Biochem. J. (2013) 449, 79–89 (Printed in Great Britain)
79
doi:10.1042/BJ20120871
Structures of trans -2-enoyl-CoA reductases from Clostridium
acetobutylicum and Treponema denticola : insights into the substrate
specificity and the catalytic mechanism
Kuan HU*†1 , Meng ZHAO†‡1 , Tianlong ZHANG*1 , Manwu ZHA*, Chen ZHONG*, Yu JIANG‡2 and Jianping DING*2
*State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yue-Yang Road,
Shanghai 200031, China, †Graduate School of Chinese Academy of Sciences, 320 Yue-Yang Road, Shanghai 200031, China, and ‡Key Laboratory of Synthetic Biology, Institute of
Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 300 Feng-Lin Road, Shanghai 200032, China
TERs (trans-2-enoyl-CoA reductases; EC 1.3.1.44), which
specifically catalyse the reduction of crotonyl-CoA to butyrylCoA using NADH as cofactor, have recently been applied in the
design of robust synthetic pathways to produce butan-1-ol as a
biofuel. We report in the present paper the characterization of
a CaTER (a TER homologue in Clostridium acetobutylicum),
the structures of CaTER in apo form and in complexes with
NADH and NAD + , and the structure of TdTER (Treponema
denticola TER) in complex with NAD + . Structural and sequence
comparisons show that CaTER and TdTER share approximately
45 % overall sequence identity and high structural similarities
with the FabV class enoyl-acyl carrier protein reductases in
the bacterial fatty acid synthesis pathway, suggesting that both
types of enzymes belong to the same family. CaTER and TdTER
function as monomers and consist of a cofactor-binding domain
and a substrate-binding domain with the catalytic active site
located at the interface of the two domains. Structural analyses of
CaTER together with mutagenesis and biochemical data indicate
that the conserved Glu75 determines the cofactor specificity,
and the conserved Tyr225 , Tyr235 and Lys244 play critical roles
in catalysis. Upon cofactor binding, the substrate-binding loop
changes from an open conformation to a closed conformation,
narrowing a hydrophobic channel to the catalytic site. A modelling
study shows that the hydrophobic channel is optimal in both width
and length for the binding of crotonyl-CoA. These results provide
molecular bases for the high substrate specificity and the catalytic
mechanism of TERs.
INTRODUCTION
Enoyl-CoA reductases, which belong to the superfamily of
oxidoreductases and exist ubiquitously in all organisms, catalyse
the reduction of enoyl-CoA to acyl-CoA using NADH or NADPH
as a cofactor with usually reversible kinetics. TERs identified
in Euglena gracilis and T. denticola utilize NADH as cofactor,
exhibit high substrate specificity for crotonyl-CoA and moderate
activity for hexanoyl-CoA, and possess no activity for the reverse
oxidation reaction [13–15]. Notably, TdTER has a much higher
activity than EgTER (E. gracilis TER) for the reduction of
crotonyl-CoA to butyryl-CoA which can be further converted
into butan-1-ol by the bifunctional butyraldehyde and butanol
dehydrogenase [12,13]. Homologues of TdTER and EgTER have
been found in many prokaryotes and it was suggested that this
distinct class of enzymes might be involved in a novel fatty
acid synthesis pathway [14,15]. Despite their great potential, the
function and the catalytic mechanism of TERs remain unclear,
limiting their usage in biosynthesis of biofuels.
We have identified CaTER (C. acetobutylicum TER) and
determined the crystal structures of CaTER in the apo form and in
complexes with NADH and NAD + , and the crystal structure
of TdTER in complex with NAD + . CaTER exhibits similar
biochemical properties as TdTER, but has a relatively weaker
activity for crotonyl-CoA. The structural and biochemical data
together reveal the key residues involved in the recognition and
In the modern world fossil fuels have been the dominant
energy resource. Due to concerns about energy shortage and
the sustainability of fossil fuels, extensive efforts have been
made in the past decade to seek alternative energy sources. One
approach is the synthesis of medium-chain volatile alcohols as
biofuels by engineered micro-organisms [1–3]. Butan-1-ol, which
is naturally synthesized from condensation of acetyl-CoA via a
series of reversible reactions in Clostridium species, has attracted
the most attention [4–6]. Several research groups have achieved
high-titre and high-yield production of butan-1-ol through genetic
manipulation of Clostridia [5,7,8] or development of recombinant
non-native butan-1-ol-producing organisms through introduction
of the genes of Clostridium and other organisms responsible
for catalysis of the butan-1-ol synthesis reaction [9–13]. In
particular, in several of those engineered pathways the high
productivity of butan-1-ol (titre of 30 g/l and yield of 70–88 %
of the theoretical value) is attributed to the replacement of C.
acetobutylicum BCD (butyryl-CoA dehydrogenase) by TdTER
[Treponema denticola TER (trans-2-enoyl-CoA reductase); EC
1.3.1.44] and the artificial build-up of NADH and acetylCoA as driving forces for the crotonyl-CoA reduction step
[12,13].
Key words: biofuel, catalytic mechanism, crystal structure,
reductase, substrate specificity, synthetic biology.
Abbreviations used: ACP, acyl-carrier protein; BCD, butyryl-CoA dehydrogenase; BmFabV, Burkholderia mallei FabV; DTT, dithiothreitol; FAS-II, fatty acid
synthesis; MR, molecular replacement; PEG, poly(ethylene glycol); RMSD, root mean square deviation; SAD, single-wavelength anomalous dispersion;
SDR, short-chain dehydrogenase/reductase; SeMet, selenomethionine; TER, trans -2-enoyl-CoA reductase; CaTER, Clostridium acetobutylicum TER;
EgTER, Euglena gracilis TER; TdTER, Treponema denticola TER; XoFabV, Xanthomonas oryzae FabV; YpFabV, Yersinia pestis FabV.
1
These authors contributed equally to this work.
2
Correspondence may be addressed to either of these authors (email [email protected] or [email protected]).
The structural co-ordinates reported in the PDB under accession codes 4EUH, 4EUE, 4EUF and 4FBG.
c The Authors Journal compilation c 2013 Biochemical Society
80
K. Hu and others
binding of the cofactor and the substrate and in the catalytic
reaction of TERs. The sequence and structural comparisons also
demonstrate that TERs and FabVs belong to the same family.
These results are valuable for engineering of TERs in the design
of more robust biosynthetic pathways to produce alcohols as
biofuels.
MATERIALS AND METHODS
Cloning, expression and purification of CaTER and TdTER
The CaTER gene was amplified by PCR from the genomic DNA
of C. acetobutylicum and cloned into the pET28a expression
vector (Novagen). The TdTER gene was synthesized by Sangon
Biotech and cloned into the pET22b expression vector (Novagen)
for the structural study and the pET28a expression vector for
the biochemical study. Each plasmid was transformed into
Escherichia coli BL21(DE3) Rosseta strain cells (Novagen), and
protein expression was induced with 0.2 mM IPTG (isopropyl
β-D-thiogalactopyranoside) at 30 ◦ C for 4 h. The cells were
lysed on ice by sonication and the supernatant was used for
protein purification. Protein purification was carried out by
affinity chromatography using a Ni-NTA (Ni2 + -nitrilotriacetate)
Superflow column (Qiagen) followed by gel filtration using a
Superdex 75 16/60 column (Amersham Biosciences). The purified
protein was of high purity (above 95 %) as shown by SDS/PAGE,
and was stored in 20 mM Hepes (pH 7.0), 50 mM NaCl and
2 mM DTT (dithiothreitol) for CaTER and in 20 mM Tris/HCl
(pH 8.0), 50 mM NaCl and 2 mM DTT for TdTER. SeMet
(selenomethionine)-substituted CaTER and TdTER proteins were
prepared as for the native protein except that the bacterial cells
grew in M9 medium. Constructs of the CaTER mutants containing
various point mutations were generated using the QuikChangeTM
Site-Directed Mutagenesis kit (Stratagene) and verified by DNA
sequencing. Expression and purification of the CaTER mutants
were the same as for the wild-type enzyme.
data for the apo SeMet CaTER and the NAD + -bound native
CaTER in space group P21 21 21 were collected to 2.1 Å and
2.7 Å resolution respectively. Crystals of the SeMet TdTER–
NAD + complex were grown from drops consisting of 1 μl of
protein solution and 1 μl of reservoir solution containing 0.1 M
Bis-Tris (pH 8.0), 22 % PEG3350 and 10 μM sacrosine, which
belong to space group P1 and contain 16 TdTER molecules in an
asymmetric unit with a solvent content of 51.7 %. The selenium
SAD diffraction data of TdTER were collected to 3.0 Å resolution
from a flash-cooled crystal at 100 K. All diffraction data were
collected at beamline 17U of Shanghai Synchrotron Radiation
Facility (SSRF), China and processed using HKL2000 [16]. The
statistics of the diffraction data are summarized in Table 1.
