Metal ion inhibition of nonenzymatic pyridoxal phosphate catalyzed decarboxylation and transamination

Article abstract of DOI:10.1021/ja0026354

Nonenzymatic pyridoxal phosphate (PLP) catalyzed decarboxylations and transaminations have been revisited experimentally. Metal ions are known to catalyze a variety of PLP-dependent reactions in solution, including transamination. It is demonstrated here that the rate accelerations previously observed are due solely to enhancement of Schiff base formation under subsaturating conditions. A variety of metal ions were tested for their effects on the reactivity of the 2-methyl-2-aminomalonate Schiff bases. All were found to have either no effect or a small inhibitory one. The effects of Al³⁺ were studied in detail with the Schiff bases of 2-methyl-2-aminomalonate, 2-aminoisobutyrate, alanine, and ethylamine. The decarboxylation of 2-methyl-2-aminomalonate is unaffected by metalation with Al³⁺, while the decarboxylation of 2-aminoisobutyrate is inhibited 125-fold. The transamination reaction of ethylamine is 75-fold slower than that of alanine. Ethylamine transamination is inhibited 4-fold by Al³⁺ metalation, while alanine transamination is inhibited only 1.3-fold. Metal ion inhibition of Schiff base reactivity suggests a simple explanation for the lack of known PLP dependent enzymes that make direct mechanistic use of metal ions. A comparison of enzyme catalyzed, PLP catalyzed, and uncatalyzed reactions shows that PLP dependent decarboxylases are among the best known biological rate enhancers: decarboxylation occurs 10¹⁸-fold faster on the enzyme surface than it does free in solution. PLP itself provides the lion's share of the catalytic efficiency of the holoenzyme: at pH 8, free PLP catalyzes 2-aminoisobutyrate decarboxylation by ～10¹⁰-fold, with the enzyme contributing an additional ～10⁸-fold.

Full text of DOI:10.1021/ja0026354

VOLUME 123, NUMBER 2
J ANUARY 17, 2001
© Copyright 2001 by the
American Chemical Society
Metal Ion Inhibition of Nonenzymatic Pyridoxal Phosphate Catalyzed
Decarboxylation and Transamination
Roger F. Zabinski and Michael D. Toney*
Contribution from the Department of Chemistry, UniVersity of CaliforniasDaVis, One Shields AVenue,
DaVis, California 95616
ReceiVed July 17, 2000
Abstract: Nonenzymatic pyridoxal phosphate (PLP) catalyzed decarboxylations and transaminations have been
revisited experimentally. Metal ions are known to catalyze a variety of PLP-dependent reactions in solution,
including transamination. It is demonstrated here that the rate accelerations previously observed are due solely
to enhancement of Schiff base formation under subsaturating conditions. A variety of metal ions were tested
for their effects on the reactivity of the 2-methyl-2-aminomalonate Schiff bases. All were found to have either
no effect or a small inhibitory one. The effects of Al3+ were studied in detail with the Schiff bases of 2-methyl-
2-aminomalonate, 2-aminoisobutyrate, alanine, and ethylamine. The decarboxylation of 2-methyl-2-aminoma-
lonate is unaffected by metalation with Al3+, while the decarboxylation of 2-aminoisobutyrate is inhibited
125-fold. The transamination reaction of ethylamine is 75-fold slower than that of alanine. Ethylamine
transamination is inhibited 4-fold by Al3+ metalation, while alanine transamination is inhibited only 1.3-fold.
Metal ion inhibition of Schiff base reactivity suggests a simple explanation for the lack of known PLP dependent
enzymes that make direct mechanistic use of metal ions. A comparison of enzyme catalyzed, PLP catalyzed,
and uncatalyzed reactions shows that PLP dependent decarboxylases are among the best known biological
rate enhancers: decarboxylation occurs 1018-fold faster on the enzyme surface than it does free in solution.
PLP itself provides the lion’s share of the catalytic efficiency of the holoenzyme: at pH 8, free PLP catalyzes
2-aminoisobutyrate decarboxylation by ∼1010-fold, with the enzyme contributing an additional ∼108-fold.
Pyridoxal phosphate (PLP1) plays a central role in the amino
acid and amine metabolism of all organisms. It is covalently
bound at the active site of a very large number of enzymes which
effect a variety of types of chemical transformations including
decarboxylation, transamination, racemization, R-â and â-γ
eliminations, and retro-aldol cleavages. The early work of Snell’s
2
group2-6 on the pure coenzyme and its analogues provided a
mechanistic foundation for understanding the enzymatic activi-
ties.
* To whom correspondence should be addressed. E-mail: toney@
chem.ucdavis.edu. Phone: (530) 754-5282. Fax: (530) 752-8995.
