A stereoselective synthesis of a key intermediate to 1β-methylcarbapenem via aziridine ring-opening reaction

Article abstract of DOI:10.1055/s-2003-39910

A stereocontrolled synthesis of azetidinone 3 as a key intermediate to 1β-methylcarbapenem 2 has been achieved via iodoamidation of trichloroacetimidate prepared from (Z)-olefinic allylic alcohol 6, aziridine ring-opening reaction with cyanide nucleophile and a tandem β-lactam formation.

Full text of DOI:10.1055/s-2003-39910

LETTER
1149
A Stereoselective Synthesis of a Key Intermediate to 1b-Methylcarbapenem via
Aziridine Ring-opening Reaction
Synthesis of
a
u
K
ey Interm
e
n
diate to 1
b
-M
g
e
thylcarbapenemHo Kang,* Mihyong Kim, Do Hyun Ryu
Center for Molecular Design and Synthesis, Department of Chemistry, School of Molecular Science (BK21), Korea Advanced Institute of
Science and Technology, Daejeon 305-701, Korea
Fax +82(42)8692810; E-mail: shkang@kaist.ac.kr
Received 14 April 2003
TBSO
HO
H
H
Abstract: A stereocontrolled synthesis of azetidinone 3 as a key
intermediate to 1b-methylcarbapenem 2 has been achieved via
iodoamidation of trichloroacetimidate prepared from (Z)-olefinic
allylic alcohol 6, aziridine ring-opening reaction with cyanide nu-
cleophile and a tandem b-lactam formation.
CO2H
2
NH
NC HNTs
OCOCMe3
OCH2OMe
O
3
4
HO
Key words: stereoselective synthesis, 1b-methylcarbapenem, io-
doamidation, aziridine ring-opening, tandem b-lactam formation
HO
N
OCOCMe3
Ts
Since the discovery of (+)-thienamycin 1 from Strepto-
myces cattleya,1 carbapenems have been received consid-
erable attention among the b-lactam antibiotics because of
their potent antibacterial activities (Figure 1). While (+)-
thienamycin is biologically unstable and metabolized by
renal dehydropeptidase I, 1b-methylcarbapenem 2 devel-
oped later by the Merck group2 was found to have the su-
perior stability against enzymatic metabolism as well as
the excellent antibiotic potency. Its prospective pharma-
ceutical utility has led many synthetic organic and medic-
inal chemists to be engaged in the synthesis of various 1b-
methylcarbapenem derivatives, in which azetidinone 3 is
evidently one of the most successfully employed key in-
termediates.3 Intrigued by its four contiguous asymmetric
centers as well as its medicinal value, we herein describe
a stereocontrolled synthesis of azetidinone 3.
5
6
Scheme 1 Retrosynthetic analysis
For the synthesis of 6, the hydroxyl group of the commer-
cially available alcohol 7 was protected with chlorometh-
yl methyl ether in 91% yield and the resulting bromide
was reacted with Ph3P to render phosphonium salt 8 in
86% yield (Scheme 2). Wittig olefination of 8 with (R)-2-
t-butyldimethylsilyloxypropanal5 was performed with n-
butyllithium in HMPA and THF to give a 21:1 unsepara-
ble mixture of cis- and trans-olefins in 90% combined
yield. After desilylation using TBAF, the mixture was
readily separated to afford the desired cis-olefinic allylic
alcohol 6 in 94% yield. Iodoamidation of (Z)-olefinic al-
lylic trichloroacetimidate has been established to generate
trans-4,5-disubstituted oxazoline by a preference for 5-
exo ring closure probably due to the steric interactions in
the transition state.4,6 Accordingly, 6 was reacted with
trichloroacetonitrile in the presence of DBU and the re-
sulting crude trichloroacetimidate was intramolecularly
cyclized using iodine monobromide at –78 °C to furnish
the expected trans-oxazoline 9 as a single stereoisomer.
The oxazoline 9 was exhaustively deprotected with meth-
anolic HCl, cyclized using sodium bicarbonate in MeOH
and tosylated in sequence to give rise to N-p-toluenesulfo-
nyl aziridine 10 in 95% yield from 9. In comparison with
N-t-butoxycarbonyl, N-benzyloxycarbonyl, N-trifluoro-
methylsulfonyl and N-nitrophenylsulfonyl aziridines,
the tosyl aziridine was found to be most reactive toward
aziridine ring-opening reaction. Also, it was revealed that
the protecting group of the primary hydroxyl group of 10
affected the ring-opening reaction significantly. Among
silyl, t-butoxycarbonyl, triphenymethyl and trimethyl-
acetyl protecting groups, the best experimental results in
terms of regioselectivity and chemical yield was attained
from trimethylacetate 5. Treatment of 5 with 5 equivalents
of lithium cyanide in HMPA produced a 7:1 separable
HO
HO
H
H
N
H
H
N
NH2
NMe2
S
S
NH
O
O
CO2H
CO2H
1
2
Figure 1
Retrosynthetically, aziridine 5 was proposed as our key
synthetic intermediate, the regioselective ring-opening re-
action of which was planned by cyanide anion to secure
all the requisite contiguous chiral centers (Scheme 1). In
addition, we envisioned that 5 could be generated via our
iodoamidation protocol using (Z)-olefinic allylic trichlo-
roacetimidate derived from the corresponding alcohol 6.4
Synlett 2003, No. 8, Print: 24 06 2003.
Art Id.1437-2096,E;2003,0,08,1149,1150,ftx,en;U07103ST.pdf.
© Georg Thieme Verlag Stuttgart · New York

