SAR-based optimization of 2-(1H-pyrazol-1-yl)-thiazole derivatives as highly potent EP1 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2013.10.065

We describe a medicinal chemistry approach for generating a series of 2-(1H-pyrazol-1-yl)thiazoles as EP₁ receptor antagonists. To improve the physicochemical properties of compound 1, we investigated its structure-activity relationships (SAR).

Full text of DOI:10.1016/j.bmcl.2013.10.065

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6569–6576
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
SAR-based optimization of 2-(1H-pyrazol-1-yl)-thiazole derivatives
as highly potent EP1 receptor antagonists
⇑
Masakazu Atobe , Kenji Naganuma, Masashi Kawanishi, Akifumi Morimoto, Ken-ichi Kasahara,
Shigeki Ohashi, Hiroko Suzuki, Takahiko Hayashi, Shiro Miyoshi
Pharmaceutical Research Center, Asahi Kasei Pharma Corporation, 632-1 Mifuku, Izunokuni-shi, Shizuoka 410-2321, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe a medicinal chemistry approach for generating a series of 2-(1H-pyrazol-1-yl)thiazoles as
EP1 receptor antagonists. To improve the physicochemical properties of compound 1, we investigated
its structure–activity relationships (SAR). Optimization of this lead compound provided small compound
25 which exhibited the best EP1 receptor antagonist activity and a good SAR profile.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 14 September 2013
Revised 21 October 2013
Accepted 29 October 2013
Available online 6 November 2013
Keywords:
EP1 antagonist
Pyrazole
Thiazole
Solubility
Optimization
Overactive bladder (OAB) is a urological condition with symp-
toms of urgency, with or without urge incontinence, and is usually
accompanied by frequency and nocturia. OAB is estimated to affect
16% of people in the United States and Europe,1 and 30% in Asia.2
Current OAB therapies consist primarily of antimuscarinic drugs,
but their use is limited by unpleasant side effects such as dry
mouth, constipation and blurred vision.3
Gradual bladder filling is associated with the activation of sen-
sory nerves, and the intensity of the stimuli on these sensory
nerves leads to a desire to void. A proposed etiology for OAB is
an alteration of the excitability of these sensory nerves. Targeting
of the hyperactive pathways represents an attractive therapeutic
approach for OAB.4
Prostaglandin E2 (PGE2) has been suggested as a possible thera-
peutic agent because of its role in the excitement of afferent nerves
via the EP1 receptor.5 This concept leads to EP1 receptor antago-
nists being one of the most promising candidates in developing
OAB drugs.6 For example, evidence for a role of the EP1 receptor
is provided by studies of EP1 receptor-selective antagonists. SC-
19220, an EP1 receptor antagonist, increased bladder capacity in
normal rats7, and in a rat spinal cord injury model of overactive
bladder, the EP1 receptor antagonist ONO-8711 produced a de-
crease in detrusor overactivity.8
antagonists.9 In this study we found that compound 1 appeared to
give the best EP1 receptor antagonistic activity and good oral
pharmacokinetics (Fig. 1). Unfortunately, 1 has low water solubility
and high protein binding to plasma proteins, mainly albumin. To
overcome these problems, we investigated the structure–activity
relationships (SAR) of this lead compound and optimized it by
characterizing its hydrophobic properties, such as the cLogP
value. Our previous results showed that the pyrazole ring plays
an important role in the expression of EP1 antagonistic activity.
We therefore aimed to replace the 3-, 4- and 5-position on the pyra-
zole with various functional groups adjusting physicochemical
property.
Table 1 shows the results of SAR studies of the 5-position on the
pyrazole. Replacement of the phenyl ring with 2-thienyl (2) or 3-
thienyl (3) resulted in moderate activity. We prepared compounds
bearing 3-pyridyl (4), 5-pyrimidyl (5), 1-N-methyl-4-pyrazolyl (6)
and 4-amino-phenyl (7) and found that these compounds had de-
creased EP1 activity, whereas the corresponding 3-fluorophenyl (8)
and 4-fluorophenyl (9) showed moderate activity. These results
suggested that the pocket at the 5 position should be hydrophobic
and shallow, so substituting with a hydrophilic group, such as a
heterocyclic group containing one or two nitrogen atoms, is incom-
patible with activity (Table 1).
