Development of a highly selective EP2-receptor agonist. Part 2: Identification of 16-Hydroxy-17,17-trimethylene 9β-chloro PGF derivatives

Article abstract of DOI:10.1016/S0968-0896(01)00370-4

Further chemical modification of 1a and 2 was undertaken to identify a more chemically stable selective EP2-receptor agonist for development as a clinical candidate. 9β-Chloro PG analogues 4a e and 5a, c e were found to be potent and selective EP2-receptor agonists. Among them, the compound 4aLy, which is a chemically stabilized lysine salt of 4a, exhibited an excellent profile both in biological activities and physicochemical properties. The agonist 4aLy was found to suppress uterine motility in anesthetized pregnant rats, while PGE₂ stimulated uterine motility. Structure activity relationships (SARs) are discussed. Copyright

Full text of DOI:10.1016/S0968-0896(01)00370-4

Bioorganic & Medicinal Chemistry 10 (2002) 1107–1114
Development of a Highly Selective EP2-receptor Agonist. Part 2:
Identification of 16-Hydroxy-17,17-trimethylene
9ꢀ-chloro PGF Derivatives
Kousuke Tani,* Atsushi Naganawa, Akiharu Ishida, Hiromu Egashira, Kenji Sagawa,
Hiroyuki Harada, Mikio Ogawa, Takayuki Maruyama, Shuichi Ohuchida, Hisao Nakai,
Kigen Kondo and Masaaki Toda
Minase Research Institute, Ono Pharmaceutical Co., Ltd., Shimamoto, Mishima, Osaka 618-8585, Japan
Received 7 August 2001; accepted 17 October 2001
Abstract—Further chemical modification of 1a and 2 was undertaken to identify a more chemically stable selective EP2-receptor
agonist for development as a clinical candidate. 9b-Chloro PG analogues 4a–e and 5a, c–e were found to be potent and selective
EP2-receptor agonists. Among them, the compound 4aLy, which is a chemically stabilized lysine salt of 4a, exhibited an excellent
profile both in biological activities and physicochemical properties. The agonist 4aLy was found to suppress uterine motility in
anesthetized pregnant rats, while PGE2 stimulated uterine motility. Structure–activity relationships (SARs) are discussed. # 2002
Elsevier Science Ltd. All rights reserved.
Introduction
Since our final goal is to identify a highly selective EP2-
receptor agonist and develop it as a clinical candidate,
we focused on the identification of a more potent and
chemically stable EP2-receptor agonist. Thus, synthesis
and biological evaluation of PGF derivatives, which do
not undergo the degradation reaction shown in Scheme
1, were continued based on the information obtained
from the preceding experimental findings. We report
here the identification and biological evaluation of a
more potent and chemically stable EP2-receptor ago-
nist. Structure–activity relationships (SARs) are also
discussed.
The diverse biological activities of PGE2 are considered
to be a hybrid of the activities that are mediated by the
four EP-receptor subtypes EP1, EP2, EP3 and EP4.1 Of
these, EP2-receptor2 has been characterized as inducing
relaxation of blood vessel, gastrointestinal tract, trachea
and uterine smooth muscle,3 and is suggested to play
some role in the production and control of cytokines4
and bone metabolism.5 Development of a potent and
selective EP2-receptor agonist is expected to be a pro-
mising approach to developing a therapeutically useful
drug. In a preceding study,6 we reported the identifi-
cation of 16-hydroxy-17,17-trimethylene PGE2 1a as a
highly selective EP2-receptor agonist. Although 1a
could be a useful research tool for the EP2-receptor
mediated pharmacological activities, its clinical utility is
limited by a weak potency relative to that of PGE2.
Moreover it is a chemically unstable compound because
it is well known that PGE derivatives I start to degrade
with the initial conversion to the corresponding PGA
derivatives II as shown in Scheme 1.
Chemistry
9b-Chloro PGF analogues 4a–e were synthesized from
the corresponding 9-keto analogues 6a–e6 as shown in
Scheme 3. Sequential reactions were: (1) stereoselective
*Corresponding author. Tel.: +81-75-961-1151; fax: +81-75-962-
9314; e-mail: k.tani@ono.co.jp
Scheme 1. Degradation pathway of PGE derivatives.
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00370-4

1108
K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
reduction of the 9-keto group of 6a–e with lithium tri-
sec-butylborohydride;7 (2) acetylation with acetic anhy-
dride; (3) deprotection of the t-butylsimethylsilyl (TBS)
groups with aqueous hydrogen fluoride;8 (4) separation
of the diastereoisomer (16R-OH) by a silica gel column
chromatography gave 8a–e, respectively. Protection of
the two hydroxy groups with t-butyldimethylsilyl tri-
fluoromethanesulfonate followed by deacetylation with
potassium carbonate in methanol gave 10a–e, respec-
tively. Tosylation of 10a–e followed by chlorination
with tetrabutylammonium chloride9 and deprotection of
the two TBS groups yielded 12a–e. The corresponding
ꢀ8,9 unsaturated side product was removed by column
chromatography. Compounds 12a–e were converted to
4a–e by the alkaline hydrolysis, respectively. Com-
pounds 5a and 5c–e were synthesized from 7a and 7c–e,6
respectively, by the same procedure as described in the
synthesis of 4a–e from 6a–e. The configuration of C-16
of 4a was tentatively assigned to 16S and reconfirmed
by the enantioselective synthesis, which will be pre-
sented in a subsequent report.
1a, 2 and 3. A series of 9b-chloro PGF analogues was
evaluated for mouse (m) EP-receptor and human (h) IP-
receptor affinities, and mEP2-receptor agonist activity
as reported previously.6 Two series of 9b-chloro PGF
analogues, 4a–e, 5a and 5c–e were identified. Of these,
4a exhibited the most promising profiles both in the in
vitro and in vivo studies. Compound 4a was further
evaluated for its ability to suppress uterine motility in
anesthetized pregnant rats after intravenous adminis-
tration.