(1) Abbreviations used: PLP, pyridoxal phosphate; PMP, pyridoxamine
phosphate; MAM, 2-methyl-2-aminomalonate; AIB, R-aminoisobutyrate;
TEA, triethanolamine; CHES, 2-[N-cyclohexylamino]ethanesulfonic acid;
NMR, nuclear magnetic resonance spectroscopy. The following metal salts
at millimolar concentrations had no effect on the rate of MAM decarboxy-
lation: CaCl2, MgCl2, CsCl, CdCl2, MnCl2, NiCl2, CoCl2, Cu(OAc)2, Pb-
(NO3)2, Cr(ClO4)3, Yt(NO3)3, and TlCl3. The decarboxylation rate constant
decreased linearly with concentration for the following metal salts: Sm-
(OAc)3, ErCl3, EuCl3, and La(NO3)3.
(2) Metzler, D. E.; Snell, E. E. J. Biol. Chem. 1952b, 198, 353-361.
(3) Snell, E. E. Physiol. ReV. 1953, 33, 509-524.
(4) Metzler, D. E.; Ikawa, M.; Snell, E. E. J. Am. Chem. Soc. 1954, 76,
648-652.
(5) Ikawa, M.; Snell, E. E. J. Am. Chem. Soc. 1954, 76, 653-655.
(6) Snell, E. E. In Transaminases;Christen, P., Metzler, D. E., Eds.;
Wiley: New York, 1984; pp 19-35.
(7) Metzler, D. E.; Snell, E. E. J. Am. Chem. Soc. 1952, 74, 979-983.
(8) Longernecker, J. B.; Snell, E. E. J. Am. Chem. Soc. 1957, 79, 142-
145.
10.1021/ja0026354 CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/19/2000

194 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zabinski and Toney
Software Associates). 1H NMR spectra were collected on a Bruker 300
MHz spectrometer.
It was observed in these early studies that metal ions promote
many of the nonenzymatic reactions.7,8 A wide variety of di-
and trivalent metal ions are active. The Al3+ ion is a well-studied
example and is a particularly appropriate comparison to the
proton because of its small ionic radius. The early studies cited
above were semiquantitative at best in their analyses. No attempt
was made to determine the relative magnitudes of the effects
of the metal ions on promotion of Schiff base formation and
on the reactivities of the metal ion bound Schiff bases.
The assumption in the literature has been that the multiple
charge of the Schiff base bound metal ion is more effective
than the unitary charge of a proton at stabilizing the carbanion
formed on heterolytic cleavage of a bond to CR of the amino
acid.6 Extensive studies from Bruice’s laboratory (ref 9 and
references therein) demonstrating and characterizing reactions
in the absence of metal ions showed their ancillary nature in
nonenzymatic systems. The absence in the literature of reports
of PLP dependent enzymes that make direct mechanistic use
of metal ions additionally suggests that they provide no
fundamental advantage to catalysis. This paper describes experi-
ments in which the reactivities of four Schiff bases of amino
acids and amines are examined in the presence and absence of
Al3+. It is shown that, in all cases, the protonated Schiff base
is at least as or more reactive in decarboxylation or transami-
nation compared to the metalated Schiff base.
Dissociation Constant Measurements. Schiff base dissociation
constants were measured by variation of amino acid or amine
concentration in a series of solutions of constant volume and PLP
concentration. Spectra (250-600 nm) were collected and analyzed
globally by SPECFIT. The reactions were allowed to proceed to >5
half-lives of the slowest kinetic phase before spectra were recorded.
The reactivity of MAM precluded equilibrium measurements, and the
apparent dissociation constant reported in Table 1 was obtained from
the hyperbolic dependence of the decarboxylation rate constant on
MAM concentration.
Dissociation constants for Al3+ binding to the Schiff bases at pH 5
were measured using the pH jump described above. Equilibrium spectra
(250-600 nm) were obtained after >5 half-lives of the slowest kinetic
phase, and were analyzed by SPECFIT to determine the values of Kd.
As with Schiff base formation, the reactivity of MAM required the
use of kinetic data to obtain an apparent Kd for Al3+. Rate constants
for the first phase of the reaction of Al3+ with MAM-PLP in pH jump
experiments (0.2 M MAM) showed a hyperbolic dependence on Al3+
concentration and were used to obtain the values reported in Table 1.
Rate Constant Measurements. Schiff base formation rate constants
reported in Table 2 were measured in reactions containing 0.2 M
potassium acetate, 0.2 M amino acid, 0.1 mM PLP, and 0.2 M
potassium chloride at 25 °C. Spectra (250-500 nm) were collected
versus time and analyzed globally using SPECFIT.
