• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Antimicrobial polymers: Peptide-mimetic design and mechanism of action

ProQuest Dissertations and Theses, 2011
Dissertation
Author: Edmund Francis Palerrmo
Abstract:
The proliferation of antibiotic-resistant bacteria and the decline in new antibiotic drug approvals have lead to an emergent need for novel antibacterial strategies. In nature, host defense peptides (HDPs) kill invading pathogens as part of the innate immune system in multicelluar organisms. HDPs are diverse in sequence and conformation, but certain physiochemical features are conserved: they are cationic (+1-5 net charge), amphiphilic (30-50% hydrophobic residues), and low molecular weight (10-50 residues). Their putative mechanism involves the disruption of bacterial membranes. Because the activity of these peptides is modulated by physiochemical characteristics, rather than receptor-specific interactions, we hypothesized that synthetic macromolecules could mimic their function. The peptide-mimetic design strategy has several advantages including diminished cost of manufacturing, lower susceptibility to degradation, and access to the wide range of chemical functionalities obtained by synthetic means. Random copolymers composed of ∼50% cationic (aminoethylmethacrylate) and ∼50% hydrophobic (methyl methacrylate) monomer units, and molecular weights of 1-4 kDa, inhibited bacterial growth at concentrations as low as ∼10μM whereas they did not lyse human red blood cells up to ∼1000µM. Polymers bearing primary amine groups in the side chains displayed more potent activity than analogous polymers containing quaternary ammonium salts, attributed the higher affinity of primary amines for phosphate lipid head groups, leading to efficient membrane permeabilization. Polymethacrylamides showed low µM antibacterial activity without harming RBCs, although they were cytotoxic to human epithelial cells. After further optimization, copolymers bearing aminobutyl side chains showed highest activity among other polymers bearing shorter (aminoethyl) and longer (aminohexyl) cationic side chains. Biophysical experiments using model lipid membranes, fluorescence imaging, flow cytometry, and molecular dynamics simulations indicated that the mechanism of antimicrobial action involves membrane disruption. Pore formation in cell membranes followed by osmotic lysis and leakage of cytoplasmic contents, results in bacterial cell death. Although synthetic polymers are heterogeneous in structure (polydispersity, random sequence, and heterotacticity) and conformation, they nevertheless effectively mimic the antimicrobial activity and mechanism of host defense peptides. Hence, this emerging class of materials can be excellent candidates for a wide variety of applications in combating the rise of drug-resistant infectious diseases.

TABLE OF CONTENTS DEDICATION ii ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF FIGURES vii LIST OF APPENDICES xi LIST OF ABBREVIATIONS xii CHAPTER I. Introduction and Background: The Convergence of Peptide and Polymer Science toward Novel Antibiotics 1 Motivation and Objectives 1 Host Defense Peptides 2 Synthetic Polymer Biocides 8 Peptide-mimetic Antimicrobial Polymers 9 II. Effect of Cationic Group Structure on Antimicrobial and Hemolytic Activity of Amphiphilic Copolymers 15 Introduction 15 Polymer Design, Synthesis and Characterization 18 Antimicrobial Activity 21 Hemolytic Activity 25 Selectivity 26 Bactericidal Activity vs. pH 28 Hemolytic Activity vs. pH 31 Conclusions 34 III. Role of Cationic Group Structure in Membrane Binding and Disruption by Amphiphilic Copolymers 36 Introduction 36 Dansyl-labeled Polymer Synthesis and Characterization 38 Water-Octanol Partition Coefficients 41 Polymer-Liposome Dissociation Constants 43 Polymer-Induced Liposome Leakage 47 Solid-state NMR 49 IV

Membrane Disruption Mechanism 51 Conclusions 53 IV. Amphiphilic Polymethacrylamides as an Antimicrobial Design Platform 55 Introduction 55 Polymer Synthesis and Characterization 56 Antimicrobial Activity 59 Hemolytic Activity 62 Cytotoxicity 64 Liposome Dye Leakage 67 Conclusions 72 V. Role of Cationic Side Chain Spacer Groups in Activity and Mechanism of Antimicrobial Action by Amphiphilic Copolymers 74 Introduction 74 Polymer Synthesis and Characterization 76 Structure-Activity Relationships 78 Synthesis of Fluorophore-labeled Polymers 81 Fluorescence Imaging of Bacteria Cells 82 Flow Cytometry Analysis 84 E. coli Membrane Permeabilization 86 Mechanism of Osmotic Lysis 88 Molecular Dynamics Simulation 91 Mechanism of Antimicrobial Action 94 Conclusions 95 VI. Conclusions and Future Directions 96 Structure-Activity Relationships 96 Future Directions in Polymer Chemistry 97 Mechanism of Antimicrobial Action 98 Future Mechanistic Work 99 Challenges Remaining 100 APPENDICES 101 REFERENCES 150 v

