Antibody-Catalyzed Benzoin Oxidation as a Mechanistic Probe for Nucleophilic Catalysis by an Active Site Lysine

Article abstract of DOI:10.1002/chem.200305034

Aldolase antibody 24H6, which was obtained by reactive immunization against a 1,3-diketone hapten, is shown to catalyze additional reactions, including H/D exchange and oxidation reactions. Comparison of the H/D exchange reaction at the α-position of a wide range of aldehydes and ketones by 24H6 and by other aldolase antibodies, such as 38C2, pointed at the significantly larger size of the 24H6 active site. This property allowed for the catalysis of the oxidation of substituted benzoins to benzils by potassium ferricyanide. This reaction was used as a mechanistic probe to learn about the initial steps of the 24H6-catalyzed aldol condensation reaction. The Hammett correlation (ρ=4.7) of log(k_cat) versus the substituent constant, σ, revealed that the reaction involves rapid formation of a Schiff base intermediate from the ketone and an active site lysine residue. The rate-limiting step in this oxidation reaction is the conversion of the Schiff base to an enamine intermediate. In addition, linear correlation (ρ= 3.13) was found between log(K_M) and σ, indicating that electronic rather than steric factors are dominant in the antibody-substrate binding phenomenon and confirming that the reversible formation of a Schiff base intermediate comprises part of the substrate-binding mechanism.

Full text of DOI:10.1002/chem.200305034

FULL PAPER
Antibody-Catalyzed Benzoin Oxidation as a Mechanistic Probe for
Nucleophilic Catalysis by an Active Site Lysine
[
a]
[a]
[a]
[a, b]
Genia Sklute, Rachel Oizerowich, Hagit Shulman,* and Ehud Keinan*
Abstract: Aldolase antibody 24H6,
which was obtained by reactive immu-
nization against a 1,3-diketone hapten,
is shown to catalyze additional reac-
tions, including H/D exchange and oxi-
dation reactions.Comparison of the
H/D exchange reaction at the a-posi-
tion of a wide range of aldehydes and
ketones by 24H6 and by other aldolase
antibodies, such as 38C2, pointed at the
significantly larger size of the 24H6
active site.This property allowed for
the catalysis of the oxidation of substi-
tuted benzoins to benzils by potassium
ferricyanide.This reaction was used as
a mechanistic probe to learn about the
initial steps of the 24H6-catalyzed aldol
condensation reaction.The Hammett
idue.The rate-limiting step in this oxi-
dation reaction is the conversion of the
Schiff base to an enamine intermediate.
In addition, linear correlation (1=
correlation (1=4.7) of log(k ) versus
3.13) was found between log(K ) and
cat
M
the substituent constant, s, revealed
that the reaction involves rapid forma-
tion of a Schiff base intermediate from
the ketone and an active site lysine res-
s, indicating that electronic rather than
steric factors are dominant in the anti-
body±substrate binding phenomenon
and confirming that the reversible for-
mation of a Schiff base intermediate
comprises part of the substrate-binding
mechanism.
Keywords: aldol reaction ¥ catalytic
antibodies ¥ reaction mechanisms ¥
Schiff bases
Introduction
verting the carbonyl donor (ketone or aldehyde) into a pro-
tonated Schiff base (I; Scheme 1), which can then tautomer-
ize to an enamine intermediate (II).The latter electron-rich
An active site nucleophilic lysine residue with a highly per-
turbed pK is an essential element of the catalytic machinery
a
[
1]
available to both the natural, type I aldolase enzymes and
the aldolase antibodies that were elicited by reactive immu-
nization with 1,3-diketones. These chemically programmed
antibodies, which include 33F12, 38C2,
4G3, were shown to catalyze aldol reactions with broad
[2]
[2,3]
[4]
24H6, 93F3, and
[5]
8
substrate scope, remarkable enantioselectivity, and high re-
action rates.The crystal structure of 33F12 indeed revealed
[6]
a lysine residue that is buried in a hydrophobic pocket.
