Discovery of CP-533536: An EP2 receptor selective prostaglandin E2 (PGE2) agonist that induces local bone formation

Article abstract of DOI:10.1016/j.bmcl.2009.01.059

Sulfonamides, exemplified by 3a, were identified as highly selective EP₂ agonists. Lead optimization led to the identification of CP-533536, 7f, a potent and selective EP₂ agonist. CP-533536 demonstrated the ability to heal fractures

Full text of DOI:10.1016/j.bmcl.2009.01.059

Bioorganic & Medicinal Chemistry Letters 19 (2009) 2075–2078
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of CP-533536: An EP2 receptor selective prostaglandin E2 (PGE2)
agonist that induces local bone formation
a
b
a
a
a
Kimberly O. Cameron a, , Bruce A. Lefker , Hua Z. Ke , Mei Li , Michael P. Zawistoski , Christina M. Tjoa ,
*
Ann S. Wright a, Shari L. DeNinno a, Vishwas M. Paralkar a, Thomas A. Owen c, Li Yu d, David D. Thompson a
a Department of Cardiovascular and Metabolic Diseases, Pfizer Global Research and Development, Groton Laboratories, Eastern Point Road, Groton, CT 06340, USA
b Amgen Inc., Thousand Oaks, CA 91320, USA
c Department of Theoretical and Applied Science, Ramapo College, Mahwah, NJ 07430, USA
d Drug Safety and Disposition, Millennium Pharmaceuticals, Inc., Cambridge, MA 02139, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Sulfonamides, exemplified by 3a, were identified as highly selective EP2 agonists. Lead optimization led
to the identification of CP-533536, 7f, a potent and selective EP2 agonist. CP-533536 demonstrated the
ability to heal fractures when administered locally as a single dose in rat models of fracture healing.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 11 December 2008
Revised 15 January 2009
Accepted 20 January 2009
Available online 23 January 2009
Keywords:
Prostaglandin
EP2 receptor
Fracture healing
Bone
Bone has the unique ability to heal, however factors such as
aging, metabolic diseases, and smoking can lead to delayed healing
or non-union of fractured bones. The ability to induce and acceler-
ate bone healing in conditions where bone repair has been im-
paired remains an unmet medical need and would significantly
reduce the cost and morbidity associated with osteoporotic frac-
tures. Fracture healing involves a cascade of events in which vari-
ous growth and differentiation factors have been shown to play a
role. Prostaglandin E2 (PGE2), is known to increase bone formation
in humans and animals when dosed systemically and enhance
bone formation and healing in animals when dosed locally.1 The
pharmacological activity of PGE2 results from its action on four
g-protein coupled cell surface receptor subtypes, EP1, EP2, EP3
and EP4. Of the four PGE2 receptor subtypes, EP2 and EP4 have been
envisaged as the two receptors most likely to be responsible for
PGE2’s actions on bone.2 We sought a non-prostanoid, subtype
selective PGE2 agonist that would induce local bone formation,
and would be better tolerated than previously studied non-selec-
tive prostanoids. We recently reported our discovery of EP2 and
EP4 selective agonists and their ability to restore bone locally as
well as systemically in an osteopenic rat model.3 Herein, we report
the SAR which led to the discovery of potent and selective EP2 ago-
nists and their ability to induce local bone healing.4
Acyclic prostanoid 2 (Fig. 1), reported previously,5 was charac-
terized in our labs as a weak, but selective EP2 agonist. Preparation
of the benzylic alcohol derivative, 3a (Table 1), provided improved
EP2 potency and maintained selectivity. Sulfonamide 3a was lo-
cally injected into the bone marrow of the proximal tibial metaph-
ysis of 6-week-old male rats.6 A single injection dose-dependently
increased new bone formation in the injected site of the marrow
cavity, thus supporting the hypothesis that EP2 agonists can pro-
mote bone formation locally. Medicinal chemistry efforts were
therefore directed towards improving potency while maintaining
selectivity. Results of our SAR efforts which led to the identification
of our clinical candidate are reported herein.
A range of aliphatic bottom chains (3-octanol replacements),
acid linkers (heptanoic acid replacements), and alkyl and aryl sul-
fonamide caps were explored. A variety of approaches were em-
O
O
H3C
S
CO2H
N
O
CO2H
HO
OH
OH
Prostaglandin E2, 1
2
rEP2 IC50 = 5.2 nM, EC50 = 0.2 nM
rEP4 IC50 = 2.1 nM, EC50 = 0.7 nM
rEP2 IC50 = 225 nM, EC50 = 134 nM
rEP4 IC50 = 2533 nM
* Corresponding author. Tel.: +1 860 441 3410; fax: +1 860 715 4706.
E-mail address: kimberly.o.cameron@pfizer.com (K.O. Cameron).
Figure 1. Structures of 1 and 2.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.059

