Dual mechanism of bacterial lethality for a cationic sequence-random copolymer that mimics host-defense antimicrobial peptides.
ABSTRACT Flexible sequence-random polymers containing cationic and lipophilic subunits that act as functional mimics of host-defense peptides have recently been reported. We used bacteria and lipid vesicles to study one such polymer, having an average length of 21 residues, that is active against both Gram-positive and Gram-negative bacteria. At low concentrations, this polymer is able to permeabilize model anionic membranes that mimic the lipid composition of Escherichia coli, Staphylococcus aureus, or Bacillus subtilis but is ineffective against model zwitterionic membranes, which explains its low hemolytic activity. The polymer is capable of binding to negatively charged vesicles, inducing segregation of anionic lipids. The appearance of anionic lipid-rich domains results in formation of phase-boundary defects through which leakage can occur. We had earlier proposed such a mechanism of membrane disruption for another antimicrobial agent. Experiments with the mutant E. coli ML-35p indicate that permeabilization is biphasic: at low concentrations, the polymer permeabilizes the outer and inner membranes; at higher polymer concentrations, permeabilization of the outer membrane is progressively diminished, while the inner membrane remains unaffected. Experiments with wild-type E. coli K12 show that the polymer blocks passage of solutes into the intermembrane space at high concentrations. Cell membrane integrity in E. coli K12 and S. aureus exhibits biphasic dependence on polymer concentration. Isothermal titration calorimetry indicates that the polymer associates with the negatively charged lipopolysaccharide of Gram-negative bacteria and with the lipoteichoic acid of Gram-positive bacteria. We propose that this polymer has two mechanisms of antibacterial action, one predominating at low concentrations of polymer and the other predominating at high concentrations.
- [Show abstract] [Hide abstract]
ABSTRACT: Antimicrobial peptides (AMPs) are multifunctional compounds that may show antimicrobial and immunomodulatory activities. With the rapid increase in the incidence of multidrug-resistant bacteria, there is an enormous interest in AMPs as templates for the production of new antibiotics. However, there are concerns that the therapeutic administration of AMPs can select resistant strains. In order to distinguish between resistant and non-resistant strains and verify resistance specificity to AMPs, in this study a magainin I-resistant Escherichia coli model was used. First, the identity of all strains was confirmed by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF)-MS, VITEK 2 and MicroScan, and the susceptible and magainin-resistant strains were successfully differentiated by MALDI-TOF-MS analysis. Furthermore, cross-resistances to a broad spectrum of antibiotics were evaluated, showing that all E. coli strains are susceptible to the drugs tested, suggesting that the resistance seems to be specific to AMPs. Finally, the specific resistance to magainin I compared with other AMPs was checked by microdilution. This experiment showed that the magainin MICs were 62 and 104 μM for susceptible and resistant strains, respectively. The other AMPs MICs were 3.4 μM to proline-arginine-rich 39-amino-acid peptide, 43 μM to porcine myeloid antimicrobial 23-amino-acid peptide-23 and 1.2 μM to cecropin P1 for all strains, demonstrating any additional resistance to peptides here evaluated, confirming that the resistance seems to be essentially specific to magainin I. In summary, the data reported here reinforce the proposal that magainin I seems not to be merely a membrane disruptor, probably showing additional molecular targets in pathogenic bacteria.The Journal of Antibiotics advance online publication, 7 May 2014; doi:10.1038/ja.2014.48.The Journal of Antibiotics 05/2014; · 2.19 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Novel synthetic mimics of antimicrobial peptides have been developed to exhibit structural properties and antimicrobial activity similar to those of natural antimicrobial peptides (AMPs) of the innate immune system. These molecules have a number of potential advantages over conventional antibiotics, including reduced bacterial resistance, cost-effective preparation, and customizable designs. In this study, we investigate a family of nylon-3 polymer-based antimicrobials. By combining vesicle dye leakage, bacterial permeation, and bactericidal assays with small-angle X-ray scattering (SAXS), we find that these polymers are capable of two interdependent mechanisms of action: permeation of bacterial membranes and binding to intracellular targets such as DNA, with the latter necessarily dependent on the former. We systemically examine polymer-induced membrane deformation modes across a range of lipid compositions that mimic both bacteria and mammalian cell membranes. The results show that the polymers' ability to generate negative Gaussian curvature (NGC), a topological requirement for membrane permeation and cellular entry, in model Escherichia coli membranes correlates with their ability to permeate membranes without complete membrane disruption and kill E. coli cells. Our findings suggest that these polymers operate with a concentration-dependent mechanism of action: at low concentrations permeation and DNA-binding occur without membrane disruption, while at high concentrations complete disruption of the membrane occurs.Biochimica et Biophysica Acta 04/2014; · 4.66 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Antimicrobial peptides (AMPs) are promising leads for novel antibiotics; however, their activity is often compromised under physiological conditions. The purpose of this study was to determine the activity of α-peptide/β-peptoid peptidomimetics and AMPs against Escherichia coli and Staphylococcus aureus in the presence of human blood-derived matrices and immune effectors. The minimum inhibitory concentration (MIC) of two peptidomimetics against E.coli decreased by up to one order of magnitude when determined in 50% blood plasma as compared to MHB media. The MIC of a membrane-active AMP, LL-I/3 also decreased, whereas two intracellularly acting AMPs were not potentiated by plasma. Blood serum had no effect on activity against E. coli and neither matrix had an effect on activity against S. aureus. Unexpectedly, physiological concentrations of human serum albumin did not influence activity. Plasma potentiation was not caused by an LL-37 analogue, lysozyme or hydrogen peroxide; however, plasma potentiation of activity was abolished when the complement system was heat-inactivated. Time-course experiments indicated that potentiation was due to plasma-mediated effects on bacterial cells prior to activities of peptidomimetics. The unexpected enhancement of antibacterial activity of peptidomimetics and AMPs under physiological conditions significantly increases the therapeutic potential of these compounds.Research in Microbiology 08/2013; · 2.89 Impact Factor
Dual Mechanism of Bacterial Lethality for a Cationic
Sequence-Random Copolymer that Mimics
Host-Defense Antimicrobial Peptides
Raquel F. Epand1, Brendan P. Mowery2, Sarah E. Lee2,
Shannon S. Stahl2, Robert I. Lehrer3, Samuel H. Gellman2
and Richard M. Epand1⁎
1Biochemistry and Biomedical
Sciences, McMaster University,
Hamilton, ON, Canada L8N
2Department of Chemistry,
University of Wisconsin,
Madison, WI 53706, USA
3Department of Medicine, David
Geffen School of Medicine,
University of California, Los
Angeles, Los Angeles, CA
Received 11 February 2008;
received in revised form
20 March 2008;
accepted 21 March 2008
28 March 2008
Flexible sequence-random polymers containing cationic and lipophilic
subunits that act as functional mimics of host-defense peptides have
recently been reported. We used bacteria and lipid vesicles to study one
such polymer, having an average length of 21 residues, that is active
against both Gram-positive and Gram-negative bacteria. At low concen-
trations, this polymer is able to permeabilize model anionic membranes
that mimic the lipid composition of Escherichia coli, Staphylococcus aureus, or
Bacillus subtilis but is ineffective against model zwitterionic membranes,
which explains its low hemolytic activity. The polymer is capable of
binding to negatively charged vesicles, inducing segregation of anionic
lipids. The appearance of anionic lipid-rich domains results in formation of
phase-boundary defects through which leakage can occur. We had earlier
proposed such a mechanism of membrane disruption for another
antimicrobial agent. Experiments with the mutant E. coli ML-35p indicate
that permeabilization is biphasic: at low concentrations, the polymer
permeabilizes the outer and inner membranes; at higher polymer
concentrations, permeabilization of the outer membrane is progressively
diminished, while the inner membrane remains unaffected. Experiments
with wild-type E. coli K12 show that the polymer blocks passage of solutes
into the intermembrane space at high concentrations. Cell membrane
integrity in E. coli K12 and S. aureus exhibits biphasic dependence on
polymer concentration. Isothermal titration calorimetry indicates that the
polymer associates with the negatively charged lipopolysaccharide of
Gram-negative bacteria and with the lipoteichoic acid of Gram-positive
bacteria. We propose that this polymer has two mechanisms of
antibacterial action, one predominating at low concentrations of polymer
and the other predominating at high concentrations.
