Determinants of substrate specificity in KdcA, a thiamin diphosphate-dependent decarboxylase

Article abstract of DOI:10.1016/j.bioorg.2006.08.005

Thiamin diphosphate-dependent decarboxylases catalyze the non-oxidative decarboxylation of 2-keto carboxylic acids. Although they display relatively low sequence similarity, and broadly different range of substrates, these enzymes show a common homotetrameric structure. Here we describe a kinetic characterization of the substrate spectrum of a recently identified member of this class, the branched chain 2-keto acid decarboxylase (KdcA) from Lactococcus lactis. In order to understand the structural basis for KdcA substrate recognition we developed a homology model of its structure. Ser286, Phe381, Val461 and Met358 were identified as residues that appeared to shape the substrate binding pocket. Subsequently, site-directed mutagenesis was carried out on these residues with a view to converting KdcA into a pyruvate decarboxylase. The results show that the mutations all lowered the K_m value for pyruvate and both the S286Y and F381W variants also had greatly increased values of k_cat with pyruvate as a substrate.

Full text of DOI:10.1016/j.bioorg.2006.08.005

Bioorganic Chemistry 34 (2006) 325–336
www.elsevier.com/locate/bioorg
Determinants of substrate specificity in KdcA,
a thiamin diphosphate-dependent decarboxylase
Alejandra Yep, George L. Kenyon, Michael J. McLeish *
College of Pharmacy, University of Michigan, 428 Church Street, Ann Arbor, MI 48109-1065, USA
Received 28 July 2006
Available online 9 October 2006
Abstract
Thiamin diphosphate-dependent decarboxylases catalyze the non-oxidative decarboxylation of 2-
keto carboxylic acids. Although they display relatively low sequence similarity, and broadly different
range of substrates, these enzymes show a common homotetrameric structure. Here we describe a
kinetic characterization of the substrate spectrum of a recently identified member of this class, the
branched chain 2-keto acid decarboxylase (KdcA) from Lactococcus lactis. In order to understand
the structural basis for KdcA substrate recognition we developed a homology model of its structure.
Ser286, Phe381, Val461 and Met358 were identified as residues that appeared to shape the substrate
binding pocket. Subsequently, site-directed mutagenesis was carried out on these residues with a
view to converting KdcA into a pyruvate decarboxylase. The results show that the mutations all low-
ered the K
values of kcat with pyruvate as a substrate.
2006 Elsevier Inc. All rights reserved.
m
value for pyruvate and both the S286Y and F381W variants also had greatly increased
ꢀ
Keywords: Benzoylformate; Phenylpyruvate; Pyruvate; 2-Keto acid; Mutagenesis; Homology model
1
. Introduction
Pyruvate decarboxylase from Zymomonas mobilis (ZmPDC) and Saccharomyces
cerevisiae (ScPDC), benzoylformate decarboxylase from Pseudomonas putida (BFD) and
*
Corresponding author. Fax: +1 734 615 3079.
E-mail address: mcleish@umich.edu (M.J. McLeish).
0
045-2068/$ - see front matter ꢀ 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2006.08.005

