Homology modelling and active-site-mutagenesis study of the catalytic domain of the pneumococcal phosphorylcholine esterase

Article abstract of DOI:10.1016/j.bmc.2005.06.060

Streptococcus pneumoniae is among the major human pathogens. Several interactions of this bacterium with its host appear to have been mediated by bacterial cell wall components. Specifically, phosphorylcholine residues covalently attached to teichoic and

Full text of DOI:10.1016/j.bmc.2005.06.060

Bioorganic & Medicinal Chemistry 13 (2005) 6404–6413
Homology modelling and active-site-mutagenesis study of the
catalytic domain of the pneumococcal phosphorylcholine esterase
a,
Nuria Eugenia Campillo,a, , Juan Antonio Paez, Laura Lagarterab and Ana Gonzalezc
*
´
a
´
´
Instituto de Quımica Medica (CSIC), Juan de la Cierva, N 28006-3, Madrid, Spain
´
Serrano 119, Madrid 28006, Spain
b
´ ´
Dpto de Quimica-Fısica de Macromoleculas Biologicas, Instituto de Quımica Fısica Rocasolano (CSIC),
´
´
c
´ ´
Departamento de Microbiologıa Molecular, Centro de Investigaciones Biologicas (CSIC), Ramiro de Maeztu 9, Madrid 28040, Spain
Received 2 February 2005; revised 28 June 2005; accepted 29 June 2005
Available online 19 August 2005
Abstract—Streptococcus pneumoniae is among the major human pathogens. Several interactions of this bacterium with its host
appear to have been mediated by bacterial cell wall components. Specifically, phosphorylcholine residues covalently attached to
teichoic and lipoteichoic acids serve as anchors for many surface-located proteins (choline-binding proteins CBPs), including
cell-adhesion and virulence factors, and are also recognized by host response components through choline-binding receptors. In this
study, we have performed modelling of the catalytic domain of pneumococcal phosphorylcholine esterase (Pce), a modular enzyme
that is capable of removing phosphorycholine residues from teichoic and lipoteichoic acids, remodelling their distribution on the
bacterial envelope. We wish to contribute to the structural knowledge of Pce. In this pursuit, 3D models of Pce have been established
by homology modelling, using the X-ray structure of enzymes from the a/b metallo-lactamase family fold as templates. Theoretical
models of pneumococcal phosphorylcholine esterase (Pce) catalytic modules obtained by homology modelling, and corresponding
docking studies employed to find out the residues involved in the binding of Zn ions, are discussed according to mutational studies
and ab initio calculations. The presence of a binuclear Zn cluster in the catalytic domain of Pce and a likely coordination model are
proposed.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
episodes of invasive pneumococcal infections in humans
are caused by 23 pneumococcal serotypes.5 There is a
limited efficacy of the available vaccines and the use of
antibiotics results in either capsular type shifting or in
the rapid appearance and spreading of antibiotic
resistance determinants.6,7
Streptococcus pneumoniae is among the most frequent
gram-positive pathogens. Pneumococcus can infect
upper respiratory tracts of adults and children, and
spread to blood, lungs, middle ear or even the nervous
system.1 Infections caused by pneumococcus are cur-
rently a leading cause of invasive bacterial disease and
a major cause of morbidity and mortality worldwide,
especially among the elderly and young children,2 and
they are the main cause of bacterial pneumonia, usually
of the lobar type, paranasal sinusitis and otitis media or
meningitis, which is usually secondary to one of the
former infections.3,4 More than 90 different pneumococ-
cal serotypes can be distinguished by their polysaccha-
ride capsules, but approximately 90% of clinical
The cell wall of S. pneumoniae is roughly six layers thick
and is composed of peptidoglycan polymeric chains with
teichoic acid substituents attached to approximately
every third N-acetylmuramic acid.1 Lipoteichoic acid is
chemically identical to teichoic acid but it is attached
to the cell membrane by a lipidic moiety. Both teichoic
and lipoteichoic acids contain phosphorylcholine; two
choline residues may be covalently bound to each penta-
saccharide repeating unit. This is an essential element in
the biology of S. pneumoniae since choline specifically
adheres to choline-binding receptors that are located
on virtually all human cells.8 Besides, they are also rec-
ognized by host response components, such as the hu-
man C-reactive protein (CRP)9 and the receptor of the
platelet-activating factor (PAF).10
Keywords: Pce; Choline-binding protein; Homology modelling; Binu-
clear zinc centre; Ab initio calculations.
*
Corresponding author. Tel.: +34 91 5622900; fax: +34 91 5644853;
e-mail: nuria@suricata.iqm.csic.es
These authors have contributed equally to this work.
0968-0896/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.06.060

N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
6405
On the basis of functional genomic analysis, it is esti-
mated that pneumococcus contains more than 500 sur-
face proteins. Some of them are membrane-associated
lipoproteins, while others are physically associated with
the cell wall. The latter includes five penicillin-binding
proteins11 (PBPs), two neuraminidases12 and an IgA
protease.13 A unique group of proteins on the pneumo-
coccal surface is the family of choline-binding proteins
(CBPs).1,14,15 S. pneumoniae codes for four murein
hydrolases, Lyt A,16 Lyt B,17 Lyt C18 and phosphorych-
oline esterase (Pce).19–21 All CBPs share an N- or C-ter-
minal module that is made up of homologous repeats of
about 20 amino acids, while the additional module/s
accounting for their specific functions are different.
