Báo cáo khoa học: Identification and characterization of the metal ion-dependent L-alanoyl-D-glutamate peptidase encoded by bacteriophage T5 pdf

Identification and characterization of the metal
ion-dependent
L-alanoyl-D-glutamate peptidase
encoded by bacteriophage T5
Galina V. Mikoulinskaia
1
, Irina V. Odinokova
2
, Andrei A. Zimin
3
, Valentina Ya. Lysanskaya
3
,
Sergei A. Feofanov
1
and Olga A. Stepnaya
3
1 Branch of Shemyakin & Ovchinnikov’s Institute of Bioorganic Chemistry RAS, Pushchino, Moscow, Russia
2 Institute of Theoretical and Experimental Biophysics RAS, Pushchino, Moscow, Russia
3 Skryabin’s Institute of Biochemistry and Physiology of Microorganisms RAS, Pushchino, Moscow, Russia
Introduction
There are two ways by which phages lyse host cells.
They can either cause ‘lysis from without’ – which
takes place when the cell is being infected – or
induce ‘lysis from within’, which occurs when the
phage progeny escape the host cell. There can be at
least two evolutionary strategies for lysis from within
[1,2]. The phages whose genome is dsDNA (like
lambda phage) usually have a two-protein lysis
system: this consists of endolysin, which destroys the
bacterial cell wall, and holin, a hydrophobic protein

providing access of endolysin to the substrate (pepti-
doglycan of the cell wall) by forming lesions in the
inner membrane. Endolysins are proteins with vari-
ous muralytic activities; as a rule, they are initially
accumulated in the cytoplasm before lysis, which is
eventually triggered by holin. However, some endoly-
sins have an N-terminal translocation domain, and
will therefore accumulate in the periplasm; in these
cases, holin is not necessary, and lysis will be medi-
ated by the Sec system of the host cell [3].
Keywords
bacteriophage T5; endolysin; Gram-negative;
holin;
L-alanoyl-D-glutamate peptidase
Correspondence
G. V. Mikoulinskaia, Branch of Shemyakin &
Ovchinnikov’s Institute of Bioorganic
Chemistry RAS, Prospekt Nauki, 6,
Pushchino, Moscow Region 142290, Russia
Fax: +7 4967 330527
Tel: +7 4967 731780
E-mail: mikulinskaya@fibkh.serpukhov.su
(Received 14 August 2009, revised
14 October 2009, accepted 16 October
2009)
doi:10.1111/j.1742-4658.2009.07443.x
Although bacteriophage T5 is known to have lytic proteins for cell wall
hydrolysis and phage progeny escape, their activities are still unknown.
This is the first report on the cloning, expression and biochemical charac-
terization of a bacteriophage T5 lytic hydrolase. The endolysin-encoding

lys gene of virulent coliphage T5 was cloned in Escherichia coli cells, and
an electrophoretically homogeneous product of this gene was obtained with
a high yield (78% of total activity). The protein purified was shown to be
an l-alanoyl-d-glutamate peptidase. The enzyme demonstrated maximal
activity in diluted buffers (25–50 mm) at pH 8.5. The enzyme was strongly
inhibited by EDTA and BAPTA, and fully reactivated by calcium ⁄ manga-
nese chlorides. It was found that, along with E. coli peptidoglycan,
peptidase of bacteriophage T5 can lyse peptidoglycans of other Gram-nega-
tive microorganisms (Pectobacterium carotovorum, Pseudomonas putida,
Proteus vulgaris, and Proteus mirabilis). This endolysin is the first example
of an l-alanoyl-d-glutamate peptidase in a virulent phage infecting Gram-
negative bacteria. There are, however, a great many sequences in databases
that are highly similar to that of bacteriophage T5 hydrolase, indicating a
wide distribution of endolytic l-alanoyl-d-glutamate peptidases. The article
discusses how an enzyme with such substrate specificity could be fixed in
the process of evolution.
Abbreviations
BAPTA, 1,2-bis-(O-aminophenoxy)ethane-N,N,N¢,N¢-tetraacetic acid; DNF, 2,4-dinitrophenyl.
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7329
Endolysins are divided into five classes according to
the type of bonds that they cleave in peptidoglycans.
The classes are as follows: (a) lysozyme-like mura-
midases, which hydrolyze the glycoside bond between
N-acetylmuramic acid and N-acetylglucosamine; (b)
lytic transglycosylases, which attack the same bonds as
muramidases but additionally catalyze the intramole-
cular transfer of the O-muramylic residue to the
C6 hydroxyl; (c) N-acetyl-b-d-glucosaminidases, which
hydrolyze the glycoside bond between N-acetylgluco-
samine and N-acetylmuramic acid; (d) N-acetylmura-

myl-l-alanine amidases, which hydrolyze bonds
between N-acetylmuramic acid and l-alanine; and (e)
peptidases, which hydrolyze peptide bonds [4,5]. Some
endolysins have multiple activities: for example, endo-
lysin of bacteriophage B30 is both a muramidase and
an endopeptidase [6].
Most endolysins have two functional domains: an
enzymatically active N-terminal domain, and a C-ter-
minal domain, which is responsible for recognition of
the substrate (peptidoglycan) and determines the speci-
ficity of the enzyme [4]. The antibacterial effect of end-
olysins is not always associated with their enzymatic
activity. For example, endolysin of bacteriophage T4
has four amphipathic C-terminal a-helices, whose basic
(positively charged) amino acids interact with nega-
tively charged components of the outer membrane,
and thus cause its degradation [7]. Endolysins may be
acid or alkaline proteins with diverse pH and ionic
strength optima.
The spectrum of antibacterial action of endolysins is
determined by the enzyme type, composition of the cell
wall components of the target bacterium, and configu-
ration of the substrate (peptidoglycan). The range of
bacteria sensitive to lysis may be wide, as in the case
of endolysin of Lactobacillus helveticus temperate bac-
teriophage u0303 [8]. There are also counterexamples:
endolysin of bacteriophage u3626 specifically degrades
only cell walls of the bacterium Clostridium perfringens
[9]. The selectivity of endolysin of bcateriophage C1
for streptococci allows one to forecast its use as a

agent against these bacteria colonizing the mucous
epithelium of the upper air passages [10]. The specific-
ity of another phage endolysin for Bacillus anthracis
suggests that it has potential for application in diag-
nostics [11]. Also, endolysins may act synergistically
with antibiotics [12]. All this makes endolysins poten-
tial candidates as antibacterial drugs. The lysozyme of
bacteriophage T4, for example, was successfully used
to protect potato from Bacillus subtilis [13].
The system of lytic proteins of bacteriophage T5 was
first found in the process of sequencing of the early
region of its genome, which is now deposited in the Gen-
Bank database (GenBank accession no. AY509815) [14].
Analysis of the phage’s primary DNA sequence revealed
hol and lys genes, which turned out to be located in the
same operon under a common promoter. A search of
protein databases for homologs of hol and lys gene
products provided evidence that they are probably
involved in the process of host cell lysis [14].
The objectives of the present work included preli-
minary biochemical characterization of the novel end-
olysin of bacteriophage T5, determination of the
peptidoglycan cleavage site, and analysis of endolysin
specificity for bacterial species.
Results
The hol and lys genes are located in the early C region
of the bacteriophage T5 genome under a common
promoter (Fig. 1). Endolysin (GenBank accession no.
AAS19387; UniProt accession no. Q6QGP7) is a
polypeptide that consists of 137 amino acids (expected

molecular mass of 15.266 kDa) and has a calculated
pK value of 8.32.
Gene cloning and protein purification
The lys gene was cloned into plasmid vector pET3a, and
the electrophoretic analysis showed that the
Escherichia coli clones selected induced synthesis of a
protein product of the expected size (about 15 kDa).
The sequence of the lys gene in the plasmid was checked
by sequencing both DNA strands. Inside the cell, all of
the protein product of lys gene was in a soluble state:
after centrifugation of cellular homogenate, the pellet
fraction did not contain the target protein, and this was
confirmed electrophoretically (data not shown).
Fig. 1. Organization of the bacteriophage T5 gene locus that carries genes coding for lytic proteins. Flags indicate promoters; T-like signs
indicate transcription terminators. Arrows indicate open reading frames. Numerals on the sides indicate the distance from the beginning of
the genome in bp.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7330 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
Endolysin of bacteriophage T5 was extracted from
the cells of the producer strain and purified to an elec-
trophoretically homogeneous state by ion exchange
chromatography on two consecutive columns, Toyo-
pearl DEAE 650M and phosphocellulose (Fig. 2;
Table 1). The activity of the final preparation was
calculated to be 8.38 · 10
3
units per mg of protein.
Storage
It was shown that, when stored in Tris ⁄ HCl buffer in
the presence of 1 mm EDTA at 4 °C, the protein

