Comparison of two metal-dependent pyruvate aldolases related by convergent evolution: Substrate specificity, kinetic mechanism, and substrate channeling

Article abstract of DOI:10.1021/bi100251u

HpaI and BphI are two pyruvate class II aldolases found in aromatic meta-cleavage degradation pathways that catalyze similar reactions but are not related in sequence. Steady-state kinetic analysis of the aldol addition reactions and product inhibition assays showed that HpaI exhibits a rapid equilibrium random order mechanism while BphI exhibits a compulsory order mechanism, with pyruvate binding first. Both aldolases are able to utilize aldehyde acceptors two to five carbons in length; however, HpaI showed broader specificity and had a preference for aldehydes containing longer linear alkyl chains or C2-OH substitutions. Both enzymes were able to bind 2-keto acids larger than pyruvate, but only HpaI was able to utilize both pyruvate and 2-ketobutanoate as carbonyl donors in the aldol addition reaction. HpaI lacks stereospecific control producing racemic mixtures of 4-hydroxy-2-oxopentanoate (HOPA) from pyruvate and acetaldehyde while BphI synthesizes only (4S)-HOPA. BphI is also able to utilize acetaldehyde produced by the reduction of acetyl-CoA catalyzed by the associated aldehyde dehydrogenase, BphJ. This aldehyde was directly channeled from the dehydrogenase to the aldolase active sites, with an efficiency of 84%. Furthermore, the BphJ reductive deacylation reaction increased 4-fold when BphI was catalyzing the aldol addition reaction. Therefore, the BphI-BphJ enzyme complex exhibits unique bidirectionality in substrate channeling and allosteric activation.

Full text of DOI:10.1021/bi100251u

3774 Biochemistry 2010, 49, 3774–3782
DOI: 10.1021/bi100251u
Comparison of Two Metal-Dependent Pyruvate Aldolases Related by Convergent
Evolution: Substrate Specificity, Kinetic Mechanism, and Substrate Channeling†
Weijun Wang,‡ Perrin Baker,‡ and Stephen Y. K. Seah*
Department of Molecular and Cellular Biology, University of Guelph, Guelph, Ontario, Canada N1G 2W1.
‡These authors contributed equally
Received February 18, 2010; Revised Manuscript Received April 1, 2010
ABSTRACT: HpaI and BphI are two pyruvate class II aldolases found in aromatic meta-cleavage degradation
pathways that catalyze similar reactions but are not related in sequence. Steady-state kinetic analysis of the
aldol addition reactions and product inhibition assays showed that HpaI exhibits a rapid equilibrium random
order mechanism while BphI exhibits a compulsory order mechanism, with pyruvate binding first. Both
aldolases are able to utilize aldehyde acceptors two to five carbons in length; however, HpaI showed broader
specificity and had a preference for aldehydes containing longer linear alkyl chains or C2-OH substitutions.
Both enzymes were able to bind 2-keto acids larger than pyruvate, but only HpaI was able to utilize both
pyruvate and 2-ketobutanoate as carbonyl donors in the aldol addition reaction. HpaI lacks stereospecific
control producing racemic mixtures of 4-hydroxy-2-oxopentanoate (HOPA) from pyruvate and acetaldehyde
while BphI synthesizes only (4S)-HOPA. BphI is also able to utilize acetaldehyde produced by the reduction of
acetyl-CoA catalyzed by the associated aldehyde dehydrogenase, BphJ. This aldehyde was directly channeled
from the dehydrogenase to the aldolase active sites, with an efficiency of 84%. Furthermore, the BphJ
reductive deacylation reaction increased 4-fold when BphI was catalyzing the aldol addition reaction.
Therefore, the BphI-BphJ enzyme complex exhibits unique bidirectionality in substrate channeling and
allosteric activation.
The aerobic degradation of aromatic compounds by micro-
organisms is important for maintenance of the global carbon
cycle and for removal of toxic industrial pollutants such as
polychlorinated biphenyls (PCBs) (1, 2). It is also important
for cholesterol utilization and hence survival of Mycobacterium
tuberculosis in host macrophages (3). In the meta-cleavage path-
way, diverse aromatic compounds can be transformed, via a
catechol intermediate, to 4-hydroxy-2-ketoacids which are sub-
sequently cleaved across the C3-C4 bond by pyruvate aldolases
to form pyruvate and an aldehyde (4, 5). For example, BphI
(EC 4.1.3.39), the aldolase in the PCBs degradation pathway of
Burkholderia xenovorans LB400, catalyzes the aldol cleavage
of 4-hydroxy-2-oxopentanoate to pyruvate and acetaldehyde
(Figure 1A) (6). HpaI (EC 4.1.2.-) (also known as HpcH), in
the 3- and 4-hydroxyphenylacetate catabolic pathway of Escher-
ichia coli strain W or C, transforms 4-hydroxy-2-oxo-1,7-hepta-
nedioate to pyruvate and succinic semialdehyde (Figure 1B) (7).
HpaI is also able to catalyze the aldol cleavage of 4-hydroxy-2-
oxopentanoate, the physiological substrate of BphI (8, 9). Both
HpaI and BphI share a similar reaction mechanism as they are
competitively inhibited by oxalate and exhibit secondary oxa-
loacetate decarboxylase activity. This can be rationalized by the
formation of a pyruvate enolate intermediate in their catalytic
mechanisms (10). This pyruvate enolate intermediate is stabilized
by divalent metal ions in the two enzymes reminiscent of class II
aldolases (11, 12) and is distinct from class I aldolases that utilize
a Schiff base mechanism (13-15).
Interestingly, HpaI and BphI do not share any sequence simi-
larities. HpaI belongs to the pyruvate/phosphoenolpyruvate
family of enzymes (16, 17), whose members also include
citrate lyase and pyruvate kinase (18), while BphI belongs to
the HMGL-like family of enzymes that includes its homologue
DmpG from the phenol degradation pathway and the enzymes
3-hydroxy-3-methylglutaryl-CoA lyase, 2-isopropylmalate syn-
thase, and transcarboxylase 5S (19). Homologues of HpaI and
BphI are clustered into two different COG groups (COG3836 for
HpaI and COG0119 for BphI) and are widely distributed in
microorganisms (20, 21). Although both the HpaI and BphI
families of aldolases adopt TIM barrel folds, the HpaI hexamer is
composed of (Rβ)8 barrels in which R8 from each subunit forms a
domain-swapped dimer with the adjacent subunit (22). In con-
trast, the BphI family of aldolases is composed of two symme-
trical half-barrels, (Rβ)1-4 and (Rβ)5-8, with helical insertions
between β1-R1 and β5-R5. They also contain an additional
R-helical domain located at the C-terminus of the TIM barrel
called the communication domain which plays an important role
in heterodimerization with an associated acetaldehyde dehydro-
genase (EC 4.1.2.10) that converts the acetaldehyde formed from
the aldolase reaction to acetyl-CoA using NADþ and coenzyme
A as cofactors (6, 23). This association enables the direct
channeling of the acetaldehyde product from the aldolase to
the dehydrogenase, as demonstrated recently in the BphI aldo-
lase-BphJ dehydrogenase complex. In addition, the amino acid
ligands of the metal cofactor differ in HpaI and BphI (Glu149 and
Asp175 in HpaI; His199, His201, and Asp18 in BphI) which may
†This research is supported by National Science and Engineering
Research Council of Canada (NSERC) Grant 238284 (to S.Y.K.S.).
