Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of classical pathway complement activation by liposomes and modulation by incorporated… Bradley, Amanda Joan 1998

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

Item Metadata

Download

Media
831-ubc_1998-271129.pdf [ 9.43MB ]
Metadata
JSON: 831-1.0088736.json
JSON-LD: 831-1.0088736-ld.json
RDF/XML (Pretty): 831-1.0088736-rdf.xml
RDF/JSON: 831-1.0088736-rdf.json
Turtle: 831-1.0088736-turtle.txt
N-Triples: 831-1.0088736-rdf-ntriples.txt
Original Record: 831-1.0088736-source.json
Full Text
831-1.0088736-fulltext.txt
Citation
831-1.0088736.ris

Full Text

C H A R A C T E R I Z A T I O N O F C L A S S I C A L P A T H W A Y C O M P L E M E N T A C T I V A T I O N B Y L I P O S O M E S A N D M O D U L A T I O N B Y I N C O R P O R A T E D P O L Y ( E T H Y L E N E G L Y C O L ) by A M A N D A J O A N B R A D L E Y B.Sc . ( H O N O U R S ) U N I V E R S I T Y O F B R I T I S H C O L U M B I A , 1991 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R T M E N T O F P A T H O L O G Y A N D L A B O R A T O R Y M E D I C I N E We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A February 1998 ® Amanda Joan Bradley, 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT Complement activation causes opsonization of foreign particles leading to particle elimination from the blood. Complement-mediated opsonization of charged and large liposomes presents a fundamental problem in their use to deliver therapeutic agents in vivo. To prolong the circulation half-lives of such liposomes, complement activation must be curtailed. The two overall aims of this study were to characterize complement activation through antibody-independent C l q binding to anionic liposomes and to assess the ability of poly(ethylene glycol)-lipids (PEG-lipids) to inhibit this complement activation. This study determined that both electrostatic and chemical forces contributed to C l q binding to anionic liposomes. Negative phospholipids in the liposome composition were required to detect C l q binding to liposomes. A t close to physiologic p H (7.2) and ionic strength (0.145 M ) and in the absence or presence of serum, anionic liposomes bound a measurable but small amount of C l q . However, as the concentration of negative phospholipid in the liposomes increased, C l q binding increased. C l q binding increased even more as the ionic strength or the p H decreased. C l q peptide studies also demonstrated the importance of electrostatics. Peptides composed of residues 14-26 of the C l q A chain ( C l q A 1 4 . 2 6 ) bearing a net charge of plus five inhibited C l q binding and complement activation by anionic liposomes. This inhibitory capacity was dependent on the peptide's five positive charges and was independent of conformation or amino acid-sequence. While electrostatics were important in determining C l q saturation binding, C l q binding affinity constants were independent of the electrostatic component suggesting that i i chemical forces were also involved. Clq-mediated complement activation was also affected by liposome size. For cardiolipin-containing liposomes, 240 nm vesicles bound more C l q and activated complement more readily than 100 nm vesicles. Multilamellar vesicles (—1-8 jum) bound 15 times more C l q than 240 nm liposomes. This study is the first to show the ability of PEG-l ipids to act as a barrier against complement activation by anionic liposomes. Incorporation of either cholesterol-PEG 6 0 0 ( C H - P E G 6 0 0 ) , cholesterol-PEG 1 0 0 0 ( C H - P E G 1 0 0 0 ) , or phosphatidylethanolamine-PEG 2 0 0 0 ( P E - P E G 2 0 0 0 ) caused P E G - l i p i d dose-dependent inhibition of C l q binding to and complement activation by large anionic liposomes. Complement activation was strongly inhibited when 15 mole % of C H - P E G 6 0 0 , 10 mole % C H - P E G 1 0 0 0 , or 5 mole % P E -PEG 2 0 oo w e r e incorporated into 100 nm anionic liposomes. i i i TABLE OF CONTENTS Abstract i i Table of Contents iv List of Tables x i i List of Figures xiv Acknowledgements xix Dedication xx Abbreviations xxi Chapter 1. Introduction 1 1.1 Complement Activation 2 1.1.1 C l q and Initiation of the Classical Pathway of Complement Wi th or Without Antibody 2 1.1.2 Classical Pathway of Complement 10 1.1.3 Alternative Pathway of Complement 11 1.1.4 Activation of Complement by Pentraxins and Lectins 12 1.2 Consequences of Complement Activation 14 1.2.1 Opsonization 14 1.2.2 Generation of Biological Response Modifiers 15 1.2.3 Membrane Attack Complex ( M A C ) Formation, iv Membrane Damage, and Ce l l Lysis 15 1.3 Endogenous Regulation of Complement 16 1.4 Liposomes 19 1.4.1 Multilamellar and Unilamellar Vesicles 19 1.4.2 Liposome Applications 20 1.4.3 Limit ing Factors for Liposome Use 21 1.4.4 A Method for Monitoring Ligand-Liposome Binding: Particle Electrophoresis 23 1.5 Interactions Between Liposomes and Complement 24 1.5.1 Physical Aspects of Liposomes that Contribute to Complement Activation 25 1.5.2 Relationship Between Complement Activation and Fate of Liposomes 30 1.6 Incorporation of Poly (ethylene glycol)-lipids into Liposomes 31 1.6.1 Characteristics of Poly (ethylene glycol) 32 1.6.2 Modification of Liposome Circulation Time by Poly(ethylene glycol) Incorporation 34 1.7 Overall Objectives 35 Chapter 2. Materials and Methods 37 2.1 Reagents 37 2.1.1 Antisera and Antibodies 37 v 2.1.2 Lipids 37 2.1.3 Peptides 37 2.1.4 Other Reagents 38 2.2 Preparation of Reagents for Experiments 39 2.2.1 Preparation of Normal Human Serum 39 2.2.2 Purification of C l q from Human Plasma 39 2.2.3 Preparation of C l q Depleted Serum 40 2.2.4 Radiolabelling C l q 41 2.2.5 Synthesis and Analysis of C H - P E G Derivatives 42 2.2.6 Synthesis of C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 and C H - P E G 1 4 0 0 - N H - C O C H 3 45 2.2.7 Preparation of Liposomes 46 2.3 Analytical Techniques 47 2.3.1 Autoradiography of 1 2 5 I -C1 q 47 2.3.2 Functional Complement Assay 48 2.3.3 C l q Hemolytic Assay 49 2.3.4 C l q E L I S A 50 2.3.5 Equil ibrium C l q Binding Assay 51 2.3.6 Clq/ l iposome Binding Inhibition Assay 53 2.3.7 Particle Electrophoresis 54 2.3.8 Analysis of C H - P E G 1 4 0 0 Incorporation into Liposomes 55 v i 2.3.9 Assessment of C H - P E G Exchange with Normal Human Serum 56 Chapter 3 . Complement Activation by Liposomes via C l q Binding and Effect of Liposome Size 57 3.1 Complement Activation by Anionic Liposomes and Not Neutral Liposomes 57 3.2 Evaluation of Purity and Function of Isolated and Radiolabelled C l q 59 3.3 C l q Binding to Anionic Liposomes: Observed and Measured 68 3.4 Effect of Liposome Size on C l q Mediated Complement Activation 74 Chapter 4. Characterization of C l q Binding to Anionic Liposomes 79 4.1 p H Dependence of C l q Binding to Anionic Liposomes 79 4.2 Binding of Purified C l q to Anionic Liposomes at L o w p H 81 4.3 Effect of L i p i d Concentration on C l q Binding to C L 2 0 Liposomes at p H 7.2 86 4.4 Effect of Ionic Strength on Clq/ l iposome Binding 89 4.5 Measurement of Clq/ l iposome Binding in Human Serum 95 v i i Chapter 5: Evaluation of the Role of a Cationic Region of Clq in Liposome Interactions with Complement 100 5.1 Inhibition of C l q Binding to C L 2 0 Liposomes by C l q A Peptides 102 5.2 Inhibition of Liposome Complement Activation by C l q A Peptides 104 5.3 Direct Interaction of Cationic Residues with Anionic Liposomes 107 Chapter 6: Modulation of Clq Mediated Complement Interactions with Anionic Liposomes by Incorporation of PEG-lipids 110 6.1 Evaluation of C H - P E G - N H - 1 4 C O C H 3 Purity 110 6.2 Evaluation of C H - P E G Incorporation into Liposomes 111 6.3 Assessment of Loss of C H - P E G from Liposomes in Human Serum 117 6.4 Effect of PEG- l ip id Incorporation on C l q Binding to 100 nm Liposomes and Complement Activation 119 6.5 Effect of P E G - l i p i d Incorporation on C l q Binding to 240 nm Liposomes and Complement Activation 125 Chapter 7. Discussion 134 7.1 Characterization of C l q Binding and Complement Activation by Anionic Liposomes 134 v i i i 7.1.1 Limitations, Assumptions, and Justification 134 7.1.2 The Requirement for Negative Charge 137 7.1.3 The Effect of Surface Charge Density on C l q Binding to Anionic Liposomes 141 7.1.4 The Effect of p H on C l q Binding to Anionic Liposomes 142 7.1.5 The Ionic Strength Dependence of C l q Binding to Anionic Liposomes 144 7.1.6 A Model for CIq/Liposome Interactions 146 7.1.7 Possible Biological Reasons for L o w C l q Binding Levels under Physiologic Conditions 153 7.1.8 C l q Aggregation: A Potential Role in Determining C l q Binding Levels 154 7.1.9 C l q Binding to Anionic Liposomes in Human Serum 157 7.2 The Effect of Liposome Size on C l q Binding and Complement Activation 158 7.3 Examination of a Cationic Region of C l q for Involvement in Complement/Liposome Interactions 161 7.3.1 C l q A Peptides Inhibit C l q Binding and Complement Activation by Anionic Liposomes 161 7.3.2 C l q A Peptide Inhibitory Capacity Lacks Conformation or Sequence Specificity 162 7.3.3 The Requirement for Cationic Residues: Charge Control ix Peptides and an Unrelated Control Peptide 163 7.4 The Effect of P E G - l i p i d Incorporation on C l q Binding and Complement Activation by Anionic Liposomes 166 7.4.1 Inhibition of C l q Binding and Complement Activation by 100 nm Anionic Liposomes As a Result of P E G - l i p i d Incorporation 166 7.4.2 The Effect of P E G - l i p i d Anchor 171 7.4.3 Inhibition of C l q Binding and Complement Activation by 240 nm Anionic Liposomes as Result of P E G - l i p i d Incorporation 174 7.5 Future Directions: Testing the Model 175 Chapter 8. Summary 177 Bibliography 182 Appendix 1 Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes at p H 7.2 in a Purified Protein System: Effect of L i p i d Concentration 204 Appendix 2 Scatchard Plots for C l q Binding to 30 and 40 mole % Cardiolipin-containing Liposomes at p H 7.2 in a Purified Protein System 206 Appendix 3 Scatchard Plots for C l q Binding to Cardiolipin-containing Liposomes at p H 4 in a Purified Protein System 207 Appendix 4 Scatchard Plots for C l q Binding to Phosphatidylglycerol-containing Liposomes at p H 4 in a Purified Protein System 208 Appendix 5 Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes at p H 7.2 in a Purified Protein System at 100 m M N a C l 209 Appendix 6 Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes in Clq-Depleted serum (1/60): Effect of L i p i d Concentration 210 x i LIST OF TABLES T A B L E 1 T A B L E 2 T A B L E 3 T A B L E 4 Substances That Bind C l q in the Absence of Antibody Regulators of Complement Interaction of Liposomes and Complement 8 18 26 Apparent Association Constants and Saturation Binding for C l q Binding to Cardiolipin-containing Liposomes at p H 7.2 in a Purified Protein System: Effect of Liposome Surface Charge Density 71 T A B L E 5 Apparent Association Constants and Saturation Binding for C l q Binding to 240 nm Anionic Liposomes at p H 4 in a Purified Protein System 85 T A B L E 6 Effect of Ionic Strength and p H on Liposome Aggregation in the Presence of C l q 91 T A B L E 7 Apparent Association Constants and Saturation Binding for C l q Binding to C L 2 0 Liposomes at p H 7.2 and 100 m M N a C l in a Purified Protein System 95 T A B L E 8A Effect of L i p i d Concentration on Apparent Association Constants and Saturation Binding for C l q Binding to C L 2 0 Liposomes in a Purified Protein System at p H 7.2 98 T A B L E 8B Effect of L i p i d Concentration on Apparent Association Constants and Saturation Binding for C l q Binding to C L 2 0 Liposomes in Human Serum Diluted 1/60 98 xn T A B L E 9 C l q A Test Peptides 102 T A B L E 10 The Amount of C l q A Peptides Required to Inhibit C l q Binding and Complement Activation by Liposomes 106 T A B L E 11 Assessment of C H - P E G _ 1 4 0 0 Incorporation into Liposomes 114 T A B L E 12 Analysis of C H - P E G _ 1 4 0 0 Exchange in Human Serum: Effect of the Initial Amount of C H - P E G _ 1 4 0 0 in the Liposomes 119 x i i i L I S T O F F I G U R E S F I G U R E 1 Activation Pathways of Complement F I G U R E 2 The C I complex F I G U R E 3 Preparation of C H - P E G - N H 2 F I G U R E 4 Complement Consumption in Human Serum by Liposomes Composed of Different P C . Requirement for Negative Surface Charge F I G U R E 5 Assessment of C l q Purity 5 A Silver Stained S D S - P A G E Analysis of C l q Purification 5B Isoelectric Focussing of C l q 5C Autoradiograph of 1 2 5 I - C l q F I G U R E 6 Hemolytic Activi ty of C l q and 1 2 5 I - C l q as a Function of Storage Time F I G U R E 7 Functional Activity of Purified C l q and 1 2 T - C l q in Relation to Storage Time 7 A C l q Binding to Human Aggregated IgG Measured by E L I S A 7B 1 2 5 I - C l q Counts Associated with Human Aggregated IgG F I G U R E 8 C l q Binding to Cardiolipin-containing Liposomes as a Function of C l q Storage Time F I G U R E 9 Autoradiograph of C l q Bound to Cardiolipin-containing Liposomes F I G U R E 10 C l q Binding to Cardiolipin-containing Liposomes at p H 7.2: 4 5 44 58 61 63 65 67 69 xiv Effect of Surface Charge Density 70 10A C l q Binding to Liposomes with 0, 20, and 30 mole % C L 10B C l q Binding to Liposomes with 30 and 40 mole % C L F I G U R E 11 Reversibility of C l q Binding to Anionic Liposomes 73 F I G U R E 12 Effect of Liposome Size on C l q Binding to Anionic Liposomes in Human Serum 75 F I G U R E 13 Effect of Liposome Size on C l q Binding Isotherms 76 F I G U R E 14 Effect of Liposome Size on Complement Activation 78 F I G U R E 15 pH-dependence of C l q Binding to Anionic Liposomes 80 F I G U R E 16 Assessment of C l q Integrity After L o w p H Treatment 82 F I G U R E 17 C l q Binding to Cardiolipin-containing Liposomes at L o w p H 83 F I G U R E 18 C l q Binding to Phosphatidylglycerol-containing Liposomes at L o w p H 84 F I G U R E 19 Effect of L i p i d Concentration on C l q / C L 2 0 Binding Isotherms at p H 7.2 87 19A Binding Isotherms with L i p i d Concentrations of 0.5 and 2 m M 19B Surface Area Dependence Transformation F I G U R E 20 Ionic Strength Dependence of C l q Binding to C L 2 0 M L V s 90 F I G U R E 21 C l q / C L 2 0 Binding Isotherms: Effect of Ionic Strength 93 21A C l q Binding to Extruded C L 2 0 Liposomes at 20, 100, and 145 m M N a C l 21B C l q Binding to C L 2 0 M L V s at 20, 100, and 145 m M N a C l xv F I G U R E 22 Effect of L i p i d Concentration on C l q Binding to C L 2 0 Liposomes in Serum (diluted 1/60) 97 22A C l q Binding Isotherms for C L 2 0 Liposomes in Serum (diluted 1/60) at L i p i d Concentrations of 0.6, 2, and 6 m M 22B Surface Area Dependence Transformation F I G U R E 23 Effect of Serum Dilution on C l q Binding to C L 2 0 Liposomes 99 F I G U R E 24 Model of C l q Showing Location of Peptide Sequences 101 F I G U R E 25 Inhibition of C l q Binding to C L 2 0 Liposomes by C l q A Peptides 103 F I G U R E 26 Inhibition of Liposome-Induced Complement Consumption by C l q A Peptides 105 F I G U R E 27 Interaction of C l q A Peptides with C L 2 0 Liposomes 108 F I G U R E 28 Ionic Strength Dependence of Peptide/CL20 Binding at p H 7.0 109 F I G U R E 29 Assessment of P E G - l i p i d Incorporation into Liposomes by Particle Electrophoresis 112 F I G U R E 30 Chromatographic Separation of Free C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 from Liposome-associated C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 115 30A Isolation of Liposome-bound C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 from Buffer 30B Isolation of Liposome-bound C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 from Serum xvi F I G U R E 31 Analysis of C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 Exchange in Human Serum 118 F I G U R E 32 Inhibition of C l q Binding to 100 nm C L 2 0 Liposomes by Incorporation of PEG-l ipids 121 32A Effect of Increasing Amounts of C H - P E G 6 0 0 on Clq/ l iposome Binding 32B Effect of Increasing Amounts of C H - P E G i 0 0 0 on Clq/ l iposome Binding 32C Effect of Increasing Amounts of P E - P E G 2 0 0 0 on Clq/ l iposome Binding F I G U R E 33 Complement Activation by 100 nm C L 2 0 Liposomes: 123 Inhibition by P E G - l i p i d Incorporation 33A Effect of Increasing Amounts of C H - P E G 6 0 0 on Complement Activation 33B Effect of Increasing Amounts of C H - P E G 1 0 0 0 on Complement Activation 33C Effect of Increasing Amounts of P E - P E G 2 0 0 0 on Complement Activation F I G U R E 34 Inhibition of C l q Binding to 240 nm C L 2 0 Liposomes by Incorporation of PEG-l ipids 127 34A Effect of Increasing Amounts of C H - P E G 6 0 0 on Clq/ l iposome Binding xvn 34B Effect of Increasing Amounts of C H - P E G 1 0 0 0 on Clq/ l iposome Binding 34C Effect of Increasing Amounts of P E - P E G 2 0 0 o on Clq/ l iposome Binding F I G U R E 35 C l q Binding to 240 nm C L 2 0 Liposomes with Incorporated 35A Effect of C H - P E G 1 0 0 0 on Clq/ l iposome Binding Isotherm 35B Effect of P E - P E G 2 0 0 0 on Clq/ l iposome Binding Isotherm F I G U R E 36 Complement Activation by 240 nm C L 2 0 Liposomes: PEG-l ip ids 129 Inhibition by P E G - l i p i d Incorporation 132 36A Effect of C H - P E G 6 0 0 on Complement Activation 36B Effect of C H - P E G 1 0 0 0 on Complement Activation 36C Effect of P E - P E G r 2 0 0 0 on Complement Activation F I G U R E 37 Clq /Liposome Interactions: A Model 151 xv i i i A C K N O W L E D G E M E N T S I am grateful for the friendship, advice, and support I have had from an amazing group of people in the Devine and Brooks labs: Katherine Serrano, Derek Sim, Elena Levin , Mar ia Issa, N i k k i Chahal, V i c k y Monsalve, Nadine Brockman, Sarah Kerrigan, Johan Jansen, Aniko Takacs-Cox, and Raymond Norris-Jones. Thank you also to Jihan Marjan who has moved on but has left a distinctive mark on the lab and w i l l always remain a good friend. Thanks to Doina Hritcu for giving me a hands-on course on polymer chemistry. Cheers to Sarah and Raymond for bringing joy into daily lab life in the Brooks lab and again to Raymond for his particle electrophoresis expertise. Immeasurable thanks go to Derek and Katherine who kept me going with their encouragement and humour. I am also grateful to my non-lab friends, particularly to K i m Durlacher, L isa Dickson, and Rob Dunlop who have understood my periodic drop-out-of-sight phases and who have on occasion been more excited about my work than I was. A big thank you to J im Sibley for his artistry, computer wizardry (and computer !), and for sharing his unbounded zest for life. Finally, thanks to my supervisors, Dana Devine and Don Brooks who have been a vast source of information, direction, and motivation. xix DEDICATION To my mother and father and to Duffy who I miss. xx A B B R E V I A T I O N S B C A : bicinchoninic acid B S A : bovine serum albumin C l q A : the A polypeptide chain of C l q C l q A (i4-26): synthetic peptide consisting of residues 14-26 of the A chain of C l q C l q A ( s c r a m b l e d ) : peptide consisting of residues 14-26 of the C l q A chain but in scrambled order C l q A ( P . A ) : peptide consisting of residues 14-26 of the C l q A chain where proline (P) residues are replaced with alanine (A) residues C l q A ( 2 + ) : peptide consisting of residues 14-26 of the C l q A chain where three of the cationic residues were replaced with glycine (G) residues C l q A ( 0 + ) : peptide consisting of residues 14-26 of the C l q A chain where all cationic residues were replaced with glycine (G) residues C l q - D S : Clq-depleted serum C I inh: C I inhibitor C4bp: C4 binding protein C H : cholesterol 3 H - C H E : tritiated cholesteryl hexadecyl ether C L : cardiolipin derived from bovine heart C L 2 0 : liposomes composed of E P C : C H : C L at 35:45:20 mole % C L 4 0 : liposomes composed of E P C : C H : C L at 15:45:40 mole % cpm: counts per minute xxi C R P : C-reactive protein D A F : decay accelerating factor; CD55 D C P : dicetylphosphate D M P C : dimyristoylphosphatidylcholine, a 14:0/14:0 P C D O P C : dioleoylphosphatidylcholine, an 18:1/18:1 P C D O T A P : 1,2-bis(oleoyloxy)-3-trimethyl-ammonio)propane D P P A : dipalmitoylphosphatidic acid D P P C : dipalmitoylphosphatidylcholine, a 16:0/16:0 P C D P P E : dipalmitoylphosphatidylethanolamine D P P G : dipalmitoylphosphatidylglycerol D S P C : distearoylphosphatidylcholine, a 18:0/18:0 P C D S P E : distearoylphosphatidylethanolamine E A : antibody sensitized sheep red blood cells E D C : 1 -ethyl-3-(3-dimethylaminopropyl)carbodiimide E L I S A : enzyme-linked immunosorbent assay E N D O : endothelial cell E P C : phosphatidylcholine (PC) derived from egg E P E : phosphatidylethanolamine (PE) derived from egg E P I : epithelial cell fB: factor B fD: factor D FIB: fibroblast xx i i Fgn: fibrinogen G C : galactosyl ceramide G M 1 : monosialoganglioside G V B 2 + . gelatin veronal-buffered saline with 0.15 m M C a 2 + and 1 m M M g 2 H D L : high density lipoprotein H E P C : hydrogenated phosphatidylcholine (PC) derived from egg H R F : homologous restriction factor H R P : horseradish peroxidase H S P C : hydrogenated soy P C H u : human I: ionic strength IEF: isoelectric focusing L P S : lipopolysaccharides L U V : large unilamellar vesicle M A C : membrane attack complex M B P : mannan-binding lectin; mannose binding protein M C P : membrane cofactor protein M I P : membrane inhibitor protein M I R L : membrane inhibitor of reactive lysis; C D 5 9 M L V : multilamellar vesicle M O P S : 3-(N-morpholino)propane-sulfonic acid N H S : normal human serum xx i i i P A : phosphatidic acid P B S : phosphate buffered saline P C : phosphatidylcholine P E : phosphatidylethanolamine P E G : poly (ethylene glycol) P G : phosphatidylglycerol derived from egg p i : isoelectric point PI: phosphatidylinositol P L T : platelet P O P E : palmitoyloleoylphosphatidylethanolamine P M N s : polymorphonuclear lymphocytes PS: phosphatidylserine R B C : erythrocytes R E S : reticuloendothelial system R T : room temperature S A : stearylamine S A P : serum amyloid P S D : standard deviation S E M : standard error of the mean S M : sphingomyelin S U V : small unilamellar vesicle T C A : trichloroacetic acid xxiv T L C : thin layer chromatography T M B : 3,3' ,5,5'-tetramethylbenzidine V B S : veronal-buffered saline W B C : leucocytes 1. I N T R O D U C T I O N The overall purpose of this project is to determine how complement is activated by anionic liposomes and how this complement activation can be blocked. The use of liposomes as carriers of therapeutic agents such as anti-cancer and anti-bacterial drugs requires that liposomes retain their stability and have a long half-life in the circulation. Complement proteins are amongst the gamut of immunoproteins responsible for opsonization and clearance of foreign particles from the blood. Because anionic liposomes are known to activate the classical pathway of complement (Chonn, A . et al. 1991), the initiation of the classical pathway, i.e. the binding of C l q to the surface of anionic liposomes, was studied in detail in the first part of this thesis. While models existed for the process of complement activation, the properties involved in direct antibody-independent C l q binding were not well understood. Despite the fact that the use of small, neutral liposomes composed of high melting point phospholipids has resulted in enhanced circulation lifetimes (Senior, J. and Gregoriadis, G . 1982), further improvements with regard to reducing liposome opsonization are required to obtain a more diverse set of liposome formulations for a wider spectrum of applications. The approach investigated in this study to reduce complement activation by liposomes is one which has been used to keep proteins off surfaces. The second part of this thesis focuses on evaluating the ability of surface-anchored poly (ethylene glycol) (PEG) to inhibit C l q binding and subsequent complement activation. 1 1.1 Complement Activation The complement system may be activated by several different initiation processes, including the classical pathway, the antibody-independent classical pathway, the alternative pathway, the pentraxin pathway, and the lectin pathway. Once complement is initiated, the cascade nature of the system becomes evident as the proteins in the pathway are sequentially activated. A n illustration of the pathways of complement is presented in Figure 1. 1.1.1 C l q and Initiation of the Classical Pathway of Complement Wi th or Without Antibody Classical pathway activation is triggered by the binding of C l q to a surface that may or may not be coated in antibody. C l q is an unusually shaped glycoprotein that consists of six globular headgroups connected by a triple helix collagen-like stalk (Reid, K . et al. 1976) (Figure 2). Human C l q is approximately 8.3 % carbohydrate by weight (Yonemasu, K . et al. 1971; Calcott, M . and Muller-Eberhard, H . 1972). C l q has a molecular mass of 465 kDa and is found in human serum at a concentration of 80-100 mg/L (McAleer , M . and S im, R . 1993). The three distinct polypeptide chains, A , B , and C which are approximately 220 amino acid residues in length each have a collagen-like region and a globular region (Porter, R. and Reid, K . 1978; Reid, K . et al. 1982). The triple helices of the collagen-like region begin close to the N-terminus and continue to about residue 89 while the remaining —131 residues of each chain fold to form the globular head domains (Reid, K . et al. 1982). Most C l q normally circulates as part of the C I 2 complex with two each of the C l r and C l s subunit proteins in a calcium-dependent association (Strang, C . et al. 1982). C l r and C l s bind C l q at the hinge region on the collagenous portion of C l q (see Figure 2). In the antibody-dependent classical pathway, C l q binds the Fc portion of IgG or I g M . For human IgG, the relative binding affinities of monomeric IgG subclasses to C l q are such that IgG3 > I g G l > IgG2 > IgG4 (Burton, D . 1993). The binding of C l q to immunoglobulins is mediated by ionic interactions (Poon, P. et al. 1985, Smith, K . et al. 1994) and the interaction is sensitive to ionic strength (Burton, D . et al. 1980). A l so , for IgG binding to C l q , the charged residues Glu-318, Lys-320 and Lys-322 in the C 7 2 domain are critical and highly conserved (Burton, D . et al. 1980 and Duncan, A . and Winter, G.1988). Immunoglobulins bind the globular headgroups of C l q and although the binding site in C l q has not been unequivocally established, positively charged residues in the headgroups are considered to be involved (Comis, A . and Easterbrook-Smith, S. 1985). 3 Figure 1: Activation Pathways of the Complement System ALTERNATIVE PATHWAY CLASSICAL PATHWAY PENTRAXIN PATHWAY LECTIN I PATHWAY I 4 Figure 2: Mode l of C I Under normal physiologic conditions, most C l q circulates in the blood as shown here, within the C I complex. C I is a calcium-dependent macromolecular complex consisting of one C l q protein (465 kDa) and two each of the C l r (85 kDa each) and C i s (85 kDa each) protein subunits. The C l r and C i s subunits interact to form a tetramer which makes contacts with the collagen-like stalk of C l q (Strang, C . et al. 1982). C l q recognition of and binding to a substance initiates the classical pathway of complement. Once C l q binding has initiated the activation of C I , the C l r and C i s zymogens are converted to active proteases. gobular domains collagenous domains C1q 5 It was initially thought that the binding of two or more of the C l q globular headgoups, requiring a dimer or larger oligomer of IgG, was absolutely necessary for activation of C I (Wright, J . et al. 1980; Cohen, S. 1968). However, C l q exhibits a measurable binding affinity for monomeric IgG. The binding of C l q to monomeric IgG results in a low level of C I activation (Tschopp, J. et al. 1980). The strength of C l q binding increases as the number of contacts between C l q and the activator increases. Consequently, as the aggregation state of IgG increases, so the binding constant of C l q increases (Burton, D . 1985). While the affinity of C l q binding to monomeric IgG is 2.5 x 10 4 M " 1 (Schumaker, V . et al. 1976), when IgG is heavily aggregated or is in immune complexes, the affinity of the interaction with C l q is 2 x 10 8 M " 1 (Hughes-Jones, N . and Gardner, B . 1978). In the absence of antibody, C l q binds directly to a variety of substances, including a number of different proteins, polyanions, cell structures, D N A , and many different cell types. A list of some non-immunoglobulin substances that bind C l q is provided in Table 1. Direct C l q binding to a surface or substance may or may not result in C I activation. While some non-immunoglobulin activators bind C l q more avidly than aggregated IgG, probably due to the multimeric structures of these activators which provide the potential of multiple binding sites, others bind weakly and fail to activate C I (Cooper, N . 1985). In the cases where substances bind C l q but fail to activate complement, C l q binding may be involved in a variety of processes such as the deposition of immune complexes along basement membranes (Bohnsack, J. et al. 1985), stimulation of oxidative metabolism in human P M N s (Tenner, A . and Cooper, N . 1982) and enhancement of 6 phagocytosis by human macrophages and monocytes (Bobak, D . et al. 1987; Ohkuro, M . et al. 1994). Involvement of C l q in these processes is made possible by the existence of C l q receptors on a wide variety of cell types including monocytes, macrophages, B cells, T cells, fibroblasts, endothelial cells, and platelets (reviewed by Schreiber, R . 1984). With the exception of the human fibroblast receptor which binds C l q at the globular domains (and activates complement) (Bordin, S. and Page, R . 1989), the rest of the C l q receptors on these cells bind the collagen like-portion of C l q . In order to do this, C l q must be free, rather than being within the C I complex. Although, under normal conditions C l q circulates in the C I complex, free C l q is generated at sites where C I activation has occurred, such as sites of tissue injury and inflammation. Upon C I activation, the C l r and C l s subunits are rapidly dissociated from C l q by C I inhibitor (Sim, R . et al. 1979). 7 Table 1: Substances that Bind C l q In the Absence of Antibody Substance K a Activates C ? Globular or Collagenous Region ? Reference Fibrinogen ** 1.2 x 10 8 M 1 8.3 x 10 6 M 1 no collagenous & globular Entwistle R. and Furcht, L.(1988) Fibr in ** 1.7 x 10 6 M 1 no n.d. Entwistle R. and Furcht, L.(1988) Fibronectin ** n.d. no collagenous Sorvil lo, J . et al. (1985) S A P n.d. yes collagenous Y i n g , S. et al. (1993) C R P ** n.d. yes collagenous Jiang, H . et al. (1992b) E. coli Strains n.d. yes n.d. Tenner, A . et al. (1984) L P S of E. coli ** 8 x 10 7 M 1 n.d. collagenous Zohair, A . et al. (1989) L i p i d A of Bacterial L P S n.d. yes n.d. Morr ison, D . and Kl ine , L . (1977); Cooper, N . and Morr ison, D . (1978) HIV-1 Retrovirus n.d. yes globular Thielens, N . et al. (1993) Some R N A Tumor Viruses n.d. yes n.d. Cooper, N . et al. (1976) Heparin n.d. yes n.d. Rent, R. et al. (1975) Decorin 1.3 x 10 8 M 1 no collagenous Krumdieck, R. et al. (1992) Monosodium Urate Crystals * n.d. yes n.d. Giclas, P . et al. (1979) H u T Cells ** n.d. no n.d. Chen, A . et al. (1994) 8 H u Diplo id Fibroblasts * 1.5 X 10 9 M " 1 yes globular Bordin, S. et al. (1989) Platelets ** 2.9 x 10 6 M 1 no collagenous Peerschke, E . et al. (1987) M y e l i n n.d. yes n.d. Vanguri , P. et al. (1982) P-Amyloid fibres n.d. yes collagenous Jiang, H . et al. (1994) Anionic Liposomes n.d. yes n.d. Marjan, J . et al. (1994) D N A A ^ _ . .i—7—n n.d. yes collagenous Jiang, H . et al. (1992a); Uwatoko, S. and Mannik, M . (1990) ** denotes that these binding measurements were made under very low ionic strength conditions (1= 0.015-0.09 M ) . * denotes that binding measurements were made at 1=0.1 M . nd= not determined. Not all cell types that bind C l q are listed. Amongst the substances which bind C l q directly and activate complement, some substances bind C l q via the globular headgroups and others bind via the collagenous region (see Table 1). The group of antibody-independent complement activators that bind the collagenous region of C l q includes serum amyloid P (SAP) (Ying, S . -C . et al. 1993), C-reactive protein (CRP) (Jiang, H . et al. 1992b), D N A (Jiang, H . et al. 1992a), and (3-amyloid fibres (Jiang, H et al. 1994). These complement activators share a common property: they all have repeating negative charges. The specific region of the collagenous stalk of C l q that is believed to mediate binding to this group of activators is contained within residues 14-26 of the C l q A polypeptide chain. These 13 amino acids include 5 cationic residues and 4 hydrophobic residues. Because anionic liposomes have repeating negative surface charges, the possibility that the C l q A 1 4 . 