remained stable. BSA (1 mgÆmL
)1
), 20% glycerol or
0.1% Triton X-100 had a positive effect on the protein
storage: the enzyme activity did not decrease over
several months.
Optimal conditions for enzyme functioning
The activity of bacteriophage T5 endolysin was found
to depend strongly on the concentration of the buffer
(Fig. 3A). The enzyme showed maximal activity in
diluted Tris ⁄ HCl buffers (25–50 mm). With increasing
pH, the enzyme activity grew, and it reached its maxi-
mum at a pH of about 8.5 (Fig. 3B).
Other components of the buffer also affected the
lytic ability of bacteriophage T5 endolysin. We mea-
sured enzyme activity at two different pH values under
the chosen standard conditions, and compared it with
the activity observed upon the addition of various sub-
stances. As can be seen in Fig. 4, the enzyme is metal
ion-dependent, because it was inhibited by EDTA.
This was confirmed by the experiments in which a pure
peptidoglycan was used as the substrate. High buffer
concentrations and high divalent metal ion concentra-
tions inhibited the enzyme activity. Interestingly, the
Fig. 2. SDS ⁄ PAGE analysis of fractions from bacteriophage T5 end-
olysin purification steps. Lanes: 1, molecular weight markers; 2,
crude extract; 3, chromatography on Toyopearl 650M; 4, chroma-
tography on phosphocellulose. Wells 2–4 were loaded with 7 lgof
total protein each.
Table 1. Purification of recombinant endolysin of bacteriophage T5. Specific activity values represent the mean ± standard deviation (n ‡ 3).

Fraction
volume (mL)
Protein
concentration (mgÆmL
)1
)
Specific
activity (UÆmg
)1
)
Total
activity (U)
Purification
factor (fold) Yield (%)
Crude extract 10.0 6.0 1830 ± 58 109 800 1 100
Chromatography on Toyopearl 650M 33.0 0.8 4000 ± 105 105 600 2.2 96
Chromatography on phosphocellulose 6 1.7 8380 ± 140 85 480 4.6 78
Buffer concentration, mM
0 50 100 150 200 250
Enzyme activity, U*10
–3
0
1
2
3
4
Enzyme activity, U*10
–3
0
1

2
3
4
pH
5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
Fig. 3. (A) The effect of buffer concentration on enzyme activity.
The activity was measured in Tris ⁄ HCl (pH 8.2) containing 0.1% Tri-
ton X-100; the reaction was initiated by the addition of 0.015 lgof
enzyme. (B) Effect of pH on enzyme activity. The activity was mea-
sured in 50 m
M Tris ⁄ HCl buffer containing 0.1% Triton X-100; the
reaction was initiated by the addition of 0.015 lg of enzyme.
G. V. Mikoulinskaia et al.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7331
activity depended on the type of buffer used: in the
potassium phosphate buffer, the enzyme worked much
worse than in the Tris ⁄ HCl buffer, especially at high
pH values. Perhaps the catalytic function of endolysin
requires the presence of ions of some metals whose
phosphate salts are poorly soluble. We supposed that
the cations needed for endolysin functioning might
be Ca
2+
,Mg
2+
,orZn
2+
: these metals form low-solu-
bility phosphate complexes, whose solubility drops

further upon alkalization of the medium.
It is also significant that the enzyme is much more
active in the presence of 0.1% Triton X-100. Caldentey
and Bamford, [15] whose work suggested to us the idea
of using this soft nonionic detergent, considered that it
enhanced endolysin activity through action on the cell
wall (Triton X-100 probably makes peptidoglycan
more accessible or facilitates enzyme binding).
We also explored the effects of various cations on
the storage of endolysin. The presence of 10 mm
Mg
2+
,Ca
2+
or Mn
2+
in the storage medium was
shown to have no effect on the enzyme activity. Other
divalent (Zn
2+
and Cu
2+
) and trivalent (Fe
3+
and
Cr
3+
) cations, at a concentration of 10 mm, were
found to completely inactivate the enzyme immediately
after addition to the storage medium.

To determine which cations are necessary for the
enzyme to function, we used 0.1 mm metal chelators
with different affinity for metal ions: EDTA, which
binds many cations; 1,2-bis-(O-aminophenoxy)ethane-
N,N,N¢,N¢-tetraacetic acid (BAPTA), which selectively
binds Ca
2+
and, to a lesser extent, Mg
2+
,Zn
2+
[16], and
Fe
2+
[17]; 1,10-phenanthroline, which has high affinities
for Zn
2+
,Ni
2+
,Co
2+
, and Cd
2+
, and which is rather
effective in binding Mn
2+
, but is ineffective in binding
Mg
2+
and Ca

2+
[18,19]; and deferoxamine (desferriox-
amine) and Tiron (4,5-dihydroxy-1,3-benzene-disulfonic
acid), both of which are chelators of Fe
3+
.
EDTA (a broad-spectrum chelator) and BAPTA (a
specific chelator for Ca
2+
) completely inhibited the
activity of endolysin at a concentration of 0.1 mm,
whereas 0.1 mm 1,10-phenanthroline did not affect the
enzyme activity – although the binding constants of
phenanthroline for Zn
2+
and of BAPTA for Ca
2+
are
very close [19]. A 1-day incubation of the protein in
10 mm 1,10-phenanthroline did not decrease its activ-
ity, in contrast to what was observed with 10 mm
EDTA or BAPTA. It should also be noted that, in the
absence of chelators, 0.5 mm ZnCl
2
completely inacti-
vated the enzyme.
We tried to restore activity of the enzyme inhibited
with 0.1 mm EDTA or BAPTA by adding various salts
at concentrations of 0.1–1 mm (zinc, magnesium, man-
ganese and calcium chlorides). Among those salts, only

calcium and manganese chlorides were found to
restore the enzyme activity completely (Table 2). Zinc
chloride and, in the case of EDTA, magnesium chlo-
ride at low (0.1–0.25 mm) concentrations partially
restored enzyme activity, probably because of binding
of the chelator. Increasing the concentration of magne-
sium or zinc chloride to 0.5 mm resulted in inhibition
of the enzyme.
There was no inhibition of enzyme activity in the
presence of Tiron and deferoxamine; moreover, defe-
roxamine even enhanced the activity by 25%, probably
by chelating some endogenous cations that could
compete with Ca
2+
and thus affect enzyme activity.
Classification of the enzyme by the type of bond
hydrolyzed
To determine whether the enzyme studied was a glyco-
syl hydrolase, peptidase, or amidase, we compared its
hydrolytic action with that of egg white lysozyme. To
do this, we analyzed the quantity of reducing and
amino groups released after peptidoglycan hydrolysis
(to exclude the effect of pre-existing groups, peptido-
glycan was either acetylated or reduced prior to hydro-
lysis). The results of the assay are presented in
Table 3.
Pre-acetylation of free amino groups in peptidogly-
can practically eliminates staining for them in the
sample hydrolyzed by egg white lysozyme. This is not
surprising, because this enzyme is a glycosyl hydrolase,

not a peptidase. In contrast, the samples treated with
Fig. 4. The effect of reaction mixture composition on enzyme
activity. Standard conditions: 50 m
M Tris ⁄ HCl containing 0.1%
Triton X-100. K-P
i
is 50 mM potassium phosphate buffer containing
0.1% Triton X-100.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7332 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
bacteriophage T5 endolysin show the appearance of
free amino groups, suggesting that the enzyme has a
peptidase activity.
The peptide subunit of E. coli peptidoglycan is
l-alanoyl-d-glutamyl-l-diaminopimelinoyl-d-alanoyl-d-
alanine [20], and about one-third of the total number
of subunits are involved in the formation of inter-
subunit cross-bridges between diaminopimelic acid
and d-alanine [21]. A TLC analysis of ether-extracted
2,4-dinitrophenyl (DNF)–amino acids (see Experi-
mental procedures) showed the presence of DNF–
glutamic acid in the bacteriophage T5 hydrolysate
(Fig. 5). Accordingly, as revealed with an amino acid
analyzer, the material remaining after extraction
showed about 70% loss in the glutamic acid content
(Table 4).
Thus, we can conclude that the enzyme studied
hydrolyzes the bond between l-alanine and d-glu-
tamic acid; that is, it is an l-alanoyl-d-glutamate
peptidase.

