Structural studies of a thermophilic esterase from a new Planctomycetes species, Thermogutta terrifontis

Article abstract of DOI:10.1111/febs.13326

Thermogutta terrifontis esterase (TtEst), a carboxyl esterase identified in the novel thermophilic bacterium T. terrifontis from the phylum Planctomycetes, has been cloned and over-expressed in Escherichia coli. The enzyme has been characterized biochemic

Full text of DOI:10.1111/febs.13326

Structural studies of a thermophilic esterase from a new
Planctomycetes species, Thermogutta terrifontis
Christopher Sayer1, Michail N. Isupov1, Elizaveta Bonch-Osmolovskaya2
and Jennifer A. Littlechild1
1 The Henry Wellcome Building for Biocatalysis, Biosciences, College of Life and Environmental Sciences, University of Exeter, UK
2 Winogradsky Institute of Microbiology of the Russian Academy of Sciences, Moscow, Russia
Keywords
Thermogutta terrifontis esterase (TtEst), a carboxyl esterase identified in the
novel thermophilic bacterium T. terrifontis from the phylum Planctomyce-
carboxyl esterase; Planctomyces; site-
directed mutagenesis; thermophile; X-ray
tes, has been cloned and over-expressed in Escherichia coli. The enzyme
has been characterized biochemically and shown to have activity towards
small p-nitrophenyl (pNP) carboxylic esters, with optimal activity for pNP-
structure
Correspondence
propionate. The enzyme retained 95% activity after incubation for 1 h at
80 °C. The crystal structures of the native TtEst and its complexes with the
substrate analogue D-malate and the product acetate have been determined
to high resolution. The bound ligands have allowed the identification of
the carboxyl and alcohol binding pockets in the enzyme active site. Com-
parison of TtEst with structurally related enzymes provides insight into
how differences in their catalytic activity can be rationalized based upon
the properties of the amino acid residues in their active site pockets. The
mutant enzymes L37A and L251A have been constructed to extend the
substrate range of TtEst towards the larger butyrate and valerate pNP-
esters. These mutant enzymes have also shown a significant increase in
activity towards acetate and propionate pNP esters. A crystal structure of
the L37A mutant was determined with the butyrate product bound in the
carboxyl pocket of the active site. The mutant structure shows an expan-
sion of the pocket that binds the substrate carboxyl group, which is consis-
tent with the observed increase in activity towards pNP-butyrate.
J. A. Littlechild, The Henry Wellcome
Building for Biocatalysis, Biosciences,
College of Life and Environmental Sciences,
University of Exeter, Stocker Road, Exeter
EX4 4QD, UK
Fax: +44 1392 723489
Tel: +44 1392 723468
E-mail: j.a.littlechild@exeter.ac.uk
(Received 27 March 2015, revised 8 May
2015, accepted 19 May 2015)
doi:10.1111/febs.13326
Database
The GenBank sequence accession number for the Thermogutta terrifontis esterase is KR002593.
The protein structures for T. terrifontis esterase and their complexes have been deposited in the
Protein Data Bank with codes: 4UHC (native), 4UHD (acetate bound), 4UHE (malate bound)
and 4UHF (L37A mutant with butyrate bound).
Introduction
The esterase enzymes are widely used as stereoselective
catalysts for the synthesis of optically pure molecules
for the pharmaceutical and agrochemical industries [1].
They are widely distributed in bacteria, archaea and
eukaryotes [2]. They have been used extensively in
industry as a result of their versatility, robustness and
stereoselectivity, as well as their ability to promote
synthetic reactions in organic solvents in the absence
of water [3]. A common use of these enzymes is the
resolution of chiral carboxylic acids, as well as pri-
Abbreviations
Agl, Aureobacterium species (–) c-lactamase; PcaD, Burkholderia xenovorans 3-oxiadipate-enol-lactonase; PDB, Protein Data Bank; PfCpo,
Pseudomonas fluorescens choroperoxidase; PfEst, Pseudomonas fluorescens esterase; pNP, p-nitrophenyl; TtEst, Thermogutta terrifontis
esterase.
2846
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
mary, secondary and tertiary alcohols. Examples
include the release of ferulic acid from plant cell wall
polysaccharides such as pectin or xylan [4] and the
catabolism of benzyl esters with the release of acid [5].
The resulting benzyl alcohols are then oxidized to ben-
zoates by dehydrogenases and further metabolized in
the b-ketoadipate pathway. The first crystal structure
of an esterase was that of acetyl-choline esterase from
Torpedo californica [6].
sia. The new organism represents a novel genus and
species in the phylum Planctomycetes and is the first
thermophilic member of this large and environmentally
important group. This organism grows up to 70 °C,
with an optimum for growth of 55–60 °C [17]. Similar
to other Planctomycetes, T. terrifontis is able to
degrade diverse natural polymeric substrates, including
a majority of polysaccharides and could be regarded
as a source of novel thermostable enzymes. The gen-
ome of T. terrifontis has been sequenced, assembled
and is now in the process of analysis (P. Menzel,
A. Elcheninov, A. Slesarev, N. Ravin, E. A. Bonch-
Osmolovskaya, S. R. Gudbersdottir, A. Krogh,
X. Peng & I. Kublanov, unpublished data). A potential
esterase/lactonase gene was identified using a bioinfor-
matics approach from the T. terrifontis genome
sequence. This gene TtEst was cloned and over-
expressed in Escherichia coli. The purified T. terrifontis
esterase (TtEst) enzyme has been biochemically charac-
terized and crystallized to determine the three-dimen-
sional structure, which provides insight into its
substrate specificity. The substrate range of the TtEst
has been extended using site-directed mutagenesis to
accept longer carboxyl chain esters.
The carboxyl esterases (EC 3.1.1.1) catalyse the
hydrolysis of the ester bond of relatively small water
soluble substrates. The carboxyl esterase NP is used for
the production of the nonsteriodal drug naproxen [7].
The racemic naproxen methylester is hydrolyzed to the
(S)-acid, which is separated from the (R)-methylester
to yield (S)-naproxen with a 99% nantiomeric excess
and a yield of 95%. Lipases (EC 3.1.1.3) catalyze a
related esterase reaction on larger water insoluble long-
chain fatty acids; however, the overall mechanism of
hydrolysis varies between the two enzyme classes [8].