© 2008 Elsevier Ltd. All rights reserved.
Edited by I. B. Holland
Keywords: cell wall; antimicrobial polymer; bacterial encapsulation;
membrane permeability; lipopolysaccharide
*Corresponding author. E-mail address: firstname.lastname@example.org.
Abbreviations used: LUV, large unilamellar vesicle; DOPE, 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine; DOPG,
1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]; CL, cardiolipin; DOPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine;
ITC, isothermal titration calorimetry; ANTS, 8-aminonaphthalene-1,3,6,trisulfonic acid; DPX, p-xylene-bis-pyridinium
bromide; SUV, small unilamellar vesicle; DSC, differential scanning calorimetry; DPPE, 1,2-dipalmitoyl-sn-glycero-3-
phosphoethanolamine; LPS, lipopolysaccharide; LTA, lipoteichoic acid; OM, outer membrane; IM, inner membrane;
ONPG, o-nitrophenyl-3-D-galactoside; MIC, minimal inhibitory concentration; MBC, minimal bactericidal concentration;
MLV, multilamellar vesicle; POPE, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine; TSB, tryptic soy broth.
doi:10.1016/j.jmb.2008.03.047 J. Mol. Biol. (2008) 379, 38–50
Available online at www.sciencedirect.com
0022-2836/$ - see front matter © 2008 Elsevier Ltd. All rights reserved.
The prevalence of pathogenic bacteria that are
resistant to conventional antibiotic therapies has
inspired a widespread search for new antibacterial
agents. Host-defense peptides are attractive in this
regard because it appears to be difficult for bac-
teria to develop resistance to these natural anti-
biotics.1–4Extensive efforts have been made to
optimize the biological activity of antibacterial
peptides5and other sequence-specific oligomers
that are designed to mimic host-defense peptides.6
However, the preparation of oligomers with a
specific sequence of subunits, including peptides,
requires stepwise synthetic methods, which are
very costly, representing a significant barrier to the
use of these materials.7
If host-defense peptide mimics could be created
via polymerization, then these materials might be
more amenable to biomedical application. Antibac-
terial polymers have been reported,8–12but it is not
clear whether these materials display the distinctive
activity profile of host-defense peptides, selectively
disrupting bacterial cell membranes in preference to
eukaryotic cell membranes. Synthetic polymers
tend to be nonselective in their membrane-disrupt-
ing effects.13–15Recently, we reported a new class of
random copolymers that reproduce the selectivity
of natural host-defense peptides for bacterial mem-
branes.16These materials are prepared by ring-
opening polymerization of β-lactams. Variation of
the cationic/lipophilic ratio, length, and other
features allowed us to identify short polymers
that are active against both Gram-positive and
Gram-negative bacteria but that exhibit a low
propensity for lysis of human red blood cells
(hemolysis). Here we characterize in greater detail
the behavior of one of these polymers with an
average length of 21 subunits and an average
molecular weight of 2800 (Mn/Mw=1.4) as well as
m:n=2:3 (referred to as 360in Ref. 16) (the structure
of this polymer is shown below). We explored the
interaction of this polymer with synthetic lipid
vesicles and with bacterial membranes. Knowledge
acquired with vesicle model systems has been
useful for elucidating the mechanism of action of
Fig. 1. Leakage of aqueous contents from vesicles (ANTS/DPX assay): Experiments were carried out at 37 °C in 10 mM
Hepes buffer (pH 7.4, 0.14 M NaCl and 1 mM EDTA). Varying amounts of polymer were added at 20 s. Upper left panel:
DOPC LUVs (20 μM) were used. Lubrol LX was added after 220 s to obtain a maximum value of vesicle permeabilization.
Upper right panel: LUVs of DOPC:DOPE:CL (12:84:4; 50 μM) were placed in the cuvette. The following concentrations of
polymer were added: curve 1, 0.63 μM; curve 2, 0.8 μM; curve 3, 1 μM; curve 4, 1.25 μM; curve 5, 5 μM. Detergent addition
is not shown. Lower left panel: LUVs of DOPE:DOPG (80:20; 50 μM) were placed in the cuvette. Added polymer
concentrations were as follows: curve 1, 0.2 μM; curve 2, 0.45 μM; curve 3, 0.63 μM; curve 4, 1.63 μM. Lubrol LX was
added after 220 s to obtain a maximum value of vesicle permeabilization. Lower right panel: LUVs of DOPG:CL (58:42;
50 μM) were placed in the cuvette. Added polymer concentrations were as follows: curve 1, 0.63 μM; curve 2, 0.8 μM;
curve 3, 1.25 μM. Detergent addition is not shown.
Dual Mechanism of Bacterial Lethality
this polymer against Gram-positive and Gram-
The polymer efficiently permeabilizes anionic
vesicles with compositions mimicking those of
The lipid compositions of bacterial membranes
differ widely and can influence susceptibility to
antimicrobial agents.17Leakage was measured with
large unilamellar vesicles (LUVs) having composi-
tions approximating those of the membranes of
three bacterial species. We used liposomes of the
following lipid mixtures to mimic the membrane
compositions of several different bacteria: Escher-
ichia coli [1,2-dioleoyl-sn-glycero-3-phosphoethano-
rac-(1-glycerol)] (DOPG)=80:20], Staphylococcus aur-
eus [DOPG:cardiolipin (CL)=58:42], and Bacillus
subtilis (DOPE:DOPG:CL=12:84:4) (Fig. 1). Each of
these types of vesicles has a net negative surface
charge because DOPG and CL have anionic head
groups. At very low concentrations of polymer (3–
5 μg/mL), these anionic bilayers are disrupted
within a very narrow range of mole fractions of
polymer (Fig. 2). Disruption occurs with similar
efficiency for all the anionic lipid compositions
studied, indicating that the polymer can act equally
well on cytoplasmic membranes of Gram-negative
as well as Gram-positive bacteria. In contrast, the
polymer is not able to permeabilize 1,2-dioleoyl-sn-
glycero-3-phosphocholine (DOPC) vesicles, which
do not have a net charge, even at a concentration as
high as 60 μg/mL.
The polymer binds anionic phospholipid
vesicles but not zwitterionic vesicles
Isothermal titration calorimetry (ITC) of small
unilamellar vesicles (SUVs) composed of DOPE:
DOPG (80:20) with the polymer is biphasic (Fig. 3,
left panel). Vesicle leakage data using the 8-
aminonaphthalene-1,3,6,trisulfonic acid (ANTS)/p-
xylene-bis-pyridinium bromide (DPX) assay with
50 μM phospholipids as LUVs indicate that addition
to vesicles of polymer at concentrations correspond-
ing to the first four to six ITC injections results in
significant bilayer permeabilization. Given that the
polymer is cationic and the vesicles contain an
anionic lipid, charge neutralization must occur
when the polymer interacts with the vesicles. Since
ion pairing should be exothermic, other compensa-
tory endothermic events must occur simultaneously
to result in a net endothermic signature in the ITC
measurement. A reorganization of the hydrogen
bonding network among phosphatidylethanola-
mine head groups could contribute to the observed
endotherm. ITC data indicate no binding of polymer
to DOPC SUVs (Fig. 3, right panel), in contrast to the
behavior seen with DOPE:DOPG SUVs. The lack of
binding to DOPC SUVs explains the lack of DOPC
LUV permeabilization seen in leakage experiments
and is consistent with the low degree of hemolysis
exhibited by the polymer.16
The polymer causes phase separation in anionic
phospholipid mixtures, clustering negative
Interactions with negatively charged lipids have
been extensively studied in antimicrobial pep-
tides.18,19Here, differential scanning calorimetry
(DSC) experiments were performed to determine if
the polymer is capable of not only binding to but
also clustering negatively charged lipid head
groups. We previously proposed this mechanism
to explain the action of another antimicrobial
Two lipid mixtures, which differed in the tempera-
tures of their phase transitions, were studied here.
