Discovery of Potent Non-Nucleoside Inhibitors of Dengue Viral RNA-Dependent RNA Polymerase from a Fragment Hit Using Structure-Based Drug Design

Article abstract of DOI:10.1021/acs.jmedchem.6b00143

The discovery and optimization of non-nucleoside dengue viral RNA-dependent-RNA polymerase (RdRp) inhibitors are described. An X-ray-based fragment screen of Novartis fragment collection resulted in the identification of a biphenyl acetic acid fragment 3, which bound in the palm subdomain of RdRp. Subsequent optimization of the fragment hit 3, relying on structure-based design, resulted in a >1000-fold improvement in potency in vitro and acquired antidengue activity against all four serotypes with low micromolar EC₅₀ in cell-based assays. The lead candidate 27 interacts with a novel binding pocket in the palm subdomain of the RdRp and exerts a promising activity against all clinically relevant dengue serotypes.

Full text of DOI:10.1021/acs.jmedchem.6b00143

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Discovery of Potent Non-Nucleoside Inhibitors of
Dengue Viral RNA-Dependent RNA Polymerase From
a Fragment Hit Using Structure-Based Drug Design
Fumiaki Yokokawa, Shahul Nilar, Christian G. Noble, Siew Pheng Lim, Ranga Rao, Stefani Tania, Gang
Wang, Gladys Lee, Juerg Hunziker, Ratna Karuna, Ujjini H. Manjunatha, Pei-Yong Shi, and Paul W. Smith
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00143 • Publication Date (Web): 17 Mar 2016
Downloaded from http://pubs.acs.org on March 20, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Potent NonꢀNucleoside Inhibitors of
Dengue Viral RNAꢀDependent RNA Polymerase
From a Fragment Hit Using StructureꢀBased Drug
Design
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
,
†
†
†
†
†
Fumiaki Yokokawa,* Shahul Nilar, Christian G. Noble, Siew Pheng Lim, Ranga Rao,
†
†
†
†
†
Stefani Tania, Gang Wang, Gladys Lee, Jürg Hunziker, Ratna Karuna, Ujjini Manjunatha,
†
†,‡
†
Pei-Yong Shi, and Paul W. Smith
†
Novartis Institute for Tropical Diseases, 10 Biopolis Road, No. 05ꢀ01, Chromos 138670,
Singapore
‡
Department of Biochemistry & Molecular Biology, Department of Phamarcology &
Toxicology, Sealy Center for Structural Biology & Molecular Biophysics, University of Texas
Medical Branch, Galveston, TX 77555, USA
KEYWORDS: Dengue, RNAꢀdependent RNA polymerase, Fragmentꢀbased drug discovery,
Nonꢀnucleoside inhibitor
ABSTRACT: The discovery and optimization of nonꢀnucleoside dengue viral RNAꢀdependentꢀ
RNA polymerase (RdRp) inhibitors are described. An Xꢀrayꢀbased fragment screen of Novartis’
fragment collection resulted in the identification of a biphenyl acetic acid fragment 3, which
bound in the palm subdomain of RdRp. Subsequent optimization of the fragment hit 3, relying
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
on structureꢀbased design, resulted in a >1000ꢀfold improvement in potency in vitro and acquired
antiꢀdengue activity against all four serotypes with low micromolar EC in cellꢀbased assays.
50
The lead candidate 27 interacts with a novel binding pocket in the palm subdomain of the RdRp
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and exerts a promising activity against all clinicallyꢀrelevant dengue serotypes.
INTRODUCTION
Dengue is a mosquitoꢀborne viral disease causing fluꢀlike symptoms, occasionally developing
into the potentially lifeꢀthreatening complications, dengue hemorrhagic fever (DHF) or dengue
shock syndrome (DSS). It is now widespread in over 100 countries, threatening 40% of the
1
world’s population. There are estimated to be approximately 390 million dengue infections
annually, 96 million of which exhibit disease symptoms, including 500,000 cases of severe
2
dengue and 22,000 deaths. Sanofi’s dengue vaccine recently gained regulatory approval in
Mexico, Brazil, and the Philippines, but has its limitations since it requires a threeꢀdose regimen
and is limited to individuals aged 9 to 45 years old. An antiviral therapy is not yet available for
treatment of dengue fever. Therefore, there is an urgent need to develop safe and effective drugs
3
for the treatment of dengue infection.
Dengue viruses (DENV) belong to the genus Flavivirus of the family Flaviviridae, and are
divided into four distinct but closely related serotypes (DENVꢀ1, DENVꢀ2, DENVꢀ3, and
DENVꢀ4). Infection with one serotype confers lifeꢀlong immunity, however a secondary
infection by a different serotype can increase the risk of developing severe dengue, because
4
crossꢀimmunity to the other serotypes is only partial and temporary. Therefore an ideal antiꢀ
dengue drug should exert panꢀserotype activity.
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The dengue viral genome is a singleꢀstranded RNA of about 11,000 nucleotides, and encodes
three structural proteins and seven nonstructural proteins. The nonstructural protein 5 (NS5) acts
as a methyltransferase and an RNAꢀdependent RNA polymerase (RdRp), both of which are
essential for viral replication. DENV NS5 RdRp performs both minusꢀstrand and plusꢀstrand
RNA synthesis during replication. Since it has no mammalian counterpart and its protein
sequence is conserved across all four serotypes with more than 65% homology, it offers an
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
attractive opportunity for discovery of new antiviral agents. The 3D crystal structure of DENV
NS5 RdRp adopts a classical polymerase shape, resembling a rightꢀhand that consists of fingers,
palm, and thumb subdomains (Figure 1). The catalytic active site of the DENV RdRp is defined
6
63
664
by a conserved GDD motif comprising two aspartic acid residues (Asp , Asp ), which are
located in the palm subdomain. These Asp residues are involved in the coordination of two metal
ions that are essential to the catalytic mechanism for incoming nucleotide incorporation. The
priming loop, which regulates RNAꢀtemplate binding and polymerization, points from the thumb
6
ꢀ8
subdomain towards the active site.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Crystal structure of 3 (magenta) bound to the RdRp domain of DENVꢀ3 protein NS5
shown with the protein as a cartoon (PDB ID 5F3T). The palm subdomain (cyan) contains the
6
63
664
activeꢀsite residues (Asp and Asp from the GDD motif, represented as orange sticks), and
the fingers and thumb subdomains are colored purple and green, respectively. The priming loop
is colored yellow.
