Rational design of stereoselectivity in the class II pyruvate aldolase BphI

Article abstract of DOI:10.1021/ja208754r

BphI, a pyruvate-specific class II aldolase, catalyzes the reversible carbon-carbon bond formation of 4-hydroxy-2-oxoacids up to eight carbons in length. During the aldol addition catalyzed by BphI, the S-configured stereogenic center at C4 is created via attack of a pyruvate enolate intermediate on the si face of the aldehyde carbonyl of acetaldehyde to form 4(S)-hydroxy-2-oxopentanoate. Replacement of a Leu-87 residue within the active site of the enzyme with polar asparagine and bulky tryptophan led to enzymes with no detectable aldolase activity. These variants retained decarboxylase activity for the smaller oxaloacetate substrate, which is not inhibited by excess 4-hydroxy-2-oxopentanoate, confirming the results from molecular modeling that Leu-87 interacts with the C4-methyl of 4(S)-hydroxy-2-oxoacids. Double variants L87N;Y290F and L87W;Y290F were constructed to enable the binding of 4(R)-hydroxy-2-oxoacids by relieving the steric hindrance between the 5-methyl group of these compounds and the hydroxyl substituent on the phenyl ring of Tyr-290. The resultant enzymes were shown to exclusively utilize only 4(R)- and not 4(S)-hydroxy-2-oxopentanoate as the substrate. Polarimetric analysis confirmed that the double variants are able to synthesize 4-hydroxy-2-oxoacids up to eight carbons in length, which were the opposite stereoisomer compared to those produced by the wild-type enzyme. Overall the k_cat/K _m values for pyruvate and aldehydes in the aldol addition reactions were affected 10-fold in the double variants relative to the wild-type enzyme. Thus, stereocomplementary class II pyruvate aldolases are now available to create chiral 4-hydroxy-2-oxoacid skeletons as synthons for organic reactions.

Full text of DOI:10.1021/ja208754r

ARTICLE
pubs.acs.org/JACS
Rational Design of Stereoselectivity in the Class II Pyruvate
Aldolase BphI
Perrin Baker and Stephen Y. K. Seah*
Department of Molecular and Cellular Biology, University of Guelph, Guelph, Ontario, Canada N1G 2W1
ABSTRACT: BphI, a pyruvate-specific class II aldolase, catalyzes the
reversible carbonÀcarbon bond formation of 4-hydroxy-2-oxoacids up to
eight carbons in length. During the aldol addition catalyzed by BphI, the
S-configured stereogenic center at C4 is created via attack of a pyruvate
enolate intermediate on the si face of the aldehyde carbonyl of acetaldehyde
to form 4(S)-hydroxy-2-oxopentanoate. Replacement of a Leu-87 residue
within the active site of the enzyme with polar asparagine and bulky
tryptophan led to enzymes with no detectable aldolase activity. These
variants retained decarboxylase activity for the smaller oxaloacetate sub-
strate, which is not inhibited by excess 4-hydroxy-2-oxopentanoate, con-
firming the results from molecular modeling that Leu-87 interacts with the
C4-methyl of 4(S)-hydroxy-2-oxoacids. Double variants L87N;Y290F and
L87W;Y290F were constructed to enable the binding of 4(R)-hydroxy-2-oxoacids by relieving the steric hindrance between the
-methyl group of these compounds and the hydroxyl substituent on the phenyl ring of Tyr-290. The resultant enzymes were shown
5
to exclusively utilize only 4(R)- and not 4(S)-hydroxy-2-oxopentanoate as the substrate. Polarimetric analysis confirmed that the
double variants are able to synthesize 4-hydroxy-2-oxoacids up to eight carbons in length, which were the opposite stereoisomer
compared to those produced by the wild-type enzyme. Overall the k /K values for pyruvate and aldehydes in the aldol addition
cat
m
reactions were affected e10-fold in the double variants relative to the wild-type enzyme. Thus, stereocomplementary class II
pyruvate aldolases are now available to create chiral 4-hydroxy-2-oxoacid skeletons as synthons for organic reactions.
1
2,13
’
INTRODUCTION
of 4-hydroxy-2-oxoacids to pyruvate and an aldehyde.
The
enzyme can also be utilized in the reverse to catalyze the reverse
aldol addition reaction. BphI exhibits strict stereochemical con-
trol as it can only utilize or produce the 4(S) enantiomer of
Approximately 60% of all pharmaceutical agents currently on
the market are chiral, and the development of enantiomerically
pure formulations is often critical to ensure high specificity for the
14,15
1À4
4-hydroxy-2-oxopentanoate.
The catalytic and kinetic mechanisms
intended biological target.
Aldolases have the potential to
of the enzyme have recently been elucidated. In the aldol addition
direction, proton abstraction from the C3 methyl generates a pyruvate
enolate that is in turn stabilized by Arg-16 and the divalent metal
catalyze the formation of CÀC bonds with stereochemical control
and are therefore powerful biocatalysts for the synthesis of
5
pharmaceutical compounds, including enzyme inhibitors and
15
6
cofactor. The enolate attacks an aldehyde with si-facial selectivity,
leading to carbonÀcarbon bond formation. His-20 serves as the
acid catalyst that protonates the aldehydic oxygen during this
attack to form 4-hydroxy-2-oxoacid. The enzyme has been
demonstrated to exhibit broad substrate specificity for linear
aldehyde moieties up to six carbons in length. Three amino acid
residues, Leu-87, Leu-89, and Tyr-290, located in the active site
complex oligosaccharides. Among aldolases, the dihydroxyace-
tone phosphate (DHAP) dependent aldolase, which possesses a
strict requirement for DHAP as the nucleophilic component and
can accept virtually any aldehyde as an electrophile, has been
7À9
extensively investigated for organic synthesis.
