Block versus Random Amphiphilic Copolymers as Antibacterial Agents

Aug 16, 2011 - ... and Materials Sciences, School of Dentistry, University of Michigan, Ann Arbor, Michigan 48109, United States. bS Supporting Inform...
6 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/Biomac

Block versus Random Amphiphilic Copolymers as Antibacterial Agents Yukari Oda,† Shokyoku Kanaoka,† Takahiro Sato,† Sadahito Aoshima,*,† and Kenichi Kuroda*,‡ † ‡

Department of Macromolecular Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan Department of Biologic and Materials Sciences, School of Dentistry, University of Michigan, Ann Arbor, Michigan 48109, United States

bS Supporting Information ABSTRACT: We examined the antibacterial and hemolytic activities in a series of amphiphilic block and random copolymers of poly(vinyl ether) derivatives prepared by base-assisting living cationic polymerization. Block and random amphiphilic copolymers with similar monomer compositions showed the same level of activity against Escherichia coli. However, the block copolymers are much less hemolytic compared to the highly hemolytic random copolymers. These results indicate that the amphiphilic copolymer structure is a key determinant of activity. Furthermore, the block copolymers induced dye leakage from lipid vesicles consisting of E. coli-type lipids, but not mammalian lipids, while the random copolymers disrupted both types of vesicles. In addition, both copolymers displayed bactericidal and hemolytic activities at concentrations 1 or 2 orders of magnitude lower than their critical (intermolecular) aggregation concentrations (CACs), as determined by light scattering measurements. This suggests that polymer aggregation or macromolecular assembly is not a requisite for the antibacterial activity and selectivity against bacteria over human red blood cells (RBCs). We speculate that different single-chain conformations between the block and random copolymers play an important role in the antibacterial action and underlying antibacterial mechanisms.

’ INTRODUCTION Development of new antibacterial agents is urgently needed due to the emergence of antibiotic-resistant bacteria. It has been a scientific challenge to create new antibacterial agents that are not susceptible to the development of resistance mechanisms in pathogenic bacteria. One approach is the implementation of naturally occurring antimicrobial host-defense peptides found in the innate immune system.1,2 These peptides fold into secondary conformations such as α-helix or β-sheet upon binding to bacterial cell membranes, segregating their hydrophobic and cationic side chains into distinct domains. Although the molecular mechanisms of these peptides remain in debate, the peptides bind to bacterial cell surfaces and insert into cell membranes, leading to disruption of membrane integrity and cell death.3 Alternatively, some peptides penetrate the membrane and interact with cellular targets such as enzymes and DNA/RNA, inhibiting macromolecular synthesis. The peptides may have multiple potential targets and complex mechanisms to exert their antimicrobial effect.2,4,5 Bacterial cell membranes contain negatively charged lipids in greater abundance than mammalian cell membranes. Hence, cationic peptides preferentially bind to bacteria by electrostatic attraction, resulting in the selective targeting of bacteria over human cells.3 The peptides are promising antibiotic candidates; however, their pharmaceutical and biomedical applications are limited due to the lack of ability to use systemically poor pharmacokinetics, susceptible to proteolysis and high manufacturing costs. To address these issues, cationic and amphiphilic polymers have been utilized as a platform to mimic the structural features r 2011 American Chemical Society

and functions of antimicrobial peptides.6,7 To that end, amphiphilic polymethacrylates,8 polynorbornenes,9 and polyamides10 have been prepared as synthetic antimicrobials. Many studies have elucidated the role of key structural parameters in controlling biological activity. For example, the effects of amphiphilicity,914 molecular weight,8,12 type of cationic charges,15 and PEGylation16,17 have been shown to influence activity. However, few examples of antibacterial activity of amphiphilic block copolymers can be found in the literature.18,19 Song et al. reported that the antibacterial and hemolytic activities of polymers prepared by alternating ring-opening metathesis polymerization can be controlled by the exact distance of ammonium groups along the backbone.20 However, the role of copolymer structures in the antibacterial action and underlying antibacterial mechanism has not yet been systematically investigated. A computational study on amphiphilic polymethacrylate derivatives indicated that the amphiphilic copolymer structures play an important role in the polymer membrane interaction; the hydrophobic segments of block copolymers inserted into the hydrophobic region of lipid bilayers, and the cationic segments remained in the aqueous environment.21 We hypothesized that the antibacterial activity and membrane disruption mechanism by amphiphilic copolymers could be controlled by the amphiphilic copolymer structures. To that end, we prepared amphiphilic vinyl ether copolymers with random and block sequences by living cationic polymerization, Received: June 9, 2011 Revised: August 15, 2011 Published: August 16, 2011 3581

dx.doi.org/10.1021/bm200780r | Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

Figure 1. Amphiphilic poly(vinyl ether)s with block and random copolymer structures.

and investigated the effect on their antibacterial and hemolytic activities (Figure 1). In this report, we wish to report the structureactivity relationship as well as to propose the potential antibacterial mechanisms of amphiphilic block and random vinyl ether copolymers.

’ EXPERIMENTAL SECTION Materials. 2-Phthalimidoethyl vinyl ether (PIVE) was prepared by the reaction of 2-chloroethyl vinyl ether with potassium phthalimide.22 It was purified by recrystallization from methanol twice, then from ethyl acetate, and vacuum-dried for more than 3 h prior to use. Isobutyl vinyl ether (IBVE; TCI; > 99.0%) and dichloromethane (Nacalai Tesque; 99%) were purified by double distillation over calcium hydride. 1-(Isobutoxy)ethyl acetate [IBEA; CH3CH(OiBu)OCOCH3] was prepared from IBVE and acetic acid as previously reported.23 1,4-Dioxane (Wako; > 95%) was distilled over calcium hydride and then lithium aluminum hydride. Commercial Et1.5AlCl1.5 (Nippon Aluminum Alkyls; 1.0 M solution in toluene) was used without further purification. All materials mentioned above for cationic polymerization except for PIVE were stored in brown ampules under dry nitrogen. The lipids L-α-phosphatidylcholine (egg; eggPC), 1-palmitoyl-2-olelyl-sn-glycero3-phosphoethanolamine (POPE), and 1-palmitoyl-2-olelyl-sn-glycero3-phospho-(10 -sn-glycerol) (POPG) were purchased as lyophilized powders from Avanti Polar Lipids (Alabaster, AL). Lipid stock solutions (10 mM) in chloroform were kept frozen at 80 C until use. Human red blood cells (red blood cells leukocytes reduced adenine saline added) were obtained from the American Red Cross Blood Services, Southeastern Michigan Region. Synthesis of Amphiphilic Copolymers.24 Polymerization was carried out under a dry nitrogen atmosphere in a glass tube with a threeway stopcock baked at 250 C for 10 min before use. A typical example for the sequential block copolymerization of IBVE and PIVE in CH2Cl2 is as follows: The reaction was initiated by the addition of an Et1.5AlCl1.5 solution (0.6 mL of 200 mM in CH2Cl2) with a dry medical syringe into a mixture of IBVE (0.08 mL), 1,4-dioxane (0.6 mL), and cationogen (IBEA) (0.3 mL of 200 mM in hexane) in CH2Cl2 at 0 C. The total volume of the reaction mixture was 6.0 mL ([IBVE]0 = 100 mM, [IBEA]0 = 10 mM, [Et1.5AlCl1.5]0 = 20 mM, and [1,4-dioxane] = 1.2 M). After 6 min, when IBVE had been consumed almost quantitatively, the second monomer PIVE (1.5 mL of 800 mM in CH2Cl2) was added to the polymerization mixture, temporarily cooled to 78 C. The polymerization was terminated by adding prechilled methanol (23 mL) containing a small amount of an aqueous ammonia solution (0.3 wt %) to the reaction solution. The quenched reaction mixture was washed with dilute hydrochloric acid to remove initiator residue. The reaction

