Structure-activity analysis of base and enzyme-catalyzed 4-hydroxybenzoyl coenzyme A hydrolysis

Article abstract of DOI:10.1016/j.bioorg.2006.07.002

In this study, the second-order rate constant k₂ of base-catalyzed hydrolysis and the values of k_cat, K_m and k_cat/K_m of wild-type Pseudomonas sp. CBS3 4-hydroxybenzoyl coenzyme A (4-HBA-CoA) thioesterase-catalyzed hydrolysis of 4-HBA-CoA and its para-substituted analogs were measured. For the base-catalyzed hydrolysis, the plot of log k₂ vs the σ value of the para-substituents was linear with a slope (ρ) of 1.5. In the case of the enzyme-catalyzed hydrolysis, the k_cat/K_m values measured for the para-substituted analogs defined substrate specificity. Asp32 was shown to play a key role in substrate recognition, and in particular, in the discrimination between the targeted substrate and other cellular benzoyl-CoA thioesters.

Full text of DOI:10.1016/j.bioorg.2006.07.002

Bioorganic Chemistry 35 (2007) 1–10
www.elsevier.com/locate/bioorg
Structure–activity analysis of base and
enzyme-catalyzed 4-hydroxybenzoyl coenzyme
A hydrolysis
1
*
Feng Song, Zhihao Zhuang , Debra Dunaway-Mariano
Department of Chemistry, University of New Mexico, Albuquerque, NM 87131, USA
Received 10 May 2006
Available online 7 September 2006
Abstract
In this study, the second-order rate constant k2 of base-catalyzed hydrolysis and the values of kcat
,
Km and kcat/Km of wild-type Pseudomonas sp. CBS3 4-hydroxybenzoyl coenzyme A (4-HBA-CoA)
thioesterase-catalyzed hydrolysis of 4-HBA-CoA and its para-substituted analogs were measured.
For the base-catalyzed hydrolysis, the plot of logk2 vs the r value of the para-substituents was linear
with a slope (q) of 1.5. In the case of the enzyme-catalyzed hydrolysis, the kcat/Km values measured
for the para-substituted analogs defined substrate specificity. Asp32 was shown to play a key role in
substrate recognition, and in particular, in the discrimination between the targeted substrate and
other cellular benzoyl-CoA thioesters.
ꢀ 2006 Elsevier Inc. All rights reserved.
Keywords: Thioester hydrolysis; Specific base catalysis; Substrate specificity; 4-HBA-CoA; Thioesterase; Struc-
ture–activity; SAR; Enzyme catalysis
1. Introduction
Thioesterases are ubiquitous enzymes that catalyze the hydrolysis of acyl-thioesters in a
variety of cellular contexts. Thioesterases play essential physiological roles in fatty acid
*
Corresponding author. Fax: +1 505 277 6202.
E-mail address: dd39@unm.edu (D. Dunaway-Mariano).
Present address: Department of Chemistry, Pennsylvania State University, 104 Chemistry Building,
1
University Park, PA 16802, USA.
0045-2068/$ - see front matter ꢀ 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2006.07.002

2
F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
Fig. 1. The reaction steps of 4-chlorobenzoate degradation of Pseudomonas sp. Strain CBS3 [17].
and polyketide biosynthesis [1–3], aromatic compound degradation [4–6], acylated protein
degradation [7,8], signal transduction [9] and lipid metabolism [10]. Most of thioesterases
evolved within the a/b-hydrolase fold superfamily [11] or within the hotdog-fold enzyme
superfamily [12]. The catalytic triad-based mechanism of amide, oxygen ester and thioester
hydrolysis mediated by the enzymes of the a/b-hydrolase fold superfamily is well charac-
terized as is the mechanism used for substrate recognition [13]. In contrast, little is known
about the mechanisms of catalysis and substrate recognition employed by the hotdog-fold
thioesterases.
