Discovery of a novel 2-(1H-pyrazolo[3,4-b]pyridin-1-yl)thiazole derivative as an EP1 receptor antagonist and in vivo studies in a bone fracture model

Article abstract of DOI:10.1016/j.bmcl.2018.06.022

We describe a medicinal chemistry approach to the discovery of a novel EP₁ antagonist exhibiting high potency and good pharmacokinetics. Our starting point is 1, an EP₁ receptor antagonist that exhibits pharmacological efficacy in cystometry models following intravenous administration. Despite its good potency in vitro, the high lipophilicity of 1 is a concern in long-term in vivo studies. Further medicinal chemistry efforts identified 4 as an improved lead compound with good in vitro ADME profile applicable to long term in vivo studies. A rat fracture study was conducted with 4 for 4 weeks to validate its utility in bone fracture healing. The results suggest that this EP₁ receptor antagonist stimulates callus formation and thus 4 has potential for enhancing fracture healing.

Full text of DOI:10.1016/j.bmcl.2018.06.022

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a novel 2-(1H-pyrazolo[3,4-b]pyridin-1-yl)thiazole derivative
as an EP1 receptor antagonist and in vivo studies in a bone fracture model
Masakazu Atobea,⁎, Kenji Naganumaa, Masashi Kawanishia, Takahiko Hayashib, Hiroko Suzukic,
Masahiro Nishidab, Hirokazu Araib
a
Laboratory for Medicinal Chemistry, Asahi Kasei Pharma Corporation, 632-1 Mifuku, Izunokuni, Shizuoka 410-221, Japan
Laboratory for Pharmacology, Asahi Kasei Pharma Corporation, 632-1 Mifuku, Izunokuni, Shizuoka 410-221, Japan
Laboratory for Safety Assessment & ADME, Pharmaceutical Research Center, Asahi Kasei Pharma Corporation, 632-1 Mifuku, Izunokuni, Shizuoka 410-2321, Japan
b
c
A R T I C L E I N F O
A B S T R A C T
Keywords:
EP1 receptor
Prostaglandin E2
Target validation
Oral administration
Bone fracture healing
We describe a medicinal chemistry approach to the discovery of a novel EP1 antagonist exhibiting high potency
and good pharmacokinetics. Our starting point is 1, an EP1 receptor antagonist that exhibits pharmacological
efficacy in cystometry models following intravenous administration. Despite its good potency in vitro, the high
lipophilicity of 1 is a concern in long-term in vivo studies. Further medicinal chemistry efforts identified 4 as an
improved lead compound with good in vitro ADME profile applicable to long term in vivo studies. A rat fracture
study was conducted with 4 for 4 weeks to validate its utility in bone fracture healing. The results suggest that
this EP1 receptor antagonist stimulates callus formation and thus 4 has potential for enhancing fracture healing.
Prostaglandin E2 (PGE2) is a metabolite of arachidonic acid and
interacts with various important bioactive ligands to exhibit physiolo-
gical functions.1 Most of these functions involve the four prostaglandin
receptors EP1, EP2, EP3 and EP4.2 EP1 is most abundant on the C fibers
of the bladder3 and EP1 antagonistic activity may be useful for the
treatment of overactive bladder (OAB).4 We previously reported pyr-
azole/indazole derivatives which have highly potent EP1 receptor an-
tagonistic activity with good selectivity against EP2-EP4,5–7 and their
anti-OAB effect was identified using a rat cystometry model following
1 mg/kg iv administration of the most potent compound 1.7 However,
clinical study results by Ono Pharmaceutical showed that their EP1
receptor antagonist ONO-8539 did not meet their clinical end-point,
leading them to conclude that EP1 receptors may not be promising
targets for OAB treatment.8 And also, KRP-EPA-605, which was co-
development by Kyorin and Kissei, was in phase I clinical trials for the
treatment of overactive bladder. However, the studies was discontinued
optimization studies of our compounds and target validation studies
with our optimum small molecule.
