Discovery of 3-hydroxy-4-cyano-isoquinolines as novel, potent, and selective inhibitors of human 11β-hydroxydehydrogenase 1 (11β-HSD1)

Article abstract of DOI:10.1016/j.bmcl.2011.09.058

Derived from the HTS hit 1, a series of hydroxyisoquinolines was discovered as potent and selective 11β-HSD1 inhibitors with good cross species activity. Optimization of substituents at the 1 and 4 positions of the isoquinoline group in addition to the co

Full text of DOI:10.1016/j.bmcl.2011.09.058

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6693–6698
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of 3-hydroxy-4-cyano-isoquinolines as novel, potent, and
selective inhibitors of human 11b-hydroxydehydrogenase 1 (11b-HSD1)
Shung C. Wu a, , David Yoon a, Janice Chin c, Katy van Kirk b, Ramakrishna Seethala a, Rajasree Golla a,
⇑
Bin He a, Thomas Harrity a, Lori K. Kunselman a, Nathan N. Morgan a, Randolph P. Ponticiello a,
Joseph R. Taylor a, Rachel Zebo a, Timothy W. Harper a, Wenying Li b, Mengmeng Wang a, Lisa Zhang a,
Bogdan G. Sleczka b, Akbar Nayeem a, Steven Sheriff b, Daniel M. Camac b, Paul E. Morin b,
John G. Everlof b, Yi-Xin Li a, Cheryl A. Ferraro c, Kasia Kieltyka c, Wilson Shou c,
Marianne B. Vath c, Tatyana A. Zvyaga c, David A. Gordon a, Jeffrey A. Robl a
a Bristol-Myers Squibb Research & Development, PO Box 5400, Hopewell, NJ 08534-5400, USA
b Bristol-Myers Squibb Research & Development, PO Box 4000, Princeton, NJ 08543-4000, USA
c Bristol-Myers Squibb Research & Development, 5 Research Parkway, Wallingford, CT 06492, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Derived fromthe HTS hit 1, a seriesofhydroxyisoquinolineswas discoveredaspotentand selective11b-HSD1
inhibitors with good cross species activity. Optimization of substituents at the 1 and 4 positions of the iso-
quinoline group in addition to the core modifications, with a special focus on enhancing metabolic stability
and aqueous solubility, resulted in the identification of several compounds as potent advanced leads.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 23 August 2011
Revised 14 September 2011
Accepted 15 September 2011
Available online 21 September 2011
Keywords:
Type 2 diabetes
11b-HSD1
Inhibitors
Hydroxyisoquinolines
Cross-species activity
Type 2 diabetes is a chronic disease, affecting 2.8% of the global
population in 2000; this is expected to rise to 4.4% in 2030.1 Cur-
rent oral treatments, either as a single agent or in combination,
do not exhibit complete or sustained glucose lowering efficacy in
type 2 diabetic patients. Thus, there is an unmet medical need
for improved antidiabetic agents which are both more effective
in safely lowering glucose levels and displaying sustained efficacy.1
The use of 11b-HSD1 inhibitors in reducing plasma glucose and
glucose output as well as increasing peripheral glucose uptake
has been supported by preclinical and limited clinical data.2
In the glucocorticoid pathway, cortisol plays an important role
in glucose homeostasis. Increased levels of intracellular cortisol
have been linked to increased glucose output by liver and de-
creased insulin uptake by adipose tissue. Intracellular cortisol lev-
els are regulated by 11b-HSD1 and 11b-HSD2. 11b-HSD1 catalyzes
the transformation of cortisone to cortisol, mainly in liver and adi-
pose tissue, while 11b-HSD2 catalyzes the transformation of corti-
sol to its inactive form cortisone in the kidney. Therefore, selective
inhibition of 11b-HSD1 in liver and adipose tissue could be a po-
tential treatment for diabetes and/or metabolic syndrome.3
Hydroxypyrimidine 1 was identified as an 11b-HSD1 inhibitor
from high-throughput screening of the BMS compound collection.
