Discovery of BI 207524, an indole diamide NS5B thumb pocket 1 inhibitor with improved potency for the potential treatment of chronic hepatitis C virus infection

Article abstract of DOI:10.1021/jm501532z

The development of interferon-free regimens for the treatment of chronic HCV infection constitutes a preferred option that is expected in the future to provide patients with improved efficacy, better tolerability, and reduced risk for emergence of drug-resistant virus. We have pursued non-nucleoside NS5B polymerase allosteric inhibitors as combination partners with other direct acting antivirals (DAAs) having a complementary mechanism of action. Herein, we describe the discovery of a potent follow-up compound (BI 207524, 27) to the first thumb pocket 1 NS5B inhibitor to demonstrate antiviral activity in genotype 1 HCV infected patients, BILB 1941 (1). Cell-based replicon potency was significantly improved through electronic modulation of the pK_a of the carboxylic acid function of the lead molecule. Subsequent ADME-PK optimization lead to 27, a predicted low clearance compound in man. The preclinical profile of inhibitor 27 is discussed, as well as the identification of a genotoxic metabolite that led to the discontinuation of the development of this compound.

Full text of DOI:10.1021/jm501532z

Subscriber access provided by DICLE UNIV
Article
Discovery of BI 207524, an indole diamide NS5B thumb pocket 1 inhibitor
with improved potency for the potential treatment of chronic HCV infection
Pierre L. Beaulieu, Paul C. Anderson, Richard Bethell, Michael Bös, Yves Bousquet,
Christian Brochu, Michael Cordingley, Gulrez Fazal, Michel Garneau, James R Gillard,
Stephen H Kawai, Martin Marquis, Ginette McKercher, Marc-André Poupart, Timothy
Stammers, Bounkham Thavonekham, Dominik Wernic, Jianmin Duan, and George Kukolj
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm501532z • Publication Date (Web): 13 Nov 2014
Downloaded from http://pubs.acs.org on November 17, 2014
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 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of BI 207524, an indole diamide NS5B thumb pocket 1
inhibitor with improved potency for the potential treatment of
chronic HCV infection
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
Pierre L. Beaulieu*, Paul C. Anderson, Richard Bethell, Michael Bös, Yves Bousquet, Christian
Brochu, Michael G. Cordingley, Gulrez Fazal, Michel Garneau, James R. Gillard, Stephen
Kawai, Martin Marquis, Ginette McKercher, Marc-André Poupart, Timothy Stammers,
Bounkham Thavonekham, Dominik Wernic, Jianmin Duan and George Kukolj.
Boehringer Ingelheim (Canada) Ltd., Research and Development, 2100 Cunard Street, Laval
(Quebec), Canada, H7S 2G5
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 54
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
ABSTRACT
The development of interferonꢀfree regimens for the treatment of chronic HCV infection
constitutes a preferred option that is expected in the future to provide patients with improved
efficacy, better tolerability and reduced risk for emergence of drugꢀresistant virus. We have
pursued nonꢀnucleoside NS5B polymerase allosteric inhibitors as combination partners with
other direct acting antivirals (DAAs) having a complementary mechanism of action. Herein, we
describe the discovery of a potent followꢀup compound (BI 207524, 27) to the first thumb pocket
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
NS5B inhibitor to demonstrate antiviral activity in genotype 1 HCV infected patients, BILB
941 (1). Cellꢀbased replicon potency was significantly improved through electronic modulation
of the pKa of the carboxylic acid function of the lead molecule. Subsequent ADMEꢀPK
optimization lead to 27, a predicted low clearance compound in man. The preclinical profile of
inhibitor 27 is discussed, as well as the identification of a genotoxic metabolite that led to the
discontinuation of the development of this compound.
2
ACS Paragon Plus Environment

Page 3 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
HCV is a bloodꢀborne pathogen that causes serious and often fatal liver damage and is the
leading cause of liver transplantation in industrialized nations. An estimated 130 – 170 million
people are infected worldwide and conditions associated with HCV infection represent a
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
1
significant economic burden on health care systems. The viral functions encoded by the HCV
genome have been the focus of intense drug discovery efforts since the isolation and
2,3
characterization of the virus 25 years ago. The recent approval in 2011 of the first HCV Direct
Acting Antivirals (DAAs) has provided improved treatment options for patients suffering from
chronic hepatitis C virus infection. First generation NS3/4A protease inhibitors
TM
TM
(
boceprevir/Victrelis from Merck and telaprevir/Incivek from Vertex Pharmaceuticals) have
resulted in improved efficacy for the treatment of genotypeꢀ1 (gt1) HCV infection when
complementing the previous standard of care (SoC) consisting of pegylated interferonꢀα
4
(
PegIFN) and ribavirin (RBV), a broadꢀspectrum antiviral agent. However, despite improved
overall outcomes, these treatments suffer from the combined side effects of the SoC and the new
drugs themselves, resulting in significant discontinuations of therapy. Furthermore, the treatment
remains ineffective in up to 20 to 30% of treatment naïve gt1 patients which predominates in
North American, Europe and Japan. The lack of efficacy in some patient populations is caused in
5
part by rapid emergence of virus that is resistant to the new agents. Secondꢀgeneration protease
inhibitors with improved efficacy, lower dosing requirements and increased tolerability such as
TM
faldaprevir (Boehringer Ingelheim, Phase 3) and recently approved simeprevir/Olysio
TM
(
Medivir/Tibotec), as well as the nucleotide NS5B inhibitor sofosbuvir/Solvadi (Gilead) have
provided patients with improved outcomes (80 to >90% cure rates in gt1), but still require
6
combination with SoC to reduce emergence of resistant virus in this difficult to treat population.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 54
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
The development of interferonꢀfree regimens would represent a preferred option and relies on
combination of potent DAAs with complementary modes of action, which through a concerted
assault on virusꢀessential functions, are expected to inhibit HCV replication while minimizing
selection of resistant variants. Such regimens would provide improved tolerability by eliminating
the necessity for interferonꢀbased immune coꢀtherapy. Examples of DAA combinations in late
stage clinical investigation include permutations of NS3/4A protease, nucleoside/nucleotide and
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
7
nonꢀnucleoside NS5B polymerase and NS5A function inhibitors.
We have been actively pursuing nonꢀnucleoside NS5B allosteric inhibitors as potential
7
a,b
combination partners with other DAAs.
The HCVꢀencoded NS5B RNAꢀdependent RNA
polymerase is a key component of the replication machinery of the virus and has provided fertile
8
,9
grounds for the discovery of antiviral agents. In addition to nucleoside and nucleotide
substrateꢀbased inhibitors, this enzyme which catalyzes the replication of viral RNA, has proven
susceptible to inhibition by agents that interfere with conformational changes that are important
in the early steps of viral RNA synthesis. Molecules that bind to one of 5 distinct allosteric sites
(denoted thumb sites 1 and 2 and palm sites 1, 2 and 3) have been identified and the mechanisms
through which they inhibit enzymatic activity have been validated both in cellꢀbased replicon
1
0
assays that reproduce replication of subgenomic HCV RNA and subsequently in the clinic for
8
all but Palm site 3. We recently described the discovery of 1 (BILB 1941, Figure 1), the first
thumb pocket 1 NS5B inhibitor to demonstrate antiviral activity in patients infected with gt1
HCV. Compound 1 is an indoleꢀbased cinnamic acid derivative that displays moderate antiviral
activity in gt1a/1b replicon assays (EC50 = 153 and 84 nM respectively) and favorable
1
1
pharmacokinetics in preclinical animal species. In a 5ꢀday multiple rising dose monotherapy
trial in gt1ꢀHCV infected patients, 1 produced significant antiviral activity (up to 2.5 log10
4
ACS Paragon Plus Environment