Our compound, 2-(3-phenyl-6-(trifluoromethyl)-1H-indazol-1-yl)
thiazole-4-carboxylic acid (1) (Fig. 1), exhibits high EP1 antagonistic
activity with excellent selectivity against EP2–EP4 and other GPCR re-
ceptors7 but its high lipophilicity complicates pharmacokinetics upon
oral administration. We therefore increased the number of sp3 atoms or
charged atoms to reduce the lipophilicity and improve the pharmaco-
kinetics of the compound and investigated the long term effects of these
derivatives in in vivo studies.
The structure activity relationship (SAR) results for the derivatives
are summarized in Table 1. 4H-Hydroindazole (2) reduced EP1 activity
and hydro-pyrane at the 3-position of the indazole (3) also showed good
EP1 activity and a lower clogP. 1H-Pyrazolo[3,4-b]pyridine (4) ex-
hibited the best EP1 activity, low clogP, and good metabolic stability.
Moving the nitrogen atom to the 4-position on the indazole to provide
1H-pyrazolo[4,3-b]pyridine (5) or the addition of a carbon at the 1-
position to generate indazole (6) reduced EP1 activity. We synthesized
several derivatives of 1H-pyrazolo[3,4-b]pyridine (7–9) but their ac-
tivities were lower than that of 4, and thus we chose 4 as the most
promising compound for evaluation in an in vivo pharmacokinetics
study (Table 1).
in 2014.9 On the other hand, Eli-Lilly reported the generation of EP1
−/
KO mice as a rat-bone fracture model10 and indicated that EP1 re-
−
ceptor is a negative regulator that acts at multiple stages of the fracture
healing process and that the inhibition of EP1 signaling may enhance
fracture healing. These results encouraged us to evaluate the utility of
EP1 receptor antagonists for the clinical treatment of bone fractures.
However, genetic target validation studies to date have been unable to
determine the interaction between KO mouse features and the effects of
small molecules, in some cases. Consequently, we have conducted
The selectivity of EP1–4 receptors of 4 is shown in Fig. 2-(a). 4
showed excellent selectivity against EP2–4. Additionally, 4 provided a
negative Ames test result. A binding panel test with 4 against other
⁎
Corresponding author.
E-mail address: atobe.mb@om.asahi-kasei.co.jp (M. Atobe).
https://doi.org/10.1016/j.bmcl.2018.06.022
Received 13 April 2018; Received in revised form 6 June 2018; Accepted 12 June 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Atobe,M.,Bioorganic&MedicinalChemistryLetters(2018),https://doi.org/10.1016/j.bmcl.2018.06.022

M. Atobe et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 1
Summary of the optimization studies of 1.
No. Structure
EP1 reporter cLog P Clint Rat/
Assay IC50
(nM)
MDCK
human (ml/ (cm/s)
min/kg)
1
0.6
5.56
4.40
3.82
186/192
170/102
232/146
17.8
Fig. 1. 2-(3-Phenyl-6-(trifluoromethyl)-1H-indazol-1-yl)thia-zole-4-carboxylic
acid (1).
GPCR and non-kinetic enzymes showed that 4 has no activity below
1 µM. The in vitro ADME profile of 4 is shown in Fig. 2-(b). The protein
binding ability of 4 was lower than that of 1, indicating that the phy-
sicochemical properties of 4 are superior to those of 1. A pharmacoki-
netic study11 showed that the unbound concentration of 4 was sig-
nificantly higher than that of 1 and thus 4 was chosen as the most
promising compound for further pharmacokinetic studies (Fig. 2-(c)).
Compound 4 was orally administered at 10 mg/kg, 30 mg/kg and
100 mg/kg and its pharmacokinetics are shown in Fig. 3. The corre-
sponding Cmax values were 24.8 µM, 75.1 µM and 179 µM, dose-de-
pendently (Fig. 3). Particularly, a top dose of 100 mg/kg showed to be
able to expose the high concentration after 24 h. Indeed, PK study for
4 weeks was performed with doses of 30 mg/kg QD, 30 mg/kg BID and
100 mg/kg QD. The result suggest that all doses showed good exposure
of 4 during 4 weeks as tool compound for long-term in vivo studies
(Supplementary material).