Previously,6 we succeeded in identifying a series of 16-
hydroxy-17,17-trimethylene PGE2 analogues, which
demonstrated highly selective EP2-receptor agonist
activity. On the basis of the reported SARs, 4a–e, 5a and
5c–e were synthesized and biologically evaluated (Table
2). All of these compounds showed only weak affinity to
the hIP-receptor up to 104 mM. 9b-Chloro-16-hydroxy-
17,17-trimethylene-o-nor-PGF2 4a exhibited a promis-
ing profile both in EP2-receptor selectivity and agonist
activity. Homologation of the o-chain of 4a provided 4b
with slightly less subtype selectivity with regards to the
EP3- and EP4-subtypes, while their potencies both in
the EP2-receptor affinity (Ki) and agonist activity (EC50)
were slightly reduced. Replacement of the n-propyl
moiety of 4b with an allyl moiety afforded 4c an increase
both in the EP2-receptor affinity and agonist activity,
while its EP1-receptor selectivity was reduced relative to
4b. The EP2-receptor affinity was maximized in the 16-
cyclopropylmethyl derivative 4d, while its EP4-receptor
affinity increased relative to 4a–c and its potent EC50
value was retained. Replacement of the 16-cyclopro-
pylmethyl moiety of 4d with an isobutyl moiety afforded
4e a 10-fold less potent EC50 value relative to 4d. The
EP1-receptor affinity of 4e was reduced compared with
4d while its EP4-receptor affinity was restored.
Results and Discussion
To block the dehydration reaction of the b-hydroxy
ketone moiety of the PGE analogues I described in
Scheme 1, 9b-chloro PGF analogues10 4a–e, 5a and 5c–e
were synthesized and biologically evaluated. The 9b-
chloro PGF analogue, which is illustrated in Schering’s
nocloprost11 and DP-receptor agonist,9,12 has been
reported to be chemically stable. In our preliminary
study, the 9-deoxy-9b-chloro PGF2 310 exhibited a high
EP2-receptor affinity (Table 1). As such, 9b-chloro PGF
analogue was expected to show a further increased EP2-
receptor affinity. As demonstrated in Scheme 2, chemi-
cal modification began with structural hybridization of
The same chemical modification as described above was
applied to the 1,6-inter-p-phenylene analogues to give
5a and 5c–e. Compound 5a exhibited EP2-receptor
selectivity with several times less potent affinity than
those of 4a–e, while its EC50 value was markedly
reduced relative to those of 4a–e. The 16-allyl analogue
5c demonstrated more potent EP2-receptor affinity and
agonist activity than 5a with good subtype selectivity.
The 16-cyclopropylmethyl analogue 5d demonstrated
more potent EP2-receptor affinity and agonist activity
than 5a and 5c, while its EP1- and EP4-receptor affi-
nities increased. The EP2-receptor affinity of the 17-iso-
butyl analogue 5e was as potent as that of 5a, while its
agonist activity was less potent. The EP2-receptor affi-
nities of 5a and 5c–e were nearly 3-fold less potent than
those of the corresponding PGF2 analogues 4a and 4c–e,
respectively, while their EC50 values were more than
100-fold less potent than those of 4a and 4c–e. With
regards to the two types of a-chain, SAR of 9b-chloro
analogues was different from that of 9-keto analogues.6
9b-Chloro analogues demonstrated a relatively big dif-
ference in their EC50 values up to their structural change
in the a-chain, although 9-keto analogues demonstrated
a relatively small difference in their EC50 values.6 In the
case of compound 5, the flexibility of the a-chain was lim-
ited compared with that of 4. Thus, the conformational
Scheme 2. Molecular design of 9b-chloro PGF analogues 4a–e, 5a and
5c–e.

K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
1109
Scheme 3. Synthesis of 9b-chloro analogues 4a–e, 5a and 5c–e. Reagent: (a) lithium tri-sec-butylborohydride, THF, À78 ꢀC; (b) Ac2O, DMAP,
pyridine; (c) aqueous HF, MeCN; (d) separation; (e) TBSOTf, 2,6-lutidine, CH2Cl2; (f) K2CO3, MeOH; (g) TsCl, pyridine; (h) n-Bu4NCl, toluene; (i)
aqueous NaOH, MeOH.
change in the 5-membered ring that was caused by
transformation of C-9 from sp2 to sp3 reduced the EP2
agonist activity of 5.
uterine motility in a dose-dependent manner. Increased
potency in the suppression of uterine motility by 4a was
observed relative to those of 9-keto analogues 1a and
1b, while the interphenylene analogue 5c demonstrated
much less potency. Complete elimination of side effects
such as hypotension could not be accomplished,
although the hypotensive effects tended to decrease by
reducing the affinity for the IP-receptor compared with
that of 1b. Consequently, the hypotension caused by 1a,
4a and 4b was considered to be one of the effects inher-
ent in an EP2-receptor agonist. Based on its highly
selective EP2-receptor affinity, potent agonist activity
and in vivo effect on spontaneous uterine motility in late
term pregnant rats, compound 4a was selected for fur-
ther evaluation.
The compound 4a displaced [3H]PGE2, bound to the
membrane of CHO cells expressing the mEP2- and
hEP2 receptors with Ki values of 2.2 and 0.74 nM,
respectively. The EP2-receptor affinity was nearly 10-
fold higher than that of PGE2, and 150- and 300-fold
higher than those of AH-1320513 and butaprost,14
respectively. Compound 4a exhibited a much lower
affinity for the other EP subtypes (mEP1, mEP3a and
EP4) as well as prostanoid receptors (mFP, hTP and
hIP) other than the mDP-receptor. The affinity of 4a for
the mDP-receptor was at least 15-fold lower than for
the mEP2-receptor. Similar findings were confirmed by
functional studies based on increases in intracellular
cAMP production and Ca2+ levels. As shown in Figure
1, 4a stimulated cAMP production in mEP2-receptor
expressing cells with an EC50 value of 2.8 nM. Its
potency was comparable to that of PGE2, and nearly
10-fold more potent than that of butaprost free acid
form 1b. Compound 4a showed much lower agonist
activity to the other prostanoid receptor subtypes. The
details of the biological profiles of 4a will be described
in a subsequent report.
