Discovery of a novel, first-in-class, orally bioavailable azaindole inhibitor (VX-787) of influenza PB2

Article abstract of DOI:10.1021/jm5007275

In our effort to develop agents for the treatment of influenza, a phenotypic screening approach utilizing a cell protection assay identified a series of azaindole based inhibitors of the cap-snatching function of the PB2 subunit of the influenza A viral polymerase complex. Using a bDNA viral replication assay (Wagaman, P. C., Leong, M. A., and Simmen, K. A. Development of a novel influenza A antiviral assay. J. Virol. Methods 2002, 105, 105-114) in cells as a direct measure of antiviral activity, we discovered a set of cyclohexyl carboxylic acid analogues, highlighted by VX-787 (2). Compound 2 shows strong potency versus multiple influenza A strains, including pandemic 2009 H1N1 and avian H5N1 flu strains, and shows an efficacy profile in a mouse influenza model even when treatment was administered 48 h after infection. Compound 2 represents a first-in-class, orally bioavailable, novel compound that offers potential for the treatment of both pandemic and seasonal influenza and has a distinct advantage over the current standard of care treatments including potency, efficacy, and extended treatment window.

Full text of DOI:10.1021/jm5007275

Article
pubs.acs.org/jmc
Discovery of a Novel, First-in-Class, Orally Bioavailable Azaindole
Inhibitor (VX-787) of Influenza PB2
Michael P. Clark,*,† Mark W. Ledeboer,† Ioana Davies,† Randal A. Byrn,† Steven M. Jones,†
Emanuele Perola,† Alice Tsai,† Marc Jacobs,† Kwame Nti-Addae,† Upul K. Bandarage,† Michael J. Boyd,†
Randy S. Bethiel,† John J. Court,† Hongbo Deng,† John P. Duffy,† Warren A. Dorsch,† Luc J. Farmer,‡
Huai Gao,† Wenxin Gu,† Katrina Jackson,† Dylan H. Jacobs,† Joseph M. Kennedy,† Brian Ledford,†
Jianglin Liang,† Francois Maltais,† Mark Murcko,† Tiansheng Wang,† M. Woods Wannamaker,†
̧
Hamilton B. Bennett,† Joshua R. Leeman,† Colleen McNeil,† William P. Taylor,† Christine Memmott,†
Min Jiang,† Rene Rijnbrand,† Christopher Bral,§ Ursula Germann,† Azin Nezami,† Yuegang Zhang,†
Francesco G. Salituro,∥ Youssef L. Bennani,‡ and Paul S. Charifson†
†Vertex Pharmaceuticals Inc., 50 Northern Ave, Boston, Massachusetts 02210, United States
‡Vertex Pharmaceuticals (Canada) Inc., 275 Armand-Frappier, Laval, Quebec H7V 4A7, Canada
§Arrowhead Research Corporation, 465 Science Drive, Suite C, Madison, Wisconsin 53711, United States
∥Sage Therapeutics, 215 First Street, Cambridge, Massachusetts 02141, United States
S
* Supporting Information
ABSTRACT: In our effort to develop agents for the treatment
of influenza, a phenotypic screening approach utilizing a cell
protection assay identified a series of azaindole based inhibitors
of the cap-snatching function of the PB2 subunit of the influenza
A viral polymerase complex. Using a bDNA viral replication
assay (Wagaman, P. C.; Leong, M. A.; Simmen, K. A.
Development of a novel influenza A antiviral assay. J. Virol.
Methods 2002, 105, 105−114) in cells as a direct measure of
antiviral activity, we discovered a set of cyclohexyl carboxylic
acid analogues, highlighted by VX-787 (2). Compound 2 shows
strong potency versus multiple influenza A strains, including
pandemic 2009 H1N1 and avian H5N1 flu strains, and shows an efficacy profile in a mouse influenza model even when treatment
was administered 48 h after infection. Compound 2 represents a first-in-class, orally bioavailable, novel compound that offers
potential for the treatment of both pandemic and seasonal influenza and has a distinct advantage over the current standard of
care treatments including potency, efficacy, and extended treatment window.
INTRODUCTION
■
The orthomyxoviridae family of RNA viruses includes influenza
A, B, and C. Influenza is responsible for more than 200 000
hospitalizations2 on a yearly basis and an annual mortality rate
that varies from 3000 to 49 000 deaths in the United States
during nonpandemic seasons.3 Both swine and avian hosts can
serve as reservoirs of influenza A. Occasionally, transmission of
Figure 1. Inhibitors of influenza virus: NA inhibitor oseltamivir (1)
and 2.
a new flu variant from other species to humans can lead to a
global epidemic, such as the 2009 H1N1 swine flu pandemic.
The emergence of the highly pathogenic H5N1 and H7N9
avian influenza strains emphasizes the continued potential
global risk of the jumping of animal influenza strains to the
human population.4 The current standard of care (SOC)
antivirals for influenza cases in the United States are the
neuraminidase inhibitors (NAIs) oseltamivir (1, Figure 1) and
zanamivir. While these agents can be effective against a variety
of type A and B influenza viruses, they suffer from two main
limitations. First, the neuraminidase inhibitors have only a
moderate impact on the severity of symptoms as well as
duration of sickness and they must be administered within 24−
48 h of infection.5 Second, resistance to this class of antivirals
Received: May 9, 2014
Published: July 14, 2014
© 2014 American Chemical Society
6668
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Figure 2. Azaindole screening hits.
Figure 3. X-ray structure of cap-binding domain (contact residues) of PB2 containing 7-methyl GTP (left) or 3 (right). Accession codes are the
following: 4NCE for 7-methyl GTP and 4NCM for 3.
has generated significant concern,6 especially with the report
that the H5N1 influenza virus has shown resistance to
oseltamivir,7 reinforcing the critical need for new anti-influenza
therapeutics with novel mechanisms of action.
the host pre-mRNA. Once bound to PB2, the PA endonuclease
subunit cleaves the host RNA strand, leaving a 10−13
nucleotide primer. The PB1 subunit contains the conserved
polymerase domain and utilizes the primer for RNA elongation.
Soon after identification of PB2 as the target for these azaindole
screening hits, the X-ray crystal structures of m7GTP and VRT-
0761704 (3) bound to the PB2 cap-binding domain were
determined (Figure 3). Though belonging to a different space
group (see Supporting Information), the structure of the PB2
cap binding fragment was consistent with that previously
described for complexes with m7GTP.9 This new structure
revealed that 3 occupies the m7GTP binding site and produces
similar interactions with the protein as the m7GTP guanine
base (Figure 3). Both m7GTP and 3 form hydrogen bonds to
the side chains of Glu361 and Lys376. The azaindole of 3 is
sandwiched between the aromatic side chains of His357 and
Phe404, while the pyrimidine π-stacks against Phe323. The
dimethylalanyl portion of 3 is located near the region occupied
by the m7GTP ribose and α-phosphate. We observed no direct
polar interactions of 3 with the positively charged residues
(Lys-339, Arg-355, and His-357) located in the m7GTP
phosphate binding region, which became an area for
exploration discussed later in this paper.
RESULTS AND DISCUSSION
■
The goal of our program was to develop a novel dual route (iv
and oral) inhibitor for the treatment of influenza that
specifically addresses the shortcomings of the current standard
of care. To this end, we started our program by utilizing a
phenotypic cell protection (CPE) assay to identify screening
hits that could overcome the viral challenge limitation of the
neuraminidase inhibitors.8 An initial screen of 1000 compounds
(selected based on their cellular activity in various assays and
the lack of cytotoxicity) identified numerous compounds with
measurable antiviral activity. In particular, a series of 3-
pyrimidineazaindoles with significant cell protective activity at
10 μM concentration emerged from the screen. Subsequent
rounds of screening identified multiple subclasses that were
selected for exploratory chemistry. Three subclasses of interest
can be exemplified by the hits shown in Figure 2.