Structure determination and refinement
The structure of the SeMet CaTER–NADH complex was solved
by the SAD method using Phenix [17]. There was strong electron
density for an NADH at the active site of CaTER. The structure
refinement was carried out against the 2.0 Å SAD data using
Phenix [17] and Refmac5 [18]. The model building was performed
using Coot [19]. The structures of the apo SeMet CaTER and
the NAD + -bound native CaTER were solved by MR (molecular
replacement) method using the structure of the SeMet CaTER–
NADH complex as the search model. There was good electron
density for an NAD + at the active site in the latter structure. The
structure of the SeMet TdTER–NAD + complex was solved by
MR using the structure of the SeMet CaTER–NADH complex as
the search model. There was weak, but evident, electron density
for an NAD + at the active site of each TdTER and we were able
to build confidently NAD + in 8 out of the 16 TdTER molecules
in the asymmetric unit. The stereochemical geometry of the
structures was analysed using Procheck [20]. The figures were
generated using PyMol (http://www.pymol.org). The statistics
of the structure refinement and the quality of the final structure
models are also summarized in Table 1.
Crystallization and diffraction data collection
Enzyme activity assay
Crystallization was performed using the sitting-drop vapour
diffusion method at 16 ◦ C. Prior to crystallization, the SeMet
CaTER was incubated with 2 mM NADH. Crystals of the
SeMet CaTER–NADH complex were grown from drops
consisting of 1 μl of protein solution and 1 μl of reservoir
solution containing 0.2 M Mg(CH3 COO)2 and 20 % PEG
[poly(ethylene glycol)] 3350, which belong to space group C2
and contain one CaTER in an asymmetric unit with a solvent
content of 46.1 %. Selenium SAD (single-wavelength anomalous
dispersion) diffraction data were collected to 2.0 Å (1 Å = 0.1 nm)
resolution from a flash-cooled crystal at 100 K. In an attempt
to obtain crystals of substrate-bound CaTER or TdTER, the
native and SeMet CaTER or TdTER were incubated with 2.5 mM
crotonyl-CoA (Sigma) in the absence and presence of
2.5 mM NAD + . In the absence of NAD + , we obtained crystals
of the apo SeMet CaTER from drops consisting of 1 μl of protein
solution and 1 μl of reservoir solution containing 0.24 M K2 HPO4
and 22 % PEG3350, which belong to space group P21 21 21 and
contain one CaTER in an asymmetric unit with a solvent content
of 51.8 %. In the presence of NAD + , we obtained crystals of
the NAD + -bound native CaTER from drops consisting of 1 μl
of protein solution and 1 μl of reservoir solution containing
0.2 M ammonium citrate (pH 7.0) and 20 % PEG3350, which
also belong to space group P21 21 21 and contain one CaTER in an
asymmetric unit with a solvent content of 42.7 %. Diffraction
The activities of CaTER and TdTER to convert crotonyl-CoA
into butyryl-CoA were assayed by monitoring the oxidation of
NADH to NAD + over time at 340 nm using a Beckman Coulter
DU800 spectrophotometer. The reaction mixtures consisted of
0.1 M K2 HPO4 buffer (pH 6.2), 0.4 μM CaTER or TdTER,
0.4 mM NADH, and a varied concentration of crotonyl-CoA
(60–500 μM) in a total volume of 100 μl. The enzyme was preincubated with NADH for 10 min before the addition of crotonylCoA. The specific activity of the enzyme was measured at a
fixed concentration of crotonyl-CoA (500 μM). The oxidation
activity for the reverse reaction was measured by monitoring
the reduction of NAD + in a reaction mixture consisting of
0.1 M K2 HPO4 buffer (pH 6.2), 0.4 μM CaTER or TdTER,
0.4 mM NAD + and 500 μM butyryl-CoA. The apparent kinetic
parameters K m and kcat (Table 2) were determined by fitting the
kinetic data to the Michaelis–Menten equation using the nonlinear regression analysis method implemented in Prism 4.0 for
Windows (GraphPad Software). All experiments were carried out
at 25 ◦ C and repeated at least twice under the same conditions.
c The Authors Journal compilation c 2013 Biochemical Society
Docking experiment
The trans-2-crotonyl-CoA substrate was docked into the CaTER–
NADH complex using AutoDock4 [21]. The co-ordinates
of crotonyl-CoA were retrieved from the crystal structure of
Substrate specificity and catalytic mechanism of CaTER and TdTER
Table 1
81
Summary of diffraction data and structure refinement statistics
Numbers in parentheses represent the highest resolution shell.
Parameters
Diffraction data
Wavelength (Å)
Space group
a (Å)
b (Å)
c (Å)
α (◦ )
β (◦ )
γ (◦ )
a (Å)
Resolution (Å)
Observed reflections
Unique reflections (I/σ (I) > 0)
Average redundancy
Average I/σ (I)
Completeness (%)
R merge (%)*
Refinement and structure model
Reflections [Fo 0σ (Fo )]
Working set
Test set
R work /R free (%)†
Number of atoms
Protein
Cofactor
Na +
Water
Average B factor (Å2 )
All atoms
Protein
Cofactor
Na +
Water
RMSDs
Bond lengths (Å)
Bond angles (◦ )
Ramachandran plot (%)
Most favoured
Allowed
Generously allowed
SeMet apo-CaTER
SeMet CaTER–NADH
Wild-type CaTER–NAD +
SeMet TdTER–NAD +
0.9791
P 21 2 1 2 1
58.0
77.6
107.2
90.0
90.0
90.0
58.0
50.0–2.10 (2.18–2.10)
164126
28964
5.7 (5.3)
16.7 (6.1)
98.3 (96.9)
10.8 (30.0)
0.9793
C2
111.2
46.0
85.4
90.0
90.7
90.0
111.2
50.0–2.00 (2.07–2.00)
214427
29209
7.3 (7.4)
28.6 (7.7)
99.4 (99.0)
12.3 (45.9)
0.9791
P 21 2 1 2 1
57.7
70.0
101.6
90.0
90.0
90.0
57.7
50.0–2.70 (2.80–2.70)
77975
11837
6.6 (7.0)
20.8 (4.5)
99.4 (98.9)
9.2 (44.4)
0.9795
P1
100.8
120.0
171.3
90.8
105.0
112.8
100.8
50.0–3.00 (3.11–3.00)
483580
132147
3.7 (3.6)
13.7 (2.4)
95.5 (95.8)
14.4 (65.8)
26935
1426
19.3/23.4
3444
3211
1
232
27660
1456
19.8/22.9
3492
3217
44
1
230
11156
585
22.9/28.4
3196
3123
44
1
28
125454
6623
23.6/29.2
49790
49438
352
–
–
37.0
36.7
22.2
37.6
41.4
23.3
22.9
56.4
31.1
29.0
44.7
44.6
93.5
37.9
27.3
67.8
67.7
–
–
–
0.008
1.1
0.007
1.0
0.008
1.2
0.008
1.1
92.9
6.8
0.3
92.6
7.1
0.3
87.9
11.2
0.9
88.2
11.7
0.1
*R merge = hkl i |Ii (hkl) − <I(hkl)>|/ hkl i Ii (hkl).
†R = hkl F o | − |F c / hkl |F o |.
Table 2
Specific activity and kinetic parameters of the wild-type and mutant CaTER and TdTER towards crotonyl-CoA
ND, the parameters could not be determined.