2 6 region of C l q may play a role in mediating Clq/ l iposome interactions was assessed as part of this dissertation. 1.1.2 Classical Pathway of Complement When C l q binding to a substance results in activation of C I , the sequence of events in the complement activation cascade is initiated. Following C l q binding, it has been hypothesised that a series of conformational changes takes place within the C I complex ( C l q C l r 2 C l s 2 ) . These movements result in the autoactivation of the two C l r subunits which are then capable of cleaving and activating the two C i s subunits (Schumaker, V . et al. 1986). The activated C i s proteases within the C l complex then cleave C4 . One of the cleavage products, C4b contains both a reactive thiolester bond which allows C4b to covalently bind exposed hydroxyl or amino groups and a binding site for C2 (Janatova, J . and Tack, B . 1981). When C2 binds C4b, C2 is cleaved by C i s . The C2a fragment which is also a serine protease forms a magnesium-dependent association with C4b in the C4b2a complex known as the classical pathway C 3 convertase. The ability of the C3 convertase, C4b2a, to cleave many molecules of C 3 serves as an amplification point for the opsonization of invading pathogens or foreign particles by complement. Cleavage of C 3 generates C3a and C3b. The C3b fragment, like C4b, contains a reactive thiolester bond which mediates the covalent attachment of C3b to exposed hydroxyl or amino groups which are found on a wide variety of substances such as proteins, carbohydrates, and lipids (Sim, R. et al. 1981). The presence of C3b marks the substance that it is bound to for removal by phagocytes which have C3b-receptors. 10 • I When C4b2a binds C3b, the resulting complex can cleave C5 to release C5a. C5b, the first component of the membrane attack complex, remains associated with C3b until C 6 and C7 bind. The formation of the membrane attack complex ( M A C ) is the same regardless of which pathway achieved activation (for classical, alternative, lectin or pentraxin pathways) and the components of the M A C assemble without further proteolysis. Upon the binding of C 7 to form the C5b-7 complex, a transition takes place resulting in enhanced hydrophobicity of the complex and improved association with the membrane. The binding of C8 then facilitates C9 binding. Up to 18 C 9 molecules can insert in the C5b-9 n complex forming a permeable channel through the membrane that leads to the lysis and death of a target cell . 1.1.3 Alternative Pathway of Complement The alternative pathway of complement is antibody-independent and plays an important role in host defense against pathogenic microorganisms (Schreiber, R. et al. 1979), virus-infected cells (Joseph, B . et al. 1975), and neoplastic cells (Budzko, D . et al. 1976). The alternative pathway of complement activation is initiated by the normal, constant low level hydrolysis of C3 generating C 3 ( H 2 0 ) . C 3 ( H 2 0 ) binds factor B which is then activated by proteolytic cleavage by factor D . This fluid phase convertase complex, C 3 ( H 2 0 ) B b , cleaves C3 into C3a and C3b. The C3b fragment possesses opsonic capabilities as it covalently binds surfaces or substances with amino or hydroxyl groups 11 via the reactive C3b thiolester bond. To trigger activation of this pathway, a proactive environment around the deposited C3b is required. There are chemical properties that determine whether a particle or surface wi l l activate the pathway, although these are not well defined. A n example of a pro-activating surface would be a microbe with particulate cell wall materials such as zymosan and endotoxin (Gewurz, H . et al. 1968). On the other hand, the presence of sialic acid on the surface of blood cells or bacteria prevents alternative pathway activation (Fearon, D . 1978). If the requirement for a proactive environment is met, factor B wi l l bind C3b in the presence of magnesium. Factor D then cleaves the C3b-bound factor B resulting in the generation of the alternative pathway C3 and C5 convertase, C3bBb. The C3bBb complex cleaves many C3 molecules providing an amplification point for the generation of C3b. The C3bBb complex also cleaves C5 to initiate the production of M A C , which assembles as described above. 1.1.4 Activation of Complement by Pentraxins and Lectins The pentraxin and lectin pathways of complement activation are two additional routes for the initiation of the classical pathway of complement in the absence of immunoglobulin. The two members of the pentraxin family found in humans, C R P and S A P , mediate activation of the classical pathway by binding C l q and activating C I (Jiang, H . et al. 1992b; Y i n g , S. et al. 1993). Alternately, mannan binding lectin, also known as mannose binding protein ( M B P ) , of the lectin pathway bypasses C l q and binds to C l r 2 C l s 2 itself promoting activation of C I (Lu, J . et al. 1990). M B P is also found in 12 serum with its own associated serum proteases, M A S P - 1 and M A S P - 2 . This M B P / M A S P complex can activate complement through C 2 and C4 in the absence of the complement proteases (Thiel, S. et al. 1997). As members of the pentraxin family, C R P and S A P have a cyclic pentameric structure composed of 5 identical non :covalently linked subunits. C R P is an acute phase reactant. The concentration of C R P may increase in response to inflammation, infection or acute tissue injury from less than 5 ptg/ml in normal serum to greater than 500 /xg/ml (Claus, D . et al. 1976). C R P binds to a variety of substances including certain phospholipids (Kaplan, M . and Volanakis, J . 1974; Volanakis, J . and Narkates, A . 1981), positively charged liposomes (Richards, R. et al. 1979), matrix proteins such a fibrinogen and laminin, nuclear D N A (Robey, F . et al. 1985) and polycations including protamine and poly-L-arginine (Claus, D . et al. 1977; DiCamel l i , R. et al. 1980). S A P is not an acute phase protein in humans. It is a glycoprotein with normal serum levels of about 40 ng/m\ (Pepys, M . et al. 1978). S A P interacts with molecules such as mannose-terminated glycoproteins (Kubak, B . et al. 1988), fibronectin (deBeer, F . et al. 1981), heparin and dermatan sulfate (Hamazaki, H . 1987), and histones (Hicks, P. et al. 1992). Both C R P and S A P bind to the collagenous region of C l q . It is believed that residues 14-26 on the A polypeptide chain of C l q comprise the binding site since C l q A 1 4 . 2 6 peptides inhibit the binding of C l q to C R P or S A P and inhibit complement activation (Jiang, H . et al. 1992b; Y i n g , S. et al. 1993). Activation of complement has thus far been described only for M B P in the lectin family, a group of carbohydrate-binding proteins (Lu , J . et al. 1990). M B P is an acute 13 phase reactant. M B P may have several configurations, the most common of which is the hexameric form which bears a striking resemblance to the structure of C l q with globular domains connected through associated collagenous portions of each subunit in a central stem-like structure. Sharing a common structure with C l q may be a factor in allowing M B P to mimic C l q in associating with C l r 2 C l s 2 and activating complement (Ohta, M . et al. 1990). Although MBP-init iated activation of complement generates C3b particles and thus provides an efficient means of opsonization, in vitro experiments have shown that little C5b-9 is generated (Super, M . et al. 1990). 1.2 Consequences of Complement Activation 1.2.1 Opsonization The deposition of C3 fragments, C3b and iC3b, generated as a result of complement activation onto a target surface marks that surface for removal by phagocytic cells bearing receptors (CR1 and CR3) which are specific for these molecules (Griffin, F . 1977). The fragment iC3b is formed when factor I (a regulatory protein) cleaves membrane bound-C3b releasing a small fragment and leaving iC3b associated with the membrane. Among the phagocytes that are equipped with C 3 fragment receptors are neutrophils, monocytes and macrophages which are in the circulation, liver Kupffer cells, and splenic macrophages. Opsonization of foreign organisms or particles is required for efficient phagocytosis. 14 1.2.2 Generation of Biological Response Modifiers In addition to fragments C3b and iC3b, other activation cleavage fragments and further breakdown products exhibit the ability to react with specific receptors. C3a, C4a, and C5a are anaphylactic peptides and as such have the ability to stimulate the release of histamine and other active products by binding to receptors on mast cells in the tissues leading to a tissue reaction and increased vascular permeability (Morgan, B . 1990). O f these, C5a is by far the most efficient anaphylactic agent. C3a and C5a also have chemotactic activity. Production of these molecules results in an influx of neutrophils into the area. The generation of these reactive fragments as a result of complement activation at a site of injury plays a major role in inflammation. 1.2.3 Membrane Attack Complex ( M A C ) Formation, Membrane Damage and Ce l l Lysis The formation of the M A C was described above for the classical pathway of complement. This lytic pathway is common for al l the pathways of complement. The M A C , C5b-9 n , forms functional pores in the target membrane allowing water, ions and small molecules to flow into the cell causing cell swelling and possible lysis. For metabolically inert target cells and for targets that do not possess any endogenous M A C inhibitory proteins, such as aged red cells or liposomes, a single M A C is sufficient to cause lysis. In nucleated, metabolically active cells, many thousands of M A C complexes may form on a cell without cell lysis occurring (Morgan, B . 1990). 15 1.3 Endogenous Regulation of Complement Complement activation must be controlled to avoid inappropriate destruction of homologous cells, inappropriate stimulation of an inflammatory response and the depletion of complement required to mediate an immune response to invading pathogens. The activation of complement is regulated by a number of proteins. Some of these inhibitors circulate in the blood (fluid phase inhibitors) and others are anchored to endogenous cell membranes. A summary of these regulatory proteins is presented in Table 2. Control of classical pathway activation at the level of C I is mediated by C I inhibitor ( C l - I N H ) and possibly by C l q inhibitor. The spontaneous activation of C I is inhibited through non-covalent interaction with C l - I N H , a serum glycoprotein (Ziccardi, R . 1982). C I inhibitor also controls activated C I by firmly binding to the catalytic sites on C l r and C i s , followed by dissociation of C l r and C i s from C l q (Ziccardi, R . et al. 1979). C l - I N H restricts the half-life of activated C I to approximately 20 seconds. The role of C l q inhibitor is less clear. C l q inhibitor is a chondroitin 4-sulfate proteoglycan which binds C l q at the collagen-like portion of the protein which also serves as the association site for C l r 2 C l s 2 (Ghebrehiwet, B . and Muller-Eberhard, H . 1978). Under physiologic conditions in vitro, C l q inhibitor has been demonstrated to bind C l q and inhibit C l q hemolytic activity in a dose-dependent manner (Conradie, J . et al. 1975 and Silvestri, L . et al. 1981). Other fluid phase regulators of complement such as factor H , C4-binding protein (C4bp), and S-protein (vitronectin) act as competitive inhibitors of complex partners. For 16 instance, C4bp can bind up to six molecules of C4b. C4bp binding to C4b blocks the C2 binding site preventing the formation of the C3 convertase, C4b2a, and accelerating the dissociation of previously formed C4b2a complexes. S-protein inhibits M A C formation by competing with membrane lipids for binding to the C5b-7 complex (Murphy, B . et al. 1988). Some plasma regulators exert their inhibitory capacity through enzymatic inactivation of complement proteins by either direct proteolytic activity or by acting as a protease cofactor. For example, factor H is a cofactor for factor I which cleaves C3b and C4b. Cell-bound inhibitory proteins target either C3 convertases or M A C on the cell surface. A n example of endogenous cell-bound protection against the C3 convertases of both the classical and alternative pathways is CD55 or decay accelerating factor ( D A F ) . CD55 prevents the association of C3b with factor B and of C4b with C 2 . D A F also acts to enhance dissociation of preformed complexes (Lublin, D . et al. 1989). M A C formation is inhibited by the action of the cell anchored protein called C D 5 9 or membrane inhibitor of reactive lysis ( M I R L ) . C D 5 9 is believed to bind C8 and C9 in the assembling M A C preventing the unfolding of C9 which is necessary for membrane insertion and preventing further binding of C9 or polymerization (Mer i , S. et al. 1990). In addition to the presence of plasma and cell-bound protein inhibitors, complement is also regulated by factors such as the geometric requirements for complement assembly and the requirement for divalent cations (Devine, D . 1991). Spatially, in order for C5 to be cleaved by the C4b2a convertase complex, C5 needs to be bound to C3b such that C3b-C5 is adjacent to the convertase. The requirement of C a 2 + 17 Table 2: Complement Regulatory Proteins P R O T E I N A C T I O N D I S T R I B U T I O N M W (kDa) M E M B R A N E - B O U N D D A F (CD55) -inhibits C3 activation -accelerates decay of C3 & C5 convertases R B C , W B C , P L T , EPI & E N D O 70 M I R L (CD59) -inhibits M A C formation R B C , W B C , P L T , EPI & E N D O 18 C R 1 -C3b, C4b, iC3b receptor -cofactor for f l -inhibits C3 convertases & alternative C5 convertase R B C & W B C 210-330 M C P -cofactor for f l W B C , EPI & E N D O , FIB & P L T 45-70 C8bp ( M I P . H R F ) -inhibits M A C formation R B C , P M N s & monocytes 65 Serum cone. (mg/L) C I inh -inhibits activated C I 150-300 110 F L U I D P H A S E C4bp -inhibits C3 activation -cofactor for f l 200-400 540 factor I - inactivates C3 & C5 convertases 30-40 88 factor H -cofactor for f l in alt.path, -accelerates decay of C3 & C5 convertases 200-700 155 Properdin -stabilizes alternative C3 & C5 convertases 20-30 220 S-protein -inhibits M A C formation 420-600 80 ww/ i n SP40,40 -inhibits M A C formation 35-105 80 * R B C , erythrocytes; W B C , leucocytes; P L T , platelets; E P I , epithelial cell ; E N D O , endothelial cell ; F I B , fibroblast; P M N s , polymorphonuclear lymphocytes 18 and M g 2 + for appropriate protein-protein complex formations indicates the natural tendency for these complexes to decay. Thus, the inherent nature of complement as well as endogenous regulatory proteins allows for strict control of complement and aims to avoid inappropriate activation or damage of "bystander" cells. 1.4 Liposomes Liposomes are vesicles composed of lipids which enclose an aqueous space. Lipids are amphiphilic; they have a hydrophilic headgroup which, for phospholipids, consists of a phosphate group, and a hydrophobic tail consisting of two long hydrocarbon chains. In aqueous solution, the l ip id headgroups orient themselves towards the aqueous environment and the acyl chains orient towards each other to exclude water resulting in the formation of a bilayer. Bangham and Home (1964) originally described liposomes, using electron microscopy to view the structures formed by a suspension of purified phospholipids derived from cells. 1.4.1 Multi lamellar and Unilamellar Vesicles Multilamellar vesicles ( M L V s ) composed of concentric bilayers are formed spontaneously when a dry fi lm of phospholipids and lipids are dispersed in excess water. These vesicles are very heterogeneous in size and lamellarity. M L V s may range in size from 0.2-10 /mi in diameter. The degree of lamellarity is such that only 5% or less of the total l ipid is in the outermost lamellae which is exposed to the environment (Cullis , P. et al. 1985). 19 Unilamellar vesicles are grouped by size such that 25-50 nm diameter vesicles are termed small unilamellar vesicles (SUVs) and 50-300 nm diameter liposomes are called large unilamellar vesicles ( L U V s ) . S U V s can be produced from M L V s by a number of different methods, including sonication (Huang, C . 1969) and use of a French press (Barenholz, Y . et al. 1979). Because S U V s are so small, they have a high bilayer curvature. The resulting packing constraints experienced by the lipids disturbs their physical properties. S U V s have a greater tendency to fuse (Wong, M . et al. 1982) and are more readily acted on by high density lipoproteins (HDLs) which remove phospholipid molecules from the bilayers of vesicles, leading to their disintegration (Scherphof, G . et al. 1984). In addition, S U V s have low trapping volumes (Huang, C . 1969). L U V s serve as more stable model membranes and have greater potential in drug delivery due to greater stability and larger trapped volumes. Procedures to produce L U V s include detergent dialysis (Mimms, L . et al. 1981), ethanol injection (Chen, C . and Schullery, S. 1979) reverse phase evaporation (Szoka, F . and Papahadjopoulos, 1978) and direct extrusion (Hope, M . et al. 1985). The extrusion technique, which is used in this thesis, results in production of largely unilamellar vesicles with sizes close to the filter pore size for liposomes extruded through filters with 200 nm or smaller pore sizes. 1.4.2 Liposome Applications Liposomes have been used for a variety of applications from serving as model membranes to the delivery of therapeutic agents. With respect to the use of liposomes as model membranes, liposomes have been used to study the roles of lipids in membranes 20 and to investigate processes such as membrane fusion (Bental, M . et al. 1987), protein-l ipid interactions (Chapman, D . 1982), membrane permeability (Deamer, D . and Nichols, J . 1983) and both the activation of complement (Kovacsovics, T . et al. 1985; Parce, J . and McConne l l , H . 1980; Parce, J. et al. 1983) and the formation of the M A C (Silversmith, R . and Nelsestuen, G . 1986; Shin, M . et al. 1977; Lachmann, P. et al. 1970). Liposomes that are designed to deliver therapeutic agents are at various stages of development, including up to advanced clinical trials and use in humans. Cl in ica l uses include vaccine transport (Antimisiaris, S. et al. 1993), diagnostic imaging (Tilcock, C . et al. 1993), drug transport (Gabizon, A . et al. 1994), hemoglobin transport as a blood surrogate (Kobayashi, K . et al. 1997), and gene and antisense transfer (L iu , Y . et al. 1995; Lappalainen, K . et al. 1997). Liposome-encapsulated drugs include drugs against neoplastic diseases or parasitic, bacterial, viral , and fungal infections (Chonn, A . and Cul l is , P . 1995; Wasan, K . etal. 1995). 1.4.3 Limit ing Factors for Liposome Use In general, injected liposomes are eliminated from the circulation by the reticuloendothelial system (RES): by hepatic Kupffer cells, splenic macrophages and by the lung, bone marrow and kidneys. The rate of elimination of liposomes from the plasma depends on the liposome composition. While over 90% of neutral liposomes end up in the liver within a few hours of injection (Woodle, M . and Lasic, D . 1992), certain negatively charged liposomes are cleared even more rapidly, with half-lives on the order of a few minutes (Chonn, A . et al. 1992). The rapid loss of liposomes from the circulation 21 presents a major limiting factor for the use of liposomes. The mechanism for the efficient uptake of injected liposomes is believed to involve association with and opsonization by plasma proteins (Moghimi, S. and Patel, H . 1992). Indeed, an inverse correlation between liposome circulation half times and the total protein bound to liposomes has been found (Chonn, A . et al. 1992). Plasma opsonins including IgG, the complement fragments, C3b and iC3b and fibronectin (Hsu, M . and Juliano, R . 1982) have been shown to coat liposomes and enhance the uptake of these liposomes by phagocytic cells which bear opsonin-specific receptors. In addition to rapid removal from the circulation, liposomes may also be destabilized by elements in the plasma. Bilayer permeability changes and release of entrapped solutes may be caused by plasma factors such as the M A C of complement, lipoproteins, apolipoproteins, and plasma phospholipid transfer proteins. For example, the interaction of high density lipoproteins ( H D L ) with liposomes results in dissolution of the liposome structure, transfer of liposomal phospholipids to H D L , and massive release of entrapped solute (Scherphof, G . et al. 1978; Ta l l , A . and Small , D . 1977). Isolated apolipoproteins such as apo-A-1 destabilize liposomes by transferring into the membranes and enhancing l ipid movement (Mendez, A . et al. 1988). The circulation lifetimes of liposomes reflect the combination of liposome destabilization and removal by circulating phagocytes as well as those in the liver and spleen. Circulation lifetimes have also been shown to depend upon factors such as the l ipid dose, liposome size, l ipid composition, and charge of the vesicles (Juliano, R . and Stamp, D . 1975). In addition, the encapsulation of drugs may affect liposome circulation 22 half-lives. For example, liposomes with entrapped doxorubicin had longer blood residency times than empty liposomes (Bally, M . et al. 1990). 1.4.4 A Method for Monitoring Ligand-Liposome Binding: Particle Electrophoresis Several different methods of measuring the binding of C l q to liposomes were used in this study. These methods include a competitive enzyme-linked immunosorbant assay ( E L I S A ) , a centrifugation based equilibrium binding assay using radiolabelled C l q , and particle electrophoresis. A s this last method is less widely used, a brief description of the capacity of particle electrophoresis is presented here. A detailed description of the technique is provided in Materials and Methods. Particle electrophoresis presents a way to measure the electrokinetic behaviour of particles such as cells, latex beads, and liposomes. It is a direct method of surface examination which is independent of the size and shape of the particle in the size range of interest here. Agents that alter the hydrodynamic properties of the surface region, such as neutral polymeric headgroups, also affect the electrophoretic mobility (Janzen, J . et al. 1996). This technique has been used to examine the modification of membrane functional groups and adsorption of ligands to membranes or surfaces since these modifications often affect the electrokinetic properties of a surface (Seaman, G . 1975). For example, particle electrophoresis has been used to study the effects of anticoagulants, storage of blood after collection, and different blood groups on the surface properties of red cells. Because the binding of proteins to liposomes potentially alters the liposome surface charge, monitoring 23 changes in liposome surface charge by particle electrophoresis can provide a measure of the amount of protein binding. The nature of the interaction can also be studied by monitoring the dependence of the electrophoretic mobility of the vesicles on factors such as p H and ionic strength. 1.5 Interaction Between Liposomes and Complement Many investigators have shown that liposomes can activate complement, depending on their composition and physical features. The ability of liposomes to activate complement has even been utilized to produce assays for the measurement of complement activity in serum (Bowden, D . et al. 1986 and Masaki , T . et al. 1989). In addition to antibody-mediated complement activation, liposomes can also initiate complement in the absence of immunoglobulins (Marjan, J . et al. 1994). The activation of complement by liposomes is an important issue as it can result in loss of liposome stability with subsequent leakage of entrapped substances (Funato, K . et al. 1992 and Hesketh, T . et al. 1971) or liposome opsonization by fragments of C3 leading to enhanced uptake by phagocytic cells (Matsuo, H . et al. 1994, Scieszka, J. et al. 1991, Harashima, H . et al. 1994, L i u , D . et al. 1995, Roerdink, F . et al. 1983, Wassef, N . and A l v i n g , C . 1987 & 1993). Complement activation also results in the release of fragments C3a and C5a which have powerful chemotactic and anaphylactic functions. The chemotactic behaviour of these fragments allows for signalling of phagocytic cells; Scieszka et al. (1991) demonstrated that the generation of C5a enhanced P M N phagocytosis of liposomes. In some species, such as swine, liposome 24 induced activation of complement and generation of C5a may result in anaphylactic shock and possibly death (Wassef, N . et al. 1989). Another example of the detrimental effects of complement activation by liposomes has been reported by Loughrey, H . et al. (1990). The transient thrombocytopenia in rats after intravenous administration of PG-containing liposomes was found to be due to the binding of C3b-opsonised liposomes to platelet C3 receptors. A s illustrated above, there are many ways that liposome efficacy can be lost due to liposomal activation of complement. 1.5.1 Physical Aspects of Liposomes that Contribute to Complement Activation A s mentioned above, whether liposomes activate complement and the extent to which they activate complement depends on physical aspects relating to the liposome surface. The activation of complement by different liposome compositions is summarized in Table 3. The capacity of liposomes to activate complement varies with l ipid saturation (Shin, M . et al. 1978; Devine, D . et al. 1994; Chonn, A . et al. 1991), the amount of cholesterol in the membrane (Alving, C . et al. 1977; Chonn, A . et al. 1991), the presence of charged phospholipids (Chonn, A . et al. 1991; Marjan, J . et al. 1994), and the size of the liposomes (Devine, D . et al. 1994; L i u , D . et al. 1995). Wi th respect to liposome composition, studies by Devine et al. (1994) and Chonn et al. (1991) demonstrated that in rat serum, complement activation was greater for anionic liposomes composed of saturated phospholipids whereas in human serum, unsaturated anionic liposomes were more potent activators. In rat, guinea pig and human serum, increasing the cholesterol content of anionic liposomes increased the level of 25 Table 3: Relationship Between Liposome Composition and Complement Activation Liposome Composition Species Activation Pathway Reference E P C : C H (55:45) E P C / D P P C / D O P C : C H (55:45) rat human guinea pig non-activating non-activating non-activating Devine, D . et al. (1994) Marjan, J . et al. (1994); Chonn, A . et al. (1991) Chonn, A . et al. (1991) E P C : C H : D P P E (35:45:20) rat human & guinea pig non-activating non-activating Devine, D . et al. (1994) Chonn, A . et al. (1991) D P P C : C H : D P P E (40:33:27) human alternative M o l d , C . (1989) E P C : C H : S A / D O T A P (35:45:20) rat human & guinea pig classical alternative Devine, D . et al. (1994) Chonn, A . et al. (1991) D M P C : C H : S A : G C (44:33:15:8) human alternative Cunningham, C . et al. (1979) E P C : C H : P G / P A / C L / P I / P S (35:45:20) rat human & guinea pig classical classical Devine, D . et al. (1994) Chonn, A . et al. (1991); Marjan, J. et al. (1994) D P P C : C H : D P P G (35:45:20) or (60:0:40) human classical Chonn, A . et al. (1991) D O P C : C H : D O P G (45:45:10) or (70:0:30) human classical Chonn, A . et al. (1991) P C : C H : P S (63:31:6) ( - 5 0 0 nm) rat alternative L i u , D . et al. (1995) D P P C : C H : D C P (26:71:3) human classical A l v i n g , C . et al. (1977) E P C : P S (90:10) human alternative Comis, A . et al. (1986) E P C : P S : E P E (80:10:10) human alternative Comis, A . et al. (1986) H E P C : C H : D C P (44.5:44.5:11) rat alternative Funato, K . et al. (1992) 26 complement activation in a dose-dependent way (Devine, D . et al. 1994; Humphries, G . and McConne l l , H . 1975; Chonn, A . et al. 1991). The mechanism by which cholesterol facilitates complement activation by these liposomes is unclear. Cholesterol is generally added to liposome compositions in order to increase the stability of liposomes in plasma or serum. Cholesterol increases the packing of phospholipid molecules, reducing bilayer permeability and preventing the loss of l ipid to high density lipoproteins (Kirby, C . et al. 1980a). The fact that cholesterol has both a stabilizing effect and the ability to enhance complement activation suggests that lipoproteins and complement interact with the liposome membrane in different ways. It should also be noted that some animal species, such as swine, have antibodies to cholesterol which contribute to complement activation and even to anaphylaxis upon injection of cholesterol-containing liposomes (Swartz, G . et al. 1988; Wassef, N . et al. 1989). Liposome charge plays a dominant role i n complement activation. Although there are some reports of complement activation by neutral liposomes (Mold , C . et al. 1989), others have not detected activation (Chonn, A . et al. 1991; Marjan, J . et al. 1994). Whether the liposome is negatively charged or positively charged appears to dictate which pathway of complement is activated. Chonn et al. (1991) showed that negatively charged liposomes activated the classical pathway of complement and positively charged liposomes activated the alternative pathway. However, there are discrepancies in the literature on this point, possibly owing to species differences, differences in systems employed and in the preparation of the plasma or serum used. The association of C R P with positively 27 charged liposomes leading to complement activation and complement-mediated liposome damage has also been reported (Richards, R. et al. 1979). For the initiation of the classical pathway of complement, the properties involved in the direct binding of C l q are not well understood. However, it has been proposed that negative surface charge and the presence of repeating binding sites are required for the binding of C l q to a particle surface. Activation of the classical pathway by negatively charged liposomes supports the electrostatic determinant in this hypothesis. Indeed, anionic liposome compositions which bind large amounts of plasma protein and have shortened circulation times activate the classical pathway of complement through the direct binding of the C l q subunit of C I (Chonn, A . et al. 1992; Marjan, J . et al. 1994). Because of complement's preference for charged particles, anionic and cationic liposomes that are used to deliver therapeutic agents are at the greatest risk of being opsonized or degraded by complement in vivo. Examples of such systems include the cardiolipin-based and the PG-based formulations of doxorubicin, PG-based amphotericin B preparations, cationic formulations for gene transfer and antisense transfer, and liposomes that are specifically targeted to R E S cells. Taking advantage of the high affinity binding of doxorubicin to cardiolipin, Rahman et al. (1985) prepared a liposomal doxorubicin formulation containing C L : P C : C H : S M (at 8:44:30:17 mole %) and bearing an overall positive charge. Doxorubicin entrapped in these liposomes retained its antitumour capacity in mice and rats and conferred reduced cardiotoxicity compared to free doxorubicin (Rahman, A . et al. 1985 and Rahman, A . et al. 1986b). However, much more doxorubicin was found in 28 the spleen and liver of mice and rats when liposomal drug was administered compared to free drug (Rahman, A . et al. 1986a; Rahman, A . et al. 1986b). This enhanced uptake by the spleen and liver suggests that liposome opsonization had occurred allowing for efficient uptake by splenic and hepatic macrophages. Similarly, the P G liposomal formulation of doxorubicin was rapidly removed from the circulation by the R E S (Gabizon, A . et al. 1994). Preparations of amphotericin B in P G or PS-containing liposomes developed by Madden et al. (1990) may also activate complement. The cationic liposomes that are used in anti-sense and anti-gene technology to facilitate the transport of oligonucleotides into cells are also potentially at risk for activating complement in vivo, probably through the alternative pathway. In cases where the aim is to have the liposome contents internalized by macrophages, complement-mediated opsonization may also play a role. Liposomes containing PS have been used to deliver a series of macrophage activating factors including a synthetic peptide analogue of gram negative bacterial cell walls, a chemotactic peptide, and interferon-gamma to phagocytic cells in order to promote the activation of monocyte and macrophage tumoricidal properties ( N i i , A . et al. 1991, Morikawa, K . et al. 1988; Fidler, I. et al. 1985). In addition, Verma et al. (1991) reported that phagocytosis of liposomes containing a recombinant malaria antigen by macrophages resulted in antigen expression at the macrophage surface. In addition to surface charge, liposome size also effects complement activation. In separate studies by Devine et al. (1994) and L i u et al. (1995), complement activation in rat serum was found to increase with increasing liposome size. Larger liposomes also 29 experienced greater complement-dependent degradation in rat serum (Harashima, H . 