Design and synthesis of spirocyclic compounds as HCV replication inhibitors by targeting viral NS4B protein

Article abstract of DOI:10.1016/j.bmcl.2014.03.080

Two novel series of spirocyclic piperidine analogs appended to a pyrazolo[1,5-a]pyridine core were designed, synthesized and evaluated for their anti-HCV activity. A series of piperidine ketals afforded dispiro 6p which showed excellent in vitro anti-HCV activities (EC₅₀ of 1.5 nM and 1.2 nM against genotype 1a and 1b replicons, respectively). A series of piperidine oxazolidinones afforded 27c which showed EC₅₀'s of 10.9 nM and 6.1 nM against 1a and 1b replicons, respectively. Both compounds 6p and 27c bound directly to non-structural NS4B protein in vitro (IC₅₀'s = 10.2 and 30.4 nM, respectively) and exhibited reduced potency in replicons containing resistance mutations encoding changes in the NS4B protein.

Full text of DOI:10.1016/j.bmcl.2014.03.080

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2288–2294
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of spirocyclic compounds as HCV replication
inhibitors by targeting viral NS4B protein
a
b
b
a
Vincent W.-F. Tai a, , Dulce Garrido , Daniel J. Price , Andrew Maynard , Jeffrey J. Pouliot ,
⇑
Zhiping Xiong a, John W. Seal III b, Katrina L. Creech b, Luz H. Kryn b, Todd M. Baughman c, Andrew J. Peat a
a GlaxoSmithKline, Antiviral Discovery Performance Unit, 5 Moore Drive, Research Triangle Park, NC 27709, USA
b GlaxoSmithKline, Platform Technology and Science, 5 Moore Drive, Research Triangle Park, NC 27709, USA
c GlaxoSmithKline, Anti Bacterial Discovery Performance Unit, 5 Moore Drive, Research Triangle Park, NC 27709, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Two novel series of spirocyclic piperidine analogs appended to a pyrazolo[1,5-a]pyridine core were
designed, synthesized and evaluated for their anti-HCV activity. A series of piperidine ketals afforded
dispiro 6p which showed excellent in vitro anti-HCV activities (EC50 of 1.5 nM and 1.2 nM against geno-
type 1a and 1b replicons, respectively). A series of piperidine oxazolidinones afforded 27c which showed
EC50’s of 10.9 nM and 6.1 nM against 1a and 1b replicons, respectively. Both compounds 6p and 27c
bound directly to non-structural NS4B protein in vitro (IC50’s = 10.2 and 30.4 nM, respectively) and exhib-
ited reduced potency in replicons containing resistance mutations encoding changes in the NS4B protein.
Ó 2014 Elsevier Ltd. All rights reserved.
Received 18 February 2014
Revised 22 March 2014
Accepted 25 March 2014
Available online 3 April 2014
Keywords:
NS4B
Hepatitis C virus (HCV)
Replication inhibitors
Spirocyclic ketals
Pyrazolo[1,5-a]pyridine
Hepatitis C is a blood-borne disease of the liver transmitted by
the Hepatitis C Virus (HCV) and is a leading cause of chronic liver
disease and liver transplantations.1,2 The World Health Organiza-
tion (WHO) estimates that about 150 million people are chronically
infected with HCV, and more than 350,000 people die every year
from hepatitis C-related liver diseases worldwide.3 Unlike hepatitis
A and B viruses, no vaccine against HCV is available. The current
standard of care for treating genotype 1 HCV patients involves the
combination of an oral protease inhibitor, boceprevir or telaprevir,
along with pegylated IFN (PegIFN) and ribavirin.4–6 These triple
therapy regimens offer an improvement in cure rate and shorten
the duration of treatment in comparison to PegIFN and ribavirin
alone. However, these therapies suffer from significant side effects,
such as anemia, serious skin reactions/rash, and fatigue as well as
drug–drug interactions that require close monitoring. Opportuni-
ties remain for improvement of the first generation of direct-acting
antiviral (DAA) agents in areas such as resistance, tolerability, and
pan-genotypic efficacy. Furthermore, identification of a regimen
with all oral agents that removes the need for IFN remains a long
sought after goal. The recent FDA approval of Sovaldi™ (sofosbuvir)
is a significant step towards these goals.
11 proteins, including 4 structural proteins (C, E1, E2, and p7) and 6
nonstructural proteins (NS2, NS3, NS4A, NS4B, NS5A, and NS5B)
and the so-called ‘F’ protein.7,8 Several of the non-structural pro-
teins have proven to be viable targets for clinical HCV interven-
tion.9,10 The non-structural protein 4B (NS4B) is less well
studied; however recent reports have shown this protein is an inte-
gral part of the replication complex, and therefore represents a no-
vel target for HCV therapy.11–16 Recently, we17–19 and others20–22
have described small molecule inhibitors of HCV that target
NS4B. These compounds generate resistant mutations in the HCV
replicon, including H94N, F98L, and V105M in the NS4B sequence
of genotype 1b, and have been shown to directly bind the NS4B
protein. The current work extends these chemical series,17–19
resulting in compounds of novel structure that exhibit low-nM po-
tency against HCV replicons.
