Discovery of a potent respiratory syncytial virus RNA polymerase inhibitor

Article abstract of DOI:10.1016/j.bmcl.2013.10.018

Targeting viral polymerases has been a proven and attractive strategy for antiviral drug discovery. Herein we describe our effort in improving the antiviral activity and physical properties of a series of benzothienoazepine compounds as respiratory syncyt

Full text of DOI:10.1016/j.bmcl.2013.10.018

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6789–6793
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a potent respiratory syncytial virus RNA polymerase
inhibitor
Hui Xiong, Melinda Foulk, Lisa Aschenbrenner, Jun Fan, Choi-Lai Tiong-Yip, Kenneth D. Johnson,
Demetri Moustakas, Paul R. Fleming, Dean G. Brown, Minli Zhang, Douglas Ferguson, Dedong Wu, Qin Yu
⇑
Infection Innovative Medicine, AstraZeneca Pharmaceuticals, 35 Gatehouse Lane, Waltham, MA 02451, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Targeting viral polymerases has been a proven and attractive strategy for antiviral drug discovery. Herein
we describe our effort in improving the antiviral activity and physical properties of a series of ben-
zothienoazepine compounds as respiratory syncytial virus (RSV) RNA polymerase inhibitors. The antiviral
activity and spectrum of this class was significantly improved by exploring the amino substitution of the
pyridine ring, resulting in the discovery of the most potent RSV A polymerase inhibitors reported to date.
Ó 2013 Published by Elsevier Ltd.
Received 30 July 2013
Revised 6 October 2013
Accepted 8 October 2013
Available online 16 October 2013
Keywords:
Respiratory syncytial virus
Antiviral
Polymerase inhibitor
O
Respiratory syncytial virus (RSV) is the leading cause of infant
hospitalization and acute lower respiratory tract infection world-
wide.1 Despite decades of research and development, effective
therapy options for RSV infection remain limited.2 Palivizumab,3
a monoclonal antibody targeting the RSV fusion (F) protein, signif-
icantly reduced RSV-related hospitalizations among high-risk in-
fants when used as prophylaxis but was not an effective
treatment. Ribavirin, the only approved small molecule therapy
against RSV, has restricted utility due to its controversial efficacy
and toxicity.4 Several potent, small-molecule fusion inhibitors have
been discontinued at preclinical or Phase 1–2 stages,5 with the
exception of inhalable MDT-6376 and GS-58067 advancing to Phase
2 clinical trials. A siRNA agent (ALN-RSV01)8 was reported to have
missed its primary endpoint in a Phase 2b clinical trial. The devel-
opment of RSV604,9 a novel benzodiazepine targeting viral N-pro-
tein, has been stopped in Phase 2 trials.
N
N
O
N
N
NH2
N
N
N
O
HO
N
OH
HO
OH
Ribavirin
BMS-433771
H
N
O
S
H
O
N
N
NH
O
NH
F
N
O
O
R
In 2005, it was reported that YM-53403, identified from random
screening, potently inhibited the replication of RSV in plaque
reduction assays10 (Fig. 1). Resistance studies showed that it
targets the viral RNA polymerase L protein, which catalyzes the
transcription and replication of the viral genome. A recent patent
disclosure11 revealed that modification of either the biphenyl or
the cyclopropyl region of YM-53403 was tolerated for antiviral
activity. We confirmed the antiviral activity of key reference com-
pound 1, and uncovered the intrinsic lack of RSV B activity and
suboptimal drug-like properties (e.g., low aqueous solubility and
N
H
X
RSV-604
YM-53403
(X = C, R = Ph)
1
(X = N, R = pyrrolidinyl)
Figure 1. Chemical structure of selective examples of RSV inhibitors.
low metabolic stability, presumably due to its high lipophilicity
and high molecular weight.).
In this letter, we report the improvement of antiviral potency,
spectrum coverage, as well as the drug-like properties of this
chemical series.
⇑
Corresponding author.
E-mail address: qin.yu@astrazeneca.com (Q. Yu).
