Distinctive molecular inhibition mechanisms for selective inhibitors of human 11β-hydroxysteroid dehydrogenase type 1

Article abstract of DOI:10.1016/j.bmc.2008.08.065

11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) catalyzes the NADPH dependent interconversion of inactive cortisone to active cortisol. Excess 11β-HSD1 or cortisol leads to insulin resistance and metabolic syndrome in animal models and in humans. Inhibiting 11β-HSD1 activity signifies a promising therapeutic strategy in the treatment of Type 2 diabetes and related diseases. Herein, we report two highly potent and selective small molecule inhibitors of human 11β-HSD1. While compound 1, a sulfonamide, functions as a simple substrate competitive inhibitor, compound 2, a triazole, shows the kinetic profile of a mixed inhibitor. Co-crystal structures reveal that both compounds occupy the 11β-HSD1 catalytic site, but present distinct molecular interactions with the protein. Strikingly, compound 2 interacts much closer to the cofactor NADP⁺ and likely modifies its binding. Together, the structural and kinetic analyses demonstrate two distinctive molecular inhibition mechanisms, providing valuable information for future inhibitor design.

Full text of DOI:10.1016/j.bmc.2008.08.065

Bioorganic & Medicinal Chemistry 16 (2008) 8922–8931
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Distinctive molecular inhibition mechanisms for selective inhibitors
of human 11b-hydroxysteroid dehydrogenase type 1
Hua Tu a, Jay P. Powers b, , , Jinsong Liu c,à, Stefania Ursu a, Athena Sudom c, Xuelei Yan b, Haoda Xu c,
*
David Meininger c, Michael DeGraffenreid b, Xiao He b, Juan C. Jaen b, , Daqing Sun b, Marc Labelle b,§
,
Hiroshi Yamamoto d, Bei Shan a, Nigel P. C. Walker c, Zhulun Wang c,
*
a Department of Metabolic Disorders, Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, USA
b Department of Medicinal Chemistry, Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, USA
c Department of Molecular Structure, Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, USA
d Chemical Research Laboratories, Central Pharmaceutical Research Institute, Japan Tobacco Inc., 1-1 Murasaki-cho, Takatsuki, Osaka 569-1125, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
11b-hydroxysteroid dehydrogenase type 1 (11b-HSD1) catalyzes the NADPH dependent interconversion
of inactive cortisone to active cortisol. Excess 11b-HSD1 or cortisol leads to insulin resistance and met-
abolic syndrome in animal models and in humans. Inhibiting 11b-HSD1 activity signifies a promising
therapeutic strategy in the treatment of Type 2 diabetes and related diseases. Herein, we report two
highly potent and selective small molecule inhibitors of human 11b-HSD1. While compound 1, a sulfon-
amide, functions as a simple substrate competitive inhibitor, compound 2, a triazole, shows the kinetic
profile of a mixed inhibitor. Co-crystal structures reveal that both compounds occupy the 11b-HSD1 cat-
alytic site, but present distinct molecular interactions with the protein. Strikingly, compound 2 interacts
much closer to the cofactor NADP+ and likely modifies its binding. Together, the structural and kinetic
analyses demonstrate two distinctive molecular inhibition mechanisms, providing valuable information
for future inhibitor design.
Received 8 April 2008
Revised 20 August 2008
Accepted 26 August 2008
Available online 29 August 2008
Keywords:
11b-Hydroxysteroid dehydrogenase (11b-
HSD1)
Small molecule inhibitor
Inhibition mechanism
Enzyme kinetics
Crystal structure
X-ray crystallography
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
of 11b-HSD1 in the fat tissue of transgenic mice produced pheno-
types similar to human metabolic syndrome, such as insulin-
Glucocorticoids are steroid hormones that are produced by the
hypothalamus–pituitary–adrenal (HPA) axis as part of a normal
stress response and have profound effects on blood glucose level
and metabolism. Glucocorticoid excess in humans results in a
variety of metabolic phenotypes, including glucose intolerance,
central obesity, hypertension, dyslipidemia, and increased cardio-
vascular mortality, as seen in Cushing’s Syndrome.1 11b-HSD1,
widely expressed in tissues, is the principal enzyme that gener-
ates active glucocorticoids (cortisol) from inactive ones (corti-
sone) in a NADPH dependent manner, and has been proposed
as an amplifier of glucocorticoid action.2,3 Indeed, overexpression
resistance, hyperlipidemia and hyperphagia.4 The therapeutic
benefit of inhibiting 11b-HSD1 activity has been demonstrated
by treatment with the non-selective 11b-HSD inhibitor carbenox-
olone in men with type 2 diabetes, in which carbenoxolone
reduced the glucose production rate.5 Furthermore, several selec-
tive small molecule 11b-HSD1 inhibitors have been recently
shown to improve insulin sensitivity and metabolic syndrome in
preclinical animal models.6,7 Taken together, these results suggest
that 11b-HSD1 is a promising pharmaceutical target for novel
therapy in the clinical treatment of prevalent type 2 diabetes
and other related metabolic disorders.8–11
The development of potent and selective small molecule inhibi-
tors for 11b-HSD1 is currently being pursued vigorously by the
pharmaceutical industry.12–18 Overall, most compounds are active
site inhibitors. However, their detailed kinetic and molecular mech-
anisms have not been reported to date. Here we present two novel
and selective inhibitors of 11b-HSD1 as well as their kinetic and
co-crystal structural analyses. The distinct molecular interactions
and kinetic profiles of the different inhibitors revealed in this report
provide the first detailed discussion, to our knowledge, of the
molecular inhibition mechanism of 11b-HSD1 by small molecules.
* Corresponding authors. Tel.: +1 650 244 2446; fax: +1 650 837 9427 (Z.W.); tel.:
+1 650 210 2949; fax: +1 650 625 8940 (J.P.P.).
E-mail addresses: jpowers@chemocentryx.com (J.P. Powers), zwang@amgen.com
(Z. Wang).
Present address: ChemoCentryx Inc., 850 Maude Avenue, Mountain View, CA
94043, USA.
à
Present address: Guangzhou Institute of Biomedicine and Health, Guangzhou
Science Park, 510663, China.
§
Present address: Lundbeck Research USA Inc., 215 College Road, Paramus, NJ
07652, USA.
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.08.065

H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
8923
ment with one equivalent of (R)-2-methylpiperazine in acetonitrile
at 70 °C, and the resulting amine was then acylated with 4-acetyl-
benzenesulfonyl chloride to give N,N-disubstituted piperazine 7 in
36% overall yield from alcohol 4 (three steps). Ketone 7 was con-
verted into a 1:1 diastereomeric mixture of trifluoromethylcarbi-
nols 8 in excellent yield by treatment with TMSCF3 and TBAF in
THF. Resolution of the resulting diastereoisomers was accom-
plished by use of a carbamate chiral auxiliary. Formation of the
mixed carbonate resulting from treatment of 8 with 4-nitrophenyl
chloroformate and DMAP was followed by displacement of the
nitrophenylphenol with commercially available (S)-(+)-2-amino-
1-butanol in a single pot. The resulting mixture of diastereoiso-
mers was separated via flash chromatography to give the pure de-
sired diastereoisomer 9 in 35% yield from 8. Treatment of 9 with
KOH in tert-butanol at 90 °C gave compound 1 in good yield via
hydrolysis of the nitrile to the primary amide with concomitant
hydrolysis of the carbamate to the desired trifluoromethylcarbinol.
Hydrolysis of the carbamate is believed to be facilitated by intra-
molecular cyclization of the auxiliary hydroxyl group to give the
oxazolidinone, and is quite facile. We found that non-hydroxyl
bearing amines used as the auxiliary carbamate may be hydrolyzed
using the above method as well, albeit at a slower rate, presumably
due to their inability to pass through the putative oxazolidinone.