CF3
CF3
N
O
N
O
N
N
N
N
Br
Cl
Cl
O
N
O
HCV is a single-stranded RNA virus in the Flaviviridae family,
encoding a polyprotein of ꢀ3000 amino acids that is processed into
N
O
1
2
OH
EC50 HCV 1a = 0.3 nM
EC50 HCV 1b = 2.0 nM
EC50 HCV 1a = 31.4 nM
EC50 HCV 1b = 45.5 nM
⇑
Corresponding author. Tel.: +1 919 483 5841.
E-mail address: vincent.w.tai@gsk.com (V.W.-F. Tai).
http://dx.doi.org/10.1016/j.bmcl.2014.03.080
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
2289
CF3
N
N
O
N
12 steps
N
CO2Me
CO2Me
OH
Cl
3
H
N
Cbz
N
CF3
c
a, b
O
N
N
3
+
O
O
R
O
Cl
R1 R2
Figure 1. (A) X-ray crystal structure of compound 1. (B). Overlay of models for
compounds 2 (cyan) and 6e (green).
6
4
5
Scheme 1. General synthetic scheme for spiro analogs. Reagents and conditions: (a)
diol (1 equiv), pTsOH (0.05 equiv), toluene, 135–140 °C, Dean–Stark apparatus; (b)
H2 (1 atm), 10% Pd/C Degussa type, MeOH; (c) Method A: amine (1–1.5 equiv), T3P
(1-propanephosphoric acid anhydride, 50% wt solution in EtOAc, 1.5 equiv), Hunig’s
base (3 equiv), DMF; or Method B: amine (1–1.5 equiv), HATU (1.2 equiv), Hunig’s
base (3 equiv), DMF. (In 6h, 2-benzyloxy-1,3-propanediol was used as starting
material which was removed during hydrogenation of Cbz group.)
Compounds 1 and 2 are potent HCV replication inhibitors targeting
NS4B (Fig. 1).17,18 An X-ray crystal structure of compound 123 re-
vealed that the piperidine and oxazolidinone rings are orthogonal
to each other (Fig. 1A). The orthogonal orientation of the two rings
is also favored in the energy minimized molecular modeling calcu-
lations for analog 2. As part of our efforts to optimize this class of
inhibitors, new chemotypes in the amide region were explored.
Spiro-bicyclic motifs offer unique rigid 3-dimensional frameworks
wherein the two rings are constrained to perpendicular planes
(Fig. 1B). Furthermore, spirocyclic ketal substructures such as spiro-
ketals are found in numerous natural products from diverse sources
such as insects, microbes, plants, fungi and marine organisms.24
Two classes of spirocyclic compounds based on the potent pyrazol-
o[1,5-a]pyridine core are reported herein as HCV replication
inhibitors.
of 1-Cbz-4-piperidone 4 with the corresponding diol followed by
hydrogenolysis (Scheme 1). The core acid 3 was coupled to amines
5 using standard amide coupling conditions such as HATU or T3P.25
Compounds 6a–6h (Table 1) were readily prepared using this gen-
eral protocol. Ketone 6i was prepared in 81% yield by oxidation of
compound 6h with Dess–Martin periodinane (DMP).
The synthesis of simple dispiro compounds 6j–6m is shown in
Scheme 2. The diols are either commercially available or can be
prepared from reduction of the 1,1-dialkyl esters.26 Oxetane analog
6m was synthesized based on a cyclization strategy employing
The pyrazolopyridine core 3 was prepared in 12 steps from di-
methyl pyrazolo[1,5-a]pyridine-2,3-dicarboxylate with 35% overall
yield.18 The amines 5 are either commercially available or can be
readily synthesized by a two-step procedure involving ketalization
diol 12, which was prepared using
a modified ketalization
procedure with benzene/DMF as solvents to aid solubility of
Table 1
HCV 1a and HCV 1b replicon efficacy of spiro ketal series of compoundsa
Compound
R
EC50 1a (nM)
EC50 1b (nM)
Compound
R
EC50 1a (nM)
9.0
EC50 1b (nM)
4.1
O
N
O
O
N
6a
>500
>500
6j
O
O
O
N
N
N
N
6b
6c
6d
299
33.8
46.3
161
6.1
8
6k
6l
5.1
2.2
7.6
14.2
3.2
O
O
O
O
N
N
N
OH
O
O
O
O
O
6m
14.4
O
O
O
OH
O
6e
6f
7.4
23.8
42.2
1.2
9.8
6n
6o
6p
6q
4.3
32.5
1.5
2.2
8.2
1.2
3.8
O
N
N
N
N
O
OH
O
O
O
N
N
N
N
O
OH
O
O
O
6g
209
O
O
O
O
O
OH
O
6h
6i
19.5
18
13.8
19.1
5.1
O
O
O
a
All compounds in Tables 1 and 2 were tested in a cell based cytotoxicity assay in Huh-7 cells (genotype 1b) and showed CC50s >50
M.
