3098
Biomacromolecules 2009, 10, 3098–3107
Structural Determinants of Antimicrobial Activity and Biocompatibility in Membrane-Disrupting Methacrylamide Random Copolymers Edmund F. Palermo,† Iva Sovadinova,‡ 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 Received July 11, 2009; Revised Manuscript Received September 16, 2009
Low molecular weight random copolymers bearing protonated primary amine groups and hydrophobic alkyl groups in the side chains were synthesized and their activities against E. coli, S. aureus, human red blood cells, and human epithelial carcinoma cells (HEp-2) were quantified. The mole fraction of alkyl side chains in the copolymers (falkyl) and the length of the alkyl chains were major determinants of the activities. Against E. coli cells, activity was diminished as falkyl was increased from 0 to about 0.2, but was then enhanced dramatically as falkyl was increased further. Activity against S. aureus was diminished continually with increasing falkyl. The cytotoxicity to human epithelial carcinoma cells also decreased with increasing falkyl. Conversely, hemolytic activity showed monotonic enhancement with increasing falkyl. The cationic homopolymer (falkyl ) 0) completely inhibited S. aureus growth at 3 µM (10.2 µg/mL) and completely inhibited metabolic activity in HEp-2 cells at 10 µM (34 µg/mL), although it did not induce any detectable hemolysis up to 645 µM (2000 µg/mL). Polymer-induced dye leakage from liposomes provided a biophysical basis for understanding the factors which modulate the polymer-membrane interactions. Disruption of Zwitterionic POPC vesicles induced by the copolymers was enhanced as falkyl increased, following trends similar to the hemolytic activity data. The ability of the polymers to permeabilize vesicles of POPE/POPG and DOPG/Lysyl-DOPG/CL displayed trends similar to trends in their activities against E. coli and S. aureus, respectively. This was interpreted as evidence that the antimicrobial mechanism employed by the polymers involves disruption of bacterial cell membranes. An investigation of leakage kinetics revealed that the cationic homopolymer induced a gradual release of contents from POPE/POPG and DOPG/Lysyl-DOPG/CL vesicles, while the more hydrophobic copolymers induced rapid dye efflux. The results are interpreted as evidence that the cationic homopolymer and hydrophobic copolymers in this study exert their antimicrobial action by fundamentally different mechanisms of membrane disruption.
Introduction Cationic, amphiphilic polymers have long attracted scientific and commercial interest due to their ability to kill bacteria in solution and on surfaces by a mechanism involving disruption of the cytoplasmic membrane.1 Ikeda et al. reported the antimicrobial activity of a cationic, amphiphilic polystyrene derivative in the 1980s,2 and many other biocidal polymer platforms have been studied since then.3-8 Recently, antimicrobials which are also nonhemolytic have been obtained from random copolymers of methacrylate derivatives,9,10 polymers of amphiphilic norbornenes,11 and random copolymers of β-lactams.12 The enhanced biocompatibility of these compounds is a feature that mimics host-defense peptides found in nature,13 rendering them potential candidates for a broad range of biomedical applications. However, a complete set of design principles, which are requisite to guide further development of antimicrobial polymers with low toxicity to human cells, have not been fully described to date. Systematic studies have provided descriptions of some parameters that modulate antimicrobial and hemolytic activity of synthetic polymers, toward a design rationale for future development. The spatial relationship between the alkyl side * To whom correspondence should be addressed. E-mail: kkuroda@ umich.edu. † Macromolecular Science and Engineering Center. ‡ Department of Biologic and Materials Sciences.
