Vol.2, No.5, 484-492 (2010)
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
The lytic mechanism of Escherichia coli α-hemolysin
associated to outer membrane vesicles
——Lytic action mechanism of OMVs-associated HlyA
Vanesa Herlax1*, Maria Florencia Henning1, Ana María Bernasconi1, Felix María Goñi3, Laura Bakás1,2
1Instituto de Investigaciones Bioquímicas La Plata (INIBIOLP), CCT-La Plata, CONICET, UNLP, Facultad de Ciencias Médicas,
La Plata, Argentina; *Corresponding Author: vherlax@atlas.med.unlp.edu.ar
2Departamento de Ciencias Biológicas, Facultad de Ciencias Exactas, Universidad Nacional de La Plata, La Plata, Argentina
3Unidad de Biofísica (Centro Mixto CSIC-UPV/EHU), and Departamento de Bioquímica, Universidad del País Vasco, Bilbao, Spain
Received 2 January 2010; revised 1 February 2010; accepted 2 February 2010.
Alpha-hemolysin (HlyA) is an extracellular toxin
secreted by Escherichia coli, targeting to plasma
membranes of eukaryotic cells. Recently it was
found that this toxin is released to external
media associated to bacterial outer membrane
vesicles (OMVs), but the hemolytic mechanism
in this way has not been studied yet. Our results
report that HlyA is the only protein present in
OMVs that is responsible for erythrocyte lysis,
and show that no fusion event is involved in the
lytic mechanism of OMVs-HlyA. Furthermore,
the specific hemolytic activity is approximately
10 fold higher than that of purified free-HlyA,
showing the same relative lysis efficiency and
specificity for erythrocytes from different
species. OMVs could be an important auxiliary
way of secretion, acting mainly as a concentra-
tion mechanism of HlyA near the target cells.
Cell lysis would occur after toxin transfer from
OMVs to target membranes, as demonstrated by
hemolysis kinetic studies, lipid mixing and
western blot assays.
Keywords: OMVs; Toxin; Lipid Mixing;
Hemolytic Activity; Erythrocytes
The hemolytic toxin -hemolysin (HlyA), member of
the RTX toxin family [1], is an important virulence fac-
tor produced by several strains of Escherichia coli. It is
involved in human extraintestinal diseases, such as uri-
nary tract infections, peritonitis, meningitis and septice-
mia [2]. As in most RTX proteins, an operon composed
of 4 genes (hlyABCD) is involved in the polypeptide
synthesis, postraslational modification and secretion of
the active HlyA to extracellular media [3,4]. HlyA is
synthesized in an inactive form (ProHlyA), which is
activated in the cytoplasm to the hemolytically active
form by HlyC, a fatty acid acyltransferase [5]. Then
HlyA is secreted across both membranes by the type I
export process employing an uncleaved C-terminal reco-
gnition signal [6,7]. Although HlyA has its own machin-
ery to be export from the bacteria, Balsalobre et al. [8]
demonstrated the presence of physiologically active
HlyA in outer membrane vesicles (OMVs) of hemolytic
strains of Escherichia coli.
Outer membrane vesicles (OMVs) are constantly
being discharged from the surface of Gram negative
bacteria during bacterial growth. All Gram negative
bacteria studied up to date, including Escherichia coli,
Neisseria meningitidis, Pseudomonas aeruginosa, Hel-
icobacter pylori, Shigella flexneri and Actinobacillus
actinomycetemcomitans produce outer membrane vesi-
cles and their release is increased when bacteria are
exposed to stressful conditions such as antibiotics or
serum. One can be tempted to think that OMVs pro-
duction is the result of membrane instability, but recent
studies have demonstrated that mutations that affect
protein synthesis, localization, envelope structure and
envelope stress response pathway alter the vesiculation
levels. In this way the release of OMVs offers to the
cell an effective mechanism for removal of material as
a macromolecular complex, allowing to discard unw-
anted material or alter the composition of the envelope
under conditions where remodeling would be advantag-
eous [9,10].
Despite that in vivo, release of OMVs could not be
monitored, the presence of particles resembling these
vesicles has been detected in plasma from patients with
different infectious processes [11-18]. OMVs serve as
secretory vehicles for proteins and lipids of Gram
negative bacteria and play roles in establishing a colon-
V. Herlax et al. / Natural Science 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/Openly accessible at
ization niche carrying or transmitting virulence factors
into host cells or modulating host defense and response,
acting also as long-range virulence factors that can
protect luminal cargo from extracellular host proteases
and penetrate into tissues more readily than larger bac-
teria [19]. An extensive review about the significance
of bacterial OMVs in the host-pathogen interaction was
published by Kuehn and Kesty [20]. Besides the toxin o
protein delivery, other roles were characterized for
OMVs, such as: interaction interspecies during multis-
pecies infections, DNA transfer, DNA uptake, commu-
nication interspcies [21].
OMVs are primarily composed of lipopolysaccharide
(LPS), phospholipids and various outer membrane (OM)
proteins and periplasmic components, but they do not
contain intra membrane (IM) or cytoplasmic components
[22], although OMVs isolated from E. coli O157:H7
contain DNA [23].