0960-894X/$ - see front matter Ó 2013 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.bmcl.2013.10.018

6790
H. Xiong et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6789–6793
H
N
O
H
N
O
S
O
OH
S
S
(a) (b) (c)
(d) (e)
N
N
O
R
R
O
N
H
O
N
N
H
Ar
NH2
3
4
2
Scheme 1. Synthesis of benzo[b]thieno[2,3-d]azepine compounds. Reagents and conditions: (a) HATU, cyclopropylamine, DIPEA, DMF; (b) 4-nitro-benzoyl chloride, DIPEA,
pyridine; (c) Pd/C, H2, 1N HCl, EtOH; (d) 2-Halogen-ArCOCl, DIPEA, pyridine; (e) amine, DIPEA, NMP, reflux.
X
H
N
O
B(OH)2
Y
Y
O
O
S
S
NH
S
Br
S
X
Boc
(f)
(a) (b) (c)
TBSO
N
N
H
H
Y
NH
(d) (e)
Boc
HO
X
N
OH
Boc
5
7
6
8
Scheme 2. Synthesis of pyrido[2,3-b]thieno[2,3-d]azepines. Reagents and conditions: (a) TBSCl, imidazole, DMF; (b) NBS, AcOH/CHCl3; (c) LDA, THF, À78 °C; CO2 (g); HATU,
cyclopropylamine, DIPEA, DMF; (d) Pd-118, K2CO3, EtOH/H2O, reflux; (e) TBAF, THF, 0 °C; (f) DIAD, PPh3, toluene, reflux.
As shown in Scheme 1, the synthesis of the benzothienoazepine
series started from commercially available benzothienoazepine
carboxylic acid 2. Installation of the cyclopropyl amide, acylation
with 4-nitro-benzoyl chloride, and catalytic hydrogenation of the
nitro group provided aniline 3, which underwent amide coupling
and subsequent aromatic substitution with a variety of secondary
amines to afford the final compounds.
To modify the tricyclic core, we employed an intramolecular
Mitsunobu cyclization strategy.12 Starting from commercially
available 2-(thiophen-3-yl) ethanol 5, silyl ether formation, bro-
mination, and 2-step introduction of the carboxyamide, provided
the fully functionalized thiophene 6 (Scheme 2). Suzuki coupling
between 6 and appropriate boronic acids formed the desired biaryl
C–C bond. Subsequent removal of the TBS ether gave the Boc pro-
tected amino alcohol 7, which was treated with DIAD and PPh3 in
hot toluene to provide the azepine ring. Replacement of the thio-
phene with either pyrazole or imidazole was accomplished via a
similar synthetic strategy.
The RSV subgenomic replicon assay was used as a primary assay
to evaluate antiviral activity of the compounds synthesized in this
study. The RSV replicon cell line contains a stable and self-
replicating RSV A2 subgenomic replicon encoding the viral pro-
teins essential for RSV replication and the Renilla luciferase
reporter.13 Antiviral activity of each compound was measured by
inhibition of the Renilla luciferase activity within the replicon-
bearing cells in a robust two-day assay.14
DMPK characteristics (permeability, metabolic clearance) of the
compounds were also measured. Since an RSV L polymerase
Table 1
Antiviral activity, cytotoxicity and solubility for 1, 9a–9f
H
N
O
S
N
O
N
O
Linker
N
N
H
Compd
Linker
Replicon
EC50/CC50
M)
ELISA
EC50 (lM)
A2/B-WST
Sol. (lM)
(
l
1
0.31/>50
3.7/>50
0.15/>50
<1
3
N
9a
2.4/>50
>50/>50
N
9b
>50/>50
3
We also employed a live virus assay in HEp-2 cells using RSV A2
and B Washington (WST) strains to confirm and investigate com-
pound activity against RSV A and B subtypes. This assay uses an
ELISA method to measure viral protein expression level after 3 days
of infection and treatment with experimental compound.
Compound activities in RSV replicon and live virus ELISA assays
were presented as 50% effective concentrations (EC50). Compound
cytotoxicity was determined similarly as 50% cytotoxicity concen-
tration (CC50). Key physical properties (solubility, lipophilicity) and
9c
>50/>50
14/>50
nda
nda
11
26
9d
N
9e
9f
>50/>50
16/>50
nda
nda
2
3
a
nd = not determined for compounds inactive in replicon assay.