Subsequently, the molecular target of this compound class
was determined to be the PB2 subunit of the influenza viral
polymerase.9 The viral polymerase contains three subunits,
PB1, PB2, and PA, that are responsible for replication and
transcription of the eight separate segments of the viral RNA
genome in the nuclei of infected cells.9,10 The heterotrimeric
viral polymerase synthesizes viral mRNAs via a unique “cap-
snatching” mechanism where the virus utilizes host pre-mRNA
as a primer for transcription.11 The PB2 subunit contains a cap
binding domain for 7-methyl GTP (m7GTP) on the 5′-end of
In the early stages of our project we established that optimal
and consistent cellular potency was achieved with chloro
substitution at the 5-position of the azaindole ring system (Cl >
F > H; see discussion later in text). On the basis of our analysis
of the crystal structure and additional azaindole screening hits,
we focused our optimization efforts on the amino substituent at
the 4-position of the pyrimidine ring system. Initial analogues,
such as cyclohexylamine, 4 (Table 1), maintained potency
relative to 3. Additionally, we observed a preference for six-
6669
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
groups with Asn-429. We also observed a stereochemical
preference: (S)-stereochemistry for the aminopiperidines and
(R)-stereochemistry for the aminomethylpiperidines. These
analogues, typically capped with an amide, urea, or sulfonamide,
consistently showed high clearance in rat PK studies
(compounds 8 and 9, Table 6). In an effort to improve
potency (through potential interactions with the His-357, Lys-
339, and Arg-355 residues in the GTP phosphate binding
region), increase solubility, and reduce clearance thought to be
associated with the amido-capped piperidines, we designed a set
of N-alkylpiperidines, highlighted by piperidines 10 and 11.
The N-alkylacetamide analogue 11 showed the strongest
cellular (bDNA) and enzyme binding potency of any
compound in this series; however, 11 also showed high
clearance in rats (182 mL min−1 kg−1). The diol 10 was the first
compound in this series to display reasonable potency in the
bDNA assay and show an acceptable improvement in clearance
and half-life (rat) as well as reasonable oral exposure in mouse
(Table 6). Our initial pharmacology studies with 10 in a mouse
pretreatment influenza model showed efficacy at high doses
(data not shown). These observations were exciting but showed
that further optimization of potency and pharmacokinetics was
still required.
Table 1. Optimization of in Vitro Potency of Cycloalkyl and
Piperidine Based 5-Chloroazaindole Inhibitors
Synthesis of the azaindolepiperidine diol 10 was accom-
plished in eight steps from the commercially available bromide
12 (Scheme 1). Protection of the azaindole NH functionality as
a tosylate followed by reaction with bis(pinacolato)diboron
afforded the boronate ester 13. Suzuki coupling with 2-chloro-
5-fluoro-4-methylthiopyrimidine, 14, provided the methyl
sulfide 15. Oxidation of the sulfide to the sulfoxide 16 allows
for facile displacement with Boc-protected (S)-3-aminopiper-
idine. Removal of the Boc-protecting group followed by
alkylation of the free piperidine NH with (S)-glycidol affords
the desired piperidine diol 14 after base-catalyzed removal of
the tosyl protecting group.
During the course of our lead optimization efforts, we
evolved from the piperidine-based analogues to 1,2- and 1,3-
diaminocyclohexyl analogues (Table 2). Our goal was to
maintain the putative interactions made by the piperidine acyl/
sulfonamide groups while improving metabolic stability by
moving the nitrogen out of the ring. These analogues showed a
more consistent boost in cellular (bDNA) potency, relative to
the aminopiperidine analogues, as well as improvement in
metabolic stability (data not shown). The preference for the
(S)-stereochemistry adjacent to the pyrimidine ring was
maintained in this subseries (compounds 18 and 19) as
reflected in both the cell protection and PB2 fluorescent
binding assays.14 While the PB2 binding data were useful in
prioritizing each of the subseries of azaindoles that were
explored, the PB2 data did not always reflect the potency trends
observed in the cellular bDNA assays,14 which became the
primary assay for driving the SAR on this project. In general,
the 1,3-diaminocyclohexyl analogues displayed a more robust
SAR tolerance than the 1,2-diaminocyclohexyl analogues,
exemplified by the >50-fold boost in potency in the bDNA
assay (20 vs 21). Therefore, we focused our attention on
derivatization of the 3-amino functionality. The preferred
stereochemical relationship of the 1,3-diamino group was
established as cis (data not shown). Capping of the 3-amino
group with simple amides resulted in a significant boost in
bDNA potency (compounds 21−23), while more solubilizing
amides were less potent (compounds 24 and 25). The most
potent compounds in this series came in the form of tertiary
a
MDCK cells were incubated with test compounds and influenza A
virus (A/PR/8/34 strain) for 72 h, and the concentration of test
compound resulting in 50% cell protection was reported as the EC50.
b
The concentration of test compound resulting in viral RNA levels
equal to that of 10% of the control wells was reported as EC90.
c
Affinity for cap-binding domain of the PB2 subunit as measured in a
fluorescence polarization competition binding assay.
membered cycloalkyl over smaller rings (compound 6). We
then identified additional potent compounds in the bDNA
assay1 having substituted cycloheteroalkylamines on the
pyrimidine ring (Table 1), highlighted by the aminopiperidine
and aminomethylpiperidine subseries (compounds 7−11). Our
hypothesis for the observed improvement in potency was due
to the hydrophobic portion of the piperidine rings more
efficiently occupying the aforementioned hydrophobic pocket
coupled with a potential interaction of the acyl/sulfonamide
6670
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthetic Scheme for 5-Chloroazaindolepyrimidinyl Piperidine Analogue 10
a
Reagents and conditions: (a) NaH, TsCl, DMF, 85%; (b) bis(pinacolato)diboron, KOAc, PdCl2(dppf), dioxane, water, reflux, 50%; (c) 14,
i
Pd(PPh3)4, Na2CO3, DME, water, 94%; (d) mCPBA, CH2Cl2, 96%; (e) (S)-3-amino-1-Boc piperidine, Pr2NEt, DMF, 90 °C, 60%; (f) TFA,
CH2Cl2, 95%; (g) (S)-(−)-glycidol, EtOH, microwave, 140 °C, 58%; (h) NaOMe, MeOH, 45%.
substituted urea analogues, such as compounds 26−28,
compared to the secondary urea 29. Additionally, N-
methylimidazoleamides 30 and 31 showed equivalent potency.
It is worth noting that these more potent ureas (27 and 28)
and amides (30 and 31) only showed a small difference in
potency between the 5-chloro and 5-fluoroazaindole sub-
stitutions. Table 3 shows a summary of SAR at the 5-position of
the azaindoles. Generally, small halo- or haloalkyl substituents
at the 5-position (compounds 27, 28, and 33) displayed the
best potency, while the methyl ketone, 34, showed the least
activity in this set. A simple 5-methyl substitution (compound
35) retains potency better than larger alkyl groups, hydrogen or
electron donating groups (compounds 36, 37 and 38) with the
most significant drop-off in potency among these small
substituents occurring with 5-substituted N,N-dimethylamine
(compound 39). A direct comparison of rat iv clearance values
of the four most potent analogues in Table 3 (compounds 27,
28, 33, and 35) showed a clear preference for the 5-fluoro
substitution. The improved clearance values, along with the
observation that the 5-fluoroazaindoles generally showed
improvements in kinase selectivity, aqueous solubility, and
oral exposure in rodents, helped better guide the program, even
though earlier, less potent analogues showed a greater
difference in potency between 5-chloro and 5-fluoro.
