Synthesis and evaluation of piperidine urea derivatives as efficacious 11β-hydroxysteroid dehydrogenase type 1 inhibitors in diabetic ob/ob mice

Article abstract of DOI:10.1016/j.bmcl.2012.02.095

11β-Hydroxysteroid dehydrogenase type 1 (11β-HSD1) has attracted considerable attention as a potential target for the treatment of diabetes and metabolic syndrome. Herein we report the design, synthesis and efficacy evaluation of novel amide and urea 11β-

Full text of DOI:10.1016/j.bmcl.2012.02.095

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2748–2752
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of piperidine urea derivatives as efficacious
11b-hydroxysteroid dehydrogenase type 1 inhibitors in diabetic ob/ob mice
⇑
⇑
Liming Zhang, Junhua Chen, Mengmeng Ning, Qingan Zou, Ying Leng , Jianhua Shen
State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zuchongzhi Road, Zhangjiang Hi-Tech Park, Shanghai 201203, China
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) has attracted considerable attention as a poten-
tial target for the treatment of diabetes and metabolic syndrome. Herein we report the design, synthesis
and efficacy evaluation of novel amide and urea 11b-HSD1 inhibitors. Structure–activity relationship
studies led to the identification of 10c, which was efficacious in a diabetic ob/ob mouse model and
reduced fasting and non-fasting blood glucose levels after ip dosing.
Received 1 December 2011
Revised 19 February 2012
Accepted 28 February 2012
Available online 3 March 2012
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
11b-HSD1 inhibitor
Diabetes
Metabolic syndrome
Amide
Urea
Glucocorticoids are steroid hormones and play a key role in reg-
ulating glucose and lipid homeostasis.1 Excessive levels of gluco-
corticoids can cause metabolic complications, including insulin
resistance, central obesity, hyperglycemia, dyslipidemia, hyperten-
sion, and type 2 diabetes.2–4 Glucocorticoid action depends on cir-
culating hormone levels and local activation or regeneration by key
enzymes, 11b-hydroxysteroid dehydrogenase isozymes (11b-
HSDs).
There are two types of 11b-HSDs, 11b-HSD1 and 11b-HSD2.
11b-HSD1 functions predominantly as an NADPH-dependent
reductase and converts inactive cortisone into the receptor-active
glucocorticoid cortisol in humans. Moreover, it is highly expressed
in the liver and adipose tissues. Thus, it locally amplifies the gluco-
corticoid action in specific tissues.5,6 11b-HSD2, on the other hand,
is a NAD-dependent enzyme that catalyzes the dehydrogenase
reaction of converting cortisol to cortisone (Fig. 1) and is mainly lo-
cated in kidney. Inhibition of 11b-HSD1 has the potential to control
cortisol concentrations in the liver and adipose without affecting
systemic circulating concentrations.
resistance, dyslipidemia and hypertension.8,9 These mouse genetic
models demonstrated the connection between 11b-HSD1 and met-
abolic syndrome as well as diabetes. As a matter of fact, a number
of small molecule inhibitors of 11b-HSD1 have been disclosed in
the past few years.10–12 In a phase 2 study, Incyte’s small molecule
inhibitor INCB13739 showed a positive trend in reducing fasting
glucose and HbA1c with a 200 mg dose.13 MK-0916 modestly im-
proved HbA1c, body weight, and blood pressure in a phase 2 trial
in patients with type 2 diabetes.14
These promising studies strongly suggest that 11b-HSD1 is a
potential therapeutic target for a broad range of disorders that
could be improved by decreased intracellular glucocorticoid levels
(i.e., cortisol), including metabolic syndrome as well as type 2
diabetes.
Recently, we have reported that (phenylsulfonamidomethyl)
benzamides and (phenylsulfonamidomethyl)nicotinamides were
OH
OH
11b-HSD1 deficient mice show enhanced glucose tolerance,
attenuated gluconeogenic responses and improved lipid and
lipoprotein profiles.7 Conversely, transgenic mice overexpressing
11b-HSD1 selectively in adipose develop many of the features of
the metabolic syndrome such as glucose intolerance, insulin
O
O
O
OH
HO
OH
H
11 -HSD1
11 -HSD2
H
H
H
H
H
O
O
cortisone
cortisol
⇑
Corresponding authors. Tel.: +86 21 20231969; fax: +86 21 20231971 (J.S.);
tel./fax: +86 21 50806059 (Y.L.).
