The discovery of potent inhibitors of aldosterone synthase that exhibit selectivity over 11-β-hydroxylase

Article abstract of DOI:10.1016/j.bmcl.2010.06.086

Aldosterone, the final component of the renin-angiotensin-aldosterone system, plays an important role in the pathophysiology of hypertension and congestive heart failure. Aldosterone synthase (CYP11B2) catalyzes the last three steps of aldosterone biosynthesis, and as such appears to be a target for the treatment of these disorders. A sulfonamide-imidazole scaffold has proven to be a potent inhibitor of CYP11B2. Furthermore, this scaffold can achieve high levels of selectivity for CYP11B2 over CYP11B1, a key enzyme in the biosynthesis of cortisol.

Full text of DOI:10.1016/j.bmcl.2010.06.086

Bioorganic & Medicinal Chemistry Letters 20 (2010) 4324–4327
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of potent inhibitors of aldosterone synthase that exhibit
selectivity over 11-b-hydroxylase
b
b
a
a
b
Christopher M. Adams a, , Chii-Whei Hu , Arco Y. Jeng , Rajeshri Karki , Gary Ksander , Dan LaSala ,
Jennifer Leung-Chu b, Guiqing Liang c, Qian Liu a, Erik Meredith a, Chang Rao a, Dean F. Rigel b, Jie Shi a,
Sherri Smith c, Clayton Springer a, Chun Zhang a
*
a Global Discovery Chemistry, Novartis Institutes for BioMedical Research, 250 Massachusetts Ave., Cambridge, MA 02139, USA
b Cardiovascular and Metabolism Disease Area, Novartis Institutes for BioMedical Research, One Health Plaza, East Hanover, NJ 07936, USA
c Metabolism and Pharmacokinetics, Novartis Institutes for BioMedical Research, 250 Massachusetts Ave., Cambridge, MA 02139, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Aldosterone, the final component of the renin–angiotensin–aldosterone system, plays an important role
in the pathophysiology of hypertension and congestive heart failure. Aldosterone synthase (CYP11B2)
catalyzes the last three steps of aldosterone biosynthesis, and as such appears to be a target for the treat-
ment of these disorders. A sulfonamide–imidazole scaffold has proven to be a potent inhibitor of
CYP11B2. Furthermore, this scaffold can achieve high levels of selectivity for CYP11B2 over CYP11B1, a
key enzyme in the biosynthesis of cortisol.
Received 11 April 2010
Revised 12 June 2010
Accepted 15 June 2010
Available online 19 June 2010
Keywords:
CYP11B2
Ó 2010 Elsevier Ltd. All rights reserved.
Aldosterone
CYP11B1
Aldosterone synthase inhibitor
Hypertension.
The renin–angiotensin–aldosterone system (RAAS) is a key reg-
ulator of blood pressure and extracellular volume. Aldosterone, the
final component of the RAAS, is a potent mineralocorticoid. Ele-
vated aldosterone levels play an important role in the pathophys-
iology of hypertension and congestive heart failure.1 As such,
aldosterone has attracted much attention from both the pharma-
ceutical and clinical communities. Until recently, the majority of
this effort has focused on mineralocorticoid receptor (MR) antago-
nists, such as spironolactone and eplerenone.2,3 However, current
evidence indicates that the deleterious effects of elevated aldoste-
rone levels are not solely mediated by the MR and that non-geno-
mic effects may play a critical role.4,5 Therefore, directly inhibiting
the synthesis of aldosterone may prove advantageous.6
Subsequent studies have shown that FAD286 is also an inhibitor
of 11-b-hydroxylase (CYP11B1). CYP11B1 is also expressed in the
adrenal gland and catalyzes the last step in the biosynthesis of cor-
tisol, the primary glucocorticoid.7 Cortisol mediates the stress re-
sponse affecting energy mobilization and the immune system.
Therefore,
a selective CYP11B2 inhibitor may be desirable.
