Design, synthesis and biological evaluation of novel aminothiazoles as antiviral compounds acting against human rhinovirus

Article abstract of DOI:10.1016/j.bmcl.2013.04.077

We describe here the design, synthesis and biological evaluation of antiviral compounds acting against human rhinovirus (HRV). A series of aminothiazoles demonstrated pan-activity against the HRV genotypes screened and productive structure-activity relationships. A comprehensive investigational library was designed and performed allowing the identification of potent compounds with lower molecular weight and improved ADME profile. 31d-1, 31d-2, 31f showed good exposures in CD-1 mice. The mechanism of action was discovered to be a host target: the lipid kinase phosphatidylinositol 4-kinase III beta (PI4KIII?). The identification of the pan-HRV active compound 31f combined with a structurally distinct literature compound T-00127-HEV1 allowed the assessment of target related tolerability of inhibiting this kinase for a short period of time in order to prevent HRV replication.

Full text of DOI:10.1016/j.bmcl.2013.04.077

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3841–3847
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and biological evaluation of novel aminothiazoles
as antiviral compounds acting against human rhinovirus
Anne Décor a, , Chantal Grand-Maître a, Oliver Hucke a, ,à, Jeff O’Meara a,§, Cyrille Kuhn a,–
,
⇑
Léa Constantineau -Forget b, Christian Brochu a, Eric Malenfant a, Mégan Bertrand-Laperle a,k
Josée Bordeleau a, Elise Ghiro a, Marc Pesant a, Gulrez Fazal a, Vida Gorys a, Michael Little a,
,
Colette Boucher a, Sylvain Bordeleau a, Pascal Turcotte a, Tim Guo a, Michel Garneau a, Catherine Spickler a,
Annick Gauthier a
a Boehringer Ingelheim (Canada) Ltd., Research and Development, 2100 Cunard Street, Laval, Québec, Canada H7S 2G5
b Université de Montréal, Department of Chemistry, PO Box 6128, Montréal, Québec, Canada H3C 3J7
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe here the design, synthesis and biological evaluation of antiviral compounds acting against
human rhinovirus (HRV). A series of aminothiazoles demonstrated pan-activity against the HRV geno-
types screened and productive structure–activity relationships. A comprehensive investigational library
was designed and performed allowing the identification of potent compounds with lower molecular
weight and improved ADME profile. 31d-1, 31d-2, 31f showed good exposures in CD-1 mice. The mech-
anism of action was discovered to be a host target: the lipid kinase phosphatidylinositol 4-kinase III beta
(PI4KIIIß). The identification of the pan-HRV active compound 31f combined with a structurally distinct
literature compound T-00127-HEV1 allowed the assessment of target related tolerability of inhibiting
this kinase for a short period of time in order to prevent HRV replication.
Received 18 February 2013
Revised 25 April 2013
Accepted 29 April 2013
Available online 7 May 2013
Keywords:
Human rhinovirus
HRV
Library design
Aminothiazoles
Design of experiment
DOE
Ó 2013 Elsevier Ltd. All rights reserved.
PI4KIIIß
Human rhinovirus (HRV) is a major cause of the common cold
and is associated with being a trigger for asthma and chronic
obstructive pulmonary disease (COPD) exacerbations in children
and adults.1,2 There are no approved drugs for the treatment of
HRV infections to date. Therefore there is a high medical need to
identify a treatment for HRV infection to prevent disease exacerba-
tion in patients with asthma and COPD. Preliminary results in a
trial to prevent asthma exacerbations using the capsid-binding
inhibitor BTA-798 trended towards being positive, however it has
been reported that it is not effective against all HRV genotypes.3,4
HRV is a non-enveloped single positive-stranded RNA virus belong-
ing to the Picornaviridae family (Enterovirus genus) with more than
133 genotypes, divided into three species A, B and C, based on
sequencing the complete genomes.5,6 HRV-A and HRV-C are more
likely to cause moderate to severe illness than HRV-B.7
Identifying compounds that had broad anti-rhinoviral activity
against species and genotypes was a key objective of our explora-
tion. In order to find novel HRV pan-active inhibitors of viral repli-
cation, a phenotypic antiviral screening strategy was chosen. A
subset of 98K compounds from the Boehringer Ingelheim collec-
tion was screened.8 The hits underwent profiling and prospecting
to eliminate those with demonstrated cytotoxicity,9 lack of selec-
tivity against other virus (HIV, CMV, HCV), lack of potential for
pan-HRV genotype activity and lack of productive structure–activ-
ity relationships (SAR). A set of 20 best HRV hits was obtained from
the 98K compounds screen. Preliminary mechanisms of action
studies were performed mainly to rule out capsid-binding inhibi-
tors due to lack of genotype coverage.10 The hit triage allowed at
the end to identify one hit: compound 1 (Table 1).
