Evolution of novel tricyclic CRTh2 receptor antagonists from a (E)-2-cyano-3-(1H-indol-3-yl)acrylamide scaffold

Article abstract of DOI:10.1016/j.bmcl.2012.12.050

(E)-2-(3-(3-((3-Bromophenyl)amino)-2-cyano-3-oxoprop-1-en-1-yl) -1H-indol-1-yl)acetic acid (1) was discovered in a HTS campaign for CRTh2 receptor antagonists. An SAR around this hit could be established and representatives with interesting activity profiles were obtained. Ring closing tactics to convert this hit series into a novel 2,3,4,5-tetrahydro-1H-pyrido[4, 3-b]indole based CRTh2 receptor antagonist series is presented.

Full text of DOI:10.1016/j.bmcl.2012.12.050

Bioorganic & Medicinal Chemistry Letters 23 (2013) 944–948
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Evolution of novel tricyclic CRTh2 receptor antagonists from
a (E)-2-cyano-3-(1H-indol-3-yl)acrylamide scaffold
Anja Valdenaire, Julien Pothier, Dorte Renneberg, Markus A. Riederer, Oliver Peter, Xavier Leroy,
⇑
Carmela Gnerre, Heinz Fretz
Actelion Pharmaceuticals Ltd, Gewerbestrasse 16, CH-4123 Allschwil, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 25 December 2012
(E)-2-(3-(3-((3-Bromophenyl)amino)-2-cyano-3-oxoprop-1-en-1-yl)-1H-indol-1-yl)acetic acid (1) was
discovered in a HTS campaign for CRTh2 receptor antagonists. An SAR around this hit could be estab-
lished and representatives with interesting activity profiles were obtained. Ring closing tactics to convert
this hit series into a novel 2,3,4,5-tetrahydro-1H-pyrido[4,3-b]indole based CRTh2 receptor antagonist
series is presented.
Keywords:
CRTh2
DP2
Prostaglandin
Asthma
Rhinitis
Ó 2012 Elsevier Ltd. All rights reserved.
Prostaglandin D2 (PGD2), a major metabolite of arachidonic
acid, is predominantly involved in the physiological response to al-
lergy. It is released at high concentration by immunoglobulin E
activated mast cells upon allergen challenge1 and was demon-
strated to activate two G-protein coupled receptors (GPCRs), the
classical DP1 receptor and the recently discovered chemoattractant
receptor-homologous molecule expressed on T-helper 2 cells
(CRTh2 also known as DP2) receptor.2 CRTh2 activation leads to
the recruitment of granulocytes and Th2 cells by chemotaxis3 to
the inflammation site and has been in almost all cases demon-
strated to be pro-inflammatory.4,5 Furthermore it is now well
established that antagonizing selectively the CRTh2 receptor could
be useful in the treatment of asthma and other inflammatory dis-
eases such as allergic rhinitis.6–10 Herein, we describe the discovery
of a novel chemotype series and its conversion into a new lead ser-
ies, which eventually should deliver potent, highly selective CRTh2
antagonists for oral application.
((3-bromophenyl)amino)-2-cyano-3-oxoprop-1-en-1-yl)-1H-in-
dol-1-yl)acetic acid (1, Fig. 1), was identified as a singleton and its
potency to antagonize PGD2 induced hCRTh2 receptor activation
with an IC50 = 0.6
to compete against [3H]PGD2 for the hCRTh2 receptor with an
IC50 = 0.5 M in a radioligand displacement assay.
