Succinct synthesis of β-amino acids via chiral isoxazolines

Article abstract of DOI:10.1021/ja0431713

β-Amino acids are important synthetic targets due to their presence in a wide variety of natural products, pharmaceutical agents, and mimics of protein structural motifs. While β-amino acids containing geminal substitution patterns have enormous potential for application in these contexts, synthetic challenges to the stereoselective preparation of this class of compound have thus far limited more complete studies. We present here a straightforward method employing chiral isoxazolines as key intermediates to access five different β-amino acid structural types with excellent selectivity. Of particular note is the use of this approach to prepare highly substituted cis-β-proline analogues. The ready access to these diversely substituted compounds is expected to facilitate future studies of the structure and function of this important class of molecules.

Full text of DOI:10.1021/ja0431713

Published on Web 03/22/2005
Succinct Synthesis of â-Amino Acids via Chiral Isoxazolines
Amelia A. Fuller, Bin Chen, Aaron R. Minter, and Anna K. Mapp*
Contribution from the Department of Chemistry, UniVersity of Michigan,
Ann Arbor, Michigan 48109-1055
Received November 12, 2004; E-mail: amapp@umich.edu
Abstract: â-Amino acids are important synthetic targets due to their presence in a wide variety of natural
products, pharmaceutical agents, and mimics of protein structural motifs. While â-amino acids containing
geminal substitution patterns have enormous potential for application in these contexts, synthetic challenges
to the stereoselective preparation of this class of compound have thus far limited more complete studies.
We present here a straightforward method employing chiral isoxazolines as key intermediates to access
five different â-amino acid structural types with excellent selectivity. Of particular note is the use of this
approach to prepare highly substituted cis-â-proline analogues. The ready access to these diversely
substituted compounds is expected to facilitate future studies of the structure and function of this important
class of molecules.
Introduction
â-Amino acids are important constituents of biologically
active natural products (e.g., taxol, dolastatin) and pharmaceuti-
cal agents as well as valuable precursors to such structures as
â-lactams (Figure 1a).1,2 In addition, oligomers of â-amino acids
(â-peptides) have attracted attention as useful peptidomimetics
because of their propensity to adopt stable secondary structures
3
(
e.g., helices, â-turns) and their proteolytic stability relative to
4
natural peptides. These features make â-peptides attractive
targets for use in a number of biological applications. There
are several recent reports, for example, detailing â-peptides with
positively charged residues that exhibit potent antimicrobial
activities, presumably through disruption of bacterial cell
5
membranes. In all of the â-amino acid applications, the
substitution pattern and the stereochemistry at the C2 and/or
C3 positions strongly influence both structural characteristics
and stability. Thus, effective stereoselective synthetic approaches
for â-amino acids remain highly sought after.2
Two particularly intriguing classes of â-amino acids are also
among the most challenging synthetic targets. The first of these
Figure 1. Structural classes of â-amino acids. (a) The basic structure of a
â-amino acid. The nomenclature of Seebach in which the superscript after
,6-8
8
“â” describes the substitution pattern of a â-amino acid is used throughout.
According to this system, for example, a â2 -amino acid has a single
substituent at the C2 position and two substituents at the C3 position. (b)
Examples of geminally disubstituted â-amino acids, a particularly chal-
lenging class to prepare in stereoisomerically pure form. (c) â-Proline
containing either a cis relationship between the amine group and the
carboxylic acid (as in 4) or a trans relationship (5) produces predictable
secondary structures upon oligomerization.
,3,3
(
1) Juaristi, E. EnantioselectiVe Synthesis of â-Amino Acids; Wiley-VCH: New
York, 1997.
(
2) Juaristi, E.; L o´ pez-Ruiz, H. Curr. Med. Chem. 1999, 6, 983-1004.
3) Seebach, D.; Matthews, J. L. Chem. Commun. 1997, 2015-2022. Hill, D.
J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S. Chem. ReV. 2001,
(
1
01, 3893-4011. Gellman, S. H. Acc. Chem. Res. 1998, 31, 173-180.
3,3
contains geminal disubstitution at C2 and/or C3 such as the â -
Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Chem. ReV. 2001, 101,
3
219-3232. Stigers, K. D.; Soth, M. J.; Nowick, J. S. Curr. Opin. Chem.
and â -amino acids depicted in Figure 1b. Given the dense
substitution and their resistance to proteolytic degradation, this
molecular class provides excellent building blocks for bioactive
molecules. Further, preliminary studies and predictive models
indicate that the secondary structures adopted by â-peptides
incorporating these amino acids will be unique and potentially
2,3,3
8
Biol. 1999, 3, 714-723.
(
4) Hook, D. F.; Gessier, F.; Noti, C.; Kast, P.; Seebach, D. ChemBioChem
2
004, 5, 691-706. Frackenpohl, J.; Arvidsson, P. I.; Schreiber, J. V.;
Seebach, D. ChemBioChem 2001, 2, 445-455. Hintermann, T.; Seebach,
D. Chimia 1997, 51, 244-247.
(5) Porter, E. A.; Wang, X. F.; Lee, H. S.; Weisblum, B.; Gellman, S. H. Nature
2
000, 404, 565-565. Arvidsson, P. I.; Frackenpohl, J.; Ryder, N. S.;
Liechty, B.; Petersen, F.; Zimmermann, H.; Camenisch, G. P.; Woessner,
R.; Seebach, D. ChemBioChem 2001, 2, 771. Hamuro, Y.; Schneider, J.
P.; DeGrado, W. F. J. Am. Chem. Soc. 1999, 121, 12200-12201.
6) Liu, M.; Sibi, M. P. Tetrahedron 2002, 58, 7991-8035.