⇑
Corresponding author. Tel.: +49 7351 54 99462; fax: +49 7351 83 99462.
E-mail address: oliver.hucke@boehringer-ingelheim.com (O. Hucke).
Current address: Bayer, Small Molecules Research, Gebäude 6510, Alfred Nobel
Strasse 50, 40789 Monheim, Germany.
à
Current address: Boehringer Ingelheim Pharma GmbH & Co. KG, Birkendorferst-
rasse 65, 88397 Biberach, Germany.
§
Current address: Alphora Research Inc., 2395 Speakman Drive, Suite 2001,
Mississauga, Ontario, Canada L5K 1B3.
–
Current address: Boehringer Ingelheim (USA), 900 Ridgebury Rd, Ridgefield, CT
This compound is quite attractive from a medicinal chemistry
perspective in that it is very modular and offers many potential
sites at which to carry out SAR. However, it contains an aniline
06877, USA.
k
Current address: Fabbrica Italiana Sintetici, 9321, Airlie Street, LaSalle, Québec,
Canada H8R 2B.
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.04.077

3842
A. Décor et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3841–3847
Table 1
Activity and ADME and physicochemical properties of compounds 1–6
Ph
Ph
R1
4
O
N
R2a
R2b
N
N
R2d
O
N
O
N
O
H2N
O
S
O
NH
X
N
Ph
R2
N
R2c
a
b
Compds
X
R1 R2 EC50 (nM)
CC50
Sol pH 2/6.8 LogD Caco-2c HLMd% CYP
(
l
M)
(
lg/mL)
pH
7.4
(10ꢀ6
cm/s)
QH
2D6/3A4
g)
(l
HRV 14 HRV 16 HRV 1a HRV 2 HRV 8 HRV 54 HRV 89 HRV 45 HRV 5
Lead
criteria
<1000
>25
>20 ꢁ EC50 >25
0–5
>0.5
<75
>1
1
2
3
4
5
6
C
N
N
C
C
C
H
H
R2a
R2a 1650
410
Me R2b 990
640
630
1350
290
905
1400
1230
1600
1200
960
930
1200
1240
790
510
na
470
915
945
210
300
260
140
130
590
1300
280
400
530
395
890
790
1700
610
680
890
1200
630
350
360
430
720
660
600
430
430
535
490
>65
>65
46
>479/<0.2 2.9
1000/3
>987/125
11.3
2.3
2.1
19
21
17
66
45
42
<25
74
47
nae/0.5
2.6/1.8
4.3/0.71
25/>30
na/0.98
>30/15
2.6
2.8
Me R2a
>65
>65
>64
>814/2.5 2.5
58/<0.1 3.2
>1000/<0.1 3.4
H
H
R2c
R2d 1080
220
1100
a
Anti-HRV activity is evaluated in a cell-culture based assay (viral replication assay) that measures the ability of compounds to protect from the cytopathic effect (CPE)
caused by HRV replication.
b
Cytotoxicity on H1-Hela cells, targeted window CC50 > 20 ꢁ EC50
.
c
Caco-2 A to B, permeability on a model using cultured polarized monolayer of human epithelial colorectal adenocarcinoma cells.
HLM: human liver microsome: in vitro metabolic clearance expressed as % of estimated human hepatic blood flow of 21 mL/min/kg.
Na: not available.
d
e
structural alert that would have to be removed.11 As shown in
Table 1, compound 1 is pan-active against the screened genotypes,
The objectives of this study were to identify the key pharmaco-
phores, complete the SAR, improve potency and ADME profile, and
to try to identify replacements for the aniline. The elucidation of
the mode of action and determination of the biotarget is described
in detail elsewhere.8 Primary SAR studies of compound 1, have
shown that the aniline can be replaced with the aminopyridine
(compound 2). This product demonstrated pan-activity within
approximatively two fold of compound 1. Additional profiling
showed significant improvement in solubility at pH 2.0 and 6.8
over the reference compound 1. Additional benefits include lower
lipophilicity and improved human liver microsome (HLM) stability.
and has
(CC50 > 65
a
l
good window with respect to the cytotoxicity
M). At the time the SAR expansion started, the mode
of action of compound 1 was not known. Initial studies showed
that the series acts later in the infectious cycle of HRV at the level
of RNA replication, but it was found not to be active against HRV
3Dpol or HRV 3C protease.8 Initial in vitro ADME profiling on com-
pound 1 highlights a good Caco-2 permeability and logD, whereas
the improvement of the metabolic stability and CYP 3A4 inhibition
profile was desirable.
