Cleavage of β-lactone ring by serine protease. Mechanistic implications

Article abstract of DOI:10.1016/S0968-0896(02)00108-6

Both enantiomers of 3-benzyl-2-oxetanone (1) were found to be slowly hydrolyzed substrates of α-chymotrypsin having k_cat values of 0.134±0.008 and 0.105±0.004 min^-1 for (R)-1 and (S)-1, respectively, revealing that α-CT is virtually unable to differentiate the enantiomers in the hydrolysis of 1. The initial step to form the acyl-enzyme intermediate by the attack of Ser-195 hydroxyl on the β-lactone ring at the 2-position in the hydrolysis reaction may not be enzymatically driven, but the relief of high ring strain energy of β-lactone may constitute a major driving force. The deacylation step is also attenuated, which is possibly due to the hydrogen bond that would be formed between the imidazole nitrogen of His-57 and the hydroxyl group generated during the acylation in the case of (R)-1, but in the α-CT catalyzed hydrolysis of (S)-1 the imidazole nitrogen may form a hydrogen bond with the ester carbonyl oxygen.

Full text of DOI:10.1016/S0968-0896(02)00108-6

Bioorganic & Medicinal Chemistry 10 (2002) 2553–2560
Cleavage of ꢀ-Lactone Ring by Serine Protease.
Mechanistic Implications
Dong H. Kim,* Jeong-il Park, Sang J. Chung, Jung Dae Park,
No-Kyung Park and Jong Hoon Han
Division of Molecular and Life Sciences, Center for Integrated Molecular Systems, National Research Laboratory for
Advanced Technology and Biomedical Microinstrumentation, Pohang University of Science and Technology,
San 31 Hyojadong, Pohang, South Korea 790-784
Received 7 January 2002; accepted 23 March 2002
Abstract—Both enantiomers of 3-benzyl-2-oxetanone (1) were found to be slowly hydrolyzed substrates of a-chymotrypsin having
kcat values of 0.134Æ0.008 and 0.105Æ0.004 minÀ1 for (R)-1 and (S)-1, respectively, revealing that a-CT is virtually unable to dif-
ferentiate the enantiomers in the hydrolysis of 1. The initial step to form the acyl-enzyme intermediate by the attack of Ser-195
hydroxyl on the b-lactone ring at the 2-position in the hydrolysis reaction may not be enzymatically driven, but the relief of high
ring strain energy of b-lactone may constitute a major driving force. The deacylation step is also attenuated, which is possibly due
to the hydrogen bond that would be formed between the imidazole nitrogen of His-57 and the hydroxyl group generated during the
acylation in the case of (R)-1, but in the a-CT catalyzed hydrolysis of (S)-1 the imidazole nitrogen may form a hydrogen bond with
the ester carbonyl oxygen. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
sition states that are generated in the processes of the
formation of acyl-enzyme intermediate and its sub-
sequent hydrolysis are stabilized through hydrogen
bonding interactions with two backbone NHs of Ser-
195 and Gly-193, which together are referred to as the
oxyanion hole.1,5 At the S1 subsite of a-CT there is pre-
sent a large hydrophobic pocket that accommodates the
P1 amino acid residue having a bulky hydrophobic side
chain such as Phe. As a prototypical enzyme for a large
family of serine proteases, a-CT has been served as a
model target for the development of inhibitor design
strategies that can be of value in designing inhibitors for
proteases of medicinal interest.6
Serine proteases that may be represented by a-chymo-
trypsin (a-CT)1 have received an increasing attention in
recent years because of their roles in the pathology of
numerous diseases such as rheumatoid arthritis,2
thrombosis,3 and pulmonary emphysema.4 These
enzymes are characterized by having a catalytically
essential serine residue at the active site. The hydroxyl
group of the serine residue attacks in collaboration with
His and Asp the scissile peptide or ester bond of sub-
strates, generating an acyl-enzyme intermediate which
is, in turn, hydrolyzed by the active site water molecule
to yield the second product with regeneration of the free
enzyme.1 These three amino acid residues which are
directly involved in the enzymic hydrolysis are generally
referred to as the catalytic triad. In the case of a-CT,
Ser-195 is the essential serine residue which together
with His-57 and Asp-102 constitute the catalytic triad.
