UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structure/function and mode of action of antimicrobial Fong, Carol L. Friedrich 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-714616.pdf [ 6.3MB ]
Metadata
JSON: 831-1.0104989.json
JSON-LD: 831-1.0104989-ld.json
RDF/XML (Pretty): 831-1.0104989-rdf.xml
RDF/JSON: 831-1.0104989-rdf.json
Turtle: 831-1.0104989-turtle.txt
N-Triples: 831-1.0104989-rdf-ntriples.txt
Original Record: 831-1.0104989-source.json
Full Text
831-1.0104989-fulltext.txt
Citation
831-1.0104989.ris

Full Text

STRUCTURE/FUNCTION AND MODE OF ACTION OF ANTIMICROBIAL CATIONIC PEPTIDES ON GRAM POSITIVE BACTERIA By C A R O L L . FRIEDRICH F O N G B.Sc. (honours microbiology), The University of Alberta, 1995  A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Microbiology and Immunology)  We accept this thesis as conforming to !he required standard  THE UNIVERSITY OF B R m S H  COUMBIA  June, 2001 © Carol L. Friedrich Fong, 2001  In  presenting this  degree  at the  thesis  in  University of  partial  fulfilment  of  of  department  this thesis for or  by  his  or  requirements  British Columbia, I agree that the  freely available for reference and study. I further copying  the  representatives.  It  publication of this thesis for financial gain shall not  is  of  ffj  QI&CJKJ  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  Library shall make it  by the  understood  that  be allowed without  permission.  Department  an advanced  agree that permission for extensive  scholarly purposes may be granted her  for  fXrf\m U-n.oloQ I  head of copying  my or  my written  ABSTRACT Antimicrobial cationic peptides are ubiquitous in nature and are thought to be a component of the first line of defense against infectious agents. It is important to study these peptides in order to use them as potential templates for new antibiotic therapeutics. The aim of this study was to determine the structure/activity relationships of selected peptides and to determine the mode of action of these peptides on Gram positive bacteria. Studies with model membrane systems using circular dichroism and fluorescence spectroscopy indicated that these active peptides were induced into a more defined structure upon binding to lipids and detergents. In general, peptide interaction with lipids and detergents were similar.  The peptides entered a hydrophobic environment, with  tryptophan residues inaccessible to the aqueous solution.  Although there were  differences between peptides in specific interactions, no correlation could be made between lipid interaction and antimicrobial specificity or activity. The activity of cationic peptides from different structural classes was determined on various Gram positive strains and the killing kinetics of these peptides were very similar at 10-fold the MIC. Electron microscopy of S. aureus and S. epidermidis treated with the peptides at 10-fold the MIC showed variability in effects on bacterial structure. Mesosome-like structures were seen to develop in S. aureus with all peptides, whereas different effects, including nuclear condensation, were seen in the case of S. epidermidis. The membrane-potential-sensitive fluorophore DiSC3(5) was utilized to assess the interaction of antimicrobial peptides with the cytoplasmic membrane of S. aureus. Studies of the kinetics of killing and membrane depolarization showed that no correlation could be made between cytoplasmic membrane depolarization and peptide activity.  ii  Thus, although cytoplasmic membrane permeabilization was a widespread ability among peptides, it did not appear to be the killing mechanism. Macromolecular  synthesis  assays  showed  that  all  peptides  studied  had  intracellular effects, and these were often seen at sub-lethal peptide concentrations. The peptides differed in their specific effects on macromolecular synthesis.  In general,  evidence presented here suggested that cytoplasmic membrane disruption is not the sole mechanism of action against bacteria, and that multiple targets may be involved. specific mechanism differs between peptides of different structural classes.  iii  The  TABLE OF CONTENTS ABSTRACT  II  T A B L E OF CONTENTS  IV  LIST OF TABLES  VI  LIST OF FIGURES  VII  LIST OF ABBREVIATIONS  X  ACKNOWLEDGEMENTS  XII  INTRODUCTION A. B. C. D. E. F. G. H.  1  Antimicrobial Cationic Peptides (Sources and General Characteristics) Structure/Function Studies with Cationic Peptides Mode of Action Studies Peptides Used in this Study Methods of Determining Antimicrobial Activity Effects of Salt on Antimicrobial Activity Therapeutic Potential and Peptide Development Aims of this Study  MATERIALS AND METHODS A. B. C. D.  28  Bacterial Strains and Growth Conditions Chemicals Liposome preparation Structural Studies - Spectroscopy  28 30 30 31  D.l. Fluorescence Spectroscopy D. 2. Circular Dichroism Measurements  31 32  E. Antimicrobial Activity  32  E. l . Minimal Inhibitory Concentration E. 2.Bacterial Killing Assays  32 33  F. Electron Microscopy  33  FA. Transmission Electron Microscopy F. 2. Scanning electron microscopy  33 34  G. Cytoplasmic Membrane Depolarization Assay H. Macromolecular Synthesis assay and Bacterial Killing Assays RESULTS  35 36 37  C H A P T E R O N E : ANTIMICROBIAL ACTIVITY AND STRUCTURE-ACTIVITY RELATIONSHIPS  A. B. C. D. E.  J 4 7 16 21 23 24 25  Introduction Verification of MIC Method MICs of Peptides on Gram Positive Bacteria Effect of Media and Cations on MICs Bacterial Killing Assays E. 1. Killing of Log Phase Bacteria E.2. Killing Assays in the Presence of CCCP  37  37 39 41 42 45 45 48  F. MICs of lndolicidin Variants G. Summary  55 58  C H A P T E R T W O : INTERACTION AND BINDING OF PEPTIDES WITH LIPIDIC SYSTEMS  A. Introduction B. Circular Dichroism Spectroscopy  60  60 61  B.l. Suspected a-helical Peptides and Variants  61  iv  B.2. Comparison of Three CD spectroscopy Analysis Programs B.3. Indolicidin and Variants B. 4. CD Scans with LPS and LTA  C. Fluorescence Spectroscopy  79  C. 1. Fluorescence Spectroscopy in the Presence of Liposomes and Detergents C.2. Potassium Iodide Quenching C.3. Spectroscopy with Spin-Labeled Lipids  D. Summary  C H A P T E R T H R E E : M E C H A N I S M OF A C T I O N O N G R A M POSITIVE B A C T E R I A  A. Introduction B. Electron Microscopy  93 96  98 100  C. Cytoplasmic Membrane Depolarization  107  C. l. Development of DiSC (5) Assay C.2. Depolarization by Various Peptides C. 3. Cell Viability Assay in Conjunction with the Depolarization Assay 3  D. Macromolecular Synthesis Studies  107 108 112  114  Characterization of the Action of Peptides on the Auxotroph S. aureus ISP 67 Macromolecular Synthesis Inhibition by Indolicidin and CPI 1CN Macromolecular Synthesis Inhibition by CPI OA Macromolecular Synthesis Inhibition by CP29 Macromolecular Synthesis Inhibition by Bac2A-NH 2  E. Summary  114 117 118 122 122  125  DISCUSSION  126  A. Overview B. Antimicrobial Activity of Various Peptides C. Peptide Design and Structure-Activity Relationships D. Interaction with Model Membranes and Structure-Activity Relationships E. Antimicrobial Mechanism against Gram Positive Bacteria E. 1. E.2. E.3. E.4.  79 83 87  96 98  B.l. Transmission Electron Microscopy (TEM) B. 2. Scanning Electron Microscopy (SEM)  D. l . D.2. D.3. D.4. D. 5.  63 64 74  Electron Microscopy Cytoplasmic Membrane Depolarization Macromolecular Synthesis Inhibition Conclusion  126 127 129 131 139 139 141 143 144  F. Potential Applications of Peptides G. Future Studies  145 147  REFERENCES  149  v  LIST OF TABLES Table 1: Amino acid sequences of antimicrobial cationic peptides used in this study....27 Table 2: Bacterial Strains  29  Table 3: MICs of peptides in different holding solutions and media  39  Table 4: MICs of peptides using the agarose method and the broth dilution method  41  Table 5: Broth dilution MICs of peptides for various Gram positive bacteria  43  Table 6: MICs of peptides for E. coli and S aureus in the presence of salts  45  Table 7: MICs of indolicidin variants on various Gram negative and Gram positive bacteria  56  Table 8: Alpha-helicity of peptides in various environments  63  Table 9: Alpha-helical content for C E M A , as predicted by three analysis programs  63  Table 10: Fluorescence emission wavelength maximum of peptides in the presence of various liposomes 80 Table 11: MICs of peptides against S. aureus wild-type and auxotroph (ISP67)  114  Table 12: Summary of effects on peptide-treated S. epidermidis, as observed by electron microscopy 147 Table 13: Summary of the antimicrobial activity, membrane depolarization and macromolecular synthesis inhibition of different peptides on S. aureus ISP 67  VI  148  LIST OF FIGURES Figure 1: Killing of S. aureus by peptides in LB broth  47  Figure 2: Killing of S. epidermidis by peptides in LB broth  49  Figure 3: Killing of E.faecalis by peptides in LB broth  50  Figure 4: Killing of S. aureus by indolicidin in the absence and presence of C C C P  51  Figure 5: Killing of S. aureus by CPI 1CN in the absence and presence of C C C P  52  Figure 6: Killing of S. aureus by CP10A in the absence and presence of CCCP  53  Figure 7: Killing of S. aureus by CP29 in the absence and presence of CCCP  54  Figure 8: C D spectra of CP201, CP208, C E M A , C E M E , CP26 and CP29 in the presence of liposomes 65 Figure 9: C D spectra of indolicidin and CP11CN in the presence of liposomes  69  Figure 10: C D spectra of indolicidin in POPC, POPG, mixed liposomes and lyso-PC....70 Figure 11: C D spectra of indolicidin in buffer, SDS and lyso-PC  71  Figure 12: C D spectra of CPI 1CN in POPC, lyso-PC, POPG and mixed liposomes  72  Figure 13: C D spectra of CPI OA in buffer, liposomes and DPC  73  Figure 14: C D spectra of CP10A in buffer and lipoteichoic acid  76  Figure 15: C D spectra of indolicidin, CPI 1CN and Bac2A-NH in buffer and lipoteichoic acid 77 2  Figure 16: C D spectra of CP26 and CP29 in buffer and lipoteichoic acid  78  Figure 17: Fluorescence spectra of indolicidin in the presence of buffer and liposomes..82 Figure 18: Fluorescence scans of indolicidin in the presence of detergent micelles and buffer 84 Figure 19: Fluorescence scans of CPI 1CN in the presence of detergent micelles, liposomes and buffer  85  Figure 20: Fluorescence scans of CPI OA in the presence of detergent (DPC) micelles, liposomes and buffer 86  vii  Figure 21: Stern-Volmer plot. KI quenching of CP26 fluorescence in buffer and liposomes  88  Figure 22: Stern-Volmer plot. KI quenching of C E M A fluorescence in buffer and liposomes  89  Figure 23: Stern-Volmer plot. KI quenching of CPI 1CN fluorescence in buffer and liposomes  90  Figure 24: Stern-Volmer plot. KI quenching of CPI OA fluorescence in buffer and liposomes  91  Figure 25: Stern-Volmer plot. KI quenching of indolicidin fluorescence in buffer and liposomes and lyso-PC  92  Figure 26: Relative fluorescence intensities of peptides in the presence of spin labeled liposomes 95 Figure 27: Electron micrographs of untreated, CPI lCN-treated, CP29-treated and Bac2A-NH2-treated S. aureus  101  Figure 28: Electron micrographs of CPI OA-treated and indolicidin-treated S. aureus...102 Figure 29: Electron micrographs of untreated, CPI lCN-treated and CP29-treated S.  epidermidis  103  Figure 30: Electron micrographs of Bac2A-NH -treated, CPlOA-treated and CP26treated S. epidermidis  104  Figure 31: Scanning electron micrographs of untreated, CPI lCN-treated and CP29treated S. epidermidis  105  2  Figure 32: Electron micrographs of untreated, CPI lCN-treated and Bac2A-NH -treated 2  E.faecalis  106  Figure 33: Depolarization of the cytoplasmic membrane of S. aureus by valinomycin, indicated by maximum fluorescence reached, as a function of KC1 concentration 109 Figure 34: Depolarization of the cytoplasmic membrane of S. aureus as a function of peptide concentration Ill Figure 35: Depolarization of the cytoplasmic membrane of S. aureus and the levels of survival (in CFU/ml) of bacteria under the dye assay conditions 113  viii  Figure 36: Depolarization of the cytoplasmic membrane of S. aureus ISP 67 as a function of peptide concentration 116 Figure 37: Effect of indolicidin on H-labelled thymidine, uridine, histidine incorporation into S. aureus ISP67 macromolecules and percent survivors under identical conditions 119 3  Figure 38: Effect of CPI 1CN on H-labelled thymidine, uridine, histidine incorporation into S. aureus ISP67 macromolecules and percent survivors under identical conditions 120 3  Figure 39: Effect of CP10A on H-labelled thymidine, uridine, histidine incorporation into S. aureus ISP67 macromolecules and percent survivors under identical conditions 121 3  Figure 40: Effect of CP29 on H-labelled thymidine, uridine, histidine incorporation into S. aureus ISP67 macromolecules and percent survivors under identical conditions 123 Figure 41: Effect of Bac2A-NH2 on H-labelled thymidine, uridine, histidine incorporation into S. aureus ISP67 macromolecules and percent survivors under identical conditions 124 Figure 42: Structures of indolicidin, CPI 1CN and CP10A determined by 2D-NMR in dodecylphophocholine (determined by Dr. Annett Rozek, UBC) 138  ix  LIST OF ABBREVIATIONS A T C C : American Type Culture Collection BSA: bovine serum albumin CCCP: carbonyl cyanide-m-chlorophenyl hydrazone CD: Circular Dichroism C E M E : cecropin/melittin hybrid peptide CFU: colony forming unit DiSC3(5): 3, 3-dipropylthiacarbocyanine DPC: dodecylphosphocholine HEPES: (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid) sodium salt LB: Luria Bertani Broth LPS: lipopolysaccharide LTA: lipoteichoic acid Lyso-PC: lysophophocholine M H : Mueller-Hinton broth MIC: minimum inhibitory concentration MRSA: methicillin-resistant Staphylococcus aureus NMR: nuclear magnetic resonance OD: optical density POPC:  l-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine  POPG: l-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol NAPS: Nucleic Acid/Protein Service Unit SDS: sodium dodecyl sulfate  X  SEM: scanning electron microscopy TCA: trichloroacetic acid T E M : transmission electron microscopy TFE: trifluoroethanol tempo-PC: 1,2-dioleoyl-sn-glycero-3-phosphotempocholine 5-doxyl-PC: l-palmitoyl-2-stearoyl-(5-doxyl)-sn-glycero-3-phosphocholine 12-doxyl-PC: 1 -palmitoyl-2-stearoyl-( 12-doxyl)-sn-glycero-3-phosphocholine  XI  ACKNOWLEDGEMENTS I thank Bob Hancock, my supervisor for his guidance and encouragement, as well as my committee members Dr. Fernandez, Dr. Warren and Dr. Chow for their support and suggestions. I thank members of the Hancock lab, both past and present, for help and support. Specifically I would like to thank Susan Farmer, for her invaluable advice and discussion as well as her help with electron microscopy. I thank Aleks Patryzkat both for help during the macromolecular synthesis assays and helpful discussion, Vally Mendoza for technical help during the summer of 1999, and Annett Rozek for collaboration and discussion on the structural studies. I thank Terry Beveridge and Dianne Moyles at the University of Guelph for training me and collaborating with me on some electron microscopy studies. The financial support of the Canadian Bacterial Diseases Network was greatly appreciated. Finally, I would like to thank my parents and my siblings for their unwavering support and encouragement throughout all those years at university. I learned a lot from my siblings Dr. Glenn, Dr. Loran and Dr. Erena, who all suffered through the realities of graduate school while simultaneously setting such high standards and thus gaining my admiration. I thank my friends, both the Edmontonians and the Vancouverites for the laughs and gripe sessions that kept me going. Most importantly, I thank my husband Fenton for his constant love, support and understanding during the good times and the bad times through the course of my Ph.D. studies (it also helped that he was a grad student as well...).  Without him there is no doubt that this thesis would  not have been made a reality.  xii  INTRODUCTION A. Antimicrobial Cationic Peptides (Sources and General Characteristics) Since the rapid emergence of antibiotic-resistant bacteria, there has been more effort put into the research and discovery of new antimicrobials. Much emphasis has been put into developing analogues of earlier antibiotic classes.  Unfortunately, these  analogues are useful only temporarily, as existing resistance mechanisms quickly adapt to these new compounds (Chopra, 1998). As a result, much hope is placed on new agents such as the oxazolidinones and antimicrobial cationic peptides.  Antimicrobial cationic  peptides are part of a growing number of proteins and peptides that have been discovered with antimicrobial properties and are thought to play a key role in both the adaptive and innate immune defense.  For various reasons, which are discussed below, these  antimicrobial cationic peptides have been receiving a lot of attention with the hope that they can be developed into useful therapeutic drugs. Antimicrobial cationic peptides are ubiquitous in nature and are thought to be a component of the first line of defense against infectious agents (Hancock et al., 1995). Cationic peptides have been isolated from a variety of sources including humans, plants, animals, insects, fish and bacteria.  Moreover, these peptides show a large chemical  diversity. In general these peptides are short and cationic, although large variations in net positive charge and peptide length can be observed.  In addition, amphipathicity  (structures with the positively charged side chains on one face of the molecule and the hydrophobic side chains on the other) is a property often associated with these antimicrobial peptides.  Amphipathicity can also vary, as calculated hydrophobic  moments can be very different between peptides (Hancock et al., 1995). At the genetic  1  level, the genes for these cationic peptides generally express a leader peptide, a propeptide and the mature cationic peptide.  The often negatively charged propeptide  likely forms an inactive complex with the mature peptide in order to protect the producing cell (Hancock et al., 1995). This pro-region is cleaved off in order to activate the mature peptide. Two evolutionarily distinct classes of peptides have been found in mammals: the defensins (Lehrer et al., 1991) and the cathelicidins (Zanetti et al., 1990). The defensins are characterized by their disulphide bonded B-sheet structure whereas the cathelecidin family of peptides is composed of a variety of different structures.  However, all the  cathelecidin peptides share a conserved N-terminal propiece. The cathelicidins are found in circulating neutrophils and at mucosal surfaces and are thus thought to be an important part of both systemic and local host defense (Gennaro and Zanetti, 2000). Defensins and cathelicidins are released from neutrophils together, and were found to act in synergy on E. coli and S. aureus (Nagaoka et al., 2000). There are many ways to classify peptides.  Peptides are often grouped together  according to their organism of origin. An example of this is the bacteriocins, a family of peptides produced by bacteria (Jack et al, 1995). Peptides can also be grouped according to the organizations of their genes. Cathelicidins are an example of such a family due to their highly homologous pro-regions, as mentioned above (Zanetti et al, However, structurally different  peptides  1990).  can have similar genetic origins while  structurally similar peptides can have different genetic origins. In general, peptides are most commonly classified by their secondary structure. There are four structural classes of cationic peptides; the disulphide bonded B-sheet peptides, including the defensins, the  2  amphipathic a-helical peptides such as the cecropins and melittins, the extended linear peptides which often have a single amino acid predominating, e.g. indolicidin and PR-39, and the loop structured peptides like bactenecin (Hancock et al, 1995). Some peptides contain unique characteristics, for example, the gramicidins (Urban et al, 1980) contain D-amino acids, and the lantibiotic peptides contain the uniquely post-translationally modified amino acids lanthionine and didehydroalanine (Jack et al, 1995). Many peptides have broad-spectrum antimicrobial activity, however, some peptides preferentially kill either Gram positive (eg. insect defensins (Matsuyama and Natori, 1988)) or Gram negative bacteria (eg. cecropin PI (Boman et al, 1991)), or fungi (eg. drosomycin (Fehlbaum et al., 1994)).  The basis for this selectivity is not well  understood. However, it is accepted in the field that an electrostatic interaction occurs between the cationic peptide and the negatively charged phospholipids in microbial membranes. In general, eukaryotic cells are less sensitive to cationic peptides, and this has been attributed to the differences in lipid composition between the membranes of eukaryotic and prokaryotic cells.  The plasma membranes of higher eukaryotes contain  sterols and the anionic phopholipids are only found on the inner surface of the plasma membrane.  In addition, eukaryotic membranes have a lower negative membrane  potential gradient (Matsuzaki et al, 1995). Some peptides have been shown to bind and neutralize components of the cell wall of bacteria that are released into the bloodstream and which are responsible for septic shock (Gough et al, components  (lipopolysaccharide  and lipoteichoic  acid)  1996). are  often  These cell wall released  with  conventional antibiotic treatment. In addition, these peptides have been shown to inhibit the induction of the cytokine tumor necrosis factor (TNF) in mice, further protecting  3  them from endotoxinaemia (Gough et al., 1996).  Specific peptides upregulate the  expression of interleukins that chemoattract inflammatory cells (Welling et al., 1998) or influence the complement system (van den Berg et al, 1998). However, some peptides are cytotoxic, but in certain cases this cytotoxicity is inhibited by host serum (Gennaro and Zanetti, 2000).  Interestingly, several peptides display antiviral activity.  For  example, defensins neutralize herpes simplex virus, vesicular stomatitis virus and influenza virus in vitro (Daher et al., 1986).  Cecropins, melittins (Wachinger et al.,  1998) and indolicidin (Robinson et al, 1998) display anti-HIV activity.  B. Structure/Function Studies with Cationic Peptides Many different structures and structural classes of antimicrobial cationic peptides exist. General statements about structure/function relationships are thus very difficult to make. It is widely believed that, regardless of structural class, amphipathicity is one of the most important structural requirements for antibacterial activity (Andreu and Rivas, 1998; Pathak et al., 1995). The large diversity in peptide structures has led to the theory that these peptides act in a non-specific manner that includes general membrane disruption. This membrane disruption is a result of an electrostatic interaction between the positively charged peptides and the negatively charged phospholipids of microbial membranes, and the subsequent induction of an amphipathic peptide structure.  This  would explain the fact that most synthetic peptides with all D-amino acids retain activity, indicating that the interaction is a non-stereo-selective physicochemical one.  Some  research has indicated that the driving force behind this membrane-peptide interaction is the membrane potential gradient across the cytoplasmic membrane (oriented interior negative).  A correlation between sensitivity to magainins and increased negative  4  potential in both bacteria and membrane vesicles (Matsuzaki et al., 1995; McAuliffe et al., 1998) as well as the decreased susceptibility of E.coli to indolicidin in the presence of the membrane uncoupler CCCP (Falla et al., 1996) support this theory. However, since variation in the properties (eg. amphipathicity and membrane potential) that appear to be necessary for membrane activity exists between peptides and bacteria, other factors are likely to be involved.  For example, some studies have suggested that strongly  amphipathic peptides tend to form relatively stable pores, whereas peptides with lower amphipathicity are more likely to migrate through the membrane (Matsuzaki et al., 1997).  The antimicrobial activity of the cathelecidin-derived peptides C A P 18 and  SMAP-29 was found to correlate with predicted hydrophobicity along the peptide backbone and net positive charge (Travis et al., 2000). The importance of a-helicity has also been studied with various peptides.  In  studies of the peptide buforin, an increase in activity correlated with an increase in ahelicity (Park et al., 2000). Correspondingly, an increase in membrane penetration was observed. In addition, the hinge region that exists in many a-helical peptides has also been studied.  These regions, often composed of proline or glycine residues, result in  helix-bend-helix structures. Peptide analogues based on CP26 (a variant of the cecropinmelittin hybrid peptide) were created with 0, 1 or 2 proline residues in this region (Zhang et al., 1999). The presence of a single proline residue resulted in peptides with increased specificity and selectivity.  In contrast, a peptide with no proline residues had increased  antimicrobial and hemolytic activities and decreased selectivity.  In addition, as the  number of proline residues increased, the ability to permeabilize the cytoplasmic membrane of E. coli decreased (Zhang et al., 1999). The hinge region of a cecropin-  magainin hybrid peptide is composed of a Gly-Ue-Gly motif.  Analogs of this peptide  were designed and it was found that the flexibility or B-turn that is provided by the hinge region is important for electrostatic interactions (Shin et al, 2000). The hinge allows the N-terminal region to reside on the membrane surface while the more hydrophobic amphipathic C-terminal region can interact with the acyl chains of the phopholipids (Shin et al., 2000). The hinge region in buforin was found to be important for penetrating the cytoplasmic membrane (Park et al, 2000).  An amino acid substitution of the proline  residue responsible for creating the hinge resulted in a peptide more like magainin, and this resulted in a different mechanism of killing action (Park et al, 2000). Recently, the importance of linearity for antimicrobial activity was examined (Oren and Shai, 2000). Cyclic analogues of a-helical peptides were designed and were found to cause an increase in selectivity between bacteria and erythrocytes. The cyclic peptides were active against both Gram negative and positive bacteria, but had impaired binding to zwitterionic membranes. In addition, cyclicization of these peptides prevented oligomerization (Oren and Shai, 2000).  Therefore it seemed likely that peptide  aggregates may lead to an increase in hydrophobic interactions and thus interaction with more neutral lipids, such as those found in the membranes of erythrocytes. Structure-function studies have been performed on peptides from other structural classes as well. An interesting class to study is that of the linear peptides with one or two amino acids predominating.  The secondary structures of the peptides that have been  studied from this class vary. For example, studies of novel peptide dodecamers indicated that these peptides were either a 3 i - or a-helix (Mayo et al., 2000). However, proline0  arginine-rich peptides such as Bac5 and PR-39 are thought to have a poly-proline II  6  helical structure or a y-helix (which is even more extended than poly-proline II helix). Other structural properties have been investigated as well. Studies with these peptides have shown that the arginine at the N-terminus was important for activity. In addition, an effect of chirality was seen, which is unusual for cationic peptides, indicating the possibility of a stereospecific interaction (Castle et al., 1999). The proline-arginine rich apidacin-type peptides were found to have a constant region that conferred the antibacterial function, and a variable region that conferred specificity (Castle et al., 1999). In this thesis, I examined the structure and activity of indolicidin, a prolinetryptophan rich peptide found in this structural class, as well as some variants of indolicidin.  I also examined cecropin-melittin hybrid analogues designed to have an  increase in a-helicity or amphipathicity. More details of these peptides are given later in this chapter.  C. Mode of Action Studies Little is known about the initial interactions of cationic peptides with the cell wall of Gram positive bacteria, whereas interactions of cationic peptides with the outer membranes of Gram negative bacteria have been studied extensively.  The initial  interactions of some cationic peptides with Gram negative bacteria are thought to involve binding to surface lipopolysaccharide (Piers and Hancock, 1994; Sawyer et al, 1988). The peptides bind the divalent cation binding sites at the base of LPS, thereby displacing the divalent cations that are essential for outer membrane integrity. Consequently the outer membrane bilayer is destabilized, resulting in transient disruptions (Peterson et al,  7  1987).  