The esterases obey classical Michaelis–Menten kinetics
and have a relatively open active site, whereas the lipas-
es require interfacial activation where the hydrophobic
lid domain moves away from the active site in the pres-
ence of a minimum concentration of the lipid substrate.
The lipases and most of the carboxyl esterases belong
to the a/b hydrolase fold family [9,10] with a catalytic
triad of Ser-His-Asp; however, esterase activity has
been reported for enzymes with different folds and cat-
alytic residues [11–13].
The proposed mechanism of catalysis for carboxyl
esterases resembles the serine protease mechanism
described by Blow et al. [14] involving the nucleophilic
attack by the catalytic serine hydroxyl on the carbonyl
carbon of the scissile bond. The carbonyl oxygen binds
to the oxyanion hole, which is usually formed by two
main chain nitrogen atoms. The attack by the catalytic
serine hydroxyl results in a tetrahedral intermediate.
Subsequently, the alcohol product is released and an
acyl-enzyme complex is formed, which is attacked by a
water molecule. Another tetrahedral intermediate is
formed, which collapses to release the carboxyl prod-
uct and the free enzyme. The catalytic serine residue in
a/b hydrolase fold esterases is usually located in a
tight nucleophilic elbow with the consensus sequence
Gly-X-Ser-X-Gly [15], although not exclusively. The
two glycine residues of this motif are not conserved in
the Alcaligenes esterase [16].
Results and Discussion
Biochemical characterization
The TtEst enzyme was successfully over-expressed in
E. coli Arctic express (RIL) strain, and purified using
a nickel affinity column and gel filtration chromatogra-
phy. The protein eluted as a monomer from a cali-
brated size exclusion column. The esterase activity was
tested using short chain p-nitrophenyl (pNP) esters as
substrates (Fig. 1). The enzyme was most active
towards pNP-acetate (0.64 lmolꢀminꢁ1ꢀmgꢁ1
)
and
In the present study, we report the discovery, clon-
ing and biochemical and structural characterization of
an esterase from a novel thermophilic bacterium Ther-
mogutta terrifontis, which has been recently isolated
from a terrestrial hot spring on Kunashir Island, Rus-
Fig. 1. The specific activity of the native TtEst and the L37A and
L251A mutant enzymes towards pNP-esters with varying carbon
chain lengths. Enzyme activity was measured monitoring pNP
production [31].
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2847

Structural studies of a thermophilic esterase
C. Sayer et al.
propionate (1.07 lmolꢀminꢁ1ꢀmgꢁ1
)
towards pNP-butyrate
esters, with
a
the pH and the presence of different ligands affect the
conformation of many residues, both charged and
hydrophobic. Effects are also observed regarding the
occupancies of different alternative conformers of both
the side and main chains and significantly different con-
formations of the visible part of the C-terminal His-tag.
much
lower
activity
(0.16 lmolꢀmin–1ꢀmgꢁ1
) and no detectable activity
towards pNP-valerate. The gene from the T. terrifontis
genome was identified using a bioinformatics approach
as an esterase or an enol-lactonase. The enzyme
showed no activity to lactonase substrates (4-hydroxy-
3-pentenoic acid c-lactone, c-butyrolactone and c-val-
erolactone) and was therefore established as an ester-
ase. The thermostability of the enzyme was tested by
incubation at increasing temperatures, cooling and
measuring the residual activity. This procedure showed
that the enzyme retained 95% of its activity after incu-
bation for 1 h at 80 °C and 22% of its activity at
85 °C. No activity was observed after incubation of
the enzyme at 90 °C or above.
Overall fold
The TtEst is a monomer in the crystal form, com-
prised of two domains: an a/b hydrolase fold core
domain (Met1-Arg129 and Arg200-Ala275) and a cap
domain (Pro130-Glu199) (Fig. 2). Six helices of the
core domain surround an eight stranded b-sheet with
connectivity
1,2,ꢁ1x,2x,1x,1x,1x
and
direction
+ ꢁ + + + + + + [18]. The active site is located at the
interface of the two domains and consists of the cata-
lytic triad of Ser101, His 250 and Asp222 . The cata-
lytic serine is located in a tight nucleophilic elbow at
the end of strand b5 with the conserved esterase signa-
ture sequence.
Structure of TtEst
Quality of the models
The TtEst and its complexes crystallized in the space
group P3221 with unit cell parameters: a = b = 43.3,
c = 227.1 A, a = b = 90°, c = 120°. The asymmetric
Active site
ꢀ
unit contained one monomer, giving a solvent content
of 39%. All residues were modelled with the exception
of Met1 because no ordered electron density was
observed for this residue. Parts of the C-terminal His-
tag were modelled in some of the structures. The struc-
The enzyme was originally crystallized in a single con-
dition containing poly(ethylene glycol) 600, sodium
Hepes at pH 7.5, which allowed the determination of a
native enzyme structure. A micro-seeding approach
[19] has produced crystals in many more conditions in
a range of precipitants and at various pH values with
different additives. These crystals were used for sub-
strate/inhibitor soaks. Although most tested ligands
failed to bind, a crystal from one of these conditions
contained the substrate analogue malate and another
crystal contained the product acetate bound in the
active site at high occupancy.
ꢀ
tures were refined to a resolution of 1.2 A or better
with a final Rfree of 14.2 or better (Table 1). The com-
plex of the L37A mutant of the TtEst with butyrate
crystallized in the same space group as the native
enzyme with similar cell parameters and was refined to
ꢀ
a resolution of 1.1 A. The D-malate and acetate
ligands in the native TtEst, and the butyrate ligand in
the L37A mutant were modelled at full occupancy.
The proline residues 13 and 36 were modelled in the
cis conformation. Many side chains and several residue
main chains were modelled in alternative conforma-
tions. The active site Ser101 is the only outlier on the
Ramachandran plot.