The His-57 functions as a general base, activating the
nucleophilic hydroxyl of Ser-195 and the water molecule
that is required for the deacylation reaction. The tran-
Recently, we have reported that 3-benzyl-4-bromo-
methyloxetan-2-one is a fast acting alternate substrate
inhibitor7 for a-CT.8 In a continued effort in this line of
study, we have synthesized enantiomerically pure (R)-
and (S)- forms of 3-benzyl-2-oxetanone (1) and investi-
gated their chemistry at the active site of a-CT. We
anticipated that (R)-1 would be a substrate for a-CT as
its stereochemistry corresponds to the stereochemistry
of substrate, but its enantiomer, i.e., (S)-1 may function
as an irreversible inactivator for the enzyme. Because of
its stereochemistry (S)-1 may bind to a-CT at the active
site with its lactone ring being rested in such a way that
*Corresponding author. Fax: +82-54-279-5877; e-mail: dhkim@post-
ech.ac.kr
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00108-6

2554
D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
the activity curve in the plot of activity remaining versus
time (Fig. 2). About 75% of the enzyme activity was
recovered in 17 h. The experimental results suggest that
compound 1 is a substrate for a-CT albeit poor, and
furthermore that the cleavage of the lactone ring occurs
possibly in two steps with formation of an intermediate
that may be formed by the cleavage of the lactone ring
of 1 by a nucleophile present at the active site of the
enzyme. This has been confirmed by the electrospray
ionization mass spectrometry (ESI-MS): The native
a-CT and the intermediate formed by the interaction of
a-CT with each enantiomer of 1 showed a peak at
25.234 and 25.397 Da, respectively (Fig. 3). The peak of
25.397 Da corresponds to a-CT+1, indicating that the
intermediate is the addition product generated most
likely by the nucleophilic attack of the catalytic Ser-195
hydroxyl group on the lactone ring as anticipated. The
nucleophilic attack may take place either at the C-2 or
C-4 position.9 However, since both enantiomers were
shown to be substrates for a-CT, the possibility that the
intermediates may have an ether linkage by the attack at
the C-4 of the lactone ring is unlikely, because if the
latter intermediates are formed, they would resists
undergoing further reactions, causing inactivation of the
enzyme.12,13 A positive active site protection shown in
the competitive binding assay with an excess substrate
established that both enantiomers of 1 bind the enzyme
at the active site (Fig. 4). The KM values14 for binding of
(R)-1 and (S)-1 to a-CT and the ka for the formation of
the intermediates were estimated from the respective
Kitz–Wilson plot15 constructed using the kobs values
obtained for the initial 150and 250s for ( S)-1 and
(R)-1, respectively, after the addition of a-CT to a solu-
tion of succinyl-Ala-Ala-Pro-Phe-p-nitroanilide16 (Suc-
AAPF-p-NA) and optically active 1 (Figs 5 and 6).
Dissociation constants (KS) of the Michaelis complexes
were estimated from the respective Dixon plot (Fig. 7).17
The stabilities of the acylated a-CT were determined by
the proflavin displacement assay method:18 A rapid
decrease of absorbance at 466 nm due to the inter-
mediate formation was followed by a slow return of the
absorption for the each enantiomer as can be seen from
Figure 8. The rate constant for the recovery of the
enzymic activity (kd) was estimated from the progress
curves (Fig. 8). The kinetic data thus obtained are col-
lected in Table 1.19
the Ser-195 hydroxyl would attack the ring at the 4-
position rather than the 2-position to result in the for-
mation of an ether bond that resists to be cleaved. The
postulated configuration-dependent reaction paths of 1
at the active site of a-CT are depicted schematically in
Figure 1. The 2-oxetanone ring is an inherently reactive
moiety by virtue of its high ring strain energy
(22.8 kcal molÀ1), and is known to undergo a facile ring
cleavage reaction by the attack either at the C-2 car-
bonyl carbon or at the C-4 position.9
Results and Discussion
The hydroxyl group in 2 that was prepared by the lit-
erature method10 was oxidized to the corresponding
carboxylate with Jones’ reagent, and then treated with
K2CO3 in aqueous methanolic solution to give 4. Race-
mic 4 was converted into its methyl ester 5, and then
was subjected to enzymatic kinetic resolution using
a-CT to obtain (R)-4 and (S)-5. The optical purity of
(R)-4 determined as its methyl ester. Thus, (R)-4 was
converted into the corresponding methyl ester under the
acidic conditions and its optical purity was determined
by chiral HPLC to be 93.6% ee. The (S)-5 from the
kinetic resolution was treated again with a-CT to
improve the optical purity of 98.6% ee. The (S)-5 was
hydrolyzed to give (S)-4. Both enantiomers of 1 were
prepared by lactonization of the optically active 4 under
the modified Mitsunobu conditions (Scheme 1).11
Vederas et al.11 reported that the lactonization proceeds
without racemination under the reaction conditions.
Racemic 1 was prepared by lactonization of 4.
When a-CT was exposed to 50-fold Molar excess of
racemic 1, about 90% of the enzyme activity was lost
within several minutes, followed by slow regaining of
the enzymic activity, as shown by the upward swing of
Figure 1. Postulated reaction paths of (a) (R)-1 and (b) (S)-1 at the active site of a-CT.