FAD286 exhibits IC50 values of 1.6 and 9.9 nM in recombinant hu-
man CYP11B2 and CYP11B1 enzyme assays, respectively.10 This
small degree of selectivity is not surprising, considering that the
amino acid sequences of these two enzymes are 95% identical in
the coding regions.11 Homology models of the two enzymes
The final three steps of aldosterone biosynthesis are catalyzed
by a single cytochrome P450 enzyme, CYP11B2, which is expressed
in zona glomerulosa cells of the adrenal gland.7 Importantly, fadroz-
ole, an inhibitor of CYP11B2, has been shown to reduce aldosterone
levels in humans.8 Fadrozole was initially developed as an aroma-
tase (CYP19) inhibitor, but it was later discovered that the R-enan-
tiomer of fadrozole (FAD286, Fig. 1) is a potent aldosterone
synthase inhibitor.9
N
N
N
N
N
O
CN
FAD286
1
CN
* Corresponding author. Tel.: +1 617 871 7724.
E-mail address: christopherM.adams@novartis.com (C.M. Adams).
Figure 1. The R-enantiomer of fadrozole (FAD286) and a representative analog
from the lactam series (1).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.06.086

C. M. Adams et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4324–4327
4325
sulfonamide analogs with the aim of greatly improving selectivity
for CYP11B2 over CYP11B1.
Sulfonamide 2 (Table 1), one of the earliest molecules synthe-
sized in this series, displays modest CYP11B1/CYP11B2 selectivity.
However, it is also plagued by poor solubility (<0.5
abolic stability in human liver microsomes (HLM, t1/2 <5 min),16
and activity against CYP19 (39% inhibition at 1
M).17 In an at-
lM), low met-
l
tempt to improve upon this profile we initially explored the space
around the nitrogen of the sulfonamide. From a synthetic perspec-
tive this required a four-step sequence starting from commercially
available aldehyde 3 (Scheme 1).
A series of reductive aminations with commercially available
primary amines permitted access to a diverse array of amino-imi-
dazoles. Sulfonylation of the resulting secondary amines with chlo-
romethanesulfonyl chloride, followed by microwave-mediated
cyclization afforded sulfonamides of type 6 in modest yields. Alkyl-
ation with the appropriate alkyl halide or epoxide completes the
R
a
O
N
H
Tr
N
N Tr
Figure 2. Homology models of CYP11B1 (pink) and CYP11B2 (cyan). Active sites
depicted; residues that lie within 10 Å of heme, and are different between CYP11B1
and CYP11B2, are labeled.
N
R
N
NH2
3
4
R
O
provided further evidence that achieving high levels of selectivity
would be challenging.12 Our modeling indicated that only four res-
idues differed between the two enzymes within 10 Å of the heme,
the presumed substrate binding site (Fig. 2).
Having recognized the ability of FAD286 to inhibit CYP11B2, we
began a program to discover novel analogs of this compound with
improved properties. Early efforts focused on modifying the satu-
rated portion of the fused ring system of fadrozole. One promising
series, which contains a lactam moiety exemplified by 113,14
(Fig. 1), was found to have excellent potency against CYP11B2
(IC50: 40 nM).15 However, compounds from this series were also
potent CYP11B1 inhibitors (e.g., IC50 = 2 nM for 1).
N
S
b
c
N
Tr
N
O
O O
S
Cl
d
Cl
5
Cl
R
R
N
N
N
N
O S
O
N
O S
O
N
X
2
R
2
R
6
7
Scheme 1. Reagents and conditions: (a) NaBH(OAc)3, DCM, rt; (b) DIPEA, DCM,
À78 °C; (c) DMF, EtOH, DIPEA (3:1:1), microwave irradiation at 200 °C for 1 h,
10–30% yield for the three steps; (d) KHMDS, THF, À78 °C to rt 10–70%.