However, since
DHAP-dependent enzymes produce a 1,3,4-trihydroxy-2-butanone
skeleton that is already highly functionalized and requires
the expensive DHAP precursor, there has been emphasis in recent
years on the development and discovery of other aldolases that
have alternative substrate specificities. One such group is pyruvate
aldolases that catalyze the aldol addition of cheaply available
precursors, pyruvate and aldehyde, to produce a 4-hydroxy-2-
oxoacid skeleton (Scheme 1) that can be further modified
chemically or enzymatically to produce amino- or hydroxyl-
15
have been implicated in substrate recognition. Leu-87 interacts
with the C4-methyl of the substrate as substitution to a shorter
alanine side chain led to a >30-fold increase in K for acetalde-
m
hyde in the aldol addition reaction. Leu-89 governs chain length
specificity of the aldehyde moiety of the substrate, and when
substituted with an alanine, the enzyme can effectively utilize
aldehydes up to five carbons in length with catalytic efficiency
15
10,11
equal to that of the native acetaldehyde. Tyr-290 functions to
substituted derivatives as chiral synthons for organic reactions.
BphI is a metal-dependent, class II pyruvate aldolase found in
the polychorinated biphenyl (PCB) degradation pathway of
Burkholderia xenovorans that catalyzes the retro-aldol cleavage
Received: September 16, 2011
Published: November 14, 2011
r 2011 American Chemical Society
507
dx.doi.org/10.1021/ja208754r
|J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
ARTICLE
a
Scheme 1. General Mechanism of Pyruvate Class II Aldolases
a
The pyruvyl moiety of the substrate coordinates the divalent metal ion via its C1-carboxylate and C2-keto oxygens. A catalytic base, denoted as B,
abstracts a proton from the C4-OH group of the substrate. This leads to CÀC bond cleavage between C3 and C4, producing an aldehyde and a pyruvate
enolate anion. A catalytic acid, denoted as AH, donates a proton to the pyruvate enolate to produce pyruvate.
position a water molecule that is proposed, on the basis of kinetic
and solvent isotope experiments, to abstract a proton from pyruvate
to form the pyruvate enolate intermediate. Y290F and Y290S
chemicals were analytical grade and were obtained from Sigma-Aldrich
and Fisher Scientific (Nepean, ON, Canada). 4(S)- and 4(R)-hydroxy-2-
oxopentanoate and racemic 4-hydroxy-2-oxoacids with aldehyde sub-
stituents up to five carbons in length were synthesized and purified as
15
variants resulted in a 10-fold decrease in K and k for the retro-
m
cat
1
4
previously described.
DNA Manipulation. DNA was purified, digested, and ligated using
aldol cleavage of 4-hydroxy-2-oxopentanoate. Surprisingly, the var-
iants also lost stereochemical control and were able to catalyze aldol
cleavage reactions of both 4(S) and 4(R) enantiomers of the
2
1
standard protocols. bphI was previously cloned into plasmid pBTL4-
1
2
15
T7. Mutations were incorporated according to a modified Quikchange
4
-hydroxy-2-oxopentanoate with similar catalytic efficiencies. This
22
(
Stratagene) method that uses nonoverlapping primers. The se-
is consistent with the proposal that the p-hydroxyl of Tyr-290 in the
wild-type enzyme also sterically restricts the binding of the 4(R)
enantiomer and its removal in the variants enables the enzyme to
utilize this enantiomer. While many class I and class II pyruvate
aldolases exist, the class I 2-keto-3-deoxy-6-phosphogalactonate
quences of the sense oligonucleotides used to construct the L87N and
L87W variants were GTCAGTGCCAATCTCTTGCCCGGCATCG-
GCACC and GTCAGTGCCTGGCTCTTGCCCGGCATCGGCACC,
respectively, with altered codons underlined. Site-directed mutagenesis to
replace Tyr-290 was carried out according to the megaprimer method
(KDPGal) aldolase is the only known member that appears to be
1
5,23
using previously generated primers.
plasmid DNA, 0.5 unit of Pfu polymerase, a 20 nM concentration of each
, and a 2 μM concentration of each
PCR reactions contained 10 ng of
stereospecific for the R enantiomer of its natural substrate. KDPGal
aldolase can catalyze aldol addition using a broad range of electro-
philic substrates, provided they incorporate polar functionality at C2,
dNTP, 1Â Pfu buffer, 1 mM MgSO
4
16
primer in a total volume of 50 μL. The following amplification profile was
followed: 95 °C for 1 min, followed by 15 cycles of 95 °C for 1 min, 45 °C
for 1 min, and 72 °C for 6 min with a final extension of 72 °C for 10 min.
Genes from positive clones were sequenced at the Guelph Molecular
Supercenter, University of Guelph.
C3, or C4. However, the enzyme is not efficient with longer
aldehydes as electrophiles, and therefore, it is desirable to engineer
other pyruvate aldolases, such as BphI, to create R-specific pre-
cursors of some compounds, such as 4-hydroxy-2-oxooctanoate for
17,18
the antifungal syringomycin.
Expression and Purification of BphI. Variants of BphI were
expressed in recombinant Escherichia coli BL21 (λDE3) cells and were
Although the structure of BphI is not available, the structure of
its orthologue, DmpG (56% sequence similarity to BphI), from
the phenol degradation pathway of Pseudomonas putida CF600,
12
expressed and purified as previously described. The concentration of
purified protein was determined by the Bradford assay using bovine
19
has been solved by X-ray crystallography (PDB 1NVM). This
structure only contains bound oxalate, a pyruvate enolate ana-
logue, but not the full substrate with the aldehyde moiety. Here
we explore the use of molecular modeling and prior knowledge
24
serum albumin as the standard.
