Convenient total synthesis of taranabant (MK-0364), a novel cannabinoid-1 receptor inverse agonist as an anti-obesity agent

Article abstract of DOI:10.1016/j.tet.2007.10.056

Being obese has various health problems that are related to type 2 diabetes mellitus, cardiovascular disease, hypertension, hyperlipidemia, and fibrinolytic abnormalities. Merck's taranabant (MK-0364), a CB1R inverse agonist, is currently in Phase 3 clini

Full text of DOI:10.1016/j.tet.2007.10.056

Available online at www.sciencedirect.com
Tetrahedron 63 (2007) 12845–12852
Convenient total synthesis of taranabant (MK-0364), a novel
cannabinoid-1 receptor inverse agonist as an anti-obesity agent
*
Min-ah Kim, Jong Yup Kim, Kwang-Seop Song, Jeongmin Kim and Jinhwa Lee
Small Molecule Group, Central Research Institute, Green Cross Corporation, 341 Bojeong-dong, Giheung-gu, Yongin,
Gyunggi-do 446-799, Republic of Korea
Received 27 September 2007; revised 13 October 2007; accepted 15 October 2007
Available online 17 October 2007
Abstract—Being obese has various health problems that are related to type 2 diabetes mellitus, cardiovascular disease, hypertension, hyper-
lipidemia, and fibrinolytic abnormalities. Merck’s taranabant (MK-0364), a CB1R inverse agonist, is currently in Phase 3 clinical trials, and is
being actively pursued by Merck toward obesity market. Merck intends to file for FDA approval of taranabant in 2008. In order to overcome
practical difficulty involved in lab-scale preparation of taranabant with a dynamic kinetic resolution procedure developed by Merck’s process
group, we developed a ‘user-friendly’ method to install stereogenic centers by adopting Evans asymmetry chemistry. This method allowed us
to prepare readily sub-gram scale of the target compound in a convenient way.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
by the FDA for the approval for psychiatric safety reasons,
although rimonabant has been approved in the Europe for
the treatment of obesity. On the other hand, Merck’s
taranabant (MK-0364), another CB1R inverse agonist, is
currently in Phase 3 clinical trials and Merck intends to
file for FDA approval of taranabant in 2008 (Fig. 1).
In recent years, obesity has become a major health problem
for many postindustrial societies. The number of deaths per
year attributable to obesity is about 30,000 in UK and nearly
400,000 in United States, where obesity is set to overtake
smoking as the main preventable cause of illness and prema-
ture death.1–3 The total direct and indirect costs of obesity
were estimated to be approximately V32,800 million per
year in EU and $99.2 billion per year in USA.3,4 Obesity
poses a major health risk for serious diet-related chronic dis-
ease, including type 2 diabetes, cardiovascular disease, hy-
pertension and stroke, and some of cancers.1 For these
reasons, the World Health Organization declared obesity
a global epidemic5,6 and obesity is now considered as dis-
ease that needs pharmacological treatments.7–9
With our efforts to discover and develop a new medicine
for the treatment of obesity, we needed to prepare up to
gram-scale of the precursors of taranabant or taranabant
itself for our obesity program. Our imminent goal was to
O
S
NH
Cl
O
N
N
N
N
O
Me
N
N
A lot of studies on causes of obesity have tried to identify
new potential targets that could be exploited to create novel
drugs. At last it was discovered that modulation of endocan-
nabinoid system by specifically blocking the cannabinoid re-
ceptor 1 (CB1) in both the brain and periphery can provide
a novel target for the treatment of obesity.10
N
Cl
Cl
Cl
Cl
rimonabant (SR141716)
SLV319
Recently, among arsenal to fight obesity, rimonabant
(SR141716), a selective cannabinoid-1 receptor (CB1R) in-
verse agonist discovered by Sanofi Aventis, was put on hold
O
O
N
NC
Cl
N
H
CF3
Keywords: Taranabant; Anti-obesity; Evans asymmetric reaction; Hydride
reduction.
1, taranabant (MK-0364)
* Corresponding author. Tel.: +82 31 260 9892; fax: +82 31 260 9870;
e-mail: jinhwalee@greencross.com
Figure 1. Structures of CB1R inverse agonists: rimonabant (SR141716),
SLV319, and taranabant (MK-0364).
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.10.056

12846
M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
O
CO2Me
Br
Br
Cl
b
a
CO2Me
Br
2
Cl
d
3
4
Me
Me
NHBoc
e
Br
Cl
OH
Br
Cl
c
5
6
O
Me
O
O
N
O
N
NC
Cl
N
NC
Cl
NH2
f
HO
H
CF3
CF3
8
1, taranabant (MK-0364)
7
Scheme 1. Synthesis of taranabant (MK-0364) by Merck’s medicinal chemistry group. (a) p-Cl-PhCH2Br, KHMDS, ꢀ78 ꢁC to rt. (b) MeONH(Me)$HCl,
Me2AlCl. (c) LiBH(sec-Bu)3, ꢀ78 ꢁC. (d) (i) MsCl, Et3N; (ii) NaN3, DMF; (iii) H2, Boc2O, PtO2. (e) (i) Zn(CN)2, Pd2(dba)3, dppf; (ii) HCl/dioxane. (f) (i)
PyBop, N-methylmorpholine, CH2Cl2; (ii) prep chiral HPLC.
design and develop a convenient, practical, and reproducible
synthetic method for medicinal chemists to utilize and to
provide reasonable amount of intermediates toward tarana-
bant whenever necessity comes along for medicinal chemis-
try purpose. Herein we wish to report an asymmetric
synthesis of taranabant (MK-0364) featuring Evans chiral
auxiliary methodology to install a pivotal benzylic stereo-
genic center.
