Cationic spacer arm design strategy for control of antimicrobial activity and conformation of amphiphilic methacrylate random copolymers

Article abstract of DOI:10.1021/bm300342u

Antimicrobial and hemolytic activities of amphiphilic random copolymers were modulated by the structure of the cationic side chain spacer arms, including 2-aminoethylene, 4-aminobutylene, and 6-aminohexylene groups. Cationic amphiphilic random copolymers with ethyl methacrylate (EMA) comonomer were prepared with a range of comonomer fractions, and the library of copolymers was screened for antimicrobial and hemolytic activities. Copolymers with 4-aminobutylene cationic side chains showed an order of magnitude enhancement in their antimicrobial activity relative to those with 2-aminoethylene spacer arms, without causing adverse hemolysis. When the spacer arms were further elongated to hexylene, the copolymers displayed potent antimicrobial and hemolytic activities. The 4-aminobutylene side chain appears to be the optimal spacer arm length for maximal antimicrobial potency and minimal hemolysis, when combined with hydrophobic ethylmethacrylate in a roughly 70/30 ratio. The copolymers displayed relatively rapid bactericidal kinetics and broad-spectrum activity against a panel of Gram-positive and Gram-negative bacteria. The effect of the spacer arms on the polymer conformation in the membrane-bound state was investigated by molecular dynamics simulations. The polymer backbones adopt an extended chain conformation, parallel to the membrane surface. A facially amphiphilic conformation at the membrane surface was observed, with the primary ammonium groups localized at the lipid phoshophate region and the nonpolar side chains of EMA comonomers buried in the hydrophobic membrane environment. This study demonstrates that the antimicrobial activity and molecular conformation of amphiphilic methacrylate random copolymers can be modulated by adjustment of cationic side chain spacer arms.

Full text of DOI:10.1021/bm300342u

Article
pubs.acs.org/Biomac
Cationic Spacer Arm Design Strategy for Control of Antimicrobial
Activity and Conformation of Amphiphilic Methacrylate Random
Copolymers
Edmund F. Palermo,† Satyavani Vemparala,*,§ and Kenichi Kuroda*,†,‡
‡
†Macromolecular Science and Engineering Center and Department of Biologic and Materials Sciences, School of Dentistry,
University of Michigan, Ann Arbor, Michigan 48109, United States
§The Institute of Mathematical Sciences, CIT Campus, Taramani, Chennai 600113, India
S
* Supporting Information
ABSTRACT: Antimicrobial and hemolytic activities of amphiphilic
random copolymers were modulated by the structure of the cationic
side chain spacer arms, including 2-aminoethylene, 4-aminobutylene, and
6-aminohexylene groups. Cationic amphiphilic random copolymers with
ethyl methacrylate (EMA) comonomer were prepared with a range of
comonomer fractions, and the library of copolymers was screened for
antimicrobial and hemolytic activities. Copolymers with 4-aminobutylene
cationic side chains showed an order of magnitude enhancement in their
antimicrobial activity relative to those with 2-aminoethylene spacer arms, without causing adverse hemolysis. When the spacer
arms were further elongated to hexylene, the copolymers displayed potent antimicrobial and hemolytic activities. The 4-
aminobutylene side chain appears to be the optimal spacer arm length for maximal antimicrobial potency and minimal hemolysis,
when combined with hydrophobic ethylmethacrylate in a roughly 70/30 ratio. The copolymers displayed relatively rapid
bactericidal kinetics and broad-spectrum activity against a panel of Gram-positive and Gram-negative bacteria. The effect of the
spacer arms on the polymer conformation in the membrane-bound state was investigated by molecular dynamics simulations.
The polymer backbones adopt an extended chain conformation, parallel to the membrane surface. A facially amphiphilic
conformation at the membrane surface was observed, with the primary ammonium groups localized at the lipid phoshophate
region and the nonpolar side chains of EMA comonomers buried in the hydrophobic membrane environment. This study
demonstrates that the antimicrobial activity and molecular conformation of amphiphilic methacrylate random copolymers can be
modulated by adjustment of cationic side chain spacer arms.
parameters such as charge and amphiphilicity, the membrane-
targeting antimicrobial function of AMPs has been successfully
emulated by synthetic peptides10 and peptoids11 that exert
antimicrobial activity with low toxicity to human cells. The
central goal of this biomimetic approach is to capture the
physiochemical features of the naturally occurring peptides
rather than reproducing the exact chemical structures. In that
light, Tew and DeGrado developed cationic, amphiphilic
arylamide oligomers, which are currently in clinical trials as
antibiotic drugs.12,13 Further mimicry of AMPs has encom-
passed amphiphilic homopolymers14 and random copoly-
mers15,16 that exert antimicrobial activity with minimal toxicity
after extensive optimization of their structural characteristics.
Membrane disruption is chiefly modulated by the molecular
weight (MW), cationic charge, and amphiphilicity of the
peptides.5,17,18 Cationic groups are initially attracted to the
anionic components of bacterial membranes.19 Membrane
binding induces secondary structure in which the cationic and
INTRODUCTION
■
With the rapid spread of antibiotic-resistant bacterial infections,
coupled to a decline in the availability of new antibiotic drugs,
there is now an urgent need for compounds that exert novel
mechanisms of antimicrobial action.1 In nature, the release of
host defense antimicrobial peptides (AMPs) by eukaryotes is a
key component of innate immunity to invasion by pathogenic
bacteria because AMPs are able to kill a broad spectrum of
bacteria, without harming host cells and with a lower likelihood
of inducing resistance.2 Hence, there is substantial interest in
utilizing them to combat infectious diseases, and many peptide
antibiotics are currently in clinical trials.3 The mechanisms of
AMP action are complex and diverse: the peptides have been
shown to permeabilize bacterial cell membranes, inhibit
intracellular targets to cause direct killing of the invading
pathogens, and modulate the host immune system.4−6 Recent
transcriptional analyses have revealed that certain AMPs induce
responses in the bacteria as a result of subinhibitory
challenges.7,8 Of these known effects, the membrane-disrupting
ability of the peptides has attracted much attention, particularly
in the field of synthetic polymer mimics.9 Because the
membrane disruption event is modulated by physiochemical
Received: March 2, 2012
Revised: April 3, 2012
Published: April 5, 2012
© 2012 American Chemical Society
1632
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
hydrophobic side chains of amino acid residues are segregated
to opposing domains, known as “facially amphiphilic”
structures.20 The facially amphiphilic structure of these peptides
is believed to be a key factor in their potent antimicrobial
activity as well as selective toxicity to bacterial cells over host
cells. As such, compounds that can adopt facially amphiphilic
conformation show substantial promise as peptide-mimics such
as β-peptides21,22 and peptoids.11 A wide variety of synthetic
amphiphilic random copolymers based on methacrylates,23
methacrylamides,24 norbornenes,14 and nylon-3 derivatives25
have been designed to mimic the physiochemical features and
functions of AMPs. These polymers have been shown to inhibit
bacterial growth with minimal harm to human red blood cells
(RBCs). These random copolymers are not designed to have
intrinsic facially amphiphilic secondary conformations. How-
ever, they may be capable of adopting irregular but facially
amphiphilic confirmations on the cell membrane given
sufficient backbone flexibility.16 Such a hypothesis has been
previously proposed by Gellman and coworkers regarding their
antimicrobial nylon-3 copolymers.25 We have demonstrated
that the ammonium groups of methacrylate copolymers are
likely to form a complex with phosphate lipid head groups by a
combination of electrostatic and hydrogen bonding effects,
which enhance the affinity of polymers for lipid membranes.26
Sum frequency generation (SFG) spectroscopy showed that the
hydrophobic side chains in the same copolymers interact with
the nonpolar lipid alkyl tails.27 This experimental evidence
suggests that the ammonium groups of the copolymers interact
with phosphate lipid head groups while the hydrophobic side
chains of comonomers are inserted into the membranes.
