Oxazolones as potent inhibitors of 11β-hydroxysteroid dehydrogenase type 1

Article abstract of DOI:10.1016/j.bmcl.2007.06.054

2,5,5-Trisubstituted oxazolones were identified as potent inhibitors of 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1). The synthesis, structure-activity relationship and metabolic stability of these compounds are presented.

Full text of DOI:10.1016/j.bmcl.2007.06.054

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4837–4840
Oxazolones as potent inhibitors of
11b-hydroxysteroid dehydrogenase type 1
Lori Sutin,* So¨ren Andersson, Lars Bergquist, Victor M. Castro, Eva Danielsson,
Stephen James, Martin Henriksson, Lars Johansson, Christina Kaiser,
´
Katarina Flyren and Meredith Williams
Biovitrum AB, SE-112 76 Stockholm, Sweden
Received 30 May 2007; revised 13 June 2007; accepted 13 June 2007
Available online 20 June 2007
Abstract—2,5,5-Trisubstituted oxazolones were identified as potent inhibitors of 11b-hydroxysteroid dehydrogenase type 1 (11b-
HSD1). The synthesis, structure–activity relationship and metabolic stability of these compounds are presented.
Ó 2007 Elsevier Ltd. All rights reserved.
11b-Hydroxysteroid dehydrogenase type 1 (11b-HSD1)
is a biological target that has attracted much interest
over the past few years.1–3 This membrane-bound en-
zyme acts as an NADPH-dependent reductase and di-
rectly controls the interconversion of inactive cortisone
to the receptor-active glucocorticoid cortisol (Fig. 1).4
OH
OH
OH
O
O
11
β
β
-HSD1
-HSD2
O
HO
OH
H
H
11
H
H
H
H
O
O
cortisone
cortisol
It has been suggested, based on both animal and human
studies, that glucocorticoid excess in tissues such as
liver, adipose and skeletal muscle might be a contribut-
ing factor towards the onset of the metabolic syn-
drome.5,6 Studies have shown that 11b-HSD1
knockout mice, compared to wild-type, demonstrate
resistance to the development of diet-induced metabolic
syndrome as well as displaying a lower fasting blood
glucose level.7 Furthermore, this phenotype was found
to be closely tied to metabolic effects in the adipose tis-
sue, where glucocorticoid-inducible transcripts encoding
for leptin and TNF-a were reduced.8 The enzyme, 11b-
HSD1, therefore is thought to play a crucial role in dis-
eases associated with the metabolic syndrome, such as
obesity and diabetes. The type II isoform, 11b-HSD2,
catalyzes the inactivation of cortisol using NAD as a
cofactor. Whereas 11b-HSD1 is primarily expressed in
the liver and adipose tissue, 11b-HSD2 is located mainly
Figure 1. Interconversion of cortisone and cortisol by 11b-HSD1 and
11b-HSD2 enzymes.
in the kidney.9 Any potential drug aimed at inhibiting
11b-HSD1 should be selective towards this isoform,
since inappropriate inhibition of 11b-HSD2 may lead
to adverse side effects.10
Work from our laboratories has demonstrated that
2,5,5-trisubstituted 1,3-thiazol-4(5H)-ones (Fig. 2) are
inhibitors of 11b-HSD1.11 In particular, highly potent
compounds were obtained with aliphatic mono-, bi- or
tricyclic groups for R1, in combination with short
alkyl chains for R2 and R3.
O
R2
N
Keywords: 11b-HSD1; Metabolic syndrome; Oxazolones.
Corresponding author. Tel.: +46 8 6973540; fax: +46 8 6972320;
e-mail: lori.sutin@biovitrum.com
Present address: S*Bio Pte Ltd, 1 Science Park Road, #05-09 The
Capricorn, Singapore Science Park II, Singapore 117528, Singapore.
R1
*
R3
N
S
H
Figure 2. General structure of 11b-HSD1 active thiazolones.
