Structure-activity relationships (SAR) and structure-kinetic relationships (SKR) of bicyclic heteroaromatic acetic acids as potent CRTh2 antagonists II: Lead optimization

Article abstract of DOI:10.1016/j.bmcl.2014.08.029

Extensive structure-activity relationship (SAR) and structure-kinetic relationship (SKR) studies in the bicyclic heteroaromatic series of CRTh2 antagonists led to the identification of several molecules that possessed both excellent binding and cellular potencies along with long receptor residence times. A small substituent in the bicyclic core provided an order of magnitude jump in dissociation half-lives. Selected optimized compounds demonstrated suitable pharmacokinetic profiles.

Full text of DOI:10.1016/j.bmcl.2014.08.029

Bioorganic & Medicinal Chemistry Letters 24 (2014) 5123–5126
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure–activity relationships (SAR) and structure–kinetic
relationships (SKR) of bicyclic heteroaromatic acetic acids as potent
CRTh2 antagonists II: Lead optimization
⇑
Juan Antonio Alonso, Miriam Andrés, Mónica Bravo, Marta Calbet, Paul R. Eastwood , Peter Eichhorn,
Cristina Esteve, Manel Ferrer, Elena Gómez, Jacob González, Marta Mir, Imma Moreno, Silvia Petit,
Richard S. Roberts, Sara Sevilla, Bernat Vidal, Laura Vidal, Pere Vilaseca, Miriam Zanuy
Almirall R&D Centre, Laureano Miró, 408-410, 08980 Sant Feliu de Llobregat, Barcelona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Extensive structure–activity relationship (SAR) and structure–kinetic relationship (SKR) studies in the
bicyclic heteroaromatic series of CRTh2 antagonists led to the identification of several molecules that
possessed both excellent binding and cellular potencies along with long receptor residence times. A small
substituent in the bicyclic core provided an order of magnitude jump in dissociation half-lives. Selected
optimized compounds demonstrated suitable pharmacokinetic profiles.
Received 4 July 2014
Revised 7 August 2014
Accepted 8 August 2014
Available online 19 August 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
CRTh2 antagonist
Receptor residence time
Structure–activity relationship (SAR)
Structure–kinetic relationship (SKR)
Small molecules which antagonize the pro-inflammatory effects
of PGD2 mediated by CRTh2 on key cell types associated with aller-
gic inflammation (basophils, eosinophils and Th2 lymphocytes)
should have a potential benefit in related pathologies. As a conse-
quence there has been considerable interest of late in the search
for CRTh2 antagonists.1
drug-target residence time is a factor which, depending on the par-
ticular pharmacokinetic properties of a drug, may be important
when considering the duration of a pharmacological effect in vivo4
and (b) at the onset of this research, only relatively short receptor
residence half-lives (t <2 h) had been reported for optimized car-
½
boxylic acid-containing CRTh2 antagonists.5 In order to maximize
the possibility of the discovery of an acidic antagonist compatible
with a once-daily dosing regimen, a molecule possessing a long
In the preceding article,2 the design and synthesis of a number
of potent bicyclic heteroaromatic acetic acids (Fig. 1) as CRTh2
antagonists, was described.
receptor residence time (t >12 h) was targeted.
½
In this work, the optimization of this series is described result-
ing in the discovery of orally available compounds possessing long
receptor residence times.
An exhaustive study of the SAR of a variety of the bicyclic
templates previously discovered2 began with the investigation of
the effects of ring substitution on potency in the indazole series
(compounds 1–12, Table 1). A general synthetic route used to
access such derivatives has been previously described2 and full
experimental details can be found elsewhere.3
Biological results for indazoles 2–12 are presented in Table 1. It
was quickly found that small substituents in R1 led to moderate
increases in both binding potency and dissociation half-life
(Table 1), fluorine being the preferred group (compound 3). Substi-
tution in R2 (compounds 7–9) led to only minimal changes in bind-
ing potency although compound 8 demonstrated the highest
potency in eosinophil shape change (ESC) functional assay of the
whole indazole series. Incorporation of the substituents which
gave the best improvements with respect to residence time in R1
(F, compound 3) and R2 (Cl, compound 7) in a single molecule
led to compound 10 with a significantly improved receptor resi-
In addition to binding potency, the optimization of the receptor
dissociation half-life was of particular interest given that: (a)
dence half-life (dissociation t 10.5 h). Comparison of this result
½
with compound 1 clearly demonstrated that binding potency was
not the only factor governing the receptor residence time. The
incorporation of even the small fluorine substituent in R3 led to a
significant loss of potency (compound 12) so no further derivatives
⇑
Corresponding author at present address: Drug Discovery Division, Almirall
R&D Centre, Laureano Miró, 408-410, 08980 Sant Feliu de Llobregat, Barcelona,
Spain. Tel.: +34 93 291 7540; fax: +34 93 291 3420.