Enzyme Assays. Assays were performed in both the cleavage
(retro-aldol) direction and the aldol addition reaction. All assays were
performed at 25 °C, in a total volume of 1 mL of 100 mM 4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, pH 8.0,
15
from the biochemical analysis of BphI to rationally design the
class II aldolase BphI to exclusively utilize or produce the
opposing 4(R) enantiomer. Our results demonstrate that rational
design to change the stereospecificity of an enzyme is possible
without the availability of detailed experimentally determined
enzymeÀsubstrate structural information.
containing 1 mM MnCl
2
. The retro-aldol and oxaloacetate decarbox-
ylase assays were performed by coupling pyruvate formation with
12
NADH oxidation using lactate dehydrogenase as previously described.
One unit of enzyme represents the amount of protein that produces
1 μmol of pyruvate from the substrate in 1 min.
Aldol addition assays contained varying concentrations of pyruvate or
aldehydes and were initiated by the addition of 100 μg of enzyme.
Following incubation for 60 min at 25 °C, the reaction was quenched
with 200 μL of concentrated HCl and incubated overnight at 25 °C to
lactonize the product. A 500 μL aliquot from each reaction was subjected
’
EXPERIMENTAL PROCEDURES
Chemicals. Sodium pyruvate, all aldehydes, L-lactate dehydrogenase
(LDH; rabbit muscle), alcohol dehydrogenase (Saccharomyces cerevisiae),
€
and Chelex 100 were from Sigma-Aldrich (Oakville, ON, Canada).
Restriction enzymes, T4 DNA ligase, and Pfu polymerase were from
Fermentas (Mississauga, ON, Canada) or New England Biolabs
to HPLC using an AKTA Explorer 100 apparatus (Amersham Pharmacia
Biotech, Baie d’Urf ꢀe , QC, Canada) equipped with an Aminex fast acid
analysis ion exchange column (100 mm  7.8 mm) with a mobile phase
of 0.1% formic acid at a flow rate of 0.6 mL/min. Products were detected
at a wavelength of 215 nm and quantified using a standard curve of pure
compounds of known concentrations. Data were fitted by nonlinear
(
Pickering, ON, Canada). NiÀnitrilotriacetic acid (NTA) Superflow
resin was obtained from Qiagen (Mississauga, ON). Recombinant HpaI
was purified according to a previously reported method. All other
20
5
08
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
ARTICLE
Figure 1. Active site of DmpG with modeled substrates. (A) The 4-methyl substituent of 4(S)-HOPA is directed toward Leu-87 at the active site of the
wild-type enzyme. (B) 4(R)-HOPA can be accommodated in the active site if Tyr-290 is replaced with phenylalanine. Numbers in parentheses
4
3
correspond to the residue numbers in BphI. Images were generated using PyMOL.
Table 1. Steady-State Kinetic Parameters of BphI and Its Variants for the Aldol Cleavage of the Stereoisomers of 4-Hydroxy-2-
a
oxopentanoate
4
(S)-HOPA
4(R)-HOPA
À2 À1
cat (10
2
À1 À1
À2 À1
2
À1 À1
enzyme
K
m,app (μM)
k
s
)
k
cat/Km,app (10 M
s
)
Km,app (μM)
kcat (10
s
)
k
cat/Km,app (10 M
s
)
3
3
WT
Y290F
89 ( 8
13 ( 1
NA
400 ( 0.7
13.2 ( 0.2
447 ( 40
126 ( 9
NA
NA
NA
NA
b
13 ( 1
757 ( 89
435 ( 42
13.1 ( 0.3
9.8 ( 0.5
4.3 ( 0.2
104 ( 6
1.3 ( 0.06
1.0 ( 0.1
c
L87N;Y290F
L87W;Y290F
NA
c
NA
NA
NA
a
Aldol cleavage activity was completed with 4(S)- and 4(R)-HOPA. Assays were performed at 25 °C and contained 0.4 mM NADH, 1 mM MnCl , and
2
b
c
À1
1
9.2 units of LDH in 100 mM HEPES (pH 8.0) in a total volume of 1 mL. NA denotes no detectable activity. Data from ref 15. Activity is <0.0001 s .
regression to the MichaelisÀMenten equation using the program
into the crystal structure of DmpG, an orthologue of BphI, by
superimposing the pyruvyl moiety to the position of the experi-
mentally determined oxalate within the active site of the
25
Leonora. In situations where the enzyme was unable to be saturated
because of high K values, the apparent second-order rate constants at
low substrate concentrations were estimated from the linear gradient of
m
15
enzyme. In this position the C1-carboxyl and C2-carbonyl
26,27
15
velocity versus substrate concentration graphs.
oxygens of HOPA coordinate the divalent metal cofactor. The
Determination of Stereospecificity. The stereoselectivity of the
wild type and variants were assessed through the degradation of racemic
mixtures of 4-hydroxy-2-oxoacids with aldehyde substituents up to five
carbons in length. Assays contained100 μM racemic4-hydroxy-2-oxoacid,
van der Waals radius of the hydroxyl oxygen of the Tyr-291
residue (corresponds to Tyr-290 in BphI) overlaps with the C4
carbon of HOPA, necessitating a clockwise rotation of the phenyl
ring of Tyr-291 by 14.5° to relieve this steric clash. Even in the
alternative confirmation of Tyr-291, there is insufficient space
proximal to this residue to accommodate the 4-methyl substi-
tuent of the R enantiomer of HOPA. The C4-methyl of 4(S)-
HOPA, on the other hand, orients toward Leu-88 (Leu-87 in
BphI) in a space opposite Tyr-291 (Figure 1A).
To engineer the enzyme to utilize the 4(R) enantiomer, the
phenolic oxygen was removed in BphI by substitution of Tyr-290
with phenylalanine, thus creating more space for the 4(R)-methyl
substituent of the substrate (Figure 1B). Replacements of Leu-87
with the polar asparagine (L87N) and the bulky tryptophan (L87W)
residue were then attempted to create unfavorable polarÀhydropho-
bic interactions or to introduce steric constraints to prevent binding
of the 4(S) enantiomer.