This lysine residue can catalyze the aldol reaction by con-
[
a] G.Sklute, R.Oizerowich, Dr.H.Shulman, Prof.E.Keinan
Department of Chemistry and
Scheme 1.Pathway for the antibody-catalyzed aldol reaction.B =base.
Institute of Catalysis Science and Technology
Technion–Israel Institute of Technology
Technion City
species may participate in variety of carbonyl transforma-
tions, such as aldol and retroaldol reactions, alkylation, deut-
eration, decarboxylation, and so forth.The aldol reaction,
for example, involves nucleophilic addition of II to a car-
bonyl acceptor to produce a new Schiff base intermediate
Haifa 32000 (Israel)
[
b] Prof.E.Keinan
Department of Molecular Biology and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
(
III) and, following hydrolysis, an aldol product (IV).
1
0550 North Torrey Pines Road
Beyond the synthetic significance of catalyzing asymmet-
La Jolla, California 92037 (USA)
+
ric carbonyl transformations by new biocatalysts, it is impor-
tant to understand the details of their catalytic machinery,
[
] Incumbent of the Benno Gitter & Ilana Ben-Ami chair of Biotech-
nology, Technion.
Chem. Eur. J. 2004, 10, 2159 ± 2165
DOI: 10.1002/chem.200305034
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
2159

FULL PAPER
particularly in comparison with the naturally occurring
counterparts.Mechanistic information and identification of
enzyme±substrate intermediates could be achieved by using
various reactions as mechanistic probes.For example, linear
[7]
free-energy relationship studies, using substituted aromatic
substrates, could teach us about charge development in the
transition state and point at rate-limiting steps.We have re-
cently carried out such Hammett correlation studies of the
[4,8]
38C2- and 24H6-catalyzed aldol and retroaldol reactions,
taking advantage of the broad substrate scope of these bio-
catalysts.We found that although the two antibodies exhibit
broad substrate specificities, they utilize slightly different
mechanisms.While antibody 38C2 adopts a mechanism that
is reminiscent of an acid-catalyzed aldol reaction, antibody
2
4H6 uses a mechanism that is similar to the base-catalyzed
reaction.
In those studies,
inhibition by excess acceptor, we searched for another reac-
tion that shares the first mechanistic steps with the aldol re-
action, but that continues irreversibly to products.
[4,8]
we used acetone with a series of sub-
stituted cinnamaldehydes to investigate the substituent
effect on the acceptor partner of the aldol/retroaldol reac-
tion.To study the aldol reaction, we employed acetone in
large excess, rendering the kinetics pseudo-first order.The
use of one of the reactants in large excess was essential not
only for the simplification of the kinetics, but, more impor-
tantly, for pushing the highly reversible aldol reaction to-
wards the products.A complementary study with a series of
substituted donor substrates is highly desirable.Unfortu-
nately, the reversibility of the aldol/retroaldol reactions
would require the employment of the acceptor aldehyde in
large excess, which would result in total inhibition of the
catalyst by quantitatively converting its active-site lysine res-
idue into a Schiff base.In addition, the aldolase antibodies,
The notion of using another reaction as a mechanistic
probe for the first steps of the biocatalyzed aldol reaction
[9]
has been demonstrated by Christen, who investigated the
rabbit muscle aldolase-catalyzed oxidation of a-hydroxycar-
bonyls to a-dicarbonyls.The aldolase-catalyzed oxidation of
various substrates, including dihydroxyacetone phosphate,
fructose-1-phosphate, and fructose-1,6-diphosphate, with tet-
ranitromethane as well as with other oxidants testified to
the formation of Schiff base and enamine intermediates.
These oxidation reactions were also used more generally as
mechanistic probes to identify other oxidizable substra-
te±enzyme intermediates with several other enzymes, includ-
ing type I and type II aldolases, aspartate aminotransferase,
pyruvate decarboxylase, and 6-phosphogluconate dehydro-
[10]
38C2, 33F12, 93F3, and 84G3 elicited against small haptens,
genase. In all cases, the enzyme-catalyzed oxidations were
found to be highly specific with respect to the substrate, but
rather nonspecific with respect to the oxidant.
such as 1, although quite indiscriminate with respect to the
acceptor molecules, do not accept large donors, such as sub-
stituted acetophenones, which are required for Hammett
studies.