326
A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
indolepyruvate decarboxylase from Enterobacter cloacae (IPDC) all belong to a group of
enzymes whose main function is the non-oxidative decarboxylation of 2-keto acids [1]. The
2
+
X-ray structures of these enzymes, which utilize Mg and thiamin diphosphate (ThDP) as
cofactor, show that they are structurally similar [2–5]. However, while the PDCs prefer
pyruvate as a substrate and, to a lesser extent, other aliphatic 2-keto acids [6], BFD prefers
aromatic 2-keto acids [7] as does IPDC [8].
One of the intriguing features of these enzymes is that, with the exception of residues
involved in binding ThDP, there is little strict conservation of other active site residues,
at least in terms of sequence identity [4,8,9]. This is somewhat surprising as the differences
between the natural substrates are mostly steric in nature (Fig. 1). However, there does
appear to be some positional conservation of residues. This suggests that the active site
architecture, as well as the chemistry of the cofactor and reaction intermediates, have
played a dominant role in the evolution of enzymes of this type [4]. In our efforts towards
understanding the evolution of substrate specificity in ThDP-dependent enzymes, site-di-
rected mutagenesis was employed in an attempt to interconvert ZmPDC and BFD [10].
Although the interconversion was not particularly successful, a single mutation in either
enzyme was able to bring about significantly enhanced decarboxylase activity with long
(
C5/C6) chain aliphatic keto acids. In fact, the I472A variant of ZmPDC was able to
decarboxylate 2-ketohexanoic acid with a higher kcat/Km value than that of the WT
enzyme with its natural substrate. Further, both this ZmPDC variant and the analogous
A460I variant of BFD were able to decarboxylate branched chain keto acids with kcat/Km
values greater than those of the WT enzymes [10].
Recently, the cloning of KdcA, a branched chain 2-keto acid decarboxylase, was report-
ed [11]. The enzyme appeared to be involved in flavor formation in cheeses brought about
by the actions of Lactococcus lactis, with 3-methyl-2-oxobutanoic acid being the preferred
substrate (Scheme 1). Here we describe the overexpression of KdcA in Escherichia coli and
a kinetic analysis of its substrate specificity. In addition, based on a homology model of
O
O
O
CH3
COOH
COOH
pyruvic acid
COOH
benzoylformic acid
CH3
3-methyl-2-oxobutanoic acid
O
O
CH3
O
COOH
COOH
phenylpyruvic acid
HO
CH3
COOH
COOH
2-oxopentanoic acid
O
4-hydroxyphenylpyruvic acid
CH3
CH3
3-methyl-2-oxopentanoic acid
H
N
CH3
O
CH3
O
O
O
COOH
indole-3-pyruvic acid
COOH
COOH
4-methyl-2-oxopentanoic acid
2-oxohexanoic acid
CH3
CH3
S
COOH 4-methylthio-2-oxo-butanoic acid
Fig. 1. Structures of substrates for the various 2-keto acid decarboxylases.

A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
327
O
O
CH3
CH3
KdcA
COOH
H
+
CO2
CH3
CH3
Scheme 1.
KdcA we have carried out a mutagenic analysis of several residues potentially involved in
substrate binding.
2
. Materials and methods
The plasmid pNZ500 [11], containing the kdcA gene was obtained from Nizo Food
Research. Thiamin diphosphate and the various 2-keto acids were purchased from
Sigma–Aldrich. Buffer salts were purchased from Fisher. All other reagents were of the
highest purity available.
2
.1. Homology modeling of KdcA
The three-dimensional structure of KdcA was modeled with Modeller 8v1 [12] using the
crystal structures of ZmPDC (PDB code 1ZPD), ScPDC complexed with pyruvamide
PDB code 1QPB), and IPDC (PDB code 1OVM), as templates. The initial multiple align-
(
ment was performed with ClustalW [13] and manually refined to include functionally con-
served residues, predicted secondary structures, and hydrophobicity profiles. Secondary
structures were predicted using the PSI-PRED program [14]. Different models of the
monomer were generated and assessed by the VERIFY_3D [15] and PROCHECK pro-
grams [16], and the best model was chosen to model the dimeric and tetrameric forms
of the enzyme. Figures were created with PyMOL (DeLano Scientific, http://
pymol.sourceforge.net/), SwissPDBViewer ([17] www.expasy.org/spdbv/) and Pov-Ray
(
Persistence of Vision Raytracer Pty. Ltd., http://povray.org/). The PDB file for the model
of the KdcA dimer is available as Supplementary material.
2
.2. Amplification and site-directed mutagenesis of kdcA
The coding region of kdcA was amplified by PCR from plasmid pNZ500 [11] using
0
0
the primers 5 -GGAGGAATGaCATGtATACAGTAGGAGATTGGAGATT-3 and
5
0
0
-CATCATCCGTTGATATCTcgagATTTTGCTCA-3 . These primers contained chang-
es from wild-type, denoted by lowercase letters, which introduced BspLU11I and XhoI
restriction sites, respectively. The amplified fragment was gel purified, digested and sub-
cloned into pET-24d (Novagen) previously digested with NcoI and XhoI. This construc-
tion added a 6· His tag to the C-terminus of the enzyme. The fidelity of the
amplification was confirmed by sequencing the resulting plasmid, pET24d-KdcA-His, at
the University of Michigan core facility.
Site-directed mutagenesis was performed using the QuikChange protocol and Pfu
polymerase (Stratagene) using pET24d-KdcA-His as a template. The forward primers
used for the mutagenesis are shown below, with the mutated codons underlined and,
again, lowercase letters indicating base changes from the wild-type gene:

