Synthesis and evaluation of the metabolites of AMG 221, a clinical candidate for the treatment of type 2 diabetes

Article abstract of DOI:10.1021/ml2001467

All eight of the major active metabolites of (S)-2-((1S,2S,4R)-bicyclo[2.2. 1]heptan-2-ylamino)-5-isopropyl-5-methylthiazol-4(5H)-one (AMG 221, compound 1), an inhibitor of 11β-hydroxysteroid dehydrogenase type 1 that has entered the clinic for the treatm

Full text of DOI:10.1021/ml2001467

LETTER
pubs.acs.org/acsmedchemlett
Synthesis and Evaluation of the Metabolites of AMG 221, a Clinical
Candidate for the Treatment of Type 2 Diabetes
Aiwen Li,† Chester C. Yuan,*,† David Chow,† Michelle Chen,‡ Maurice G. Emery,§ Clarence Hale,‡
Xiping Zhang,§ Raju Subramanian,§ David J. St. Jean, Jr.,† Renee Komorowski,‡ Murielle Vꢀeniant,‡
Minghan Wang,‡ and Christopher Fotsch†
†Departments of Chemistry Research and Discovery, ‡Metabolic Disorders, and §Pharmacokinetics and Drug Metabolism, Amgen Inc.,
One Amgen Center Drive, Thousand Oaks, California 91320, United States
S Supporting Information
b
ABSTRACT: All eight of the major active metabolites of (S)-
2-((1S,2S,4R)-bicyclo[2.2.1]heptan-2-ylamino)-5-isopropyl-5-
methylthiazol-4(5H)-one (AMG 221, compound 1), an inhibitor
of 11β-hydroxysteroid dehydrogenase type 1 that has entered the
clinic for the treatment of type 2 diabetes, were synthetically
prepared and confirmed by comparison with samples generated
in liver microsomes. After further profiling, we determined that metabolite 2 was equipotent to 1 on human 11β-HSD1 and had
lower in vivo clearance and higher bioavailability in rat and mouse. Compound 2 was advanced into a pharmacodynamic model in
mouse where it inhibited adipose 11β-HSD1 activity.
KEYWORDS: AMG 221, 11β-HSD1 inhibitors, metabolites, clinical candidate
ince the publication of the Food and Drug Administration's
Guidance for Industry: Safety Testing of Drug Metabolites in
fully characterize these compounds and further confirm their
structures.
S
2008 (FDA MIST guidelines),1 there has been a renewed focus
on studying the metabolites of clinical candidates during the drug
development process throughout the pharmaceutical industry.2 While
the FDA MIST guidelines recommend that certain metabolites
be subjected to additional safety studies, metabolites of clinical
candidates are also needed for pharmacokinetic and pharmaco-
logical studies. In fact, there are examples where a metabolite,
after being profiled further, replaced the parent compound as
the drug candidate.3,4 Therefore, the development of synthetic
routes to supply sufficient amounts of metabolites for their com-
plete characterization has become an important objective for
many drug discovery and development programs.
Turning first to the synthesis of the metabolites formed on the
norbornyl ring (2À5), we devised a route (Scheme 1) that
allowed us to prepare three of these from a common intermediate
16. We envisioned using chiral norbornyl diol 10,7 as an
intermediate for the preparation of analogs 2À4. Preparation
of norbornyl diol 10 by Hiyashi's method7 followed by mono-
benzylation delivered compound 11. Compound 11 was treated
with tert-butylchlorodiphenylsilane in DMF to give compound
12, which was then debenzylated by hydrogenation to give
compound 13. The chirality of the hydroxyl group was inverted
through a two step sequence by oxidation to the ketone, followed
by reduction with L-selectride to give compound 14. A Mitsunobu
reaction with phthalimide followed by deprotection with hydra-
zine gave a primary amine.8 Addition of benzoyl isothiocyanate
to the amine followed by hydrolysis produced thiourea 15.
Heating thiourea 15 with ethyl 2-bromo-isovalerate in a micro-
wave reactor delivered 5-isopropylthiazolone. Alkylation at the
5-position of the thiazolone was achieved with LDA and methyl
iodide, and after removal of the TBDPS protecting group, the
desired thiazolone 16 was obtained as a mixture of two diaster-
eomers. The metabolite 2 was obtained through chiral super-
critical fluid chromatography (SFC) separation of compound
16. The exo-hydroxyl group of 16 was converted to the endo-
hydroxyl group through oxidation and stereoselective reduc-
tion with L-selectride,9 followed by chiral SFC separation to give
Recently, we reported on the discovery of AMG 221 (1), an
inhibitor of 11β-hydroxysteroid dehydrogenase type 1 (11β-
HSD1) that has entered clinical trials for the treatment of type 2
diabetes.5,6 During our development efforts, we discovered that 1
was metabolized into eight common oxidation products among
rat, mice, dog, monkey, and human (2À9) (Chart 1). The
norbornyl and isopropyl groups of 1 underwent metabolic
oxidation to give hydroxylation products 2, 3, 5À8, ketone 4,
and alkene 9. For structure elucidation (see the Supporting
Information for more details), these metabolites were isolated
from in vitro dog/human liver microsomal incubations supple-
mented with NADPH or isolated from rat urine obtained
following a single dose administration of AMG 221 (1). How-
ever, the bioactivity and pharmacokinetics of these metabolites
could not be determined with the small amount of material
generated from microsomes. Therefore, we set out to prepare
these metabolites by synthetic methods so that we might more
Received: June 18, 2011
Accepted: September 13, 2011
Published: September 13, 2011
r
2011 American Chemical Society
824
dx.doi.org/10.1021/ml2001467 ACS Med. Chem. Lett. 2011, 2, 824–827
|