The potency of the lactam series led us to explore a bioisosteric
sulfonamide series. This Letter describes the development of cyclic
Table 1
Exploration of the space around the sulfonamide nitrogen
R
N
N
O S
O
N
hCYP11B2 IC50
(nM) (Ref. 10)
,
hCYP11B1 IC50
(nM) (Ref. 10)
,
hCYP19 % Inhib
HLM t½, (min)
(Ref. 15)
@ 1
lM (Ref. 16)
2, R = 4-F-Ph-CH2–
8, R = 4-Cl-Ph-CH2–
9, R = Ph-CH2–
10, R = 4-CF3O-Ph-CH2–
11,19R = 4-CF3-Ph-CH2–
12, R = c-C3H5–
2.5
5.2
26.1
50.2
103
544
566
930
122
39
51
14
13
0
5
8
3.
18.6
104
121
>1000
758
7
13
>405
55
9
19
13, R = c-C3H5-CH2–
S
14, R=
117
628
55.7
119
13
71
2
15, R=
19
Cl
16, R=
Cl
9.4
76.0
204
56
24
6
5
17, R=
Cl
61.7

4326
C. M. Adams et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4324–4327
synthesis. The resulting stereoisomers could be readily separated
via HPLC employing a chiral stationary phase.
groups (e.g., 22) reintroduced CYP19 activity and did not address
the solubility concerns. The greatest levels of selectivity were
achieved by alkylating intermediate 6 with styrene epoxide. Reac-
tion with (R)-styrene epoxide furnished 24, which was greater than
600-fold selective for CYP11B2 over CYP11B1, and exhibited negligi-
ble CYP19 inhibition.
Despite the selectivity achieved over CYP11B1 and CYP19, this
particular scaffold exhibited strong inhibition of CYP3A4 in vitro.
Inhibition of CYP3A4-mediated oxidation of both testosterone
and midazolam is observed (Table 3).20 The SAR relating to midaz-
olam competition is particularly flat. Modifications to the northern
hemisphere, including replacement of the benzyl moiety with a
thiophene (e.g., 14) failed to improve selectivity over CYP3A4. Add-
ing polarity to the southern hemisphere also proved ineffective
(e.g., 18, 19).
With this synthesis in hand, we generally found that one enan-
tiomer proved to be more potent against CYP11B2 and/or more
selective.14 We also determined that benzylic moieties are pre-
ferred and that halogen substitution at the para position affords
maximal CYP11B2 potency and selectivity over CYP11B1 (e.g., 2
and 8, Table 1). Unfortunately, changes in this region were not able
to significantly improve metabolic stability without dramatic de-
creases in potency. Of particular note are compounds 12 and 13,
which exhibit vastly improved metabolic stability, but only weakly
inhibit CYP11B2. Based on the superior metabolic stability of com-
pounds 12 and 13, we theorized that the benzylic position of com-
pounds such as 8 may be metabolically labile. This was confirmed
by metabolite identification studies.18 We next synthesized com-
pounds 15–1713 with the intention of decreasing the rate of ben-
zylic oxidation (Table 1). Only 15 showed any improvement in
metabolic stability, and only 16 maintained single digit nanomolar
potency against CYP11B2.
Although attenuating potency against CYP3A4 has been a chal-
lenge for this particular scaffold, this work further6 demonstrates
that outstanding selectivities over CYP11B1 can be achieved, in
spite of its very high homology with CYP11B2.
Surprised by the lack of improvement in metabolic stability,
we initiated further metabolite identification studies with 16.
These studies indicated that the installation of the benzylic
methyl group did in fact block metabolism at this site, but that
the isobutyl group was now the major site of metabolism. We
therefore began exploring replacements for the isobutyl group.
We anticipated that incorporating polar functional groups in this
region may not only serve to slow metabolism, but would also
improve solubility and attenuate CYP19 activity. To our delight,
not only did this strategy lead to improvements in all of these
areas, but they also greatly improved selectivity over CYP11B1
(Table 2).
