The roles of Ser-36, Asp-132 and Asp-201 in the reaction of Pseudomonas fluorescens Kynureninase

Article abstract of DOI:10.1016/j.bbapap.2019.05.005

Kynureninase from Pseudomonas fluorescens (Pfkynase)catalyzes the pyridoxal-5′-phosphate (PLP)dependent hydrolytic cleavage of L-kynurenine to give anthranilate and L-alanine. Asp-132 and Asp-201 are located in the structure near the pyridine NH of the PLP, with Asp-201 forming a hydrogen bond. Mutation of Asp-132 to alanine and glutamate and Asp-201 to glutamate results in reduced catalytic activity with L-kynurenine and β-benzoyl-L-alanine, but not O-benzoyl-L-serine. D132A, D132E D201E and S36A mutant Pfkynases all can form quinonoid and vinylogous amide intermediates with β-benzoyl-L-alanine, similar to wild-type enzyme. D132A, D132E, and D201E Pfkynase react more slowly with β-benzoyl-L-alanine and benzaldehyde to form an aldol product absorbing at 490 nm than wild-type, with D132E reacting the slowest. The ¹H NMR spectra of wild-type and D201E Pfkynase are very similar in the low field region from 10 to 18 ppm, but that of D132A Pfkynase is missing a resonance at 13.1 ppm. These results show that these residues modulate the reactivity of the PLP at different stages during the reaction cycle. Ser-36 is located near the expected location of the carbonyl oxygen of the substrate. Mutation of Ser-36 to alanine results in a 230-fold reduction of k_cat and 30-fold reduction in k_cat/K_m with L-kynurenine, but very little effect on the reaction of O-benzoyl-L-serine. Thus, the rate-determining step in the reaction of S36A Pfkynase is the C_β-C_γ bond cleavage. These results support the hypothesis that Ser-36 together with Tyr-226 is part of an oxyanion hole that polarizes the carbonyl of the substrate in the catalytic mechanism of Pfkynase.

Full text of DOI:10.1016/j.bbapap.2019.05.005

BBA - Proteins and Proteomics 1867 (2019) 722–731
Contents lists available at ScienceDirect
BBA - Proteins and Proteomics
journal homepage: www.elsevier.com/locate/bbapap
The roles of Ser-36, Asp-132 and Asp-201 in the reaction of Pseudomonas
fluorescens Kynureninase
a,b,⁎
b
a
b
b
Robert S. Phillips
, Mori Crocker , Richard Lin , O. Elijah Idowu , David K. McCannon ,
b,1
Santiago Lima
a
b
Department of Chemistry, University of Georgia, Athens, GA 30602, United States of America
Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602, United States of America
A R T I C L E I N F O
A B S T R A C T
Keywords:
Kynureninase from Pseudomonas fluorescens (Pfkynase) catalyzes the pyridoxal-5′-phosphate (PLP) dependent
hydrolytic cleavage of L-kynurenine to give anthranilate and L-alanine. Asp-132 and Asp-201 are located in the
structure near the pyridine NH of the PLP, with Asp-201 forming a hydrogen bond. Mutation of Asp-132 to
alanine and glutamate and Asp-201 to glutamate results in reduced catalytic activity with L-kynurenine and β-
benzoyl-L-alanine, but not O-benzoyl-L-serine. D132A, D132E D201E and S36A mutant Pfkynases all can form
quinonoid and vinylogous amide intermediates with β-benzoyl-L-alanine, similar to wild-type enzyme. D132A,
D132E, and D201E Pfkynase react more slowly with β-benzoyl-L-alanine and benzaldehyde to form an aldol
Pyridoxal-5′-phosphate
Enzyme mechanism
Site-directed mutagenesis
Stopped-flow kinetics
NMR spectroscopy
1
product absorbing at 490 nm than wild-type, with D132E reacting the slowest. The H NMR spectra of wild-type
and D201E Pfkynase are very similar in the low field region from 10 to 18 ppm, but that of D132A Pfkynase is
missing a resonance at 13.1 ppm. These results show that these residues modulate the reactivity of the PLP at
different stages during the reaction cycle. Ser-36 is located near the expected location of the carbonyl oxygen of
the substrate. Mutation of Ser-36 to alanine results in a 230-fold reduction of kcat and 30-fold reduction in kcat
/
K with L-kynurenine, but very little effect on the reaction of O-benzoyl-L-serine. Thus, the rate-determining step
m
in the reaction of S36A Pfkynase is the C
β
-C bond cleavage. These results support the hypothesis that Ser-36
γ
together with Tyr-226 is part of an oxyanion hole that polarizes the carbonyl of the substrate in the catalytic
mechanism of Pfkynase.
1
. Introduction
The mechanism of kynureninase is proposed to proceed through a
ketimine intermediate formed rapidly from a quinonoid species [2,3].
Kynureninase [EC 3.7.1.3] is a pyridoxal-5′-phosphate (PLP) de-
The quinonoid intermediate is formed within the dead time (ca. 2 ms)
−1
pendent enzyme in the kynurenine pathway of tryptophan metabolism,
found in animals, fungi, and some bacteria [1]. It catalyzes the retro-
Claisen type hydrolytic cleavage of L-kynurenine to give L-alanine and
anthranilate (Eq. (1)).
of the stopped-flow instrument, and decays at ~700 s
to give the
ketimine intermediate (Scheme 1) [3]. The high rate of substrate α-
deprotonation in kynureninase suggests that the PLP is strongly elec-
trophilic. The electrophilicity of PLP-dependent enzymes is controlled
(
1)
⁎
Corresponding author at: Department of Chemistry, University of Georgia, Athens, GA 30602, United States of America.
E-mail address: plp@uga.edu (R.S. Phillips).
1
Present address: Department of Biology, Virginia Commonwealth University, 1000 West Cary Street Suite 126, Richmond, VA 23284-2012, USA.
https://doi.org/10.1016/j.bbapap.2019.05.005
Received 15 November 2018; Received in revised form 6 May 2019; Accepted 10 May 2019
Available online 14 May 2019
1570-9639/ © 2019 Elsevier B.V. All rights reserved.