Page 5 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
12
reduction in viremia) when dosed at 450 mg q8h. Unfortunately, GI intolerance precluded
testing of the compound at higher doses but this proofꢀofꢀconcept trial suggested that more
potent analogs may improve antiviral response. We describe herein the discovery of a followꢀup
candidate from this class, which displayed improved potency compared to 1 in both gt1a/1b
replicons while maintaining a similarly favorable PK profile.
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 1. Structure of the first thumb pocket 1 NS5B inhibitor with antiviral activity in man
O
H
N
N
N
H
O
N
COOH
1
(BILB 1941)
1b NS5B IC50 = 60 nM
a/1b replicon EC50 = 153 / 84 nM
1
RESULTS and DISCUSSION
As mentioned above, a followꢀup candidate to 1 requires potency improvement while
maintaining a similar or better PK profile (e.g., longer human t ). Since the preclinical
1
/2
toxicological profile of 1 in animals was not predictive of the GI intolerance side effects that
were observed at doses required to provide a uniform response across HCV gt1a and gt1b
infected subjects, we postulated that further structural optimization leading to increased potency
could translate into improved response, increased tolerability and/or reduced dosing
requirements. SAR which led to the discovery of compound 1 provided the foundation for an
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 54
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
optimization strategy of this lead. For example, the cyclopentyl ring at the 3ꢀposition of the
indole scaffold was maintained even though replacement by a cyclohexyl ring was known to
provide 2 – 3 fold potency improvements, but at the expense of metabolic stability due 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
11
CYP450ꢀmediated oxidative metabolism of that substituent. Likewise, no advantage was
1
3
anticipated from expanding further from the indole nitrogen methyl substituent. On the other
hand, the Cꢀ2 heterocyclic substituent, the central amino acid linker and the rightꢀhandꢀside
cinnamic acid moiety were shown to be determinants and modulators of potency and ADMEꢀPK
1
1
profiles. Since the presence of the cinnamic acid moiety was beneficial for potency and
solubility (compound 1 apparent pKa = 4.5 as derived from a pHꢀtitration profile) and did not
compromise Cacoꢀ2 permeability, cellular activity or oral absorption, we investigated the effect
of adding substituents on the cinnamic acid aromatic ring in an attempt to improve the profile of
1
. Both the steric and electronic impact of substituents on this part of the molecule (e.g.,
conformational rigidification, pKa of carboxylic acid group) were investigated and initial
1
4
outcomes are reported in Table 1.
As can be seen in Table 1, substitution in the orthoꢀposition relative to the nitrogen atom
compounds 2 – 7) reduced potency compared to unsubstituted analog 1, particularly in the case
(
of bulky substituents (e.g., CF , 5) suggesting that substitutions affecting coꢀplanarity of the
3
15
amide linkage relative to the phenyl ring were detrimental to interactions with NS5B. On the
other hand, introduction of substituents in the orthoꢀposition relative to the acrylic acid moiety
(
compounds 8 – 14) were well tolerated, even for bulky groups such as CF that remained within
3
3ꢀfold potency (compound 11) of 1. The introduction of electron donating alkoxy groups ortho
relative to the acrylic acid function provided the most benefit. For example, a methoxy
substituent produced a 2.5ꢀfold improvement in replicon potency (compound 13). Furthermore,
6
ACS Paragon Plus Environment

