Highly diastereoselective alkylation of aziridine-2-carboxylate esters: Enantioselective synthesis of LFA-1 antagonist BIRT-377

Article abstract of DOI:10.1002/anie.200500923

The benzhydryl group is the key: Efficient alkylation of 3-substituted aziridine-2-carboxylates is only possible with N-benzhydryl-protected aziridines and occurs with complete retention of the configuration at the 2-position. Sequential catalytic asymmet

Full text of DOI:10.1002/anie.200500923

Angewandte
Chemie
Synthetic Methods
DOI: 10.1002/anie.200500923
Highly Diastereoselective Alkylation of
Aziridine-2-carboxylate Esters: Enantioselective
Synthesis of LFA-1 Antagonist BIRT-377**
Aniruddha P. Patwardhan, V. Reddy Pulgam, Yu Zhang,
and William D. Wulff*
Aziridines are important synthons in organic chemistry as
they provide convenient entry to optically pure amines, a-
amino acids, amino alcohols, diamines, and a variety of other
amino compounds that are useful both in industrial and
academic endeavors. In the past, most optically pure azir-
idines were derived from acyclic members of the chiral pool,
however, methods are emerging for the direct synthesis of
optically pure aziridines through catalytic asymmetric reac-
tions.[1] We have developed a process for the catalytic
asymmetric synthesis of aziridines from the reaction of
benzhydryl imines with diazo compounds mediated by a
chiral boron Lewis acid prepared from the VAPOL and
[2–4]
VANOLligands.
This asymmetric aziridination (AZ)
proved general for a range imines including those prepared
from a variety of aryl aldehydes and also from primary,
secondary, and tertiary aliphatic aldehydes (90–99% ee).
Much lower enantioselectivities were observed with N-
benzyl imines.[2d]
a-Amino acids which are tetrasubstituted at the a-carbon
are very popular tools that are used to control conformation
in peptides, and hence their biological and pharmaceutical
properties. A large number of methods have been developed
for the synthesis of tetrasubstituted a-amino acids, and this
subject has been reviewed.[5] Interestingly, aziridines have
rarely been used for the synthesis of tetrasubstituted a-amino
acids and this may be partly due to the fact that the alkylation
of aziridine-2-carboxylates is virtually an unknown reaction.
Typically, attempts to alkylate aziridine-2-carboxylate esters
leads to either ring opening or to self-condensation.[6,7] The
only known examples involve the use of either thioesters of
aziridine-2-carboxylates[6] or the use of a nitrogen substituent
on the aziridine that can chelate a metal enolate.[7]
A direct application of the AZ reaction to the synthesis of
tetrasubstituted a-amino acids could be envisioned through
the asymmetric aziridination of imine 1 with diazo compound
[*] Dr. A. P. Patwardhan, Dr. V. R. Pulgam, Y. Zhang, Prof. W. D. Wulff
Department of Chemistry
Michigan State University
East Lansing, MI 48824 (USA)
Fax: (+1)517-353-1793
E-mail: wulff@cem.msu.edu
[**] This work was supported by a grant from the NIH (NIGMS 63019).
We thank Dr. Rui Hangfor his help with the X-ray analysis.
Supportinginformation for this article, includingpreparative
procedures and spectral data for all new compounds and X-ray
diffraction data for 5a and 10, is available on the WWW under
http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 6169 –6172
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6169

Communications
lamide (LDA). The optimal conditions involved the treat-
ment of 3a with 2.0 equivalents of LDA at ꢀ788C followed by
addition of 3 equivalents of methyl iodide, and then allowing
the reaction mixture to warm to room temperature to give 5a
in 82% yield (1.1 equiv of LDA gave a 48% yield of 5a). We
screened the reactions in Table 1 with a mixture of DME (1,2-
dimethoxyethane) and diethyl ether (5:1) as solvent,[7] but
have subsequently found that comparable yields of 5a can be
achieved with either DME (79%) or THF (85%). Interest-
ingly, the use of diethyl ether led to a complicated reaction
mixture with the formation of only a trace amount of the
product. The aziridine 5a was formed as a single diaste-
reomer, as determined by 1H NMR spectroscopy (> 99%
d.r.). The assignment of the stereochemistry of 5a was made
on the basis of NOE experiments and by an X-ray diffraction
analysis of a single crystal of 5a.