P.B. is a recipient of a NSERC PGS-D scholarship.
*Corresponding author. E-mail: sseah@uoguelph.ca. Phone: (519)
824 4120 ext 56750. Fax: (519) 837 1802.
r
pubs.acs.org/Biochemistry
Published on Web 04/05/2010
2010 American Chemical Society

Article
Biochemistry, Vol. 49, No. 17, 2010 3775
F
IGURE 1: Comparison of the last steps of (A) biphenyl degradation and (B) 4-hydroxyphenylacetate and pathways. The hydratases HpaH and
BphH share 35% sequence identity. However, HpaI and BphI do not share any sequence similarities. The biphenyl degradation contains one
additional enzyme, an acetaldehyde dehydrogenase, BphJ.
contribute to differences in metal specificity observed in the two
enzymes (6, 8). Since all of TIM barrel enzymes are thought to
arise from a common ancestor, the evolutionary relationship
between HpaI and BphI can be best described as functional
convergence after evolutionary sequence divergence.
genase (ADH, Saccharomyces cerevisiae), and Dowex 1X8-200
ion-exchange resin were from Sigma-Aldrich (Oakville, Ontario,
Canada). Ni-NTA superflow resin was obtained from Qiagen
(Mississauga, Ontario, Canada). HpaI and BphI-J were over-
expressed and purified according to the methods previously
described (6, 8). All other chemicals were of analytical grade
and were obtained from Sigma-Aldrich and Fisher Scientific
(Nepean, Ontario, Canada).
Besides their importance in aromatic degradation pathways,
pyruvate aldolases are potentially useful for synthesis of natural
products and bioactive molecules due to their ability to catalyze
C-C bond formation (24, 25). However, compared to class I
pyruvate aldolases (13), very little is known about class II
pyruvate aldolases. Substrate specificity and kinetic analysis have
only been determined in the aldol cleavage direction with a
limited number of compounds (6, 8). Although aldol addition
assays have been reported for MhpE (80% and 56% sequence
identity to BphI and DmpG, respectively), the enzyme analyzed
might be a class I aldolase contaminant since it did not require
metal ions for activity and was sensitive to sodium borohydride
treatment (26). It is therefore unclear whether pyruvate class II
aldolases could utilize substituted aldehydes or other 2-ketoacids
besides pyruvate as substrates. The stereochemistry of the reac-
tion catalyzed by the enzymes has also not been determined.
Here, we describe a detailed comparison of the substrate speci-
ficity, stereospecificity, and kinetic mechanisms between HpaI
and BphI in the aldol addition reaction. The structure-function
relationship analysis of these enzymes provides an important
framework, not only for future utilization and engineering of
these enzymes as biocatalysts but also for modifying aromatic
degradation pathways for environmental bioremediation.
Aldol Cleavage Activity Assay. Aldol cleavage assays
were performed as previously described (6, 8). To rule out the
possibility that assays may be influenced by contaminating metal
ions carried over from the purified enzyme, specific activities of
the purified enzymes and EDTA-treated apoenzymes were
determined and confirmed to be the same. For inhibition assays
in the aldol cleavage reaction, concentrations of each inhibitor
were varied from 0.5Ki to 3Ki. Reactions were coupled with
alcohol dehydrogenase (ADH) or LDH, depending on the nature
of the inhibitor. Data were fitted to either a competitive or a
mixed noncompetitive inhibition equation using Leonora (27).
Aldol Addition Activity Assay. Standard assays contained
80 mM pyruvate or 2-ketobutyric acid, 2 mM divalent metal ion,
3 μg of HpaI or 100 μg of BphI-BphJ, and aldehyde concentra-
tions between 0.1Km and to 2Km in a final volume of 1.0 mL of
100 mM HEPES buffer, pH 8.0. The reaction was initiated by the
addition of enzyme and proceeded for a duration of 10 min for
HpaI and 120 min for BphI 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 to HPLC using an AKTA Explorer 100
ꢀ
apparatus (Amersham Pharmacia Biotech, Baie d’Urfe, Quebec,
EXPERIMENTAL PROCEDURES
Canada) equipped with an Aminex fast-acid analysis ion-ex-
change 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. The identity of the
synthesized 4-hydroxy-2-ketoacid compounds was further con-
firmed by ESI-qTOF MS analysis performed at the Advanced
Analysis Centre, University of Guelph. Due to similar retention
Chemicals. Sodium pyruvate, sodium oxalate, 2-ketobutano-
ate, 2-ketopentanoate, 4-methyl-2-ketopentanoate, aldehydes,
L
-lactate dehydrogenase (LDH,1 rabbit muscle), alcohol dehydro-
1Abbreviations: ADH, alcohol dehydrogenase; DDG, 2-dehydro-3-
deoxygalactarate; HEPES, 4-(2-hydroxyethyl)-1-piperazinepropanesul-
fonic acid; HOPA, 4-hydroxy-2-oxopentanoate; IPTG, isopropyl β-
thiogalactopyranoside; LDH, -lactate dehydrogenase; MOPS, 3-(N-
morpholino)propanesulfonic acid; NMR, nuclear magnetic resonance.
D
-
L
times for 2-dehydro-3-deoxy-L-pentonate (addition product of

3776 Biochemistry, Vol. 49, No. 17, 2010
Wang et al.
pyruvate with glycoaldehyde) and 2-keto-3-deoxygluconate
(KDG, addition product of pyruvate with glyceraldehyde), reac-
tion rates were determined enzymatically using the HpaI cleavage
reaction coupled with LDH. Following each reaction, a 500 μL
aliquot of the reaction solution was filtered through a YM10
filter to remove the enzyme. Dowex 1X8-200 ion-exchange
resin (sulfate form) (50 mg) was added to the flow-through and
mixed for 10 min to remove pyruvate before enzymatic quanti-
fication of product formation. For substrate specificity, data
were fitted to the Michaelis-Menten equation using Leonora.
For kinetic mechanism assays using two varying substrates, data
were fitted to a ternary complex bisubstrate equation (eq 1) using
Leonora.
20 units of ADH, respectively. The reaction was initiated by
addition of acetyl-CoA. Reactions were quenched with 1/5 (v/v)
concentrated HCl at 7 and 15 min and incubated overnight, and
HOPA was quantified by HPLC as described earlier. Reverse
channeling was also demonstrated by measuring NADH con-
sumption at 370 nm (extinction coefficient of 2600 M-1 cm-1).
