Structure-based design and synthesis of 1,3-oxazinan-2-one inhibitors of 11β-hydroxysteroid dehydrogenase type 1

Article abstract of DOI:10.1021/jm2005354

Structure based design led directly to 1,3-oxazinan-2-one 9a with an IC₅₀ of 42 nM against 11β-HSD1 in vitro. Optimization of 9a for improved in vitro enzymatic and cellular potency afforded 25f with IC ₅₀ values of 0.8 nM for the enzyme and 2.5 nM in adipocytes. In addition, 25f has 94% oral bioavailability in rat and >1000× selectivity over 11β-HSD2. In mice, 25f was distributed to the target tissues, liver, and adipose, and in cynomolgus monkeys a 10 mg/kg oral dose reduced cortisol production by 85% following a cortisone challenge.

Full text of DOI:10.1021/jm2005354

ARTICLE
pubs.acs.org/jmc
Structure-Based Design and Synthesis of 1,3-Oxazinan-2-one
Inhibitors of 11β-Hydroxysteroid Dehydrogenase Type 1
Zhenrong Xu,† Colin M. Tice,*,† Wei Zhao,† Salvacion Cacatian,† Yuan-Jie Ye,† Suresh B. Singh,†
Peter Lindblom,† Brian M. McKeever,† Paula M. Krosky,† Barbara A. Kruk,† Jennifer Berbaum,†
Richard K. Harrison,† Judith A. Johnson,† Yuri Bukhtiyarov,† Reshma Panemangalore,† Boyd B. Scott,†
Yi Zhao,† Joseph G. Bruno,† Jennifer Togias,† Joan Guo,† Rong Guo,† Patrick J. Carroll,‡
Gerard M. McGeehan,† Linghang Zhuang,† Wei He,† and David A. Claremon†
†Vitae Pharmaceuticals, 502 West Office Center Drive, Fort Washington, Pennsylvania 19034, United States
‡Department of Chemistry, University of Pennsylvania, 250 South 33rd Street, Philadelphia, Pennsylvania 19104, United States
S Supporting Information
b
ABSTRACT: Structure based design led directly to 1,3-oxazi-
nan-2-one 9a with an IC50 of 42 nM against 11β-HSD1 in vitro.
Optimization of 9a for improved in vitro enzymatic and cellular
potency afforded 25f with IC50 values of 0.8 nM for the enzyme
and 2.5 nM in adipocytes. In addition, 25f has 94% oral bioavail-
ability in rat and >1000ꢁ selectivity over 11β-HSD2. In mice,
25f was distributed to the target tissues, liver, and adipose, and
in cynomolgus monkeys a 10 mg/kg oral dose reduced cortisol
production by 85% following a cortisone challenge.
’ INTRODUCTION
and activates the glucocorticoid receptor, resulting in increased
expression of a wide range of genes involved in metabolism, immune
response, bone formation, memory, and reproduction. More
specifically, cortisol drives gluconeogenesis in the liver and
adipogenesis in adipose tissue, which when elevated may con-
tribute to metabolic syndrome. Circulating levels of cortisol are
tightly controlled through the HPA axis, as well as by binding to
corticosteroid-binding globulin and serum albumin and by
inactivation to cortisone by 11β-hydroxysteroid dehydrogenase
type 2 (11β-HSD2), an NAD-dependent oxidase expressed
primarily in the kidney.8
However, tissue-specific and intracellular concentrations of
cortisol vary from plasma levels primarily because of the expres-
sion and activity of 11β-HSD1 and 11β-HSD2.9 NADPH-
dependent 11β-HSD1 is primarily expressed in liver and adipose
tissue, and its elevated expression in adipose tissue has been
linked to obesity, insulin resistance, diabetes, and cardiovascular
disease in humans.10ꢀ14 There is considerable overlap between
the symptoms of hypercortisolism (Cushing’s syndrome) and
those of the metabolic syndrome.15 Transgenic mice selectively
overexpressing 11β-HSD1 in adipose tissue exhibit visceral
obesity, insulin resistance, and hypertension,16,17 while 11β-
HSD1 knockout mice are resistant to diet induced obesity and
have increased insulin sensitivity.18ꢀ20 Thus, a selective inhibitor
of 11β-HSD1 may lower tissue-specific levels of cortisol and may
Diabetes is an abnormal state marked by an inability to make
sufficient insulin (type 1 diabetes) or by an inability to appro-
priately respond to insulin (type 2 diabetes). Both conditions
lead to elevated concentrations of glucose in the blood. The
estimated global diabetes prevalence in 2010 is 285 million and is
predicted to rise to 438 million by 2030 (IDF, Diabetes Atlas).1
When left untreated, diabetes results in serious complications
including cardiovascular disease, renal failure, and retinal da-
mage. Moreover, diabetes is often associated with comorbidities
of obesity, dyslipidemia, and hypertension. This assembly of
diseases is known as metabolic syndrome.2 Current therapies
include life style intervention, oral antidiabetes drugs, and
injections of insulin or incretin mimetics. The need for novel,
safe, and easy to administer alternatives persists because of
inadequate improvement in glycemic control and/or intolerable
side effects.3 In recent years, advances in understanding the
fundamental biology underlying the diseases comprising meta-
bolic syndrome have raised possibilities for pharmacological
interventions that directly impact their etiology and progression.
One enzyme proposed to be critical for the development of
metabolic syndrome is 11β-hydroxysteroid dehydrogenase type
1 (11β-HSD1), a member of the short chain dehydrogenase/
reductase (SDR) superfamily4,5 responsible for the local activa-
tion of inactive cortisone to the active glucocorticoid cortisol.
Plasma cortisol is primarily synthesized de novo in the zona
fasciculata of the adrenal cortex in response to stimuli from the
hypothalamicꢀpituitaryꢀadrenal axis (HPA).6,7 Cortisol binds to
Received: May 3, 2011
Published: July 25, 2011
r
2011 American Chemical Society
6050
dx.doi.org/10.1021/jm2005354 J. Med. Chem. 2011, 54, 6050–6062
|

Journal of Medicinal Chemistry
ARTICLE
Figure 1. Literature 11β-HSD1 inhibitors.
be useful for the treatment of diabetes and other morbidities
associated with metabolic syndrome.
Carbenoxolone (1, Figure 1), a semisynthetic derivative of the
natural product 18β-glycyrrhetinic acid, is a nonselective inhi-
bitor of 11β-HSD1 and of 11β-HSD2. Despite its lack of
selectivity for 11β-HSD1, 1 has been investigated in animal
and human studies where it improved insulin sensitivity.21,22
Since 2002,23 several classes of more selective, synthetic inhibi-
tors of 11β-HSD1 have been reported.24ꢀ26 Triazole 2 increased
insulin sensitivity and decreased fasting glucose, cholesterol, and
adipose tissue mass in mice.27 Thiazolone 3 dosed at 25 mg/kg b.
i.d. for 13 days to mice with DIO reduced fed blood glucose and
plasma insulin levels.28 Oral dosing of thiazolone 4 has been
reported to increase plasma adiponectin levels and decrease
fasting glucose levels in KKAγ mice.29,30 Sulfonamide 5 has been
shown to reduce fed glucose and fasted insulin in mice when
incorporated into a high fat diet at 30 (mg/kg)/day.31 The activity
of these compounds in animal models further validates the
potential for selective inhibitors of 11β-HSD1 in the treatment
of diabetes.
Several 11β-HSD1 inhibitors are reported to have entered
clinical trials, including adamantylamide 6.32 When added to
ongoing metformin therapy in a phase 2 trial, a 200 mg dose of
INCB13739 (structure not disclosed) reduced HBA1C by 0.6%
and fasting plasma glucose by 24 mg/dL.33 A 6 mg/day dose of
MK-0916 (structure not disclosed) in a phase 2 trial in patients
with type 2 diabetes did not improve fasting plasma glucose;
however, modest improvements in HBA1C, body weight, and
blood pressure were observed.34
Figure 2. Design of oxazinanone 11β-HSD1 inhibitors.
cycle. In the mouse 11β-HSD1 structure 1Y5R,35 the side chain
hydroxyl groups of Ser170 and Tyr183 both donate hydrogen
bonds to the hydroxyl oxygen at C11 of corticosterone, the
murine analogue of cortisol. 1n 2BEL, a human 11β-HSD1
structure, Ser170 and Tyr183 donate hydrogen bonds to the
ketone and ring E carboxylate of 1, respectively. 11β-HSD1 has
large, primarily hydrophobic pockets on both sides of the key
hydrogen-bonding residues, Ser170 and Tyr183. One of these
(pocket I) is formed by the cofactor, Thr124, Leu126, and
Val180. The second hydrophobic pocket (pocket II) is formed
by Leu126, Val180, and Tyr177. In 2BEL, Pocket II has a large
opening to solvent due to the truncated C-terminus of the
protein construct used for crystallization. Subsequently, a full
length dimeric human 11β-HSD1 X-ray structure, 2IRW,36 was
published. This structure revealed that the hydrophobic residues
of the C-terminal tail of the second monomer substantially close
off pocket II, leaving a much smaller opening to solvent than is
present in 2BEL. Our early design work employed 2BEL, but
once the more physiologically relevant 2IRW became available,
this was adopted.
