UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigating the structure-function relationship of cationic antimicrobial peptides and lipopeptides Cheng, John Tien Jui 2010

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

Item Metadata

Download

Media
24-ubc_2010_fall_cheng_john_tien_jui.pdf [ 18.19MB ]
Metadata
JSON: 24-1.0060257.json
JSON-LD: 24-1.0060257-ld.json
RDF/XML (Pretty): 24-1.0060257-rdf.xml
RDF/JSON: 24-1.0060257-rdf.json
Turtle: 24-1.0060257-turtle.txt
N-Triples: 24-1.0060257-rdf-ntriples.txt
Original Record: 24-1.0060257-source.json
Full Text
24-1.0060257-fulltext.txt
Citation
24-1.0060257.ris

Full Text

I VESTIGATI G THE STRUCTURE-FU CTIO RELATIO SHIP OF CATIO IC A TIMICROBIAL PEPTIDES A D LIPOPEPTIDES by John Tien Jui Cheng M.Sc., Simon Fraser University, Canada, 2006 B.Sc., Simon Fraser University, Canada, 2004  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in THE FACULTY OF GRADUATE STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2010  © John Tien Jui Cheng, 2010  ABSTRACT Antibiotics have been playing a major role in combating bacterial infections for centuries. Since the discovery of modern antibiotics, numerous derivatives have been designed and developed to treat different bacterial infections.  Recently, antibiotic resistance has been  continuously and increasingly reported. The lack of antibiotic alternatives makes these resistant bacteria become more difficult to eliminate. Antimicrobial peptides constitute a major part of the innate immune system of an organism. Their high activity and little resistance make them ideal candidates for novel antibiotic development. This dissertation focuses on aurein peptides, a class of amphibian cationic antimicrobial peptides from Litoria aurea, and daptomycin, a lipopeptide. We have examined the structurefunction relationship of two aurein peptides, aurein 2.2 and aurein 2.3. They were found to adopt α-helices and perturb membrane bilayers via mechanisms similar to toroidal pore or toroidal pore/liposome formation in model membranes.  We have also designed and inspected the  structure-activity correlation of different aurein 2.2 analogues by residue 13-substitutions and Nand/or C-terminal truncations. We have found that residue 13 and N-terminus are required for antimicrobial activity, whereas an N-terminal truncation gives rise to a peptide analogue with immunomodulatory activity in vitro.  The effects of membrane composition and model  membrane choice have been further investigated. We have found that the peptide behaviour is dependent on different model membranes.  We have examined the importance of solvent  accessibility in the mechanism of action for daptomycin and found that daptomycin molecules are indeed solvent-exposed in apo- and Ca2+-form and insert slightly into lipid membranes.  ii  Taken together, we have developed a set of references for future design of new antibiotics based on aurein peptides. By using this set of references as a starting point, we hope to gain a better understanding of how antimicrobial peptides function from structural and membrane perspectives and design novel antimicrobial agents to combat increasing antibiotic resistance in the future.  iii  PREFACE The work presented in this dissertation is the outcome of various collaborations with other scientists. Their contributions are summarized as follows. Chapter 2 has been published as: “Pan, Y.-L., Cheng, J. T. J., Hale, J. Pan, J. Hancock, R. E. W. and Straus, S. K. 2007. Characterization of the Structure and Membrane Interaction of the Antimicrobial Peptides Aurein 2.2 and 2.3 from Australian Southern Bell Frogs. Biophys. J. 92: 2854-2864”. My supervisor, Dr. S. K. Straus, designed the experiments. Dr. J. D. Hale performed MIC assays and Dr. J. Pan helped Y.-L. Pan with NMR analyses. I refined and optimized all sample preparation and experimental techniques. Both Y.-L. Pan and I are equally contributing authors. Dr. S. K. Straus wrote the manuscript with inputs from both Y.-L. Pan and me. I wrote this chapter and Dr. S. K. Straus edited this chapter. Chapter 3 has been published as: “Cheng, J. T. J., Hale, J. D., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2009. Effect of Membrane Composition on Antimicrobial Peptides Aurein 2.2 and 2.3 From Australian Southern Bell Frogs”. Biophys. J. 96: 552–565. My supervisor, Dr. S. K. Straus, and I designed the experiments. Dr. J. D. Hale conducted DiSC35 assays and M. Elliot performed calcein release assays. I performed all other experiments and analyzed all data. I wrote the manuscript and Dr. S. K. Straus edited the manuscript. Chapter 4 has been submitted as: “Cheng, J. T. J., Hale, J. D., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2010. The Importance of Bacterial Membrane Composition on the Structure/Function of Aurein 2.2 and some of its Mutants”. My supervisor, Dr. S. K. Straus, and I conceived the research.  Dr. H. Jenssen designed the two C-terminally truncated peptide  iv  analogues. Both Dr. J. D. Hale and M. Elliot conducted MIC assays on P. aeruginosa and E. coli. I performed all other experiments and analyzed all data. I wrote the manuscript and Dr. S. K. Straus edited the manuscript. Chapter 5 has been published as: “Cheng, J. T. J., Hale, J. D., Kindrachuk, J., Jenssen, H., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2010. The Importance of Residue 13 and the C-terminus on the Structure and Activity of the Antimicrobial Peptide Aurein 2.2”.  My  supervisor, Dr. S. K. Straus, and I conceived the research. I designed the three residue 13substituted mutants and Dr. H. Jenssen designed the two C-terminally truncated peptide analogues. Both Dr. J. D. Hale and M. Elliot performed MIC assays on P. aeruginosa and E. coli. I conducted all other experiments and analyzed all data. I wrote the manuscript and Dr. S. K. Straus edited the manuscript. Chapter 6 was a collaborative project. My supervisor, Dr. S. K. Straus, designed the research. Dr. H. Jenssen designed the two N-terminally truncated peptide analogues and I designed L13A mutant.  Dr. J. Kindrachuk conducted immunomodulatory activity assays,  cytotoxicity assays, and cell membrane permeation assays. Dr. J. D. Hale performed MIC assays. I conducted all other experiments and analyzed all data. I wrote this chapter and both Dr. S. K. Straus and Dr. J. Kindrachuk edited this chapter. Chapter 7 has been completed by both Dr. S. K. Straus and me. Dr. S. K. Straus and I designed the experiments. I performed all the NMR experiments with Dr. S. K. Straus and analyzed all data. I wrote this chapter and Dr. S. K. Straus edited this chapter. Overall, 85% of the work presented in this dissertation is from me with Dr. S. K. Straus’ professional guidance, and 15% of the work presented in this dissertation is from collaborations with Hancock lab.  v  TABLE OF CO TE TS Abstract .............................................................................................................................. ii Preface ............................................................................................................................... iv Table of contents .............................................................................................................. vi List of tables....................................................................................................................... x List of figures ................................................................................................................... xii List of symbols and abbreviations ................................................................................ xvi Acknowledgements ........................................................................................................ xix Dedication ........................................................................................................................ xx Chapter 1: Introduction ................................................................................................... 1 1.1 Antibiotics ...................................................................................................... 1 1.1.1 History of antibiotics .................................................................................. 1 1.1.2 Categories and classes of antibiotics .......................................................... 2 1.1.3 Examples of conventional antibiotics ......................................................... 4 1.1.4 Antibiotic resistance ................................................................................... 6 1.1.5 Structure of bacteria .................................................................................. 12 1.2 Alternative antibiotics – antimicrobial peptides ........................................... 14 1.2.1 Cationic antimicrobial peptides (CAPs) ................................................... 15 1.2.2 Acidic lipopeptides ................................................................................... 18 1.2.3 Mechanisms of action ............................................................................... 19 1.3 Amphibian antimicrobial peptides................................................................ 21 1.3.1 Citropins ................................................................................................... 22 1.3.2 Aurein peptides ......................................................................................... 23 1.4 Studying antimicrobial peptides in model systems and real bacteria ........... 24 1.4.1 CD spectroscopy ....................................................................................... 24 1.4.2 Nuclear magnetic resonance (NMR) spectroscopy .................................. 29 1.4.3 Membrane leakage assays and antimicrobial activity .............................. 33 1.5 Aim of this dissertation ................................................................................ 35 1.5.1 Structure-function relationship of aurein 2 peptides ................................ 36 1.5.2 Choice of lipid composition for model membranes ................................. 37 1.5.3 Designing CAP analogues ........................................................................ 38 1.5.4 Lipopeptides ............................................................................................. 39 1.5.5 Conclusion ................................................................................................ 39 1.6 References .................................................................................................... 40 Chapter 2: Characterization of the structure and membrane interaction of the antimicrobial peptides aurein 2.2 and 2.3 from Australian Southern Bell frogs ...... 44 2.1 Introduction .................................................................................................. 44 2.2 Results .......................................................................................................... 45 vi  2.2.1 2.2.2 2.2.3 2.3 2.4 2.5 2.5.1 2.5.2 2.6  Determining the structure of the aurein peptides ...................................... 46 Determining the membrane insertion states and the membrane perturbation mechanisms of the aurein peptides ...................................... 54 Antimicrobial activity of the aurein peptides ........................................... 65 Discussion..................................................................................................... 66 Summary and conclusion ............................................................................. 69 Materials and methods .................................................................................. 70 Materials ................................................................................................... 70 Methods .................................................................................................... 70 References .................................................................................................... 77  Chapter 3: The effect of membrane composition on the antimicrobial peptides aurein 2.2 and 2.3 ............................................................................................................ 79 3.1 Introduction .................................................................................................. 79 3.2 Results .......................................................................................................... 81 3.2.1 Secondary structure of the aurein peptides ............................................... 81 3.2.2 Membrane insertion states of the aurein peptides..................................... 83 3.2.3 Perturbation of lipid headgroups by the aurein peptides .......................... 86 3.2.4 Perturbation of lipid chains by the aurein peptides .................................. 88 3.2.5 Membrane leakage induced by the aurein peptides in model membranes 89 3.2.6 Aurein-induced membrane leakage in S. aureus ...................................... 92 3.3 Discussion..................................................................................................... 93 3.4 Summary and conclusion ............................................................................. 97 3.5 Materials and methods .................................................................................. 98 3.5.1 Materials ................................................................................................... 98 3.5.2 Methods .................................................................................................... 98 3.6 References .................................................................................................. 103 Chapter 4: The importance of bacterial membrane composition: CD and MR studies of the structure and activity of aurein 2.2 and aurein 2.3 in CL/POPG and POPE/POPG model membranes ................................................................................. 106 4.1 Introduction ................................................................................................ 106 4.2 Results ........................................................................................................ 109 4.2.1 Secondary structure of the aurein peptides ............................................. 110 4.2.2 Membrane insertion states of the aurein peptides................................... 113 4.2.3 Lipid headgroup perturbation by the aurein peptides ............................. 116 4.2.4 Antimicrobial activity of the aurein peptides ......................................... 119 4.3 Discussion................................................................................................... 121 4.4 Summary and conclusion ........................................................................... 124 4.5 Materials and methods ................................................................................ 124 4.5.1 Materials ................................................................................................. 124 4.5.2 Methods .................................................................................................. 125 4.6 References .................................................................................................. 127 Chapter 5: The importance of residue 13 and the C-terminus on the structure and activity of the amphibian antimicrobial peptide aurein 2.2 ...................................... 130 5.1 Introduction ................................................................................................ 130 5.2 Results ........................................................................................................ 132 5.2.1 Antimicrobial activity of aurein mutants ................................................ 133 5.2.2 Secondary structure of aurein mutants ................................................... 134 vii  5.2.3 5.2.4 5.2.5 5.2.6 5.3 5.4 5.5 5.6 5.6.1 5.6.2 5.7  Membrane insertion state of aurein mutants ........................................... 138 Lipid headgroup perturbation by aurein mutants.................................... 142 Using DiSC35 assay to observe the bacterial membrane leakage induced by aurein mutants .................................................................................... 148 Antimicrobial activity of aurein mutants against bacteria containing PE ................................................................................................................ 150 Discussion................................................................................................... 151 Summary and conclusion ........................................................................... 156 Supplemental figures .................................................................................. 157 Materials and methods ................................................................................ 158 Materials ................................................................................................. 158 Methods .................................................................................................. 159 References .................................................................................................. 161  Chapter 6: ew immunomodulatory peptides arise from -terminal truncation of the amphibian antimicrobial peptide aurein 2.2 ........................................................ 163 6.1 Introduction ................................................................................................ 163 6.2 Results ........................................................................................................ 166 6.2.1 Immunomodulatory activity in vitro ....................................................... 166 6.2.2 Cytotoxicity activity in vitro ................................................................... 168 6.2.3 Antimicrobial activity ............................................................................. 169 6.2.4 Secondary structure by solution CD spectroscopy ................................. 170 6.2.5 Secondary structure by solution 1H NMR spectroscopy ........................ 175 6.2.6 Membrane insertion states ...................................................................... 176 6.2.7 Membrane permeation ability ................................................................. 178 6.2.8 Immunomodulatory activity of select aurein mutants in a murine model of bacterial infection ................................................................................... 179 6.3 Discussion................................................................................................... 179 6.4 Summary and conclusion ........................................................................... 184 6.5 Materials and methods ................................................................................ 184 6.5.1 Materials ................................................................................................. 184 6.5.2 Methods .................................................................................................. 185 6.6 References .................................................................................................. 190 Chapter 7: The importance of solvent accessibility in the mechanism of action of the lipopeptide daptomycin ................................................................................................ 192 7.1 Introduction ................................................................................................ 192 7.2 Results and discussion ................................................................................ 194 7.2.1 Daptomycin structure.............................................................................. 194 7.2.2 Daptomycin dynamics/solvent accessibility ........................................... 195 7.2.3 Daptomycin location in DHPC micelles in the presence of Ca2+ ........... 197 7.2.4 Daptomycin in the presence of Mn2+ ...................................................... 199 7.3 Implications for the mechanism of action of daptomycin and general conclusions ................................................................................................. 203 7.4 Materials and methods ................................................................................ 205 7.4.1 Materials ................................................................................................. 205 7.4.2 Methods .................................................................................................. 205 7.5 References .................................................................................................. 208 Chapter 8: Summary, conclusion, and future work .................................................. 210 viii  8.1 8.2 8.2.1 8.2.2 8.2.3 8.2.4 8.3 8.3.1 8.3.2 8.4 8.5  Thesis summary .......................................................................................... 210 Experimental conclusions ........................................................................... 212 Perspectives on the peptide structure ...................................................... 212 Perspectives on the peptide-membrane interaction ................................ 214 Perspectives on the model membranes ................................................... 215 Daptomycin............................................................................................. 216 Future work ................................................................................................ 217 Structures and functions ......................................................................... 217 Beyond antimicrobial peptides ............................................................... 219 Final remarks .............................................................................................. 220 References .................................................................................................. 223  Appendices ..................................................................................................................... 225 Appendix A: Experimental protocols ..................................................................... 225 Appendix A.1: Peptide workup .............................................................................. 225 Appendix A.2: HPLC purification ......................................................................... 226 Appendix A.3: Solution CD sample preparation .................................................... 228 Appendix A.4: Oriented CD sample preparation ................................................... 229 Appendix A.5: 31P NMR sample preparation ......................................................... 229 Appendix B: Peptide sequence and molecular weights of all aurein peptides and analogues ................................................................................................ 231  ix  LIST OF TABLES Table 1.1.  Table 1.2.  Table 2.1. Table 2.2.  Table 2.3. Table 3.1. Table 3.2. Table 3.3.  Table 4.1. Table 4.2. Table 5.1. Table 5.2.  Table 5.3. Table 5.4.  Mammalian cationic antimicrobial peptides (Table reproduced from Hancock, R. E. and Diamond, G. 2000 (29)). These peptides adopt one of the four common secondary structures. Note that disulfide bridges are labeled with numbers in subscript, at the relevant positions. ...................... 16 Amino acid sequences, molecular weights (MW, g/mol), charges and minimal inhibitory concentrations (MICs, µg/ml) of aurein 1.2, aurein 2.2, aurein 2.3 and citropin 1.1, according to Rozek, T. et al., 2000 (72).......... 36 Summary of the membrane insertion states of the three aurein peptides in the presence of different lipid bilayers as a function of P/L molar ratio. .... 57 Percentage of calcein released relative to 0.1% Triton X (defined as 100%) from 3:1 DMPC/DMPG (mol/mol) liposomes, in the presence of aurein 2.2, aurein 2.3 and aurein 2.3-COOH at 1:15 P/L molar ratio. (Calcein release assay results are courtesy of M. Elliot). ...................................................... 64 Minimal inhibitory concentrations (MICs) of the three aurein peptides. .... 65 % α-helix of the three aurein peptides in the presence of ddH2O/TFE and different SUVs at different P/L molar ratios. .............................................. 83 Membrane insertion states of the three aurein peptides in the presence of different lipid bilayers at different P/L molar ratios.................................... 85 Percentage of calcein released relative to 0.1% Triton X (defined as 100%) from 1:1 and 3:1 POPC/POPG (mol/mol) liposomes, in the presence of aurein 2.2, aurein 2.3, and aurein 2.3-COOH. The errors associated with the measurements have been determined for the P/L molar ratio of 1:15 in 1:1 and 3:1 POPC/POPG from repeat measurements and are given in the table. (Calcein release assay results are courtesy of M. Elliot)............................. 90 Membrane insertion states of the three aurein peptides in the presence of different lipid bilayers at different P/L molar ratios.................................. 115 MIC results of aurein 2.2, aurein 2.3 and aurein 2.3-COOH against B. cereus, P. aeruginosa and E. coli. ............................................................. 120 Peptide sequences and molecular weights of the two aurein parent peptides and the five aurein mutant peptides. .......................................................... 131 Minimal inhibitory concentrations (MICs) in µg/ml of the five aurein mutants toward S. aureus and S. epidermidis. MICs are given as the most frequently observed value obtained from repeat experiments. (MIC assay results are courtesy of Dr. J. D. Hale). ...................................................... 133 Membrane insertion states of the five aurein mutants in the presence of different lipid bilayers at different P/L molar ratios.................................. 142 Minimal inhibitory concentrations (MICs) in µg/ml of the five aurein mutants and gramicidin S (control) toward B. cereus, P. aeruginosa and E. x  Table 6.1.  Table 8.1.  coli. MICs are given as the most frequently observed value obtained from repeat experiments. (MIC values obtained in P. aeruginosa Pa01 and E. coli C500 are courtesy of Dr. J. D. Hale and M. Elliot). ........................... 151 Minimal inhibitory concentrations (MICs) in µg/ml of the three aurein mutants toward S. aureus and S. epidermidis (Gram positive bacteria), and E. coli and P. aeruginosa (Gram negative bacteria). MICs are given as the most frequently observed value obtained from repeat experiments. (MIC assay results are courtesy of Dr. J. D. Hale and M. Elliot). ...................... 170 Summary of the effects of peptide modification on the aurein peptides. The amino acids that exist in the peptide sequence are shown in green, whereas the ones that do not exist are shown in black. AM = antimicrobial; IM = immunomodulatory; T = toroidal pore/liposome formation. .................... 220  xi  LIST OF FIGURES Figure 1.1. Figure 1.2.  Figure 1.3. Figure 1.4. Figure 1.5. Figure 1.6.  Figure 1.7.  Figure 1.8.  Figure 1.9. Figure 1.10. Figure 1.11. Figure 1.12. Figure 1.13.  Figure 1.14.  Figure 1.15. Figure 1.16.  The molecular (left panel) and 3D (right panel) structures of penicillin and its derivatives:................................................................................................ 4 The (a) molecular and (b) 3D structures of gentamicin. (Figures adapted from “Wikipedia” with minor modification, http://en.wikipedia.org/wiki/Gentamicin). .................................................... 5 The three processes that bacteria acquire antibiotic resistance: .................... 8 Five mechanisms that bacteria develop against antibiotics:.......................... 9 A micrograph of Gram stains of mixed Staphylococcus aureus (Gram positive cocci) and Escherichia coli (Gram negative bacilli): .................... 11 The cross section schematic diagram depicts the major components of Gram positive and Gram negative bacteria. (Figure adapted from “Bacterial morphology” with minor modification, http://micro.digitalproteus.com/morphology2.php). ................................... 12 Differences in the membrane structure between (a) Gram positive and (b) Gram negative bacteria. (Figure adapted from “MJCR/MBMB 403 Microbiology, Southern Illinois University Carbondale” with minor modification, http://www.cehs.siu.edu/fix/medmicro/pix/walls.gif). ......... 13 Microscopic image of Staphylococcus aureus, the superbug (green graphlike colonies). (Figure adapted from http://swampie.files.wordpress.com/2008/02/staphylococcus-aureus.jpg). 14 Various secondary structures of antimicrobial peptides, as selected from Protein Data Bank: ...................................................................................... 17 The main mechanisms of action of antimicrobial peptides against bacterial membranes: ................................................................................................. 20 Standard curves of α-helix, β-sheet and random coil based on solution CD spectra of poly-lysine. ................................................................................. 25 Solution CD spectra of magainin 2 and its analogues in the presence of 1:1 TFE/buffer (v/v): ......................................................................................... 26 Reconstructed spectra of the parallel and perpendicular components of the CD spectrum of an α-helix. The summation represents isotropic distribution of helices. The dashed line is the reference spectrum of polylysine as given by Greenfield and Fasman (86). .................................. 27 OCD spectra of (a) alamethicin and (b) protegrin-1 demonstrate that the peptides adopt the inserted (state I) and surface state (state S) in DPhPC multilayers: .................................................................................................. 28 The energy difference between the two levels in protons (I = ½). .............. 30 31 P NMR spectra of lipid/s in different physical states: .............................. 32 xii  Figure 1.17.  31  P NMR spectra of mechanically aligned 3:1 POPC/POPG (mol/mol) bilayers in the absence and presence of LL7-27 at 37°C: ........................... 33 Figure 2.1. Solution CD spectra of the aurein peptides in ddH2O and ddH2O/TFE mixtures: ...................................................................................................... 47 Figure 2.2. Solution CD spectra of the aurein peptides in DMPC small unilamellar vesicles (SUVs): .......................................................................................... 48 Figure 2.3. Solution CD spectra of the aurein peptides in 1:1 DMPC/DMPG (mol/mol) small unilamellar vesicles (SUVs): ............................................................. 49 Figure 2.4. Fingerprint region of solution-state 1H NMR NOESY spectra of (a) aurein 2.2, (b) aurein 2.3 and (c) aurein 2.3-COOH: ............................................. 51 Figure 2.5. Fingerprint region of solution-state 1H NMR NOESY spectrum of aurein 2.2: ............................................................................................................... 52 Figure 2.6. NMR-derived evidence indicating that the aurein peptides are α-helical: .. 53 Figure 2.7. Oriented CD spectra of a) & d) aurein 2.2, b) & e) aurein 2.3 and c) & f) aurein 2.3-COOH (P/L molar ratios = 1:15 (blue), 1:30 (green), 1:40 (red), 1:80 (black), and 1:120 (grey)) in DMPC bilayers (left panel) and 1:1 DMPC/DMPG (mol/mol) bilayers (right panel): ........................................ 55 Figure 2.8. Solid-state 31P NMR spectra of mechanically aligned DMPC bilayers with and without the three aurein peptides: ......................................................... 58 Figure 2.9. Solid-state 31P NMR spectra of the mechanically aligned 4:1 DMPC/DMPG (mol/mol) bilayers with and without the three aurein peptides: .................. 61 Figure 2.10. DSC thermograms of 1:1 DMPC/DMPG (mol/mol) liposomes in the absence (light grey solid line) and presence of aurein 2.2 (black line), aurein 2.3 (dark grey line), and aurein 2.3-COOH (grey line) at 1:15 P/L molar ratio. ............................................................................................................. 62 Figure 2.11. Calcein release profiles of 3:1 DMPC/DMPG (mol/mol) liposomes in the presence of aurein 2.2 (black solid line), aurein 2.3 (black dotted line), and aurein 2.3-COOH (grey solid line) at 1:15 peptide:lipid molar ratio. (Calcein release assay results are courtesy of M. Elliot)............................. 64 Figure 3.1. Solution CD spectra of the aurein peptides in 3:1 DMPC/DMPG (left panel), 1:1 POPC/POPG (centre panel), and 3:1 POPC/POPG (right panel) (mol/mol) small unilamellar vesicles (SUVs): ............................................ 82 Figure 3.2. Oriented CD spectra of the aurein peptides in 3:1 DMPC/DMPG (left panel), 1:1 POPC/POPG (centre panel), and 3:1 POPC/POPG (right panel) (mol/mol) bilayers: ...................................................................................... 84 Figure 3.3. Solid-state 31P NMR spectra of the mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers in the absence and presence of the three aurein peptides: ..................................................................................................................... 87 Figure 3.4. DSC thermograms of 1:1 POPC/POPG (mol/mol) liposomes in the absence (solid black line) and presence of aurein 2.2 (dotted black line), aurein 2.3 (solid grey line), and aurein 2.3-COOH (dotted grey line) at 1:15 P/L molar ratio. Note that the exotherm at 0°C ~ 5°C is due to water melting transition. ..................................................................................................... 89  xiii  Figure 3.5.  Figure 3.6.  Figure 3.7.  Figure 4.1. Figure 4.2.  Figure 4.3. Figure 4.4. Figure 4.5. Figure 5.1. Figure 5.2. Figure 5.3. Figure 5.4. Figure 5.5. Figure 5.6. Figure 5.7.  Figure 5.8.  Calcein release assay results of 1:1 and 3:1 POPC/POPG (mol/mol) small unilamellar vesicles (SUVs) in the presence of aurein peptides at 1:15 P/L molar ratio. .................................................................................................. 91 Membrane depolarization of S. aureus C622 induced by aurein 2.2 at 1X, 5X MIC; aurein 2.3 at 1X, 5X MIC; and finally, aurein 2.3-COOH at 1X MIC, where the MIC values used were as reported in (1). Gramicidin S (1X MIC) was used as a control. Results are a representative of 2 ~ 3 experiments. The arrow represents the time point where peptide was added. (DiSC35 assay results are courtesy of Dr. J. D. Hale). ................................ 92 Possible models for the perturbation of (a) DMPC/DMPG and (b) POPC/POPG by the aurein peptides. To keep this model simple, some finer points, such as the distinction in the ability of aurein 2.2 to induce membrane depolarization in S. aureus C622 as compared to aurein 2.3 and aurein 2.3-COOH, as well as the possibility of toroidal pore/liposome formation, are obviously not taken into account. The arrows represent leakage. ........................................................................................................ 95 Solution CD spectra of the aurein peptides in 1:1 CL/POPG (mol/mol) (left panel) and 1:1 POPE/POPG (mol/mol) (right panel) SUVs: .................... 111 Structure content of the aurein peptides in the presence of (a) 1:1 POPC/POPG (mol/mol), (b) 1:1 CL/POPG (mol/mol), and (c) 1:1 POPE/POPG (mol/mol) SUVs: ................................................................. 112 Oriented CD spectra of the aurein peptides in 1:1 CL/POPG (mol/mol) (left panel) and 1:1 POPE/POPG (mol/mol) (right panel) bilayers: ................. 114 Solid-state 31P NMR spectra of mechanically aligned 1:1 CL/POPG (mol/mol) bilayers with and without the three aurein peptides: ................ 117 Solid-state 31P NMR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without the three aurein peptides: ................ 118 Solution CD spectra of aurein mutants in 1:1 POPC/POPG (mol/mol) SUVs: ........................................................................................................ 134 Solution CD spectra of aurein mutants in 1:1 POPE/POPG (mol/mol) SUVs: ........................................................................................................ 135 Structural content of aurein mutants in (a) 1:1 POPC/POPG (mol/mol) and (b) 1:1 POPE/POPG (mol/mol) SUVs: ..................................................... 137 Oriented CD spectra of aurein mutants in 1:1 POPC/POPG (mol/mol) bilayers: ..................................................................................................... 139 Oriented CD spectra of L13A in 1:1 POPC/POPG (mol/mol) bilayers as a function of P/L ratio: ................................................................................. 140 Oriented CD spectra of aurein mutants in 1:1 POPE/POPG (mol/mol) bilayers: ..................................................................................................... 141 Solid-state 31P NMR spectra of mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: ..................................................................................................... 143 Solid-state 31P NMR spectra of mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: ..................................................................................................... 144 xiv  Figure 5.9.  Figure 5.10.  Figure 5.11.  Figure 5.12.  Figure 5.13.  Figure 6.1. Figure 6.2. Figure 6.3. Figure 6.4. Figure 6.5. Figure 6.6. Figure 6.7. Figure 6.8.  Figure 7.1. Figure 7.2. Figure 7.3. Figure 7.4. Figure 7.5.  Solid-state 31P NMR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: ..................................................................................................... 146 Solid-state 31P NMR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: ..................................................................................................... 147 Membrane depolarization of S. aureus C622 induced by aurein mutants at 1X and 5X MIC. Gramicidin S (1X MIC) was used as control. Results are representative of 3 experiments. ................................................................ 149 Solid-state 31P NMR spectra of mechanically aligned 4:1 POPE/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: ..................................................................................................... 157 Solid-state 31P NMR spectra of mechanically aligned 4:1 POPE/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: ..................................................................................................... 158 Monocyte chemoattractant protein (MCP-1) induction in the presence of different peptides: ...................................................................................... 167 Cytosolic lactate dehydrogenase (LDH) release in the presence of different peptides:..................................................................................................... 169 Solution CD spectra of aurein mutants in ddH2O/TFE mixtures (vol/vol): ................................................................................................................... 171 Solution CD spectra of aurein mutants in DPC (left panel) and SDS (right panel) micelles:.......................................................................................... 172 Secondary structure content of the three aurein mutants in (a) ddH2O/TFE (vol/vol), (b) DPC, and (c) SDS micelles:................................................. 174 Fingerprint region of solution-state 1H NMR NOESY spectrum of aurein 2.2-∆4N: .................................................................................................... 176 Oriented CD spectra of N-terminally truncated aurein mutants in 1:1 POPC/POPG (mol/mol) bilayers: .............................................................. 177 The ability of N-terminally truncated aurein mutants to permeate across real cell membranes was assessed as a function of time at 20 µg/ml and 100 µg/ml of peptides. 0.5% Triton-X was used as control. AM01 = aurein 2.2∆4N and AM04 = aurein 2.2-∆4N∆3C. (Membrane permeation assay results are courtesy of Dr. J. Kindrachuk). ................................................ 178 1D 1H NMR spectra of apo-daptomycin in 100 mM KCl buffer: ............. 196 1D 1H NMR spectra of Ca2+-daptomycin in 100 mM KCl buffer: ........... 197 1 H NOESY spectra of Ca2+-daptomycin in 100 mM KCl buffer in (a) DHPC-d40 and (b) DHPC-d40/DHPC (9:1 mol/mol): ................................ 199 1 H signals of selected intra- (left panel) and inter-residue (right panel) NOEs of daptomycin in various conditions: ............................................. 200 Individual intra- (left panel) and inter-residue (right panel) NOE intensity as a function of Mn2+ concentration: ............................................................. 202  xv  LIST OF SYMBOLS A D ABBREVIATIO S π  Pi (≈ 3.141592654)  γ  Gyromagnetic ratio of the nucleus [Hz⋅T-1]  ν0  Larmor frequency of the nucleus in [Hz]  ω0  Larmor frequency of the nucleus in [rad⋅s-1]  °C  Degree Celsius  µg  Microgram (10-9 gram)  µl  Microliter (10-6 liter)  µM  Micromole Per Litre (10-6 mole per litre)  µm  Micrometer (10-6 meter)  µmol  Micromole (10-6 mole)  Å  Angstrom (10-10 meter)  B0  External magnetic field [Tesla]  CAP  Cationic Antimicrobial Peptide  CD  Circular Dichroism  CHCl3  Chloroform  CL  Bovine Heart Cardiolipin  cm  Centimeter (10-2 meter)  CSA  Chemical Shift Anisotropy  D2O  Deuterated Water  ddH2O  Double Distillated Water (or Distillated Deionized Water)  DHPC  1,2-dihexanoyl-sn-glycero-3-phosphocholine  DHPC-d40  Perdeuterated 1,2-dihexanoyl-sn-glycero-3-phosphocholine  DiSC35  3,3'-dipropylthiadicarbocyanine iodide  DMPC  1,2-dimyristoyl-sn-glycero-3-phosphocholine  DMPG  1,2-dimyristoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]  DPC  Dodecylphosphocholine  DSC  Differential Scanning Calorimetry  EDTA  Ethylenediaminetetraacetic acid  ESI  Electrospray Ionization  EtOH  Ethanol (ethyl alcohol) xvi  GMS  Gramicidin S  h  Planck constant [6.62606896(33) × 10−34 J⋅s]  HDP  Host Defense Peptide  HEPES  4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid  HPLC  High Performance Liquid Chromatography  hr  Hour  Hz  Hertz  I-state  Inserted state  K  Kelvin  LDH  Lactose Dehydrogenase  M  Mole per litre  MALDI  Matrix-Assisted Laser Desorption/Ionization  MBC  Minimal Bactericidal Concentration  MCP  Monocyte Chemoattractant Protein  mdeg  Millidegree (10-3 degree)  MeOH  Methanol (methyl alcohol)  mg  Milligram (10-3 gram)  MHz  Megahertz (106 hertz)  MIC  Minimal Inhibitory Concentration  min  Minute  ml  Milliliter (10-3 liter)  mM  Millimole Per Litre (10-3 mole per litre)  mm  Millimeter (10-3 meter)  mmol  Millimole (10-3 mole)  mol  Mole  mol/mol  Mole to mole ratio  ms  Millisecond (10-3 second)  2 (g)  nm  Nitrogen gas Nanometer (10-9 meter)  MR  Nuclear Magnetic Resonance  OE  Nuclear Overhauser Effect  OESY  Nuclear Overhauser Enhancement Spectroscopy  OCD  Oriented Circular Dichroism  OD  Optical Density  P/L  Peptide-to-Lipid ratio  PBMC  Peripheral Blood Mononuclear Cell xvii  pg  Picogram (10-12 gram)  POPC  1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine  POPE  1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine  POPG  1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]  ppm  Part Per Million  PRE  Paramagnetic Relaxation Enhancement  REDOR  Rotational Echo Double Resonance  RF  Radio Frequency  RP HPLC  Reverse-Phase High Performance Liquid Chromatography  s  Second  SDS  Sodium Dodecylsulfate  S-state  Surface-adsorbed state  SUV  Small Unilamellar Vesicle  TFE  2,2,2-trifluoroethanol  TFE-d3  Perdeuterated 2,2,2-trifluoroethanol  Tm  The solid/lamellar to liquid-crystalline phase transition temperature  TOCSY  Total Correlation Spectroscopy  vol  Volume  ћ  Reduced Planck constant [1.054571628(53) × 10−34 J⋅s]  1  Proton Nuclear Magnetic Resonance  H MR  31  P MR  Phosphorus 31 Nuclear Magnetic Resonance  xviii  ACK OWLEDGEME TS First of all, as a Christian, I would like to thank my Lord, Jesus Christ, for His love, guidance, help and support for the past four years of my Ph.D. program in the Department of Chemistry at the University of British Columbia. He has proven it to me again that being a Christian and performing scientific research are not conflicting.  I would like to take this  opportunity to thank my senior supervisor, Dr. S. K. Straus for her professional guidance and expertise for the past four years, and for offering this great opportunity for me to work and achieve great accomplishment on these projects. I would also like to thank my supervisory committee members, Dr. H. Li, Dr. C. Fyfe and Dr. R. Krems, for their great help and support whenever I seek assistance, particularly Dr. H. Li whose support never fails. I would want to thank all of my co-workers for their cooperation and support in the lab. I would also want to thank my friends from Evangelical Chinese Bible Church, church fellowship, badminton club and all other friends for their continuous comfort and company. I must not forget to thank my parents, my brother and my fiancée for their endless love and support, as I cannot have completed this study without them. Finally, but not the least, I would like to dedicate my special appreciation to the University of British Columbia for providing me with this fantastic opportunity to accomplish my Ph.D. program in Biological and Physical Chemistry, where I have received rigorous and extensive training to become a professional in this field.  xix  DEDICATIO To my Lord, Jesus Christ, the Saviour, who was, is and is to come, and will come to judge this world.  To my dear fiancée, Tina Chia Wei Jao.  To my dear parents, Jackson Kuang Fu Cheng and Lillian Li Ling Hung, and my dear brother, Peter Tien Hao Cheng.  For God so loved the world that He gave His one and only Son, that whoever believes in Him shall not perish but have eternal life. (John 3:16, The Holy Bible)  xx  CHAPTER 1  CHAPTER 1: Introduction  1.1 Antibiotics 1.1.1 History of antibiotics Antibiotics refer to a class of therapeutic agents that are capable of killing bacteria or preventing bacterial growth. The record of the first antibiotic use can be dated back to 2500 B.C. Ancient Chinese, Egyptians and Greeks utilized herbal compounds (plants or molds) with antimicrobial functions to treat microbial infections (1-3). In the early days, Louis Pasteur was one of the first recognized physicians who observed that bacteria could be used to kill other bacteria. In the 19th century, not only did he change the hospital practice with sterile techniques, he also discovered vaccination (using weakened microbes) and developed pasteurization (sterilization of food by heating). These new revolutions laid an important foundation for the discovery and development of modern antibiotics. (4) The discovery of the first modern antibiotic can be traced back to the development of the narrow-spectrum antibiotic, salvarsan, by Paul Ehrlich in 1909 for the treatment of syphilis. Later in 1928, penicillin was discovered by Sir Alexander Fleming, a Scottish bacteriologist, and was further developed in purified form by Ernst Chain and Howard Florey (5). In 1935, Domagk discovered synthetic antimicrobial chemicals (sulfonamides). During World War II, further isolation and tests by animal injection demonstrated that penicillin is an extremely useful antibiotic to meet the urgent needs. Penicillin usage greatly increased based on these findings, which also encouraged the further search for other chemical agents of similar use. From the late 1  CHAPTER 1  1940's through to the early 1950's, streptomycin, chloramphenicol, and tetracycline were discovered and introduced as antibiotics, adding choices to the expanding antibiotic arsenal (6). In the years that followed, researchers continuously discovered many more derivatives from current antibiotics. These new discoveries have started a new era of therapeutic research and industry against bacterial infections.  1.1.2 Categories and classes of antibiotics 1.1.2.1 Categories of antibiotics Antibiotics can be divided into categories according to their uses and activities. Three factors are considered generally: target specificity (narrow- or wide-spectrum antibiotics), antibacterial action (bactericidal or bacteriostatic activities), and drug administration (ingestion or injection). The ones that target only specific bacterial species are termed narrow-spectrum antibiotics; whereas, the ones that target a wide range of bacterial species are called broadspectrum antibiotics. If antibiotics kill bacteria directly, they possess bactericidal activities; whereas, if antibiotics prevent bacterial growth, they are said to have bacteriostatic activities. Antibiotics can be administered orally, intravenously, or topically, depending on the effectiveness of the drugs and the location of the infection. When developing an antibiotic, three factors are usually considered. First, selectivity toward the microbe is critical. A new antibiotic must have low toxicity toward the host but high toxicity toward the bacterium. Second, the therapeutic index, i.e. the ratio of the effective therapeutic dose to the dose toxic to the host, must be high. The higher the therapeutic index is, the better the antibiotic becomes. Lastly, antibiotic susceptibility tests should always be conducted to obtain the actual antimicrobial efficiency of the newly designed drug. This basic quantitative measure usually involves two methodologies: minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC)  2  CHAPTER 1  assays. MIC is the lowest concentration of the antibiotic that results in inhibition of visible growth (colonies on a plate or turbidity in broth culture) under standard conditions. MBC is the lowest concentration of the antibiotic that kills 99.9% of the original inoculum in a given time.  1.1.2.2 Classes of antibiotics In today’s definition, antibiotics include both natural isolates and synthetic compounds (5). Examples include fungal extracts (penicillin and its derivatives) and synthetic sulfonamides (mefanide and sulfacetamide). These antibiotics are active against various bacterial infections. For example, macrolides (azithromycin and erythromycin) are active against streptococcal and respiratory  infections,  whereas  sulfonamides  (sulfamethizole  and  trimethoprim-  sulfamethoxazole) are active against urinary tract infections. Generally, different classes of antibiotics have different target sites. For example, aminoglycosides target the bacterial 30S ribosomal subunit (some work by binding to the 50S subunit), inhibiting the translocation of the peptidyl-tRNA from the A-site to the P-site, and causing misreading of mRNA. These actions usually prevent the bacterium from synthesizing proteins vital to its growth. Another example is macrolides, which binds irreversibly to the bacterial 50S ribosomal subunit and thus prevents peptidyl tRNA translocation. This, in turn, inhibits bacterial protein biosynthesis. Sulfonamides exhibit their antimicrobial functions by inhibiting folate synthesis in bacteria. The conversion of PABA (para-aminobenzoate) to dihydropteroate is a key step in folate synthesis, which requires the enzyme dihydropteroatesynthetaseto catalyze this conversion.  Sulfonamides act as  competitive inhibitors of this enzyme, which can disrupt its enzymatic function and stops the folate production. Folate production is necessary for bacterial survival as folate is involved in bacterial nucleic acid synthesis.  Since DNA and RNA syntheses require nucleic acids as  building blocks, these processes would not take place if nucleic acid synthesis is inhibited. Bacteria will not be able to undergo cell division if DNA and RNA materials are not duplicated. 3  CHAPTER 1  In turn, bacterial growth would stop and successful bacterial inhibition would be obtained. Listed here are just few examples of the many classes of antibiotics discovered and developed to target different sites in order to treat various bacterial infections.  1.1.3 Examples of conventional antibiotics Since the discovery of the first modern antibiotics in 1909, a large number of antibiotics from different sources have been found and functional derivatives have been designed and synthesized to improve antibiotic efficacy. All these antibiotics fall in the class of conventional antibiotics, which includes isolates from natural sources, semi-synthetic derivatives, and synthetic analogues. Figure 1.1 shows the molecular and 3D structures of the first widely developed and used antibiotics, penicillin, and its derivatives, ampicillin and amoxicillin.  Figure 1.1.  The molecular (left panel) and 3D (right panel) structures of penicillin and its derivatives: (a) & (d) penicillin; (b) & (e) ampicillin; (c) & (f) amoxicillin. Note the functional group difference among the three antibiotics at the N-terminus.  Penicillin and its derivatives belong to the class of β-lactam antibiotics. These antibiotics kill bacteria by directly acting on the bacterial cell wall and inhibiting the peptidoglycan cross4  CHAPTER 1  link formation. The binding of the β-lactam moiety of penicillin to DD-transpeptidase, an enzyme that links the peptidoglycan molecules in bacteria, weakens the cell wall of the bacterium. This causes cytolysis or death due to osmotic pressure. Moreover, peptidoglycan precursor accumulation can activate bacterial cell wall hydrolases and autolysins, which further digest the bacteria's existing peptidoglycan. Penicillin can work with aminoglycosides synergistically. Aminoglycosides kill bacteria by disrupting bacterial protein synthesis, which requires bacterial cell wall penetration. The presence of penicillin inhibits peptidoglycan synthesis, which gives aminoglycosides an easier entry into the bacteria. This results in a lowered MBC value for susceptible organisms. (7) An example of aminoglycosides is gentamicin, an isolate from Micromonospora genus bacteria. Gentamicin acts on the 30S subunit of the bacterial ribosome, interrupting protein synthesis. Figure 1.2 shows the molecular and 3D structure of gentamicin.  Figure 1.2.  The (a) molecular and (b) 3D structures of gentamicin. (Figures adapted from “Wikipedia” with minor modification, http://en.wikipedia.org/wiki/Gentamicin).  Gentamicin is normally used to treat Gram-negative bacterial infections including Pseudomonas, Proteus, Serratia, and Gram-positive Staphylococcus. However, gentamicin is 5  CHAPTER 1  generally not used for eisseria gonorrhoeae, eisseria meningitidis or Legionella pneumophila bacterial infections due to potential risk from toxicity. (8)  1.1.4 Antibiotic resistance Antibiotic resistance has long been observed since the discovery of antibiotics. For example, as early as 1953, the first strain of dysentery bacillus resistant to chloramphenicol, tetracycline, streptomycin, and the sulfanilamides was discovered. In addition, in the 1950's, tuberculosis bacteria were found to acquire resistance rapidly to streptomycin, a commonly prescribed antibiotic for tuberculosis treatment. (6) Misuses of antibiotics can easily lead to the emergence of antibiotic-resistant bacterial strains. Common misuses include the administration of antibiotics for viral infections, failure to complete the entire course of antibiotic treatment, and excessive uses of antibiotics in the food industry (5,9). These misuses enable infectious bacteria to become resilient and develop ways to survive various drugs.  