Accordingly, it is possible for the flexible polymethacrylae
chains to adopt facially amphiphilic conformation in the
membrane, despite their lack of defined monomer sequence
and intrinsic secondary structures.
One of the key parameters that has not been systematically
studied is the chemical identity and length of cationic spacer
arms in the side chains of copolymers, likely to affect the
amphiphilic properties of polymers. The variation of spacer
arms may also affect the conformation of polymers and
insertion into cell membranes: a related concept is the
snorkeling effect for peptides with long ammonium side-chain
spacer arms, in which the ammonium spacer arms reach to the
water−lipid interface, allowing nonpolar peptide helices to
localize in the hydrophobic domains of lipid bilayers.28,29 The
position and orientation of hydrophobic transmembrane helices
can be modulated by the snorkel effect. Insertion of
transmembrane helices, possessing mismatch between peptide
length and the thickness of hydrophobic core in a lipid bilayer,
was facilitated by electrostatic interaction between lipid
phosphate groups and lysine segment at either helix end.30
Similarly, snorkeling of an arginine residue at the helix end
reduces the energy for insertion of helices, allowing trans-
membrane configuration.31 The position of transmembrane
helices in lipid bilayers was also controlled by incorporation of
lysine or arginine in the middle of sequence, which snorkels
toward the lipid−water interface.32
cell uptake as the spacer lengths were increased.33 However,
this design concept has never been applied to antimicrobial
polymers to the best of our knowledge. The advantage of such a
design approach is that the spacer group parameter can be
tuned for controlling the antimicrobial activity without
changing the ratio of the number of cationic ammonium
groups to hydrophobic comonomers.
To that end, we demonstrate the role played by the spacer
groups in terms of antimicrobial and hemolytic activities. The
cationic spacer arms have various lengths and structures
(ethylene, butylene, hexylene, and cyclohexylene) to assess
their effects on the activities systematically. Furthermore, we
employed molecular dynamics simulations to investigate the
conformations of the copolymers when bound to anionic lipid
bilayers.
MATERIALS AND METHODS
■
Materials. 2,2′-Azobisisobutyronitrile (AIBN) was purchased from
Sigma-Aldrich, and ethyl methacrylate (EMA), methyl 3-mercapto-
propionate (MMP), ethanolamine, 4-butanolamine, 6-hexanolamine,
and di-tert-butyldicarbonate were purchased from Acros and used
without further purification. Methacryloyl chloride was purchased from
Acros and was freshly distilled prior to each use. Trifluoroacetic acid
and reagent grade solvents were purchased from Fisher, and the bee
venom toxin melittin (>85%) was purchased from Sigma. Gel
permeation chromatography (GPC) was performed using a Waters
1515 pump and RI detector. Gas chromatography was performed on a
Shimadzu GC-2010. Human RBCs (leukocytes reduced adenine saline
added) were obtained from the American Red Cross Blood Services
Southeastern Michigan Region and used prior to the out date
indicated on each unit. The bacteria were obtained from ATCC:
Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 25923),
Enterococcus faecalis (ATCC 29212), Bacillus subtilis (ATCC 6633),
Acinetobacter baumannii (ATCC 17978), Pseudomonas aeruginosa
(ATCC 27853), and Salmonella enterica (ATCC 14028). MRSA
LAC 1263 was provided by Dr. Boles at the University of Michigan.
Monomer Synthesis. To a solution of ethanolamine (42.7 mmol,
2.6 g) in a biphasic mixture of THF (30 mL) and NaOH(aq) (1 M, 50
mL) was added a solution of di-tert-butyldicarbonate (42.7 mmol, 9.3
g) in THF (20 mL) dropwise, and the mixture was stirred at room
temperature overnight. The resulting N-Boc-protected alcohol was
extracted in ethylacetate by washing with water, saturated NaCO3H-
(aq), and brine. The same procedure was followed in the case of 4-
butanolamine, 6-hexanolamine, and trans-4-aminocyclohexanol.
Freshly distilled methacryloyl chloride (23 mmol, 2.2 mL) was diluted
with dichloromethane (5 mL) and added dropwise to a solution of the
N-Boc-protected alcohol (23 mmol) and triethylamine (25 mmol, 3.5
mL) in dichloromethane (50 mL) at 0 °C ice bath, and the mixture
was allowed to stir overnight. The solution was then filtered, and the
filtrate was concentrated under reduced pressure, extracted in
ethylacetate by washing with water, saturated NaCO3H(aq), and
brine. Figure 1 shows the monomers 1−4 prepared in this work.
The monomer 2-(tert-butoxycarbonylamino)ethyl methacrylate, 1,
was recrystallized from hexanes at −20 °C to give a white solid in 72%
yield. 1H NMR (400 MHz, CDCl3): δ 6.09 (s, 1H), 5.55 (s, 1H), 4.73
(bs), 4.17 (t, 2H), 3.40 (q, 2H), 1.91 (s, 3H), 1.41 (s, 9H). 13C NMR
(100 MHz, CDCl3): δ 167.27, 156.04, 136.00, 125.89, 79.56, 63.94,
39.67, 28.32, 18.27. Mass spectrometry: ESI calcd for [C11H19NO4 +
Na]+, 252.1212; found, 252.1209.
The monomer 4-(tert-butoxycarbonylamino)butyl methacrylate, 2,
was purified by silica gel column chromatography (eluent: hexane/
ethyl acetate 3:1) to give a colorless, viscous oil in 81% yield. Thin
layer chromatography (eluent: hexane/ethyl acetate 3:1) Rf = 0.54, I2
In the case of heterogeneous synthetic polymers, longer alkyl
spacer arms may facilitate membrane insertion by analogy to
the snorkel effect. These putative effects of spacer arms on the
polymer properties and conformations in lipid bilayers could
furthermore modulate their antibacterial activity. Recently,
adjustment of spacer arms has also been utilized for tuning the
activity of cell-penetrating peptoids, which showed improved
1
stained. H NMR (300 MHz, CDCl3): δ 6.09 (s, 1H), 5.55 (s, 1H),
4.54 (bs), 4.15 (t, 2H), 3.15 (q, 2H), 1.93 (s, 3H), 1.8−1.5 (m, 4H),
1.43 (s, 9H). 13C NMR (100 MHz, CDCl3): δ 167.35, 155.92, 136.34,
125.73, 79.16, 64.25, 40.11, 28.26, 26.73, 25.97, 18.21. Mass
1633
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
The monomer 6-(tert-butoxycarbonylamino)hexyl methacrylate, 3,
was purified by silica gel column chromatography (eluent: hexane/
ethyl acetate 9:1) to give a colorless, viscous oil in 68% yield. Thin
layer chromatography (eluent: hexane/ethyl acetate 9:1) Rf = 0.90, I2
1
stained. H NMR (300 MHz, CDCl3): δ 6.09 (s, 1H), 5.54 (s, 1H),
4.52 (bs), 4.13 (t, 2H), 3.10 (q, 2H), 1.93 (s, 3H), 1.67 (m, 2H), 1.5−
1.3 (m, 6H), 1.43 (s, 9H). 13C NMR (100 MHz, CDCl3): δ 167.49,
155.95, 136.46, 125.19, 79.03, 64.58, 40.45, 29.96, 28.52, 28.39, 26.41,
26.67, 18.30. Mass spectrometry: ESI calcd for [C15H27NO4 + Na]+,
308.1838; found, 308.1833.