’ DISCUSSION OF ENZYME STRUCTURE
Prior to initiation of our medicinal chemistry program, we
examined the crystal structures of 11β-HSD1 that were publicly
available at that time (PDB codes 1XSE, 1XU7, 1XU9, 1Y5M,
1Y5R, 2BEL). In 11β-HSD1, Ser170 and Tyr183, along with the
cofactor NADPH, are key elements involved in the catalytic
6051
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
’ TEMPLATE DESIGN
templates (Figure 2). Fragment 7 was docked into the binding
site of the 2BEL structure of 11β-HSD1 with the carbonyl
oxygen of the cyclic carbamate positioned to accept a hydrogen
bond from the hydroxyl of Ser170 and the ring oxygen positioned
to accept a hydrogen bond from the hydroxyl of Tyr183. This
binding mode was suggested by the hydrogen bonding network
of the 1,2,4-triazole ligand in fungal SDR structure 1YBV.37 By
use of a proprietary structure based design tool,38 more than
5000 molecules were grown in silico by adding fragments directly
to the N3 and C6 positions of 7. Diphenyloxazinanone 8 was
selected from this set of de novo grown molecules based on its
excellent and complementary fit to the binding pocket of the
enzyme (Figure 3). In the model of 8, both the carbonyl oxygen
and the ring oxygen remained in good hydrogen bonding
distance to the hydroxyl groups of Ser170 and Tyr183. The
3-phenyl substituent enjoyed excellent hydrophobic interactions
with the enzyme. However, to achieve a good fit to the enzyme
binding site, the 6-phenyl substituent had to adopt an energeti-
cally unfavorable pseudoaxial orientation on the oxazinanone
ring. To reduce the energetic penalty for the pseudoaxial
orientation, we appended a methyl group at the 6-position of
the oxazinanone. This methyl group was also expected to block
oxidative metabolism at the benzylic position. In addition, to fill
pocket II better, we appended a methyl group at the meta
position of the 3-phenyl ring. These changes gave compound
(R)-9a which retained the hydrogen bond network to the
carbamate oxygens modeled for 8. The added methyl groups
better filled pockets I and II, and an improved fit of the 6-phenyl
group within pocket I was observed. Compound 9a and analogues
of general structure 9 were selected for synthesis.
The 1,3-oxazinan-2-one ring system 7 was one of a small
number of cyclic, carbonyl-containing rings selected as starting
Figure 3. Model of 8 bound to the 2BEL structure of human 11β-
HSD1. The second monomer chain is not shown. Ligand 8 interacts
with the catalytic site through hydrogen bonds between the carbonyl
oxygen and the side chain hydroxyl of Ser170 and between the ring
oxygen and the side chain hydroxyl of Tyr183.
Table 1. Aniline SAR
compd
R3
X
enzyme IC50 a (nM)
9a
Me
Me
Me
H
3-Me
3-Me
3-Me
3-Me
3-Me
H
42
23
9a1b
9a2c
9b1d
9b2e
9c
152
999
353
160
40
’ CHEMISTRY
Compounds of general structure 9 (Table 1) were prepared as
shown in Scheme 1. Addition of allyl Grignard to acetophenone
(10) followed by ozonolysis and selective tosylation of the
primary alcohol gave hydroxytoslyate 11a. Enantiopure hydro-
xytosylates 11b were prepared from the corresponding commer-
cially available (R) and (S) enantiomers of 12. The oxazinanone
ring was formed in one pot by treatment of hydroxytosylates 11
with aryl isocyanates in the presence of DBU. The enantiomers of
9a were separated by HPLC on a chiral column.
H
Me
Me
Me
Me
Me
Me
Me
9d
2-Cl
3-Cl
3-Br
4-Cl
4-CF3
9e
103
23
9f
9g
229
805
631
9h
16f
a Average of at least two replicates. b Shorter tR isomer. Stereochemical
configuration not assigned. c Longer tR isomer. Stereochemical configuration
not assigned. d (R) isomer. e (S) isomer. f Structure is shown in Scheme 2.
N-(4-Bromobenzyl) analogue 16 was prepared as shown in
Scheme 2. Hydroxynitrile 13, prepared by addition of acetonitrile
anion to acetophenone (10), was reduced with borane to amino
Scheme 1. Synthesis of Analogues 9a
a See Table 1 for definitions of R3 and X. Reagents and conditions: (a) H2Cd=CHCH2MgBr, THF, ꢀ70 °C to room temp; (b)O3, CH2Cl2, ꢀ70 °C,
then NaBH4, ꢀ20 °C; (c)TsCl, Et3N, CH2Cl2, 0 °C to room temp; (d) X-C6H4NCO, DBU, THF, CH2Cl2, 0 °C to room temp.
6052
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of 16a
a Reagents and conditions: (a)MeCN, n-BuLi, THF, ꢀ20 °C to room temp; (b) BH3 Me2S, THF, reflux, 16 h; (c) triphosgene, Et3N, CH2Cl2, 0 °C, 1 h;
3
(d) 4-bromobenzyl bromide, NaH, THF, room temp, 16 h.
Scheme 3. Synthesis of Analogues 18a
a See Table 3 for definitions of R3 and Y. Reagents and conditions: (a) Y-PhB(OH)2, Pd(PPh3)2Cl2, NaHCO3, H2O, THF, reflux; (b) O3, CH2Cl2, ꢀ78 °C,
then NaBH4, ꢀ20 °C to room temp; (c) BH3 THF, 0 °C to room temp, then NaBO3 H2O, H2O, 0 °C to room temp.
3
3
Scheme 4. Synthesis of Intermediate 17a
a Reagents and conditions: (a) Me2NH, HCHO, HCl, EtOH, 70 °C, 16 h; (b) 3-Br-C6H3NH2, 1:1 H2O/EtOH, 80 °C, 16 h; (c) MeOCOCl, K2CO3,
THF, 0 °C to room temp, 2 h; (d) H2CdCHCH2MgBr, THF, ꢀ70 °C to room temp, 2 h.
alcohol 14. Treatment of 14 with triphosgene afforded oxazinanone
15 which was N-alkylated with 4-bromobenzyl bromide to give 16.
Compounds of general structure 18 (Table 3) were prepared as
shown in Scheme 3. Starting bromide 9f was prepared as shown in
Scheme 1, while 17 was accessed as shown in Scheme 4. Suzuki
coupling of fluorophenylboronic acids with 9f and 17 afforded
18aꢀf. TheallylR3 substituent of compound 18f was converted to a
2-hydroxyethyl group in 18g by ozonolysis and borohydride reduc-
tion. Hydroboration effected the conversion of the allyl R3
substituent in 18f to the 3-hydroxypropyl substituent in 18h. The
enantiomers of two compounds, 18d and 18h, were separated by
preparative SFC on a chiral column.
Bromides 22 (Scheme 5) were used as key intermediates to
access analogues of general structure 25 (Table 4). Reaction of
3-chloropropiophenones 20 with allyl Grignard in the presence
of CeCl3 gave chloro alcohols 21, which were treated with
(S)-4-bromo-R-methylbenzyl isocyanate to give bromooxazina-
nones 22 as mixtures of diastereomers. Suzuki coupling of
6053
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Scheme 5. Synthesis of Biphenylethyl Analogues 25a
a Definitions of R3, Y, and Z in 25 are given in Table 4. Reagents and conditions: (a) H2CdCHCH2MgBr, CeCl3, THF, ꢀ78 °C to room temp, 16 h;
(b) (S)-4-Br-C6H4CHMeNCO, DBU, THF, reflux, 1 day; (c) Y-C6H5B(OH)2, (PPh3)2PdCl2, Na2CO3, H2O, THF; (d) O3, CH2Cl2, ꢀ78 °C, then
NaBH4, ꢀ20 °C to room temp; (e) NaIO4, OsO4, 1:1 H2O/THF, room temp, 2 h, then NaBH4, MeOH, room temp, 1 h; (f) BH3 THF, 0 °C to room
3
temp, then NaBO3 H2O, H2O, 0 °C to room temp.