Page 7 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
an additive effect was observed upon methoxylation of both orthoꢀpositions (compound 14)
which led to an analog that exhibited close to a 10ꢀfold improvement in potency (EC = 8 nM)
50
relative to our previous clinical candidate 1. The improved antiviral potency of alkoxyꢀcinnamic
acid derivatives such as 13 and 14 cannot be rationalized on the basis of their intrinsic potency as
measured in our biochemical assays or lipophilicity (calculated logP = 5.4 and 5.3 for 1 and 13
respectively). On the other hand, the electronic impact of electronꢀdonating alkoxy substituents
on the ionizability (i.e., pKa as derived from pH titration curves) of the cinnamic acid is
significant (measured pKa = 4.5 and 6.5 for 1 and 15 respectively) and may account for the
increased ability of alkoxyꢀderivatives to permeate hepatocyte membranes and inhibit replication
of subgenomic RNA in cellꢀbased replicon assays. This hypothesis is corroborated by cell
permeability measurements in a Cacoꢀ2 assay where apparent apical → basolateral permeability
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
ꢀ6
values of 11.2 X 10 cm/sec and 19.0 X 10 cm/sec were observed for 1 and 15 respectively.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 54
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
Table 1. Initial exploration of substituted right-hand-side cinnamic acid derivatives
8
ACS Paragon Plus Environment

Page 9 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
H
X
N
N
N
H
O
N
COOH
1
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
IC50
ꢀM)a,b (ꢀM)a,c
EC50
IC50
EC50
Entry
cinnamic acid
Entry
cinnamic acid
(ꢀM)a,b (ꢀM)a,c
(
H
N
H
N
1
0.060
0.084
1.03
8
0.069
0.061
0.079
0.13
H
N
H
N
F
2
0.37
9
F
H
H
N
N
Cl
3
4
0.23
0.30
0.56
10
11
0.10
0.18
0.11
0.23
Cl
H
H
N
N
CF3
CF3
H
H
N
N
OH
5
6
7
2.25
> 2.5
12
13
14
0.044
0.17
F
H
H
N
N
OMe
OMe
1.32
0.44
>2.8
0.25
0.047
0.34
0.032
0.008
F
OH
H
H
N
N
OMe
a
values are an average of at least two determinations. Full length NS5B enzymatic assay.14a cLuc
b
14c,d
reporter assay
Although compound 14 had reasonable stability in the presence of human and rat liver
microsomes (HLM/RLM T1/2 = 121/42 min) and good permeability as measured in a Cacoꢀ2 cell
ꢀ6
assay (apical → basolateral = 17.6 x 10 cm/sec) 14 was poorly absorbed when administered
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 54
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
orally to rats in a cassette screen (mixture of 4 compounds, see experimental section for details)
at 4 mg/kg, with plasma inhibitor concentrations reaching 0.4 and 0.3 ꢀM at 1 and 2 h time
points respectively. Monoꢀsubstituted methoxy analog 13, with EC50 = 32 nM and C1h/2h
=
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
1.1/0.5 ꢀM in a 4 mg/kg rat cassette screen, became a prototype for further investigation of
alkoxycinnamic acid derivatives as shown in Table 2.
Table 2. Alkoxy cinnamic acid SAR
1
0
ACS Paragon Plus Environment

Page 11 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
H
N
OR
N
N
H
O
N
COOH
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
IC50
ꢀM)a,b (ꢀM)a,c
EC50
calcd plasma C
1h/2h
log P
HLM T1/2
Compound
OR
e
(ꢀM)f
(
min)d
(
1
3
5
OMe
OEt
0.047
0.032
0.024
82
64
5.3
5.7
1.1/0.5
0.9/0.4
1
0.048
0.13
0.11
1
6
7
OnPr
OiPr
0.036
0.036
124
114
6.1
6.2
0.4/0.2
1.3/0.8
1
18
OiBu
0.42
0.25
0.055
0.046
>300
231
6.4
6.0
O
19
20
O
0.28
0.065
0.027
50
87
6.2
5.1
O
21
0.056
OMe
a
b
values are an average of at least two determinations. Full length NS5B
14a c
14c,d
d
enzymatic assay.
Luc reporter assay
. Human liver microsome
stability. e JChem 5.0.0 (http://www.chemaxon.com). Following oral
administration to rats as mixtures of four compounds at a dose of 4 mg/kg
each.
f
Variations in intrinsic potency of compounds (Table 2) in the enzymatic NS5B assay apparently
were not reflected in the cellꢀbased replicon assay as all compounds had comparable potency
(
EC = 24 – 65 nM). Thus, increasing the size of the alkoxy substituent (e.g., 16 – 20) did not
50
provide advantages over smaller groups such as methoxy and ethoxy (compounds 13 and 15). In
addition to simple alkyl ethers, functionalized ethers were also examined. For example,
methoxyethoxy analog 21 had reduced lipophilicity and a comparable profile to 13 and 15. Ether
side chains containing basic substituents such as ethylamino, picolinyl and ethylmorpholino were
also investigated but did not provide any potency advantage and in some cases, compromised
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 54
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
other key parameters such as Cacoꢀ2 permeability when zwitterionic species were generated
(results not shown). As cellꢀbased potency improved, the optimization strategy shifted some
focus to identifying analogs that also displayed favorable ADME and PK properties, thus
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
advancing their overall profile toward eventual selection of a followꢀup compound to 1.
The metabolic stability of the analogs presented in Table 2 ranged from satisfactory to excellent
when incubated with human liver microsomes (t1/2 = 50 – >300 min) and inhibitors 15 – 17 had
ꢀ6
excellent Cacoꢀ2 permeability (Cacoꢀ2A→B = 19 – 23 x 10 cm/sec). A cassette of 4 compounds
(13, 15 – 17) was administered orally to rats at a dose of 4 mg/kg each and compound plasma
exposures were measured at 1 h and 2 h time points. Consistent with the Cacoꢀ2 permeability
data, alkoxycinnamic acid derivatives displayed promising oral absorption in rats (plasma C =
1h
0.9 – 1.3 ꢀM), except for nPr analog 17 which showed reduced exposure and was eliminated
from further consideration. While both methoxy and ethoxy cinnamic acid derivatives were
subsequently pursued in parallel for some time, analogs bearing an ethoxy substituent often
displayed improved properties over corresponding methoxy analogs and efforts were eventually
focused on further optimization of 15.
Studies that lead to the discovery of 1 had revealed the ability of indole Cꢀ2 substituents to
11
modulate the physicochemical properties and ADMEꢀPK profile of this class of inhibitors. This
position thus became the next focus of exploration in our attempts to identify potent analogs that
maintain an overall PK profile minimally comparable 1. Approximately 30 variations were
examined, much of the selection building on SAR previously established in the unsubstituted
1
1
cinnamic acid series. These included alkyl, aryl, 5ꢀ and 6ꢀmembered heterocycles and a
selection of those analogs is presented in Table 3. Although a variety of Cꢀ2 heterocycles
provided compounds with good antiviral potency in the gt1b luciferase replicon assay (EC < 30
5
0
1
2
ACS Paragon Plus Environment