ARTICLE

mixture was further washed with sodium hydroxide solution and water. The solvent was removed under reduced pressure, and the resultant polymers were dried under vacuum overnight. To remove the pendant phthalimide groups, the phthalimide-containing copolymer (0.40 g) was dissolved in a mixture of 1,4-dioxane/ methanol (2:1 v/v; 30 mL) and hydrazine monohydrate (0.8 mL, 10 equiv to the imide units in the polymer) was added. The solution was heated to reflux with stirring for 3 h. After the volume of solvent and unreacted hydrazine was decreased under vacuum, a white powder (polymeric ammonium salt of phthalylhydrazine) was precipitated. Then 0.5 M hydrochloric acid (16 mL, 5 equiv to the imide groups of the starting polymer) was added to the residue. The mixture was heated to reflux for 15 min with vigorous stirring. The solution was diluted with water and kept heating for an additional 45 min. The resulting insoluble byproduct (phthalylhydrazide) was removed by filtration. The filtrate was neutralized with 1.0 M aqueous sodium hydroxide (8 mL, equiv to the hydrochloric acid employed), and the polymer product was purified by dialysis against distilled water for at least 2 days and then Milli-Q water for a day (MWCO of dialysis tubes are 2000 or 3500 for the polymers with low MW < 7100, or 12000 for MW ∼ 13200). Preparation of FITC-Labeled Block Copolymer. Amphiphilic block copolymer (10 mg, [NH2] = 0.081 mmol) was dissolved in DMF (10 mL) and triethylamine (30 μL, 0.24 mmol). Fluorescein isothiocyanate (FITC, Aldrich) (3.2 mg, 8.1  103 mmol) in DMF (2 mL) was added into the reaction mixture. The reaction mixture was stirred for 4 h at room temperature in the dark. After the solvent was removed under reduced pressure, the residue was dissolved in methanol. Unreacted FITC was removed by size exclusion chromatography (Sephdex LH-20 gel, Amersham Biosciences, Uppsala, Sweden) using methanol. Comparing the molar absorbance coefficient of F-B3826 and free fluorescein, the average number of FITC molecule per block copolymer chain is 0.45 (Supporting Information). This corresponds to one FITC molecule in 62 amine groups. Characterization of Copolymers. The molecular weight distribution (MWD) of the polymers was measured by gel permeation chromatography (GPC) in chloroform at 40 C with three polystyrene gel columns [Tosoh; TSK gel G-4000HXL, G-3000HXL, and G-2000HXL; exclusion limit molecular weight = 4  105, 6  104, and 1  104, respectively; column size = 7.8 mm (internal diameter)  300 mm; flow rate = 1.0 mL/min] connected to a Tosoh DP-8020 pump, a CO-8020 column oven, a UV-8020 ultraviolet detector, and an RI-8020 refractive-index detector. The number-average (Mn) and weight-average molecular weight (Mw) and polydispersity index (Mw/Mn) were calculated based on 16 polystyrene standards (Tosoh; Mn = 5771.09  106, Mw/Mn < 1.1). Nuclear magnetic resonance (NMR) spectra were recorded using a JEOL JNM-ECA 500 spectrometer (500 MHz), a Varian MR400 spectrometer (400 MHz), a Varian Inova 400 spectrometer (400 MHz), or a Varian Inova 500 spectrometer (500 MHz). Static and Dynamic Light Scattering Measurements. Static and dynamic light scattering measurements were carried out with an ALV/DLS/SLS-5000 light scattering system at 37 C. Vertically polarized light with the wavelength of 532 nm emitted from an Nd:YAG laser was used as the incident light, and the scattering system was calibrated using toluene as the reference material. Polymer solutions were prepared to 501000 μg/mL and 5002000 μg/mL in HEPES buffer (10 mM HEPES, 150 mM NaCl, pH 7) for B3826 and R4025, respectively. The solution was filtrated with a 0.45 μm cellulose acetate membrane filter. Antibacterial Assay.25,26. Polymer stock solutions were prepared in DMSO (10 or 20 mg/mL) or in 2-propanol (20 mg/mL) for polymers with low solubility to DMSO (R3879 and B3977). The stock solution was serially diluted 16 2-fold by 0.01% acetic acid. A single colony of Escherichia coli was incubated in Mueller-Hinton (MH) broth at 37 C with gentle shaking overnight. The bacterial suspension was then diluted by MH broth to OD600 = 0.1 and incubated again for 3582