The electron density calculated for crystals grown
from racemic imidazole malate and poly(ethylene gly-
col) 4000 at pH 6.0 contained clearly defined features
corresponding to a D-malate molecule in the active site
(Fig. 3a). The malate is bound in the active site by its
two carboxyl groups; the proximal one is hydrogen-
bonded to the backbone oxygen of Phe35 and to the
side chain of His250. The distal carboxyl group forms
a salt bridge to Arg139 and is hydrogen bonded to the
phenolic oxygen of Tyr105 (Fig. 3b,c). The hydroxyl
Influence of pH and ligands on the enzyme structure
The four TtEst structures reported in the present study,
as well as many others not reported here, were obtained
by crystallization at both low and physiological pH in
the presence of ligands. These crystals all belonged to
the same space group P3221 and diffracted beyond a res-
ꢀ
oxygen of D-malate is in close proximity (3.5 A) to the
side chains of His250 and the catalytic Ser101 and
makes hydrogen bonds with water. The other residues
in the vicinity of the D-malate molecule (Ile224,
Arg127, Tyr105, Met197 and Phe35) map out the alco-
hol binding pocket where the pNP part of the ester
substrate binds for hydrolysis.
ꢀ
olution of 1.2 A. They were subsequently soaked with
different ligands at both physiological and low pH. It
appears that, although the TtEst retains its overall fold,
2848
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
Table 1. Summary of the data processing and refinement statistics.
Crystal
Native
Acetate
Malate
L37A butyrate
Beamline
I03
I04
I04
I02
ꢀ
Wavelength (A)
Space group
0.9763
0.9795
0.9795
0.9795
P3221
P3221
P3221
P3221
Unit cell a, b, c
ꢀ
43.2, 43.2, 226.6,
90, 90, 120
1
43.3, 43.3, 227.
1, 90, 90, 120
1
43.3, 43.3, 226.6,
90, 90, 120
1
43.2, 43.2, 227.8,
90, 90, 120
1
(A) a, b, c (°)
Number of protein
monomers per
asymmetric unit
Solvent content %;
38.7; 2.02
40.3; 2.06
40.4; 2.06
39.2; 2.04
3
–1
ꢀ
Vm (A ꢀDa
)
ꢀ
Resolution range (A)
75.5–1.03
(1.06–1.03)a
8.0 (2.9)
121070
98.4 (84.0)
6.1 (62.2)
18.5 (2.4)
7.9
37.48–1.07
(1.10–1.07)
9.1 (5.5)
99937
37.52–1.16
(1.19–1.16)
9.6 (9.6)
86054
37.97–1.08
(1.11–1.08)
9.3 (8.8)
107910
100 (99.9)
7.1 (111.8)
14.8 (2.0)
8.9
Multiplicity
Unique reflections
Completeness (%)
b
Rsym (%)
90.5 (51.1)
3.4 (28.2)
33.3 (5.6)
7.7
98.6 (95.6)
9.3 (96.5)
12.2 (2.2)
8.4
I/r (I)
Wilson B factor (A )
2
ꢀ
Refined residues
Refined
278
274
272
278
438
366
463
324
water molecules
Refined
ligandc molecules
Rcryst (%)d
2 9 acetate
1 9 malate,
1 butyrate
10.9
12.9
9.4
10.7
14.2
11.5
13.6
Rfree (%) (5%
of total data)
11.2
0.010 (0.019)e
1.6 (1.9)
0.014 (0.019)
1.8 (2.0)
0.015 (0.019)
1.8 (2.0)
0.014 (0.019)
ꢀ
rmsd bond lengths (A)
rmsd bond angles (o)
Occupancy of ligand
1.8 (2.0)
Acetate 1 = 1.0
Acetate 2 = 0.7
Malate = 1.0
Butyrate = 0.75 + 0.25
2
ꢀ
Mean B factor (A )
Protein
Solvent
Ligands
13.3
27.7
13.0
26.6
13.5
13.7
30.6
11.4
15.9
32.7
15.3
Ramachandran plot analysis, residues in (%)
Most favoured regions
Additional allowed regions
Generously allowed regions
Disallowed regions
91.8
7.8
0
90.8
8.7
0
93.0
6.6
0
92.2
7.3
0
0.4
0.4
0.4
0.4
Wilson B-factor was estimated by SFCHECK [46]. Ramachandran plot analysis was performed using PROCHECK [44].
aValues in parentheses are given for the outer resolution shell.
bRsym = ΣhΣJ|<Ih> ꢁ IJ(h)|/ΣhΣJIJ(h), where I(h) is the intensity of reflection h. Σh is the sum over all reflections and ΣJ is the sum over J
measurements of the reflection.
cOnly molecules within the active site cavity are considered ligands.
dRcryst = Σ||Fo| ꢁ |Fc||/Σ|Fo|.
eTarget values are given in parenthesis.
The proximal carboxyl of the D-malate binds with
one oxygen in the oxyanion hole. The remainder of
the D-malate ligand corresponds to the leaving alcohol
group of the putative substrate, (3R)-3,4-dihydroxybu-
tyrate carboxyl ester, which is structurally similar to
L-b-homoserine carboxyl ester.
which is present in equal amounts in the crystallization
medium. This could be attributed to the formation of
a hydrogen bond between the hydroxyl oxygen of D-
malate and a water molecule. The hydroxyl oxygen of
the L-malate would have no partners for hydrogen
bond formation because it would point towards the
hydrophobic side chains of Phe35 and Met158. The
preferred binding of D-malate suggests that TtEst has
The electron density clearly shows preference of
binding of D-malate rather than its L-enantiomer,
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2849

Structural studies of a thermophilic esterase
C. Sayer et al.
A
B
Fig. 2. A cartoon representation of the TtEst structure showing the
core and the cap domains. The carbon atoms for the catalytic triad
residues are shown in blue, the residues of the alcohol binding
pocket are shown in navy and the residues of the carboxyl binding
pocket are shown in magenta.
the potential to be used for the resolution of racemic
substrates such as b-homoserine esters.
An electron density corresponding to an acetate
molecule bound in the active site (Fig. 4a) was found
in the crystals grown from ammonium acetate and
poly(ethylene glycol) 10 000 at pH 4.0. The acetate
molecule is bound with one carboxyl oxygen atom in
close proximity to the catalytic serine, whereas the sec-
ond carboxyl oxygen is hydrogen-bonded to the back-
bone amides of Phe35 and Met102, which form the
oxyanion hole (Fig. 4b). This corresponds to the
expected orientation that the corresponding ester sub-
strate would occupy during the course of the reaction.
