Design, synthesis, and SAR studies of novel polycyclic acids as potent and selective inhibitors of human 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD-1)

Article abstract of DOI:10.1016/j.bmcl.2011.09.055

Starting from high throughput screening hit 2-adamantyl acetic acid 3, a series of polycyclic acids have been designed and synthesized as novel, potent, and selective inhibitors of human 11β-HSD-1. Structure-activity relationships of two different regions

Full text of DOI:10.1016/j.bmcl.2011.09.055

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6699–6704
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, and SAR studies of novel polycyclic acids as potent
and selective inhibitors of human 11b-hydroxysteroid dehydrogenase type
1 (11b-HSD-1)
Xiang-Yang Ye a, , Stephanie Y. Chen a, Akbar Nayeem b, Rajasree Golla c, Ramakrishna Seethala c,
⇑
Mengmeng Wang d, , Timothy Harper d, Bogdan G. Sleczka e, Yi-Xin Li e, Bin He f, Mark Kirby f,
David A. Gordon f, Jeffrey A. Robl a,
⇑
a Department of Chemistry, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
b CADD, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
c Lead Evaluation, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
d MAP, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
e BDAS, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
f Department of Biology, Research & Development, Bristol-Myers Squibb, PO Box 5400, Princeton, NJ 08543-5400, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Starting from high throughput screening hit 2-adamantyl acetic acid 3, a series of polycyclic acids have
been designed and synthesized as novel, potent, and selective inhibitors of human 11b-HSD1. Structure–
activity relationships of two different regions of the chemotype (polycyclic ring and substituents on qua-
ternary carbon) are discussed.
Received 18 August 2011
Revised 13 September 2011
Accepted 15 September 2011
Available online 21 September 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
11b-Hydroxysteroid dehydrogenase type 1
Inhibitor
Polycyclic acids
According to the World Health Organization, the number of type
2 diabetes patients had reached 154 million in 2000, and is growing
rapidly with the projection of over 330 million diabetics worldwide
by 2030.1 This chronic progressive disease, characterized by eleva-
tion in plasma glycosylated hemoglobin (HbA1c), is driven in part
by insulin resistance exacerbated by abdominal obesity and insulin
deficiency due to progressive deterioration of pancreatic beta-cell
function. Metabolic syndrome, a cluster of metabolic abnormalities
(insulin resistance, obesity, dyslipidemia, hyperglycemia and
hypertension) is increasingly recognized as a major cause of cardio-
vascular diseases and type 2 diabetes. Selective inhibition of
11b-hydroxysteroid dehydrogenase type 1 (11b-HSD-1) represents
a promising approach for the potential treatment of type 2 diabetes
and metabolic syndrome and as such has received intensive atten-
tion from both academia and the pharmaceutical industry.2
11b-HSD-1 catalyzes the inter-conversion of inactive cortisone to
the active steroid cortisol, a principle circulating glucocorticoid in
humans. Excess glucocorticoids are implicated in increased glucose
output in the liver, dampened glucose-dependent insulin sensitiv-
ity in adipose tissue, and reduced insulin secretion from the
pancreas. The concept of selective inhibition of 11b-HSD-1 as a
viable strategy for treating type 2 diabetes and metabolic syndrome
is supported by genetic and pharmacological studies in various
animal models, as well as preliminary clinical studies in type 2 dia-
betic (T2D) patients. For example, 11b-HSD-1 knockout mice are
resistant to developing diet induced obesity and hyperglycemia,3
while transgenic mice that overexpress 11b-HSD-1 in adipose exhi-
bit a phenotype characterized by a variety of metabolic disorders,
including obesity, insulin resistance, glucose intolerance, and
hyperglycemia.4 Several papers have reported that adipose tissue
from obese individuals exhibits elevated 11b-HSD-1 activity.
Chronic administration of inhibitors in animal models of diabetes
has demonstrated positive effects on glucose, body weight, and
plasma lipids.5 Furthermore, Incyte has recently demonstrated that
the selective 11b-HSD-1 inhibitor INCB-13739 improves hepatic
and peripheral insulin sensitivity and lipid profiles in T2D patients
after 1À3 months of treatment.6
⇑
Corresponding authors. Tel.: +1 609 818 4130; fax: +1 609 818 3550 (X.-Y. Ye),
tel.: +1 609 818 5048; fax: +1 609 818 3550 (J.A. Robl).