Table 3. Release of amino and reducing groups during enzymatic
hydrolysis of E. coli peptidoglycan. Values represent the mean ±
standard deviation (n ‡ 3).
Substrate Enzyme
Reducing
groups
(nmolÆmg
)1
)
Amino
groups
(nmolÆmg
)1
)
Acetylated
peptidoglycan
T5 endolysin 23.06 ± 2.8 229.2 ± 13.3
Egg white lysozyme 36.09 ± 2.9 15.7 ± 0.6
Reduced
peptidoglycan
T5 endolysin 19.02 ± 1.6 550.6 ± 27.8
Egg white lysozyme 77.04 ± 5.9 382.0 ± 22.1
Fig. 5. Analysis of amino acids released by peptidase of bacterio-
phage T5. TLC was performed on Kieselgel 60 F
254
plates as
described in Experimental procedures. Lanes: 1, DNF–alanine
(Sigma); 2, DNF–glutamic acid (Sigma); 3, control peptidoglycan; 4,
peptidoglycan hydrolyzed by bacteriophage T5 peptidase; 5, pepti-
doglycan hydrolyzed by egg white lysozyme. The thin lines at the

bottom and the top indicate the initial and final eluent fronts,
respectively.
Table 2. Restoration of lytic activity of the enzyme by cations after
its inhibition with EDTA or BAPTA. The activity was measured turbi-
dimetrically (see Experimental procedures). Before measurements,
0.1 m
M EDTA or BAPTA and enzyme (0.015 lg) were added to the
cell suspension. The reaction was initiated by the addition of a chlo-
ride of the corresponding metal. The activity was calculated as a
percentage of the initial activity, measured in the absence of chela-
tors. Values represent the mean ± standard deviation (n ‡ 3). ND,
not determined.
Concentration
of the metal
ion (m
M)
Relative activity (%)
EDTA BAPTA
Control 0.000 2.8 ± 0.2 3.2 ± 0.2
Zn
2+
0.110 51.5 ± 4.1 47.0 ± 3.2
0.125 53.7 ± 4.5 42.9 ± 3.0
0.250 31.32 ± 1.7 21.9 ± 1.8
0.500 10.8 ± 0.8 5.5 ± 0.5
Mg
2+
0.110 11.7 ± 0.8 ND
0.125 14.1 ± 1.0 1.1 ± 0.2
0.250 26.4 ± 2.1 4.2 ± 0.3

0.500 6.5 ± 0.3 3.8 ± 0.3
Mn
2+
0.110 106.6 ± 2.9 105.6 ± 1.2
0.125 116.5 ± 5.4 106.9 ± 1.7
0.250 111.5 ± 4.9 122.9 ± 8.4
0.500 91.9 ± 5.7 101.1 ± 7.5
1.000 68.1 ± 5.7 66.2 ± 5.4
Ca
2+
0.110 120.5 ± 13.6 92.1 ± 7.8
0.125 117.6 ± 8.3 103.0 ± 9.5
0.250 84.6 ± 6.9 101.6 ± 8.7
0.500 79.5 ± 6.0 95.8 ± 6.3
1.000 69.8 ± 10.8 90.8 ± 10.0
Table 4. Content of Glu relative to Ala in the samples of enzymati-
cally hydrolyzed peptidoglycan of Ps. putida. The contents of
individual amino acids were determined after acidic hydrolysis of
peptidoglycan samples using an amino acid analyzer (see
Experimental procedures).
Component
Peptidoglycan
Untreated
Treated
with egg
white
lysozyme
Treated with
bacteriophage
T5 endolysin

Ala 1.00 1.00 1.00
Glu 0.57 0.54 0.17
G. V. Mikoulinskaia et al.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7333
Amino acid sequence analysis
A psi-blast analysis of bacteriophage T5 l-alanoyl-
d-glutamate peptidase showed that there are many
similar proteins, of both phage and bacterial origin
(217 hits after the third iteration). The best hits with
low E-values ( 10
)45
–10
)30
) were sequences of pro-
teins from Gram-negative bacteria of the genera Yer-
sinia, Shigella, Escherichia, Photorhabdus, Providencia,
Agrobacterium, Proteus, Serratia, Hahella, Haemophi-
lus, Chromobacterium, Vibrio, Vibrionales, Aliivibrio,
Brevundimonas, Erwinia, and Bordetella, as well as
bacteriophage proteins. In the E-value region of 10
)27
–
10
)20
, one can find protein sequences of Gram-positive
bacteria. On the basis of multiple alignment of 53
sequences, which were selected among 95 of the best
hits after removing repetitions, a phylogenetic tree was
constructed under conditions of minimal evolution

using the program mega v. 4.0 (Fig. 6). On this tree,
the proteins of enterobacteria and close species, as well
as their bacteriophages, form a branch separated
from the branches of Gram-positive bacteria and their
bacteriophages. Figure 7 shows an alignment of seven
sequences constructed with the program clustalx.
The sequences belong to the branch of proteins that
Phage ST64T
Phage PS3
Proteus mirabilis
Yersinia bercovieri
Phage RB43
Phage T5
Phage EPS7
Haemophilus influenzae
Brevundimonas sp.
Phage phiEcoM
Chloroherpeton thalassium
Aliivibrio salmonicida
Vibrio fischeri
Phage Xp15
Bordetella bronchiseptica
Bordetella parapertussis
Photorhabdus asymbiotica
Erwinia tasmaniensis
Providencia stuartii
Chromobacterium violaceum
Yersinia enterocolitica
Yersinia mollaretii
Yersinia pseudotuberculosis

Yersinia pestis Pestoides F
Proteus penneri
Vibrio splendidus
Vibrionales bacterium
Phage PY100
Yersinia frederiksenii
Agrobacterium tumefaciens
Hahella chejuensis
Phage phiJL001
Phage RB49
Phage Phi1
Serratia proteamaculans
Phage phiP27
Shigella boydii
Escherichia
coli
Phage A500
Listeria innocua
Phage A006
Phage P35
Phage SPO1
Exiguobacterium sibiricum
Geobacillus sp.
Geobacillus kaustophilus
Anoxybacillus flavithermus
Clostridium acetobutylicum
Exiguobacterium sp.
Bacillus cereus
Phage B025
Listeria monocytogenes

Phage A118
100
100
100
96
100
100
99
99
87
99
99
95
88
90
88
87
85
84
79
83
79
78
73
29
70
53
50
29
37