2076
K. O. Cameron et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2075–2078
Table 1
In vitro results for methyl sulfonamide heptanoic acid analogs.
O
H3C
S
CO2H
N
O
R1
Compound
R1
IC50 rEP2 (nm)
225
IC50 rEP4 (nm)
2533
EC50 (cAMP) rEP2 (nm)
134
2
OH
3a
3b
3c
3d
3e
3f
93
2503
636
502
160
289
590
>3200
>3200
>3200
>3200
>3200
>3200
2467
44
NT
OH
OH
887
423
94
O
OH
457
NT
Ph
3g
O
Method A:
O
O
O
H3C
1. MeSO2Cl, TEA, CH2Cl2
H3C
S
S
H
H
Et
R
N
O
CO2H
(
)
(
)
N
O
H2N
O
O
5
5
n
O
2. NaH, DMF
3. R1-CH2-[Cl,Br]
4. NaOH, H2O
R1
OH
O
2: n=4, rEP IC50 = 225 nM
O
1. R1CHO, reductive amination
R2
2
Method B:
H2N
O
S
4: n=3, rEP IC50 = 910 nM
2
L
N
O
L
O
2. R2SO2Cl, TEA, CH2Cl2
3. Ester deprotection
5: n=2, rEP IC50 >3200 nM
2
R1
Figure 2. In vitro potency of truncated acids.
Method C:
O
1. R1CH2NH2, reductive amination
O
O
R2
H
O
S
H
L
N
O
R
Compounds prepared were tested at the rat EP2 and EP4 receptor
2. R2SO2Cl, TEA, CH2Cl2
3. Ester deprotection
L
O
O
subtypes for their ability to bind and to stimulate release of cAMP
and data is presented in Tables 1–3.7 The in vitro numbers reflect
the average of at least n of three determinations. All compounds with
functional activity were full agonists as compared to PGE2 in the
cAMP assay. In addition to EP2 potency and selectivity, our objectives
for a fracture healing agent included single dose delivery to the frac-
ture site. We also desired a compound that would be highly cleared
from the systemic circulation if leakage from the site occurred.
SAR studies on the bottom chain (Table 1) demonstrated that the
benzylic alcohol was not required for binding or functional activity.
A variety of alkyl and aryl groups could be incorporated without sub-
stantial loss in binding potency or selectivity versus EP4.
Changes to the heptanoic acid side-chain were explored for po-
tency improvements. Truncation of the 6-carbon linker provided
significant loss in activity (Fig. 2). Incorporation of a heteroatom
into the 6-carbon linker also led to significant loss in binding activ-
ity (6a, Table 2). Based on these data, a series of analogs was pre-
R1
Scheme 1. General methods for the preparation of sulfonamides.
ployed to access these analogs4a and general methods are provided
in Scheme 1. For the preparation of heptanoic acids, Method A was
followed. Formation of the methyl sulfonamide by treatment of
ethyl 7-aminoheptanoate with methane sulfonyl chloride, alkyl-
ation with the appropriate alkyl halide and subsequent saponifica-
tion provided the desired sulfonamide acids (Table 1). A similar
approach was employed for the preparation of compounds in
Fig. 2. For compounds where a linker was incorporated into the al-
kyl chain, either Method B or C was followed depending on avail-
ability of the amine or aldehyde starting intermediates. In either
case, reductive amination, sulfonamide formation, and ester depro-
tection provided analogs depicted in Tables 2 and 3.