In the field of 11b-HSD1 inhibitors, compound 1 contains a unique
hydroxypyrimidine functionality which appears to act as a novel
H-bonding pharmacophore. As an advantage it has good cross spe-
cies 11b-HSD1 inhibitory activity for the mouse, cynomolgus
CN
CN
4
OH
OH
3
N
N
1
N
1
S
S
Cl
Cl
2a
11 β-HSD1 IC50 (nM):
human = 27nM
mouse = 58 nM
cyno = 53 nM
11 β-HSD1 IC50 (nM):
human = 5 nM
mouse = 5 nM
cyno = 1 nM
⇑
Corresponding author. Tel.: +1 609 818 5228; fax: +1 609 818 3560.
E-mail address: shung.wu@bms.com (S.C. Wu).
Figure 1. Generation of hydroxytetrahydroisoquinoline 11b-HSD1 inhibitors based
on HTS hit hydroxypyrimidine 1.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.058

6694
S. C. Wu et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6693–6698
R2
CN
R2
O
X
OH
NH2
OH
a
b
c
X
N
X
S
X
N
S
S
SH
R1
2, X = CH2; R2 = CN
7, X = CH2; R2 = H
13, X = O or NBn; R2 = CN
3
5, R2 = CN
6, R2 = H
4
R2
Br
OH
OH
OH
d
e or f
N
S
N
N
S
7
S
Cl
9, R2 = 4-CNPh
10, R2 = CO2Me
11. R2 = CONH2
Cl
Cl
8
g
Scheme 1. Synthetic route utilized for the generation of tetrahydroisoquinoline derivatives 2, 7, 8, 9, 10 and 13. Reagents and conditions: (a) NCCH2CN, CS2, Et3N, MeOH,
80 °C; (b) 1 N NaOH, 150 °C, 45%; (c) ArCH2Br or ArCH2Cl, K2CO3, EtOH, 90 °C, 60–80%; (d) NBS, CCl4, 100 °C, 70%; (e) 4-CN-PhB(OH)2, Pd(PPh3)4, dioxane, 80 °C, 60%; (f) CO
(25 psi), Pd(OAc)2, dppp, THF/MeOH, 70 °C, 65%; (g) (i) NaOH, MeOH/H2O; (ii) (COCl)2, DMF, dichloromethane; (iii) NH4OH, dichloromethane/MeOH, 50%.
monkey, and human enzymes (IC50 = 58, 53, and 27 nM, respec-
tively). Seeking to develop greater chemical novelty in the chemo-
type, we fused the phenyl ring onto the pyrimidine core, providing
novel compounds such as the hydroxytetrahydroisoquinoline 2a as
shown in Figure 1. This compound showed greatly improved activ-
ity versus 1 against the 11b-HSD1 enzyme while exhibiting good
potency across all species as well as selectivity over 11b-HSD2
of this series, as well as structural studies with inhibitors bound
to the enzyme.
Initial work was focused on exploring the effect of the periphe-
ral substituents on the hydroxytetrahydroisoquinoline core. The
synthesis of these compounds is outlined in Scheme 1.5 The
tetrahydroisoquinoline core was generated from commercially
available cyclic ketone 3 (X = CH2, O, or N–Bn). Condensation with
malononitrile and carbon disulfide in the presence of triethylamine
in methanol gave cyano thioxoisothiochromene 4. Treatment of 4
with aqueous sodium hydroxide at 150 °C provided cyanohydroxy-
(58% inhibition @ 10 lM). However, this particular compound
exhibited metabolic and aqueous solubility issues which have pla-
gued other 11b-HSD1 inhibitors reported in the literature.4
Addressing these issues was a focus of the medicinal chemistry
program. As such we have examined the effects of structural vari-
ation of the core ring structure and the peripheral substituents on
11b-HSD1 activity and development liabilities. Herein we report
the SAR leading to the enhancement of the biological properties
tetrahydro-isoquinoline
