N-sulfonylanthranilic acid derivatives as allosteric inhibitors of dengue viral RNA-dependent RNA polymerase

Article abstract of DOI:10.1021/jm901044z

A novel class of compounds containing N-sulfonylanthranilic acid was found to specifically inhibit dengue viral polymerase. The structural requirements for inhibition and a preliminary structure-activity relationship are described. A UV cross-linking experiment was used to map the allosteric binding site of the compound on the viral polymerase.

Full text of DOI:10.1021/jm901044z

7934 J. Med. Chem. 2009, 52, 7934–7937
DOI: 10.1021/jm901044z
N-Sulfonylanthranilic Acid Derivatives
as Allosteric Inhibitors of Dengue Viral
RNA-Dependent RNA Polymerase
Zheng Yin,†,§ Yen-Liang Chen,†
Ravinder Reddy Kondreddi,† Wai Ling Chan,† Gang Wang,†
Ru Hui Ng,† Joanne Y. H. Lim,† Wan Yen Lee,†
Duraiswamy A. Jeyaraj,† Pornwaratt Niyomrattanakit,†
Daying Wen,† Alex Chao,† J. Fraser Glickman,‡
Hans Voshol,‡ Dieter Mueller,‡ Carsten Spanka,‡
Sigmar Dressler,‡ Shahul Nilar,† Subhash G. Vasudevan,†
Pei-Yong Shi,† and Thomas H. Keller*,†
Figure 1. High throughput hit 1 and N-sulfonylanthranilic acid
lead 2.
Scheme 1. Synthesis of Inhibitors 2 and 6-17a
†Novartis Institute for Tropical Diseases, 10 Biopolis Road, No. 05-01
Chromos, 138670, Singapore and ‡Novartis Pharma AG, Forum 1,
Novartis Campus, CH-4056 Basel, Switzerland. §Current address:
College of Pharmacy and The State Key Laboratory of
Elemento-Organic Chemistry, Nankai University,
Tianjin 300071, P. R. China.
Received July 14, 2009
Abstract: A novel class of compounds containing N-sulfonylan-
thranilic acid was found to specifically inhibit dengue viral poly-
merase. The structural requirements for inhibition and
a
preliminary structure-activity relationship are described. A UV
cross-linking experiment was used to map the allosteric binding site
of the compound on the viral polymerase.
a Reagents and conditions: (a) 3-bromobenzenesulfonyl chloride,
pyridine; (b) Pd(Ph3)4, K2CO3, DMF; (c) LiOH, MeOH.
The World Health Organization has estimated that the four
serotypes of the dengue virus (DENVa) cause 50-100 million
human infections annually worldwide.1 Approximately
500 000 of the infected individuals progress to dengue hemor-
rhagic fever, leading to 22 000 deaths.2 DENV is mosquito-
borne and threatens up to 2.5 billion people in more than
100 endemic countries. Although the morbidity of dengue
infection is low, its socioeconomic impact is high.3 Currently
there is no vaccine or effective antiviral treatment available.4
As a result, development of a safe, effective therapeutic agent
for DENV is urgently needed.
DENV is a member of the genus flavivirus in the family
Flaviviridae.5 The viral genome is a single-stranded RNA
(∼10.7 kb) of positive polarity. The viral genome encodes
three structural proteins and seven nonstructural proteins.
The nonstructural protein 5 (NS5) functions as a methyltrans-
ferase and an RNA-dependent RNA polymerase (RdRp) that
are essentialfor viral replication.6 The RdRpis responsiblefor
the synthesis of both minus-sense and positive-sense viral
RNAs during replication. Similar to other polymerases, the
structure of DENV RdRp exhibits a canonical right-hand
conformation and consists of fingers, palm, and thumb
domains.7 The active site of the enzyme is defined by a
conserved GDD motif in the palm domain.
The DENV RdRp is a preferred antiviral target because of
the established druggability of this class of viral enzymes,8 its
essentiality for viral replication, its conservation among all
four serotypes, and the lack of counterparts in mammalian
cells.9 To date, no non-nucleoside inhibitor of the DENV
RdRp has been reported.
