Insights into the mechanism of pseudomonas dacunhae aspartate β-decarboxylase from rapid-scanning stopped-flow kinetics

Article abstract of DOI:10.1021/bi100272g

The mechanism of wild-type and R37A mutant Pseudomonas dacunhae aspartate β-decarboxylase (ABDC) was studied by rapid-scanning stopped-flow spectrophotometry. Mixing wild-type ABDC with 50 mM disodium l-Asp resulted in the formation of a 325 nm absorption peak within the dead time of the stopped-flow instrument, likely the ketimine of pyridoxamine 5′-phosphate and oxaloacetate or pyruvate. After consumption of the l-Asp, the 360 nm feature of the resting enzyme was restored. Thus, the 325 nm species is a catalytically competent intermediate. In contrast, mixing wild-type ABDC with the disodium salt of either threo- or erythro-β-hydroxy-dl-Asp at 50 mM resulted in a much slower formation of the 325 nm complex, with an apparent rate constant of ～1 or 0.006 s^-1, respectively. When wild-type ABDC is mixed with disodium succinate, a nonreactive analogue of l-Asp, formation of a new peak at 425 nm is observed. The apparent rate constant for formation of the 425 nm band exhibits a hyperbolic dependence on succinate concentration, showing that there is a rapid binding equilibrium, followed by a slower reaction in which the internal aldimine is protonated on the Schiff base N. Hydrostatic pressure shifts the spectrum from the 425 nm form to the 360 nm form, consistent with a conformational change. It is likely that the binding of substrate or analogues induces a conformational change that releases strain in the Lys pyridoxal 5′-phosphate Schiff base and increases the pK_a, resulting in protonation of the Schiff base to initiate transaldimination. Mixing of R37A mutant ABDC with 50 mM l-Asp also results in the formation of the 325 nm complex, but with an apparent rate constant of 0.2 s^-1, at least 5000-fold slower than the rate of wild-type ABDC. In contrast to wild-type ABDC, R37A ABDC shows no change in the cofactor spectrum when mixed with disodium succinate. These results suggest that Arg-37, a conserved active site residue in ABDC, plays a role in modulating the pK_a of the pyridoxal 5′-phosphate complexes during catalysis.

Full text of DOI:10.1021/bi100272g

5
066 Biochemistry 2010, 49, 5066–5073
DOI: 10.1021/bi100272g
Insights into the Mechanism of Pseudomonas dacunhae Aspartate β-Decarboxylase from
Rapid-Scanning Stopped-Flow Kinetics
,
‡,§
§
‡
‡
Robert S. Phillips,* Santiago Lima, Roman Khristoforov, and Bakthavatsalam Sudararaju
‡
§
Department of Chemistry, University of Georgia, Athens, Georgia 30602, and Department of Biochemistry and Molecular Biology,
University of Georgia, Athens, Georgia 30602
Received February 23, 2010; Revised Manuscript Received May 14, 2010
ABSTRACT: The mechanism of wild-type and R37A mutant Pseudomonas dacunhae aspartate β-decarboxylase
(ABDC) was studied by rapid-scanning stopped-flow spectrophotometry. Mixing wild-type ABDC with
5
0 mM disodium
L
-Asp resulted in the formation of a 325 nm absorption peak within the dead time of the
0
stopped-flow instrument, likely the ketimine of pyridoxamine 5 -phosphate and oxaloacetate or pyruvate.
After consumption of the -Asp, the 360 nm feature of the resting enzyme was restored. Thus, the 325 nm
L
species is a catalytically competent intermediate. In contrast, mixing wild-type ABDC with the disodium salt
of either threo- or erythro-β-hydroxy-DL-Asp at 50 mM resulted in a much slower formation of the 325 nm
-
1
complex, with an apparent rate constant of ∼1 or 0.006 s , respectively. When wild-type ABDC is mixed with
disodium succinate, a nonreactive analogue of -Asp, formation of a new peak at 425 nm is observed. The
L
apparent rate constant for formation of the 425 nm band exhibits a hyperbolic dependence on succinate
concentration, showing that there is a rapid binding equilibrium, followed by a slower reaction in which the
internal aldimine is protonated on the Schiff base N. Hydrostatic pressure shifts the spectrum from the 425 nm
form to the 360 nm form, consistent with a conformational change. It is likely that the binding of substrate or
0
analogues induces a conformational change that releases strain in the Lys pyridoxal 5 -phosphate Schiff base
and increases the pK , resulting in protonation of the Schiff base to initiate transaldimination. Mixing of
a
R37A mutant ABDC with 50 mM -Asp also results in the formation of the 325 nm complex, but with an
L
-
1
apparent rate constant of 0.2 s , at least 5000-fold slower than the rate of wild-type ABDC. In contrast to
wild-type ABDC, R37A ABDC shows no change in the cofactor spectrum when mixed with disodium
succinate. These results suggest that Arg-37, a conserved active site residue in ABDC, plays a role in
0
modulating the pK of the pyridoxal 5 -phosphate complexes during catalysis.