chains and the cationic groups in poly(vinyl pyridine)s,14 the charge density in polynorbornenes15 and the structure of cationic groups in polystyrenes16 have been shown to profoundly impact the antimicrobial and hemolytic activities. The biocompatibility of antimicrobial poly(vinyl pyridine) derivatives was enhanced by incorporating hydrophilic poly(ethylene glycol) side chains.17,18 Our laboratory recently reported a direct comparison of polymethacrylate derivatives containing either protonated amines or quaternary ammonium units, which revealed that the protonated primary amines are more effective at conferring cell-selective antimicrobial action.19 In these studies, a variety of polymer backbone structures were employed and activities were discussed largely in terms of the side chain properties. We wondered whether the properties of the polymer backbone structure would significantly affect the observed activities. Specifically, we hypothesized that a more hydrophilic backbone would improve the hemocompatibility of antimicrobial polymers, while side chain modification could be exploited to tune antimicrobial activity. Polymethacrylate random copolymers, a prototypical class of antimicrobials,9,10,19 contain ester groups which could be replaced with more hydrophilic amide linkages to demonstrate the effect of increased polarity near the polymer backbone. Kopecˇek and co-workers have extensively studied the polymerization of methacrylamides in the presence of thiols to control molecular weight.20 They have obtained a variety of biocompatible N-(2-hydroxypropyl)methacrylamide polymers conjugated to various flourophores, peptides, and other biologically
10.1021/bm900784x CCC: $40.75 2009 American Chemical Society Published on Web 10/05/2009
Membrane-Disrupting Methacrylamide Random Copolymers
active compounds for applications including cancer therapy.20-22 Hence, it seemed plausible that hemocompatible antimicrobials could be developed using a polymethacrylamide platform including cationic and hydrophobic side chains. To that end, we prepared methacrylamide copolymers containing primary ammonium chloride groups and hydrophobic alkyl groups. We investigated the chemical stability of the copolymers exposed to basic conditions by means of NMR spectroscopy and established their ionization behavior by potentiometric titration. Furthermore, we studied the effect of copolymer composition when the alkyl groups were butyl or hexyl on the antimicrobial and hemolytic potencies, as well as on the viability of human epithelial cells from larynx carcinoma (HEp-2 cells). The ability of the polymers to induce dye leakage from model lipid vesicles was also investigated to gain a biophysical basis for understanding the membrane-disrupting activity of these copolymers.
Materials and Methods Materials. N-(3-Aminopropyl)methacrylamide hydrochloride (APMAm · HCl, >98%) was purchased from PolySciences (Warrington, PA). 2,2′-Azobisisobutyronitrile (AIBN) was purchased from Sigma-Aldrich. Methyl 3-mercapto-propionate (MMP), butylamine, hexylamine, and methacryloyl chloride were purchased from Acros Organics. Reagent grade solvents were purchased from Fisher Scientific. Standardized solutions of 0.100 N sodium hydroxide and hydrogen chloride were purchased from Ricca Chemical Company. All reagents were used without further purification. The lipids 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1′-sn-glycerol) (POPG), 1,2dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (DOPG), 1,2-dioleoylsn-glycero-3-[phospho-rac-(3-lysyl(1-glycerol))] (Lysyl-DOPG), and 1, 1′,2,2′-tetraoleoyl cardiolipin (CL) were purchased as lyophilized powders from Avanti Polar Lipids (Alabaster, AL). Lipid stock solutions (10 mM) in chloroform were kept frozen at -80 °C, under a nitrogen blanket, until use. Solutions containing Lysyl-DOPG were used immediately after preparation according to the recommendations of Avanti. 