ACS Medicinal Chemistry Letters
LETTER
metabolite 3. Metabolite 4 was isolated after the oxidation of 16
and chiral separation.
Metabolite 5 was prepared by the route shown in Scheme 2.
Compound 1710 was converted to 18 through alkylation of the
enolate with isopropyl iodide. Hydroboration and oxidation of
compound 18 gave metabolite 5 after chiral SFC.
Chart 1. Structures of 1 (AMG 221) and Its Metabolitesa
The metabolites 6 and 7 were obtained by the route shown in
Scheme 3. The thiazolone 1910 was alkylated with LDA and 2-(2-
bromopropoxy)tetrahydro-2H-pyran, and then, the THP group
was removed to give compound 20 as a mixture of four dia-
stereomers. After chiral SFC separation, 6 and 7 were obtained
Treating thiazolone 19 with LDA and acetone at low tem-
perature gave compound 21 as a mixture of two diastereomers.
Upon chiral separation, metabolite 8 was obtained. When com-
pound 21 was treated with DAST at room temperature, the
fluorinated product 22 was obtained as the major product11
along with a small amount of the elimination product 23,
which was further purified by chiral chromatography to give
metabolite 9.
With enough quantities of all identified metabolites in hand,
the in vitro potencies and preclinical pharmacokinetic properties
were measured (see Tables 1 and 2). We were intrigued to find
that all metabolites tested were active in the human 11β-HSD1
biochemical assay. The in vivo pharmacokinetics of some of the
more potent metabolites (2, 3, 4, and 8) were assessed in rats,
with compound 2 having the lowest clearance and highest
bioavailability of the metabolites tested. Compound 2 was also
tested in mouse pharmacokinetic studies, and as was found in rat,
Scheme 2. Preparation of Compound 5a
a (a) (i-Pr)2NLi, 2-iodopropane, THF, À78 °C, 71%. (b) (1) BH3-THF,
THF, 0 °C; (2) H2O2, NaOH, 65 °C, 37% (2 steps). (c) Chiral
chromatography.
a*The stereochemistry is arbitrarily assigned.
Scheme 1. Preparation of Compounds 2À4a
a Reagents and conditions: (a) NaOH, BnBr, 15-Crown-5, THF, 10 °C, 43%. (b) TBDMSCl, imidazole, DMF, room temperature, 96%. (c) H2, 10%
PdÀC, CH3OH, room temperature, quant. (d) PCC, CH2Cl2, 0 °C, 98% (crude). (e) L-Selectride, THF, À78 °C, 76%. (f) DEAD, Ph3P, phthalimide,
THF, room temperature, 68%. (g) (1) Hydrazine, ethanol, reflux; (2) benzoyl isothiocyanate, CHCl3, room temperature; (3) K2CO3, CH3OH, room
temperature, 74% (3 steps). (h) Ethyl 2-bromoisovalerate, (i-Pr)2EtN, ethanol, microwave, 150 °C, 87%. (i) (i-Pr)2NLi, CH3I, THF, À78 °C, 74%.
(j) TBAF, THF, 0 °C, 98%. (k) PCC, CH2Cl2, 0 °C, 87%. (l) Chiral chromatography, 36%. (m) L-Selectride, THF, À78 °C, 87%. (n) Chiral
chromatography, 41%. (o) Chiral chromatography, 48%.
825
dx.doi.org/10.1021/ml2001467 |ACS Med. Chem. Lett. 2011, 2, 824–827