1994b; L i u , Z . et al. 1988). In these studies, the release of carboxyfluorescein from liposomes was attributed to the formation of membrane pores by C5b-9 complexes ( M A C s ) . 1.5.2 Relationship Between Complement Activation and Fate of Liposomes Many of the same factors which enhance complement activation by liposomes also increase the rate of liposome uptake and removal from the circulation. This correlation as well as direct evidence of complement involvement in liposome uptake suggests that complement plays a significant role in liposome removal by the R E S . Examples of factors that affect both complement activation and liposome elimination from the circulation are liposome size and surface charge. Larger liposomes are eliminated from the circulation by the R E S more rapidly than smaller liposomes (Juliano, R. and Stamp, D . 1975; Gabizon, A . and Papahadjopoulos, D . 1992; Senior, J . et al. 1985). The impact of surface charge on the fate of liposomes was first demonstrated by Juliano and Stamp (1975) who showed that negatively charged liposomes are cleared from the circulation of rats more rapidly than uncharged or positively charged liposomes. Some studies have shown enhancement of phagocytic uptake by inclusion of negative phospholipids (Fujiwara, M . et al. 1996; Lee, K . - D . et al. (1992 & 1993); Dijkstra, J . et al. 1985; Hsu, M . and Juliano, R. 1982). Lee et al. (1992 & 1993) found that the addition of phosphatidylserine (PS), phosphatidic acid (PA) , or phosphatidylglycerol (PG) resulted in a 10-20 fold increase in liposome uptake compared 30 with neutral liposomes depending on the cell line. However, other investigators have found a suppressive effect of certain anionic phospholipids. Roerdink et al. (1983) and Wassef et al. (1991) reported that addition of PS , phosphatidylinositol (PI), dicetylphosphate ( D C P ) , or dipalmitoylphosphatidic acid ( D P P A ) resulted in suppression of complement-opsonized liposome uptake. Wi th regard to liposome uptake by phagocytic cells, the effect of inclusion of negatively charged phospholipids appears to differ depending on the type of phospholipid and on the experimental system used. Direct evidence of complement involvement in liposome removal from the circulation is supplied by several investigators. While complement activation has been shown to destabilize liposomes causing leakage of entrapped solutes in vitro (Funato, K . et al. 1992; Hesketh, T. et al. 1971), the main effect of complement activation by liposomes in vivo is probably related to the opsonic activity of complement. Opsonization of liposomes by complement has been shown to greatly enhance uptake of liposomes by macrophages (Roerdink, F . et al. 1983; Wassef, N . and A l v i n g , C . 1987 & 1993). C3 plays a critical role in hepatic uptake of liposomes (Matsuo, H . et al. 1994; Harashima, H . et al. 1994a; L i u , D . et al. 1995) and iC3b was found to be responsible for opsonization and phagocytosis of liposomes containing 20 mole % P G by human neutrophils (Scieszka, J . et al. 1991). 1.6 Incorporation of Poly (ethylene glycol)-lipids into Liposomes Liposome opsonization and subsequent elimination from the circulation presents a significant complication for the potential use of charged or large liposomes in vivo. If 31 sites other than the R E S are to be targeted, the circulation lifetimes of liposomes must be increased. In addition to designing liposomes to be neutral and small (between 100 and 200 nm in diameter), another approach to increasing liposome circulation half-life has been to incorporate poly (ethylene glycol) (PEG) into the liposomal bilayer. This approach has met with considerable success, increasing circulation half-lives of both neutral and charged liposomes (Blume, G . and Cevc, G . 1990; Klibanov, A . et al. 1990; Woodle, M . et al. 1992; Senior, J . et al. 1991; A l l en , T. et al. 1991). In this study, we assess whether PEG-l ip ids are capable of specifically inhibiting complement activation by liposomes. 1.6.1 Characteristics of Poly(ethylene glycol) P E G is a non-toxic, neutral polyether which is soluble in water and in many organic solvents and has a large exclusion volume for most macromolecules (Harris, J . 1991). The exclusion volume is the volume from which the centre of mass of a second molecule is excluded. The chemical structure of P E G is HO-(CH2CH20)n-CH2CH2OH. The polymer chain is highly mobile in aqueous solution (Nagaoka, S. et al. 1985). A t high concentrations, P E G can be used to precipitate proteins and nucleic acids and to cause cell fusion (Atha, D . and Ingham, K . 1983; Ahkong, Q. et al. 1975). P E G excludes other polymers and in so doing may form aqueous two phase systems when paired with certain other polymers such as dextran above a set of critical concentrations. When attached to other molecules or surfaces, P E G provides a biocompatible, protective coating: proteins are rendered non-immunogenic and surfaces are rendered 32 protein-rejecting. Because of these abilities, P E G has been attached to many different types of molecules and surfaces including proteins, arterial replacements, blood contacting devices, and liposomes. Covalent attachment of P E G to proteins such as B S A , lactopherin, o^-macroglobulin, and recombinant proteins interleukin-2 and granulocyte colony-stimulating factor results in a decrease in immunogenicity and an increase in the protein circulation lifetime (Abuchowski, A . et al. 1977; Beauchamp, C . et al. 1983; Knauf, M . et al. 1988; Satake-Ishikawa, R . et al. 1992). The mechanisms behind the ability of PEG-coated surfaces or of P E G in solution to exhibit very low degrees of protein adsorption are not entirely understood. Most theories proposed have been related to the unique aqueous solution properties of P E G molecules. A s noted above, in aqueous solution, P E G is a heavily hydrated, highly mobile, neutral molecule with a large exclusion volume. Theories to explain the ability of P E G to keep proteins off surfaces include the rapid mobility of hydrated P E G chains giving an approaching protein little time in which to form an interaction (Nagaoka, S. et al. 1985), the large excluded volume of P E G (Hermans, J . 1982), a repulsive force resulting from a loss of configurational entropy when a protein approaches P E G resulting in reduced motion of the chain (Nagaoka, S. et al. 1985), and the lack of binding sites (ionic and hydrophobic) on P E G (Golander, C . et al. 1986). The attachment of P E G to a surface may also alter the electrical interactions associated with the surface since surface charges become buried beneath this viscous, hydrated, neutral layer (Harris, J . 1991). 33 1.6.2 Modification of Liposome Circulation Time by Poly(ethylene glycol) Incorporation Poly(ethylene glycol) is introduced into a liposome by including l ipid derivatives such as cholesterol-PEG ( C H - P E G ) and phosphatidylethanolamine-PEG (PE-PEG) into the l ipid mixture. Poly (ethylene glycol) can also be attached to the surface of pre-formed liposomes resulting in a polymeric coating only on the outside surface (Senior, J. et al. 1991). Many studies have shown increased circulation times for liposomes with P E G -lipids (Blume, G . and Cevc, G . 1990; Klibanov, A . et al. 1990; Woodle, M . et al. 1992; Senior, J . et al. 1991; A l l en , T. et al. 1991). Liposomes with incorporated PEG-l ipids may remain in the blood with half-lives as long as 48 hours (Zalipsky, S. 1995). The ability of P E G to protect liposomes from uptake by phagocytic cells in the circulation and in the liver and spleen is likely due to the ability of P E G to act as a surface barrier to plasma proteins (Allen, T . et al. 1991b). Senior et al. (1991) have shown that plasma components which adsorbed instantaneously ( t < l min) to PEG-free, control liposomes, adsorbed more slowly to liposomes with P E G ; no plasma induced alteration in partitioning was detected for up to 6 hours for liposomes with 20 mole % P E - P E G 5 0 0 0 (Senior, i. etal. 1991; Blume, G and Cevc, G . 1993). The capacity of PEG-l ipids to increase the circulation lifetimes of liposomes depends upon the P E G surface density and on the P E G chain length (Klibanov, A . et al. 1991; Maruyama, K . etal. 1991 and 1992; Woodle, M . etal. 1992; Torchil in, V . etal. 1992; M o r i , A . et al. 1991; Litzinger, D . et al. 1992). Between 5-10 mole % of P E -P E G 1 0 0 0 . 5 0 0 0 appears to give the best results (Woodle, M . and Lasic , D . 1992). Wi th 34 respect to l ip id anchor, D S P E has been cited as the most effective anchor for P E G in terms of stability of the linkage and minimization of l ipid exchange (Parr, M . et al. 1994; Al l en , T. et al. 1991). However, a recent study by Vertut-Doi et al. (1996) has shown that cholesterol-anchored P E G was superior to P E - P E G in its ability to inhibit uptake of liposomes by macrophages. A s a result of the prolongation of liposome blood circulation times, liposomes with incorporated PEG-l ip ids containing anti-cancer drugs have demonstrated improved efficacy in animal models (Gabizon, A . and Papahadjopoulos, D . 1988; Papahadjopoulos, D . et al. 1991). PEG-liposomes have also been shown to provide prolonged delivery of antibiotics and peptide hormones (Bakker-Woudenberg, I. et al. 1993; Woodle, M . et al. 1992). 1.7 Overall Objectives The overall aim of this project was to define the character of the interactions between liposomes and complement and to determine whether these interactions could be decreased by the addition of P E G to the liposome surface. This general aim has been divided into three separate objectives. 1. The first objective of this study was to characterize the antibody-independent Clq-mediated complement activation by anionic liposomes. The effects of liposome charge density, p H , and ionic strength on C l q binding to anionic liposomes were determined to assess the electrostatic component of the interaction. The binding of 35 C l q to anionic liposomes was then measured in human serum in order to determine whether the presence of the rest of the serum components has an affect on the binding affinity or on the amount of C lq bound. Since liposome size effects liposome residency times in vivo, the effect of liposome size on C l q binding and complement activation was also investigated. These studies were designed to provide insight into the requirements for C l q binding and subsequent complement activation on a model cell. 2. The second objective was to assess the role of a highly cationic region of C l q for involvement in liposome binding. This region, comprised of residues 14-26 of the ClqA polypeptide chain, has been shown to mediate C l q binding to other immunoglobulin-independent activators of the classical pathway of complement. 3. The third objective of this study was to assess the effect of liposomal surface PEG on C l q binding and total complement activation by liposomes. While the ability of PEG to keep proteins off of liposomes has been demonstrated, this study is the first to investigate the ability of incorporated PEG to act as a barrier against a specific component of human plasma, the complement system. 36 2. MATERIALS A N D METHODS 2.1 Reagents: 2.1.1 Antisera and Antibodies Goat-anti-human C l q antiserum was purchased from Quidel (San Diego, C A ) . HRP-conjugated secondary antibody (rabbit anti-goat) was either from Jackson Immunoresearch (WestGrove, P A ) or from Chemicon (Temeculla, C A ) . Hemolysin for sensitizing sheep red blood cells was from Diamedix Corp. (Miami , F L ) . The source of whole human IgG was Jackson Immunoresearch. 2.1.2 Lipids . Phosphatidylcholines ( E P C , D P P C , D M P C , D O P C ) , phosphatidylglycerol (PG), and cardiolipin (CL) were purchased from Avanti Polar Lipids , Inc. (Alabaster, A L ) . Cholesterol (CH) was purchased from Sigma (St. Louis , M O ) . 1,2-distearoyl-sn-glycero-3-phosphatidylethanolamine-mPEG 2 0 0 0 ( D S P E - P E G 2 0 0 0 ) was from Shearwater Polymers Inc. (Huntsville, A L ) . Liposome markers: 3 H - C H E (cholesterol hexadecyl ether) and 3 H - D P P C were from Dupont (Boston, M A ) and Amersham (Little Chalfont, U K ) , respectively. 2.1.3 Peptides C l q A peptides were synthesized by the Biotechnology N A P S Unit protein laboratory, University of British Columbia. The following peptides were made: authentic 37 ClqA ( 1 4 . 2 6 ) (A-G-R-P-G-R-R-G-R-P-G-L-K); C lqA ( s c r a m b l e d ) (K-P-R-G-L-G-G-A-G-R-R-P-R); ClqA ( P . A ) in which both proline residues were replaced with alanine residues (A-G-R-A-G-R-R-G-R-A-G-L-K); C l q A ( 2 + ) i n which three of the cationic residues were replaced with glycine residues (A-G-G-P-G-R-G-G-R-P-G-L-G); and C l q A ( 0 + ) in which all cationic residues were replaced with glycine residues (A-G-G-P-G-G-G-G-G-P-G-L-G). A 13 amino acid peptide which was related to the terminal sequence of fibrinogen, (Fgn) (C-H-H-L-G-G-A-K-Q-A-G-N-V), was used as a control. A comparison of the properties of these peptides is presented in Table 9. 2.1.4 Other Reagents 14C-acetic acid and Na1 2 5I were purchased from Amersham. Enzymobeads were purchased from BioRad (Hercules, CA) but are no longer available. Iodogen was from Pierce (Rockford, IL). Sheep red blood cells used in the hemolytic assay were initially from Diamedix Corp. and were later supplied by Cederlane (Hornby, Ont.). Vacutainer tubes for blood collection were from Becton Dickinson (Fairview, NJ). For gel electrophoresis and Western blotting, the mini-electrophoresis chamber, mini trans blot apparatus and molecular weight markers were purchased from BioRad. Five percent polyacrylamide precast IEF phast gels with pharmalyte 3-10 were purchased from Pharmacia Biotech. (Uppsala, Sweden). ECL hyperfilm was from Amersham. Pure human C l q was purchased from Quidel and from Calbiochem (San Diego, CA). Immulon-2 microtitre plates used for ELISA experiments were purchased from Dynatech Laboratories (Chantilly, VA). The substrate used to detect peroxidase-bound antibody, 38 3,3',5,5'-tetramethylbenzidine (TMB), was from Sigma. Polyallomer centrifugation tubes (5 x 20 mm) used in the equilibrium binding assays were from Beckman (Palo Alto, CA) . The P E G 6 0 0 and P E G 1 0 0 0 which were used to synthesize C H - P E G 6 0 0 and C H - P E G 1 0 0 0 were purchased from Aldrich (Milwaukee, WI). P E G _ 1 4 0 0 used to synthesize radiolabeled C H -PEG-NH-COCH3 was from Sigma. 2.2 Preparation of Reagents for Experiments 2.2.1 Preparation of Normal Human Serum Normal human serum (NHS) pools were prepared from venous blood collected from healthy donors in Vacutainer tubes. When blood was collected in SST serum separation tubes, the tubes were left at room temperature (RT) for 20 min and were then centrifuged at 1900 x g for 10 min. The serum was poured off, pooled and stored at -80°C. When blood was collected in whole blood Vacutainers, cells and platelets were spun down at 1900 x g for 15 min and the plasma was removed. A clot was produced by adding CaCl 2 to give a final concentration of 20 m M in the pooled plasma and incubating at 37°C for 45 min. The clotted plasma was poured through gauze and the resulting serum was diluted into aliquots and stored at -80 °C. 2.2.2 Purification of C l q from Human Plasma C l q was purified from acid citrate dextrose-anticoagulated human plasma obtained from the Canadian Red Cross using the method of Tenner et al. (1981) with modifications. Four units of plasma were pooled and treated with 20 m M CaCl 2 to initiate 39 clot formation. Serum was obtained by pouring the clotted plasma through gauze. To remove lipids, the serum was centrifuged at 12 500 x g for 30 min at 4°C and the serum was removed by pipette from underneath the floating l ipid layer. The clarified serum was then passed through a 500 ml column of 100-200 mesh size BioRex 70 (BioRad). In order to increase retention of C l q on the column, the Biorex column was run at p H 7.1 rather than p H 7.3. After washing with starting buffer (50 m M sodium phosphate, 2 m M E D T A , and 82 m M N a C l , p H 7.1), C l q was eluted from the column in 50 m M sodium phosphate buffer with 2 m M E D T A and a N a C l gradient starting with 82 m M N a C l and finishing at 300 m M N a C l . Fractions containing immunoreactive and functional (with respect to IgG binding) C l q as detected by C l q E L I S A were pooled and concentrated by precipitation on ice with 33% ammonium sulfate saturated at R T , p H 7.1. After redissolving the pellet in 50 m M Tris with 1 m M E D T A and 500 m M N a C l , p H 7.2, the semi-pure C l q was applied to a Biogel A 5 m gel filtration column (2 x 140 cm). Fractions containing functional C l q as detected by C l q E L I S A were again pooled and concentrated by precipitation on ice with 33% ammonium sulfate saturated at R T , p H 7.1. The pellet was redissolved in a Tris buffer with 50 m M Tris , 500 m M N a C l , p H 7.2. The removal of brownish insoluble material (probably a contaminant from the ammonium sulfate) was facilitated by centrifugation at 3000 x g for 10 min. Purified C l q was divided into 50 /*1 aliquots and stored at -80°C. 2.2.3. Preparation of Clq-Depleted Serum Clq-depleted serum (C lq -DS) was generated during the isolation of C l q from 40 human plasma. The serum effluent from the BioRex 70 column was pooled, divided into 0.3 m l aliquots, and stored at -80°C. To assure that the pooled serum was deficient of C l q , samples run on S D S - P A G E were transferred to nitrocellulose and probed for C l q using goat anti-human C l q antiserum. N o C l q bands were detected in lanes with C l q - D S samples in the Western blots. In addition, C l q - D S samples did not support any hemolytic activity in the C l q hemolytic assay. 2.2.4 Radiolabelling of C l q Purified human C l q was radiolabeled using either the Enzymobead method or the Iodogen method (Tenner, A . et al. 1981 and Tenner, A . personal communication). C l q is very sensitive to the method of iodination such that a loss of C l q hemolytic activity may occur depending on the radiolabelling method used and on the ratio of Na 1 2 5 I to C l q (Tenner, A . et al. 1981 and Heusser, C . et al. 1973). Two hundred fig of purified or commercial C l q was labelled at a time. Commercial C l q was dialysed into high salt, C l q storage buffer (either 50 m M Tris or 1.8 m M sodium barbital and 3.1 m M barbituric acid with 500 m M N a C l , p H 7.2) to remove the glycerol in the commercial storage buffer prior to radiolabelling. The Enzymobead method involved the incubation of 200 ^g of C l q and 100 u-Ci of N a 1 2 5 I with 25 /xl of Enzymobeads (lactoperoxidase glucose oxidase-coated beads) suspended in distilled water. After the addition of 25 ul of substrate, a 1% (3-D-glucose solution, the labelling reaction proceeded for 15-20 min at R T . Enzymobeads exhibited a batch to batch variation and they are now no longer available. The iodogen method involved precoating a glass test tube with 10 /xg of 41 iodogen according to the Pierce Iodogen insert, addition of 200 pig of C l q in C l q storage buffer with the desired amount of N a 1 2 5 I , and a 10 min R T incubation. For both methods, the reaction product was checked by precipitating the protein with trichloroacetic acid ( T C A ) . The protein precipitate and the solution were assessed for 1 2 5 I counts using an L K B Wallac gamma counter (Compugamma model 1282). Separation of free 1 2 5 I from 1 2 5 I - C l q was achieved by running the reaction mixture through a PD-10 column (Pharmacia Biotech.) which had been precoated with 0.5 mg B S A . Fractions were tested for 1 2 5 I incorporation by T C A precipitation as above and fractions in which > 98 % of the 1 2 5 I was incorporated into the protein were pooled. The molar ratio of N a 1 2 5 I to C l q was either 1:4 or 1:10, depending on the specific activity desired. 2.2.5 Synthesis and Analysis of C H - P E G Derivatives C H - P E G derivatives with 600 and 1000 mean M W P E G s were synthesized by Dr . Johan Janzen according to Patel (Patel, K . et al. 1984) with the addition of a preliminary distillation from 1,4-dioxane in order to dry the P E G stock. C H - P E G _ 1 0 0 0 - N H 2 was provided by Dr . Steven Ansel l . C H - P E G _ 1 0 0 0 - N H 2 was synthesized in order to facilitate the production of radiolabeled C H - P E G _ 1 0 0 0 . The synthesis required four steps and is shown in diagrammatic form in Figure 3. First, P E G 1 0 0 0 was reacted with methanesulfonylchloride (MsCl ) in benzene and in the presence of triethylamine (Et 3 N) to form the PEG 1 0 0 0 -mesylate ( P E G 1 0 0 0 - O M s , P E G 1 0 0 0 - ( S O 2 C H 3 ) ) (personal communication Harris, J.). CH-alkoxide was then prepared by treating cholesterol dissolved in benzene with potassium hydride (KH) (personal communication 42 Harris, J.)- The PEG 1 0 0 0 -mesylate was then reacted with the cholesterol-alkoxide to form C H - P E G _ 1 0 0 0 . Final ly, to produce C H - P E G - N H 2 , a vast excess of concentrated ammonia was added to the dried C H - P E G mixture residue and the reaction was allowed to proceed for 1 week. The reaction mixture was passed twice down a silica column to isolate the C H -P E G - N H 2 product from side reaction products. T L C was run on silica column fraction samples. C H - P E G - N H 2 stained positive with molybdate reagent and with fluorescamine, allowing for the distinction of product fractions. The product was dissolved in ethanol, treated with activated charcoal to remove discoloured contaminants, and filtered through celite. The solvent was removed on a Rotovap and the residue was dissolved in water and lyophilized. The average molecular weight of the P E G was determined to be 1400 g/mole by 400 M H z ' H - N M R using chloroform-d (Cambridge Isotope Laboratories, Andover, M A ) as the solvent. 43 FIGURE 3: Preparation of CH-PEG-NH 2 MsC!/Et3N KH/Cholesterol 44 2.2.6 Synthesis of C H - P E G , ^ - N H - ^ C O C H , and C H - P E G ^ n - N H - C O C H , Labelled C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 and unlabelled C H - P E G 1 4 0 0 - N H - C O C H 3 were prepared with the guidance and assistance of Dr . Steven Ansel l . Radiolabeled C H - P E G was produced by reacting C H - P E G 1 4 0 0 - N H 2 with 1 4C-acetic acid in the presence of N-hydroxysuccinimide (NHS) and l-ethyl-3-(3-dimethylaminopropyl)carbodiimide ( E D C ) at a mole ratio of 0.58 : 1 : 2.15 : 3.17 ( C H -P E G - N H 2 : C H 3 1 4 C O O N a : N H S : E D C ) at room temperature for 30 min under basic conditions. To determine whether the reaction was complete, T L C was run with the product and starting material ( C H - P E G - N H 2 ) on silica using a methanol:chloroform (15:85) solvent system. N o residual starting material remained after 30 min. The reaction mixture was washed with water twice. Organic phases were pooled. The water phases were also washed with chloroform and organic phases were added to the organic pool. The crude product pool was dried with magnesium sulphate and the filtered solution was then concentrated under dry nitrogen gas. T L C on silica plates in methanol:chloroform (15:85) was performed and visualized by ammonium molybdate in acetic acid or fluorescamine solution to determine the degree of substitution generated. The reaction mixture was purified by silica chromatography. Collected fractions were assessed by T L C . Pooled fractions containing one spot for C H - P E G 1 4 0 0 N H - 1 4 C O C H 3 were dried under nitrogen and were then lyophilized. C o l d C H - P E G 1 4 0 0 N H - C O C H 3 was synthesized in a similar but simpler manner. C H - P E G 1 4 0 0 - N H 2 was reacted with acetic anhydride at a ratio of 1: 2 ( C H - P E G i 4 0 0 - N H 2 : acetic anhydride) for 20 min at room temperature. To extract the crude product, the 45 reaction mixture was water washed and water phases were further mixed with dichloromethane. The organic phases were pooled and product was isolated by silica chromatography as above. 2.2.7 Preparation of Liposomes Liposomes were made as previously described (Hope, M . et al. 1985). P E G - l i p i d containing liposomes were prepared by adding the P E G - l i p i d derivative in chloroform to the rest of the constituent lipids. Dried lipids were resuspended in buffer and following five freeze-thaw cycles, the resulting multilamellar vesicles ( M L V s ) were extruded under pressure through two stacked 100 nm or 400 nm diameter pore size polycarbonate filters (Costar, Cambridge, M A ) (extruder from Lipex Biomembranes, Vancouver, Canada). Various buffers for liposome resuspension were used in different assays. For the hemolytic assay and C l q E L I S A , liposomes were resuspended in isotonic veronal-buffered saline ( V B S : 1.8 m M sodium barbital, 3.1 m M barbituric acid, 145 m M N a C l , p H 7.4). For liposomes used in the equilibrium binding assay in the presence of serum and to visualize liposome-bound C l q by autoradiography, the resuspension buffer was 1.8 m M sodium barbital, 3.1 m M barbituric acid, 50 m M N a C l and 13.4% sucrose, p H 7.3. For liposomes used in the equilibrium binding assay in the absence of plasma proteins other than C l q , the liposome buffer contained 1.8 m M sodium barbital, 3.1 m M barbituric acid, 11% sucrose, and either 20, 100, or 145 m M N a C l depending on the ionic strength required for the assay. A second liposome buffer containing 11 m M Tris , 46 m M N a C l , and 7.0 % sucrose was also used for this purified system assay. 46 Large unilamellar vesicle ( L U V ) size distributions were analyzed using the Nicomp Submicron Particle Sizer (Model 270) (Particle Sizing Systems, Inc. Santa Barbara, C A ) in vesicle Gaussian distribution mode. Liposomes extruded through 100 nm pore size filters were 110 + 20 nm in diameter; liposomes extruded through 400 nm pore size filters were 240 + 38 nm in diameter. The l ipid composition had no effect on the liposome size. Liposome phospholipid content was determined using a colourimetric phosphate assay (Fiske, C . and Subbarow, Y . 1924). The total l ipid concentration was determined from the concentration of phosphate in the liposome sample as determined by the phosphate assay and the molar ratio of phosphate to the phospholipids (eg. 2:1 for cardiolipin and 1:1 for other phospholipids). The calculation follows: total l ipid concentration =( phosphate concentration / mole % of phosphate in the liposomes). 2.3 Analytical Techniques 2.3.1 Autoradiography of ^ I - C l q Autoradiography was carried out on either labelled C l q alone or on C l q which had been incubated with liposomes. In the case of untreated 1 2 5 I - C l q , reduced and non-reduced samples were run next to molecular weight markers on 7.5% S D S - P A G E . Protein bands were transferred onto nitrocellulose paper using a mini transblot apparatus. The paper was allowed to dry and was then loaded into an X-ray cassette with E C L hyperfilm. 47 To visualize 1 2 5 I - C l q on liposomes, 1 2 5 I - C l q was incubated with 240 nm diameter liposomes in a mixture consisting of 2 m M total l ipid, C l q depleted serum diluted 1/60 and 1 2 5 I - C l q at 1.7 ttg/ml (an equivalent concentration for physiologic C l q concentration diluted 1/60) for 20 min at R T . Liposome-bound C l q was separated from free C l q by loading the reaction mixture onto an intermediate density buffer (3.2% sucrose, 145 m M dextrose, 1.8 m M sodium barbital, 3.1 m M barbituric acid, and 145 m M NaCl ) and spinning for 30 min at 166,300 x g in a Beckman ultracentrifuge, model TL-100 using a swinging bucket rotor. The polyallomer 5 x 20 mm centrifugation tubes were frozen and sliced. After resuspending the liposome pellets in surrounding buffer, pellet samples were subjected to 10% S D S - P A G E and autoradiography as described above. 2.3.2 Functional Complement Assay A modified hemolytic assay was used to measure the amount of complement consumed by liposomes in human serum. Since only one optimal serum concentration of human serum was used, this assay does not provide C H 5 0 (Z) values. Liposomes were serially diluted in V B S containing 0.15 m M C a C l 2 , 1 m M M g C l 2 , and 0.1% gelatin ( G V B 2 + ) . A n equal volume of diluted normal human serum (1:4) was then added to each tube. Liposome lysis controls (colour blanks) were run in parallel with the test samples and consisted of liposomes incubated with buffer and without serum. Following a 30 min incubation at 37°C, G V B 2 + was added to the serum/liposome mixture. The total residual complement content of the liposome-treated serum was then measured by adding antibody sensitized sheep red blood cells ( E A cells) to the liposome/serum mixture. Briefly, sheep 48 erythrocytes that were sensitized with rabbit I g M anti-sheep R B C antibody (hemolysin) by incubating washed sheep red blood cells with hemolysin at a ratio of 1:1000 for 45 min at 37°C were suspended at a concentration of 10 9 cells/ml in G V B 2 + . In duplicate, an equal volume of E A cells was incubated with liposome-treated serum for 30 min at 37°C. The hemolytic reaction was stopped by addition of V B S containing 0.1% gelatin and 20 m M E D T A ( G V B - E D T A ) . Unlysed cells were pelleted by centrifugation and the amount of hemoglobin released was quantified spectrophotometrically at A 4 1 5 n m . The amount of complement remaining in each tube was compared with that of serum incubated in the absence of liposomes (100% lysis control). Percent E A cell lysis was calculated using average O D values for the duplicates as follows: % E A lysis = ( O D 4 1 5 f o r t e s t s a m p ] e - O D c o l o u r b l a n k ) / ( O D 4 1 5 f o r m % - O D C O | 0 u r b l a n k ) x 100. Percent complement depletion was calculated as 100 - % E A lysis. To measure the capacity of C l q peptides to inhibit complement activation by 100 nm diameter anionic liposomes, the above hemolytic assay was used with the additional incubation of peptides with liposomes prior to the addition of serum. Peptides were first serially diluted in G V B 2 + . A n equal volume of E P C : C H : C L (35:45:20 mole %) liposomes (5mM) was added to the peptides and the mixture was incubated at R T for 40 min. 2.3.3 C l q Hemolytic Assay To assess the hemolytic capacity of purified C l q , the standard hemolytic assay (CH50) was altered such that Clq-depleted serum (C lq -DS) was used in the place of 49 N H S . Purified or labelled C l q was then added back at 50 /xg /ml (although C l q is present in human serum at 80-100 / ig/ml, a titration of C l q into the C l q hemolytic assay showed a plateau of activity at 50 /xg/ml). After serially diluting the C l q - D S / C l q mixture, E A cells were added. The amount of E A lysis was determined as described above and C H 5 0 values were derived by regression. C l q - D S alone had no hemolytic activity. When purified C l q was added back, the activity achieved was compared with that of N H S . The C l q hemolytic assay was carried out in duplicate for each test sample. C l q activity was expressed as follows: % N H S C H 5 0 = C H 5 0 ( s a m p , e C l q ) / C H 5 0 ( N H S ) x 100 using average C H 5 0 values for the duplicates. 