The pharmacokinetic profile of 28 showed encouraging oral
exposure in both rat and mouse (Table 6). This compound was
advanced into a mouse influenza model (Figure 4) showing
75% survival in a +48 h delay to treatment model (30 mpk b.i.d.
× 10 days). The corresponding Cmin following the 30 mpk b.i.d.
regimen was 370 μM, approximately 30-fold over the in vitro
EC90. The efficacy observed with 28 was an improvement over
the current SOC, oseltamivir. When dosing oseltamivir at the
clinically relevant human equivalent dose (10 mpk b.i.d.) in the
same +48 h delay-to-treatment model, oseltamivir provided no
survival benefit (single dose PK of oseltamivir carboxylate (10
mpk in healthy Balb/C mouse) AUC = 3.8 μg·h/mL, Cmax =
2520 ng/mL).
A concerted effort was initiated to explore chemical diversity
on the aminocyclohexyl ring system. A breakthrough was
achieved during our exploration of the 1,2-diaminocyclohexane
analogues. We reversed the 2-acetamido functionality of 20 to
give the β-aminocarboxamide 40 (Table 4), which showed
similar potency in the bDNA assay. While moderate improve-
ments in potency were observed by exploring amide
substitutions (data not shown), the turning point in this series
came when we designed the carboxylic acid analogue 41. The
structural information shown in Figure 2 had suggested that if
we could incorporate a negatively charged moiety into the
phosphate-binding region of GTP, this might result in strong
interactions with the corresponding basic residues (His-357,
Lys-339, and Arg-355). Compound 41 proved to be the most
potent ligand we had seen in the PB2 fluorescent polarization
binding assay, exceeding the sensitivity limits of the assay (Kd <
0.002 nM). The compound also exhibited a significant increase
in cellular potency (16-fold in bDNA assay relative to 40).
While this compound also showed exquisite selectivity vs
kinases and receptors, it showed limited oral exposure in rats
(Table 6) as well as an unfavorable efflux ratio (Caco-2 efflux
ratio of 13). In an attempt to increase lipophilicity around the
carboxylic acid functionality, in order to increase cell
penetration and consequently improve cell potency as well as
increase oral exposure, we designed geminal substituted
carboxylic acid analogues 42−44, all of which either maintained
or slightly increased cellular potency. The simple geminal-
methyl analogue 42 showed reduced efflux (Caco-2 ratio of 2),
presumably due to increasing lipophilicity and bulk around the
acid moiety15 and improved oral exposure in rats (Table 6). A
significant breakthrough in this optimization path was the
design of the bridged cyclohexyl analogues. We further
optimized the van der Waals interactions in the hydrophobic
portion in the PB2 binding site and achieved better balance
6671
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Table 2. Optimization of in Vitro Potency of Diamino
Cyclohexyl-Based Inhibitors
Table 3. SAR of 5-Substituted Azaindole Inhibitors
a
b
CPE EC50
(μM)
bDNA EC90
rat iv Cl
compd
R
(μM)
(mL min−1 kg−1
)
27
28
33
34
35
36
37
38
39
Cl
F
0.002
0.004
0.003
>10
0.005
0.012
0.003
>3
59
22
81
CF3
C(O)Me
CH3
0.010
0.025
0.095
0.098
1.260
0.013
0.260
0.170
0.200
>3
78
CH2CH3
OMe
H
NMe2
a
MDCK cells were incubated with test compounds and influenza A
virus (A/PR/8/34 strain) for 72 h, and the concentration of test
compound resulting in 50% cell protection was reported as the EC50.
b
The concentration of test compound resulting in viral RNA levels
equal to that of 10% of the control wells was reported as EC90.
Compound 2, designated VX-787,12 also demonstrates
potent, antiviral activity in vitro (CPE assay) against a broad
range of influenza type A strains,13 including neuraminidase
inhibitor and amantadine resistant isolates and current
pandemic H1N1 and H5N1 strains (Table 5). The activity
appears to be specific for influenza A virus, as neglible activity
against influenza B virus was observed (data not shown).
The X-ray crystal structure of 2 bound to PB2 confirmed that
the azaindole ring system makes the expected interactions with
the protein residues Lys-376 and Glu-361 as well as stacking of
the three aromatic rings of 2 between the side chains of His-
357, Phe-323, and Phe-404 (Figure 5). The carboxylic group of
the ligand makes two water-mediated interactions with the ε
nitrogen of His 357 and with Gln-406 as well as the main chain
carbonyl of Arg-355. Additional interactions with Lys-339 and
Arg-355 in the adjacent region were expected, consistent with
the substantial increase in binding affinity imparted by the
carboxylic group. However, no direct interactions between the
ligand and these basic residues were observed in the structure.
We hypothesized that the low pH and high salt concentration
of the crystallographic buffer (3 M sodium formate, pH 4.5)
may have masked these interactions. While SPR experiments
showed that the binding affinity of the related acid 41 was not
significantly affected by these factors, no confirmatory crystal
structure at low salt and neutral pH could be obtained.
Scheme 2 depicts the synthesis of 2 using a more convergent
process of coupling the azaindole boronate species toward the
end of the sequence. The synthesis of the enantiomerically pure
β-amino ester pyrimidine fragment 55 was accomplished in five
steps, followed by a chromatographic resolution (Scheme 2).
The sequence begins with a Diels−Alder cycloaddition between
maleic anhydride and 1,3-cyclohexadiene. The resulting endo
cycloaddition product 50 was stirred for 4 days under basic
conditions to open the anhydride and epimerize the resulting
ester to afford the desired trans ester carboxylic acid product 51
as a racemic mixture. Curtius rearrangement with subsequent
trapping with benzyl alcohol provided the protected cyclic
amino acid 52. Hydrogenation of 52 yielded the bridged
a
MDCK cells were incubated with test compounds and influenza A
virus (A/PR/8/34 strain) for 72 h, and the concentration of test
compound resulting in 50% cell protection was reported as the EC50.
b
The concentration of test compound resulting in viral RNA levels
equal to that of 10% of the control wells was reported as EC90.
c
Affinity for cap-binding domain of the PB2 subunit as measured in a
fluorescence polarization competition binding assay.
between the lipophilicity of the molecule vs the acidic
functionality. The single carbon bridged analogues 45 and 46
showed a further increase in cellular potency, with a slight
preference for the endo-product 46. The [2,2,2]bicyclooctane
analogues 47 and 2 afforded the most potent compounds in the
program with bDNA EC90 values of <5 nM as well as the added
advantage of lack of endo and exo stereoisomers with the
symmetric [2,2,2] system.
6672
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Figure 4. In vivo activity of 28 and oseltamivir in mouse influenza A model when administered 48 h after infection: survival and body weight curves
of male BALB/C mice (8 mice/group) inoculated with mouse-adapted influenza viruses A/PR/8/34 (5e3 TCID50/mouse) by intranasal instillation.