lM was confirmed. Furthermore, 1 was shown
l
A library of approximately 200 compounds with the general
structure 2 (Fig. 1) was prepared in a semi-automated high-
throughput mode in order to confirm that singleton 1 is a true
hit belonging to a hit cluster, and to demonstrate a clear struc-
ture–activity relationship (SAR) within this cluster. Eventually,
replacements for the structural alerts like the cyano group, the
a
-cyanoacrylamide motif, and the N-aryl amide bond should be
envisaged in order to reduce the presumed toxicity potential.12
The carboxylic acid of 1 was kept throughout this validation pro-
cess since it was recognized as a common functionality to all
endogenous prostanoids as well as many of the published prosta-
noid receptor agonists and antagonists.13
As already described in a preceding publication,11 our search for
CRTh2 receptor antagonists started by screening about 80,000
GPCR biased compounds from our in-house compound collection
by means of an intracellular Ca2+ liberation (FLIPR™, Molecular De-
vices) assay with HEK-293 cells stably expressing the human (h)
CRTh2 receptor. This effort provided two distinct screening hits
that could be confirmed as valuable starting points for a lead dis-
covery program. The hit-to-lead evolution of 2-(2-((2-(4-chloro-
phenoxy)ethyl)thio)-1H-benzo[d]imidazol-1-yl)acetic acid, was
discussed in an earlier publication.11 A second hit, (E)-2-(3-(3-
The synthetic access of such analogues is outlined in Scheme 1.
Commercially available 1H-indole-3-carbaldehyde 3a, accordingly
R1 substituted, was condensed with pyrrolidine in a Dean–Stark
apparatus to form (Z)-3-(pyrrolidin-1-ylmethylene)-3H-indole,
which then was N-alkylated with either ethyl bromoacetate
(n = 1), ethyl 3-bromopropanoate (n = 2) or ethyl 4-bromobutano-
ate (n = 3). Knoevenagel condensation of indolium bromide 4 with
tert-butylcyanoacetate provided cyanoacrylate 5.14 Literature pre-
cedence suggests that stereo-controlled E-selectivity in the forma-
tion of 5 relies on steric effects of the bulky carboxylate versus the
less bulky cyano.15 The Knoevenagel condensation provided exclu-
⇑
Corresponding author. Tel.: +41 61 565 6208; fax: +41 61 565 8901.
E-mail address: heinz.fretz@actelion.com (H. Fretz).
sively E-configured
a-cyanoacrylate analogues as confirmed by
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.12.050

A. Valdenaire et al. / Bioorg. Med. Chem. Lett. 23 (2013) 944–948
945
R1
H
O
A
O
O
NRR'
OR
R2
N
H
Br
N
R3
X
A
CN
B
N
N
X
N
(CH2)n
COOH
COOH
1
(I)
2
FLIPR (Ca2+ flux) hCRTH2 IC
Binding
A= H, Me
B= H, Cl, CN
n= 1,2,3
50= 0.6 µM
hCRTH2 IC
50= 0.5 µM
Figure 1. Structure of high-throughput screening hit 1, general formula 2 indicating investigated positions to establish an SAR, and query formula (I) employed for the data
base search of the Cambridge Crystallographic Data Centre (CCDC).
O
R2
R2
COOH
N
O
CO2tBu
R3
N
R3
(e)
N
CN
N
N
CN
(d), (e)
R1
R1
R1
(c)
COOEt
COOEt
N Br
(CH2)n
N
N
21a
22a
(CH2)n
CO2Et
(CH2)n
CO2Et
(f)
(f)
CO2Et
(j)
4
5
6
COOH
R2
(a) (b)
A
O
N
N
H
O
R3
O
16
2
COOH
CN
R1
20a
(f)
R1
(b)
(j),(e), (f)
R1
(g)
CO2Et
Examples
13a-f
14a-h
15a-l
17
N
H
N
N
N
CO2Et
NBoc
H
3a (A= H)
18, 19a-d
7
3b (A= Me)
8
20b
R1 = H, F, Cl, Me
N
H
1:1 mixture of E/Z isomers
23
(h)
R2
R2
O
O
(f)
(f)
(b), (k)
N
N
COOR
O
R3
R3
HCl
NH
R
B
B
B
(b)
N
(e)
(l)
R1
R1
R1
N
H
N
H
N
N
N
CO2Et
12a (B = Cl)
12b
9a (B = Cl, R = Et)
(B = Cl, R = H)
(B = H, R = H)
COOEt
(i)
COOEt
10a
10b
11a (B = Cl)
11b
25
24
(B = H)
(B = H)
Scheme 1. Reagents and conditions: (a) pyrrolidine, toluene, reflux (87%); (b) Br(CH2)nCO2Et, acetone, reflux, rt (70–93%); (c) NCCH2CO2tBu, NaOEt, EtOH, CHCl3, rt; (d) 50%
TFA in CH2Cl2, (97%); (e) (COCl)2, DMF(cat.), CH2Cl2 (quant.); then HNR2R3, DIEA, CH2Cl2, 0 °C to rt (48–73%); (f) 1 M aqueous NaOH in THF, rt, 15 min to 2 h (35–100%); (g)
NCCH2CONR2R3, NH4OAc, AcOH, toluene, 110 °C, 1:1 mixture of E/Z isomers (33%); (h) CCl3CO2Et, CrCl2, THF, E isomer (31%); (i) LiOH in H2O/THF (81%); (j) BrCH2CO2Et, NaH,
DMF, rt (53%); (k) 4 M HCl in dioxane, rt, overnight (88%); (l) RNCO, DIEA, DCM, rt, 5 h (80%); or RCOCl, TEA, DCM, 0 °C to rt (44–89%).