These disruptions allow the peptide to pass through the outer membrane and  access the cytoplasmic membrane where an electrostatic  interaction between the  negatively charged lipid head groups and the cationic peptides occurs. This process has been termed self-promoted uptake (Hancock, 1997). This ability to cause disruptions in the outer membrane of bacteria is thought to be an explanation for the synergy that is often seen between cationic peptides and classic antibiotics (Hancock, 1997). It was generally accepted in the field that cationic peptides electrostatically interact with the cytoplasmic membranes of both Gram negative and Gram positive bacteria.  In general, the wide antimicrobial spectrum and speed of action of cationic  peptides, as well as the activity of all-D isomers, suggested a non-stereospecific killing mechanism. Therefore, it has been commonly accepted that membrane disruption was the sole mechanism of killing action. There are different theories as to what happens at the surface of these membranes. It is increasingly disputed as to whether peptide channel formation leads to dissolution of the proton motive force and leakage of essential molecules (Cociancich et al., 1993; Juretic et al., 1989; Wu et al., 1999), or whether the action of peptides reflects a more general solubilizing effect of the peptides (Epand et al., 1995).  However, more recently evidence has been presented that peptide-membrane  interaction is an intermediate step in the uptake of peptide into the cytoplasm where it inhibits an essential function by e.g. binding to polyanionic D N A (Park et al., 1998; Zhang etal., 1999). Initially, cationic antimicrobial peptides have been demonstrated to insert into the membrane interface and adopt a position parallel to the membrane surface.  For  subsequent membrane interactions three general models have been proposed. The barrel-  8  stave model is one of the models suggested for peptide-membrane interaction. In this model, regular clusters of amphipathic peptides form pores (Perez-Paya et al., 1995). The result is proposed to be a water-filled barrel through which essential ions and small molecules escape from the cell, causing cell death. Depending on channel lifetime, this would be expected to cause complete dissipation of membrane potential. The peptide alamethicin is an example of a peptide that can form aqueous pores, as shown by neutron in-plane scattering (He et al., 1995). However, the formation of pores would be less favorable with cationic peptides because of the electrostatic repulsion that would occur between positively charged residues, and channel-like events have been observed even with very small cyclic peptides like bactenecin that would not be able to span the bilayer (Wu etal, 1999). This led to the proposition of the carpet model (Epand et al, 1995). In this model the peptides are proposed to cover the surface of the membrane in a carpet-like cluster. Once a saturation point is reached, there is a general perturbation of the membrane that leads to abrupt lysis of the microbial cell. membrane potential.  This would cause a rapid dissipation of the  Magainin is an example of a peptide believed to interact with  membranes in this fashion (Matsuzaki etal., 1998). Many peptides do not appear to cause such gross detergent-like effects on the membranes of bacteria, nor do they appear to form long-lived, defined pores. To explain these observations, the aggregate channel model has been postulated (Hancock and Chappie, 1999).  This model proposes that peptides form unstructured micelle-like  aggregates as opposed to the structured pores suggested in the barrel-stave model. These aggregates could act as short-lived channels for the passage of ions and upon dissociation  9  the peptides could locate to the surface of either monolayer. peptide to reach the cytoplasm.  This would allow the  This model would explain any lack of complete  membrane depolarization or gross membrane effects when bacteria die. The electrostatic nature of the peptide-membrane interaction has led some researchers to believe that peptide charge is a discriminating factor. However, charge is not the only factor since peptides often show a preference for different pathogens. The variation of lipid composition is a second factor that may explain peptide selectivity. The "two-state model" states that at low peptide/lipid ratios the peptide adsorbs in the head group region in a functionally inactive state, however, above a threshold concentration the peptide forms a multi-pore state that is lethal to the cell (Huang, 2000).  This  threshold peptide/lipid ratio is a function of the peptide binding energy, the energy level of the pore, and the lipid composition. Membrane potential allows for a lower energy of the insertion state but cannot explain the transition as a function of peptide concentration (Huang, 2000). Other studies with oriented circular dichroism spectroscopy (OCD) show how alignment and orientation of peptide helices with respect to the planar lipid are coupled to lipid phase state and critical peptide surface density (Clayton and Sawyer, 2000). Studies with synthetic lipid vesicles have emphasized the cytoplasmic membrane as the site of antimicrobial action.  These experiments have also demonstrated that  differences in membrane interactions exist between peptides.  However, results from  synthetic model membranes should be interpreted cautiously because they are imperfect models of biological membranes. As a result, assays that determine the depolarization and permeabilization of whole bacteria have been used. For example, the ONPG assay  10  has been commonly used (Friedrich et al., 1999; Lehrer et al., 1989) with E. coli. This assay measures the accessibility of a normally membrane-impermeable substrate (orthonitrophenyl galactosidase, ONPG) to cytoplasmic B-galactosidase (Lehrer et al., 1989). It suffers from the use of a bulky substrate (ONPG) which may be a poor indicator of the leakage of small molecules or depolarization.  Therefore, the membrane-potential  sensitive dye diSC (5) was introduced into use more recently. This dye distributes 3  between cells and the external medium according to the electrical potential gradient. The dye enters energized cells and concentrates in the cytoplasmic membrane where it quenches its own fluorescence.  Thus if cytoplasmic membrane depolarization occurs in  the presence of peptides it is measured as an increase in fluorescence.  This assay has  been used with E. coli (Wu et al., 1999) and was adapted in this thesis work for use with S. aureus.  Flow cytometry has also been used with various fluorescent dyes that  discriminate between depolarized and non-depolarized bacteria (Yeaman et al., 1998). Despite the existence of these assays using whole bacteria, synthetic lipid vesicles have been more often used to study peptide-membrane interactions. In fact, some differences can be seen between studies with synthetic lipid vesicles and whole bacteria.  For  example, when studying the mechanism of action of cecropin A, Silvestro et al. (2000) found differences in concentration dependencies of membrane permeabilization when studying membrane vesicles and whole cells using the ONPG and the DiSC3(5) assays. However, a close relationship between bactericidal and permeabilizing activity could be seen and the conclusion remained that the mechanism of action of cecropin involved membrane disruption.  11  More recently, no correlation was observed between membrane depolarization and killing activity of several different cationic peptides, and this prompted the reevaluation of the mechanism of action of these peptides (Wu et al, 1999). For example, buforin 2, a peptide with more potent antibacterial activity than magainin, was shown to have a weaker permeabilizing activity but was more efficiently translocated across the membrane than magainin (Kobayashi et al, 2000).  The translocation mechanisms  between the two peptides were shown to be different, as buforin did not cause lipid flipflop. The authors thus speculated that buforin had an intracellular target (Kobayashi et al., 2000).  Indeed, studies with biotin-labeled buforin analogues showed that peptide  potency correlated with the ability of the peptide to penetrate the cytoplasmic membrane of E. coli and accumulate in the cytoplasm (Park et al, 2000). More evidence that the bactericidal effect of some peptides is not due to cytoplasmic membrane depolarization was presented recently (Daugelavicius et al., 2000; Zhang et al, 2000b).  Their work  with polymyxin B showed that the bactericidal effect was at concentrations below that needed for pore formation. The capability of E. coli to form colonies was lost in the presence of non-depolarizing concentrations of polymyxin B.  Thus the investigators  concluded that pore formation and depolarization of the cytoplasmic membrane were not obligatory for the killing action of the peptide. Recent work by Castle et al. (1999) with 14  C-labeled apidaecin-type peptides (proline and arginine-rich) showed that peptide  uptake in E. coli was stereospecific and mediated by a transporter mechanism. transport was necessary but not sufficient for bacterial killing.  This  In their study, some  peptide analogues were taken up by the cell but did not exhibit antibacterial action, and evidence was presented that supported inhibition of protein synthesis machinery as being  12  the downstream target of these peptides (Castle et al, 1999). Otvos et al. (2000b) used fluorescein-labeled  peptides  to identify biopolymers that bound to the  peptides  pyrrhocoricin, apidaecin and drosocin. Through this work they discovered that these proline-rich peptides bound DnaK, a bacterial heat shock protein, and the chaperone GroEL in a stereospecific manner. They determined that the binding of these proteins and bacterial killing were related events. Another example of a specific peptide-protein interaction is the selective inactivation of microbial serine protease by the equine neutrophil peptide eNAP-2 (Couto et al, 1993). Other studies have demonstrated inhibition of macromolecular synthesis. Subbalakshmi and Sitaram (1998) showed that indolicidin permeabilized the membrane of E. coli and then inhibited D N A synthesis.  Boman et al. demonstrated that PR-39  killed bacteria by stopping both D N A and protein synthesis (Boman et al,  1993).  Regardless of the recent evidence of alternate mechanisms of action that include an intracellular target, the interaction between antimicrobial peptides and the cytoplasmic membranes of bacteria are important in order for the peptide to access the cytoplasm. Most studies discussed thus far have been on Gram negative bacteria or have used synthetic model membranes.  Studies of the effects of cationic peptides on the  membranes of Gram positive bacteria have also been conducted. Some cationic peptides have been shown to interact with the cytoplasmic membrane in a voltage-dependent manner (Falla et al, 1996; Kordel et al, 1988). However, other studies have indicated that the requirements for transmembrane potential vary among cationic peptides. For example, Koo et al. (1996) showed that defensins are either transmembrane potential independent or they have a low threshold A T for activity compared with thrombin-  13  induced platelet microbicidal protein-1 (tPMP) when studying a related pair of S. aureus strains differing in A^F generation. Using flow cytometry, Yeamen et al. (1998) showed that  human neutrophil  defensin-1  (HNP-1) depolarized  and permeabilized the  cytoplasmic membrane of S. aureus in vitro, whereas tPMP-1 did not depolarize but did permeabilize the membrane.  It has also been shown that the properties of the  cytoplasmic membrane itself may influence bacterial susceptibility to cationic peptides. Protoplasts from resistant and stationary phase S. aureus were less susceptible to lysis and were more intact after treatment with tPMP-1 (Koo et al., 1997). More recently it was shown that the membranes of S. aureus strains resistant to tPMP-1 had elevated levels of longer chain and unsaturated membrane lipids, resulting in increased fluidity and membrane disorder (Bayer et al., 2000).  The presence of qacA, a proton motive  force-dependent pump, has also been linked to tPMP resistance (Kupferwasser et al., 1999). Ultrastructural studies of S. aureus treated with defensins (Shimoda et al., 1995) and platelet microbicidal proteins (PMPs) (Yeaman et ai, 1998) showed cell membrane damage followed by cell death. The defensins caused mesosome-like structures to appear before the bacteria lost their viability, but no remarkable effects on the cell wall were seen (Shimoda et al., 1995).  However, D-mastoparan, a peptide isolated from hornet  venom, resulted in bleb-like extrusions on the surface of S. aureus, as visualized by scanning electron microscopy (Li et al., 2000). Through the use of the same DiSC3(5) assay used in this thesis work, Krijgsveld et al. (2000) discovered that thrombocidin could not dissipate the membrane potential of Lactobacillus lactis despite its antibacterial activity, leading them to the conclusion that the mechanism of action did not involve pore  14  formation. These studies have collectively suggested that for some peptides, membrane disruption is an important but not necessarily lethal event. Recent studies have indicated that some peptides may indeed have an intracellular target. Xiong et al. (1999) found that S. aureus, pretreated with inhibitors of D N A gyrase or protein synthesis, demonstrated decreased or blocked killing by HNP-1 and tPMP-1, whereas pretreatment with bacterial cell wall synthesis inhibitors enhanced bacterial killing.  The authors concluded that cytoplasmic membrane effects occurred prior to  effects on protein and D N A synthesis. Certain bacteriocins with selective activity against Gram positive bacteria can damage the cell membrane and cause lysis of the cell. However, it has been proposed that lysis is a secondary effect caused by deregulation of the autolytic system that results in destruction of the peptidoglycan layer (Martinez-Cuesta et al., 2000).  These  conclusions arise from the observations that lysis occurred only when active autolysins were present.  Depletion of cellular energy would cause an imbalance in control of  autolysins resulting in cell wall degradation and thus cell lysis.  Also, regulation of  autolysins by electrical potential of the cell membrane, lipoteichoic acid or extracellular proteinases has been shown.  Autolysin activation has been implicated as a mode of  action for the cationic bacteriocins nisin and Pep5 on S. aureus (Bierbaum and Sahl, 1985) but was not found to be a significant mechanism for HNP-1 and tPMP-1 (Xiong et al., 1999).  In addition, a direct correlation was observed between tolerance to  antibacterial cationic peptides and the D-ala content of teichoic acids, a polymer in the peptidoglycan layer of Gram positive bacteria (Peschel et ai, 1999).  15  However other  studies (Scott et al., 1999) found no correlation between binding to lipoteichoic acid and minimal inhibitory concentration. Other mechanisms have been proposed for the Gram positive-specific lantibioticbacteriocins.  Lantibiotics are thought to be inactive against Gram negative bacteria  because of the outer membrane barrier. There are 2 types of lantibiotics: type A, which are believed to kill by pore formation, and type B, which are thought to inhibit peptidoglycan synthesis resulting in decreased cell wall thickness and lysis (Brotz and Sahl, 2000).  Nisin is believed to act against Lactobacilli by binding the lipid-bound  peptidoglycan precursor Lipid II (Brotz and Sahl, 2000). These studies with Gram positive bacteria have shown much diversity in the effects of different antimicrobial cationic peptides. However, at the time this thesis study was initiated, much less was known about the effects and potential mechanisms employed by cationic peptides on Gram positive bacteria, and indeed results presented here were published before some of the above-cited papers (Friedrich et al, 2000). This thesis study investigated the effects of a set of diverse peptides on the membranes and on macromolecular synthesis by S. aureus.  D. Peptides Used in this Study Cecropins were originally isolated from the immune hemolymph of the North American silk moth Hyalophora cecropia (Hultmark et al., 1980). Cecropins have been well studied and characterized with respect to structure and function (Fink et al., 1989). Based on model membrane studies, the broad spectrum of antimicrobial activity of cecropins has been attributed to an ability to form large pores in bacterial cell membranes  16  (Christensen et al., 1988).  A series of hybrid peptides was created consisting of the  amphipathic, a-helical N-terminal region of cecropin A followed by the hydrophobic N terminal a-helix of the bee venom peptide melittin (Wade et al., 1992). These hybrids form ion-permeable channels in model lipid membranes (Wade et al, 1990). To understand the structure-function relationships of these peptides, analogues based on cecropin (1-8) melittin (1-18) hybrid (CEME) were studied.  The general  conclusions from these studies were that the hybrids should have a hydrophilic and a hydrophobic domain linked by a hinge region (Boman et al, 1989).  A hinge region  provides conformational flexibility due to the presence of glycine and proline residues (Andreu et al, 1992). The aromatic residue at position 2 and the a-helical region in the first 11 amino acids have been indicated as necessary components for antimicrobial action (Andreu et al, 1985). Piers et al. (Piers et al, 1994) further modified the hybrid peptide C E M E (also called MBI-27, (Gough et al, 1996)) by adding two extra positively charged residues to the C-terminus in order to assess the role of charges in the interactions with bacteria. The resulting peptide (CEMA; also called M B 1-28, (Gough et al, 1996)) had similar MICs to C E M E but had an increased ability to permeabilize the outer membrane of Gram negative bacteria and an increased affinity for LPS. The 26-amino acid a-helical peptides CP26 and CP29 (Friedrich et al,  1999)  (Table 1) are derived from this series of hybrid peptides consisting of the amphipathic, ahelical N-terminal region of cecropin A followed by the hydrophobic N-terminal a-helix of the bee venom peptide melittin (Wade et al,  1992).  The C E M E sequence was  modified to increase the a-helical content in the first 14 amino acids, resulting in CP29. Another related peptide, CP26, had the same first 10 residues as CP29, but the C-  17  terminus was predicted to be more hydrophilic with the addition of an extra positively charged lysine.  These peptides (CEME, C E M A , CP26 and CP29) (Table 1) were  analyzed, in this study using spectroscopy; as well, CP26 and CP29 were included in the mode of action experiments. The  13 amino acid peptide indolicidin was isolated from bovine neutrophils  (Selsted et al., 1992).  Indolicidin and its improved derivative CP11CN (Falla and  Hancock, 1997) (Table 1) have a unique amino acid composition consisting of very high percentage of tryptophan and proline and are amidated at their C-termini. CPI 1CN was designed to have a greater positive charge and enhanced amphipathicity and was found to have improved activity against Gram negative bacteria. This correlated with an increased ability to bind LPS and to permeabilize the outer membrane of E.coli (Falla and Hancock, 1997). Indolicidin was shown to form pores in model membranes (Falla et al, 1996) as well as to inhibit D N A synthesis in E. coli (Subbalakshmi and Sitaram, 1998). These peptides have been shown to have an extended structure distinct from a-helical and Pstructured peptides (Falla et al, 1996; Rozek et al, 2000). However, there has been debate over the secondary structure of indolicidin. Falla et al. (1996) suggested that the secondary structure of indolicidin was that of a poly-proline U helix, similar to that seen previously with PR-39 (Cabiaux et al, 1994). It was then suggested that this C D analysis was biased because the contribution of tryptophan masked certain spectral features, and therefore the structure included a P-turn (Ladokhin et al, 1999). Tritrypticin, a 13-residue peptide derived from the cathelecidin peptide C12, contains the aromatic-prolinearomatic motif similar to that seen in indolicidin (Nagpal et al, 1999). Interestingly, C D spectroscopy studies of this peptide indicated that the structure is a cluster of poly-proline  18  II helix and P-turn (Nagpal et ah, 1999) and is an amphipathic molecule with segregation of cationic and aromatic residues. These researchers have suggested a possible functional activation through conformational transition from P-turn to poly-proline II helix. Further structural studies were done in the presence of various lipid environments in this thesis study in order to gain further insight into the structure of indolicidin and CP11CN.  My work contributed to and complemented the 2D-NMR studies of these  peptides undertaken by Dr. Annett Rozek in our lab. Dr. Rozek determined the structures of indolicidin, CP11CN and CP10A (indolicidin variant with prolines substituted with alanines) in the presence of dodecylphosphocholine (DPC) micelles.  Indolicidin was  found to have a "boat" shape with the elements of both poly-proline II helix and P-turn (Rozek et al., 2000). CPI 1CN was found to have a very similar structure to indolicidin with a larger charged region at the ends of the peptide (unpublished results). CP10A, on the other hand, was found to be a short a-helix with a large hydrophobic face. Because of its unique sequence and structure, I chose indolicidin as the focus of my structure/activity analysis and produced various indolicidin analogues.  I looked at activity on both Gram  negative and Gram positive bacteria, with emphasis on selectivity.  Subbalakshmi et al.  (1996) studied the activity of some indolicidin variants on E. coli and S. aureus, as well as their hemolytic activities.  The substitution of all tryptophan residues  with  phenylalanines resulted in a peptide that retained activity against E. coli. Further analysis by the same group showed that single tryptophan analogues (again, with W-> F substitutions) resulted in peptides with only a slight decrease in potency against E. coli but with abolished hemolytic activity (Subbalakshmi et al., 2000).  These peptide  analogues all had a p-turn structure in T F E and micelles, as shown by C D spectroscopy.  19  An interesting recent development in the structure/function studies of indolicidin was the creation of X-indolicidin, an analogue with Trp-Trp cross-links (Osapay et al., 2000). This peptide was found to be similar in both amphiphilicity and activity to the wild-type, yet was resistant to proteolysis (Osapay et al., 2000). CP10A (Table 1) is a peptide with the 3 proline residues of indolicidin replaced with alanine (Subbalakshmi et ai, 1996). This peptide was found to have activity against E. coli and S. aureus and was used in this thesis. Indolicidin, CP11CN and CP10A were all included in the mechanism of action studies presented here. Bactenecin (dodecapeptide) is a 12 amino acid peptide originally isolated from bovine neutrophils (Romeo et al., 1988).  Bactenecin has 2 cysteine residues that are  thought to form an intramolecular disulfide bond, resulting in a type I P-turn structure. Indeed, the solution structure of chemically synthesized, monomeric bactenecin was determined using 2D-NMR and was found to be a loop-like structure with a type I P-turn and two anti-parallel P-strands (Raj et al., 2000). However, other observations have led some researchers to believe that the cysteine residues may be involved in intermolecular disulfide bridges in nature, resulting in a peptide dimer (Storici et al., 1996). Bac2A-NH  2  (Table 1) is a linearized, N-terminally amidated version of the synthetic cyclic peptide bactenecin with better activity against Gram positive bacteria than the native bactenecin (Wu and Hancock, 1999). The linearization of the loop may actually result in a peptide that is more like the natural dimeric form of the peptide (Storici et al., 1996). Because of the improved activity against Gram positive bacteria, Bac2A-NH was chosen as a 2  peptide to be used in some of the studies presented here.  20  The peptides described above were all very different from each other structurally. It was thus of interest to determine characteristics of these peptides such as their structure/activity relationships and mechanism of action on Gram positive bacteria. This would give a more general view of cationic peptides and would help determine if certain characteristics are widespread among the structural classes. Most studies focus on one peptide or one set of similar peptides, so this unique approach was used in the hope that it would provide a more "global" view of antimicrobial cationic peptides.  E . Methods of Determining Antimicrobial Activity The National Committee on Clinical Laboratory Standards (NCCLS) broth dilution method of determining MICs is commonly used in antibiotic field (Amsterdam, 1996). In general, this method involves making serial dilutions of the peptide in the wells of multi-well plates and adding a known amount of bacteria. The MIC is then recorded as the concentration of peptide that results in inhibition of bacterial growth when compared with the control. However, researchers have adapted the method to suit the peptide field, thus variations on the method exist. As expected, there are limitations to this method. MICs are generally read after 24 h of incubation.  In this time period, a fast-acting  peptide may kill most bacteria within the first hour, however, if only one bacterium survives and divides during the remaining time of incubation, a falsely high MIC reading may result.  As well, some peptides have been found to precipitate or bind to plastics  such as polystyrene, resulting in a decrease in effective peptide concentration. This has led to some debate as to the most appropriate method of determining peptide activity. It has been suggested that peptides in 0.1% serum albumin will prevent peptide adherence  21  to plastic and 0.01% acetic acid will prevent the peptide from precipitating (Cole and Ganz, 2000).  This, in conjunction with using polypropylene multi-well plates, would  minimize these problems and give a more accurate MIC. This revised method is posted on the Hancock Lab website (http://www.cmdr.ubc.ca/bobh/MIC.htm). These concerns led me to do some comparison studies of MIC methods in this work.  Recently,  Giacometti et al. compared the NCCLS and the "Hancock method" and found that the NCCLS method consistently resulted in 2-4 fold higher MICs (Giacometti et al, 2000). This group also found that killing was more rapid in polypropylene tubes than in polystyrene tubes, further evidence of a potential binding reaction occurring between peptides and polystyrene plastic (Giacometti et al., 2000). In addition to the broth dilution method, various methods involving the use of solid media are sometimes used to determine MIC.  Often this involves the addition of the  peptide into a well or a filter (ie. disk diffusion) on the surface of solid media, and the subsequent zone of inhibition on a lawn of bacterial growth is measured (Acar and Goldstein, 1996). However, agar carries an overall negative charge due to acidic and sulfate groups on the polysaccharides., The size of the zone of inhibition depends on the ability of the cationic peptides to diffuse rapidly through the gel. However, interaction of the positively charged peptide with the negatively charged agar will likely impede diffusion making it difficult to use this method to assess the MICs of peptides. Another variation of the solid media method involves adding peptide to molten agarose before solidification, followed by spotting a known amount of bacteria on the surface. The amount of colonial growth is estimated in order to determine MIC. Agarose is often used in favor of agar because of its more defined nature and more neutral charge (Acar and  22  Goldstein, 1996).  Additional disadvantages of these solid media methods include the  amount of time and labor needed, as well as the large amount of peptide required. In this study, I compared this latter agarose method with values obtained by the broth dilution method.  F. Effects of Salt on Antimicrobial Activity The physiological concentrations of salt are approximately 100 m M of NaCl, 1 mM M g S 0 and 2 m M CaCl (Amsterdam, 1996). It is therefore important to analyze 2  4  2  the activity of cationic peptides under these conditions. We reported the salt-resistance of cecropin-melittin hybrid peptide analogues when looking at activity against Gram negative bacteria (Friedrich et al., 1999). These peptides were active in the presence of and therefore resistant to concentrations of 150 mM NaCl, but were antagonized by 3 to 5 mM M g C l . 2  Mg  This antagonism was attributed to competition with the peptides for the  binding sites at the base of LPS on the outer membranes of Gram negative bacteria.  + +  This displacement of M g  + +  is believed to be part of the self-promoted uptake process of  cationic peptides across the outer membrane (Hancock et al., 1995). Some cathelicidins are resistant to physiological concentrations of salt (150 mM NaCl), whereas defensins are sensitive to the presence of salt (Nagaoka et al, 2000). The recombinant peptide ASABF, originally isolated from nematodes, is a very active peptide against S. aureus but is antagonized by concentrations of 40-80 mM NaCl, and is even more inhibited by CaCl (Zhang et al., 2000a). 