Currently, approximately 70% of bacteria that cause infections in  hospitals are found to be resistant to one or more conventional antibiotics used commonly (9). In some serious cases, traditional therapeutic agents such as sulfa drugs are administered, even though sulfa drugs are highly discouraged due to their high toxicity. The immediate need to find new therapeutic agents against the growing antibiotic resistance has become critical. Antibiotic resistance can be categorized into three types: natural or intrinsic resistance, mutational resistance, and extrachromosomal or acquired resistance. Bacteria have natural or intrinsic resistance usually by altering their structure or cellular components to make the target sites inaccessible or unavailable. Some bacteria have a natural multidrug efflux system to pump antibiotics out of the cell, or direct drug inactivation to disable antibiotics in the cell. Mutational resistance usually involves genetic change in response to prolonged antibiotic application. For  6  CHAPTER 1  example, streptomycin resistance arises as a result of bacterial modification on the target site, rDNA (rpsL) gene mutation, and β-lactam resistance as a result of change in penicillin binding proteins (PBPs). These mutations generally lead to reduced permeability or uptake of the drug. Finally, bacteria can adopt extrachromosomal or acquired resistance, through dissemination by plasmids or transposons. These acquired resistances allow bacteria to inactivate drugs by new enzymes, for example, aminoglycoside-modifying enzymes (β-lactamases, chloramphenicol acetyltransferase). Other processes include acquired efflux system (e.g. tetracycline efflux), target site modification (e.g. methylation in the 23S component of the 50S ribosomal subunit by ermmethylases), and metabolic by-pass (e.g. trimethoprim resistance by resistant DHF reductase).  1.1.4.1 Different ways to acquire antibiotic resistance There are three major ways that bacteria acquire antibiotic resistance: transformation, conjugation and transduction.  Figure 1.3 illustrates these three processes.  First,  “transformation” allows the transfer of free unbound antibiotic-resistant genes between bacteria. One bacterium can acquire these free unbound DNA materials typically either released from a dead bacterium or directly from another bacterium through pili, a microbial sexual organ-like structure. Second, “conjugation” allows plasmids encoding multiple antibiotic-resistant genes to be transferred between different bacterial strains. Plasmid is a type of circular DNA capable of encoding multiple antibiotic-resistant genes, which can be used as a medium to transfer antibiotic-resistant genes between different bacterial strains. Third, “transduction” allows the transfer of antibiotic-resistant genetic material through uptake or exchange of random DNA pieces that can be spliced into bacterial chromosomes. These random DNA pieces may not encode antibiotic-resistant genes against the current drugs, but may encode genes resistant against drugs to be developed in the future. Transduction can be accomplished through virus or  7  CHAPTER 1  bacteriophage.  In addition, spontaneous mutation of bacterial DNA can also give rise to  antibiotic resistance. All of these contribute to growing antibiotic resistance (10,11).  Figure 1.3.  The three processes that bacteria acquire antibiotic resistance: (1) Transformation: transferring free DNA materials encoding antibiotic resistant genes between bacteria. (2) Conjugation: transferring free DNA plasmid encoding antibiotic resistant genes between bacteria. (3) Transduction: transferring free DNA materials encoding antibiotic resistant genes between bacteria via virus or bacteriophage. (Figure adapted from “Yim, G., 2009” with minor modification (12)).  1.1.4.2 Different mechanisms of antibiotic resistance There are several mechanisms of antibiotic resistance: efflux, membrane plug formation, ribosomal blockade, enzymatic destruction, and alteration of the drug (12). Bacteria adopt these mechanisms to counterattack antibiotics, as described above, either intrinsically or through acquisition. Figure 1.4 shows how bacteria react against antibiotic treatment. First, bacteria can develop methods to export antibiotics out of the cell (efflux). Bacteria utilize a variety of proteins as efflux pumps to export any harmful substance. These protein pumps exist in both Gram positive and Gram negative bacteria and as well as in eukaryotes. 8  CHAPTER 1  These pumps are usually specific for one foreign substance or can export various structurally different substances. For example, some pumps can transport antibiotics of multiple classes and are thus a key feature of multiple drug resistance. There are five major families of efflux transporters in prokaryotes: MF (major facilitator), MATE (multidrug and toxic efflux), RND (resistance-nodulation-division), SMR (small multidrug resistance) and ABC (ATP binding cassette). Studies have suggested that 5% ~ 10% of all bacterial genes encode proteins for transport functions and a large number of these proteins are efflux pumps (13).  Figure 1.4.  Five mechanisms that bacteria develop against antibiotics: (1) Efflux; (2) membrane plug formation; (3) ribosomal blockade; (4) enzymatic destruction; (5) alternation of the drug. All these mechanisms allow a bacterium to actively block an antibiotic from entering the cell or deactivate antibiotic to become no longer functional. (Figure adapted from “Yim, G., 2009” with minor modification (12)).  Second, bacteria can alter a protein to block antibiotics from entering the cell (membrane plug formation).  As many small hydrophilic antibiotics (such as β-lactams, tetracycline,  chloramphenicol and fluoroquinolones) enter a bacterium through porins, a class of membrane proteins that allow the entry of substances through the outer membrane into the cell, bacteria were found to block antibiotic entry by changing their outer membrane profiles (loss/severe 9  CHAPTER 1  reduction of porins or replacement of major porins by other non-functional analogues), or by modifying functions through specific mutations to reduce the outer membrane permeability (14). Third, bacteria can synthesize a protein or alter the ribosomal structure to block antibiotics from binding to ribosomes (ribosomal blockade). For example, a recent study has found that a class of E. coli mutants is able to survive macrolide antibiotic application. These mutants had a 3-residue deletion (∆MKR) in ribosomal protein L22, which distorts the L22 loop structure to allow mRNA translation (protein synthesis) even in the presence of macrolides. This mutation is also found to reduce the binding of antibiotics to the target site on the ribosome (15). Other modification include methylation of the erm gene and substitution of adenine to guanine, which increases the size of the target site (position 2058 of 23S ribosome) and thereby eliminate the binding of the desosamine sugar group of erythromycin, a macrolide antibiotic (16). Fourth, bacteria can produce an enzyme to destroy the antibiotic prior to displaying its function (enzymatic destruction). These include various chemical processes such as hydrolysis, group transfer, and redox mechanisms. For example, β-lactam rings of penicillin and its related antibiotics are particularly susceptible to enzymatic (hydrolytic enzyme β-lactamase) hydrolysis by bacteria. Bacteria can also destroy an antibiotic through group transfer modification, which includes acyltransfer, phosphorylation, glycosylation, nucleotidylation, ribosylation, and thiol transfer. These processes can actively lower the drug concentration locally. Another commonly used method is the so-called redox mechanism. For example, TetX, a Bacteroides fragilis enzyme, is responsible for the oxidation of tetracycline which abolishes the antibiotic function of tetracycline antibiotics. It catalyzes the monohydroxylation of tetracycline antibiotics at position 11a, which disrupts the Mg2+-binding site of the antibiotic that is required for antibacterial activity (17).  10  CHAPTER 1  Finally, bacteria can synthesize an enzyme to alter the administered antibiotic so as to deactivate it (alteration of the drug). This process can be directed toward antibiotics that do not have groups vulnerable to hydrolysis. Bacteria possess various enzymes that can modify or add chemical constituents to an antibiotic in order to deactivate its function.  For example,  aminoglycoside-resistance enzymes can add adenosine monophosphate (AMP) moieties (using adenylyl transferases), phosphate groups (using phosphoryltranferase), or acetylate the amino group (using acetyl transferases).  All these additional chemical substituents interrupt and  prevent aminoglycoside antibiotics from binding to the RNA targets in the ribosome. Thus, the subsequent protein synthesis can continue without interruption (18).  Figure 1.5.  A micrograph of Gram stains of mixed Staphylococcus aureus (Gram positive cocci) and Escherichia coli (Gram negative bacilli): Gram positive bacteria such as S. aureus will show purple, whereas Gram negative bacteria such as E. coli will show pink when Gram stain is applied. Note the shape difference between the two bacterial strains, which is also indicated by their names. (Figure adapted from “Wikipedia” with minor modification, http://en.wikipedia.org/wiki/File:Gram_stain_01.jpg).  These five major mechanisms enable bacteria to adapt to different antibiotics and become resistant to them. Nowadays, more and more bacterial proteins and enzymes are found to be responsible for the increasing antibiotic resistance. An obvious example of antibiotic resistance comes from Staphylococcus aureus, a bacterial species that was once susceptible to penicillin in  11  CHAPTER 1  the 1940s and 1950s. Almost immediately after penicillin was introduced, resistance in certain strains of staphylococci was noticed. S. aureus belongs to the class of Gram positive bacteria, which show purple stain when undergoing Gram stain treatment (Figure 1.5). Gram positive and Gram negative bacteria have several structural differences. Not all antibiotics are effective on both Gram positive and Gram negative bacteria. Some antibiotics work better on one but not the other. As mentioned earlier, different antibiotics target different cellular components of a bacterium. Thus, it is important to understand the structural differences between the two bacterial classes as well as what organelles exist within a bacterial cell.  1.1.5 Structure of bacteria Gram positive bacteria are generally characterized by the absence of an outer membrane and periplasmic space, and the presence of a thicker peptidoglycan layer. Figure 1.6 gives the pictorial representation of Gram positive and Gram negative bacteria, showing the structural differences between them.  Figure 1.6.  The cross section schematic diagram depicts the major components of Gram positive and Gram negative bacteria. (Figure adapted from “Bacterial morphology” with minor modification, http://micro.digitalproteus.com/morphology2.php). 12  CHAPTER 1  Both Gram positive and Gram negative bacteria have unbound/unpackaged DNA materials floating inside the cytoplasm.  Common organelles include inclusion bodies,  ribosomes, a peptidoglycan layer, flagella and a capsule. The most obvious difference between the two bacterial types lies at the peptidoglycan layer/outer membrane arrangement. Gram positive bacteria have a thick peptidoglycan cross-link network layer interconnected with lipoteichoic acids (LTA).  Gram negative bacteria, on the other hand, has a much thinner  peptidoglycan layer surrounded by an outer lipid membrane. In between the lipid membranes and peptidoglycan layer is the periplasmic space, where various lipoproteins are located. The outer membrane of Gram negative bacteria is also characterized by the presence of lipopolysaccharides (LPS).  Figure 1.7 shows schematic diagrams of the outer membrane  structure of Gram positive and Gram negative bacteria.  Figure 1.7.  Differences in the membrane structure between (a) Gram positive and (b) Gram negative bacteria. (Figure adapted from “MJCR/MBMB 403 Microbiology, Southern Illinois University Carbondale” with minor modification, http://www.cehs.siu.edu/fix/medmicro/pix/walls.gif).  13  CHAPTER 1  Today, many Gram positive bacteria, including most S. aureus strains, are found resistant to penicillin, nafcillin, oxacillin and gentamicin, leaving a narrow selection of antibiotics such as methicillin and vancomycin (5,19). In the past decade, methicillin-resistant and vancomycinresistant S. aureus strains have been continuously reported, which make S. aureus even more difficult to eliminate (20-22). Currently, the antibiotic resistance of S. aureus has become a major threat in the hospital environment, causing infections via hospital beds, needles, and open wounds (9). Conventional antibiotics, including their second- or third-generation derivatives, seem to be inefficient against the wide-spreading resistance (5). It is necessary to find and develop other antimicrobial agents such as antimicrobial peptides in order to win this battle. Figure 1.8 shows the microscopic image of this vicious bacterium, depicting its characteristic grape-like colonies.  Figure 1.8.  Microscopic image of Staphylococcus aureus, the superbug (green graph-like colonies). (Figure adapted from http://swampie.files.wordpress.com/2008/02/staphylococcus-aureus.jpg).  1.2 Alternative antibiotics – antimicrobial peptides Antimicrobial peptides, also part of host defense peptides (HDPs), comprise an important constituent of the innate immune response among all classes of life. They are found capable of targeting Gram positive and Gram negative bacteria, mycobacteria, enveloped viruses, fungi and  14  CHAPTER 1  cancerous cells (23,24).  Cationic antimicrobial peptides (CAPs), a major class of charged  antimicrobial peptides, have been found to cause little or no antimicrobial resistance to date (2527). Another major class of antimicrobial peptides is lipopeptides, which are also shown to be very effective and have little resistance (28). To date, this lack of resistance makes them ideal candidates for antibiotic development.  1.2.1 Cationic antimicrobial peptides (CAPs) CAPs are also found in all forms of life including bacteria, fungi, plants, insects, birds, crustaceans, amphibians and mammals. They are ubiquitous and diversify into many classes. It is usually found that different classes of peptides and many variants in a given class can exist within a single organism. For example, cattle are known to contain at least 38 antimicrobial peptides. These include α-helical BMAP peptides, β-stranded α- and β-defensins, the loop peptide bactenecin, and the extended peptide indolicidin. Hancock and Diamond suggested four possible reasons for this variety (29). First, a given peptide normally has a limited antimicrobial spectrum. This suggests that multiple peptides are necessary to cover and complete the whole antimicrobial spectrum. Second, many peptides were discovered to work synergistically to display an antimicrobial function against microorganisms. Third, these peptides also possess some non-antimicrobial (for example, anti-endotoxic, chemotactic and pro-inflammatory, i.e. immunomodulatory) activities, which may vary and complement each other. Fourth, different cell types are usually responsible for producing different peptides. As a result, a given tissue might express only a subset. Table 1.1 lists examples of mammalian peptides commonly studied.  15  CHAPTER 1 Table 1.1.  Mammalian cationic antimicrobial peptides (Table reproduced from Hancock, R. E. and Diamond, G. 2000 (29)). These peptides adopt one of the four common secondary structures. ote that disulfide bridges are labeled with numbers in subscript, at the relevant positions.  Peptide  Structure  Sequence  Rabbit α-defesin (NP-1)  β-sheet  VVC1AC2RRALC3LPRERRAGFC3RIRGRIHLC2C1RR  Human β-defensin 1  β-sheet  DHYNC1VSSGQC2LYSAC3PIFTKIQGTC2YRGKAKC1C2K  Pig protegrin-1  β-sheet  RGGRLC1YC2RRRFC2VC1VGR#  Human LL-37  α-helix  LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES#  Human histatin 5  α-helix  DSHAKRHHGYKRKFHEKHHSHRGY#  Cattle indolicidin  Extended  ILPWKWPWWPWRR#  Pig PR39  Extended  RRRPRPPYLPRPRPPPFFPPRLPPRIPPGFPPRFPPRFP#  Cattle bactenecin  Loop  RLC1RIVVIRVC1R  #  The peptide is known to be amidated at its carboxy terminus.  CAPs are generally 12 to 50 amino acids long with a net positive charge of +2 to +7 due to a large number of basic residues, for example, arginine, histidine and lysine. Proteolytic digestion of larger cationic proteins can also give rise to many other CAPs. These larger cationic proteins, interestingly, also have an implicated role in innate immunity. For example, lactoferrin is an 80 kDa bactericidal protein that exists in milk and many mucosal secretions such as tears and saliva (30). Cathepsin G is another protein proposed to participate in the killing and digestion of engulfed pathogens (31). All these larger proteins share some characteristics and activities that are reminiscent of cationic peptides (29). CAPs are generally unstructured in free solution and adopt their final configuration in the presence of biological membranes (23,32-34). They are found to adopt common secondary structures, such as α-helices, β-sheets, loops and extended structures, where their amphiphilicity is formed with hydrophilic and hydrophobic residues on opposite sides (23,26). In some cases, some peptides can adopt mixed conformation. β-sheet peptides are usually stabilized by 2 ~ 4 disulfide bonds, whereas loop peptides normally contain only one. Extended peptides usually carry an excessive number of one or two particular amino acids (e.g. proline, tryptophan or  16  CHAPTER 1  histidine) (29). Figure 1.9 shows the most commonly adopted structures of various widely studied antimicrobial peptides in the presence of lipid membranes.  Figure 1.9.  Various secondary structures of antimicrobial peptides, as selected from Protein Data Bank: (a) α-helix (magainin); (b) β-sheet (human β-defensin); (c) extended (indolicidin); (d) mixed (protegrin-1). Antimicrobial peptides are normally unstructured in aqueous solution but adopt structures in membrane-mimicking environment. (Figures adapted from “Wikipedia” with minor modification, http://upload.wikimedia.org/wikipedia/en/e/ef/Various_AMPs.png).  In general, the most potent cationic peptides fold into molecules either as amphipathic structures or as cationic extended structures with a hydrophobic core separating two charged segments. This allows CAPs to associate/interact with and integrate into the outer leaflet, and cause the destabilization of bacterial membranes through a membrane thinning effect (35). Then, the bacterial membrane lyses occur as a result of various mechanisms of action of the peptides.  17  CHAPTER 1  1.2.2 Acidic lipopeptides Acidic lipopeptides refer to a class of antimicrobial peptides that comprise of 11 ~ 13 amino acid residues that are typically cyclized into a peptide ring and a long-chain fatty acid invariably attached to the macrocyclic peptide core. The cyclization occurs through two major ways: macrolactonization (e.g. A21978C/daptomycin and A54145) and macrolactamization (e.g. amphomycin/friulimicin and laspartomycin/glycinocin). Most of these listed cyclic lipopeptides have a characteristic ten-membered ring. The fatty acid tails, however, vary significantly in branching, degree of saturation and oxidation states. These variations lead to highly diversified structures found in this class of antibiotics. The highly varied lipid tails can actually impose great impact on the antimicrobial behaviour and toxicity of these peptides. These lipid tails range from 6 ~ 16 carbons, with the shortest chain of six carbons found in calcium-dependent antibiotics (CDAs), and the longest chain of 16 carbons found in glycinocin B (36). The lipopeptides cyclized by macrolactonization are classified as cyclic lipodepsipeptides. The most well-known of them is A21978C, also known as daptomycin. Daptomycin also has a ten-membered peptide ring connected between Thr4 and Kyn13, one of the several nonproteinogenic residues found in the daptomycin sequence. Other similar residues and sterically uncommon D-amino acids include D-Asn2, Orn6 (ornithine), D-Ser11 and MeGlu12 (3methylglutamic acid). Daptomycin has a saturated, decanoic acid lipid tail attached to Trp1. A54145, another group of lipodepsipeptides, has a slightly larger number of uncommon amino acids and a variety of decanoyl chains. Seven of the 13 residues are nonproteinogenic and three of them are D-amino acids. More differences between A54145 and A21978C come from the incorporation of either Ile or Val at position 13. In addition, the lipid tails diversify into isodecanoyl, n-decanoyl, and anteiso-decanoyl chains. Finally, another class of lipodepsipeptides is  18  CHAPTER 1  CDAs. These peptides are cyclized between Thr2 and Trp11. However, the lipid tails are invariant and exclusively found to be 2,3-epoxy-hexanoyl (36). Cyclic lipopeptides, in which cyclization of the peptide chain occurs through macrolactamization, can be divided into two groups. One of them is amphomycin/friulimicin, which is divided into two subgroups again: A-1437 and friulimicin. Amphomycin, the first discovered antibiotic from the subgroup A-1437, and other similar antibiotics, such as tsushimycin and aspartocin, all have the characteristic rigid peptide ring connected between Dab2 (diaminobutyrate) and Pro11. The only structural difference comes from the different fatty acid substituents. Friulimicin, on the other hand, only differs in the exocyclic amino acid (Asn1 instead of Asp1). Other uncommon amino acids were also discovered at various positions. Another group of cyclic lipopeptides is laspartomycin/glycinocin.  Laspartomycins were  discovered as an antimicrobial peptide related to amphomycin. Its structure remained unknown until the first structure of laspartomycin C was determined. Two major differences were found. The ten-membered ring is connected between Dap2 (diaminopropionate) instead of Dab2. The lipid tail is 2,3-unsaturated. The peptide ring also has different residues substituted at different positions, including D-Thr9, Gly4 and Ile10. Finally, glycinocins have an almost identical structure as laspartomycin, except some carry different fatty acid chains, and some have Val instead of Ile at position 11 (36).  1.2.3 Mechanisms of action Since the discovery of antimicrobial peptides, researchers have been trying to understand how and why these peptides are active against specific bacteria. Many studies were conducted to describe the behaviour of these peptides in intact bacteria or model membranes, and different mechanisms of action were proposed based on these observations. These mechanisms of action include the carpet model (33,34), toroidal model (33), barrel-stave model (32), a micellar 19  CHAPTER 1  aggregate channel model (37,38) or a detergent-like mechanism (39). These mechanisms of action cause membrane depolarization, degradation of cell walls, membrane micellarization and eventually, bacterial death (40).  Figure 1.10 shows the main mechanisms of action as  summarized by Papo and Shai (41).  Figure 1.10. The main mechanisms of action of antimicrobial peptides against bacterial membranes: (a) Barrel-stave model; (b) carpet model; (c) toroidal model. The peptide molecules bind to the membrane surface initially and insert into the lipid bilayers at a peptide concentration greater than the threshold value. The peptide insertion can create membrane micellarization, toroidal pores, or peptide channels. (Figure reproduced from “Papo, N.and Shai, Y., 2005” with minor modification (41)).  The most commonly observed mechanisms of action are the carpet model, toroidal model and barrel-stave model. Generally, peptide molecules, when exposed to lipid membranes, will bind to the membrane surface. When the peptide concentration increases above a threshold value, the peptide molecules will start inserting into the membrane core to adopt a thermodynamically more favourable state.  This is due to membrane surface tension, for  example, positive or negative curvature strain on the membrane surface. This insertion usually leads to three different mechanisms of membrane perturbation. First, peptide molecules can 20  CHAPTER 1  extract lipids from the membrane bilayers, completely disintegrating the membrane and destabilizing its structural integrity (membrane micellarization, carpet model). Second, peptide molecules can also align themselves along the lipid headgroups to form a pore-like structure, creating membrane defects (toroidal pore formation, toroidal model). Last, peptide molecules can insert directly, if the peptide is sufficiently long to span the lipid bilayers, to form channels, causing additional membrane openings (channel formation, barrel-stave model). Peptide binding to the membrane surface is not necessarily the initial step in the latter case. Free peptide molecules in an aqueous environment may also insert into the lipid bilayers to form channels directly, without first binding to the membrane surface. These three mechanisms of action comprise the famous model proposed for antimicrobial peptides, the “Shai-Matsuzaki-Huang model” (23).  1.3 Amphibian antimicrobial peptides As mentioned earlier, many animals and plants have acquired CAPs as constituents of their immune defense systems. Amphibians are found to possess several families of CAPs as a major component in their host-defense mechanism (42,43). The skin of amphibians constitutes an important element of innate defense against microbial infections. Secretions on the surface of the skin contain a variety of peptides acting as natural antibiotics to counterattack bacterial and fungal invasion, which are common in the natural environment of amphibians. When these animals are exposed to various stresses or stimuli, granular glands, the dermal structures where these host-defense compounds are produced and stored, release their content onto the frog’s skin (44-48). Many studies have reviewed (23,42,43) or characterized these peptides extensively, which include magainins (49-58), maculatins (59-64), brevinins (65-70), and others, such as  21  CHAPTER 1  citropin 1.1 and aurein 1.2 from the Australian tree frogs Litoria citropa and Litoria aurea (44,59-62,64,71,72), respectively. The latter peptide is part of a larger family of peptides known as the aurein peptides, which range in length from 13 ~ 25 residues. The following sections will briefly introduce some of the commonly studied amphibian antimicrobial peptides as examples.  1.3.1 Citropins Citropins are another class of amphibian antimicrobial peptides isolated from Australian green tree frogs Litoria citropa. This class consists of 19 peptides, and their sequences were previously determined by electrospray mass spectrometry (73). These peptides are produced by both the dorsal and submental glands of the frog (74,75). Citropin 1.1, the most well studied peptide in citropin peptide family (48), was found to be active against Leuconostac lactis (6 µg/ml), Micrococcus luteus (12 µg/ml) and Staphylococcus epidermidis (12 µg/ml) (76). Citropin 1.1 is a 16-residue peptide that has some basic and highly hydrophobic amino acids. It is currently one of the simplest, wide-spectrum amphibian antimicrobial peptide studied to date. It was proposed that these peptides adopt an amphiphilic α-helical conformation in the presence of lipid environment (59).  Using the lipid monolayer approach, citropins were  demonstrated to interact preferentially with anionic phospholipids (59).  Marcotte I. et al.  suggested that citropin 1.1 inserts into the lipid bilayer partially, as its length does not allow for complete spanning of the membrane to form a pore or pore-like structure (62). Due to its short peptide length, it was suggested that citropin 1.1 disorders lipid membranes via a carpet-like mechanism (60,61,77).  In this case, citropin 1.1 mostly likely binds predominately to the  membrane surface (34), according to Shai-Matsuzaki-Huang model, then disrupts the membrane bilayers and finally, lyses the cell (78).  22  CHAPTER 1  1.3.2 Aurein peptides Aurein peptides exhibit antimicrobial activities against Gram positive bacteria, such as Bacillus cereus, Leuconostoc lactis, S. aureus, and S. epidermis, and cancerous cells (72). There are five families of aurein peptides, ranging from the short-length active aurein 1 to 3 families to the longer and typically inactive aurein 4 and 5 families (72). Two different C-terminal groups are found in the aurein peptide families, namely, amidated (-CONH2) and carboxylic (-COOH) groups. Most active aurein peptides are found to have the amidated C-terminus, whereas aurein peptides carrying the carboxylic C-terminus are mostly inactive (72). Currently, aurein 1.2 is the most well studied peptide in the aurein peptide family (60,62,71). It is a 13-residue peptide with a net charge of +2. Model membrane studies using solution-state circular dichroism (CD) and NMR spectroscopy have shown that aurein 1.2 adopts an α-helical structure in the presence of either 70% trifluoroethanol (TFE)/30% water (72) or SDS micelles (79). Aurein 1.2 is proposed to interact predominantly with the bacterial membrane surface due to its insufficient length (59,62). Surface absorption into bacterial membranes is, therefore, the proposed way that aurein 1.2 disrupts the phospholipid bilayers through a detergent-like or carpet-like mechanism. Recently, aurein 1.2 was shown to be an effective bactericidal agent against staphylococci and streptococci (80). Moreover, it was found to have relatively low cytotoxicity and to act in synergy with other antibiotics such as minocycline or clarithromycin. Since these latter antibiotics are hydrophobic, it is believed that the membrane perturbation induced by aurein 1.2 facilitates the entry of minocycline or clarithromycin into the membrane, making them more effective (81,82). The ability of aurein 1.2 to perturb membranes has recently been examined in detail using differential scanning calorimetry and Fourier transform infrared spectroscopy studies (48).  23  CHAPTER 1  1.4 Studying antimicrobial peptides in model systems and real bacteria Many methods have been developed and used to study the structure-function relationships of CAPs in model membrane systems, including circular dichroism (CD) spectroscopy and various nuclear magnetic resonance (NMR) experiments which probe proton (1H), deuterium (2H), nitrogen-15 (15N) and phosphorus (31P) signals. Solution CD spectroscopy provides solution-state information on the secondary structures of the peptides, such as αhelices, β-sheets, or random coils. Oriented CD spectroscopy provides solid-state information on the orientation of the peptides, such as whether the peptide is surface-adsorbed (S-state) or inserted (I-state or T-state if tilted) into the phospholipid bilayers.  1  H NMR and  15  N NMR  spectroscopy give information on the microscopic structure of the peptides, where the identification of amino acid peaks, the sequential spectral assignments, and the insertion angle determination are possible. 2H and 31P NMR spectroscopy give information on how the peptides perturb the motion of the acyl chains and the orientation of the headgroups of the phospholipid bilayers, respectively, and provide an indication of the overall fluidity and order of the bilayer phase structure.  Together with other tools such as calcein release, DiSC35 and minimum  inhibitory concentration (MIC) assays, these techniques provide information on the effects and the changes induced by CAPs on the phospholipid bilayers and intact bacteria. The following sections will introduce these techniques briefly, starting with CD (Section1.4.1) and NMR (Section1.4.2) spectroscopy, followed by membrane leakage and MIC assays (Section 1.4.3).  1.4.1 CD spectroscopy 1.4.1.1 Solution CD spectroscopy Solution CD spectroscopy has been widely used to explore the global structure and to gain the first insight on the secondary structure of a particular peptide/protein (83-85). CD 24  CHAPTER 1  spectroscopy is based on differential absorption of left- and right-handed circularly polarized light. A protein with a specific secondary structure will absorb left- and right-handed circularly polarized light differently. Thus, different protein secondary structures will give rise to different but characteristic CD spectra. Figure 1.11 shows the standard curves for different secondary structures (α-helix, β-sheet and random coil) based on solution CD spectra of poly-lysine (86).  Figure 1.11. Standard curves of α-helix, β-sheet and random coil based on solution CD spectra of poly-lysine. The standard spectra are used to fit solution CD data of the peptide/protein under investigation to quantify the extent of different secondary structures that exist in the peptide/protein. (Figure reproduced from “Circular Dichroism Spectroscopy” with minor modification, http://www.ruppweb.org/cd/cdtutorial.htm).  Since each of the three basic secondary structures of a polypeptide chain (α-helix, βsheet, random coil) show a characteristic CD spectrum, if a peptide/protein consists of any or all of these elements, it should be possible to deconvolute its CD spectrum into the three individual contributions. By using computer programs designed specifically to extract the information on the percentage of each type of secondary structure from solution CD spectra, one is able to quantify the extent of different secondary structures present in an antimicrobial peptide.  25  CHAPTER 1  However, solution CD spectra can only be used to approximate the percentage of α-helix, βsheet/strand/turn and random coil. They do not provide information on the exact location of the detected secondary structure in a peptide.  With this limitation, one needs to use NMR  spectroscopy or other methods to pinpoint the exact location of these structures, in order to obtain the microscopic structural information.  Even with these restraints, solution CD  spectroscopy is still a very powerful tool to investigate the effect of temperature, ionic strength, pH, and binding partner on the secondary structure of a given peptide. Most of all, one can use this tool to quickly gain initial information on the secondary structure of a newly discovered or derived antimicrobial peptide and to use it as a confirmation of structural results found from other methods. Many studies have characterized the secondary structure of different CAPs, including aurein 1.2 and citropin 1.1 (α-helix) (62), arenicin (β-sheet) (87), protegrin-1 (complex) (88), and indolicidin (extended chain) (89). Figure 1.12 shows solution CD spectra of magainin 2, an amphibian CAP, and its analogues. The spectra demonstrate that these peptides adopt an α-helical conformation (90).  Figure 1.12. Solution CD spectra of magainin 2 and its analogues in the presence of 1:1 TFE/buffer (v/v): All spectra show a maximum at 195 nm and two minima at 207 nm and 222 nm, showing that these peptides adopt an α-helical structure. The buffer components are: 10 mM Tris, 154 mM NaCl, and 0.1 mM EDTA (pH 7.4). (Figure adapted from “Wieprecht, T. et al., 1997” (90)). 26  CHAPTER 1  1.4.1.2 Oriented CD (OCD) spectroscopy OCD spectroscopy has been recently used to examine the orientation of CAPs in mechanically aligned lipid bilayers.  As most CAPs kill bacteria by disrupting bacterial  membranes through concentration-dependent effect, it would be important to inspect the peptidemembrane interaction as a function of peptide concentration. The Shai-Matsuzaki-Huang model demonstrates that at low concentration, peptide molecules bind to the membrane surface (surface-adsorbed state, or S-state). At high concentrations, peptide molecules insert into the membrane bilayers, inducing toroidal pore or channel formation (inserted or tilted state, or I-state or T-state). Each physical state of the peptide can give rise to a characteristic OCD spectrum (91), as shown in Figure 1.13.  Figure 1.13. Reconstructed spectra of the parallel and perpendicular components of the CD spectrum of an α-helix. The summation represents isotropic distribution of helices. The dashed line is the reference spectrum of polylysine as given by Greenfield and Fasman (86). The OCD spectrum for an α-helical peptide in the surface-adsorbed state (//) displays a single maximum at 195 nm and two minima at 207 nm and 222 nm. The OCD spectrum for peptide in the inserted state (⊥) displays one maximum at 195 nm and one minimum at 222 nm, where minimum at 207 nm is greatly reduced. (Figure reproduced from “de Jongh, H. H. J. et al., 1994” with minor modification (92)).  The threshold concentration for a peptide to go from the S-state to I- or T-state indicates how effective the peptide is at disrupting a membrane. If a CAP changes its orientation from the S-state to I-/T-state at lower concentration, it is more effective in inserting into the membrane 27  CHAPTER 1  bilayers, or vice versa. Thus, one can easily monitor the membrane insertion state of a given antimicrobial peptide as a function of peptide concentration and deduce a peptide insertion profile. Moreover, one can compare the insertion profiles among various peptides under study in the membrane bilayers with the same lipid composition, or a specific peptide in the membrane bilayers with different lipid composition. With this information, one can determine whether the peptide-membrane interaction examined is peptide-dependent and/or lipid-specific.  Figure 1.14. OCD spectra of (a) alamethicin and (b) protegrin-1 demonstrate that the peptides adopt the inserted (state I) and surface state (state S) in DPhPC multilayers: The spectra show the two extreme cases of alamethicin and protegrin-1 in different orientation. (a) The solid spectrum shows a single maximum at 195 nm and two minima at 207 and 222 nm which indicates that the peptide is in the surface-adsorbed state. The dotted spectrum shows reduced minimum at 207 nm, which indicates that the peptide is in the inserted state. (b) The solid spectrum shows a single maximum at 195 nm and one small minimum at 215 nm which indicates that the peptide is in the inserted state. The dotted spectrum shows only one minimum at 200 nm, which indicates that the peptide is in the surface-adsorbed state. (Figure 1.14(a) reproduced from “Wu, Y. et al., 1990” (91); Figure 1.14(b) reproduced from “Heller, W. T. et al., 1998” (93)).  Note that this technique still has some limitations. For example, OCD spectroscopy can give information on the dynamics of the membrane insertion states only at the equilibrium state, but cannot provide information on the kinetics of the membrane insertion process. OCD spectra also do not provide precise information on the actual insertion angle of the peptide in the lipid 28  CHAPTER 1  bilayers. Despite these limitations, OCD spectroscopy is still a very useful tool to assess how effectively a CAP can create membrane defects. Many studies have examined the membrane insertion ability of CAPs in model membrane bilayers, including protegrin-1 (93), aurein 1.2 and citropin 1.1 (62). Figure 1.14 shows the OCD spectra of alamethicin (an α-helical CAP) and protegrin-1 (a β-sheet CAP) in 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC) bilayers, as examples (93).  1.4.2  uclear magnetic resonance ( MR) spectroscopy  1.4.2.1 Brief overview of MR spectroscopy NMR spectroscopy is a technique that makes use of the fact that nuclei have magnetic moments (µ µ). When nuclei are placed in an external magnetic field B0, they will precess according to Equation 1.1:  E = hν 0 = − γhB0 Equation 1.1.  Relationship between the energy and precession frequency of magnetic moment.  where E is the energy [J], h is the Planck constant [6.62606896(33)×10−34 J⋅s], ν0 is the precession frequency or the Larmor frequency of the nucleus [Hz], γ is the gyromagnetic ratio of the nucleus [Hz⋅T-1], ћ is the reduced Planck constant (reduced by 2π) [1.054571628(53)×10−34 J⋅s], and B0 is the external magnetic field [Tesla]. For nuclei with spin ½, there are two energy levels when these nuclei are placed in a constant magnetic field. The energy difference between the two states is ∆E = 2µ µB. According to Boltzmann distribution, the ratio of nuclei in the upper and lower energy level is given by the Boltzmann equation: 29  CHAPTER 1 1  =e  − ∆E kT  2  Equation 1.2. The Boltzmann distribution  where N1 is the number of unuclei in the upper energy level and N2 is that in the lower, ∆E is the energy difference between the two levels, k is the Boltzmann’s constant [1.38066 × 10-23 J⋅K-1] and T is the temperature [K]. Figure 1.15 illustrates this energy difference.  Figure 1.15. The energy difference between the two levels in protons (I = ½). mI = -½ is the higher energy level with spins antiparallel to the external magnetic field B0. mI = ½ is the lower energy level with spins parallel to the external magnetic field B0.  Boltzmann distribution predicts that nuclei will favour the lower energy level in order to reach the equilibrium state when they are placed in B0. If they are irradiated with a radio frequency (RF), they will be excited to the higher energy level. When they return to the lower energy level, they will emit signals that give information about their chemical environment. There are various types of NMR spectroscopy.  The most commonly used are 1H  (proton), 2H (deuterium), and 13C (carbon-13) NMR spectroscopy. Others include 3He (helium), 15  N (nitrogen), and  31  P (phosphorus) NMR spectroscopy. In our studies, we used 1H and  31  P  NMR spectroscopy. 30  CHAPTER 1 1  1.4.2.2 H MR spectroscopy In this thesis, solution-state 1H NMR spectroscopy was used to determine the structures of the aurein peptides and the solvent accessibility of daptomycin in different membrane environments. 1H NMR spectroscopy does not have sample preparation restraints as those of Xray crystallography, where suitable crystals are necessary to yield good diffraction data. An NMR experiment usually involves five steps: sample preparation, data collection, resonance assignment, restraint generation, and structure calculation. When a specific pulse sequence is applied to a sample, different information on the chemical environment around the examined molecules can be obtained.  We used pulse sequences from total correlation spectroscopy  (TOCSY) (94) and nuclear Overhauser enhancement spectroscopy (NOESY) (95) to probe bond and spatial connectivities between 1H nuclei, in order to assign resonances and to calculate a structure. In addition, we also used NOESY to investigate paramagnetic relaxation effects (PRE) on the structure and oligomerization state of daptomycin in different chemical environments. These techniques allow us to obtain structural information on these systems.  1.4.2.3 31P MR spectroscopy 31  P NMR spectroscopy is a powerful technique to study the membrane perturbation  mechanism of a given antimicrobial peptide. Since most antimicrobial peptides interact with the membrane surface initially and are in contact with membrane headgroups directly, it is crucial to establish how these peptides interact with membrane headgroups and whether these interactions lead to membrane perturbation.  31  P NMR signals are sensitive to the orientations of the  phosphorus headgroups of lipid membranes. Thus, if the addition of a given antimicrobial peptide disrupts the membrane integrity, then it would usually disrupt the lipid headgroup alignment, which can easily result in orientational changes of the lipid headgroups, giving rise to a different 31P NMR spectrum. 31  CHAPTER 1 31  P NMR studies can be conducted in various lipid mixtures in different physical forms:  unaligned multilayers, bicelles, or mechanically aligned bilayers. Each physical form of the lipid mixture gives rise to a characteristic 31P NMR spectrum. Figure 1.16 shows 31P NMR spectra of lipid dispersions in each physical state.  Figure 1.16.  31  P MR spectra of lipid/s in different physical states:  (a) A theoretical 31P powder pattern spectrum, showing orientational dependence of 31P signals on the phosphorus headgroups of the phospholipid/s examined; (b) unaligned perdeuterated DMPC (DMPCd54) multilayers; (c) magnetically aligned DMPC/DHPC bicelles; (d) mechanically aligned POPC/POPG bilayers. (Figure 1.16(a) reproduced from “Chia, B. C. S. et al., 2000” (96), (b) from “Balla, M. S. et al., 2004” (61), (c) from “Marcotte, I. et al., 2003” (62), (d) from “Lu, J. et al., 2006” (97) with minor modification).  Unaligned lipid multilayers usually display a powder-pattern, which covers all orientations. Bicelles, on the other hand, normally exhibit a spectrum consisting of one or more major peaks, depending on the number of lipids in the mixture. Mechanically aligned lipid bilayers, due to uniaxial orientation of the phosphorus headgroups, mostly show one single, slightly broad peak. Several  31  P NMR studies of antimicrobial peptides were conducted in  mechanically aligned lipid bilayers.  Due to different shapes of different phospholipids, 32  CHAPTER 1  phosphatidylcholine and phosphatidylethanolamine are commonly used to achieve a better alignment owing to a close to cylindrical shape. In the absence of peptide, most mechanically aligned lipid bilayers will give a single peak at a characteristic chemical shift position, depending on the lipid mixture used. In the presence of peptide, different disordering effects will give rise to different 31P signals as a result of orientational changes of the headgroups. Figure 1.17 shows 31  P NMR spectra of mechanically aligned POPC/POPG bilayers in the absence and presence of  LL7-27, a very recently studied LL-37 analogue (98).  Figure 1.17.  31  P MR spectra of mechanically aligned 3:1 POPC/POPG (mol/mol) bilayers in the absence and presence of LL7-27 at 37°C: (A) POPC/POPG bilayers alone; POPC/POPG bilayers containing (B) 3 mol % of LL7-27; and (C) 3 mol % of LL7-27 and 15 mol % of cholesterol. (Figure reproduced from “Thennarasu, S. et al., 2010”(98)).  1.4.3 Membrane leakage assays and antimicrobial activity 1.4.3.1 Calcein release assay and DiSC35 assay Both calcein release and DiSC35 assays monitor the real time release of fluorescent dye from model membranes or real bacteria, respectively. These two methods are particularly useful to examine the ability of an antimicrobial peptide to induce membrane leakage.  As most 33  CHAPTER 1  antimicrobial peptides are found to act on bacterial membranes, these assessments correlate with antimicrobial  activity.  The  calcein  release  assay  monitors  the  release  of  Bis[ , −bis(carboxymethyl)aminomethyl]fluorescein (calcein) from model membranes upon the addition of an antimicrobial peptide. In the absence of peptide, calcein is trapped and selfquenched inside liposomes, so no fluorescence should be observed. If the presence of peptide which can create membrane defects, membrane disruption or disintegration occurs, and therefore calcein is released from liposomes, which results in an increase in fluorescence. Similarly, DiSC35 assay monitors the release of 3,3'-dipropylthiadicarbocyanine iodide (DiSC35), but from bacterial membranes. The real time property of these assays allows one to inspect the membrane disrupting efficiency of the peptides studied. Generally, the more efficient a peptide is to induce membrane leakage, the faster the appearance and the greater the magnitude of fluorescence observed. If a peptide can disrupt the membrane at a multiple magnitude of its MIC value, then one should be able to observe a dramatic increase in fluorescence when the peptide concentration increases. These fluorescence data can be compared as a function of different peptides with respect to a control peptide or detergent, or as a function of lipid composition or peptide-to-lipid ratio. In addition, a unique part of these assays is that one is able to perform membrane leakage kinetic studies on a given antimicrobial peptide, whereas OCD and  31  P NMR spectroscopy  provide mostly information at equilibrium.  1.4.3.2 Minimum inhibitory concentration (MIC) assay All the methods mentioned above give observations on the behaviour of an antimicrobial peptide in model systems, or bacterial membranes only. Whether these observations correlate with those in intact bacteria as a whole must still be examined. The determination of the MIC is a commonly used method that assesses the antimicrobial activity of a given antimicrobial peptide. The antimicrobial activity is defined by the minimal inhibitory concentration (MIC), 34  CHAPTER 1  which is the minimal amount of peptide required to inhibit bacterial growth. Generally, an antimicrobial peptide has an MIC of 2 ~ 16 µg/ml (very active), 32 ~ 64 µg/ml (moderately active), or 128 ~ 256 µg/ml (marginally active) against a specific bacterium. An MIC greater than 256 µg/ml usually indicates that the peptide has negligible antimicrobial activity.  1.5 Aim of this dissertation Even though many studies have been conducted to characterize the structure-function relationship of various antimicrobial peptides, there are still many unknowns about how and why an antimicrobial peptide is active or behaves in a certain way. It is well known that an animal can contain numerous antimicrobial peptides of conservative sequences with minor differences, yet it has not been examined extensively how these minor differences in the sequence could affect antimicrobial functions. Since many studies were conducted in model membranes, the effect of membrane lipid composition on the behaviour of peptides has not been examined in detail. Many studies have inspected analogues of various antimicrobial peptides, hoping to find a more efficient peptide. However, very few studies have actually investigated why Nature selects a particular sequence for an antimicrobial peptide, and whether specific residues are removable or not. This is of particular importance, as designing a better peptide analogue also requires the knowledge on the essential existence of specific residues in the sequence. Immunomodulatory peptides have been recently discovered to play a significant role in antimicrobial activities (99), and we will show here that some aurein peptide analogues actually exhibit immunomodulatory functions. Lastly, more and more studies have been conducted to characterize the structures of lipopeptides. However, there are still a lot of unknowns about how lipopeptides interact with bacterial membranes at the molecular level.  35  CHAPTER 1  In this dissertation, I will mainly address four major perspectives of our studies on the aurein 2.2 and aurein 2.3 peptides and their analogues (CAPs), and daptomycin (lipopeptides).  1.5.1 Structure-function relationship of aurein 2 peptides As previously mentioned, aurein 1.2 is one of the most widely studied peptides in the aurein family. We were interested in characterizing the structure-function relationship of the aurein-2-family peptides. Particularly, we were interested in aurein 2.2 and aurein 2.3, since they have nearly identical sequences but yet have very different reported antimicrobial activities. Both aurein 2.2 and aurein 2.3 are 16-residue peptides with a net charge of +2, but they differ only by one amino acid residue at position 13 (L13 vs. I13). Table 1.2 shows the amino acid sequences and MIC values of the three aurein peptides and citropin 1.1 (62,72).  Table 1.2.  Peptide Aurein 1.2 Aurein 2.2 Aurein 2.3 Citropin 1.1  Amino acid sequences, molecular weights (MW, g/mol), charges and minimal inhibitory concentrations (MICs, µg/ml) of aurein 1.2, aurein 2.2, aurein 2.3 and citropin 1.1, according to Rozek, T. et al., 2000 (72).  Peptide information Sequence MW (g/mol) GLFDIIKKIAESF-NH2 1480.76 GLFDIVKKVVGALGSL-NH2 1614.98 GLFDIVKKVVGAIGSL-NH2 1615.97 GLFDVIKKVASVIGGL-NH2 1614.98  Charge +2 +2 +2 +2  MIC (µ µg/ml) S. aureus S. epidermidis 50 50 25 25 100 100 25 25  Previous research reported that aurein 2.2 is more active than aurein 1.2, whereas aurein 2.3 is only marginally active (72). How a single conservative amino-acid mutation leads to this big difference in the antimicrobial activity of the two peptides needs to be examined. Our goal was to determine whether this difference in antimicrobial activity is caused by a difference in structure or ability to interact with bacterial membranes.  By understanding how small  perturbations in the sequence relate to changes in structure and function, we hoped to gain a better insight into design rules for development of future antibiotics. We have investigated by determining the difference in the structure of the three aurein peptides (aurein 2.2, aurein 2.3 and  36  CHAPTER 1  its carboxy C-terminal analogue, aurein 2.3-COOH) and then correlating any difference in the structure to the difference in the antimicrobial activity.  This would provide important  information on how a slightest change in the sequence could lead to a significant difference in the activity, which will be demonstrated in Chapter 2.  1.5.2 Choice of lipid composition for model membranes Selecting the relevant model membrane for conducting a study on a given antimicrobial peptide is equally important. Previous studies have carried out experiments in various model membranes. The choice of model membranes evolved from pure (62) to PC/PG (100) mixtures to PE/PG (101,102) mixtures. Different acyl chain lengths and degrees of saturation also play a role in membrane models. 1,2-dimyristoyl chains were traditionally chosen for the ease of lipid bilayer alignment (103), even though 1-palmitoyl-2-oleoyl chains were also highly used (57). Despite higher lamellar to liquid crystalline phase transition temperature (Tm), several studies also used 1,2-dipalmitoyl chains (for example, DPPC/DPPG, Tm = 41°C) to have a model membrane consisting of long-chained lipids (104-106). Some studies used 1,2-dioleoyl chains (for example, DOPC/DOPG, Tm = -20°C ~ -18°C) instead to compensate the high Tm issue while not sacrificing the need for long-chained lipids (107,108). All these models were used to account for different bilayer thickness and fluidity for studying antimicrobial peptides. Some studies have reported that lipid composition has pronounced effects on the behaviour of antimicrobial peptides, such as protegrin-1 (109) and pardaxin (110). Not until recently, the importance of using a relevant model membrane has caught major attention. Since different bacteria have very different membrane lipid composition, it is crucial to use model membranes that best mimic lipid membranes of the target bacteria. To our best knowledge, not many studies have investigated this aspect quantitatively. Our aim was to quantitatively describe the behaviour of the aurein peptides from the lipid membrane’s point of view. This would 37  CHAPTER 1  provide important perspectives on how different headgroups, membrane surface charge, or acyl chain length impose different effects and establish a reference for choosing a relevant model membrane for future antimicrobial studies. These findings will be reported in Chapter 3 and Chapter 4.  1.5.3 Designing CAP analogues Various studies have designed mutant analogues from currently known CAPs to probe for a better, more efficient antimicrobial alternative. Most analogues were designed by changing the structure through cyclization, loop formation or oligomerization, as shown for a magainin 2 dimer analogue (111), or by having a different number of positively charged residues (112). Very recently, citropin 1.1 analogues were studied using 1H NMR spectroscopy to examine the thermodynamic stability of each analogue and determine the structural perturbation of placing different residues at designated positions (113). To our best knowledge, not many studies have been conducted to determine the essential residues of aurein peptides, and whether substituting or removing certain residues would enhance or abolish antimicrobial activities and peptidemembrane interactions. Since aurein 2.2 and aurein 2.3 differ only by one residue at position 13 and most aurein peptides differ at the C-terminus, our goal was to understand the importance of residue 13 and the C-terminus of aurein 2.2, and their impact on the peptide-membrane interaction and activity (Chapter 5). This information would also allow us to assess whether future design of a better antimicrobial peptide based on aurein 2.2 can or should be approached from changing residue 13 or C-terminal sequence. Since most aurein peptides have a conserved N-terminus, we also wanted to determine whether removing N-terminal residues would improve or obliterate the antimicrobial activity.  As expected, N-terminal truncation eliminated the  antimicrobial function but in turn, to our surprise, gives peptides an enhanced immunomodulatory activity in vitro (Chapter 6). 38  CHAPTER 1  1.5.4 Lipopeptides Lipopeptides such as daptomycin have become a major interest to many pharmaceutical companies.  