10 μg for the wild-type BphI or 500 μg for the BphI variants, 0.4 mM
NADH, 1 mM MnCl , and 19.2 units of LDH in 100 mM HEPES buffer,
2
pH 8.0. Reactions were measured continuously at 340 nm until comple-
tion (e120 min). Both the wild type and variants were also used to
synthesize the 4-hydroxy-2-oxoacids in assays containing 2 M aldehyde,
100 mM pyruvate, and 4 mg of the wild-type BphI or variants in 100 mM
HEPES, pH 8.0. After 3 h, reactions were quenched by the addition of
2+
Chelex to remove the Mn cofactor and subsequently filtered through a
YM10 filter (Millipore) to remove the enzyme. Samples were lyophilized
and resuspended in 4 mL of water. The pH was adjusted to 2.0 using HCl,
allowing 4-hydroxy-2-oxoacids to lactonize overnight. Samples were
loaded on an Aminex HPX-87H column (300 Â 7.8 mm) with a mobile
phase of 0.1% formic acid at a flow rate of 0.6 mL/min. Products were
detected at 215 mm, and retention times were consistent with products
14
previously confirmed by LCÀMS, collected, and lypholized. Synthe-
sized substrates were resuspended in 100 mM HEPES, pH 8.0, and the
optical rotation of purified substrates was determined using a Rudolph
Autopol IV automatic polarimeter at 25 °C using a wavelength of 589 nm.
Single variants (L87N, L87W) and double variants (L87N;
Y290F and L87W;Y290F) were successfully created by site-
specific mutagenesis, and they could be purified to homogeneity
with yields similar to those of the wild-type enzyme.
Stereospecificity for HOPA in the Retro-Aldol Reaction.
Variants were initially tested for their ability to catalyze the retro-
aldol cleavage of 4(S)- or 4(R)-HOPA. The double variants
L87N;Y290F and L87W;Y290F were inactive toward 4(S)-
HOPAbut were able to catalyze thealdol cleavage of theopposing
’
RESULTS
Molecular Modeling and Creation of Enzyme Variants.
-Hydroxy-2-oxopentanoate (HOPA) was previously modeled
4
5
09
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
ARTICLE
Table 2. Steady-State Kinetic Parameters of BphI and Its
Variants for Oxaloacetate Decarboxylation
Table 3. Degradation of Racemic Mixtures of 4-Hydroxy-2-
oxoacids
a
a
À1 À1
cat (10
À1 À1
enzyme
K
m,app (mM)
k
s
)
kcat/ Km,app (M
s
)
3
degradation of racemic mixtures of 4-hydroxy-2-oxoacids (%)
WT
7.4 ( 0.4
163 ( 4
2200 ( 100
254 ( 30
4
-hydroxy-2- 4-hydroxy-2- 4-hydroxy-2- 4-hydroxy-2-
L87N
0.21 ( 0.02
0.5 ( 0.01
enzyme oxopentanoate oxohexanoate oxoheptanoate oxooctanoate
b
L87W
5.8 ( 0.3
WT
51.6 ( 1.0
99.0 ( 3.3
50.6 ( 0.8
98.6 ( 3.2
49.6 ( 1.2
48.8 ( 1.8
51.2 ( 0.7
97.9 ( 2.2
50.6 ( 1.9
51.2 ( 2.1
48.6 ( 0.8
98.2 ( 2.7
49.3 ( 1.2
50.6 ( 1.9
Y290F
0.92 ( 0.07
0.84 ( 0.06
6.2 ( 0.5
2.7 ( 0.1
0.8 ( 0.02
4 ( 0.1
530 ( 40
98 ( 7
Y290F
L87N;Y290F
L87W;Y290F
L87N;Y290F 48.5 ( 1.3
L87W;Y290F 46.2 ( 1.6
64 ( 5
a
Assays were performed with 50 μg of enzyme, 0.4 mM NADH, 19.2
units of LDH, and oxaloacetate varying between at least 0.1Kb and 10K
in 100 mM HEPES buffer, pH 8.0, in a total volume of 1 mL. Due to the
high apparent K for the substrate, the specificity constant can be
a
Assays contained 20 μM 4-hydroxy-2-oxoacid, 0.4 mM NADH, 1 mM
MnCl , and 19.2 units of LDH in 100 mM HEPES buffer, pH 8.0. The
amount of enzyme added was 10 μg for the wild type and 500 μg for the
m
m
2
m
estimated from the linear gradient of specific activity vs substrate
variants.
concentration graph.
Table 4. Specific Rotation of 4-Hydroxy-2-oxoacids Synthe-
sized by BphI and Its Variants
a
4
(R) enantiomer (Table 1). The catalytic efficiencies of the retro-
aldol activities for these variants were however compromised due
2
5
to a ∼40-fold decrease in k and 30À60-fold higher K values
specific rotation of synthesized 4-hydroxy-2-oxoacids [α]D
cat
m
compared to those of the wild-type enzyme. The decrease in k is
cat
4-hydroxy-2- 4-hydroxy-2- 4-hydroxy-2- 4-hydroxy-2-
possibly associated with the loss of the phenolic oxygen of Tyr-
2
15
enzyme oxopentanoate oxohexanoate oxoheptanoate oxooctanoate
90 that hydrogen bonds with a catalytic water molecule. No
À1
detectable activity (activity of <0.0001 s ) was observed in single
variants L87N and L87W for either enantiomer of HOPA.
Oxaloacetate (OAA) Decarboxylase Activity. To determine
if the inactivity of L87N and L87W single variants in the retro-
aldol reaction above is due to catalytic inactivation of the variant
enzymes or the inability of the enzymes to bind to HOPA, we
tested the enzymes for OAA decarboxylase activity. Oxaloace-
tate decarboxylation is a secondary reaction of pyruvate
WT
+15.4
0
+6.7
À0.8
À7.9
À7.9
+14.0
À0.2
+33.3
0
Y290F
L87N;Y290F
L87W;Y290F
+13.6
+14.0
À14.7
À13.7
À33.1
À34.0
a
Samples were synthesized by each respective enzyme, lactonized and
purified with an Aminex HPX-87H column with a mobile phase of 0.1%
formic acid, and lyophilized. The optical rotation of purified substrates
was determined using a Rudolph Autopol IV automatic polarimeter at
1
2,20,28
aldolases
that proceeds through CÀC bond cleavage
25 °C at a wavelength of 589 nm.
between C3 and C4, followed by proton donation to a pyruvate
enolate intermediate, analogous to the retro-aldol reaction.