D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
2555
The above kinetic and mass spectral data together with
well established catalytic mechanism1 for a-CT suggest
strongly that both enantiomers of 1 form acyl-enzyme
intermediates with concurrent cleavage of the lactone
ring by the Ser-195 hydroxyl upon exposing 1 to a-CT,
and the intermediates thus formed hydrolyze at a much
slower rate than that for the acylation reaction. Figure 9
shows computer generated models of a-CT that is acy-
lated at the Ser-195 hydroxyl by each enantiomer of 1.
Both deacylation rate constants are drastically atte-
nuated compared with those of normal substrates
(kd=ꢀ4000 minÀ1).20 The slow deacylation reaction
may be ascribed to the reduced basicity of the imidazole
moiety in His-57 as a result of hydrogen bonding inter-
actions in the acyl-intermediates: As can be seen in Fig-
ure 9a, in the acyl-intermediate formed with (R)-1 the
hydroxyl group generated in the ring cleavage reaction
is positioned within the distance of a hydrogen bonding
(3.09 A). On the other hand, when the Ser-195 hydroxyl
is acylated by (S)-1, the imidazole nitrogen may form a
hydrogen bond with the ester carbonyl oxygen; the two
atoms are separated by 2.99 A (Fig. 9b). Recently,
Mobashery and associates exploited such hydrogen
bonding to develop a novel type of a-CT inhibitor, 6.21
Their molecular modeling study suggested that the
hydrogen bond between the imidazole of His-57 and the
hydroxyl b to the ester carbonyl of the acyl-enzyme
intermediate may reduce the basicity of the imidazole of
His-57, leading to impair the catalytic machinery for the
deacylation reaction.21 The molecular modeling further
suggested that the hydrogen bonding is not sensitive to
the spatial orientation of the hydroxyl group.21 Our
kinetic results and the molecular models of the acyl-
intermediates are in accordance with the proposition
stipulated by Mobashery and associates.212
the a-CT catalyzed hydrolysis of the b-lactone of 1. It
appears that this may be the first example that a-CT
fails to exhibit stereospecificity in its enzymic reactions,
although there are precedences that some of the unna-
tural substrates and inhibitors for a-CT display reversed
stereospecificity, i.e., the d-stereospecificity.22 The ka
values obtained with the lactone substrates are drasti-
cally reduced, in spite of its inherent high chemical
reactivity owing to the high ring strain, compared with
that reported for structurally comparable acyclic ester
substrates such as N-acetyltryptophan ethyl ester
(kcat=1620min À1),23 suggesting that the tetrahedral
transition states generated in the acylation reaction of a-
CT with 1 may not be stabilized by the oxyanion hole,
and thus the relief of high ring strain energy associated
with the lactone ring cleavage may constitute the driving
force for the acylation reaction. The kobs for the hydro-
lysis of unsubstituted b-lactone in pH 7.5 buffer at room
temperature was reported to be 2.3Â10À3 minÀ1 24
.
This
One of the distinctive features of enzymic reactions is
the substrate specificity including specificity due to the
spatial configuration of substrate molecule. The S1 sub-
site hydrophobic pocket is situated in such a way that it
can preferentially accommodate l-amino acid residue
having a hydrophobic side chain, in forming a-CT sub-
strate complex thus to manifest substrate stereo-
chemistry in favor of the l-stereochemistry. Surprisingly,
however, no noticeable stereoselectivity was observed in
Figure 2. Spontaneous reactivation of a-CT inactivated by 1 at 25 ꢁC
in 0.05 M Tris buffer pH 7.5 containing 20% DMSO. The final con-
centrations of a-CT and 1 are 25 nM and 1.25 mM, respectively. The
enzymic activity was assayed using Suc-AAPF-pNA (125 mM) as the
substrate.
Scheme 1. Reagents, conditions, and (yields): (a) Jones’ reagent, 0 ꢁC
(73%); (b) K2CO3, MeOH/H2O (90%); (c) H+, MeOH (86%); (d) a-
CT, 0.05 M phosphate buffer (pH 7.5); (e) (i) a-CT, 0.05 M phosphate
buffer (pH 7.5); (ii) LiOH, MeOH/H2O (70%); (f) Ph3P, DMAD,
À78 ꢁC (75%).
Figure 3. Superimposed electrospray mass spectra of a-CT (m/z
25234 Da) and a-CT complexed with (R)-1 (m/z 25397 Da).

2556
D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
rate constant for the uncatalyzed b-lactone cleavage
compares favorably with the ka values obtained with 1
and tends to support the notion that the acylation reac-
tions of the Ser-195 hydroxyl by 1 are not enzymatically
driven.25 It may thus be concluded that the unique
chemical property of b-lactone in 1 is responsible, at
least in part, for the lack of stereospecificity shown in
the a-CT catalyzed hydrolysis of 1.