2  This sensitivity has been attributed to the inhibition of  electrostatic interactions in the presence of high ionic strength medium. The structural characteristics of peptides that confer resistance or sensitivity to salt antagonism are  23  unknown. However, the sensitivity to ionic strength may depend on the target organism. It was therefore of interest in this thesis to examine the activity of the peptides of various structural classes in the presence of different salts and to compare the effects on E. coli  and S. aureus.  G . Therapeutic Potential and Peptide Development The emergence of antibiotic-resistant bacteria has led to the search for new antibiotic therapeutics.  Antimicrobial cationic peptides, which are found in almost all  living organisms, are potential candidates. However, the development of these peptides for therapeutic use is problematic. For example, the balance of toxicity and antibacterial potency is a major issue. In addition, antimicrobial cationic peptides are often sensitive to proteolytic degradation. This characteristic may allow these peptides to be used safely as fast degradation will not induce an immune response; however, this rapid clearance may result in a lack of antimicrobial action. Unnatural or all-D amino acids incorporated into these peptides may circumvent this problem of proteolysis (Blondelle et al., 1994). Despite these problems, the minimal tendency of antimicrobial peptides to induce bacterial resistance has made the prospect of developing these peptides into therapeutics more attractive and is one of its main advantages over conventional antibiotics. There have been different methods used to determine which peptides are worthy of further investigation.  Combinatorial libraries have been shown to produce a large  number of random peptides for screening (Houghten et al., 1991). The results have been quite promising, however, this method requires a large initial investment.  Another  approach is the designing of peptides de novo based on repeating motifs resulting in an  24  amphipathic peptide (Blondelle and Houghten, 1992).  More commonly, attempts at  developing new peptide antibiotics are made from existing peptides.  There is also the  high production cost of peptides and other safety issues that need to be considered. Economically viable large-scale production of these antimicrobial peptides is a whole separate issue currently being addressed.  Recombinant expression methods using  bacteria (Baneyx, 1999; Piers et al., 1993) and yeast (Sudbery, 1996) as well as production using transgenic plants or animals have been explored.  Indeed, elevated  levels of peptides have been produced in the leaves and seeds of transgenic plants (Jach et al, 1995). This has potential for use in protecting crops and reducing the use of crop protecting chemicals. Gene therapy with cationic peptide sequences is another area with potential. Using animal models, histatin genes have been successfully incorporated into salivary glands (O'Connell etal., 1996). Despite all of the issues surrounding the development of cationic peptides as therapeutics, there have been a few peptides brought to clinical trials as topical medicines.  For example, peptide candidates are in clinical trials for prevention of  catheter-associated and acne infections, the treatment of oral mucositis, as well as N. meningitidis infections and H. pylori infections (Hancock and Lehrer, 1998).  H . Aims of this Study The majority of the studies of antimicrobial cationic peptides have been done with Gram negative bacteria, with some exceptions, such as the work done on Gram positive specific bacteriocins. In addition, most studies have been done with one peptide or set of peptides of a particular structural class. This thesis study was initiated to gain insight into  25  mechanism of action against Gram positive bacteria of peptides from different structural classes. The interaction and binding of different peptides with lipids as well as the activity of peptide variants were investigated in order to determine if a correlation existed between structure/lipid interaction and antibacterial activity. I hypothesized that different peptides would have different structures and lipid interactions and that this would lend insight into peptide selectivity between Gram positive and Gram negative bacteria. However, a lack of correlation in this area led to further investigation into the mode of action.  Therefore, various techniques were used with a selection of peptides from  different structural classes in order to further investigate the mechanism of action of these peptides on Gram positive bacteria, using mainly S. aureus as the model organism. More specifically, it was of interest to determine if the widely accepted theory of membrane disruption as the mechanism of action could be applied over a range of structurally different peptides on Gram positive bacteria.  I hypothesized that there would be  differences in the mode of action of structurally-different peptides, and that these mechanisms may include targets other than the cytoplasmic membrane.  26  Table 1: Amino acid sequences of antimicrobial cationic peptides used in this study  Peptide  Amino A c i d  Sequence  Length  Charge  % Hydrophobic amino acids  CP26  KWKSFIKKLTSAAKKWTTAKPLISS  26  +7  46  CP29  KWKSFIKKLTTAVKKVLTTGLPALIS  26  +6  50  CEME  KWKLFKKIGIGAVLKVLTTGLPALIS  26  +5  58  KWKLFKKIGIGAVLKVLTTGLPALKLTK  28  +6  54  CP201  KWKSFIKNLTKGGSKILTTGLPALIS  26  +5  42  CP208  KKKSFIKLLTSAKVSVLTTAKPLISS  26  +6  46  INDOL-  ILPWKWPWWPWRR-NH  2  13  +3  77  CP11CN  ILKKWPWWPWRRK-NH  2  13  +5  62  CP10A  ILAWKWAWWAWRR-NH  2  13  +3  77  Bac2A-  RLARIWIRVAR-NH  12  +4  67  (CP27) CEMA (CP28)  ICIDIN  2  NH  2  27  MATERIALS AND METHODS  A. Bacterial Strains and Growth Conditions  All bacterial strains used in these studies are listed in Table 2.  Most bacterial  strains used for antimicrobial activity studies were grown in no salt Luria broth (LB; 10 g/L bacto-tryptone and 5 g/L bacto-yeast extract) unless specified. Streptococcus strains were grown in Todd Hewitt broth (500 g/L beef heart infusion, 20 g/L bacto neopeptone, 2 g/L bacto-dextrose, 2 g/L sodium chloride, 0.4 g/L disodium phosphate, 2.5 g/L sodium carbonate).  In some specified cases Mueller-Hinton (MH) broth (beef extract, acid  hydrolysate of casein, and starch in a ratio of 3:17.5:1.5 as well as approximately 4.84 mg C a / L and 4.24 mg M g / L ) was used. ++  M H , L B and Todd Hewitt broth were all  ++  purchased from Difco Laboratories, Detroit, Michigan. S. aureus ISP67 was grown in a defined synthetic medium (which contains, 0.05 g/L M g S 0 , 7.0 g/L K 2 H P 0 , 2.0 g/L 4  4  K H P 0 , 0.5 g/L Na citrate, 1.0 g/L (NH )S0 and 2.5 g/L glucose 20 mg/L thymidine, 2  4  4  4  1.0 mg/L thiamine, 1.2 mg/L nicotinic acid, 0.25 mg/L D-pantothenate (calcium salt), 0.005 mg/L biotin, 1.0 mg/L niacin, 5.0 mg/L adenine, 5.0 mg/L guanine, 5.0 mg/L cytosine, 5.0 mg/L uridine and all amino acids ranging in concentration from 10-100 mg/L (including 20 mg/L histidine)).  Thymidine, uridine and histidine were purchased  from Sigma Chemical Co., St. Louis, MO. All strains were grown on a shaker at 37 °C.  28  Table 2: Bacterial Strains Strains Escherichia coli UB1005  Description  Reference/S ource  nal , met"  (Grinsted etal, 1972)  Pseudomonas aeruginosa H103  wild-type  (Hancock, 1979)  Salmonella typhimurium C590 C610  wild-type phoP/phoQ mutant; defensin sens.  (Fields etal, 1989) (Fields etal, 1989)  Staphylococcus aureus A T C C 25923 SAP0017 ISP 67  wild-type methicillin-resistant (MRSA) auxotroph for thy, uri and his  A T C C collection T. Chow, U B C J. Iandolo (Kuhl etal, 1978)  Staphylococcus epidermidis C621  clinical isolate  T. Chow, U B C  r  Staphylococcus haemolyticus Isogenic strains clinical isolate vancomycin-resistant  T. Chow, U B C T. Chow, U B C  Enterococcus faecalis A T C C 29212  wild-type  A T C C collection  Listeria monocytogenes N C T C 7973  wild-type  N C T C collection  Corynebacterium xerosis  isolate  U B C collection  Streptococcus pneumoniae A T C C 49619  wild-type/clinical isolate  A T C C collection  Streptococcus pyogenes A T C C 19615  wild-type/clinical isolate  A T C C collection  Streptococcus mitis  clinical isolate  U B C collection  29  B . Chemicals  All peptides (Table 1) were synthesized by N-(9-fluorenyl)methoxy carbonyl (Fmoc) chemistry at the Nucleic Acid Protein Service (NAPS) unit at the University of British Columbia. Bovine Serum Albumin (BSA) Fraction V lyophilisate was purchased from Boehringer Mannheim (Germany). Dipropylthiacarbocyanine (DiSC (5)) was purchased 3  from Molecular Probes (Eugene, Oregon). Valinomycin was purchased from Sigma (St. Louis,  Missouri).  l-Pamitoyl-2-oleoyl-sn-glycero-3-phosphocholine  palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol Lipids Inc., Vancouver, B C , Canada. phosphotempocholine phosphocholine  (tempo  (5-doxyl  PC), PC),  (POPC)  and 1-  (POPG) were purchased from Northern  Head group labeled  l,2-dioleoyl-sn-glycero-3-  l-palmitoyl-2-stearoyl-(5-doxyl)-sn-glycero-3l-palmitoyl-2-stearoyl-(12-doxyl)-sn-glycero-3-  phosphocholine (12-doxyl PC) and lysophophocholine (lyso-PC) were obtained from Avanti Polar Lipids Inc., Alabaster, A L , USA. Dodecylphosphocholine purchased from Cambridge Isotopes Laboratories, Andover, M A . , USA.  (DPC) was Carbonyl  cyanide-ra-chlorophenyl hydrazone (CCCP), E. coli 011B4 smooth lipopolysaccharide (LPS) and S. aureus lipoteichoic acid (LTA) were all purchased from Sigma Chemicals Co (St Louis, MO).  Radiolabeled H-thymidine, H-uridine and H-histidine were 3  3  3  purchased from Sigma Chemicals C. (St Louis, MO).  C. Liposome preparation A chloroform solution of lipid was first dried under a stream of N and then put under 2  vacuum to completely remove the solvent. The resulting lipid film was re-suspended in  30  10 mM phosphate buffer (pH 7.0).  When the liposomes were prepared with peptide  present, a methanol solution of the peptide was mixed in with the lipid-chloroform solution before drying.  The suspension was put through five cycles of freeze-thaw to  produce multilamellar liposomes, followed by extrusion through 0.1 um double-stacked Poretics filters (AMD Manufacturing Inc., Mississauga, ON, Canada) using an extruder device (Lipex Biomembranes, Vancouver, B.C.).  For some circular dichroism (CD)  experiments, liposomes were further extruded through 0.05 um filters. In the case of the spin-labeled lipids, liposomes were prepared as above, using 2.6 umol (10 mol % of total lipid) of the spin labeled lipid (phosphotempocholine, 5-doxyl phosphocholine and 12doxyl phosphocholine) with 16.4 umol POPC and 7 umol POPG.  D . Structural Studies - Spectroscopy D.l. Fluorescence  Spectroscopy  Fluorescence emission spectra were recorded on an LS 50B spectrofluorometer (Perkin Elmer (Canada) Ltd., Markham, ON, Canada). Measurements were performed between 300-450 nm at 1-nm increments using a 5 mm quartz cell at room temperature. The excitation wavelength was set to 280 nm with both the excitation and emission slit widths set to 4 nm. Spectra were baseline corrected by subtracting blank spectra of the corresponding lipid or detergent solutions without peptide. The samples contained 2 uM peptide and 0.5 mM lipid or 10 mM detergent in lOmM HEPES buffer (pH 7.2). The aqueous quencher potassium iodide was added in 20 mM increments in order to assess tryptophan accessibility to the aqueous buffer. Spin labeled lipids were used in order to estimate tryptophan position in the liposomes.  31  D. 2. Circular Dichroism Measurements Circular Dichroism (CD) spectra were obtained using a J-720 spectropolarimeter (Japan Spectroscopic Company, Tokyo, Japan). Each spectrum (190-250 nm) was the average of 4 scans using a quartz cell of 1 mm path length at room temperature. The scanning speed was 50 nm/min at a step size of 0.1 nm, 2 s response time and 1.0 nm bandwidth. All samples were 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentrations of lipid or detergent were 2 mM or 10 mM, respectively. The concentration of lipoteichoic acid used was 0.5 mg/ml.  A range of LPS amounts, from  15 fig to 500 pg in 200 ul, were used. Spectra were base line corrected by subtracting a blank spectrum of a sample containing all components except the peptide. After noise correction, ellipticities were converted to mean residue molar ellipticities [0] in units of degxcm /dmol.  E . Antimicrobial Activity  E.l. Minimal Inhibitory Concentration The minimal inhibitory concentration (MIC) of each peptide was determined using a broth microdilution assay modified from the method of Amsterdam (Amsterdam, 1996). Briefly, serial dilutions of each peptide were made in 0.2% BSA/0.01% acetic acid (unless otherwise  specified)  solution in 96-well  polypropylene (Costar, Corning  Incorporated, New York) microtiter plates. Each well contained 10 ul of peptide and was inoculated with 100 ul of the test organism in LB broth (or Todd Hewitt broth in the case of Streptococci) to a final concentration of approximately 10 cfu/ml. The MIC was 5  32  taken as the lowest peptide concentration at which growth was visibly inhibited (as judged by eye) after 24 h incubation at 37°C. The agarose MIC method involved pouring plates with media containing agarose and the desired concentration of peptide. A 1 to 2 ul sample of a 10 CFU/ml suspension of 7  bacteria (resulting in 10 CFU) was then spotted onto the surface of the media using a 4  replica plater.  The spots were allowed to dry before placing the plates in the 37°C  incubator for 24 h.  The MIC was defined as the concentration of peptide where the  amount of visible colony growth was approximately 10% (or less) of the confluent growth seen with the control.  E. 2.Bacterial Killing Assays Overnight cultures were diluted 10" in LB broth and allowed to grow to exponential 2  phase (OD o of 0.6) and then diluted in fresh medium to give a working concentration of 60  6  7  10 to 10 cells/ml. The peptides or antibiotics were added at 10 times their MIC and this suspension was incubated at 37°C. At regular intervals after peptide addition, samples were removed, diluted and plated on to L B agar plates to obtain a viable count. When killing curves were done in the presence of CCCP, lOuM, 50uM or lOOuM C C C P was added to the working solution of cells for 10 minutes at 37°C before addition of peptide.  F. Electron Microscopy F.L Transmission Electron Microscopy Two methods were used in the preparation of thin sections. In the case of S. aureus, S. epidermidis and E. faecalis in the presence of CP11CN, CP29, Bac2A-NH and 2  gramicidin, the facilities at T.J. Beveridge's lab at the University of Guelph were used  33  along with the following protocol.  Exponential phase bacteria were treated with the  peptide (at 10 times the MIC) for either 10 or 30 minutes at 37°C. After treatment, the bacterial pellets were enrobed in 2% Noble Agar (Difco) and fixed with 2.5% buffered glutaraldehyde for 1 hour.  The cells were then post-fixed in 1% buffered osmium  tetroxide for 1 hour, stained en bloc with 1 % uranyl acetate, dehydrated in a graded series of ethanol, and embedded in L.R. white resin. cacodylate, pH 7.4.  The buffer used was 0.1M sodium  Thin sections were prepared on Formvar, carbon stabilized copper  grids and stained with 1% uranyl acetate and lead citrate.  The resin and grids were  purchased from Marivac (Halifax, N.S., Canada). Microscopy was performed with a Philips EM300 microscope under standard operating conditions. In the case of S. aureus and S. epidermidis in the presence of CPI OA and CP26 and S. aureus in the presence of indolicidin, the above protocol was used at the U B C Biosciences E M lab facility, with the following exceptions.  Spur resin was used, the copper grids used were not copper-  stabilized and 2% uranyl acetate was used instead of 1% for staining of the thin sections. The resin and grids were purchased from Canemco (Toronto, ON, Canada). Microscopy was performed with a Zeiss Stem 10C microscope under standard operating conditions.  F.2. Scanning electron microscopy The preparation of samples for S E M was similar to that of T E M with the following exceptions. The samples were pushed through a syringe in order to collect the sample on filters. These filters were then placed in the wells of 24-well polystyrene plate (Costar, Corning Incorporated, New York) and fixed according to the protocol described for T E M . Once the sample was dehydrated and in 100% ethanol, the filters were placed in a  34  critical point dryer (CPD 020) to completely dry the filters by exchange with liquid CO2. The samples were then coated with a thin layer of gold using a Nanotech Semprep II Sputter Coater. Microscopy was performed with a Hitachi 4100 high resolution field emission scanning electron microscope at the Department of Physics, U B C , under standard operating conditions.  G. Cytoplasmic Membrane Depolarization Assay The depolarization of the cytoplasmic membrane of S. aureus A T C C 25923 by the peptides was determined with the membrane potential sensitive cyanine dye DiSC (5) 3  (Sims et al, 191 A) using a modification of the method of Wu and Hancock (Wu et al, 1999).  Briefly, exponential phase bacteria were washed and resuspended in 5 mM  Hepes, 20 mM glucose buffer (pH 7.2) to an OD of 0.05. This cell suspension was incubated with 100 mM KC1 (to equilibrate cytoplasmic and external K concentration) +  and 0.4 uM DiSC (5) until there was a stable (approximately 90%) reduction in 3  fluorescence due to DiSC (5) uptake and quenching in the cell in response to an intact 3  membrane potential. A 1 ml aliquot of cell suspension was placed in a cuvette and the desired concentration of peptide was added. Fluorescence was monitored in a PerkinElmer model 650-1 OS fluorescence spectrometer (Perkin-Elmer Corp., Norwalk, CT) at an excitation wavelength of 622 nm and an emission wavelength of 670 nm. Aliquots were removed at time intervals in order to obtain a viable count.  35  H. Macromolecular Synthesis assay and Bacterial Killing Assays Overnight cultures of S. aureus ISP67 were diluted 10" in synthetic media and allowed to grow to exponential phase (OD o of 0.3). The cultures were spun down and 60  resuspended in warm synthetic media. 15 pi of 1 mCi/ml H-thymidine, H-uridine or 3  3  3  H-histidine (at specific activities of 25 Ci/mmol, 26 Ci/mmol and 46 Ci/mmol  respectively) was added to 0.5 ml of the resuspended culture. incubation at 37°C, CP10A was added at 2X and 10X its MIC.  After 5 minutes of Samples (50 pi) were  removed at 0 (before peptide), 5, 10, 20 and 40 minutes and added to cold 5% trichloroacetic acid (TCA) (purchased from Fischer Scientific, Fair Lawn, NJ, USA) with 10-fold unlabelled precursors in order to precipitate the macromolecules.  After 40  minutes on ice and 15 minutes at 37°C the samples were collected over vacuum on Whatman 47 mm GF/C glass microfibre filters (VWR Canlab, Mississauga, ON, Canada) and washed with cold T C A .  The filters were collected and put into scintillation vials  with ReadySafe liquid scintillation cocktail (Beckman, Fullerton, CA) and counted on a scintillation counter to measure precursor incorporation into macromolecules.  At the  same time points, 5 ul samples were removed from non-radioactive parallel cultures and diluted in 1 mL of buffer and plated on to LB plates with added supplements in order to obtain a viable count.  36  RESULTS  C H A P T E R O N E : Antimicrobial Activity and Structure-Activity Relationships A . Introduction  In this study, the antimicrobial activity of a group of peptides was determined on a range of Gram positive strains. Most previous work had relied on the activity of peptides on a few Gram positive strains. By using a large panel of bacteria, more information on general peptide activity and specificity was gathered. This provided information on the peptides that was required for further mode of action and structure/activity studies. However, an initial verification of the method used to determine antimicrobial activity was required. A variety of methods have been used to assess the antimicrobial activity of peptides and their minimal inhibitory concentrations (MICs).  Methods that involve the use of  solid media to determine inhibition of growth in the presence of peptide may be limited by peptide diffusion rates through the agar or agarose media. Methods that involve the use of broth dilution are limited in that they provide only an approximate MIC value and not a bactericidal concentration value.  Furthermore, the MIC value may be easily  overestimated due to growth of a few surviving bacteria. Here, the broth micro- dilution assay of Amsterdam (Amsterdam, 1996), modified for peptides, was used to determine the relative activities of various peptides. A number of variables which could influence the assay results were investigated in order to optimize the assay.  Some of these  variables included the age of peptide, type of storage solution, type of multi-well plate  37  and type of media used.  An alternative assay for MIC determination, the agarose  method, was used and the results compared to the broth dilution method. Many studies have investigated the effects of salt on the activities of cationic peptides. For example, the cathelecidin peptides CAP18 and SMAP29 have been shown to be salt-insensitive (Travis et al., 2000), whereas defensins are inactive in the presence of salts (Goldman et al, 1997). Here, salts were included in the MIC assay broth in order to mimic the ionic conditions of a physiological environment. As a result, the effects of the presence of salts on the activity of the peptides from various structural classes were investigated. Another aspect of peptide activity is the rate at which bacterial viability decreases upon exposure to a peptide. Killing kinetics were determined using the viable plate count method. Peptides from different structural classes were compared in this manner. As well, the effect of CCCP, which is an energy uncoupler, on the antimicrobial action of the various peptides on S. aureus was studied.  Previous studies have shown that CCCP  protected E. coli from the antimicrobial action of indolicidin (Falla et al., 1996). This was attributed to a peptide requirement for electrical potential for antibacterial action. Insight into peptide-structure and activity relationships is important and necessary in order to design future antimicrobials for therapeutic use. For this reason, structureactivity relationships for one peptide, indolicidin, were investigated in this study. Indolicidin was chosen because of its short and unique amino acid sequence; it contains the largest percentage of tryptophan of any known peptide or protein (Selsted et al., 1992).  In order to determine the importance of specific amino acids and peptide  38  properties of indolicidin, variants were created and the activities compared using the broth dilution method.  Table 3: M I C s of indolicidin, C P 1 1 C N and gramicidin in previously prepared water solution (old), freshly prepared water solution (new) and freshly prepared acetic acid/BSA solution (aa/bsa) in the presence of L B no salt and Mueller-Hinton broth. The results shown are the median values from 3 separate experiments.  MIC ng/ml Indolicidin (old)  Indolicidin (new)  LB  32  MH  Indolicidin  CP-  CP-  CP-  Gram-  Gram-  Gram-  11CN (new)  11CN  (aa/bsa)  11CN (old)  (aa/bsa)  icidin (old)  icidin (new)  (aa/bsa)  32  32  16  16  16  16  8  8  8  8  8  16  16  16  16  16  16  LB  4* (>64)  8* (>64)  8* (>64)  4*  4*  4*  (>64)  (>64)  (>64)  16  16  16  MH  8* (>64)  16* (>64)  16* (>64)  4*  4*  4*  (32)  (16)  (16)  32  16  32  LB  4  4  4  16  8  8  2  2  2  MH  4  8  8  16  8  8  2  2  2  LB  2  4  4  4  4  4  2  1  1  MH  2  2  2  2  2  2  2  1  1  Bacteria and medium  icidin  E.coli  P. aeruginosa  S. aureus  S. epidermidis  * = concentration resulting in approximately 50% of bacterial growth compared to the control without peptide B. Verification of M I C Method Initially, I wanted to determine how certain variables could influence the results of the broth dilution assay. Table 3 shows the M I C s of peptides of different ages and in different holding solutions. The results shown are the median values from 3 experiments  39  done on different days. The MIC values recorded in this thesis are the concentrations at which no growth could be seen with the naked eye, unless otherwise specified.  A  difference of one two-fold dilution value is not considered a significant difference. The intrinsic variability of the assay is generally never more than one two-fold dilution, and this was the case with all MIC data reported in this thesis. The results show that there was no significant difference in MICs between old solutions of peptide in water that had been frozen and re-thawed multiple times, and fresh solutions of peptide in water. As well no significant difference was seen with fresh peptide solutions made up in acetic acid/BSA solution. This solution is thought to minimize peptide aggregation and peptide adhesion to the sides of the multi-well plate. With the exception of indolicidin with E.coli, no significant differences were seen between L B and M H broth. Indolicidin had a four-fold higher MIC in the presence of LB-no salt than in M H . The agarose method requires addition of peptide to molten agarose. This is followed by spotting a known concentration of bacteria on the agarose surface.  The  concentration of peptide at which little or no colony growth was observed was designated the MIC. I was interested in how the results of this approach compared to that of the broth dilution assay.  Table 4 shows the MICs obtained with the agarose method  compared to the broth dilution method.  With the exception of the MICs on 5.  epidermidis and the MICs of gramicidin on Gram positive bacteria, the values obtained with the agarose method were all above 64 ug/ml, significantly higher than those obtained with the broth dilution method. Thus no correlation was seen between the two methods. Because the agarose method was much more labor-intensive, the broth dilution method was used.  40  Table 4: MICs (ug/ml) of peptides using the agarose method (A) and the broth dilution method (B). The results are the median values from 3 separate experiments.  CP11CN  Indolicidin  E. coli S. typhimurium S. typhimurium Defensin sens. P. aeruginosa H103 S. epidermidis E. faecalis S. aureus 25923  S. aureus K147  S. aureus SAP0017  Gramicidin  A  B  A  B  A  B  >64  32  >64  16  >64  8  >64  >64  >64  >64  >64  16  >64  16  >64  4  32  8  >64  8* >64  >64  >64  16  4  4  16  8  0.5  1  >64  >64  >64  >64  2  4  >64  4  >64  64  2  2  >64  >64  >64  >64  1  2  >64  64  >64  >64  1  2  4* >64  * = concentration resulting in approximately 50% of bacterial growth compared to the control without peptide  C. MICs of Peptides on Gram Positive Bacteria In order to determine if certain peptide properties were required for activity against Gram positive bacteria, a variety of different cationic peptides was used in the MIC assays. The amino acid sequences and characteristics of the peptides used in this study are shown in Table 1.  The MICs of the peptides for a large number of Gram  positive bacteria are shown in Table 5. The results shown are the median values of 3 separate experiments. CP26 was the least effective peptide, with MICs of 16 ug/ml or  41  greater, for all bacteria except Corynebacterium xerosis. CP29, an a-helical peptide closely related to CP26, had MICs 2-8 fold better than those for CP26 against most bacteria.  CP11CN was previously shown to be more active than-its parent peptide  indolicidin against Gram negative bacteria (Falla and Hancock, 1997), but had equivalent or worse activity against Gram positive bacteria. However, replacing the three proline residues of indolicidin with alanines resulted in a peptide (CP10A) that was more active against most Gram positive bacteria, and was the only indolicidin variant with appreciable activity against Entereococcus faecalis.  In general, CPI OA and the  linearized bactenecin Bac2A-NH had the best activity against Gram positive bacteria. 2  The MICs of the peptides for methicillin-resistant S. aureus (MRSA) and vancomycinresistant S. haemolyticus were not significantly different than those for the parent strains. In a separate study, MICs of indolicidin and CP-11CN were done on strains of Streptococcus pneumoniae in Todd Hewitt broth and incubated under low C 0 conditions. Indolicidin had MICs of 16 to 32 ug/ml whereas CPI 1CN did not have MICs below 128 ug/ml. All MICs were repeated a minimum of 3 times.  D. Effect of Media and Cations on MICs MICs were performed in the presence of different media and salts in order to determine any effects of these two variables on the activity of peptides on Gram positive and Gram negative bacteria. MIC experiments were done in the presence of L B normal salt, L B no salt and Mueller-Hinton. Results showed that the media composition caused no significant difference in the MICs (data not shown). L B normal salt contained 5 g/L NaCl. In order to be consistent, I continued to use L B no salt (termed LB throughout) as  42  2  Table 5: Broth Dilution MIC Values (ug/ml) of Cationic Peptides for various Gram positive bacteria. The results are the median values from 3 separate experiments.  