The choline-binding module (CBM) recognizes choline
residues that are present in teichoic acids, allowing an
attachment of the enzyme to the bacterial surface. Lyt
A, Lyt B and Lyt C constitute the enzymatic machinery
that is necessary for cell wall degradation, whereas Pce,
a phosphorylcholine esterase, is involved in remodelling
the distribution of choline residues that regulate the
activity of other autolysins on the bacterial envelope
by hydrolysing the phosphorylcholine residues from tei-
choic and lipoteichoic acid.22
recognition methods were applied using FUGUE,29
a program for recognizing distant homologues by
sequence structure comparison and 3D-PSSM.30
2.1.2. Alignment and adjustment. The most significant
step of a homology model process is to obtain the cor-
rect sequence alignment of the target sequence with
the homologues. Thus, this reliable sequence-structure
mapping has allowed us to identify important structural
features of the residues. The structure-based alignments
for the found homologues were generated with COM-
PARER.31,32 The target sequence Pce was added using
the profile alignment facility of CLUSTALW,33
followed by a manual adjustment of the multiple align-
ment sequence with the SEAVIEW34 program guided by
Pce secondary structure prediction with JPRED.35 The
alignment was annotated with JOY.36
2.1.3. Comparative modelling. We adopted a recursive
approach comprising sequence alignment and model
building. From the best alignment of template structures
to the target sequence, 3D models containing all non-
hydrogen atoms were automatically obtained using the
method implemented in MODELLER (version 4.0).37
The program deduces distance and angle constraints
from a template structure and combines them with ener-
gy terms for adequate stereochemistry in an objective
function that is later optimized in Cartesian space with
conjugate gradients and molecular dynamic (simulated
annealing) methods. Fifteen models of Pce were generat-
ed, but only the one with fewest restraintsÕ violation and
lowest energy was evaluated. Several cycles of realign-
ment and model building were repeated until no further
optimization of the model could be obtained.
Pce comprises a catalytic module located at the N-termi-
nal part of the molecules and a C-terminal choline-bind-
ing module made up of 10 homologous repeating units,
which are similar to sequence repeats present in other
CBPs. It would be of great interest to gain additional
information about Pce and the biological role of choline
during infection by this pathogen as a potential way for
developing new treatments of pneumococcal diseases.
To this end, it is crucial to define the molecular structure
and mechanism of this enzyme. On the basis of sequence
similarity, the catalytic domain of Pce has been included
into the metallo-b-lactamase fold family (PF00753) that
comprises, among other enzymes, metallo-b-lactamases,
glyoxalases, thiolesterases and phosphodiesterases.23–27
The enzymes from this family usually bind two zinc ions
per molecule as cofactor, and the presence of just one or
both metal ions appears to be required for full catalytic
activity.25
The final model was evaluated with PROCHECK38 and
VERIFY3D.39 JOY and COMPARER programs were
used to generate structure-based alignments between
the target and the template in every cycle of the model
building process.
2.1.4. Docking of Zn ions into the active site. In the pres-
ent study, the Zn atoms were placed according to the
atomic coordinates of Zn–O–Zn system of the corre-
sponding template. The Zn-coordinate systems were
moved onto a Pce model using the option Fit Monomer
of Sybyl.40 Sybyl has not been completely parameterized
for Zn ion, the atomic parameters of Zn ion being
assigned according to Sybyl parameters.
To gain insights that will increase our knowledge of Pce,
not only from a structural point of view, but also con-
cerning the role of Zn ions in the catalytic mechanism,
we report herein a discussion of two possible tridimen-
sional models of Pce. In addition, the potential ligands
for the zinc binuclear cluster are proposed.
Conformational search using Genetic Algorithm (GA)
implemented in Sybyl package was performed to study
the binding site of the Pce model. This program uses
Genetic Algorithm, inspired by natural selection in evo-
lution, as a global optimizer to locate the energy minima
of molecules. Judson and Haydon41 among other
authors have shown that GAs provide a good, if not bet-
ter, performance than other methods of conformational
search. They scale well, making them suitable for large
molecules. Several runs were carried out with a popula-
tion size of 100 individuals, a set of constraints being
used. Only the side chains of residues involved in the
2. Materials and methods
2.1. Computational methods
2.1.1. Selection of templates. The identification of homo-
logues of Pce (Swissprot code Q9AJT3) was carried out
by performing sequence database searches with standard
tools, such as PSI-BLAST28, scanning the non-redun-
dant protein sequence database at NCBI with an E-
value cut-off of 10ꢀ4 and retaining only those matches
with the most significant E-values. Besides, protein-fold

6406
N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
active site were considered in this conformational
search. During conformational search, distance re-
straints were also applied to the Zn–O distance. After
examination of the docking model, structures were
selected for further analyses using LPC42 to generate
protein–Zn ion interaction data.
tides for each mutation: primer PceHA50 (50-CCACAG
TGATGCTATTGGAAATGTT-30) and primer PceH
A30 (50-AACATTTCCAATAGCATCACTGTGG-30)
for H90A mutant and primer PceDA50 (50-CAGTTT
GGGGCCATGGATATT-30) and primer PceDA30 (50-
AATATCCATGGCCCCAAACTG-30) for D158A
mutant. All these PCR-amplified fragments were cloned
into plasmid pT7-7 and the DNA sequences of the
mutants were confirmed by the dideoxy chain-termina-
tion method with an automated Abi Prism 3700TM
DNA sequencer (Applied Biosystems). The recombinant
strain cultures were expressed, using 50 lM isopropyl-
thio-b-D-galactopyranoside as inducer, in E. coli BL21
(DE3) strain, and purified as wild-type Pce. All primers
were synthesized on a Beckman model Oligo 1000M
synthesizer. Restriction enzymes and other DNA modi-
fying enzymes were purchased from Amersham.