5 and descyanohydroxytetrahydroiso-
quinoline 6 in an ꢀ6:4 ratio. Simple alkylation of thiols 5 or 6
with benzyl halides afforded the corresponding final products 2,
7, or 13, depending upon the identity of the starting materials. Bro-
mination of 7 followed by Suzuki reaction or carbonylation of the
Table 1
SAR of tetrahydroisoquinoline analogs 2, 7, 8, 9, and 11
R2
OH
N
S
11b-HSD1 IC50 (nM)a
R1
Cmpd
R1
R2
MetStab rateb h, m, r (nmol/min mg)
Aq sol. (lg/mL)
h
m
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
7
2-Cl
H
3-Cl
4-Cl
2-OMe
2-Ph
2-SO2Me
2-CN
2,6 DiCl
2-F-5-CN
2-Cl
2-Cl
2-Cl
2Cl
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
H
Br
4-CN-Ph
CO-NH2
5
15
16
85
135
225
49
7
4
5
0.09, 0.17, 0.16
NTc
6
16
39
25
9
145
10
6
3
9
9247
2266
3831
NTc
NTc
1
NTc
0.09, 0.30, 0.30
1
NTc
NTc
NTc
32
7
NTc
0.00, 0.11, 0.19
0.15, 0.23, 0.25
0.00, 0.00, Nulld
0.03, 0.12, Nulld
0.27, 0.30, 0.30
NTc
1
11
NTc
NTc
NTc
1
6904
269
93
8
9
11
0.19, 0.18, 0.30
>10
NTc
NTc
a
b
c
IC50 values refer to biochemical human or mouse 11b-HSD1 assay data.6 All compounds have >200-fold selectivity over 11b-HSD2.
MetStab rate <0.1 is considered to be desirable (more stable); maximal rate in this assay is reported as 0.3.7
NT = not tested.
Null = failed.
d

S. C. Wu et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6693–6698
6695
Table 2
SAR of tetrahydroisoquinoline analogs 2 and 12
CN
OH
N
R
A
L
Cmpd
AL
R
11b-HSD1 IC50 (nM)a
MetStab rateb h, m, r (nmol/min mg)
Aq sol. (lg/mL)
h
m
2a
–SCH2–
–S(O)2CH2–
–S–
2-Cl-Ph
2-Cl-Ph
2-Cl-Ph
Ph
2-Cl-Ph
2-Cl-Ph
Cyclobutyl
2-CNPy-3-yl
5
5
0.09, 0.17, 0.16
0.07, 0.12, 0.14
0.07, 0.17, 0.15
NTc
6
12a
12b
12c
12d
12e
2k
240
31
2816
12
3
267
95
4348
6
6
33
10
NTc
1
—
NTc
1
–SCH2CH2–
–OCH2–
–SCH2–
–SCH2–
0.18, 0.27, 0.24
Nulld
1
9
3
0.00, 0.16, 0.24
0.04, 0.05, 0.05
NTc
38
2l
a
b
c
IC50 values refer to biochemical human or mouse 11b-HSD1 assay data. All compounds have >200-fold selectivity versus 11b-HSD2.
MetStab rate <0.1 is considered to be desirable (more stable); maximal rate in this assay is reported as 0.3.7
NT = not tested.
Null = failed.
d
intermediate bromide 8 afforded the 4-aryl-substituted tetrahy-
droisoquinoline 9 or 4-carboxymethyl tetrahydroisoquinoline 10,
respectively. Hydrolysis of the ester 10 and amidation of the result-
ing acid using oxalyl chloride and ammonium hydroxide gave
amide 11.
highly preferred. All the analogs described in Table 1 continued to
have good cross species inhibitory activity, except for the analogs
with bromo and aryl substitution at C-4 (8 & 9). In addition, all
the analogs demonstrated >200-fold selectivity versus human
11b-HSD2 (data not shown). With few exceptions, the relatively
Table 1 outlines the SAR of the peripheral substituents with
respect to 11b-HSD1 inhibitory activity. Ortho substituents on the
S-benzyl aryl group were preferred versus meta and para substitu-
ents, as exemplified by the 2-, 3- and 4-chloro substituted phenyl
analogs 2a, 2c, and 2d, respectively. The unsubstituted phenyl ana-
log 2b has comparable activity to the 3-chlorophenyl analog 2c.
