Copying nature's mechanism for the decarboxylation of β-keto acids into catalytic antibodies by reactive immunization

Article abstract of DOI:10.1021/ja9620797

Reactive immunization was used to generate catalytic antibodies that use the enamine mechanism common to the natural class I aldolase enzymes. In order to investigate the possibility of exploiting the imine and enamine intermediates programmed into antibo

Full text of DOI:10.1021/ja9620797

11720
J. Am. Chem. Soc. 1996, 118, 11720-11724
Copying Nature’s Mechanism for the Decarboxylation of
â-Keto Acids into Catalytic Antibodies by Reactive
Immunization
Robert Bjo1rnestedt, Guofu Zhong, Richard A. Lerner,* and Carlos F. Barbas III*
Contribution from The Skaggs Institute for Chemical Biology and the Department of
Molecular Biology, The Scripps Research Institute, 10666 North Torrey Pines Road,
La Jolla, California 92037
ReceiVed June 20, 1996X
Abstract: Reactive immunization was used to generate catalytic antibodies that use the enamine mechanism common
to the natural class I aldolase enzymes. In order to investigate the possibility of exploiting the imine and enamine
intermediates programmed into antibody catalysts by reactive immunization and the features which antibody aldolases
share with naturally evolved catalysts, we have studied their ability to catalyze the decarboxylation of structurally
related â-keto acids. Both aldolase antibodies were shown to efficiently catalyze the decarboxylation of two hapten-
related â-keto acids with rate enhancements (kcat/kuncat) between 4959 and 14 774. Inhibition studies support the
role of an essential lysine residue in the active site of the antibodies and the formation of a cyanide accessible imine
intermediate in the mechanism. Investigation of the decarboxylation reaction of 2-{3′-(4′′-acetamidophenyl)propyl}-
acetoacetic acid, 4, to 6-(4′-acetamidophenyl)-2-hexanone, 12, in the presence of 18O-labeled water by electrospray
mass spectrometry revealed obligatory incorporation of 18O in the antibody-catalyzed reaction consistent with
decarboxylation proceeding via an imine intermediate. These studies demonstrate that reactive immunization may
be utilized to program in fine detail the mechanism of catalytic antibodies and our ability to exploit the programmed
reaction coordinate for different catalytic tasks.
Introduction
nism of these aldolase antibodies was programmed using
reactive immunization3 with the â-diketone hapten 1 as shown
in Scheme 1. The â-diketone 1 is a chemical trap. Reaction
of an ꢀ-amino group of lysine within the active site of the
antibody with a keto group of 1 results in the formation of a
tetrahedral carbinolamine, which is dehydrated to an imine and
subsequently tautomerized to the stable vinylogous amide 3
shown in Scheme 1. The imine and enamine intermediates
developed along the reaction coordinate of the reaction between
the antibody and the diketone and the aldol reaction are also
found along the reaction coordinates of other reactions, for
example, decarboxylations, racemizations, and alkylation reac-
tions to name a few. Indeed, it is known that some naturally
occurring class I aldolases are bifunctional. In addition to
catalyzing the aldol reaction, these bifunctional catalysts also
catalyze the decarboxylation of â-keto acids.5 In order to
investigate the possibility of exploiting common intermediates
programmed into antibody catalysts by reactive immunization
and the features which antibody aldolases share with naturally
evolved catalysts, we have studied their ability to catalyze the
decarboxylation of structurally related â-keto acids.
The study of amine-catalyzed decarboxylations of â-keto
acids occupies a special place in the history of bioorganic
chemistry and enzymology. In a series of elegant studies
Westheimer and colleagues elucidated in detail the mechanism
by which the enzyme acetoacetate decarboxylase catalyzes the
decarboxylation of acetoacetic acid.1 These studies demon-
strated that the reaction proceeds by the formation of a Schiff
base between the ꢀ-amino group of a lysine residue in the
enzyme and acetoacetate, followed by decarboxylation to form
an enamine which is then tautomerized to a Schiff base and
finally hydrolyzed to release acetone and the free enzyme. The
enzymatic mechanism is analogous to that originally proposed
by Pederson in 1934 for simple amine-catalyzed decarboxyla-
tions.2 Herein we report the programming of the chemical
mechanism used by the natural enzyme acetoacetate decarboxy-
lase into catalytic antibodies by the process of reactive im-
munization.
Previously, antibodies capable of catalyzing intermolecular
aldol reactions with control of stereochemistry in both Cram
and anti-Cram directions were produced.3a These catalytic
antibodies were shown to use the enamine mechanism common
to the natural class I aldolase enzymes.4 The reaction mecha-
Results and Discussion
Two antibody aldolases, 38C2 and 33F12, were available for
study.3a,6 These antibodies catalyze the retroaldol reaction of
X Abstract published in AdVance ACS Abstracts, November 15, 1996.