2.1.5. Theoretical calculations. Ab initio and functional
density theory (DFT) calculations on a simplified model
of the active site have been carried out where Asp and
His residues were replaced by CH3COOꢀ and NH3 mol-
ecules, respectively. This simplified model has been
tested in the active site of a phosphodiesterase and
thus, with this approach, optimized geometric parame-
ter involving zinc ions did not show any significant
changes.43
The ab initio molecular orbital and functional density
theory (DFT) calculations were performed using the
Gaussian 98 package.44 The standard 6-31G* basis set
at the Hartree–Fock level45,46 (HF) and with the density
2.2.1. Activity assays. Pce activity was assayed at 25 °C
in Hepes/Zn2+ buffer using NPPC (p-nitro-
phenylphosphorycholine) as substrate by measuring
the absorbance of the product, p-nitrophenol, at
410 nm.21 Pce activity almost disappeared on extensive
dialysis against Hepes buffer supplemented with 2 mM
EDTA but was restored upon appropriate dilution into
zinc-containing buffers.
functional calculation (DFT), B3LYP functional47,48
,
were used. All structures were fully optimized without
any symmetry restrictions in gas phase, until default
convergence criteria of the program were satisfied.
2.2. Protein purification and site-directed mutagenesis
Pce phosphorylcholine esterase21 was purified, from the
cell extracts of E. coli BL21 (DE3) (pRGR12), by affinity
chromatography on DEAE cellulose equilibrated with
phosphate/Zn buffer (20 mM phosphate, 3 lM ZnCO3,
pH 7.0), following a general procedure described else-
where.49 The protein was specifically eluted using a linear
gradient (0–70 mM choline in 2 h; elution rate 1 ml/min)
in phosphate/Zn buffer with 0.05 M NaCl. Homogeneity
of protein preparations was confirmed by SDS-PAGE.
Before use, protein solutions were extensively dialyzed
3. Results and discussion
3.1. Comparative modelling
Because of its recent isolation, there is little information
about Pce phosphorycholine esterase and its 3D structure
has not yet been elucidated. To understand the binding
characteristics of a Pce protein at the molecular level,
we first conducted homology modelling studies. The
PSI-Blast search for the catalytic domain of Pce did not
give any close homologues as possible templates. There-
fore, protein-fold recognition methods were applied
(Table 1).
against Hepes/Zn buffer (20 mM Hepes, 3 lM Zn2+
;
I = 0.05, pH 7.0) and protein concentration was deter-
mined spectrophotometrically using a molar absorption
coefficient of 194,020 Mꢀ1 cmꢀ1 at 280 nm.
All templates show a metallo-b-lactamase fold and in-
clude three Zn-metallo-b-lactamases,51–53 the human
glyoxalase II54and the oxygen oxydoreductase (ROO)
from Desulfovibrio gigas,55 a Fe-containing enzyme.
The metallo-b-lactamase L1 (code pdb 1sml),51 secreted
by pathogenic Stenotrophomonas maltophilia, and the
human glyoxalase II (code pdb 1qh5)54 provided the
best ranking scores. The active sites of both enzymes
contain a binuclear zinc-binding site, essential for full
enzymatic activity, although they have different modes
of zinc coordination.
Pce H90A and D158A mutants were constructed by
PCR site-directed mutagenesis.50 According to this pro-
tocol, we used oligonucleotides complementary to the
ends of the mature Pce encoding gene primer LytDN2
(50-CCGAATTCAAGGAGATTAACATATGCAAGA
AAGTTCAGGAAATAAAAT-30, where EcoRI restric-
tion site is underlined) and primer LytD-C (50-TTGGA
TCCCTACTACTGTTCTGATTCCGATTTG-30, where
BamHI restriction site is underlined) for the 50 and 30
ends, respectively, and two complementary oligonucleo-
Table 1. Ranking of fold using FUGUE and 3D-PSSM
Fold
Protein activity
From
PDB code
FUGUE score
3D-PSSM
Metallo-hydrolase
Metallo-hydrolase
Flavodoxin-like
Zn metallo-b lactamase
Glyoxalase II
Stenotrophomonas maltophilia
Human
1sml
1qh5
1e5d
1a7t
1jt1
9.99
7.96
5.30
4.20 eꢀ3
3.75 eꢀ4
2.11 e0
3.08 eꢀ2
3.42 eꢀ2
Oxygen oxydoreductase
Zn metallo-b lactamase
Zn metallo-b lactamase
Desulfovibrio gigas
Bacteroides fragilis
Fluoribacter Gormanii
a
Metallo-hydrolase
Metallo-hydrolase
—
a
—
a FUGUE does not recognize it as distant homologue.

N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
6407
b-strands (1–11) (Fig. 3c), shows one more layer in the
sandwich fold.
H
H
H
O
H
H
Zn1
Zn2
Zn1
Zn2
Taking into account the alignment shown in Figure 2,
the residues involved in the coordination of the zinc
would be His 85, His 87, Asp 89, His 90, Asp 158 and
His 228. Residues His 85 and His 87 are located between
b4 and helix B, while Asp 89 and His 90 participate in
the formation of helix B. Of the other two residues,
Asp 158 is located in a loop between b8 and b9, and
His 228 is located just at the beginning of the loop
following b11 (Fig. 3c).
H
O
H/C
H
D
H
H
H
D
b
a
Figure 1. Coordination models of metallo-hydrolases of the metallo-b-
lactamase fold: (a) coordination of metallo-b-lactamases and (b)
coordination of human glyoxalase II.
The three metallo-b-lactamases show two closely spaced
zinc ions that bind six protein residues with tetrahedral
(Zn1) and distorted trigonal bipyramidal (Zn2) coordina-
tion,51 and both ions are bridged by a water molecule
(coordination A, Fig. 1a). In contrast, the two zinc ions
of glyoxalase II,54 coordinated by seven protein residues,
show an octahedral coordination and are double bridged
by an aspartic acid and a water molecule (coordination B,
Fig. 1b).