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Scheme 1. Proposed mechanism of kynureninase.
by the protonation state of the pyridine nitrogen [4,5]. Pseudomonas
2. Experimental methods
1
fluorescens kynureninase (Pfkynase) has a conserved aspartate, Asp-
2
01, which accepts a hydrogen bond from the pyridinium NH, thus
2.1. Materials
raising the pK of the cofactor and ensuring that the nitrogen remains
a
protonated under catalytic conditions. This aspartate is strictly con-
served in all members of the aminotransferase superfamily [6] and is
critical for activity [4,7]. In addition, in kynureninase a strictly con-
served Asp-132 is located very close to Asp-201 and the pyridinium NH,
and may assist in stabilization of the protonated nitrogen. This Asp-132
residue is not conserved in other aminotransferase family members;
however, it is often replaced by other residues with polar side chains,
such as histidine. Thus, we were interested in the role of Asp-132 in the
mechanism of kynureninase. The carbon‑carbon bond cleavage step
occurs via a gem-diolate intermediate (Scheme 1) [8,9], which may be
stabilized by hydrogen bonding and/or acid/base catalysis with Tyr-
O-Benzoyl-L-serine [12] and S-(2-aminophenyl)-L-cysteine S,S-di-
oxide [9] were prepared as described previously. Buffers, solvents and
common reagents were products of Fisher Scientific Co.
2.2. Enzymes
D132A, D132E, D201E mutant Pfkynase were prepared as pre-
viously described [13]. The kynU gene was amplified by PCR from
pTZKYN [14] and ligated into pLATE11 to allow for expression in E. coli
BL21(DE3) host cells. S36A mutant Pfkynase was prepared using a
modification of the Quick-Change protocol with partially overlapping
primers [15]. The mutated positions are underlined below.
2
26 and the phosphate of PLP [10]. Ser-36 is strictly conserved and
located near the substrate binding pocket, in a potential position to
form a hydrogen bond with the substrate carbonyl oxygen [11]. Hence,
we postulated that Ser-36, together with Tyr-226, may form an “oxy-
anion hole” to accelerate water addition to the carbonyl to form the
critical gem-diolate intermediate. We have now studied the roles of Ser-
Forward: 5’-CTACGCCATCGGTTGCACTTTCAAAT.
Reverse: 5’-GACCGCCATTGAGGTATTTGAAAGTG.
The resulting S36A mutant plasmid was transformed into E. coli
DH5α, colonies were selected and plasmids isolated, sequenced to
confirm the mutation, and transformed into E. coli BL21(DE3) cells for
expression. Cells were grown at 37 °C for 24 h using Studier auto-
induction medium [16], and the Pfkynases were purified by Phenyl-
Sepharose chromatography as described previously [14]. The con-
centration of enzyme was estimated from the absorbance at 280 nm,
3
6, Asp-132 and Asp-201 in the mechanism of Pfkynase by site-directed
mutagenesis, rapid-scanning stopped-flow spectrophotometry, and nu-
clear magnetic resonance spectrophotometry. The results presented
herein are consistent with the proposed roles for these residues in
catalysis.
1
%
using A280 = 14.3 [17]. The PLP content was determined by the
absorption of the cofactor at 388 nm in 0.1 M NaOH [18].
723

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
2
.3. Kinetic studies
values within 1 order of magnitude of wild-type kynureninase. The
D201A and D132A mutations decrease PLP binding about 10-fold.
4
Kinetics measurements were performed on a Cary 1 UV–vis spec-
However, kcat/K
m
is reduced by 10 for D201A, but only about 10-fold
trophotometer equipped with a 6 × 6 Peltier temperature-controlled
cell compartment. The kinetic studies were performed at 37 °C in
for D132A [13]. Thus, Asp-201 is primarily responsible for binding PLP
and maintaining electrophilicity of the PLP during catalysis, as ex-
pected since this aspartate is conserved in all members of the amino-
transferase family. It is interesting that the glutamate mutants, D132E
and D201E, show significantly better PLP binding than wild-type en-
zyme [13]. This suggests that these aspartate residues may introduce
strain in the internal aldimine that contributes to catalysis. In the
crystal structure of human kynureninase, the imine bond is rotated 39°
out of plane with the PLP, and the C4’-N-Cα bond angle is 127°, rather
4
0 mM potassium phosphate, pH 8.0, 50 μM PLP, with varying amounts
of substrate, following the absorbance decrease at 360 nm
−
1
−1
(
Δε = −4500 M cm ) as kynurenine is converted to anthranilic
−1 −1
acid, or at 246 nm (Δε = −4500 M cm ) as β-benzoyl-L-alanine is
converted to benzoic acid [19]. For the reaction of O-benzoyl-L-serine,
the formation of pyruvate was followed by addition of 20 μg/mL rabbit
muscle lactate dehydrogenase (USBiochemicals) and 0.1 mM NADH, at
−
1
−1
2
3
40 nm (Δε = −6220 M cm ). The kinetic parameters were ob-
than the expected 120° for an sp nitrogen. It has been found that other
tained by fitting the data to Eqs. (2) and (3), using HYPERO and
COMPO [20] for kcat and kcat/K and K , respectively. The kcat and kcat
parameters were obtained from Vmax and Vmax/K using the enzyme
molecular weight of 47 kDa [14].
members of the aminotransferase superfamily, tyrosine phenol-lyase
[5], aspartate aminotransferase [24] and aspartate β-decarboxylase
[25] show similar strain in the internal aldimines.
m
i
/
K
m
m
In the crystal structure of human kynureninase with a bound sub-
strate analogue inhibitor, 3-hydroxyhippurate, we found that the car-
bonyl oxygen of the ligand was located ~4 Å from the OH of Tyr-275
and Ser-75 (Fig. 1) [11]. This tyrosine and serine are found to be strictly
conserved in all kynureninase sequences. We predicted that these re-
sidues could be donating hydrogen bonds to polarize the carbonyl
oxygen during catalysis. Mutation of the corresponding tyrosine in
Pfkynase to phenylalanine, Y226F, resulted in very low activity, about
V = Vmax S/(K
V = Vmax S/(K
m
+ S)
(2)
m
(1 + I/K
i
) + S)
(3)
2
.4. Stopped-flow kinetics
Stopped-flow kinetics experiments were performed on an OLIS RSM-
000 and an Applied Photophysics SX-18 stopped-flow spectro-
0
.1%, with L-kynurenine, in agreement with this prediction [10]. Fur-
1
thermore, we suggested that Tyr-226, which hydrogen bonds to the
phosphate of PLP, may be involved in transferring a proton between the
photometer. The enzymes were diluted to ca. 2 mg/mL in 0.04 M po-
tassium phosphate, pH 8.0, before mixing. Ligand solutions were pre-
pared in the same buffer. The kinetic data were collected at room
temperature, 20–22 °C. The data were analyzed with OLIS Global Works
software [21] to obtain the kinetic parameters.