Metal ion binding rate constants were measured using a pH jump
from 8 to 5. This allows saturation of PLP with amino acid at high pH
followed by rapid metal ion binding and slow Schiff base hydrolysis
at low pH. Reaction conditions were identical to those for Schiff base
formation except various concentrations of Al3+ were included in the
potassium acetate buffer. The Al3+ binding rate constants reported in
Table 2 are maximal values obtained from hyperbolic fits of observed
rate constants versus metal ion concentration.
MAM decarboxylation in the presence of Zn2+ or Ga3+ was measured
using the pH jump method with 0.2 M MAM and 0.2 M potassium
acetate, pH 5 and 4.5, respectively. The Ga3+ reactions could only be
followed at pH 4.5 or lower due to precipitation of gallium.
Alanine and ethylamine transamination and AIB decarboxylation
were conducted at pH 5 using 1 M amino acid or amine in the presence
or absence of 15 mM Al3+. No pH jump was used since full
equilibration of Schiff base formation and metal ion binding occurs on
the time scale of transamination and decarboxylation. The AIB and
ethylamine reactions were extremely slow and were followed for only
∼10% of the reaction. The rate constants were obtained by fitting
exponential curves in which the amplitude of the slow kinetic phase
was fixed at the change in absorbance expected for complete PLP
reaction.
The rate constants reported herein for decarboxylation of AIB
and transamination of alanine and ethylamine in the absence of
metal ions allow the calculation of the catalytic ratios for the
nonenzymatic and enzymatic PLP dependent reactions. These
calculations provide a measure of the catalytic prowess of PLP
dependent enzymes relative to others that have been character-
ized.
Experimental Procedures
Materials. Methylaminomalonic acid was synthesized as previously
described10,11 and stored at -20 °C. D,L-Alanine and PLP were
purchased from Sigma. TEA and zinc acetate were obtained from
Fisher. Ethylamine was obtained as a 70% aqueous solution from
Sigma. Aluminum chloride was from Mallinckrodt. Gallium chloride
and AIB were purchased from Aldrich. All references to metal ion
concentrations refer to the total metal ion concentration. MAM and
PLP solutions were prepared immediately before use, and the PLP
solutions were kept in the dark to prevent decomposition.
Reaction Conditions. Reactions contained 0.1 mM PLP, 0.2 M KCl,
and 0.2 M potassium acetate (pH 5.0), 0.2 M TEA (pH 8.0), or 0.2 M
CHES (pH 9.0). They were maintained at 25 °C in the dark in quartz
cuvettes, which were covered with Teflon caps. Reactions that took
longer than 2 h to complete were additionally sealed with high-vacuum
grease to prevent evaporation. Controls with PLP in the absence of
amine were maintained identically to the reaction mixtures. These
controls demonstrated that evaporation over the long reaction times is
insignificant and that PLP is not degraded under the experimental
conditions.
Product Analysis. Product analysis was performed by quenching
reaction aliquots with sufficient potassium hydroxide to neutralize
reactants and buffer and give a final concentration of 0.5 M KOH.
The base quench rapidly hydrolyzes any Schiff base present. Absor-
bance spectra were measured to determine the amount of PLP or PMP
by comparison to the measured extinction coefficients of the pure
compounds under similar conditions.
1H NMR (D2O, 0.5 M KOD) was used in some cases to validate the
UV-vis product analysis. For PLP, 1H NMR (0.5 M KOD, D2O at 4.8
pH Jump. Schiff bases were preformed at pH 8 in 10 mM TEA-
HCl buffer, followed by addition of the metal ion in 0.2 M potassium
acetate buffer at low pH to bring the solution to the desired pH. This
pH jump was done to maximize Schiff base formation using high pH
and to use low pH to maximize the rate of metal ion binding and metal
ion solubility.
1
ppm): 10.4 (C4′-H), 7.64 (C6-H), and 2.34 (C2′-H3). For PMP, H
NMR (0.5M KOD, D2O at 4.8 ppm): 3.76 (C4′-H2), 7.57 (C6-H),
and 2.32 (C2′-H3).
Results
Instruments. Absorbance spectra were collected from 250 to 600
nm with a Hewlett-Packard 8453 diode-array spectrophotometer. The
absorbance data were analyzed globally using SPECFIT (Spectrum
Schiff Base and Metal Ion Dissociation Constants. Table
1 summarizes the dissociation constants for Schiff base forma-
tion between amino acids or amines and PLP, as well as for
the binding of Al3+ to these Schiff bases. The stability of the
Schiff bases decreases with decreasing pH as previously
observed.12 On the other hand, the apparent affinity of Al3+ for
the AIB Schiff base increases with decreasing pH, in agreement
(9) Auld, D. S.; Bruice, T. C. J. Am. Chem. Soc. 1967, 89, 2098-2106.
(10) Sun, S.; Zabinski, R. F.; Toney, M. D. Biochemistry 1998, 37, 3865-
3875.