Ph
R3 O
O
HO
O
N
O
Ph
b
OEt
+
12, R3=Et
13, R3=H
NH
c
O
a
N
N
10
11
R1
N
R1
H2N
NH
Ph
O
S
HN
NH
N
+
d
H2N
N
O
for R1=H
S
NH2
Cl
8, R1=H or Me
S
O
NH2
9, R1=H or Me
1
7
Ph
N
HOOC
14, R5=Boc
N
R5
N
e
15, R5=Boc
16, R5=H
6, R5=
H2N
f
g
Ph
R5
for R1=H
S
NH
O
N
N
O
COOH
17
N
Scheme 1. Synthesis of compounds 1 and 6. Reagents and conditions: (a) for R1 = H, H2O/HCl 1 N (83/17), 100 °C, overnight; for R1 = Me, MeOH/HCl 1 N (83/17), 40 °C,
overnight, both used as a crude; (b) DMF, HATU, DIPEA, 30 min, amine, room temperature, overnight, 68%; (c) DMSO, NaOH 10 N, room temperature, overnight, 40%; (d)
MeCN, HATU, DIPEA, room temperature 1 h, 22%; (e) MeCN, HATU, DIPEA, used as crude; (f) DCM, TFA, overnight, 60%; (g) DMF, HATU, DIPEA, carboxylic acid 17 room
temperature.

A. Décor et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3841–3847
3843
A similar CYP 3A4 inhibitory activity was observed and the Caco-2
permeability is still in the targeted window. Introduction of a
methyl group R1 in position 4 of the aminothiazole heterocycle
(compound 3) resulted in an improvement in potency on the stud-
ied genotypes’ panel with good solubility, logD and HLM stability.
Having substituents in position 4 and 5 is also a way known to pre-
vent metabolic activation. Less conservative SAR were also ex-
plored. Compounds 4 and 5 with a simplified, rigidified right
handside (RHS) maintained potency comparable to the hit com-
pound 1. Compound 4 addresses the CYP3A4 inhibition observed
with the original hit while maintaining a very good permeability.
Initial SAR studies on the hit compound 1 have shown that the
aniline and the aminothiazole are key pharmacophores, and that
these parts are very sensitive to minor changes.12 However, the
aniline can be replaced by the aminopyridine, and introduction
of the methyl in position 4 improved potency.13 Compound 4
and 5 have shown that the RHS of the molecule tolerates signifi-
cant structure modifications. This finding prompted us to search
for new compounds with novel RHS structures with lower molec-
ular weight, improved potency and ADME properties, and a com-
prehensive investigational library was designed and synthesized
for this purpose.
The general synthesis procedure of designed compounds is
illustrated in Schemes 1–3. In order to prepare the aniline com-
pounds 1, 4–6, the key scaffold 9 was obtained by heating reagents
7 and 8 (with R1 = Me or H) using HCl 1 N.14 Carboxylic acid 13 was
prepared by amide coupling of 10 with 11 followed by ester hydro-
lysis. Selective amide formation on the aminothiazole group of
scaffold 9 (with R1 = H) using carboxylic acids 13 or 14 in acetoni-
trile, HATU and DIPEA provided compound 1 or compound 15.
After removal of the protecting Boc group and coupling with acid
17, compound 16 led to analog 6. As shown in Scheme 2, selective
amide formation on the aminothiazole group of 9 led to intermedi-
ate 19 using carboxylic acid 18. Removal of the Boc group in the
left-hand side and acylation provided analog 5. For the synthesis
of analog 4, the scaffold 20, which was protected by a Boc group
on the aniline (R1 = Me) was used as well as the carboxylic acid
24. Nucleophilic substitution of 21 with 22 followed by ester
hydrolysis provided intermediate 24. Analog 4 was prepared from
20 by amide coupling followed by Boc deprotection of the aniline.