We have previously reported hit-to-lead optimization of 2-
(1H-pyrazol-1-yl)-thiazole derivatives as novel EP1 receptor
Table 2 shows SAR studies of the thiazole part. Replacement of
the thiazole ring with 2-oxazolyl (10), 2-furyl (11), 2-pyridyl (12)
or 3-pyridyl (13) resulted in decreased activity. Only the 2-thienyl
substitution (14) was tolerated when the 5-methyl-thiazolyl (15)
and carboxylethylene group (16) were retained. These data
⇑
Corresponding author. Tel.: +81 558 76 8493; fax: +81 558 76 5755.
E-mail address: atobe.mb@om.asahi-kasei.co.jp (M. Atobe).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.10.065

6570
M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
a. hEP1 rep.: ‘human EP1 reporter assay’
b. hCLint.: ‘human intrinsic clearance’
c. rCLint.: ‘rat intrinsic clearance’
d. Sol: ‘solubility’
e. MDCK: ‘Madin-Darby canine kidney’
Figure 1. Profile of 1.
Table 1
Table 2
Optimization of the 5-position on the pyrazole
Optimization of the thiazole part
Compd R2
hEP1 rep. 1
inhibition
lM
hEP1 rep. IC50
(nM)
cLogP
Compd R1
hEP1 rep. 1
inhibition
l
M
hEP1 rep. IC50
(nM)
cLogPa
1
50
5.48
4.63
1
50
5.48
2
3
4
120
80
5.34
5.13
3.98
10
33%
11
12
24%
48%
5.22
5.23
34%
24%
5
3.98
13
14
15
26%
5.03
5.99
5.68
6
7
8%
3.60
4.26
10%
1000
330
8
9
50
5.48
5.62
100
16
500
5.11
a
clogP value was calculated with ChemBioDraw 12.0.
suggested that the thiazole ring might fit into a small pocket, so the
SAR studies were narrowed to focus on this moiety (Table 2).
SAR of the 4-position on the pyrazole was investigated (Table 3).
When R3 = Et (17), EP1 activity was decreased. When R3 = CF3 (18),
EP1 activity was retained. We prepared halogen derivatives, such as
fluorine (19), chlorine (20), bromine (21) and iodine (22), and
found that 20 showed highly potent EP1 receptor antagonist activ-
ity. When R3 = OH (23), EP1 activity was retained. Substitution with
the methoxy group (24) provided decreased EP1 activity. Finally,
the compound bearing an amine (25) appeared to give the best
EP1 receptor antagonist activity (IC50: 20 nM) and the lowest LogP.
These data suggested that this position is critical for controlling the
physicochemical properties of the antagonist.
The results in Tables 1–3 indicated that decreasing the size of
the molecule increases potency. We then focused on optimizing
the pyrazole ring by fusing this ring with both phenyl rings

M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
6571
(Table 4). 2-(3-Phenylindeno[1,2-c]pyrazol-1(4H) (26) exhibited
significantly increased EP1 receptor antagonist activity (IC50
Table 3
Optimization of the 4-position on the pyrazole
:
0.9 nM). 2-(3-Phenyl-1H-benzofuro[3,2-c]pyrazol-1-yl) (27)
showed good EP1 activity and lower cLogP. 3-Phenyl fused (28),
2-(3-phenyl-4,5-dihydro-1H-benzo[g]indazol-1-yl), showed de-
creased EP1 receptor antagonist activity. However, 5-phenyl fused
(29), 2-(3-phenyl-4,5-dihydro-2H-benzo[g]indazol-2-yl), exhibited
significantly increased EP1 receptor antagonist activity (IC50
2.0 nM).