7) Cole, D. C. Tetrahedron 1994, 50, 9517-9582. Sewald, N. Angew. Chem.,
Int. Ed. 2003, 42, 5794-5795.
8
quite stable. However, due to the relatively small number of
(
(
such structures available in stereoisomerically pure form, more
9,10
complete studies have not been forthcoming. A second class
(8) Abele, S.; Seebach, D. Eur. J. Org. Chem. 2000, 1-15.
of â-amino acids of particular interest includes the cyclic proline
5376
9
J. AM. CHEM. SOC. 2005, 127, 5376-5383
10.1021/ja0431713 CCC: $30.25 © 2005 American Chemical Society

Succinct Synthesis of
â-Amino Acids
A R T I C L E S
1
4
2
3
Figure 2. Isoxazoline strategy for the synthesis of â-amino acids. If R and R or R and R are covalently linked, access to cyclic â-amino acid analogues
such as the â-prolines can be envisioned.
analogues (“â-proline”), cis-2-aminocyclopentanoic acid (cispen-
tacin, 4) and trans-2-aminocyclopentanoic acid (5) (Figure
natively, â-amino acids can be accessed via a conjugate addition
of an amine into an acrylate; this transformation has been
6
11,12
1
c).
Illustrative of the relationship between stereochemistry
carried out asymmetrically by employing a chiral amine, a chiral
13
19
and structure, oligomers of 4 form strands, whereas oligomers
ester, or a chiral Lewis acid. However, extending this method
1
4
of 5 form helices. Due to their predictable and well-defined
structural characteristics, this class of â-amino acids has
enormous potential for the formation of higher order structures,
transitioning to protein-like structure and function for catalyst
development and pharmaceutical applications.15 As with their
R-amino acid counterparts, the formation of higher order
assemblies (discrete helical bundles or â-barrels, for example)
will be dependent upon interactions between side-chain func-
tional groups such as hydrophobic packing, salt bridges, and
to the synthesis of highly substituted targets is problematic as
it involves overcoming both the steric hindrance and the reduced
reactivity of the corresponding acrylates. Asymmetric catalytic
reduction of â-amino acrylates has provided a highly efficient
20
entry to various â-amino acids in high enantiomeric excesses,
although this method is limited by its intrinsic inability to access
geminally disubstituted â-amino acids such as 2, 3, or 6. Perhaps
the most versatile addition to the repertoire of synthetic
approaches is the addition of enolates to chiral imines,2 and
this has provided a method for the facile preparation of a range
of â-amino acids including a recent report of the parent trans-
1,22
1
6
hydrogen bonding. Efficient synthetic approaches to single
stereoisomers of both 4 and 5 have been reported,11,12 but the
introduction of additional substituents (e.g., 6) remains a
substantial synthetic challenge. Thus, the further development
of both classes of â-amino acids is contingent upon synthetic
strategies that will provide access to single stereoisomers of
these compounds.
23
â-proline (5) prepared via a chiral sulfinyl imine. Nonetheless,
this approach does not provide ready entry into certain classes
of highly substituted â-amino acids, particularly those with
sterically similar substituents at C3 (e.g., 2 and 3), for example,
or members of the cis-substituted â-proline class (e.g., 4 and
6). Thus, access to the highly substituted and/or conformationally
constrained â-amino acids remains a significant obstacle to the
study of â-peptides incorporating these building blocks.
To address this synthetic challenge, we developed an
orthogonal approach employing chiral isoxazolines as key
intermediates for the preparation of a diverse array of â-amino
acids, including the particularly challenging cyclic and highly
The essential synthetic challenge for the â-amino acid classes
described above is one shared by many synthetic targets,
including a diverse group of natural products and other
biologically active agents: stereocenters bearing an amine
17
substituent, in particular tertiary and quaternary stereocenters.
Thus, synthetic advances in highly substituted â-amino acid
preparation also impact a variety of fields. The traditional
method of â-amino acid synthesis, the Arndt-Eistert homolo-
2
4
substituted variants. Isoxazolines are readily accessible as
single stereoisomers via a 1,3-cycloaddition reaction between
a nitrile oxide and a chiral allylic alcohol employing the
1
3
gation, is a powerful approach for the preparation of â -amino
acids but is relatively ineffective for more substituted versions
because the yields and selectivities of the reaction suffer
significantly with increased substitution.18 In addition, cyclic
â-amino acids cannot be prepared using this approach. Alter-
25
conditions originally described by Kanemasa et al. and further
26
expanded by Carreira and co-workers (Figure 2). We reasoned
that nucleophilic addition to the C3 CdN bond of the isoxazo-
(
9) Abele, S.; Seiler, P.; Seebach, D. HelV. Chim. Acta 1999, 82, 1559-1571.
Seebach, D.; Abele, S.; Sifferlen, T.; H a¨ nggi, M.; Gruner, S.; Seiler, P.
HelV. Chim. Acta 1998, 81, 2218-2243.
(19) Doi, H.; Sakai, T.; Iguchi, M.; Yamada, K.; Tomioka, K. J. Am. Chem.
Soc. 2003, 125, 2886-2887. Hata, S.; Iguchi, M.; Iwasawa, T.; Yamada,
K.; Tomioka, K. Org. Lett. 2004, 6, 1721-1723. Hamashima, Y.; Somei,
H.; Shimura, Y.; Tamura, T.; Sodeoka, M. Org. Lett. 2004, 6, 1861-1864.
Yamagiwa, N.; Matsunaga, S.; Shibasaki, M. J. Am. Chem. Soc. 2003, 125,
16178-16179. Sibi, M. P.; Prabagaran, N.; Ghorpade, S. G.; Jasperse, C.
