4,5-Disubstituted cis-pyrrolidinones as inhibitors of type II 17β-hydroxysteroid dehydrogenase. Part 3. Identification of lead candidate

Article abstract of DOI:10.1016/j.bmcl.2006.06.041

A series of 4,5-disubstituted cis-pyrrolidinones was investigated as inhibitors of 17β-HSD II for the treatment of osteoporosis. Biochemical data for several compounds are given. Compound 42 was selected as the lead candidate.

Full text of DOI:10.1016/j.bmcl.2006.06.041

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4965–4968
4,5-Disubstituted cis-pyrrolidinones as inhibitors of type II
17b-hydroxysteroid dehydrogenase. Part 3.
Identification of lead candidate
Jill Wood, Cedo M. Bagi, Christiana Akuche, Antonietta Bacchiocchi, Jeremy Baryza,
Marie-Luise Blue, Catherine Brennan, Ann-Marie Campbell, Soongyu Choi,
James H. Cook, Patricia Conrad, Brian R. Dixon, Paul P. Ehrlich, Todd Gane,
David Gunn,* Ted Joe, Jeffrey S. Johnson, Jerold Jordan, Richard Kramss, Peiying Liu,
´
Joan Levy, Derek B. Lowe, Ian McAlexander, Reina Natero, Aniko M. Redman,
William J. Scott, Christopher Town, Ming Wang, Yamin Wang and Zhonghua Zhang
Bayer Research Center, Bayer HealthCare, Pharmaceutical Division, 400 Morgan Lane, West Haven, CT 06516, USA
Received 24 May 2006; revised 9 June 2006; accepted 12 June 2006
Available online 27 June 2006
Abstract—A series of 4,5-disubstituted cis-pyrrolidinones was investigated as inhibitors of 17b-HSD II for the treatment of osteo-
porosis. Biochemical data for several compounds are given. Compound 42 was selected as the lead candidate.
Ó 2006 Elsevier Ltd. All rights reserved.
The goal of this project was to identify an orally active
inhibitor of 17b-hydroxysteroid dehydrogenase, type II
(17b-HSD II),1 an enzyme known to oxidize many
steroid substrates. We aimed to test the hypothesis that
such an inhibitor could increase local bone concentra-
tions of estradiol resulting in maintenance of bone qual-
ity and strength in an osteoporosis model.2 We
previously reported on the discovery of 4,5-disubstituted
cis-pyrrolidinone 1 as a 17b-HSD II inhibitor.3 Here, we
discuss further refinement of the series to identify a suit-
able compound for in vivo studies.
of interest were resolved into enantiomers by chiral
chromatography, and evaluated in plasma exposure
studies.
The relative stereochemistry illustrated in Figure 1 was
critical for activity against 17b-HSD II.3 Further inves-
tigations were needed to understand the SAR involving
the two phenyl groups. Related regioisomers of the
2-phenylthiophene show only slightly lower activity
against 17b-HSD II, suggesting that the thiophene is
in a large pocket or exposed to solvent (Table 1).
Although benzothiophene 4 showed moderate activity,
it was rapidly degraded by liver microsomes and was
not pursued.
The screening cascade began with a biochemical assay
that measured conversion of NAD to NADPH by
recombinant human 17b-HSD II. Next, inhibition of
estradiol to estrone conversion by intact osteoblast
MG63 cells was measured. Advanced compounds were
evaluated against 11b-HSD II, the closest structurally
related hydroxysteroid dehydrogenase, and for absence
of binding in several hormone receptor assays. Analogs
Due to the relatively large tolerance to modifications of
the D-ring, we sought to improve potency along with
OH
N
O
S
D
A
Keywords: Pyrrolidinone; Hydroxysteroid dehydrogenase; Inhibitor;
Osteoporosis.
1
*
Corresponding author. Tel.: +1 203 812 6296; fax: +1 203 812
6182; e-mail: david.gunn.b@bayer.com
Figure 1. Pyrrolidinone 1 shown with relative stereochemistry.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.041