Although the synthetic route appears to provide small
amounts of related target compounds, it is quite inefficient
and time-consuming to adopt multiple column chromato-
graphy in addition to chiral HPLC for resolution for a reason-
able scale-up. Then the process group at Merck developed
a dynamic kinetic resolution based on a ruthenium-catalyzed
enantioselective reduction to transform a racemic ketone in-
termediate into a single enantiomer as depicted inScheme 2.12
As shown in Scheme 1, the original route to taranabant em-
ployed by medicinal chemistry group at Merck has tedious
column chromatography works for the isolation of interme-
diate alcohol 5 as well as use of preparative chiral HPLC
technique to resolve taranabant (1) at the final stage.11
The process provides a very elegant way to introduce chiral-
ity of taranabant and is amenable to large-scale production.
However, when we actually try to perform the reaction for
medicinal chemistry laboratory scale, we have found it
rather inconvenient to employ super-high pressure of hydro-
gen gas (90–150 psi) in the presence of Noyori’s catalyst
(xyl-BINAP/DAIPEN) or Dow catalyst (xyl-PHANE-
PHOS/DPEN). It is not likely to adopt the dynamic kinetic
resolution technology especially for medicinal chemists
who usually deal with multi-gram scale reaction, but not
up to 100-g scale for their routine laboratory projects. We en-
visioned that the key intermediate, the chiral bromo alcohol
9 in Scheme 3, could be derived from the corresponding chi-
ral bromo ketone 12 upon diastereoselective reduction using
L-Selectride at low temperature.13 In turn, the requisite
O
OH
H2 (90 psi)
(S)-Noyori's catalyst
Br
Cl
Br
Cl
KOtBu, IPA
4
9
Scheme 2. Enantioselective reduction of ketone 4 by Merck’s process chem-
istry group.
O
O
O
O
Br
Cl
N
O
Evans asymmetry reaction
Br
N
O
Ph
11
Ph
10
O
OH
Me
Br
Br
Cl
L-Selectride
Cl
12
9
Scheme 3. Synthesis of (R,S)-bromo alcohol.

M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
12847
(S)-benzyl bromo ketone 12 could be obtained by adopting
Evans chiral auxiliary technology installing the first stereo-
genic center for taranabant.14 We fully recognized the fact
that the first stereogenic center of (S)-benzyl bromo ketone
12 might be vulnerable to potential epimerization along
the synthesis, but we have undertaken the synthetic study
considering potential benefits secured from this route if suc-
cessful: user-friendly synthesis: i.e., any skilled lab person-
nel can synthesize requisite intermediates readily for either
discovery of a new scaffold or structure–activity relationship
study without being intimidated by using high-pressure hy-
drogen gas or chiral HPLC for resolution. Also it would pro-
vide another synthetic choice for rapid access to taranabant,
an anti-obesity agent.
2. Results and discussion
2.1. Total synthesis of taranabant (MK-0364)
We began our synthesis with a classical Evans asymmetric
reaction route as in Scheme 4.14
Figure 2. X-ray structure of compound 11.
Thus, 3-bromophenyl acetic acid 13 was coupled with lithi-
ated (S)-4-benzyloxazolidin-2-one via pivaloyl mixed anhy-
dride prepared from pivaloyl chloride in the presence of
a base such as triethylamine to produce N-acyloxazolidinone
10 in 79% yield. The mixed anhydride method usually pro-
vides the better yield of N-acyloxazolidinone 10 than routine
acyl chloride method. Next, alkylation of 10 using NaHMDS
with 1-(bromomethyl)-4-chlorobenzene at ꢀ78 ꢁC affords
the alkylated product 11 in 72% yield. LiHMDS can be
also adopted as a suitable base for the conversion, but we
found that use of NaHMDS gives more favorable results
the chiral auxiliary of acyloxazolidinone 11 was cleaved
by standard conditions (LiOOH, 0 ꢁC) to produce the corre-
sponding acid 14 uneventfully. The specific optical rotation
of 14 was measured to be +10.5. Then the next three steps to
prepare methyl ketone 12 proceeded in a straightforward
fashion. Thus, treatment of the bromo acid 14 with oxalyl
chloride provided the corresponding acyl chloride, which
was directly reacted with N,O-dimethylhydroxylamine in
the presence of triethylamine to give the Weinreb amide in
91% yield. The resulting Weinreb amide was treated with
MeMgBr to generate the desired methyl ketone 12 in 99%
yield. The transformation from bromo acid 14 to methyl
ketone progressed without any purification by column chro-
matography. We tried to control the neighboring stereochem-
istry by converting the methyl ketone 12 to the alcohol with
hydride reduction. We briefly screened hydride reduction
conditions. The results are summarized in Table 1.
1
consistently. The H NMR spectrum of 11 did not show
any notable isomer peaks, confirming seemingly very high
de value. The relative and absolute stereochemistries of the
desired alkylated product 11 were further secured through
single-crystal X-ray analysis as demonstrated in Figure 2.
Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the Cam-
bridge Crystallographic Data Centre as supplementary pub-
lication no. CCDC661320.