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules 90 min. The bacterial culture in the midlogarithmic phase (OD600 ∼ 0.50.6) was diluted to OD600 = 0.005 with HEPES buffer (10 mM HEPES, 150 mM NaCl, pH 7), corresponding to ∼106 cfu/mL. This bacterial suspension (90 μL) was then mixed with the serial dilutions of polymer (10 μL) in each well of a 96-well polypropylene microplate, which was not treated for tissue culture (Corning #3359). After a 4 h incubation period at 37 C, the biocidal concentration for 99.9% killing (BC99.9) was determined by inoculating the aliquot of the bacteria suspension (100 μL) with 102-fold dilution by HEPES buffer on an agar plate. After incubation at 37 C overnight, the number of colonies was counted. The BC99.9 value was determined as the polymer concentration which showed no colony formation (3-log reduction or 99.9% killing). When bactericidal activity was determined in MH broth, the bacterial culture in the midlogarithmic phase (OD600 ∼ 0.50.6) was diluted with MH broth to be OD600 = 0.001, corresponding to ∼2  105 cfu/mL. This bacterial suspension (90 μL) was then mixed with the serial dilutions of polymer (10 μL) in each well of a 96-well polypropylene microplate. After an 18 h incubation period at 37 C, the BC99.9 was determined by inoculating the aliquot of the bacteria suspension (10 μL) with 104-fold dilution by HEPES buffer on an agar plate. Colony counting was performed in duplicate. The BC99.9 values are reported as an average of the two trials except B5826 and B7826 for which only one assay was performed. Hemolysis Assay. Hemolytic activity of polymers was assessed by monitoring hemoglobin release from human red blood cells (RBCs). The same polymer dilutions prepared for the antibacterial assay were used in the hemolysis assay. RBCs (1 mL) were diluted into HEPES buffer (9 mL; 10 mM HEPES, 150 mM NaCl, pH 7) and then centrifuged at 1000 rpm for 5 min. The supernatant was carefully removed using a pipet. The RBCs were then washed with HEPES buffer two additional times. The resulting RBC dispersion (10% v/v RBC) was diluted 3-fold in HEPES buffer to give the assay stock (3.3% v/v RBC). The assay stock (90 μL) was then mixed with each of the polymer dilutions (10 μL) on a sterile 96-well polypropylene microplate to give a final solution of 3% v/v RBC, which corresponds to approximately 2  108 RBCs/mL based on counting in a hemocytometer. HEPES buffer (10 μL) was added instead of polymer solution as negative hemolysis control. Triton X-100 (10 μL, 1% v/v) or melittin (10 μL, 3.1 μM) was used as positive hemolysis control. The assay plate was incubated for 60 min at 37 C in an orbital shaker at 100 rpm. The plate was then centrifuged at 1000 rpm for 10 min. The supernatant (10 μL) was diluted into HEPES buffer (90 μL) and the absorbance at 405 nm was recorded using a microplate reader (Thermo Scientific Varioskan Flash). The fraction of hemolysis was defined as H = (A  A0)/(A100  A0), where A is the absorbance reading of the sample well, A0 is the negative hemolysis control (buffer), and A100 is the positive hemolysis control (Triton X-100 or melittin). Hemolysis was plotted as a function of polymer concentration and the HC50 was defined as the polymer concentration, which causes 50% hemolysis relative to the positive control. We estimated this value by fitting the sigmoidal data to empirical Hill equation, H = 1/{(HC50/[P])n + 1}, where [P] is the total concentration of polymer. The fitting parameters were n and HC50. In some cases, hemolysis did not reach 50% up to the highest polymer concentration tested (1000 μg/mL) and the HC50 was not determined. All experiments were performed three times in triplicate. The HC50 values and errors are reported as averages and standard deviations of three independent experiments, respectively. Hemolysis curves for each polymer presented in this report are representative data from the three experiments. Dye Leakage from Liposomes.27 A solution of lipid (250 μL, 10 mM) in chloroform was slowly evaporated under a gentle nitrogen stream and subsequently dried under vacuum for 12 h. An aqueous buffer (10 mM HEPES, 50 mM sulforhodamine B (SRB), pH 7.4) was adjusted to an osmolarity of 280 ( 5 mmol/kg by addition of saturated

ARTICLE

NaCl and measured using a vapor-pressure osmometer. The dry lipid film was resuspended in this buffer, vigorously vortexed for 5 min, and subjected to 10 freeze/thaw cycles between dry ice in acetone and a 50 C water bath. Then it was passed 21 times through a mini-extruder equipped with two stacked polycarbonate membranes of 400 nm average pore size. Unincorporated dye was removed by size exclusion chromatography over Sepharose Cl-4B gel (Amersham Biosciences, Uppsala, Sweden) using buffer (10 mM HEPES, 150 mM NaCl, pH 7). The concentration of lipid in the obtained suspension was determined by a colorimetric phosphorus assay.28 This solution was diluted in the same buffer to a lipid concentration of 11.11 μM. The suspension (90 μL) was then mixed with polymer stock solutions (10 μL) on a 96-well black microplate to give a final lipid concentration of 10 μM in each well. The assay buffer (10 μL) and Triton X-100 (0.1% v/v, 10 μL) were employed as the negative and positive control, respectively. After a 1 h incubation at 37 C with orbital shaking (100 rpm), the fluorescence intensity in each well was recorded using a microplate reader (Thermo Scientific Varioskan Flash) with excitation and emission wavelengths of 565 and 586 nm, respectively. The fraction of leaked SRB in each well was calculated according to the expression L = (F  F0)/(FTX  F0), where F is fluorescence intensity recorded in the well, F0 is the intensity in the negative control well, and FTX is the intensity in the positive control well. Fluorescence Microscopy. B3826 was dissolved in DMSO (20 mg/mL) with a small amount of F-B3826 to prepare a homogeneous mixture of these polymers. This stock solution was diluted by HEPES buffer to give 200 μg/mL solution containing 1% DMSO. This solution was further diluted by HEPES buffer containing 1% v/v DMSO to give polymer solutions at concentrations of 50, 25, and 10 μg/mL. Comparing the maximum absorbance of polymer solution at 25 μg/mL with that of free fluorescein, FITC content in this mixture was estimated to be 5.3 mol % relative to the total number (mole) of B3826 polymer chains. Confocal fluorescence microscopy images of the mixtures were recorded using Eclipse Ti Confocal Microscope C1 (Nikon, EZ-C1 software). FITC was excited at 488 nm.

’ RESULTS AND DISCUSSION Polymer Design and Synthesis. To investigate the effect of amphiphilic copolymer structures on their antibacterial activity, we have prepared a series of amphiphilic block and random poly(vinyl ether) derivatives (Figure 1) by base-assisting living cationic polymerization (Scheme 1). This polymerization method has been previously used to prepare various poly(vinyl ether)s, including block and star-shaped polymers with precisely controlled molecular weight (MW) and sequence of block segments.29 In this study, we have synthesized low MW block poly(vinyl ether)s bearing protected primary amino groups in side chains by this polymerization method, and the following deprotection gave water-soluble cationic copolymers (MW ∼ 4000).24 We targeted low molecular weights because the low MW polymers were expected to be less hemolytic compared to high MW polymer counterparts according to the previous report on polymethacrylate derivatives.8 For the block copolymer synthesis, the hydrophobic monomer isobutyl vinyl ether (IBVE) was polymerized first, followed by addition of phthalimide-protected amine vinyl ether (PIVE). Random copolymers were synthesized under the same conditions by polymerizing mixtures of these monomers. GPC analysis showed these protected polymers had a narrow molecular weight distribution (Table 1). 1H NMR analysis indicated that the final monomer compositions and the degree of polymerization (DP) were close to the value calculated from the monomer feed 3583

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

Scheme 1. Synthesis of Amphiphilic Block Copolymers by Living Cationic Polymerizationa

a Reaction conditions for copolymers with DP ∼ 40: [IBVE]0 = 0.100.30 M, [PIVE]add = 0.100.30 M (0.80 M solution in CH2Cl2), [IBEA]0 = 10 mM, [Et1.5AlCl1.5]0 = 20 mM, [1,4-dioxane] = 1.2 M, in toluene at 0 C.