MIC (ng/ml) Species  Strain  Staphylococcus A T C C aureus  25923  Staphylococcus SAP0017 aureus  MRSA  Staphylococcus clinical haemolyticus  isolate  Staphylococcus vancohaemolyticus  mycin  isolate  Enterococcus  ATCC  faecalis  29212  Listeria monocytogenes  CP29  Indol-  CP11-  CP 10-  icidin  CN  A  Bac2A  64  8  8  16  4  4  64  8  8  16  4  16  64  8  2  2  2  1  >64  16  1  1  1  1  32  16  4  8  2  1  >64  64  >64  >64  8  2  16  4  4  16  1  0.25  16  8  4  8  1  2  1  2  0.5  0.5  1  0.25  R  Staphylococcus clinical epidermidis  CP26  N C T C 7973  Streptococcus  ATCC  pyogenes  19615  Corynebacterium xerosis  lab strain  43  the media in this assay. The MICs of the same peptides seen in Table 5 in the presence of L B and different salts are seen in Table 6. The results shown are median values of 3 separate experiments. The MICs were done on S. aureus and E. coli. As was seen previously, the MICs of all peptides regardless of structural class increased 4-fold or more in the presence of 5 m M M g C l and 5 mM CaCl on E. coli. However, with the exception of 2  2  Bac2A-NH , the peptides were not significantly affected by the presence of divalent 2  cations when tested on S. aureus. When monovalent cations in the form of NaCl and KC1 were added to the media, the peptides were more affected when tested on S. aureus. There was a four-fold or greater increase in MICs of all peptides tested in the presence of KC1. NaCl had less of an effect than KC1, generally increasing the MICs by 2- to 8-fold, with the exception of Bac2A-NH , which was the peptide most affected by all the salts 2  tested. The a-helical peptides CP26 and CP29 were not affected by the presence of either KC1 or NaCl when tested on E. coli. Indolicidin, the indolicidin variants and Bac2A-NH  2  exhibited 2 to 4-fold increases in MIC in the presence of monovalent cations when tested on E. coli.  44  Table 6 : MICs of peptides for E. coli and S aureus in the presence of LB and salts. The results are the median values from 3 separate experiments.  Strain + additive S. aureus + 5 mM Mg++ + 5 mM Ca++ + 100 mM K+ + 100 mM Na+ E. coli + 5 mM Mg++ + 5 mM Ca++ + 100 mM K+ + 100mMNa+  CP26 64 64 >64 >64 >64 1 16 64 1 1  CP29 8 8 16 32 16 2 8 16 2 2  MIC (ug/ml) Indolicidin CP11CN 4 8 4 16 32 8 64 >64 64 8 16 16 >64 >64 >64 >64 64 32 32 64  CPI OA 4 4 4 16 8 16 >64 >64 32 32  Bac2A 8 32 >64 >64 >64 8 >64 >64 32 16  E. Bacterial Killing Assays E.L Killing of Log Phase Bacteria The killing kinetics of the different peptides were compared using a bacterial viability assay.  The results shown here were the average values from 2-4 separate  experiments, with the numbers normalized to log 6 at time point zero.  The assay  variability, inevitable due to slightly different inoculum amounts and the nature of the biological assay, did not exceed 20% and was generally less than 10%.  Killing curve  data revealed more than approximately 3 log orders of killing of S. aureus over 90 minutes with 10-fold the MIC of indolicidin (80 ug/ml), CP11CN (160 ug/ml), Bac2AN H (40 jig/ml) and CP29 (80 ug/ml) (Fig. 1). Most killing occurred within the first 10 2  to 30 minutes, and CPI OA resulted in between 4-5 log orders of killing of S. aureus in this time frame.  Control cultures with no peptide added showed continued bacterial  45  growth (data not shown) and this was consistent for all the killing experiments in this thesis. The same peptides also caused an approximate decrease of between 3 and 4 log orders for S. epidermidis (Fig. 2). However, 10-fold the MIC of Bac2A-NH (20 ug/ml) 2  did not cause significant killing of E. faecalis over 90 minutes (Fig. 3), and 10-fold the MIC of CPI OA (80 ug/ml) resulted in 1 to 2 log orders of kill in the same time frame. Consistent with the MICs, E. faecalis was more resistant to these cationic peptides.  46  0 0  30  60  90  Time (min) Figure 1: Killing of S. aureus (as determined by cfu counts) by 160 ug/ml CP11CN (+), 40 ug/ml Bac2A-NH ( A ) , 80 ug/ml CP29 (o), 80 2  ug/ml indolicidin (V), and 40 ug/ml CPI OA (•) in L B broth. The results shown are the average of 2-4 separate experiments. The assay variability did not exceed 20% and was generally less than 10%.  47  E.2. Killing Assays in the Presence of CCCP In order to see if the energy uncoupler CCCP had an effect on the action of these peptides, killing curves of S. aureus were done in the presence of different concentrations of CCCP.  Previous work revealed that CCCP protected E. coli from the action of  indolicidin (Falla et al., 1996).  The data shown are the average of 2 separate  experiments. Although the variability of this assay at some time points was up to 2 log orders different, this was seldom and the trends seen (ie. bacterial protection or enhanced killing) were reproducible and the graphs shown are representative of these trends. Interestingly, indolicidin and its derivatives CP10A and CP11CN had different activities in the presence of CCCP.  When lOuM and 50 uM of CCCP was added, there  was no significant difference in killing seen from indolicidin in the absence of CCCP (Fig. 4).  However, the addition of lOOuM of CCCP caused a significant increase in  killing effects over the time period of 90 minutes.  All concentrations of CCCP were  bacteriostatic in control experiments (data not shown), indicating that CCCP at lOOuM enhanced the killing by indolicidin. In contrast, S. aureus was protected from the action of CP11CN at all three concentrations of CCCP used (Fig. 5). complicated response to CCCP (Fig. 6).  CP10A had a more  At lOuM CCCP, there was significant  protection from killing by CPI OA. However, at 50uM there was significant antibacterial enhancement. When the concentration was further increased to lOOuM, protection was again seen. However, not to the same extent as at lOuM. CP29, the a-helical peptide used in these experiments, was found to have better activity with CCCP at all 3 concentrations used (Fig. 7), with killing increasing at higher concentrations of CCCP. All peptides were used at 10-fold the MIC, except CP10A, which was at 5-fold the MIC.  48  0  30  60  90  Time (min) Figure 2: Killing of S. epidermidis (as determined by cfu counts) by 80 ng/ml CP11CN (+), 10 ng/ml Bac2A-NH (A), and 160 ug/ml CP29 (o) in L B broth. The results shown are the averages from 2-4 separate experiments. The assay variability did not exceed 20% and was generally less than 10%. 2  49  0 0  30  60  T i m e (min)  Figure 3 : Killing of E. faecalis (as determined by era counts) by 20 ug/ml'Bac2A-NH (+) and 80 ug/ml CP10A ( A ) in L B broth. The results shown are the averages from 2-4 separate experiments. The assay variability did not exceed 20% and was generally less than 10%. 2  50  90  0  30  60  Time (min) Figure 4: Killing of S. aureus (as determined by cfu counts) by 80 ug/ml indolicidin in the absence of CCCP (+) and the presence of 10 uM CCCP ( A ) , 50 uM CCCP (o) and 100 uM CCCP (•) in L B broth. The results shown are the averages of 2 separate experiments. The variability of this assay at some time points was up to 2 log orders, however this was seldom and the trends seen were reproducible and the graphs shown are representative of these trends.  51  90  0  30  60  Time (min) Figure 5: Killing of S. aureus (as determined by cfu counts) by 160 ug/ml CP11CN in the absence of CCCP (+) and the presence of 10 uM CCCP ( A ) , 50 uM and 100 uM CCCP (•) in L B broth. The results shown are the averages of 2 separate experiments. The variability of this assay at some time points was up to 2 log orders, however this was seldom and the trends seen were reproducible and the graphs shown are representative of these trends.  52  90  0  30  60  90  Time (min) Figure 6: Killing of S. aureus (as determined by cfu counts) by 20 ug/ml CPI OA in the absence of CCCP (+) and the presence of 10 uM CCCP (A), 50 uM CCCP (o) and 100 uM CCCP (•) in L B broth. The results shown are the averages of 2 separate experiments. The variability of this assay at some time points was up to 2 log orders, however this was seldom and the trends seen were reproducible and the graphs shown are representative of these trends.  53  0  30  60 Time (min)  Figure 7 : Killing of S. aureus (as determined by cfu counts) by 80 ug/ml CP29 in the absence of CCCP (+) and the presence of 10 uM CCCP (A), 50 uM CCCP (o) and 100 uM CCCP (•) in LB broth. The results shown are the averages of 2 separate experiments. The variability of this assay at some time points was up to 2 log orders, however this was seldom and the trends seen were reproducible and the graphs shown are representative of these trends.  54  90  F. MICs of Indolicidin Variants Peptide variants based on indolicidin and CP11CN (designed by Dr. Tim Falla, (Falla and Hancock, 1997)) were synthesized and their activities compared using the broth microdilution method. The amino acid sequences of the peptide variants are shown in Table 7.  The tryptophan residues of indolicidin were replaced with phenylalanines,  resulting in CPlOFa.  The IL residues of CPlOFa were replaced with FF, resulting in  CPlOFb. CP 10W is indolicidin with IL replaced with WW. Similar to CP10W, CPI OR is identical to indolicidin with the exception of the first 2 residues, which in this case contains the positively charged amino.acid arginine. These peptides were designed to determine the roles and importance of tryptophan, isoleucine and leucine in indolicidin. CP10A, as described earlier, has the proline residues replaced with alanine residues. CP10ARR is CP10A with the first two residues replaced by two arginine residues. Similar amino acid replacements were done using CP11CN as a template, resulting in CPI IF. CPI IR, however, is a peptide resulting from the replacement of the first 4 amino acids of CPI 1CN with RWRR. In order to increase the hydrophobicity of the N-terminus of CPI 1CN, an additional leucine residue is added along with an acetyl group on the N terminal isoleucine.  CP11AR is CP11R with the prolines replaced with alanines. All  peptides were amidated at the C-terminus.  55  Table 7: M I C s of indolicidin variants on various Gram negative and Gram positive bacteria. The results are the median values from 3 separate experiments.  M I C (ng/ml) Peptide  S.typhiS.typhimurium murium def  E.c oli 32  A m i n o acid sequences  Indol  ILPWKWPWWPWRR-NH  lOFa  ILPFKFPFFPFRR-NH2  lOFb  FFPFKFPFFPFRR-NH  10W  WWPWKWPWWPWRR-NH  10R  RRPWKWPWWPWRR-NH  P.aerug  -inosa  32  16  8  32  64  16  >64  16  32  16  16  64  >64  16  >64  2  4  64  2  2  10A  ILAWKWAWWAWRR-NH2  8  16  8  16  10ARR  RRAWKWAWWAWRR-NH  4  16  2  64  8  32  4  32  32  32  4  32  16  64  2  2  11CN  2  2  ILKKWPWWPWRRK-NH  2  2  2  11F  ILKKFPFFPFRRK-NH2  11R  RWRRWPWWPWRRK-NH  11L  acetyl-LLKKWPWWPWRRKNH  8  32  8  16  RWRRWAWWAWRRK-NH  4  64  2  >64  2  2  11AR  2  M I C (ng/ml) Peptide  S. aureus E. S.epidermidis S.aureus met faecalis  A m i n o acid sequences  Indol  ILPWKWPWWPWRR-NH2  4  8  8  64  lOFa  ILPFKFPFFPFRR-NH  8  16  >64  >64  lOFb  FFPFKFPFFPFRR-NH  4  8  16  32  10W  WWPWKWPWWPWRR-NH  8  8  8  64  10R  RRPWKWPWWPWRR-NH  1  4  8  64  10A  ILAWKWAWWAWRR-NH2  2  4  4  8  10ARR  RRAWKWAWWAWRR-NH  1  4  16  >64  ILKKWPWWPWRRK-NH2  8  16  16  64  1 IF  ILKKFPFFPFRRK-NH2  8  .32  64  >64  11R  RWRRWPWWPWRRK-NH  1  4  8  >64  11L  acetyl-LLKKWPWWPWRRK  2  8  16  32  11AR  RWRRWAWWAWRRK-NH  1  8  16  >64  11CN  2  2  2  2  2  2  2  56  The MICs (Table 7) were done on a panel of 4 Gram negative and 4 Gram positive bacteria in order to see potential differences in selectivity. The results are the median value of 3 experiments. In general, it was found that the variants with a higher positive charge (additional arginine residues) and those with proline to alanine substitutions had the best activities against both the Gram negative and the Gram positive bacteria tested. CPI OR had MICs 2- to 8-fold better than indolicidin, with the exception of S. typhimurium.  CPI OA was more active against S. typhimurium and E. faecalis than  CPI OR, but was 4- to 8-fold less active against defensin-sensitive S. typhimurium and P. aeruginosa.  CP10ARR was not significantly different than CP10R in MICs, with the  exception of being 4-fold better against S. typhimurium and 32-fold worse against P. aeruginosa.  CP11R had better activity than CP11CN against P. aeruginosa and the  Gram positive organisms (with the exception of E. faecalis).  This activity against  P.aeruginosa was lost upon substitution of prolines with alanines, resulting in CP11AR. With the exception of P. aeruginosa and E. coli, where CP11AR had 4-fold better activity, CP11AR and CP11R were not significantly different in activity. Peptides with tryptophan to phenylalanine substitutions or alterations in the first two amino acids were in general not significantly different in peptide activity than the peptide templates. Exceptions to this included a loss in activity of CPlOFa on P. aeruginosa and MRS A and of CP10W on P. aeruginosa. CPI IF was also not significantly different than CP11CN with the exception of a 4-fold increase in MIC against E. coli and MRSA. Aside from a 4-fold lower MIC against S. epidermidis, CP11L was not significantly different from CPI 1CN. These results taken together indicated that an increase in positive charge or a  57  significant change in structure (eg. pro-> aia) affects the activity of these peptides and that the N-terminal hydrophobic residues did not contribute a significant role in antibacterial activity.  G . Summary The broth dilution method used for MIC determination was reliable and not significantly affected by factors such as media or age/type of peptide solution. The MICs of peptides of various structural classes revealed some improved anti-Gram positive peptide variants such as CP10A and Bac2A-NH . 2  In addition, an a-helical peptide  (CP26) was found to have significantly reduced antimicrobial activity on Gram positive bacteria compared to Gram negative bacteria. The addition of salts had a significant effect on the activity of the peptides, but the effects depended on the salt, peptide and type of organism involved.  In general, M g C l  2  and CaCl  2  had greater effects on the  antimicrobial activity against Gram negative bacteria and KC1 and NaCl had greater effects on the activity against Gram positive bacteria. Kill assays showed that at 10-fold the MIC most peptides, regardless of structural class, had very similar killing kinetics over a 90 minute period. The peptides generally killed between 2 to 4 log orders of S. aureus and S. epidermidis in the first 10 to 30 minutes.  CCCP acted as an antagonist (in the case of CP11CN) and acted in synergy  with other peptides depending on the concentration used.  CCCP  alone had a  bacteriostatic effect on S. aureus. The antimicrobial activity of indolicidin variants showed that increasing the positive charge of the peptide often increased its activity, and that the tryptophan residues  58  as well as the first 2 hydrophobic residues of indolicidin did not play as important of role in antimicrobial action.  59  C H A P T E R T W O : Interaction and Binding of Peptides with Lipidic Systems A. Introduction Many studies have been done with liposomes and detergents in order to determine the interaction of peptides with the cytoplasmic membrane. Some investigators believe that specific  peptide-lipid interactions  explain peptide  selectivity  of prokaryotic  organisms over eukaryotic organisms, since eukaryotic membranes contain sterols and more neutrally charged lipids than prokaryotes (Matsuzaki et al., 1995). There has been speculation that this may also contribute to selectivity of some peptides for certain types of bacteria (ie Gram positive or Gram negative). Bacteria such as E. coli and S. aureus contain  phosphotidylethanolamine  (neutrally  charged),  phosphotidylglycerol  and  cardiolipin (negatively charged) at varying levels in their cytoplasmic membranes. Suitable methods for comparing the interaction and secondary structure of peptides in different membrane environments include C D and fluorescence spectroscopy.  In this  study we examined the level and type of structure of peptides with different primary structures and activity profiles. These experiments would also determine if a correlation existed between structure and peptide design.  The tryptophan fluorescence technique  was used to investigate the amount of peptide interaction and binding with lipids of various compositions, as well as the approximate position of the peptide when bound to lipid.  These experiments were done with liposomes containing a mixture of POPC  (neutrally charged) and POPG (negatively charged) as the model for the bacterial membrane. POPC was used as the neutral lipid instead of POPE because of previously established protocols and the tendency of POPC to form stable liposomes and POPE to form hexagonal phase lipid. A correlation (or lack of correlation) between the peptides'  60  structure and binding affinity and the peptides' activity would give insight into the mode of action of these peptides on bacteria.  B. Circular Dichroism Spectroscopy B.l. Suspected a-helical Peptides and Variants In previous studies our lab had examined the antimicrobial activity of the cecropin-melittin hybrid C E M E and its variant C E M A .  However, the secondary  structures of these peptides had not been determined and analyzed with respect to activity. The C E M E sequence was modified to increase the a-helical content in the first 14 amino acids, resulting in CP29, using an Edmundson helical wheel projection (Schiffer and Edmundson, 1967). This involved inserting hydrophilic amino acids at residues 4,-8, 10, 11 and 14, whereas amino acids 6 and 9 were changed to hydrophobic residues. The maintenance of cationic charge was achieved by placing lysine residues at positions 8 and 14. Another related peptide, CP26, had the same first 10 residues as CP29, but the C-terminus was predicted to be more hydrophilic with the addition of an extra positively charged lysine. CP201 was a peptide that was very similar to CP29, with the exception of the inclusion of a flexible hinge region consisting of two glycines in the center of the peptide. CP208 was similar to CP26, with four amino acid replacements, including the replacement of the tryptophan at residue number two with a lysine. In general, these peptides had only slight differences in charge, length and hydrophobicity (Table 1). CD spectra were measured in 10 mM sodium phosphate buffer in the presence and absence of POPC:POPG (7:3) liposomes, as well as in the membrane-mimicking  61  environments provided by the addition of SDS and trifluoroethanol (TFE; also considered a helix-inducing solvent). The concentrations of peptide and lipid in the liposomes were 50 uM and 2 mM, respectively.  All C D spectra were collected 3 times and the results  shown are representative of these experiments.  In buffer, all peptides exhibited a  spectrum characteristic of an unordered structure. The spectra of CP26, C E M E , C E M A , CP29 and CP201 in the presence of liposomes showed the typical appearance of a-helix rich structures with minimal mean residue molar ellipticity values at 207 and 222 nm (Fig. 8). The spectrum of CP208 was essentially a random coil. Similar spectra were obtained with both 60 and 90 nm liposomes (made by extruding liposomes through 0.05 urn filters and 0.1 um filters, respectively) indicating that light scattering by the liposomes did not affect these results. An estimate of percent a-helicity in the various solutions was obtained using the K2D algorithm (Andrade et al, 1993) (Table 8). This program predicted that the peptides contained only a-helix or random coil secondary structures with no p-sheet structure. In general, CP29 appeared to have the most cc-helical structure in liposomes (approximately 50%), C E M A , CP26 and C E M E had about one third a-helix, whereas CP201 and especially CP208 failed to become substantially ahelix in the presence of liposomes. Most researchers have not examined a-helicity in liposomes but rather in the so-called membrane-mimicking solvents T F E and SDS. Generally speaking, these solvents also revealed that the peptides were a-helical, but the relative a-helicities varied between the different peptides. For example, in SDS CP208 was as a-helical as was CP29. This stresses the importance of microenvironment in peptide structure formation.  62  Table 8 : Alpha-helicity of peptides in various environments as assessed by circular dichroism spectroscopy interpreted according to the K2D algorithm Condition  CP26 6 26-35 42 30  Phosphate buffer Liposomes SDS" 50% T F E 3  a  b  C  CP29 8 50-57 23 20  CP201 8 20 13 ND  C  % a-helix CP208 7 8 26 ND  CEMA 4 17-33 50 32  CEME 7 29-30 35 25  The first number is for 90 nm liposomes and the second for 60 nm liposomes SDS at a ratio of 40:1 with peptide ND = not determined  B.2. Comparison of Three CD spectroscopy Analysis  Programs  There are many programs available for determining secondary structure content based on C D spectroscopy. Here, the K2D (Andrade et al., 1993), LINCOMB (Perczel et al., 1992) and Variable Selection (Dicroprot) (Manavalan and Johnson, 1987) programs were used with the peptide C E M A in order to determine a-helical content and to determine the reliability of the K2D program when compared to older, more frequently used (and more labor-intensive) programs. Table 9 shows the percentage values obtained  Table 9 : Alpha helical content for C E M A , as predicted by three C D spec analysis programs - K2D (K), LINCOMB (L) and Variable Selection (V) % a-helical content buffer  T F E (50%)  SDS (1:10)  SDS (1:40)  Liposomes (90 nm)  Liposomes (60 nm)  K  L  V  K  L  V  K  L  V  K  L  V  K  L  V  K  L  V  5  28  0  32  22  40  36  35  37  42  35  39  15  10  19  30  12  17  with these programs with C E M A under the different conditions used.  In general, the  values obtained were consistent between the 3 programs. The one obvious inconsistency  63  is that the LDSfCOMB program predicted 28% helicity in the presence of buffer, when the scan appeared to be representative of a random coil. In the 60 nm liposomes the K2D program predicted approximately 2-fold higher helicity than the other 2 programs, however in the presence of T F E , SDS and 90 nm liposomes the values obtained were consistent between all three. These results gave credibility to the K2D program and the values obtained with the other peptides designed to be a-helical.  B.3. Indolicidin and Variants The peptide indolicidin, which has a high percentage of proline and tryptophan residues, was modified to have a greater positive charge by the replacement of one PW motif with a lysine, and the addition of a lysine at the C-terminus. The resulting peptide, CPI 1CN, was found to have an increased activity against Gram negative bacteria and an increased affinity for binding to LPS (Falla and Hancock, 1997). Previous work with indolicidin in our lab revealed an interesting C D scan in the presence of POPC:POPG (7:3) liposomes. This spectra indicated that indolicidin may contain elements or characteristics of a poly-L-proline type U helix, a relatively rare secondary structure. Another C D study indicated that indolicidin contained P-turns (Ladokhin et al, 1999). My thesis research went further by looking at the secondary structure of indolicidin and its variants CPI 1CN and CPI OA in the presence of liposomes of different composition as well as some detergents in order to gain insight into structural changes and binding affinities.  In addition, the peptides were prepared and extruded with the liposomes, as  opposed to the passive addition of the peptide to liposomes that was done previously.  64  Wavelength[nm] Figure 8: C D spectra of cationic peptides in the presence of 90 nm liposomes (POPC-POPG, 7:3) and CP26 in buffer (random coil). The peptides are represented as follows: CP201, round dots; CP208, solid lines; C E M A , dashdots; C E M E , long dashes; CP26, dashes; and CP29, square dots. All samples contained 50 uM peptide in 10 m M sodium phosphate buffer (pH 7.0). The concentration of lipid was 2 mM.  65  The C D spectra of free and bound indolicidin, as well as bound CP11CN, are shown in Fig.9.  All C D spectra were collected 3 times and the results shown are  representative of these experiments. In aqueous solution the spectrum was characterized by a broad negative band at -200 nm. This minimum is normally characteristic of an unordered structure, but has also been observed for the poly-L-proline II helix (Bovey and Hood, 1967) and B-turns (Perczel, 1991).  Upon binding to liposomes, a strong  negative band appeared at -228 nm with a relative maximum at -218 nm. The band at 228 nm has been attributed to the tryptophan side chains (Ladokhin et al,  1997;  Ladokhin et al., 1999). Below 200 nm the light scatter caused by the large lipid vesicles obscured the data (Wallace and Mao, 1984). The spectrum of CPI 1CN was very similar to indolicidin, with the exception of a notable shift to positive ellipticities. The spectra of indolicidin in liposomes of varying composition are shown in Fig. 10. The spectra of indolicidin in POPC liposomes (neutral), POPG liposomes (negatively charged) and mixed liposomes (POPCPOPG 7:3) were very similar. The minima in the presence of mixed liposomes were slightly more negative than in the others, indicating a potential increased affinity for mixed liposomes. Also in this figure is the spectra in the presence of lyso-PC, a micelle-forming lipid similar to the detergent DPC used in 2DN M R structural studies. The scan was similar to those in the presence of liposomes with the exception of the maximum at ~220nm being shifted to a position that was a few nm lower. This could potentially be explained by the fact that the light scatter seen at lower wavelengths with lipids was absent in the presence of detergent and thus this maximum could be more clearly defined. In Fig. 11 the scan of indolicidin in the presence of SDS is shown alongside those in buffer and lyso-PC. There was no major difference between  66  the scans in the two detergents, although a more distinct minimum at 228 nm and a less distinct minimum around -200 nm in the presence of SDS was observed.  These data  indicated that, as compared to SDS, the conformation of indolicidin in DPC/Lyso-PC more closely resembled the conformation of the peptide bound to liposomes. Fig. 12 shows the C D scans of CP11CN in the presence of the different liposomes. Interestingly, the scan in the presence of POPC liposomes showed that there was either a less defined structure or a lower binding affinity of CP11CN with the neutrally charged lipid. It appeared that the presence of some negatively charged lipid was necessary for the complete binding and/or structure formation of CP11CN, as shown by the nearly identical scans in the presence of POPG and mixed liposomes. This seemed consistent with the fact that CP11CN was more positively charged than indolicidin, resulting in a higher binding affinity for negatively charged LPS and potentially a lower affinity for neutral lipids. The scan of CP11CN in the presence of lyso-PC was very similar to the scan of indolicidin in the same detergent. The positive ellipticity at -220 nm seen in the presence of lipids was below zero in lyso-PC, thus making it more similar to indolicidin. The C D spectra of free and lipid-bound CP10A are shown in Fig. 13. The peptide when bound to POPC, POPG, P O P C P O P G (7:3), and DPC exhibited similar spectra with minima at approximately 207 nm and 218 nm, and a maximum at approximately 230 nm. The negative bands were more intense in the presence of POPC and mixed liposomes.  There was also a maximum at approximately 195 nm seen in the presence of  DPC and buffer that could not be observed in the presence of liposomes, due to interference by light scattering below 200 nm (Wallace and Mao, 1984). The double minimum at 207 and 218 nm, along with the maximum at 195 nm, were indicative of a-  67  helical structure (Holzarth and Doty, 1965). The maximum at 230 nm was likely due to the presence of tryptophan residues (Woody, 1994). The similarity of C D curves verified that the structure of CPI OA in the presence of DPC was comparable to that seen in the presence of lipids.  In aqueous solution, the C D bands at 207, 218 and 195 nm were also  observed, albeit with reduced intensity, indicating that the helical structure was also present in the absence of lipids.  However, the maximum at 230 nm was not present  suggesting a less defined tryptophan side chain orientation. These results indicated that CPI OA formed a more ordered structure in the presence of lipids.  68  2  Wavelength[nm] Figure 9 : C D spectra of indolicidin (dashes) and CP11CN (dots) in 90 nm liposomes (POPC-POPG, 7:3) and indolicidin in buffer (solid). All samples contained 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentration of lipid was 2 mM.  69  1  Figure 10: C D spectra of indolicidin in POPC and POPG (solid), Lyso-PC (dots), and mixed (7:3) P O P C P O P G (dashes). All samples contained 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentrations of lipid or detergent were 2 mM or 10 mM, respectively.  