We anticipated that the donor size problem could be
solved with antibody 24H6, which was induced by the large
dicarbonyl haptens, 2a and 2b.To address the problem of
Here we focus on our recently reported aldolase antibody
24H6, showing that it indeed accepts large substrate mole-
cules and catalyzes oxidation reactions that are not cata-
lyzed by other aldolase antibodies.More importantly, we
report on a Hammett correlation study that sheds light on
the initial steps of the catalytic aldol mechanism with this
antibody.The specific reaction used as the mechanistic
probe was the oxidation of a-ketoalcohols to 1,2-diketones
by potassium ferricyanide.
Abstract in Hebrew:
Results and Discussion
Antibody 24H6, which was raised against a mixture of 2a
and 2b, catalyzes aldol and retroaldol reactions with a broad
variety of substrates.In order to learn more about the sub-
strate preferences of this antibody, we studied the relative
rates of the H/D exchange reaction at the a-position of a
wide range of aldehydes and ketones (Figure 1).As seen in
the figure, aldehydes are very reactive substrates with this
antibody, and cyclic ketones are generally better substrates
than the acyclic ones.This tendency and the general prefer-
ence of 24H6 for aromatic substrates in the aldol/retroaldol
[4]
reactions agree with a highly hydrophobic active site that
is lined up with many aromatic residues.
2160
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2159 ± 2165

Antibody-Catalyzed Benzoin Oxidation
2159 ± 2165
Scheme 2.Cross aldol reaction between hexanal and p-nitrocinnamalde-
hyde.
À1
with kcat =0.002 min and K =40mm.By contrast, 38C2 is
M
more limited with respect to the donor size.For example,
3
8C2 can use propanal and butanal as donors, but not hepta-
[11]
nal and higher aldehydes.
It appears that despite the
broad substrate range, which characterizes antibodies pro-
duced through reactive immunization, the general shape and
size of the hapten and the positioning of the 1,3-diketone
functionality in it play significant roles in determining the
dimensions and topography of the resultant antibody active
site.Although both 38C2 and 24H6 were elicited against di-
ketone haptens, they show different reactivity and substrate
range.The significantly larger size of the 24H6 active site
suggests that this antibody could catalyze the above-men-
tioned oxidation of substituted benzoins to benzils.
Figure 1.Ketone and aldehyde substrates used for 24H6-catalyzed deute-
rium exchange reactions at the a-carbon.The numbers represent relative
rates.A) Acyclic ketones and aldehydes.B) Cyclic ketones.
Use of the deuteration reaction to compare the substrate
[11]
[12]
profile of 24H6 (Figure 1) with that of 38C2 (Figure 2)
For the Hammett correlation studies we synthesized a
series of substituted benzoin substrates and benzil products
as described in Scheme 3.The substituted desoxybenzoins,
indicates that 38C2 is more selective for small substrates,
Scheme 3.Syntheses of substrates and products.
Figure 2.Ketone and aldehyde substrates used for 38C2-catalyzed deute-
3
a and 3c were prepared from the appropriate benzalde-
rium exchange reactions at the a-carbon.The numbers, taken from refer-
À1
hydes by reaction with phenyldiazomethane (formed in situ
from the sodium salt of phenyltosylhydrazone in THF ).
Compound 3d was prepared from phenylacetic acid and 4-
anisaldehyde. Hydroxylation of 3d with N-phenylsulfon-
yloxaziridine (Davis reagent)
The substituted benzils 5a and 5c were obtained by direct
oxidation with SeO2. 4-Methoxybenzil (5d) was obtained
by oxidation of 3d with copper sulfate in refluxing pyri-
dine.
ence [12], are rate constants [min ].When two numbers are given for
[13]
one substrate, they refer to the a- and a’-positions.A) Acyclic ketones
and aldehydes.B) Cyclic ketones.
[14]
[15]
particularly for monocyclic ketones.For example, with 38C2
decalone is 3000 times less reactive than cyclohexanone.By
contrast, with 24H6 decalone is only five times less reactive
than cyclohexanone, indicating that this antibody is more in-
discriminate towards large substrates than 38C2.
afforded 4a, 4c, and 4d.