Despite the fact that daptomycin has passed clinical trials and become  commercialized as Cubicin for the last line of defense against S. aureus infection, the mechanism of action of other lipopeptides is still unknown. Particularly, daptomycin’s interactions with bacterial membranes have not been fully characterized yet. Our aim was to examine how tightly packed daptomycin molecules are in the absence and presence of Ca2+ and lipid membranes. We want to know whether aggregates of daptomycin and Ca2+ insert directly into lipid membranes or first dissociate into monomers, insert and reassociate to form aggregates in lipid membranes. This would provide further evidence to clarify some of the intermediate steps in the mechanism for daptomycin-membrane interactions (40). These observations will be given in Chapter 7.  1.5.5 Conclusion Taken together all these studies will ultimately serve to help develop novel antibiotics based on these CAPs and lipopeptides in the future. In order to do so, we have dissected the structure-activity relationship of aurein 2 peptides and daptomycin in depth, hoping to establish useful references for future designs of different peptide analogues, in order to fight the fast growing antibiotic resistance.  39  CHAPTER 1  1.6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.  12.  13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.  Forrest, R. D. J R Soc Med 1982, 75, 198-205. Lindblad, W. J. Int J Low Extrem Wounds 2008, 7, 75-81. Wainwright, M. Mycologist 1989, 3, 21-23. Chew, R., "Louis Pasteur, Chemist, 1822 - 1895," Lucidcafe Library, 2010, http://www2.lucidcafe.com/lucidcafe/library/95dec/pasteur.html "Antibiotic," Wikipedia, The Free Encyclopedia, 2010, http://en.wikipedia.org/wiki/Antibiotic Lu, Y. and Haeberlin, A., "History of Antibiotics," ORACLE ThinkQuest, Education Foundation, 2010, http://library.thinkquest.org/25462/history.html Kasten, B. and Reski, R. Journal of Plant Physiology 1997, 150, 137-140. "Gentamicin," Wikipedia, The Free Encyclopedia, 2010, http://en.wikipedia.org/wiki/Gentamicin "Antibiotic Resistance," U.S. Food and Drug Administration, 2010, http://www.fda.gov/ Administration., U. S. F. a. D. 2010. Devitt, T., Gen, Y., Schulte, D., Tenenbaum, D., Toburen, A. and Trebach, S., "What doesn’t kill bacteria makes them stronger. It’s all in the teamwork.," An NISE Project founded by NSF. The Why Files., 2007, http://whyfiles.org/038badbugs/mechanism.html Yim, G., "ATTACK OF THE SUPERBUGS: ANTIBIOTIC RESISTANCE," The Science Creative Quarterly, 2009, http://www.scq.ubc.ca/attack-of-the-superbugsantibiotic-resistance/ Webber, M. A. and Piddock, L. J. J Antimicrob Chemother 2003, 51, 9-11. Delcour, A. H. Biochim Biophys Acta 2009, 1794, 808-16. Moore, S. D. and Sauer, R. T. Proc atl Acad Sci U S A 2008, 105, 18261-6. Auerbach, T., Bashan, A. and Yonath, A. Trends Biotechnol 2004, 22, 570-6. Wright, G. D. Adv Drug Deliv Rev 2005, 57, 1451-70. Walsh, C. ature 2000, 406, 775-81. Tenover, F. C. Am J Med 2006, 119, S3-10; discussion S62-70. Alfaro, C., Fergie, J. and Purcell, K. Pediatr Infect Dis J 2005, 24, 274-6. Buckingham, S. C., McDougal, L. K., Cathey, L. D., Comeaux, K., Craig, A. S., Fridkin, S. K. and Tenover, F. C. Pediatr Infect Dis J 2004, 23, 619-24. Prevention, C. f. D. C. a. Morb. Mortal. Wkly. Rep. 1999, 48, 707-710. Zasloff, M. ature 2002, 415, 389-95. "Antimicrobial peptides," Wikipedia, The Free Encyclopedia, 2010, http://en.wikipedia.org/wiki/Antimicrobial_peptides Devine, D. A. and Hancock, R. E. W. Curr Pharm Des 2002, 8, 703-14. Hancock, R. E. W. Lancet Infect Dis 2001, 1, 156-64. Hancock, R. E. W. Lancet Infect Dis 2005, 5, 209-18. Tally, F. P. and DeBruin, M. F. J Antimicrob Chemother 2000, 46, 523-6. Hancock, R. E. and Diamond, G. Trends Microbiol 2000, 8, 402-10. "Lactoferrin," Wikipedia, The Free Encyclopedia, 2010, http://en.wikipedia.org/wiki/Lactoferrin "Cathepsin G," Wikipedia, The Free Encyclopedia, 2010, http://en.wikipedia.org/wiki/Cathepsin_G Huang, H. W. Biochemistry 2000, 39, 8347-52. Matsuzaki, K. Biochim Biophys Acta 1999, 1462, 1-10. 40  CHAPTER 1  34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.  45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.  Shai, Y. Biochim Biophys Acta 1999, 1462, 55-70. Chen, F. Y., Lee, M. T. and Huang, H. W. Biophys J 2003, 84, 3751-8. Strieker, M. and Marahiel, M. A. Chembiochem 2009, 10, 607-16. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother 1999, 43, 1317-23. Wu, M., Maier, E., Benz, R. and Hancock, R. E. W. Biochemistry 1999, 38, 7235-42. Bechinger, B. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1529-1539. Straus, S. K. and Hancock, R. E. W. Biochim Biophys Acta 2006, 1758, 1215-1223. Papo, N. and Shai, Y. Cell Mol Life Sci 2005, 62, 784-90. Pukala, T. L., Bowie, J. H., Maselli, V. M., Musgrave, I. F. and Tyler, M. J. at Prod Rep 2006, 23, 368-93. Rinaldi, A. C. Curr Opin Chem Biol 2002, 6, 799-804. Apponyi, M. A., Pukala, T. L., Brinkworth, C. S., Maselli, V. M., Bowie, J. H., Tyler, M. J., Booker, G. W., Wallace, J. C., Carver, J. A., Separovic, F., Doyle, J. and Llewellyn, L. E. Peptides 2004, 25, 1035-54. Barra, D. and Simmaco, M. Trends Biotechnol 1995, 13, 205-9. Bevins, C. L. and Zasloff, M. Annu Rev Biochem 1990, 59, 395-414. Lazarus, L. H. and Attila, M. Prog eurobiol 1993, 41, 473-507. Seto, G. W., Marwaha, S., Kobewka, D. M., Lewis, R. N., Separovic, F. and McElhaney, R. N. Biochim Biophys Acta 2007, 1768, 2787-800. Giovannini, M. G., Poulter, L., Gibson, B. W. and Williams, D. H. Biochem J 1987, 243, 113-20. Zasloff, M. Proc atl Acad Sci U S A 1987, 84, 5449-53. Zasloff, M., Martin, B. and Chen, H. C. Proc atl Acad Sci U S A 1988, 85, 910-3. Bechinger, B. Biochim Biophys Acta 2005, 1712, 101-8. Bechinger, B., Zasloff, M. and Opella, S. J. Protein Sci 1993, 2, 2077-84. Duclohier, H., Molle, G. and Spach, G. Biophys J 1989, 56, 1017-21. Hallock, K. J., Lee, D. K. and Ramamoorthy, A. Biophys J 2003, 84, 3052-60. Jacob, L. and Zasloff, M. Ciba Found Symp 1994, 186, 197-216; discussion 216-23. Ramamoorthy, A., Thennarasu, S., Lee, D. K., Tan, A. and Maloy, L. Biophys J 2006, 91, 206-16. Yang, L., Weiss, T. M., Lehrer, R. I. and Huang, H. W. Biophys J 2000, 79, 2002-9. Ambroggio, E. E., Separovic, F., Bowie, J. and Fidelio, G. D. Biochim Biophys Acta 2004, 1664, 31-7. Ambroggio, E. E., Separovic, F., Bowie, J. H., Fidelio, G. D. and Bagatolli, L. A. Biophys J 2005, 89, 1874-81. Balla, M. S., Bowie, J. H. and Separovic, F. Eur Biophys J 2004, 33, 109-16. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids 2003, 122, 107-20. Niidome, T., Kobayashi, K., Arakawa, H., Hatakeyama, T. and Aoyagi, H. J Pept Sci 2004, 10, 414-22. Boland, M. P. and Separovic, F. Biochim Biophys Acta 2006, 1758, 1178-83. Chen, T., Li, L., Zhou, M., Rao, P., Walker, B. and Shaw, C. Peptides 2006, 27, 42-8. Conlon, J. M., Al-Ghaferi, N., Abraham, B., Jiansheng, H., Cosette, P., Leprince, J., Jouenne, T. and Vaudry, H. Peptides 2006. Kumari, V. K. and Nagaraj, R. J Pept Res 2001, 58, 433-41. Morikawa, N., Hagiwara, K. and Nakajima, T. Biochem Biophys Res Commun 1992, 189, 184-90. Simmaco, M., Mignogna, G., Barra, D. and Bossa, F. FEBS Lett 1993, 324, 159-61.  41  CHAPTER 1  70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98.  Won, H. S., Kim, S. S., Jung, S. J., Son, W. S., Lee, B. and Lee, B. J. Mol Cells 2004, 17, 469-76. Rozek, T., Bowie, J. H., Wallace, J. C. and Tyler, M. J. Rapid Commun Mass Spectrom 2000, 14, 2002-11. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem 2000, 267, 5330-41. Wabnitz, P. A., Bowie, J. H., Wallace, J. C. and Tyler, M. J. Rapid Commun Mass Spectrom 1999, 13, 1724-32. Chia, B. C., Carver, J. A., Mulhern, T. D. and Bowie, J. H. Eur J Biochem 2000, 267, 1894-908. Wegener, K. L., Wabnitz, P. A., Carver, J. A., Bowie, J. H., Chia, B. C., Wallace, J. C. and Tyler, M. J. Eur J Biochem 1999, 265, 627-37. Bradford, A. M., Bowie, J. H., Tyler, M. J. and Wallace, J. C. Australian Journal of Chemistry 1996, 49, 1325-1331. Chia, C. S., Torres, J., Cooper, M. A., Arkin, I. T. and Bowie, J. H. FEBS Lett 2002, 512, 47-51. Chia, B. C. S., Lam, Y. H., Dyall-Smith, M., Separovic, F. and Bowie, J. H. Lett. Pept. Sci. 2000b, 7, 151-156. Wang, G., Li, Y. and Li, X. J Biol Chem 2005, 280, 5803-11. Giacometti, A., Cirioni, O., Riva, A., Kamysz, W., Silvestri, C., Nadolski, P., Vittoria, A. D., Lukasiak, J. and Scalise, G. Antimicrob Agents Chemother 2007. Vaara, M. and Saxen, H. Duodecim 1999, 115, 2259-64. McCafferty, D. G., Cudic, P., Yu, M. K., Behenna, D. C. and Kruger, R. Curr Opin Chem Biol 1999, 3, 672-80. Fasman, G. D. Circular Dichroism and the Conformational Analysis of Biomolecules; Plenum Press: New York, 1996. Johnson, W. C., Jr. Proteins 1990, 7, 205-14. Woody, R. W. Methods Enzymol 1995, 246, 34-71. Greenfield, N. and Fasman, G. D. Biochemistry 1969, 8, 4108-16. Andra, J., Jakovkin, I., Grotzinger, J., Hecht, O., Krasnosdembskaya, A. D., Goldmann, T., Gutsmann, T. and Leippe, M. Biochem J 2008, 410, 113-22. Fahrner, R. L., Dieckmann, T., Harwig, S. S., Lehrer, R. I., Eisenberg, D. and Feigon, J. Chem Biol 1996, 3, 543-50. Rozek, A., Friedrich, C. L. and Hancock, R. E. Biochemistry 2000, 39, 15765-74. Wieprecht, T., Dathe, M., Beyermann, M., Krause, E., Maloy, W. L., MacDonald, D. L. and Bienert, M. Biochemistry 1997, 36, 6124-32. Wu, Y., Huang, H. W. and Olah, G. A. Biophys J 1990, 57, 797-806. de Jongh, H. H., Goormaghtigh, E. and Killian, J. A. Biochemistry 1994, 33, 14521-8. Heller, W. T., Waring, A. J., Lehrer, R. I. and Huang, H. W. Biochemistry 1998, 37, 17331-8. Davis, D. G. and Bax, A. Journal of the American Chemical Society 1985, 107, 28202821. Jeener, J., Meier, B. H., Bachmann, P. and Ernst, R. R. The Journal of Chemical Physics 1979, 71, 4546-4553. Chia, B. C. S., Lam, Y. H., Dyall-Smith, M., Separovic, F. and Bowie, J. H. Letters in Peptide Science 2000, 7, 151-156. Lu, J. X., Blazyk, J. and Lorigan, G. A. Biochim Biophys Acta 2006, 1758, 1303-13. Thennarasu, S., Tan, A., Penumatchu, R., Shelburne, C. E., Heyl, D. L. and Ramamoorthy, A. Biophys J 2010, 98, 248-57. 42  CHAPTER 1  99. 100. 101.  102. 103. 104. 105.  106. 107. 108. 109. 110. 111. 112. 113.  Easton, D. M., Nijnik, A., Mayer, M. L. and Hancock, R. E. Trends Biotechnol 2009, 27, 582-90. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry 2006, 45, 5793-9. Chekmenev, E. Y., Vollmar, B. S., Forseth, K. T., Manion, M. N., Jones, S. M., Wagner, T. J., Endicott, R. M., Kyriss, B. P., Homem, L. M., Pate, M., He, J., Raines, J., Gor'kov, P. L., Brey, W. W., Mitchell, D. J., Auman, A. J., Ellard-Ivey, M. J., Blazyk, J. and Cotten, M. Biochim Biophys Acta 2006, 1758, 1359-72. Chen, T., Scott, C., Tang, L., Zhou, M. and Shaw, C. Regul Pept 2005, 128, 75-83. Lewis, R. N., Zhang, Y. P. and McElhaney, R. N. Biochim Biophys Acta 2005, 1668, 203-14. Klajnert, B., Janiszewska, J., Urbanczyk-Lipkowska, Z., Bryszewska, M. and Epand, R. M. Int J Pharm 2006, 327, 145-52. Zweytick, D., Pabst, G., Abuja, P. M., Jilek, A., Blondelle, S. E., Andra, J., Jerala, R., Monreal, D., Martinez de Tejada, G. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1426-35. Leidy, C., Linderoth, L., Andresen, T. L., Mouritsen, O. G., Jorgensen, K. and Peters, G. H. Biophys J 2006, 90, 3165-75. Wimmer, R., Andersen, K. K., Vad, B., Davidsen, M., Molgaard, S., Nesgaard, L. W., Kristensen, H. H. and Otzen, D. E. Biochemistry 2006, 45, 481-97. Epand, R. F., Schmitt, M. A., Gellman, S. H. and Epand, R. M. Biochim Biophys Acta 2006, 1758, 1343-50. Ishitsuka, Y., Pham, D. S., Waring, A. J., Lehrer, R. I. and Lee, K. Y. Biochim Biophys Acta 2006, 1758, 1450-60. Hallock, K. J., Lee, D. K., Omnaas, J., Mosberg, H. I. and Ramamoorthy, A. Biophys J 2002, 83, 1004-13. Tencza, S. B., Creighton, D. J., Yuan, T., Vogel, H. J., Montelaro, R. C. and Mietzner, T. A. J Antimicrob Chemother 1999, 44, 33-41. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett 2001, 501, 146-50. Sikorska, E., Greber, K., Rodziewicz-Motowidlo, S., Szultka, L., Lukasiak, J. and Kamysz, W. J Struct Biol 2009, 168, 250-8.  43  CHAPTER 2  CHAPTER 2: Characterization of the structure and membrane interaction of the antimicrobial peptides aurein 2.2 and 2.3 from Australian Southern Bell frogs†  2.1 Introduction The previous chapter introduced one of the most widely studied amphibian cationic antimicrobial peptides, aurein 1.2, whose structure-function relationship has been examined extensively. In this chapter, we report data on the structure, membrane interaction, and activity of the two 16-residue peptides from the same family: aurein 2.2 (net charge +2) and aurein 2.3 (net charge +2). Aurein 2.2 is active against a number of Gram positive bacteria. For example, it displays an minimal inhibitory concentration (MIC) of 25 µg/ml against Staphylococcus aureus and Staphylococcus epidermidis (1). This makes aurein 2.2 more active than aurein 1.2  and equally active to citropin 1.1. Aurein 2.3, on the other hand, is generally only marginally active, with a typical MIC of 100 µg/ml (1). Interestingly, the difference in amino acid sequence between aurein 2.2 and 2.3 is only a conservative mutation of a leucine to an isoleucine at position 13. In order to assess how this mutation may be correlated to activity, we have determined the structure of aurein 2.2 and aurein 2.3 in TFE, in 1,2-dimyristoyl-sn-glycero-3phosphocholine (DMPC) small unilamellar vesicles (SUVs) and in 1:1 DMPC/1,2-dimyristoylsn-glycero-3-[phospho-rac-(1-glycerol)] (DMPG) SUVs, using circular dichroism (CD). The †  A version of this chapter has been published as follows:  Pan, Y.-L*, Cheng, J. T. J.*, Hale, J., Pan, J., Hancock, R. E. W. and Straus, S. K. 2007. Characterization of the Structure and Membrane Interaction of the Antimicrobial Peptides Aurein 2.2 and 2.3 from Australian Southern Bell Frogs. Biophys. J. 92: 2854-2864. *  Both authors contributed equally to this work. 44  CHAPTER 2 1  structures of the peptides in 25% TFE (v/v) were also determined using solution-state H NMR. In order to further examine how activity is modulated by sequence, we have also studied a modified version of aurein 2.3 in which the amidated C-terminus is replaced by a carboxyl group (aurein 2.3-COOH). In addition, we have investigated the interaction of these three peptides with model zwitterionic (DMPC) membranes, using oriented CD (OCD) and solid-state  31  P  NMR, in order to determine how they perturb the membrane bilayers. We have also inspected how these peptides interact with model bacterial membranes consisting of a 1:1 molar mixture of DMPC/DMPG using OCD. Finally, we have tested the antimicrobial activity of the three peptides against S. aureus (strain C622, ATCC 25923) and S. epidermidis (strain C621, clinical isolate). Overall, these data should enable us to correlate structure and membrane interaction of these three marginally different peptides with antimicrobial activity and lead to a better understanding of how sequence modulates function in the aurein peptide family. This chapter will give an overview of the studies on the structure and membrane interaction of the three aurein peptides. This chapter will be divided into three sections. Section 2.2 outlines the structural and functional analyses of the three aurein peptides. Section 2.3 discusses the significance of these findings and whether a conservative mutation can lead to activity differences on a molecular basis. Section 2.4 summarizes the findings and presents conclusions from this study.  2.2 Results This section will be divided into three main subsections. Section 2.2.1 describes whether the three aurein peptides adopt different structures or undergo different conformational changes in the presence of aqueous and lipid environment. Section 2.2.2 explains how the three aurein peptides behave in the lipid bilayers and their possible mode of action against model bacterial  45  CHAPTER 2  membranes. Finally, Section 2.2.3 connects the results obtained in the live bacteria to the observations in model membranes.  2.2.1 Determining the structure of the aurein peptides Characterizing the structure is a first step toward understanding the structure-function relationship of antimicrobial peptides. By using solution CD spectroscopy, we were able to find the conformational changes of the three aurein peptides in model membranes. By using solutionstate 1H NMR spectroscopy, we were able to confirm the sequential and spatial connectivities among the amino acids of individual aurein peptides in model membranes.  2.2.1.1 Using solution CD spectroscopy to identify the structure of the aurein peptides in the presence of ddH2O, TFE and phospholipids To determine which structure aurein 2.2, aurein 2.3, and aurein 2.3-COOH adopt in water and in the presence of model membranes, solution CD experiments were performed. Figure 2.1 shows solution CD spectra of the three aurein peptides in the presence of ddH2O and in 25%, 50%, and 75% (v/v) TFE. In the presence of 100% ddH2O, the spectra of the three aurein peptides exhibited a minimum at 197 nm which is characteristic of the spectra of peptides adopting a random-coil conformation. In the presence of ddH2O/TFE, the spectra of the three aurein peptides showed two minima at 207 nm and 222 nm and a maximum at 190 nm which are characteristic of the spectra of peptides adopting an α-helical conformation. The spectra indicate that all three aurein peptides were not structured in the presence of water, but underwent structural change and adopted an α-helical conformation in the presence of TFE. The increase in the concentration of TFE did not change the α-helical conformation of the three aurein peptides, which is evident from the identical spectra of the three aurein peptides at all concentrations of TFE.  This 46  CHAPTER 2  indicates that the conformational change of the three aurein peptides from random coil to α-helix was independent of the TFE concentrations examined.  Figure 2.1.  Solution CD spectra of the aurein peptides in ddH2O and ddH2O/TFE mixtures: (a) aurein 2.2, (b) aurein 2.3 and (c) aurein 2.3-COOH (solid black line: 100% ddH2O; dotted black line: 75% ddH2O/25% TFE; solid grey line: 50% ddH2O/50% TFE; grey dotted line: 25% ddH2O/75% TFE). The spectra indicate that, in all cases, the peptides were unstructured in ddH2O, but adopted an α-helical conformation upon addition of TFE. (Figure 2.1(b) and (c) are courtesy of Y. L. Pan).  As TFE has been shown to promote the formation of α-helices (for a recent example see (2)), solution CD experiments of aurein 2.2, aurein 2.3, and aurein 2.3-COOH were also conducted in DMPC and 1:1 DMPC/DMPG (mol/mol) small unilamellar vesicles (SUVs) to determine the true conformation in a membrane environment. Figure 2.2 illustrates solution CD 47  CHAPTER 2  spectra of the three aurein peptides in the presence of DMPC SUVs as a function of peptide-tolipid (P/L) molar ratio.  Figure 2.2.  Solution CD spectra of the aurein peptides in DMPC small unilamellar vesicles (SUVs): (a) aurein 2.2, (b) aurein 2.3, and (c) aurein 2.3-COOH (solid black line: P/L = 1:15; dotted black line: P/L = 1:50; solid grey line: P/L = 1:100). The spectra indicate that the peptides adopted an α-helical conformation in the presence of DMPC SUVs. (Figure 2.2(b) and (c) are courtesy of Y. L. Pan).  In the presence of DMPC SUVs, the spectra of the three aurein peptides also showed two minima at 207 nm and 222 nm and a maximum at 190 nm, which are characteristic of the spectra of peptides adopting an α-helical conformation. The spectra indicate that the three aurein peptides adopted an α-helical conformation in the presence of DMPC SUVs. Since DMPC 48  CHAPTER 2  SUVs are a model for mammalian membranes, we also determined the structure of the three aurein peptides in 1:1 DMPC/DMPG (mol/mol) SUVs, which are a model for bacterial membranes. Figure 2.3 shows solution CD spectra of the three aurein peptides in the presence of 1:1 DMPC/DMPG (mol/mol) SUVs as a function of P/L molar ratio.  Figure 2.3.  Solution CD spectra of the aurein peptides in 1:1 DMPC/DMPG (mol/mol) small unilamellar vesicles (SUVs): (a) aurein 2.2, (b) aurein 2.3, and (c) aurein 2.3-COOH (solid black line: P/L = 1:15; dotted black line: P/L = 1:50; solid grey line: P/L = 1:100). The spectra indicate that the peptides adopted an α-helical conformation in the presence of 1:1 DMPC/DMPG (mol/mol) SUVs.  In the presence of DMPC/DMPG SUVs, the spectra of the three aurein peptides again showed two minima at 207 nm and 222 nm and a maximum at 190 nm which are characteristic  49  CHAPTER 2  of the spectra of peptides adopting an α-helical conformation. The spectra indicate that the three aurein peptides also adopted an α-helical conformation in the presence of DMPC/DMPG SUVs. From solution CD spectra, we have found that the three aurein peptides adopted a random-coil conformation in the presence of water, but adopted an α-helical conformation in the presence of TFE and phospholipids. The similar spectra in the presence of ddH2O/TFE, DMPC and 1:1 DMPC/DMPG (mol/mol) SUVs at all P/L molar ratios indicate that the structural change and the α-helical conformation of the three aurein peptides did not change with decreasing P/L molar ratios and were thus independent of the molar concentrations/types of phospholipids examined.  2.2.1.2 Using solution-state 1H MR spectroscopy to identify the structure and amino acid residues of the aurein peptides in the presence of water/TFE mixture In order to determine whether the peptides adopt a continuous α-helical structure or whether the peptides are bent, solution-state 1H NMR experiments were performed. As no change in solution CD spectra was observed in the TFE concentration range used here, solutionstate 1H NMR spectra were collected using 25% TFE-d3 (v/v).  1  H NMR spectra for the three  peptides were assigned using the TOCSY and NOESY data sets, using TOPSPIN. Figure 2.4 shows the fingerprint region of 1H NMR NOESY spectra of the three aurein peptides in the presence of 25% TFE-d3.  50  CHAPTER 2  Figure 2.4.  Fingerprint region of solution-state 1H and (c) aurein 2.3-COOH:  MR  OESY spectra of (a) aurein 2.2, (b) aurein 2.3  The spectra were acquired using a phase sensitive NOESY experiment, with excitation sculpting with gradients for water suppression (see text). All spectra were acquired at 25°C, using 64 scans and a mixing time of 150 ms. The spectra were referenced to the residual methylene protons present in TFE-d3 (3.918 ppm). In (b), arrows indicate some of the connectivities use to perform the sequential assignment. In ambiguous cases, the HN-HN region was also used to confirm i to i+1 connectivities. The 2D data set consisted of 4096 data points in t2 and 256 points in t1. (Figure 2.4(b) and (c), are courtesy of Y. L. Pan).  Of all three peptides, the spectra for aurein 2.3 were the least overlapped. The 1H NMR spectra of the three aurein peptides were nearly identical and the spectral assignments did not vary significantly. This gives us a first indication that the three aurein peptides adopted similar structures. In the fingerprint region, (i, i+4) NOE connectivities were observed, suggesting that the three aurein peptides adopted a continuous α-helical structure, not a bent structure. 51  CHAPTER 2  Furthermore, we completed the sequential assignment of amino acid residues using the fingerprint region of the TOCSY spectrum (spectrum not shown). Note that the spectra and assignments of aurein 2.3 and aurein 2.3-COOH were the work of Y. L. Pan and are, therefore, not discussed further. Figure 2.5 illustrates the sequential assignment of the fingerprint region of the NOESY spectrum of aurein 2.2 in the presence of 25% TFE-d3.  Figure 2.5.  Fingerprint region of solution-state 1H MR OESY spectrum of aurein 2.2: The spectrum was acquired using a phase sensitive NOESY experiment, with excitation sculpting with gradients for water suppression. The spectrum was acquired at 30°C, using 64 scans and a mixing time of 150 ms. The spectrum was referenced to the residual methylene protons present in TFE-d3 (3.918 ppm). In the spectrum, green lines indicate the amino acid connectivities used to perform the sequential assignment. In ambiguous cases, the HN-HN region was also used to confirm i to i+1 connectivities.  52  CHAPTER 2  In the fingerprint region, a large number sequential (i to i+1) connectivities were observed. At a low contour level, additional i to i+2 connectivities were observed in the HN-HN region, as summarized in Figure 2.6.  Figure 2.6.  MR-derived evidence indicating that the aurein peptides are α-helical: a) Hα chemical shift differences for aurein 2.2 (hashed), aurein 2.3 (solid black) and aurein 2.3-COOH (white); b) typical NOE connectivities observed for these peptides – shown here for aurein 2.3 only. Solid black bars represent unambiguous NOEs, whereas grey bars represent connectivities which are present but are ambiguous due to overlap.  Other NOE connectivities observed included: dαN(i,i+4), dαN(i,i+3), and dαβ(i,i+3), as summarized in Figure 2.6b). Together with the observed chemical shift differences of the measured Hα chemical shifts with respect to random coil values (3) (Figure 2.6a)), these data suggest that aurein 2.2 adopts a continuous α-helical structure. Similar finding were also found for aurein 2.3 and aurein 2.3-COOH.  53  CHAPTER 2  2.2.2 Determining the membrane insertion states and the membrane perturbation mechanisms of the aurein peptides Since both solution CD and 1H NMR results showed that the C-terminal groups and a single amino-acid mutation did not alter the structure of the three aurein peptides significantly, we have taken the next step to examine whether the C-terminal groups and a single amino-acid mutation would cause a difference in the membrane insertion states and the membrane perturbation mechanisms of the three aurein peptides. To determine how aurein 2.2, aurein 2.3, and aurein 2.3-COOH interact with the lipid bilayers, OCD and solid-state 31P NMR experiments were carried out. By using OCD spectroscopy, we were able to find the threshold concentrations at which the three aurein peptides insert into model membranes. Using solid-state  31  P NMR  spectroscopy allowed us to extract information on the perturbation mechanisms of the three aurein peptides in model membranes. For both methods, samples were prepared in almost identical fashion so that the data sets could be compared directly and also to verify that the samples were aligned. All experiments were conducted at 30°C (liquid crystalline phase) for DMPC and DMPC/DMPG. In addition, experiments were repeated at least twice to ensure reproducibility of the results.  2.2.2.1 Using oriented CD (OCD) spectroscopy to determine the membrane insertion states of the aurein peptides Figure 2.7 shows OCD spectra of the three aurein peptides in DMPC bilayers (left panel) and 1:1 DMPC/DMPG (mol/mol) bilayers (right panel) as a function of P/L molar ratio. The spectra were normalized and scaled at 222 nm for ease of comparison.  54  CHAPTER 2 40  40 a) Aurein 2.2-CONH2  30  20  10  195 200  205  210 215  220  225 230  235  240 245  250  Ellipticity (mdeg)  Ellipticity (mdeg)  20  0 190 -10  -20 -30 -40 40  Wavelength (nm) b) Aurein 2.3-CONH2  30  Wavelength (nm) e) Aurein 2.3-CONH2  30 20  10  195 200  205 210  215 220  225  230 235  240 245  250  Ellipticity (mdeg)  20  Ellipticity (mdeg)  0 190 195 200 205 210 215 220 225 230 235 240 245 250 -10  -30 -40 40  10 0 190 195 200 205 210 215 220 -10  225 230  235 240  245 250  -20  -20 -30  -30 Wavelength (nm) Wavelength (nm)  -40 20  -40 20 c) Aurein 2.3-COOH  15  Wavelength (nm) f) Aurein 2.3-COOH  15 10  5 0 190 195 200 205 210 215 220 225 -5  230 235 240 245 250  Ellipticity (mdeg)  10  Ellipticity (mdeg)  10  -20  0 190 -10  d) Aurein 2.2-CONH2  30  5 0 190 -5  -10  -10  -15  -15  -20  Wavelength (nm)  Figure 2.7.  -20  195 200  205  210 215  220  225 230  235  240 245  250  Wavelength (nm)  Oriented CD spectra of a) & d) aurein 2.2, b) & e) aurein 2.3 and c) & f) aurein 2.3-COOH (P/L molar ratios = 1:15 (blue), 1:30 (green), 1:40 (red), 1:80 (black), and 1:120 (grey)) in DMPC bilayers (left panel) and 1:1 DMPC/DMPG (mol/mol) bilayers (right panel): The spectra were normalized such that the intensities of all spectra at 222 nm are the same. The spectra (left panel) show that the peptides inserted into DMPC bilayer at threshold P/L molar ratios between 1:15 and 1:30 for aurein 2.2 and aurein 2.3, and 1:30 and 1:40 for aurein 2.3-COOH. The spectra (right panel) show that the peptides inserted into 1:1 DMPC/DMPG (mol/mol) bilayers at all P/L molar ratios for aurein 2.2 and aurein 2.3, and became not inserted at a P/L molar ratio between 1:80 and 1:120 for aurein 2.3-COOH. (Figure 2.7(b) and (c) are courtesy of Y. L. Pan).  At 1:120, 1:80 and 1:40 P/L molar ratios, the spectra of the three aurein peptides exhibited two minima at 207 nm and 222 nm and a maximum at 190 nm which are characteristic of the spectra of peptides adopting the surface-adsorbed or not inserted state (S-state) in DMPC bilayers (Figure 2.7, left panel). At 1:30 P/L molar ratio, the spectra of aurein 2.2 and aurein 2.3 still showed two minima; whereas, the spectrum of aurein 2.3-COOH showed a decreased minimum at 207 nm, which is characteristic of the spectra of peptides adopting the inserted state 55  CHAPTER 2  (I-state). At 1:15 P/L molar ratio, the spectra of the three aurein peptides all exhibited a similar change in the spectrum. These results indicate that both aurein 2.2 and aurein 2.3 were surfaceadsorbed at P/L molar ratios from 1:30 to 1:120, but inserted into DMPC bilayers at a P/L molar ratio between 1:15 and 1:30; whereas, aurein 2.3-COOH was surface-absorbed at P/L molar ratios from 1:40 to 1:120, and inserted into DMPC bilayers at a P/L molar ratio between 1:30 and 1:40. These threshold concentrations are similar to those observed for aurein 1.2 and citropin 1.1 (6), where a change is observed to occur between P/L molar ratios of 1:15 and 1:50. Interestingly, aurein 2.3-COOH, which is presumably the least active of the three peptides, appeared to insert into phosphatidylcholine bilayers slightly more readily than its amidated Cterminus counterpart. Since no significant difference in the membrane insertion states of the three aurein peptides was observed in DMPC bilayers, we used 1:1 DMPC/DMPG (mol/mol) model membranes to investigate whether the presence of DMPG in the bilayers would cause any difference in the membrane insertion state. At P/L molar ratios from 1:15 to 1:80, the spectra of the three aurein peptides exhibited one maximum at 195 nm and only one minimum at 222 nm (Figure 2.7, right panel).  The minimum at 207 nm was significantly reduced, which is  characteristic of the spectra of peptides adopting the inserted state (I-state). At 1:120 P/L molar ratio, the spectra of aurein 2.2 and aurein 2.3 remained the same; whereas, the spectrum of aurein 2.3-COOH showed two minima at 207 nm and 222 nm, which are characteristic of the spectra of peptides adopting the surface-adsorbed state (S-state). These results indicate that both aurein 2.2 and aurein 2.3 inserted into DMPC/DMPG bilayers at all P/L molar ratios; whereas, aurein 2.3COOH inserted into DMPC/DMPG bilayers at P/L molar ratios from 1:15 to 1:80, but became surface-adsorbed at a P/L molar ratio between 1:80 and 1:120.  56  CHAPTER 2  Comparing to the peptide insertion profiles in DMPC bilayers, the amidated peptides were able to insert into DMPC/DMPG bilayers at all concentration ranges shown. Indeed, even at very low peptide concentrations (P/L ~ 1:200), aurein 2.2 and aurein 2.3 remained in the Istate (data not shown).  For aurein 2.3-COOH, on the other hand, the threshold P/L*  concentration was between 1:80 and 1:120, indicating that comparatively high peptide concentrations were needed for insertion to take place.  This is most likely due to the  unfavourable electrostatic interactions between the negatively charged C-terminus and the negatively charge PG headgroups (11). Table 2.1 summarizes the membrane insertion states of the three aurein peptides in DMPC and DMPC/DMPG bilayers for close comparison.  Table 2.1.  Summary of the membrane insertion states of the three aurein peptides in the presence of different lipid bilayers as a function of P/L molar ratio.  Lipid bilayers P/L molar ratio 1:15 1:30 1:40 1:80 1:120  A2.2 I S S S S  DMPC A2.3 I S S S S  OH I I S S S  1:1 DMPC/DMPG (mol/mol) A2.2 A2.3 I I I I I I I I I I  OH I I I I S  *A2.2 = aurein 2.2-CONH2; A2.3 = aurein 2.3-CONH2; OH = aurein 2.3-COOH. *P/L molar ratio = peptide-to-lipid ratio (mol/mol); I = inserted state; S = surface-adsorbed state.  In summary, in the presence of DMPC bilayers, the three aurein peptides did not show high membrane insertion abilities. When DMPG was present in the bilayers, the membrane insertion abilities of the three aurein peptides increased greatly. The presence of DMPC did not cause significant changes in the membrane insertion abilities of the three aurein peptides, but the presence of DMPG did. Note that the gradual decrease of the minimum at 207 nm with the increase in P/L molar ratio (Figure 2.7, left panel) may be an indication that the membrane insertion states of the three aurein peptides were concentration dependent in DMPC bilayers. The similar spectra of aurein 2.2 and aurein 2.3 at all P/L molar ratios (Figure 2.7, right panel)  57  CHAPTER 2  may suggest that the membrane insertion states of both peptides were independent of peptide concentrations examined in 1:1 DMPC/DMPG (mol/mol) bilayers.  2.2.2.2 Using solid-state 31P MR spectroscopy to determine the membrane perturbation mechanisms of the aurein peptides Since there were differences in the membrane insertion states of the three aurein peptides in zwitterionic versus anionic lipid bilayers, it is important to establish whether there are also differences in the membrane perturbation mechanisms of the three aurein peptides for different lipid bilayers. Figure 2.8 shows solid-state  31  P NMR spectra of DMPC bilayers in the absence  and presence of the three aurein peptides as a function of P/L molar ratio.  Figure 2.8.  Solid-state 31P MR spectra of mechanically aligned DMPC bilayers with and without the three aurein peptides: The spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. The spectra were processed without any linebroadening. (Spectra of aurein 2.3 and aurein 2.3-COOH of Figure 2.8 are courtesy of Y. L. Pan).  58  CHAPTER 2 31  In the absence of the aurein peptides, a single P spectral peak was observed at 30 ppm. This indicates that the phosphorus headgroups of DMPC bilayers were well aligned. In the presence of the aurein peptides at 1:120 P/L molar ratio, similar  31  P spectral peaks were also  observed at 30 ppm. This indicates that the phosphorus headgroups of DMPC bilayers were not affected orientationally at low P/L molar ratios. At 1:15 P/L molar ratio, spectra of DMPC bilayers containing either aurein 2.2 or aurein 2.3 exhibited an additional This peak corresponds to the  31  P peak at 12 ppm.  31  P signal from perturbed headgroups, possibly as a result of the  peptide insertion into DMPC bilayers. At 1:15 P/L molar ratio, the spectrum of DMPC bilayers containing aurein 2.3-COOH remained similar to that at 1:120 P/L molar ratio, which suggests that the orientation of DMPC bilayer headgroups were not affected at high P/L molar ratios. This may indicate that aurein 2.3-COOH inserted into DMPC bilayers without effectively altering the orientation of the bilayer headgroups. The data illustrate that for the most part, the lipids remained aligned with increasing peptide concentration. The peptides appeared to disorder the headgroups slightly, as evidenced by the scaling of the 31P chemical shift anisotropy (CSA). In addition, the peptides affected the dynamics of the lipid headgroups, as shown by a decrease in T2, leading to line broadening. Both of these effects have also been previously observed for aurein 1.2 and citropin 1.1 (12). In addition, a small proportion of the lipid headgroups were significantly perturbed, as seen by the appearance of an additional  31  P NMR resonance (at 12 ppm) with increasing peptide  concentration. The presence of a peak near the isotropic position has previously been observed in  31  P NMR spectra of aurein 1.2 and has been attributed to membrane disruption (12).  Generally, the presence of a peak at or near the isotropic position has been observed in solid-state NMR studies of other antimicrobial peptides (13-16) and has been attributed to either the formation of small lipid vesicles/micelles, the formation of a different lipid phase (16), or from toroidal pore defects within the bilayers (14,17). In order to clearly identify which of these 59  CHAPTER 2  mechanisms is relevant here, additional data from experiments such as differential scanning calorimetry (to determine changes in phase) or  15  N NMR (to determine the orientation of the  peptide in the bilayer) would be needed. As the aim of this study was to determine whether the aurein peptides behave differently from aurein 1.2 and citropin 1.1 (which promote bilayer damage via a detergent-like mechanism (18), resulting in turn in membrane leakage (19)) and to determine whether this can be correlated to activity, the exact orientation in which aurein 2.2, aurein 2.3, and aurein 2.3-COOH perturb DMPC membranes will not be characterized further at this point in time. The  31  P NMR data suggest, in corroboration with the OCD results, that the  interaction of these three peptides with DMPC bilayers is identical. The abilities of the three aurein peptides to perturb the headgroups of model bacterial membranes were also monitored. Figure 2.9 shows  31  P NMR spectra of the mechanically  aligned 4:1 DMPC/DMPG (mol/mol) bilayers with and without the three aurein peptides as a function of P/L molar ratio, with the bilayer normal parallel to the magnetic field.  4:1  DMPC/DMPG (mol/mol) bilayers were used for the ease of alignment. In the absence of the aurein peptides, a single  31  P spectral peak was observed at ca. 30  ppm (Figure 2.9(a)). This indicates that the phosphorus headgroups of DMPC/DMPG bilayers were well aligned. In the presence of aurein 2.2 and aurein 2.3 at 1:15 and 1:80 P/L molar ratios, consistent single and narrower peaks (with small powder-pattern signals) were observed, where the chemical shifts were shifted upfield to 0 ppm. This indicates that the phosphorus headgroups of DMPC/DMPG mixtures may not align into bilayers (possibly micelle formation) in the presence of aurein 2.2 and aurein 2.3. In the presence of aurein 2.3-COOH at 1:15 P/L molar ratio, the spectrum displayed a similar single, narrow, upfield-shifted peak. However, the spectra at 1:80 and 1:120 P/L molar ratios showed much bigger powder-pattern signals, which suggests that the orientation of the phosphorus headgroups of DMPC/DMPG bilayers were not affected as  60  CHAPTER 2  significantly at low P/L molar ratios. This is consistent with our OCD result (Figure 2.7) where aurein 2.3-COOH did not insert readily into DMPC/DMPG bilayers at 1:120 P/L molar ratio.  (a)  1:15 Peptide:Lipid (mol/mol) 1:80 Peptide:Lipid (mol/mol) 1:120 Peptide:Lipid (mol/mol)  (b)  (c)  (d)  50  0  ppm (t1)  Figure 2.9.  50 ppm (t1)  0  50  0  ppm (t1)  Solid-state 31P MR spectra of the mechanically aligned 4:1 DMPC/DMPG (mol/mol) bilayers with and without the three aurein peptides: (a) DMPC/DMPG bilayers alone, containing (b) aurein 2.2, (c) aurein 2.3 and (d) aurein 2.3-COOH. For the 1:15 P/L molar ratio, 1.03 mg of peptide was used. For the 1:80 P/L molar ratio, 0.19 mg of peptide was used. And finally, for the 1:120 P/L molar ratio, 0.13 mg of peptide was used. The spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. The spectra were processed without any line broadening.  The presence of peptides significantly changed the physical state of DMPC/DMPG bilayers (Figure 2.9(b) ~ (d)).  This indicates that the phosphorus headgroups of 4:1  DMPC/DMPG (mol/mol) mixtures were highly curved in the presence of aurein 2.2 and aurein  61  CHAPTER 2  2.3. Indeed, a number of other solid-state NMR studies of antimicrobial peptides (14,16,20,21) have demonstrated that the presence of a peak at 0 ppm is indicative of either small lipid vesicle/micelle formation or a different lipid phase formation. However, at low concentrations of aurein 2.3-COOH, the peak at 30 ppm did not disappear completely. This suggests that partial alignment was still maintained and complete destabilization of the lipid bilayers did not occur.  2.2.2.3 Using differential scanning calorimetry (DSC) to determine lipid chain perturbation of the aurein peptides Observing phase transition changes provides information on the overall integrity of phase structures of lipid membranes in the presence of antimicrobial peptides. Since both aurein 2.2 and aurein 2.3 had a pronounced disordering effect on the headgroups of DMPC/DMPG bilayers, differential scanning calorimetry (DSC) experiments were performed to determine whether the three aurein peptides affect the acyl chains and observe how these peptides disrupt the phase structure of lipid membranes.  0.013  Cp (Cal/°C)  0.011 0.009 0.007 0.005 0.003 0.001 -0.001 5  15  25  35  45  Temperature (°C)  Figure 2.10. DSC thermograms of 1:1 DMPC/DMPG (mol/mol) liposomes in the absence (light grey solid line) and presence of aurein 2.2 (black line), aurein 2.3 (dark grey line), and aurein 2.3-COOH (grey line) at 1:15 P/L molar ratio. 62  CHAPTER 2  Figure 2.10 shows DSC thermograms of 1:1 DMPC/DMPG (mol/mol) liposomes in the absence and presence of aurein 2.2, aurein 2.3, and aurein 2.3-COOH (P/L = 1:15). In the absence of peptides, the thermogram consisted of a pretransition peak at 17.5°C and a main phase transition peak at 24.5°C and was consistent with literature findings (e.g. (22)). In the presence of the amidated peptides, the pretransition peak remained similar, whereas the main phase transition peak was broadened and was almost completely abolished. This indicates that both aurein 2.2 and aurein 2.3 disrupted the lipid membranes severely, whereby only a small lamellar-liquid crystalline coexistence regime existed. For aurein 2.3-COOH, the main transition peak was severely broadened as well, but not to the extent of the amidated peptides. The broadened main phase transition peak was also an indication of membrane curvature in the presence of peptides, consistent with the 31P NMR data presented above for 4:1 DMPC/DMPG. A similar pretransition peak for the amidated peptides indicates that the low-temperature phase domains remained intact, which suggests that these peptides did not affect the low-temperature phase domains significantly.  2.2.2.4 Using calcein release assays to determine membrane leakage induced by the aurein peptides in model membranes Given the results above, calcein release assays were used to further determine to what extent the aurein peptides cause membrane disruption. In general, the assay probes the increase in fluorescence when the fluorophore (calcein) is released as a result of membrane leakage. Gramicidin S acted as the positive control in these assays. The units of fluorescence were arbitrary and set to a range from 0 to 1000. The assay was carried out for a minimum of 300 seconds. Figure 2.11 shows the calcein release assay results of the aurein peptides in 3:1 DMPC/DMPG (mol/mol) liposomes at 1:15 P/L molar ratio. 63  CHAPTER 2  900  Fluorescence  800 700 600 500 400 0  100  200  300  Time (sec)  Figure 2.11. Calcein release profiles of 3:1 DMPC/DMPG (mol/mol) liposomes in the presence of aurein 2.2 (black solid line), aurein 2.3 (black dotted line), and aurein 2.3-COOH (grey solid line) at 1:15 peptide:lipid molar ratio. (Calcein release assay results are courtesy of M. Elliot).  Table 2.2 summarizes the relative percentage of aurein peptide-induced calcein release from DMPC/DMPG liposomes with reference to the positive control 0.1% Triton X (set to 100%) at 1:15 P/L molar ratios.  Table 2.2.  Percentage of calcein released relative to 0.1% Triton X (defined as 100%) from 3:1 DMPC/DMPG (mol/mol) liposomes, in the presence of aurein 2.2, aurein 2.3 and aurein 2.3COOH at 1:15 P/L molar ratio. (Calcein release assay results are courtesy of M. Elliot).  Peptides 3:1 DMPC/DMPG  Aurein 2.2 92  Aurein 2.3 87  Aurein 2.3-COOH 93  All three aurein peptides caused nearly 100% calcein release from DMPC/DMPG liposomes at high peptide concentrations. This indicates that the presence of the aurein peptides could induce severe membrane leakage. This suggests that the aurein peptides had significant membrane disruption effects on DMPC/DMPG liposomes. These data are consistent with the DSC data that the presence of the aurein peptides resulted in an overall lipid phase structure destabilization.  64  CHAPTER 2  2.2.3 Antimicrobial activity of the aurein peptides Given that all three aurein peptides adopted α-helical structure regardless of membrane environment (DMPC versus DMPC/DMPG) and given that the peptides interacted with the membranes in a manner which could not be directly correlated to activity, minimal inhibitory concentrations (MICs) of all three peptides against two Gram positive bacteria (S. aureus and S. epidermidis) were determined. The MICs, reported in Table 2.3, indicate that the amidated  peptides had very similar activities under conditions used here, contrary to what was reported in the literature (1). Aurein 2.2 and aurein 2.3 had similar MICs of 15 µg/ml and 25 µg/ml against the wild type S. aureus strain C622, respectively. Likewise, these two peptides had identical MICs of 8 µg/ml against S. epidermidis strain C621. Aurein 2.3-COOH, on the other hand, showed much less activity with MICs of greater than 100 µg/ml against both types of bacteria. Peptides with charged C-termini have been found to be much less active or inactive (1,23) than their amidated counterparts. Wells containing polymyxin B, culture only and broth only, were used as controls. The MICs observed for polymyxin B are reported in Table 2.3 and agree with literature findings (24).  Table 2.3.  Minimal inhibitory concentrations (MICs) of the three aurein peptides. MICs of aurein 2.2, aurein 2.3, aurein 2.3-COOH, and Polymyxin B (control) towards S. aureus and S. epidermidis given as the most frequently observed value obtained from repeat experiments. (MIC assay results are courtesy of Dr. J. D. Hale).  Peptides Aurein 2.2 Aurein 2.3 Aurein 2.3-COOH Polymyxin B  S. aureus Strain C622 15 25 ≥ 100 50  S. aureus++ 25 100    S. epidermidis Strain C621 8 8 > 128 55  S. epidermidis++ 25 100    ++  MIC results are adapted from Rozek et al., 2000 (1).  *  = data not available.  65  CHAPTER 2  2.3 Discussion Determining the structure of antimicrobial peptides and characterizing their interaction with the lipid bilayers is essential to understanding how they function and kill bacteria. By elucidating the mode of action of antibiotics, it is possible to, on the one hand, better understand how microbes develop resistance (25), and on the other, develop modified versions of these agents to mitigate this development. An approach which has received much attention recently is to search for naturally occurring antibiotic molecules derived from the plant and animal kingdoms (18,26-29), which have net positive charge and typically adopt amphiphilic structures to maximize their interactions with bacterial membranes. In order to elucidate the mode of action of a cationic antimicrobial peptide, one typically picks a highly active peptide and determines its structure by CD and/or NMR in a membrane or membrane mimetic environment, and then characterizes its interaction with model membrane bilayers such as POPC (e.g. MSI-78 and MSI-594 (30)), DPhPC (e.g. alamethicin (17,31)), DMPC (e.g. aurein 1.2 (6)), and other diacylphosphatidylcholine membranes (e.g. K2(LA)xK2 (32)). That is, the studies presented here were conducted in lipids which are good models for probing the hemolytic activity of the peptides, or lipid mixtures, such as POPC/POPG (e.g. MSI78 and MSI-594 (30)), DMPC/DMPG (e.g. PGLa (8)), which are good models for bacterial membranes.  To completely describe the peptide-lipid interactions, one needs to take into  account a range of parameters such as peptide-to-lipid ratio, membrane composition, temperature, hydration, buffer composition (18), and lipid phase (17). Once this is taken into consideration, one typically generates a model by which the peptide inserts into the lipid bilayers: namely, via the carpet mechanism (33), barrel-stave (34) or toroidal (35) pore formation, or simply a detergent-like mechanism (18,36).  66  CHAPTER 2  Here we have taken a slightly different approach in that we have studied three peptides which have essentially the same amino acid sequence, but have very different reported activities (1) with respect to different microbes. Aurein 2.2 (GLFDIVKKVVGALGSL-CONH2) was reported to be the most active of the three peptides investigated. It shares its first 9 residues in common with citropin 1.1 (though two residues have slightly different order), which is also 16 amino acid residues in length. Presumably this sequence similarity may explain why aurein 2.2 and citropin 1.1 display similar activities against Leuconostoc lactis, S. aureus, and S. epidermidis (37). Aurein 2.3, on the other hand, with a single point mutation L13→I13, is only  marginally active (1). We also investigated a modified version of aurein 2.3, with a carboxy Cterminus. Since most active members of the aurein peptide family have an amidated C-terminus, with only one aurein (aurein 5.2) with a -COOH terminus being active, and then only marginally so (37), it is expected that aurein 2.3-COOH will at best be only marginally active. The data presented here showed that despite the difference in sequence and in reported activity, all three peptides adopted continuous α-helical structures. The results from solution CD and NMR were, in fact, analogous to those reported in the literature for aurein 1.2 and citropin 1.1 (6,37). In all cases, these antimicrobial peptides were unstructured in solution and then folded in the presence of membranes or membrane mimetics. In other words, all peptides followed the first step of the Shai-Matsuzaki-Huang model (33,34,38-40) regardless of whether the peptides were in TFE, DMPC SUVs, or DMPC/DMPG SUVs. Once folded, the aurein peptides studied here then interacted with phosphatidylcholine (PC) membranes by predominantly associating with the surface. At high concentrations, the peptides realigned from a surface-bound S-state to a tilted T-state (insert at a tilt angle). The exact value of this tilt angle has yet to be determined (e.g. by labelling one residue with  15  N (6,12)), but is expected to be  similar to that of citropin 1.1 (6), given the same length of the peptides (future studies). The transition from the S to the T state occurred at P/L* between 1:15 and 1:30 for the amidated 67  CHAPTER 2  peptides and between 1:30 and 1:40 for aurein 2.3-COOH.  The slightly more favourable  insertion of aurein 2.3-COOH into PC membrane was most likely due to electrostatic interactions (11). Repulsive interactions between C-termini which were in close proximity when the peptide was surface-associated were presumably minimized when the peptides inserted. The  31  P NMR  spectra of mechanically aligned DMPC bilayers containing peptides showed that at high peptide concentrations, a proportion of the lipid headgroups were perturbed. This is again similar to what was observed for aurein 1.2 and citropin 1.1 (12). The similar OCD and NMR spectra observed for all peptides suggest that the interaction of the aurein peptides with phosphatidylcholine membranes did not depend on the sequence or the nature of the C-terminus. The interactions of the aurein peptides with bacterial model membranes consisting of 1:1 DMPC/DMPG (mol/mol), on the other hand, showed that the nature of the C-terminus modulated peptide insertion. Aurein 2.2 and aurein 2.3 displayed similar behaviour and inserted readily into PC/PG membrane, even at low peptide concentrations (i.e. P/L* < 1:200). Aurein 2.3-COOH inserted into PC/PG membranes at P/L* between 1:120 and 1:80. In other words, it inserted more readily into PC/PG membranes than in PC alone, but did not insert as easily as the amidated peptides did. This is consistent with the 31P NMR and DSC results, where aurein 2.3COOH showed less pronounced effect than the amidated counterpart on the phosphorus headgroups and acyl chains of DMPC/DMPG bilayers, respectively. The 31P NMR peak at ~ 0 ppm indicates that the three aurein peptides may perturb DMPC/DMPG bilayers via detergentlike or carpet-like mechanisms. Clearly, the charge interactions between the positively charged Lys side-chains and the negatively charged lipid headgroups drove all the aurein peptides to interact, insert, and disorder PC/PG bilayers more readily. The charge repulsions between the COOH terminus and the PG headgroups resulted in higher aurein 2.3-COOH peptide concentration needed for the T state to be achieved.  68  CHAPTER 2  Overall, the structural and membrane interaction data indicate that the single point mutation L13→I13 in going from aurein 2.2 to aurein 2.3 did not affect how these peptides folded and interacted with DMPC and DMPC/DMPG membranes. This is consistent with the new activity measurement reported here, which showed that these two peptides had similar antimicrobial properties. This indicates that small changes in the overall hydrophobicity of a peptide (i.e. leucine and isoleucine have slightly different hydrophobicity scales (41)) are not likely to have an effect on the activity of a cationic antimicrobial peptide. In addition, the data indicates that the nature of the C-terminus, specifically its charge, did not affect the structure a cationic antimicrobial peptide adopted in the presence of membrane, but rather its interaction with charged lipid headgroups. The MICs obtained for the aurein 2.3-COOH peptide clearly showed that a charged C-terminus can destroy the antibiotic activity.  2.4 Summary and conclusion In conclusion, we have demonstrated that in order to elucidate the mode of action of a family of cationic antimicrobial peptides, it may be useful to compare peptides with similar sequences but different activities in order to determine whether structure and/or peptidemembrane interactions are important for activity. We have also shown that it is important to study these peptides in model bacterial membranes (DMPC/DMPG) and not DMPC alone, as electrostatic interactions are an important driving force for peptide-lipid interactions. Finally, now that we have determined under which conditions the aurein 2.2 and aurein 2.3 peptides perturb the lipid bilayers and how that is correlated with activity, we will now determine the exact mechanism by which these peptides bring about membrane disruption (Chapter 3). We will verify whether the detergent-like mechanism proposed for aurein 1.2 and citropin 1.1 based on data obtained in DMPC (6,12) is also relevant for the aurein 2.2 and 2.3 peptides studied in other model membrane systems (Chapter 3). 69  CHAPTER 2  2.5 Materials and methods 2.5.1 Materials Fmoc-protected amino acids, Wang and Rink resin, 2-(1H-benzotriazol-1-yl)-1,1,3,3tetramethyluronium hexafluorophosphate (HBTU) were purchased from Advanced ChemTech (Louisville, KY, USA).  -hydroxybenzotiazhole (HOBt) was obtained from Novabiochem.  , -dimethylformamide (DMF), dichloromethane (DCM), acetonitrile (AcN) and potassium  nitrate  were  purchased  from  Fisher  Chemicals  (Nepean,  ON,  Canada).  , -  diisopropylethylamine (DIEA), trifluoroacetic acid (TFA), ethane dithiol (EDT), triethylsilane (TES) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Mylar plates were made by cutting Melinex Teijin films from Dupont (Wilton, Middlesbrough, UK). 1,2-dimyristoyl-snglycero-3-phosphocholine (DMPC) and 1,2-dimyristoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (DMPG) were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and obtained dissolved in chloroform.  2.5.2 Methods 2.5.2.1 Peptide synthesis Aurein 2.2 and aurein 2.3 were synthesized using Rink resin. Aurein 2.3-COOH was synthesized using Wang resin. In all cases, the synthesis was started by first pre-coupling the first residue (Fmoc-Leu) to the appropriate resin. Briefly, 0.25 mmol of the resin was pre-soaked for about an hour in 10 ml DMF. Then Fmoc-Leu was pre-activated by dissolving in 0.5 M DMF with 1 mmol HBTU, 1 mmol HOBT and 2 mmol DIEA. The Fmoc-Leu was then added to the resin. The mixture was spun overnight.  70  CHAPTER 2  Next, the peptides were synthesized using an Applied Biosystems 431A peptide synthesizer (Foster City, CA, USA) by in situ neutralization Fmoc chemistry, using the preloaded Leu-Wang resin or Leu-Rink resin, as appropriate. Side chains were protected as follows: Asp(OtBu), Lys(Boc) and Ser(tBu).  