LIST OF TABLES TABLE 1-1 Structural Features and Activity of Antimicrobial Peptides and Polymers 9 II-1 Characterization of methacrylate random copolymers containing different ammonium functional groups. R = methyl 22 II-2 Characterization of methacrylate random copolymers containing different ammonium functional groups. R = butyl 23 III-l Characterization of random copolymers with dansyl end groups 39 IV-1 Characterization of methacrylamide random copolymers 57 V-l Characterization of methacrylate random copolymers containing different cationic side chain spacer groups 77 B-1 Characterization of the polymers bearing primary amines and methyl groups 117 B-2 Characterization of the polymers bearing tertiary amines and methyl groups 119 B-3 Characterization of the polymers bearing quaternary ammonium salt groups and methyl groups 120 B-4 Characterization of the polymers bearing primary amines and butyl groups 120 B-5 Characterization of the polymers bearing tertiary amines and butyl groups 121 B-6 Characterization of the polymers bearing quaternary ammonium salt groups and butyl groups 121 B-7 Characterization of the FITC-labeled polymers 132 VI

LIST OF FIGURES FIGURE 1-1 Sequence and secondary structure of magainin-2, a prototypical host defense peptide 3 1-2 Schematic drawing of the bacteria-selective membrane binding exerted by antimicrobial peptides due to a combination of electrostatic attraction and the hydrophobic effect 5 1-3 Proposed mechanisms of membrane disruption exerted by antimicrobial peptides 6 1-4 Structures of (A) peptides, (B) P-peptides, and (C) peptoids 7 1-5 Polymer disinfectants based on benzyalkonium chlorides 8 1-6 Chemical structures which possess facially amphiphilic (FA) conformations on bacterial membranes 10 1-7 Synthesis of a peptide-mimetic antimicrobial polymer 11 1-8 Peptide-mimetic antimicrobial polymers are obtained by the convergence of peptide and polymer science 13 II-1 Synthesis of amphiphilic random copolymers bearing primary, tertiary, or quaternary ammonium salts 18 II-2 Titration of amphiphilic random copolymers bearing primary, tertiary, or quaternary ammonium salts 19 II-3 Antimicrobial activities (MIC) of random copolymers 24 II-4 Hemolytic activities (HC50) of random copolymers 26 II-5 Selectivity Indices (HC50/MBC) of random copolymers 27 II-6 Bactericidal activity (MBC) as a function of pH and extent of ionization 29 II-7 Hemolysis as a function of pH 32 II-8 Hemolytic activity (HC50) as a function of pH and extent of ionization 33 III-1 Synthesis of amphiphilic random copolymers bearing primary, tertiary, or quaternary ammonium salts and dansyl end groups 39 III-2 Potentiometric titration curves for the representative copolymers 40 III-3 Partition coefficients of each polymer between octanol and aqueous phase 42 III-4 Fluorescence enhancement of dansyl-labeled polymers upon partitioning into the hydrophobic membrane environment 43 vii