lM except compounds 6d, 6f with CC50s
>25
l

2290
V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
O
O
provide larger quantities of the desired R-isomer 6p required for
HO
HO
a
in vitro and in vivo studies. The availability of chiral amine 19 also
allowed quick access to analogs with different cores. Chiral resolu-
tion of amine 16 was achieved by crystallization of the diastereo-
b
EtO
EtO
6k
8
meric salts30 of amine 16 and (+)-O,O0-di-p-toluoyl-
D-tartaric acid
7
in ethanol (Scheme 3, Method B). The desired R-isomer was ob-
tained as a 2:1 stoichiometry of amine/acid in 26% yield. The abso-
lute stereochemistry of the 3-O-benzoate group was determined
by X-ray crystallography of diastereomeric salt 18 (Fig. 2).23 Wash-
ing with aqueous sodium bicarbonate provided the free base 19 in
88% yield with 97% enantiomeric excess. Using this diastereomeric
crystallization methodology, 10 g of chiral amine 19 was readily
obtained. Similar to Method A, the amine 19 was coupled to acid
3 and the benzoate saponified to give compound 6p in 93% yield
over 2 steps.
A second series of spiro compounds with five-membered oxazo-
lidinone and lactam rings was also identified through focused
screening efforts.31 Amines 21a and 21b were obtained by depro-
tection of commercially available N-Boc derivatives 20a and 20b
respectively. The amines 21a and 21b were coupled to acid 3 under
T3P conditions to give compounds 22a and 22g in 89% and 85%
yield, respectively (Scheme 4). The N-alkyl analogs 22b–22f, 22h
were prepared by simple alkylation of compounds 22a and 22g
with the corresponding halides.
Hydroxyl piperidine analogs 27a–d were prepared as shown in
Scheme 5. Treatment of ketone 14 with potassium cyanide under
basic conditions gave cyanohydrin 23 as a mixture of diastereoiso-
mers in 84% yield. The nitrile group in 23 was reduced under
hydrogen in the presence of platinum oxide and Boc anhydride
to give a mixture of syn- and anti-isomers (rac-24a and rac-24b)
which were easily separated by silica gel chromatography. The rel-
ative stereochemistry of each compound was determined at a later
stage (vide infra). Acid hydrolysis of the Boc group in rac-24a fol-
lowed by oxazolidinone formation with carbonyldiimidazole (CDI)
gave spiro oxazolidinone rac-25a in 77% yield.
Methylation of rac-25a followed by purification using chiral
supercritical fluid chromatography (SFC) gave enantiomers 26a
and 26b each in 25% yield. The absolute stereochemistry of com-
pounds 26a and 26b was determined by vibrational circular
dichroism (VCD) and infrared (IR) analyses (vide supra). Removal
of the protecting groups in 26a and 26b followed by coupling with
core acid 3 gave compounds 27a and 27b, respectively. Similarly,
O
HO
HO
d
c
MeO
MeO
6l
OBn
OBn
O
10
9
Cbz
N
Cbz
N
OH
OH
OH
f
g
e
O
O
O
O
6m
4
+
OH
O
OH OH
11
13
12
Scheme 2. Synthesis of compounds 6k, 6l, 6m. Reagents and conditions: (a) LAH
(4 equiv), Et2O, 0 °C to rt (88%); (b) same steps as in Scheme 1 (66% for 3 steps); (c)
LAH (4 equiv), Et2O, 0–45 °C (47%); (d) same steps as in Scheme 1 (45% for 3 steps);
(e) pentaerythritol (1.5 equiv), pTsOH (0.05 equiv), benzene/DMF (4:6 v/v) heated
to 80 °C for 30 min, 1-Z-piperidone (1 equiv) in benzene/DMF (4:6 v/v) added,
125 °C for 4 days (67%); (f) nBuLi (1.1 equiv), THF, ꢁ25 °C to rt, 30 min; p-TsCl,
ꢁ20 °C to 0 °C for 1.5 h; nBuLi (1.1 equiv) 0 °C to rt, overnight (40% based on 11%
recovery of starting material); (g) last 2 steps as in Scheme 1 (86% for 2 step).
pentaerythritol.27 The diol 12 was converted to mono-tosylate fol-
lowed by treatment with nBuLi in the same pot to give oxetane 13
in 27% yield.28
Hydroxyl piperidine analogs 6n, 6o, 6p and keto analog 6q were
prepared as shown in Scheme 3. Benzoylation of 1-Cbz-4-piperi-
done 4 at the 3-position was carried out using Tomkinson’s proto-
col29 with N-methyl-O-benzoylhydroxylamine hydrochloride in
DMSO to give 14 in 97% yield. Ketalization of ketone 14 with com-
mercially available 1,1-bis(hydroxymethyl)cyclopropane followed
by hydrogenolysis gave racemic amine 16 in 81% yield over 2 steps.