For double coupling of serine, leucine and  isoleucine, one extra step of coupling was performed for each amino acid with only DMF washes in between. After chain assembly was completed, the peptide was deprotected and cleaved simultaneously from the resin using a cleavage mixture of 81% trifluoroacetic acid (TFA), 5% ddH2O, 2.5% ethane dithiol (EDT) and 1% triethylsilane (TES) for 5 hours. TFA was then removed from the mixture by rotary evaporation. Chilled diethyl ether was used to precipitate the crude product. Finally, the resulting peptide was dissolved in water and lyophilized.  2.5.2.2 Purification The crude peptide product was purified by preparative RP-HPLC on a Waters 600 system (Waters Limited, Mississauga, ON, Canada) with 229 nm UV detection using a Phenomenex (Torrance, CA, USA) C4 preparative column (20 µm, 2.1× 25 cm) at a flow rate of 20 ml/min and linear gradient of 0 to 50% buffer B (10% ddH2O, 90% acetonitrile (AcN) containing 0.1% TFA) in buffer A (90% ddH2O, 10% AcN containing 0.1% TFA) over 80 minutes. The identity of the products was verified using electrospray ionization mass spectrometry: aurein 2.2, purity ≥ 99%, MW = 1614.95 g/mol (from MALDI-TOF); aurein 2.3, purity ≥ 99%, MW = 1614.95 g/mol (from MALDI-TOF); aurein 2.3-COOH, purity ≥ 98%, MW = 1615.95 g/mol (from MALDI-TOF).  2.5.2.3 Solution CD sample preparation Solution CD samples with constant peptide concentration of 200 µM were prepared in different compositions of water and trifluoroethanol (TFE): 100% water, 75% water with 25% 71  CHAPTER 2  TFE, 50% water with 50% TFE and 25% water with 75% TFE. Different peptide-to-lipid molar ratios of samples were also prepared: 1:15, 1:50, and 1:100. Appropriate amounts of lipid/lipids in chloroform were vacuum dried in a 25 ml round bottom flask overnight followed by addition of peptide in water. The mixture was sonicated in a water bath for at least 30 minutes (i.e. until the solution was no longer turbid) to ensure lipid vesicle formation.  For all samples,  corresponding background samples without peptides were prepared for spectral subtraction.  2.5.2.4 Solution MR sample preparation The peptides were dissolved in 65% water, 25% TFE-d3 and 10% D2O. Each peptide sample had a final concentration of 2.0 mM, with a total volume of 600 µl.  2.5.2.5 Mechanically oriented sample preparation For solid-state NMR analysis, samples were prepared for four different peptide-to-DMPC or 4:1 DMPC/DMPG (mol/mol) lipid molar ratios of 1:15 and 1:120. The amount of lipid/s (dissolved in chloroform) was kept constant at 19.18 µmol. The lipid was dried using a stream of air to remove most of the chloroform. Then, the appropriate amount of peptide was added and the mixture was redissolved in 800 µl of methanol (DMPC) or 400 µl of ddH2O (DMPC/DMPG). The mixture was deposited in 5µl (DMPC) or 10 µl (DMPC/DMPG) portions repeatedly onto 9 Mylar plates, which were placed in a Petri dish. Between depositions, most of the methanol or ddH2O evaporated before the next portion was deposited onto the plate. The plated samples were then covered and left to dry overnight on the bench. Next, the samples were placed in a 93% relative humidity chamber and were indirectly hydrated by incubating at 37ºC for 4 days (DMPC) or 5 days (DMPC/DMPG). The slides, on which the samples were still slightly humid (ca. 1 µl of water per slide), were stacked. The humidity of the samples was verified by visual inspection. The degree of alignment was verified by solid-state  31  P NMR. 72  CHAPTER 2  Consistent sample preparation was verified by preparing 2 ~ 3 samples for each lipid composition and peptide concentration. Finally, the plated samples were wrapped in a thin layer of parafilm before data acquisition. For oriented circular dichroism analysis, samples were prepared in a similar fashion as described above. The peptide amount was kept constant at 0.5 µmol and mixed with appropriate molar ratios of DMPC or 1:1 DMPC/DMPG (mol/mol) and sonicated in 2 ml of ddH2O. Each mixture was deposited in 90 µl portions with a syringe onto 3 × 1 cm and 1mm thick quartz slides, which were cleaned thoroughly with ethanol. After indirect hydration of the samples, clear layers of samples were obtained on the slides. Each sample was covered with a second slide with a spacer in between. Spacers were made by cutting 6 layers of stacked parafilm into a rectangular 3 × 1 cm frame with 2 mm width. To hold the slides in place, a thin layer of parafilm was wrapped around the edges of the slides.  2.5.2.6 Circular dichroism Circular dichroism (CD) experiments were carried out using a JASCO J-810 spectropolarimeter (Victoria, BC, Canada) at 30oC.  The spectra were obtained over a  wavelength range of 190 nm ~ 250 nm. Continuous scanning mode with a response of 1 second with 0.5 nm steps, bandwidth of 1.5 nm and a scan speed of 20 nm/min were used. The signalto-noise ratio was increased by acquiring each spectrum over an average of 3 scans. Finally, each spectrum was corrected by subtracting the background from the sample spectrum. Solution CD samples were placed in a cell (0.1 cm in length) in 200 µl portions, while oriented CD samples on quartz slides were directly placed in the sample compartment. The temperature was kept constant by means of a water bath.  73  CHAPTER 2  2.5.2.7 MR spectroscopy Solution-state NMR data was acquired on a Bruker 500 MHz instrument (Milton, ON, Canada), operating at a 1H frequency of 500.17 MHz. The parameters for the experiments were chosen to match previously reported parameters for aurein 1.2 (37), as much as possible. All spectra were collected at 25°C. Spectra were acquired using total correlation spectroscopy (TOCSY) and nuclear Overhauser enhancement spectroscopy (NOESY) experiments, in phasesensitive mode using time proportional phase incrementation (TPPI) (42) in the indirect dimension. The TOCSY experiment used the MLEV17 sequence for mixing (mixing time = 70 ms) and excitation sculpting with gradients for water suppression (43).  The 2D data set  consisted of 4096 data points in t2 and 256 points in t1. The NOESY experiment was acquired with a mixing time of 150 ms and also used excitation sculpting for water suppression. The data size for this data set was the same as for the TOCSY spectrum. Signals were averaged using 32 scans for the TOCSY and 64 scans for the NOESY experiments, respectively. The spectra were referenced to the residual methylene protons present in TFE-d3 (3.918 ppm). Spectra were processed to result in 1k x 1k points. Solid-state 31P NMR experiments on mechanically aligned DMPC or 4:1 DMPC/DMPG (mol/mol) samples were carried out on the same Bruker 500-MHz NMR spectrometer, operating at a phosphorus frequency of 202.48 MHz with proton decoupling. obtained at 30°C with a single  31  achieved using SPINAL-16 (44).  31  P ΝΜΡ spectra were  P pulse/1Η decoupling sequence where the decoupling was The 90o pulse was set to 9.75 µs (DMPC) or 12.5 µs  (DMPC/DMPG) and a 3 second recycle delay was used. Each spectrum was acquired using 2048 scans, with no linebroadening applied.  74  CHAPTER 2  2.5.2.8 Minimal Inhibitory Concentration (MIC) Determination Minimal Inhibitory Concentrations (MIC) for aurein 2.2, aurein 2.3, and aurein 2.3COOH were determined based on the previously described modified methodology (45). Briefly, 18 hour cultures of S. aureus C622 (ATCC 25923) and S. epidermidis C621 (clinical isolate generously donated by D. Speert) grown in Mueller Hinton (MH) medium (Difco, Oakville, ON, Canada) were diluted to ~ 2 × 105 CFU/ml. 90 µl of diluted culture was then dispensed into a 96 well polystyrene microtitre plate (Costar, Cambridge, MA, USA). Separately, two fold serial dilutions in sterile MH broth of the respective peptide were carried out at 10X final concentration before 10 µl of each dilution was transferred to the culture and grown for 18 hours at 37°C before being read. The MIC was recorded as the lowest concentration of peptide in which no visible growth could be observed. Controls included the peptide antibiotic Polymyxin B (Sigma, St. Louis, MO, USA). In addition, culture only and broth only wells were used.  2.5.2.1 Calcein release assays Appropriate amount of 3:1 DMPC/DMPG (mol/mol) lipid mixture were weighed and dissolved in chloroform (1 ml ~ 2 ml) in a glass vial. Chloroform was then evaporated under a stream of nitrogen and the lipid mixture was further dried under vacuum for at least 2 hours. Calcein resuspending buffer was prepared by dissolving 62.0 mg of calcein in 1 ml of 5.0 mM HEPES buffer, pH 7.5 (final concentration is equal to 100 mM). NaOH was added in small aliquots until calcein dissolved to yield a dark orange solution. Calcein release buffer was then added to the lipid mixture which underwent 5 cycles of freezing and thawing. The liposomes were extruded through two double stacked 0.1 µm membranes. The extruded calcein entrapped liposomes were then separated from free calcein in solution using a Sephadex G50 column which was rehydrated overnight in 20 mM HEPES, 150 mM NaCl, and 1.0 mM EDTA, pH 7.4. The rehydrating buffer was also used as the eluting buffer. 75  CHAPTER 2  Calcein free liposomes were prepared using the same procedure without using calcein in the resuspending buffer. 1.5 ml of the calcein free liposomes was prepared due to greater usage during the assay. The lipid mixture was resuspended in 20 mM HEPES, 150 mM NaCl, and 1.0 mM EDTA, pH 7.4 and extruded as above, without running through a Sephadex G50 column. The calcein release assay was performed by combining 2 ml of 20 mM HEPES, 150 mM NaCl, 1.0 mM EDTA, pH 7.4, 3.75 µl of calcein entrapped liposomes, and 7.5 µl of calcein free liposomes in a cuvette, with slow stirring. Fluorescence was measured using a Perkin-Elmer 640-10S spectrofluorimeter (Waltham, MA, USA) with excitation wavelength of 490 nm and emission wavelength of 520 nm. A slit width of 6 nm was normally used but could be adjusted to achieve maximum fluorescence. The baseline fluorescence was established for approximately 100 seconds. The maximum fluorescence was determined by adding 0.1 % Triton X-100 as a control. After establishing the baseline and maximum fluorescence, the peptides of interest was added to perform the assay. Gramicidin S was used as a positive control.  2.5.2.2 Differential scanning calorimetry Each aurein peptide was added at 1:15 aurein:lipid molar ratio to multilamellar vesicles of 1:1 DMPC/DMPG (25 mg/ml) resuspended in HEPES buffer (20 mM HEPES, pH 7.0, 100 mM NaCl). Samples were degassed for 5 min prior to loading the sample into a VP-DSC or multicell DSC (Calorimetry Sciences, South Provo, UT, USA), located at the University of British Columbia Centre for Biological Calorimetry.  The samples were heated over a  temperature range of 1ºC to 70ºC at a rate of 1.00ºC/minute and then cooled. The resulting data were converted to units of molar heat capacity after baseline correction by subtracting a blank buffer scan.  76  CHAPTER 2  2.6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.  Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem 2000, 267, 5330-41. Perham, M., Liao, J. and Wittung-Stafshede, P. Biochemistry 2006, 45, 7740-9. Wishart, D. S., Bigam, C. G., Holm, A., Hodges, R. S. and Sykes, B. D. J Biomol MR 1995, 5, 67-81. Chen, F. Y., Lee, M. T. and Huang, H. W. Biophys J 2002, 82, 908-14. Wu, Y., Huang, H. W. and Olah, G. A. Biophys J 1990, 57, 797-806. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids 2003, 122, 107-20. de Jongh, H. H., Goormaghtigh, E. and Killian, J. A. Biochemistry 1994, 33, 14521-8. Tremouilhac, P., Strandberg, E., Wadhwani, P. and Ulrich, A. S. J Biol Chem 2006, 281, 32089-32094. Kucerka, N., Kiselev, M. A. and Balgavy, P. Eur Biophys J 2004, 33, 328-34. Mecke, A., Lee, D. K., Ramamoorthy, A., Orr, B. G. and Banaszak Holl, M. M. Biophys J 2005, 89, 4043-50. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett 2001, 501, 146-50. Balla, M. S., Bowie, J. H. and Separovic, F. Eur Biophys J 2004, 33, 109-16. Hallock, K. J., Lee, D. K., Omnaas, J., Mosberg, H. I. and Ramamoorthy, A. Biophys J 2002, 83, 1004-13. Mani, R., Buffy, J. J., Waring, A. J., Lehrer, R. I. and Hong, M. Biochemistry 2004, 43, 13839-48. Henzler Wildman, K. A., Lee, D. K. and Ramamoorthy, A. Biochemistry 2003, 42, 654558. Hallock, K. J., Lee, D. K. and Ramamoorthy, A. Biophys J 2003, 84, 3052-60. Dave, P. C., Billington, E., Pan, Y. L. and Straus, S. K. Biophys J 2005, 89, 2434-42. Bechinger, B. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1529-1539. Ambroggio, E. E., Separovic, F., Bowie, J. H., Fidelio, G. D. and Bagatolli, L. A. Biophys J 2005, 89, 1874-81. Henzler-Wildman, K. A., Martinez, G. V., Brown, M. F. and Ramamoorthy, A. Biochemistry 2004, 43, 8459-69. Buffy, J. J., McCormick, M. J., Wi, S., Waring, A., Lehrer, R. I. and Hong, M. Biochemistry 2004, 43, 9800-12. Jung, D., Powers, J. P., Straus, S. K. and Hancock, R. E. Chem Phys Lipids 2008. Pukala, T. L., Bowie, J. H., Maselli, V. M., Musgrave, I. F. and Tyler, M. J. at Prod Rep 2006, 23, 368-93. Falla, T. J. and Hancock, R. E. Antimicrob Agents Chemother 1997, 41, 771-5. Peschel, A. and Sahl, H. G. at Rev Microbiol 2006, 4, 529-36. Hancock, R. E. W. Lancet Infect Dis 2005, 5, 209-18. Hancock, R. E. W. Lancet Infect Dis 2001, 1, 156-64. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother 1999, 43, 1317-23. Zasloff, M. Engl J Med 2002, 347, 1199-200. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry 2006, 45, 5793-9. Wu, Y., He, K., Ludtke, S. J. and Huang, H. W. Biophys J 1995, 68, 2361-9. 77  CHAPTER 2  32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.  Harzer, U. and Bechinger, B. Biochemistry 2000, 39, 13106-14. Shai, Y. Biochim Biophys Acta 1999, 1462, 55-70. Huang, H. W. Biochemistry 2000, 39, 8347-52. Matsuzaki, K., Murase, O., Fujii, N. and Miyajima, K. Biochemistry 1996, 35, 11361-8. Bechinger, B. Biochim Biophys Acta 2005, 1712, 101-8. Rozek, T., Bowie, J. H., Wallace, J. C. and Tyler, M. J. Rapid Commun Mass Spectrom 2000, 14, 2002-11. Papo, N. and Shai, Y. Cell Mol Life Sci 2005, 62, 784-90. Matsuzaki, K. Biochim Biophys Acta 1999, 1462, 1-10. Yang, L., Weiss, T. M., Lehrer, R. I. and Huang, H. W. Biophys J 2000, 79, 2002-9. Kyte, J. and Doolittle, R. F. J Mol Biol 1982, 157, 105-32. Marion, D. and Wuthrich, K. Biochem Biophys Res Commun 1983, 113, 967-74. Hwang, T. L. and Shaka, A. J. Journal of Magnetic Resonance Series A 1995, 112, 275279. Sinha, N., Grant, C. V., Wu, C. H., De Angelis, A. A., Howell, S. C. and Opella, S. J. J Magn Reson 2005, 177, 197-202. Wu, M. and Hancock, R. E. J Biol Chem 1999, 274, 29-35.  78  CHAPTER 3  CHAPTER 3: The effect of membrane composition on the antimicrobial peptides aurein 2.2 and 2.3†  3.1 Introduction To determine the mode of action of cationic antimicrobial peptides in general, it is important to establish the nature of the interaction of the peptides with model membrane bilayers. Over the years, a number of lipids have been used for such studies: 1-palmitoyl-2oleoyl-sn-glycero-3-phosphocholine (POPC) (e.g. MSI-78 and MSI-594 (2)), DMPC (e.g. aurein 1.2 (3)), 1,2-diphytanoyl-sn-glycero-3-phosphatidylcholine (DPhPC) (e.g. alamethicin (4,5)), and other diacylphosphatidylcholine membranes (e.g. K2(LA)xK2 (6)), or lipid mixtures, such as POPC/POPG (e.g. MSI-78 and MSI-594 (2)), DMPC/DMPG (e.g. PGLa (7)), etc.  To  completely describe the peptide-lipid interactions, a range of parameters such as peptide-to-lipid ratio (P/L), membrane composition, temperature, hydration, buffer composition (8), and lipid phase (5) need to be taken into account. And most importantly, the results from such studies need to be correlated with assays performed on intact bacteria, for instance the DiSC35 assay, that assesses depolarization of cytoplasmic membranes, to determine biological relevance. In order to elucidate a more comprehensive understanding of mechanisms of action for aurein 2.2 and aurein 2.3, we have examined the effect of lipid bilayer thickness/fluidity and molar PG content on the ability of the peptides to perturb membranes, using widely used lipid  †  A version of this chapter has been published as follows:  Cheng, J. T. J., Hale, J. D., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2009. The Effect of Membrane Composition on the Antimicrobial Peptides Aurein 2.2 and 2.3 from Australian Southern Bell Frogs. Biophys. J. 96(2): 552-565. 79  CHAPTER 3  mixtures, i.e. 1:1 and 3:1 mixtures of DMPC/DMPG and POPC/1-palmitoyl-2-oleoyl-sn-glycero3-[phospho-rac-(1-glycerol)] (POPG). Given that the two peptides differ in sequence only at position 13, we wanted to establish whether the modest change in going from leucine to isoleucine had any effect on peptide-lipid interactions. This was of particular interest since the original reports indicated that aurein 2.2 was four times more active than aurein 2.3, as determined by minimal inhibitory concentrations (MIC) (9). Although in our hands the MICs for these two peptides were closer to one another (1) and it is well known that some variability exists in the determination of MICs (e.g. variations by factors of 2), it is still important to understand what effect the amino acid sequence (e.g. hydrophobicity, electrostatics, etc.) has on structure and membrane interaction and how that, in turn, affects antimicrobial activity (10-14). As in the study described in Chapter 2, we have used aurein 2.3-COOH, an inactive version of aurein 2.3, as a benchmark. In this chapter, we have used solution CD spectroscopy to determine whether any structural changes of the three aurein peptides occur in different lipid vesicles. To assess how activity may be influenced by different membrane composition, we have determined the interaction between the three aurein peptides and various lipid bilayers using OCD and 31P NMR spectroscopy, and differential scanning calorimetry (DSC). We have conducted calcein release assays (using POPC/POPG model membranes) and DiSC35 assays (using S. aureus C622) to further examine whether leakage is sequence and/or lipid composition dependent. Overall, by correlating with our previous results (Chapter 2), these data should allow us to determine what the best bacterial model membranes are to study this family of cationic antimicrobial peptides and to better understand how sequence modulates function. This chapter will outline the important details of the studies on the effect of the membrane composition on the three aurein peptides. This chapter will be divided into three  80  CHAPTER 3  sections. Section 3.2 gives the structural analyses and deduces the peptide-membrane interaction mechanisms in different model bacterial membranes. Section 3.3 discusses the findings and how they correlate with the structure-function relationship of the aurein peptides.  Section 3.4  summarizes and presents the conclusions from this study.  3.2 Results 3.2.1 Secondary structure of the aurein peptides In the previous chapter, we showed that the aurein 2.2, aurein 2.3, and aurein 2.3-COOH peptides adopted an α-helical structure in the presence of TFE, DMPC and 1:1 DMPC/DMPG (mol/mol) SUVs (1). In order to verify that the peptides remain structured or whether the peptides undergo different conformational changes in various lipid environments probed here, solution CD experiments were performed using different lipid mixtures and peptide-to-lipid (P/L) ratios. Figure 3.1 shows solution CD spectra of the three aurein peptides in 3:1 DMPC/DMPG, 1:1 and 3:1 POPC/POPG (mol/mol) small unilamellar vesicles (SUVs) as a function of P/L molar ratio. All spectra consisted predominantly of a maximum at 190 nm and two minima at 210 nm and 222 nm, which are spectral characteristic of peptides adopting α-helical structure. All spectra were fitted using three different programs (CDSSTR (15), CONTINLL (16), and SELCON3 (17-19)), using either the full data set or with half the data set (using only points at every 1 nm in the range between 190 and 260 nm) to estimate error. Table 3.1 shows % α-helix of the three aurein peptides in the presence of ddH2O/TFE and different SUVs as a function of P/L molar ratio.  81  CHAPTER 3 60  60  60 ( d ) A u r e in 2 .2 - C O N H  2  20 0 -20 -40  40 20 0 -20 -40  190  210  230  250  W a v e le n g t h (n m )  230  ( e ) A u r e in 2 .3 - C O N H  -20 -40  190  20 0 -20  W a v e le n g t h ( n m )  210  230  -40  W a v e le n g t h (n m )  250  210  230  250  W a v e le n g t h (n m ) 60 ( i) A u r e in 2 .3 - C O O H  40 20 0 -20 -40  230  -20  190  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  -20  210  0  (f ) A u r e in 2 .3 -C O O H  0  190  20  250  60 ( c ) A u r e in 2 .3 - C O O H  20  2  40  W a v e le n g t h (n m )  40  250  -40 190  60  230  (h ) A u r e in 2 .3 -C O N H 2  40  250  210  W a v e le n g t h (n m )  -40 230  -20  250  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  0  210  0  60  2  20  190  20  W a v e le n g t h (n m )  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  ( b ) A u r e in 2 .3 - C O N H  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  210  60  40  2  40  -40 190  60  Figure 3.1.  (g ) A u r e in 2 .2 -C O N H 2  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  40  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  M e a n R e s id u e E llip t ic it y ( 1 0 - 3 d e g ⋅c m 2 / d m o l )  ( a ) A u r e in 2 .2 - C O N H  40 20 0 -20 -40  190  210  230  W a v e le n g t h (n m )  250  190  210  230  250  W a v e le n g t h ( n m )  Solution CD spectra of the aurein peptides in 3:1 DMPC/DMPG (left panel), 1:1 POPC/POPG (centre panel), and 3:1 POPC/POPG (right panel) (mol/mol) small unilamellar vesicles (SUVs): (Top row) aurein 2.2; (middle row) aurein 2.3; and (bottom row) aurein 2.3-COOH (solid black line: P/L = 1:15; dash black line: P/L = 1:50: dotted black line: P/L = 1:100). The spectra indicate that the aurein peptides adopt an α-helical conformation in the presence of POPC/POPG SUVs.  The results demonstrated that all three aurein peptides adopted close to 100% α-helical conformation at high P/L molar ratios (as also found in 50% ddH2O/TFE, Table 3.1). As previously noted (1), similar intensities were observed for all peptide-to-lipid molar ratios studied (P/L = 1:15, 1:50, and 1:100) for 3:1 DMPC/DMPG, indicating that maximum binding of the peptide to the lipid vesicles occurred. Saturation would be observed with a combination of 82  CHAPTER 3  signals from both α-helical and random coil structures (3). For POPC/POPG (1:1 or 3:1), on the other hand, helical content increased with increased peptide concentration, indicating that high concentrations were needed to achieve maximum folding. Overall the data showed that the structures of the three aurein peptides were thus dependent of the molar concentrations/types of phospholipids examined.  % α-helix of the three aurein peptides in the presence of ddH2O/TFE and different SUVs at different P/L molar ratios.  Table 3.1.  P/L molar ratio 1:15 1:50 1:100  1:1 ddH2O/TFE (vol/vol) A2.2 A2.3 OH  3:1 DMPC/DMPG (mol/mol) A2.2 A2.3 OH  1:1 POPC/POPG (mol/mol) A2.2 A2.3 OH  3:1 POPC/POPG (mol/mol) A2.2 A2.3 OH  99.2a  98.0 97.5 98.8  98.6 85.2 74.6  100.0 79.1 73.1  92.8a  99.2a  99.5 99.3 99.3  99.7 99.8 98.8  97.5 98.8 70.3  97.4 97.5 75.1b  97.5 97.5 70.8b  98.9 98.3 65.0  *A2.2 = aurein 2.2-CONH2; A2.3 = aurein 2.3-CONH2; OH = aurein 2.3-COOH. a  Values independent of P/L molar ratio.  b  The error for these is estimated to be ± 20, whereas for the other values it is ± 4. The higher error is due to the fact that these CD spectra were noisier than the others.  3.2.2 Membrane insertion states of the aurein peptides Understanding how the aurein peptides interact with different model membranes is crucial to elucidating the effect of the membrane composition on the extent of peptide insertion into and peptide disordering on the lipid bilayers.  OCD experiments were conducted to  investigate the membrane insertion states of the three aurein peptides in different lipid bilayers. Solid-state  31  P NMR experiments were performed to determine the membrane perturbation  mechanism of the three aurein peptides in different model membranes. For both OCD (Section 3.2.2) and solid-state 31P NMR (Section 3.2.3), samples were prepared in similar fashion so that the data sets could be directly compared and also to verify that the samples were aligned. All experiments were conducted at 30°C (liquid crystalline phase) for consistent comparison with our previous studies (Chapter 2). In addition, experiments were repeated at least twice to ensure 83  CHAPTER 3  reproducibility of the results. Figure 3.2 shows OCD spectra of the three aurein peptides in 3:1 DMPC/DMPG, 1:1 and 3:1 POPC/POPG (mol/mol) bilayers as a function of P/L molar ratio.  30  (a ) A u r e in 2 .2 -C O N H  20  30 2  -20 -30  0  -20 -30 -40  -50  -50  30  (b ) A u r e in 2 .3 -C O N H  20  30 2  20  -50 190 200 210 220 230 240 250  ( e ) A u r e in 2 .3 - C O N H  W a v e le n g t h ( n m ) 30 2  20  -20 -30  0 -10 -20 -30  -40  -40  -50  -50 190 200 210 220 230 240 250  (c ) A u r e in 2 .3 -C O O H  20  190 200 210 220 230 240 250  W a v e le n g t h ( n m ) 30  ( f ) A u r e in 2 .3 - C O O H  20  -30  0 -10 -20 -30  -40  -40  -50  -50 190 200 210 220 230 240 250  ( i) A u r e in 2 .3 - C O O H  10  E llip t ic it y ( m d e g )  -20  -30  -50  10  E llip t ic it y ( m d e g )  10 0  -20  -40  30  -10  0  W a v e le n g t h ( n m )  30  2  -10  190 200 210 220 230 240 250  W a v e le n g t h (n m )  ( h ) A u r e in 2 .3 - C O N H  10  E llip t ic it y ( m d e g )  E llip t ic it y ( m d e g )  -10  Figure 3.2.  -30 -40  10  0  W a v e le n g t h (n m )  -20  W a v e le n g t h ( n m )  10  20  0 -10  190 200 210 220 230 240 250  W a v e le n g t h (n m )  2  10  -10  -40  ( g ) A u r e in 2 .2 - C O N H  20  E llip t ic it y ( m d e g )  0 -10  190 200 210 220 230 240 250  E llip t ic it y ( m d e g )  2  10  E llip t ic it y ( m d e g )  E llip t ic it y ( m d e g )  10  E llip t ic it y ( m d e g )  30  (d ) A u r e in 2 .2 - C O N H  20  0 -10 -20 -30 -40 -50  190 200 210 220 230 240 250  W a v e le n g t h ( n m )  190 200 210 220 230 240 250  W a v e le n g t h ( n m )  Oriented CD spectra of the aurein peptides in 3:1 DMPC/DMPG (left panel), 1:1 POPC/POPG (centre panel), and 3:1 POPC/POPG (right panel) (mol/mol) bilayers: (Top row) aurein 2.2; (middle row) aurein 2.3; and (bottom row) aurein 2.3-COOH [P/L molar ratios = 1:15 (blue), 1:30 (green), 1:40 (red), 1:80 (black), and 1:120 (grey)]. The spectra were normalized such that the intensities of all spectra at 222 nm are the same. The spectra show that the peptides insert into 1:1 POPC/POPG (mol/mol) bilayers at threshold P/L molar ratios between 1:80 and 1:120 for aurein 2.2, between 1:40 and 1:30 for aurein 2.3, and 1:15 and 1:30 for aurein 2.3-COOH.  The spectra were scaled so that the minimum at 222 nm had the same intensity. In the presence of 1:1 POPC/POPG bilayers, each peptide inserted differently: the threshold P/L molar ratio was between 1:80 and 1:120 for aurein 2.2, between 1:40 and 1:30 for aurein 2.3, and 1:15 84  CHAPTER 3  and 1:30 for aurein 2.3-COOH. The spectra in Figure 3.2(a) ~ (c) showed that the three aurein peptides inserted (inserted, I-state or tilt, T-state) into 3:1 DMPC/DMPG bilayer at threshold P/L molar ratios between 1:40 and 1:80, and became surface-adsorbed (S-state) at P/L molar ratios greater than 1:80. This is in contrast to previously reported OCD data in 1:1 DMPC/DMPG (mol/mol) bilayers (1), where all three peptides were already in the inserted or tilted state at P/L molar ratios of 1:120.  Finally, the spectra in Figure 3.2(g) ~ (i) show that the three aurein  peptides inserted into 3:1 POPC/POPG bilayer at threshold P/L molar ratios between 1:40 and 1:80. Table 3.2 summarizes the membrane insertion states of the three aurein peptides in the presence of different lipid bilayers as a function of P/L molar ratio.  Table 3.2.  Membrane insertion states of the three aurein peptides in the presence of different lipid bilayers at different P/L molar ratios.  Lipid bilayers P/L molar ratio 1:15 1:30 1:40 1:80 1:120  3:1 DMPC/DMPG A2.2 A2.3 OH I I I I I I I I I S S S S S S  1:1 POPC/POPG A2.2 A2.3 OH I I I I I S I I/S S I I/S S S S S  3:1 POPC/POPG A2.2 A2.3 OH I I I I I I I I I S S S S S S  *A2.2 = aurein 2.2-CONH2; A2.3 = aurein 2.3-CONH2; OH = aurein 2.3-COOH. *P/L = peptide-to-lipid; I = inserted state; S = surface-adsorbed state.  The data illustrated that both molar PG content and bilayer thickness/fluidity played a role in peptide insertion profiles. A decrease in molar PG content and increase in bilayer thickness/fluidity progressively reduced the insertion ability of the amidated peptides. High molar PG content and increased bilayer thickness/fluidity resulted in an inability of aurein 2.3COOH to insert into 1:1 POPC/POPG bilayers except at high peptide concentrations. Since the DMPC/DMPG and POPC/POPG bilayer hydrophobic thickness are ca. 26.5 Å (20,21) and ca. 39 Å (22) in the liquid crystalline phase, respectively, the three aurein peptides would not have sufficient peptide length (hydrophobic length = ca. 24 Å) to span the lipid bilayer entirely, particularly in the long-chained POPC/POPG bilayers. Unfavourable electrostatic interactions 85  CHAPTER 3  between the negatively charged C-terminus and the negatively charged PG headgroups were likely the explanation for the need of high aurein 2.3-COOH peptide concentrations to permit insertion into the lipid bilayers at increased molar PG content (10).  3.2.3 Perturbation of lipid headgroups by the aurein peptides The  31  P NMR experiments were conducted to investigate whether the insertion of the  peptides was accompanied by a perturbation of the lipid headgroups and whether this membrane disruption occurred via a barrel-stave, carpet, or toroidal pore (23-27); a micellar aggregate channel (28,29); or a detergent-like mechanism (8).  31  P NMR spectra were recorded for all  peptides in 4:1 POPC/POPG (mol/mol) bilayers, to compare with our previous results in 4:1 DMPC/DMPG (mol/mol) bilayers (Chapter 2) and our current OCD results.  These lipid  composition were chosen because in both cases, all three peptides showed similar concentration dependent insertion profiles (as in the 3:1 cases presented above). In contrast to our observations in DMPC/DMPG bilayers (Chapter 2), the aurein peptides affected the thicker/more fluidic 4:1 POPC/POPG (mol/mol) bilayers differently. Figure 3.3 shows 31P NMR spectra of POPC/POPG bilayers in the absence and presence of the three aurein peptides as a function of P/L molar ratio. In the absence of peptides, the spectra consisted primarily of a single resonance at 30 ppm, which again indicated that the lipid bilayers were aligned with their normal parallel to the magnetic field (Figure 3.3(a)). The minor peak at -10 ppm represented the signal from a small percentage of unaligned bilayer headgroups and has also been observed in other studies (e.g. (30,31)). In the presence of the amidated aurein peptides, the spectra showed an increased contribution from unaligned  31  P headgroups and a broadened peak  at 30 ppm. In addition, a powder-pattern signal was also observed in the range of -10 to 30 ppm, indicative of random headgroup orientations (Figure 3.3(b) and (c)).  86  CHAPTER 3  (a)  1:15 Peptide:Lipid (mol/mol) 1:80 Peptide:Lipid (mol/mol) 1:120 Peptide:Lipid (mol/mol)  (b)  (c)  (d) 50 ppm (t1)  Figure 3.3.  0  50 ppm (t1)  0  50  0  ppm (t1)  Solid-state 31P MR spectra of the mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers in the absence and presence of the three aurein peptides: POPC/POPG bilayers (a) alone or containing (b) aurein 2.2, (c) aurein 2.3 and (d) aurein 2.3-COOH. For the 1:15 P/L molar ratio, 1.03 mg of peptide was used. For the 1:80 P/L molar ratio, 0.19 mg of peptide was used. And finally, for the 1:120 P/L molar ratio, 0.13 mg of peptide was used. The spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. The spectra were processed without any line broadening.  These changes in the spectra occurred for all P/L molar ratios and both aurein 2.2 and aurein 2.3, suggesting no obvious dependence on peptide sequence and concentration. When aurein 2.3-COOH was added at high concentrations, the spectrum looked similar to the amidated counterparts. At low peptide concentrations, however, the two individual peaks at -10 ppm and 30 ppm disappeared and broadened powder-pattern signals were observed. This may indicate  87  CHAPTER 3  that aurein 2.3-COOH perturbed the phosphorus headgroups in a slightly different way from the amidated peptides at low peptide concentrations (Figure 3.3(d)). The underlying powder pattern indicates that the aurein peptides may disorder the bilayer headgroups by mechanisms similar to toroidal pore (23-27) or toroidal pore/liposome formation.  The extent of the membrane  perturbation was found to be neither concentration-dependent nor peptide-specific (at least for the amidated versions).  3.2.4 Perturbation of lipid chains by the aurein peptides Observing phase transition changes provides information on the overall phase structural integrity of lipid membranes in the presence of antimicrobial peptides. Since both aurein 2.2 and aurein 2.3 disordered the lipid chains of DMPC/DMPG extensively (Chapter 2), DSC experiments were performed to examine whether aurein 2.2 and aurein 2.3 affect longer lipid acyl chains and observe how these peptides disrupt the phase structure of thicker/more fluidic lipid membranes. Figure 3.4 shows the thermogram of 1:1 POPC/POPG (mol/mol) liposomes in the absence and presence of aurein 2.2, aurein 2.3, and aurein 2.3-COOH at 1:15 P/L molar ratio. In the absence of peptides, the thermogram exhibited a single main phase transition peak at -2°C. In the presence of the aurein peptides, however, the transition temperature did not change significantly. For all peptides, the transition peak intensity was slightly different when the peptides were present.  This indicated that both amidated aurein peptides did not have  pronounced effects on the chain-melting event and most likely perturbed the headgroups more than the acyl chains. Taken together with the 31P NMR data, this would indicate that the aurein peptides may induce toroidal pore formation (either a distorted toroidal or localized membrane aggregate model) or toroidal pore/liposome formation in POPC/POPG bilayers.  88  CHAPTER 3  140 120  Cp (Cal/°C)  100 80 60 40 20 0 -20  -20  -10  0  10  20  30  Temperature (°C)  Figure 3.4.  DSC thermograms of 1:1 POPC/POPG (mol/mol) liposomes in the absence (solid black line) and presence of aurein 2.2 (dotted black line), aurein 2.3 (solid grey line), and aurein 2.3-COOH (dotted grey line) at 1:15 P/L molar ratio. ote that the exotherm at 0°°C ~ 5°°C is due to water melting transition.  3.2.5 Membrane leakage induced by the aurein peptides in model membranes Given the results above, calcein release assays were used to further assess to what extent the aurein peptides cause membrane disruption. In general, the assay probes the increase in fluorescence when the fluorophore (calcein) is released as a result of membrane leakage. Gramicidin S, a cyclic antimicrobial peptide, acted as the positive control in these assays. The units of fluorescence were arbitrary and set to a range from 0 to 1000. The assay was carried out for a minimum of 300 seconds. Table 3.3 summarizes the relative percentage of aurein peptide-induced calcein released with reference to 0.1% Triton X (set to 100%) from different model membranes as a function of P/L molar ratio.  89  CHAPTER 3 Table 3.3.  Percentage of calcein released relative to 0.1% Triton X (defined as 100%) from 1:1 and 3:1 POPC/POPG (mol/mol) liposomes, in the presence of aurein 2.2, aurein 2.3, and aurein 2.3COOH. The errors associated with the measurements have been determined for the P/L molar ratio of 1:15 in 1:1 and 3:1 POPC/POPG from repeat measurements and are given in the table. (Calcein release assay results are courtesy of M. Elliot).  Peptides P/L molar ratio 1:1 POPC/POPG 3:1 POPC/POPG  Aurein 2.2 1:15 1:80 1:120 27 ± 2% 4% 2% 36 ± 2% 4% 1%  Aurein 2.3 1:15 1:80 1:120 18 ± 2% 3% 3% 12 ± 3% 6% 1%  Aurein 2.3-COOH 1:15 1:80 1:120 21 ± 5% 1% 2% 16 ± 5% 1% 2%  *Errors associated with measurement were determined for P/L molar ratio of 1:15 in 1:1 and 3:1 POPC/POPG from repeated measurement as given here.  The three aurein peptides caused significant calcein release at 1:15 P/L molar ratios, and only minimal (≤ 6%) release at lower peptide concentrations. All three aurein peptides caused nearly 100% calcein release from 3:1 DMPC/DMPG (mol/mol) liposomes at high peptide concentrations (Chapter 2). In POPC/POPG liposomes, on the other hand, the aurein peptides were less effective at causing membrane leakage. Indeed, at 1:15 P/L molar ratio, aurein 2.2 induced the highest calcein release (27% and 36% in 1:1 and 3:1 POPC/POPG, respectively), while aurein 2.3 caused a slightly lower calcein release than aurein 2.3-COOH from POPC/POPG liposomes (Table 3.3). It is interesting to note that in 3:1 POPC/POPG, aurein 2.2 was a factor of 2 ~ 3 times more effective at perturbing liposomes than aurein 2.3 and aurein 2.3COOH, which were found equally effective. This suggests that changing the nature of the Cterminus had little effect on the capacity of these peptides to induce membrane leakage. At 1:80 and 1:120 P/L molar ratios, all aurein peptides did not significantly disrupt 1:1 and 3:1 POPC/POPG liposomes, consistent with the OCD results presented above. Figure 3.5(a) and (b) show the calcein release profiles of the three aurein peptides in 1:1 and 3:1 POPC/POPG (mol/mol) liposomes at 1:15 P/L molar ratio, respectively. When added to POPC/POPG membranes, aurein 2.2 induced higher calcein release than the two aurein 2.3related peptides. Surprisingly, aurein 2.3-COOH induced more leakage than aurein 2.3 at the steady-state (t > 200s), although aurein 2.3 perturbed membranes more initially. Figure 3.5(c) and (d) show the concentration-dependence of aurein-induced calcein release for 1:1 and 3:1 90  CHAPTER 3  POPC/POPG liposomes, respectively.  The percentage of calcein released increased as the  peptide concentration increased for all aurein peptides.  At low peptide concentrations, no  difference was observed between the peptides within the margin of error (± 2 ~ 5%) of the experiment.  (a)  (c)  200  40  % Calcein Release  35  Fluorescence  150 100 50 0 0  100  200  30 25 20 15 10 5 0 0.00  300  Time (sec)  (b)  0.05  0.10  Peptide/Lipid Molar Ratio  (d)  200  40  % Calcein Release  35  Fluorescence  150 100 50 0 0  100  200  Time (sec)  Figure 3.5.  300  30 25 20 15 10 5 0 0.00  0.05  0.10  Peptide/Lipid Molar Ratio  Calcein release assay results of 1:1 and 3:1 POPC/POPG (mol/mol) small unilamellar vesicles (SUVs) in the presence of aurein peptides at 1:15 P/L molar ratio. Calcein release spectra of (a) 1:1 and (b) 3:1 POPC/POPG (mol/mol) small unilamellar vesicles (SUVs) in the presence of aurein 2.2 (solid black line), aurein 2.3 (dotted black line), and aurein 2.3COOH (solid grey line) at 1:15 P/L molar ratio. Percentage of calcein released from (c) 1:1 and (d) 3:1 POPC/POPG (mol/mol) SUVs in the presence of aurein 2.2 (solid black line), aurein 2.3 (dotted black line), and aurein 2.3-COOH (solid grey line) as a function of P/L molar ratio, with respect to 0.1 % Triton X (set to 100%). The fluorescence scale was arbitrary and set from 0 (minimum) to 1000 (maximum), with respect to 0.1% Triton X (set to 100%). Note that the spikes in calcein release spectra were artefacts due to pipet tips inserting into the cuvette. (Calcein release assay results are courtesy of M. Elliot).  91  CHAPTER 3  3.2.6 Aurein-induced membrane leakage in S. aureus Antimicrobial peptides can kill bacteria through many different mechanisms of which cytoplasmic membrane depolarization is a common target. To observe what effects the aurein peptides had on the cytoplasmic membrane of S. aureus strain C622, we carried out membrane depolarization experiments using the membrane sensitive dye DiSC35.  All peptides were  compared to the membrane acting cyclic peptide gramicidin S (GMS). Figure 3.6 shows DiSC35 release profiles of the aurein peptides in S. aureus at 1X and 5X MIC.  400 350 300  Fluorescence  GMS  250  Aurein 2.2-CONH2 1X MIC Aurein 2.2-CONH2 5X MIC  200  Aurein 2.3-CONH2 1X MIC Aurein 2.3-CONH2 5X MIC  150  Aurein 2.3-COOH 1X MIC  100 50 0 -50  0  50  100  150  200  250  300  350  400  Time (sec)  Figure 3.6.  Membrane depolarization of S. aureus C622 induced by aurein 2.2 at 1X, 5X MIC; aurein 2.3 at 1X, 5X MIC; and finally, aurein 2.3-COOH at 1X MIC, where the MIC values used were as reported in (1). Gramicidin S (1X MIC) was used as a control. Results are a representative of 2 ~ 3 experiments. The arrow represents the time point where peptide was added. (DiSC35 assay results are courtesy of Dr. J. D. Hale).  The DiSC35 results showed that at 1X and 5X MIC, aurein 2.2 demonstrated greater depolarization of the membrane than any of the other peptides including GMS. Aurein 2.3 demonstrated similar depolarization to GMS at 1X MIC, whereas aurein 2.3-COOH was much  92  CHAPTER 3  less efficient than GMS. Increasing the level of aurein peptide added to 5X MIC showed an increased depolarization effect for both amidated peptides.  3.3 Discussion Understanding how membrane composition modulates peptide-lipid interactions can be an important step in unlocking the mechanism of action of a given antimicrobial peptide. In the studies discussed in this chapter, we further examined how two antimicrobial peptides from the Australian Southern Bell frog Litoria aurea, namely aurein 2.2 and aurein 2.3, interact with a range of model membranes and the membrane from S. aureus C622. As a negative control, we also investigated the peptide-lipid interaction of a non-active analogue, aurein 2.3-COOH. Bacterial membrane composition varies from bacteria to bacteria and also as a response to changing environment (32,33) and exposure to antibiotics (34-36).  Phosphatidyl-  ethanolamine (PE) is known as a major membrane component (up to 80%) in Gram negative bacteria (36-38), while PG is identified as a major membrane component (up to 58%) in Gram positive bacteria (39-42). Incorporating PG in model membranes is thus necessary to replicate the lipid bilayers of Gram positive bacteria. Previous studies were commonly conducted in model membranes having 33% or 50% molar PG content (10,43-48). DMPC/DMPG bilayers (3,21,22,44,46,48) and POPC/POPG bilayers (3,10,43,45,49) are the most commonly used bacterial membrane models. Other models include PG only or mixtures of DOPC/DOPG (50,51) or DPPC/DPPG (22,52,53) to account for different bilayer thickness and fluidity. Our study has made use of the most widely used models: DMPC/DMPG (ca. 26 Å (20)) and POPC/POPG (ca. 39 Å (22)) bilayers to determine whether the bilayer thickness/fluidity or the PG content has an effect on aurein peptide-lipid interactions.  93  CHAPTER 3  It is widely accepted that cationic antimicrobial peptides only adopt secondary structure in the presence of membranes (23,54,55) and that the adoption of secondary structure is a key first step to membrane interaction. Our solution CD results showed that the structures of the three aurein peptides were similar in the presence of 1:1 DMPC/DMPG (1) and, as demonstrated here, in 3:1 DMPC/DMPG. Maculatin 1.1 and citropin 1.1, α–helical antimicrobial peptides from tree frogs, have also been shown to retain their structures in DMPC, DMPG, POPC and POPG membranes (3,56). We have demonstrated here, however, that helical content does depend on membrane bilayer thickness/fluidity, with differences in percent α-helical structure being observed between DMPC/DMPG and POPC/POPG. This demonstrates that secondary structure is independent of PG content, but depends on bilayer thickness/fluidity. In other words, membrane bilayer thickness/fluidity appears to have an effect on the ability of a peptide to bind the membrane and, consequently, to have an effect on the ability of a lipid bilayer to induce peptide secondary structure. Many studies have demonstrated how electrostatics and bilayer thickness play an important role in peptide-lipid interactions (e.g. (10-12,43,48,57)). In the latter case, it has been generally shown that longer antimicrobial peptides, such as caerin 1.1 and maculatin 1.1 (3), insert more readily into lipid bilayers and experience less impact from increased bilayer thickness. On the other hand, shorter antimicrobial peptides insert less readily and compensate by inserting into only one leaflet or tilting to minimize hydrophobic mismatch. In addition, it has recently been shown that the tilt of helices, which increases with decreasing bilayer thickness and has an impact on the ability of a peptide to oligomerize (58), is not necessarily accompanied by a change in the phase, order, or structure of the lipid bilayers (59). In this study, a clear difference in the insertion profile of the aurein peptides was observed as a function of bilayer thickness/fluidity. In the thinner DMPC/DMPG bilayers, much lower concentrations of the  94  CHAPTER 3  amidated peptides were required for insertion (P/L ~ 1:200 for 1:1 DMPC/DMPG) (Chapter 2). In the thicker POPC/POPG bilayers, higher aurein peptide concentrations were required to observe a change from the surface-adsorbed state to the inserted or tilted state. Viewed in a different way, one might argue that the thinner DMPC/DMPG membranes might result in a larger portion of the peptides interacting with the entire lipid molecules (headgroups and chains), resulting in more significant membrane disruption, as shown in a possible model depicted in Figure 3.7(a) and as evidenced by the 31P NMR, DSC, and calcein release assay data.  Figure 3.7.  Possible models for the perturbation of (a) DMPC/DMPG and (b) POPC/POPG by the aurein peptides. To keep this model simple, some finer points, such as the distinction in the ability of aurein 2.2 to induce membrane depolarization in S. aureus C622 as compared to aurein 2.3 and aurein 2.3-COOH, as well as the possibility of toroidal pore/liposome formation, are obviously not taken into account. The arrows represent leakage.  In the case of thicker POPC/POPG membranes (Figure 3.7(b)), the aurein peptides may be forced to interact comparatively more with the surface (as area/molecule is more), resulting in a toroidal pore or related, such as distorted toroidal, toroidal pore/liposome, or localized membrane aggregate mechanism, where the lipid headgroups were perturbed (as seen by the 31P NMR data) but the lipid acyl chains were probably not disordered extensively (as seen by the DSC data). The order of acyl chains in the absence and presence of the aurein peptides can be 95  CHAPTER 3 2  further examined by H NMR. Since in this case, many more peptides would be required to line the defects in the membrane bilayers (23-27), the amount of calcein released for a given P/L molar ratio would be expected to be lower than in the case where micellarization occurs. In terms of electrostatics, the increase in the amount of PG in DMPC/DMPG bilayers clearly has the effect that fewer peptides are required to promote insertion into the membrane. Indeed, the threshold P/L molar ratio in 1:1 DMPC/DMPG was around 1:200 (data not shown), whereas in 3:1 DMPC/DMPG it was between 1:40 and 1:80. Presumably the increased negative charge in a 1:1 DMPC/DMPG favoured the binding of the aurein peptides to the surface, enabling them to insert and perturb the membrane bilayers more readily.  In the case of  POPC/POPG, the increased negative charge in the 1:1 versus the 3:1 case resulted in a more complex effect and insertion profile.  Here the increased PG content had the effect of  differentiating the behaviour of aurein 2.2 from that of aurein 2.3 and also from that of aurein 2.3-COOH. Presumably in the POPC/POPG case, where distorted toroidal pores or localized membrane aggregates might be formed and peptides pack (aggregate?) to line these defects, not only would peptide-lipid electrostatic interactions be important but also peptide-peptide electrostatic interactions. A number of studies have indeed shown that Leu might be more hydrophobic in peptides than Ile (60-63), suggesting that aurein 2.2 would more likely aggregate. To verify whether aurein 2.2 is more likely to form oligomers in solution, the retention times of aurein 2.2 and aurein 2.3 on a reversed-phase HPLC were determined. Pure peptides were injected and eluted under a gradient of water/acetonitrile, as used to purify the crude peptides (see Materials and Methods, Section 3.5). The retention times for both peptides were nearly identical indicating that both peptides behaved the same in solution (data not shown). To further probe whether aurein 2.2 has a higher propensity to oligomerize in membranes, NOESY spectra were acquired for aurein 2.2 in DPC micelles and for aurein 2.3 under the same conditions.  96  CHAPTER 3  Preliminary data suggest that aurein 2.2 formed dimers in DPC micelles, due to the larger number of long range NOE cross-peaks found, as was reported for MSI-78 (2). A comparison of the calcein release and the DiSC35 data clearly shows the merits and limitations of using model membranes. By using the same model membranes in the calcein release assay as in the other experiments (OCD, 31P NMR, and DSC), a comprehensive model of the interaction of the aurein peptides with well-defined DMPC/DMPG and POPC/POPG membranes can be elucidated. For instance, the calcein release data clearly showed that all three aurein peptides were equally efficient at perturbing 3:1 DMPC/DMPG liposomes, a finding which was supported by the OCD, 31P NMR, and DSC data (Chapter 2). The calcein release data also demonstrated that aurein 2.2 was better at inducing leakage at high concentrations as compared to the other two peptides, a finding which was supported by the DiSC35 data. On the other hand, the calcein release data suggested that the inactive aurein 2.3-COOH was more efficient at perturbing both 1:1 and 3:1 POPC/POPG liposomes than the amidated version, a finding which was clearly not supported by much of the other data, and especially by the biological data which is consistent with the inactivity of this peptide.  3.4 Summary and conclusion Overall the results presented here suggest that the aurein peptides interact with membranes in a manner that is dependent on lipid composition and is also dependent on sequence. The DiSC35 data show that aurein 2.2 is better at inducing membrane depolarization than aurein 2.3, which is itself as efficient as gramicidin S. The OCD data of the three peptides in 1:1 POPC/POPG model membranes show similar distinction in the behaviour of the peptides suggesting that perhaps this lipid composition is the optimal one for studying this subset of the aurein peptide family. Finally, if one assumes that 1:1 POPC/POPG is a more relevant model  97  CHAPTER 3  membrane and that the peptides function by inducing distorted toroidal pores, toroidal pores/liposomes or localized membrane aggregates, then the sequence dependence observed in the DiSC35 results may be related to differences in the propensity of aurein 2.2 to form oligomers relative to the other two peptides in model membrane.  3.5 Materials and methods 3.5.1 Materials Peptide synthesis materials were purchased as described in Chapter 2 (1). Mylar plates were made by cutting Melinex Teijin films from Dupont (Wilton, Middlesbrough, UK). 1,2dimyristoyl-sn-glycero-3-phosphocholine (DMPC), 1,2-dimyristoyl-sn-glycero-3-[phospho-rac(1-glycerol)] (DMPG), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (POPG) were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and obtained dissolved in chloroform. Bis[ , −bis(carboxymethyl)aminomethyl]fluorescein (calcein) was purchased from SigmaAldrich (St. Louis, MO, USA), as was 3,3'-dipropylthiadicarbocyanine iodide (DiSC35).  3.5.2 Methods 3.5.2.1 Peptide synthesis Aurein 2.2, aurein 2.3 and aurein 2.3-COOH were synthesized as previously described (1), using an CS Bio Co. peptide synthesizer (Menlo Park, CA, USA) by in situ neutralization Fmoc chemistry, using Rink or Wang resin, as appropriate. The C-terminal Leu was doublecoupled (i.e. allowed to couple to the resin for 60 minutes, washed, and then allowed to couple to resin for a further 60 minutes before the next step in the peptide synthesis) to improve the yield.  98  CHAPTER 3  3.5.2.2 Purification The crude peptide product was purified by preparative RP-HPLC on a Waters 600 system (Waters Limited, Mississauga, ON, Canada) with 229 nm UV detection using a Phenomenex (Torrance, CA) C4 preparative column (20.0 µm, 2.1 cm × 25.0 cm) as previously (1). The identity of the products was verified using electrospray ionization (ESI) mass spectrometry and MALDI-TOF as previously described (1) and confirmed to be 98 ~ 99% pure.  3.5.2.3 Solution circular dichroism (CD) sample preparation Solution CD samples with constant peptide concentration of 2.0 mM were prepared in different peptide-to-lipid (P/L, DMPC/DMPG or POPC/POPG) molar ratios: 1:15, 1:50, and 1:100 (or 6.7, 2.0, and 1.0 mole % of peptide with respect to lipids). Appropriate amounts of lipids in chloroform were dried using a stream of air to remove most of the chloroform and vacuum dried overnight in a 5.0 ml round bottom flask. After adding 450.0 µl of ddH2O and 0.1 µmol (0.16 mg) of peptide to dried lipids, the mixture was sonicated in a water bath for a  minimum of 30 min (until the solution was no longer turbid) to ensure lipid vesicle formation. For all samples, corresponding background samples without peptides were prepared for spectral subtraction.  3.5.2.4 Mechanically oriented sample preparation Solid-state NMR samples were prepared for three different P/L molar ratios: 1:15, 1:80 and 1:120, following procedures similar to those reported (64,65).  The amount of lipids  (dissolved in chloroform) was kept constant at 9.59 µmol. The lipid was dried using a stream of air to remove most of the chloroform and vacuum dried overnight in a 5.0 ml round bottom flask. Then, the appropriate amount of peptide was added and the mixture was redissolved in 400 µl of  99  CHAPTER 3  ddH2O by sonication. The mixture was deposited in 10 µl portions repeatedly onto 9 Mylar plates placed in a Petri dish. Between depositions, most of ddH2O was evaporated before the next portion was deposited onto the plate. The plated samples were then placed in a 93% relative humidity chamber and were indirectly hydrated by incubating at 37ºC for 7 days. The humidity of the samples was verified by visual inspection (well hydrated samples are translucent). The degree of alignment was verified by solid-state  31  P NMR. Consistent sample preparation was  verified by preparing 2 ~ 3 samples for each lipid composition and peptide concentration. Finally, the plated samples were wrapped in a thin layer of parafilm before data acquisition. Oriented CD samples were prepared in a similar fashion as described above. The peptide amount was kept constant at 0.5 µmol (0.81 mg) and mixed with appropriate molar ratios of lipids: 1:15, 1:30, 1:40, 1:80 and 1:120 P/L molar ratios (or 6.67, 3.33, 2.50, 1.25, and 0.83 mole % of peptide with respect to lipids) and sonicated in 2.0 ml of ddH2O. Each mixture was deposited in 90 µl portions onto 3 cm × 1 cm and 1 mm thick quartz slides, cleaned thoroughly with ddH2O and ethanol prior to sample preparation. Clear layers of samples were observed on the slides after indirect hydration of the samples. Prior to CD spectral acquisition, each sample was covered with a second slide with a spacer (6 layers of stacked parafilm in a rectangular 3 cm × 1 cm frame with 2 mm width) in between.  3.5.2.5 CD and MR spectroscopy Solution and oriented CD experiments were carried out using a JASCO J-810 spectropolarimeter (Victoria, BC, Canada) at 30oC as previously described (1). Solid-state  31  P NMR experiments on mechanically aligned lipid bilayer samples were  carried out on the Bruker 500-MHz NMR spectrometer at 30°C, operating at a phosphorus frequency of 202.48 MHz with proton decoupling as previously reported (1). The 90o pulse was 100  CHAPTER 3  set to 10.50 µs and a 3 second recycle delay was used. The spectra were acquired using 2048 scans and processed without any line broadening.  