For the aminopyridine analogs (such as compound 2, 3 and the
ones from the library), reagent 26 is protected with a Boc group,
leading to 27 (Scheme 3). A Suzuki coupling is performed between
intermediate 27 and the boronate ester 25. The key intermediate
R1
Boc
R1
Boc
N
Ph
18
HOOC
N
Ph
Ph
19, R1=H
N
N
R2=
H2N
a
H2N
S
O
NH
f
S
NH2
N
R2
5, R1=H
9, R1=H or Me
O
R2=R2c
e
iPr
4, R1=Me
R2=R2b
O
O
R1
N
b for R1=Me
iPr
N
BocHN
iPr
HN
S
NH2
4O
R
OMe
O
c
O
N
20
Br
+
O
O
O
O
21
22
23, R4=Me
d
24, R4=H
Scheme 2. Synthesis of compounds 4 and 5. Reagents and conditions: (a) MeCN, HATU, DIPEA, room temperature 2 h, 59%; (b) DMSO, Et3N, Boc2O, room temperature,
overnight, 42%; (c) THF, 0 °C, NaH, 0 °C for 5 min, room temp for 30 min, then electrophile, room temperature overnight, 33%; (d) THF/ MeOH (1/1), NaOH 1 N, room
temperature overweekend, 53%; (e) MeCN, HATU, DIPEA, room temperature overnight then, concentration to dryness, HCl 4 N/dioxane, 1 h, concentration to dryness 49%; (f)
HCl 4 N/dioxane, 1 h, concentration to dryness, then MeCN, HATU, DIPEA, CH3CO2H, room temperature overnight, 26%.
1
R
1
R
N
1
R
H2N
O
O
b
d
N
S
+
Br
N
NH
N
H2N
B
H2N
c
5
O
R
S
N
5
N
H
R
N
S
2
O
2
N
H
R
2
HO
R
5
30
5
1
2a
25
28, R =Boc
26, R =H
2, R =H, R = R
a
5
5
29, R =H
27, R =Boc
1
2
2a
3, R =Me, R = R
1
2
2e
2f
31a, R =Me, R = R
31c, R2g
31d, R2h
31e, R2i
31b, R =Me, R = R
1
2
31a, R2e
OMe
CF3
O
N
1
2
2g
2h
31c, R =Me, R = R
1
2
Ph
31d, R =Me, R = R
31b, R2f
1
2
2i
31e, R =Me, R = R
CF3
Scheme 3. Synthesis of compounds 2, 3 and 31a to e. Reagents and conditions: (a) for R1 = H, DCM, Boc2O, Et3N, DMAP, room temperature, overnight, 73%; for R1 = Me, DCM,
Boc2O, Et3N, DMAP, room temperature, 6 h., 66%; (b) for R1 = H, Dioxane (0.07 M), H2O (0.5 M), 27 (1.5 equiv), Na2CO3 (3 equiv), Pd(PPh3)4 (0.2 equiv), 15 min,
lw-high,
115 °C, 46%; for R1 = Me, DMF (0.06 M), H2O (0.6 M), 27 (1.5 equiv), Ba(OH)2 (3 equiv), Pd(PPh3)4 (0.2 equiv), 20 min, MW-high, 115 °C, 62%; (c) for R1 = H, TFA/DCM(1/1)
(0.17 M), room temperature, 1.5 h, 100%; for R1 = Me, HCl 4 N/dioxane (0.2 M), 65 °C, 3 h, 97%; (d) for compound 2 and 3, DMF, HATU, DIPEA, room temperature overnight,
34%; for compounds 31a–e, DCM, HATU, DIPEA, room temperature overnight, 19–66%.

Table 2
Activity and ADME and physicochemical properties of best 3D similar and diverse compounds from the comprehensive investigational library (31) compared to compound 1
N
N
MeO
MeO
31a, R2e
31d, R2h
H2N
N
N
Ph
H2N
S
NH
OMe
CF3
S
NH
N
H
N
N
O
N
O
O
R2
N
31e, R2i
31b, R2f
31c, R2g
CF3
O
N
1
T-00127-HEV1
31
O
N
Ph
a
b
c
d
Compd
EC50 (nM)
PI4K IIIb
(nM)
PI4K III
(nM)
a
CC50
sol pH 2/6.8
logD pH
7.4
Caco-2
HLM
QH
%
CYP 3A4
(lM)
(l
M)
(lg/mL)
10ꢀ6 cm/s)
HRV-14
HRV-16
HRV-1a
HRV-2
HRV-8
HRV-54
HRV-89
HRV-45
HRV-5
1
640
500
290
325
450
730
630
285
140
145
450
795
1600
680
300
340
390
360
790
710
295
335
390
430
210
160
125
84
63
190
1300
360
235
235
320
430
790
295
200
190
370
330
1200
200
135
na
390
320
660
nae
145
130
180
490
100
22
53
40
35
67
>10,000
>10,000
>10,000
3600
2800
4500
>64
22
33
39
24
479/<0.2
>1000/14
>1000/1.1
>1000/8.7
>1000/16
57/<0.1
2.9
2.8
3.8
3.8
2.9
4.0
11.3
18
22
20
20
66
<24
34
76
31
48
0.5
>30
>30
17
>30
7.7
31a (R2e
31b R2f
31c R2g
31d R2h
31e R2i
)
>64
26
a
b
c
Anti-HRV activity is evaluated in a cell-culture based assay (viral replication assay) that measures the ability of compounds to protect from the cytopathic effect (CPE) caused by HRV replication.