When the electron withdrawing chloro group was replaced with
an electron donating methoxy group (2e), the inhibitory activity
was attenuated 27-fold compared to 2a. Smaller substituents at
the ortho position of the phenyl group were also preferred, as
exemplified by a reduction of 40- and 10-fold in inhibitory activity
with the phenyl (2f) and sulfone (2g) analogs, respectively. Other
small electron withdrawing groups, such as 2-cyano (2h), and di-
substitution, such as 2,6-dichloro and 2-fluoro-5-cyanophenyl (2i
and 2j, respectively) were well tolerated. Various substituents at
C-4 (i.e., R2) were also surveyed, including hydrogen, bromo, amide,
cyano, and aryl groups. The cyano group from the initial lead was
low aqueous solubility (most analogs 6 6 lg/mL) and metabolic
stability of this series of analogs continued to be problematic.
Biotransformation studies of 2a showed that oxidative metabo-
lism was occurring at several sites, including the benzylic group,
the sulfide atom, and the fused cyclohexane ring.8 Thus, attention
was then focused on examining the effect of the benzyl linker and
variation of the benzyl aryl group in order to improve metabolic
stability. Generation of the compounds in Table 2 was effected
by oxidation of the sulfide 2a to the corresponding sulfone 12a fol-
lowed by displacement of the sulfone with various nucleophiles.
Oxidation of the sulfide in the lead 2a to the sulfone (12a) resulted
in 50-fold loss of inhibitory activity. Shortening the –SCH2– linker
to –S– (12b) or completely removing the linker (12c) attenuated
the activity by 6- and 600-fold, respectively. Elongation of the thio-
methyl linker to thioethyl (12d) or replacement with –OCH2– (12e)
maintained 11b-HSD1 inhibitory activity, at 12 nM and 3 nM,
respectively. However, none of these structural modifications
N
OH
N
N
OH
a
OH
b, c
N
R3
A
N
Br
O
R3
O
R3
A
16, A = S
17, A = O
14
15
OH
OH
N
N
N
N
d
e
R
AH
R3
R3
18, A = S
19, A = O
20a-g, A = S
21, A = O
Scheme 2. Synthetic route utilized for the generation of isoquinoline derivatives 20–21. Reagents and conditions: (a) NCCH2CO2Et, NaH, CuBr, toluene, 110 °C, 30–50%; (b)
Ac2O, 4-MeO(Ph)CH2SH or 4-MeO(Ph)CH2OH, CH3CN, 75 °C; (c) K2CO3, MeOH, 45–60% over two steps; (d) TFA, CF3SO3H, dichloromethane, 70 °C, 65–70%; (e) ArylCH2Br,
K2CO3, EtOH, 90 °C, 70–80%.

6696
S. C. Wu et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6693–6698
Table 3
SAR of isoquinoline analogs 13, 20, and 21
CN
OH
Ring
N
R
A
Cmpd
2a
Ring
A
S
S
R
11b-HSD1 IC50 (nM)a
MetStab rateb h, m, r (nmol/min mg)
Aq sol. (lg/mL)
h
m
2-Cl-Ph
Ph
5
5
0.09, 0.17, 0.16
6
13a
5
50
0.04, 0, 0.07
NTc
13b
13c
S
S
2-Cl-Ph
2-Cl-Ph
104
13
4348
21
0.08, 0.04, 0
NTc
8
N
Bn
O
0.12, 0.07, 0.04
13d
20a
S
S
2-CN-Py-3-yl
2-Cl-Ph
12
4
105
19
0, 0.01, 0
238
1
O
0.17, 0.20, 0.23
F
20b
20c
S
S
2-Cl-Ph
2-Cl-Ph
8
6
7
0.24, 0.08, 0.13
012, 0.08, 0.05
NTc
NTc
10
F
F
20d
20e
20f
20g
21
S
S
S
S
O
2-Cl-Ph
8
14
62
7
0.12, 0.10, 0.05
0.25, 0.05, 0.04
0.09, 0, 0.04
NTc
8
2-Cl-Ph
10
2
CN
F
2-Cl-Ph
1
2-CNPy-3-yl
2-Cl-Ph
8
24
21
0, 0.02, 0.02
3
F
8
0.05, 0.02, 0.04
2
F
a
b
c
IC50 values refer to biochemical human or mouse 11b-HSD1 assay data. All compounds have >200-fold selectivity versus 11b-HSD2, except for 13a.
