Formal Synthesis of Ezetimibe Using a Proline-mediated, Asymmetric, Three-component Mannich Reaction

Article abstract of DOI:10.1246/cl.150916

The formal total synthesis of ezetimibe was accomplished using a proline-mediated, asymmetric, three-component Mannich reaction as the key step. The two stereogenic centers on the β-lactam skeleton of ezetimibe were controlled by the syn-selective asymmetric Mannich reaction, followed by isomerization.

Full text of DOI:10.1246/cl.150916

CL-150916
Received: October 1, 2015 | Accepted: October 22, 2015 | Web Released: January 5, 2016
Formal Synthesis of Ezetimibe Using a Proline-mediated,
Asymmetric, Three-component Mannich Reaction
Yasuharu Shimasaki, Seitaro Koshino, and Yujiro Hayashi*
Department of Chemistry, Graduate School of Science, Tohoku University, Aoba-ku, Sendai, Miyagi 980-8578
(E-mail: yhayashi@m.tohoku.ac.jp)
The formal total synthesis of ezetimibe was accomplished
using proline-mediated, asymmetric, three-component
Mannich reaction as the key step. The two stereogenic centers
on the β-lactam skeleton of ezetimibe were controlled by the
syn-selective asymmetric Mannich reaction, followed by isomer-
ization.
OH
OBn
OH
a
O
known
3R 4S
Step C
S
HO
N
F
N
O
1) epimerization
2) oxidation or
hydrolysis
O
F
2
ezetimibe (1)
F
OBn
OBn
Step A
Step B
Ezetimibe (ZETIAμ) (1) is a drug that shows strong
cholesterol inhibitory activity by blocking the transport of
cholesterol across the intestinal wall.1 1 has been commercial-
ized for the treatment of hyperlipidemia. The annual worldwide
sales of ZETIAμ in 2013 were $2.7 billion as well as $1.6
billion for VYTORINμ (a combination drug of ezetimibe and
simvastatin).2
R
O
R
HN
OH
N
O
lactamization
1) proline-mediated
Mannich reaction
2) oxidation
F
F
3 R = CH2CO2Me
4 R = CH2CH2OTHP
5 R = CH=CH2
6 R = CH2CO2Me
7 R = CH2CH2OTHP
8 R = CH=CH2
OBn
F
1 possesses a β-lactam skeleton and three stereogenic
centers, including two contiguous chiral centers on the β-lactam
ring. Various syntheses of 1 have been reported,3-6 and the key
issue for the synthesis of 1 is how to establish two contiguous
stereogenic centers on the β-lactam. Most of the previous
syntheses employed a chiral auxiliary-based strategy4 or a chiral
pool-derived strategy5 to control the stereogenic centers at C3
and C4, or separation of the desired enantiomer by chromatog-
raphy using a chiral phase.6 However, if the stereogenic centers
of the β-lactam ring of 1 can be controlled using a chiral catalyst,
this can become a powerful method to synthesize 1. There is
only one catalytic method for the synthesis of the β-lactam ring
of 1, viz. that by Chmielewski and co-workers using an (R,R)-
Cr(salen) catalyst.3
We reported a proline-mediated, asymmetric, three-compo-
nent, cross-Mannich reaction in 2003,7 wherein a three-compo-
nent, one-pot reaction of two different aldehydes and p-anisidine
proceeded in an excellent syn-selective and enantioselective
manner (eq 1). If this method can be applied to the synthesis of
the β-lactam core of 1, an efficient and eco-friendly synthetic
method of 1 would be established.
R
O
CHO
NH2
9
R = CH2CO2Me
10 R = CH2CH2OTHP
11 R = CH=CH2
12
13
Scheme 1. Retrosynthetic analysis of ezetimibe (1).
lactamization from carboxylic acids 6, 7, and 8. The cis-lactam
precursors 6, 7, and 8 would be prepared via an L-proline-
mediated, asymmetric, cross-Mannich reaction,7,8 followed by
oxidation.
One of the key reactions is the proline-mediated Mannich
reaction, but the effect of the substituent on aniline is hardly
known from the literatures.7,8 Generally, the electron-rich p-
methoxyphenyl substituent has been employed. Therefore, it is
unclear whether the Mannich reaction with 12 possessing an
electron-deficient p-fluorophenyl substituent would proceed
successfully.
Herein, we report the efficient synthesis of the known key
intermediate 2.