E-mail addresses: xiang-yang.ye@bms.com (X.-Y. Ye), Jeffrey.Robl@bms.com
(J.A. Robl).
Over the past few years, several classes of non-steroidal 11b-
HSD-1 inhibitors have been reported in the literature.7,8 These in-
clude sulfonamides and sulfones, amides and lactams, triazoles,
Current address: PDM, Pfizer Inc., 1 Burtt Rd., Andover, MA 01810, USA.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.055

6700
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6699–6704
H
N
S
H
N
N
N
HO
O
N
N
N
O
CF3
2
(IC50 = 120 nM)
1a (h-HSD1 Ki = 8 nM)
CO2H
H
N
O
N
Br
N
H
HO
O
O
1c
1b (h-HSD1 IC50 = 1.1 nM)
Figure 1. 2-Adamantamine and 1-adamantyl carboxylic acid based 11b-HSD-1 inhibitors.
thiazolones, and others inhibitors. Within the amide/lactam cate-
gory, researchers from Abbott and several other organizations
have incorporated 2-adamantamine into their chemotypes, as
exemplified by compounds 1a–c.9a Researchers from Novo Nordisk
have disclosed amides derived from 1-adamantyl carboxylic acid,
as exemplified by compound 2 (Fig. 1).9b In both cases, the amido
functionalities are presumed to interact with the catalytic Ser170
and Tyr183 residues in the active site of the enzyme. In this Letter,
we wish to report the discovery of novel adamantyl/polycyclic
carboxylic acids as potent and selective inhibitors of human
11b-HSD-1. While several teams have reported inhibitors which
contain a carboxylic acid group (e.g., compounds 1b and 1c),10
other functional moieties present in the molecule (ureas, amides,
etc.) are presumed to serve as the hydrogen-bonding pharmaco-
phore with the Ser170 and Tyr183 residues. To the best of our
knowledge, the compounds represented by structures 4, 5 and 6
are the first in which a carboxylic acid functionality is proposed
to serve as the primary pharmacophore to the enzyme’s catalytic
binding site.11
itory activity against human 11b-HSD-1 (IC50 = 860 nM). Molecular
docking12 of 3 using a published crystal structure of human
11b-HSD-113 showed a key hydrogen bond interaction of the car-
boxylic group of the ligand with the catalytic residues Ser170 and
Tyr183 on the enzyme. The adamantyl moiety positions well in the
enzyme’s hydrophobic pocket (Fig. 3). Based on this docking mod-
el, we speculated that additional of a lipophilic substituent (e.g.,
aryl) at the C-2 position of adamantyl should further enhance
the potency by more optimally filling the proximal hydrophobic
binding site. This was supported by the identification of the highly
potent C-2 4-F-phenyl derivative 4a in the screening deck. The
finding demonstrated that the hydrophobic groups at C-2 position
were not only well tolerated, but dramatically enhanced potency.
In contrast, we found that introducing a polar substituent (such
as –CN in 4b) in the C-2 position significantly decreased inhibitory
potency.
Using 4a as a starting point for our SAR program, we subse-
quently varied the linker length between the carboxylate group
and the adamantane ring to find that the acetic acid based analog
(n = 1) was optimal (compare compound 4a with 5a and 5b). These
initial findings set the stage for subsequent optimization studies in
Our initial lead, 2-adamantyl acetic acid 3 (Fig. 2), was identi-
fied via high throughput screening and exhibited moderate inhib-
optimization
of linker
HO
HO
2
H
1
2
R
1
O
O
4a (R = 4-F-phenyl, IC50 =12 nM)
4b (R = CN, IC50 = 8950 nM)
3 (IC50 = 860 nM)
other bicyclic
ring systems
HO
HO
2
n
1
n = 1 is optimal
O
O
R
6
F
= bicylic ring systems
5a (n =0, IC50 =156 nM)
4a (n =1, IC50 = 12 nM)
5b (n = 2, IC50 = 234 nM)
Figure 2. HTS hit and the evolution of bicyclic carboxylic acids as novel 11b-HSD-1 inhibitors.