MetStab rate <0.1 is considered to be desirable (more stable); maximal rate in this assay is reported as 0.3.7
NT = not tested.
resulted in a significant improvement in the metabolic stability of
2a (except for 12e). Replacement of the aryl group with a cyclobu-
tyl group (2k) was tolerated for 11b-HSD1 activity and resulted in
enhanced stability in human but not rodent microsomes. In con-
trast, when the 2-chlorophenyl group was replaced with the more
polar 2-cyanopyridin-3-yl (2l), 11b-HSD1 activity was retained and
cross-species metabolic stability and aqueous solubility were both
improved. This suggested that increasing the polarity of the aryl
group can be tolerated for 11b-HSD1 inhibitory activity and may
address some of the metabolism and solubility issues of the initial
lead.
Modification of the cyclohexane ring in 2a was chemically less
straightforward and thus were more challenging to prepare. Some
of the analogs in this series, such as 13a–d, were prepared by the
chemistry outlined in Scheme 1 starting with the appropriate ke-
tone. Preparation of the hydroxyisoquinoline analogs 20a–g and
21 is outlined in Scheme 2. The synthesis began with the reaction
of oxime 14 with ethyl 2-cyanoacetate in the presence of NaH and
CuBr in toluene at 110 °C to give hydroxyisoquinoline N-oxide
15.9a Acylation of the N-oxide with excess acetic anhydride, addi-
tion of 4-methoxybenzylthiol or 4-methoxybenzyl alcohol, fol-
lowed by subsequent hydrolysis of the intermediate acetate

S. C. Wu et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6693–6698
6697
afforded 4-methoxybenzylthio hydroxyisoquinoline 16 or 4-
methoxybenzyloxy hydroxyisoquinoline 17, respectively.9b Depro-
tection of the 4-methoxybenzyl group with TFA/triflic acid
followed by alkylation with the corresponding aryl bromide gave
the desired isoquinolines 20a–g or 21.
The effects of various modifications on the core ring structure
are outlined in Table 3. Benzofusion of 2a afforded compound
13a, which exhibited superior in vitro microsomal stability and po-
tent human 11b-HSD1 inhibitory activity. Unfortunately both
selectivity versus 11b-HSD2 (IC50 = 386 nM) and mouse activity
were attenuated. Introduction of the more polar N-benzyl func-
tionality (13b) resulted in an improvement in metabolic stability
versus 2a; however, activity versus the human and mouse enzyme
was attenuated 20- and 800-fold, respectively. Incorporation of
oxygen into the cyclohexyl-fused ring (13c) was generally well
tolerated. Further optimization of the benzyl substituent by
replacement of the 2-chlorophenyll group with the aforemen-
tioned 2-cyanopyridin-3-yl group afforded 13d, which in general
provided an excellent balance of human 11b-HSD1 activity, meta-
bolic stability and aqueous solubility.
Concurrently, we also explored the SAR of the isoquinoline ser-
ies, which seemed to parallel that of tetrahydroisoquinoline series
with respect to inhibitory potency of the 11b-HSD1 enzyme. Sur-
prisingly the metabolic stability of 20a was attenuated compared
to its saturated analog 2a. However, it was found that metabolic
stability could be modulated by judicious substitution on the iso-
quinoline ring as shown by analogs 20b–f. 7-Fluoro substitution
(20f) provided overall optimal 11b-HSD1 activity and metabolic
stability. In an attempt to improve the aqueous solubility of 20f,
the 2-chlorophenyl group was replaced with the 2-cyanopyridin-
3-yl group to give 20g. Unfortunately, the aqueous solubility was
only moderately improved. Replacement of the S-benzyl group
with O-benzyl (21) resulted in a compound with a superior meta-
bolic stability profile, but was still sub-optimal with respect to
aqueous solubility.
Several analogs with good potency, acceptable metabolic stabil-
ity and aqueous solubility were evaluated for pharmacokinetic
exposure studies in Balb/C mouse in the hope of finding a com-
pound suitable for in vivo pharmacodynamic testing. The results
are summarized in Table 4. Unfortunately, all the ‘optimized’ ana-
logs (in terms of the in vitro metabolic stability) exhibited reduced
exposure as compared to the HTS hit 1. Despite these findings, all
of the compounds in Table 4 were tested (30 mpk, po) for acute
pharmacodynamic activity in a mouse dehydrocorticosterone to
corticosterone conversion assay.10 None of these compounds
exhibited a response in this model.