Table 4. Optimization of in Vitro Potency of Cyclohexyl
Carboxylate Based Inhibitors
Table 5. Compound 2 Maintains Antiviral Potency vs
Adamantine and NAI Resistant Viruses
a
mean CPE EC50 (μM)
oseltamivir
carboxylate
virus
2
amantadine
d
A/Georgia/20/2006
0.0026
>10
>10
>10
b
d
A/Texas/48/2009
0.0028
>10
d
A/Puerto Rico/8/34
A/California/07/2009
A/Mexico/4108/2009
0.000 32
0.0018
>10
>10
>10
>10
0.16
b
b
d
>10
d
0.0027
>10
b
d
A/New York/18/2009
0.000 59
<0.00015
>10
c
e
f
f
A/Viet Nam/1203/2004
ND
ND
a
b
Mean MDCK cell 3-day CPE assay EC50, N = 3 or greater. Influenza
c
A (H1N1) 2009 pandemic strain. Highly pathogenic avian influenza
(H5N1) strain. Resistant, determined via externally validated
phenotypic or genetic analysis. Performed at Southern Research
Institute. ND: not determined.
d
e
f
cyclohexylamino ester 53. After coupling with 2,4-dichloro-5-
fluoropyrimidine, chromatography on a chiral support was
utilized to separate the enantiomers. With the (S,S) bridged
cyclohexylamino-chloropyrimidine 55 in hand, a Suzuki cross-
coupling with the 5-fluoroazaindole boronate ester fragment,
56,12 gives the fully assembled scaffold 57, which upon removal
of the tosylate protecting group and hydrolysis of the ester
provided 2 in eight overall steps.
As mentioned earlier, one of the advantages of the carboxylic
acid analogues is a superior selectivity profile. PB2 inhibitor 2
was tested against a panel of kinases (65) and only showed
measurable inhibition (Ki = 1.6 μM) against a single kinase,
GSK3β (see Supporting Information). Additionally, 2 was
evaluated for interaction with a variety of receptors, ion
channels, and other targets (68 total) and was shown to bind
competitively to only two of these targets at 10 μM: adenosine
A1 (96% inhibition) and A2a (62% inhibition). No inhibition
was observed with 2 against either hERG (manual patch, >30
μM) or a panel of CYPs (>30 μM). The pharmacokinetic
profile of 2 showed good oral exposure in both rat and mouse.
Compound 2 was advanced into a mouse influenza model
where it showed 100% survival in the +48 h delay to treatment
model at 10, 3, or 1 mpk (b.i.d. × 10 days), whereas the SOC
oseltamivir provided no survival benefit in this model at 10 mpk
(Figure 6). Compound 2 also showed a marked improvement
in observed body weight loss (BWL) with a loss of 7.1% at 10
mpk, 10.9% at 3 mpk, and 22.5% at 1 mpk on study day
compared to 28 administered at an even higher dose (30 mpk)
(∼30% BWL). Again, oseltamivir was run in the same +48 h
delayed treatment model and showed 0% survival and greater
a
MDCK cells were incubated with test compounds and influenza A
virus (A/PR/8/34 strain) for 72 h, and the concentration of test
compound resulting in 50% cell protection was reported as the EC50.
b
The concentration of test compound resulting in viral RNA levels
equal to that of 10% of the control wells was reported as EC90.
c
Affinity for cap-binding domain of the PB2 subunit as measured in a
fluorescence polarization competition binding assay.
6673
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Figure 5. X-ray structure of 2 bound to PB2. Accession code is 4P1U.
a
Scheme 2. Synthetic Scheme for 2
Table 6. Mouse and Rat PK Parameters for Selected PB2
Influenza Inhibitors
rat (3 mg/kg)
mouse (30 mg/kg)
d
b
c
b
c
compd
iv Cl, T1/2(h)
po AUC, Cmax, % F
AUC
Cmax
8
152, 0.4
88, 0.7
56, 2.3
22, 1.7
22, 2
9
10
28
31
6.5
180
19
1.2
1.1, 0.2, 46
3.7, 0.6, 100
0.2, 0.03, 7
2.2, 0.2, 82
0.3, 0.03, 24
18.2
13
a
41
16, 1
a
42
24, 3.2
32, 3.9
26, 3.5
2.4
10
22
1.3
3.7
18
47
2
0.5, 0.08, 26
a
b
Rat PK data obtained from racemic mixture. AUC measured (μg·h/
c
d
mL). Normalized to actual dose (μg/mL). Cl measured (mL min−1
kg−1).
improvement is more apparent when compared on an
equivalent exposure basis (Table 6). Additionally, 2 shows
efficacy beyond +48 h in this treatment model, results that will
be described in an upcoming publication.13
CONCLUSIONS
■
Starting from a phenotypic screening effort using a cell
protection assay, a class of azaindole-based compounds was
identified that was found to be inhibitors of the PB2 subunit of
the influenza A viral polymerase complex. Structure guided
optimization led to the design of a set of cyclohexylcarboxylic
acid analogues, highlighted by 2. Compound 2 was found to be
an orally bioavailable, extremely potent, and efficacious
inhibitor with a clean selectivity profile. Further testing showed
2 was extremely potent versus all influenza A strains tested,
including pandemic H1N1 and avian H5N1 flu strains. The
overall profile of 2, coupled with an early acting mechanism in
the influenza viral life cycle, translated into unprecedented
efficacy in a +48 h treatment mouse influenza model. This data
clearly indicate that 2 is a first-in-class influenza inhibitor that is
superior to the current standard-of-care agent oseltamivir with
a
Reagents and conditions: (a) CHCl3, 78%; (b) NaOMe, MeOH, 4
days, 85%; (c) DPPA, Et3N, BnOH, 77%; (d) H2, Pd/C, THF/
MeOH, 99%; (e) 2,4-dichloro-5-fluoropyrimidine, iPr2NEt, THF,
77%; (f) SFC chiral separation; (g) 56, Pd2(dba)3, K3PO4, 2-MeTHF,
water, 120 °C, 95%; (h) HCl, dioxane, MeCN, 95%; (i) NaOH, THF,
MeOH, 95%.
than 30% BWL loss at 10 mpk b.i.d. The efficacy observed with
2 not only represents a significant improvement over earlier
compounds, such as 28, but a dramatic improvement over the
current SOC, oseltamivir, at this higher viral challenge dose.
Compound 2 exhibited greater efficacy than 28 in its protection
against the drastic body weight loss in the infected animals. The
6674
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Figure 6. In vivo activity of 2 and oseltamivir in mouse influenza A model when administered 48 h after infection: survival and body weight curves of
male BALB/C mice (8 mice/study) inoculated with mouse-adapted influenza viruses A/PR/8/34 (5e3 TCID50/mouse) by intranasal instillation.
5-Chloro-3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-1-
tosyl-1H-pyrrolo[2,3-b]pyridine (56). Using the same procedure
described above and substituting 3-bromo-5-fluoro- lH-pyrrolo[2,3-
b]pyridine instead of 3-bromo-5-chloro-lH-pyrrolo[2,3-b]pyridine
afford the desired product, 56: 1H NMR (CDCl3) 8.45 (s, 1H),
8.15 (s, 2H), 8.10 (d, 2H), 7.85 (d, 1H), 7.45 (d, 2H), 2.32 (s, 3H),
1.33 (s, 9H); MS(ES+) 417.
5-Chloro-3-(5-fluoro-4-(methylthio)pyrimidin-2-yl)-1-tosyl-
1H-pyrrolo[2,3-b]pyridine (15). 2-Chloro-5-fluoro-4-methylsulfa-
nylpyrimidine, 14 (0.64 g, 3.60 mmol), 5-chloro-3-(4,4,5,5-tetrameth-
yl-1,3,2-dioxaborolan-2-yl)-1-tosyl-1H-pyrrolo[2,3-b]pyridine, 13
(1.30 g, 3.0 mmol), and aqueous Na2CO3 (4.5 mL of 2 M solution,
9.0 mmol) were dissolved in DME (14 mL). The mixture was purged
with nitrogen for 15 min, treated with Pd(PPh3)4 (0.17 g, 0.15 mmol),
and heated in a pressure tube at 125 °C for 15 min. After the mixture
was cooled to room temperature, the resulting precipitate was filtered
and washed with methanol to afford 15 (1.27 g, 94%): 1H NMR (400
MHz, CDCl3) δ 8.80 (d, J = 2.4, 1H), 8.60 (s, 1H), 8.40 (d, J = 2.4,
1H), 8.25 (d, J = 1.7, 1H), 8.19−8.03 (m, 2H), 7.30 (d, J = 8.1, 2H),
2.74 (s, 3H), 2.39 (s, 3H); MS (ES+) 449.3.
respect to overall potency, efficacy, and extended treatment
window. It is believed that 2 possesses great therapeutic
potential and is being evaluated in clinical trials.