ROESY NMR experiments (data not shown). A data base search of
the Cambridge Crystallographic Data Centre (CCDC) for compounds
with the generic formula (I) yielded 82 X-ray crystal structures, 61
comprising the alkyl 3-aryl-2-cyanoacrylate and 21 the 3-aryl-2-
cyanoacrylamide motif, respectively (Refcodes listed in Supple-
mentary Table S1). All the structures displayed (E)-configured sub-
acid 13d (first eluting peak, Rt = 3.04 min) and the corresponding Z
isomer (second eluting peak, Rt = 3.28 min) were separated by pre-
parative RP-HPLC.17 E and Z configuration could be unequivocally
assigned by 1D selective gradient NOESY NMR experiments (data
not shown). As expected, unambiguous interactions between the
amide proton and the hydrogen atoms of the methyl group were
observed for 13d, confirming E configuration, whereas no such
interactions were detectable for the second isomer. At this stage
of the project, no effort was made to further explore the A site
by other than methyl substitution due to the lack of a satisfying
stereocontrolled access to pure E isomers with geminal substitu-
tion at C(3).
stitution of the
a-cyanoacrylate double bond. The X-ray crystal
structures revealed co-planarity of the double bond with ring A
at C(3) of the acrylate moiety in all the cases where there is no sub-
stituent connected to the ring atom X. In particular, the recently
published X-ray crystal structure of methyl (E)-2-cyano-3-(1H-in-
dol-3-yl)acrylate (VEGMAW) displays co-planarity of the double
bond with the adjacent indole ring.16 Hydrolysis of tert-butylester
5 with 50% TFA, and subsequent amide formation applying an acid
chloride procedure gave 6. Pure final compounds 13a–c, 14a–h and
15a–l were obtained after saponification of ethyl ester 6 and pre-
parative reverse-phase (RP) HPLC purification.
A stereoselective Cr(II) mediated Z-olefination protocol pro-
vided ethyl (Z)-a-chloroacrylate 9a as a single geometrical isomer
by reacting 3a with ethyl trichloroacetate.18 Alkaline ester hydro-
lysis gave the free carboxylic acid 10a which was converted to
amide 11a using the acid chloride method. Alkylation with ethyl
bromoacetate in the presence of Cs2CO3, saponification of ethyl es-
ter 12a followed by acidic work-up provided the final product 13e
as the free acid. The preparation of 13f started with the amidation
of commercially available (E)-3-(1H-indol-3-yl)-2-propenoic acid
Alkylation of 1-(1H-indol-3-yl)ethanone 3b with ethyl bromo-
acetate followed by the Knoevenagel condensation of 7 with 2-cy-
ano-N-phenylacetamide gave a 1:1 mixture of (E)- and (Z)-8.15 This
mixture was subjected to saponification of the ethyl ester. The free

946
A. Valdenaire et al. / Bioorg. Med. Chem. Lett. 23 (2013) 944–948
Table 2
(10b) with aniline. Subsequent N-alkylation of 11b and deprotec-
tion of 12b finally provided 13f.