a
1
0
Aspartate β-decarboxylase (ABDC) is a pyridoxal 5 -phos-
PLP-dependent aminotransferase family, most of which form
dimers or tetramers. However, despite these quaternary structur-
al differences, the active site of ABDC is remarkably similar to
that of aspartate aminotransferase (AAT) (4). One notable
difference between the structure of ABDC and AAT is the
presence of an additional arginine residue, Arg-37, in the active
site, which is conserved in all known ABDC sequences. We pre-
viously prepared the mutant R37A ABDC and found that it
retains significant activity, apparently ruling out a role for Arg-37
as an essential active site acid-base catalyst (4). We have now
examined the pre-steady state kinetics of the reactions of wild-
phate (PLP)-dependent enzyme found mainly in bacteria (1),
although it has also been reported in mammalian tissues (2).
ABDC catalyzes the irreversible β-decarboxylation of
to give -alanine and carbon dioxide (eq 1)
L-aspartate
L
In bacteria, this reaction is coupled with an Asp/Ala
antiporter in the cell membrane, resulting in a proton
gradient and net ATP synthesis (3). The mechanism of
ABDC is intriguing, since it is the only known PLP-
dependent enzyme that catalyzes β-decarboxylation,
rather than R-decarboxylation.
type and R37A mutant ABDC with -Asp, erythro- and threo-β-
L
hydroxy-DL-Asp, and succinate. These results suggest that Arg-37
plays an important role in modulating the ionization state of the
PLP cofactor during catalysis.
Recently, the three-dimensional structures of Pseudomonas
dacunhae and Alcaligenes faecalis ABDC were determined by
X-ray crystallography (4, 5). The enzyme exists as a dodecame-
ric quaternary structure unique among the members of the
MATERIALS AND METHODS
Enzymes. Recombinant wild-type and R37A ABDC from
P. dacunhae were prepared as previously described (4). Enzyme
assays were performed using a coupled assay with alanine
dehydrogenase as previously described (4), measuring the increase
in absorbance at 340 nm (Δε = 6200 M cm ) as NAD is con-
verted to NADH.
*To whom correspondence should be addressed: Department of
-
1
-1
þ
Chemistry, University of Georgia, Athens, GA 30602. Phone: (706)
5
42-1996. Fax: (706) 542-9454. E-mail: rsphillips@chem.uga.edu.
1
Abbreviations: ABDC, aspartate β-decarboxylase; AAT, aspartate
0
0
Materials. L-Aspartate was a product of Sigma-Aldrich
aminotransferase; PLP, pyridoxal 5 -phosphate; PMP, pyridoxamine 5 -
phosphate; β-hydroxyAsp, β-hydroxy-DL-aspartic acid.
Chemical Co. Succinic acid was obtained from Fisher Scientific.
pubs.acs.org/Biochemistry
Published on Web 05/14/2010
r 2010 American Chemical Society

Article
Biochemistry, Vol. 49, No. 24, 2010 5067
Threo- and erythro-β-hydroxy-DL-Asp (6) were generous gifts of
C. H. Stammer. Other common chemicals were obtained from
Fisher Scientific.
is the pressure-independent value of the apparent equilibrium
constant, and ΔV is the reaction volume.
o
Instruments. Stopped-flow kinetics experiments were per-
formed on an RSM-1000 instrument from OLIS, Inc. (Bogart,
GA), equipped with a stopped-flow cell compartment. This
instrument has a dead time of ca. 2 ms and can collect
UV-visible scans at speeds up to 1000 Hz. The stopped-flow
experiments were performed at ambient temperature, generally
RESULTS
Reaction of Wild-Type ABDC with L-Aspartate. When
wild-type ABDC is mixed with 50 mM disodium L-Asp in the
stopped-flow spectrophotometer, there is a rapid change in the
spectrum within the dead time of the instrument, ca. 2 ms, from
the 360 nm form of the resting enzyme to a spectrum with peaks
at 325 and 360 nm and a small shoulder at 425 nm (spectrum 1,
Figure 1A). As the Asp substrate is consumed, in ∼1 min, the
spectrum gradually reverts to one similar to that of the resting
form, with a λmax of 360 nm (Figure 1A,B).
Reaction of Wild-Type ABDC with threo- and erythro-
β-Hydroxy-DL-aspartate. Mixing wild-type ABDC with 50 mM
disodium threo-β-hydroxy-DL-Asp results in a slow reaction, with
formation of a spectrum with a peak at 325 nm, a shoulder at
22-25 ꢀC. The enzyme assays were performed on a Cary 1 UV-
visible spectrophotometer equipped with a Peltier-controlled six-
cell changer. The effects of hydrostatic pressure on the absorption
spectra of the ABDC-succinate complex were measured using
a Cary 14 UV-vis spectrophotometer modified by OLIS, Inc., to
contain a high-pressure cell from ISS (Champaign, IL), equipped
with a manual pressure pump, and maintained at 25 ꢀC with an
external circulating water bath.