3
1
carbonyl oxygen of substrate and the phosphate. P NMR spectra of
kynureninase with and without an inhibitor bound showed that the
phosphate is in the dianionic state in the free enzyme, but becomes
protonated to the monoanion in the complex with 5-bromodihydro-L-
kynurenine [10]. Thus, this suggested that the phosphate of PLP is
playing an active role in the catalytic mechanism as an acid-base cat-
alyst. In this paper, we have examined the proposition that Ser-36 may
be assisting Tyr-226 in activating the carbonyl group for nucleophilic
addition of water in the catalytic mechanism. We prepared the S36A
2
.5. NMR spectra
The enzymes were concentrated to a monomer concentration of
7 mg/mL (1 mM) in 0.04 M potassium phosphate, pH 8 containing
4
1
0% D
2
O. The data were collected at 499.8 MHZ in a Varian Innova
mutant Pfkynase, with the results shown in Table 1. The value of k for
cat
instrument at 25 °C. Water suppression was performed with the Jump-
return pulse sequence. The spectrum width was 8 kHz collected in 18 k
data points. Acquisition time was 1.125 s, and 512 transients were
collected.
L-kynurenine is reduced about 230-fold, K is reduced 7-fold, and k
/
m
cat
K
m
is reduced about 30-fold, as a result of the mutation. However, since
the rate-determining step for reaction of wild-type Pfkynase with L-
kynurenine is release of the second product, L-alanine [2], the actual
effect on the C
β
-C bond cleavage step can be estimated from the rapid-
γ
3
. Results and discussion
3.1. Steady-state kinetics
The active site of enzymes in the aminotransferase family has
strictly conserved aspartate and lysine residues, that are involved in
cofactor binding and catalysis [6]. Sequence alignments initially pre-
dicted that Asp-201 and Lys-227 are these residues in Pfkynase. The
crystal structure of Pfkynase confirmed that these residues are involved
in PLP binding [13]. Lys-227 forms a covalent Schiff's base linkage with
the cofactor in the resting enzyme, and likely plays a role as a catalytic
base while free during the catalytic cycle, as it does in aspartate ami-
notransferase [22]. The function of Asp-201 is to maintain the proto-
nation state of the PLP during catalysis by forming an ion pair and
hydrogen bond with the pyridinium NH [5]. Thus, this aspartate can
modulate the electrophilicity of the cofactor during catalysis by ad-
justing the strength of the hydrogen bond in different intermediates.
Usually, replacement of this critical aspartate residue with a non polar
amino acid like alanine results in an inactive or very weakly active
enzyme [4,5,23]. In contrast, mutation of aspartate to glutamate would
be expected to preserve the critical ion pairing and hydrogen bonding.
We have previously mutated Asp-132 and Asp-201 in Pfkynase to Ala
and Glu [13]. With the exception of D201A, the other mutant enzymes
Fig. 1. Crossed-eye stereo view of the overlay of active sites of Pfkynase and
human kynureninase. The human Kynase structure (PDB ID: 3E9K) is in cyan
and that of PfKynase (PDB ID: 1QZ9) is in green. The numbers are those of
PfKynase and in parentheses are the corresponding residue numbers of human
Kynase. The carbonyl oxygen of 3-hydroxyhippurate (3H) is located between
Ser36(75) and Tyr226(275). The blue dashes indicate possible hydrogen bonds.
(
For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)
have readily measurable activity with L-kynurenine, and have kcat/K
m
724

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Table 1
Activity of wild-type and S36A mutant Pfkynases.
Enzyme
Substrate
k
cat, s−1
a
kcat/Km, M−1 s−1
m
K , μM
5a
4b
Wild-type
Wild-type
Wild-type
Y226F
L-kynurenine
16
6.3 × 10
25
9
b
β-benzoyl-L-alanine
O-benzoyl-L-serine
L-kynurenine
0.7
8.0 × 10
c
4c
2.5 ± 0.2
(1.1 ± 0.1) × 10
230 ± 30
−
3
c
(5.8 ± 0.3) × 10
0.10 ± 0.02
0.069 ± 0.019
< 0.001
530 ± 85
11 ±
2
c
Y226F
O-benzoyl-L-serine
L-kynurenine
41 ± 11
2400 ± 600
3.5 ± 1.1
S36A
(2.0 ± 0.3) × 104
< 714
S36A
β-benzoyl-L-alanine
O-benzoyl-L-serine
i
1.4 (K )
S36A
1.7 ± 0.1
(2.4 ± 0.2) × 104
71 ±
7
a
b
c
From reference [2].
From reference [19].
From reference [10].
quench data to be about 1000-fold [2]. The OG of Ser-36 must donate a
hydrogen bond to the substrate γ‑carbonyl in the transition state for
water addition that provides reduction of the activation energy by
about 4.5 kcal/mol. The S36A mutation has a similar effect on activity
and β-benzoyl-L-alanine is 3.5–25 μM (Table 1) [2,19], so these reac-
tions are fully saturated at 1 mM.
The reaction of wild-type Pfkynase with β-benzoyl-L-alanine shows
the very fast formation of a peak at about 348 nm, with it being par-
tially formed within the dead time (ca. 2 ms) of the stopped-flow in-
strument (Fig. 2A, B). This peak has been assigned previously to a vi-
nylogous amide not on the direct catalytic pathway [26]. There is a
small peak of a quinonoid complex at 490 nm in the first scan which
decays rapidly as the 348 nm species is formed. All of the mutant en-
zymes studied react with 1 mM β-benzoyl-L-alanine to form a similar
complex with a strong absorption peak at about 345 nm (Figs. 3–6A, B).
However, there are significant differences in the rates of formation of
as the Y226F mutation [10]. Thus, the ‘oxyanion hole’ created by Tyr-
6
2
26 and Ser-36 contributes about 10 to the rate acceleration of Pfky-
nase. β-Benzoyl-L-alanine is an alternative substrate for kynureninase,
with a kcat/K
m
value about 10% that of L-kynurenine for wild-type
Pfkynase [19,26]. With β-benzoyl-L-alanine, we did not observe any
activity in steady state kinetic assay for S36A Pfkynase, following the
absorbance decrease of the compound at 246 nm, but we could measure
a K value of 1.4 μM. However, O-benzoyl-L-serine shows activity with
i
S36A Pfkynase for elimination of benzoate comparable to that of the
wild-type enzyme (Table 1), and significantly better than found pre-
viously for Y226F Pfkynase (Table 1) [10].
the complex. The rate constants, k
1
and k , for this reaction are sum-
2
marized in Table 2.