36
36
26
22
20
20
19
17
4
9
4
2
2
40
37
17
47
15
41
0.1
Fig. 6. Phylogenetic tree of amino acid sequences of bacterial and
phage endolysins constructed by the minimum evolution method.
Numbers at the nodes represent the bootstrap values with 500 rep-
lications. The scale bar indicates an evolutionary distance of
0.1 amino acid substitutions per site. Bootstrap values are indicated
above or below the branches. GenBank accession numbers are as
follows: bacteriophage ST64T, NP_720320.1; bacteriophage PS3,
CAA09701.1; P. mirabilis, YP_002151708.1; Y. bercovieri, ZP_
00822433.1; bacteriophage RB43, YP_239135.1; bacteriophage T5,
YP_006868.1; bacteriophage EPS7, YP_001836966.1; Haemophi-

lus influenzae, YP_248988.1; Brevundimonas sp., YP_002588905.1;
bacteriophage phiEcoM, YP_001595416.1; Chloroherpeton thalas-
sium, YP_001995193.1; Aliivibrio salmonicida, YP_002262290.1;
Vibrio fischeri, YP_205400.1; bacteriophage Xp15, YP_239293.1;
Bordetella bronchiseptica, NP_889694.1; Bordetella parapertussis,
NP_885037.1; Photorhabdus asymbiotica, CAR67777.1; Erwinia tas-
maniensis, YP_001907932.1; Providencia stuartii, ZP_02961079.1;
Chromobacterium violaceum, NP_903215.1; Yersinia enterocolitica,
AAT90759.1; Yersinia mollaretii, ZP_00825275.1; Yersinia pseudotu-
berculosis, YP_001399405.1; Yersinia pestis pestoides F, YP_
001161675.1; Proteus penneri, ZP_03801884.1; Vibrio splendidus,
ZP_00991905.1; Vibrionales bacterium, ZP_01814817.1; bacterio-
phage PY100, CAJ28446.1; Yersinia frederiksenii, ZP_00828831.1;
Agrobacterium tumefaciens, NP_353494.2; Hahella chejuensis, YP_
435691.1; bacteriophage phiJL001, YP_224014.1; bacteriophage
RB49, NP_891673.1; bacteriophage Phi1, YP_001469446.1; Serra-
tia proteamaculans, YP_001471697.1; bacteriophage phiP27, NP_
543082.1; Shigella boydii, YP_001880486.1; E. coli, ZP_03049236.
1; bacteriophage A500, YP_001468411.1; Listeria innocua, NP_
469473.1; bacteriophage A006, YP_001468860.1; bacteriophage
P35, YP_001468812.1; bacteriophage SPO1, YP_002300379.1;
Exiguobacterium sibiricum, YP_001814297.1; Geobacillus sp., ZP_
02914525.1; Geobacillus kaustophilus, YP_145852.1; Anoxybacillus
flavithermus, YP_002315045.1; Clostridium acetobutylicum, NP_
347645.1; Exiguobacterium sp., ZP_02992216.1; Bacillus cereus,
YP_002446097.1; bacteriophage B025, YP_001468664.1; L. mono-
cytogenes, ZP_03669234.1; bacteriophage A118, NP_463486.1.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7334 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
are the closest to bacteriophage T5 peptidase; they are

encoded by the genomes of bacteriophages ST64T,
PS3, phiEcoM, and RB43, and the bacteria Yer-
sinia bercovieri and Proteus mirabilis. In multialign-
ment, we did not use the bacteriophage EPS7 protein
sequence, which is the most similar to the sequence of
bacteriophage T5 peptidase, because it seems to be
shortened. The high similarity of two bacterial proteins
(from Y. bercovieri and P. mirabilis) with the phage
proteins could indicate their coding by prophages. As
a result of the analysis, we found the amino acid
sequence motifs HXXXXXXD and DXXH, which
are typical for peptidases of the C subfamily of
family M15 ( The former
sequence is a part of the phage consensus sequence
SK(R)HI(L,M)T(S)GD(N)AI(V,L)DI(L,F)I(L,A,Y),P,
consisting of 13 amino acids, six of which are identical
and the others of which are highly conserved.
Testing the lytic ability of endolysin on
heterologous microorganisms
To assess the spectrum of bacteriolytic action of bacte-
riophage T5 endolysin, we tested its ability to hydro-
lyze various substrates prepared from the cells of
selected bacteria according to a standard protocol
(Table 5). Among the bacteria selected were Gram-
positive and Gram-negative species, which differed in
the structure of peptidoglycan and the composition of
the cell wall.
All of the tested Gram-negative bacteria were sub-
jected to rapid lysis. Among the Gram-positive cells
examined, only B. subtilis cells were lysed relatively

well; however, the rate of their lysis was about 10
4
-fold
lower than that of Gram-negative cells. In Listeria
cells, the rate of lysis was an order of magnitude lower
than that in bacilli (Table 5). The rest of the Gram-
positive cells were not lysed.
The specificity of bacteriophage T5 endolysin
towards Gram-negative bacteria suggests that it might
be used as a selective antibacterial agent. However, to
make peptidoglycan accessible to the enzyme, it would
be necessary to perturb the outer cell membrane. We
therefore used polymyxin B, an antibiotic that can
bind phospholipids of the outer membrane and destroy
membrane integrity, as a putative destructive agent.
Figure 8 shows that, at a concentration of 40 lgÆmL
)1
,
polymyxin B inhibited the growth of cells but did not
lyse them (the attenuance of the cell mixture did not
decrease). At the same time, endolysin with polymyxin
B lysed the cells completely. The nontreated and end-
olysin-treated E. coli cells grew well on a nutrient agar
medium, forming a dense bacterial lawn (Fig. 8A,D);
treatment of cells with polymyxin B resulted in difficul-
ties in cell growth (Fig. 8B); and combined endoly-
sin ⁄ polymyxin B treatment led to complete lysis of all
living cells – there was no further growth (Fig. 8C).
Discussion
In this study, we cloned the gene of a novel enzyme

and then purified and characterized this enzyme, which
turned out to be endolysin of bacteriophage T5, a
component of the phage cell lysis system. Our experi-
Fig. 7. Multiple alignment of protein
sequences of bacteriophages T5, ST64T,
PS3d, RB43, and phiEcoM, and the bacteria
Y. bercovieri and P. mirabilis, constructed
using the program
CLUSTAL X. Amino acids
common to all the sequences are marked in
gray; conservative amino acids, which par-
ticipate in metal ion binding and catalysis,
are marked in black.
G. V. Mikoulinskaia et al.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7335
ments have shown that the protein is a metal ion-
dependent l-alanoyl-d-glutamate peptidase. The
enzyme is strongly inhibited by EDTA and BAPTA,
and completely reactivated by Ca
2+
and Mn
2+
.
Considering BAPTA as a specific chelator of Ca
2+
,we
calculated the concentration of free Ca
2+
in the med-

ium, using the program cabuf, and concluded that, to
cleave peptidoglycan, the enzyme would require free
Ca
2+
at a concentration of 0.05–0.4 mm. Probably,
in vitro,Ca
2+
can be replaced with Mn
2+
.
It is known that bacteriophage T5 requires 0.1 mm
Ca
2+
to produce phage progeny in E. coli cells [22].
Calcium ions play a key role at certain stages of phage
development: upon infection, when the phage DNA
penetrates into cells [22]; and during the synthesis of
phage RNA [23] and proteins [24]. Interestingly, the
enzyme that is directly involved in the final stage of
development (lysis) is also activated by the same con-
centrations of Ca
2+
. There is another example of such
activation: aminopeptidase A (EC 3.4.11.7), a Zn
2+
-
containing enzyme (gluzincin) of metallopeptidase
family M1, is also activated by Ca
2+
and can be reac-

tivated by Ca
2+
and Mn
2+
[25]. The Ca
2+
, which
binds to the enzyme through aspartic acids [26,27],
contributes to its substrate specificity: it forms a bridge
between an aspartic acid of the enzyme and an acidic
N-terminal amino acid of the substrate [26]. The affin-
ity of bacteriophage T5 endolysin for d-glutamic acid
of peptidoglycan and the ability of endolysin to be
activated by Ca
2+
may be related as well. On the basis
of data on enzyme metal ion dependence, we can
include l-alanoyl-d-glutamate peptidase of bacterio-
phage T5 in the sub-subclass of metalloendopeptidases
[EC 3.4.24 (probable); the enzyme can be listed in EC
after publication of evidence that it catalyzes this reac-
tion)]. The analysis of the primary amino acid
sequence of bacteriophage T5 peptidase revealed con-
served amino acids (His66, Asp73, Asp130, and
His133) that are typical for metallopeptidases of the
M15 family and take part in the binding of the metal
ion in the process of catalysis [28].
This is the first example of an l-alanoyl-d-glutamate
peptidase to be found in a virulent phage infecting
Gram-negative bacteria. Enzymes of this class, Ply118

and Ply500, were first found in two temperate phages,
A118 and A500, which infect Gram-positive rods of
the Listeria genus [29]. Since then, there has been only
one l-alanoyl-d-glutamate peptidase – from the Gram-
positive bacteria B. subtilis – whose existence has been
proved biochemically [30]. It is interesting that Listeria
and Bacillus are not close relatives of E. coli, but their
peptidoglycan is also of the A1c type. Ply118 and
Ply500 show rather high substrate specificities: apart
from Listeria, they affect only three species of the
Bacillus genus (which also have peptidoglycan of the
A1c type) [29]. l-Alanoyl-d-glutamate peptidase of
bacteriophage T5 is similar to the N-terminal, enzy-
matically active domain of Ply118 (25% identity). The
C-terminal domain of Ply118 is responsible for sub-
strate recognition [31]. The enzyme of bacteriophage
T5 is much shorter than Ply118 (137 and 281 amino
acids, respectively); evidently, in bacteriophage T5 end-
olysin, the functions of substrate binding and catalysis
Table 5. The effect of bacteriophage T5 peptidase on heterologous
microorganisms. The rate of cell lysis was measured turbidimetri-
cally (see Experimental procedures), using autoclaved bacterial cells
as a susbstrate. Values represent the mean ± standard deviation
(n ‡ 3).
Organism Relative rate of lysis (UÆmg
)1
enzyme)
E. coli K-12 (1.12 ± 0.12) · 10
4
Pe. carotovorum (1.01 ± 0.15) · 10

