Cyclodextrin-mediated deacylation of amino acid esters with marked stereoselectivity.

Article abstract of DOI:10.1248/cpb.50.1283

With respect to the hydrolysis (deacylation) of Z-D(L)-amino acid esters (N-(benzyloxycarbonyl)-D(L)-amino acid p-nitrophenyl esters) mediated by alpha-, beta- and gamma-cyclodextrins (CyDs), a remarkably high enantioselectivity (L/D=9.0) was observed for the deacylation of Ala substrate with gamma-CyD. The kinetic results on the basis of the Michaelis-Menten principle indicate that the enantioselectivity should be mainly originated in the deacylation process of substrates following the formation of gamma-CyD-substrate (1 : 1) complexes. The computer modeling (molecular mechanics) studies on the inclusion complexes are also described.

Full text of DOI:10.1248/cpb.50.1283

September 2002
Notes
Chem. Pharm. Bull. 50(9) 1283—1285 (2002)
1283
Cyclodextrin-Mediated Deacylation of Amino Acid Esters with Marked
Stereoselectivity
Koichi GOTO,a Kentaro NAKASHIMA,a Osamu TANOUE,a Satoshi NUKUSHINA,a Isao TOUDO,a
b
b
a
,a
*
Chikara IMAMURA, Yasuji IHARA, Yo ko M ATSUMOTO, and Ryuichi UEOKA
a Division of Applied Chemistry, Graduate School of Sojo University (former name: Kumamoto Institute of Technology);
4–22–1 Ikeda, Kumamoto 860–0082, Japan: and b Department of Human Environment, Yamaguchi Prefectural University;
3–2–1 Sakurabatake, Yamaguchi 753–8502, Japan.
Received May 1, 2002; accepted June 28, 2002; published online July 3, 2002
With respect to the hydrolysis (deacylation) of Z-D(L)-amino acid esters (N-(benzyloxycarbonyl)-D(L)-amino
acid p-nitrophenyl esters) mediated by a-, b- and g-cyclodextrins (CyDs), a remarkably high enantioselectivity
(L/D؍9.0) was observed for the deacylation of Ala substrate with g-CyD. The kinetic results on the basis of the
Michaelis–Menten principle indicate that the enantioselectivity should be mainly originated in the deacylation
process of substrates following the formation of g-CyD–substrate (1 : 1) complexes. The computer modeling
(molecular mechanics) studies on the inclusion complexes are also described.
Key words cyclodextrin; enantioselective hydrolysis; amino acid ester
Stereoselective hydrolysis of amino acid esters have at- Results and Discussion
tracted considerable attention in connection with understand-
With respect to the enantioselective hydrolysis (deacyla-
ing the origins of the stereoselectivity observed with prote- tion) of Z-D(L)-amino acid esters (Z-D(L)-Ala, Leu, Phe and
olytic enzymes. In the course of our study on the enantiose- Trp-PNP) mediated by unmodified a-, b- and g-CyDs, the
lective hydrolysis (deacylation) of amino acid esters with the kinetic parameters obtained on the basis of the Michaelis–
functional molecular assemblies composed of surfactants and Menten principle, the binding constant (Kb) for the formation
reactive species, we emphasized that the stereochemical con- of CyD-substrate complex and the rate constant (k2) for the
trol could be attained by changing the composition of the co- deacylation of substrate by CyD, are summarized in Table
aggregates1—4) and regulating ionic strength5,6) and tempera- 1—3. The noteworthy aspects are as follows: (a) Overall, the
ture.7,8)
deacylation mediated by a-CyD was favorable for all of the
Furthermore, since cyclodextrins (CyDs) have been noted L-enantiomeric substrates (reflected in k2Kb, L/Dϭ1.5—1.7),
as useful enzyme mimics,9—12) we employed CyDs as host though the Kb for D-isomer of Phe substrate and k2 values for
molecules and a markedly high stereoselectivity was attained D-isomer of Ala substrate were larger than those for the cor-
in the diastereoselective deacylation of dipeptide esters.13) On responding L-isomers. (b) In the case of the reaction by b-
the other hand, Ueno reported the high D-enantiomer-selec- CyD, L-enantiomer-selectivity was also observed for the dea-
tive deacylation of amino acid esters mediated by modified cylation of Ala, Leu and Phe substrates (L/Dϭ1.5—3.2),
b-CyDs.14) However, little has been known about the signifi- while D-enantiomeric selectivity (D/Lϭ1.4) was observed for
cant enantioselectivity for the deacylation of amino acid es- the deacylation of Trp substrate. (c) Relatively higher L-enan-
ters with unmodified CyDs.15)
tioselectivity was obtained in the deacylation mediated by g-
In this study, we report the successful experimental results CyD (L/Dϭ2.6—9.0). Most remarkably, the highest enantio-
with marked enantioselectivity for the deacylation of Z-D(L)- selectivity (L/Dϭ9.0) was attained for the deacylation of Ala
amino acid esters (N-(benzyloxycarbonyl)-D(L)-amino acid p- substrate. Furthermore, it was also attractive that this fairly
nitrophenyl esters) as mediated by unmodified CyDs, and the high enantioselectivity (L/Dϭ9.0) could be mainly originated
computer modeling (molecular mechanics) studies on the in- in the deacylation process of substrates, because it was re-
clusion complexes of g-CyD with the specific substrates are flected by a fairly large k2L/k2D value (7.4) and a small KbL/KbD
also presented.