Stopped-Flow Experiments. The stopped-flow experiments
were performed in 0.1 M sodium acetate buffer (pH 6.0). The
enzyme was exchanged into the reaction buffer by rapid gel
filtration on a PD-10 column (GE Corp.) equilibrated with 0.1 M
sodium acetate (pH 6.0) immediately before use. Solutions of
wild-type or R37A mutant ABDC (2-3.2 mg/mL) were mixed
3
60 nm, and a very weak shoulder at 430 nm (Figure 2A) that
resembles the spectrum formed with -Asp. There is an isosbestic
point in the spectra at 335 nm. In contrast to the reaction with
-Asp, these spectroscopic changes take place over 3 s (Fig-
L
L
-1
ure 2B), with a rate constant of ∼1 s . Since the reaction with
Asp occurs within the dead time of the stopped-flow instrument
e2 ms), the corresponding reaction with -Asp must occur with
L-
with disodium salts of
L
-Asp, erythro- or threo-β-hydroxyAsp
(
L
(
0.1 M), or various concentrations of disodium succinate (0.02-
-1 2
a rate constant greater than 1000 s . Thus, the reaction of threo-
β-hydroxy-DL-Asp with ABDC is at least 1000-fold slower than
0.3 M). We prepared the disodium salts by dissolving the acids
in 2 equiv of 1 M NaOH before dilution with acetate buffer.
Scans were collected from 250 to 800 nm for various periods
of time and scanning rates after mixing, depending on the rate
of the individual reaction monitored. The stopped-flow data
were analyzed using Global Works (7) provided by OLIS, Inc.
The data were fitted to models of either one or two exponentials
as necessary to obtain reasonable fits. The concentration depen-
dence for reactions of succinate was fitted to eq 2
that with
L
-Asp. The reaction with 50 mM disodium erythro-β-
hydroxy-DL-Asp produces a spectrum similar to those of
and threo-β-hydroxy-DL-Asp, with an absorbance peak at 325 nm,
a shoulder at 360 nm, and a very weak shoulder at 430 nm, albeit
L-Asp
much more slowly, over a period of ∼3 min, with a rate constant of
-1
∼
0.006 s (Figure 2C,D). This shows that the reaction of erythro-
β-hydroxy-DL-Asp with ABDC is at least 16700-fold slower than
2
that with
35 nm.
Reaction of Wild-Type ABDC with Succinate. When
L-Asp. These spectra also exhibit an isosbestic point at
3
K
obs ¼ ðk
f
½succinateꢀÞ=ðKeq þ ½succinateꢀÞ þ k
r
ð2Þ
where Keq is the dissociation constant, k is the rate constant for
wild-type ABDC is mixed with disodium succinate, there is a
rapid decrease in the 360 nm absorbance peak and an increase in
an absorbance peak at 425 nm (Figure 3A,B), with an isosbestic
point at 386 nm. The apparent rate constant for the reaction is
dependent on the concentration of succinate, over the range from
10 to 150 mM, with a hyperbolic dependency (Figure 4). Fitting
f
the forward reaction, and k is the rate constant for the reverse
r
reaction (8).
Hydrostatic Pressure Experiments. The enzyme solution
þ
contained 1 mg/mL ABDC in 0.1 M MES-Na (pH 5.3)
containing 0.1 M disodium succinate in 1.2 mL quartz bottles
with a 9 mm path length, capped with Teflon tubing and
immersed in spectroscopic-grade ethanol as the pressurizing
fluid. MES buffer was used rather than acetate since buffers
-1
these data to eq 2 gives the following values: k = 71.6 ( 1.2 s
,
f
-
1
kr = 37.3 ( 1.5 s , and Keq = 21.9 ( 1.7 mM.
When the ABDC-succinate complex is subjected to increasing
hydrostatic pressure, the spectrum shifts, with a decrease in the
425 nm absorbance and an increase in the 360 nm absorbance,
and with an isosbestic point at 386 nm (Figure 5A), so there are
only two absorbing species contributing to the spectra. The
absorbance change at 425 nm with pressure is shown in Figure
6B, with the empty circles indicating the data obtained during
the compression and the filled circles the corresponding data
obtained during decompression. These data show that the
changes in the spectrum with pressure are completely reversible,
indicating that there is a simple reversible equilibrium between
the two states. Fitting the 425 nm data in Figure 5B to eq 3 gives
with anionic conjugate bases have a high ΔpK with pressure (9).
a
þ
A buffer blank containing 0.1 M MES-Na (pH 5.3) and 0.1 M
disodium succinate at 100 bar was used to obtain a baseline
reading. The pressure was increased or decreased in 100 bar
increments to a maximum of 1900 bar, and the enzyme solutions
were scanned immediately after each increase or decrease in
pressure. The total time to measure the spectra through the range
of increasing and decreasing pressures was approximately 2 h.