The first hotdog-fold thioesterase structure reported was that of the 4-hydroxybenzoyl-
CoA2 (4-HBA-CoA) thioesterase [14,15]. This enzyme catalyzes the final step of 4-chloro-
benzoate degradation pathway in the soil-dwelling bacterium Pseudomonas sp. strain
CBS3 [4,16] (Fig. 1). The native structure is a dimer of dimers. Each monomer consists
of a long a-helix wrapped by a 6-stranded b-sheet. Two monomers align to form a func-
tional dimer having a 12-stranded sheet structure with the two a-helices aligned anti-par-
allel in a slight off set. The active site (Fig. 2) is located at the subunit interface. Two
dimers associate to form the tetrameric structure. The nucleotide region of the substrate
CoA moiety rests on the protein surface while the pantenoate arm extends into the active
site crevice, desolvating the 4-hydroxybenzoyl thioester unit [15,17]. The substrate thioes-
ter C@O is positioned to engage in hydrogen bond interaction with the N-terminus of the
monomer A a-helix, while the aromatic ring extends into the active site and approaches
the a-helix of the opposing monomer B. The active site contains two carboxylate residues.
Asp17, which functions in general base or nucleophilic catalysis, is located at the loop
region that terminates in the N-terminus of a-helix A. Previous studies of site-directed
mutants have shown that Asp17 is essential for catalysis [17]. Asp32, located on the a-helix
B binds the ring C(4)OH via two bridging water molecules (Fig. 2).
In this paper, we report on the mechanism of action of the Pseudomonas sp. CBS3 4-
HBA-CoA thioesterase with special focus on comparative substrate–activity relationships
(SARs) in the specific base-catalyzed and enzyme-catalyzed hydrolysis of the C(4) substi-
tuted benzoyl-CoA substrates. In this manner, the role of the indirect interaction between
Asp32 and the substrate benzoyl para-OH in substrate recognition is demonstrated.
2
Abbreviations used: CoA, coenzyme A; 4-HBA-CoA, 4-hydroxybenzoyl-coenzyme A; SARs, substrate–
activity relationships; PCR, polymerase chain reaction; IPTG, isopropyl-b-D-galactopyranoside; DTT,
dithiothreotal; 4-FBA-CoA, 4-fluorobenzoyl-CoA; 4-CF3BA-CoA, 4-trifluorobenzoyl-CoA; 4-CBA-CoA, 4-
chlorobenzoyl-CoA; 4-MBA-CoA, 4-methylbenzoyl-CoA; 4-MeOBA-CoA, 4-methoxybenzoyl-CoA; BA-CoA,
benzoyl-CoA; DTNB, 5,50-dithio-bis(2-nitrobenzoic acid); K+Hepes, potassium salt of N-(2-hydroxyethyl)pip-
erzine-N0-2-ethanesulfonic acid; Caps, 3-(cyclohexylamino)-1-propanesulfonate; 4-HP-CoA, 4-hydroxyphenacyl-
CoA; 4-CBA, 4-chlorobenzoate; 4-HBA, 4-hydroxybenzoate.

F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
3
Fig. 2. Stereo active site picture of the complex of Pseudomonas sp. strain CBS3 4-HBA-CoA thioesterase (blue)/
4-HP-CoA (green, truncated) superimposed with the complex of D17N (cyan)/4-HBA-CoA (red, truncated). The
residue-residue and residue-substrate interactions are shown on the complex of WT/4-HP-CoA as black dash line
[15]. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of
this paper.)
2. Materials and methods
2.1. Materials
The recombinant Pseudomonas sp. strain CBS3 4-HBA-CoA thioesterase [15,17] was
prepared from the Escherichia coli clone as described in ref 16. Mutagenesis was carried
out using a PCR-based strategy [18] with the wild-type thioesterase gene as template, com-
mercial primers from Invitrogen, the PCR kit supplied by Stratagene, and the Power
Block IITM System thermal cycler manufactured by ERICOMP. PCR-amplified DNA
was cloned into the pET-3a vector (Novagen) for expression in E. coli HMS174 (DE3).
The mutated gene was verified by DNA sequencing. The mutant protein was purified to
homogeneity (as judged by SDS–PAGE analysis) using the same procedure as used for
purification of the wild-type thioesterase. 4-HBA-CoA, 4-fluorobenzoyl-CoA (4-FBA-
CoA), 4-trifluorobenzoyl-CoA (4-CF3BA-CoA), 4-chlorobenzoyl-CoA (4-CBA-CoA),
4-methylbenzoyl-CoA (4-MBA-CoA) and 4-methoxybenzoyl-CoA (4-MeOBA-CoA) were
synthesized as reported in [15,19]. Benzoyl-CoA (BA-CoA) was purchased from Sigma.