Cytotoxicity on H1-Hela cells, targeted window CC50 > 20 ꢁ EC50
.
Caco-2 A to B, permeability on a model using cultured polarized monolayer of human epithelial colorectal adenocarcinoma cells.
HLM: human liver microsome: in vitro metabolic clearance expressed as % of estimated human hepatic blood flow of 21 mL/min/kg.
Na: not available.
d
e

A. Décor et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3841–3847
3845
Table 3
Rat PK profilea of compounds 31a and 31b versus T-00127-HEV1
Compd
EC50 (nM) HRV-14 EC50 (nM) HRV-16 CC50 (nM) PI4KIIIb (nM) Cmax
(l
M) AUC (
l
M ꢁ h) %F
77
160b
Cltot (mL/min/kg) T1/2 (h) Vss
31a
31d
500
450
285
450
22,000
24,000
22
35
48
1.5
1.9
2.0
2.65
9.7
10.0
70
40
0.38
1.7
1.88
4.2
4.4
T-00127-HEV1c 2300
1800
>65,000
91 18
2.8
a
b
c
Rat PK at 5 mg/kg po and at 2 mg/kg iv.
>100 due to large extrapolation of the AUC po at 5 mg/kg and from AUC iv at 2 mg/kg.
For T-00127-HEV1, sol (at pH 2/6.8) = 1000 and 42
l
g/mL; logD (at pH 7.4) = 2.7; Caco-2 = 27 ꢁ 10ꢀ6 cm/s; HLM QH < 25%; IC50 CYP 3A4 > 30
lM.
Table 4
Activity on enteroviruses and on a panel of kinases of compounds 31d-1, 31d-2, 31f and T-00127-HEV1
N
MeO
MeO
31d-1
31f
N
H2N
CF3
CF3
CF3
S
NH
N
N
H
N
N
O
R2
N
31d-2
31
T-00127-HEV1
O
Compds
EC50 (nM)
EC50 HRV-14b
PI4K PI4K
IIIb III
(nM) (nM)
Lipid
kinase
specifi city specificity
Protein
kinase
a
HRV-B14 HRV-A16 HRV-C15 HRV-14b HRV-16b EV-71 Echo Polio CV-B4 CV-B3 CC50
replicona replicona replicona
virus
(lM)
31d-1
31d-2
31-f
310
83
120
115
33
44
260
45
72
995
290
470
1200
360
190
1300
240
170
450
170
<23
860
1100 380
280 61
130 120
1100
520
280
>64
>66
>66
>65
225
53
48
>10,000
0/8
2/277c
3/277e
2/277g
0/277
2750 3/8d
>10,000 2/8f
>10,000
T-00127-HEV1 590
250
200
2300
1800
4400
1200 720
4900
137
0/8
a
Replicon assay, measuring inhibition of the RNA replication of HRV via a luciferase reporter replicon. HRV-B14 and HRV-A16 are subgenomic reporter replicons and HRV-
C15 is a full-length reporter replicon.
b
Cell-culture based assay (viral replication assay) that measures the ability of compounds to protect from the cytopathic effect (CPE)caused by HRV replication.
c
STK17A (IC50 = 2.9
l
l
M); MAP4K2 (IC50 = 9.7
M), PIK3C2B (IC50 = 8.9
M), DYRK3 (IC50 = 10 M), MAP4K4 (IC50 = 8.0
M), PIK3C2B (IC50 = 5.6 M).
M), MSK2 (IC50 = 4.6 M).
l
l
M).
d
e
f
PIK3C2A (IC50 = 4.8
CDK9 (IC50 = 5.3
PIK3C2A (IC50=12
M), PIK3CD (IC50 = 11
l
M).