Oxaloacetate is however one carbon shorter than HOPA and
more highly oxidized so that CÀC bond cleavage leads to
production of carbon dioxide rather than an aldehyde. Both
L87N and L87W were found to possess OAA decarboxylase
activity, confirming that they are catalytically active. The catalytic
efficiencies of OAA decarboxylase in the L87NandL87W variants
were however lower than those of the wild-type enzyme by 10-
and 380-fold, respectively (Table 2). The L87N variant exhibited
a 35-fold decrease in K relative to the wild type, while the K for
Stereochemical Control for Longer Aldehyde Substrates.
To determine if the stereospecificity for the opposing enantio-
mer to the wild-type enzyme is preserved in the double variants
for aldehydes longer than acetaldehyde, variants were tested for
activity toward racemic mixtures of 4-hydroxy-2-oxoacids con-
taining alkyl chains eight or more carbons in length (pyruvate
addition products with propionaldehyde, butyraldehyde, and
pentaldehyde). These racemic compounds were synthesized
using HpaI, an enzyme that does not exhibit stereochemical
m
m
14
control. The wild type and L87N;Y290F and L87W;Y290F
variants were only able to degrade ∼50% of racemic 4-hydroxy-2-
oxoacids (Table 3). In addition, the ∼50% 4-hydroxy-2-oxoacids
remaining after treatment of the wild-type enzyme could be
further degraded to completion by the L87N;Y290F or L87W;
Y290F variants, showing that the variants utilize the opposite
enantiomer that the wild-type enzyme cannot utilize. Similarly,
when racemic 4-hydroxy-2-oxoacids were treated with Y290F;
L87N or Y290F;L87W variants first, the wild-type enzyme was
able to degrade the remaining substrate. In contrast, the Y290F
variant was able to cleave ∼98% of these racemic mixtures,
confirming that this single variant is not stereospecific.
the L87W variant was much higher than that of the wild type as
the enzyme could not be saturated with the substrate under the
assay conditions. The kcat values for OAA decarboxylase activity
were also lower than those of the wild type, possibly due to the
absence of the phenolic oxygen of Tyr-290 that is important in
positioning the catalytic water that protonates the pyruvate
enolate. Both enantiomers of HOPA at a concentration of
1
0 mM did not inhibit the OAA decarboxylase activities of the
two single leucine variants, confirming that the lack of aldolase
activity is due to the inability of the enzymes to bind HOPA.
OAA decarboxylase activity was not dramatically affected in
the L87N;Y290F variant relative to the single L87N variant. In
comparison, the K for the L87W;Y290F variant is lower than
The aldol addition products of pyruvate and various aldehydes
synthesized by the wild-type enzyme and each variant enzyme
were also subjected to polarimetric analysis. The specific rota-
tions of 4-hydroxy-2-oxoacids produced by L87N;Y290F and
L87W;Y290F variants were opposite those of 4-hydroxy-2-
oxoacids synthesized by the wild-type enzyme (Table 4). The
m
that of the L87W variant, providing evidence that Trp-87 also
sterically impedes the proper binding of the carboxylate of
oxaloacetate in the single L87W variant, and this is alleviated
upon the removal of the phenolic oxygen of Tyr-290 in the
double variant.
5
10
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
ARTICLE
Y290F single variant produced substrates with an optical rotation
of approximately 0, further confirming that the enzyme exhibits
poor facial selectivity. Together, these results demonstrate that
constructed variants exhibit the stereoselectivity opposite that of
the wild type and that the alkyl chain length of the aldehyde does
not affect stereochemical control.
’ DISCUSSION
Pyruvate aldolases are a diverse group of enzymes that differ
in their substrate specificities, stereospecificities, and catalytic
mechanisms. For example, some pyruvate aldolases that utilize
catalytic lysine residues in their reaction mechanisms (class I
2
9
aldolases), such as N-acetylneuraminic acid (NAL) aldolase
and 2-keto-3-deoxygluconate (KDG) aldolase, lack stereochemical
Kinetic Analysis of the Aldol Addition Reaction. Steady-
state kinetic parameters were determined in the aldol addition
reaction using pyruvate and various aldehydes. The Km,app values
for pyruvate were approximately 2À4-fold higher in the variants
than in the wild-type enzyme (Table 5), suggesting that pyruvate
binding is largely unaffected. All variants exhibited similar kcat
values that were 3À10-fold lower than that of the wild type. In
30,31
control.
Other class I aldolases such as 2-keto-3-deoxy-6-
phosphogluconate-6-phosphate (KDPG) aldolase and 2-keto-3-
deoxy-6-phosphogalactonate (KDPGal) aldolase catalyze the iden-
tical reaction, but with si- and re-facial selectivity at C4, respectively,
1
0
providing access to both stereochemical sequences. However,
they only utilize aldehydes with polar groups at C2, C3, or C4 as
16
substrates. Pyruvate aldolases that are divalent metal dependent
class II aldolases) include HpaI and 4-hydroxy-4-methyl-2-oxo-
glutarate/4-carboxy-4-hydroxy-2-oxoadipate (HMG/CHA) aldo-
addition, the variants exhibited specificity constants (k /K
)
cat m,app
(
toward acetaldehyde and propionaldehyde similar to those of the
wild-type enzyme, suggesting that the C2 of the aldehyde forms
favorable hydrophobic interactions with the phenyl ring of Phe-
14,28,32
lases that, unlike BphI, lack stereospecificity.