2.3.4 C l q Enzyme Linked Immunosorbant Assay (ELISA) A C l q E L I S A was used to detect C l q binding to liposomes as indicated by the depletion of C l q from liposome-treated serum. Whole human IgG was aggregated by heating at 62°C for 40 min in 50 m M T r i s - H C l , p H 8.6. Immulon-2 microtitre plate wells were then coated with 130 [A of aggregated human IgG at 20 / ig /ml . The wells were blocked with Tris buffer containing 1% gelatin, p H 8.0, for 2 hr at R T . Liposomes were serially diluted in G V B 2 + and were then incubated with an equal volume of either neat or diluted (1/30 in G V B 2 + ) serum at 37°C for 30 min. Liposome-treated serum samples were then diluted with an equal volume of G V B 2 + . A 130 i i i aliquot of each diluted liposome-treated serum mixture was applied to the wells in duplicate. Following incubation for 60 min at R T , the wells were washed with PBS containing 0.05% Tween 20 to remove liposomes, free C l q and liposome-bound C l q . The amount of C l q bound to the IgG-50 coated wells was determined using a two-step detection system where the primary antibody was goat anti-human C l q and the secondary antibody was peroxidase-conjugated rabbit anti-goat IgG. Both antibodies were diluted to 1:3000 in PBS containing 3% powdered skim milk. The substrate used was T M B . Colour development was stopped by the addition of 2 M HC1 after 2 min and was measured at A 4 5 0 . Max ima l C l q binding to the well occurred when serum incubated in the absence of liposomes was used (100%). The percent of C l q bound to the IgG coated well was calculated using average values of O D for the duplicates as follows: % C l q bound to well = ( O D 4 5 0 t e s t s a m p l e -b lank) / (OD 4 5 0 1 0 0 % -blank) . The percent of C l q depleted from the serum was then calculated as 100% -% C l q bound to wel l . To assess whether negatively charged liposomes were competing with free C l q for binding to the IgG coated wells, the E L I S A was carried out on samples which had the liposomes removed by centrifugation before addition to the wel l . Results were the same for C l q depletion by liposomes whether or not the liposomes were removed prior to sample addition to the wel l . 2.3.5 Equil ibrium C l q Binding Assay Equil ibrium binding measurements were made with liposomes suspended in pure C l q and in diluted human serum. In the purified protein system using barbital buffers, reaction mixtures consisted of 15 /xl of C l q / 1 2 5 C l q mixture, 5 u\ of sucrose-containing ~ 240 nm liposomes, and 80 /tl of diluting buffer. To obtain a final ionic strength of 145 m M N a C l , the dilution buffer consisted of 1.8 m M sodium barbital, 3.1 m M barbituric 51 acid, 77.9 m M N a C l , and 7.2% D-glucose. Altered ionic strength conditions required dilution of buffers to give different salt concentrations such that for the 100 m M reaction, 25 m M N a C l was required in the dilution buffer and for 20 m M reaction, there was no salt in the dilution buffer. In the purified protein system using Tris buffers, reaction mixtures consisted of 30 ul of C l q / 1 2 5 I - C l q mixture, 5 ul of sucrose-containing ~ 240 nm liposomes, and 65 ul of diluting buffer (10 m M Tris). In the diluted serum system, the reaction mixture consisted of 25 ul sucrose-containing ~ 240 nm liposomes and 75 / i l of a mix of G V B 2 + , C l q - D S , and C l q / 1 2 5 I - C l q . Following a 20 min reaction at R T , liposome-bound C l q was separated from free C l q by centrifugation. A 40 ul sample of reaction mixture was layered onto 180 ul of an intermediate density separating buffer (buffers described below) in 5 x 20 mm polyallomer tubes. This was done in duplicate. Tubes were centrifuged for 30 min at 166,300 x g in a Beckman ultracentrifuge, model TL-100 using a swinging bucket rotor. This spin resulted in pelleting of > 90 % of the liposomes. The separation buffer for the barbital purified protein system contained 1.8 m M sodium barbital, 3.1 m M barbituric acid, 4.5% sucrose and 2.9% D-glucose with N a C l at the desired concentration for the experiment (i.e. 20 m M , 100 m M , or 145 m M ) . The separation buffer for the Tris purified protein system contained 10 m M Tris , 110 m M N a C l , and 3.4 % sucrose. For the serum system, the separating buffer was 1.8 m M sodium barbital, 3.1 m M barbituric acid, 3.5% sucrose and 2.9% D-glucose with N a C l at 145 m M . After centrifugation, the tubes were frozen and sliced into two pieces. While the pellet slice contained liposome-bound C l q , the supernatant slice provided a measure of 52 the equilibrium C l q concentration. Tube slices were counted in an L K B Wallac gamma counter (Compugamma model 1282). A s a control, to determine the amount of C l q spinning down in the absence of liposomes, tubes were run in which no liposomes were present ( C l q control-also done in duplicate). The amount of C l q spun down in this control tube was subtracted from the binding values for the l iposome/Clq reaction tubes. The amount of C l q bound to liposomes was calculated as follows: fig C l q bound = (cpm p e l l e t - c p m ^ c o n t r o l )/specific activity of C l q mixture. 2.3.6 Clq/ l iposome Binding Inhibition Assay Additional experiments were conducted to measure the inhibition of C l q binding to anionic liposomes by incorporation of PEG-l ipids and by competition with C l q A peptides. To assess the effect of P E G - l i p i d incorporation on C l q binding, liposomes with PEG-l ipids were used in the equilibrium C l q binding assay. The percent inhibition of C l q binding was determined by comparison with the amount of C l q binding to liposomes containing cardiolipin and no PEG- l ip id . Percent inhibition of C l q binding was calculated as follows: % inhibition = 100 - ( ( C l q b o u n d ( C L 2 0 + P E G . l i p i d ) / C l q b o u n d ( C L 2 0 ) ) x 100) where Clqb 0und = (% of cpm in pellet) s a m p l e - (% of cpm in pellet) c lq control' To measure the inhibition of C l q binding to liposomes by C l q A peptides, peptides were added to the l iposome/Clq reaction and residual C l q binding to liposomes was measured using the equilibrium C l q binding assay. Reaction mixtures consisted of 74 jul pure system diluting buffer, 15 /xl of 1 2 5 I - C l q , 5 of liposomes, and 6 /xl of peptide or peptide buffer. The % inhibition of C l q binding was calculated as follows: 53 % inhibition = 100 - ( C l q b o u n d ( C L 2 0 1 i p o + p e p t i d e ) / C l q b o u n d ( C L 2 0 ) x 100) where C l q b o u n d = (% of cpm in pel let) s a m p l e - (% of cpm in p e l l e t ) c l q c o m r o l . 2.3.7 Particle Electrophoresis A Rank Mark I (Rank Bros. , Bottisham, U . K . ) electrophoresis apparatus with a quartz cylindrical chamber fitted with silver chloride electrodes was used to measure electrophoretic mobilities of M L V s . A horizontal microscope equipped with a water immersion objective allows for the observation of the migration of individual particles at constant temperature in an applied electric field at a magnification of 320 times. The velocity of individual liposomes was measured at 25°C at the stationary level (Seaman, G . 1975). The electrophoresis chamber, which holds a volume of 2.6 ml and needs to be filled such that no air bubbles are present, was flooded with liposomes and each particle selected visually was timed manually in each direction with reversal of the electric field. Making velocity measurements in both directions largely eliminates any errors due to drift. Ten or more particles were timed as they moved across the eye piece reticule per experimental condition. The mobility was calculated from the averaged velocities, the applied voltage, and the chamber electrical length according to the following calculation: electrophoretic mobility (u) = velocity of the particle (/xm/sec) / electric field strength (E) (volt/cm) where E = (voltage/1) and 1 = the effective distance between the electrodes (Seaman, G . 1975). For experiments assessing the interaction of C l q or C l q peptides with liposomes, 54 liposomes (0.36 m M total lipid) were pre-incubated with C l q or peptides for 40 min at R T in a total volume of 780 /xl. Just prior to flooding the electrophoresis chamber with the reaction mixture, buffer was added to bring the mixture volume to 2.8 m l . The requirement to f i l l the chamber presents a disadvantage for these binding experiments which w i l l be addressed with respect to the experimental data in the Discussion. The advantages of this method include the fact that measurements are made rapidly, with the timing for each vesicle being from 5-15 seconds. The method also permits direct observation of the shape and size of the liposomes. In addition, the migration rate of individual vesicles can be compared so that any differences in the electrokinetic behaviour of liposomes may be established. 2.3.8 Analysis of C H - P E G 1 4 n n Incorporation into Liposomes A blend of labelled and unlabelled C H - P E G 1 4 0 0 - N H C O C H 3 was used in the preparation of liposomes composed of E P C : C H : C L : C H - P E G at 35:45-n:20:n mole % where n is the mole % of C H - P E G 1 4 0 0 - N H C O C H 3 . Either 3H-cholesterol hexadecyl ether ( C H E ) or 3 H - D P P C was used as a liposome marker at 2000-2500 cpm//xl in liposomes at 5 m M total l ipid. Liposomes were extruded through 100 nm diameter pore size filters. Any foam present in the extruded liposomes was collapsed back into the liposome solution by applying a stream of nitrogen onto the surface. These liposomes were then passed down a 3 cc syringe Sepharose C L 2 B (Pharmacia Biotech.) column in order to separate liposomes from free C H - P E G micelles. One hundred /xl fractions were collected. Fraction samples were dissolved in CytoScint scintillation fluid ( ICN Biomedicals, Inc. Costa 55 Mesa, C A ) and counted in a Philips scintillation counter (model PW4700). Counts for the three peak fractions were used to determine the ratio of 1 4 C - P E G to 3H-liposomes. This 1 4 C / 3 H ratio was compared with the initial ratio for unchromatographed liposomes to determine the percent of the initial ratio. The % incorporation was calculated as follows: % incorporation = (mole % C H - P E G added) x ((peak 1 4 C / 3 H ratio)/(initial 1 4 C / 3 H ratio)). 2.3.9 Assessment of C H - P E G Exchange with Normal Human Serum Liposomes used in this assay were composed of E P C : C H : C L : C H - P E G 1 4 0 0 at 35:30:20:15 mole % with a trace amount of 3 H - D P P C (3500 cpm/^l in liposomes at 5 m M total lipid) where C H - P E G 1 4 0 0 was a mixture of labelled and unlabelled C H - P E G 1 4 0 0 -N H C O C H 3 . Liposomes were extruded through 100 nm diameter pore size filters and any foam present was collapsed back into the liposome solution by applying a stream of nitrogen onto the surface. Liposomes were incubated with an equal volume of human serum or with buffer (either V B S or H E P E S : 10 m M Hepes, 145 m M N a C l , p H 7.4) at 37°C. Fifty u\ samples were removed at time zero (immediately after adding serum to the liposomes), 30 min, 1 hr, 2 h r , 4 h r , 8 h r , 12hr , and 24 hours. Samples were run down Sepharose C L 2 B columns in 3 cc syringes. Fractions were collected and sampled for scintillation counting. Counts from the top three fractions were used to determined the C H - P E G - 1 4 C to 3 H -liposome ratio. The % of initial 1 4 C / 3 H ratio was calculated as follows: % initial 1 4 C / 3 H ratio = (peak 1 4 C / 3 H ratio)/(initial 1 4 C / 3 H ratio) x 100. 56 Chapter 3. Complement Activation by Liposomes via C l q Binding and Effect of Liposome Size 3.1 Complement Activation by Anionic Liposomes and Not Neutral Liposomes While previous studies in human serum demonstrated that neutral liposomes composed of E P C : C H failed to activate complement (Marjan, J. et al. 1994), many other neutral compositions using well defined phosphatidylcholines had not previously been tested for their complement activating potential in human serum. A modified hemolytic assay was used to test the ability of 100 nm diameter liposomes composed of 100 % D P P C , D S P C : C H (55:45 mole %) or D M P C : C H (55:45 mole %) to activate complement (Figure 4). While anionic liposomes composed of E P C : C H : C L ( 35:45:20 mole %) showed liposome dose-dependent complement activation, none of 100% D P P C , D S P C : C H , or D M P C : C H liposomes activated complement under these experimental conditions. 57 FIGURE 4 : Complement Consumption in Human Serum by Liposomes Composed of Different PC: Requirement for Negative Surface Charge. Complement consumption by 100 nm liposomes was assessed by hemolytic assay. Liposomes were composed of E P C : C H (55:45 mole %) (O) , E P C : C H : C L (35:45:20 mole %) ( • ) , D M P C C H (55:45 mole %) (A), D S P C : C H (55:45 mole %) ( • ) , and D P P C (100 %) (•). The liposome concentration was the concentration of total l ip id and was determined as described in Materials and Methods. For 100 nm liposomes, the total exposed surface area and number of vesicles are related to the total l ipid concentration in the following ways: one cm 2 = 4.75 x 10"4 ptmole total l ipid; one vesicle = 1.5 x 10~13 /xmole total l ipid. The liposome/serum reaction volume was 100 a\. Data symbols represent the average of duplicates. 3.2 Evaluation of Purity and Function of Isolated and Radiolabelled C l q Four different C l q preparations were used throughout this work. Each of the C l q preparations was assessed for purity and function at the time of purification, after labelling, and at various times after storage. Isolated C l q was assessed for purity by S D S - P A G E with silver staining, Western blotting, and isoelectric focusing (IEF). Figure 5 A shows a silver stained S D S - P A G E evaluation of the purity of crude C l q and C l q final product. While the lanes with partially purified C l q (lanes 1 and 2) exhibited several higher molecular weight bands, only the three bands correlating to C l q subunits A , B , and C were visible in the product lane (lane 5). The enhanced staining of the C l q C chain band is an artifact of silver staining. Coomassie staining of reduced C l q on S D S - P A G E shows an equal staining of the C l q A , B , and C chain bands. Lane 6 contained a high molecular weight contaminant that was removed from the C l q preparation by Biogel size exclusion. Western blots of the purified product showed only C l q subunit bands. A n I E F "phast" gel was run for samples of purified and commercial C l q and was then silver stained (Figure 5B). Lanes 5 and 6 show the p i band for purified C l q and lanes 7 and 8 show the commercial C l q p i band. A p H calibration curve was plotted and the p i range for C l q was determined to be 6.8 - 7.1. This p i determination was carried out twice and is within the p i range previously cited in the literature (6.1 to 7) (Rosano, C . and Hurwitz, C . 1977). It should be noted, however, that there is some controversy with regard to the p i of C l q . This w i l l be discussed in detail in section 7.1.1. Radiolabelled C l q was assessed in a similar manner for purity. Figure 5C shows the autoradiograph of 1 2 5I-labelled C l q . Under reducing conditions (lanes 1 and 2), three 59 bands for the A , B , and C polypeptide chains of C l q were visible. The C chain band was darkest because 90.1% of the 1 2 5 I attaches to the tyrosine-rich C l q C chain (Tenner, A . et al. 1981). Under non-reduced conditions (lanes 4 and 5), the A - B and C - C dimers were visualized. The band at - 2 2 0 k D likely represents aggregates of C l q chain dimers. Because the function of purified protein is questionable after lengthy storage, C l q function was tested not only after purification and radiolabelling, but also at several times throughout its storage and prior to its use. The following three assays were used: the C l q hemolytic assay (Figure 6), a C l q E L I S A (Figure 7), and the C l q equilibrium binding assay which measures C l q binding to anionic liposomes (Figure 8). The first test of C l q function involved adding purified or labelled C l q back to Clq-depleted serum in the Clq-hemolytic assay. While Clq-depleted serum alone showed no hemolytic activity, when freshly isolated or labelled C l q was added back, 85% of the activity achieved with N H S was attained (Figure 6). The loss of 15% of N H S activity was likely due to the manipulation required to produce Clq-depleted serum or to the dilution effect of the serum which had been run through the cation exchange column. For purified C l q , the original level of hemolytic activity was maintained for at least 6 months. After one year, 77 % of the original activity remained. For labelled C l q , the original activity was maintained for 1.5 months. After 6 months, 76 % of that activity remained. Two samples of non-functional C l q were also tested using the Clq-hemolytic assay. Less than 20 % of N H S activity was measured when C l q which had been labelled using a vast excess of Iodogen was tested. Purified C l q which was boiled for 2 min showed no hemolytic activity. 60 Figure 5: Assessment of C l q Purity Semi-pure and pure C l q were run on a 12 % S D S - P A G E under reducing conditions. Bands were visualized by silver staining (panel A ) . Lanes 1 and 2 contained partially purified C l q from after the Biorex cation exchange column. Lanes 3 and 4 contained broad and low range M W markers, respectively. Lane 5 contained isolated C l q product. Lane 6 contained a contaminant removed from the C l q preparation by Biogel size exclusion. A 5 % polyacrylamide precast I E F phast gel with pharmalyte 3-10 was run with samples of purified and commercial C l q (panel B) . Bands were visualized by silver staining. Lanes 1 and 2 contain broad range p i calibration markers. Lanes 5 and 6 show the p i band for purified C l q and lanes 7 and 8 show the commercial C l q p i band. 1 2 5I-labelled C l q was run on a 10 % S D S - P A G E and was visualized by autoradiography (panel C ) . Lanes 1 and 2 contain reduced 1 2 5 I - C l q and lanes 4 and 5 contain non-reduced 1 2 5 I - C l q . 61 A -«-C1qA —C1qB - * C 1 q C B 1 2 3 4 5 6 -*-Load Load 1 2 3 4 5 6 7 8 kDa 220 • 66 • 46 • 30 • 21.5 • 1 2 3 4 5 62 FIGURE 6 : Hemolytic Activity of C l q and 1 2 S I-Clq as a Function of Storage Time. The C l q hemolytic assay was used to monitor C l q activity. Fifty / ig/ml of C l q or 1 2 5 I - C l q was added to C l q - D S and the hemolytic activity was compared with that of N H S . Isolated C l q ( • ) and radiolabelled C l q (O) batches were tested over time. 1 2 5 I - C l q labelled using too much iodogen (+) and boiled C l q ( • ) were tested on the same day only. Data is composite for three different batches of purified C l q and for three different labelled C l q batches. Data symbols represent the average of duplicates. 100 200 Time (days) 300 400 63 The second test of C l q function, a C l q E L I S A , assessed the ability of purified or labelled C l q to bind aggregated whole human IgG. Pure C l q and 1 2 5 I - C l q binding profiles were superimposable indicating that these C l q preparations maintained their ability to bind IgG for at least one year (Figure 7A) . The capacity to bind aggregated IgG was greatly diminished for 1 2 5 I - C l q which had been labelled using too much Iodogen illustrating that this assay distinguishes between functional and denatured C l q . The ability of different batches of 1 2 5 I - C l q to bind human IgG was further evaluated by measuring the 1 2 5 I counts associated with the E L I S A plate wells (Figure 7B). 1 2 5 I - C l q labelled using 1 2 5 I : C l q ratios of 1:4 and 1:10 showed dose-dependent cpm profiles which were similar to the total C l q profiles in panel A . The 1 2 5 I - C l q batch labelled using too much Iodogen showed very few cpm associated with the IgG coated wells and was not used in this study. C l q quality was further monitored using the equilibrium C l q binding assay (Figure 8). For up to one year, the same level of C l q binding to C L 2 0 liposomes was observed. However, after 14 months at -80°C, the C l q binding level increased more than ten times. A s a result, C l q which was over a year old was not used. 64 F I G U R E 7: Functional Activity of Purified C l q and 1 2 5 I - C l q in Relation to Storage Time. A C l q E L I S A was used to monitor the ability of isolated C l q and 1 2 5 I - C l q to bind heat aggregated human IgG. Panel A shows E L I S A results for purified C l q batches which had been stored for 2 months (v) and 1 year ( • ) . Results for an incorrectly labelled C l q batch (+) and for 1 month old properly labelled batches which used 1 2 5 I : C l q ratios of 1:4 (A) (specific activity = 1.65 x 1 0 6 cpm//^g) or 1:10 (O) (specific activity = 2.4 x 10 5 cpm//ig) are also shown. To assess non-specific C l q binding, isolated C l q was added to B S A coated wells ( • ) . Following the colourimetric detection above, E L I S A plate wells which contained labelled C l q were broken off and counted in a gamma counter. Panel B shows 1 2 5 I counts on the E L I S A plate wells as a function of the amount of C l q added. Data symbols show average values of the duplicates. 65 A 1.50 A d d e d C1q C o n e , (ug/ml) B A d d e d C1q C o n e , (ug/ml) 6 6 FIGURE 8: C l q Binding to Cardiolipin-containing Liposomes as a Function of C l q Storage Time. The equilibrium C l q binding assay was used to monitor the effect of storage time on C l q binding to C L 2 0 liposomes. Binding was measured at a C l q equilibrium concentration of ~ 0.01 uM using a purified system at p H 7.2. The total l ipid concentration in the reaction was 0.5 m M . Data points represent the mean of 3 experiments done in duplicate. Error bars represent one standard deviation. •a o. o E n. "D c o O o E ZL 0.80 0.60 h w 0.40 0.20 h 0.00 Storage Time (months) 67 3.3 C l q Binding to Anionic Liposomes: Observed and Measured To visualize the binding of C l q to anionic liposomes, autoradiography was performed on liposomes incubated with, then separated from a human serum mixture containing 1 2 5 I - C l q (Figure 9). C l q was found on E P C : C H : C L (35:45:20 mole %) liposomes but not on neutral E P C : C H (55:45 mole %) liposomes. Quantitative measurements of C l q binding to anionic liposomes were then made using the C l q equilibrium binding assay. A s the negative surface charge density increased, a consequent increase in the amount of C l q binding to liposomes was observed (Figure 10). Increasing the amount of cardiolipin in the liposome composition from 20 to 30 mole % resulted in a 4 fold increase in the amount of C l q binding. Doubling the cardiolipin content resulted in 40 times more C l q binding. A summary of the binding data for C L 2 0 , C L 3 0 , and C L 4 0 liposomes is presented in Table 3. Apparent association constants and values for the amount of C l q bound at saturation were derived from Scatchard plots (see Appendix). Where the Scatchard analysis yielded a curvilinear plot, a K a value was derived from the steep linear portion of the plot and the amount of C l q bound at saturation was obtained from the direct binding plot. 68 Figure 9: Autoradiograph of C l q bound to liposomes. Liposomes were incubated with Clq-depleted serum doped with 1 2 5 I - C l q for 20 min at R T . Liposome-bound C l q was separated from free C l q by centrifugation as described in Materials and Methods. Following resuspension of liposome pellets, samples were run under non-reducing conditions on a 10 % S D S - P A G E . Purified, radiolabeled C l q was run in lane 1 as a control. Samples from serum/liposome reactions with E P C : C H : C L (35:45:20 mole %) liposomes and E P C : C H (55:45 mole %) liposomes were run in lanes 3 and 4 respectively. The single band at approximately 54 k D represents the dimer of the C l q C polypeptide chain. The very low molecular weight band at the bottom of the gel, at the dye front, may be free iodine. kDa 200 • 97.4 • 69 • 46 • 30 • 1 2 69 FIGURE 10: C l q Binding to Cardiolipin-containing Liposomes at pH 7.2: Effect of Surface Charge Density. C l q binding to 240 nm liposomes at p H 7.2 in a purified system was measured using the C l q equilibrium binding assay. Liposomes were composed of E P C : C H : C L at (55-n):45:n mole % where n = 0 (O) , 20 ( • ) , 30 (v), or 40 (A) mole % of C L . A constant total l ip id concentration of 0.5 m M was used in the reaction. Binding isotherms for C L 0 , C L 2 0 , and C L 3 0 liposomes are shown in panel A . Because far greater levels of C l q binding were measured for C L 4 0 liposomes, the vertical axis was adjusted in panel B in order to show the C L 4 0 binding isotherm in relation to the C L 2 0 and C L 3 0 isotherms. 5 Q. O E "D C o CT O o E 0.70 0.60 0.50 5T 0.40-1 LU — 0.30-0.20-0.10-0.00 OO, ° J 1—^-0.00 0.05 0.10 0.15 0.20 0.25 0 30 (E-1) Equilibrium C1q Cone. (pM) B 2 Q. "5 E c o cr O o 0.70 0.60-0.50-0.40-LU 0.30-0.20" 0.10-0.00 wV^1 ft »dtfl* T<B 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Equilibrium C1q Cone. (pM) < E 1 ) 70 T A B L E 4: Apparent Association Constants and Saturation Values for C l q Binding to Cardiolipin-containing Liposomes at pH 7.2 in a Purified Protein System: Effect of Liposome Surface Charge Density. Saturation binding and association constants for C L 3 0 and C L 4 0 liposomes were derived from linear Scatchard plots (Appendix 1). For C L 2 0 liposomes, a curvilinear Scatchard plot was obtained. In this case, the K a value was derived from the steep linear portion of the Scatchard plot and the C l q saturation value was from the isotherm. To convert yumole total l ipid to number of vesicles, the surface area per vesicle and the number of vesicles per /imole total l ipid were first estimated. The surface area per vesicle was calculated as S A v e s i c l e = 47tr2 and was 1.8 x 10' 9 cmVvesicle for our system. The number of 240 nm vesicles per /imole total l ipid was estimated to be 8.0 x 10 1 1 vesicles//u.mole l ipid based on the following calculations and assumptions: assuming that 35% of the total l ipid is exposed and that the average area per l ipid is 0.7 nm 2 , the number of l ipid molecules per vesicle = (2.9 x 180,960 nm 2)/0.7 nm 2 = 750,000 molecules. The number of vesicles per ^mole of l ipid = (6.022 x 10 1 7 molecules/ jumole)/750,000 molecules = 8.0 x 10 1 1 . Saturation Binding Liposome Composition Scatchard Profile K a ( M 1 ) ± S . E . M . /rniole C lq b o u n d / /imole total lipid molecules of ClqbounAesicle E P C : C H : C L (35:45:20) curvi-linear 2 ± 0.16 X 1 0 9 2.0 x l O " 6 3/2 E P C : C H : C L (25:45:30) linear 5.8 + 0.5 X 1 0 8 6.4 x l O " 6 5/1 E P C : C H : C L (15:45:40) linear 3.8 + 0.44 x l O 8 6.7 x l O 5 50/1 71 To address whether the interaction between C l q and anionic liposomes was reversible, C l q binding to C L 2 0 M L V s was monitored before and after washing by particle electrophoresis (Figure 11). This experiment was carried out at subphysiologic ionic strength ([NaCl] = 0.1 M ) in order to enhance Clq/ l iposome interactions. The effect of ionic strength on C l q binding to anionic liposomes is presented in section 4.4 (page 89). In the absence of C l q , C L 2 0 M L V s had a high mobility in the applied electric field. When C l q was incubated with liposomes prior to vesicle electrophoresis, liposome mobilities decreased drastically due to neutralization of the negative surface charge by C l q binding. After one buffer wash, liposome mobilities returned to the original level for C L 2 0 M L V s alone indicating that liposome-bound C l q had dissociated. These results demonstrated that C l q binding to anionic liposomes under these experimental conditions was reversible. 72 F I G U R E 11: Reversibility of C l q Binding to Anionic Liposomes. Particle electrophoresis was used to measure the mobility of C L 2 0 M L V s alone, with C l q , and after washing. The Clq/ l iposome reaction consisted of 0.134 /xM (61.5 /xg/ml) C l q , C L 2 0 M L V s at 0.36 m M total l ip id , and barbital buffer with 100 m M N a C l at p H 7.0. After 40 min at R T , the reaction mixture volume was made up to 2.8 m l and mobility measurements were made. The diluted reaction mixture was removed from the electrophoresis chamber and M L V s were washed by adding 2 ml of fresh buffer and spinning the liposomes down at 1800 x g for 6 min. After removing the supernatant, the liposome pellet was reconstituted in 2.8 ml fresh buffer and liposome mobilities were measured as before. A second wash was performed in the same manner and the mobility of the washed C L 2 0 vesicles was measured. Mobi l i ty measurements were made on ten vesicles in the mixture. The complete experiment was carried out twice; one data set is shown with the solid line and the other with the dotted line. Error bars represent one standard deviation for the mobilities of the ten vesicles. 3.4 Effect of Liposome Size on Clq-mediated Complement Activation The C l q E L I S A and hemolytic assay were used to compare C l q binding and complement activating capacities of liposomes with - 1 0 0 and - 2 4 0 nm diameters in normal human serum. Quantitative Clq/ l iposome binding measurements for 240 nm diameter liposomes and M L V s were also compared. T o assess the validity of the competitive C l q E L I S A results, the E L I S A was carried out on samples which had the liposomes removed by centrifugation before addition to the wel l ; this was done to determine whether negatively charged liposomes were competing with free C l q for binding to the IgG coated wells. Results were the same for C l q depletion by liposomes whether or not the liposomes were removed prior to sample addition to the well . The C l q E L I S A i n neat human serum showed that 240 nm diameter C L 2 0 liposomes bound more C l q than 100 nm diameter liposomes at l ipid concentrations below 1 m M (Figure 12). This liposome size dependence was more pronounced in the equilibrium binding assay results which compared C l q binding capacities of 240 nm C L 2 0 liposomes and C L 2 0 M L V s (Figure 13). The level of C l q binding was 15 times greater for M L V s than for 240 nm extruded liposomes. 74 FIGURE 12: Effect of Liposome Size on C l q Binding to Anionic Liposomes in Human Serum. The C l q E L I S A was used to measure C l q binding to C L 2 0 liposomes (expressed as C l q depletion) in 50 % human serum. Liposomes were extruded through 100 nm and 400 nm pore size filters to produce 100 nm (O) and 240 nm ( • ) vesicles respectively. Each point represents the mean of 3 experiments done in duplicate. Error bars represent one S . E . M . The liposome concentration was the concentration of total l ip id and was determined as described in Materials and Methods. For 100 nm liposomes, the total exposed surface area and number of vesicles are related to the total l ipid concentration in the following ways: one c m 2 = 4.75 x 10"4 pimole total l ipid; one vesicle = 1.5 x 10"13 ttmole total l ipid. For 240 nm liposomes, one cm 2 = 6.9 x 10"4 /xmole total l ipid; one vesicle = 1.25 x 1 0 1 2 /xmole total l ipid. The liposome/serum reaction volume was 100 ul. 1 0 0 Liposome Cone. (mM) 75 F I G U R E 13: Effect of Liposome Size on C l q Binding Isotherms. The equilibrium binding assay was used to measure C l q binding to extruded and non-extruded C L 2 0 liposomes in a purified system at p H 7.2 and 145 m M N a C l . M L V s (A) or 240 nm diameter liposomes (O) were reacted with varying amounts of C l q . The total l ipid concentration in the reaction was 0.5 m M . Since the amount of exposed l ipid was much less for M L V s compared with extruded liposomes, values for the amount of C l q bound to vesicles was expressed as pinole C l q bound per /xmole of exposed l ipid. For 240 nm extruded liposomes, the amount of exposed lipid was taken to be 35 % of the total l ipid (based on 3 1 P - N M R data: Devine, D . et al. 1994). For M L V s , the amount of exposed lipid was taken to be 5 % of the total l ipid (based on an estimate in Biochemistry of Lipids and Membranes) (Cullis , P. and Hope, M . 