X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6699–6704
6701
Table 1
this series. In addition to variation of the adamantyl group, we also
SAR of polycyclic carboxylic acids18
examined other bicyclic ring systems and substituents on the rings
with the aim to enhance metabolic stability and in vitro potency of
the series. As a result, we synthesized a series of bicyclic carboxylic
acids with the general structure shown as 6 in Figure 2.
HO
We developed a general route to access polycyclic acids 6 as de-
scribed in Scheme 1. Knoevenagel condensation of polycyclic ke-
O
tones
7 (commercially available or obtained via literature
procedures) and Meldrum’s acid was carried out in anhydrous pyr-
idine in the presence of catalytic piperidine to afford the disubsti-
tuted alkylidene 8.14 1,4-Conjugate addition of nucleophiles such
alkyl or aryl Grignard (RMgX) or other organometallic (RLi) re-
agents to 8 was carried out in the presence of Cu+ salts (such as
CuBr, CuCN, etc.) at room temperature or at 40 °C to afford 9.15
The isopropylidene malonate 9 was converted to 6 in generally
high yields via hydrolysis and subsequent decarboxylation in wet
organic solvent at 110 °C.16 This general scheme allowed variation
of the R group as well as the polycyclic ring system for rapid SAR
elucidation.
F
6
% Remainingb
Hu Mu
a
IC50 (nM)
Hu
Compd
Mu
4a
12
120
78
48
Overall, compound 4a represented an excellent starting point
for the chemical program. Based on the presence of the polar car-
boxylic acid, the lead exhibited favorable in vitro properties that
have been typically problematic with known/reported inhibitors.
In addition to good intrinsic potency versus human 11b-HSD-1
6a
6b
6c
694
193
63
9545
2621
2381
NTc
83
NT
81
100
100
OMe
6d
594
153
>10K
1016
NT
87
NT
50
6e
6f
3
204
174
184
3954
6849
100
NT
73
NT
O
6g
6h
OH
NT
NT
Figure 3. Docking model of compounds 3 (blue) and 4a (green) with human
11b-HSD-1.12 The enzyme’s hydrophobic pocket is shown in green wire mesh, while
the hydrophilic pocket is shown in blue wire mesh.
OH
6i
6j
127
354
NT
NT
NT
b
O
a
O
O
O
O
O
2203
6508
100
7
8
O
O
O
HO
c
6k
6l
13
21
1616
4570
87
94
F
O
R
R
9
O
6
R = aryl or alkyl
100
100
OMe
Scheme 1. Synthesis of bicyclic carboxylic acids 6. Reagents and conditions: (a)
Meldrum’s acid, anhydrous pyridine, cat. piperidine, room temperature, 5 days
(50À90%); (b) RLi or RMgX, CuBr, THF, À20 °C to room temperature (or to 40 °C), 6 h
(30À85%); (c) NaCl (saturated, aqueous), DMSO, 110 °C, 12 h (ꢀ90%).
(continued on next page)

6702
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6699–6704
microsomes. Finally, additional attempts to incorporate a polar hy-
Table 1 (continued)
droxyl (6m–6q) directly to the adamantyl ring, while routinely
enhancing metabolic stability, often led to a diminution in potency
versus the human enzyme and a near abolishment of potency ver-
sus the mouse.
a
IC50 (nM)
% Remainingb
Compd
Hu
Mu
Hu
Mu
Table 2 highlights a limited SAR set directed towards variation
of the hydrophobic R group in structure 6. Not surprisingly, when
R is unsubstituted phenyl (6x), the inhibitory activities against
both human and mouse enzymes decrease slightly compared to
its 4-fluoro phenyl analog 4a. In general, substituents (whether po-
6m
6n
6o
6p
4705
189
425
160
NT
100
100
NT
89
100
NT
OH
>10K
>10K
776
Table 2
OH
SAR of adamantyl carboxylic acids18
OH
OH
HO
O
R
NT
NT
6
% Remainingb
a
Compd
R
IC50 (nM)
Hu
Mu
Hu
Mu
6q
48
4446
100
100
F
4a
6r
12
7
120
78
48
OH
a
IC50 refers to biochemical assay versus the human or mouse 11b-HSD-1
enzyme.
b
Represents percentage of compound remaining after incubation in human or
mouse microsomes after a 10-minute incubation period.