Absorbance values at 425 nm were corrected for solvent com-
pressibility using a modified Tait equation (10). The corrected
absorbance values were analyzed by fitting to eq 3
AP ¼ ðA - A Þ½K ꢁ expð - PΔV =RTÞꢀ=½1 þ K
¥
o
eq
o
eq
2
The reaction of ABDC with L-Asp is over within the dead time of the
ꢁ
expð - PΔV
o
=RTÞꢀ þ A
¥
ð3Þ
stopped-flow instrument, which is ∼2 ms. Assuming that the reaction is
not seen because at least three half-lives have passed in the dead time, the
half-life would then be less than 0.67 ms. The lower limit on the rate
constant is then given by 0.693/6.7 ꢁ 10 s = 1034 s
where AP is the observed absorbance at pressure P, A is the
o
-4
-1
absorbance at 0 bar, A is the absorbance at infinite pressure, K
.
¥
eq

5
068 Biochemistry, Vol. 49, No. 24, 2010
Phillips et al.
F
3
IGURE 1: Reaction of wild-type ABDC with 50 mM disodium L-Asp in 0.1 M sodium acetate (pH 6.0). (A) Scans are shown at 0.002, 2.69, 8.9,
7.1, and 60.5 s. The dotted line (spectrum 0) is the spectrum of the enzyme before the reaction. (B) Time courses at 360 and 425 nm.
F
0
IGURE 2: (A) Reaction of wild-type ABDC with 50 mM disodium threo-β-hydroxy-DL-Asp in 0.1 M sodium acetate (pH 6.0). Scans are shown at
.0087, 0.249, 0.601, 1.19, 2.38, and 5.31 s. The dotted line (spectrum 0) is the spectrum of the enzyme before the reaction. (B) Time courses for the
reaction of threo-β-hydroxy-DL-Asp at 360 and 425 nm. (C) Reaction of wild-type ABDC with 50 mM disodium erythro-β-hydroxy-DL-Asp in 0.1
M sodiumacetate (pH6.0). Scans are shown at0.411, 7.19, 35.9, 71.8, 107.7, and 179.7 s. The dottedline(spectrum0) isthe spectrumofthe enzyme
before the reaction. (D) Time courses for the reaction of erythro-β-hydroxy-DL-Asp at 360 and 425 nm.
the following values: Keq = 0.0169 ( 0.0072, and ΔV = -50 (
β-decarboxylation reaction, and there are only a few PLP-depen-
dent enzymes that catalyze similar reactions involving β-γ bond
cleavage with resulting electrophilic rather than nucleophilic
substitution at the β-carbon, including kynureninase (11), cy-
steine desulfurase (12), and selenocysteine lyase (13). It has been
suggested that these enzymes utilize PMP-ketimine intermediates
to polarize the β-γ bond and provide an electron sink for the
reaction forming a formal β-carbanion (11-13). PMP-ketimine
intermediates are also commonly found in aminotransferase
reactions and generally exhibit absorption peaks at ∼325 nm
arising from the PMP (14). Novogrodsky and Meister reported
that a 325 nm complex is formed in the steady state from ABCD
20 mL/mol.
Reaction of R37A ABDC with L-Asp and Succinate.
Mixing R37A ABDC with 50 mM disodium L-Asp results in a
slow decrease in the magnitude of the 360 nm absorption peak,
concomitant with an increase at 330 nm (Figure 6A,B). The
absorbance changes take ∼10 s to complete, with an apparent rate
-
1
constant of ∼0.2 s . Thus, the reaction of R37A ABDC is ∼5000-
2
fold slower than that with wild-type ABDC. In contrast to wild-
type ABDC, there is no significant shoulder at 425 nm in the
spectra. Mixing R37A ABDC with 100 mM disodium succinate
results in no significant change in the spectrum (data not shown).
and
transamination resulting in PMP and oxaloacetate (15). In our
work, we find that mixing ABDC with -Asp results in the
formation of a 325 nm intermediate within the dead time of the
L-Asp, but they assumed that it was the result of the abortive
DISCUSSION
ABDC catalyzes the β-decarboxylation of
L
-Asp to give
L-Ala
L
and CO (1). This reaction is the only known PLP-dependent
2

Article
Biochemistry, Vol. 49, No. 24, 2010 5069
F
IGURE 3: Reaction of wild-type ABDC with 100 mM disodium succinate in 0.1 M sodium acetate (pH 6.0). (A) Scans are shown at 0.0055,
0
3
.0105, 0.0175, 0.0265, 0.0405, and 0.0995 s. The dotted line (spectrum 0) is the spectrum of the enzyme before the reaction. (B) Time courses at
60 and 425 nm.
gel filtration, and it decays slowly to release
L-Ser (17). Thus, the
3
25 nm complex is also catalytically competent for the reactions
of β-hydroxyAsp. In our studies, we also found that the 325 nm
species is formed from the β-hydroxyAsp diastereomers, albeit
much more slowly than from
hydroxyAsp reacts 167-fold faster than the erythro isomer
Figure 2) to form the 325 nm complex. This is consistent with
L-Asp (Figure 2). However, threo-β-
(
the much weaker binding of the erythro diastereomer reported by
Miles and Meister (17). A proposed mechanism for the reaction
of the β-hydroxyAsp diastereomers with ABDC is shown in
Scheme 1. Formation of an external aldimine is followed by
deprotonation to a quinonoid intermediate, which is then repro-
tonated to give a β-hydroxyoxaloacetate ketimine of PMP.