1150
S. H. Kang et al.
LETTER
HO
OH
PPh3Br
Br
a,b
c,d
4
c,d
a,b
H2NOC HNTs
OH
12
OCH2OMe
OH
OCH2OMe
TBSO
O
TBSO
7
8
6
H
H
N
I
HO
C
NH
OTBS
OTBS
f
e
O
Ts
H
OCH2OMe
t-Boc NH Ts
N
N
O
OP
13
14
Ts
CCl3
P = H
10
TBSO
TBSO
9
g
H
H
H
P = COCMe3
5
e
ref 10b
CO2H
NH OTBS
NH
HO
HO
O
O
h
15
3
+
NC HNTs OCOCMe3
OCOCMe3
TsNH CN
Scheme 3 Reagents and conditions: (a) DIBAL, CH2Cl2, –78 °C,
91%; (b) 30% H2O2, EtOH, NH4OH, r.t., 89%; (c) TBSCl, imidazole,
DMF, r.t., 87%; (d) t-Boc2O, DMAP, Et3N, CH2Cl2, r.t., 90%; (e) Na-
naphthalide, DME, –78 °C, 95%.
4
11
Scheme 2 Reagents and conditions: (a) MeOCH2Cl, i-Pr2NEt,
CH2Cl2, r.t., 91%; (b) PPh3, K2CO3, MeCN, 90 °C, 86%; (c) n-BuLi,
HMPA, THF, –15°C, then (R)-2-t-butyldimethylsilyloxypropanal,
–78 °C to –15 °C, 90%; (d) TBAF, THF, r.t., 94% for cis-olefin;
(e) Cl3CCN, DBU, MeCN, 0 °C, then IBr, DBU, CH2Cl2, MeCN, –78
°C to –50 °C, 91%; (f) 6 N HCl, MeOH, r.t., then NaHCO3, MeOH,
r.t., then TsCl, r.t., 95%; (g) Me3CCOCl, DMAP, Et3N, CH2Cl2, 0 °C,
94%; (h) LiCN, HMPA, r.t., 91%.
dine ring-opening reaction and the tandem intramolecular
cyclization.
Acknowledgment
This work was supported by CMDS and the Brain Korea 21 Project.
mixture of the desired b-amino cyanide 4 and its dia-
stereomer 11 in 91% combined yield.
References
(1) (a) Albers-Schönberg, G.; Arison, B. H.; Hensens, O. D.;
Hirshfield, J.; Hoogsteen, K.; Kaczka, E. A.; Rhodes, R. E.;
Kahan, J. S.; Kahan, F. M.; Ratcliffe, R. W.; Walton, E.;
Ruswinkle, L. J.; Morin, R. B.; Christensen, B. G. J. Am.
Chem. Soc. 1978, 100, 6491. (b) Kahan, J. S.; Kahan, F. M.;
Goegelman, R.; Currie, S. A.; Jackson, M.; Stapley, E. O.;
Miller, T. W.; Miller, A. K.; Hendlin, D.; Mochales, S.;
Hernandez, S.; Woodruff, H. B.; Birnbaum, J. J. Antibiot.
1979, 32, 1.
b-Lactam ring formation was regarded as the crucial step
in the next sequence of our synthesis. Our initial effort for
the transformation focused on direct or indirect hydrolysis
of the cyanide derivatives into carboxylic acids to prove
abortive. Alternatively, the intramolecular cyclization
was conceived to be attained using amide instead of car-
boxylic acid. Accordingly, the trimethylacetyl group of 4
