Kinetic mechanism and structural requirements of the amine-catalyzed decarboxylation of oxaloacetic acid

Article abstract of DOI:10.1021/jo8014648

(Chemical Equation Presented) The kinetic and chemical mechanism of amine-catalyzed decarboxylation of oxaloacetic acid at pH 8.0 has been reevaluated using a new and versatile assay. Amine-catalyzed decarboxylation of oxaloacetic acid proceeds via the formation of an imine intermediate, followed by decarboxylation of the intermediate and hydrolysis to yield pyruvate. The decrease in oxaloacetic acid was coupled to NADH formation by malate dehydrogenase, which allowed the rates of both initial carbinolamine formation (as part of the imination step) and decarboxylation to be determined. By comparing the rates observed for a variety of amines and, in particular, diamines, the structural and electronic requirements for diamine-catalyzed decarboxylation at pH 8.0 were identified. At pH 8.0, monoamines were found to be very poor catalysts, whereas some diamines, most notably ethylenediamine, were excellent catalysts. The results indicate that the second amino group of diamines enhances the rate of imine formation by acting as a proton shuttle during the carbinolamine formation step, which enables diamines to overcome high levels of solvation that would otherwise inhibit carbinolamine, and thus imine, formation. The presence of the second amino group may also enhance the rate of the carbinolamine dehydration step. In contrast to the findings of previous reports, the second amino group participates in the reaction by enhancing the rate of decarboxylation via hydrogen-bonding to the imine nitrogen to either stabilize the negative charge that develops on the imine during decarboxylation or preferentially stabilize the reactive imine over the unreactive enamine tautomer. These results provide insight into the precise catalytic mechanism of several enzymes whose reactions are known to proceed via an imine intermediate.

Full text of DOI:10.1021/jo8014648

Kinetic Mechanism and Structural Requirements of the
Amine-Catalyzed Decarboxylation of Oxaloacetic Acid
Nabil K. Thalji,†,‡ William E. Crowe,‡ and Grover L. Waldrop*,†
DiVision of Biochemistry and Molecular Biology, Department of Chemistry, Louisiana State UniVersity,
Baton Rouge, Louisiana 70803
gwaldro@lsu.edu
ReceiVed July 7, 2008
The kinetic and chemical mechanism of amine-catalyzed decarboxylation of oxaloacetic acid at pH 8.0
has been reevaluated using a new and versatile assay. Amine-catalyzed decarboxylation of oxaloacetic
acid proceeds via the formation of an imine intermediate, followed by decarboxylation of the intermediate
and hydrolysis to yield pyruvate. The decrease in oxaloacetic acid was coupled to NADH formation by
malate dehydrogenase, which allowed the rates of both initial carbinolamine formation (as part of the
imination step) and decarboxylation to be determined. By comparing the rates observed for a variety of
amines and, in particular, diamines, the structural and electronic requirements for diamine-catalyzed
decarboxylation at pH 8.0 were identified. At pH 8.0, monoamines were found to be very poor catalysts,
whereas some diamines, most notably ethylenediamine, were excellent catalysts. The results indicate
that the second amino group of diamines enhances the rate of imine formation by acting as a proton
shuttle during the carbinolamine formation step, which enables diamines to overcome high levels of
solvation that would otherwise inhibit carbinolamine, and thus imine, formation. The presence of the
second amino group may also enhance the rate of the carbinolamine dehydration step. In contrast to
the findings of previous reports, the second amino group participates in the reaction by enhancing the
rate of decarboxylation via hydrogen-bonding to the imine nitrogen to either stabilize the negative charge
that develops on the imine during decarboxylation or preferentially stabilize the reactive imine over the
unreactive enamine tautomer. These results provide insight into the precise catalytic mechanism of several
enzymes whose reactions are known to proceed via an imine intermediate.
Introduction
ꢀ-ketocarboxylic acid, the ꢀ-decarboxylation reaction has
received the most attention. ꢀ-Decarboxylation can be catalyzed
by a number of agents including divalent cations,1 protons,2
and enzymes (OAA decarboxylase,3 phosphoenolpyruvate car-
boxykinase,4 malic enzyme,5 and pyruvate kinase6), each of
which serves as an electron sink that stabilizes the incipient
carbanion formed with the loss of CO2.
Oxaloacetic acid (OAA) plays a central role as an intermediate
in several metabolic pathways such as the tricarboxylic acid
cycle, gluconeogenesis, fatty acid biosynthesis, amino acid
degradation, and amino acid biosynthesis. As such, the chemistry
of OAA has been studied extensively, and because OAA is a
* To whom correspondence should be addressed. Tel: 225-578-5209. Fax:
225-578-7258.
Amines have also been shown to catalyze the decarboxylation
of OAA.7 It has been widely postulated that such decarboxy-
lations proceed through an imine intermediate formed from
† Division of Biochemistry and Molecular Biology.
‡ Department of Chemistry.
144 J. Org. Chem. 2009, 74, 144–152
10.1021/jo8014648 CCC: $40.75 2009 American Chemical Society
Published on Web 11/26/2008

Amine-Catalyzed Decarboxylation of Oxaloacetic Acid
SCHEME 1. General Mechanism of Amine-Catalyzed Decarboxylation of OAA
condensation of the ketocarbonyl of OAA and the amine.8
Following imination of OAA, the intermediate decarboxylates
to yield the corresponding pyruvate imine, which subsequently
undergoes hydrolysis to generate pyruvate and the amine catalyst
as shown in Scheme 1.9 The imine formation step can be further
divided into nucleophilic attack by the amine on OAA to form
a carbinolamine and dehydration of the carbinolamine to form
the imine intermediate.
While amine-catalyzed OAA decarboxylation has been known
for a number of years,11 surprisingly, there has never been any
attempt to correlate the structure of different amines with the
rates of the different steps in Scheme 1. Several enzymes, such
as aldolase in glycolysis, transaldolase in the pentose phosphate
pathway, and amino acid transaminases involved in amino acid
catabolism all utilize imine formation to catalyze their respective
reactions.12 In addition, studies of amine-catalyzed decarboxy-
lation are particularly relevant to acetoacetate decarboxylase
which has two proximal lysine residues in the active site that
work in concert to decarboxylate acetoacetate via imine forma-
tion.13 Thus, an understanding of the amine-catalyzed decar-
boxylation of OAA will provide insight into enzymatic trans-
formations that utilize imine formation as part of their catalytic
cycle. Herein, we report on the kinetic mechanism and structural
requirements for the ꢀ-decarboxylation of OAA. Furthermore,
we introduce a novel, enzyme-coupled assay that is more
versatile than previous assays because it can be used either for
presteady-state kinetics to analyze imine formation or for steady-
state kinetics to examine the decarboxylation step.