Step A: Mannich reaction and oxidation step: In all cases of
aldehydes 9, 10, and 11, proline-mediated cross-Mannich
reaction in N-methylpyrrolidone (NMP) proceeded successfully
to afford the desired aldehydes (Table 1). We confirmed that in
the case of the electron-deficient p-fluorophenyl substituent, the
asymmetric Mannich reaction proceeded smoothly. The obtained
aldehydes were unstable; hence, they were immediately purified
through a short silica gel column and subjected to Pinnick-
Kraus oxidation.9 Even when 2-methyl-2-butene was used as a
scavenger for chlorine, which is produced during the Pinnick-
Kraus oxidation step, some by-product that was chlorinated at
the p-fluoroaniline moiety was obtained. For this reason, we
added the more electron-rich N,N-dimethylaniline as a scav-
enger, which could completely suppress the side reaction to
provide the desired carboxylic acids 6, 7, and 8.
OMe
OMe
1) 10 mol%
O
L
-proline
HN
R1
R2
R1CHO
ð1Þ
H
OH
2) NaBH4
NH2
R2
syn:anti > 95:5, 71~ >99% ee
The retrosynthesis of 1 is outlined in Scheme 1. The
synthetic route from 2 to 1 has already been reported by
Schering-Plough Co.4a and Dr. Reddy’s Laboratories Ltd.4e The
trans-lactam 2 could be obtained via epimerization and func-
tional group transformation of the side chain of cis-lactams 3, 4,
and 5. We chose -CH2CO2Me, -CH2CH2OTHP, and -CH=CH2
as the R substituents of lactams 3, 4, and 5, which can be
easily converted to a carboxylic acid moiety by oxidation or
hydrolysis. cis-Lactams 3, 4, and 5 could be synthesized by
30 | Chem. Lett. 2016, 45, 30–32 | doi:10.1246/cl.150916
© 2016 The Chemical Society of Japan

Table 1. Synthesis of trans-lactams 14, 15, and 16
OBn
OBn
Step A
OBn
Step C
Step B
F
OBn
1) 15 mol%
L
-proline
C
tBuOK
THF, rt
NMP, -20 o
R
SOCl2, Et3N
CH2Cl2, rt
R
R
O
R
N
N
HN
OH
O
O
O
2) NaClO2, NaH2PO4
N,N-dimethylaniline
tBuOH/H2O, 0 o
NH2
CHO
F
F
F
C
12
13
6 R = CH2CO2Me
7 R = CH2CH2OTHP
8 R = CH=CH2
3 R = CH2CO2Me
4 R = CH2CH2OTHP
5 R = CH=CH2
14 R = CH2CO2Me
15 R = CH2CH2OTHP
16 R = CH=CH2
9
R = CH2CO2Me
10 R = CH2CH2OTHP
11 R = CH=CH2
Step Aa
Step Bc
Step Cd
Entry Aldehyde Time/h Yield/%f syn/antig ee/%h SMe Time/h Yield/%j cis/transk SMe Time/h Yield/%j cis/transk
1
2
3b
9
10
11
26
20
41
55
49
67
84/16
71/29
92/8
n.d.i
n.d.i
82
6
7
8
3
5
3
71
43
99
89/11
65/35
>95/5l
3
4m
5
®
0.5
1.5
0
53
82
®
23/77
22/78
aMannich reaction: 4-(benzyloxy)benzaldehyde (1.0 equiv), p-fluoroaniline (1.1 equiv), L-proline (15 mol %) in NMP (0.5 M) at rt for 10-24 h; aldehyde
(3.0 equiv) at ¹20 °C for the indicated time; then the obtained β-amino aldehyde was immediately purified through a short column of silica gel. Pinnick-
Kraus oxidation: NaClO2, NaH2PO4, N,N-dimethylaniline, t-BuOH/H2O at 0 °C. See the Supporting Information for details. bMannich reaction: 4-
benzyloxybenzaldehyde (1.0 equiv), p-fluoroaniline (1.1 equiv), L-proline (15 mol %) in i-PrOH (1.4 M) at 55 °C for 12 h; concentration of reaction
mixture; then aldehyde (3.0 equiv) at ¹20 °C for the indicated time; the obtained β-amino aldehyde was immediately purified through a short column of
c
d
silica gel. Lactamization: SOCl2, Et3N in CH2Cl2 at room temperature. See the Supporting Information for details. Epimerization: t-BuOK, MS3¡ in
THF at room temperature. eSM: starting molecule. fYield of isolated β-amino carboxylic acid as a mixture of diastereomers. gDetermined by analysis by
1H NMR spectroscopy. Determined by HPLC analysis on a chiral phase after NaBH4 reduction. See the Supporting Information for details. n.d.: not
determined. Yield of isolated β-lactam as a mixture of cis/trans isomers. Determined by analysis by H NMR spectroscopy. ee value of trans-isomer
was 60%. Determined by HPLC analysis on a chiral phase. mEpimerization: t-BuOK (1.0 equiv), MS3¡ in THF/DMI (5/1) at room temperature.
h
i
j
k
1
l
H
t-BuOK in THF (eq 2); however, in the case of cis-lactam 3, the
epimerization did not proceed at all, even in the presence of an
excess amount of t-BuOK.