Page 13 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
nM), only a limited set of analogs were identified that combined good cellꢀbased potency with a
satisfactory in vitro ADME profile and maintained lipophilicity in a satisfactory range. In
particular, haloꢀpyridine and haloꢀpyrimidine derivatives (e.g., 24, 27, 28) displayed good to
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
excellent metabolic stability upon incubation with human or rat liver microsomes (HLM t1/2
=
273 – >300 min; RLM t1/2 = 104 – >300 min) and excellent Cacoꢀ2 permeability (A → B >
ꢀ6
1
3x10 cm/sec). The in vitro ADME profiles translated into good oral plasma exposure at the 1
h time point when compounds were screened in rats as a mixture of 4 compounds, each dosed at
mg/kg. In particular, haloꢀpyrimidine analogs 27 and 28, as well as fluoropyridine 23 displayed
high exposures (C = 3.4 and 1.8 µM for pyrimidines 27 and 28 and 1.3 µM for fluoropyridine
4
1
h
23).
Table 3. C-2 optimization of 2-ethoxycinnamic acid analogs
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 54
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
O
H
N
OEt
N
N
Aryl, HET
H
O
COOH
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
IC50
EC50
ꢀM)a,b (ꢀM)a,c
calcd Cacoꢁ2 plasma C1h/2h
HLM T1/2
min)d
Compound
Cꢁ2
N
e
(cm/sec)f
log P
(ꢀM)g
(
(
15
0.048 0.024
64
5.7 19 x 10ꢁ6
0.9/0.4
22
0.41
0.021
6.9
23
F
0.11
0.33
0.33
0.016
0.026
0.010
72
81
78
5.9 17 x 10ꢁ6
6.3 18 x 10ꢁ6
5.6 9.4 x 10ꢁ6
1.3/0.8
0.5/0.3
0.1/0.1
N
24 Cl
N
2
5
6
N
H N
2
N
2
0.048 0.021
46
5.0
N
N
27 Cl
0.11
0.46
0.013
0.019
102
154
5.3 13 x 10ꢁ6
3.4/2.3
N
N
13 x 10ꢁ6
28
Br
5.5
1.8/2.1
0.4/0.1
N
N
29
0.19
0.15
0.012
0.012
44
36
5.4
5.1
N
N
30
S
a
b
values are an average of at least two determinations. Full length NS5B enzymatic
. Human liver microsome stability. JChem 5.0.0
http://www.chemaxon.com). Apical to basolateral direction. Following oral
14a c
14c,d d
e
assay.
Luc reporter assay
f
g
(
administration to rats as mixtures of four compounds at a dose of 4 mg/kg each.
Previous studies performed in the unsubstituted cinnamic acid series had shown that a
cyclobutane 2ꢀaminocarboxylic acid linker in the central portion of the molecule provided
1
4
ACS Paragon Plus Environment

Page 15 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
11
superior PK profiles (e.g., compound 1). Once again in this series, while gemꢀdimethyl linkers
had excellent cellꢀbased potency (e.g., 31 and 33), these analogs displayed significantly reduced
oral exposure in rat cassettes when compared to the corresponding cyclobutane derivatives (27
and 34). Similarly, cyclopropane derivatives (e.g. 32) exhibited reduced potency as a result of
slightly altered dihedral angles (as determined by NMR studies) in the central portion 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
15
unbound molecule compared to the NS5Bꢀbound conformation of inhibitors. As shown in
Table 4, inhibitor 34 had an overall wellꢀbalanced profile encompassing potency, in vitro ADME
properties and in vivo rat exposure.
Table 4. Central amino acid linker modifications
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 54
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
O
R
R
H
N
OR'
N
N
N
N
Cl
H
O
COOH
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
R
R
HLM T1/2
calcd Cacoꢁ2 plasma C1h/2h
Compound
IC50
EC50
ꢀM)a,b (ꢀM)a,c
OR'
N
log Pe (cm/sec)f
(ꢀM)g
(
min)d
(
H
O
5.3 24 x 10ꢁ6
3
1
2
OEt
OEt
N
0.027 0.008
0.5/0.1
H
O
3
0.38
0.034
195
129
4.9
N
H
O
O
O
O
4.7 12 x 10ꢁ6
4.7 17 x 10ꢁ6
33
N
H
0.20
0.20
0.010
0.013
0.2/0.0
1.7/1.1
OMe
34
>300
OMe
N
H
O
a
values are an average of at least two determinations. Full length NS5B enzymatic assay.14a cLuc
. Human liver microsome stability. JChem 5.0.0 (http://www.chemaxon.com).
Apical to basolateral direction. Following oral administration to rats as mixtures of four compounds
b
14c,d d
e
reporter assay
f
g
at a dose of 4 mg/kg each.
As depicted in Figure 2, additional structural modifications to the rightꢀhandꢀside
alkoxycinnamic acid moiety were explored, including substitution of the cinnamic acid double
bond with methyl, ethyl or fluorine groups (35) in an attempt to explore steric shielding of the
acrylic acid moiety and electronic effects, as well as conformational rigidification (36). While
these modifications were relatively well tolerated, none provided an advantage over the original
alkoxycinnamic acid moiety (results not shown).
Figure 2. Cinnamic acid double bond substitution and rigidification
1
6
ACS Paragon Plus Environment