(11) Bailey, G. B.; Chotamangsa, O.; Vuttivej, K. Biochemistry 1970,
9, 3243-3248.

PLP Catalyzed Decarboxylation and Transamination
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 195
Table 1. Dissociation Constants for Schiff Base Formation and
hand, the data for MAM show that Schiff base hydrolysis is
only 2-fold slower than metal ion binding. The goal was to
measure the rate constants for decarboxylation of PLP fully
bound by amino acid and metal ion. Therefore, higher concen-
trations (1.8 M MAM, 1 M AIB) were used in the experiments
in which metal ion dependent decarboxylation rate constants
were measured (Table 3) to ensure nearly complete complex-
ation of PLP as the metal ion bound MAM Schiff base at low
pH.
The binding of metal ions (Al3+, Zn2+, or Ga3+) to the Schiff
bases exhibited multiple kinetic phases between pH 5 and pH
9, suggesting the fast formation of an encounter complex
followed by slow isomerization. Figure 1 presents data obtained
in the reaction of the AIB Schiff base with Al3+ at pH 5. The
kinetic traces given in Figure 1A clearly show the two-
exponential nature of the reaction. The spectra of the three
species involved in the two-exponential process are shown in
Figure 1B. The Schiff base (species A) has a typical absorbance
maximum of ∼420 nm, while the intermediate metalated Schiff
base has a major peak at ∼310 nm, and the final isomerized
metalated Schiff base absorbs maximally at ∼370 nm, in
agreement with previous findings.13
Al3+ Bindinga
reactants
Kd (mM)
A
B
pH 5
pH 9
26(2)
AIB
PLP
3000(100)
650(100)b
1700(40)
310(10)
MAM
PLP
PLP
PLP
Alanine
Ethylamine
Al3+
1.2(0.03)
0.34(0.01)
0.44(0.3)b
AIB-PLP
0.023(0.007)
2(1)b
Al3+
Al3+
Al3+
MAM-PLP
Ala-PLP
EA-PLP
2.3(2)b
0.034(0.009)
2.0(0.1)
a Errors are given in parentheses. Reaction conditions: 25 °C in the
dark, 0.1 mM PLP, 0.2 M KCl, variable concentration of reactant A,
and either 0.2 M potassium acetate (pH 5) or 0.1 M CHES (pH 9).
Titrations with Al3+ at pH 9 contained 1 M AIB, 1 M ethylamine, or
0.2 M MAM. Titrations with Al3+ at pH 5 were performed via pH
jumps from pH 8 with 0.2 M amino acids or amine. b Kinetically
determined values. The apparent Kd for MAM was determined from
hyperbolic fits of observed decarboxylation rate constants versus
concentration (Figure 3). Al3+ Kd values were obtained from hyperbolic
fits of the rate constants for the first metal ion binding kinetic phase
versus concentration. The Al3+/MAM data at pH 5 were obtained in
pH jump experiments in which Schiff base formation was maximized
by preforming it at pH 8.
The binding of Ga3+ (5 mM) to the AIB Schiff base exhibited
three kinetic phases: k1met ) 6.0(0.4) × 10-2 s-1, k2met ) 4.5-
(0.1) × 10-3 s-1, k3met ) 1.5(0.1) × 10-3 s-1. Two Ga3+ binding
Table 2. Rate Constants for Schiff Base Formation and Al3+
Bindinga
met
phases were distinguished from MAM decarboxylation: k1
reactants
met
met
kSB × 102
k1 × 102
k2 × 102
met
) 3.0(0.1) × 10-2 s-1, k2 ) 1.8(0.1) × 10-3 s-1
.
A
B
(s-1
)
(s-1
)
(s-1
)
The first phase in Zn2+ binding to either the AIB or MAM
Schiff base is complete in the mixing time (5-10 s). The rate
constant for the slower phase for Zn2+ (5 mM) binding to the
AIB Schiff base was 4.0(0.1) × 10-3 s-1, while that for binding
AIB
PLP
PLP
AIB-PLP
MAM-PLP
1.4 (0.1)
2.2 (0.1)
MAM
Al3+
Al3+
7.9 (0.6)
3.9 (0.8)
0.10 (0.01)
0.4 (0.1)
to the MAM Schiff base was 2.3(0.1) × 10-2 s-1
.
a Reaction conditions: 25 °C in the dark, 0.1 mM PLP, 0.2 M amino
acid, 0.2 M KCl, 0.2 M potassium acetate, pH 5. Metal ion binding
kinetics were measured using a pH jump method that involved
preforming the Schiff base at pH 8, followed by addition of potassium
acetate and various concentrations of Al3+ to give pH 5. The metalation
rate constants are kmax values obtained from plots of kobs versus Al3+
concentration. The kmax value for AIB Schiff base formation at pH 5 is
Decarboxylation and Transamination Rate Constants.