The formation of the 345 nm intermediate occurs in two phases for
wild-type, D132A, D132E, and S36A enzymes. The fast phase (k in
1
Table 2) possibly corresponds to the formation of a ketimine inter-
mediate (Scheme 1), which should have an absorption maximum at
about 325 nm, by reprotonation at C-4′ of the initial quinonoid inter-
mediate formed within the dead time of the instrument [3]. This is 3–5-
fold slower for the mutants, except for D201E. The rate constant of the
3
.2. Reaction of wild-type, D132A, D132E, and D201E Pfkynase with β-
benzoyl-L-alanine and benzaldehyde
Since the UV spectrum of β-benzoyl-L-alanine does not overlap with
the enzyme PLP spectrum, unlike L-kynurenine, it is a useful substrate
for spectroscopic investigation of the reaction mechanism [19,26]. We
have now examined the reaction of wild-type, D132A, D132E, D201E,
and S36A Pfkynase with 1 mM β-benzoyl-L-alanine, without and with
benzaldehyde, by rapid-scanning stopped-flow spectrophotometry.
second phase (k in Table 2) of the 345 nm absorbance change varies
2
only about 4-fold for all the enzymes, indicating that none of the mu-
tations significantly affects this step of the mechanism. We have sug-
gested that this step corresponds to adventitious deprotonation of the β-
C to give the vinylogous amide [Scheme 1], most likely by Lys-227,
Note that the K
m
of wild-type and mutant Pfkynase for L-kynurenine
Fig. 2. Reaction of wild-type Pfkynase with β-ben-
zoyl-L-alanine without and with benzaldehyde. A.
Reaction of 20 μM Pfkynase with 1 mM β-benzoyl-L-
alanine. Scans are shown at 0.002 (black), 0.010
(
red), 0.020 (green), 0.040 (blue), 0.080 (orange),
0
.160 (maroon), 0.320 (violet) and 0.640 (indigo)
seconds. The dashed line is the spectrum of the free
enzyme. B. Time courses at 345 (black) and 424
(
red) nm for the reaction shown in A. C. Reaction of
2
1
0 uM Pfkynase with 1 mM β-benzoyl-L-alanine and
0 mM benzaldehyde. Scans are shown at 0.025
(
(
black), 1 (red), 2 (green), 4 (blue), 8 (orange), 16
maroon), and 30 (violet) seconds. D. Time courses
at 345 (black) and 488 (red) nm for the reaction
shown in C.(For interpretation of the references to
color in this figure legend, the reader is referred to
the web version of this article.)
725

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Fig. 3. Reaction of D132A Pfkynase with β-benzoyl-
L-alanine without and with benzaldehyde. A.
Reaction of 20 μM D132A Pfkynase with 1 mM β-
benzoyl-L-alanine. Scans are shown at 0.002 (black),
0
.020 (red), 0.040 (green), 0.080 (blue), 0.160 (or-
ange), 0.320 (maroon), 0.640 (violet) and 0.920
indigo) seconds. The dashed line is the spectrum of
the free enzyme. B. Time courses at 344 (black) and
(
4
25 (red) nm for the reaction shown in A. C.
Reaction of 20 μM D132A Pfkynase with 1 mM β-
benzoyl-L-alanine and 10 mM benzaldehyde. Scans
are shown at 0.20 (black), 1 (red), 2 (green), 4
(
blue), 8 (orange), 16 (maroon), and 30 (violet)
seconds. D. Time courses at 345 (black) and 488
(
red) nm for the reaction shown in C.(For inter-
pretation of the references to color in this figure le-
gend, the reader is referred to the web version of this
article.)
since we observed a primary isotope effect on its formation with a β-
deuterated substrate [26].
slower than D132A and 170-fold slower than wild-type Pfkynase
(Table 2). Thus, the mutation of Asp-132 to glutamate is surprisingly
worse than mutation to alanine for the carbon bond cleavage step. This
is likely due to unfavorable steric effects of the mutation affecting Asp-
201. Asp-132 must be involved in controlling the orientation of Asp-201
to interact with the PLP nitrogen, and this may change during the
catalytic cycle.
We have previously studied D132A and D132E Pfkynases and found
them to have measurable activity, around 10% of wild-type enzyme
[
13]. Thus, this aspartate is not essential for activity, but is likely to
affect the cofactor properties through its contact with Asp-201. The
cofactor spectrum of D132A is altered from that of wild-type Pfkynase
(
dashed line in Fig. 3), with peaks at 340 and 420 nm, showing that the
The visible spectrum of D201E Pfkynase is similar to that of wild-
type Pfkynase, with a peak at about 420 nm (Fig. 5, dashed line). Thus,
this mutation seems to have minimal effects on the cofactor environ-
ment in the internal aldimine. However, in the formation of the 345 nm
vinylogous amide intermediate (Fig. 5A), D201E exhibits only one
phase rather than two (Fig. 5B, Table 2). This suggests that the faster
phase, which is likely to be reprotonation of the quinonoid at C-4′ [2,3],
has been slowed by the D201E mutation so that it can no longer be
separately observed. As mentioned previously, D201E Pfkynase binds
PLP about 10-fold more tightly than wild-type enzyme [13]. The
cofactor equilibrium between enolimine and ketoenamine tautomers
27] has been affected by the mutation. Despite this difference, the
[
reaction of D132A Pfkynase with β-benzoyl-L-alanine to form the
3
45 nm intermediate is only modestly affected (Fig. 3A, B, Table 1).
D132E Pfkynase also shows an altered PLP spectrum with a shoulder
at about 340 nm (Fig. 4A), indicating a change in the electrostatic en-
vironment of the cofactor. In contrast, although D132E Pfkynase forms
the 345 nm intermediate (Fig. 4A) with similar kinetics as D132A
(
Fig. 4B, Table 2), it forms the 490 nm intermediate about 25-fold
Fig. 4. Reaction of D132E Pfkynase with β-benzoyl-
L-alanine without and with benzaldehyde. A.
