Content uploaded by Pedro García
Author content
All content in this area was uploaded by Pedro García on Jun 19, 2017
Content may be subject to copyright.
Content uploaded by Roberto Díez-Martínez
Author content
All content in this area was uploaded by Roberto Díez-Martínez on Mar 02, 2015
Content may be subject to copyright.
A novel chimeric phage lysin with high in vitro and in vivo bactericidal
activity against Streptococcus pneumoniae
Roberto Dı
´ez-Martı
´nez1–3,He
´ctor D. De Paz1†, Esther Garcı
´a-Ferna
´ndez1,3, Noemı
´Bustamante3,4,
Chad W. Euler2,5, Vincent A. Fischetti2, Margarita Menendez3,4‡ and Pedro Garcı
´a1,3*‡
1
Departamento de Microbiologı
´a Molecular y Biologı
´a de las Infecciones, Centro de Investigaciones Biolo
´gicas, CSIC, Madrid, Spain;
2
Laboratory of Bacterial Pathogenesis and Immunology, The Rockefeller University, New York, NY, USA;
3
CIBER de Enfermedades
Respiratorias, Madrid, Spain;
4
Departamento de Quı
´mica-Fı
´sica Biolo
´gica, Instituto Quı
´mica-Fı
´sica Rocasolano, CSIC, Madrid, Spain;
5
Department of Medical Laboratory Sciences, Hunter College, CUNY, New York, NY, USA
*Corresponding author. Tel: +34-91-8373112; Fax: +34-91-5360432; E-mail: pgarcia@cib.csic.es
†Present address: Infectious Diseases Research Unit, Department of Molecular Microbiology, Hospital Sant Joan de Deu, Barcelona, Spain.
‡M. M. and P. G. contributed equally to this article.
Received 31 October 2014; returned 2 December 2014; revised 9 December 2014; accepted 29 January 2015
Objectives: Streptococcus pneumoniae is becoming increasingly antibiotic resistant worldwide and new antimi-
crobials are urgently needed. Our aim was new chimeric phage endolysins, or lysins, with improved bactericidal
activity by swapping the structural components of two pneumococcal phage lysozymes: Cpl-1 (the best lysin
tested to date) and Cpl-7S.
Methods: The bactericidal effects of four new chimeric lysins were checked against several bacteria. The purified
enzymes were added at different concentrations to resuspended bacteria and viable cells were measured after
1 h. Killing capacity of the most active lysin, Cpl-711, was tested in a mouse bacteraemia model, following mouse
survival after injecting different amounts (25–500 mg) of enzyme. The capacity of Cpl-711 to reduce pneumo-
coccal biofilm formation was also studied.
Results: The chimera Cpl-711 substantially improved the killing activity of the parental phage lysozymes, Cpl-1
and Cpl-7S, against pneumococcal bacteria, including multiresistant strains. Specifically, 5 mg/mL Cpl-711 killed
≥7.5 log of pneumococcal R6 strain. Cpl-711 also reduced pneumococcal biofilm formation and killed 4 log of the
bacterial population at 1 mg/mL. Mice challenged intraperitoneally with D39_IU pneumococcal strain were pro-
tected by treatment with a single intraperitoneal injection of Cpl-711 1 h later, resulting in about 50% greater
protection than with Cpl-1.
Conclusions: Domain swapping among phage lysins allows the construction of new chimeric enzymes with high
bactericidal activity and a different substrate range. Cpl-711, the most powerful endolysin against pneumococci,
offers a promising therapeutic perspective for the treatment of multiresistant pneumococcal infections.
Keywords: S. pneumoniae, antimicrobial therapy, bacteriophages, bacterial biofilm, animal infections
Introduction
Streptococcus pneumoniae, the pneumococcus, is a Gram-positive
bacterium usually found in asymptomatic nasopharyngeal car-
riage, but is also the leading cause of many infections worldwide,
ranging from very common and usually mild clinical conditions,
such as acute otitis media and rhinosinusitis, to less frequent but
potentially life-threatening invasive diseases, such as sepsis,
community-acquired pneumonia and meningitis.
1
Classical treat-
ment of these infections has been the use of certain well-tolerated
drugs, especially b-lactam antibiotics, which significantly changed
the outcome of pneumococcal diseases and substantially reduced
their clinical and economic impact. Unfortunately, in recent years
the lack of new antibiotics and the emergence of MDR bacteria
haveledtoahealthcarecrisisfrequentlyassociatedwiththe
so-called ‘post-antibiotic era’. This situation has provoked the US
CDC to call for an aggressive and immediate action to halt the
spread of drug-resistant pathogens.
2
In this context, phage therapy
is becoming one of the most promising alternative weapons to fight
dangerous multiresistant pathogens, with the use of not only iso-
lated virions but also some of their products, like endolysins. These
specialized enzymes are peptidoglycan hydrolases synthesized as
#The Author 2015. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com
J Antimicrob Chemother 2015; 70: 1763– 1773
doi:10.1093/jac/dkv038 Advance Access publication 1 March 2015
1763
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
part of the tightly controlled late genes that provoke the lysis of the
bacterial cells for phage progeny release. Endolysins have been
shown to be very effective in killing a variety of Gram-positive bac-
teria when added exogenously as purified proteins.
3
Although the
typical architecture consists of a two-domain structure, in which
the N-terminal part is responsible for the catalytic activity (catalytic
domain or CD) and the C-terminal part functions as a cell wall-
binding domain (CWBD),
4
some peptidoglycan hydrolases possess
multiple hydrolytic domains.
5,6
Cpl-1
7
and Cpl-7
8
lysozymes are encoded by the virulent Cp-1
and Cp-7 pneumococcal bacteriophages, respectively, belonging
to the Podoviridae family.
9
These enzymes are very similar in
their CDs but differ completely in their CWBDs. The N-terminal
CD of both enzymes belongs to the GH-25 family of glycosylhydro-
lases sharing 85.6% sequence identity (95.2% similarity). This
domain folds into an irregular (b/a)
5
b
3
barrel, characteristic of
lysozymes of the same family.
7,10 –12
In its CWBD, Cpl-1 displays
the typical choline-binding repeats that allow anchoring to the
phosphocholine residues of (lipo)teichoic acids,
7
whereas Cpl-7
contains three perfect repeats forming three tandemly arranged
helical bundles that recognize the cell wall in a choline-
independent manner.
8,13
Linkers of both endolysins are different
in amino acid sequence and length, suggesting distinct interac-
tions between their respective modules, but both are highly
hydrophilic and share a highly charged N-terminal motif.
Recently, it has been demonstrated that a synthetic variant of
Cpl-7, named Cpl-7S, was capable of improving the bactericidal
effect of Cpl-7 not only against pneumococci but also against
other important pathogens like Streptococcus pyogenes and
Enterococcus faecalis.
14
In this work, we have constructed chimeric proteins by shuffling
the structural elements, i.e. the two functional domains and the
linkers, from Cpl-1 and Cpl-7S endolysins, giving rise to four new
enzymes. One of them, Cpl-711, turned out to be clearly more
active than Cpl-1 in killing several pneumococcal strains grown
in either planktonic or sessile (biofilm) conditions. The increased
lethal action of Cpl-711 has been validated in a mouse model
of bacteraemia, confirming that this new chimeric lysin is the
most potent enzyme against pneumococci described so far.
Materials and methods
Ethics statement
All experiments were performed under strict adherence to the NIH guide-
lines for the ethical treatment of animals. Signs of infection were moni-
tored three times daily throughout the experimental time course.
Moribund mice were euthanized by cervical dislocation under isoflurane
anaesthesia. Approval for these studies was granted by the Rockefeller
University Institutional Animal Care and Use Committee (Protocol:
11005). Experiments were carried out at the Rockefeller University’s ani-
mal housing facility, an AAALAC-accredited research facility, with all
efforts made to minimize animal suffering.
Bacterial strains, media and growth conditions
The bacterial strains used in this study are listed in Table S1 (available as
Supplementary data at JAC Online). Pneumococcal strains were grown at
378C in C medium supplemented with yeast extract (0.8 mg/mL; Difco
Laboratories) (C+Y)
25
or on tryptic soy agar plates containing 5% defibri-
nated sheep blood.