Of particular interest in this study were diamines, which are
effective catalysts at neutral pH. A previous study of amine-
catalyzed imine formation and ꢀ-decarboxylation of OAA
suggested that increased catalyst basicity would promote tau-
tomerization of the OAA-imine to the unreactive enamine, thus
decreasing the rate of decarboxylation.14 Since aliphatic di-
amines with greater separation between the two amino groups
have been shown to be poorer catalysts than diamines with
amino groups in close proximity,15 the authors concluded that
the decreased catalytic ability of long-chain diamines was due
to their higher basicity.14 However, this assumption does not
address the possibility that the second amino group is directly
involved in the decarboxylation reaction. Through systematic
and detailed kinetic analysis of a variety of diamines and
diamine analogs, we provide strong evidence that the second
amino group of diamines plays a crucial role in diamine-
catalyzed ꢀ-decarboxylation.
Results and Discussion
(1) (a) Pedersen, K. J. Acta Chem. Scand. 1952, 6, 285. (b) Gelles, E.;
Clayton, J. P. Trans. Faraday Soc. 1956, 52, 353. (c) Grissom, C. B.; Cleland,
W. W. J. Am. Chem. Soc. 1986, 108, 5582.
(2) (a) Grissom, C. B.; Cleland, W. W. J. Am. Chem. Soc. 1986, 108, 5582.
(b) Pedersen, K. J. Acta Chem. Scand. 1952, 6, 285.
Assay for Amine-Catalyzed Oxaloacetic Acid Decarboxy-
lation. Early assays of amine-catalyzed OAA decarboxylation
involved measuring the products of the reaction, CO2, or
pyruvate. The amount of CO2 released was measured mano-
metrically17 while formation of pyruvate was coupled to the
NADH-requiring enzyme lactate dehydrogenase.18 A more
recent study utilized absorbance at 280 nm corresponding to
the formation of the OAA-enamine.14 All three of these
experimental approaches have significant shortcomings. The
manometric assay is a fixed-time assay that relies on the
(3) (a) Waldrop, G. L.; Braxton, B. F.; Urbauer, J. L.; Cleland, W. W.; Kiick,
D. M. Biochemistry 1994, 33, 5262. (b) Piccirilli, J. A.; Rozzell, J. D. J.; Benner,
S. A. J. Am. Chem. Soc. 1987, 109, 8084. (c) Jetten, M. S.; Sinskey, A. J. Antonie
Van Leeuwenhoek 1995, 67, 221. (d) Labrou, N. E.; Clonis, Y. D. Arch. Biochem.
Biophys. 1999, 365, 17.
(4) Holyoak, T.; Sullivan, S. M.; Nowak, T. Biochemistry 2006, 45, 8254.
(5) (a) Grissom, C. B.; Cleland, W. W. Biochemistry 1988, 272, 2927. (b)
Park, S.-H.; Harris, B. G.; Cook, P. F. Biochemistry 1986, 25, 3572.
(6) Kiick, D. M.; Cleland, W. W. Arch. Biochem. Biophys. 1989, 270, 647.
(7) (a) Wohl, A.; Oesterlin, C. Ber. Dtsch. Chem. Ges. 1901, 34, 1139. (b)
Hay, R. W. Aust. J. Chem. 1965, 18, 337. (c) Pedersen, K. J. Acta Chem. Scand.
1954, 8, 710. (d) Leussing, D. L.; Raghavan, N. V. J. Am. Chem. Soc. 1980,
102, 5635.
(13) (a) Highbarger, L. A.; Gerlt, J. A. Biochemistry 1996, 35, 41. (b)
Hamilton, G. A.; Westheimer, F. H. J. Am. Chem. Soc. 1959, 81, 6332. (c)
Fridovich, I.; Westheimer, F. H. J. Am. Chem. Soc. 1962, 84, 3208. (d)
Westheimer, F. H. Tetrahedron 1995, 51, 3. (e) Kokesh, F. C.; Westheimer,
F. H. J. Am. Chem. Soc. 1971, 93, 7270.
(8) (a) Hay, R. W. Aust. J. Chem. 1965, 18, 337. (b) Pedersen, K. J. Acta
Chem. Scand. 1954, 8, 710. (c) Leussing, D. L.; Raghavan, N. V. J. Am. Chem.
Soc. 1980, 102, 5635.
(9) An analogous mechanism has also been proposed for amine catalyzed
decarboxylation of acetoacetate, another ꢀ-ketocarboxylic acid.10
(10) (a) Guthrie, J. P.; Jordan, F. J. Am. Chem. Soc. 1972, 94, 9136. (b)
Pedersen, K. J. J. Am. Chem. Soc. 1938, 60, 595.
(11) Wohl, A.; Oesterlin, C. Ber. Dtsch. chem. Ges. 1901, 34, 1139.
(12) McMurry, J.; Begley, T. The Organic Chemistry of Biological Pathways;
Roberts and Company: Englewood, CO, 2005; 172-173, 208-210, 223-
226.
(14) Leussing, D. L.; Raghavan, N. V. J. Am. Chem. Soc. 1980, 102, 5635.
(15) Munakata, M.; Matsui, M.; Tabushi, M. Bull. Chem. Soc. Jpn. 1970,
43, 114.
(16) Cleland, W. W. Methods Enzymol. 1979, 63, 103.
(17) (a) Hay, R. W. Aust. J. Chem. 1965, 18, 337. (b) Kaneko, S. J. Biochem.
(Japan) 1938, 28, 1. (c) Gelles, E.; Clayton, J. P. Trans. Faraday Soc. 1956, 52,
353.
(18) Spetnagel, W. J.; Klotz, I. M. J. Am. Chem. Soc. 1976, 98, 8199.
J. Org. Chem. Vol. 74, No. 1, 2009 145

Thalji et al.