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.054

4838
L. Sutin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4837–4840
As an extension of this work, the analogous 1,3-oxazol-
4(5H)-ones (henceforth abbreviated as oxazolones) were
investigated. Synthesis of the 2-alkylamino-5,5-dial-
kyloxazolones was achieved in 5 steps from the corre-
sponding ketones (Scheme 1). Addition of potassium
cyanide to the ketones in 20% sulfuric acid gave the cya-
nohydrins.12 The a-hydroxy esters were formed by
hydrolysis of the cyanohydrins followed by esterifica-
tion, and were then reacted with guanidine in the
presence of base to give the corresponding 2-amino-
5,5-dialkyloxazolones.13 These key intermediates were
then treated with primary amines14 under microwave
irradiation (typically at 180 °C for 1 h) to give the final
products in modest yields (10–40%).15
bi- or tricycloalkylamines in the 2-position. The results,
summarized in Table 1, include binding and inhibitory
activity in human 11b-HSD1 SPA16 and adipocyte17 as-
says, respectively, as well as human and rat microsomal
intrinsic clearance data.18
Results from the SPA assay reveal that several of the
oxazolones are highly potent. For R1, it appears that
the potency is slightly enhanced with increased lipo-
philicity, manifested by the high activities of the ada-
mantyl derivatives 8, 10 and 12–14. In accordance
with this trend, the introduction of a 3-hydroxy sub-
stituent on the 1-adamantyl group (16 and 17) resulted
in decreased activities when compared to the corre-
sponding 1-adamantyl compounds 12 and 13. For
R2 and R3, 5,5-diethyl, 5-spiropentyl, 5-methyl-5-triflu-
oromethyl and 5-isopropyl-5-methyl substitution give
essentially equipotent species within each R1-series.
The compounds employing heterospiro substitution
at the oxazolone C5 position (2, 3 and 15) exhibit
lower binding activities, most markedly the N-benzyl-
4-piperidine analogue 2. To confirm selectivity, com-
pound 5 was tested against 11 b-HSD2 and showed
no appreciable activity (Ki > 10 lM).
Based on our experience of the SAR from the thiazol-
ones (vide supra), a series of oxazolones (1–17) were syn-
thesized incorporating alkyl, fluoroalkyl, alkylspiro or
heteroalkylspiro moieties in the 5-position and mono-,
O
R2
R3
O
a
b, c
NC
HO
R2
R3
EtO
HO
R2
R3
The activities in adipocytes follow a similar trend as the
SPA assay potencies, however, more dramatic
differences are observed. This might indicate that some
of the compounds are susceptible to poor cell permeabil-
ity and/or efflux. For a majority of the compounds
in Table 1, intrinsic clearance in rat microsomes is
significantly lower than in human microsomes. How-
ever, an exception to this trend is observed with the
O
O
N
R2
R3
d
e
N
R2
R3
R1
N
O
H2N
O
H
Scheme 1. Reagents and conditions: (a) KCN, 20% H2SO4, rt; (b)
concd HCl, reflux; (c) EtOH, HCl, reflux; (d) guanidine hydrochloride,
K2CO3, EtOH, reflux; (e) R1NH2, EtOH, MW 180 °C.
Table 1. SAR of 2-cycloalkylaminooxazolones
O
N
R2
R3
R1
N
O
H
Compound R1
R2
R3
R2, R3
SPA Kia (nM) Adipocytes IC50 (nM)
a
a,b
CLint (lL/
min/mg)
Human Rat
1
2
Cycloheptyl
Me iPr
22
–(CH2)2–N(Bn)–(CH2)2– >10,000
1866
nd
5
52
nd
nd
62
nd
125
50
29
nd
81
51
108
94
nd
28
5
Cycloheptyl
Cycloheptyl
Cyclooctyl
nd
nd
6
3
–(CH2)2–O–(CH2)2–
1433
17
10
17
14
7
nd
4
Me CF3
Et
117
nd
5
Cyclooctyl
Cyclooctyl
Cyclooctyl
Et
Me iPr
nd
10
9
6
61
7
–(CH2)4–
454
59
8
2-Adamantyl
2-Adamantyl
2-Adamantyl
2-Adamantyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
Me CF3
Et
5
9
Et
Me iPr
12
6
76
14
nd
28
13
26
20
nd
6
10
11
12
13
14
15
16
17
–(CH2)4–
18
4
214
nd
Et
Et
Me iPr
6
24
nd
–(CH2)4–
–(CH2)2–O–(CH2)2–
4
43
39
25
273
358
66
1-(3-Hydroxyadamantyl) Et
1-(3-Hydroxyadamantyl) Me iPr
Et
5
3
9
a For chiral compounds, data reported for mixture of stereoisomers.