(1) The following constitutes a partial listing of the most relevant
references: (a) Hamilton, G. A.; Westheimer, F. H. J. Am. Chem. Soc. 1959,
81, 6332. (b) Fridovich, I.; Westheimer, F. H. J. Am. Chem. Soc. 1962, 84,
3208. (c) Laursen, R. A.; Westheimer, F. H. J. Am. Chem. Soc. 1966, 88,
3426. (d) Tagaki, W.; Guthrie, J. P.; Westheimer, F. H. Biochemistry 1968,
7, 905. (e) Frey, P. A.; Kokesh, F. C.; Westheimer, F. H. Biochemistry
1971, 14, 7266. (f) Kokesh, F. C.; Westheimer, F. H. J. Am. Chem. Soc.
1971, 93, 7270. (g) Schmidt, Jr., D. E.; Westheimer, F. H. Biochemistry
1971, 10, 1249. (h) Autor, A. P.; Fridovich, I. J. Biol. Chem. 1970, 245,
5214. (i) Westheimer, F. H. Proceedings of the Robert A. Welch Foundation
Conferences on Chemical Research; Robert A. Welch Foundation, Houston,
1971; Vol. 15, pp 7-50. (j) Westheimer, F. H. Tetrahedron 1995, 51, 3.
(2) Pederson, K. J. J. Phys. Chem. 1934, 38, 559.
(3) (a) Wagner, J.; Lerner, R. A.; Barbas, C. F., III Science 1995, 270,
1797. (b) Wirshing, P.; Ashley, J. A.; Lo, C. -H. L.; Janda, K. D.; Lerner,
R. A. Science 1995, 270, 1775.
(4) Rutter, W. J. Fed. Proc. Am. Soc. Exp. Biol. 1964, 23, 1248. Morris,
A. J.; Tolan, D. R. Biochemistry 1994, 33, 12291.
(5) Kobes, R. D.; Dekker, E. E. Biochem. Biophys. Res. Commun. 1967,
27, 607. Nishihara, H.; Dekker, E. E. J. Biol. Chem. 1972, 247, 5079.
Vlahos, C. J.; Dekker, E. E. J. Biol. Chem. 1986, 261, 11049.
(6) Nucleic acid sequencing of the genes coding for these antibodies
has revealed that they are somatic mutants of one another and differ by 18
amino acids in their variable regions.
S0002-7863(96)02079-3 CCC: $12.00 © 1996 American Chemical Society

Decarboxylation of â-Keto Acids
J. Am. Chem. Soc., Vol. 118, No. 47, 1996 11721
Scheme 1
in the aldol reaction. Other antibodies which bound hapten 1
with high affinity albeit in a noncovalent fashion were also
studied and were found not to catalyze the decarboxylation
reactions.
The current study differs from decarboxylations of activated
model compounds catalyzed previously10 with antibodies in
studies designed to elucidate medium effects in enzyme-
catalyzed reactions. Our goal is to create an enzyme that uses
a mechanism common to nature’s catalyst, thereby testing the
ability of the experimenter to program a detailed catalytic
mechanism into antibodies. To test whether this was achieved,
the mechanism by which antibodies 38C2 and 33F12 operate
was studied further. Inhibition of decarboxylation activity, as
well as aldolase activity,3a was complete in the presence of the
hapten 2 or 2,4-pentanedione when spectrophotometric titration
indicated the formation of a single vinylogous amide species
per active site. This is analogous to the inhibition of the enzyme
acetoacetate decarboxylase by acetopyruvate.1d Additionally,
incubation of antibody 38C2 and 4 in the presence of 1% v/v
acetone and 50 µM cyanide resulted in 88% inhibition of
decarboxylase activity. Activity decreases of 23% and 15%
were observed when the study was performed with the addition
of either acetone or cyanide alone. The synergistic effect of
cyanide and acetone suggests that inhibition is due to attack of
the acetone-antibody imine by cyanide to form a covalent
aminonitrile adduct. These data are similar to those reported
for acetoacetate decarboxylase and the bifunctional enzyme
2-keto-4-hydroxyglutarate aldolase.1h,5 Cyanide alone inhibits
the antibody-catalyzed decarboxylation reaction in a concentra-
tion dependent manner, indicating that imine 15 is accessible
to attack by cyanide. These experiments confirm that decar-
boxylation and aldol reactions occur at the same site and are
dependent on an essential ꢀ-amino group of a lysine residue in
the active site.