(A) Specific activity
Enzyme
NADH (units·mg − 1 )
EgTER
TdTER
TdTER
CaTER
CaTERE75A
1.6 +
− 0.02
43 +
− 4.8
455.8 +
− 9.6
30.8 +
− 2.6
8.2 +
− 0.3
NADPH (units·mg − 1 )
0.7 +
− 0.09
ND
15.3 +
− 1.7
2.0 +
− 0.2
10.0 +
− 0.5
Ratio
Reference
2.3:1
[14]
[15]
–
–
–
29.8:1
15.4:1
1:1.2
(B) Kinetic parameters
Enzyme
k cat (s − 1 )
K m (NADH) (μM)
k cat /K m (s − 1 ·M − 1 )
Relative activity (%)
TdTER
CaTER
CaTERF11K
CaTERY225A
CaTERY235F
CaTERK244A
CaTERK245A
385.9 +
− 0.4
28.2 +
− 0.7
16.5 +
− 0.2
+ 0.1
2.3 −
ND
ND
23.6 +
− 0.5
69.7 +
− 5.2
105.4 +
− 7.9
31.6 +
− 2.6
+ 16.5
129.9 −
ND
ND
132.3 +
− 8.0
5.5 +
− 0.4×105
2.7 +
− 0.2×105
5.2 +
− 0.4×104
+ 0.2×10
1.8 −
ND
ND
5
1.8 +
− 0.1 × 10
2037
100
198
6.7
ND
ND
67
6
c The Authors Journal compilation c 2013 Biochemical Society
82
K. Hu and others
its complex with Clostridium symbiosum glutaconyl-coA
decarboxylase A subunit (PDB code 3GLM). All hydrogen
atoms were added to the substrate and charges were assigned
by the Gasteiger calculation using AutoDockTools4 [21]. Polar
hydrogen atoms were added to the CaTER–NADH complex after
the removal of the water molecules. The substrate was then
docked in a 40 Å cube with a spacing of 0.375 Å encompassing
the active site. The docking calculation was carried out using the
Lamarckian genetic search algorithm [22] with a standard setup
of an initial population of 50 randomly chosen orientations, a
maximum of 250 000 energy evaluations, a mutation range of
0.02, a cross-over rate of 0.80 and an elitism value of 1.0 for each
run. A total of ten independent docking runs were performed and
ranked according to their mean docking energy by the scoring
function of AutoDock4. Docking results were clustered using a
cut-off of 2 Å RMSDs (root mean square deviations). The docking
models were further optimized by energy minimization using
GROMACS 4 [23] with the GROMOS 53a6 force field [24].
RESULTS AND DISCUSSION
Biochemical characterization of CaTER and TdTER
To identify TER homologues in other organisms as alternatives
in the development of synthetic pathways for the production of
butan-1-ol, we carried out a sequence search in GenBank® and
identified a TER homologue in C. acetobutylicum (GenBank®
accession number AE001437). A comparison of the amino acid
sequences of CaTER and TdTER shows a moderate sequence
homology (45 % identity and 62 % similarity). Biochemical
analysis shows that, as expected, TdTER prefers NADH rather
than NADPH as a cofactor and has high activity for crotonoylCoA, but no detectable activity for the reverse oxidation reaction.
The specific activity of TdTER for crotonoyl-CoA was determined
to be 455.8 units/mg, which is about 10-fold higher than that
reported by Tucci and Martin [15] (Table 2). This discrepancy
might be due to the difference in the purities of the enzyme which
was purified by one-step affinity chromatography by Tucci and
Martin [15], but purified by affinity chromatography followed
by gel filtration in the present study. CaTER possesses similar
enzymatic properties as TdTER; however, the specific activity
of CaTER for crotonoyl-CoA is about 14.8-fold lower than
that of TdTER (30.8 units/mg compared with 455.8 units/mg).
Specifically, CaTER has a slightly higher K m value for NADH
(105.4 μM compared with 69.7 μM) and a kcat value 13.7-fold
lower than that of TdTER (28.2 s − 1 compared with 385.9 s − 1 )
(Table 2).
Overall structures of CaTER and TdTER
To understand the molecular basis of the substrate specificity
and the catalytic mechanism of TERs, we solved the structures
of CaTER in apo form at 2.1 Å resolution, in complex with
NADH at 2.0 Å resolution and in complex with NAD + at
2.7 Å resolution, and the structure of TdTER in complex with
NAD + at 3.0 Å resolution (Table 1). CaTER consists of 398
residues with a theoretical molecular mass of 45.7 kDa and
TdTER comprises 397 residues with a theoretical molecular
mass of 44.8 kDa. The full-length CaTER is well defined in the
apo and the NADH-bound structures, whereas residues 1–9 and
147–148 are disordered in the NAD + -bound structure. In the
CaTER–NADH complex, the bound NADH is well defined with
strong electron density (Supplementary Figure S1 at http://www.
biochemj.org/bj/449/bj4490079add.htm); in the CaTER–NAD +
c The Authors Journal compilation c 2013 Biochemical Society
complex the bound NAD + has evident, but relatively weaker,
electron density compared with NADH (Supplementary Figure
S1) and a relatively higher average B factor (56.4 Å2 ) compared
with the protein (44.6 Å2 ) (Table 1), indicating a partial occupancy
and/or a high flexibility. In the TdTER–NAD + structure the fulllength protein is well defined for all 16 TdTER molecules in
the asymmetric unit; however, only eight TdTER molecules were
modelled as the NAD + -bound form and others such as the apo
form due to the poor density of the cofactor (Supplementary
Figure S1). As in the CaTER–NAD + complex, the bound
NAD + has a higher average B factor (93.5 Å2 ) compared with
the protein (67.7 Å2 ) (Table 1), indicating a low occupancy
and/or a high flexibility. It is noteworthy that in the CaTER–
NADH structure, there is an evident spherical density near the
pyrophosphate moiety of NADH which is interpreted as a Na +
ion (Supplementary Online data at http://www.biochemj.org/
bj/449/bj4490079add.htm). A similar metal ion is also observed
at the equivalent position in the apo and NAD + -bound CaTER
structures. The biological significance of this metal ion is elusive.
TERs belong to the SDR (short-chain dehydrogenase/
reductase) superfamily. Both CaTER and TdTER adopt the
typical architecture of SDR enzymes [25] and comprise two
domains: a cofactor-binding domain and a substrate-binding
domain (Figure 1A). The cofactor-binding domain assumes a
typical Rossmann fold consisting of a six-stranded parallel βsheet (β3–β6 and β11–β12) flanked by five α-helices (α1–α3,
α10 and α14) on one side and three α-helices (α4, α6 and α7) on
the other. The substrate-binding domain consists of five α-helices
(α8, α9 and α11–α13) on one side, two α-helices (α5 and the
N-terminal part of α7), a short 310 α-helix (η1) and a β-hairpin
(β7 and β8) on the other, and a β-hairpin (β9 and β10) covering
the top. The catalytic active site is located at the interface between
the two domains.
Structural comparisons of the three CaTER structures show
no significant conformational differences in the overall structures
(an RMSD of 0.82 Å for the 392 Cα atoms between the apo
and the NADH-bound forms, an RMSD of 0.57 Å for the 382
Cα atoms between the apo and the NAD + -bound forms, and an
RMSD of 0.89 Å for the 381 Cα atoms between the NADHbound and the NAD + -bound forms) (Figure 1B). However,
notable conformational changes are observed in helix α8 and
the two flanking loops (from Val275 to Pro286 ) which are denoted
as the substrate-binding loop in analogy to other SDR enzymes.
Compared with the apo form, the substrate-binding loop in the
NADH-bound form moves closer towards the active-site pocket
by an average distance of 3.0 Å (calculated on the basis of the
positions of the Cα atoms), narrowing the hydrophobic channel
leading to the active site (Figure 1B). In addition, several residues
at the active site particularly Tyr235 and Thr276 change their sidechain conformations to interact with the cofactor (see results
below). Interestingly, the conformation of the substrate-binding
loop in the NAD + -bound form is more similar to that in the apo
form than that in the NADH-bound form (Figure 1B). Analysis
of crystal packing in these three structures indicates that the
substrate-binding loop is not involved in intermolecular contacts,
and thus its conformation is not constrained by the crystal lattices.
Thus the conformational difference in the substrate-binding loop
between the NADH- and NAD + -bound forms is biologically
relevant and reflects the structural difference of the two enzymatic
states in the catalytic reaction.
Structural comparison of the 16 TdTER molecules in the
asymmetric unit shows no notable conformational differences in
the overall structure or in the substrate-binding loop (RMSDs
of <0.40 Å for all Cα atoms) (Supplementary Figure S2 at
http://www.biochemj.org/bj/449/bj4490079add.htm). Thus one
Substrate specificity and catalytic mechanism of CaTER and TdTER
83
c The Authors Journal compilation c 2013 Biochemical Society
84
K. Hu and others
NAD + -bound TdTER was chosen as the representative in the
following structural analysis and discussion. Although CaTER
and TdTER share only a moderate sequence homology (45 %
identity and 62 % similarity), the two enzymes show a high
structural similarity with identical secondary structure elements
(Figure 1C). A detailed structural comparison shows that the
NAD + -bound TdTER is more similar to the apo or NAD + -bound
CaTER (an RMSD of 1.16 Å for the 377 Cα atoms) than the
NADH-bound CaTER (an RMSD of 1.26 Å for the 374 Cα atoms),
particularly in the substrate-binding loop (Figure 1C). The open
conformation of the substrate-binding loop in the TdTER–NAD +
complex is consistent with the low occupancy of NAD + at the
active site and the weak binding of NAD + with the surrounding
residues (see below).