H. Xiong et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6789–6793
6791
Table 2
Table 3
Antiviral activity, cytotoxicity and solubility for 1, 10a–10d
Antiviral activity, cytotoxicity and solubility for 1, 11a–11j
H
O
H
N
O
N
S
Ar1
X
Y
N
N
O
O
R1
O
N
O
N
H
N
N
N
H
Compd R1
Replicon ELISA
EC50 EC50
HEp-
2
M) CC50
Sol. (lM)
Compd
X/Y
Ar1
Replicon
EC50/CC50
ELISA
EC50 (lM)
A2/B-WST
Sol. (lM)
(
lM)
(
l
(
lM)
A2/B-
WST
(lM)
S
S
S
1
10a
10b
10c
10d
C/C
N/C
C/N
C/C
C/C
0.31/>50
2.2/>50
8.6/>50
9/>50
0.15/>50
<1
1
1
N
N
0.31
0.63
0.15/>50
0.23/17
>50
>50
<1
36
11a
O
1.1/>50
7.8/>50
20/50
HO
11b
13
12
8.0/>50
>50
40
N
N
N
10
11c
11d
S
O
9.6/>50
2.2/>50
>50
>50
26
4
O
O
N
S
3.6
N
32
121
OH
N
N
11e
11f
11g
>50
> 50/>50
0.01/1.0
34/17
>50
>50
34
>100
2
N
N
NMe2
O
0.01
N
22/>50
50/50
N
N
8.5
9
55
O
11h
5.3/45
39
4
N
N
N
N
O
O
structure and robust biochemical assay were unavailable, we relied
on RSV replicon and live-virus assay readouts to guide the chemis-
try campaign.
11i
11j
0.04
0.03/2.2
0.01/0.5
>50
>50
<1
<1
<0.007
We first explored stepwise truncation of various regions of
compound 1, and found that these changes generally led to the loss
of antiviral activity (details not shown). We next attempted replac-
ing the phenyl linker (Table 1), and found that the pyridyl linker in
compound 9a led to ꢀ10-fold loss of antiviral activity in both
replicon and ELISA assays. The isomeric pyridyl compound 9b led
to even greater loss of antiviral activity. Insertion of saturated link-
ers (compounds 9c–9f) resulted in large losses of antiviral activity,
while aqueous solubility was improved (compounds 9c, 9d).
As shown in Table 2, several heterocyclic replacements of the
tricyclic benzothienoazepine core were designed to lower the
lipophilicity. The pyridyl compound 10a showed ꢀ7-fold loss of
antiviral activity than 1, and the isomeric pyridyl compound 10b
was even less active in both assays. Replacement of the thiophene
ring with either pyrazole (compound 10c) or imidazole (compound
10d) gave promising improvements in aqueous solubility, but at
the cost of losing antiviral activity.
Various secondary amines substituted on the pyridine ring were
also examined (Table 3). Compared to pyrrolidine in compound 1,
morpholine (11a) led to improved aqueous solubility, comparable
antiviral activity in replicon assay, and weak activity towards RSV
B in the ELISA assay. Several other attempts at improving physical
properties (11b–11e) led to loss of antiviral activity. Interestingly,
introduction of an oxetane spiro-fused piperidine (2-oxa-7-azaspi-
ro[3.5]nonane, 11f) gave significantly improved antiviral activity of
10 nM against RSV A2, and 1 lM against RSV B-WST, whereas com-
pound 11g with 2-methyl-2,7-diazaspiro[3.5]nonane showed
much weaker antiviral activity, which was thought to be due to
the protonation of the basic tertiary amine in 11g under physiolog-
ical pH environment. However, removal of the basic azetidine
nitrogen through the formation of the corresponding acetamide
(11h) did not reverse the activity loss. Incorporation of either
3-oxa-9-azaspiro[5.5]undecane (11i) or 7-oxa-2-azaspiro[3.5]non-
ane (11j) led to good antiviral activity against both RSV A2 and B
strains, with compound 11j as the most potent RSV A polymerase
inhibitor reported to date.
Further modification of the aromatic ring bearing the amine
substitution was also explored. As shown in Table 4, among all four
possible pyridyl isomers and the phenyl ring, the pyridine ring in
compound 11f was superior. Furthermore, a methoxyl substitution
on the pyridine ring (11o) was also tolerated, indicating a possible
position for further modification.
Considering the physical properties of compound 11f (molecu-
lar weight of 638, low aqueous solubility of 2 lM and low meta-
bolic stability), and the site of RSV infection in respiratory tract,
we evaluated the lung retention and distribution properties of

6792
H. Xiong et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6789–6793
Table 4
group, plasma and lung tissues were collected and analyzed
(Fig. 2).