This positions the carbon atom of acetate into the car-
boxyl binding pocket located at the bottom of the
active site (Fig. 4c). This area is clearly defined by the
hydrophobic side chain residues of Leu37, Met41,
Leu100 and Leu251, in addition to the backbone of
residues His33-Phe35. The size of this pocket limits the
activity of TtEst to smaller carboxyl group substrates
and explains the low activity observed for pNP-buty-
rate. In particular, Leu251 and Leu37 would sterically
hinder the binding of pNP-butyrate because structural
rearrangements of these side chains would be required
for substrate binding (Fig. 4b).
C
Fig. 3. The interactions of the bound substrate analogue D-malate
in the active site cavity of the TtEst, which define the alcohol
binding pocket. (A) Electron density maps for the malate binding
site clearly show the preference of TtEst for the D-isomer of the
racemic malate present in the crystallization droplet. The 2Fo ꢁ Fc
map (blue) is contoured at 1.6 r and the Fo ꢁ Fc map is
contoured at 3.1 r (green) and ꢁ3.1 r (red). The ligand and amino
acid residues are shown as stick models. The solvent molecules
are shown as red stars. (B) Structural features of the alcohol
binding site. The ligand and the side chains of the key amino acid
residues are shown as stick models and the protein backbone is
shown as a cartoon. Hydrogen bonds and ionic interactions are
shown as dashed lines. (C) The key interactions of the D-malate
Interestingly, multiple attempts to obtain complexes
of the native TtEst with different carboxylic acids,
including acetate, at higher pH have produced struc-
tures with the carboxyl group of the ligand bound in
the same orientation as the proximal carboxyl of
molecule in the TtEst active site are shown using
diagram [47].
a LIGPLOT+
2850
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
Fig. 4. The active site cavity of the TtEst with the product acetate
bound, highlighting the carboxyl binding pocket. (A) The electron
density maps showing the location of the acetate binding site. The
2Fo ꢁ Fc map (blue) is contoured at 1.6 r and the Fo ꢁ Fc map is
contoured at 3.1 r (green) and ꢁ3.1 r (red). The ligand and amino
acid residues are shown as stick models. (B) Structural features of
the carboxyl binding site responsible for restricting the size of the
substrate acyl group. The acetate molecule and the side chains of
the key amino acid residues are shown as stick models and the
protein backbone is shown as a cartoon. Hydrogen bonds are
shown as dashed lines. (C) The electrostatic potential surface of
the TtEst viewed from the outside towards the active site cavity.
The positive charge is shown in blue and the negative charge is
shown in red. The acetate molecule was not used for the surface
calculation and is shown as a stick model. The carbons of the
acetate molecule bound at the bottom of the active site are shown
in yellow and the oxygens are shown in red. The image was
generated using CCP4 MG [48].
A
B
Manual positioning of pNP-carboxyl substrates in
the active site of the TtEst was performed on the basis
of the enzyme D-malate and acetate complexes. It
shows that the nitro group of the substrate occupies
the same position as the distal carboxyl of the malate
making a salt bridge to Arg139 and hydrogen bond to
Tyr105. The acetate group of the pNP-acetate super-
imposes closely with the acetate molecule in the corre-
sponding structure and fits into the carboxyl binding
pocket with the ester bond in the catalytic position.
The active site cavity of the TtEst can therefore be
divided into the carboxyl binding pocket, which
defines the length of the hydrolysed carboxyl ester,
and the alcohol binding pocket, which is located closer
to the edge of the active site funnel.
C
Site-directed mutagenesis
The largest pNP carboxyl ester for which limited TtEst
esterase activity was detected is pNP-butyrate (Fig. 1).
It was suggested that the length of the cleaved carboxyl
group is restricted by the size of the TtEst carboxyl
binding pocket. The side chains of Leu37 and Leu251
were identified as sterically hindering the binding of sub-
strates with longer carboxyl groups and were selected as
targets for mutagenesis to expand the carboxyl binding
pocket. Two mutant enzymes with either of these resi-
dues replaced by alanine were prepared (L37A and
L251A). Both mutant enzymes showed an approximate
five- to seven-fold increase in activity for pNP-butyrate
and some activity towards pNP-valerate (Fig. 1). Unex-
pectedly, both mutants also showed at least a two-fold
increase in activity towards pNP-acetate and a signifi-
cant increase in activity towards pNP-propionate.
malate. We attribute this to the charge of the carbox-
ylic acid at physiological pH. At an acidic pH of 4.0,
the acetate molecule carries no charge and successfully
mimics the substrate/product binding. This also sug-
gests a mechanism of product removal from the
enzyme active site at physiological pH where the
charged product loses affinity for the active site and is
displaced.
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2851

Structural studies of a thermophilic esterase
C. Sayer et al.
The structure of the L37A mutant enzyme from a
crystal soaked in butyrate at pH 4.0 revealed electron
density in the active site corresponding to this ligand
at high occupancy (Fig. 5a). The butyrate side chain in
the expanded carboxyl binding pocket is shown with
the C4 atom occupying the equivalent position of the
leucine side chain in the native enzyme. This confirms
the role of this pocket in the binding of the ester car-
boxyl group (Fig. 5b) and rationalizes the observed
increase in activity towards pNP-butyrate observed for
this mutant enzyme.
Structural comparison
A
Structures have been reported for many enzymes of
the a/b hydrolase family that catalyse a wide diversity
of enzymatic reactions [20,21]. A BLAST [22] search of
TtEst against the structural database reveals its closest
homologues with known structure. These are 3-oxoadi-
pate-enol lactonase from Burkholderia xenovorans
Lb400, which was used as a MR model [PcaD; 29%
sequence identity; Protein Data Bank (PDB) code
2XUA] [23], Pseudomonas fluorescens esterase (PfEst;
30%; PDB code: 3HEA) [24], Aureobacterium species
(ꢁ) c-lactamase (Agl; 30%; PDB code: 1HKH) [25] in
addition to several bacterial nonheme haloperoxidases,
such as P. fluorescens choroperoxidase (PfCpo; 28%;
PDB code: 1A8S) [26]. Esterase activity towards pNP-
acetate has been reported for most of these proteins.