Although 9b-chloro PGF2 4a was expected to be che-
mically stable, it demonstrated a similar degree of
instability to the 9-keto analogues 1a–b. However, it
was found to be markedly stabilized by forming a salt of
4a with a base such as lysine. The physicochemical
properties of the lysine salt 4aLy in handling and water-
solubility, etc. were excellent. As shown in Table 4, both
of the compounds 1a and 4a were viscous oils, and
unstable at ambient temperatures. According to the
The modifications resulted in the discovery of highly
selective EP2-receptor agonists 4a–e, 5a and 5c–e. The
in vivo activities of the test compounds 1a, 1b, 4a, 4c
and 5c were evaluated using anesthetized pregnant rats
(on days 18–20 of pregnancy: n=5–6). Whereas PGE2
stimulated uterine motility,15 1a, 1b, 4a, 4c and 5c sup-
pressed uterine motility (Table 3). Intravenous adminis-
tration of 4a and 4c showed significant suppression of
Table 1. The Ki values of PGE2 and 3
Compd
Binding Ki (nM)
mEP1
mEP2
mEP3
mEP4
PGE2
3
18
0.3
38
1.4
5.0
0.8
3.1
0.2
Figure 1. Effect of PGE2, 1b and 4a on cAMP production in CHO
cells expressing mouse EP2-receptors.

1110
K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
HPLC analysis, the percentage of compound 1a
remaining after storage at ambient temperatures for 8
weeks was 8%. The compound 4a showed low stability
(after 1 month at 60 ꢀC, the percentage of compound
remaining was 21%) while the sodium salt 4aNa showed
quite good stability at 60 ꢀC. However 4aNa was still
not good enough to be developed as a clinical candidate
because it was hygroscopically very unstable. The lysine
salt 4aLy demonstrated excellent physicochemical
properties such as stability, solubility in water and
descriptive features. The percentage remaining of 4aLy
after storage at 60 ꢀC for one month was 100%, and its
solubility in water was more than 100 mg. Additionally,
4aLy was a good crystalline product with a melting
point of 166–168 ꢀC.
In summary, the highly selective EP2-receptor agonist
4aLy was identified by chemical modification of the
chemical lead 1a, which was derived from butaprost.
Using compound 4aLy, its biological activities, which
are mediated by the EP2-receptor, were investigated
regarding uterine contractile activity. The selective EP2-
receptor agonist 4aLy was found to suppress uterine
motility in anesthetized pregnant rats, while PGE2 sti-
Table 2. Biological evaluation of 4a–e, 5a and 5c–e
Compd
R1
R2
Binding Ki (nM)
EC50 EC50 (nM)b
mEP1
mEP2
mEP3
mEP4
6100
EP2
3.8
4a
4b
4c
4d
4e
5a
5c
5d
5e
>104
3.3
4.2
1.7
0.9
2.7
10
>104
>104
1300
1100
740
2800
2700
2700
5200
>104
>104
>104
>104
3300
2300
400
6.0
1.8
2.5
21
2600
>104
>104
1100
2900
2100
>104
660
310
240
180
790
5.1
3.0
10
>104
aUsing membrane fractions of Chinese hamster ovary (CHO) cells expressing prostanoid receptors, Ki values were determined by the competitive
binding assay, which was performed according to the method of Kiriyama et al.17 with some modifications. When the test compound did not dis-
place binding of radioligands by 50% even at a concentration of 104 nM, the Ki value was not determined (expressed >104).
bWith regards to the subtype-receptor agonist activity, EC50 values were determined based on the effects of the test compounds on the increases in
intracellular cAMP production in the mouse (m) EP2 receptor.
Table 3. Pharmacological effects of selective EP2-receptor agonists in rat
Compd
Uterine activitya
Hypotensive effect
ED50 (mg/kg iv)
Relative potency
(1b=1.0)
ꢀBP (maximal
response, mmHg)
Relative potency
(1b=1.0)
PGE2
1a
1b
4a
4c
Stimulateb
274.7
455.2
32.9
33.5
>300
—
1.7
1.0
13.8
13.6
<1.5
—
—
0.57
1.0
0.61
0.48
—
22.8 (1000)c
39.8 (1000)
24.4 (100)
19.2 (100)
15.8 (300)
5c
aThe test compounds were intravenously administrated to anesthetized pregnant rats (on days 18–20 of pregnancy, n=5–6). Uterine motility was
evaluated according to the Montevideo method.16 Uterine activity was calculated from the uterine motility for the 5-minute period before and after
commencement of the administration, and post-dose uterine activity was calculated as percentage of inhibition to pre-dose uterine activity.
bPGE2 stimulated uterine motility at >1.8 mg/kg.15
cHypotensive effects (ꢀBP) showed maximal efficacy at roughly 3-fold dosage of ED50 values of the uterine activity. Numbers in parentheses indicate
the dose of maximal hypotensive response (mg/kg iv).

K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
Table 4. Physicochemical properties of EP2-receptor agonists
1111
Compd
Stabilitya (%)
Solubility
(mg/mL of H2O)
Description
Mp (ꢀC)
1a
4a
4aNa
4aLy
8b
N.T.d
N.T.
>10
Viscous oil
Viscous oil
Hygroscopic amorphous
Colorless crystal
—
—
—
166–168
21c
97c
100c
>100
aPercentage of the remaining test compound was determined by HPLC analysis.
bPercentage after 8 weeks at ambient temperatures.
cPercentage after 1 month at 60 ꢀC.
dNot tested.
mulated uterine motility. Other biological activities of
4aLy are being investigated in our laboratory.
water. The organic layer was washed with aqueous HCl
water and brine, dried over magnesium sulfate and
concentrated in vacuo to give a crude 9a-acetoxy deri-
vative as an oil.