Assays were combined as mentioned above in the presence
and absence of 100 mM pyruvate, 20 units of ADH, or 10 mM
EDTA. The specific activity of ADH is unaffected by the small
amounts of EDTA added. The productive channeling efficiency
(CE) was defined as the ratio of V3 to V1, which can be deter-
mined by the equation:
V3
V1 - V2
CE ð%Þ ¼
ꢀ 100 ¼
ꢀ 100
V1
V1
VAB
v ¼
ð1Þ
ꢀ
ꢁ
KiAKmB þ KmAB þ KmBA þ AB
V1 þ V2
¼
2 -
ꢀ 100 ¼ ð2 - RÞ ꢀ 100
ð2Þ
V1
where v and V are the initial and maximal velocities, respectively,
A and B are concentrations of pyruvate and acetaldehyde, KmA
and KmB are Michaelis constants of substrates pyruvate and
acetaldehyde, and KiA is the dissociation constant for pyru-
vate (28).
where V1 is the rate of conversion of acetyl-CoA to acetaldehyde
by BphJ, V2 is the rate of conversion of acetaldehyde to ethanol in
the bulk solvent by ADH, V3 is the rate of aldol addition of
pyruvate with acetaldehyde which is channeled from BphJ, R is
the ratio of NADH oxidation with ADH compared to no ADH
which is (V1 þ V2)/V1 (Figure 2).
Determination of Stereospecificity. The 4-hydroxy-2-oxoacids
were synthesized by HpaI (2.5 mg) and BphI (20 mg), respec-
tively, using 0.3 M pyruvate and 1.8 M of acetaldehyde or
0.3 mM succinic semialdehyde and 10 mM pyruvate in 10 mL
of 0.1 M HEPES buffer (pH 8.0) containing 2 mM MnCl2.
Following incubation for 60 min, 0.5 g of Chelex 100 was added
to remove all metal ions and to stop the enzyme reaction. The
reaction mixture was then filtered through a YM10 filter to
remove the enzyme and then separated in a Dowex 1X8-200 ion-
exchange resin using a linear 0-0.1 M H2SO4 gradient. Fractions
containing the product were neutralized to pH ∼7.0 with 5.0 M
NaOH and lyophilized. (4S)-HOPA was also generated enzyma-
tically from catechol using XylE, TodF, and BphH as previously
described (8). (R)-HOPA was generated by treatment of racemic
HOPA with 8 mg of BphI in 0.1 M HEPES buffer (pH 8.0)
containing 2 mM MnCl2 to remove the majority of (4S)-HOPA.
The product was purified using Dowex 1X8-200 resin and
HPLC. The concentration of each enantiomer of HOPA was
determined by aldol cleavage with BphI, followed by HpaI, both
in a coupled assay using LDH. Optical rotation of enantiomers
was determined using an Autopol III automatic polarimeter at
25 °C. The NMR spectrum was recorded with a Bruker Avance
600 MHz at 25 °C containing a cyroprobe.
Active Site Architecture Analysis. The structure of HpaI
was analyzed using coordinates from the PDB (2V5K). The
model of BphI, residues 4-340, was generated by ModWeb (29),
an online server based on MODELLER (30, 31), using the crystal
structure of the homologous aldolase DmpG (1NVM:A) which
has 56% sequence identity and 70% sequence similarity to
BphI (23). A total of 117 models were generated, and the top
predicted protein structure was selected and validated based on
sequence identity, discrete optimized protein energy (DOPE),
ModPipe protein quality score (MPQS), and MetaMQAPII (32).
The final model had a z-DOPE of -1.27 and a MPQS of 1.65999
which is >1.1. MetaMQAPII showed that the generated model
˚
has an rmsd of 1.472 A and GDT-TS (global distance test total
score), which is a CASP measure of model quality, of 82.493. All
images were generated using PyMOL (33).
RESULTS
Carbonyl Donor and Aldehyde Acceptor Specificity.
HpaI and BphI were tested for their ability to catalyze the
enolization of pyruvate analogues by monitoring the R-proton
1
Proton Exchange of 2-Ketoacids Catalyzed by HpaI and
BphI. Proton exchange reactions were performed in a 5.0 mm
NMR tube with an assay solution of 600 μL containing 30 mM
2-ketoacids, 2 mM divalent metal ion, and 0.2 mg of BphI-BphJ
in 20 mM deuterated MOPS buffer (pD 8.0). The chemical shift
of each R-proton of 2-ketoacid is as follows: pyruvate (s, 2.38
ppm), 2-ketobutyric acid (m, 2.70-2.80 ppm), 2-ketopentanoate
(m, 2.69-2.72 ppm), and 4-methyl-2-oxopentanoate (m, 2.59-
2.62 ppm). Enzyme was added to initiate the reaction, and the
1H NMR spectrum was recorded four times every 3 min over a
period of 30 min using a Bruker Avance 600 MHz spectrometer
at 25 °C using previously described methods (6, 8). A solution
without enzyme was measured and used as a negative control.
Reverse Substrate Channeling and Channeling Effi-
ciency. Assays contained 1 mM acetyl-CoA or acetaldehyde,
100 mM pyruvate, 1.5 mM NADH, 2 mM MnCl2, and 180 μg of
BphI-BphJ in 100 mM HEPES buffer (pH 8.0) with or without
exchange rate of these compounds in deuterated buffer by H
NMR. Interestingly, HpaI was able to catalyze the proton
exchange of 2-ketobutanoic acid, which is one carbon longer than
the physiological carbonyl donor, pyruvate, although the kexchange
is 60-fold lower (kexchange of 12.7 ( 0.3 μmol min-1 mg-1 versus
3
3
769 ( 22 μmol min-1 mg-1). BphI was only able to catalyze the
3
3
proton exchange of pyruvate with a kexchange of 0.32 ( 0.02
μmol min-1 mg-1. Neither enzyme was able to catalyze proton
3
3
exchange of 2-ketopentanoic acid and 4-methyl-2-ketopentanoic
acid. To determine whether the aldolases can bind these pyruvate
analogues, steady-state kinetic inhibition assays were performed.
All 2-ketoacids tested were able to competitively inhibit the aldol
cleavage of 4-hydroxy-2-oxopentanoate by HpaI and BphI,
demonstrating that active sites of both enzymes are able to
accommodate 2-ketoacids larger than pyruvate (Table 1). The
inhibition constants (Kic), however, increase with increasing size of
the 3-subsutituent.

Article
Biochemistry, Vol. 49, No. 17, 2010 3777
FIGURE 2: Overview of reverse channeling assay in the aldolase-dehydrogenase complex BphI-BphJ. V1 is the rate of conversion of acetyl-CoA
to acetaldehyde by BphJ, V2 is the rate of conversion of acetaldehyde to ethanol in the bulk solvent by ADH, and V3 is the rate of aldol addition of
pyruvate with acetaldehyde by BphI.
obtained for acetaldehyde and propionaldehyde are consistent
Table 1: Inhibition Constants of 2-Ketoacids in the Aldol Cleavage of
HOPAa
with the similar specificity of the enzyme toward the corresponding
4-hydroxy-2-oxopentanoate and 4-hydroxy-2-oxohexanoate sub-
strates in the aldol cleavage direction (6). Longer chain aldehydes
resulted in dramatic decrease in catalytic efficiencies, due to the
high Km values for these substrates. Unlike HpaI, aldehydes
substituted with C2- and C3-OH are not substrates of BphI.