Figure 2 presents data for the reaction of MAM with PLP in
the presence of Al3+. Three methods of analysis (continuous
spectrophotometric monitoring, base quench followed by NMR,
and base quench followed by spectrophotometric coenzyme
determination) yield rate constants for the decrease of PLP or
the increase of PMP that are in good agreement. Decarboxy-
lation thus occurs as the slowest kinetic process in reactions
containing metal ion and the most reactive amino acid. The data
in Figure 3 show the effect of 15 mM (saturating) Al3+ on the
rate constant for decarboxylation of the MAM Schiff base. The
inclusion of the metal ion has essentially no effect on the rate
of MAM decarboxylation.
0.05 ( 0.01 s-1
.
with previous studies.13 The kinetically determined apparent Kd
values for Al3+ binding to the MAM Schiff base were obtained
from the first kinetic process, and thus probably represent
dissociation constants of initial encounter complexes that
subsequently isomerize to the final forms. In agreement with
this proposition, the kinetically determined apparent Kd at pH
5 for Al3+ binding to the AIB Schiff base is much larger than
the equilibrium value and is, within error, identical to that for
MAM. The Kd for Al3+ binding to the ethylamine Schiff base
is ∼100-fold larger than that for the AIB Schiff base, in accord
with previous observations.14 This is undoubtedly due to the
absence of the carboxylate ligand to the metal ion in the Schiff
base.
Schiff Base Formation and Metal Ion Binding Rate
Constants. The data in Table 2 show that, at concentrations of
AIB (0.2 M) high enough to saturate PLP at pH 8 (the pH at
which Schiff base formation is allowed to occur before jumping
the pH to 5 in reactions containing metal ion), metal ion binding
at pH 5 is fast relative to Schiff base hydrolysis. On the other
Spectrophotometric data for the decarboxylation of AIB and
the transamination of ethylamine are shown in Figure 4. Data
obtained from base-quenched aliquots are in agreement with
those obtained directly for alanine, AIB, and ethylamine. The
inclusion of Al3+ in the reactions greatly slows AIB decarboxy-
lation (125-fold) and slows ethylamine transamination to a lesser
extent (4-fold). The transamination of alanine is slightly inhibited
by Al3+ (1.3-fold). The rate of MAM decarboxylation decreased
slightly with increasing concentrations of Ga3+, Zn2+, and
several other metal ions examined.1
Table 3 collects the rate constants for decarboxylation and
transamination in the presence and absence of Al3+. The rate
constants for the metal ion free reactions have been corrected
for subsaturating amounts of amino acid or amine on the basis
of the dissociation constants reported in Table 1. The reported
values are thus for the Schiff bases. Previous reports demon-
strated that dilute solutions of PLP analogues in strongly
buffered solutions, as here, yield pyruvate quantitatively from
(12) Felty, W. L.; Ekstrom, C. G.; Leussing, D. L. J. Am. Chem. Soc.
1970, 92, 3006-3011.
(13) Davis, L.; Roddy, F.; Metzler, D. E. J. Am. Chem. Soc. 1960, 83,
127-134.
(14) Weng, S. H.; Leussing, D. L. J. Am. Chem. Soc. 1983, 105, 4082-
4090.

196 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zabinski and Toney
Table 3. Rate Constants for Decarboxylation and Transamination in the Presence and Absence of Al3+ a
k × 106 (s-1
)
k × 106 (s-1
)
pH 5
pH 5
pH 8
Decarboxylation
MAM-PLPc
AIB-PLPb
3.9(0.3)
0.031(0.003)
10000(1000)
9800(800)
460(20)
450(10)
AIB-PLP + Al3+
MAM-PLP + Al3+ c
Transamination
alanine-PLPb
70(1)
54(1)
ethylamine-PLPb
ethylamine-PLP + Al3+
0.93(0.01)
0.25(0.04)
alanine-PLP + Al3+
a Reaction conditions: 25 °C in the dark, 0.1 mM PLP, 0.2 M KCl, 0.2 M potassium acetate (pH 5) or 0.2 M TEA (pH 8), and 15 mM Al3+ for
the metal ion containing reactions. Errors are given in parentheses. b Rate constants for the reactions of AIB, alanine, and ethylamine (each present
c
at 1 M) without Al3+ have been adjusted to 100% PLP saturation on the basis of the dissociation constants given in Table 1. kmax values from
hyperbolic fits of observed rate constants versus either MAM or Al3+ concentration. In the presence of Al3+, MAM was present at 1.8 M at pH 5
and 0.2 M at pH 8. Data at pH 5 were measured using a pH jump from 8 to 5 to maximize Schiff base formation.
Figure 2. Time course for absorbance changes at 325 nm (PMP
formation) and 392 nm (PLP loss) and of PMP formation detected by
NMR and PLP loss detected by base quenching with spectrophotometric
quantification for MAM decarboxylation in the presence of 4 mM Al3+
at pH 5.