The scope of the alkylation reaction was investigated with
a number of different aziridines and electrophiles (Table 1).
In all cases a single C2 epimer was observed and, on the basis
of the structure of 5a, was assigned as that formed from
retention at the 2-position. Attempts to epimerize the enolate
of 3a failed. Deprotonation of 3a at ꢀ788C and then warming
to 08C for 2 h before quenching with water at ꢀ788C led to
98% recovery of 3a with complete retention of stereochem-
istry. Primary alkyl iodides gave moderate yields, while
aldehydes gave high yields but with no selectivity at the
alcohol sterogenic center. The only electrophile that gave a
mixture of isomers at the aziridine was tributyltin chloride,
and in this case the minor product was assigned as the O-
alkylated product.
Scheme 1. Application of the AZ reaction to the synthesis of tetrasub-
stituted a-amino acids.
4 followed by the reductive opening of the aziridine ring
(route a, Scheme 1). However, we found that substituted
diazoesters of the type 4 (R2 ¼ H) are sluggish substrates in
the AZ reaction. Thus route B, which involves the less readily
available and sterically hindered diazo compound 4, is not
viable. Herein, we report the development of an alternative
method (route A, Scheme 1) that utilizes commercially
available ethyl diazoacetate and a subsequent stereoselective
alkylation of an aziridine-2-carbox-
The alkylations of aziridine-2-carboxylates with a benz-
hydryl group on the nitrogen have never been reported, nor
have the alkylations of aziridine-2-carboxylates with a
Table 1: Alkylation of 3.[a]
ylate. This strategy has led to the
first examples of alkylations of
aziridine-2-carboxylate esters with
nonchelating N substituents. The
synthetic utility of the sequential
AZ reaction and aziridine alkyl-
ation is illustrated in the asymmet-
ric catalytic synthesis of the leu-
kointegrin LFA-1 antagonist BIRT-
377.[8]
Entry
R
Substrate
EX
Product
Yield [%]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
3a
3a
3a
3a
3a
3a
3a
3a
3a
3a
3b
3c
3d
3e
3 f
H2O
H2O
CH3I
n-C8H17I
3a
3a
5a
5b
5c
5d
5e
5 f
5g
5h
5i
98
98[b]
82
Aside from thioesters,[6] the only
known examples of aziridine-2-car-
boxylate alkylations employ a 2-
methoxy-1-phenylethyl substituent
on nitrogen to stabilize the metal
enolate intermediate.[7] Thus it was
not clear if the desired alkylation of
3 to 5 would be feasible. Indeed,
initial attempts to alkylate the azir-
idine 3a with methyl iodide met
with failure. Both the lithium and
50
61
33
63
=
CH2 CHCH2Br
PhCH2Br
MOMCl
PhCHO
n-C3H7CHO
Bu3SnCl
CH3I
CH3I
CH3I
CH3I
MOMCl
95[c]
89[c]
73[d]
70
Ph
Ph
2-naphthyl
p-PhC6H4
p-BrC6H4
c-C6H11
tBu
5j
5k
5l
64
86
70
66
5m
[a] Unless otherwise specified, the reaction was performed with 2 equivalents of LDA in a solution of 3
(0.06m) in DME/Et2O (5:1). Only one isomer of 5 was observed except for entries 8–10. [b] Reaction
mixture warmed to 08C for 2 h and then quenched at ꢀ788C. [c] A 1:1 mixture of diastereomers at the
carbinol carbon. [d] Includes an 18% yield of a product tentatively assigned as an O-alkylated isomer.
MOM=methoxymethyl.
potassium
amides
of
hexa-
methyldisilazane failed to deproto-
nate aziridine 3a. Methylation was
successful with lithium diisopropy-
6170
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6169 –6172

Angewandte
Chemie
73% ee. Thus, there does not seem to be a time
dependence on the loss of optical activity. The
normal aging time of the enolate is 30 minutes,
and in this case its optical purity is not lower but
in fact slightly higher (81% ee). Increasing the
time of addition of 11 to LDA leads to only a
slight increase in the proportion of the alkyl-
ated product 12 to the Claisen product 13.