Page 17 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
H
H
N
O
N
OR
R'
COOH
COOH
35
36
R = Me, Et
R' = Me, Et, F
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
Three analogs which had improved replicon potency compared to 1 and also showed promising
preliminary ADMEꢀPK characteristics were selected for advanced profiling. Individual rat PK
profiles for compounds 23, 27 and 34 are shown in Table 5 and compared to that of clinical
candidate 1. Compound 27 displayed comparable exposure in rat to compound 1 albeit with
lower bioavailability, but clearance was reduced and halfꢀlife was increased from 1.5 to 2.4 h in
this rodent species. Characteristic of carboxylic acid containing molecules, the volume of
distribution remained low but comparable to 1. In rats, compound 27 showed favorable
distribution from plasma to the liver target organ with liver/plasma ratio ~5 at 6 h post dosing.
Table 5. Single compound Rat PK parameters.
a
b
Oral parameters (10 mg/kg)
IV parameters (2 mg/kg)
Cmax
AUC
(µM*h)
23
MRT
(h)
4.1
Vss
(L/kg)
0.48
CL
(mL/kg/min)
8.3
T1/2
(h)
1.5
F%
(µM)
1
7.5
59
17
19
16
c
c
c
2
3
7
4
2.2
6.3
4.1
6.1
19
3.0
3.6
4.1
0.42
0.38
0.29
6.2
2.6
3.7
1.5
2.4
2.1
2
3
10.6
a
b
Compound dosed as an oral suspension in 0.5% Methylcellulose and 0.3% Tweenꢀ80. Bolus
c
injection prepared in 70% PEGꢀ400:30% water. 1% NꢀMethylpyrrolidone (NMP) was added to
the oral dosing solution.
Compound 27 was profiled in dog and monkey to provide crossꢀspecies data for human PK
predictions and the results are compared to 1 in Table 6. Compound 27’s PK parameters in dogs
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 54
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
and monkeys were comparable to compound 1. As in rats, a low volume of distribution was
observed but clearance was slightly reduced almost twoꢀfold in both species relative to 1. While
plasma exposures following oral administration were superior to compound 1 (likely as a result
of reduced clearance) bioavailability for 27 was lower in dogs and comparable in monkeys 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
11
those previously reported for compound 1. Based on its crossꢀspecies PK profile and following
1
6
allometry principles, compound 27 was predicted to be a low clearance compound in human
estimated IV Cl = 0.7 mL/min/kg, estimated t1/2 ~8.5 h, suitable for BID or TID dosing
(
regimens).
Table 6. Dog and rhesus monkey PK profiles for 1 and 27.
a
b
Oral parameters (10 mg/kg)
IV parameters (2 mg/kg)
Cmax
(µM)
15.3
AUC
(µM*h)
99
MRT
(h)
5.4
Vss
(L/kg)
0.54
CL
(mL/kg/min)
1.8
T1/2
(h)
4.3
F%
1
(dog)
64
29
70
72
c
27 (dog)
8.1
6.4
9.6
76
20.5
53
8.0
2.9
6.1
0.38
0.70
0.60
1.0
9.5
3.6
4.5
1.7
3.0
1
(monkey)
c
27 (monkey)
a
b
Compound dosed as an oral suspension in 0.5% Methylcellulose and 0.3% Tweenꢀ80. Bolus
c
injection prepared in 70% PEGꢀ400:30% water. 1% NꢀMethylpyrrolidone (NMP) was added to
the oral dosing solution.
Additional profiling data for compound 27 (BI 207524) is summarized in Table 7. In a
1
4a
biochemical assay using a NS5B∆21 polymerase construct, compound 27 had an IC = 84 nM
5
0
(comparable to 1) and was selective against poliovirus RNAꢀdependent RNA polymerase and a
mammalian DNAꢀdependent RNA polymerase isolated from calf thymus (IC = 124 and 70 ꢀM
5
0
1
4a
respectively). In a gt1b replicon assay where subgenomic RNA was quantified by Taqman
1
4b
assay, 27 was approximately 7ꢀfold more potent than 1 with an EC = 11 nM. As expected for
5
0
1
8
ACS Paragon Plus Environment