Decarboxylation would give an enolamine, which is reprotonated
on the β-carbon to give a β-hydroxypyruvate ketimine. Depro-
0
tonation of this ketimine at C-4 gives the serine quinonoid
F
type ABDC with disodium succinate. The curve is calculated from
IGURE 4: Dependence of the rate constant for the reaction of wild-
intermediate which is reprotonated at R-C to give the external
aldimine of L-Ser, which releases L-Ser. The very slow formation
-1
f r
fittingofthe datatoeq2 withthe followingvalues:k =71.6s , k =
-1
3
7.3 s , and Keq = 21.9 mM.
of the ketimine from the β-hydroxyAsp residues suggests that
either the external aldimine formation or the aldimine-ketimine
interconversion is slow. Miles and Meister were surprised by the
stability of the 325 nm intermediate formed from β-hydroxyAsp
residues and proposed that either a protein nucleophile formed
stopped-flow instrument (Figure 1). This intermediate spectrum
remains constant during the steady state, but the spectrum
rapidly reverts back to the resting 360 nm form when substrate
is depleted (Figure 1). This result suggests that the 325 nm species
is a catalytically competent intermediate rather than a branch
species arising from abortive transamination. It is reasonable to
assign the structure of the 325 nm species formed from ABDC
and Asp to a PMP-ketimine, either with oxaloacetate, before
decarboxylation, or with pyruvate, after decarboxylation. Since
an adduct with the
by intramolecular addition of the β-hydroxy group of the nascent
-Ser to the aldimine (17). Since -Asp forms a similar transient
L-Ser aldimine or an oxazolidine was formed
L
L
intermediate from either diastereomer of β-hydroxyAsp, we
believe it is more likely that the 325 nm species is a relatively
stable ketimine. An alternative explanation for the stability of
the complex is that it is an enolamine formed subsequent to
β-decarboxylation, which may be only slowly reprotonated on
the β-carbon because of either steric effects or altered electro-
nics (Scheme 1). It is interesting in this regard that threo-β-
hydroxyAsp is a normal substrate for AAT, while erythro-β-
hydroxyAsp forms a quasi-stable quinonoid complex that
turns over very slowly (18). Since the active sites of ABDC and
AAT are nearly identical in structure (4), it seems likely that there
are similar interactions of the β-hydroxyAsp diastereomers
in both active sites. The structure of the quinonoid complex of
1
3
C isotope effects indicated that decarboxylation is not rate-
limiting (16), it seems more likely that the 325 nm complex in the
steady state is the ketimine of pyruvate with PMP.
Miles and Meister studied the reaction of ABDC with the threo
and erythro diastereomers of β-hydroxyAsp, and they found that
both isomers are slow substrates, undergoing β-decarboxylation
at ∼0.02% of the rate of
also noticed formation of a transient complex absorbing at
25 nm with both diastereomers of β-hydroxyAsp. The 325 nm
complex of β-hydroxyAsp is stable enough to be isolated by
L-Asp to form L-Ser (17). These authors
3

5
070 Biochemistry, Vol. 49, No. 24, 2010
Phillips et al.
F
IGURE 5: Effect of hydrostatic pressure on the ABDC-succinate complex. The solution contained ABDC (1 mg/mL), 0.1 M MES (pH 5.3),
and 0.1 M disodium succinate. (A) Spectra are shown at (1) 300, (2) 600, (3) 1100, (4) 1300, (5) 1500, (6) 1700, and (7) 1900 bar. (B) Absorbance at
25 nm of the ABDC-succinate complex at various pressures: (O) increasing pressure and (b) decreasing pressure. The line through the points is
the result of fitting the data to eq 3.
4
FIGURE 6: Reactionof R37A ABDC with50 mMdisodiumL-Asp in0.1 M sodium acetate (pH 6.0). (A) Scans are shown at0.005, 3.08, 5.49, 9.05,
and 15.80 s. The dotted line (spectrum 0) is the spectrum of the enzyme before the reaction. (B) Time courses at 360 and 425 nm.
erythro-β-hydroxyAsp with AAT shows a hydrogen bond be-
tween Tyr-70 and the β-OH group of the β-hydroxyAsp (19).
Mutation of Tyr-70 to Phe in AAT reduces the affinity of erythro-
β-hydroxyAsp and destabilizes the quinonoid complex, showing
that the hydrogen bond contributes significantly to binding (20).
In ABDC, the much weaker binding and slower reaction of
erythro-β-hydroxyAsp suggest that the hydrogen bond to Tyr-
reaction in which the PLP becomes partially protonated on the
PLP-lysine Schiff base (eq 4).