In recent years, there has been a growing interest in DENV NS5, with several groups reporting
9
ꢀ14
RdRp inhibitors.
Our own research efforts on the discovery of DENV NS5 RdRp inhibitors
have led to the identification of nucleos(t)ide inhibitors, which act as chain terminators during
15,16
RNA synthesis by their 5’ꢀtriphosphate metabolites.
As a complementary activity, highꢀ
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
throughput screening (HTS) of the Novartis corporate compound archive was conducted using
1
7
18
either a radioactive scintillation proximity or fluorescentꢀcoupled assays measuring the
elongation activity of the enzyme. Previously we reported that our HTS campaigns led to the
identification of an Nꢀsulfonyl anthranilic acid hit 1 with an IC50 of 7.2 ꢀM. Subsequent
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
structureꢀactivity relationship (SAR) exploration improved the potency to an IC of 0.26 ꢀM.
5
0
However this medicinal chemistry optimization ended up with a highly lipophilic molecule 2
(clogP 6.6), which had potentially unfavorable drugꢀlike properties as a lead candidate. In
1
7, 19
addition, compound 2 did not exhibit any antiviral activity in cell culture (Figure 2).
Figure 2. Nꢀsulfonyl anthranilic derivatives as nonꢀnucleoside inhibitors of the dengue RdRp
from an HTS hit.
As reported in the previous publications, several hits from the HTS campaigns were rigorously
followed up, however we failed to identify specific DENV NS5 RdRp inhibitors with good
physicochemical properties, because of false positives due to impurities or binding to the large
2
0
RNA template tunnel space. The fact that multiple HTS campaigns to identify RdRp enzyme
inhibitors did not provide viable hits, led us to consider fragment screening to identify novel and
tractable hits that could be optimized to potent leads with good drugꢀlike properties. As a part of
fragment screening campaigns, we undertook an Xꢀray crystallographic screen, which provides
unambiguous proof of binding to the target site and reveals the binding mode of the hits, thus
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
providing clear direction for their optimization. Herein, we describe our approach to fragmentꢀ
based screening followed by structureꢀbased drug design efforts to identify potent nonꢀ
nucleoside inhibitors of the dengue NS5 polymerase.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
A fragment screen of the Novartis fragment collection using Xꢀray crystallography identified a
single hit; a biphenyl acetic acid fragment, 3 (IC 734 ꢀM, SPRꢀKd 613 ꢀM, LE 0.24), which
5
0
2
1
bound in the palm subdomain (Figure 1, 3). A related analog from the Novartis archive, 4 (IC
5
0
769 ꢀM, SPRꢀKd >200 ꢀM, LE 0.26) was also found to bind in the same pocket using Xꢀray
crystallography but it bound with the opposite orientation of the carboxylic acid moiety,
suggesting that the binding was mainly from the biphenyl moiety. Each carboxylic acid formed
an Hꢀbond interaction with the priming loop (Figure 3).
In the course of our research for panꢀserotype DENV RdRp inhibitors, we primarily used DENV
serotype 4. DENVꢀ4 was used for the biochemical enzymeꢀinhibition assay and for biophysical
binding assays using isothermal titration calorimetry (ITC) or surfaceꢀplasmon resonance (SPR)
for investigating SAR. DENVꢀ3 was used for Xꢀray crystallography since the methodology is
2
0
well established and as yet there is no DENVꢀ4 RdRp crystal structure. In addition, selected
compounds were evaluated by RdRp biochemical and cellꢀbased assays in all four DENV
serotypes to confirm panꢀserotype activity.
For biophysical binding analyses we used a combination of ITC and SPR. Generally the data
between the two assays correlated, although SPR consistently showed a higher affinity, probably
because the SPR assay was performed at 4 °C. Once the SPR assay was developed it was used as
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the biophysical assay of choice to develop SAR because of its higher throughput and because of
the limitations of using ITC for highꢀaffinity binding.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Overlay of coꢀcrystal structures of 3 (magenta) and 4 (cyan) bound to the RdRp
domain of DENVꢀ3 NS5. The compounds are shown as sticks and the protein as lines. The
residues in the palm and thumb subdomains are green and in the priming loop yellow. Hꢀbond
interaction of carboxylic acid groups with the priming loop is shown by dotted lines.
To initiate elaboration of the fragment hits, the biphenyl acetic acid moiety was merged to
provide the bisꢀacid 5 (IC 177 ꢀM, ITCꢀKd 154 ꢀM). The initial optimization efforts were
50
focused on filling the narrow and deep cavity from the distal aromatic ring of the bisꢀacid 5. The
SAR of the distal benzene ring of 5 was explored as shown in Table 1. Halogenated phenyl rings
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
ꢀ8 and the fiveꢀmembered furan ring 9 did not improve potency, however the thiophene ring 10
was found to provide a 10ꢀfold increase in potency (IC 15 ꢀM, ITCꢀKd 28 ꢀM). The coꢀcrystal
5
0
structure of 10 with the NS5 polymerase domain suggested that this higher affinity is potentially
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
due to the noncovalent interaction of the sulfur of the thiophene with the oxygen of the OH sideꢀ
7
96
22
chain of Ser (Figure 4A). In addition, the Xꢀray structure suggested that the 2ꢀpostion of the
thiophene ring provided the correct vector to grow the compound toward the narrow cavity
(Figure 4B).