Considering that HlyA was found associated to
OMVs of some strain of E. coli and that erythrocytes
have been the cell type classically used in studies on the
action mechanism of this toxin, in this research we
studied the hemolytic mechanism of OMVs-associated
HlyA. The results demonstrate that the lytic action of
OMVs-associated HlyA like free-HlyA does not in-
volve fusion events, instead the toxin is transferred
from OMVs to target cells, as demonstrated by hemo-
lysis kinetic studies, lipid mixing and western blot as-
2.1. Isolation of Outer Membrane Vesicles
E. coli strains used as a source of HlyA were WAM 1824,
an overproducing hemolysin strain [24], and WAM 783,
for the unacylated inactive toxin ProHlyA [25]. Both of
them were kindly provided by R. A.Welch (University of
Wisconsin, Madison, Wisconsin).
OMVs were isolated from bacterial culture supern-
atants. Briefly, bacteria were grown at 37ºC in Luria
Bertani broth (LB). Culture samples from the decelerat-
ing growth phase were centrifuged at 10000 rpm for 20
min at 4ºC. The resulting suspension was then centri-
fuged at 47500 rpm for 2 h at 4ºC to collect the pellet
containing OMVs.
2.2. Protein Purification
Free-HlyA was purified from the supernatants resulting
from OMVs. It was concentrated and partially purified
by precipitation with 20% cold ethanol. A precipitate
containing HlyA was collected by centrifugation (1 h,
10000 rpm in a Sorvall rotor SSA 34), then it was
resuspended in 20 mM Tris, 150 mM NaCl, pH 7.0 (TC
buffer). This preparation showed, on SDS-PAGE, a main
band at 110 kDa corresponding to more than 90% of the
total protein. This band was assigned to free-HlyA. The
protein could be stored at –70°C.
2.3. OMVs Composition
2.3.1. Quantitative Assays
Quantitative assays were performed using Lowry method
for proteins [26], KDO 2-keto-3-deoxyoctonate for
LPS [27] and inorganic phosphorous for phospholipids
2.3.2. Analysis of Proteins by SDS-PAGE
Samples were electrophoresed on 10% acrylamide gels
in the presence of SDS according to Laemmli et al. [29].
Protein bands were visualized by Coomassie Blue stain-
ing [30].
2.3.3. Immunoblotting Analysis
Samples from 10% SDS-polyacrylamide gels were
transferred to nitrocellulose by the method of Towbin et
al. [31]. Blots were blocked with 3% skim milk in TBS
buffer (10 mM Tris-HCl, 150 mM NaCl, pH 7.5) at
room temperature for 2 h. They were then incubated
with a solution containing a polyclonal rabbit anti-
hemolysin antibody (1:1000) in 3% skim milk-TBS at
4°C overnight, washed with TBS buffer, and finally
reacted with peroxidase-conjugated anti-rabbit Ig anti-
body (Sigma) (1:1000) in TBS buffer with 3% skim
milk at room temperature for 2 h. After incubation and
washing as stated above, the nitrocellulose was trans-
ferred to a peroxidase substrate solution containing 6
mg 4-chloro-1-naphthol (Sigma) in 1 ml methanol, 1
ml TBS, 3 ml H2O and 8 l H2O2 for the detection of
horseradish peroxidase-conjugate antibodies on the
2.4. Hemolytic Assays
For the hemolytic assays, free-HlyA or OMVs associ-
ated-HlyA were serially diluted in TC cold buffer con-
taining 10 mM CaCl2 on a 96-well microtiter plate. One
hundred L of the diluted suspensions were mixed with
100 μl of standardized horse or rabbit erythrocytes. The
mixture was incubated at 37°C for 30 min. The absorb-
ance of supernatants was read at 412 nM [32]. One
hemolytic unit/mL (HU/mL) is defined as 10 fold the
dilution of toxin preparation producing 50% lysis of the
erythrocyte suspension.
The standardization of erythrocytes was done just bef-
ore the assay. The erythrocytes were washed in 0.9%
NaCl and then diluted to 12.5 L in 1 mL of distilled
water to give a reading of 0.6 absorbance unit at 412 nM
V. Herlax et al. / HEALTH 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/Openly accessible at
2.5. Hemolysis Kinetics
Hemolysis kinetics was determined by measuring the
decrease in turbidity (absorbance at 650 nM) of a stan-
dardized horse erythrocyte suspension exposed to free-
HlyA or OMVs associated-HlyA as a function of time at
37ºC. A series of hemolytic reactions were set up starting
with an amount of free-HlyA or OMVs associated-HlyA
that would produce 100% of hemolysis in 96-well plate
dilution assays. The real time kinetics assays for hemo-
lysis were performed in the following manner: 500 µL of
standardized erythrocytes and 500 µL of hemolysis buffer
containing 10 mM CaCl2 were placed into a cuvette.
Hemolysis started when the toxin was injected into the
cuvette, and the absorbance at 650 nM was measured.
2.6. LUVs Preparation
Large unilamellar vesicles (LUVs) composed of PC/PE/Chol
2:1:1 molar ratio were prepared by extrusion using 0.1
μM pore size membranes as described by Mayer et al.
2.7. Ghost Erythrocytes Preparation
Five mililitres of horse erythrocytes were washed with TC
buffer and osmotically lysed in 10 mM tris-HCl PH = 7.4
buffer at 4ºC for 30 min. The membranes were pelleted by
centrifugation (10 min at 10000 rpm), and washed until
the supernatant remained clear. The membranes were
finally resuspended in 3 mL of TC buffer. The phosphate
concentration of these samples was 0.53 ± 0.05 mg
phospholipid/mL (n = 3).