Table 3
CYP3A4 assay results against both midazolam and testosterone
Compounds
hCYP11B2
IC50, (nM)
CYP3A4 (m)a
IC50, (nM)
CYP3A4 (t)b
IC50, (nM)
8
9
5.2
<68
100
<68
<68
<68
<68
<68
<68
<68
<68
<68
<68
100
<68
87
18.6
104
121
117
9.4
61.7
11.4
7.8
0.7
0.8
1.3
2.4
1400
490
280
1160
<68
NA
<68
<68
530
120
180
900
1920
10
11
14
16
17
18
19
20
21
22
23
24
The hydroxyl of 18 specifically blocked the most likely site of
metabolism, resulting in an improvement in the HLM t1/2 to greater
than 300 min. In addition, the added polarity improved solubility to
500 lM (pH 6.8), abrogated CYP19 activity, and afforded CYP11B1/
CYP11B2 selectivity of 42-fold while maintaining adequate potency.
Ethers, such as 20and 21, also improvedCYP11B1/CYP11B2 selectiv-
ity but failed to provide significant benefits with respect to meta-
bolic stability. Incorporating more metabolically stable alkyl
0.3
a
m = Midazolam.
t = Testosterone; see Ref. 19 for assay details.
b
Table 2
Incorporating polarity into the southern hemisphere
Cl
hCYP11B2 IC50
(nM) (Ref. 10)
,
hCYP11B1 IC50
(nM) (Ref. 10)
,
hCYP19 % Inhib
HLM t½, (min)
(Ref. 15)
@ 1
lM (Ref. 16)
N
O S
O
N
N
R
18, R = –CH2C(CH3)2OH
11.4
7.8
471
340
4
>300
76
19, R=
O
35
Me
20, R = –CH2C(CH3)2OMe
21, R = –CH2-O-CH(CH3)2
22, R = –CH2-c-C3H5
0.9
0.8
1.3
20.6
222
245
17
23
39
12
5
25
23, R=
OH
Ph
OH
Ph
2.4
0.3
234
182
6
6
24
15
24, R=

C. M. Adams et al. / Bioorg. Med. Chem. Lett. 20 (2010) 4324–4327
4327
construction was done using Prime version 1.5, Schrödinger, LLC, New York,
NY. In the generated models, the loops were first refined using the PRIME
structure refinement tool followed by stepwise minimization using the OPLS
forcefield. The parameters for heme were generated using the hetgrp_ffgen
utility available from Schrödinger.
Acknowledgements
We thank Dr. Bing Wang, and Karl Gunderson for assistance in
securing and interpreting high-field NMR spectra, and Patrick
Cronan for assistance in obtaining high resolution mass spectra.
13. All compounds depicted are single isomers (i.e., >98% de and/or >98% ee)
unless otherwise noted. The separation of stereoisomers was accomplished via
HPLC, employing columns with chiral stationary phases.
14. Typically, the enantiomeric configuration was not determined. The data
reported in this manuscript relates to the enantiomer with the preferred
in vitro profile.
15. (a) McKenna, J. Preparation of imidazopyrazines and imidazodiazepines as
agents for the treatment of aldosterone mediated conditions. WO
2004014914A1.; (b) A more detailed discussion of the development of this
series will be reported in due course.
References and notes
1. Williams, J. S.; Williams, G. H. J. Clin. Endocrinol. Metab. 2003, 88, 2364.
2. (a) Hu, X.; Bolton, C. W. Drug Dev. Res. 2006, 67, 871; (b) Pratt-Ubunama, M. N.;
Nishizaka, M. K.; Calhoun, D. A. Curr. Hypertens. Rep. 2005, 7, 186.
3. Two clinical trials, RALES and EPHESUS have shown these agents to be
beneficial for the treatment of heart failure; see: (a) Pitt, B.; Zannad, F.; Remme,
W. J.; Cody, R.; Castaigne, A.; Perez, A.; Palensky, J.; Wittes, J. N. Engl. J. Med.
1999, 341, 709; (b) Pitt, B.; Remme, W. J.; Zannad, F.; Neaton, J.; Martinez, F.;
Roniker, B.; Bittman, R.; Hurley, S.; Kleiman, J.; Gatlin, M. N. Engl. J. Med. 2003,
348, 1309.