3.5.2.6 Differential scanning calorimetry Each aurein peptide was added at 1:15 P/L molar ratio to multilamellar vesicles of 1:1 POPC/POPG (25 mg/ml) resuspended in HEPES buffer (20 mM HEPES, pH 7.0, 100 mM NaCl). Samples were degassed for 5 min prior to loading the sample into a VP-DSC or multicell DSC (Calorimetry Sciences, South Provo, UT, USA), located at the University of British Columbia Centre for Biological Calorimetry. The samples were heated and cooled over a temperature range of -20ºC ~ 70ºC at a rate of 0.33ºC/minute. The resulting data were converted to units of molar heat capacity after baseline correction by subtracting a blank buffer scan.  3.5.2.7 Calcein release assays Calcein release assays for different aurein peptides in specific lipid mixtures were carried out as previously described in Chapter 2.  3.5.2.8 DiSC35 assays The ability of the aurein peptides to depolarize the cytoplasmic membrane of S. aureus C622  was  determined  using  the  membrane  potential  sensitive  dye  3,3'-  dipropylthiadicarbocyanine iodide (DiSC35). C622 was grown to mid-logarithmic phase in LB media, centrifuged, washed in 5 mM HEPES and 20 mM glucose and then resuspended in the same buffer to a final OD600 of 0.05. A final concentration of 200 mM KCl was added to the cells and left for 30 minutes at room temperature to equilibrate cytoplasmic and external K+ concentrations, before the DiSC35 was added at a final concentration of 0.8 µM for 30 minutes.  101  CHAPTER 3  Changes in fluorescence resulting from disruption of the membrane potential were measured up to 5 minutes using a Perkin-Elmer 640-10S spectrofluorimeter (Waltham, MA, USA) using an excitation wavelength of 622 nm and a emission wavelength of 670 nM following the addition of each aurein peptide at 1X, 2X and 5X MIC (as previously ascertained (1)) to 2 ml cell suspension in a 1 cm quartz cuvette. Aurein 2.3-COOH was only tested at 1X MIC due to its high MIC. All membrane permeabilization results were compared to gramicidin S, which acts as a positive control.  102  CHAPTER 3  3.6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.  15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.  Pan, Y. L., Cheng, J. T. J., Hale, J. D., Pan, J., Hancock, R. E. W. and Straus, S. K. Biophys J 2007, 92, 2854-64. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry 2006, 45, 5793-9. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids 2003, 122, 107-20. Wu, Y., He, K., Ludtke, S. J. and Huang, H. W. Biophys J 1995, 68, 2361-9. Dave, P. C., Billington, E., Pan, Y. L. and Straus, S. K. Biophys J 2005, 89, 2434-42. Harzer, U. and Bechinger, B. Biochemistry 2000, 39, 13106-14. Tremouilhac, P., Strandberg, E., Wadhwani, P. and Ulrich, A. S. J Biol Chem 2006, 281, 32089-32094. Bechinger, B. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1529-1539. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem 2000, 267, 5330-41. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett 2001, 501, 146-50. Wieprecht, T., Dathe, M., Krause, E., Beyermann, M., Maloy, W. L., MacDonald, D. L. and Bienert, M. Febs Letters 1997, 417, 135-140. Dathe, M., Nikolenko, H., Klose, J. and Bienert, M. Biochemistry 2004, 43, 9140-50. Hilpert, K., Elliott, M. R., Volkmer-Engert, R., Henklein, P., Donini, O., Zhou, Q., Winkler, D. F. and Hancock, R. E. Chem Biol 2006, 13, 1101-7. Lewis, R. N., Liu, F., Krivanek, R., Rybar, P., Hianik, T., Flach, C. R., Mendelsohn, R., Chen, Y., Mant, C. T., Hodges, R. S. and McElhaney, R. N. Biochemistry 2007, 46, 1042-54. Johnson, W. C. Proteins 1999, 35, 307-12. Provencher, S. W. and Glockner, J. Biochemistry 1981, 20, 33-7. Sreerama, N., Venyaminov, S. Y. and Woody, R. W. Protein Sci 1999, 8, 370-80. Sreerama, N. and Woody, R. W. J Mol Biol 1994, 242, 497-507. Sreerama, N. and Woody, R. W. Anal Biochem 1993, 209, 32-44. Kucerka, N., Kiselev, M. A. and Balgavy, P. Eur Biophys J 2004, 33, 328-34. Mecke, A., Lee, D. K., Ramamoorthy, A., Orr, B. G. and Banaszak Holl, M. M. Biophys J 2005, 89, 4043-50. Leidy, C., Linderoth, L., Andresen, T. L., Mouritsen, O. G., Jorgensen, K. and Peters, G. H. Biophys J 2006, 90, 3165-75. Shai, Y. Biochim Biophys Acta 1999, 1462, 55-70. Papo, N. and Shai, Y. Cell Mol Life Sci 2005, 62, 784-90. Matsuzaki, K. Biochim Biophys Acta 1999, 1462, 1-10. Yang, L., Weiss, T. M., Lehrer, R. I. and Huang, H. W. Biophys J 2000, 79, 2002-9. Huang, H. W. Biochemistry 2000, 39, 8347-52. Wu, M., Maier, E., Benz, R. and Hancock, R. E. W. Biochemistry 1999, 38, 7235-42. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother 1999, 43, 1317-23. Mani, R., Buffy, J. J., Waring, A. J., Lehrer, R. I. and Hong, M. Biochemistry 2004, 43, 13839-48. Buffy, J. J., McCormick, M. J., Wi, S., Waring, A., Lehrer, R. I. and Hong, M. Biochemistry 2004, 43, 9800-12. 103  CHAPTER 3  32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.  47. 48. 49. 50. 51. 52. 53.  54. 55. 56. 57. 58. 59. 60.  Frerman, F. E. and White, D. C. J Bacterio. 1967, 94, 1868-74. Joyce, G. H., Hammond, R. K. and White, D. C. J Bacterio 1970, 104, 323-30. Appelbaum, P. C. and Bozdogan, B. Clin Lab Med 2004, 24, 381-402. Bozdogan, B., Esel, D., Whitener, C., Browne, F. A. and Appelbaum, P. C. J Antimicrob Chemother 2003, 52, 864-8. Conrad, R. S. and Gilleland, H. E., Jr. J Bacteriol 1981, 148, 487-97. Gmeiner, J. and Martin, H. H. Eur J Biochem 1976, 67, 487-94. Morein, S., Andersson, A., Rilfors, L. and Lindblom, G. J Biol Chem 1996, 271, 6801-9. Clejan, S., Krulwich, T. A., Mondrus, K. R. and Seto-Young, D. J Bacteriol 1986, 168, 334-40. Beining, P. R., Huff, E., Prescott, B. and Theodore, T. S. J Bacteriol 1975, 121, 137-43. Trombe, M. C., Laneelle, M. A. and Laneelle, G. Biochim Biophys Acta 1979, 574, 290300. Haque, M. A. and Russell, N. J. Microbiology 2004, 150, 1397-404. Ramamoorthy, A., Thennarasu, S., Lee, D. K., Tan, A. and Maloy, L. Biophys J 2006, 91, 206-16. Lewis, R. N., Zhang, Y. P. and McElhaney, R. N. Biochim Biophys Acta 2005, 1668, 203-14. Yamaguchi, S., Huster, D., Waring, A., Lehrer, R. I., Kearney, W., Tack, B. F. and Hong, M. Biophys J 2001, 81, 2203-14. Chekmenev, E. Y., Vollmar, B. S., Forseth, K. T., Manion, M. N., Jones, S. M., Wagner, T. J., Endicott, R. M., Kyriss, B. P., Homem, L. M., Pate, M., He, J., Raines, J., Gor'kov, P. L., Brey, W. W., Mitchell, D. J., Auman, A. J., Ellard-Ivey, M. J., Blazyk, J. and Cotten, M. Biochim Biophys Acta 2006, 1758, 1359-72. Epand, R. F., Savage, P. B. and Epand, R. M. Biochim Biophys Acta 2007, 1768, 2500-9. Henzler-Wildman, K. A., Martinez, G. V., Brown, M. F. and Ramamoorthy, A. Biochemistry 2004, 43, 8459-69. Lu, J. X., Damodaran, K., Blazyk, J. and Lorigan, G. A. Biochemistry 2005, 44, 1020817. Wimmer, R., Andersen, K. K., Vad, B., Davidsen, M., Molgaard, S., Nesgaard, L. W., Kristensen, H. H. and Otzen, D. E. Biochemistry 2006, 45, 481-97. Epand, R. F., Schmitt, M. A., Gellman, S. H. and Epand, R. M. Biochim Biophys Acta 2006, 1758, 1343-50. Klajnert, B., Janiszewska, J., Urbanczyk-Lipkowska, Z., Bryszewska, M. and Epand, R. M. Int J Pharm 2006, 327, 145-52. Zweytick, D., Pabst, G., Abuja, P. M., Jilek, A., Blondelle, S. E., Andra, J., Jerala, R., Monreal, D., Martinez de Tejada, G. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1426-35. Zasloff, M. Engl J Med 2002, 347, 1199-200. Hancock, R. E. W. Lancet Infect Dis 2001, 1, 156-64. Wegener, K. L., Wabnitz, P. A., Carver, J. A., Bowie, J. H., Chia, B. C., Wallace, J. C. and Tyler, M. J. Eur J Biochem 1999, 265, 627-37. Henzler Wildman, K. A., Lee, D. K. and Ramamoorthy, A. Biochemistry 2003, 42, 654558. Ramamoorthy, A., Lee, D. K., Santos, J. S. and Henzler-Wildman, K. A. J Am Chem Soc 2008, 130, 11023-9. Ramamoorthy, A., Kandasamy, S. K., Lee, D. K., Kidambi, S. and Larson, R. G. Biochemistry 2007, 46, 965-75. Wimley, W. C. and White, S. H. at Struct Biol 1996, 3, 842-8. 104  CHAPTER 3  61. 62. 63. 64. 65.  Bull, H. B. and Breese, K. Arch Biochem Biophys 1974, 161, 665-70. Guy, H. R. Biophys J 1985, 47, 61-70. Wilson, K. J., Honegger, A., Stotzel, R. P. and Hughes, G. J. Biochem J 1981, 199, 3141. Moll, F., 3rd and Cross, T. A. Biophys J 1990, 57, 351-62. Hallock, K. J., Lee, D. K., Omnaas, J., Mosberg, H. I. and Ramamoorthy, A. Biophys J 2002, 83, 1004-13.  105  CHAPTER 4  CHAPTER 4: The importance of bacterial membrane composition: CD and MR studies of the structure and activity of aurein 2.2 and aurein 2.3 in CL/POPG and POPE/POPG model membranes†  4.1 Introduction One approach to determining the mode of action of cationic antimicrobial peptides (CAPs) in previous chapters (Chapter 2 and Chapter 3) is to establish the nature of the interaction of the peptides with bacterial membranes. However, given that real bacterial membranes consist of many different types of lipids and protein complexes, most studies establish the mechanism of action for the peptide-membrane interaction using model membranes. Over the years, a number of lipids have been used for such studies: 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) (e.g. aurein 1.2 (1)), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) (e.g. MSI-78 and MSI-594 (2)), 1,2-diphytanoyl-sn-glycero-3-phosphatidylcholine (DPhPC) (e.g. alamethicin (3,4)), and other diacylphosphatidylcholine membranes (e.g. K2(LA)xK2(5)). Lipid mixtures such as DMPC/1,2-dimyristoyl-sn-glycero-3-phospho-(1'-rac-glycerol) (DMPG) (e.g. PGLa (6)), POPC/1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(1'-rac-glycerol) (POPG) (e.g. MSI-78 and MSI-594 (2)), or 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE)/POPG (7-12) to name but a few, have also been used. The choice of lipids used as a bacterial model membrane depends on a number of factors. These include the properties of lipids, such as hydrophobic bilayer thickness, phase transition †  A version of this chapter has been submitted as follows:  Cheng, J. T. J., Hale, J. D., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2010. The Importance of Bacterial Membrane Composition on the Structure/Function of Aurein 2.2 and some of its Mutants. 106  CHAPTER 4  temperature, and miscibility. For example, DMPC and DMPG are often used to investigate the effect of CAPs on chain melting temperatures, because these lipids undergo a phase transition close to room temperature (24°C) (13). The properties of lipids are also linked to the techniques used to investigate antimicrobial peptide-lipid interactions. For instance, differential scanning calorimetry (DSC) experiments are often performed using DMPC, DMPG, or a mixture of these lipids (13-15), because routine calorimeters operate in the range of 0 ~ 90°C. DSC experiments that aim to probe whether antimicrobial peptides induce membrane curvature will be performed using, for instance, DiPoPE (13,16). In some cases, the choice of lipids also depends on the ease with which the lipids form aligned bilayers, when placed on solid glass or polymer supports. POPC/POPG mixtures have traditionally been used for 31P NMR studies in mechanically aligned bilayers because of the ease of preparing aligned samples (17). Even so, some studies utilize other materials (e.g. naphthalene (18)) or direct rehydration (e.g. 1 µl of ddH2O per slide prior to sample stacking (19)), to help resolve alignment issues in order to improve lipid bilayer alignment.  Finally, to completely describe peptide-membrane interactions, factors such as  peptide-to-lipid ratio, hydration, buffer composition (20), and the nature of the lipid phase (4) also need to be considered. Ultimately, the choice of lipids should be dictated by biological relevance: the results obtained from the biophysical studies should correlate to some extent with assays performed on intact bacteria that assess the actual antimicrobial activity/mode of killing (e.g. MIC, membrane depolarization). Given that only a small number of CAPs have been studied extensively in different membrane environments, the ultimate bacterial model membrane which is at the same time robust and applicable for a range of methods remains elusive. The quest for good lipids requires a series of peptides for which a large number of biophysical and activity data are available. Our initial studies (Chapter 2 and Chapter 3) focused on the structure and membrane interaction of aurein 2.2, aurein 2.3, and aurein 2.3-COOH in DMPC/DMPG and POPC/POPG bacterial model 107  CHAPTER 4  membranes (21). The biophysical component of these studies were conducted in POPC/POPG model membranes, which was then proposed to be an ideal model bacterial membrane for S. aureus in our previous studies (21) and in many studies of other antimicrobial peptides (22-24).  Although phosphatidylcholine/phosphatidylglycerol (PC/PG) model membranes have been used extensively over the years (1,25-33), a number of recent studies have adopted other more “relevant” lipids mixtures, such as phosphatidylethanolamine (PE)/PG (7-12) and cardiolipin (CL)/PG (34,35). Indeed, recent studies have been conducted in model membranes consisting of different POPE/POPG ratios (36) (mol/mol or w/w), including 3:1 (7,8,11,12,18), 7:3 (10,37,38), and 3:2 (19), to name but a few, and CL/PG. Therefore, it is important to assess whether these lipid mixtures are better model membranes than POPC/POPG for the aurein 2 peptide family. In the study reported in this chapter, we have therefore chosen to investigate the structure and membrane interactions of the aurein peptides in 1:1 POPE/POPG and 1:1 CL/POPG (mol/mol) membranes. Since 1:1 CL/POPG (4:6 CL/PG in S. aureus) is deemed to be a model (39) for Gram positive bacteria such as S. aureus (40), the structure and membrane interaction of aurein 2.2 and aurein 2.3 in this environment was determined using solution circular dichroism (CD) spectroscopy, oriented CD (OCD) and solid-state  31  P NMR methods.  The results were compared to the biophysical data obtained in POPC/POPG and to the MIC data and membrane depolarization data using S. aureus C622. Next, the structure and membrane interaction of aurein 2.2, aurein 2.3, and aurein 2.3-COOH were also investigated in 1:1 POPE/POPG (mol/mol) lipids, a possible model for other Gram positive bacteria, such as Bacillus cereus. The conformation of these peptides was determined using CD. Oriented CD  (OCD) and solid-state  31  P NMR studies were again used to examine the interaction between  these peptides and POPE/POPG. The minimal inhibitory concentration (MIC) of each peptide against B. cereus strain C737 was determined in order to correlate the behaviour of the aurein peptides observed in model membranes to the ones obtained in the presence of intact bacteria. In 108  CHAPTER 4  addition, the MICs were determined against P. aeruginosa (Pa01) and E. coli (C500), which contain ~ 60% and ~ 80% PE, respectively. Overall, these data should allow us to determine the relevant model bacterial membranes for studying the behaviour of the aurein peptides in different bacteria, and to better understand how sequence modulates structure-function relationships in different model membrane environments. This chapter will give a detailed overview on the studies of the behaviour of the aurein peptides in 1:1 CL/POPG and 1:1 POPE/POPG (mol/mol) membranes. This chapter will be divided into three sections. Section 4.2 will be divided into three subsections. Section 4.2.1, Section 4.2.2 and Section 4.2.3 provide the structural and functional analyses on the three aurein peptides in the presence of different model bacterial membranes, respectively. Section 4.2.4 outlines the antimicrobial activity of the three aurein peptides in bacteria containing PE as a major membrane lipid component, including Bacillus cereus (strain C737), Pseudomonus aeruginosa (strain Pa01) and Escherichia coli (strain C500), as a comparison with the previously  obtained MIC results in S. aureus. Section 4.3 gives the detailed explanation on how these peptides interact with and perturb different lipid bilayers differently, and whether structural differences play a role in peptide-membrane interactions. Section 4.4 summarizes the findings and provides the conclusion on this study.  4.2 Results Determining the structure is an essential step to give the first insight into whether a given antimicrobial peptide remains structured or undergoes a different conformational change when exposed to different lipid environment. Whether a difference in structure causes a difference in peptide-membrane interaction is an important step to deduce the structure-function relationship of the given antimicrobial peptide in a model membrane system and can help explain why an  109  CHAPTER 4  antimicrobial peptide is active or inactive against a target bacterial strain. This section will describe the structural and functional studies on the three aurein peptides in different model membranes, 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol).  4.2.1 Secondary structure of the aurein peptides In order to verify that the three aurein peptides remain structured, or whether they changed structure in the presence of different lipid membranes, solution CD experiments were performed in 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) small unilamellar vesicles (SUVs). Figure 4.1 shows solution CD spectra of the aurein peptides in 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) SUVs as a function of P/L molar ratio. All spectra consisted of a maximum at 195 nm and two minima at 207 nm and 222 nm, which are spectral characteristic of peptide that adopt α-helical structure. This demonstrates that all aurein peptides adopted an α-helical conformation regardless of different membrane environment and P/L molar ratios. This indicates that the structures of the aurein peptides were independent of membrane type and peptide concentration. As previously observed (41), similar intensities were found for all peptide-to-lipid molar ratios (P/L = 1:15, 1:50, and 1:100) studied, indicating that maximum peptide folding occurred.  Saturation would be observed with a  combination of signals from both α-helical and random coil structures (1). Since all aurein peptides adopted α-helical conformation and not an entirely different structure from each other, determining structure content differences reveals whether these peptides are better structured (% α-helix) in a specific membrane environment. This information may help one understand whether a higher % helical content is necessary for a higher antimicrobial activity, which will be discussed in details in Section 4.3.  110  100 80  (d)  60 40 20 0 -20 -40 -60 190  210  230  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (a)  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  CHAPTER 4 100 80 60 40 20 0 -20 -40 -60  250  190  100 80  (e)  60 40 20 0 -20 -40 -60 190  210  230  20 0 -20 -40 -60 190  60 40 20 0 -20 -40 -60 230  Figure 4.1.  250  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (f)  210  210  230  250  Wavelength (nm)  80  Wavelength (nm)  250  60 40  250  100  190  230  100 80  Wavelength (nm) (c)  210  Wavelength (nm)  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (b)  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  Wavelength (nm)  100 80 60 40 20 0 -20 -40 -60 190  210  230  250  Wavelength (nm)  Solution CD spectra of the aurein peptides in 1:1 CL/POPG (mol/mol) (left panel) and 1:1 POPE/POPG (mol/mol) (right panel) SUVs: (a) & (d) aurein 2.2; (b) & (e) aurein 2.3; and (c) & (f) aurein 2.3-COOH (solid line, P/L = 1:15; dash line, P/L = 1:50; dotted line, P/L = 1:100). Spectra indicate that the aurein peptides adopted an αhelical conformation in the presence of 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) SUVs.  To examine the % structure content of aurein peptides in CL/POPG and POPE/POPG SUVs, all spectra were fitted using three different programs (CDSSTR (42), CONTINLL (43), and SELCON3 (44-46)), as previously (21). Figure 4.2 shows % structure content of the aurein peptides in the presence of different lipid membranes as a function of P/L molar ratio.  111  CHAPTER 4  Figure 4.2.  Structure content of the aurein peptides in the presence of (a) 1:1 POPC/POPG (mol/mol), (b) 1:1 CL/POPG (mol/mol), and (c) 1:1 POPE/POPG (mol/mol) SUVs: % α-helix and % random coil were plotted as a function of P/L molar ratios and aurein peptides (light grey = aurein 2.2, medium grey = aurein 2.3, dark grey = aurein 2.3-COOH). The graphs show that % α-helix decreased and % random coil increased as P/L molar ratio decreased. This indicates that decreased % α-helix contributed to increased % random coil when peptide concentration decreased. Direct observation on % α-helix of the aurein peptides in the presence of different lipids suggests that the aurein peptides showed generally a higher helical content in 1:1 POPE/POPG (mol/mol) than in 1:1 POPC/POPG (mol/mol) and 1:1 CL/POPG (mol/mol) SUVs.  112  CHAPTER 4  The data for POPC/POPG SUVs were previously obtained results (Chapter 3) (21). The results showed that % α-helix decreased and % random coil increased as P/L molar ratio decreased (Figure 4.2(a) ~ (c)). This indicates that decreased % α-helix contributed to increased % random coil when peptide concentration decreased. These changes occurred for all peptides in all membrane environment examined, which suggests that these changes were dependent on P/L molar ratios but independent of membrane types. At high peptide concentration, the aurein peptides adopted close to 100% α-helical conformation. This shows that the helical content increased with increased peptide concentration. This indicates that high concentrations were needed to achieve maximum peptide binding to membranes, as previously observed (21). At 1:50 P/L molar ratio, both aurein 2.2 and aurein 2.3-COOH showed reduced % α-helix, except that aurein 2.3 still maintained close to 100% α-helix in the presence of POPC/POPG and CL/POPG SUVs (Figure 4.2(a) & (b)). However, aurein 2.2 and aurein 2.3-COOH showed no significant % α-helix decrease in the presence of POPE/POPG SUVs.  At low peptide  concentration, all peptides showed pronounced % α-helix decrease. Direct observation on the % α-helix of the aurein peptides in the presence of different lipids suggests that the aurein peptides  showed generally a higher helical content in POPE/POPG than in POPC/POPG and CL/POPG SUVs.  4.2.2 Membrane insertion states of the aurein peptides Since the structure of the three aurein peptides did not change significantly in response to different membrane environment, assessing the extent of peptide insertion into the lipid bilayers of different composition would be critical to understanding how the aurein peptides interact with different model bacterial membranes. OCD experiments were conducted to investigate the membrane insertion states of the three aurein peptides in 1:1 CL/POPG (mol/mol) and 1:1 113  CHAPTER 4  POPE/POPG (mol/mol) bilayers.  For both OCD (Section 4.2.2) and solid-state  31  P NMR  (Section 4.2.3), samples were prepared in similar ways so that the data sets could be directly compared and also to verify that the samples were aligned. All experiments were conducted at 30°C (liquid crystalline phase) for consistent comparison with our previous study (Chapter 2 and Chapter 3). In addition, experiments were repeated at least twice to ensure reproducibility.  30  (d)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  (a)  10 0 -10 -20 -30  30 20 10 0 -10 -20 -30  190  200  210  220  230  240  250  190  Wavelength (nm) (e)  20 10 0 -10 -20 -30 200  210  220  230  240  240  250  240  250  240  250  20 10 0 -10 -20  250  190  200  210  220  230  Wavelength (nm)  30  (f)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  230  30  Wavelength (nm)  10 0 -10 -20 -30  30 20 10 0 -10 -20 -30  190  200  210  220  230  Wavelength (nm)  Figure 4.3.  220  -30 190  (c)  210  Wavelength (nm)  30  Ellipticity (mdeg)  Ellipticity (mdeg)  (b)  200  240  250  190  200  210  220  230  Wavelength (nm)  Oriented CD spectra of the aurein peptides in 1:1 CL/POPG (mol/mol) (left panel) and 1:1 POPE/POPG (mol/mol) (right panel) bilayers: (a) & (d) aurein 2.2; (b) & (e) aurein 2.3; and (c) & (f) aurein 2.3-COOH (P/L molar ratios = 1:15 (blue), 1:80 (red), and 1:120 (green)). Spectra were normalized at 222 nm as in Chapter 2. The spectra show that the aurein peptides inserted into 1:1 CL/POPG (mol/mol) bilayers in similar behaviour as in 1:1 POPC/POPG (mol/mol) bilayers. In the presence of 1:1 POPE/POPG (mol/mol) bilayers, the aurein peptides inserted at slightly different threshold P/L molar ratios: between 1:15 and 1:80 for aurein 2.2 and aurein 2.3, and possibly less than 1:120 for aurein 2.3-COOH. 114  CHAPTER 4  Figure 4.3 shows OCD spectra of the aurein peptides in the presence of 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) bilayers as a function of P/L molar ratio. The spectra were normalized such that the intensities of all spectra at 222 nm are the same. The spectra showed that the aurein peptides demonstrated similar insertion profiles in CL/POPG bilayers as in 1:1 POPC/POPG (mol/mol) bilayers (21), but demonstrated different insertion profiles in POPE/POPG bilayers. In the presence of CL/POPG bilayers, aurein 2.2 inserted (inserted, Istate or tilt, T-state) at a threshold P/L molar ratio between 1:80 and 1:120, where aurein 2.3 displayed a gradual insertion profile, and aurein 2.3-COOH inserted at a threshold P/L molar ratio between 1:15 and 1:80. In the presence of POPE/POPG bilayers, both aurein 2.2 and aurein 2.3 showed a gradual insertion profile, but aurein 2.3-COOH inserted at a threshold P/L molar ratio less than 1:120. Since PE is a better H+ donor than PC (47), it is most likely that aurein 2.3COOH with a negatively charged C-terminus interacts better with PE than PC headgroup electrostatically.  These data indicate that the three aurein peptides behaved similarly in  POPC/POPG and CL/POPG bilayers, but not in POPE/POPG bilayers.  The OCD results  presented here illustrate that different model bacterial membranes can have significant effects on the aurein peptide insertion profiles. Table 4.1 summarizes the membrane insertion states of the three aurein peptides in the presence of different lipid bilayers as a function of P/L molar ratio.  Table 4.1.  Membrane insertion states of the three aurein peptides in the presence of different lipid bilayers at different P/L molar ratios.  Lipid bilayers P/L molar ratio 1:15 1:80 1:120  1:1 CL/POPG (mol/mol) A2.2 A2.3 OH I I I I I/S S S S S  1:1 POPE/POPG (mol/mol) A2.2 A2.3 OH I/S I I I/S I/S I S S I  *A2.2 = aurein 2.2-CONH2; A2.3 = aurein 2.3-CONH2; OH = aurein 2.3-COOH. *P/L = peptide-to-lipid; I = inserted state; S = surface-adsorbed state.  115  CHAPTER 4  4.2.3 Lipid headgroup perturbation by the aurein peptides Since the three aurein peptides showed different insertion profiles in different lipid bilayers, whether these differences were as a result of different membrane perturbation mechanisms remained to be solved.  31  P NMR experiments were conducted to determine whether  the peptide insertion is accompanied by a lipid headgroup perturbation and whether this membrane disruption occurs via a barrel-stave model, carpet model, or toroidal pore model (4852); a micellar aggregate channel model (53,54); or a detergent-like mechanism (20).  31  P NMR  spectra were recorded for 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) bilayers with and without the three aurein peptides, as our previous studies (21). Figure 4.4 shows  31  P NMR spectra of mechanically aligned 1:1 CL/POPG (mol/mol)  bilayers in the absence and presence of the three aurein peptides as a function of P/L molar ratio. In the absence of the aurein peptides, the spectra of CL/POPG bilayers consisted of a single peak at ~ 8.7 ppm. This indicates that the lipid bilayers were aligned with normal parallel to the magnetic field. In the presence of the aurein peptides, the spectra showed an increased powderpattern signal in the range of -5 ~ 30 ppm, indicative of random headgroup orientations. The powder-pattern signal increased as peptide concentration increased. This indicates that the membrane disordering effect was concentration-dependent for all three aurein peptides. At high peptide concentrations, all aurein peptides showed relatively similar disordering effect. At low peptide concentrations, aurein 2.2 exhibited more disruptive effect than aurein 2.3 and aurein 2.3-COOH. In the presence of aurein 2.3-COOH at low concentrations, only slightly broadened peaks were observed. This suggests that aurein 2.3-COOH did not have a significant disordering effect on CL/POPG bilayers, which is consistent with our previous  31  P NMR results in 4:1  POPC/POPG (mol/mol) bilayers (21).  116  CHAPTER 4  Figure 4.4.  Solid-state 31P MR spectra of mechanically aligned 1:1 CL/POPG (mol/mol) bilayers with and without the three aurein peptides: (a) CL/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of aurein 2.2 (left panel), aurein 2.3 (centre panel), and aurein 2.3-COOH (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  Examining the behaviour of the aurein peptides in 1:1 POPE/POPG (mol/mol) bilayers would reveal how and whether these peptides interact differently with model bacterial membranes containing similar headgroups (PC versus PE). Figure 4.5 shows  31  P NMR spectra  of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers in the absence and presence of the three aurein peptides as a function of P/L molar ratio.  117  CHAPTER 4  Figure 4.5.  Solid-state 31P MR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without the three aurein peptides: (a) POPE/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of aurein 2.2 (left panel), aurein 2.3 (centre panel), and aurein 2.3-COOH (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  In the absence of the aurein peptides, the spectra of POPE/POPG bilayers consisted primarily of two single resonances at 30 ppm (PE) and 15 ppm (PG). This indicates that the lipid bilayers were aligned with their normal parallel to the magnetic field. In the presence of the aurein peptides, the spectra showed slightly broadened peaks at 30 ppm, indicating increased contribution from unaligned phosphorus headgroups. In addition, powder-pattern signal was also observed in the -10 ~ 30 ppm range, indicative of random headgroup orientations. The membrane disordering effect was found to be concentration-dependent and peptide-specific. At high peptide concentrations, the spectra exhibited enlarged underlying powder pattern, and the 118  CHAPTER 4  effect was more prominent in the presence of aurein 2.2. At low peptide concentrations, the spectra still showed underlying powder pattern, but powder-pattern signal decreased with decreased peptide concentration. Contrary to our observed 31P NMR results in POPC/POPG (21) and CL/POPG (Figure 4.4) bilayers, the presence of aurein 2.3-COOH actually resulted in an increased underlying powder pattern in POPE/POPG bilayers. This suggests that aurein 2.3COOH displayed similar disordering effect, although to a slightly smaller magnitude, as aurein 2.2 and 2.3. This is opposite to our previous MIC results, where aurein 2.3-COOH is inactive or much less active against S. aureus (41). The underlying powder pattern indicates that these aurein peptides may disorder the bilayer headgroups to some extent (probably via mechanisms similar to toroidal pore (48-52) or toroidal pore/liposome formation). The extent of the membrane perturbation was found to be concentration-dependent and peptide-specific, but not for the mechanism. These data indicate that the membrane perturbation mechanism was selective on the presence of PC/CL or PE headgroups in the lipid bilayers (47).  4.2.4 Antimicrobial activity of the aurein peptides Given that the aurein peptides adopted structures with higher % α-helix and showed different peptide insertion profiles and membrane perturbation mechanisms (aurein 2.3-COOH only) in POPE/POPG membrane, the minimum inhibitory concentrations (MICs) of the three aurein peptides were determined against selected bacteria that have PE as a major membrane lipid: B. cereus (~ 43:40 PE/PG, % total lipids) strain C737, Pseudomonus aeruginosa (~ 60% PE, % total lipids) strain Pa01 and Escherichia. coli (~ 80% PE, % total lipids) strain C500, in order to correlate the observations in POPE/POPG membrane to the activity in live bacteria. The MICs of the three aurein peptides were determined using the method as previously described  119  CHAPTER 4  (41), with gramicidin S (GMS) as the positive control.  Rozek, T. et. al. has previously  determined MICs of aurein 2.2 and aurein 2.3 against B. cereus (55), but not for the inactive version of the peptides. Table 4.2 shows MIC values of the three aurein peptides against B. cereus C737, P. aeruginosa Pa01, and E. coli C500.  Table 4.2.  MIC results of aurein 2.2, aurein 2.3 and aurein 2.3-COOH against B. cereus, P. aeruginosa and E. coli. Minimal inhibitory concentrations (MICs) in µg/ml of the three aurein peptides and gramicidin S (GMS) (control) against B. cereus, P. aeruginosa and E. coli (see text for experimental details). MICs are given as the most frequently observed value obtained from repeat experiments. (MIC results in P. aeruginosa Pa01 and E. coli C500 are courtesy of Dr. J. D. Hale and M. Elliot).  Peptide  Sequence  Aurein 2.2 Aurein 2.3 Aurein 2.3-COOH GMS  GLFDIVKKVVGALGSL GLFDIVKKVVGAIGSL GLFDIVKKVVGAIGSL-OH L(D-)FPVOL(D-)FPVO  B. cereus C737 128 128 > 128 2  P. aeruginosa Pa01 > 128 > 128 > 128   E. coli C500 64 32 > 128   *(D-)F = D-Phe; O = Orn;  = data not available.  The MIC results indicate that all aurein peptides had very low to no activities against B. cereus, opposite to our observation in S. aureus under the same conditions as previously (41).  Rozek, T. et al. reported that both aurein 2.2 and 2.3 are not active against B. cereus (55). Aurein 2.3-COOH had an MIC of > 128 µg/ml against B. cereus, which renders it inactive or much less active (55,56), as expected from an inactive peptide analogue.  Given that 1:1  POPE/POPG (mol/mol) bilayers mimics the membrane composition of B. cereus, the MIC results of the aurein peptides against B. cereus are somewhat correlated with the OCD and  31  P  NMR results presented above. In other bacteria containing higher % PE, all aurein peptides still showed low antimicrobial activities. This indicates that the aurein peptides were much less active against any bacteria containing PE, and this decreased activity was not quantitatively dependent on the presence of PE in the bacterial membranes.  120  CHAPTER 4  4.3 Discussion A critical step to unlocking the mechanism of action of a given antimicrobial peptide is to understand how membrane composition modulates structure and peptide-lipid interactions. In this chapter, we have further investigated how two antimicrobial peptides from the Australian Southern Bell frog Litoria aurea, namely aurein 2.2 and aurein 2.3, interact with a range of currently used model bacterial membranes. We have also examined the peptide-lipid interaction of the inactive analogue aurein 2.3-COOH. Bacterial membrane composition varies from bacteria to bacteria and also as a response to changing environment (57,58) and to exposure to antibiotics (59-61).  It is well known that  different bacterial strains can have unique membrane compositions (25,61-66). For example, Gram positive bacteria have a membrane composition which is different from their Gram negative counterparts (40).  Even among the Gram positive bacteria themselves, different  membrane compositions have been identified. Both S. aureus and S. pneumonia have ~ 50:50 CL:PG (% of total lipids), whereas B. cereus and B. anthracis have ~ 43:40 PE:PG (% of total lipids). A slightly higher percentage of PE in the total membrane lipids was found for B. polymyxa (~60%), whereas as high as 70% PG was detected for B. subtilis. For Gram negative  bacteria, E. coli, E. cloacae, P. mirabilis, and K. pneumonia have ~ 80% PE, whereas Y. kristensenii and P. aeruginosa have ~ 60% PE (with PG being the next highest constituent).  Most Gram negative bacteria have a relatively low percentage of CL. Since S. aureus is the main antimicrobial target for the aurein peptides, incorporating CL and PG into model membranes is necessary to replicate biologically relevant membranes. However, synthetic CL has only become commercially available very recently.  Due to the high Tm of  tetramyristoylcardiolipin (TMCL) (~ 47°C) and tetrapalmitoylcardiolipin (TPCL) (~ 62°C) (67), using a mixture of either with POPG may not be biologically relevant. Bovine heart CL has  121  CHAPTER 4  unsaturated acyl chains ranging from 16C to 22C and a Tm below 0°C. Thus, a bovine heart CL:POPG mixture may be a more feasible/ideal model membranes for S. aureus (68) to conduct biophysical studies. Alternatively, a 1:1 mixture of POPE/POPG may be another ideal model membrane, in particular to mimic bacteria such as B. cereus and B. anthracis. It is widely accepted that cationic antimicrobial peptides only adopt secondary structure in the presence of membranes (48,69,70) and that the adoption of secondary structure is a key first step to membrane interaction. Our solution CD results demonstrated that the three aurein peptides tested here maintained their predominantly α-helical structure in 1:1 POPC/POPG (21) and, as demonstrated here, in 1:1 CL/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) SUVs. This is particularly true for P/L ratios of 1:50 or higher. Very few studies have demonstrated that amphibian antimicrobial peptides preserve their main conformation in both POPC/POPG and POPE/POPG membranes. To the best of our knowledge, only one molecular dynamics (MD) simulation and lipid vesicle study has shown that buforin II (BF2), an α–helical antimicrobial peptide from the Asian toad Bufo bufo gargarizans, retains its main structure in both POPC/POPG and POPE/POPG membranes (71). The presence of PE does not significantly alter secondary structure, nor does it abolish the ability of a peptide to bind to the lipid membranes. A recent review of protegrin-1 (PG-1) has demonstrated similar observations that PG-1 preserves its β-sheet structure in POPC/Chol bilayers as well as in POPE/POPG bilayers (72). The solution CD data on the aurein peptides demonstrated that the choice of PC versus CL versus PE in bacterial model membranes had small consequences on the structural content, but not on the overall structure of the peptide. Indeed, at low peptide concentrations, the aurein peptides were generally less helical in CL/PG, followed by PC/PG, and finally, PE/PG (Figure 4.2; P/L = 1:100). The study of BF2 has also demonstrated that the presence of PE induces a higher percentage of helical content in BF2 than PC (71). The higher percentage of helicity, however,  122  CHAPTER 4  does not necessarily correlate with higher antimicrobial activity or greater ability to disorder membranes. Indeed, a recent study of four α-helical antimicrobial peptides has shown that flexibility in the structure can actually promote better peptide binding to toroidal pores and, thus, better toroidal pore formation (73). Overall, the structural findings suggest that the secondary structure alone cannot necessarily be used as an indicator of how well a given antimicrobial peptide will interact with different lipids. Cationic antimicrobial peptides interact with bacterial membranes mainly through electrostatic interactions with the PG headgroups (26,74). If, however, the interaction with the PG headgroups were the only governing factor, then one would expect that the exact nature of the other lipid component would be irrelevant. The OCD data presented here suggest that the choice of PC versus CL versus PE plays a significant role in the peptide insertion ability of the aurein peptides. Indeed, aurein 2.2, aurein 2.3, and aurein 2.3-COOH inserted less readily into POPE/POPG bilayers, than POPC/POPG or CL/POPG membranes, which were roughly equivalent. Studies have suggested that PE/PG bilayers have stronger headgroup-headgroup interactions than PC/PG (and presumably also CL/PG), because the shapes of PE and PG are more complimentary. Tighter lipid packing will of course impede peptide insertion (47). The choice of PC versus CL versus PE appears to play less of a role though in how the aurein peptides perturb the lipid headgroups. The  31  P NMR data suggest that all aurein peptides  disordered the bilayer headgroups via mechanisms similar to toroidal pore (48-52) or toroidal pore/liposome formation, be that in POPC/POPG, CL/POPG, or POPE/POPG membranes. The extent of membrane perturbation was only found to be strongly concentration-dependent and partly dependent on the nature of the peptide. In the literature, several antimicrobial peptides perturb membranes similarly (48,50,52,75,76), be that in POPC, POPC/POPG, POPE, and POPE/POPG bilayers. Examples include MSI-78 (25,77,78), KIGAKI (79), and PG-1 (7-9).  123  CHAPTER 4  As the MIC data show, the aurein peptides were not very active against bacteria that contain 40% ~ 60% PE, namely bacteria like B. cereus and P. aeruginosa. The biophysical data correlate with this low activity since for 1:1 POPE/POPG, all the aurein peptides were found to insert into PE/PG membranes to a lesser extent than PC/PG.  The high helicity found in  POPE/POPG suggests that secondary structure alone cannot be used as an indicator of activity. Furthermore, the MIC data determined previously against S. aureus (41) show that the aurein peptides are quite effective at killing these bacteria and that 1:1 POPC/POPG or 1:1 CL/POPG bilayers are a good choice as model membranes for this bacteria.  4.4 Summary and conclusion Understanding the behaviour of a given amphibian antimicrobial peptide in both real bacteria and model membranes can reveal how it is active against certain bacteria and is important in any future design of better antimicrobial agents. For the aurein peptides, a good model for understanding how these peptides act on S. aureus is 1:1 POPC/POPG or 1:1 CL/POPG. For B. cereus, model membranes of 1:1 POPE/POPG are a good choice. In conclusion, specific model membranes should be used to model specific bacteria, given the unique membrane composition of individual bacterial species.  4.5 Materials and methods 4.5.1 Materials Peptide synthesis materials were purchased as previously described (21,41). Mylar plates were made by cutting Melinex Teijin films from Dupont (Wilton, Middlesbrough, UK). 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine phosphoethanolamine (POPE),  (POPC),  1-palmitoyl-2-oleoyl-sn-glycero-3-  1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] 124  CHAPTER 4  (POPG), and bovine heart cardiolipin (CL) were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and obtained dissolved in chloroform.  4.5.2 Methods 4.5.2.1 Peptide synthesis Aurein peptides were synthesized as previously described (21,41), using an CS Bio Co. peptide synthesizer (Menlo Park, CA, USA) by in situ neutralization Fmoc chemistry, using Rink or Wang resin with the double-coupling scheme, as appropriate.  4.5.2.2 Purification The crude peptide product was purified by preparative RP-HPLC on a Waters 600 system (Waters Limited, Mississauga, ON, Canada) with 229 nm UV detection using a Phenomenex (Torrance, CA) C4 preparative column (20.0 µm, 2.1 cm × 25.0 cm) as previously described (21,41). The identity of the products was verified using electrospray ionization (ESI) mass spectrometry and MALDI-TOF as previously described (21,41) and confirmed to be ≥ 99% pure.  4.5.2.3 Solution circular dichroism (CD) sample preparation Solution CD samples with a constant peptide concentration of 2.0 mM were prepared in different peptide-to-lipid (P/L) molar ratios of 1:15, 1:50, and 1:100, using 1:1 molar lipid mixtures of CL/POPG or POPE/POPG. All samples were prepared as previously described in Chapter 3.  125  CHAPTER 4  4.5.2.4 Mechanically oriented sample preparation Solid-state NMR samples were prepared for three different P/L molar ratios of 1:15, 1:80 and 1:120, as previously described in Chapter 3 (21). The plated samples were incubated at 37ºC for 6 days (CL/POPG) or 7 days (POPE/POPG), as previously described in Chapter 3. The humidity of the samples, degree of alignment, and consistent sample preparation were verified as previously described in Chapter 3. Oriented CD samples were prepared and deposited in a similar fashion as described above and in Chapter 3.  4.5.2.5 Circular dichroism Solution and oriented CD experiments were carried out using a JASCO J-810 spectropolarimeter (Victoria, BC, Canada) at 30oC as previously described (41).  4.5.2.6 MR spectroscopy Solid-state  31  P NMR experiments on mechanically aligned lipid bilayer samples were  carried out on a Bruker 500-MHz NMR spectrometer at 30°C, operating at a phosphorus frequency of 202.40 MHz with proton decoupling as previously reported (41). The 90o pulse was set to 6.00 µs (CL/POPG) and 11.25 µs (POPE/POPG), and a 3 second recycle delay was used. The spectra were acquired using 2048 scans and processed with 50 Hz line broadening.  4.5.2.7 MIC determination Minimal Inhibitory Concentrations (MIC) for aurein 2.3-COOH was determined based on the previously described modified methodology (80), with gramicidin S as the positive control.  126  CHAPTER 4  4.6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.  16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.  Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids. 2003. 122, 107-20. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry. 2006. 45, 5793-9. Wu, Y., He, K., Ludtke, S. J. and Huang, H. W. Biophys J. 1995. 68, 2361-9. Dave, P. C., Billington, E., Pan, Y. L. and Straus, S. K. Biophys J. 2005. 89, 2434-42. Harzer, U. and Bechinger, B. Biochemistry. 2000. 39, 13106-14. Tremouilhac, P., Strandberg, E., Wadhwani, P. and Ulrich, A. S. J Biol Chem. 2006. 281, 32089-32094. Mani, R., Cady, S. D., Tang, M., Waring, A. J., Lehrer, R. I. and Hong, M. Proc atl Acad Sci U S A. 2006. 103, 16242-7. Tang, M., Waring, A. J. and Hong, M. J Am Chem Soc. 2007. 129, 11438-46. Wi, S. and Kim, C. J Phys Chem B. 2008. 112, 11402-14. Patch, J. A. and Barron, A. E. J Am Chem Soc. 2003. 125, 12092-3. Picas, L., Carretero-Genevrier, A., Montero, M. T., Vazquez-Ibar, J. L., Seantier, B., Milhiet, P. E. and Hernandez-Borrell, J. Biochim Biophys Acta. 1798, 1014-1019. Picas, L., Montero, M. T., Morros, A., Vazquez-Ibar, J. L. and Hernandez-Borrell, J. Biochim Biophys Acta. 1798, 291-6. Jung, D., Rozek, A., Okon, M. and Hancock, R. E. W. Chem Biol. 2004. 11, 949-57. Lohner, K., Staudegger, E., Prenner, E. J., Lewis, R. N., Kriechbaum, M., Degovics, G. and McElhaney, R. N. Biochemistry. 1999. 38, 16514-28. Zweytick, D., Pabst, G., Abuja, P. M., Jilek, A., Blondelle, S. E., Andra, J., Jerala, R., Monreal, D., Martinez de Tejada, G. and Lohner, K. Biochim Biophys Acta. 2006. 1758, 1426-35. Thennarasu, S., Lee, D. K., Tan, A., Prasad Kari, U. and Ramamoorthy, A. Biochim Biophys Acta. 2005. 1711, 49-58. Ramamoorthy, A., Marassi, F. M., Zasloff, M. and Opella, S. J. J Biomol MR. 1995. 6, 329-34. Hallock, K. J., Lee, D. K., Omnaas, J., Mosberg, H. I. and Ramamoorthy, A. Biophys J. 2002. 83, 1004-13. Doherty, T., Waring, A. J. and Hong, M. Biochemistry. 2008. 47, 1105-16. Bechinger, B. and Lohner, K. Biochim Biophys Acta. 2006. 1758, 1529-1539. Cheng, J. T., Hale, J. D., Elliot, M., Hancock, R. E. and Straus, S. K. Biophys J. 2009. 96, 552-65. Gregory, S. M., Pokorny, A. and Almeida, P. F. Biophys J. 2009. 96, 116-31. Gupta, S. M., Aranha, C. C., Bellare, J. R. and Reddy, K. V. Contraception. 2009. 80, 299-307. Lee, J., Park, C., Park, S. C., Woo, E. R., Park, Y., Hahm, K. S. and Lee, D. G. J Pept Sci. 2009. 15, 601-6. Ramamoorthy, A., Thennarasu, S., Lee, D. K., Tan, A. and Maloy, L. Biophys J. 2006. 91, 206-16. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett. 2001. 501, 146-50. Yamaguchi, S., Huster, D., Waring, A., Lehrer, R. I., Kearney, W., Tack, B. F. and Hong, M. Biophys J. 2001. 81, 2203-14. 127  CHAPTER 4  28. 29.  30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.  Lu, J. X., Damodaran, K., Blazyk, J. and Lorigan, G. A. Biochemistry. 2005. 44, 1020817. Chekmenev, E. Y., Vollmar, B. S., Forseth, K. T., Manion, M. N., Jones, S. M., Wagner, T. J., Endicott, R. M., Kyriss, B. P., Homem, L. M., Pate, M., He, J., Raines, J., Gor'kov, P. L., Brey, W. W., Mitchell, D. J., Auman, A. J., Ellard-Ivey, M. J., Blazyk, J. and Cotten, M. Biochim Biophys Acta. 2006. 1758, 1359-72. Henzler-Wildman, K. A., Martinez, G. V., Brown, M. F. and Ramamoorthy, A. Biochemistry. 2004. 43, 8459-69. Leidy, C., Linderoth, L., Andresen, T. L., Mouritsen, O. G., Jorgensen, K. and Peters, G. H. Biophys J. 2006. 90, 3165-75. Lewis, R. N., Zhang, Y. P. and McElhaney, R. N. Biochim Biophys Acta. 2005. 1668, 203-14. Mecke, A., Lee, D. K., Ramamoorthy, A., Orr, B. G. and Banaszak Holl, M. M. Biophys J. 2005. 89, 4043-50. Dennison, S. R., Harris, F. and Phoenix, D. A. Protein Pept Lett. 2005. 12, 27-9. Xiong, Y. Q., Mukhopadhyay, K., Yeaman, M. R., Adler-Moore, J. and Bayer, A. S. Antimicrob Agents Chemother. 2005. 49, 3114-21. Domenech, O., Francius, G., Tulkens, P. M., Van Bambeke, F., Dufrene, Y. and Mingeot-Leclercq, M. P. Biochim Biophys Acta. 2009. 1788, 1832-40. Andrushchenko, V. V., Aarabi, M. H., Nguyen, L. T., Prenner, E. J. and Vogel, H. J. Biochim Biophys Acta. 2008. 1778, 1004-14. Thennarasu, S., Lee, D. K., Poon, A., Kawulka, K. E., Vederas, J. C. and Ramamoorthy, A. Chem Phys Lipids. 2005. 137, 38-51. Epand, R. F., Savage, P. B. and Epand, R. M. Biochim Biophys Acta. 2007. 1768, 2500-9. Epand, R. M. and Epand, R. F. Mol Biosyst. 2009. 5, 580-7. Pan, Y. L., Cheng, J. T. J., Hale, J. D., Pan, J., Hancock, R. E. W. and Straus, S. K. Biophys J. 2007. 92, 2854-64. Johnson, W. C. Proteins. 1999. 35, 307-12. Provencher, S. W. and Glockner, J. Biochemistry. 1981. 20, 33-7. Sreerama, N., Venyaminov, S. Y. and Woody, R. W. Protein Sci. 1999. 8, 370-80. Sreerama, N. and Woody, R. W. J Mol Biol. 1994. 242, 497-507. Sreerama, N. and Woody, R. W. Anal Biochem. 1993. 209, 32-44. Arouri, A., Dathe, M. and Blume, A. Biochim Biophys Acta. 2009. 1788, 650-9. Shai, Y. Biochim Biophys Acta. 1999. 1462, 55-70. Papo, N. and Shai, Y. Cell Mol Life Sci. 2005. 62, 784-90. Matsuzaki, K. Biochim Biophys Acta. 1999. 1462, 1-10. Yang, L., Weiss, T. M., Lehrer, R. I. and Huang, H. W. Biophys J. 2000. 79, 2002-9. Huang, H. W. Biochemistry. 2000. 39, 8347-52. Wu, M., Maier, E., Benz, R. and Hancock, R. E. W. Biochemistry. 1999. 38, 7235-42. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother. 1999. 43, 1317-23. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem. 2000. 267, 5330-41. Pukala, T. L., Bowie, J. H., Maselli, V. M., Musgrave, I. F. and Tyler, M. J. at Prod Rep. 2006. 23, 368-93. Frerman, F. E. and White, D. C. J Bacterio. 1967. 94, 1868-74. Joyce, G. H., Hammond, R. K. and White, D. C. J Bacterio. 1970. 104, 323-30. Appelbaum, P. C. and Bozdogan, B. Clin Lab Med. 2004. 24, 381-402. Bozdogan, B., Esel, D., Whitener, C., Browne, F. A. and Appelbaum, P. C. J Antimicrob Chemother. 2003. 52, 864-8. 128  CHAPTER 4  61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.  Conrad, R. S. and Gilleland, H. E., Jr. J Bacteriol. 1981. 148, 487-97. Clejan, S., Krulwich, T. A., Mondrus, K. R. and Seto-Young, D. J Bacteriol. 1986. 168, 334-40. Beining, P. R., Huff, E., Prescott, B. and Theodore, T. S. J Bacteriol. 1975. 121, 137-43. Trombe, M. C., Laneelle, M. A. and Laneelle, G. Biochim Biophys Acta. 1979. 574, 290300. Haque, M. A. and Russell, N. J. Microbiology. 2004. 150, 1397-404. Morein, S., Andersson, A., Rilfors, L. and Lindblom, G. J Biol Chem. 1996. 271, 6801-9. Lewis, R. N. and McElhaney, R. N. Biochim Biophys Acta. 2009. 1788, 2069-79. Fleming, B. D. and Keough, K. M. Can J Biochem Cell Biol. 1983. 61, 882-91. Zasloff, M. Engl J Med. 2002. 347, 1199-200. Hancock, R. E. W. Lancet Infect Dis. 2001. 1, 156-64. Fleming, E., Maharaj, N. P., Chen, J. L., Nelson, R. B. and Elmore, D. E. Proteins. 2008. 73, 480-91. Tang, M. and Hong, M. Mol Biosyst. 2009. 5, 317-22. Mihajlovic, M. and Lazaridis, T. Biochim Biophys Acta. Papo, N. and Shai, Y. Biochemistry. 2003. 42, 458-66. Matsuzaki, K., Mitani, Y., Akada, K. Y., Murase, O., Yoneyama, S., Zasloff, M. and Miyajima, K. Biochemistry. 1998. 37, 15144-53. Matsuzaki, K., Murase, O., Fujii, N. and Miyajima, K. Biochemistry. 1996. 35, 11361-8. Hallock, K. J., Lee, D. K. and Ramamoorthy, A. Biophys J. 2003. 84, 3052-60. Gottler, L. M. and Ramamoorthy, A. Biochim Biophys Acta. 2009. 1788, 1680-6. Lu, J. X., Blazyk, J. and Lorigan, G. A. Biochim Biophys Acta. 2006. 1758, 1303-13. Wu, M. and Hancock, R. E. J Biol Chem. 1999. 274, 29-35.  129  CHAPTER 5  CHAPTER 5: The importance of residue 13 and the C-terminus on the structure and activity of the amphibian antimicrobial peptide aurein 2.2†  5.1 Introduction We have demonstrated the structure and peptide-membrane interaction of aurein 2.2 and aurein 2.3 in Chapter 2 and the dependence of their behaviour on lipid composition in Chapter 3 and Chapter 4. To elucidate a more comprehensive understanding of the mechanism of action of aurein 2.2 and aurein 2.3, and to better understand how a subtle difference in sequence at residue 13 may have an impact on mode of action, peptides with mutations at position 13 and peptides with C-terminal truncations were investigated and will be discussed in this chapter. Although a large number of peptides would be needed for a complete and comprehensive study, we have used prior knowledge of the effect of mutations on other peptides (1-5) and limited our focus to five peptides. The first three peptides consisted of a single conservative mutation at residue 13 from one hydrophobic residue (L) to another. The peptides with L13A, L13F, and L13V were chosen specifically to probe the effect of steric hindrance (L13A mutation), aromatic ring-lipid interaction (L13F mutation) and hydrophobicity (L13V mutation) on structure and activity. The last two peptides consisted of the native aurein 2.2 sequence, with the last three (aurein 2.2-∆3) or six (aurein 2.2-∆6) residues missing. In aurein 2.2-∆3, residue 13 became the last residue in sequence, and in aurein 2.2-∆6, residue 13 was completely removed. Since most of the peptides †  A version of this chapter has been published as follows:  Cheng, J. T. J., Hale, J. D., Kindrachuk, J., Jenssen, H., Elliot, M., Hancock, R. E. W. and Straus, S. K. 2010. The Importance of Residue 13 and the C-terminus on the Structure and Activity of the Antimicrobial Peptide Aurein 2.2. Biophys. J. (Accepted for publication on August 30, 2010. D.O.I. is currently unavailable). 130  CHAPTER 5  in the aurein family have conserved N-termini, we anticipated that a modest C-terminal truncation might have little effect on activity.  Table 5.1 lists the peptide sequences and  molecular weights of aurein 2.2, aurein 2.3 and the five derived aurein mutant peptides.  Table 5.1.  Peptide sequences and molecular weights of the two aurein parent peptides and the five aurein mutant peptides.  Peptide Aurein 2.2 Aurein 2.3 L13A L13F L13V Aurein 2.2-∆3 Aurein 2.2-∆6  Sequence GLFDIVKKVVGALGSL GLFDIVKKVVGAIGSL GLFDIVKKVVGAAGSL GLFDIVKKVVGAFGSL GLFDIVKKVVGAVGSL GLFDIVKKVVGAL GLFDIVKKVV  MW (g/mol) 1614.95 1614.95 1572.90 1648.99 1600.95 1357.70 1116.40  To determine the mode of action of the five peptides mentioned above, their structure and activities were determined. Solution CD spectroscopy was used to examine whether these peptides have different α-helical content in the membrane environment. To assess how these mutants interact with the lipid bilayers, OCD and  31  P NMR spectroscopy were used. 3,3'-  dipropylthiadicarbocyanine iodide (DiSC35) assays (using S. aureus C622) were performed to examine whether changes in sequence affect the ability of peptides to disrupt intact bacterial membranes. Finally the minimal inhibitory concentration (MIC) of each peptide was determined against S. aureus strain C622 and S. epidermidis strain C621 to correlate observations made in model membranes to the behavior of these aurein peptides in the presence of intact bacteria. Overall, these data have allowed us to determine the role of specific residues in structure and activity, and to better understand how sequence variation modulates structure-function relationships. In Chapter 4, we showed that the three aurein parent peptides behaved differently in different model bacterial membranes and the behaviour was similar in POPC/POPG and CL/POPG, but not in POPE/POPG. Whether the effects of the modification on residue 13 and 131  CHAPTER 5  C-terminus on these derived mutant peptides are still consistent or different in POPE/POPG membrane should also be examined and verified. As in Chapter 4, we have also included our mutant study results in 1:1 POPE/POPG (mol/mol) membrane in this chapter for consistent comparison with the results in 1:1 POPC/POPG (mol/mol) membrane. This chapter will outline the detailed studies on the structure-function relationship of the five aurein mutants in S. aureus and S. aureus-mimicking model membranes, 1:1 POPC/POPG (mol/mol), as well as 1:1 POPE/POPG (mol/mol) for comparison. This chapter will be divided into three sections. Section 5.2 will be divided into five subsections. Section 5.2.1 and Section 5.2.5 sum up the activity assay results of the aurein mutants in real bacteria (S. aureus). Section 5.2.2, Section 5.2.3 and Section 5.2.4 give the structural and functional analyses on the five aurein mutants in model bacterial membranes, respectively. Section 5.3 provides the detailed explanation on how these peptides interact with and perturb the lipid bilayers differently, and whether structural differences play a role in the peptide-membrane interaction. Section 5.4 reviews the findings and provides the conclusion on this study.  