P. J. Am. Chem. Soc. 2003, 125, 11796-11797. Palomo, C.; Oiarbide, M.;
Halder, R.; Kelso, M.; G o´ mez-Bengoa, E.; Garcia, J. M. J. Am. Chem.
Soc. 2004, 126, 9188-9189. Guerin, D. J.; Miller, S. J. J. Am. Chem. Soc.
2002, 124, 2134-2136. Myers, J. K.; Jacobsen, E. N. J. Am. Chem. Soc.
1999, 121, 8959-8960.
(
10) Seebach, D.; Sifferlen, T.; Mathieu, P. A.; H a¨ ne, A. M.; Krell, C. M.;
Bierbaum, D. J.; Abele, S. HelV. Chim. Acta 2000, 83, 2849-2864.
11) F u¨ l o¨ p, F. Chem. ReV. 2001, 101, 2181-2204.
(
(
12) LePlae, P. R.; Umezawa, N.; Lee, H. S.; Gellman, S. H. J. Org. Chem.
2
001, 66, 5629-5632.
(
(
(
13) Martinek, T. A.; T o´ th, G. K.; Vass, E.; Holl o´ si, M.; F u¨ l o¨ p, F. Angew. Chem.,
Int. Ed. 2002, 41, 1718.
14) Appella, D. H.; Christianson, L. A.; Klein, D. A.; Powell, D. R.; Huang,
X. L.; Barchi, J. J.; Gellman, S. H. Nature 1997, 387, 381-384.
15) Kimmerlin, T.; Namoto, K.; Seebach, D. HelV. Chim. Acta 2003, 86, 2104-
(20) Hsiao, Y.; Rivera, N. R.; Rosner, T.; Krska, S. W.; Njolito, E.; Wang, F.;
Sun, Y. K.; Armstrong, J. D.; Grabowski, E. J. J.; Tillyer, R. D.; Spindler,
F.; Malan, C. J. Am. Chem. Soc. 2004, 126, 9918-9919. Tang, W. J.; Wu,
S. L.; Zhang, X. M. J. Am. Chem. Soc. 2003, 125, 9570-9571.
(21) Tang, T. P.; Ellman, J. A. J. Org. Chem. 2002, 67, 7819-7832.
(22) Jacobsen, M. F.; Ionita, L.; Skrydstrup, T. J. Org. Chem. 2004, 69, 4792-
4796. C o´ rdova, A. Acc. Chem. Res. 2004, 37, 102-112. Matsunaga, S.;
Kumagai, N.; Harada, S.; Shibasaki, M. J. Am. Chem. Soc. 2003, 125,
4712-4713. Kobayashi, S.; Matsubara, R.; Nakamura, Y.; Kitagawa, H.;
Sugiura, M. J. Am. Chem. Soc. 2003, 125, 2507-2515.
2
109. Gademann, K.; Kimmerlin, T.; Hoyer, D.; Seebach, D. J. Med. Chem.
2
001, 44, 2460-2468. Gelman, M. A.; Richter, S.; Cao, H.; Umezawa,
N.; Gellman, S. H.; Rana, T. M. Org. Lett. 2003, 5, 3563-3565. Gademann,
K.; Ernst, M.; Hoyer, D.; Seebach, D. Angew. Chem., Int. Ed. 1999, 38,
1
223-1226. Br u¨ ckner, A. M.; Chakraborty, P.; Gellman, S. H.; Dieder-
ichsen, U. Angew. Chem., Int. Ed. 2003, 42, 4395-4399. Kritzer, J. A.;
Lear, J. D.; Hodsdon, M. E.; Schepartz, A. J. Am. Chem. Soc. 2004, 126,
9
468-9469. Raguse, T. L.; Lai, J. R.; LePlae, P. R.; Gellman, S. H. Org.
Lett. 2001, 3, 3963-3966.
(23) Schenkel, L. B.; Ellman, J. A. Org. Lett. 2004, 6, 3621-3624.
(24) Minter, A. R.; Fuller, A. A.; Mapp, A. K. J. Am. Chem. Soc. 2003, 125,
6846-6847. Fuller, A. A.; Chen, B.; Minter, A. R.; Mapp, A. K. Synlett
2004, 1409-1413.
(
16) DeGrado, W. F.; Summa, C. M.; Pavone, V.; Nastri, F.; Lombardi, A. Annu.
ReV. Biochem. 1999, 68, 779-819. Micklatcher, C.; Chmielewski, J. Curr.
Opin. Chem. Biol. 1999, 3, 724-729.
(
17) Corey, E. J.; Guzman-Perez, A. Angew. Chem., Int. Ed. 1998, 37, 388-
(25) Kanemasa, S.; Nishiuchi, M.; Kamimura, A.; Hori, K. J. Am. Chem. Soc.
1994, 116, 2324-2339.
4
01.
(
18) Yang, H.; Foster, K.; Stephenson, C. R. J.; Brown, W.; Roberts, E. Org.
Lett. 2000, 2, 2177-2179.
(26) Bode, J. W.; Fraefel, N.; Muri, D.; Carreira, E. M. Angew. Chem., Int. Ed.
2001, 40, 2082-2085.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5377

A R T I C L E S
Fuller et al.
Scheme 1. Reduction of the Isoxazoline
Figure 3. Directed hydride addition to isoxazolines.
line (7) would be strongly influenced by the nearby C5
substituent, enabling us to establish the C3 stereochemistry in
a controlled manner. Addition of hydride at this stage would
3
2,3
provide key intermediates for the synthesis of â - and â -amino
acids, whereas the inter- or intramolecular addition of carbon
nucleophiles would create the nitrogen-bearing stereogenic
centers present in more substituted acyclic and cyclic variants.
As there existed few examples of the addition of nucleophiles
other than hydride to isoxazolines,2 this was the primary point
of investigation throughout the development of the approach.