1985). 0.16 0.13 h 0.10 h 0.06 0.03 h 0.00 0.00 0.12 0.24 0.36 Equilibrium C1q Cone. (pM) 0.48 0.60 (E-1) 76 The ability of C L 2 0 liposomes to activate human complement was then measured for 100 and 240 nm liposomes (Figure 14). Approximately 20 % more complement activation was detected for 240 nm liposomes compared to 100 nm liposomes at each l ipid concentration tested below 0.6 m M . In this concentration range, there was a 75 fold greater amount of l ipid required for 100 nm particles to achieve the same level of complement activation (>20%) as the 240 nm liposomes. It should be noted that for both the extruded, 240 nm liposomes and for the M L V s , vesicle size was not altered as a consequence of protein binding under these conditions (pH 7.2 and 145 m M NaCl ) . M L V s ranged in size from 1-10 fxm in diameter as viewed in the microscope during particle electrophoresis. 77 FIGURE 14: Effect of Liposome Size on Complement Activation. The hemolytic assay was used to measure complement activation by 100 nm (O) and 240 nm ( • ) diameter C L 2 0 liposomes and by 240 nm P C : C H (55:45 mole %) liposomes (*) as a control. Each point represents the mean of 3 experiments done in duplicate. Error bars represent one S . E . M . The liposome concentration was the concentration of total l ipid and was determined as described in Materials and Methods. For 100 nm liposomes, the total exposed surface area and number of vesicles are related to the total l ipid concentration in the following ways: one cm 2 = 4.75 x 10"4 /imole total l ipid; one vesicle = 1.5 x 10"13 ptmole total l ip id . For 240 nm liposomes, one c m 2 = 6.9 x 10"4 umo\e total l ipid; one vesicle = 1.25 x 10"12 /xmole total l ipid. The liposome/serum reaction volume was 100 a\. Liposome Cone. (mM) 78 Chapter 4. Characterization of C l q Binding to Anionic Liposomes C l q binding to anionic liposomes was further characterized with respect to the effects of p H , ionic strength, l ipid concentration, and the presence or absence of human serum. Wi th respect to binding kinetics, time courses were carried out for C l q binding to C L 2 0 liposomes with purified protein in Tris or barbital buffers at either p H 4 or p H 7.2 and in diluted serum. Maximal binding occurred within one minute under all assay conditions. The C l q binding experiment was considered to provide equilibrium binding data since the amount of C l q bound per amount of l ipid was independent of the centrifugation separation step. Whether the liposomes were spun down at 166,300 x g for 30 min or at 99,800 x g for 45 min, the amount of C l q bound per amount of l ipid was the same. 4.1 p H Dependence of C l q Binding to Anionic Liposomes In five separate experiments using four different preparations of C l q and two different buffers (Tris and barbital), C l q binding to cardiolipin or phosphatidylglycerol-containing liposomes was found to be highly pH-dependent (Figure 15). C l q binding was optimal at p H 4 regardless of the buffer or anionic l ip id used. 79 FIGURE 15: pH-dependence of C l q Binding to Anionic Liposomes. C l q binding to C L or PG-containing liposomes was measured as a function of p H using the equilibrium C l q binding assay in a purified system at 145 m M N a C l . The reaction mixture consisted of 0.5 m M total l ipid and 0.06 uM (27 /xg/ml) C l q . Binding data for C L 2 0 liposomes in Tris and barbital buffers are represented by (O) and ( • ) respectively. Data for PG40 liposomes in barbital buffer is represented by (v). The figure is a composite of three experiments and is representative of a total of 5 experiments. The experiment showing the pH-dependence of C l q binding to C L 2 0 liposomes in Tris buffer was carried out an additional two times (data not shown). 0.80 80 4.2: Binding of Purified C l q to Anionic Liposomes at L o w p H To determine whether C l q binding was altered after low p H treatment, C l q / C L 2 0 liposome interactions were monitored at physiologic ionic strength (0.145 M NaCl ) by particle electrophoresis under cycled p H conditions. For a single reaction mixture of C l q and C L 2 0 M L V s , C l q binding was assessed at p H 5, 7, and then at p H 5 again (Figure 16). In the absence of C l q , vesicle mobilities did not change when the p H was adjusted from 5 to 7. In the presence of C l q , liposome mobilities varied enormously from p H 5 to p H 7. A t p H 5, C l q interacted strongly with C L 2 0 liposomes. When the p H was raised to 7, C l q binding was no longer detected in this system. A s the p H was lowered to 5, the initial level of C l q binding was observed. These results argue that the region of C l q involved in binding anionic liposomes was not altered as a result of low p H treatment. Since C l q binding to anionic liposomes was optimal at p H 4, binding assays were conducted at p H 4 to obtain isotherms under these favourable conditions. C l q binding isotherms for liposomes with 20 or 40 mole % C L and for liposomes with 40 or 55 mole % P G are shown in Figures 17 and 18. Even at p H 4, neutral E P C : C H liposomes failed to bind C l q . For anionic liposomes, the level of C l q binding was dependent on the liposome surface charge density. When the liposome C L content was raised from 20 mole % to 40 mole %, a 4 fold increase in C l q binding was observed. Increasing the liposome P G content from 40 to 55 mole % resulted in almost twice the amount of C l q bound at saturation. A summary of C l q binding levels at p H 4 is presented in Table 5. Comparing p H 4 results (Figure 17 and Table 5) and p H 7.2 results (Figure 10 and Table 4), C L 2 0 liposomes bound 100 x more C l q at p H 4 than at p H 7.2. C L 4 0 81 liposomes bound 12 x more C l q at p H 4 compared with binding at p H 7.2. F I G U R E 16: Assessment of C l q Integrity After L o w p H Treatment. Particle electrophoresis was used to measure the binding of C l q to C L 2 0 M L V s in 145 m M N a C l under cycled p H conditions. A s a control, C L 2 0 M L V mobilities were measured in 1.8 m M sodium barbital/3.1 m M barbituric acid buffer (O) at p H 5 and then at p H 7.0. C L 2 0 M L V s were incubated with 0.134 m M (61.5 Mg/ml) C l q ( • ) at p H 5 and vesicle mobilities were measured. The mixture was removed from the electrophoresis chamber, overall p H was adjusted to p H 7 by addition of 0.5 M N a O H , and vesicle mobilities were measured. Again , the mixture was removed from the electrophoresis chamber, overall p H was adjusted back to p H 5 by addition of 0.2 M HC1, and mobility was measured. Mobi l i ty measurements were made on ten vesicles in the mixture. The complete experiment was carried out twice; one data set is shown by the solid lines and the other by the dotted lines. Error bars represent one S D for the mobilities of ten vesicles. Where no error bars are shown, one S D was smaller than the symbol used for the mean. pH 5 pH 7 pH 5 82 FIGURE 17: C l q Binding to Cardiolipin-containing Liposomes at Low pH. The equilibrium C l q binding assay was used to obtain binding isotherms for 240 nm liposomes composed of P C : C H (O) , P C : C H : C L (35:45:20 mole % ) ( • ) , and P C : C H : C L (15:45:40 mole %) ( • ) . Purified protein at p H 4 was used with 145 m M N a C l . Since no l ipid dependence was observed, data for total l ipid concentrations from 0.5 m M to 5 m M in the reaction were combined. Q_ C o .Q O o E CO LU 0.20--®—i 0.00 0.05 0.10 0.15 0.20 Equ i l i b r i um C 1 q C o n e . ( L I M ) 0.25 83 FIGURE 18: C l q Binding to Phosphatidylglycerol-containing Liposomes at Low pH. The equilibrium C l q binding assay was used to obtain binding isotherms for 240 nm liposomes composed of P C : C H : P G (15:45:40 mole %) (v) and C H : P G (45:55 mole %) {*•). Purified protein at p H 4 was used with 145 m M N a C l . Since no l ipid dependence was observed, data for total l ip id concentrations from 0.5 m M to 5 m M in the reaction were combined. Q. O E ZL • D C 13 o -Q o o £ CO LU 0 . 5 0 0 .40 -1 0 . 3 0 0 . 2 0 -0 . 1 0 -A A A V 0 . 0 0 v v V 0 . 0 0 0 . 0 6 0 . 1 2 A A A V V V 0 . 1 8 0 . 2 4 0 . 3 0 Equil ibrium C1q Cone. ( IJM) 84 Table 5: Apparent Association Constants and Saturation Binding for C l q Binding to 240 nm Anionic Liposomes at pH 4 in a Purified Protein System. The estimate of saturation binding and K a values for C L 2 0 , PG40 and PG55 liposomes were derived from linear Scatchard plots. For C L 4 0 liposomes, a curvilinear Scatchard plot was obtained; a K a value was derived from the steep linear portion of the Scatchard plot and the C l q saturation value was derived from the isotherm. Scatchard plots are in Appendices 3 and 4. To convert //jnole total l ip id to number of vesicles, the surface area per vesicle and the number of vesicles per /xmole total l ipid were first estimated. The surface area per vesicle, calculated as S A v e s i c ] e = 47ir 2 , was 1.8 X l O " 9 cm 2/vesicle for our system. The number of 240 nm vesicles per /rniole total l ipid was estimated based on the following: assuming that 35% of the total l ip id is exposed and that the average area per l ipid is 0.7 nm 2 , the number of l ipid molecules per vesicle = (2.9 x 180,960 nm 2)/0.7 nm 2 = 750,000 molecules. The number of vesicles per itmole of l ipid = (6.022 X l O 1 7 molecules//xmole)/750,000 molecules = 8.0 x l O 1 1 vesicles/iimole l ipid. Saturation Binding Liposome Composition Scatchard Profile K a (M" 1) + ' S . E . M . umole C l q b o u n d / jitmole total l ipid # molecules C l q b o u n d / v e s i c l e E P C : C H (55:45) ni l ni l E P C : C H : C L (35:45:20) linear 6 ± 0.6 x l O 8 1.9 x l O " 4 143 E P C : C H : C L (15:45:40) curvi-linear 4.5 +0.7 x l O 8 8.0 x l O ' 4 602 E P C : C H : P G (15:45:40) linear 4.2 + 0.4 x l O 8 2.2 x l O " 4 166 C H : P G (45:55) linear 4.5 + 0.3 x l O 8 4.0 x l O 4 301 85 4.3: Effect of L i p i d Concentration on C l q Binding to C L 2 0 Liposomes at p H 7.2 in a Purified Protein System C l q binding to anionic liposomes at p H 7.2 was dependent upon l ipid concentration. Figure 19A shows two separate binding isotherms obtained when 0.5 m M and 2 m M total l ip id was provided in the reaction. C l q binding was inversely related to l ipid concentration such that C L 2 0 liposomes bound twice as much C l q at saturation when four times less surface area was provided. A summary of this binding data w i l l be provided later in this chapter (Table 8A, page 98). We explored the possibility that this surface area dependence was due to the existence of different affinity species of C l q at p H 7.2. To test this hypothesis, the amount of C l q bound was plotted as a function of the amount of free C l q per unit of surface area in solution (Figure 19B). The transformed data collapsed onto one curve suggesting, as addressed later in the discussion, that C l q may exist not only in monomeric form, but also as some higher affinity species in the purified protein system at p H 7.2. 86 F I G U R E 19: Effect of L i p i d Concentration on C l q / C L 2 0 Binding Isotherms at p H 7.2. The equilibrium C l q binding assay was used to measure C l q binding to 240 nm C L 2 0 liposomes in a purified protein system at p H 7.2 (Panel A ) . The total l ipid concentration in the reaction was either 0.5 m M ( • ) or 2 m M (A). Binding data was then plotted as a function of equilibrium C l q concentration per unit surface area ( L x V / S A , where L is the equilibrium C l q concentration, V is the supernatant volume, and S A is the total surface area of liposomes in the reaction in cm 2) (Panel B) . The S A was calculated as the total pmoles of l ip id in the sample multiplied by the number of vesicles per /xmole of l ipid and by the surface area per vesicle. The number of 240 nm vesicles per /miole total l ipid was estimated to be 8.0 x 10 1 1 vesicles//imole l ipid. The surface area per vesicle was calculated as S A v e s i d e = 47tr2 where r = 1.2 x 10' 5 cm and was 1.8 x 10"9 cm 2/vesicle for our system. 87 A o E 0.25 0.20 0.15H •o _ c LO o5 B J=I 0.10H O E 0.05-fA A ft 0.00 — i 1 1 1 1 r — i 1 1 1 1 1 r 0.00 0.10 0.20 0.30 0.40 0.50 (E-1) Equil ibrium C1q C o n e . (uM) B T3 Q . "o E ZL -a c o O "o E L O L U 0.00 0.00 0.08 ' ' 0.16 0.24 ' ' 0.32 ""0.40 (LxV) /SA (umol/cm 2 ) ( E " ? ) 88 4.4: Effect of Ionic Strength on Clq/ l iposome Binding C l q / C L 2 0 liposome interactions were monitored at p H 5.0 and p H 7.0 by particle electrophoresis under different ionic strength conditions (Figure 20). The mobilities of M L V s alone decreased as the ionic strength went from 20 m M N a C l up to 145 m M as expected. Liposome mobilities were identical whether measurements were made at p H 5 or p H 7. When C l q was reacted with C L 2 0 M L V s , M L V mobilities shifted. The magnitude of the liposome mobility shift after incubation with C l q was inversely related to ionic strength indicating that C L 2 0 M L V s bound more C l q at lower ionic strength. This trend was observed whether the p H of the system was 5 or 7 but was more pronounced at p H 7. When C l q was added at p H 7, liposome mobilities decreased more than 5 mobility units at 20 m M N a C l and did not change at all at 145 m M N a C l . C L 2 0 M L V s bound more C l q at p H 5 than at p H 7 under every ionic strength condition tested. While liposome mobilities were being measured, the aggregation state of the liposomes was monitored. Liposome aggregation was observed under conditions where C l q binding to anionic liposomes was enhanced, at low ionic strength and at low p H (Table 6). 89 F I G U R E 20: Ionic Strength Dependence of C l q Binding to C L 2 0 M L V s . Particle electrophoresis was carried out on C L 2 0 M L V s alone (open symbols) or in the presence of C l q (closed symbols) under different N a C l concentration conditions at p H 5 (A) and 7 (O) . Where liposomes were preincubated with C l q , the total l ip id in the reaction was 0.36 m M and the C l q concentration was 0.134 /xM (61.5 jug/ml). Prior to loading the reaction mixture into the electrophoresis chamber, buffer was added to bring the total volume from 0.78 ml to 2.8 m l . Mobi l i ty measurements were made on ten vesicles in the mixture each time. The complete experiment was carried out twice; one data set is shown by the solid lines and the other by the dotted lines. Error bars represent one S D for the mobilities of the ten vesicles. Where no error bars are shown, one S D was smaller than the symbol used for the mean. 50 100 150 NaCl Cone. (mM) 90 T A B L E 6: Effect of Ionic Strength and pH on Liposome Aggregation in the Presence of C l q . M L V s were examined for aggregation under 320 times magnification during particle electrophoresis. M L V s (0.36 m M total lipid) were incubated with C l q at 0.134 uM (61.5 jug/ml) for 40 min at R T . Prior to loading the sample into the electrophoresis chamber, the reaction mixture was diluted with buffer to bring the volume from 0.78 m l to 2.8 ml . Liposome Composition N a C l Cone. (mM) p H Aggregation E P C : C H (55:45) 20 5 + E P C : C H : C L (35:45:20) 20 5 + + + + + 20 7 + + + ++ 100 5 + + + + 100 7 + + + 145 5 + + + 145 7 no aggregation 91 The dependence of Clq/ l iposome binding on ionic strength was further demonstrated using the equilibrium C l q binding assay on both extruded liposomes (average diameter 240 nm) and on M L V s (Figure 21). Neutral E P C : C H extruded and M L V liposomes failed to bind C l q even at low ionic strength (in 20 m M NaCl ) . A s suggested by the electrophoresis results, the amount of C l q binding to anionic liposomes was inversely related to the ionic strength of the environment. Binding isotherms for extruded liposomes (Panel A ) and for C L 2 0 M L V s (Panel B) illustrated this inverse relationship. For both extruded vesicles and M L V s , small, but non-zero levels of C l q binding were measured at 145 m M N a C l . A t 100 m M N a C l , extruded liposomes bound 80 times more C l q and M L V s bound 14 times more C l q at saturation than at 145 m M . A t 20 m M N a C l , virtually al l the C l q added bound to C L 2 0 vesicles such that saturation was not achieved at the concentrations tested. A t a C l q equilibrium concentration of 0.002 fjM, 275 times more C l q bound to extruded, 240 nm liposomes at 20 m M N a C l than at 145 m M N a C l . For C L 2 0 M L V s at 0.02 j i M equilibrium C l q concentration, 75 times more C l q bound at 20 m M N a C l than at 145 m M N a C l . Saturation binding values and apparent association constants describing the binding of C l q to 240 nm C L 2 0 liposomes and to C L 2 0 M L V s in 100 m M N a C l are presented in Table 7. The evaluation of C l q binding to anionic liposomes presented thus far has relied on data collected with purified protein. To determine whether these results reflect the behaviour of C l q under more physiologic conditions, C l q binding to anionic liposomes was measured in diluted human serum. 92 F I G U R E 21: C l q / C L 2 0 Binding Isotherms: Effect of Ionic Strength. The equilibrium C l q binding assay was used to measure C l q binding to extruded C L 2 0 liposomes ( ~ 240 nm diameter) (Panel A ) and to C L 2 0 M L V s (Panel B) at 20 (A), 100 (v), and 145 m M ( • ) N a C l . E P C : C H (55:45 mole%) extruded liposomes and M L V s were also assessed for C l q binding at 20 m M N a C l (O) . Reaction mixtures contained varied amounts of C l q and a fixed amount of total l ipid (0.5 m M ) . Because the lamellarity differed between extruded and non-extruded vesicles, the amount of C l q bound to vesicles was expressed as tunole C l q bound per /xmole of exposed l ipid. For the 240 nm diameter extruded liposomes, the amount of exposed l ip id was taken to be 35 % of the total l ip id (based on 3 1 P - N M R data in Devine, D . et al. 1994). For M L V s , the amount of exposed l ipid was taken to be 5 % of the total l ipid (Cull is , P . and Hope, M . 1985). 93 A ~ 1.00 Equil ibrium C1q Cone, (urn) B 3. ' 0.00 0.04 0.0B 0.12 0.16 0.20 Equil ibrium C1q Cone. (uM) 94 Table 7: Apparent Association Constants and Saturation Binding for C l q Binding to CL20 Liposomes at pH 7.2 and at 100 mM NaCl in a Purified Protein System. The estimate of saturation binding and K a values for C L 2 0 M L V s was derived from a linear Scatchard plot. For 240 nm C L 2 0 liposomes, a curvilinear Scatchard plot was obtained; a K a value was derived from the linear portion of the Scatchard plot and the C l q saturation value was derived from the isotherm. The Scatchard plots are presented in Appendix 5. Saturation Binding Liposome Size Scatchard profile K a (M" 1) ± S . E . M . /zmole C l q b o u n d / jumole total l ipid /miole C l q b o u n d / /unole exposed lipid 240 nm curvi-linear 2.2 ± 0.5 X l O 9 1.5 x 10"4 4.4 x l O " 4 M L V linear 1.12 + 0.4 x l O 8 1.6 x 10"4 3.2 x l O " 3 4.5 Measurement of Clq/ l iposome Binding in Human Serum The C l q equilibrium binding assay was altered as described in Materials and Methods in order to maintain isotonic conditions in the presence of serum. Human C l q -depleted serum was used as the human serum source and 1 2 5 I - C l q was added at known concentrations. Buffers were altered to balance densities such that > 90 % of liposomes and < 10 % of free C l q were pelleted upon centrifugation. Ideal experimental conditions were first secured for the system in which human Clq-depleted serum was diluted to 1/60 in the overall reaction. C l q binding isotherms for C L 2 0 liposomes in serum diluted 1/60 are presented in Figure 22 (panel A ) . E P C : C H liposomes failed to bind C l q . For C L 2 0 liposomes, the 95 level of C l q binding in this serum system was the same as the binding level in the purified protein system at similar l ipid concentrations. Different isotherms were obtained at different l ipid concentrations (0.6, 2, and 6 m M ) indicating once again that more than one C l q species is present. C l q binding was inversely related to the l ipid concentration such that C L 2 0 liposomes bound approximately ten times more C l q at saturation when ten times less surface area was provided. Plotting the amount of C l q bound as a function of free C l q per unit area in solution did not collapse the three isotherms to one curve (panel B) . Values for the saturation binding and association constants describing the binding of C l q to C L 2 0 liposomes in human serum (1/60) are shown in Table 8B. The effect of serum dilution on C l q binding to C L 2 0 liposomes was then evaluated. C l q binding to C L 2 0 liposomes was measured in Clq-depleted serum at overall dilutions of 1/60, 1/20, 1/10, and 1/4 (Figure 23). For serum dilutions from 1/60-1/10, C l q binding isotherms were superimposable. For serum diluted 1/4, C l q binding was similar in magnitude but slightly less than that measured under more dilute serum conditions. The fact that the amount of C l q spinning down in the absence of liposomes was approximately 30 % for the 1/4 serum system indicated that the assay limitation had been met and that data in this system should be considered as qualitative rather than quantitative. A t accessible concentrations, the level of C l q binding was independent of the serum dilution. These binding levels were also similar to values in the purified protein system at p H 7.2, suggesting that binding measurements made in the purified protein system were indicative of behaviour in human serum. 96 FIGURE 22: Effect of Lipid Concentration on C l q Binding to CL20 Liposomes in Serum (diluted 1/60). The equilibrium C l q binding assay was used to measure C l q binding to 240 nm C L 2 0 liposomes in human Clq-depleted serum at 1/60 in the reaction. C l q binding isotherms for C L 2 0 liposomes at l ipid concentrations of 0.6 ( • ) , 2 (A), and 6 m M (O) are shown in panel A. Binding data was then plotted as a function of equilibrium C l q concentration per unit of surface area ( L x V/SA, where L is the equilibrium C l q concentration, V is the supernatant volume, and SA is the total surface area of liposomes in the reaction) (panel B) . Q_ "5 E ZL T J C Zl o .Q O O E 0.25 0.20 ~ 0.15' 0.10 0.05-c § o v p © o O o o 0.00 1 1 r 0.00 0.03 0.06 0.08 0.11 0.14 (E-1) Equilibrium C1q Cone. (uM) B Q. "o E ZL T J C =1 o O o E ZL 0.25 0.00 l 1 1 r 0.00 0.10 0.20 0.30 0.40 0.50 (LxV)/SA (umol/cm ) ( E " 8 ) 97 Table 8A: Effect of Lipid Concentration on Apparent Association Constants and Saturation Binding for C l q Binding to CL20 Liposomes in a Pure System at pH 7.2. Curvilinear Scatchard plots were obtained at both l ipid concentrations. Values of the apparent K a were derived from the linear portion of the Scatchard plots and C l q saturation values were derived from the isotherms. The Scatchard plots are presented in appendix 1. Saturation Binding L i p i d Concentration in reaction (mM) Scatchard Profile K a + S . E . M . (M- 1) /xmole C l q b o u n d / jumole total l ipid # molecules C l q b o u n d / v e s i c l e 0.5 curvi-linear 2 .0 + 0.16 X l O 9 2 x l O " 6 3/2 2.0 curvi-linear 1.7 + 0.16 X l O 9 1 x l O " 6 3/4 Table 8B: Effect of Lipid Concentration on Apparent Association Constants and Saturation Binding for C l q Binding to CL20 Liposomes in Human Serum Diluted 1/60. Saturation binding and K a value estimates were derived from linear Scatchard plots (appendix 6). Saturation Binding L i p i d Concentration in reaction (mM) Scatchard profile K a + S . E . M . (M- 1) /imole C l q bound/ ^mole total l ip id # molecules c lqbound/vesicle 0.6 linear 1.8 + 0.12 x l O 9 2.3 x l O " 6 2/1 2 linear 1.1 ± 0.9 x l O 9 0.8 x l O " 6 2/3 6 linear 4.7 ± 0.54 x l O 8 0.4 x l O " 6 1/3 98 FIGURE 23: Effect of Serum Dilution on Clq Binding to CL20 Liposomes. The equilibrium C l q binding assay was used to measure C lq binding to 240 nm CL20 liposomes in human Clq-depleted serum at various concentrations. Incubation mixtures contained G V B 2 + buffer, 0.5 mM liposomes and a mixture of C l q and 1 2 5 I-Clq with Clq-depleted serum at a 1/60 (•), 1/20 (O), 1/10 (+), or 1/4 (*) dilution in the overall reaction. T 3 Q_ O E ZL T 3 C O . Q CT O O E 0.25 0.20h + 0.15H LO LU o.ioH 0.05 0.00 8 • -e-o o • • T I I i | I i i I | 1 1 1 1 1 1 1 1 r 0.00 0.25 0.50 0.75 1.00 (E-2) Equilibrium C1q C o n e . ( L I M ) 99 Chapter 5: Evaluation of the Role of a Cationic Region of Clq in Liposome Interactions with Complement Amongst the immunoglobulin-independent activators of complement, a group of activators which includes C R P , D N A , S A P , and beta-amyloid has been found to bind C l q at a highly cationic region of the complement protein. This site is composed of amino acid residues 14-26 on the C l q A polypeptide chain and is near the amino terminus on the collagenous portion of C l q (see Figure 24). Five of the thirteen residues are positively charged, making this region of C l q the most cationic part of the protein. Anionic liposomes share a crucial property with the above mentioned group of activators which bind C l q directly: they all have repeating negative surface charges. Because of this similarity, we investigated whether the cationic region of C l q was involved in the binding of C l q to anionic liposomes. Six different peptides were tested for their ability to inhibit C l q binding, to inhibit complement activation, and to interact directly with anionic liposomes. In addition to the authentic C l q A H . 2 6 peptide and the two C l q A control peptides (the scrambled sequence peptide and the conformational control peptide which has had proline residues substituted by alanine residues), two additional C l q A - l i k e peptides were synthesized and tested. The first, C l q A ( 0 + ) , had all five cationic residues replaced with glycine residues. The second, C l q A ( 2 + ) , had three of the cationic residues replaced by glycine residues. The sixth peptide was not related to C l q A ; it consisted of residues 400-411 of the N-terminal sequence of fibrinogen and contained one arginine and two histidine residues. For a summary of the properties of these peptides, see Table 9. 100 Figure 24: M o d e l of C l q This model of C l q denotes the region of C l q which has been investigated in this study for its involvement i n Clq /anionic liposome interaction. C l q consists of three distinct polypeptide chains, A , B , and C which together form the collagen-like triple helix region and the adjoining region of globular head domains. While the triple helixes of the collagen-like region contain residues from the N terminus up to about residue 89, the remaining —131 residues of each chain fold to form the globular head domains. IgG and I g M bind C l q via the globular portion. The C l q receptor binds at the hinge region as shown. Antibody-independent activators of complement including C R P , S A P , D N A , and beta amyloid fibres bind to the collagen-like portion of C l q near the N-terminus. 4.5nm 101 Table 9: C l q A Test Peptides Peptide length (#aa's) cationic residues hydrophobic residues Prolines authentic C l q A ( 1 4 . 2 6 ) 13 5 (4R, I K ) 4 2 C I Unscrambled) 13 5 (4R, I K ) 4 2 C l q A ( P . A ) 13 5 (4R, I K ) 4 0 C l q A ( 0 + ) 13 0 4 2 C l q A ( 2 + ) 13 2 (2R) 4 2 Fgn 13 3 ( I K , 2H) 4 0 5.1 Inhibition of C l q Binding to C L 2 0 Liposomes by C l q A Peptides C l q A , C l q A control, and unrelated peptides were examined for their ability to compete with whole C l q in the C l q equilibrium binding assay (Figure 25). A t a dose of 220 uM, the C l q A 1 4 . 2 6 authentic peptide completely blocked C L 2 0 / C l q binding. When 45 jitM of authentic C l q A 1 4 . 2 6 peptide was added, 70 % of C l q binding was inhibited. C l q A control peptides, C l q A ( s c r a m b l e d ) a n d C l q A ( P . A ) , were equally successful at inhibiting C l q binding to C L 2 0 liposomes. These results indicated that a conformational or sequence specificity in the peptide was not required. The charge control peptides, C l q A ( 0 + ) a n d C l q A ( + 2 ) , were incapable of inhibiting C l q / C L 2 0 binding indicating that more than two cationic residues were required to interact with anionic liposomes. Wi th respect to the unrelated fibrinogen peptide, an unexpected ability to inhibit the C l q / C L 2 0 interaction was observed. 70% of C l q binding was inhibited by adding 240 uM of the fibrinogen peptide. 102 F I G U R E 25: Inhibition of C l q Binding to C L 2 0 Liposomes by C l q A Peptides. The C l q equilibrium binding assay was used to measure C l q binding to C L 2 0 liposomes in the presence of peptides at p H 7.2. The reaction mixture consisted of 0.5 m M total l ipid, 0.009 pM C l q , and peptides at different concentrations. The % inhibition of C l q binding is shown for the authentic C l q A ( 1 4 _ 2 6 ) 0 ) , C l q A ( s c r a m b ] e d ) (O) , C l q A ( P . A ) (V) , C l q A ( 0 + ) ( • ) , C l q A ( 2 + ) (A), and the unrelated Fgn (0) peptides. Data points show the mean of three experiments. Error bars represent one S D . Where no error bars are shown, one S D was smaller than the symbol used for the mean. 110 Peptide Cone. (uM) 103 While C l q A 1 4 . 2 6 , C l q A ( s c r a m b , e d ) and C l q A ( P _ A ) peptides were capable of totally inhibiting Clq/ l iposome binding, relatively large amounts of these peptides were required for this effect. Mola r ratios of peptide to C l q in these binding inhibition experiments are presented in Table 10. 5.2 Inhibition of Liposome Complement Activation by C l q A Peptides A modified hemolytic assay was used to measure the ability of C l q A and control peptides to inhibit complement activation by liposomes in human serum (Figure 26). C l q A ( 1 4 . 2 6 ) completely blocked complement activation by C L 2 0 liposomes at a concentration of 3 m M in our experimental system. Forty percent inhibition was achieved at approximately 1.7 m M peptide concentration. Control C l q A peptides (scrambled, and P - A peptides) inhibited complement activation to the same extent as the authentic peptide verifying that the interaction lacks sequence and conformational specificity. Charge control peptides, C l q A ( 0 + ) and C l q A ( 2 + ) , failed to inhibit complement activation by C L 2 0 liposomes confirming that more than two positively charged residues are required for the interaction of C l q with anionic liposomes. The unrelated fibrinogen peptide which had shown an ability to inhibit Clq/ l iposome binding, failed to inhibit complement activation. Assay controls showed that peptides incubated with E A cells alone did not cause cell lysis and peptides incubated with N H S in the absence of liposomes did not activate complement. 104 FIGURE 26: Inhibition of Liposome Induced Complement Consumption by C l q A Peptides. A modified hemolytic assay was used to measure the amount of complement consumed (activated) by E P C : C H : C L (35:45:20 mole %) liposomes (5mM) in human serum (diluted 1:4 with G V B 2 + ) after an initial incubation of liposomes and peptides. The % inhibition of complement activation is shown for C l q A 0 4 . 2 6 ) (*), C l q A ( s c r a m b ] e d ) (O) , C l q A ( P . A ) (v), C l q A ( 0 + ) ( • ) , C l q A ( 2 + ) (A), and the unrelated Fgn (0) peptides. Data points show the mean of three experiments. Error bars represent one S D . Where no error bars are shown, one S D was smaller than the symbol used for the mean. 0 1 2 3 4 5 Peptide Cone . (mM) 105 A s described for the inhibition of C l q binding by C l q A 1 4 _ 2 6 , C l q A ( s c r a m b i e d ) and C l q A ( P . A ) peptides, large amounts of these peptides were required to inhibit complement activation by anionic liposomes. In fact, far more peptides were required to inhibit complement activation than were required to inhibit C l q binding. Mola r ratios of peptide to C l q for these complement activation experiments are presented in Table 10. T A B L E 10: The Amount of C l q A Peptide Required to Inhibit C l q Binding and Complement Activation by Anionic Liposomes. Molar Ratio of Pept ide/Clq Required Inhibition Assay For 50 % Inhibition For 90 % Inhibition C l q Binding 3,450 18,390 Complement Activation 40,700 61,600 106 5.3 Direct Interaction of Cationic Peptides with Anionic Liposomes The direct interaction between anionic liposomes and C l q peptides was assessed by particle electrophoresis at p H 7 under interaction-promoting low ionic strength conditions (Figure 27). Authentic C l q A ( 1 4 . 