III-5 Binding isotherms of polymers to POPC vesicles as a function of pH 44 III-6 Dissociation constants for polymer-POPC vesicle binding monitored by dansyl fluorescence 45 III-7 Partition coefficients for the copolymers in the presence of anionic detergent 46 III-8 Schematic of polymer-induced leakage of entrapped, self-quenching fluorophore from within liposomes 48 III-9 Sulforhodamine B (SRB) leakage from POPC liposomes induced by the copolymers as a function of pH 48 III-10 Experimental 31P chemical shift spectra of mechanically-aligned POPC bilayers with the copolymers incorporated 50 IV-1 Comparison of chemical structures for polymethacrylates and polymethacrylamides 56 IV-2 Synthesis of the amphiphilic methacrylamide random copolymers 56 IV-3 Potentiometric titration data for the cationic homopolymer 58 IV-4 Antimicrobial activities (MIC) of methacrylamide random copolymers 59 IV-5 Hemolysis dose-response curves for the methacrylamide copolymers 62 IV-6 Hemolysis as a function of hydrophobic comonomer content 63 IV-7 Cytotoxicity of the copolymers against HEp-2 cells by the XTT assay 65 IV-8 Cytotoxicity as a function of hydrophobic comonomer content 66 IV-9 Sulforhodamine B (SRB) leakage from liposomes induced by the copolymers as a function of lipid composition 68 IV-10 Kinetics of sulforhodamine B (SRB) leakage from liposomes 70 V-l Spacer arm design strategy and hypothesis 75 V-2 Generalized synthetic route and chemical structures of the amphiphilic polymethacrylate derivatives with different side chain groups connecting the amines to the polymer backbone 76 V-3 Optimization of the antimicrobial efficacy in methacrylate copolymers 78 V-4 Generalized synthetic route and chemical structures of the amphiphilic polymethacrylate derivatives with different spacer arms and a fluorophore in the polymer end group 81 V-5 Chemical structure of propidium iodide 83 V-6 Fluorescence microscopy images of fluorophore-labeled polymer incubated with live bacteria cells 83 V-7 Flow cytometry histograms of fluorophore-labeled polymer incubated with live bacteria cells 84 V-8 Two-dimensional flow cytometry analysis showing the binding and membrane damage to E. coli cells exerted by polymers 85 vin

V-9 Permeabilization of E. coli outer membrane (OM) and inner membrane (IM) induced by the polymers 87 V-10 Effect of external osmolyte diameter on the MBC of copolymers in phosphate buffered saline 89 V-ll Molecular dynamics simulation 92 B-1 H NMR spectrum of copolymer with primary amines 118 B-2 !H NMR spectrum of copolymer with tertiary amines 119 B-3 H NMR spectrum of copolymer with quaternary ammonium salts 121 B-4 !H NMR spectrum of dansyl-labeled copolymers 123 B-5 'l i NMR spectrum of copolymers with varying spacer arms 124 B-6 Base-induced isomerization of the methacrylate containing a primary amine in the side chain 125 B-7 }¥L NMR of the methacrylate monomer 125 B-8 H NMR of the isomerized methacrylamide monomer 126 B-9 !H NMR spectrum of methacrylamide copolymer 126 B-10 Calculated mole fractions of hydrophobic comonomer and degree of polymerizations based on interpretation of the 'H NMR spectra 127 B-ll MALDI-TOF-MS chromatographs of representative copolymers 129 B-12 MALDI-TOF-MS chromatographs of fluorophore-labeled copolymers 130 B-13 Normalized emission of the dansyl-labeled polymers 131 B-14 Fluorescence intensity versus concentration of the polymers in methanol 131 B-15 Absorbance and Emission spectra of the FITC-labeled polymers 132 B-16 Epifluorescence image of E. coli cells incubated with 10 |a.M F-E2 133 B-17 Epifluorescence image of E. coli cells incubated with 10 uM F-E4 134 B-18 Epifluorescence image of is. coli cells incubated with 10 uM F-E6 135 B-19 Epifluorescence image of E. coli cells incubated with 10 uM F-Ec6 136 B-20 Epifluorescence image of is. coli cells incubated with 10 uM F-B2 137 B-21 Epifluorescence image of S. aureus cells incubated with 10 uM F-E2 138 B-22 Epifluorescence image of & aureus cells incubated with 10 uM F-E4 139 B-23 Epifluorescence image of & aureus cells incubated with 10 uM F-E6 140 B-24 Epifluorescence image of & aureus cells incubated with 10 uM F-Ec6 141 B-25 Epifluorescence image of S. aureus cells incubated with 10 uM F-B2 142 B-26 Chemical structures of model copolymers 143 B-27 Snapshots of the four systems at the end of-50 ns simulations 144 B-28 Z-density profiles of various components of lipid-polymer systems averaged over last 10 ns of MD simulation 145 IX

29 The distance between center of mass of the hydrophobic and amine groups 146 30 The stretching of the polymer is measured as the distance between the ester carbons of first and last monomers 147 31 The stretching of the E6 polymer is measured as the distance between the ester carbons of first and last monomers 148 32 The radial distribution function of interactions of spacer groups of polymers with charged groups of lipid bilayers 149 x