2.8. Labeling of Ghost Erythrocytes with
(Rh-Pe) and N-(7-Nitro-2, 1, 3-Benzoxa-
Ghost erythrocytes were labeled with Rh-PE and NBD-PE
(Molecular Probes, Eugene, O.R.). Both fluorophores
are coupled to the free amino group of phosphatidyle-
thanolamine to provide analogues which can be incur-
porated into a lipid bilayer. 20 L of a fresh solution of
each probe (1 mg/ml in ethanol) was added to 75 L of
ghost erythrocytes. After incubating the suspension for 1 hr
at room temperature, 6 washes with TC were performed
so as to separate unbound labeled lipids. Finally the la-
beled ghost erythrocytes were resuspended in 5.4 ml of
2.9. Lipid Mixing
Membrane fusion was analyzed by lipid mixing between
OMVs and ghost erythrocytes labeled with NBD-PE and
Rh-PE. The fusion assay involves resonance energy
transfer between NBD as donor and Rh as acceptor.
When both fluorescent lipids are in ghost erythrocytes,
efficient energy transfer is observed [35]. After lipid
mixing by fusion with a population of unlabeled memb-
ranes, in this case OMVs, a decrease in efficiency of
resonance energy transfer should be observed followed
by an increase of the donor fluorescence or decrease of
the acceptor fluorescence. In our experimental design,
we followed the increase of donor (NBD-PE) fluoresc-
ence as a function of time using the excitation and emis-
sion monochromators set at 465 nM and 530 nM, res-
pectively. Slits for excitation and emission were 4 nM.
The extent of lipid mixing was determined according
to the following equation:
% lipid mixing = [(Ft – F0) / (Fmax – F0)] × 100
where F0 is the initial fluorescence value of erythrocytes
labeled with both probes, Ft is the value of fluorescence
after t minutes of incubation with OMVs and Fmax (100%
fluorescence) is the value of fluorescence after addition
of Triton × 100 to disperse maximally the probes.
3.1. Characterization of OMVs Produced by
E. coli WAM 1824
Recently, Balsalobre et al. demonstrated that physiologi-
cally active HlyA is associated with OMVs produced by
laboratory strains and also from natural and clinical E.
coli isolates. However, as the amount of OMVs and the
percentage of the soluble and OMVs associated HlyA
were likely to vary markedly depending on the strain [8],
we characterized OMVs obtained from E. coli WAM
1824 strain.
OMVs were purified from the culture filtrate of E. coli
WAM 1824. Many bilayered spherical vesicles with di-
ameters ranging from 50 to 200 nm were observed by
electron microscopy (data not shown). The lipid compo-
sition of these vesicles was analyzed by thin layer chro-
matography, showing that the two predominant lipid spe-
cies are phosphatidylethanolamine (PE) and cardiolipin
(CL), demonstrating that these vesicles arise indeed from
the bacterial outer membrane (data not shown).
The presence of HlyA in OMVs produced by E. coli
WAM 1824 is shown in Figure 1(a). This figure shows the
SDS-PAGE polypeptide profile of OMVs. A band with
an electrophoretic mobility corresponding to HlyA (MW
110 kDa) was observed whose identity was confirmed by
Western blot analysis employing polyclonal antibodies
directed against purified HlyA (Figure 1(b)). The others
bands that appear in SDS-PAGE stained by Comassie
stain, correspond to periplasmic and outer membrane
proteins also present in OMVs.
V. Herlax et al. / Natural Science 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/
Openly accessible at
(a) (b)
Figure 1. HlyA is present in WAM 1824 outer membrane vesi-
cles. (a) SDS-PAGE 10%, stained by Coomasie. 1: LMW, 2:
WAM 1824 OMVs (10μg prot), 3: WAM 783 OMVs (10 μg);
(b) Western-Blot, using anti-HlyA polyclonal antibodies. 1:
Rainbow molecular weight, 2: OMV’s (100 ng prot), 3:
Standard HlyA (50 ng).
3.2. Comparison of Hemolytic Activity
between Free-HLYA and OMVs
The following experiments were done so as to compare
the hemolytic activity between the OMVs associated-
HlyA and free toxin, which was quantified by serial dilu-
tion assays against horse erythrocytes. The hemolytic ac-
tivity was 40960 and 2560 HU/ml for OMVs associated-
HlyA and free toxin, respectively. If we consider that
total protein concentration in the OMVs is 2.7 mg/mL
and 70% corresponds to HlyA (estimated from the inten-
sity of the 110 kDa band in a 10% SDS-PAGE gel, ana-
lyzed by Kodak digital Science 1D Software), so the
approximate concentration of OMVs associated- HlyA is
1.7 10-5 M. These values allow us to calculate the spe-
cific activity which is 2.37 1012 HU/mol and 2.04 1011
HU/mol for OMVs associated-HlyA and free-HlyA, re-
In order to confirm that the hemolytic activity of
OMVs is solely due to HlyA and not to other proteins
present in the OMVs, E. coli WAM 783 (strain with a
deletion of hlyC gen that encodes HlyC, the protein
responsible for HlyA acylation) was used as a source of
OMVs. The presence of a protein corresponding to
110 kDa in 10% SDS-PAGE confirms that inactive
unacylated toxin is also associated to OMVs (Figure 1,
lane 3). OMVs obtained from this strain do not have any
hemolytic activity confirming that this activity in OMVs
is a particular property due largely or totally to the pres-
ence of HlyA.