was removed with DIBAL and then its cyanide group was
hydrolyzed into amide 12 in 81% overall yield using hy-
drogen peroxide in ethanolic ammonium hydroxide
(Scheme 3).7 After bissilylation of 12 with TBSCl in 87%
yield, the resulting amide was subjected to mesyl chloride,
tosyl chloride, triflic anhydride, benzyloxycarbonyl chlo-
ride, i-butoxycarbonyl chloride and di-t-butyl dicarbon-
ate. Only di-t-butyl dicarbonate turned out to be
successfully applied for the cyclization. Treatment of the
bissilylated amide with di-t-butyl dicarbonate in the pres-
ence of DMAP and Et3N gave the desired azetidinone 14
in 90% yield evidently via t-butoxycarbonylamide 13.
(2) Shih, D. H.; Baker, F.; Cama, L.; Christensen, B. G.
Heterocycles 1984, 21, 29.
(3) (a) Kang, S. H.; Lee, H. S. Tetrahedron Lett. 1995, 36,
6713. (b) Berks, A. H. Tetrahedron 1996, 52, 331; and
references cited therein. (c) Kondo, K.; Seki, M.; Kuroda,
T.; Yamanaka, T.; Iwasaki, T. J. Org. Chem. 1997, 62,
2877. (d) Oh, C.-Y.; Ham, W.-H. Chem. Commun. 1999,
2365.
(4) Kang, S. H.; Kim, G. T. Tetrahedron Lett. 1995, 36, 5049.
(5) Kwon, H.; Lee, M.; Lee, S.; Hwang, T. Bull. Kor. Chem.
Soc. 1997, 18, 463.
(6) Bongini, A.; Cardillo, G.; Orena, M.; Sandri, S.; Tomasini,
C. J. Org. Chem. 1986, 51, 4905.
(7) Groziak, M. D.; Chern, J.; Townsend, L. B. J. Org. Chem.
1986, 51, 1065.
(8) Tanner, D.; Somfai, P. Tetrahedron 1988, 44, 619.
(9) Tsukada, N.; Shimada, T.; Gyoung, Y. S.; Asao, N.;
Yamamoto, Y. J. Org. Chem. 1995, 60, 143.
(10) (a) Kawabata, T.; Kimura, Y.; Ito, Y.; Terashima, S.; Sasaki,
A.; Sunagawa, M. Tetrahedron 1988, 44, 2149. (b) Kaga,
H.; Kobayashi, S.; Ohno, M. Tetrahedron Lett. 1989, 30,
113.
Finally, 14 was desulfonylated by sodium naphthalide
15
in DME8 uneventfully to produce b-lactam 159 ([a]D
=
–8.32, c 1.00, CHCl3) in 95% yield, which was converted
into a known key intermediate 3 to 1b-methylcarbapenem
according to Ohno’s procedure.10
In summary, we have established a stereoselective syn-
thesis of azetidinone 3 as a key intermediate to 2 via the
diastereoselective iodoamidation, the regioselective aziri-
Synlett 2003, No. 8, 1149–1150 ISSN 1234-567-89 © Thieme Stuttgart · New York