Figure 2. Co-crystal structure of compound 13d complexed to human 11b-HSD1
and cofactor NADP+. For clarity, the NADP+ has been omitted from this represen-
tation, and only a few of the surrounding residues are included in the figure. The
initial 2Fo-Fc electron density prior to fitting 13d contoured at 1
r is shown in
magenta caged contours. The hydrogen bonds between Tyr 183 OH and Ser 170 OH
and the carbonyl oxygen of 13d are shown as a series of short prolate ellipsoids and
the average distances in Ångstroms for the four crystallographically independent
molecules are reported in the figure.
were studied by R1-MP2 calculations. It revealed no energy differ-
ence between the two tautomeric forms (keto/enol) for the
hydroxypyrimidine chemotype, but predicted that the enol form
of the hydroxyisoquinolines were ꢀ3 kcal more stable than the
keto form.12 While not definitive, enhanced glucuronidation of
the hydroxyisoquinoline analogs may have resulted in attenuated
exposures in the pharmacokinetic studies. To address this glucu-
ronidation issue, both the N and O-methylated analogs of 2a were
prepared.13 Unfortunately, 11b-HSD1 inhibitory activity was dra-
matically attenuated for both analogs. This loss of activity can be
explained by the X-ray structure of a close analog, 13d, bound to
the human 11b-HSD1 enzyme (Fig. 2).14 The structure shows a crit-
ical H-bonding interaction between the OH pharmacophore of 13d
and the side chains of S170 and Y183 residues at the catalytic site
of the enzyme. A methyl group on either the nitrogen or oxygen
atom of the hydroxyisoquinoline pharmacophore would likely dis-
rupt this binding interaction, thus leading to a loss of activity.
Also potentially contributing to a lack of in vivo efficacy by these
compounds is their reduced ability to inhibit the enzyme in whole
cell assays. Despite relatively robust inhibition of both the mouse
and human 11b-HSD1 enzymes in isolated microsomal fractions,
these compounds exhibited significantly reduced potency (50–
500-fold) in the murine adipose 3T3L1 and human derived HEK
Attenuated exposures observed in the mouse pharmacokinetic
studies might in part be attributed to glucuronidation of the hydro-
xyl group in these analogs. In vitro stability studies11 in the pres-
ence of uridine 50-diphospho-glucuronosyltransferase (UGT)
revealed that all these analogs, with the exception of hydroxypy-
rimidine 1, were rapidly glucuronidated. The data is summarized
in Table 4. The relative stabilities of the ketone versus enol tauto-
meric forms of the hydroxypyrimidine and hydroxyisoquinolines
Table 4
Pharmacokinetic exposure, in vitro glucuronidation half-life studies, and cellular assays of representative compounds
Cmax/Tmaxa ((
l
M)/h)
AUC(0-8) ha
(l
M h)
UGT T (min)
HEK IC50 (nM)
3T3L1 mIC50 (nM)
b
Cmpd
½
1
7.1/1
.03/1
0.5/4
2.9/4
2.3/4
37
0.2
2.9
15
12
>200
<1
3
1102
78
270
332
396
2854
3508
3966
2254
2162
2a
13d
20f
21
NTc
5
a
b
c
Pharmacokinetic studies were performed in mice (po dosing) using 0.5% methocel, and 0.1% Tween 80 in water as vehicle. The dose was 10 lmol/kg.
Mouse UGT.
NT = not tested.