The Li+-containing reducing agents tend to give better dia-
stereoselectivity than Na+- or K+-containing counterparts
(entries 1, 3, and 4). Steric factors appear to influence the
stereoselectivity of reduction. Thus, use of bulky reducing
At this juncture, we began to have confidence in the route,
which could lead to taranabant without tainting potentially
labile stereogenic center of acyloxazolidinone 11. Next,
O
O
O
O
Br
Cl
N
b
O
a
Br
N
O
CO2H
Br
13
Ph
Ph
10
11
O
O
OH
Me
Br
Cl
OH
c
Br
Cl
Br
Cl
d
e
14
12
9
Scheme 4. Preparation of bromo alcohol 9. (a) (i) pivaloyl chloride, Et3N, THF, 0 ꢁC; (ii) n-BuLi, (S)-4-benzyloxazolidin-2-one, THF, ꢀ78 ꢁC. (b) 1-(Bromo-
methyl)-4-chlorobenzene, NaHMDS, THF, ꢀ78 ꢁC. (c) LiOOH, THF, H2O, 0 ꢁC. (d) (i) (COCl)2, CH2Cl2; (ii) MeONHMe, Et3N, CH2Cl2; (iii) MeMgBr, THF.
(e) LiBH(sec-Bu)3, THF, ꢀ78 ꢁC.

12848
M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
Table 1. Diastereoselective reduction of bromo ketone 14
O
OH
Me
OH
Me
Br
Cl
Me
Reducing agents
Br
Cl
Br
Cl
12
9'
9
Entry
Reaction conditions
Ratio (9/90)a
1
2
3
4
5
LiBH4, 0 ꢁC, MeOH
89:11
91:9
80:20
56:44
>99:<1
LiBH4, ꢀ78 ꢁC, THF
NaBH4, 0 ꢁC, MeOH
K-Selectride, ꢀ78 ꢁC, THF
L-Selectride, ꢀ78 ꢁC, THF
a
All the above reaction conditions provide about the same yield (w90% yield) of alcohols 9 and 90, albeit in a different ratio. Selectivity was measured on
GilsonÔ reverse-phase preparative HPLC.
agent such as L-Selectride magnifies such steric factors to
achieve even greater stereoselectivity (entry 2 and 5).15
The specific optical rotation of bromo alcohol 9 thus
obtained was +36.8.
2.2. Analog of taranabant (MK-0364)
In view of the potential of taranabant or related structure as
a lead compound for anti-obesity studies, we sought to pre-
pare a few of scaffolds with similar or hopefully improved
developability characteristics including biological activity.
At the outset, we chose to evaluate the importance of CN
group on taranabant.
As described in Scheme 5, the cyano group was next intro-
duced into bromo alcohol 5 via a Pd-catalyzed cyanation
following Merck’s protocol.12 Thus, we carried out the cyan-
ation of bromide 5 to nitrile 9 under the conditions of using
Zn(CN)2 and in situ generated Pd [P(o-tol)3]4 as a catalyst in
DMF. The yield for the conversion was approximately 70%
in our hands.
Construction of an analog 20 began with bromo alcohol 9.
The bromo alcohol 9 was converted to bromo azide 18
through mesylation followed by displacement with azide
in a similar way previously described toward taranabant.
The Staudinger reaction, followed by purification by
GilsonÔ reverse-phase preparative HPLC using CH3CN/
water containing 0.1% TFA was adopted successfully for
transformation of the azide 18 to amine 19. Finally, cou-
pling of amine TFA salt 19 with acid 8 under conditions
of DMAP, EDC, and HOBt in DMF, followed by isolation
by GilsonÔ preparative HPLC readily produced analog 20
(Scheme 6).
Next, the cyano alcohol 15 was converted to cyano azide 16
through mesylation followed by displacement with azide.
The reactions went smoothly to produce azide 16 in 94%
for two steps. Transformation of the azide 16 to amine 17
was conducted successfully utilizing the Staudinger condi-
tions (PPh3 in mildly heated toluene/water).16 Purification
was performed efficiently by GilsonÔ reverse-phase prepar-
ative HPLC using CH3CN/water system containing 0.1%
TFA. Thus, the desired amine was obtained as a TFA salt
form (17) in 79% yield. The specific optical rotation of 17
was estimated to be +22.6.
2.3. Biological evaluation
Taranabant (MK-0364, 1) and the structurally similar analog
20 were screened via in vitro rat cannabinoid CB-1 binding
assay. Both taranabant (MK-0364) and analog 20 demon-
strated subnanomolar binding affinities. Taranabant turned
out to be slightly less potent than the bromo analog 20
(CB1R IC50¼0.861 nM vs CB1R IC50¼0.583 nM) via in-
house assay. As illustrated, bromophenyl not only retained
Finally, coupling of amine TFA salt 17 with acid 8 under con-
ditions of DMAP, EDC, and HOBt in DMF, followed by iso-
lation by GilsonÔ preparative HPLC produced taranabant
without difficulty. Our synthetic taranabantobtained had iden-
tical 1H NMR as well as LC–MS data reported by Merck.11
The specific optical rotation of 1 was measured to be +35.5.17
OH
N3
OH
Me
a
Br
Cl
Me
b
NC
Cl
NC
Cl
Me
c
9
16
15
Me
NH2
Me
O
O
O
N
TFA
d
NC
Cl
O
N
NC
Cl
N
HO
H
CF3
CF3
8
17
1, taranabant (MK-0364)
Scheme 5. Preparation of taranabant 1 from bromo alcohol 9. (a) 2% cat. Pd(OAc)2, P(o-tol)3, Et2Zn, Zn(CN)2, DMF. (b) (i) MsCl, Et3N; (ii) NaN3, DMF,
120 ꢁC. (c) (i) PPh3, tol/H2O; (ii) 0.1% TFA on GilsonÔ prep HPLC. (d) DMAP, EDC, DMF.