2.2. Determination of steady-state kinetic constants for Pseudomonas thioesterase wild-type
and D32S-catalyzed hydrolysis of substrate and analogs
The kcat and Km values were determined from initial velocity data. The wild-type
enzyme-catalyzed hydrolysis of 4-HBA-CoA was measured as described in ref 17. The
enzyme-catalyzed hydrolysis of 4-HBA-CoA, 4-MeOBA-CoA, 4-MBA-CoA, BA-CoA,

4
F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
4-FBA-CoA, 4-CBA-CoA, and 4-CF3BA-CoA were monitored at 25 ꢁC by measuring the
absorbance of 5-thio-2-nitrobenzoate at 412 nm (De = 13.6 mMꢀ1 cmꢀ1), which was
formed through the reaction of the 5,50-dithio-bis(2-nitrobenzoic acid) (DTNB) with the
CoASH product. Each reaction mixture (200 lL) contained a catalytic amount of thioes-
terase (0.22 lM D32S for 4-HBA-CoA, 5.4 lM wild-type or 2.3 lM D32S for 4-MeOBA-
CoA, 5.4 lM wild-type or 2.3 lM D32S for 4-MBA-CoA, 2.2 lM wild-type or 0.52 lM
D32S for BA-CoA, 0.052 lM wild-type or 0.23 lM D32S for 4-FBA-CoA, 2.2 lM
wild-type or 0.52 lM D32S for 4-CBA-CoA, and 21.6 lM wild-type or 2.3 lM D32S
for 4-CF3BA-CoA, respectively) and a varied concentration of substrate (0.5 Km ꢁ 5
Km), 2 mM DTNB, and 50 mM K+Hepes, pH 7.5. The initial velocity data were analyzed
using the equation V = Vmax [S]/([S] + Km) (V = initial velocity, Vmax = maximum veloc-
ity, [S] = substrate concentration, Km = Michaelis constant). The kcat was calculated from
Vmax/[E] where [E] is the total enzyme active site concentration (protein concentration was
determined using the Bradford method [20]) and the KinetAsyst program (IntelliKinetics,
State College, PA).
2.3. Determination of the rate constant for base-catalyzed hydrolysis of the para-substituted
benzoyl-CoA thioesters
A Beckman DU640 spectrophotometer was used to monitor the absorbance changes in
the reaction solutions of the para-substituted benzoyl-CoA thioesters. Typically a 5 lL ali-
quot of the para-substituted benzoyl-CoA thioester stock solution was added to 195 lL of
50 mM Caps (for pH 11–12) or NaOH solution (for pH above and equal to 12) contained
in a quartz cuvette (1 cm light path). The ionic strength of the reaction solution was adjust-
ed to 3 M with NaCl. The molar extinction coefficients of 4-HBA-CoA and its analogs at
300 nm were determined as a function of reaction solution pH. The reaction was moni-
tored at 300 nm at 25 ꢁC and the first-order rate constant was obtained by fitting the data
to the single exponential equation At = A0 + P exp(ꢀkobst), where At is the absorbance at
time t, A0 is initial absorbance, P is the total absorbance change, kobs is the observed rate
constant. The kobs was used to calculate the second-order rate constant k2 with the equa-
tion k2 = kobs/[OHꢀ]. The reported k2 is the average of several measures at different pH
values.
2.4. Calculation of DDGbind and DDGà
Values for DDGbind (binding energy at the ground state) and DDGà (binding energy at
the transition state) were calculated according to ref [21] using the equations:
DDGbind ¼ RT lnðKdobs; without group present=Kodbs; with group present
Þ
obs; with group present
DDGz ¼ ꢀRT ln½ðkcat=KmÞobs; without group present=ðkcat=KmÞ
ꢂ
where R is the gas constant (1.987 cal molꢀ1 Kꢀ1) and T is temperature in Kelvin (298 K).
Kodbs; without group present and Kodbs; with group present are the dissociation constants of the
enzyme–ligand complex measured in the absence and in the presence of the functional
group that contributes to substrate binding, respectively. For slow substrates, Km values
were considered as an approximation of Kd (when chemical steps are rate-limiting,
without group present
with group present
Km = kꢀ1/k1 = Kd). (kcat/Km)obs,
and (kcat/Km)obs,

F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
5
represent the kcat/Km ratio of reactions in the absence of the functional group and in the
presence of the functional group.