6698
S. C. Wu et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6693–6698
6. 11b-HSD1 microsomes isolated from HEK 293 cells over-expressing human
11b-HSD1 were incubated with the substrate cortisone and cofactor NADPH at
room temperature. The reactions were terminated with the addition of a
nonspecific 11b-HSD1 inhibitor (18b-glycyrrhetinic acid). The product, cortisol
was quantified in an immuno-competition SPA wherein the [3H]-cortisol
bound to anti-rabbit antibody Yttrium silicate SPA beads coated with
polyclonal anti-cortisol antibody was competed by cortisol produced in the
reaction and the reaction mixture was read in a scintillation plate reader
(TopCount). The IC50 value was then determined by amount of cortisol formed
from a cortisol standard curve.
cells. The reason for reduced cellular activity at present is unclear
since pampa and caco-2 permeability studies do not suggest re-
duced potential for permeability (data not shown). This attribute al-
most certainly contributed to the poor response in vivo and may be
an endemic property of this particular chemotype.
In summary, we have discovered a structurally unique series of
hydroxyisoquinolines as novel 11b-HSD1 inhibitors. These com-
pounds in general exhibited good cross species activity (human,
mouse, cynomolgus monkey) along with other desirable develop-
ment attributes such as weak CYP450 inhibition, no PXR activation,
and little ion channel activity. We were able to improve the meta-
bolic stability and aqueous solubility for a number of analogs (e.g.,
13d). The primary challenge for this series of compounds is their
propensity to undergo rapid glucuronidation in liver microsomes
and their weak activity in cellular assays. These issues required fur-
ther investigation and the results will be published in due course.
7.
5
l
L of 300
added to 450
microsomal protein in the reaction buffer (100 mM NaPO4, pH 7.4, 5 mM
MgCl2). Following 5 min pre-incubation at 37 °C, 50 L of 10 mM NADPH was
l
M compound in acetonitrile: DMSO (98.6:1.4 by volume), was
lL of pre-warmed microsomal mixtures, containing 1.1 mg/mL
l
added to the mixtures to initiate metabolic reactions. Reactions were
terminated immediately after NADPH addition (0 min sample) and after
10 min incubation with NADPH (10 min sample), by transferring 150
lL
reaction aliquots into 300 L acetonitrile. Denatured microsomal protein was
l
precipitated by centrifugation. The supernatant was then analyzed using LC–
MS/MS. The percent of parent compound remaining in the reaction mixture
was calculated by comparing LC–MS/MS peak areas from the 10 min samples
to those from the 0 min samples for each compound. The rate of metabolism
was then calculated using the following formula: metabolic rate (nmol/
min mg) = ((1-fract remaining) ꢁ substrate conc. (nmol/mL))/(time inc
(min) ꢁ protein conc. (mg/mL)), where substrate conc. = 3 nmol/mL, prot
conc = 1 mg/mL, and incubation time = 10 min. Values of 0.00 represent no
measurable metabolism over the course of the experiment whereas values of
0.30 represent complete in vitro metabolism.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.09.058.
8. Vial containing sodium phosphate buffer (0.1 M, pH 7.2), test compound
(30 M), human liver microsome (2 mg/mL) and NADPH (2 mM) was incubated
l
References and notes
in a 37 °C water bath for 1 h. Ice cold acetonitrile was added to the incubation
mixture. The mixture was centrifuged and the supernatant was analyzed with
LC/MS/MS. Metabolites were identified based on the observed mass and UV
spectra. Levels of metabolites were estimated based on UV peak area at 320–
330 nm, assuming same extinction coefficient for metabolites and the parent.
9. (a) McKillop, A.; Rao, D. P. Synthesis 1977, 760; (b) Robl, J.A.; Wu, S.C.; Yoon, D.,
WO2008005910, 2008.
10. DIO mice were bled via tail tip in the fed state and dosed orally (7.5 ml/kg) with
vehicle or drug (0.1% Tween 80, 0.5% methocel in water). At 30 min post
dosing, mice were bled via tail tip and dosed orally (7.5 ml/kg) with DHC
10 mg/kg. All animals were bled at 30, 60 and 120 minutes post DHC dosing.
1. (a) Wild, S.; Roglic, G.; Green, A.; Sicree, R.; King, H. Diabetes Care 2004, 27,
1047; (b) Patti, M.-E.; Corvera, S. Endocrine Rev. 2010, 31, 364. and references
cited therein.
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) Fotsch, C.; Wang, M. J. Med. Chem. 2008, 51, 4851; (f) Stewart, P. M.;
Tomlinson, J. W. Diabetes 2009, 58, 14; (g) Boyle, C. D.; Kowalski, T. J. Expert
Opin. Ther. Pat. 2009, 19, 801; (h) Ge, R.; Huang, Y.; Liang, G.; Li, X. Curr. Med.