The product of the nucleophilic addition, an isoxazolidine (8),
contains all of the functionality and stereochemical relationships
present in the final â-amino acid product (10). Conversion to
the â-amino acid could be accomplished through reductive
cleavage of the N-O bond followed by protection of the
resulting amine and subsequent cleavage of the diol intermediate.
We provide here a full account of this approach and its
This was particularly surprising in the case of Zr(BH4)2 because
this reagent has been used to reduce the CdN bond selectively
3
1
in the context of an oxime. We reasoned that in our system
the C5′ hydroxyl, deprotonated under the reaction conditions,
and the ring oxygen might be forming a chelate with the
reducing agent, thus activating the N-O bond for cleavage as
well as attenuating the selectivity of the hydride delivery.
Identification of conditions that would disrupt such a chelate
could then lead to enhanced selectivity for the reduction.
Following this rationale, the use of LiAlH4 in THF to reduce
the isoxazoline followed by immediate protection of the amine
provided the desired balance of reactivity and selectivity with
the formation of a 13:1 ratio of diastereomers (Scheme 1). To
verify the facial selectivity of the hydride addition, the protected
amine diol was converted to â-leucine by oxidative cleavage
of the diol moiety and the optical rotation of the final product
7,28
3
2,3
3,3
application to the synthesis of a diverse array of â -, â -, â -,
and â-2,3,3-amino acids, significantly expanding the diversity of
â-amino acids available for study and application. In addition,
we describe the further development of the strategy to include
the synthesis of a series of densely substituted cis-â-proline
analogues previously inaccessible by conventional means.
3
3
was compared to that reported in the literature. This provided
confirmation that the hydride approach was from the same face
of the isoxazoline ring as the C5 hydroxyethyl substituent. The
same series of reactions was then carried out on isoxazolines
bearing different substituents at C3 and C4 to explore the scope
of the approach (Table 1). Increasing the steric bulk at C3 (20
and 26) has little effect on the key reaction. Further, less robust
functional groups such as the alkyne in 29 are also well tolerated
Results and Discussion
3
2,3
Synthesis of â - and â -Amino Acids. To develop the
approach outlined above, our initial efforts were directed toward
3
2,3
less substituted â-amino acid targets, the â - and the â -amino
acids that contain either one or two tertiary stereogenic centers
and are commonly used in â-peptide applications. Addition of
hydride to the CdN bond would provide the key stereocenter
in the final â-amino acid product (Figure 3). We identified one
example in a report by J a¨ ger in which a proximal hydroxymethyl
substituent apparently served to direct the reduction of an
isoxazoline to provide a â-amino alcohol with modest diaste-
reoselectivity (1.7:1).29 In addition, there are several examples
of oximes (stereoelectronically similar to isoxazolines) undergo-
ing reductions directed by a nearby hydroxyl group with agents
such as Me4NBH(OAc)3.30 Given this precedent and the
presence of the hydroxyl on the C5 substituent of the isoxazo-
line, we anticipated that hydride attack directed by this group
would be the most effective means to set the C3 stereocenter.
To investigate this, isoxazoline 14, the precursor to â-leucine
3
to provide a â -amino acid typically more difficult to obtain.
Substituents at C4 of the isoxazoline can be incorporated by
employing a disubstituted olefin in the cycloaddition, as in
25,26
3
3
2.
Conversion of all of the amine diols to the final â - and
2,3
â -amino acid targets also proceeded without incident. Overall,
this approach provides a rapid, selective entry into these
commonly employed classes of â-amino acids.
Synthesis of â3,3- and â2,3,3-Amino Acids. As noted in the
3,3
2,3,3
Introduction, â - and â -amino acids are targets of particular
interest for the preparation of structurally defined oligomers but
are also among the most difficult to prepare stereoselectively
2,3,3
due to the C3 quaternary stereocenter and, in the case of â
-
8
amino acids, the vicinal substitution. To access the C3 geminal
disubstitution present in this interesting and synthetically
challenging class of compounds, we sought conditions for the
selective addition of carbon nucleophiles to C3 of appropriately
(
16), was prepared as a single stereoisomer under standard
conditions and subjected to a variety of reduction conditions
Scheme 1). NaCNBH4, Me4NBH(OAc)3, and DIBAL-H were
(
ineffective in this reaction, with no conversion to product
observed under several different conditions. The more reactive
agent Zr(BH4)2, previously shown to effect directed reduction,
(
27) Uno, H.; Terakawa, T.; Suzuki, H. Chem. Lett. 1989, 1079-1082. Huang,
K. S. L.; Lee, E. H.; Olmstead, M. M.; Kurth, M. J. J. Org. Chem. 2000,
65, 499-503.
31
(
(
(
28) Uno, H.; Terakawa, T.; Suzuki, H. Bull. Chem. Soc. Jpn. 1993, 66, 2730-
2737.
did produce the desired product but in low yield and poor
diastereoselectivity. The conditions of J a¨ ger (LiAlH4/Et2O)
provided good conversion of the isoxazoline to the product
29) J a¨ ger, V.; Schwab, W.; Buss, V. Angew. Chem., Int. Ed. Engl. 1981, 20,
601-603.
30) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. ReV. 1993, 93, 1307-
1
370. Williams, D. R.; Osterhout, M. H. J. Am. Chem. Soc. 1992, 114,
2
9,32
amine diols, but the selectivity was modest (1.5:1).
It was
8750-8751.
(31) Williams, D. R.; Benbow, J. W.; Sattleberg, T. R.; Ihle, D. C. Tetrahedron
noted during this study that, regardless of the reaction conditions,
the intermediate isoxazolidine (e.g., 12) was never observed.
Lett. 2001, 42, 8597-8601.
