Assembly state of catalytic modules as chiral switches in asymmetric strecker amino acid synthesis

Article abstract of DOI:10.1021/ja060841r

Self-assembled chiral polymetallic complexes often demonstrate novel properties as asymmetric catalysts. We report the three-dimensional structures of two such asymmetric catalysts (crystals A and B) for Strecker α,α-disubstituted amino acid synthesis. Th

Full text of DOI:10.1021/ja060841r

Published on Web 05/10/2006
Assembly State of Catalytic Modules as Chiral Switches in Asymmetric
Strecker Amino Acid Synthesis
Nobuki Kato,† Tsuyoshi Mita,† Motomu Kanai,*,† Bruno Therrien,‡,∇ Masaki Kawano,‡
Kentaro Yamaguchi,§ Hiroshi Danjo,§ Yoshihisa Sei,§ Akihiro Sato, Sanae Furusho, and
Masakatsu Shibasaki*,†
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-0033,
Japan, Graduate School of Engineering, The UniVersity of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-0033, Japan,
Faculty of Pharmaceutical Sciences at Kagawa Campus, Tokushima Bunri UniVersity, Shido, Sanuki-city, Kagawa
769-2193, Japan, and JASCO International Co., Ltd., 1-11-10 Myojin-cho, Hachioji, Tokyo 192-0046, Japan
Received February 4, 2006; E-mail: mshibasa@mol.f.u-tokyo.ac.jp
The functions of biomacromolecules are intimately related to
their three-dimensional structures, which are constructed through
hierarchically assembling individual modules. The fact that the
higher-order structures of biomacromolecules are determined by
noncovalent bond interactions is responsible for the flexibility of
life. This feature at a molecular level, however, is also thought to
be the cause of several diseases, especially neurodegenerative
diseases such as Creutzfeldt-Jacob disease and Alzheimer’s
disease.1 Proteins with multiple, thermodynamically stable, higher-
order structures of local energy minima could exhibit either normal
or aberrant functions, depending on the higher-order structure,
despite the fact that the primary structures are the same.
Figure 1. X-ray structure of crystal A (a) and its chemical structure (b).
Stereochemistry of the ligands and chloride atoms on the catechol moieties
are omitted in (b) for clarity.
Inspired by functional biomacromolecules, many artificial ma-
terials have been recently developed using a modular approach.2
Our work focuses on developing enzyme-like artificial bifunctional
asymmetric catalysts that can promote novel carbon-carbon bond-
forming reactions via dual activation of substrates and reagents at
defined positions using a catalyst.3 We previously reported that a
catalyst prepared from Gd(O-i-Pr)3 and a D-glucose-derived ligand
2 in a 1:2 ratio (Gd-2) promoted the cyanation of ketoimines 4
with high enantioselectivity.4,5 This reaction has made it possible
to access a wide range of enantiomerically enriched R,R-disubsti-
tuted amino acids, versatile chiral building blocks for biologically
active compounds.
On the basis of preliminary structural studies of the catalyst using
ESI-MS, both a 2:3 complex of Gd and the ligand and a 4:5+oxo
complex were found to be present in a catalyst solution, from which
the 2:3 complex was proposed to be the active catalyst.6 The two
Gd metals in the asymmetric catalyst should cooperate as a Lewis
acid and a nucleophile activator, respectively.3 This information
led to the recent finding that a new catalyst preparation method
using Gd(HMDS)3 (HMDS ) hexamethyldisilazane) and 2 in a
2:3 ratio (Gd*-2) resulted in the formation of the 2:3 complex as
the sole catalytically active species, which improved enantioselec-
tivity and catalyst activity.6 Here, we present the three-dimensional
structural characterization of two catalyst complexes, one (crystal
A) producing the opposite enantiomer with excellent selectivity
compared to that obtained by the catalyst prepared in situ, and the
other (crystal B) containing a 2:3 complex as a subunit and thus
reflecting the optimized Gd catalyst in solution. Analogous to
common phenomena in biological system, higher-order structural
differences in an artificial asymmetric catalyst led to a dramatic
change in its function.