Experimental
To a stirred solution of above-described crude 9a-ace-
toxy derivative and 10 mL of acetonitrile, 0.5 mL of
48% aqueous HF was added at 0 ꢀC and stirring con-
tinued for 1.5 h at room temperature. The reaction
mixture was poured into saturated aqueous sodium
bicarbonate. The aqueous layer was extracted with
EtOAc. The combined organic layers were washed with
brine, dried over magnesium sulfate and concentrated in
vacuo. The residue was purified by column chromato-
graphy (Lobar1 pre-packed column, size B, 60%
EtOAc/n-hexane) to give a less polar product (16R–OH,
142 mg, 39%) and a more polar product (16S–OH, 8b,
148 mg, 40%). Less polar product; Colorless oil; TLC
Rf 0.30 (n-hexane/EtOAc, 1/1); 1H NMR (200 MHz,
CDCl3) d 5.66 (ddd, J=15.0, 7.8, 6.0 Hz, 1H), 5.45–5.30
(m, 3H), 5.15–5.05 (m, 1H), 4.00–3.85 (m, 1H), 3.67 (s,
3H), 3.55 (dd, J=10.0, 2.4 Hz, 1H), 2.58–2.40 (m, 1H),
2.40–1.30 (m, 23H), 2.31 (t, J=7.4 Hz, 2H), 2.06 (s,
3H), 0.94 (t, J=7.2 Hz, 3H). More polar product 8b;
General directions
Analytical samples were homogeneous as confirmed by
TLC, and afforded spectroscopic results consistent with
the assigned structures. All 1H NMR spectra were
obtained using a Varian Gemini-200, VXR-200s or
Mercury300 spectrometer. Mass spectra were obtained
on a Hitachi M1200H, JEOL JMS-DX303HF or Per-
Septive Voyager Elite spectrometer. IR spectra were
measured on a Perkin–Elmer FT-IR 1760X or Jasco
FT/IR-430 spectrometer. Elemental analyses for car-
bon, hydrogen, nitrogen and sulfur were carried out on
a Perkin–Elmer PE2400 SeriesII CHNS/O analyzer.
Optical rotations were measured using a Jasco DIP-
1000 polarimeter. Column chromatography was carried
out on silica gel [Merck silica gel 60 (0.063–0.200 mm)
or Wako Gel C200]. Thin layer chromatography was
performed on silica gel (Merck TLC or HPTLC plates,
silica gel 60 F254).
1
Colorless oil; TLC Rf=0.23 (n-hexane/EtOAc, 1/1); H
NMR (200 MHz, CDCl3) d 5.65 (ddd, J=14.8, 8.0, 6.2
Hz, 1H), 5.43–5.25 (m, 3H), 5.15–5.05 (m, 1H), 3.95–
3.82 (m, 1H), 3.67 (s, 3H), 3.55 (dd, J=10.0, 2.4 Hz,
1H), 2.60–2.40 (m, 1H), 2.40–1.20 (m, 23H), 2.30 (t,
J=7.4 Hz, 2H), 2.06 (s, 3H), 0.94 (t, J=6.7 Hz, 3H).
(16S)-9-O-Acetyl-15-deoxy-16-hydroxy-17,17-trimethyl-
ene PGF2ꢁ methyl ester (8b). To a stirred solution of 6b
(740 mg, 1.17 mmol) in 20 mL of THF was added
lithium tri-sec-butylborohydride (1.0 M in THF, 1.76
mL, 1.76 mmol) at À78 ꢀC under an argon atmosphere,
and stirring continued for 30 min. The reaction was
quenched with 1 mL of 30% aqueous hydrogen perox-
ide and allowed to warm up to 0 ꢀC. The reaction mix-
ture was extracted with EtOAc. The organic layer was
washed with aqueous HCl, then brine, dried over mag-
nesium sulfate and concentrated in vacuo. The residue
was purified by column chromatography on silica gel
(10% EtOAc/n-hexane) to yield the 9a-hydroxy deriva-
tive (558 mg, 75%) as a colorless oil. TLC Rf 0.35 (n-
hexane/EtOAc, 9/1); 1H NMR (200 MHz, CDCl3) d
5.60–5.10 (m, 4H), 4.15–3.90 (m, 2H), 3.66 (s, 3H), 3.55
(t, J=5 Hz, 1H), 2.70–2.50 (m, 1H), 2.40–1.20 (m, 24H),
1.00–0.80 (m, 21H), 0.10–0.00 (m, 12H).
(16S)-11-O-(t-Butyldimethylsilyl)-15-deoxy-16-(t-butyldi-
methylsilyloxy)-17,17-trimethylenePGF2ꢁ methyl ester
(10b). To a stirred solution of 8b (148 mg, 0.329 mmol)
and 2,6-lutidine (0.18 mL, 1.58 mmol) in methylene
chloride (3 mL) was added dropwise t-butyldimethyl-
silyl trifluoromethanesulfonate (0.22 mL, 0.948 mmol)
at 0 ꢀC under argon atmosphere. After stirring for 1 h at
room temperature, the reaction was quenched with
saturated aqueous sodium bicarbonate. The aqueous
layer was extracted with EtOAc. The combined organic
layer was dried over magnesium sulfate and con-
centrated in vacuo. The residue was dissolved in
methanol (2 mL) after the addition of potassium car-
bonate (87 mg, 0.632 mmol), then the suspension was
stirred for 2 h at 40 ꢀC. After cooling in an ice-bath, the
reaction was quenched with water and extracted with
EtOAc. The organic layer was washed with water, then
brine, dried over magnesium sulfate and concentrated in
vacuo. The residue was purified by column chromato-
A solution of the 9a-hydroxy derivative (518 mg, 0.813
mmol), acetic anhydride (0.15 mL, 1.62 mmol), 4-
(dimethylamino)pyridine (3 mg) and 1 mL of pyridine
was stirred for 12 h at room temperature. The reaction
mixture was diluted with EtOAc and washed with

1112
K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
graphy on silica gel to yield 10b (202 mg, 96%) as a
1
3H). Compounds 4a, 4c–e, 5a and 5c–e were synthesized
from 6a, 6c–e, 7a and 7c–e, respectively, according to
the same procedure described above.