EXPERIMENTAL SECTION
■
All commercially available reagents and anhydrous solvents were used
without further purification. Purity assessment for final compounds
was based on analytical HPLC: 4.6 mm × 50 mm Waters YMC Pro-
C18 column, 5 μm, 120A. Mobile phases are as follows: A, H2O with
0.2% formic acid; B, acetonitrile with 0.2% formic acid; gradient, 10−
90% B in 3 min with 5 min run time. The flow rate is 1.5 mL/min.
Unless specified otherwise, all compounds were ≥95% pure. Mass
samples were analyzed on a Micro Mass ZQ, ZMD, Quattro LC, or
Quatro II mass spectrometer operated in a single MS mode with
electrospray ionization. Samples were introduced into the mass
spectrometer using flow injection (FIA) or chromatography. The
mobile phase for all mass analysis consisted of acetonitrile−water
mixtures with either 0.2% formic acid or ammonium formate. 1H
NMR spectra were recorded using either a Bruker Avance 400 (400
MHz) or a Bruker Avance II-300 (300 MHz) instrument. The column
chromatography was performed using Teledyne ISCO RediSep
normal phase (35−70 μm) or RediSep Gold normal phase (25−40
μm) silica flash columns using a Teledyne ISCO Combiflash
Companion or Combiflash Rf purification system. Preparative reversed
phase chromatography was carried out using a Gilson 215 liquid
handler coupled to a UV−vis 156 Gilson detector, an Agilent Zorbax
SB-C18 column, 21.2 mm × 100 mm, with a linear gradient from 10%
to 90% CH3CN in H2O over 10 min (0.1% trifluoroacetic acid). The
flow rate was 20 mL/min.
5-Chloro-3-(5-fluoro-4-(methylsulfinyl)pyrimidin-2-yl)-1-
tosyl-1H-pyrrolo[2,3-b]pyridine (16). To a cold (0 °C) solution of
5-chloro-3-(5-fluoro-4-(methylthio)pyrimidin-2-yl)-1-tosyl-1H-
pyrrolo[2,3-b]pyridine, 15 (56.0 g, 124.7 mmol), in dichloromethane
(2.5 L) was added portionwise 3-chlorobenzenecarboperoxoic acid
(31.0 g, 131.1 mmol). The internal temperature was maintained below
5 °C. The resulting reaction mixture was slowly allowed to warm to
ambient temperatue over 4 h. The reaction mixture was quenched with
10% aqueous NaHSO3 solution (150 mL) and stirred for 30 min. The
organic layer was separated, washed with aqueous saturated NaHCO3
solution (2 × 600 mL), brine (600 mL), dried over Na2SO4, filtered,
and concentrated under reduced pressure. The resulting crude residue
was triturated with MTBE (2 × 150 mL) to afford the desired product
5-Chloro-3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-1-
tosyl-1H-pyrrolo[2,3-b]pyridine (13). Sodium hydride (2.2 g, 54.0
mmol) was added portionwise to a cold (0 °C) solution of 3-bromo-5-
chloro-lH-pyrrolo[2,3-b]pyridine, 12 (10.5 g, 45.0 mmol), in
dimethylformamide (70 mL) under a nitrogen atmosphere. After the
mixture was stirred at room temperature for 30 min, tosyl chloride (8.7
g, 46.0 mmol) was added to the mixture and the mixture was stirred at
room temperature for 18 h. The reaction mixture was quenched with
water (150 mL) and the resulting brown solid was filtered and dried
under vacuum to afford 3-bromo-5-chloro-1-tosyl-1H-pyrrolo[2,3-
1
16 (55.4 g. 96% yield) as a white solid: H NMR (DMSO-d6) δ 9.12
(s, 1H), 8.89 (d, J = 3 Hz, 1H), 8.67 (s, 1H), 8.53 (d, J = 3 Hz, 1H),
8.12 (d, J = 9 Hz, 2H), 7.46 (d, J = 9 Hz, 2H), 3.06 (s, 3H), 2.36 (s,
3H).
(S)-2-(5-Chloro-1-tosyl-1H-pyrrolo[2,3-b]pyridin-3-yl)-5-fluo-
ro-N-(piperidin-3-yl)pyrimidin-4-amine (17). To a solution of 5-
chloro-3-(5-fluoro-4-(methylsulfinyl)pyrimidin-2-yl)-1-tosyl-1H-
pyrrolo[2,3-b]pyridine, 16 (4.0 g, 8.6 mmol), and tert-butyl (3S)-3-
aminopiperidine-1-carboxylate (2.1 g, 10.3 mmol) in DMF (50 mL)
was added diisopropylethylamine (3.0 mL, 17.2 mmol). The reaction
mixture was heated at 80 °C for 18 h. The mixture was cooled to room
temperature and diluted into aqueous saturated NH4Cl solution and
extracted twice with EtOAc. The combined organic phases were
washed with brine (3 times), dried (MgSO4), filtered, and
concentrated under vacuum. The resulting residue was purified via
silica gel chromatography (0%−80% EtOAc/hexanes) to afford the
intermediate piperidine: ESI-MS m/z calcd 601.5, found 601.43 (M +
1).
1
b]pyridine (14.8g, 85%): H NMR (CDCl3) 8.4 (s, 1H), 8.1−82 (d,
2H), 7.8−7.9 (d, 2H), 7.3−7.4 (d, 2H), 2.4 (3H, s); MS(ES+) 387.
A 500 mL round-bottom flask under a nitrogen atmosphere was
charged with 3-bromo-5-chloro-1-tosyl-1H-pyrrolo[2,3-b]pyridine (6.9
g, 18.0 mmol), bispinacolatodiboron (6.9 g, 27.0 mmol), PdCl2(dppf)2
(1.5 g, 2.0 mmol), potassium acetate (5.3 g, 54.0 mmol), and
dimethoxyethane (100 mL). The reaction mixture was stirred at 90 °C
for 18 h. The reaction mixture was diluted with ethyl acetate (200 mL)
and washed with brine. The organic phase was dried over magnesium
sulfate, filtered, and concentrated in vacuo. The resulting residue was
purified by silica gel chromatography, using as eluent pentane/EtOAc
1
To a solution of tert-butyl (3S)-3-[[2-[5-chloro-1-(p-tolylsulfonyl)-
pyrrolo[5,4-b]pyridin-3-yl]-5-fluoropyrimidin-4-yl]amino]piperidine-
1-carboxylate (2.1 g, 3.5 mmol) in CH2Cl2 (30 mL) was added
0−20% and then triturated with pentane to afford 13 (4g, 50%): H
NMR (CDCl3) 8.30 (s, 1H), 8.10 (s, 2H), 8.00−8.05 (d, 2H), 7.20 (d,
2H), 2.40 (s, 3H), 1.40 (s, 9H); MS(ES+) 433.