hCRTh2 receptor binding IC50 values of compounds 14a–h
In an initial study the relevance of substituents A, B, and the
length of the carboxylic acid bearing alkyl chain (CH2)n was inves-
tigated. The data are summarized in Table 1. In comparison with
13a, a considerable decrease in potency was observed with longer
alkane chains than methylene, for example, propionic acid 13b
H
O
R2
N
CN R3
N
COOH
(IC50 = 5.9
l
M) and butanoic acid 13c (IC50 = 2.6
l
M) were 13 and
14
six times less potent, respectively. As demonstrated with 13d, no
change in potency was caused if A represents a methyl group in-
stead of hydrogen and no significant change in affinity was ob-
a
Compd.
R2
R3
hCRTh2 binding (buffer) IC50 (nM)
a
b
c
d
e
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Methyl
n-Butyl
Cyclohexyl
Phenylmethyl
2-Phenylethyl
15
20
30
30
10
served with 13e where a chlorine atom is replacing
a-cyano.
However, if B represents hydrogen as in 13f a tenfold drop in po-
tency was observed, indicating the importance of a substituent B.
As depicted in Table 2, very potent antagonists were obtained
with N,N-disubstituted 14a–h. If compared with 13a for example,
an almost 30-fold increase in potency was achieved with N-methyl
analogue 14a. Large hydrophobic groups were beneficial for good
hCRTh2 affinity. Within this study, the most potent antagonist
14h was identified with an IC50 = 7 nM. All of the compounds dis-
cussed so far were shown to act as full antagonists in the Ca2+ lib-
eration (FLIPR) assay (data not shown).
f
60
14
g
h
7
a
All IC50 values are reported throughout the article as mean of at least three
experiments.
Next, the effects of R1 substituents at the indole core were
investigated (Table 3). The positional scanning, whereupon one
hydrogen atom was substituted with fluorine, chlorine, and
methyl, each at C(4), C(5), C(6) and C(7) as outlined with 15a–l,
uncovered that such groups had a beneficial effect at positions
C(4) and C(5) providing antagonists with an IC50 between 5 and
10 nM (15a, b, e, f, i, j). Significantly less potent compounds were
obtained if the same substituents were attached at C(6) and C(7).
As demonstrated with 15g and 15k, chlorine or methyl at C(6) de-
creased the IC50 by a factor >50.
In order to further evaluate the 2-cyano-3-(1H-indol-1-yl)acryl-
amide series, potent representatives were investigated with re-
spect to their in vitro pharmacological, physicochemical, ADME
and in vivo pharmacokinetic characteristics. Anolog 14g is pre-
sented in Table 4 as a representative example with an IC50 = 14 nM.
Human serum albumin (HSA, 0.5%) was added to the assay buffer
in order to identify those antagonists that potentially retain po-
tency under conditions where binding to blood proteins occurs.
No SAR and no clear correlations with physico-chemical parame-
ters could be established concerning this frequently observed plas-
ma shift effect. So far, one might conclude that compounds with
small and less hydrophobic NR2R3, clogP <3 and a negative LogD7.4
tend to be less affected. No plasma shift was observed for 14g
(IC50 = 10 nM). An antagonistic effect with an IC50 = 54 nM was ob-
tained for 14g in the functional Ca2+ release assay. Finally, 14g was
shown to inhibit in a functional assay dose-dependently the shape
change of human eosinophils (hESC) with an IC50 = 160 nM.19,20
No cross-selectivity with other prostanoid receptors like the hu-
man prostaglandin E2 (PGE2) receptor subtypes hEP1–4, the human
thromboxane A2 (TXA2) receptor TP2, and DP2, the other PGD2
receptor was noticed. A DP1/CRTh2 selectivity factor of >500 was
found.
The low molecular weight (MW <400) of 14g and its measured
LogD7.4 value of À0.4 gave rise to a compound with excellent sol-
ubility (>600
g mLÀ1) in buffered aqueous solution at neutral pH.