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 Random Copolymers. Alkyl methacrylamides were prepared by reaction of hexylamine or butylamine with methacryloyl chloride in dichloromethane with triethylamine. The products were purified by washing with water, saturated NaHCO3(aq), and brine and then by silica gel chromatography using ethyl acetate/hexanes (1:1). Polymerization was carried out using a modified literature procedure.20 APMAm · HCl and butyl- or hexylmethacrylamide (0.5 mmol total monomer, various ratios) were dissolved in methanol/ethanol (1:1, 0.5 mL) in borosilicate glass test tubes. AIBN (0.005 mmol, 0.82 mg) and MMP (0.025 mmol, 2.7 µL) were added from concentrated stock solutions and the tubes were sealed with rubber septa and copper wire. After degassing with N2 for 5 min each, the tubes were submerged in an oil bath set to 60 °C and left to stir for 24 h. Then, the polymers were purified by repeated precipitation from methanol into diethylether. They were subsequently dried under vacuum and then lyophilized to afford white powders. Potentiometric Titration. Polymers were titrated according to the procedure of Thomas and Tirrell.23 Briefly, polymers (∼10 mg) were dissolved in aqueous saline (10 mL, [NaCl] ) 150 mM) in a glass scintillation vial to give a polymer concentration of ∼0.3 mM and purged with N2. A nitrogen blanket was maintained over the sample throughout the titration and the temperature was held at 24 ( 0.2 °C. Aliquots of standardized 0.100 N sodium hydroxide or 0.100 N hydrochloric acid (5 or 10 µL) were injected using a Hamilton syringe and the solution was stirred for 5 min to allow thermal and chemical
Biomacromolecules, Vol. 10, No. 11, 2009
3099
equilibrium. The pH was recorded after each injection using an Accumet Basic AB15 pH meter. The degree of ionization of the polymer was derived from the charge neutrality condition as
R≡
[-NH3+] [-NH2] + [-NH3+]
)
[Cl-] + [HO-] - [Na+] - [H+] [-NH2] + [-NH3+]
where [-NH2] is the concentration of neutral amine groups in solution and [-NH3+] is the concentration of amine groups that are protonated. Antimicrobial Activity. The lowest polymer concentration required to completely inhibit growth of bacteria, defined as the minimum inhibitory concentration (MIC), was determined by a turbidity-based microdilution assay in Muller Hinton broth (MH) according to the procedure approved by The National Committee for Clinical Laboratory Standards (NCCLS)24 with modifications recommended on the Web site of Prof. R. E. W. Hancock.25 Briefly, Escherichia coli ATCC 25922 or Staphylococcus aureus ATCC 25923 in the midlogarithmic growth phase were diluted to OD600 ) 0.001 in MH broth. This stock suspension of bacteria (90 µL) was mixed with serial dilutions of a polymer stock solution (10 µL) in each well of a 96-well polypropylene microplate, which was not treated for tissue culture (Corning #3359). 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 was observed relative to the negative growth control, sterile 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. Hemolytic Activity. Toxicity to human red blood cells (RBCs) was assessed by a hemoglobin release assay. RBCs (1 mL) were diluted into phosphate-buffered saline (9 mL; PBS: 10 mM phosphate, 150 mM NaCl, pH 7.4) and then centrifuged at 1000 rpm for 5 min. The supernatant was carefully removed using a pipet. The RBCs were then washed with PBS two additional times. The resulting stock (10% v/v RBC) was diluted 3-fold in PBS to give the assay stock (3.3% v/v RBC). The assay stock (90 µL) was then mixed with each of the polymer serial dilutions (10 µL) on a sterile 96-well round-bottom polypropylene microplate to give a final solution of 3% v/v RBC, which corresponds to approximately 108 red blood cells per mL based on counting in a hemacytometer. PBS (10 µL) or Triton X-100 (10 µL, 1% v/v) were added instead of polymer solution as negative and positive hemolysis controls, respectively. The microplate was secured in an orbital shaker at 37 °C and 180 rpm for 60 min. The plate was then centrifuged at 1000 rpm for 10 min. The supernatant (10 µL) was diluted into PBS (90 µL) and the 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). 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 the empirical Hill equation, H ) 1/((HC50/[P])n + 1), where [P] is the total concentration of polymer. The fitting parameters were n and HC50. All experiments were performed three times in triplicate. Cell Culture. Cytotoxicity experiments were carried out using the HEp-2 cell line, which are human epithelial cells isolated from larynx carcinoma. It should be noted that the HEp-2 cell line is likely contaminated with HeLa cells derived from cervical cancer.26 Cells were grown in minimum essential medium (MEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 1 mM pyruvate, and 0.1 mM nonessential amino acids at 37 °C, 5% CO2, and 95% relative humidity. The doubling time of the HEp-2 cells in this supplemented medium is about 22-24 h. In medium without added serum, the cells generally proliferate only with difficulty.