Figure 1. Interconversion of cortisone and cortisol mediated by 11b-HSD1 and 11b-
E-mail addresses: yleng@mail.shcnc.ac.cn (Y. Leng), jhshen@mail.shcnc.ac.cn
HSD2.
(J. Shen).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.02.095

L. Zhang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2748–2752
2749
O
O
O
O
N
S
Cl
Cl
S
N
H
N
H
N
H
R
N
H
O
R
O
O
bicyclic analogues
R = alkyl
Initial leads
Figure 2. Strategic approach from (phenylsulfonamidomethyl)benzamides to bicyclic analogues.
potent and selective inhibitors of 11b-HSD1.15,16 Molecular model-
Table 1
ing experiments demonstrated that the N-cycloalkyl amide moiety
was the essential binding part of these molecules.16 Preliminary
structure–activity relationship (SAR) studies showed that alkyl
substitution on the nitrogen of sulfonamide is beneficial to im-
prove potency. We presumed that the enhanced potency might
be influenced by the steric effect of the substituent. Given the
above observations, we hypothesized the possibility of constrain-
ing the sulfonamide moiety into a six-membered ring to form bicy-
clic analogues (Fig. 2). We aimed to improve biological potency and
increase structural novelty by incorporating a suitable bicyclic
moiety. As our initial leads were generally highly lipophilic (clogP
>5),15 we also hoped to reduce the lipophilicity of these molecules
to values considered more suitable for orally administered drugs.
Herein we report the discovery of potent 11b-HSD1 amide and
urea inhibitors with good in vivo efficacy in diabetic ob/ob mice.
The synthesis of compound 5 is depicted in Scheme 1. Com-
pound 1 was alkylated with methyl 4-(bromomethyl)benzoate 2
to give 3, which was subsequently hydrolysed to yield carboxylic
acid 4. Finally, compounds 5 were obtained by coupling of various
11b-HSD1 inhibition for 4-substituted benzamides 5
O
O
N
Cl
S
R1
N
H
O
O
Compound
R1
11b-HSD1 IC50 (nM)
Human
686
Mouse
H
N
5a
5b
5c
529
145
180
H
N
802
443
H
N
amines with the acid
4 using standard methods for amide
formation.17
N
N
5d
5e
60
90
185
210
Compounds were evaluated for inhibition of human and mouse
11b-HSD1 enzymes by the scintillation proximity assay (SPA)18
and selected compounds were evaluated in human and mouse
11b-HSD2 assays (for details, see Supplementary data). Glyc-
yrrhizic acid and carbenoxolone were used as reference
compounds.
First, the effect of modifying alkyl substitution in the amide re-
gion of compounds 5 was investigated and the results were shown
in Table 1. Generally, 11b-HSD1 inhibition activity was maintained
after the incorporation of a thiadiazine ring. On the distal alkyl
groups, various lipophilic amides were well tolerated. The N-cyclo-
hexyl amide 5a showed a moderate potency with a human IC50 of
686 nM. As compared to 5a, introduction of cycloheptyl (5b) or
cyclooctyl (5c) only resulted in improvement of potency in mouse
assays. When N-cyclohexyl amide (5a) was replaced with the more
lipophilic 1-adamantanyl amide group (5g), the potency increased
in both human and mouse enzymatic assays. Introduction of alkyl
N
5f
55
47
84
14
H
N
5g
N
5h
70
10
55
8
Glycyrrhizic acid
O
O
O
O
Cl
S
Cl
S
b
a
Br
N
NH
+
O
COOCH3
N
H
O
N
H
O
O
2
1
3
O
O
O
O
cord
Cl
S
Cl
S
N
N
R1
OH
O
N
H
O
N
H
O
O
5
4
Scheme 1. Synthesis of 4-substituted benzamides 5. Reagents and conditions: (a) K2CO3, DMF, rt; (b) 1 mol/L NaOH, MeOH, reflux; (c) (i) SOCl2, CH2Cl2, reflux; (ii) Et3N,
amine, CH2Cl2; (d) EDCI, HOBt, CH2Cl2, amine.