Antiviral activity, cytotoxicity and solubility for 11f, 11k–11o
H
O
Compound 11f was absorbed rapidly through lungs into blood
circulation in solution form (absorption t1/2 <1 min). Only 2.5% of
N
the administered 10
at 5 min post dosing. When the dose was normalized, the 10
l
g/kg IT dose remained in the lung and BALF
g/kg
S
l
IT plasma and the 3 mg/kg IV plasma profiles appeared to be iden-
tical, suggesting rapid and complete absorption from the IT dose
site to the systemic circulation. The nano-suspension formulation
did not show significant delay in absorption from lung. Only 4.5%
of the administered dose remained in lung and BALF at 5 min post
dosing.
These experiments suggested that compound 11f was not re-
tained to a significant extent in the lung and BALF, and was instead
rapidly absorbed into the systemic circulation following IT dosing.
Further improvement of the lung retention properties is needed to
obtain an inhalable RSV antiviral from this series.
O
N
O
O
N
N
H
Ar2
Compd
Ar2
Replicon
EC50
ELISA
EC50 (lM)
A2/B-WST
HEp-2
CC50
(lM)
Sol. (lM)
(l
M)
NR2
We have identified novel modifications that delivered potent
and broad-spectrum RSV L polymerase inhibitors without cytotox-
icity. This provides new knowledge that may pave the way towards
effective RSV therapeutics and new tool compounds to interrogate
RSV L protein function.
11f
0.01
0.16
0.01/1.0
>50
2
N
NR2
11k
<0.21/5.1
0.68/14
4.8
<1
N
NR2
Acknowledgments
The authors would like to thank Camil Joubran, Nancy DeGrace,
Natascha Bezdenejnih-Snyder, Ziling Lu, Ying Liu and Bridget Rea-
ney for analytical chemistry support; Steven Hsu, Ken Grime, Britt-
Marie Fihn and Anders Lundqvist for in vitro and in vivo DMPK
support; Nancy Ulbrandt and Hong Ji for RSV strains and protocols
support; Ruben Tommasi, Martin Hentemann, Giuseppe Ciaramella
and Scott Butler for scientific discussions.
11l
0.57
34
9
N
NR2
11m
11n
0.35
0.19
0.47/49
43
<1
<1
N
NR2
<0.21/>50
>50
References and notes
NR2
N
1. Hall, C. B.; Weinberg, G. A.; Iwane, M. K.; Blumkin, A. K.; Edwards, K. M.; Staat,
M. A.; Auinger, P.; Griffin, M. R.; Poehling, K. A.; Erdman, D.; Grijalva, C. G.; Zhu,
Y.; Szilagyi, P. N. Eng. J. Med. 2009, 360, 588.
2. (a) Collins, P. L.; Melerob, J. A. Virus Res. 2011, 162, 80; (b) Welliver, R. C. Curr.
Opin. Pharmacol. 2010, 10, 289; (c) Shigeta, S. Exp. Opin. Invest. Drugs 2000, 9,
221; (d) Carter, M.; Cockerill, G. S. Annu. Rep. Med. Chem. 2008, 43, 229.
3. Null, D. et al Pediatrics 1998, 102, 531.
11o
0.02
<0.21/1.8
>50
<1
OMe
4. Rodriguez, W. J.; Hall, C. B.; Welliver, R.; Simoes, E. A.; Ryan, M. E.; Stutman, H.;
Johnson, G.; Van Dyke, R.; Groothuis, J. R.; Arrobio, J.; Schnabel, K., et al J.
Pediatr. 1994, 125, 129.
5. (a) Sun, Z.; Pan, Y.; Jiang, S.; Lu, L. Viruses 2013, 5, 211; (b) Meanwell, N. A.;
Langley, D. R. Top. Med. Chem. 2012, 8, 167; (c) Meanwell, N. A.; Cianci, C. W.;
Krystal, M. R. In Antiviral Drugs: From Basic Discovery Through Clinical Trials;
Kazmierski, W. M., Ed.; Wiley: Hoboken, 2011; p 353; (d) Bonfanti, J.-F.; Ispas,
G.; Van, V. F.; Olszewska, W.; Gevers, T.; Roymans, D. In Antiviral Drugs: From
Basic Discovery Through Clinical Trials; Kazmierski, W. M., Ed.; Wiley: Hoboken,
2011; p 341.