A low potential for drug-drug interactions was expected, since
the IC50 values of 14g on three major relevant cytochrome P450 en-
l
zymes (CYP 2C9, 2D6, 3A4) were >10 lM. Stability in different
media was assessed by incubating 14g in rat and human plasma
for up to 4 h, in simulated gastric fluid (SGF) for 1 h and in simu-
lated intestinal fluid (SIF) for 4 h.21 Recovery of unchanged parent
was determined by LC–MS. The elimination half-life (T1/2) of 14g in
rat as well as in human plasma was >4 h, with respective 90% and
95% recovery of the parent compound after 4 h incubation time.
After incubation in SGF for 1 h, more than 90% of unchanged parent
14g was recovered, whereas after exposure in SIF for 4 h, less than
70% was detectable, indicating that the compound is not stable if
exposed to the latter conditions for a longer time. Metabolic stabil-
ity of 14g was assessed with human liver microsomes (HLM) and
with rat liver microsomes (RML) as well as with rat hepatocytes
(RHepa). Respective intrinsic clearance (CLint) values of 6 and
32 l
L minÀ1 mgÀ1 protein in HLM and in RLM preparations were
determined, indicating a low metabolic conversion potential. Also,
a good metabolic stability was observed in RHepa with a Clint of
11 l .
L minÀ1 106 cellsÀ1
An in vivo pharmacokinetic study was carried out with 14g in
Wistar rats at single intravenous and oral doses of 1 and 10 mg/
kg, respectively (Table 4). After intravenous dosing, a low plasma
clearance (CL) of 14 mL minÀ1 kgÀ1, a short terminal half-life (T1/
2) of 1 h, and common to carboxylic acids, an expected low volume
of distribution (Vss) of 0.7 L kgÀ1 was determined.
An exposure (AUC0–last) of 2200 ng h mLÀ1 was found after oral
administration. The resulting oral bioavailability of only 18% could
be attributed to incomplete absorption.
Table 1
hCRTh2 receptor binding IC50 values of compounds 13a–f
So far, a cluster around singleton 1 was generated and a SAR
could be established with analogues displaying potencies in the
single digit nM range. Due to the limited oral bioavailability not
exceeding the 20% threshold, and due to the fact that the structural
alerts mentioned at the beginning still were present, a scaffold
hopping exercise was initiated with the goal to identify novel core
structures devoid of these unwanted features. Consequently,
the vicinal substituted double bond motif of the (E)-2-cyano-3-
(1H-indol-3-yl)acrylamide series was thoroughly studied in a
Compd.
A
B
n
hCRTh2 binding
(buffer) IC50 (nM)
a
b
c
d
e
f
H
H
H
CN
CN
CN
1
2
3
1
1
1
430
5900
2600
510
560
A
O
N
H
Me CN
H
H
B
N
Cl
H
(CH2)n
COOH
4400
13

A. Valdenaire et al. / Bioorg. Med. Chem. Lett. 23 (2013) 944–948
947
Table 3
hCRTh2 receptor binding IC50 values of compounds 15a–l
Compd.
R1
hCRTh2 binding
Compd.
R1
hCRTh2 binding
Compd.