K ¼21:9 mM
Eð360 nmÞ þ succinate
F
R
1
k
f
¼71:6 s-
E succinateð360 nmÞ
F
E succinateð425 nmÞ ð4Þ
3
3
R
- 1
k
r
¼37:3 s
1
34 (homologous to Tyr-70 in AAT) is not able to form, and
A similar change in the PLP spectrum was observed with
noncovalent ligand binding to Escherichia coli AAT (21) and
was attributed to a conformational change between the open and
closed states. Thus, this slow reaction of ABDC with succinate is
likely a similar conformational change which changes the PLP
environment and favors the protonated ketoenamine by increas-
there may be a steric clash.
Reaction of ABDC with succinate, an unreactive Asp analo-
gue lacking the R-amino group, results in the formation of an
absorption peak at 425 nm, corresponding to a protonated
ketoenamine tautomer of the PLP-lysine Schiff base, from the
360 nm resting enzyme (Figure 3). The rate constant for this
ing the pK of the aldimine N (Scheme 2). The succinate likely
a
reaction is a function of succinate concentration, with a hyper-
bolic dependency (Figure 4). This result demonstrates that there
is rapid equilibrium binding of succinate, followed by a slower
binds by forming salt bridges to Arg-375 and Arg-497
(Scheme 2), which are homologous with the Arg residues in
AAT (Arg-292 and Arg-386, respectively, in E. coli AAT) that

Article
Biochemistry, Vol. 49, No. 24, 2010 5071
Scheme 1: Proposed Mechanism of Decarboxylation of β-HydroxyAsp by ABDC
bind the R- and β-carboxylates of Asp (22). This hypothesis that a
conformational change occurs in ABDC is supported by the
effects of hydrostatic pressure on the protonation equilibrium
of the succinate complex (Figure 5). In previous studies, we
demonstrated that increasing hydrostatic pressure changes the
allosteric equilibrium of tryptophan synthase to favor the open
conformation (23-25). In this study, the 425 nm absorption
shoulder of the protonated internal aldimine of ABDC in 0.1 M
succinate shifts under high pressure to the unprotonated cofactor
form [360 nm (Figure 5A)]. On the basis of our work with try-
ptophan synthase (23-25), this species is the open state, thus
confirming that ABDC must undergo a conformational change
to accommodate substrate and initiate transaldimination. The
observed pressure-dependent changes in the spectra are not likely
to arise from subunit dissociation for the following reasons. (1)
The highest pressures used in our studies (1.9 kbar) are lower than
those used for subunit dissociation studies, generally greater than
(i.e., first-order) upon decompression. In contrast, subunit dis-
sociation would result in hysteresis, with the reverse reaction
obeying second-order kinetics upon decompression. The ΔV for
o
the ABDC conformational equilibrium is relatively small (-50
mL/mol), much lower than that for tryptophan synthase (-125
to -200 mL/mol) (23) but similar to that seen for the equilibrium
between the 420 and 338 nm forms of the cofactor of tryptophan
indole-lyase, -38 mL/mol (27). Since at pH 5.3, in the presence of
succinate, the equilibrium still favors the neutral imine absorbing
at 360 nm (Figure 5A), the pK of ABDC must be e5.3 even in
a
the succinate complex. In contrast, for aspartate aminotransfer-
ase, the pK of the unliganded internal aldimine is 6.8, and it
a
increases to 8.8 in Michaelis complexes (21). In solution, the pKa
values of PLP aldimines are ∼12 (28). Since ABDC functions
physiologically when bacteria are subjected to acidic stress, the
low pK of the internal aldimine may be important in regulation
a
of enzyme activity in vivo.
2
.5 kbar (26). (2) The isosbestic point at 386 nm in the pressure
The structure of ABDC (4) shows that the geometry of the
PLP-lysine Schiff base is distorted, with the lysine ε-amino group
out of the plane of the PLP ring and a C-CdN bond angle of
∼152ꢀ (Figure 7), far from the ideal trigonal bond angle of 120ꢀ.
studies (Figure 5A) is the same as that seen in the stopped-flow
kinetic experiments (Figure 3A), indicating that the same two
spectroscopic species observed in the stopped-flow kinetic experi-
ment are involved in the pressure-sensitive equilibrium. (3) There
is no hysteresis in the data when the pressure is decreased, indi-
cating that the spectral changes are readily and rapidly reversible
0
˚
The distance from ε-N of Lys-315 to 3 -O of PLP is 3.68 A,
indicating that intramolecular H-bonding would be very weak
0
even if the N were protonated. Instead, the 3 -O of PLP is strongly

5
072 Biochemistry, Vol. 49, No. 24, 2010
Phillips et al.
Scheme 2: Effect of Succinate on the ABDC PLP Spectrum
FIGURE 7: Structure of PLP in the ABDC active site. The coordinates were taken from Protein Data Bank entry 3FDD. Hydrogen bonds are
represented by dashed green lines. Only heavy atoms (C, N, O, and P) are shown. This figure was prepared using SPDBV 4.01 (29).