4
Ps. putida (1.35 ± 0.18) · 10
4
P. vulgaris (1.19 ± 0.20) · 10
4
P. mirabilis (1.04 ± 0.15) · 10
4
B. subtilis 0.48 ± 0.05
L. monocytogenes 0.016 ± 0.001
S. aureus 0.0
C. xerosis 0.0
M. luteus 0.0
AB
CD
Fig. 8. Analysis of E. coli cell viability after endolysin action. Frag-
ments of plates containing cells preincubated with: (A) pure endoly-
sin (40 lgÆmL
)1
); (B) polymyxin B (40 lgÆmL
)1
); (C) polymyxin B
(40 lgÆmL
)1
) and pure endolysin (40 lgÆmL
)1
); and (D) control cells.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7336 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
are carried out by a single domain. Interestingly, a
recent crystallographic examination of the Ply500

enzyme [28] showed the presence of Zn
2+
in the active
center – a phenomenon that is typical for a number of
proteins of the peptidase M15 family [32]. The inhibi-
tion of bacteriophage T5 peptidase by the specific
Ca
2+
chelator BAPTA and by ZnCl
2
does not agree
with those data, and nor does the absence of an inhibi-
tory effect of 1,10-phenanthroline, which has a high
affinity for zinc. However, the above-mentioned Zn
2+
-
containing Ca
2+
-activated aminopeptidase A is inhib-
ited by Zn
2+
too [33], and such a feature can also be
found among other Zn
2+
-containing metalloenzymes
[34]. In addition, the failure to observe inhibition with
a particular chelating agent need not be an absolute
gauge of the absence of Zn
2+
[19]. Perhaps the

question of which cation is bound at the active center
of bacteriophage T5 endolysin will be answered
after crystallographic examination. Nevertheless, the
stimulatory effect of Ca
2+
on the enzyme is evident.
The blast analysis shows that there are more than
100 proteins – both from phages and from bacteria
(mainly Gram-negative bacteria) – that are much
more similar to bacteriophage T5 endolysin than
Ply500 and Ply118. Probably, all of them are l-ala-
noyl-d-glutamate peptidases and have a common ori-
gin, and bacteriophage T5 peptidase is a typical
enzyme of the bacterial cell lysis system. It is possi-
ble that such a wide distribution of endolytic l-ala-
noyl-d-glutamate peptidases is related to the frequent
occurrence of the l-alanine–d-glutamate bond in pep-
tidoglycan.
The hypothesis that phage l-alanoyl-d-glutamate
peptidases have a common origin is supported by the
results of comparison of phage holin sequences. In
most cases, holin sequences demonstrate little similar-
ity to each other, and these proteins are considered to
have evolved independently of endolysins [35]. How-
ever, holin of bacteriophage T5 shows significant
similarity to holins of bacteriophages RB43 (E =
3 · 10
)16
) and RB49 (E =9· 10
)21

), as does bacte-
riophage T5 l-alanoyl-d-glutamate peptidase towards
RB43 and RB49 putative endolysins (E =7· 10
)19
and E =2· 10
)16
, respectively). It is interesting that
bacteriophages RB43 and RB49 belong to the group
of pseudo T-even phages. Holin of bacteriophage T4,
a product of gene t, also resembles holin of bacterio-
phage T5 (E =6· 10
)13
). However, endolysin E of
bacteriophage T4 is a muramidase, and has no relation
to bacteriophage T5 peptidase. It is possible that there
might have been horizontal gene transfer between
pseudo-T-even phages and bacteriophage T5. It should
be noted that in bacteriophages RB43, RB49, and T4,
endolysin and holin are located in different genome
regions – not in the same operon under a common
promoter, as in bacteriophage T5. It can be supposed,
therefore, that the variant present in bacteriophage T5
(colocalization of the lytic system genes) brings more
evolutionary advantages in terms of their coordinated
expression.
In addition to holin and endolysin, some dsDNA
phages have another pair of proteins that are involved
in cell lysis and provide a competitive advantage to the
phage under unfavorable growth conditions (analogs
of the products of the Rz and Rz1 genes of bacterio-

phage k) [36]. One of these proteins is a lipoprotein,
and the other is a transmembrane protein; defects in
their genes result in Mg
2+
-dependent disorders in the
process of lysis. Analogs of these proteins have
recently been found in quite a large number of phages,
including T5, T4, RB32, RB43, RB49, and RB69 [36].
All of these proteins are located separately from holin
and endolysin. The products of genes T5p045 (Rz ana-
log) and T5p044 (Rz1 analog) show a slight resem-
blance to the products of genes PseT.3 and PseT.2 of
bacteriophages T4 and RB32. Interestingly, the operon
in which genes T5p045 and T5p044 are located in is
the upstream neighbor of the holin–endolysin gene
locus (Fig. 1), but the coding sequences of these genes
are included in the early transcript. At the same time,
the bacteriophage T5 holin–endolysin promoter is
probably late, although it is located in the early gen-
ome region. It contains more GC pairs, particularly in
the region +10 to +20, which is typical for late
promoters of bacteriophage T5 [37]; in addition, the
conservative region )33 contains a TTnAnA sequence
(typical for late bacteriophage T5 promoters) and does
not contain TTGCTn, which is a sign of early promot-
ers [38]. It should be noted that holin T and endolysin
E of bacteriophage T4 are late proteins, although gene
e is located in the early region [39].
To assess the spectrum of bacteriolytic action of
bacteriophage T5 endolysin, we tested the bacteria of

Gram-positive and Gram-negative species, which differ
in the structure of peptidoglycan and the composition
of the cell wall. For example, Gram-negative cells con-
tains peptidoglycan of the A1c type (not amidated,
with c-mesodiaminopimelic acid in the third position);
in B. subtilis, the peptidoglycan is of the same type but
amidated; Listeria monocytogenes also has peptido-
glycan of the A1c type, but its cell wall structure has
some features (for instance, teichoic acid of the Liste-
ria cell wall contains two substituents, N-acetylglucos-
amine and dirhamnosyl, whereas teichoic acids of the
other Gram-positive bacteria include only one type of
substituting group) [40]; Staphylococcus aureus has an
G. V. Mikoulinskaia et al. L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7337
A3a-type peptidoglycan (with l-lysine in the third
position); the type of peptidoglycan of Micrococcus
luteus is A2 (l-lysine in the third position; teichuronic
acids instead of teichoic acids); and the cell walls of
corynebacteria contain lipids, like those of Gram-nega-
tive microorganisms [20]. The action of bacteriophage
T5 endolysin seemed to be directed towards peptido-
glycan type A1c, although the cell wall composition of
Gram-positive bacteria (L. monocytogenes and B. sub-
tilis) protected them from rapid lysis. Thus, our preli-
minary results suggest that the peptidase of
bacteriophage T5 is specific for the cell walls of Gram-
negative microorganisms, containing A1c peptidogly-
can and lacking either teichoic or teichuronic acids,
which are typical of the cell walls of the Gram-positive