High-melting phospholipid mixture
The uppermost and lowermost curves in Fig. 4
show heating and cooling cycles, respectively, for
the 2:1 dipalmitoyl-sn-glycero-3-phosphoethanola-
mine (DPPE):CL lipid mixture in the absence of
polymer. These data are shown as reference points
for interpretation of the traces in the middle of Fig. 4,
which were acquired in the presence of the polymer.
DPPE is a zwitterionic lipid that in pure form has
a thermal transition at 62 °C; addition of the
anionic lipid CL, a lipid that by itself has no phase
transition in the measurable temperature range,
results in a lipid mixture with a broad transition at
Fig. 2. Polymer concentration dependence of leakage
at 200 s in 50 μM vesicles composed of (▪) DOPG:DOPE
CL (12:84:4) LUVs, (●) DOPE:DOPG (80:20) LUVs, and
(▴) DOPG:CL (58:42) LUVs.
Dual Mechanism of Bacterial Lethality
lower temperatures. Subsequent addition of the
polymer to this lipid mixture (the polymer has no
phase transition by itself) causes the appearance of
a more cooperative phase transition at higher
temperature and of higher enthalpy, corresponding
closely to that of pure DPPE (Fig. 4). This dramatic
alteration in the phase transition of the DPPE:CL
lipid mixture results from segregation of the
negatively charged CL from the zwitterionic
DPPE as a result of binding to the cationic polymer.
We recently reported an analogous lipid segrega-
tion phenomenon in another system.20
The polymer effect is irreversibly lost when the
highest temperature is reached in the first heating
cycle. The origin of this loss of polymer effect is
unclear; it is unlikely that heating to 70 °C causes
chemical decomposition of this polymer.
Low-melting phospholipid mixture
Because of the irreversibility of the DSC scans with
the high-melting lipid mixture, similar experiments
were performed with a lower melting mixture, 4:1
lamine (POPE):CL. Pure POPE has a phase transition
at 25 °C, but the mixture of POPE and CL has a
Addition of the polymer causes dramatic changes in
the DSC curves, resulting in the appearance of a
transitioncomponent athighertemperature thanthat
observed with the pure lipid mixture (Fig. 5). These
more enriched in phosphatidylethanolamine.
As a consequence of polymer-induced phase
separation, the clustering of anionic charge would
lead to phase-boundary defects in the membrane
through which leakage can occur. Interaction of the
polycationic peptide poly-L-Lys causes increased
ordering of the acyl chains of anionic lipids.21This
would result in a mismatch in bilayer thickness
between anionic lipid clustered by the polymer and
the remainder of the membrane. This is analogous
to gel and liquid crystalline phase-boundary defects
that have been suggested to be responsible for the
increased leakage of liposomes at the phase
transition temperature where domains coexist.22
We previously presented evidence for such a
mechanism for a discrete oligomer that functions
as an antimicrobial peptide mimic.20It can be
anticipated that leakage through phase-boundary
defects would be more restrictive with regard to the
size of the solute that could traverse the membrane,
but it would be sufficient to alter ion and
electrochemical gradients. These phase-boundary
defects are different from those in which a defect
was caused by insertion of a peptide or polymer
into a membrane,23and it would be an indirect
consequence of changes in the arrangement of
membrane lipids. Studies are underway to further
characterize this phenomenon in natural host-
Fig. 3. ITC of SUVs titrated with 100 μM polymer at 30 °C. Upper curves show heat flow for each injection
(microcalories per second) as a function of time (minutes), and lower curves show the integration of each peak
(kilocalories per mole injectant) as a function of polymer-to-lipid molar ratio. Left panel: ITC of 50 μM SUVs composed of
DOPE:DOPG (80:20) titrated with polymer in the syringe. Right panel: ITC of 50 μM SUVs composed of DOPC titrated
with polymer in the syringe.
Dual Mechanism of Bacterial Lethality
The ability shown by this flexible sequence-
random polymer to cluster anionic charge supports
the idea that formation of conformationally irre-
gular but globally amphiphilic molecular surfaces
can be induced in the presence of an anionic
ITC studies show strong exothermic interactions
between the polymer and lipopolysaccharide or
ITC titrations were performed to assess the mode
of interaction between the polymer and lipopoly-
saccharide (LPS), which is the main component of
the outer wall in Gram-negative bacteria, and
between the polymer and lipoteichoic acid (LTA),
which is the main component of the peptidoglycan
layer that surrounds Gram-positive bacteria. (The
LPS and LTA structures can be found in Refs. 24
and 25, respectively.) It has been suggested that ITC
analysis of LPS reports on the state of order of the
acyl chains in lipid A.26The polymer interactions
with LPS and LTA result in very different thermo-
LPS from E. coli 0111:B4 was titrated with polymer
at 38 °C (Fig. 6, left panel). This temperature is above
the phase transition of this LPS sample, which was
determined by DSC to occur at 27 °C (data not
shown). Binding of the polymer to LPS results in a
biphasic exothermic process.
LTA from S. aureus does not exhibit a thermotropic
phase transition. The titration of LTA with the
polymer at 30 °C (Fig. 6, right panel) shows a large
exothermic process at the start, likely resulting from
charge interactions, and an appearance of an
endothermic process in the end. LTA is part of the
negatively charged peptidoglycan matrix of Gram-
positive bacteria, which is much thicker than the
peptidoglycan layer in E. coli. LTA forms supramo-
lecular micellar structures in aqueous dispersions,
unlike LPS, which favors a lamellar arrangement.27
ITC shows that the polymer binds strongly to both
LPS and LTA, but the thermodynamics of the
binding are different for these two substances.
The polymer permeabilizes the outer membrane
of E. coli ML-35p in a biphasic manner
The mutant E. coli ML-35p is constitutive for
cytoplasmic β-galactosidase, lacks lac permease, and
expresses a plasmid-encoded periplasmic β-lacta-
mase. Two chromogenic reporter molecules were
used to monitor permeabilization of the outer
membrane (OM) and the inner membrane (IM) in a
single assay.28–30Nitrocefin, a chromogenic cepha-
losporin, cannot cross the OM and is excluded from
the periplasmic space.However,permeabilization of
the OM allows nitrocefin to enter the periplasm
Fig. 5. DSC scans with MLVs of POPE:CL (80:20).
Same conditions as in Fig. 4. Addition of polymer to an
aliquot of MLVs that was removed from the cell after three
scan cycles. The mixture was returned to the calorimeter
cell and rescanned for three more cycles. Lipid/polymer
molar ratio was 32.
Fig. 4. DSC carried out with 2.5 mg/mL of DPPE:CL
(2:1) at a scan rate of 0.75°/min in the absence and
presence of polymer in 20 mM Pipes (pH 7.4, 0.14 M NaCl
and 1 mM EDTA). After three cycles of heating and
cooling with the lipid mixture alone, the sample was
removed from the calorimeter cell and the polymer was
added and then placed back in the cell for three more
cycles of heating and cooling. Odd numbers represent
heating scans, and even numbers represent cooling scans.