26
The other Gram-positive bacteria were grown in
Todd–Hewitt broth containing 0.5% yeast extract (THY) at 378C, without
shaking. Escherichia coli strains, used for gene cloning and producing
recombinant proteins, were grown in LB medium with shaking at 378C.
Synthesis and cloning of chimeric genes
The plasmids and oligonucleotides used in this work are listed in Table S1.
The synthetic DNA fragments encoding Cpl-117, Cpl-177 and Cpl-711 were
purchased from ATG:Biosynthesis (Merzhausen, Germany) as E. coli
codon-optimized pUC derivatives. Codon optimization was performed
using the web server Optimizer (http://genomes.urv.es/OPTIMIZER). The
NdeI+PstI-digested fragments carrying the chimeric genes were sub-
cloned into the pT7-7 expression vector, previously treated with the
same restriction enzymes, and the resulting plasmids pTRD760,
pTRD761 and pTRD762 were transformed into E. coli DH10B. cpl-771
was constructed by a two-step process. The first required the incorporation
of a conservative mutation in cpl-7S at the end of the linker using a modi-
fied PCR-driven overlap extension technique.
27
This mutation consists of
the addition of a HindIII restriction site by two independent PCRs using
A+B and C+D primers, respectively, and pTRD750 as template. Primers
B and C are partially complementary and contain the HindIII restriction
site. The resulting PCR products and the primers A and D were used for a
third PCR to amplify cpl-7S that included the desired mutation. Parameters
for the two first independent PCRs were: 948C for 30 s (denaturation), 658C
for 30 s (annealing) and 728C for 2 min(extension). The parameters for the
third PCR were: 948C for 30 s (denaturation), 658C for 30 s (annealing) and
728C for 4 min (extension). Then the modified cpl-7S was digested with
NdeI and ClaI and cloned into pT7-7 treated with the same enzymes, ren-
dering pHP200. The second step consisted of the fusion of the part of the
cpl-1 encoding CWBD of Cpl-1 (CWBD
1
) to that coding for the CD and linker
region of Cpl-7S. To accomplish this, the portion of cpl-1 encoding CWBD
1
was amplified from pCIP100 using E+F primers containing HindIII and
ClaI restriction sites. PCR parameters used were: 948C for 30 s (denatur-
ation), 658C for 30 s (annealing) and 728C for 2 min (extension). The
HindIII +ClaI-digested PCR product was cloned into pHP200, previously
digested with the same enzymes, resulting in pHP771. The composition
of chimeric genes was confirmed by DNA sequencing (Secugen,
Madrid, Spain).
Production and purification of chimeric proteins
For overproduction of chimeric enzymes, transformed E. coli BL21(DE3)
cells with pTRD760, pTRD761, pTRD762 or pHP771 were incubated in LB
medium containing ampicillin (0.1 mg/mL) to an OD
600
of 0.6. IPTG
(0.1 mM) was then added, and the incubation was continued overnight
at 308C. Cells were harvested by centrifugation (10 000 g, 5 min), resus-
pended in 20 mM sodium phosphate buffer (pH 6.0) (Pi buffer), disrupted
in a French pressure cell and centrifuged (50000 g,45 min) to remove cell
debris. Streptomycin sulfate (Sigma; 2%, w/v) was added to the protein
extract, incubating for 15 min at 48C with slow stirring to facilitate DNA
precipitation. The insoluble fraction was removed by ultracentrifugation
(50000 g, 45 min) at 48C. Cpl-711 and Cpl-771 proteins were purified by
affinity chromatography using DEAE-cellulose, as previously described.
24
Cpl-117 and Cpl-177 were purified following the procedure reported for
Cpl-7.
8
The purity of the isolated proteins was checked using 10% SDS–
PAGE and MS (MALDI-TOF). Protein concentrations were determined spec-
trophotometrically using the corresponding molar absorption coefficients
at 280 nm. Before use, all proteins were equilibratedby dialysis in Pi buffer.
In vitro cell wall activity assay
Purified enzymes were checked in vitro for cell wall degradation using
[methyl-
3
H]choline-labelled pneumococcalcellwallsasthesubstrate
and following a previously described method.
28
Briefly, 10 mL of enzyme
Dı
´ez-Martı
´nez et al.
1764
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
at the appropriate dilution was added to the reaction sample containing
240 mL of Pi buffer and 10 mL of radioactively labelled cell walls
(35000 cpm), as described previously.
29
After 15 min of incubation at
378C, the reaction was stopped by adding 10 mL of formaldehyde (37%,
v/v) and 10 mL of BSA (4%, w/v). The pellet was removed by centrifugation
(12000 g, 15 min), and the enzymatic activity was quantified by measur-
ing the radioactivity in 200 mL of the supernatant with a liquid scintillation
counter (LKB Wallac). One unit of enzymatic activity (U) was defined as the
amount of enzyme that catalyses the hydrolysis (solubilization) of 1 mgof
cell wall material in 10 min.
28
MIC determinations
MICs of Cpl-711, Cpl-771, Cpl-117 and Cpl-177 were determined by the
microdilution method approved by the CLSI (2006) using CAMHB supple-
mented with 5% lysed horse blood. Modal values from three separate
determinations were considered. Pneumococcal strain ATCC 49619 was
used as a quality control strain for susceptibility testing (http://www.
lgcstandards-atcc.org/Products/All/49619.aspx).
Bacteriolytic and bactericidal assays
Bacteria were grown to logarithmic phase to an OD
550
of about 0.3, and
then cultures were centrifuged and washed twice with PBS [137 mM
NaCl, 2.7 mM KCl, 10 mM Na
2
HPO
4
and 1.8 mM KH
2
PO
4
(pH 6.0)], and
the final OD
550
was adjusted to 0.6 in PBS, using a Helios Epsilon spectro-
photometer (Thermo Scientific). Afterwards, resuspended bacteria were
transferred into plastic tubes, and the tested enzyme was added.
Samples were incubated at 378C for 1 h, and the turbidity decrease
(OD
550
) was measured at selected intervals. Controls replacing the
added enzyme with Pi buffer were always run in parallel. Measurement
of viable cells was carried out in C +Y or blood agar plates. For each sam-
ple, a 10-fold dilution series was prepared in PBS, and 10 mL of each dilu-
tion was plated. Colonies were counted after overnight incubation at 378C.
Biofilm assays
Pneumococcal biofilms were produced using a P046 strain that is deficient
in LytA and LytC autolysins.
18
Biofilms were grown in Costar 3595 96-well
PST microtitre plates (Corning, New York, USA).
18,30
Briefly, cells were grown
in C+YmediumtoanOD
595
of 0.5–0.6, sedimented by centrifugation,
resuspended in C+Y medium and aliquots of 200 mLcontaining
4.25×10
6
cfu were dispensed into each well in two microtitre plates.
After 16 h of incubation at 348C, the biofilm formed was washed with
PBS and treated for 2 h at 378C with either Pi buffer or different concentra-
tions of enzymes in Pi buffer. Finally, one of the plates was stained with a
0.2% (w/v) crystal violet solution for 15 min and then washed three times
with distilled water to remove non-adherent bacteria. After solubilizing the
biofilm in 95% ethanol (200 mL per well), the OD
595
was determined using
a VersaMax microplate absorbance reader (Molecular Devices). The other
plate was used to determine viability. To do this, a 10-fold dilution series
was prepared in PBS and 10 mL of each dilution was plated in blood agar
plates. Colonies were counted after overnight incubation at 378C. For
observation by confocal laser scanning microscopy (CLSM), biofilms were
grown at 348C for 16 h on glass-bottomed WillCo-dish dishes (WillCo
Wells). Then, they were washed with PBS and treated for 2 h at 378C
with either Pi buffer or different amounts of the corresponding enzymes.
Biofilms were then washed three times with PBS to remove cell debris, and
stainedwiththeLIVE/DEADBacLight bacterial viability kit L-13152
(Invitrogen-Molecular Probes) for monitoring the viability of bacterial
populations. Cells with a compromised membrane—considered dead or
dying—stained red, whereas those with an intact membrane stained
green.
18
The biofilms were observed at ×63 magnification using a Leica
TCS-SP5-AOBS-UVCLSMequippedwithanargonionlaser.The
excitation/emission maxima were around 488/500 to 561 nm. Images
were acquired using LCS software (Leica). Projections were obtained
through the x–yplane (individual scans at 0.5 mm intervals) and x–z
plane (images at 5 mm intervals).