SCHEME 2. Reaction Catalyzed by Malate Dehydrogenase
TABLE 1. Reagent Combinations of the OAA Decarboxylation
Assaya
initial velocity ×
reagents in Cuvette
malate, NAD+,
10-1 (dA340/min)
0
ethylenediamine
malate, malate dehydrogenase,
ethylenediamine
0
FIGURE 1. Representative time course of the malate dehydrogenase
assay for amine-catalyzed decarboxylation of OAA showing (a) the
burst phase corresponding to rapid combination of OAA and the amine
and (b) the slow phase representing the steady-state decarboxylation
of the imine followed by hydrolysis of the product and generation of
another molecule of the OAA-imine. Reagent concentrations: 2.0 M
Tris-HCl, pH 8.0, 76 U/mL malate dehydrogenase, 20 mM malic acid,
50 mM NAD+, and 545 mM ethylenediamine.
malate dehydrogenase,
NAD+, ethylenediamine
malate, malate
0
0.02 ( 0.01
2.97 ( 0.09
dehydrogenase, NAD+
malate, malate dehydrogenase,
NAD+, ethylenediamine
a Reagent concentrations: 20 mM malate, 50 mM NAD+, 76 U/mL
malate dehydrogenase, 70 mM ethylenediamine. All assays performed in
2.0 M Tris-HCl, pH 8.0. The velocities are the average ( S.D. of three
measurements.
used. Thus, if rapid initial decarboxylation were occurring, only
one phase would be observed. The likely alternative is that the
rapid phase in Figure 1 must correspond to the formation of a
covalent intermediate that cannot be reduced by malate dehy-
drogenase, thus driving the equilibrium toward increased
production of OAA and, in turn, toward increased production
of NADH. Since the rapid phase is the first phase observed,
the most likely possibility is that it corresponds to the nucleo-
philic attack of OAA by the amine to yield a carbinolamine. It
is plausible that the rapid phase may correspond to the rate of
carbinolamine dehydration, but only if carbinolamine formation
is faster than dehydration and the equilibrium between OAA
and the carbinolamine lies far toward OAA such that the first
step cannot be seen. The slow phase, on the other hand,
represents the rate of decarboxylation (which appears to be rate-
limiting), hydrolysis of the pyruvate imine, and formation of a
new OAA-imine, which drives the coupling enzyme to produce
more NADH.22 While it is possible that the rate-limiting step
of the reaction is, in fact, the hydrolysis of the pyruvate imine,
this is unlikely because it is chemically analogous to the rapid
reaction observed in the burst phase, and thus decarboxylation
of the OAA imine is likely to be the rate-limiting step. The
biphasic time course is consistent with the conclusions of
Leussing and Raghavan,14 who observed that above pH 7.5
decarboxylation is much slower than imine formation when
assumption that all of the released CO2 is in the gas phase and
not in solution. Moreover, measuring the production of CO2 or
pyruvate only provides information on the overall rate of
decarboxylation. Measuring the absorbance at 280 nm is
complicated by the fact that OAA (enol tautomer) also absorbs
at 280 nm. Thus, the previous assays only allow for effective
measurement of the rate of decarboxylation and provide little
information on the rate of imine formation. Therefore, we
developed a novel approach for measuring the rates involved
in amine-catalyzed decarboxylation.
Instead of using OAA directly as the substrate for the reaction,
it was generated in situ via the malate dehydrogenase-catalyzed
oxidation of malate by NAD+ (Scheme 2). Upon incubating
the enzyme, NAD+, and malate, the reaction rapidly comes to
equilibrium in favor of malate (equilibrium constant is 2.86 ×
10-5 ) [NADH][OAA]/[NAD+][malate]).19 Reaction of an
amine with OAA to form a carbinolamine intermediate results
in oxidation of malate to form OAA to restore the equilibrium.
Thus, the rate of malate oxidation corresponds to the rate of
carbinolamine formation and is monitored by following the
increase in NADH which absorbs at 340 nm. Since ethylene-
diamine has been previously reported to catalyze the decar-
boxylation of OAA at near-neutral pH conditions,14 it was used
to characterize the assay. Different combinations of the reagents
of the assay were used to exclude the possibility of alternative
reactions and to show that the assay does, in fact, measure the
rate of OAA decarboxylation. The results of this study are
summarized in Table 1.
(20) The choice of pH was motivated by several factors. The studies of
Leussing and Raghavan14 found that decarboxylation was slower than imine
formation above pH 7.5. Therefore, pH 8.0 afforded the ability to observe the
fast consumption of oxaloacetate as well as the slower steps involving
decarboxylation. Moreover, measuring malate dehydrogenase activity at pH 8.0
using Tris buffer was well established, and from a practical standpoint malate
dehydrogenase is a homodimer which dissociates into monomers as the pH
decreases.21 Lastly, control experiments in HEPES buffer showed that while
Tris is an amine it did not react with oxaloacetate, whereas ethylenediamine
did.
A representative time course for the reaction measured using
this assay at pH 8.020 is shown in Figure 1. The time course is
biphasic with an initial rapid increase in absorbance followed
by a slower linear phase. The fast phase cannot be explained as
a rapid initial decarboxylation of the equilibrium concentration
of OAA because the OAA concentration is maintained by the
coupling enzyme, which is not rate-limiting at the concentrations
(21) Wood, D. C.; Jurgensen, S. R.; Geesin, J. C.; Harrison, J. H. J. Biol.
Chem. 1981, 256, 2377.
(22) Two other assays were used to confirm that pyruvate was indeed being
produced by the decarboxylation reaction. NMR analysis confirmed the product
was pyruvate, but the rates of decarboxylation could not be compared to the
assay used here because the analysis was performed in DMSO. In addition, lactate
dehydrogenase also detected the production of pyruvate. However, the very low
concentrations of oxaloacetate used in the malate dehydrogenase assay were
not feasible for the lactate dehydrogenase assay and thus the rates from the two
assays cannot be compared directly, which underscores the sensitivity as well
as the versatility of the malate dehydrogenase assay used in this study.
(19) Guynn, R. W.; Gelberg, H. J.; Veech, R. L. J. Biol. Chem. 1973, 248,
6957.
146 J. Org. Chem. Vol. 74, No. 1, 2009

Amine-Catalyzed Decarboxylation of Oxaloacetic Acid
the imine intermediate, which allows for the development and
maintenance of a steady-state concentration of the OAA-imine.
This results in saturation of the amine since the decarboxylation
of the OAA-imine must occur before the catalyst is released
to form the imine with a new molecule of OAA. Several factors
confirm that the saturation observed is not a consequence of
simply saturating malate dehydrogenase: First, the reaction rate
increases linearly with increasing amounts of the catalyst.
Second, when a variety of amines were tested as catalysts for
OAA decarboxylation, saturation curves were obtained despite
the fact that the lowest concentration of malate used was 1.0
mM. Since the KM value of malate for malate dehydrogenase is
1.0 mM,25 only the upper portion of the saturation curve would
have been observed at the malate concentrations used. Third,
when the reaction was run with three different concentrations
of malate dehydrogenase with malate held at a saturating
concentration of 200 mM the rate did not change, indicating
that the saturation phenomena were involved with the amine
catalysis of OAA.