TERs and FabVs belong to the same enzyme family
A structural similarity search against the PDB using the
Dali Server [26] reveals that CaTER and TdTER share
very high structural similarities with XoFabVs (Xanthomonas
oryzae FabVs) [27] and YpFabVs (Yersinia pestis FabVs) [28].
Superimposition of the apo-CaTER with the apo-XoFabV (PDB
code 3S8M) yields an RMSD of 1.39 Å for the 375 Cα
atoms, and superimposition of the NADH-bound CaTER with
the NADH-bound YpFabV (PDB code 3ZU3) yields an RMSD
of 1.30 Å for the 378 Cα atoms (Supplementary Figure S3
at http://www.biochemj.org/bj/449/bj4490079add.htm). FabVs
were characterized as a novel class of enoyl-ACP (acyl-carrier
protein) reductases [29,30]. Enoyl-ACP reductases catalyse
reduction of the double bond of enoyl-ACP to produce acyl-ACP
in the last, and rate-limiting, step of the bacterial FAS-II (fatty acid
synthesis) pathway, and are divided into four classes: FabI, FabL,
FabV and FabK. The first three classes use NADH or NADPH
as a cofactor and belong to the SDR superfamily [25,31]. Within
these three classes, only FabI and FabV have a typical Rossmann
fold motif. However, the FabV class is distinct from the FabI class
in several aspects: FabVs are relatively larger than FabIs (∼ 400
residues compared with ∼ 260 residues) and the majority of the
extra residues are located in and around the active site; FabVs exist
as a monomer rather than a dimer or tetramer as adopted by FabIs;
and FabVs contain an active-site consensus sequence (YX8 K)
different from that of FabIs (YX6 K) [28,30]. Sequence alignment
of TERs and FabVs from different species demonstrates that these
enzymes share about 45 % sequence identity among members of
each family and between members of the two families, which is
much higher than their similarities to the other classes of enoylACP reductases, including FabI, FabL and FabK (about 15 %
identity). Particularly, the functionally important residues that are
involved in substrate binding, cofactor binding and catalysis
are almost strictly conserved among the identified TERs and
FabVs (Figure 1D). These results suggest that TERs and FabVs
belong to the same enzyme family.
Previously, TERs were defined as a unique family of enoylCoA reductases in prokaryotes that catalyse reduction of the
Figure 1
double bond of enoyl-CoA to produce acyl-CoA using NADH
as cofactor, and were suggested to function in a novel fatty acid
synthesis pathway that uses enoyl-CoAs rather than ACP-bound
enoyl intermediates as the substrate [14,15,32]. On the other hand,
it was shown that VcFabV (Vibrio cholerae FabV) can use both
crotonyl-CoA and crotonyl-ACP as the substrate with a slightly
higher activity for the latter (K m of 1178 μM compared with
195 μM and kcat /K m of 9×106 s − 1 ·M − 1 compared with 4.1×107
s − 1 ·M − 1 ) and functions in the bacterial FAS-II pathway [29].
Although the activities of TERs against crotonoyl-ACP were not
examined [14,15], with the high similarities in both sequence and
structure between TERs and FabVs, it is possible that TERs might
also use both enoyl-CoAs and enoyl-ACPs as the substrate and
function in the bacterial FAS-II pathway to catalyse the reduction
of enoyl-ACPs to acyl-ACPs. In particular, CaTER is very likely
to exert such function as no other SDR enoyl-ACP reductases
have been identified in C. acetobutylicum.
Cofactor binding and specificity
TERs and FabVs contain a strictly conserved cofactor-binding
motif, GxxxGxG, between β3 and α2 in the Rossmann fold
(Figure 1D) which is different from some other members
of the SDR superfamily [25]. The cofactor-binding site of
CaTER resides at the bottom of the active-site pocket. In
the CaTER–NADH structure, the bound NADH assumes an
extended conformation which is stabilized by hydrogen-bonding
interactions with several conserved residues of the surrounding
loops (Figure 2A). Specifically, the nicotinamide moiety
of NADH is stabilized by the main-chain amide and carbonyl of
Leu274 and the side-chain hydroxyl of Thr276 via hydrogen bonds.
The hydroxyls of the nicotinamide ribose form hydrogen bonds
directly with the side chain of Lys244 and indirectly with the
side chain of Tyr235 via a water molecule, both of which play
critical roles in catalysis. The pyrophosphate moiety is stabilized
by hydrogen-bonding interactions with the side-chain hydroxyls
of Thr276 and Ser50 and the main-chain amides of Gly51 and
Phe52 . The adenine moiety inserts into a hydrophobic pocket
formed by Tyr74 , Phe113 , Leu139 , Ala140 and Ala141 , and is stabilized
by hydrogen-bonding interactions with the side-chain carboxyl
of Asp111 and the main-chain amides of Tyr74 and Ala112 . The
hydroxyls of the adenine ribose form hydrogen bonds with the
main-chain amide of Tyr74 , the side-chain carboxyl of Glu75 and
the side-chain hydroxyl of Ser49 . Most of these interactions are
also observed in the NADH-bound YpFabV structure [28] and
the majority of the involved residues are strictly conserved in
TERs/FabVs (Figure 1D). The functional importance of some of
these conserved residues has been confirmed by the biochemical
data showing that mutations of the corresponding residues in
XoFabV and BmFabV (Burkholderia mallei FabV; equivalent to
Ser50 , Phe52 , Asp111 , Tyr235 and Lys244 of CaTER) either abolish or
significantly impair the enzymatic activity [27,30]. These results
suggest that the recognition and binding of NADH might be
conserved in TERs and FabVs.
Structures of CaTER and TdTER
(A) A stereo view of the overall structure of the CaTER–NADH complex. The bound NADH is shown with a stick model in yellow, α-helices are coloured cyan, β-strands in magenta and loops
in violet. The substrate-binding loop is coloured red. (B) Superimposition of the apo (orange), NADH-bound (cyan) and NAD + -bound (pink) CaTER structures. The bound NADH and NAD + are
coloured accordingly. Although the three structures are similar in the overall conformation, there are notable differences in the conformation of the substrate-binding loop as shown in the inset.
(C) Superimposition of the NAD + -bound CaTER structure (pink) and the NAD + -bound TdTER structure (yellow). The bound NAD + is coloured accordingly. There is no notable difference in the
conformation of the substrate-binding loop as shown in the inset. (D) Structure-based sequence alignment of TERs and FabVs from different species. The secondary structures of CaTER and YpFabV
are placed on the top and at the bottom of the alignment respectively. Strictly conserved residues are highlighted in shaded red boxes and conserved residues in open red boxes. The active-site
residues Tyr225 , Tyr235 and Lys244 in CaTER are marked by red stars.
c The Authors Journal compilation c 2013 Biochemical Society
Substrate specificity and catalytic mechanism of CaTER and TdTER
Figure 2
85
Structure of the cofactor-binding site
(A) Interactions of NADH with the surrounding residues in the CaTER–NADH complex. The bound Na + ion is shown as a purple sphere and conserved water molecules as red spheres. The
hydrogen-bonding interactions are shown with dashed lines and distances. (B) Interactions of NAD + with the surrounding residues in the CaTER–NAD + complex. (C) Interactions of NAD + with the
surrounding residues in the TdTER–NAD + complex. (D) Comparison of the cofactor-binding site in the apo (orange), NADH-bound (cyan) and NAD + -bound (pink) CaTER structures. The bound
NADH and NAD + are coloured accordingly. The surrounding residues are shown with stick models. Thr276 and Tyr235 have notable conformational changes. (E) Comparison of the cofactor-binding
site in the NAD + -bound CaTER (pink) and TdTER (yellow) structures. The bound NAD + is coloured accordingly. The surrounding residues are shown with stick models. The labels for CaTER and
TdTER are coloured pink and black respectively. There is no notable difference in the cofactor-binding site except for the side chains of Lys244 of CaTER and Lys249 of TdTER. (F) Electrostatic surface
of the cofactor-binding site in the CaTER–NADH complex. Mutation of Glu75 to an alanine creates space to accommodate the 2 -phosphate of the adenine ribose of NADPH. NADH and a modelled
NADPH are shown with stick models. A scale bar for the electrostatic potential is provided.