HpaI and BphI
exhibit broad specificity for the aldol reaction of pyruvate with
nonphosphorylated nonsubsituted aldehydes and are therefore
useful for synthesis of 4-hydroxy-2-ketoacid precursors for organic
synthesis.
During the aldol addition catalyzed by BphI, the S-configured
stereogenic center at C4 is created via attack of a pyruvate enolate
intermediate onto the si face of the aldehyde carbonyl of
acetaldehyde. The availability of an X-ray crystal structure of
2
90 (Table 6). The Y290F single variant exhibited similar
specificities for all aldehydes and had notably higher specificity
for pentaldehyde compared to the wild type and double variants.
Similar to the wild-type enzyme, the double variants exhibited a
reduction in catalytic efficiencies as the aldehyde chain length
increased to more than three carbons. This is primarily due to an
increase in Km,app for longer aldehydes. Overall the k /K for the
cat
m
double variants in the aldol addition reaction was affected e10-
fold relative to that of the wild-type enzyme.
19
the orthologous DmpG (56% amino acid identity), containing
the pyruvate enolate analogue oxalate, previously enabled us to
propose a model for the binding of 4(S)-hydroxy-2-oxopentano-
Table 5. Steady-State Kinetic Parameters for the Utilization
of Pyruvate in the BphI Wild Type (WT) and L87N;Y290F
and L87W;Y290F Variants in the Aldol Addition Reaction
15
ate in the active site. A critical residue is Tyr-290, which
prevents the binding of the 4(R) isomer due to steric clash
between the C4-methyl substituent of the substrate and the
hydroxyl oxygen of the tyrosyl residue. Hence, in the Y290F
variant, steric constraint at the stereogenic carbon is alleviated,
resulting in a loss of stereochemical control. Since Tyr-290
functions to position a catalytic water in the reaction mechanism,
the activity of the Y290F variant is, however, also slightly
compromised. Interestingly, a T161V substitution in KDPG
aldolase also eliminated the stereochemical control in the
enzyme, allowing the enzyme to accept either KDPG or KDPGal
a
À1
À1 À1
enzyme
Km,app (mM)
kcat,app (s
)
(kcat/K
m
)
app (M
s
)
3
WT
13.0 ( 0.6
20.0 ( 1.9
30.2 ( 1.6
38.2 ( 3.1
1.20 ( 0.02
0.19 ( 0.01
0.11 ( 0.01
0.16 ( 0.01
91.5 ( 4.7
8.4 ( 0.8
3.6 ( 0.2
4.3 ( 0.4
Y290F
L87N;Y290F
L87W;Y290F
a
Assays were performed at 25 °C and contained 120 μg of BphI, 1 mM
MnCl , and 240 mM propionaldehyde with varying concentrations of
2
pyruvate in 100 mM HEPES buffer, pH 8.0. The 4-hydroxy-2-oxohex-
anoate product formed was lactonized and detected at 215 nm.
11
as the substrate with roughly equal efficiencies. However, unlike
Table 6. Steady-State Kinetic Parameters for the Utilization of Various Aldehydes of the BphI Wild Type and Variants in the Aldol
a
Addition Reaction
enzyme
WT
kinetic parameter
acetaldehyde, two carbons propionaldehyde, three carbons butryaldehyde, four carbons pentaldehyde, five carbons
K
m,app (mM)
64.2 ( 3.7
0.860 ( 0.020
13.4 ( 0.80
81.0 ( 6.8
0.20 ( 0.005
2.5 ( 0.2
136 ( 9.6
1.79 ( 0.07
13.2 ( 1.1
81.2 ( 7.0
0.16 ( 0.006
2.0 ( 0.2
À1
k
cat,app (s
)
À1 À1
b
b
k
cat/Km,app (M
s
)
)
)
)
2.19 ( 0.12
104 ( 8.3
0.470 ( 0.01
80.5 ( 7.6
0.14 ( 0.003
1.7 ( 0.1
3
Y290F
K
m,app (mM)
À1
k
cat,app (s
)
0.16 ( 0.005
1.5 ( 0.1
À1 À1
k
cat/Km,app (M
s
3
L87N;Y290F
L87W;Y290F
K
m,app (mM)
70.7 ( 3.7
0.30 ( 0.003
4.3 ( 0.2
54.9 ( 3.8
0.10 ( 0.003
1.8 ( 0.1
À1
k
cat,app (s
)
À1 À1
b
b
b
k
cat/Km,app (M
s
0.16 ( 0.008
0.31 ( 0.004
0.040 ( 0.002
3
K
m,app (mM)
86.3 ( 8.5
0.59 ( 0.003
6.9 ( 0.7
24.4 ( 1.9
0.14 ( 0.003
5.7 ( 0.5
À1
k
cat,app (s
)
À1 À1
b
k
cat/Km,app (M
s
0.051 ( 0.003
3
a
Assays were performed at 25 °C and contained 120 μg of BphI, 1 mM MnCl
00 mM HEPES buffer, pH 8.0. The 4-hydroxy-2-oxoacid products formed were lactonized and detected at 215 nm via HPLC. Due to the high
2
, and 100 mM pyruvate with varying concentrations of aldehyde in
b
1
apparent K
m
for the substrate, the specificity constant can be estimated from the linear gradient of specific activity vs substrate concentration graph.