colorless oil. TLC Rf=0.35 (n-hexane/EtOAc, 9/1); H
NMR (200 MHz, CDCl3) d 5.60–5.10 (m, 4H), 4.15–
3.90 (m, 2H), 3.66 (s, 3H), 3.55 (t, J=5 Hz, 1H), 2.70–
2.50 (m, 1H), 2.40-1.20 (m, 24H), 1.00–0.80 (m, 21H),
0.10–0.00 (m, 12H).
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylene-!-norPGF2 (4a). Pale yellow viscous oil;
optical rotation [a]D25=À24.4 (c 1.00, EtOH); TLC
Rf=0.33 (EtOAc/n-hexane/AcOH, 60/30/1); IR (neat)
3351, 2936, 1709, 1432, 1243, 1069, 968, 866 cmÀ1; MS
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylenePGF2 methyl ester (12b). To a stirred solu-
tion of 10b (137 mg, 0.215 mmol) in pyridine (1 mL) was
added p-toluenesulfonyl chloride (435 mg, 2.28 mmol)
at room temperature under argon atmosphere. After
stirring for 12 h at room temperature, the reaction mix-
ture was quenched with water and extracted with
EtOAc. The organic layer was washed with aqueous
HCl, water, then brine, dried over magnesium sulfate
and concentrated in vacuo to give a tosylate as an oil.
To a stirred solution of the tosylate in toluene (5 mL)
was added tetrabutylammonium chloride (598 mg, 2.15
mmol). The reaction mixture was stirred for 6 h at 50 ꢀC
under an argon atmosphere. After cooling in an ice-
bath, the reaction mixture was quenched with water and
extracted with EtOAc. The organic layer was washed
with brine, dried over magnesium sulfate and con-
centrated in vacuo to give a crude product. To a stirred
solution of this crude product and 5 mL of acetonitrile,
0.25 mL of 48% aqueous HF was added at 0 ꢀC. The
reaction mixture was stirred for 1.5 h at room tempera-
ture, and then poured into saturated aqueous sodium
bicarbonate–EtOAc. The aqueous layer was extracted
with EtOAc. The combined organic layers were washed
with brine, dried over magnesium sulfate and con-
centrated in vacuo. The residue was purified by column
chromatography (Lobar1 pre-packed column, size B,
5% 2-propanol/toluene) to remove an eliminated pro-
duct (ꢀ8,9 unsaturated product) to yield 12b (62 mg,
67%) as a colorless oil. TLC Rf=0.26 (n-hexane/EtOAc,
2/1); MS (APCI, Pos, 20V) m/z 409 (M+HÀH2O)+, 391
(M+HÀ2H2O)+, 355 (M+HÀ2H2OÀHCl)+; IR
(neat) 3369, 2954, 1739, 1436, 1222, 1071, 968 cmÀ1; 1H
NMR (200 MHz, CDCl3) d 5.58 (ddd, J=15.0, 8.2, 5.6
Hz, 1H), 5.50-5.32 (m, 3H), 4.18-3.95 (m, 2H), 3.67 (s,
3H), 3.53 (dd, J=10.4, 2.2 Hz, 1H), 2.76 (br, 1H), 2.40–
1.20 (m, 23H), 2.33 (t, J=7.3 Hz, 2H), 0.94 (t, J=6.8
Hz, 3H).
1
(FAB, Neg) m/z 397 (MÀH)À, 361 (MÀHÀHCl)À; H
NMR (200 MHz, CDCl3) d 5.56 (ddd, J=15.4, 8.2, 5.2
Hz, 1H), 5.55–5.30 (m, 3H), 5.60–5.00 (br, 3H), 4.20–
3.96 (m, 2H), 3.56 (dd, J=10.2, 2.0 Hz, 1H), 2.42–1.30
(m, 20H), 2.35 (t, J=7.0 Hz, 2H), 0.92 (t, J=7.4 Hz,
3H).
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylene-19,20-didehydroPGF2 (4c). Pale yellow vis-
cous oil; TLC Rf 0.74 (n-hexane/EtOAc/AcOH, 1/3/
0.05); MS (APCI, Neg, 20V) m/z 409 (MÀH)À; IR
(neat) 3363, 2936, 1709, 1639, 1435, 1245, 1072, 996,
969, 914, 734 cmÀ1; 1H NMR (300 MHz, CDCl3) d 5.94
(ddt, J=17.1, 10.2, 7.2 Hz, 1H), 5.56 (ddd, J=15.3, 7.8,
6.0 Hz, 1H), 5.50–5.36 (m, 3H), 5.17–5.07 (m, 2H), 4.11
(q, J=7.2 Hz, 1H), 4.04 (m, 1H), 3.57 (dd, J=10.2, 2.1
Hz, 1H), 2.40–1.63 (m, 25H).