3. Results and discussion
3.1. Chemical model of 4-HBA-CoA catalysis
The chemical reaction catalyzed by 4-HBA-CoA thioesterase is rather simple: nucleo-
philic displacement of a thiolate by a water molecule (Fig. 3). Oxygen ester and thioester
hydrolysis both occur through an addition–elimination pathway, but only the hydrolysis
of the oxygen ester benefits from proton assisted departure of the leaving group [22–24].
The thioester hydrolysis in solution should occur faster at alkaline pH than at neutral
pH because hydroxide is present to serve as the nucleophile in place of the water molecule.
Indeed, whereas solution hydrolysis of 4-HBA-CoA and its analogs was too slow to detect
at 25 ꢁC, pH 7, the hydrolysis at >pH 10 was not. Thus, the values of first-order rate con-
stant kobs for 4-HBA-CoA and the varied para-substituted benzoyl-CoA thioesters (MeO,
Me, H, F, Cl and CF3) were measured at pH 11.0–13.5. The second-order rate constant k2
for each pH value was calculated by the equation k2 = kobs/[OHꢀ] and the final k2 values
for the natural substrate and its analogs were derived from the average of the data mea-
sured at several pH values (Table 1).
The correlation between the rate constant (logk2) of base-catalyzed hydrolysis of ben-
zoyl-CoA with various para-substituents and the para-substituent constant r [25], which
Fig. 3. The reaction pathway for hydroxide ion displacement of the CoA thiolate from 4-HBA-CoA.
Table 1
The steady-state kinetic constants determined for wild-type 4-HBA-CoA thioesterase-catalyzed hydrolysis of the
natural substrate 4-HBA-CoA, and the substrate analogs 4-MeOBA-CoA, 4-MeBA-CoA, BA-CoA, 4-FBA-
CoA, 4-CBA-CoA, and 4-CF3BA-CoA measured at pH 7.5 and 25 ꢁC
Substrate
kcat (sꢀ1
)
Km (lM)
kcat/Km (Mꢀ1 sꢀ1
)
k2 (Mꢀ1 sꢀ1
)
ASC (rp)a
4-HBA-CoA
4-MeOBA-CoA
4-MeBA-CoA
BA-CoA
4-FBA-CoA
4-CBA-CoA
4-CF3BA-CoA
(1.83 0.03) · 1ꢀ011
(2.7 0.2) · 10
(4.7 0.1) · 10ꢀ2
1.00 0.03
6.0 0.4
3.1 · 106
4.8 · 103
2.0 · 102
2.0 · 103
1.9 · 104
2.0 · 103
1.3 · 102
1.8 · 10ꢀ3
6.0 · 10ꢀ2
8.5 · 10ꢀ2
1.2 · 10ꢀ1
2.5 · 10ꢀ1
3.2 · 10ꢀ1
9.6 · 10ꢀ1
ꢀ0.22
ꢀ0.27
ꢀ0.14
0.00
+0.15
+0.24
+0.53
(5.6 0.6) · 101
(2.3 0.2) · 102
(5.1 0.3) · 102
(5.2 0.6) · 102
(5.5 0.1) · 102
(3.0 0.3) · 102
9.7 0.7
1.10 0.01
(3.8 0.1) · 10ꢀ2
See Section 2 for details.
a
Average substituent constant for the substrate benzoyl ring C(4) substituent taken from Ref [25].

6
F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
8
6
4
2
0
CF3
Cl
F
H
0
Me
MeO
-2
OH
-0.2
-4
-0.4
0.2
0.4
0.6
Sigma Value (σ)
Fig. 4. The dependence of base-catalyzed hydrolysis rate constant (Logk2 (Mꢀ1 sꢀ1), j, (r is for OH)) or
thioesterase-catalyzed hydrolysis rate constant (log(kcat/Km) (Mꢀ1 sꢀ1), d) of benzoyl-CoAs with different para-
substituents upon the para-substituent r value.
was listed in Table 1, was measured. The plot of logk2 vs the r value was linear with a
slope (q) of 1.5 (see Fig. 4). The base-catalyzed hydrolysis of 4-HBA-CoA deviated from
the fitted line when the r value of para-hydroxyl group was plotted. This is because at
alkaline pH the C(4)OH is ionized (pKa of the 4-HBA-CoA hydroxyl group is 8.6 [26]).