3.3. Lipid Mixing Assays
In order to determine whether the hemolytic action mec-
hanism of OMVs associated-HlyA implies membrane
fusion, we analyzed the lipid mixing between OMVs and
ghost erythrocytes labeled with NBD-PE (donor) and
Rh-PE (acceptor). The ghost erythrocytes concentration
used in lipid mixing assays is equal to the erythrocyte
suspension used in hemolytic experiments.
6.9% FRET decrease was detected between OMVs
and labeled erythrocytes promoted by OMVs associated-
HlyA. When free-HlyA was added to labeled erythro-
cytes, 8.1% FRET decrease was obtained (Figure 2),
which cannot be assigned to lipid mixing because unla-
beled erythrocytes were absent in this experiment, so it
should be attributed to protein insertion into lipid bilayer,
increasing the distance between donor and acceptor [36].
The decrease in energy transfer in both experiments
was similar, indicating that fusion events are not inv-
olved in the action mechanism of this toxin when it is
OMVs associated; instead, a transfer process from
OMVs to the target membrane must be involved. The
presence of a high affinity specific receptor is not strictly
necessary, as the same experiments were repeated with
labeled large unilamellar vesicles (LUVs), giving similar
FRET decrease (data not shown).
3.4. HlyA Transfer from OMVs to
To study HlyA transfer from OMVs to erythrocytes, a
serial diluted hemolytic assay was carried out as desc-
ribed previously in Experimental Procedures. Samples
were centrifuged at 10,000 rpm to separate ghost erythr-
ocytes from OMVs. The pellets containing the ghosts
were resuspended in SDS sample buffer, and electropho-
resed in a 10% SDS-PAGE. Finally this gel was trans-
time (sec)
0200 400 600 800
Fluorescence at 530 nm
Figure 2. OMVs associated-HlyA does not induce membrane
fusion. Lipid mixing assays between labeled erythrocytes with
Rh-PE/NBD-PE and OMVs () or free-HlyA (). Solid
arrows correspond to the addition of OMVs or free-HlyA.
Dotted arrows correspond to the addition of 0.2% Triton X-100
(Fmax). Total lipid concentration was 0.1 mM. 40 HU for
free-HlyA and OMVs associated-HlyA were used.
V. Herlax et al. / HEALTH 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/
Openly accessible at
ferred to nitrocellulose and HlyA was detected using a
polyclonal rabbit anti-HlyA antibody. Figure 3(a) shows
that HlyA is effectively transferred to erythrocytes. The
amount of HlyA decreases with the decrease of the
OMVs concentration, according to the decrease in the
hemolytic activity observed (data not shown).
Another fact that confirms that fusion events are not
involved in the hemolytic process mediated by OMVs is
the absence of LPS in the ghost erythrocytes membranes
after their incubation with OMVs. The detection of LPS
was analyzed by a specific silver stain for LPS (BioRad
silver stain kit, catalog 161-0445) of a 16% SDS-PAGE. A
typical LPS pattern is not observed in this sample (Figure
3(b), lane 1), instead some bands of glycoproteins of
erythrocytes are observed.
Figure 3. HlyA is transfered from OMVs to erythrocytes. (a)
HlyA transference from OMVs to erythrocytes analyzed by
western blot: lane 1: standard HlyA, 2-7: samples of ghosts
erythrocytes previously incubated at 37°C during 30 min with
serially diluted OMVs, 8: ghost erythrocytes; (b) Analysis of
LPS content in ghost erythrocytes: lane 1: ghosts obtained
from the hemolysis assay with OMVs, 2: ghosts, 3: OMVs and
4: Standard LPS from E. coli 011:B4.
3.5. Hemolysis Kinetics
Typical hemolysis time courses are shown in Figure 4.
For both, OMVs associated-HlyA and free-HlyA, the
time to reach 50% of hemolysis increases as the toxin
concentration decreases. The hemolytic curves exhibit a
define lag period for both free-HlyA and OMVs associated-
HlyA. This lag period lengthens noticeably at decreasing
concentrations of OMVs associated-HlyA, but it remains
practically constant for free-HlyA as seen in Figure 5(a),
indicating that the rate limiting step in the hemolysis
process induced by OMVs associated-HlyA is the diffus-
ion of vesicles in the aqueous medium, due to their higher
molecular weight and volume compared to free-HlyA.
Initial rate of hemolysis obtained from the linear por-
tion of the kinetic curve are shown in Figure 5(b). Vir-
02004006008001000 12001400 1600180020002200 24002600 2800
> OMV HlyA
DO 650 nm
time (sec)
02004006008001000 1200 1400 1600 1800 2000 22002400 2600
DO 650 nm
Figure 4. Kinetics of hemolysis induced by OMVs associ-
ated-HlyA (a) and free-HlyA (b). The real time kinetic assays
for hemolysis were performed in the following manner: 500 µl
of standardized erythrocytes and 500 µl of hemolysis buffer
containing 10 mM CaCl2 were placed into a cuvette. Hemolysis
started when the toxin was injected into the cuvette, and the
absorbance at 650 nm was measured. The arrow indicates toxin
concentration increase.