:
Table 5 shows the antagonist activity of highly potent com-
pounds 1, 25, 26 and 29 as measured using an EP1 reporter-gene
assay and a Ca assay. Stimulation of the EP1 receptor by PGE2
was evaluated by a reporter-gene assay and an intracellular Ca2+
release assay using human EP1 receptor-expressed recombinant
cells. Table 5 also shows the intrinsic metabolic clearance (CLint)
of 1, 25, 26 and 29 in rat and human hepatic microsomes. CLint
Compd R3
hEP1 rep. 1 lM inhibition hEP1 rep. IC50 (nM) cLogP
1
50
5.48
6.01
6.28
5.73
5.80
5.85
5.71
4.33
4.53
4.74
17
18
19
20
21
22
23
24
25
ꢀ4%
150
30
60
20
130
60
was calculated from the elimination ratio (0/20 min at 0.5 lM).
These results suggest that 1 and 25 would be stable in both species.
The solubility of 1, 25, 26 and 29 in aqueous buffer at pH 1.2 and
pH 6.8 showed that 25 exhibited the best solubility under these
conditions.
18%
20
To evaluate the epithelial membrane permeability of compound
25, we used Madin–Darby canine kidney (MDCK) cells as an
in vitro cell model Figure 2.10
The permeability of compound 25 through the MDCK cell
monolayer was significantly higher than through the atenolol,
showing that compound 25 exhibits good permeability. Next, the
safety of compound 25 was evaluated by checking CYP inhibition
and genetic toxicity using the Ames test. Compound 25 did not in-
Table 4
Comparison of fused pyrazole ring compounds
hibit the major human CYP enzymes below 5
ure 2. The IC50 values obtained are above 5
l
l
M, as seen in Fig-
M for all human
CYPs. A bacterial mutation test was conducted using an amino
acid-requiring strain of Salmonella typhimurium (TA98) to detect
the point mutations. No genetic toxicity was observed for com-
pound 25.
The pharmacokinetic parameters of 25 in rats are shown in
Figure 2. A single oral administration of compound 25 to rats at
Compd
R4
hEP1 rep. IC50 (nM)
cLogP
1
50
5.48
10 mg/kg (n = 3) provided a Cmax of 6.33
17.4 M/L hr. The bioavailability (BA) inratswascalculated to be 26%.
Compound 25 was tested against a battery of 57 receptors and 4
enzymes at a concentration of 10 M (Cerep, Celle l’Evescault
France). No significant activity (<25%) was seen (data not shown).
Thus, compound 25 shows excellent selectivity over a range of
off-targets with the exception of the TP receptor (IC50 = 0.2 lM).
These data suggested that compound 25 exhibits high potency
and a good pharmacokinetic profile.
The synthesis of the 2-(1H-pyrazol-1-yl)-thiazole derivatives
2–9 listed in Table 1 is shown in Scheme 1. Ethyl benzoate and
propionitrile were condensed with sodium ethoxide to provide
b-ketonitrile 31. Compound 31 and thiosemicarbazide in ethanol
under reflux produced pyrazolylthiourea, which was then reacted
in situ with ethyl 2-bromopyruvate to give 2-(1H-pyrazol-1-yl)thi-
azole 32 in 18% yield in a two-step, one-pot reaction. Iodization of
32 under the Sandmeyer condition gave iodide 33 as a common
intermediate. Suzuki–Miyaura coupling of compound 33 was used
to prepare the heterocyclic moieties 35a–h, followed by hydrolysis
lM and an AUC(0–1) of
l
26
27
28
29
0.9
20
5.41
5.42
5.82
5.82
l
>100
2
Table 5
Comparison of the profile of 1, 25, 26 and 29
Compd
hEP1 rep. IC50
(nM)
Ca assay IC50
(nM)
Human CLint
(ml/min/kg)
Rat CLint
(ml/min/kg)
Protein
bound (%)
Sol. H2O
g/ml)
Sol. pH 1.2
Sol. pH 6.8
(lg/ml)
(
l
(
lg/ml)
1
50
20
0.9
2
500
60
40
35
19
425
149
15
29
927
236
99.71
99.00
99.68
99.88
129
865
4
0
867
0
401
1012
57
25
26
29
20
20
0
10

6572
M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
a. The peak plasma concentration of a drug after administration.
b. BA: ‘bioavailability’
Figure 2. Profile of 25.