3
Figure 4. Diastereomers of 25f.
bromooxazinanones 22 with 4-fluoro and 2,4-difluorophenyl-
boronic acids afforded biphenyl intermediates 23. The allyl side
chains were elaborated to the 2-hydroxyethyl and 3-hydroxypro-
pyl side chains, as described above for 18g and 18h, to give
diastereomeric products 24 and 25. Compound 25e (Table 4),
with an ethyl group at the benzylic position, was prepared by an
analogous sequence using (S)-4-BrC6H4CHEtNCO in the second
step. In one case (25k), the allyl group of bromo intermediate 23d
was elaborated to 2-hydroxyethyl prior to Suzuki coupling. The
desired, potent compounds 25 were readily separated from the
undesired diastereomers 24 by chromatography on silica gel.
Separation of the diastereomers of 22 and 23 on silica gel could
also be achieved routinely. In all cases, the potent diastereomers
25, and the diastereomers of precursors 22 and 23 that could be
converted to 25, were characterized by a consistent downfield shift
1
of the benzylic methyl doublet (∼1.5 ppm) in the H NMR
6054
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Scheme 6. Elaboration of 2-Hydroxyethyl to Other Two-Carbon Side Chainsa
a Reagents and conditions: (a) Jones reagent, Me2CO, 0 °C to room temp, 16 h; (b) NH3, EDC, HOBt, i-Pr2NEt, CH2Cl2, room temp, 16 h; (c) MsCl,
Et3N, CH2Cl2, 0 °C to room temp, (d) NaN3, DMF, 70 °C, 16 h; (e) PPh3, 20:1 THF/H2O, room temp, 16 h; (f) Ac2O, NaOAc, room temp, 2 h;
(g) H2CONH2, HCl, H2O, reflux, 16 h; (h) MeNH2, CDI, i-Pr2NEt, CH2Cl2, room temp, 16 h; (i) EtNH2, CDI, i-Pr2NEt, CH2Cl2, room temp, 16 h;
(j) MeSO2Cl, Et3N, CH2Cl2, room temp, 2 h; (k) H2NSO2NH2, dioxane, reflux, 16 h.
Scheme 7. Elaboration of 3-Hydroxypropyl to Other Three-Carbon Side Chainsa
0
a Reagents and conditions: (a) Jones reagent, Me2CO, 0 °C to room temp, 16 h; (b) R3 NH2, EDC, HOBt, i-Pr2NEt, CH2Cl2, room temp, 16 h;
(c) MsCl, Et3N, CH2Cl2, 0 °C to room temp; (d) NaN3, DMF, 70 °C, 16 h; (e) PPh3, 20:1 THF/H2O, room temp, 16 h; (f) H2CONH2, HCl, H2O,
reflux, 16 h; (g) MeNH2, CDI, i-Pr2NEt, CH2Cl2, room temp, 16 h; (h) MeSO2Cl, Et3N, CH2Cl2, room temp, 2 h.
spectrum compared tothe sameresonance in24(∼1.2 ppm). The
absolute configuration of 25f (Table 4) was established by X-ray
crystallography, and the configurations of other analogues were
methyl doublet. Compounds 27 and 28 (Figure 4), the enantio-
mers of 24f and 25f, were prepared by the sequence shown in
Scheme 5 using (R)-4-bromo-R-methylbenzyl isocyanate in the
second step.
1
assigned based on the H NMR chemical shift of the benzylic
6055
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
The syntheses of analogues 26 (Table 5) in which the
2-hydroxyethyl or 3-hydroxypropyl side chains in 25f and
25g were further elaborated are depicted in Schemes 6 and 7,
respectively. Oxidation of alcohols 25f and 25g to carboxylic
acids followed by amide formation afforded 26a, 26i, and 26j.
Alcohols 25f and 25g were converted to the corresponding
primary amines 26b and 26k which were used to access a number
of acylated derivatives (26cꢀh, Scheme 6; 26lꢀn, Scheme 7).
were incubated at 37 °C for 2 h, and [3H]cortisol production
was quantitated by HPLC. IC50 values represent the mean of at
least duplicate assays and were generated from an eight-point
concentrationꢀresponse curve.
’ RESULTS AND DISCUSSION
Designed compound 9a had enzyme and adipocyte IC50 of 42
and 166 nM as a racemate. The enantiomers of 9a were separated
by chiral chromatography, and isomer 9a1 was substantially
more potent than 9a2. The absolute configuration of 9a1 was
not determined. The importance of the methyl group at the
6-position of the oxazinanone ring, introduced to favor the
desired pseudoaxial conformation of the 6-phenyl ring in pocket
I, was demonstrated by preparation of 9b1 and 9b2, both of
which were less potent than 9a1. Substitution on the N-phenyl
ring was briefly examined. Deletion of the methyl group at the
meta-position decreased potency (9c versus 9a). Introduction of
chlorine at the ortho-position (9d) or bromine at the meta-
position (9f) restored potency levels comparable to that of 9a,
while chlorine substitution at the para-position (9g) was un-
favorable. A rat PK study showed that 9a1 had low oral
bioavailability and was subject to rapid clearance (Table 2).
Inspection of a model of 9a bound to 11β-HSD138 suggested
that the m-methyl group of 9a could be replaced with a phenyl
ring to give 18a which was expected to better fill hydrophobic
pocket II. With an enzyme IC50 of 4.7 nM, compound 18a
(Table 3) was 9ꢁ more potent than 9a. The adipocyte
potency of 18a exhibited a 2.5ꢁ improvement over that of
9a; however, its t1/2 in rat liver microsomes (RLM t1/2) was
only 16 min. Introduction of fluorine substituents on the
aromatic rings was expected to improve metabolic stability.
In fact, fluorine (Table 3) and other small lipophilic substit-
uents (data not shown) were generally tolerated on the distal
ring of the biphenyl system, and fluorine in the para position
(18d) increased the RLM t1/2 to >60 min. The 2,4-difluoro
substitution in 18e proved to be especially favorable with
’ BIOLOGICAL ASSAYS
Compounds were assayed for inhibition of 11β-HSD1 in
enzyme- and cell-based assays. Both assays measured the con-
version of [3H]cortisone to [3H]cortisol, which was quantified
using SPA beads and a microscintillation plate reader.39 Enzyme
assays used recombinant 11β-HSD1 isolated as a microsomal
preparation from CHO cells. Assays were performed in 25 mM
HEPES, pH 7.4, 50 mM KCl, 2.5 mM NaCl, 1 mM MgCl2, 1 mM
NADPH, and 80 nM [3H]cortisone at room temperature for 1 h.
Cell-based potency was assessed in differentiated human adipo-
cytes by the addition of cortisone (80 nM [3H]cortisone). Cells
Table 2. Rat PK Dataa
po AUC(inf)
po t1/2
iv CL
compd
(ng h/mL)
(h)
(mL/(min kg))
F (%)
3
3
9a1b
18g1c
25bd
25dd
25fe
99
1045
915
0.8
4.7
3.6
1.6
3.0
3.3
71
71
48
86
41
44
4
44
19
23
94
83
450
3870
3168
25kd
a iv dose of 2 mg/kg, po dose of 10 mg/kg. b po vehicle: 10% ethanol/
c
10% Tween 80/80% PEG 400 (solution). po vehicle: 1% DMSO qs
with 10% ethanol/10% Tween 80/80% PEG 400 (solution). d po vehicle:
0.5% methyl cellulose in deionized water (homogeneous suspension).
e 10% ethanol/10% polysorbate 80/80% PEG 400 (solution).
Table 3. Biphenyl SAR
compd
R3
Y
enzyme IC50 a (nM)
adipocyte IC50 a (nM)
RLM t1/2 (min)
18a
Me
Me
Me
Me
Me
Me
Me
H
4.7
12
64
40
50
16
18b
2-F
18c
3-F
4.2
3.3
3.3
29
18d
4-F
62
75
47
18d1b
18d2c
18e
4-F
36
4-F
2,4-di-F
2,4-di-F
2,4-di-F
2,4-di-F
2,4-di-F
2,4-di-F
1.6
3.2
1.2
0.8
140
4.4
24
18f
CH2dCHCH2
HOCH2CH2
HOCH2CH2
HOCH2CH2
HOCH2CH2CH2
50
18g
4.4
1.9
570
6.9
18g1d
18g2e
18h
a Average of at least two replicates. b Longer tR isomer on a ChiralCel-AS column. c Shorter tR isomer on a ChiralCel-AS column. d Longer tR isomer on a
Chiralpak IA column. e Shorter tR isomer on a Chiralpak IA column.