FIGURE 2. Plot of ethylenediamine concentration versus initial
velocity of the decarboxylation step showing linear dependence of the
rate of decarboxylation on catalyst concentration. Reagent concentra-
tions: 2.0 M Tris-HCl, pH 8.0, 76 U/mL malate dehydrogenase, 20
mM malic acid, and 50 mM NAD+ with varying ethylenediamine
concentration.
Thus, the malate dehydrogenase assay is unique for measuring
amine-catalyzed OAA decarboxylation because it measures
starting material consumption which gives more information
because the point of measurement is prior to the rate-determining
step. The assay provides evidence to support the proposed
mechanism: the burst phase indicates initial OAA combination
with the amine to form the corresponding carbinolamine, and
the slow phase shows both the irreversible decarboxylation step
and the hydrolysis of the pyruvate imine to release an amine
that can then form a new OAA imine for another catalytic cycle.
Formation of this new OAA imine results in the production of
more NADH, which would not have been observed without both
the irreversible step and the regeneration of the amine. The fast
rate of carbinolamine formation can be monitored by presteady-
state methods while the rectangular hyperbola derived from the
rates of the slow phase can be fitted to the Michaelis-Menten
equation to yield a maximal velocity corresponding to the rate
of decarboxylation. Because the decarboxylation rates observed
in the steady-state analysis were obtained at saturating concen-
trations, they allow for direct comparison of the abilities of the
imine complexes to undergo decarboxylation. The presteady-
state rates of decarboxylation were not used in analysis of
decarboxylation, and can be found in Supporting Table 1
(Supporting Information). An experimental system is now in
place for determining the structural requirements for how amines
affect the rates of carbinolamine formation, carbinolamine
dehydration, and the decarboxylation step.
Structural Requirements for Amine-Catalyzed Decar-
boxylation of Oxaloacetic Acid. Table 2 shows the amines
tested along with their fast phase (carbinolamine formation) rates
measured from the pre-steady-state analysis. Table 3 contains
the same amines and the measured maximal slow-phase
(decarboxylation) velocities at steady state. For the pre-steady-
state analysis, the level of amine and malate was the same for
each amine, which allows for a direct comparison of the rates
of carbinolamine formation for each amine. Similarly, the results
of the steady-state study were useful in the comparison of the
ability of the imine complexes to decarboxylate because all rates
were measured at saturating concentrations.
FIGURE 3. Plot of OAA concentration versus initial velocity of the
decarboxylation step. The line represents the best fit of the data to the
Michaelis-Menten equation. Malic acid concentration was converted
to OAA concentration prior to regression using the equilibrium
expression for malate dehydrogenase: 2.86 × 10-5 ) [NADH][OAA]/
[NAD+][malate].
ethylenediamine is used as the catalytic amine.23 In addition,
Leussing and Raghavan’s pH profile of the rates of imine
formation suggested that, with ethylenediamine, the rate of
carbinolamine formation was faster than its subsequent dehydra-
tion at pH 8.0. Thus, the measured rates of carbinolamine
formation do not also correspond to the rate of imine formation
as long as carbinolamine formation is the fastest step.
The rate of the slow phase varied in a linear fashion with
increasing amounts of the amine (Figure 2) but exhibited
saturation kinetics when the concentration of malate was varied
(Figure 3). The Michaelis-Menten model for enzyme-catalyzed
reactions was used to derive the parameters KM and Vmax to
describe the saturation observed. It is important to note that the
KM values determined are only analogous to the corresponding
enzymatic parameter in that they represent the substrate
concentrations needed to achieve half-maximal velocity. The
Michaelis-Menten equation does not apply to this reaction
because, unlike enzyme-catalyzed reactions, where enzyme
concentration is much lower than substrate concentration, in
our assay, the concentration of the amine catalyst is much greater
than that of OAA. However, the reason amine-catalyzed
decarboxylation exhibits saturation kinetics at high catalyst
concentrations is due to the relatively low turnover rate of the
decarboxylation step with respect to the rate of formation of
Requirements for Carbinolamine Formation. The monoam-
ines aniline26 and aminoacetonitrile14 have been reported to
(23) Even at acidic pH values, the decarboxylation step for amine catalysis
of acetoacetate is still partially rate-limiting.24
(24) O’Leary, M. H.; Baughn, R. L. J. Am. Chem. Soc. 1972, 94, 626–630.
(25) Zimmerle, C. T.; Alter, G. M. Biochemistry 1993, 32, 12743.
J. Org. Chem. Vol. 74, No. 1, 2009 147

Thalji et al.
TABLE 2. Rates of Burst Phase
TABLE 4. Rates of Carbinolamine Formation with Monoaminesa
rate of burst phase ×
rate of carbinolamine formation ×
catalyst
10-3 (dA/min)
amine
10-2 (dA/min)
1,2-diaminopropane
ethylenediamine
1,3-diaminopropane
1,4-diaminobutane
1,5-diaminopentane
1,6-diaminohexane
1,7-diaminoheptane
trans-1,4-diaminocyclohexane
cis-1,2-diaminocyclohexane
trans-1,2-diaminocyclohexane
2-dimethylaminoethylamine
N,N′-dimethylethylenediamine
ethanolamine
23.7 ( 0.4
28.7 ( 0.2
41.2 ( 0.9
48.4 ( 0.3
52.7 ( 0.5
49.5 ( 0.4
49.4 ( 0.3
44.7 ( 1.1
38.1 ( 1.2
23.8 ( 0.5
30.9 ( 0.8
29.0 ( 0.8
19.7 ( 0.5
16.8 ( 0.1
44.8 ( 0.9
0
ethylamine
propylamine
butylamine
pentylamine
hexylamine
isopropylamine
tert-butylamine
cyclohexylamine
0
1.59 (0.05
2.26 (0.04
2.66 (0.09
2.57 (0.05
0.96 (0.14
1.26 (0.10
3.54 (0.05
a Reaction conditions: 2.0 M Tris-HCl, pH 8.0, 76 U/mL malate
dehydrogenase, 20 mM malic acid, 50 mM NAD+, and 545 mM amine.
The velocities are the average ( S.D. of three measurements.
cysteamine
diethylenetriamine
2-aminoethyltrimethylammonium chloride
phenylenediamine
aminoacetonitrile
distinct burst phase associated with carbinolamine formation,
we demonstrated that, at pH 8.0, the burst phase was noticeably
absent when aminoacetonitrile and aniline were used as catalysts.