R
O
N
H
O
N
H
H
R1
OMe
OMe
R = MeO, F R2
20 mol% tBuOK
ð2Þ
Figure 1. Transition-state model.
N
THF, 0 o
C
N
O
O
18
trans/cis = 80/20
17
~100%
When 4-pentenal (11) was used as a Mannich donor, the
diastereoselectivity of 8 was better than those of 6 and 7, and the
enantioselectivity of 8 was good (syn/anti = 92/8, 82% ee). The
reason for the lower diastereoselectivity of 6 and 7 would be the
bulkiness of the R group on 9 and 10. When the R group is bulky,
it is difficult to form the rigid transition state shown in Figure 1.
Compared with the case using a p-methoxyphenyl substitu-
ent in eq 1, the stereoselectivity of 8 is slightly lower. In case
a p-methoxyphenyl substituent is used, a strong hydrogen
bond can be formed between the formed imine and enamine to
give high stereoselectivity. On the other hand, in case a p-
fluorophenyl substituent is used, the hydrogen bond would be
weaker than that in the p-methoxyphenyl substituent case, which
might decrease the stereoselectivity of 8. With the desired
carboxylic acid 8 in hand, we attempted further transformations.
Step B: Lactamization step: Carboxylic acid 8 was
lactamized with SOCl2 and Et3N to give cis-β-lactam 5 along
with trans-lactam 16 (99%, cis/trans >95/5), which was
removed by column chromatography because trans-lactam 16
had a low ee value (60% ee). Lactamization of carboxylic acids
6 and 7 gave 3 and 4 in poor yield, probably because of the
bulkiness of the R group.
OMe
OMe
In the case of cis-lactam 4, the addition of 1.0 equiv and
3.3 equiv t-BuOK in THF resulted in no reaction. The addition
of t-BuOH as a proton source was not effective at all. In contrast,
the addition of 1,3-dimethyl-2-imidazolidinone (DMI) promoted
epimerization to afford trans-lactam 15 as the major product.
However, the yield is not sufficient because of side reactions
such as hydrolysis of the β-lactam ring. The cause of the low
reactivity of 3 and 4 would be the bulkiness of the R substituent.
The epimerization of cis-lactam 5, which has a less bulky R
substituent, successfully proceeded to provide the desired trans-
lactam 16 as the major compound (cis/trans = 22/78) without
hydrolysis. trans-Lactam 16 was separated from cis-5 by column
chromatography.
Further transformation to carboxylic acid 2 was as follows
(Scheme 2): hydroboration of 16 with 9-BBN followed by
TEMPO oxidation11 successfully gave the known intermediate
2. The ee value of 2 was determined after conversion to the
corresponding methyl ester (80% ee, see Supporting Informa-
tion) and our previous work on the Mannich reaction7a predicts
1
that the absolute stereochemistry of 2 is 3R, 4S. The H NMR
Step C: Epimerization step: The substituent at C3 on the β-
lactam ring greatly affects isomerization. According to Burnett’s
report,10 cis-lactam 17 possessing a 3-phenylpropyl substituent
is easily converted to trans-lactam 18 with a catalytic amount of
spectrum is in good agreement with that of the literature.12 It was
confirmed that the enantiomeric excess did not change during
the transformation from 8 to 2. Conversion of 2 to 1 is known.
Thus, the formal synthesis of 1 has been achieved.
Chem. Lett. 2016, 45, 30–32 | doi:10.1246/cl.150916
© 2016 The Chemical Society of Japan | 31

OBn
9-BBN
OBn
TEMPO
OBn
B. A. Mckittrick, WO1995008532A1, 1995. b) B. B. Shankar,
M. P. Kirkup, S. W. McCombie, J. W. Clader, A. K. Ganguly,
Tetrahedron Lett. 1996, 37, 4095. c) T. K. Thiruvengadam, X.
Fu, C.-H. Tann, T. L. Mcallister, J. S. Chiu, C. Colon, Patent
WO2000034240A1, 2000. d) J. Talley, E. Martinez, D.
Zimmer, R. Lundrigan-Soucy, Patent WO2006102674A2,
2006. e) C. H. V. A. Sasikala, P. R. Padi, V. Sunkara, P.
Ramayya, P. K. Dubey, V. B. R. Uppala, C. Praveen, Org.
Process Res. Dev. 2009, 13, 907. f) M. Sova, J. Mravljak, A.