4966
J. Wood et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4965–4968
Table 1. Thiophene substitution pattern4
Table 2. D-ring analogs4
O
N
O
N
OH
Ar
OH
S
R
Compound
Ar
17b-HSD II
IC50 (lM)
Compound
R
17b-HSD II
IC50 (lM)
S
1
0.10
1
8
0.10
0.12
0.28
0.58
0.34
S
2
0.58
F
S
9
3
4
0.32
0.59
F3C
10
11
S
F
OMe
some of the biopharmaceutical properties, specifically
solubility, by varying this position (Table 2). These ana-
logs were prepared either by direct addition of an aryl
thiophene to 5 as previously reported,5 or, via Suzuki
coupling of 6 with various boronic acids (Scheme 1).
12
13
14
15
16
17
18
19
20
21
22
23
0.15
1.32
0.48
4.11
0.14
0.03
1.28
0.15
0.30
0.38
0.33
1.19
NMe2
CO2H
OH
A wide variety of substituents such as mono-, di-, and
tri-substituted phenyls, naphthyls, and biphenyls were
investigated. Loss of activity was seen with some large
or more lipophilic groups, however reasonable activity
was maintained over a very diverse set of analogs.
CF3
F
Substituents at the meta-position of the phenyl ring did
not greatly affect activity. Both electron-withdrawing
(10) and electron-donating (11 and 12) groups caused
a slight drop-off in activity. At the para-position, hydro-
philic substituents generally showed much better activity
than lipophilic groups. The para-position also seemed to
be readily oxidized by liver microsomes. In comparing
the active enantiomers of 1 (para-H) and 16 (para-F)
after 20 min in rat microsomes, we found only 11% of
(À)-1 remaining while there was 94% of (À)-16 remain-
ing. The putative metabolite, phenol 14, maintains mod-
erate activity.
CN
Cl
N
Most of the pyrrolidinones were poorly soluble in water.
To improve water solubility, nitrogen heterocycles and
acid derivatives were synthesized. Acids 13 and 21 were
more soluble but unfortunately were much less active in
the cell assay. The 2- and 3-pyridyl compounds, 19 and
20, respectively, maintained activity in the cell however,
N
CO2H
they were still poorly soluble as
a free base.
S
Nevertheless, the pyridyl analogs had the option for salt
formation at a later stage to aid solubility.
F
Linking groups were investigated between the thiophene
and the phenyl group (22–26). Again we found that a
wide variety of groups were tolerated, however sulfone
(continued on next page)

J. Wood et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4965–4968
4967
Table 2 (continued)
activity. An ortho-fluorine generally improved the activ-
ity 10-fold (30–35) (Table 3).
Compound
R
17b-HSD II
IC50 (lM)
In summary, biochemical potency against 17b-HSD II
was achieved with analogs from a novel, non-steroidal
structural class (Tables 1–3). The majority of enzyme
active compounds maintained activity in the cell estradiol
conversion functional assay. The pyrrolidinones did not
show appreciable binding to the estrogen, androgen or
other steroid receptors and no tangible activity was seen
in the counter-screens (11 b-HSD II, 17b-HSD I, and
17b-HSD III).
O
S
24
25
26
0.16
0.07
0.95
O
O
S
N
O
O
O
The SAR in the pyrrolidinone series differed greatly
between human and rat enzyme; this was not unexpected
since rat 17b-HSD II has only a 60% homology to the
human protein.7 Since SAR from biochemical assays
using monkey enzyme correlated with the human re-
sults, it was felt a proof of principle experiment should
be conducted in Cynomolgus macaques. In preparation
for this study, select analogs were separated into enanti-
OH
O
N
N
O
O
S
Br
S
Br
H
Br
THF, n-BuLi
-70 o
(67%)
C
5
6
OH
R
Table 3. Substituted A-ring analogs4
N
O
S
B(OH)
2
HO
S
N
Pd(Ph3P)4
Na2CO3, DME
R1
O
R
7
Scheme 1. Late stage modification of D-ring.
R2
Example
R1
R2
17b HSD II
IC50 (lM)
linkers were the most potent in the biochemical assay.
Phenyl sulfones (24) gave compounds that were potent
in the biochemical and cell assays. Sulfonamide 25 was
very active against the enzyme, but lost activity 10-fold
in the cellular assay.
30a
2-F
0.01
0.06
0.06
0.12
0.04
0.19
0.48
2.55
0.76
0.88
31
32
33
34
35
36
37
38
39
2-F
F
Turning to the preparation of A-ring analogs, few
changes could be made using the initial route.5 However,
an alternative synthesis based on an intramolecular
Michael addition was developed that allows for the pro-
duction of A-ring analogs (Scheme 2).6
O
S
2-F
O
2-OCH3
2-OCH3
2-OCH3
3-F
F
In vitro evaluation of this series indicated that substitu-
tion at the ortho-position by small groups was neutral to
beneficial while substitutents at the meta- and para-posi-
tions, for the most part diminished the 17b-HSD II
O
S
O
O
S
O
Cl
O
OH
N
H
CO Et
2
HCl
1)
R2
F
F
EDCI, DMAP
2) LiHMDS
27
O
3-OCH3
4-F
O
O
N
N
1) H2, Pd/C
2) LiBH4
OEt
OH
R2
R2
4-OCH3
28
Scheme 2. Intramolecular Michael addition route.
29
a Single enantiomer.