active site of a-CT to find that contrary to our expecta-
tion both are slowly hydrolyzed substrates of a-CT hav-
ing kcat values of 0.134Æ0.008 and 0.105Æ0.004 minÀ1
for (R)-1 and (S)-1, respectively. a-CT is virtually unable
to differentiate the enantiomers in the catalytic hydro-
lysis of the b-lactone ring in 1. It appears that the acy-
lation step in the hydrolysis of the b-lactone is not
enzymatically driven, and the relief of high ring strain
energy of the b-lactone would constitute the driving
force for the reaction. The deacylation step is atte-
nuated, which is possibly due to the reduced basicity of
His-57 imidazole ring as a consequence of the hydrogen
bond that would be formed between the imidazole ring
of His-57 and the hydroxyl group generated in the acy-
lation step. The imidazole moiety of His-57 functions as
a general base, activating the water molecule that
attacks the acyl-enzyme intermediate. The results of
present study may be of value in designing a novel type
of inhibitors for serine proteases. We are currently
exploiting the findings made in the present study to
develop a new design strategy that can be useful for
designing inhibitors effective against serine proteases.
It is interesting to note that the present observation that
both enantiomers of 1 are hydrolyzed by a-CT is in con-
trast to the recent finding of Lall et al.26 who reported
that (R)-N-benzyloxycarbonylamino-2-oxetanone inac-
tivates irreversibly the 3C cysteine protease of hepatitis
A virus (HAV). The thiol of Cys-172 in the cysteine
protease was shown to attack at the 4-position of the
b-lactone of the compound to form a thioether linkage
that resists hydrolysis.26 Although serine and cysteine
proteases are members of evolutionary distant proteases,
they share mechanistic similarities.27 Thus, both enzyme
families bear similar active site geometries and form cova-
lent acyl-enzyme intermediates.27 The reason why the two
mechanistically similar enzymes show such different reac-
tivities toward b-lactone is not apparent to us.
Experimental
General
Conclusion
We have synthesized both enantiomers of 3-benzyl-2-
oxetanone (1) and investigated their chemistry at the
Melting points were measured on a Thomas-Hoover
capillary melting point apparatus and are uncorrected.
1H and 13C NMR spectra were recorded on a Bruker
AM 300 (300 MHz) instrument using tetramethylsilane
as the internal standard. IR spectra were recorded on a
Bruker Equinox 55 FT-IR spectrophotometer. High
Figure 4. Active site protection tests for (a) (R)-1 ([a-CT]=25 nM,
[(R)-1]=500 mM, and [Succ-AAPF-pNA]=125, 180, and 250 mM), and
(b) (S)-1 ([a-CT]=25 nM, [(S)-1]=200 mM, and [Succ-AAPF-
pNA]=125, 180, and 250 mM).
Figure 5. Progress curves showing the a-CT catalyzed hydrolysis of
Succ-AAPF-pNA in the presence of varied concentrations of (a) (R)-1
and (b) (S)-1. Assay conditions are described in Experimental.

D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
2557
resolution mass spectra were obtained at Korea Basic
Science Center, Daejeon, Korea. ESI-MS spectrometric
analysis was performed using a triple quadrupole mass
spectrometer (Quattro, Micromass, Manchester, UK;
0.1% resolution, 0.01% accuracy) equipped with an ESI
source. Flash chromatography was performed on Silica
gel 60 (230–400 mesh) from Merck. Optical rotations
were measured on a Rudolph Research Autopol III
digital polarimeter. Elemental analyses were performed
at the Center for Biofunctional Molecules, Pohang
University of Science and Technology, Pohang, Korea.
All chemicals were obtained from Aldrich Chemical Co.
and solvents were purified before use.
containing K2CO3 (1.4 g) and several drops of water.
After stirring for 2 h at room temperature, the reaction
mixture was treated with several drops of acetic acid,
and concentrated in vacuo. The residue was purified by
flash column chromatography (silica gel, EtOAc:hex-
ane=1:3) to give 4 (90%): mp 58–59 ꢁC; lit29 mp 63–
65ꢁC: IR (KBr) 3300, 1710 cmÀ1; H NMR (300 MHz,
1
CDCl3) d 7.16–7.31 (m, 5H), 4.8 (br, 2H), 3.68–3.80(br,
2H), 3.00–3.07 (m, 1H), 2.80–2.92 (m, 2H).