To confirm not only that decarboxylation is dependent on a
lysine residue but that imine formation is obligatory for
decarboxylation, we studied the 38C2-catalyzed decarboxylation
of 4 in buffered 18O-labeled water. If the reaction proceeds as
described in Scheme 4, the product ketone 12 must incorporate
18O in the keto functionality which is obtained following
decarboxylation and hydrolysis of the imine 15. While 4 proved
unstable to GC-MS analysis, electrospray MS allowed for the
study of this reaction. Initially we determined that purified
substrate 4 and product 12 containing 16O produced M+ ions at
the expected m/z of 278 for the protonated acid of 4 and m/z
256 for the product 12, corresponding to a (12 + Na+) ion. A
reaction mixture containing 4, 2% of the 16O-containing product
12 as an internal standard, and antibody 38C2 in 18O water was
followed by direct analysis of the reaction using electrospray
MS. The antibody-catalyzed conversion of 4 to 12 results in
an obligatory increase of two mass units to yield a new peak at
m/z of 258 corresponding to the incorporation of 18O into 12,
Figure 2. The appearance of a peak at m/z 280 at 160 min
corresponds to the exchange of 18O into the substrate 4. Further
study of the simple exchange reaction in the absence of protein
or in the presence of a noncatalytic antibody showed that the
incorporation of 18O into the m/z 280 species is not significantly
catalyzed by antibody 38C2 and constitutes the background
exchange rate. From these experiments we conclude that the
reaction proceeds as described in Scheme 4 where the decar-
boxylation step is rapid compared to formation and hydrolysis
of 15.
4-hydroxy-6-phenyl-2-hexanone but not 2-hydroxy-6-phenyl-
4-hexanone, suggesting preferential attack of the lysine at the
2 position. This reactivity and the hapten structure suggested
4 would be a suitable â-keto acid substrate if 38C2 and 33F12
could tolerate substitution at the 3 position of the hapten. In
order to map the geometry of the active sites of these antibodies,
we studied their reactions with a series of structurally related
â-diketones. Both antibodies 38C2 and 33F12, shown to be
capable of reacting with the linear diketones 2 and 2,4-
pentanedione,3a also reacted with the branched diketones
3-methyl-2,4-pentanedione, 2-acetylcyclopentanone, and 2-ace-
tylcyclohexanone to form stable vinylogous amides which
exhibited intense ultraviolet absorptions, λmax, at 328, 335, and
335 nm, respectively, in a spectral region clear of absorption
from protein. These experiments suggested an active site
geometry which tolerates considerable substitution at the 3
position of the hapten molecule. On the basis of these
experiments, the â-keto acids 4 and 5 were synthesized.
The lithium salt 11 of compound 4 was prepared in six steps
from 4-iodoaniline as shown in Scheme 2. Acetylation of
4-iodoaniline (85%) followed by palladium-catalyzed arylation
of allyl alcohol under Heck conditions7 generated compound 7
(69%). Reduction of compound 7 followed by reaction with
triphenylphosphine and bromine provided compound 9 (84%).
R-Alkylation of ethyl acetoacetate with 9 under basic conditions8
afforded the ester 10 (61%) which upon saponification with
LiOH provided compound 11 (94%). The ketone 12 was
prepared by saponification of compound 10 followed by acid-
catalyzed decarboxylation of the resulting â-keto acid. The
lithium salt 14 of compound 5 was prepared from compound 6
as shown in Scheme 3. In this synthesis, palladium-catalyzed
arylation7 of 1-methyl-2-propen-1-ol with compound 6 produced
the ketone 13 (81%). Direct carboxylation of ketone 13 with
magnesium methyl carbonate9 in DMF afforded 14 (51%) upon
workup with LiOH and purification by RP-HPLC.
The â-keto acids 4 and 5 prepared by acidification of their
respective lithium salts 11 and 14 were tested as substrates for
decarboxylation by antibodies 38C2 and 33F12. The initial rates
of decarboxylations catalyzed by 38C2 and 33F12 were studied
as a function of the concentration of substrates 4 and 5. The
kinetic parameters are summarized in Table 1. As shown in
Figure 1, the decarboxylation of 4 by antibody 38C2 followed
Michaelis-Menten saturation kinetics. For both antibodies,
catalysis of decarboxylation of 4 or 5 was efficient and exhibited
multiple turnovers and no product inhibition. Catalysis of the
decarboxylation of 5 by 38C2 at a rate lower than that observed
for substrate 4 is indicative of the preferential reactivity of the
antibody with the keto functionality at the 2 position as observed
(7) Heck, R. F.; Nolley, J. P. J. Org. Chem. 1972, 37, 2320.
(8) Benetti, S.; Romagnoli, R; De Risi, C; Spalluto, G.; Zanirato, V.