The k2 of the unionized 4-HBA-CoA predicted by the logk2 vs r plot of Fig. 4, is
6.5 · 10ꢀ2 Mꢀ1 sꢀ1
.
The q value is the reaction constant which measures the susceptibility of the reaction to
the influence of the substituents and the extent to which negative charge builds in the rate-
limiting transition state. The q value measured for alkaline hydrolysis of para-substituted
benzoyl-ethyl esters is 2.19 (measured in 75% methanol at 25 ꢁC) [27]. The hydrolysis of
the oxygen ester and the thioester proceeds via a rate limiting transition-state in which
the build-up of negative charge at the reaction center is significant, yet not as large as
observed with the oxygen ester.
3.2. Substituents effects in thioesterase-catalyzed hydrolysis of 4-HBA-CoA analogs
The requirement for precise steric and electronic complementation between substrate
and enzyme active site as the complex proceeds along the reaction coordinate precludes
the application of SAR analysis to define transition state charge. Instead, the SAR analysis
provides information about the specificity of the reaction for the substrate (substituent
varied in substrate analog) and for the catalytic site (amino acid replaced by site-directed
mutagenesis). Examination of the active site region surrounding the para-hydroxyl in the
X-ray structure of the wild-type thioesterase-4-HP-CoA complex (see Fig. 2) reveals that
the hydroxyl group is involved in a network of H-bonds. The H-bond branches can serve
to amplify the binding energy derived from the formation of the primary H-bond with the
substrate in the ground state and/or in the transition state. From Fig. 2 we can see that the

F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
7
˚
nucleophile Asp17 is positioned close (3.2 A) to the thioester C@O carbon, and that the
thioester C@O is engaged in hydrogen bond formation with the backbone NH of Tyr24
located at the N-terminus of the active site a-helix. The benzoyl ring para-hydroxyl group
is engaged in hydrogen bond formation with the Thr59 backbone NH, and with an active
site-bound water molecule. If the para-benzoyl ring hydroxyl group was eliminated or
replaced, the ring might shift in position, and the thioester C@O may no longer be opti-
mally aligned with the Asp17 carboxylate.
To examine the importance of the hydrogen bond interaction, the kcat and kcat/Km of
the wild-type thioesterase-catalyzed hydrolysis of benzoyl-CoA substrates having the dif-
ferent para-substituents MeO, Me, H, F, Cl, and CF3 were measured (see Table 1). The
removal of the para-hydroxyl in the BA-CoA substrate resulted in an 18-fold and a
1500-fold decrease in kcat and kcat/Km values, respectively. This observation agrees with
the important role of para-hydroxyl in substrate binding. The loss of binding interaction
with the para-hydroxyl substituent could allow the thioester C@O to shift slightly in posi-
tion relative to the nucleophile, as the ring is given more freedom to move within the con-
fines of the active site.
For 4-FBA-CoA, the kcat/Km value is 160-fold smaller than that of 4-HBA-CoA, which
suggests that the fluoride does not function as well as the hydroxyl group in facilitating
substrate binding. Reduced hydrogen bond interaction between the para-F and the
Thr59 backbone amide NH, and the active site bound water, might be responsible for this
lowered binding affinity. The kcat for thioesterase-catalyzed hydrolysis of 4-CBA-CoA is
18-fold smaller in value than that of 4-HBA-CoA and the kcat/Km value is 1500-fold small-
er than that measured for 4-HBA-CoA. Apparently, both binding affinity and binding ori-
entation are impaired by the inability of the Cl substituent to engage in hydrogen bond
interaction with the Thr59 backbone amide NH, and with the bound water molecule.
The CF3 and CH3 para-substituted substrates were the least reactive. The kcat/Km value
for thioesterase-catalyzed 4-CF3BA-CoA hydrolysis is 24,000-fold smaller than that mea-
sured for 4-HBA-CoA, while that for the 4-MBA-CoA is reduced 15,000-fold. Notably,
the kcat/Km value of 4-MeOBA-CoA is only 640-fold smaller than the kcat/Km value of
4-HBA-CoA. This comparison of substituent effects suggests that interaction with the
C(4)OH contributes to productive binding whereas the substitution of the C(4)OH proton
with a methyl group impairs it.