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylene-!-dinor-18-cyclopropylPGF2 (4d). Color-
less viscous oil; TLC Rf 0.31 (n-hexane/AcOEt/AcOH,
3/2/0.05); IR (neat) 3367, 2932, 1708, 1433, 1248,
1019, 969 cmÀ1; MS (FAB, Pos) m/z 425 (M+H)+,
407 (M+HÀH2O)+, 389 (M+HÀ2H2O)+, 353
1
(M+HÀ2H2OÀHCl)+; H NMR (200 MHz, CDCl3) d
5.60 (ddd, J=15.4, 7.6, 5.4 Hz, 1H), 5.55–5.35 (m, 3H),
4.20–3.98 (m, 2H), 4.20–3.00 (br, 3H), 3.71 (dd, J=10.4,
2.2 Hz, 1H), 2.40–1.60 (m, 8H), 2.36 (t, J=6.9 Hz, 2H),
1.51 (dd, J=14.2, 6.8 Hz, 1H), 1.37 (dd, J=14.2, 6.2
Hz, 1H), 0.90–0.65 (m, 1H), 0.57–0.45 (m, 2H), 0.15–
0.05 (m, 2H).
(16S)-9-deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-tri-
methylene-19-methylPGF2 (4e). Colorless viscous oil;
TLC Rf=0.34 (n-hexane/AcOEt/AcOH, 3/2/0.05); IR
(neat) 3367, 2952, 1708, 1435, 1251, 1074, 969 cmÀ1; MS
(FAB, Pos) m/z=427 (M+H)+, 391 (M+HÀ2H2O)+,
355 (M+HÀ2H2OÀHCl)+; 1H NMR (200 MHz,
CDCl3) d 5.60 (ddd, J=15.4, 8.2, 5.6 Hz, 1H), 5.55–5.35
(m, 3H), 4.20–3.98 (m, 2H), 4.20–3.00 (br, 3H), 3.65
(dd, J=10.2, 2.2 Hz, 1H), 2.40–1.65 (m, 19H), 2.36 (t,
J=7.1 Hz, 2H), 1.55 (dd, J=14.2, 6.6 Hz, 1H), 1.33 (dd,
J=14.2, 6.2 Hz, 1H), 0.92 (d, J=6.6 Hz, 3H), 0.91 (d,
J=6.6 Hz, 3H).
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylene-PGF2 (4b). To a stirred solution of 12b (45
mg, 0.110 mmol) in methanol (3 mL), 2 N sodium
hydroxide (1 mL) was added at room temperature.
After stirring for 1 h, the reaction mixture was acidified
with aqueous hydrochloric acid and extracted with
EtOAc. The organic layer was washed with water, then
brine, dried over magnesium sulfate and concentrated in
vacuo. The residue was purified by column chroma-
tography on silica gel to yield 4b (44 mg, quant.) as a
pale yellow oil. TLC Rf 0.31 (n-hexane/EtOAc/AcOH,
3/2/0.05); MS (APCI, Neg, 20V) m/z 411 (MÀH)À; IR
(neat) 3369, 2931, 1709, 1435, 1249, 1070, 968 cmÀ1; 1H
NMR (200 MHz, CDCl3) d 5.58 (ddd, J=15.4, 7.6, 5.4
Hz, 1H), 5.55–5.35 (m, 3H), 4.20–4.00 (m, 2H), 4.00–
3.00 (br, 3H), 3.57 (dd, J=10.2, 2.2 Hz, 1H), 2.40–1.20
(m, 22H), 2.36 (t, J=6.9 Hz, 2H), 0.94 (t, J=6.8 Hz,
(1R,2R,3R,4R)-1-Chloro-2-[2-(4-carboxyphenyl)ethyl]-3-
[(1E,4S)-4-hydroxy-5,5-trimethylene-1-heptenoyl]-4-hy-
droxycyclopentane (5a). Colorless viscous oil; TLC
Rf=0.14 (n-hexane/AcOEt, 1/1); MS (APCI, Neg, 20V)
m/z 419 (MÀH)À; IR (neat) 3391, 2931, 1694, 1611,
1575, 1421, 1278, 1179, 1074 cmÀ1; 1H NMR (200 MHz,
CDCl3) d 8.00 (d, J=8.0 Hz, 2H), 7.27 (d, J=18.0 Hz,
2H), 5.58 (ddd, J=15.4, 8.4, 6.6 Hz, 1H), 5.40 (dd,
J=15.4, 8.2 Hz, 1H), 4.40–3.00 (br, 3H), 4.18–4.03 (m,

K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
1113
2H), 3.56 (dd, J=10.2, 2.2 Hz, 1H), 2.79 (t, J=7.6
Hz,2H), 2.38–1.30 (m, 16H), 0.91 (t, J=7.8 Hz, 3H).
insoluble substances were removed by filtration. After
cooling, 5 L of EtOAc was added to the filtrate and the
resulting precipitates were collected by filtration and
dried in vacuo to yield the l-lysine salt 4aLy (54.6 g,
91%) as a colorless crystal. Mp 166–168 ꢀC (dec.);
Optical rotation [a]D25=À20.4 (c 1.00, H2O); MS (FAB,
Pos.) m/z 545 (M+H)+, 421 (MÀLys+Na)+; IR (KBr)
3424, 2934, 1618, 1542, 1406, 1307, 965 cmÀ1; 1H NMR
(200 MHz, CD3OD) d 5.70–5.30 (m, 4H), 4.02 (q,
J=7.1 Hz, 2H), 3.58–3.45 (m, 2H), 2.92 (t, J=7.3 Hz,
2H), 2.40–1.30 (m, 28H), 0.92 (t, J=7.5 Hz, 3H). Anal.
calcd for C28H49ClN2O6; C, 61.69; H, 9.06; N, 5.14;
found; C, 61.72; H, 9.21; N, 5.21.
