Discovery and initial optimization of alkoxyanthranilic acid derivatives as inhibitors of HCV NS5B polymerase

Article abstract of DOI:10.1016/j.bmcl.2016.11.054

Alkoxyanthranilic acid derivatives have been identified to inhibit HCV NS5B polymerase, binding in an allosteric site located at the convergence of the palm and thumb regions. Information from co-crystal structures guided the structural design strategy. Ultimately, two independent structural modifications led to a similar shift in binding mode that when combined led to a synergistic improvement in potency and the identification of inhibitors with sub-micromolar HCV NS5B binding potency.

Full text of DOI:10.1016/j.bmcl.2016.11.054

Accepted Manuscript
Discovery and Initial Optimization of Alkoxyanthranilic Acid Derivatives as
Inhibitors of HCV NS5B Polymerase
Kyle Parcella, Andrew Nickel, Brett R. Beno, Steven Sheriff, Changhong Wan,
Ying-Kai Wang, Susan B. Roberts, Nicholas A. Meanwell, John F. Kadow
PII:
S0960-894X(16)31206-9
DOI:
http://dx.doi.org/10.1016/j.bmcl.2016.11.054
BMCL 24449
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
27 October 2016
18 November 2016
19 November 2016
Please cite this article as: Parcella, K., Nickel, A., Beno, B.R., Sheriff, S., Wan, C., Wang, Y-K., Roberts, S.B.,
Meanwell, N.A., Kadow, J.F., Discovery and Initial Optimization of Alkoxyanthranilic Acid Derivatives as
Inhibitors of HCV NS5B Polymerase, Bioorganic & Medicinal Chemistry Letters (2016), doi: http://dx.doi.org/
10.1016/j.bmcl.2016.11.054
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
Discovery and Initial Optimization of Alkoxyanthranilic Acid Derivatives as
Inhibitors of HCV NS5B Polymerase
Kyle Parcellaa*,†, Andrew Nickela,†,‡ , Brett R. Benoa, Steven Sheriffb, Changhong Wanb,#,Ying-Kai Wanga,
Susan B. Robertsa, Nicholas A. Meanwella, John F. Kadowa,§
aBristol-Myers Squibb, Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA. bBristol-Myers Squibb, Research and Development,
PO Box 4000, Princeton, NJ 08540, USA.
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Alkoxyanthranilic acid derivatives have been identified to inhibit HCV NS5B polymerase,
binding in an allosteric site located at the convergence of the palm and thumb regions.
Information from co-crystal structures guided the structural design strategy. Ultimately, two
independent structural modifications led to a similar shift in binding mode that when combined
led to a synergistic improvement in potency and the identification of inhibitors with sub-
micromolar HCV NS5B binding potency.
Keywords:
2016 Elsevier Ltd. All rights reserved.
Hepatitis C Virus (HCV)
NS5B Polymerase
Allosteric Inhibitors
Primer Grip
It is estimated that about 3% of the world’s population is
infected with hepatitis C virus (HCV).1 Chronic HCV infection
has been one of the leading causes of liver disease accounting for
approximately fifteen thousand deaths annually in the US.2 HCV
is a positive single stranded RNA virus which encodes for a
single polypeptide, which is subsequently cleaved into structural
and non-structural proteins.3 The non-structural proteins, which
have been shown to play critical roles in viral replication, have
emerged as popular targets for HCV chemotherapy. In recent
years, the standard of care has evolved rapidly from a pegylated
interferon-α/ribavirin-based treatment to a combination of direct-
acting antivirals.4 Of these targets, the RNA-dependent RNA
polymerase NS5B has generated particular attention from the
medicinal chemistry community.
Several allosteric sites on the enzyme have been investigated
in the context of small molecule inhibitors.3 About 10 Å away
from the active site in NS5B, a large 15 Å wide and 20 Å deep
cleft is known to bind a number of structurally diverse small
molecule inhibitors, including HCV-796 (1) which was the first
non-nucleoside NS5B inhibitor to demonstrate clinical proof of
concept (Fig 1).3 Generally speaking, these inhibitors are
characterized by varying degrees of resistance to the mutated
C316Y NS5B enzyme.
Figure 1. Literature allosteric inhibitors of HCV NS5B polymerase that bind
a large cleft 10-20 Å away from the active site.3,5
In an effort to identify a unique inhibitor that binds to the same
region of the NS5B enzyme, we initiated a high throughput
screening program. Representative compounds from the BMS
proprietary compound collection were screened for their ability
to inhibit the production of viral RNA by a truncated NS5B
genotype 1b enzyme at 50 M concentration. The compounds
that were active at this concentration were subjected to a counter
screen using an NS5B mutant containing the single point
__________
* Corresponding author. Tel.:203-415-8843;
e-mail address: kyle.parcella@bms.com
†These authors contributed equally.
‡Current address: Materia, 60 N San Gabriel Blvd. Pasadena, CA 91107.
# Current address: Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury Rd
Ridgefield, CT 06877
§Current address: ViiV Healthcare, 5 Research Parkway,
c/o Bristol-Myers Squibb Company, Wallingford CT, 06492