In our search for DENV RdRp inhibitors, we developed a
scintillation proximity assay (SPA), using a DENV-2 full-
lengthrecombinant NS5, polyG asa primer, and biotinylated-
oligo C as a template.10 The assay measured the incorporation
of radioactive 3H-GTP into the polyC/biotinylated-oligo
G complex that was captured on the streptavidin-coated
scintillationbeads. High throughput screening of the Novartis
corporate compound archive led to the identification of
2-[3-(3-methyl-5-oxo-4,5-dihydropyrazol-1-yl)benzenesulfonyl-
amino]benzoic acid (1) withanIC50 of 7.2 μM. Derivatization of
1 rapidly produced N-sulfonylanthranilic acid (2), which
had a IC50 of 0.7 μM (Figure 1). In this communication, we
report the synthesis and biological activity of this new series of
N-sulfonylanthranilic acids.
Scheme 1 depicts the general procedure for the synthesis of
2 and derivatives with a fixed benzylpyrazole fragment.
3-Bromobenzenesulfonyl chloride was reacted with appro-
priate anilines 3 to form sulfonamides 4, and subsequent
coupling with commercially available benzyl pyrazole boronic
ester (5) through a palladium catalyzed Suzuki reaction11
gave the desired N-sulfonylanthranilic acid analogues.
*To whom correspondence should be addressed. Phone: þ65 6722
2908. Fax: þ65 6722 2915. E-mail: thomas.keller@novartis.com.
a Abbreviations: DENV, dengue virus; RdRp, RNA-dependent
RNA polymerase; NS5, nonstructural protein 5; SPA, scintillation
proximity assay; GTP, guanine triphoshpate.
r
pubs.acs.org/jmc
Published on Web 11/06/2009
2009 American Chemical Society

Letter
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 24 7935
Table 2. Replacement of Benzyl on Pyrazole
Scheme 2. Synthesis of Inhibitors 22-29a
compd
R2
R3
IC50 (μM)
2
H
H
H
H
H
H
H
F
benzyl
H
Me
0.7
a Reagents and conditions: (a) CsCO3, CH3CN; (b) pyridine;
(c) Pd(Ph3)4, K2CO3, DMF; (d) LiOH, MeOH.
22
23
24
25
26
27
28
>20
>20
>20
6.77
>20
0.37
0.26
isobutyl
isopentyl
ethylmorpholine
Table 1. Substitution Effect on the Phenyl Ring of N-Sulfonylanthra-
nilic Acids
methyl-2-naphthalene
methyl-2-naphthalene
while substitution at the 4- (13), 6- (11), and 3- (14) position
led to progressively inferior inhibition. On the basis of this
initial structure-activity information, further work focused
on the 5-position of the phenyl ring. Replacing the chlorine
with the stronger electron-withdrawing fluorine (15)
improved the activity slightly, while an electron-donating
methoxy group (16) reduced the activity significantly. These
results suggest that electron-withdrawing groups may
be generally preferred at the 5-position of the phenyl ring.
The results suggest that the electron withdrawing group
increases the acidity of the carboxylic acid which may be
critical for the interaction with the polymerase enzyme. In
agreement withthis hypothesis, significant lossof potency was
observed when a second methoxy group was introduced at the
4-postion (17) (Table 1).
With 2-[3-(1H-pyrazol-4-yl)benzenesulfonylamino]benzoic
acid as a fixed fragment, replacement of the benzyl group on
the pyrazole ring was investigated to identify the minimum
scaffold capable of inhibiting the dengue NS5 polymerase
(Table 2). Removal (22) or replacement of the benzyl group
with small hydrophobic substituents such as methyl (23) on
pyrazole was detrimental to the activity. Interestingly, while
the isobutyl derivative (24) was not able to inhibit the dengue
polymerase, the compound withanisopentyl replacement(25)
(one carbon longer aliphatic chain) showed activity at low
micromolar range. Moreover, no activity at the concentration
tested was observed when a hydrophilic morpholine (26) was
introduced.
compd
R1
R2
IC50 (μM)
2
2-COOH
H
H
0.7
6
H
>20
>20
>20
>20
>20
18.02
0.77
2.56
>20
0.37
9.99
>20
7
2-COOMe
2-CONH2
3-COOH
4-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
2-COOH
H
8
H
9
H
10
11
12
13
14
15
16
17
H
6-Cl
5-Cl
4-Cl
3-Cl
5-F
5-methoxy
4,5-dimethoxy
The carboxylic acidtargetswere produced byhydrolysis ofthe
corresponding methyl esters.
Compounds 22-29 were synthesized as shown in Scheme 2.