2-Benzyl-3-hydroxypropanoic acid methyl ester (5). To
an ice-chilled stirred solution of 4 (1.3 g, 7 mmol) in
MeOH (20mL) was added acetyl chloride (2 mL). The
mixture was stirred for 24 h at room temperature, and
then evaporated under reduced pressure. The residue
was dissolved in EtOAc (30mL) and aqueous NaHCO 3
solution added to neutralize the mixture. The organic
layer was washed with copious water, and brine, dried
over MgSO4, and evaporated to give an oily residue,
which was purified by column chromatography (silica
gel, EtOAc:Hexane=1:5) to yield the product 530 as an
3-Acetoxy-2-benzylpropanoic acid (3). To an ice-chilled
solution of 210 (3.1 g, 15 mmol) in 10mL of acetone was
added slowly the Jones’ reagent until brownish color of
the solution remains over 20min, then 2-propanol was
added until the solution became clear. The precipitate
that separated was filtered using a Celite pad and the
filtrate was evaporated under reduced pressure. The
residue was diluted with 5 mL of 1 N HCl then extracted
with ethyl acetate (30mL Â3). The extract was dried
over anhydrous MgSO4 and evaporated under reduced
pressure to give a yellowish syrup. The residue was pur-
ified by flash column chromatography (silica gel,
EtOAc:hexane=3:1) to give 328 as an oil (73%): IR (KBr)
1741, 1712 cmÀ1; 1H NMR (300MHz, CDCl3) d 7.16–7.32
(m, 5H), 4.10–4.23 (m, 2H), 2.96–3.09 (m, 2H), 2.81–2.94
(m, 1H), 1.98 (s, 3H); MS (CI) m/z 223 (M+1)+.
oil (86%): IR (KBr) 3300, 1735 cmÀ1
;
1H NMR
(300 MHz, CDCl3) d 7.19–7.35 (m, 5H), 3.65–3.80(m,
2H), 3.70(s, 3H), 3.03 (dd, J=16.9, 9.2 Hz, 1H), 2.88
(m, 2H), 2.18 (t, J=7.7 Hz, 1H); MS (EI) m/z 195
(M+1)+.
(R)-2-Benzyl-3-hydroxypropanoic acid ((R)-4). Racemic
5 (3.48 g, 18 mmol) was suspended in 50 mL of 0.01 M
phosphate buffer solution. To the mixture was added
a-CT (300 mg), and stirred slowly. The pH of the reac-
tion mixture was maintained at pH 7.5 by addition of
sodium hydroxide solution (0.2 N) using a pH-stat
2-Benzyl-3-hydroxypropanoic acid (4). Compound 3
(1.8 g, 8 mmol) was dissolved in methanol (10mL)
Figure 7. Dixon plots for determinations of KS values in the acylation
reactions of a-CT with (a) (R)-1 and (b) (S)-1. Assay conditions are
found in Experimental.
Figure 6. Kitz–Wilson plots for determinations of KM and ka values
for (a) (R)-1 and (b) (S)-1.

2558
D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
reduced pressure to give optically pure (S)-530 whose
optical purity was determined by HPLC using a chiral
column as described above for (R)-5 to obtain the % ee
value of 98.6. To the solution of (S)-5 (0.9 g, 4.9 mmol)
in tetrahydrofuran (5 mL) and MeOH (1.5 mL) at 0 ꢁC,
was added dropwise aqueous solution of LiOH (0.13 g,
5.4 mmol). The resulting solution was stirred at room
temperature for 12 h. The solution was acidified to pH 2
with 1 N HCl, extracted with ether (3Â20mL), dried
over MgSO4, and evaporated under reduced pressure.
The residue was recrystallized to give (S)-429 (0.6 g,
70%): mp 68.5 ꢁC; [a]D=À22.3ꢁ (c=1.5, MeOH).
(S)-3-Benzyl-2-oxetanone ((S)-1). To a solution of tri-
phenylphosphine (2.65 g, 10mmol) in THF (30mL) was
added dropwise dimethyl azodicarboxylate (3.7 mL of
40% solution in toluene, 10 mmol) at À78 ꢁC and stirred
for 20min. ( S)-2-Benzyl-3-hydroxypropionic acid (1.3 g,
7.21 mmol) in THF (15 mL) was added. The resulting
mixture was stirred at À78 ꢁC for 1 h, and then allowed
to warm to room temperature under stirring for 4 h. The
mixture was evaporated under reduced pressure to give
a reddish oil which was purified by column chromato-
graphy (hexane:ethyl acetate=5:1) to yield (S)-1 as an
oil (880mg, 75%): [ a]D +95.0ꢁ (c=0.88, CHCl3); IR
(KBr) 1821 cmÀ1 1H NMR (300 MHz, CDCl3) 7.18–
;
Figure 8. The determinations of kd values for (a) (R)-1 and (b) (S)-1
by the proflavin displacement assay method.