Scheme 1. Reagents and conditions: (a) propionitrile, sodium ethoxide, EtOH, rt, (84%); (b) thiosemicarbazide, EtOH, 80 °C then ethyl bromopyrvate, 80 °C, (18%); (c) tBuONO,
I2, MeCN, reflux, (66%); (d) Pd2dba3, (o-tol)3P, K3PO4, 34a–h, DMF, 80 °C, (35a: 65%, 35b: 70%, 35c: 68%, 35d: 62%, 35e: 53%, 35f: 67%, 35g: 65%, 35h: 68%); (e) 5 M NaOH,
EtOH, rt, (2: 61%, 3: 72%, 4: 52%, 5: 48%, 6: 62%, 7: 74%, 8: 80%, 9: 78%); (f) TFA, CH2Cl2, (39%).

M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
6573
Scheme 2. Reagents and conditions: (a) Pd(OAc)2, cBRIDP, K3PO4, 38a, mesitylene, 180 °C (30%) or CuI, K3PO4, 37, 38b–f, mesitylene, 185 °C, (39b: 10%, 39c: 42%, 39d: 63%,
39e: 32%, 39f: 40%); (d) 5 M NaOH, EtOH, rt, (10: 67%, 11: 72%, 12: 32%, 13: 62%, 14: 53%, 15: 54%).
Scheme 3. Reagents and conditions: (a) LiAlH4, THF, rt, (97%); (b) Dess–Martin periodinane, CH2Cl2, rt, (58%); (c) ethyldiethylphosphonic acid, KHMDS, THF, 0 °C, (52%); (c)
5 M NaOH, EtOH, (78%).
to give 2–6, 8 or 9. In the case of 35f, after deprotection of the Boc
group with TFA, hydrolysis of 35i afforded the corresponding car-
boxylic acid 7 (Scheme 1).
copper mediated N-arylation was used to give the thiazole or het-
erocyclic moiety, followed by hydrolysis to provide 11–15
(Scheme 2).
Compounds 10, 11 and 13–16 listed in Table 2 were synthesized
according to Scheme 2. For compound 10, N-arylation of 36 with
chloride 38a in the presence of cBRIDP, manufactured by Takasag-
o11, was used to introduce the oxazole-ring moiety, followed by
hydrolysis to give 10. When bromide 38b–f was used, Buchwald
For compound 16, the synthetic route started from previously
reported compound 40 (Scheme 3), which was first reduced to
the corresponding primary alcohol 41 with LiAlH4. Alcohol 41
was oxidized with Dess–Martin periodinane to give aldehyde 42,
and subsequent reaction under the Horner–Wadsworth–Emmons

6574
M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
Scheme 4. Reagents and conditions: (a) propionitrile, sodium ethoxide, EtOH, rt, (74%); (b) (NH2)2-H2O, EtOH, rt, (98%); (c) CuI, K3PO4, 37, ethyl 2-bromothiazolecarboxylate,
mesitylene, 185 °C, (16%); (d) 5 M NaOH, EtOH, rt, (68%).