The synthesis of compound 2 (Scheme 2) began with the cyclo-
propanation of commercially available benzonitrile 10 to give 11 in
good yield. Amide formation via alkaline hydrolysis of the nitrile
proceeded in excellent yield to give 12, followed by acid hydrolysis
to give acid 13. Fischer esterification of the acid gave ester 14 in
excellent overall yield. Treatment of ester 14 with hydrazine hy-
drate in refluxing ethanol afforded hydrazide 15. Commercially
available acid chloride 16 afforded benzamide 17 after treatment
with 2-aminopropane in chloroform, and treatment of 17 with
2. Results and discussion
2.1. Chemistry
In order to develop potent and selective small molecule inhibi-
tors of human 11b-HSD1 enzyme activity, we performed high
throughput screening of a small molecule compound library and
optimized selected leads.17–19 This effort resulted in two synthetic
inhibitors, shown in Figure 1, which were derived from two dis-
tinct chemical series and were extensively characterized. Com-
pound 1 is a representative of the sulfonamide series,17 while
compound 2 has the triazole moiety as its central chemical fea-
ture.19 The synthesis of compound 1 (Scheme 1) began with reduc-
tion of the commercially available cyclopropyl ester 3 with sodium
borohydride in a mixed (DME/MeOH) solvent system to give alco-
hol 4 in good yield. Alcohol 4 was then converted to its tosylate 5
by treatment with p-TsCl and triethylamine in dichloromethane.
The crude tosylate 5 was directly converted to amine 6 by treat-
CF3
O
HO
N
NH2
1
2
N
O
S
O
F
N
N
F3CO
N
Figure 1. Chemical structures of compounds 1 and 2.
b
a
EtO
CN
HO
HN
TsO
CN
CN
O
3
4
5
NH
HN
CN
N
c
d
6
O
CF3
HO
CN
N
CN
e
N
N
S
O
N
O
S
O
O
7
8
f
H
N
CF3
O
CF3
HO
Et
HO
O
g
CN
N
NH2
N
O
N
N
S
S
O
O
O
O
1
9
Scheme 1. Synthesis of compound 1. Reagents: (a) NaBH4, DME/MeOH, 74%. (b) p-TsCl, Et3N, CH2Cl2. (c) (R)-(-)-2- Methylpiperazine, CH3CN. (d) 4-Acetylbenzenesulfonyl
chloride, Et3N, CH2Cl2, yield 36% for steps b-d (4 to 7). (e) TMS-CF3, TBAF, THF, yield 84%. (f) i. 4-Nitrophenyl chloroformate, DMAP, CH3CN. ii. (S)-(+)-2-Amino-1-butanol.
iii. Separation of the diastereoisomers via flash chromatography; yield of 9 from 8 35%. (g) Potassium hydroxide, tert-butyl alcohol, 79%.

8924
H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
a
b
CN
CN
CONH2
F
F
F
10
11
12
c
H
N
e
NH2
R
O
F
F
13: R = CO2H
14
15
d
: R = CO2Me
O
O
S
g
f
Cl
N
H
N
H
F3CO
F3CO
F3CO
16
17
18
h
F
SMe
N
N
N
i
F3CO
N
F3CO
2
19
Scheme 2. Synthesis of compound 2. Reagents and conditions: (a) 1-Bromo-2-chloroethane, benzyltriethylammonium chloride, NaOH, H2O, 50 °C, 68%. (b) H2O2, NaOH,
acetone/H2O, 93%. (c) 6 N HCl, 1,4-dioxane, reflux, 98%. (d) Concentrated H2SO4, MeOH, reflux 8 h, quantitative. (e) Hydrazine hydrate, refluxing EtOH, 76%. (f) Iso-
propylamine, CHCl3, 0 °C, 93%. (g) Lawesson’s reagent, THF, reflux, quantitative. (h) Methyl iodide, THF, 75%. (i) Compound 15, pyridine, reflux, 5 h, 35%.
Lawesson’s reagent gave thioamide 18, which was readily con-
verted to thioimidate 19 in good yield. Combining thioimidate 19
in refluxing pyridine with hydrazide 15 afforded triazole inhibitor
2 in fair yield.
2.2. Inhibition of 11b-HSD1
Physiologically, 11b-HSD1 functions as a reductase enzyme
(Figure 2A). We reconstituted an in vitro reductase assay with
Figure 2. (A) Reaction scheme catalyzed by 11b-HSD1 reductase activity in vivo. Concentration–response curves for compounds 1 and 2 as inhibitors of 11b-HSD1 reductase
(B) and dehydrogenase (C) activity. The reductase reaction was constituted with cortisone and cofactor NADPH, while the dehydrogenase reaction was made with cortisol and
cofactor NADP+.

H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
8925
recombinant full-length human 11b-HSD1 purified from a baculo-
virus expression system (Invitrogen) to measure the inhibitory po-
tency of compounds 1 and 2. Various concentrations of compound
1 and 2 were incubated with 11b-HSD1, [3H]-cortisone and cofac-
tor NADPH at 37 °C for 90 min in 96-well plate format. At the end
of the incubation period, the reaction was stopped with the non-
specific 11b-HSD1 inhibitor carbenoxolone, and the reaction prod-
uct [3H]-cortisol was captured and quantified by anti-cortisol
monoclonal antibody in a scintillation proximity assay (SPA). Per-
cent inhibition was calculated based on reference controls, and
IC50 values were determined using a curve fitting algorithm based
on percent inhibition and the corresponding compound concentra-
tions. As shown in Figure 2B, both compound 1 and 2 could po-
tently and completely inhibit the reductase reaction of 11b-
HSD1. The biochemical IC50 is 1.2 0.7 nM for compound 1, and
1.6 0.5 nM for compound 2. Under similar assay conditions, nei-
ther compound 1 nor 2 significantly inhibited several related ste-
roid enzymes, including human 11b-HSD2 and 17b-HSD1 at
value of 1/V less than zero), indicating compound 2 decreased both
the Vmax and Km of NADPH. Interestingly, in a substrate competi-
tion experiment (Figure 3F), compound 2 also showed a mixed
inhibition profile with the value of alpha >1 (lines intersecting at
a value of 1/V greater than zero), indicating compound 2 decreased
Vmax but increased Km of substrate. These two complementary sets
of kinetic data suggest that inhibition of 11b-HSD1 by compound 2
is much more complex than that of compound 1, and cannot be
completely overcome by an excess of either cofactor or substrate.
2.4. Co-crystal structures of 11b-HSD1 with inhibitors
To further elucidate the molecular mechanism of inhibition, we
carried out structural analysis of 11b-HSD1 co-crystallized with
inhibitors using X-ray crystallography. A truncated version of re-
combinant human 11b-HSD1 (residues 24–292 with a single muta-
tion of C272S) was produced from E. coli. This 11b-HSD1 was
purified in the presence of the steroidal detergent CHAPS (3-[(3-
cholamidopropyl)dimethylammonio]-1-propanesulfonate). Crys-
tals produced from this protein contained cofactor NADP+ and
CHAPS bound to the enzyme, although neither NADP+ nor NADPH
was supplemented during the purification and crystallization
steps. Co-crystals of 11b-HSD1 with each inhibitor were then gen-
erated by soaking the 11b-HSD1 crystal with the inhibitors individ-
ually. The crystals belong to space group P21 with four protein
molecules in the asymmetric unit which correspond to two dimers.