bacteria.
Some Gram-negative microorganisms are pathogenic
for animals and plants (Pseudomonas, Klebsiella, Pro-
teus, and Agrobacterium), making bacteriophage T5
peptidase a potential candidate for the development of
drugs that could be of use in biotechnology and plant
cultivation.
Successful application of bacteriophage T5 endolysin
as a selective lytic agent against Gram-negative cells
requires the presence of factors that disturb outer
membrane permeability. There are a number of these:
EDTA, sodium tripolyphosphate, heat, pH [41], ultra-
sound [42], peptidolipids [43], and polymyxin B [44]. In
this work, we have demonstrated that E. coli cell lysis
by bacteriophage T5 endolysin is possible after poly-
myxin B treatment. This constitutes an example prov-
ing the effectiveness of this approach, and other
permeabilizing agents could also be applicable.
Experimental procedures
Materials
E. coli strains B, Z85 and BL21(DE3) and bacteriophage
T5
+
were taken from the collection of the Laboratory of
Molecular Microbiology of the Institute of Biochemistry
and Physiology of Microorganisms (IBPM RAS). Strains
of E. coli K12, B. subtilis, M. luteus, Pseudomonas putida,
S. aureus, Pectobacterium carotovorum, Proteus vulgaris,
P. mirabilis, L. monocytogenes and Corynebacterium xerosis
were obtained from the All-Russian Collection of Micro-

organisms (IBPM RAS). Plasmid pET3a was provided
by Novagen (Madison, WI, USA). Bacteria and phages
were grown either in liquid LB broth or on agarized LB
medium. Selection of clones was performed on plates with
ampicillin (50 lgÆmL
)1
). Crystalline egg white lysozyme was
purchased from Serva (Heidelberg, Germany). Restriction
endonucleases were from Fermentas (Vilnius, Lithuania).
All other chemicals were purchased, unless otherwise stated,
from either ICN (Irvine, CA, USA) or Sigma (St Louis,
MO, USA).
Cloning of the lys gene
The lys gene of bacteriophage T5 was amplified by PCR
with primers LysF (5¢-gtcgagacATATGAGTTTTAAAT
TTGGT-3¢) and LysR (5¢-ctggatccATTAAACTAGTTCG
ACATG-3¢), which contain sites hydrolyzed by restriction
endonucleases NdeI and BamHI. The PCR fragment was
cloned into plasmid vector pET3a (into the region con-
trolled by the promoter of gene 10 of bacteriophage T7),
using standard molecular biology techniques. The clones
carrying the insert were selected after treatment with restric-
tion endonucleases and electrophoresis in 1% agarose. The
construct obtained was named pT5lys. Plasmid pT5lys was
further used to transform cells of E. coli strain BL21(DE3).
The synthesis of endolysin was induced with 0.5 mm isopro-
pyl-thio-b-d-galactoside at a culture density corresponding
to attenuance D
550 nm
= 1.0; the cells were harvested by

centrifugation (6000 g, 10 min) 2.5 h later.
Isolation and purification of endolysin
Cells of E. coli BL21(DE3) from 200 mL of culture (1.1 g)
carrying plasmid pT5lys were suspended in 10 mL of
25 mm Tris ⁄ HCl (pH 8.0), containing 40 mm NaCl and
1mm EDTA, and disrupted by sonication for 1 min (two
30 s treatments at a power of 75 W). The suspension was
centrifuged at 20 000 g for 30 min. The supernatant
(9.5 mL) was passed through an 11.4 mL column with
Toyopearl DEAE 650M (TosoHaas, Stuttgart, Germany)
and then applied to a 10 mL phosphocellulose column
equilibrated with the same buffer. Proteins were eluted by a
linear gradient of sodium chloride (0.05–0.50 m)in25mm
Tris ⁄ HCl buffer (pH 8.0) containing 1 mm EDTA (total
volume, 100 mL). Fractions (2 mL) were analyzed by
PAGE in a 15% polyacrylamide gel. The target protein was
eluted with 0.3 m NaCl.
Enzyme activity assay
The substrate for enzyme activity assay was prepared as
follows. An overnight culture of E. coli B cells was treated
with chloroform (added to 5% of total volume) for 15 min,
and the cells were then were washed twice with water and
stored frozen. Just before the measurement, the cells were
suspended in a reaction buffer (50 mm Tris ⁄ HCl, pH 8.2,
containing 0.1% Triton X-100). The activity was deter-
mined spectrophotometrically, by the decrease of attenu-
ance at 450 nm, in 1 cm acrylic cuvettes at room
temperature. An activity unit was defined as the quantity of
enzyme that provides the rate of attenuance decrease of 1.0
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.

7338 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
optical unitÆmin
)1
. All activity data were calculated from
three independent measurements.
Inhibition of enzyme activity by chelators
The substrate-providing cells were suspended in the reaction
buffer (50 mm Tris ⁄ HCl, pH 8.2, containing 0.1% Triton
X-100), and the chelators were added. The reaction was ini-
tiated by the addition of a standard amount of the enzyme
(15 ng mL
)1
).
Extraction of peptidoglycan
Peptidoglycan was extracted by Streshinskaya’s method
[45]. E. coli or Ps. putida cells (7 g, wet weight) were resus-
pended in 14 mL of 10 mm Tris-HCl (pH 8.0), and 0.4 mg
of DNase was then added. After addition of 20 mL of 4%
SDS, the mixture was boiled for 30 min and kept at 4 °C
for 12 h. The suspension was then sonicated (75 W, 1 min)
and centrifuged at 20 000 g for 1 h. The pellet was resus-
pended in 14 mL of water and, after addition of 4% SDS
(20 mL), the suspension was boiled for 15 min and
incubated at 4 °C for 12 h. After centrifugation at 20 000 g
for 1 h, the pellet was washed three times with distilled
water, resuspended in 10 mm Tris ⁄ HCl (pH 8.0), treated
with trypsin (100 lgÆmL
)1
) for 3 h at 37 °C, boiled for
15 min, and centrifuged at 20 000 g for 1 h. The pellet was

resuspended in 0.5 m trichloroacetic acid (1 : 100, v ⁄ v) and
incubated at 4 °C under stirring for 48 h; this was followed
by centrifugation at 15 000 g for 30 min. The pellet was
resuspended in 0.5 m trichloroacetic acid (1 : 100, v ⁄ v),
incubated at 37 °C for 48 h, and centrifuged at 15 000 g
for 30 min. The pellet was washed five times with water
and lyophilized.
Cell walls were acetylated with acetic anhydride as
described previously [46]. The oxidized redox groups were
reduced with NaBH
4
by Ward’s method [47].
Peptidoglycan hydrolysis
The substrate was acetylated or reduced peptidoglycan from
E. coli (2 mgÆmL
)1
); enzymes (egg white lysozyme or bac-
teriophage T5 endolysin) were added at a concentration
0.1 lgÆmL
)1
; the reaction was performed in 0.05 m Tris ⁄ HCl
buffer (pH 8.2) at room temperature. After hydrolysis, the
samples were centrifuged at 8000 g for 5 min, and the super-
natant was assayed for reducing and amino groups released.
Determination of reducing groups
The quantity of reducing groups was determined by Park
and Johnson’s method [48]. A 40 lL aliquot of the syspen-
sion of hydrolyzed peptidoglycan (see above) was placed in a
test-tube, and this was followed by the addition of ferricya-
nide (0.5 g of potassium ferricyanide per liter, stored in a

dark glass) and carbonate cyanide (5.3 g of sodium carbon-
ate and 0.65 g of KCN per liter) reagents (200 lL each). The
sample was boiled in a water bath for 15 min, and 1 mL of
the ferric iron solution (1.5 g of ferric ammonium sulfate and
1 g of SDS per liter of 0.025 m H
2
SO
4
) was added. After
15 min, absorbance was measured at 690 nm. The quantity
of reducing groups was calculated using a calibration curve
constructed for glucose. The control was the same acety-
lated ⁄ reduced peptidoglycan, except that it was not hydro-
lyzed but was rather supplemented with 0.1 lgÆmL
)1
bacteriophage T5 endolysin just before the measurements.
Determination of free NH
2
groups
The quantity of free NH
2
groups was measured according to
Ghuysen’s protocol [49]. An ethanol solution of 2,4-dinitro-
fluorobenzene (13 lL per 1 mL of ethanol) was added to a
sample of cell wall hydrolysate (1 : 10, v ⁄ v). The mixture was
kept in a water bath at 60 °C for 30 min. Then, four volumes
of 2 m HCl were added, and absorbance was measured at
420 nm. The quantity of free amino groups was calculated
from an alanine-based calibration curve. The control was the
same acetylated ⁄ reduced peptidoglycan, except that it was

