Synthesis and in vitro evaluation of human FP-receptor selective prostaglandin analogues

Article abstract of DOI:10.1016/S0960-894X(00)00273-0

The in vitro evaluation of a series of saturated prostaglandins revealed that compounds with omega chin aromatic rings retain nanomolar potency for the human prostaglandin F receptor (hFP receptor), exemplified by compound 8. In contrast, the double bonds are required for activity in the series with an acyclic omega chain as in PGF(2α). (C) 2000 Published by Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00273-0

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1519±1522
Synthesis and In Vitro Evaluation of Human FP-receptor
Selective Prostaglandin Analogues
Mitchell A. deLong,* Jack Amburgey, Cynthia Taylor, John A. Wos,
David L. Soper, Yili Wang and Renee Hicks
Procter & Gamble Pharmaceuticals, Mason, OH 45040, USA
Received 7 December 1999; accepted 25 April 2000
AbstractÐThe in vitro evaluation of a series of saturated prostaglandins revealed that compounds with omega chain aromatic rings
retain nanomolar potency for the human prostaglandin F receptor (hFP receptor), exempli®ed by compound 8. In contrast, the
double bonds are required for activity in the series with an acyclic omega chain as in PGF2a. # 2000 Published by Elsevier Science
Ltd.
Chemistry
Introduction
All aromatic compounds were synthesized using a mod-
The naturally-occurring prostaglandin PGF2a has a
wide range of pharmacological properties.1 The primary
mediator of its activity is thought to be the prostanoid
receptor F, the FP receptor, following the nomenclature
of Coleman.2 FP-mediated events are thought to include
the reduction of intraocular pressure,3 the induction of
estrus in some species and the induction of labor. However,
while naturally-occurring prostaglandins are characterized
by their activity against a particular prostaglandin recep-
tor, they are not speci®c for any given receptor. Further,
con¯icting data exist in the literature on both the tissue
and receptor selectivity of PGF2a analogues.4 The confu-
sion may now be dispelled with the cloning and expression
of the cell-surface receptors which are thought to mediate
the e€ects of the prostaglandins.5
i®cation of the original Corey prostaglandin route as out-
lined in the Scheme for compounds 7 and 8. Thus meta-
tri¯uoromethylcinnamic acid (9) was saturated, then ester-
i®ed with TMS-diazomethane to give the ester 10 in good
yield. Treatment of methyl dimethylphosphonate with n-
butyllithium followed by 10 provided ketophosphonate 11
in yields ranging from 50±80%. Ketophosphonate 11 was
then treated with sodium hexamethyl-disilazane followed
by condensation with Corey Lactone (12) to give the a,b-
unsaturated ketone 13 in approximately 50% yield. Ste-
reoselective reduction of C15 was realized by treatment
of 13 with a premixed solution of BINAL and lithium
aluminum hydride to give the C15 alcohol 14 as the only
isomer isolated. Deprotection of the benzoate ester was
accomplished with potassium carbonate in methanol to
give diol 15. The C11 and C15 alcohols were subsequently
protected as their tetrahydropyranyl (THP) ethers to
give 16. Reduction of 16 with diisobutylaluminum
hydride gave the corresponding lactol which was, without
puri®cation, condensed with the sodium salt of the stabi-
lized phosphonium ylide of 4-carboxybutyl triphenyl-
phosphonium bromide to install the prostaglandin a-
chain with high Z-stereoselectivity (9:1 Z:E) providing 17.
Removal of the THP ethers with acetic acid and water
gave 7. Saturation of the 5,6 and 13,14 double bonds
gave the tetrahydro analogue 8.
As part of a pharmacophore minimization e€ort in our
bone anabolics group,6 we examined the contribution that
the double bonds had to the potency of PGF2a and its
analogues. Known hFP receptor ligands were thus con-
verted to their corresponding saturated counterparts and
tested in hFP receptor functional7 (EC50) and radioligand-
displacement binding8 (IC50) assays. Data are averages
of triplicate measurements.