The co-crystal structures of 11b-HSD1 in complex with compound
1 or compound 2 were determined by molecular replacement to
2.5 and 2.55 Å, respectively, using the previously reported 11b-
HSD1 structure as a search model (PDB code: 1XU9).20 The final
electron density maps of 11b-HSD1 are clear, except for some dis-
ordered regions including the loop of residues 228 to 233 and the
C-terminal loop of residues 281 to 292. The X-ray data statistics are
shown in Table 1.
10
lM, suggesting that both compounds are specific for 11b-HSD1.
Since 11b-HSD1 also has dehydrogenase activity in vitro, we
measured the inhibitory activity of compounds 1 and 2 in the
dehydrogenase assay as well. Reaction mixtures containing 11b-
HSD1, [14C]-cortisol, cofactor NADP and a range of compound con-
centrations were incubated at 37 °C for 1 h before being extracted
with ethyl acetate. [14C]-Cortisol and reaction product [14C]-corti-
sone in the organic extract were then separated by thin layer chro-
matography (TLC), and were quantified with a phosphor-imager
system. Percent inhibition was calculated based on references,
and plotted against compound concentrations. As shown in Figure
2C, although compound 1 and 2 could both inhibit the dehydroge-
nase reaction of 11b-HSD1, the two compounds displayed very dif-
ferent inhibitory potency. Compound 2 appeared to be about 50-
fold less potent than compound 1 in inhibiting the dehydrogenase
reaction, while, as shown before, the two compounds had compa-
rable potency in inhibiting the reductase reaction of the same en-
zyme. This set of data suggests that the two compounds have very
different inhibition characteristics.
The overall structure of the 11b-HSD1 protein is similar to the
one reported previously by Hosfield et al.20 Resembling other short
chain dehydrogenase/reductase (SDR) enzymes, the core structure
of 11b-HSD1 (Figure 4) adopts a Rossmann fold with a central par-
2.3. Kinetic analysis of inhibitors
allel 6-stranded b-sheet having 3
addition, 11b-HSD1 has an insertion of b6-
tended-strand conformation, an extra b-strand (b7) packed in par-
allel with the b6 strand, and two C-terminal -helices appended to
a
-helices at each side.21–23 In
6 in an -helix and ex-
a
a
Kinetic analysis was performed on the two compounds in vitro
to understand their respective mechanisms of inhibition of 11b-
HSD1. We used the same conditions as the aforementioned
biochemical assays, but with either the cofactor NADPH or the sub-
strate cortisol as a variable. Figure 3A shows the Lineweaver–Burk
plot of compound 1 using NADPH as a variable, in which com-
pound 1 decreased the apparent Vmax of NADPH without affecting
the Km, indicating the inhibitor is non-competitive with respect to
NADPH. As shown in Figure 3B, the secondary plot for reciprocal
Vmax as a function of compound 1 concentration yielded a Ki of
1.1 nM, similar to the IC50 value determined above. Figure 3C
shows the Lineweaver–Burk plot of compound 1 in the presence
of various concentrations of cortisol. Reciprocally, compound 1 in-
creased the Km for cortisol while leaving the Vmax unchanged, dem-
onstrating that enzyme inhibition by compound 1 is substrate-
competitive and reversible. The secondary plot (Figure 3D) of
apparent Km as a function of compound 1 concentration yielded
a Ki of 2.7 nM. These two sets of kinetic data complement each
other and clearly suggest that compound 1 inhibits 11b-HSD1 by
occupying its substrate binding site without affecting the cofactor
NADPH binding.
a
the core structure. The cofactor NADP+ shows well-defined electron
density in the structure. The binding of NADP+ is similar to that of
other SDR enzymes, with the molecule in the same extended con-
formation as previously described, forming both extensive hydro-
philic and hydrophobic interactions with the protein.20,24,25
The co-crystal structure of the 11b-HSD1 with inhibitor com-
pound 1 showed well resolved electron density for compound 1
in all four protein molecules (Figure 5A). Compound 1 binds in a
V-shaped conformation, snugly in the narrow steroid binding
channel walled by the b6–a6 insertion at one side and the loop
b5– 5 on the other side, with the trifluoromethyl hydroxyl phenyl
a
moiety approaching the catalytic end of the substrate binding
pocket and the cyclopropane carboxamide towards the solvent
area where the C-terminal of an adjacent molecule in the crystal
lattice packs. The inhibitor binding is almost identical in all four
protein molecules. Compound 1 makes one direct hydrogen bond
with 11b-HSD1 with one oxygen atom of the sulfonamide moiety
accepting a proton from the backbone amide of Ala172 in loop
b5–a5 (Figure 5B). The rest of the compound makes numerous
A similar set of kinetic characterizations with compound 2,
however, showed considerably different profiles. As shown in
Figure 3E of the Lineweaver–Burk plot of compound 2 with NADPH
as a variable, compound 2 displayed a mixed inhibition profile with
respect to NADPH with the value of alpha <1 (lines intersecting at a
van der Waals contacts with the protein, including residue
Tyr183 from the catalytic Tyr-X-X-X-Lys motif. At the substrate
recognition end, compound 1 also interacts with NADP+, where
the phenyl ring makes
a van der Waals contact with the
nicotinamide ring. In addition, the hydroxyl moiety forms a

8926
H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
Figure 3. The primary (Lineweaver–Burk) and secondary plots of the kinetic assays. (A) Lineweaver–Burk plot of compound 1 with cofactor NADPH. (B) Secondary plot of Ki
determination for compound 1 with NADPH. (C) Lineweaver–Burk plot of compound 1 with substrate cortisol. (D) Secondary plots of Ki determination for compound 1 with
substrate cortisol. (E) Lineweaver–Burk plot of compound 2 with NADPH and (F) with substrate cortisol.
water-mediated hydrogen bond with one pyrophosphate oxygen
of NADP+. However, compound 1 does not change the binding state
of NADP+ to 11b-HSD1, which is consistent with our kinetic find-
ings that compound 1 is non-competitive with NADPH.
cepts a hydrogen bond from the hydroxyl moieties of Tyr183 and
Ser170, respectively. In the parental crystal structure without
inhibitor, where the steroid detergent molecule CHAPS occupies
the substrate-binding site, the side chain of Ser170 points to
Tyr183 as we previously observed in the compound 1 co-crystal
structure (Figure 5B), with the hydroxyl moiety forming a hydro-
gen bond with Tyr183. The side chain of Tyr183 points to the cofac-
tor NADP+ and makes a hydrogen bond with one hydroxyl of the
ribose. Here, the binding of compound 2 results in the Ser170 side
chain rotating 180° and pointing away from Tyr183. The rest of the
ligand–protein interactions are mainly hydrophobic in nature. The
In the co-crystal structure of 11b-HSD1 with compound 2, the
inhibitor also showed well-defined electron density in all four pro-
tein molecules (Figure 5C). Compound 2 is positioned in the steroid
substrate binding site in an L-shaped conformation with the flu-
oro-phenyl group as the short arm, the cyclopropane moiety being
the turning point, with triazole and trifluoromethoxy phenyl being
on the long line (Figure 5D). Interestingly, the triazole moiety is sit-
uated right at the catalytic center, and makes two direct hydrogen
bonds with residues Tyr183 and Ser170 of the catalytic triad
(Ser170, Tyr183, and Lys187) responsible for substrate reduction.
Each of the two unsubstituted nitrogen atoms of the triazole ac-
fluoro-phenyl moiety points towards the b4–a4 loop that passes
on the top of the substrate-binding cleft. The phenyl ring makes
van der Waals contacts with the side chains of Thr124 and
Leu126, and the fluorine atom is in van der Waals contact with

H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
8927
Table 1
than the conventional van der Waals distance between C and N
atoms. While compound 1 occupies only the substrate site and is
in van der Waals contact with NADP+, compound 2 not only resides
in the substrate site and engages in interactions with two residues
of the catalytic triad, but also impinges partially on the cofactor
site. Interactions of the triazole ring of compound 2 with the two
catalytic residues evidently lessen the interactions of these resi-
dues with NADP+. However, this is most likely compensated for
by the interactions between NADP+ and the anchored compound
2. Together, this set of interactions likely increased the affinity of
NADP(H) to 11b-HSD1, a conclusion which is supported by the ki-
netic analysis.