The plot of log(kcat/Km) vs r (see Fig. 4) indicated that no correlation exists between
electron withdrawing capacity of the ring substituent and substrate activity (kcat or kcat
/
Km). Thus, we can surmise that the electron withdrawing ability of the para-substituent
does not determine the reactivity of the substrate. Instead, the interaction between this
substituent and the enzyme active site seems to be the controlling factor.
The para-substituent of BA-CoA is a hydrogen atom. Unlike the para-hydroxyl of the
4-HBA-CoA substrate, the para-hydrogen of BA-CoA does not interact with water sol-
vent in the unbound state nor with the polar active site groups in the enzyme complex.
Because catalysis is slow with this substrate, the substrate binding step is likely to be at
rapid equilibrium. According, the BA-CoA Km = 510 lM is an estimate of its Kd (the val-
ue proved to be too large to measure using fluoresence titration methods). Based on the
Kd = 0.15 lM for 4-HBA-CoA bound with D17N mutant thioesterase, we can calculate
that the para-hydroxyl contributes a DDGbind = 4.8 kcal/mol for binding in the ground
state. In view of the fact that this value represents the differential in H-bond interaction
of the para-hydroxyl with enzyme and water solvent, it is impressively large. Based on

8
F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
the ratio of kcat/Km values, the para-hydroxyl contributes a DDG = 4.3 kcal/mol for
transition state stabilization. Thus, the para-hydroxyl group functions in substrate binding
and not in catalysis (as would be the case if it was bound more tightly at the transition
state than at the ES complex).
3.3. The role of the active site Asp32 was revealed by site-directed mutagenesis
˚
The active site Asp17, which is positioned 3.2 A from the substrate thioester carbonyl
carbon, is essential for catalysis as evidenced by the 100,000-fold decrease in activity that
occurs upon replacement by Asn [17]. The Asp32 on the other hand, is positioned near the
4-hydroxybenzoyl ring where it binds to the para-hydroxyl through two bridging water
molecules. The turnover rate of the D32S mutant is only 33-fold less than that of the
wild-type thioesterase. This suggests that Asp32 plays only a supportive role in catalysis.
In this work, the specificity of the D32S mutant towards the para-substituted substrate
analogs was measured to determine if the Asp32 carboxylate contributes to substrate
discrimination. The kcat, Km and kcat/Km values obtained are reported in Table 2.
4-HBA-CoA was found to be the best substrate for the D32S mutant. The kcat/Km for
D32S-catalyzed 4-HBA-CoA hydrolysis is reduced 220-fold compared to that of the
wild-type enzyme. Although the wild-type thioesterase catalyzes the hydrolysis of each
of the 4-HBA-CoA analogs faster than does the D32S mutant, the difference in the rate
of catalysis is substantially less than it is when 4-HBA-CoA serves as substrate. For exam-
ple, the decrease in the kcat/Km value for the hydrolysis of 4-MeOBA-CoA is 37-fold, and
the decrease in the kcat/Km for the hydrolysis of 4-CF3BA-CoA is only 3-fold. Thus, for
the wild-type thioesterase-catalyzed hydrolysis of 4-HBA-CoA, the interaction between
the para-benzoyl ring hydroxyl and the carboxyl group of the residue Asp32 is significant.
If the substrate is modified to remove the C(4)OH for H-bond interaction, the
contribution of Asp32 to catalysis is significantly reduced, as is the ability of the enzyme
to discriminate substrates (Table 2).
