1. Discovery of 11b-hydroxysteroid dehydrogenase type 1 inhibitor
Sung Pyo Hong a
, Ky Youb Nam b
, Young June Shin c
, Kil Won Kim a
, Soon Kil Ahn a,⇑
a
Institute for New Drug Development, Division of Life Sciences, Incheon National University, Incheon 406-772, Republic of Korea
b
Center for Development and Commercialization of Anti Cancer Therapeutics, Asan Medical Center, University of Ulsan College of Medicine, 138-736, Republic of Korea
c
R & D Center, Ahn Gook Pharm., Suwon 443-766, Republic of Korea
a r t i c l e i n f o
Article history:
Received 29 January 2015
Revised 15 June 2015
Accepted 30 June 2015
Available online 4 July 2015
Keywords:
11-b-Hydroxysteroid dehydrogenase
(11b-HSD1) inhibitor
Antidiabetes
Adamantane sulfonamide
Structure-based molecular modeling
Optimization
a b s t r a c t
Various adamantane sulfonamides showed potent inhibitory activity against 11b-hydroxysteroid dehy-
drogenase type 1 (11b-HSD1). In continuation of our efforts to discover a more potent, selective and
metabolically stable 11b-HSD1 inhibitor in mice as well as in humans, we optimized the adamantane
sulfonamide using structure-based molecular modeling. Compound 3, which has alkyl side chains on
the linker, demonstrated a potent inhibitory activity against human and mouse 11b-HSD1 (IC50 of
0.6 nM and 26 nM, respectively) and good physicochemical properties as a new anti-diabetes drug
candidate.
Ó 2015 Elsevier Ltd. All rights reserved.
Until now, many anti-diabetes drugs, such as sulfonylureas,
metformin, glitazones, DPP IV inhibitors and SGLT 2 inhibitors,
have been developed.1
11b-Hydroxysteroid dehydrogenase type 1
(11b-HSD1) inhibitors have been actively studied as novel
therapeutics for the treatment of type II diabetes.2–4
In tissues,
the cortisol level is responsible for the increase of blood glucose
and is regulated by two 11b-hydroxysteroid dehydrogenase iso-
zymes, 11b-HSD1 and 11b-hydroxysteroid dehydrogenase type 2
(11b-HSD2). 11b-HSD1 converts inactive cortisone to active corti-
sol, whereas 11b-HSD2 catalyzes the reverse reaction. It has been
experimentally proved that the activity of cortisol is controlled
not only by cortisol excretion but also by the interconversion of
active cortisol to inactive cortisone at the tissue level by
11b-HSD1 and 11b-HSD2.5
It has been suggested that the increased
glucocorticoid activity in the white adipose tissue by 11b-HSD1 is a
key player in the development of visceral obesity, insulin resis-
tance, diabetes, type 2 diabetes, dyslipidemia and hypertension
in mice.6
Therefore, the inhibition of 11b-HSD1 is expected to
decrease plasma glucose levels and typical diabetes-related syn-
dromes without appreciable side effects.
In continuation of our efforts to discover new drug candidates
with higher potency, selectivity and metabolic stability in mice
as well as in humans, we studied systematic modification of
chemical structure of N-(adamantanyl-2-yl)-(phenylsulfon-
amido)alkanamide (1) using molecular modeling studies.
H
N
O
H
N
S
O O
X
Y
n
R
1
H
N
O
H
N
S
O OF
NH2
O
2
http://dx.doi.org/10.1016/j.bmcl.2015.06.099
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.
⇑ Corresponding author. Tel.: +82 31 888 6821; fax: +82 31 888 6833.
E-mail address: skahn@incheon.ac.kr (S.K. Ahn).