The first heating scan of the lipid mixture with the
polymer shows that it has segregated negative charge,
causing the appearance of a phase transition correspond-
ing to a domain highly enriched in DPPE. Lipid/polymer
molar ratio was 32.
Dual Mechanism of Bacterial Lethality
where itscleavageby a β-lactamase produces acolor
change that can be monitored spectrophotometri-
cally at 486 nm. The reaction is described below:31
Permeabilization begins a few minutes after
treatment with either the polymer or melittin
(Fig. 7). Experiments with nitrocefin carried out
with the host-defense peptide LL-37, a human
cathelicidin-associated antimicrobial agent,32sho-
wed that it also permeabilized the OM of E. coli
ML-34p within a few minutes after its addition.
The extent of permeabilization increases as the
amount of polymer increases up to 25 μg/mL, but
permeabilization is decreased at higher polymer
concentrations; in contrast, there is no significant
decrease in permeabilization at higher concentra-
tions of melittin (Fig. 8). Control experiments show
that the enzymatic activity of β-lactamase in
hydrolyzing nitrocefin is not affected by 100 μg/
mL of polymer (data not shown).
This assay is usually followed for only 20–30 min
after adding nitrocefin, since at longer times a small
progressive increase in the amount of nitrocefin
hydrolyzed is seen in the absence of polymer (Fig.
7), which could be caused by trace amounts of β-
lactamase appearing in the medium. The blank
corresponds to bacteria alone, with no nitrocefin or
polymer present, and is used to monitor any
absorbance change due to increased turbidity
resulting from the proliferation of bacteria. The
reaction was monitored for up to 60 min in our work
because the IM permeabilization is slow and does
not begin until after 30 min (see below). However,
the biphasic effect of polymer concentration is
observed at all times throughout the full 60 min.
this imide is similar to a β-lactam. It is unlikely that
has been discussed in our previous publication.14
Low polymer concentrations permeabilize the IM
of E. coli ML-35p
Because the mutant E. coli ML-35p has no
lactose permease, o-nitrophenyl-3-D-galactoside
(ONPG) cannot traverse its IM to be cleaved by
cytoplasmic β-galactosidase to o-nitrophenol,
unless permeabilization of the IM occurs. ONPG
cleavage produces a color change that can be
Fig. 7. OM permeabilization of E. coli ML-35p caused
by the polymer (upper panel) or melittin (lower panel) as a
function of time at different concentrations (0–100 μg/mL)
and at 37 °C. Hydrolysis of nitrocefin by β-lactamase was
followed by absorbance at 486 nm measured every 2 min.
Fig. 6. ITC of LPS or LTA, with 100 μM polymer in the syringe, in 10 mM Hepes and 0.14 M NaCl, pH 7.4, 10 μL/
injection. Plots show heat flow (microcalories per second) as a function of time (minutes). Left panel: 400 μg/mL of LPS
from E. coli 0111:B4. Right panel: 400 μg/mL of LTA from S. aureus.
Dual Mechanism of Bacterial Lethality
measured spectrophotometrically at 420 nm, and
the reaction is as follows:
Simultaneous measurement of permeabilization
of the OM and IM in E. coli ML-35p indicates that the
IM is permeabilized only at very low concentrations
of polymer (Fig. 9, upper panel). Thus, the extent of
the reaction of ONPG with β-galactosidase is
reduced compared with the membrane-disrupting
peptide melittin (Fig. 9, lower panel). The bulk
concentration of polymer in the medium is signifi-
cantly higher than that which reaches the IM
because some polymer is excluded by the cell wall
and/or bound to LPS. In addition, the process is
much slower than with melittin, as it takes ∼30 min
for the polymer to begin permeabilizing the IM,
compared with 5–10 min for melittin and ∼20 min
for LL-37 (at 5 μg/mL).32Although there is only a
small amount of ONPG influx, it does demonstrate
that the polymer is capable of damaging the IM at
low bulk polymer concentrations. Furthermore, we
demonstrate that the effectiveness of the polymer
decreases with increasing concentration.
Control experiments show that the enzymatic
activity of β-galactosidase with ONPG to produce o-
nitrophenol is not inhibited in the presence of
100 μg/mL of polymer. Therefore, the lower extent
of the permeabilization reaction of the IM of E. coli
ML-35p with increasing concentrations of polymer
cannot be attributed to inhibition of the β-galacto-
sidase activity (data not shown).
After 30 min, there is a small amount of hydrolysis
of ONPG in the absence of cationic peptides that
could be caused by some dying bacteria increasing
the permeability of the IM with time and allowing
some diffusion of molecules through the IM.
However, at polymer concentrations N25 μg/mL,
the time dependence curves of the polymer are
superimposed to those obtained for a sample that
does not contain polymer, showing lack of polymer-
induced ONPG uptake across the IM. The blank
included in the graph corresponds to bacteria alone,
with no ONPG or polymer present.
High concentrations of the polymer can block
active transport of ONPG in wild-type E. coli K12
This assay was carried out to further probe the
ability of the polymer to prevent IM permeabiliza-
tion at high concentrations. A wild-type E. coli in a
medium containing nutrients and energy sources
was employed. Wild-type E. coli K12 was grown
under conditions that allow ONPG to cross the IM
by active transport via lactose permease. When
bacteria are given ONPG, one can determine
whether the polymer inhibits active transport of
this solute into E. coli. Bacteria treated with up to
25 μg/mL of polymer show enzymatic cleavage of
ONPG, but at higher polymer concentrations, this
of ONPG is detected at all the concentrations of
melittin examined (Fig. 10). All the concentrations of
melittin and polymer at which ONPG is transported
produce the same overlapping curves because the
same amount of ONPG is transported across the IM.
Hence, for these conditions, the passage of ONPG
across the OM is not the rate-determining step and
flux across the IM is determined by the lactose
permease and is independent of the presence of
Fig. 9. IM permeabilization of E. coli ML-35p caused by
the polymer (upper panel) or melittin (lower panel) as a
function of time (for 60 min) at different concentrations (0–
100 μg/mL) and at 37 °C. Reaction of ONPG with β-
galactosidase was measured every 2 min by absorbance at
Fig. 8. OM permeabilization of E. coli ML-35p caused
by the polymer (▪) or melittin (▴) as a function of polymer
concentration (0–100 μg/mL) after 60 min. Hydrolysis of
nitrocefin by β-lactamase was followed by absorbance at
Dual Mechanism of Bacterial Lethality
polymer or melittin. However, at concentrations
N25 μg/mL, the polymer effectively blocks transport
of ONPG. Lack of hydrolysis of ONPG at high
concentrations of polymer is not caused by meta-
bolic inhibition of the permease because the experi-
ments described above with the mutant ML-35p,
which lacks lac permease, showed the same effect.
Rather,itis the result ofONPG notbeingable to pass
through the OM to access the periplasmic space,
similarly to what was observed above with nitroce-
fin at high concentrations of polymer.
This assay confirms the observations made in the
mutant E. coli ML-35p. Since permeabilization of the
OM at concentrations N25 μg/mL of polymer
becomes progressively diminished (Fig. 7, upper
panel), ONPG is prevented from gaining access to
the intermembrane space and thus from crossing the
IM to undergo a cleavage reaction.
Membrane integrity in Gram-positive and
Gram-negative bacteria in the presence of
increasing concentrations of polymer
We used well-known dyes to assess bacterial
membrane integrity. Membrane-permeant SYTO 9
labels DNA of all bacterial cells, those with intact cell
membranes and those with damaged cell mem-
branes. Propidium iodide penetrates only bacteria
with damaged membranes causing displacement of
SYTO 9 from DNA. However, loss of SYTO 9 signal
under these conditions may also arise from quench-
ing by fluorescence resonance energy transfer from
SYTO 9 to propidium iodide,33augmenting the
signal that results from loss of membrane integrity.