3.4. Biological significance of substrate specificity
The purpose of Pseudomonas thioesterase is to function in the 4-chlorobenzoate
(4-CBA) degradation pathway. In the first step of this pathway, 4-CBA-CoA is formed
Table 2
The steady-state kinetic constants determined for Pseudomonas 4-HBA-CoA thioesterase mutant D32S-catalyzed
hydrolysis of the natural substrate 4-HBA-CoA, and the substrate analogs 4-MeOBA-CoA, 4-MeBA-CoA, BA-
CoA, 4-FBA-CoA, 4-CBA-CoA and 4-CF3-CoA measured at pH 7.5 and 25 ꢁC
Substrate
kcat (sꢀ1
)
Km (lM)
kcat/Km (Mꢀ1 sꢀ1
)
ASC (rp)
4-HBA-CoA
4-MeOBA-CoA
4-MeBA-CoA
BA-CoA
4-FBA-CoA
4-CBA-CoA
4-CF3BA-CoA
(5.6 0.2) · 10ꢀ1
(3.2 0.2) · 10ꢀ2
(1.4 0.1) · 10ꢀ2
(5.1 0.2) · 10ꢀ2
(2.38 0.04) · 10ꢀ1
(1.35 0.05) · 10ꢀ1
(1.45 0.04) · 10ꢀ2
(4.1 0.6) · 101
(2.5 0.3) · 102
(7 1) · 102
1.4 · 104
1.3 · 102
2.0 · 10
1.4 · 102
8.6 · 102
1.3 · 102
4.2 · 10
ꢀ0.22
ꢀ0.27
ꢀ0.14
0.00
+0.15
+0.24
+0.53
(3.7 0.4) · 102
(2.8 0.1) · 102
(2.1 0.2) · 102
(3.7 0.3) · 102
See Section 2 for details.

F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
9
from 4-CBA and CoA at the expense of ATP (Fig. 1). Following dehalogenation of the 4-
CBA-CoA to 4-HBA-CoA, the thioester linkage is hydrolyzed by the thioesterase releas-
ing CoA and the pathway product, 4-HBA. The thioesterase must refrain from hydrolyz-
ing the thioester unit of the 4-CBA-CoA in order to avoid a futile cycle of ATP hydrolysis.
The para-benzoyl ring substituent might therefore, play an important role in substrate rec-
ognition. Indeed, the kcat/Km value for benzoyl-CoA, and for each para-substituted ben-
zoyl-CoA substrate tested (MeO, Me, H, F, Cl, and CF3; Table 1), is significantly lower
than that measured for 4-HBA-CoA. The kcat/Km value of 2 · 103 Mꢀ1 sꢀ1 for thioester-
ase-catalyzed 4-CBA-CoA hydrolysis is 1500-fold smaller than the kcat/Km of
3.1 · 106 Mꢀ1 sꢀ1 for thioesterase-catalyzed 4-HBA-CoA hydrolysis and 100-fold smaller
than the kcat/Km of 1.7 · 105 Mꢀ1 sꢀ1 for dehalogenase-catalyzed 4-CBA-CoA dehalogen-
ation [28] measured under the same conditions (pH 7.5, 25 ꢁC). This rate difference pre-
vents loss of the 4-CBA-CoA to thioesterase-catalyzed hydrolysis under physiological
conditions, provided that the cellular dehalogenase concentration is equal to or greater
than the cellular thioesterase concentration.
Recent studies have shown that benzoate is converted to benzoyl-CoA (at the expense
of ATP) before degradation [5,29]. In bacteria that use 4-chlorobenzoate and benzoate as
fuel, it is important that the benzoyl-CoA not be degraded by the 4-HBA-CoA thioester-
ase before it could be funneled in the benzoyl-CoA degradation pathway. Unless the 4-
HBA-CoA thioesterase can discriminate between 4-HBA-CoA and benzoyl-CoA, the
two pathways cannot coexist. The kcat/Km benzoyl-CoA is low (2 · 103 Mꢀ1 sꢀ1), and
much lower than the kcat/Km (3.1 · 106 Mꢀ1 sꢀ1) for thioesterase-catalyzed hydrolysis of
4-HBA-CoA. Thus, at the enzyme concentration required to process the 4-HBA-CoA,
the benzoyl-CoA pool should not be effected.
4. Conclusion
In conclusion, the reaction between hydroxide ion and the thioester moiety of 4-HBA-
CoA proceeds through a transition state which accumulates less negative charge than does
the rate limiting transition state formed in the reaction of hydroxide ion and the para-
hydroxybenzoyl-ethyl ester. In the 4-HBA-CoA thioesterase-catalyzed reaction, the induc-
tive effect of the substrate para-substituent is masked by steric and hydrogen bonding
effects. Recognition of the C(4)OH in the targeted substrate 4-HBA-CoA is important
for prevention of unwanted hydrolysis of the pathway substrate 4-CBA-CoA, as well as
the metabolite benzoyl-CoA [5,29].