2750
L. Zhang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2748–2752
Table 2
Table 3
11b-HSD1 inhibition for 3-substituted benzamides 6
11b-HSD1 inhibition for urea analogues 10
O
O
O
O
O
Cl
S
Cl
S
N
N
R1
1
R
N
N
H
O
N
H
O
O
R1
11b-HSD1 IC50 (nM)
Compound
10a
R1
11b-HSD1 IC50 (nM)
Compound
Human
283
Mouse
Human
Mouse
H
N
17
37
N
N
6a
6b
3
58
N
N
10b
121
21
4
1
401
10c
0.41
N
N
6c
258
N
N
10d
517
10
6d
4
69
8
Glycyrrhizic acid
10
10e
158
10
2
8
Glycyrrhizic acid
substituent at the nitrogen of the amide proved to be beneficial for
enzyme activity (5a, 5b vs 5d, 5f respectively). Compound 5d
showed 11-fold improvement in potency against human 11b-
HSD1 when compared to 5a. Compound 5g which contained a 1-
adamantanyl amide substituent exhibited the greatest potency in
this series, (human-1 IC50 = 47 nM, mouse-1 IC50 = 14 nM) and
was 14-fold more potent than 5a. These results suggest that large
lipophilic amides could better accommodate the hydrophobic
space in the enzyme binding pocket, so these groups were kept
in further optimization.
Next, 3-substituted benzamide analogues 6 were prepared to
determine the effect of amide location on 11b-HSD1 potency. The
synthesis of those analogues was similar to 5 by replacement of
intermediate 2 with methyl 3-(bromomethyl)benzoate. The inhib-
itory activities of compounds 6 against 11b-HSD1 are compiled in
Table 2. In accord with previous studies,19 ‘V-shaped’ isomers 6
were more potent than linear amide 5 in human assays. All com-
pounds of 6 showed high binding affinities to human 11b-HSD1
enzyme (human IC50 <5 nM). However, change of amide location
was found to be detrimental to potency in mouse assays, and the
mouse IC50 of these analogues were 20- to 260-fold lower. Such
discrepancies in 11b-HSD1 activity between species might be
anticipated, because the mouse shares only 79% homology with
the human enzyme.20 However, different species activity presents
a challenge for identifying predictive in vivo models.
To narrow species discrepancies, we then turned our attention
to the central phenyl region of the molecule. We replaced the phe-
nyl ring with piperidine and synthesized piperidine-1-carboxam-
ide analogues 10. The synthesis of 10 was summarized in
Scheme 2. Compound 8 was prepared by alkylation reaction be-
tween 1 and 7 in the presence of NaH in DMF. Conversion of 8 to
9 was achieved by removing the Boc group via TFA. Finally, treat-
ment of intermediate 9 with triphosgene followed by correspond-
ing amines generated the target analogues 10.17
Table 3 displays limited examples from SAR study of 10.
Replacement of the phenyl ring with piperidine led to improved
potency against mouse 11b-HSD1, but reduced potency against
O
O
O
O
b
Cl
S
a
Cl
S
NH
Br
N
+
NBoc
NBoc
N
H
O
N
H
O
7
8
1
O
O
O
O
Cl
S
c
Cl
S
N
N
N
R1
NH
N
H
O
N
H
O
O
10
9
Scheme 2. Synthesis of piperidine ureas 10. Reagents and conditions: (a) NaH, DMF, 60 °C; (b) TFA, 1,4-dioxane, 60 °C; (c) triphosgene, Et3N, amine, CH2Cl2.

L. Zhang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2748–2752
2751
Table 4
Based on these SAR studies, compound 6c and 10c were se-
lected for pharmacokinetic evaluation (Table 5). Unfortunately
both compounds exhibited poor pharmacokinetics in rodent mod-
els upon oral dosing. As an example, compound 10c produced poor
plasma exposures upon dosing in rats at a 5 mg/kg oral dose with
an AUC of 147 ng h/mL. Because of the poor rat pharmacokinetic
profiles, 10c was dosed by intraperitoneal injection for in vivo
pharmacodynamic evaluation.
11b-HSD1 and 11b-HSD2 inhibition for compound 10
Compounds
11b-HSD1 IC50 (nM)
11b-HSD2 IC50 (nM)
Human
Mouse
Human
Mouse
10b
10c
10e
121
21
158
10
37
0.41
2
2200
860
720
1
>100,000
>100,000
>100,000
Glycyrrhizic acid
Carbenoxolone
8
82
The in vivo efficacy of compound 10c was investigated in ob/ob
mice, an animal model of type 2 diabetes. As shown in Figure 3,
after 5 days of the treatment with compound 10c (50 mg/kg) once
daily by intraperitoneal injection, the fasting blood glucose level in
ob/ob mice was reduced by 37.1%, compared with that in vehicle
control mice. The fasting blood glucose level in ob/ob mice was
maintained the reduction by the rate 33.3%, 39.4%, and 36.1% at
days 10, 15, and 20, respectively (Fig. 3a), which reflected the sta-
ble performance of compound 10c on glucose control in ob/ob
mice. Similarly, the non-fasting blood glucose level in ob/ob mice
was reduced by 25% after 5 days treatment with compound 10c.