5.2 Results Determining the antimicrobial activity of newly designed mutant peptides is essentially the first step to establish whether a modification could enhance or weaken the antimicrobial function of the given mutant peptide. Section 5.2.1 will give a brief overview of the MIC results of the five aurein mutants in various bacteria and whether they were active or less active against these bacterial strains.  132  CHAPTER 5  5.2.1 Antimicrobial activity of aurein mutants MICs of the five aurein mutants against two Gram-positive bacteria (S. aureus and S. epidermidis) were determined. Table 5.2 reports the MIC assay results of the five aurein mutants  in S. aureus (strain C622) and S. epidermidis (strain C621).  Table 5.2.  Minimal inhibitory concentrations (MICs) in µg/ml of the five aurein mutants toward S. aureus and S. epidermidis. MICs are given as the most frequently observed value obtained from repeat experiments. (MIC assay results are courtesy of Dr. J. D. Hale).  Peptide Aurein 2.2-CONH2 L13A L13F L13V Aurein 2.2-∆3 Aurein 2.2-∆6  S. aureus (C622) 16 16 16 16 32 128  S. epidermidis (C621) 8 8 16 8 8 128  The MIC values indicate that all residue 13-substituted mutants had very similar activities to the parent peptide, aurein 2.2 (6), against both Staphylococcus strains. These mutants had simliar MICs of 16 µg/ml against the wild-type S. aureus strain C622. Aurein 2.2-∆3 still showed a considerable activity (MIC = 32 µg/ml), whereas aurein 2.2-∆6 had an MIC of 128 µg/ml which makes it inactive or much less active (7,8).  However, MICs of residue 13-  substituted mutants against the wild-type S. epidermidis strain C621 were slightly different. L13A and L13V showed similar MICs of 8 µg/ml as aurein 2.2, whereas L13F had a slightly higher MIC of 16 µg/ml. Likewise, MICs of C-terminally truncated mutants were different. Aurein 2.2-∆3 had an identical MIC of 8 µg/ml as residue 13-substituted mutants, whereas aurein 2.2-∆6 again showed little or no activity with an MIC of 128 µg/ml. Wells containing culture only and broth only were used as controls. Since MICs of aurein mutants were very similar except aurein 2.2-∆6, whether this difference is caused by a difference in structure or membrane interaction will be outlined in Section 5.2.2 and Section 5.2.3/Section 5.2.4, respectively. 133  CHAPTER 5  5.2.2 Secondary structure of aurein mutants Once the MIC of a derived antimicrobial peptide mutant is determined, characterizing the structure is fundamentally the first step toward understanding the structure-function relationship of the given mutant peptide. By using solution CD spectroscopy, we were able to find the conformational changes of the five aurein mutants in model membranes. This section will  80  (d)  60 40 20 0 -20 -40 190  210  230  250  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (a)  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  describe the structural studies of the five aurein mutants in various lipid membranes.  80 60 40 20 0 -20 -40 190  (e)  60 40 20 0 -20 -40 190  210  230  250  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  Wavelength (nm) (c)  230  250  Wavelength (nm)  80  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  Wavelength (nm) (b)  210  80 60 40 20 0 -20 -40 190  210  230  250  Wavelength (nm)  80 60 40 20 0 -20 -40 190  210  230  250  Wavelength (nm)  Figure 5.1.  Solution CD spectra of aurein mutants in 1:1 POPC/POPG (mol/mol) SUVs: (a) L13A; (b) L13F; (c) L13V; (d) aurein 2.2-∆3; and (e) aurein 2.2-∆6 (solid black line, P/L = 1:15; dashed black line, P/L = 1:50; dotted black line, P/L = 1:100). Spectra indicate that aurein mutants adopted an α-helical conformation in the presence of POPC/POPG SUVs.  Our previous studies discussed in Chapter 3 have demonstrated that 1:1 POPC/POPG (mol/mol) membranes may be the best representative model for S. aureus. In order to ascertain  134  CHAPTER 5  that aurein mutants remain structured and uncover whether they experience different structural changes in the presence of different lipid membranes, solution CD experiments were performed in 1:1 POPC/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) small unilamellar vesicles (SUVs). Figure 5.1 and Figure 5.2 show solution CD spectra of the five aurein mutants in the presence of 1:1 POPC/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) SUVs as a function of P/L molar ratio, respectively.  (d)  80 60 40 20 0 -20 -40 -60 190  210  230  250  100  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  100  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (a)  80 60 40 20 0 -20 -40 -60 190  (e)  80 60 40 20 0 -20 -40 -60 190  210  230  250  250  80 60 40 20 0 -20 -40 -60 190  210  230  250  Wavelength (nm)  Wavelength (nm) 100  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (c)  230  100  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  100  M ean R es idue Ellipticity [ 1 0 -3deg.cm2.dmo l-1 ]  (b)  210  Wavelength (nm)  Wavelength (nm)  80 60 40 20 0 -20 -40 -60 190  210  230  250  Wavelength (nm)  Figure 5.2.  Solution CD spectra of aurein mutants in 1:1 POPE/POPG (mol/mol) SUVs: (a) L13A; (b) L13F; (c) L13V; (d) aurein 2.2-∆3; and (e) aurein 2.2-∆6 (solid black line, P/L = 1:15; dashed black line, P/L = 1:50; dotted black line, P/L = 1:100). Spectra indicate that aurein mutants adopted an α-helical conformation in the presence of POPE/POPG SUVs.  All spectra consisted of a maximum at 190 nm and two minima at 207 nm and 222 nm, which are spectral characteristics of peptide adopting α-helical structure. This demonstrates that 135  CHAPTER 5  all five aurein mutants adopted an α-helical conformation in the presence of POPC/POPG and POPE/POPG SUVs, and the peptide structure was independent of the peptide concentration. As previously observed (6), similar intensities were found for all peptide-to-lipid molar ratios (P/L = 1:15, 1:50, and 1:100). This indicates that maximum binding of peptides to lipid vesicles occurred. Saturation would be observed with a combination of signals from both α-helical and random coil structures (9). All spectra were fitted using three different programs (CDSSTR (10), CONTINLL (11), and SELCON3 (12-14)) in order to examine the % helicity of each mutant in POPC/POPG and POPE/POPG SUVs. Figure 5.3 shows the % α-helix and % random coil of the five aurein mutants as a function of P/L molar ratio in the presence of 1:1 POPC/POPG (mol/mol) and 1:1 POPE/POPG (mol/mol) SUVs. The results showed that % α-helix decreased and % random coil increased as P/L molar ratio decreased. This indicates that decreased % α-helix contributed to increased % random coil when peptide concentration decreased. The results demonstrated that all five aurein mutants adopted close to 100% α-helical conformation at high peptide concentration (1:15 P/L molar ratio) in all lipid membranes. As previously observed (15), helical content increased with increased peptide concentration. This indicates that high concentrations were needed to achieve maximum peptide binding to lipid membranes.  At low peptide  concentrations, however, differences in helical content were observed in the presence of POPC/POPG. At 1:50 P/L molar ratio, L13A, L13V and aurein 2.2-∆3 showed higher helical content, whereas L13F and aurein 2.2-∆6 showed lower helical content. At 1:120 P/L molar ratio, L13V and aurein 2.2-∆3 maintained high % α-helix, whereas L13F and aurein 2.2-∆6 showed significantly reduced helical content. L13A, on the other hand, retained ~ 50% of its helical content. Direct observation on the % α-helix of aurein mutants in the presence of different lipids suggests that aurein mutants showed generally a higher helical content in 136  CHAPTER 5  POPE/POPG than in POPC/POPG and CL/POPG SUVs. Overall, the data showed that the structures of the five aurein mutants were retained but the structural content were dependent of peptide concentration and membrane type, and differences were significant at lower peptide concentrations.  Figure 5.3.  Structural content of aurein mutants in (a) 1:1 POPC/POPG (mol/mol) and (b) 1:1 POPE/POPG (mol/mol) SUVs: % α-helix and % random coil were plotted as a function of P/L molar ratio and aurein mutant; colour legend from white to black: aurein 2.2-∆6, L13F, L13A, L13V, aurein 2.2-∆3. The plots show that % α-helix decreased and % random coil increased as P/L molar ratio decreased. This indicates that decreased % α-helix contributed to increased % random coil at low peptide concentrations. All aurein mutants showed generally higher % α-helix in POPE/POPG than in POPC/POPG SUVs.  137  CHAPTER 5  In the presence of POPC/POPG, both L13F and aurein 2.2-∆6 mutants showed lower helical content at 1:50 P/L molar ratio, but L13F mutant retained its activity whereas aurein 2.2∆6 did not. Whether or not this would affect the mutant’s ability to insert into the lipid bilayers is crucial to understanding the effect of peptide modification on antimicrobial mechanisms. Next section will give a brief overview of the membrane insertion states of the five aurein mutants in POPC/POPG and POPE/POPG bilayers.  5.2.3 Membrane insertion state of aurein mutants Examining the interaction of aurein mutants with model membranes is crucial to elucidating the effect of peptide modification on the extent of peptide insertion into the lipid bilayers. OCD experiments were conducted to investigate the peptide insertion profiles in 1:1 POPC/POPG (mol/mol) bilayers. Differences found in the peptide insertion profiles could reveal whether a specific peptide modification could enhance the peptide-membrane interaction or not. Similarly, same set of experiments were performed in 1:1 POPE/POPG (mol/mol) bilayers to observe whether the modification effects were influenced by different model membrane systems. This section will describe our observations on the peptide-membrane interaction of the five aurein mutants in model membranes, followed by activity assays in real bacteria. Figure 5.4 shows OCD spectra of the five aurein mutants in the presence of 1:1 POPC/POPG (mol/mol) bilayers as a function of P/L molar ratio.  138  CHAPTER 5 30  30  (d)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  (a)  10 0  -10  20 10 0  -10  -20  -20  -30  -30 190  200  210  220  230  240  250  190  200  Wavelength (nm) 30  220  230  240  250  240  250  30  (e)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  (b)  210  Wavelength (nm)  10 0  -10  20 10 0  -10  -20  -20  -30  -30 190  200  210  220  230  240  250  Wavelength (nm)  190  200  210  220  230  Wavelength (nm)  30  Ellipticity (mdeg)  (c)  20 10 0  -10 -20 -30 190  200  210  220  230  240  250  Wavelength (nm)  Figure 5.4.  Oriented CD spectra of aurein mutants in 1:1 POPC/POPG (mol/mol) bilayers: (a) L13A; (b) L13F; (c) L13V; (d) aurein 2.2-∆3; and (e) aurein 2.2-∆6. P/L molar ratios = 1:15 (blue), 1:80 (red), and 1:120 (green). Spectra were normalized such that intensities of all spectra at 222 nm are the same. The spectra show that peptides inserted into 1:1 POPC/POPG (mol/mol) bilayers at threshold P/L molar ratios between 1:15 and 1:80 for L13A and aurein 2.2-∆6, and between 1:80 and 1:120 for L13F, L13V and aurein 2.2-∆3.  The spectra were normalized so that the minimum at 222 nm had the same intensity. The spectra (Figure 5.4(b) ~ (d)) showed that L13F, L13V and aurein 2.2-∆3 inserted (inserted, Istate or tilt, T-state) into POPC/POPG bilayers at threshold P/L molar ratios between 1:80 and 1:120, and became surface-adsorbed (S-state) at P/L molar ratios greater than 1:120. L13A showed less insertion at 1:80 P/L molar ratio and a more gradual insertion profile (Figure 5.4(a)). 139  CHAPTER 5  Aurein 2.2-∆6, on the other hand, was unable to insert into POPC/POPG bilayers (Figure 5.4(e)) at 1:80 P/L molar ratio already. The data illustrate that changing L13 to Phe or Val or removing the last three residues of aurein 2.2 did not change the insertion profile of the peptides, as the parent aurein 2.2 also showed an insertion threshold P/L molar ratio between 1:80 and 1:120 (15). Changing L13 to Ala, however, shifted the insertion profile more towards that of aurein 2.3 (Figure 5.5), which displayed a threshold P/L molar ratio between 1:40 and 1:30 (15). Finally, complete removal of L13 by means of the 6-residue truncation drastically perturbed the insertion profile of the peptide, making the peptide behave more like aurein 2.3-COOH (15). Overall, this suggests that either removing the long hydrophobic leucine (as in aurein 2.2-∆6) or shortening the side-chain (as in L13A) reduces the ability of the peptide to anchor into POPC/POPG membranes. This suggests that residues 11-13 may play an important role in aurein peptide insertion into the lipid bilayers.  Figure 5.5.  Oriented CD spectra of L13A in 1:1 POPC/POPG (mol/mol) bilayers as a function of P/L ratio: P/L = 1:15 (blue), 1:30 (orange), 1:40 (purple), 1:80 (red), and 1:120 (green). normalized such that intensities of all spectra at 222 nm are the same.  Spectra were  The above observations were made in POPC/POPG bilayers. Whether these aurein mutants behave similarly in POPE/POPG bilayers was also investigated. Figure 5.6 shows OCD 140  CHAPTER 5  spectra of the five aurein mutants in the presence of 1:1 POPE/POPG (mol/mol) bilayers as a function of P/L molar ratio.  30  (d)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  (a)  10 0 -10 -20 -30  30 20 10 0 -10 -20 -30  190  200  210  220  230  240  250  190  200  Wavelength (nm) 30  (e)  20  Ellipticity (mdeg)  Ellipticity (mdeg)  (b)  10 0 -10 -20 -30 200  210  220  230  240  250  Wavelength (nm)  Ellipticity (mdeg)  220  230  240  250  240  250  30 20 10 0 -10 -20 -30  190  (c)  210  Wavelength (nm)  190  200  210  220  230  Wavelength (nm)  30 20 10 0 -10 -20 -30 190  200  210  220  230  240  250  Wavelength (nm)  Figure 5.6.  Oriented CD spectra of aurein mutants in 1:1 POPE/POPG (mol/mol) bilayers: (a) L13A; (b) L13F; (c) L13V; (d) aurein 2.2-∆3; and (e) aurein 2.2-∆6. P/L molar ratios = 1:15 (blue), 1:80 (red), and 1:120 (green). Spectra were normalized such that intensities of all spectra at 222 nm are the same. The spectra show that peptides inserted into 1:1 POPE/POPG (mol/mol) bilayers at threshold P/L molar ratios between 1:15 and 1:80 for L13A and aurein 2.2-∆6, and between 1:80 and 1:120 for L13F, L13V and aurein 2.2-∆3.  The spectra show that L13A, L13F, and L13V inserted into POPE/POPG bilayer at threshold P/L molar ratios between 1:15 and 1:80, and became surface-adsorbed (S-state) at P/L 141  CHAPTER 5  molar ratios greater than 1:80. Aurein 2.2-∆3 could not insert at all P/L molar ratios examined. This is opposite to what we observed previously in 1:1 POPC/POPG (mol/mol) bilayers, where aurein 2.2-∆3 inserted at low peptide concentrations, which is consistent to its high activity against S. aureus. Aurein 2.2-∆6, on the other hand, had similar insertion profile as previously in POPC/POPG bilayers, unable to insert at 1:80 P/L molar ratio already. Generally, all aurein mutants showed less insertion, except aurein 2.2-∆6. These data illustrated that different model bacterial membranes could have significant effects on the aurein peptide insertion profiles. Table 5.3 summarizes the membrane insertion states of the five aurein mutants in the presence of different lipid bilayers as a function of P/L molar ratio to illustrate the effect of different model bacterial membranes on the peptide insertion profiles.  Table 5.3.  Lipid bilayers P/L molar ratio 1:15 1:80 1:120  Membrane insertion states of the five aurein mutants in the presence of different lipid bilayers at different P/L molar ratios.  1:1 POPC/POPG (mol/mol) L13A L13F L13V I I/S S  I I S  I I S  Aurein 2.2-∆3 I I S  1:1 POPE/POPG (mol/mol) Aurein 2.2-∆6 I S S  L13A L13F L13V I I/S S  I S S  I S S  Aurein 2.2-∆3 S S S  Aurein 2.2-∆6 I S S  *P/L = peptide-to-lipid; I = inserted state; S = surface-adsorbed state.  5.2.4 Lipid headgroup perturbation by aurein mutants Since aurein mutants demonstrated different insertion profiles in different lipid bilayers examined, it is important to inspect whether difference in peptide insertion profiles also results in difference in lipid headgroup perturbation.  The  31  P NMR experiments were conducted to  determine whether this membrane disruption occurs via a barrel-stave model, carpet model, or toroidal pore model (16-20); a micellar aggregate channel model (21,22); or a detergent-like mechanism (23).  31  P NMR spectra were recorded for all mutants in 4:1 POPC/POPG (mol/mol)  (Figure 5.7 and Figure 5.8) bilayers, as our previous studies described in Chapter 3 (15). 142  CHAPTER 5  Additional  31  P NMR experiments were also conducted for all mutants in 1:1 POPE/POPG  (mol/mol) (Figure 5.7 and Figure 5.8) bilayers for close comparison in different model membrane systems. Figure 5.7 shows 31P NMR spectra of 4:1 POPC/POPG (mol/mol) bilayers in the absence and presence of residue 13-substituted mutants as a function of P/L molar ratio.  (a)  (b)  (c)  (d)  50 ppm (t1)  0  50 ppm (t1)  Figure 5.7.  0  50  0  ppm (t1)  Solid-state 31P MR spectra of mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: (a) POPC/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of L13A (left panel), L13F (centre panel), and L13V (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  Figure 5.8 shows 31P NMR spectra of 4:1 POPC/POPG (mol/mol) bilayers in the absence and presence of C-terminally truncated mutants as a function of P/L molar ratio. 143  CHAPTER 5  (a)  (b)  (c)  (d)  50 ppm (t1)  Figure 5.8.  0  50  0  ppm (t1)  Solid-state 31P MR spectra of mechanically aligned 4:1 POPC/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: (a) POPC/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of aurein 2.2∆3 (left panel) and aurein 2.2-∆6 (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  In the absence of aurein mutants, the spectra consisted primarily of two single resonances at 30 ppm (PC) and 15 ppm (PG), which indicates that the lipid bilayers were aligned with their normal parallel to the magnetic field (Figure 5.7 and Figure 5.8). Note that compared to the spectra shown in Chapter 3 (15), the 31P NMR spectra were better resolved. This has to do with improved alignment achieved by incubating the samples for 8 rather than 7 days, drying the lipids under a stream of N2 rather than air, and resuspending the lipids in a slightly larger volume of water prior to depositing them on the slides. In the presence of aurein mutants, the spectra showed a slightly broadened peak at 30 ppm, indicating increased contribution from unaligned 144  CHAPTER 5 31  P headgroups. In addition, a powder-pattern signal was also observed in the range of -10 ~ 30  ppm, indicative of random headgroup orientations possibly due to unaligned lipid headgroups (Figure 5.7 and Figure 5.8). Different effects on the lipid bilayer headgroups were observed when different mutants were added. When residue 13-substituted mutants were added at high concentrations, the spectra showed enlarged underlying powder pattern (Figure 5.7). At low peptide concentrations, the sizes of the powder pattern were relatively the same. Note that in the presence of L13V at high concentrations, the peak at 30 ppm decreased substantially and the powder pattern increased significantly. This suggests that L13V may be more disruptive to the bilayer headgroup alignment than other mutants, at this concentration.  When C-terminally  truncated mutants were added at high concentrations, the spectra again showed enlarged underlying powder pattern (Figure 5.8). However, at low peptide concentrations, the sizes of the powder pattern decreased as the peptide concentration decreased.  Comparing residue 13-  substituted mutants to C-terminally truncated mutants, at high peptide concentrations, no significant difference was found. At low peptide concentrations, L13A, L13F, L13V and Aurein 2.2-∆3 resulted in greater powder-pattern signals than Aurein 2.2-∆6.  These results were  consistent with our OCD results, where Aurein 2.2-∆6 appeared to be unable to insert into the lipid bilayers at low concentrations. These changes in the spectra occurred for all aurein mutants at all P/L molar ratios examined, suggesting no obvious dependence on peptide sequence and concentration. The underlying powder pattern indicates that these aurein mutants may disorder the bilayer headgroups by formation of toroidal pores (16-20) or toroidal pores/liposomes. The extent of the membrane perturbation was found to be concentration-dependent and peptide-specific, but not for the mechanism.  145  CHAPTER 5  How these aurein mutants perturb the headgroups of POPE/POPG bilayers was also examined. Figure 5.9 shows  31  P NMR spectra of 1:1 POPE/POPG (mol/mol) bilayers in the  absence and presence of residue 13-substituted mutants as a function of P/L molar ratio.  (a)  (b)  (c)  (d) 50 ppm (t1)  Figure 5.9.  0  50 ppm (t1)  0  50 ppm (t1)  0  Solid-state 31P MR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: (a) POPE/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of L13A (left panel), L13F (centre panel), and L13V (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  Figure 5.10 shows  31  P NMR spectra of 1:1 POPE/POPG (mol/mol) bilayers in the  absence and presence of C-terminally truncated mutants as a function of P/L molar ratio. When the five different aurein mutants were added to POPE/POPG bilayers, different effects on the bilayer headgroups were observed. When residue 13-substituted mutants were added at high  146  CHAPTER 5  concentrations, the spectra showed an increased contribution from an underlying powder pattern. At low peptide concentrations, the sizes of the powder pattern were relatively similar. Note that in the presence of L13V mutant at high concentrations, the powder pattern increased significantly. This suggests that at this concentration, L13V mutant may be more disruptive to the bilayer headgroup alignment than other mutants, which was also observed in POPC/POPG bilayers.  (a)  (b)  (c)  (d) 50 ppm (t1)  0  50 ppm (t1)  0  Figure 5.10. Solid-state 31P MR spectra of mechanically aligned 1:1 POPE/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: (a) POPE/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of aurein 2.2∆3 (left panel) and aurein 2.2-∆6 (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  When C-terminally truncated mutants were added at high concentrations, the spectra again showed enlarged underlying powder pattern. At low peptide concentrations, the sizes of 147  CHAPTER 5  the powder pattern decreased as aurein 2.2-∆3 concentration decreased, and stayed relatively the same for aurein 2.2-∆6. Comparing the two C-terminally truncated mutants, aurein 2.2-∆3 resulted in a smaller powder-pattern signal than aurein 2.2-∆6 at all peptide concentrations. Comparing residue 13-substituted mutants to C-terminally truncated mutants, no significant difference was found at all peptide concentrations, except that aurein 2.2-∆3 resulted in a slightly smaller underlying powder pattern than other mutants. These results are consistent with our OCD results, where aurein 2.2-∆3 was less able to insert fully into the lipid bilayers at all concentrations. Overall, the trends observed for the mutant peptides in POPE/POPG were reminiscent of those seen in POPC/POPG bilayers. This suggests perhaps that membrane perturbation was mainly driven by electrostatic interactions between the positively charged peptides and the negatively charged PG headgroups, and was less dependent on the composition of the partner lipid, i.e. be it PC or PE (24). Taken together, all the presence of powder patterns in the 31P NMR data suggests that all aurein mutants disordered the bilayer headgroups via mechanisms similar to toroidal pore (1620) or toroidal pore/liposome formation, be that in POPC/POPG, CL/POPG, or POPE/POPG membranes. The extent of the membrane perturbation was found to be strongly concentrationdependent and partly dependent on the nature of the peptide, but not for the mechanism.  5.2.5 Using DiSC35 assay to observe the bacterial membrane leakage induced by aurein mutants Since cytoplasmic membrane is a common target site for many antimicrobial peptides, performing membrane leakage assay on these aurein mutants in real bacteria can reveal whether these mutant peptides kill bacteria through membrane depolarization and to what extent these mutant peptides can induce membrane leakage. To observe what effects aurein mutants have on the cytoplasmic membrane of S. aureus C622, we carried out membrane depolarization 148  CHAPTER 5  experiments using the membrane sensitive dye DiSC35. Since aurein mutants did not have pronounced disordering effect on POPE/POPG membranes, it was expected to observe minimal membrane disruption from membrane leakage assays in live bacteria containing PE/PG as the major membrane lipids. All aurein mutants were compared to the membrane acting cyclic peptide gramicidin S (GMS). Figure 5.11 shows the DiSC35 assay results of the five aurein mutants in S. aureus strain C622 at 1X and 5X MIC.  Figure 5.11. Membrane depolarization of S. aureus C622 induced by aurein mutants at 1X and 5X MIC. Gramicidin S (1X MIC) was used as control. Results are representative of 3 experiments. DiSC35 release profile of aurein mutants as a function of time at (a) 1X and (b) 5X MIC; (c) % membrane depolarization induced by aurein mutants at 300 sec after peptide addition with respect to Gramicidin S as the control; (d) membrane depolarization efficiency at 300 sec after peptide addition at 5X MIC with respect to 1X MIC. [Legend for Figure 5.11(a) and (b): blue square = L13A, black diamond = L13F, green triangle = L13V, orange cross = aurein 2.2-∆3, red circle = aurein 2.2-∆6, purple cross = GMS]. 149  CHAPTER 5  The DiSC35 assay data showed that at 1X MIC (Figure 5.11(a)), both L13F and aurein 2.2-∆3 demonstrated greater membrane depolarization than the other peptides. L13A and L13V both showed a similar depolarization effect, whereas aurein 2.2-∆6 was much less efficient than any other peptides. At 5X MIC (Figure 5.11(b)), all aurein mutants showed increased DiSC35 release, but with different magnitude. Both L13A and aurein 2.2-∆6 showed a much greater fluorescence than other mutants.  Figure 5.11(c) illustrates the percentage of membrane  depolarization induced by aurein mutants with respect to gramicidin S at 300 seconds after the peptide addition. The efficiency of membrane depolarization of the five aurein mutants in the increasing order is: aurein 2.2-∆6, L13A/L13V, and L13F/aurein 2.2-∆3. Increasing the level of aurein peptide added to 5X MIC showed an increased depolarization effect for L13A and aurein 2.2-∆6, but not for other mutants (Figure 5.11(d)). This suggests that most mutants were at their maximum depolarization efficiency already at 1X MIC, except for L13A and aurein 2.2-∆6.  5.2.6 Antimicrobial activity of aurein mutants against bacteria containing PE Given that all five aurein mutants studied here adopted α-helical structure but interacted with different model bacterial membranes very differently, MICs of the five aurein mutants were tested further against bacteria which has relatively equivalent % PE and % PG as its major membrane lipids, for example B. cereus, to correlate the results obtained in 1:1 POPE/POPG (mol/mol) membrane with the activity in real bacteria. MICs of the five aurein mutants were also determined against two other bacteria, P. aeruginosa (~ 60% PE) and E. coli (~ 80% PE), containing higher % PE as their major membrane lipids. These results would give an idea how active these aurein mutants are against real bacteria containing PE as the main membrane lipid, to correlate with results obtained in POPE/POPG membranes.  150  CHAPTER 5  MIC values indicate that all aurein mutants had very low to no activities against B. cereus strain C737, in contrast to our observation in S. aureus C622 (6). The mutants had MICs of 128 µg/ml against B. cereus C737, which classifies them as inactive or much less active (7,8), except that aurein 2.2-∆3 showed very little activity (MIC = 64 µg/ml). The MIC results suggest that the aurein peptides were ineffective against bacteria with high PE content in their membranes. In addition, the MIC results correlate somewhat with the oriented CD and 31P NMR results obtained in POPE/POPG model membranes.  Table 5.4.  Minimal inhibitory concentrations (MICs) in µg/ml of the five aurein mutants and gramicidin S (control) toward B. cereus, P. aeruginosa and E. coli. MICs are given as the most frequently observed value obtained from repeat experiments. (MIC values obtained in P. aeruginosa Pa01 and E. coli C500 are courtesy of Dr. J. D. Hale and M. Elliot).  Peptide Aurein 2.2-CONH2 L13A L13F L13V Aurein 2.2-∆3 Aurein 2.2-∆6 GMS  B. cereus (C737) 128 128 128 128 64 128 2  P. aeruginosa (Pa01) > 128 > 128 > 128 > 128 > 128 > 128   E. coli (C500) 64 128 64 64 32 128   *GMS = gramicidin S;  = data not available.  5.3 Discussion Determining which residue is crucial to antimicrobial activity can be a key step in understanding the mechanism of action of a given antimicrobial peptide. In this study, five mutant peptides from aurein 2.2, an antimicrobial peptide from the Australian Southern Bell frog Litoria aurea, were investigated to examine how residue-13 substitutions and C-terminal  truncations had an effect on the structure, membrane interaction, and activity. Our previous studies demonstrated that aurein 2.2 and aurein 2.3 showed similar antimicrobial activity (6), but perturbed membranes to a different extent (15). Aurein 2.2 and  151  CHAPTER 5  aurein 2.3 differ by only one residue at position 13. Residue specificity at this position may, therefore, be a key to the difference in the peptide-lipid interactions. When examining a specific residue, there are several factors to consider. First, steric hindrance usually plays an important role in peptide/protein activities or folding patterns (25-30). Second, aromatic residues, especially tryptophan, have been shown to promote peptide insertion into the lipid bilayers (1,31). Lastly, hydrophobicity is widely accepted as a key factor for achieving high antimicrobial activity (32,33). Here, we substituted Leu13 with Ala13 to test the effect of steric hindrance, Phe13 to test the importance of aromatic residues and Val13 to test the effect of hydrophobic chain length on the structure-function relationship of the aurein 2 family peptides. Our solution CD results showed that the three aurein mutants L13A, L13F and L13V adopted α-helical structure in the presence of 1:1 POPC/POPG membranes. Several studies have also demonstrated that mutated antimicrobial peptides either adopt similar structures (e.g. human β defensin-1, bacteriocin and dermcidin (34-37)), similar structures but with varying structural  content (e.g. mesentericin Y105 and MSI-78/LL37 analogs (32,38)), or entirely different conformations (e.g. truncated human β defensin-3 (39)). The extent of the change in structure depends on the nature and the position of the amino acid which is substituted. We showed here that for relatively conservative substitutions of the aurein peptides, the overall structure was preserved but the helical contents of the three mutants were different, particularly at low peptide concentrations. This indicates that secondary structure is dependent on the nature of the amino acid at residue 13, analagous to what was observed in studies on a mesentericin Y105 mutant (38). Comparing the helical content at a P/L ratio of 1:100, the results showed that L13V was the most helical, followed by aurein 2.2 and aurein 2.3, then L13A, and finally L13F. This trend does not parallel the hydrophobicity trend in the Kyte and Doolittle scale (40), whereby Ile is more hydrophobic than Val, followed by Leu, Phe, and finally Ala, or in any other available  152  CHAPTER 5  hydrophobicity scale (e.g. (41) or (42)). This indicates that hydrophobicity alone cannot account for structural differences, but that other factors, such as the bulkiness of the side-chain, its ability and interactions with neighboring residues, must play a structural role. Hydrophobicity alone does not account for how the peptides interacted with POPC/POPG membranes either. The OCD results suggested that L13F, L13V, and aurein 2.2 were the most effective at inserting into POPC/POPG bilayers, with threshold P/L ratios being between 1:80 and 1:120. L13A, on the other hand, was slightly less effective, and aurein 2.3, with an Ile at position 13, was the least effective. Insertion of all the peptides was accompanied by perturbations similar to toroidal pore or toroidal pore/liposome formation, as seen in the  31  P  NMR results here and the results previously reported for aurein 2.2 and 2.3 (15). At low peptide concentrations, the extent of membrane disruption, as indicated by the  31  P signals from  unoriented headgroups, was similar in the presence of all three mutants. Interestingly however, at high peptide concentrations (1:15 P/L ratio) L13V significantly perturbed the membrane bilayers. Perhaps the high helical content of L13V as seen from the CD data could account for this observation. The results with the L13A, L13F, and L13V mutants demonstrated that overall it is difficult to clearly correlate subtle changes in sequence with subtle changes in structure and membrane interaction. What the data suggest in broad terms, however, is that structure and membrane interaction of the aurein peptides correlate reasonably well with MIC and membrane depolarization. Indeed, aurein 2.2 and L13V, which were the most helical and effective at inserting into and perturbing POPC/POPG bilayers, were effective antimicrobial peptides and depolarized bacterial membranes well. On the other hand, aurein 2.3, which did not readily insert into POPC/POPG bilayers (as seen from the OCD data), also did not depolarize membranes well (15) and was not as good an antimicrobial peptide as aurein 2.2 (6). Of course, comparing data  153  CHAPTER 5  obtained from a bacterial membrane model system to that obtained in bacteria is fraught with danger, but to gain insight into how the aurein peptides function, models are a necessary evil. Since the residue-13 substitutions studied here did not significantly influence the structure and peptide-membrane interactions, the more dramatic effect of truncating the Cterminus of aurein 2.2 was also investigated. Two C-terminally truncated mutants were examined: one with the last three residues in aurein 2.2 removed (loss of G14S15L16) and the other with the last six residues in aurein 2.2 removed (loss of G11A12L13G14S15L16). Both of these deletion mutants retained a predominantly α-helical structure, as seen in the solution CD results (Figure 5.1). However, aurein 2.2-∆6 showed a much lower helical content at all P/L ratios. The extensively truncated aurein 2.2-∆6 may be too short to form an ordered conformation, as observed for the bioactive gastrin-17 analogs (43) and the (KIGAKI)n peptide (44). A CD study of magainin-2 also demonstrated that a minimum of 12 residues (a helical length of 24 ~ 34Å) is required for antimicrobial activity (45). Both the decreased the helical content and the increased random coil content would contribute to the inability of the peptide to perturb lipid membranes greatly (44). Previous studies on cationic antimicrobial peptides have suggested that positively charged residues (e.g. Lys and Arg) are important for activity, whereas negatively charged or neutral amino acid (e.g. Asp and Ser) are refractory (46). Aurein 2.2 has two Lys residues located at position 7 and 8, and a Ser at position 15. Truncating the first three C-terminal residues removed Ser15. Our MIC and OCD results showed that this truncation did not have a considerable effect on the antimicrobial activity and the insertion profile of the peptide (Aurein 2.2-∆3). This suggests that the peptide segment “G14S15L16” is not necessary for antimicrobial activity. Indeed, Aurein 2.2-∆3 has identical activity to its parent, aurein 2.2. Removal of a larger segment of aurein 2.2, on the other hand, had a significant effect on the antimicrobial activity and  154  CHAPTER 5  insertion profile of the peptide, as seen from our MIC and OCD results. This implies that the peptide segment “G11A12L13” is crucial for antimicrobial activity and membrane interactions. The GAL segment may be needed as a longer C-terminal hydrophobic segment for a better peptidemembrane interactions or may offer more effective hydrophobic side-chain packing, as proposed for PMAP-23 (a cathelicidin) (47) and aurein 1.2 (48). It is also possible, however, that aurein 2.2-∆6 is simply of insufficient peptide length (~ 15 Å) to span the thicker POPC/POPG bilayers (~ 39 Å) (49), as was observed for the short-chain peptide ampullosporin A (50). This would explain why very high concentrations of aurein 2.2-∆6 are needed for insertion to occur at all (or carpet-like mechanism?). Once inserted, both aurein 2.2-∆3 and aurein 2.2-∆6 perturb the lipids by possibly forming (distorted) toroidal pores or toroidal pores/liposomes. Investigating the behaviour of aurein mutants in POPE/POPG bilayers would give a better insight into how aurein mutants interact with different model bacterial membranes, and whether these interactions are subject to differences in membrane composition. According to the MIC results, the aurein peptides were not very active against bacteria that contain 40% ~ 60% PE, such as B. cereus and P. aeruginosa, respectively. The OCD and  31  P NMR data correlate  somewhat with this low activity, as all aurein mutants were found to insert into and perturb the headgroups of PE/PG membranes to a lesser extent than PC/PG. The high helicity found in POPE/POPG suggests that secondary structure cannot be used as a reference for activity. For completeness, the MIC of each aurein peptide was determined for E.coli (80% PE), as well as the effect of the peptides on 4:1 POPE/POPG headgroups using  31  P NMR on aligned bilayers  (Figure 5.12 and Figure 5.13, Supplemental figures). Interestingly, the data here suggest that some of the aurein peptides had some minimal amount of activity, yet did not perturb the lipid headgroups as well as when the peptides were in 1:1 POPE/POPG. Therefore, in this case, one would have to conclude that a membrane model of 4:1 POPE/POPG (or even 1:1 POPE/POPG) is inadequate for determining the mode of action of the aurein peptides that show some minimal 155  CHAPTER 5  activity against E. coli. Finally, the MIC data determined previously against S. aureus (6) suggest that the aurein peptides are effective at killing these bacteria and that 1:1 POPC/POPG or 1:1 CL/POPG bilayers may be a good choice as model membranes to represent this bacteria. These observations are consistent with those stated in Chapter 4.  5.4 Summary and conclusion Overall, the results presented here demonstrate the importance of residue 13 for the structure and function of aurein 2.2.  Removal of L13 has a dramatic effect on structure,  membrane interaction, toroidal pore or toroidal pore/liposome formation, MIC and membrane depolarization. This clearly indicates that the presence of residue 13 is critical for activity. The exact nature of residue 13, however, is less important, as long as it is a hydrophobic residue. Indeed, mutation of L13 in aurein 2.2 to Ala, Phe, Val or Ile (as in aurein 2.3) results in antimicrobial peptides which display similar activities.  Presumably, Nature has a built-in  tolerance to the exact type of amino acid residue at a given position in the cationic antimicrobial peptide, as long as its essential character (e.g. hydrophobicity (51)) is preserved. The fact that the activities are similar, but that the extent of membrane depolarization that the L13 mutants display is different, further suggests that Nature has a tolerance in the exact mechanism that cationic antimicrobial peptides use to kill bacteria (52).  156  CHAPTER 5  5.5 Supplemental figures  (a)  (b)  (c)  (d) 50 ppm (t1)  0  50 ppm (t1)  0  50 ppm (t1)  0  Figure 5.12. Solid-state 31P MR spectra of mechanically aligned 4:1 POPE/POPG (mol/mol) bilayers with and without three residue 13-substituted aurein mutants: (a) POPE/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of L13A (left panel), L13F (centre panel), and L13V (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  157  CHAPTER 5  (a)  (b)  (c)  (d) 50 ppm (t1)  0  50 ppm (t1)  0  Figure 5.13. Solid-state 31P MR spectra of mechanically aligned 4:1 POPE/POPG (mol/mol) bilayers with and without two C-terminally truncated aurein mutants: (a) POPE/POPG bilayers alone, P/L = (b) 1:120, (c) 1:80, and (d) 1:15 in the presence of aurein 2.2∆3 (left panel) and aurein 2.2-∆6 (right panel). Spectra were recorded using 2048 scans at 30°C, oriented such that the bilayer normal was parallel to the external magnetic field. Spectra were processed with 50 Hz line-broadening.  5.6 Materials and methods 5.6.1 Materials Peptide synthesis materials were purchased as previously described in Chapter 2. Mylar plates were made by cutting Melinex Teijin films from Dupont (Wilton, Middlesbrough, UK). 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3[phospho-rac-(1-glycerol)] (POPG) were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and obtained dissolved in chloroform. 3,3'-dipropylthiadicarbocyanine iodide (DiSC35) was purchased from Sigma-Aldrich (St. Louis, MO, USA). 158  CHAPTER 5  5.6.2 Methods 5.6.2.1 Peptide synthesis Aurein 2.2 mutants were synthesized as previously described (6,15), using an CS Bio Co. peptide synthesizer (Menlo Park, CA, USA) by in situ neutralization Fmoc chemistry, using Rink or Wang resin with the Leu double-coupling scheme, as appropriate.  5.6.2.2 Purification The crude peptide product was purified by preparative RP-HPLC on a Waters 600 system (Waters Limited, Mississauga, ON, Canada) with 229 nm UV detection using a Phenomenex (Torrance, CA, USA) C4 preparative column (20.0 µm, 2.1 cm × 25.0 cm) as previously reported (6,15). The identity of the products was verified using electrospray ionization (ESI) mass spectrometry and MALDI-TOF as previously mentioned (6,15) and confirmed to be ≥ 99% pure. The molecular weights of the five aurein mutant peptides can be found in Table 5.1.  5.6.2.3 Solution circular dichroism (CD) sample preparation Solution CD samples with a constant peptide concentration of 2.0 mM were prepared in different peptide to lipid (P/L) molar ratios of 1:15, 1:50, and 1:100, using 1:1 molar lipid mixtures of POPC/POPG or POPE/POPG, as previously informed in Chapter 4 and in (6,15).  5.6.2.4 Mechanically oriented sample preparation Solid-state NMR samples were prepared and plated for three different P/L molar ratios of 1:15, 1:80 and 1:120, and incubated at 37ºC for 8 days (POPC/POPG or POPE/POPG), as previously described (6,15). The humidity of the samples, degree of alignment, and consistent sample preparation were verified as previously reported (6,15). Finally, the plated samples were 159  CHAPTER 5  wrapped in a thin layer of parafilm and placed in plastic sheathing before data acquisition. Oriented CD samples were prepared and deposited in a similar fashion as mentioned above.  5.6.2.5 Circular dichroism Solution and oriented CD experiments were carried out using a JASCO J-810 spectropolarimeter (Victoria, BC, Canada) at 30oC as previously described (6,15).  5.6.2.6 MR spectroscopy Solid-state  31  P NMR experiments on mechanically aligned lipid bilayer samples were  carried out on the Bruker 500-MHz NMR spectrometer at 30°C, operating at a phosphorus frequency of 202.40 MHz with proton decoupling as previously reported (6,15). The 90o pulse was set to 11.25 µs, and a 3 second recycle delay was used (POPC/POPG and POPE/POPG). The spectra were acquired using 2048 scans and processed with 50 Hz line broadening.  5.6.2.7 DiSC35 assay The ability of the aurein mutants to depolarize the cytoplasmic membrane of S. aureus C622  was  determined  using  the  membrane  potential  sensitive  dye  3,3'-  dipropylthiadicarbocyanine iodide (DiSC35), as previously reported in Chapter 3 (15).  5.6.2.8 MIC determination Minimal Inhibitory Concentrations (MIC) for L13A, L13F, L13V, aurein 2.2-∆3, and aurein 2.2-∆6 were determined against various bacterial strains based on the previously described modified methodology (53) as in Chapter 2 and Chapter 4.  160  CHAPTER 5  5.7 References 1. 2. 3. 4.  5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.  29. 30.  Dathe, M., Nikolenko, H., Klose, J. and Bienert, M. Biochemistry. 2004. 43, 9140-50. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett. 2001. 501, 146-50. Hilpert, K., Elliott, M. R., Volkmer-Engert, R., Henklein, P., Donini, O., Zhou, Q., Winkler, D. F. and Hancock, R. E. Chem Biol. 2006. 13, 1101-7. Lewis, R. N., Liu, F., Krivanek, R., Rybar, P., Hianik, T., Flach, C. R., Mendelsohn, R., Chen, Y., Mant, C. T., Hodges, R. S. and McElhaney, R. N. Biochemistry. 2007. 46, 1042-54. Wieprecht, T., Dathe, M., Krause, E., Beyermann, M., Maloy, W. L., MacDonald, D. L. and Bienert, M. FEBS Lett. 1997. 417, 135-40. Pan, Y. L., Cheng, J. T. J., Hale, J. D., Pan, J., Hancock, R. E. W. and Straus, S. K. Biophys J. 2007. 92, 2854-64. Pukala, T. L., Bowie, J. H., Maselli, V. M., Musgrave, I. F. and Tyler, M. J. at Prod Rep. 2006. 23, 368-93. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem. 2000. 267, 5330-41. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids. 2003. 122, 107-20. Johnson, W. C. Proteins. 1999. 35, 307-12. Provencher, S. W. and Glockner, J. Biochemistry. 1981. 20, 33-7. Sreerama, N., Venyaminov, S. Y. and Woody, R. W. Protein Sci. 1999. 8, 370-80. Sreerama, N. and Woody, R. W. J Mol Biol. 1994. 242, 497-507. Sreerama, N. and Woody, R. W. Anal Biochem. 1993. 209, 32-44. Cheng, J. T., Hale, J. D., Elliot, M., Hancock, R. E. and Straus, S. K. Biophys J. 2009. 96, 552-65. Shai, Y. Biochim Biophys Acta. 1999. 1462, 55-70. Papo, N. and Shai, Y. Cell Mol Life Sci. 2005. 62, 784-90. Matsuzaki, K. Biochim Biophys Acta. 1999. 1462, 1-10. Yang, L., Weiss, T. M., Lehrer, R. I. and Huang, H. W. Biophys J. 2000. 79, 2002-9. Huang, H. W. Biochemistry. 2000. 39, 8347-52. Wu, M., Maier, E., Benz, R. and Hancock, R. E. W. Biochemistry. 1999. 38, 7235-42. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother. 1999. 43, 1317-23. Bechinger, B. and Lohner, K. Biochim Biophys Acta. 2006. 1758, 1529-1539. Arouri, A., Dathe, M. and Blume, A. Biochim Biophys Acta. 2009. 1788, 650-9. Fowler, S. M., Riley, R. J., Pritchard, M. P., Sutcliffe, M. J., Friedberg, T. and Wolf, C. R. Biochemistry. 2000. 39, 4406-14. Holst, B., Zoffmann, S., Elling, C. E., Hjorth, S. A. and Schwartz, T. W. Mol Pharmacol. 1998. 53, 166-75. Marana, S. R., Terra, W. R. and Ferreira, C. Eur J Biochem. 2002. 269, 3705-14. Runkel, L., De Dios, C., Karpusas, M., Baker, D., Li, Z., Zafari, M., Betzenhauser, M., Muldowney, C., Miller, S., Redlich, P. N., Grossberg, S. E., Whitty, A. and Hochman, P. S. J Interferon Cytokine Res. 2001. 21, 931-41. Sekine, O., Sugo, T., Ebisawa, K., Umeyama, H., Iwahana, H., Ruiz-Saez, A., de Bosch, N. and Matsudai, M. Thromb Haemost. 2002. 87, 282-7. Wei, Z. and Swiedler, S. J. J Biol Chem. 1999. 274, 1966-70. 161  CHAPTER 5  31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.  Lazaridis, T. Proteins. 2005. 58, 518-27. Radzishevsky, I. S., Rotem, S., Zaknoon, F., Gaidukov, L., Dagan, A. and Mor, A. Antimicrob Agents Chemother. 2005. 49, 2412-20. Shafer, W. M., Katzif, S., Bowers, S., Fallon, M., Hubalek, M., Reed, M. S., Veprek, P. and Pohl, J. Curr Pharm Des. 2002. 8, 695-702. Chandrababu, K. B., Ho, B. and Yang, D. Biochemistry. 2009. 48, 6052-61. Pazgier, M., Prahl, A., Hoover, D. M. and Lubkowski, J. J Biol Chem. 2007. 282, 181929. Sanchez-Hidalgo, M., Martinez-Bueno, M., Fernandez-Escamilla, A. M., Valdivia, E., Serrano, L. and Maqueda, M. J Antimicrob Chemother. 2008. 61, 1256-65. Steffen, H., Rieg, S., Wiedemann, I., Kalbacher, H., Deeg, M., Sahl, H. G., Peschel, A., Gotz, F., Garbe, C. and Schittek, B. Antimicrob Agents Chemother. 2006. 50, 2608-20. Morisset, D., Berjeaud, J. M., Marion, D., Lacombe, C. and Frere, J. Appl Environ Microbiol. 2004. 70, 4672-80. Antcheva, N., Morgera, F., Creatti, L., Vaccari, L., Pag, U., Pacor, S., Shai, Y., Sahl, H. G. and Tossi, A. Biochem J. 2009. 421, 435-47. Kyte, J. and Doolittle, R. F. J. Mol. Biol. 1982. 157, 105-32. Miyazawa, S. and Jernigen, R. L. Macromolecules. 1985. 18, 534-552. Abraham D.J. and A.J., L. Proteins. 1987. 2, 130-52. Copps, J., Murphy, R. F. and Lovas, S. Peptides. 2009. doi:10.1016/j.peptides.2009.09.016. Meier, M. and Seelig, J. J Am Chem Soc. 2008. 130, 1017-24. Patch, J. A. and Barron, A. E. J Am Chem Soc. 2003. 125, 12092-3. Wegener, K. L., Wabnitz, P. A., Carver, J. A., Bowie, J. H., Chia, B. C., Wallace, J. C. and Tyler, M. J. Eur J Biochem. 1999. 265, 627-37. Yang, S. T., Jeon, J. H., Kim, Y., Shin, S. Y., Hahm, K. S. and Kim, J. I. Biochemistry. 2006. 45, 1775-84. Wang, G. Biochim Biophys Acta. 2009. doi:10.1016/j.bbamem.2009.07.028. Leidy, C., Linderoth, L., Andresen, T. L., Mouritsen, O. G., Jorgensen, K. and Peters, G. H. Biophys J. 2006. 90, 3165-75. Salnikov, E. S., Friedrich, H., Li, X., Bertani, P., Reismann, S., Hertweck, C., O'Neil, J. D., Raap, J. and Bechinger, B. Biophys J. 2008. 96, 86-100. Jiang Z, Kullberg BJ, van der Lee H, Vasil AI, Hale JD, Mant CT, Hancock RE, Vasil ML, Netea MG and RS., H. Chem Biol Drug Des. 2008. 72, 483-95. Hale, J. D. and Hancock, R. E. Expert Rev Anti Infect Ther. 2007. 5, 951-9. Wu, M. and Hancock, R. E. J Biol Chem. 1999. 274, 29-35.  162  CHAPTER 6  CHAPTER 6: ew immunomodulatory peptides arise from terminal truncation of the amphibian antimicrobial peptide aurein 2.2  6.1 Introduction Host defense peptides (HDPs) form an important part of the immune defense system of many plants and animals (1). Their unique ability to modulate the host defense mechanism against microbial invasion has triggered great interest in developing these peptides as a novel defense arsenal, namely immunomodulatory therapy, in addition to traditional antibiotic therapy (2). As these peptides exhibit a wide range of biological activities from direct pathogen killing to immune modulation, various HDP applications have been suggested, for example, as new broadspectrum anti-infectives, anti-inflammatory agents, wound-healing/anti-tumor agents, and novel vaccine adjuvants (3). Using the highly diverse peptide structures as a flexible template, one can design more potent and desirable synthetic peptides and derivatives for clinical and commercial applications (4). Examples of these peptides include a range of cationic antimicrobial peptides (CAPs) secreted by mammals or amphibians, which also possess immunomodulatory activities, as part of their host defense mechanism (3,5). Many of these peptides have common structural features including a large number of positively charged and hydrophobic residues. Although highly diverse in terms of amino acid sequence, many HDPs adopt common conformations, such as α-helices, β-sheets or circular structure (6), in the presence of lipids or lipid mimetics. Currently, human defensins, a major family of naturally occurring CAPs, are the most widely studied among all HDPs. Defensins have an average length of 30 residues with a net 163  CHAPTER 6  positive charge and adopt an amphiphilic triple-stranded antiparallel β-sheet structure (7). These peptides usually become active after trypsin-catalyzed removal of an N-terminal segment from an inactive precursor (8,9). The defensins can be grouped into three subfamilies, the α-, β-, and γ-defensins, based on their disulfide-bond arrangement. The α-defensins are predominantly  found in humans and are stored in the secretory granules of neutrophils and other leukocytes (10,11). These peptides are responsible for maintaining immune homeostasis in the gut (12). Unlike α-defensins, β-defensins are present in the mucosal secretions of many epithelial cells in response to proinflammatory stimuli and infection, occurring in respiratory, gastrointestinal or urogenital tracts, and in inflamed skin (13-16). Several types of cells in the human immune system produce β-defensins, including monocytes, macrophages, and dendritic cells (7). γdefensins are much rarer and only found in rhesus macaques, but not in humans or other mammals. They are produced in neutrophils and monocytes as cyclic molecules and exhibit some moderate antimicrobial activities (17). Cathelicidins are the second major family of HDPs in mammals. Bovine and porcine immune systems produce a large variety of cathelicidins, including bactenecin, indolicidin, PR39, and many others (18). LL-37, a human cathelicidin, has been extensively studied (19) and adopts an amphipathic α-helix in the presence of dodecylphosphocholine (DPC) micelles (20). In addition to its antimicrobial activities, LL-37 also has broad immunomodulatory activities. These include modulation of chemokine/cytokine production (21-23), chemoattraction of immune cells (23,24) and the selective down-regulation of TNF-α responses to pro-inflammatory stimuli including bacterial lipopolysaccharide (LPS) (25).  Similarly, LL-37 is normally  produced as an inactive precursor protein hCAP18 (26) and becomes active following proteolytic cleavage of an N-terminal cathelin-like domain (18).  Several studies have reviewed or  characterized other HDPs that were found to possess immunomodulatory functions, such as 164  CHAPTER 6  magainin 2 amide (27), peptide leucine arginine (pLR) (5), melittin (28), Bac2A (linear derivative of bactenecin) and indolicidin (29). Both magainin 2 amide and pLR are amphibian CAP derivatives but interestingly, to the best of our knowledge, no study has yet reported any immunomodulatory activity from aurein peptides, another widely studied family of amphibian CAPs from Australian Southern Bell frogs Litoria aurea (30). We have demonstrated the structure-function relationship of aurein 2.2, aurein 2.3 and aurein 2.3-COOH in Chapter 2 and some of their derivatives in Chapter 5. In addition to these derivatives, we designed two more mutant analogues of aurein 2.2, one with the first four Nterminal residues truncated (aurein 2.2-∆4N), and the other with an additional three C-terminal residues truncated (aurein 2.2-∆4N∆3C). As defensins and cathelicidins are activated following proteolytic processing of the N-terminal region, truncating the N- and/or C-terminus of aurein 2.2 may perhaps improve/abolish its immunomodulatory activity. This information could give us the first insight to which N-terminal residues are required for antimicrobial activity in the aurein 2 family peptides and whether an N-terminal truncation could give rise to an immunomodulatory peptide such as LL-37 and others.  