Coupling of core acid 3 with amine 16 gave compound 17 which
was deprotected to give racemic compound 6n. The racemate 6n
was resolved by chiral HPLC to give enantiomers 6o and 6p. Addi-
tionally, the hydroxyl group in racemate 6n was oxidized by DMP
to give ketone 6q. An alternative method was also developed to
Method A
Cbz
N
H
N
CF3
Cbz
N
O
N
N
N
c
b
g
a
d
4
OBzl
OBzl
6q
O
O
O
O
OBzl
Cl
OR
O
O
O
14
17
: R = Bzl
e
f
15
16
6n
: R = H
6o 6p
+
Method B
H
H
H
N
N+
d, e
-
h
i
CO2
O
R
R
6p
16
O
OBzl
OBzl
O
O
O
O
O
-
O
CO2
2
18
19
Scheme 3. Reagents and conditions: Method A: (a) MeHNOC(O)Ph (1 equiv), DMSO, rt (97%); (b) 1,1-bis(hydroxymethyl)cyclopropane (2.5 equiv), pTsOH (0.05 equiv),
toluene, 140 °C, Dean–Stark apparatus (83%) ; (c) H2 (1 atm), 10% Pd/C Degussa type, MeOH (98%); (d) 3 (1 equiv), 16 (1.4 equiv), T3P (1.5 equiv), Hunig’s base (4 equiv), DMF,
0 °C (97%); (e) LiOH.H2O (3 equiv), THF/MeOH/H2O (92%); (f) separation by chiral column, RegisCell™ ODH (25 cm ꢂ 30 mm), 20% IPA in hexanes, 6o (36%), 6p (39%); (g) DMP
(1.2 equiv), DCM, rt (49%). Method B: (h) di-O,O0-di-p-toluoyl-
D-tartaric acid (1 equiv), EtOH, 50 °C to room temperature (26%); (i) aq NaHCO3, EtOAc (23%, 97% ee); (d) 3
(1 equiv), 19 (1.4 equiv), T3P (1.5 equiv), Hunig’s base (4 equiv), DMF (98%); (e) LiOH (3 equiv), THF, MeOH, water (95%).

V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
2291
Figure 2. An X-ray crystal structure showing the cation and anion of compound 18. Anisotropic atomic displacement ellipsoids for the non-hydrogen atoms are shown at the
20% probability level. Hydrogen atoms are displayed with an arbitrarily small radius.
that the six- and seven-membered rings (6c and 6d) were equipo-
tent and superior to the five-membered ring (6b). The six-mem-
bered ring system was chosen for the development of further
SAR. Simple methyl substitution on the terminal spiro ring (6e) im-
proved the potency on both 1a and 1b replicons, while the di-
methyl and phenyl analogs (6f, 6g) were less active. Compounds
with polar groups, such as hydroxyl- (6h) and keto-groups (6i),
were tolerated. The interesting structural feature of the dispiro
cyclopropyl compound 6j prompted us to further examine other
dispiro analogs. Both the cyclopropyl and cyclobutyl dispiro com-
pounds (6j, 6k) gave excellent potency versus genotype 1a and
1b replicons (EC50 <10 nM). Compound 6j was then tested in a
pharmacokinetic (PK) study in rats, revealing a moderate rate of
clearance (37.8 mL/min/kg) (Table 4). A reactive metabolite was
also detected upon incubation in rat microsomes although it was
not observed in human microsomes.33 We envisioned that addition
of polar functionalities at the amide region may help improve the
PK profile, as previously reported.19 Oxetane compound 6m and
the hydroxyl cyclobutyl analog 6l were tolerated similarly to the
cyclobutyl analog 6k. In a related effort, a polar substituent (hydro-
xyl or ketone) was installed at the 3-position of the piperidine ring,
reminiscent of the carbonyl in piperazinone 2. Racemic-6n and ke-
tone 6q showed similar potency in genotype 1a and 1b replicons.
Racemic-6n was further resolved to give two enantiomers in which
the R-isomer 6p (EC50 1a = 1.5 nM, EC50 1b = 1.2 nM) was similar in
potency to our previously reported lead molecule 2.18 One of the
oxygen atoms in the spiroketals is proposed to act as a hydrogen
bond acceptor, similar to the carbonyl in oxazolidinone as in com-
pound 1 or the carbonyl in piperazinone as in compound 2. Addi-
tion of a R-hydroxy group at the 3-position of the piperidine ring
(6p) may provide an alternative oxygen atom as hydrogen bond
acceptor. The superior HCV 1a potency of compound 2 is attributed
to the hydroxy group on the cyclohexyl ring which is also observed
in hydroxyl cyclobutyl analog 6l.34 Compound 6p was further eval-
uated in rat PK studies (Table 4).