Murine bacteraemia model
The infection model was based on methods described previously,
31,32
using 4–6-week-old female BALB/c mice (weight range, 15–20 g)
obtained from Charles River Laboratories (Wilmington, MA, USA). Briefly,
after a period of acclimatization, mice were injected intraperitoneally (ip)
with 0.5 mL of a 5.5×10
5
cfu/mL suspension of S. pneumoniae strain
D39_IU cells. Bacterial inoculation titres were calculated by serial dilution
and plating onto Columbia blood agar plates for each experiment. To con-
firm the bacteraemic state of mice at the time of treatment (i.e. 1 h after
bacterial challenge), multiple organs, including spleen, liver, kidney and
heart/blood, were removed aseptically from three mice, homogenized
and plated for bacterial counts. One hour post-infection, the remaining
animals were divided into four to five treatment groups per lysin and
were injected ip with 0.5 mL containing 500, 200, 50 or 25 mg of either
Cpl-711 or Cpl-1 lysins, or PBS buffer as control. The survival rate for
each experimental group was monitored every 12 h for the first 24 h
and then every 24 h for up to 7 days post-infection. The results from
four independent experiments were combined to evaluate a total of 20
mice for controls and 17 mice for each lysin-treated group.
MS
Purified samples of chimeric proteins were analysed by MALDI-TOF, as
described elsewhere.
33
A grid voltage of 93%, a 0.1 ion guide wire voltage,
and a delay time of 350 ns in the linear positive ion mode were used.
External calibration was performed with carbonic anhydrase (29 024 Da)
and enolase (46672 Da) from Sigma, covering an m/z range of 16 000–
50000 units.
Circular dichroism
Circular dichroism spectra were recorded at 208C in a J-810 spectropolari-
meter (Jasco Corporation) equipped with a Peltier cell holder.
Measurementswereperformedin1mmand1cmpath-lengthcells
(far- and near-circular dichroism spectra, respectively) using the experi-
mental conditions previously described.
13
Buffer contribution was sub-
tracted from protein spectra and corrected data were converted into
mean residue ellipticities using the corresponding average molecular
mass per residue. Choline titration curves were obtained by measuring
the increase in ellipticity at 225 nm as a function of the ligand concentra-
tion. The dissociation constants were calculated by fitting the Hill equation
to the experimental curves.
34
Analytical ultracentrifugation
Sedimentation velocity experiments were run at 45000 rpm using cells
with double sector Epon-charcoal centrepieces. Differential sedimentation
coefficients were calculated by least-squares boundary modelling of the
experimental data with the program SEDFIT and corrected to s
20,w
values.
35
Sedimentation equilibrium experiments were carried out at dif-
ferent rotor speeds, from 10000 to 17000 rpm, as described previously,
34
and weight-average molecular weights were calculated using the
Heteroanalysis program, version 1.1.2 (http://www.biotech.uconn.edu/
auf/). All measurements were performed in Pi buffer at 208C using protein
concentrations of 0.2 mg/mL in an Optima XL-A analytical ultracentri-
fuge (Beckman Coulter) with an AN50-Ti rotor. The ratio of the frictional
translational coefficients of the protein particle and the equivalent rigid
sphere sedimenting like the protein (f/f
0
) and Stockes’s radio (R
s
), related
Chimeric lysins against pneumococcus
1765
JA
C
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
to the protein hydrodynamic shape, were calculated from the protein
molecular masses and s
20,w
values using the partial specific volumes
and hydration coefficients estimated from the amino acid sequence
with the SEDNTERP program.
36
Modelling of the Cpl-711 structure
The three-dimensional model of the Cpl-711 chimera was built automat-
ically using the Cpl-1 structure as a template (PDB access code; 1H09) with
the Swiss-Model server in the alignment mode (http://swissmodel.expasy.
org). Model quality was stereochemically and energetically validated with
Procheck
37
and VERIFY 3D,
38
respectively, comparing well with Cpl-1.
StructurefiguresweremadewithPYMOL(Schro
¨dinger LLC, Cambridge
MA, USA).
Statistical analysis
All in vitro results are representative of data obtained from repeated inde-
pendent experiments, and eachvalue represents the mean+SD for three or
four replicates. Statistical analysis was performed using the two-tailed
Student t-test (for two groups), whereas analysis of variance (ANOVA)
was chosen for multiple comparisons. For all in vivo data, the log-rank
(Mantel– Cox) and Gehan– Breslow– Wilcoxon tests were used to draw, ana-
lyse and compare the survival curves. GraphPad InStat version 5.0
(GraphPad Software, San Diego, CA, USA) was used for statistical analysis.
Results
Design of chimeric proteins
With the aim of creating endolysins with improved bactericidal
activity against pneumococcal and non-pneumococcal species
when added exogenously to bacterial cultures, we decided to
shuffle and combine the three structural elements (CD, linker
and CWBD) of Cpl-1 and Cpl-7S lysozymes. Cpl-1 is strictly
dependent on the presence of choline residues in teichoic acid
to fully hydrolyse the bacterial cell wall, whereas Cpl-7S is an
engineered variant of the WT Cpl-7 (the exact construct is
described below), which degrades susceptible cell walls in a cho-
line-independent manner. Cpl-7S has an improved antimicrobial
activity compared with Cpl-7, due to the inversion of the net
charge of the CWBD.
14
Figure 1shows a schematic representa-
tion of the four synthetic chimeric proteins (Cpl-711, Cpl-771,
Cpl-117 and Cpl-177) and their parental proteins (Cpl-1
and Cpl-7S).
Cloning, overproduction and purification of the new enzymes
were carried out as detailed in the Materials and methods section.
The three-number code of these chimeric lysozymes corresponds
to the origin of the CD (first number), linker region (second num-
ber) and CWBD (third number) of both enzymes: 1 from Cpl-1 and
7 from Cpl-7S. Thus, Cpl-711 and Cpl-771 contain the CD of Cpl-7S
at the N-terminal region, and the CWBD
1
at the C-terminal region.
The only difference between them lies in the origin of the linker
connecting both domains, coming from either Cpl-1 (in Cpl-711)
or Cpl-7S (in Cpl-771). Following the same rule, proteins Cpl-117
and Cpl-177 share the CD of Cpl-1 and the CWBD of Cpl-7S
(CWBD
7S
), the difference being in the linkers coming from either
Cpl-1 (in Cpl-117) or Cpl-7S (in Cpl-177).
The amino acid substitutions introduced in Cpl-7S with respect
to Cpl-7 (five residues in each one of the three CW_7 repeats of the
CWBD
7
) led to a drastic change in total charge of the domain
(Z
CWBD
)from214.93 to +3.00.
14
It is worth noting that, despite
the sequence differences, the net charges of the CDs (Z
CD
)of
Cpl-1 and Cpl-7S are identical, as are those of their linkers (Z
link
),
meaning that the differences in the net charge of the four chi-
meric proteins reside in the CWBD. Therefore, the chimeras
Cpl-711 and Cpl-771 present a net charge identical to that of
Cpl-1, whereas those of chimeras Cpl-117 and Cpl-177 are identi-
cal to that of Cpl-7S (Figure 1). Charge has been demonstrated to
be a key factor in the optimal lytic activity of endolysins when
added exogenously.
14,39
Nevertheless, the specific activities of
the four purified chimeras and parental proteins were very similar,
Net charge
Protein
Specific
activity
(U/mg)
Origin
ZCD Zlink ZCWBD ZTotal
−10.86 −3.98 −0.98 −15.82 Cpl-1 8.00 × 104
−10.86 −3.98 +3.00 −11.84 Cpl-7S 6.50 × 104
−10.86 −3.98 −0.98 −15.82 Cpl-711 7.25 × 104
−10.86 −3.98 −0.98 −15.82 Cpl-771 9.50 × 104
−10.86 −3.98 +3.00 −11.84 Cpl-117 4.65 × 104
−10.86 −3.98 +3.00 −11.84 Cpl-177 6.75 × 104
400
Nt Ct
Amino acid number
0 100 200 300
Synthetic
Lytic phage
CW_binding_1
(PF01473)
GH_25 family (Cpl-7)
(PF01183)
GH_25 family (Cpl-1)
(PF01183)
Cpl-7 Linker
Cpl-1 Linker
Modified CW_binding_7
Figure 1. Schematic representation and characteristics of chimeric and parental murein hydrolases. The synthetic chimeric proteins Cpl-711, Cpl-771,
Cpl-117 and Cpl-177, and their parental proteins Cpl-1 and Cpl-7S, are shown together with their origin, specific activity (U/mg) and net charge. Nt, N-
terminal; Ct, C-terminal; Z
CD
, catalytic domain charge; Z
link
, linker charge; Z
CWBD
, cell wall-binding domain charge; Z
Total
, total charge.