The aliphatic compounds PGF2a (1) and 16,16, dime-
thyl PGF2a (3) were purchased (Cayman Chemical, Ann
Arbor, MI) and tested without further puri®cation. The
*Corresponding author. Tel.: +1-513-627-2626; fax: +1-513-627-
0026; e-mail: delong.ma@pg.com
0960-894X/00/$ - see front matter # 2000 Published by Elsevier Science Ltd.
PII: S0960-894X(00)00273-0

1520
M. A. deLong et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1519±1522
saturated aliphatic compounds 2 and 4 were synthesized
by hydrogenation of the appropriate precursor with H2
over 5% Pd/C in ethanol.
in the straight chain series the two double bonds may be
necessary for high anity binding.
Cloprostenol (5),9 a 16-phenoxy-substituted ligand of
the hFP receptor, upon hydrogenation, provides tetra-
hydrocloprostenol, (6) which maintains its potency at
the human FP receptor with an EC50 of 9 nM and has
an even better selectivity pro®le for the hFP receptor
versus the hTP, hEP1 and hEP3 receptors (Table 1).
Similarly, the 17-phenyl analogue, (7), and its tetra-
hydro version (8) gave hFP EC50's between 64 and 85
nM; again the saturated analogue has the better receptor
selectivity pro®le. The syntheses of these analogues was
accomplished as outlined in the Scheme.
Results
As shown in Table 1, both PGF2a(1), and 16,16-dimethyl
PGF2a(3) have nanomolar activity at the hFP receptor, in
agreement with literature values on tissues. Also con-
sistent, the saturated analogue of PGF2a(2) has an IC50 of
522 nM and an EC50 of 5000 nM, a 200-fold decrease in
activity at the receptor. A similar decrease is observed
upon hydrogenation of 16,16-dimethyl PGF2a(4). Thus,
Table 1. hFP Binding and excitatory receptors' functional data for PGF2a and derivatives
Compound
IC50 hFP
2.5 nM
EC50 hFP
EC50 hTP
EC50 hEP1
EC50 hEP3
79 nM
(75%a)
1400 nM
(95%b)
600 nM
(80%c)
2300 nM
(50%d)
522 nM
16 nM
5000 nM
(50%a)
30,000 nM,
(100%b)
>10,000 nM,
(<50%c)
>20,000 nM,
(<50%d)
90 nM
(150%a)
24 nM,
(95%b)
300 nM,
(80%c)
2000 nM,
(130%d)
>1000 nM
1.0 nM
1.0 nM
10 nM
>20,000 nM
(<20%a)
>20,000 nM
(<20%b)
>20,000 nM
(<20%c)
>20,000 nM
(<20%d)
10 nM
(100%a)
2400 nM
(100%b)
760 nM
(80%c)
5000 nM
(75%d)
9 nM
(100%)
8000 nM
(90%b)
19,000 nM
(100%c)
>20,000 nM
(<50%d)
64 nM
(87%a)
2100 nM
(50%b)
>20,000 nM
(<50%c)
>20,000 nM
(<50%d)
8.6 nM
85 nM
(90%a)
>20,000 nM
(<50%b)
30,000 nM
(100%c)
>20,000 nM
(< 50%d)
a% of control ligand, ¯uprostenol.
b% of control ligand, U46619.
c% of control ligand, 17-f-PGE2.
d% of control ligand, sulprostone.

M. A. deLong et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1519±1522
1521
Structure±Activity Relationship
tion in all cases resulted in molecules that did not
appreciably bind to the hTP receptor, and only one
example had measurable activity at the hEP3 receptor, in
contrast to the unsaturated series,10 where they were found
to activate both of these receptors in tissue preparations.