Table 1
In vitro inhibitory activity of compounds with various link lengths against human and
mouse 11b-HSD1
Compound Structure 11b-HSD1 IC50
(nM)
Human Mouse
2 2 21
3
H
N
N
H
S
O
NH2
O
O O
F
0.6 26
4
H
N
NH2
O
S
F O O
28 244
Bioorganic & Medicinal Chemistry Letters 25 (2015) 3501–3506
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
2. Table 2
In vitro inhibitory activity of compounds with various alkyl substituents on the linker
against human and mouse 11b-HSD 1
Compound Structure 11b-HSD1 IC50
(nM)
Human Mouse
3 0.6 26
5
H
N
N
H
S
O
NH2
O
O O
F
>1000 >1000
6
H
N
N
H
S
O
NH2
O
O O
F
25 193
7
H
N
N
H
S
O
NH2
O
O O
F
OH
>1000 >1000
8
H
N
N
H
S
O
NH2
O
O O
F
OH
>1000 >1000
Table 3
In vitro inhibitory activity of compounds with various alkyl substituents on the linker
against human and mouse 11b-HSD1
Compound Structure 11b-HSD1 IC50
(nM)
Human Mouse
9
H
N
O
H
N
S
O OF
NH2
O
>1000 >1000
10
H
N
O
H
N
S
O OF
NH2
O
9 204
11
H
N
O
H
N
S
O OF
NH2
O
6 13
Figure 2. Connolly surface model of compound 3 (a) and compound 5 (b).
Figure 1. The docking models of compound 3 and compound 5. (a) Compound 3 is depicted in cyan with yellow binding site residues. The dashed line indicates hydrogen
bonding interactions. (b) Compound 5 is presented in magenta.
3502 S. P. Hong et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3501–3506
3. Table 4
In vitro inhibitory activity of substituted phenyl derivatives against human and mouse 11b-HSD1
H
N
N
H
S
Ar
O
O O
NH2
O
Compound Structure 11b-HSD1 IC50 (nM) Compound Structure 11b-HSD1 IC50 (nM)
Human Mouse Human Mouse
3
F
0.6 26 3g
F
Cl >1000 13
3a 2 28 3h
Cl
F
0.1 10
3b
Cl 1 12 3i
F
F
F
14 207
3c
Cl
34 130 3j
Cl
Cl
0.3 172
3d
t-Bu
2 56 3k
F
F
0.4 12
3e
HO
44 84 3l
F
F
10 18
3f
ON
>1000 130 3m
Cl
Cl
0.2 8
H
N
N
H
S
O
NH2
O
O OH
N
N
H
S
O
OMe
O
O O
H
N
H2N
O
OMe
O
H2N
OMe
O
H
N
NH
O
OMe
O
Boc
CO2HNHBoc +
(a)
(b) (c)
(d), (e)
F F
HCl
12 13
14 15
16 3
Scheme 1. Reagents and conditions: (a) EDCI, HOBt, TEA, DCM, rt, overnight; (b) 4 M HCl in dioxane, EtOAc, rt, overnight; (c) 2-fluorobenzene sulfonyl chloride, TEA, DCM, rt,
overnight; (d) 2 N NaOH, THF/EtOH (1:1), rt, overnight; (e) EDCI, HOBt, 35% aq NH3, rt, 20 h.
S. P. Hong et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3501–3506 3503
4. We previously selected (1s,3R,4s,5S,7s)-4-(3-(2-fluorophenylsulfon-
amido)-3-methylbutanamido) adamantane-1-carboxamide (2) as a
lead compound. Compound 2 showed a potent inhibitory activity
against human and mouse 11b-HSD1 (IC50 values of 2 nM and
21 nM, respectively).7
First, we studied the effect of the linker length on the 11b-HSD1
inhibitory activity. Derivatives of methylpropanoic acid with link-
ers that were one carbon shorter 3 and two carbons shorter 4 were
synthesized and evaluated for their 11b-HSD1 inhibitory activities
in comparison to compound 2 (Table 1).7
The propanoic acid derivative (3) showed comparable inhibi-
tory activities to the butanoic acid derivative (2), whereas the sul-
fonic acid derivative (4) demonstrated moderate inhibitory
activity. We determined that compound 3 was a new lead because
compound 3 showed better physicochemical and biological prop-
erties than compound 2.8
Next, we investigated the effect of an alkyl group in the linker
on 11b-HSD1 inhibitory activity. Derivatives of acetic acid (5),
cyclopropanecarboxylic acid (6), and 3-hydroxypropanoic acid (7)
and (8) were prepared and their 11b-HSD1 inhibitory activities
were compared with that of compound 3 (Table 2).