Moreover, the effect of compound 10c in lowering the blood glu-
cose level in ob/ob mice was also sustained at the rate of 27.9%,
34.6%, and 35.6% at day 10, 15, and 20, respectively (Fig. 3b). Dur-
ing the 20 days treatment, compound 10c reduced the fasting and
non-fasting blood glucose level in ob/ob mice by an average rate of
31.4% and 25.7%, respectively (Fig. 3a and b). In our studies, the ef-
fects of stress response on fasting and non-fasting glucose levels
were not observed (for details, see Supplementary data). These re-
sults showed that treatment with 11b-HSD1 inhibitor 10c im-
proved blood glucose control, indicating that compound 10c
exhibited good in vivo efficacy against type 2 diabetes in a mouse
model.
Though 10c showed favorable profiles in an in vivo mouse mod-
el, two key issues proved difficult to address in this series. First, 10c
exhibited poor PK profiles. The highly lipophilic nature of alkyl
groups likely contributes to the generally poor metabolic stability
of these compounds. Another major concern is the marginal selec-
tivity against the human 11b-HSD2 enzyme. As seen in Table 4,
compound 10c exhibited about 40-fold selectivity for human
11b-HSD1 versus human 11b-HSD2 (IC50 = 860 nM) which needs
further improvement.
In conclusion, we have identified a series of amide and urea
derivatives as novel and potent 11b-HSD1 inhibitors. Several com-
pounds exhibited high in vitro activity toward human 11b-HSD1.
Compound 10c was selected for an in vivo experiment. It exhibited
good in vivo efficacy in diabetic ob/ob mice and reduced fasting
and non-fasting blood glucose levels after ip dosing. 10c was
Table 5
Pharmacokinetic profiles for 6c and 10c in male Sprague–Dawley ratsa
Compounds Tmax
Cmax
AUC0–t
AUC0–1
MRT
t1/2
(h)
(h)
(ng/mL)
(ng h/mL)
(ng h/mL)
(h)
6c
10c
2.3
2.0
3.86
46.8
15.5
173
16.1
209
4.02
4.24
2.02
2.30
a
Mean values (n = 3) from three rats for oral dosing.
the human enzyme. The mouse IC50 of 10 is 3- to 80-fold higher
than human IC50. Compound 10c was the best inhibitor among this
series and exhibited a human 11b-HSD1 IC50 of 21 nM. This series
of analogues demonstrated that the phenyl ring was not essential
for 11b-HSD1 potency and could be replaced with more flexible
ring.
Lipophilicity is a key physicochemical property for drug candi-
dates.21 According to generally accepted criteria of drug-likeliness,
the calculated logP (clogP) value of a drug candidate should be
smaller than five.22,23 As hoped, introduction of a bicyclic ring
had a major impact on the overall lipophilicity. Compared with
our initial leads, compounds 5, 6 and 10 demonstrated favorable
clogP profiles (clogP <5),24 suggesting that they might have im-
proved drug-like properties.
11b-HSD2 is an NAD-dependent dehydrogenase that catalyzes
the reverse reaction of 11b-HSD1 to inactivate cortisol to cortisone.
Inhibition of 11b-HSD2 might cause hypokalemia and hyperten-
sion due to kidney mineralocorticoid receptor activation.1,25 There-
fore, 11b-HSD1 inhibitors would need to demonstrate selectivity
over 11b-HSD2 to avoid undesirable side effects. Selected com-
pounds of 5 and 6 exhibited high selectivity against human 11b-
HSD2 with IC50 selectivity ratio >1000. Many of the compounds
10 exhibited marginal selectivity and left an area for further
improvement (Table 4).
Figure 3. Antidiabetic efficacy of compound 10c in ob/ob mice. Ob/ob mice were administered with compound 10c (50 mg/kg) once daily by intraperitoneal injection for
20 days. Fasting (a) and non-fasting (b) blood glucose levels were measured every five days. Values are expressed as means SEM for n = 7 mice. ⁄P <0.05 versus vehicle
control mice.

2752
L. Zhang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2748–2752
10. Webster, S. P.; Pallin, T. D. Expert Opin. Ther. Pat. 2007, 17, 1407.
11. Hughes, K. A.; Webster, S. P.; Walker, B. R. Expert Opin. Investig. Drugs 2008, 17,
481.