E-mail address: paul.eastwood@almirall.com (P.R. Eastwood).
http://dx.doi.org/10.1016/j.bmcl.2014.08.029
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

5124
J. A. Alonso et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5123–5126
Table 2
SAR 7-azaindole derivatives
N
O
O
HO
R4
N
O
O
HO
R4
Bicyclic
Heterocycle
N
Figure 1. Bicyclic heteroaromatic CRTh2 antagonists.
N
R2
Table 1
Ring substitution in the indazole seriesa
Compd
R2
R4
GTP
IC50 (nM)
c
binding
ESC whole
blood IC50
(nM)
hCRTh2
a
a
dissociation
a
t
(h)
½
N
132
14
15
16
17
18
19
20
21
22
23
24
25
26
H
Me
Et
CF3
CF3
CF3
CF3
CF3
CF3
CF3
F
F
H
Cl
Me
OCF3
9
10
10
49
11
54
135
17
4.7
5.0
5.2
5.9
5.8
40
34
2.3
21
O
O
HO
7.8
ND
ND
8.0
ND
ND
11
6.4
9.5
18
CF3
9.9
5.4
29b
ND
ND
1.3
11
8.8
31b
24
N
N
cPr
CHF2
CF3
OMe
H
Me
Me
Me
Me
Me
Me
R1
R3
R2
Compd R1
R2
R3 GTP
c
ESC whole
blood IC50
(nM)6
hCRTh2
binding IC50
dissociation
(nM)6
t
(h)b
13
6.7
ND
½
16
ND
12
2
3
4
5
6
7
8
9
H
Cl
F
OMe
Me
CN
H
H
H
H
H
H
H
Cl
Me
F
H
H
H
H
H
H
H
H
H
H
H
F
19
6.3
4.3
23
17
33
16
9
15
15
10
127
47
160
23
1.6
3.2
5.0
2.4
2.3
1.2
2.8
1.8
1.4
10.5
3.3
ND
SO2Me
Bold numbers for dissociation indicate t >12 h.
½
a
27
All mean data values reported are derived from at least two experiments.
Extrapolated from remaining % inhibition at last time point after washout.
b
108
103
23
13
34
18
71
ND
H
H
10
11
12
F
Cl
H
Cl
F
H
N
N
O
O
O
CF3
CF3
Br
N
N
NH
N
a
OEt
(i)
29
(ii)
All mean data values reported are derived from at least two experiments.
For details of the biological assays, see Ref. 6. ND = not determined.
b
N
N
Cl
27
28
30
were prepared. Although only slight potency gains had been
achieved in this series, it was now evident that the desired resi-
dence half-life target of >12 h was likely to be achieved.
N
O
In parallel, structural activity relationships in an alternative ser-
ies based upon the 7-azaindole template (Table 2) was under
investigation. Previously described compound 13, unsubstituted
in the central core, had shown superior activity and pharmacoki-
netic properties when compared to the equivalent indazole deriv-
ative 1.2 Compounds synthesised (14–26) and biological results are
given in Table 2 and a representative synthesis,7 as exemplified by
the preparation of 14 is shown in Scheme 1.
6-Methyl-7-azaindole 28 was prepared from the known chloro
azaindole 278 via a Suzuki coupling with trimethylboroxine fol-
lowed by base-promoted hydrolysis of the carbamate protecting
group. Intermediate 28 was then subjected to a copper-catalyzed
coupling reaction with aryl bromide 29 to give 30 in good yield.
Bromination of the azaindole nucleus went smoothly with NBS to
give 31 which was subjected to a palladium-catalyzed Negishi-type
coupling reaction with a commercially available organozinc deriva-
tive (tBuO2CCH2ZnCl) to install the acetic ester side chain. Finally,
deprotection with trifluoroacetic acid led to the compound 14.