In addition to studying these two  peptides, we also investigated the immunomodulatory activity of an aurein mutant, L13A, discussed in Chapter 5. In addition to in vitro and in vivo assays, we utilized various biophysical methods to characterize the structure and peptide-membrane interaction of these three aurein peptides. To assess whether these peptides have immunomodulatory activity, we performed assays on human peripheral blood mononuclear cells (PBMCs) and monitored the production of monocyte chemoattractant protein (MCP-1), a chemokine from mammalian T cells, in the presence of these aurein mutants. Previously, it has been demonstrated that MCP-1 induction in PBMCs correlates with the immunomodulatory activity of HDPs (31). To examine whether these peptides are toxic 165  CHAPTER 6  to host cells, we conducted cytotoxicity assays on PBMCs to monitor the release of cytosolic lactate dehydrogenase (LDH) in the presence of these peptides. To determine the antimicrobial activity of the N-terminally truncated aurein mutants, we determined the minimum inhibitory concentration (MIC) against Staphylococcus aureus strain C622 and Staphylococcus epidermidis strain C621 (Gram positive bacteria), and Escherichia coli strain C500 and Pseudomonas aeruginosa strain Pa01 (Gram negative bacteria). In addition, we conducted cell membrane  permeabilization assays to examine whether changes in sequence affect the ability of peptides to permeate through lipid membranes. We also performed solution CD and 1H NMR experiments to examine the structure of these peptides in aqueous, trifluoroethanol (TFE) and lipid environments.  We also conducted oriented CD (OCD) experiments to inspect whether N-  terminal truncation has an effect on the peptide-membrane interaction. Finally, these results will be correlated to the results from murine models of bacterial infection in order to determine whether these peptides impart protection in an animal model. Overall, these data should allow us to determine which N-terminal residues are essential for antimicrobial activity, to assess whether N-terminal truncation can give rise to immunomodulatory activity, and to better understand how sequence modulates structure-function relationships.  6.2 Results 6.2.1 Immunomodulatory activity in vitro Monocyte chemoattractant protein (MCP-1) is a chemokine that shows chemotactic activity for various immune cells including monocytes/macrophages, T cells, and neutrophils. It has been used as a chemokine screen for immunomodulatory peptides and for adjuvant development, such as HH2 and CpG (32), or their complex (33). For these chemokine screens,  166  CHAPTER 6  HH18, an immunomodulatory peptide with minimal induction of MCP-1, was used as the control. In addition, both aurein parent peptides were used for comparison. Human PBMCs were stimulated with the three aurein mutants and monitored for induction of MCP-1 by enzyme-linked immunosorbent assay (ELISA). Figure 6.1(a) shows the relative amount of MCP-1 production as a function of the different peptides.  Figure 6.1.  Monocyte chemoattractant protein (MCP-1) induction in the presence of different peptides: (a) MCP-1 production in pg/ml as a function of different peptides. Note that the Y-axis unit is in arbitrary units. (b) MCP-1 induction efficiency with respect to aurein 2.3. Note that both negative control (HH18) and aurein 2.2 did not induce any observable MCP-1 production. (Immunomodulatory activity assay results are courtesy of Dr. J. Kindrachuk).  Both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C led to potent induction of MCP-1 production in vitro as compared to the parent peptides aurein 2.2 and aurein 2.3. This indicates  167  CHAPTER 6  that the presence of the N-terminally truncated mutants gave rise to significantly improved immunomodulatory activities. L13A also enhanced MCP-1 production, but not as effectively as the N-terminally truncated mutants. Figure 6.1(b) shows the MCP-1 induction efficiency for the three aurein mutants with respect to aurein 2.3. Note that aurein 2.2 did not induce MCP-1 production, so it was not used further for comparison purposes. Compared to aurein 2.3, both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C improved chemokine induction by a factor of ~ 150, whereas L13A improved it by a factor of ~ 55. These results were particularly striking for aurein 2.2-∆4N and aurein 2.2-∆4N∆3C. They showed greatly enhanced immunomodulatory capability and the potential ability to recruit immune cells, as compared to the parent peptides. This suggests that these peptides may be ideal templates for the generation of novel innate defense regulars (IDRs) for therapeutic application.  6.2.2 Cytotoxicity activity in vitro It is important that an HDP with potent immunomodulatory activity is not toxic to host cells. In order to assess possible toxicities, cytotoxicity assays were performed to monitor cell membrane permeabilization by measuring release of LDH in human PBMCs in the presence of the three aurein mutants. Figure 6.2(a) shows LDH release in PBMCs for each of the peptides. Many cationic peptides have been shown to have strong membrane lytic properties. Interestingly, both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C resulted in minimal to no release of LDH from PBMCs. This suggests that both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C were not highly toxic to mammalian cells, the target host. L13A, on the other hand, resulted in a slightly higher release of LDH from PBMCs. When compared to aurein 2.2 (100%), both aurein 2.2∆4N and aurein 2.2-∆4N∆3C showed much less cytotoxicity with only 27% and 0%,  respectively, whereas L13A was slightly more cytotoxic with 37% (Figure 6.2(b)). The presence 168  CHAPTER 6  of aurein 2.3, interestingly, resulted in a slightly less cytotoxic effect with only 67% of that for aurein 2.2.  Figure 6.2.  Cytosolic lactate dehydrogenase (LDH) release in the presence of different peptides: (a) LDH release monitored at OD492 as a function of different peptides. Note that the Y-axis unit is arbitrary. (b) LDH release efficiency with respect to aurein 2.2 (100%). (Cytotoxicity assay results are courtesy of Dr. J. Kindrachuk).  6.2.3 Antimicrobial activity Previous studies have reported that cationic peptides, which show immunomodulatory activity, generally have moderate or weak antimicrobial activity (3). In order to verify whether this is true for the two N-terminally truncated aurein mutants, MICs of the two aurein mutants against two Gram-positive (S. aureus and S. epidermidis) and two Gram-negative (E. coli and P. aeruginosa) bacteria were determined. The MICs, reported in Table 6.1, indicate that both 169  CHAPTER 6  aurein mutants had identical MICs of ≥ 128 µg/ml against all bacteria tested here. This suggests that both aurein mutants showed minor or no activity against these bacterial strains. The MIC results for L13A were adapted from Chapter 5 and confirmed that the N-terminal region is needed for the peptide to exhibit antimicrobial activity against S. aureus and S. epidermidis. Wells containing culture alone and broth alone were used as controls.  Table 6.1.  Minimal inhibitory concentrations (MICs) in µg/ml of the three aurein mutants toward S. aureus and S. epidermidis (Gram positive bacteria), and E. coli and P. aeruginosa (Gram negative bacteria). MICs are given as the most frequently observed value obtained from repeat experiments. (MIC assay results are courtesy of Dr. J. D. Hale and M. Elliot).  Peptide Aurein 2.2-∆4N Aurein 2.2-∆4N∆3C L13A  S. aureus (C622) 128 128 16  S. epidermidis (C621) 128 128 8  E. coli (C500) 128 128 128  P. aeruginosa (Pa01) >128 >128 >128  *MIC value of L13A was adapted from Chapter 5.  6.2.4 Secondary structure by solution CD spectroscopy In order to characterize the secondary structure of the two aurein mutants in the presence of a lipid or lipid-mimicking environment, solution CD experiments were performed in ddH2O/trifluoroethanol  (TFE)  (vol/vol),  dodecylphosphocholine  (DPC)  and  sodium  dodecylsulfate (SDS). Figure 6.3 shows solution CD spectra of the three aurein mutants as a function of ddH2O/TFE ratio (vol/vol). In the presence of ddH2O, the spectra of aurein 2.2-∆4N and aurein 2.2-∆4N∆3C consisted of one minimum at 190 nm which is characteristic of random coil. As the % TFE increased, the minimum at 190 nm gradually shifted toward 200 nm, the minimum at 222 nm became more prominent, and a maximum at 190 nm was evident. These features demonstrate that the two N-terminally truncated aurein mutants formed a mixture of αhelices and random coils, which may indicate saturation in peptide binding (34). Increase in the % TFE did not result in different spectra for L13A, which indicates that a maximal α-helix and minimal random coil contribution already occurred at 25% TFE. 170  CHAPTER 6  Figure 6.3.  Solution CD spectra of aurein mutants in ddH2O/TFE mixtures (vol/vol): (a) aurein 2.2-∆4N; (b) aurein 2.2-∆4N∆3C; (c) L13A (solid black line = 0% TFE; dashed black line = 25% TFE; solid grey line = 50% TFE; dashed grey line = 75% TFE). Spectra indicate that aurein 2.2∆4N and aurein 2.2-∆4N∆3C adopted mostly random coils in the presence of aqueous environment or low TFE concentrations, but adopted a mixture of random coil and α-helical conformation at high TFE concentrations. L13A, on the other hand, underwent structural change from random coil to αhelix when TFE was present.  To determine actual structures of these peptides in lipid membranes, solution CD experiments were performed in DPC (mimicking a neutral mammalian membrane environment) and SDS (mimicking a negatively charged bacterial lipid membrane environment). Figure 6.4 shows solution CD spectra of the three aurein mutants in DPC and SDS micelles as a function of P/L molar ratio.  171  CHAPTER 6  Figure 6.4.  Solution CD spectra of aurein mutants in DPC (left panel) and SDS (right panel) micelles: (a) & (d) aurein 2.2-∆4N; (b) & (e) aurein 2.2-∆4N∆3C; (c) & (f) L13A (solid line, P/L = 1:15; dashed line, P/L = 1:50; dotted line, P/L = 1:100). Spectra indicate that the aurein mutants adopted a mixed conformation of α-helix and random coil or an α-helical conformation in the presence of DPC and SDS micelles, except aurein 2.2-∆4N∆3C at 1:15 P/L molar ratio.  In the presence of DPC micelles at high peptide concentrations, the spectrum of aurein 2.2-∆4N consisted of one minimum at 200 nm. This indicates that the secondary structure of aurein 2.2-∆4N remained mostly random coil, contrary to the other aurein peptides. As peptide concentration decreased, the spectra gradually transitioned to show two minima at 207 nm and 222 nm and one maximum at 190 nm. This indicates that aurein 2.2-∆4N adopted a mixed conformation of α-helix and random coil, and the α-helical spectral characteristics were more obvious and dominant at 1:100 P/L molar ratio. The spectra of aurein 2.2-∆4N∆3C, on the other 172  CHAPTER 6  hand, mostly showed one minimum at 200 nm. This indicates that aurein 2.2-∆4N∆3C did not form a specific structure except at low peptide concentrations, and the random coil conformation dominated at all peptide concentration examined. The spectra of L13A exhibited two minima and one maximum, which are spectral characteristics of an α-helical structure. This indicates that L13A formed an α-helix in the presence of DPC and the structural conformation was independent of peptide concentration. Note that the spectra of aurein 2.2-∆4N and aurein 2.2∆4N∆3C were acquired from 180 nm to 260 nm to highlight the spectral characteristics of mixed  conformation. In the presence of SDS micelles, all spectra consisted of a maximum at 190 nm and two minima at 207 nm and 222 nm, which are characteristic of α-helical structure.  This  demonstrates that the three aurein mutants adopted an α-helical conformation in this environment. The only exception is aurein 2.2-∆4N∆3C in SDS at a 1:15 P/L ratio, where aurein 2.2-∆4N∆3C adopted a mixture of α-helix (~ 36%) and β-sheet/β-turn (~ 37%). As previously observed (35), similar intensities were found for all peptide-to-lipid molar ratios (P/L= 1:15, 1:50, and 1:100) studied, indicating that maximum binding of the peptide to the lipid vesicles occurred. Saturation would be observed with a combination of signals from both α-helical and random coil (34). To examine the percentage of structure content of each aurein mutant in lipid membranes and the membrane mimetics studied, all spectra were fitted using three different programs (CDSSTR (36), CONTINLL (37), and SELCON3 (38-40)). Figure 6.5 shows the percentage of structure content as a function of % TFE (vol/vol) or P/L molar ratio for the three aurein mutants.  173  CHAPTER 6  Figure 6.5.  Secondary structure content of the three aurein mutants in (a) ddH2O/TFE (vol/vol), (b) DPC, and (c) SDS micelles: The percentage of α-helical and random coil content were plotted as a function of % TFE, or P/L molar ratio, and aurein mutants (as indicated). The plots show that the percentage of α-helical content increased and the percentage of random coil content decreased as % TFE increased or peptide/DPC molar ratio decreased for the two N-terminally truncated aurein mutants, and stayed relatively constant as peptide/SDS molar ratio increased. % α-helix did not change significantly and was close to 100% for L13A in TFE or lipid environment. 174  CHAPTER 6  In the presence of TFE, both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C showed an increased % α-helix and decreased % random coil as % TFE increased. In the presence of DPC, interestingly, both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C showed an increased % α-helix and decreased % random coil as the DPC concentration increased. This is contrary to the previously observed behaviour of the aurein peptides in lipid membranes (35,41). L13A, on the other hand, showed high helical content at all concentrations examined. In the presence of SDS, all aurein mutants showed no obvious change in the structure content, except that aurein 2.2-∆4N∆3C had a lower % α-helix but a higher % β-sheet at 1:15 P/L molar ratio. These results demonstrate that aurein 2.2-∆4N and aurein 2.2-∆4N∆3C required a high percentage of TFE to form ~ 37% helical structure, but remained relatively structured in SDS (~ 80% and ~ 60% α-helix, respectively). The presence of DPC at low concentrations, surprisingly, did not induce obvious conformational changes of the N-terminally truncated aurein mutants, but a high DPC concentration did.  L13A, as observed previously, still remained close to 100% α-helical  regardless of TFE or different membrane environments.  Overall, the data show that the  structures of the two N-terminally truncated aurein mutants were dependent of the molar concentrations examined, but not the structure of L13A.  6.2.5 Secondary structure by solution 1H MR spectroscopy In order to determine whether the peptides adopt an α-helical structure at atomic level, solution-state 1H NMR experiments were performed and the spectra were collected using 75% TFE-d3 (vol/vol).  1  H NMR spectra for aurein 2.2-∆4N were assigned using the TOCSY and  NOESY data sets, using TOPSPIN. Figure 6.6 shows the fingerprint region of 1H NMR NOESY spectrum of aurein 2.2-∆4N in the presence of 75% TFE-d3. Furthermore, we have completed  175  CHAPTER 6  the sequential assignment of amino acid residues using the fingerprint region of the TOCSY spectrum (spectrum not shown).  Figure 6.6.  Fingerprint region of solution-state 1H MR OESY spectrum of aurein 2.2-∆ ∆4 : The spectrum was acquired using a phase sensitive NOESY experiment, with excitation sculpting with gradients for water suppression. The spectrum was acquired at 30°C, using 64 scans and a mixing time of 150 ms. The spectrum was referenced to the residual methylene protons present in TFE-d3 (3.918 ppm). In the spectrum, green lines indicate the amino acid connectivities used to perform the sequential assignment.  6.2.6 Membrane insertion states In order to understand how the peptides interact with membranes, OCD experiments were conducted to investigate peptide insertion profiles in 1:1 POPC/POPG (mol/mol) bilayers. All experiments were conducted at 30°C (liquid crystalline phase). In addition, experiments were repeated at least twice to ensure reproducibility of the results. Figure 6.7 shows the OCD spectra 176  CHAPTER 6  for the two N-terminally truncated aurein mutants in 1:1 POPC/POPG (mol/mol) bilayers as a function of P/L molar ratio. The OCD spectra for L13A can be found in Chapter 5.  Figure 6.7.  Oriented CD spectra of -terminally truncated aurein mutants in 1:1 POPC/POPG (mol/mol) bilayers: (a) aurein 2.2-∆4N and (b) aurein 2.2-∆4N∆3C. The P/L molar ratios are 1:15 (blue), 1:80 (red), and 1:120 (green). Spectra were normalized at 222 nm. The spectra show that peptides inserted into 1:1 POPC/POPG (mol/mol) bilayers at threshold P/L molar ratios between 1:15 and 1:80.  The spectra were normalized at 222 nm as previously described (35,41). The spectra showed that both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C inserted (inserted, I-state or tilt, Tstate) into POPC/POPG bilayers at threshold P/L molar ratios between 1:15 and 1:80, and were surface-adsorbed (S-state) at P/L molar ratios greater than 1:80 (mol/mol). The data illustrate that removing the last three residues and/or the four N-terminal residues of aurein 2.2 has a drastic impact on the peptide insertion profile, making the peptide behave like aurein 2.3-COOH (35). This indicates that removing N-terminal residues reduced the ability of the peptide to insert  177  CHAPTER 6  into POPC/POPG membranes. Overall, this suggests that the N-terminal residues may play a critical role in aurein peptide insertion into the lipid bilayers.  6.2.7 Membrane permeation ability Since both N-terminally truncated mutants could not insert into 1:1 POPC/POPG (mol/mol) bilayers effectively, we wanted to know whether they showed similar behaviour in live host cells. In order to assess the ability of these peptides to permeate across biological membranes, cell membrane permeabilization assays were conducted.  Figure 6.8 shows  membrane permeation assay spectra of aurein 2.2-∆4N and aurein 2.2-∆4N∆3C at 20 µg/ml and 100 µg/ml, with 0.5% Triton-X as the control.  Figure 6.8.  The ability of -terminally truncated aurein mutants to permeate across real cell membranes was assessed as a function of time at 20 µg/ml and 100 µg/ml of peptides. 0.5% Triton-X was used as control. AM01 = aurein 2.2-∆ ∆4 and AM04 = aurein 2.2-∆ ∆4 ∆3C. (Membrane permeation assay results are courtesy of Dr. J. Kindrachuk). 178  CHAPTER 6  The results indicate that both N-terminally truncated aurein mutants were unable to permeate across biological membranes even at high peptide concentrations. This suggests that N-terminal residues are necessary to facilitate peptide permeation through the lipid bilayers.  6.2.8 Immunomodulatory activity of select aurein mutants in a murine model of bacterial infection The ability of N-terminally truncated mutants to protect animals was examined in murine model of S. aureus infection. CD-1 mice were pre-treated with the various peptides or saline control 4 hours prior to infection. Surprisingly, pre-treatment of the mice with the aurein mutants did not improve bacterial clearance as compared to the control mice. The inability of these peptides to protect the mice indicates that despite their highly enhanced immunomodulatory activities in vitro, the N-terminally truncated mutants were not able to exhibit their activities in vivo (possibly excreted?). A similar test is currently under way to determine whether L13A provides protection in vivo.  6.3 Discussion In Chapter 5, we demonstrated that determining which residue is crucial to the antimicrobial activity can be a key step in understanding the mechanism of action of a given antimicrobial peptide. In this chapter, we created another two mutant peptides from aurein 2.2, an antimicrobial peptide from Australian Southern Bell frog Litoria aurea, to inspect how Nterminal truncation would have an effect on the peptide structure. We then examined how changes in the structure modulate the antimicrobial activity and/or give rise to immunomodulatory activities within these mutant peptides. Many immunomodulatory peptides (IMPs) were discovered when some antimicrobial peptide derivatives were developed as novel antibiotics, such as MX226 and hLF1-11 (3). These 179  CHAPTER 6  discoveries have shown that it is possible to develop new IMPs based on currently found antimicrobial peptides. To design a new antimicrobial peptide, mutating specific residues or truncation through N- or C-terminal deletions are common practices for screening peptides for antimicrobial regions that may be exploited for further design strategies. To date, few studies have been conducted to examine the effect of terminal truncation on these peptides. These include removing residues of specific length from either terminus, as demonstrated by the truncation study of human β defensin-3 (42), LL-37 (43) and lactoferrampin (44). By correlating the results from these model membrane studies (for example, solution and oriented CD) with the results obtained from assays performed in vitro (for example, immunomodulatory activity and cytotoxicity), we can gain a better picture of how peptide truncation influences the antimicrobial/immunomodulatory activities and the structure-function relationship of these peptides under biologically relevant conditions. Our solution CD results demonstrated that both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C adopted similar structures, forming primarily a mixture of α-helices and random coils in the presence of TFE, DPC or SDS micelles. Several studies have also demonstrated that some Nterminally modified cationic peptides (for example, lactoferrampin) still adopt similar, but less well defined, structures than their parent counterparts (44). We showed here that the main structures were preserved but the helical contents of the two mutants were dependent on the TFE or lipid concentration. Both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C showed increased % αhelix as TFE or DPC concentration increased, and decreased % random coil contributed to increased % α-helix. Previous studies of LL-37 also demonstrated that it adopts an α-helical conformation in the presence of DPC micelles (20), but the dependence of helical content on lipid or peptide concentration was not examined extensively. As TFE is known to induce peptide secondary structures, it would be anticipated that a higher TFE concentration would have more 180  CHAPTER 6  stabilizing effect and induce a higher helical content of these extensively truncated peptides. However, it was surprising to observe an increased DPC concentration could also induce a higher helical content of these peptides, which has not previously been, to the best of our knowledge, reported quantitatively. This suggests two possibilities. One is that saturated peptide binding to DPC micelles already occurred at low DPC concentrations. Another could be that N-terminus was necessary for forming a complete α-helical structure at low lipid concentrations, as demonstrated by the study of N-terminally truncated lactoferrampin (44). The peptides adopted better defined structures in the presence of SDS. Since both aurein mutants still retain a net +2 charge, one would expect the peptides to bind more effectively to SDS than neutral lipids, explaining why the structure is better defined in this case. Overall, these peptides displayed a high degree of structural plasticity.  As long as they adopt a certain degree of secondary  structure, they may be able to exhibit their immunomodulatory functions. L13A, on the other hand, due to the presence of four N-terminal residues, formed > 99% α-helical structure. This once again confirms that N-terminal residues were necessary for forming a more defined structure. However, one must note that an increase in % α-helix may not correlate with increase in immunomodulatory activity (Figure 6.1 and Figure 6.5). This is supported by the similar antimicrobial activities of aurein peptides and mutant analogues with different helical content described in Chapter 5. Similarly, we demonstrated here that structure alone may not be used as an indicator to assess the ability of a peptide to modulate immune responses. The model antimicrobial peptide membrane perturbation (see Chapter 1) requires that a peptide adopt a structure upon contact with bacterial membranes and then perturbs the lipid bilayer. A model of how immunomodulatory peptides function, on the other hand, may not involve the same steps. First, immunomodulatory peptides were found to be very effective in interacting with DPC micelles, which would easily lead to saturation in peptide binding at low  181  CHAPTER 6  peptide concentration. This is of interest as the N-terminal segment “G1L2F3D4” does not carry any positive charge and could promote better peptide binding to the neutral membrane surface. In contrast, we observed that these peptides did not need this segment for improved peptide binding. Second, these peptides may not need to form well defined structures, or may form extended structures as indolicidin (45), to interact with lipid membranes. In this case, extended structures or structures with high random coil content may be the more ideal conformation for these peptides to interact with neutral lipid membranes or modulate immune responses. In other words, these peptides may not need to form a specific structure, most likely after proteolytic cleavage of the N-terminus, in order to modulate immune activities. The correlation between the helical content and the immunomodulatory activity was further investigated. Our immunomodulatory activity assay results illustrated that helical content did not have a significant impact on the immunomodulatory efficiency. Both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C displayed a similar degree of MCP-1 induction, despite having a slightly lower % α-helix for aurein 2.2-∆4N∆3C due to its shortened C-terminus. This indicates that Nterminal, not C-terminal, truncation of aurein 2.2 resulted in an enhanced immunomodulatory activity.  This further suggests that C-terminal segment “G14S15L16” was not required for  displaying immunomodulatory function. The high level of MCP-1 induction by aurein 2.2-∆4N (~ 15,000 pg/ml) and aurein 2.2-∆4N∆3C (~ 14,000 pg/ml) appeared to be much stronger than the parent peptides aurein 2.2 (0 pg/ml) and aurein 2.3 (~ 97 pg/ml). L13A, even though with a slightly less enhanced immunomodulatory activities (~ 5400 pg/ml), was still 55 times more effective than aurein 2.3. This makes these peptides ideal candidates for IDR development. Another criterion was that these peptides were not toxic to host cells. Our cytotoxicity assay results demonstrated that aurein 2.2-∆4N∆3C did not induce a significant cytotoxic response or LDH release from human PBMCs. However, both aurein 2.2-∆4N and L13A showed a slightly 182  CHAPTER 6  higher LDH release. This suggests that the presence of the C-terminal segment “G14S15L16” may be related to cytotoxicity. We demonstrated in Chapter 5 that L13A behaved similarly to aurein 2.3 in model membranes. Here, we observed that both L13A and aurein 2.3 were more cytotoxic than the two N-terminally truncated mutants, but less than aurein 2.2. This suggests that by branching off the Leu13 side chain into alkyl constituents or by shortening the aliphatic chain, the cytotoxic effect of the peptide could possibly be reduced. Interestingly, previous study of the mammalian LL-37 has shown that LL-37 analogues with two or six N-terminal residues truncated still retain their antimicrobial activity and display their immunomodulatory activity to the same extent as the parent peptide (43). This is in sharp contrast to what was observed here. Both aurein 2.2-∆4N and aurein 2.2-∆4N∆3C showed similar MIC values as aurein 2.2∆6 (Chapter 5). This suggests that aurein 2.2 requires both the N-terminus and the segment  containing residue 13 to exhibit antimicrobial activity. In order to determine whether this low activity is caused by a difference in peptide-membrane interaction, we examined the peptide insertion profiles for the two aurein mutants. Both mutants showed an insertion threshold between 1:15 and 1:80 P/L molar ratios, similar to that found for aurein 2.2-∆6. The presence of the C-terminus did not promote better peptide insertion. This was also demonstrated by aurein 2.2-∆3 (Chapter 5) where it inserted into the lipid bilayers as efficiently as aurein 2.2. This further confirms that C-terminal segment “G14S15L16” is not required for peptide-membrane interaction, but the N-terminus is more important.  Further evidence was obtained from  membrane permeation assay results, where both mutants could not permeate through lipid membranes even at high peptide concentrations. These model membrane study results correlate with murine model study results, where the inability of N-terminally truncated aurein mutants to permeate across lipid membranes could explain the inability of these peptides to protect the mice.  183  CHAPTER 6  Sequence conservation has been evolutionarily important for many proteins and peptides to preserve their functions. For aurein peptides, the N-terminal residues are usually conserved (30). Thus, N-terminal truncation could easily abolish the antimicrobial activity. Furthermore, among the aurein 2 family peptides, aurein 2.1 differs from aurein 2.2 by a Phe3-Leu13 swap, and is found to be much less active (30). This suggests that Phe3 is needed for the N-terminal sequence conservation, and it has been shown that Phe or other aromatic residues, especially Trp, are necessary at the N-terminus to facilitate peptide insertion into the lipid bilayers (46,47).  6.4 Summary and conclusion Overall, the results presented here suggest that the N-terminus is required for aurein 2.2 to exhibit antimicrobial activity and for peptide insertion into or permeation across lipid membranes to take place. Furthermore, N-terminal truncation by four residues can give rise to peptides with enhanced immunomodulatory activities, and L13A mutant may be an ideal candidate for future IMP development due to its ability to show some activity and to insert into lipid membranes. Finally, this study has again shown that the peptide sequence can modulate activities and are well correlated with the structure-function relationship of aurein 2.2 peptide.  6.5 Materials and methods 6.5.1 Materials Peptide synthesis materials were purchased as previously informed in Chapter 2. 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3[phospho-rac-(1-glycerol)] (POPG) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). 2,2,2-trifluoroethanol was purchased from ACROS Organics (New Jersey, USA) and TFE-d3 was obtained from Cambridge Isotope Laboratories (Andover, MA, USA). 184  CHAPTER 6  6.5.2 Methods 6.5.2.1 Peptide synthesis Aurein 2.2 mutants were synthesized as previously described (33,35), using an CS Bio Co. peptide synthesizer (Menlo Park, CA) by in situ neutralization Fmoc chemistry, using Rink resin with the C-terminal double-coupling scheme. HH18 (IWVIWRR-NH2) was synthesized using solid phase Fmoc chemistry, purified to a purity > 90% using reverse phase HPLC and analyzed by mass spectrometry, at Gen-Script (Piscataway, NJ, USA).  6.5.2.2 Purification The crude aurein 2.2 mutant peptide product was purified by preparative RP-HPLC on a Waters 600 system (Waters Limited, Mississauga, ON, Canada) with 229 nm UV detection using a Phenomenex (Torrance, CA, USA) C4 preparative column (20.0 µm, 2.1 cm × 25.0 cm) as previously described (35).  The identity of the products was verified using electrospray  ionization (ESI) mass spectrometry and MALDI-TOF as previously described (35) and confirmed to be ≥ 99% pure.  6.5.2.3 Solution circular dichroism (CD) sample preparation Solution CD samples with constant peptide concentration of 2.0 mM were prepared in different % TFE: 0, 25, 50, and 75% (vol/vol), or in different peptide-to-lipid (P/L, DPC or SDS) molar ratios: 1:15, 1:50, and 1:100. Appropriate amounts of lipid in chloroform or MeOH were dried using a stream of nitrogen gas to remove most of chloroform or MeOH and vacuum dried overnight in a 5.0 ml round bottom flask. After adding 450.0 µl of ddH2O and 0.5 µmol of peptide in ddH2O to dried lipid, the mixture was sonicated in a water bath for a minimum of 30  185  CHAPTER 6  min (until the solution was no longer turbid) to ensure lipid vesicle formation. For all samples, corresponding background samples without peptides were prepared for spectral subtraction.  6.5.2.4 Mechanically oriented sample preparation Oriented CD samples were prepared using three different P/L molar ratios of 1:15, 1:80 and 1:120.  The lipids were dried using a stream of nitrogen gas to remove most of the  chloroform and vacuum dried overnight in a 5.0 ml round bottom flask. The peptide amount was kept constant at 0.5 µmol and mixed with appropriate molar ratios of lipids and sonicated in 2.0 ml of ddH2O. Each mixture was deposited in 90.0 µl portions onto 3.0 cm × 1.0 cm and 1.0 mm thick quartz slides, cleaned thoroughly with ddH2O and ethanol prior to sample preparation. The plated samples were then placed in a 93% relative humidity chamber and were indirectly hydrated by incubating inside a dessicator at 37ºC for 8 days. Clear layers of samples were observed on the slides after indirect hydration of the samples. The humidity of the samples was verified by visual inspection. The degree of alignment was previously verified by solid-state 31P NMR. Prior to CD spectral acquisition, each sample was covered with a second slide with a spacer (6 layers of stacked parafilm in a rectangular 3.0 cm × 1.0 cm frame with 2.0 mm width).  6.5.2.5 Circular dichroism Solution and oriented CD experiments were carried out using a JASCO J-810 spectropolarimeter (Victoria, BC, Canada) at 30oC as previously described (41). Briefly, the spectra were obtained over a wavelength range of 190 ~ 250 nm, using continuous scanning mode with a response of 1 s with 0.5-nm steps, a bandwidth of 1.5 nm, and a scan speed of 50 nm/min. The signal/noise ratio was increased by acquiring each spectrum over an average of three scans. Finally, each spectrum was corrected by subtracting the lipid background from the sample spectrum. Solution CD samples were placed in a cell (0.1 cm in length) in 200 µl 186  CHAPTER 6  portions, while oriented CD samples on quartz slides (as described above) were directly placed in the sample compartment. The temperature of the sample compartment was kept constant by means of a water bath. The CD experiments were repeated twice (solution CD) or 3 ~ 4 times (OCD). Linear dichroism effects were tested for and found not to contribute to the signal.  6.5.2.6 MR spectroscopy Solution-state NMR data was acquired on a Bruker 500 MHz instrument, operating at a 1  H frequency of 500.17 MHz. The parameters for the experiments were chosen to match  previously reported parameters for aurein 1.2 (30), as much as possible. All spectra were collected at 30°C. Spectra were acquired using total correlation spectroscopy (TOCSY) and nuclear Overhauser enhancement spectroscopy (NOESY) experiments, in phase-sensitive mode using time proportional phase incrementation (TPPI) (48) in the indirect dimension.  The  TOCSY experiment used the MLEV17 sequence for mixing (mixing time = 70 ms) and excitation sculpting with gradients for water suppression (49). The 2D data set consisted of 4096 data points in t2 and 256 points in t1. The NOESY experiment was acquired with a mixing time of 150 ms and also used excitation sculpting for water suppression. The data size for this data set was the same as for the TOCSY spectrum. Signals were averaged using 32 scans for the TOCSY and 64 scans for the NOESY experiments, respectively. The spectra were referenced to ddH2O (4.800 ppm). Spectra were processed to result in 1k x 1k points.  6.5.2.7 Immunomodulatory assay 6.5.2.7.1  Cell isolation and peptide stimulation  Venous blood from healthy volunteers was collected in Vacutainer® collection tubes containing sodium heparin as ananticoagulant (BD Biosciences, Sparks, MD, USA) in accordance with the University of British Columbia ethical approval and guidelines. Blood was 187  CHAPTER 6  diluted with an equal volume of complete RPMI 1640 medium, supplemented with 10% (vol/vol) heat-inactivated FBS, 2.0 mM L-glutamine, and 1.0 mM sodium pyruvate (all from Invitrogen Life Technologies, Carlsbad, CA, USA) and separated by centrifugation over a FicollPaque Plus (Amersham Biosciences, Piscataway, NJ, USA) density gradient. The buffy coat was collected and, washed twice in RPMI 1640 complete medium, and the number of PBMCs was determined by trypan blue exclusion. PBMC (1 x 105) were seeded into 96-well tissue culture dishes (Falcon; BD Biosciences, Sparks, MD, USA) at 1 x 106 cells/ml at 37°C in 5% CO2, and rested for 1 hr. The cells where then exposed to peptide at 20 or 100 µg/ml for 24 hrs. All experiments involved at least three biological replicates. 6.5.2.7.2  MCP-1 detection  Following 24 hrs of peptide exposure, the tissue culture supernatants were centrifuged at 16,000 × g (13,000 rpm) at 4°C for 5 minutes in an IEC MicroMax centrifuge to obtain cell-free samples. Supernatants were aliquoted and then stored at –20°C before assaying for chemokine secretion. MCP-1 secretion in the tissue culture supernatants was detected by sandwich ELISA (BioSource International and eBiosciences, respectively). All assays were performed in triplicate. The concentration of MCP-1 in the culture medium was quantified by establishing a standard curve with serial dilutions of recombinant human MCP-1.  6.5.2.8 Cytotoxicity assay PBMCs were isolated as previously described (33). To test toxicity of the peptide, PBMC (2 × 105) were seeded into 96-well plates (Sarstedt, NC, USA) and incubated at 37°C in 5% CO2 overnight. The release of cytosolic lactate dehydrogenase (LDH) was then assessed after 24 hr of incubation with the peptide (33). Treatment of cells with 0.5% triton was used as a positive control. All experiments were done in triplicate.  188  CHAPTER 6  6.5.2.9 Immunization of mice The animal model infection studies were performed in accordance with UBC animal care ethics approval and guidelines, as per animal care certificate #A04-0020. In brief, female CD-1 mice (Jackson Laboratories, Bar Harbor, Maine, USA) were infected i.p. with 200 µl (approximately 5.0 × 108 colony forming units) of a S. aureus (ATCC#25923) culture in mid-log phase diluted 1/10 in Porcine Mucin (Sigma# M2378, 5% w/v in saline [0.9% NaCl in water]). Peptide (8 mg/kg dissolved in sterile saline) was administered through same route 4 hr prior to infection. After 24 hr the mice were euthanized via CO2 asphyxiation and cervical dislocation. Peritoneal lavage was collected and serial diluted in PBS and plated onto Muller Hinton plates for bacterial enumeration.  6.5.2.10 Membrane permeation assays The SYTOX green assay for cell membrane permeability was performed as per the manufacturer’s instructions (Promega, Madison, WI, USA). Briefly, 105 PBMCs were washed twice in PBS then incubated in the dark with 1 µM SYTOX Green in PBS for 15 minutes. Fluorescence was measured every 5 minutes after peptide addition for up to 2 hrs in a microplate reader, with excitation and emission wavelengths of 485 and 520 nm, respectively.  The  maximum fluorescence control was demonstrated by the addition of 0.5% Triton X-100.  189  CHAPTER 6  6.6 References 1. 2. 3. 4. 5.  6. 7. 8. 9. 10. 11. 12.  13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.  Zasloff, M. ature. 2002. 415, 389-95. Garlapati, S., Facci, M., Polewicz, M., Strom, S., Babiuk, L. A., Mutwiri, G., Hancock, R. E., Elliott, M. R. and Gerdts, V. Vet Immunol Immunopathol. 2009. 128, 184-91. Easton, D. M., Nijnik, A., Mayer, M. L. and Hancock, R. E. Trends Biotechnol. 2009. 27, 582-90. Hancock, R. E. and Sahl, H. G. at Biotechnol. 2006. 24, 1551-7. Salmon, A. L., Cross, L. J., Irvine, A. E., Lappin, T. R., Dathe, M., Krause, G., Canning, P., Thim, L., Beyermann, M., Rothemund, S., Bienert, M. and Shaw, C. J Biol Chem. 2001. 276, 10145-52. Powers, J. P. and Hancock, R. E. Peptides. 2003. 24, 1681-91. Ganz, T. at Rev Immunol. 2003. 3, 710-20. Ghosh, D., Porter, E., Shen, B., Lee, S. K., Wilk, D., Drazba, J., Yadav, S. P., Crabb, J. W., Ganz, T. and Bevins, C. L. at Immunol. 2002. 3, 583-90. Wilson, C. L., Ouellette, A. J., Satchell, D. P., Ayabe, T., Lopez-Boado, Y. S., Stratman, J. L., Hultgren, S. J., Matrisian, L. M. and Parks, W. C. Science. 1999. 286, 113-7. Agerberth, B., Charo, J., Werr, J., Olsson, B., Idali, F., Lindbom, L., Kiessling, R., Jornvall, H., Wigzell, H. and Gudmundsson, G. H. Blood. 2000. 96, 3086-93. Zeya, H. I. and Spitznagel, J. K. Science. 1963. 142, 1085-7. Wehkamp, J., Salzman, N. H., Porter, E., Nuding, S., Weichenthal, M., Petras, R. E., Shen, B., Schaeffeler, E., Schwab, M., Linzmeier, R., Feathers, R. W., Chu, H., Lima, H., Jr., Fellermann, K., Ganz, T., Stange, E. F. and Bevins, C. L. Proc atl Acad Sci U S A. 2005. 102, 18129-34. Bals, R., Wang, X., Wu, Z., Freeman, T., Bafna, V., Zasloff, M. and Wilson, J. M. J Clin Invest. 1998. 102, 874-80. Goldman, M. J., Anderson, G. M., Stolzenberg, E. D., Kari, U. P., Zasloff, M. and Wilson, J. M. Cell. 1997. 88, 553-60. Harder, J., Bartels, J., Christophers, E. and Schroder, J. M. ature. 1997. 387, 861. Stolzenberg, E. D., Anderson, G. M., Ackermann, M. R., Whitlock, R. H. and Zasloff, M. Proc atl Acad Sci U S A. 1997. 94, 8686-90. Tang, Y. Q., Yuan, J., Osapay, G., Osapay, K., Tran, D., Miller, C. J., Ouellette, A. J. and Selsted, M. E. Science. 1999. 286, 498-502. Zanetti, M. J Leukoc Biol. 2004. 75, 39-48. Finlay, B. B. and Hancock, R. E. W. at Rev Micro. 2004. 2, 497-504. Porcelli, F., Verardi, R., Shi, L., Henzler-Wildman, K. A., Ramamoorthy, A. and Veglia, G. Biochemistry. 2008. 47, 5565-72. Bowdish, D. M., Davidson, D. J., Speert, D. P. and Hancock, R. E. J Immunol. 2004. 172, 3758-65. Filewod, N. C., Pistolic, J. and Hancock, R. E. FEMS Immunol Med Microbiol. 2009. 56, 233-40. Scott, M. G., Davidson, D. J., Gold, M. R., Bowdish, D. and Hancock, R. E. J Immunol. 2002. 169, 3883-91. Niyonsaba, F., Iwabuchi, K., Someya, A., Hirata, M., Matsuda, H., Ogawa, H. and Nagaoka, I. Immunology. 2002. 106, 20-6.  190  CHAPTER 6  25.  26. 27. 28. 29. 30. 31.  32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.  45. 46. 47. 48. 49.  Mookherjee, N., Brown, K. L., Bowdish, D. M., Doria, S., Falsafi, R., Hokamp, K., Roche, F. M., Mu, R., Doho, G. H., Pistolic, J., Powers, J. P., Bryan, J., Brinkman, F. S. and Hancock, R. E. J Immunol. 2006. 176, 2455-64. Bowdish, D. M., Davidson, D. J., Lau, Y. E., Lee, K., Scott, M. G. and Hancock, R. E. J Leukoc Biol. 2005. 77, 451-9. Cross, L. J., Ennis, M., Krause, E., Dathe, M., Lorenz, D., Krause, G., Beyermann, M. and Bienert, M. Eur J Pharmacol. 1995. 291, 291-300. Vlasak, R., Unger-Ullmann, C., Kreil, G. and Frischauf, A. M. Eur J Biochem. 1983. 135, 123-6. Bowdish, D. M., Davidson, D. J., Scott, M. G. and Hancock, R. E. Antimicrob Agents Chemother. 2005. 49, 1727-32. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem. 2000. 267, 5330-41. Nijnik, A., Madera, L., Ma, S., Waldbrook, M., Elliott, M. R., Easton, D. M., Mayer, M. L., Mullaly, S. C., Kindrachuk, J., Jenssen, H. and Hancock, R. E. J Immunol. 2010. 184, 2539-50. Lee, J. G., Lee, S. H., Park, D. W., Lee, S. H., Yoon, H. S., Chin, B. R., Kim, J. H., Kim, J. R. and Baek, S. H. Cell Signal. 2008. 20, 105-11. Kindrachuk, J., Jenssen, H., Elliott, M., Townsend, R., Nijnik, A., Lee, S. F., Gerdts, V., Babiuk, L. A., Halperin, S. A. and Hancock, R. E. Vaccine. 2009. 27, 4662-71. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids. 2003. 122, 107-20. Cheng, J. T., Hale, J. D., Elliot, M., Hancock, R. E. and Straus, S. K. Biophys J. 2009. 96, 552-65. Johnson, W. C. Proteins. 1999. 35, 307-12. Provencher, S. W. and Glockner, J. Biochemistry. 1981. 20, 33-7. Sreerama, N., Venyaminov, S. Y. and Woody, R. W. Protein Sci. 1999. 8, 370-80. Sreerama, N. and Woody, R. W. Anal Biochem. 1993. 209, 32-44. Sreerama, N. and Woody, R. W. J Mol Biol. 1994. 242, 497-507. Pan, Y. L., Cheng, J. T. J., Hale, J. D., Pan, J., Hancock, R. E. W. and Straus, S. K. Biophys J. 2007. 92, 2854-64. Antcheva, N., Morgera, F., Creatti, L., Vaccari, L., Pag, U., Pacor, S., Shai, Y., Sahl, H. G. and Tossi, A. Biochem J. 2009. 421, 435-47. Ciornei, C. D., Sigurdardottir, T., Schmidtchen, A. and Bodelsson, M. Antimicrob Agents Chemother. 2005. 49, 2845-50. van der Kraan, M. I. A., Nazmi, K., Teeken, A., Groenink, J., van 't Hof, W., Veerman, E. C. I., Bolscher, J. G. M. and Nieuw Amerongen, A. V. Biological Chemistry. 2005. 386, 137-142. Hancock, R. E. and Diamond, G. Trends Microbiol. 2000. 8, 402-10. Lazaridis, T. Proteins. 2005. 58, 518-27. Dathe, M., Nikolenko, H., Klose, J. and Bienert, M. Biochemistry. 2004. 43, 9140-50. Marion, D. and Wuthrich, K. Biochem Biophys Res Commun. 1983. 113, 967-74. Hwang, T. L. and Shaka, A. J. Journal of Magnetic Resonance Series A. 1995. 112, 275279.  191  CHAPTER 7  CHAPTER 7: The importance of solvent accessibility in the mechanism of action of the lipopeptide daptomycin  7.1 Introduction In recent years, lipopeptides have received much attention as a novel class of antibiotics because they are typically highly active against multi-resistant bacteria. A number of lipopeptides consist of cationic amphiphilic peptides with an acylated N-terminus (C8-C18 fatty acid chain length) (1-6). Examples include members of the polymyxin family, which have been studied extensively (7-11). In the anionic lipopeptide class, the first naturally occurring member to be discovered more than fifty years ago was amphomycin (12). Additional members of this class of compounds include crystallomycin (13,14), aspartocin (15-18), glumamycin (19-21), laspartomycin (22), tsushimycin (23,24), and, by far the most-studied, daptomycin (25-30). Because of their unique composition, lipopeptides function in a manner that is atypical for most antibacterial agents, making them a unique class of antibiotics. Daptomycin is a cyclic anionic tridecapeptide, with a number of D-amino acids, 3 uncommon amino acids, and an N-terminus that is acylated with an n-decanoyl fatty acid side chain. It requires high concentrations of calcium (greater than 1000 times the daptomycin concentration), in the form of Ca2+, for activity (31-33). The mode of action of daptomycin is still not fully characterized (for recent work and reviews see (30,34-39)), but many of the initial and later steps in the mechanism are known. In a first step, daptomycin is combined with calcium in solution in a 1:1 Ca2+/daptomycin molar ratio to form an oligomer. This has recently been confirmed using equilibrium sedimentation (40) and solution-state NMR (40-42) experiments, 192  CHAPTER 7 2+  where upon addition of one molar equivalent of Ca , a micelle consisting of 14 ~ 16 daptomycin molecules was formed. Next, the daptomycin/Ca2+ complex interact with bacterial membranes and perturb the latter. This process is likely mediated by a high concentration of calcium ions, which are known to interact more strongly with negatively charged lipid headgroups than do, say magnesium ions (43). DSC and solid-state 31P NMR studies of daptomycin in the presence of 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine  and  1-palmitoyl-2-oleoyl-sn-glycero-3-  phospho-rac-1-glycerol (POPC/POPG) (1:1) lipid bilayers and Ca2+ have demonstrated the ability of daptomycin to perturb membranes (44) by inducing positive curvature strain on the lipids. Specifically, the data clearly indicate that the addition of daptomycin had a strong effect on the structural organization of acidic membranes in a manner dependent on Ca2+. At the late stages of the mechanism, daptomycin is proposed to oligomerize in the membrane. As a result, an efflux of potassium from the bacterial cell occurs, which in turn is proposed to lead to cell death, as this loss of potassium leads to dysfunction of macromolecular synthesis (36). It is, however, also possible that daptomycin aggregation in the membrane would simply interfere with membrane-associated processes, including synthesis of cell wall components, energetics, cell division, and so on. Recent studies on the structure of daptomycin at different stages in the mechanism have suggested that the structure of the lipopeptide does not change significantly and that the peptide shows changes in dynamic in going from one step to another. In order to understand the importance of mobility in relation to its rapid bactericidal activity, we have further investigated the dynamics/solvent accessibility of daptomycin. In this chapter, we will briefly discuss the different NMR structures of daptomycin which have been elucidated in the apo-, Ca2+- and DHPC/Ca2+-forms. In subsequent sections, we will present recent solution-state NMR data obtained for daptomycin under the three conditions listed above. The implications of our results  193  CHAPTER 7  for the intermediate steps in the mechanism of action of daptomycin will be discussed in the following sections.  7.2 Results and discussion 7.2.1 Daptomycin structure The NMR structure of daptomycin in the apo-form has been solved by three different research groups, each proposing a considerably different structure for daptomycin (41,42,45). The backbone Cα RMSD between the 1T5M consensus structure (45) and 1XT7 (42) is 3.2 Å. The corresponding RMSDs between 1T5M and the Rotondi and Gierasch structures (> 4 Å, for each of the six structures reported in (41)) are also large. The differences are even larger if the lipid tail and side chains are taken into account. In (44), our lab suggested that this wide spread of proposed structures may be related to the fact that daptomycin is very dynamic in this form. Indeed, if a peptide is highly flexible then the number of observable long-range NOEs is reduced. As a consequence, while it is possible to satisfy the observed experimental NOE restraints with a small set of protein structures, any such set cannot be truly representative of the sampled phase space in solution (46). The structure of daptomycin in the presence of calcium was first determined by Jung et al. (45). They later demonstrated that when Ca2+ is added in a 1:1 Ca2+/daptomycin molar ratio  that a micelle is formed and that the formation of this aggregate is not accompanied by a change in the structure of daptomycin (47). Indeed, when the divalent cation magnesium is added to daptomycin, the NOESY spectra closely resemble the NOESY spectra obtained for apodaptomycin. Furthermore, re-interpreting the appearance of new NOE cross-peaks in the NOESY spectra of daptomycin in the presence of 1 equivalent of Ca2+ as being due to intermolecular contacts in the daptomycin micelle resulted in new structures, which were also 194  CHAPTER 7  quite similar to the apo-form (40). This suggests that the interaction of daptomycin with Ca2+ is weak and non-perturbing. Given these findings, we proposed that the first step in the mechanism of action of daptomycin is the formation of a loose micelle, which serves to deliver daptomycin to the bacterial membrane (30) in a “detergent-like” form. This form would have a large membrane-disruptive potential, thereby allowing daptomycin to insert into the membrane rapidly and effectively. Finally, the structure of daptomycin in the presence of DHPC micelles and Ca2+ is an extended ring, much like apo-daptomycin (44). A comparison of the backbone Cα RMSDs between the representative structure of daptomycin in lipid micelles and the structures of apodaptomycin yields values (3.0 ~ 4.7 Å) that are similar to the RMSDs obtained when comparing the different apo-daptomycin structures amongst themselves (see above). Since the RMSDs are smallest when calculated relative to 1T5M and 1XT7 and since the NOE violations calculated relative to the apo-form dataset are lowest, it was suggested that daptomycin undergoes only a minor conformational rearrangement upon binding with DHPC in the presence of Ca2+. This would indicate that daptomycin has a high degree of plasticity, allowing it to readily adapt to one environment or another at each step in the mechanism.  7.2.2 Daptomycin dynamics/solvent accessibility In order to test whether daptomycin in the apo-form is highly dynamic, we performed hydrogen/deuterium (H/D) exchange experiments. A sample consisting of 1.5 mM daptomycin was prepared in 100 mM KCl buffer (volume = 500 µl), as previously described (45). A comparison of a 1D spectrum of 1.5 mM daptomycin in KCl buffer, 92% H2O/8% D2O with the spectra acquired in the presence of 100% D2O showed that the amide hydrogens exchanged immediately, as shown in Figure 7.1.  195  CHAPTER 7  1 0 0 % D 2 O , 7 5 m in 1 0 0 % D 2 O , 2 0 m in 1 0 0 % D 2 O , 1 5 m in  1 0 0 % D 2 O , 5 m in 1 0 0 % D 2 O , 0 m in 8 % D 2O , C o n tro l 10.0  5.0  0.0  ppm (t1)  Figure 7.1.  1D 1H MR spectra of apo-daptomycin in 100 mM KCl buffer: The fast H/D exchange could be observed immediately after the addition of D2O at 0 minute. HN signals (8 ~ 9 ppm) disappeared and were not recovered at 5, 15, 20 and 75 minutes after the addition of D2O.  No amide signals were detected at the 5 minute time point or later. It is interesting to note that a similar experiment performed on a 1.5 mM daptomycin sample, in the presence of 100 mM KCl, 0.2 mM EDTA, and 5 mM CaCl2 (conditions very close to the Ca2+-daptomycin sample used in other studies (45,47)) also resulted in the amide H’s exchanging within the first 5 minute time point, which is shown in Figure 7.2. This indicates that both the apo- and Ca2+-forms of daptomycin were possibly dynamic and/or adopted structures where the amides were fully solvent exposed.  196  CHAPTER 7  1 0 0 % D 2 O , 7 5 m in  1 0 0 % D 2 O , 2 0 m in 1 0 0 % D 2 O , 1 5 m in  1 0 0 % D 2 O , 5 m in 1 0 0 % D 2 O , 0 m in 8 % D 2O , C o n tro l 10.0  5.0  0.0  ppm (t1)  Figure 7.2.  1D 1H MR spectra of Ca2+-daptomycin in 100 mM KCl buffer: The fast H/D exchange could be observed immediately after the addition of D2O at 5 minutes. HN signals (8 ~ 9 ppm) disappeared and were not recovered at 5, 15, 20 and 75 minutes after the addition of D2O.  7.2.3 Daptomycin location in DHPC micelles in the presence of Ca2+ In order to understand the intermediate steps in the mechanism of action of daptomycin, it is important to determine where daptomycin is located in a lipid membrane bilayer. Given that daptomycin was found to perturb 1:1 POPC/POPG lipid bilayers (44) by affecting the structural organization of acidic membranes in a manner dependent on Ca2+, it would be most biologically relevant to determine the location daptomycin under these conditions. Unfortunately, in the presence of POPC/POPG and Ca2+, daptomycin causes membrane fusion (44), making the use of many of the methods to measure insertion depth of a peptide into membrane bilayers (for example, spin label measurements (48), fluorescence (49), paramagnetic Mn2+ (50), and NOESY 197  CHAPTER 7  spectroscopy (51)) difficult. As discussed in (44), however, it is possible to use DHPC in the presence of Ca2+ as a reasonable surrogate for membrane bilayers that are critical determinants of daptomycin action. We have, therefore, chosen in a first step to determine the location of daptomycin in DHPC micelles using solution-state NMR. Two daptomycin/DHPC/Ca2+ samples were prepared. The first was as described in (44) and consisted of 1.5 mM daptomycin, 75 mM DHPC-d40, and buffer consisting of 100 mM KCl, 0.2 mM EDTA, 5mM CaCl2, pH 6.70, in 92% H2O/8 % D2O (total volume of 500 µl). The second sample was identical, except that rather than using 75 mM DHPC-d40, 67.5 mM DHPCd40 and 7.5 mM DHPC (protonated) were used (52). The spectrum obtained for the first sample is shown in Figure 7.3(a). The slightly lower number of cross-peaks found compared to those in the NOESY spectrum reported in (44) are attributed to the lower temperature used here. The assignment of the cross-peaks was verified by comparing the spectrum in Figure 7.3(a) with the one in (44). In the presence of 10% protonated DHPC, additional NOE cross-peaks were found between daptomycin and the lipid resonances (53). These are indicated in Figure 7.3(b). As can be seen, new NOEs were found between DHPC headgroups and D-Ala8, Gly10, Ser11 and MeGlu12. Additional NOEs were also found between DHPC acyl chains and Trp1, Asp3, Orn6, Asp7 and Kyn13.  This indicates that daptomycin molecules were inserted into DHPC  membrane, with the residues closer to the decanoyl chain possibly inserted deeper into the hydrophobic core, and the residues closer to the outer ring likely more exposed to contact with the headgroups.  198  CHAPTER 7  Figure 7.3.  1  H OESY spectra of Ca2+-daptomycin in 100 mM KCl buffer in (a) DHPC-d40 and (b) DHPCd40/DHPC (9:1 mol/mol):  The spectra showed selected examples of additional NOEs found between daptomycin molecules and DHPC membrane. The spectra suggested that daptomycin molecules were in contact with and likely inserted into DHPC membrane.  