-
1
˚
H-bonded to Asn-256 and Tyr-289, with distances of ∼2.8 A
325 nm intermediate (0.2 s ) is comparable to the steady state
-
1
(
Figure 7). Arg-37 is H-bonded to the bridging O of the
kcat (0.7 s ) (4), suggesting that aldimine formation may be rate-
determining for R37A ABDC. Furthermore, in contrast to wild-
type ABDC, mixing of R37A ABDC with disodium succinate at
pH 6 did not result in any change in the enzyme PLP spectrum
phosphate group of PLP (Figure 7). The protonated ketoena-
mine tautomer of the PLP-Lys-315 Schiff base is the one that is
expected to be reactive toward transaldimination with a sub-
strate, so a conformational change is likely to occur after
L-Asp
(data not shown), indicating that the pK of the Schiff base is very
a
binding to activate the PLP, followed by transaldimination,
low even in the succinate complex.
which is found for AAT (21). Indeed, a transient shoulder at
Conclusion. ABDC forms a PMP-ketimine intermediate
during turnover with substrates. These data suggest that Arg-
37 in ABDC influences the pKa of the PLP-Lys-315 Schiff
base. The electrostatic effect of this additional Arg residue in
∼
425 nm is observed in the reactions of L-Asp and both dia-
stereomers of β-hydroxyAsp (Figures 1A, 2A, and 3A). However,
it is interesting that R37A ABDC does not show the 425 nm
shoulder in the reaction with
25 nm complex with
wild-type ABDC (Figure 7). This very slow rate of reaction of
R37A ABDC may reflect the unfavorable protonation of the
internal aldimine. In fact, the rate constant for formation of the
L
-Asp, and the formation of the
the proximity of the PLP may contribute to the low pK of the
a
3
L
-Asp is ∼5000-fold slower than that for
aldimine N in the unliganded state. However, Arg-37 also has the
paradoxical effect of promoting protonation when ligands bind.
Possibly, the dianionic substrate neutralizes two of the three
arginine guanidinium cations and reduces the positive charge of

Article
Biochemistry, Vol. 49, No. 24, 2010 5073
the active site. Arg-37 may also swing away from the cofactor to
allow the substrate access. Thus, our data demonstrate that Arg-
16. Rosenberg, R. M., and O’Leary, M. H. (1985) Aspartate β-decarbox-
ylase from Alcaligenes faecalis: Carbon-13 kinetic isotope effect and
deuterium exchange experiments. Biochemistry 24, 1598–1603.
37, which is a conserved residue in ABDC, controls the proton-
17. Miles, E. W., and Meister, A. (1967) The mechanism of the reaction of
ation state, and hence the reactivity, of the PLP of ABDC,
β-hydroxyaspartate with
pyridoxal 5 -phosphate-enzyme inhibition. Biochemistry 6, 1734–
1743.
8. Jenkins, T. (1969) Glutamic-Aspartic Transaminase VI. The Reaction
with Certain β-Substituted Aspartic Acid Analogues. J. Biol. Chem.
L
-aspartate β-decarboxylase. A new type of
0
possibly by opposing the strain introduced by the H-bonds of
0
3
-O of PLP with Asn-256 and Tyr-289.
1
REFERENCES
236, 1121–1125.
1
. Tate, S. S., and Meister, A. (1971)
L
-Aspartate-β-decarboxylase:
Structure, catalytic activities, and allosteric regulation. Adv. Enzymol.
19. von Stosch, A. G. (1996) Aspartate aminotransferase complexed with
erythro-β-hydroxyaspartate: Crystallographic and spectroscopic
identification of the carbinolamine intermediate. Biochemistry 35,
15260–15268.
Relat. Areas Mol. Biol. 35, 503–543.
2
. Rathod, P. K., and Fellman, J. H. (1985) Identification of mammalian
aspartate-4-decarboxylase. Arch. Biochem. Biophys. 238, 435–446.
. Abe, K., Ohnishi, F., Yagi, K., Nakajima, T., Higuchi, T., Sano, M.,
Machida, M., Sarker, R. I., and Maloney, P. C. (2002) Plasmid-
encoded asp operon confers a proton motive metabolic cycle catalyzed
by an aspartate-alanine exchange reaction. J. Bacteriol. 184, 2906–
20. Hayashi, H., and Kagamiyama, H. (1995) Reaction of aspartate
aminotransferase with
3
L-erythro-3-hydroxyaspartate: Involvement
of Tyr70 in stabilization of the catalytic intermediates. Biochemistry
34, 9413–9423.
21. Hayashi, H., Mizuguchi, H., and Kagamiyama, H. (1998) The imine-
0
2
913.
pyridine torsion of the pyridoxal 5 -phosphate Schiff base of aspartate
4
. Lima, S., Sundararaju, B., Huang, C., Khristoforov, R., Momany, C.,
and Phillips, R. S. (2009) The crystal structure of the Pseudomonas
dacunhae aspartate-β-decarboxylase dodecamer reveals an unknown
oligomeric assembly for a pyridoxal-5 -phosphate-dependent enzyme.