Dı
´ez-Martı
´nez et al.
1766
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
ranging from 4.65×10
4
to 9.50×10
4
U/mg, when tested on radio-
actively labelled pneumococcal cell walls (Figure 1).
Bactericidal activity of chimeric enzymes against
pneumococcal strains
The bacteriolytic/bactericidal activity of the four chimeras was
tested against several pneumococcal strains. Initially, the antibac-
terial activity was analysed against S. pneumoniae R6 cells using
5mg/mL of each enzyme. Experiments were performed measuring
the OD
550
for 60 min and plating for bacterial counting. Cpl-711 and
Cpl-771, which possess the CWBD
1
characteristic of a choline-
dependent enzyme, showed dramatic bacteriolytic effects with a
rapid decrease in cell density, similar to that observed with Cpl-1.
On the other hand, Cpl-117 and Cpl-177, which contain the
CWBD
7S
, displayed lower bacteriolytic activity than Cpl-711 and
Cpl-771, but comparable to that of the parental Cpl-7S lysozyme
(Figure 2a). These results correlated well with data for bacterial sur-
vival after endolysin treatment. Cpl-711 and Cpl-771 nearly sterilized
the cultures, causing a decrease in R6 viability of ≥7.5 log, whereas
Cpl-117 and Cpl-177 reduced viability by 4.7 log (Figure 2b).
After demonstrating the extreme effect of 5 mg/mL Cpl-711
and Cpl-771 on R6 cells, we tested their efficacy at lower concen-
trations to determine the differences in bactericidal activity com-
pared with Cpl-1, considered to date as the most active endolysin
against pneumococcal infections.
40
To accomplish this, bacteri-
cidal assays were repeated using the three lysins at 0.1 and
0.01 mg/mL. Cpl-711 and Cpl-771 produced greater decreases in
OD
550
than Cpl-1 at both concentrations (Figure 2cande).
Furthermore, cell viability analysis after 1 h of treatment revealed
that chimera Cpl-711 showed the highest bactericidal activity
against R6, reducing the number of viable cells by 2 and 2.6 log
at 0.01 and 0.1 mg/mL, respectively. This reduction was about
1 log more than Cpl-771 at both concentrations. This large lethal
effect,achievedwithsuchalowconcentrationofchimeric
enzyme, largely exceeded the killing capacity of Cpl-1, which at
0.1 mg/mL reduced cell viability by 90% and at 0.01 mg/mL by
only 15% (Figure 2d and f).
To check whether Cpl-711 was also the most active chimera
against encapsulated pneumococci, similar experiments were
done using the S. pneumoniae strains D39 (serotype 2), P007
(serotype 3) and P008 (serotype 4), and the multiresistant clinical
strains 1515/97 (serotype 6B) and 48 (serotype 23F). As in the pre-
vious assays, the difference in terms of bactericidal activity
among chimeric and parental proteins was clearer when the
tests were performed with reduced concentrations of the
enzymes. A summary of the bactericidal experiments with these
strains using 0.1 mg/mL of each protein is shown in Table 1.
Cpl-711 was also the most potent bactericidal agent against
P007 and P008 strains, decreasing cell viability by 2 log, whereas
the chimera Cpl-771 and the parental protein Cpl-1 reduced via-
bility by ≤1 log when tested with these strains. Furthermore,
Cpl-711 was the only endolysin that showed a marked effect on
the multiresistant strain 48, reducing viability by 1 log. As for
the other strains tested, D39 and the multiresistant 1515/97,
both Cpl-711 and Cpl-771 chimeras exhibited a similar bacteri-
cidal effect, which was 1 log higher than that of Cpl-1. These
results confirmed that, despite their similar specific activities
in vitro, the four chimeras do not display the same bactericidal
properties when acting from the outside of living bacteria.
Moreover, Cpl-711 was the most lethal enzyme independent of
the size and composition of the pneumococcal capsule. In add-
ition, results derived from these bactericidal assays were confirmed
by MICs determined for the S. pneumoniae ATCC 49619 strain,
which were 4.5+0.5 mg/L for Cpl-711, 8.5+0.5 mg/L for Cpl-771,
255+5 mg/L for Cpl-117 and 350+35 mg/L for Cpl-177, whereas
those of Cpl-1 and Cpl-7S were 16+4and54+10 mg/L, respect-
ively.
14
To visualize the extraordinary killing effect of the chimeric
Cpl-711 lysin on pneumococcal strains, we recorded a video under
the microscope (see Video S1), which clearly demonstrates in real
time the rapid lytic process of purified Cpl-711 on strain R6.
Bactericidal activity of chimeric proteins against other
streptococcal species
As previously described, the presence of CW_7 repeats in Cpl-7
and Cpl-7S confers the ability to efficiently kill several non-
pneumococcal bacteria, with the engineered variant Cpl-7S
being the most powerful lysin against them.
14
To test the capacity
of the four newly constructed chimeras to kill other streptococcal
species, bactericidal assays against the human pathogens S.
pyogenes and Streptococcus mitis SK137 were performed. In the
case of S. pyogenes, 0.1 mg/mL of any chimera showed a poor bac-
tericidal effect (data not shown), so the experiments were
repeated using 5 mg/mL protein. Under these conditions, Cpl-7S
displayed the highest lethal effect (4 log decrease in viability) fol-
lowed by Cpl-117 and Cpl-177 with 2 log lethality (Figure S1A).
These values confirmed that the chimeras harbouring the
CWBD
7S
did not improve the bactericidal activity of the parental
Cpl-7S enzyme. Cpl-1 did not kill S. pyogenes, in accordance with
previous data
14
and, similarly, the same negative results were
obtained for Cpl-711 and Cpl-771 chimeras (Figure S1A). As
expected from the presence of choline in its cell wall,
21
S. mitis
SK137 revealed a greater susceptibility to Cpl-1, Cpl-711 and
Cpl-771, with Cpl-711 being the most active, killing 4.7 log of bac-
teria, whereas Cpl-771 and Cpl-1 reduced viability by 3.6 log
(Figure S1B). In this case, Cpl-7S did not produce the same lethal
effect as with S. pyogenes, causing a decrease in viability of
3.3 log, similar to Cpl-1. Chimeras Cpl-117 and Cpl-177 were the
least active enzymes against this species, as they killed about
2.3 log of cells (Figure S1B). These results reinforce the idea that
differences in susceptibility to the action of these endolysins are
directly related to the specificity and affinity of the corresponding
CWBDs, although sequence composition of CDs and modular
organization are also relevant, as well as the composition and
structure of the whole bacterial envelope.
Dispersing and lethal activity of Cpl-711 against
pneumococcal biofilms
As with many bacterial species, S. pneumoniae is capable of grow-
ing as stable biofilms, a microbial community associated with up
to 80% of all chronic infections,
41
including those recalcitrant to
antibiotics.
42,43
Several bacterial and phage murein hydrolases
(LytA, Ejl, Pal, Cpl-1 and Cpl-7) efficiently destroy pneumococcal
biofilms.
30
Thus, we also examined the ability of Cpl-711 to dis-
persepneumococcalcellswhengrownasbiofilms.At1mg/mL
Cpl-711 reduced biofilm formation by 92%, whereas Cpl-1 and
Cpl-7S reached 50% and 60%, respectively (Figure 3a). Moreover,
the loss of viable cells in the same experiment was more dramatic
Chimeric lysins against pneumococcus
1767
JA
C
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
since Cpl-711 killed about 4 log of the bacterial population, in con-
trast to the 1.5 log drop caused by Cpl-1 and Cpl-7S (Figure 3b).