(32) J a¨ ger, V.; Buss, V.; Schwab, W. Tetrahedron Lett. 1978, 3133-3136.
5378 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Succinct Synthesis of
â-Amino Acids
A R T I C L E S
Table 1. Reduction of Isoxazolines and Conversion to â-Amino
Acids
Table 2. Addition of Stabilized Grignard Reagents to Isoxazolines
a
1
Diastereomeric ratios determined by H NMR spectral integration prior
b
a Diastereomer ratio determined by 1H NMR analysis of the crude
reaction mixture. b Isolated as the HCl salt.
to separation. Combined yield of diastereomers; only a single diastereomer
was carried on to the acid. Due to poorly dispersed resonances in the H
NMR spectra, dr was determined by mass comparisons of isolated
c
1
d
compounds. Treatment with NaIO4 followed by NaClO2 provided 31.
addition to isoxazolines are even more rare,27,28 making this a
largely unexplored reaction.
functionalized isoxazolines. In contrast to the hydride addition
described in the previous section, it was anticipated that the
facial selectivity of these nucleophiles would be sterically driven,
with approach opposite the C5 ring substituent being more
favorable. However, while numerous examples of the selective
addition of organometallic nucleophiles to imines have been
Organometallic reagents commonly used for additions to
imines (organocerium, organozinc, and cuprate reagents, for
example) provided poor conversion to product (0-15% yield)
under a variety of conditions. In contrast, the more reactive
Grignard reagent allylmagnesium chloride was an effective
nucleophile, providing the desired adduct in good to excellent
yields (entries 1-3, Table 2). Optimal diastereoselectivity was
obtained using THF as solvent and upon complexation of the
isoxazoline with BF3‚OEt2 to activate the CdN bond prior to
addition of the Grignard reagent. Less polar solvents such as
diethyl ether provided comparable product yields but reduced
diastereoselectivity (2-3:1). The facial selectivity of the addition
was verified by NOE difference experiments on a derivative of
3
4
documented, only a few examples of carbon nucleophiles
added to analogous acyclic ketoxime ethers can be found in
3
5
the literature. References documenting carbon nucleophile
(
33) Elmarini, A.; Roumestant, M. L.; Viallefont, P.; Razafindramboa, D.;
Bonato, M.; Follet, M. Synthesis 1992, 1104-1108.
(
34) Bloch, R. Chem. ReV. 1998, 98, 1407-1438. Enders, D.; Reinhold, U.
Tetrahedron: Asymmetry 1997, 8, 1895-1946. Kobayashi, S.; Ishitani, H.
Chem. ReV. 1999, 99, 1069-1094.
(
35) Moody, C. J.; Gallagher, P. T.; Lightfoot, A. P.; Slawin, A. M. Z. J. Org.
Chem. 1999, 64, 4419-4425. Hanessian, S.; Yang, R. Y. Tetrahedron Lett.
3
6 in which the allyl group was oxidatively transformed into a
1
996, 37, 8997-9000. Greven, R.; J u¨ tten, P.; Scharf, H. D. Carbohydr.
Res. 1995, 275, 83-93. Marco, J. A.; Carda, M.; Murga, J.; Gonz a´ lez, F.;
Falomir, E. Tetrahedron Lett. 1997, 38, 1841-1844. Marco, J. A.; Carda,
M.; Murga, J.; Rodr ´ı guez, S.; Falomir, E.; Oliva, M. Tetrahedron:
Asymmetry 1998, 9, 1679-1701. Nicolaou, K. C.; Mitchell, H. J.; Jain, N.
F.; Bando, T.; Hughes, R.; Winssinger, N.; Natarajan, S.; Koumbis, A. E.
Chem.-Eur. J. 1999, 5, 2648-2667.
primary alcohol to maximize signal dispersion (Figure 4). As
indicated, the signals for the protons on the primary alcohol
chain were enhanced upon irradation of the H4 proton syn to
H5, whereas the isobutyl methylene was enhanced upon
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5379

A R T I C L E S
Fuller et al.
Table 3. Addition of Organolithium Reagents to Isoxazolines
Figure 4. NOE difference experiment to confirm facial selectivity of
addition.
irradiation of H4 syn to the C5 hydroxyethyl substituent,
consistent with the stereochemical assignment indicated.
Similar reaction conditions were successfully employed with
several different stabilized Grignard reagents (entries 4-6, Table
2
) to provide isoxazolidines containing the requisite quaternary
center at C3 in good to excellent yields (83-96% yield).
Predictably, increasing the steric bulk of the nucleophile
corresponded to an increase in diastereoselectivity observed (up
to >20:1 dr). To obtain â2 -amino acid precursors, an
additional substituent at C4 of the isoxazoline can readily be
incorporated, as in entries 7 and 8. When the C4 group is syn
to the C5 substituent, only a single diastereomer is observed in
the nucleophilic addition reaction (entry 7). When the substit-
uents are anti, however, the diastereoselectivity is further
influenced by the C3 substituent (entry 8). In the case of 43a
in which both C3 and C4 bear an ethyl group, a 1:1 mixture of
diastereomeric isoxazolidines is isolated. In contrast, an increase
in the size of C3 as in 43b leads to a completely diastereose-
lective reaction; an NOE difference study of an acetylated
version of 44b confirmed exclusive approach of the nucleophile
from the si face (see Supporting Information for details).
Analogous selectivity was observed with larger C4 substituents
such as benzyl (C3 ) Et, 1.2:1 dr; C3 ) i-Bu, single
diastereomer). This suggests that the steric interplay between
the C3 and C4 substituents selectively shields the re face of
the isoxazoline. These findings indicate that, through judicious
choice of substituents at these positions, the stereochemistry at
C3 can be set independently of C4, enabling access to an even
,3,3
a
1
Determined by H NMR spectral integration of crude product mixture.