334
83
47
c
F
NT = not tested.
OMe
Cl
6s
6t
16
5
134
103
105
NT
NT
NT
NT
(IC50 = 12 nM) and selectivity versus human 11b-HSD-2 (<50%
inhibition at 10 lM), compound 4a also exhibited reasonable aque-
ous solubility, good permeability, weak CYP450 inhibition (1A2,
2C9, 2C19, 2D6, 3A4) and hERG inhibition, as well as negligible
NT
NT
PXR transactivation (EC50 > 50 lM). Unfortunately, microsomal
6u
4
stability (assessed using isolated liver microsomes) was marginal
and mouse enzymatic inhibitory activity was weak, preventing
evaluation of the lead in our in vivo pharmacodynamic and efficacy
models. Thus, attention was primarily directed towards optimiza-
tion of potency and metabolic stability in this chemotype.
6v
4
111
216
NT
NT
Cl
The known propensity of the adamantyl ring system to undergo
microsomal oxidation prompted us to first examine the effect of
different polycyclic ring systems on potency and metabolic stabil-
ity. Table 1 summarizes the inhibitory activity of these compounds
tested against both human and mouse 11b-HSD-1. In vitro stabil-
ity in the presence of human and mouse liver microsomes is re-
ported when available. The simple cyclohexyl derivative 6a
exhibits only moderate activity against human 11b-HSD-1, while
the carboxylic acids bearing more sophisticated polycyclic rings
such as [2,2,1]-bicycloheptane and [3,2,1]-bicyclooctane (6b and
6c) enhance potency 3À10-fold, highlighting the critical need for
hydrophobic space-filling at this position of the molecule. Notably,
the inhibitor bearing a [3,3,0]-bicyclooctane ring motif (6f) was
the most potent among this set of compounds, exhibiting single
digit nanomolar activity versus the human enzyme, but only mod-
est potency versus the mouse.17 Metabolic stability was also
enhanced relative to 4a. Attempts to modify 6f by either incorpo-
rating a polar hydroxyl functional group (6h and 6i) or building an
oxygen into the ring as an ether linker (6g and 6j) resulted in a
substantial loss of potency. In order to directly block putative met-
abolic oxidation of the adamantyl ring, we also incorporated a
fluorine or methoxy group at the bridgehead position (6k and
6l). Both compounds maintained in vitro potency for the human
enzyme while exhibiting enhanced metabolic stability in liver
6w19
12
87
100
CN
6x
6y
20
24
172
113
NT
NT
100
100
Cl
6z
4.9
155
66
7.4
Cl
6aa
6bb
14
82
54
55
98
43
5
2.2
NT = not tested.
a
IC50 refers to biochemical assay in human or mouse 11b-HSD-1 enzyme.
Represents percentage of compound remaining after incubation in human or
mouse microsomes after a 10-minute incubation.
b

X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6699–6704
6703
Table 3
of CYP450 enzymes, no PXR activation and weak ion channel
activities.22 Unfortunately, the propensity of these compounds to
undergo glucuronidation represented a significant hurdle for phar-
macokinetic optimization of this chemotype. As a result, additional
efforts were subsequently directed to the identification of suitable
replacements for the carboxylic acid functionality, including tetra-
zoles, sulfonamides and amide based functionalities. These efforts
will be the topic of future disclosures.
Coarse mouse pharmacokinetic studies and in vitro glucuronidation half-lives of
representative bicyclic acids
AUCa
(0–8 h) (
UGT T1/2
b
Compound
Cmax
(lM)
lMÃh)
(human/mouse) (min)
4a
6y
6w
6k
6l
6o
6f
6q
0.48
1.4
2
4.1
0.3
0.22
<LLQ
<LLQ
0.3
NT
1.9/2.1
4.2/0.8
NT
8.7/2.2
35/8.8
NT
0.07
0.04
<LLQ
<LLQ
0.06
0.04
Supplementary data
0.244
11/2.8
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.09.055.