3100
Biomacromolecules, Vol. 10, No. 11, 2009
Palermo et al.
Scheme 1. Synthesis of the Random Copolymers
XTT Assay. HEp-2 cells were seeded into the 96-well cell culture plates (Falcon #3072, U.S.A.) at a density of 104 cells per well. After a 24 h incubation, the cell confluence was about 50-60%, and the culture medium was replaced with serial dilutions of polymer stock solutions in an antibiotic- and serum-free MEM. The viability of cells exposed to the polymers was assessed using a commercial kit (Cell Proliferation Kit II, Roche, U.S.A.). After a 4 or 24 h exposure to polymers, the cells were washed with PBS and MEM without phenol red and serum was added (75 µL). A solution of sodium 3′-[1(phenylaminocarbonyl)-3,4-tetrazolium]-bis-(4-methoxy-6-nitro) benzene sulfonic acid hydrate (XTT) and an electron-coupling reagent N-methyl dibenzopyrazine methyl sulfate (PMS) were added to each well to give the final concentration of 0.2 mg/mL and 1.5 µg/mL, respectively. To assess the effect of polymers on the conversion of substrate, control wells containing only medium, polymer, and XTT with PMS were included. After a 4 h incubation at 37 °C in the presence of 5% CO2, the amount of the orange colored formazan derivative produced by the metabolic cellular activity was determined by the absorbance at 450 nm (test wavelength) and 750 nm (background wavelength). The spectrophotometer was calibrated to zero absorbance using medium without cells. The cell viability was determined relative to that of a control containing intact cells, which were exposed to only solvent. All samples were run in triplicate. We present the polymer concentration data in µM units because we are interested in the activity of individual polymer chains. The MIC, hemolysis, and cytotoxicity data in µg/mL units can be found in the Supporting Information (Figures S5, S6, and S7). Because the molecular weights do not vary significantly between polymers in this study, the same general conclusions can be drawn from either set of figures. Liposome Dye Leakage. A solution of lipid (100 µL, 10 mM) in chloroform was slowly evaporated under a gentle N2 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 NaCl and measured using a vapor-pressure osmometer, so that liposomes could be prepared without initial osmotic pressure across the membranes. 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 from Amersham Biosciences (Uppsala, Sweden) using a buffer containing no dye (10 mM HEPES, 150 mM NaCl, pH 7.4, 280 ( 5 mmol/kg). The concentration of lipid in the obtained suspension was determined by a colorimetric phosphorus assay.27 This solution was diluted in the same buffer to a lipid concentration of 11.11 µM. This suspension (90 µL) was 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 (0.1% v/v, 10 µL) were employed as the negative and positive controls. After 1 h, the fluorescence intensity in each well was recorded using a Thermo Fisher microplate reader 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 the fluorescence intensity recorded in the well, F0 is the intensity in the negative control well, and FTX is the intensity in the positive
Table 1. Characterization of the Random Copolymers polymer P0 PB20 PB36 PB54 PB78 PH18 PH33 PH51 PH63
R)
falkyl
DP
Mn (kDa)
butyl ′′ ′′ ′′ hexyl ′′ ′′ ′′
0.00 0.20 0.36 0.54 0.78 0.18 0.33 0.51 0.63
17 20 14 15 14 19 17 16 16
3.1 3.6 2.4 2.4 2.3 3.5 3.0 2.9 2.9
control well. For the measurement of dye leakage kinetics, the liposome stock was diluted to 10 µM in a 2 mL quartz cuvette and fluorescence signal was monitored. Then, an aliquot of polymer solution was injected into the cuvette via a syringe port in the spectroflourometer lid and fluorescence intensity was recorded every 2 s for 10 min. Triton X-100 solution (4% v/v, 5 uL) was injected as the positive lysis control at the end of the time course.