7.36 (m, 5H), 4.32 (m, 1H), 3.98–4.08 (m, 2H), 3.08–3.21
(m, 1H); 13C NMR (75 MHz, CDCl3) 171.5, 137.2, 129.3,
129.0, 127.6, 64.8, 53.4, 34.1; HRMS-FAB m/z: (M+1)+
calcd for C10H10O2, 163.0759, found, 163.0761.
potentiometer. When 35 mL of the alkaline solution was
consumed (18 h), the reaction mixture was acidified with
1 N HCl solution and saturated with sodium chloride
then extracted with ethyl acetate. The extract was dried
over anhydrous MgSO4, filtered, and evaporated under
reduced pressure. The residue was purified by column
chromatography (silica gel, EtOAc:hexane=1:4 to 1:1)
to give (R)-429 and crude (S)-5. The spectral data of
these compounds were identical with those of respective
racemic compounds, i.e., 4 and 5. (R)-4: [a]D=+23.5ꢁ
(c=1.00, MeOH). In order to determine the optical
purity of (R)-4 thus obtained, the product was con-
verted into the corresponding methyl ester under acidic
conditions (MeOH+SOCl2), and the product was ana-
lyzed by HPLC using a chiral column (Chiralcel OD,
Diacel Chemical Ind.) to find that it has the optical
purity of 93.6% ee. The HPLC was performed under
the following conditions: eluent, n-hexane (98)/2-propa-
nol (2); flow rate, 1 mL minÀ1; detection at 217 nm.
(R)-3-Benzyl-2-oxetanone ((R)-1). was prepared in a
similar fashion starting with (R)-2-benzyl-3-hydroxy-
propionic acid and obtained in an overall yield of 78%:
[a]D À99.0ꢁ (c=0.58, CHCl3); HRMS-FAB m/z:
(M+1)+ calcd for C10H10O2, 163.0759, found, 163.0754.
Kinetics
General. a-CT and Suc-Ala-Ala-Pro-Phe-pNA were
purchased from Sigma Chemical Co. Tris buffer
(0.05 M, pH 7.5) containing 0.05 M CaCl2 was used for
the enzyme assay. All solutions were prepared by dis-
solving in doubly distilled and deionized water. Stock
assay solutions were filtered through a membrane filter
before use. The enzyme activity was measured by mon-
itoring the change in absorbance at 410nm caused by
p-nitroanilide that was generated by the enzymic
hydrolysis of Suc-Ala-Ala-Pro-Phe-pNA (substrate).
The stock solutions of the substrate (5 mM) and 1 were
prepared in DMSO solution.
(S)-2-Benzyl-3-hydroxypropanoic acid ((S)-4). The crude
(S)-5 from the enzymic resolution of 5 was treated again
with a-CT until 15 mL of 0.2 N NaOH was consumed.
The solution was extracted with ether several times. The
combined organic layer was washed with 5% NaHCO3,
brine, dried over MgSO4, and then evaporated under
Competitive substrate assay for the determination of KM
and ka values. a-CT (15 nM) was added to a 20%
DMSO solution of the Tris buffer (pH 7.5) containing
Table 1. Kinetic parametersa
Compound
ka (minÀ1
)
KM (mM)
KS (mM)
ka/KM (MÀ1 minÀ1
)
kd (minÀ1
)
kcat (minÀ1
)
(R)-1
(S)-1
1.15Æ0.05
1.64Æ0.03
167Æ32
196Æ11
83Æ6.6
.152
68900
41,600
Æ0.005
0.134Æ0.008
0.105Æ0.004
39.4Æ4.1
0.112Æ0.001
aData represent the mean of three experiments.

D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
2559
the substrate (400 mM) and 1 of specified concentrations
(112 to 672 mM). The change in absorbance at 410nm
was recorded on a UV spectrophotometer (HP 8453) for
150–250 s and plotted against time to obtain progress
curves (Fig. 5). The pseudo-first order rate constants
(kobs) were calculated by fitting the progress curves to
equation 1. Values of KM, and ka were obtained from
the plots of 1/kobs versus 1/[I]0 (Fig. 6) according to eq.
2:
concentrations of proflavin, a-CT, and 1 were 9, 90, and
90 mM, respectively). The change in absorbance at
466 nm was followed and plotted against time to have
progress curve (Fig. 8). Replot of the increasing absor-
bance versus time in a semilogarithmic manner gave a
straight line the slope of which corresponds kd.
Determination of KS values
KS values were estimated from the semireciprocal plot
of the initial velocity versus the concentration of 1 (Fig.
7) according to the method of Dixon. Two concentra-
tions of the substrate were used (100 and 200 mM).
Typically, the enzyme stock solution was added to var-
ious concentrations of 1 (0to 200 mM) in 20% DMSO
solution of the Tris buffer (pH 7.5) (1 mL cuvette), and
the initial rates were measured immediately using a
microcomputer-interfaced UV spectrometer.
À
Á
A410 nm ¼ vS:t þ ðv0 À vSÞÁ 1 À eÀkobsÁt =kobs þ c
ð1Þ
ð2Þ
1=kobs ¼ 1=ka þ ½KM Áð1 þ ½Sꢀ0=KMÞꢀ=ðka Á½Iꢀ0Þ
Determination of kd values
Active site protection test
To a mixture of proflavin solution in the pH 7.5 Tris
buffer (100 mL, 100 mM) and a-CT stock solution in the
same Tris buffer (1.0mL, 100 mM) was added a solution
To premixed solutions of 1 (500 and 200 mM for (R)-1
and (S)-1, respectively) and varying concentrations of
Suc-AAPF-pNA (125, 180, and 250 mM) in 20% DMSO
solution of the Tris buffer (pH 7.5) was added a-CT
(25 nM), and the remaining enzymic activity was mon-
ꢁ
of 1 in DMSO (10 mL, 10mM) at 25 C (the final
ꢁ
itored at 410nm at 25 C, and the activity remaining
was plotted against time (Fig. 4). The rate of inactiva-
tion was decreased as the concentration of the substrate
increased.