Data and refinement statisticsa
Complex
11b-HSD1 + compound 1
11b-HSD1 + compound 2
Crystal wavelength
(Å)
Space group
Cell constants
1.54
1.0
P21
P21
a = 56.6 Å, b = 153.8 Å,
c = 73.9 Å, b = 93.1
2.5
100,110
40,775
a = 56.7 Å, b = 153.3 Å,
c = 73.6 Å, b = 92.2
2.55
147,508
40,780
Resolution (Å)
Total reflections
No. of unique
reflections
Completeness (%)
(outer shell)
93.7 (89.7)
99.8 (100.0)
Rsym (%) (outer shell) 12.3 (16.4)
4.3 (44.2)
10.2 (1.6)
21.8
3. Conclusion
I/
r
(outer shell)
3.9 (3.8)
21.0
Rcryst (%)
Rfree (%)
RMS deviations:
Bonds (Å)
RMS deviations:
Angles (deg)
Total non-H atoms
28.9
0.014
27.8
0.020
In summary, we have reported two novel 11b-HSD1 inhibitors
and their molecular interactions with 11b-HSD1 by kinetic analysis
and X-ray crystallography. The structures revealed that a set of
amino acid residues in the active site, including Ala172 and the
catalytic triad Ser170, Tyr183, and Lys187, makes important inter-
actions with inhibitors and is therefore the key determinant of
binding affinity for these small molecule inhibitors. The two com-
pounds presented here are of different inhibitor classes, as shown
by their chemical structures and kinetic profiles, as well as two dif-
ferent binding modes, even though they both occupy the steroid
substrate binding site. Compound 1 behaves as a simple substrate
competitive inhibitor. Structurally, it occupies the substrate-bind-
ing site and neither perturbs the catalytic site nor affects the
NADP(H) binding. In contrast, compound 2 inhibits 11b-HSD1 in
a mixed inhibitor mode. Structurally, it not only occupies the sub-
strate binding site but also alters the catalytic site residues and
NADP(H) binding. Furthermore, compound 2 imposes tight interac-
tions with NADP+, enlightening the observed kinetic profile of a
mixed inhibitor.
1.65
2.10
8758
8447
P
P
Rsym
=
|Iavg À Ij |/ Ij.
P
P
R = |Fo À Fc|/ Fo, where Fo and Fc are observed and calculated structure factors,
respectively, Rfree was calculated from a randomly chosen 5% of reflections excluded
from the refinement, and Rcryst was calculated from the remaining 95% of
reflections.
a
RMS deviation is the root-mean-square deviation from ideal geometry. Num-
bers in parentheses are for the highest resolution shell.
NADPH is an essential cofactor for the 11b-HSD1 reduction
reaction. Recent literature has shown that local NADPH concentra-
tion generated by hexose-6-phosphate dehydrogenase (H6PD)
might be critical for the reductase activity of 11b-HSD1, which is
confined to the ER lumen.26 In fact, H6PD deficiency results in a
marked reduction of 11b-HSD1 activity.26 Thus, the coupled mod-
ulation of compound 2 on both substrate and NADP(H) binding is
noteworthy for further investigation. Conceivably, the two inhibi-
tion mechanisms might yield distinct in vivo efficacy profiles,
which might be an interesting subject for future studies. Our re-
sults provide important insight into both the inhibition mechanism
of 11b-HSD1 and future inhibitor design.
4. Experimental
Figure 4. Overall co-crystal structure of 11b-HSD1 with compound 1. The protein is
shown as a ribbon diagram, spectrum color coded with N-terminal colored in blue
to C-terminal in red. The cofactor NADP+ and compound 1 are shown in a molecular
surface representation which is color coded red for oxygen, blue for nitrogen,
orange for phosphorus, magenta for fluorine and cyan for carbon in compound 1
and grey for carbon in NADP+. The dimer partner is shown with a purple ribbon.
Reagents and solvents used below were obtained from commer-
cial sources and when required were purified in the accepted fash-
ion. 1H NMR spectra were obtained on a Varian Gemini 400 MHz
NMR spectrometer. Electrospray ionization (ESI) mass spectrome-
try analysis was performed using a Hewlett-Packard 1100 MSD
electrospray mass spectrometer using the HP1100 HPLC for sample
delivery. Combustion analyses were performed by Atlantic Micro-
lab Inc. (Norcross, GA). Air and/or moisture sensitive reactions
were carried out under N2 using flame dried glassware and stan-
dard syringe/septa techniques.
the backbones of Ser125 and Leu126. The trifluoromethoxy phenyl
is oriented towards the solvent area with the trifluoromethoxy
group occupying the space normally occupied by the side chain
of Tyr177. As a result, Tyr177 swings up and adopts a different side
chain conformation as compared with the parental structure and
the compound 1 co-crystal structure.
In addition to the above ligand–protein interactions, we ob-
served that the electron density of compound 2 connects to the
density of NADP+ with the triazole edge stacking on the nicotin-
amide ring (Figure 6). The distances between the two unsubstitut-
ed nitrogen atoms of the triazole to the C4 and C5 atoms of the
nicotinamide ring are merely 3.0 Å, respectively, which are less
4.1. Synthesis of (1-(((R)-3-methyl-4-(4-((S)-1,1,1-trifluoro-2-
hydroxypropan-2-yl)phenylsulfonyl)-piperazin-1-yl)methyl)-
cyclopropanecarboxamide) (1)
1-(Hydroxymethyl)cyclopropanecarbonitrile (4): To a solution
of 48.19 g 1-cyanocyclopropanecarboxylic acid ethyl ester

8928
H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
Figure 5. The inhibitor binding mode of 11b-HSD1. (A) Compound 1, shown in stick representation, in the active site. Electron density maps, shown in blue and magenta for
compound 1 and NADP+, respectively, are contoured at a level of 1
from a 2FoÀFc map. (B) Protein–ligand interactions for compound 1. (C) Compound 2, shown in stick
representation with carbon atoms in light green, in the active site. Electron density maps, shown in blue and magenta for compound 2 and NADP+, respectively, are contoured
at a level of 1
from a 2FoÀFc map. (D) Protein–ligand interactions for compound 2. The hydrogen bonds are shown in magenta dotted lines. Coloring is as in Figure 4.
r
r
Figure 6. The interactions between inhibitor and cofactor NADP+, shown in molecular surface representation. (A) Compound 1 and (B) compound 2. Coloring is as in Figure 5.
(0.35 mol, 1.0 equiv) in 200 mL ethylene glycol dimethyl ether and
20 mL MeOH was added 26.2 g NaBH4 (0.69 mol, 2.0 equiv) in por-
tions at 0 °C. The reaction mixture was stirred at room temperature
for 24 h, followed by the dropwise addition of 50 ml sat. NH4Cl to
the reaction mixture at 0 °C. The resulting suspension was stirred
at room temperature for 24 h, filtered, and the solid was washed
with 10% MeOH/DME. The solid was then acidified with 2 N HCl
at 0 °C to pH 6–7 and further extracted with 10% MeOH/CH2Cl2.
The organic layers were then combined, dried (MgSO4), and con-
centrated under reduced pressure. Purification by distillation (bp
85–87 °C @ 5–10 mmHg) gave 25.0 g of the product as a colorless
oil (0.26 mol, 74%): 1H NMR (DMSO-d6, 400 MHz) d 3.39 (d,
J = 6.0 Hz, 2 H), 1.14 (m, 2H), 0.92 (m, 2H).