Table 1. Characterization of Copoly(IBVE-PIVE)s protected precursor polymer DPa entry

copolymer structure

PIVE

IBVE

MPIBVEb (mol %)

Mn (NMR)c

Mw/Mn (GPC)d

deprotected polymere

1

homopolymer

44 (40)

0 (0)

0

9600

1.11

H44

2

random copolymer

30 (30)

10 (10)

25

7500

1.13

R4025

3

18 (20)

20 (20)

53

5900

1.14

R3853

4

8 (10)

30 (30)

79

4700

1.31

R3879

5

28 (30)

10 (10)

26

7100

1.16

B3826

6

diblock copolymer

19 (20)

20 (20)

51

6100

1.11

B3951

7

9 (10)

30 (30)

77

5000

1.37

B3977

8 9

43 (45) 58 (60)

15 (15) 20 (20)

26 26

10800 14600

1.19 1.14

B5826 B7826

10

92 (100)

50 (50)

35

25000

1.14

B14235

a

Calculated by comparing the integral area of signals from the phthalimide group relative to that of the methyl group of the isobutyl group in 1H NMR spectra. The theoretical DP based on the feed monomer composition is presented in the parentheses. b Mole percentage of IBVE relative to the total number of monomers in a polymer chain. c Calculated based on DP and molecular weight of monomeric units. d Determined by GPC in CHCl3, polystyrene calibration. e The polymers after deprotection are denoted by R/BXy (R: random, B: block, X: total DP, y: mol % of IBVE).

compositions (Table 1). These results suggest that the cationic polymerization proceeded in a living fashion to give amineprotected copolymers with low molecular weights (MW < 10000). The phthalimide protecting groups were removed by treating the polymers with hydrazine, and the crude polymers with primary amine groups were further purified by dialysis against water to remove low MW impurities. 1H NMR analysis indicated the yield of deprotection was >99% (Supporting Information). The DPs of these deprotected polymers are ∼40, corresponding to MW of 35004000, which was calculated based on the molecular weight of monomers and DP. The deprotected polymers are denoted as R/BXy (R: random copolymers, B: block copolymers, X: total DP, and y: mole percentage of IBVE units relative to the total number of monomers in a polymer chain); for

example, the block copolymer containing DP = 38 and 26 mol % IBVE units is referred to as B3826. Antibacterial Activity. To examine the effect of amphiphilic copolymer structures and compositions on their bactericidal activity against E. coli, a reduction in the number of viable cells was determined after incubation with the polymers. Although a turbidity-based assay has been used as the accepted standard assay method in this field to determine the minimum inhibitory concentration (MIC) of antibacterial polymers, it requires using a growth medium, which contains nutrient proteins and fragments. These components in a medium may bind nonspecifically to the polymers and interfere with the physicochemical characterization of polymers in solution. Therefore, to elucidate the relationship between polymer aggregation and antibacterial activity, 3584

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

Figure 2. Bactericidal activity (BC99.9) against E. coli of block copolymers (filled squares) and random copolymers (empty circles) with DP ∼ 40 as a function of the mole percentage of hydrophobic isobutyl side chains (MPIBVE) in HEPES buffer (A) and MH broth (B). Each data point and error bar represents the average and standard deviation of two independent experiments in duplicate. The BC99.9 value of B3977 in MH broth was over 1000 μg/mL.

we tested the polymers in an antibacterial assay using HEPES buffer containing only defined salts and concentrations. Because the HEPES buffer is a nongrowth medium, a reduction in the number of viable E. coli cells was determined by inoculating aliquots on agar plates and counting the number of bacterial colonies. The bactericidal concentration (BC99.9) of the polymers was determined as the lowest polymer concentration to cause at least a 3-log reduction (99.9% killing) in the number of viable E. coli cells after incubation with the polymers for 4 h at 37 C. In the absence of polymer (positive viability control), the number of viable E. coli cells increased from 3  106 cfu/mL to 5  107 cfu/mL after 4 h incubation, indicating that the bacteria cells were viable in this condition. The cationic homopolymer H44 displayed bactericidal activity against E. coli with a BC99.9 value of 1.6 μg/mL. The random and block copolymers displayed bactericidal effects against E. coli in a BC99.9 value of 163 μg/mL, depending on the monomer compositions (Figure 2A). Interestingly, there is no significant difference in the BC99.9 values between the block and random copolymers. This might suggest that the monomer composition is responsible for bactericidal activity of polymers rather than the amphiphilic copolymer structures. The BC99.9 values of block and random copolymers both had no significant change up to 50% mole percentages of hydrophobic isobutyl groups (MPIBVE), and the polymers with ∼80% MPIBVE showed high BC99.9 values (3163 μg/mL). In addition, the bactericidal activity of these polymers was also determined in growth medium Mueller-Hinton (MH) broth after an 18 h incubation to determine if the activity profile of polymers depends on the bacterial growth condition (Figure 2B). The BC99.9 values in MH broth are generally in the range of 161000 μg/mL, which is over 10 times higher than in HEPES buffer. This is likely due to competition of bactericidal action of polymers with bacterial growth as well as aggregation of polymers with components in the broth, resulting in the lower activity (higher BC99.9 values). The BC99.9 values of the random copolymers showed a minimum for R4025 and increased with increasing MPIBVE (>53%) while the BC99.9 values of the block copolymers increased at high MPIBVE, which is similar to the trend shown in HEPES buffer. This suggests that the activity of copolymers also generally decreases at high MPIBVE in the growth assay condition.