Chem. ReV. 1995, 95, 1065.
(9) Tirpak, R. E.; Olsen, R. S.; Rathke, M. W. J. Org. Chem. 1985, 50,
4877.
(10) Lewis, C.; Kra¨mer, T.; Robinson, S.; Hilvert, D. Science 1991, 253,
1019. Tarasow, T. M.; Lewis, C.; Hilvert, D. J. Am. Chem. Soc. 1994, 116,
7959. Lewis, C.; Paneth, P., O’Leary, M. H.; Hilvert, D. J. Am. Chem.
Soc. 1993, 115, 1410. Ashley, J. A.; Lo, C.-H.; McElhaney, G. P.;
Wirsching, P.; Janda, K. D. J. Am. Chem. Soc. 1993, 115, 2515.

11722 J. Am. Chem. Soc., Vol. 118, No. 47, 1996
Bjo¨rnestedt et al.
Scheme 2a
a Conditions: (a) Ac2O, Et3N, CH2Cl2, rt, 2 h. (b) Allyl alcohol, PdCl2, Bu4NCl, NaHCO3, DMF, rt. (c) NaBH4, CH3OH. (d) Br2, Ph3P, DMF.
(e) Ethyl acetoacetate, CH3ONa, CH3OH, reflux. (f) LiOH, H2O, rt. (g) (i) NaOH, H2O. (ii) 0.05 M H2SO4, 50 °C.
Scheme 3a
a Conditions: (a) 1-Methyl-2-propen-1-ol, PdCl2, Bu4NCl, NaHCO3,
DMF, rt. (b) (i) MMC, DMF, 120 °C, 5 h. (ii) HCl, 0 °C. (c) LiOH,
H2O.
Table 1. Kinetic Parameters for Antibody-Catalyzed
Decarboxylation Reactionsa
Figure 1. Michaelis-Menten plot for the conversion of 4 to 12
substrate
kcat (min-1
)
Km (mM)
kun (min-1)
1.11 × 10-5
1.23 × 10-5
kcat/kun
catalyzed by antibody 38C2. Product formation was monitored by high-
performance liquid chromatography. The curve was fit to the experi-
mental data by nonlinear regression analysis using GraFit software.17
4
0.164
0.125
0.082
0.061
0.95
1.82
0.85
0.20
14774
11261
6666
4b
5
5b
4959
of AAN-catalyzed decarboxylation of acetoacetate have dem-
onstrated that the imine formed in this reaction decarboxylates
at a rate 300 000 times faster than acetoacetic acid itself.11b,d
The rate of acetoacetate decarboxylase-catalyzed decarboxyla-
tion is only 156 times that of the rate of decarboxylation of the
N-cyanomethylimine of acetoacetate allowing most of the
catalytic power of the enzyme to be assigned to the enzyme’s
ability to facilitate imine formation with the amine group of
the active site lysine.11b,13 Since imine formation is optimal at
a given pH11c with an amine whose pKa approximates the given
pH, phenylalanine ethyl ester (PheOEt) whose amine has a pKa
of 7.212a is an appropriate model compound for the study of
amine catalyzed decarboxylations at neutral pH. We determined
the second-order rate constant for PheOEt-catalyzed decarboxy-
lation of 4 to be 6.9 × 10-6 min-1 mM-1. PheOEt has been
studied as a model compound for peptide-catalyzed decarboxy-
lation of oxaloacetate, and the second-order rate constant for
a Assay conditions: 10 mM phosphate, 150 mM NaCl, pH 7.4, in
the presence of 1.25 µM antibody. Reaction using antibody 33F12.
b
The catalytic efficiency of antibody 38C2 as a decarboxylase
can be brought into perspective by comparison with decarboxy-
lation reactions catalyzed by simple amines,11 designed12a and
combinatorially optimized peptides,12b and the enzyme acetoac-
etate decarboxylase.1,13 The most effective simple amine
catalyst of acetoacetate decarboxylation is aminoacetonitrile,
AAN.11b This amine has served as a model for acetoacetate
decarboxylase catalyzed decarboxylations since its pKa ap-
proximates that of the active-site lysine of the enzyme. Studies
(11) (a) Guthrie, J. P.; Jordan, F. J. Am. Chem. Soc. 1972, 94, 9132. (b)
Guthrie, J. P.; Jordan, F. J. Am. Chem. Soc. 1972, 94, 9136. (c) Hine, J.;
Via, F. A. J. Am. Chem. Soc. 1972, 94, 190. (d) O’Leary, M. H.; Baughn,
R. L. J. Am. Chem. Soc. 1972, 94, 626.
(12) (a) Johnsson, K.; Allemann, R. K.; Widmer, H.; Benner, S. A. Nature
1993, 365, 530. (b) Pe´rez-Paya´, E.; Houghten, R. A.; Blondelle, S. E. J.