328
A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
0
V461I: 5 -CATAAATAATGATGGTTATACAaTTGAAAGAGAAATCCAtGGAC
0
CTACTC-3
0
0
S286Y: 5 -GCTTGGAGTGAAGtTaACGGACTaCTCAACAGGTGCATTC-3
F381W: 5 -GCTGAACAAGGAACCTCATggTTTGGAGCgTCAACAATTTTC-3
M538W: 5 -GCGCCAAAATTACTGAAAAAAtgGGGTAAgcTtTTTGCTGAGC-3 .
0
0
0
0
In all primers additional silent mutations were introduced, generating new restriction
sites for NcoI, HincII, HindIII and HincII, respectively. These were used for the initial
screening for mutants by restriction analysis. The coding regions of the resulting plasmids
were sequenced to ensure that only the intended mutations had occurred.
2
.3. Protein expression and purification
The wild-type and mutant plasmids pET24d-KdcA-His were transformed into
BL21(DE3)pLysS chemically competent cells (Promega). Cultures (1 L) were grown
until OD600 ꢀ 0.8 and then induced with 1 mM IPTG. After overnight growth at room
temperature, the cells were harvested, resuspended in buffer A (50 mM potassium phos-
phate buffer, pH 8.0, 500 mM NaCl, 10 mM imidazole), disrupted by sonication and
cell debris removed by centrifugation. The cell free extract was applied to a
Ni–NTA column connected to a BioLogic LP system (Bio-Rad) and eluted in buffer
A containing 250 mM imidazole. The fractions of highest purity were pooled and
the buffer was exchanged for storage buffer (100 mM potassium phosphate buffer,
pH 6.0, 1 mM MgSO , 0.5 mM ThDP, and 10% glycerol) using Econo-Pac 10 DG
4
desalting columns (Bio-Rad). The protein samples were concentrated with Amicon
Ultra centrifugal filters (Millipore).
The WT enzyme and all KdcA variants were homogeneous as evaluated by
SDS–PAGE. Protein concentrations were determined with the Bradford assay [18] using
bovine serum albumin as standard.
2
.4. Assay of decarboxylase activity
The decarboxylation of 2-keto acids was measured at 30 ꢁC using a modified coupled
enzymatic assay previously described for benzoylformate decarboxylase activity [7]. The
assay mixture contained 0.2 mM NADH, 0.5 units/mL horse liver alcohol dehydrogenase
and the 2-keto acid in assay buffer (50 mM potassium phosphate buffer, pH 6.0, 1 mM
MgSO , 0.5 mM ThDP) in a total volume of 1 mL. The reaction was initiated by adding
4
0
.5–100 lg of enzyme diluted in storage buffer, and the consumption of NADH was mon-
itored at 340 nm with a Cary 50 Bio spectrophotometer (Varian).
One unit is defined as the amount of enzyme that catalyzes the decarboxylation of
1
lmol of 2-keto acid per minute at 30 ꢁC under standard conditions.
2
.5. Kinetic analysis
Initial velocity data were fitted to the Michaelis–Menten equation using the
program Originꢂ 5.0. Kinetic parameters were determined per monomer using a
molecular mass of 60,884 Da. Additionally, data were fitted to the Hill equation,
and in all cases n ꢀ 1.0.
H