b Microsomal clearance.

L. Sutin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4837–4840
4839
Table 2. 11b-HSD1 SPA binding affinities of oxazolones with aryl-
containing amines in the 2-position
derivative (19) is more active than its one-carbon homo-
logue (18), but both are considerably less potent than
their cycloalkyl analogues presented in Table 1.
Increased activity is observed with the derivatives incor-
porating indane-2-amine (20), and more significantly,
(S)-methylbenzylamine (21).
Compound Ki (nM)
O
N
R2
R3
R1
N
O
H
A direct comparison of the (S)-methylbenzylamine
derivatives 22 and 23 to their (R)-stereoisomers 24 and
25 clearly indicates that there is a stereochemical prefer-
ence at the binding site for the (S)-configuration. This
observation was also made for corresponding com-
pounds in the thiazolone series.20 Encouraged by this re-
sult, a series of oxazolones based on commercially
available (S)-alkylbenzylamines were investigated.
Notably, replacing the methyl group in 21 with either
ethyl (26) or hydroxymethyl (27) results in loss of
activity. Poor potency is also shown for the naphthyl
derivative 28. The series of substituted (S)-methylben-
zylamines 29–31 reveal that the binding affinity is highly
dependent on the substituent. The most potent
compounds in the series are the ortho-CF3 derivatives
31–33. Notably, the N-benzyl-4-piperidine compound
33 exhibits surprisingly high activity when compared
to the analogous species 23 and 25. Even more striking
is the difference in activity between 33 and its cycloheptyl
analogue 2 (Table 1). This suggests that the SAR for the
compounds with aryl-containing amines in the oxazo-
lone 2-position cannot be directly extrapolated from
that of the 2-cycloalkylamine series presented in Table 1.
R1
R2
R3
N
N
Ph
Ph
Ph
18
nd
nd
N
6451
19
1761
nd
nd
nd
nd
N
2-Indane
20
879
N
21
369
22a
77
23
676
Ph
Ph
N
N
nd
24a
871
25
1418
26
>10,000
nd
nd
nd
nd
Ph
N
N
HO
Compounds 31–33 were investigated further with regard
to 11 b-HSD1 potency in human adipocytes and for
microsomal intrinsic clearance (Table 3). The results
show that 11b-HSD1 activity in cells is high for 32
and 33, and somewhat lower for 31. The table also
shows that the 5,5-dialkyl-substituted oxazolones 31
and 32 are metabolically more stable in both rat and hu-
man microsomes when compared to the spiro-N-benz-
ylpiperidine analogue 33. Furthermore, in comparing
31 and 32 to the corresponding 2-cyclo-
alkylaminooxazolones (1, 6, 10, 12 and 13 in Table 1),
rat microsomal metabolic stability is clearly improved.
This implies that benzylic amines on the oxazolone 2-po-
sition might provide a versatile alternative to the cyclo-
alkylamines in identifying compounds with a beneficial
preclinical profile.
27
>10,000
Ph
28
>10,000
nd
nd
nd
nd
nd
nd
Naph
N
29
>10,000
p-MeOPh
p-BrPh
N
30
2875
N
31
19
32a
46
33
20
o-CF3Ph
N
a Racemate.