5
11
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
ARTICLE
same time, these Leu-87 single variants are unable to accommodate
the 4(R) enantiomer due to the steric blockage by Tyr-290. Thus,
combining mutations at two positionsinthedoublevariants(L87N;
Y290F and L87W;Y290F) led to enzymes with stereoselectivity for
the 4(R) enantiomer. Km,app values for pyruvate binding were
unaffected in these variants, consistent with our model indicating
that these amino acid substitutions are sufficiently removed from the
carbonyl donor. The L87W;Y290F variant had a marginally higher
k /K in both the retro-aldol and aldol addition reactions relative
cat
m
to the L87N;Y290F variant possibly due to the more favorable
accommodation of the hydrophobic tail of 4-hydroxy-2-ox-
oacids if the active site is also hydrophobic. The stereoselec-
tivity is retained for longer aldehyde substrates, suggesting
that Phe-290 does not sterically hinder their binding. Inspec-
tion of the crystal structure of DmpG suggests that there is
sufficient space at the distal end of the active site pocket to
Figure 2. Model of 4(R)-hydroxy-2-oxooctanoate in the substrate
binding site of DmpG. Tyr-291 and Leu-88 of the active site are replaced
with phenylalanine and asparagine, respectively, in the model. The distal
end of the substrate can adopt different conformations to fit within the
solvent-accessible areas of the active site (mesh). Carbon atoms of the
substrates are colored cyan, and the residues in the substrate binding
sites are colored gray. Corresponding residue numbers in BphI are given
in parentheses.
1
3
accommodate these aldehydes (Figure 2).
There are few examples of successful rational design of
stereoselectivity among aldolases. In fact, directed evolution
based on random mutagenesis has been the preferred method
to alter the stereoselective profiles of aldolases and other
36À40
biocatalysts.
Although no prior structural information on
the enzyme is necessary for directed evolution, its main limitation
is the requirement for robust screening methods to select for
4
1
the BphI Y290F variant, this substitution was accompanied by a
significant loss of kcat (10 ) against both the natural and epimeric
substrates. The reason why the T161V variant loses stereospe-
cificity is not clear, but it suggests catalytic involvement in
stereochemical control.
desired variants from a pool of thousands of mutants. An
alternative method to modify the stereoselectivity of aldolases is
by rational design, an approach that has been attempted with
4
42
2
-keto-3-deoxygluconate aldolase. The natural enzyme pro-
duces the diasteroisomersD-2-keto-3-deoxygluconate (D-KDGlu)
and D-2-keto-3-deoxygalactonate (D-KDGal) in almost equal
mixtures using pyruvate and D-glyceraldehyde as substrates.
Detailed structures of the enzyme containing bound D-KDGlu
or D-KDGal were required to modify the enzyme by site-specific
mutagenesis to preferentially produce one or the other enantio-
mer, exploiting the subtle differences in hydrogen-bonding
interactions between the enzyme and the C4 (the stereogenic
center), C5, and C6 oxygens/hydroxyls of the D-glyceraldehyde
moiety of the respective enantiomer.
Aromatic residues within the active sites of other enzymes have
been implicated in stereoselectivity. Steric occlusion by Phe-297
within the active site of acetylcholinesterase is the primary deter-
33
minant of both enantiomeric preference and substrate selectivity.
Similarly, Phe-77 and Phe-138 in aryl sulfotransferase IV, which
catalyzes the sulfation of many drugs, carcinogens, and xenobiotics,
was shown by alanine substitutions to provide significant steric
interactions that both impart stereospecificity and enhance catalytic
34
efficiency. Among aldolases, fructose-1,6-bisphosphate aldolase
was engineered to catalyze the formation of the opposite enana-
tiomer, tagatose, by replacing a histidine residue (His-26) with an
aromatic tyrosine (H26Y). The subtle change in conformation of
the active site in this variant affects the position of the bound
glyceraldehyde 3-phosphate (G3P) plane relative to the dihydroxy-
acetonephosphate (DHAP) enediolate, permitting the attack of
’
CONCLUSION
For many aldolases, the mechanism for stereochemical
control is not fully understood. We have shown how structure
modeling can be used to rationally design a stereocomple-
mentary variant of BphI. The results presented herein showed
that both Leu-87 and Tyr-290 govern stereochemical control
and that relatively few amino acid substitutions within the
active site can result in large changes in the stereoselectivity of
an enzyme. This approach compares favorably and can com-
plement directed evolution approaches to create biocatalysts
of desired stereospecificity.
35
DHAP on the opposite face of the incoming G3P molecule.
While Tyr-290 in BphI prevents binding of the 4(R) isomer of
the substrate, Leu-87, positioned opposite Tyr-290 in the active site
of BphI, appeared to be important for interaction with the 4-methyl
of the S enantiomer of the substrate. This is supported by previous
studies where replacementof Leu-87 with a shorteralaninereduced
the specificity constants of the enzyme toward acetaldehyde 32-fold
while maintaining a similar specificity constant for pyruvate in the
15
’ AUTHOR INFORMATION
aldol addition reaction. As demonstrated through inhibition
studies, Leu-87 variants lost the ability to bind and catalyze the
cleavage of HOPA, although they were still catalytically active as
they retained OAA decarboxylase activities. OAA differs from
HOPA in that it lacks the C5-alkyl group. Therefore, in the
L87N variant, the incompatibility between the hydrophobic C5-
alkyl of HOPA and the polar side chain of Asn-87 is a possible factor
that has a negative effect on HOPA binding and catalysis. In the
L87W variant, an unfavorable steric clash is thought to occur
between the C5-alkyl and the bulky side chain of Trp-87. At the
Corresponding Author
sseah@uoguelph.ca
’
ACKNOWLEDGMENT
This research is supported by the National Science and
Engineering Research Council of Canada (NSERC), Grant
238284 (to S.Y.K.S.). P.B. is a recipient of an NSERC PGS-D
scholarship.
5
12
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513

Journal of the American Chemical Society
REFERENCES
ARTICLE
’
(40) Wada, M.; Hsu, C. C.; Franke, D.; Mitchell, M.; Heine, A.;
Wilson, I.; Wong, C. H. Bioorg. Med. Chem. 2003, 11, 2091.
(
(
(1) Patel, R. N. Adv. Appl. Microbiol. 2000, 47, 33.
41) Panke, S.; Wubbolts, M. Curr. Opin. Chem. Biol. 2005, 9, 188.