The previous kinetic data have shown that the catalytic reaction
of BmFabV follows a sequential Bi Bi mechanism with NADH
binding first and NAD + dissociating last [30]. In the NAD + bound CaTER and TdTER, NAD + is defined with relatively
weaker electron density (Supplementary Figure S1) and maintains
fewer interactions with the surrounding residues compared with
NADH in the NADH-bound CaTER structure, although most of
the residues involved in the interactions with NADH and NAD
maintain similar conformations (Figures 2A–2C). Additionally,
the substrate-binding loop assumes a conformation similar
to the open one in the apo-CaTER rather than the closed one
in the NADH-bound CaTER (Figure 1B), and Thr276 on the loop
and Tyr235 at the active site also assume conformations similar
to those in the apo-CaTER and do not interact with NAD +
(Figure 2D). There are no notable differences in the cofactorbinding site between the CaTER–NAD + and TdTER–NAD +
complexes, except for the side chains of Lys244 in CaTER and
Lys249 in TdTER (Figure 2E and Supplementary Online data). It
is possible that the open conformation of the substrate-binding
loop and the weaker interactions of NAD + with the surrounding
c The Authors Journal compilation c 2013 Biochemical Society
86
K. Hu and others
residues in the CaTER–NAD + complexes might allow the product
to dissociate easily from the enzyme.
TERs possess high activity for the reduction reaction, but no
activity for the reverse oxidation reaction ([13–15] and the present
study). Previously, Bond-Watts et al. [12] reported that TdTER
has a much higher affinity for NADH than NAD + , which is in
agreement with the structural data showing that NAD + has weaker
interactions with the enzymes compared with NADH. These
results suggest that there is a large equilibrium constant favouring
the reduction reaction, thereby resulting in the irreversibility of the
reduction reaction. The large equilibrium constant favouring
the reduction reaction for TERs could be explained by the
remarkable difference in the redox potentials of the reduction
reaction and the oxidative reaction: the NAD + /NADH pair
possesses a much lower redox potential ( − 320 mV) than that
of the crotonyl-CoA/butyryl-CoA pair ( − 125 mV), keeping the
acyl-CoA derivative largely in the reduced state [33].
Interestingly, although most of the TERs and FabVs use NADH
as a cofactor, EgTER can use either NADH or NADPH as
a cofactor [14,15,29,30] (Table 2). NADH and NADPH are
differentiated only by the 2 -phosphate of the adenine ribose
of NADPH. Structural and sequence analyses of TERs/FabVs
from various species indicate that a conserved glutamic acid
(Glu75 in CaTER and YpFabV and Glu80 in TdTER) plays an
important role in discriminating NADH against NADPH. In the
NADH- and NAD + -bound CaTER structures, the side chain
of Glu75 recognizes the 2 -OH of the adenine ribose of NADH
via a hydrogen bond. A similar hydrogen-bonding interaction
is also observed in the TdTER–NAD + and YpFabV–NADH
complexes. Intriguingly, EgTER has an alanine residue at the
equivalent position (Figure 1D). A modelling study indicates that
the side chain of Glu75 would have steric conflict with the 2 phosphate of the adenine ribose of NADPH, and substitution of
Glu75 with an alanine residue would create space to accommodate
the 2 -phosphate of NADPH (Figure 2F), hence conferring the
enzyme an ability to use both NADH and NADPH as a
cofactor. This hypothesis is verified by the biochemical data
that the E75A CaTER mutant can indeed use both NADH and
NADPH as cofactor (Table 2). The specific activity of the E75A
CaTER mutant was determined to be 8.2 units/mg using NADH
as cofactor (3.8-fold lower than the wild-type enzyme), and
10.0 units/mg using NADPH as cofactor (5.0-fold higher than
the wild-type enzyme). These results provide the molecular basis
for the distinct cofactor specificities of different TERs/FabVs.
Substrate specificity
TdTER and EgTER exhibit high activity for crotonyl-CoA, but
weak activity for hexenoyl-CoA [14,15]. Our biochemical data
show that CaTER also displays high activity for crotonyl-CoA.
To investigate the substrate binding mode and substrate specificity
of TERs, extensive efforts were made to obtain structures of
CaTER and TdTER in complex with the crotonyl-CoA substrate;
however, none were obtained by co-crystallization or soaking
experiments. As described above, the binding of NADH induces
a conformational change of the substrate-binding loop by about
3.0 Å, narrowing the hydrophobic channel connecting the active
site (Figure 1D). The hydrophobic channel in the CaTER–NADH
structure appears to be appropriate in both width and length for
binding the crotonyl and pantetheine moieties of crotonyl-CoA
(Figure 3A). In addition, there is a large surface groove near the
entrance to the channel which could be the binding site of
the CoA moiety of the substrate. Thus we docked a trans-2crotonyl-CoA into the active site of the CaTER–NADH complex
using AutoDock4 [21], which reveals some insight into the
c The Authors Journal compilation c 2013 Biochemical Society
substrate-binding mode and the high substrate specificity of
CaTER (and possibly other TERs).
In the docking model, crotonyl-CoA assumes an extended
conformation and adopts a binding mode similar to that of the fatty
acyl substrate in the structure of the FabI class Mycobacterium
tuberculosis InhA (enoyl-ACP reductase) in complex with a C16
fatty acyl substrate (PDB code 1BVR) [34] (Figures 3A and 3B).
The crotonyl and pantetheine moieties of the substrate insert into
the hydrophobic channel without obvious steric conflicts. The
crotonyl moiety lies next to NADH and is stabilized by Ile157 ,
Ile240 , Tyr225 , Tyr235 , Ile282 and Phe285 mainly via hydrophobic
interactions. In particular, the hydroxyl of Tyr235 is positioned to
form hydrogen-bonding interactions with the thioester carbonyl
of the crotonyl moiety and the 2 -hydroxyl of the nicotinamide
ribose of NADH; and the side chain of Tyr225 is positioned near
the C2 atom of the crotonyl moiety and has a π–π stacking
interaction with the crotonyl moiety. The pantetheine moiety has
both hydrophobic and hydrophilic interactions with Ala141 , Pro142 ,
Ile157 , Met196 , Ala278 , Ser279 , Tyr281 and Ile282 . It is interesting to
note that most of the residues involved in the interactions with
the crotonyl and pantetheine moieties are highly conserved in
TERs/FabVs except Phe285 (Figure 1D). The side chain of Phe285
is positioned at the deep end of the substrate-binding channel and
appears to block further entry of the enoyl moiety. Substitution of
Phe285 with a residue containing a smaller side chain might allow
the binding of a slightly longer chain substrate, suggesting that
with the variance at this position, TERs/FabVs might be able to
catalyse the reduction of substrates with different chain lengths.
The pyrophosphate and 3 -phosphate-adenosine moieties of the
substrate are accommodated in the large surface groove formed
mainly by Phe11 , Ile12 , Arg13 , Val15 , Arg82 , Thr85 , Arg143 , Lys277 and
Asn375 , and have relatively fewer hydrophilic and hydrophobic
interactions with the protein (Figures 3A and 3B). The adenosine
ribose is stabilized by hydrogen-bonding interactions with
the side chain of Arg82 and the main chain of Ile12 . Interestingly, the
docking model suggests that substitution of Phe11 with a lysine
could introduce a hydrogen-bonding interaction with the 3 phosphate of the adenosine moiety. Indeed, the biochemical data
show that the F11K CaTER mutant has an increased affinity for
crotonyl-CoA (3.3-fold), but no marked change in the kcat value
(Table 2), providing supporting evidence for the validity of the
docking model.
Functional roles of the active-site residues in the catalytic
mechanism
The catalytic mechanism of the SDR superfamily members has
been studied extensively; it consists of the formation of an enolate
intermediate through direct transfer of a hydride ion from NADH
(or NADPH) via a nucleophilic addition to the C3 atom of
the substrate followed by protonation of the thioester carbonyl
[25,34–36]. It is well established that the cofactor donates a
hydride to the C3 atom of the substrate, but no consensus has
been reached yet about where a proton comes from. Additionally,
a consensus sequence at the active site is implicated to play critical
roles in the catalytic reaction.