6056
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Table 4. SAR of Biphenylmethyl Analogues 25
a
enzyme
IC50 a (nM)
adipocyte
IC50 a (nM)
enzyme/plasma
IC50 (nM)
CYP3A4 IC50
RLM t1/2
compd
R1
R3
Y
Z
(μM)
(min)
25a
25b
25c
25d
25e
25f
Me
Me
Me
Me
Et
CH2CH2OH
2,4-di-F
2,4-di-F
4-F
H
0.7
0.6
0.5
0.6
0.6
0.8
1.1
0.9
0.9
0.6
0.5
0.9
0.7
0.8
0.4
0.9
2.5
2.1
1.2
1.3
0.7
1.5
8.4
17.9
5.7
16
18
CH2CH2CH2OH
CH2CH2OH
H
14
H
28
CH2CH2CH2OH
CH2CH2CH2OH
CH2CH2OH
4-F
H
11.9
18.7
17.5
18.0
8.6
11
1.5
12
>30
6
28
4-F
H
>60
119
21
Me
Me
Me
Me
Me
Me
2,4-di-F
2,4-di-F
4-F
4-F
4-F
2-F
3-F
4-F
4-F
25g
25h
25i
CH2CH2CH2OH
CH2CH2CH2OH
CH2CH2CH2OH
CH2CH2CH2OH
CH2CH2OH
32
4-F
11.8
14.7
18.5
7
63
25j
4-F
8
29
25k
4-F
21
58
a Average of at least two replicates.
Figure 5. Model of 25f bound to the 2IRW structure of human 11β-HSD1. The side chain hydroxyl of Ser170 forms a hydrogen bond to the carbonyl
oxygen of the ligand, while the phenolic hydroxyl of Tyr183 lies within hydrogen bonding distance of both the carbonyl and ring oxygen atoms of the
ligand. The primary alcohol of the ligand forms a hydrogen bond to a phosphate oxygen of NADP. The 4-fluorophenyl ring stacks with Tyr183, and its
para carbon fits into a hydrophobic subpocket formed by Thr124, Ser125, Leu126, and Val180. The 2,4-difluorophenyl ring interacts with the Tyr284
side chain of the partner chain in this model.
enzyme and adipocyte potencies 3ꢁ better than 18a. Despite
the attractive potency of 18e, its hydrophobicity (alogP > 5;
PSA = 30 Å2) was outside the desirable range for druglike
molecules.
Two approaches were explored to reduce hydrophobicity by
introduction of polar functionality groups. First, a library of 34
analogues of general structure 18 with polar Y substituents was
prepared by coupling commercially available boronic acids to 9f.
As predicted by our model, polar Y substitutuents were not
tolerated (data shown in Supporting Information).
The second approach resulted from examination of a model of
(R)-18e bound to the protein. This model suggested the
possibility of forming hydrogen bonds to the pyrophosphate
portion of the NADP cofactor using the 6-methyl substituent on
the oxazinanone ring as an attachment point. Models suggested
that replacement of the methyl group with a 2-hydroxyethyl ((S)-
18g) or 3-hydroxypropyl ((R)-18h) moiety would position a
hydroxyl group within hydrogen bonding distance of the NADP
cofactor pyrophosphate. These compounds were prepared as
their racemates. The 2-hydroxyethyl compound 18g retained the
enzyme potency of 18e, while 18h was 2.5ꢁ less potent. In the
adipocyte assay, both 18g and 18h were markedly more potent
than 18e. The isomers of 18g were separated by chromatography
on a chiral column, and isomer 18g1 was shown to be responsible
for the activity of 18g. In a rat PK study, 18g1 demonstrated
substantially greater exposure than 9a1, although it remained
subject to high clearance (Table 2). The calculated physical
properties of 18g1 (alogP = 4.6; PSA = 50 Å2) were more
favorable than those of 18e.
Despite its improved physical properties and PK profile, the
aniline functionality of 18g1 was considered undesirable in a
chronically administered drug.40,41 Replacement of the N-phenyl
group with an N-benzyl group was investigated. Examination of a
model of 16 suggested that an (S)-methyl group at the benzylic
carbon would gain hydrophobic interactions but that transposi-
tion of the distal phenyl ring from the meta to para position
would be necessary to properly occupy pocket II.
Thus, 25a and 25b, (S)-R-methylbenzyl analogues of 18g and
18h, were prepared and exhibited excellent enzyme and cell
potencies (Table 4). The 3-hydroxypropyl compound 25b
suffered a greater loss in potency in the presence of human plasma
than 25a. Analogues25c and 25d, in which the Y substitutuentwas
6057
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Table 5. Side Chain SAR
compd
R3
enzyme IC50 a (nM) adipocyte IC50 a (nM) enzyme/plasma IC50 a (nM) CYP3A4 IC50 a (μM) RLM t1/2 (min)
25f
CH2CH2OH
CH2CONH2
0.8
2.8
2.7
7.1
0.8
3.1
22
2.5
12
18
62
12
>30
3
119
26a
26b
26c
26d
26e
26f
CH2CH2NH2
66
CH2CH2NHAc
34
164
29
CH2CH2NHCONH2
CH2CH2NHCONHMe
CH2CH2NHCONHEt
CH2CH2NHSO2Me
CH2CH2NHSO2NH2
CH2CH2CH2OH
3.5
16
5
1
69
26g
26h
25g
26i
1.0
2.0
1.1
1.0
22
9.0
11
45
37
18
42
9
12
2.0
5.0
145
16
>30
>30
21
25
CH2CH2CONH2
26j
CH2CH2CONHMe
CH2CH2CH2NH2
26k
26l
5.7
1.4
4.1
1.8
2
CH2CH2CH2NHCONH2
CH2CH2CH2NHCONHMe
CH2CH2CH2NHSO2Me
19
53
32
4
26m
19
0.6
7
26n
3.7
>60
a Average of at least two replicates.
changed from 2,4-difluoro to 4-fluoro, showed similar trends.
Changing the R1 substituent from methyl to ethyl increased both
plasma shift and CYP3A4 inhibition (25e versus 25d). Introduc-
tion of a 4-fluoro substituent on the left-hand phenyl ring gave 25f
and 25g, both of which had excellent enzyme and cell potency,
combined with limited CYP3A4 inhibition. Hydroxyethyl com-
pound 25f was distinguished by its superior metabolic stability in
rat liver microsomes. Analogues of 25d with fluoro substitution at
the 2-position (25h), 3-position (25i), and 4-position (25j) on the
left-hand phenyl ring all had excellent potency but had IC50 < 10
μM against CYP3A4. Introduction of a 4-fluoro substituent onto
the left-hand phenyl ring of 25c raised the CYP3A4 IC50 above
10 μM and gave 25k with good RLM stability. Rat PK parameters
were determined for 25b, 25d, 25f, and 25k. The monofluoro,
3-hydroxypropyl analogue 25d demonstrated the highest clear-
ance (Table 2). Both 2-hydroxyethyl compounds, 25f and 25k,
possessed excellent oral bioavailability.
Protein-bound models of 25f (Figure 5) and 25g indicated
that other functional groups capable of donating a hydrogen
bond to the cofactor pyrophosphate would be accommodated
within the protein binding site. Table 5 shows that certain
analogues of this type retained enzyme IC50 < 5 nM; however,
in no case did the adipocyte potencies of these compounds
exceed those of 25f and 25g. Furthermore, with the exception of
primary amides 26a and 26i, these compounds were more potent
inhibitors of CYP3A4 than 25f and 25g.