Results and Discussion Polymer Synthesis and Characterization. Two series of random copolymers containing hydrophobic alkyl groups and primary amine groups were prepared in one step by free radical polymerization in methanol/ethanol mixtures (Scheme 1). This scheme is slightly modified from a report by Kopecˇek and coworkers on the polymerization of N-(2-hydroxypropyl)methacrylamide.20 The thiol compound was employed as a chain transfer agent to control the number average degree of polymerization (DP). The feed ratio of comonomers was varied to tailor the average mole fraction of alkyl side chains in the copolymers, falkyl. Peak integrations in the 1H NMR spectra were analyzed to determine the values of DP and falkyl for each polymer (Supporting Information, Figure S1). The copolymers were obtained with DP values in the range of 14-20 (Table 1). This low molecular size range (2.4-3.6 kDa) is similar to previously described polymers10-12 and peptides,28 which are known to possess antimicrobial and nonhemolytic properties. The falkyl values ranged from 0 to 0.78 in the PB series (R ) butyl) and 0 to 0.63 in the PH series (R ) hexyl), in good correlation with the comonomer feed compositions used in the polymerizations (Supporting Information, Figure S2). The chemical stability of the polymers in basic aqueous solution was examined by 1H NMR. We found that the aminefunctionalized side chains in P0 were unaltered by stirring in pH 10 for 24 h (Supporting Information, Figure S3). In that condition, primary amine-functionalized polymethacrylates have been previously found to undergo multiple chemical changes due to isomerization and hydrolysis.19,29,30 Hence, this new class of cationic, amphiphilic polymethacrylamides possesses improved chemical stability relative to the well-studied polymethacrylate derivatives.
Membrane-Disrupting Methacrylamide Random Copolymers
Biomacromolecules, Vol. 10, No. 11, 2009
3101
of E. coli and S. aureus in the micromolar concentration range, with MIC values that depend on falkyl and the length of the alkyl group (Figure 2). In the same assay conditions, the host defense peptide magainin-2 and the bee venom toxin peptide melittin showed MIC values against E. coli of 51 and 4 µM, respectively. Against S. aureus, the MIC values of magainin-2 and melittin were found to be >100 µM and 2 µM, respectively. Hence, the copolymers in this study showed antimicrobial potency comparable to that of peptides found in nature.
Figure 1. Potentiometric titration data for the cationic homopolymer P0 in aqueous 150 mM NaCl solution. Solid symbols denote the forward titration with NaOH and empty symbols denote the backtitration with HCl.
Potentiometric Titration. Before testing the antimicrobial and hemolytic activity, the ionization behavior of the polymers in this study was examined by means of potentiometric titration using a previously described procedure.23 Establishing the extent of ionization in the polymers is requisite for the proper interpretation of structure-activity data because the protonated (cationic) and deprotonated (neutral) amine groups may play different roles in their antimicrobial or hemolytic action. It is well-known that pH influences the activity of membrane-lytic peptides31 and polymers32-34 bearing ionizable groups in the side chains. Hence, we assess the ionization curves of the polymers in this study before proceeding to measure their biological activities. The cationic homopolymer P0 was titrated to the equivalence point with NaOH and subsequently backtitrated with HCl in good agreement (Figure 1). The extent of ionization in this polymer is R > 0.98 in the pH range of 6-8. This indicates that nearly all of the primary amine groups in the side chains of P0 are protonated (cationic) in the physiologically relevant pH range. The R value of this polymer reaches 0.5 when the pH is increased to about 9.7. Hence, the polymer can be considered a two-component system of cationic ammonium groups and hydrophobic alkyl groups in buffer of pH 7.4. Relative to P0, low molecular weight polymers of aminoethylmethacrylate, poly(AEMA), were previously shown to be significantly less basic: whereas 99% of the amine groups in P0 are cationic at pH 7.4, only 72% of those in poly(AEMA) are cationic in that condition.19 This difference may be related to the difference in polymer polarity. The more polar amide groups in P0 are likely better