Figure 1. (A, top) Time courses for the reaction of the preformed
AIB-PLP Schiff base with 15 mM Al3+ at pH 5. Biphasic kinetics
are clear from the data. (B, bottom) Spectra of the species observed in
the Al3+ binding reaction: A, Schiff base (λmax ) 414 nm), B,
intermediate metalated Schiff base (λmax ) 374 nm); C, final Al3+
Schiff base complex (λmax ) 368 nm).
-
Figure 3. Observed rate constants for MAM decarboxylation at pH 5
in the presence (solid circles) and absence (open circles) of 19 mM
Al3+
.
alanine via transamination9 and pyruvate and alanine from the
decarboxylation of MAM.15
pyridine-4-aldehyde catalyzed transamination of alanine from
which one can calculate an expected value of 7 × 10-5 s-1 in
half-neutralized 0.2 M acetate buffer. The transamination of
γ-aminobutyrate, mechanistically similar to that with ethylamine,
was previously reported to be much slower than that of alanine,16
consistent with the 75-fold difference in rate constants reported
in Table 3.
The rate constants reported in Table 3 for both MAM
decarboxylation and alanine transamination catalyzed by PLP
are in excellent agreement with those previously measured under
similar conditions. Thanassi15 reported a value of ∼4 × 10-3
s-1 for the 5-deoxypyridoxal catalyzed decarboxylation of
MAM. Auld and Bruice9 reported rate constants for 3-hydroxy-
(16) Rossi, C. S.; Olivo, F.; Rabassini, A.; Siliprandi, N. Arch. Biochem.
Biophys. 1962, 96, 650-652.
(15) Thanassi, J. W. Biochemistry 1972, 11, 2909-2914.

PLP Catalyzed Decarboxylation and Transamination
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 197
Thus, loss of CO2 from the final PLP-MAM:Al3+ complex is
the rate-determining step in MAM decarboxylation. By exten-
sion, CO2 loss or deprotonation of the PLP-amino acid:Al3+
complexes of AIB, alanine, and ethylamine are assumed to be
rate-determining since these substrates are much less reactive.
In support of this assumption, no absorption bands above 450
nm were observed in any of these reactions.
Al3+ Inhibits Decarboxylation and Transamination of
Metalated Schiff Bases. Metal ion catalysis of transamination7,8
has been reported, while Kalyankar and Snell22 reported a ∼2-
fold reduction in AIB decarboxylation by metal ions. Metal ions
were reported to inhibit slightly â-elimination reactions.23 The
rate constants for the reactions of metal ion free Schiff bases
and their 1:1 Al3+ chelates reported in Table 3 give a direct
measure of the effect of Al3+ solely on the central chemical
step (i.e., CO2 loss or deprotonation), unlike the previous studies
cited above.
The data indicate that protonated Schiff bases are generally
more reactive than Al3+-bound Schiff bases. Ethylamine reacts
4-fold faster and alanine reacts 1.3-fold faster in the absence of
Al3+ than in its presence. Decarboxylation of the AIB Schiff
base shows a larger (125-fold) effect due to the chelation of
the reactive carboxylate group by the metal ion. Carboxylate
chelation would promote deprotonation (i.e., transamination)
stereoelectronically by orienting the CR-H bond more nearly
perpendicular to the pyridine ring plane, but inhibit CO2 loss
both stereoelectronically and electrostatically. The MAM Schiff
base, with two carboxylate groups, would incur a combination
of favorable stereoelectronic effects and unfavorable electrostatic
effects for CO2 loss, which is apparent in the lower rate constant
ratio of 1 in the absence and presence of Al3+, as compared to
125 for AIB.
Figure 4. Time courses for the decarboxylation of AIB and transami-
nation of ethylamine in the absence and presence of 16 mM Al3+
.
Discussion
Quinonoid Intermediate Stability. Several reports have
described the observation of dihydropyridine tautomers of Schiff
bases (quinonoid intermediates) between amino acids and PLP
analogues by their characteristic long-wavelength (∼500 nm)
absorbance. Schirch and Slotter17 and Abbott and Bobrik18
reported observing in water a quinonoid intermediate from
diethylaminomalonate and pyridoxal, while Maley and Bruice19
found a similar species in water during the reaction of 1-methyl-
4-formylpyridine with glycine, alanine, and glycine ethyl ester.
This species has been observed in reactions of the ethyl esters
of alanine in methanol in the presence of Al3+ by Matsumoto
and Matsushima.20
The previously observed rate enhancements due to the
presence of metal ions can now safely be ascribed simply to
the enhancement of Schiff base formation by the metal ions in
solutions containing subsaturating concentrations of amino acids.