Table 6. In Vitro Characterization of 25f
enzyme/receptor
IC50 (μM)
11β-HSD2a
17β-HSD1b
3β-HSD2c
CYP3A4d
CYP2D6d
CYP2C9d
hERGe
>10
>10
4.7
>10
>10
4.9
>10
receptor
EC50 (μM)
FXRf
GRg
MRh
>10
>10
>5
a Compounds were diluted in DMSO and dispensed to 96-well microtiter
plates. Substrate (35 mM 3H-cortisol and 1 mM NAD+ in 50 mM potassium
phosphate buffer, pH 7.4) and microsomes isolated from CHO cells
overexpressing 11B-HSD2 were added, and the plate was incubated at room
temperature for 60 min. The reaction was quenched with ACN, and
cortisone product was detected by HPLC. b Puranen, T.; Poutanen, M.;
Ghosh, D.; Vihko, P.; Vihko, R. Mol. Endocrinol. 1997, 11, 77. c Frye, S. V.;
Haffner, C. D.; Maloney, P. R.; Mook, R. A.; Dorsey, G. F.; Hiner, R. N.;
Cribbs, C. M.;Wheeler, T. N.;Ray, J. A.J. Med. Chem. 1994, 37, 2352. d Assay
kit was purchased from Invitrogen. e Assay was performed by MDS Pharma
Services. f Parks, D. J.; Blanchard, S. G.; Bledsoe, R. K.; Chandra, G.; Consler,
T. G.; Kliewer, S. A.; Stimmel, J. B.; Willson, T. M.; Zavacki, A. M.; Moore,
D. D.; Lehmann, J. M. Science 1999, 284, 1365. g Assay kit was purchased
from Panvera/Invitrogen. h Grover, G. S.;Turner, B. A.;Parker, C. N.;Meier,
J.; Lala, D. S.; Lee, P. H. J. Biomol. Screening 2003, 8, 239.
Compounds 24f, 27, and 28, the three diastereomers of 25f,
were much less potent than 25f (Figure 4), supporting the
validity of our model.
Profile of 25f. Compound 25f was selected for further in
vitro and in vivo characterization. The compound displayed
>1000ꢁ selectivity for 11β-HSD1 over three related hydroxysteroid
dehydrogenases, three CYP isozymes, hERG, and three nuc-
lear receptors (MR, FXR, and GR) (Table 6). Among this group
6058
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Table 7. Pharmacokinetics of 25f in Three Species
microsomes and was bioavailable in rat and to a lesser extent in
dog and monkey (Table 7). When dosed orally at 30 mg/kg to
mice, 25f was distributed into adipose and liver, key target tissues
for the inhibition of 11β-HSD1 (Table 8).
dose iv/po
(mg/kg)
AUC(inf)
iv CL
species
(ng h/mL) t1/2 (h) (mL/(min kg)) F (%)
3
3
rat
2/10a
1/2b
1/2b
3870
584
3.0
41
94
16
26
In Vivo Biology. The poor potency of 25f against mouse
11β-HSD1 (IC50 = 670 nM) precluded the use of rodent pharma-
codynamic models. However, the compound was equally potent
against 11β-HSD1 from cynomolgus monkeys (IC50 = 0.97 nM),
and this species was selected to measure in vivo inhibition of
11β-HSD1. In this model, endogenous plasma glucocorticoids
were suppressed with dexamethasone, and the animals were
challenged with a bolus of cortisone 21-acetate 5 h after compound
administration. A single 10 mg/kg dose of 25f, administered by oral
gavage, reduced cortisol production ∼85% (Figure 6), consistent
with substantial inhibition of 11β-HSD1 in vivo.
dog
3.7
9.2
monkey
1500
10.9
5.8
a po vehicle:10% ethanol/10% polysorbate 80/80% PEG 400 (solution).
b po vehicle: 0.5% methylcellulose in deionized water.
Table 8. Mouse Tissue Distribution of 25f a
drug level (ng/mL)
time (h)
liver
adipose
plasma
’ CONCLUSION
1
4
8
15317
9813
1898
5519
3165
632
5927
3260
573
We have described the structure-based design of a novel class
of 11β-HSD1 inhibitors built around a 1,3-oxazinan-2-one ring.
Optimization (Figure 7) furnished 25f which combined excellent
potency against 11β-HSD1 in human adipocytes with >1000ꢁ
selectivity over three other hydroxysteroid dehydrogenases,
including 11β-HSD2. This compound was orally bioavailable
in three species and distributed to adipose tissue in mouse. Oral
administration to cynomolgus monkeys substantially reduced
plasma cortisol levels following a cortisone challenge.
a Animals (n = 3) were dosed po with 30 mg/kg dissolved in 10%
ethanol/10% Tween 80/80% PEG 400. Tissues were collected upon
sacrifice at 1, 4, and 8 h and snap frozen. Later, thawed tissues were
homogenized in 0.05 M BES, pH 7, and compound was extracted into
acetonitrile. Concentrations were determined by LCꢀMS/MS by
comparison to a standard curve generated independently in each tissue
matrix.
’ EXPERIMENTAL SECTION
1-Chloro-3-(4-fluorophenyl)hex-5-en-3-ol (21d). A 250 mL
flask was charged with anhydrous CeCl3 (5.58 g, 22.6 mmol) and THF
(40 mL). The mixture was vigorously stirred for 3.5 h at room
temperature. The suspension was cooled to ꢀ78 °C, and a solution of
allylmagnesium bromide (1.0 M in THF, 21 mL, 21.0 mmol) was added.
After the mixture was stirred for 2 h at ꢀ78 °C, a solution of 3-chloro-
1-(4-fluorophenyl)propan-1-one (2.52 g, 13.5 mmol) in THF (30 mL)
was added via cannula. The reaction mixture was allowed to slowly warm
to 8 °C while stirring overnight (18 h). The reaction was quenched with
saturated aqueous NaHCO3, and the aqueous layer was extracted with
EtOAc. The organic layer was dried over Na2SO4. After the solvents
were evaporated, the residue was purified by chromatography on silica
gel, eluting with hexanes/EtOAc to afford 21d (3.00 g, 97%) as an oil. 1H
NMR (400 MHz, CDCl3) δ 7.37ꢀ7.32 (m, 2H), 7.07ꢀ7.02 (m, 2H),
5.57ꢀ5.47 (m, 1H), 5.20ꢀ5.19 (m, 1H), 5.16 (m, 1H), 3.59ꢀ3.52 (m,
1H), 3.24ꢀ3.18 (m, 1H), 2.70 (dd, J = 13.8, 5.9 Hz, 1H), 2.50 (dd, J =
13.8, 8.5 Hz, 1H), 2.29 (t, J = 7.9 Hz, 2H), 2.22 (s, 1H); 19F NMR (376
MHz, CDCl3) δ ꢀ116.52 (m); LCꢀMS method 1, tR = 1.79 min, m/z
213, 211 (M ꢀ OH)+.
Figure 6. Pharmacodynamic effects of 25f in cynomolgus monkeys.
Animals were dosed with dexamethasone 8 h prior to compound
administration (t = ꢀ8 h) to suppress endogenous plasma cortisol
concentrations. The animals received vehicle or compound (10 mg/kg, po)
at time zero (t = 0 h) and 5 h later (t = 5 h) and were challenged
with cortisone 21-acetate (25 mg, po). Blood samples were drawn every
30 min for 6 h to measure cortisol production. The animals were crossed
over after a 7-day washout.
6-Allyl-3-((S)-1-(4-bromophenyl)ethyl)-6-(4-fluorophenyl)-
1,3-oxazinan-2-one (22d). A mixture of 21d (0.413 g, 1.8 mmol,
1.0 equiv), (S)-(ꢀ)-1-(4-bromophenyl)ethyl isocyanate (0.501 g,
2.2 mmol, 1.2 equiv), and DBU (0.738 g, 4.8 mmol, 2.7 equiv) in
THF (10 mL) was heated at reflux for 25 h. The mixture was diluted with
EtOAc and washed with 1 N aqueous HCl. The aqueous phase was
extracted with EtOAc (2ꢁ). The combined organic phase was dried over
Na2SO4. After the solvents were evaporated, the crude product was
directly used in the next step without further purification.
of antitargets, IC50 values for 3β-HSD2 and CYP2C9 of 4.7 and
4.9 μM were determined. Profiling against 165 enzymes and
receptors (MDS PanLabs Spectrum Screen) showed 25f to be
>5000ꢁ selective over all targets tested. The compound is >99%
protein bound in dog, monkey, and human plasma, consistent
with a pronounced shift in enzyme potency when assayed in
the presence of 50% human plasma. 25f was stable in rat liver
A portion of the crude product was purified by chromatography on
silica gel, eluting with hexanes/EtOAc to afford analytical samples.