7.2.4 Daptomycin in the presence of Mn2+ Mn2+ is a paramagnetic compound that is able to quench an NMR signal through paramagnetic relaxation enhancement (PRE) effect. Depending on the exposure to the aqueous surroundings, signals from different residues would be quenched at different rates. We compared the rate of quenching of the signal for daptomycin under three different conditions:  apo-  daptomycin, Ca2+-daptomycin, Ca2+-daptomycin in DHPC. If daptomycin inserts into the membrane (as a monomer or as an oligomer), then we expect to see a difference in the Mn2+quenching rate of 1H signals between Ca2+-daptomycin and Ca2+-daptomycin in DHPC. 199  CHAPTER 7 1  2+  Figure 7.4 shows H signals of selected NOEs as a function of Mn  concentration with  respect to daptomycin.  Figure 7.4.  1  H signals of selected intra- (left panel) and inter-residue (right panel) various conditions:  OEs of daptomycin in  (a) & (d) apo-daptomycin; (b) & (e) Ca2+-daptomycin ; (c) & (f) Ca2+-daptomycin in DHPC. All spectra were acquired at 35°C. The molar amount of Mn2+was titrated into the sample and was negligible in volume. Y-axis units are arbitrary.  In the absence of Ca2+, intra-residue NOE intensities decreased drastically as the Mn2+ concentration increased. This indicates that daptomycin molecules did not form tightly bound 200  CHAPTER 7  aggregates, with residues mostly exposed to the aqueous environment. The consistent decrease in NOE intensities also suggests that the residues inspected were likely to be equally affected by the Mn2+ PRE effect. Different inter-residue NOE intensities, on the other hand, were affected differently depending on the contact. This indicates that daptomycin molecules were in the highly dynamic state, which was consistent with observations from H/D exchange experiments. In the presence of Ca2+, both intra- and inter-residue NOE intensities decreased less dramatically, and the decreases were at consistent rates among most NOEs. This indicates that Ca2+ presence led to loosely bound daptomycin aggregates. Mn2+ had less accessible surface, which resulted in relatively similar NOE intensity decay curves. However, since Mn2+ could still access the peptide ring, the oligomeric structure would most likely be not very tightly bound. At low Mn2+concentrations, all NOE intensities showed less consistent rates. One intra- (4N, 4β) and one inter-residue (5N, 4β) NOE exhibited a slightly slower intensity decrease at Mn2+ concentration greater than 15%. This indicates that Thr4 was slightly more shielded from the Mn2+ PRE effect compared to other residues in the Ca2+-form. This also suggests that Ca2+daptomycin molecules may undergo some minor structural rearrangement which shielded Thr4 more from the aqueous surrounding. When Ca2+-daptomycin oligomers were exposed to DHPC micelles, all NOE intensities showed a more gradual decrease as Mn2+concentration increased. The rates of the decrease in intensity were similar and consistent for all NOEs. This indicates that daptomycin molecules were more shielded from the Mn2+ PRE effect in the presence of DHPC. This suggests that daptomycin inserted in DHPC membranes and were less exposed to the aqueous environment. (4N, 4β) and (5N, 4β) again exhibited a slow, less drastic intensity drop than other NOEs. (5N, 4γ), in contrast, did not show any decrease in intensity. This indicates that Thr4 could be more inserted in DHPC membrane and thus more shielded from the solvent.  201  CHAPTER 7  Figure 7.5.  Individual intra- (left panel) and inter-residue (right panel) OE intensity as a function of Mn2+ concentration: Apo-daptomycin (blue), Ca2+-daptomycin (red), and Ca2+-daptomycin in DHPC (green). The plots showed that both apo- and Ca2+-daptomycin were more exposed to the Mn2+ PRE effect and Ca2+daptomycin oligomers were most likely loosely bound. Ca2+-daptomycin oligomers in the presence of DHPC were, on the other hand, less influenced by the Mn2+ PRE effect. This suggests that Ca2+daptomycin oligomers may be inserted in DHPC membranes and were therefore less exposed to the aqueous environment. Y-axis units are arbitrary. 202  CHAPTER 7 2+  The responses of individual NOE intensities to the Mn PRE effect were also examined. In general, the intensity decrease was faster for apo-daptomycin, followed by Ca2+-daptomycin, and finally the slowest for Ca2+-daptomycin in DHPC. Figure 7.5 shows the intensity decrease of selected NOEs as a function of Mn2+ concentration. For all the examined NOEs, the decreases in intensity were more similar between apo- and Ca2+-daptomycin, but slightly less sharp for Ca2+-daptomycin (intra-residue NOEs), and much more gradual for Ca2+-daptomycin in DHPC. This indicates that Ca2+-daptomycin molecules were loosely bound (faster intensity decreases for inter-residue NOEs) and were inserted into membranes in the presence of DHPC. If Ca2+daptomycin molecules extracted membrane lipids (micellarization), then peptide-lipid contact should still be observed, but the shielding effect would be less. In the presence of DHPC membrane, four intra-residue NOEs ((12N, 12γ2*), (11N, 11β*), (10N, 10α) and (8N, 8β)) showed generally similar intensity decreases, whereas (4N, 4β) showed much slower intensity decrease. Similarly, four inter-residue NOEs ((13N, 12γ2*), (9N, 8β), (7N, 6β) and (5N, 4β)) exhibited similar intensity decreases with respect to each other and intra-residue NOEs, whereas (5N, 4γ) did not display an intensity decrease at all Mn2+ concentrations. This suggests again that Thr4 was highly protected from the aqueous surrounding.  7.3 Implications for the mechanism of action of daptomycin and general conclusions Much remains to be understood about how the rapid bactericidal activity of daptomycin is related to its dynamic nature and its interaction with the cytoplasmic membrane and whether oligomerization in the membrane is crucial. Previous studies have demonstrated that daptomycin molecules interact with lipid membranes, but the exact nature of this interaction has not been clarified fully (49). We have shown that daptomycin molecules are indeed dynamic/solventaccessible and possibly loosely bound in Ca2+-form until exposed to lipid environment. We have 203  CHAPTER 7  also demonstrated that daptomycin molecules are in contact with DHPC micelles and possibly insert into lipid membranes. These results may provide some insight on the initial steps of the peptide-membrane interaction. The intermediate to later steps of the peptide-membrane interaction are yet to be elucidated. We will attempt to answer two follow-up questions: 1. Do most/all parts of daptomycin molecules insert into lipid membranes? 2. Do daptomycin molecules oligomerize in lipid membranes? The first question deals with how daptomycin molecules insert into lipid membranes. We have shown that daptomycin molecules were in close contact with DHPC membrane, but these results did not pinpoint the actual location of daptomycin molecules in lipid membranes. In order to determine the insertion depth of selected regions of daptomycin molecules, we propose to conduct rotational-echo double resonance (REDOR) experiments. REDOR is a solid-state NMR technique previously developed to measure the distance between two nuclei (54), using additional π pulses to re-introduce dipolar coupling under magic angle spinning (MAS) condition. Thus, the distances between selected labeled sites can be measured using REDOR dephasing curves. For example, by measuring the distances between the lipid  31  P headgroups  and labeled residues (e.g. 15N labeled) on the peptide, the insertion depth of daptomycin can be assessed quantitatively. The second question deals with whether daptomycin molecules insert into lipid membranes as monomers or oligomers. In particular, it is yet to be clarified whether Ca2+daptomycin oligomers fall apart, insert, and oligomerize in lipid membranes, or insert directly as oligomers in the intermediate to later part of the daptomycin mechanism (30). In order to determine the oligomerization state of daptomycin in membrane bilayers, we propose to conduct solid-state 19F centerband-only detection of exchange (CODEX) (55) experiment. A recent study has also demonstrated that the oligomerization state of antimicrobial peptide PG-1 could be 204  CHAPTER 7  determined using  19  F CODEX technique (56). Other techniques may also be used to further  characterize the intermolecular distances in these oligomers, such as 13CO-19F REDOR study of K3, a synthetic analog of PGLa (57), and 13C-19F/1H-13C/15N-13C REDOR study of PG-1 (58). These additional NMR studies may be able to provide some further insights on the dynamics of daptomycin in its interaction with lipid membranes. We hope by answering the two questions above, we will gain a more comprehensive understanding of daptomycin-lipid interactions and further uncover the antimicrobial mechanism of daptomycin in the future.  7.4 Materials and methods 7.4.1 Materials Daptomycin was a generous gift from Cubist Pharmaceuticals (Lexington, MA, USA). Calcium chloride, potassium chloride and ethylenediaminetetraacetic acid (EDTA) were obtained from Fisher Scientific (Fair Lawn, New Jersey, USA). Manganese chloride solution was obtained from Sigma Aldrich (St. Louis, MO, USA).  1,2-Dihexanoyl-sn-Glycero-3-  Phosphocholine (DHPC) was obtained from Avanti Polar Lipids (Alabaster, AL, USA). DHPCd40 was obtained from Cambridge Isotope Laboratories (Andover, MA, USA).  7.4.2 Methods 7.4.2.1 MR sample preparation Apo-daptomycin samples were prepared by adding 1.2 mg of daptomycin (final concentration = 1.5 mM) to 100 mM KCl (pH 7) and 8 % D2O in a total volume of 500 µl. Ca2+daptomycin samples were prepared in the same manner but included 0.2 mM EDTA (pH 8) and 5 mM CaCl2 (pH 7). For 1:50 daptomycin/DHPC-d40 (mol/mol) samples, 18.5 mg of DHPC-d40 was weighed and transferred to an Eppendorf tube, dissolved in MeOH, dried using N2 (g), and 205  CHAPTER 7  finally vacuum dried overnight. For the same sample containing 10 molar % protonated DHPC (16.7 mg of DHPC-d40 and 1.7 mg of DHPC) were weighed, co-dissolved in MeOH and dried using N2 (g) and under vacuum overnight. The same amount of daptomycin and reagents were added as for Ca2+-daptomycin samples.  7.4.2.2 Hydrogen/deuterium (H/D) exchange For H/D exchange experiments, the apo- and Ca2+-daptomycin samples were prepared as above. The samples were lyophilized overnight to remove ddH2O prior to experiments. Each sample was added 500 µl of D2O, vortexed, spun down, and transferred to an NMR tube immediately (~ 5 min). One-dimensional solution-state NMR data were acquired at 0, 5, 15, 20 and 75 min intervals to observe H/D exchange.  7.4.2.3 Mn2+ titration For Mn2+ titration experiments, all samples were vortexed briefly and then sonicated for at least an hour prior to transferring to an NMR tube. Successful addition of Mn2+ was achieved by adding 0.375 µl of 0.1 M MnCl2 stock solution (5 molar % increment of Mn2+) and 0.15 µl of 1.0 M MnCl2 stock solution (20 molar % increment of Mn2+) to ensure minimal dilution effect. After each addition, the NMR tube was gently turned upside down for 2 ~ 3 times to mix thoroughly. Two-dimensional solution-state NOESY data were acquired at 0 %, 5 %, 10 %, 15 %, 20 %, 40 % and 60 % Mn2+ (mol/mol) intervals to observe the paramagnetic relaxation effect.  7.4.2.4 MR spectroscopy Both one- and two-dimensional solution-state NMR data were acquired on a Bruker 500MHz instrument (Milton, Ontario, Canada), operating at a 1H frequency of 499.981 MHz. All spectra were collected at 35°C.  2D spectra were acquired using nuclear Overhauser 206  CHAPTER 7  enhancement spectroscopy (NOESY) experiments in phase-sensitive mode using time proportional phase incrementation (TPPI) in the indirect dimension. The two-dimensional data set consisted of 4096 data points in t2 and 256 points in t1. The NOESY experiment was acquired with a mixing time of 150 ms and used excitation sculpting with gradients for water suppression. Signals were averaged using 128 scans for the NOESY experiments. The spectra were referenced to D2O (4.800 ppm). Spectra were processed to result in 1k ×1k points. Two-dimensional  NOESY  spectra  for  1:50  daptomycin/DHPC-d40  and  1:50  daptomycin/DHPC-d40/DHPC (9:1 mol/mol) samples were recorded on a Varian Unity 500 spectrometer operated by the UBC Laboratory for Molecular Biophysics. Homonuclear NOESY (τm = 150 ms) (59) spectra were collected at 308K. WATERGATE (60) was used for water suppression and the data matrices were 4k by 256. The spectra were referenced to external DSS and processed to yield a 4k x 256 point matrix.  207  CHAPTER 7  7.5 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.  Chu-Kung, A. F., Bozzelli, K. N., Lockwood, N. A., Haseman, J. R., Mayo, K. H. and Tirrell, M. V. Bioconjug Chem. 2004. 15, 530-5. Lockwood, N. A., Haseman, J. R., Tirrell, M. V. and Mayo, K. H. Biochem J. 2004. 378, 93-103. Wakabayashi, H., Matsumoto, H., Hashimoto, K., Teraguchi, S., Takase, M. and Hayasawa, H. Antimicrob Agents Chemother. 1999. 43, 1267-9. Mak, P., Pohl, J., Dubin, A., Reed, M. S., Bowers, S. E., Fallon, M. T. and Shafer, W. M. Int J Antimicrob Agents. 2003. 21, 13-9. Epand, R. M. Biopolymers. 1997. 43, 15-24. Thennarasu, S., Lee, D. K., Tan, A., Prasad Kari, U. and Ramamoorthy, A. Biochim Biophys Acta. 2005. 1711, 49-58. Zhang, L., Dhillon, P., Yan, H., Farmer, S. and Hancock, R. E. W. Antimicrob Agents Chemother. 2000. 44, 3317-21. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother. 1999. 43, 1317-23. Storm, D. R., Rosenthal, K. S. and Swanson, P. E. Annu Rev Biochem. 1977. 46, 723-63. Miller, I. R., Bach, D. and Teuber, M. J Membr Biol. 1978. 39, 49-56. Teuber, M. and Bader, J. Antimicrob Agents Chemother. 1976. 9, 26-35. Heinemann, B., Kaplan, M. A., Muir, R. D. and Hooper, I. R. Antibiot Chemother. 1953. 3, 1239-1242. Lomakina, N. N. and Brazhnikova, M. G. Biokhimiia. 1959. 24, 425-31. Gol'Dberg, L. E. Antibiotiki. 1959. 4, 63-7. Darken, M. A., Jensen, A. L. and Shu, P. Antibiot Annu. 1959. 7, 199-204. Kirsch, E. J., Dornbush, A. C. and Backus, E. J. Antibiot Annu. 1959. 7, 205-12. Redin, G. S. and Mc, C. M. Antibiot Annu. 1959. 7, 213-9. Shay, A. J., Adam, J., Martin, J. H., Hausmann, W. K., Shu, P. and Bohonos, N. Antibiot Annu. 1959. 7, 194-8. Fujino, M. Bull Chem Soc Jpn. 1965. 38, 517-22. Fujino, M., Inoue, M., Ueyanagi, J. and Miyake, A. Bull Chem Soc Jpn. 1965. 38, 515-7. Shibata, M., Kanzaki, T., Nakazawa, K., Inoue, M., Hitomi, H., Mizuno, K., Fujino, M. and Akira, M. J Antibiot (Tokyo). 1962. 15, 1-6. Naganawa, H., Hamada, M., Maeda, K., Okami, Y. and Takeushi, T. J Antibiot (Tokyo). 1968. 21, 55-62. Shoji, J. and Otsuka, H. J Antibiot (Tokyo). 1969. 22, 473-9. Shoji, J. I., Kozuki, S., Okamoto, S., Sakazaki, R. and Otsuka, H. J Antibiot (Tokyo). 1968. 21, 439-43. Hodinka, R. L., Jack-Wait, K., Wannamaker, N., Walden, T. P. and Gilligan, P. H. Eur J Clin Microbiol. 1987. 6, 100-3. Allen, N. E., Hobbs, J. N. and Alborn, W. E., Jr. Antimicrob Agents Chemother. 1987. 31, 1093-9. Bayer, A. S., Yih, J. and Hirano, L. Antimicrob Agents Chemother. 1988. 32, 141-3. Botha, P. L., van Heerden, C. J., Vorster, E. and Venter, E. C. S Afr Med J. 1988. 74, 168. Ehlert, F. and Neu, H. C. Eur J Clin Microbiol. 1987. 6, 84-90. Straus, S. K. and Hancock, R. E. W. Biochim Biophys Acta. 2006. 1758, 1215-1223.  208  CHAPTER 7  31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.  Andrew, J. H., Wale, M. C., Wale, L. J. and Greenwood, D. J Antimicrob Chemother. 1987. 20, 213-21. Chow, A. W. and Cheng, N. Antimicrob Agents Chemother. 1988. 32, 788-90. Eliopoulos, G. M., Willey, S., Reiszner, E., Spitzer, P. G., Caputo, G. and Moellering, R. C., Jr. Antimicrob Agents Chemother. 1986. 30, 532-5. Abbanat, D., Macielag, M. and Bush, K. Expert Opin Investig Drugs. 2003. 12, 379-99. Alder, J. D. Drugs Today (Barc). 2005. 41, 81-90. Silverman, J. A., Perlmutter, N. G. and Shapiro, H. M. Antimicrob Agents Chemother. 2003. 47, 2538-44. Hancock, R. E. W. Lancet Infect Dis. 2005. 5, 209-18. Steenbergen, J. N., Alder, J., Thorne, G. M. and Tally, F. P. J Antimicrob Chemother. 2005. 55, 283-8. Shah, P. M. Clin Microbiol Infect. 2005. 11 Suppl 3, 36-42. Teriete, P., Franzin, C. M., Choi, J. and Marassi, F. M. Biochemistry. 2007. Rotondi, K. S. and Gierasch, L. M. Biopolymers. 2005. 80, 374-85. Ball, L. J., Goult, C. M., Donarski, J. A., Micklefield, J. and Ramesh, V. Org Biomol Chem. 2004. 2, 1872-8. Garidel, P. and Blume, A. Chem Phys Lipids. 2005. Scott, W. R., Baek, S. B., Jung, D., Hancock, R. E. and Straus, S. K. Biochim Biophys Acta. 2007. 1768, 3116-26. Jung, D., Rozek, A., Okon, M. and Hancock, R. E. W. Chem Biol. 2004. 11, 949-57. Daura, X., Antes, I., van Gunsteren, W. F., Thiel, W. and Mark, A. E. Proteins. 1999. 36, 542-55. Ho, S. W., Jung, D., Calhoun, J. R., Lear, J. D., Okon, M., Scott, W. R., Hancock, R. E. and Straus, S. K. Eur Biophys J. 2008. 37, 421-33. Liang, B., Bushweller, J. H. and Tamm, L. K. J Am Chem Soc. 2006. 128, 4389-97. Jung, D., Powers, J. P., Straus, S. K. and Hancock, R. E. W. Chemistry and Physics of Lipids. 2008. 154, 120-128. Buffy, J. J., Hong, T., Yamaguchi, S., Waring, A. J., Lehrer, R. I. and Hong, M. Biophys J. 2003. 85, 2363-73. Siarheyeva, A., Lopez, J. J. and Glaubitz, C. Biochemistry. 2006. 45, 6203-11. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry. 2006. 45, 5793-5799. Fernández, C. s., Hilty, C., Wider, G. and Wüthrich, K. Proceedings of the ational Academy of Sciences of the United States of America. 2002. 99, 13533-13537. Gullion, T. and Schaefer, J. Journal of Magnetic Resonance (1969). 1989. 81, 196-200. deAzevedo, E. R., Hu, W. G., Bonagamba, T. J. and Schmidt-Rohr, K. Journal of the American Chemical Society. 1999. 121, 8411-8412. Buffy, J. J., Waring, A. J. and Hong, M. J Am Chem Soc. 2005. 127, 4477-83. Toke, O., O'Connor, R. D., Weldeghiorghis, T. K., Maloy, W. L., Glaser, R. W., Ulrich, A. S. and Schaefer, J. Biophysical Journal. 2004. 87, 675-687. Mani, R., Tang, M., Wu, X., Buffy, J. J., Waring, A. J., Sherman, M. A. and Hong, M. Biochemistry. 2006. 45, 8341-8349. Jeener, J., Meier, B. H., Bachmann, P. and Ernst, R. R. J. Chem. Phys. 1979. 71, 45464553. Piotto, M., Saudek, V. and Sklenar, V. J Biomol MR. 1992. 2, 661-5.  209  CHAPTER 8  CHAPTER 8: Summary, conclusion, and future work  8.1 Thesis summary This thesis presented our work on the studies of the structure and function relationships of two members of the aurein 2 family cationic antimicrobial peptides (CAPs), aurein 2.2 and aurein 2.3 and their analogues. In addition to the studies on CAPs, this thesis also presented our initial work on the dynamics/solvent accessibility of daptomycin, a lipopeptide - another class of antimicrobial peptides. Chapter 1 provided the basic introduction on the history, classes, and categories of antibiotics. It further described the current problems of antibiotic resistance and how bacteria acquire resistance. Later in the chapter, antimicrobial peptides were introduced as a possible alternative to conventional antibiotics, given their limited resistance, relatively low cytotoxicity, and biodegradable properties. Given these unique properties, many studies have been conducted to characterize the structure-function relationships of these peptides, in order to develop an alternative antimicrobial arsenal to combat the continuously growing antibiotic resistance. In addition, Chapter 1 also introduced various methodologies commonly used to study the structure and peptide-membrane interaction dynamics/mechanism of antimicrobial peptides. These include different biophysical techniques such as CD and NMR spectroscopy (model system), as well as DiSC35 and MIC (intact bacteria) assays. These techniques allow us to elucidate the relationship between the sequence and the structure, and to understand the correlation between the structure and the antimicrobial mechanism, both qualitatively and quantitatively.  210  CHAPTER 8  Chapter 2 introduced the fundamental studies on the structure-function relationships of aurein 2.2 and aurein 2.3. Various biophysical studies were conducted to determine whether an L13I mutation could have an effect on the structure and function of these peptides significantly. In this chapter, we also investigated whether the nature of the C-terminus was the determining cause for the difference in peptide-membrane interaction and antimicrobial activity. In Chapter 3, we inspected the effect of lipid bilayer thickness/fluidity and molar PG content on the behaviour of peptides in model membranes. These results in model membranes were correlated with the results in intact bacteria to determine which mixture, namely 1:1 or 3:1 DMPC/DMPG (mol/mol) or 1:1 or 3:1 POPC/POPG (mol/mol), was a more relevant model membrane to study aurein peptides against Staphylococcus aureus. The peptide behaviour in different model membranes were also examined and discussed in details in Chapter 4, which demonstrated the importance of a relevant model membrane choice for studying a CAP and relating the data back to a specific bacterial strain. Chapter 5 and Chapter 6 presented studies on the importance of residue 13 and the N- and C-termini on the structure-function relationship of aurein 2.2. These studies were performed to investigate whether residue 13 was selective for only Leu and Ile or could be substituted for other amino acids. The importance of the N- and/or C-terminus was assessed by truncating either or both ends by three to six residues. In addition, we also examined the structure and function of daptomycin, a lipopeptide, in Chapter 7 to further our understanding on antimicrobial peptides. The peptide-membrane interaction and dynamics/solvent accessibility were characterized further in a lipid membrane environment.  Finally, with all the information  obtained from studying the aurein and daptomycin peptides, we now are in a better position for designing novel antimicrobial and/or immunomodulatory peptides to combat the increasing antibiotic resistance.  211  CHAPTER 8  8.2 Experimental conclusions Studying an antimicrobial peptide usually involves three steps:  characterizing the  structure, examining the peptide-membrane interaction, determining the activity/function in intact bacteria, as demonstrated by many studies (1-8). In this section, we will draw conclusions based on the experiments conducted both in model membranes and intact bacteria. We will conclude with discussions on three different perspectives: peptide structure, peptide-membrane interaction, and model membranes.  8.2.1 Perspectives on the peptide structure Most CAPs adopt any of the four secondary structures upon exposure to membrane environment: α-helix (e.g. magainin 2 (9)), β-sheet (e.g. arenicin (10)), complex extended structure, (protegrin-1 (11)), or extended chain (e.g. indolicidin (12)). In particular, aurein 1.2, a widely studied CAP from the aurein family, and citropin 1.1, another 16-residue CAP from the citropin family with similar length and sequence as aurein 2.2, were also found to adopt α-helical structures (4). It is known that many amphibian CAP families, for example aurein family (7), contain a variety of peptides that differ only by few residues. Whether this subtle difference in peptide sequence correlates with difference in peptide structure was not examined extensively. As most studies have drawn comparisons among CAPs from different classes or families, we have determined whether a conservative L13I mutation could result in structural difference between aurein 2.2 and aurein 2.3, both from the same family (7). We have found that this conservative mutation as well as the nature of the C-terminus did not change the structures of the two aurein peptides. The presence of a Leu residue at position 13 only resulted in a peptide structure with a slightly less % α-helix at low peptide concentrations.  Further structural  examination using solution-state 1H NMR spectroscopy indicated that all three aurein peptides  212  CHAPTER 8  adopted a continuous α-helix, like aurein 1.2 (7), not altered to a bent helix such as spinigerin (13) or helix-hinge-helix such as maculatin 1.1 (6).  We have demonstrated here that a  conservative residue mutation may not have a significant impact on the peptide structure but affects activity. This could be an important contribution to the studies on the structure-function relationship of various CAPs. Some studies have attempted to design more efficient analogues based on CAPs. These include changing the structure through cyclization, loop formation or oligomerization, as shown for a magainin 2 dimer analogue (14), or by having a different number of positively charged residues (15). To the best of our best knowledge, the essential residues or peptide segments for the aurein peptides from the structural perspective have not yet been clarified. We have found that residue 13-stubstitution did not change the peptide conformation significantly, but only varied the helical content. We have discovered that shortening the aliphatic side chain improved peptide folding significantly, whereas removing most of the aliphatic side chain or substituting with an aromatic ring did not. The importance of peptide length was also inspected by truncating four residues from the N-terminus and/or three or six residues from the C-terminus, respectively. Truncating the C-terminus by three residues did not change the secondary structure, but removing the C-terminus by six residues, on the other hand, had an impact on the helical content considerably. Truncating the N-terminus by four residues imposed even more drastic disruption on the structural formation of the peptide. These observations suggest that both steric and hydrophobicity factors played a role in peptide folding and the size of the side chain group had an impact on how well defined a structure the peptide could form. Additional observations also suggest that the four residues “G1L2F3D4” and three residues “G11A12L13” form the minimal hydrophobic N- and C-terminal segments required for the proper formation of an α-helix. This also suggests that an aurein 213  CHAPTER 8  peptide required more than ten residues to maintain a prominent helical content. These studies may contribute to a better understanding on whether substituting or removing certain residues would alter the peptide structure.  8.2.2 Perspectives on the peptide-membrane interaction Another important aspect in studying an antimicrobial peptide is to understand how it interacts with lipid membranes and whether differences in the sequence or structure can lead to differences in the peptide-membrane interaction. CAPs normally interact with lipid membranes via different mechanisms of action, including the carpet model (16,17), toroidal model (16), barrel-stave model (18), a micellar aggregate channel model (19,20) or a detergent-like mechanism (21). Previous studies have proposed that aurein 1.2 and citropin 1.1 perturb lipid membranes via carpet-like or detergent-like mechanism (1,21). The exact nature of membrane perturbation of aurein 2 family peptides has not been investigated comprehensively. Given the similar peptide sequence and length, these aurein peptides may disorder lipid membranes in similar mechanisms as aurein 1.2 and citropin 1.1. Here, we have found that both aurein 2.2 and aurein 2.3 interacted with lipid membranes as their primary target, via mechanisms that were dependent on the model membranes in which the studies were conducted i.e. carpet-/detergentlike mechanism in short-chain lipid membranes and possibly toroidal-like (or toroidal/liposomelike) mechanism in long-chain lipid membranes. The abilities of these aurein peptides to interact with lipid membranes were not greatly influenced by a conservative L13I mutation, but more by an amidated to carboxy C-terminal change. The effect of L13I mutation was more observable in the abilities of the aurein peptides to insert into lipid bilayers or induce membrane leakage. Very few studies have been conducted to examine the effect of residue substitution and terminal truncation on the peptide-membrane interaction. One of the most common practices is to increase the number of positive charges on a CAP, as shown by a previous study of PG-1 (22). 214  CHAPTER 8  We have demonstrated based on residue 13-substitution studies that a larger aliphatic side chain group was needed at position 13 for the peptide to insert/disorder the lipid bilayers more efficiently, but the degree of membrane perturbation was not selective (or strongly dependent) on the nature of the side chain group. In addition, we have also shown that C-terminal truncation by three residues did not have significant effect, but removing the C-terminus by six residues, in contrast, abolished the ability of the peptide to insert efficiently and disrupt the lipid bilayers extensively. Given that the peptide-membrane interaction was found to be selective on the Cterminus but not on residue 13, the effect of N-terminal truncation was then inspected. We have found that N-terminal truncation by four residues nearly abolished the antimicrobial function but gave rise to a highly enhanced immunomodulatory function of these peptides in vitro. As many CAPs were found to display immunomodulatory activities after proteolytic cleavage of their Nterminal region (23-25), we have demonstrated that these N-terminally truncated aurein mutant peptides may be possible candidates for novel immunomodulatory drug development. Overall, we have demonstrated that residue 13, N-terminal segment “G1L2F3D4” and Cterminal segment “G11A12L13” are essential for aurein peptides to interact with lipid membranes and exhibit antimicrobial activity.  These studies may add further information toward  understanding how/why a CAP is active against a bacterium.  8.2.3 Perspectives on the model membranes Choosing the relevant model membrane system to study an antimicrobial peptide is equally important for characterizing the structure-function relationship of a given antimicrobial peptide.  The choice of model membranes depends on various factors such as bilayer  thickness/fluidity (e.g. short 1,2-dimyristoyl chains (26) and long 1,2-dipalmitoyl chains (27)), lamellar to liquid crystalline phase transition temperature (Tm) (e.g. lower Tm of 1,2-dioleoyl chain (28)), molar PG content (e.g. 33% (5) or 50% (29)), and ease of sample preparation (e.g. 215  CHAPTER 8 31  DSC (30) and P NMR (31)). In terms of headgroups, these choices have transitioned from pure PC (4) to PC/PG (32) mixtures to PE/PG (33,34) mixtures. The effects of model membrane composition have been reported previously for protegrin-1 (35) and pardaxin (36), but not comprehensively for other CAPs. We have demonstrated that bilayer thickness/fluidity, PG content and headgroup choice have significant effects on the behaviour of aurein peptides in model membrane environment.  Different membrane insertion states and perturbation  mechanisms were discovered for the aurein peptides in different model membranes. Overall, we have demonstrated in these studies that it is important to study a CAP from the lipid membrane’s perspective, and choosing a relevant model membrane system is a must. These studies may offer additional references for one to select a model membrane to examine a specific CAP in the future.  8.2.4 Daptomycin We have further examined the mechanism of action of daptomycin from the lipopeptide family.  These peptides have been highly investigated recently and have shown promising  treatment in clinical environment (37,38). daptomycin has yet not been fully clarified.  However, the structure-function relationship of Here, we further investigated the initial and  intermediate steps in the peptide-membrane interaction, particularly the dynamic state/solvent accessibility of daptomycin molecules in the presence of lipid membranes (39). We discovered from paramagnetic relaxation enhancement (PRE) studies that daptomycin molecules were loosely bound together in Ca2+-bound aggregate form, but more tightly shielded in the membrane-bound form. This indicates that daptomycin molecules do indeed insert into lipid membranes and peptide-membrane contact was evident through additional NOEs found between lipid headgroups and daptomycin molecules. This also suggests that daptomycin molecules may  216  CHAPTER 8  not insert very deeply into lipid membranes but stayed mostly bound to or slightly inserted into the membrane surface.  8.3 Future work Structural and functional studies on aurein and daptomycin peptides have greatly helped reveal some of the unknowns on the structure-function relationships of these peptides. Using various biophysical tools to study synthetic antimicrobial peptides in model membrane systems greatly helps elucidating the link between the sequence and the structure and how this link modulates the function of the given antimicrobial peptide. Even though many studies were conducted to uncover structural diversities and various models were proposed to explain antimicrobial mechanisms of these peptides, much still remains to be studied to further characterize these peptides in order to develop a novel antibiotic in the future.  8.3.1 Structures and functions By using OCD spectroscopy, we were able to assess the ability of aurein peptides to insert into the lipid bilayers. However, as OCD experiments only provide information as peptide insertion profiles, the limitation of OCD spectroscopy does not allow the determination of the peptide insertion angle in the lipid bilayers. In order to determine the peptide insertion angle, we will need to use solid-state  15  N NMR spectroscopy. As  15  N-1H dipolar coupling is sensitive to  the orientation, the 15N signal will appear at different chemical shift values if the peptide lies on the surface of inserts into the lipid bilayers at different angles. Our preliminary 15N NMR results of aurein 2.2 with  15  N-labeled G11 and A12 at 1:15 P/L molar ratio in 4:1 DMPC/DMPG  (mol/mol) bilayers indicated that aurein 2.2 lied/inserted in multiple orientations (data not shown). This suggests that DMPC/DMPG bilayers were severely disordered in the presence of aurein 2.2, which was consistent with our observation that aurein 2.2 perturbed DMPC/DMPG 217  CHAPTER 8  bilayers through mechanisms similar to membrane micellarization. To further examine the insertion angle as a function of P/L molar ratio, we will conduct solid-state 1D 15N NMR and 2D 1  H-15N PISEMA experiments (40). These results may further confirm the OCD results that  aurein 2.3 showed a more gradual insertion profile than aurein 2.2, and aurein 2.3-COOH was unable to insert until at high peptide concentration (Chapter 3). In addition, these results can help us understand better how the peptide insertion angle is related to the peptide-membrane interaction, so that we can create another reference for designing a novel antibiotic in the future. Both OCD and 15N NMR spectroscopy do not provide information on the transmembrane region of an antimicrobial peptide in the lipid bilayers. In order to determine this region, we will conduct proton/deuterium (H/D) exchange experiments initially to determine which part of the peptide lies closer to the membrane surface or inserts deeply into the hydrophobic core. In order to pinpoint the actual peptide insertion depth, we will conduct REDOR experiments. REDOR is a powerful technique to determine the heteronuclear distance (41), by re-introducing dipolar coupling through additional π pulses under magic angle spinning (MAS) condition. Thus, selecting appropriate labeling sites will allow the distances between labels to be measured through REDOR dephasing curves. By measuring the distances between the lipid headgroups and labeled residues (for example, 15N labeled) on the peptide, the insertion depth of aurein and daptomycin can be assessed quantitatively. To further characterize the oligomerization state of aurein peptides, we will conduct solution-state 1H NOESY experiments to observe whether there are additional NOEs between aurein 2.2 molecules, or vice versa. Our preliminary results showed that there were indeed more NOEs present for aurein 2.2 in DPC micelles (data not shown).  We will examine this  observation further in different peptide concentrations and lipid membrane environment. If  218  CHAPTER 8  peptide aggregation state is directly correlated with the antimicrobial activity, here we would establish another useful reference for future design of a novel antibiotic. Immunomodulatory therapy has recently become very important as it has the potential to become a complementary therapeutic approach.  We observed that L13A mutant displayed  improved immunomodulatory function and relatively low cytotoxicity. Thus, L13A mutant is a possible candidate for novel immunomodulatory drug development. We will continue studying this peptide in a murine model to assess its ability to induce antibody responses against bacterial treatment.  8.3.2 Beyond antimicrobial peptides So far, most studies have been conducted to examine antimicrobial peptides alone. To the best of our knowledge, very few studies have looked at the synergistic effect of antimicrobial peptides with other biological molecules/chemical compounds. One example is a recent study of the N-terminal fragment of the frog skin peptide esculentin-1b [Esc(1–18)] in combination with clinically used antimicrobial agents evaluated against Stenotrophomonas maltophilia, which has shown an improved antimicrobial activity (42). Chemical penetration enhancers (CPEs) such as isopropyl myristate (IPM) (43) have been known to enhance transdermal delivery/topical absorption.  We may be able to include such a compound with topical application of an  antimicrobial peptide such as daptomycin to increase the drug effectiveness on bacterial infections on the skin, or transdermal antibiotic delivery into the host body. These further studies may provide more insights on better drug combinations to improve drug efficiency.  219  CHAPTER 8  8.4 Final remarks Taken together all these results and future studies, we hope that we have established and will continue expanding this set of references for the future design of new antibiotic and immunomodulatory therapeutics. Some of these references are summarized in Table 8.1. From these studies, we have established three general rules. First, structure or secondary structure content alone may not be used to assess whether a given antimicrobial peptide is active  Table 8.1.  Summary of the effects of peptide modification on the aurein peptides. The amino acids that exist in the peptide sequence are shown in green, whereas the ones that do not exist are shown in black. AM = antimicrobial; IM = immunomodulatory; T = toroidal pore/liposome formation.  Residue position Amino acid AM activity (S. aureus C622)  Aurein peptide sequence 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 G L F D I V K K V V G A L I A F V G S L  High High High High High Medium Low Low Low  IM activity (PBMCs)  Low Low Medium High High  Cytotoxicity High (PBMCs) Medium Medium Low Low  220  CHAPTER 8  Aurein peptides sequence Residue position  1  Amino acid  G L F D  Membrane insertion ability1  2 3 4 5 6 I  7  8  9 10 11 12  V K K V  V  G  A  13 L I A F V  14 15 16 G  S  L  High Medium Medium High High High Low Low Low  Membrane T perturbation T mechanism1 T T T T T Membrane leakage (S. aureus C622)2  High Medium Low Medium Medium Medium Medium  Secondary structure3  High High Medium Low High High Low Low Low  1  Based on OCD and 31P NMR results in POPC/POPG bilayers.  2  Based on DiSC35 assay results compared to gramicidin S.  3  Based on helical content at 1:100 P/L molar ratio in the presence of 1:1 POPC/POPG (mol/mol) SUVs.  221  CHAPTER 8  or marginally active against a specific bacterial strain. Second, residue specificity and peptide length/segment are critical to peptide-membrane interaction and activity/function. Last, specific peptide segments are essential for antimicrobial and/or immunomodulatory activity. Thus, when designing a new antibiotic based on an antimicrobial peptide, one has to be careful not to truncate an essential peptide segment, over-truncate the peptide sequence, or placing an irrelevant residue substitution. In addition, one must evaluate/justify the antimicrobial efficacy of a new antimicrobial peptide derivative based on multiple factors such as peptide-membrane interaction, membrane leakage induction, and membrane perturbation mechanism, but not on the structural factor alone. We have also established three general rules for choosing a model membrane to study a given antimicrobial peptide. First, it is important to choose a lipid/lipid mixture with relevant acyl chain length and headgroup ratio. Second, it is critical to select a lipid composition that best mimics the lipid membrane of a given bacterium. Last, various factors such as Tm and ease for sample preparation also need to be taken into account to compensate the limitation of using model membranes for studying antimicrobial peptides. Thus, when selecting a model membrane, one would be careful to choose a mixture that best represents the bacterium examined yet not to neglect biologically relevant condition and limitation. I hope that from this thesis, I can and will continue inspiring researchers and future generations to enter this field and dedicate themselves in this continuous and ongoing battle against bacterial infection and antibiotic resistance. I hope that this thesis has illustrated the importance of antimicrobial research and yes, it is possible to fight against the antibiotic resistance as long as we are willing to!  222  CHAPTER 8  8.5 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.  Ambroggio, E. E., Separovic, F., Bowie, J. H., Fidelio, G. D. and Bagatolli, L. A. Biophys J 2005, 89, 1874-81. Balla, M. S., Bowie, J. H. and Separovic, F. Eur Biophys J 2004, 33, 109-16. Hallock, K. J., Lee, D. K. and Ramamoorthy, A. Biophys J 2003, 84, 3052-60. Marcotte, I., Wegener, K. L., Lam, Y. H., Chia, B. C., de Planque, M. R., Bowie, J. H., Auger, M. and Separovic, F. Chem Phys Lipids 2003, 122, 107-20. Ramamoorthy, A., Thennarasu, S., Lee, D. K., Tan, A. and Maloy, L. Biophys J 2006, 91, 206-16. Chia, B. C., Carver, J. A., Mulhern, T. D. and Bowie, J. H. Eur J Biochem 2000, 267, 1894-908. Rozek, T., Wegener, K. L., Bowie, J. H., Olver, I. N., Carver, J. A., Wallace, J. C. and Tyler, M. J. Eur J Biochem 2000, 267, 5330-41. Wegener, K. L., Wabnitz, P. A., Carver, J. A., Bowie, J. H., Chia, B. C., Wallace, J. C. and Tyler, M. J. Eur J Biochem 1999, 265, 627-37. Wieprecht, T., Dathe, M., Krause, E., Beyermann, M., Maloy, W. L., MacDonald, D. L. and Bienert, M. FEBS Lett 1997, 417, 135-40. Andra, J., Jakovkin, I., Grotzinger, J., Hecht, O., Krasnosdembskaya, A. D., Goldmann, T., Gutsmann, T. and Leippe, M. Biochem J 2008, 410, 113-22. Fahrner, R. L., Dieckmann, T., Harwig, S. S., Lehrer, R. I., Eisenberg, D. and Feigon, J. Chem Biol 1996, 3, 543-50. Rozek, A., Friedrich, C. L. and Hancock, R. E. Biochemistry 2000, 39, 15765-74. Landon, C., Meudal, H., Boulanger, N., Bulet, P. and Vovelle, F. Biopolymers 2006, 81, 92-103. Tencza, S. B., Creighton, D. J., Yuan, T., Vogel, H. J., Montelaro, R. C. and Mietzner, T. A. J Antimicrob Chemother 1999, 44, 33-41. Dathe, M., Nikolenko, H., Meyer, J., Beyermann, M. and Bienert, M. FEBS Lett 2001, 501, 146-50. Matsuzaki, K. Biochim Biophys Acta 1999, 1462, 1-10. Shai, Y. Biochim Biophys Acta 1999, 1462, 55-70. Huang, H. W. Biochemistry 2000, 39, 8347-52. Hancock, R. E. W. and Chapple, D. S. Antimicrob Agents Chemother 1999, 43, 1317-23. Wu, M., Maier, E., Benz, R. and Hancock, R. E. W. Biochemistry 1999, 38, 7235-42. Bechinger, B. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1529-1539. Tang, M., Waring, A. J. and Hong, M. Biochimica et Biophysica Acta (BBA) Biomembranes 2009, 1788, 514-521. Ghosh, J. K., Shaool, D., Guillaud, P., Ciceron, L., Mazier, D., Kustanovich, I., Shai, Y. and Mor, A. J Biol Chem 1997, 272, 31609-16. Wilson, C. L., Ouellette, A. J., Satchell, D. P., Ayabe, T., Lopez-Boado, Y. S., Stratman, J. L., Hultgren, S. J., Matrisian, L. M. and Parks, W. C. Science 1999, 286, 113-7. Zanetti, M. J Leukoc Biol 2004, 75, 39-48. Lewis, R. N., Zhang, Y. P. and McElhaney, R. N. Biochim Biophys Acta 2005, 1668, 203-14. Zweytick, D., Pabst, G., Abuja, P. M., Jilek, A., Blondelle, S. E., Andra, J., Jerala, R., Monreal, D., Martinez de Tejada, G. and Lohner, K. Biochim Biophys Acta 2006, 1758, 1426-35. 223  CHAPTER 8  28. 29. 30. 31. 32. 33.  34. 35. 36. 37. 38.  39. 40. 41. 42.  43.  Wimmer, R., Andersen, K. K., Vad, B., Davidsen, M., Molgaard, S., Nesgaard, L. W., Kristensen, H. H. and Otzen, D. E. Biochemistry 2006, 45, 481-97. Matsuzaki, K., Mitani, Y., Akada, K. Y., Murase, O., Yoneyama, S., Zasloff, M. and Miyajima, K. Biochemistry 1998, 37, 15144-53. Lohner, K., Staudegger, E., Prenner, E. J., Lewis, R. N., Kriechbaum, M., Degovics, G. and McElhaney, R. N. Biochemistry 1999, 38, 16514-28. Ramamoorthy, A., Marassi, F. M., Zasloff, M. and Opella, S. J. J Biomol MR 1995, 6, 329-34. Porcelli, F., Buck-Koehntop, B. A., Thennarasu, S., Ramamoorthy, A. and Veglia, G. Biochemistry 2006, 45, 5793-9. Chekmenev, E. Y., Vollmar, B. S., Forseth, K. T., Manion, M. N., Jones, S. M., Wagner, T. J., Endicott, R. M., Kyriss, B. P., Homem, L. M., Pate, M., He, J., Raines, J., Gor'kov, P. L., Brey, W. W., Mitchell, D. J., Auman, A. J., Ellard-Ivey, M. J., Blazyk, J. and Cotten, M. Biochim Biophys Acta 2006, 1758, 1359-72. Tang, M., Waring, A. J. and Hong, M. J Am Chem Soc 2007, 129, 11438-46. Ishitsuka, Y., Pham, D. S., Waring, A. J., Lehrer, R. I. and Lee, K. Y. Biochim Biophys Acta 2006, 1758, 1450-60. Hallock, K. J., Lee, D. K., Omnaas, J., Mosberg, H. I. and Ramamoorthy, A. Biophys J 2002, 83, 1004-13. Arbeit, R. D., Maki, D., Tally, F. P., Campanaro, E. and Eisenstein, B. I. Clin Infect Dis 2004, 38, 1673-81. Fowler, V. G., Jr., Boucher, H. W., Corey, G. R., Abrutyn, E., Karchmer, A. W., Rupp, M. E., Levine, D. P., Chambers, H. F., Tally, F. P., Vigliani, G. A., Cabell, C. H., Link, A. S., DeMeyer, I., Filler, S. G., Zervos, M., Cook, P., Parsonnet, J., Bernstein, J. M., Price, C. S., Forrest, G. N., Fatkenheuer, G., Gareca, M., Rehm, S. J., Brodt, H. R., Tice, A. and Cosgrove, S. E. Engl J Med 2006, 355, 653-65. Straus, S. K. and Hancock, R. E. W. Biochim Biophys Acta 2006, 1758, 1215-1223. Park, S. H. and Opella, S. J. J Mol Biol 2005, 350, 310-8. Gullion, T. and Schaefer, J. Journal of Magnetic Resonance (1969) 1989, 81, 196-200. Maisetta, G., Mangoni, M. L., Esin, S., Pichierri, G., Capria, A. L., Brancatisano, F. L., Di Luca, M., Barnini, S., Barra, D., Campa, M. and Batoni, G. Peptides 2009, 30, 16221626. Liu, H., Li, S., Wang, Y., Yao, H. and Zhang, Y. International Journal of Pharmaceutics 2006, 311, 182-186.  224  APPENDICES  APPE DICES Appendix A: Experimental protocols Appendix A.1: Peptide workup The following protocol is for the peptide workup procedure after the completion of automatic peptide synthesis using CS Bio Co. peptide synthesizer (model # CS136XT), located in Straus Lab in the Department of Chemistry at the University of British Columbia. 1. Cleavage from resin and deprotection of side chains: o  Transfer resin into a reaction funnel.  o  Rinse the reaction vessel few times with DCM or MeOH to wash remaining resin into the reaction funnel.  o  Wash the resin (3 × 5 ml of DCM, 3 × 5 ml of MeOH, 3 × 5 ml of Et2O).  2. Put in vacuum dessicator overnight: o  Prepare the trap by filling up the dewar with dry ice and acetone to the top.  o  Cover the dewar with Styrofoam or paper towels to keep the temperature low.  o  This will prevent Et2O from entering and damaging the vacuum pump.  3. Cleave the peptide from the resin: o  Add the cleavage mixture to the reaction vessel.  o  Cleavage mixture for cleaving peptides with amidated C-terminus: 23.75 ml TFA, 0.625 ml ddH2O, 0.625 ml TES.  o  Cleavage mixture for cleaving peptides with carboxyl C-terminus: 23.60 ml TFA, 1.25 ml ddH2O, 0.125 ml TES, 1.25 ml EDT.  o  Stir for 5 hours.  4. Precipitate the peptide: 225  APPENDICES  o  Filter off the cleaved peptide mixture into a 100 ml round bottom flask.  o  Wash the reaction funnel with 3 × 5 ml of TFA. Collect filtrates as well.  o  Reduce volume to oily droplets by rotary evaporation.  o  Place the round bottom flask in ice bath.  o  Add cold diethyl ether drop-wise for first 10 ml ~ 20 ml, and then pipet-wise to 100 ml in total.  o  Filter the precipitate (peptide) using vacuum suction filter.  o  Decant solvent.  o  Wash again with diethyl ether, filter, and decant solvent.  o  Dissolve the precipitate (peptide) in ddH2O.  o  Store in freezer or freeze in N2 (l) and lyophilize.  5. Characterization by MALDI-TOF MS.  6. Purification by RP HPLC using C4 column.  Appendix A.2: HPLC purification The following protocol is for RP HPLC purification of crude peptide product using Waters 600 system HPLC (model 600 controller and model 2996 photodiode array detector), located in Sherman Lab in the Department of Chemistry at the University of British Columbia. 1. Prepare Buffer A (90% ddH2O, 10% can, 0.1% TFA) and Buffer B (10% ddH2O, 90% can, 0.1% TFA). 2. Filter Buffer A and Buffer B by vacuum suction filtration. 3. Dissolve the peptide crude product in minimum amount of Buffer A or ddH2O. Sonicate to dissolve if necessary. 4. Each HPLC purification allows 4 ml of crude product to be purified. 5. Place the C4 column on the rack (uncap both ends of the column, screw the metal tubing to the inlet and outlet of the column).  226  APPENDICES  6. Position Buffer A (Solvent C tubing) and Buffer B (Solvent D tubing) and waste bottle. Switch on He  (g).  Switch on power (two black switches, one for the HPLC machine and  the other for the lamp). Wait for the status light to stop flashing. 7. On the screen: o  Setup  A: disable (0, enter)  B: disable (0, enter)  C: enable (1, enter)  D: enable (1, enter). o  Direct  sparge 100 ml  enter (Observe bubbles in Buffer A and Buffer B  bottles)  wait for 10 minutes  sparge 0 ml  enter.  8. On the computer: o  Empower System-System login: user = system, password = manager quickstart  o  Instrument aurein 2.3  Erica  HPLC_manual  open file control panel  John  OK  No (Save changes to untitled)  switch ml  enter  turn to “draw” turn to “run”  aurein 2.2 or  No (Save changes)  9. Flip the metal switch to right to allow solvent flow through tubing C: 100%, D: 0%  advance  flow rate, 5 ml  enter  draw solvent (5 ml)  check composition,  attach a syringe to the black  turn to “inject”  inject and leave 1  remove the syringe and discard the remaining 1 ml of solvent  repeat the same procedure one time with solvent C and one time with solvent D the metal switch to left to allow solvent flow back to through the column  flip  flow rate, 0  ml. 10. Front of the column rack: o  Turn inlet switch to semi-prep position (Arrow pointing to “semi-prep”  turn  the load/injection switch up to the “load” position. 11. On the computer: o  Equilibrate (bottom left)  John aurein 2.2 or aurein 2.3  (Want baseline ~ 0) for ~ 10 minutes  Equilibrate/Monitor  Observe the line to be flat  bottom one  abort. 12. Draw 5 ml of peptide crude product solution using the injection syringe (No bubbles) semi-prep  take off lid and insert the syringe all the way in  check load position.  13. On the computer:  227  APPENDICES  o  Press “make single injection” (first one bottom left) and volume)  press “inject”  14. Manually inject the peptide solution position  enter information (name  wait until ready screen pop up. turn the load/injection switch down to the “inject”  after the HPLC starts running, turn the load/injection switch back to the  “load” position  remove the syringe and place the lid back.  15. Collect every 0.5 minute (30 sec) for each tube. Observe intensity at least 0.1 for collection. Then wait until 80 min run is finished. 16. After the run, check to see the abort button (bottom left) is off  stop flow  repeat  from step 11 to begin next HPLC purification. 17. If not continuing HPLC, then switch off both the HPLC machine and the lamp, and then switch off He (g). 18. On the computer: o  Browse project peaks  view  update  select “filename”  integrate  select  right click on the data and select all columns.  o Right click on the HPLC spectrum to print the spectrum to PDF file  save in  both the computer and a floppy disk for printout.  Appendix A.3: Solution CD sample preparation The following protocol is for the preparation of solution CD samples of aurein peptides in model lipid membranes. 1. Add calculated amount of lipid/s (dissolved in CHCl3:MeOH) into 5 ml round bottom flask. 2. Dry using N2 (g). 3. Dry in vacuum dessicator overnight or ~ 8 hours (store at -20°C if not used right away). 4. Redissolve the lipid in 450 µl ddH2O. 5. Add 50 µl of 2 mM peptide solution (or 50 µl of ddH2O for control samples). 6. Sonicate the mixture for ~ 1.5 hours (Or until the lipid or peptide/lipid mixture dissolves completely in ddH2O).  228  APPENDICES  7. Transfer the lipid or peptide/lipid mixture into a 600 µl ependorf tube. Add ddH2O to 500 µl level to compensate for loss of ddH2O due to evaporation during sonication. 8. Finger-vortex to mix and spin down using the table centrifuge.  9. The sample is now ready for CD experiment.  Appendix A.4: Oriented CD sample preparation The following protocol is for the preparation of oriented CD samples of aurein peptides in model lipid membranes. 1. Add calculated amount of lipid/s (Dissolved in CHCl3:MeOH) into 5 ml round bottom flask. 2. Dry in vacuum dessicator overnight or ~ 8 hours (store at -20°C if not used right away). 3. Redissolve the lipid in 2000 µl ddH2O. 4. Add 250 µl of 2.0 mM peptide solution (or 250 µl of ddH2O for control samples). 5. Sonicate the mixture for ~ 2 hours (or until the lipid or peptide/lipid mixture dissolves completely in ddH2O). 6. Deposit 90 µl of lipid or peptide/lipid mixture on quartz slides placed in a Petri dish. Deposit on the centre toward the bottom of the quartz slides. Be careful not to deposit the sample over where the spacer is located. 7. Dry the samples using air or N2 (g). 8. Place the Petri dish containing the sample on the quartz slide in a rehydration chamber filled with K3PO4 (s) and ddH2O. Incubate at 37ºC for at least 72 hours (usually 4 days for DMPC, 5 days for 4:1 DMPC/DMPG (mol/mol), 6 days for 1:1 CL/POPG (mol/mol), 8 days for 4:1 POPC/POPG (mol/mol) and POPE/POPG (mol/mol), and 9 days for 1:1 POPE/POPG (mol/mol)).  9. Sample is then ready for the CD experiment.  Appendix A.5:  31  P MR sample preparation  The following protocol is for the preparation of  31  P NMR samples of aurein peptides in  model lipid membranes. 229  APPENDICES  1. Add calculated amount of lipid/s (dissolved in CHCl3:MeOH) into 5 ml round bottom flask. 2. Dry using N2 (g). 3. Dry in vacuum dessicator overnight or ~ 8 hours (store at -20°C if not used right away). 4. Redissolve the lipid/s in 500 µl ddH2O. 5. Add calculated amount of peptide solution (Or without peptide for control samples). 6. Sonicate the mixture for ~ 2 hours. 7. Deposit lipid or peptide/lipid mixture in 10 µl aliquots on 9 mylar plates placed in a Petri dish (mylar plate dimension depends on the coil size of the solid-state NMR probe). 8. Dry between each deposit. 9. Dry in vacuum dessicator overnight or a minimal of ~ 8 hours (if using CHCl3 or MeOH to dissolve and deposit the sample). 10. Place the Petri dish in a rehydration chamber filled with K3PO4  (s)  and ddH2O. Incubate  at 37ºC for at least 48 hours (usually 4 days for DMPC, 5 days for 4:1 DMPC/DMPG (mol/mol), 6 days for 1:1 CL/POPG (mol/mol), 8 days for 4:1 POPC/POPG (mol/mol) and POPE/POPG (mol/mol), and 9 days for 1:1 POPE/POPG (mol/mol)). 11. Stack and wrap the mylar plates in a piece of thin parafilm. Secure the sample by pulling and folding the two loose ends of the parafilm very gently. 12. Sample is now ready for NMR experiment.  230  APPENDICES  Appendix B: Peptide sequence and molecular weights of all aurein peptides and analogues Peptide name  Peptide sequence  Molecular weight (g/mol)  Aurein 2.2  GLFDIVKKVVGALGSL  1614.95  Aurein 2.3  GLFDIVKKVVGAIGSL  1614.95  Aurein 2.3-COOH  GLFDIVKKVVGAIGSL-COOH  1615.95  L13A  GLFDIVKKVVGAAGSL  1572.90  L13F  GLFDIVKKVVGAFGSL  1648.99  L13V  GLFDIVKKVVGAVGSL  1600.95  Aurein 2.2-∆3  GLFDIVKKVVGAL  1357.70  Aurein 2.2-∆6  GLFDIVKKVV  1116.40  Aurein 2.2-∆4N  IVKKVVGALGSL  1182.50  Aurein 2.2-∆4N∆3C  IVKKVVGAL  925.20  231  

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.24.1-0060257/manifest

Comment

Related Items