mutation of C316Y. The subset of compounds that were active in
the wild-type enzyme and inactive in the mutant were identified
as binding to NS5B in the vicinity of Cys316. This screening
cascade identified the 3-alkoxylanthranilic acid 5 (Fig 2) as a
single-digit micromolar inhibitor of NS5B.
Table 1. Removal of the Benzamide NH.
Figure 2. HTS screening hit against wild-type and C316Y mutant HCV
NS5B genotype 1b enzyme.6
Cmpd
IC50 (M)
> 50
6
7
8
8.7
32
This lead was soaked into crystals of a truncated NS5B enzyme
and X-ray quality crystals were obtained showing the compound
binding in a pocket often referred to as Palm Site 2. As shown in
Fig. 3, 5 binds in the region proximal to Cys316, with the central
aromatic ring oriented toward this residue while the chlorobenzyl
group projects into a large hydrophobic pocket.7 The carboxylate
and benzamide carbonyls form hydrogen-bonds with the
backbone NHs of Gly449 and Tyr448, respectively.
Co-crystallization of the 7-NS5B complex showed that the ligand
binds in a similar orientation as the acyclic amide, with the
benzamide carbonyl and hydrophobic chlorophenyl groups
residing in the same regions (Fig. 4)7. The ligand shifts
somewhat, however, pulling the carboxylate far enough away to
allow for a water molecule to bridge between Gly449 and the
carboxylate oxygen, rather than forming a direct hydrogen-bond
as seen in the carboxylate of 5.
Figure 3. Co-crystal structure of 5 bound to NS5B (PDB 5TRH). Compound
5 is depicted in ball and stick representation with orange carbon atoms.
Protein is shown with a white backbone cartoon and specific residues are
displayed in stick representation with green carbon atoms. Hydrogen bonds
are denoted as blue prolate ellipsoids. The omit Fo-Fc electron density is
contoured at 2 rmsd in magenta mesh. Image generated with The PyMOL
Molecular Graphics System (v. 1.8, Schrödinger, LLC).
Figure 4. Bound co-crystal structure of 7 and NS5B (PDB 5TRI). Compound
7 is depicted in ball and stick representation with orange carbon atoms.
Protein is shown with a white backbone cartoon and specific residues are
displayed in stick representation with green carbon atoms while water
molecules are shown as red spheres. Hydrogen bonds are denoted as blue
prolate ellipsoids. The omit Fo-Fc electron density is contoured at 2.5 rmsd in
magenta mesh. Image generated with The PyMOL Molecular Graphics
System (v. 1.8, Schrödinger, LLC).
We focused our initial optimization efforts on the central
benzamide portion of the molecule. The X-ray data suggested
that the benzamide NH played a role in binding affinity and N-
methylation resulted in a >6-fold decrease in potency (6, Table
1). However, this was attributed to a stabilization of the bio-
inactive cis-amide rotamer.8 To eliminate the issue of rotamers,
the cyclic benzamides 7 and 8 were prepared. Of these, the
isoindolinone was the preferred ring size and was approximately
equipotent to the parent benzamide.
The proximity of the benzamide phenyl ring in both 5 and 7 to
Arg386 and its directionality prompted us to consider appending
a carboxylate to that region of the molecule with the potential of
forming a salt bridge.