R1
hCRTh2 binding
(buffer) IC50 (nM)
(buffer) IC50 (nM)
(buffer) IC50 (nM)
R1
6
a
b
c
d
F–C(4)
F–C(5)
F–C (6)
F–C(7)
10
10
45
e
f
g
Cl–C(4)
Cl–C(5)
Cl–C(6)
Cl–C(7)
5
7
290
50
i
j
k
l
Me–C(4)
Me–C(5)
Me–C(6)
Me–C(7)
10
7
690
140
5
4
H
O
N
100
h
15
7
CN Me
N
COOH
Table 4
Physicochemical, in vitro pharmacological and ADME, and in vivo pharmacokinetic characteristics of representative 14g and new lead structure 19d
14g
19d
14g
19d
Pharmacological properties
hCRTh2 Bdg. (buffer)
hCRTh2 Bdg. (0.5% HSA)
hCRTh2 (Ca2+ flux)
Physicochemical properties
MW
IC50 (nM)
IC50 (nM)
IC50 (nM)
IC50 (nM)
14
10
54
160
8.2
>10
>25
17
15
(Da)
385
À0.4
334
À1.2
LogD7.4
Solubility in
Water (pH 4.6)
170
110
>10
>10
>25
b
hESC (plasma)
(
(
(
l
l
l
g mLÀ1
g mLÀ1
g mLÀ1
)
)
)
17
7
670
93
68
64
hDP1 (PGD2 binding)
IC50
IC50
IC50
(
(
(
lM)
lM)
lM)
Buffer pH 4
Buffer pH 7
Stability in
SGF (parent after 1 h)
SIF (parent after 4 h)
515
810
100
100
a
hEP1–4
hTP2 (Ca2+ flux)a
(%)
(%)
In vitro ADME properties
Inhibition of
CYP 2C9/2D6/3A4
Stability in plasma:
T1/2 rat/human
(% parent, 4 h)
CLint (HLM)/(RLM)
CLint (RHepa)
In vivo PK propertiesc
AUC0–last (po)
Cmax
CL
T1/2
Vss
F
(ng h mLÀ1
)
2200
213
14
1
0.7
18
6300
600
4.6
2.3
0.5
IC50
(h)
(
l
M)
>10/>10/>10
>15/>50/>50
(ng mLÀ1
)
(mL minÀ1 kgÀ1
(h)
)
>4/>4
(90/95)
6/32
>4/>4
90/90
5/<4
0.9
(L kgÀ1
(%)
)
(
(
l
l
L minÀ1 mgÀ1 prot.)
L minÀ1 106 cellsÀ1
14
)
11
a
b
c
hEP2, hEP4: PGE2 binding; hEP1, hEP3: PGE2 induced Ca2+ mobilization (FLIPR); hTP: U46619 induced Ca2+ mobilization (FLIPR).
In brackets: pH of final aqueous solution.
Oral administration (po) 10 mg/kg, intravenous (iv) 1 mg/kg as a solution in 20% propyleneglycol/80% buffer, to three male Wistar rats, respectively.
conformational analysis using 13a as a representative example.
Obtained low energy conformations should then serve as a basis
for the design of novel scaffolds. A hypothetical diene structure II
reflecting the substitution pattern of the corresponding (E)-2-cya-
no-3-(1H-indol-3-yl)acrylamide motif is depicted in Figure 2 (A).
Such a diene may adopt the s-cis or s-trans conformation. Gener-
ally, s-trans conformations of conjugated dienes are assumed ther-
modynamically more stable due to steric repulsion (1,3-allylic
strain) between the two inside hydrogen atoms (or substituents)
of the s-cis conformation. However, in the present example steric
repulsion caused by the phenyl ring at C(4) forces the anticipated
diene to adopt the s-cis conformation. A geometry and conforma-
tional energy analysis of 13a was performed using semi-empirical
PM6 and a restricted Hartree–Fock SCF calculation formalism (Pu-
lay DIIS + Geometric Direct Minimization) with a 3-21G⁄⁄ basis set
(Spartan ’10). The result of this energy minimization calculation re-
vealed that low energy conformations were predominantly s-cis
configured whereas the s-trans conformers were classified as high
energy conformations. The low energy conformation pictured in
Figure 2 (B) displays co-planarity of the double bond with the adja-
cent indolyl ring. This is in agreement with the recently published
X-ray crystal structure of methyl (E)-2-cyano-3-(1H-indol-3-
yl)acrylate (VEGMAW).16
bazole 17, and tetrahydro-pyridoindole 18 were envisaged. Energy
minimization calculations of the proposed tricyclic cores were per-
formed and the low energy conformations were compared with
13a (Supplementary Table S2). In general, the proposed molecules
seem to exhibit a more bent and less planar shape than 13a. Ulti-
mately, the compounds were prepared according to established
procedures (Scheme 1) from commercially available 20a, 20b and
23 in four steps, and tested in the above described CRTh2 binding
and FLIPR assays. Carbazole 16 was found equipotent with 13a,
whereas fortuitously racemic 17 was three times more potent than
13a, thus opening up another avenue to a novel and potent CRTh2