solvated in water relative to the polymethacrylates, which contain less polar ester linkages. P0 may therefore adopt a more extended chain conformation, minimizing Coulombic repulsion of cationic charges and thereby increasing basicity.35 The cationic groups in P0 are attached to the polymer backbone by a propyl linker, whereas those of poly(AEMA) are separated by an ethyl linkage, which would also likely relieve Coulombic repulsions. Also, the increased dielectric of the microenvironment surrounding the amine groups in the polymethacrylamides would increase basicity relative to the lower dielectric environment in polymethacrylates.23 Utilizing amide groups instead of esters in the random copolymers enhances water solubility but also furthermore affects the number of charged groups at a given pH due to the aforementioned factors. Hence, it should be noted that the polarity of the polymerizable units and the ionization behavior of the side chains are interdependent design parameters. Antimicrobial Activity. The amphiphilic, cationic copolymers in this study were able to completely inhibit the growth
The copolymers in the PH series, which contain hexyl side chains, showed MIC values as much as an order of magnitude lower than those of the PB copolymers, which have butyl groups. This difference in MIC values increased with increasing falkyl. Against E. coli, PB54 had a MIC value of 340 µM (833 µg/mL), while PH51 had a MIC of 14 µM (42 µg/mL). These two polymers have similar DP and falkyl, which indicates that longer hydrophobic alkyl side chains impart more potent antimicrobial activity. Hence, it appears that increasing the hydrophobicity of side chains in the antimicrobial polymers enhances activity, as previously discussed in the literature.10-12 In both polymer series, the MIC values against E. coli went through local maxima as falkyl was varied in the range of 0-0.8. When falkyl was increased from 0 to ∼0.2, the MIC values increased, which implies a loss of antimicrobial activity. Apparently, the antimicrobial action depends on the cationic charges in the polymer, which are diluted in the polymer chains by including a small fraction of hydrophobic repeat units. Increasing falkyl beyond 0.2 caused a decrease in MIC, implying that increasing the hydrophobicity enhances antimicrobial activity in the higher falkyl range. Hence, we consider that the cationic homopolymers and the hydrophobic copolymers may exert different mechanisms of antimicrobial action. Against S. aureus, the MIC values went through a maximum as falkyl was increased in the PB series. In the PH series, the MIC values against S. aureus approached a plateau with increasing falkyl. Interestingly, the maximum in the PB series against S. aureus occurred at higher falkyl (∼0.55) compared to the E. coli MIC data, which went through a maximum near falkyl ) 0.2. This suggests that S. aureus cells are more sensitive to the effect of cationic charge density in the copolymers, while E. coli cells are more susceptible to the action of the hydrophobic groups. This distinction may arise from the differences in membrane structures between the two tested microorganisms. In the context of other primary amine containing polymers discussed in the literature, the polymethacrylamides in this work displayed anomalous activity minima (MIC maxima) as the hydrophobic comonomer content was increased. Kuroda and DeGrado found that the antimicrobial activity of polymethacrylates was enhanced by increasing the mole fraction of butyl groups in the copolymer side chains.9 Kuroda et al. further reported that antimicrobial activity of polymethacrylates increased sigmoidally with increasing fraction of hydrophobic comonomer, and was fit to the empirical Hill equation.10 Mowery et al. described random copolymers of beta lactams which showed increasing activity as the fraction of hydrophobic comonomer was increased.12 Ilker et al. likewise demonstrated that the antimicrobial activity of polynorbornenes was enhanced by increasing the hydrophobicity of the side chains.11 The MIC values often display local minima with increasing hydrophobicity, which can be ascribed to the formation of microaggregates in the aqueous environment in the case of excessively hydrophobic polymers. The MIC values going through a local maximum (antimicrobial activity at a minimum) with increasing
3102
Biomacromolecules, Vol. 10, No. 11, 2009
Palermo et al.