(1-Cyanocyclopropyl)methyl-4-methylbenzenesulfonate (5): To
150 g of the 1-(hydroxymethyl)-cyclopropanecarbonitrile 4 (pre-
pared above, 154 mmol, 1.0 equiv) in 3.0 L CH2Cl2 was added
311.7 g triethylamine (3.08 mol, 2.0 equiv) and 324.7 g p-toluene-
sulfonyl chloride (1.7 mol, 1.1 equiv) at 0 °C. The solution was al-
lowed to stir at 0 °C for 15 min, and then at room temperature
for an additional 18 h. The solution was then re-cooled and 20 ml

H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
8929
2-(dimethylamino)ethanol was added to the reaction mixture at
0 °C, which was then allowed to stir for an additional 2 h. 1.5 L
1 N HCl was added and the resulting solution was stirred for an
additional 30 min. The organic phase was separated, washed
(1Â 1 N HCl, 1Â H2O, 1Â sat. NaHCO3), dried (Na2SO4), and concen-
trated under reduced pressure to give 254 g of the crude product
tosylate 5 as light brown solid.
(R)-1-((3-Methylpiperazin-1-yl)methyl)cyclopropanecarbonitrile
(6): The crude tosylate 5 obtained above (252 g, 1.0 mol, 1.0 equiv)
in 1.2 L CH3CN was combined in a 3 L flask with 200 g (R)-(À)-2-
methylpiperazine (2.0 mol, 2.0 equiv). The resulting mixture was
then placed into a preheated 70 °C bath. After stirring for 12 h,
the mixture was cooled to room temperature and concentrated un-
der reduced pressure. The resulting residue was diluted with sat.
NaHCO3, extracted (3Â 10% MeOH/CH2Cl2), dried (MgSO4), and
concentrated under reduced pressure. The resulting amine (6)
was taken directly to the next step without any additional
purification.
was added 67.9 g DMAP (55.6 mol, 1.5 equiv) at 0 °C, followed by
the addition of 4-nitrophenyl chloroformate (89.7 g, 44.5 mol, 1.2
equiv) in portions. The resulting mixture was allowed to stir for
15 min at 0 °C followed by 5.5 h at room temperature. 56.2 g (S)-
(+)-2-amino-1-butanol (63.1 mmol, 1.7 equiv) was added dropwise
via addition funnel. After the addition was complete, the solution
was allowed to stir for an additional 12 h. Most of the CH3CN
was removed under reduced pressure and the residue was diluted
with EtOAc. The solid was filtrated off and washed with EtOAc. The
combined liquids were then washed with sat. NH4Cl, dried
(MgSO4), and concentrated under reduced pressure. The two dia-
stereomers obtained were purified and separated by flash chroma-
tography (SiO2, 50% EtOAc /hexanes). The first portion of the two
close spots was collected and concentrated under reduced pressure
to give 70 g of the desired pure diastereomer 9 as colorless oil
(12.8 mmol, 34.5%).
1-(((R)-3-Methyl-4-(4-((S)-1,1,1-trifluoro-2-hydroxypropan-2-
yl)phenylsulfonyl)piperazin-1-yl)methyl)cyclopropanecarboxa-
mide (1): A 500 mL flask containing product 9 obtained above
(70.0 g, 12.8 mmol) was charged with 300 mL t-BuOH and 50.0 g
KOH. The resulting mixture was then placed into a preheated
90 °C bath. After stirring for 12 h, the mixture was diluted with
H2O, and extracted (5Â 10% MeOH/CH2Cl2). The organics were then
dried (MgSO4) and concentrated under reduced pressure. Purifica-
tion by flash chromatography (SiO2, 90% EtOAc /hexanes) gave
50.0 g of the product as white solid (11.1 mmol, 87%, 88% ee).
The white solid was then dissolved in 800 mL boiling CH3CN and
the resulting solution was allowed to cool overnight in an open
flask. The crystals which formed overnight were filtered. The fil-
trate was then concentrated under reduced pressure to give
46.0 g of the final product as a white solid (10.2 mmol, 79% overall
yield, 94% ee): mp 169–172 °C; 1H NMR (DMSO-d6, 400 MHz) d
8.02 (s, 1H), 7.83 (m, 4H), 7.00 (s, 1H), 6.86 (s, 1H), 4.02 (m, 1H),
3.59 (m, 1H), 3.17 (t, J = 12.0 Hz, 1H), 2.90 (d, J = 11.0 Hz, 1H),
2.76 (d, J = 11.5 Hz, 1H), 2.42 (d, J = 13.0 Hz, 1H), 2.23 (d,
J = 13.0 Hz, 1H), 1.93 (dd, J = 3.5, 11.5 Hz, 1H), 1.79 (m, 1H), 1.73
(s, 3H), 1.03 (d, J = 6.5 Hz, 3H), 0.96 (d, J = 3.0 Hz, 2H), 0.43 (d,
J = 3.0 Hz, 2H); 13C NMR (100 MHz, DMSO-d6) d 174.3, 144.4,
140.2, 127.6, 126.5, 73.4, 73.1, 61.5, 56.7, 51.7, 48.8, 22.6,19.7,
15.1, 13.4, 13.0; IR (neat, cmÀ1) 3448, 3138, 1647, 1589, 1421,
1338, 1281, 1169, 1150, 1094, 981, 761, 713; MS (ESI): m/z = 450
(R)-1-((4-(4-Acetylphenylsulfonyl)-3-methylpiperazin-1-yl)-
methyl)cyclopropanecarbonitrile (7): The crude amine 6 ob-
tained above was dissolved in 1.5 L CH2Cl2 followed by the
addition of 202 g triethylamine (2.0 mol) at 0 °C. Two hundred
grams of 4-acetylbenzenesulfonyl chloride was then added to
the reaction mixture in portions. After the reaction appeared
complete as monitored via TLC, the resulting mixture was di-
luted with sat. NaHCO3 and stirred for an additional 2 h. The or-
ganic phase was separated, and the aqueous solution was
extracted (4Â 10% MeOH/CH2Cl2). The organics were then com-
bined, dried (MgSO4), and concentrated under reduced pressure.
The crude product was passed through a plug of SiO2 (50% EA/
hexanes), followed by purification of the resulting white solid
by crystallization (1.2 L 50% EtOAc/hexanes) to give 200 g of
the product 7 as a white solid (55.3 mol, 36% for three steps):
1H NMR (DMSO-d6, 500 MHz) d 8.14 (d, J = 8.0 Hz, 2H), 7.95 (d,
J = 8.5 Hz, 2H), 4.05 (m, 1H), 3.63 (m, 1H), 3.18 (ddd, J = 12.5,
6.0, 3.0 Hz, 1H), 2.85 (d, J = 11.5 Hz, 1H), 2.74 (d, J = 11.5 Hz,
1H), 2.65 (s, 3H), 2.46 (d, J = 13.0 Hz, 1H), 2.21 (d, J = 13.0 Hz,
1H), 2.04 (dd, J = 11.0, 3.5 Hz, 1H), 1.94 (ddd, J = 11.5, 5.8,
3.5 Hz, 1H), 1.20 (m, 2H). 1.10 (dd, J = 15.0, 7.5 Hz, 3H), 0.85
(m, 2H); MS (ESI): m/z = 362.2 [M+H]+.