Reaction of 20 μM D132E Pfkynase with 1 mM β-
benzoyl-L-alanine. Scans are shown at 0.002 (black),
0
.020 (red), 0.040 (green), 0.080 (blue), 0.160 (or-
ange), 0.320 (maroon), 0.640 (violet) and 0.920
indigo) seconds. The dashed line is the spectrum of
the free enzyme. B. Time courses at 344 (black) and
(
4
25 (red) nm for the reaction shown in A. C.
Reaction of 20 μM D132E Pfkynase with 1 mM β-
benzoyl-L-alanine and 10 mM benzaldehyde. Scans
are shown at 0.20 (black), 1 (red), 2 (green), 4
(
blue), 8 (orange), 16 (maroon), and 30 (violet)
seconds. D. Time courses at 345 (black) and 488
(
red) nm for the reaction shown in C.(For inter-
pretation of the references to color in this figure le-
gend, the reader is referred to the web version of this
article.)
726

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Fig. 5. Reaction of D201E Pfkynase with β-benzoyl-
L-alanine without and with benzaldehyde. A.
Reaction of 20 μM D201E Pfkynase with 1 mM β-
benzoyl-L-alanine. Scans are shown at 0.002 (black),
0
.020 (red), 0.040 (green), 0.080 (blue), 0.160 (or-
ange), 0.320 (maroon), 0.640 (violet) and 0.920
indigo) seconds. The dashed line is the spectrum of
the free enzyme. B. Time courses at 344 (black) and
(
4
25 (red) nm for the reaction shown in A. C.
Reaction of 20 μM D201E Pfkynase with 1 mM β-
benzoyl-L-alanine and 10 mM benzaldehyde. Scans
are shown at 0.20 (black), 1 (red), 2 (green), 4
(
blue), 8 (orange), 16 (maroon), and 30 (violet)
seconds. D. Time courses at 345 (black) and 488
(
red) nm for the reaction shown in C.(For inter-
pretation of the references to color in this figure le-
gend, the reader is referred to the web version of this
article.)
internal aldimines of PLP-dependent enzymes are often strained by
distortion of the torsion angle of the imine bond [24,25], and the strain
is enforced by the aspartate hydrogen bonding to the PLP nitrogen [7].
This strain may increase the reactivity of the internal aldimine with the
incoming amino acid substrate [24]. The stronger PLP binding in the
D201E mutant and reduced reactivity in the fast phase may be the re-
sult of reduced strain of the internal aldimine. The rate of formation of
the vinylogous amide is within a factor of 2 of wild-type Pfkynase.
Thus, the effect of the D201E mutation is selectively observed on dif-
ferent steps in the mechanism, indicating that Asp-201 can actively
modulate the reactivity of the cofactor throughout the catalytic cycle.
The spectrum of the internal aldimine of S36A Pfkynase is similar to
wild-type enzyme (Fig. 6A, dashed line). S36A Pfkynase shows the
formation of the 345 nm intermediate from β-benzoyl-L-alanine, de-
spite its lack of catalytic activity with this ligand (Fig. 6A). The rate of
formation of this species is only about 3-fold slower than wild-type
Pfkynase (Table 2).
Table 2
Rate constants for reaction of kynureninase with β-benzoyl-L-alanine and
benzaldehyde.
, 345 nm, s−1
, 345 nm, s−1
, 490 nm, s−1
Enzyme
k
1
k
2
k
3
Wild-type
D132A
D132E
D201E
S36A
157 ± 21
7.8 ± 0.1
1.8 ± 0.1
3.9 ± 0.1
4.3 ± 0.1
3.0 ± 0.3
0.66 ± 0.02
35 ±
37 ±
5
5
0.10 ± 0.01
0.0039 ± 0.0002
0.012 ± 0.004
Not observed
Not observed
59 ±
7
after the release of the carboxylate product is trapped by an aldol-type
reaction to give a γ-phenylhomoserine product [8,28], which forms a
quasi-stable quinonoid complex with a prominent peak at about 490 nm
(
Figs. 1–4C, D). The wild-type and all mutant Pfkynases except S36A
form a detectable peak at 490 nm with β-benzoyl-L-alanine and ben-
zaldehyde under these conditions. With wild-type Pfkynase, the rate
In the presence of benzaldehyde, the enamine intermediate formed
Fig. 6. Reaction of S36A Pfkynase with β-benzoyl-L-
alanine. A. Reaction of 20 μM S36A Pfkynase with
1
mM β-benzoyl-L-alanine. Scans are shown at 0.002
(
black), 0.020 (red), 0.040 (green), 0.080 (blue),
0
0
.160 (orange), 0.320 (maroon), 0.640 (violet) and
.920 (indigo) seconds. The dashed line is the spec-
trum of the free enzyme. B. Time courses at 344
(
black) and 425 (red) nm for the reaction shown in
A.(For interpretation of the references to color in this
figure legend, the reader is referred to the web ver-
sion of this article.)
727

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Fig. 7. Reaction of S36A Pfkynase with L-kynurenine
without and with benzaldehyde. A. Reaction of
2
0 μM S36A Pfkynase with 1 mM L-kynurenine.
Scans are shown at 0.002 (black), 0.020 (red), 0.040
(
(
green), 0.080 (blue), 0.160 (orange), 0.320
maroon), 0.640 (violet) and 0.920 (indigo) seconds.
B. Time course at 488 nm for the reaction shown in
A. C. Reaction of 20 μM S36A Pfkynase with 1 mM
kynurenine and 10 mM benzaldehyde. Scans are
shown at 0.20 (black), 1 (red), 2 (green), 4 (blue), 8
(
orange), 16 (maroon), and 30 (violet) seconds. D.
Time course at 488 nm for the reaction shown in C.
(
For interpretation of the references to color in this
figure legend, the reader is referred to the web ver-
sion of this article.)
constant for formation of this intermediate from benzaldehyde is
0.07 = 957-fold. Thus, the effect of the S36A mutation is specifically on
the chemical step in the mechanism. Furthermore, the reaction of O-
benzoyl-L-serine shows no effect of the mutation (Table 1). This sub-
strate undergoes a mechanistically simpler β-elimination reaction from
a quinonoid intermediate, and the Ser-36 hydroxyl would not be re-
quired to activate the substrate. These results support our hypothesis
that Ser-36 specifically donates a hydrogen bond to the kynurenine
γ‐carbonyl oxygen (Fig. 1), in order to polarize it and activate it for
nucleophilic addition of water, by forming an “oxyanion hole”, re-
miniscent of serine proteases [29].
identical to kcat for β-benzoyl-L-alanine, indicating that C
β
-C bond
γ
cleavage is rate-determining for this substrate in the steady-state
compare Tables 1 and 2) [26]. Thus, the rate constants for formation of
(
the 490 nm peak with benzaldehyde (k in Table 2) allows us to com-
3
pare directly the effects of the mutations on the chemical step. It is
interesting that D132A is the fastest of the mutants to form this inter-
−
1
mediate, at 0.1 s
(Fig. 2C, D, Table 2), while D132E and D201E are
much slower to form the 490 nm species (Figs. 3 and 4C, D), indicating
that the C -C bond cleavage step is much slower for these mutants,
possibly due to stabilization of the vinylogous amide.