The dimethyl group looked to be crucial for the 11b-HSD 1 inhi-
bitory activity. Based on molecular modeling studies, we inferred
that the dimethyl group hydrophobically interacted with Tyr177
and Ala172 of human 11b-HSD1. The hydroxymethyl compounds
7 and 8 are sterically hindering for the hydrophobic interaction
(Figs. 1 and 2).9
We further confirmed the effect of the alkyl group in other
derivatives. As shown in Table 3, the alkyl group plays a critical
role for the hydrophobic interaction and, consequently, the
11b-HSD1 inhibitory activity. As we had already expected from
the comparable inhibitory activity of the butanoic acid derivative
(2) and the propanoic acid derivative (3), the position of the alkyl
group in the butanoic acid derivatives (10) and (11) was marginal.
This result could be anticipated from the modeling studies.
We also tried to optimize the inhibitory activity through mak-
ing various substituents to the phenyl ring. The inhibitory activity
of substituted phenyl derivatives against 11b-HSD 1 are summa-
rized in Table 4. Most of the substituted phenyl derivatives showed
potent to moderate inhibitory activities. Based on the molecular
modeling studies, we speculated that these tendencies arouse from
S
O O
+
CO2HNH2
OH (a)
Cl
S
O O
F
CO2H
OH
N
H
+
H2N
OMe
O
HCl
19
19
(b), (c), (d)
H
N
N
H
S
O
NH2
O
O O
F
F
OH
17 18
20 8
Scheme 2. Reagents and conditions: (a) Na2CO3, H2O, rt, 24 h; (b) DCC, HOBt, TEA, rt, overnight; (c) 2 N NaOH, THF/EtOH (1:1), rt, overnight; (d) EDCI, HOBt, 35% aq NH3, rt,
20 h.
Table 5
In vitro inhibitory activity of substituted adamantane derivatives against human and mouse 11b-HSD1
X
H
N
N
H
S
O
O O
F
Compound Structure 11b-HSD1 IC50 (nM) Compound Structure 11b-HSD1 IC50 (nM)
Human Mouse Human Mouse
3 –CONH2 0.6 26 21d –CH2COOH >1000 >1000
21 –COOH >1000 >1000 21e –CH2CN >1000 >1000
21a –CONHNH2 >1000 >1000 21f –C(NH)NHOH 55 >1000
21b –CN >1000 >1000 21g –CH2CONH2 15 301
21c –CH2OH 2 21 21h –CH2CH2CONH2 >1000 543
Table 6
Selectivity, ex vivo 11b-HSD1 activity and metabolic stability of compound 3
Compound 11b-HSD2 (lM) Ex vivo Metabolic
stability (remain%
@ 30 min)
Liver (%) Fat (%) Human Mouse
3 >10 57 38 75 92
H
N
N
H
S
O
OMe
O
O O
(a)
H
N
N
H
S
O
NHNH2
O
O O
F
F
16 21a
Scheme 3. Reagents and conditions: (a) hydrazine hydrate, MeOH, rt, overnight.
3504 S. P. Hong et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3501–3506
5. the location of phenyl ring residue within a solvent accessible sur-
face area.
The synthetic scheme of representative compound 3 is
described in Scheme 1. Compounds 1 to 11 except for compounds
5, 7 and 8 were synthesized according to Scheme 1. A Boc-pro-
tected aminoacid (12) was condensed with optically pure
adamantane carboxylate (13) to give compound 14, which was
deprotected and coupled with substituted benzene sulfonyl chlo-
ride to give compound 16, which in turn was converted to an
amide (3).10
Compounds 5, 7 and 8 were synthesized following Scheme 2.
Unlike Scheme 1, the coupling sequence was changed, so that sub-
stituted benzene sulfonyl chloride (17) was coupled with the
amino acid (18) first and then condensed with optically pure
adamantane carboxylate (20) to finally give compound 8. The same
reaction conditions were followed as in Scheme 1.
Next we studied the effects of substituents on the 11b-HSD1
inhibitory activity of the adamantane derivatives. The 11b-HSD1
inhibitory activities of the substituted adamantane derivatives
are summarized in Table 5.