11a or the acyl sulfonamides 11c and 12c maintained measurable
activity but were generally weaker than the parent acids.
Transposition of the carboxylic acid was also investigated and
was found to have relatively minor effects on potency. This was
surprising considering the fact that extended carboxylates such as
meta-glycinamide 12d were not anticipated to be able to form the
same H-bond to Gly449 as seen in 5. Subsequent co-
crystallization of the NS5B-12d complex (Fig. 6) revealed that
12d binds in the same region of the enzyme proximal to Cys316,
with the ligand shifting slightly relative to 5 to allow for the
extended carboxylate to accept a hydrogen-bond from the OH of
Ser2887. This new interaction compensates for the lost hydrogen-
bond to Gly449.
Table 2. Dicarboxylate SAR
Cmpd
R
IC50 (M)
Table 3. Benzoate Alternatives
4.6
1.2
9a
9b
2.1
6.6
4.4
9c
10a
10b
11
10c
10d
1.7
Cmpd
R
IC50 (M)
Co-crystallization studies of 9b, the most potent of the
dicarboxylates, with the NS5B enzyme afforded X-ray quality
crystals7. The co-crystal structure shows that the appended
acetate indeed forms a salt bridge with Arg386, as anticipated,
and additionally establishes a hydrogen-bond with the OH of
Tyr415 (Fig. 5). It was anticipated that dicarboxylic acids of this
type would be poorly absorbed, so we turned our attention to the
possibility of replacing the benzoic acid with a more readily
absorbed surrogate that would be able to maintain the ability to
accept a hydrogen bond from Gly449.
25
11a
> 50
11b
24
10
11c
11d
12a
8.3
> 50
15
12b
12c
6.2
12d
13
0.88
Figure 5. Bound co-crystal structure of 9b and NS5B (PDB 5TRJ).
Compound 9b is depicted in ball and stick representation with orange carbon
atoms. Protein is shown with a white backbone cartoon and specific residues
are displayed in stick representation with green carbon atoms. Hydrogen
bonds are denoted as blue prolate ellipsoids. The omit Fo-Fc electron density
is contoured at 2 rmsd in magenta mesh. Image generated with The PyMOL
Molecular Graphics System (v. 1.8, Schrödinger, LLC).
A number of H-bond accepting isosteres were investigated as
benzoate alternatives (Table 3). Neutral species (i.e. amides 11b
and 12b) were not tolerated. Diffuse anions, such as tetrazole