Figure 4. (A) Overlay of coꢀcrystal structures of 5 (purple) and 10 (grey) bound to the RdRp
domain of DENVꢀ3 NS5. A water molecule is displayed as a red sphere, forming Hꢀbonds with
800
802
His and Gln . The compounds are shown as sticks and the amino acid residues as lines. The
palm and thumb subdomains are green and the priming loop is yellow. (B) Alternate view of 5
and 10 from the opposite site (180 degree rotation) compared to (A) shown in surface
representation. The green arrow highlights the vector off the 2ꢀposition of the thiophene to grow
toward the inner cavity.
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Various substituents such as Cl, Me, and Br moieties were introduced in the 2ꢀposition of the
thiophene to fill the cavity, however none of them improved potency (compounds 11ꢀ13). It was
noted that a water molecule was bound deeper inside the narrow cavity forming two Hꢀbond
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
802
800
interactions with Gln
and His . To interact with the water, a nitrile group was added,
however cyanoꢀthiophene 14 significantly decreased potency. As an alternative approach we
attempted to displace the water molecule. A propargyl alcohol moiety was introduced to the 2ꢀ
thiophene ring to provide 15 (IC 1.7 ꢀM, ITCꢀKd 1.4 ꢀM), which gave a 100ꢀfold increase in
5
0
potency and improved the ligand efficiency (LE) to 0.35 as compared to 5. The homopropargyl
alcohol analog 16 was found to be less potent than 15, suggesting that the propargyl alcohol had
the optimal length to displace the water molecule. Coꢀcrystallization of 15 with the NS5 RdRp
domain confirmed that the propargyl alcohol filled the narrow cavity and displaced the water
8
00
molecule to form the Hꢀbond interactions with the backbone of His and the side chain of
8
02
Gln as predicted (Figure 5). Amino acid residues of the binding pocket are strictly conserved
800
802
across all four serotypes except for residues Thr and Glu
in DENVꢀ2 (see supporting
information Figure S3). The propargyl alcohol moiety in 15 is likely to form similar Hꢀbond
interactions with residues 800 and 802 in DENV 1ꢀ4, generating panꢀserotype activity with IC50
values of 0.3 to 2.2 ꢀM (Table 2).
Table 1. Enzyme inhibition and LE data for bis-acid analogs 5 – 16.
c
Compound
R1
DENVꢀ4 NS5
DENVꢀ4 NS5
LE
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
b
IC (ꢀM)
ITCꢀKd (ꢀM)
5
0
5
6
7
8
9
Ph
177
154
0.26
0.25
0.26
0.28
0.27
0.36
0.32
0.28
0.31
0.25
2ꢀClꢀPh
192
141
62
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3ꢀClꢀPh
4ꢀClꢀPh
2ꢀfuryl
211
15
552
28
1
0
2ꢀthienyl
11
5ꢀClꢀ(2ꢀthienyl)
5ꢀMeꢀ(2ꢀthienyl)
5ꢀBrꢀ(2ꢀthienyl)
5ꢀCNꢀ(2ꢀthienyl)
26
33
1
1
1
2
3
4
96
39
46
199
134
1
1
5
6
1.7
29
1.4
34
0.35
0.27
a
b
Enzyme IC50 values were determined as described in the Experimental Section. Kd values were
c
determined as described in the Experimental Section. LE = (1.4*pIC )/heavy atoms.
5
0
Table 2. Enzyme inhibition of compound 15 against other dengue serotypes.
Dengue serotype
IC (ꢀM)
5
0
1
2
3
0.3
2.2
0.5
ACS Paragon Plus Environment

Page 11 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
1.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Overlay of coꢀcrystal structures of 10 (grey) and 15 (cyan) with the RdRp domain of
DENVꢀ3 NS5. Residues from the palm and thumb subdomains are shown as green lines and the
priming loop as yellow lines. As predicted, the hydroxyl group of the propargyl 15 is seen to
substitute the water molecule present in the complex with 10.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Further, we investigated the optimization of binding to the outside of the pocket. The propargyl
alcohol analog 15 showed poor membrane permeability in the Cacoꢀ2 cells (Papp = 0.51 cm/s x
ꢀ6
1
0 ), likely due to the existence of two negativelyꢀcharged carboxylic acids. To improve cellular
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
permeability, we revisited the monoꢀacid original hit 3 for the further optimization. Replacement
of the distal benzene ring with the thiopheneꢀpropargyl alcohol moiety afforded 17 (IC 7.5 ꢀM,
5
0
ITC Kd 14 ꢀM, SPR Kd 2.2 ꢀM). The monoꢀacid 17 showed 5ꢀfold increased Cacoꢀ2
ꢀ6
permeability (Papp = 2.73 cm/s x 10 ) as compared to the bisꢀacid 15. Removing the methylene
group of the phenyl acetic acid moiety of 17 provided 18, which resulted in 3.7 fold
improvement in potency (IC = 2.3 ꢀM, ITC Kd 3.8 ꢀM, SPR Kd 1.7 ꢀM). Introduction of the
50
methyl group on the 6ꢀposition of the methoxy benzene ring in 18 gave 19, which inhibited the
RdRp with an IC of 0.5 ꢀM, causing a further increase in LE to 0.42 (IC = 0.53 ꢀM, ITC Kd
5
0
50
1.6 ꢀM, SPR Kd 0.11 ꢀM).
2
3
Subsequently, we sought to identify carboxylic acid bioisosteres, which could have a similar
geometrical topology of Hꢀbond donor and acceptor with the CO H. Replacing the carboxylic
2
acid in 18 with a wellꢀknown isostere, the tetrazole, afforded 20, which retained inhibitory
activity against RdRp (IC = 2.4 ꢀM), however the Cacoꢀ2 permeability of 20 was not improved
50
ꢀ6
(
Papp = 1.39 cm/s x 10 ). The diazaoxazolone 21 was found to be more lipophilic and less
acidic than the corresponding carboxylic acid 19, however it showed 5ꢀfold less potency. The
nonꢀacidic imidazole 22 lost potency significantly. Conversion of the carboxylic acid to methyl
acylsulfonamide 23 offered comparable potency and binding affinity in the biochemical and the
biophysical analyses, suggesting that the acylsulfonamide functionality engages similar
interactions with the RdRp (IC = 0.34 ꢀM, ITC Kd 2.0 ꢀM, SPR Kd 0.12 ꢀM). The reverse
5
0
acylsulfonamide 24 was found to be less potent than 23. Unfortunately, despite potent inhibition
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
against the RdRp with submicromolar IC , both 19 and 23 showed no antiꢀdengue activity up to
5
0
50 ꢀM in A549 cells, which may reflect their inability to permeate cells efficiently and to reach
the intracellular enzyme target. This is supported by the measured acid dissociation constant
value (pKa 4.0ꢀ4.2) and distribution coefficient value (logD 0.8ꢀ1.2) at pH 7.4 for 19 and 23.