A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
329
3
. Results and discussion
.1. Expression and characterization of KdcA
When first reported, KdcA had been overexpressed in L. lactis and crude extracts were
3
used to determine its pH optimum (ca. 6.0). In addition, its relative activity with a spec-
trum of potential substrates was measured by static headspace gas chromatography
[
11]. The results indicated that 3-methyl-2-oxobutanoic acid, the product of the catabolic
transamination of valine, was the preferred substrate. 4-Methyl-2-oxopentanoic acid and
-methyl-2-oxopentanoic acid, the products of leucine and isoleucine catabolism, respec-
3
tively, were also readily accepted. Overall, the substrate spectrum of KdcA looked very
similar to that of a-ketoisovalerate decarboxylase from L. lactis, with which it shares
almost 90% sequence identity [19]. Of particular interest to us was the observation that
2
-oxopentanoic acid and 2-oxohexanoic acid were also good substrates for KdcA. Previ-
ously, the latter had been shown to be the preferred substrate for the I472A variant of
ZmPDC and both were also substrates for BFD A460I [10]. It was thought that comparing
the three dimensional structures and the substrate specificities of ZmPDC, KdcA, IPDC,
and BFD could provide insight into the evolution of substrate specificity across the ThDP-
dependent decarboxylases.
Accordingly, the kdcA gene was cloned and KdcA expressed as its 6· His tagged var-
iant (hereafter referred to as the wild-type (WT) enzyme) and was assayed with a variety of
substrates (Table 1). In line with the earlier study, the highest kcat values were observed
with the branched chain 2-keto acids. However, the kcat value for phenylpyruvate was
about 60% that of the branched chain derivatives. The kcat value decreased to about 6%
of WT for the tyrosine catabolite, 4-hydroxyphenylpyruvate. In some respect there were
more surprises in the K data than the k data (Table 1). Broadly speaking, the aliphatic
m
cat
2
-keto acids had Km values in the 1–3 mM range. By contrast, phenylpyruvate had a K
m
ꢁ
1 ꢁ1
value of 0.2 mM and, at 129 mM
s
, by far the highest value of kcat/K . It is interesting
m
to note that, although 4-hydroxyphenylpyruvate had the second lowest Km value, its kcat
value was 10-fold lower than that of phenylpyruvate, suggesting that its binding was not
Table 1
Substrate specificity of KdcA
a
ꢁ1
ꢁ1 ꢁ1
(mM s )
Substrate
K
m
(mM)
kcat (s
)
k
cat/K
m
Phenylpyruvic acid
0.21 ± 0.06
0.76 ± 0.05
0.60 ± 0.09
2.8 ± 0.3
3.7 ± 0.3
1.3 ± 0.2
27 ± 3
41 ± 1
13 ± 1
48 ± 2
49 ± 2
129
54
21
17
13
7.8
4.6
3.6
3
2
3
4
2
4
4
-Methyl-2-oxopentanoic acid
-Oxohexanoic acid
-Methyl-2-oxobutanoic acid
-Methyl-2-oxopentanoic acid
-Oxopentanoic acid
9.9 ± 0.4
2.9 ± 0.1
8.7 ± 1.3
7.3 ± 0.2
-Hydroxyphenylpyruvic acid
-Methylthio-2-oxobutanoic acid
0.63 ± 0.03
2.4 ± 0.9
Benzoylformic acid
Pyruvic acid
7.5 ± 0.4
0.97
0.085
b
b
c
n.d.
n.d.
a
Results are the average ± standard deviation of three independent experiments.
Not determined.
Measured under V/K conditions (i.e., [S] > K ).
m
b
c