1-(((R)-3-Methyl-4-(4-(1,1,1-trifluoro-2-hydroxypropan-2-yl)-
phenylsulfonyl)piperazin-1-yl)methyl)cyclopropanecarbonitrile (8):
To a 2 L flask containing a portion of ketone 7 obtained above
(165 g, 45.6 mmol, 1.0 equiv) in 900 mL THF was added 97.2 g
TMS-CF3 (68.4 mmol, 1.5 equiv) at 0 °C. The solution was allowed
to stir for 30 min. Tetrabutylammonium fluoride (684 mL, 1.0 M
in THF, 68.4 mmol, 1.5 equiv) was added dropwise via addition
funnel over a period of 3 h. After the addition was complete, the
solution was allowed to stir for an additional 30 min and warm
to room temperature, followed by the addition of sat. NaHCO3,
and the resulting mixture was allowed to stir overnight. The or-
ganic phase was separated and the aqueous solution was ex-
tracted (4Â 10% MeOH/CH2Cl2). The organics were combined,
dried (MgSO4), and concentrated under reduced pressure. Purifi-
cation by flash chromatography (SiO2, 50% EtOAc/hexanes) gave
165 g of the product 8 as white solid (38.2 mol, 84%): 1H NMR
(DMSO-d6, 500 MHz) d 7.84 (m, 4 H), 6.86 (s, 1H), 4.01 (m, 1H),
3.58 (m, 1H), 3.16 (ddd, J = 13.0, 6.5, 3.0 Hz, 1H), 2.85 (d,
J = 11.0 Hz, 1H), 2.73 (d, J = 11.0 Hz, 1H), 2.45 (d, J = 13.0 Hz, 1H),
2.21 (d, J = 13.0 Hz, 1H), 2.05 (dd, J = 11.5, 3.5 Hz, 1H), 1.95 (ddd,
J = 11.5, 5.8, 3.5 Hz, 1H), 1.73 (s, 3H), 1.20 (m, 2H). 1.09 (d,
J = 7.0 Hz, 3H), 0.86 (m, 2H); MS (ESI): m/z = 432.1 [M+H]+.
(S)-2-(4-((R)-4-((1-Cyanocyclopropyl)methyl)-2-methylpipera-
zin-1-ylsulfonyl)phenyl)-1,1,1-trifluoropropan-2-yl (S)-1-hydrox-
ybutan-2-ylcarbamate (9): To a 2 L flask containing product 8
obtained above (160 g, 37.1 mmol, 1.0 equiv) in 465 mL CH3CN
[M+H]+; Anal. calcd for C19H26F3N3O4S: C, 50.77; H, 5.83; N, 9.35.
25
Found: C, 50.52; H, 5.88; N, 9.14; [a] D-38° (c 0.50, CH2Cl2).
4.2. Synthesis of (3-(1-(4-fluorophenyl)cyclopropyl)-4-
isopropyl-5-(4-(trifluoromethoxy)phenyl)-4H-1,2,4-triazole) (2)
To
a mixture of 1-(4-fluorophenyl)acetonitrile 10 (20.3 g,
150 mmol), 1-bromo-2-chloroethane (25 mL, 300 mmol) and ben-
zyltriethylammonium chloride (683 mg, 3.00 mmol) was added
50% aqueous NaOH (84 g, 1.05 mol), and the resulting mixture
was heated at 50 °C overnight. After cooling, the mixture was
poured into water and extracted with diisopropyl ether. The organ-
ic layer was washed sequentially with water, 1 N aqueous HCl, and
brine, and dried over MgSO4. Filtration, concentration in vacuo and
purification by silica gel flash chromatography (n-hexane/EtOAc
6:1) gave 16.4 g (68%) of 11 as a yellow oil: 1H NMR (CDCl3,
400 MHz) d 7.25–7.33 (m, 2H), 7.01–7.10 (m, 2H), 1.70–1.76 (m,
2H), 1.35–1.41 (m, 2H).
To a solution of 11 (16.4 g, 102 mmol) in acetone (140 mL) was
added 4 N aqueous NaOH (100 mL) at room temperature. 30% H2O2
(150 mL) was added dropwise to the solution with cooling in an
ice-water bath. The mixture was allowed to stand at room temper-
ature and stirred for an additional 2 h. The reaction mixture was
cooled in an ice-water bath, and aqueous Na2SO3 (10% in water,
159 mmol) was added to the mixture. The solvent was removed

8930
H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
by evaporation in vacuo, and the precipitated solid was collected
by filtration and washed with water and n-hexane to give 17.0 g
(93%) of 12 as a white solid.
lovirus protein expression system. Cells infected with the 11b-
HSD1-expressing baculovirus were extracted with lysis buffer
(50 mM Tris, pH 7.5, 150 mM NaCl, 0.5% Triton). FLAG-tagged
11b-HSD1 was captured by anti-FLAG tagged beads (Sigma), eluted
with a FLAG peptide (Sigma), and used as the enzyme source in the
biochemical assays for potency determination and kinetic analysis.
To determine the inhibitory activity and IC50 values (the inhib-
itor concentrations giving 50% inhibition) in the reductase assay,
A mixture of 12 (17.0 g, 94.8 mmol) in 6 N aqueous HCl (95 mL)
and 1,4-dioxane (150 mL) was heated at reflux temperature over-
night. The solvent was removed by evaporation in vacuo, and the
residue extracted with EtOAc. The organic layer was washed with
brine and dried over MgSO4. Filtration and concentration in vacuo
gave 16.8 g (98%) of the product 13 as a white solid: 1H NMR
(DMSO-d6, 400 MHz) d 12.3 (broad s, 1 H), 7.30–7.38 (m, 2H),
7.05–7.14 (m, 2H), 1.42–.47 (m, 2H), 1.10–1.15 (m, 2H); MS
(ESI): m/z = 179.1 [MÀH]À.
inhibitors of various concentrations were incubated with 100 ll/
well PBS buffer containing 10 ng of purified human 11b-HSD1,
0.1% BSA (Sigma), 16 nM [3H]-cortisone (GE Healthcare, 1.8TBg/
mol, 49 Ci/mmol) and 500
at 37 °C for 90 min. At the end of reaction incubation, 40
buffer containing 175 M 11b-HSD1 inhibitor carbenoxolone (Sig-
ma), 0.16 g of anti-cortisol monoclonal antibody (East Coast Biol-
ogics) and 365 g of SPA PVT anti-mouse antibody-binding beads
l
M NADPH (Sigma) in 96-well plates
A mixture of 13 (11.8 g, 65.5 mmol) and concentrated H2SO4
(1.5 mL) in MeOH (100 mL) was heated at reflux temperature for
8 h. The solvent was removed by evaporation in vacuo. The residue
was diluted with water and extracted with EtOAc. The organics
were washed sequentially with sat. aqueous NaHCO3, water, and
brine, and dried over MgSO4. Filtration and concentration in vacuo
gave 12.7 g (quantitative) of 14 as yellow oil.
ll PBS
l
l
l
(GE Healthcare) were added to each well to stop the enzyme reac-
tion and capture [3H]-cortisol to the SPA beads. The reaction prod-
uct [3H]-cortisol in each well was determined with a TopCount
microscintillation plate reader (Packard Instrument). Percent inhi-
bition was calculated based on references set with no compound as
A mixture of 14 (12.7 g, 65.5 mmol) and hydrazine hydrate
(12.7 mL, 328 mmol) in EtOH (40 mL) was heated at reflux temper-
ature overnight. The solvent was removed by evaporation in vacuo,
and the residue was crystallized from water and collected by filtra-
tion to give 9.64 g (76%) of 15 as a pale yellow solid: 1H NMR
(DMSO-d6, 300 MHz) d 8.10 (s, 1H), 7.30–7.38 (m, 2H), 7.08–7.17
(m, 2H), 4.16 (s, 2H), 1.28–1.35 (m, 2H), 0.90–0.97 (m, 2H).