The quaternary structures of these proteins are not
conserved: although haloperoxidases, PfEst and Agl
are trimeric proteins, PcaD forms a stable dimer and
the TtEst reported here is monomeric in both solution
and in the crystal form.
B
Because the catalytic triad residues and the position
of the oxyanion hole are conserved between these
enzymes, it appears that the type of catalytic reaction
and the substrate specificity depend on the residues of
the active site pockets. An overall comparison of the
alcohol binding pocket in TtEst with the equivalent
pocket in PcaD and Agl shows that the TtEst pocket
has a much more polar and charged environment in
the active site, which allows the binding of organic
acids such as D-malate where the distant carboxyl is
coordinated by Arg139 and Tyr105. The PcaD and
Agl enzymes have more hydrophobic substrate binding
pockets, with the residues Arg139 and Tyr105 in TtEst
replaced by Trp135 and Ile129 in PcaD and Trp204
and Leu125 in Agl. In both PcaD and Agl, the active
site is better suited for the binding of structures such
as lactone and c-lactam rings. Similarly, the alcohol
pocket in the PfEst structure is lined with several
hydrophobic phenylalanine side chains that should
have affinity for the lactone ring. This would explain
its lactonase activity towards caprolactone [27] in addi-
tion to the esterase activity.
Fig. 5. The binding of butyrate in the extended carboxyl binding
pocket of the L37A TtEst mutant enzyme. (A) Electron density map
for the butyrate molecule bound in the active site is shown in the
same orientation as in Fig. 3a. The mutation of Leu37 to Ala allows
the larger carboxyl group to fit into the smaller carboxylic pocket
explaining its increased activity towards longer carbon chain esters
compared to the native enzyme. The 2Fo ꢁ Fc map (blue) is
contoured at 1.5 r and the Fo ꢁ Fc map is contoured at 3.2 r
(green) and ꢁ3.2 r (red). The amino acid residues are shown as
stick models. (B) The electrostatic potential surface of the L37A
mutant TtEst enzyme butyrate complex viewed from the outside
towards the active site cavity from the same viewpoint as in
Fig. 3c. The positive charge is shown in blue and the negative
charge is shown in red. The butyrate molecule was not used for
the surface calculation and is shown as a stick model. The carbons
of the butyrate molecule bound at the bottom of the active site are
shown in green and the oxygens are shown in red. The image was
generated using CCP4 MG [48].
The carboxyl pocket of the TtEst has been described
above. The trapped tetrahedral intermediate complex
structure of the PfEst mutant enzyme between the cat-
2852
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
alytic serine and ethyl acetate [24] reveals the orienta-
tion of the acetate group consistent with that of the
free acetate in the TtEst structure (Fig. 6). The car-
boxyl pocket in the PfEst is lined with the side chains
of Phe93 and Phe162 and appears to be more restric-
tive with respect to the size of the carboxyl group than
the equivalent pocket in the TtEst. The side chain of
Phe166 in Agl, which replaces Leu162 in TtEst, lies in
close proximity to the methyl group of the acetate
ligand in the TtEst structure. This would prevent the
propionate ester binding and restrict the acetate ester
binding. Indeed, only limited pNP-acetate esterase
activity has been reported for the Agl enzyme [28].
The side chain of Ser34 in PcaD, which replaces gly-
cine in other related structures, may inhibit its esterase
activity altogether.
esterase and related hydrolase enzymes. The high ther-
mal stability of the TtEst enzyme will make it a good
addition to the existing toolbox of esterase enzymes
used for industrial biocatalysis.
Materials and methods
Cloning and over-expression
Potential hydrolase sequences were identified in the T. terri-
fontis genome using the ANASTASIA galaxy pipeline
[29,30]. The putative TtEst gene was amplified by PCR
from genomic DNA under standard conditions using Pfu
polymerase (Thermo Scientific, Hemel Hempstead, UK)
with the primers: forward 50-AGAAGGAGATATAACT
ATGGCACAGCGTG-30 and reverse 50-GTGGTGGTGA
TGGTGATGGCCTTACTCAGTATGT-30. The DNA was
annealed into the pLATE31 vector (which contains a C-ter-
minal His tag) using the aLICator Ligation Independent
Cloning and Expression System (Thermo Scientific). The
vector harbouring the TtEst gene was transformed into
ArcticExpress (DE3)RIL E. coli cells (Agilent, Santa Clara,
CA, USA) that contain rare codons and chaperones to
assist protein folding at lower temperatures.
Interestingly, PfCpo has a propionate ligand bound
in the active site, with this complex being obtained by
co-crystallization at pH 6.6 [26]. Its position is equiva-
lent to the position of the proximal carboxyl of D-
malate in TtEst with the hydrophobic ethyl ‘handle’
pointing away from the smaller carboxyl binding
pocket. We have observed similar binding for a range
of carboxylic acids in the course of our attempts to
soak/co-crystallize different carboxylic acids into TtEst
at physiological pH (not shown). It would appear that
the carboxylic acid has to be uncharged to bind to the
carboxyl site in its catalytic orientation. This is sup-
ported by binding of the butyrate ligand to the L37A
mutant enzyme at pH 4.0.
The E. coli harbouring the TtEst plasmid was cultured
overnight in LB medium (100 mL). The overnight culture
was used to inoculate 1 L of LB media and incubated at
37 °C with shaking at 250 r.p.m. until an D600 of 0.4 was
reached. The temperature of the incubator was lowered to
11 °C and the cultures were cooled for 1 h prior to being
induced with the addition of 1 mM IPTG. The cultures
were left shaking at 12 °C for ~ 48 h. The cells were then
harvested at 4700 g at 4 °C for 20 min. The pellet was then
The present study provides insight into the differing
substrate specificity that has evolved in this class of
Fig. 6. A stereo diagram showing the superimposition of the acetate complex of TtEst (green) and the mutant PfEst (blue) with a trapped
ethyl acetate tetrahedral intermediate complex. Despite the differences in the carboxyl binding pockets between TtEst and PfEst, the
orientation of the methyl group of acetate is similar. The key residue side chains are shown as sticks and the protein backbone is shown as
a cartoon.