FabVs contain a consensus YX8 K sequence at the active
site preceded by another strictly conserved tyrosine residue
(corresponding to Tyr225 in YpFabV and BmFabV and Tyr226
in XoFabV), which is, however, different from the other SDR
enoyl-ACP reductases [29,30]. The functional importance of
these residues of FabVs in catalysis has been verified by the
mutagenesis, biochemical and structural data [27,28,30]. Like
FabVs, TERs from different species contain the same sequence
motif, corresponding to Tyr235 , Lys244 and Tyr225 in CaTER
Substrate specificity and catalytic mechanism of CaTER and TdTER
87
(Figure 1D). Our mutagenesis data confirm that these residues
play important roles in catalysis. Specifically, replacements of
Tyr235 with phenylalanine and Lys244 with alanine completely
abolish the activity and mutation of Tyr225 to alanine decreases
the kcat by 12-fold, but has no significant effect on the K m value
of NADH (Table 2).
The biochemical data are consistent with the CaTER–NADH
structure and the docking model of CaTER in complex with
crotonyl-CoA. The side-chain hydroxyl of Tyr235 is not only
involved in the binding of NADH, but also forms a hydrogen
bond with the thioester carbonyl of the substrate, suggesting that
Tyr235 may function in both protonation of the crotonyl moiety
and stabilization of the enolate intermediate in catalysis. The side
chain of Lys244 is involved in direct hydrogen-bonding interactions
with both 2 - and 3 -hydroxyls of the nicotinamide ribose of
NADH, suggesting that Lys244 may primarily function in binding
of NADH. Mutation of Lys244 to an alanine would severely impair
the cofactor binding and consequently destroys the enzymatic
activity. On the other hand, Tyr225 has no direct interaction with
the cofactor, but has a π–π stacking interaction with the crotonyl
moiety of the substrate, suggesting that it may mainly function
in stabilization of the substrate. Mutation of Tyr225 to an alanine
would not affect the cofactor binding, but have a marked effect
on the substrate binding and consequently the catalytic reaction.
BmFabV contains a lysine residue (Lys245 ) following the activesite residue Lys244 which is suggested to play a role in the substrate
binding [30]. However, this lysine residue is not conserved in
FabVs, and XoFabV contains a valine residue at the equivalent
position whose mutation to alanine has no evident effect on the
enzymatic activity [27]. In the YpFabV structure, the side chain
of the equivalent Lys245 points away from the substrate-binding
pocket and is unlikely to interact directly with the substrate [28].
In our CaTER and TdTER structures, the corresponding residue
(Lys245 of CaTER or Lys250 of TdTER) also orients its side chain
away from the substrate-binding pocket and is not involved in
the substrate binding; this is in agreement with the kinetic data
showing that mutation K245A in CaTER has no significant effect
on the enzymatic activity (Table 2). These results indicate that
Lys245 of CaTER or the equivalent in other TERs is not a key
residue in substrate binding or catalysis.
On the basis of the structure of β-ketoacyl-ACP reductase
FabG of the SDR superfamily, it is proposed that a hydrogenbonding network formed by the highly conserved tyrosine and
lysine residues at the active site and several water molecules
functions as a proton relay system to replenish a proton from the
solvent to the thioester carbonyl of the substrate in catalysis [37].
Intriguingly, a similar hydrogen-bonding network is observed in
the NADH-bound CaTER structure. Specifically, the side chain
of Tyr235 forms a hydrogen bond indirectly with the 2 -hydroxyl
of the nicotinamide ribose via a water molecule; the side chain of
Lys244 forms two hydrogen bonds with both 2 - and 3 -hydroxyls
of the nicotinamide ribose; and a water molecule mediates a
network of hydrogen bonds between the side chain of Lys244 , the
main-chain carbonyl of Met196 , and three additional water
Figure 3
Structure of the catalytic active site
(A) Electrostatic surface of the catalytic active site in the CaTER–NADH complex. The bound
NADH (yellow) and the docked crotonyl-CoA (grey) are shown as stick models. The crotonyl and
pantetheine moieties of the substrate insert into the hydrophobic channel and the pyrophosphate
and 3 -phosphate-adenosine moieties are accommodated in the surface groove near the
channel. (B) Interactions of crotonyl-CoA with the surrounding residues in the docked model of
CaTER–NADH in complex with the substrate. NADH is shown in yellow, crotonyl-CoA in grey
and the surrounding residues in cyan. The potential hydrogen-bonding interactions between the
substrate and the surrounding residues are indicated with broken lines. (C) Comparison of the
catalytic active site in the docked model of CaTER–NADH in complex with crotonyl-CoA and
the structure of InhA in complex with NAD + and a substrate (PDB code 1BVR). The colour
scheme for the CaTER–NADH complex is the same as in (B). The InhA–NAD + complex is
coloured as NAD + in green, the substrate in blue and the surrounding residues in magenta. The
potential hydrogen-bonding network between the substrate, the cofactor and the surrounding
residues is indicated with broken lines. The labels for CaTER and InhA are coloured black
and magenta respectively. (D) A schematic diagram of the proposed catalytic mechanism of
CaTER showing the functional roles of the active-site residues and the cofactor. NADH donates
a hydride to the C3 atom of the substrate. A potential hydrogen-bonding network acts a proton
relay system to replenish a proton from the solvent to the thioester carbonyl of the substrate.
c The Authors Journal compilation c 2013 Biochemical Society
88
K. Hu and others
molecules (Figure 2A). Furthermore, in the docking model of
CaTER in complex with the substrate, the position of the water
molecule bridging the interaction of Tyr235 and the nicotinamide
ribose is occupied by the thioester carbonyl of the substrate, and
the hydroxyl of Tyr235 is in a position to make hydrogen-bonding
interactions with both the thioester carbonyl of the substrate
and the 2 -hydroxyl of the nicotinamide ribose (Figures 3B
and 3C). These results suggest that CaTER may use a similar
proton relay system to replenish a proton from the solvent to the
thioester carbonyl of the substrate and that both Tyr235 and Lys244
participate in the proton relay system and play important roles in
the protonation of the carbonyl during the catalytic reaction.
In several engineered butan-1-ol biosynthesis pathways, the
replacement of BCD by TdTER in the crotonyl-CoA reduction
step is considered as one of the driving forces for the high
productivity of butan-1-ol [12,13]. Although both BCD and TER
can catalyse the reduction of crotonoyl-CoA to butyryl-CoA,
sequence alignment of TERs and BCDs from different species
shows that the two families of enzymes share no detectable
sequence homology with an overall sequence identity of <10 %
(results not shown). Consistently, structure comparison of CaTER
and Megasphaera elsdenii BCD [38] shows that the overall
structures and the active-site structures of the two enzymes are
completely different (results not shown). These results suggest
that TER and BCD might employ different catalytic mechanisms
for the reduction of crotonyl-CoA to butyryl-CoA.
Taking the structural, modelling and biochemical data of
CaTER and TdTER together we can propose the functional roles
of the active-site residues in the catalytic mechanism of CaTER
and probably other TERs (Figure 3D). Specifically, Lys244 plays
important roles in both cofactor binding and proton transfer; Tyr235
takes part in both protonation of the thioester carbonyl of the
substrate and stabilization of the enolate intermediate; and Tyr225
participates primarily in binding and stabilization of the crotonyl
moiety of the substrate.
AUTHOR CONTRIBUTION
Kuan Hu carried out the biochemical and crystallization experiments. Meng Zhao
participated in the biochemical study. Tianlong Zhang and Manwu Zha determined the
structures and Tianlong Zhang performed the structural analysis. Chen Zhong participated
in data analyses. Yu Jiang supervised Meng Zhao’s work and participated in data analyses.
Jianping Ding conceived of the study, participated in the experimental design and data
analyses, and wrote the paper.
ACKNOWLEDEGMENTS
We thank the staff members at beamline 17U of Shanghai Synchrotron Radiation Facility
(SSRF), China for technical support in diffraction data collection and other members of
our groups for helpful discussion.
FUNDING
This work was supported by the Ministry of Science and Technology of China [grant
numbers 2011CB966301, 2011CB911102 and 2012BAD32B07], the National Natural
Science Foundation of China [grant numbers 31170690 and 30800027] and the Science
and Technology Commission of Shanghai Municipality [grant number 10JC1416500].