The different degree of biofilm disintegration was fully confirmed
by CLSM, which clearly showed the thickness of corresponding
biofilms as well as the remaining live and dead cells (Figure 3c).
Activity of Cpl-711 in a mouse infection model
To validate the in vitro bactericidal results of Cpl-711 on pneumo-
coccal cells, we employed a mouse bacteraemia model and com-
pared the in vivo activity of the enzyme with the parental Cpl-1.
The most appropriate bacterial challenge dose that was lethal
0.6
0.4
0.2
OD550
OD550
0
Time (min)
15 30 45 60410
0.6
0.4
0.2
0
Time (min)
15 30 45 60410
2
4
6
8
Control
Cpl-1
Cpl-711
Cpl-771
Control
Cpl-1
Cpl-711
Cpl-771
2
4
6
8
**
10
*
**
(a)
(b)
(c) (d)
(e)
(f)
Log10 cfu/mLLog10 cfu/mLLog10 cfu/mL
Cpl-771
Cpl-711
Cpl-7S
Cpl-1
Control
Cpl-117
Cpl-177
Cpl-771
Cpl-1
Cpl-7S
Cpl-117
Cpl-177
Control
Cpl-711
Cpl-771
Cpl-1
Control
Cpl-711
Cpl-771
Cpl-1
Control
Cpl-711
15 30 45
410
Time (min)
60
OD550
0.6
0.4
0.2
0
0.8
2
4
6
8
**
**
**
**
** **
Figure 2. Bacteriolytic and bactericidal effects of different chimeric and parental enzymes against S. pneumoniae strain R6. Enzymes were added at
5mg/mL (a and b), 0.1 mg/mL (c and d) and 0.01 mg/mL (e and f). Decay of the bacterial suspension (OD
550
) after addition of the selected enzyme
(a, c and e). Incubation was performed at 378C for 60 min. Controls with PBS are included. Data correspond to a representative experiment. Viable
cells determined on blood agar plates after 60 min of incubation at 378C (b, d and f). Data are means from four independent experiments. Error
bars represent standard deviations, and asterisks indicate that results are statistically significant compared with the control in the absence of
enzymes (one-way ANOVA with a post hoc Dunnet test; *P,0.01, **P,0.001).
Dı
´ez-Martı
´nez et al.
1768
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
to mice in 2–3 days was found to be 5.5×10
5
cfu/mL for pneumo-
coccal strain D39_IU, when administered by ip injection. One hour
after bacterial challenge, animals were treated with a single ip injec-
tion containing different amounts of either Cpl-711 or Cpl-1 (in the
range of 25 –500 mg per mouse), or PBS as control. Mouse survival
was followed over 7 days and the results of four different experi-
ments were combined and represented as a Mantel –Cox curve.
Rescue of the animals treated with Cpl-711 varied depending on
the concentration used: 100% survival (20/20) when treated with
500 mg of the enzyme (not shown); 70% survival (14/20) when trea-
tedwith200mg; 50% survival (10/20) when treated with 50 mg; and
45% survival (9/20) when treated with 25 mg. These results show
that treatment with the Cpl-711 chimera resulted in about 50%
greater protection than treatment with Cpl-1. For example, only
20% of mice (4/20) survived when treated with 50 mgofCpl-1com-
pared with 50% survival with the same amount of Cpl-711 (Figure4).
Cpl-711 and Cpl-771 retain the global structure and
choline-binding capacity of parental Cpl-1
In Cpl-1 folding, the relative positions of the CD and the CWBD
1
are
restrained by the interactions between both modules and the
Table 1. Bactericidal activity of the different chimeras and their parental
enzymes against S. pneumoniae
Strain
(serotype)
a
Viability decrease after treatment with
b
Cpl-1 Cpl-7S Cpl-711 Cpl-771 Cpl-117 Cpl-177
S. pneumoniae
R6 (none) +2+++ ++ 22
D39 (2) +2++ ++ 22
P007 (3) 22++ + 22
P008 (4)
c
+2++ + 22
1515/97 (6B) 22++22
48 (23F) 22+222
a
Bacteria were incubated at 378C in PBS (OD
550
0.6) with the indicated
enzyme at 0.1 mg/mL. Viability was determined after 1 h of incubation
by plating samples onto blood agar plates and analysing cfu.
b
2indicates no effect, and +,++ and +++ indicate a decrease in viability
of 1, 2 or 3 log, respectively.
c
Constructed as strain P007, but with DNA froma serotype 4 pneumococcal
strain.
2
4
6
8**
Control
Cpl-711
Cpl-1
Cpl-7S
**
Cpl-711
Cpl-1
Cpl-7S
*
1 µg/mL 0.1 µg/mL
Log10 cfu/mL
**
** **
20
40
60
80
*
Control
Cpl-711
Cpl-1
Cpl-7S
*
Cpl-711
Cpl-1
Cpl-7S
*
100
(a)
(c)
(b)
Biofilm formation (%)
1 µg/mL 0.1 µg/mL
**
**
**
4
2 31
Figure 3. Degradation of pneumococcal biofilms upon addition of Cpl-711. (a) Percentage of crystal violet stainingremaining in the S. pneumoniae P046
biofilm after treatment for 2 h at 378C with 1 or 0.1 mg/mL Cpl-711 and the parental proteins Cpl-1 and Cpl-7S. The values of biofilm were normalized to
the highest absorbance reading, and the percentage was calculated in relation to untreated strain P046. (b) Viable cells of S. pneumoniae P046 growing
as biofilms determined on blood agar plates after treatment for 2 h at 378C with 1 and 0.1 mg/mL Cpl-711, Cpl-1 and Cpl-7S. (c) CLSM image of the
viability of biofilm-grown S. pneumoniae P046 treated for 2 h with 0.1 mg/mL Cpl-711 (panel 2), Cpl-1 (panel 3) or Cpl-7S (panel 4) or without any
enzyme (panel 1). In all panels with calculated data, the data represent the mean of four independent experiments. Error bars represent standard
deviations, and asterisks indicate that results are statistically significant compared with the control in the absence of enzymes (one-way ANOVA with
a post hoc Dunnet test; *P,0.05, **P,0.01). Bar, 25 mm.
Chimeric lysins against pneumococcus
1769
JA
C
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
tightness of the linker, which contains a DDEEDD motif that could
be responsible for the enzyme’s extended conformation (Figure S2).
In addition, the fold adopted by the region of CWBD
1
that interacts
with the CD also determines the total number of choline-binding
functional sites.
7
The CWBD and linker of Cpl-1 and Cpl-711 are
identical and their CDs differ in only 27 residues, but those involved
in the interface with the CWBD
1
(Ala160, Phe162, Glu163, Phe165,
Trp174, Tyr176, Ile185 and Leu187 in strands b6–b8) are all con-
served (Figure S2). Therefore, the modular structure and the
choline-binding capacity of Cpl-1 are likely preserved in Cpl-711,
whose three-dimensional structure has been modelled using the
X-ray structure of Cpl-1 as template (Figure S2C). The same prob-
ably applies to the Cpl-771 chimera, whose linker conserves five
of the six initial acidic residues (Figure S2A), although some differ-
ences might be expected in the linker conformation because of its
length (16 residues in Cpl-771 versus 11 in Cpl-711). To check these
assumptions, the binding of choline to Cpl-711 and Cpl-771 was
measured and the structural organization of both chimeras exam-
ined with and without choline using analytical ultracentrifugation
and circular dichroism. As shown in Figure S3(A and B), the circular
dichroism spectra of Cpl-1, Cpl-711 and Cpl-771 are very similar,
both in the far- and near-UV regions, confirming that their second-
ary and tertiary structures are comparable in the free and choline-
bound forms. The small differences in the regions sensitive to the
contribution of aromatic side chains are probably due to the pres-
ence of one more tryptophan (Trp63) and the change F77Y in the
CD of Cpl-7. Indeed, the spectra of the two chimeras are identical
within experimental error. Variations of the circular dichroism spec-
tra induced by choline addition were also alike and the titration
curves obtained by monitoring the ellipticity changes at 225 nm
as a function of ligand concentration showed that choline-binding
affinities were also comparable (Figure S3C). The dissociation con-
stants estimated by fitting the Hill equation to the titration data
varied from 4.1+0.2 mM (Cpl-711) to 3+1 mM (Cpl-771) and
3.6+0.2 mM (Cpl-1), and the Hill coefficients were also similar
(Table S2).