Page 19 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
thumb pocket 1 NS5B inhibitors, a ~3ꢀfold shift was observed in a gt1a replicon assay (EC =
5
0
2
9 nM) relative to gt1b. In Huhꢀ7 cells, the compound was not significantly cytotoxic (CC = 38
50
ꢀM in a MTT assay). This class of compound is typically highly protein bound (>99% for 27 and
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
1
in human plasma) but only slight shifts in EC50 values were observed in replicon assays
performed in the presence of up to 40% human serum (3.8ꢀfold for 27). Compound 27 exhibited
moderate lipophilicity as reflected by a measured logD (pH 7.4) = 4.1 and was only weakly
acidic (pKa = 6.5) which is in line with the observed improvements in cellꢀbased potency and the
ꢀ6
high Cacoꢀ2 permeability value (13 x 10 cm/sec). Metabolic stability in human liver
microsomes was moderate (t1/2 = 102 min) and comparable to that of compound 1 while the
CYP450 inhibition profile was improved with the lowest IC values measured against the 2C9
5
0
isozyme (IC = 13.5 ꢀM) .
5
0
Table 7. Structure and profile of compound 27 (BI 207524).
O
H
N
O
N
N
N
N
Cl
H
O
COOH
27 (BI 207524)
Compound 1 Compound 27
1
4a
Gt1b NS5B∆21 IC (nM)
67
84
29 / 11
38 µM
3.8 fold
99.6%
5
0
1
4b
Gt 1a/1b replicon EC (nM)
153 / 84
93 µM
3 fold
99.2%
3.3 / 4.5
90
5
0
MTT TC50
Serum shift in 40% human serum
Human plasma protein binding
logD (pH 7.4) / pKa
4.1 / 6.5
102
HLM t1/2 (min)
CYP450 inhibition: IC (µM)
13/3.5/2.9
>30/>30
11 x 10
>30/13.5/>30
>30/>30
5
0
1A2/2C9/2C19/2D6/3A4
a
ꢀ6
ꢀ6
Cacoꢀ2 permeability (cm/sec)
13 x 10
a
apical pH 6.0/basolateral pH 7.4
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 54
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
Compound 27 was nonꢀmutagenic in the Ames test in the presence or absence of metabolic
activation at dose levels up to maximum solubility. The compound showed no inhibition of the
hERG channel at 0.8 ꢀM (solubility limit) and no effect in the action potential duration assay in
isolated guinea pig papillary muscle at 90% repolarization (APD90). In conscious telemeterized
dogs, no effects were observed on cardiovascular function and electrocardiographic parameters
up to the highest dose of 200 mg/kg (plasma Cmax = 18 ꢀM). Results from in vivo general
pharmacology studies performed in rats and a receptor binding assay screen did not lead to
findings that would preclude advancement into more advanced toxicology studies and further
development.
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
We recently reported a model to predict the pharmacokineticꢀpharmacodynamic relationship for
direct acting antiꢀHCV agents in humans, based on the liver partition coefficientꢀcorrected
inhibitory quotient (LCIQ) derived from in vitro hepatocyte partition coefficients and animal in
1
7
vivo liver/plasma exposures. Available clinical data for HCV DAAs, suggests that a C LCIQ
min
≥
500 is generally predictive of a strong antiviral response. According to this model and based on
its gt1a antiviral potency and predicted human liver partitioning coefficient, regimens of 400 mg
of compound 27 administered twice a day or 230 mg administered threeꢀtimes a day should
maintain Cmin liver concentrations 500ꢀfold gt1a EC (LCIQ ) predictive of a strong antiviral
50
500
response in humans.
However, during in vitro and in vivo crossꢀspecies metabolite studies performed in support of the
development of 27, small amounts of a genotoxic 4ꢀaminoꢀ2ꢀethoxycinnamic acid metabolite
were observed. This metabolite resulted from cleavage of the anilide bond and was also detected
in animal plasma samples from PK studies and upon incubation of 27 with simulated gastric
2
0
ACS Paragon Plus Environment

Page 21 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
fluids (SGF). In human liver microsomes, the metabolite was formed at a rate of ~1 ppm/min in
a NADPHꢀindependent manner. In light of the challenges associated with the progression of such
1
8
a compound, development of 27 was discontinued and efforts focused on compounds without
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
liabilities associated with the formation of genotoxic metabolites.
SYNTHESIS OF INHIBITORS
Inhibitors were synthesized in a convergent fashion through amide bond coupling between
indoleꢀ6ꢀcarboxylic acid derivatives and amine fragments as shown in scheme 1 and as
1
1
previously described for compound 1.
6ꢀIndolecarboxylic acid derivatives 38 were prepared from protected 2ꢀbromoindole 37 and
commercially available aryl or heteroaryl halides using standard crossꢀcoupling protocols as
1
1
previously described. Amide bond coupling with amines 40, themselves derived from
condensation of α,αꢀdisubstituted amino acids with 4ꢀaminocinnamate esters 39 provided
directly inhibitors 41 after deprotection of the carboxylic acid function by saponification.
Alternatively, Indoles 38 could be coupled in a stepwise fashion to protected α,αꢀdisubstituted
amino acid esters to provide azalactones 42 and then condensed with 4ꢀamino cinnamate esters
to provide 41 after deprotection.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 54
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
Scheme 1. General synthetic scheme
COOMe
N
1
Br
R
2
R
N
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
37
O
1
2
N
O
R R
Ar, Het
H N
COOP
39
2
42
1
2
O
R R
H
R
N
N
N
COOH
N
Ar, Het
H
O
Ar, Het
COOH
1
2
4
1
R R
H
R
38
N
H N
2
O
COOMe
40
R
1
2
H N
2
R R
+
H N
COOH
2
COOMe
39
Cinnamate esters 39 for inhibitors 2 – 11 were prepared using a palladiumꢀcatalyzed Heck crossꢀ
coupling reaction using commercially available substituted 4ꢀbromo or 4ꢀiodoanilines and
methyl acrylate as shown in Equation 1.
Equation 1. Preparation of substituted 4-aminocinnamate ester building blocks for the
synthesis of inhibitors 2 – 11
R
Br
COOMe
R
COOMe
H N
2
Et N / DMF
3
or
Pd acetate
or
H N
2
R
39
I
Herrmann's catalyst
H N
2
2
2
ACS Paragon Plus Environment