When increasing amounts of polymer were added to
Gram-negative (E. coli) or Gram-positive (S. aureus)
bacteria, there was an initial increase in membrane
of up to 25 μg/mL of polymer, but above this
concentration, much less propidium iodide appears
to reach the bacterial DNA. Fluorescence resonance
energy transfer could contribute to the observed
fluorescence quenching but would not result in a
of ONPG, even though it may not be a good
quantitative measure of loss of bacterial integrity.
low concentrations of polymer, but at concentrations
of polymer N25 μg/mL, bacterial toxicity occurs by a
different mechanism that we show is through
blockage of transport through the OM.
Biphasic curves were obtained in wild-type Gram-
negative as well as Gram-positive bacteria. We have
shown that the polymer is capable of interacting
electrostatically with LTA, a component of the thick
peptidoglycan layer of Gram-positive bacteria. This
Fig. 10. Entry of ONPG in E. coli K12, grown in the
absence of lactose, in the presence of increasing concen-
trations of polymer (upper panel) or melittin (lower panel)
at 37 °C. In the upper panel, concentrations were as
follows: (1) 0, 10, and 25 μg/mL of polymer (super-
imposed); (2) 50 μg/mL of polymer; and (3) 100 μg/mL of
polymer. In the lower panel, concentrations were (1) 0, 10,
25 50, and 100 μg/mL of melittin (superimposed).
Fig. 11. Percentage of membrane integrity in E. coli K12
(upper panel) and S. aureus (lower panel) measured with
the BacLight assay in the presence of increasing concen-
trations of polymer, as described in Methods. Values were
obtained from a standard curve.
Dual Mechanism of Bacterial Lethality
assay shows that high concentrations of polymer are
retained in this layer and do not reach the cytoplas-
Minimal inhibitory concentration and minimal
The minimal inhibitory concentration (MIC) for
the polymer in the mutant E. coli ML-35p of 7 μg/
mL was determined as described in Methods (i.e.,
this MIC value corresponds to the concentration of
polymer required for 50% inhibition of bacterial
growth). Previously determined values of MIC for E.
coli K12 and S. aureus were 13 and 25 μg/mL,
respectively16; in that study, MIC was defined as the
lowest concentration that causes 100% inhibition of
growth.Theexpression of a periplasmic β-lactamase
and the lack of lactose permease do not greatly affect
the MIC of the E. coli mutant ML-35p. Minimal
bactericidal concentration (MBC) was found to be 13
and 15 μg/mL for E. coli ML-35p and E. coli K12,
respectively. MIC values for melittin were recently
reported34and defined as the lowest concentration
that causes 50% inhibition of growth. The MIC
(MBC) values are 4 (4) μg/mL for E. coli K12 ATCC
23716 and ∼2 (2) μg/mL for S. aureus ATCC 43300.
The polymer we studied displays significant
antimicrobial activity against both Gram-positive
and Gram-negative bacteria. We have shown that,
like host-defense peptides, this polymer disrupts
model vesicles composed of anionic lipids with
compositions mimicking those of Gram-negative or
Gram-positive bacteria, at concentrations compar-
able with the MIC. The polymer does not disrupt
model zwitterionic vesicles, in accord with its low
hemolytic capacity (Figs. 1 and 2).
Analysis of polymer–vesicle interactions with
DSC suggests that the polymer can induce segrega-
tion of lipids in mixtures, resulting in the formation
of domains (Figs. 4 and 5). It has been suggested that
bilayer leakage can be induced by the formation of
boundary defects at the borders of lipid domains.35
The polymer's ability to cause lipid domain forma-
tion could be the basis for the polymer-induced
leakage from vesicles. Our hypothesis that mem-
brane leakage arises from induced domain forma-
tion and the resulting defects at domain boundaries
is distinct from other mechanisms that have been
proposed to explain the permeability induced by
host-defense peptides in bilayer membranes.18Stu-
dies are underway to determine whether clustering
of anionic lipids and concomitant boundary defects
play a role in membrane disruption by natural host-
Our studies suggest that the polymer can exert
toxic effects on bacteria via a second mechanism as
well. Most of our mechanistic experiments have
involved the Gram-negative bacterium E. coli, and
our interpretation of the results from these studies
must account for the double membrane surrounding
these bacteria. The OM is asymmetrical. The
external leaflet is composed largely of LPS, which
consists of an oligosaccharide portion linked cova-
lently mainly to lipid A, while the inner leaflet of the
OM is a monolayer with a lipid composition similar
to that of the IM. A thin peptidoglycan layer
separates the OM and IM. LPS molecules in the
OM are interconnected by Mg2+ions to form an
ordered structure. In addition, the OM contains
porins, proteins that allow passage of small solutes
but exclude hydrophilic molecules larger than
We observed two distinct components of the
antimicrobial action of the polymer in Gram-nega-
tive bacteria, one being disruption of the IM and the
second resulting from interactions with the OM. We
can distinguish between changes in the permeability
to substrates for each of the two membranes of E. coli
using the specially constructed mutant ML-35p.29
We compared the polymer results with those
obtained simultaneously with melittin, a very lytic
and well-studied peptide toxin that permeabilizes
efficientlyboth the IM and the OM of Gram-negative
bacteria and with published results on LL-37.32It is
clear that at low concentrations (2.5–25 μg/mL of
polymer), there is permeabilization of the IM (Fig. 9),
and thus the polymer must have also traversed the
OM. At 25 μg/mL, there is a maximal flux across the
OM but the rate of permeabilization of the IM is
already less than maximal. This may be a conse-
quence of differences in the rates of permeation
across the OM of the nitrocefin versus the polymer. In
addition, there is some heterogeneity in the size of
the polymer that could result in the selective
transport of the lower-molecular-weight species.
Furthermore, not all of the polymer that passes the
OM necessarily reaches the IM. In fact, we would
expect some of the polymer to be trapped by binding
to the proteoglycan layer in Gram-negative bacteria.
contribution of blockage of the OM to the reduced
permeability across the IM at intermediate polymer
concentrations, but it is clear that at high polymer
concentration there is no permeationof the IM bythe
polymer, and this can be accounted for by blockage
of transport across the OM. Thus, the bactericidal
action of the polymer is a consequence of IM
permeabilization at low concentrations but is caused
by blockage of transport across the OM at high
polymer concentrations where the bacteria are killed
even though the cell membrane is not ruptured.
ITC results indicate that the cationic polymer
binds to LPS (Fig. 6, left panel). It has been pos-
tulated36that polycationic molecules associate elec-
trostatically with LPS and competitively displace
divalent cations, leading to a disorganization of the
OM, which then makes the IM accessible to these
Nitrocefin is the substrate used to detect induced
permeability of the OM of the mutant E. coli ML-
35p. Our results with nitrocefin, which normally
cannot penetrate the OM, show that the OM
Dual Mechanism of Bacterial Lethality
becomes permeable at polymer concentrations up
to 25 μg/mL (i.e., at concentrations comparable
with the MIC and MBC). However, when this
threshold concentration of polymer is reached, its
ability to cause OM permeability declines. We
suggest that high polymer concentrations lead to a
strong association with the LPS layer, which results
in a more compact OM, impairing movement of
water-soluble substances across the OM.
With E. coli ML-35p, ONPG hydrolysis indicates
that at concentrations of polymer b25 μg/mL,
permeabilization of the IM has occurred (at con-
centrations equivalent to those obtained for the MIC
and MBC). This E. coli strain contains no lactose
permease, so ONPG can reach β-galactosidase in the
cytoplasm only if another agent (e.g., the polymer or
melittin) compromises IM integrity. ONPG must
move from the medium across the OM into the
periplasmic space in order to gain access to the IM.