REFERENCES
1 Zaslavskaia, L. A., Lippmeier, J. C., Shih, C., Ehrhardt, D., Grossman, A. R. and Apt, K. E.
(2001) Trophic conversion of an obligate photoautotrophic organism through metabolic
engineering. Science 292, 2073–2075
2 Atsumi, S., Hanai, T. and Liao, J. C. (2008) Non-fermentative pathways for synthesis of
branched-chain higher alcohols as biofuels. Nature 451, 86–89
c The Authors Journal compilation c 2013 Biochemical Society
3 Steen, E. J., Kang, Y., Bokinsky, G., Hu, Z., Schirmer, A., McClure, A., Del Cardayre, S. B.
and Keasling, J. D. (2010) Microbial production of fatty-acid-derived fuels and chemicals
from plant biomass. Nature 463, 559–562
4 Durre, P. (2007) Biobutanol: an attractive biofuel. Biotechnol. J. 2, 1525–1534
5 Lee, S. Y., Park, J. H., Jang, S. H., Nielsen, L. K., Kim, J. and Jung, K. S. (2008)
Fermentative butanol production by Clostridia . Biotechnol. Bioeng. 101, 209–228
6 Li, F., Hinderberger, J., Seedorf, H., Zhang, J., Buckel, W. and Thauer, R. K. (2008)
Coupled ferredoxin and crotonyl coenzyme A (CoA) reduction with NADH catalyzed by the
butyryl-CoA dehydrogenase/Etf complex from Clostridium kluyveri . J. Bacteriol. 190,
843–850
7 Scotcher, M. C., Rudolph, F. B. and Bennett, G. N. (2005) Expression of abrB310 and
SinR, and effects of decreased abrB310 expression on the transition from acidogenesis to
solventogenesis, in Clostridium acetobutylicum ATCC 824. Appl. Environ. Microbiol. 71,
1987–1995
8 Papoutsakis, E. T. (2008) Engineering solventogenic clostridia . Curr. Opin. Biotechnol.
19, 420–429
9 Steen, E. J., Chan, R., Prasad, N., Myers, S., Petzold, C. J., Redding, A., Ouellet, M. and
Keasling, J. D. (2008) Metabolic engineering of Saccharomyces cerevisiae for the
production of n-butanol. Microb. Cell Fact. 7, 36
10 Inui, M., Suda, M., Kimura, S., Yasuda, K., Suzuki, H., Toda, H., Yamamoto, S., Okino, S.,
Suzuki, N. and Yukawa, H. (2008) Expression of Clostridium acetobutylicum butanol
synthetic genes in Escherichia coli . Appl. Microbiol. Biotechnol. 77, 1305–1316
11 Berezina, O. V., Zakharova, N. V., Brandt, A., Yarotsky, S. V., Schwarz, W. H. and Zverlov,
V. V. (2010) Reconstructing the clostridial n-butanol metabolic pathway in Lactobacillus
brevis . Appl. Microbiol. Biotechnol. 87, 635–646
12 Bond-Watts, B. B., Bellerose, R. J. and Chang, M. C. (2011) Enzyme mechanism as a
kinetic control element for designing synthetic biofuel pathways. Nat. Chem. Biol. 7,
222–227
13 Shen, C. R., Lan, E. I., Dekishima, Y., Baez, A., Cho, K. M. and Liao, J. C. (2011) Driving
forces enable high-titer anaerobic 1-butanol synthesis in Escherichia coli . Appl. Environ.
Microbiol. 77, 2905–2915
14 Hoffmeister, M., Piotrowski, M., Nowitzki, U. and Martin, W. (2005) Mitochondrial
trans -2-enoyl-CoA reductase of wax ester fermentation from Euglena gracilis defines a
new family of enzymes involved in lipid synthesis. J. Biol. Chem. 280, 4329–4338
15 Tucci, S. and Martin, W. (2007) A novel prokaryotic trans -2-enoyl-CoA reductase from
the spirochete Treponema denticola . FEBS Lett. 581, 1561–1566
16 Otwinowski, Z. and Minor, W. (1997) Processing of X-ray diffraction data collected in
oscillation mode. Methods Enzymol. 276, 307–326
17 Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd,
J. J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W. et al. (2010) PHENIX: a
comprehensive Python-based system for macromolecular structure solution. Acta
Crystallogr. D66, 213–221
18 Murshudov, G. N., Vagin, A. A. and Dodson, E. J. (1997) Refinement of macromolecular
structures by the maximum-likelihood method. Acta Crystallogr. D53, 240–255
19 Emsley, P. and Cowtan, K. (2004) Coot: model-building tools for molecular graphics. Acta
Crystallogr. D60, 2126–2132
20 Laskowski, R. A., Macarthur, M. W., Moss, D. S. and Thornton, J. M. (1993) PROCHECK:
a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr.
26, 283–291
21 Morris, G. M., Huey, R., Lindstrom, W., Sanner, M. F., Belew, R. K., Goodsell, D. S. and
Olson, A. J. (2009) AutoDock4 and AutoDockTools4: automated docking with selective
receptor flexibility. J. Comput. Chem. 30, 2785–2791
22 Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew, R. K. and
Olson, A. J. (1998) Automated docking using a Lamarckian genetic algorithm and an
empirical binding free energy function. J. Comput. Chem. 19, 1639–1662
23 Hess, B., Kutzner, C., van der Spoel, D. and Lindahl, E. (2008) GROMACS 4: algorithms
for highly efficient, load-balanced, and scalable molecular simulation. J. Chem. Theory
Comput. 4, 435–447
24 Oostenbrink, C., Villa, A., Mark, A. E. and van Gunsteren, W. F. (2004) A biomolecular
force field based on the free enthalpy of hydration and solvation: the GROMOS force-field
parameter sets 53A5 and 53A6. J. Comput. Chem. 25, 1656–1676
25 Kavanagh, K. L., Jornvall, H., Persson, B. and Oppermann, U. (2008) The SDR
superfamily: functional and structural diversity within a family of metabolic and regulatory
enzymes. Cell. Mol. Life Sci. 65, 3895–3906
26 Holm, L. and Rosenstrom, P. (2010) Dali server: conservation mapping in 3D. Nucleic
Acids Res. 38, W545–W549
27 Li, H., Zhang, X., Bi, L., He, J. and Jiang, T. (2011) Determination of the crystal structure
and active residues of FabV, the enoyl-ACP reductase from Xanthomonas oryzae . PLoS
ONE 6, e26743
Substrate specificity and catalytic mechanism of CaTER and TdTER
28 Hirschbeck, M. W., Kuper, J., Lu, H., Liu, N., Neckles, C., Shah, S., Wagner, S., Sotriffer,
C. A., Tonge, P. J. and Kisker, C. (2012) Structure of the Yersinia pestis FabV
enoyl-ACP reductase and its interaction with two 2-pyridone inhibitors. Structure 20,
89–100
29 Massengo-Tiasse, R. P. and Cronan, J. E. (2008) Vibrio cholerae FabV defines a new
class of enoyl-acyl carrier protein reductase. J. Biol. Chem. 283, 1308–1316
30 Lu, H. and Tonge, P. J. (2010) Mechanism and inhibition of the FabV enoyl-ACP
reductase from Burkholderia mallei . Biochemistry 49, 1281–1289
31 Massengo-Tiasse, R. P. and Cronan, J. E. (2009) Diversity in enoyl-acyl carrier protein
reductases. Cell. Mol. Life Sci. 66, 1507–1517
32 Inui, H., Miyatake, K., Nakano, Y. and Kitaoka, S. (1984) Fatty acid synthesis in
mitochondria of Euglena gracilis . Eur. J. Biochem. 142, 121–126
33 Matthies, C. and Schink, B. (1992) Energy conservation in fermentative glutarate
degradation by the bacterial strain WoG13. FEMS Microbiol. Lett. 79, 221–225
89
34 Rozwarski, D. A., Vilcheze, C., Sugantino, M., Bittman, R. and Sacchettini, J. C. (1999)
Crystal structure of the Mycobacterium tuberculosis enoyl-ACP reductase, InhA, in
complex with NAD + and a C16 fatty acyl substrate. J. Biol. Chem. 274, 15582–15589