The monomer trans-4-(tert-butoxycarbonylamino)cyclohexyl meth-
acrylate, 4, was purified by recrystallization from ethyl acetate at −20
°C to give a white solid in 77% yield. Thin layer chromatography
(eluent: hexane/ethyl acetate 9:1) Rf = 0.41, I2 stained. 1H NMR (300
MHz, CDCl3): δ 6.08 (s, 1H), 5.54 (s, 1H), 4.74 (tt, 1H), 4.40 (bs),
3.47 (m, 1H), 2.02 (m, 4H), 1.92 (s, 3H), 1.6−1.4 (m, 2H), 1.44 (s,
9H), 1.35−1.15 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 166.88,
155.20, 136.63, 125.21, 79.26, 72.05, 48.46, 30.73, 29.87, 28.37, 18.27.
Mass spectrometry: ESI calcd for [C15H25NO4 + Na]+, 306.1681;
found, 306.1681.
Polymer Synthesis. Random copolymers were prepared by a
previously described technique23 with minor alterations. EMA or butyl
methacrylate (BMA) was dissolved with N-Boc-protected aminoalkyl
methacrylate 1, 2, 3, or 4 (1 mmol total monomers, various ratios) in
acetonitrile (0.5 mL). AIBN (0.01 mmol) and MMP (0.1 mmol) were
added from concentrated stock solutions. The mixtures were
deoxygenated by nitrogen flushing for 2 min each and then submerged
in a 70 °C oil bath overnight. After concentration under reduced
pressure, the Boc-protecting groups were cleaved using trifluoroacetic
acid and precipitated from methanol (0.5 mL) into diethyl ether (25
mL) twice. Obtained precipitates were lyophilized to afford the
random copolymers as fine white powders.
Figure 1. Conversion−time plots for the copolymerization of EMA
with Boc-protected aminoalkylmethacrylates (A) 1, (B) 2, and (C) 3.
The monomer ratio of EMA to Boc-protected aminoalkylmethacry-
lates is 1:1. Gel-permeation chromatographs for each of the obtained
copolymers, in THF relative to PMMA standards. Mn values are 4.9,
5.2, and 5.9 kDa, and PDI values are 1.52, 1.55, and 1.64 for the
copolymers containing EMA and 1, 2, and 3, respectively.
spectrometry: ESI calcd for [C13H23NO4 + Na]+, 280.1525; found,
280.1529.
a
Table 1. Characterization and Biological Activity Data for the Library of Copolymers and Peptides
b
MIC (μg/mL)
HC50/MIC
S. aureus
c
c
c
d
f HB
Homopolymers (f HB = 0)
HDP
Mn (kDa)
E. coli
S. aureus
HC50 (μg/mL)
E. coli
E20
E40
E60
Ec60
B20
0
0
0
0
0
12.3
10.6
11.0
10.1
15.3
3.1
3.0
3.4
3.1
3.9
2000
2000
10
100
250
7.8
>2000
>2000
2.6
>1.0
>1.0
0.26
>1.5
>2.0
>20
>8.0
0.33
>9.5
>32
1300
1000
210
63
>2000
>2000
Copolymers (f HB = 0.2 to 0.3)
E228
E429
E627
Ec628
B226
0.28
0.29
0.27
0.28
0.26
13.5
12.5
12.6
11.5
13.6
2.9
2.9
3.3
2.9
3.1
250
21
125
63
>2000
1300
2.8
>8.0
62
>16
21
7.8
21
7.8
31
0.36
41
0.36
28
860
16
16
34
2.1
2.1
Copolymers (f HB = 0.4 to 0.5)
E245
0.45
0.46
0.40
0.44
0.40
16.2
17.6
14.5
12.7
19.6
3.1
3.6
3.4
2.9
4.1
13
10
6.5
7.8
16
42
21
10
21
16
76
53
2.3
29
9.0
5.8
1.8
E446
5.3
2.5
E640
0.35
3.7
0.23
1.4
Ec644
B240
0.56
0.56
Peptides
magainin-2
melittin
2.5
2.8
125
12
>250
5.6
>250
1.7
>2
0.14
0.30
a
Comprehensive biological activity data for the entire library of copolymers in this work is given in the Supporting Information (Table S1).
b
Minimum inhibitory concentration (MIC), the concentration of polymer which completely inhibited growth of the bacteria after incubation in MH
c
broth for 18 h at 37 °C. The initial challenge was ∼5 × 105 cfu/mL. The data are rounded to two significant digits. Mole fraction of hydrophobic
repeat units EMA or BMA (f HB), the average degree of polymerization (DP), and the number-average molecular weight (Mn) were by end-group
d
analysis of the 1H NMR spectra and confirmed by MALDI-TOF-MS (Supporting Information). Polymer concentration that induced 50% leakage
of hemoglobin from human red blood cells (HC50) relative to the positive control surfactant Triton-X-100. The data are rounded to two significant
digits.
1634
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
Polymer Characterization. 1H NMR confirmed that the
copolymers were obtained and that there was no detectable amount
of unreacted monomer or chain transfer agent remaining in the
samples. Comparison of the integrated peak areas enabled calculation
of the number-average degrees of polymerization (DPs), based on end
group analysis, and the average mole fraction of hydrophobic repeating
units (f HB) was determined. MALDI-TOF-MS wasperformed with a
Micromass TofSpec-2E spectrometer in linear positive ion mode with
a 20 kV potential. The matrix was α-cyano-4-hydroxycinnamic acid in
ethanol/acetonitrile (10 mg/mL). Polymer solutions (1 mM, 1 uL)
and matrix solution (10 uL) were spotted onto the target plate. MW
distributions matched the NMR calculation (Supporting Information).
Antimicrobial Activity. The lowest polymer concentration
required to inhibit growth of bacteria completely, defined as the
MIC, was determined by a turbidity-based microdilution assay in
Muller Hinton (MH) broth according to the procedure approved by
The National Committee for Clinical Laboratory Standards (NCCLS)
with the modifications proposed by Weigand et al.34 An overnight
culture of bacterial strains was regrown to exponential phase (OD600 of
0.5 to 0.6) and diluted to OD600 = 0.001 in MH broth. This stock (90
μL) was mixed with a polymer solution (10 μL) in a 96-well
polypropylene microplate (Corning no. 3359) to give the final
concentration of bacteria on the microplate of approximately 5 × 105
CFU/mL. After incubating for 18 h at 37 °C, the OD600 in each well
was recorded using a microplate reader (Perkin-Elmer Lambda
Reader). The MIC was defined as the lowest polymer concentration
at which no turbidity increase was observed relative to the negative
control, MH broth. All experiments were performed three times in
triplicate, and the MIC values reported are the average of the three
trials. The MIC values were determined below the solubility limit of
the polymers in MH broth in every case.
subsequent runs. All simulations were performed by NAMD2.7.35 The
forcefield parameters for polymers were adopted from our previous
simulations,36 the latest forcefield of CHARMM3637 was used for
lipids, and water was modeled by TIP3P model.38
RESULTS AND DISCUSSION
■
Polymer Synthesis and Characterization. Random
copolymers of Boc-protected aminoalkylmethacrylates and
EMA were prepared by thermally initiated free radical
polymerization in the presence of the chain transfer agent
methyl mercaptopropionate to afford low-MW (<10 kDa)
copolymers. The procedure was modified from the previously
described protocol.23 As a preliminary experiment, the
copolymerization was monitored by drawing aliquots at regular
intervals and determining the conversion of each comonomer
by gas chromatography, relative to the internal standard
mesitylene. For each pair of comonomers (1:1 ratio), both
methacrylates were consumed at nearly identical rates,
suggesting the random incorporation of monomers to the
polymer chain regardless of the side-chain spacer length of
aminoalkylmethacrylates (Figure 1). GPC revealed that the
MWs (5 to 6 kDa) and polydispersity index (PDI) values (1.5
to 1.6) of the obtained copolymers with different spacer arms
are comparable. Hence, we synthesized an extensive library of
random copolymers containing various ratios of hydrophobic to
amino-functionalized methacrylate repeating units for bio-
activity assessment (Figure 2).