The most likely explanation for this observation is that the
electron-withdrawing nitrile group of aminoacetonitrile renders
it an extremely poor nucleophile, while aniline is a poor
nucleophile because of the electron-withdrawing nature of the
aromatic ring, and thus, it is likely that an acid catalyst is needed
to promote carbinolamine formation. In fact, previous reports
of aniline-26 and aminoacetonitrile-catalyzed14 OAA decarboxy-
lation observed catalysis only under acidic conditions, which
is consistent with our findings that carbinolamine, and subse-
quent imine, formation (which is necessary for decarboxylation)
does not occur at pH 8.0, and which supports the conclusion
that acid catalysis is needed for imine formation with these
weakly nucleophilic amines.28
Ethylamine also failed to react with OAA to form a
carbinolamine, despite the fact that the ethylamine amino group
should be much more nucleophilic than that of aminoacetonitrile
or aniline due to the lack of the electron-withdrawing nitrile or
aromatic group. It appears that the absence of imine formation
with ethylamine can be attributed to solvation effects resulting
in a significant decrease in nucleophilicity that prevents carbino-
lamine formation. This was demonstrated by measuring the rate
of the burst phase with the monoamines propylamine, buty-
lamine, pentylamine, hexylamine, isopropylamine, cyclohexy-
lamine, and tert-butylamine. As the length of the alkyl chain
increases from ethylamine to hexylamine, solvation of the amino
group should decrease due to the presence of a longer
hydrophobic chain. As shown in Table 4, the rate of carbino-
lamine formation increases as solvation of the amino group
decreases from ethylamine to hexylamine, suggesting that the
nucleophilicity of the amino group is heavily dependent on the
degree of solvation of the nucleophilic amine.
0
0
0
aniline
Reaction conditions: 2.0 M Tris-HCl, pH 8.0, 76 U/mL malate
dehydrogenase, 20 mM malic acid, 50 mM NAD+, and 545 mM amine.
The velocities are the average ( S.D. of three measurements.
TABLE 3. Maximal Steady-State Decarboxylation Ratesa
decarboxylation rate ×
catalyst
1,2-diaminopropane
ethylenediamine
1,3-diaminopropane
1,4-diaminobutane
10-5 (min-1
)
336.2 ( 48.5
156.1 ( 17.7
16.7 ( 1.4
7.8 ( 0.7
5.6 ( 1.0
7.9 ( 1.5
2.4 ( 0.1
0
100.5 ( 15.3
127.0 ( 6.3
18.4 ( 2.6
0
6.9 ( 0.9
8.8 ( 0.8
154.9 ( 34.0
0
0
0
1,5-diaminopentane
1,6-diaminohexane
1,7-diaminoheptane
trans-1,4-diaminocyclohexane
cis-1,2-diaminocyclohexane
trans-1,2-diaminocyclohexane
2-dimethylaminoethylamine
N,N′-dimethylethylenediamine
ethanolamine
cysteamine
diethylenetriamine
2-aminoethyltrimethylammonium chloride
phenylenediamine
aminoacetonitrile
aniline
0
a Reaction conditions: 2.0 M Tris-HCl, pH 8.0, 76 U/mL malate
dehydrogenase, and 50 mM NAD+, with varying malate concentration
from 1 to 200 mM. Amines were generally used at 35 mM, except for
1,4-diaminobutane, 1,5-diaminopentane, and 1,7-diaminoheptane, which
were held at 200 mM, ethanolamine, which was held at 140 mM,
cysteamine, which was held at 50 mM, 1,6-diaminohexane, which was
held at 100 mM, and 2-dimethylaminoethylamine, which was held at 70
mM. The maximal velocities were determined by fitting a plot of OAA
concentration (as determined by the equilibrium constant for the malate
A similar solvation trend can be seen among diamines. The
rate of carbinolamine formation was found to increase signifi-
cantly as the size of the diamine increased, with ethylenediamine
exhibiting the slowest rate and with 1,5-diaminopentane, 1,6-
diaminohexane, and 1,7-diaminoheptane exhibiting much faster
rates. Furthermore, for diethylenetriamine, which was expected
to be more poorly solvated than ethylenediamine because of
the increased ratio of nonpolar methylene groups to polar amino
groups, the rate of the burst phase was faster than that of
ethylenediamine.
dehydrogenase
reaction)
versus
initial
velocities
to
the
Michaelis-Menten equation. Velocities were converted from dA340/min
to min-1 by dividing by the product of the extinction coefficient of
NADH and the amine concentration, which enabled direct comparison
of rates despite differences in amine concentration.
catalyze the decarboxylation of OAA under certain conditions.
Aminoacetonitrile has also been reported to be a very effective
catalyst for acetoacetate decarboxylation.27 Both of these
decarboxylations have been proposed to proceed through an
imine intermediate. Therefore, we investigated the ability of
these monoamines to form an imine with OAA at pH 8.0. Using
a stopped-flow spectrophotometer to measure the rate of the
(26) Hay, R. W. Aust. J. Chem. 1965, 18, 337.
(27) Guthrie, J. P.; Jordan, F. J. Am. Chem. Soc. 1972, 94, 9136.
(28) A helpful reviewer has pointed out that the lack of reactivity of aniline
and aminoacetonitrile could also be the consequence of unfavorably low
equilibrium constants for imine formation at pH 8.0.
148 J. Org. Chem. Vol. 74, No. 1, 2009

Amine-Catalyzed Decarboxylation of Oxaloacetic Acid
SCHEME 3. Intramolecular Assistance during
Carbinolamine Formation
ethylenediamine, N,N′-dimethylethylenediamine, and 2-dim-
ethylaminoethylamine react in their neutral, fully deprotonated
form, instead of in their monoprotonated forms. This explanation
is attractive because the neutral forms of these amines are
expected to be much more nucleophilic than the monoprotonated
forms. Since 2-aminoethyltrimethylammonium chloride has a
permanent positive charge, it cannot exist in a more nucleophilic
form, thus explaining its lack of reactivity.
Thus, it is clear that cooperativity between the amino group
and a second polar functional group is not a requirement for
carbinolamine formation, but if solvation is significant, as in
the case of ethylenediamine, cooperativity seems to be necessary
to stabilize the tetrahedral intermediate formed during carbino-
lamine formation, either by hydrogen bonding or else by proton
transfer. As solvation diminishes, the requirement for a second
polar group drops. In addition, most diamines probably react
in their monoprotonated forms, although it is possible that
ethylenediamine, which is a weaker base than the other
diamines, reacts in its more nucleophilic neutral form.