6675
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
trifluoroacetic acid (20 mL). After the reaction mixture was stirred at
room temperature for 75 min, the mixture was concentrated under
vacuum. The crude residue was diluted with EtOAc and neutralized
with 1 N sodium hydroxide solution. The aqueous phase was separated
and extracted again with EtOAc. The combined organic phases were
dried (MgSO4), filtered, and concentrated under vacuum to afford the
g, 475.7 mmol), was dissolved in toluene (1.0 L). Diphenylphosphor-
ylazide (112.8 mL, 523.3 mmol) and triethylamine (72.9 mL, 523.3
mmol) were added. The reaction mixture was heated to 90 °C for 2 h.
Benzyl alcohol (49.2 mL, 475.7 mmol) was added, and the mixture
was heated to 90 °C over 3 days. The mixture was cooled to room
temperature and diluted with EtOAc (500 mL) and aqueous saturated
NaHCO3 solution. The organic phase was washed with brine, dried
(MgSO4), filtered, and concentrated in vacuo. The resulting crude
material was purified by silica gel chromatography with dichloro-
1
desired product, 17, as a light yellow solid: H NMR (300 MHz,
DMSO-d6) δ 8.76 (d, J = 2.5, Hz, 1H), 8.50 (d, J = 2.5 Hz, 1H), 8.44
(s, 1H), 8.27 (d, J = 4.0 Hz, 1H), 8.06 (d, J = 8.5 Hz, 2H), 7.66 (d, J =
6.9 Hz, 1H), 7.45 (d, J = 8.2 Hz, 2H), 4.17 (m, 1H), 3.17 (dd, J = 3.1,
11.8 Hz, 1H), 2.99−2.94 (m, 1H), 2.67−2.60 (m, 1H), 2.38−2.34 (m,
1H), 2.06−2.02 (m, 1H), 1.77−1.73 (m, 1H), and 1.63−1.50 (m, 2H)
ppm; LCMS tR = 2.1 min, (M + 1) 501.5, (M − 1) 499.5.
(R)-3-((S)-3-((2-(5-Chloro-1-tosyl-1H-pyrrolo[2,3-b]pyridin-3-
yl)-5-fluoropyrimidin-4-yl)amino)piperidin-1-yl)propane-1,2-
diol (17b). To a solution of (S)-2-(5-chloro-1-tosyl-1H-pyrrolo[2,3-
b]pyridin-3-yl)-5-fluoro-N-(piperidin-3-yl)pyrimidin-4-amine, 17 (2.22
g, 4.43 mmol), in ethanol was added (S)-glycidol (0.66 g, 8.86 mmol).
The reaction mixture was heated at 80 °C overnight. The reaction
mixture was concentrated in vacuo. The resulting residue was purified
by silica gel chromatography (1−15% MeOH/CH2Cl2 gradient) to
afford 1.48 g (58%) of the desired product: LCMS tR = 2.05 min, (M +
H) 576.47.
1
methane to afford 115 g of oil (77%). H NMR show it contains
BnOH (about 0.05 equiv). Product was used without further
purification: 1H NMR (300 MHz, CDCl3) δ 7.40−7.24 (m, 5H),
6.41 (t, J = 7.4 Hz, 1H), 6.21−6.04 (m, 1H), 5.15−4.94 (m, 2H),
4.63−4.45 (m, 1H), 4.30−4.18 (m, 1H), 3.70 (s, 2H), 3.49 (s, 1H),
2.81 (br s, 1H), 2.68 (br s, 1H), 2.08 (s, 1H), 1.76−1.56 (m, 1H),
1.52−1.35 (m, 1H), 1.33−1.14 (m, 1H), 1.12−0.87 (m, 1H) ppm.
( )-trans-Methyl 3-Aminobicyclo[2.2.2]octane-2-carboxy-
late (53). A solution of racemic trans-methyl 3-(((benzyloxy)-
carbonyl)amino)bicyclo[2.2.2]oct-5-ene-2-carboxylate, 52 (115.0 g,
364.7 mmol), in THF (253 mL) and MeOH (253 mL) was placed
under 40 psi of hydrogen overnight. Some exotherm was observed.
The reaction mixture was filtered through Celite and washed with
MeOH. The filtrate was concentrated in vacuo to afford 69 g of the
desired product as an oil: 1H NMR (400 MHz, CDCl3) δ 3.63 (d, J =
5.6 Hz, 3H), 3.30 (d, J = 6.7 Hz, 1H), 2.11 (d, J = 6.6 Hz, 1H), 1.91 (t,
J = 7.3 Hz, 1H), 1.80−1.64 (m, 1H), 1.63−1.38 (m, 6H), 1.36−1.23
(m, 2H) ppm.
(2S,3S)-Methyl 3-((2-Chloro-5-fluoropyrimidin-4-yl)amino)-
bicyclo-[2.2.2]octane-2-carboxylate (55). To a solution of
racemic-trans-methyl 3-aminobicyclo[2.2.2]octane-2-carboxylate, 53
(1.30 g, 7.09 mmol), and 2,4-dichloro-5-fluoropyrimidine (1.77 g,
10.64 mmol) in DMF (20 mL) was added N,N-diisopropylethylamine
(4.94 mL, 28.38 mmol). The reaction mixture was stirred at room
temperature for 100 min. The mixture was diluted into aqueous
saturated NH4Cl solution and extracted twice with EtOAc. The
combined organic phases were washed three times with brine, dried
(MgSO4), filtered, and concentrated in vacuo. The crude residue was
purified via silica gel chromatography (0−10% MeOH/CH2Cl2
gradient) to afford 1.41 g of the racemic product, 39: LC/MS
gradient 10−90%, 0.1% formic acid, 5 min, C18/ACN tR = 1.14 min,
(M + H) 314.11.
(R)-3-((S)-3-((2-(5-Chloro-1H-pyrrolo[2,3-b]pyridin-3-yl)-5-
fluoropyrimidin-4-yl)amino)piperidin-1-yl)propane-1,2-diol
(10). To a solution of (R)-3-((S)-3-((2-(5-chloro-1-tosyl-1H-pyrrolo-
[2,3-b]pyridin-3-yl)-5-fluoropyrimidin-4-yl)amino)piperidin-1-yl)-
propane-1,2-diol, 17b (1.48 g, 2.57 mmol), in methanol (20 mL) was
added sodium methoxide (10 mL of 25% w/v solution, 46.28 mmol).
After 5 min, the reaction mixture was diluted into water and extracted
twice with EtOAc. The combined organic phases were dried (MgSO4),
filtered, and concentrated in vacuo. The resulting residue was purified
by silica gel chromatography (1−20% MeOH/CH2Cl2 gradient) to
1
afford 0.66 g (45%) of the desired product: H NMR (300 MHz,
DMSO-d6) δ 12.31 (s, 1H), 8.72 (d, J = 2.4 Hz, 1H), 8.28 (d, J = 2.4
Hz, 1H), 8.20 (s, 1H), 8.17 (d, J = 4.0 Hz, 1H), 7.33 (d, J = 7.6 Hz,
1H), 4.51 (m, 1H), 4.37 (s, 1H), 4.25 (m, 1H), 3.64 (m, 1H), 3.35 (s,
2H), 3.08−2.95 (m, 1H), 2.80−2.70 (m, 1H), 2.47−2.25 (m, 2H),
2.22−2.12 (m, 2H), 1.99−1.90 (m, 1H), 1.70−1.60 (m, 2H), and 1.45
(m, 1H) ppm; HRMS (ESI) of C19H23ClFN6O2 [M + H] calcd,
421.154 95; found, 421.154 54.
meso-endo-Tetrahydro-4,7-ethanoisobenzofuran-1,3-dione
(50). To a cold (0 °C) solution of maleic anhydride (210.0 g, 2142.0
mmol) in CHCl3 (2.3 L) was added cyclohexa-1,3-diene (224.5 mL,
2356.0 mmol) slowly over 50 min. The mixture was warmed to room
temperature and stirred overnight in the dark. After removal of the
solvent under reduced pressure, 2.1 L of MeOH was added to the
mixture and the mixture was heated to 50 °C for 10 min and then
cooled to 0 °C. The resulting precipitate was filtered and dried in an
oven at 45 °C overnight to afford 283 g of a white solid. The resulting
endo (meso) Diels−Alder cycloaddition product was used without
further purification.