ACS Medicinal Chemistry Letters
LETTER
Scheme 3. Preparation of Compounds 6À9a
a (a) (i-Pr)2NLi, CH3CHBrCH2OTHP, THF, À78 °C. (b) p-TsOH, CH3OH, room temperature, 13% (2 steps). (c) Chiral chromatography, 34% (6),
8% (7). (d) (i-Pr)2NLi, acetone, THF, À78 °C, 30%. (e) Chiral chromatography, 41%. (f) DAST, CH2Cl2, room temperature. (g) Chiral
chromatography, 3.5% (two steps). *The stereochemistry is arbitrarily assigned.
Table 1. Potency of Compounds Against 11β-HSD1 of Dif-
Table 3. Pharmacokinetics of Compounds 1 and 2 in Micea
ferent Species
male CD1 mouse
human
mouse
compd
1
2
SPAaKi (
cellb IC50
SEMc (nM)
(
SPAaKi (
SEMc (nM)
CL (L/h/kg)
Vss (L/kg)
t1/2 (h)
3.31
0.90
3.32
31
1.48
1.43
no.
SEMc (nM)
1
2
3
4
5
6
7
8
9
12.8 ( 9.1
16.5 ( 4.8
17.9 ( 5.0
23.5 ( 11.8
449 ( 134
37.3 ( 9.6
27 ( 1.4
10.1 ( 6.2
12.7 ( 10.2
15.0 ( 12.6
10.2 ( 2.4
237 ( 181
71.6 ( 34.7
87
70.8 ( 51.1
152 ( 80
181 ( 88
230 ( 20
5.72
F (%)
>100
a iv dose and formulation: 2 mg/kg in 20% captisol, 80% water for 1, and
10% dimethylacetamide, 23% water, 67% PEG 400 for 2; oral dose and
formulation: 10 mg/kg in 10% Pluronic F108 90% OraPlus for 1, 0.1%
Tween-80, and 0.5%CMC in water for 2.
4 h for the 10 mg/kg group, and at 2, 4, 6, and 8 h for the 30 mg/
kg group. The magnitude of inhibition for the 10 mg/kg group
was 38% at 2 h and 31% at 4 h. For the 30 mg/kg group, the
magnitude of inhibition was 51% for 2 h, 46% for 4 h, and ca. 36%
for 6 and 8 h. The free plasma concentrations of 2 that achieved
46% inhibition of 11β-HSD1 in the adipose at 4 h was 2.1 μM
(mouse plasma protein binding = 40.9%), a value 14-fold higher
than the mouse 11β-HSD1 Ki. In contrast, compound 1 achieved
47% inhibition of 11β-HSD1 in the fat at 4 h5 with a free plasma
concentration of 0.24 μM (mouse plasma protein binding =
90.1%), a value just 3-fold higher than the mouse 11β-HSD1 Ki.
One possible explanation for why metabolite 2 may require
higher free plasma concentrations than compound 1 to achieve a
similar pharmacodynamic is that 2 is less lipophilic and may not
penetrate adipose tissue as well. Indeed, at the 4 h time point, the
fat to plasma ratio was 0.27 for 2 (30 mg/kg), whereas the fat to
plasma ratio was 1.34 for 1 (50 mg/kg). However, because
metabolite 2 does show some inhibition of 11β-HSD1in the fat
and because compound 2 has lower in vivo clearance in the
mouse, these data suggest that 2 may contribute to pharmaco-
dynamic effects observed with compound 1.
17.5 ( 7.7
6.0 ( 1.4
21.3 ( 20
16
a Ki was derived from a SPA using the species 11β-HSD1 expressed in
E. coli cells and H-cortisone as the substrate. b IC50 was determined
3
from a whole cell assay using CHO cells overexpressing human 11β-
HSD1. c SEM for at least two independent determinations.
Table 2. Pharmacokinetics of Selected Compounds in Male
SpragueÀDawley Ratsa
compd
1
2
3
8
CL (L/h/kg)
Vss (L/kg)
t1/2 (h)
1.16
1.53
12.18b
1.53
34
0.44
0.97
2.17
2.22
58
0.74
1.43
2.41
1.94
NA
1.92
0.53
0.24
0.28
53
MRT
F (%)
a iv dose and formulation: (2 mg/kg) in DMSO; oral dose and
formulation: 10 mg/kg, 0.1% Tween-80, 0.5% CMC, and 99.4% water.
b Terminal t1/2
.
In summary, we have successfully prepared and profiled all
eight metabolites identified from compound 1. These synthetic
compounds were found to match the isolated metabolites in the
coelution experiments, respectively. Therefore, the chemical
structures of these metabolites were further confirmed (see the
Supporting Information). These metabolites all showed activity
in the 11β-HSD1 biochemical assay with metabolites 2, 3, and 8
it had a lower clearance and higher bioavailability than com-
pound 1 (Table 3).
Compound 2 was then tested in a mouse pharmacodynamic
(PD) model measuring the 11β-HSD1 activity in the adipose
ex vivo (see Figure 1).5 In a timeÀcourse study at 10 and 30 mg/
kg (po), compound 2 significantly inhibited 11β-HSD1 at 2 and
826
dx.doi.org/10.1021/ml2001467 |ACS Med. Chem. Lett. 2011, 2, 824–827