The association state and the hydrodynamic features of
Cpl-711 and Cpl-771 and their complexes with choline were
examined by analytical ultracentrifugation in comparison with
Cpl-1. In the absence of choline, all of the enzymes sediment as
single species, whose almost identical sedimentation coefficients
correspond to protein monomers (Figure S3D), as verified by sedi-
mentation equilibrium experiments (Table S2). Like in Cpl-1, cho-
line binding regulates the self-association of the two chimeras
that sediment as dimers under choline-saturating conditions
(Figure S3D and Table S2), a characteristic that has been shown
to be relevant for LytA and Cpl-1 activities.
34,44–46
Additionally,
neither the linker nor the CD exchange has perceptible conse-
quences on the overall protein shape since the hydrodynamic
parameters (f/f
0
and R
s
) calculated from sedimentation data for
each association (monomer or dimer) state were alike for the
three proteins (Table S2). Together these results confirmed that
the overall structural features of Cpl-1 are preserved in Cpl-711
and Cpl-771, in agreement with our hypothesis.
Discussion
Endolysins, like all phage-encoded proteins, have been subjected
to a functional optimization process through joint evolution with
their obligate bacterial hosts over millions of years. The final
results are enzymes selected to cause rapid lysis of the host for
efficient release of phage progeny, thus assuring phage survival.
This evolutionary process has generated a wide variety of lysins
that originate from the continuous recombination of different
functional modules, from bacteria and phages, and their subse-
quent horizontal gene transfer.
47
Many of the natural enzymes
encoded by these chimeric genes have been well studied, the
putative origin of each module has been suggested and different
fusions or deletions constructed and tested to determine their
activities against the susceptible bacteria.
48
Additionally, func-
tional synthetic endolysins have been engineered at the genetic
level by module recombination to suit specific lytic requirements.
For example, fusion of the lysostaphin gene from Staphylococcus
simulans to the B30 endolysin gene from a Streptococcus agalactiae
phage, or to its endopeptidase-coding module, created new
enzymes active against S. agalactiae,Streptococcus uberis and
Staphylococcus aureus, the main pathogens responsible for mas-
titis in cows.
49
It has also been demonstrated that pneumococcal
murein hydrolases from either the host or phages harbour a CWBD
essential for activity
50
that can be exchanged with other pneumo-
coccal or clostridial CWBDs to recreate functional chimeric
lysin.
51,52
Nonetheless, the overall published data suggest that
natural endolysins are optimal enzymes to lyse and kill certain
bacteria via their natural route. However, it is plausible to consider
the possibility for enzymatic improvement through genetic engin-
eering, at least when endolysins are added exogenously.
Detailed knowledge of the structural and functional properties of
lysozymes Cpl-1 and Cpl-7 prompted us to explore the effect of
combining their CDs and CWBDs as well as the linkers connecting
them. To this aim, we took into account the information provided
by the crystal structure of the entire Cpl-1 (free and ligand-bound)
and the available structural models of Cpl-7.
7,8,53
We chose t he sy n-
thetic Cpl-7S lysozyme instead of the natural Cpl-7 to construct the
25
50
75
100
Survival (%)
0
Time post-infection (days)
1234567
**
*
*
Cpl-1 200 µg
Cpl-711 200 µg
Cpl-711 50 µg
Cpl-1 50 µg
Cpl-1 25 µg
Control
Cpl-711 25 µg
Figure 4. Mice were rescued from lethal S. pneumoniae D39_IU infection by
Cpl-711. Mice were injected ip with 5.5×10
5
cfu/mL S. pneumoniae D39_IU
strain (time 0). One hour later, mice received one ip injection of either PBS
(control) or lysin. The lines represent the survival of mice treated (or not)
with different concentrations of Cpl-711 or Cpl-1. Mice were monitored
for survival over a period of 7 days and results plotted as Kaplan –
Meier survival curves. Survival curves were compared with the log-rank
(Mantel–Cox) and Gehan– Breslow–Wilcoxon tests (*P,0.01; **P,0.001).
Dı
´ez-Martı
´nez et al.
1770
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
new chimeric enzymes because of its superior bactericidal activ-
ity.
14
The most remarkable (and unexpected) result of our approach
was the notable bactericidal capacityof Cpl-711, which was clearly
greater than that exhibited by the natural endolysin Cpl-1, consid-
ered to date as the most powerful lytic enzyme against pneumococ-
cal infections. This improved activity of Cpl-711 was evident not only
against several pneumococcal strains grown as planktonic cultures,
including some antibiotic-resistant strains, but also against pneumo-
coccal biofilms in vitro, and gave superior protection compared with
other lysins in a mouse model of bacteraemia. Therefore, the even-
tual application of this potent chimeric protein to treat multiresistant
pneumococcal infections deserves future investigation.
The noticeable improvement of the bactericidal activity of
Cpl-711, compared with that of Cpl-1, is interesting in itself
since, to our knowledge, despite the numerous endolysin fusion
constructs reported so far, only the Ply187-KSH3 chimera
showed a higher bactericidal effect over its parental enzyme.
54
Nevertheless, Ply187, encoded by a phage infecting S. aureus,
was atypical among endolysins from Gram-positive bacteria
because of the poor activity displayed by the recombinant
enzyme, which was probably due to the absence of a CWBD.
Thus, the addition of an SH3b domain from LysK endolysin,
known to act as a CWBD, enhanced its lytic activity.
54
In contrast,
Cpl-711 is the result of swapping of catalytic and substrate-
binding domains from natural bimodular lysozymes that already
behave as very potent bactericidal agents.
Any explanation of the observed differences in activity among
the four constructed chimeras and the two parental enzymes
must consider the concept of modular evolution of proteins
55
and the fact that the activity of the complete enzyme will be
determined by the specific features of the isolated domains as
well as by those conferred by the interactions established
between them and with the optimized linkers. In addition, the
lytic action of endolysins from the outside of intact cells is prob-
ably a complex process controlled by many factors, like the com-
position and charges of the endolysin and the bacterial envelope.
In this respect, the lack of correlation between the activities mea-
sured using purified cell wall fragments or intact cultures of R6
strain as substrates, as well as the influence of capsule compos-
ition on the killing activity of Cpl-1, Cpl-711 and Cpl-771 (Table 1),
clearly demonstrated that substrate fragmentation and/or the
inherent removal of components attached to the bacterial enve-
lope minimize the potential impact of sequence differences on the
enzymatic activity. This lack of correlation may be due to differ-
ences in substrate accessibility and bond cleavage.
Another general conclusion drawn from this study is that the
type of CWBD is a major factor in the lethality of endolysins, i.e.
Cpl-117 and Cpl-177 are very similar to the parental Cpl-7S,
whereas Cpl-711 and Cpl-771 are rather similar to Cpl-1. It is
well established that the nature of the CWBD influences substrate
specificity and cell-wall binding affinity as well as the overall endo-
lysin structure. In this respect, Cpl-1 structure differs from that of
Cpl-7, which, according to a SAXS-based model, adopts an
extended structure with the CWBD
7
horizontally packed to the
catalytic barrel.
7,8
Such divergence, enhanced by choline-
mediated dimerization of Cpl-711, Cpl-771 and Cpl-1, would likely
modify protein diffusion as well as substrate recognition and lytic
efficiency. Interestingly, a covalently stabilized dimer of Cpl-1 was
shown to have increased antipneumococcal activity and
decreased plasma clearance.
46
Since Cpl-1 and Cpl-711 onlydiffer in the CD, the first clue tohelp
explain the large increase in lethal activityof Cpl-711 may be found
in the 27 distinct residues of both domains (Figure S2). These amino
acid differences affect neither the catalytic residues, which are
sequentially and structurally conserved,
7,8
nor the modular organ-
ization and choline-binding capacity. Non-conservative changes
are located mostly between the loop connecting strand b3to
helix-3 and the beginning of helix-4, which includes the region con-
taining the proton donor Glu94 (Figure S2). However, Cpl-1 and
Cpl-711 have the same overall net charge. Consequently, this fac-
tor, so relevant for the activity of Cpl-7S and other endolysins,
14,39
does not explain the improved activity of Cpl-711 compared with
natural Cpl-1. Nevertheless, the amino acid changes modify the
electrostatic potential of the catalytic surface (Figure S4), which
could affect the unfavourable electrostatic interactions with the
negative charge of the cell wall. This can also slightly alter the pat-
tern of the contacts observed in the model proposed for the com-
plex between Cpl-1 and a pentasaccharide dipentapeptide.