Although Gd*-2 produced optimum results in the enantioselective
Strecker reaction, we screened various crystallization conditions
to obtain single crystals of the catalytically active species, com-
bining several related chiral ligands (e.g. 1-3) and rare earth metal
sources. Colorless, air-stable prisms were obtained from a propi-
onitrile-hexane (2:1) solution of the complex prepared from Gd-
(O-i-Pr)3 and ligand 3 in a 2:3 ratio (crystal A: 80% yield). X-ray
crystallographic analysis revealed that crystal A was a 4:5 complex
of Gd and 3 with a µ-oxo atom surrounded by four Gd atoms
(Figure 1).7 The tetranuclear structure was maintained in a solution
state,8 and the sole peak observed by ESI-MS corresponded to
crystal A. Thus, crystal A was distinct from catalysts prepared in
situ, such as Gd*-2, containing Gd and the chiral ligand in a 2:3
ratio. The difference in the assembly state of the chiral modules
dramatically affected the catalytic function. When crystal A was
used as a catalyst in the Strecker reaction of ketoimines, enanti-
oselectivity was completely reversed compared to the catalyst
prepared in situ (Table 1, entries 4, 10, 13, 15, and 17). The reaction
rate was approximately 5-50 times slower than that using the
catalyst prepared in situ.9 The enantioswitching was not attributed
to the chiral ligand 3, because the in situ prepared Gd-3 gave the
“normal” (S)-products (Table 1, entries 3, 9, 12, and 16). Thus,
the “aberrant” (R)-product formation using crystal A was attributed
to the change in the assembly mode of the chiral modules through
the crystallization process.
† Graduate School of Pharmaceutical Sciences, The University of Tokyo.
‡ Graduate School of Engineering, The University of Tokyo.
§ Faculty of Pharmaceutical Sciences at Kagawa Campus, Tokushima Bunri
University.
Further efforts to obtain single crystals relevant to the catalyst
prepared in situ led to the isolation of a second crystal type.
Colorless, air-stable prisms (crystal B) grew from a THF solution
of La(O-i-Pr)3 and ligand 1 mixed in a 2:3 ratio (47% yield). The
X-ray diffraction study revealed the structure to be a unique pseudo
JASCO International Co., Ltd.
∇ Current address: Universite´ de Neuchaˆtel, case postale 2, CH-2007 Neuchaˆtel,
Switzerland.
9
6768
J. AM. CHEM. SOC. 2006, 128, 6768-6769
10.1021/ja060841r CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Catalytic Enantioselective Strecker Reaction of
Ketoimines
ligand ) 2:3 and 4:5+oxo complexes were observed (Figure 2b).
Crystal B (8 mol % La) promoted the catalytic, enantioselective
Strecker reaction of 2a, giving the product with 26% ee (Table 1,
entry 5). Configuration of the major product was the same (S) as
when using a catalyst prepared in situ (e.g., Gd*-2). Crystal B
produced a similar level of enantioselectivity as when using a
catalyst prepared in situ from La(O-i-Pr)3 and 1 (La-1: entry 6).
Therefore, crystal B represents one of the structures of the
lanthanum catalyst in solution.
Although crystallization of the 2:3 Gd complex that gives optimal
enantioselectivity in the Strecker reaction has not been successful,
the ESI-MS fragment patterns of crystal B provides insight into
the relevance of crystal B to the optimized Gd catalyst in solution.11
The observed fragments (2:3 and 4:5+oxo) can be viewed as build-
ing domains of the whole polymetallic assembly. The X-ray struc-
ture of crystal B supports this consideration: the distance between
La2 and La3 atoms (or La4 and La5 atoms) was significantly longer
than that between La1 and La2 (or La5 and La6) (Figure 2b). Thus,
the 6:8 complex (crystal B) is likely to be constructed via assembly
of the 2:3 subunit (shaded red in Figure 2b) and the 4:5+oxo subunit
(shaded blue). We previously reported that there is a correlation
between the enantioselectivity and population of the 2:3 Gd complex
in ESI-MS analysis.6 Together, this previous observation along with
the present structural information suggests that the terminal 2:3
subunits of crystal B correspond to the optimized active catalyst
(Gd*-2) of the enantioselective Strecker reaction.