M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
12849
N3
OH
Me
b
a
Br
Cl
Me
Br
Cl
18
9
Me
Me
O
O
O
N
TFA
NH2
c
Br
Cl
O
N
Br
N
HO
H
CF3
CF3
8
20
Cl
19
Scheme 6. Preparation of an analog 20. (a) (i) MsCl, Et3N; (ii) NaN3, DMF, 120 ꢁC. (b) (i) PPh3, tol/H2O; (ii) 0.1% TFA on GilsonÔ prep HPLC. (c) DMAP,
EDC, DMF.
biological activity, but also even improved in vitro CB1
receptor binding affinity. This observation suggests that
the cyano group on phenyl ring may be important, but not
critical for activity. Presumably its function is to improve
physical properties of taranabant for better developability
characteristics.
silica gel 60 (230–400 mesh). Most of the reactions were
monitored by thin layer chromatography on 0.25 mm
E. Merck silica gel plates (60 F254), visualized with UV light
using a 5% ethanolic phosphomolybdic acid or p-anisalde-
hyde solution.
4.1.1. (S)-4-Benzyl-3-(2-(3-bromophenyl)acetyl)oxazo-
lidin-2-one (10). To a stirred solution of 15.5 g (72 mmol)
of 2-(3-bromophenyl)acetic acid (13) in 290 mL of anhy-
drous THF was added 11 mL (79 mmol) of triethylamine un-
der an atmosphere of argon. The mixture was cooled to
ꢀ78 ꢁC, and trimethylacetylchloride (11.5 mL, 94 mmol)
was then added using a cannula. The resulting white suspen-
sion was stirred for 10 min at ꢀ78 ꢁC, 1 h at 0 ꢁC, and re-
cooled to ꢀ78 ꢁC. Meanwhile, in a different flask, a solution
of lithiated S-oxazolidinone was prepared by the dropwise
addition of 29 mL n-butyllithium (2.5 M in hexane) to
a ꢀ78 ꢁC solution of 15.5 g of the auxiliary in 290 mL of an-
hydrous THF. The mixture was stirred for 20 min at ꢀ78 ꢁC.
The lithiated chiral auxiliary was transferred via a cannula
into the reaction flask containing the preformed mixed anhy-
dride at ꢀ78 ꢁC. The mixture was stirred at 0 ꢁC for 1 h and
was allowed to warm to 23 ꢁC in 16 h. The mixture was then
quenched with 200 mL of saturated ammonium chloride so-
lution. THF was evaporated in vacuo. The product was ex-
tracted with (3ꢂ200 mL) ethyl acetate. The organic layer
was washed with 1 N sodium hydroxide (2ꢂ100 mL) and
1 N sodium bisulfate (1ꢂ100 mL), dried by anhydrous mag-
nesium sulfate, filtered, and evaporated. Purification by
silica gel chromatography (elution with 15–30% ethyl
acetate in hexane) gave 21.4 g (79%) of the desired com-
pound as a white solid. 1H NMR (CDCl3, 400 MHz):
d 7.51–7.50 (m, 1H), 7.45–7.42 (m, 1H), 7.33–7.20 (m,
5H), 7.15–7.12 (m, 2H), 4.72–4.66 (m, 1H), 4.33–4.17 (m,
4H), 3.27 (dd, 1H, J¼13.6, 3.6 Hz), 2.78 (dd, 1H, J¼9.8,
13.2 Hz); LC–MS: m/e 374 (M++H).
3. Summary
We have achieved a convenient total synthesis of taranabant
(MK-0364), a novel cannabinoid-1 receptor inverse agonist
for the treatment of obesity by employing Evans asymmetry
reaction. We introduced a pivotal asymmetric center via the
Evans chiral auxiliary methodology and accomplished
subsequent diastereoselective reduction to install adjacent es-
sential asymmetric center for taranabant. Using this synthetic
route we prepared a structurally similar analog, which retains
significant cannabinoid CB-1 receptor binding affinity.
4. Experimental
4.1. General
All references to ether are to diethyl ether; brine refers
to a saturated aqueous solution of NaCl. Unless otherwise
indicated, all temperatures are expressed in ꢁC (degrees cen-
tigrade). All reactions are conducted under an inert atmo-
sphere at room temperature unless otherwise noted, and all
solvents are of the highest available purity unless otherwise
indicated. Microwave reaction was conducted with a Biotage
microwave reactor. 1H NMR spectra were recorded on either
a Jeol ECX-400 or a Bruker DPX-300 spectrometer. Chemi-
cal shifts were expressed in parts per million (ppm, d units).
Coupling constants are in units of hertz (Hz). Splitting pat-
terns describe apparent multiplicities and are designated as
s (singlet), d (doublet), t (triplet), q (quartet), quint (quintet),
m (multiplet), and br (broad). Mass spectra were obtained
with either Micromass, Quattro LC Triple Quadruple Tan-
dem Mass Spectometer, ESI, or Agilent, 1100LC/MSD,
ESI. Optical rotation data were obtained on a JASCO
P-1030 automatic polarimeter. For preparative HPLC, ca.