M).
l
l
l
l
l
g
STK17A (IC50 = 2.1
l
l
28 is deprotected. The scaffold 29 offers two sites for the amide for-
mation: one at the aminothiazole and the other one at the amino-
pyridine. Selective amide formation on the aminothiazole group,
was obtained using HATU, DIPEA and dichloromethane, solvent
in which the scaffold is poorly soluble.15
with the software MODDE. Three variables were studied: the palla-
dium catalyst [Pd(PPh3)4 or Pd(dppf)2Cl2 or Pd(OAc)2 with the
DavePhos ligand], the base [Na2CO3, Cs2CO3, Ba(OH)2, tBuOK or
K3PO4] and the solvent [dioxane or DMF or a mixture DME-MeOH].
Running 18 experiments planned with the DOE software MODDE
allowed us to improve the reaction yield from 42% to 62% (see Sup-
plementary data Fig. S1 for the table with the 18 experiments of
the DOE). The best results were found using barium hydroxide as
The synthesis of the building block 28 with R1 = Me, necessary
for the comprehensive investigational library has been optimized,
using for the Suzuki step a design of experiment (DOE) approach
Table 5
Profiling of compounds 31d-1, 31d-2, 31f and T-00127-HEV1: ADME assays and dose escalating PK in CD-1 mice
Compds
EC50 (nM)
Sol pH
2/6.8
(lg/mL)
LogD
pH 7.4
Caco-2b
H LMc%
QH
CYP 3A4
M)
Cmax
(
l
M)
AUC_ (
l
M ꢁ h)
Cltot
(mL/min/kg)
iv at 2 mg/kg
(10ꢀ6 cm/s)
(
l
at 5 mg/kg
50 mg/kg
250 mg/kg
at 5 mg/kg
50 mg/kg
250 mg/kg
HRV-B14
replicona
HRV-A16
replicona
HRV-C15
replicona
31d-1
31d-2
31-f
T-00127-HEV1
310
83
120
590
115
33
44
260
45
72
1000/82
>1000/51
>1000/13
1000/2
3.2
3.2
3.4
2.7
44
42
41
11
<24
48
27
>30
>30
17
1.8 18 55
2.2 20 26
0.5 12 59
1.9 15 52
2.7 44 176
4.0 38 12
208 514 732
3.6 45 307
91
71
58
250
200
25
>30
ndd
a
b
c
Replicon assay, measuring inhibition of the RNA replication of HRV via a luciferase reporter replicon.
Caco-2 A to B, permeability on a model using cultured polarized monolayer of human epithelial colorectal adenocarcinoma cells.
HLM: human liver microsome: in vitro metabolic clearance expressed as % of estimated human hepatic blood flow of 21 mL/min/kg.
Nd: not done.
d

3846
A. Décor et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3841–3847
a base, Pd(PPh3)4 for the catalyst, and DMF as the solvent. The reac-
tion was performed successfully on a 7g scale.
profile (for potency and CD-1 mouse PK at 5 mg/kg) seen with
the best enantiomer 31d-2.
For the design of the comprehensive investigational library of
the RHS variations (Scheme 3, compounds 31) 3D pharmaco-
phore hypotheses and diversity calculations were employed.
One part of this library (set 1) was selected based on 3D shape
and chemical similarity to bioactive conformation hypotheses
that were generated for compounds 4 and 6, using a 3D align-
ment approach. The other part of the library (set 2) was based
purely on chemical diversity and was complementary to set 1
(refer to Supplementary data S2 for details on the computational
design process). The library was planned to generate ꢂ400 ana-
logs. The list of suitable in-house and commercially available re-
agents (carboxylic acids 30) was pre-filtered according to
physico-chemical properties of the enumerated compounds via
Pipeline Pilot and SpotFire,16 undesirable (toxic or reactive)
groups were removed, and supplier availability was considered.
This process led to 3564 acids that were prioritized with the
above mentioned 3D similarity and diversity based approach.
In total, 285 representative analogs were synthesized (81%
Compounds 31d-1, 31d-2, 31f and the literature compound
T-00127-HEV1 were completely profiled in a panel of 8 lipid
kinases and 277 protein kinases available at Invitrogen. An
excellent specificity for PI4KIIIß versus other lipid and protein
kinases was found. As shown in Table 4, the few instances where
there is a minor inhibitory activity, the IC50 is in the order of
2000–10,000 nM (vs an IC50 PI4KIIIß of 53–225 nM).