These results demonstrate, for the first time with three-dimen-
sional structural elucidation, that function (enantioselectivity) of
an artificial asymmetric catalyst is tunable, depending on the assem-
bly mode of the chiral modules. This finding will accelerate the
development of on-demand artificial asymmetric catalysts with
switchable functions, via controlling the higher-order modular
assembly.
a In entry 2, the catalyst was prepared from Gd(HMDS)3 and ligand 2
in a 2:3 ratio. In other entries, the catalyst was prepared from a lanthanide
(either Gd or La) isopropoxide. b Determined by chiral HPLC. c Absolute
configuration was determined as shown. In other entries, the absolute
configuration was temporarily assigned.
Acknowledgment. Financial support was provided by a Grant-
in-Aid for Specially Promoted Research of MEXT.
Supporting Information Available: Experimental procedures,
X-ray diffraction data, and ESI-MS charts. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Dobson, C. M. Nature 2003, 426, 884.
(2) For example, see: Kato, T.; Mizoshita, N.; Kishimoto, K. Angew. Chem.,
Int. Ed. 2006, 45, 38.
(3) Kanai, M.; Kato, N.; Ichikawa, E.; Shibasaki, M. Synlett 2005, 1491.
(4) (a) Masumoto, S.; Usuda, H.; Suzuki, M.; Kanai, M.; Shibasaki, M. J.
Am. Chem. Soc. 2003, 125, 5634. (b) Kato, N.; Suzuki, M.; Kanai, M.;
Shibasaki, M. Tetrahedron Lett. 2004, 45, 3147. (c) Kato, N.; Suzuki,
M.; Kanai, M.; Shibasaki, M. Tetrahedron Lett. 2004, 45, 3153.
(5) Catalytic enantioselective Strecker reaction of ketoimines reported from
other groups, see: (a) Vachal, P.; Jacobsen, E. N. Org. Lett. 2000, 2,
867. (b) Vachal, P.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124, 10012.
(c) Chavarot, M.; Byrne, J. J.; Chavant, P. Y.; Valle´e, Y. Tetrahedron:
Asymmetry 2001, 12, 1147.
(6) See Supporting Information (SI) for details. For advantages in using Gd*-2
to Gd-2, see: (a) Fukuda, N.; Sasaki, K.; Sastry, T. V. R. S.; Kanai, M.;
Shibasaki, M. J. Org. Chem. 2006, 71, 1220. (b) Suzuki, M.; Kato, N.;
Kanai, M.; Shibasaki, M. Org. Lett. 2005, 7, 2527.
(7) For detailed X-ray and MS data, see SI.
(8) Studies to elucidate the complex structures in solution using NMR did
not give any useful information even when nonparamagnetic metals such
as Y or La were used (no significant signals were observed).
(9) Due to this significant difference in catalyst activity between the 2:3
complex and the 4:5+oxo complex, the Strecker products were obtained
with high enantioselectivity in (S)-configuration even using Gd-1, 2, or 3
prepared from Gd(O-i-Pr)3, which contain both the 2:3 and 4:5+oxo
complexes (Table 1).
Figure 2. X-ray structure of crystal B (a) and its chemical structure (b).
C2-symmetric 6:8 complex of La and 1 (Figure 2a).7 Six La atoms
were arrayed in line as a backbone ornamented with eight chiral
ligands. This is the longest linear homopolymetallic lanthanide array
in one molecule (maximum distance ) 2.75 nm) among the
enantiomerically pure lanthanide complexes reported to date.10 This
structure was maintained in solution, based on the fact that the
parent peak corresponding to crystal B was observed by ESI-QFT-
MS.7 In addition, two major fragment peaks corresponding to metal:
(10) For reviews, see (a) Buu¨nzli, J. G.; Piguet, C. Chem. ReV. 2002, 102,
1897. (b) Aspinall, H. C. Chem. ReV. 2002, 102, 1807.
(11) The lower enantioselectivity of crystal B than that of Gd-1 or 2 might be
due to differences in the metal character and not to the basic structure of
the catalytically active subunit. Very similar ESI-QFT-MS fragment
patterns were observed from crystal B, Gd-1, and Gd-2. See SI.
JA060841R
9
J. AM. CHEM. SOC. VOL. 128, NO. 21, 2006 6769