Acknowledgment
This work was supported in part by NIH Grant GM28688.
References
[1] L. Katz, S. Donadio, Annu. Rev. Microbial. 47 (1993) 875–912.
[2] S. Smith, FASEB J. 8 (1994) 1248–1259.
[3] J.W. Trauger, R.M. Kohli, H.D. Mootz, M.A. Marahiel, C.T. Walsh, Nature 407 (2000) 215–218.
[4] J.D. Scholten, K.-H. Chang, P.C. Babbitt, H. Charest, M. Sylvestre, D. Dunaway-Mariano, Science 253
(1991) 182–185.
[5] J. Gescher, A. Zaar, M. Mohamed, H. Schagger, G. Fuchs, J. Bacteriol. 184 (2002) 6301–6315.

10
F. Song et al. / Bioorganic Chemistry 35 (2007) 1–10
[6] E.R. Olivera, B. Minambres, B. Garcia, C. Muniz, M.A. Moreno, A. Ferrandez, E. Diaz, J.L. Garcia, J.M.
Luengo, Proc. Natl. Acad. Sci. USA 95 (1998) 6419–6424.
[7] L.A. Camp, S.L. Hofmann, J. Biol. Chem. 268 (1993) 22566–22574.
[8] J.Y. Lu, L.A. Verkruyse, S.L. Hofmann, Proc. Natl. Acad. Sci. USA 93 (1996) 10046–10050.
[9] J.A. Duncan, A.G. Gilman, J. Biol. Chem. 273 (1998) 15830–15837.
[10] R. Hertz, J. Magenheim, I. Berman, J. Bar-Tana, Nature 392 (1998) 512–516.
[11] D.L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, F. Frolow, S.M. Franken, M. Harel, S.J. Remington, I. Silman,
J. Schrag, et al., Protein Eng. 5 (1992) 197–211.
[12] S.C. Dillon, A. Bateman, BMC Bioinformatics 12 (2004) 109.
[13] J. Pleiss, M. Fischer, R.D. Schmid, Chem. Phys. Lipids 93 (1998) 67–80.
[14] M.M. Benning, G. Wesenberg, R. Liu, K.L. Taylor, D. Dunaway-Mariano, H.M. Holden, J. Biol. Chem.
273 (1998) 33572–33579.
[15] J.B. Thoden, H.M. Holden, Z. Zhuang, D. Dunawy-Mariano, J. Biol. Chem. 277 (2002) 27468–27476.
[16] K.-H. Chang, P.-H. Liang, W. Beck, J.D. Scholten, D. Dunaway-Mariano, Biochemistry 31 (1992) 5605–
5610.
[17] Z. Zhuang, F. Song, W. Zhang, K. Taylor, A. Archambault, D. Dunaway-Mariano, J. Dong, P.R. Carey,
Biochemistry 41 (2002) 11152–11160.
[18] H.A. Erlich, PCR Technology Principles and Applications for DNA Amplification, W.H. Freeman and Co.,
New York, 1992.
[19] Z. Zhuang, F. Song, H. Takami, D. Dunaway-Mariano, J. Bacteriol. 186 (2004) 393–399.
[20] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[21] G.J. Narlikar, D. Herschlag, Biochemistry 37 (1998) 9902–9911.
[22] A.C. Hengge, W.W. Cleland, J. Am. Chem. Soc. 112 (1990) 7421–7422.
[23] M.L. Bender, H.d’A. Heck J. Am. Chem. Soc. 89 (1967) 1211–1220.
[24] F.R. Fedor, T.C. Bruice, J. Am. Chem. Soc. 87 (1965) 4138–4145.
[25] N.S. Isaacs, Physical Organic Chemistry, second ed., Longman Scientific & Technical, New York, 1995.
[26] L.T. Webster Jr., J.J. Mieyal, U.A. Siddiqui, J. Biol. Chem. 249 (1974) 2641–2645.
[27] H.H. Jaffe, Chem. Rev. 53 (1953) 191–261.
[28] P.-H. Liang, G. Yang, D. Dunaway-Mariano, Biochemistry 32 (1993) 12245–12250.
[29] A. Zaar, W. Eisenreich, A. Bacher, G. Fuchs, J. Biol. Chem. 276 (2001) 24997–25004.