LIST OF APPENDICES APPENDIX A. Materials and Methods 102 Materials 102 Monomer Synthesis 103 Polymer Synthesis 104 Potentiometric Titration 106 Water-Octanol Partition Coefficients 107 Antimicrobial Activity Assays 108 Hemolytic Activity Assays 110 Liposome-Polymer Binding 111 Liposome Dye Leakage 112 Cell Culture and XTT Assay 113 Fluorescence Imaging 114 Flow Cytometry Analysis 114 E. coli Membrane Permeabilization 115 Molecular Dynamics Simulation 115 B. Polymer Characterization Characterization of Copolymers by !H NMR Analysis 117 Characterization of Copolymers by MALDI-TOF-MS 129 Absorbance and Emission of Dye-Labeled Polymers 131 Supplemental Fluorescence Images 133 Supplemental Molecular Dynamics Simulation Data 143 XI

LIST OF ABBREVIATIONS AEMA AIBN AMP BMA CL CMC CTAB DDP DMAEMA DOPG DP EMA FITC HC5o HDP IC50 Lysl-DOPG logP MALDI-TOF-MS MBC MIC MMA MMP MW NMR PBS PEG PI POPC POPE POPG QAS RBC SRB TFA aminoethylmethacrylate azo-bis-isobutyronitrile Antimicrobial Peptide butylmethacrylate cardiolipin Critical Micelle Concentration cetyltrimethylammonium bromide dodecylphosphate N,N-dimethylaminoethylmethacrylate l,2-dioleoyl-s«-glycero-3-[phospho-rac-(l-glycerol)] degree of polymerization ethylmethacrylate fluorescein isothiocyanate Hemolytic Concentration Host Defense Peptide Cytotoxic Concentration Dissociation constant l,2-dioleoyl-s«-glycero-3-[phospho-rac-(3-lysyl(l-glycerol))] water-octanol partition coefficient Matrix-Assisted Laser Desorption Ionization Time of Flight Mass Spectroscopy Minimum Bactericidal Concentration Minimum Inhibitory Concentration methylmethacrylate methylmercaptopropionate molecular weight Nuclear Magnetic Resonance phosphate buffered saline polyethylene glycol propidium iodide l-palmitoyl-2-oleoyl-s«-glycero-3-phosphocholine l-palmitoyl-2-oleoyl-OT-glycero-3-phosphoethanolamine l-palmitoyl-2-oleoyl-sft-glycero-3-phospho-(r-s_y«-glycerol) quaternary ammonium salt Red Blood Cell sulforhodamine B trifluoroacetic acid xn

CHAPTER I Introduction and Background: The Convergence of Peptide and Polymer Science toward Novel Antibiotics Motivation and Objectives Since the serendipitous discovery of penicillin by Alexander Fleming in 1928,l antibiotic drugs have been the standard in treatment for infectious disease, saving countless lives worldwide. Unfortunately, widespread use has led to evolution of microbial strains that resist the action of antibiotics. ' Notable examples of antibiotic- resistant bacteria that currently pose a significant threat to public health include methicillin-resistant Staphylococcus aureus (MRSA), extensively drug-resistant tuberculosis (XDR-TB), ' and vancomycin-resistant Enterococci (VRE). A steady supply of new antibiotics, which work by novel mechanisms of action, is therefore requisite to keep pace with the constant proliferation of antibiotic-resistant bacteria. However, an alarming trend has recently emerged: while resistance continues to spread prodigiously, the number of new antibiotic drugs approvals has been declining.8 As a result, there is currently an urgent need for new anti-infective strategies, which are not susceptible to the existing mechanisms of resistance. The chief objective is to obtain compounds that would fulfill the following design objectives: • Potent activity against a broad spectrum of bacterial strains, to ensure applicability against a wide variety of infections • Rapid bactericidal kinetics, to avoid loss of activity by degradation • Little or no toxicity to human cells, to minimize side effects during treatment • Reduced likelihood of inducing bacterial resistance, to ensure that the drugs will remain useful for an extended period of time • Reasonable manufacturing cost, to ensure affordability 1