In the 7-azaindole series, keeping the substituent at R4 constant
as a trifluoromethyl residue (Table 2, compounds 13–19), it was
found that placement of small substituents at R2 led to major
changes in receptor dissociation half-life (compounds 14 and 17),
although binding affinities were, in general, little affected. The
most striking difference was seen between compounds 13 and 14
CF3
(iv, v)
(iii)
Br
N
14
N
31
Scheme 1. Reagents and conditions: (i) trimethylboroxine (1.2 equiv), PdCl2-
dppfꢀDCM (5 mol %), K2CO3 (3 equiv), 1,4-dioxane, argon, 120 °C, 18 h then NaOH,
MeOH, rt, 2 h, 80%; (ii) bromide 29 (1 equiv), CuI (10 mol %), trans-N1,N2-dimeth-
ylcyclohexane-1,2-diamine (15 mol %), K3PO4 (2 equiv), 1,4-dioxane, argon, 130 °C,
18 h, 72%; (iii) NBS (1 equiv), EtOAc, 0 °C, 1 h, 78%; (iv) tBuO2CCH2ZnCl (3 equiv),
QPhos (5 mol %), Pd(dba)2 (5 mol %), THF, argon, reflux, 1 h, 84%; (v) TFA, DCM, rt,
2.5 h, 78%.
which differ by only a single methyl group—whilst both com-
pounds had a very similar binding potency, the latter compound
showed an improvement in dissociation half-life by an order of
magnitude (dissociation t >20 h).9 This is in marked contrast to
½
the equivalent pair of indazole derivatives compounds 1 and 8
(Table 1) in which no difference was seen with regard to residence
time. The same trend was seen when comparing derivatives where
R4 had been modified—the incorporation of the methyl substituent
at R2 dramatically improved residence time (compounds 20 and

J. A. Alonso et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5123–5126
5125
21) in the 7-azaindole series. When R2 was fixed as a methyl sub-
Table 4
SAR of amide region in the 7-azaindole seriesa
stituent, a variety of groups could be placed at R4 and the resultant
derivatives (compounds 14, 21–26) maintained excellent binding
potency and residence times (residence time ranking from low to
high H < F < OCF3 < CF3 < Me < Cl). The exception was the sulfone
derivative 26 which was significantly less potent than the unsub-
stituted compound 22.
R5
R6
N
O
O
HO
CF3
N
N
In general, the incorporation of a substituent (either Me or CF3)
in the same relative position as the 6-position of the 7-azaindole
series in a variety of other bicyclic heterocycles6 led to similar dra-
matic increases in receptor residence time (Table 3, compounds
32–42). Clear differences were seen comparing the dissociation
half-lives of substituted compounds with their un-substituted
counterparts.10 The origin of this general effect remains unex-
plained as no attempts to dissect the kinetic features of receptor
binding with radiolabelled ligands11 were made.
Results of the preliminary investigations into the SAR of the
amide portion of the molecule in the 7-azaindole series are pre-
sented in Table 4 (compounds 43–55). A typical synthetic route
is illustrated by the preparation of 44 (Scheme 2). Compound 28
was coupled with the Boc-protected derivative 56 to give 57. Bro-
mination of 57 followed by installation of the ester side chain using
the protocols described previously for the synthesis of 14 gave the
intermediate 58 in good yield. Exhaustive deprotection of 58 fol-
lowed by re-esterification of the acid moiety proceeded in quanti-
tative yield to give key intermediate 59 which was used for the
synthesis of several derivatives. For example, treatment of 59
with cyclobutane carbonyl chloride followed by ester hydrolysis
provided 44 in good yield.
Compd R5
R6
GTP
c
ESC whole
blood IC50
(nM)
hCRTh2
dissociation
t (h)
½
binding
IC50 (nM)
14
43
44
45
46
47
48
49
50
51
52
53
54
55
cPr
Me
cBu
CH2cPr
CH2OMe
CH2Ph
Et
Et
Et
Et
Et
Et
10
10
10
7
9
6
7.8
ND
8.5
29
5
25
ND
4.7
ND
ND
22
60
ND
ND
21
6.4
24
22
21b
24b
ND
26
2-pyridyl Et
NHCH2Ph Et
CH2NMe2 Et
97
5
163
28
11
4
ND
9.2
32b
40b
7.0
7.8
cPr
cPr
cPr
cPr
cPr
iPr
CH2cPr
CH2CH2OPh
CH2CH2OMe
CH2CH2OH
4
14
Bold numbers for dissociation indicate t >12 h.
½
a
All mean data values reported are derived from at least two experiments.