The Decarboxylation Step. Given a reasonable understand-
ing of the requirements for carbinolamine formation, a thorough
structural analysis of the rate of the decarboxylation step was
then performed to evaluate whether or not the second amino
group of diamines plays a significant role in the catalysis of
CO2 loss from OAA. The rate of decarboxylation of OAA by
diamines was found to be dependent on the length of the diamine
chain, with ethylenediamine exhibiting the fastest decarboxy-
lation rates, and longer diamines such as 1,6-diaminohexane
and 1,7-diaminoheptane catalyzing decarboxylation at lower
rates. In addition, both cis-1,2-diaminocyclohexane and trans-
1,2-diaminocyclohexane exhibited catalytic activity only slightly
lower than that of ethylenediamine. This is reasonable since
the proximity of the two amino groups of the cyclic diamines
should be comparable to that of ethylenediamine.
A burst phase was also observed with the cyclic diamines
trans-1,4-diaminocyclohexane, trans-1,2-diaminocyclohexane,
and cis-1,2-diaminocyclohexane. The rate of this burst was
fastest for trans-1,4-diaminocyclohexane and was comparable
to that of 1,4-diaminobutane. However, both of these rates were
slightly slower than that of the monoamine cyclohexylamine.
This can be rationalized by noting that both of these diamines,
in their monoprotonated forms, have an electron-withdrawing
ammonium group that decreases the nucleophilicity of the other
nitrogen. On the other hand, cyclohexylamine, which can only
react from its neutral form (even though the protonated form
predominates at equilibrium), is more nucleophilic and thus
forms the carbinolamine more rapidly. This is also consistent
with the lower rates observed for trans-1,2-diaminocyclohexane
and cis-1,2-diaminocyclohexane, which, when monoprotonated,
would be less nucleophilic than trans-1,4-diaminocyclohexane
because the electron-withdrawing ammonium group is closer
to the nucleophilic amino group.
It was rather surprising that ethylenediamine formed the
carbinolamine at all since it would be expected to be solvated
at least as well as ethylamine, if not better because of the
presence of the second polar amino group. The fact that it does
form the carbinolamine suggests the possibility that the second
amino group of ethylenediamine plays a role in carbinolamine
formation, either by hydrogen bonding to stabilize the oxyanion
generated by nucleophilic attack of the carbonyl as shown in
Scheme 3 or else by donating a proton to the oxyanion to
stabilize it. Further evidence for this effect is seen with
ethanolamine and cysteamine, both of which exhibit the burst
kinetics, albeit slowly. With ethanolamine, it is probable that
the hydroxyl group acts as a hydrogen bond donor to stabilize
the incipient oxyanion, whereas with cysteamine the reasonably
acidic thiol group can donate a proton to the oxyanion.
Three ethylenediamine analogues were assayed to help
elucidate the factors affecting carbinolamine formation. N,N′-
Dimethylethylenediamine and 2-dimethylaminoethylamine both
exhibited a burst phase at nearly the same rate as ethylenedi-
amine. For N,N′-dimethylethylenediamine, the amino groups are
secondary, and thus although the burst phase still corresponds
to the formation of a carbinolamine, the dehydration product is
an enamine. For 2-dimethylaminoethylamine, which is both a
primary and a tertiary amine, the burst can only represent
carbinolamine formation via reaction of the OAA with the
primary amino group because attack by the tertiary amine cannot
form a stable product. In contrast to the first two analogues,
2-aminoethyltrimethylammonium chloride, the trimethylated
quaternary ammonium salt of ethylenediamine, did not exhibit
burst kinetics. One possible explanation for this is that the
quaternary ammonium salt is reasonably well solvated and, thus,
like ethylamine, lacks a hydrogen-bonding or acidic group that
can enhance imine formation by stabilizing the high energy
oxyanionic intermediate. However, another possibility is that
The effect of increasing chain length on the rate of the
decarboxylation step had been noted previously,15 and Leussing
and Raghavan14 offered an explanation based on basicity of the
imine. They suggested that, with increasing basicity of the imine,
the unreactive enamine would begin to predominate, decreasing
the overall effectiveness of the diamine catalyst. Since basicity
of diamines increases with increasing separation of the two
amino groups, this is an enticing explanation for the significant
decrease in catalytic rate observed. However, the decarboxy-
lation rates of 1,4-diaminobutane and trans-1,4-diaminocyclo-
hexane were found to be significantly different, as trans-1,4-
diaminocyclohexane failed to catalyze decarboxylation. This is
surprising because both compounds possess similar electronic
properties, as evidenced by the fact that both are aliphatic
diamines with amino groups separated by four carbons. Fur-
thermore, the pKa values for the two molecules are nearly
identical (pK1 ) 9.24, pK2 ) 10.74 for trans-1,4-diaminocy-
clohexane and pK1 ) 9.40, pK2 ) 10.79 for 1,4-diaminobutane),
and the rates of carbinolamine formation for both catalysts are
virtually identical. Thus, according to the mechanism proposed
by Leussing and Raghavan, both compounds should catalyze
the decarboxylation step at the same rate (since the basicities
of the imine nitrogens formed by both catalysts should be
similar). Because this was not observed, this suggests that the
role of the second amino group in the decarboxylation step is
not simply to depress the pKa of the imine to inhibit tautomer-
ization to the enamine. Considering the lack of electronic
differences between the two catalysts, the most reasonable
J. Org. Chem. Vol. 74, No. 1, 2009 149

Thalji et al.