Scheme 5. Reagents and conditions: (a) sodium nitrite, urea, DMSO, 0 °C, (53%); (b) thiosemicarbazide, EtOH, 80 °C then ethyl bromopyruvate, 80 °C, (14%); (c) 5 M NaOH,
EtOH, rt, (25: 52%, 20: 74%, 21: 52%, 22: 83%, 18: 74%); (d) tBuONO, CuCl2 (R2 = Cl) or CuBr2 (R2 = Br) or I2 (R2 = I), MeCN, 80 °C, (51a: 46%, 51b: 32%, 51c: 40%); (e) 2,2-difluoro-
2-(fluorosulfonyl)-acetic acid, CuI, DMF, 100 °C, (43%).
condition resulted in the formation of acrylate 43. Finally, hydroly-
sis of 43 afforded the corresponding carboxylic acid 16.
For compound 17, the synthetic route started from the commer-
cially available 1-phenylbutan-1-one 44 (Scheme 4), which was
with a reducing nitroso to give 2-(1H-4-amino-pyrazole-1-yl)
thiazole 50. Finally, hydrolysis of 50 afforded the corresponding
carboxylic acid 25 (Scheme 5).
Chlorination, bromination or iodation of 4-amino-pyrazole 50
under typical Sandmeyer conditions gave the corresponding 4-Cl
(51a), 4-Br (51b) and 4-I (51c), followed by hydrolysis to give
20–22. Trifluoromethylation of 51c under copper-mediated condi-
tions gave 4-CF3 (52), followed by hydrolysis to give 18.
Compounds 19 and 24 listed in Table 3 were synthesized
according to Scheme 6. 2-Fluoro-1,3-diphenylpropane-1,3-dione
(53), reported by Stavber et al.13, was reacted with thiosemicarba-
zide and ethyl bromopyravate under one-pot pyrazole-thiazole
cyclization conditions to give 54, followed by hydrolysis to provide
19. In the same manner 2-hydroxy-1,3-diphenylpropane-1,3-
dione14 (55) was converted to 56, then hydrolysis of 56 gave the
corresponding carboxylic acid 23. Methylether formation of 56
first
a-acylated to the corresponding dione 45 with LHMDS
and PhCOCl at room temperature and then condensed with
(NH2)2-H2O to give 3,5-diphenyl-4-methyl pyrazole 46. Buchwald
N-arylation of compound 47 was used to give the thiazole-ring
moiety, followed by hydrolysis to give 17.
Compounds 18, 20–22 and 25 listed in Table 3 were synthesized
according to Scheme 5. The synthetic route started from compound
48 reported by Quyen et al.12, which was first converted to the cor-
responding a-nitrosodiketone 49 with NaNO2 and urea. Compound
49 was reacted with thiosemicarbazide and ethyl bromopyruvate
in ethanol under reflux conditions in a one-pot pyrazole-thiazole
cyclization to form 2-(1H-pyrazol-1-yl)thiazole, which was reacted

M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
6575
Scheme 6. Reagents and conditions: (a) thiosemicarbazide, EtOH, 80 °C then ethyl bromopyruvate, 80 °C, (54: 21%, 56: 31%); (b) 5 M NaOH, EtOH, rt, (19: 42%, 23: 60%); (c)
MeI, NaH, DMF, 0 °C, (62%).
Scheme 7. Reagents and conditions: (a) LHMDS, PhCOCl, toluene, rt, 10 min; (b) (NH2)2-H2O, AcOH, rt, 3 h, (2 steps, 60a: 72%, 60b: 53%, 64: 85%); (c) CuI, K3PO4, ethyl 2-
bromo-thiazolecarboxylate, 37, mesitylene, 185 °C, (61a: 10%, 61b: 11%, 65a: 6%, 65b: 5%); (d) 5 M NaOH, EtOH, rt, on, (26: 73%, 27: 75%, 28: 72%, 29: 75%).

6576
M. Atobe et al. / Bioorg. Med. Chem. Lett. 23 (2013) 6569–6576
under typical alkylation conditions provided 57, and finally hydro-
lysis of 57 afforded 24 (Scheme 6).
OAB drugs. The continuation of this study will be reported
elsewhere.