receptor antagonist series.22 Tetrahydro-pyridoindole 18 was
found equipotent with 13a (IC50 = 430 nM). Although less potent
than 17, this achiral core was further explored with priority be-
cause analogues were easily accessible (Scheme 1) and no cumber-
some chiral separation or development of an enantioselective
synthesis was needed, as for example, for the production of 3,4-
dihydro-1H-carbazole analogues of 17. A preliminary SAR with a
few analogues (Table 5) already indicated the potential of this
new series, for example, replacing phenylamino in 18 with a more
hydrophobic R (benzyl, 19a, n-pentyl, 19c) resulted in a twofold
improvement in binding affinity, whereas a substantial loss was
observed with 19b bearing a small R (methyl). In comparison to
13a and 18, an unexpected 25-fold lower IC50 value was measured
for 19d (R = phenyl) in the binding assay (IC50 = 17 nM), and 19d
antagonized PGD2 induced intracellular Ca2+ liberation with an
IC50 = 170 nM. Based on these intriguing potencies, it was decided
The calculated 3D structure of 13a was used as a template for
the molecular design of tricyclic core structures in order to lock
the molecule in the s-cis conformation (Fig. 2(C)). By applying this
ring closing tactic the respective carbazole 16, 3,4-dihydro-1H-car-

948
A. Valdenaire et al. / Bioorg. Med. Chem. Lett. 23 (2013) 944–948
(A)
(D)
(B)
(C)
H
O
E
H
O
O
H
H
H
E
N
H
E
H
NH
N
H
N
H
R
R
N
N
H
N
COOH
II (s-cis)
II (s-trans)
13a
13a
O
O
O
O
N
H
R
N
N
H
N
H
N
N
N
N
N
COOH
COOH
COOH
COOH
19a-d
16
IC50 490 nM
17
IC50 145 nM
18
IC50 430 nM
Figure 2. (A) Substituted diene II reflecting the (E)-2-cyano-3-(1H-indol-3-yl)acrylamide motif in its s-cis and s-trans conformation; (B) low energy conformation of structure
13a, purple arrows indicate sites of ring closure; (C) molecular design: ring closing tactic (green arrow) locking the molecule in the s-cis conformation; (D) ‘proof-of-concept’
variants 16, 17, 18 and 19a–d as envisaged from ring closing tactic.
providing DMPK data, M. Holdener and K. Hilpert for supportive
discussions, and J. Williams for proofreading the manuscript.
Table 5
hCRTh2 receptor binding IC50 values of compounds 19a–d
Compd.
R
hCRTh2 binding
(buffer) IC50 (nM)
Supplementary data
R1
6
a
b
c
d
Benzyl
Methyl
n-Pentyl
200
1600
210
17
5
4
H
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
12.050. These data include MOL files and InChiKeys of the most
important compounds described in this article.
O
N
Phenyl
7
CN Me
N
COOH
19
References and notes
1. Murray, J. J.; Tonnel, A. B.; Brash, A. R. N. Engl. J. Med. 1986, 315, 800.
2. Gervais, F. G.; Cruz, R. P.; Chateauneuf, A.; Gale, S.; Sawyer, N.; Nantel, F.;
Metters, K. M.; O’Neill, G. P. J. Allergy Clin. Immunol. 2001, 108, 982.
3. Hirai, H.; Tanaka, K.; Yoshie, O.; Ogawa, K.; Kenmotsu, K.; Takamori, Y.;
Ichimasa, M.; Sugamura, K.; Nakamura, M.; Takano, S.; Nagata, K. J. Exp. Med.
2001, 193, 255.
to fully characterize 19d and to benchmark it with 14g (Table 4).
Comparable potency in the inhibition of the eosinophil shape
change (IC50 = 110 nM) and high selectivity against the other pro-
stanoid receptors were obtained. Improved PK characteristics, for
example, a threefold higher exposure (AUC0–last 6300 ng h mLÀ1),
a three times lower clearance (4.6 mL minÀ1 kgÀ1) and a prolonged
T1/2 (2.3 h) was found for 19d in Wistar rats at single intravenous
and oral doses of 1 and 10 mg/kg, respectively. The lack of obvious
ADME liabilities (no CYP inhibition, low metabolic clearance), im-
proved stability in SGF and SIF were the basis to further exploit
19d as a novel drug-like lead structure, devoid of any structural
alerts. Improving potency and oral bioavailability was the goal of
a subsequent lead optimization program which would finally cul-
minate in the selection of a clinical candidate (Fretz et al., in
preparation).