not hydrolyzed but rather supplemented with 0.1 lgÆmL
)1
bacteriophage T5 endolysin just before the measurements.
Determination of enzyme specificity of
bacteriophage T5 endolysin
Peptidoglycan from Pseudomonas sp. (2 mg) was suspended
in 1 mL of 15 mm Tris ⁄ HCl (pH 8.2). Three 200 lL aliquots
were taken: the first was supplemented with bacteriophage
T5 peptidase (0.2 lg); the second was supplemented with
the same amount of egg white muramidase; and the
third was used as a control. After 24 h, the samples were
treated with dinitrofluorobenzene, and then subjected to
acid decomposition and ether extraction, as described above
[49]. DNF derivatives of amino acids were analyzed by
TLC on Kieselgel 60 F
254
plates (Merck, Darmstadt,
Germany) in a system of chloroform ⁄ methanol ⁄ benzyl
alcohol ⁄ concentrated ammonia ⁄ water (30 : 30 : 30 : 6 : 2)
[49]. The qualitative and quantitative analysis of amino
acids remaining in the solution was performed with a
Microtechna T-339 amino acid analyzer (Microtechna,
Prague, Czech Republic).
Testing the lytic ability of endolysin on
heterologous microorganisms
Cells of overnight cultures of bacteria were killed by autoclav-
ing, washed, and suspended in the reaction buffer (50 mm
Tris ⁄ HCl, pH 8.2, containing 0.1% Triton X-100) to a con-
centration of  2 · 10
8

cellsÆmL
)1
. The reaction was initiated
G. V. Mikoulinskaia et al. L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7339
by addition of enzyme (15 ng for the Gram-negative cells;
15 lg for the Gram-positive cells). The lytic activity was
determined by the decrease in attenuance at 450 nm, in 1 cm
acrylic cuvettes at room temperature, as described above.
Analysis of viability of endolysin-treated E. coli
cells
Cells of the E. coli overnight culture were suspended in the
reaction buffer (50 mm Tris ⁄ HCl, pH 8.2, containing 0.1%
Triton X-100) to a concentration of  2 · 10
8
cellsÆmL
)1
.
The first sample was supplemented with polymyxin B (final
concentration, 40 lgÆmL
)1
); the second sample was supple-
mented with pure endolysin (final concentration,
40 lgÆmL
)1
); and the third sample was supplemented with
both polymyxin B and endolysin at the same concentra-
tions. Untreated cells were used as a control. All samples
were incubated for 24 h at 37 °C, and the attenuance was
then measured. Aliquots corresponding to 10

7
initial cells
were grown on LB agar plates overnight.
Standard analytical techniques
The concentration of protein was determined by the Brad-
ford method [50], using egg white lysozyme as the standard.
Protein samples were analyzed by denaturing electrophoresis
in a 12% polyacrylamide gel by the Laemmli method [51],
using a standard kit of marker proteins containing b-lactal-
bumin (14.2 kDa), soybean trypsin inhibitor (20.1 kDa),
carboanhydrase (29.0 kDa), ovalbumin (45.0 kDa), BSA
(66.0 kDa), and phosphorylase B (94.0 kDa). Electrophore-
sis was performed for 1.0 h at room temperature (field
intensity, 15 VÆcm
)1
). Gels were stained with Coomassie
Brilliant Blue G-250 (Sigma, St Louis, MO, USA) and
washed with distilled water.
Software
Nucleotide and amino acid sequences were analyzed using
gene runner (v. 3.0) (Hastings Software, Inc., Hastings,
NY, USA) and clustalx software [52]. For analysis of
amino acid sequences, we used the blast search program [53]
from the server of the National Center of Biotechnology
Information (National Library of Medicine, USA; http://
www.ncbi.nlm.nih.gov/blast/). Free and total cation concen-
trations in the presence of multiple ligands were calculated
by the cabuf program ( />buf). The phylogenetic analysis was conducted with the aid
of mega4 [54].
References

1 Young R (1992) Bacteriophage lysis: mechanism and
regulation. Microbiol Rev 56, 430–481.
2 Bernhardt TG, Wang IN, Struck DK & Young R
(2002) Breaking free: ‘protein antibiotics’ and phage
lysis. Res Microbiol 153, 493–501.
3 Xu M, Struck DK, Deaton J, Wang IN & Young R
(2004) A signal–arrest–release sequence mediates export
and control of the phage P1 endolysin. Proc Natl Acad
Sci USA 101, 6415–6420.
4 Borysowski J, Weber-Dabrowska B & Go
´
rski A (2006)
Bacteriophage endolysins as a novel class of antibacte-
rial agents. Exp Biol Med (Maywood) 231, 366–377.
5 Loessner MJ (2005) Bacteriophage endolysins – current
state of research and applications. Curr Opin Microbiol
8, 480–487.
6 Pritchard DG, Dong S, Baker JR & Engler JA (2004)
The bifunctional peptidoglycan lysin of Streptococ-
cus agalactiae bacteriophage B30. Microbiology 150,
2079–2087.
7Du
¨
ring K, Porsch P, Mahn A, Brinkmann O & Gieffers
W (1999) The non-enzymatic microbicidal activity of
lysozymes. FEBS Lett 449, 93–100.
8 Deutsch SM, Guezenec S, Piot M, Foster S & Lortal S
(2004) Mur-LH, the broad-spectrum endolysin of Lacto-
bacillus helveticus temperate bacteriophage phi-0303.
Appl Environ Microbiol 70, 96–103.

9 Zimmer M, Vukov N, Scherer S & Loessner MJ (2002)
The murein hydrolase of the bacteriophage phi3626 dual
lysis system is active against all tested Clostridium
perfringens strains. Appl Environ Microbiol 68,
5311–5317.
10 Nelson D, Loomis L & Fischetti VA (2001) Prevention
and elimination of upper respiratory colonization of
mice by group A streptococci by using a bacteriophage
lytic enzyme. Proc Natl Acad Sci USA 98, 4107–4112.
11 Schuch R, Nelson D & Fischetti VA (2002) A bacterio-
lytic agent that detects and kills Bacillus anthracis.
Nature 418, 884–889.
12 Djurkovic S, Loeffler JM & Fischetti VA (2005)
Synergistic killing of Streptococcus pneumoniae with the
bacteriophage lytic enzyme Cpl-1 and penicillin or
gentamicin depends on the level of penicillin resistance.
Antimicrob Agents Chemother 49, 1225–1228.
13 de Ahrenholtz I, Harms K, de Vries J & Wackernagel
W (2000) Increased killing of Bacillus subtilis on the
hair roots of transgenic T4 lysozyme-producing
potatoes. Appl Environ Microbiol 66, 1862–1865.
14 Mikoulinskaia GV, Zimin AA, Feofanov SA &
Miroshnikov AI (2004) Identification, cloning,
and expression of bacteriophage T5 dnk gene
encoding a broad specificity deoxyribonucleoside
monophosphate kinase (EC 2.7.4.13). Protein Expr
Purif 33, 166–175.
15 Caldentey J & Bamford DH (1992) The lytic enzyme of
the Pseudomonas phage phi 6. Purification and biochemi-
cal characterization. Biochim Biophys Acta 1159, 44–50.