The synthetic route of compounds 26–29 listed in Table 4 is
shown in Scheme 7. In the case of phenylindeno[1,2-c]pyrazol
compounds 26 and 27, the synthetic route started from the com-
mercially available 2,3-dihydro-1H-inden-1-one (58a) and benzo-
furan-3(2H)-one (58b) (Scheme 7). These compounds were
acylated with LHMDS and PhCOCl at room temperature to give
the corresponding diones 59a and 59b, and then condensed with
(NH2)2-H2O to give tricyclic-pyrazole 60a and 60b. Buchwald
N-arylation of compound 60a and 60b was used to give the
thiazole-ring moiety as a single isomer, followed by hydrolysis to
give 26 and 27. NOESY correlations observed between 8-H and
the ethylester proton confirmed the structure of 61a. In a similar
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.
10.065.
References and notes
1. Abrams, P.; Cardozo, L.; Fall, M.; Griffiths, D.; Rosier, P.; Ulmsten, U.; van
Kerrebroeck, P.; Victor, A.; Wein, A. Neurourol. Urodyn. 2002, 21, 167.
2. Yamaguchi, O. Clinics Drug Ther. 2002, 21, 2.
3. Andersson, K. E. Pharmacol. Rev. 1993, 45, 253.
4. Khan, M. A.; Thompson, C. S.; Mumtaz, F. H.; Jeremy, J. Y.; Morgan, R. J.;
Mikhailidis, D. P. Prostaglandins Leukot. Essent. Fatty Acids 1998, 59, 415.
5. Ikeda, M.; Kawatani, M.; Maruyama, T.; Ishihama, H. Biomed. Res. 2006, 27, 49.
6. Palea, S.; Toson, G.; Pietra, C.; Trist, D. G.; Artibani, W.; Romano, O.; Corsi, M. Br.
J. Pharmacol. 1998, 124, 865.
7. Maggi, C. A.; Giuliani, S.; Patacchini, R.; Conte, B.; Furio, M.; Santicioli, P. Eur. J.
Pharmacol. 1988, 152, 273.
8. Yoshida, M.; Inadome, A.; Takahashi, W.; Yono, M.; Miyamoto, Y.; Murakami, S.,
et al J. Urol. 2000, 163, 44. abstract 191.
sequence, the
a-acylation of a-tetlarone (62) afforded 63,
which underwent (NH2)2-H2O pyrazole cyclization to provide
the 3-phenyl-4,5-dihydro-1H-benzo[g]indazol (64). Buchwald
N-arylation of compound 64 was used to give the thiazole-ring
moiety as a mixture of regioisomers, followed by separation by
HPLC to provide regioisomers 65a and 65b. The configuration of
65a and 65b was determined from NOE difference data. Hydrolysis
of the ethyl esters 65a and 65b with 5 M NaOH provided 28 and 29,
respectively.
9. Atobe, M.; Naganuma, K.; Kawanishi, M.; Morimoto, A.; Kasahara, K.; Ohashi, S.;
Suzuki, H.; Hayashi, T.; Miyoshi, S. Bioorg. Med. Chem. Lett. 2013, 23, 6064.
10. Avdeef, A.; Tam, K. Y. J. Med. Chem. 2010, 53, 3566.
In conclusion, we have designed and synthesized a series of 2-
(1H-pyrazol-1-yl)thiazoles as EP1 antagonists. The results suggest
that small compound 25 gives the best EP1 antagonistic activity
and exhibits a good pharmacokinetic profile. Compound 25 may
therefore serve as a lead compound for further development of
11. Suzuki, K.; Hori, Y.; Kobayashi, T. Adv. Synth. Catal. 2008, 350, 652.
12. Quyen, T. H. Le; Umetani, S.; Suzuki, M.; Matsui, M. J. Chem. Soc., Dalton Trans.
1997, 643.
13. Stavber, S.; Šket, B.; Zajc, B.; Zupan, M. Tetrahedron 1989, 45, 6003.
14. Blatt, A. H.; Hawkins, W. L. J. Am. Chem. Soc. 1936, 58, 81.