Therefore, the lack of cellular activity is likely to be the result of poor cellular permeability due
to their negativelyꢀcharged carboxylic acid/acylsulfonamide functionalities and their high
polarity. The Cacoꢀ2 data for 23 also indicated that it has low cellular permeability (Papp = 0.50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ6
cm/s x 10 ).
Table 3. Enzyme inhibition and LE data for mono-acid and acid bioisoster analogs 17 – 24.
d
D4 IC50
D4 ITC
D4 SPR
logD
Cacoꢀ2
c
cpd
R1
R2
pKa LE
a
b
b
ꢀ6
(ꢀM)
Kd (ꢀM) Kd (ꢀM) (pH 7.4)
(cm/s x10 )
0.51
1
1
1
5
7
8
1.7
7.5
1.4
0.35
0.34
0.39
0.42
CH
COOH
H
H
14
3.8
1.6
2.2
1.7
2.73
2
COOH
COOH
2.3
0.79
1.2
4.2
4.2
19
Me
0.53
0.11
2
0
H
2.4
0.36
1.39
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
1
2
Me
Me
3.2
44
1.9
3.6
5.5
6.6
0.32
0.26
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
3
4
Me
Me
0.34
2.5
2
0.12
0.8
4.0
4.8
0.36
0.31
0.5
0.78
a
b
Enzyme IC50 values were determined as described in the Experimental Section. Kd values were
c
d
determined as described in the Experimental Section. LE = (1.4*pIC )/heavy atoms. Apical to basal
5
0
permeability at pH 7.4. D4 = DENVꢀ4.
In order to gain cellular activity, our next objective was to modify the physicochemical
properties of the compounds to enhance their cellular permeability as shown in Table 4. The Xꢀ
ray coꢀcrystal structure of 23 with the DENVꢀ3 NS5 RdRp domain confirmed that it was binding
in the same pocket, as shown in Figure 6. The propargyl alcohol projected into the narrow cavity
8
00
802
and formed two Hꢀbond interactions with His and Glu . The acylsulfonamide formed three
7
94
729
795
Hꢀbond interactions with the side chains of Thr and Arg and the backbone of Trp . The
methyl group of the acylsulfonamide moiety did not make contact with the enzyme surface but
was exposed to the solvent space, suggesting that modification of the methyl group could serve
to influence the overall physicochemical properties of inhibitors without interfering with their
affinity to the enzyme.
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. (A) Xꢀray coꢀcrystal structure of 23 (purple) bound to the RdRp domain of DENVꢀ3
NS5. The compound is shown as sticks and the protein residues as lines. Residues in the palm
and thumb subdomains are green and in the priming loop are yellow. (B) Twoꢀdimensional
ligandꢀinteraction map generated using Molecular Operating Environment. Hydrophobic
residues are colored green, polar residues are colored light purple and charged residues have an
additional blue ring. The degree of solvent exposure is shown by the blue halos. Hꢀbond
interactions to the amino acid mainchain or sidechain are shown as dashed blue or green arrows
respectively, pointing towards the Hꢀbond acceptor. Waterꢀmediated contacts are shown as gold
dashed lines.
Changing the methyl to benzene sulfonamide 25 did not affect the potency against RdRp (IC50
0.17 ꢀM) and increased the lipophilicity (logD 1.7) as expected. It showed moderate cellular
activity (EC 18ꢀ41 ꢀM), whereas its acidity (pKa 3.8) was similar to those of the carboxylic
5
0
acid 19 (pKa 4.2) and the methyl sulfonamide 23 (pKa 4.0). The Cacoꢀ2 data suggested that the
ꢀ6
cellular permeability of 25 was still low (Papp = 0.5 cm/s x 10 ). In an attempt to improve the
cellular permeability by reducing ionization of the acylsulfonamide by modifying its pKa, the 6ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
methyl group of the methoxy benzene ring of 25 was replaced with the electronꢀdonating
ꢀ6
methoxy group to afford 26, which improved Cacoꢀ2 cell permeability (Papp = 10.8 cm/s x 10 )
with similar lipophilicity (logD 1.5), reflecting reduced acidity (pKa 4.8) of the acyl
sulfonamide. However, compound 26 showed only modest improvement in cellular potency
(EC50 7.3ꢀ37 ꢀM) despite its increased cellular permeability. On the other hand, 3ꢀ
methoxyphenyl sulfonamide 27 displayed the most potent antiꢀdengue activity in this series with
EC of low micromolar against all four serotypes (EC 1.8ꢀ2.3 ꢀM) without cytotoxicity up to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
50
5
0 ꢀM (CC >50 ꢀM in HepG2 cells), while it had similar physicochemical properties (log D
5
0
ꢀ6
1
.6, pKa 4.7) and lower Cacoꢀ2 permeability (Papp = 3.91 cm/s x 10 ) as compared to 26. The
exact cause of the significant improvement in the cellular potency of 27 is not known, but it may
be due to difference of accessibility of the compounds to the binding site in the presence of viral
and cellular proteins. Compound 27 was also confirmed to be active against all four serotypes in
the biochemical assays (Table 5). When the OMe at the 4ꢀposition of the benzene ring of 27 was
replaced with the electronꢀwithdrawing chlorine, the acidity of 28 was increased to pKa 3.8,
resulting in slightly reduced cellular activity (EC 5.5ꢀ13 ꢀM). We found that the 8ꢀquinolyl
5
0
sulfonamide 29 exhibited the highest biochemical potency and binding affinity with an IC of
5
0
0.023 ꢀM and an SPRꢀKd of 0.007 ꢀM, although it was 2ꢀ to 6ꢀfold less potent than 27 in the
cellꢀbased assays (EC 3.8ꢀ14 ꢀM). The 8ꢀquinolyl acylsulfonamide moiety of 29 may take a
5
0
different binding conformation to the RdRp in the biochemical and SPR assay systems, resulting
in the highest potency and binding in this series.