β
γ
3.4. Reaction with S-(2-aminophenyl)-L-cysteine S,S-dioxide
3.3. Reaction of S36A Pfkynase with L-kynurenine and benzaldehyde
S-(2-Aminophenyl)-L-cysteine S,S-dioxide (SAPCO) is a potent
S36A Pfkynase shows no detectable formation of the 490 nm in-
competitive inhibitor of Pfkynase, with a K of 30–70 nM, that was
i
termediate from β-benzoyl-L-alanine and benzaldehyde under the
conditions of this experiment (data not shown), neither does it show
any detectable steady-state activity with β-benzoyl-L-alanine, although
it has measurable activity with L-kynurenine (Table 1). Hence, we ex-
amined its reaction with L-kynurenine and benzaldehyde in the rapid
scanning stopped-flow spectrophotometer. Similar to wild-type Pfky-
nase, a peak at 494 nm is formed in the dead-time (ca. 2 ms) of the
instrument, and then rapidly decays (Fig. 7A) by reprotonation [3]. The
originally designed by us as a potential transition state analogue [9,30].
Since the S36A mutation has a large effect on the activity, we reasoned
that it would also affect the inhibition by SAPCO. However, we could
not determine K of the mutant S36A by steady state kinetics with this
i
compound, since the low activity required enzyme concentrations
comparable to the inhibitor concentration. At the lowest concentration
of S36A that could give reliable kinetic results, 100 nM, we saw nearly
−
1
100% inhibition at 100 nM SAPCO. Hence, we can conclude that the K
rate constant for the decay is 300 ± 20 s (Fig. 7B), only slightly less
i
−
1
for the S36A mutant enzyme is < 100 nM, and likely similar to that of
wild type Pfkynase, 30 nM [31]. This is different than Y226F Pfkynase,
than that of wild-type Pfkynase, 720 s
[3]. When benzaldehyde is
included in the reaction of S36A Pfkynase with L-kynurenine, a new
peak at 490 nm grows in slowly (Fig. 7C). The formation of this com-
plex is biphasic (Fig. 6D), similar to wild-type Pfkynase (Fig. 1D). The
which shows a K value about 30-fold higher than wild-type for SAPCO
i
[
8]. Since there is a much larger effect of the mutations on the catalytic
−
1
mechanism than on the binding of SAPCO, it is unlikely that SAPCO is a
transition-state analogue, as we thought originally [9]. SAPCO forms a
quinonoid complex with an intense peak at 494 nm with wild-type
Pfkynase [30]. We examined the reaction of S36A, D132A, D132E, and
D201E mutants with SAPCO by rapid scanning stopped-flow spectro-
photometry, with the results shown in Fig. 8. The formation of the
quinonoid intermediates from SAPCO occurs rapidly, almost completely
within the dead time of the instrument, so the rate constant is > 300
fast phase shows a rate constant of 0.07 ± 0.002 s , in excellent
agreement with kcat of L-kynurenine for this mutant (Table 1). The
slower phase probably represents the gradual increase in concentration
of the inhibitory γ-phenylhomoserine aldol product [8], favoring the
equilibrium for formation of the quinonoid complex. In contrast to
wild-type Pfkynase [2], the cleavage of the C
β
γ
-C bond is rate-de-
termining with L-kynurenine for S36A Pfkynase. The rate-determining
step in steady-state for the reaction of wild-type Pfkynase with L-ky-
nurenine is the release of the second product, L-alanine [2]. Since the
−
1
s
, for all of these enzymes except D132A (Fig. 8). Thus, the muta-
−
1
tions do not have significant effects on quinonoid intermediate forma-
tion.
C
β
-C bond cleavage occurs with k = 67 s for wild-type Pfkynase [2],
γ
the S36A mutation affects the intrinsic chemical step by 67/
728

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
Fig. 8. Reaction of Pfkynases with S-(2-amino-
phenyl)-L-cysteine S,S-dioxide. A. Reaction of 20 μM
D132A Pfkynase with 1 mM S-(2-aminophenyl)-L-
cysteine S,S-dioxide. Scans are shown at 0.002
(
black), 0.020 (red), 0.040 (green), 0.080 (blue),
0
0
.160 (orange), 0.320 (maroon), 0.640 (violet) and
.920 (indigo) seconds. B. Time courses at 494 nm
for D132A (black), D201E (red), D132E (blue) and
S36A(green) for the reaction shown in A.(For inter-
pretation of the references to color in this figure le-
gend, the reader is referred to the web version of this
article.)
3.5. NMR spectra of wild-type, D132A and D201E Pfkynase
The downfield region of the 1H NMR spectra of PLP-dependent
enzymes exhibits resonances associated with the bound cofactor [31].
Since Asp-132 and Asp-201 are in close contact with the PLP, we ex-
pected to see some changes in this region of the NMR spectra in the
mutants. The distance between the OGs of Asp-132 and Asp-201 in the
structure of Pfkynase is 2.5 Å (Fig. 9); thus, one of these carboxylates
must be protonated and form a hydrogen bond to avoid severe elec-
trostatic repulsion. Since Asp-201 is in hydrogen-bonding contact
(
2.7 Å) with the pyridinium nitrogen of PLP, it is reasonable to propose
that it is Asp-132 that is protonated. The NMR spectra of wild-type,
D132A and D201E Pfkynase are shown in Fig. 10. The spectra of wild-
type and D201E Pfkynase are nearly identical, except that two peaks at
about 12 ppm have merged in D201E Pfkynase. In contrast, the peak at
1
3.1 ppm is almost absent in the spectrum of D132A mutant, suggesting
that this proton is associated with Asp-132. In aspartate amino-
transferase, there is a corresponding histidine, His-143, that hydrogen
1
bonds to Asp-222 and gives a H resonance at about 14 ppm [31]. Re-
Fig. 10. 500 MHz 1H NMR spectra of wild-type, D132A and D201E Pfkynases.
cent high resolution x-ray structures and density functional theory
calculations show that this histidine affects the pK of the pyridinium
a
Fig. 9. Crossed-eye stereo view of the active site of Pfkynase (1QZ9.pdb). The hydrogen bonds are indicated with blue dashes.(For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
729

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
N1 and the position of the proton between Asp-222 and the nitrogen
Funding
[
32]. It is likely that Asp-132 has a similar function in Pfkynase. The
farthest downfield resonance at 17.1 ppm present in all the NMR
spectra is likely to be the proton on the PLP nitrogen, by comparison
with aspartate aminotransferase [31].