2
12
26.77
3
0.7
39.02
A femoral fracture study12 was performed with 30 mg/kg twice a
day (BID), 30 mg/kg once a day (OD) and 100 mg/kg OD of 4 as an in
vivo validation study of bone fracture healing. Closed femur fractures
were created in 8-week-old SD mice. The most significant effect of 4
was observed in the callus area at all doses (Fig. 4-(a)): after 4 weeks
administration, the callus area was significantly broader than in the
vehicle control and large gaps in the bone were filled by mineralization,
demonstrating that 4 stimulates the formation of callus. Photos of bone
following the administration of 100 mg/kg OD are shown in Fig. 4-(b).
The result was that 9 bodies out of 19 bodies showed completely cured
the bone fracture (red rectangle highlighted). However, in some mice
there was a delay in bone fracture healing, in contrast to EP1−/− mice.
Thus, EP1 receptor antagonists may aid fracture healing but their effi-
cacy and mechanism of action via EP1 receptor antagonists remain
unclear. Mechanistic studies are currently underway and the results will
be reported in the near future.
4
0.3
4.26
94/70
23.47
5
12
4.47
The synthesis of compound 2 is shown in Scheme 1. Claisen con-
densation of 10 with benzoyl chloride was followed by (NH2)2-H2O to
give fused-pyrazole 11 in 11% yield following a literature protocol.13
Next, Ullmann coupling with copper and amine catalyst led to N-ar-
ylation on the pyrazole nitrogen.14 Finally, the hydrolysis of 12 pro-
vided 2 in 76% yield (Scheme 1).
The synthesis of compound 3 is shown in Scheme 2. The indazole
cyclization via oxime ether based on a literature methodology provided
the indazole 14 in 37% yield. After iodation and Boc-protection, Suzuki
coupling with 2H-dihydro-pyran boronic acid to give the 3-2H-dihy-
dropyranyl indazole in 39% yield. Finally, Ullmann type coupling with
ethyl 2-bromo thiazole carboxylate and hydrolysis yielded compound 3
(Scheme 2).
6
7
12
12
5.39
3.94
111/48
19.66
The synthesis of compound 4 and 5 is shown in Scheme 3. LDA-
mediated lithiation of 2-fluoro-6-trifluoromethyl pyridine 1915 was
followed by Weinreb amide to give 3-benzoyl pyridine, and subse-
quently treated with (NH2)2-H2O to give fused-pyrazole 21 in 62% yield
as 2 steps. Next, Ullmann coupling led to N-arylation on the pyrazole
nitrogen to give 22. Finally, hydrolysis of 22 provided 4 in 90% yield.
In a similar manner, the Ullmann type coupling of commercially
available 23 gave 24 in 19% yield, followed by hydrolysis to give 5
(Scheme 3).
(continued on next page)
2

M. Atobe et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 1 (continued)
No. Structure
EP1 reporter cLog P Clint Rat/
MDCK
Assay IC50
(nM)
human (ml/ (cm/s)
min/kg)
8
40
3.49
9
11%
4.05
inhibition
@100 nM
Fig. 3. The pharmacokinetics of 4 following oral administration.
(a)
Cpd. EP1
EP2
EP3
EP4
Reporter-gene
Binding assay Ki (nM)
assay IC50 (nM)
4
0.3
8.7
>8700 >5000 >2700
(b)
Cpd. cLogP
Cmax
(μM)
AUC(0-8)
(μM/L•h)
Protein
Binding
F (%)
31%
26%
1
4
5.56
4.26
5.34
9.27
99.95
2.62
5.71
97.40
(c)
Fig. 4. (a) Time course of thickening in the callus area; (b) Photos of bone
following the administration of 100 mg/kg OD.