Novel compounds which fulfill the above criteria are expected to advance the field of antimicrobials by circumventing the current resistance mechanisms, and will thereby aide in the treatment of infectious disease. This dissertation focuses on using new design principles that mimic the antibacterial function of natural peptides to optimize the activity of polymers as antimicrobial agents which fulfill the criteria listed above. The central goal of this bio-mimetic design approach is to capture the specific structural features of the peptides which control their activity using synthetic polymers, rather than the conventional approach of exactly copying their chemical structure. Host Defense Peptides Host defense peptides (HDPs) are a class of compounds found in nature that function as a part of the innate immune system to defend the host against invasion by harmful bacteria.9 These peptides primarily act by direct killing of bacteria cells, and by induction of immune response.10 Hence, the term "host-defense" is derived from their protective function in the body. Focusing on the ability of these natural peptides to directly kill bacteria, they are often referred to as "antimicrobial peptides (AMP)" in the literature.11 Because the AMPs kill bacteria without harming host cells, they are particularly attractive as potential new antibiotics. The peptides can kill a broad spectrum of bacterial strains, exert rapid bactericidal kinetics, are non-toxic to host cells. Additionally, they are less prone to resistance by bacterial cells due to their mechanism of action, which putatively relies on membrane disruption.12 Since the 1950's, it has been known that normal tissues and secretions possess the ability to inhibit the growth of microbes and that these substances play a role in the innate immunity of host cells.13 The identity of the active antimicrobial compounds in these tissues remained elusive for many decades. In 1987, Zasloff discovered one of the earliest known HDPs in the skin of the African clawed frog Xenopus laevis.14 Xenopus does not show signs of infection in microbially contaminated pond water, even after undergoing surgery without antibiotics, which prompted Zasloff to search for an active antimicrobial compound in their skin. He isolated two short peptides from the frog skin secretions and determined their sequences, which are rich in hydrophobic and cationic residues. The peptides displayed potent activity against a panel of fourteen gram-positive 2

and gram-negative bacteria, including E. coli and S. aureus. They were named magainin- 1 and -2 (from the Hebrew magain, meaning "shield"). The sequence and secondary structure of magainin-2 are shown in Figure 1-1. Hydrophilic Jt V ^ > £ / Hydrophobic J 7 « ^ T ^ ^ ^ ^ Magainin-2 sequence: GIGKFLHSAKKFGKAFVGEIMNS Figure 1-1. Sequence and secondary structure of magainin-2, a prototypical host defense peptide. Other early pioneers in the HDP field include Boman and Lehrer, who independently isolated antimicrobial cepropins from insects15 and defensins from humans,1 '17 respectively. Since these early discoveries, it has been shown that HDPs are present in all multi-cellular organisms,9 as a part of immune system to combat invading pathogens.14 Since their discovery, HDPs have attracted intense scientific and commercial interest.18'19 For example, omiganan (MBI-226) is a cationic, amphiphilic peptide isolated from bovine neutrophils which recently passed Phase III clinical trial for preventing catheter-related infections.20 This demonstrates the potential for host defense peptides to serve as novel antibiotic drugs. Host defense peptides are characterized by extraordinary diversity in origin, primary sequence, and secondary structure.9 Despite the lack of homology, there appear to be certain features which are common to a vast majority of antimicrobial peptides.18 Wang and co-workers have compiled an exhaustive database of more than seventeen hundred known antimicrobial peptides (AMPs).21'22 The most up-to-date version of this database can be accessed at the following URL: http://aps.unmc.edu/AP/main.php. Analysis of the AMP database reveals that the vast majority of antimicrobial peptides possess the following characteristics in terms of their amino acid sequences: • Net cationic charge at neutral pH (average +3.8) • Amphiphilicity (average 44% non-polar residues) • Low molecular weight (average 30.6 amino acids) These properties are thought to play a key role in the mechanism of antimicrobial action exerted by the peptides. Determining the mechanism by which HDPs exert their 3

bactericidal effects has been a main focus in the field. ' Once the active peptides were isolated, it was found that the all-D and all-L enantiomeric forms express similar antimicrobial activity, which indicates the absence of a specific receptor-mediated interaction (not a "lock-and-key" type mechanism).25 Considering their amphiphilic nature, Matsuzaki and others hypothesized that the mechanism of antimicrobial action exerted by HDPs involves disruption of the bacterial cell membrane.26'27 Even before the discovery of magainins, it was proposed that the antimicrobial components in normal tissue act by damaging bacterial membranes.28'29 Indeed, magainin-2 induces leakage of entrapped contents from within model lipid vesicles, which mimic the phospholipid composition of bacterial cell membranes.26 A large body of work has been directed at analysis of the details of this membrane i n -in disruption event. " Magainin-2, an extensively studied prototypical HDP, exists as a random coil in solution. Because bacterial membranes display a greater density of anionic lipids on their membranes, relative to mammalian cells, electrostatic attraction favors selective binding of the cationic peptides to bacteria versus mammalian cells (Figure 1-2). Additionally, mammalian cells contain cholesterol, which increases the rigidity of the membranes and are therefore believed resist the insertion by antimicrobial peptides.38 Upon binding to bacterial cell membranes, it adopts an a-helical structure, in which cationic residues are displayed on one "face" of the helix, while the non-polar side chains are on the opposing face.38'39 This motif is known as "facial amphiphilicity" and has been observed in a variety of a-helical antimicrobial peptides.40"43 The helix formation of peptides is induced by the complexation of cationic groups with lipid headgroups, whereas the hydrophobic face of the helix is localized in the non-polar membrane core (Figure I-3).42'44 Other antimicrobial peptides form P-sheet structures, in which cationic and hydrophobic residues are displayed on opposite sides, and the backbone is stabilized by disulfide bonds.45 In addition, certain peptides can form irregular conformations (lacking the helical or sheet motif) which nevertheless segregate cationic and hydrophobic residues into different domains.46 This suggests that the facially amphiphilic nature, rather than specific structural elements such the a-helix, is the chief determinant of antimicrobial activity. 4