330
A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
optimal for catalysis. Overall, pyruvate and benzoylformate, the substrates of ZmPDC
and BFD, respectively, had the lowest values of kcat/K . While some activity was found
m
with pyruvate, the levels were very low and saturation could not be reached even at sub-
strate concentrations as high as 70 mM. Therefore kcat/K , the value of the apparent sec-
m
ond order rate constant for reaction with pyruvate, was determined from plots of reaction
rate versus substrate concentration under conditions such that the pyruvate concentration
was much lower than its Km value. Conversely, the kcat value for benzoylformate was in
the range of the slower aliphatic substrates, but its value of Km was the highest of those
measured.
Many PDCs, including ScPDC, display the sigmoidal kinetics indicative of substrate acti-
vation [20–22]. KdcA showed no evidence for substrate activation, and fitting kinetic data to
the Hill equation provided n values ꢀ1.0 (data not shown). In this respect KdcA was similar
H
to ZmPDC, IPDC and BFD all of which obey Michaelis–Menten kinetics [7,8,23].
3
.2. Homology modeling
A homology model of KdcA was built based on the solved crystal structures of three
other ThDP-dependent decarboxylases. KdcA shares 36%, 31%, and 41% sequence iden-
tity with ZmPDC, ScPDC and IPDC, respectively. Modeling based on templates with
>40% sequence identity is generally guaranteed to be successful, but in the case of lower
identities, errors can be significantly reduced by employing an accurate sequence align-
ment [12,24,25]. In this case, the structure of IPDC was chosen as the initial template
because it had the highest sequence identity with KdcA. The structure of ZmPDC was also
included as a template because, despite its lower sequence identity with KdcA, it improved
the quality of the model in some specific regions. In particular, ZmPDC was used to model
the loop between residues 340–360, which aligns with a disordered area in the IPDC crys-
tal structure (residues 342–355) (Fig. 2). Later the ScPDC structure was also incorporated
Fig. 2. Alignment of ThDP-dependent 2-keto acid decarboxylases. The residues studied in this work are indicated
by black triangles. The boxes designate residues involved in catalysis and cofactor binding. The disordered region
in the IPDC structure is indicated by a gray line.

A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
331
in the building of the model because, unlike the structures of IPDC and ZmPDC, it has an
inhibitor (pyruvamide) bound in the active site in addition to ThDP.
Using Modeller 8v2, after iterative manual refinements of the alignment to accommo-
date insertions and deletions of the query sequence respect to the templates in the optimal
manner, 32 models of the monomer were generated. Verify_3D and PROCHECK were
used to evaluate the models, and to identify the one with the best structural compatibility.
The final model of KdcA, which was used in structural analysis, is dimeric and includes
both ThDP and pyruvamide.
The model of the monomer (Fig. 3A) is compatible with the a/b topology observed in
all ThDP-dependent decarboxylases crystallized to date [26]. Each monomer comprises the
N-terminal (Pyr) domain, a middle domain and the C-terminal (PP) domain. Two mono-
mers are required to form the active site, which includes contributions from the Pyr and
PP domains of each monomer. These bind the pyrimidine ring and the diphosphate tail
of ThDP, respectively [26]. The region of the active site comprises the ThDP binding motif
(
Fig. 2), the catalytic glutamic acid residue (Glu49) and the hydrophobic residue (Ile404)
which holds the cofactor in its typical ‘‘V’’-conformation. There is also a pair of catalytic
histidine residues. This pair is formed by two contiguous histidines in IPDC and the PDCs,
but in BFD the two histidines come from loops in two different monomers. In KdcA, the
two histidines are contiguous (Fig. 3B) suggesting that KdcA has evolved from a PDC-like
precursor. Conceivably, it should be possible to alter the binding site of KdcA sufficiently
for it to accept pyruvate as a reasonable substrate.
A
B
Middle domain
ThDP
His281 (BFD)
His70 (BFD)
His115 (IPDC)
His113 (ZmPDC)
His112 (KdcA)
His116 (IPDC)
His114 (ZmPDC)
His113 (KdcA)
Pyr domain
PP domain
Fig. 3. Structural model of Lactococcus lactis KdcA. (A) Cartoon representation of the monomer highlighting the
three domain structure. The molecules of ThDP and pyruvamide are included as ball-and-stick models. (B)
Superimposition of active site histidines from BFD (green), ZmPDC (gold), IPDC (blue) and KdcA (gray). The
figure was obtained using the coordinates from 1BFD, 1ZPD, 1OVM and the homology model of KdcA. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