Modifications of the pyrazole were achieved by direct alkyla-
tion of commercially available pyrazole boronic ester (18) in
the presence of cesium carbonate. Suzuki coupling reaction of
the modified pyrazole boronic esters 19 with 3-bromobenze-
nesulfonamides 21 followed by hydrolysis furnished the
desired N-sulfonylanthranilic acids.
Efforts to understand the structure-activity relationships
of the lead molecule started with the carboxylic acid moiety.
As shown in Table 1, removal (6) or replacement of
carboxylic acid withfunctional groups such asester and amide
(7 and 8) resulted in lower potency compounds, indicating
that the carboxylic acid is essential for the activity. Moreover,
compounds having the carboxylic acid functional group in
meta- and para-position (9 and 10) were inactive against
RdRp, demonstrating the importance of the position of the
carboxylic acid (Table 1).
Therefore, the structural feature that gives rise to potency is
likely to be a hydrophobic terminal group linked to the
headgroup by a defined chain length. This could be demon-
strated by the potency increase of the methyl-2-naphthalene
replacement (27) that extends the hydrophobic substituent.
A further (albeit small) improvement in potency was made
through a combination of structural fragments showing the
best activity. Compound 28 with fluoro substitution on the
phenyl ring of carboxylic acid and methyl-2-naphthalene on
pyrazole displayed an IC50 of 0.26 μM against the dengue
polymerase (Table 2).
Inhibitors 2, 15, and 28 were assessed for selectivity
against several other polymerases, including human DNA
polymerase R and β and hepatitis C RdRp. The inhibitors
did not inhibit these polymerases, exhibiting IC50 of >20 μM
The effect of substituents on the phenyl ring adjacent to
the carboxylic acid was explored subsequently (Table 1).
A chlorine atom was well tolerated in the 5-position (12),

7936 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 24
Yin et al.
Table 3. Selectivity Panel
IC50 (μM)
dengue
RdRp
HCV
RdRp
human
DNA Pol R
human
DNA Pol β
compd
2
0.7
>20
>20
>20
>20
>20
>20
>20
>20
>20
15
28
0.37
0.26
Figure 3. Ligand interaction map for inhibitor 29 in the binding
pocket. The red dotted lines show the hydrogen bonding interac-
tions between 29 and T413, R737. The backbone of the specific
residues are shown as red lines.
Figure 2. Structure of inhibitor 29.
(Table 3). These results demonstrate that these novel inhibi-
tors specifically suppress DENV RdRp.
N-Sulfonylanthranilic acid derivatives seem to inhibit the
dengue RdRp during elongation; they can bind to the RdRp
alone or to the RNA-RdRp complex, suggesting that com-
pound binding does not interfere with template binding and
vice versa (unpublished results). In an effort to map the site of
association between the compound and the polymerase, a low
micromolar inhibitor (29) (Figure 2) (RdRp IC50 = 1.5 μM)
tethered with the photoalkylating group12 benzophenone was
synthesized. Under ultraviolet light irradiation, the com-
pound reacted with the protein and irreversibly inhibited the
RdRp activity in our assay. On the basis of MALDI MS, a
peptide spanning from amino acid position 313 to 326 of
DENV-2 NS5 polymerase (QTGSASSMVNGVVR) was
covalently linked to 29 ([M þ H]þ m/z 2048.8), resulting in a
mass increment of 656.17 (data not shown). Since unmodified
N-terminal b5 and C-terminal y6 ions were observed in the
corresponding MS-MS spectrum, the modification is located
within the sequence SSM of the above amino acid sequence.
On the basis of the known reactivity of the n,π* triplet state
of benzophenone13 and the presence of methionine in the
identified peptide sequence (e.g., m/z 1376.7, -CH2CH2Sþ),
we assumed that amino acid M - 320 is most likely the cross-
linking site.
Figure 4. Proposed allosteric binding site of N-sulfonylanthranilic
acid derivative 29 in the dengue polymerase domain.
protein structure was subjected to a molecular dynamics study
in which only the loops of the protein and the ligand were
allowed tomove. Asthe photo-cross-linking experimentswere
done at 277 K, a NVT ensemble was used with an initial
heating phase of 10 ps. The AMBER9917 force field in MOE
was used for the dynamics simulation with a time step of 1 fs
with sampling at every 0.5 ps over a 1 ns equilibration period.