Seebach and co-workers previously observed
that the enolates of aziridine thioesters can also
be alkylated with retention and that the enolate
must either exist as the C-metalloenolate 14a or
the O-metalloenolate 14b, which is substan-
tially pyramidal at the enolate carbon (Sche-
me 3).[6b] We interpret the above results as
involving a configurationally stable enolate
(drawn as the C-enolate 15a but could also be
the pyramidal O-enolate 15b) that does not
epimerize with time but rather reacts with
either retention or inversion, the proportion
of which is electrophile dependent. This would
be consistent with the observation that all of the
cis-3-substituted aziridines shown in Table 1
give exclusively alkylated products with reten-
tion of configuration. The R group in enolate
16a would be expected to disfavor alkylation
with inversion by approach of the electrophile
from the rear side.
Scheme 2. Influence of substituents on aziridine-2-carboxylates in alkylation reactions.
substituent in the 3-position. Thus, we decided to investigate
the alkylations of aziridines 9 and 11 in an effort to see which
has the greater influence (Scheme 2). The results indicate that
the benzhydryl group has the most significant impact.
Whereas the cyclohexyl-substituted aziridine 3e is methyl-
ated cleanly in 70% yield with methyl iodide (Table 1,
entry 14), the benzyl analogue 9 gives a complicated reaction
mixture under the same conditions and none of the expected
alkylated aziridines could be detected or isolated from the
crude reaction mixture (all starting material was consumed).
The only product that was isolated and identified from this
reaction mixture was the pyrrole 10, whose origin is unclear
but whose structure was confirmed by X-ray diffraction
analysis. It is also clear that the presence of a substituent at
the 3-position of the aziridine is important. Aziridine 11 which
has no substituent in the 3-position gives only a 15% yield of
the alkylated aziridine 12 while all of the aziridines in Table 1
can be alkylated with methyl iodide in 64–82% yield under
the same conditions. The major product in the methylation of
11 is the Claisen condensation product 13, which is the only
product that has been seen from the attempted alkylation of
ethyl esters of aziridine-2-carboxylates.[7]
Interestingly, the optical purity of the methylated azir-
idine 12 (81% ee) is less than the starting aziridine 11 (94%
ee). The loss of optical activity (47% ee) is greater for the
protonation of the enolate of 11 which occurs with retention
of configuration. This result suggests that the loss of optical
activity is dependent on the electrophile. The enolate of 11
was allowed to age for only 2 minutes before methyl iodide
was added, and the optical purity of 12 was found to be
Scheme 3. Alkylation of aziridine thioesters. E=electrophile.
The utility of the alkylation of aziridine-2-carboxylates in
the synthesis of tetrasubstituted a-amino acids is demon-
strated in the synthesis of BIRT-377 (Scheme 4). BIRT-377
has been developed as an agent for the treatment of
inflammatory and immune disorders.[8] The asymmetric syn-
thesis was achieved through the aziridine 3d, which was
prepared by the AZ reaction with 1 mol% catalyst in 87%
yield and with 94% ee with greater than 50:1 selectivity for
the cis isomer. A single recrystallization gave material with
over 99% ee (72% recovery, first crop). The methylation of
3d followed from the above procedure to give 5k in 86%
yield as a single diastereomer. Reductive ring opening was
performed with borane–trimethylamine complex in the
presence of trifluoroacetic acid (TFA) to give the amino
ester 19 in 87% yield (BH3·Me3N with TFA gave a 6:1
mixture of 19 and 22).[9] Ring opening by hydrogenolysis
under a variety of conditions including Pd-C/HCO2H occur-
red with simultaneous bromide reduction. Surprisingly, the
Angew. Chem. Int. Ed. 2005, 44, 6169 –6172
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6171

Communications
W. D. Wulff, Org. Lett. 2001, 3, 3675 – 3678; d) J. Antilla, W. D.
Wulff, Y. Zhang, unpublished results.
[3] The catalyst is prepared by heating a solution of (S)-VAPOLand
triphenylborate (3 equiv) in CCl4 at 858C for 1 h and then by
removing the volatiles under
30 minutes.
a high vacuum at 858C for
[4] For other examples of aziridines from the reactions of imines with
diazo compounds, see references [2a], [2b], and references
therein, and the following: a) S. H. Lee, I. W. Song, Bull.