12. Boyle, C. D.; Kowalski, T. J. Expert Opin. Ther. Pat. 2009, 19, 801.
13. Rosenstock, J.; Banarer, S.; Fonseca, V. A.; Inzucchi, S. E.; Sun, W.; Yao, W.;
Hollis, G.; Flores, R.; Levy, R.; Williams, W. V.; Seckl, J. R.; Huber, R. Diabetes
Care 2010, 33, 1516.
identified as a highly advanced lead and further optimization of its
PK profile and selectivity toward 11b-HSD2 is currently under
progress.
Acknowledgment
14. Feig, P. U.; Shah, S.; Hermanowski-Vosatka, A.; Plotkin, D.; Springer, M. S.;
Donahue, S.; Thach, C.; Klein, E. J.; Lai, E.; Kaufman, K. D. Diabetes Obes. Metab.
2011, 13, 498.
15. Zhang, X.; Zhou, Z.; Yang, H.; Chen, J.; Feng, Y.; Du, L.; Leng, Y.; Shen, J. Bioorg.
Med. Chem. Lett. 2009, 19, 4455.
We gratefully acknowledge financial support from the National
Natural Science Foundation of China (30901838).
16. Zhang, X.; Zhou, Y.; Shen, Y.; Du, L.; Chen, J.; Leng, Y.; Shen, J. Acta Pharmacol.
Sin. 2009, 30, 1344.
Supplementary data
17. For general procedures and characterization data, see the Supplementary data.
18. Mundt, S.; Solly, K.; Thieringer, R.; Hermanowski-Vosatka, A. Assay Drug Dev.
Technol. 2005, 3, 367.
19. Wang, H.; Ruan, Z.; Li, J. J.; Simpkins, L. M.; Smirk, R. A.; Wu, S. C.; Hutchins, R.
D.; Nirschl, D. S.; Van Kirk, K.; Cooper, C. B.; Sutton, J. C.; Ma, Z.; Golla, R.;
Seethala, R.; Salyan, M. E. K.; Nayeem, A.; Krystek, S. R., Jr.; Sheriff, S.; Camac, D.
M.; Morin, P. E.; Carpenter, B.; Robl, J. A.; Zahler, R.; Gordon, D. A.; Hamann, L.
G. Bioorg. Med. Chem. Lett. 2008, 18, 3168.
20. Barf, T.; Vallgårda, J.; Emond, R.; Häggström, C.; Kurz, G.; Nygren, A.; Larwood,
V.; Mosialou, E.; Axelsson, K.; Olsson, R.; Engblom, L.; Edling, N.; Rönquist-Nii,
Y.; Öhman, B.; Alberts, P.; Abrahmsén, L. J. Med. Chem. 2002, 45, 3813.
21. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881.
22. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Delivery Rev.
1997, 23, 3.
23. Christopher, A. L. Drug Discovery Today: Technol. 2004, 1, 337.
24. ClogP calculation based on Pipeline Pilot Webport.
25. White, P. C.; Mune, T.; Rogerson, F. M.; Kayes, K. M.; Agarwal, A. K. Pediatr. Res.
1997, 41, 25.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2012.02.095.
References and notes
1. Tomlinson, J. W.; Walker, E. A.; Bujalska, I. J.; Draper, N.; Lavery, G. G.; Cooper,
M. S.; Hewison, M.; Stewart, P. M. Endocr. Rev. 2004, 25, 831.
2. Wang, M. Nutr. Metab. 2005, 2, 3.
3. Wang, M. Drug Dev. Res. 2006, 67, 567.
4. Stulnig, T. M.; Waldhäusl, W. Diabetologia 2004, 47, 1.
5. Oppermann, U. C. T.; Persson, B.; Jörnvall, H. Eur. J. Biochem. 1997, 249, 355.
6. Seckl, J. R.; Chapman, K. E. Eur. J. Biochem. 1997, 249, 361.
7. Kotelevtsev, Y.; Holmes, M.; Burchell, A.; Houston, P.; Schmoll, D.; Jamieson, P.;
Best, R.; Brown, R.; Edwards, C.; Seckl, J.; Mullins, J. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 14924.
8. Masuzaki, H.; Paterson, J.; Shinyama, H.; Morton, N. M.; Mullins, J. J.; Seckl, J. R.;
Flier, J. S. Science 2001, 294, 2166.
9. Masuzaki, H.; Yamamoto, H.; Kenyon, C.; Elmquist, J.; Morton, N.; Paterson, J.;
Shinyama, H.; Sharp, M.; Fleming, S.; Mullins, J.; Seckl, J.; Flier, J. J. Clin. Invest.
2003, 112, 83.