Page 23 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
ꢀAminoꢀethoxycinnamic acid methyl ester 46 was prepared from 2ꢀethoxyꢀ4ꢀnitrobenzoic acid
43 using a sequence involving reductionꢀoxidation to the corresponding aldehyde 44 followed by
HornerꢀWittig to give nitrocinnamate 45 and reduction of the nitro group to the corresponding
aniline using iron as described in scheme 2. Cinnamate ester containing αꢀalkyl substituents (e.g.,
35) were prepared in the same manner using substituted phosphonoacetates.
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
Scheme 2. Preparation of 4-amino-2-ethoxycinnamic acid methyl ester 46
O N
OEt
O
O N
OEt
OH
a, b
2
c
2
44
H
O
43
H N
OEt
2
O N
OEt
2
d
OMe
OMe
O
45
O
46
a
(
a) BH .Me S, THF 0 °C to RT; (b) MnO , CHCl , reflux;
3
2
2
3
(
c) trimethylphosphonoacetate, NaH, THF, RT; (d) Fe
powder, NH Cl, EtOH, reflux.
4
Alkoxycinnamic acid building blocks with variations of the alkoxy moiety for the synthesis of
inhibitors 16 – 21 were prepared from 2ꢀhydroxyꢀ4ꢀnitrobenzoic acid 47 through the sequence
depicted in scheme 3 using a similar reductionꢀoxidation sequence to generate benzaldehyde 48
which served as a key intermediate providing access to chromene analog 49 or protected 2ꢀ
hydroxycinnamate 52 following protection of 48 as the pivaloate ester and HornerꢀWittig
condensation. Phenol 52 was etherified to alkoxy derivatives 53 by direct alkylation using alkyl
halides and Cs CO or via a Mitsunobu protocol. Subsequent reduction with sodium hydrosulfite
2
3
or iron powder provided the corresponding aniline building blocks 54.
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 54
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
Scheme 3. Preparation of chromene 49 and 2-alkoxycinnamate building blocks (inhibitors
a
16 – 21) by alkylation and Mitsunobu etherification protocols
H
COOH
COOMe
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, b
c, d
O
O N
OH
H N
O
2
O N
OH
2
2
4
9
48
47
e
H
O
O
COOMe
COOMe
O N
O
2
O N
2
O N
OH
g
2
f
52
O
O
51
50
h or i
COOMe
COOMe
j
H N
OR
2
O N
OR
2
54
53
a
(a) BH .Me S, THF 0 °C to RT; (b) MnO , CHCl , reflux; (c)methylacrylate, DABCO, 50°C; (d) Fe
3
2
2
3
powder, NH4Cl, EtOH, reflux; (e) PivꢁCl, Et N, CH Cl , RT; (f) trimethylphosphonoacetate, NaH, THF,
3
2
2
RT; (g) conc. H SO ; (h) ROH, DEAD, Ph P, THF, RT; (i) RI or RBr, Cs CO , DMF, RT or 60 °C; (j) Fe,
2
4
3
2
3
NH Cl, EtOHꢁwatre, 80 °C or K CO , NaHSO , EtOHꢁwater, RT.
4
2
3
3
2,6ꢀDimethoxycinnamate analog 14 was prepared from amine 58 prepared as shown in scheme 4
19
from methyl 4ꢀaminoꢀ2,6ꢀdimethoxybenzoate 55,
by condensation with NꢀBocꢀ1ꢀ
aminocyclobutane carboxylic acid to provide protected amino acid derivative 56 followed by the
usual reductionꢀoxidationꢀHorner/Wittig and deprotection to access 58 as the hydrochloride salt
via intermediate benzaldehyde 57.
2
4
ACS Paragon Plus Environment

Page 25 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 4. Preparation of 2,6-dimethoxycinnamate building block for the synthesis of
a
inhibitor 14
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
OH
BocHN
H
H N
OMe
2
b, c
N
OMe
O
BocHN
O
COOMe
a
COOMe
OMe
OMe
56
55
H
N
H
N
OMe
OMe
O
d, e
H N
BocHN
2
O
O
COOMe
57
58 OMe
OMe
a
(a) TBTU, Et N, DMSO, RT; (b) LiAlH , THF, 0 °C to RT; (c) MnO , THF, reflux;
(d) trimethylphosphonoacetate, NaH, THF, RT; (e) 4N HCl, dioxane, RT
3
4
2
Inhibitors 16 – 21 were prepared using an alternative assembly sequence illustrated in Scheme 5.
1
1
6ꢀIndole carboxylic acid 59 was condensed with the ethyl ester of 1ꢀaminocyclobutane
carboxylic acid and the resulting amide product was saponified to provide carboxylic acid 60.
Treatment with acetic anhydride provided azalactone 61 as a stable material that was used in
condensation reactions with a variety of 4ꢀaminocinnamic ester building blocks 39 (e.g. 62) to
provide final inhibitors (e.g. 21) following deprotection of the cinnamate ester.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 54
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 5. Alternative assembly of inhibitors
O
NH3
COOEt
O
O
O
Cl
N
N
N
OH
N
COOH
N
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
H
N
N
N
a, b
c
60
61
59
d, e
O
COOMe
H
N
OMe
N
N
H
H N
OMe
2
O
N
6
2
O
13
OH
a) TBTU, Et N, DMF, RT; (b) 1N NaOH, MeOHꢁTHFꢁwater, RT; (c) Ac O, 100 °C; (d) Aniline 62, toluene, AcOH, 85 °C;
a
(
3
2
(
e)1N NaOH, DMSO.
CONCLUSION
We have described the discovery of a potent HCV NS5B thumb pocket 1 followꢀup compound to
our initial clinical candidate, 1. Cellꢀbased potency in the subgenomic replicon was improved 7
to 8ꢀfold by decreasing the acidity of 1 by ~100x (pKa = 4.5 → 6.5) through electronic effects
mediated by the introduction of alkoxy substituents in conjugation to the acrylic acid moiety.
Subsequent ADMEꢀPK optimization led to the discovery of 27, an analog that was predicted to
be a low clearance compound in man based on its crossꢀspecies PK profile. Based on its
improved potency, crossꢀspecies PK profile and predicted human liver partitioning, doses of 400
mg BID or 230 mg TID of 27 (corresponding to LCIQ = 500) are predicted to provide a stronger
1
7
antiviral response in man than achieved with 1 at 450 mg TID (LCIQ ~ 100). However, the
identification of a genotoxic aniline metabolite released through the action of liver peptidases
2
6
ACS Paragon Plus Environment