Korean Chem. Soc. 2005, 26, 223 – 224; b) M. Redlich, M. M.
Hossain, Tetrahedron Lett. 2004, 45, 8987 – 8990; c) B. Vander-
hoydonck, C. V. Stevens, Synthesis 2004, 722 – 734; d) A. L.
Williams, J. N. Johnston, J. Am. Chem. Soc. 2004, 126, 1612 –
1613; e) Y. Li, P. W. H. Chan, N.-Y. Zhu, C.-M. Che, H.-L.
Kwong, Organometallics 2004, 23, 54 – 66; f) J. R. Krumper, M.
Gerisch, J. M. Suh, R. G. Bergman, T. D. Tilley, J. Org. Chem.
2003, 68, 9705 – 9710; g) J. S. Yadav, B. V. S. Reddy, P. N. Reddy,
M. S. Rao, Synthesis 2003, 1387 – 1390; h) T. Akiyama, S. Ogi, K.
Fuchibe, Tetrahedron Lett. 2003, 44, 4011 – 4013; i) J. S. Yadav,
B. V. S. Reddy, M. S. Rao, P. N. Reddy, Tetrahedron Lett. 2003, 44,
5275 – 5278; j) W. Sun, C.-G. Zia, H.-W. Wang, Tetrahedron Lett.
2003, 44, 2409 – 2411; k) E. Hevia, J. Perez, V. Riera, D. Miguel, S.
Kassel, A. Rheingold, Inorg. Chem. 2002, 41, 4673 – 4679; l) M. F.
Mayer, M. M. Hossain, J. Organomet. Chem. 2002, 654, 202 – 209;
m) V. K. Aggarwal, E. Alonso, M. Ferrara, S. E. Spey, J. Org.
Chem. 2002, 67, 2335 – 2344; n) D. Morales, J. Perez, L. Riera, V.
Riera, R. Corzo-Suarez, S. Garcia-Granda, D. Miguel, Organo-
metallics 2002, 21, 1540 – 1545; o) M. P. Spanedda, B. Crousse, S.
Narizuka, D. Bonnet-Delphon, J.-P. Begue, Collect. Czech. Chem.
Commun. 2002, 67, 1359 – 1365; p) K.-D. Lee, J.-M Suh, J.-H.
Park, H.-J. Ha, H. G. Choi, C. S. Park, J. W. Chang, W. K. Lee, Y.
Dong, H. Yun, Tetrahedron 2001, 57, 8267 – 8276; q) V. K.
Aggarwal, M. Ferrara, C. J. OꢀBrien, A. Thompson, R. V. H.
Jones, R. Fieldhouse, J. Chem. Soc. Perkin Trans. 1 2001, 1635 –
1643; r) B. Crousse, S. Narizuka, D. Bonnet-Delpon, J.-P. Begue,
Synlett 2001, 5, 679 – 681; s) M. P. Doyle, W. Hu, D. J. Timmons,
Org. Lett. 2001, 3, 933 – 935; t) M. F. Mayer, Q. Wang, M. M.
Hossain, J. Organomet. Chem. 2001, 630, 78 – 83; u) V. K.
Aggarwal, E. Alonso, G. Fang, M. Ferrara, G. Hynd, M.
Porcelloni, Angew. Chem. 2001, 113, 1482 – 1485; Angew. Chem.
Int. Ed. 2001, 40, 1433 – 1436; v) S. H. Lee, T. D. Han, K. Yu, K. H.
Ahn, Bull. Korean Chem. Soc. 2001, 22, 449 – 450; w) V. K.
Aggarwal, M. Ferrara, Org. Lett. 2000, 2, 4107 – 4110; x) R. Hori,
T. Aoyama, T. Shioiri, Tetrahedron Lett. 2000, 41, 9455 – 9458;
y) S. Sengupta, S. Mondal, Tetrahedron Lett. 2000, 41, 6245; z) T.
Kubo, S. Sakaguchi, Y. Ishii, Chem. Commun. 2000, 625 – 626;
aa) W. Xie, J. Fang, J. Li, P. G. Wang, Tetrahedron 1999, 55,
12929 – 12938; ab) D. L. Wright, M. C. McMills, Org. Lett. 1999,
1, 667 – 670; ac) K. Juhl, R. G. Hazell, K. A. Jorgensen, J. Chem.