3.3. Docking of Zn ion into the active site
Docking studies were carried out to investigate the
metallic centre of Pce_1 model. An initial model of the
metallic centre was obtained by transference of Zn–
water–Zn system coordinates from 1sml structure to
Pce_1. Then, a conformational search of the residues in-
volved in the binding of Zn ions was carried out using
GA. The 3D structure of a Zn-binding model for
Pce_1 is shown in Figure 4 and the distances of Zn
residues of the final model for Pce_1 are shown in Table
2. Most of the distances are within the range of experimen-
tal data in the templates, except for Asp 158 and Asp 89.
A comparison of the structural alignment of predicted
folds with the Pce sequence is shown in Figure 2 where
the Zn-binding residues have been identified. There are
some significant insertions but some of the key regions,
such as those containing residues involved in zinc bind-
ing, are well-conserved. Besides, a number of residues
involved in the right orientation of Zn- binding residues
are also conserved. With regard to zinc donors, His 85,
His 87, Asp 89, His 90 and His 228 (numbering from Pce
sequence) are well-conserved among the sequences.
Some structures, such as 1aj7t, show a cysteine instead
of His 90 (numbering from Pce sequence). Asp 158
and Asp 203 (numbering of Pce) are not fully conserved
(Fig. 2), with the latter involved in metal binding usually
associated to coordination B.25
Taking into account the conservation of the Zn-binding
residues among members of the metallo-b-lactamase
fold family (Fig. 2a) and the resulting distances of Zn
ions to the putative metal-binding residues, there are
two doubtful residues: His 90 and Asp 158. Both of
them are not fully conserved among the different en-
zymes and the distance of Asp 158 to Zn1 ion seems
to be rather long. To establish the role of both these
residues clearly, mutagenesis studies were carried out.
3.4. Mutational study
3.2. Model based on 1sml structure (Pce_1)
Enzyme kinetics and activities of the two mutants
(H90A and D158A) of Pce were measured and com-
pared with those of the wild-type enzyme (Table 3).
The D158A mutant did not show a significant change
in the catalytic constant, kcat, although catalytic efficien-
cy (kcat/Km) somewhat worsened. These results suggest
that Asp 158 is not involved in the binding of zinc
atoms. In contrast, mutation H90A decreases the activ-
ity to 12% of wild-type Pce and the catalytic efficiency by
two orders of magnitude, proving that His 90 is actually
involved in zinc coordination.
Considering the best-obtained ranking, a model using
coordination A as pattern of coordination of Zn ions
for Pce sequence was developed. Thus, we chose 1sml
and 1a7t as representative members of Zn-metallob-
lactamases to use them as templates for homology
modelling. Three models of Pce were built based on
found homologues according to the final alignment
(Fig. 2a). Model 1a was based on 1sml structure with
14.5% of identity, model 1b was based on 1a7t struc-
ture with 12.6% of identity and model 1c was built
using both structures as template. The three final mod-
els were analyzed on the basis of the values of the
Ramachandran plot and of Verify 3D. The model 1a
(Pce_1) showed the best validation values. Procheck
summary of model Pce_1 showed that 87.9% of the res-
idues were in the allowed region and only 2.4% in the
disallowed region. The average score obtained with
Verify 3D was 0.7, comparable to the value of 1sml
(0.8). A comparison of the protein topologies is shown
in Figure 3. The topographical diagrams show that sec-
ondary structure elements exhibit an abba fold in all
members of the metallo-b-lactamase family fold, whilst
the Pce_1 model, having six a-helices (A–F) and eleven
3.5. Model based on 1qh5 structure(Pce_2)
Disagreement between mutagenesis studies and theoret-
ical model of coordination of Zn1 ion (Pce_1) could
have arisen either due to a misalignment or/and a wrong
coordination of Zn1 ion. The experimental data avail-
able suggest that both factors are likely contributing to
this disagreement. The model was, therefore, recon-
structed modifying both facts: the type of Zn-coordina-
tion and the alignment. In this sense, the Pce catalytic
module was remodelled using as a representative
template of the type of coordination B (Fig. 1b) the

6408
N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
1a7t (
1sml (
1qh5 (
1e5d (
4 )
2 )
1 )
2 )
svkIsddIsItqLsdkVyTYvSlaeie----gw-
evplpqlraytvdaswlq----pmaplqIAdhTwQIGTe-----------
A
MkVevlpAl-----------
qAtkIidgFhLVGAidwnsrdFhgyt
1jt1 ( 36 )
yp--mPn----pfppfrIAgnLyYVGTd-----------
PCE