Biol. Chem. 1996, 271, 4120.
(13) Highbarger, L. A.; Gerlt, J. A.; Kenyon, G. L. Biochemistry 1996,
35, 41.

Decarboxylation of â-Keto Acids
J. Am. Chem. Soc., Vol. 118, No. 47, 1996 11723
Scheme 4
PheOEt-catalyzed decarboxylation of oxaloacetate has been
reported. This together with the new data provided here and
the data on acetoacetate decarboxylase-catalyzed decarboxyla-
tions allows the relative efficiencies over amine catalysis (kcat/
Km)/kNH2, the rate enhancement over background (kcat/kuncat), and
Figure 2. Mass spectra from electrospray MS showing the antibody-
catalyzed conversion of â-keto acid 4 to ketone 12 in buffered 98%-
18O-labeled water. The m/z of 256 corresponds to ketone 12 already
present at t ) 0 whereas the antibody catalyzed conversion of 4 (m/z
278) to 12 incorporated 18O to produce an m/z of 258. The reaction
contained 1.5 mM â-keto acid 4, 11 mg/mL antibody 38C2, and
phosphate-buffered saline in 18O-labeled water. Immediately before
analysis the samples were diluted 10-fold in methanol.
the effective molarity (kcat/kNH ) of the active site amines of
2
peptide and protein catalysts to be compared (Table 2). The
antibody 38C2 provides a 25 000-fold enhancement of the
efficiency of the decarboxylation of 4 as compared to the
nonenzymic PheOEt amine-catalyzed process. This is a sig-
nificantly greater enhancement than observed for the peptide
catalysts oxaldie-112a and Ox-Opt12b and much lower than that
observed for acetoacetate decarboxylase.1,13 The effectiveness
of the active site amine of 38C2 is also indicated by an effective
molarity of 24 M which exceeds that observed in most model
systems14 and other catalytic antibodies.15 With respect to
catalysis of the aldol reaction by 38C2, steps following imine
formation are rate limiting in contrast to the decarboxylase
activity described here where imine formation is rate limiting.
Table 2. Kinetic Parameters for Antibody-Catalyzed
Decarboxylation Reactionsa
catalyst
kcat/kuncat
(kcat/Km)/kamine
kcat/kamine (M)
AAD
38C2
5.2 × 109 a
14774
423d
9.5 × 106 b
2.5 × 104 c
48.5d
7.8 × 107
23.8c
oxaldie I
0.42d
Ox-Opte
1159e
205e
1.16e
a kcat and Km of the acetoacetate decarboxylase (AAD)-catalyzed
b
decarboxylation of acetoacetate from ref 13, kuncat from ref 1i. kamine
,
aminoacetonitrile-catalyzed decarboxylation of acetoacetate from ref
1i. c Amine ) phenylalanine ethyl ester. d 14-mer polypeptide catalyzed
decarboxylation of oxaloacetate from ref 12a, amine ) phenylalanine
ethylester. e 18-mer polypeptide (YKLLKELLAKLKWLLRKL-NH2)-
catalyzed decarboxylation of oxaloacetate from ref 12b, amine )
phenylalanine ethyl ester.
Conclusions
One of the goals of bioorganic chemistry is to create new
catalysts that utilize defined chemical mechanisms. These
studies highlight the ability of reactive immunization to ac-
complish this.3 The programming of the identity of amino acid
side chains involved in covalent catalysis constitutes a unique
feature of this approach and is a significant advance over the
use of transition state analogs alone to induce catalytic
antibodies.3a,14 Ultimately, one would wish to combine both
methods wherein the experimenter controls both the global
aspects and mechanistic details of the reaction coordinate. This
approach offers significant advantages over peptide model
systems which use simple electrostatic interactions to bind their
substrates and are thus very limited with respect to their ability
to achieve the concert of effects that characterize an efficient
Experimental Section
General Procedures. Solvents were ACS grade from Fischer.
Methanol and methylene chloride were dried over powdered magnesium
and CaH2, respectively. Reagents were purchased from Aldrich or
Fluka and used without further purification, unless otherwise stated.
NMR was recorded on a Brucker AM-300 spectrometer for samples
in CDCl3 solution. Chemical shifts δ are given in parts per million,
and coupling constant J values are given in hertz. Mass spectra were
provided by the Scripps Research Institute facility (G. Siuzdak). Flash
chromatography was performed with Merck silica gel 60 (230-400
mesh).