Figure 2. Minimum inhibitory concentration (MIC) values as functions of falkyl for the random copolymers against (A) E. coli and (B) S. aureus. A lower MIC value implies that the polymer is a more potent antimicrobial agent. We present the data in µM units because we are interested in the activity of individual polymer chains. The MIC data expressed in µg/mL units can be found in the Supporting Information (Figure S5).
Figure 3. Percentages of hemolysis induced by polymers in (A) the PB series and (B) the PH series as functions of total polymer concentration. The symbols represent experimental data and the curves are best fits to the Hill equation. Error bars represent the standard deviation from triplicate measurements.
hydrophobic comonomer in random copolymers bearing primary amines has not been previously reported to the best of our knowledge. Decreasing MIC values with increasing hydrophobicity has been reported in the case of polymers bearing quaternary ammonium salts (QAS), such as copolymers of quaternized pyridine and alkylmethacrylates.14 Additionally, studies on surface-bound QAS polymers have shown that potent antimicrobial activity requires high cationic charge density.36-38 Reduction of the cationic charge density with increasing falkyl may be related to the observed loss of activity in the random copolymers presented here. However, whether these polymers exert their antimicrobial effects by a mechanism similar to that of QAS-polymer surfaces remains unclear at present. A significant factor that distinguishes polymethacrylamides in this work from previously studied polymethacrylates is their ionization behavior: in pH 7.4, only about 72% of the amine groups in the polymethacrylates are protonated,19 while 100% of the amines in P0 are cationic. In our prior studies on polymethacrylates, we found that the antimicrobial and hemolytic activities were significantly reduced when the experiments were performed in pH 6, in which 100% of the amine groups are protonated.19 It is possible that the presence of neutral amines enhances activity, which implies that we can tune the activity of amphiphilic polymers by choosing ionizable groups with various pKa values. Hemolytic Activity. To examine the biocompatibility of the polymers, hemoglobin release from human red blood cells was measured after incubation with each of polymers at various concentrations. The extent to which copolymers in this study
lysed human red blood cells depended on the polymer concentration as well as the identity and amount of the hydrophobic side chains (Figure 3). In the PB series, polymers with falkyl < 0.55 showed no significant hemolysis up to roughly 570-670 µM (2000 µg/ mL), which is comparable to the nonhemolytic property of magainin-2, which induced less than 10% hemolysis up to 100 µM in the same assay conditions. This demonstrates the veracity of our initial hypothesis: indeed, nonhemolytic antimicrobials could be obtained by using a polar methacrylamide design platform. On the other hand, increasing the hydrophobicity of the side chains did eventually lead to increased hemolysis. PB78 induced a significant amount of hemolysis (HC50 ) 79 µM, Figure 3A) and the polymers in the PH series were even more hemolytic. PH33 showed an HC50 value of 36 µM (108 µg/mL) and PH51 and PH63 showed HC50 values lower than 5 µM (54 against E. coli and >640 against S. aureus. In the context of other polymers in the literature,41 the latter value is remarkably high, which implies that P0 is an excellent nonhemolytic staphylocidal agent. As we hypothesized, the more polar methacrylamide platform afforded polymers which do not lyse human red blood cells. On the other hand, the HC50/MIC values of the copolymers decrease with increasing hydrophobic content. While the HC50/MIC values of PH18 were >11 for E. coli and >170 for S. aureus, those of PH33 were 0.92 for E. coli and 2.2 for S. aureus. This shows that increasing the hexylmethacrylamide content by only 15% precipitously negates the selectivity for bacteria vs red blood cells. Despite the hemocompatibility of P0, the polycation is highly cytotoxic to HEp-2 cells. The value of IC50/MIC for P0 was 54 >3.6 >2.8 >5.6 0.71 >11 0.92 0.15 0.15
>640 >86 >19 >12 1.1 >170 2.2 0.15 0.09