2
,3,3
wider range of â -amino acid precursors.
Initial attempts to expand this reaction beyond stabilized
Grignard reagents were unsuccessful as the use of aryl or alkyl
organometallic agents provided only recovered starting material.
Given the sites available for deprotonation within the isoxazo-
with this modification, nonstabilized Grignard reagents remained
ineffective. Again employing BF3‚OEt2 as a Lewis acid,
methyllithium and phenyllithium as well as heteroaryllithium
reagents were suitable nucleophiles for this transformation. In
contrast to the reactions with Grignard reagents in Table 2, the
use of toluene as solvent provided the best combination of yield
and selectivity. As shown in entries 6 and 7, additional
substitution could be incorporated into the isoxazoline to prepare
vicinally substituted isoxazolidines in excellent diastereoselec-
tivity; in the case of entry 7, however, the yield was lower
despite prolonged reaction times, likely due to additional steric
hindrance about the reactive center. An NOE difference study
on a peracetylated derivative of 48 provided evidence that the
approach of the nucleophile is opposite the sterically bulky C5
substituent (see Supporting Information for details).
2
8,36
line, competitive deprotonation appeared a likely culprit.
In support of this, a deuterium quench (D2O) of a reaction with
phenyllithium/BF3‚OEt2 and 23 produced an isoxazoline labeled
with deuterium at the C4 position. Because intramolecular proton
transfer from the C5 hydroxyethyl group was a possible
contributor, we explored protection of this group. This would
also have the added advantage of increasing the steric bulk at
this position and potentially further increasing the diastereose-
lectivity of addition. Of the various protecting groups screened
to serve this purpose, the TBS ether was most effective because
of both the stability of this protecting group under the Lewis
acidic reaction conditions and the ease of its removal in
subsequent steps.
It remained for us to convert these diverse isoxazolidines,
with all of the requisite substitution and stereochemical relation-
ships installed, to the target â-amino acids (Figure 5). The
synthetic plan entailed reduction of the N-O bond followed
by protection of the nitrogen with a carbamate protecting group.
In several of the substrates, the nitrogen is benzylic; thus we
sought conditions that were mild enough such that these
compounds would not be over-reduced. Use of 10% Pd/C with
With the side-chain hydroxyl masked, organolithium reagents
could be employed as nucleophiles to provide the isoxazolidine
adducts in excellent diastereoselectivity (Table 3), although even
(
36) Annunziata, R.; Cinquini, M.; Cozzi, F.; Raimondi, L. Tetrahedron 1986,
4
2, 2129-2134. Shatzmiller, S.; Shalom, E.; Lidor, R.; Tartkovski, E.
Liebigs Ann. Chem. 1983, 906-912. Curran, D. P.; Chao, J. C. J. Am.
Chem. Soc. 1987, 109, 3036-3040. Kozikowski, A. P.; Ghosh, A. K. J.
Org. Chem. 1984, 49, 2762-2772. Grund, H.; J a¨ ger, V. Liebigs Ann. Chem.
37
38
NH4O2CH, H2-Pd(OH)2 in HCl-MeOH, or SmI2 conditions
gave incomplete conversions and/or inconsistent yields. How-
1980, 80-100. J a¨ ger, V.; Schwab, W. Tetrahedron Lett. 1978, 3129-3132.
5380 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Succinct Synthesis of
â-Amino Acids
A R T I C L E S
Figure 6. Retrosynthetic analysis of a cis-â-proline analogue. In this
example, the latent nucleophile is an aryl bromide. By choosing a different
electrophile in the alkylation reaction, the synthesis of cis-â-proline variants
containing a range of substitution patterns can be envisioned.
desirable. Also noteworthy among the structures we have
synthesized are the substituents that can readily be transformed
into a variety of other functional groups using straightforward
means. The allyl and methylallyl groups, for example, are readily
oxidized to hydroxyl and carbonyl groups (for example, see
Figure 4), themselves excellent precursors to amines and
guanidinium functionality. The methylfuryl group is also readily
converted to a carboxylic acid under mild oxidative conditions.42
Thus, a full range of substituents and substitution patterns can
be accessed using this synthetic strategy.
Synthesis of cis-â-Prolines: Conformationally Constrained
â2 -Amino Acids. As outlined in the Introduction, oligomers
of cispentacin (cis-2-aminocyclopentanoic acid, 4) form
,3,3
Figure 5. Conversion of isoxazolidines to â3,3- and â2,3,3-amino acids.
Reaction conditions: (a) (i) LiAlH4, Et2O, 0 °C to room temperature, 3 h;
13
â-strands, a common protein fold. Cispentacin itself is a natural
4
3
(ii) BOC2O, THF/H2O, pH 10, 12 h. (b) (i) NaIO4, CH3CN/H2O (1:1), room
product, originally isolated independently from Bacillus cereus
and Streptomyces setonii
temperature, 0.5 h; (ii) NaClO2, 2-methyl-2-butene, KH2PO4, t-BuOH, H2O,
room temperature, 2 h. (c) N-O bond cleavage effected by 10% Pd/C,
NH4O2CH. (d) Amine protected as benzyl carbamate following N-O bond
reduction: CbzCl, Na2CO3, THF/H2O (5:1), 12 h.
43,44
45
and possessing antifungal activity.
From both a structural and a biological activity standpoint,
cispentacin analogues are thus synthetic targets of great inter-
4
6
est. The most common synthetic approach employs a [2+2]-
ever, treatment of the isoxazolidines with LiAlH4 in Et2O cleanly
provided the target amine diols that were immediately protected
to facilitate isolation. In most cases, a BOC group was used,
but in the case of 37 and 54, the amine diol proved resistant to
protection, probably because of the significant steric encum-
brance of the nitrogen. For these examples, the sterically less
demanding benzyl carbamate (CBZ) protecting group proved
successful. In contrast to the conditions employed earlier for
cycloaddition to form a â-lactam that is then hydrolyzed to the
11,47
desired â-amino acid.