[16]
[17]
Benzoin (4b) and benzil (5b) were commercially
In accordance with these results, which point at a larger
active site in 24H6, we also found that 24H6 can catalyze
aldol condensation reactions with relatively large aliphatic
aldehydes as donors.For example, butanal, valeraldehyde,
hexanal, and heptanal were found to be good aldol donors,
as exemplified by the cross aldol reaction between hexanal
and p-nitrocinnamaldehyde (Scheme 2), which proceeded
available.
Antibody 24H6 was found to catalyze the oxidation of all
substituted benzoins (4a±d) to the corresponding benzil de-
rivatives (5a±d) in the presence of K [Fe(CN) ].Catalysis
followed Michaelis±Menten saturation kinetics.Antibody
38C2 failed to catalyze this reaction, thus serving for control
experiments that confirmed specific catalysis at the binding
3
6
Chem. Eur. J. 2004, 10, 2159 ± 2165
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
2161

E.Keinan et al.
FULL PAPER
site of 24H6.Apparently, such substituted aromatic ketone
substrates are too large to fit within the active site of 38C2.
Another control experiment that supported catalysis by the
active-site lysine residue was carried out with 24H6 in the
presence of 10% acetone as a co-solvent instead of DMSO.
Expectedly, since acetone can form a Schiff base with the
antibody lysine residue, the reaction was inhibited by ap-
proximately 30% under these conditions.The antibody-cata-
lyzed oxidation reactions were carried out at 258C with a
constant concentration of 24H6 (10.7mm) and K [Fe(CN) ]
3
6
(
2mm) and variable concentrations of substrate in phosphate
buffered saline (PBS, 50mm phosphate, 100mm NaCl,
pH 9.0) containing 10% DMSO. The progress of the reac-
tion was monitored by HPLC equipped with a UV detector.
The kinetic parameters, kcat and KM (Table 1), were extract-
Table 1.Kinetic parameters of the antibody 24H6-catalyzed oxidation re-
actions with substrates 4a±d.The s values were taken from refer-
Figure 3.Hammet correlation for the 24H6-catalyzed oxidation reaction.
ence [14].
À1
R
s
k
cat [min
]
K
M
[mm]
kcat/K
M
À2
À3
À4
À4
À3
À4
À4
À4
4
4
4
4
a
b
c
Cl
H
Me
OMe
0.24
0.00
9.10î10
1.96î10
4.45î10
1.20î10
50.05
7.98
3.76
0.52
1.82î10
2.50î10
1.18î10
2.30î10
À0.16
d
À0.28
ed from a Lineweaver±Burk analysis of the kinetic data.
Originally, we attempted to increase the range of substrates
for the Hammett correlation study and include three more
substrates with higher and lower s values.To that end we
synthesized 4e (R=NMe ), 4 f (R=CF ), and 4 g (R=NO )
2
3
2
using the above-described procedures.However, the 24H6-
catalyzed oxidation reaction with 4e produced the corre-
sponding benzil along with other products.Substrates 4 f
and 4g, which bear electron-withdrawing substituents, were
found to be too unstable under the antibody-catalyzed reac-
tion conditions.Fortunately, the quality of the correlation
line and the relatively steep slope made a four-point line
sufficient for this study.
Plotting the values of log(k ) against the values of the
cat
[18]
Hammett substituent constant, s (Figure 3),
afforded a
linear correlation with a positive slope (1=4.7). This Ham-
mett correlation coefficient suggests that a positive charge is
diminished in the transition state of the rate-limiting step.A
similarly large coefficient (1=5.3), which has been reported
for the alkaline hydrolysis of 2-methoxy tropones, was also
interpreted in terms of formation of a significant negative
charge in the transition state.
A plausible mechanistic pathway of the 24H6-catalyzed
oxidation reaction (Scheme 4) starts with a nucleophilic
Scheme 4.Mechanistic pathway of the antibody-catalyzed oxidation reac-
tion.