2 6 ) , C l q A ( s c r a m b | e d ) , and C l q A ( P . A ) peptides all interacted strongly with C L 2 0 M L V s . In contrast, no interaction was detected with the C l q A ( 0 + ) peptide. While some pept ide /MLV binding was observed at the highest peptide concentration (900 /ig/ml) with C l q A ( 2 + ) peptide, the level of interaction was slightly higher for the fibrinogen peptide. The magnitude of these direct interactions between C l q A peptides and C L 2 0 M L V s mirrored the inhibitory capacities of the peptides. To further characterize peptide/CL20 M L V interactions, the effect of ionic strength was investigated using particle electrophoresis. Figure 28 shows liposome interactions with C l q A ( 1 4 . 2 6 ) and control peptides at 20 and 100 m M N a C l . C l q A ( 1 4 . 2 6 ) and the sequence and conformational control peptides, C l q A ( s c r a m b l e d ) and C l q A ( P . A ) , exhibited a high level of liposome binding with more binding at lower ionic strength. While C l q A ( 0 + ) and C l q A ( 2 + ) peptides did not bind to C L 2 0 M L V s at the concentration tested (90 tig/ml), the Fgn peptide exhibited a slight binding capacity but only at 20 m M N a C l . 107 F I G U R E 27: Interaction of C l q A Peptides with C L 2 0 Liposomes. Particle electrophoresis was used to monitor the direct interaction of C l q A and control peptides with C L 2 0 M L V s at p H 7 and 20 m M N a C l . Peptides at 9, 45, 90, or 900 /xg/ml were incubated with M L V s (0.36 m M total l ipid in reaction) for 40 minutes at R T . Liposome mobilities in the presence of increasing amounts of C l q A 1 4 . 2 6 ( A ) , C l q A { s c r a m b ] e d ) (O) , C l q A ( P . A ) (v), C l q A ( 0 + ) ( • ) , C l q A ( 2 + ) (A), and unrelated Fgn (0) peptides are presented. Mobi l i ty measurements were made on ten vesicles in the mixture each time. The complete experiment was carried out three times. Error bars represent one S D . Where no error bars are shown, one S D was smaller than the symbol used for the mean. 0 200 400 600 800 1000 1200 Peptide Cone. (uM) 108 F I G U R E 28: Ionic Strength Dependence of Peptide/CL20 Binding at p H 7. Particle electrophoresis was used to measure C L 2 0 M L V interactions with C l q A , C l q A control, and unrelated control peptides at p H 7 and ionic strengths of either 20 or 100 m M N a C l . C L 2 0 M L V mobilities alone ( • ) and in the presence of 90 tig/ml of C l q A 1 4 . 2 6 0 ) , C l q A ( s c r a m b l e d ) (O) , C l q A ( P . A ) ( v ) , C l q A ( 0 + ) ( • ) , C l q A ( 2 + ) (A), and unrelated Fgn (0) peptides are presented. Mobi l i ty measurements were made on ten vesicles in the mixture each time. The complete experiment was carried out three times. Error bars represent one S D . Where no error bars are shown, one S D was smaller than the symbol used for the mean. 20 40 60 80 N a C l Cone . (mM) 100 120 109 Chapter 6 : Modulation of Clq-Mediated Complement Interactions with Anionic Liposomes by Incorporation of PEG-Lipids We investigated the capacity of incorporated PEG-l ipids to inhibit Clq-mediated complement activation by anionic liposomes. The derivatives used in this study were C H -P E G 6 0 0 , C H - P E G ) 0 0 0 and P E - P E G 2 0 0 0 . While P E - P E G has previously been evaluated in terms of its incorporation and maintenance in the liposome membrane in the presence of serum (Parr, M . et al. 1994), C H - P E G derivatives had not been equivalently assessed. To obtain quantitative measurements of the percent incorporation and of the amount of C H -P E G maintained within liposomes in the presence of human serum, we synthesized a radiolabelled C H - P E G _ ] 0 0 0 and used the assay system developed by Parr et al. (1994) to assess the maintenance of P E - P E G in mouse serum. 6 T Evaluation of C H - P E G - N H - ^ C O C H , Purity Prior to using the labelled C H - P E G - N H - 1 4 C O C H 3 as a tool to assess C H - P E G derivative incorporation and maintenance in liposome membranes, the purity was evaluated. ' H - N M R carried out on C H - P E G - N H 2 showed only the series of proton resonance peaks for P E G and for cholesterol. Samples of C H - P E G - N H - 1 4 C O C H 3 were run on silica and reverse phase C - l 8 T L C s with iodine, molybdenum, and fluorescamine staining. A single spot was detected and no free cholesterol or free P E G was detected. Fluorescamine staining and visualization under U V failed to illuminate any spots, indicating that no free amine groups ( C H - P E G - N H 2 ) were present. 110 To ensure that introduction of the amino acetate unit did not alter the C H - P E G , the ability of C H - P E G 1 4 0 0 - N H - C O C H 3 to inhibit complement activation by anionic liposomes was tested side by side with C H - P E G 1 0 0 0 - O H . The slightly higher complement inhibitory capacity of C H - P E G 1 4 o o - N H - C O C H 3 compared to C H - P E G 1 0 0 0 - O H was consistent with the longer P E G chain producing enhanced exclusion of complement. These results indicated that the synthesized C H - P E G 1 4 0 0 - N H - C O C H 3 was pure and appropriate to use for evaluation of C H - P E G derivative stability. 6.2 Evaluation of C H - P E G Incorporation into Liposomes The incorporation of C H - P E G derivatives into liposomes was first assessed by particle electrophoresis (Figure 29). Due to an increase in drag force associated with the presence of P E G extending from the liposome surface, liposomes with P E G had slower mobilities than plain C L 2 0 M L V s . Comparing different P E G molecular weights, a direct relationship existed between the length of the P E G and the associated decrease in liposome mobility. For a given P E G molecular weight, liposome mobilities decreased with increasing mole % of C H - P E G indicating that indeed as more C H - P E G was added to the l ipid mixture, more C H - P E G was incorporated into liposomes. I l l FIGURE 29: Assessment of PEG-lipid Incorporation into Liposomes by Particle Electrophoresis. Particle electrophoresis was carried out on M L V s containing E P C : C H : C L : C H - P E G 6 0 0 (A) and E P C : C H : C L : C H - P E G 1 0 0 0 (O) at (35:45-n:20:n) mole % where n = 0, 5, 10, and 15 mole % and E P C : C H . C L : P E - P E G 2 0 0 0 ( • ) at (30:45:20:5) mole %. Liposomes were made in 1 m M M O P S with 0.5 m M E D T A and 100 m M N a C l . Mobil i ty measurements were made on ten vesicles in the mixture each time. Error bars representing one standard deviation are plotted but covered by data point symbols. o CD W - -2 o > "E o E -3 E -4 o • O O O A A -A 3 i I 10 mole % PEG-lipid 15 20 112 In order to quantitatively assess the amount of C H - P E G in liposomes, radiolabeled CH-PEG-containing liposomes were prepared and the proportion of free C H - P E G was determined. The separation of free C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 from liposome-associated C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 was achieved by Sepharose chromatography. Figure 30, panel A shows the separation of C L 2 0 liposomes prepared with C H - P E G 1 4 0 0 at 30 mole % of the total l ipid from unincorporated C H - P E G 1 4 0 0 in buffer. While liposome tracer counts showed one sharp peak, C H - P E G counts showed a second peak after the liposome-associated peak indicating that free C H - P E G was present. Comparing the average 1 4 C - P E G / 3H-liposome ratio from liposome peak fractions 12-14 with the initial 1 4 C - P E G / 3H-liposome ratio provided a measure of the amount of C H -P E G i 4 0 0 - N H - 1 4 C O C H 3 incorporated into the liposomes. Table 11 shows the amount of C H - P E G ] 4 0 0 actually incorporated into E P C : C H : C L : C H - P E G (35:45-n:20:n) liposomes as a function of the amount added in the lipid mix. The vast majority of the C H - P E G 1 4 0 0 added to the l ipid mixture was incorporated into the liposomes. In fact, up to 19.5 mole % of C H - P E G 1 4 0 0 was incorporated when 30 mole % was added to the l ip id mix. For experiments testing the inhibitory effect of C H - P E G on complement activation, a maximum of 15 mole % C H -P E G was used, most of which is in the membrane. In an additional experiment, C L 2 0 liposomes made without any C H - P E G were incubated with free C H - P E G in solution (an amount equivalent to 5 mole % of the total l ipid amount for the liposomes) prior to chromatography. 3 H-liposome and 1 4 C - P E G counts came out in completely overlapping profiles indicating that virtually all of the 5 113 mole % CH-PEGi4o 0 was incorporated into the liposomes (data not shown). Panel B of Figure 30 shows the separation of free and liposome-associated C H -P E G in the presence of human serum. Liposomes were made up to contain E P C : C H : C L : C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 at 35:30:20:15 mole % and were incubated with neat human serum or buffer as a control. In buffer, the liposome and C H - P E G profiles were the same shape and no secondary C H - P E G peak was evident. In the presence of human serum, there was a single liposome peak and two C H - P E G peaks. The second C H - P E G peak from fractions 18 to 28 indicated that a portion of the C H - P E G had become dissociated from the liposomes in serum. Table 11 : Assessment of CH-PEG_ 1 0 0 0 Incorporation into Liposomes mole % C H - P E G Added to L i p i d M i x % of Initial 1 4 C / 3 H Ratio mole % Incorporated 5 96.6 +/- 4.7 4.8 7.5 93.1 +/- 1.0 7.0 10 92.6 +/- 0.6 9.3 15 90.9 +/- 2.4 13.6 30 64.9 +/- 5.2 19.5 114 FIGURE 30: Chromatographic Separation of Free C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 from Liposome-associated C H - P E G i 4 0 0 - N H - 1 4 C O C H 3 . Free and liposome-associated C H -P E G 1 4 0 0 - N H - 1 4 C O C H 3 were separated on 3 ml columns of Sepharose C L 2 B . Panel A shows chromatographic profiles for 100 nm liposomes made using the l ipid mixture E P C : C H : C L : C H - P E G 1 4 0 0 at 35:15:20:30 mole % with trace 3 H - D P P C as the liposome marker and run in buffer. Panel B shows chromatographic profiles for 100 nm liposomes made using the l ip id mixture E P C : C H : C L : C H - P E G 1 4 0 0 at 35:30:20:15 mole % with trace 3 H - D P P C and incubated in neat human serum or in buffer at 37°C for 30 minutes. For the buffer system, 3 H-liposome counts and C H - P E G - 1 4 C counts are indicated by (v) and ( • ) respectively. For the serum system, 3H-liposome counts and C H - P E G - 1 4 C counts are indicated by (A) and (O) respectively. 115 A 6.3 Assessment of Loss of C H - P E G from Liposomes in Human Serum Figure 31 shows the effect of prolonged incubation of C L 2 0 liposomes bearing approximately 15 mole % C H - P E G 1 4 0 0 with human serum at 37°C on the amount of C H -P E G maintained in the liposomes. Immediately upon introduction of human serum to liposomes, the ratio of P E G to l ipid marker decreased by about 35 %, indicating that C H -P E G had dissociated from the liposomes. After this initial loss, the C H - P E G level in the liposomes was maintained for at least 24 hours. Controls in the absence of serum showed no loss of C H - P E G from the membrane until the 24 hour time point. A t 24 hours, 30 % of C H - P E G was lost from the membrane presumably due to oxidation of lipids at 37°C in buffer. While loss of C H - P E G 1 4 0 0 from liposomes was also observed for liposomes that were made with 5 and 10 mole % C H - P E G 1 4 0 0 , less C H - P E G exchanged out of these liposomes. These results are presented in Table 12. Once again, the loss of C H - P E G was immediate and following the initial loss, the C H - P E G level in the liposomes was maintained for at least 24 hours. 117 F I G U R E 31: Analysis of C H - P E G 1 4 0 0 - N H - 1 4 C O C H 3 Exchange i n Human Serum. Liposomes composed of E P C . C H : C L : C H - P E G 1 4 0 0 - N H - , 4 C O C H 3 at 35:30:20:15 mole % with trace 3 H - D P P C were incubated either with neat human serum (A) or with H E P E S buffer (10 m M H E P E S with 145 m M N a C l , p H 7.2) (O) as a control at 37°C. A t various time points, samples were removed and were run down a 3 ml Sepharose C L 2 B column. Each point represents the mean of three experiments. Error bars show one S D . Where no error bars are shown, one S D was smaller than the symbol used for the mean. 1 1 0 100C3GO--Q Q. Q , X 90 -o o 20 -1 0 -o I i - ' 1 1 1 0 4 8 1 2 16 2 0 2 4 Time (hours) 118 Table 12 : Analysis of C H - P E G _ 1 4 0 0 Exchange in Human Serum: Effect of the Initial Amount of C H - P E G _ 1 4 0 0 in the Liposomes % of Initial 1 4 C / 3 H Ratio mole % C H - P E G Added to L i p i d M i x 0 hour 8 hour 24 hour 5 89.6 81.1 84.2 10 82.0 75.4 80.4 15 65.7 69.1 72.9 Having determined that C H - P E G derivatives were incorporated and maintained in liposomes in the presence of human serum to a satisfactory degree, the effect of C H - P E G on C l q mediated complement activation by anionic liposomes was assessed and compared to that of P E - P E G . 6.4 Effect of P E G - l i p i d Incorporation on C l q Binding to 100 nm Liposomes and Complement Activation The ability of incorporated PEG-l ipids to inhibit C l q binding to 100 nm anionic liposomes as determined by the E L I S A assay is illustrated in Figure 32. Both C l q binding to anionic liposomes and the inhibition of this binding were l ipid concentration-dependent. For each P E G - l i p i d derivative, increasing P E G surface density resulted in an increased inhibitory effect on C l q binding. For instance, while no significant difference was observed between C l q binding to C L 2 0 liposomes and C L 2 0 liposomes with C H - P E G 6 0 0 119 up to 7.5 mole % (panel A ) , at 10 mole % C H - P E G 6 0 0 , approximately 20 % of the binding is blocked. A t 15 mole %, 50-100 % of C l q binding was inhibited, depending on the l ip id concentration. P E G chain length was also important for determining the degree of inhibition. C l q binding to liposomes was inversely related to the P E G chain length. While incorporation of 15 mole % C H - P E G 6 0 0 (panel A ) resulted in a 50-100 % reduction of C l q binding, 15 mole % C H - P E G 1 0 0 0 incorporation (panel B) resulted in virtually total inhibition at all l ip id concentrations. Ten mole % P E - P E G 2 0 0 0 (panel C) resulted in 100% inhibition of C l q binding at all l ipid concentrations. The effect of P E G incorporation on the complete activation of complement by 100 nm diameter anionic vesicles is shown in Figure 33. A s observed for C l q binding inhibition, the inhibition of complement activation was greater for each of the PEG-l ipids when the surface density of P E G - l i p i d was higher. Using C H - P E G 6 0 0 (panel A ) as an example, while 5 mole % C H - P E G 6 0 0 had no effect on complement activation, incorporation of 15 mole % C H - P E G 6 0 0 resulted in the complete abolition of complement activation by these anionic liposomes. Also parallel to C l q binding inhibition results for 100 nm liposomes, the inhibition of complement activation was enhanced as the P E G M W increased. While 15 mole % of C H - P E G 6 0 0 was required for abolition of complement activation, 10 mole % of C H - P E G 1 0 0 0 (panel B) or 5-7.5 mole % P E - P E G 2 0 0 0 (panel C) was sufficient to virtually abolish complement activation. 120 FIGURE 32: Inhibition of C l q Binding to 100 nm CL20 Liposomes by Incorporation of PEG-lipids. The effect of C H - P E G 6 0 0 (panel A ) , C H - P E G 1 0 0 0 (panel B) and P E - P E G 2 0 0 0 (panel C ) on C l q binding to C L 2 0 liposomes was determined by measuring C l q depletion from human serum (diluted 1/60 in the reaction) using the C l q E L I S A . Liposomes were composed of E P C : C H (55:45 mole %) (O) , E P C : C H : C L (35:45:20 mole %) ( • ) , E P C : C H : C L : ( C H - P E G 6 0 0 or C H - P E G 1 0 0 0 ) at (35:45-n:20:n) and E P C : C H : C L : ( P E -P E G 2 0 0 0 ) at (35-n:45:20:n). Values for n were as follows: n=2.5% (0), n = 5 % (*), n=7.5% ( • ) , n=10% (v), and n=15% ( • ) . Each point represents the mean of 3 experiments done in duplicate; bars represent one S . E . M . 121 100 Liposome Cone . (mM) 122 FIGURE 33: Complement Activation by 100 nm CL20 Liposomes: Inhibition by PEG-lipid Incorporation. The effect of C H - P E G 6 0 0 (panel A ) , C H - P E G 1 0 0 0 (panel B) and P E -PEG20oo (panel C) on complement activation by C L 2 0 liposomes was determined by complement hemolytic assay in human serum (diluted 1/10 in the reaction). Liposomes were composed of E P C . C H (55:45 mole %) (O) , E P C . C F L C L (35:45:20 mole %) ( • ) , E P C : C H : C L : ( C H - P E G 6 0 0 or C H - P E G 1 0 0 0 ) at (35:45-n:20:n) and E P C : C H : C L : ( P E -P E G 2 0 0 0 ) at (35-n:45:20:n). Values for n were as follows: n=2.5% (0), n = 5% (A) , n=7.5% ( • ), n=10% (v), and n=15% ( • ) . Each point represents the mean of 4 experiments done in duplicate; bars represent one S . E . M . 123 L i p o s o m e C o n e . (mM) 124 6.5 Effect of P E G - l i p i d Incorporation on C l q Binding by 240 nm Liposomes and Complement Activation Figures 34 and 35 illustrate the effect of PEG-l ipids on C l q binding to larger extruded liposomes with a mean diameter of 240 nm. The inhibition of C l q binding by incorporation of PEG-l ipids depended on the surface density and chain length of the P E G . Using C H - P E G 6 0 0 as an example, at 5 mole % C H - P E G 6 0 0 , C l q was inhibited by 55-75 % depending on the l ip id concentration; at 7.5 mole %, 60-80 % inhibition was achieved; and, at 15%, C l q / C L 2 0 liposome interactions were abolished. The direct relationship between P E G chain length and inhibitory capacity is illustrated by the observation that half of the amount of P E - P E G 2 0 0 0 , 2.5 mole%, was required to achieve the same effect as C H - P E G 1 0 0 0 at 5 mole %. Comparing C l q binding results for 240 nm liposomes with those for the 100 nm liposomes (Figure 32), a smaller amount of P E G - l i p i d was required to achieve a dramatic decrease in the C l q bound (Figure 34). A 55 to 70 percent decrease in C l q depletion was achieved by addition of the smallest P E G concentrations, independent of the type of P E G -lipid anchor. These C l q binding profiles for 240 nm liposomes bearing PEG-l ipids were very different from profiles for 100 nm liposomes. C l q binding to 240 nm diameter C L 2 0 liposomes with varying amounts of either P E - P E G 2 0 0 o o r C H - P E G i 0 0 0 was also quantified using the equilibrium C l q binding assay (Figure 35). Binding profiles for liposomes with C H - P E G 1 0 0 0 and P E - P E G 2 0 0 0 were very similar. Total blockage of C l q binding was achieved by adding 10 mole % C H - P E G 1 0 0 0 or 7.5 mole % P E - P E G 2 0 0 0 to the l ip id composition at C l q concentrations tested up to 125 0.041 uM (19 jug/ml) C l q . A t 0.066 i t M (30 /ig/ml) C l q , 90% of C l q binding was blocked in both cases. Either 7.5 mole % C H - P E G 1 0 0 0 or 5 mole % of P E - P E G 2 0 0 0 was sufficient to achieve 80-90% inhibition of C l q / C L 2 0 binding at all C l q concentrations tested. For the lowest amounts of PEG-l ipids tested: 5 mole % C H - P E G 1 0 0 0 and 2.5 mole % PE-PEG 2 0 oo; less efficient blockage of C l q / C L 2 0 binding was observed. Inhibition of C l q binding to these liposomes was C l q concentration dependent such that at lower amounts of C l q , 80-90% of C l q binding was blocked. A t the highest C l q concentration tested (0.066 uM), only 30-45% of C l q binding was blocked. These binding assay results were similar to results from the C l q E L I S A and provide further evidence that fairly small amounts of PEG-l ip ids were sufficient for virtually total inhibition of C l q mediated complement activation by larger liposomes. 126 F I G U R E 34: Inhibition of C l q Binding to 240 nm C L 2 0 Liposomes by Incorporation of PEG-l ip ids . The effect of C H - P E G 6 0 0 (panel A ) , C H - P E G 1 0 0 0 (panel B) and P E - P E G 2 0 0 0 (panel C) on C l q binding to C L 2 0 liposomes was determined by measuring C l q depletion from human serum diluted 1/60 ([Clq] ~ 1.67 Mg/ml or 0.0036 /AM) using a C l q E L I S A . Liposomes were composed of E P C : C H (55:45 mole %) (O) , E P C : C H : C L (35:45:20 mole %) ( • ) , E P C : C H : C L : ( C H - P E G 6 0 0 or C H - P E G 1 0 0 0 ) at (35:45-n:20:n) and E P C : C H : C L : ( P E - P E G 2 0 0 0 ) at (35-n:45:20:n). Values for n were as follows: n=2.5% (0), n = 5% (*), n=7.5% ( • ), n=10% (v), and n=15% ( • ) . Each point represents the mean of 3 experiments done in duplicate; bars represent one S . E . M . 127 100 Liposome Cone. (mM) 128 F I G U R E 35: C l q Binding to 240 nm C L 2 0 Liposomes with Incorporated PEG-l ip ids . The C l q equilibrium binding assay was used to measure C l q binding to 240 nm C L 2 0 liposomes with incorporated PEG-l ipids at p H 7.2 in a purified system. The effect of C H -PEGIOQO (panel A ) and P E - P E G 2 0 0 0 (panel B) on C l q binding was expressed as the % inhibition of binding. Liposomes were composed of either E P C . C H : C L : C H - P E G 1 0 0 0 at 35:(45-n):20:n mole % or E P C : C H : C L : P E - P E G 2 0 0 0 at (35-n):45:20:n mole % where n = 2.5% (0), n=5% (A), n=7.5% ( • ) or n=10% (v) mole %. The total l ipid concentration in the reactions was 0.5 m M . Each point represents the mean of 3 experiments done in duplicate; bars represent one standard deviation. 129 A 130 The inhibition of complement activation by 240 nm anionic liposomes by the incorporation of PEG-l ipids is shown in Figure 36. Here again, the capacity of P E G -lipids to inhibit complement activation by 240 nm liposomes was dependent upon the P E G surface density and M W . For example, while 5-7.5 mole % C H - P E G 6 0 0 provided 60-70 % inhibition, 15 mole % completely abolished complement activation. Wi th respect to P E G chain length, the same inhibitory capacity was achieved by either 2.5 mole% of P E - P E G 2 0 0 0 or 5 mole % of C H - P E G 1 0 0 0 . In each case, less P E G - l i p i d derivative was required to produce a significant effect on complement activation for 240 nm liposomes (Figure 36) than for 100 nm liposomes (Figure 34). This enhanced effect was most noticeable at the lowest P E G concentrations. Addit ion of 5 mole % C H - P E G , ^ (panel A ) , which had no effect on 100 nm liposomes, blocked 50% of complement activation in the 240 nm liposomes. Similarly, 5 mole % C H - P E G 1 0 0 0 (panel B) showed a much greater inhibitory effect in the larger liposomes. 131 FIGURE 36: Complement Activation by 240 nm CL20 Liposomes: Inhibition by PEG-lipid Incorporation. The effect of C H - P E G 6 0 0 (panel A ) , C H - P E G 1 0 0 0 (panel B) and P E -P E G 2 0 0 0 (panel C) on complement activation by C L 2 0 liposomes was determined by complement hemolytic assays. Liposomes were composed of E P C : C H (55:45 mole %) (O) , E P C : C H : C L (35:45:20 mole %) ( • ) , E P C : C H : C L : ( C H - P E G 6 0 0 or C H - P E G 1 0 0 0 ) at (35:45-n:20:n) and E P C : C H : C L : ( P E - P E G 2 0 0 0 ) at (35-n:45:20:n). Values for n were as follows: n=2.5% (0), n=5% n=7.5% ( • ) , n=10% (v), andn=15% ( • ) . Each point represents the mean of 3 experiments done in duplicate; bars represent one S . E . M 132 Liposome Cone. (mM) 133 Chapter 7: Discussion 7.1 Characterization of C l q Binding and Complement Activation by Anionic Liposomes Although it is generally accepted that complement plays a role in the clearance of liposomes from the circulation, the interaction of complement with liposomes has not been studied in detail. In this thesis, results which characterize the activation of complement by the binding of C l q to the liposome surface are presented. 7.1.1 Limitations, Assumptions, and Justification Purified protein systems were used in this study in order to characterize C l q binding to anionic liposomes. Purified protein systems allow for binding measurements to be made under a variety of known conditions and for the most direct interpretation of binding data. The drawback to the use of purified protein systems is that many other protein-lipid interactions occur in whole serum and these may affect the binding of C l q to liposomes in vivo. For this reason, C l q binding to anionic liposomes was also measured in diluted human serum. The equilibrium C l q binding assay was the primary assay used to measure C l q binding to liposomes under different conditions. The binding measurements made with this assay were independent of the experimental conditions used to separate liposome-bound C l q from free C l q (i.e. the centrifugation speed and time). This suggests that the equilibrium established upon incubation of the reaction mixture is not altered by the experimental design. 134 In order to attempt to correlate our results in vitro with clearance observations in vivo, the in vitro assays presented here were carried out with l ip id concentrations that spanned a range relevant to the in vivo studies of others. Most experiments in mouse models have been done using total l ipid concentrations in the blood ranging from 0.1-2.5 m M (Woodle, M . et al. 1992, Chonn, A . et al. 1992 and Al l en , T . et al. 1991). The amount of C l q bound to liposomes was expressed most often throughout this work as /xmoles of C l q bound per ttmole of total l ipid, since these were the quantities measured. Wi th respect to converting total l ipid concentrations used in these studies to numbers of vesicles or to units of surface area, two noteworthy assumptions were made. First, 100 nm liposomes were unilamellar and 240 nm liposomes (extruded through 400 nm pore size filters) were multilamellar to some degree. The proportion of exposed lipid for 240 nm vesicles was taken to be 35 %, based on 3 1 P - N M R data for liposomes with compositions similar to the ones used in this study (Devine, D . et al. 1994). The amount of exposed l ip id was taken to be 5 % for unextruded M L V s based on an estimate in Biochemistry of Lipids and Membranes (Cullis, P. and Hope, M.1985). These estimates were used since specific 3 1 P - N M R data for C L 2 0 liposomes could not be obtained due to the interaction of C L 2 0 vesicles with M n 2 + , the quenching agent (vesicles aggregated). It is possible that the use of these values underestimates the amount of exposed l ipid for C L 2 0 vesicles since there may be fewer lamellae due to the greater amount of negative charge on these vesicles. Second, since these liposomes contained at least 20 mole % of C L which has two phosphate groups and four acyl chains, the number of vesicles per tunole of total l ipid was estimated assuming that the average surface area per l ipid was 135 0.7 nm 2 . With regard to the p i of C l q , results from both the current study and the study by Rosano and Hurwitz (1977) are in agreement. However, the neutral p i value presented in these studies differs from results of the early characterization of C l q where the immunoelectrophoretic migration of C l q and IgG were compared (Calcott, M . and Muller-Eberhard, H . 1972). Complement literature generally sites the p i of C l q as > 9 despite the fact that little direct evidence is available. In light of isoelectric focussing data repeated twice in this thesis and of similar values obtained in the study by Rosano and Hurwitz (1977), the p i of C l q was taken to be 6.8-7.1. The cationic nature of C l q with respect to binding anionic substances can be explained by regions of positive charge on the protein surface. Indeed, C l q is a large and complex protein. Consequently, while net charge may play a role in non-specific electrostatic interactions, the overall net charge of the protein is not the most relevant feature in specific interactions; rather, it is the surface charge on the region of interest which may play a role in specific interactions. Final ly, the nature of the interaction between C l q and anionic liposomes under equilibrium conditions was assessed by Scatchard analysis. This method of analysis provides information about the strength and amount of binding and may also indicate the complexity of binding interactions. While the method was originally used to describe the binding of ligands to macromolecules, the same equations for this simple binding model apply for the binding of proteins to surfaces. A straight line Scatchard plot infers a simple interaction between the protein and theo surface in which the protein binds to the surface at binding sites which are identical and independent. Non-linear Scatchard plots suggest that 136 the interaction is more complicated. The non-linear plots, such as those generated for some of the data in these studies which curve inward towards the axes origin suggest either negative cooperativity or multiple classes of independent sites. Because non-linear Scatchard plots were obtained, a combination of sources of information regarding the interaction between C l q and anionic liposomes was drawn upon to form the model presented in section 7.1.6. C l q binding to anionic liposomes may be complicated by the strong electrostatic component of the interaction (discussed later in detail) or by the possible presence of C l q aggregates discussed in section 7.1.8. For the curvi-linear Scatchard plots, the steep linear portion of the curve was assumed to describe the major contributing part of the interaction. Consequently, an apparent association constant was derived from this portion of the plot. 7.1.2 The Requirement for Negative Charge Measurement of complement activation by liposomes has been previously shown to depend on the liposome surface charge with activation of the classical pathway of complement requiring negative charge (Chonn, A . et al. 1991; Kovacsovics, T. et al. 1985; Marjan, J . et al. 1994; Devine, D . et al. 1994). The lack of detectable classical pathway complement activation in vitro by neutral liposomes containing E P C , D P P C , D M P C , D S P C and C H shown in this study (Figure 4) supports the requirement for negative surface charge. A s wel l , throughout this study, under varied conditions of p H and ionic strength and with liposomes of different sizes, C l q binding to neutral P C : C H (55:45 mole %) liposomes was not detected. These results are in agreement with the 137 study by Kovacsovics et al. (1985) who reported that neutral liposomes composed of 100 % P C , P C : C H at 60:40 mole %, and P C : P E (50:50) failed to bind C l q or to activate C I . In contrast to these findings, others have suggested that complement does play a role in the fate of neutral liposomes. Detection of C 3 fragments, C3b and iC3b, on neutral liposomes would indicate an opsonic role for complement. While some studies did not find C3 associated with neutral liposomes (Marjan, J . et al. 1994), others have detected fragments of C3 on E P C : C H liposomes although the amount was much less than on anionic liposomes (Chonn, A . et al. 1991 and Chonn, A . et al. 1992). Additional evidence is supplied by Wassef and A l v i n g (1987 & 1993) who have shown that neutral liposomes which are pre-opsonized with complement are rapidly taken up by macrophages in vitro. A study by Gaber et al. (1995) demonstrated that complement was involved in the destabilization of neutral thermosensitive liposomes in bovine serum, but not in human serum. Reports by M o l d (1989) and Comis and Easterbrook-Smith (1986) showed alternative pathway activation by P C : P E (80:20) liposomes in human serum. Further, the fact that even neutral liposomes, while having longer half-lives than charged liposomes, still have relatively short lifetimes in the circulation may indicate complement involvement. Several labs have shown that less than eight percent of the initial injection of liposomes composed of E P C : C H , D S P C : C H , and S M : P C : C H is found in the blood of mice or rats at 24 hours post injection (Allen, T . et al. 1991; Lasic , D . et al. 1991; Woodle, M . et al. 1992; Gabizon, A . and Papahadjopoulos, D . 1988). For D S P C : C H liposomes at a saturating dose, a blood half-life of 20 hours has been reported; below saturation, however, the half-life is less than 10 hours (Woodle, M . and Lasic, D . 