(1R,2R,3R,4R)-1-Chloro-2-[2-(4-carboxyphenyl)ethyl]-3-
[(1E,4S)-4-hydroxy-5,5-trimethylene-1,7-octadienoyl]-4-
hydroxycyclopentane (5c). Colorless viscous oil; TLC Rf
0.15 (n-hexane/AcOEt, 1/1); MS (APCI, Neg, 20V) m/z
431 (MÀH)À; IR (neat) 3390, 2930, 1695, 1640, 1611,
1575, 1422, 1279, 1179, 1078 cmÀ1; 1H NMR (200 MHz,
CDCl3) d 8.00 (d, J=8.0 Hz, 2H), 7.27 (d, J=18.0 Hz,
2H), 5.93 (ddt, J=17.4, 10.0, 7.2 Hz, 1H), 5.65–5.50 (m,
1H), 5.38 (dd, J=15.4, 8.4 Hz, 1H), 5.15–5.06 (m, 2H),
4.80–3.60 (br, 3H), 4.18–4.00 (m, 2H), 3.55 (bd, J=8.4
Hz, 1H), 2.78 (t, J=7.6 Hz, 2H), 2.41–1.60 (m, 16H).
Prostanoid EP receptor binding assay
(1R,2R,3R,4R)-1-Chloro-2-[2-(4-carboxyphenyl)ethyl]-3-
[(1E,4S)-4-hydroxy-5,5-trimethylene-6-cyclopropyl-1-
hexenoyl]-4-hydroxycyclopentane (5d). Colorless viscous
oil; TLC Rf 0.26 (n-hexane/EtOAc, 1/2); MS (APCI,
Neg. 20V) m/z 445 (MÀH)À, 409 (MÀHÀHCl)À; IR
(neat) 3369, 2930, 1694, 1611, 1423, 1284, 1179, 1083,
Membranes from CHO cells expressing the prostanoid
receptors were incubated with radioligand (2.5 nM of
[3H]PGE2) and the test compounds at various con-
centrations in assay buffer (10 mM Kpi (KH2PO4,
KOH; pH 6.0), 1 mM EDTA and 0.1 mM NaCl).
Incubation was carried out at 25 ꢀC for 60 min except
for EP1-receptor (20 min). The incubation was termi-
nated by filtration through Whatman GF/B filters. The
filters were then washed with ice-cold buffer [10 mM
Kpi (KH2PO4, KOH; pH 6.0), 0.1 mM NaCl], and the
radioactivity on the filter was measured in 6mL of liquid
scintillation (ACSII) mixture with a liquid scintillation
counter. Nonspecific binding was determined by incu-
bation of 10 mM unlabeled PGE2 with assay buffer.
1019, 970, 909, 858, 734 cmÀ1 1H NMR (200 MHz,
;
CDCl3) d 8.01 (d, J=8.0 Hz, 2H), 7.28 (d, J=8.0 Hz,
2H), 5.57 (m, 1H), 5.39 (m, 1H), 4.38 (br, 3H), 4.10 (m,
2H), 3.70 (bd, J=8.4 Hz, 1H), 2.79 (t, J=7.6 Hz, 2H),
2.40–1.62 (m, 14H), 1.52 (dd, J=14, 6.6 Hz, 1H), 1.36
(dd, J=14, 6.4 Hz, 1H), 0.78 (m, 1H), 0.49 (m, 2H),
0.10 (m, 2H).
(1R,2R,3R,4R)-1-Chloro-2-[2-(4-carboxyphenyl)ethyl]-3-
[(1E,4S)-4-hydroxy-5,5-trimethylene-7-methyl-1-oct-
enoyl]-4-hydroxycyclopentane (5e). Colorless viscous oil;
TLC Rf 0.33 (n-hexane/EtOAc, 1/2); MS (APCI, Neg.
20V) m/z 447 (MÀH)À, 411 (MÀHÀHCl)À; IR (neat)
3368, 2952, 1695, 1611, 1422, 1315, 1283, 1179, 1080,
970, 909, 859, 735 cmÀ1; 1H NMR (200 MHz, CDCl3) d
8.00 (d, J=8.2 Hz, 2H), 7.27 (d, J=8.2 Hz, 2H), 5.60
(ddd, J=15, 8.8, 5.2 Hz, 1H), 5.42 (dd, J=15, 7.8 Hz,
1H), 4.12 (m, 2H), 3.69 (br, 3H), 3.63 (dd, J=10, 1.8
Hz, 1H), 2.79 (t, J=7.9 Hz, 2H), 2.38–1.62 (m, 15H),
1.54 (dd, J=14, 6.8 Hz, 1H), 1.33 (dd, J=14, 6.4 Hz,
1H), 0.91 (d, J=6.6 Hz, 3H), 0.90 (d, J=6.6 Hz, 3H).
Measurement of cAMP production
CHO cells expressing EP2-receptor were cultured in 24-
well plates (1Â105 cells/well). After 2 days, the media
were removed and cells were washed with 500 mL of
Minimum Essential Medium (MEM) and preincubated
for 10 min in 450 mL of assay buffer (MEM containing 1
mM of IBMX, 1% of BSA) at 37 ꢀC. Then reaction was
started with the addition of each test compound in 50
mL of assay buffer. After incubation for 10 min at 37 ꢀC,
the reaction was terminated by addition of 500 mL of
ice-cold 10% trichloroacetic acid. The cAMP produc-
tion was measured by radioimmunoassay using a cAMP
assay kit (Amersham).
(16S)-9-Deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-
trimethylene-!-norPGF2 sodium salt (4aNa). To a stir-
red solution of 4a (2.33 g, 5 84 mmol) in 1,4-dioxane (10
mL) was added aqueous 1 N sodium hydroxide (5.84
mL). The reaction mixture was stirred for 10 min, con-
centrated in vacuo and lyophilized to yield the sodium
salt 4aNa (2.39 g, 97%) as a white amorphous. MS
(FAB, Pos.) m/z 421 (M+H)+; IR (neat) 3398, 2937,
1562, 1407, 1074 cmÀ1; 1H NMR (200 MHz, DMSO-d6)
d 6.00–4.00 (br, 2H), 5.60 (ddd, J=15.4, 8.0, 5.8 Hz,
1H), 5.51–5.23 (m, 3H), 4.09 (q, J=7.3 Hz, 1H), 3.87 (q,
J=6.5Hz, 1H), 3.36 (dd, J=9.2, 2.8 Hz, 1H), 2.30–1.20
(m, 22H), 0.85 (t, J=7.3 Hz, 3H).