16. Human Liver Microsomal Stability Assay: The half life (t1/2) in human liver
microsomes (HLM) (BD Gentest, Woburn, MA) was determined based on
published method (vide infra). In brief, 1
pooled HLM (containing 0.5 mg microsomal protein per mL) in the presence of
25 g of alamethacin per mg protein, 1.0 mM NADPH, 1.0 mM UDPGA, and
lM test compound was incubated in
l
4. Wehling, M. J. Mol. Med. 1995, 73, 439.
5. Kushibiki, M.; Yamada, M.; Oikawa, K.; Tomita, H.; Osanai, T.; Okumura, K. Eur.
J. Pharmacol. 2007, 572, 182.
2 mM MgCl2. The half-life (t1/2) was derived by monitoring the disappearance
of test compound over a short period of time (e.g., 30 min). See: (a) Obach, R. S.
Drug Metab. Dispos. 1999, 27, 1350; (b) Kalvass, J. C.; Tess, D. A.; Giragossian, C.;
Linhares, M. C.; Maurer, T. S. Drug Metab. Dispos. 2001, 29, 1332; (c) Chauret, N.;
Gauthier, A.; Nicoll-Griffith, D. A. Drug Metab. Dispos. 1997, 26, 1; (d) Fisher, M.
B.; Campanale, K.; Ackermann, B. L.; VandenBranden, M.; Wrighton, S. A. Drug
Metab. Dispos. 2000, 28, 560.
6. This approach has also been pursued by Hartmann and co-workers; see: (a)
Hartmann, R. W.; Muller, U.; Ehmer, P. B. Eur. J. Med. Chem. 2003, 38, 363; (b)
Ulmschneider, S.; Negri, M.; Voets, M.; Hartmann, R. W. Bioorg. Med. Chem. Lett.
2006, 16, 2; (c) Voets, M.; Antes, I.; Scherer, C.; Müller-Viera, U.; Biemel, K.;
Marchais-Oberwinkler, S.; Hartmann, R. W. J. Med. Chem. 2006, 49, 2222; (d)
Lucas, S.; Heim, R.; Ries, C.; Schewe, K. E.; Birk, B.; Hartmann, R. W. J. Med.
Chem. 2008, 51, 8077; (e) Ries, C.; Lucas, S.; Heim, R.; Birk, B.; Hartmann, R. W. J.
Steroid Biochem. Mol. Biol. 2009, 116, 121. and references therein.
7. (a) Rainey, W. E. Mol. Cell Endrocrinol. 1999, 151, 151. and references therein;
(b) Bureik, M.; Lisurek, M.; Bernhardt, R. Biol. Chem. 2002, 383, 1537.
8. Trunet, P. F.; Mueller, P.; Girard, F.; Aupetit, B.; Bhatnagar, A. S.; Zognbi, F.;
Ezzet, F.; Menard, J. J. Clin. Endocrinol. Metab. 1992, 74, 571.
17. The human CYP19 assay was performed according to the method of: Stresser,
D. M.; Turner, S. D.; McNamara, J.; Stocker, P.; Miller, V. P.; Crespi, C. L.; Patten,
C. J. Anal. Biochem. 2000, 284, 427.
18. Metabolite identification: Test articles were incubated in pooled human liver
microsomes (BD Gentest, Woburn, MA) in the presence of cofactors, such as
NADPH (oxidative metabolism) and UDPGA (conjugated metabolism).
Following in vitro incubation and protein precipitation, samples were
analyzed by LC/MS/MS for detection and characterization of the metabolites
formed.
9. Ménard, J.; Pascoe, L. J. Hypertens. 2006, 24, 993.
10. LaSala, D.; Jeng, A.; Shibanaka, Y. Anal. Biochem. 2009, 394, 56.
11. Mornet, E.; Dupont, J.; Vitek, A.; White, P. C. J. Biol. Chem. 1989, 264, 20961.
12. The human CYP11B1 and CYP11B2 homology models were constructed using
19. This compound was profiled as the racemate.
20. Bell, L.; Bickford, S.; Nguyen, P. H.; Wang, J.; He, T.; Zhang, B.; Friche, Y.;
Zimmerlin, A.; Urban, L.; Bojanic, D. J. Biomol. Screen. 2008, 13, 343.
crystal structure of CYP3A4 (PDB code 1TQN) as
a template. Model