70  190  200  220  240  Wavelength[nm] Figure 11: C D spectra of indolicidin in buffer (solid), SDS (dots) and Lyso-PC (dashed). All samples contained 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentration of detergent was 10 mM.  71  250  Wavelength[nm] Figure 12: C D spectra of CP11CN in POPC (solid), Lyso-PC (dots), POPG and 7:3 POPC:POPG (dashes). All samples contained 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentrations of lipid or detergent were 2 mM or 10 mM, respectively.  72  15  Wavelength[nm] Figure 13: C D spectra of CPI OA in buffer (solid lines) and in complexes with POPC liposomes (long dash), POPG liposomes (dash) and DPC (dots). The spectra in the presence of POPC:POPG (7:3) liposomes was identical to that in POPC liposomes. All samples contained 50 uM peptide in 10 m M sodium phosphate buffer (pH 7.0). The concentrations of lipid or detergent were 2 mM or 10 mM, respectively.  73  B.4. CD Scans with LPS and LTA Lipoteichoic acid is a negatively charged polymer present in the peptidoglycan layer of Gram positive cell walls and is anchored in the cytoplasmic membrane. LPS is a negatively charged polymer extending out from the outer membranes of Gram negative bacteria. It has been shown from a dansylpolymyxin displacement assay that peptides will bind to LPS (Piers et al, 1994) and that this binding is part of the self-promoted uptake process in Gram negative bacteria. In addition, the dansylpolymyxin assay was modified to show that some peptides could also bind to L T A (Scott et al, 1999). It was hypothesized that in the presence of these polymers, C D spectroscopy would give further insight into the binding affinities and structure of these peptides upon interaction with the cell envelope.  A range of LPS amounts, from 15 jag to 500 ug in 200 ul of phosphate  buffer, was used with 50uM of the peptide CP26. The C D scans of the peptide in the presence of LPS were not different in shape from that of the scan in buffer; however, the signals were less intense (ie. the band at ~ 200 nm was less negative).  At higher LPS  concentrations there was a large amount of light scatter thus causing the signal to be too noisy for interpretation. When liposomes were added to a mixture of LPS and peptide, the CD scan changed, indicating that the peptide had become a-helical. Therefore, although binding to LPS may have occurred, no secondary structure induction was evident with LPS alone. The C D scans in the presence of L T A yielded different results to that of LPS. As seen in Figs 14, 15, and 16, L T A induced structural changes similar to that seen with liposomes or detergent. All C D spectra were collected 3 times and the results shown are representative of these experiments. CP26 and CP29 became a-helical, and indolicidin,  74  CPI 1CN and CPI OA had scans similar to those seen in liposomes. Bac2A-NH became a 2  P-sheet structure, similar to that was observed previously with liposomes (Wu and Hancock, 1999). It was possible that L T A formed micelles, which resulted in peptide structural changes similar to those seen with detergents and lipids.  75  Wavelength[nm] Figure 14: C D spectra of CPI OA in buffer (solid) and L T A (dashed) All samples contained 50 uM peptide in 10 m M sodium phosphate buffer (pH 7.0). The concentration of lipoteichoic acid used was 0.5 mg/ml.  76  Figure 15: C D spectra of indolicidin (solid), CPI 1CN (dots) and Bac2A-NH (dashed) in L T A . All samples contained 50 uM peptide in 10 mM sodium phosphate buffer (pH 7.0). The concentration of lipoteichoic acid used was 0.5 mg/ml. 2  77 I  190  200  220  240  Wavelength[nm] Figure 16: C D spectra of CP26 (solid) and CP29 (dashed) in L T A . All samples contained 50 uM peptide in 10 m M sodium phosphate buffer (pH 7.0). The concentration of lipoteichoic acid used was 0.5 mg/ml.  78  C. Fluorescence Spectroscopy  C.l. Fluorescence Spectroscopy in the Presence of Liposomes and Detergents The tryptophanfluorescenceemission scans of the peptides were analyzed in the presence of liposomes of varying composition. A shift in the wavelength maximum to a shorter wavelength (blueshift), as well as an increase in fluorescence intensity when excited at 280 nm indicated that the tryptophan residue or residues have shifted to a more hydrophobic environment (Lakowicz, 1983). Table 10 shows the wavelength maxima of the peptides as well as the controls in the presence of buffer and liposomes with different ratios of neutral (POPC) to negatively charged (POPG) lipids. All fluorescence spectra were collected 3 times and the results shown are representative of these experiments. Tryptophan alone did not display any change in wavelength maxima with liposomes. Tryptophan octyl ester served as a positive control in this experiment. A l l peptides and tryptophan octyl ester had a blueshift in the presence of liposomes, however this shift was smaller with POPC than in the presence of any POPG-containing liposomes, although this difference is more obvious for the a-helical peptides (which all have one tryptophan residue in the N-terminus) and tryptophan octyl ester. Tryptophan octyl ester had a shift of 16 nm with POPC and a further 4 to 6 nm shift in the presence of POPG. C E M E and C E M A were similar in that a shift of between 10 to 12 nm occurred with POPC and then a further 3 to 5 nm shift when POPG was present.  The largest change in maximum  wavelength observed for both of these peptides was 15 nm. CP29, which was found to have the greatest content of a-helix (as judged by C D spectrometry), underwent the  79  largest shift with POPC, a shift of 20 nm. A further 4 to 5 nm shift took place when POPG was present. This shift, which was larger than that for C E M E and C E M A , may  Table 10: Fluorescence emission wavelength maximum of the cationic peptides, tryptophan octyl ester (positive control) and tryptophan (negative control) in the presence of various liposomes. The change in max X is the difference between the emission maximum wavelength in the presence of HEPES and in the presence of POPG. PC=POPC, PG=POPG  Wavelength of Emission Maximum (nm) HEPES  POPC  PC:PG  PC:PG  PC:PG  7:3  5:5  3:7  POPG  Change in max A,  CP26  352  342  326  326  326  326  26  CEME  352  342  338  337  337  337  15  CEMA  352  340  339  337  337  337  15  CP29  352  332  328  328  327  327  25  CP11CN  354  346  344  344  344  344  10  Indolicidin  350  343  341  340  340  340  10  TOE  352  336  332  330  330  330  22  TRYP  352  352  352  352  352  352  0  indicate a greater interaction or binding of CP29 to the lipid bilayer. CP26, which has the same N-terminus as CP29, was also similar in that the greatest change in emission wavelength maximum was around 25 to 26 nm. However, CP26 demonstrated only half the shift observed with CP29 in the presence of POPC (10 nm) and then a further large shift of 16 nm when POPG was present.  In light of the fact that the positive control,  tryptophan octyl ester, had a greater shift in the presence of POPG than POPC alone, it  80  was difficult to interpret the interactions of the peptides with these lipids. However, it can be seen that there were differences between the peptides. Indolicidin and CP11CN have 5 and 4 tryptophan residues, respectively. These peptides had blueshifts similar to each other (7 to 8 nm with POPC and a further 2-3 nm with POPG).  However, this maximum change of 10 nm was significantly less than the  total change of 22 to 26 nm for CP26, CP29 and T O E .  Because of the multiple  tryptophan residues, it seems possible to conclude that at least some of the tryptophan residues were present in a hydrophobic environment.  Fig. 17 shows the scans for  indolicidin in the presence of POPC, POPG and P O P C P O P G (7:3) liposomes. There was an increase in fluorescence intensity when POPG was present. The blueshift seen in the presence of liposomes was reproducible between the experiments. Many studies with peptides were done in the presence of detergent micelles as the lipid-mimicking environment.  2D-NMR was being used to determine the tertiary  structure of some of these antimicrobial peptides in the presence of detergents, thus it was considered important to establish that the induced structure of the peptides was similar in both environments.  For this reason, fluorescence spectroscopy was performed with  indolicidin and its derivatives in the presence of DPC and lyso-PC.  Fig. 18 shows  indolicidin in the presence of these detergents. A similar blueshift of approximately 10 nm, as well as similar intensities, occurred in the presence of these detergents.  In  particular, the scans in the presence of DPC and POPC (the neutral molecules) were very similar. Fig. 19 shows the scans of CPI 1CN in the presence of liposomes and DPC. The wavelength maxima were 355 nm in buffer, 350 nm for POPC and DPC, and 344 nm for  81  500  300  350 400 Wavelength (nm)  450  Figure 17: Fluorescence spectra of indolicidin in the presence of buffer and liposomes. The samples contained 2 uM peptide and 0.5 mM lipid in lOmM HEPES buffer (pH 7.2).  7:3 mixed liposomes and POPG. In Fig. 20 the emission scans of CPI OA in the same liposomes and detergents are shown. These scans were all very similar, especially in the presence of DPC and P O P C P O P G (7:3).  The emission maximum of the peptide in  buffer alone was around 355 nm. With all lipid environments, an 8 nm blue shift in the emission maximum as well as an increase in fluorescence intensity occurred.  The  emission spectra in POPC, POPG, P O P C P O P G (7:3) and DPC were all very similar with  82  peak wavelengths of around 347 nm, indicating that the environment experienced by the peptide in lipids and DPC was comparable. This data gave us confidence that the structural elements identified in the presence of detergents were very similar to those that would be found in liposomes or in the membranes of bacteria.  C.2. Potassium Iodide Quenching To determine if the tryptophan residue(s) were accessible to the aqueous environment of the buffer, fluorescence spectroscopy in the presence of the aqueous quencher KI was done. These experiments were done with both liposomes and detergents (in the case of the indolicidins).  A Stern-Volmer plot was used to express the data, plotting KI  concentration on the X-axis against Fo/F (where Fo is the fluorescence in the absence of quencher and F is fluorescence in the presence of quencher) on the y-axis. CP26, one of the a-helical peptides with a very large blueshift in the presence of liposomes containing the negatively charged POPG (POPCPOPG 7:3, 5:5, 3:7 and POPG alone), showed no decrease in fluorescence with KI in these liposomes as seen by the virtually straight line seen on the graph (Fig. 21). However, in the presence of POPC liposomes, some quenching of the tryptophan residue occurred, although when compared the quenching of the peptide in buffer it was obvious that POPC liposomes were offering some protection to the tryptophan. Similar results were seen with C E M A (Fig. 22), a a-helical peptide with a different blueshift profile than CP26. There was somewhat better protection of the tryptophan residue in C E M A with POPC than what was seen with CP26. These results indicated that the tryptophan residue in position number 2 is deep in a hydrophobic  83  500 — \ SDS  400 300  /  L  P  C  \  H / / /  200 H 100  f — X  A  / /  300  /  DPC:SDS  >Rf V v C  /  buffer  ^>  350  400  Wavelength (nm)  Figure 18: Fluorescence scans of indolicidin in the presence of detergent micelles and buffer. The samples contained 2 uM peptide and 10 mM detergent in lOmM HEPES buffer (pH 7.2).  84  450  300  320  340  360  380  400  wavelength Figure 19: Fluorescence scans of CP11CN in the presence of detergent micelles, liposomes and buffer. The samples contained 2 uM peptide and 0.5 m M lipid or 10 mM detergent in lOmM HEPES buffer (pH 7.2).  85  500  300  320  340  360  380  400  wavelength Figure 20: Fluorescence scans of CPI OA in the presence of detergent (DPC) micelles, liposomes and buffer. The samples contained 2 uM peptide and 0.5 mM lipid or 10 mM detergent in lOmM HEPES buffer (pH7.2).  86  environment, inaccessible to the aqueous solution, and that this interaction likely requires some negatively charged lipid for greater binding. These same experiments were done with indolicidin, CP11CN and CP10A, all yielding similar results.  In the presence of all types of lipid there was insignificant  quenching by KI and this is evident in Fig. 23 for CP11CN and in Fig. 24 for CP10A. These results indicated that most or all the tryptophan residues were hidden within a hydrophobic shield. In order to determine if detergent protects tryptophan from the quenching of KI, the experiments were done in the presence of lyso-PC. As seen in Fig. 25, indolicidin was protected equally in lyso-PC and POPC, although slightly less so than in the presence of liposomes containing POPG.  This result indicated that the peptides may  interact more intimately with the negatively charged phopholipid POPG. However, this was evidence that similar environments were provided by detergents and liposomes.  C.3. Spectroscopy with Spin-Labeled Lipids In order to estimate the depth of the tryptophan residue(s) within the phospholipid bilayer, fluorescence studies were done in the presence of liposomes with incorporated spin-labels at various positions. The liposomes consisted of 10 % (2.6 umol) of the spin labeled  lipid  (phosphotempocholine,  5-doxyl  phosphocholine  and  12-doxyl  phosphocholine) with 16.4 umol POPC and 7 umol POPG. Fluorescence quenching is inversely proportional to the distance of the tryptophan residue(s) from the spin label. Fig. 26 shows the relative fluorescence intensities in the presence of the spin labels, where 100 represented the fluorescence in the absence of spin label. All spectra were  87  1.0  0.5  0  20  40  60  80  KI concentration (mM)  Figure 21: Stern-Volmer Plot. KI quenching of CP26 fluorescence in buffer (solid triangle), POPC liposomes (open triangle) and POPQPOPG liposomes (3:7, 5:5 and 7:3) and POPG liposomes (solid square). The samples contained 2 uM peptide and 0.5 mM lipid in lOmM HEPES buffer (pH 7.2).  88  100  0  20  40  60  80  KI concentration (mM) Figure 2 2 : Stern-Volmer Plot. KI quenching of C E M A fluorescence in buffer (solid triangle), POPC liposomes (circles) and P O P C P O P G liposomes (3:7, 5:5 and 7:3) and POPG liposomes (solid square and inverted triangles). The samples contained 2 uM peptide and 0.5 m M lipid in lOmM HEPES buffer (pH 7.2).  89  100  2.0  KI concentration (mM) Figure 23: Stern-Volmer Plot. KI quenching of CP11CN fluorescence in buffer ( A ) , POPC, P O P C P O P G liposomes (3:7, 5:5 and 7:3) and POPG liposomes (remaining dashed lines). The samples contained 2 uM peptide and 0.5 mM lipid in lOmM HEPES buffer (pH 7.2).  90  I  J  0  20  40  60  80  100  KI concentration (mM) Figure 24: Stern-Volmer Plot. KI quenching of CPI OA fluorescence in buffer (cross), POPC (circles), POPC:POPG liposomes (3:7, 5:5 and 7:3) and POPG liposomes (triangles). The samples contained 2 uM peptide and 0.5 mM lipid in lOmM HEPES buffer (pH 7.2).  91  -  I  2  —O  1  a I  a  1  1  0  20  9  *  —  o  n  i  40  60  80  KI concentration (mM) Figure 25: Stern-Volmer Plot. KI quenching of indolicidin fluorescence in buffer (cross), POPC:POPG liposomes (3:7, 5:5 and 7:3) and POPG liposomes (triangles) and lyso-PC and POPC (circles). The samples contained 2 uM peptide and 0.5 mM lipid or 10 mM detergent in lOmM HEPES buffer (pH 7.2).  92  100  collected 3 times and the results shown are the average of these experiments. In general, when the spin labels were on the 5 or 12 position of the hydrocarbon chain, there was a greater reduction in fluorescence than when the label was on the phospholipid head group.  This was consistent for all seven peptides studied.  The a-helical peptides  consistently had a greater reduction in fluorescence with 5-doxyl than 12-doxyl, an indication that the tryptophan residue at position 2 was not positioned too deeply within the membrane. The indolicidin peptides did not show such an obvious difference in fluorescence between 5- and 12-doxyl-PC. The 5 tryptophan residues present may have been positioned throughout this region of the hydrocarbon chain. These results suggested that on average the tryptophan side chains resided in the bilayer interface.  D. Summary Peptides designed to be more a-helical (on the Edmundson helical wheel) were found to have a higher helical content through C D spectroscopy studies. These peptides were random coils in buffer and helical in detergents and liposomes. The K2D analysis program predicted helicity in percentage values and these values were reasonably consistent with those obtained from two other analysis programs, LINCOMB and Variable Selection.  Indolicidin and CP11CN had similar C D spectra, with a notable  increase in ellipticity in the case of CP11CN.  The differences in spectra between  indolicidin in the presence of different liposomes and lyso-PC could be attributed to light scatter below 200 nm in the presence of liposomes. The scan in the presence of SDS was slightly different. The C D scans of CPI 1CN in the presence of negatively charged lipids showed deeper minima than the scans in the presence of POPC, suggesting the possibility  93  of a more defined structure in the presence of negative charge.  CPI OA was not  disordered in buffer and had a structure that included a-helical components and was thus completely different from indolicidin and CP11CN.  The C D spectra of CP26 in the  presence of LPS were no different in shape from the spectra in buffer.  C D for all  peptides in the presence of L T A showed scans similar to those seen with lipids. Fluorescence spectroscopy showed that all peptides interacted with liposomes and detergents in such a way that the tryptophan residues were buried within a hydrophobic environment and were largely inaccessible to the aqueous solution.  This interaction  appeared to be stronger with negatively charged lipids than with neutral lipids. Some small differences in wavelength and blueshift existed between the peptides. the interactions  of peptides  with liposomes  In general  and detergents were very similar.  Fluorescence spectroscopy with spin-labeled lipids suggested that on average the tryptophan side chains resided in the lipid bilayer interface.  94  CP-10A  CP11CN  Indolicidin  CP26  CP27  CP28  CP29  Figure 26: Relative fluorescence intensities of peptides in the presence of spin labeled liposomes. 100 represents the intensity in the absence of spin-labels. The black bar represents head-group labelled tempo-PC, hatched bars represent 5-doxyl PC and the white bars represent 12-doxyl PC.  95  CHAPTER THREE: Mechanism of Action on Gram Positive Bacteria A. Introduction It is widely believed that the mechanism of action of cationic peptides on both Gram negative and Gram positive bacteria involves the disruption of the cytoplasmic membrane (Duclohier et al., 1989). There has been evidence of membrane disruption by cationic peptides using model membrane systems (Silvestro et al., 1997). These studies have been interpreted as demonstrating that peptides form pores or have a "detergentlike" effect on model membranes, leading to lysis or gross membrane leakage (Cociancich et al., 1993; Epand et al., 1995). One concern in these studies is the high peptide to lipid ratios used.  Nevertheless, different theories of peptide-membrane  interactions have arisen from these studies and have contributed to a general view in the field that all peptides kill bacteria by one of these two mechanisms (Epand et ah, 1995; Perez-Paya et al., 1995). In addition, evidence of permeabilization and depolarization of the E. coli cytoplasmic membrane by cationic peptides has been shown using the ONPG assay (Friedrich et al, 1999; Lehrer et al, 1989) and the DiSC (5) assay (Wu et al, 3  1999). Membrane depolarization is the result of a dissipation of the electrical potential across the membrane resulting from the proton motive force. This results in an inside negative orientation across the membrane.  The electrical potential is necessary for  cellular processes and energy production. Using flow cytometry, Yeamen et al. (1998) showed that human neutrophil defensin-1 (HNP-1) depolarized and permeabilized the cytoplasmic membrane of S. aureus in vitro, whereas tPMP-1 did not depolarize but did permeabilize the membrane. However, these and other recent studies have shown a lack of correlation between membrane effects and antimicrobial activity at concentrations  96  around the MIC, indicating that membrane disruption may not be the sole mechanism of action (Daugelavicius et al, 2000; Wu et al, 1999; Zhang et al, 2000b). In addition, other intracellular targets have been identified (Boman et al, 1993; Couto et al, 1993; Otvos et al,  2000b).  In my thesis, electron microscopy, a cytoplasmic membrane  depolarization (DiSC (5)) assay and macromolecular synthesis studies were used to 3  define the mechanism of action of peptides of different structural classes and to determine if there is a universal mechanism of action of cationic peptides on Gram positive bacteria. It is notable that most other studies have been done with Gram negative bacteria (most commonly E. coli) and with only one peptide or class of cationic peptides. Both transmission and scanning electron microscopy were used to visualize ultrastructural changes to both the cell surface and interior. The bacteria were treated with 10-fold the MIC of the peptide in order to see an effect on a greater percentage of cells.  The cytoplasmic membrane depolarization assay uses the membrane-potential  sensitive dye DiSC (5) which distributes between cells and the medium depending on the 3  cytoplasmic membrane gradient. quenches.  Inside the cells, the dye is concentrated and self-  If the peptides cause a membrane disruption that dissipates the membrane  potential, dye is released  into the medium, resulting in increased  fluorescence.  Intracellular effects (specifically effects on D N A , R N A and protein synthesis) were investigated by quantifying the incorporation of H-labelled thymidine, uridine and 3  histidine into D N A , R N A and proteins, respectively.  In these experiments, a S. aureus  strain auxotrophic for thymidine, uridine and histidine was used.  97  B. Electron Microscopy B.l. Transmission Electron Microscopy (TEM) In order to determine a possible alternative mechanism of action of peptides on Gram positive bacteria, transmission electron microscopy was used to visualize thin sections of bacteria that had been treated with the peptide at 10-fold the MIC for 30 minutes. CP29 (80 ng/ml), Bac2A-NH (40 ug/ml), indolicidin (80 ng/ml) and CPI 1CN 2  (160 ng/ml) (Figs. 27, 28) showed similar effects on S. aureus. Laminar mesosomes (intracellular lamellar membranes) were seen arising from the septa and cell wall. Untreated control bacteria did not have detectable mesosome structures or cell lysis. The mesosomal structures caused by Bac2A-NH appeared not to be as large as those caused 2  by CP11CN and indolicidin. Thin sections of indolicidin- (Fig. 28) and KCl-treated S. aureus showed that in the presence of KC1 (which increases the MIC of indolicidin 16fold) there were significantly less mesosome structures seen (and no mesosomes seen with cells treated with KC1 only). Lysis was rarely seen (<10%), but with Bac2A-NH , 2  lysis was seen occurring at the septal site. With S. epidermidis, greater effects were seen (Figs 29, 30). CP29 (160 [ig/ml), CP11CN (80 ng/ml) and Bac2A-NH (10 ng/ml) all caused cell wall effects, which 2  included fibres extending from the cell surface.  CP11CN appeared to have the most  severe effects on the cell wall, including cell wall breaks and variability in wall thickness. CPI 1CN also resulted in the formation of mesosome structures and condensation of D N A (which was seen as light clusters within the cytoplasm) in 5. epidermidis.  Bac2A-NH  2  showed similar effects to those caused by CP11CN and, in addition, abnormal septum formation (ie. the appearance of a new septum before completion of cell division or septa  98  forming in parallel) was seen. The cytoplasmic membrane could be seen separating from the cell wall in some Bac2A-NH -treated cells, in addition to the observed formation of 2  mesosomes. CP29 appeared to cause the most lysis when compared to the other peptides, as seen by lytic debris, but intact cells showed no signs of either D N A condensation or mesosome formation. However, the overall amount of lysis was still approximately less than 10% of the total cell population. Interestingly, CP29, like Bac2A-NH , caused 2  abnormal septal wall formation as well as cell wall disintegration. Thin sections of CPlOA-treated S. epidermidis (Fig. 30) and S. aureus were prepared (Fig. 28).  The  bacteria were treated with 10-fold the MIC of CPI OA (20 ug/ml and 40 ug/ml for S. epidermidis and S. aureus, respectively) for 10 minutes and 30 minutes before fixing. These electron micrographs showed the formation of mesosomes in peptide-treated cells only. Often these mesosomes occurred around the septum. There was no difference seen between 10 min and 30 min treated cells.  In contrast with other peptide-treated S.  epidermidis cells, no nuclear condensation was seen and there appeared to be minimal cell wall effects.  As well, no apparent lysis or gross leakage of cellular cytoplasmic  contents was observed.  In contrast, CP26-treated cells, (S. aureus at 640 and S.  epidermidis at 320 ng/ml of CP26) (Fig. 30) underwent some lysis (approximately <10% and 25% for S. aureus and S. epidermidis, respectively), with most of the residual intact cells not showing any gross ultrastructural effects. mesosomes were seen. treated cells.  The exception was that a few  Again, no difference was seen between 10 min and 30 min  This is noteworthy as CP26 has the highest MICs on Gram positive  bacteria, and the killing caused by CP26 appeared to be distinctly different from that caused by the other peptides.  99  Thin sections of E. faecalis treated with 20 ug/ml of Bac2A-NH looked identical 2  to the control cells (Fig. 32).  However, after treatment with 640 ug/ml of CP11CN,  effects were seen at the septa (Fig. 32).  There appeared to be separation between the  cytoplasmic membrane and the septal wall, followed by septum breakdown as it matures.  B.2. Scanning Electron Microscopy (SEM) S E M was done in order to see what peptide effects, if any, could be observed on the outer surface of S. aureus and S. epidermidis.  CP11CN and CP29 were chosen  because of their different structures and the different effects of these peptides on S. epidermidis as seen with transmission electron microscopy. Fig. 31 shows the electron micrographs of the control (A), CPI lCN-treated (B) and CP29-treated (C) S. epidermidis. These were done at both 10 and 30 minutes. There was clearly a difference between the smooth surfaces of the control and those of the peptide-treated bacteria. CPI lCN-treated bacteria had bleb-like structures protruding out from the surface of the cell wall, and these structures were often seen arising from the septum. CP29-treated cells did not have these bleb-like structures, but the cell wall did appear to have holes and a rough surface.  These results were consistent with those seen with T E M , and were  further evidence of the different effects resulting from different peptides. The SEMs of S. aureus did not show any difference in the outer cell surface between the control and peptide-treated cells. This is also consistent with the results seen with the T E M , as the major effects observed were mesosome-like structures in the interior of the cell.  100  B  Figure 27 : Electron micrographs of untreated (A), CPI lCN-treated (B), CP29treated (C) and Bac2A-NH -treated (D) S. aureus. A l l peptides were at concentrations of 10-fold the MIC. The cells are approximately 750 nm in diameter. The arrows point to mesosome-like structures. 2  101  Figure 28: Electron micrographs of CPI OA-treated (A) and indolicidin-treated (B) S. aureus. A l l peptides were at concentrations of 10-fold the MIC. The cells are approximately 750 nm in diameter. The arrows point to mesosome-like structures.  102  A  B  Figure 29: Electron micrographs of untreated (A), CPllCN-treated (B), and CP29-treated (C) S. epidermidis. A l l peptides were at concentrations of 10-fold the MIC. The cells are approximately 750 nm in diameter.  103  A  Figure 30: Electron micrographs of Bac2A-NH -treated (A), CPlOA-treated (B), and CP26-treated (C) S. epidermidis. All peptides were at concentrations of 10-fold the MIC. The cells are approximately 750 nm in diameter. 2  104  Figure 31: Scanning electron micrographs of untreated (A), CPllCN-treated (B), and CP29-treated (C) S. epidermidis. All peptides were at concentrations of 10-fold the MIC. The cells are approximately 750 nm in diameter.  105  Figure 32:  Electron micrographs o f untreated (A), C P I lCN-treated (B), and  Bac2A-NH -treated (C) E.faecalis. A l l peptides were at concentrations o f 102  fold the M I C .  