NT = not tested.
a
Pharmacokinetic studies were performed in mice using 0.5% methocel, and 0.1%
mole/kg (30 mole/kg for 4a).
tween 80 in water as vehicle. The dose used was 10
l
l
Plasma concentration was measured at 1, 4, and 8 h time points. AUC (0–8 h) was
References and notes
calculated based on the concentrations measured at these time points.
b
Glucuronidation half-lives were measured by incubating substrate (10 lM)
1. Wild, S.; Roglic, G.; Green, A.; Sicree, R.; King, H. Diabetes Care 2004, 27, 1047.
2. (a) Tomlinson, J. W.; Stewart, P. M. Nat. Clin. Pract. Endocrinol. Metab. 2005, 1,
92; (b) Wang, M. Curr. Opin. Investig. Drugs 2006, 7, 319; (c) Saiah, E. Curr. Med.
Chem. 2008, 15, 642; (d) Schnackenberg, C. G. Curr. Opin. Investig. Drugs 2008, 9,
295; (e) Stewart, P. M.; Tomlinson, J. W. Diabetes 2009, 58, 14; (f) Boyle, C. D.;
Kowalski, T. J. Exp. Opin. Ther. Patents 2009, 19, 801; (g) Ge, R.; Huang, Y.; Liang,
G.; Li, X. Curr. Med. Chem. 2010, 17, 412. and references cited therein.
3. (a) Kotelevtsev, Y.; Holmes, M. C.; Burchell, A.; Houston, P. M.; Schmoll, D.;
Jamieson, P.; Best, R.; Brown, R.; Edwards, C. R. W.; Seckl, J. R.; Mullins, J. J.
Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 14924; (b) Morton, N. M.; Holmes, M. C.;
Fievet, C.; Staels, B.; Tailleux, A.; Mullins, J. J.; Seckl, J. R. J. Biol. Chem. 2001,
276, 41293.
4. (a) Masuzaki, H.; Paterson, J.; Shinyama, H.; Morton, N. M.; Mullins, J. J.; Seckl, J.
R.; Flier, J. S. Science 2001, 294, 2166; (b) 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. J. Clin. Invest. 2003, 112, 83.
5. (a) Wang, S. J.; Birtles, S.; de Schoolmeester, J.; Swales, J.; Moody, G.; Hislop, D.;
O’Dowd, J.; Smith, D. M.; Turnbull, A. V.; Arch, J. R. Diabetologia 2006, 49, 1333;
(b) Joanisse, D. R.; Thieringer, R.; Deshaies, Y. Endocrinology 2007, 148, 2391; (c)
Walker, B. R.; Connacher, A. A.; Lindsay, R. M.; Webb, D. J.; Edwards, C. R. J. Clin.
Endocrinol. Metab. 1995, 80, 3155; (d) 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; (e) Schnackenberg, C. G.; Doe, C. P.; Costell, M. H.; Turner, A. R.;
Cui, J.; Bentley, R.; Maniscalco, K.; Eybye, M. E.; Willette, R. N.; Yue, T.
Hypertension 2007, 50, 87; (f) Andrews, R. C.; Rooyackers, O.; Walker, B. R. J. Clin.
Endocrinol. Metab. 2003, 88, 285.
with either human or mouse liver microsomes (1 mg/mL) in the presence of UDGPA
(5 mM) over a 0, 5, 15, and 30 min period. The incubation mixture was quenched at
each time point and analyzed by LC–MS to obtain the percentage of substrate
remaining for each time point. UGT T1/2 was obtained from the curve plotted from
the above data.
lar or non-polar group) in either the 3-position or 4-position of
phenyl ring appeared to be neutral or slightly increase potency
(6r–6v). The inhibitory activity of benzyl analog 6y decreased
slightly versus 4a but was similar to the unsubstituted phenyl
6x. Clearly substitution at this position is well tolerated with the
enzyme’s ability to accommodate a variety of diverse hydrophobic
groups. For reasons that are unclear, the metabolic stability of the
benzyl analog 6y appears to be enhanced versus the 4-F-phenyl
(4a) and other substituted benzyl derivatives (6z and 6aa). The
most potent inhibitor in this comparison is the phenethyl analog
6bb, which exhibits superior inhibitory activity against both hu-
man (IC50 = 2.2 nM) and mouse (IC50 = 54 nM) 11b-HSD-1,
although its metabolic stability in mouse liver microsomes is quite
poor.