(R)-6-Allyl-3-((S)-1-(4-bromophenyl)ethyl)-6-(4-fluoro-
phenyl)-1,3-oxazinan-2-one (RS)-22d. 1H NMR (400 MHz,
CDCl3) δ 7.25ꢀ7.20 (m, 4H), 7.05ꢀ7.01 (m, 2H), 6.71 (d, J = 8.5 Hz,
2H), 5.74ꢀ5.64 (m, 1H), 5.58 (q, J = 7.0 Hz, 1H), 5.09ꢀ4.99 (m, 2H),
2.92ꢀ2.87 (m, 1H), 2.63ꢀ2.50 (m, 2H), 2.33ꢀ2.16 (m, 3H), 1.47
6059
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
Figure 7. Evolution of 25f from (R)-9a.
(d, J = 7.0 Hz, 3H); 19F NMR (376 MHz, CDCl3) δ ꢀ114.91 (m).
LCꢀMS method 1, tR = 9.71 min, m/z = 418, purity = 86%. Anal. Calcd
for C21H21BrFNO2: C, 60.30; H, 5.06; N, 3.35. Found: C, 60.25; H,
4.87; N, 3.24.
to ꢀ78 °C, and ozone was bubbled in until the mixture turned blue.
NaBH4 (9.8 g, 258 mmol) was added, and the mixture was allowed to
warm to room temperature and stirred overnight. Saturated aqueous
NH4Cl was added, and the mixture was extracted with CH2Cl2 (3ꢁ).
The combined organic layer was washed with brine, dried over Na2SO4,
and concentrated under reduced pressure. The residue was purified by
flash chromatography on silica gel, and the more polar isomer was
collected. Recrystallization from ethanol afforded 25f (6.0 g, 25%). 1H
NMR (400 MHz, CDCl3) δ 7.24ꢀ7.17 (m, 5H), 7.00ꢀ6.79 (m, 6H),
5.62 (q, J = 7.0 Hz, 1H), 3.74ꢀ3.68 (m, 1H), 3.51ꢀ3.45 (m, 1H),
2.91ꢀ2.88 (m, 1H), 2.30ꢀ2.01 (m, 5H), 1.96 (br s, 1H), 1.48 (d, J=7.0Hz,
3H); 19F NMR (376 MHz, CDCl3) δ ꢀ111.61 (m), ꢀ114.05 (m),
ꢀ114.90 (m); 13C NMR (100 MHz, CDCl3) δ 162.28 (dd, J = 250.0,
12.3 Hz), 162.17 (d, J = 247.7 Hz), 159.66 (dd, J = 250.7, 12.3 Hz),
153.03, 138.53, 137.09 (d, J = 3.1 Hz), 134.06, 131.24 (dd, J = 9.2, 4.6
Hz), 128.70 (d, J = 2.3 Hz), 127.09, 126.58 (d, J = 8.4 Hz), 124.57 (dd,
J = 13.8, 3.8 Hz), 115.74 (d, J = 21.5 Hz), 111.59 (dd, J = 21.5, 3.8 Hz),
104.37 (t, J = 26.1 Hz), 82.83, 58.09, 53.18, 44.82, 36.11, 31.75, 15.16;
LCꢀMS method 1, tR = 8.61 min, m/z = 456, purity = 100%. Anal. Calcd
for C26H24F3NO3: C, 68.56; H, 5.31; F, 12.51; N, 3.08; O, 10.54. Found:
C, 68.53; H, 5.21; N, 3.06.
(S)-6-Allyl-3-((S)-1-(4-bromophenyl)ethyl)-6-(4-fluoro-
phenyl)-1,3-oxazinan-2-one (SS)-22d. 1H NMR (400 MHz,
CDCl3) δ 7.46 (d, J = 8.2 Hz, 2H), 7.31ꢀ7.28 (m, 2H), 7.17 (d, J =
8.2 Hz, 2H), 7.07 (t, J = 8.5 Hz, 2H), 5.76ꢀ5.66 (m, 2H), 5.10ꢀ4.99 (m,
2H), 2.75ꢀ2.52 (m, 4H), 2.23ꢀ2.19 (m, 1H), 2.08ꢀ2.00 (m, 1H), 1.24
(d, J = 7.0 Hz, 3H); 19F NMR (376 MHz, CDCl3) δ ꢀ115.07 (m);
LCꢀMS method 2, tR = 2.03 min, m/z 420, 418 (MH+).
6-Allyl-3-((S)-1-(20,40-difluoro-[1,10-biphenyl]-4-yl)ethyl)-
6-(4-fluorophenyl)-1,3-oxazinan-2-one (23f). A mixture of 22d
(10.0 g, 24.0 mmol), 2,4-difluorophenylboronic acid (4.74 g, 28.7
mmol), PdCl2(PPh3)2 (672 mg, 0.96 mmol), 2 M aqueous Cs2CO3
(30 mL, 600 mmol), and dioxane (120 mL) was heated at reflux under
N2 for 2 h. The mixture was cooled to room temperature and filtered.
The filtrate was extracted with EtOAc (3ꢁ). The combined organic
layer was washed with brine, dried over Na2SO4, and concentrated to
afford crude 23f which was used in the next step without further
purification. A portion of the crude product was purified by chroma-
tography on silica gel, eluting with hexanes/EtOAc to provide analytical
samples.
The less polar diastereomer 24f exhibited the following spectral
1
characteristics. H NMR (400 MHz, CDCl3) δ 7.48ꢀ7.30 (m, 7H),
(S)-6-Allyl-3-((S)-1-(20,40-difluorobiphenyl-4-yl)ethyl)-6-
(4-fluorophenyl)-1,3-oxazinan-2-one (SS)-23f. 1H NMR
(CDCl3) δ 7.47 (d, J = 8.2 Hz, 2H), 7.42ꢀ7.30 (m, 5H), 7.08 (t,
J = 8.2 Hz, 2H), 6.98ꢀ6.88 (m, 2H), 5.82ꢀ5.68 (m, 2H), 5.08 (d, J =
10.2 Hz, 1H), 5.02 (d, J = 17.0 Hz, 1H), 2.78ꢀ2.71 (m, 2H),
2.66ꢀ2.54 (m, 2H), 2.25ꢀ2.20 (m, 1H), 2.13ꢀ2.05 (m, 1H), 1.30
(d, J = 7.0 Hz, 3H). LCꢀMS method 1, tR = 10.68 min, m/z = 452.
purity = 96%. HRMS calcd for C27H24F3NO2 + Na 454.1896, found
454.1891.
7.13ꢀ6.88 (m, 4H), 5.78 (q, J = 7.0 Hz, 1H), 3.78ꢀ3.74 (m, 1H),
3.57ꢀ3.50 (m, 1H), 2.79ꢀ2.67 (m, 2H), 2.31ꢀ2.11 (m, 4H), 1.81 (br s,
1H), 1.31 (d, J = 7.3 Hz, 3H); 19F NMR (376 MHz, CDCl3) δ ꢀ111.48
(m), ꢀ113.88 (m), ꢀ114.67 (m); LCꢀMS method 2, tR = 1.86 min,
m/z 456 (MH+).
(R)-3-((S)-1-(20,40-Difluoro-[1,10-biphenyl]-4-yl)ethyl)-6-
(4-fluorophenyl)-6-(3-hydroxypropyl)-1,3-oxazinan-2-one
(25g). To a stirred solution of (RS)-23f (3.05 g, 6.7 mmol) in THF
(40 mL) at 0 °C was added 1 M BH3 in THF (8.1 mL, 8.1 mmol). The
mixture was stirred at room temperature for 2 h, recooled to 0 °C, and
treated with water (2 mL), 3 M aqueous NaOH (2 mL, 6 mmol), and
30% aqueous H2O2 (2 mL). The mixture was stirred overnight at room
temperature, and the pH was adjusted to 6ꢀ7 with aqueous HCl. The
mixture was extracted with EtOAc (3ꢁ). The combined organic layer
was washed with saturated aqueous NaHCO3 and brine and dried over
Na2SO4. Removal of the solvent left a residue which was purified by
chromatography on silica gel, followed by preparative HPLC to afford
25g (700 mg, 22%). 1H NMR (CDCl3) δ 7.34ꢀ7.22 (m, 4H),
7.07ꢀ6.98 (m, 4H), 6.96ꢀ6.85 (m, 3H), 5.69 (q, J = 7.02 Hz, 1H),
(R)-6-Allyl-3-((S)-1-(20,40-difluorobiphenyl-4-yl)ethyl)-6-(4-
1
fluorophenyl)-1,3-oxazinan-2-one (RS)-23f. H NMR (CDCl3)
δ 7.33ꢀ7.23 (m, 5H), 7.03 (t, J = 8.2 Hz, 2H), 6.96ꢀ6.86 (m, 4H),
5.77ꢀ5.67 (m, 2H), 5.10 (d, J = 10.3 Hz, 1H), 5.04 (d, J = 17.3 Hz, 1H),
2.99ꢀ2.94 (m, 1H), 2.66ꢀ2.54 (m, 2H), 2.41ꢀ2.34 (m, 1H), 2.30ꢀ2.17
(m, 2H), 1.55 (d, J = 7.0 Hz, 3H); LCꢀMS method 2, tR = 2.13 min,
m/z = 452.