N-(4-Iodophenyl)acetamide (6). A solution of 4-iodoaniline (10.0
g, 45 mmol) and triethylamine (19.2 mL, 137 mmol) in 100 mL of
methylene chloride was cooled to 0 °C. Acetic anhydride (13.0 mL,
135 mmol) was added dropwise. The reaction mixture was left at 0
°C for 10 min, and then it was allowed to reach room temperature.
The mixture was kept at ambient temperature for 2 h. The volatile
materials were removed under vacuum, and the solid residue was
dissolved in 30 mL of hot methylene chloride. The white solid product
6 (9.9 g) was obtained by filtration, yield 85%.
and specific catalyst.12,1j The rate enhancement (kcat/kuncat
)
14 774) for antibody 38C2-catalyzed decarboxylation of 4, while
significant, is 105-fold lower than that reported for acetoacetate
decarboxylase.1i,13 Thus, antibody 38C2 represents a catalyst
which having solved the problem of chemical mechanism awaits
fine tuning of interactions along the reaction coordinate of
decarboxylation. The evolution of catalysis of this antibody to
even greater levels may shed light on the factors responsible
for the enormous rate enhancements observed for natural
decarboxylase enzymes.16
3-(4′-Acetamidophenyl)propanal (7). N-(4-Iodophenyl)acetamide
(6) (3.6 g, 14 mmol) was added to 16 mL of dried DMF, and then
tetrabutylammonium chloride (3.9 g, 14 mmol), sodium bicarbonate
(14) Kirby, A. J. AdV. Phys. Org. Chem. 1980, 17, 183.
(15) Lerner, R. A.; Benkovic, S. J.; Schultz, P. A. Science 1991, 252,
659. Schultz, P. G.; Lerner, R. A. Science 1995, 269, 1835.
(16) Preliminary studies suggest that the catalytically active lysine
possesses an ꢀ-amino group with a highly perturbed pKa analogous to that
observed in acetoacetate decarboxylase.1e-g

11724 J. Am. Chem. Soc., Vol. 118, No. 47, 1996
Bjo¨rnestedt et al.
(2.9 g, 35 mmol), and allyl alcohol (1.4 mL, 21 mmol) were added.
The mixture was stirred for 10 min, and then palladium chloride (0.57
g, 3.2 mmol) was added. The reaction mixture was kept stirring at
room temperature for 36 h under an atmosphere of nitrogen. It was
then diluted with 100 mL of ethyl acetate, washed with 25 mL of 5%
hydrochloric acid and 2 × 25 mL of brine, and dried over magnesium
sulfate. Evaporation of solvent gave a crude product, which was
purified by column chromatography on silica gel (ethyl acetate/hexane
) 70/30), 1.85 g of pure product 7 was obtained, yield 69%. 1H NMR
(300 MHz, CDCl3): δ 9.32 (t, J ) 1.3, 1 H), 7.43 (d, J ) 8.4, 2 H),
7.27 (s, br, 1 H), 7.13 (d, J ) 8.4, 2 H), 2.92 (t, J ) 7.5, 2 H), 2.77 (t,
J ) 7.5, 2 H), 2.17 (s, 3 H). MS: m/z (rel intens) 192 (M + H+, 2),
171 (100), 149 (62), 121 (12), 107 (34). C11H13O2N (191.23).
3-(4′-Acetamidophenyl)-1-propanol (8). At 0 °C, sodium boro-
hydride (0.12 g, 3 mmol) was portionly added to 3-(4′-acetamidophe-
nyl)propanal (7) in 25 mL of dried methanol. The reaction mixture
was kept at 0 °C for 1 h, and then it was poured into 200 mL of
ammonium chloride-saturated ice-water. The reaction mixture was
extracted with 3 × 70 mL of ethyl acetate. The combined organic
phases were dried over sodium sulfate. Evaporation of solvent gave
0.47 g of alcohol product 8, yield 91%. 1H NMR (300 MHz, CDCl3):
δ 7.43 (s, br, 1 H), 7.40 (d, J ) 8.5, 2 H), 7.13 (d, J ) 8.5, 2 H), 3.66
(t, J ) 6.4, 2 H), 2.68 (t, J ) 7.7, 2 H), 2.17 (s, 3 H), 1.90 (m, 2 H).
MS: m/z (rel intens) 216 (M + Na+, 52), 194 (M + H+, 27), 176
(100), 154 (46), 136 (39), 107 (16); C11H15O2N (193.25).