Soc. Perkin Trans. 1 1999, 2293 – 2297.
Scheme 4. Reagents and conditions: a) Ph2CHNH2 (1.0 equiv), CH2Cl2,
MgSO4, room temperature, 24 h; b) ethyl diazoacetate (1.1 equiv), (S)-
VAPOL-B catalyst (1 mol%; see footnote 3), CCl4, room temperature,
20 h; c) LDA (2 equiv), DME/Et2O (5:1), ꢀ788C, 0.5 h; MeI (3 equiv),
ꢀ78 to RT; d) BH3·Me3N (12 equiv), TFA (9 equiv), CH2Cl2, 08 to RT,
36 h; e) Et3SiH (3 equiv), TFA, reflux, 5 h; f) 3,5-dichloro-phenyl isocya-
nate (1.1 equiv), DMSO, Na2CO3, 1208C, 1.5 h; g) NaHMDS
(1.2 equiv), DMF, room temperature; MeI (1.2 equiv), 2 h. DMSO=di-
methyl sulfoxide; HMDS=hexamethyldisilazide; DMF=N,N-dimethyl-
formamide.
reductive ring opening of 5k with triethylsilane and TFA gave
mainly the amino alcohol 22, which results from ring opening
with trifluoroacetate. Cleavage of the benzhydryl group in 19
with triethylsilane in the presence of trifluoroacetic acid gave
the amine 20 in 95% yield. The conversion of amine 20 into
BIRT-377 (21) follows methods employed in previous syn-
theses.[8d]
The fact that benzhydryl-protected aziridine-2-carboxy-
lates can be readily alkylated at the 2-position greatly
enhances the synthetic utility of the asymmetric aziridination
(AZ) reaction, as illustrated by the synthesis of BIRT-377.
Additional studies with other electrophiles and applications
in other syntheses will be reported in due course.
[5] a) C. Cativiela, M. D. Diaz-de-Villegas, Tetrahedron: Asymmetry
1998, 9, 3517 – 3599; b) C. Cativiela, M. D. Diaz-de-Villegas,
Tetrahedron: Asymmetry. 2000, 11, 645 – 732.
[6] a) D. Seebach, R. Hꢁner, Chem. Lett. 1987, 49 – 52; b) R. Hꢁner,
B. Olano, D. Seebach, Helv. Chim. Acta 1987, 70, 1676 – 1691.
[7] a) V. Alezra, M. Bonin, L. Micouin, H.-P. Husson, Tetrahedron
Lett. 2000, 41, 651 – 654; b) V. Alezra, M. Bonin, L. Micouin, C.
Policar, H.-P. Husson, Eur. J. Org. Chem. 2001, 2589 – 2594.
[8] a) N. K. Yee, Org. Lett. 2000, 2, 2781 – 2784; b) E. Napolitano, V.
Farina, Tetrahedron Lett. 2001, 42, 3231 – 3234; c) R. P. Frutos, S.
Stehle, L. Nummy, N. Yee, Tetrahedron: Asymmetry 2001, 12,
101 – 104; d) S. R. Kapadia, D. M. Spero, M. Eriksson, J. Org.
Chem. 2001, 66, 1903 – 1905; e) N. S. Chowdari, C. F. Barbas III,
Org. Lett. 2005, 7, 867 – 870.
Received: March 11, 2005
Published online: August 24, 2005
Keywords: asymmetric catalysis · aziridines · BIRT-377 · enols ·
.
synthetic methods
[1] P. Müller, C. Fruit, Chem. Rev. 2003, 103, 2905 – 2920.
[9] a) E. Vedejs, A. Klapars, D. L. Warner, A. H. Weiss, J. Org. Chem.
2001, 66, 7542 – 7546; b) B. Carboni, L. Monnier, Tetrahedron
1999, 55, 1197 – 1248.
[2] a) J. C. Antilla, W. D. Wulff, J. Am. Chem. Soc. 1999, 121, 5099 –
5100; b) J. C. Antilla, W. D. Wulff, Angew. Chem. 2000, 112, 4692 –
4695; Angew. Chem. Int. Ed. 2000, 39, 4518 – 4521; c) C. Loncaric,
6172
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6169 –6172