100 mg of a product was injected in 1 mL of DMSO onto
a SunFireÔ Prep C18 OBD 5 mm 19ꢂ100 mm Column
with a 10 min gradient from 10% CH3CN to 90% CH3CN
in H2O. Flash chromatography was carried out using Merck
4.1.2. (S)-4-Benzyl-3-((S)-2-(3-bromophenyl)-3-(4-chloro-
phenyl)propanoyl)oxazolidin-2-one (11). To a solution
of oxazolidinone 10 (20.4 g, 55 mmol) and 4-chlorobenzyl
bromide (17.5 g, 85 mmol) in 180 mL of anhydrous THF
at ꢀ78 ꢁC was added sodium hexamethyldisilazide (1 M in
THF, 85 mL, 85 mmol). The reaction was allowed to warm
to room temperature overnight. The volatile materials were
removed on a rotary evaporator, and the resulting mixture
was partitioned between saturated ammonium chloride
(200 mL) and ethyl acetate (200 mL). The organic layer

12850
M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
was separated and the aqueous layer was extracted with ethyl
acetate (2ꢂ200 mL). The combined organic extracts were
dried over anhydrous sodium sulfate, filtered, and concen-
trated to dryness. Purification by silica gel chromatography
(elution with 15–30% ethyl acetate in hexane) gave 19.6 g
(t, J¼7.4 Hz), 3.35 (dd, J¼13.9, 7.5 Hz), 2.84 (dd, 1H,
J¼13.9, 7.3 Hz), 2.04 (s, 3H).
4.1.5. (2R,3S)-3-(3-Bromophenyl)-4-(4-chlorophenyl)bu-
tan-2-ol (9). L-SelectrideÒ (1 M solution in THF, 9.5 mL)
was added to a solution of ketone (12, 2.13 g, 6.31 mmol)
in anhydrous THF (32 mL) under N2 at ꢀ78 ꢁC. The mixture
was stirred at ꢀ78 ꢁC for 1.5 h and warmed to ꢀ40 ꢁC. The
aqueous NaOH (3 N, 9.6 mL) and 30% H2O2 (4.9 mL) were
added slowly and stirred vigorously for 2 h at 0 ꢁC. The re-
action mixture was diluted with ethyl acetate and the organic
phase was separated. The organic phase was washed with
water, saturated Na2S2O3, and brine, and dried over
MgSO4. After evaporation, the residue was purified using
column chromatography on silica gel (hexane/ethyl
acetate¼6:1) to give desired products (1.98 g, 92%) as color-
1
(72%) of the desired compound as a white solid. H NMR
(CDCl3, 400 MHz): d 7.60–7.59 (m, 1H), 7.42–7.39 (m,
1H), 7.34–7.31 (m, 1H), 7.27–7.23 (m, 5H), 7.20–7.16 (m,
3H), 6.99–6.96 (m, 2H), 5.40 (dd, 1H, J¼9.6 Hz), 4.61–
4.58 (m, 1H), 4.07–4.06 (m, 2H), 3.46 (dd, 1H, J¼9.6,
13.6 Hz), 3.06–2.98 (m, 2H), 2.61 (dd, 1H, J¼9.0,
13.6 Hz); LC–MS: m/e 498 (M++H).
4.1.3. (S)-2-(3-Bromophenyl)-3-(4-chlorophenyl)prop-
anoic acid (14). A 0.05 M solution of oxazolidinone 11
(19 g, 38 mmol) in a 3:1 mixture of THF and H2O was
treated with 30% H2O (35 mL, 306 mmol), followed by
LiOH (3.2 g, 76 mmol) at 0 ꢁC. The resulting mixture was
stirred and allowed to warm to room temperature overnight.
THF was then removed under vacuum. The residue was
washed with dichloromethane (500 mLꢂ2) to remove (S)-
4-benzyloxazolidin-2-one. The desired product was isolated
by ethyl acetate extraction of the acidified (pH 1–2) aqueous
phase. No further purification was required. Standing under
high vacuum yielded 13 g (98%) of the title compound.
[a]24.7 +10.5 (c 0.815, CHCl3).
1
less oil. H NMR (DMSO-d6, 500 MHz): d 7.39 (m, 1H),
7.30–7.28 (m, 1H), 7.20–7.09 (m, 6H), 4.67 (br, 1H),
3.83–3.76 (m, 1H), 3.02 (dd, 1H, J¼6.0, 13.5 Hz), 2.90–
2.78 (m, 2H), 0.88 (d, 3H, J¼6.5 Hz). 13C NMR (DMSO-
d6, 300 MHz): d 144.8, 140.0, 132.3, 131.3, 130.7, 130.1,
129.3, 129.0, 128.4, 121.5, 68.2, 53.7, 37.2, 21.9. Reduction
of ketone (12) with NaBH4 provided the mixture of diaste-
reomers. Major diastereomer was identified as 9. Minor
diastereomer (90): 1H NMR (DMSO-d6, 500 MHz): d 7.30–
7.29 (m, 2H), 7.19–7.13 (m, 3H), 7.10–7.08 (m, 1H), 6.99 (d,
2H, J¼8.0 Hz), 4.84 (br, 1H), 3.79–3.73 (m, 1H), 3.32–3.28
(m, 1H), 2.83–2.75 (m, 2H), 0.89 (d, 3H, J¼6.0 Hz); [a]24.3
+36.8 (c 0.585, CHCl3).
4.1.4. (S)-3-(3-Bromophenyl)-4-(4-chlorophenyl)butan-
2-one (12). To a solution of acid 14 (2.85 g, 8.39 mmol) in
dichloromethane (25 mL) at 0 ꢁC were added dimethylform-
amide (one drop) and oxalyl chloride (1.5 mL, 16 mol)
dropwise. The reaction was allowed to warm to room tem-
perature for 2 h and concentrated to dryness to give the crude
acyl chloride, which was used without further purification.
To a solution of the acyl chloride in dichloromethane
(40 mL) were added N-methoxy-N-methylamine hydrochlo-
ride (2.5 g, 25 mmol) and triethylamine (5.8 mL, 42 mmol)
at 0 ꢁC. After stirring at room temperature for 4 h, the reac-
tion mixture was diluted with ether (100 mL) and succes-
sively washed with water, dilute aqueous sodium hydrogen
sulfate, and brine, dried over anhydrous magnesium sulfate,
filtered, and concentrated to dryness to afford (S)-2-(3-bro-
mophenyl)-3-(4-chlorophenyl)-N-methoxy-N-methylpropan-
amide (91%), which was used without further purification.