A further simulation phase of 0.5 ns was calculated using the
same time parameters as before. The distance from the
carbonyl O atom in the benzophenone moiety to the S in
M320 in the protein was measured for each of the saved
structures. It was found that this distance varied between
A computational study was performed to rationalize the
photoaffinity labeling result and to identify the possible
compound binding site. In the X-ray crystallographic struc-
ture of the DENV RdRp (PDB code 2J7U),7 residues
311-317, 408-415, and 454-466 were not observed. As a
consequence, the three-dimensional atomic coordinates for
these amino acid sequences were obtained by constructing a
homology structure for the DENV RdRp using MOE.14 The
structure 2J7U was used as the primary template, and
the Cartesian coordinates for the backbone atoms of the
missing amino acids were obtained from the crystal structure
of the West Nile virus RdRp (PDB code 2HCN).15 The rmsd
for the backboneatomsbetweenthe X-ray structure 2J7Uand
˚
4.0 and 5.0 A for 82% of the complex structures during the
dynamics, and for 18% of the time of observation, this
˚
distance fluctuated between 3.0 and 4.0 A. The average
distance between the O and S atoms during the dynamics
˚
was 4.21 A (Figure 3). The side chain of R737 formed a strong
hydrogen bonding donor interaction with the carbonyl bond
˚
of the carboxamide linker moiety of 29 at a distance of 2.03 A.
˚
Another such interaction, at a distance of 2.4 A, was observed
between the carboxylate/acid group and the hydroxyl group
in the side chain of T413. The hydrogen bond between
carobxylate/acid and the hydroxyl group of T413 is likely
strengthened by electron withdrawing effect of the additional
group at the 5-position of phenyl ring, which is consistent with
initial structure-activity relationship study. These two inter-
actions were maintained during the course of the dynamics
˚
the homology structure was 1.03 A. Inhibitor 29 (with the
attached benzophenone moiety) was docked into the poly-
merase binding pocket of the homology model using
GOLD.16 The docked pose that best explained the overall
SAR presented above was chosen as a reasonable initial
orientation. This modeled complex of the docked ligand and

Letter
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 24 7937
simulations, indicating an importance in the stabilization of
the compound in the allosteric binding pocket. On the basis of
these results, we hypothesize that the reported compounds
bind to the dengue RdRp at an allosteric pocket that is located
between the finger and the thumb regions as shown in Figure 4.
We have identified N-sulfonylanthranilic acid derivatives
that can selectively inhibit dengue viral polymerase. Our
results suggest that these compounds may bind to a novel
allosteric site located between the finger and the thumb
regions in dengue RdRp.
using Kunjin virus replicon-expressing BHK Cells. J. Virol. 1998,
72 (9), 7270–7279.
(7) Yap, T. L.; Xu, T.; Chen, Y. L.; Malet, H.; Egloff, M. P.; Canard,
B.; Vasudevan, S. G.; Lescar, J. Crystal structure of the dengue
virus RNA-dependent RNA polymerase catalytic domain at 1.85-
angstrom resolution. J. Virol. 2007, 81 (9), 4753–4765.
(8) (a) De Clercq, E. Recent highlights in the development of
new antiviral drugs. Curr. Opin. Microbiol. 2005, 8, 552–560.
(b) De Francesco, R.; Carfi, A. Advances in the development of new
therapeutic agents targeting the NS3-4A serine protease or the NS5B
RNA-dependent RNA polymerase of the hepatitis C virus. Adv. Drug
Delivery Rev. 2007, 59, 1242–1262.
(9) Rawlinson, S. M.; Pryor, M. J.; Wright, P. J.; Jans, D. A. Dengue
virus RNA polymerase NS5: a potential therapeutic target? Curr.
Drug Targets 2006, 7, 1623–1638.
(10) The reaction was performed in 96-well, half area plates. Com-
pounds were incubated at various concentrations with 50 nM
purified recombinant dengue 2 NS5 polymerase and annealed
poly(C)/biotin oligo G as the template, in 50 mM Tris-HCl, pH
7.0, 10 mM KCl, 2 mM MgCl2, 2 mM MnCl2, 1 mM β-mercap-
toethanol, and 0.05% CHAPS buffer for 1 h at room temperature.
The reaction was initiated by addition of 0.4 μM GTP and 0.5 μCi
[3H]GTP in a total reaction volume of 50 μL for 1 h at room
temperature. The reaction was stopped by adding equal volumes of
25 mM Tris-HCl, pH 7.5, 75 mM NaCl, 20 mM EDTA, and 0.15
mg of streptavidine coated scintillation beads per well. After 1 h,
the reaction was quantified using a scintillation counter. Each
concentration was tested in duplicate, and the data were reported
as an average of at least two experiments.