However, the stereoselective instal-
lation of additional substituents within the â-proline backbone,
particularly at the â-position, is difficult by this or other
48
methods. To directly address these points, we envisioned that
isoxazolines bearing a substituent at C4 containing a nascent
nucleophilic group would be versatile intermediates for the
preparation of this structural class as an intramolecular nucleo-
philic attack would generate the requisite cis-substituted cyclo-
pentane with a nitrogen-bearing quaternary center at C3. A
retrosynthetic analysis of a benzofused cispentacin in accordance
with this strategy is depicted in Figure 6. Isoxazolines bearing
3
2,3
the â - and â -amino acids, we found that a two-step procedure
employing NaIO4 to cleave the diol followed by standard
NaClO2 conditions to oxidize the intermediate aldehyde pro-
ceeded with shorter reaction times and greater overall yields of
the suitably protected â-amino acids, perhaps due to the shorter
exposure of potentially sensitive side chains to the oxidizing
C4 substituents are readily available by alkylation of the core
28,36
ring structure as shown,
with substitution along the alkylating
3
9
conditions.
As illustrated by the â - and â2,3,3-amino acids in Figure 5,
the isoxazoline-based synthetic strategy is remarkably effective
for the preparation of a wide range of structures. One of the
most notable features of this approach is that the C3 substituents
in the â-amino acid are not required to be sterically dissimilar
3,3
(41) For elegant approaches to â-amino acids in which substituents are sterically
dissimilar, see: Tang, T. P.; Ellman, J. A. J. Org. Chem. 1999, 64, 12-
1
3. Gais, H.-J.; Loo, R.; Roder, D.; Das, P.; Raabe, G. Eur. J. Org. Chem.
2
003, 1500-1526. Kobayashi, K.; Matoba, T.; Irisawa, S.; Takanohashi,
A.; Tanmatsu, M.; Morikawa, O.; Konishi, H. Bull. Chem. Soc. Jpn. 2000,
7
3, 2805-2809. Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc.
2
000, 122, 8180-8186.
(
42) Borg, G.; Chino, M.; Ellman, J. A. Tetrahedron Lett. 2001, 42, 1433-
40
for stereoselective synthesis to be achieved, as is most often
1436.
the case with other synthetic strategies.2
1,41
(43) Konishi, M.; Nishio, M.; Saitoh, K.; Miyaki, T.; Oki, T.; Kawaguchi, H.
J. Antibiot. 1989, 42, 1749-1755.
An excellent example
of this is 59 with a benzyl and isobutyl group at C3; in the key
stereocenter-forming step, the diastereoselectivity achieved is
(44) Iwamoto, T.; Tsujii, E.; Ezaki, M.; Fujie, A.; Hashimoto, S.; Okuhara, M.;
Kohsaka, M.; Imanaka, H.; Kawabata, K.; Inamoto, Y.; Sakane, K. J.
Antibiot. 1990, 43, 1-7.
1
8:1. Many of the side chains present in natural R-amino acids
(45) Oki, T.; Hirano, M.; Tomatsu, K.; Numata, K. I.; Kamei, H. J. Antibiot.
1
989, 42, 1756-1762.
are sterically similar, and thus this class of targets is particularly
(
46) Ohki, H.; Inamoto, Y.; Kawabata, K.; Kamimura, T.; Sakane, K. J. Antibiot.
1991, 44, 546-549.
(
37) Beccalli, E. M.; Broggini, G.; Farina, A.; Malpezzi, L.; Terraneo, A.; Zecchi,
(47) F u¨ l o¨ p, F.; Palk o´ , M.; K a´ m a´ n, J.; L a´ z a´ r, L.; Sillanp a¨ a¨ , R. Tetrahedron:
Asymmetry 2000, 11, 4179-4187. K a´ m a´ n, J.; Forr o´ , E.; F u¨ l o¨ p, F.
Tetrahedron: Asymmetry 2000, 11, 1593-1600. Szakonyi, Z.; F u¨ l o¨ p, F.;
Bern a´ th, G.; Evanics, F.; Riddell, F. G. Tetrahedron 1998, 54, 1013-1020.
(48) Gees, T.; Schweizer, W. B.; Seebach, D. HelV. Chim. Acta 1993, 76, 2640-
2653. Hanselmann, R.; Zhou, J. C.; Ma, P.; Confalone, P. N. J. Org. Chem.
2003, 68, 8739-8741. Aggarwal, V. K.; Roseblade, S. J.; Barrell, J. K.;
Alexander, R. Org. Lett. 2002, 4, 1227-1229. Konosu, T.; Oida, S. Chem.
Pharm. Bull. 1993, 41, 1012-1018.
G. Eur. J. Org. Chem. 2002, 2080-2086.
(
(
38) Bode, J. W.; Carreira, E. M. Org. Lett. 2001, 3, 1587-1590.
39) Additional confirmation of the stereochemical assignments was obtained
through X-ray crystallographic analysis of acid 67. See ref 24 for details.
40) For rare examples, see ref 10 and: Espino, C. G.; Wehn, P. M.; Chow, J.;
Du Bois, J. J. Am. Chem. Soc. 2001, 123, 6935-6936. Satoh, T.; Fukuda,
Y. Tetrahedron 2003, 59, 9803-9810. Davis, F. A.; Deng, J. H.; Zhang,
Y. L.; Haltiwanger, R. C. Tetrahedron 2002, 58, 7135-7143.
(
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5381

A R T I C L E S
Fuller et al.