Stereospecificity Control. Besides their distinct substrate
specificity, HpaI and BphI differ in the stereospecific control of
the aldol addition and cleavage reactions. HOPA synthesized
using Mn2þ-HpaI or Mn2þ-BphI have identical structures as
competitive inhibition constant Ki (mM)
2-ketoacid
HpaI
BpaI
glyoxylate
0.40 ( 0.04
0.51 ( 0.27
0.50 ( 0.04
3.6 ( 0.20
6.98 ( 0.33
0.21 ( 0.01
0.55 ( 0.03
0.34 ( 0.04
1.09 ( 0.10
2.71 ( 0.17
pyruvate
2-ketobutanoate
2-ketopentanoate
4-methyl-2-oxopentanoate
aAssays were performed by varying the inhibitor and substrate concen-
trations and contained 0.4 mM NADH, 2 mM divalent metal ions (Co2þ for
HpaI and Mn2þ for BphI), and 20 units of ADH in 100 mM HEPES buffer,
pH 8.0 at 25 °C, in a total volume of 1 mL.
1
determined by H and 13C NMR but differ in optical rotation
(Table S2). The optical rotation of HOPA produced by BphI is
similar to the (4S)-HOPA produced from catechol using catechol
dioxygenase (XylE), 2-hydroxy-6-oxo-2,4-heptadienoate hydro-
lase (TodF), and 2-hydroxypenta-2,4-dienoate hydratase (BphH)
Consistent with its ability to catalyze enolate formation in
pyruvate and 2-ketobutanoic acid, HpaI is able to catalyze cross-
aldol addition reactions between these carbonyl donors and
various aldehydes (Table 2). In particular, the enzyme is able
to utilize straight chain and branched chain aldehydes ranging
from two to five carbons. The highest catalytic efficiency
(kcat/Km) was obtained for succinic semialdeyde, the physiologi-
cal substrate of HpaI. A ∼2-fold decrease in Km and similar
kcat values were observed for glycoaldehyde compared to acet-
aldehyde, which lacks the C2-OH substituent. In comparison, an
11- and 170-fold decrease in catalytic efficiencies was observed
([R]25 = 15.4 and 12.3, respectively) (34), indicating that the
D
aldol addition reaction catalyzed by BphI only produces (4S)-
HOPA. On the other hand, the optical rotation of HOPA
synthesized by HpaI is 0.11, indicative of a racemic mixture of
(4R)- and (4S)-HOPA in approximately equal concentrations.
In addition, BphI was only able to cleave approximately 53% of
the HOPA synthesized by HpaI, leaving a HOPA product
(R-isomer) with an optical rotation of -11.3. On the other
hand, HpaI was able to cleave both (4R)- and (4S)-HOPA. The
4-hydroxy-2-oxo-1,7-heptanedioate, synthesized by HpaI, had
an optical rotation of 0.05, which indicates that the enzyme is
also not stereospecific for its physiological substrate.
Determination of Kinetic Mechanism of the Aldol Addi-
tion Reaction. Kinetic assays were performed using varying
concentrations of pyruvate and acetaldehyde in the aldol addi-
tion reaction. Lineweaver-Burk plots of the results show a series
of converging lines consistent with a classical ternary complex
bi-uni mechanism in both HpaI and BphI (Figure S2) (35). The
kcat of HpaI in the addition reaction is ∼200 times higher than
BphI (Table 3).
for isobutyraldehyde and (D,L)-glyceraldehyde, which have C3-
methyl and C2,C3-OH substitutions, respectively, compared to
the nonsubstituted propionaldehyde. Thus, C3-OH substitution
in aldehyde acceptors results in a dramatic decrease in catalytic
efficiency. Using 2-ketobutanoic acid as a carbonyl donor and
acetaldehyde as acceptor, HpaI showed a 10-fold decrease in
catalytic efficiency relative to pyruvate. This decrease is the result
of a lower kcat value, as Km differed by less than 2-fold.
BphI, on the other hand, can only utilize pyruvate as a
carbonyl donor and preferentially utilizes straight chain alde-
hydes as carbonyl acceptors. The enzyme has the highest
specificity constants (kcat/Km) for aldehydes of two to three
carbons in length. The similar specificity constants (kcat/Km)
In order to determine the order of substrate binding in the two
aldolases, product inhibition assays were further performed.
Both pyruvate and acetaldehyde competitively inhibit the aldol

3778 Biochemistry, Vol. 49, No. 17, 2010
Wang et al.
Table 2: Aldehyde Specificity of HpaI and BphI in the Aldol Addition Reactiona
HpaI
BphI
carbon
(kcat/Km)app
(ꢀ103 M-1 s-1
(kcat/Km)app
(M-1 s-1
)
length
structure
Km,app (mM) kcat,app (s-1
)
)
Km,app (mM) kcat,app (s-1
)
pyruvate as carbonyl donor
2
3
4
acetaldehyde
CH3CHO
62.9 ( 4.7
33.3 ( 6.1
32.9 ( 0.9
205.4 ( 5.2
175.5 ( 8.8
358.4 ( 3.5
5.6 ( 0.42
132.5 ( 4.6
64.9 ( 2.7
3.3 ( 0.3
5.27 ( 0.3
10.9 ( 0.3
0.063 ( 0.005 ND
9.9 ( 1.2
64.28 ( 3.72 0.86 ( 0.02 13.4 ( 0.80
glycolaldehyde
propionaldehyde
CH2(OH)CHO
CH3CH2CHO
124.6 ( 9.6
135.9 ( 9.6
0.40 ( 0.01 3.7 ( 0.31
1.79 ( 0.07 13.2 ( 1.1
(
D
,L)-glyceraldehyde
CH2(OH)CH(OH)CHO 88.6 ( 12.0
ND
ND
butyraldehyde
CH3(CH2)2CHO
(CH3)2CHCHO
COOH(CH2)2CHO
CH3(CH2)3CHO
13.4 ( 1.5
73.8 ( 6.2
9.1 ( 0.9
2.19 ( 0.12b
0.50 ( 0.04b
ND
isobutyraldehyde
succinic semialdehyde
pentaldehyde
0.9 ( 0.1
203.8 ( 7.7
22.5 ( 2.3
ND
ND
5
20.0 ( 0.3b
0.47 ( 0.01b
2-ketobutanoate as carbonyl
donor
2
acetaldehyde
CH3CHO
50.1 ( 4.3
21.9 ( 0.6
0.4 ( 0.04
ND
ND
ND
aAssays were performed at 25 °C and contained 7.7 μg of HpaI or 120 μg of BphI, 2 mM CoCl2 for HpaI and 2 mM MnCl2 for BphI, and 100 mM pyruvate
with varying concentrations of aldehydes. Reactions were 10 min in duration for HpaI and 120 min in duration for BphI. 4-Hyroxy-2-ketoacids formed were
lactonized and detected at 215 nm by HPLC analysis using an Aminex fast-acid analysis ion-exchange column. ND denotes no detectable activity. bDue to low
substrate solubility and the high apparent Km for this substrate, the specificity constant can be estimated from only the gradient of the specific activity vs
substrate concentration graph. All synthesized products were confirmed using HPLC and ESI-qTOF-MS (Table S1).