1992). 138 The lack of detection of C l q binding and complement activation in the assays used in this thesis and by others may be related to in vitro experimental conditions and species differences. In the human serum experiments in this thesis, the maximum time that liposomes were exposed to human serum at a maximum concentration of 1/4 was 1.5 hours. In purified system experiments, C l q was incubated with liposomes for 20 min at R T . It is possible that complement interactions with neutral liposomes require prolonged incubation times in vitro. A l so , complement may interact with neutral liposomes after they have acquired a negative charge upon incubation in plasma or in the circulation due to the binding of negatively charged plasma proteins. Arguing against this hypothesis that complement interactions with liposomes may be time-dependent, is the observation that neutral liposomes adsorb proteins almost instantaneously upon incubation in plasma (Senior, J . et al. 1991; Chonn, A . et al. 1991). Difficulties in detecting the interaction of neutral liposomes with complement and resulting discrepancies in the literature indicate that a different mechanism may be involved in neutral liposome/complement interactions other than direct C l q binding; perhaps C R P is involved (Volanakis, J . and Narkates, A . 1981). A s wel l , for these neutral liposomes, non-complement opsonins such as fibrinogen or fibronectin or opsonin-independent attack by macrophages may play a more important role in clearance. Liposomes which bear a negative surface charge have been used in this study to characterize complement activation in vitro. Negatively charged liposomes exhibit shorter circulation times (Juliano, R. and Stamp, D . 1975) and bind more total protein than neutral or positively charged liposomes (Chonn, A . et al. 1992). Several studies have 139 shown enhancement of phagocytic uptake by inclusion of negative phospholipids (Fujiwara, M . et al. 1996; L i u , D . et al. 1995; Lee, K . - D . et al. 1992 & 1993; Dijkstra, J . et al. 1985; Hsu, M . and Juliano, R. 1982). The contribution of complement to this rapid uptake was clearly described by Harashima et al. (1994b), Matsuo et al. (1994) and L i u et al. (1995) who showed that the enhanced uptake of anionic liposomes by liver phagocytic cells in a perfused rat liver model was due to liposome opsonization with iC3b particles. While some negatively charged liposomes exhibit these characteristics, it appears that not all anionic lipids are equal in their effects and that in addition to net negative charge, there are other factors, such as the chemical nature of the phospholipids, that are important. Several studies have shown that different amounts of total protein and different types of proteins become associated with liposomes that have the same net charge imparted by different anionic components (Chonn, A . et al. 1992; Hernandez-Caselles, T . et al. 1993). The rapid clearance of anionic liposomes appears also to be a function not only of negative surface charge but of the nature of the negative phospholipids used. While it is generally agreed that liposomes containing PS , P A , and P G have enhanced clearance (Gabizon, A . and Papahadjopoulos, D . 1992), liposomes bearing negative charge imparted by plant PI have been reported to have relatively long half-lives (Gabizon, A . and Papahadjopoulos, D . 1992; Chonn, A . et al. 1992). The negative charge on the liposomes used in this study was imparted by inclusion of C L at 20-40 mole %. While other anionic phospholipids such as P G and PS have one negatively charged phosphate group, cardiolipin has two. Consequently, C L has a net 140 charge of -2 which is twice the charge on P G and PS. Cardiolipin-containing liposomes have consistently been demonstrated to be strong activators of the classical pathway of complement with appreciable levels of C3 detected on the membrane after only two minutes in the circulation of mice (Chonn, A . et al. 1992). Liposomes containing E P C : C H : C L at 35:45:10 mole % have half-lives of less than two minutes (Chonn, A . et al. 1992). Thus, C L and the closely related PG-containing liposomes serve as suitable model membranes on which to study complement activation by the binding of C l q in detail and for studying the complement-modifying potential of P E G - l i p i d incorporation into the membrane. 7.1.3 The Effect of Surface Charge Density on C l q Binding to Anionic Liposomes Protein interactions with liposomes can be mediated by electrostatic, hydrophobic, and van der Waals forces or by covalent bonds. In studying the binding of isolated C l q with C L and P G containing liposomes, the electrostatic component was dominant. The negative surface charge density was the first obvious factor which affected C l q binding. As the proportion of anionic phospholipid in the liposomes increased, consequent increases in C l q binding at physiologic ionic strength in a purified protein system at p H 7.2 (Figure 10 & Table 4) or p H 4 (Figures 17,18 & Table 5) were measured. This relationship between surface charge density and C l q binding parallels the effect of anionic phospholipid concentration on overall complement activation which has been shown in several studies (Marjan, J . etal. 1994; Chonn, A . etal. 1991; Devine, D . etal. 1994) lending further support to the hypothesis that the direct, immunoglobulin-free binding of 141 C l q to liposomes leads to complement activation. A previous study by Kovacsovics et al. (1985) also showed that the antibody-independent binding of C l q and activation of C I by cardiolipin-containing liposomes was dependent upon the amount of negative charge. It is noteworthy that while the saturation level of C l q binding to anionic liposomes presented in this study increased with increasing concentrations of the negatively charged phospholipid in the liposome composition, the binding affinities decreased. This is the opposite to what is expected for a purely electrostatic interaction, suggesting that the actual binding of C l q to the liposome surface relies on chemical energy as well . A model for the interaction of C l q with anionic liposomes is discussed in section 7.1.6. 7.1.4 The Effect of p H on C l q Binding to Anionic Liposomes In addition to the effect of negative surface charge density, C l q binding to anionic liposomes was highly dependent upon the p H of the environment. Optimal C l q binding occurred at p H 4 (Figure 15) where C L 2 0 liposomes bind 100 times more C l q than at p H 7. Similarly, C l q has been demonstrated to directly bind d s D N A at low p H (pH 4.45). The same study showed that these low p H conditions abolished C l q binding to heat aggregated human IgG (Uwatoko, S. and Mannik, M . 1990). This observation suggests that a conformational change in C l q may occur at low p H which affects the globular head groups of the protein (the IgG binding sites) but not the collagen-like region which is involved in D N A binding. The results summarized in Figure 16 demonstrate the reversibility of the effect of exposing C l q to low p H , thus implying that the region involved in binding anionic liposomes, likely the collagen-like portion of C l q , was not 142 modified in a permanent way at low p H . ^ The results showing a huge p H effect on C l q binding to anionic liposomes resembles work by Bergers, J . et al. (1993) who studied the adsorption of other proteins, including myoglobin which has a PI similar to C l q (PI m y o g l o b i n = 6 . 9 - 7 . 4 and P I c l q = 6.8 - 7.1), to anionic liposomes containing 10 mole % P G as a function of p H under low ionic strength conditions (I = 0.01 M ) . Bergers found that adsorption of proteins to anionic liposomes was only found at p H values where the proteins had an overall positive charge. The plateau level reached for protein-liposome binding depended on the p H of the incubation medium and this p H dependency was ascribed to the magnitude of the net positive charge of the protein. For myoglobin, adsorption to liposomes occurred only when the p H was dropped to 5.5 and below; at this p H the protein was positively charged. Adsorption was not detected at p H 7.3 since the protein charge was zero to slightly negative. For myoglobin, then, it was determined that adsorption to anionic liposomes was critically dependent upon electrostatic attraction between membrane- and protein-associated charges as the initial driving force. Values for C l q binding to C L and PG-containing liposomes are similar with respect to the enhanced plateau binding levels at low p H compared to high p H and to the very low but detectable levels of C l q binding at p H 7.2, close to the PI of the protein. Clearly, electrostatics play a critical role in the potential of C l q binding to anionic liposomes. However, the fact that C l q binding to anionic liposomes was detected at p H 7.2, where C l q bears no net charge, suggests that other forces may also play a role. Work on the binding of cationic peptides to CL-containing liposomes by De Kroon et al. 143 (1990) supports this hypothesis. De Kroon reported that both hydrophobic and electrostatic interactions determined the binding affinity. In addition, the region of C l q which is thought to be responsible for the binding of non-immunoglobulin activators of complement, ( C l q A 1 4 . 2 6 ) , contains both cationic and hydrophobic residues. 7.1.5 The Ionic Strength Dependence of C l q Binding to Anionic Liposomes Another indicator of the importance of electrostatics in Clq/ l iposome interactions is the ionic strength dependence of C l q binding to anionic liposomes. C l q binding to anionic extruded liposomes and M L V s was enhanced enormously when the ionic strength was lowered from physiologic (0.145 M ) to 0.02 M (Figures 20 and 21). Lowering the ionic strength from 0.145 M to 0.1 M resulted in 80 times more C l q binding to 240 nm liposomes and 14 times more binding to M L V s in the equilibrium C l q binding assay. A n additional assessment of the effect of ionic strength on Clq/ l iposome binding was provided by the particle electrophoresis technique. The electrophoretic mobility, the velocity of a vesicle in an applied electric field, is proportional to the surface charge density of the liposome. When C l q is added, the binding of C l q to liposomes is detected as a decrease in the liposome electrophoretic mobility due to neutralization of the liposome surface charge and to the charges on the liposome-bound proteins. Consequently, the decrease in liposome electrophoretic mobility is directly related to the amount of C l q bound. Figure 20 confirmed that C l q binding to C L 2 0 M L V s was enhanced as the ionic strength was decreased because as the ionic strength decreased, the electrophoretic mobility of the liposomes decreased in the presence of C l q . 144 It is important to note that due to the experimental design for this particle electrophoresis technique, Clq-l iposome interactions may be underestimated by these measurements. For the Rank Mark I apparatus used in this dissertation work, the sample chamber, which needs to be filled, holds a volume of 2.6 ml . The requirement to f i l l the chamber means that either a homogenous mixture of liposomes and proteins in buffer is loaded into the chamber or the chamber is filled with buffer and the liposome/protein reaction mixture is injected in the chamber at the position of the eyepiece. The injection method is technically very challenging and measurements of liposome-protein binding are time-dependent due to the dilution effect and the potentially reversible nature of the reaction. For this work, the chamber was flooded with a homogenous mixture. In order to conserve purified C l q , the reaction mixture volume was kept to 780 jul and this mixture was then diluted with buffer to bring the volume up to 2.8 m l . This large dilution factor prior to making the electrophoretic mobility measurements may have effectively washed some of the C l q off of the liposomes depending on the reversibility of the C l q -liposome binding under the conditions of the assay. Reversibility of the reaction has been shown (Figure 11) under 0.1 M N a C l and p H 7 conditions. This dilution effect explains why C l q binding to M L V s at physiologic p H and ionic strength which was measured in the equilibrium binding assay was not detected by particle electrophoresis. The enhancement of C l q binding by lowering the ionic strength has been commonly used throughout studies investigating the nature of C l q binding to other substances. For example, the binding of single stranded D N A to C l q was found to be inversely related to ionic strength (Uwatoko, S. and Mannik, M . 1990). A large number 145 of the Clq- l igand affinity constants reported in the literature have been determined under subphysiologic ionic strength conditions. Examples of Clq- l igand binding studies using buffers of ionic strength in the range of 0.015 M - 0.07 M to optimize binding include investigations into C l q binding to fibrinogen and fibrin (Entwistle, R . and Furcht, L . 1988), fibronectin (Bing, D . et al. 1982 and Sorvil lo, J . et al. 1985), laminin (Bohnsack, J . et al. 1985), collagen (Menzel, E . et al. 1981), bacterial lipopolysaccharides (Zohair, A . et al. 1989) and monomer and polymer forms of mouse I g M (Taylor, B . et al. 1994). Even C l q binding to specific receptors on cells and platelets was found to be ionic strength dependent (Schreiber, R . 1984). For instance, Peerschke and Ghebrehiwet (1990) showed enhanced adhesion of C l q to platelet receptors at 20 m M N a C l compared to 150 m M . Bordin and Page (1989) demonstrated that the K a for C l q binding to a human fibroblast subtype decreased 2.7 fold from 100 m M to 150 m M N a C l . Chen et al. (1994) measured receptor-specific C l q binding to human T cells at 90 m M N a C l and suggested that their conditions were not optimal since binding to other cells, such as B cells was measured under even lower ionic strength conditions. Hence, these studies concur with our data which suggests at physiological ionic strength, the fluid phase interaction between C l q and many of its binding partners is very low. 7.1.6 A Mode l for Clq /Liposome Interactions Under physiologic conditions, the binding of C l q to anionic liposomes was very low. The saturation binding value for C L 2 0 liposomes was 2 x 10"6 ptmole C l q bound per /xmole of total l ipid. Converting this value to the number of C l q molecules bound per 146 C L 2 0 vesicle gave 3 C l q molecules for every 2 liposomes. To provide an explanation for this low level of C l q saturation binding at p H 7.2 and 145 m M N a C l and for the observations that decreasing the p H and ionic strength and increasing the concentration of anionic phospholipid in the membranes increased the saturation binding levels without increasing the affinity constants, a model for C l q binding to anionic liposomes is presented. The binding of a charged protein to an oppositely charged surface is always affected by two factors. First, the non-specific electrostatic attraction of the protein for the surface results in accumulation of ligand near the surface. Second, the actual intrinsic binding of proteins to surfaces then depends on the specific electrostatic, hydrophobic, and van der Waals interactions. Before discussing the electrostatic model with respect to C l q binding to anionic liposomes, it is useful to first consider the distinction between apparent equilibrium constants (Ka) and intrinsic or microscopic equilibrium constants (Ki) . Apparent equilibrium constants for C l q binding reflect both the electrostatic attraction which allows for the increased concentration of C l q near the liposome surface as well as the actual intrinsic binding of C l q to the liposome which is determined by the specific bonding energies. When the liposome surface C l q concentration is higher than the bulk concentration due to electrostatic attraction, apparent C l q binding constants reflect the combination of both the electrostatic recruitment and the specific binding components. Intrinsic equilibrium constants do not include the bulk electrostatic contribution which increases the local concentration of C l q near the liposomes surface. When the non-147 specific electrostatic contribution is absent, such as when C l q has no net charge, only the intrinsic binding component is measured. In this case, the apparent binding constants are equal to the intrinsic binding constants. The fact that p H , ionic strength, and liposome surface charge density strongly affect the level of C l q binding to liposomes indicates that electrostatic attraction is necessary for C l q to bind to liposomes. The electrostatic component may in fact be responsible for the curvi-linear nature of some of the Scatchard plots obtained for Clq/anionic liposome interactions. Negative cooperativity would be expected since the negative surface charge on the liposome decreases as ligand binding proceeds. However, the electrostatic or Stern model of binding states that the apparent binding constant increases exponentially with an increase in the negative surface charge (Shaw, D . 1970). For the simplest case of a monovalent ion binding an oppositely charged surface, the relationship is K a = K i exp(-F4 / ( 0 ) /RT) where F = Avagadro's number x the electron charge, is the magnitude of the negative surface potential, R is the gas constant, and T is the absolute temperature. Because Clq/ l iposome association constants decreased as the surface potential was increased, the opposite behaviour to that predicted by this model, the electrostatic model was considered insufficient for describing C l q binding to anionic liposomes. While the electrostatic component is important, this electrostatic model alone does not explain the saturation behaviour or the relative affinities of C l q binding to anionic liposomes under different conditions. The model for C l q binding to anionic liposomes proposed here and illustrated in Figure 38 has three components. First, the electrostatic potential produced by the acidic 148 phospholipids in the liposomes attracts C l q , thereby increasing the concentration of C l q in the aqueous phase at the liposome surface. Similar increases in the concentration of cationic peptides at negatively charged liposome surfaces has been described by Mosier and McLaughl in (1992). Further support is drawn from observations that specific interactions between enzymes and substrate-bearing liposomes are enhanced by the addition of acidic phospholipids in the liposome composition. Here again, it is suggested that the electrostatic potential at the membrane produced by the acidic phospholipid results in an increase in the enzyme concentration at the liposome surface (Rebecchi, M . etal. 1992). Second, it is proposed that the binding sites for C l q are created by the electrostatically-induced recruitment and concentration of the acidic phospholipids at a position in the liposome next to the C l q at the surface. Enhancement of the local concentration of a given l ipid due to electrostatic lipid-protein interactions has been reported by other investigators (Cullis, P . et al. 1985). For instance, the introduction of glycophorin into liposomes promotes the transbilayer movement (flip-flop) of P C (De Kruijff, B . et al. 1978). Poly lysine has been shown to affect the lateral segregation of acidic lipids to such an extent that for CL-containing liposomes, poly lysine can actually precipitate the cardiolipin (De Kruijff, B . and Cul l i s , P . 1980). In a study of the aggregation of anionic liposomes by addition of a polycationic polymer, Yaroslavov et al. (1994) hypothesized that the interaction of the polycation with CL-containing liposomes induced the migration of cardiolipin from the inner to the outer leaflet of the liposome membrane. 149 The third element of the model proposed is that C l q binds anionic liposomes at these created sites through a combination of intrinsic chemical and electrostatic forces. Other studies have shown that both electrostatic and hydrophobic interactions determine the intrinsic affinity of positively charged peptides for negatively charged vesicles (De Kroon, A . et al. 1990). A n example of a protein which has a significant number of basic residues but adsorbs to membranes by means of hydrophobic interactions is the protein melittin. For melittin, its structure prevents the cluster of basic residues from interacting with the acidic lipids of the membrane (Beschiaschvili, G . and Seelig, J . 1990). Another example of proteins interacting and in fact inserting into membranes is provided by the complement proteins which make up the membrane attack complex. A s C7 and C8 jo in with C5b6, the resulting complexes become more hydrophobic and insert further into the lipid membrane. The saturation binding levels and affinity constants for C l q binding to anionic liposomes can now be explained in a physical sense by applying this model. The low amount of C l q bound at saturation to anionic liposomes at neutral p H and physiologic ionic strength was due to the lack of binding sites created. A t neutral p H , C l q has little or no average net charge (pi 6.8-7.1). Consequently, the concentration of C l q at the liposome membrane was very low. Because there was very little C l q close to the liposome and because the protein has no net charge, the electrostatic induction of binding sites was rare. Upon creation of the occasional binding site, C l q binding occurred as a result of chemical interactions ( C l q still has pockets of positive charge, even though the overall net charge is zero). 150 F I G U R E 38: Clq /L iposome Interactions: A M o d e l . Three steps to the model for C l q binding to anionic liposomes are presented. First, C l q is electrostatically attracted to the liposome surface. Second, binding sites are created through electrostatic attraction between the positively charged C l q and the anionic phospholipids of the liposome. This results in the segregation of anionic phospholipids and in the formation of binding sites. Third, the actual intrinsic binding of C l q to the liposome is mediated by chemical forces such as hydrophobic interactions. Electrostat ic Crea t ion recruitment of Binding of C 1 q binding sites 151 Under physiologic conditions, when the acidic l ipid concentration is increased in the liposome composition, C l q binding saturation levels, although still low, increased (5 C l q molecules per C L 3 0 vesicle and 50 C l q molecules per C L 4 0 vesicle). Because C l q was neutral overall under these conditions, the concentration of C l q at the liposome surface was still low. However, since there was a greater concentration of C L in the liposomes, the creation of binding sites occurred more readily in response to the proximity of the C l q protein. Once the binding sites were created, C l q binding occurred The observation that the binding affinity actually decreased when the acidic phospholipid concentration increased suggested that the binding sites created on these liposomes were lower affinity sites. A purely electrostatically-mediated binding would have increased in apparent affinity with increasing negative surface charge. Therefore, these results suggested that the actual intrinsic binding of C l q to anionic liposomes involved chemical interactions, such as hydrophobic and hydrogen bonding interactions. The requirement for the electrostatically-induced creation of binding sites and for both chemical and electrostatic forces to be involved in the actual binding also offers an explanation for the increased saturation binding upon decreasing the p H or ionic strength and the apparent independence of C l q binding affinity or association constants under these conditions. A s described earlier, lowering the p H of the system increases the net positive charge on C l q . Because of this increased positive charge, the concentration of C l q at the liposome surface would be greatly enhanced. The ability of C l q to induce the creation of binding sites on the liposomes would also be greatly increased. These enhanced electrostatic effects resulted in a large increase in C l q saturation binding. The 152 apparent association constants decreased with decreasing p H indicating that while these binding sites were more plentiful, they were also somewhat lower affinity sites. Lowering the ionic strength also resulted in an increase in C l q saturation binding values. While long-range electrostatic effects remained small due to the lack of average net positive charge on C l q , local electrostatic effects were enhanced by lowering the ionic strength. The positively charged pockets on C l q provided the electrostatic requirement for the formation of binding sites on the liposomes. The affinity constant for C l q binding to C L 2 0 liposomes at 100 m M N a C l was the same as the constant for C l q binding at 145 m M N a C l (2 x 10 9 M" 1 ) suggesting that a similar extent of movement of phospholipids in the membrane was induced resulting in higher affinity sites. 7.1.7 Possible Biological Reasons for L o w C l q Binding Levels Under Physiologic Conditions From a biological perspective, the low level of binding of C l q to anionic liposomes at p H 7.2 and physiologic ionic strength was somewhat surprising since the total complement activation assay showed strong activation by these liposomes. Others have made similar observations, however. Kovacsovics et al. (1985) found that C I activation was more readily detected than C l q binding despite the fact that they used assay conditions which promote the binding interaction, namely measuring C l q binding to large unilamellar vesicles of 1.3 am in diameter at the subphysiologic ionic strength of 100 m M N a C l . This is no doubt due to the cascade nature of complement activation which allows for amplification once the recognition process, the binding of C l q , has 153 initiated the pathway. This amplification results in the triggering of a greater full complement response than the level of initiation, C l q binding, might appear to warrant. Complement activation may also be detected more readily than C l q binding because while activation occurs instantaneously upon binding of C I to the activator, to measure C l q binding, the interaction must be stable. In addition, although liposomes have been found to activate complement in immunoglobulin-free human serum, more anionic phospholipid was required in the liposome composition to allow detection of complement activation in the absence of immunoglobulins (Marjan, J . et al. 1994) indicating that while C l q binds directly to anionic liposomes, more C l q binds when immunoglobulins are also present. The complement activation assays carried out i n this thesis used whole normal human serum and thus measured effects of the combination of direct C l q binding and immunoglobulin-mediated complement activation. 7.1.8 C l q Aggregation: A Potential Role in C l q Binding It is also possible that when measuring the binding of isolated C l q , separated from the rest of the C I complex, the aggregation state of C l q may contribute to the different binding levels. C l q in an aggregated state would bind more strongly due to multivalent interactions with liposomes and would probably give higher saturation binding levels than monomeric C l q because of a larger area per molecule with which to recruit acidic l ipid. The aggregation of purified C l q is known to occur under some conditions. For instance, reducing the ionic strength to below 100 m M N a C l results in C l q aggregation 154 (Tenner, A . personal communication; Ziccardi , R . and Cooper, N.1979). C l q aggregation also occurs upon ultrafiltration, repeated freeze-thawing, and upon storage (Tenner, A . et al. 1981; Tenner, A . personal communication). A s part of the study presented in this thesis, the increase in purified C l q binding with time in storage was likely due to C l q aggregation (Figure 8). In fact, the use of an intermediate density separating buffer was required for the equilibrium binding assay because spinning C l q alone resulted in a gradient of C l q counts from the top of the tube to the bottom. This suggests either a centrifugation artifact or that purified C l q contains a mixture of aggregated and non-aggregated material. In addition, the observation recorded in Table 6 that liposomes aggregated under conditions where higher levels of C l q binding occurred (low p H and low ionic strength) suggests that C l q aggregation may mediate the liposome aggregation. This phenomenon has been shown for polymers binding to surfaces (Israelachvili, J . 1992). The neutralization of liposome negative surface charge as a result of C l q binding may also contribute to this liposome aggregation. The aggregation state of C l q may contribute to its binding capacity under physiologic conditions. This hypothesis is supported by the l ipid concentration dependence of C l q binding that was illustrated in the purified protein system (Figure 19). If only a single protein species were binding to the liposomes, the amount of protein adsorbed per unit area (i.e. per amount of liposomes) at a given equilibrium C l q concentration should be the same at al l concentrations of liposomes. The fact that the amount of C l q bound per amount of l ipid decreased with increasing amounts of total surface area (liposomes) suggested that C l q was poly disperse. A n example of the binding of a poly disperse protein 155 to a surface is provided in the study of B S A adsorption onto latex beads by Olal (1990). Results of this study confirmed that oligomeric species of B S A bound more strongly to latex than monomeric B S A (Olal , A . 1990). Based on the assumption that due to multiple contact points a larger binding species would bind preferentially to a surface compared to a smaller one, Stuart et al. (1980) suggested that when a small surface area is available, the larger species that have a higher affinity outcompete the lower affinity monomeric species. Wi th the addition of more surface area, the binding of the high affinity species is followed by an additional weaker adsorption of the lower affinity species. The combination of binding of a high and a low affinity species would appear as a smaller average amount of C l q bound per amount of surface area. The existence of a poly disperse binding species can be determined by plotting the amount adsorbed against the solution volume multiplied by the equilibrium concentration divided by the surface area (Stuart, M . et al. 1980). If the surface area dependence was due to the poly dispersion of the binding species, the binding curves should collapse into one curve. The transformed data for the two liposome concentration-dependent isotherms for C l q binding to C L 2 0 liposomes did collapse into one curve. This suggested that higher affinity species, perhaps oligomers of C l q , were present as well as lower affinity monomers. C l q in an aggregated state would bind more strongly due to multivalent interactions with liposomes. However, care must be taken with this interpretation. The facts that the saturation level of C l q binding was so low and that the binding affinity failed to significantly increase upon decreasing the ionic strength, a condition which 156 enhances C l q aggregation, implies that while the aggregation of C l q may play a role in determining the binding level of C l q , it is not the sole factor. In addition, the implied aggregation of C l q appears to be driven by forces other than electrostatics. If simple electrostatic energy was dominant, C l q would not aggregate at low ionic strength or at low p H as the enhanced charge effects on the protein would lead to C l q - C l q repulsion. C l q aggregation may be driven by other forces such as hydrophobic forces or by conformational changes in the protein under different conditions. The C l q species which is predominant in binding to anionic liposomes, C l q aggregates possibly, would still follow the model presented earlier to describe the binding results obtained in this study. In blood, C l q has been displayed as clusters or aggregates on the components of the extracellular matrix (Bordin, S. et al. 1990). For C l q binding to cells with C l q receptors, the interaction of monomeric C l q fails to trigger a biologic response whereas a response is noted when several C l q molecules are fixed to the same surface. Bordin et al. (1990) speculated that aggregated rather than monomeric C l q may be the ligand for the fibroblast C l q receptor. 