To a solution of isopropylamine (6.83 mL, 80.2 mmol) in chloro-
form (15 mL) was added 4-trifluoromethoxybenzoyl chloride 16
(3.60 g, 16.0 mmol) over 5 min with cooling in an ice-water bath.
The reaction mixture was stirred for 30 min and concentrated in
vacuo. The residue was washed in sat. aqueous NaHCO3 and col-
lected by filtration to give 3.69 g (93%) of 17 as a white solid.
A mixture of 17 (3.69 g, 14.9 mmol) and Lawesson’s reagent
(4.83 g, 11.9 mmol) in THF (40 mL) was heated at reflux tempera-
ture for 1 h. After cooling, the mixture was diluted with EtOAc and
the insoluble material was removed by filtration. The filtrate was
washed with sat. aqueous NaHCO3 and brine, and dried over
MgSO4. Filtration, concentration in vacuo, and purification by silica
gel chromatography (n-hexane/EtOAc 5:2) gave 4.00 g (quantita-
tive yield) of 18 as a yellow solid.
0% inhibition and 100 lM carbenoxolone as 100%. IC50 values were
determined using curve fitting algorithm Prism (GraphPad) based
on percent inhibition and the corresponding inhibitor
concentrations.
To determine the inhibitory activity in the dehydrogenase as-
say, a range of concentrations of each inhibitor were added to wells
containing 100
0.1% BSA and 100
and 0.25
14C]-cortisol (Perkin-Elmer Life Sciences, 1.67–
2.22 GBg/mmol, 45–60 mCi/mmol). After 1-h incubation at
ll of PBS buffer containing 10 ng human 11b-HSD1,
l
M NADP, 2.25 M unlabeled cortisol (Sigma)
l
lM [
a
37 °C, the reaction mix was extracted with ethyl acetate. [14C]-Cor-
tisone and [14C]-cortisol in the organic phase were separated by
thin layer chromatography, and were quantified with a phos-
phor-imager system (Fujifilm). Percent inhibition was calculated
based on references set with no compound as 0% inhibition and
100 lM carbenoxolone as 100%.
4.4. Kinetic analysis of NADPH and substrate competition
To a solution of 18 (4.00 g, 15.2 mmol) in THF (40 mL) was
added methyl iodide (3.78 mL, 60.7 mmol). The solution was stir-
red for 60 h at room temperature and concentrated in vacuo. The
residue was triturated in EtOAc and collected by filtration to give
4.67 g (75%) of 19 as a solid.
To determine the mechanism of kinetic action using NADPH
concentration as a variable, various concentrations of NADPH were
incubated in each well of a 96-well plate with 100 ll of PBS buffer
containing 10 ng human 11b-HSD1, 0.1% BSA, and 16 nM of [3H]-
cortisone with or without inhibitors in the aforementioned 96-well
assay. After 90-min incubation at 37 °C, [3H]-cortisol was captured
by the anti-cortisol monoclonal antibody and quantified with the
scintillation proximity assay. Lineweaver–Burk double-reciprocal
plots were generated from the range of NADPH concentrations
and the corresponding reaction velocity of cortisol formation to
A mixture of 15 (1.00 g, 5.15 mmol) and 19 (2.50 g, 6.17 mmol)
in pyridine (10 mL) was heated at reflux temperature for 5 h. The
solvent was removed by evaporation in vacuo, and the residue ex-
tracted with EtOAc. The organic layer was washed with sat. aque-
ous NaHCO3 and brine, and dried over MgSO4. Filtration,
concentration in vacuo and purification by silica gel chromatogra-
phy (CHCl3/acetone 5/1:1/1) gave 745 mg (35%) of product 2 as a
white solid: mp 144–146 °C;1H NMR (DMSO-d6, 400 MHz) d 7.65
(m, 2H), 7.51 (d, J = 7.8 Hz, 2H), 7.22–7.12 (m, 4H), 4.52 (m, 1H),
1.57 (m, 2H), 1.46 (m, 2H), 0.97 (d, J = 6.6 Hz, 6H); 13C NMR
(100 MHz, DMSO-d6) d 161.9, 159.6, 156.2, 149.3, 137.6, 132.5,
129.1, 128.2, 128.1, 120.7, 115.5, 115.2, 47.8, 21.7, 20.2, 15.3; IR
(neat, cmÀ1) 3001, 2362, 2334, 1655, 1510, 1473, 1420, 1258,
1222, 1165, 830, 810; MS (ESI): m/z = 406 [M+H]+; Anal. calcd for
C21H19F4N3O: C, 62.22; H, 4.72; N, 10.37. Found: C, 62.19; H,
4.70; N, 10.42.
produce apparent Km and Vmax
.
Kinetic studies examining the dehydrogenase activity of 11b-
HSD1 were performed by incubating various concentrations of
[
14C]-labeled cortisol with 10 ng human 11b-HSD1, 0.1% BSA and
100 lM NADP, with or without inhibitors in 100 ll/well of PBS buf-
fer. After 1-h incubation at 37 °C, the reaction mix was extracted
with ethyl acetate. [14C]-Cortisone and [14C]-cortisol in the organic
phase were separated by thin layer chromatography, and were
quantified with a phosphor-imager system (Fujifilm). Apparent
Km and Vmax values were derived from Lineweaver–Burk double-
reciprocal plot analysis.
4.3. Biochemical assays
4.5. Protein preparation and crystallization
Full-length human 11b-HSD1 was cloned, tagged with a FLAG
sequence at the C-terminus, and expressed in Hi5 cells via a bacu-
The protein preparation was carried out according to a previously
reported protocol.20 In brief, the truncated form of human 11b-HSD1

H. Tu et al. / Bioorg. Med. Chem. 16 (2008) 8922–8931
8931
5. Andrews, R. C.; Rooyackers, O.; Walker, B. R. J. Clin. Endocrinol. Metab. 2003, 88,
285–291.
6. Alberts, P.; Nilsson, C.; Selen, G.; Engblom, L. O.; Edling, N. H.; Norling, S.;
Klingstrom, G.; Larsson, C.; Forsgren, M.; Ashkzari, M.; Nilsson, C. E.; Fiedler,
M.; Bergqvist, E.; Ohman, B.; Bjorkstrand, E.; Abrahmsen, L. B. Endocrinology
2003, 144, 4755–4762.
7. 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. J. Exp. Med. 2005, 202, 517–527.
8. Tomlinson, J. W. Minerva Endocrinol. 2005, 30, 37–46.
(residue 24–292) with a single mutation C272S was cloned into a
pBAD-ThioE vector (Invitrogen) by PCR. The recombinant protein
with a N-terminal affinity tag of MKHQHQHQHQHQHQQPL was ex-
pressed in E. coli Rossetta1(DE3) cells (Novagen), grown in 2YT med-
ium at 30 °C and induced by 0.2% arabinose and 0.25 mM
corticosterone overnight at 20 °C before harvesting. The protein
was purified by a Ni+–NTA column (Qiagen), and eluted with
200 mM imidazole after a wash with 40 mM imidazole. The Ni+ elu-
ant was then applied to a superdex 200 column (Pharmacia). The
protein eluted as a dimer in a buffer of 250 mM NaCl, 25 mM Tris–
HCl, pH 7.9, and 4 mM CHAPS and was then concentrated to
30 mg/mL for crystallization. The 11b-HSD1 crystals were grown
9. Stulnig, T. M.; Waldhausl, W. Diabetologia 2004, 47, 1–11.
10. Walker, B. R.; Seckl, J. R. Expert Opin. Therap. Targets 2003, 7, 771–783.
11. Oppermann, U. Endocr. Metab. Immune Disord. Drug Targets 2006, 6, 259–269.
12. Olson, S.; Aster, S. D.; Brown, K.; Carbin, L.; Graham, D. W.; Hermanowski-
Vosatka, A.; LeGrand, C. B.; Mundt, S. S.; Robbins, M. A.; Schaeffer, J. M.;
Slossberg, L. H.; Szymonifka, M. J.; Thieringer, R.; Wright, S. D.; Balkovec, J. M.