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2853

Structural studies of a thermophilic esterase
C. Sayer et al.
re-suspended in 25 mM Tris-HCl (pH 8.0), 0.5 M NaCl and
20 mM imidazole. Cells were disrupted by sonication at
10 lm (Soniprep 150; MSE (UK) Ltd, London, UK) on ice
for 4 min and the cell debris was removed by centrifugation
at 20 000 g at 4 °C for 30 min. The clarified cell lysate was
then heat treated at 65 °C for 30 min before being centri-
fuged at 20 000 g at 4 °C for 30 min to remove any dena-
tured proteins.
of silicon and paraffin oils) before being stored at 20 °C, and
was regularly checked for growth of crystals using a light
microscope.
Crystals appeared within 1 week. Native crystals were
grown from 50 mM sodium Hepes (pH 7.5) and 18%
poly(ethylene glycol) 600 and were frozen directly from
the droplet. Remaining crystals from this droplet were
used in microbatch microseeding for further crystalliza-
tion screening [19]. Crystals were broken up in the micro-
batch well using a tip, and pipetted in to a solution
(50 lL) containing 50 mM sodium Hepes (pH 7.5) and
18% poly(ethylene glycol) 600 and vortexed briefly. This
‘seed stock’ was used in the next round of crystal screen-
ing (The Stura Footprint ScreenTM + MacroSolTM HT-96
screen) with volumes of 0.3 lL of protein, 0.2 lL of
screen and 0.1 lL of seeds.
Microbatch seed screening produced crystals containing
bound acetate or D-malate, which were used as crystalliza-
tion additives. Acetate complex crystals were grown from
33 mM ammonium acetate (pH 4.0) and 7% poly(ethylene
glycol) 10 000, and were frozen in a cryo-protectant con-
taining 50 mM ammonium acetate, 11% poly(ethylene gly-
col) 10 000 and 30% poly(ethylene glycol) 400. Malate
complex crystals were grown from 33 mM racemic imidaz-
ole malate (pH 6.0) and 10% poly(ethylene glycol) 4000,
and frozen in a cryo-protectant containing 50 mM imidaz-
ole malate, 15% poly(ethylene glycol) 4000 and 30% poly
(ethylene glycol) 400. Mutant L37A crystals were grown
from 50 mM sodium cacodylate (pH 6.5) and 18% poly
(ethylene glycol) 2000 monomethyl ether. Mutant crystals
were soaked for 30 min and subsequently frozen in a cryo-
protectant containing 50 mM sodium butyrate, 50 mM
potassium hydrogen phthalate buffer, 33 mM sodium caco-
dylate (pH 4.0) 12% poly(ethylene glycol) 2000 monometh-
yl ether, 50 mM NaCl and 35% poly(ethylene glycol) 400.
The TtEst mutants L37A and L251A were prepared
using the QuikChange Lightning Site-Directed Mutagenesis
Kit (Agilent) in accordance with the manufacturer’s
instructions. The mutant constructs were over-expressed in
the same way as the native protein.
The native and mutant proteins were purified using a 1-
mL His-Trap FF crude column (GE Healthcare, Little Chal-
font, UK) using an elution gradient from 20 to 500 mM imid-
azole in 20 mM Tris-HCl (pH 8.0) and 0.5 M NaCl. The
enzyme was then applied to a calibrated Superdex 200 Hi-
Load 16/60 gel filtration (GF) column (GE Healthcare) and
eluted with 1 column volume in a buffer of 20 mM Hepes and
0.5 M NaCl (pH 8.0) at 1.0 mLꢀminꢁ1
.
Activity assays
Activity for the hydrolysis of an ester was determined at
25 °C measuring pNP production from its carboxylic esters:
pNP-acetate, pNP-propionate, pNP-butyrate and pNP-valer-
ate in a modification of the method described by Armstrong
et al. [31]. Reactions were carried out in a total volume of
100 lL containing a final concentration of 20 mM Hepes buf-
fer, 100 mM NaCl (pH 7.5), 50 lgꢀmLꢁ1 enzyme and 1 mM
substrate, and the change in D410 was recorded.
The thermal stability of TtEst was tested by incubating
the enzyme at a concentration of 1 mgꢀmLꢁ1 in 20 mM He-
pes, 0.5 M NaCl (pH 8.0) at 50, 60, 70, 80, 90 and 100 °C
for 1 h. Enzyme aliquots (100 lL) were withdrawn at
appropriate times and cooled on ice before the residual
activity was measured using the method described above.
Lactonase activity was tested using the method previously
described by Hiblot et al. [32]. Assays were carried out in
Bicine 2.5 mM (pH 8.3), NaCl 150 mM, Cresol purple
0.25 mM. Cresol purple (pKa 8.3 at 25 °C) was used as a
pH indicator to follow the acidification related to the
lactone ring hydrolysis.
X-ray data collection and structure solution
Data were collected on beamlines I02, I03 and I04 at the
Diamond Synchrotron light source (Didcot, UK) at 100 K
in a stream of gaseous nitrogen using a Pilatus detector
(Dectris Ltd, Baden, Switzerland). Data were processed
and scaled using XDS [33] and AIMLESS [34] in the XIA2 [35]
pipeline. All further data and model manipulation was car-
ried out using the CCP4 suite [36].
Phases for the native structure were determined using the
molecular replacement method implemented in MOLREP [37]
using the PcaD structure, which shares 29% sequence iden-
tity to TtEst as a model [23].
Crystallization
The TtEst was concentrated to ~ 15 mgꢀmLꢁ1 using a 10-
kDa membrane Vivaspin (Vivaproducts, Littleton, MA,
USA) and microbatch crystallization trials were set up using
an Oryx 6 crystallization robot (Douglas Instruments, Hung-
erford, UK) and The Stura Footprint ScreenTM + Macro-
SolTM HT-96 screen (Molecular Dimensions, Newmarket,
UK). The droplet contained a 50 : 50 ratio of protein
solution to screen and was covered with Al’s oil (50 : 50 mix
The rotation function was calculated with an integration
ꢀ
ꢀ
radius of 32 A at a resolution of 2.5 A using the monomeric
2XUA model, which gave a prominent rotation peak of
7.6 r when the MOLREP model modification based on
sequence alignment was used [38]. The translation search was
ꢀ
conducted at a resolution of 4 A. The monomer model was
2854
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
positioned with a correlation coefficient of 0.27, whereas
background peaks had a correlation no higher than 0.21.