This mechanistic scheme suggests that the first two steps,
formation of intermediates I and II, involve accumulation of
a negative charge in the benzylic position in the transition
state, which would be reflected by a positive 1 value.Both
oxidation steps involve partial formation of a positive
charge in the benzylic position, which would be reflected by
a negative 1 value.Thus, based on the large, positive value
of 1 in our case, we conclude that both oxidation steps
which form intermediates V and VI are fast.Thus, the rate-
limiting step would be either formation of I or II.Mechanis-
tic studies of Schiff base formation suggest that the rate-de-
[19]
attack of a low-pK amine residue on the substrate carbonyl
a
to form a Schiff base (or a protonated Schiff base) inter-
mediate (I).Tautomerization of the latter to an enamine in-
termediate (II) occurs by means of a base-mediated depro-
tonation at the a-carbon atom.The highly electron-rich
double bond in II is prone to oxidation, which probably
occurs through two consecutive one-electron oxidation steps
[20]
termining step depends on the pH. In acidic solution the
reversible attack of free amine on the carbonyl carbon to
form a tetrahedral intermediate is rate-limiting; this is char-
III
by Fe to produce intermediates V and VI.Finally, hydroly-
sis of VI would release the product, 5.
2162
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2159 ± 2165

Antibody-Catalyzed Benzoin Oxidation
2159 ± 2165
acterized by Hammett constants ranging between 1.5 and
[
20]
2.
Conversely, in basic solution, the attack of free amine is
fast, while the loss of water (or hydroxide ion) becomes the
slow step.The much larger value of 4 7. found in our case
suggests that formation of I is not rate-limiting, pointing at
the deprotonation of I to produce II as the rate-limiting
step.
These findings agree with our previously reported 38C2-
[
12]
catalyzed isotope-exchange experiments
formation of the Schiff base intermediate in a rapid pre-
that supported
Scheme 5.Mechanistic pathway of the uncatalyzed oxidation reaction.
1
6
18
equilibrium.The antibody-catalyzed O/ O exchange of the
1
8
carbonyl oxygen of cycloheptanone in H O, was found to
2
À1
be very fast (kcat =418 min , K =21mm).Comparison of
M
these results with the kinetic parameters for the 38C2-cata-
lyzed H/D exchange with the same substrate (kcat =13 min ,
than the enolization step.The irreversibility of the oxidation
steps renders the enolization step practically irreversible as
well, thus providing an opportunity of directly measuring
the Hammett correlation coefficient of the enolization pro-
cess.
À1
K =16mm) indicated that the formation of the enamine in-
M
termediate is approximately 32 times slower than the forma-
tion of the Schiff base.These studies have also indicated
that in the 38C2-catalyzed aldol condensation reaction, the
To examine the substituent effect on the binding constant
4
CÀC bond-forming step is approximately 10 times slower
we plotted the log(K ) values against the Hammett s sub-
M
[12]
than the formation of the enamine. Accordingly, we be-
lieve that this bond-forming step is rate-limiting in the
stituents constants (Figure 5).Interestingly, a nearly linear
24H6-catalyzed aldol reaction as well.
It is interesting to compare the Hammett correlation coef-
ficient of the 24H6-catalyzed oxidation of 4 (1=4.7) with
that of the uncatalyzed reaction (1=1.7, Figure 4), which
M
Figure 5.Log( K ) versus s for the oxidation reaction.
correlation with a positive slope (1=3.13) was found. In our
previously reported Hammett correlation studies with aldol
reaction catalyzed either by 38C2 or by 24H6 we employed
substituted benzaldehyde acceptors rather than substituted
Figure 4.Hammet correlation for the uncatalyzed oxidation reaction.
donors.Very little effect of the substituent on binding ( K )
M
[4,8]
was performed under the same conditions in the absence of
the antibody.The proposed mechanistic pathway of the
latter reaction, which is probably general base-catalyzed, is
outlined in Scheme 5.Again, as is the case in the antibody-
catalyzed reaction, the positive 1 value of the uncatalyzed
reaction indicates that the rate-limiting step is the formation
of the enol intermediate (VII).As in the catalyzed reaction,
the two irreversible one-electron oxidation steps that form
intermediate VIII and the final product (5) are much faster
was observed in those cases.