H
N
N
H
S
O
NH2
O
O O
CN
H
N
N
H
S
O
O O
(a) (b)
H
N
N
H
S
O
O O
NHOH
NH
F
F
F
3 21b
21f
Scheme 4. Reagents and conditions: (a) TFAA, Py, dioxane, rt, 5 h; (b) NH2OH HCl, NaHCO3, MeOH, reflux, 4 h.
H
N
N
H
S
O
OMe
O
O O
(a)
H
N
N
H
S
O
OH
O O
(b)
H
N
N
H
S
O
OTs
O O
(c)
H
N
N
H
S
O
CN
O O
(d)
H
N
N
H
S
O
CONH2
O O
F F
F
F
F
16 21c
22 21e
21g
Scheme 5. Reagents and conditions: (a) LAH, THF, 0 °C to rt, 1 h; (b) TsCl, Py, DCM, rt, 12 h; NaCN, KI, DMSO, 90 °C, 24 h; (d) 30% H2O2, 0.2 N NaOH, MeOH/DMSO, 50 °C,
overnight.
H
N
N
H
S
O
OH
O O
(a), (b), (c)
(e)
H
N
N
H
S
O
O O
(d)
CO2H
H
N
N
H
S
O
O O
CO2H
H
N
N
H
S
O
O O
CONH2
F F
F
F
21c 23
24 21h
Scheme 6. Reagents and conditions: (a) N-methylmorpholine N-oxide, tetrapropylammonium perruthenate, molecular sieve, ClCH2CH2Cl, rt, overnight; (b) trimethyl
phosphonoacetate, NaH, THF, rt, 5 h; (c) 2 N NaOH, THF/EtOH, rt, 5 h; (d) H2, Pd/C, rt, overnight; (e) EDCI, HOBt, 35% aq NH3, TEA, CH3CN, rt, 20 h.
S. P. Hong et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3501–3506 3505
6. It is interesting that the effect of a 1-subutituent of adamantane
on the 11b-HSD1 inhibitory activity was volatile, although the
molecular modeling studies did not show any particular interac-
tions between 1-substituents of adamantane and 11b-HSD1.
The selectivity, ex vivo 11b-HSD1 activity and metabolic stabil-
ity of compound 3 was tested. As shown in Table 6, compound 3
showed a good selectivity for human 11b-HSD1, with a much
higher IC50 of >10 lM for 11b-HSD2. Moreover, oral administration
of compound 3 inhibited 11b-HSD1 activity by 57% and 38% in
mouse liver and epididymal fat tissues with high metabolic
stability.
The synthetic scheme for compound 21a is depicted in
Scheme 3. An adamantane ester (16) was condensed with hydra-
zine hydrate to give compound 21a.11
The synthesis of compound 21b and 21f were carried out
according to Scheme 4. Adamantane amide (3) was dehydrated
with trifluoroacetic anhydride to give adamantane nitrile (21b),
which was reacted with hydroxylamine hydrochloride to give
compound 21f.12
Compounds 21c, 21e and 21g were prepared following
Scheme 5. Adamantane carboxylate (16) was reduced to adaman-
tane alcohol (21c), substituted with cyanide to produce adaman-
tane nitrile (21e), and finally converted to adamantane amide
(21g).
The synthesis of compound 21h was performed as described in
Scheme 6. Adamantane alcohol (21c) was converted to aldehyde,
which was followed by the Wittig reaction and hydrolysis to give
an unsaturated carboxylic acid (23). This was reduced and con-
verted to adamantane amide (21h).13
In conclusion, the 11b-HSD1 inhibitory activity of N-(adaman-
tanyl-2-yl)- (phenylsulfonamido) alkanamide (1) derivatives is
the most potent in acetic acid linker and greatly dependent on
the hydrophobic interactions of the alkanoic acid linker. Various
substitutions with phenyl derivatives are tolerable for the
11b-HSD1 inhibitory activity. The effect of a 1-substituent of
adamantane on the 11b-HSD1 inhibitory activity needs further
investigation.