LC–MS: m/e 382 (M++H).
4.1.6. 3-((2S,3R)-1-(4-Chlorophenyl)-3-hydroxybutan-2-
yl)benzonitrile (15). Zn(CN)2 (212 mg, 1.80 mol) was
added to a solution of bromo alcohol (9, 1.02 g, 3.00 mmol)
and DMF (4 mL). The atmosphere of the mixture was then
exchanged to nitrogen gas for 20 min, and the batch was
heated to 56 ꢁC. In a separate flask, DMF (3 mL) was charged
and exchanged to nitrogen gas. Pd(OAc)2 (13.5 mg,
0.06 mmol) and P(o-tol)3 (75.3 mg, 0.24 mmol) were
charged to this flask, and solution was heated to 56 ꢁC over
40 min and held for an additional 15 min. ZnEt2
(0.082 mL, 0.09 mmol) was then added to the catalyst solu-
tion over 15 min, and the resulting slurry was aged at 56 ꢁC
for 1 h. The Zn(CN)2 mixture was transferred to the catalyst
slurry over 1 h at 56 ꢁC, and the mixture was aged an
additional 2 h. The reaction was cooled with an ice bath to
5–10 ꢁC. Concentrated NH4OH (5 mL) was added, keeping
the batch temperature at <30 ꢁC, and the batch was aged
for 1 h. The slurry was then filtered over a pad of silica gel
that was wetted with toluene. The cake was washed with
20 L of toluene. To the filtrate were added 3 mL of 20%
aqueous NH4OH and toluene (3 mL). After mixing well,
the layers were separated, and the organic layer was washed
with 10 mL each of 15% aqueous NaCl solution and water.
This solution of alcohol 15 (71% yield) was carried forward
to the mesylation reaction. 1H NMR (DMSO-d6, 400 MHz):
d 7.65 (s, 1H), 7.59 (d, 1H, J¼7.6 Hz), 7.50 (d, 1H,
J¼7.6 Hz), 7.38 (t, 1H, J¼7.6 Hz), 7.19 (d, 2H, J¼8.0 Hz),
7.09 (d, 2H, J¼8.0 Hz), 4.70 (d, 1H, J¼4.4 Hz), 3.85–3.83
(m, 1H), 3.06–3.01 (m, 1H), 2.95–2.87 (m, 2H), 0.90 (d,
3H, J¼6.4 Hz).
To a solution of (S)-2-(3-bromophenyl)-3-(4-chlorophenyl)-
N-methoxy-N-methylpropanamide (10.2 g, 27 mmol) in
anhydrous tetrahydrofuran (95 mL) at 0 ꢁC was added meth-
ylmagnesium bromide (3 M in ether, 18 mL, 53 mmol). Af-
ter stirring at 0 ꢁC for 2 h, the reaction was quenched with
methanol (4 mL) and 2 M hydrochloric acid (40 mL). The
volatile materials were removed by concentrating on a rotary
evaporator and the residue was partitioned between saturated
ammonium chloride (150 mL) and ether (150 mL). The
organic layer was separated, and the aqueous layer was
extracted with ether (2ꢂ150 mL). The combined organic
extracts were dried over anhydrous magnesium sulfate, fil-
tered, and concentrated to dryness to afford (S)-3-(3-bromo-
phenyl)-4-(4-chlorophenyl)butan-2-one (99%), which was
used without further purification. 1H NMR (CDCl3,
300 MHz): d 7.42–7.39 (m, 1H), 7.33–7.32 (m, 1H), 7.21–
7.16 (m, 3H), 7.09–7.06 (m, 1H), 7.00–6.95 (m, 2H), 3.82
4.1.7. 3-((2S,3S)-3-Azido-1-(4-chlorophenyl)butan-2-yl)-
benzonitrile (16). To a solution of alcohol 15 (1.7 g,

M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
12851
5.9 mmol) in ethyl acetate (16 mL) at 0 ꢁC were added
triethylamine (1.1 mL, 7.6 mmol) and methanesulfonyl
chloride (0.6 mL, 7.6 mmol). After stirring at 0 ꢁC for
1.5 h, the reaction was quenched by addition of saturated
aqueous sodium bicarbonate (15 mL). After stirring at
room temperature for 1 h, the organic layer was separated,
dried over anhydrous sodium sulfate, filtered, and concen-
trated to dryness to afford (2R,3S)-4-(4-chlorophenyl)-3-
(3-cyanophenyl)butan-2-yl methanesulfonate, which was
used without further purification. 1H NMR (CDCl3,
400 MHz): d 7.55–7.52 (m, 1H), 7.45 (m, 1H), 7.43–7.39
(m, 2H), 7.18–7.14 (m, 2H), 6.99–6.94 (m, 2H), 5.08–
5.03 (m, 1H), 3.2–3.0 (m, 3H), 2.89 (s, 3H), 1.33 (d, 3H,
J¼6.4 Hz).
J¼8.4 Hz), 6.86 (d, 1H, J¼8.8 Hz), 6.72 (d, 2H, J¼8.4 Hz),
5.83 (d, 1H, J¼8.8 Hz), 4.38–4.28 (m, 1H), 3.11 (dd, 1H,
J¼3.2, 12.8 Hz), 2.87–2.74 (m, 2H), 1.75 (s, 3H), 1.71 (s,
3H), 0.87 (d, 3H, J¼6.8 Hz); LC–MS: m/e 516 (M++H);
[a]25.0 +35.5 (c 0.455, CHCl3).