(S)-3-((S)-1-(20,40-Difluoro-[1,10-biphenyl]-4-yl)ethyl)-6-(4-
fluorophenyl)-6-(2-hydroxyethyl)-1,3-oxazinan-2-one (25f).
A solution of 23f (23 g, 43 mmol) in CH2Cl2 (250 mL) was cooled
6060
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
3.58 (t, J = 6.24 Hz, 1H), 3.10ꢀ2.87 (m, 4H), 2.39ꢀ2.16 (m, 3H),
2.14ꢀ1.89 (m, 2H), 1.56 (d, J = 7.02 Hz, 3H); 19F NMR (CDCl3)
ꢀ111.60, ꢀ114.06, ꢀ114.62. LCꢀMS method 1, tR = 8.79 min, m/z =
470, purity = 99%. HRMS calcd for C27H26F3NO3 + Na. 492.1762,
found 492.1745.
(13) Walker, B. R.; Andrew, R. Tissue production of cortisol by
11beta-hydroxysteroid dehydrogenase type 1 and metabolic disease.
Ann. N.Y. Acad. Sci. 2006, 1083, 165–184.
(14) Tomlinson, J. W.; Stewart, P. M. Modulation of glucocorticoid
action and the treatment of type-2 diabetes. Best Pract. Res., Clin.
Endocrinol. Metab. 2007, 21, 607–619.
(15) Bertagna, X.; Guignat, L.; Groussin, L.; Bertherat, J. Cushing's
disease. Best Pract. Res., Clin. Endocrinol. Metab. 2009, 23, 607–623.
(16) Masuzaki, H.; Paterson, J.; Shinyama, H.; Morton, N. M.;
Mullins, J. J.; Seckl, J. R.; Flier, J. S. A transgenic model of visceral
obesity and the metabolic syndrome. Science (Washington, DC, U. S.)
2001, 294, 2166–2170.
(17) Masuzaki, H.; Yamamoto, H.; Kenyon, C. J.; Elmquist, J. K.;
Morton, N. M.; Paterson, J. M.; Shinyama, H.; Sharp, M. G. F.; Fleming,
S.; Mullins, J. J.; Seckl, J. R.; Flier, J. S. Transgenic amplification of
glucocorticoid action in adipose tissue causes high blood pressure in
mice. J. Clin. Invest. 2003, 112, 83–90.
(18) Kotelevitsev, 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. 11beta-Hydroxysteroid dehydrogenase type 1 knock-
out mice show attenuated glucocorticoid-inducible responses and resist
hyperglycemia on obesity or stress. Proc. Natl. Acad. Sci. U.S.A. 1997,
94, 14924–14929.
(19) 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.
Novel adipose tissue-mediated resistance to diet-induced visceral obe-
sity in 11beta-hydroxysteroid dehydrogenase type 1-deficient mice.
Diabetes 2004, 53, 931–938.
(20) Morton, N. M.; Holmes, M. C.; Fievet, C.; Staels, B.; Tailleux,
A.; Mullins, J. J.; Seckl, J. R. Improved lipid and lipoprotein profile,
hepatic insulin sensitivity, and glucose tolerance in 11beta-hydroxysteroid
dehydrogenase type 1 null mice. J. Biol. Chem. 2001, 276, 41293–
41300.
(21) Andrews, R. C.; Rooyackers, O.; Walker, B. R. Effects of the
11beta-hydroxysteroid dehydrogenase inhibitor carbenoxolone on in-
sulin sensitivity in men with type 2 diabetes. J. Clin. Endocrinol. Metab.
2003, 88, 285–291.
(22) Sandeep, T. C.; Andrew, R.; Homer, N. Z. M.; Andrews, R. C.;
Smith, K.; Walker, B. R. Increased in vivo regeneration of cortisol in
adipose tissue in human obesity and effects of the 11beta-hydroxysteroid
dehydrogenase type 1 inhibitor carbenoxolone. Diabetes 2005, 54,
872–879.
(23) Barf, T.; Vallgårda, J.; Emond, R.; Haeggstroem, C.; Kurz, G.;
Nygren, A.; Larwood, V.; Mosialou, E.; Axelsson, K.; Olsson, R.;
Engblom, L.; Edling, N.; Roenquist-Nii, Y.; Oehman, B.; Alberts, P.;
Abrahmsen, L. Arylsulfonamidothiazoles as a new class of potential
antidiabetic drugs. Discovery of potent and selective inhibitors of the
11beta-hydroxysteroid dehydrogenase type 1. J. Med. Chem. 2002, 45,
3813–3815.
’ ASSOCIATED CONTENT
S
Supporting Information. Synthetic procedures and
b
NMR, MS, and HPLC data on new compounds; assay protocols;
crystallographic data for 25f. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: +1-215-461-2042. Fax: +1-215-461-2006. E-mail: ctice@
vitaerx.com.
’ ABBREVIATIONS USED
HPA, hypothalamicꢀpituitaryꢀadrenal; 11β-HSD1, 11β-hydroxy-
steroid dehydrogenase type 1; 11β-HSD2, 11β-hydroxysteroid
dehydrogenase type 2;DIO, diet-induced obesity; HBA1C, glyco-
sylated hemoglobin
’ REFERENCES
(1) Diabetes Atlas, 4th ed.; IDF: Brussels, Belgium, 2009.
(2) Alberti, K. G. M. M.; Eckel, R. H.; Grundy, S. M.; Zimmet, P. Z.;
Cleeman, J. I.; Donato, K. A.; Fruchart, J.-C.; James, W. P. T.; Loria,
C. M.; Smith, S. C., Jr. Harmonizing the metabolic syndrome: a joint
interim statement of the International Diabetes Federation Task Force
on Epidemiology and Prevention; National Heart, Lung, and Blood
Institute; American Heart Association; World Heart Federation; Inter-
national Atherosclerosis Society; and International Association for the
Study of Obesity. Circulation 2009, 120, 1640–1645.
(3) Mohler, M. L.; He, Y.; Wu, Z.; Hwang, D. J.; Miller, D. D. Recent
and emerging anti-diabetes targets. Med. Res. Rev. 2009, 29, 125–195.
(4) Tanaka, N.; Nonaka, T.; Nakamura, K. T.; Hara, A. SDR:
structure, mechanism of action, and substrate recognition. Curr. Org.
Chem. 2001, 5, 89–111.
(5) Bray, J. E.; Marsden, B. D.; Oppermann, U. The human short-
chain dehydrogenase/reductase (SDR) superfamily: a bioinformatics
summary. Chem.-Biol. Interact. 2009, 178, 99–109.
(6) Jacobson, L. Hypothalamicꢀpituitaryꢀadrenocortical axis
regulation. Endocrinol. Metab. Clin. North Am. 2005, 34, 271–292.
(7) Harno, E.; White, A. Will treating diabetes with 11beta-HSD1
inhibitors affect the HPA axis? Trends Endocrinol. Metab. 2010, 21,
619–627.
(8) Albiston, A. L.; Obeyesekere, V. R.; Smith, R. E.; Krozowski, Z. S.
Cloning and tissue distribution of the human 11beta-hydroxysteroid
dehydrogenase type 2 enzyme. Mol. Cell. Endocrinol. 1994, 105,
R11–R17.
(9) Cooper, M. S.; Stewart, P. M. 11beta-Hydroxysteroid dehydro-
genase type 1 and its role in the hypothalamusꢀpituitaryꢀadrenal axis,
metabolic syndrome, and inflammation. J. Clin. Endocrinol. Metab. 2009,
94, 4645–4654.
(10) Wamil, M.; Seckl, J. R. Inhibition of 11beta-hydroxysteroid
dehydrogenase type 1 as a promising therapeutic target. Drug Discovery
Today 2007, 12, 504–520.