8,56
On
the other hand, it seems unlikely that the glycopeptide chain bound
to the catalytic cleft could make contact with other regions of the
CD. However, the possibility that endolysins could interact with
adjacent muropeptide chains (or other molecules present in the
intact cell wall) cannot be completely ruled out. It might help
explain their different behaviour with cell wall fragments or intact
cells, as well as the distinct bactericidal activities reported here.
Interestingly, the structure of the Cpl-1 E94D mutant complexed
with a disaccharide pentapeptide showed an additional molecule
of the substrate analogue bound outside the catalytic groove
(PDB access code: 2J8F). For similar reasons, a certain rearrange-
ment of a longer linker in the Cpl-771 chimera could account for
its decrease in lethality and MIC increase compared with Cpl-711.
The absence of a structural model of the entire Cpl-7 at atomic
resolution precludes a thorough analysis of the differences found
among Cpl-7S, Cpl-177 and Cpl-117, although the SAXS-based
model of Cpl-7 suggests that, like in Cpl-1, the hydrophobic cavity
formed by strands b6–b8 might interact with the CWBD
7S
.
8
Of
note, the increase in the MIC values of Cpl-177 and Cpl-117 with
respect to Cpl-7S is comparable to the reduction observed on pas-
sing from Cpl-1 to Cpl-711 or Cpl-771 (i.e. the inverse modifica-
tion) and further supports the notion that sequence divergences
make the CD of Cpl-7 more efficient. Elucidation of the three-
dimensional structures of the full-length chimeras together with
that of Cpl-7S and their complexes with peptidoglycan analogues
could provide interesting, although partial, insights into the
molecular basis of the differences in catalytic activity and/or sub-
strate binding of this collection of lysozymes.
In summary, this study discloses the complexity of mechanisms
underlying the exogenous bacteriolytic/bactericidal action of endo-
lysins and demonstrates the feasibility of constructing new and
more lethal phage lysins from WTor modified preexisting enzymes.
Taking into account the high diversity within bacterial and bacterio-
phage populations, there are great opportunities for designing
novel antimicrobials that might reduce the increasing incidence
of antibiotic resistance obstructing therapeutic efforts worldwide.
Acknowledgements
We thank Ernesto Garcı
´a for helpful comments and for critically reading
the manuscript prior to submission, and E. Cano, S. Ruiz and V. Lo
´pez for
Chimeric lysins against pneumococcus
1771
JA
C
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
skillful technical assistance. We are indebted to M. Domenech,
M. Seisdedos and G. E. Serrano for their advice and help with CLSM
imaging, and to G. Garcı
´a for help in protein purification.
Funding
This research was funded by grants from the Ministerio de Economı
´ay
Competitividad (MINECO) to P. G. (SAF2009-10824 and SAF2012-39444-
C02-01), and from the MINECO (BFU2009-10052 and BF4U2012-36825)
and the Consejerı
´adeEducacio
´n de la Comunidad de Madrid (S2010/
BMD/2457) to M. M. and Rockefeller University laboratory funds to
V. A. F. Additional funding was provided by the CIBER de Enfermedades
Respiratorias (CIBERES), an initiative of the Instituto de Salud Carlos III
(ISCIII). R. D.-M. was the recipient of one fellowship from the MINECO
(FPI programme). The funders had no role in study design, data collection
and analysis, decision to publish or preparation of the manuscript.
Transparency declarations
None to declare.
Supplementary data
Table S1, Table S2, Video S1 and Figures S1 to S4 are available as
Supplementary data at JAC Online (http://jac.oxfordjournals.org/).
References
1Andrade AL, Toscano CM, Minamisava R et al. Pneumococcal disease
manifestation in children before and after vaccination: what’s new?
Vaccine 2011; 29 Suppl 3: C2–14.
2McCarthy M. CDC calls for urgent action to combat rise of drug resistant
pathogens. BMJ 2013; 347: f5649.
3Fischetti VA. Bacteriophage lysins as effective antibacterials. Curr Opin
Microbiol 2008; 11: 393– 400.
4Loessner MJ. Bacteriophage endolysins—current state of research and
applications. Curr Opin Microbiol 2005; 8: 480– 7.
5Rigden DJ, Jedrzejas MJ, Galperin MY. Amidase domains from bacterial
and phage autolysins define a family of g-d,l-glutamate-specific amidohy-
drolases. Trends Biochem Sci 2003; 28: 230 – 4.
6Nelson DC, Schmelcher M, Rodriguez-Rubio L et al. Endolysins as antimi-
crobials. Adv Virus Res 2012; 83: 299– 365.
7Hermoso JA, Monterroso B, Albert A et al. Structural basis for selective
recognition of pneumococcal cell wall by modular endolysin from phage
Cp-1. Structure 2003; 11: 1239– 49.
8Bustamante N, Campillo NE, Garcı
´aEet al. Cpl-7, a lysozyme encoded by
a pneumococcal bacteriophage with a novel cell wall-binding motif. J Biol
Chem 2010; 285: 33184– 96.
9Lo
´pez R, Garcı
´a E. Recent trends on the molecular biology of pneumo-
coccal capsules, lytic enzymes, and bacteriophage. FEMS Microbiol Rev
2004; 28: 553–80.
10 Rau A, Hogg T, Marquardt R et al. A new lysozyme fold. Crystal structure
of the muramidase from Streptomyces coelicolor at 1.65 A
˚resolution. J Biol
Chem 2001; 276: 31994– 9.
11 Porter CJ, Schuch R, Pelzek AJ et al. The 1.6 A
˚crystal structure of the
catalytic domain of PlyB, a bacteriophage lysin active against Bacillus
anthracis.J Mol Biol 2007; 366: 540– 50.
12 Hermoso JA, Garcı
´a JL, Garcı
´a P. Taking aim on bacterial pathogens:
from phage therapy to enzybiotics. Curr Opin Microbiol 2007; 10: 461– 72.
13 Bustamante N, Rico-Lastres P, Garcı
´aEet al. Thermal stability of Cpl-7
endolysin from the Streptococcus pneumoniae bacteriophage Cp-7; cell
wall-targeting of its CW_7 motifs. PLoS One 2012; 7: e46654.
14 Dı
´ez-Martı
´nez R, de Paz H, Bustamante N et al. Improving the lethal
effect of Cpl-7, a pneumococcal phage lysozyme with broad bactericidal
activity, by inverting the net charge of its cell wall-binding module.
Antimicrob Agents Chemother 2013; 57: 5355– 65.
15 Hoskins J, Alborn WE Jr, Arnold J et al. Genome of the bacterium
Streptococcus pneumoniae strain R6. J Bacteriol 2001; 183: 5709–17.
16 Lanie JA, Ng W-L, Kazmierczak KM et al. Genome sequence of Avery’s
virulent serotype 2 strain D39 of Streptococcus pneumoniae and compari-
son with that of unencapsulated laboratory strain R6. JBacteriol
2007; 189: 38– 51.
17 Domenech M, Garcı
´a E, Moscoso M. Versatility of the capsular genes
during biofilm formation by Streptococcus pneumoniae.Environ Microbiol
2009; 11: 2542– 55.
18 Moscoso M, Garcı
´aE,Lo
´pez R. Biofilm formation by Streptococcus
pneumoniae: role of choline, extracellular DNA, and capsular polysacchar-
ide in microbial accretion. J Bacteriol 2006; 188: 7785– 95.
19 YusteJ,JadoI,FenollAet al.b-Lactam modification of the bacter-
aemic profile and its relationship with mortality in a pneumococcal
mouse sepsis model. J Antimicrob Chemother 2002; 49:331– 5.
20 Soriano F, Cafini F, Aguilar L et al. Breakthrough in penicillin resistance?
Streptococcus pneumoniae isolates with penicillin/cefotaxime MICs of
16 mg/L and their genotypic and geographical relatedness. J Antimicrob
Chemother 2008; 62: 1234– 40.