Table 4. Enzyme inhibition, SPR-Kd, cellular potency, logD, pKa, and Caco-2 data for acyl
sulfonamide analogs 23, 25 – 29.
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
c
D4
SPR
Kd
EC (ꢀM, A547 cells)
5
0
d
Cacoꢀ2
(cm/s
D4 IC50
logD
D4 (pH 7.4)
Cpd
R1
R2
R3
pKa
a
(
ꢀM)
D1
D2
D3
ꢀ6
x10 )
b
(
ꢀM)
0.12
ND
23
Me
Ph
Ph
Me
Me
OMe
OMe
0.34
0.17
0.25
0.17
0.14
>50 >50 >50 >50
0.8
1.7
1.5
1.6
0.9
4.0
3.8
4.8
4.7
3.8
0.5
0.5
2
2
2
2
5
6
7
8
31
15
34
37
2.3
13
18
7.3
1.8
5.5
41
14
OMe OMe
0.09
0.07
0.01
10.8
3.91
ND
3ꢀMeOPh OMe OMe
1.8
6.8
1.8
7.0
3ꢀMeOPh OMe
Cl
2
9
Me
OMe
0.023
0.007
6.3
14
3.8
10
1.5
4.4
2.57
a
b
Enzyme IC50 values were determined as described in the Experimental Section. Kd values were
c
determined by SPR as described in the Experimental Section. EC values were determined as described
5
0
d
in the Experimental Section. Apical to basal permeability at pH 7.4. D1 = DENVꢀ1, D2 = DENVꢀ2, D3 =
DENVꢀ3, D4 = DENVꢀ4.
Table 5. Enzyme inhibition of compound 27 against other dengue serotypes.
Dengue serotype
IC (ꢀM)
5
0
1
2
0.068
0.11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
4
0.048
0.17
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSION
In summary, we have described a fragmentꢀbased drug discovery approach to identify novel,
potent nonꢀnucleoside inhibitors of dengue viral RNA polymerase. An Xꢀrayꢀbased fragment
screen resulted in the identification of a biphenyl acetic acid hit 3 with an IC of ~700 ꢀM,
50
which is bound to a novel pocket in the palm subdomain of the RdRp. Subsequent growing and
optimization of the fragment hit 3 using crystallography and computerꢀaided structureꢀbased
design achieved >1,000ꢀfold improvement in the potency. Replacement of the carboxylic acid
moiety with an isosteric acylsulfonamide followed by optimization of physicochemical
properties led to the identification of 27, which displayed antiviral activity in the cellꢀbased
assays for all four dengue serotypes at low micromolar concentrations. Additional biological
characterization of 27 and 29 have confirmed their onꢀtarget cellular activities, which is reported
2
4
elsewhere. To our knowledge, this is the first class of compounds to show antiviral activity that
correlates with direct evidence for interacting with a specific site of the DENV NS5 RdRp. Thus,
these analogs are promising leads for further optimization and development.
CHEMISTRY
Bisꢀacid derivatives 5-16 described herein were prepared as shown in Scheme 1. Benzylic
bromination of 3,5ꢀdimethyl bromobenzene (30) followed by nucleophilic displacement of the
25
resulting benzylic bromide 31 with potassium cyanide gave the bisꢀcyanide 32, which was
converted to the corresponding carboxylic acid 33 and carboxylic acid ester 34. The aromatic
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26
ring was installed by the SuzukiꢀMiyaura crossꢀcoupling methodology between 33 and
aromatic boronic acids to afford compounds 5ꢀ8 and 10ꢀ11. Alternatively, the crossꢀcoupling
with the bisꢀcyanide 32 with aromatic boronic acids followed by hydrolysis of the cyanide gave
compounds 9 and 12. Compounds 13ꢀ15 were prepared by palladium catalyzed borylation of the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
7
aryl bromide 34 with bis(pinacolato)diboron, crossꢀcoupling with 2,5ꢀdibormothiophene, 2ꢀ
cyanoꢀ5ꢀbromothiophene and 3ꢀ(5ꢀbromothiophenꢀ2ꢀyl)propꢀ2ꢀynꢀ1ꢀol (37), followed by
2
8
alkaline hydrolysis of the ester 38. Sonogashira reaction of the bromothiophene 38 (R = Br)
1
with homopropargyl alcohol gave compound 39, which underwent saponification to yield
compound 16.
a
Scheme 1. Synthesis of bis-acid derivatives 5-16
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) NBS, AIBN, CH CN, reflux; (b) KCN, KI, 18ꢀcrownꢀ6, CH CN;
3
3
o
o
(
c) conc. HCl, 80 C; (d) H SO , MeOH, 50 C; (e) PdCl (dppf)
—
CH Cl , K PO , 1,4ꢀdioxane, 80
2
4
2
2
2
3
4
o
o
C; (f) PdCl (dppf), CsF, DMF, 80 C; (g) bis(pinacolato)diboron, PdCl (dppf)—CH Cl , KOAc,
2 2
2
2
o
o
DMSO, 90 C; (h) PdCl (dppf)—CH Cl , K PO , 1,4ꢀdioxane, 80 C; (i) aq. KOH or aq. LiOH,
2 2 3 4
2
o
THF, MeOH; (j) (Ph P) Pd, CuI, i-Pr NEt, toluene, 80 C.
3
4
2
Compound 17 was prepared from the commercially available acid 40 as shown in Scheme 2.