Bactericidal kinetics was determined by inoculation of the polymers
with bacteria in MH broth and colony counting on agar plates. An
overnight culture of E. coli ATCC 25922 was regrown to exponential
phase (OD600 of 0.5 to 0.6) and diluted to an OD600 of 0.001 in 8 mL
of MH broth, and polymer was added from stock solution to a final
concentration of two times the MIC value (Table 1). The tubes were
incubated at 37 °C with orbital shaking and 10 μL aliquots were drawn
every 15 or 30 min. After dilution by 103, 102, or 10-fold in buffer, the
cells were streaked onto agar plates and incubated overnight. The
colonies were then counted to determine the cfu/mL as a function of
time. Each data point was done in duplicate, and the averages are
reported plus or minus the range.
Hemolytic Activity. Toxicity to human RBCs was assessed by a
hemoglobin release assay. RBCs (1 mL) were diluted into HEPES
buffered saline (9 mL; HBS = 10 mM HEPES, 150 mM NaCl, pH 7)
and washed with HBS three times. The stock (10% v/v) was diluted
three-fold in HBS (3.3% v/v) and was then mixed (90 μL) with
polymer solution (10 μL) on a 96-well microplate (final 3% v/v, 108
RBC/mL). After 60 min of incubation at 37 °C and 100 rpm, the plate
was centrifuged at 1000 rpm for 10 min, and the supernatant (10 μL)
was diluted into PBS (90 μL). Absorbance at 405 nm was recorded
using a microplate reader (Perkin-Elmer Lambda Reader). The
fraction of hemolysis was defined as H = (A − A0)/(ATX − A0),
where A is the absorbance reading of the sample well, A0 is the
negative hemolysis control (buffer), and ATX is the positive hemolysis
control (Triton X-100). HC50 was defined as the polymer
concentration that causes 50% hemolysis by curve-fitting with the
empirical Hill equation, H = 1/((HC50/[P])n + 1), where [P] is the
total concentration of polymer and the fitting parameters were n and
HC50. All experiments were performed three times in triplicate.
Molecular Dynamics Simulations. Model copolymers with DP =
10 and f HB = 0.3 (Figure 7) were inserted in aqueous phase above pre-
equilibrated POPE-POPG (7:3) bilayers, and each system (three in
total) was simulated for 100 ns. Counterions were added to each
system for overall charge neutrality. All three systems were simulated
under isothermal−isobaric ensemble conditions (305 K temperature
and 1 atm pressure). The long-range Coulomb interactions were dealt
with particle mesh Ewald method. An integration time step of 1.0 fs
was used for the first 10 ns, and a time step of 2 fs was used for all
Figure 2. Synthesis of cationic, amphiphilic methacrylate random
copolymers with various cationic side chain spacer arms. AIBN: 2,2′-
Azobis(2-methylpropionitrile). MMP: Methyl mercaptopropipnate.
MeCN: Acetonitrile.
Deprotection of the copolymers was achieved using TFA
(Figure 2), and the obtained cationic, amphiphilic copolymers
were characterized by 1H NMR. The average DPs were
calculated based on comparison of the integrated peak areas
arising from the methyl ester end groups and the side chains
(Supporting Information Figure S1). Although GPC could not
be performed on the cationic copolymers because of their
insolubility in THF, the MALDI-TOF-MS of representative
deprotected copolymers corroborates the NMR end-group
analysis (Supporting Information Figure S2). We intentionally
tailored the average DPs to the range of 12−15 (Table 1)
because such low-MW polymers in this range have been
previously shown to give potent antimicrobial activity with
1635
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
minimal hemolysis, whereas increasing MW generally increases
hemolysis.15
evidenced by more closely matched forward and back-titration
data curves. This may be due to the fact that longer linkages
separate the amine groups from the esters in the backbone,
reducing the likelihood of intramolecular nucleophilic addition.
Antimicrobial Activity. The antimicrobial activities of the
polymers were quantified in terms of MIC, the lowest polymer
concentration that completely inhibits bacterial growth. The
MIC values were determined in a turbidity-based microdilution
assay. See the Supporting Information (Table S1) for MIC
values of the entire library of copolymers in this work. In
general, the copolymers in this study exhibited inhibitory effects
against Gram negative E. coli and Gram positive S. aureus to
varying extents depending on their cationic side-chain spacer
arm groups and their fHB values (Table 1 and Figure 4).
The f HB values were obtained by comparison of signals
1
arising from the two different side chains in the H NMR
spectra. For each copolymer series, the fraction of hydrophobic
repeat units (EMA or BMA) in the copolymer, f HB, was varied
in the range of 0.0 to 0.6. Copolymers with higher f HB values
were not studied because of insolubility in the aqueous assay
media. The ammonium side-chain spacer arms were varied
from ethylene (E2), butylene (E4), hexylene (E6), to
cyclohexylene (Ec6). The fHB values of each of the polymers
are indicated by the subscript for each sample (Table 1). For
example, the copolymer in the E4 series containing 29 mol %
EMA units is referred to as E429.
Potentiometric Titration. The ionization behavior of the
polymers in this study was assessed by potentiometric titration
(Figure 3).39−41 The adjustment of the spacer groups
Figure 3. Ionization behavior of the cationic homopolymers
containing various “spacer arms” by potentiometric titration. Forward
titration with NaOH (filled symbols) and back-titration with HCl
(open symbols).
significantly impacted the titration curves: about 30% of the
primary amine groups in the homopolymer E20 were
deprotonated at pH 8, whereas <5% of the amine groups in
E40 and E60 were deprotonated at pH 9. The effect of
increasing apparent pKa (pH at α = 0.5) with elongated spacer
length is in agreement with a previous report on related amino-
functionalized methacrylates.42 Columbic repulsion of neigh-
boring charges localized in the polymer chains increases the
repulsive energy between side chains, reducing the degree of
protonation compared with the monomers in solution.43
Elongation of the spacer arms allows for greater spatial
separation of the ammonium groups on neighboring monomer
units, relieving the Columbic repulsion and thereby increasing
the apparent pKa values. The elongation of cationic side chain
groups has also been utilized to control side-chain interactions,
resulting in enhancement of the formation of secondary
structures of synthetic peptides.44 The biological activity assays
are performed at pH 7, in which >93% of the amino groups in
all of the homopolymers are protonated (cationic).
The forward- and back-titration data deviate as α is increased
beyond about 0.5. On the basis of our previous work39 and
others,43 this deviation is likely a result of isomerization of the
aminoalkylmethacrylate units to hydroxyalkylmethacrylamide
units or hydrolysis of the side-chain ester groups, which reduces
the number of amine groups in the polymer and curtails
activity. Despite the instability of the amino-methacrylates at
pH >9, the polymers are chemically stable up to pH 8 at 37 °C
overnight. Moreover, the elongation of spacer groups appears
to suppress the isomerization reactions to some extent, as
Figure 4. Antimicrobial activity of the copolymers with various spacer
arms and comonomer compositions, against E. coli and S. aureus.