However the MR solution did not refine at a resolution of
2 Montella IR, Schama R & Valle D (2012) The
classification of esterases: an important gene family
involved in insecticide resistance – a review. Mem Inst
Oswaldo Cruz 107, 437–449.
ꢀ
1.1 A using the ARP/WARP procedure [39] or REFMAC5 [40]
with Rfree as high as 0.56 after 40 cycles of refinement. There-
fore, the MR solution was subjected to eight cycles of SHELXE
phase extension procedure [41]. The resulting model, which
contained 269 alanine residues, had a correlation for the par-
tial structure against the native data of 48.34%, with only
39% of starting Ca atoms of the MR model being within
3 Bornscheuer UT (2002) Microbial carboxyl esterases:
classification, properties and application in biocatalysis.
FEMS Microbiol Rev 26, 73–81.
4 de Vries RP & Visser J (1999) Regulation of the
feruloyl esterase (faeA) gene from Aspergillus niger.
Appl Environ Microbiol 65, 5500–5503.
5 Jones RM, Collier LS, Neidle EL & Williams PA (1999)
Are ABC genes determine the catabolism of aryl esters
in Acinetobacter sp. Strain ADP1. J Bacteriol 181,
4568–4575.
6 Sussman JL, Harel M, Frolow F, Oefher C, Goldman
A, Toker L & Silman I (1991) Atomic structure of
acetylcholinesterase from Torpedo californica: a
prototypic acetylcholine-binding protein. Science 253,
872–879.
7 Quax WJ & Broekhuizen CP (1994) Development of a
new Bacillus carboxyl esterase for use in the resolution
of chiral drugs. Appl Microbiol Biotechnol 41, 425–431.
8 Schrag J & Cygler M (1997) Lipases and alpha/beta
hydrolase fold. Methods Enzymol 284, 85–107.
9 Ollis DL, Cheah E, Cygler M, Dijkstra B, Frolow F,
Franken SM, Harel M, Remington SJ, Silman I,
Schrag J et al. (1992) The a/b hydrolase fold. Protein
Eng 5, 197–211.
10 Heikinheimo P, Goldman A, Jeffries C & Ollis DW
(1999) Of barn owls and bankers: a lush variety of a/b
hydrolases. Structure 7, 141–146.
11 Wei Y, Schottel JL, Derewenda U, Swenson L, Patkar
S & Derewenda ZS (1995) A novel variant of the
catalytic triad in the Streptomyces scabies esterase. Nat
Struct Biol 2, 218–223.
ꢀ
1 A of their positions in the SHELXE model.
The resulting SHELXE structure was subjected to the ARP/
WARP procedure, prior to manual model building in COOT
[42], followed by refinement with REFMAC5. The dictionary
definitions for the ligands were created using JLIGAND [43].
The statistics of the data processing and parameters of the
final refined models are provided in Table 1. After refine-
ment, the quality of the model was checked using PROCHECK
[44]. Images were created using PYMOL [45] unless otherwise
stated.
Acknowledgements
The present study was supported by the Hotzyme pro-
ject (grant agreement number 265933) financed by the
European Union 7th Framework Programme FP7/
2007–2013. MI would like to thank the BBSRC and
the University of Exeter for support. The authors
would like to thank the Diamond Synchrotron Light
Source for access to beamlines I02, I03 and I04 (pro-
posal number MX8889) and the beamline scientists for
assistance. In addition, we would like to thank Nikolai
Ravin, Andrey Mardanov and Vitaly Kadnikov from
the Russian Academy of Sciences for the sequencing
of the T. terrifontis genome. Other partners in the
Hotzyme project are also thanked for the establish-
ment of the ANASTASIA platform for bioinformatic
analysis and for their support of this project.
12 Wagner UG, Petersen EI, Schwab H & Kratky C
(2002) EstB from Burkholderia gladioli: a novel esterase
with a b-lactamase fold reveals steric factors to
discriminate between esterolytic and b-lactam cleaving
activity. Protein Sci 11, 467–478.
13 Isupov MN, Brindley AA, Hollingsworth EJ,
Murshudov GN, Vagin AA & Littlechild JA (2004)
Crystallization and preliminary X-ray diffraction
studies of a fungal hydrolase from Ophiostoma
novo-ulmi. Acta Crystallogr D Biol Crystallogr 60,
879–882.
14 Blow DM, Birktoft JJ & Hartley BS (1969) Role of a
buried acid group in the mechanism of action of
chymotrypsin. Nature 221, 337–340.
Author contributions
C. Sayer cloned, over-expressed and carried out bio-
chemical and structural characterisation of the enzyme.
M. Isupov was involved in structural determination of
the enzyme. E. Bonch-Osmolovskaya was responsible
for isolation of the novel genome. J. Littlechild over-
saw the project. All authors contributed to the writing
of the manuscript.
15 Nardini M & Dijkstra BW (1999) Alpha/beta hydrolase
fold enzymes: the family keeps growing. Curr Opin
Struct Biol 9, 732–737.
References
16 Bourne PC, Isupov MN & Littlechild JA (2000) The
atomic-resolution structure of a novel bacterial esterase.
Structure 8, 143–151.
1 Faber K (1997) Biotransformations of non-natural
compounds: state of the art and future developments.
Pure Appl Chem 69, 1613–1632.
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2855

Structural studies of a thermophilic esterase
C. Sayer et al.
17 Slobodkina GB, Kovaleva OL, Miroshnichenko ML,
Slobodkin AI, Kolganova TV, Novikov AA, van
Heerden E & Bonch-Osmolovskaya EA (2014)
Thermogutta terrifontis gen. nov., sp. nov. and
Thermogutta hypogea sp. nov., novel thermophilic
anaerobic representatives of the phylum
31 Armstrong JM, Myers DV, Verpoorte JA & Edsall
JT (1966) Purification and properties of human
erythrocyte carbonic anhydrases. J Biol Chem 241,
5137–5149.
32 Hiblot J, Gotthard G, Chabriere E & Elias M (2012)
Structural and enzymatic characterization of the
lactonase SisLac from Sulfolobus islandicus. PLoS One
7, e47028.
Planctomycetes. Int J Syst Evol Microbiol 65, 760–765.
18 Richardson JS (1981) The anatomy and taxonomy of
protein structure. Adv Prot Chem 34, 167–339.