6. (a) Douglas, J. L.; Panis, M. L.; Ho, E.; Lin, K. Y.; Krawczyk, S. H.; Grant, D. M.; Cai,
R.; Swaminathan, S.; Cihlar, T. J. Virol. 2003, 77, 5054; (b) Press release,
MicroDose Therapeutx.
7. http://www.clinicaltrials.gov/ct2/show/NCT01756482.
8. (a) DeVincenzo, J.; Lambkin-Williams, R.; Wilkinson, T.; Cehelsky, J.; Nochur, S.;
Walsh, E.; Meyers, R.; Gollob, J.; Vaishnaw, A. Proc. Natl. Acad. Sci. U.S.A. 2010,
107, 8800; (b) Press release, Alnylam Pharma.
9. (a) Henderson, E. A.; Alber, D. G.; Baxter, R. C.; Bithell, S. K.; Budworth, J.;
Carter, M. C.; Chubb, A.; Cockerill, G. S.; Dowdell, V. C. L.; Fraser, I. J.; Harris, R.
A.; Keegan, S. J.; Kelsey, R. D.; Lumley, J. A.; Stables, J. N.; Weerasekera, N.;
Wilson, L. J.; Powell, K. L. J. Med. Chem. 2007, 50, 1685; (b) Chapman, J.;
Abbott, E.; Alber, D. G.; Baxter, R. C.; Bithell, S. K.; Henderson, E. A.; Carter, M.
C.; Chambers, P.; Chubb, A.; Cockerill, G. S.; Collins, P. L.; Dowdell, V. C.;
Keegan, S. J.; Kelsey, R. D.; Lockyer, M. J.; Luongo, C.; Najarro, P.; Pickles, R. J.;
Simmonds, M.; Taylor, D.; Tyms, S.; Wilson, L. J.; Powell, K. L. Antimicrob.
Agents Chemother. 2007, 51, 3346; (c) Chapman, J.; Cockerill, G. S. In Antiviral
Drugs: From Basic Discovery Through Clinical Trials; Kazmierski, W. M., Ed.;
Wiley: Hoboken, 2011; p 367.
10. Sudo, K.; Miyazaki, Y.; Kojima, N.; Kobayashi, M.; Suzuki, H.; Shintani, M.;
Shimizu, Y. Antiviral Res. 2005, 65, 125.
11. Mackman, R. L.; Sperandio, D.; Yang, H. WO 2011/005842 A1, 2011.
12. Ohtani, T.; Kawano, Y.; Kitano, K.; Matsubara, J.; Komatsu, M.; Uchida, M.;
Tabusa, F.; Nagao, Y. Heterocycles 2005, 66, 481.
Figure 2. Plasma and lung tissue exposure of 11f in rat.
11f via intratracheal (IT) and intravenous (IV) dosing.15 Compound
11f was dosed to rats as a solution at 10 g/kg, or a nanoparticle
suspension at 100 and 3000 g/kg doses. Following IT dosing, plas-
l
l
ma, lung tissue and bronchoalveolar lavage fluid (BALF) were col-
lected and drug concentrations were determined. In the IV dose

H. Xiong et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6789–6793
6793
13. Malykhina, O.; Yednak, M. A.; Collins, P. L.; Olivo, P. D.; Peeples, M. E. J. Virol.
2011, 85, 4792.
of antibody-bound RSV antigen was quantified by adding peroxidase substrate
and measuring absorbance at 450 nm. Compound EC50 values were calculated
using the percent activity of the maximal signal (Replicon-only or virus-only
infection) at multiple compound concentrations using the XLFit model.
Compound CC50 were determined in the same cell lines using a cell viability
assay kit.
14. Tiong-Yip, C.; Plant, H.; Sharpe, P., Fan, J.; Rich, K; Gorseth, E.; Yu, Q. (in press).
The RSV replicon cell line (BHK21-based APC126) was obtained from Dr. Mark
Peeples (Nationwide Children’s Hospital) via Apath, LLC. The RSV ELISA used a
primary monoclonal antibody recognizing the RSV F-protein and a goat anti-
mouse secondary antibody conjugated to horseradish peroxidase. The presence
15. Cooper, A. E.; Ferguson, D. D.; Grime, K. K. Curr. Drug Metab. 2012, 13, 457.