The inhibitory effect of metal ions on decarboxylation and
transamination of the Schiff bases examined here provides a
viable explanation for the lack of observation of PLP dependent
enzymes that make direct mechanistic use of metal ions in the
central chemical steps.
Abbott and Martell21 examined the reaction of the 1:1 Al3+
-
bound alanine Schiff bases by NMR. These authors proposed
that a stable quinonoid intermediate is formed in the course of
the reaction due to the strong electrostatic effect of the trivalent
metal ion as compared to the proton. The possibility that the
decarboxylation of MAM in the presence of Al3+, which is much
faster than the transamination of alanine, produces a stable
quinonoid intermediate was therefore considered.
Spectra collected in the course of the PLP-MAM:Al3+
reaction have no detectable absorbance band at wavelengths
longer than 450 nm (indicative of a quinonoid intermediate),
as was observed in the references cited above. Formation of a
stable quinonoid intermediate after rapid CO2 loss equates to
protonation of the quinonoid intermediate being the rate-
determining step in the conversion of the PLP-MAM:Al3+
chelate to either the pyruvate ketimine or the alanine aldimine
(which corresponds to protonation on C4′ of the coenzyme or
CR of the amino acid, respectively). If a stable quinonoid that
does not absorb above 450 nm is very rapidly formed and
protonation of it is rate-determining, then PLP loss and PMP
formation as determined by the quench methods would cor-
respond to one of the early processes observed in the spectral
data obtained directly from the unquenched reaction. This is
not the case; rate constants from quench data coincide with those
obtained spectroscopically from the slowest kinetic process.
Pyridine Nitrogen Protonation. The pKa of the pyridine
nitrogen of Schiff bases formed between PLP and R-amino acids
is in the range 5.9-6.7,24 while that for metalated Schiff bases
is ∼7.4.12 The MAM data in Table 3 at pH values of 5 and 8
thus provide a rough comparison of the reactivities of the
pyridine nitrogen protonated and unprotonated Schiff bases. For
both the metal ion free and Al3+-bound forms, the protonation
of the pyridine nitrogen has only a ∼20-fold effect, which is
small compared to the total rate acceleration due to PLP in
nonenzymatic reactions (see below).
A similar 20-fold effect was observed by Maley and Bruice25
in the comparative transamination of alanine by 3-hydroxypy-
ridine-4-aldehyde vs 1-methyl-3-hydroxypyridine-4-aldehyde,
where quaternization in the latter enforces a positive charge on
the nitrogen. In their studies of the transamination of alanine
by 3-hydroxypyridine-4-aldehyde, Auld and Bruice9 made the
(22) Kalyankar, G. D.; Snell, E. E. Biochemistry 1962, 1, 594-600.
(23) Tatsumoto, K.; Martell, A. E. J. Am. Chem. Soc. 1981, 103, 6203-
6208.
(24) Kallen, R. G.; Korpela, T.; Martell, A. E.; Matsushima, Y.; Metzler,
C. M.; Metzler, D. E.; Morozov, Y. V.; Ralston, I. M.; Savin, F. A.;
Torchinsky, Y. M.; Ueno, H. In Transaminases; Christen, P., Metzler, D.
E., Eds.; Wiley: New York, 1984.
(17) Schirch, L.; Slotter, R. A. Biochemistry 1966, 5, 3175.
(18) Abbott, E. H.; Bobrick, M. A. Biochemistry 1973, 12, 846-851.
(19) Maley, J. R.; Bruice, T. C. J. Am. Chem. Soc. 1968, 90, 2843-
2847.
(20) Matsumoto, S.; Matsushima, Y. J. Am. Chem. Soc. 1974, 96, 5228-
5232.
(21) Abbott, E. H.; Martell, A. E. J. Am. Chem. Soc. 1973, 95, 5014-
5019.
(25) Maley, J. R.; Bruice, T. C. Arch. Biochem. Biophys. 1970, 136,
187-192.

198 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zabinski and Toney
interesting observation that protonation of the R-carboxyl group
has an effect equal to protonation of the pyridine nitrogen,
amounting to a ∼75-fold rate enhancement. Thus, the strict
conservation in a very large number of PLP dependent enzymes
of an arginine residue that interacts intimately with the substrate
R-carboxyl group26 may reflect not simply binding interactions,
but also mechanistic requirements. A quantitatively coincidental
factor of 75-fold is observed here for the increase in Schiff base
reactivity toward transamination due to the presence of an
unprotonated R-carboxylate function (alanine vs ethylamine;
Table 3).