V. Herlax et al. / Natural Science 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/
Openly accessible at
g protein
0510 15 20 25 30 35 40
lag period (sec)
free HlyA
g protein
0510 15 20 25 30 35 40
hemolysis rate (DO/sec)
free HlyA
Figure 5. Initial hemolytic rate (a) and lag time (b) as a func-
tion of OMVs associated-HlyA () and free-HlyA () concen-
tration. Initial rates (linear portion) and lag time were obtained
from Figure 5 at different toxin concentrations.
tually constant values were obtained with OMVs associ-
ated-HlyA for all the concentrations studied, while a
decrease in initial rates as concentrations decrease was
observed for free-HlyA.
3.6. Erythrocytes Specificities of Free-HLYA
and OMVs Associated-HlyA
To characterize the hemolytic efficiency of OMVs assoc-
iated-HlyA on different species of mammalian cells, we
chose rabbit and horse erythrocytes, which were the
most commonly used and characterized in previous stu-
dies related with the hemolytic action mechanism of this
toxin. In Figures 6(a) and (b) it can be seen that in both
cases, rabbit erythrocytes are more sensitive than those
of horse, as demonstrated in earlier results on the char-
acterization of the lytic action of HlyA [37]. The D50
values (amount of toxin that produces 50% of hemolysis)
are 0.46 and 2.81 nM for OMVs associated-HlyA for
rabbit and horse erythrocytes respectively.
The dose-response curves indicate a higher lytic effic-
iency of OMVs associated-HlyA (Figure 6(a)) as com-
pared to free-HlyA (Figure 6(b)), shown by the D50 on
horse erythrocytes of 2.81 and 22.5 nM, respectively.
Moreover, analysis of curves of hemolytic activity
against concentration using SigmaPlot (Jandel Scientific,
San Rafael, CA) fitted a sigmoidal curve in both cases.
In the case of horse cells, data fit a sigmoid 4 parameter
curve, while the rabbit ones fit a logistic 3 parameter
curve, demonstrating a different cooperativity in the
hemolysis process. The same behavior is observed for
OMVs associated-HlyA as for free-HlyA.
Recently, it was demonstrated the presence of active
HlyA associated with OMVs produced by natural and
clinical E. coli isolates. Despite a great deal of published
log g Hly in OMV
0.0001 0.0010.010.11101001000
% hemolysis
log μg HIy in OWVs
log μg HIy in OMVs
log g free-Hly
0.0001 0.0010.010.11101001000
% hemolysis
Figure 6. Hemolysis induced by OMVs associated-HlyA and
free-HlyA on erythrocytes of different species. Hemolytic act-
ivity of OMVs associated-HlyA (a) and free-HlyA (b), tested
with rabbit erythrocytes () and horse erythrocytes ().
V. Herlax et al. / HEALTH 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/Openly accessible at
data regarding the activity of soluble HlyA secreted by E.
coli, nothing was known regarding the action mechan-
ism of OMVs associated HlyA. First, we demonstrated
in this research, the presence of HlyA associated to
OMVs produced by a laboratory E. coli strains (WAM
1824). We demonstrated the presence of HlyA associated
to these vesicles as seen in SDS-PAGE polypeptide prof-
ile and Western blot analysis employing polyclonal ant-
ibodies directed against purified HlyA (Figure 1). The
toxin associated to OMVs has a specific activity one
order of magnitude higher than the free-HlyA. It is imp-
ortant to note that this protein is the only one respons-
able of the hemolytic activity as OMVs produced by
strain WAM 783 (strain encoding the inactive form of
HlyA, ProHlyA) are not hemolytic.
In order to determine whether the hemolytic action
mechanism of OMVs associated-HlyA implies mem-
brane fusion, we analyzed the lipid mixing between
OMVs and labeled ghost erythrocytes. The small FRET
decrease obtained could be attributed to protein insertion
into lipid bilayer, as similar values were obtained in the
presence of free-HlyA (Figure 2). These values are sim-
ilar to those obtained in the study of HlyA insertion area
using also FRET experiments with labeled membranes
[36]. As for free-HlyA [38], fusion events are not inv-
olved in the action mechanism of the toxin associated
with OMVs; instead, a transfer process from OMVs to
the target membrane must be involved. The transference
of HlyA as a mechanism was confirmed by the presence
of HlyA and by the absence of LPS in the pellet obtained
after incubation of ghost erythrocytes with OMVs (Fig-
ure 3). These results effectively demonstrate that the
hemolysis is caused by the transference of HlyA and not
OMVs-cell membranes fusion.
Although fusion events have been described for leuk-
otoxin from A. actinomycetemcomitans, another RTX
toxin, this could be specific for this RTX toxin [39], as
the secretion of leukotoxin differs from that of other
RTX proteins in that it remains associated with intact
cells even in the presence of functional ltxBD genes [40].
The common mechanism of the lytic action of HlyA
implies the diffusion of a soluble secreted toxin from
bacteria, through aqueous medium, to reach the target
cells. Now, our results demonstrate that a large number
of toxin molecules are concentrated and transported by
OMVs and then transfer to target cells.