The data in Fig. 9 show that concentrations of
polymer up to 25 μg/mL allow ONPG to reach β-
galactosidase within E. coli ML-35p, but not higher
polymer concentrations. This biphasic concentration
effect is similar to what we observed with nitrocefin,
and the data in Fig. 9, as well as the data in Fig. 10
with wild-type E. coli, indicate that progressively
higher concentrations of the polymer cause dimin-
ished solute movement across the E. coli OM. Thus,
at high polymer concentrations, no ONPG reaction
occurs because the substrate is excluded from
reaching the IM. The BacLight assay provides
additional evidence for a biphasic effect of IM
permeabilization as a function of polymer concen-
tration (Fig. 11) in the absence of ONPG. Since the
bacteria fail to recover at high concentrations of
polymer, a second mechanism is required to explain
its activity in the absence of IM permeabilization.
Controls demonstrated that this polymer does not
inhibit galactosidase or permease and that the
polymer is not susceptible to the action of proteases
Gram-positive bacteria lack an OM but have a
much thicker peptidoglycan layer that includes LTA.
ITC studies with the polymer indicate that it binds
electrostatically to LTA (Fig. 6, right panel). With
Gram-positive bacteria, it appears that a dual
mechanism also occurs, as seen from the biphasic
response of membrane integrity in S. aureus (Fig. 11),
which could result from the polymer's electrostatic
interaction with LTA. Although the polymer is able
to reach and disrupt the cytoplasmic membrane at
low concentrations, at concentrations N25 μg/mL, a
tight association with the peptidoglycan layer could
result in impaired access of substrates to the cyto-
Some cationic polymers have been found to block
transport through porins.37,38Several porins have
been identified in E. coli.39The polymer we used in
this work is a heterogeneous mixture of molecular
weights. The lower-molecular-weight end of the
polymer is just at the exclusion limit of porins. It is
possible that the polymer could also block porin
access after a threshold concentration is reached.
Thus, blocking access to porins could be, in addition
to interaction with the LPS layer, a factor in
preventing passage of small molecules to the IM.
However, in Gram-positive bacteria, which lack
porins, a biphasic behavior as a function of polymer
concentration is also observed (Fig. 11), pointing to
strong association of the polymer with the pepti-
doglycan layer as likely cause for the observed
effects at higher polymer concentrations.
One of the most primitive host-defense mechan-
isms involves enveloping the pathogen to deprive it
of nutrients. This happens in trees (amber trees, for
example, encapsulate microbial or insect pests) and
insects, such as Drosophila melanogaster, which
exhibits a number of host-defense systems involving
blocking access of nutrients to microbial invaders.40
Thus, a mechanism involving the blockade of the
passage of nutrients at high concentrations of
polymer would be reminiscent of a known host-
In summary, the polymer we studied appears to
have two distinct mechanisms for killing bacteria.
At low concentrations, comparable with the MIC
and MBC, the polymer can cross the OM of
Gram-negative bacteria or the peptidoglycan layer
of Gram-positive bacteria to disrupt the cytoplas-
mic membrane. At higher concentrations, the
cytoplasmic membrane is not disrupted but the
bacteria do not recover viability. Instead, the
polymer associates strongly with the negatively
charged components of the cell wall or peptido-
glycan layer to make the bacteria impermeable to
solutes, a blockade with lethal consequences. Both
mechanisms are compatible with antimicrobial
Materials and Methods
Phospholipids were purchased from Avanti Polar
Lipids (Alabaster, AL). LPS O111:B4 from E. coli and
LTA from S. aureus were purchased from Sigma Chemical
Co. ANTS, DPX, and a BacLight Bacteria Viability Kit
were obtained from Molecular Probes (Invitrogen);
ONPG, penicillinase type IV from Enterobacter cloacae,
and β-galactosidase from E. coli were purchased from
Sigma Chemical Co. The polymer was prepared as pre-
Preparation of phospholipid vesicles
Lipid films were made by dissolving appropriate
amounts of lipids in chloroform/methanol 2:1 (v/v),
followed by solvent evaporation under nitrogen to
deposit the lipid as a film on the walls of a tube. Final
traces of solvent were removed in a vacuum chamber
attached to a liquid nitrogen trap for 3–4 h. Dried films
were kept under argon gas at −20 °C until used. Films
were hydrated with buffer and vortexed extensively to
make multilamellar vesicles (MLVs). To obtain LUVs, we
Dual Mechanism of Bacterial Lethality
subjected them to five cycles of freezing and thawing and
further processed them with 10 passes through two
stacked 100-nm polycarbonate filters (Nucleopore Filtra-
tion Products, Pleasanton, CA) in a high-pressure barrel
extruder (Lipex Biomembranes, Vancouver, BC, Canada).
LUVs were kept on ice and used within a few hours of
preparation. Lipid phosphorus was determined using the
method of Ames.41
Leakage of aqueous contents from liposomes
Lipid films were hydrated with 12.5 mM ANTS,
45 mM DPX, 68 mM NaCl, and 10 mM Hepes at
pH 7.4.42The osmolarity of the liposome solution was
adjusted to be equal to that of the buffer as measured by
a cryo-osmometer (Advanced Model 3 Plus Micro-
Osmometer, Advanced Instruments Inc., Norwood,
MA). After passage through a 2.5×20-cm column of
Sephadex G-75, the fractions collected in the void
volume were stored on ice and the concentration of
phospholipid was determined by phosphate analysis.
Fluorescence measurements were made in a quartz
cuvette containing 2 mL of Hepes buffer [pH 7.4,
10 mM Hepes, 140 mM NaCl, and 1 mM ethylenedia-
minetetraacetic acid (EDTA)] at 37 °C, with stirring.
LUVs were added to the cuvette to a final concentration
of 50 μM, and fluorescence was recorded as a function of
time with an SLM Aminco Bowman Series II spectro-
fluorimeter, using an excitation wavelength of 360 nm
and an emission wavelength of 530 nm with 8-nm
bandwidth slits. A 500-nm cutoff filter was placed in the
emission path. Leakage was initiated with addition of
polymer. The 100% value was obtained with the addition
of 20 μL of 20% Lubrol LX, followed by vigorous mixing
and brief sonication.
Measurements were made in a Nano II Differential
Scanning Calorimeter (Calorimetry Sciences Corp., Lin-
den, UT). The method was adapted from that reported
by Epand et al.17Briefly, MLVs were prepared by
hydration of films as described above. The lipid
suspension was placed in the calorimeter cell, scanned
for several cycles of heating and cooling at 0.75°/min,
and removed from the cell; a volume of polymer solution
was then added to an aliquot of lipid mixture. It was
placed again in the cell and scanned once more for
several cycles of heating and cooling. In control experi-
ments, Pipes buffer (pH 7.4, 20 mM Pipes, 140 mM NaCl,
and 1 mM EDTA) was added instead of polymer
solution. Curves were plotted with the program Origin
v.5.0 and analyzed with the fitting program DA-2
provided by MicroCal Inc. (Northampton, MA).
Titrations were performed in a VP-ITC instrument
(MicroCal Inc.). Polymer was placed in the syringe at a
concentration of 100 μM in Hepes buffer and 0.14 M
NaCl, pH 7.4. SUVs of DOPE and DOPG in an 80:20
molar ratio, or DOPC, all prepared by sonicating MLVs to
clarity, were placed in the calorimeter cell at a concentra-
tion of 50 μM. They were titrated with 10-μL injections of
polymer at 30 °C.
LPS and LTA at a concentration of 400 μg/mL in
10 mM Hepes and 0.14 M NaCl, pH 7.4, were titrated with
10-μL injections of polymer at 30 °C for LTA and 38 °C
A detailed thermodynamic analysis of the LPS and LTA
titrations was not carried out because of the high degree of
heterogeneity of the sugar chains in the preparations.