35 Quemard, A., Sacchettini, J. C., Dessen, A., Vilcheze, C., Bittman, R., Jacobs, Jr, W. R.
and Blanchard, J. S. (1995) Enzymatic characterization of the target for isoniazid in
Mycobacterium tuberculosis . Biochemistry 34, 8235–8241
36 Jornvall, H., Persson, B., Krook, M., Atrian, S., Gonzalez-Duarte, R., Jeffery, J. and Ghosh,
D. (1995) Short-chain dehydrogenases/reductases (SDR). Biochemistry 34, 6003–6013
37 Price, A. C., Zhang, Y. M., Rock, C. O. and White, S. W. (2004) Cofactor-induced
conformational rearrangements establish a catalytically competent active site and a proton
relay conduit in FabG. Structure 12, 417–428
38 Djordjevic, S., Pace, C. P., Stankovich, M. T. and Kim, J. J. (1995) Three-dimensional
structure of butyryl-CoA dehydrogenase from Megasphaera elsdenii . Biochemistry 34,
2163–2171
Received 27 May 2012/10 October 2012; accepted 11 October 2012
Published as BJ Immediate Publication 11 October 2012, doi:10.1042/BJ20120871
c The Authors Journal compilation c 2013 Biochemical Society
Biochem. J. (2013) 449, 79–89 (Printed in Great Britain)
doi:10.1042/BJ20120871
SUPPLEMENTARY ONLINE DATA
Structures of trans -2-enoyl-CoA reductases from Clostridium
acetobutylicum and Treponema denticola : insights into the substrate
specificity and the catalytic mechanism
Kuan HU*†1 , Meng ZHAO†‡1 , Tianlong ZHANG*1 , Manwu ZHA*, Chen ZHONG*, Yu JIANG‡2 and Jianping DING*2
*State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yue-Yang Road,
Shanghai 200031, China, †Graduate School of Chinese Academy of Sciences, 320 Yue-Yang Road, Shanghai 200031, China, and ‡Key Laboratory of Synthetic Biology, Institute of
Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 300 Feng-Lin Road, Shanghai 200032, China
RESULTS
In the CaTER–NADH structure, there is an evident spherical
density near the pyrophosphate moiety of NADH which is
interpreted as a Na + ion with a reasonable B factor of 31.1 Å2 .
This metal ion has seven co-ordination ligands with a bipyramidal
geometry including two pyrophosphate oxygens, the main-chain
carbonyls of Gly47 and Ser50 , the main-chain amide of Gly53 ,
and the main-chain carbonyl and side-chain hydroxyl of Ser138 .
Interestingly a Na + ion is also observed at the same position
with similar co-ordination ligands in our apo and NAD + -bound
CaTER structures and in the YpFabV–NADH structure [1]. In
the apo XoFabV structure, a water molecule was identified at the
equivalent position with four co-ordination ligands which could
be a Na + ion as well [2]. The biological significance of this metal
ion is unknown and may deserve further investigation. As the
metal ion exists in both apo and cofactor-bound TERs/FabVs,
it may play some role in stabilization of the cofactor-binding
site.
Previously, it was predicted that EgTER contains a FADbinding motif (GxGxxG) at the C-terminus [3,4]. However, this
sequence motif is not conserved in the other TERs and in the
structures of both CaTER and TdTER, the corresponding region
(residues 376–381 of CaTER and residues 379–384 of TdTER)
forms a surface exposed loop (the α13–α14 connecting loop)
which cannot be a binding site for FAD (Figures 1A and 1D
of the main text). This is consistent with the biochemical data
showing that the enzymatic activity of TERs does not require
FAD [4,5].
The biochemical data show that TdTER has a higher activity
than CaTER (Table 2 of the main text). We carried out the
comparison between the NAD + -bound TdTER and CaTER
structures and tried to give an explanation for the difference on
the structural basis. The structural comparison shows that there
are no notable differences in the overall structure, in the substratebinding loop and in the cofactor-binding site between the CaTER–
NAD + and TdTER–NAD + complexes. In addition, sequence
comparison indicates that the catalytic active-site residues and the
majority of the residues involved in the interactions with NADH
and NAD + are strictly conserved in CaTER and TdTER (and other
TERs). Nonetheless, there is a conformational difference in the
side chains of Lys244 in CaTER and Lys249 in TdTER (Figure 2E
of the main text). Lys244 in CaTER orients its side chain towards
the active site to interact with the hydroxyls of the nicotinamide
ribose of NAD + . In contrast Lys249 in TdTER points its side
chain away from the active site, which might result in a weaker
binding of NAD + and hence easier dissociation of NAD + when
the catalytic reaction is completed, thereby contributing in part
to the higher activity of TdTER. The conformational difference
of the two lysine residues could be due to the differences of the
nearby residues (such as Pro229 and Tyr232 in CaTER and Glu234
and Gln237 in TdTER at the equivalent positions; Figure S4).
1
These authors contributed equally to this work.
Correspondence may be addressed to either of these authors (email [email protected] or [email protected]).
The structural co-ordinates reported in the PDB under accession codes 4EUH, 4EUE, 4EUF and 4FBG.
2
c The Authors Journal compilation c 2013 Biochemical Society
K. Hu and others
Figure S2 Structural comparison of TdTER with and without an NAD + bound
in the structure of the TdTER-NAD + complex
In an asymmetric unit there are 16 TdTER molecules that are arranged in two layers and each
layer comprises two pseudotetramers. There was fairly good electron density for a bound NAD +
at the active site in eight TdTER molecules, but poor density in the others; thus, only eight
TdTER molecules were modelled as the NAD + -bound form and the others as the apo form.
Structural comparison of the 16 TdTER molecules with or without an NAD + bound in the
asymmetric unit shows no notable conformational differences in the overall structure and in the
conformation of the substrate-binding loop with RMSDs of <0.40 Å for all Cα atoms. Shown is
the superimposition of one typical TdTER with an NAD + bound (yellow) and one typical TdTER
without an NAD + bound (blue).
Figure S1
Simulated annealing omit F o –F c maps for the bound cofactors
(A) A representative simulated annealing omit F o –F c map (1.0 σ contour level) for the bound
NADH in the CaTER–NADH complex. The final co-ordinates of NADH are shown as a stick
model. (B) A representative simulated annealing omit F o –F c map (1.0 σ contour level) for the
bound NAD + in the CaTER–NAD + complex. The final co-ordinates of NAD + are shown as a
stick model. (C) A representative simulated annealing omit F o –F c map (1.0 σ contour level) for
the bound NAD + in the TdTER–NAD + complex. The final co-ordinates of NAD + are shown
as a stick model.
c The Authors Journal compilation c 2013 Biochemical Society
Substrate specificity and catalytic mechanism of CaTER and TdTER
Figure S3
Structural comparisons of CaTER with FabVs
(A) Stereo view of the superimposition of the apo CaTER (green) and the apo XoFabV (PDB code 3S8M, orange) [2]. (B) Stereo view of the superimposition of the NADH-bound CaTER (cyan) and the
NADH-bound YpFabV (PDB code 3ZU3, light blue) [1]. CaTER shares a very high structural similarity with XoFabV and YpFabV. Superimposition of the apo CaTER with the apo XoFabV reveals an
RMSD of 1.39 Å for the 375 Cα atoms, and superimposition of the NADH-bound CaTER with the NADH-bound YpFabV yields an RMSD of 1.30 Å for the 378 Cα atoms with all secondary structure
elements superimposable.
c The Authors Journal compilation c 2013 Biochemical Society
K. Hu and others
Figure S4 Structural comparison of the cofactor-binding site in the NAD + bound CaTER (pink) and TdTER (yellow) structures
The bound cofactors and the residues which might contribute to the conformational difference
between Lys244 in CaTER and Lys249 in TdTER are shown as stick models and coloured
accordingly. The labels for CaTER and TdTER are coloured pink and black respectively.
REFERENCES
1 Hirschbeck, M. W., Kuper, J., Lu, H., Liu, N., Neckles, C., Shah, S., Wagner, S., Sotriffer,
C. A., Tonge, P. J. and Kisker, C. (2012) Structure of the Yersinia pestis FabV enoyl-ACP
reductase and its interaction with two 2-pyridone inhibitors. Structure 20, 89–100
2 Li, H., Zhang, X., Bi, L., He, J. and Jiang, T. (2011) Determination of the crystal structure
and active residues of FabV, the enoyl-ACP reductase from Xanthomonas oryzae . PLoS
ONE 6, e26743
3 Inui, H., Miyatake, K., Nakano, Y. and Kitaoka, S. (1986) Purification and some properties
of short chain-length specific trans -2-enoyl-CoA reductase in mitochondria of Euglena
gracilis . J. Biochem. 100, 995–1000
4 Hoffmeister, M., Piotrowski, M., Nowitzki, U. and Martin, W. (2005) Mitochondrial
trans -2-enoyl-CoA reductase of wax ester fermentation from Euglena gracilis defines a new
family of enzymes involved in lipid synthesis. J. Biol. Chem. 280, 4329–4338
5 Tucci, S. and Martin, W. (2007) A novel prokaryotic trans -2-enoyl-CoA reductase from the
spirochete Treponema denticola . FEBS Lett. 581, 1561–1566
Received 27 May 2012/10 October 2012; accepted 11 October 2012
Published as BJ Immediate Publication 11 October 2012, doi:10.1042/BJ20120871
c The Authors Journal compilation c 2013 Biochemical Society