332
A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
Ala460 (BFD)
Ile472 (ZmPDC)
Val461 (KdcA)
Phe464 (BFD)
Ile476 (ZmPDC)
Ile465 (KdcA)
ThDP
Fig. 4. Putative substrate binding residues of ZmPDC (green), BFD (gold), and KdcA (gray). The corresponding
residues in ZmPDC and BFD were previously studied by site-directed mutagenesis [10]. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
According to this model, Val461 in KdcA is structurally equivalent to Ala460 in BFD
and Ile472 in ZmPDC (Fig. 4). This position has been studied in the two latter enzymes,
and it has been shown that swapping this residue generates PDC and BFD enzymatic vari-
ants with increased ability to decarboxylate branched chain 2-keto acids [10]. Therefore,
reverse engineering would suggest that the V461I mutation should result in an increase
in affinity for pyruvate.
The fact that ZmPDC has a much more limited substrate spectrum than ScPDC is well
established [27]. In the former, the bulky aromatic residues, Tyr290, Trp392, and Trp551,
are thought to restrict the size of the binding pocket and thereby prevent the binding of larger
substrates [2]. In a multiple alignment of plant and bacterial PDCs aromatic side chains in the
equivalent positions are conserved (data not shown). The exceptions are the yeast PDCs
which have smaller residues in those positions. In ScPDC, for example, Trp392 is replaced
by alanine and, when the W392A variant of ZmPDC was prepared, it was found to have
an expanded range of activity [27,28]. In the IPDC structure the corresponding residues
are Thr290, Ala387, and Leu542. In toto these substitutions account for the difference in vol-
3
˚
ume of the active site cavity observed between the ZmPDC and IPDC structures (130 A in
3
˚
IPDC compared to 85 A in ZmPDC, [5]). The structural homologues in the KdcA model are
Ser286, Phe381, and Met538 (Fig. 5). Each of these residues is smaller than its ZmPDC coun-
terpart and it was thought that conversion of the individual residues, in turn, to its ZmPDC
analogue, would decrease the volume of the active site cavity of KdcA and may provide a
variant with increased pyruvate decarboxylase activity.
3
.3. Site-directed mutagenesis
The KdcA variants, V461I, S286Y, F381W and M538W, were constructed, expressed
and purified to apparent homogeneity as described under Section 2. During purification

A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
333
Ala387 (IPDC)
Trp392 (ZmPDC)
Phe381 (KdcA)
Leu542 (IPDC)
Trp551(ZmPDC)
Met538 (KdcA)
ThDP
Pyruvamide
Thr290 (IPDC)
Tyr290 (ZmPDC)
Ser286 (KdcA)
Fig. 5. Residues in the putative substrate binding pocket of ZmPDC (gold), IPDC (blue), and KdcA (gray). The
molecule of pyruvamide is an indicator of the putative position of the KdcA substrate. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
the four mutant enzymes showed physical properties essentially identical to those of WT
KdcA. However, all of the variants showed an increase in their ability to decarboxylate
pyruvate (Table 2).
Pyruvate is not a good substrate for WT KdcA and, while there is evidence for
decarboxylase activity, the enzyme is not saturated even at high pyruvate concentra-
tions. By contrast all the mutants showed saturation kinetics, with Km values between
2
0–65 mM. Perhaps more impressive was the increase in k values, particularly for
cat
ꢁ
1
ꢁ1
the S286Y and F381W variants which had values of 17 s and 26 s , respectively.
To put this into context, the kcat value for WT KdcA with phenylpyruvate, its opti-
ꢁ
1
mum substrate based on kcat/Km values, was 27 s . The results show clearly that,
while there is room for an improvement in binding affinity for pyruvate, the trend
is in the right direction.
Presumably, if the KdcA substrate binding site has been made smaller, this should be
reflected in decreased affinity for the larger substrates. The results in Table 2, in general
confirm this hypothesis. This was particularly evident for the S286Y variant which exhib-
ited a relatively low K value for pyruvate but showed 25- and 30-fold increases in K val-
m
m
ues for 3-methyl-2-oxopentanoic acid and phenylpyruvate, respectively. There were also
some anomalies: the M538W variant, for example, had the lowest K value for pyruvate.
m
It also showed 6-fold increases in Km values for 3-methyl-2-oxopentanoic acid and ben-
zoylformate, yet had a lower Km value for phenylpyruvate than that of WT KdcA. The
F381W variant was also surprising, in that it had the highest kcat value for pyruvate
but, essentially, an unchanged value for phenylpyruvate. It also showed a 4.5-fold increase
in kcat for 3-methyl-2-oxopentanoic acid. The reasons for the anomalous results remain
unclear.