CF3
Boc
N
H
N
O
N
N
N
a
b
Cl
X
X
X
HN
HN
O
O
N
R
O
20a
: X = O
20b: X = CH2
21a
: X = O
21b: X = CH2
c
22b
22c
: X = O, R = Me
d
:X = O, R = Et
e
f
22a
: X = O,
R =H
22d
:X = O, R = CH2CH2OH
:X = O, R = CH2cPr
:X = O, R = Bn
22e
g
22f
22h
c
22g
:X = CH2, R = Me
: X = CH2,
R =H
Scheme 4. Reagents and conditions: (a) HCl in 1,4-dioxane (10 equiv), DCM (21a,
quantitative; 21b, quantitative); (b) (1 equiv), 20a or 21b (1.5 equiv), T3P
3
(1.5 equiv), Hunig’s base (4 equiv), DMF, 0 °C (89% for 22a, 85% for 22g); (c) NaH
(1.5 equiv), MeI (1.5 equiv), DMF, 0 °C (88% for 22b, 85% for 22h); (d) NaH
(1.2 equiv), EtI (1.5 equiv), DMF, 0 °C (79%); (e) (i) NaH (1.2 equiv), TBDMSO(CH2)2-
Br (1.5 equiv), DMF, 0 °C; (ii) TBAF (2 equiv), THF (32%); (f) NaH (1.2 equiv),
cPrCH2Br (1.5 equiv), DMF, 0 °C (75%); (g) NaH (1.2 equiv), BnBr (1.5 equiv), DMF,
0 °C (81%).
compounds 27c and 27d were obtained in a 6-step sequence from
rac-24b.
Compounds were evaluated using subgenomic replicons of
genotypes 1a (HCV 1a) and 1b (HCV 1b).32 As shown in Table 1,
spiro ring attachment at the 4 position of the piperidine amide
was preferred relative to the 3-position (compound 6c vs 6a). A
quick evaluation of the ring size of the spiro system demonstrated

2292
V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
Cbz
N
Cbz
N
Cbz
N
CF3
N
CF3
N
N
O
N
a
b
N
O
N
c - h
14
+
OBzl
OBzl
OBzl
OH
NC OH
OH
NHBoc
OH
NHBoc
OH
Cl
Cl
O
O
23
-24a
-24b
rac
rac
O
O
N
N
Me
Me
c, d
27c
27d
Cbz
N
Cbz
N
Cbz
N
e
+
OBzl
OBzl
OBzl
O
O
O
O
O
O
HN
N
N
Me
N
Me
-25a
rac
26a
26b
f - h
f - h
O
CF3
CF3
N
O
N
N
N
N
OH
OH
Cl
Cl
O
O
O
O
N
Me
N
Me
27a
27b
Scheme 5. Reagents and conditions: (a) KCN (1.2 equiv), NaHCO3 (2 equiv), diethyl ether/water (84%); (b) H2 (50 psi), Pt2O (0.22 equiv), Boc2O (2.5 equiv), iPrOH (22% for rac-
24a, 17% for rac-24b); (c) HCl in 1,4-dioxane (55 equiv), DCM; (d) CDI (6 equiv), DMF (77% for rac-25a, 64% for rac-25b); (e) (i) NaH (1.1 equiv), MeI (1 equiv), DMF, 0 °C; (ii)
chiral separation of 26a and 26b by OJH chiral column (25 cm ꢂ 30 mm), 20% iso-propanol in hexanes (25% for 26a, 25% for 26b), chiral separation of 26c and 26d by ADH
chiral column under SFC condition with 20% MeOH in CO2 (26% for 26c, 28% for 26d); (f) LiOHꢃH2O (3 equiv), THF/MeOH/water; (g) H2 (1 atm), 10 wt% Pd/C (Degussa type)
(0.05 equiv); (h) 3 (1 equiv), spirocyclic amine (1.1), HATU (1.2 equiv), Hunig’s base (4 equiv), DMF (yields for 3 steps: 78% for 27a, 80% for 27b, 68% for 27c, 72% for 27d).
Table 2
improve the PK in the spiro ketal series, a hydroxyl group was
HCV 1a and HCV 1b replicon efficacy of spiro oxazolidinone and spiro lactam series of
incorporated at the 3-position of the piperidine ring. All four dia-
compounds
stereoisomers were prepared and three (27a, 27c, 27d) of them
were tolerated. The most potent diastereoisomer 27c was further
profiled in a rat PK study.