L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7340 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS
16 Aballay A, Sarrouf MN, Colombo MI, Stahl PD &
Mayorga LS (1995) Zn2+ depletion blocks endosome
fusion. Biochem J 312, 919–923.
17 Britigan BE, Rasmussen GT & Cox CD (1998) Binding
of iron and inhibition of iron-dependent oxidative cell
injury by the ‘calcium chelator’ 1,2-bis(2-aminophen-
oxy)ethane n,n,n¢,n¢-tetraacetic acid (BAPTA). Biochem
Pharmacol 55, 287–295.
18 Auld DS (1988) Use of chelating agents to inhibit
enzymes. Methods Enzymol 158, 110–114.
19 Auld DS (1995) Removal and replacement of metal ions
in metallopeptidases. Methods Enzymol 248, 228–242.
20 Schleifer KH & Kandler O (1972) Peptidoglycan types
of bacterial cell walls and their taxonomic implications.
Bacteriol Rev 36 , 407–477.
21 Quintela JC, Caparro
´
s M & de Pedro MA (1995) Vari-
ability of peptidoglycan structural parameters in gram-
negative bacteria. FEMS Microbiol Lett 125, 95–100.
22 Bonhivers M & Letellier L (1995) Calcium controls
phage T5 infection at the level of the Escherichia coli
cytoplasmic membrane. FEBS Lett 374, 169–173.
23 Moyer RW & Buchanan JM (1970) Effect of calcium
ions on synthesis of T5-specific ribonucleic acid. J Biol
Chem 245, 5904–5913.
24 Moyer RW & Buchanan JM (1970) Effect of calcium
ions on synthesis of T5-specific proteins. J Biol Chem

245, 5897–5903.
25 Danielsen EM, Noren O, Sjostrom H, Ingram J &
Kenny AJ (1980) Aspartate aminopeptidase: purifica-
tion by immunoadsorbent chromatography and proper-
ties of the detergent- and proteinase-solubilized forms.
Biochem J 189, 591–603.
26 Goto Y, Hattori A, Mizutani S & Tsujimoto M (2007)
Aspartic acid 221 is critical in the calcium-induced mod-
ulation of the enzymatic activity of human aminopepti-
dase A. J Biol Chem 282, 37074–37081.
27 Claperon C, Rozenfeld R, Iturrioz X, Inguimbert N,
Okada M, Roques B, Maigret B & Llorens-Cortes C
(2008) Asp218 participates with Asp213 to bind a
Ca2+ atom into the S1 subsite of aminopeptidase A: a
key element for substrate specificity. Biochem J 416,
37–46.
28 Korndo
¨
rfer IP, Kanitz A, Danzer J, Zimmer M, Loess-
ner MJ & Skerra A (2008) Structural analysis of the
l-alanoyl-d-glutamate endopeptidase domain of Listeria
bacteriophage endolysin Ply500 reveals a new member
of the LAS peptidase family. Acta Crystallogr 64, 644–
650.
29 Loessner MJ, Wendlinger G & Scherer S (1995) Hetero-
geneous endolysins in Listeria monocytogenes bacterio-
phages: a new class of enzymes and evidence for
conserved holin genes within the siphoviral lysis cas-
settes. Mol Microbiol
16, 1231–1241.

30 Fukushima T, Yao Y, Kitajima T, Yamamoto H &
Sekiguchi J (2007) Characterization of new L,D-endo-
peptidase gene product CwlK (previous YcdD) that
hydrolyzes peptidoglycan in Bacillus subtilis. Mol Genet
Genomics 278, 371–383.
31 Loessner MJ, Kramer K, Ebel F & Scherer S (2002)
C-terminal domains of Listeria monocytogenes bacterio-
phage murein hydrolases determine specific recognition
and high-affinity binding to bacterial cell wall carbohy-
drates. Mol Microbiol 44, 335–349.
32 Rawlings ND & Barrett AJ (1995) Evolutionary fami-
lies of metallopeptidases. Methods Enzymol 248,
183–228.
33 Wang J & Cooper MD (1993) Histidine residue in the
zinc-binding motif of aminopeptidase A is critical for
enzymatic activity. Proc Natl Acad Sci USA 90, 1222–
1226.
34 Larsen KS & Auld DS (1989) Carboxypeptidase A:
mechanism of zinc inhibition. Biochemistry 28, 9620–
9625.
35 Martı
´
n AC, Lo
´
pez R & Garcı
´
a P (1998) Functional
analysis of the two-gene lysis system of the pneumococ-
cal phage Cp-1 in homologous and heterologous host
cells. J Bacteriol 180, 210–217.

36 Summer EJ, Berry J, Tran TA, Niu L, Struck DK &
Young R (2007) Rz ⁄ Rz1 lysis gene equivalents in
phages of Gram-negative hosts. J Mol Biol 373,
1098–1112.
37 Brunel F, Thi VH, Pilaete MF & Davison J (1983)
Transcription regulatory elements in the late region of
bacteriophage T5 DNA. Nucleic Acids Res 11, 7649–
7658.
38 Wang J, Jiang Y, Vincent M, Sun Y, Yu H, Wang J,
Bao Q, Kong H & Hu S (2005) Complete genome
sequence of bacteriophage T5. Virol 332, 45–65.
39 McPheeters DS, Christensen A, Young ET, Stormo G
& Gold L (1986) Translational regulation of expression
of the bacteriophage T4 lysozyme gene. Nucleic Acids
Res 14, 5813–5826.
40 Fiedler F (1988) Biochemistry of the cell surface of Liste-
ria strains: a locating general view. Infection 16, 592–597.
41 Ananou S, Ga
´
lvez A, Martı
´
nez-Bueno M, Maqueda M
& Valdivia E (2005) Synergistic effect of enterocin AS-
48 in combination with outer membrane permeabilizing
treatments against Escherichia coli O157:H7. J App.
Microbiol 99, 1364–1372.
42 Rapoport N, Smirnov AI, Timoshin A, Pratt AM &
Pitt WG (1997) Factors affecting the permeability of
Pseudomonas aeruginosa cell walls toward lipophilic
compounds: effects of ultrasound and cell age. Arch

Biochem Biophys 344
, 114–124.
43 Besson F, Peypoux F, Quentin MJ & Michel G (1984)
Action of antifungal peptidolipids from Bacillus subtilis
on the cell membrane of Saccharomyces cerevisiae.
J Antibiot(Tokyo) 37 , 172–177.
44 Begunova EA, Stepnaia OA, Tsfasman IM & Kulaev IS
(2004) The effect of the extracellular bacteriolytic
G. V. Mikoulinskaia et al. L-Alanoyl-D-glutamate peptidase of bacteriophage T5
FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS 7341
enzymes of Lysobacter sp. on gram-negative bacteria.
Microbiology (Russia) 73, 320–325.
45 Streshinskaia GM, Naumova IB & Panina LI (1979)
Chemical composition of the cell wall of Streptomyces
chrysomallus which produces the antibiotic aurantin.
Microbiology (Russia) 48, 814–819.
46 Riordan JF & Vallee BL (1967) Acetylation. Methods
Enzymol 11, 565–570.
47 Ward JB (1973) The chain length of the glycans in
bacterial cell walls. Biochem J 133, 395–398.
48 Park KT & Johnson MJ (1949) A submicrodetermina-
tion of glucose. J Biol Chem 181, 149–151.
49 Ghuysen JM, Tipper DJ & Strominger JL (1966)
Enzymes that degrade bacterial cell walls. Methods
Enzymol 8, 685–711.
50 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein
utilizing the principle of protein-dye binding. Anal
Biochem 72, 248–254.
51 Laemmli UK (1968) Cleavage of structural proteins

during the assembly of the head of bacteriophage T4.
Nature 170, 5259–5268.
52 Jeanmougin F, Thompson JD, Gouy M, Higgins DG &
Gibson TJ (1998) Multiple sequence alignment with
Clustal X. Trends Biochem Sci 23, 403–405.
53 Altschul SF, Madden TL, Scha
¨
ffer AA, Zhang J,
Zhang Z, Miller W & Lipman DJ (1997) Gapped
BLAST and PSI-BLAST: a new generation of protein
database search programs. Nucleic Acids Res 25,
3389–3402.
54 Tamura K, Dudley J, Nei M & Kumar S (2007)
MEGA4: Molecular Evolutionary Genetics Analysis
(MEGA) software version 4.0. Mol Biol Evol 24,
1596–1599.
L-Alanoyl-D-glutamate peptidase of bacteriophage T5 G. V. Mikoulinskaia et al.
7342 FEBS Journal 276 (2009) 7329–7342 ª 2009 The Authors Journal compilation ª 2009 FEBS