21 Bergstro
¨mN,JanssonPE,KilianMet al.Structuresoftwocell
wall-associated polysaccharides of a Streptococcus mitis biovar 1 strain.
A unique teichoic acid-like polysaccharide and the group O antigen
whichisaC-polysaccharideincommonwithpneumococci.Eur J
Biochem 2000; 267: 7147–57.
22 Sambrook J, Russell DW. Molecular Cloning, A Laboratory Manual. Cold
Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 2001.
23 Tabor S. Expression using the T7 RNA polymerase/promoter system. In:
Ausubel FM, Brent R, Kingston RE et al., eds. Current Protocols in Molecular
Biology. New York: Green and John Wiley & Sons, 1990; 16.2.1–11.
24 Sanz JM, Garcı
´a JL. Structural studies of the lysozyme coded by the
pneumococcal phage Cp-1. Conformational changes induced by choline.
Eur J Biochem 1990; 187: 409– 16.
25 Lacks S, Hotchkiss RD. A study of the genetic material determining an
enzyme activity in Pneumococcus.Biochim Biophys Acta 1960; 39:
508–18.
26 MacFaddin JF. Media for Isolation –Cultivation –Identification –Maintenance
of Medical Bacteria. Baltimore, MD: Williams & Wilkins, 1985.
27 Ho SN, Hunt HD, Horton RM et al. Site-directed mutagenesis by
overlap extension using the polymerase chain reaction. Gene 1989; 77:
51–9.
28 Ho
¨ltje JV, Tomasz A. Purification of the pneumococcal N-acetylmuramyl-
l-alanine amidase to biochemical homogeneity. JBiolChem1976; 251:
4199 – 207.
29 Mosser JL, Tomasz A. Choline-containing teichoic acid as a structural
component of pneumococcal cell wall and its role in sensitivity to lysis
by an autolytic enzyme. J Biol Chem 1970; 245: 287 – 98.
30 Domenech M, Garcı
´a E, Moscoso M. In vitro destruction of
Streptococcus pneumoniae biofilms with bacterial and phage peptidogly-
can hydrolases. Antimicrob Agents Chemother 2011; 55: 4144–8.
31 Jado I, Lo
´pez R, Garcı
´aEet al. Phage lytic enzymes as therapy of
antibiotic-resistant Streptococcus pneumoniae infection in a murine sepsis
model. J Antimicrob Chemother 2003; 52: 967– 73.
Dı
´ez-Martı
´nez et al.
1772
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
32 Gilmer DB, Schmitz JE, Euler CW et al. Novel bacteriophage lysin with
broad lytic activity protects against mixed infection by Streptococcus pyo-
genes and methicillin-resistant Staphylococcus aureus.Antimicrob Agents
Chemother 2013; 57: 2743– 50.
33 Moreno FJ, Quintanilla-Lo
´pez JE, Lebro
´n-Aguilar R et al. Mass spectro-
metric characterization of glycated b-lactoglobulin peptides derived from
galacto-oligosaccharides surviving the in vitro gastrointestinal digestion.
J Am Soc Mass Spectrom 2008; 19: 927–37.
34 Monterroso B, Sa
´iz JL, Garcı
´aPet al. Insights into the structure-
function relationships of pneumococcal cell wall lysozymes, LytC and
Cpl-1. J Biol Chem 2008; 283: 28618–28.
35 Schuck P. Size-distribution analysis of macromolecules by sedimenta-
tion velocity ultracentrifugation and Lamm equation modeling. Biophys J
2000; 78: 1606– 19.
36 Laue TM, Shah BD, Ridgeway TM et al. Computer-aided interpretation
of analytical sedimentation data for proteins. In: Harding SE, Rowe AJ,
Horton JC, eds. Analytical Ultracentrifugation in Biochemistry and Polymer
Science. Cambridge: Royal Society of Chemistry, 1992; 90 – 125.
37 Laskowski RA, MacArthur MW, Moss DS et al.PROCHECK: a program to
check the stereochemical quality of protein structures. J Appl Crystallogr
1993; 26: 283–91.
38 Lu¨thy R, Bowie JU, Eisenberg D. Assessment of protein models with
three-dimensional profiles. Nature 1992; 356:83–5.
39 Low LY, Yang C, Perego M et al.Roleofnetchargeonthecatalytic
domain and the influence of the cell-wall binding domain on the bacteri-
cidal activity, specificity, and host-range of phage lysins. JBiolChem
2011; 286: 34391– 403.
40 Grandgirard D, Loeffler JM, Fischetti VA et al. Phage lytic enzyme Cpl-1
for antibacterial therapy in experimental pneumococcal meningitis. J
Infect Dis 2008; 197:1519–22.
41 Wolcott RD, Ehrlich GD. Biofilms and chronic infections. JAMA
2008; 299: 2682– 4.
42 Moscoso M, Domenech M, Garcı
´a E. Vancomycin tolerance in
Gram-positive cocci. Environ Microbiol Rep 2011; 3: 640– 50.
43 Domenech M, Garcı
´a E, Moscoso M. Biofilm formation in Streptococcus
pneumoniae.Microb Biotechnol 2012; 5: 455– 65.
44 Varea J , Sa iz JL , Lo
´pez-Zumel C et al.Dosequencerepeatsplayan
equivalent role in the choline-binding module of pneumococcal LytA ami-
dase? J Biol Chem 2000; 275: 26842–55.
45 Romero P, Lo
´pez R, Garcı
´a E.Key role of amino acid residues in the dimer-
ization and catalytic activation of the autolysin LytA, an important virulence
factor in Streptococcus pneumoniae.JBiolChem2007; 282: 17729– 37.
46 Resch G, Moreillon P, Fischetti VA. A stable phage lysin (Cpl-1) dimer
with increased antipneumococcal activity and decreased plasma clear-
ance. Int J Antimicrob Agents 2011; 38: 516– 21.
47 Hendrix RW. Bacteriophages: evolution of themajority. Theor Popul Biol
2002; 61: 471–80.
48 Sheehan MM, Garcı
´a JL, Lo
´pez R et al. The lytic enzyme of the pneumo-
coccal phage Dp-1: a chimeric lysin of intergeneric origin. Mol Microbiol
1997; 25: 717–25.
49 Donovan DM, Dong S, Garrett W et al. Peptidoglycan hydrolase fusions
maintain their parental specificities. Appl Environ Microbiol 2006; 72: 2988– 96.
50 Garcia JL, Diaz E, Romero A et al. Carboxy-terminal deletion analysis of
the major pneumococcal autolysin. J Bacteriol 1994; 176: 4066– 72.
51 Dı
´az E, Lo
´pez R, Garcı
´a JL. Chimeric phage-bacterial enzymes: a clue to
the modular evolution of genes. Proc Natl Acad Sci USA 1990; 87: 8125–9.
52 Croux C, Ronda C, Lo
´pez R et al. Interchange of functional domains
switches enzyme specificity: construction of a chimeric pneumococcal-
clostridial cell wall lytic enzyme. Mol Microbiol 1993; 9: 1019– 25.
53 Silva Martin N. Biologı
´a estructural de las interacciones pato
´geno - hos-
pedador en Streptococcus pneumoniae: mecanismos de reconocimiento y
lisis de la pared bacteriana. PhD Thesis. Madrid: Universidad Auto
´noma de
Madrid, 2011.
54 Mao J, Schmelcher M, Harty WJ et al. Chimeric Ply187 endolysin kills
Staphylococcus aureus more effectively than the parental enzyme. FEMS
Microbiol Lett 2013; 342:30–6.
55 Copley RR, Ponting CP, Schultz J et al. Sequence analysis of multido-
main proteins: past perspectives and future directions. Adv Protein Chem
2003; 61: 75– 98.
56 Pe
´rez-Dorado I, Campillo NE, Monterroso B et al. Elucidation of the
molecular recognition of bacterial cell wall by modular pneumococcal
phage endolysin CPL-1. J Biol Chem 2007; 282: 24990– 9.
Chimeric lysins against pneumococcus
1773
JA
C
at CSIC on June 2, 2015http://jac.oxfordjournals.org/Downloaded from