Table 3: Steady-State Kinetic Parameters of HpaI and BphI for the Aldol
Addition Reactiona
Table 4: HOPA Formation during an ADH Competition Assaya
HOPA concn (μM)
enzyme
kcat (s-1
)
KiA (mM)
KmA (mM)
KmB (mM)
[substrate]
acetyl-CoA
acetaldehyde
reaction time (min)
without ADH
with ADH
HpaI
BphI
219.5 ( 5.9
1.04 ( 0.05
2.01 ( 0.40
10.3 ( 0.10
5.6 ( 0.5
14.8 ( 1.4
62.1 ( 4.0
67.0 ( 6.3
7
15
7
175.4 ( 2.6
198.1 ( 1.0
13.0 ( 1.5
25.9 ( 0.5
140.8 ( 0.01
160.6 ( 5.6
ND
aAssays were performed at 25 °C and contained 7.7 μg of HpaI or 120 μg
of BphI, 2 mM CoCl2 for HpaI and 2 mM MnCl2 for BphI, and varying
concentrations of pyruvate (substrate A) and acetaldehyde (substrate B).
Reactions were 10 min in duration for HpaI and 120 min in duration for
BphI. Kinetic parameters were determined using a discontinuous assay
method as described in Experimental Procedures.
15
ND
aAssays were performed at 25 °C and contained 1 mM acetyl-CoA,
100 mM pyruvate, 1.5 mM NADH, and 2 mM MnCl2 in the presence and
absence of 20 units of ADH in 100 mM HEPES buffer (pH 8.0) in a total
volume of 1 mL. Lactonized HOPA was detected and quantified using
absorbance at 215 nm. ND denotes no detectable product.
cleavage of HOPA by HpaI, indicating that either compound is
able to bind to the free enzyme (Figure S3). Therefore, HpaI
follows a rapid equilibrium random order mechanism. On the
other hand, BphI exhibits competitive inhibition for pyruvate but
mixed inhibition for acetaldehyde, indicating that the enzyme
obeys a compulsory or preferred order mechanism in the aldol
addition reaction with pyruvate binding to the active site prior to
acetaldehyde. Similar results were obtained using Mn2þ-HpaI,
confirming that the distinct kinetic mechanisms between HpaI
and BphI are not due to the differences in metal cofactor used in
the assays.
Reverse Substrate Channeling in the BphI-BphJ Com-
plex. BphI forms a complex with the acetaldehyde dehydrogen-
ase, BphJ, permitting their active sites to be linked by a molecular
tunnel. We previously demonstrated that acetaldehyde, produced
from aldol cleavage of HOPA, can be directly channeled from the
aldolase to the dehydrogenase, where it is then converted to
acetyl-CoA in the presence of NADþ and coenzyme A (6). Since
BphJ can catalyze the reverse reaction, transforming acetyl-CoA
to acetaldehyde, we investigated if aldehyde channeling can occur
in reverse, and whether this would increase the efficiency of
HOPA synthesized by BphI. Reverse substrate channeling was
assessed by including alcohol dehydrogenase (ADH) in the assay
mixture, which would convert any acetaldehyde that is released to
the bulk solvent to ethanol. Table 4 shows a summary of the
results obtained from HPLC analyses of reaction products.
Control reactions using acetaldehyde in place of acetyl-CoA
showed that ADH effectively competes with BphI for exogenous
acetaldehyde since no HOPA can be detected. When acetyl-CoA
is used in place of acetaldehyde, HOPA was synthesized, and
addition of ADH in the reaction resulted in only a ∼20%
reduction of HOPA produced, suggesting that the majority of
acetaldehyde formed during the BphJ reductive deacylation
reaction is not released into the bulk solvent but is channeled
directly to BphI (∼80% channeling efficiency). Furthermore, a
16-fold greater concentration of HOPA is formed using acetyl-
CoA than an equivalent amount of acetaldehyde added exogen-
ously without ADH in the reaction mixture. The high Km for
acetaldehyde in BphI appears to be compensated by the direct
channeling of acetaldehyde from BphJ, allowing BphI to experi-
ence a higher concentration of this substrate.
Using NADH oxidation assay it is also possible to confirm and
quantify the reverse channeling efficiency (Figure 2). In the
absence of pyruvate, NADH oxidation rate by BphJ during
the conversion of acetyl-CoA to acetaldehyde is 3.85 ꢀ 10-2
μmol min-1 mg-1, and this rate increased 2-fold following
3
3
addition of ADH, as ADH will also oxidize NADH when
transforming the acetaldehyde released by BphJ into ethanol

Article
Biochemistry, Vol. 49, No. 17, 2010 3779
(Table 5). However, when pyruvate is present in the same assay,
NADH oxidation rates only differ by 1.l6-fold, whether ADH
was present or not, providing evidence that acetaldehyde is
channeled from BphJ to BphI to form HOPA and is unavailable
for oxidation by ADH. Upon addition of EDTA, which chelates
the Mn2þ cofactor and inactivates BphI, a 2-fold increase in
NADH oxidation is again observed upon addition of ADH.
From these results and eq 2, channeling efficiency was deter-
mined to be 84%, which is in agreement with our earlier estimate
from analysis of amounts of HOPA produced.
We also observed that the rate of reductive deacylation
of acetyl-CoA to acetaldehyde, catalyzed by BphJ, inc-
reased ∼4-fold upon addition of 100 mM pyruvate, but not
the pyruvate enolate analogue, oxalate, to the BphI-BphJ com-
plex. When BphI activity was inhibited through addition of
EDTA, this allosteric activation was abolished. Thus, bound
pyruvate or other intermediates in the aldol addition reaction
catalyzed by BphI activates BphJ reductive deacylation activity.
Active Site Architecture Analysis. The structures of HpaI
and BphI were compared using the crystal structure of HpaI from
the PDB (2V5K) and a model of BphI, built by using the
functionally similar homologous aldolase, DmpG, as a template
(56% sequence identity) (23). The active site of HpaI is formed
Table 5: Rates of NADH Oxidation in an ADH Competition Assay To
Test for Substrate Channeling in the BphI-BphJ Complexa
NADH oxidation activity (ꢀ10-2 μmol min-1 mg-1
)
3
3
0 mM
100 mM
pyruvate
100 mM pyruvate þ
pyruvate
10 mM EDTA
no ADH
3.85 ( 0.17
7.85 ( 0.12
16.6 ( 0.35
19.2 ( 0.67
3.92 ( 0.13
7.83 ( 0.19
with ADH
ratio (R)
2.03
1.16
84
2.00
0
CE (%)
-3 (∼0)
aAssays were performed at 25 °C and contained 1 mM acetyl-CoA,
100 mM pyruvate, 1.5 mM NADH, and 2 mM MnCl2 in the presence and
absence of 20 units of ADH in 100 mM HEPES buffer (pH 8.0) in a total
volume of 1 mL. R is the ratio ofNADH oxidation activity in the presence and
absence of ADH. CE is the aldehyde channeling efficiency calculated by eq 2.