CF3
N
O
N
N
Two sets of compounds (6j and 6p, 22b and 27c) were selected
for further in vitro profiling. In addition to their potent binding
(EC50’s = 10.2–32 nM) to NS4B as measured by displacement of a
radioligand (Table 3), all four compounds showed shifts in replicon
potency versus two stable replicons bearing mutations (H94N and
V105M) previously shown to confer resistance to analogs in a re-
lated chemical series. These data suggest the analogs are interact-
ing directly with NS4B. While these compounds showed excellent
potency in HCV 1a and HCV 1b replicons, the potency against a
genotype 2a replicon remained poor (in the micro-molar range).
In general, this class of inhibitors showed excellent drug-like-
ness, as demonstrated by the potencies and physicochemical prop-
erties such as low clogP, good solubility and permeability
(Table 3). In vivo profiling of the two pairs of compounds suggested
improved PK properties with the addition of the hydroxyl group at
the 3-position of the piperidine ring (Table 4). In the spiro ketal
series (6j vs 6p), addition of the hydroxyl group in compound 6p
had little effect on the moderate clearance of compound 6j after
IV administration to rats, but it resulted in significant improve-
ment of the oral exposure (greater than 4-fold improvement in
Cmax and AUC). In addition, the reactive metabolite observed in
compound 6j was not observed in compound 6p. In the spiro oxa-
zolidinone series (22b vs 27c), addition of hydroxyl group led to
improvement in clearance (ꢀ2ꢂ) while oral exposure only in-
creased by 1.5-fold.
R1
X
6
Cl
5
N
R
O
Compound
R
Stereo-
chemistry
X
R1
EC50 1a
(nM)
EC50 1b
(nM)
22a
22b
22c
22d
22e
22f
22g
22h
27a
27b
27c
27d
H
Me
Et
CH2CH2OH
CH2cPr
Bn
H
Me
Me
Me
Me
Me
O
O
O
O
O
O
CH2
CH2
O
H
H
H
H
H
H
H
H
166
2.0
55.4
3.4
5.9
78.5
23.4
457
1575
69.2
7.0
20.6
23.2
18.1
427
441
184
(5R,5R)
(5S,6S)
(5R,6S)
(5S,6R)
OH 14.5
OH 3558
OH 10.9
OH 16.7
O
O
O
3992
6.1
7.2
In the spiro [6.5] series, an initial assessment of alkyl substitu-
tion on the lead compound 22a identified the N-methyl derivative
22b (EC50 1a = 2 nM, EC50 1b = 3.4 nM) as the most potent com-
pound in the spiro oxazolidinone series (Table 2). SAR revealed that
the methyl substitution appears optimal. Replacement of the oxa-
zolidinone ring with a lactam ring (O ? C) resulted in significant
loss of potency, in particular for compound 22h. This is consistent
with previous amide modifications that demonstrate the need for
at least one oxygen atom between the piperidine and the adjacent
ring.18,19 Unfortunately, metabolite ID studies of 22b showed a
significant amount of N-demethylation. In an analogous effort to
In summary, two new series of spiro piperidine analogs were
rationally designed which inhibit HCV replication with EC50s
<11 nM for genotypes 1a and 1b replicons. However, the emer-
gence of more potent analogs from other series18 and the low
HCV 2a potency for these compounds has put these series on hold.

V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
2293
Table 3
7. Bartenschlager, R.; Frese, M.; Pietschmann, T. Adv. Virus Res. 2004, 63, 71.
8. Chevaliez, S.; Pawlotsky, J.-M. Horizon Biosci. 2006, 5.
In vitro and physical properties of matched compound pairs
9. Sofia, M. J.; Chang, W.; Furman, P. A.; Mosley, R. T.; Ross, B. S. J. Med. Chem.
2012, 55, 2481.
Compound
6j
6p
22b
27c
10. Yang, P. L.; Gao, M.; Lin, K.; Liu, Q.; Villareal, V. A. Curr. Opin. Virol. 2011, 1, 607.
11. Hepatitis C Viruses; Sklan, E. H., Glenn, J. S., Eds.; Horizon Bioscience, 2006. p.
245.
EC50 1a (nM)
EC50 1b (nM)
EC50 1b H94N (nM)
(fold shift)
EC50 1b V105M (nM)
(fold shift)
9
4.1
1.5
1.2
2
3.4
10.9
6.1
175
(27.3ꢂ)
1,128
(185ꢂ)
38,019
30.4
310
N/A
472.9
123
132
2.88
87
24.3
(5.9ꢂ)
67.6
(16.5ꢂ)
6,839
32
105
432
469.9
30
35.9
(29.9ꢂ)
171.8
(143ꢂ)
11,253
10.2
323
407
485.9
50
42.7
(12.5ꢂ)
20.4
(6ꢂ)