J. Mol. Biol. 388, 98–108.
a
aminotransferase lowers its pK in the unliganded enzyme and is
crucial for the successive increase in the pK during catalysis. Bio-
a
chemistry 37, 15076–15085.
0
22. J €a ger, J., Moser, M., Sauder, U., and Jansonius, J. N. (1994) Crystal
structures of Escherichia coli aspartate aminotransferase in two
conformations. Comparison of an unliganded open and two liganded
closed forms. J. Mol. Biol. 239, 285–305.
5
. Chen, H. J., Ko, T. P., Lee, C. Y., Wang, N. C., and Wang, A. H.
(
2009) Structure, assembly, and mechanism of a PLP-dependent
dodecameric -aspartate β-decarboxylase. Structure 17, 517–529.
L
23. Phillips, R. S., Miles, E. W., Holtermann, G., and Goody, R. S. (2005)
Hydrostatic Pressure Stabilizes the Open Conformation of Salmonella
typhimurium Tryptophan Synthase. Biochemistry 44, 7921–7928.
24. Phillips, R. S., McPhie, P., Miles, E. W., Marchal, S., and Lange, R.
(2008) Quantitative Effects of Allosteric Ligands and Mutations on
Conformational Equilibria in Salmonella typhimurium Tryptophan
Synthase. Arch. Biochem. Biophys. 470, 8–19.
25. Phillips, R. S., Miles, E. W., Marchal, S., and Lange, R. (2008)
Pressure and Temperature Jump Relaxation Kinetics of Salmonella
typhimurium Tryptophan Synthase L-Serine Complex Reveal Large
6
. Jones, C. W., III, Leyden, D. E., and Stammer, C. H. (1969) threo and
erythro β-hydroxy-DL-aspartic acids. Can. J. Chem. 47, 4363–4366.
. Matheson, I. B. C. (1990) A critical comparison of least absolute
deviation fitting (robust) and least-squares fitting: The importance of
error distributions. Comput. Chem. 14, 49–57.
. Strickland, S., Palmer, G., and Massey, V. (1975) Determination of
dissociation constants and specific rate constants of enzyme-substrate
(or protein-ligand) interactions from rapid reaction kinetic data.
J. Biol. Chem. 250, 4048–4052.
7
8
9. Kitamura, Y., and Itoh, T. (1987) Reaction of protonic ionization for
buffering agents. J. Solution Chem. 16, 715–725.
Transition State Compressibility and Heat Capacity Changes. J. Am.
Chem. Soc. 130, 13580–13588.
1
1
0. Li, Y.-H. (1967) Equation of state of water and sea water. J. Geophys.
Res. 72, 2665–2678.
1. Soda, K., and Tanizawa, K. (1979) Kynureninases: Enzymological
properties and regulation mechanism. Adv. Enzymol. Relat. Areas
Mol. Biol. 49, 1–40.
26. Seifert, T., Bartholmes, P., and Jaenicke, R. (1982) Reconstitution of
the isolated β2-subunit of tryptophan synthase from Escherichia coli
after dissociation induced by high hydrostatic pressure. Equilibrium
and kinetic studies. Biophys. Chem. 15, 1–8.
27. Kogan, A., Gdalevsky, G. Y., Cohen-Luria, R., Goldgur, Y., Phillips,
R. S., Parola, A. H., and Almog, O. (2009) Conformational changes
and loose packing promote E. coli tryptophanase cold lability. BMC
Struct. Biol. 9, 65.
28. Kallen, R. G., Korpela, T., Martell, A. E., Matsushima, Y., Metzler,
C. M., Metzler, D. E., Morozov, Yu. V., Ralston, I. M., Savin, F. A.,
Torchinsky, Yu. M., and Ueno, H. (1985) in Transaminases
(Christen, P., and Metzler, D. E., Eds.) pp 37-108, John Wiley &
Sons, New York.
29. Guex, N., and Peitsch, M. C. (1997) SWISS-MODEL and the Swiss-
PdbViewer: An environment for comparative protein modeling.
Electrophoresis 18, 2714–2723.
1
1
1
1
2. Zheng, L., White, R. H., Cash, V. L., and Dean, D. R. (1994)
Mechanism for the desulfurization of L-cysteine catalyzed by the nifS
gene product. Biochemistry 33, 4714–4720.
3. Esaki, N., Karai, N., Nakamura, T., Tanaka, H., and Soda, K. (1985)
Mechanism of reactions catalyzed by selenocysteine β-lyase. Arch.
Biochem. Biophys. 238, 418–423.
4. Metzler, C. M., and Metzler, D. E. (1987) Quantitative description of
absorption spectra of a pyridoxal phosphate-dependent enzyme using
lognormal distribution curves. Anal. Biochem. 166, 313–327.
5. Novogrodsky, A., and Meister, A. (1964) Control of aspartate
β-decarboxylase activity by transamination. J. Biol. Chem. 239, 879–888.