Antimicrobial Peptide Hydrophobic V @\ Electrostatic and hydrophobic Membrane Phospholipids o -. a H O" I ^ 0 0,0 0 0 0 0 O0OO o o o Mammalian cell membrane 0OO0OOOG OO0OO0OO Bacterial cell membrane POPC, Zwitterionic (mammalian cells) <=> cholesterol ~DCO0 Zwitterionic lipids •^^^O Zwitterionic lipids w w © anionic lipid POPG, anionic (bacteria cells) 0AH 0 POPE, Zwitterionic (bacteria cells) NH3* Figure 1-2. Schematic drawing of the bacteria-selective membrane binding exerted by antimicrobial peptides due to a combination of electrostatic attraction and the hydrophobic effect. Chemical structures of phospholipids abundant in mammalian and bacterial cell membranes are shown on the right. Subsequent to initial binding, the peptides are believed to disrupt the bacterial cell membranes. There are several mechanisms proposed to explain the observed membrane disruption events (Figure 1-3), which are the subject of ongoing debate in the literature.23 Accumulation of peptides on the outer surface causes an imbalance in the chemical potential between the proximal and distal membrane leaflets (i.e. the lipid monolayers presented on the outer and inner faces of the membrane bilayers, respectively). This imbalance may be relieved by disrupting the barrier function of the membrane by the formation of pores or other permeable defects in the membrane.47 The chemical imbalance moreover causes translocation of the peptides across the membranes, resulting in transient pores.48 The pore life time is dependent on the lipids, and estimated to be -40 Usee.49 The pores are then closed after the translocation of peptides into the cytoplasm.50 There are several molecular-level pore structures proposed.35 For the leakage of aqueous components from cytosol, the interior of the pore must be hydrophilic. The pore interior may be lined by the hydrophilic groups of the peptides and phospholipid headgroups reoriented to form a "toroidal pore" across the membrane (Figure 1-3). Another possibility is the formation of "barrel stave" type pores, which are lined by peptides alone 5

without lipid reorientation.37'45'51 Finally, peptides may accumulate on the outer surface of the membranes and completely disrupt the lipids via the "carpet model".35 For large a quantity of peptide bound to the membrane, solubilization of the membranes could be dominant (Figure 1-3). It is generally accepted that the increased membrane permeability leads to loss of the transmembrane potential, leakage of cytoplasmic contents or osmotic lysis, and ultimately cell death. Solubilization oooooooooooeo / i \ Toroidal pore Barrel-stave Carpet 3OOOO0O V V OOC Translocation ooooo Figure 1-3. Proposed mechanisms of membrane disruption exerted by antimicrobial peptides While HDPs are very promising candidates in pharmaceutical applications, several obstacles continue to plague the realistic outlook for their widespread use as drugs.52 The methods required to synthesize sequence specific peptides are often cost- and labor-intensive relative to traditional antibiotics.52 Furthermore, HDPs being considered for clinical trials have encountered obstacles including proteolysis, reduced activity in vivo, and poor pharmacokinetics.52 Proteolysis, the degradation of peptide chains by enzymes in the body, reduces the number of active compounds available to kill 6