Data were analyzed with the program Origin v.5.0.
Permeabilization of the OM and IM of Gram-negative
The mutant E. coli ML-35p was employed in this assay.
This engineered strain is constitutive for cytoplasmic β-
galactosidase, lacks lac permease, and expresses a
plasmid-encoded periplasmic β-lactamase. Two chromo-
genic reporter molecules were used to monitor permea-
bilization of OM and IM in a single assay, nitrocefin and
Bacteria were grown in tryptic soy broth (TSB) from a
single colony, overnight, at 37 °C. After three washings in
phosphate buffer (pH 7.4, 10 mM phosphate buffer and
0.1 M NaCl), it was diluted to 106colony-forming units
(CFU)/mL in incubation buffer (phosphate buffer, pH 7.4,
containing 300 μg/mL of TSB) and added to all wells in a
non-culture-treated polystyrene microplate, together with
increasing concentrations of polymer or melittin (0–
100 μg/mL), in duplicate. The solution of polymer was
also made in phosphate buffer. Each well also contained
30 μM nitrocefin in phosphate buffer or 2.5 mM ONPG in
phosphate buffer. Absorbance was followed simulta-
neously at 486 nm and at 420 nm for 60 min, taking
readings every 2 min at 37 °C, in a SpectraMax Pro
microplate reader equipped with SoftMax Plus software
(Molecular Devices, Sunnyvale, CA), with shaking.
β-Galactosidase kinetic reaction with ONPG
β-Galactosidase from E. coli, 750 U/mg enzyme, was
added to phosphate buffer, pH 7.4, to make a 1.3-mg/mL
solution. One unit of enzyme is 1 μmol ONPG hydro-
lyzed/min at 25 °C. Enzymatic activity was tested by
absorption at 420 nm in a cuvette containing 1 mL of total
volume with phosphate buffer, 2.5 mM ONPG, 100 μg/
mL of polymer, 1 mM mercaptoethanol, and 1 mM MgCl2.
Reaction was started with the addition of 5 μL of β-
galactosidase solution at 25 °C and followed for 5 min at
420 nm. A control run replacing polymer with buffer was
also done. Readings were performed in a Cary-50 Bio UV-
visible spectrophotometer (Varian Instruments, Walnut
β-Lactamase kinetic reaction with nitrocefin
Penicillinase type IV from E. cloacae, 0.4 U/mg enzyme,
was added to phosphate buffer, pH 7.4, to make a 1.6-mg/
mL solution. One unit of enzyme is 1 μmol substrate
hydrolyzed/min at 25 °C. Enzymatic activity was tested
by absorption at 486 nm in a cuvette containing 1 mL of
total volume with phosphate buffer, 28 μM nitrocefin, and
100 μg/mL of polymer. Reaction was started with the
addition of 20 μL of penicillinase solution at 25 °C and
followed for 5 min at 486 nm. A control run replacing
polymer with buffer was also done. Readings were
performed in a Cary-50 Bio UV-visible spectrophotometer
IM permeability to ONPG in wild-type E. coli
The organism was grown overnight at 37 °C in TSB to
probe permeability to ONPG in wild-type E. coli K12
(JP109). Two hundred microliters was added to the wells
Dual Mechanism of Bacterial Lethality
of a polystyrene microplate, together with 10 μL of a
30 mM solution of ONPG (1.5 mM) in 10 mM phosphate
buffer as well as progressive concentrations of either
melittin or polymer solution in phosphate buffer, each in
duplicate, after adjusting the culture to an OD at 420 nm of
1.0 in TSB. The plate was read at 420 nm in a SpectraMax
microplate reader with SoftMax Pro software (Molecular
Devices) at 37 °C with shaking, for 45 min, taking readings
every 2 min.
A Live/Dead BacLight Bacteria Viability Kit from
Molecular Probes, using SYTO 9 and propidium iodide
stains, was used to assess membrane integrity, following a
protocol provided by the manufacturer. Briefly, bacteria
were grown in TSB to stationary phase, washed with
140 mM NaCl three times, and then diluted either in
140 mM NaCl or in 70% isopropanol (dead bacteria) to an
OD at 670 nm of 0.03 for E. coli K12 (108bacteria/mL) or
that of 0.15 for S. aureus ATCC 29213 (107bacteria/mL). A
standard curve was set up with a total volume of 1 mL,
containing mixtures of bacteria in either NaCl or
isopropanol together with the dye mixture. Fluorescence
emission was read in an SLM Aminco Bowman Series II
spectrofluorimeter at 490–700 nm, exciting at 470 nm,
following a 15-min incubation at room temperature in the
dark. Another set of cuvettes containing a total volume of
1 mL of bacteria washed in 140 mM NaCl and increasing
concentrations of polymer and the dye mixture were
treated the same way as the standard curve.
A broth microdilution assay in Mueller–Hinton Broth
was used, according to the procedures outlined by the
National Committee for Clinical Laboratory Standards
[Methods for Dilution Antimicrobial Susceptibility Tests for
Bacteria that Grow Aerobically. Approved Standard M7-A6
(2003), National Committee for Clinical Laboratory
Standards: Wayne, PA]. Briefly, 200 μL of solution
containing 106CFU/mL of bacteria was placed in each
well of noncoated polystyrene 96-well plates. Polymer
was dissolved in 0.01% acetic acid containing 0.2%
bovine serum albumin, and serial dilutions were made
and kept in silanized glass containers until applied to
the microplate. After incubation for 18 h at 37 °C with
gentle shaking, its absorbance was read in a Spectra-
Max microplate reader with SoftMax Pro software
(Molecular Devices) at 620 nm. MIC was taken as the
lowest concentration of drug at which there was 100%
reduction of growth. After absorbance readings, con-
centrations above the MIC were plated in agar to
determine the MBC following another 18 h of incuba-
tion at 37 °C.
This work was supported by the Canadian
Institutes of Health Research (Grant MOP-86608),
by a Collaborative Research in Chemistry grant
from the US National Science Foundation (CHE-
0404704), and by the University of Wisconsin–
Madison Nanoscale Science and Engineering Center
1. Zasloff, M. (2002). Antimicrobial peptides of multi-
cellular organisms. Nature, 415, 389–395.
2. Marr, A. K., Gooderham, W. J. & Hancock, R. E. (2006).
Antibacterial peptides for therapeutic use: obstacles
and realistic outlook. Curr. Opin. Pharmacol. 6,
3. Durr, U. H., Sudheendra, U. S. & Ramamoorthy, A.
(2006). LL-37, the only human member of the
cathelicidin family of antimicrobial peptides. Biochim.
Biophys. Acta, 1758, 1408–1425.
4. Dhople, V., Krukemeyer, A. & Ramamoorthy, A.
(2006). The human beta-defensin-3, an antibacterial
peptide with multiple biological functions. Biochim.
Biophys. Acta, 1758, 1499–1512.
5. Tossi, A., Sandri, L. & Giangaspero, A. (2000).
Amphipathic, alpha-helical antimicrobial peptides.
Biopolymers, 55, 4–30.
6. Goodman, C. M., Choi, S., Shandler, S. & DeGrado,
W. F. (2007). Foldamers as versatile frameworks for
the design and evolution of function. Nat. Chem. Biol.
7. Hancock, R. E. W. & Sahl, H. G. (2006). Antimicrobial
and host-defense peptides as new anti-infective
therapeutic strategies. Nat. Biotechnol. 24, 1551–1557.
8. Ikeda, T., Tazuke, S. & Suzuki, Y. (1984). Biologically
active polycations: 4. Synthesis and antimicrobial
activity of poly(trialkylvinylbenzylammonium chlor-
ide)s. Makromol. Chem.–Macromol. Chem. Phys. 185,
9. Kawabata, N. & Nishiguchi, M. (1988). Antibacterial
activity of soluble pyridinium-type polymers. Appl.