Acknowledgement
This work was supported by the Incheon National University
Research Grant in 2013.
References and notes
1. Rochester, C. D.; Akiyode, O. World J. Diabetes 2014, 5, 305.
2. Anagnostis, P.; Katsiki, N.; Adamidou, F.; Athyros, V. G.; Karagiannis, A.; Kita,
M.; Mikhailidis, D. P. Metab. Clin. Exp. 2013, 62, 21.
3. Chapman, K.; Holmes, M.; Seckl, J. Physiol. Rev. 2013, 93, 1139.
4. Majumdar, S. K.; Inzucchi, S. E. Endocrine 2013, 44, 47.
5. Sandeep, T. C.; Walker, B. R. Trends Endocrinol. Metab. 2001, 12, 446.
6. Morton, N. M.; Paterson, J. M.; Masuzaki, H.; Holmes, M. C.; Staels, B.; Fievet, C.;
Walker, B. R.; Flier, J. S.; Mullins, J. J.; Seckl, J. R. Diabetes 2004, 53, 931.
7. Lee, Y.; Shin, Y. J.; Ahn, S. K. Bioorg. Med. Chem. Lett. 2014, 24, 1421.
8. The water solubility of compound 3 is 0.3 mg/ml, which is five times as much
as that of compound 2. After 3 weeks of daily oral administration, fasting
plasma glucose levels had decreased by 39% and 28% in KKAy mice treated with
30 mg/kg compound 3 and 2, respectively.
9. The binding models of the 11b-HSD1/compound 3 and 5 complexes were
described in Figure 1. In docking studies, the structure of 11b-HSD1 was taken
from PDB 2ILT (http://www.rcsb.org), in which the structure was solved in a
complex bound to adamantane sulfone inhibitor. Based on the adamantane
amide coordinates, the compound structure was superimposed to an
adamantane amide inhibitor in the X-ray crystal complex. The initial
complex was optimized with 1000 steps of steepest decent and 3000 steps of
conjugate gradient while holding the 11b-HSD1 heavy atoms restrained to
their initial positions by means of a harmonic force constant of
1 kcal molÀ1
ÅÀ2
using CHARMm in Accelrys Discovery Studio 4.1.
10. All novel synthetic compounds gave satisfactory analytical and spectral data.
Selected data for 3: 1
H NMR (400 MHz, CDCl3) d 8.40 (s, 1H), 7.81 (t, 1H), 7.72–
7.70 (m, 1H), 7.46–7.38 (m, 2H), 7.15 (d, J = 7.6 Hz, 1H), 7.00 (br s, 1H), 6.72 (br
s, 1H), 3.68 (d, J = 6 Hz, 1H), 1.92–1.76 (m, 11H), 1.49 (d, J = 12.0 Hz, 2H), 1.25
(s, 6H); 13
C NMR (100 MHz, DMSO-d6) d 178.7, 172.9, 159.3, 156.8, 135.2,
130.7, 129.2, 124.9, 117.2, 59.0, 52.5, 39.3, 38.7, 38.5, 31.0, 30.1, 26.7, 25.1;
HRMS (ESI) m/z: Calcd for C21H29FN3O4S 438.1863; found 438.1857.
11. Wu, J.; Zhang, D.; Chen, L.; Li, J.; Wang, J.; Ning, C.; Yu, N.; Zhao, F.; Chen, D.;
Chen, X.; Chen, K.; Jiang, H.; Liu, H.; Liu, D. J. Med. Chem. 2013, 56, 761.
12. Pace, P.; Di Francesco, M. E.; Gardelli, C.; Harper, S.; Muraglia, E.; Nizi, E.;
Orvieto, F.; Petrocchi, A.; Poma, M.; Rowley, M.; Scarpelli, R.; Laufer, R.;
Gonzalez Paz, O.; Monteagudo, E.; Bonelli, F.; Hazuda, D.; Stillmock, K. A.;
Summa, V. J. Med. Chem. 2007, 50, 2225.
13. Hiroi, K.; Watanabe, T.; Kawagishi, R.; Abe, I. Tetrahedron: Asymmetry 2000, 11,
797.
3506 S. P. Hong et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3501–3506