(1.2).
On the other hand, Table 4 shows the results of energies
for the inclusion complexes on Z-D(L)-Ala-PNP having dif-
ferent shapes with g-CyD calculated by molecular mechanics
with water solvent effects. It was suggested that the confor-
mation of a hairpin shape (Fig. 2a) should be more favorable
as compared with the straight shape (Fig. 2b) for the inclu-
sion complex on D- or L-isomer and g-CyD. Furthermore, the
energy of the inclusion complex with L-isomer was lower
than that with D-isomer, which is in harmony with the Kb
value (L/Dϭ1.2) for the binding process on Z-D(L)-Ala-PNP
and g-CyD. Therefore, it was estimable that Z-D(L)-Ala-PNP
having a hairpin shape was encapsulated by g-CyD as in Fig.
3. Plausibly, the complex between g-CyD and L-isomer hav-
ing a hairpin shape should be more efficient for the next dea-
Fig. 1. Cyclodextrins (CyDs) and Z-D(L)-Amino Acid Esters
To whom correspondence should be addressed. e-mail: ueoka@life.sojo-u.ac.jp
© 2002 Pharmaceutical Society of Japan

1284
Vol. 50, No. 9
Table 1. Kinetic Parameters for the Deacylation of Z-D(L)-Amino Acid p-Nitrophenyl Esters Mediated by a-CyD
Ϫ1
k2 (sϪ1
L-Isomer
)
Kb (M
)
k2Kb (MϪ1 sϪ1
)
Substrate
ks (sϪ1
)
L/D
L/D
L/D
D-Isomer
L-Isomer
D-Isomer
L-Isomer
D-Isomer
Z-Ala-PNP
Z-Leu-PNP
Z-Phe-PNP
Z-Trp-PNP
0.00291
0.00155
0.00419
0.00215
0.0947
0.0118
0.106
0.166
1.8 (D/L)
1.1
3.5
1.99
11.2
8.12
2.35
0.778
4.99
6.97
4.41
2.6
1.4
2.1 (D/L)
1.6
0.188
0.132
0.249
0.210
0.129
0.0909
0.151
0.124
1.5
1.5
1.6
1.7
0.0112
0.0302
0.0282
0.0302
1.1
25 °C, pH 9.5, 0.02 M carbonate buffer (0.05 M KCl), 10% (v/v) CH3CN–H2O, [substrate]ϭ7.0ϫ10Ϫ6 M, [a-CyD]ϭ(1.0Ϫ10)ϫ10Ϫ3 M.
Table 2. Kinetic Parameters for the Deacylation of Z-D(L)-Amino Acid p-Nitrophenyl Esters Mediated by b-CyD
Ϫ1
k2 (sϪ1
L-Isomer
)
Kb (M
)
k2Kb (MϪ1 sϪ1
)
Substrate
ks (sϪ1
)
L/D
L/D
L/D
D-Isomer
L-Isomer
D-Isomer
L-Isomer
D-Isomer
Z-Ala-PNP
Z-Leu-PNP
Z-Phe-PNPa)
Z-Trp-PNP
0.00291
0.00155
0.00419
0.00215
0.140
0.0198
0.109
0.112
0.154
1.1 (D/L)
1.4
2.3
32.9
124
95.6
86.9
9.25
62.5
148
3.6
2.0
1.6 (D/L)
1.3 (D/L)
4.61
2.45
10.4
9.73
1.42
0.888
7.16
3.2
2.8
1.5
0.0142
0.0484
0.124
1.1 (D/L)
114
14.1
1.4 (D/L)
25 °C, pH 9.5, 0.02 M carbonate buffer (0.05 M KCl), 10% (v/v) CH3CN–H2O, [substrate]ϭ7.0ϫ10Ϫ6 M, [b-CyD]ϭ(1.0Ϫ10)ϫ10Ϫ3 M. a) [b-CyD]ϭ(0.10Ϫ10)ϫ10Ϫ3 M.