Environ. Microbiol. 54, 2532–2535.
10. Li, G., Shen, J. & Zhu, Y. (1998). Study of pyridinium-
type functional polymers: II. Antibacterial activity of
soluble pyridinium-type polymers. J. Appl. Polym. Sci.
11. Chen, C. Z., Beck-Tan, N. C., Dhurjati, P., Van
Dyk, T. K., LaRossa, R. A. & Cooper, S. L.
(2000). Quaternary ammonium functionalized poly
(propylene imine) dendrimers as effective antimi-
cules, 1, 473–480.
12. Tiller, J. C., Liao, C. J., Lewis, K. & Klibanov, A. M.
(2001). Designing surfaces that kill bacteria on contact.
Proc. Natl. Acad. Sci. USA, 98, 5981–5985.
13. Gelman, M. A., Weisblum, B., Lynn, D. M. &
Gellman, S. H. (2004). Biocidal activity of polystyr-
enes that are cationic by virtue of protonation. Org.
Lett. 6, 557–560.
14. Ilker, M. F., Nusslein, K., Tew, G. N. & Coughlin, E. B.
(2004). Tuning the hemolytic and antibacterial
activities of amphiphilic polynorbornene derivatives.
J. Am. Chem. Soc. 126, 15870–15875.
15. Kuroda, K. & DeGrado, W. F. (2005). Amphiphilic
polymethacrylate derivatives as antimicrobial agents.
J. Am. Chem. Soc. 127, 4128–4129.
16. Mowery, B. P., Lee, S. E., Kissounko, D. A., Epand,
R. F., Epand, R. M., Weisblum, B. et al. (2007). Mimicry
of antimicrobial host-defense peptides by random
copolymers. J. Am. Chem. Soc. 129, 15474–15476.
17. Epand, R. F., Savage, P. B. & Epand, R. M. (2007).
Bacterial lipid composition and the antimicrobial
efficacy of cationic steroid compounds (Ceragenins).
Biochim. Biophys. Acta, 1768, 2500–2509.
18. Ramamoorthy, A., Thennarasu, S., Lee, D. K., Tan, A.
& Maloy, L. (2006). Solid-state NMR investigation of
the membrane-disrupting mechanism of antimicrobial
Dual Mechanism of Bacterial Lethality
peptides MSI-78 and MSI-594 derived from magainin
2 and melittin. Biophys. J. 91, 206–216.
19. Thennarasu, S., Lee, D. K., Tan, A., Prasad, K. U. &
Ramamoorthy, A. (2005). Antimicrobial activity and
membrane selective interactions of a synthetic lipo-
peptide MSI-843. Biochim. Biophys. Acta, 1711, 49–58.
20. Epand, R. F., Schmitt, M. A., Gellman, S. H. & Epand,
R. M. (2006). Role of membrane lipids in the
mechanism of bacterial species selective toxicity by
two alpha/beta-antimicrobial peptides. Biochim. Bio-
phys. Acta, 1758, 1343–1350.
21. Laroche, G., Dufourc, E. J., Pezolet, M. & Dufourcq,
J. (1990). Coupled changes between lipid order and
polypeptide conformation at the membrane surface.
A deuterium-NMR and Raman study of polyly-
sine–phosphatidic acid systems. Biochemistry, 29,
22. Hays, L. M., Crowe, J. H., Wolkers, W. & Rudenko, S.
(2001). Factors affecting leakage of trapped solutes
from phospholipid vesicles during thermotropic
phase transitions. Cryobiology, 42, 88–102.
23. Shai, Y., Makovitzky, A. & Avrahami, D. (2006). Host
defense peptides and lipopeptides: modes of action
and potential candidates for the treatment of bacterial
24. Ferguson, A. D., Welte, W., Hofmann, E., Lindner, B.,
Holst, O., Coulton, J. W. & Diederichs, K. (2000). A
conserved structural motif for lipopolysaccharide
recognition by procaryotic and eucaryotic proteins.
Structure, 8, 585–592.
25. Fournier, B. & Philpott, D. J. (2005). Recognition of
Staphylococcus aureus by the innate immune system.
Clin. Microbiol. Rev. 18, 521–540.
26. Brandenburg, K., David, A., Howe, J., Koch, M. H. J.,
Andra, J. & Garidel, P. (2005). Temperature depen-
dence of the binding of endotoxins to the polycationic
peptides polymyxin B and its nonapeptide. Biophys. J.
27. Gutberlet, T., Frank, J., Bradaczek, H. & Fischer, W.
(1997). Effect of lipoteichoic acid on thermotropic
membrane properties. J. Bacteriol. 179, 2879–2883.
28. Ericksen, B., Wu, Z., Lu, W. & Lehrer, R. I. (2005).
Antibacterial activity and specificity of the six
human α-defensins. Antimicrob. Agents Chemother.
29. Lehrer, R. I., Barton, A. & Ganz, T. (1988). Con-
current assessment of inner and outer membrane
permeabilization and bacteriolysis in E. coli by
multiple-wavelength spectrophotometry. J. Immunol.
Methods, 108, 153–158.
30. Lehrer, R. I., Barton, A., Daher, K. A., Harwig, S. S.,
Ganz, T. & Selsted, M. E. (1989). Interaction of human
defensins with Escherichia coli. Mechanism of bacter-
icidal activity. J. Clin. Invest. 84, 553–561.
31. Buckner, F. S. & Wilson, A. J. (2005). Colorimetric
assay for screening compounds against Leishmania
amastigotes grown in macrophages. Am. J. Trop. Med.
Hyg. 72, 600–605.
32. Turner, J., Cho, Y., Dinh, N. N., Waring, A. J. & Lehrer,
R. I. (1998). Activities of LL-37, a cathelin-associated
antimicrobial peptide of human neutrophils. Antimi-
crob. Agents Chemother. 42, 2206–2214.
33. Stocks, S. M. (2004). Mechanism and use of the
commercially available viability stain, BacLight. Cyto-
metry A, 61, 189–195.
34. Andra, J., Monreal, D., de Tejada, G. M., Olak, C.,
Brezesinski, G., Gomez, S. S. et al. (2007). Rationale for
the design of shortened derivatives of the NK-lysin-
derived antimicrobial peptide NK-2 with improved
activity against Gram-negative pathogens. J. Biol.
Chem. 282, 14719–14728.
35. Allende, D., Simon, S. A. & McIntosh, T. J. (2005).
Melittin-induced bilayer leakage depends on lipid
material properties: evidence for toroidal pores. Bio-
phys. J. 88, 1828–1837.
36. Hancock, R. E. & Chapple, D. S. (1999). Peptide anti-
biotics. Antimicrob. Agents Chemother. 43, 1317–1323.
37. Basle, A. & Delcour, A. H. (2001). Effect of two
polyamine toxins on the bacterial porin OmpF.
Biochem. Biophys. Res. Commun. 285, 550–554.
38. Iyer, R., Wu, Z., Woster, P. M. & Delcour, A. H. (2000).
Molecular basis for the polyamine–OmpF porin
interactions: inhibitor and mutant studies. J. Mol.
Biol. 297, 933–945.
39. Nikaido, H. (1994). Porins and specific diffusion
channels in bacterial outer membranes. J. Biol. Chem.
40. Lemaitre, B. & Hoffmann, J. (2007). The host defense
of Drosophila melanogaster. Annu. Rev. Immunol. 25,
41. Ames, B. N. (1966). Assay of inorganic phosphate,
total phosphate and phosphatases. Methods Enzymol.
induced fusion and destabilization of liposomes.
Biochemistry, 24, 3099–3106.
Dual Mechanism of Bacterial Lethality