6,839
23.4
267
N/A
456.9
66
12. Rai, R.; Deval, J. Antiviral Res. 2011, 90, 93.
13. Gouttenoire, J.; Penin, F.; Moradpour, D. Rev. Med. Virol. 2010, 20, 117.
14. Li, S.; Yu, X.; Guo, Y.; Kong, L. Cell. Microbiol. 2012, 14, 994.
15. Gu, Z.; Graci, J. D.; Lahser, F. C.; Breslin, J. J.; Jung, S. P.; Crona, J. H.; McMonagle,
P.; Xia, E.; Liu, S.; Karp, G.; Zhu, J.; Huang, S.; Nomeir, A.; Weetall, M.; Almstead,
N. G.; Peltz, S. W.; Tong, X.; Ralston, R.; Colacino, J. M. Antimicrob. Agents
Chemother. 2013, 57, 3250.
16. Zhang, X.; Zhang, N.; Chen, G.; Turpoff, A.; Ren, H.; Takasugi, J.; Morrill, C.; Zhu,
J.; Li, C.; Lennox, W.; Paget, S.; Liu, Y.; Almstead, N.; George, N.; Gu, Z.;
Komatsu, T.; Clausen, V.; Espiritu, C.; Graci, J.; Colacino, J.; Lahser, F.; Risher, N.;
Weetall, M.; Nomeir, A.; Karp, G. M. Bioorg. Med. Chem. Lett. 2013, 23, 3947.
17. Banka, A.; Baskaran, S.; Catalano, J.; Chong, P.; Dickson, H.; Fang, J.; Maung, J.;
Neitzel, M. L.; Peat, A.; Price, D.; Rai, R.; Roberts, C. D.; Shotwell, B.; Tai, V.;
Zhang, H. WO 2010091411A1.
18. Miller, J. F.; Chong, P. Y.; Shotwell, J. B.; Catalano, J. G.; Tai, V. W.-F.; Fang, J.;
Banka, A. L.; Roberts, C. D.; Youngman, M.; Zhang, H.; Xiong, Z.; Mathis, A.;
Pouliot, J. J.; Hamatake, R. K.; Price, D. J.; Seal, J. W.; Stroup, L. L.; Creech, K. L.;
Carballo, L. H.; Todd, D.; Spaltenstein, A.; Furst, S.; Hong, Z.; Peat, A. J. J. Med.
Chem. 2014, 57, 2107.
19. Shotwell, J. B.; Baskaran, S.; Chong, P.; Creech, K. L.; Crosby, R. M.; Dickson, H.;
Fang, J.; Garrido, D.; Mathis, A.; Maung, J.; Parks, D. J.; Pouliot, J. J.; Price, D. J.;
Rai, R.; Seal, J. W.; Schmitz, U.; Tai, V. W. F.; Thomson, M.; Xie, M.; Xiong, Z. Z.;
Peat, A. J. ACS Med Chem. Lett. 2012, 3, 565.
EC50 2a (nM)
IC50 NS4B 1b (nM)a
Solubility (
l
g/mL)b
Papp MDCK (nm/s)c
MW
T
T
human (min)d
rat (min)d
½
<15
2.96
56
28
2.68
76
80
2.58
67
½
cLogPe
TPSAf
g
LLEcalc
4.5
5.3
5.0
4.6
NS4B binding assay as measured by displacement of a radioligand.19
Kinetic solubility measured at pH 7.4.35
Passive Papp is defined as the A ? B Papp value in the presence of P-gp inhibitor
a
b
c
GF 120918.
Half-life of compounds in human and rat hepatocytes.18
d
e
cLogPs are calculated using the Daylight/BioByte v4.3 algorithm, BioByte Corp.,
Claremont, CA.
Topological polar surface areas (TPSAs) are calculated using Ertl’s method.36
f
20. Phillips, B.; Cai, R.; Delaney, W.; Du, Z.; Ji, M.; Jin, H.; Lee, J.; Li, J.; Niedziela-
Majka, A.; Mish, M.; Pyun, H.-J.; Saugier, J.; Tirunagari, N.; Wang, J.; Yang, H.;
Wu, Q.; Sheng, C.; Zonte, C. J. Med. Chem. 2014, 57, 2161.
g
Lipophilicity ligand efficiency (LLE) is defines as pIC50 (or pKi) – clogP (or
logD).37
21. Kakarla, R.; Liu, J.; Naduthambi, D.; Chang, W.; Mosley, R. T.; Bao, D.; Steuer, H.
M. M.; Keilman, M.; Bansal, S.; Lam, A. M.; Seibel, W.; Neilson, S.; Furman, P. A.;
Sofia, M. J. J. Med. Chem. 2014, 57, 2136.
22. Einav, S.; Glenn, J. S.; McDowell, R.; Yang, W. WO 2009039248A2.
23. X-ray crystallographic data of compounds 1 and 18 can be found at Cambridge
Structural Database (CSD), the deposition numbers are CCDC 970777 and CCDC
970778, respectively.