4.1.10. N-((2S,3S)-4-(4-Chlorophenyl)-3-(3-bromophe-
nyl)butan-2-yl)-2-methyl-2-(5-(trifluoromethyl) pyridin-
2-yloxy)propanamide (20). The title compound was
prepared according to the similar procedures described
above. Purification by preparative HPLC yielded the title
1
compound. H NMR (CDCl3, 400 MHz): d 8.35 (s, 1H),
7.81 (dd, 1H, J¼2.0, 8.4 Hz), 7.30 (d, 1H, J¼7.6 Hz), 7.16
(t, 1H, J¼2.0 Hz), 7.10–7.03 (m, 3H), 6.89 (d, 1H,
J¼7.6 Hz), 6.84 (d, 1H, J¼8.0 Hz), 6.77 (d, 2H, J¼8.8 Hz),
5.81 (d, 1H, J¼9.2 Hz), 4.32–4.26 (m, 1H), 3.03 (d, 1H,
J¼12 Hz), 2.84–2.75 (m, 2H), 1.73 (s, 3H), 1.71 (s, 3H),
0.88 (d, 3H, J¼6.8 Hz); LC–MS: m/e 569 (M++H).
To a solution of (2R,3S)-4-(4-chlorophenyl)-3-(3-cyanophe-
nyl)butan-2-yl methanesulfonate in dimethylformamide
(5 mL) was added sodium azide (1.9 g, 29 mmol). After stir-
ring at 120 ꢁC for 2 h, the reaction mixture was poured into
water (20 mL), and the product was extracted with ether
(2ꢂ25 mL). The combined organic extracts were washed
with water, dried over anhydrous magnesium sulfate, fil-
tered, and concentrated to dryness. The residue was purified
on a silica gel column to yield azide 16 (94% for two steps),
which was used without further purification. 1H NMR
(CDCl3, 400 MHz): d 7.50 (d, 1H, J¼10.4 Hz), 7.40–7.33
(m, 2H), 7.26–7.22 (m, 1H), 7.14–7.11 (d, J¼8.4 Hz), 6.82
(d, 2H, J¼8.4 Hz), 3.72–3.65 (m, 1H), 3.29 (dd, 1H,
J¼2.4, 12 Hz), 2.86–2.76 (m, 2H), 1.15 (d, 3H, J¼6.4 Hz);
LC–MS: m/e 285 (M++H).
4.2. Pharmacological test: in vitro activity analysis
The compounds of the present invention were analyzed
for their binding characteristics for CB1 and CB2 and the
pharmacological activity thereof in accordance with the
method disclosed in Ref. 18. The analysis was performed
using [3H]CP-55940, which is a selectively radioactivity-
labeled 5-(1,1-dimethyheptyl)-2[5-hydroxy-2-(3-hydroxy-
propyl)-cyclohexyl]-phenol, purchased from Perkin–Elmer
Life Sciences, Inc. (Boston, Massachusetts, USA), through
a rat CB-1 receptor binding protocol as follows.
4.1.8. 3-((2S,3S)-3-Amino-1-(4-chlorophenyl)butan-2-
yl)benzonitrile trifluoroacetic acid (17). To the toluene
(2 mL) solution of azide 16 (430 mg, 1.38 mmol) from the
previous step was added water (1 mL), and the batch was
heated to 70 ꢁC. A solution of PPh3 (543 mg, 2.08 mol) in
toluene (1 mL) was added to the batch for over 1 h (slowly
in order to control nitrogen evolution). The batch was aged
for an additional 7 h and then cooled to ambient temperature.
Purification by GilsonÔ HPLC system (elution with 0.1%
TFA of H2O and acetonitrile) gave 438 mg (79%) of the
The tissue obtained from the brain of SD rats was homoge-
nized with a Dounce homogenate system in TME (50 mM
Tris, 3 mM MgCl2 and 1 mM EDTA, pH 7.4) at 4 ꢁC, and
the homogenate was centrifuged at 48,000g for 30 min at
4 ꢁC. The pellet was re-suspended in 5 mL of TME and the
suspension was divided into aliquots and stored at ꢀ70 ꢁC
until its use in the following assay.
Test compound (2 ml) was diluted in dimethylsulfoxide and
added to a deep well of a polypropylene plate, to which 50 ml
of [3H]CP-55940 diluted in a ligand buffer solution (0.1%
bovine serum albumin (BAS)+TME) was added. The tissue
concentrations were determined by Bradford protein analy-
sis, and 148 ml of brain tissue of the required concentration
was added to the plate. The plate was covered and placed
in a 30 ꢁC incubator for 60 min, and then transformed on
GF/B filtermat pretreated in polyethylenimine (PEI) using
a cell harvester. Each filter was washed five times and dried
at 60 ꢁC for 1 h. Then, the degree of radioactivity retained by
the filter was measured using Wallac MicrobetaÔ (Perkin–
Elmer Life Sciences, Inc., Massachusetts, USA) and the
activity of the compound for inhibiting CB1 receptor was
determined therefrom.
1
TFA salt form. H NMR (CD3OD, 400 MHz): d 7.58 (m,
1H), 7.54 (m, 1H), 7.45 (m, 2H), 7.14 (d, 2H, J¼8.4 Hz),
6.96 (d, 2H, J¼8.4 Hz), 3.67–3.64 (m, 1H), 3.22–3.15 (m,
1H), 2.98–2.92 (m, 1H), 1.16 (d, 3H, J¼6.8 Hz); LC–MS:
m/e 285 (M++H), [a]24.4 +22.6 (c 0.790, CHCl3).