It has been previously reported that the antibacterial activity of amphiphilic random copolymers including polymethacrylates,11 polynorbornenes,9 and polyamides10 were enhanced as the hydrophobicity of copolymers was increased, and the MIC values reach a constant at higher compositions of hydrophobic groups. In general, increasing hydrophobicity of amphiphilic polymers enhances the antibacterial activity because the hydrophobic groups are likely to insert into the hydrophobic regions of lipid membranes, facilitating membrane disruption. Excess hydrophobicity of polymer chains however, induces their self-association in aqueous media, which would reduce the number of active polymer chains and result in no activity enhancement for polymers with higher compositions of hydrophobic groups.11 Therefore, we expected that the BC99.9 of amphiphilic poly(vinyl ether)s would also decrease or the activity is enhanced with increasing MPIBVE. However, the activity of the random poly(vinyl ether)s decreased at the high MPIBVE values, which contrasts to our previous results on polymethacrylates11 and other reports.9,10 To explain this, BC99.9 dependence on MPIBVE, we first speculated that the poly(vinyl ether)s are more hydrophobic than other polymers, and addition of isobutyl groups to the poly(vinyl ether)s would not increase the activity, but rather enhance the intramolecular or intermolecular hydrophobic aggregation of polymers in solution, resulting in the reduced activity. However, examining the chemical structures of other polymers reported previously,10,11 the poly(vinyl ether) backbone and isobutyl side chains are not necessarily more hydrophobic than other polymers, and the intrinsic hydrophobicity of the poly(vinyl ether) backbone may not explain the observed activity dependence on MPIBVE. We further speculated that the decrease in activity of the random poly(vinyl ether)s could be related to the flexible backbone of poly(vinyl ether)s, which would allow the hydrophobic side chains to associate in aqueous solution more easily compared to more rigid polymers. Therefore, the flexible poly(vinyl ether) chains could facilitate the formation of intramolecular aggregates, which reduce the number of active polymer chains against bacteria. We speculate here that the polymers are likely in the form of single chains rather than intermolecular aggregates in the concentration ranges of BC99.9 because the BC99.9 is orders of magnitude lower than the critical (intermolecular) aggregation concentrations (CACs) determined by light scattering measurements (discussed below). The activity reduction by the 3585

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

Figure 3. Percentages of hemolysis induced by homopolymer (A), copolymers with MPIBVE = ∼25% (B), 50% (C), and 80% (D). The data points and error bars represent the average and standard deviations of a representative measurement in triplicate. The sigmoidal curves of homopolymer (A) and random copolymers (BD) are best fits to the Hill equation.

intramolecular hydrophobic aggregation appears to be more dominant for the random poly(vinyl ether)s than the activity enhancement by hydrophobicity, as found in other polymers. On the other hand, the more rigid polymers (polymethacrylates, polynorbornenes, and polyamides) are likely to have more extended conformations, and the hydrophobic groups would be more exposed to the surrounding aqueous environment. As a result, the more rigid polymers reflect their hydrophobicity more directly on the membrane disruption action, increasing their antibacterial activity. In addition, it has been also reported that the activity of quaternized poly(2-vinly pyridine)s was enhanced by copolymerizing hydrophilic monomers with PEGs side chains.30 The activity enhancement was explained by the increased hydrophilicity of polymers due to the PEG modifications, which possibly reduces protein adsorption on the polymers. The random poly(vinyl ether)s with low or no hydrophobic isobutyl groups could have a similar effect of hydrophilicity on their antibacterial activity. Hemolytic Activity. As an initial assessment of toxic effect, the lytic activity of the copolymers against human red blood cells (hemolysis) was also examined. In general, the hemolytic activity of the copolymers studied here was strongly dependent on their amphiphilic copolymer structures although the copolymers displayed the same level of bactericidal activity regardless the structural differences (Figure 3). The cationic homopolymer H44 induced no hemolysis at low polymer concentrations (1000 (max 42.5 ( 6.3%)c

0.98

R4025 R3853

1.6 ( 0.0 3.1 ( 0.0

15.6 ( 0.0 62.5 ( 0.0

>1000 >500

0.49 ( 0.17 1.8 ( 0.24

R3879

31.3 ( 0.0 1000 ( 0.0 18.9 ( 1.3

>1000

B3826

2.4 ( 0.91 62.5 ( 0.0

>1000 (max 37.7 ( 2.8%)c

2.0

B3951

3.1 ( 0.0

>1000 (max 12.9 ( 6.3%)c

0.49

B3977

62.5 ( 0.0 >1000

>1000 (max 22.8 ( 8.5%)c

0.78

250 ( 0.0

a

Determined in HEPES buffer or MH broth against E. coli. b The HC50 values and errors were reported as average and standard deviations of the three independent experiments, respectively. c Local maximum values of hemolysis induced by each polymer. d The lowest polymer concentration to induce hemagglutination.

Table 3. Effect of Polymer Length on Bactericidal Activity of Block Copolymers BC99.9a in HEPES

a

in MH broth

polymer

(μg/mL)

(μM)

(μg/mL)

(μM)

B3826

2.4

0.66

62.5

17.5

B5826

0.78

0.15

62.5

11.9

B7826

1.6

0.22

31.3

4.4

Determined in HEPES buffer or MH broth against E. coli.

In addition to the hemolytic activity, the cationic homopolymer H44 and block copolymers induced agglutination of RBCs (hemagglutination) although the random copolymers did not induce any hemagglutination (Table 2). Synthetic polycations are previously reported to induce hemagglutination,31 and it is speculated that the cationic groups of polymer surfaces bind to anionic groups of RBC surfaces, which cross-link the cells and induce cluster formation of RBCs. It is likely that the cationic ammonium groups of the copolymers studied here are also responsible for the hemagglutination; however, there appears to be no clear relationship between the lowest polymer concentration to induce hemagglutination (CH) and the number of amine groups in a polymer chain (Table 2). The hemagglutination mechanism might involve other interactions between polymers and RBC surfaces, including the hydrophobic association of polymers and biopolymers on the cell surfaces.32 Molecular Weight Effect. We further investigated the effects of DP of block copolymers on their bactericidal and hemolytic activities (Table 3, Figure 4). The copolymers with the same MPIBVE of 26%, but with different DPs (DP = 3878), displayed slightly different BC99.9 values (Table 3). B5826 showed the lowest BC99.9 value, indicating that there is an optimal length of polymers for high efficacy. Comparing the BC99.9 values given in μM, which reflects the activity of individual polymer chain, the BC99.9 values of B5826 and B7626 are similar. Similarly, the BC99.9 values given in μg/mL determined in MH broth decreased by only 2-fold as the polymer length was increased. These results

Figure 4. Percentages of hemolysis induced by B3826 (empty circles), B5826 (empty squares), and B7826 (filled triangles) as a function of polymer concentration. Error bars represent the standard deviations from triplicate measurements.