Table 4
In vivo rat DMPK profile of compounds 6j, 6p, 22b and 27d
Compound
6j
6p
22b
27c
24. Perron, F.; Albizati, K. F. Chem. Rev. 1989, 89, 1617.
IV Clearancea,b
37.8
4.5
254
693
29
35.7
0.8
1134
2792
>100
56.5
2.1
269
1059
72
25.7
2.0
344
1528
49
25. Wissmann, H.; Kleiner, H. J. Angew. Chem. 1980, 92, 129.
26. Tavares, F. X.; Deaton, D. N.; Miller, L. R.; Wright, L. L. J. Med. Chem. 2004, 47,
5057.
27. Murguia, M. C.; Vaillard, S. E.; Grau, R. J. Synthesis 2001, 1093.
28. Loy, R. N.; Jacobsen, E. N. J. Am. Chem. Soc. 2009, 131, 2786.
29. Beshara, C. S.; Hall, A.; Jenkins, R. L.; Jones, K. L.; Jones, T. C.; Killeen, N. M.;
Taylor, P. H.; Thomas, S. P.; Tomkinson, N. C. O. Org. Lett. 2005, 7, 5729.
30. Palovics, E.; Faigl, F.; Fogassy, E. InTech 2012, 3.
31. The corresponding six membered 1,3-oxazinan-2-one analogs are less active.
32. Detailed description of the HCV replicon assay and NS4B binding assay can be
found in Ref. 19. The replicon data represents the average of experimental data
with n = 2 or more.
33. A proposed mechanism for glutathione conjugation of compound 6j is drawn
below. The addition of hydroxyl group in compound 6p may have altered the
routes of metabolism.
Vdss
Cmax
c
PO AUC 0–24
%F
a
All studies were conducted in accordance with the GSK Policy on the Care,
Welfare and Treatment of Laboratory Animals and were reviewed the Institutional
Animal Care and Use Committee either at GSK or by the ethical review process at
the institution where the work was performed.
b
Dosed as a solution of DMSO/20%HPbCD (10:90); IV dosed 1 mg/kg; PO dosed
5 mg/kg; clearance (mL/min/kg); Vdss (L/kg); Cmax (ng/mL); PO AUC 0–24 (hꢄng/mL).
c
Concentrations of all
4 compounds taken at 24 h were below level of
quantitation.
CF3
CF3
The unique spirocyclic amines, which were developed through
modeling studies, offer new fragments that target different chem-
ical space for future drug discovery.
O
N
N
hydrolysis
[H]
O
N
N
N
N
Cl
Cl
O
Acknowledgments
O
OH
CF3
6j
The authors acknowledge the contributions of the following
GSK scientists to the manuscript: Christine Talarico (permeability),
Rachel Schmidt (chiral purification), Douglas Minick (vibrational
circular dichroism spectroscopy), and Brad Shotwell (review of
manuscript).
O
N
N
N
HO
O
P450 [O]
GSH
Cl
S
OH
H2N
HO2C
O
N
H
References and notes
NH
1. Davis, G. L.; Albright, J. E.; Cook, S. F.; Rosenberg, D. M. Liver Transplant. Off Publ.
Am. Assoc. Study Liver Dis. Int. Liver Transplant. Soc. 2003, 9, 331.
2. Armstrong, G. L.; Wasley, A.; Simard, E. P.; McQuillan, G. M.; Kuhnert, W. L.;
Alter, M. J. Ann. Intern. Med. 2006, 144, 705.
CO2H
3. World Health Organization fact sheet 164, July 2013, http://www.who.int/
mediacentre/factsheets/fs164/en/index.html.
4. Kwong, A. D.; Kauffman, R. S.; Hurter, P.; Mueller, P. Nat. Biotechnol. 2011, 29,
993.
5. Venkatraman, S. Trends Pharmacol. Sci. 2012, 33, 289.
6. Chen, K. X.; Njoroge, F. G. Prog. Med. Chem. 2010, 49, 1.
34. We previously19 showed that addition of hydroxyl group to the cyclohexyl ring
improves HCV 1a potency as demonstrated by the following pair of compounds
with imidazopyridine core. The corresponding cyclohexyl analog for
compound 2 has not been made.

2294
V. W.-F. Tai et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2288–2294
CF3
CF3
35. Hill, A. P.; Young, D. J. Drug Discovery Today 2010, 15, 648.
36. Ertl, P.; Rohde, B.; Selzer, P. J. Med. Chem. 2000, 43, 3714.
37. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881.
O
N
O
N
N
N
N
N
O
O
O
O
Cl
Cl
N
N
OH
EC50 HCV 1a = 11 nM
EC50 HCV 1b = 3.5 nM
EC50 HCV 1a = 0.9 nM
EC50 HCV 1b = 9.7 nM