(11) Fotsch, C.; Wang, M. Blockade of glucocorticoid excess at the
tissue level: inhibitors of 11beta-hydroxysteroid dehydrogenase type 1 as
a therapy for type 2 diabetes. J. Med. Chem. 2008, 51, 4851–4857.
(12) Saiah, E. The role of 11beta-hydroxysteroid dehydrogenase in
metabolic disease and therapeutic potential of 11beta-HSD1 inhibitors.
Curr. Med. Chem. 2008, 15, 642–649.
(24) St. Jean, D. J., Jr.; Wang, M.; Fotsch, C. Inhibitors of 11beta-
HSD1: a potential treatment for the metabolic syndrome. Curr. Top.
Med. Chem. (Sharjah, United Arab Emirates) 2008, 8, 1508–1523.
(25) Boyle, C. D. Recent advances in the discovery of 11beta-HSD1
inhibitors. Curr. Opin. Drug Discovery Dev. 2008, 11, 495–511.
(26) Schnackenberg, C. G. 11beta-Hydroxysteroid dehydrogenase
type 1 inhibitors for metabolic syndrome. Curr. Opin. Invest. Drugs
(Thomson Sci.) 2008, 9, 295–300.
(27) Hermanowski-Vosatka, A.; Balkovec, J. M.; Cheng, K.; Chen,
H. Y.; Hernandez, M.; Koo, G. C.; Le Grand, C. B.; Li, Z.; Metzger, J. M.;
Mundt, S. S.; Noonan, H.; Nunes, C. N.; Olson, S. H.; Pikounis, B.; Ren,
N.; Robertson, N.; Schaeffer, J. M.; Shah, K.; Springer, M. S.; Strack,
A. M.; Strowski, M.; Wu, K.; Wu, T.; Xiao, J.; Zhang, B. B.; Wright, S. D.;
Thieringer, R. 11beta-HSD1 inhibition ameliorates metabolic syndrome
and prevents progression of atherosclerosis in mice. J. Exp. Med. 2005,
202, 517–527.
(28) Veniant, M. M.; Hale, C.; Hungate, R. W.; Gahm, K.; Emery,
M. G.; Jona, J.; Joseph, S.; Adams, J.; Hague, A.; Moniz, G.; Zhang, J.;
Bartberger, M. D.; Li, V.; Syed, R.; Jordan, S.; Komorowski, R.; Chen,
6061
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062

Journal of Medicinal Chemistry
ARTICLE
M. M.; Cupples, R.; Kim, K. W.; St. Jean, D. J.; Johansson, L.;
Henriksson, M. A.; Williams, M.; Vallgarda, J.; Fotsch, C.; Wang, M.
Discovery of a potent, orally active 11.beta-hydroxysteroid dehydrogenase
type 1 inhibitor for clinical study: identification of (S)-2-((1S,2S,4R)-
bicyclo[2.2.1]heptan-2-ylamino)-5-isopropyl-5-methylthiazol-4(5Hlic>)-
one (AMG 221). J. Med. Chem. 2010, 53, 4481–4487.
(29) Sundbom, M.; Kaiser, C.; Bjoerkstrand, E.; Castro, V. M.;
Larsson, C.; Selen, G.; Nyhem, C. S.; James, S. R. Inhibition of 11beta
HSD1 with the S-phenylethylaminothiazolone BVT116429 increases
adiponectin concentrations and improves glucose homeostasis in dia-
betic KKAy mice. BMC Pharmacol. 2008, 8, 3.
(30) Hale, C.; Veniant, M.; Wang, Z.; Chen, M.; McCormick, J.;
Cupples, R.; Hickman, D.; Min, X.; Sudom, A.; Xu, H.; Matsumoto, G.;
Fotsch, C.; St. Jean, D. J., Jr.; Wang, M. Structural characterization and
pharmacodynamic effects of an orally active 11beta-hydroxysteroid
dehydrogenase type 1 inhibitor. Chem. Biol. Drug Des. 2008, 71, 36–44.
(31) Wan, Z.-K.; Chenail, E.; Xiang, J.; Li, H.-Q.; Ipek, M.; Bard, J.;
Svenson, K.; Mansour, T. S.; Xu, X.; Tian, X.; Suri, V.; Hahm, S.; Xing,
Y.; Johnson, C. E.; Li, X.; Qadri, A.; Panza, D.; Perreault, M.; Tobin, J. F.;
Saiah, E. Efficacious 11β-hydroxysteroid dehydrogenase type I inhibitors
in the diet-induced obesity mouse model. J. Med. Chem. 2009, 52,
5449–5461.
(32) Whittamore, P. R. O. Discovery of a Potent, Selective and
Orally Bioavailable Acidic 11beta-Hydroxysteroid Dehydrogenase Type
1 (11beta-HSD1) Inhibitor. Presented at EFMC-ISMC 2010, Brussels,
Belgium, 2010.
(33) Rosenstock, J.; Banarer, S.; Fonseca, V. A.; Inzucchi, S. E.; Sun,
W.; Yao, W.; Hollis, G.; Flores, R.; Levy, R.; Williams, W. V.; Seckl, J. R.;
Huber, R. The 11-beta-hydroxysteroid dehydrogenase type 1 inhibitor
INCB13739 improves hyperglycemia in patients with type 2 diabetes
inadequately controlled by metformin monotherapy. Diabetes Care
2010, 33, 1516–1522.
(34) Feig, P. U.; Shah, S.; Hermanowski-Vosatka, A.; Plotkin, D.;
Springer, M. S.; Donahue, S.; Thach, C.; Klein, E. J.; Lai, E.; Kaufman,
K. D. Effects of an 11β-hydroxysteroid dehydrogenase type 1 inhibitor,
MK-0916, in patients with type 2 diabetes mellitus and metabolic
syndrome. Diabetes, Obes. Metab. 2011, 13, 498–504.
(35) Zhang, J.; Osslund, T. D.; Plant, M. H.; Clogston, C. L.; Nybo,
R. E.; Xiong, F.; Delaney, J. M.; Jordan, S. R. Crystal structure of murine
11beta-hydroxysteroid dehydrogenase 1: an important therapeutic
target for diabetes. Biochemistry 2005, 44, 6948–6957.
(36) Patel, J. R.; Shuai, Q.; Dinges, J.; Winn, M.; Pliushchev, M.;
Fung, S.; Monzon, K.; Chiou, W.; Wang, J.; Pan, L.; Wagaw, S.;
Engstrom, K.; Kerdesky, F. A.; Longenecker, K.; Judge, R.; Qin, W.;
Imade, H. M.; Stolarik, D.; Beno, D. W. A.; Brune, M.; Chovan, L. E.;
Sham, H. L.; Jacobson, P.; Link, J. T. Discovery of adamantane ethers as
inhibitors of 11beta-HSD-1: synthesis and biological evaluation. Bioorg.
Med. Chem. Lett. 2007, 17, 750–755.
(37) Andersson, A.; Jordan, D.; Schneider, G.; Lindqvist, Y. Crystal
structure of the ternary complex of 1,3,8-trihydroxynaphthalene reduc-
tase from Magnaporthe grisea with NADPH and an active-site inhibitor.
Structure (London) 1996, 4, 1161–1170.
(38) “Contour” is a structure-based drug design tool that grows
molecules in silico, in the context of a protein binding site, by addition of
fragments to a user selected starting template: Singh, S. B. Manuscript in
preparation.
(39) Solly, K.; Mundt, S. S.; Zokian, H. J.; Ding, G. J.-F.;
Hermanowski-Vosatka, A.; Strulovici, B.; Zheng, W. High-throughput
screening of 11beta-hydroxysteroid dehydrogenase type 1 in scintilla-
tion proximity assay format. Assay Drug Dev. Technol. 2005, 3, 377–384.
(40) Nagao, M.; Yahagi, T.; Honda, M.; Seino, Y.; Matsushima, T.;
Sugimura, T. Demonstration of mutagenicity of aniline and o-toluidine
by norharman. Proc. Jpn. Acad., Ser. B 1977, 53, 34–37.
(41) Hecht, S. S.; El-Bayoumy, K.; Tulley, L.; LaVoie, E. Structureꢀ
mutagenicity relationships of N-oxidized derivatives of aniline, o-toluidine,
20-methyl-4-aminobiphenyl, and 3,20-dimethyl-4-aminobiphenyl. J. Med.
Chem. 1979, 22, 981–987.
6062
dx.doi.org/10.1021/jm2005354 |J. Med. Chem. 2011, 54, 6050–6062