Table 3. Kinetic Parameters for the Deacylation of Z-D(L)-Amino Acid p-Nitrophenyl Esters Mediated by g-CyD
Ϫ1
k2 (sϪ1
L-Isomer
)
Kb (M
)
k2Kb (MϪ1 sϪ1
)
Substrate
ks (sϪ1
)
L/D
L/D
L/D
D-Isomer
L-Isomer
D-Isomer
L-Isomer
D-Isomer
Z-Ala-PNP
Z-Leu-PNP
Z-Phe-PNP
Z-Trp-PNP
0.00291
0.00155
0.00419
0.00215
0.264
0.0827
0.171
0.0357
0.0146
0.0529
0.0384
7.4
5.7
3.2
2.5
101
147
171
109
83.4
121
89.2
105
1.2
1.2
1.9
1.0
26.7
12.2
29.2
10.4
2.98
1.77
4.72
4.03
9.0
6.9
6.2
2.6
0.0957
25 °C, pH 9.5, 0.02 M carbonate buffer (0.05 M KCl), 10% (v/v) CH3CN–H2O, [substrate]ϭ7.0ϫ10Ϫ6 M, [g-CyD]ϭ(1.0Ϫ10)ϫ10Ϫ3 M.
Table 4. Energies on the Basis of Molecular Mechanics Calculations for
the Inclusion Complexes on Z-D(L)-Ala-PNP Having Different Shapes with
g-CyD in Water
Inclusion complex energy (kJ/mol)
Hairpin shape
Straight shape
L-Isomer
D-Isomer
254.70
261.78
262.42
274.99
Fig. 3. The Calculated Stable Structures of Inclusion Complex on Z-L-
Ala-PNP with g-CyD
Fig. 2. Calculated Stable Structures of Hairpin Shape (a) and Straight
Shape (b) on Z-L-Ala-PNP

September 2002
1285
tries for the molecular mechanics calculations with water solvent effects.
The molecular mechanics calculations were performed by the conjugate gra-
dient algorithm with OPLS all atom force fields20) as implemented in Macro-
Model 7.1.21) All of the calculations were carried out under the condition of
Max Iterationϭ10000 steps, Converge Thresholdϭ0.0010. Solvation was
achieved by using GB/SA solvation model22) inside MacroModel 7.1. Fur-
thermore, preliminary calculations for the formation of the intermediates in
deacylation process have been attempted by the molecular dynamics and
molecular mechanics methods.
cylation step as compared with that between g-CyD and the
corresponding D-isomer. More detailed modeling studies are
being continued in order to gain further insight into the ori-
gin of the marked enantioselectivity.
In conclusion, the remarkably high enantioselectivity
(L/Dϭ9.0) was observed for the hydrolysis (deacylation) of Z-
D(L)-Ala-PNP mediated by unmodified g-CyD for the first
time. The kinetic results on the basis of the Michaelis–
Menten principle indicate that this L-superior enantioselectiv-
ity should be mainly originated in the deacylation process of
substrates following the formation of g-CyD–substrate (1 : 1)
complexes. The computer modeling (molecular mechanics)
studies supported the L enantioselectivity on the formation of
inclusion complex with the more stable conformer of the
guest.
References and Notes
1) Ueoka R., Matsumoto Y., Moss R. A., Sugii A., Harada K., Kikuchi J.,
Murakami Y., J. Am. Chem. Soc., 110, 1588—1595 (1988).
2) Goto K., Matsumoto Y., Ueoka R., J. Org. Chem., 60, 3342—3346
(1995).
3) Ueoka R., Moss R. A., Swarup S., Matsumoto Y., Strauss G., Mu-
rakami Y., J. Am. Chem. Soc., 107, 2185—2186 (1985).
4) Ueoka R., Dozono H., Matsumoto Y., Moss R. A., Cho M., Kitahara
K., Kato Y., Chem. Pharm. Bull., 38, 219—220 (1990).
5) Tanoue O., Baba M., Tokunaga Y., Goto K., Matsumoto Y., Ueoka R.,
Tetrahedron Lett., 40, 2129—2132 (1999).