Chem. 2010, 17, 412; (i) Tiwari, A. IDrugs 2010, 13, 266. 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) Yeh, V. S. C.; Kurukulasuriya, R.; Fung, S.; Monzon, K.; Chiou, W.; Wang, J.;
Stolarik, D.; Imade, H.; Shapiro, R.; Knourek-Segel, V.; Bush, E.; Wilcox, D.;
Nguyen, P. T.; Brune, M.; Jacobson, P.; Link, J. T. Bioorg. Med. Chem. Lett. 2006,
16, 5555; (b) Patel, J. R.; Shuai, Q.; Dinges, J.; Winn, M.; Pliushcher, M.; Fung, S.;
Monzon, K.; Chiou, W.; Wang, J.; Pan, L.; Wagaw, S.; Engstrom, K.; Kerdesky, F.
A.; Longenecker, K.; Judge, R.; Qin, W.; Imade, H. M.; Stolarik, D.; Beno, D. W. A.;
Brune, M.; Chovan, L. E.; Sham, H. L.; Link, J. T. Bioorg. Med. Chem. Lett. 2007, 17,
750; (c) Julian, L. D.; Wang, Z.; Bostick, T.; Caille, S.; Choi, R.; DeGraffenreid, M.;
Di, Y.; He, X.; Hungate, R. W.; Jaen, J. C.; Liu, J.; Monshouwer, M.; McMinn, D.;
Rew, Y.; Sundon, A.; Sun; Daqing, S.; Tu, H.; Ursu, S.; Walker, N.; Yan, X.; Ye, Q.;
Power, J. P. J. Med. Chem. 2008, 51, 3953; (d) Cheng, H.; Hoffman, J.; Le, P.; Nair,
S. K.; Cripps, S.; Matthews, J.; Smith, C.; Yang, M.; Kupchinsky, S.; Dress, K.;
Edwards, M.; Cole, B.; Walters, E.; Loh, C.; Ermolieff, J.; Fanjul, A.; Bhat, G. B.;
Herrera, J.; Pauly, T.; Hosea, N.; Paderes, G.; Rejto, P. Bioorg. Med. Chem. Lett.
2010, 20, 2897.
35 lL of whole blood were collected per time point in microvette tubes coated
with EDTA and kept on ice. Samples were centrifuged at 40 °C for 10 min at
2500 RPM. Plasma corticosterone concentration was measured using EIA. AUC
of corticosterone concentrations were calculated using Graphpad.
11. To a solution of 2
10 L of 200 mM of MgCl2, 40
liver microsome and 2 L of 2.5 mg/mL of alamethicine in water was added
34 L of 30 mM of UDPGA and 66 L of water. The total 200 L solution was
incubated for 0, 5, 15 and 30 min. At each time point, the solution was added
400 L of acetonitrile, which contained 1% of formic acid, to precipitate the
l
L of 1 mM 2–8 AcCN/water solution of the substrate and
l
l
L of 500 mM Tris buffer, 10 L of 20 mg/mL
l
l
l
l
l
l
protein at time. The supernatant was centrifuged and remaining substrate was
determined by LC–MS. T1/2 was determined from the standard curve of the
substrate concentration versus time.
12. A conformational search was first carry out for each compound to locate the
global minimum energy conformation in aqueous phase using the OPLS 2005
force field with mixed-torsional low mode sampling using the Macromodel
program. The RI-MP2 energy was calculated from the minimum-energy
structure using the QChem package version 3.2.
13. The O-methylated analog was prepared by treatment of 2 with MeI and K2CO3.
The N-methylated analog was prepared by the procedure described in
Bassindale, A. R.; Parker, D. J.; Patel, P.; Taylor, P. G. Tetrahedron Lett. 2000,
41, 4933.
5. El-Khawaga, A. M.; El-Naggar, G. M.; Hassan, K. M.; El-Dean, A. M. Phosphorus,
Sulfur Silicon Relat. Elem. 1989, 44, 203.
14. PDB Deposition number is 3TFQ.