Minimum inhibitory concentration (MIC) is defined the lowest
polymer concentration, which completely inhibits the growth of
bacteria based on turbidity. The initial challenge was ∼5 × 105 cfu/
mL. Error bars represent the standard deviation from three
independent trials, each done in triplicate.
The spacer groups strongly impacted the antimicrobial
activity of the polymers against E. coli (Table 1 and Figure 4).
The cationic homopolymers E20 and E40 showed only weak
antimicrobial activity against E. coli (MIC = 2000 μg/mL),
whereas E60 exhibited an MIC value of 10 μg/mL. This
suggests that elongation of spacer length enhances the activity
of the cationic homopolymers without incorporation of
hydrophobic comonomers. For comparison, the bee venom
toxin peptide melittin and natural AMP magainin-2 displayed
an MIC value of 12 and 125 μg/mL against E. coli, respectively.
It should be noted that homopolymers E20 and B20 have the
same chemical structures, but the B20 displayed lower MIC
values. This is likely to reflect the higher MW of B20.39
The weak activity of E2 and E4 homopolymers may suggest
that the hydrophobic properties of polymers are not sufficient
1636
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
to exert their antibacterial effect.23 Hence, we further examined
the effect of increasing the fraction of hydrophobic comonomer
(f HB) content on the activity. The MIC values generally
decreased with increasing f HB, except for the E6 polymers,
which displayed no dependence of MIC on f HB. When the
hydrophobic content is increased beyond f HB = 0.4, the MIC
values reached a plateau of ∼10 μg/mL for all polymers. This
indicates that the hydrophobic properties of the comonomer
EMA dominate the activity over the effect of spacer arm length
when the fraction of hydrophobic comonomer exceeds ∼0.4.
The effect of spacer arm length is quite evident when 0.1 < f HB
< 0.4, where the MIC values of copolymers decreased by orders
of magnitude as the spacer arms were elongated from two (E2)
to six (E6) carbon linkages. For example, elongating the spacer
arms from two (E228) to four (E429) carbon atoms led to a
decrease in MIC from 250 to 20 μg/mL, an activity
enhancement of about 12-fold (Table 1). Further extension
of the spacer arms to six carbons (E627) yielded an MIC value
of ∼8 μg/mL. We speculate that the hydrophobic property of
the 6-aminohexyl side chains dominates the activity, whereas
the comomoner EMA has no impact on the activity in this
particular case. In the moderate range of hydrophobic
comonomer EMA content (f HB = 0.27 to 0.29) the relative
activity ranking is E627 > E429 > E228, supporting the notion
that the antibacterial activity can be controlled by the
combination of spacer arm elongation and the hydrophobic
comomomers.
Figure 5. Hemolytic activity of the copolymers with various spacer
arms and comonomer compositions against human red blood cells.
Error bars represent the standard deviation from three independent
trials, each done in triplicate.
groups. It appears that the hemolytic activity of copolymers
may be approaching a plateau region at high hydrophobicity,
although it was not possible to confirm this with polymers of
higher f HB because of their insolubility in the assay media.
The copolymer E429 showed potent antimicrobial activity
(MIC = 21 μg/mL for E. coli) but was relatively nonhemolytic
(HC50 = 1300 μg/mL), giving a selectivity index (HC50/MIC)
of 63 (Table 1 and Figure 6). In the E2 polymer series, the best
Interestingly, the MICs of Ec6, which possesses constrained
cyclohexyl spacer arms, are orders of magnitudes weaker than
those of E6, whereas these polymers contain the same number
of carbon atoms (C6) in their spacer groups, This indicates that
the antibacterial mechanism is sensitive to structural difference
and properties of spacer arms (cyclic vs linear alkyl).
Against S. aureus, similar to the case of E. coli, the MIC values
decreased with increasing f HB and reached a plateau of ∼10−20
μg/mL for all the copolymers (Figure 4B). The MIC values of
hompolymers and copolymers with low fHB for S. aureus are
lower than E. coli, indicating that S. aureus is more susceptible
to the cationic polymers. For the copolymers with hydrophobic
content in the range of fHB = 0.27 to 0.29, the antimicrobial
activity increased with increasing spacer group length, giving
the relative activity ranking: E627 > E429 > E228. These results
also demonstrate that the activity against Gram positive E. coli
as well as Gram negative S. aureus can be controlled by tuning
the spacer arm groups in the cationic side chains.
Hemolytic Activity. Damage to human RBCs was
quantified as the polymer concentration that induces 50%
hemoglobin release (HC50), which was determined from dose−
response curves, that is, the percentage of hemolysis measured
as a function of polymer concentration. (See Figure S4 in the
Supporting Information for hemolysis dose−response curves of
all of the copolymers in this study.)
The homopolymers in all but the E6 series were relatively
nonhemolytic (HC50 > 2000 ug/mL) (Figure 5). For
comparison, the bee venom toxin peptide melittin showed an
HC50 value of 1.7 μg/mL under this assay condition. The
hemolysis increased by orders of magnitude with increasing f HB,
from 0 to ∼0.5, for all copolymers except E6 polymers, which
showed constant HC50 value of ∼5 μg/mL regardless of the f HB
values. When the hydrophobic comonomer EMA composition
was increased excessively (f HB > 0.5), the hemolytic activity of
all copolymers was pronounced, with HC50 < 20 μg/mL
without a large dependence on the identity of the spacer arm
Figure 6. Selectivity of the copolymers containing various cationic side
chain spacer arms, for E. coli and S. aureus relative to human red blood
cells.
example was E237, which showed potent antimicrobial activity
(MIC = 21 μg/mL) and relatively weak hemolytic activity
(HC50 = 1100 μg/mL), giving a selectivity index of 52. The E6
copolymers showed high hemolytic activity (HC50 < 5 μg/mL)
and very low selectivity (HC50/MIC < 1) regardless of the f HB
values. The B2 polymers, which contain BMA comonomers as
the hydrophobic groups, have been previously reported23 to
show the potent activity with relatively low selectivity toward
bacteria over human RBCs. The MIC values of B226 are similar
to those of E429 (MIC = 15−20 μg/mL), but B226 is highly
hemolytic (HC50 = 34 μg/mL). These results demonstrate that
the copolymers can be tuned by selecting ethylene or butylene
spacer arms and relatively low fHB to inhibit bacterial growth
selectively without causing adverse toxicity to human RBCs.
We speculate that the spacer arms would enable fine-tuning
of the hydrophobic property of polymers, which provides an
additional design strategy to obtain balanced amphiphilic
properties such that these polymers exert selective activity
against bacteria over RBCs. We also speculate that the spacer
arms play an important role in the binding of polymers to cell
1637
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
membranes and polymer conformations in membranes, which
will be discussed in the following section.
Table 2. Spectrum of Antimicrobial Activity by
Representative Copolymers
Bactericidal Kinetics. We further examined the bactericidal
kinetic exerted by representative copolymers, which displayed
potent antibacterial activity but varying degrees of selectivity to
bacteria over RBCs (Figure 6) to assess the potential of thee
copolymers as antibacterial agents. Accordingly, we monitored
the number of viable E. coli cells as a function of exposure time
to the copolymers at twice their respective MIC concentrations.
E429 and E627 caused three-log reductions in the viable cell
population (99.9% killing) within 60 and 15 min, respectively,
whereas E228 required 120 min for the same level of reduction
(Figure 7). This demonstrates that the rate of killing was
MIC or HC50 (μg/mL)
magainin-
b
b
bacterium
gram
(+)
E228
125
E429
63
E627 melittin
2
Staphylococcus
aureus
7.8
5.6
>250
Staphylococcus
(+)
(+)
125
250
31
31
7.8
n.d.
n.d.
aureus (CA-
a
MRSA)
Enterococcus
faecalis
7.8
16
>500
Bacillus subtilis
Escherichia coli
(+)
(−)
(−)
31
7.8
21
10
3.9
7.8
3.9
5.6
12
63
250
63
125
n.d.