Fig. 2. (a) Selectivity of EP receptor of 1 and 4; (b) in vitro ADME profile of 4, F
(%): bioavailability; (c) Relative ratio of the plasma unbound concentration to
the in vitro IC50 of 4 (pharmacokinetics (PK): rat, p.o., 10 mg/kg, IC50: reporter-
gene assay).
Scheme 1. Reagents and conditions: (a) LHMDS, PhCOCl, THF, rt, then (NH2)2-
H2O, rt, (11%); (b) ethyl 2-bromothiazole-4-carboxylate, (1S,2S)-N1,N2-di-
methylcyclohexane-1,2-diamine, CuI, K3PO4, 185 °C (41%); (c) 5 M NaOH(aq)
EtOH, rt (76%).
,
3

M. Atobe et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 2. Reagents and conditions: (a) MeONH2, (NH2)2-H2O, THF, 80 °C,
(37%); (b) I2, KOH, DMF (83%); (c) Boc2O, Et3N, DMAP, CH2Cl2, rt (quant); (d)
3,4-dihydro-2H-pyran-6-boronic acid pinacol ester, Pd2dba3, K3PO4, (o-tolyl)3P,
toluene, 120 °C (39%); (e) ethyl 2-bromothiazole-4-carboxylate, (1S,2S)-N1,N2-
dimethylcyclohexane-1,2-diamine, CuI, K3PO4, 185 °C (27%); (f) 5 M NaOH(aq)
,
EtOH, rt (83%).
Scheme 5. Reagents and conditions: (a) LDA, THF, −78 °C, then N-methoxy-N-
methylbenzamide, then dimethyl amine/azetidine/pyrrolidine, (NH2)2-H2O,
THF, rt, (28: 24%, 29: 79%, 30: 6%); (b) ethyl 2-bromothiazole-4-carboxylate,
(1S,2S)-N1,N2-dimethylcyclo-hexane-1,2-diamine, CuI, K3PO4, 185 °C (31: 51%,
32: 51%, 33: 43%); (c) 5 M NaOH(aq), EtOH, rt (7: 32%, 8: 40%, 9: 30%).
disclose our findings from molecular biology mechanistic studies.
Acknowledgement
The authors thank Rika Komatsu and Sachiko Amishiro for analy-
tical data, and Yoko Miyata for purification support.
Scheme 3. Reagents and conditions: (a) LDA, THF, −78 °C, then N-methoxy-N-
methylbenzamide (64%); (b) (NH2)2-H2O, EtOH, 70 °C, (62%); (c) ethyl 2-
bromothiazole-4-carboxylate,
A. Supplementary data
(1S,2S)-N1,N2-dimethylcyclo-hexane-1,2-dia-
mine, CuI, K3PO4, 185 °C (22: 74%. 24: 19%); (d) 5 M NaOH(aq), EtOH, rt (4:
90%, 5: 88%).
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.bmcl.2018.06.022.
References
1. Smyth EM, Grosser T, Wang M, Yu Y, FitzGerald GA. J Lipid Res. 2009;50:S423.
2. Breyer RM, Bagdassarian CK, Myers SA, Breyer MD. Annu Rev Pharmacol Toxicol.
2001;41:661.
3. Ikeda M, Kawatani M, Maruyama T, Ishihama H. Biomed Res. 2006;27:49.
4. Yamaguchi O. Folia Pharmacol. 2003;121:331.
5. Atobe M, Naganuma K, Kawanishi M, et al. Bioorg Med Chem Lett. 2013;23:6064.
6. Atobe M, Naganuma K, Kawanishi M, et al. Bioorg Med Chem Lett. 2013;23:6569.
7. Atobe M, Naganuma K, Kawanishi M, et al. Bioorg Med Chem Lett. 2014;24:1327.
8. Chapple CR, Abrams P, Andersson KE, et al. J Urol. 2014;191:253.
9. https://www.pharmacodia.com/yaodu/html/v1/chemicals/
Scheme 4. Reagents and conditions: (a) NaH, DMF, ethyl 2-(bromomethyl)
thiazole-4-carboxylate, 0 °C, (14%); (b) 5 M NaOH(aq), EtOH, rt (63%).
df74c9ff1542f1574271e759b260543f.html. See related papers; Nishigaya Y, Umei K,
Saito Y, et al. Bioorg Med Chem Lett. 2017;25:4044.