The different behavior in the kinetics of hemolysis
observed in Figure 4 for OMVs associated-HlyA and
free-HlyA can be explained considering that, in the first
case, the limiting step is the slow diffusion of OMVs in
the aqueous medium that increases the lag time. Once
OMVs reach the erythrocyte, a large number of toxin
molecules which are concentrated on OMVs, are all tog-
ether transferred to target cells either as a monomer or as
a preassembled oligomer. On the contrary, the limiting
step for free-HlyA is the number of toxin molecules that
reach the target cells, due to the fact that they decrease
as a function of toxin concentration. This can explain the
high specific hemolytic activity for OMVs-associated
HlyA in comparison with the hemolytic activity of free-
HlyA. The fact that HlyA can be transfer in an oligo-
meric structure is possible because we have recently
demonstrated that an oligomer is necessary for the occu-
rrence of hemolysis [41]. What’s more, the adoption of a
cytolytically active oligomeric conformation in OMVs
was described for enterobacterial ClyA cytotoxin [18].
The transference is direct from the OMVs to the eryt-
hrocytes without the delivery of the toxin to the medium.
This is demonstrated by the differences found in the
hemolysis kinetic studies (Figure 4). This conclusion is
also supported by results published by Balsalobre et al.,
who demonstrated that HlyA is tightly associated to
vesicles appearing in a soluble form only after the mem-
brane structure of the OMVs is disrupted by detergents
[8]. This transfer process does not need the presence of a
specific high affinity receptor because the same effect
was observed when LUVs labeled with Rh-PE and
NBD-PE were used (data not shown).
Rabbit and horse erythrocytes, were the most comm-
only erythrocytes used and characterized in previous
studies related with the hemolytic action mechanism of
this toxin. In Figures 6(a) and (b) it can be seen that in
both cases, rabbit erythrocytes are more sensitive than
those of horse, as demonstrated in earlier results on the
characterization of the lytic action of HlyA [37]. For
rabbit erythrocytes a smoother hemolytic curve was ob-
tained in comparison to that of horse erythrocytes. This
fact may be due to the presence of high proportion of
high affinity receptors in the horse erythrocytes in com-
parison to those of rabbit. The higher lytic efficiency of
the toxin with rabbit cells could be due to the presence
of a large number of low affinity binding sites i.e. mem-
brane phospholipids, which facilitate the concentration
of the toxin in the membrane (Herlax et al., unpublished
Currently, the presence of a receptor is a contradictory
point in the elucidation of the action mechanism of this
toxin, probably due that experiments were done with
erythrocytes from different mammalian species. Glyco-
phorin was described as a receptor in horse and human
erythrocytes [42], while another group proposed that the
hemolytic process does not depend on the receptor pres-
ence [43]. Anyway, the same behavior for HlyA against
to erythrocytes of different species is observed for
OMVs associated-HlyA as for free-HlyA.
Finally, it is important to remember that our previous
results support the hypothesis that HlyA is secreted to
the external medium as a LPS-HlyA complex, in which
the main action of LPS is to maintain the protein stabil-
ity in solution, while it only indirectly affects HlyA lytic
V. Herlax et al. / Natural Science 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/
Openly accessible at
activity [44]. The presence of LPS in the toxin sample
used in those studies, which was obtained by precipita-
tion techniques from a culture supernatant without the
ultracentrifugation step, could be due to some OMVs
contamination. This is also supported by the presence of
phospholipid in toxin samples purified by methods such
as ammonium sulphate precipitation and exclusion chro-
matography [45]. Due to the fact that OMVs associated-
HlyA represents a very high percentage of toxins in the
culture supernatants, results on biological effects of
HlyA published up to date, mainly those in which sublytic
concentrations were used, should be revised as small
amounts of LPS present in the OMVs can stimulate the
release of inflammatory mediators [46].
In conclusion, our results demonstrate that OMVs
constitute an alternative secretion mechanism for HlyA
that result in the toxin reaching higher concentrations
without altering its lytic action mechanism.
The observation that certain virulence factors are enri-
chhed in vesicles suggests that OMVs may have a key
role in bacterial pathogenesis by mediating transmission
of active virulence factors and other bacterial envelop-
ment component to host cells. Numerous OMVs associ-
ated virulence factors have been shown to induce cyto-
toxicity confer vesicle binding to and invasion to host
cells and modulate the immune response. On the other
hand, similar to bacterial-host cell interactions, OMVs-
cell interactions can be altered by manipulating the exp-
ression of outer membrane proteins in bacteria. The ma-
nipulation of OMVs adherence and the possibility to
create chimeras between the N-terminal half of HlyA,
which contains the pore forming domain, and binding
domains to specific proteins in the surface of a cell of
interest (e.g. cancerous cells) should be useful for num-
erous applications, redirecting this engineering OMVs to
specific cell types in order to achieve a desired therapeutic
This work was partially supported by grants from UNLP, CICPBA and
ANPCyT-Argentina. LSB is a member of the “Carrera del Investigador”,
CICPBA, Argentina; VH is a member of the “Carrera del Investigador”,
Consejo Nacional de Investigaciones Científicas y Tecnológicas
(CONICET), Argentina and MFH is a Fellow from CONICET, Argentina.
AMB is a member the “Carrera del Técnico y Profesional de Apoyo”,
[1] Coote, J.G. (1992) Structural and fuctional relationships
among the RTX toxin determinants of gram-negative bacte-
ria. FEMS Microbiology Reviews, 88, 137-162.
[2] Cavalieri, S., Bohach, G. and Synder, I. (1984) Escherichia
coli a-hemolysin characteristics and probable role in patho-
genicity. Microbiology Reviews, 48, 326-343.