Protection of the carboxylic acid 40 with methyl ester, palladium catalyzed borylation followed
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
by crossꢀcoupling with 37 gave compound 43. Subsequent hydrolysis of the methyl ester 43
provided compound 17.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 2. Synthesis of compound 17
a
Reagents and conditions: (a) SOCl , MeOH; (b) bis(pinacolato)diboron, PdCl (dppf)—CH Cl ,
2 2
2
2
o
o
KOAc, 1,4ꢀdioxane, 100 C; (c) PdCl (dppf)
—
CH Cl , K PO , 1,4ꢀdioxane, 80 C; (d) aq. LiOH,
2
2
2
3
4
THF.
The other monoꢀacid analogs 18ꢀ19 containing the thiophene propargyl alcohol moiety were
prepared as described in Scheme 3. SuzukiꢀMiyaura crossꢀcoupling of the aryl halide 44 with 2ꢀ
thiophene boronic acid afforded 45. Bromination of the thiophene 45 using Nꢀbromosuccinimide
(
NBS) provided the 2ꢀbromoꢀthiophene 46, which was reacted with propargyl alcohol under the
Sonogashira coupling condition to provide 50 (R = H). Alternatively, the palladium catalyzed
2
borylation of the aryl halide 48 followed by crossꢀcoupling with 37 yielded 50 (R = Me). The
2
resulting esters 50 were hydrolyzed to give compounds 18ꢀ19.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 3. Synthesis of compounds 18-19
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) 2ꢀthiophene boronic acid, Pd(dba) , Ph P, K CO , 1,4ꢀdioxane; (b)
2
3
2
3
o
NBS, DMF; (c) propargyl alcohol, PdCl (dppf)
—
CH Cl , CuI, Et N, DMF, 90 C; (d) Br , Fe,
2
2
2
3
2
CHCl ; (e) SOCl , MeOH; (f) bis(pinacolato)diboron, PdCl (dppf)
—
CH Cl , KOAc, 1,4ꢀdioxane,
3
2
2
2
2
o
o
9
0 C; (g) 37, PdCl (dppf)
—
CH Cl , K CO , 1,4ꢀdioxane, H O, 80 C; (h) aq. LiOH, THF,
2
2
2
2
3
2
MeOH.
Synthesis of the tetrazole analog 20 was started by hydrolysis of the ester 46 as shown in Scheme
4. Amide coupling of the acid 51 with ammonium chloride afforded the primary carboxamide 52,
which underwent dehydration to the nitrile 53. The nitrile 53 was converted into the tetrazole 54
by treatment with sodium azide. Subsequent Sonogashira coupling of 54 with propargyl alcohol
gave compound 20.
a
Scheme 4. Synthesis of compounds 20
ACS Paragon Plus Environment

Page 23 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) aq. LiOH, THF, MeOH; (b) NH Cl, HATU, i-Pr NEt, DMF; (c)
4
2
o
(
CF CO) O, Et N, CH Cl ; (d) NaN , amberlystꢀ15, DMSO, 90 C; (e) propargyl alcohol,
3 2 3 2 2 3
o
PdCl (dppf)—CH Cl , CuI, Et N, DMF, 80 C.
2 2 3
2
The diazaoxazolone 21 was synthesized from the amidoxime 58, which was prepared by addition
of hydroxylamine to the corresponding cyano derivative 57. Cyclization of the amidoxime 58
with carbonyldiimidazole (CDI) to the diazaoxazolone 59 followed by Sonogashira coupling
with propargyl alcohol yielded compound 21. The imidazole ring 61 was prepared by cyclization
of the aldehyde 60 with glyoxal in the presence of aqueous ammonia. Subsequent introduction of
the propargyl alcohol provided compound 22 (Scheme 5).
a
Scheme 5. Synthesis of compounds 21-22
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
o
Reagents and conditions: (a) (Ph P) Pd, K PO , aq. 1,4ꢀdioxane, 80 C; (b) NBS, DMF; (c)
3
4
3
4
.
o
o
NH OH HCl, NaHCO , MeOH, 90 C; (d) CDI, DBU, 110 C; (e) propargyl alcohol, (Ph P) Pd,
2
3
3
4
o
CuI, i-Pr NEt, THF, 80 C; (f) DIBALꢀH, CH Cl ; (g) glyoxal, liq. NH , MeOH; (h) propargyl
2
2
2
3
o
alcohol, PdCl (dppf)—CH Cl , CuI, Et N, 1,4ꢀdioxane, 80 C.
2 2 3
2
The reverse acylsulfonamide derivative 24 was prepared as described in Scheme 6. Reaction of
2ꢀmethoxyꢀ4ꢀmethyl bromobenzene (62) with chlorosulfonic acid gave the aryl sulfonyl chloride
6
3, which was treated with ammonia in methanol followed by acetyl chloride to afford the
reverse acylsulfonamide 65. Crossꢀcoupling with 2ꢀthiophene boronic acid, bromination of the
thiophene 66, followed by introduction of propargyl alcohol yielded compound 24.
a
Scheme 6. Synthesis of the reverse acylsulfonamide derivative 24
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) ClSO H, CH Cl ; (b) NH , MeOH; (c) acetyl chloride, DMAP,
3
2
2
3
Et N, CH Cl ; (d) thiopheneꢀ2ꢀboronic acid, (Ph P) Pd, Na CO , aq. 1,2ꢀdimethoxyethane, 80
3
2
2
3
4
2
3
o
o
C; (e) NBS, DMF; (f) propargyl alcohol, PdCl (dppf)—CH Cl , CuI, Et N, DMF, 80 C.
2 2 3
2
Acylsulfonamide derivatives 23, 25ꢀ29 were prepared by coupling of the benzoic acid
intermediates 68 with various sulfonamides using EDCI·HCl (1ꢀethylꢀ3ꢀ(3ꢀ
dimethylaminopropyl)carbodiimide hydrochloride) as a coupling reagent in the presence of
HOBt (1ꢀhydroxybenzotriazol) or DMAP (4ꢀdimethylaminopyridine) as shown in Scheme 7. The
resulting bromo thiophenes 69 were treated with the Sonogashira reaction to provide compounds
2
3, 25ꢀ29.
a
Scheme 7. Synthesis of acylsulfonamide derivatives 23, 25-29
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Reagents and conditions: (a) EDCI·HCl, HOBt or DMAP, DMF or CH Cl ; (b) propargyl
2
2
o
alcohol, PdCl (dppf)
—
CH Cl , CuI, Et N or Cs CO , 1,4ꢀdioxane, 80 C.