Acinetobacter
baumannii
n.d.
Pseudomonas
aeruginosa
(−)
(−)
210
16
16
280
140
1.7
500
Salmonella
enterica
125
16
7.8
2.8
500
human RBC
(HC50)
>2000
1300
>250
a
Community acquired methicillin-resistant S. aureus (MRSA) strain
LAC 1263. MIC values were previously reported.47
b
we modeled the interaction of three representative copolymers
(E2, E4, and E6) with lipid membranes by molecular dynamics
simulations. The structures used for these simulations are
copolymers with DP = 10 and f HB = 0.3 (Figure 8). The mole
Figure 7. Bactericidal kinetics by copolymers. The concentration of
viable E. coli cells given in colony-forming units per milliliter (CFU/
mL) (left y-axis) or the percentages of killed E. coli (right y-axis) after
incubation with the copolymers was plotted as a function of time.
Three representative copolymers (E228, E429, and E627) containing
different spacer arms with f HB in the range of 0.27 to 0.29 were used in
this assay. The initial E. coli concentration was ∼4.5 × 105 CFU/mL,
and the polymer concentrations were twice their respective MIC
values (MIC = 250, 21, and 7.8 μg/mL for E228, E429, and E627,
respectively). Error bars represent the range from duplicate measure-
ments.
Figure 8. Monomer sequence and structures of model copolymers.
The copolymers have different length of cationic spacer arms: ethylene
(m = 2) for E2model, butylene (m = 4) for E4model, and hexane (m = 6)
for E6model
.
accelerated by the elongation of the spacer groups, in
corroboration with the observed decrease in MIC values. By
comparison, Pexiganin, a potent synthetic derivative of
magainin, was reported to cause a 3-log reduction in the viable
E. coli cell population within 1 h at a concentration of twice the
MIC.10 The copolymers in this study appear to exert similar
bactericidal effect.
Activity Spectrum. To assess the spectrum of activity, we
also tested the representative copolymers against a panel of
Gram positive and Gram negative bacteria (Table 2). The MIC
values of E429 are equal to or less than 100 μg/mL for all of the
tested strains, which is substantially lower than the HC50 value
(1300 μg/mL). These results suggest that the copolymer has
the broad spectrum activity with selective activity to bacteria
over human RBCs. For all of the strains tested, elongation of
the spacer arm side-chain length caused an improvement in the
antimicrobial activity. Importantly, the E4 copolymers showed
inhibitory effects against A. baumannii, which is inherently
resistant to many antibiotic drugs,45 and a community-acquired
strain of methicillin-resistant S. aureus (MRSA)46 at low
microgram per milliliter concentrations.
fraction f HB = 0.3 was selected because the polymers that
showed antimicrobial and hemolytic activities in this range were
sensitive to the spacer arms in the side chains. We modeled the
copolymer structures with an isotactic sequence in which
cationic monomers and EMA comonomers were unevenly
distributed, or scrambled, in a ratio of 7:3.
The model membranes selected to understand the
interactions of the three model polymers with lipid bilayers
are composed of POPE and POPG lipids in a 7:3 ratio, which is
a common formulation used to mimic bacterial cell
membranes.48 In all three simulations, the polymer backbones
are seen to be parallel to the membrane surface, with the
hydrophobic groups projected into the membrane core and the
ammonium groups projected to the water−membrane interface
(Figure 9). It appears that the depth of insertion depends on
the spacer arm length. The copolymer E2model (the model
polymer with m = 2 in Figure 8) was found to be largely near
the lipid membrane surface, and most of the ammonium groups
(colored orange) appear to remain in water phase even at the
end of 100 ns of simulation. E2model also adopted a stable
compact conformation such that the polymer chain is folded in
the XY plane (perpendicular to membrane normal) (Figure
9B). The cationic ammonium groups in the polymer side chains
(colored red) of E4model and E6model are localized at the lipid−
Molecular Dynamics Simulation. Although molecular-
level details of the bactericidal action exerted by polymers and
peptides have been elusive experimentally, simulations have
provided a wealth of information on this fine scale. Accordingly,
1638
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
Figure 9. (A) Snapshots of the three systems at the end of 100 ns simulations. Water is not shown for clarity. The ammonium spacer arm groups and
hydrophobic ethyl groups of EMA comonomers are colored orange and green, respectively. (B) Average conformation of the copolymers in the XY
plane (perpendicular to membrane normal) at the end of 100 ns of MD simulations. The cationic and EMA ethyl side chains are colored red and
blue, respectively. The polymer backbones are colored green. (C) Average conformation of the copolymers oriented parallel to the membrane
normal.
water interface, and the nonpolar ethyl side chains of
comonomer EMA (colored green) are inserted into the
hydrophobic membrane core (Figure 8). The end-to-end
distances of the polymers in the last 20 ns of simulation are
∼20 Å for E4model and E6model (Figure S5 of the Supporting
Information), indicating that E4model and E6model are in highly
stretched conformations, considering that the theoretical length
for a fully extended conformation is 23 to 24 Å. The cationic
spacer arms and hydrophobic ethyl groups of EMA comonomer
in the polymer side chains are segregated to opposite faces of
the polymer backbone (Figure 9C). The average distances
between the center of mass of the ethyl and amine groups in
the polymers are ∼7 and ∼9 Å for E4model and E6model (Figure
S6 of the Supporting Information), which are substantial given
that the theoretical maximum distances for full extension are
∼16 and ∼18 Å, respectively.
The Z-density profiles of the copolymers also illustrate
amphiphilic conformations in the lipid bilayer upon polymer
insertion. The ammonium groups of the polymer side chains
(Figure 10, black line) are well-aligned with the phosphate
heads groups of both lipids (blue line) and the hydroxyl groups
of POPG glycerol (purple line), suggesting that the ammonium
groups are localized near the water−membrane interface. The
peaks of the ethyl side chains (green line) and the lipid acyl
chains (orange) also overlap, indicating that the hydrophobic
side chains of the polymer are inserted into the hydrophobic
membrane core. The ammonium groups in the side chains and
the ethyl side chains are spatially segregated with respect to the
Figure 10. Z-density profiles of various components of lipid−polymer
systems averaged over last 20 ns of MD simulations. The green lines
represent the polymer hydrophobic side chains, whereas the black lines
denote the polymer amino groups. Clearly, the segregation into facially
amphiphilic conformations is evident in the case of the E2 and E4
model copolymers.
as α-helix and β-sheet found in naturally occurring AMPs.
Mondal et al. previously investigated the interaction of helical
β-peptides and random copolymer models of β-peptides with a
model membrane49 and suggested that these peptides bind to
the membrane by forming facially amphiphilic structures. The
analysis of peptide sequence and conformation in that study
implied that well-defined secondary structure is not necessary
for peptide binding, and the ability of peptide sequence to
segregate the cationic and hydrophobic groups is rather
ester groups of side chains (red line) for E4model and E6model
,
indicating that the cationic arms and ethyl side chains are
presented on opposite faces of the polymer backbone.