bacteria. Pharamcokinetics, the time-dependent distribution of a drug in the body after administration, are hampered by removal of the peptides from the site of infection prior to killing the bacteria. Activity can be further hindered in vivo due to screening of the electrostatic attraction by salts or inactivation of the peptides by serum. It has been hypothesized that the antimicrobial mechanism of HDPs can be generalized to non-natural (synthetic) peptides, because their function is controlled by facial amphiphilicity and cationic charge, rather than highly specific receptor-mediated interactions. Synthetic a-helical peptides which can adapt facially amphiphilic structures, such as the magainin derivative MSI-78, have shown superior antimicrobial activity relative to the natural peptide magainin-2.12' 53 In addition to the peptides based on naturally-occurring a-amino acids (Figure 1-4, structure A), peptidomimetics have been also examined. For example, P-peptides consisting of P-amino acids (Figure 1-4, structure B), have been shown to kill bacteria without harming human cells.54'55 Peptoids are another class of peptidomimetics in which the side chains are attached to the amide nitrogen atom (Figure 1-4, structure C). Peptoids lack a chiral center in the backbone, but the side chains can be designed to induce facially amphiphilic conformations upon membrane binding.43 R H H R A B C Figure 1-4. Structures of (A) peptides, (B) p-peptides, and (C) peptoids Accordingly, P-peptides56 and peptoids43 were shown to kill bacteria without harming human cells, by mimicking the facially amphiphilic helical structure of host defense peptides. These non-natural platforms afford additional advantages such as resistance to proteolysis (degradation by enzymes in the body). Hence, researchers can design the peptides for pharmaceutical applications and indeed many peptide antibiotics such as Omiganan are currently in clinical trails.11'18'57 An alternative approach is to utilize synthetic polymers which are not easily recognized by enzymes in the body, thus potentially solving the problem of degradation encountered in the case of peptide drugs. It is quite possible that synthetic polymers, with 7

access to a nearly limitless range of chemical structures, will be particularly useful in circumventing some of the obstacles which complicate the application of membrane- active antimicrobial peptides. Synthetic Polymer Biocides In household cleaning products, benzalkonium chloride surfactant compounds have been used as the active disinfectant since the 1950's.58'59 The chemical structure of these surfactants features a quaternary ammonium salt (QAS) unit attached to a long alkyl chain. In 1983, flceda and co-workers reasoned that the activity of these surfactants might be enhanced by incorporation into the side chains of a synthetic polymer, in which the active compound could be locally concentrated into the many repeating units of the polymer chains.60 This was readily achieved by introducing a vinyl moiety onto the benzyl group of the surfactant and then polymerizing the obtained styrene by conventional free radical procedures (Figure 1-5). Figure 1-5. Polymer disinfectants based on benzyalkonium chlorides, which are surfactants used in household cleaning products. The polymerization of this styrene derivative is a standard thermally initiated free radical process. Figure adopted from Ikeda et al (1983). These QAS-functionalized polystyrenes were shown to effectively kill bacteria, with potencies dependent on the alkyl side chain length and molecular weight. The field of disinfecting polymers has expanded significantly, with a wide variety of polymer platforms including quaternized vinyl pyridines, vinyl alcohols, and methacrylate derivatives.61' 62 It has been asserted that the mechanism of action employed by these polymers involves damage of bacterial cell membranes, causing leakage of cytoplasmic contents or complete lysis of the cells. In contrast to host defense peptides, these surfactant molecules are not expected to adapt facially amphiphilic conformations or to form discrete pores in the membranes.63 The application of these materials involves 8

immobilization of the polymer chains on surfaces (Figure 1-8, structure G) for disinfecting coatings.6 ' These polymer biocides are not commonly viewed as potential pharmaceuticals, however, because their surfactant characteristic would likely induce toxicity to human cells as well as combating bacteria. There is no evidence in the early literature of any conceptual connection between the discovery of host defense peptides (HDPs) and the initial development of polymer disinfectant materials to the author's knowledge. Such a connection is now emerging because, curiously, the polymers appear to have certain features in common with HDPs. ' In addition, the membrane-disrupting mode of action exerted by the polysurfactants seems analogous to peptide-induced permeabilization, although the former are not cell- type selective. Indeed, despite some similarities, several key differences between host defense peptides and polymer surfactants should be highlighted (Table 1-1). Table 1-1. Structural Features and Activity of Antimicrobial Peptides and Polymers Host Defense Peptides Polymer Biocides Peptide- mimetic Polymers Molecular weight 2-5 kDa 10-100 kDa 1-10 kDa Cationic Groups Primary amines (K), guanidines (R) Quaternary ammonium (QAS) Primary amines Hydrophobic Groups Non-polar amino acids (A,V,L, etc.) Long alkyl chains (Cfi-Ciz) Short alkyl chains (CrC4) Activity Antimicrobial, non-toxic Biocidal Antimicrobial, lower toxicity It would be ideal to identify polymers which selectively kill of bacteria cells, without harming human cells, toward novel antibacterial drugs. Polymers which capture the essential features of HDPs are expected to show similar biological activity, because their activity is modulated by physiochemical determinants. Hence, we focus on designing polymers to mimic the salient structural features of HDPs. Peptide-mimetic Antimicrobial Polymers Mimicry of host defense peptides using synthetic P-peptides and peptoids has demonstrated that the antibacterial properties typical of HDPs can also be achieved by 9