Table 4. Intramolecular Nucleophilic Addition
Figure 7. Conversion of isoxazolidines to conformationally constrained
2,3,3
â
-amino acids.
uents were treated with LDA to form an internal azaenolate
followed by addition of an electrophile such as 2-bromobenzyl
bromide (Table 4). The alkylation proceeds with excellent regio-
and diastereoselectivity with only a single isomer observed in
1
the crude product mixture by either GC or H NMR analysis.
An NOE study carried out on isoxazoline 71 confirmed that, as
expected, the alkylation event takes place on the face of the
isoxazoline opposite the sterically bulky C5 subsitutent, resulting
in a C4,C5-anti relationship in the product. We anticipated that
the aryl bromide would be converted upon treatment with tert-
butyllithium into a potent nucleophile poised for intramolecular
addition to the adjacent CdN bond. Indeed, this reaction was
found to proceed smoothly, and the triethylsilyl protecting
groups in the product were removed conveniently by quenching
the reaction with acid to provide tricyclic isoxazolidines (Table
4
). In contrast to the intermolecular nucleophilic additions
described in previous sections, addition of an intramolecular
nucleophile does not require a Lewis acid, and the yields are
nearly identical with or without added BF3‚Et2O. The addition
1
is similarly tolerant of R side chains of varying steric and
electronic characteristics. Further, the nucleophile itself can be
electron-deficient (entry 5) or bear electron-donating substituents
(
entry 6) without deleterious effects on the yield. The example
a
b
(
i) LDA, THF, -78 °C, 1.5 h; (ii) RCH2Br, 2 h. (i) t-BuLi, THF,
+
c
in entry 5 is particularly noteworthy as the pyridine ring serves
to increase water solubility of the product, useful for later
biological applications, and also provides a hydrogen-bond
acceptor within the backbone for assembly purposes.
-
78 °C; (ii) H3O , 0 °C to room temperature. To facilitate isolation of
this very polar isoxazolidine, the TES group was not removed after the
nucleophilic addition reaction.
agent backbone providing additional functional groups within
the â-proline product.49 As with the acyclic â-amino acid
variants, the absolute stereochemistry of the final â-amino acid
product is dictated by the initial choice of allylic alcohol absolute
stereochemistry, enabling ready access to either enantiomer.
Using the conditions previously identified, it was straight-
forward to convert the highly functionalized, tricyclic isoxazo-
lidines to the corresponding â-amino acids. The isoxazolidine
N-O bond was reduced with LiAlH4, and, due to the steric
hindrance about the resulting amine, protection of this functional
group was best accomplished using benzyl chloroformate to
provide a benzyl carbamate moiety. A two-step oxidation
procedure (NaIO4 then NaClO2) was likewise effective for
oxidative cleavage of the diol moiety, providing the â-amino
acid products in good to excellent overall yields (Figure 7). The
ready synthesis of these â-proline examples illustrates the
versatility of the isoxazoline-based synthetic approach for this
functionally dense compound class. With the additional substitu-
tion, these structures are promising candidates for developing
â-peptides that form higher order structures.
The cis-â-proline variants selected as synthetic targets contain
both a quaternary center as well as a fused aryl or heteroaryl
group, lending further conformational rigidity to the final
product in addition to added positions for functionalization. To
initiate the synthesis, isoxazolines bearing various C3 substit-
(
49) Alternatively, employing a 1,2-disubstituted olefin in the initial 1,3-dipolar
cycloaddition to form the isoxazoline can be used to install a C4 substituent.
This approach is generally less desirable for producing a diversity of
isoxazolines, however, because the yields of the cycloaddition often decrease
as olefin substitution increases, and, further, the allylic alcohols are less
readily available.
5382 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Succinct Synthesis of
â-Amino Acids
A R T I C L E S
Conclusions
both structural and biological studies. In addition to the valuable
â-amino targets addressed here, we also anticipate that the
general strategy will find even broader use. In particular, we
have shown that isoxazolines are useful precursors for targets
containing tertiary and quaternary amine-bearing stereocenters;
thus one can envision using isoxazolines en route to other
synthetic targets including various nitrogen-containing phar-
maceuticals and alkaloid natural products.
We have described here a versatile and efficient approach
for the synthesis of a diverse array of â-amino acids. The
approach employs chiral isoxazolines as the centerpiece,
translating the high diastereoselectivity available from a 1,3-
dipolar cycloaddition reaction into the formation of structures
with up to four contiguous stereocenters. The initial implemen-
3
2,3
tation provided access to â - and â -amino acids, classes of
â-amino acids that have been widely employed in pharmaceuti-
cal targets and peptidomimetics. When applied to the synthesis
of highly substituted â-amino acids, targets that demand the
installation of an amine-bearing quaternary stereocenter, the
method has proven to be especially valuable; the novel targets
described here significantly expand the set of C3-disubstituted
â-amino acids available for structure/function studies of â-pep-
tide oligomers. Elaborating on the same general strategy, we
have demonstrated that appropriately substituted isoxazolines
can be used to efficiently generate a high degree of molecular
complexity toward the preparation of cis-â-proline variants. Such
densely functionalized structures should find immediate use in
Acknowledgment. A.K.M is grateful for financial support
from the Burroughs Wellcome Fund (New Investigator in the
Toxicological Sciences) and the March of Dimes (Basil
O’Connor Starter Scholar). A.A.F. is an Eli Lilly & Co.
Graduate Fellow. A.R.M. was supported by the Chemistry/
Biology Interface Training Program (GM09597) and a Pfizer
graduate fellowship.
Supporting Information Available: Experimental procedures
and characterization data for all new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
JA0431713
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5383