106  C. Cytoplasmic Membrane Depolarization  Cl. Development of DiSC3(5) Assay The membrane-potential sensitive dye diSC3(5), which distributes between cells and the external medium according to the electrical potential gradient, was used to determine if cytoplasmic membrane depolarization occurred in the presence of peptides. In order to ensure that the response of the dye was due solely to the change in membrane electrical potential, the amount of KC1 needed to equilibrate cytoplasmic and external K concentrations under these assay conditions was determined.  +  The potassium gradient  opposes the electrical gradient in order to maintain the membrane potential gradient Av|/ across the membrane (Letellier and Shechter, 1979). Valinomycin is an ionophore that transports K across the cytoplasmic membrane according to the magnitude of the Ai|/ and +  the potassium concentration gradient. To ensure that we are just measuring A\|/, and not the K concentration gradient with diSC 5, sufficient K +  3  +  must be added externally to  balance the internal K concentration (Harold et ah, 1970) Therefore, valinomycin (at a +  final concentration of 1 uM) was used in the presence of varying concentrations of external KC1 to observe the movement of K according to the K gradient to alter the +  +  equilibrium A y , which was manifested as a change in fluorescence intensity (Fig. 33). As was expected, there was a direct relationship between increasing KC1 concentration in the buffer and increase in fluorescence intensity as a result of the dye being released into the buffer and a decrease in self-quenching.  This relationship was linear up to a  concentration of 80mM to lOOmM KC1, at which point the increase in fluorescence intensity reached its maximum. At this concentration an equilibrium of K ions existed +  107  between the inside and outside of the cells, and the K concentration gradient no longer +  was a factor. Therefore, for the subsequent depolarization assays involving peptides, a concentration of lOOmM KC1 in the assay buffer was used.  C.2. Depolarization by Various Peptides Cytoplasmic membrane permeabilization has been implicated as the mode of action of peptides against Gram negative bacteria in part from the results of experiments using model membrane systems (Silvestro et al, 1997) and from the ONPG assay using intact cells (Falla and Hancock, 1997; Friedrich et al, 1999; Lehrer et al, 1989). The ONPG assay measures the accessibility of a normally membrane-impermeable substrate (ortho-nitrophenyl galactosidase, ONPG) to cytoplasmic P-galactosidase (Lehrer et al, 1989).  The major limitation of this assay is the use of a bulky substrate.  Generally  speaking, this assay uses concentrations well above the MIC. However, more recently, the results from experiments with the DiSC (5) assay indicated that the interaction of 3  some peptides with cytoplasmic membrane does not correlate with antibacterial activity on E. coli (Wu et al, 1999). We were interested in determining the effects of the peptides on the cytoplasmic membrane of S. aureus and therefore adapted the DiSC (5) assay used 3  previously by Wu and Hancock (Wu et al,  1999).  Fig. 34 shows membrane  depolarization, demonstrated here by an increase in fluorescence units, as a function of peptide concentration. The results shown are the averages of 2-3 separate experiments. The variability between the experiments most often did not exceed 2-3 fluorescence units, and seldom varied by 5-6 units. A value of approximately 20 fluorescence units was equivalent to complete depolarization as assessed by the addition of excess  108  gramicidin S as well as the fluorescence of the dye in the absence of cells.  All of the  peptides studied here had the ability to depolarize the cytoplasmic membrane of S.aureus,  S3 13 o  0  20  40  60  80  100  KC1 (mM) Figure 33: Depolarization of the cytoplasmic membrane of S. aureus by 1 uM valinomycin, indicated by maximum fluorescence reached, as a function of KC1 concentration in the assay buffer (5 mM Hepes, 20 mM glucose, pH 7.2).  109  120  however, peptides with different structures had different concentration:activity profiles. CP26 and CP29 completely depolarized the membrane at lower concentrations than the other peptides studied, with 50% depolarization at between 1 and 2 ug/ml. It is important to note that CP26 had very poor antimicrobial activity against S. aureus and even CP29 had a MIC well above such concentrations. Conversely, the most active peptide, Bac2AN H , did not depolarize the membrane at low concentrations, with 50% depolarization at 2  between 8 and 16 ug/ml.  Indolicidin and CP 11 C N depolarized the membrane at  relatively low concentrations, but at a maximum, only 50 to 75% depolarization was observed. Because the assay conditions included 100 mM KC1 in the buffer, MICs were determined in the presence of 100 mM KC1 (Table 6). In general, the addition of lOOmM KC1 increased the MIC of the peptides on S. aureus. The MIC of CP29 was increased 4fold from 16 to 64 ug/ml. Indolicidin and CPI 1CN both demonstrated an 8-fold increase in MIC while Bac2A-NH had a 16-fold increase in MIC. However, the MICs of these 2  peptides in the presence of KC1 were all well above the amount needed to cause permeabilization.  CP10A  almost  completely  depolarized  the  membrane  at a  concentration of 1 ug/ml (data not shown) within 5 minutes. These results were similar to those found with the a-helical peptides CP26 and CP29, which completely depolarized the membrane at low concentrations. However, CP10A alone quenched the fluorescence of the dye so higher concentrations of CPI OA gave unreliable results.  110  0 0.5  1  10  Peptide concentration (jxg/ml)  Figure 34: Depolarization of the cytoplasmic membrane of S. aureus as a function of peptide concentration, indicated by maximum fluorescence reached within 5 minutes. CP11CN ( • ) , indolicidin (•), CP29 ( T ) , CP26 (•), and Bac2A-NH2 (+). The assay buffer consisted of 5 mM Hepes and 20 mM glucose, pH 7.2. The results shown are the averages from 2-3 separate experiments. The variability between the experiments most often did not exceed 2-3 fluorescence units, and seldom varied by 5-6 units.  Ill  70  C.3. Cell Viability Assay in Conjunction with the Depolarization Assay To determine if a correlation existed between cell death and membrane depolarization under these conditions, cell viability assays were carried out in conjunction with the depolarization assay. The viability of cells taken directly from the assay tube, at one peptide concentration, is shown in Fig. 35.  The a-helical peptides (CP26 and CP29)  (Fig. 35) and indolicidins were chosen for these studies.  In Fig. 35 (A), the  concentrations of the peptides were chosen to give similar cytoplasmic membrane permeabilization profiles, resulting in a concentration of CP26 4-fold higher than CP29. However, at this concentration, CP26 caused 1 log order killing of bacteria, whereas CP29 resulted in almost four log orders of killing. More important, however, was the observation that for all peptides, 90% or more of the killing was complete at a time when either little to no depolarization (A) or less than 50% of complete depolarization had occurred (B). Indolicidin and CPI 1CN depolarized the membrane at a faster rate than the a-helical peptides. However for both types of peptides the maximum was reached by 5 minutes. Indolicidin and CPI 1CN continued to show killing after 5 minutes. There were distinctly different effects seen due to the peptides and no consistent correlation could be seen between depolarization and antimicrobial activity.  112  Time (min) Figure 35: Depolarization of the cytoplasmic membrane of S. aureus (dashed lines) as indicated by the kinetics of fluorescence intensity changes in the presence of 8 ug of CP26 ( A ) per ml or 2 pg of CP29 (•) per ml (A) and in the presence of 32 ug of indolicidin ( A ) or CPI 1CN (•) per ml (B). Shown are the levels of survival (in CFU/ml) of bacteria under the dye assay conditions (solid lines). Results are representative of 2-3 separate experiments.  113  D. Macromolecular Synthesis Studies  D.l. Characterization of the Action of Peptides on the Auxotroph S. aureus ISP 67 i) MICs Macromolecular synthesis studies were carried out with S. aureus ISP67, a strain auxotrophic for thymidine, uridine and histidine. To ensure that the auxotrophic strain ISP67 was reasonably similar to the wild type S. aureus strain A T C C 25923, (on which other studies had been done), MICs of the peptides were measured on both strains in the presence of both L B and the synthetic media (Table 11).  In general there was no  significant difference between the MICs of these peptides on the 2 strains. The MICs on ISP 67 were often one 2-fold dilution less than the MICs on the wild type, which by convention is not considered significant.  There was also no significant difference  between the results in the two media, with the exception of Bac2A-NH , which had 4- to 2  8-fold higher MICs in the presence of synthetic media.  This may indicate that  something in the synthetic media is either inhibiting or competing with Bac2A-NH . 2  Table 11: MICs of peptides against S. aureus A T C C 25923 (WT) and 5. aureus ISP 67 in the presence of LB no salt media (LB) and synthetic media (SM). The results are the median values from 3 separate experiments.  Strain WT WT ISP67 ISP67  Medium LB SM LB SM  CP26 >64 >64 64 64  CP29 8 32 8 8  MIC( ug/ml) indolicidin CP11CN 4 16 4 16 4 8 2 8  114  CPI OA 4 2 2 1  Bac2A 8 64 8 32  ii) Cytoplasmic Membrane Depolarization Assay The peptides were used in the depolarization assay described previously to determine the relative abilities of the peptides to depolarize the cytoplasmic membrane of auxotrophic S. aureus. Fig. 36 shows the maximum depolarization reached within 5 minutes at various peptide concentrations. The a-helical peptides CP26 and CP29 resulted in nearly complete depolarization at low concentrations and were once again the best permeabilizers. Indolicidin and CP11CN did not cause 100% depolarization and the amount of depolarization that took place at the higher concentrations was similar to that seen at lower concentrations.  Bac2A-NH  2  did not permeabilize well at lower  concentrations but at higher concentrations (>8 ug/ml) resulted in almost complete depolarization. These general patterns were very similar to those seen with the wild type, however, it was notable that there was higher depolarization seen at lower concentrations. This indicated that the ISP67 cytoplasmic membrane was more readily depolarized than the membrane of the wild type. The  depolarization assay was  done  under the  same conditions  as  the  macromolecular synthesis assays with the same concentrations of the peptides (2X MIC). All of the peptides permeabilized the membrane under these conditions but complete depolarization was not reached, even after 45 minutes. CP29, CPI 1CN and Bac2A-NH  2  achieved a maximum of around 50% depolarization while indolicidin depolarized membranes only up to 30%. Indolicidin and CPI 1CN achieved maximum depolarization quickly, within the first 5 minutes, whereas CP29 and Bac2A-NH maximum depolarization until after 20 minutes.  115  2  did not reach  I i i i i I  0.5  i  i  1  L  i i i i i I  i  i  i  i i I  10 Peptide concentration (|ig/ml)  Figure 36: Depolarization of the cytoplasmic membrane of S. aureus ISP 67 as a function of peptide concentration, indicated by maximum fluorescence reached within 5 minutes. CP11CN ( A ) , indolicidin (•), CP29 ( • ) , CP26 (•), and Bac2A-NH2 (+). The results are the averages from 2-3 separate experiments.  116  70  D.2. Macromolecular Synthesis Inhibition by Indolicidin and CP11CN Indolicidin and CP11CN are related peptides that are rich in proline and tryptophan. The effects of indolicidin and CPI 1CN on DNA, R N A and protein synthesis were studied. Incorporation of radiolabeled thymidine into DNA, uridine into R N A and histidine into protein was measured. In the presence of indolicidin, protein synthesis was affected within the first 5 minutes after treatment (Fig. 37), and R N A synthesis was affected within 10 minutes, with no significant difference observed between 2X and 10X the MIC. There was an effect on D N A synthesis seen at 10 minutes, however, the rate of thymidine incorporation differed proportionally between 2X and 10X MIC.  At 2X the  MIC of indolicidin there was no killing detected within 40 min, whilst at 10X MIC there were only around 10% survivors, with a steady decrease over the 40 minutes. Similar results were observed with CPI 1CN (Fig. 38), for which protein and R N A synthesis was affected within the first 5 to 10 minutes and there was no significant difference between 2X and 10X MIC. D N A synthesis was affected after 10 minutes with 10X MIC. Little effect was seen on D N A synthesis in the presence of 2X MIC. The killing curves were also similar to indolicidin, with no killing seen at 2X the MIC, and around 80% killing at 10X MIC. All macromolecular synthesis results shown in this thesis are representative of 3 separate experiments.  The counts per minute were not reproducible between  experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible.  117  D.3. Macromolecular Synthesis Inhibition by CPI OA CPI OA, a derivative of indolicidin, contained alanine residues substituted for proline residues.  This peptide had a better antimicrobial activity against S. aureus.  Killing assays done in conjunction with the macromolecular synthesis experiments showed that at 2x MIC there were 50% or greater survivors over the first 40 minutes, whereas at lOx MIC there were less than 10% survivors in the same time frame. Nonetheless, there was no major difference between the two in effects on R N A or protein synthesis (Fig. 39), which demonstrated nearly complete inhibition, indicating that the effect on synthesis at low peptide concentration was probably not the result of dead or dying cells. Macromolecular synthesis did not cease simultaneously, as expected if this was membrane disruption resulting in leakage of essential molecules. Protein and RNA synthesis (Fig. 39B, 39C) appeared to be affected first with differences observable within the first 5 minutes after peptide addition. D N A synthesis (Fig. 39A) was not affected until after 10 min.  118  120000 80000 40000  0  5  10  15  20  25  30  35 40  -5  0  5  10  15  20  25  30  35  Time (min)  Time (min) D 100  c D O  O  IX 10  15  20  25  30  35  40  Time (min)  Time (min)  Figure 3 7 : Effect of indolicidin at 2X the MIC (triangle) and 10X the MIC (square) on H-labelled thymidine (A), uridine (B), histidine (C) incorporation into S. aureus ISP67 macromolecules, measured as counts per minute over time. The control (no peptide) is shown in diamonds. Percent survivors (D) in the presence of 10X the MIC (square) and 2X the MIC (triangle) of indolicidin under conditions identical to the macromolecular synthesis assay. The results are representative of 3 separate experiments. The counts per minute were not reproducible between experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible. 3  119  40  e D.  10  15  20  25  30  35  40  10  Time (min)  D  15  20  25  30  35  Time (min)  100  o >  6 O  c a u  OH  10  15  20  25  30  35  40  Time (min)  Figure 38: Effect of CPI 1CN at 2X the MIC (triangle) and 10X the MIC (square) on H-labelled thymidine (A), uridine (B), histidine (C) incorporation into S. aureus ISP67 macromolecules, measured as counts per minute over time. The control (no peptide) is shown in diamonds. Percent survivors (D) in the presence of 10X the MIC (square) and 2X the MIC (triangle) of CPI 1CN under conditions identical to the macromolecular synthesis assay. The results are representative of 3 separate experiments. The counts per minute were not reproducible between experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible. 3  120  40  0  5  -5  10 15 20 25 30 35 40  0  5  10 15 20 25 30 35 40  Time (min)  Time (min)  D  300000  100  o >  200000 A 100000  c  o  <u  CL,  0  5  10 15 20 25 30 35 40  10  20  30  40  Time (min)  Time (min)  Figure 3 9 : Effect of CP10A at 2X the MIC (triangle) and 10X the MIC (square) on H-labelled thymidine (A), uridine (B), histidine (C) incorporation into S. aureus ISP67 macromolecules, measured as counts per minute over time. The control (no peptide) is shown in diamonds. Percent survivors (D) in the presence of 10X the MIC (square) and 2X the MIC (triangle) of CP10A under conditions identical to the macromolecular synthesis assay. The results are representative of 3 separate experiments. The counts per minute were not reproducible between experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible. 3  121  D.4. Macromolecular Synthesis Inhibition by CP29 CP29, an a-helical peptide derived from a cecropin-melittin hybrid peptide (Fig. 40), had an effect on protein and D N A synthesis at 10 minutes, and this appeared to be concentration dependent since at 10X the MIC there was an even greater effect. R N A synthesis was affected later, with significant differences seen only at 40 minutes. Interestingly, CP29 had the greatest effect on killing compared to the other peptides used in this study. The peptide at 2X the MIC caused killing of approximately half the cells, and at 10X MIC there were 0.01% or less survivors, all within the first 5 to 10 minutes.  D.5. Macromolecular Synthesis Inhibition by Bac2A-NH2 Bac2A-NH , a linear derivative (C3A, C11A) of the cyclic peptide bactenecin, 2  with increased activity on Gram positive bacteria (Fig. 41), affected the synthesis of protein, R N A and D N A within the first 5 minutes after addition of the peptide. There was no significant difference between results at 2X and 10X the MIC, although killing curves showed that at 2X the MIC there was very little killing (80% or greater survivors) and at 10X MIC there was a decrease to less than 10% survivors. The majority of the killing occurred in the first 5 minutes after peptide addition.  122  5  10  15  20  25  30 35 40  10  T i m e (min)  15  20  25  30  35 40  T i m e (min) D 100  o >  E a,  e  <u o CL,  5  10  15 20  25  30  35 40  T i m e (min)  Figure 40: Effect of CP29 at 2X the MIC (triangle) and 10X the MIC (square) on H-labelled thymidine (A), uridine (B), histidine (C) incorporation into S. aureus ISP67 macromolecules, measured as counts per minute over time. The control (no peptide) is shown in diamonds. Percent survivors (D) in the presence of 10X the MIC (square) and 2X the MIC (triangle) of CP29 under conditions identical to the macromolecular synthesis assay. The results are representative of 3 separate experiments. The counts per minute were not reproducible between experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible. 3  123  60000 40000  0  5  10  15  20  25  30  35  40  5  Time (min)  10  15  20  25  30  35 40  Time (min) D 100  300000 o >  200000 D. o  c  100000  o o u u CL, 10  15  20  25  30  35 40  Time (min)  Figure 41: Effect of Bac2A-NH at 2X the MIC (triangle) and 10X the MIC (square) on H-labelled thymidine (A), uridine (B), histidine (C) incorporation into S. aureus ISP67 macromolecules, measured as counts per minute over time. The control (no peptide) is shown in diamonds. Percent survivors (D) in the presence of 10X the MIC (square) and 2X the MIC (triangle) of Bac2A-NH under conditions identical to the macromolecular synthesis assay. The results are representative of 3 separate experiments. The counts per minute were not reproducible between experiments, as is to be expected with a biological assay of this type, so no error bars were used. However, the trends seen were reproducible. 2  2  124  E . Summary Electron microscopy showed varying effects of peptides on cells depending on both the peptide and the type of bacterium.  Peptide-treated S. aureus exhibited  mesosome structures in the cytoplasm, and this was consistent among the different peptides. These mesosome structures were found to be less prevalent in bacteria in the presence of KC1, which increases the MIC of many peptides. However, peptide-treated S. epidermidis underwent various structural effects depending on the peptide. The effects of the a-helical peptides CP26 and CP29 were distinctly different from those of Bac2AN H , CP11CN and CP10A. 2  CP26 and CP29 resulted in mainly lysis and cell wall  disintegration, whereas CP11CN and Bac2A-NH had numerous effects including nuclear 2  condensation, the formation of mesosomes, and abnormal septation and cell wall effects. CPI OA, on the other hand, resulted in mesosome structure formation only. CPI 1CN had effects on the septal wall of E. faecalis, but Bac2A-NH (which had better activity against 2  E. faecalis) did not have any obvious effects on cell structure or integrity.  Scanning  electron microscopy verified the findings from transmission electron microscopy. All peptides had the ability to depolarize the cytoplasmic membrane of S. aureus. However, the peptides also had different effects on cytoplasmic membrane depolarization, and these effects did not correlate with MICs or killing. Macromolecular synthesis assays showed that intracellular effects did occur, but these too differed depending on the peptide. Intracellular effects were seen at sub-lethal concentrations, and DNA, RNA, and protein synthesis did not cease simultaneously (with the exception of Bac2A-NH ). 2  125  DISCUSSION A . Overview Antimicrobial cationic peptides have been under study for over a decade and still speculation remains as to what the mechanism of action on bacteria is, as well as what peptide structure and function relationships exist. Many authors and reviewers to date have stated that the cytoplasmic membrane is the site of action whereby the peptides cause a major breach in cytoplasmic membrane permeability or lead to cell lysis (Hancock et al., 1995).  Since this study was initiated, more evidence, including that  presented in this work, has surfaced in favor of alternative mechanisms of action. The structural elements that result in peptide specificity and selectivity are being investigated in many studies.  In general, peptide characteristics such as charge, hydrophobicity and  amphipathicity have been examined. It is believed that amphipathicity is one of the most important structural requirements for antibacterial activity (Andreu and Rivas, 1998; Pathak et al., 1995). Often these studies use model membranes in order to gain insight into peptide structure and interaction with the membranes of bacteria. The mechanism of action against Gram positive bacteria has not been studied as intensively as the mechanism on Gram negative bacteria. The aim of this study was to gain further insight into peptide structure/activity relationships as well as into the mode of action against Gram positive bacteria. Our approach initially involved testing the activity of peptides of different structural classes on Gram positive bacteria under various conditions. As well, in  this  study  it  was  investigated  if  a  correlation  existed  between  peptide  structure/interaction with model membranes and peptide activity and specificity. When no correlation was discovered, further investigation into the mode of action was required.  126  Electron microscopy, cytoplasmic membrane depolarization assays and macromolecular synthesis assays were used to more closely analyze the interaction of these peptides with Staphylococcus  aureus.  It is important to note that previous studies relied on a single  class of peptides, whilst here a range of different peptides were studied in order to gain a more global view of the antimicrobial mechanisms of cationic peptides.  Indeed, some  results presented in this thesis have been published and have contributed to the growing pool of evidence  that cationic peptides have targets other than the cytoplasmic  membrane.  B. Antimicrobial Activity of Various Peptides Initial work was necessary to verify and optimize our method of determining MICs. The work showed that, in general, the age and type of holding solution, as well as the media used in the broth dilution method, did not have a significant effect on MICs. The agarose MIC method was found to be more dependent on both peptide and bacteria used. The peptides displayed a range of antibacterial activity against a panel of Gram positive bacteria. CP26, an a-helical peptide with excellent antibacterial activity against Gram negative bacteria and comparable to CP29 (Friedrich et al., 1999), had little to no activity against Gram positive bacteria, as judged in an MIC assay, with the exception of Corynebacterium  xerosis.  It was however able to cause partial killing below its MIC.  Presumably, the lack of complete killing influences the final value of the MIC. CP26 is similar to CP29 in that CP26 shares the same N-terminal amino acids as CP29, but the middle hinge region was made more amphipathic and flexible with an added alanine, and the C-terminus was modified to be more a-helical and hydrophilic. Despite their similar  127  sequences (6 amino acids different), CP29 was apparently more active against most Gram positive bacteria. CPI 1CN, a variant of indolicidin, had improved activity against Gram negative bacteria (Falla and Hancock, 1997) but was not better against Gram positive bacteria. In contrast, CPI OA had the best activity of the indolicidin variants against Gram positive bacteria, and its activity against Gram negative bacteria was approximately equal to that of CP11CN (data not shown).  Interestingly, CP10A had significant activity  against E. faecalis, which was normally resistant to many cationic peptides. All peptides killed methicillin-resistant S. aureus and vancomycin-resistant S. haemolyticus to a similar extent as the parent strains. This indicated that the mechanisms of methicillin and vancomycin resistance did not affect the peptides, further evidence that the mechanism of action of these peptides was likely very different from that of conventional antibiotics. Thus, the best peptides studied here had activity against a broad range of Gram positive bacteria.  Through this preliminary work I discovered peptides with differences in  activity and selectivity to use in my experiments. The presence of monovalent cations in the form of KC1 and NaCl had a negative effect on the activity of the peptides against S. aureus, whereas the presence of divalent cations in the form of M g C l and CaCl did not have a significant effect. The opposite 2  2  was true for E. coli. Previous studies (Piers and Hancock, 1994) have suggested that these peptides displace M g  + +  at the binding sites at the base of LPS, thereby destabilizing  the outer membrane and allowing self-promoted uptake.  Divalent cations act as  competitors for the binding sites therefore resulting in a decrease in activity against E. coli. This is consistent with data from other peptides, including magainin 2 (Matsuzaki et al., 1999). The effect of monovalent cations on the activity of these peptides on Gram  128  positive bacteria is likely due to sensitivity to the charge-screening effect of the high ionic strength solutions. The killing kinetics of these peptides, regardless of structural class, were similar at 10-fold the MIC on both S. epidermidis and S. aureus. However, C C C P had markedly different effects on the killing kinetics of these peptides. Previous work revealed that CCCP protected E. coli from the action of indolicidin (Falla et al, 1996). In this study, however, C C C P was observed to act in synergy at certain concentrations with indolicidin, CP10A and CP29. CCCP alone had a bacteriostatic effect on S. aureus. Some property of CCCP may have enhanced the action of peptide. Alternatively, the dissipation of the electrical potential may, in the case of S. aureus and these specific peptides, have aided in the antimicrobial action (eg. inhibition of peptide efflux, synergistic inhibition of a specific target, etc.). This is in contrast to previously held theories that peptides require an intact electrical potential in order to form channels in the cytoplasmic membrane and kill the cell (Falla et al,  1996) and is further evidence that the mode of action of  antimicrobial peptides may be both bacteria and peptide specific.  C. Peptide Design and Structure-Activity Relationships The MICs of a panel of indolicidin variants led to some general conclusions. It was found that an increase in positive charge led to an increase in peptide activity. For example, CP11CN, CP10R and CP11R all had an increase in cationicity. However, the backbone structures of these peptides were likely to be very similar. This increase in positive charge may lead to an increase in selectivity for the outer membranes of Gram negative bacteria or the more anionic phopholipids of the cytoplasmic membranes of  129  bacteria.  However, the similar backbone structure of these three variants allowed the  peptides to exert a killing action similar to that of the parent peptide, indolicidin. The exchange of tryptophan residues did not cause a significant effect, which was consistent with a previous study in which the tryptophan residues of indolicidin were exchanged with phenylalanine and the resulting peptide had comparable activity to the parent peptide against E. coli (Subbalakshmi et al., 1996). A more recent study where single tryptophan analogs (with one tryptophan residue replaced with leucine) were created demonstrated that these analogs retained antimicrobial activity, but did not lyse erythrocytes (Subbalakshmi et al., 2000). (isoleucine  The hydrophobic N-terminal amino acids  and leucine) did not play an important role in antimicrobial activity.  