synthetic methods without the need to isolate them from natural sources. ' ' Entirely non-biological molecules can also exert a membrane-disrupting antimicrobial mechanism.69'70 Small, rigid oligomers displaying cationic and hydrophobic groups were among the first examples of entirely abiotic compounds which effectively mimic HDP activity (Figure 1-8, structure B).69'71 The conformation of these molecules are fixed by rigid arylamide structures stabilized by hydrogen bonding between amide nitrogen atoms and thioether groups in neighboring side chains. The oligomers display their cationic and hydrophobic components on opposite sides of the molecules. This design is intended to reproduce the facial amphiphilicity of HDPs by the simplest chemical structure. Remarkably, an aryl amide oligomer based on the concept of facial amphiphilicity has already passed a phase I clinical trial as an intravenous antibiotic drug to combat MRSA.72 Induced a- pjxec| p/\ Induced helix FA irregular FA Figure 1-6. Chemical structures which possess facially amphiphilic (FA) conformations on bacterial membranes, via induction of the a-helix, fixed rigid structures, and irregular conformations. Synthetic polymers, including polymethacrylates, nylons and polynorbornenes, are another platform with great promise for mimicry of HDPs.66'67 Unlike disinfecting polymers (surfactant-based biocides), the HDP-mimetic design approach has the potential to endow excellent antimicrobial potency without harming human cells. Although facially amphiphilic secondary structures cannot be readily programmed into synthetic polymers by current methods, a sufficiently flexible polymer chain may adapt irregular but facially amphiphilic conformation upon binding at the membrane-water interface (Figure 1-6). This can be achieved by random copolymers with flexible backbones, such as nylons or polymethacrylates, in which the side chains display pendant cationic groups 10

Full document contains 172 pages
Abstract: The proliferation of antibiotic-resistant bacteria and the decline in new antibiotic drug approvals have lead to an emergent need for novel antibacterial strategies. In nature, host defense peptides (HDPs) kill invading pathogens as part of the innate immune system in multicelluar organisms. HDPs are diverse in sequence and conformation, but certain physiochemical features are conserved: they are cationic (+1-5 net charge), amphiphilic (30-50% hydrophobic residues), and low molecular weight (10-50 residues). Their putative mechanism involves the disruption of bacterial membranes. Because the activity of these peptides is modulated by physiochemical characteristics, rather than receptor-specific interactions, we hypothesized that synthetic macromolecules could mimic their function. The peptide-mimetic design strategy has several advantages including diminished cost of manufacturing, lower susceptibility to degradation, and access to the wide range of chemical functionalities obtained by synthetic means. Random copolymers composed of ∼50% cationic (aminoethylmethacrylate) and ∼50% hydrophobic (methyl methacrylate) monomer units, and molecular weights of 1-4 kDa, inhibited bacterial growth at concentrations as low as ∼10μM whereas they did not lyse human red blood cells up to ∼1000µM. Polymers bearing primary amine groups in the side chains displayed more potent activity than analogous polymers containing quaternary ammonium salts, attributed the higher affinity of primary amines for phosphate lipid head groups, leading to efficient membrane permeabilization. Polymethacrylamides showed low µM antibacterial activity without harming RBCs, although they were cytotoxic to human epithelial cells. After further optimization, copolymers bearing aminobutyl side chains showed highest activity among other polymers bearing shorter (aminoethyl) and longer (aminohexyl) cationic side chains. Biophysical experiments using model lipid membranes, fluorescence imaging, flow cytometry, and molecular dynamics simulations indicated that the mechanism of antimicrobial action involves membrane disruption. Pore formation in cell membranes followed by osmotic lysis and leakage of cytoplasmic contents, results in bacterial cell death. Although synthetic polymers are heterogeneous in structure (polydispersity, random sequence, and heterotacticity) and conformation, they nevertheless effectively mimic the antimicrobial activity and mechanism of host defense peptides. Hence, this emerging class of materials can be excellent candidates for a wide variety of applications in combating the rise of drug-resistant infectious diseases.