Further, activity at the hEP1 receptor was greatly
reduced for all of the molecules in this series. Thus the
initial selectivity trend observed was continued in these
analogues. The potency at the hFP receptor, however,
varied widely from compound to compound. The ability
To investigate the breadth of this ®nding, we initiated a
limited exploration of the structure±activity relationship
(SAR) of the saturated prostaglandin analogues. Using
the synthetic route outlined in Scheme 1, a number of
aromatic prostaglandin analogues were synthesized and
tested. Compounds with initial binding IC50s >2000 nM
(hFP) were considered inactive and not further tested.
Data are shown in Table 2. Removal of the unsatura-
Scheme 1. (a) H2, 5% Pd/C, EtOAc; then TMSCHN2, MeOH (b) n-BuLi, THF, 0 ꢀC (c) NaHMDS, DME, (d)(S)-BINAL-H, LiAlH4, THF, (e)
K2CO3, MeOH overnight (f) 2.2 equiv DHP, cat. p-TSA, CH2Cl2 (g) DIBAL-H, 78 ꢀC, then 2.1 equiv NaHMDS and Ph3P+(CH2)4COOH Br (h)
1:1:1 THF: H2O: HOAc; (i) H2, 5% Pd/C, EtOAc.
Table 2. Human-FP receptor data for 17-phenyl analogues
Entry
X
IC50 hFP
(Binding)a
IC50 hEP1
(Binding)b
IC50 hEP3
(Binding)b
IC50 TP
(Binding)c
18
19
20
21
22
23
24
25
26
27
28
29
30
31
H
o-F
m-F
p-F
82 nM
76 nM
42 nM
210 nM
2140 nM
8.6 nM
3630 nM
630 nM
4400 nM
5890 nM
1620 nM
290 nM
540 nM
210 nM
1100 nM
3100 nM
10,000
>10,000
>10,000
3700 nM
>10,000
Not Tested
>10,000
Not Tested
>10,000
Not Tested
Not Tested
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
Not Tested
>10,000
Not Tested
>10,000
Not Tested
Not Tested
>10,000
3450
o-CF3
m-CF3
p-CF3
o-CH3
p-CH3
o-OCH3
m-OCH3
p-OCH3
m-CH2CH3
m-CN
Not Tested
>10,000
Not Tested
10,000
Not Tested
Not Tested
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000

1522
M. A. deLong et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1519±1522
of the molecule to displace PGF2a from its receptor
diminishes rapidly with the increasing aryl substituent
size. The compound class containing a single ¯uoro
substituent, (19, 20 and 21) while active at all three
positions on the aromatic ring, was slightly less potent at
the para position where it also shows some undesirable
hEP1 activity. The other substituents show a divergent
SAR; e.g. the ortho methyl is more potent than the para,
while the para-methoxy is more potent than its ortho
analogue. Finally, electronics may play a role, as the m-F
(42 nM) is more potent than H (82 nM), but the m-CN at
210 nM is not. Thus the characteristics of the terminal
''pocket'' where the aromatic ring of the ligands ®t plays
a major role in determining their anity.
100, 1125. (b) Crankshaw, D. J.; Dyal, R. Can. J. Physiol.
Pharmacol. 1994, 72, 870. (c) Crankshaw, D. J.; Dyal, R. J.
Clin. Endrocrinol. Metab. 1992, 75, 1027. (d) Garcia-Villar, R.;
Marnet, P. G.; Laurentie, M. P.; Toutain, P. L. Am. J. Vet.
Res. 1985, 46, 841. (e) Featherstone, R. L.; Robinson, C.;
Holgate, S. T.; Church, M. K. Naunyn-Schiedebergs Arc.
Pharmacol. 1990, 341, 439. (f) McKenni€, M. G.; Norman, P.;
Cuthbert, N. J.; Gardiner, P. J. Br. J. Pharmacol. 1991, 104,
585.
2. (a) Coleman, R. A.; Humphrey, P. P. A.; Kennedy, I.;
Lumley, P. Trends Pharmacol. Sci. 1984 5, 303. (b) Coleman,
R. A.; Smith, W. L. Narumiya, S. Pharmacol. Rev. 1994, 46,
205.