K. O. Cameron et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2075–2078
2077
Table 2
In vitro results for methyl sulfonamide analogs
O
H3C
S
R1
N
O
R2
Compound
R1
R2
IC50 rEP2 (nM)
701
IC50 rEP4 (nM)
>3200
EC50 (cAMP) rEP2 (nM)
407
O
O
CO2H
6a
n-Butyl
6b
6c
6d
6e
6f
n-Butyl
n-Butyl
n-Butyl
n-Butyl
n-Butyl
n-Butyl
t-Butyl
n-Butyl
t-Butyl
3127
833
>3200
>3200
>3200
>3200
>3200
>3200
>3200
>3200
>3200
NT
CO2H
413
NT
CO2H
>3200
494
O
CO2H
102.7
587
NT
CO2H
654
CO2H
6g
6h
6i
>3200
792
CO2H
203
34.4
48
O
CO2H
270
CO2H
6j
379
Table 3
In vitro results for aryl sulfonamide analogs
O
R4
O
A
CO2H
S
N
tBu
Compound
R4
A
IC50 rEP2 (nM)
IC50 rEP4 (nM)
EC50 (cAMP) rEP2 (nM)
7a
7b
7c
7d
Phenyl
2-Pyridyl
1-Methylimidazolyl
4-Chlorophenyl
2-Thiazolyl
CH2
CH2
0
CH2
CH2
0
278
52
326
>3200
63
>3200
>3200
>3200
>3200
>3200
>3200
0.2
2.3
58
NT
1.1
0.3
7e
7f CP-533536
3-Pyridyl
50
pared where a phenyl linker was incorporated into the chain with
the goal of exploring various vectors while trying to maintain the
appropriate chain length (Table 2). More promising linkers from
Table 2 (e.g., 6h and 6j) which provided an attractive balance of po-
tency and selectivity were further optimized by modification of the
sulfonamide cap.
Because our objective was a compound that provided a short
systemic half-life, we chose the metabolically labile tert-butyl
group for our sulfonamide optimization studies. Replacement of
the methyl sulfonamide with aryl provided significant improve-
ments in functional potency (Table 3). The 3-pyridyl sulfonamide
7f, CP-533536, demonstrated excellent in vitro potency against