These results suggest that these polymers adopt facially
amphiphilic conformations upon binding to the membranes
despite their lack of defined intrinsic secondary structures such
1639
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
important for efficient binding. Despite the distinctive differ-
ences in the chemical structures and conformational restriction
(backbone flexibility) between the peptides and the meth-
acrylate copolymers studied here the same principle appears to
govern the polymer-membrane binding.
cell surface to disintegrate the membrane integrity.4 Interest-
ingly, a computational study on the action of magainin
derivatives in lipid bilayer showed that they were able to
form distorted toroidal pores even when the peptides were not
in a stable α-helix form.50 This indicates that an α-helical
conformation is not a prerequisite for pore formation. Further
investigation on the polymer binding and cooperativity
between multiple synthetic polymer chains would therefore
enhance our understanding of the membrane disruption
mechanism and biological relevance. Elucidating the relation-
ship between chemical structures of the copolymers and their
interactions with membranes is requisite for the rational design
of optimal copolymer structures for potent activity.
The copolymer chains are positioned more toward the
hydrophobic membrane core as the spacer arms are elongated
(Figure 9). By analogy to the snorkel effect observed with
transmembrane peptide helices,30−32 we expected that
increased spatial separation between the cationic ammonium
and hydrophobic ester groups would facilitate deeper
membrane insertion. Indeed, the distribution of ester groups
(red line) also shifted toward the center of the lipid bilayer as
the spacer side chains were elongated (Figure 10). The longer
alkyl chain in the spacer arms could provide more hydro-
phobicity for membrane insertion as well as additional distance
between the cationic side chains and the hydrophobic polymer
backbone, by analogy to the snorkeling effect observed in
peptides.29 The snorkeling effect reduces the energetic barrier
to peptide insertion for transmembrane helices. Studies on such
peptides indicated that the butylene side-chain spacer arms of
lysine residues facilitate such insertion. In this study, variation
of spacer arm alkyl length enables control of polymer insertion
as well. It has been suggested that the smaller conformational
restriction on the cationic segments of β-peptides results in the
lower free energy for attachment to membranes.49 The longer
spacer arms of methacrylate copolymers may facilitate the
arrangement into facially amphiphilic coils due to an increase in
conformational freedom in the cationic side chains. This may in
turn enable the deeper insertion of polymer chains into the
membranes as well as lowering the free energy barrier to
membrane binding. However, the E4model and E6model
copolymers do not appear to penetrate entirely to the
membrane core. One possibility is that the hydrophobic ethyl
groups of comonomers are not inserted into the membrane to
significant extent because the ethyl ester groups of EMA are not
hydrophobic enough for deeper insertion. We also speculate
that the ester groups of methacrylate are not optimal for
membrane insertion because the polarity of ester groups costs
energy, not favorable to be buried into the hydrophobic
membrane core.
CONCLUSIONS
■
We developed a new strategy to modulate the antimicrobial and
hemolytic activities of amphiphilic random copolymers by
adjustment of the cationic side-chain spacer arms. In terms of
potent antimicrobial activity (MIC in low μg/mL) and minimal
hemolytic activity, the four carbon spacer arms in 4-amino-
butylmethacrylate afford the best profile in this study. This
polymer exerts relatively rapid bactericidal kinetics and shows
broad-spectrum activity against a panel of bacteria, including
MRSA and A. baumannii.
Preliminary studies by molecular dynamics simulations
showed that the copolymers adopt segregation of cationic
side chains and ethyl groups of comonomer EMA or facially
amphiphilic conformation when bound to a bacteria-type lipid
membrane. The formation of such amphiphilic polymer
conformations might be a key determinant for optimal
antimicrobial efficacy, although exact details of the antibacterial
mechanism of polymers remain unclear at this point. In the case
of designing AMPs, MD simulations have been called upon to
explain and predict biological activity profiles.49 In the future,
MD simulations may also aid in the understanding of the
biological activities observed in synthetic polymers as well as
rational design of membrane-active antimicrobial polymers.
This spacer arm approach will provide new molecular design
strategy to improve the antibacterial activity of amphiphilic
random copolymers as well as to shed light to the mechanism
of antibacterial action of polymers.
It is interesting to note that E2model adopted a folded
conformation in the XY plane, whereas E4model and E6model
showed more extended backbones (Figure 9B). This is likely
due to the polarity of the lipid−water interface, which
determines the polymer conformations. Because E2model is
located at the lipid−water interface, the polymer chain folds
into the compact conformation to minimize the exposure of
hydrophobic polymer backbone and side chains to the aqueous
phase, similar to micelle formation in aqueous solution. E4model
and E6model are inserted into the hydrophobic membrane core,
which in turn makes the polymer chains extend to maximize the
interaction between the polymer backbone and acyl chains of
lipids. The detailed analysis of polymer conformation and
membrane insertion will be discussed in a forthcoming report.
These results suggest that the spacer arms could modulate the
polymer binding to membranes and amphiphilic conformation.
The applicability of this polymer-membrane binding model to
the biological activity remains tentative at this time. Several
mechanisms have been proposed to explain the membrane-
disrupting action of AMPs. The proposed models suggest that
multiple α-helical peptides either insert into cell membranes to
form transmembrane pores cooperatively or accumulate on the
ASSOCIATED CONTENT
■
S
* Supporting Information
Polymer synthesis and characterization data and supplementary
biological activity data. This material is available free of charge
via the Internet at http://pubs.acs.org
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: vani@imsc.res.in (S.V.) and kkuroda@umich.edu
(K.K.).
Notes
The authors declare the following competing financial
interest(s): K. Kuroda is a coinventor on a patent application
filed by the University of Pennsylvania covering “Antimicrobial
Copolymers and Uses Thereof”. The patent application has
been licensed to PolyMedix, Inc. (Radnor, PA). PolyMedix did
not play a role in the design and conduct of this study, in the
collection, analysis, or interpretation of the data, or in the
preparation, review, or approval of the article.
1640
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641

Biomacromolecules
Article
(26) Palermo, E. F.; Lee, D.-K.; Ramamoorthy, A.; Kuroda, K. J. Phys.
Chem. B 2011, 115, 366−375.
ACKNOWLEDGMENTS
■
This research was supported by the Department of Biologic and
Materials Sciences, University of Michigan School of Dentistry
and NSF CAREER Award (DMR-0845592 to K.K.). For the
MD simulations, computing time on the “Annapurna” cluster at
IMSc is greatly appreciated. We thank Dr. Robertson
Davenport, University of Michigan Hospital, for providing the
RBCs, Prof. Eric Krukonis, University of Michigan School of
Dentistry for use of the microplate reader, Prof. Blaise Boles at
University of Michigan for providing MRSA LAC 1263, and
Prof. Anne J. McNeil, University of Michigan Chemistry
Department, for use of the GPC and GC.
(27) Avery, C. W.; Palermo, E. F.; McLaughin, A.; Kuroda, K.; Chen,
Z. Anal. Chem. 2011, 83 (4), 1342−1349.
(28) Monne, M.; Nilsson, I.; Johansson, M.; Elmhed, N.; von Heijne,
G. J. Mol. Biol. 1998, 284 (4), 1177−1183.
(29) Mishra, V. K.; Palgunachari, M. N.; Segrest, J. P.;
Anantharamaiah, G. M. J. Biol. Chem. 1994, 269 (10), 7185−7191.
(30) Jaud, S.; Fernandez-Vidal, M.; Nilsson, I.; Meindi-Beinker, N.
M.; Hubner, N. C.; Tobias, D. J.; von Heijne, G.; White, S. H. Proc.
Natl. Acad. Sci. U.S.A. 2009, 106 (28), 11588−11593.
(31) Schow, E. V.; Freites, J. A.; Cheng, P.; Bernsel, A.; von Heijne,
G.; White, S. H.; Tobias, D. J. J. Membr. Biol. 2011, 239 (1−2), 35−48.