Table 2
a,b
Kinetic constants for decarboxylase activity of KdcA variants
KdcA variant
3-Methyl-2-oxopentanoic acid
Pyruvic acid
Benzoylformic acid
Phenylpyruvic acid
k
cat
ꢁ
K
m
k
cat/K
m
k
cat
ꢁ1
K
m
k
cat/K
m
k
cat
ꢁ1
K
m
k
cat/K
m
k
cat
ꢁ1
K
m
k
cat/K
m
1
ꢁ1 ꢁ1
ꢁ1 ꢁ1
ꢁ1 ꢁ1
ꢁ1 ꢁ1
(s
)
(mM)
(mM
s
)
(s
)
(mM)
(mM
s
)
(s
)
(mM)
(mM
s
)
(s
)
(mM)
(mM
s
)
WT
40
14
35
177
12
0.76
1.1
19
1.7
5.5
53
13
1.8
104
ND
ND
18
34
65
22
0.085
0.13
0.5
0.4
0.11
7.3
3.1
2.1
9.4
4.7
7.5
2.2
6.9
6.2
46
1.0
1.4
0.3
1.5
0.1
27
0.21
2.3
6.3
0.17
0.1
128
1.3
0.4
117
7.0
V461I
S286Y
F381W
M538W
a
2.3
17
26
2.5
2.9
2.5
20
0.7
2
Reactions were carried out at 30 ꢁC as described under Experimental Procedures.
Errors are estimated to be ± 10% from triplicate measurements.
b