Kovač, S. Pečar, Z. Časar, S. Gobec, Synthesis 2010, 3433.
g) C. C. Sánchez, R. A. Dobarro, C. J. Bosch, D. D. García,
Patent WO2012004382A1, 2012. h) Y. Ren, Y.-J. Duan, R.-J.
Li, Y. Deng, L. Hai, Y. Wu, Chin. Chem. Lett. 2014, 25, 1157.
a) G. Wu, Y. Wong, X. Chen, Z. Ding, J. Org. Chem. 1999,
64, 3714. b) R. C. Aryan, K. K. G. Karooti, Y. Kumar, P. D.
Rathod, Patent WO2004099132A2, 2004. c) M. Michalak,
M. Stodulski, S. Stecko, A. Mames, I. Panfil, M. Soluch, B.
Furman, M. Chmielewski, J. Org. Chem. 2011, 76, 6931.
S. B. Rosenblum, T. Huynh, A. Afonso, H. R. Davis, Jr., N.
Yumibe, J. W. Clader, D. A. Burnett, J. Med. Chem. 1998, 41,
973.
O
R
S
PhI(OAc)2
THF, rt;
HO
HO
N
N
N
NaBO3, H2O
rt
H2O/CH2Cl2
rt
O
O
O
19
2
16
F
F
F
77%, 80% ee[a]
80%
Scheme 2. Synthesis of key intermediate 2. [a] ee value was
determined after conversion to corresponding methyl ester. See
Supporting Information for details.
In conclusion, we have accomplished the formal total
synthesis of ezetimibe (1) using a proline-mediated, asymmetric,
three-component Mannich reaction and isomerization as key
steps. This is the first organocatalyst-mediated asymmetric
synthesis of the β-lactam ring in ezetimibe (1).
5
This work was supported by a Grant-in-Aid for Scientific
Research on Innovative Areas “Advanced Molecular Trans-
formations by Organocatalysts” from The Ministry of Educa-
tion, Culture, Sports, Science and Technology, Japan.
6
7
a) Y. Hayashi, W. Tsuboi, I. Ashimine, T. Urushima, M.
Shoji, K. Sakai, Angew. Chem., Int. Ed. 2003, 42, 3677. b) Y.
Hayashi, T. Urushima, W. Tsuboi, M. Shoji, Nat. Protoc.
2007, 2, 113.
Supporting Information is available on http://dx.doi.org/
10.1246/cl.150916.
References
8
9
a) W. Notz, K. Sakthivel, T. Bui, G. Zhong, C. F. Barbas, III,
Tetrahedron Lett. 2001, 42, 199. b) A. Córdova, Synlett 2003,
1651. c) W. Notz, S. Watanabe, N. S. Chowdari, G. Zhong,
J. M. Betancort, F. Tanaka, C. F. Barbas, III, Adv. Synth.
Catal. 2004, 346, 1131. d) A. Córdova, Chem.®Eur. J. 2004,
10, 1987.
a) G. A. Kraus, B. Roth, J. Org. Chem. 1980, 45, 4825.
b) G. A. Kraus, M. J. Taschner, J. Org. Chem. 1980, 45, 1175.
c) B. S. Bal, W. E. Childers, Jr., H. W. Pinnick, Tetrahedron
1981, 37, 2091.
1
a) M. Van Heek, C. F. France, D. S. Compton, R. L. Mcleod,
N. P. Yumibe, K. B. Alton, E. J. Sybertz, H. R. Davis, Jr., J.
Pharmacol. Exp. Ther. 1997, 283, 157. b) M. Van Heek, C.
Farley, D. S. Compton, L. Hoos, K. B. Alton, E. J. Sybertz,
H. R. Davis, Jr., Br. J. Pharmacol. 2000, 129, 1748. c) C.
Gagné, H. E. Bays, S. R. Weiss, P. Mata, K. Quinto, M.
Melino, M. Cho, T. A. Musliner, B. Gumbiner, Am. J.
Cardiol. 2002, 90, 1084.
2
3
Merck & Co., Annual Report 2013.
M. Śnieżek, S. Stecko, I. Panfil, B. Furman, M. Chmielewski,
J. Org. Chem. 2013, 78, 7048.
10 D. A. Burnett, Tetrahedron Lett. 1994, 35, 7339.
11 R. K. Rej, P. Pal, S. Nanda, Tetrahedron 2014, 70, 4457.
12 Y. Tu, CN1931838A, 2006.
4
a) S. B. Rosenblum, S. Dugar, D. A. Burnett, J. W. Clader,
32 | Chem. Lett. 2016, 45, 30–32 | doi:10.1246/cl.150916
© 2016 The Chemical Society of Japan