-------------------------QESSGNKIHFINVQEGGS-------
1a7t ( 33 )
1sml ( 37 )
1qh5 ( 10 )
1e5d ( 28 )
1jt1 ( 67 )
PCE
---gmvpsNGMIVinn--hqAALLDTPin-------daqTemLvnwVtds
------dLTALLVqTp--dgAVLLDGGM--------pqMAshLldnMka-
-----tdNYMYLVIDdetkeAAIVDPV-----------qpqkVVdaark-
lSpmGTTYNAYLVeDe---kTTLFDTVk--------aeykgeLlcGIas-
------dLASYLIvTp--rgNILINSDl--------eaNVpmIkasIkk-
-DAIILESNGHFAMVDTGEDYDFPDGSDSRYPWREGIETSYKHVLTDRVF
* * **
1a7t ( 71 )
1sml ( 70 )
1qh5 ( 43 )
1e5d ( 66 )
1jt1 ( 102 )
PCE
lhAk------VttFIPNHwhgdCIgGLgyLqrk--gVqSyANqmTidlAk
rgVtpr---dLrlILLSHAHAdHAGPVaeLkrrT-gAkVAANaeSavlLa
hgVk------LtTVLTTHhhwdHAgGNekLvklesgLkVyGGd---d---
v-idpk---kIdyLVIQHLelDHAgALpaLieaCqPekIFTSslGqkaMe
lgfkFs---dTkiLLISHAhfdHAaGSelIkqqT-kAkYmVMdeDvsvIl
RRLKELSVQKLDFILVTHTHSDHIGNVDELLSTYPVDRVYLKKYSDSRIT
1a7t ( 113 )
1sml ( 116 )
1qh5 ( 81 )
1e5d ( 112 )
1jt1 ( 148 )
PCE
ekg--------------------lpvPehGftdsltvsLdgMpLqCyylg
rGGsdDlhfg-------dgitYppanAdriVmdgevItVggivFtAhfM-
----------------------rIgalthkithlstlqVgsLnVkCLaT-
shfhyk------------------dwpVqvVkhgetlsLgkrtVtFyeTr
sGgksDFhya-----ndsstyftqstVdkvLhdgerVelggtvLtAhlT-
NSERLWDNLYGYDKVLQTATETGVSVIQNITQGD-----------AHFQF
*
*
1a7t ( 143 )
1sml ( 158 )
1qh5 ( 108 )
1e5d ( 144 )
1jt1 ( 194 )
PCE
GGhAtDN-------------IVVwLp-------------tenILFGGCML
aGhTpGS-------------TAWTWtDtrng-------kpvrIAYADsLs
pCHTsgH-------------ICYfVskpg-g-------seppAVFTGDTL
mLHwpDS-------------MVSWFa-------------dekVLISNDiF
PGhTrGC-------------TTWTMkLkdhg-------kqyqAVIIGSIg
-GDMDIQLYNYENETDSSGELKKIWDDNSNSLISVVKVNGKKIYLGGDL-
1a7t ( 167 )
1sml ( 188 )
1qh5 ( 137 )
1e5d ( 168 )
1jt1 ( 224 )
PCE
kDn---qttsig----------nis----dA--dvtaWpkTLdkVkakF-
Ap----gY-qLq----------gnp----ryphLiedYrrSFatVraL--
FVA---GCGk-------------fy----eg--tAdeMckALlevLGrlp
GQnIAaserFSdqIpvhtLerAMreYYANiVNpyapqTlkaietlvgag-
vnp---gY-kLv----------dni----tYpkIaedYkhSikvLesm--
DNVH--GA--------------------------EDKYGPLIGKVD----
*
1a7t ( 197 )
1sml ( 217 )
1qh5 ( 165 )
1e5d ( 217 )
1jt1 ( 255 )
PCE
psAryVVPGhg------------------------------------n--
-pÇdvLLTPhPgaS--nwdYaaga-------------rAga------k--
pdT-rVyCGheyTi--nNLkFArhVEpgnaaIrekLawAkekysigep--
VapefICPDHG-VIFrgadqCtfAvqkYveYAeQkpt------nkVVIFY
-rÇdIFLGShAgmF--dLknkyvllskgq-------nnpFv------d--
----LMKFNHHHDTNKSNTKDFIKNLSPSLIVQTSDSLPWKN---GVDSE
1a7t ( 209 )
1sml ( 243 )
1qh5 ( 210 )
1e5d ( 260 )
1jt1 ( 287 )
PCE
ygg--t-eLIeh-TkqiVn--qyiests
alt--ÇkaYAda-AeqkFd--gqlaket---------a--g
TVpStLaeEftYNPFMrVrektVqqhaget----dpvt--TMraVRrekd
DSm---whsTek-MArvLa--esFrdegCtvklmwCkachhSqIMSeIsd
ptg--ÇknyIeq-kAndFy--telkkqe---------t--g
YVNWLKERGIERIN-------AASKDYDATVFDIRKDG-FVNISTSYKPI
1a7t
1sml
1qh5 ( 254 )
1e5d ( 304 )
1jt1
qfkmprd
AGAVIVGSpThnngILpyVagTLqyIkglrPqnKIGGAFGsfgwsGeSTk
PCE
PSFQAGWHKSAYGNWWYQAPDSTGEYAV
alpha helix
red
x
x
x
x
B
beta strand
blue
310 helix
maroon
lower case
solvent accessible
solvent inaccessible
hydrogen bond to main-chain amide
UPPER CASE X
bold
x
x
ç
x
hydrogen bond to main-chain carbonyl underline
disulfide bond
cedilla
positive phi torsion angle
italic
Figure 2. (A) Sequence alignment of predicted folds with Pce sequence generated by JOY. Asterisks ( ) indicate the residues involved in the binding
*
of the zinc atoms in a PCE1 model. (B) Key alignment of JOY.

N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
6409
Figure 3. Topographical diagram of the arrangement of secondary structure elements of (a) metallo-b-lactamases (1sml (I), 1a7t (II) and 1jt1 (III)),
(b) glyoxalase II (1qh5 (IV), C-terminal domain is included), (c) Pce_1 model and (d) Pce_2 model. Triangles represent b-strands and circles represent
a-helices, the N- and C-termini being marked. The first and last helices and b-strands of Pce_1 (c) and Pce_2 (d) are indicated with letter (helices) and
with number (b-strands). The positions of residues involved in the cofactor binding in Pce_1 and Pce_2 models are pointed out with black dots.
structure of human glyoxalase II (1qh5) and the align-
ment of the region comprising Asp 158 was modified
(Fig. 5). The final model, Pce_2, was analyzed using val-
ues of the Ramachandran plot and Verify 3D. Procheck
summary of Pce_2 showed that 86.1% of the residues
were in the allowed region and only 4.0% in the disal-
lowed region. The average score obtained with Verify
3D was 0.6 comparable to 0.7 of 1qh5 structure. The ter-
tiary structure of Pce_2 shows the abba folding motif
characteristic of the metallo-b-lactamase family, with
10 a-helices (A–J) and nine b-strands (1-9) (Fig. 3d).
data of X-ray geometries, in particular the theoretical
distances Asp–Zn are longer, especially the distance of
Asp171–Zn in Pce_2 (Table 2).