Extrapolated from remaining % inhibition at last time point after washout.
b
Table 3
Impact of ring substitution on dissociation half-lifea
N
O
O
HO
R4
A
R2
Compd
A
R4
R2
GTPc
binding IC50 (nM)
ESC whole blood IC50 (nM)
hCRTh2 dissociation t (h)
½
32
33
F
F
H
CF3
11
14
11
17
2.0
20
∗
∗
N
N
R2
34
35
36
CF3
CF3
CF3
H
Me
CF3
20
17
15
15
10
82
1.9
21
25
∗
∗
∗
∗
∗
N
R2
37
38
39
CF3
CF3
H
H
Me
CF3
37
119
11
84
ND
7.7
2.5
ND
27
∗
N
R2
40
41
CF3
CF3
H
Me
13
1.4
24
2.0
19
∗
7.0
4.2
N
N
R2
∗
N
42
CF3
Me
7.3
26
N
R2
Bold numbers for dissociation indicate t >12 h.
½
a
All mean data values reported are derived from at least two experiments.

5126
J. A. Alonso et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5123–5126
plasma levels as a consequence of moderate clearance and high
Boc
N
volumes of distribution (for acidic species). Compound 14, which
showed the best overall profile, was also found to have reasonable
oral bioavailability in the rat (F = 27%, 10 mg/kg) and was selected
for further profiling.
Boc
Boc
N
N
CF3
Br
CF3
CF3
(ii)
(iii)
56
28
Br
N
N
(i)
In summary, optimization of the bicyclic heteroaromatic series
of CRTh2 antagonists led to the identification of several molecules
that possess both excellent binding and cellular potencies along
with long receptor residence times. The incorporation of a small
substituent (methyl or similar) in the core in many cases led to
an order-of-magnitude increase in dissociation half-lives. The suc-
cessful demonstration that the long drug-target residence times
seen in this series translated into sustained in-vivo pharmacologi-
cal effects will be reported in due course. Further optimization of
the binding kinetics of this series is reported in the following
article.
N
N
57
Boc
N
NH
O
O
O
MeO
CF3
CF3
(iv)
(v-vi)
44
N
N
N
N
59
58
References and notes
Scheme 2. Reagents and conditions: (i) bromide 56 (1 equiv), CuI (10 mol %),
trans-N1,N2-dimethylcyclohexane-1,2-diamine (15 mol %), K3PO4 (2 equiv), 1,4-
dioxane, argon, 130 °C, 48 h, 23%; (ii) NBS (1.3 equiv), EtOAc, ꢁ10 °C, 2 h, 90%;
(iii) t-BuO2CCH2ZnCl (4 equiv), QPhos (5 mol %), Pd(dba)2 (5 mol %), THF, argon,
microwave irradiation, 100 °C, 2 h, 68%; (iv) 4 M HCl in dioxane, rt, 18 h, then
MeOH, 30 °C, 1.5 h, 100%; (v) cyclobutanecarbonyl chloride (1.9 equiv), Et3N
(3.8 equiv), DCM, rt, 30 min, 81%; (vi) LiOHꢀH2O (4 equiv), THF, H2O, rt, 30 min, 82%.
1. Several reviews have recently appeared highlighting the progress in the search
for small molecule CRTh2 antagonists: (a) Norman, P. Expert Opin. Invest. Drugs
2014, 23, 55; (b) Pettipher, R.; Whittaker, M. J. Med. Chem. 2012, 55, 2915; (c)
Chen, J. J.; Budelsky, A. L. Prog. Med. Chem. 2011, 50, 49; (d) Burgess, L. Annu.
Rep. Med. Chem. 2011, 46, 119; (e) Ulven, T.; Kostenis, E. Expert Opin. Ther. Pat.
2010, 20, 1505; (f) Norman, P. Expert Opin. Invest. Drugs 2010, 19, 947; (g)
Medina, J. C.; Liu, J. Annu. Rep. Med. Chem. 2006, 41, 221.
2. Alonso, J. A.; Andrés, M.; Bravo, M.; Buil, M. A.; Calbet, M.; Castro, J.; Eastwood,
P. R.; Eichhorn, P.; Esteve, C.; Gómez, E.; González, J.; Mir, M.; Petit, S.; Roberts,
R. S.; Vidal, B.; Vidal, L.; Vilaseca, P.; Zanuy, M. Bioorg. Med. Chem. Lett. Accepted
for publication.
Table 5
Pharmacokinetic profiles of selected derivatives in Wistar rat (1 mg/kg)a
3. Eastwood, P. R.; Gomez, E.; Gonzalez, J.; Lozoya, M. A.; Matassa, V. G.; Mir, M.;
Roberts, R. S.; Vidal, B. PCT Int. Appl.WO 2013/010881, 2013.