Substitutions of these amino acids and an increase in hydrophobicity in the N-terminus both resulted in equally active peptides. Again, an alteration of these amino acids would likely not result in a significant change in the backbone structure, the parameter which likely has the most significant effect on antimicrobial action.  These results were  generally consistent between Gram positive and Gram negative bacteria. The exchange of the proline residues in indolicidin for alanine residues resulted in CPI OA, a peptide more active against Gram positive bacteria. This substitution resulted in a drastic change in overall structure, as will be discussed in further detail later in this discussion. Thus, structure/function relationships with respect to the parent peptide could not be derived from CPI OA because of these new peptide properties.  130  D. Interaction with Model Membranes and Structure-Activity Relationships A combination of appropriate chain length, amino acid composition and positioning of apolar and positively charged residues is required for the antibacterial activity of cationic antimicrobial peptides; however, the exact nature of this combination is still under study. Pathak et al. (1995) hypothesized that the amphiphilicity of antimicrobial peptides is the most important factor governing activity, compared to either mean hydrophobicity or a-helix content. In another study it was reported that the antimicrobial activity of the cecropin-melittin hybrid peptides depended on their helical nature (Andreu et al., 1985). The already established cecropin-melittin hybrid peptide C E M E , and C E M A , a variant peptide with 2 extra amino acids and positive charges in the C-terminus were used as the templates for further design. Piers et al. (1994) showed that C E M A was more efficient at binding the divalent cation binding sites on surface LPS, and more efficient at destabilizing the outer membrane of gram negative bacteria. However, C E M A did not display notably better antibacterial killing or MICs than C E M E . In this study we showed that the altered C-terminus of C E M A made it somewhat less a-helical than C E M E , which raised the question as to whether the degree of a-helicity is a factor in antibacterial activity.  Recent work with cathelecidin-derived peptides  and their  analogues indicated that the degree of helicity was not a simple predictor of antimicrobial activity (Travis et al., 2000). C E M E was modified to have an increase in amphipathic ahelical content in its first 14 amino acids, resulting in CP29. This was accomplished by spacing out the lysine and hydrophobic amino acids to better fit a helical wheel projection, as well as replacing the glycine residues in the middle of the peptide with the threonine which is more easily accommodated in an a-helix. CP26 was designed using  131  CP29 as the parent peptide. The middle hinge region was made more amphipathic and flexible with an added alanine, and the C-terminus was modified to become more ahelical and hydrophilic with the addition of an extra positive charge. CP201 was also based on CP29, with one less charge and lower hydrophobicity as a result of the 2 glycines added to make the central region more flexible, like that of C E M E . CP208 was very similar to CP26, with the main difference being the substitution of a lysine for the conserved tryptophan residue at position 2. This peptide was also designed to maintain an a-helical structure. In general, the a-helical, amphipathic cationic peptides have been estimated to contain between 30-90% helix, depending on the peptide and the environment. Most of these peptides are unordered in aqueous solution, with the exception of PMAP-37 and LL-37, which were helical in buffer (likely due to oligomerization). A peptide analogue of BMAP-28, in which hydrophobic amino acids at the C-terminus were replaced with hydrophilic residues, showed an increase in activity against Gram negative bacteria, but a decrease in activity against Gram positive bacteria (Gennaro and Zanetti, 2000). This is somewhat similar to the results presented here for CP26. The C D spectral analysis of these peptides showed that CP29 had the highest helicity of all the peptides. Although CP26 was designed to have a more helical C-terminus, the changes around the bend region (i.e. in the vicinity of the proline residue at position 22), including the additional charge seemed to have a detrimental effect on a-helix formation. CP201 was similar to CP26 in terms of its C D spectra. CP208, however, was essentially a random coil in the presence of liposomes, possibly owing to the lack of the important hydrophobic residue tryptophan in position 2 since this amino acid is known to be important for interaction of  132  proteins with lipid membranes (Meers, 1990). This was despite the fact that CP208 had a similar charge and percentage of hydrophobic residues as the other peptides, and was designed to have a a-helical structure. CP26, C E M E , C E M A and CP29 had the same number of amino acids, but charges ranging from +5 to +7, hydrophobic amino acid contents from 46 to 58%, and ahelix contents ranging from 17 to 57% in lipid environments. It was thus of interest to note the similarities and differences between these peptides. All had similar and good activities against the Gram negative bacteria tested, and were able to rapidly kill logarithmic phase bacteria (Friedrich et al, 1999). The increase in a-helicity of CP29 did not make the peptide more active in vitro (Friedrich et al, 1999). The MICs of CP26 against Gram negative bacteria were also similar indicating that the altered bend region and extra charge did not affect its MIC or that it had a different killing mechanism. However, CP26 did have decreased activity against Gram positive bacteria. CP201, which fell within the ranges of most of the physical properties of the above 4 peptides, was not a good antibacterial, possibly due to the combination of two detrimental factors, lower hydrophobicity (42%) and less charge (+5). CP208, which also had physical properties similar to the 4 active peptides had virtually no antibacterial activity, possibly because it lacked a tryptophan residue at position 2 which I propose to be required for insertion of the peptide into the lipid membrane. This was supported by the observation that although CP208 was designed to be a-helical, it was unable to form a-helix upon interaction with liposomes, but was able to in the presence of SDS. Fluorescence spectroscopy of CP26, C E M E , C E M A and CP29 indicated that all 4 peptides entered the hydrophobic environment of the lipid bilayer and the tryptophan  133  residues were inaccessible to the aqueous solution. However, CP26 and CP29, with their similar N-termini, had greater blueshifts upon liposome interaction. Quenching of the fluorescence emission of the bilayer-bound peptides was greatest for spin labels at position 5 in the hydrophobic chain. These results indicate that there is a slight difference in interaction with lipids between these peptides, however no correlation could be made with activity. The C D spectra of indolicidin suggested that slight structural variations existed depending on the micelle type or liposomes used. Previous C D studies of indolicidin have led to controversial interpretations and suggestions that indolicidin is composed predominantly of poly-proline II helix elements or p-turns in the backbone dihedral angles (Falla et al, 1996; Ladokhin et al, 1999). 2D-NMR studies, from our lab, of indolicidin in the presence of DPC and SDS have verified that there were elements of both poly-proline U and p-turn present in indolicidin, and that indeed the structure differed somewhat between DPC and SDS. The structure of indolicidin in DPC has been shown to be a "boat" shape which comprised of a hydrophobic core flanked by two positively charged regions (Rozek et al, 2000) (Fig. 42).  The C D and fluorescence  spectrum of indolicidin in lyso-PC/DPC most closely resembled that of the spectrum in liposomes of both neutral and negatively charged lipids. The blue shift of the emission fluorescence maximum with a concomitant increase in intensity in the presence of lipids indicated a movement of the tryptophan residues into a more hydrophobic environment (Lakowicz, 1983). The small blue shift of 8 nm was consistent with experiments using membrane spin label probes suggesting that the tryptophan residues on average were close to the polar lipid head group. Nevertheless, the tryptophan fluorescence of the  134  peptide in lipid membranes was not quenched by the aqueous quencher KI, indicating that these residues became inaccessible to the buffer. Therefore, it is very likely that the 3D-structure determined in DPC is similar to the structure that would be induced upon interaction with the cytoplasmic membrane of bacteria. However, 2D-NMR studies have shown that indolicidin formed a more extended and less defined structure in the presence of SDS (Rozek et al., 2000), which explained the difference in C D spectra. This may be explained by a difference in the shape of the micelles formed by SDS due to the negative head group charge on this molecule.  The 3D-structure of CP11CN has also been  determined in DPC and was found to be similar to the structure of indolicidin (Fig. 42). The charged regions on the ends of CP11CN were larger and the hydrophobic central region was smaller, however, the overall backbone structure was very similar between the two peptides. The C D and fluorescence data presented here indicated that the structures of indolicidin and CP11CN would be similar. However, CP11CN was shown to have a decreased interaction with POPC.  This could be explained by the more positively  charged peptide causing increased repulsion upon binding to a neutral liposome, resulting in an overall decrease in peptide binding. CPI 1CN did have better activity against Gram negative bacteria, likely because the increased cationicity allowed for greater interaction with LPS in the outer membrane of Gram negative bacteria (Falla and Hancock, 1997). The unique sequence and activity of CPI OA made it an interesting candidate for structural analysis. The C D spectrum of CP10A in buffer was not that of a random coil. This is in contrast with many other cationic peptides in this study, such as the a-helical peptides C E M E and C E M A and indolicidin. This is unusual for a very short peptide. However, upon interaction with lipids the minima at 205 nm and 218 nm were greatly  135  enhanced indicating that there was a further induction of structure upon binding. In both lipids and detergent a maximum appeared at 230 nm that was not present in buffer. The latter was likely due to the contribution of tryptophan (Woody, 1994).  Probably the  minimum at 222 nm, expected for helical structure, was partially cancelled by the maximum at 230 nm. The shapes of the C D curves were similar in the presence of both lipids and detergents, showing the same minima and maxima.  The spectrum in the  presence of POPC and mixed liposomes showed the strongest minima, indicating the highest content of helical secondary structure. The fluorescence spectra and spin label data were similar to those observed with indolicidin and CPI 1CN. The similar blue shift of CPI OA in lipid vesicles and DPC suggested that the tryptophan residues in CPI OA were in a similar environment.  Tryptophan has a strong preference for the membrane  interface when compared with other hydrophobic amino acids (Wimley and White, 1996). Furthermore, the tryptophan bands at 230 nm in the C D spectra were identical indicating a similar orientation of tryptophan side chains with respect to the backbone. The structure of CPI OA in DPC micelles determined from N M R data (by Annett Rozek) (Fig. 42) was mainly a-helical with elements of 3i  0  helical geometry. The  structural differences between indolicidin (extended) and CPI OA (helical) were brought about by the substitution of the three proline residues for alanine. Alanine has the largest helix-forming propensity of all natural amino acid residues, which has been attributed in part to its small side chain (Chakrabartty et al, 1994). The spacing of alanines in CP10A three and four residues apart, may have had an additional helix-forming influence, since these small side chains lined up on the hydrophobic side of the peptide, thereby providing ample space for the bulky tryptophan side chains. This change in structure resulted in a  136  peptide with increased activity against Gram positive bacteria.  CPI OA had a more  compact structure than indolicidin, which was more extended. Compactness in addition to increased amphiphilicity may be a factor in the difference in selectivity/mechanism of action of CPI OA. Because of its more coiled structure, CPI OA was too small to span the membrane, therefore it was unlikely to be a pore former as suggested for certain other ahelical peptides such as cecropin (Christensen et al,  1988).  This leads to further  speculation that the peptide may diffuse through the membrane in order to reach an intracellular target. The C D spectra of the peptides in the presence of L T A were identical to the spectra in the presence of liposomes.  This indicated that the peptides did bind L T A ,  which verified previous work showing that some peptides bound L T A in a modified dansyl-polymyxin displacement assay (Scott et al., 1999).  The C D spectra in the  presence of LPS did not give evidence of peptide structure induction upon LPS binding. Recent work by Matsuzaki et al. (1999) showed that the peptide magainin formed a helical structure upon interaction with LPS-containing liposomes. With the exception of CP208, a peptide with no activity and very little structure induced  upon  lipid  structure/interaction  interaction,  with  membrane  no  correlation  and  differences between the peptides.  137  could  be  antimicrobial activity  made  between  despite  obvious  Figure 42: Structures of indolicidin (A), CP11CN (B) and CP10A (C) determined by 2D-NMR in dodecylphophocholine. Shown on the left are the backbone structures and the structures with all heavy atoms are shown on the right. The N-terminus is shown to the left. All positively charged side chains are shown in gray. Structures were determined by Dr. Annett Rozek of the Department of Microbiology and Immunology, University of British Columbia.  138  E . Antimicrobial Mechanism against Gram Positive Bacteria The exact mode of action of cationic antimicrobial peptides on Gram positive bacteria is unknown. However, it has been proposed and is widely believed that the peptides interact with and disrupt the cytoplasmic membrane, leading to the dissolution of the proton motive force and leakage of essential molecules, resulting in cell death. Recently there has been evidence, including the results presented in this thesis, of an intracellular or alternative target (Xiong et al,  1999) (Otvos et al, 2000b) as well as important  interactions with the cell wall (Peschel et al, 1999). Here, these issues were investigated using peptides with different structures.  E.l. Electron Microscopy Electron microscopy showed that frequently there was an effect on the S. aureus membrane by all peptides studied.  This was demonstrated by the appearance of  mesosome structures similar to those seen with defensins (Shimoda et al,  1995),  trimethoprim (Nishino  1993).  et al,  1987)  and rifampin  (Gottfredsson  et al,  Mesosomes, which are intracytoplasmic membrane inclusions, have been regarded as structural artifacts induced by the chemical fixatives used on the cells prior to plastic embedding and thin sectioning (Beveridge, 1989). Yet, mesosomes must be regarded as being indicative of cytoplasmic membrane structural alterations, in this case induced by the cationic peptides, since untreated cells did not contain them, and KCl-treated cells contained them less frequently.  Furthermore, since the cytoplasmic membrane is  instrumental in cell wall synthesis and turnover, a perturbation to this membrane could  139  also  affect  cell wall integrity and autolysin regulation (Kemper et al,  1993).  Accordingly, the very fact that mesosome-like structures were seen in most treated cells is indicative of cytoplasmic membrane effects and (possibly) uncoupling of the synthesis and turnover of cell wall polymers. Clearly, lysis was evident only in a small proportion of cells, and in the case of Bac2A-NH this lysis was apparently initiated at the septal 2  site. With S. epidermidis, however, more diverse effects were seen. CP29, CPI 1CN and Bac2A-NH caused cell wall effects such as cell wall breaks, thinning and disintegration 2  as well as abnormal septation. interestingly, trimethoprim also causes defects in cell wall formation and irregular cross wall formation similar to those seen here (Nishino et al., 1987). CP29, unlike Bac2A-NH and CPI 1CN, did not cause mesosome-like structures 2  or nuclear condensation in S. epidermidis. Notably, CPI OA caused mesosomes but did not appear to cause extensive damage to the cell wall or nuclear condensation on S. epidermidis. CP26 caused lysis of S. epidermidis, but intact cells did not appear to have any of the structural changes mentioned above, with the exception of a few smaller and infrequent mesosome structures. It is worthwhile to note that the effects caused by the two a-helical peptides, CP26 and CP29, were similar in that intact treated cells did not demonstrate nuclear condensation or frequent mesosomes. Thin sections of Bac2A-NH 2  treated E. faecalis did not reveal any obvious structural changes, which correlated with the minimal killing seen over 90 minutes, despite the good MIC of Bac2A-NH on E. 2  faecalis.  Interestingly, CP11CN, a peptide with a high MIC on E. faecalis, showed  effects on the cell wall around the septum. These various effects indicated that these peptides might kill cells in different ways.  140  The S E M data gave further evidence of these peptides differing in mechanism of action and verified some of the findings with T E M . The blebbing of the cell wall seen in the S E M of CPI lCN-treated S. epidermidis corresponds to the cell wall fibres extending from the surface of the cell in the thin sections. The holes seen in the cell walls of CP29treated S. epidermidis correspond to the disintegrating cell wall seen in the thin sections. Moreover, the SEMs of S. aureus in the presence of the same peptides did not look any different than the control, corresponding to the thin sections where only mesosome structures were seen inside the cell.  A summary of the different effects seen on S.  epidermidis as a result of the different peptides is shown in Table 12.  E.2. Cytoplasmic Membrane Depolarization Previously, the DiSC3(5) assay was used to demonstrate disruption of the E. coli cytoplasmic membrane by cationic peptides (Wu et al., 1999). Here, we further adapted the DiSC (5) assay for use with S. aureus. The peptides showed considerable differences 3  in this assay, and as shown with E. coli, there was no obvious correlation between cytoplasmic membrane depolarization and antimicrobial activity (Wu et al, 1999). For example, CP26, which had the lowest activity of all the peptides, had one of the best abilities to permeabilize at low concentrations, and along with the other a-helical, CP29, caused complete depolarization by 8 ug/ml.  CP29, in 100 mM KC1, had MICs well  above the concentration needed for complete depolarization, consistent with the results for indolicidin, CPI 1CN and Bac2A-NH . Bac2A-NH , the peptide with the best MIC in 2  2  the absence of KC1, did not permeabilize well at low concentrations.  CPI OA almost  completely depolarized the membrane at a concentration of 1 ug/ml, indicating that it  141  was more similar in action to the a-helical peptides than to its parent peptide indolicidin. There have been reports of the rapid permeabilization of the cytoplasmic membranes of susceptible bacteria and artificial membranes at concentrations comparable to the MIC values (Gennaro and Zanetti, 2000; Wu et al,  1999).  It is important to note that  depolarization of the cytoplasmic membrane is not, per se, a lethal event as the depolarizer valinomycin in the presence of 100 mM KC1 is in fact bacteriostatic not bactericidal. Killing curves done in conjunction with the depolarization assay indicated that similar cytoplasmic membrane permeabilization profiles did not correspond with similar killing rates.  This was shown for the closely related a-helical peptides (Fig. 35A),  whereby CP29 killed significantly more bacteria than CP26 without a significant difference in permeabilization. As well, a significant reduction in numbers of bacteria (90-99%) occurred within the first minute after addition of the peptide.  At this point  there was very little permeabilization of the cytoplasmic membrane by the a-helical peptides (Fig. 35A), and only ~ 50% of the maximum possible permeabilization by indolicidin and CPI 1CN (Fig. 35B). Killing by indolicidin and CPI 1CN continued after the depolarization had reached a plateau at approximately 5 minutes. This is in contrast to the results with HNP-1 and tPMP-1, where the cytoplasmic membrane of S. aureus was depolarized within minutes but cell death occurred 1 to 2 hours later (Yeaman et al, 1998).  The passage of the peptide through the membrane in order to reach an  intracellular target would be expected to cause an increase in membrane permeability and this might account for the lag time between depolarization and killing.  Membrane  permeabilization occurring after cell death could be a secondary or subsidiary effect of  142  the peptides. These results are thus consistent with the view that cytoplasmic membrane permeabilization is not the primary target for bacterial killing.  E.3. Macromolecular Synthesis Inhibition The results of the MICs and the cytoplasmic membrane depolorization assay indicated that the peptides used in this study had similar activities against the auxotrophic strain ISP 67 and S. aureus A T C C 25923.  The auxotroph may be slightly more  susceptible to peptides than the wild-type, perhaps due to the more fragile nature of auxotrophic strains, but in general the MICs were not significantly different and the patterns of depolarization by the peptides were similar. The MICs of peptides in the presence of LB and in the presence of synthetic media containing histidine, uridine and thymidine were not significantly different, suggesting that the precursors were not acting as competitive inhibitors of the peptides. Thus I have confidence that the results obtained from the macromolecular studies are applicable to the wild-type strain. The peptides used in this study have been shown to have very different structures as well as effects on S. aureus. In addition, these peptides also differed in intracellular effects. A summary is shown in Table 13. Bac2A-NH affected DNA, R N A and protein 2  synthesis at a concentration where 80% or more of the bacteria were viable. CP29, on the other hand, was the most efficient bactericidal peptide but failed to have obvious effects on R N A synthesis until 40 minutes, a time at which 50% or more bacteria were nonviable. These results may indicate that the effects seen on R N A synthesis resulted from dead or dying cells. Exposure to indolicidin and CPI 1CN resulted in different effects on macromolecular synthesis. Both protein and RNA synthesis were rapidly affected despite  143  the fact that no killing was seen over 40 minutes at 2X the MIC.  Intracellular effects  were seen although the cells were still viable. Consistent in part with this, a prior study had proposed that indolicidin inhibited D N A synthesis in Sitaram, 1998).  E. coli (Subbalakshmi and  Overall, it appeared that CP29 and indolicidin/CPl 1CN had quite  different effects on cells.  Indolicidin and CP11CN could potentially penetrate the  cytoplasmic membrane quickly and reach intracellular targets, resulting in a decrease in macromolecular synthesis. However, a threshold concentration would need to be reached in order to cause cell death.  CP29, however, might be exerting its lethal action in a  different manner, causing a delayed effect on macromolecular synthesis.  CP10A had  effects on macromolecular synthesis similar to those observed with indolicidin, but with more rapid lethal effect. Macromolecular synthesis studies verified that intracellular effects occurred, even at sub-lethal peptide concentrations.  Cessation of all macromolecular synthesis did not  occur simultaneously with all peptides, as would be expected if large cytoplasmic lesions occurred that resulted in loss of precursors or destruction of cellular ATP. It is thus likely that some peptides interacted with the cytoplasmic membrane of  S. aureus in order to  reach an intracellular target and that the observed effects on the cytoplasmic membrane reflects this passage across the membrane to access their target.  E.4. Conclusion I propose that there are multiple possible anionic targets of peptide action against bacteria including the D N A , RNA, cell wall, cytoplasmic membrane and various enzymes and that these targets are differentially accessed depending on the peptide and  144  bacteria in question. The rapid killing action of the peptides, once thought to be the result of membrane disruption, may indeed be the result of the peptide attacking multiple targets resulting in the inhibition of eg. cell division or D N A synthesis, as well as membrane effects. The observation of loss of colony forming ability at any given time reflects the association of peptide with cells in such a manner as to eventually prevent cell multiplication and thus represents a "commitment" to die rather than cell death per se. Although the exact mechanism of action of peptides on Gram positive bacteria has not been resolved, the evidence reported here is consistent with a multiple-hit model, as discussed for polycationic aminoglycosides (Hancock, 1981) and with the multimodal model presented for tPMP-1 and HNP-1 (Xiong et al, 1999). Further insight into the mechanism of action will aid in the production of cationic antimicrobial peptides as future therapeutics.  F. Potential Applications of Peptides These cationic peptides have been shown to have a broad spectrum of activity, including activity against bacteria that are resistant to most known antibiotics. There are some strains of bacteria that are resistant to cationic peptides.  These include a S.  typhimurium strain with an altered LPS structure (Groisman et ai, 1997) and a. Neisseria gonorrhoeae strain capable of active efflux of antimicrobial peptides (Shafer et al., 1998). The  sap operon in V. fisheri has been shown to be responsible for the resistance  of this bacterium to some antimicrobial peptides.  This operon is believed to be  associated with peptide transport (Chen et al., 2000). The consecutive exposure of P.  aeruginosa to indolicidin resulted in a stable mutant with decreased susceptibility to the  145  peptide (Giacometti et al., 1999).  However, thus far most peptides, unlike classic  antibiotics, have not been shown to elicit rapid resistance in susceptible bacteria (Hancock and Lehrer, 1998). This may be due to the fact that simultaneous inhibition of more than one target is thought to render emergence of resistance less likely (Chopra, 1998). This is a very desirable trait for future antibiotics as the emergence of antibioticresistant bacteria is increasing at an alarming rate. Further knowledge of the structure and function of these peptides, as well as the structural factors involved in toxicity and antimicrobial selectivity, will hopefully allow modification of these peptides into protease-resistant, non-toxic antibiotics.  Recently, the cross-linking of tryptophan  residues in indolicidin resulted in an active but protease-resistant peptide (Osapay et al., 2000), and unnatural amino acids incorporated into an insect peptide resulted in increased stability and decreased toxicity (Otvos et al., 2000a).  The fact that some of these  peptides are endogenous in humans and are part of the innate immune system makes them more desirable candidates for therapeutic drugs.  Penetration into the cell,  appropriate antibacterial spectrum, favorable pharmacokinetics and low toxicity are requirements for new drugs.  Some peptides have already entered clinical trials. For  example, Micrologix Biotech Inc. (Vancouver, BC) has a peptide candidate in clinical trials for prevention of catheter-associated and acne infections. Intrabiotics has a peptide candidate in phase DT clinical trials for treatment of oral mucositis.  Other peptide  candidates from various companies are in different stages of clinical trials for indications such as N. meningitidis and H. pylori infections (Hancock and Lehrer, 1998).  146  G. Future Studies We are only now beginning to understand the complexity and diversity of these peptides and their antibacterial action. This study and others have indicated that there is no universal mode of action, and as a result, targets specific to certain peptides need to be identified along with the structural elements required for these interactions. The lethal interactions need to be distinguished from the non-lethal effects and there should be a clearer understanding of what elements result in selectivity between different bacterial classes. This will ensure that peptides designed for therapeutic purposes will maintain their killing ability at non-toxic levels.  