Page 27 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and/or chemical hydrolysis of the anilide bond in biological fluids led to discontinuation of the
development of 27.
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
EXPERIMENTAL SECTION
General experimental. All commercially obtained solvents and reagents were used as received
without further purification. All reactions were carried out under an atmosphere of argon.
Temperatures are given in degrees Celsius. Solution percentages express a weight to volume
relationship, and solution ratios express a volume to volume relationship, unless stated
1
otherwise. NMR spectra were recorded on a Bruker AVANCEII (400 MHz for H NMR)
spectrometer and were referenced to either DMSOꢀd (2.50 ppm) or CDCl (7.27 ppm). Data is
6
3
reported as follows: chemical shift (ppm), multiplicity (s = singlet, d = doublet, t = triplet, br =
broad, m = multiplet), coupling constant (J ꢀ reported to the nearest 0.5 Hz), and integration.
Low resolution mass spectra were obtained on a Micromass Platform LCZ model ZMD 4000 in
electrospray mode. Highꢀresolution mass spectra (HRMS) were obtained on a Bruker
micrOTOFꢀQ II in electrospray positive (ES+) ionization mode. Purification of crude material
was performed either by flash column chromatography or by using a CombiFlash Companion
using RediSep Silica or SilicaSep columns according to preꢀprogrammed gradient and flow rate
separation conditions in Hexane / EtOAc or DCM / MeOH. The final compounds were purified
by preparative HPLC on a Waters 2767 Sample Manager with Pumps 2525, Column Fluidics
Organizer (CFO), PDA Detector 2996 and MassLynx 4.1 using either a Whatman Partisil 10ꢀ
ODSꢀ3 column , 2.2 X 50 cm or a YMC CombiꢀPrep ODSꢀAQ column, 50 x 20 mm ID, S – 5
µm, 120 Å, and a linear gradient program from 2 to 100% AcCN /water (0.06% TFA). Fractions
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 54
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
were analyzed by analytical HPLC, and the pure fractions were combined, concentrated, frozen
and lyophilized to yield the desired compound as a neutral entity or the trifluoroacetate salt for
basic analogues. Inhibitor HPLC purity was measured by using a Waters Alliance 2695
Separation Module with a Waters TUV 2487 UV detector. The column was a Combiscreen
ODSꢀAQ, 5 µm, 4.6 x 50 mm, linear gradient from 5% to 100% ACN/H2O + 0.06% TFA in
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
10.5 minutes, detection at 220 nm. All final inhibitors had HPLC homogeneity ≥ 95% unless
noted otherwise (see supporting information).
2-Iodo-5-chloropyrimidine. 2ꢀHydroxypyrimidine hydrochloride (100 g, 0.754 mole) was
dissolved in water (180 mL) and conc. H SO (42 mL, 0.788 mole) was added dropwise with
2
4
vigorous stirring. After stirring for an additional 30 min, water was removed under reduced
pressure at 70°C and the orange residue dried under high vacuum (146 g). The residue was
transferred into a 4 L flask and water (500 mL) was added. A suspension of Ba(OH) (129 g,
2
0.752 mole) in water (1200 mL) was added and the cloudy suspension stirred for 30 min. The
TM
mixture was filtered through Celite and the water removed under reduced pressure to provide
2ꢀhydroxypyrimidine free base as a bright yellow solid (66.4 g).
2ꢀHydroxypyrimidine free base (42 g, 0.44 mole) was added to AcOH (500 mL) and the mixture
heated to 120°C. NꢀChlorosuccinimide (67 g, 0.5 mole, 1.15 equivalent) was added cautiously
(15 min) in small amounts to the hot solution. Stirring was continued for an additional 5 min and
the reaction mixture cooled to RT. The material was concentrated under reduced pressure and the
residue was stirred overnight with CH Cl (200 mL). The suspended solid was removed by
2
2
2
8
ACS Paragon Plus Environment

Page 29 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
filtration and the filtrate evaporated under reduced pressure to give 2ꢀhydroxyꢀ5ꢀ
chloropyrimidine as a beige solid (17.3 g).
The 5ꢀchloroꢀ2ꢀhydroxypyrimidine from above (8.0 g, 0.06 mole) was placed in a dry 500 mL
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
flask under an Ar atmosphere, and POCl (79.4 mL) was added followed by N,Nꢀdimethylaniline
3
(2.6 g). The mixture was heated to 120°C and stirred for 1 h. The dark brown mixture was
concentrated under reduced pressure at 50°C. The residue was quenched carefully with ice water
and the precipitated material was extracted with pentane (3 X 200 mL). The extract was washed
with water and aqueous NaHCO solution, and dried (Na SO ). Volatiles were removed under
3
2
4
reduced pressure with no external heating to prevent sublimation of the volatile product, to
provide 2,5ꢀdichloropyrimidine as a white solid (6 g).
A flask was charged with 57% HI (48 mL) and cooled to 0°C in an iceꢀsalt mixture. The crude
2,5ꢀdichloropyrimidine (6 g) was added and the mixture stirred for 4 h. The yellow suspension
was treated carefully with K CO (32 g) in water (60 mL) and the pale yellow solid was
2
3
collected by filtration. The solid was washed with water and dried to give the title compound (8
o
1
13
g): mp 71ꢀ73 C; H NMR (CDCl ) δ 8.47 (s, 2H); C NMR (101 MHz, DMSOꢀd ) δ 157.1,
3
6
132.2, 125.1.
2
-(5-Chloro-pyrimidin-2-yl)-3-cyclopentyl-1-methyl-1H-indole-6-carboxylic acid 38
(Het = 5-chloro-2-pyrimidinyl). 2ꢀBromoindole 37 was converted to the corresponding
1
1
tributylstannane derivative as previously described. The stannane (37.50 g, 69 mmol, 1 equiv)
was charged in an ovenꢀdried 2 L threeꢀnecked flask equipped with a mechanical stirrer and
1
dissolved in dry DMF (300 mL). 5ꢀChloroꢀ2ꢀiodopyrimidine (91% content by H NMR, 20.0 g,
2
9
ACS Paragon Plus Environment