Compd
Terminal
(h)
Cmax
(ng/mL)
Cl
AUC0–1
(ng*h/mL)
Vss
(L/kg)
MRT
(h)
t
(mL/min/kg)
½
4. Recent references which discuss the impact and importance of residence time
in drug discovery include: (a) Dahl, G.; Akerud, T. Drug Discovery Today 2013,
18, 697; (b) Pan, A. C.; Borhani, D. W.; Dror, R. O.; Shaw, D. E. Drug Discovery
Today 2013, 18, 667; (c) Copeland, R. A. Future Med. Chem. 2011, 3, 1491; (d)
Copeland, R. A. Expert Opin. Drug Disc. 2010, 5, 305; (e) Swinney, D. C. Curr. Opin.
Drug Discov. Devel. 2009, 12, 31.
14
32
35
39
40
4.9
1.9
0.5
6.6
4.6
1610
431
366
1900
910
14
76
134
16
1225
222
126
1074
639
1.8
5.4
2.7
3.2
4.2
2.3
1.2
0.3
3.4
2.6
26
5. Stock, N.; Volkots, D.; Stebbins, K.; Broadhead, A.; Stearns, B.; Roppe, J.; Parr, T.;
Baccei, C.; Bain, G.; Chapman, C.; Correa, L.; Darlington, J.; King, C.; Lee, C.;
Lorrain, D. S.; Prodanovich, P.; Santini, A.; Evans, J. F.; Hutchinson, J. H.; Prasit,
P. Bioorg. Med. Chem. Lett. 2011, 21, 1036; (b) Mathiesen, J. M.; Christopoulos,
A.; Ulven, T.; Royer, J. F.; Campillo, M.; Heinemann, A.; Pardo, L.; Kostenis, E.
Mol. Pharmacol. 2006, 69, 1441; (c) Gervais, F. G.; Sawyer, N.; Stocco, R.; Hamel,
M.; Krawczyk, C.; Sillaots, S.; Denis, D.; Wong, E.; Wang, Z.; Gallant, M.;
Abraham, W. M.; Slipetz, D.; Crackower, M. A.; O’Neill, G. P. Mol. Pharmacol.
2011, 79, 69.
6. Andrés, M.; Bravo, M.; Buil, M. A.; Calbet, M.; Castro, J.; Domènech, T.; Eichhorn,
P.; Ferrer, M.; Gómez, E.; Lehner, M. D.; Moreno, I.; Roberts, R. S.; Sevilla, S.
Bioorg. Med. Chem. Lett. 2013, 23, 3349.
7. For full synthetic details see: Vidal, B.; Alonso, J. A.; Buil, M. A.; Eastwood, P. R.;
Esteve, C.; Lozoya, M. A.; Roberts, R. S.; Vidal, L.; Gonzalez, J.; Mir, M. PCT Int.
Appl. WO 2013/010880, 2013.
8. Aronov, A.; Lauffer, D. J.; Li, P.; Tomlinson, R. C. PCT Int. Appl. WO 2005/028475,
2005.
9. A similar effect on residence time had been observed in our previous series of
pyrrolopiperidinone acetic acids: Andrés, M.; Buil, M. A.; Calbet, M.; Casado, O.;
Castro, J.; Eastwood, P.; Eichhorn, P.; Ferrer, M.; Forns, F.; Moreno, I.; Petit, S.;
Roberts, R. S.; Bioorg. Med. Chem. Lett. 2014, 24, 5111.
10. The addition of a single methyl group can cause profound increases in potency,
leading to the term ‘magic methyl’ Leung, C. S.; Leung, S. S. F.; Tirado-Rives, J.;
Jorgensen, W. L. J. Med. Chem. 2012, 55, 4489.
a
Formulations: 40% PEG.
Several conclusions can be drawn from the biological data
(Table 4): (i) a variety of changes made at R5 and R6 led to com-
pounds which maintained excellent binding potency, including
the urea 49. (ii) Incorporation of a 2-pyridyl ring or a basic centre
in R5 led to a loss in binding potency (compounds 48 and 50). (iii)
In certain cases receptor residence times were adversely affected
by small changes in both R5 (compound 43) and R6 (compounds
51 and 54).
The iv pharmacokinetic profiles in rat were determined for ana-
logues selected on the basis of binding/cellular potencies and res-
idence times and the results are given in Table 5.
As anticipated based on previous observations,2 the volumes of
distribution for all compounds were higher than expected for
acidic compounds. Extremely high clearance (Phepatic blood
flow) was seen for compounds 32 and 35 which resulted in poor
to moderate terminal half-lives and low plasma levels. Compounds
14, 39, and 40 showed good terminal half-lives and significant
11. Tummino, P. J.; Copeland, R. A. Biochemistry 2008, 47, 5481.