Select compounds from Tables 1 and 2 were tested for enzy-
matic assay in a HEK cell line expressed with human 11b-HSD-1
and were found to exhibit reasonable potency.20 The result sug-
gests that this class of compounds have ability to penetrate cells
and reach the target enzyme in the mitochondria.
6. Tiwari, A. IDrugs 2010, 13, 266. and references cited therein.
7. Some of the most recent publications in this field are: (a) Wang, H.; Robl, J. A.;
Hamann, L. G.; Simpkins, L.; Golla, R.; Li, Y.-X.; Seethala, R.; Zvyaga, T.; Gordon,
D. A.; Li, J. J. Bioorg. Med. Chem. Lett. 2011, 21, 4146; (b) Xu, Z.; Tice, C. M.; Zhao,
W.; Cacatian, S.; Ye, Y.-J.; Singh, S. B.; Lindblom, P.; McKeever, B. M.; Krosky, P.
M.; Kruk, B. A.; Berbaum, J.; Harrison, R. K.; Johnson, J. A.; Bukhtiyarov, Y.;
Panemangalore, R.; Scott, B. B.; Zhao, Y.; Bruno, J. G.; Togias, J.; Guo, J.; Guo, R.;
Carroll, P. J.; McGeehan, G. M.; Zhuang, L.; He, W.; Claremon, D. A. J. Med. Chem.
2011, 54, 6050; (c) 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, M. M.; Cupples, R.; Kim,
K. W.; St. Jean, D. J.; Johansson, L.; Henriksson, M. A.; Williams, M.; Vallgarda,
J.; Fotsch, C.; Wang, M. J. Med. Chem. 2010, 53, 4481; (d) 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. J. Med. Chem. 2009, 52, 5449; (e) Xiang, J.; Wan, Z.-K.;
Li, H.-Q.; Ipek, M.; Binnun, E.; Nunez, J.; Chen, L.; McKew, J. C.; Mansour, T. S.;
Xu, X.; Suri, V.; Tam, M.; Xing, Y.; Li, X.; Hahm, S.; Tobin, J.; Saiah, E. J. Med.
Chem. 2008, 51, 4068.
Despite the marginal mouse potency generally observed for this
chemotype, several compounds having reasonable human in vitro
potency and good human/mouse metabolic stability were assessed
in a coarse mouse pharmacokinetic model and the results are sum-
marized in Table 3. Unfortunately all compounds exhibited low
plasma exposure in mice, despite their reasonably good PAMPA
permeability, solubility, and in vitro microsomal metabolic stabil-
ity profiles (shown in Tables 1 and 2). To rationalize these results,
we looked to see if phase 2 metabolism could be contributing to
the low exposure and rapid clearance, particularly since the pres-
ence of the carboxyl groups enable glucuronidation as a route of
elimination. While glucuronidation rates varied in the human
in vitro assay, half-lives in the mouse assay were quite short, sug-
gesting that rapid glucuronidation (perhaps coupled with some
oxidative metabolism) was responsible for the poor pharmacoki-
netic exposures of these compounds upon oral dosing.
8. Refs. 2c,f and the references cited therein.
9. (a) Rohde, J. J.; Pliushchev, M. A.; Sorensen, B. K.; Wodka, D.; Shuai, Q.; Wang, J.;
Fung, S.; Monzon, K. M.; Chiou, W. J.; Pan, L.; Deng, X.; Chovan, L. E.; Ramaiya,
A.; Mullally, M.; Henry, R. F.; Stolarik, D. F.; Imade, H. M.; Marsh, K. C.; Beno, D.
W. A.; Fey, T. A.; Droz, B. A.; Brune, M. E.; Camp, H. S.; Sham, H. L.; Frevert, E.;
Uli, ; Jacobson, P. B.; Link, J. T. J. Med. Chem. 2007, 50, 149; (b) Ebdrup, S.