The large value of 3.13, as
well as the remarkable difference between the substituent
effects on the binding of either donor or acceptor, suggests
that electronic rather than steric factors are dominant in
binding.These findings support the assumption that a chem-
ical bond is formed between the antibody and the donor
molecule during the catalytic event.This notion has been
[6]
previously suggested for all aldolase antibodies, including
[4]
24H6, based on chemical modifications (reductive alkyla-
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
2163
Chem. Eur. J. 2004, 10, 2159 ± 2165
www.chemeurj.org

E.Keinan et al.
FULL PAPER
tion of the antibody lysine residue with acetone and borohy-
dride) and on spectral data (formation of a new UV absorp-
tion band at 316±318 nm, which was attributed to the forma-
tion of a vinylogous amide upon binding of acetylacetone to
the lysine residue).
The solution was then poured into water and extracted with diethyl
ether, washed with saturated aqueous K CO , and dried over MgSO .
2 3 4
The solvent was removed under reduced pressure and the residue was
purified by flash chromatography.
1
4
-Chlorobenzil (5a): H NMR: d=7.93 (m, 4H), 7.48 ppm (m, 5H); MS
À
(
CI): m/z: 244 [M] .
In conclusion, oxidation of a series of substituted benzoins
to benzils by potassium ferricyanide was used as a mechanis-
tic probe to better understand the initial steps of the 24H6-
catalyzed aldol condensation reaction.Hammett correlation
1
4
-Methylbenzil (5c): H NMR: d=7.95 (brd, J=7.4 Hz, 2H), 7.85 (d, J=
8.1 Hz, 2H), 7.61 (d, J=7.4 Hz, 1H) 7.48 (t, J=7.4 Hz, 2H), 7.29 (d, J=
.1 Hz, 2H), 2.42 ppm (s, 3H); MS (CI): m/z: 244.1 [M] .
À
8
4
-Methoxybenzil (5d): 4-Methoxybenzoin (0.2 g, 0.83 mmol) was refluxed
for 5 h with a solution of copper sulfate (0.4 g) and pyridine (5 mL) in
water (1 mL).Aqueous HCl (2 m, 5 mL) was added to the cooled mixture
and the aqueous layer was extracted three times with dichloromethane.
The combined organic extracts washed twice with brine and dried over
MgSO4. The solvent was removed under reduced pressure and the resi-
due was purified by flash chromatography. H NMR: d=7.97 (brd, J=
.4 Hz, 2H), 7.94 (brd, J=8.4 Hz, 2H), 7.61 (d, J=7.4 Hz, 1H), 7.50 (t,
J=7.4 Hz, 2H), 6.96 (dd, J=8.0, 1.9 Hz, 2H), 3.87 ppm (s, 3H); MS (CI):
(
1=4.7) of log(k ) versus the s substituent constant re-
cat
vealed that the rate-limiting step in this oxidation reaction is
the conversion of the Schiff base (initially formed from the
ketone and an active site lysine residue) to an enamine in-
termediate.In addition, linear correlation ( 1=3.13) was
1
8
found between log(K ) and the Hammett s substituents
M
constants.This correlation suggests that electronic rather
than steric factors are dominant in the antibody±substrate
binding phenomenon, consistent with reversible formation
of a Schiff base intermediate.
+
m/z: 241 [M+H] .
Antibody-catalyzed reactions: All antibody-catalyzed reactions were car-
ried out in PBS (50 mm phosphate, 100 mm NaCl, pH 9.0) containing
0% DMSO as organic co-solvent.The progress of the reactions was
monitored by HPLC using an RP Supelcosil LC18 column, with the ini-
tial rates calculated by regression analysis.Antibody concentration was
1
1
0.7 mm for 24H6.The rate of the uncatalyzed reactions was subtracted.