SCHEME 4. Catalysis by Intramolecular Stabilization of the Developing Negative Charge
alternative is that the observed discrepancy in catalytic ability
arises because the second amino group is directly involved in
the catalytic mechanism.29
SCHEME 5. Catalysis by Preferential Stabilization of the
Reactive Imine over the Unreactive Enamine
We suggest three possible roles for the second amino group
in the decarboxylation step. The first possibility (Scheme 4) is
that the second amino group, which will exist primarily in its
protonated form at pH 8.0,14 could function to stabilize the anion
formed on the imine nitrogen during decarboxylation by either
donating a proton to the imine nitrogen as it develops anionic
character or else by hydrogen-bonding to stabilize the anionic
nitrogen via the protonated second amino group, which is an
excellent hydrogen-bond donor. Since trans-1,4-diaminocyclo-
hexane is conformationally locked by the cyclohexane ring such
that the second amino group cannot come into close proximity
with the imine nitrogen to stabilize it, the second amino group
cannot participate in the decarboxylation step, which explains
its lack of reactivity. On the other hand, the flexible 1,4-
diaminobutane can orient the second amino group adjacent to
the imine nitrogen, resulting in rapid catalysis. Having the
second amino group as a proton donor in the decarboxylation
step also allows for a better understanding of why increased
amino group separation results in a decreased rate of decar-
boxylation. By increasing the length of the chain, there is a
greater entropic cost associated with bringing the second amino
group in close proximity to the imine nitrogen. This reasoning
also accounts for 1,2-diaminopropane having a higher maximal
velocity of decarboxylation than ethylenediamine because the
methyl group which constitutes the difference between ethyl-
enediamine and 1,2-diaminopropane causes the gauche orienta-
tion of the two amino groups to predominate. Since this
orientation in the OAA-imine places the second amino group
and the imine nitrogen in the closest proximity, the reaction
proceeds at a higher rate. This mechanism is similar to a
mechanism proposed by Ogino and associates to explain the
low catalytic activity of trans-oriented conformationally locked
cyclic diamines with respect to their cis counterparts.29
Leussing and Raghavan14 postulated that the diprotonated
OAA-imine (protonated at the second amino group and on the
labile carboxyl group) is the reactive species in the decarboxy-
lation step. If this is the case, no negative charge would develop
on the imine nitrogen because the carboxyl group would donate
its proton to the nitrogen as CO2 is evolved, and thus, the
mechanism suggested above would be invalid. An alternative
function of the terminal amino group in the decarboxylation
step could be to depress the pKa of the imine nitrogen by
hydrogen bonding as shown in Scheme 5. Lowering the basicity
of the imine nitrogen with respect to that of the R-carbon shifts
imine/enamine equilibrium toward the reactive imine tautomer,
which enhances catalysis.30
A final possible explanation for the trend in rates as diamine
chain length increases is that the rate-limiting step of the reaction
changes. Because the burst phase is evident at all times,
carbinolamine formation can never be the slowest step. How-
ever, there is a distinct possibility that the rate of carbinolamine
dehydration becomes the overall rate-limiting step as diamine
chain length increases. In this case, we suggest a role for the
second amino group that is quite similar to the role it might
play if the rate-limiting step were always decarboxylation.
Instead of hydrogen-bonding or protonating the imine nitrogen
as the negative charge develops, the second amino group may
protonate the hydroxyl group of the carbinolamine (Scheme 6),
making it a better leaving group and enhancing the rate of
(30) In addition to enamine tautomerization, other possible unreactive
intermediates include imidazolidines (from ring-closing with diamines), thiazo-
lidines (from cysteamine ring-closing), and oxazolidines (from ethanolamine ring-
closing). The formation of these intermediates would decrease catalytic rates in
much the same way as enamine tautomerization. However, imizadolidine
formation does not appear to be as significant as enamine tautomerization in
diamine-catalyzed decarboxylation because, otherwise, increasing diamine chain
length (which would decrease the stability of the corresponding unreactive cyclic
intermediates) would have resulted in increased rates of decarboxylation. On
the other hand, thiazolidine and oxazolidine ring-closing may be significant factors
in the low decarboxylation rates observed for cysteamine and ethanolamine.
(29) Ogino, K.; Tamiya, H.; Kimura, Y.; Azuma, H.; Tagaki, W. J. Chem.
Soc., Perkin Trans. 2 1996, 979.
150 J. Org. Chem. Vol. 74, No. 1, 2009

Amine-Catalyzed Decarboxylation of Oxaloacetic Acid
TABLE 5. Inhibition Studiesa
SCHEME 6. Catalysis of Carbinolamine Dehydration by
Intramolecular Protonation of Hydroxyl Group
inhibitor
initial velocity × 10-1 (dA340/min)
none
1.50 ( 0.05
1.53 ( 0.06
1.54 ( 0.06
1.55 ( 0.08
1.48 ( 0.03
100 mM methyl acetoacetate
200 mM methyl acetoacetate
100 mM acetone
200 mM acetone
a Reagent concentrations used: 76 U/mL malate dehydrogenase, 50
mM NAD+, 20 mM malic acid, 35 mM ethylenediamine, and 2.0 M
Tris-HCl, pH 8.0. The velocities are the average ( S.D. of three
measurements.
dehydration. As the separation between the two amino groups
increases, the ability of the second amino group to protonate
the hydroxyl group decreases, causing the dehydration step to
become rate-limiting.
possibility that diamines catalyze OAA decarboxyation by
simply stabilizing the oxyanion via hydrogen bonding. This was
achieved by assaying several compounds which cannot form
an imine (or an enamine) but which can act as hydrogen-bond
donors. If amino groups are hydrogen bonding to the carbonyl
oxygen of OAA, then diols analogous to the diamines might
also hydrogen bond and catalyze decarboxylation. Therefore,
1,2-propanediol (analogue of 1,2-diaminopropane), 1,3-pro-
panediol (analogue of 1,3-diaminopropane), cis-1,2-cyclohex-
anediol (analogue of cis-1,2-diaminocyclohexane), and cis-1,2-
pentanediol were examined as possible catalysts. The diols did
not exhibit any catalytic activity when compared to ethylene-
diamine (data not shown). However, the hydrogens in the
oxyanion holes of enzymes are usually provided by amides,34
not hydroxyl groups, and therefore, urea and N,N′-diacetyleth-
ylenediamine (bis-acetylated ethylenediamine) were also tested
as catalysts of OAA decarboxylation. It was found that these
amides also did not catalyze OAA decarboxylation. It is
important to note that Munakata et al.15 reported that urea did
catalyze OAA decarboxylation. However, catalysis was only
observed under acidic conditions (pH 5.0) instead of at pH 8.0,
and the reported rate constant for urea-catalyzed decarboxylation
was more than 10-fold lower than that of ethylenediamine. As
a final test to determine if the amino groups are hydrogen
bonding to the carbonyl oxygen of OAA, acetone and methyl
acetoacetate, both of which contain carbonyls that could compete
with OAA for hydrogen bonding to the amine, were tested as
inhibitors of the ethylenediamine-catalyzed reaction. As can be
seen in Table 5, neither acetone nor methyl acetoacetate
inhibited catalysis by ethylenediamine. Thus, diamine-catalyzed
decarboxylation of OAA does not occur by stabilization of the
oxyanion through hydrogen bonding of the amino groups.
As expected, N,N′-dimethylethylenediamine did not exhibit
decarboxylation since it only forms an unreactive enamine with
OAA. Given the fact that imine formation is necessary for
decarboxylation, it becomes clear why phenylenediamine,
aniline, aminoacetonitrile, ethylamine, and 2-aminoethyltrim-
ethylammonium chloride do not exhibit significant rates of
decarboxylation since none of these molecules form the imine.