4.1.9. N-((2S,3S)-4-(4-Chlorophenyl)-3-(3-cyanophenyl)-
butan-2-yl)-2-methyl-2-(5-(trifluoromethyl)pyridin-2-
yloxy)propanamide (taranabant, MK-0364, 1). To a
solution of 2-methyl-2-(5-(trifluoromethyl)pyridin-2-yloxy)-
propanoic acid11 (8, 410 mg, 1.65 mmol) in dichlorome-
thane(5 mL)wereaddedaminesalt(17, 438 mg, 1.10 mmol),
DMAP (268 mg, 2.20 mmol), and then EDC (253 mg,
1.32 mmol). The mixture was stirred at room temperature
overnight. The product was extracted with dichloromethane
(2ꢂ10 mL). The combined organic extracts were washed
with water, dried over anhydrous magnesium sulfate, filtered,
and concentrated to dryness. The residue was purified on Gil-
sonÔ HPLC system to yield taranabant (MK-0364: 560 mg,
Acknowledgements
Financial support was provided by Green Cross Corporation.
We thank Professor Eun Lee (Seoul National University) for
allowing us to use JASCO P-1030 automatic polarimeter for
optical rotation data. We thank Mr. Sung-Han Lee (GCC
Central Research Institute) for the biological assays of
1
0.77 mmol). H NMR (CDCl3, 400 MHz): d 8.35 (s, 1H),
7.83 (dd, 1H, J¼2.4, 8.8 Hz), 7.45 (d, 1H, J¼7.6 Hz), 7.31
(t, 1H, J¼8.0 Hz), 7.25–7.21 (m, 2H), 7.08 (d, 2H,

12852
M.-ah Kim et al. / Tetrahedron 63 (2007) 12845–12852
10. Pagotto, U.; Vicennati, V.; Pasquali, R. Ann. Med. 2005, 37,
270–275.
taranabant and its analog 20. In addition, we thank Dr.
Hoseop Yun (Ajou University) for solving the X-ray struc-
ture of compound 11. Finally, we would like to acknowledge
Dr. Chong-Hwan Chang for his inspiration to create small
molecule programs at GCC.
11. Lin, L. S.; Lanza, T. J.; Jewell, J. P.; Liu, P.; Shah, S. K.; Qi, H.;
Tong, X.; Wang, J.; Xu, S. S.; Fong, T. M.; Shen, C.-P.; Lao, J.;
Xiao, J. C.; Shearman, L. P.; Stribling, D. S.; Rosko, K.; Strack,
A.; Marsh, D. J.; Feng, Y.; Kumar, S.; Samuel, K.; Yin, W.; Van
der Ploeg, L. H. T.; Goulet, M. T.; Hagmann, W. K. J. Med.
Chem. 2006, 49, 7584–7587.
Supplementary data
12. Chen, C.-y.; Frey, L. F.; Shultz, S.; Wallace, D. J.; Marcantonio,
K.; Payack, J. F.; Vazquez, E.; Springfield, S. A.; Zhou, G.; Liu,
P.; Kieczykowski, G. R.; Chen, A. M.; Phenix, B. D.; Singh, U.;
Strine, J.; Izzo, B.; Krska, S. W. Org. Process Res. Dev. 2007,
11, 616–623.
Supplementary data associated with this article can be found
in the online version, at doi:10.1016/j.tet.2007.10.056.
References and notes
13. Iwasawa, Y.; Shibata, J.; Nonoshita, K.; Arai, S.; Masaki, H.;
Tomimoto, K. Tetrahedron 1996, 52, 13881–13894.
14. Dharanipragada, R.; VanHulle, K.; Bannister, A.; Bear, S.;
Kennedy, L.; Hruby, V. J. Tetrahedron 1992, 48, 4733–4748.
15. Yamamoto, Y.; Matsuoka, K.; Nemoto, H. J. Am. Chem. Soc.
1988, 110, 4475–4476.
1. Patel, P. N.; Pathak, R. Am. J. Health Syst. Pharm. 2007, 64,
481–489.
2. Bray, G. A. J. Med. Chem. 2006, 49, 4001–4007.
3. Haslam, D. W.; James, W. P. T. Lancet 2005, 366, 1197–1209.
4. Fry, J.; Finley, W. Proc. Nutr. Soc. 2005, 64, 359–362.
5. Report of a WHO Consultation on Obesity: Obesity-Preventing
and Managing a Global Epidemic; World Health Organization:
Geneva, 1997.
6. Antel, J.; Gregory, P. C.; Nordheim, U. J. Med. Chem. 2006, 49,
4008–4016.
7. Aronne, L. J.; Thornton-Jones, Z. D. Clin. Pharmacol. Ther.
2007, 81, 748–752.
16. Yokokawa, F.; Asano, T.; Shioiri, T. Org. Lett. 2000, 2, 4169–
4172.
17. As far as we are concerned, Merck has not published the spe-
cific optical rotation data of taranabant. However, we believe
that our compound should be just the identical with Merck’s
based on comparison of NMR data in addition to rat canna-
binoid CB1 receptor binding affinity data.11
8. Padwal, R. S.; Majumdar, S. R. Lancet 2007, 369, 71–77.
9. Bishop, M. J. J. Med. Chem. 2006, 49, 3999–4000.
18. Devane, W. A.; Dysarz, F. A., 3rd; Johnson, M. R.; Melvin,
L. S.; Howlett, A. C. Mol. Pharmacol. 1998, 34, 605–613.