Experimental Section
1
General methods: Most H NMR spectra were recorded on a Bruker
Acknowledgment
AM200 spectrometer, operating at 200 MHz using CDCl
3
as a solvent
(
unless otherwise specified).Positive ion mass spectra, using the fast
We thank the Israel±US Binational Science Foundation, the German±Is-
raeli Project Cooperation (DIP), and the Skaggs Institute for Chemical
Biology for financial support.
atom bombardment (FAB) technique, were obtained on a VG ZAB-VSE
double-focusing, high-resolution mass spectrometer equipped with either
a cesium or sodium ion gun.EI-MS spectra were measured on a Finnigan
MAT-711 spectrometer.CI-MS spectra were measured on a Finnigan
TSQ-70 spectrometer.UV/Vis spectra were recorded on a Shimadzu UV-
1
601 spectrometer.Long period reactions were maintained at 25 8C by
[1] C.-H. Wong, T. D. Machajewski, Angew. Chem. 2000, 112, 1406±
1430; Angew. Chem. Int. Ed. 2000, 39, 1352±1374, and references
therein.
using a Friocell incubator.TLC was performed on glass sheets pre-coated
with silica gel (Merck, Kieselgel 60, F254, Art.5715).Column chromato-
graphic separations were performed on silica gel (Merck, Kieselgel 60,
[2] J.Wagner, R.A.Lerner, C.F.Barbas III,
Science 1995, 270, 1797.
2
30±400 mesh, Art.9385) under pressure (flash chromatography).Dry
[3] a) T.Hoffmann, G.Zhong, B.List, D.Shabat, J.Anderson, S.Gra-
matikova, R.A. Lerner, C.F. Barbas III, J. Am. Chem. Soc. 1998,
120, 2768; b) B.List, D.Shabat, C F. .Barbas III, R A. .Lerner,
Chem. Eur. J. 1998, 4, 881; c) G.Zhong, D.Shabat, B.List, J.Ander-
THF (Sure-seal) was purchased from Aldrich.HPLC analyses were car-
ried out with a Merck±Hitachi Lachrom system equipped with an L-7100
pump, an L-7400 UV/Vis detector, and a D-7000 system manager with a
Supelco RP LC-18 analytical column.All starting materials and reagents,
including benzoin (4b) and benzil (5b), were purchased from Aldrich.
son, S.C.Sinha, R.A.Lerner, C.F.Barbas III,
110, 2609; Angew. Chem. Int. Ed. 1998, 37, 2481; d) S.C. Sinha, J.
Sun, G.Miller, G., C.F.Barbas III, R.A.Lerner, Org. Lett. 1999, 1,
Angew. Chem. 1998,
General procedure for a-hydroxylation: The appropriate ketone
(
1 mmol) in dry THF (6 mL) was added dropwise to a freshly prepared,
1623e) J .M .Turner, T.Bui, R .A .Lerner, C .F .Barbas III, B.List,
Chem. Eur. J. 2000, 6, 2772.
[4] H.Shulman, C.Makarov, A.K.Ogawa, F.Romesberg, E.Keinan
Am. Chem. Soc. 2000, 122, 10743.
cold (À788C) solution of LDA (1.1 mmol) in dry THF (2 mL), and the
mixture was stirred at the same temperature for 30 min.A solution of
phenylsulfonyloxaziridine (1.4 mmol in 6 mL THF) was added dropwise,
and the progress of the reaction was monitored by TLC.Upon comple-
J.
[5] G.Zhong, R .A .Lerner .C .F .Barbas III,
Angew. Chem. 1999, 111,
tion the reaction was quenched with saturated aq NH
was then allowed to warm to room temperature, was extracted with di-
ethyl ether, was washed with saturated aq Na (2î15 mL) and brine
2î15 mL), and was dried over MgSO .The solvent was removed under
4
Cl; the mixture
3957; Angew. Chem. Int. Ed. 1999, 38, 3738.
[6] C.F.Barbas III, A.Heine, G.Zhong, T.Hoffmann, S.Gramatikova,
R.Bjornestedt, B.List, J.Anderson, E.A.Stura, I.A.Wilson, R.A.
Lerner, Science 1997, 278, 2085.
2 2 3
S O
(
4
reduced pressure and the residue was purified by flash chromatography.