˚
by ∼30 residues from adjacent dimers and consists of an ∼15 A
˚
deep bell-shaped cleft with an ∼12 A wide mouth. This broad
entrance to the active site is predominantly lined with noncharged
residues and a few positively charged residues (Figure 3A).
F
IGURE 3: Comparison of active site architecture between HpaI and BphI. (A) Top view surface representation of the entrance to the active site of
HpaI. The image is colored using generated electrostatics. Oxamate, an analogue of pyruvate enolate, is depicted in ball and stick representation.
The metal ion, Mg2þ, is shown as a green sphere. (B) Side view of the HpaI active site cavity is depicted in mesh showing generated electrostatics.
The metal ion ligands Glu149 and Asp175 are shown in sticks ligated to Mg2þ. (C) The active site of BphI, buried in the C terminus of the TIM
˚
˚
barrel, has a diameter of 9.0 A and width of 2-4 A. Entrance to the active site occurs through two molecular tunnels shown as gray mesh. (D) The
hydroxyl group of Tyr290 in BphI, shown in stick form, causes one side the active site to be narrowed. Based on the position of oxalate binding in
DmpG, pyruvate was modeled in ball and stick representation into the innermost part of the substrate binding site. The metal ligands Asp18,
His199, and His 201 are shown as sticks interacting with the Mn2þ, shown as a purple sphere. All images were generated in PyMOL.

3780 Biochemistry, Vol. 49, No. 17, 2010
Wang et al.
Oxamate, an analogue of pyruvate enolate, is coordinated to
the metal ion and is positioned on one side of the cleft, an
equivalent position as pyruvate in the YfaU structure (PDB
ordered movement of Tyr291 (Tyr290 in BphI) at the entrance of
the channel and the combined ordered movements of Ile172
,
Ile196, and Met198 (equivalent to Ile171, Leu195, and Met197 in
BphJ) at the exit of the channel (23). Our results indicate that
acetaldehyde can be efficiently channeled in the reverse direction,
providing the first known example of bidirectional substrate
channeling using the native substrate and demonstrating either
that gating through the channel does not occur or that the
proposed gating mechanism is reversible. Enzymes that exhibit
substrate channeling are often allosterically regulated (38-41),
ensuring that reactions occurring in separate active sites are
coupled. In the reverse channeling reaction, BphI activity was
demonstrated to stimulate the activity of BphJ by ∼4-fold. A
similar relationship was previously observed in forward channel-
ing where catalysis of BphJ stimulates activity of BphI ∼5-
fold (6). This suggests that allosteric activation may be bidirec-
tional or that the presence of pyruvate leads to the rapid
conversion of acetaldehyde to HOPA resulting in faster release
of products from the BphJ active site.
2þ
˚
2VWT, 53% sequence identity, rmsd 0.6 A to HpaI-Mg
-
oxamate complex) (18). There is space distal to the pyruvate
methyl carbon that can accommodate large aldehydes
(Figure 3B).
The model of BphI displays a small internal active site deeply
buried in the C-terminus of the TIM barrel, with a diameter of
˚
˚
9.0 A and width of 2-4 A. Noticeably, the hydroxyl group of
Tyr290 penetrates into the active site causing the side perpendi-
˚
cular to the metal ligands to be narrowed. A 10 A long narrow
tunnel leads to the bulk solvent, allowing substrates to enter. A
second tunnel, composed of hydrophobic residues, presumably
connects to the active site of BphJ, to allow for channeling of
aldehydes from the aldolase to the dehydrogenase (Figure 3C).
Based on the position of oxalate binding in DmpG, pyruvate is
proposed to bind to the innermost part of the substrate bind-
ing site, with the two carboxylate oxygen atoms (O1 and O3)
of pyruvate forming equatorial ligands to the metal ion
(Figure 3D) (23). This leaves a small space for aldehyde binding
close to the entrance.
Aldolases homologous to HpaI are not restricted to aromatic
catabolic pathways. These homologues include YfaU (18), in-
volved in rhamnose degradation, and 2-dehydro-3-deoxygalac-
tarate aldolase (12), involved in the galactarate degradation
pathway in E. coli. The products of these aldolases are pyruvate
and long aliphatic and hydrophilic aldehydes. The large active
sites of HpaI and homologues allow them to utilize diverse
substrates which may explain the recruitment of these enzymes
in diverse metabolic pathway. In comparison, BphI and its
homologues are mainly found in aromatic degradative path-
ways (42-44). The strict substrate specificity resulting from small
active sites of the enzymes may limit the recruitment of these
enzymes in aromatic degradation pathways, where aldol cleavage
products are pyruvate and a small volatile and toxic aldehyde,
such as acetaldehyde and propionaldehyde. Therefore, these
aldolases are always associated with a dehydrogenase that
converts toxic aldehydes to acyl-CoA. Although none of the
other HMGL-like family members associate with another en-
zyme or exhibit substrate channeling, most of them possess an
additional domain at the C-terminus of the TIM barrel domain,
which can perhaps evolve to form associations with other
enzymes or to regulate catalysis (45).
The apparent broad substrate specificity of HpaI, combined
with a high turnover number, is an advantage for its use as a
biocatalyst for carbon-carbon bond formation reactions in
organic synthesis. However, this enzyme shows poor stereospe-
cificity. Our results demonstrate that the use of BphI or HpaI and
BphI in tandem is a promising method to produce compounds
with only (S)- or (R)-conformations at C4. In addition, the
reverse substrate channeling exhibited by the BphI-BphJ com-
plex poses a unique strategy for organic synthesis, through the
use of acyl-CoAs in place of aldehydes as cosubstrates to achieve
higher product transformation efficiency. The comparison of
kinetic mechanisms and substrate specificities will guide ongoing
structural and mutagenesis studies aimed at designing pyruvate
aldolases with novel substrate specificities and stereospecificities.
DISCUSSION
HpaI and BphI are class II pyruvate aldolases that catalyze
similar reactions in aromatic meta-cleavage pathway but exhibit
no sequence similarities and differ in detailed structural topology.
Analysis of the two enzymes in the aldol addition reactions
revealed intriguing differences in their kinetic mechanisms,
stereospecificity, and substrate specificity which can be reason-
ably correlated to the architecture of their respective active sites.
HpaI has a wide entrance leading from the bulk solvent to the
active site which may explain the rapid equilibrium random order
kinetic mechanism observed in the aldol addition reaction. In
contrast, the active site of BphI is smaller, with the aldehyde
binding site closer to the entrance and pyruvate binding site in the
interior. Therefore, pyruvate binding must precede aldehyde
binding in this enzyme in order for the aldol addition to occur.
Similarly, aldehyde produced in the aldol cleavage reaction must
dissociate from the active site and be channeled prior to
dissociation of pyruvate. The small size of the BphI active site
may also cause steric constraints which allow pyruvate enolate to
only attack the re face of acetaldehyde, thereby forming (4S)-
HOPA (Figure S1). In contrast, the larger substrate binding site
of HpaI enables the enzyme to utilize 2-ketobutyrate in place of
pyruvate as the carbonyl donor and bulky aldehydes as carbonyl
acceptors and permits aldehydes to bind in alternative conforma-
tions, leading to formation of racemic products. The aldol
addition reaction catalyzed by 2-keto-3-deoxygluconate aldolase
from Solfolobus solfataricus has also been reported to be non-
stereospecific (36). Thus, the poor facial selectivity observed in
HpaI is not unique among aldolases.