Disagreement between the geometrical experimental
data and theoretical model of penta-coordination of
Zn ions in Pce_2 could be due to a wrong coordination
of Zn ion in the model or due to the method of calcu-
lations used. Thus, with the two models (Pce_1 and
Pce_2) on hand we decided to study the stability of
both coordination modes of Zn ions by means of ab ini-
tio and DFT calculations on a simplified model of each
active site of Pce_1 and Pce_2 to clarify the coordina-
tion environment of the zinc ions. The study in gas
phase was performed using a molecular orbital ab initio
method at the Hartree–Fock level with the 6-31G* ba-
sis set and a density functional theory (DFT), B3LYP/
6-31G*. The results from the ab initio and DFT calcu-
lations are similar, indicating that both proposed sim-
plified coordination models correspond to energy
minima and that the geometries of these minimalist
models are fairly in good agreement with the experi-
mental data of X-ray geometries. In fact, both quan-
tum models of the environment of Zn ions (Table 2)
are compatible with the geometries of coordination
models A (Pce_1) and B (Pce_2) obtained with GA
calculations, indicating that in the simplified model
Pce_2, coordination of the residue Asp171 with the
Zn ion is possible.
In accordance with the alignment shown in Figure 5, the
residues involved in the coordination of the Zn atoms
are His 85, His 87, Asp 89, His 90, Asp 171, Asp 203
and His 228. The transference of the Zn–water–Zn sys-
tem coordinates from 1qh5 structure to Pce_2 model
was carried out, followed by a conformational search
of the residues involved in metal binding. The final 3D
model for Pce_2 bimetallic cluster is shown in Figure
6. The distances of the residues to Zn ions of the final
model (Table 2) are within the range of experimental
data in 1qh5, except for Asp 171.
3.6. Ab initio calculation
The results obtained from Conformational Search using
Genetic Algorithm (GA) implemented in Sybyl package
are not completely in agreement with the experimental

6410
N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
Figure 4. Three-dimensional structural model of the bimetallic zinc cluster of Pce_1. Only residues potentially involved in the cofactor binding are
displayed in stick representation. Zinc ions are shown as red spheres, water as a magenta sphere, respectively. Hydrogens are omitted for clarity. The
figure was generated using InsightII program.57
˚
Table 2. Distances (A) of the residues (aa) involved in the coordination of Zn atoms of X-ray structures (1sml and 1qh5) and both models of Pce
(Pce_1 and Pce_2)
Structure
d Zn1–aa1 d Zn1–aa2 d Zn1–aa3 d Zn1–aa4 d Zn2–aa4 d Zn2–aa1 d Zn2–aa2 d Zn2–aa3 d Zn1–O–Zn2
His84
2.03
His 86
2.12
His 160
2.04
Asp 88
2.07
His 89
2.02
His 225
2.07
1SML
—
—
—
—
2.06, 1.87
His85
2.29
2.12
His 87
1.66
2.11
Asp 158
3.30
1.97
Asp 89
3.26
1.99
His 90
2.02
2.13
His 228
1.94
2.12
Pce_1a
Pce_1(HF/6-31G*)b
2.24, 2.10
1.94, 1.96
1.91, 1.95
Pce_1(B3LYP/6-31G*)c 2.07
2.06
1.97
2.07
1.95
2.04
His 54
2.15
His 56
2.21
His 110
2.12
Asp 134
2.57
Asp 134
2.08
His 54
2.09
Asp 59
2.28
His 173
2.13
1QH5
2.15, 2.02
His85
2.32
2.29
His 87
2.32
2.16
Asp 171
4.82
1.98
Asp 203
1.61
2.13
Asp 203
3.24
2.28
Asp 89
3.05
2.03
His 90
2.55
2.13
His 228
1.66
2.12
Pce_2a
2.15, 2.02
1.95, 1.96
1.93, 1.95
Pce_2(HF/6-31G*)b
Pce_2(B3LYP/6-31G*)c 2.24
2.09
1.99
2.11
2.38
2.01
2.05
2.05
a Distances of Pce models from conformational search using GA.
b Distances of simplified models from HF/6-31G* calculation.
c Distances of simplified models from B3LYP/6-31G* calculation.

N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
6411
Table 3. Hydrolytic activity of S. pneumoniae wild-type Pce and the
H90A and D158A mutants on NPPCa
by substitution of His 90, a ligand of Zn2, by alanine
showed a highly reduced activity supports its implication
in Zn coordination and the existence of two Zn atoms in
the catalytic centre. However, the Pce_1 model fails in the
prediction of Asp 158 participating in the binding of Zn1.
Km (mM)
kcat (sꢀ1
)
kcat/Km (mMꢀ1 sꢀ1
58
0.7 0.1
15
)
Wild type
H90A
D158A
0.16 0.02
1.7 0.4
0.47 0.06
9.3 0.2
1.13 0.07
7.1 0.2
6
2
The second model (Pce_2) obtained by homology mod-
elling from 1qh5 shows a binuclear Zn-binding site,
according to the mutational study. Both zinc ions show
a penta-coordination and they would be bridged, as in
human glyoxalase II, Desulfovibrio gigas oxygen oxido-
reductase and other metallo-hydrolases containing binu-
clear zinc cocatalytic centres,56 by a water molecule and
Asp 203. The Pce_2 model shows a fold similar to those
of metallo-b-lactamase family members and the spatial
distribution of residues involved in the binding of zinc
ions is consistent with the experimental data.
a Measurements were made in Hepes/Zn buffer at 25 °C (I = 0.05).