WO2007107550.
10. (a) Tice, C. M.; Zhao, W.; Xu, Z.; Cacatian, S. T.; Simpson, R. D.; Ye, Y.-J.; Singh, S.
B.; McKeever, B. M.; Lindblom, P.; Guo, J.; Krosky, P. M.; Kruk, B. A.; Berbaum, J.;
Harrison, R. K.; Johnson, J. J.; Bukhtiyarov, Y.; Panemangalore, R.; Scott, B. B.;
Zhao, Y.; Bruno, J. G.; Zhuang, L.; McGeehan, G. M.; He, W.; Claremon, D. A.
Bioorg. Med. Chem. Lett. 2010, 20, 881; (b) WO2004/056745.
11. Though the attempts to grow co-crystal structure of ligands 4–6 with human
11b-HSD-1 failed, we did obtain several co-crystal structures of their
corresponding amide derivatives bound to human 11b-HSD-1. The
conformation of adamantyl moiety and carbonyl group in co-crystal
In summary, we have designed and synthesized a series of poly-
cyclic carboxylic acids as potent and selective21 11b-HSD-1 inhibi-
tors. These compounds were generally potent against the human
enzyme, but less potent against the mouse (better analogs exhib-
ited IC50 values in the 100–300 nM range). Though not fully
reported, the chemotype in general exhibited favorable develop-
ment attributes such as good metabolic stability, weak inhibition

6704
X.-Y. Ye et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6699–6704
structures are superimposed to those shown in docking model (Fig. 3). The
results will be reported elsewhere.
12. The software used for docking is Glide-SP implemented in Maestro version 9.0;
http://www.schrodinger.com.
in human, but E226 in mouse. In both cases, the interaction of the ligand with
the human enzyme is more favored than with the mouse enzyme. This
difference can be especially dependent upon the chemotype. For example, see
our recent publication in Ref. 7a.
13. PDB code 2BEL, resolution 2.11 Å. See http://www.rcsb.org/pdb/
explore.do?structureId=2BEL for details.
18. In general, compounds in Tables 1 and 2 were generated via the reaction
sequences in Scheme 1 using different ketones 7 as the starting material and
appropriate RMgX or RLi as the nucleophile. Some acids in Table 1 require
additional steps for functional group transformation. For typical experimental
procedure, see Supplementary data.
14. (a) Baty, J. D.; Jones, G.; Moore, C. J. Org. Chem. 1969, 34, 3295; Other methods
involve molecular sieves or dehydrative condensation with TiCl4. See: (b) Vogt,
P. F.; Molino, B. F.; Robichaud, A. J. Synth. Commun. 2001, 31, 679; (c) Brown, R.
F. C.; Coulston, K. J.; Eastwood, F. W.; Gatehouse, B. M.; Guddatt, L. W.; Luke,
W.; Pfenninger, M.; Rainbow, I. Aust. J. Chem. 1984, 37, 2509.
19. Compound 6w was synthesized from compound 6v. See Supplementary data
for details.
15. Haslego, M. L.; Smith, F. X. Synth. Commun. 1980, 10, 421.
20. In general, the cellular assay IC50 values were 5–10-fold less potent than the
corresponding IC50 values in the biochemical assay.
21. Though not reported, all inhibitors tested exhibited little/no inhibition (<50% at
10 lM) versus human 11b-HSD-2.
22. The lipophilic nature of these carboxylic acids resulted in a concern for PPAR
16. Knöpfel, T. F.; Carreira, E. M. J. Am. Chem. Soc. 2003, 125, 6054.
17. In general, the higher potency of the inhibitors for the human enzyme versus
the mouse can be attributed to the higher lipophilic nature of the binding
pocket of human 11b-HSD-1. Significant differences in the binding pocket
include, A226 and Y177 in human versus E226 and Q177 in mouse. The
docking model (Fig. 3) shows the adamantyl group pointing towards Y177 in
human, but Q177 in mouse, and the 4-F-phenyl group pointing towards A226
activity. We tested several compounds from this class, including 4a, and no
PPAR activity across
30 M.
a, c, d subtypes was detected up to a concentration of
l