indicate that increasing the polymer length does not significantly improve the bactericidal activity. On the other hand, the hemolytic activity of copolymers depends on the polymer length (Figure 4). B3826 caused a local maximum of 30% hemolysis at ∼100 μg/mL, but the block polymers with increased DP, B5826, and B7826, induced no hemolysis up to 1000 μg/mL. This indicates that the hemolytic activity of the block copolymers was reduced by increasing the DP. Liposome Dye Leakage. To quantify the ability of polymers to disrupt cell membranes, polymer-induced leakage of the small fluorescent dye molecule sulforhodamine B (SRB) from large unilamellar lipid vesicles (LUVs) was examined (Figure 5). LUVs with POPE/POPG (4:1) and eggPC are simplified model membranes of the E. coli and RBC cell membranes, respectively. These model vesicles are limited by their lipid compositions, as the bacterial cell and RBC membranes are a complex of proteins, lipids, and polysaccharides on the cell surface with membrane asymmetry. Although it is difficult to recapitulate them in LUVs, the leakage experiments will provide a gauge of the membrane permeabilizing activity of the polymers. The polymers H44, B3826, and R4025 were tested, which have the similar BC99.9 values of 1.62.4 μg/mL against E. coli. These copolymers all induced dye-leakage from LUVs with POPE/POPG in the similar leakage curves although the leakage levels at high concentrations are somewhat different (Figure 5A). The similar ability of these polymers to disrupt lipid bilayers appears to mirror the similar BC99.9 values of these polymers (Table 2). These results may suggest that the polymers exert bactericidal effect by disrupting E. coli cell membranes. However, dye leakage from model membranes does not necessarily relate to permeabilization of cytoplasmic membranes by the polymers, but the polymers may bind to potential targets on membrane surfaces or penetrate the cytoplasmic membranes and interact with cytoplasmic targets, causing cell death. The molecular details of antibacterial action and mechanisms of these polymers are, however, unclear at this point. The leakage from eggPC LUVs depended on the amphiphilic copolymer structures (Figure 5B). The leakage induced by the block copolymer B3826 was lower than ∼20% although the random copolymer R4025 induced >80% leakage in the entire concentration range (0.0293.8 μg/mL). These results suggest 3587

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

Figure 5. Percentages of SRB leakage from LUVs composed of (A) POPE/POPG (4:1) and (B) eggPC. The LUVs were incubated with the polymers B3826 (empty circles), R4025 (empty squares), and H44 (filled triangles) at 37 C for 1 h. The error bars represent the standard deviations from triplicate measurement.

Figure 6. Confocal fluorescent microscopy image (single layer image) of 50 μg/mL B3826 containing F-B3826 in HEPES buffer (1% v/v DMSO; FITC: 5.3 mol % relative to the total number (mole) of B3826 polymer chains).

that the block copolymer selectively disrupts the bacteria-type membranes over the erythrocyte-type membranes, which appears to reflect the selective activity of polymers against E. coli over RBCs although the antibacterial actions of polymers remain unclear. The homopolymer H44 also induced dye leakage up to 40% at 4.0 μg/mL, which also seems to reflect the hemolytic activity (Figure 3A) although the eggPC liposomes are very simplified models for RBCs. Polymer Aggregation and Antibacterial Activity. The data presented above indicate that the block and random copolymer structures play an important role in the antibacterial efficacy and selectivity to bacteria over RBCs. To examine the role of the copolymer aggregation in the antibacterial mechanisms, we investigated the structures of polymer aggregates by confocal fluorescent microscopy and static and dynamic light scattering (SLS and DLS) methods. The fluorescence image of B3826 aggregates doped with a FITC-labeled polymer (F-B3826) showed the formation of particles with a diameter of ∼500 nm above 50 μg/mL (Figure 6). The SLS analysis for aqueous solutions of the block copolymer B3826 showed that the radius of gyration ÆS2æ1/2, weight-average molecular mass Mw, and aggregation number (i.e., the number of polymer chains in an aggregate) of the block copolymer aggregate were 200 nm, 8.1  107 g/mol, and 20000,

respectively, at the copolymer concentration above 100 μg/mL (Table 4 and Supporting Information). The hydrodynamic radius of the block copolymer aggregate obtained by DLS was as large as 250 nm. The aggregate sizes estimated by the light scattering experiments are consistent with those found in the fluorescence image. Because the molecular weight of B3826 is less than 104, the maximum end-to-end distance of the single copolymer chain would be ∼3 nm.33 Therefore, the large particle size and aggregation number may indicate that the spherical B3826 aggregate is not likely a single star-like micelle, but a large spherical aggregate or a vesicle.34 On the other hand, light scattering intensities from aqueous solutions of the random copolymer R4025 were much weaker than B3826 solutions, and the radius of gyration and the aggregation number were 27 nm and 22, respectively, in the copolymer concentration range 5002000 μg/mL (Table 4 and Supporting Information). The hydrodynamic radius of the R4025 aggregate was not estimated because scattering intensities were too low to obtain accurate autocorrelation functions. These results indicate that the hydrophobicity of isobutyl groups randomly distributed along the R4025 copolymer chain is much weaker than that of the IBVE block in B3826. At the polymer concentration of 10 μg/mL, we observed no particles in fluorescence imaging of aqueous B3826 solutions (data not shown). Furthermore, we determined the critical (intermolecular) aggregation concentration (CAC) of the aqueous B3826 solution by light scattering.35 As shown in Table 4 and Supporting Information, the CAC of the aqueous block copolymer B3826 solution was 36 μg/mL, being consistent with the fluorescence imaging. On the other hand, the CAC for the aqueous solution of the random copolymer R4025 estimated by light scattering was 380 μg/mL (Table 4 and Supporting Information), which is much higher than that for the aqueous block copolymer B3826 solution due to the weaker hydrophobicity. Interestingly, the CAC values are more than 1 order of magnitude higher than the BC99.9 values in HEPES buffer for both B3826 and R4025 (Table 4). The HC50 value for R4025 is also significantly lower than the CAC value. In addition, the copolymers induced dye leakage from liposomes at the polymer concentrations lower than their CACs (Figure 5). These results indicate that these copolymers are membrane active below CACs, suggesting that intermolecular aggregation or macromolecular assembly is not requisite for the antibacterial activity and selectivity against bacteria over RBCs. In other words, the conformation of individual polymer chains is responsible for the bactericidal activity. 3588

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

Table 4. Characterization of Polymer Aggregates in HEPES Buffer polymer

ÆS2æz1/2 a (nm)

RHb (nm)

Mwc

Naggd

B3826

200

250

8.1  107

20000

36

2.4 ( 0.91

>1000

9.4  104

22

380

1.6 ( 0.0

0.49 ( 0.17

R4025

27

g

n.d.