4. Conclusions
To summarize, two spatial models of the catalytic module
of pneumococcal phosphorylcholine esterase (Pce) are de-
scribed. These models are based on templates belonging
to hydrolases of the metallo-b-lactamase fold family
(1sml, 1a7t, 1jt1 and 1qh5) displaying two different coor-
dination modes of the zinc ions acting as cofactor. The
drastic drop in Pce activity observed upon dialysis against
EDTA-containing buffer and its recovery upon addition
of Zn salts (see Materials and methods) show that Pce is
a zinc-dependent phosphorylcholine esterase. Docking
studies with zinc ions were employed to find out the resi-
dues involved in coordination of the metallic cofactor and
to understand the role of particular residues in metal
binding.
The relevance of His 90 in Pce activity suggests that both
metal ions are directly involved in the catalytic process
and support the presence of a binuclear zinc cocatalytic
centre in the active site of Pce. Double bridging of zinc
atoms by a solvent molecule and an aspartic or glutamic
acid is a characteristic feature of cocatalytic centres
in zinc metallo-hydrolases61 and would be consistent
with the coordination type proposed in a Pce_2 model.
In contrast, the coordination in a Pce_1 model seems
to be found in systems, such as b-lactamases, whose
The first model (Pce_1) was built by homology modelling
of the 1sml structure. The fact that the mutant generated
Figure 5. JOY output of the structural superposition of the template 1qh5 and Pce_2, model.

6412
N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
Figure 6. Three-dimensional structural model of the bimetallic zinc cluster of Pce_2. Only residues potentially involved in the cofactor binding are
displayed in stick representation. Zinc ions are shown as blue spheres, water as a red sphere, respectively. Hydrogens are omitted for clarity. The
figure was generated using InsightII program.57
9. Hirschfield, G. M.; Pepys, M. B. QJM 2003, 96, 793–
807.
10. Cundell, D. R.; Weiser, J. N.; Shen, J.; Young, A.;
Tuomanen, E. I. Infect. Immun. 1995, 63, 757–761.
11. Popham, D. L.; Young, K. D. Curr. Opin. Microbiol.
2003, 6, 594–599.
12. Achyuthan, K. E.; Achyuthan, A. M. Comp. Biochem.
Physiol. B Biochem. Mol. Biol. 2001, 129, 29–64.
13. Parsons, H. K.; Vitovski, S.; Sayers, J. R. Biochem. Soc.
Trans. 2004, 32, 1130–1132.
activity depends on the presence of a first zinc ion (Zn1),
while binding of a second metal equivalent (Zn2) may
have different effects on activity.23 Since the pneumococ-
cal phosphorylcholine esterase is an interesting thera-
peutic target, these studies contribute to imparting
knowledge of a Pce catalytic mechanism.
References and notes
´
´
´
´
14. Lopez, R.; Garcıa, E.; Garcıa, P.; Garcıa, J. L. Microb.
´ ´
1. Lopez, R.; Garcıa, E. FEMS Microbiol. Rev. 2004, 28,
Drug Resist. 1997, 3, 199–211.
553–580.
2. Binder, S.; Levitt, A. M.; Sacks, J. J.; Hughes, J. M.
Science 1999, 284, 1311–1313.
15. Rosenow, C.; Ryan, P.; Weiser, J. N.; Johnson, S.;
Fontan, P.; Ortqvist, A.; Masure, H. R. Mol. Microbiol.
1997, 25, 819–829.
´
´
´ ´
16. Garcıa, E.; Garcıa, J. L.; Ronda, C.; Garcıa, P.; Lopez, R.
´
3. Hirst, R. A.; Kadioglu, A.; Ocallaghan, C.; Andrew, P. W.
Clin. Exp. Immunol. 2004, 138, 195–201.
4. Casal, J.; Tarrago, D. Curr. Opin. Infect. Dis. 2003, 16,
219–224.
5. Neralla, S.; Meyer, K. C. Drugs Aging 2004, 21, 851–864.
6. Dagan, R. Curr. Opin. Otolaryngol. Head Neck Surg.
2004, 12, 488–494.
Mol. Gen. Genet. 1985, 201, 225–230.
´
Garcıa, J. L. Mol. Microbiol. 1999, 31, 1275–1277.
´ ´ ´ ´
18. Garcıa, P.; Gonzalez, M. P.; Garcıa, E.; Lopez, R.;
´
´
´
17. Garcıa, P.; Gonzalez, M. P.; Garcıa, E.; Lopez, R.;
´
´
Garcıa, J. L. Mol. Microbiol. 1999, 33, 128–138.
´
´
´
19. Garcıa, J. L.; Dıaz, E.; Romero, A.; Garcıa, P. J.
Bacteriol. 1994, 176, 4066–4072.
20. Tomasz, A.; Moreillon, P.; Pozzi, G. J. Bacteriol. 1988,
170, 5931–5934.
7. McCormick, A. W.; Withney, C. G.; Farley, M. M.;
Lynfield, R.; Harrison, L. H.; Bennett, N. M.; Schaffner,
W.; Reingold, A.; Hadler, J.; Cieslak, P.; Samore, M. H.;
Lipsitch, M. Nat. Med. 2003, 9, 424–430.
8. Bozdogan, B.; Bogdanovich, T.; Kosowska, K.; Jacobs,
M. R.; Appelbaum, P. C. Curr. Drug Targets Infect.
Disord. 2004, 4, 169–176.