The relatively small effect of pyridine nitrogen protonation
in the nonenzymatic reactions suggests that, as has been found
in computational studies,27,28 the protonated azaallylic function
(Schiff base) may be the dominant factor in stabilization of
carbanions generated on the substrate carbon bearing the
nitrogen atom of the Schiff base. In this regard, it is significant
that a very wide variety of carbonyl compounds,29 including
formaldehyde,30 can function as catalysts of amino acid decar-
boxylation. The existence of the pyruvoyl-dependent decar-
boxylases also highlights the ability of a Schiff base derived
from a simple carbonyl function to stabilize well an azaallylic
carbanion.
constants for a variety of reactions that are catalyzed by enzymes
and showed that OMP decarboxylase provides an enormous rate
enhancement of ∼1017-fold. The rate constant for either alanine
or glycine spontaneous decarboxylation at neutral pH and 25
°C, obtained by extrapolation from high temperatures, is ∼2 ×
10-17 s-1 and rather insensitive to pH in the neutral range
(Snider and Wolfenden, J. Am. Chem. Soc., in press). One can
calculate from this and the data in Table 3 that, in the absence
of enzyme, PLP catalyzes the decarboxylation of AIB by ∼1011-
fold at pH 5 and by ∼1010-fold at pH 8. This impressive rate
acceleration by PLP alone is equal to or better than that of many
enzymes.32
Thanassi33 reported a rate constant of ∼10-7 s-1 for the
spontaneous decarboxylation of aminomalonate at pH 5. Com-
pared to the value of 10-2 s-1 for the PLP catalyzed decar-
boxylation of 2-methylaminomalonate at the same pH, this
represents only a ∼105-fold rate acceleration by PLP for this
intrinsically much more decarboxylation reactive amino acid,
in accord with expectation.
The total decarboxylation rate enhancement at pH 8 due to
the enzyme-PLP complex (i.e., holoenzyme) is ∼1018-fold.
Thus, PLP dependent decarboxylases and OMP decarboxylase
are currently the best known biological rate enhancers, as
determined from first-order rate constant ratios. The ∼108-fold
component of this due to the protein portion of the holoenzyme
is in the middle of the known range of coenzyme independent
enzymes;32 in other words, the protein portion does a rather
average job at enhancing the decarboxylation rate constant. The
enormous rate acceleration of the holoenzyme is only possible
due to the high efficiency of the PLP coenzyme itself, assuming
its efficacy in the enzyme active site is at least equal to that in
solution.
Comparison of Enzymatic and Nonenzymatic Rates. The
decarboxylation of AIB by the PLP dependent enzyme dialkyl-
glycine decarboxylase occurs with a maximal rate constant of
25 s-1 at pH 8 and 25 °C.31 This value is ∼108-fold greater
than the PLP dependent nonenzymatic reaction would be,
assuming a similar 20-fold lower decarboxylation rate constant
at pH 8 as found with MAM. The same enzyme catalyzes the
transamination of alanine with a maximal rate constant of 22
s-1 31
,
giving a ∼107-fold increase over the PLP dependent
nonenzymatic rate. The transamination of γ-aminobutyrate by
the E. coli isozyme of γ-aminobutyrate aminotransferase occurs
with a kcat value of ∼50 s-1 (J. Langston and M. D. Toney,
unpublished results), giving a ∼109-fold increase over the PLP
dependent nonenzymatic rate.
An additional and possibly equally important role of the
protein in PLP dependent enzymes is to direct the reaction of
the external aldimine intermediate (i.e., the enzyme-bound PLP-
substrate Schiff base) along a specific reaction pathway, since
this intermediate has multiple reaction pathways available to
it. Dunathan’s34 elegant proposal of stereoelectronic control of
reaction specificity combined with specific placement of
auxiliary catalytic functional groups, mechanisms not available
to the nonenzymatic reactions, likely accounts for the observed
high specificities of PLP dependent enzymes.
Radzicka and Wolfenden32 measured nonenzymatic rate
(26) Mehta, P. K.; Hale, T. I.; Christen, P. Eur. J. Biochem. 1993, 214,
549-561.
(27) Bach, R. D.; Canepa, C. J. Am. Chem. Soc. 1997, 119, 11725-
11733.
(28) Bach, R. D.; Canepa, C.; Glukhovtsev, M. N. J. Am. Chem. Soc.
1999, 121, 6542-6555.
(29) Ardill, H.; Grigg, R.; Sridharan, V.; Surendrakumar, S. Tetrahedron
1988, 44, 4953-4966.
JA0026354
(30) Joucla, M.; Mortier J. Chem. Soc., Chem. Commun. 1985, 1566-
1567.
(32) Radzicka, A.; Wolfenden, R. Science 1995, 267, 90-93.
(33) Thanassi, J. W. Biochemistry 1970, 9, 525-532.
(34) Dunathan, H. C. Proc. Natl. Acad. Sci. U.S.A. 1966, 55, 712-716.
(31) Sun, S.; Bagdassarian, C. K.; Toney, M. D. Biochemistry 1998, 37,
3876-3885.