(32) Krishnakumar, S. S.; London, E. J. Mol. Biol. 2007, 374 (5),
1251−1269.
(33) Wender, P. A.; Mitchell, D. J.; Pattabiraman, K.; Pelkey, E. T.;
Steinman, L.; Rothbard, J. B. Proc. Natl. Acad. Sci. U. S. A. 2000, 97
(24), 13003−13008.
REFERENCES
■
(1) Boucher, H. W.; Talbot, G. H.; Bradley, J. S.; Edwards, J. E.;
Gilbert, D.; Rice, L. B.; Scheld, M.; Spellberg, B.; Bartlett, J. Clin. Infect.
Dis. 2009, 48 (1), 1−12.
(34) Wiegand, I.; Hilpert, K.; Hancock, R. E. W. Nat. Protoc. 2008, 3
(2), 163−175.
(2) Zasloff, M. Nature 2002, 415 (6870), 389−395.
(3) Marr, A. K.; Gooderham, W. J.; Hancock, R. E. W. Curr. Opin.
Pharmacol. 2006, 6 (5), 468−472.
(35) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid,
E.; Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K. J. Comput.
Chem. 2005, 26 (16), 1781−1802.
(4) Brogden, K. A. Nat. Rev. Microbiol. 2005, 3 (3), 238−250.
(5) Marcos, J. F.; Gandia, M. Expert Opin. Drug Discovery 2009, 4 (6),
659−671.
(36) Ivanov, I.; Vemparala, S.; Pophristic, V.; Kuroda, K.; DeGrado,
W. F.; McCammon, J. A.; Klein, M. L. J. Am. Chem. Soc. 2006, 128 (6),
1778−1779.
(37) Klauda, J. B.; Venable, R. M.; Freites, J. A.; O’Connor, J. W.;
Tobias, D. J.; Mondragon-Ramirez, C.; Vorobyov, I.; MacKerell, A. D.;
Pastor, R. W. J. Phys. Chem. B 2010, 114 (23), 7830−7843.
(38) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R.
W.; Klein, M. L. J. Chem. Phys. 1983, 79 (2), 926−935.
(39) Palermo, E.; Kuroda, K. Biomacromolecules 2009, 10 (6), 1416−
1428.
(40) Thomas, J. L.; Chung, J.; Gross, D.; Tirrell, D. Abstr. Pap. Am.
Chem. Soc. 1995, 210, 56-POLY.
(41) Thomas, J. L.; You, H.; Tirrell, D. A. J. Am. Chem. Soc. 1995,
117 (10), 2949−2950.
(6) Yeaman, M. R.; Yount, N. Y. Pharmacol. Rev. 2003, 55 (1), 27−
55.
(7) Bader, M. W.; Navarre, W. W.; Shiau, W.; Nikaido, H.; Frye, J.
G.; McClelland, M.; Fang, F. C.; Miller, S. I. Mol. Microbiol. 2003, 50
(1), 219−230.
(8) Gooderham, W. J.; Bains, M.; McPhee, J. B.; Wiegand, I.;
Hancock, R. E. W. J. Bacteriol. 2008, 190 (16), 5624−5634.
(9) Matsuzaki, K. Biochim. Biophys. Acta, Biomembr. 2009, 1788 (8),
1687−1692.
(10) Ge, Y. G.; MacDonald, D. L.; Holroyd, K. J.; Thornsberry, C.;
Wexler, H.; Zasloff, M. Antimicrob. Agents Chemother. 1999, 43 (4),
782−788.
(42) Geurts, J. M.; Gottgens, C. M.; Van Graefschepe, M. A. I.;
Welland, R. W. A.; Van Es, J. J. G.; German, A. L. J. Appl. Polym. Sci.
2001, 80 (9), 1401−1415.
(11) Patch, J. A.; Barron, A. E. J. Am. Chem. Soc. 2003, 125 (40),
12092−12093.
(43) He, L. H.; Read, E. S.; Armes, S. P.; Adams, D. J. Macromolecules
2007, 40 (13), 4429−4438.
(12) Tew, G. N.; Scott, R. W.; Klein, M. L.; Degrado, W. F. Acc.
Chem. Res. 2010, 43 (1), 30−39.
(44) Zhang, Y. F.; Lu, H.; Lin, Y.; Cheng, J. J. Macromolecules 2011,
44 (17), 6641−6644.
(13) Liu, D. H.; Choi, S.; Chen, B.; Doerksen, R. J.; Clements, D. J.;
Winkler, J. D.; Klein, M. L.; DeGrado, W. F. Angew. Chem., Int. Ed.
2004, 43 (9), 1158−1162.
(45) Adams, M. D.; Chan, E. R.; Molyneaux, N. D.; Bonomo, R. A.
Antimicrob. Agents Chemother. 2010, 54 (9), 3569−3577.
(46) Boles, B. R.; Thoendel, M.; Roth, A. J.; Horswill, A. R. PloS One
2010, 5 (4), e10146.
(14) Ilker, M. F.; Nusslein, K.; Tew, G. N.; Coughlin, E. B. J. Am.
Chem. Soc. 2004, 126 (48), 15870−15875.
(15) Kuroda, K.; Caputo, G. A.; DeGrado, W. F. Chem.Eur. J.
(47) Sovadinova, I.; Palermo, E. F.; Huang, R.; Thoma, L. M.;
Kuroda, K. Biomacromolecules 2011, 12 (1), 260−268.
(48) Wilkinson, S. G. Microbial Lipids; Academic Press: London,
1988; Vol. 1.
(49) Mondal, J.; Zhu, X.; Cui, Q. A.; Yethiraj, A. J. Phys. Chem. B
2010, 114 (42), 13585−13592.
(50) Leontiadou, H.; Mark, A. E.; Marrink, S. J. J. Am. Chem. Soc.
2006, 128 (37), 12156−12161.
2009, 15 (5), 1123−1133.
(16) Mowery, B. P.; Lee, S. E.; Kissounko, D. A.; Epand, R. F.;
Epand, R. M.; Weisblum, B.; Stahl, S. S.; Gellman, S. H. J. Am. Chem.
Soc. 2007, 129 (50), 15474−15476.
(17) Rathinakumar, R.; Walkenhorst, W. F.; Wimley, W. C. J. Am.
Chem. Soc. 2009, 131 (22), 7609−7617.
(18) Wimley, W. C. ACS Chem. Biol. 2010, 5 (10), 905−917.
(19) Matsuzaki, K.; Harada, M.; Funakoshi, S.; Fujii, N.; Miyajima, K.
Biochim. Biophys. Acta 1991, 1063 (1), 162−170.
(20) Wakamatsu, K.; Takeda, A.; Tachi, T.; Matsuzaki, K. Biopolymers
2002, 64 (6), 314−327.
(21) Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Chem. Rev. 2001,
101 (10), 3219−3232.
(22) Porter, E. A.; Wang, X.; Lee, H. S.; Weisblum, B.; Gellman, S. H.
Nature 2000, 405 (6784), 298−298.
(23) Kuroda, K.; DeGrado, W. F. J. Am. Chem. Soc. 2005, 127 (12),
4128−4129.
(24) Palermo, E. F.; Sovadinova, I.; Kuroda, K. Biomacromolecules
2009, 10 (11), 3098−3107.
(25) Mowery, B. P.; Lindner, A. H.; Weisblum, B.; Stahl, S. S.;
Gellman, S. H. J. Am. Chem. Soc. 2009, 131 (28), 9735−9745.
1641
dx.doi.org/10.1021/bm300342u | Biomacromolecules 2012, 13, 1632−1641