´
´
´
21. De las Rivas, B.; Garcıa, J. L.; Lopez, R.; Garcıa, P.
Microb. Drug Resist. 2001, 7, 213–222.
22. Vollmer, W.; Tomasz, A. Mol. Microbiol. 2001, 39, 1610–
1622.

N. E. Campillo et al. / Bioorg. Med. Chem. 13 (2005) 6404–6413
6413
23. Helfand, M. S.; Bonomo, R. A. Curr. Drug Targets Infect.
Disord. 2003, 3, 9–23.
24. Bonfiglio, G.; Russo, G.; Nicoletti, G. Expert Opin.
Investig. Drugs 2002, 11, 529–544.
25. Rasia, R. M.; CeolÕin, M.; Vila, A. J. Protein Sci. 2003, 12,
1538–1546.
26. Bush, K. Clin. Infect. Dis. 2001, 32, 1085–1089.
27. Daiyasu, H.; Osaka, K.; Ishino, Y.; Toh, H. FEBS Lett.
2001, 503, 1–6.
28. Altschul, S. F.; Madden, T. L.; Schaffer, A. A.; Zhang, J.;
Zhang, Z.; Miller, W.; Lipman, D. J. Nucleic Acids Res.
1997, 1, 3389–3402.
29. Shi, J.; Blundel, T. L.; Mizuguchi, K. J. Mol. Biol. 2001,
310, 243–257.
Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G.A.; Ayala, P. Y.;
Cui, Q.; Morokuma, K.; Salvador, P.; Dannenberg, J. J.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A.
G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R.L.; Fox, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian, Inc.,
GAUSSIAN 98, Pittsburgh, PA, 2001.
30. Kelley, L. A.; MacCallum, R. M.; Sternberg, M. E. J.
Mol. Biol. 2000, 299, 499–520.
45. Cramer, C. J.; Truhlar, D. G. Chem. Rev. 1999, 99, 2161–
2200.
31. Sali, A.; Blundell, T. J. Mol. Biol. 1990, 20, 403–428.
32. Zhu, Z. Y.; Sali, A.; Blundell, T. Protein Eng. 1992, 5, 43–
51.
46. Francl, M. M.; Pietro, J.; Hehre, W. J.; Binkley, J. S.;
Gordon, M. S.; De Frees, D. J.; Pople, J. A. J. Chem.
Phys. 1982, 77, 3654–3665.
33. Thompson, J. D.; Gibson, T. J.; Plewniak, F.; Jeanmou-
gin, F.; Higgins, D. G. Nucleic Acids Res. 1997, 25, 4876–
4882.
47. Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
48. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B. 1988, 37,
785–789.
34. Galtier, N.; Gouy, M.; Gautier, C. Comput. Appl. Biosci.
1996, 12, 543–548.
49. Sambrook, J.; Russell, D. W. Molecular Cloning: A
Laboratory Manual, 3rd ed.; Cold Spring Harbor Labo-
ratory Press: New York, 2001.
35. Cuff, J. A.; Clamp, M. E.; Siddiqui, A. S.; Finlay, M.;
Barton, G. J. Bioinformatics 1998, 14, 892–893.
36. Mizuguchi, K.; Deane, C.; Blundell, T.; Johnson, M. S.;
Overington, J. P. J. Bioinformatic 1998, 14, 617–623.
37. Sali, A.; Blundell, T. J. Mol. Biol. 1993, 234, 779–815.
38. Luthy, R.; Bowie, J. U.; Eisenberg, D. Nature 1992, 5,
583–585.
39. Laskowski, R. A.; McArthur, W.; Moss, D.; Thornton, J.
M. J. Appl. Crystallogr. 1993, 26, 283–291.
40. SYBYL 6.9, Tripos Inc., 1699 South Hanley Rd., St.
Louis, MO 63144, USA.
´
´
50. Sanchez-Puelles, J. M.; Sanz, J. M.; Garcıa, J. L. Gene
1990, 89, 69–75.
51. Ullah, J. H.; Walsh, T. R.; Taylor, I. A.; Emery, D. C.;
Verma, C. S.; Gamblin, S. J.; Spencer, J. J. Mol. Biol.
1998, 284, 125–136.
52. Frazao, C.; Silva, G.; Gomes, C. M.; Matias, P.; Coelho,
R.; Sieker, L.; Macedo, S.; Liu, M. Y.; Oliveira, S.;
Teixeira, M.; Xavier, AV.; Rodrigues-Pousada, C.; Carr-
ondo, M. A.; Le Gall, J. Nat. Struct. Biol. 2000, 7, 1041–
1045.
41. Judson, O. P.; Haydon, D. J. Mol. Evol. 1999, 49, 539–
550.
42. Sobolev, V.; Sorokine, A.; Prilusky, J. E.; Abola, E.;
Edelman, M. Bioinformatics 1999, 15, 327–332.
43. Koca, J. V.; Zhan, C. G.; Rittenhouse, R. C.; Ornstein, R.
L. J. Comput. Chem. 2003, 24, 368–378.
44. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.
E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A.; Stratmann, R. E.; Burant, J. C.;
53. GarcÕia-Saez, I.; Mercuri, P. S.; Papamicael, C.; Kahn, R.;
Frere, J. M.; Galleni, M.; Rossolini, G. M.; Dideberg, O.
J. Mol. Biol. 2003, 325, 560–651.
54. Cameron, A. D.; Ridderstrom, M.; Olin, B.; Mannervik,
B. Struct. Fold. Des. 1999, 7, 1067–1078.
55. Fitzgerald, P. M.; Wu, J. K.; Toney, J. H. Biochemistry
1998, 37, 6791–6800.
56. Auld, D. S. BioMetals 2001, 14, 271–313.
57. InsightII. Biosym, San Diego, CA, USA.