4968
J. Wood et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4965–4968
Table 4. Resolved enantiomers9
parent. Based on its overall in vitro profile and the
Cynomolgus macaque exposure profile, compound 42
was selected as the lead to establish non-human primate
proof-of-concept.
HO
S
N
R1
O
R2
Acknowledgments
Example
R1
R2
H
17b-HSD
II IC50 (lM)
MG63 cell
IC50 (lM)
We thank L. Musza for his structural NMR work. We
also thank Dr. Rolf Grosser and Dr. J. Brice for chiral
HPLC method development and support. We thank Dr.
D. Mueller and G. Medvedeff for providing ample hu-
man 17b-HSD II enzyme for the biochemical assay.
We thank T. Alebic, S. Jaworski, and M. Prevost for
plasma data and D. Young and L. Zadjura for micro-
somal stability data. We thank Dr. Y.-C. Tseng and J.
Luong for formulation support. We thank Dr. Robert
Brommage and his team at Wake Forest University
for collecting the primate plasma samples.
40
41
42
43
44
0.12
0.22
0.05
0.03
0.06
0.10
0.06
0.09
0.03
0.48
F
F
2-F
H
N
N
2-F
H
Supplementary data
O
S
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.06.041.
O
Table 5. Exposure studies
References and notes
Sprague–Dawley
rat 20 mg/kg, po
Cynomolgus macaque
30 mg/kg, po
1. Poirer, D. Curr. Med. Chem. 2003, 10, 453.
2. Takano-Yamamoto, T.; Rodan, G. A. Proc. Natl. Acad.
Sci. U.S.A. 1990, 87, 2172.
3. Gunn, D.; Akuche, C.; Baryza, J.; Blue, M.-L.; Brennan,
C.; Campbell, A.; Choi, S.; Cook, J.; Conrad, P.; Dixon, B.;
Dumas, J.; Ehrlich, P.; Gane, T.; Joe, T.; Johnson, J.;
Jordan, J.; Kramss, R.; Liu, P.; Levy, J.; Lowe, D.; Mc
Alexander, I.; Natero, R.; Redman, A. M.; Scott, W.; Seng,
T.; Sibley, R.; Wang, M.; Wang, Y.; Wood, J.; Zhang, Z.
Bioorg. Med. Chem. Lett. 2005, 15, 3053.
AUCnorm (g h/L)
Cmax, norm (g/L)
76.1
28.7
179
36.4
Compound 42 was dosed as a suspension in CMC/Tween 80/Rasp-
berry crystal light (0.5/0.5/99).
omers using chiral HPLC.8 Compounds as single enan-
tiomers were prioritized in rat serum exposure studies
(po) before moving to Cynomolgus macaque expo-
sure studies. Representative compounds are shown in
Table 4.
4. In Tables 1–3 biochemical data are given, compounds are
racemic unless otherwise noted.
5. (a) Hartwig, W.; Born, L. J. Org. Chem. 1987, 52, 4352; (b)
Wood, J. E.; Baryza, J. L.; Brennan, C. R.; Choi, S.; Cook,
J. H.; Dixon, B. R.; Ehrlich, P. P.; Gunn, D. E.;
McAlexander, I.; Liu, P.; Lowe, D. B.; Redman, A. M.;
Scott, W. J.; Wang, Y. US Patent 6,784,167 B2, 2004.
6. Cook, J. H.; Baryza, J.; Brennan, C.; Lowe, D.; Wang, Y.;
Redman, A.; Scott, W. J.; Wood, J. E. Tetrahedron Lett.
2005, 46, 1525.
7. The human and the rat 17b-HDS II (Accession # NM_
002153 and BC088134, respectively) are 60% homologous
at the protein level using Omiga software.
8. Alcohol 29 could also be resolved by HPLC and used as a
late stage intermediate.
Compound 43, when dosed as the hydrochloride salt,
had very low exposure level in the rats and so was not
considered further. Compounds 40, 41, 42, and 44 ad-
vanced to plasma exposure studies in macaques. In all
cases except for 42, the exposure level was 2–10-fold
lower in macaques than in rats, based on AUC. A par-
allel trend was observed in microsomal incubations, that
is, compounds 40, 41, and 44 were more stable in rat
microsomal assays than in the macaque, while 42 was
more stable in the macaque assay. Table 5 outlines the
plasma levels obtained for 42. The hydrochloride salt
of 42 was also profiled in plasma studies and did not
show significantly different plasma levels than the
9. Absolute stereochemistry as shown was determined by
NMR studies on the methoxy-phenylacetic esters of both
enantiomers of one analog in the cis-pyrrolidinone series.