Figure 7. Overlay of crystal structures NS5B bound to 5 and 1 which are
depicted in ball and stick representation with orange and green carbon atoms,
respectively.
Figure 6. Bound co-crystal structure of 12d and NS5B (PDB 5TRK).
Compound 12d is depicted in ball and stick representation with orange
carbon atoms. Protein is shown with a white backbone cartoon and specific
residues are displayed in stick representation with green carbon atoms while
water molecules are shown as red spheres. Hydrogen bonds are denoted as
blue prolate ellipsoids. The omit Fo-Fc electron density is contoured at 2
rmsd in magenta mesh. Image generated with The PyMOL Molecular
Graphics System (v. 1.8, Schrödinger, LLC).
Acknowledgments
I’d like to thank all those who helped make this paper possible,
including Mark Witmer and especially Kyle Eastman for his time
editing and mentorship. Use of the IMCA-CAT beamline 17-ID
and 17-BM at the Advanced Photon Source was supported by the
companies of the Industrial Macromolecular Crystallography
Association through a contract with Hauptman-Woodward
Medical Research Institute. This research used resources of the
Advanced Photon Source, a U.S. Department of Energy (DOE)
Office of Science User Facility operated for the DOE Office of
Science by Argonne National Laboratory under Contract No. DE-
AC02-06CH11357.
The process of relocating the carboxylic acid of 5 led to a shifted
binding mode in 12d that is similar to what was seen in the co-
crystal structure of the NS5B-7 complex (Fig. 4). Combining the
two structural modifications into a hybrid molecule 13 resulted in
a synergistic effect on potency with the IC50 value = 0.88 M.
Despite the improvement in NS5B enzyme potency compared to
the initial lead 5, 13 lacked whole cell activity in the HCV
replicon assay (EC50 > 100 µM)9 presumably due to the presence
of the carboxylic acid moiety which may limit cell permeability.
Further work is clearly needed to gain whole cell activity in this
series but this was not pursued in deference to an alternative
chemotype (vide infra).
In summary, functionalized anthranilic acid derivatives have
been identified to bind in an allosteric site located at the junction
of the thumb and palm regions of NS5B polymerase. This initial
SAR work guided by co-crystallographic studies resulted in
compound 13 with a 10-fold improvement in potency compared
to the initial HTS hit 5. Unfortunately, 13 was inactive in a
replicon assay, indicating that considerably additional
References and notes
1. Hanafiah M, Flasman G, Wiersma S. Global epidemiology of hepatitis
C virus infection: new estimates of age-specific antibody to HCV
seroprevalence. Hepatology. 2013 57, 4, 1333-42.
2. Centers for Disease Control and Prevention: Viral Hepatitis- Hepatitis
C. www.cdc.gov/hepatitis/hcv. Accessed 10/26/15.
3. Sofia, M. J.; Chang, W.; Furman, P. A.; Mosley, R. T. Ross, B. S.
Nucleoside, Nucleotide, and Non-Nucleoside Inhibitors of Hepatitis C
Virus NS5B RNA-Dependent RNA-Polymerase J. Med. Chem., 2012,
55, 2481–2531.
4. Kwong, A. The HCV Revolution Did Not Happen Overnight. ACS Med.
Chem. Lett., 2014, 5, 214–220.
optimization may be required.
However, the X-ray
crystallographic observations were exploited after an overlay of
the bound forms of anthranilic acid 5 with the benzofuran 1
suggested an interesting avenue for drug design (Fig. 7). The
central benzene ring of 5 projects in a direction that would
correspond to the C5 position of the benzofuran, a relationship
that suggested the potential for merging elements of both
chemotypes by exploring substitution at the C5 position of 1. An
exploration based on this hypothesis as a means of improving
activity ultimately led to an inhibitor which was advanced into
phase 1 clinical studies10.
5. Pfefferkorn, J.; Greene, M.; Nugent, R.; Gross, R.; Mitchell, M.; Finzel,
B.; Harris, M.; Wells, P.; Shelly, J.; Anstadt, R.; Kilkuskie, R.; Kopta,
L.; Schwende, F. Inhibitors of HCV NS5B polymerase. Part 1:
Evaluation of the southern region of (2Z)-2-(benzoylamino)-3-(5-
phenyl-2-furyl)acrylic acid. Bioorg.Med. Chem. Lett. 2005, 15, 2481-
2486.
6. IC50 values were determined in a HCV NS5B RdRP enzyme assay
described in WangY-K.; Rigat K.; Roberts S.; Gao M. Anal. Biochem.
2006, 359, 106-111.
7. Coordinates and data for structures of the following compounds in
complex with NS5B 1b WT have been deposited in the wwPDB with
the following IDs: 5 (5TRH), 7 (5TRI), 9b (5TRJ), 12d (5TRK).
8. Itai, A; Toriumi, Y; Tonrioka, N; Kagechika, H; Isao Azumaya, I;
Shudo, K.Tet. Lett; 1989, 30, 45, 6177-6180.
9. EC50 values were determined in a HCV replicon luciferase whole cell
assay, as described in Eastman, K.; Parcella, K. Kadow, J. US Patent
20140275154.
10. Yeung, K-S; Parcella, K.; Beno, B.; Bender, J; Grant-Young, K.; Rigat,
K; Wang, Y-K; Liu, M.; Lemm, J.; Mosure, K.; Hanumegowda, U.;
Zhuo, X.; Parker, D.; Sinz, M.; Santone, K.; Smith, D.; Li, J.;
Fraunhoffer, K.; Delmonte, A.; Colston, E.; Pasquinelli, C.; Gao, M.;
Meanwell, N.; Roberts, S.; Knipe, J.; Kadow, J. Discovery of BMS-
929075 an HCV NS5B replicase allosteric inhibitor advanced to phase 1
clinical studies. 250th ACS National Meeting & Exposition, Boston,
MA, United States, August 16-20, 2015, MEDI-32. Manuscript in
preparation.
Supplemental Materials
Supplemental information associated with this article can be
found
online.

Discovery and Initial Optimization of
Alkoxyanthranilic Acid Derivatives as Inhibitors
of HCV NS5B Polymerase
Leave this area blank for abstract info.
Kyle Parcella, Andrew Nickel, Brett R. Beno, Steven Sheriff, Changhong Wan, Ying-Kai Wang,
Susan B. Roberts, Nicholas A. Meanwell, John F. Kadow