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
References
3
.6. The mechanism of kynureninase
[1] K. Soda, K. Tanizawa, Kynureninases: enzymological properties and regulation
mechanism, Adv. Enzymol. Relat. Areas Mol. Biol. 49 (1979) 1–40.
[
[
[
2] S.V. Koushik, J.A. Moore 3rd, B. Sundararaju, R.S. Phillips, The catalytic me-
chanism of kynureninase from Pseudomonas fluorescens: insights from the effects of
pH and isotopic substitution on steady-state and pre-steady-state kinetics,
Biochemistry 37 (1998) 1376–1382.
The proposed mechanism of Pfkynase is shown in Scheme 1.
Binding of L-kynurenine to the internal aldimine gives the external al-
dimine. The phosphate of PLP accepts a proton from the substrate
during transaldimination [11]. Deprotonation of the α-CH of the ex-
3] R.S. Phillips, B. Sundararaju, S.V. Koushik, The catalytic mechanism of kynur-
eninase from Pseudomonas fluorescens: evidence for transient quinonoid and keti-
mine intermediates from rapid-scanning stopped-flow spectrophotometry,
Biochemistry 37 (1998) 8783–8789.
−
1
ternal aldimine occurs with k > 2000 s , since the quinonoid inter-
mediate, absorbing at about 490 nm, is formed within the dead time of
the stopped-flow instrument [3]. The mutations studied here do not
significantly affect the formation of the quinonoid intermediates
4] T.V. Demidkina, N.G. Faleev, A.I. Papisova, N.P. Bazhulina, V.V. Kulikova,
P.D. Gollnick, R.S. Phillips, Aspartic acid 214 in Citrobacter freundii tyrosine phenol-
lyase ensures sufficient C–H-acidity of the external aldimine intermediate and
proper orientation of the cofactor at the active site, Biochim. Biophys. Acta (BBA) -
Protein Proteomics 1764 (2006) 1268–1276.
(
Fig. 8). The quinonoid intermediate of wild-type Pfkynase then is ra-
−1
pidly reprotonated, with k~720 s , at the 4′‑carbon to give a ketimine
[
5] D. Milić, T.V. Demidkina, L.N. Zakomirdina, D. Matković-Čalogović, A.A. Antson,
Crystal structure of Citrobacter freundii Asp214Ala tyrosine phenol-lyase reveals that
Asp214 is critical for maintaining a strain in the internal Aldimine, Croat. Chem.
Acta 85 (2012) 283–288.
intermediate [3]. The ketimine can either react with water at the
γ‑carbonyl to form a gem-diol, or be deprotonated on C to give a vi-
β
nylogous amide, absorbing at about 345 nm, which is off the catalytic
pathway. The D132A, D132E and S36A mutations have 3–5-fold slower
rates of formation of this intermediate. D201E mutant Pfkynase is about
[
[
6] F.W. Alexander, E. Sandmeier, P.K. Mehta, P. Christen, Evolutionary relationships
among pyridoxal-5′-phosphate-dependent enzymes: Regiospecific α, β and γ fa-
milies, Eur. J. Biochem. 219 (1994) 953–960.
7] T. Yano, S. Kuramitsu, S. Tanase, Y. Morino, H. Kagamiyama, Role of Asp222 in the
catalytic mechanism of Escherichia coli aspartate aminotransferase: the amino acid
residue which enhances the function of the enzyme-bound coenzyme pyridoxal 5′-
phosphate, Biochemistry 31 (1992) 5878–5887.
3
5-fold slower than wild-type Pfkynase (Table 2). This suggests that the
D201E mutation stabilizes the ketimine structure more than the other
mutations. Formation of the enamine intermediate from the C
β
-C in-
γ
[
[
8] R.S. Phillips, R.K. Dua, Stereochemistry and mechanism of aldol reactions catalyzed
by Kynureninase, J. Am. Chem. Soc. 113 (1991) 7385–7388.
termediate occurs in two steps, first addition of water to form a tetra-
hedral gem-diol(ate) intermediate [8,26], followed by the retro-Claisen
9] R.K. Dua, E.W. Taylor, R.S. Phillips, S-aryl-L-cysteine S,S-dioxides: design and
evaluation of a new class of mechanism based inhibitors of Kynureninase, J. Am.
Chem. Soc. 115 (1993) 1264–1270.
reaction to cleave the C
β
-C bond. It is this hydration of the carbonyl
γ
that is expected to be sensitive to the S36A and Y226F mutations. We
find that each mutation has about a 1000-fold effect on this step with
kynurenine, consistent with an “oxyanion hole” formed by hydrogen
bonding to the substrate carbonyl facilitating the nucleophilic addition
of hydroxide. The effect of the S36A mutation is even larger on the
reaction of β-benzoyl-L-alanine, since it rapidly forms the 345 nm in-
termediate (Fig. 6), yet shows no detectable formation of benzoate, or
the 490 nm aldol product with benzaldehyde. Collapse of the tetra-
[
10] R.S. Phillips, I. Scott, R. Paulose, A. Patel, T.C. Barron, The phosphate of pyridoxal-
5
′-phosphate is an acid/base catalyst in the mechanism of Pseudomonas fluorescens
kynureninase, FEBS J. 281 (2014) 1100–1109.
[11] S. Lima, S. Kumar, V. Gawandi, C. Momany, R.S. Phillips, Crystal structure of the
Homo sapiens kynureninase-3-hydroxyhippuric acid inhibitor complex: insights into
the molecular basis of kynureninase substrate specificity, J. Med. Chem. 52 (2009)
3
89–396.