In view of the limitation of the current HRV pharmacological
animal models, it was proposed to test compounds from two struc-
turally diverse classes in the related enterovirus Coxsackie virus
(CV-B4) mouse pancreatitis model in order to distinguish com-
pound related from target related toxicity. Therefore, compounds
31d-1, d-2, f and T-00127-HEV1 were profiled in a panel of ente-
roviruses including CV-B4. As shown in Table 4, they showed activ-
ity against this panel of viruses.
In order to identify compounds for the in vivo evaluation, ana-
logs 31d-1, 31d-2, 31f and T-00127-HEV1 were evaluated in a dose
escalation study in CD-1 mice at 5, 50 and 250 mg/kg (Table 5).19
A good exposure was achieved for these compounds, but at
250 mg/kg the gem-dimethyl carrying 31f showed a superior
exposure and a higher Cmax when compared to 31d-2. Therefore
compound 31f was selected together with compound T-00127-
HEV1 (structurally diverse) based on their overall profile (potency,
ADME properties, PK profile and selectivity). When these
compounds underwent repeat-dosing in SJL mice that are used in
the CVB-4 pancreatitis model, deleterious effects were seen with
mortality observed at all tested doses.8
In summary, compounds meeting lead criteria were obtained:
31f for example showed potency and pan-activity against the
screened HRV genotypes and against other enteroviruses such
as CV-B4. Compounds with selectivity for PI4KIIIß were obtained
and they reached acceptable ADME and PK profiles with good
exposure in CD-1 mice at 5, 50 and 250 mg/kg. This allowed
us to perform repeated dosing studies in-house in SJL mice with
two structurally diverse PI4KIIIß inhibitiors: 31f and T-00127-
HEV1, and to show that short term inhibition of PI4KIIIb is
deleterious. This data raises doubts on the safety of inhibiting
PI4KIIIb in order to reduce HRV-induced asthma and COPD
exacerbations.
chemical success rate). 46 analogs had EC50 values below 2
and 21 below 1 M versus HRV14. The active analogs were
found in both sets the similar and diverse ones.
lM
l
During the design and synthesis of these potent analogs, the
mode of action of this series was identified to be preventing HRV
replication via inhibiting the host lipid kinase phosphatidylinositol
4-kinase III beta (PI4KIIIb).8 At the same time this work was being
conducted, a study was published showing that the target of envi-
roxime-like compounds is PI4KIIIb.17 The compound identified in
this prior study was T-00127-HEV1.
For the most potent analogs (31a–e) the bioresults, ADME
and physchem properties are presented in Table 2. Compounds
31a–c are belonging to set 1 of 3D similar structures and com-
pounds 31d and 31e are belonging to the set 2 of diverse struc-
tures. Compounds 31a–e were pan-active against the screened
genotypes, had
a decreased molecular weight and showed
improvement for their microsome stability (HLM), their perme-
ability (Caco-2 A to B) and their CYP 3A4 when compared with
compound 1. Compounds 31a–d showed also an improved solu-
bility at pH 2 and 6.8. This library allowed the identification of
novel, improved RHS regarding not only potency but also phy-
schem and ADME properties. Based on HLM and their overall
profile compound 31a and 31d were selected for rat PK evalua-
tion (Table 3). Compound 31d showed an improved exposure
Acknowledgments
(AUC = 9.7
l
M ꢁ h) and a lower total clearance when compared
We thank Michael Cordingley and Richard Bethell for their
support for the HRV project. We also would like to thank
Ywe Looper, Stephanie Warner and Julie Edwards for their
assistance with patents and scientific administration and Angelo
Filosa for analytical support. We thank Dr. Johan Neyts and
colleagues at the Katholieke Universiteit Leuven in Belgium for
testing compounds against their panel of enteroviruses (CV-B3,
CV-B4, Echovirus, EV-71, and Poliovirus) for the results
presented in Table 4.
to compound 31a. For comparison, the data for compound
T-00127-HEV1 are shown in Table 3. For the most potent
analogs against HRV, identified from the library (31a–31e), the
activity against PI4KIIIß and
a is reported in Table 2.
PI4KIIIß is involved in a membrane trafficking between the
endoplasmic reticulum, the Golgi and the secretory pathway and
is expressed widely in all tissues in the body. Hsu et al.18 proposed
that the enterovirus protein 3A promotes the recruitment of PI4-
KIIIß to the Golgi membrane. The recruited PI4KIIIß is suggested
to remodel the membrane through phosphorylation and this re-
cruits 3Dpol to the membrane and thereby creating a site that pro-
motes viral RNA synthesis. The effect of inhibiting selectively
PI4KIIIß over a shorter duration was not known.