[3] Welch, R.A. (1991) Pore-forming cytolysins of gram-nega-
tive bacteria. Molecular Microbiology, 5(3), 521-528.
[4] Ludwig, A. and Goebel, W. (1999) The family of the multi-
genic encoded RTX toxin. In: Alouf, J.E. and Freer, J.H.,
Eds., The Comprehensive Source-Book of Bacterial Protein
Toxins, Academic Press, 330-348.
[5] Stanley, P., Korornakis, V. and Hughes, C. (1998) Acylation
of Escherichia coli hemolysin: A unique protein lipidation
mechanism underlying toxin fuction. Microbiology and
Molecular Biology Reviews, 62(2), 309-333.
[6] Jarchau, T., Chakraborty, T., Garcia, F. and Goebel, W.
(1994) Selection for transport competence of C-terminal
polypeptides derived from Escherichia coli hemolysin: The
shortest peptide capable of autonomous HlyB/HlyD-
dependent secretion comprises the C-terminal 62 amino ac-
ids of HlyA. Molecular and General Genetics, 245, 53-60.
[7] Koronakis, V., Koronakis, E. and Hughes, C. (1989) Isola-
tion and analysis of the C-terminal signal directing export of
Escherichia coli hemolysin protein across both bacterial
membranes. European Molecular Biology Organization
Journal, 8(2), 595-605.
[8] Balsalobre, C., Silvan, J.M., Berglund, S., Mizunoe, Y.,
Uhlin, B.E. and Wai, S.N. (2006) Release of the type I sec-
reted alpha-haemolysin via outer membrane vesicles from
Escherichia coli. Molecular Microbiology, 59(1), 99-112.
[9] McBroom, A.J., Johnson, A.P., Vemulapalli, S. and Kuehn,
M.J. (2006) Outer membrane vesicle production by Es-
cherichia coli is independent of membrane instability. The
Journal of Bacteriology, 188(15), 5385-5392.
[10] McBroom, A.J. and Kuehn, M.J. (2007) Release of outer
membrane vesicles by Gram-negative bacteria is a novel
envelope stress response. Molecular Microbiology, 63(2),
[11] Beveridge, T.J. (1999) Structures of Gram-negative cell
walls and their derived membrane vesicles. The Journal of
Bacteriology, 181(16), 4725-4733.
[12] Horstman, A.L. and Kuehn, M.J. (2000) Enterotoxigenic E.
coli secretes active heat-labile enterotoxin via outer mem-
brane vesicles. Journal of Biological Chemistry, 275(17),
[13] Devoe, I.W. and Gilchrist, J.E. (1973) Release of endotoxin
in the form of cell wall blebs during in vitro growth of Neis-
seria meningitis. Journal of Experimental Medicine, 13 8,
[14] Keenan, J., Day, T., Neal, S., Cook, B., Perez-Perez, G.,
Allardyce, R. and Bagshaw, P. (2000) A role for the bacte-
rial outer membrane in the pathogenesis of Helicobacter py-
lori infection. FEMS Microbiology Letters, 182(2), 259-264.
[15] Fiocca, R., Necchi, V., Sommi, P., Ricci, V., Telford, J.
Cover, T.L. and Solcia, E. (1999) Release of Helicobacter
pylori vacuolating cytotoxin by both a specific secretion
pathway and budding of outer membrane vesicles. Uptake of
released toxin and vesicles by gastric epithelium. Journal of
Pathol ogy, 188, 220-226.
[16] Kato, S., Kowashi, Y. and Demuth, D.R. (2002) Outer
membrane-like vesicles secreted by Actinobacillus action-
mycetemcomitans are enriched in leukotoxin. Microbial
Pathogenes, 32, 1-13.
[17] Wai, S.N., Takade, A. and Amako, K. (1995) The release of
outer membrane vesicles from the strains of enterotoxigenic
Escherichia coli. Microbiology and Immunology, 39(7),
[18] Wai, S.N., Lindmark, B., Soderblom, T., Takade, A., Weste-
V. Herlax et al. / HEALTH 2 (2010) 484-492
Copyright © 2010 SciRes. http://www.scirp.org/journal/HEALTH/Openly accessible at
rmark, M., Oscarsson, J., Jass, J., Richter-Dahlfors,V., Mi-
zunoe, Y. and Uhlin, B.E. (2003) Vesicle-mediated export
and assembly of pore-forming oligomers of the enterobacte-
rial ClyA cytotoxin. Ce ll, 115(1), 25-35.
[19] Chi, B., Qi, M. and Kuramitsu, H.K. (2003) Role of dentil-
isin in Treponema denticola epithelial cell layer penetration.
Research in Microbiology, 154(9), 637-643.
[20] Kuehn, M. and Kesty, N. (2005) Bacterial outer membrane
vesicles and the host-pathogen interaction. Genes and Dev-
elopment, 19(22), 2645-2655.
[21] Mashburn-Warren, L. and Whiteley, M. (2006) Special
delivery: Vesicle trafficking in prokaryotes. Molecular Mic-
robiology, 61(4), 839-846.
[22] McBroom, A. and Kuehn, M.J. (2005) Outer membrane
vesicles. In: Curtiss, R. III, et al., Eds., EcoSal-Esche-
richia coli and Salmonella: Cellular and Molecular Bio-
logy, ASM Press, Washington, DC.