[12] R.S. Phillips, Reactions of O-acyl-L-serines with tryptophanase, tyrosine phenol-
lyase, and tryptophan synthase, Arch. Biochem. Biophys. 256 (1987) 302–310.
[
13] C. Momany, V. Levdikov, L. Blagova, S. Lima, R.S. Phillips, Three-dimensional
structure of kynureninase from Pseudomonas fluorescens, Biochemistry 43 (2004)
1193–1203.
hedral intermediate results in C
β
-C bond cleavage and formation of a
γ
carboxylic acid and the enamine. This step is affected by the D132A,
D132E, and D201E mutations (Table 2), even though these residues are
not in direct contact with the substrate. Possibly these mutations
change the orientation of the PLP-substrate adduct in the active site and
reduce the rate of breakdown of the gem-diolate. The first product,
anthranilate or benzoate, is then released. The enamine intermediate
can be subsequently trapped with benzaldehyde to give an aldol pro-
[
14] S.V. Koushik, B. Sundararaju, R.A. McGraw, R.S. Phillips, Cloning, sequence, and
expression of kynureninase from Pseudomonas fluorescens, Arch. Biochem. Biophys.
3
44 (1997) 301–308.
[15] H. Liu, J.H. Naismith, An efficient one-step site-directed deletion, insertion, single
and multiple-site plasmid mutagenesis protocol, BMC Biotechnol. 8 (2008) 91.
[
16] F.W. Studier, Protein production by auto-induction in high-density shaking cul-
tures, Prot. Express. Purif. 41 (2005) 207–234.
[17] M. Moriguchi, T. Yamamoto, K. Soda, Properties of crystalline kynureninase from
duct [8,28], or be protonated on C
β
to give a pyruvate ketimine. Finally,
Pseudomonas marginalis, Biochemistry 12 (1973) 2969–2974.
[
[
18] E.A. Peterson, H.A. Sober, Preparation of crystalline phosphorylated derivatives of
the deprotonation of C-4′ of the ketimine and reprotonation at C gives
α
6
vitamin B , J. Am. Chem. Soc. 76 (1954) 169–175.
the external aldimine of L-alanine, which is then released to complete
19] V.B. Gawandi, D. Liskey, S. Lima, R.S. Phillips, Reaction of Pseudomonas fluorescens
kynureninase with beta-benzoyl-L-alanine: detection of a new reaction intermediate
and a change in rate-determining step, Biochemistry 43 (2004) 3230–3237.
20] W.W. Cleland, Statistical analysis of enzyme kinetic data, Methods Enzymol. 63
the catalytic cycle.
[
(
1979) 103–138.
3
.7. Conclusions
[21] I.B.C. Matheson, L.J. Parkhurst, R.J. DeSa, Efficient integration of kinetic differ-
ential equation sets using matrix exponentiation, Methods Enzymol. 384 (2004)
1
8–39.
Asp-201 forms a hydrogen bond to the pyridinium NH of PLP, and
[
[
[
22] M.D. Toney, J.F. Kirsch, Lysine 258 in aspartate aminotransferase: enforcer of the
Circe effect for amino acid substrates and the general-base catalyst for the 1, 3-
prototropic shift, Biochemistry 32 (1993) 1471–1479.
modulates the properties of the PLP at various steps during catalysis.
Asp-132 also influences the reactivity of the PLP, through hydrogen
23] T. Yano, Y. Hinoue, V.J. Chen, D.E. Metzler, I. Miyahara, K. Hirotsu,
H. Kagamiyama, Role of an active site residue analyzed by combination of muta-
genesis and coenzyme analog, J. Mol. Biol. 234 (1993) 1218–1229.
bonding to Asp-201 and modulating the pK of the PLP. Ser-36 plays a
a
role in the hydration of the substrate γ‑carbonyl, as part of an oxyanion
6
24] H. Hayashi, H. Mizuguchi, I. Miyahara, M.M. Islam, H. Ikushiro, Y. Nakajima,
K. Hirotsu, H. Kagamiyama, Strain and catalysis in aspartate aminotransferase,
Biochim. Biophys. Acta, Proteins Proteomics 1647 (2003) 103–109.
hole with Tyr-226, that collectively provide ~10 -fold rate acceleration
to the catalytic mechanism of Pfkynase.
730

R.S. Phillips, et al.
BBA - Proteins and Proteomics 1867 (2019) 722–731
[
[
25] S. Lima, B. Sundararaju, C. Huang, R. Khristoforov, C. Momany, R.S. Phillips, The
crystal structure of the Pseudomonas dacunhae aspartate-β-decarboxylase dode-
camer reveals an unknown oligomeric assembly for a pyridoxal-5′-phosphate-de-
pendent enzyme, J. Mol. Biol. 388 (2009) 98–108.
consequences of perturbing Gly-193 in the oxyanion hole of serine proteases, J.
Biol. Chem. 280 (2005) 25644–25650.
[30] C. Heiss, J. Anderson, R.S. Phillips, Differential effects of bromination on substrates
and inhibitors of kynureninase from Pseudomonas fluorescens, Org. Biomol. Chem. 1
(2003) 288–295.
26] S. Kumar, V.B. Gawandi, N. Capito, R.S. Phillips, Substituent effects on the reaction
of beta-benzoylalanines with Pseudomonas fluorescens kynureninase, Biochemistry
[31] A. Kintanar, C.M. Metzler, D.E. Metzler, R.D. Scott, NMR observation of ex-
changeable protons of pyridoxal phosphate and histidine residues in cytosolic as-
partate aminotransferase, J. Biol. Chem. 266 (1991) 17222–17229.
[32] S. Dajnowicz, J.M. Parks, X. Hu, K. Gesler, A.Y. Kovalevsky, T.C. Mueser, Direct
evidence that an extended hydrogen-bonding network influences activation of
pyridoxal 5′-phosphate in aspartate aminotransferase, J. Biol. Chem. 292 (2017)
5970–5980.
4
9 (2010) 7913–7919.
[
[
27] C.M. Metzler, R. Viswanath, D.E. Metzler, Equilibria and absorption spectra of
tryptophanase, J. Biol. Chem. 266 (1991) 9374–9381.
28] G.S. Bild, J.C. Morris, Detection of β-carbanion formation during kynurenine hy-
drolysis catalyzed by Pseudomonas marginalis kynureninase, Arch. Biochem. Bophys.
2
35 (1984) 41–47.
[
29] K.M. Bobofchak, A.O. Pineda, F.S. Mathews, E. Di Cera, Energetic and structural
731