Electrospray ionization mass spectrometry
Electrospray ionization mass spectra were recorded on a
VG Quattro quadrupole mass spectrometer equipped
with an electrospray interface. a-CT (80 mM) was incu-
bated in the Tris buffer (pH 7.5) with a 50-fold excess of
1 for 10min, and then diluted with a 50% aqueous ace-
tonitrile solution containing 0.5% formic acid to have
40pmol L À1 of a-CT substrate covalent complex (an
intermediate in the a-CT catalyzed hydrolysis of 1). A
sample solution (10 mL) thus obtained was introduced
into the electrospray source using a loop injector. Mass
spectra were typically obtained from 20scans over the
range of 900 to 1600 Mr. Mass spectra were acquired
with a cone voltage of 31 V and a source temperature of
75 ꢁC. The instrument was calibrated with horse heart
myoglobin. Electrospray ionization mass spectra of
both a-CT and 1 complexes in Figure 3 were calibrated
with a-CT.
Molecular modeling
The initial geometries of the complexes were generated
by superimposing hydrocinnamaldehyde on the X-ray
crystal structure of l-[1-acetamido-2-(p-chlorophenyl)
ethyl]boronic acid31 bound to a-CT and a hydroxy-
methyl group was added to the hydrocinnamaldehyde at
the 2-position. The hydroxyl group of Ser-195 and the
carbonyl carbon of the hydrocinnamaldehyde were
covalently bound to form an ester bond and then the
l-[1-acetamido-2-(p-chlorophenyl)ethyl]boronic
was removed from the complexes.
acid
Figure 9. Computer generated models of the active site of a-CT whose
Ser-195 hydroxyl is acylated by (a) (R)-1 and (b) (S)-1.

2560
D. H. Kim et al. / Bioorg. Med. Chem. 10 (2002) 2553–2560
The simulations were performed with the Discover pro-
gram, version 2.9.0(MSI, San Diego, CA, USA) on a
Silicon Graphics Indigo 2 computer, using the CVFF
force field. A dielectric constant of 1 was used in all
calculations. The energy of the system was minimized
with respect to all 3N Cartesian coordinates until the
maximum derivative of 0.1 kcal molÀ1 AÀ1 was reached.
The resulting structure was used as the starting point for
the molecular dynamics calculations. The complex was
subjected to a simulated annealing in which the complex
was heated from 0to 300K in 10ps, equilibrated for 10
ps, and then cooled to 0K in 10 ps. The selected struc-
tures were minimized using steepest descents until the
active site (possibly His-57) to form an a,b-unsaturated acid
with regeneration of the catalytic activity of the a-CT. How-
ever, this possibility is thought to be unlikely, considering the
precedents reported in the literature: Li, M.; Luo, W.; White,
E. H. Arch. Biochem. Biophys. 1995, 320, 135. (b) Kim, D. H.;
Li, Z.-H. Bioorg. Med. Lett. 1994, 4, 2297. (c) Kim, Y. J.; Li,
Z.-H.; Kim, D. H.; Hahn, J. H. Bioorg. Med. Chem. Lett.
1996, 13, 1449.
14. KM denotes the binding constant for the overall acylation
process defined by k+1/(kÀ1+ka).
15. Kitz, R.; Wilson, I. B. J. Biol. Chem. 1962, 237, 3245.
16. DelMar, E. G.; Largman, C.; Brodrick, J. W.; Giokas,
M. C. Anal. Biochem. 1979, 99, 316.
17. Dixon, M. Biochem, J. 1953, 55, 170.
18. Bernhard, S. A.; Lee, B. F.; Tashjian, Z. H. J. Mol. Biol.
1966, 18, 405.
maximum derivative was less than 5.0kcal mol À1 AÀ1
,
and then using a conjugate gradient, until the maximum
19. The a-CT catalyzed b-lactone ring cleavage reaction may
be represented by the Scheme 2, in which KM represents the
Michaelis–Menten constant (k+1/(kÀ1+ka), Ks the dissocia-
derivative was less than 0.1 kcal molÀ1 AÀ1
.
.
tion constant of the E S complex (k+1/kÀ1), ka the rate con-
Acknowledgements
stant for acylation reaction, and kd the rate constant for
deacylation reaction. The kcat denotes the second order rate
constant for the overall reaction.