On the other hand, 2-dimethylaminoethylamine, which can form
an imine with OAA, does catalyze decarboxylation. However,
it does not catalyze the reaction nearly as rapidly as ethylene-
diamine. It is likely that two effects contribute to this decrease:
First, the dimethyl substitution of one of the amino groups causes
it to possess greater steric bulk which hinders the approximation
of the amino group and the imine. Second, the protonated tertiary
amino group, which is the only form of the amino group that
can hydrogen-bond or serve as a proton donor in the reaction,
would possess only one N-H bond capable of interacting with
the imine nitrogen, as opposed to the three N-H bonds in
protonated ethylenediamine. Thus, there is an entropic cost
associated with orienting the lone N-H bond in a reactive
conformation.
The data for ethanolamine and cysteamine strongly suggest
that hydrogen bonding, not proton transfer, is the mechanism
by which the second amino group (or other polar group) assists
in the decarboxylation step. Both ethanolamine and cysteamine
exhibit very low levels of decarboxylation. Cysteamine (thiol
pKa ) 8.631) is more acidic than monoprotonated ethylenedi-
amine (ammonium pKa ) 9.57), but it decarboxylates OAA very
slowly. In addition, an ammonium group (as in monoprotonated
ethylenediamine) is a much better hydrogen bond donor than a
neutral alcohol or thiol.32 Catalysis by ethanolamine and
cysteamine, which are analogs of the amino acids serine and
cysteine, respectively, explains why some amino acids catalyze
the decarboxylation of OAA.33
Conclusions
In this report, a new assay for amine-catalyzed decarboxy-
lation of OAA is described. The new assay is a significant
improvement over previous assays in that it allows simultaneous
examination of the rates of carbinolamine formation and
decarboxylation. It also allows kinetic data to be obtained using
very low OAA concentrations, providing a more complete
description of the kinetic profile. Structural analysis of the rate
of carbinolamine formation suggests that imine formation at pH
8.0 is inhibited if the nucleophilic amino group is strongly
solvated, but that this inhibition can be overcome by the presence
of a second group capable of either intramolecularly hydrogen-
bonding with or protonating the oxyanion of the tetrahedral
intermediate of carbinolamine formation to stabilize it. Further-
more, analysis of decarboxylation rates shows that diamine-
catalyzed decarboxylation requires the approximation of the two
amino groups to achieve high rates of decarboxylation. Thus,
Oxyanion Hole Mimics. Some enzyme-catalyzed decarboxy-
lation and transcarboxylation reactions occur via active-site
structures known as oxyanion holes, which consist of two
hydrogen bond donors (usually peptidic N-H bonds of small
amino acid residues) which stabilize the oxyanion formed in
transition states during these reactions.34 While the burst phase
of the kinetics observed with the malate dehydrogenase assay
strongly suggests imine formation, it was worth ruling out the
(31) Svensson, B. E. Biochem. J. 1988, 253, 441.
(32) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural
Chemistry and Biology; Oxford University Press: Oxford, 1999; pp 12-16.
(33) Bessman, S. P.; Layne, E. C. Arch. Biochem. 1950, 26, 25.
(34) Dodson, G.; Wlodawer, A. Trends Biochem. Sci. 1998, 23, 347–352.
J. Org. Chem. Vol. 74, No. 1, 2009 151

Thalji et al.
flow spectrophotometer. One of the drive syringes contained 0.816
M Tris-HCl pH 8.0 and 1.09 M of the amine, while the other
drive syringe contained 3.184 M Tris-HCl, pH 8.0, 152 U/mL,
malate dehydrogenase, 40 mM malic acid, and 100 mM NAD+.
Equal volumes of these solutions were mixed to give the final
concentrations: 2.0 M Tris-HCl, pH 8.0, 76 U/mL malate dehy-
drogenase, 20 mM malic acid, 50 mM NAD+, and 545 mM of the
amine. Data were acquired at 340 nm for up to 20 s. The
temperature was maintained at 22 °C with a circulating water bath.
Between injections, the flow cell was washed with dH2O. The fast
and slow phases were fitted separately by linear least-squares
analysis to determine the velocity as ∆A/min. The reported velocity
is the average of three separate injections (measurements).
Determination of pKa Values. Approximately 1.0 g of amine
was dissolved in approximately 150 mL of dH2O and titrated against
1.0 M HCl until the final end point was reached. pKa’s were
determined as the pH halfway between end points. For 2-amino-
ethyltrimethylammonium chloride hydrochloride, 0.5 g of amine
was dissolved in approximately 150 mL of dH2O and the mixture
titrated against 1.0 M NaOH until the end point was reached. pKa’s
were determined as the pH halfway between the beginning of the
titration and the end point.
the role of the second amino group in efficient diamine-catalyzed
decarboxylation of OAA is 3-fold: First, it enables well-solvated
molecules such as ethylenediamine to form carbinolamines (and
thus imines), and second, it assists in decarboxylation by either
stabilizing the negative charge that develops on the imine
nitrogen during the course of the decarboxylation, or else by
preferentially stabilizing the imine tautomer over the enamine,
thereby increasing the concentration of the reactive species and
enhancing overall decarboxylation rates. Finally, the second
amino group may serve to catalyze the carbinolamine dehydra-
tion step by protonating the hydroxyl group to enhance its
leaving group ability.
Experimental Section
Steady-State Kinetic Assay for Decarboxylation of OAA.
Amine-catalyzed decarboxylation of OAA was measured using a
malate dehydrogenase-coupled enzyme assay. Each reaction con-
tained 76 U/mL malate dehydrogenase, 50 mM NAD+, 20 mM
malic acid, and 2.0 M Tris-HCl pH 8.0. These components were
combined and allowed to reach equilibrium (which occurred within
seconds). The reaction was started by the addition of the amine
and followed by measuring the increase in absorbance at 340 nm
associated with production of NADH. Data were collected using a
Uvikon 810 (Kontron Instruments) spectrophotometer interfaced
to a PC equipped with a data acquisition program. The temperature
was maintained at 25 °C by a circulating water bath with the
capacity to heat and cool the thermospace of the cell compartment.
The initial velocities are reported as dA340/min. When malate
concentration was varied, saturation kinetics were observed and
the parameters KM and Vmax were determined by fitting the data to
the Michaelis-Menten equation using the nonlinear regression
programs of Cleland.16
Acknowledgment. The authors thank Dr. Reema Thalji and
Dr. W.W. Cleland for their expert advice in the preparation of
the manuscript.
Supporting Information Available: Table of the decar-
boxylation rates from the pre-steady-state study and KM values
as obtained from the steady-state analysis for most amines. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Rapid Reaction Kinetic Assay for Decarboxylation of OAA.
Rapid reaction kinetics were measured with a SX.18MV-R stopped-
JO8014648
152 J. Org. Chem. Vol. 74, No. 1, 2009