In conclusion, a novel class of potent 11b-HSD1 inhib-
itors has been identified. Oxazolone compounds incor-
porating cycloalkylamines or aryl-containing amines in
1-(3-hydroxyadamantyl) derivatives 16 and 17 for which
both human and rat clearance are low. Moreover, the
clearance values of 16 and 17 are significantly lower than
those of the corresponding 1-adamantyl analogues 12
and 13. This result is not unexpected since it is known
that increased polarity of the substrate often decreases
the affinity to metabolizing CYP enzymes.19 Unfortu-
nately, the 11b-HSD1 activities for 16 and 17 in adipo-
cytes are modest.
Table 3. 11b-HSD1 activity and intrinsic clearance data of oxazolones
incorporating the o-(trifluoromethyl)phenyl group
Compound
SPA
Ki (nM)
Adipocytes
IC50 (nM)
CLint (lL/min/
mg)
Human
Rat
31
19
46
20
95
12
29
6
23
6
15
To further investigate the SAR at the 2-position of the
oxazolone, a series of compounds featuring aryl-con-
taining substituents on the exocyclic nitrogen were syn-
thesized. As shown in Table 2, the 2-benzylamine
32a
33
193
293
a Racemate.

4840
L. Sutin et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4837–4840
the addition of 25 lL 11b-HSD1 as raw microsome
fraction from recombinant Pichia pastoris (the final
concentration of protein used was varied between 0.057
and 0.11 mg/mL, depending on the batch). Following
mixing, the plates were incubated on a shaker for 30–
60 min at rt. The reactions were terminated with 10 lL
stop solution (1 mM GA in ethanol). Monoclonal mouse
antibody was then added (10 lL of 1.92 lM working
solution) followed by 50 lL of YSi SPA beads. As
reference substance, carbenoxolone was run in each plate.
The plates were sealed with plastic film (Perkin-Elmer,
Top Seal-A) and incubated on a shaker for 30 min at rt
before counting. The amount of [3H]cortisol captured on
the beads was determined in a microplate liquid scintilla-
tion counter. Kinetic constants were calculated employing
the Microsoft Excel integrated application XLfit (Version
5.3.0.19, ID Business Solutions Ltd) using the sigmoidal
dose–response model 205 which is based on the non-linear
curve fitting based on Levenberg–Marquardts algorithm.
17. Compounds were dissolved in 100% DMSO to a final
concentration of 10 mM and diluted in DMSO followed
by a dilution in adipocyte medium. Cells were subjected to
compounds serially diluted in eight steps (nine concentra-
tions) ranging from 10 lM to 0.15 nM. The resulting
compound solutions were added to cells in presence of
100 nM cortisone. All samples were made and analyzed in
triplicate. Human primary subcutaneous adipocytes from
ZenBio (#SP-F-1) were propagated and differentiated
according to the protocol. Induction with compounds was
made for 5 h and the cortisol level for each compound in
the harvested media was determined with a Cortisol
Immuno Assay Kit from Assay Designs (Correlate #ADI-
901-071). Percent inhibition was calculated from absor-
bance (A405) raw data of the samples relative to the
positive control. Dose–response curves were generated by
plotting percent inhibition against compound concentra-
tions and IC50 values were calculated as the inflection
point at 50% inhibition, using 4-Parameter Logistic Model
in ExcelFit.
the 2-position and alkyl, fluoroalkyl, spiroalkyl or het-
erospiroalkyl substituents in the 5-position have been
investigated for human 11b-HSD1 activity in SPA and
adipocyte assays. Potent compounds with low intrinsic
clearance in human and rat microsomes have been
demonstrated.
References and notes
1. Walker, B. R.; Seckl, J. R. Exp. Opin. Ther. Targets 2003,
7, 771.
2. Fotsch, C.; Askew, B. C.; Chen, J. J. Exp. Opin. Ther.
Patents 2005, 15, 289.