In order to identify the best compounds with a profile which
would allow an in vivo risk assessment of this kinase, the two
enantiomers of compound 31d were separated by supercritical
fluid chromatography (SFC) and profiled. The enantiomer 31d-2
was more potent than the enantiomer 31d-1 versus all genotypes
tested. 31d-2 and 31d-1 had a similar CD-1 mouse PK at 5 mg/kg
with a good exposure. The gem-dimethyl analog 31f was also syn-
thesized and profiled and this compound maintained the good
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.04.077.
References and notes
1. Gern, J. E. J. Virol. 2010, 84, 7418.
2. Rohde, G. Infect. Disord. Drug Targets 2009, 9, 126.
3. Feil, S. C.; Hamilton, S.; Krippner, G. Y.; Lin, B.; Luttick, A.; McConnell, D. B.;
Nearn, R.; Parker, M. W.; Ryan, J.; Stanislawski, P. C.; Tucker, S. P.; Watson, K.
G.; Morton, C. J. ACS Med. Chem. Lett. 2012, 303.

A. Décor et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3841–3847
3847
4. http://www.biota.com.au/uploaded/154/
Capsid binders are characterized by acting early in the time of addition
experiment, by their inactivity against HRV45 and are known to protect HRV
from the heat treatment by binding and stabilizing the core protein VP1.
11. Edwards, P. J.; Sturino, C. Curr. Med. Chem. 2011, 18, 3116.
12. Other analogs synthesized are not disclosed in this publication.
13. This was shown on several examples, with varied R2 groups bringing activity
for example with R = CH2N(CH2Ph)COMe.
1021819_20hrvphaseiibstudyachieve.pdf; date when the reference was last
accessed: 01.02.2013.
5. Palmenberg, A.; Spiro, D.; Kuzmickas, R.; Wang, S.; Djikeng, A.; Jennifer, A.;
Rathe, J. A.; Fraser-Liggett, C. M.; Liggett, S. B. Science 2010, 324, 55.
6. Simmonds, P.; McIntyre, C.; Savolainen-Kopra, C.; Tapparel, C.; Mackay, I. M.;
Hovi, T. J. Gen. Virol. 2010, 91, 2409.
7. Lee, W. M.; Lemanske, R. F.; Evans, M. D.; Vang, F.; Pappas, T.; Gangnon, R.;
Jackson, D. J.; Gern, J. E. Am. J. Respir. Crit. Care Med. 2012, 186, 886.
8. Spickler, C.; Lippens, J.; Laberge, M-K.; Desmeules, S.; Bellavance, E.; Garneau,
M.; Guo, T.; Hucke, O.; Leyssen, P.; Neyts, J.; Vaillancourt, F. H.; Décor, A.;
O’Meara, J.; Franti, M.; Gauthier, A. Antimicrob. Agents Chemother 2013, http://
14. Boong Won, L.; Seung Dal, L. Tetrahedron Lett. 2000, 41, 3883.
15. Structures of some compounds were confirmed by 2D NMR (ROESY).
16. MW 6 500, TPSA 6 140,
H bond donors 6 10, H acceptor 6 5, 1 6 ClogP,
ClogD 6 3 rot bonds 6 10.
17. Arita, M.; Kojima, H.; Nagano, T.; Okabe, T.; Wakita, T.; Shimizu, H. J. Virol.
2011, 85, 2364.
dx.doi.org/10.1128/AAC.00303-13,
AAC.00303-13v1?etoc.
http://aac.asm.org/cgi/content/abstract/
18. Hsu, N.-Y.; Ilnyska, O.; Belov, G.; Santiana, M.; Chen, Y.-H.; Takavorian, P. M.;
Pau, C.; Vand der Schaar, H.; Kaushik-Basu, N.; Balla, T.; Caneron, C. E.;
Ehrenfeld, E.; van Kuppeveld, F. J. M.; Altan-Bonnet, N. Cell 2010, 141, 799.
19. The highest dose (250 mg/kg) was chosen high on purpose in order to clearly
identify a potential safety issue due to the target populations.
9. Cytotoxicity in H1-Hela, Huh7.5, MRC-5 and C8166.
10. There are several biotargets known for HRV: inhibitors of attachment (eg
Tremacamra), capsid binders (e.g., pleconaril, pirodavir, BTA-798), replication
inhibitors (e.g., enviroxime) and protein processing inhibitors (e.g. rupintrivir).