2
2
2
3
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EXPERIMENTAL SECTION
All materials and reagents used were of the best commercially available grade and used without
1
further purification. H NMR spectra were determined on a Bruker Ultrashield 400 MHz
spectrometer or a Bruker AVANCE III HD 400 MHz or a Varian Mercury 300 Plus 300 MHz
NMR. Compound purity was determined by the following methods. Method 1: Waters Acquity
UPLC equipped with Acquity UPLC BEH Shield RP18 column, 1.7ꢁm, 2.1 x 50 mm using a
gradient of 95:5 H O (0.1% formic acid)/CH CN over at 2.5 min, at 2.60 min then ramp to 0:100
2
3
H O (0.1% formic acid)/CH CN, hold until 4.0 min, return to 95:5 H O (0.1% formic
2
3
2
acid)/CH CN at 4.20 min until end of run with a 0.7 mL/min flow rate. Method 2: Waters
3
Acquity UPLC equipped with Acquity UPLC HSS T3 column, 1.8 ꢁm, 2.1 × 50 mm using a
gradient of 95:5 H O (0.1% formic acid)/CH CN to 2:98 H O (0.1% formic acid)/CH CN for 2
2
3
2
3
min run time with a 1.0 mL/min flow rate. Method 3: Agilent 1100series IonꢀTrap Mass detector
equipped with a Xbridge C18 column, 3.5 ꢁm, 4.6 x 75 mm using a gradient of 95:5 5mM
ammonium carbonate/CH CN to 20:80 5mM ammonium carbonate/CH CN in 4 min and hold up
3
3
to 7 min with 20:80 5mM ammonium carbonate/CH CN with a 1.0 mL/min flow rate. Method 4:
3
Waters Quattro Micro UPLCꢀLCMS equipped with a Acquity BEH C18 column, 1.7 ꢁm, 2.1 ×
1
00 mm using a gradient of 100:0 H O (0.025% trifluoro acetic acid)/CH CN (0.025% trifluoro
2
3
acetic acid) to 20:80 H O (0.025% trifluoro acetic acid)/CH CN (0.025% trifluoro acetic acid)
2
3
for 6 min run time with a 0.4 mL/min flow rate. The purity of all compounds screened in the
biological assays was examined by HPLC analysis and was found to be >95 % at 254 nm.
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
HPLC purifications were performed using Waters Prep HPLC System with wavelength detection
of 254 and 214 nm and flow rate of 20 mL/min using the following column and mobile phase
conditions: A) Atlantis Cꢀ18 5 ꢀm 19 x 150 mm column; mobile phase system of 0.1% formic
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
acid/acetonitrile, B) XBridge Cꢀ18 5 ꢀm 19 x 100 mm column; mobile phase system of 0.1%
formic acid/acetonitrile, C) XBridge Cꢀ18 5 ꢀm 19 x 150 mm column; mobile phase system of
10 mM ammonium bicarbonate/acetonitrile, D) Atlantis T3 5 ꢀm 30 x 250 mm column; mobile
phase system of 10 mM ammonium acetate/acetonitrile, E) Sunfire Cꢀ18 5 ꢀm 30 x 250 mm
column; mobile phase system of 0.1% formic acid/acetonitrile, F) XBridge Cꢀ18 5 ꢀm 19 x 150
mm column; mobile phase system of 0.1% formic acid/acetonitrile. HRMS ESIꢀMS data were
recorded using a Thermo Scientific LTQ Orbitrap XL mass spectrometer.
2,2'-([1,1'-Biphenyl]-3,5-diyl)diacetic acid (5).
A mixture of 1ꢀbromoꢀ3,5ꢀdimethylbenzene (30) (1 g, 5.4 mmol), NBS (2.02 g, 11.35 mmol) and
a catalytic amount (2ꢀ5 %) of AIBN in CH CN was refluxed under argon for 3 h. The solvent
3
was evaporated under vacuum and 25 mL of CCl was added. The mixture was heated in order to
4
dissolve the crude product and, after cooling, the insoluble succinimide was eliminated by
filtration and washed with more CCl . Evaporation of the solvent in the filtrates gave a crude
4
product which was purified by silica gel column chromatography to yield 1ꢀbromoꢀ3,5ꢀ
bis(bromomethyl)benzene (31) (890 mg, y. 48%) as a white solid. A further purification could
1
be achieved by crystallization from ethanol. mp: 97ꢀ98.5 ºC. H NMR (CDCl , 400 MHz): δ 4.41
3
(s, 4H), 7.34 (brs, 1H), 7.47 (d, J = 1.5 Hz, 2H).
To a solution of 1ꢀbromoꢀ3,5ꢀbis(bromomethyl)benzene (31) (1.35 g, 3.94 mmol) in CH CN (20
3
mL) were added KCN (0.56 g, 8.66 mmol), 18ꢀcrownꢀ6 (0.21 g, 0.79 mmol), KI (0.065 g, 0.39
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mmol), and water (1.6 mL) at room temperature. After being stirred for 24 h, the mixture was
diluted with EtOAc, washed with water and brine, dried over Na SO , filtered, and concentrated
2
4
in vacuo. The residue was purified by silica gel column chromatography to give 2,2'ꢀ(5ꢀbromoꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
,3ꢀphenylene)diacetonitrile (32) (650 mg, y. 70%). H NMR (CDCl , 400 MHz): δ 3.74 (s, 4H),
3
ꢀ
7.23 (s, 1H), 7.47 (s, 2H); LCꢀMS: m/z 235.0 (M ).