3. Barf, T.; Williams, M. Drugs Future 2006, 31, 231.
4. Krozowski, Z. Mol. Cell. Endocrinol. 1992, 84, C25.
5. Rosmond, R. Psychoneuroendocrinology 2005, 30, 1.
6. Asensio, C.; Muzzin, P.; Rohner-Jeanrenaud, F. Int. J.
Obes. Relat. Metab. Disord. 2004, 28(Suppl. 4), S45.
7. Kotelevtsev, Y.; Holmes, M. C.; Burchell, A.; Houston, P.
M.; Schmoll, D.; Jamieson, P.; Best, R.; Brown, R.;
Edwards, C. R. W.; Seckl, J. R.; Mullins, J. J. Proc. Natl.
Acad. Sci. U.S.A. 1997, 94, 14924.
8. Morton, N. M.; Paterson, J. M.; Masuzaki, H.; Holmes,
M. C.; Staels, B.; Fievet, C.; Walker, B. R.; Flier, J. S.;
Mullins, J. J.; Seckl, J. R. Diabetes 2004, 53, 931.
9. Kataoka, S.; Kudo, A.; Hirano, H.; Kawakami, H.;
Kawano, T.; Higashihara, E.; Tanaka, H.; Delarue, F.;
Sraer, J.-D.; Mune, T.; Krozowski, Z. S.; Yan, K. J. Clin.
Endocrinol. Metab. 2002, 87, 877.
10. Steward, P. M.; Wallace, A. M.; Valentino, R.; Burt, D.;
Shackleton, C. H. L.; Edwards, C. R. W. Lancet 1987, 2,
821.
˚
11. Henriksson, M.; Homan, E.; Johansson, L.; Vallgarda, J.;
Williams, M.; Bercot, E.; Fotsch, C. H.; Li, A.; Cai, G.;
Hungate, R. W.; Yuan, C. C.; Tegley, C.; St. Jean, D.;
Han, N.; Huang, Q.; Liu, Q.; Bartberger, M. D.; Moniz,
G. A.; Frizzle, M. J. WO 05116002, 2005.
12. Decourcy, M. S. Eur. Patent 1,371,632, 2003.
13. Traube, W.; Ascher, R. Chem. Ber. 1913, 46, 2077.
14. Those primary amines not commercially available were
synthesized from the corresponding carboxylic acids via
a modified Curtius rearrangement according to Nin-
omiya et al., followed by acid-mediated deprotection.
Ninomiya, K.; Thioiri, S.; Yamada, S. Tetrahedron
1974, 30, 2151.
15. Lee, C.-M.; Horrom, B. W.; Michaels, R. J.; Rasmussen,
R. M.; Plotnikoff, N. P. J. Med. Chem. 1972, 15, 1252.
16. Human 11b-HSD1 scintillation proximity assay (SPA)
was carried out in replicon plates each replica containing
10 lL of diluted compound, 40 lL assay buffer, 10 lL of
11 mM G-6-P, and 25 lL substrate mixture [3H]cortisone/
NADPH (175 nM/200 lM). Reactions were initiated by
18. Incubations were performed at 37 °C using 1 lM com-
pound concentration, 1 mg/mL microsomal protein and
1 mM NADPH in a total volume of 150 lL of 100 mM
KPO4 buffer, pH 7.4. Parent compound remaining (at six
different time points from 0 to 40 min) is measured with
LC-MS/MS after incubation with human and rat liver
microsomes (microsomal protein concentration in incuba-
tion 1 mg/mL). In vitro half-life of the compound was
calculated and from this CLint (lL/min/mg) was derived.
19. Testa, B. Biochemistry of Redox Reactions; Academic
Press, 1995, pp 31–32.
20. St. Jean, D. J., Jr.; Yuan, C.; Bercot, E. A.; Cupples, R.;
Chen, M.; Fretland, J.; Hale, C.; Hungate, R. W.;
Komorowski, R.; Veniant, M.; Wang, M.; Zhang, X.;
Fotsch, C. J. Med. Chem. 2007, 50, 429.