4. Pettipher, R. Br. J. Pharmacol. 2008, 153, S191.
5. Kostenis, E.; Ulven, T. Trends Mol. Med. 2006, 12, 148.
6. Pettipher, R.; Hansel, T. T.; Armer, R. Nat. Rev. Drug Disc. 2007, 6, 313.
7. Chen, J. J.; Budelsky, A. L. Prog. Med. Chem. 2011, 50, 49.
8. Burgess, L. E. Annu. Rep. Med. Chem. 2011, 46, 119.
9. Schuligoi, R.; Sturm, E.; Luschnig, P.; Konya, V.; Philipose, S.; Sedj, M.;
Waldhoer, M.; Peskar, B. A.; Heinemann, A. Pharmacology 2010, 85, 372.
10. Pettipher, R.; Hansel, T. T. Prog. Respir. Res. 2010, 39, 193. Karger: Basel.
11. Pothier, J.; Riederer, M. A.; Peter, O.; Leroy, X.; Valdenaire, A.; Gnerre, C.; Fretz,
H. Bioorg. Med. Chem. Lett. 2012, 22, 4660.
12. Such structural alerts are not per se prohibitive for further drug development,
as exemplified with approved Entacapone and Teriflunomide, both drugs
comprising features related to this hit series.
13. Hirata, T.; Narumiya, S. Chem. Rev. 2011, 111, 6209.
14. Knoevenagel, E. Chem. Ber. 1894, 27, 2345.
In summary, we demonstrated the validation and characteriza-
tion of a screening hit series. Attempts to eliminate assumed crit-
ical structural features led to the successful design of novel
CRTh2 antagonist scaffolds. For example, the unexpected excellent
properties of 19d provided an excellent starting point for a lead
optimization program.
15. (a) Schwarz, M. J. Chem. Soc., Chem. Commun. 1969, 212; (b) Zabicky, J. J. Chem.
Soc., Chem. Commun. 1961, 683.
16. Sonar, V. N.; Parkin, S.; Crooks, P. A. Acta Cryst. 2006, E62, o1077.
17. Preparative reverse phase HPLC on a Atlantis Prep T3, 10
lm, 30 Â 75 mm
column; eluent A: acetonitrile, eluent B: water containing 0.5% formic acid;
gradient: 10–90% A in 4 min.; flow rate: 75 ml/min.
18. Barma, D. K.; Kundu, A.; Zhank, H.; Mioskowski, Ch.; Falck, J. R. J. Am. Chem. Soc.
2003, 125, 3218.
19. Stubbs, V. E.; Schratl, P.; Hartnell, A.; Williams, T. J.; Peskar, B. A.; Heinemann,
A.; Sabroe, I. J. Biol. Chem. 2002, 277, 26012.
Acknowledgements
20. Sabroe, I.; Hartnell, A.; Jopling, L. A.; Bel, S.; Ponath, P. D.; Pease, J. E.; Collins, P.
D.; Williams, T. J. J. Immunol 1999, 162, 2946.
21. Asafu-Adjaye, E. B.; Faustino, P. J.; Tawakull, M. A.; Anderson, L. W.; Yu, L. X.;
Kwon, H.; Volpe, D. A. J. Pharm. Biomed. Anal. 2007, 43, 1854.
22. Fecher, A.; Fretz, H.; Riederer M. WO2006070325 (Actelion Pharmaceuticals Ltd,
Switz.), PCT Int. Appl. (2006).
The authors thank S. Koch, D. Lotz, A. Jonuzi, P. Risch, B. Butscha,
J. Giller, B. Lack, S. Brand, for technical support, B. Capeleto for
physico-chemical and Francois Le Goff for stability measurements
in different solution media, R. Bravo, S. Delahaye, and H. Kletzl for