We have previously demonstrated that acetaldehyde and
propionaldehyde, products of the BphI aldol cleavage reaction,
are directly channeled to the dehydrogenase (6). Substrate
channeling in other bifunctional complexes with a molecular
tunnel, such as tryptophan synthase, has only been shown
experimentally to occur in one direction (37). Structural analysis
of the homologous aldolase-dehydrogenase complex (DmpG-
DmpF) from the phenol degradation pathway has suggested that
acetaldehyde channeling is finely controlled and gated by the
ACKNOWLEDGMENT
The authors thank Dan Pan for assistance in BphI-BphJ
purification and Dyanne Brewer from the University of Guelph’s
Advanced Analysis Centre for mass spectrometry spectra acqui-
sition.

Article
Biochemistry, Vol. 49, No. 17, 2010 3781
SUPPORTING INFORMATION AVAILABLE
18. Rea, D., Hovington, R., Rakus, J. F., Gerlt, J. A., Fulop, V., Bugg,
T. D., and Roper, D. I. (2008) Crystal structure and functional
assignment of YfaU, a metal ion dependent class II aldolase from
Escherichia coli K12. Biochemistry 47, 9955–9965.
19. Forouhar, F., Hussain, M., Farid, R., Benach, J., Abashidze, M.,
Edstrom, W. C., Vorobiev, S. M., Xiao, R., Acton, T. B., Fu, Z., Kim,
J. J., Miziorko, H. M., Montelione, G. T., and Hunt, J. F. (2006)
Crystal structures of two bacterial 3-hydroxy-3-methylglutaryl-CoA
lyases suggest a common catalytic mechanism among a family of TIM
barrel metalloenzymes cleaving carbon-carbon bonds. J. Biol. Chem.
281, 7533–7545.
20. Tatusov, R. L., Natale, D. A., Garkavtsev, I. V., Tatusova, T. A.,
Shankavaram, U. T., Rao, B. S., Kiryutin, B., Galperin, M. Y.,
Fedorova, N. D., and Koonin, E. V. (2001) The COG database: new
developments in phylogenetic classification of proteins from complete
genomes. Nucleic Acids Res. 29, 22–28.
Facial selection of the carbonyl donor pyruvate enolate in
aldol addition reaction of 4-hydroxy-2-oxopentanoate (HOPA)
by HpaI and BphI (Figure S1), Lineweaver-Burk plots of HpaI
and BphI in the aldol addition reaction (Figure S2), Lineweaver-
Burk plots of the product inhibition of HpaI and BphI in the
HOPA aldol cleavage reaction (Figure S3), HPLC and Q-Tof-
MS analysis of products synthesized by aldolases (Table S1), and
optical rotation and NMR analysis of HOPA produced in aldol
addition reaction by HpaI and BphI (Table S2). This material is
available free of charge via the Internet at http://pubs.acs.org.
21. Tatusov, R. L., Galperin, M. Y., Natale, D. A., and Koonin, E. V.
(2000) The COG database: a tool for genome-scale analysis of protein
functions and evolution. Nucleic Acids Res. 28, 33–36.
22. Rea, D., Fulop, V., Bugg, T. D., and Roper, D. I. (2007) Structure and
mechanism of HpcH: a metal ion dependent class II aldolase from the
homoprotocatechuate degradation pathway of Escherichia coli. J.
Mol. Biol. 373, 866–876.
23. Manjasetty, B. A., Powlowski, J., and Vrielink, A. (2003) Crystal
structure of a bifunctional aldolase-dehydrogenase: sequestering a
reactive and volatile intermediate. Proc. Natl. Acad. Sci. U.S.A. 100,
6992–6997.
24. Takayama, S., McGarvey, G. J., and Wong, C. H. (1997) Microbial
aldolases and transketolases: new biocatalytic approaches to simple
and complex sugars. Annu. Rev. Microbiol. 51, 285–310.
25. Samland, A. K., and Sprenger, G. A. (2006) Microbial aldolases as
C-C bonding enzymes;unknown treasures and new developments.
Appl. Microbiol. Biotechnol. 71, 253–264.
REFERENCES
1. Furukawa, K., and Kimura, N. (1995) Biochemistry and genetics of
PCB metabolism. Environ. Health Perspect. 103 (Suppl. 5), 21–23.
2. Diaz, E. (2004) Bacterial degradation of aromatic pollutants: a
paradigm of metabolic versatility. Int. Microbiol. 7, 173–180.
3. Van der Geize, R., Yam, K., Heuser, T., Wilbrink, M. H., Hara, H.,
Anderton, M. C., Sim, E., Dijkhuizen, L., Davies, J. E., Mohn, W. W.,
and Eltis, L. D. (2007) A gene cluster encoding cholesterol catabolism
in a soil actinomycete provides insight into Mycobacterium tubercu-
losis survival in macrophages. Proc. Natl. Acad. Sci. U.S.A. 104,
1947–1952.
4. Furukawa, K., Hirose, J., Suyama, A., Zaiki, T., and Hayashida, S.
(1993) Gene components responsible for discrete substrate specificity
in the metabolism of biphenyl (bph operon) and toluene (tod operon).
J. Bacteriol. 175, 5224–5232.
5. Furukawa, K., Simon, J. R., and Chakrabarty, A. M. (1983) Common
induction and regulation of biphenyl, xylene/toluene, and salicylate
catabolism in Pseudomonas paucimobilis. J. Bacteriol. 154, 1356–1362.
6. Baker, P., Pan, D., Carere, J., Rossi, A., Wang, W., and Seah, S. Y.
(2009) Characterization of an aldolase-dehydrogenase complex that
exhibits substrate channeling in the polychlorinated biphenyls degra-
dation pathway. Biochemistry 48, 6551–6558.
26. Pollard, J. R., Rialland, D., and Bugg, T. D. (1998) Substrate
selectivity and biochemical properties of 4-hydroxy-2-keto-pentanoic
acid aldolase from Escherichia coli. Appl. Environ. Microbiol. 64,
4093–4094.
27. Cornish-Bowden, A. (1995) Analysis of Enzyme Kinetic Data, Oxford
University Press, New York.
7. Prieto, M. A., Diaz, E., and Garcia, J. L. (1996) Molecular character-
ization of the 4-hydroxyphenylacetate catabolic pathway of Escher-
ichia coli W: engineering a mobile aromatic degradative cluster.
J. Bacteriol. 178, 111–120.
8. Wang, W., and Seah, S. Y. (2005) Purification and biochemical
characterization of a pyruvate-specific class II aldolase, HpaI.
Biochemistry 44, 9447–9455.
9. Wang, W., and Seah, S. Y. (2008) The role of a conserved histidine
residue in a pyruvate-specific class II aldolase. FEBS Lett. 582, 3385–
3388.