N-[4-(3′-Bromopropyl)phenyl]acetamide (9). 3-(4′-Acetamido-
phenyl)-1-propanol (8) (0.40 g, 2 mmol) and triphenylphosphine (0.54
g, 2.1 mmol) were added to 10 mL of dried DMF. At room
temperature, a solution of bromine (0.37 g, 2.3 mmol) in 6 mL of DMF
was added dropwise to the reaction mixture until it became a red
solution (a small excess of bromine). Then the reaction mixture was
kept at room temperature for 1 h. Evaporation of the solvent gave a
residue which was separated by column chromatography (ethyl acetate/
hexane ) 60/40) on silica gel to afford 0.47 g of product 9, yield 92%.
1H NMR (300 MHz, CDCl3): δ 7.58 (s, br, 1 H), 7.43 (d, J ) 8.3, 2
H), 7.14 (d, J ) 8.3, 2 H), 3.39 (t, J ) 6.4, 2 H), 2.74 (t, J ) 7.3, 2
H), 2.17 (s, 3 H), 2.14 (pent, J ) 7.0, 2 H). MS: m/z (rel intens) 257
(M + H+, 19). C11H14OBrN (256.14).
Ethyl 2-{[3′-(4′′-Acetamido)phenyl)propyl}acetoacetate (10). So-
dium (15 mg, 0.65 mmol) was added to 1 mL of dried methanol. After
5 min of stirring, ethyl acetoacetate (0.075 mL, 0.63 mmol) was added
to the sodium methoxide solution. The reaction mixture was warmed
to slowly reflux. The solution of N-[4-(3′-bromopropyl)phenyl]-
acetamide (9) (130 mg, 0.5 mmol) in 2 mL of dried methanol was
then added dropwise to the reaction mixture in 10 min. The reflux
was kept for an additional 4 h. Evaporation of solvent under vacuum
gave a residue which was separated by column chromatography (ethyl
acetate/hexane ) 60/40) on silica gel to afford 93 mg of product 10,
yield 61%. 1H NMR (300 MHz, CDCl3): δ 7.53 (s, br, 1 H), 7.40 (d,
J ) 8.3, 2 H), 7.09 (d, J ) 8.3, 2 H), 4.18 (q, J ) 7.0, 2 H), 3.42 (t,
J ) 7.5, 1 H), 2.59 (t, J ) 7.5, 2 H), 2.21 (s, 3 H), 2.15 (s, 3 H), 1.85
(m, 2 H), 1.59 (m, 2 H), 1.27 (t, J ) 7.0, 3 H). MS: m/z (rel intens)
289 (M+, 42); C17H23O4N (305.37).
Lithium Salt 11 of 2-{3′-(4′′-Acetamidophenyl)propyl}acetoacetic
Acid (4). Ethyl 2-{3′-(4′′-acetamidophenyl]propyl}acetoacetate (10)
(70 mg, 0.23 mmol) and lithium hydroxide (5.4 mg, 0.22 mmol) were
added to 7.5 mL of water. The mixture was stirred at room temperature,
and after 24 h, the milk-like suspension turned into a clear solution.
The mixture was extracted with 3 × 5 mL of ethyl acetate. The water
phase was evaporated using a freezer-vacuum evaporator, resulting
in a white solid salt product, 11, yield 61 mg, 94%. 1H NMR (300
MHz, D2O/DSS): δ 7.47 (s, br, 1 H), 7.35 (d, J ) 8.3, 2 H), 7.05 (d,
J ) 8.3, 2 H), 3.36 (t, J ) 7.4, 1 H), 2.54 (t, J ) 7.4, 2 H), 2.16 (s, 3
H), 2.11 (s, 3 H), 1.81 (m, 2 H), 1.53 (m, 2 H). MS: m/z (rel intens)
283 (M+, 22); C15H18O4LiN (283.25). After a solution of 11 was
acidified by TFA, 4 was obtained. HR-MS of 4: 278.1391; C15H20O4N+
(calcd 278.1392).
was warmed to 50 °C for 30 min. After cooling to room temperature,
the reaction mixture was extracted with 3 × 15 mL of ethyl acetate
and dried over magnesium sulfate. Evaporation of solvent gave 20
mg of solid product 12, yield 95%. 1H NMR (300 MHz, CDCl3): δ
7.52 (s, br, 1 H), 7.41 (d, J ) 8.2, 2 H), 7.09 (d, J ) 8.2, 2 H), 2.55
(t, J ) 7.2, 2 H), 2.45 (t, J ) 7.0, 2 H), 2.21 (s, 3 H), 2.14 (s, 3 H),
1.64 (m, 4 H). MS: m/z (rel intens) 233 (M+, 46). C14H19O2N
(233.31).
4-(4′-Acetamidophenyl)-2-butanone (13). Synthesis was performed
as described for compound 7 with the exception that allyl alcohol was
substituted with 1-methyl-2-propen-1-ol. Yield 81%.