While the reported compounds are not potent enough
for development, they have revealed an allosteric binding site
that may be explored for future dengue NS5 polymerase
inhibitors.
Acknowledgment. We thank Peiting Zeng for analytical
support.
Supporting Information Available: Structures, experimental
methods, and characterization data for compounds 2-29,
UV cross-linking experiment, computational experimental
methods, and 3D structure (PDB data file). This material is
available free of charge via the Internet at http://pubs.acs.org.
(11) Miyaura, N; Suzuki, A. Palladium-catalyzed cross-coupling reac-
tions of organoboron compounds. Chem. Rev. 1995, 95, 2457–
2483.
(12) Al Mawsawi, L. Q.; Fikkert, V.; Dayam, R.; Witvrouw, M.; Burke,
T. R., Jr.; Borchers, C. H.; Neamati, N. Discovery of a small-
molecule HIV-1 integrase inhibitor-binding site. Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 10080–10085.
(13) Prestwich, G. D.; Dorman, G.; Elliott, J. T.; Marecak, D. M.;
Chaudhary, A. Benzophenone photoprobes for phosphoinositides,
peptides and drugs. Photochem. Photobiol. 1997, 65, 222–234.
(14) Molecular Operating Environment, version 2007.0902; Chemical
Computing Group, Inc.: Montreal, Quebec, Canada, 2007.
(15) Malet, H.; Egloff, M.-P.; Selisko, B.; Butcher, R. E.; Wright, P. J.;
Roberts, M.; Gruez, A.; Sulzenbacher, G.; Vonrhein, C.; Bricogne,
G.; Mackenzie|, J. M.; Khromykh, A. A.; Andrew, D.; Davidson,
A. D.; Canard, B. Crystal structure of the RNA polymerase
domain of the West Nile virus non-structural protein 5. J. Biol.
Chem. 2007, 282 (14), 10678–10689.
References
(1) (a) Mackenzie, J. S.; Gubler, D. J.; Petersen, L. R. Emerging
flaviviruses: the spread and resurgence of Japanese encephalitis,
West Nile and dengue viruses. Nat. Med. 2004, 10 (12, Suppl.), S98–
S109. (b) Gubler, D. J. Epidemic dengue/dengue hemorrhagic fever as a
public health, social and economic problem in the 21st century. Trends
Microbiol. 2002, 10 (2), 100–103. (c) Jacobs, M. G.; Young, P. R.
Dengue: a continuing challenge for molecular biology. Curr. Opin.
Infect. Dis. 1998, 11, 319. (d) Monath, T. P. Dengue: the risk to
developed and developing countries. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 2395.
(2) (a) Guzman, M. G.; Kouri, G. Dengue: an update. Lancet Infect.
Dis. 2002, 2 (1), 33–42. (b) Gubler, D. J. The global emergence/
resurgence of arboviral diseases as public health problems. Arch. Med.
Res. 2002, 33 (4), 330–342.
(3) Clark, D. V.; Mammen, M. P., Jr.; Nisalak, A.; Puthimethee, V.;
Endy, T. P. Economic impact of dengue fever/dengue hemorrhagic
fever in Thailand at the family and population levels. Am. J. Trop.
Med. Hyg. 2005, 72 (6), 786–791.
(16) Jones, G.; Willett, P.; Glen, R. C. Molecular recognition of
receptor sites using a genetic algorithm with a description of
desolvation. J. Mol. Biol. 1995, 245, 43–53.
(4) (a) Gubler, D. J. Dengue and dengue hemorrhagic fever. Clin.
Microbiol. Rev. 1998, 11, 480–496. (b) Halstead, S. B. Dengue. Curr.
Opin. Infect. Dis. 2002, 15 (5), 471–476.
(17) Cieplak, P.; Caldwell, J.; Kollman, P. Molecular mechanical
models for organic and biological systems going beyond the atom
centered two body additive approximation: aqueous solution free
energies of methanol and N-methyl acetamide, nucleic acid base,
and amide hydrogen bonding and chloroform/water partition
coefficients of the nucleic acid bases. J. Comput. Chem. 2001, 22,
1048–1057.
(5) Blok, J. Genetic relationships of the dengue virus serotypes. J. Gen.
Virol. 1985, 66, 1323–1325.
(6) Khromykh, A. A.; Kenney, M. T.; Westaway, E. G. Trans-
complementation of Flavivirus RNA Polymerase Gene NS5 by