28. Cornish-Bowden, A. (1995) Fundamentals of Enzyme Kinetics,
Portland Press, London.
29. Eswar, N., John, B., Mirkovic, N., Fiser, A., Ilyin, V. A., Pieper, U.,
Stuart, A. C., Marti-Renom, M. A., Madhusudhan, M. S., Yerkovich,
B., and Sali, A. (2003) Tools for comparative protein structure
modeling and analysis. Nucleic Acids Res. 31, 3375–3380.
30. Eswar, N., Webb, B., Marti-Renom, M. A., Madhusudhan, M. S.,
Eramian, D., Shen, M. Y., Pieper, U., Sali, A. (2006) Comparative
protein structure modeling using Modeller, Current Protocols in
Bioinformatics, Chapter 5, Unit 5 6.
10. Ash, D. E., Emig, F. A., Chowdhury, S. A., Satoh, Y., and Schramm,
V. L. (1990) Mammalian and avian liver phosphoenolpyruvate
carboxykinase. Alternate substrates and inhibition by analogues of
oxaloacetate. J. Biol. Chem. 265, 7377–7384.
11. Gallo, A. A., and Sable, H. Z. (1973) Rate enhancement of pyruvate
aldolization by divalent cations: a model for class II aldolases.
Biochim. Biophys. Acta 302, 443–456.
12. Izard, T., and Blackwell, N. C. (2000) Crystal structures of the metal-
dependent 2-dehydro-3-deoxy-galactarate aldolase suggest a novel
reaction mechanism. EMBO J. 19, 3849–3856.
13. Choi, K. H., Shi, J., Hopkins, C. E., Tolan, D. R., and Allen, K. N.
(2001) Snapshots of catalysis: the structure of fructose-1,6-(bis)-
phosphate aldolase covalently bound to the substrate dihydroxyace-
tone phosphate. Biochemistry 40, 13868–13875.
31. Eswar, N., Eramian, D., Webb, B., Shen, M. Y., and Sali, A. (2008)
Protein structure modeling with MODELLER. Methods Mol. Biol.
426, 145–159.
32. Pawlowski, M., Gajda, M. J., Matlak, R., and Bujnicki, J. M. (2008)
MetaMQAP: a meta-server for the quality assessment of protein
models. BMC Bioinf. 9, 403.
33. DeLano, W. L. (2002) The PyMOL Molecular Graphics System,
DeLano Scientific, San Carlos, CA.
34. Collinsworth, W. L., Chapman, P. J., and Dagley, S. (1973) Stereo-
specific enzymes in the degradation of aromatic compounds by
Pseudomonas putida. J. Bacteriol. 113, 922–931.
35. Cleland, W. W., and Cook, P. F. (2007) Enzyme Kinetics and
Mechanism, Garland Science, New York, NY.
36. Lamble, H. J., Heyer, N. I., Bull, S. D., Hough, D. W., and Danson,
M. J. (2003) Metabolic pathway promiscuity in the archaeon Sulfo-
lobus solfataricus revealed by studies on glucose dehydrogenase and
2-keto-3-deoxygluconate aldolase. J. Biol. Chem. 278, 34066–34072.
37. Dunn, M. F., Aguilar, V., Brzovic, P., Drewe, W. F., Jr., Houben,
K. F., Leja, C. A., and Roy, M. (1990) The tryptophan synthase
bienzyme complex transfers indole between the alpha- and beta-sites
14. Heine, A., Luz, J. G., Wong, C. H., and Wilson, I. A. (2004) Analysis
of the class I aldolase binding site architecture based on the crystal
˚
structure of 2-deoxyribose-5-phosphate aldolase at 0.99A resolution.
J. Mol. Biol. 343, 1019–1034.
15. Morse, D. E., and Horecker, B. L. (1968) The mechanism of action of
aldolases. Adv. Enzymol. Relat. Areas Mol. Biol. 31, 125–181.
16. Schmitzberger, F., Smith, A. G., Abell, C., and Blundell, T. L. (2003)
Comparative analysis of the Escherichia coli ketopantoate hydroxy-
methyltransferase crystal structure confirms that it is a member of the
(betaalpha)8 phosphoenolpyruvate/pyruvate superfamily. J. Bacter-
iol. 185, 4163–4171.
17. Narayanan, B. C., Niu, W., Han, Y., Zou, J., Mariano, P. S.,
Dunaway-Mariano, D., and Herzberg, O. (2008) Structure and func-
tion of PA4872 from Pseudomonas aeruginosa, a novel class of
oxaloacetate decarboxylase from the PEP mutase/isocitrate lyase
superfamily. Biochemistry 47, 167–182.
˚
via a 25-30 A long tunnel. Biochemistry 29, 8598–8607.
38. Casino, P., Niks, D., Ngo, H., Pan, P., Brzovic, P., Blumenstein, L.,
Barends, T. R., Schlichting, I., and Dunn, M. F. (2007) Allosteric
regulation of tryptophan synthase channeling: the internal aldimine
probed by trans-3-indole-30-acrylate binding. Biochemistry 46, 7728–7739.
39. Ngo, H., Kimmich, N., Harris, R., Niks, D., Blumenstein, L., Kulik,
V., Barends, T. R., Schlichting, I., and Dunn, M. F. (2007) Allosteric
regulation of substrate channeling in tryptophan synthase: modula-
tion of the L-serine reaction in stage I of the beta-reaction by alpha-site
ligands. Biochemistry 46, 7740–7753.

3782 Biochemistry, Vol. 49, No. 17, 2010
Wang et al.
40. Hawkins, A. R., and Lamb, H. K. (1995) The molecular biology of
multidomain proteins. Selected examples. Eur. J. Biochem. 232, 7–18.
41. Bacher, I., Zidar, A., Kratzel, M., and Hohenegger, M. (2004)
Channelling of substrate promiscuity of the skeletal-muscle ADP-
ribosyl cyclase isoform. Biochem. J. 381, 147–154.
42. Lee, S. J., Ko, J. H., Kang, H. Y., and Lee, Y. (2006) Coupled
expression of MhpE aldolase and MhpF dehydrogenase in Escher-
ichia coli. Biochem. Biophys. Res. Commun. 346, 1009–1015.
4-hydroxy-2-oxovalerate aldolase in Pseudomonas sp. DJ77 and their
evolutionary implications. Biochem. Biophys. Res. Commun. 256,
469–473.
44. Powlowski, J., Sahlman, L., and Shingler, V. (1993) Purification and
properties of the physically associated meta-cleavage pathway enzymes
4-hydroxy-2-ketovalerate aldolase and aldehyde dehydrogenase (acylating)
from Pseudomonas sp. strain CF600. J. Bacteriol. 175, 377–385.
45. Koon, N., Squire, C. J., and Baker, E. N. (2004) Crystal structure of
LeuA from Mycobacterium tuberculosis, a key enzyme in leucine
biosynthesis. Proc. Natl. Acad. Sci. U.S.A. 101, 8295–8300.
43. Hwang, S., Kim, S. J., Kim, C. K., Kim, Y., and Kim, Y. C. (1999) The
phnIJ genes encoding acetaldehyde dehydrogenase (acylating) and