5-(4′-Acetamidophenyl)-3-oxopentanoic Acid (5) and Its Salt 14.
4-(4′-Acetamidophenyl)-2-butanone (13) (100 mg, 0.48 mmol) was
added to magnesium methyl carbonate (MMC; 2 M in DMF, 2.4 mL).
The mixture was heated to 120 °C for 5 h. The mixture was cooled to
0 °C, and then 30 mL of hydrochloric acid was added dropwise. The
lithium salt 14 was obtained by adding an excess of lithium hydroxide
to the reaction mixture. After filtration, a part of the filtrate was purified
by RP-HPLC to obtain a pure salt, 14. MS: m/z (rel intens) 255 (M+,
11). C13H14O4LiN (255.20). The yield of 14 from 13 was 54%. A
sample of 5 was obtained by acidification of its salt 14. 5-(4′-
Acetamidophenyl)-3-oxopentanoic acid (5): 1H NMR (300 MHz,
CDCl3): δ 9.51 (s, br, 1 H), 7.44 (d, J ) 8.2, 2 H), 7.19 (d, J ) 8.2,
2 H), 3.42 (s, 2 H), 2.60 (t, J ) 7.1, 2 H), 2.44 (t, J ) 7.1, 1 H), 2.20
(s, 3 H). HR-MS: 250.1081; C13H16O4N+ (calcd 250.1079).
Assays. (a) Antibody Assays. The kinetic measurements were
performed in phosphate-buffered saline (10 mM phosphate, 150 mM
NaCl, pH 7.4) in the presence of 2.5 µM antibody active sites (based
on two active sites per 150 kDa). The product formation was followed
by RP-HPLC (Microsorb MV, C-18, 300 Å pore size, 0.45 × 22 cm,
flowrate 1.5 mL/min) monitored at 243 nm. The following isocratic
HPLC conditions were used to separate the substrates from the products
in the antibody assays: 17.5% acetonitrile/82.5% water (0.1% trifluo-
roacetic acid), tR(4) ) 8.3 min and tR(12) ) 16.1 min; 15% acetonitrile/
85% water (0.1% trifluoroacetic acid), tR(5) ) 6.8 min and tR(13) )
9.1 min. The kinetic parameters kcat and Km were determined by
nonlinear regression analysis of experimental data using the GraFit
program package.17 The remaining decarboxylation activity of antibody
38C2 in the presence of 2,4-pentanedione or cyanide was determined
as above. The antibody was preincubated with the diketone for 10
min before addition of substrate.
(b) Amine-Catalyzed Decarboxylation. The primary amine-
catalyzed decarboxylations were performed under the same experimental
conditions as the antibody assay. The second-order rate constants were
determined from plots of experimental data where the concentration
of the amine was varied.
(c) Enamine Formation. The enamine formation between the
catalysts and â-diketones (2, 2,4-pentanedione, 3-methyl-2,4 pentanedi-
one, 2-acetylcyclopentanone, and 2-acetylcyclohexanone) was followed
spectrophotometrically at 318-335 nm in thermostated (20 °C) 100
µL cuvetts using a Cary 3 spectrophotometer.
(d) 18O Incorporation. The electron spray ionization (ESI) mass
spectrometry used to monitor 18O incorporation into 12 was performed
on an API III Perkin Elmer SCIEX triple quadropole mass spectrometer.
In preparation of the sample, lyophilized antibody 38C2 was resus-
pended in 18O-labeled water (95-98% 18O, Cambridge Isotope
Laboratories, Andover, MA) to give a final concentration of 11 mg/
mL. The reaction was started by addition of â-keto acid 4 (1.5 mM),
and aliquots were taken out for analysis of 18O incorporation over time.
Immediately before analysis the samples were diluted 10-fold in
methanol.
Acknowledgment. We are grateful to Brian Bothner and
Gary Suizdak for assistance with MS and Ju¨rgen Wagner for
early contributions to this work. This study was supported by
the NIH (Grant CA27489) and the Skaggs Institute for Chemical
Biology. C.F.B. acknowledges an Investigator Award from the
Cancer Research Institute.
6-(4′-Acetamidophenyl)-2-hexanone (12). Ethyl 2-{3′-(4′′-acet-
amido)phenyl)propyl}acetoacetate (10) (30 mg, 0.10 mmol) was added
to 1 mL of sodium hydroxide (0.25 M, 0.25 mmol). The reaction
mixture was stirred until all starting material 10 was dissolved in water.
Then sulfuric acid (0.05 M, 3.0 mL) was added, and the reaction mixture
JA9620797
(17) Leatherbarrow, R. J. GraFit, Version 3.0, Erithacus Software Ltd.:
Staines, U.K., 1992.