The authors express their thanks to the Ministry of
Education and Human Resources for the BK21 fellow-
ship, Ministry of Science and Technology, and Korea
Science and Engineering Foundation for financial sup-
port of this work.
Scheme 2.
References and Notes
20. Fersht, A. Enzyme Structure and Mechanism, 2nd ed.; W.
H. Freeman and Co.: New York, 1985; p 201.
1. Wharton, C. W. In Comprehensive Biological Catalysis;
Sinnott, M. Ed.; Academic Press: New York, 1998; Vol. 1,
Chapter 9.
2. Bernstein, P. R.; Edwards, P. E.; Williams, J. C. Prog. Med.
Chem. 1994, 31, 59.
3. Ripka, W. C.; Vlasuk, G. R. Ann. Rep. Med. Chem. 1997,
32, 71.
4. Taylor, J. C.; Mittman, C., Eds; Pulmonary Emphysema
and Proteolysis: 1986; Academic Press: New York, 1987; p 1.
5. Warshel, A.; Naray-Szabo, G.; Sussman, F.; Hwang, J.-K.
Biochemistry 1989, 28, 3629.
6. For example: Silverman, R. B. The Organic Chemistry of
Drug Design and Drug Action, Academic Press: New York,
1992; p 172.
7. (a) Kranz, A.; Spencer, R. W.; Tam, T. F.; Thomas, E.;
Copp, L. J. J. Med. Chem. 1987, 30, 591. (b) Reed, P. E.;
Katzenellenbogen, J. A. J. Biol. Chem. 1991, 266, 13.
8. Kim, D. H.; Ryoo, J. J. Bioorg. Med. Chem. Lett. 1995, 5,
1287.
21. Li, Z.-H.; Bulychev, A.; Kotra, L. P.; Massova, I.;
Mobashery, S. J. Am. Chem. Soc. 1998, 120, 13003.
22. (a) Hein, G.; Niemann, C. J. Am. Chem. Soc. 1962, 84,
4495. (b) Lanonde, J. J.; Bergbreiter, D. E.; Wong, C.-H. J.
Org. Chem. 1988, 53, 2323. (c) Baek, D.-J.; Reed, P. E.;
Daniels, S. B.; Katzenellenbogen, J. A. Biochemistry 1990, 29,
4305.
23. Zerner, B.; Bond, R. P. M.; Bender, M. L. J. Am. Chem.
Soc. 1964, 86, 3674.
24. Blackburn, G. M.; Dodds, H. L. H. J. C. S. Perkin 2 1974,
377.
25. The b-lactone ring cleavage reaction that takes place at
the active site of a-CT is faster by three orders of magnitude
than that occurring in aqueous medium, but this may not be
unreasonable considering that reactions at the enzyme active
site might be treated as an intramolecular in nature (Fersht, A.
Enzyme Structure and Mechanism, 2nd ed.; W. H. Freeman
and Co.: New York, 1985; p 56).
26. Lall, M. S.; Karvellas, C.; Vederas, J. C. Org. Lett. 1999,
1, 803.
27. Otto, H.-H.; Schirmeister, T. Chem. Rev. 1997, 97, 133.
28. Fincham, C. I.; Higginbottom, M.; Hill, D. R.; Horwell,
D. C.; O’Toole, J. C.; Ratcliffe, G. S.; Rees, D. C.; Roberts, E.
J. Med. Chem. 1992, 35, 1472.
29. Monteil, T.; Danvoy, D.; Plaquevent, J.-C.; Duhamel, L.;
Duhamel, P.; Gros, C.; Schwartz, J.-C.; Lecomte, J.-M. Synth.
Commun. 2001, 31, 211.
30. Peet, N. P.; Lentz, N. L.; Dudley, M. W.; Ogden, A. M. L.;
McCarty, D. R.; Racke, M. M. J. Med. Chem. 1993, 36, 4015.
31. Stoll, V. S.; Eger, B. T.; Hynes, R. C.; Martichonok, V.;
Jones, J. B.; Pai, E. F. Biochemistry 1998, 37, 451.
9. Yang, H. W.; Romo, D. Tetrahedron 1999, 55, 6403.
10. Lee, J.; Lee, J.; Kim, J.; Kim, S. Y.; Chun, M. W.; Cho,
H.; Hwang, S. W.; Oh, U.; Park, Y. H.; Marquez, V. C.;
Beheshti, M.; Szabo, T.; Blumberg, P. M. Bioorg. Med. Chem.
2001, 9, 19.
11. Arnold, L.; Kalantar, T. H.; Vederas, J. C. J. Am. Chem.
Soc. 1985, 107, 7105.
12. It is highly unlikely that the intermediates detected by the
ESI-MS would be Michaelis–Menten complexes considering
the high KS values estimated for the complexes (Table 1).
13. (a) One of the referees has called our attention to the
possibility that the product generated by the C4 attack may
undergo a retro-aldol type reaction catalyzed by a base at the