6) Ueoka R., Cho M., Matsumoto Y., Goto K., Kato Y., Harada K., Sugii
A., Tetrahedron Lett., 31, 5335—5338 (1990).
7) Matsumoto Y., Ueoka R., J. Org. Chem., 55, 5797—5799 (1990).
8) Ueoka R., Yamada E., Yamashita O., Matsumoto Y., Kato Y., Tetrahe-
dron Lett., 32, 6597—6600 (1991).
Experimental
Materials The ester substrates were prepared from N-(benzyloxycar-
bonyl)-D(or L)-amino acids by the esterification of the COOH group with p-
nitrophenol and dicyclohexylcarbodiimide as described in ref 1. These enan-
tiomeric isomers were fully characterized. Commercially available a-, b-
and g-CyDs were recrystallized from deionized water and dried in vacuum
at 40 °C for 24 h.
Kinetic Measurements Under the conditions [CyD]ϾϾ[substrate],
pseudo-first-order rate constants (kt in the presence of CyD and ks in the ab-
sence of CyD) for the deacylation of amino acid esters were evaluated from
monitoring p-nitrophenolate liberation from the esters at 400 nm. The stoi-
chiometry for the complexation of CyDs and substrates could not be con-
firmed by spectroscopic examinations (1H-NMR, UV and fluorescence mea-
surements). It was assumed that the reaction proceeds via the formation of
1 : 1 complex as shown in eq 1, and the Kb (ϭk1/kϪ1) and k2 values were de-
termined by the least-squares method from Lineweaver–Burk plots between
1/(ktϪks) and 1/[CyD] in eq 2 (correlation coefficients for the plots were
Ͼ0.994) as described in ref 16.
9) Komiyama M., Bender M. L., J. Am. Chem. Soc., 100, 4576—4579
(1978).
10) Trainor G. L., Breslow R., J. Am. Chem. Soc., 103, 154—158 (1981).
11) Ueno A., Moriwaki F., Osa T., Ikeda T., Toda F., Hattori K., Bull.
Chem. Soc. Jpn., 59, 3109—3112 (1986).
12) Kitaura Y., Bender M. L., Bioorg. Chem., 4, 237—249 (1975).
13) Ueoka R., Matsumoto Y., Harada K., Akahoshi H., Kato Y., J. Am.
Chem. Soc., 114, 8339—8340 (1992).
14) Hamasaki K., Ueno A., Chem. Lett., 1995, 859—860.
15) Ihara Y., Nakanishi E., Nango M., Koga J., Bull. Chem. Soc. Jpn., 59,
1901—1905 (1986).
k1
k2
ks
16) Taniguchi Y., Makimoto S., Suzuki K., J. Phys. Chem., 85, 3469—
3472 (1981).
→
→
CyDϩS
CyD-S
→acyl-CyDϩ products, S
→products
kϪ1
17) Halgren T. A., J. Comp. Chem., 17, 490—519, 520—552, 553—586,
616—641 (1996).
18) Halgren T. A., Nachbar R. B., J. Comp. Chem., 17, 587—615 (1996).
19) Halgren T. A., J. Comp. Chem., 20, 720—729, 730—748 (1996).
20) Jorgensen W. L., Maxwell D. S., Tirado-Rives J., J. Am. Chem. Soc.,
118, 11225—11236 (1996).
21) Mohamadi F., Richards N. G. J., Guida W. C., Liskamp R., Lipton M.,
Caufield C., Chang G., Hendrickson T., Still W. C., J. Comput. Chem.,
11, 440—467 (1990).
22) Still W. C., Tempczyk A., Hawley R. C., Hendrickson T., J. Am. Chem.
Soc., 112, 6127—6129 (1990).
(1)
(2)
1/(ktϪks)ϭ1/(k2Ϫks)ϩ1/{Kb(k2Ϫks)[CyD]}
Calculation Method Conformer search for the structures of Z-D(L)-Ala-
PNP and the inclusion complex with g-CyD was performed on the basis of
Monte Carlo method in conjunction with MMFF17—19) molecular mechanics.
Two minimum energy conformers of Z-D(L)-Ala-PNP were found, one had a
hairpin shape with the p-nitrophenyl group located parallel to the Z group
and the other had a straight shape with the p-nitrophenyl and Z groups lo-
cated in opposite directions. These conformers of Z-D(L)-Ala-PNP were
placed in the center of the cavity of g-CyD, and the most stable structures of
the inclusion complex were searched in order to obtain the starting geome-