It will also be important to study how these  peptides interact with biofilm environments, such as those present on catheters.  These  situations represent some of the more persistent antibiotic-resistant infections.  These  studies are important if these peptides are to be developed into a new class of antibiotics.  Table 12: Summary of the effects seen as a result of treatment of S. epidermidis with different peptides, as observed by electron microscopy  Cell wall Peptide  effects  CP29 (a-helical)  Partial Disintegration  Septum effects  Nuclear effects  Abnormal  No effect  CP11CN (extended)  Blebs  Rare effect  Condensation  CP10A (small (xhelical)  No effect  No effect  No effect  Blebs  Abnormal  Condensation  Bac2A-NH (P-sheet)  2  * mesosomes seen with all peptides on S. aureus  147  Table 13. Summary of the antimicrobial activity, membrane depolarization and macromolecular synthesis inhibition of different peptides (at 1-to 2-fold MIC) on S. aureus ISP 67 in synthetic media  Cytoplasmic membrane  Time of inhibition of synthesis (min)  depolarization % survival  Time to  Concentration  over 40  Max  max  (1-2-fold MIC)  min  (%)  (min)  DNA  RNA  Protein  16  30-40%  50  25  10  40  10  indolicidin (extended)  4  100%  33  5  After 10  10  5  CP11CN (extended)  16  100%  50  5  Little effect  5  5  2  50%  *N/D  *N/D  After 10  5  5  32  60-80%  50  15  5  5  5  Peptide CP29 (a-helical)  CP10A (small ahelical) Bac2A-NH ( 0-sheet)  2  *N/D Not done  148  References Acar, J.F. and Goldstein, F.W. (1996) Disk Susceptibility Test. In Lorian, V . (ed.) Antibiotics in Laboratory Medicine. Williams and Wilkins, Baltimore, pp. 1-52. Amsterdam, D. (1996) Susceptibility testing of antimicrobials in liquid media. In Lorian, V. (ed.) Antibiotics in laboratory medicine. Williams and Wilkins, Baltimore, pp. 52-111. Andrade, M.A., Chacon, P., Merelo, J.J. and Moran, F. (1993) Evaluation of secondary structure of proteins from U V circular dichroism using unsupervised learning neural network. Prot Eng, 6, 383-390. Andreu, D., Merrifield, R.B., Steiner, H. and Boman, H.G. (1985) N-terminal analogues of cecropin A: synthesis, antibacterial activity, and conformational properties. Biochemistry, 24, 1683-8. Andreu, D. and Rivas, L. (1998) Animal antimicrobial peptides: an overview. Biopolymers, 47, 415-33. Andreu, D., Ubach, J., Boman, A., Wahlin, B., Wade, D., Merrifield, R.B. and Boman, H.G. (1992) Shortened cecropin A-melittin hybrids. Significant size reduction retains potent antibiotic activity. FEBS Lett, 296, 190-4. Baneyx, F. (1999) Recombinant protein expression in Escherichia coli. Curr Opin Biotechnol, 10, 411-21. Bayer, A.S., Prasad, R., Chandra, J., Koul, A., Smriti, M . , Varma, A., Skurray, R.A., Firth, N., Brown, M.H., Koo, S.P. and Yeaman, M.R. (2000) In vitro resistance of Staphylococcus aureus to thrombin-induced platelet microbicidal protein is associated with alterations in cytoplasmic membrane fluidity. Infect Immun, 68, 3548-53. Beveridge, T.J. (1989) The structure of bacteria. In Leadbetter, E.R. and Poindexter, J.S. (eds.), Bacteria in nature: a treatise on the interaction of bacteria and their habitats. Plenum Pub. Co., New York, Vol. 3, pp. 1-65. Bierbaum, G. and Sahl, H.G. (1985) Induction of autolysis of staphylococci by the basic peptide antibiotics Pep 5 and nisin and their influence on the activity of autolytic enzymes. Arch Microbiol, 141, 249-54.  149  Blondelle, S.E. and Houghten, R.A. (1992) Design of model amphipathic peptides having potent antimicrobial activities. Biochemistry, 31, 12688-94. Blondelle, S.E., Takahashi, E., Weber, P.A. and Houghten, R.A. (1994) Identification of antimicrobial peptides by using combinatorial libraries made up of unnatural amino acids. Antimicrob Agents Chemother, 38, 2280-6. Boman, H.G., Agerberth, B. and Boman, A. (1993) Mechanisms of action on Escherichia coli of cecropin PI and PR-39, two antibacterial peptides from pig intestine. Infect Immun, 61, 2978-84. Boman, H.G., Faye, I., Gudmundsson, G.H., Lee, J.Y. and Lidholm, D.A. (1991) Cellfree immunity in Cecropia. A model system for antibacterial proteins. Eur J Biochem, 201, 23-31. Boman, H.G., Wade, D., Boman, A., Wahlin, LA. and Merrifield, R.B. (1989) Antibacterial and antimalarial properties of peptides that are cecropin-melittin hybrids. FEBS Lett, 259, 103-106. Bovey, F.A. and Hood, F.P. (1967) The circular dichroism spectrum of poly-Lacetoxyproline. Biopolymers, 5, 915-9. Brotz, H. and Sahl, H.G. (2000) New insights into the mechanism of action of lantibiotics-diverse biological effects by binding to the same molecular target [In Process Citation]. J Antimicrob Chemother, 46, 1-6. Cabiaux, V., Agerberth, B., Johansson, J., Homble, F., Goormaghtigh, E. and Ruysschaert, J.M. (1994) Secondary structure and membrane interaction of PR-39, a Pro+Arg-rich antibacterial peptide. Eur J Biochem, 224, 1019-27. Castle, M . , Nazarian, A., Y i , S.S. and Tempst, P. (1999) Lethal effects of apidaecin on Escherichia coli involve sequential molecular interactions with diverse targets. / Biol Chem, 21 A, 32555-64. Chakrabartty, A., Kortemme, T. and Baldwin, R.L. (1994) Helix propensities of the amino acids measured in alanine-based peptides without helix-stabilizing side-chain interactions. Protein Sci, 3, 843-52. Chen, H.Y., Weng, S.F. and Lin, J.W. (2000) Identification and analysis of the sap genes from Vibrio fischeri belonging to the ATP-binding cassette gene family required for  150  peptide transport and resistance to antimicrobial peptides. Biochem Biophys Res Commun, 269, 743-8. Chopra, I. (1998) Research and development of antibacterial agents. Curr Opin Microbiol, 1,495-501. Christensen, B., Fink, J., Merrifield, R.B. and Mauzerall, D. (1988) Channel-forming properties of cecropins and related model compounds incorporated into planar lipid membranes. Proc Natl Acad Sci USA, 85, 5072-6. Clayton, A . H . and Sawyer, W.H. (2000) Oriented circular dichroism of a class A amphipathic helix in aligned phospholipid multilayers. Biochim Biophys Acta, 1467, 124-30. Cociancich, S., Ghazi, A., Hoffman, J.A., Hetrus, C. and Letellier, C. (1993) Insect defensin, an inducible antibacterial peptide, forms voltage-dependent channels in Micrococcus luteus. J Biol Chem, 268, 19239-19245. Cole, A . M . and Ganz, T. (2000) Human antimicrobial peptides: analysis and application [In Process Citation]. Biotechniques, 29, 822-6, 828, 830-1. Couto, M.A., Harwig, S.S. and Lehrer, R.I. (1993) Selective inhibition of microbial serine proteases by eNAP-2, an antimicrobial peptide from equine neutrophils. Infect Immun, 61, 2991 -4. Daher, K.A., Selsted, M . E . and Lehrer, R.I. (1986) Direct inactivation of viruses by human granulocyte defensins. J Virol, 60, 1068-74. Daugelavicius, R., Bakiene, E. and Bamford, D.H. (2000) Stages of polymyxin B interaction with the escherichia coli cell envelope [In Process Citation]. Antimicrob Agents Chemother, 44, 2969-78. Duclohier, H , Molle, G. and Spach, G. (1989) Antimicrobial peptide magainin I from Xenopus skin forms anion-permeable channels in planar lipid bilayers. Biophys I, 56, 1017-21. Epand, R.M., Shai, Y., Segrest, J.P. and Anantharamaiah, G.M. (1995) Mechanisms for the modulation of membrane bilayer properties by amphipathic helical peptides. Biopolymers, 37, 319-38. Falla, T.J. and Hancock, R.E. (1997) Improved activity of a synthetic indolicidin analog. Antimicrob Agents Chemother, 41, 771-5.  151  Falla, T.J., Karunaratne, D.N. and Hancock, R.E.W. (1996) Mode of action of the antimicrobial peptide indolicidin. J Biol Chem, 271, 19298-19303. Fehlbaum, P., Bulet, P., Michaut, L . , Lagueux, M . , Broekaert, W.F., Hetru, C. and Hoffmann, J.A. (1994) Insect immunity. Septic injury of Drosophila induces the synthesis of a potent antifungal peptide with sequence homology to plant antifungal peptides. J Biol Chem, 269, 33159-63. Fields, P.L, Groisman, E.A. and Heffron, F. (1989) A Salmonella locus that controls resistance to microbicidal proteins from phagocytic cells. Science, 243, 1059-1062. Fink, J., Boman, A., Boman, H.G. and Merrifield, R.B. (1989) Design, synthesis and antibacterial activity of cecropin-like model peptides. Int J Pep Protein Res, 33, 412-421. Friedrich, C , Scott, M . G . , Karunaratne, N., Yan, H. and Hancock, R.E. (1999) Saltresistant alpha-helical cationic antimicrobial peptides. Antimicrob Agents Chemother, 43, 1542-8. Friedrich, C.L., Moyles, D., Beveridge, T.J. and Hancock, R.E. (2000) Antibacterial action of structurally diverse cationic peptides on gram-positive bacteria.  Antimicrob Agents Chemother, 44, 2086-92. Gennaro, R. and Zanetti, M . (2000) Structural features and biological activities of the cathelicidin-derived antimicrobial peptides. Biopolymers, 55, 31-49. Giacometti, A., Cirioni, O., Barchiesi, F., Del Prete, M.S., Fortuna, M . , Caselli, F. and Scalise, G. (2000) In vitro susceptibility tests for cationic peptides: comparison of broth microdilution methods for bacteria that grow aerobically. Antimicrob Agents Chemother, 44, 1694-6. Giacometti, A., Cirioni, O., Barchiesi, F., Fortuna, M . and Scalise, G. (1999) In-vitro activity of cationic peptides alone and in combination with clinically used antimicrobial agents against Pseudomonas aeruginosa. I Antimicrob Chemother, 44, 641-5. Goldman, M.J., Anderson, G.M., Stolzenberg, E.D., Kari, U.P., Zasloff, M . and Wilson, J.M. (1997) Human beta-defensin-1 is a salt-sensitive antibiotic in lung that is inactivated in cystic fibrosis. Cell, 88, 553-560.  152  Gottfredsson, M . , Erlendsdottir, H., Kolka, R., Gudmundsson, A. and Gudmundsson, S. (1993) Ultrastructural alterations of bacteria during the postantibiotic effect. Chemotherapy, 39, 153-62. Gough, M . , Hancock, R.E. and Kelly, N.M. (1996) Antiendotoxin activity of cationic peptide antimicrobial agents. Infect Immun, 64, 4922-7. Grinsted, J., Saunders, J.R., Ingram, L.C., Sykes, R.B. and Richmond, M.H. (1972) Properties of a R factor which originated in Pseudomonas aeruginosa 1822. J Bacterial, 110, 529-37. Groisman, E.A., Kayser, J. and Soncini, F.C. (1997) Regulation of polymyxin resistance and adaptation to low-Mg2+ environments. J Bacteriol, 179, 7040-5. Hancock, R.E. (1981) Aminoglycoside uptake and mode of action-with special reference to streptomycin and gentamicin. U. Effects of aminoglycosides on cells. / Antimicrob Chemother, 8, 429-45. Hancock, R.E. (1997) Antibacterial peptides and the outer membranes of gram-negative bacilli [editorial]. J Med Microbiol, 46, 1-3. Hancock, R.E. and Carey, A . M . (1979) Outer membrane of Pseudomonas aeruginosa: heat- 2-mercaptoethanol- modifiable proteins. J Bacteriol, 140, 902-10. Hancock, R.E. and Chappie, D.S. (1999) Peptide antibiotics. Antimicrob Agents Chemother, 43, 1317-23. Hancock, R.E. and Lehrer, R. (1998) Cationic peptides: a new source of antibiotics.  Trends Biotechnol, 16, 82-8. Hancock, R.E.W., Falla, T.J. and Brown, M . (1995) Cationic Antimicrobial Peptides. Advances in Microbial Physiology, 37, 136-175. Harold, F.M., Baarda, J.R. and Pavlasova, E. (1970) Extrusion of sodium and hydrogen ions as the primary process in potassium ion accumulation by Streptococcus faecalis. J Bacteriol, 101, 152-9. He, K., Ludtke, S.J., Huang, H.W. and Worcester, D.L. (1995) Antimicrobial peptide pores in membranes detected by neutron in-plane scattering. Biochemistry, 34, 15614-8. Holzarth, G.M. and Doty, P. (1965) The ultraviolet circular dichroism of polypeptides. J. Am. Chem. Soc, 87, 218-28.  153  Houghten, R.A., Pinilla, C , Blondelle, S.E., Appel, J.R., Dooley, C.T. and Cuervo, J.H. (1991) Generation and use of synthetic peptide combinatorial libraries for basic research and drug discovery. Nature, 354, 84-6. Huang, H.W. (2000) Action of antimicrobial peptides: two-state model. Biochemistry, 39, 8347-52. Hultmark, D., Steiner, H., Rasmusen, T. and Boman, H.G. (1980) Insect immunity. Purification and properties of three inducible bactericidal proteins from hemolymph of immunized pupae of Hyalophora cecropia. European Journal of Biochemistry, 106, 7-16. Jach, G., Gornhardt, B., Mundy, J., Logemann, J., Pinsdorf, E., Leah, R., Schell, J. and Maas, C. (1995) Enhanced quantitative resistance against fungal disease by combinatorial expression of different barley antifungal proteins in transgenic tobacco. Plant J, 8, 97-109. Jack, R.W., Tagg, J.R. and Ray, B. (1995) Bacteriocins of gram-positive bacteria. Microbiol Rev, 59, 171-200. Juretic, D., Chan, H.C., Brown, J.H., Morell, J.L., Hendler, R.W. and Westerhoff, H . (1989) Magainin 2 amide and analogues, antimicrobial activity, membrane depolarization and susceptibility to proteolysis. FEBS Lett, 249, 219-223. Kemper, M.A., Urrutia, M . M . , Beveridge, T.J., Koch, A.L. and Doyle, R.J. (1993) Proton motive force may regulate cell wall-associated enzymes of Bacillus subtilis. J Bacteriol, 175, 5690-6. Kobayashi, S., Takeshima, K., Park, C.B., Kim, S.C. and Matsuzaki, K. (2000) Interactions of the novel antimicrobial peptide buforin 2 with lipid bilayers: proline as a translocation promoting factor. Biochemistry, 39, 8648-54. Koo, S.P., Bayer, A.S., Sahl, H.G., Proctor, R.A. and Yeaman, M.R. (1996) Staphylocidal action of thrombin-induced platelet microbicidal protein is not solely dependent on transmembrane potential. Infect Immun, 64, 1070-4. Koo, S.P., Yeaman, M.R., Nast, C C . and Bayer, A.S. (1997) The cytoplasmic membrane is a primary target for the staphylocidal action of thrombin-induced platelet microbicidal protein. Infect Immun, 65, 4795-800.  154  Kordel, M . , Benz, R. and Sahl, H.G. (1988) Mode of action of the staphylococcinlike peptide Pep 5: voltage- dependent depolarization of bacterial and artificial membranes. J Bacteriol, 170, 84-8. Krijgsveld, J., Zaat, S.A., Meeldijk, J., van Veelen, P.A., Fang, G., Poolman, B., Brandt, E., Ehlert, J.E., Kuijpers, A.J., Engbers, G.H., Feijen, J. and Dankert, J. (2000) Thrombocidins, microbicidal proteins from human blood platelets, are C-terminal deletion products of C X C chemokines. J Biol Chem, 275, 20374-81. Kuhl, S.A., Pattee, P.A. and Baldwin, J.N. (1978) Chromosomal map location of the methicillin resistance determinant in Staphylococcus aureus. J Bacteriol, 135, 4605. Kupferwasser, L.I., Skurray, R.A., Brown, M.H., Firth, N., Yeaman, M.R. and Bayer, A.S. (1999) Plasmid-mediated resistance to thrombin-induced platelet microbicidal protein in staphylococci: role of the qacA locus. Antimicrob Agents Chemother, 43, 2395-9. Ladokhin, A.S., Selsted, M . E . and White, S.H. (1997) Bilayer interactions of indolicidin, a small antimicrobial peptide rich in tryptophan, proline, and basic amino acids. Biophys J, 72, 794-805. Ladokhin, A.S., Selsted, M . E . and White, S.H. (1999) C D spectra of indolicidin antimicrobial peptides suggest turns, not polyproline helix. Biochemistry, 38, 12313-9. Lakowicz, J.R. (1983) Principles of fluorescence spectroscopy. Plenum Press, New York. Lehrer, R.L, Barton, A., Daher, K.A., Harwig, S.S., Ganz, T. and Selsted, M . E . (1989) Interaction of human defensins with Escherichia coli. Mechanism of bactericidal activity. J Clin Invest, 84, 553-61. Lehrer, R.L, Ganz, T. and Selsted, M . E . (1991) Defensins: endogenous antibiotic peptides of animal cells. Cell, 64, 229-30. Letellier, L . and Shechter, E . (1979) Cyanine dye as monitor of membrane potentials in Escherichia coli cells and membrane vesicles. Eur J Biochem, 102, 441-7. Li, M.L., Liao, R.W., Qiu, J.W., Wang, Z.J. and Wu, T . M . (2000) Antimicrobial activity of synthetic all-D mastoparan M . Int J Antimicrob Agents, 13, 203-8.  155  Manavalan, P. and Johnson, W . C , Jr. (1987) Variable selection method improves the prediction of protein secondary structure from circular dichroism spectra. Anal Biochem, 161,76-85. Martinez-Cuesta, M.C., Kok, J., Herranz, E., Pelaez, C , Requena, T. and Buist, G. (2000) Requirement of autolytic activity for bacteriocin-induced lysis [In Process Citation]. Appl Environ Microbiol, 66, 3174-9. Matsuyama, K. and Natori, S. (1988) Purification of three antibacterial proteins from the culture medium of NIH-Sape-4, an embryonic cell line of Sarcophaga peregrina. J Biol Chem, 263, 17112-6. Matsuzaki, K., Sugishita, K., Fujii, N. and Miyajima, K. (1995) Molecular basis for membrane selectivity of an antimicrobial peptide, magainin 2. Biochemistry, 34, 3423-9. Matsuzaki, K., Sugishita, K., Ishibe, N., Ueha, M . , Nakata, S., Miyajima, K. and Epand, R.M. (1998) Relationship of membrane curvature to the formation of pores by magainin 2. Biochemistry, 37, 11856-63. Matsuzaki, K., Sugishita, K. and Miyajima, K. (1999) Interactions of an antimicrobial peptide, magainin 2, with lipopolysaccharide-containing liposomes as a model for outer membranes of gram-negative bacteria. FEBS Lett, 449, 221-4. Matsuzaki, K., Yoneyama, S. and Miyajima, K. (1997) Pore formation and translocation of melittin. Biophys J, 73, 831-8. Mayo, K.H., Haseman, J., Young, H.C. and Mayo, J.W. (2000) Structure-function relationships in novel peptide dodecamerswith broad-spectrum bactericidal and endotoxin-neutralizing activities. Biochem J, 349, 717-728. McAuliffe, O., Ryan, M.P., Ross, R.P., Hill, C , Breeuwer, P. and Abee, T. (1998) Lacticin 3147, a broad-spectrum bacteriocin which selectively dissipates the membrane potential. Appl Environ Microbiol, 64, 439-45. Meers, P. (1990) Location of tryptophans in membrane-bound annexins. Biochem, 29, 3325-3330. Nagaoka, I., Hirota, S., Yomogida, S., Ohwada, A. and Hirata, M . (2000) Synergistic actions of antibacterial neutrophil defensins and cathelicidins. Inflamm Res, 49, 739.  156  Nagpal, S., Gupta, V., Kaur, K.J. and Salunke, D . M . (1999) Structure-function analysis of tritrypticin, an antibacterial peptide of innate immune origin. J Biol Chem, 21 A, 23296-304. Nishino, T., Wecke, J., Kruger, D. and Giesbrecht, P. (1987) Trimethoprim-induced structural alterations in Staphylococcus aureus and the recovery of bacteria in drugfree medium. J Antimicrob Chemother, 19, 147-59. O'Connell, B.C., Xu, T., Walsh, T.J., Sein, T., Mastrangeli, A., Crystal, R.G., Oppenheim, F.G. and Baum, B.J. (1996) Transfer of a gene encoding the anticandidal protein histatin 3 to salivary glands. Hum Gene Ther, 1, 2255-61. Oren, Z. and Shai, Y. (2000) Cyclization of a cytolytic amphipathic alpha-helical peptide and its diastereomer: effect on structure, interaction with model membranes, and biological function. Biochemistry, 39, 6103-14. Osapay, K., Tran, D., Ladokhin, A.S., White, S.H., Henschen, A . H . and Selsted, M . E . (2000) Formation and characterization of a single Trp-Trp cross-link in indolicidin that confers protease stability without altering antimicrobial activity. J Biol Chem, 275, 12017-22. Otvos, L., Jr., Bokonyi, K., Varga, I., Otvos, B.L, Hoffmann, R., Ertl, H.C., Wade, J.D., McManus, A . M . , Craik, D.J. and Bulet, P. (2000a) Insect peptides with improved protease-resistance protect mice against bacterial infection. Protein Sci, 9, 742-9. Otvos, L., Jr., O, I., Rogers, M.E., Consolvo, P.J., Condie, B.A., Lovas, S., Bulet, P. and Blaszczyk-Thurin, M . (2000b) Interaction between Heat Shock Proteins and Antimicrobial Peptides. Biochemistry, 39, 14150-14159. Park, C.B., Kim, H.S. and Kim, S.C. (1998) Mechanism of action of the antimicrobial peptide buforin H: buforin II kills microorganisms by penetrating the cell membrane and inhibiting cellular functions. Biochem Biophys Res Commun, 244, 253-7. Park, C.B., Yi, K.S., Matsuzaki, K., Kim, M.S. and Kim, S.C. (2000) Structure-activity analysis of buforin n, a histone H2A-derived antimicrobial peptide: the proline hinge is responsible for the cell-penetrating ability of buforin U. Proc Natl Acad Sci U S A, 91, 8245-50.  157  Pathak, N., Salas-Auvert, R., Ruche, G., Janna, M.H., McCarthy, D. and Harrison, R.G. (1995) Comparison of the effects of hydrophobicity, amphiphilicity, and alphahelicity on the activities of antimicrobial peptides. Proteins, 22, 182-6. Perczel, A. (1991) J. Am. Chem. Soc., 113, 9772-84. Perczel, A., Park, K. and Fasman, G.D. (1992) Analysis of the circular dichroism spectrum of proteins using the convex constraint algorithm: a practical guide. Anal Biochem, 203, 83-93. Perez-Paya, E., Houghten, R.A. and Blondelle, S.E. (1995) The role of amphipathicity in the folding, self-association and biological activity of multiple subunit small proteins. J Biol Chem, 270, 1048-56. Peschel, A., Otto, M . , Jack, R.W., Kalbacher, H., Jung, G. and Gotz, F. (1999) Inactivation of the dlt operon in Staphylococcus aureus confers sensitivity to defensins, protegrins, and other antimicrobial peptides. J Biol Chem, 274, 8405-10. Peterson, A . A . , Fesik, S.W. and McGroarty, E.J. (1987) Decreased binding of antibiotics to lipopolysaccharide from polymyxin-resistant strains of Escherichia coli and Salmonella typhimurium. Antimicrob Agents Chemother, 31, 230-237. Piers, K.L., Brown, M . H . and Hancock, R.E. (1993) Recombinant D N A procedures for producing small antimicrobial cationic peptides in bacteria. Gene, 134, 7-13. Piers, K.L., Brown, M . H . and Hancock, R.E.W. (1994) Improvement of outer membranepermeabilizing and lipopolysaccharide- binding activities of an antimicrobial cationic peptide by C-terminal modification. Antimicrob Agents Chemother, 38, 2311-6. Piers, K.L. and Hancock, R.E.W. (1994) The interaction of a recombinant cecropin/melittin hybrid peptide with the outer membrane of Pseudomonas aeruginosa. Mol Microbiol, 12, 951-8. Raj, P.A., Karunakaran, T. and Sukumaran, D.K. (2000) Synthesis, microbicidal activity and solution structure of the dodecapeptide from bovine neutrophils. Biopolymers, 53,281-92. Robinson, W.E., Jr., McDougall, B., Tran, D. and Selsted, M . E . (1998) Anti-HIV-1 activity of indolicidin, an antimicrobial peptide from neutrophils. J Leukoc Biol, 63, 94-100.  158  Romeo, D., Skerlavaj, B., Bolognesi, M . and Gennaro, R. (1988) Structure and bactericidal activity of an antibiotic dodecapeptide purified from bovine neutrophils. J Biol Chem, 263, 9573-5. Rozek, A., Friedrich, C.L. and Hancock, R.E. (2000) Structure of the bovine antimicrobial peptide indolicidin bound to dodecylphosphocholine and sodium dodecyl sulfate micelles. Biochemistry, 39, 15765-74. Sawyer, J.G., Martin, N.L. and Hancock, R.E. (1988) Interaction of macrophage cationic proteins with the outer membrane of Pseudomonas aeruginosa. Infect Immun, 56, 693-8. Schiffer, M . and Edmundson, A.B. (1967) Use of helical wheels to represent the structures of protein and to identify segments with helical potential. Biophys J, 7, 121-135. Scott, M . G . , Gold, M.R. and Hancock, R.E.W. (1999) Interaction of Cationic Peptides with Lipoteichoic Acid and Gram-Positive Bacteria. Infect Immun, 67, 6445-53. Selsted, M . E . , Novotny, M.J., Morris, W.L., Tang, Y.Q., Smith, W. and Cullor, J.S. (1992) Indolicidin, a novel bactericidal tridecapeptide amide from neutrophils. / Biol Chem, 267, 4292-5. Shafer, W . M . , Qu, X., Waring, A.J. and Lehrer, R.I. (1998) Modulation of Neisseria gonorrhoeae susceptibility to vertebrate antibacterial peptides due to a member of the resistance/nodulation/division efflux pump family. Proc Natl Acad Sci USA, 95, 1829-33. Shimoda, M . , Ohki, K., Shimamoto, Y . and Kohashi, O. (1995) Morphology of defensintreated Staphylococcus aureus. Infect Immun, 63, 2886-91. Shin, S.Y., Kang, J.H., Jang, S.Y., Kim, Y., Kim, K.L. and Hahm, K.S. (2000) Effects of the hinge region of cecropin A(l-8)-magainin 2(1-12), a synthetic antimicrobial peptide, on liposomes, bacterial and tumor cells. Biochim Biophys Acta, 1463, 20918. Silvestro, L., Gupta, K., Weiser, J.N. and Axelsen, P.H. (1997) The concentrationdependent membrane activity of cecropin A [published erratum appears in Biochemistry 1999 Mar 23;38(12):3850]. Biochemistry, 36, 11452-60.  159  Silvestro, L., Weiser, J.N. and Axelsen, P.H. (2000) Antibacterial and antimembrane activities of cecropin A in Escherichia coli. Antimicrob Agents Chemother, 44, 6027. Sims, P.J., Waggoner, A.S., Wang, C H . and Hoffman, J.F. (1974) Studies on the mechanism by which cyanine dyes measure membrane potential in red blood cells and phosphatidylcholine vesicles. Biochemistry, 13, 3315-30. Storici, P., Tossi, A., Lenarcic, B. and Romeo, D. (1996) Purification and structural characterization of bovine cathelicidins, precursors of antimicrobial peptides. Eur J Biochem, 238, 769-76. Subbalakshmi, C , Bikshapathy, E., Sitaram, N. and Nagaraj, R. (2000) Antibacterial and hemolytic activities of single tryptophan analogs of indolicidin. Biochem Biophys Res Commun, 274, 714-6. Subbalakshmi, C , Krishnakumari, V., Nagaraj, R. and Sitaram, N. (1996) Requirements for antibacterial and hemolytic activities in the bovine neutrophil derived 13-residue peptide indolicidin. FEBS Lett, 395, 48-52. Subbalakshmi, C. and Sitaram, N. (1998) Mechanism of antimicrobial action of indolicidin. FEMS Microbiol Lett, 160, 91-6. Sudbery, P.E. (1996) The expression of recombinant proteins in yeasts. Curr Opin Biotechnol, 1, 517-24. Travis, S.M., Anderson, N.N., Forsyth, W.R., Espiritu, C , Conway, B.D., Greenberg, E.P., McCray, P.B., Jr., Lehrer, R.L, Welsh, M.J. and Tack, B.F. (2000) Bactericidal activity of mammalian cathelicidin-derived peptides. Infect Immun, 68, 2748-55. Urban, B.W., Hladky, S.B. and Haydon, D.A. (1980) Ion movements in gramicidin pores. An example of single-file transport. Biochim Biophys Acta, 602, 331-54. van den Berg, R.H., Faber-Krol, M . C , van Wetering, S., Hiemstra, P.S. and Daha, M.R. (1998) Inhibition of activation of the classical pathway of complement by human neutrophil defensins. Blood, 92, 3898-903. Wachinger, M . , Kleinschmidt, A., Winder, D., von Pechmann, N., Ludvigsen, A., Neumann, M . , Holle, R., Salmons, B., Erfle, V. and Brack-Werner, R. (1998) Antimicrobial peptides melittin and cecropin inhibit replication of human  160  immunodeficiency virus 1 by suppressing viral gene expression. J Gen Virol, 79, 731-40. Wade, D., Andreu, D., Mitchell, S.A., Silveira, A . M . , Boman, A., Boman, H.G. and Merrifield, R.B. (1992) Antibacterial peptides designed as analogs or hybrids of cecropins and melittin. Int J Pept Protein Res, 40, 429-36. Wade, D., Boman, B., Wahlin, B., Drain, C M . , Andreu, D., Boman, H.G. and Merrifield, R.B. (1990) All-D amino acid-containing channel-forming antibiotic peptides. Proc Natl Acad Sci USA, 87, 4761-4765. Wallace, B.A. and Mao, D. (1984) Circular dichroism analyses of membrane proteins: an examination of differential light scattering and absorption flattening effects in large membrane vesicles and membrane sheets. Anal Biochem, 142, 317-28. Welling, M . M . , Hiemstra, P.S., van den Barselaar, M.T., Paulusma-Annema, A., Nibbering, P.H., Pauwels, E.K. and Calame, W. (1998) Antibacterial activity of human neutrophil defensins in experimental infections in mice is accompanied by increased leukocyte accumulation. J Clin Invest, 102, 1583-90. Wimley, W . C and White, S.H. (1996) Experimentally determined hydrophobicity scale for proteins at membrane interfaces. Nat Struct Biol, 3, 842-8. Woody, R.W. (1994) Contributions of tryptophan side chains to the far-ultraviolet circular dichroism of proteins. Eur Biophys J, 23, 253-62. Wu, M . and Hancock, R.E.W. (1999) Interaction of the cyclic antimicrobial cationic peptide bactenecin with the outer and cytoplasmic membrane. J Biol Chem, 274, 29-35. Wu, M . , Maier, E., Benz, R. and Hancock, R.E.W. (1999) Mechanism of interaction of different classes of cationic antimicrobial peptides with planar bilayers and with the cytoplasmic membrane of Escherichia coli. Biochem, 38, 7235-42. Xiong, Y.Q., Yeaman, M.R. and Bayer, A.S. (1999) In vitro antibacterial activities of platelet microbicidal protein and neutrophil defensin against Staphylococcus aureus are influenced by antibiotics differing in mechanism of action. Antimicrob Agents Chemother, 43, 1111-7. Yeaman, M.R., Bayer, A.S., Koo, S.P., Foss, W. and Sullam, P.M. (1998) Platelet microbicidal proteins and neutrophil defensin disrupt the Staphylococcus aureus  161  cytoplasmic membrane by distinct mechanisms of action. J Clin Invest, 101, 17887. Zanetti, M . , Litteri, L., Gennaro, R., Horstmann, H. and Romeo, D. (1990) Bactenecins, defense polypeptides of bovine neutrophils, are generated from precursor molecules stored in the large granules. J Cell Biol, 111, 1363-71. Zhang, H., Yoshida, S., Aizawa, T., Murakami, R., Suzuki, M . , Koganezawa, N., Matsuura, A., Miyazawa, M . , Kawano, K., Nitta, K. and Kato, Y. (2000a) In vitro antimicrobial properties of recombinant ASABF, an antimicrobial peptide isolated from the nematode ascaris suum [In Process Citation]. Antimicrob Agents Chemother, 44, 2701-5. Zhang, L., Benz, R. and Hancock, R.E.W. (1999) Influence of proline residues on the antibacterial and synergistic activities of alpha-helical peptides. Biochem, 38, 81028111. Zhang, L., Dhillon, P., Yan, H., Farmer, S. and Hancock, R.E. (2000b) Interactions of bacterial cationic peptide antibiotics with outer and cytoplasmic membranes of pseudomonas aeruginosa [In Process Citation]. Antimicrob Agents Chemother, 44, 3317-21.  162  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0104989/manifest

Comment

Related Items