CACe (μg/mL)

BC99.9f (μg/mL)

HC50 (μg/mL)

a

Radius of gyration, determined by SLS. b Hydrodynamic radius, determined by DLS. c Weight-average molecular mass, determined by SLS. Aggregation number, calculated from the weight-average molecular weight of polymer chain and Mw. e Critical (intermolecular) aggregation concentration, determined by SLS. f Determined in HEPES buffer. g The scattering intensities were too low to obtain accurate autocorrelation functions. d

Figure 7. Schematic presentation of proposed antibacterial and hemolytic activities.

Proposed Antibacterial Mechanism. The relationship between CAC and BC99.9 values of random and block copolymers indicates that single polymer chains are likely responsible for bactericidal activity and selectivity against E. coli over RBCs. Based on these results, we propose the role of the copolymer conformations in the antibacterial mechanism (Figure 7). The block copolymers could form intramolecular aggregates with a hydrophobic core wrapped by the cationic segment in water, which can be regarded as single-molecular cationic particles. The cationic properties of these intramolecular aggregates would increase their electrostatic binding to anionic lipopolysaccharides (LPS) on the E. coli cell surface and may replace the divalent cations, which stabilize the outer membrane structure. The replacement of these divalent cations destabilizes the membrane structure and increases permeability of the outer membrane, which would promote uptake of the polymers into the cell surfaces as found in antimicrobial peptides.36 The polymers could diffuse through the peptidoglycan layer to reach cytoplasmic membranes with anionic lipids and cause membrane disruption, leading to lethal leakage of cellular contents. The cationic groups of polymers could induce the segregation of anionic lipids37 or cause strong perturbation of the cell wall integrity.38 Alternatively, the polymers might transit the cytoplasmic membrane and interact with potential targets in the cytoplasma,

inhibiting macromolecular synthesis as found for some antimicrobial peptides.2,4,5 As for the hemolytic activity, the hydrophobic segment is shielded by the cationic surfaces, reducing the hydrophobic binding of polymers to the cell membranes of RBCs, thus, causing no significant hemolytic activity. However, the high cationic charge density of the intramolecular aggregates could interact strongly with the anionic biopolymers on the RBC surfaces, which cause hemagglutination. On the other hand, random copolymers would take a randomcoil or slightly shrunk conformation because the hydrophobic groups randomly distributed along the polymer chain may not form the hydrophobic core (Figure 7), if the hydrophobic content is low enough.39 These hydrophobic groups would interact with RBC cell membranes and cause cell lysis. When the hydrophobic content is high, the random copolymer may form a flower micelle.39 Even if the case, many hydrophobic groups are outside the hydrophobic core,39 which can interact with RBC cell membranes and cause cell lysis. In addition, the cationic functionality of random copolymers would also enhance the binding of polymers to bacterial cell membranes and cause membrane disruption. Light scattering experiments and fluorescence imaging (Figure 6) indicated the formation of particles with a diameter of 400500 nm above the CAC. It was suggested that intermolecular association 3589

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules of polymers or peptides could decrease the antibacterial activity due to the difficulties of polymer or peptide aggregates in traversing the bacterial cell wall.12,4042 However, the block polymer displayed at least the same level of activity (99.9% killing) above the CAC compared to the polymer activity at lower concentrations. One possible explanation is that the single polymer chains or intramolecular aggregates coexisting with large particles of intermolecular aggregates in solution above the CAC exert an antibacterial effect. In general, above the CAC, the concentration of single polymer chains remains constant to that at the CAC.43 Because the BC99.9 value (2.4 μg/mL) of the block copolymer B3826 is smaller than the CAC value (36 μg/mL), the single polymer chains could still be responsible for their antibacterial activity of 99.9% killing at polymer concentrations above the CAC even if the intermolecular aggregates are presumably not active. However, the intermolecular aggregates also could contribute to the bactericidal activity, although it is not evident if they are active at this point. Liu et al. speculated that cationic polymer nanoparticles with a diameter of 177 nm cause steric hindrance and cross-linking of peptidoglycans in the cell wall or disrupt cell membranes, resulting in bacterial cell death.38 Although the large cationic particles of intermolecular aggregates reported here are not likely to penetrate into the cell wall structure because of their large size (diameter = 400500 nm), they could bind to the bacterial cell surfaces, and their high density of cationic charges could disrupt the cell wall integrity and cell membranes for killing E. coli on contact. The block copolymers could have these multiple antibacterial mechanisms exerted by single polymer chains and large particles, which may be complementary to each other.

’ CONCLUSION The amphiphilic copolymer structure is a key determinant in a polymer’s antibacterial and hemolytic activities. The block and random copolymers with similar polymer lengths and monomer compositions displayed the same level of bactericidal activity against E. coli. The block copolymers displayed selective activity against E. coli over RBCs while the random copolymers did not and were hemolytic. Liposome dye leakage experiments showed that the block copolymers selectively disrupt the bacteria-type membranes over the erythrocyte-type membranes while the random copolymer did not show any selectivity. The difference in the single-chain conformation between the amphiphilic block and random copolymers in aqueous solution may play an important role in the different interaction with the cell membranes. So far, the relationship between the polymer structure and antibacterial activity has been studied mainly using amphiphilic random copolymers being heterogeneous with respect to the composition and degree of polymerization. On the other hand, the block copolymers studied here have definite compositions and relatively narrow molecular weight distributions, which facilitates the determination of the optimal monomer composition and polymer length for potent activity. This study highlights the potential of amphiphilic copolymer structures as a new design motif to improve the activity as well as to understand the mechanism of antibacterial actions of membrane-active antibacterial polymers. ’ ASSOCIATED CONTENT

bS

Supporting Information. Molecular weight distribution curves for the precursor polymers, 1H NMR spectra for the

ARTICLE

precursor and the deprotected polymers, SLS data, characterization of F-B3826. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]; [email protected].