( )-trans-3-(Methoxycarbonyl)bicyclo[2.2.2]oct-5-ene-2-car-
boxylic Acid (51). A solution of meso-endo-tetrahydro-4,7-ethanoiso-
benzofuran-1,3-dione, 50 (74.5 g, 418.1 mmol), was stirred in NaOMe
(764.9 mL of 25% w/w solution in MeOH, 3345.0 mmol). The
reaction mixture was stirred at room temperature for 4 days, yielding a
white suspension. The reaction mixture was concentrated in vacuo to
remove approximately 300 mL of MeOH. In another flask, HCl (315.9
mL of 36.5% w/w, 3763.0 mmol) in 300 mL of water was cooled to 0
°C. The reaction mixture was added into this HCl solution slowly, and
a white solid precipitated. The remaining methanol was removed
under reduced pressure. The mixture was cooled to 0 °C and stirred
for 30 min. The precipitate was filtered, washed with water 3 times,
giving an off-white solid. The remaining water was removed under
reduced pressure to afford 82 g of a white solid that was used without
further purification.
The racemic product 54 was submitted for SFC chiral separation to
afford the single enantiomer 55.
(2S,3S)-Methyl 3-((5-Fluoro-2-(5-fluoro-1-tosyl-1H-pyrrolo-
[2,3-b]pyridin-3-yl)pyrimidin-4-yl)amino)bicyclo[2.2.2]octane-
2-carboxylate (57). A mixture of 5-fluoro-1-(p-tolylsulfonyl)-3-
(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyrrolo[2,3-b]pyridine,
56 (10.22 g, 24.54 mmol), (2S,3S)-methyl 3-((2-chloro-5-fluoropyr-
imidin-4-yl)amino)bicyclo[2.2.2]octane-2-carboxylate, 55, (5.5 g,
17.53 mmol), K3PO4 (0.95 g, 2.92 mmol) in 2-methyl THF (100
mL), and H2O (20 mL) was degassed under a stream of nitrogen for
15 min. To the mixture was added X-Phos (1.00 g, 2.10 mmol) and
1,5-diphenylpenta-1,4-dien-3-one, palladium (0.40 g, 0.44 mmol). The
reaction vessel was sealed and heated to 115 °C for 3 h. The aqueous
phase was removed, and the organic phase was filtered through Celite
and concentrated in vacuo. The resulting residue was purified by silica
gel chromatography (20% EtOAc/hexanes) to afford 9.5 g (95%) of
the desired product: LCMS gradient 10−90%, 0.1% formic acid, 5 min,
C18/ACN, tR = 4.2 min, (M + H) 567.84.
(2S,3S)-Methyl 3-((5-Fluoro-2-(5-fluoro-1H-pyrrolo[2,3-b]-
pyridin-3-yl)pyrimidin-4-yl)amino)bicyclo[2.2.2]octane-2-car-
boxylate (58). To a mixture of (2S,3S)-methyl 3-((5-fluoro-2-(5-
fluoro-1-tosyl-1H-pyrrolo[2,3-b]pyridin-3-yl)pyrimidin-4-yl)amino)-
bicyclo[2.2.2]octane-2-carboxylate, 57 (7.44 g, 13.11 mmol), in
CH3CN (40 mL) at room temperature was added HCl (19.66 mL
of 4 M solution in dioxane, 78.66 mmol). The resulting solution was
heated to 65−70 °C for 6.5 h and then cooled to room temperature
where stirring was maintained overnight. The mixture was diluted with
MTBE, and the resulting precipitate was filtered. The solid was
suspended in MBTE and filtered twice more. The solid was dried
(
)-trans-Methyl 3-(((Benzyloxy)carbonyl)amino)-
bicyclo[2.2.2]oct-5-ene-2-carboxylate (52). Racemic trans-3-
methoxycarbonylbicyclo[2.2.2]oct-5-ene-2-carboxylic acid, 51 (100.0
6676
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
under vacuum to afford 6.36 g (95%) of the desired product as the
hydrochloride salt: H NMR (400 MHz, MeOD) δ 8.57 (dd, J = 9.1,
2.7 Hz, 1H), 8.50 (s, 1H), 8.35 (s, 1H), 8.31 (d, J = 5.6 Hz, 1H), 5.18
(d, J = 6.5 Hz, 1H), 3.70 (s, 3H), 2.97 (d, J = 6.7 Hz, 1H), 2.15 (s,
1H), 2.05 (s, 1H), 2.01−1.50 (m, 8H).
ACKNOWLEDGMENTS
■
1
We thank Barry Davis, Nigel Ewing, and Ravikanth Veluri for
high-resolution mass spectrometric determination of key
compounds. We thank Brian Krueger for invaluable SFC chiral
purifications. In addition, we thank Yuegang Zhang for small
molecule X-ray crystallography.
(2S,3S)-3-((5-Fluoro-2-(5-fluoro-1H-pyrrolo[2,3-b]pyridin-3-
yl)pyrimidin-4-yl)amino)bicyclo[2.2.2]octane-2-carboxylic
Acid (2). A solution of NaOH (119.5 mL of 2 N solution, 239.0
mmol) was added to a solution of (2S,3S)-methyl 3-((5-fluoro-2-(5-
fluoro-1H-pyrrolo[2,3-b]pyridin-3-yl)pyrimidin-4-yl)amino)bicyclo-
[2.2.2]octane-2-carboxylate, 58 (24.7 g, 59.8 mmol), in methanol (98
mL) and THF (98 mL). The reaction mixture was stirred at 30 °C for
2 h. The mixture was cooled to room temperature and stirred
overnight. The organic solvents were concentrated under vacuum. 2-
Methyl THF was added (150 mL), and the organics were
concentrated in vacuo. Additional 2-methyl THF was added, and the
mixture was neutralized with 6 N HCl while stirring the biphasic
mixture vigorously. The organic layer was separated, and the aqueous
layer was extracted with 2-methyl THF (3 × 200 mL), adjusting the
pH between each extraction to ∼pH 6. The combined organic phases
were dried over Na2SO4, filtered, and concentrated in vacuo. The solid
was taken up again in a large volume of warm 2-methyl THF and
filtered again through Celite. The clear solution was concentrated in
vacuo and dried in a vacuum oven overnight at 50 °C, followed by an
additional night at 80−85 °C to afford 22.75 g of the desired product:
1H NMR (300 MHz, DMSO-d6) δ 12.71 (br s, 1H), 8.58 (s, 1H), 8.47
ABBREVIATIONS USED
■
bDNA, branched DNA; SAR, structure−activity relationship;
AUC, area under curve; CPE, cytopathic effect assay; b.i.d.,
twice a day dosing; SOC, standard of care; m7GTP, methyl-7-
guanosine 5′-triphosphate; NaHCO3, sodium bicarbonate;
i
EtOH, ethanol; Pr2NEt, diisopropylethyl amine; NaOMe,
sodium methoxide; KOAc, potassium acetate; PdCl2(dppf),
[1,1′-bis(diphenylphosphino)ferrocene]dichloropalladium(II),
complex with dichloromethane; Pd(PPh3)4, tetrakis-
(triphenylphosphine)palladium(0); DPPA, diphenylphosphoryl
azide; BnOH, benzyl alcohol; Cmax, maximum concentration
that a drug achieves after dosing; Cmin, minimum concentration
that a drug achieves after dosing and prior to following dose;
TCID50, 50% tissue culture infective dose
(dd, J = 9.6, 2.8 Hz, 1H), 8.41 (d, J = 4.8 Hz, 1H), 8.39−8.34 (m, 1H),
4.89−4.76 (m, 1H), 2.94 (d, J = 6.9 Hz, 1H), 2.05 (br s, 1H), 1.96 (br
s, 1H), 1.68 (complex m, 7H); 13C NMR (300 MHz, DMSO-d6) δ
174.96, 157.00, 155.07, 153.34, 152.97, 145.61, 142.67, 140.65, 134.24,
133.00, 118.02, 114.71, 51.62, 46.73, 28.44, 28.00, 24.90, 23.78, 20.88,
18.98; LCMS gradient 10−90%, 0.1% formic acid, 5 min, C18/ACN,
tR = 2.24 min, (M + H) 400.14; HRMS (ESI) of C20H20F2N5O2 [M +
H] calcd, 400.157 95; found, 400.157 56.