[23] Kolling, G.L. and Matthews, K.R. (1999) Export of viru-
lence genes and shiga toxin by membrane vesicles of
Escherichia coli O157:H7. Applied and Environmental
Microbiology, 65(5), 1843-1848.
[24] Moayeri, M. and Welch, R. (1997) Prelytic and lytic
conformation of erythrocyte-associated Escherichia coli
hemolysin. Infection and Immunity, 65(6), 2233-2239.
[25] Boehm, D., Welch, R. and Snyder, I. (1990) Calcium is
requiered for binding of Escherichia coli hemolysin
(HlyA) to erythrocyte membrane. Infection and Immunity,
58(6), 1951-1958.
[26] Markwell, M.A., Haas, S.M., Bieber, L.L. and Tolbert, N.E.
(1978) A modification of the lowry procedure to simplify
protein determination in membrane and lipoprotein samples.
Analytical Biochemistry, 87(1), 206-210.
[27] Karkhanis, Y.D., Zeltner, J.Y., Jackson, J.J. and Carlo, D.J.
(1978) A new and improved microassay to determine 2-
keto-3-deoxyoctonate in lipopolisaccharide of Gram-nega-
tive bacteria. Annals of Clinical Biochemistry, 85, 595-601.
[28] Chen, P., Toribara, T. and Warner, H. (1956) Microdeter-
mination of phosphorus. Annals of Chemistry, 28, 1756-
[29] Laemmli, U.K. (1970) Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature,
227(5259), 680-685.
[30] Fomsgaard, A., Freudenberg, M. and Galanos, C. (1990)
Modification of the silver staining technique to detect
lipopolysaccharide in polyacrylamide gels. Journal of
Clinical Microbiology, 28(12), 2627-2631.
[31] Towbin, H., Staehelin, T. and Gordon, J. (1979) Electro-
phoretic transfer of proteins from polyacrylamide gels to
nitrocellulose sheets: Procedure and some applications.
Proceedings of the National Academy of Sciences, 76(9),
[32] Snyder, I.S. and Zwadyk, P. (1969) Some factors affecting
production and assay of Escherichia coli hemolysin. Journal
of General Microbiology, 5, 133-143.
[33] Soloaga, A., Ramírez, J.M. and Goñi, F.M. (1998) Rever-
sible denaturation, self-aggregation, and membrane activity
of Escherichia coli alpha-hemolysin, a protein stable in 6 M
urea. Biochemistry, 37(18), 6387-6393.
[34] Mayer, L.D., Hope, M.J. and Cullis, P.R. (1986) Vesicles of
variable sizes produced by a rapid extrusion procedure.
Biochimica et Biophysica Acta, 858, 161-168.
[35] Struck, D.K., Hoekstra, D. and Pagano, R.E. (1981) Use of
resonance energy transfer to monitor membrane fusion.
Bioc hemistry, 20, 4093-4099.
[36] Herlax, et al, unpublished results.
[37] Rennie, R.P. and Arbuthnott, J.P. (1974) Partial charac-
terization of Escherichia coli haemolysin. Journal of Medi-
cal Microbiology, 7(2), 179-188.
[38] Ostolaza, H., Bartolome, B., Ortiz de Zarate, I., de la Cruz, F.
and Goñi, F.M. (1993) Release of lipid vesicle contents by
the bacterial protein toxin alpha-haemolysin. Biochimica et
Biophysica Acta, 1147, 81-88.
[39] Demuth, D.R., James, D., Kowash, Y. and Kato, S. (2003)
Interaction of Actinobacillus actinomycetemcomitans outer
membrane vesicles with HL60 cells does not require leuko-
toxin. Cellular Microbiology, 5(2), 111-121.
[40] Lally, E.T., Golub, E.E. and Kieba, I.R. (1991) Structure and
function of the B and D genes of the Actinobacillus action-
mycetemcomitans leukotoxin complex. Microbial Patho-
genesis, 11, 111-121.
[41] Herlax, V., Maté, S., Rimoldi, O. and Bakás, L. (2009)
Relevance of fatty acid covalently bound to Escherichia coli
alpha-hemolysin and membrane microdomains in the oligo-
merization process. Journal of Biological Chemistry, 284,
[42] Cortajarena, H., Goñi, F. and Ostolaza, H. (2001) Glycoph-
orin as a receptor for Escherichia coli alphahemolysin in
erythrocytes. Journal of Biological Chemistry, 276(16),
[43] Valeva, A., Walev, I., Kemmer, H., Weis, S., Siegel, I.,
Boukhallouk, F., Wassenaar, T., Chavakis, T. and Bhakdi, S.
(2005) Binding of Escherichia coli hemolysin and Activa-
tion of the target cells is not receptor-dependent. Journal of
Biological Chemistry, 280(44), 36657-36663.
[44] Herlax, V., Tacconi de Alani, M. and Bakás, L. (2005) Role
of lipopolysaccharide on the structure and function of alpha-
hemolysin from Escherichia coli. Chemistry and Physics of
Lipids, 135(2), 107-115.
[45] Ostolaza, H., Bartolome, B., Serra, J.L., Cruz, F. and Goñi,
F.M. (1991) Alpha-haemolysin from E. coli. Purification
and self-aggregation properties. Microbiology Letters,
280(2), 195-198.
[46] Czuprynski, C.J. and Welch, R.A. (1995) Biological effects
of RTX toxins: The possible role of lipopolysaccharide.
Trends in Microbiology, 3(12), 480-483.