Bioorg. Med. Chem. Lett. 2005, 15, 4359–4362.
13. Richards, S.; Sorensen, B.; Jae, H.-s. J.; Winn, M.; Chen, Y.; Wang, J.; Fung, S.;
Monzon, K.; Frevert, E. U.; Jacobson, P.; Sham, H.; Link, J. T. Bioorg. Med. Chem.
Lett. 2006, 16, 6241–6245.
14. Gu, X.; Dragovic, J.; Koo, G. C.; Koprak, S. L.; LeGrand, C.; Mundt, S. S.; Shah, K.;
Springer, M. S.; Tan, E. Y.; Thieringer, R.; Hermanowski-Vosatka, A.; Zokian, H.
J.; Balkovec, J. M.; Waddell, S. T. Bioorg. Med. Chem. Lett. 2005, 15, 5266–5269.
15. Rohde, J.; Pliushchev, M.; Sorensen, B.; Wodka, D.; Shuai, Q.; Wang, J.; Fung, S.;
Monzon, K.; Chiou, W.; Pan, L.; Deng, X.; Chovan, L.; Ramaiya, A.; Mullally, M.;
Henry, R.; Stolarik, D.; Imade, H.; Marsh, K.; Beno, D.; Fey, T.; Droz, B.; Brune,
M.; Camp, H.; Sham, H.; Frevert, E.; Jacobson, P.; Link, J. J. Med. Chem. 2007, 50,
149–164.
at 16 °C in a sitting drop with 3 ll of the protein solution and 3 ll
of the well solution containing 16-18% (w/v) PEG3350 and 0.1 M
MES (pH 6.2–6.4). The inhibitors were introduced into the crystals
by a soaking method. The soaked crystals were transferred into the
mother liquor with additional 20% ethylene glycol and flash frozen
using liquid nitrogen.
4.6. Data collection, structure determination, and refinement
The X-ray diffraction data sets were collected on a RU-H3RHB
generator/Raxis-IV++ detector (Rigaku) to a resolution of 2.5 Å for
compound 1 co-crystal, and on a synchrotron radiation beamline
(5.0.2) at the Advanced Light Source (ALS) in Berkeley, California,
to a resolution of 2.55 Å for compound 2 co-crystal. The data were
integrated using MOSFLM27 and scaled in CCP4.28 The structures
were solved by molecular replacement with MOLREP29 using a
monomer of the published 11b-HSD1 structure (PDB code: 1XU9)
as a search model. Several rounds of model building were done
using Quanta (Accelrys). The program REFMAC in CCP4 was used
for structural refinement. The atomic coordinates and structure
factors have been deposited in the RCSB Protein Data Bank under
accession code 3D4N for compound 1 and 3D5Q for compound 2.
16. 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, 429–432.
17. Sun, D.; DeGraffenreid, M.; He, X.; Jaen, J. C.; Powers, J. P.; Yan, X.; Di, Y.; Tu, H.;
Ursu, S.; Ma, J.; Miao, S.; Tang, L.; Ye, Q.; Sudom, A.; Wang, Z. Discovery and
optimization of arylsulfonamides as a novel class of 11b-HSD1 inhibitors;
234th ACS National Meeting, 2007, Boston, MA, United States.
18. Julian, J. D.; Bostick, T.; Caille, S.; Chan, H.; Degraffenreid, M.; He, X.; Hungate,
R. W.; Jaen, J. C.; Jiang, B.; Kaizerman, J.; Liu, J.; McMinn, D.; Powers, J. P.; Rew,
Y.; Sudom, A.; Sun, D.; Tu, H.; Ursu, S.; Wang, Z.; Yan, X.; Ye, Q. Discovery and
biological evaluation of novel benzamide derivatives as potent 11b-HSD1
inhibitors for the treatment of type II diabetes; 234th ACS National Meeting,
2007, Boston, MA, USA.
19. Cardozo, M. G.; Powers, J. P.; Goto, H.; Harada, K.; Imamura, K.; Kakutani, M.;
Matsuda, I.; Ohe, Y.; Yata, S. PCT Int. Appl. 2005. pp. WO-05044192.
20. Hosfield, D. J.; Wu, Y.; Skene, R. J.; Hilgers, M.; Jennings, A.; Snell, G. P.;
Aertgeerts, K. J. Biol. Chem. 2005, 280, 4639–4648.
21. Ghosh, D.; Pletnev, V. Z.; Zhu, D. W.; Wawrzak, Z.; Duax, W. L.; Pangborn, W.;
Labrie, F.; Lin, S. X. Structure 1995, 3, 503–513.
Acknowledgments
We thank Donglin Guo for assistance in the construction and
purification of recombinant human 11b-HSD1 from a baculovirus
expression system. The Advanced Light Source (ALS) is supported
by the Director, Office of Science, Office of Basic Sciences, Materials
Sciences Division, of the U.S. Department of Energy under Contract
No. DE-AC03-76SF00098 at the Lawrence Berkeley National
Laboratory.
22. Nahoum, V.; Gangloff, A.; Legrand, P.; Zhu, D. W.; Cantin, L.; Zhorov, B. S.; Luu-
The, V.; Labrie, F.; Breton, R.; Lin, S. X. J. Biol. Chem. 2001, 276, 42091–42098.
23. Sawicki, M. W.; Erman, M.; Puranen, T.; Vihko, P.; Ghosh, D. Proc. Natl. Acad. Sci.
U.S.A. 1999, 96, 840–845.
24. Ogg, D.; Elleby, B.; Norstrom, C.; Stefansson, K.; Abrahmsen, L.; Oppermann, U.;
Svensson, S. J. Biol. Chem. 2005, 280, 3789–3794.
25. Zhang, J.; Osslund, T. D.; Plant, M. H.; Clogston, C. L.; Nybo, R. E.; Xiong, F.;
Delaney, J. M.; Jordan, S. R. Biochemistry 2005, 44, 6948–6957.
26. Draper, N.; Walker, E. A.; Bujalska, I. J.; Tomlinson, J. W.; Chalder, S. M.; Arlt,
W.; Lavery, G. G.; Bedendo, O.; Ray, D. W.; Laing, I.; Malunowicz, E.; White, P.
C.; Hewison, M.; Mason, P. J.; Connell, J. M.; Shackleton, C. H.; Stewart, P. M.
Nat. Genet. 2003, 34, 434–439.
References and notes
27. Leslie, A. G. W. Recent changes to the MOSFLM package for processing film and
image plate data; Joint CCP4 + ESF-EAMCB Newsletter on Protein
Crystallography, 1992.
28. Collaborative Computational Project, No 4. Acta Crystallogr. 1994, D50, 760–
763.
1. Mariniello, B.; Ronconi, V.; Rilli, S.; Bernante, P.; Boscaro, M.; Mantero, F.;
Giacchetti, G. Eur. J. Endocrinol. 2006, 155, 435–441.
2. Bujalska, I. J.; Kumar, S.; Stewart, P. M. Lancet 1997, 349, 1210–1213.
3. Seckl, J. R.; Walker, B. R. Endocrinology 2001, 142, 1371–1376.
4. Masuzaki, H.; Paterson, J.; Shinyama, H.; Morton, N. M.; Mullins, J. J.; Seckl, J. R.;
Flier, J. S. Science 2001, 294, 2166–2170.
29. Vagin, A.; Teplyakov, A. J. Appl. Crystallogr. 1997, 30, 1022–1025.