In Vitro Antiviral Assays. Details are presented in the Supporting
Information. MDCK cell protection and cytotoxicity assays were
performed using a modification of standard methods, essentially the 3-
day CPE-based approach employing an ATP cell viability end point.
Replication of influenza virus RNA, either positive-strand or negative-
strand polarity, was measured using strand-specific probes using the
bDNA QuantiGene method (Affymetrix), as previously described.
Influenza virus sensitivity to the neuraminidase inhibitors oseltamivir
carboxylate and zanamivir was determined by the chemiluminescent
neuraminidase inhibitor assay using the NA-XTD kit (Applied
Biosystems; Foster City, CA) per the manufacturer’s recommenda-
tions.
REFERENCES
■
(1) Wagaman, P. C.; Leong, M. A.; Simmen, K. A. Development of a
novel influenza A antiviral assay. J. Virol. Methods 2002, 105, 105−114.
(2) Thompson, W. W.; Shay, D. K.; Weintraub, E.; Brammer, L.;
Bridges, C. B.; Cox, N. J.; Fukuda, K. Influenza-associated hospital-
izations in the United States. JAMA, J. Am. Med. Assoc. 2004, 292,
1333−1340.
(3) CDC. Seasonal Influenza: Key Facts about Influenza (Flu) and
Flu Vaccine. http://cdc.gov/flu/keyfacts.htm (accessed July 30, 2012).
(4) (a) Cohen, J. Surprising twist in debate over lab-made H5N1.
Science 2012, 335, 1155−1156. (b) Herfst, S.; Schrauwen, J. A.;
Linster, M.; Chutinimitkul, S.; deWit, E.; Munster, V. J.; Sorrell, E. M.;
Bestebroer, T. M.; Burke, D. F.; Smith, D. J.; Rimmelzwaan, G. F.;
Osterhaus, A. D. M. E.; Fouchier, R. A. M. Airborne transmission of
influenza A/H5N1 virus between ferrets. Science 2012, 336, 1524−
1541.
(5) Moscona, A. Neuraminidase inhibitors for influenza. N. Engl. J.
Med. 2005, 353, 1363−1373.
(6) Thorland, K.; Awad, T.; Boivin, G.; Thabane, L. Systematic
review of influenza resistance to the neuraminidase inhibitors. BMC
Infect. Dis. 2011, 11, 134.
ASSOCIATED CONTENT
(7) Le, Q. M.; Kiso, M.; Someya, K.; Sakai, Y. T.; Nguyen, T. H.;
Nguyen, K. H. L.; Pham, N. D.; Ngyn, H. H.; Yamada, S.; Muramoto,
Y.; Horimoto, T.; Takada, A.; Goto, H.; Suzuki, Y.; Kawaoka, Y. Avian
flu: isolation of drug-resistant H5N1 virus. Nature 2005, 437, 1108.
(8) Noah, J. W.; Severson, W.; Noah, D. L.; Rasmussen, L.; White, E.
L.; Jonsson, C. B. A cell-based luminescence assay is effective for high-
throughput screening of potential influenza antivirals. Antiviral Res.
2007, 73, 50−59.
(9) (a) Guilligay, D.; Tarendeau, F.; Resa-Infante, P.; Coloma, R.;
Crepin, T.; Sehr, P.; Lewis, J.; Ruigrok, R. W.; Ortin, J.; Hart, D. J.;
Cusack, S. The structural basis for cap binding by influenza virus
polymerase subunit PB2. Nat. Struct. Mol. Biol. 2008, 15, 500−506.
(b) Pautus, S.; Sehr, P.; Lewis, J.; Fortune, A.; Wolkerstorfer, A.;
Szolar, O.; Guilligay, D.; Lundardi, T.; Decout, J.-L.; Cusack, S. New 7-
methylguanine derivatives targeting the influenza polymerase PB2 cap-
binding domain. J. Med. Chem. 2013, 56, 8915−8930.
■
S
* Supporting Information
1H NMR, MS, and HRMS spectral data for all of the final
compounds; experimental details for 7, 9−11, 18, 20, 21, 26,
28, 29, 31, 40, 42, 45, and 46; protocols for PB2 binding assay;
CPE and bDNA cellular assays; mouse influenza pharmacology
models. This material is available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: 617-961-7727. E-mail: michael_clark@vrtx.com.
Author Contributions
(10) Fechter, P.; Mingay, L.; Sharps, J.; Chambers, A.; Fodor, E.;
Brownlee, G. G. Two aromatic residues in the PB2 subunit of
influenza A RNA polymerase are crucial for cap binding. J. Biol. Chem.
2003, 278, 20381−20388.
(11) Palese, P.; Shaw, M. L. Orthomyxoviridae: The Viruses and
Their Replication. In Fields Virology; Knipe, D. M., Howley, P. M.,
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
6677
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678

Journal of Medicinal Chemistry
Article
Eds.; Lippincott Williams & Wilkins: Philadelphia, PA, 2007; pp
1647−1689.
(12) Charifson, P. S.; Clark, M. P.; Bandarage, U. K.; Bethiel, R. S.;
Boyd, M. J.; Davies, I.; Deng, H.; Duffy, J. P.; Farmer, L. J.; Gao, H.;
Gu, W.; Kennedy, J. M.; Ledford, B.; Ledeboer, M. W.; Maltais, F.;
Perola, E.; Wang, T. Inhibitors of Influenza Viruses Replication. WO
2013/019828, February 2, 2013.
(13) Byrn, R. A.; Jones, S.; Asmal, M.; Bennett, H. B.; Bral, C.; Clark,
M. P.; Jacobs, M.; Kwong, A.; Ledeboer, M.; Leeman, J. R.; McNeil,
C.; Murcko, M.; Nezami, A.; Perola, E.; Rijnbrand, R.; Saxena, K.;
Travejo, J.; Tsai, A.; Zhou, Y.; Charifson, P. S.; Preclinical activity of
VX-787, a first-in-class, orally bioavailable inhibitor of the influenza
viral polymerase PB2 subunit. Manuscript submitted for publication.
(14) Memmott, C.; Taylor, W. Manuscript in preparation.
(15) Kerns, E.; Di, L. Transporters. In Drug-like Properties: Concepts,
Structure, Design and Methods; Elsevier: Burlington, MA, 2008; pp
103−121.
6678
dx.doi.org/10.1021/jm5007275 | J. Med. Chem. 2014, 57, 6668−6678