mastrangelo E, Mattevi A, Milani M,
Moureau G, Neyts J, Owens RJ, Ren J, Selisko B, Speroni S, Steuber H,
Stuart DI, Unge T, Bolognesi M. Structure and functionality in flavivirus
NS-proteins: perspectives for drug design. Antiviral Res 2010; 87: 125–
148
4 Malet H, Massé N, Selisko B, Romette JL, Alvarez K, Guillemot JC, Tolou H,
Yap TL, Vasudevan S, Lescar J, Canard B. The flavivirus polymerase as a
target for drug discovery. Antiviral Res 2008; 80: 23–35
5 Cerutti H, Casas-Mollano JA. On the origin and functions of RNA-medi-
ated silencing: from protists to man. Curr Genet 2006; 50: 81–99
Coulerie P et al. Structure-Activity Relationship Study… Planta Med 2013; 79: 1313–1318

Copyright of Planta Medica is the property of Georg Thieme Verlag Stuttgart and its content
may not be copied or emailed to multiple sites or posted to a listserv without the copyright
holder's express written permission. However, users may print, download, or email articles for
individual use.