2078
K. O. Cameron et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2075–2078
observed after a single dose of compound. These data support
N
O
the hypothesis that local administration of an EP2 agonist will pro-
S
O
CO2H
mote bone formation.3b,3c
N
O
In summary, optimization of the weak, but selective non-pro-
stanoid EP2 agonist 3 led to the discovery of CP-533536, a non-pro-
stanoid, highly potent and selective EP2 agonist that promotes
bone formation and improves fracture healing in rat models. Fur-
ther evaluation is underway to assess the potential of CP-533536
to improve bone healing in humans.
tBu
7f, CP-533536
Rat t1/2 = 0.33 h, Cl = 56 mL/min/kg, Vss = 0.49 L/kg
Figure 3. Pharmacokinetic parameters of CP-533536 after i.v. administration of
1 mg/kg in the male Sprague–Dawley rat.
References and notes
Table 4
1. (a) Jee, W. S.; Ma, Y. F. Bone 1997, 21, 297–304; (b) Marks, S. C., Jr.; Miller, S. C.
Endocr. J. 1993, 1, 337–344; (c) Ke, H. Z.; Shen, V. W.; Qi, H.; Crawford, D. T.; Wu,
D. D.; Liang, X. G.; Chidsey-Frank, K. L.; Pirie, C. M.; Simmons, H. A.; Thompson,
D. D. Bone 1998, 23, 239.
2. (a) Raisz, L. G. Clin. Rev. 2006, 4, 123; (b) Zaidi, M.; Zaidi, N.; Sun, L. Drugs Future
2007, 32, 833.
PQCT parameters from rat tibia (n of 10) after a single injection with CP-533536,
Example 7f. Values are given as percent changes as compared to vehicle
Measurement
0.3 mg/kg
1.0 mg/kg
3.0 mg/kg
Total bone area
Bone mineral content
Bone mineral density
À4.7
2.5
4.4
20.6*
53.5*
27.5*
26.3*
20.7*
3. (a) Cameron, K. O.; Lefker, B. A.; Chu-Moyer, M. Y.; Crawford, D. T.; DaSilva-
Jardine, P.; DeNinno, S. L.; Gilbert, S.; Grasser, W. A.; Ke, H.; Lu, B.; Owen, T. A.;
Paralkar, V. M.; Qi, H.; Scott, D. O.; Thompson, D. D.; Tjoa, C. M.; Zawistoski, M. P.
Bioorg. Med. Chem. Lett. 2006, 16, 1799–1802; (b) Paralkar, V. M.; Borovecki, F.;
Ke, H. Z.; Cameron, K. O.; Lefker, B. A.; Grasser, W. A.; Qi, H. L.; Owen, T. A.; Li, M.;
DaSilva-Jardine, P.; Dumont, F.; Korsmeyer, R.; Krasney, P.; Brown, T. A.;
Plowchalk, D.; Vukicevic, S.; Thompson, D. D. Proc. Nat. Aacd. Sci. 2003, 100,
6736; (c) Li, M.; Ke, H. Z.; Hong, Q.; Healy, D.; Li, Y.; Crawford, T.; Paralkar, V.;
Owen, T. A.; Cameron, K. O.; Lefker, B. A.; Brown, T. A.; Thompson, D. D. J. Bone
Min. Res 2003, 18, 2033.
4. (a) Cameron, K. O.; Lefker, B. A.; Rosati, R. L. U.S. Patent 6,498,172, 1999.; (b)
Lefker, B. A.; Cameron, K. O.; Crawford, D. T.; DaSilva-Jardine, P. A.; DeNinno, S.
L.; Gao, H.; Grasser, W. A.; Ebbinghaus, C. F.; Healy, D. R.; Elliott, N. C.; Ke, H. Z.;
Li, M.; Lu, B.; Owen, T. A.; Paralkar, V. M.; Qi, H.; Rosati, R. L.; Tjoa, C. M.; Wright,
A. S.; Yu, L.; Zawistoski, M. P.; Thompson, D. D. Abstracts of Papers, 224th
National Meeting of the American Chemical Society, Boston, MA, 2002; Abstract
MEDI-370.
5. Jones, J. H.; Holtz, W. J.; Bicking, J. B.; Cragoe, E. J., Jr.; Mandel, L. R.; Kuehl, F. A.
J. Med. Chem. 1977, 20, 1299.
6. All procedures were reviewed and approved by the Institutional Animal Care &
Use Committee.
7. Receptor binding assay: membranes were prepared from stably transfected
HEK-293 cells expressing the rat EP2 or rat EP4 receptor subtypes. Determination
of cAMP: cellular cAMP levels were determined in the HEK-293 rat EP2 or rat EP4
cell line. Rat in vitro data was collected because we used rat as our pre-clinical
efficacy model.
7.5*
*
p < 0.001 versus vehicle.
EP2 and selectivity against a broad panel of other targets. This com-
pound was therefore selected for in vivo evaluation.
To determine if 7f met our half-life objectives pharmacokinetic
parameters were determined in the rat. This compound indeed
demonstrated high i.v. clearance as well as low volume of distribu-
tion leading to a very short half-life (Fig. 3).
In order to meet the desired single dose requirement for a frac-
ture healing agent, it was necessary to identify a formulation that
would maintain drug locally at the fracture site for an extended
time period. To achieve this objective we found that administration
of the EP2 agonist as a matrix with poly(D,L-lactide-co-glycolide)
(PLGH) provided an attractive profile.8 The CP-533536 PLGH ma-
trix was directly injected into the marrow cavity of the tibia in a
rat to assess the potential for local bone growth. Seven days after
the single injection, the bone was analyzed cross-sectionally using
peripheral quantitative computerized tomography (PQCT) for bone
formation (Table 4).9 Dose dependent increases in bone was
8. Atrix Laboratories, Inc., Fort Collins, CO.
9. Ke, H. Z.; Qi, H.; Chidsey-Frank, K. L.; Crawford, D. T.; Thompson, D. D. J. Bone
Miner. Res. 2001, 16, 765.