19 Stewart PDS, Kolek SA, Briggs RA, Chayen NE &
Baldock PFM (2011) Random microseeding: a theoretical
and practical exploration of seed stability and seeding
techniques for successful protein crystallization. Cryst
Growth Des 11, 3432–3441.
20 Carr PD & Ollis DL (2009) Alpha/beta hydrolase fold:
an update. Protein Pept Lett 16, 1137–1148.
21 Marchot P & Chatonnet A (2012) Enzymatic activity
and protein interactions in alpha/beta hydrolase fold
proteins: moonlighting versus promiscuity. Protein Pept
Lett 19, 132–143.
22 Altschul S, Gish W, Miller W, Myers E & Lipman D
(1990) Basic local alignment search tool. J Mol Biol
215, 403–410.
23 Bains J, Kaufman L, Farnell B & Boulanger MJ
(2011) A product analog bound form of 3-
oxoadipate-enol-lactonase (PcaD) reveals a
multifunctional role for the divergent cap domain. J
Mol Biol 406, 649–658.
33 Kabsch W (2010) XDS. Acta Crystallogr D Biol
Crystallogr 66, 125–132.
34 Evans PR & Murshudov GN (2013) How good are my
data and what is the resolution? Acta Crystallogr D
Biol Crystallogr 69, 1204–1214.
35 Winter G, Lobley CMC & Prince SM (2013) Decision
making in xia2. Acta Crystallogr D Biol Crystallogr 69,
1260–1273.
36 Winn MD, Ballard CC, Cowtan KD, Dodson EJ,
Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie
AG, McCoy A et al. (2011) Overview of the CCP4
suite and current developments. Acta Crystallogr D Biol
Crystallogr 67, 235–242.
37 Vagin A & Teplyakov A (2010) Molecular replacement
with MOLREP. Acta Crystallogr D Biol Crystallogr 66,
22–25.
38 Lebedev AA, Vagin AA & Murshudov GN (2008)
Model preparation in MOLREP and examples of
model improvement using X-ray data. Acta Crystallogr
D Biol Crystallogr 64, 33–39.
24 Yin DL, Bernhardt P, Morley KL, Jiang Y, Cheeseman
JD, Purpero V, Schrag JD & Kazlauskas RJ (2010)
Switching catalysis from hydrolysis to perhydrolysis in
P. fluorescens esterase. Biochemistry 49, 1931–1942.
25 Line K, Isupov MN & Littlechild JA (2004) The crystal
structure of a (-) gamma-lactamase from an
Aureobacterium species reveals a tetrahedral
intermediate in the active site. J Mol Biol 338, 519–532.
39 Langer GG, Hazledine S, Wiegels T, Carolan C &
Lamzin VS (2013) Visual automated macromolecular
model building. Acta Crystallogr D Biol Crystallogr 69,
635–641.
ꢁ
40 Murshudov GN, Skubak P, Lebedev AA, Pannu NS,
Steiner RA, Nicholls RA, Winn MD, Long F & Vagin
AA (2011) REFMAC5 for the refinement of
macromolecular crystal structures. Acta Crystallogr D
Biol Crystallogr 67, 355–367.
€
26 Hofmann B, Tolzer S, Pelletier I, Altenbuchner J, van
ꢁ
Pee KH & Hecht HJ (1998) Structural investigation of
41 Thorn A & Sheldrick GM (2013) Extending molecular-
replacement solutions with SHELXE. Acta Crystallogr
D Biol Crystallogr 69, 2251–2256.
the cofactor-free chloroperoxidases. J Mol Biol 279,
889–900.
27 Cheeseman JD, Tocilj A, Park S, Schrag JD &
Kazlauskas RJ (2004) Structure of an aryl esterase from
Pseudomonas fluorescens. Acta Crystallogr D Biol
Crystallogr 60, 1237–1243.
28 Line K (2004) Studies on novel industrial ?-lactamase
enzymes. PhD thesis. Exeter University, UK.
29 Pilalis E, Ladoukakis E, Kolisis F & Chatziioannou A
(2012) A Galaxy workflow for the functional
annotation of metagenomic samples. In: ‘Artificial
Intelligence: Theories and Applications.’ (Maglogiannis
I, Plagianakos V & Vlahavas I, eds), pp. 247–253.
Springer, Berlin Heidelberg, 7297.
42 Emsley P, Lohkamp B, Scott WG & Cowtan K (2010)
Features and development of Coot. Acta Crystallogr D
Biol Crystallogr 66, 486–501.
43 Lebedev AA, Young P, Isupov MN, Moroz OV,
Vagin AA & Murshudov GN (2012) JLigand: a
graphical tool for the CCP4 template-restraint
library. Acta Crystallogr D Biol Crystallogr 68,
431–440.
44 Laskowski RA, MacArthur MW, Moss DS &
Thornton JM (1993) PROCHECK: a program to check
the stereochemical quality of protein structures. J Appl
Cryst 26, 283–291.
30 Ladoukakis E, Pilalis E, Chatziioannou A & Kolisis F
(2014) ANASTASIA a versatile web platform for
metagenomic analysis. New Biotechnol 31, S170.
45 DeLano WL (2002) The PyMOL Molecular Graphics
System. DeLano Scientific, San Carlos, CA, USA.
http://www.pymol.org.
2856
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS

C. Sayer et al.
Structural studies of a thermophilic esterase
46 Vaguine AA, Richelle J & Wodak SJ (1999)
SFCHECK: a unified set of procedures for evaluating
the quality of macromolecular structure-factor data and
their agreement with the atomic model. Acta
drug discovery. J Chem Inf Model 51,
2778–2786.
48 Potterton L, McNicholas S, Krissinel E, Gruber J,
Cowtan K, Emsley P, Murshudov GN, Cohen S,
Perrakis A & Noble M (2004) Developments in the
CCP4 molecular-graphics project. Acta Crystallogr D
Biol Crystallogr 60, 2288–2294.
Crystallogr D Biol Crystallogr 55, 191–205.
47 Laskowski RA & Swindells MB (2011) LigPlot+:
multiple ligand-protein interaction diagrams for
FEBS Journal 282 (2015) 2846–2857 ª 2015 FEBS
2857