A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
335
Overall, the mutagenesis results confirmed predictions based on the KdcA homology
model, i.e., that Val461, Ser286, Phe381, and Met538 each contributes to substrate binding
and catalysis by KdcA. However, the results also showed that mutation of a single residue
was unlikely to provide an ‘‘instant’’ conversion to KdcA into a PDC. Therefore, in pre-
liminary experiments, combinations of double mutants as well as the triple mutant were
generated for the S286Y, F381W, and M538W variants. Unfortunately, these mutant
enzymes had impaired solubility and stability, and preliminary experiments suggest that
they have extremely low activity towards the substrates tested in this work (data not
shown). We have now turned our attention towards saturation mutagenesis to identify
candidates with improved PDC activity.
Acknowledgments
This paper is dedicated to the memory of Miriam S. Hasson, a collaborator and friend,
whose presence is sorely missed. The work was supported by a Frontiers in Integrative
Biology Research Grant (EF 0425719) from the National Science Foundation (M.J.M.
and G.L.K.).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, at
doi:10.1016/j.bioorg.2006.08.005.
References
[
[
[
1] A. Schellenberger, Biochim. Biophys. Acta 1385 (1998) 177–186.
2] D. Dobritzsch, S. Konig, G. Schneider, G. Lu, J. Biol. Chem. 273 (1998) 20196–20204.
3] P. Arjunan, T. Umland, F. Dyda, S. Swaminathan, W. Furey, M. Sax, B. Farrenkopf, Y. Gao, D. Zhang, F.
Jordan, J. Mol. Biol. 256 (1996) 590–600.
[
[
4] M.S. Hasson, A. Muscate, M.J. McLeish, L.S. Polovnikova, J.A. Gerlt, G.L. Kenyon, G.A. Petsko, D.
Ringe, Biochemistry 37 (1998) 9918–9930.
5] A. Sch u¨ tz, T. Sandalova, S. Ricagno, G. H u¨ bner, S. K o¨ nig, G. Schneider, Eur. J. Biochem. 270 (2003) 2312–
2321.
[
[
[
6] M. Pohl, P. Siegert, K. Mesch, H. Bruhn, J. Grotzinger, Eur. J. Biochem. 257 (1998) 538–546.
7] P.M. Weiss, G.A. Garcia, G.L. Kenyon, W.W. Cleland, P.F. Cook, Biochemistry 27 (1988) 2197–2205.
8] A. Sch u¨ tz, R. Golbik, K. Tittmann, D.I. Svergun, M.H. Koch, G. H u¨ bner, S. K o¨ nig, Eur. J. Biochem. 270
(2003) 2322–2331.
[
9] M.J. McLeish, G.L. Kenyon, E.S. Polovnikova, A.S. Bera, N.L. Anderson, M.S. Hasson, in: F. Jordan,
M.S. Patel (Eds.), Catalytic mechanisms and role in normal and disease states, Marcel Dekker Inc., New
York, 2004, pp. 131–141.
[
[
[
[
10] P. Siegert, M.J. McLeish, M. Baumann, M.M. Kneen, G.L. Kenyon, M. Pohl, Prot. Eng. Des. Sel. 18 (2005)
45–357.
11] B.A. Smit, J.E.T. van Hylckama Vlieg, W.J.M. Engels, L. Meijer, J.T.M. Wouters, G. Smit, Appl. Environ.
Microbiol. 71 (2005) 303–311.
12] R. Sanchez, A. Sali, in: D.M. Webster (Ed.), Protein structure prediction, methods and protocols, Humana
Press, New Jersey, 2000, pp. 97–129.
13] D. Higgins, J. Thompson, T. Gibson, J.D. Thompson, D.G. Higgins, T.J. Gibson, Nucleic Acids Res. 22
3
(1994) 4673–4680.
[
[
[
14] L.J. McGuffin, K. Bryson, D.T. Jones, Bioinformatics 16 (2000) 404–405.
15] R. Luthy, J.U. Bowie, D. Eisenberg, Nature 356 (1992) 83–85.
16] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Crystallogr. 26 (1993) 283–291.

336
A. Yep et al. / Bioorganic Chemistry 34 (2006) 325–336
[
[
[
[
[
[
[
[
[
[
[
[
17] N. Guex, M.C. Peitsch, Electrophoresis. 18 (1997) 2714–2723.
18] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
19] M. de la Plaza, P. Fernandez de Palencia, C. Pelaez, T. Requena, FEMS Microbiol. Lett. 238 (2004) 367–374.
20] G. H u¨ bner, R. Weidhase, A. Schellenberger, Eur. J. Biochem. 92 (1978) 175–181.
21] F. Krieger, M. Spinka, R. Golbik, G. H u¨ bner, S. K o¨ nig, Eur. J. Biochem. 269 (2002) 3256–3563.
22] Q. Wang, P. He, D. Lu, A. Shen, N. Jiang, J. Biochem. (Tokyo) 136 (2004) 447–455.
23] S. Bringer-Meyer, K.L. Schimz, H. Sahm, Arch. Microbiol. 146 (1986) 105–110.
24] R. Sanchez, A. Sali, Curr. Opin. Struc. Biol. 7 (1997) 206–214.
25] A. Tramontano, Methods 14 (1998) 293–300.
26] R. Duggleby, Acc. Chem. Res. in press.
27] H. Iding, P. Siegert, K. Mesch, M. Pohl, Biochim. Biophys. Acta 1385 (1998) 307–322.
28] H. Bruhn, M. Pohl, J. Gr o¨ tzinger, M.-R. Kula, Eur. J. Biochem. 234 (1995) 650–655.