A mixture of 2,2'ꢀ(5ꢀbromoꢀ1,3ꢀphenylene)diacetonitrile (32) (236 mg, 1.00 mmol),
phenylboronic acid (147 mg, 1.21 mmol), PdCl (dppf)—CH Cl (41 mg, 0.05 mmol), and
2 2
2
potassium phosphate, tribasic (210 mg, 1.21 mmol) in 1,4ꢀdioxane (2.5 mL) and MeOH (0.63
mL) was refluxed for 2 h. The mixture was directly purified by silica gel column
1
chromatography to give 2,2'ꢀ([1,1'ꢀbiphenyl]ꢀ3,5ꢀdiyl)diacetonitrile (35) (176 mg, y. 74%). H
NMR (CDCl , 400 MHz): δ 3.82 (s, 4H), 7.26 (s, 1H), 7.36ꢀ7.41 (m, 1H), 7.43ꢀ7.47 (m, 2H),
3
ꢀ
7
.50 (s, 2H), 7.53ꢀ7.56 (m, 2H); LCꢀMS: m/z 231.1 (MꢀH) .
A mixture of 2,2'ꢀ([1,1'ꢀbiphenyl]ꢀ3,5ꢀdiyl)diacetonitrile (35) (66 mg, 0.28 mmol) in conc. HCl
(1 mL) was refluxed for 12 h. The mixture was purified by the Prep HPLC (condition B) to give
1
2
,2'ꢀ([1,1'ꢀbiphenyl]ꢀ3,5ꢀdiyl)diacetic acid (5) (42 mg, y. 55%). H NMR (DMSOꢀd , 400 MHz)
6
δ 3.62 (s, 4H), 7.14 (s, 1H), 7.37 (t, J = 7.3 Hz, 1H), 7.41 – 7.50 (m, 4H), 7.59 – 7.65 (m, 2H),
ꢀ
ꢀ
1
2.43 (s, 2H). LCꢀMS: m/z 269.3 (MꢀH) . HRMS (m/z) calcd for C H O (MꢀH) 269.0819,
16 14 4
found 269.0816.
1
2
,2'-(2'-Chloro-[1,1'-biphenyl]-3,5-diyl)diacetic acid (6). Prepared as described for 5. H NMR
(
DMSOꢀd , 400 MHz) δ 3.62 (s, 4H), 7.21 (d, J = 6.7 Hz, 3H), 7.40 (qt, J = 4.2, 6.0 Hz, 3H),
6
ꢀ
7.51 – 7.61 (m, 1H), 12.39 (s, 2H). LCꢀMS: m/z 303.0 (MꢀH) . HRMS (m/z) calcd for
ꢀ
C H ClO (MꢀH) 303.0430, found 303.0425.
16
13
4
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
2
,2'-(3'-Chloro-[1,1'-biphenyl]-3,5-diyl)diacetic acid (7). Prepared as described for 5. H NMR
(
DMSOꢀd , 400 MHz) δ 3.63 (s, 4H), 7.18 (s, 1H), 7.41 – 7.53 (m, 4H), 7.58 – 7.63 (m, 1H),
6
ꢀ
7
.68 (t, J = 1.8 Hz, 1H), 12.44 (s, 2H). LCꢀMS: m/z 303.0 (MꢀH) . HRMS (m/z) calcd for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
C H ClO (MꢀH) 303.0430, found 303.0426.
16
13
4
1
2,2'-(4'-Chloro-[1,1'-biphenyl]-3,5-diyl)diacetic acid (8). Prepared as described for 5. H NMR
(
DMSOꢀd , 400 MHz) δ 3.63 (s, 4H), 7.17 (s, 1H), 7.44 (s, 2H), 7.50 – 7.56 (m, 2H), 7.62 – 7.70
6
ꢀ
ꢀ
(
m, 2H), 12.40 (s, 2H). LCꢀMS: m/z 303.0 (MꢀH) . HRMS (m/z) calcd for C H ClO (MꢀH)
16 13 4
303.0430, found 303.0425.
2
,2'-(5-(Furan-2-yl)-1,3-phenylene)diacetic acid (9).
A mixture of 2,2'ꢀ(5ꢀbromoꢀ1,3ꢀphenylene)diacetonitrile (32) (112 mg, 0.28 mmol) in conc. HC1
(1.2 mL) was refluxed for 18 h. The mixture was purified by the Prep HPLC (condition B) to
1
give 2,2'ꢀ(5ꢀbromoꢀ1,3ꢀphenylene)diacetic acid (33) (112 mg, y. 86%). H NMR (DMSOꢀd , 400
6
MHz) δ 3.58 (s, 4H), 7.15 (s, 1H), 7.37 (s, 2H), 12.45 (brs, 2H).
A mixture of 2,2'ꢀ(5ꢀbromoꢀ1,3ꢀphenylene)diacetic acid (33) (50 mg, 0.18 mmol), furanꢀ2ꢀ
boronic acid (62 mg, 0.55 mmol), PdCl (dppf)—CH Cl (33 mg, 0.04 mmol), and CsF (152 mg,
2 2
2
o
1
.00 mmol) in DMF (2.6 mL) was stirred at 80 C for 2 h. The mixture was filtered and washed
with MeOH. The filtrate was concentrated in vacuo. The residue was purified by the Prep HPLC
1
(
condition B) to give 2,2'ꢀ(5ꢀ(furanꢀ2ꢀyl)ꢀ1,3ꢀphenylene)diacetic acid (9) (25 mg, y. 52%). H
NMR (DMSOꢀd , 400 MHz) δ 3.59 (s, 4H), 6.59 (dd, J = 1.8, 3.4 Hz, 1H), 6.91 (d, J = 2.8 Hz,
6
1
H), 7.06 (s, 1H), 7.49 (d, J = 1.3 Hz, 2H), 7.74 (d, J = 1.2 Hz, 1H), 12.41 (s, 2H). LCꢀMS: m/z
ꢀ
ꢀ
259.0 (MꢀH) . HRMS (m/z) calcd for C H O (MꢀH) 259.0612, found 259.0611.
1
4
12
5
ACS Paragon Plus Environment