42) Royer, S. F.; Haslett, L.; Crennell, S. J.; Hough, D. W.; Danson,
M. J.; Bull, S. D. J. Am. Chem. Soc. 2010, 132, 11753.
43) DeLano, W. L. The PyMOL Molecular Graphics System; DeLano
Scientific: San Carlos, CA, 2002.
(
(
2) Patel, R. N. Curr. Opin. Biotechnol. 2001, 12, 587.
3) Patel, R.; Hanson, R.; Goswami, A.; Nanduri, V.; Banerjee, A.;
Donovan, M. J.; Goldberg, S.; Johnston, R.; Brzozowski, D.; Tully, T.;
Howell, J.; Cazzulino, D.; Ko, R. J. Ind. Microbiol. Biotechnol. 2003, 30, 252.
(
(4) Patel, R. N. Curr. Opin. Drug Discovery Dev. 2006, 9, 741.
(5) Gijsen, H. J.; Qiao, L.; Fitz, W.; Wong, C. H. Chem. Rev. 1996,
96, 443.
(
(
6) Koeller, K. M.; Wong, C. H. Chem. Rev. 2000, 100, 4465.
7) Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. Biotechnol.
Bioeng. 2000, 70, 349.
(
(
8) Kajimoto, T. Yakugaku Zasshi 2000, 120, 42.
9) Calveras, J.; Egido-Gabas, M.; Gomez, L.; Casas, J.; Parella, T.;
Joglar, J.; Bujons, J.; Clapes, P. Chemistry 2009, 15, 7310.
(10) Walters, M. J.; Toone, E. J. Nat. Protoc. 2007, 2, 1825.
(11) Walters, M. J.; Srikannathasan, V.; McEwan, A. R.; Naismith,
J. H.; Fierke, C. A.; Toone, E. J. Bioorg. Med. Chem. 2008, 16, 710.
12) Baker, P.; Pan, D.; Carere, J.; Rossi, A.; Wang, W.; Seah, S. Y.
Biochemistry 2009, 48, 6551.
(
(13) Carere, J.; Baker, P.; Seah, S. Y. Biochemistry 2011, 50, 8407.
(14) Wang, W.; Baker, P.; Seah, S. Y. Biochemistry 2010, 49, 3774.
(15) Baker, P.; Carere, J.; Seah, S. Y. Biochemistry 2011, 50, 3559.
(16) Shelton, M. C.; Cotterill, I. C.; Novak, S. T. A.; Poonawala,
R. M.; Sudarshan, S.; Toone, E. J. J. Am. Chem. Soc. 1996, 118, 2117.
17) Stock, S. D.; Hama, H.; Radding, J. A.; Young, D. A.; Takemoto,
J. Y. Antimicrob. Agents Chemother. 2000, 44, 1174.
18) Cheriyan, M.; Walters, M. J.; Kang, B. D.; Anzaldi, L. L.; Toone,
E. J.; Fierke, C. A. Bioorg. Med. Chem. 2011, 19, 6447.
19) Manjasetty, B. A.; Powlowski, J.; Vrielink, A. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 6992.
(
(
(
(
20) Wang, W.; Seah, S. Y. Biochemistry 2005, 44, 9447.
(21) Sambrook, J., Fritsch, E. F., Maniatis, T. Molecular Cloning: A
Laboratory Manual; Cold Spring Harbor Laboratory Press: Plainview,
NY, 1989.
(
(
(
(
22) Liu, H.; Naismith, J. H. BMC Biotechnol. 2008, 8, 91.
23) Sarkar, G.; Sommer, S. S. Biotechniques 1990, 8, 404.
24) Bradford, M. M. Anal. Biochem. 1976, 72, 248.
25) Cornish-Bowden, A. Analysis of Enzyme Kinetic Data; Oxford
University Press: New York, 1995.
26) Cornish-Bowden, A.; Eisenthal, R. Biochem. J. 1974, 139,
21.
(
7
(
(
27) Eisenthal, R.; Cornish-Bowden, A. Biochem. J. 1974, 139, 715.
28) Wang, W.; Mazurkewich, S.; Kimber, M. S.; Seah, S. Y. J. Biol.
Chem. 2010, 285, 36608.
29) Williams, G. J.; Woodhall, T.; Farnsworth, L. M.; Nelson, A.;
Berry, A. J. Am. Chem. Soc. 2006, 128, 16238.
30) Lamble, H. J.; Heyer, N. I.; Bull, S. D.; Hough, D. W.; Danson,
M. J. J. Biol. Chem. 2003, 278, 34066.
31) Theodossis, A.; Walden, H.; Westwick, E. J.; Connaris, H.;
Lamble, H. J.; Hough, D. W.; Danson, M. J.; Taylor, G. L. J. Biol. Chem.
(
(
(
2
004, 279, 43886.
(32) Maruyama, K. J. Biochem. 1990, 108, 327.
(33) Hosea, N. A.; Berman, H. A.; Taylor, P. Biochemistry 1995,
3
3
4, 11528.
(
34) Sheng, J. J.; Saxena, A.; Duffel, M. W. Drug Metab. Dispos. 2004,
2, 559.
(
35) Williams, G. J.; Domann, S.; Nelson, A.; Berry, A. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 3143.
36) Clapes, P.; Fessner, W. D.; Sprenger, G. A.; Samland, A. K. Curr.
Opin. Chem. Biol. 2010, 14, 154.
37) Bolt, A.; Berry, A.; Nelson, A. Arch. Biochem. Biophys. 2008,
74, 318.
38) Jaeger, K. E.; Eggert, T.; Eipper, A.; Reetz, M. T. Appl. Microbiol.
Biotechnol. 2001, 55, 519.
39) Fong, S.; Machajewski, T. D.; Mak, C. C.; Wong, C. Chem. Biol.
000, 7, 873.
(
(
4
(
(
2
5
13
dx.doi.org/10.1021/ja208754r |J. Am. Chem. Soc. 2012, 134, 507–513