10. Zhang M, Ho H-c, Sheu T-j, et al. J Bone Miner Res. 2011;26:132.
11. PK Study: Compounds 1 and 4 were subjected to PK studies in 7 weeks female
Sprague–Dawley rats. Test compounds were administered orally to rats (n = 3) at a
dose of 10, 30, 100 mg/kg in 10%v/v DMSO, 5%v/v Cremophor EL, 85%v/v D.
water (dose volume 5 ml/kg). Blood samples were collected via the subclavian vein
at 0.5, 1, 2, 4, 8, and 24 h post dose. Plasma was separated by centrifugation. Plasma
compound concentrations were determined using LC-MS/MS.
12. Femoral fracture model: Closed femur fractures were created in 8-week-old Sprague
Dawley rats (Charles-River Laboratories, Inc.) were used for this study. Under an-
esthesia induced by intraperitoneal (ip) injection with sodium pentobarbital (50 mg/
kg per ml; Tokyo Chemical Industry Co., Ltd., Tokyo, Japan), a medial parapatellar
arthrotomy was made, and the patella was then dislocated laterally. The 1mm hole
was made by drill at intercondylar fossa and a 21-gaude needle (Terumo
Corporation, Tokyo, Japan) inserted into the length of medullary canal of the femur
from the distal end, and a mid-diaphyseal fracture was created via three-point
bending with a three-point bending device “EZ Test, EZ-L-1kN” (SHIMADZU, Kyoto,
N-alkylation of indazole 25 with NaH led to 26, followed by hy-
drolysis to give compound 6 (Scheme 4).
Compounds 7–9 were synthesized using the Merck protocol16 for
generation of the 1H-pyrazolo[3,4-b]pyridine ring to provide 28–30 in
24%, 79% and 6% yield, respectively. Then, Ullmann coupling with
31–33 and subsequent hydrolysis provided compounds 7–9 in 32%,
40% and 30% yield, respectively (Scheme 5).
In conclusion, we identified 4 as the lead compound exhibiting the
best EP1 antagonistic activity, physicochemical properties, and oral
pharmacokinetics. A rat bone fracture study with 4 showed that 4
broadened the callus area but did not hasten healing. In future we will
4

M. Atobe et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Japan). This is based on; Bonnarens F, Einhorn TA. J Orthop Res. 1984;2:97. Vehicle
callus was measured by tracing the callus portion with “Image J” image analysis
(0.5% MC, OD), 30 mg/kg (OD) of 4, 30 mg/kg (BID) of 4 and 100 mg/kg (OD) of 4
were administrated via oral for 28 days after the operation. Blood samples were
collected via the subclavian vein at 30 min before administration of 4 (trough value)
or 1 h after administration of 4 (Tmax) at 1, 2, 3, 7, 14, 21 and 28 days post dose.
Plasma was separated by centrifugation. Plasma compound concentrations were
determined using LC-MS/MS. Healing of the femur fracture was monitored using
radiographs, which were obtained at 0, 7, 14, 21 and 28 days under anesthesia using
a NAOMI digital X-ray system (RF system lab, Japan). The area of the outer hard
software (https://imagej.nih.gov/ij/:NIH). Statistics; Results are shown as the
mean
SEM. Statistical significance was identified by t tests and two-way ANOVA
followed by Dunnett’s tests; p values of less than .05 were considered significant.
13. Vina D, Olmo E, López-Pérez JL, Feliciano A. Org Lett. 2007;9:525.
14. Antilla JC, Baskin JM, Barder TB, Buchwald SL. J Org Chem. 2004;69:5578.
15. Mongin F, Tognini A, Cottet F, Schlosser M. Tetrahedron Lett. 1998;39:1749.
16. Tucker TJ, Sisko JT, Tynebor RM, et al. J Med Chem. 2008;51:6503.
5