ACS Medicinal Chemistry Letters
LETTER
Figure 1. Inhibition of 11β-HSD1 activity with 2 (dosed PO) in the mesenteric fat of mouse.
having biochemical potency similar to parent compound 1. Four
metabolites, 2À4 and 8, were assessed in rat PK studies, and
metabolites 2 and 3 had in vivo clearance less than parent
compound 1. Because compound 2 had the lowest in vivo
clearance and was nearly equipotent to 1, it was further evaluated
in mouse studies. Metabolite 2 had superior mouse PK to 1 and
inhibited adipose 11β-HSD1 in a mouse PD study, suggesting
that metabolite 2 may contribute to PD effects seen with 1. These
results have compelled our clinical team to explore how the
metabolites of 1 may contribute to human PD. These results will
be reported in the near future.
Type 1 Inhibitor for Clinical Study: Identification of (S)-2-((1S,2S,4R)-
Bicyclo[2.2.1]heptan-2-ylamino)-5-isopropyl-5-methylthiazol-4(5H)-one
(AMG 221). J. Med. Chem. 2010, 53, 4481–4487.
(6) Gibbs, J. P.; Emery, M. G.; McCaffery, I.; Smith, B.; Gibbs, M. A.;
Akrami, A.; Rossi, J.; Paweletz, K.; Gastonguay, M. R.; Bautista, E.;
Wang, M.; Perfetti, R.; Daniels, O. Population Pharmacokinetic/Phar-
macodynamic Model of Subcutaneous Adipose 11β-Hydroxysteroid
Dehydrogenase Type 1 (11β-HSD1) Activity After Oral Administration
of AMG 221, a Selective 11β-HSD1 Inhibitor. J. Clin. Pharmacol. 2011,
51, 830–841.
(7) Hayashi, T.; Ueyama, K.; Tokunaga, N.; Yoshida, K. A Chiral
Chelating Diene as a New Type of Chiral Ligand for Transition Metal
Catalysts: Its Preparation and Use for the Rhodium-Catalyzed Asym-
metric 1,4-Addition. J. Am. Chem. Soc. 2003, 125, 11508–11509.
(8) Sen, S. E.; Roach, S. L. A Convenient Two-Step Procedure for
the Synthesis of Substituted Allylic Amines from Allylic Alcohols.
Synthesis 1995, 756–758.
(9) Mayo, P.; Tam, W. Ring-opening metathesisÀcross-metathesis
reactions (ROM-CM) of substituted norbornadienes and norbornenes.
Tetrahedron 2002, 58, 9513–9525.
’ ASSOCIATED CONTENT
S
Supporting Information. Details about the structural
b
elucidation and experimentals for the preparation of all eight
metabolites. This material is available free of charge via the
Internet at http://pubs.acs.org.
(10) Moniz, G. A.; Frizzle, M. J.; Bernard, C.; Martinelli, M.; Faul,
M.; Nani, R.; Larrow, J. Process for making 2-aminothiazolines. PCT Int.
Appl. WO 2009002445 A1, 68 pp.
(11) Kim, K.-Y.; Kim, B. C.; Lee, H. B.; Shin, H. Nucleophilic Fluo-
rination of Triflates by Tetrabutylammonium Bifluoride. J. Org. Chem.
2008, 73, 8106–8108.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: cyuan@amgen.com.
’ REFERENCES
(1) http://www.fda.gov/downloads/Drugs/GuidanceCompliance-
RegulatoryInformation/Guidances/ucm079266.pdf.
(2) Nedderman, A. N. R. Metabolites in safety testing: Metabolite
identification strategies in discovery and development. Biopharm. Drug
Dispos. 2009, 30, 153–162.
(3) Fura, A. Role of pharmacologically active metabolites in drug
discovery and development. Drug Discovery Today 2006, 11, 133–142.
(4) Fura, A.; Shu, Y.-Z.; Zhu, M.; Hanson, R. L.; Roongta, V.;
Humphreys, W. G. Discovering Drugs through Biological Transforma-
tion: Role of Pharmacologically Active Metabolites in Drug Discovery.
J. Med. Chem. 2004, 47, 4339–4351.
(5) Vꢀeniant, 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., Jr.; Johansson, L.;
ꢀ
Henriksson, M. A.; Williams, M.; Vallgarda, J.; Fotsch, C.; Wang, M.
Discovery of a Potent, Orally Active 11β-Hydroxysteroid Dehydrogenase
827
dx.doi.org/10.1021/ml2001467 |ACS Med. Chem. Lett. 2011, 2, 824–827