’ ACKNOWLEDGMENT This research was supported by the Global Center Of Excellence (G-COE) Program “Global Education and Research Center for Bio-Environmental Chemistry” at Osaka University (to YO) and Grant-in-Aid for Scientific Research on Innovative Areas of “Fusion Materials: Creative Development of Materials and Exploration of Their Function through Molecular Control” (no. 2206) from the Ministry of Education, Culture, Sports, Science and Technology, Japan (MEXT). This research was supported by NSF CAREER Award (DMR-0845592) (to KK) and Department of Biologic and Materials Sciences, University of Michigan School of Dentistry. We thank Professor Robert Davenport at the University of Michigan Hospital for the red blood cells, and Dr. Eric Krukonis, University of Michigan School of Dentistry for the use of the microplate reader. We also thank Dr. Takeshi Suwabe at University of Michigan School of Dentistry for his help on the fluorescence microscopy experiment. ’ CONFLICT OF INTEREST 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. ’ REFERENCES (1) Hancock, R. E. W.; Lehrer, R. Trends Biotechnol. 1998, 16, 82–88. (2) Zasloff, M. Nature 2002, 415, 389–395. (3) Matsuzaki, K. Biochim. Biophys. Acta 1999, 1462, 1–10. (4) Hancock, R. E. W.; Sahl, H.-G. Nat. Biotechnol. 2006, 24, 1551–1557. (5) Brogden, K. A. Nat. Rev. Microbiol. 2005, 3, 238–250. (6) Parelmo, E. F.; Kuroda, K. Appl. Microbiol. Biotechnol. 2010, 87, 1605–1615. (7) Tew, G. N.; Scott, R. W.; Klein, M. L.; DeGrado, W. F. Acc. Chem. Res. 2010, 43, 30–38. (8) Kuroda, K.; DeGrado, W. F. J. Am. Chem. Soc. 2005, 127, 4128–4129. (9) Ilker, M. F.; N€usslein, K.; Tew, G. N.; Coughlin, E. B. J. Am. Chem. Soc. 2004, 126, 15870–15875. (10) 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, 15474–15476. (11) Kuroda, K.; Caputo, G. A.; DeGrado, W. F. Chem.—Eur. J. 2009, 15, 1123–1133. (12) Mowery, B. P.; Lindner, A. H.; Weisblum, B.; Stahl, S. S.; Gellman, S. H. J. Am. Chem. Soc. 2009, 131, 9735–9745. (13) Lienkamp, K.; Tew, G. N. Chem.—Eur. J. 2009, 15, 11784–11800. (14) Lienkamp, K.; Madkour, A. E.; Musante, A.; Nelson, C. F.; N€usslein, K.; Tew, G. N. J. Am. Chem. Soc. 2008, 130, 9836–9843. 3590

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591

Biomacromolecules

ARTICLE

(15) Palermo, E. F.; Kuroda, K. Biomacromolecules 2009, 10, 1416–1428. (16) Chakrabarty, S.; King, A.; Kurt, P.; Zhang, W.; Ohman, D. E.; Wood, L. F.; Lovelace, C.; Rao, R.; Wynne, K. Biomacromolecules 2011, 12, 757–769. (17) Allison, B. C.; Applegate, B. M.; Youngblood, J. P. Biomacromolecules 2007, 8, 2995–2999. (18) Sauvet, G.; Fortuniak, W.; Kazmierski, K.; Chojnowski, J. J. Polym. Sci., Part A: Polym. Chem. 2003, 41, 2939–2948. (19) Park, E. S.; Kim, H. S.; Kim, M. N.; Yoon, J. S. Eur. Polym. J. 2004, 40, 2819–2822. (20) Song, A.; Walker, S. G.; Parker, K. A.; Sampson, N. S. ACS Chem. Biol. 2011, 6, 590–599. (21) Ivanov, I.; Vemparala, S.; Pophristic, V.; Kuroda, K.; DeGrado, W. F.; McCammon, J. A.; Klein, M. L. J. Am. Chem. Soc. 2006, 128, 1778–1779. (22) Hashimoto, T.; Ibuki, H.; Sawamoto, M.; Higashimura, T. J. Polym. Sci., Part A: Polym. Chem. 1988, 26, 3361–3374. (23) Aoshima, S.; Higashimura, T. Macromolecules 1989, 22, 1009–1013. (24) Oda, Y.; Kanaoka, S.; Aoshima, S. J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 1207–1213. (25) National Committee for Clinical Laboratory Standards Methods for dilution antimicrobial susceptibility for bacteria that grow aerobically; Villanova University: Villanova, PA, 2003. (26) Hancock, R. E. W. Minimal Inhibitory Concentration (MIC) Determination for Cationic Antimicrobial Peptides by Modified Microtitre Broth Dilution Method; Department of Microbiology and Immunology, University of British Columbia: British Columbia, Canada, http://www. cmdr.ubc.ca/bobh/methods.htm, June 2001 (posting date online; date last accessed, December 2008). (27) Palermo, E. F.; Sovadinova, Iva.; Kuroda, K. Biomacromolecules 2009, 10, 3098–3107. (28) Fiske, D. H.; Subbarow, Y. J. Biol. Chem. 1925, 66, 375–400. (29) Aoshima, S.; Kanaoka, S. Chem. Rev. 2009, 109, 5245–5287. (30) Sellenet, P. H.; Allison, B.; Applegate, B. M.; Youngblood, J. P. Biomacromolecules 2007, 8, 19–23. (31) Katchalsky, A.; Danon, D.; Nevo, A.; De Vries, A. Biochim. Biophys. Acta 1959, 33, 120–138. (32) Sevastianov, V. I.; Drushlak, I, V.; Eberhart, R. C.; Kim, S. W. Macromol. Symp. 1996, 103, 1–4. (33) Matsuda, Y.; Kawata, T.; Sugihara, S.; Aoshima, S.; Sato, T. J. Polym. Sci., Part B: Polym. Phys. 2006, 44, 1179–1187. (34) Sato, T.; Matsuda, Y. Polym. J. 2009, 41, 241–251. (35) Debye, P. Ann. N.Y. Acad. Sci. 1949, 51, 575–592. (36) Hancock, R. E. W. Lancet 1997, 349, 418–422. (37) Epand, R. F.; Mowery, B. P.; Lee, S. E.; Stahl, S. S.; Lehrer, R. I.; Gellman, S. H.; Epand, R. M. J. Mol. Biol. 2008, 379, 38–50. (38) Liu, L.; Xu, K.; Wang, H.; Tan, P. K. J.; Fan, W.; Venkatraman, S. S.; Li, L.; Yang, Y. Y. Nat. Nanotechnol. 2009, 4, 457–463. (39) Tominaga, Y.; Mizuse, M.; Hashidzume, A.; Morishima, Y.; Sato, T. J. Phys. Chem. B 2010, 114, 11403–11408. (40) Chen, Y.; Mant, C. T.; Farmer, S. W.; Hancock, R. E. W.; Vasil, M. L.; Hodges, R. S. J. Biol. Chem. 2005, 280, 12316–12329. (41) Chen, Y.; Guarnieri, M. T.; Vasil, A. I.; Vasili, M. L.; Mant, C. T.; Hodges, R. S. Antimicrob. Agents Chemother. 2007, 51, 1398–1406. (42) Meng, H.; Kumar, K. J. Am. Chem. Soc. 2007, 129, 15615–15622. (43) Israelachvili, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic Press: London, 1992.

3591

dx.doi.org/10.1021/bm200780r |Biomacromolecules 2011, 12, 3581–3591