[7] a) L.P.Hammett, J. Am. Chem. Soc. 1937, 59, 96; b) C.D.Johnson,
The Hammett Equation; Cambridge University Press, Cambridge
1
4
8
7
-Chlorobenzoin (4a): H NMR: d=7.83 (d, J=8.6 Hz, 2H) 7.31 (d, J=
1
973.
[8] H.Shulman, E.Keinan, Bioorg. Med. Chem. Lett. 1999, 9, 1745.
9] a) P.Christen, J F. .Riordan, Biochemistry 1968, 7, 1531; b) M.J.
Healy, P.Christen, J. Am. Chem. Soc. 1972, 94, 7911.
.6 Hz, 2H), 7.28 (m, 5H), 5.86 (brd, J=7 Hz, 1H), 4.45 ppm (brd, J=
Hz, 1H); MS (CI): m/z: 247 [M+H] .
+
[
1
4
-Methylbenzoin (4c): H NMR (300 MHz): d=7.76 (d, J=8.4 Hz, 2H)
7
.25 (m, 5H), 7.22 (d, J=8.4 Hz, 2H), 5.87 (brd, J=5 Hz, 1H) 4.55 (brd,
[
[
10] M.J.Healy, P.Christen, Biochemistry 1973, 12, 35.
11] H.Shulman, Ph D. .Dissertation, Technion–Israel Institute of Tech-
nology, Haifa, Israel, 2000.
+
J=5 Hz, 1H) 2.29 ppm (s, 3H); MS (CI): m/z: 227 [M+H] .
1
4
5
3
1
-Methoxybenzoin (4d): H NMR: d=7.89 (d, J=8.6 Hz, 2H) 7.29 (m,
H), 6.84 (d, J=8.6 Hz, 2H), 5.86 (brs, 1H) 4.64 (brs, 1H) 3.79 ppm (s,
[
12] A.Shulman, D.Sitry, H.Shulman, E.Keinan,
29.
Chem. Eur. J. 2002, 8,
1
3
H); C NMR: d=196.5, 164.0, 139.6, 131.5, 129.0, 128.3, 127.6, 126.5,
13.9, 75.7, 55.4 ppm; MS (CI): m/z: 243 [M+H] .
2
+
[
13] V.K.Aggarwal, X.de Vicente, B.Pelotier, I.P.Holmes, R.V.Bon-
General procedure for selenium dioxide oxidation: A solution of either
nert, Tetrahedron Lett. 2000, 41, 10327.
3
a or 3c (1 mmol) and selenium dioxide (1.1 mmol) were dissolved in
[14] M.S.Malik, S.N.Rastogi, M.M.Singh, S.Sreenivasulu, A.Agniho-
7
0% acetic acid (5 mL) and the mixture was heated to 908C for 12 h.
tri, V.Kalra, V.P.Kaboj, Eur. J. Med. Chem. 1989, 24, 193.
2164
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 2159 ± 2165

Antibody-Catalyzed Benzoin Oxidation
2159 ± 2165
[
15] F .A .Davis, A .C .Sheppard, B- .C .Chen, S.Haque,
J. Am. Chem.
[19] C.D.Johnson, The Hammett Equation, Cambridge University Press,
Cambridge, 1973, pp.181±182.
Soc. 1990, 112, 6679.
[
[
16] E.J.Corey, J.P.Schaefer, J. Am. Chem. Soc. 1960, 82, 918.
[20] a) W.P.Jencks, Prog. Phys. Org. Chem. 1964, 2, 63; b) W.P.Jencks,
Catalysis in Chemistry and Enzymology, Dover, New York, 1987,
pp.4.
17] R.J.Cremlyn, F.J.Swinbourne, O.O.Shode, J.Lynch,
Chem. 1987, 24, 117.
J. Heterocycl.
[
18] Correlation Analysis in Chemistry. Recent Advances (Eds.: N. B.
Chapman, J.Shorter), Plenum, New York, 1978, pp.455±481.
Received: April 8, 2003
Revised: December 29, 2003 [F5034]
Chem. Eur. J. 2004, 10, 2159 ± 2165
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
2165