3. Toris, C. B.; Camras, C. B.; Yablonski, M. E. Ophthalmol-
ogy 1993, 100, 1297.
4. Our compilation was published: Hop®nger, A. J. Wang, S.;
Tokarski, J. S.; Jin, B. Q.; Albuquerque, M.; Madhav, P. J.;
Duraiswami, C. J. Am. Chem. Soc. 1997, 119, 10509.
5. Kliewer, S. A.; Lenhard, J. M.; Willson, T. M.; Patel, I.;
Morris, D. C.; Lehmann, J. M. Cell 1995, 83, 813.
Conclusions
In conclusion, it has been found that the double bonds
present in the naturally-occurring prostaglandin PGF2a
are required for its nanomolar binding at the hFP recep-
tor. Surprisingly, neither of these double bonds are
necessary for similar potency levels in analogues wherein
an aromatic ring has replaced the last four atoms of the
prostaglandin skeleton. While compounds containing
substituted aromatic rings were also found to be active at
the hFP receptor, the SAR shows a relatively narrow
scope for reasonable potency. The data further suggest
an improvement in the compounds' selectivity is simul-
taneously achieved by removal of the unsaturation,
especially toward the thromboxane (hTP) receptor. The
SAR is not unambiguous, however, as both size and
electronic interactions appear to be contributing to the
®nal binding values. Finally, receptor modeling studies
of the seven transmembrane domain region of the hFP
receptor may help to rationalize this SAR. These data
suggest that new FP-receptor selective analogues with a
simpli®ed skeleton can be found in a saturated series
which might also show improved selectivity toward the
hFP receptor over the other excitatory receptors.
6. (a) Hartke, J. R.; Jankowsky, M. L.; deLong, M. A.;
Soehner, M. E.; Jee, W. S. S.; Lundy, M. W J. Bone Min. Res.,
1999, 14, S207. (b) deLong, M. A.; Miley, C. J.; Amburgey, J.
S.; Roof, S. L.; Pierce, S.; Lundy, M. W. Abstracts Of Papers,
Part 2, 213th National Meeting of the American Chemical
Society, San Francisco, CA; American Chemical Society:
Washington, DC, 1997, Abstract 135-MEDI. (c) Soper, D. L.;
Brann, M.; Amburgey, J.; Miley, C.; Wos, J.; Wang, Y.;
Hartke, J.; deLong, M. A. Abstracts Of Papers, Part 1, 215th
National Meeting of the American Chemical Society, Dallas,
TX; ACS: Washington, DC, 1998, Abstract A050-MEDI.
7. Compounds evaluated for functional activity in RAT-1
cells, transiently-transfected with a human prostanoid receptor
(hFP, hEP3, hEP1, or hTP) having a constitutively-expressed
b-galactosidase reporter gene stably-transfected therein. EC50
shown is after timed hydrolysis of galactosidase pseudsub-
strate. For details: Messier, T.; Dorman, C.M.; Brauner, H.;
Eubanks, D.; Brann, M. R. Pharmacol. Toxicol. 1995, 76, 308.
8. Binding assays: Compounds evaluated for their ability to
displace [3H]PGF2a (hFP receptor assay) [3H]PGE2 (hEP1;
EP3) [3H]U46619 (hTP) in membrane preparations isolated
from CHO-KI cells stably transfected with the receptor. The
radiolabel (5 nM) membrane and test compound were incu-
bated together 1h; plates harvested and DPM's counted.
9. Dukes, M.; Russell, W.; Walpole, A. L. Nature 1974, 250,
330.
10. No, K.; Resul, B.; Nilsson, B.; Thor, M.; Liljebris, C.;
Selen, G.; Sternschantz, J. XIV International Symposium on
Medicinal Chemistry, Maastricht, Netherlands, 1996, Abstract
#2.29.
References and Notes
1. (a) Rudolph, M. I.; Reinicke, K.; Cruz, M. A.; Gallardo,
V.; Gonzalez, C.; Bardisa, L. Br. J. Obsetet. Gynaecol. 1993,