Evaluation of uterine motility suppression activity of
EP2 agonists
Fed pregnant rats were anesthetized with urethane (1.5
g/5 mL/kg, sc) and fixed in the dorsal position. A mid-
line incision was made in the lower abdomen and a
small incision was made near the cervical area of the
right or left uterine horn, and a balloon catheter (Oka-
moto Medical Industry, for rats) was inserted between
the uterine wall and the amnion. After suturing the
abdominal incision, the intraballoon pressure was loa-
ded to approximately 5–15 mmHg. A catheter (POLY-
ETHYLENE TUBE SP10, Natsume Seisakusho) was
placed into the femoral vein for the administration of
the test compound. Uterine motility was recorded on a
recticoder (WR3320 or WR3701, GRAPHTEC) via a
pressure transducer (Life Kit DX-360, Nihon Kohden
(16S)-9-deoxy-9ꢀ-chloro-15-deoxy-16-hydroxy-17,17-tri-
methylene-!-norPGF2 L-lysine salt (4aLy). To a stirred
solution of 4a (45.9 g, 110 mmol) in 460 mL of ethanol
was added l-lysine (16.0 g, 110 mmol). To the resulting
suspension, 1600 mL of ethanol was added. The mixture
was heated at 80 ꢀC until the precipitates dissolved, then

1114
K. Tani et al. / Bioorg. Med. Chem. 10 (2002) 1107–1114
Corp.) and a strain pressure amplifier (AP-601G, Nihon
Kohden Corp.). After spontaneous uterine motility was
kept at a stable level, each experiment was started.
Briefly, the vehicle and then the test compound were
intravenously administered successively starting from
lower doses at 10-minute or larger intervals. Before each
dosing, it was confirmed that uterine motility was kept
stable for at least 5 min.
Sugimoto, Y.; Ichikawa, A. Biochem. Biophys. Res. Commun.
1998, 251, 727.
5. Kasugai, S.; Oida, S.; Iimura, I.; Arai, N.; Takeda, K.;
Ohya, K.; Sasaki, S. Bone 1995, 17, 1.
6. Tani, K.; Naganawa, A.; Ishida, A.; Sagawa, K.; Harada,
H.; Ogawa, M.; Maruyama, T.; Ohuchida, S.; Nakai, H.;
Kondo, K.; Toda, M. Bioorg. Med. Chem. (this issue).
7. Okamoto, S.; Katayama, S.; Ono, N.; Sato, F. Tetra-
hedron: Asymmetry 1992, 3, 1525.
8. Newton, R. W.; Raynolds, D. P.; Finch, M. A. W.; Kelly,
D. R.; Roberts, S. M. Tetrahedron Lett. 1979, 41, 3981.
9. Buchmann, B.; Skuballa, W.; Vorbruggen, H. Tetrahedron
Lett. 1990, 31, 3425.
10. Arroniz, C. E.; Gallina, J.; Martınez, E.; Muchowski,
J. M.; Velarde, E.; Rooks, W. H. Prostaglandins 1978, 16, 47.
11. Drugs Future 1986, 11, 660
Uterine motility was evaluated according to the Mon-
tevideo method.16 In the intravenous injection study,
uterine activity was calculated for 5-min periods before
and after administration of the test compound, and
expressed as a percentage relative to pre-treatment values:
Uterine activity (% of pre-value)=(5-min post-dose uter-
ine activity/5-min pre-dose uterine activity)Â100.
12. Thierauch, K. H.; Sturzebecher, C.St.; Schillinger, E.;
¨
¨
Rehwinkel, H.; Raduchel, B.; Skuballa, W.; Vorbruggen, H.
¨
Prostaglandins 1988, 35, 855.
13. Nials, A. T.; Vardey, C. J.; Denyer, L. H.; Thomas, M.;
Sparrow, S. J.; Shepherd, G. D.; Coleman, R. A. Cardiovas-
cular Drug Review 1993, 11, 165.
References and Notes
1. Coleman, R. A.; Smith, W. L.; Narumiya, S. Pharmacol.
Rev. 1994, 46, 205.
14. Gardiner, P. J. Br. J. Pharmac. 1986, 87, 45.
15. Kato, K.; Kawashima, T.; Ishida, M.; Higashi, Y.;
Nakanishi, S.; Oshio, Y.; Kuroda, K.; Shu, B.; Matsumoto,
K.; Sakai, T.; Tsuda, T. Acta. Obst. Gynaec. Jpn. 1973, 25,
476.
16. Caldeyro-Barcia, R.; Sica-Blanco, Y.; Poseiro, J. J.; Gon-
zalez Panizza, V.; Mendez-Bauer, C.; Fielitz, C.; Alvarez, H.;
Pose, S. V.; Hendricks, C. H. J. Pharmacol. Exp. Ther. 1957,
121, 18.
2. (a) Katsuyama, M.; Nishigaki, N.; Sugimoto, Y.; Mor-
imoto, K.; Negishi, M.; Narumiya, S.; Ichikawa, A. FEBS
Lett. 1995, 372, 151. (b) Regan, J. W.; Bailey, T. J.; Pepperl,
D. J.; Pierce, K. L.; Bogardus, A. M.; Donello, J. E.; Fair-
bairn, C. E.; Kedzie, K. M.; Woodward, D. F.; Gil, D. W.
Mol. Pharmacol. 1994, 46, 213.
3. Narumiya, S.; Sugimoto, Y.; Ushikubi, F. Physiol. Rev.
1999, 79, 1193.
4. Katsuyama, M.; Ikegami, R.; Karahashi, H.; Amano, F.;
17. Kiriyama, M.; Ushikubi, F.; Kobayashi, T.; Hirata, M.;
Sugimoto, Y.; Narumiya, S. Br. J. Pharmac. 1997, 122, 217.