7.1.9 C l q Binding to Anionic Liposomes in Serum C l q binding to anionic liposomes was also measured in the presence of the rest of the serum components (Figures 22 and 23). The binding of C l q to CL-containing liposomes, when the reaction occurred in the presence of Clq-depleted serum, was virtually the same as purified C l q binding to liposomes in buffer. A s a l ipid concentration dependence was again observed, C l q aggregates may be present, although the presence of 157 the rest of the serum components could also play a role in the observed lipid dependence. It is not known whether C l q , when added back to Clq-depleted serum in the presence of anionic liposomes, becomes complexed with C l r and C i s to form CI and binds as part of the complex or whether free C l q interacts with the liposomes. It is demonstrated here, however, that binding measurements of isolated C l q predict accurately the binding of C l q in the presence of serum components. The fundamental measurements of complement activation via C l q binding to anionic liposomes have been discussed. The effects of ionic strength and surface charge density on C l q binding reported here compare well with the influence of ionic strength and amount of negative charge on liposome stability observed by Comiskey and Heath (1990). Their study showed that decreasing the ionic strength or increasing the anionic phospholipid concentration results in enhanced serum-induced leakage of negatively charged liposomes. Enhanced complement activity may play a role in this. 7.2 The Effect of Liposome Size on C l q Binding and Complement Activation by Anionic Liposomes In addition to manipulating the environment, thereby altering C l q net charge and potentially C l q aggregation form, and altering the charge density of the membrane in order to characterize the interaction between C l q and anionic liposomes, liposomes of different sizes were made to determine the effect of liposome size on complement activation. Results of the in vitro complement activation and C l q binding assays indicate that liposome size is very important. In 50 % human serum, the C l q E L I S A detected 158 greater C l q binding for 240 nm liposomes compared to 100 nm liposomes at l ipid concentrations below 1 m M (Figure 12). A n even more apparent difference in C l q binding was obtained when the binding capacity of 240 nm extruded liposomes was compared with that of M L V s of the same composition both at physiologic ionic strength (Figure 13) and at subphysiologic ionic strength (Figure 21). Approximately 15 times more C l q bound to C L 2 0 M L V s than to extruded 240 nm liposomes under physiologic conditions. Complement activation also occurred more readily on 240 nm liposomes compared to 100 nm liposomes. Seventy-five times more l ipid was required for 100 nm particles to achieve the same level of complement activation (>20%) as the 240 nm liposomes. This direct relationship between the level of complement activation and liposome size agrees with similar observations of complement activation in rat serum (Devine, D . et al. 1994 and L i u , D . et al. 1995). Vesicle size has also been shown to be an important factor in complement-dependent immune damage to liposomes. Marker leakage experiments monitoring 1 4 C -glucose or carboxyfluorescein release have indicated that liposome size is important for complement-dependent liposome degradation in rat serum and in a purified complement protein system (Harashima, H , et al. 1994a; L i u , Z . - Y . and H u , V . 1988). The in vitro results presented here correlate well with results of in vivo clearance studies. Larger liposomes have been found to be removed from the circulation much faster than small liposomes with the exception of very small liposomes ( < ~ 80 nm) (Juliano, R . and Stamp, D . 1975; Gabizon, A . and Papahadjopoulos, D . 1992; Senior, J . et al. 1985). For anionic liposomes, evidence suggests that complement activation plays a 159 role in the elimination of liposomes. Size dependent liposome opsonization by complement (C3b) resulting in alteration of liver phagocytic cell uptake was demonstrated in a rat perfused liver model (Harashima, H . et al. 1994; L i u , D . et al. 1995). These studies suggested that larger liposomes were more effectively recognized by complement than smaller ones and that liposomes were taken up by the liver depending on the extent of opsonization (Harashima, H . et al. 1994; L i u , D . et al. 1995). The fact that Clq-mediated complement activation takes place more readily on larger liposomes suggests that there may be geometric requirements for the assembly of C I and activation of C l q which are better met on a larger surface. Geometry is known to be a factor in the assembly of other complement complexes, such as the classical pathway C5 convertase. In order for the complex to be able to cleave C 5 , C3b-bound C5 must be directly adjacent to the C4b2a complex on the membrane. However, the differences in C l q binding to 100 nm and 240 nm vesicles seems unlikely to be due to geometric requirements and'different surface areas because 100 nm liposomes are already sufficiently large such that an estimated 100 molecules of C l q would be required to cover the surface of the vesicle. Instead, the increase in C l q binding to larger liposomes, may be explained according to the C l q binding model. A greater number of C l q binding sites may be induced on a larger liposome due to the greater number of l ipid and phospholipid molecules per vesicle from which anionic phospholipids may potentially be recruited. The enhanced activation of complement by larger liposomes is significant since it contributes to the overall fate of intravenously administered liposomes. 160 7.3 Examination of a Cationic Region of C l q for Involvement in Complement Interactions with Liposomes The next goal of this thesis was to look more closely at a candidate region of C l q which could mediate the interaction of C l q with anionic liposomes. A s mentioned previously, anionic liposomes share a critical determinant for Clq-mediated complement activation with several other antibody-independent activators: a repeating negative surface charge. The major C l q binding site for this group of antibody-independent activators which includes S A P (Ying, S . - C , et al. 1993), C R P (Jiang, H . et al. 1992a), D N A (Jiang, H . et al. 1992b), and beta-amyloid fibers (Jiang, H et al. 1994), has been previously localized to residues 14-26 within the collagen-like region of the C l q A chain because synthetic C l q A ( 1 4 . 2 6 ) peptides inhibited C l q binding and complement activation. To determine i f anionic liposomes also interact with this region of the C l q A chain, we investigated the inhibitory effect of C l q A peptides on C l q binding and complement activation by anionic liposomes as well as the direct binding of C l q A peptides to anionic liposomes. 7.3.1 C l q A Peptides Inhibit C l q Binding and Complement Activation by Anionic Liposomes A s described for the above mentioned group of complement activators, C l q A ( ] 4 . 2 6 ) was capable of completely inhibiting C l q binding and complement activation by anionic liposomes. Inhibition of C l q binding to anionic liposomes required approximately the same relative amount of C l q A ( 1 4 _ 2 6 ) peptide as that required for 161 inhibition of C l q binding to C R P or to D N A . However, the amount of peptide required to inhibit C l q binding was very large; a molar ratio of 18 390 to one for peptides to C l q was required to inhibit C l q binding by 90 % (Table 10). With respect to the inhibition of complement activation by anionic liposomes, higher concentrations of peptide were required to observe inhibition. The peptide to C l q molar ratio was 61 600 to one to achieve 90 % inhibition of anionic liposome-induced complement activation. This was also the case in other reports, where, for instance, 100 times more C R P was required to observe inhibition of complement activation compared with the amount required to block C l q binding (Jiang, H . et al. 1992a). A higher concentration of peptide may have been necessary to inhibit liposome-induced complement activation because of differences between the binding assay and the complement activation (hemolytic) assay, a major one being that the hemolytic assay was performed in whole human serum (diluted 1/10 in the reaction) and the binding assay used purified protein. 7.3.2 C l q A Peptide Inhibitory Capacity Lacks Conformation and Sequence Specificity In marked contrast to the studies on other direct complement activators, the capacity of the C l q A ( 1 4 _ 2 6 ) peptide to inhibit C l q binding and complement activation by anionic liposomes lacked any sequence or conformational specificity. C l q A ( s c r a m b l e d ) and C l q A ( P . A ) peptides exhibited the same inhibitory abilities as the authentic C l q A ( 1 4 . 2 6 ) peptide (Figures 25 and 26). Further proof of the lack of sequence and conformational dependence was provided by the observation that these three peptides interacted directly with anionic liposomes to virtually the same extent (Figure 27). Differences in assay 162 systems (we have used a fluid phase binding assay, these other studies used E L I S A ) do not explain the disparate results with respect to specificity. Based on the results in Chapter 5, it appears that C l q binding and complement activation on a membrane differs from binding and activation by protein, D N A or fibre aggregates. The lack of sequence and conformation specificity in the interaction between C l q and anionic liposomes indicates that the charge component is of prime importance. This hypothesis is supported by the observation that peptides with no charge failed to bind liposomes while peptides with one to two cationic charges bound slightly at very low ionic strength and peptides with five positive charges bound aggressively (Figure 27). 7.3.3 The Requirement for Cationic Residues: Charge Control Peptides and an Unrelated Control Peptide The fact that only peptides with a large net positive charge were capable of interacting with anionic liposomes and inhibiting C l q binding and complement activation is consistent with the concept that electrostatics provide a dominant energy in the initial interaction. C l q A ( 1 4 _ 2 6 ) , C l q A ( s c r a m 5 | e d ) and C l q A ( P . A ) showed high levels of direct binding to anionic liposomes and consequently effectively competed with C l q in the C l q binding assay and in the hemolytic assay. In addition, the inverse relationship between ionic strength and C l q A peptide/liposome binding mirrored the ionic strength dependence shown for Clq/ l iposome binding (Figure 20) and for C l q / D N A interactions (Jiang, H . et al. 1992). The neutral peptide showed no binding to liposomes and no inhibitory capacity. C l q A ( 2 + ) peptide also showed no binding to anionic M L V s in 100 m M N a C l , closer to 163 physiologic salt concentration (Figure 28) and no inhibitory effects. However, as previously mentioned, while sufficient positive charge on the peptides translated into the ability to bind anionic liposomes and inhibit C l q binding, the amount of peptides required to inhibit C l q was huge. The fact that this molar ratio was so large suggests that while electrostatic energy plays an important role, it is not the only force involved in C l q binding to liposomes. This is consistent with the binding model presented earlier. The unrelated Fgn peptide results were anomalous in that this peptide which has one arginine and 2 histidine residues (contributing an overall charge of ~ +1.2 at p H 7.2) was found to inhibit C l q binding to anionic liposomes (up to 70 % with the highest concentration tested (~ 220 uM)). The fact that the Fgn peptide failed to directly bind anionic M L V s at 100 m M N a C l (Figure 28) and failed to inhibit complement activation indicated that the mechanism by which the Fgn peptide blocked Clq/ l iposome binding was different from the mechanism by which C l q A peptides achieved their inhibition of C l q binding and complement activation. It is known that fibrinogen binds directly to C l q and that there are at least two sites, one each on the collagenous and the globular regions, on C l q which are important for binding (Entwistle, R. and Furcht, L . 1988). Perhaps the Fgn peptide binds to C l q near the site of interaction between C l q and anionic liposomes and thus interferes with Clq/ l iposome binding. Since binding experiments were carried with purified protein and the complement activation assay was carried out in human serum, the fact that complement activation by anionic liposomes was not blocked by the Fgn peptide may be due to alteration of the C l q / F g n peptide interaction in the presence 164 of serum. While a large amount of peptide is required, C l q A peptides of sufficient charge (possibly > 2 and certainly 5 cationic residues) are capable of successfully inhibiting C l q -mediated complement activation by anionic liposomes regardless of the primary sequence or conformation of the peptide. In addition to residues 14-26, other regions of the protein, which have three or four cationic residues within a similar expanse of residues may also be involved. Another cationic region of the C l q A chain containing residues 76-92, at the hinge region of the structure has been shown to be of secondary importance in mediating the binding of C l q to D N A (Jiang, H . et al. 1992b), to C R P trimers (Jiang, H . et al. 1992a) and to S A P trimers (Ying, S . -C. et al. 1993). Residues 76-92 include 4 cationic amino acids and one anionic residue, giving a net charge for the 17 amino acid long region at physiologic p H of plus three. This region is therefore a candidate site on C l q for interaction with anionic liposomes in an electrostatic manner. These experiments characterizing the binding of C l q to anionic liposomes collectively have shown that electrostatics plays a dominant role in the potential for interaction between C l q and liposomes. Clq/ l iposome binding was found to be dependent on liposome surface charge and on the p H and ionic strength of the environment. In addition, competitive C l q binding and complement, activation experiments with peptides at physiologic p H and ionic strength as well as direct measurements of peptide/liposome interactions identified the electrostatic component of the Clq/ l iposome interaction as being important. However, other chemical forces are thought to contribute, specifically to the actual binding energy. Hydrophobic or hydrogen bonding interactions may participate. 165 Previous work by De Kroon (1990) suggested that both electrostatic and hydrophobic forces were involved in the binding of positively charged peptides to CL-containing vesicles. 7 .4 The Effect of PEG-lipid Incorporation on Clq Binding and Complement Activation by Anionic Liposomes Attempts to reduce liposome opsonization have been made focusing on the use of smaller, neutral liposomes and the inclusion of PEG-coupled lipids. Small neutral liposomes, however, are limited in their applications due to small entrapped volumes and particle instability due to a lack of charge stabilization (Woodle, M . and Lasic, D . 1992, Szoka, F . and Papahadjopoulos, D . 1981). Therefore, rendering large and/or charged liposomes less attractive for opsonization would be advantageous. As complement activation plays a significant role in liposome opsonization, particularly for anionic and large liposomes, complement activation must be curtailed in order to increase the circulation lifetime of liposomes. 7.4.1 Inhibition of C l q Binding and Complement Activation by 100 nm Anionic Liposomes Through P E G - l i p i d Incorporation The effect of incorporated P E G specifically on Clq-mediated activation of the complement system by both small and large anionic liposomes was investigated. P E G -lipids were chosen as a non-specific but previously successful method of inhibiting adsorption of al l proteins to liposome surfaces (Senior, J . et al. 1991). Previous studies 166 on the interaction of complement and PEG-containing surfaces have reported conflicting results, however. While Gambotz (1988) reported that immobilization of P E G onto films activated complement in human plasma, M e r r i l l (1992) showed that P E G on a surface protected against complement activation. In our experiments, both with purified protein and in human serum, the incorporation of sufficient PEG-l ipids into anionic liposomes prevented C l q binding and overall complement activation. The inhibition of C l q binding and complement activation, like the reactions themselves, were l ip id concentration dependent. For each of the three PEG-l ipids examined, C H - P E G 6 0 0 , C H - P E G 1 0 0 0 , and D S P E - P E G 2 0 0 0 , increasing the P E G surface density resulted in an increased inhibitory effect on Clq-mediated complement activation. The inhibition of C l q binding and complement activation was also enhanced as the M W of the P E G increased. Optimal conditions, where complement activation by anionic liposomes was completely blocked required 15 mole % of C H - P E G 6 0 0 , 10-15 mole % of CH-PEGiooo, or 5-10 mole % of D S P E - P E G 2 0 0 0 incorporated into 100 nm liposomes (Figure 33). The inhibition of liposome induced Clq-mediated complement activation in human serum by the incorporation of P E G into liposomes has not been previously demonstrated. The only previous studies into C l q / P E G interactions used P E G in solution. Wines and Easterbrook-Smith (1988) and Hack et al, (1981) reported that P E G of M W 6000 - 8000 in solution enhanced the binding of C l q to immune complexes. This enhancement was likely a consequence of the P E G excluded volume effect which increases the effective concentration of C l q in solution. Hence, any concentration-dependent reaction which C l q 167 undergoes, such as binding to immune complexes, is enhanced in the presence of free P E G . When P E G is tethered to a liposome surface with the chain extending into the aqueous medium, a different effect is expected: the excluded volume effect should reduce the interaction of C l q with the surface in the vicinity of the P E G . With respect to overall complement activation, the study by Chonn et al. (1992) investigating the complement inhibiting potential of monosialoganglioside (GM1) , the predecessor to PEG-l ip ids , supports our results. G M 1 is a glycolipid that has a large headgroup with a shielded negative charge. The incorporation of G M 1 into liposomes contributes a surface glycocalyx to the liposome membrane and increases the circulation half-life of liposomes (Allen, T. et al. 1989; Gabizon, A . and Papahadjopoulos, D . 1988). Chonn et al. (1992) reported that addition of 10 mole % G M 1 to complement activating liposomes inhibited complement activation in both guinea pig and human serum. In this same study, incorporation of G M 1 also resulted in a decrease in total protein adsorbed onto the liposomes. These liposomes expressed enhanced circulation lifetimes in mice. This enhanced circulation time was correlated to the overall decrease in protein binding but may also be related to the suppression of complement activation. In a separate study, G M 1 was also found to inhibit complement-dependent phagocytosis of liposomes (Wassef, N . and A l v i n g , C . 1993). The observation that the inhibitory effect of PEG-l ipids on complement activation is P E G concentration and M W dependent in vitro parallels findings in mice, rats, and rabbits where circulation lifetimes and biodistribution of PEG-containing liposomes have been similarly shown to depend on P E G surface density and chain length. Klibanov et al. 168 (1991), Maruyamak et al. (1992), Litzinger and Huang (1992) and Torchi l in et al. (1992) among others have shown that liposome circulation times increased with increasing P E G surface density. The inhibition of liposome uptake by macrophages was also shown to be directly dependent upon the P E - P E G concentration (Allen, T. et al. 1991). Recognizing that the assays differ in the context of species, system, and liposome composition, it is still noteworthy that this study by Al l en et al. (1991) using P E - P E G 1 9 0 0 to inhibit the uptake of liposomes by bone marrow macrophages provided P E - P E G 1 9 0 0 dose information which exhibits a striking resemblance to our results. Comparing the two studies, while incorporation of D S P E - P E G 1 9 0 0 . 2 0 0 0 into 100 nm liposomes at 2.5 mole % resulted in little effect on complement activation and no effect on macrophage uptake, addition of 5 mole %, resulted in obvious inhibitory effects for both complement activation (50-90%) and macrophage uptake (80%). Addit ion of 7-7.5 mole % P E - P E G showed a slight enhancement over the already high inhibitory effect in both assays and 10 mole % showed complete abrogation of complement activation and a 95 % inhibition of macrophage uptake. Opsonization of particles by complement promotes uptake by macrophages. Although no direct evidence is provided here, the hypothesis that the inhibition of complement activation contributes to the inhibition of macrophage uptake is not unreasonable. Complement-dependent macrophage and neutrophil uptake (C3b-mediated) has been shown in other studies (Wassef, N . and Alv ing , C . 1993, Scieszka, J . et al. 1991, L i u , D . et al. 1995 and Harashima, H . et al. 1994). Two studies using a perfused rat 169 liver model showed that the enhanced hepatic uptake of liposomes was mediated by the interaction between deposited C3 fragments on the liposomes and liver phagocytic C3 receptors (Matsuo, H . et al. 1994; Harashima, H . et al. 1994). Wi th respect to P E G M W , studies have shown enhancement of circulation lifetimes for liposomes in mice and rats upon increasing P E G chain length (Maruyama, K . et al. 1991 and 1992; M o r i , A . et al. 1991; Woodle, M . et al. 1992; and Vertut-Doi, A . et al. 1996). Circulation times increased as P E G size increased from 750 to 1000 to 2000. The effect of further increasing the M W to 5000 is not clear as M o r i found enhancement of circulation time but Woodle found no enhanced effect over that of P E -PEG 2 0oo- Essentially, between 5 and 10 mole% of P E - P E G 1 0 0 0 . 5 0 0 0 appears to give the best results in terms of increasing blood circulation times (Woodle, M . and Lasic , D . 1992). The results presented here are in agreement for the optimal range of P E G - l i p i d required to inhibit complement activation. Since liposomes with a large negative surface charge were used in this work, it should be noted that several studies have shown that the ability of P E - P E G to increase circulation times in vivo is independent of the surface charge of the liposomes (Klibanov, A . et al. 1991; Woodle, M . et al. 1992). These studies included up to 15 mole % of P G in the liposome mixture. We have now demonstrated that PEG-l ip ids are effective at blocking complement activation even in the presence of 20 mole % C L (i.e. 40 mole % of phosphate). 170 7.4.2 The Effect of the P E G - l i p i d Anchor With respect to the type of l ipid anchor used to add P E G to liposomal formulations, under the conditions of our assays, the cholesterol anchor and the P E anchor behave similarly, with the major differences in effect being attributable to the P E G M W and concentration. Similarly, in a study by Yoshioka (1991) comparing C H - P E G and P E - P E G lipids, cholesterol-anchored P E G was found to be equally as efficient as P E anchored P E G at preventing liposome aggregation in 83 % human plasma. In a recent study by Vertut-Doi et al. (1996), C H - P E G 4 4 0 0 was found to be superior to P E - P E G 5 0 0 0 in its ability to reduce the binding and uptake of liposomes by mouse macrophages for at least 4 hours at 37°C. In contrast, the study by A l l e n et al. (1991) in which radiolabelled liposomes containing 6.25 mole % different l i p i d - P E G i 9 0 0 derivatives were assessed for circulation lifetime, showed that C H - P E G 1 9 0 0 had only a moderate ability to decrease the uptake of liposomes by the R E S in a mouse model. While C H - P E G 1 9 0 0 liposomes had blood levels as high as the levels observed for D S P E - P E G 1 9 0 0 liposomes at 2 hours post injection, at 24 hours, the C H - P E G 1 9 0 0 liposome level in the blood was much lower than the level of D S P E - P E G 1 9 0 0 liposomes. Although no direct evidence was provided, it was hypothesised that the C H - P E G derivative exchanged from the bilayer, similar to the way in which cholesterol exchanges. To address whether the C H - P E G derivatives exchanged from the bilayer, an exchange assay in human serum was carried out on E P C : C H : C L : C H - P E G _ 1 4 0 0 at 35:30:20:15 mole %. After an initial loss of ~ 35% of the cholesterol-anchored P E G _ 1 4 n f l 171 from the membrane, the remaining 65% was maintained in the liposomes for at least 24 hours at 37°C in 50% human serum (Figure 31). The time independence of this C H - P E G loss was strikingly different to either cholesterol or P E G - P E conjugate exchange from donor vesicles in buffer to acceptor vesicles, or to the loss of P O P E - P E G 2 0 0 0 in mouse serum, which are all roughly exponentially time dependent (Rodrigueza, W . 1994; Silvius, J . and Zuckermann, M . 1993; Holland, J . et al. 1996; Parr, M . et al. 1994). Wi th regard to cholesterol exchange rates, Rodrigueza (1994) reported that rates of cholesterol flux from 100 nm diameter donor vesicles to cholesterol-free acceptor vesicles in vitro were such that at two hours, 15 % of the cholesterol was lost from the donors and by 24 hours, 60 % of the cholesterol was lost. P E - P E G conjugates were also found to diffuse from donor to P E - P E G free acceptor vesicles. The rate of conjugate exchange was inversely related to the acyl chain length of the P E anchor (Silvius, J . and Zuckermann, M . 1993; Holland, J . et al. 1996). Testing liposomes with incorporated PEG-l ipids in serum for P E G - l i p i d loss using the same exchange assay that was used in this study, Parr et al. (1994) reported that for 100 nm liposomes containing 5 mole % DSPE-PEG 2 0 oo» m e P E G to liposome marker ratio was maintained at 88 -100 % for at least 24 hours in 50% normal mouse serum. P O P E -P E G 2 0 0 0 , however, was lost from the liposome membrane, such that at 24 hours, the P E G to liposome marker ratio dropped to 55 % (Parr, M . et al. 1994). This loss of P O P E -P E G 2 0 0 0 from the liposomes was considered to be due to exchange of the entire P O P E -P E G molecule. Parr et al. (1994) also demonstrated that the chemical breakdown of the PEG- l ip id succinate linkage may account for loss of P O P E - P E G from the liposomes. The 172 time profile for the loss of P O P E - P E G 2 0 0 0 from the liposomes was very different from our C H - P E G 1 0 0 0 profile discussed above. The mechanism for the immediate loss of C H - P E G from the liposomes in the presence of human serum is not known. The chemical or enzymatic breakdown of the C H - P E G linkage is considered unlikely as ether linkages are generally stable. Differences in the concentrations of various lipoproteins in human serum compared to mouse serum may be involved. A simple explanation for the immediate C H - P E G loss is that this portion of the C H - P E G was associated with but not incorporated into the liposome membranes. In the presence of human serum, the unincorporated C H - P E G dissociated from the liposomes. This hypothesis is supported by two separate observations. First, the presence of foam observed during the extrusion of liposomes containing large amounts of C H - P E G 1 4 0 0 (eg. 10 or 15 mole %) indicated that not all of the C H - P E G ] 4 0 0 was incorporated. Second, the amount of C H - P E G lost from liposomes in the presence of human serum was proportional to the initial C H - P E G concentration associated with the liposomes (Table 12). For liposomes with 5 mole % C H - P E G 1 4 0 0 , approximately 15% of the C H - P E G was lost immediately. N o additional C H - P E G loss from the liposomes in 50 % human serum for up to 24 hours at 37°C. The maintenance of 85% of the C H - P E G in liposomes composed of 5 mole % C H - P E G 1 4 0 0 is similar to the D S P E - P E G maintenance observed by Parr et al. (1994). For the assays conducted in this thesis where PEG-containing liposomes were incubated in diluted human serum (1/60 to 1/10) for 20 minutes at R T to 1.5 hours at 37°C, the loss of C H - P E G from liposomes is likely not an issue. The fact that D S P E -173 PEG 2 0oo w a s somewhat more effective on a molar basis at inhibiting C l q binding and complement activation than C H - P E G 1 0 0 0 is consistent with the higher molecular weight headgroup on the PE-anchored species producing a greater exclusion of C l q and suggests that the difference in l ip id anchor had no effect in vitro. The inhibition data reported here indicate that C H - P E G may be more successful at providing protection to liposomes in humans than previously considered. 7.4.3 Inhibition of C l q Binding and Complement Activation by 240 nm Anionic Liposomes Through P E G - l i p i d Incorporation The ability of incorporated PEG-l ip ids to inhibit Clq-mediated complement activation by anionic liposomes extended to larger liposomes as wel l . Even on larger liposomes (240 nm), the addition of PEG-l ipids resulted in blockage of C l q binding and complement activation. In fact, P E G - l i p i d incorporation in larger liposomes had a greater inhibitory effect on Clq-mediated complement activation than the inhibition seen for smaller liposomes. While the mechanism for the enhanced effect of P E G on larger liposomes is not known, the protective effect of PEG-l ipids on the circulation times of large liposomes (200-300 nm) has been previously reported (Maruyama, K . et al. 1992; A l l en , T . et al. 1991; Litzinger, D . et al. 1994). In addition, Vertut-Doi et al. (1996) reported that the incorporation of C H - P E G into very large liposomes (1.5 jum) inhibited macrophage uptake. Thus, although an increase in liposome size predisposes for a more rapid clearance, larger liposomes with P E G incorporated were still capable of remaining in the 174 blood for extended times. In contrast, Klibanov et al. (1991) reported that liposomes containing P E - P E G 5 0 0 0 that are larger than 200 nm in diameter were cleared from the circulation very rapidly and accumulated in the spleen. Hence, there is some debate over the effect of liposome size on clearance and the cut-off size for liposomes that can be saved by P E G . Results presented in this thesis support the observations that PEG-l ipids are capable of protecting liposomes of up to 240 nm diameter. This last section has demonstrated that the incorporation of P E G into liposomes, whether by the addition of P E or CH-anchored P E G , results in inhibition of liposome/complement interactions for both small and larger liposomes (~ 240 nm). These P E G incorporation effects correlate well with the trends observed in vivo regarding circulation half-lives of liposomes. Consequently these results support the view that complement plays a major role in the recognition and clearance of anionic liposomes. 7 . 5 Future Directions: Testing the Model Our investigation into the nature of C l q binding to anionic liposomes has led us to suspect that the presence of C l q near the liposome surface causes the formation of membrane domains enriched in anionic phospholipids. To test this hypothesis, N B D -labelled phosphatidylglycerol ( N B D - P G ) could be incorporated into large unilamellar liposomes (5-15 um) and the fluorescence intensity across the liposome surface would be visualized directly by fluorescence microscopy (Haverstock, D . and Glaser, M , 1987; 175 Yang, L . and Glaser, M . 1995). Liposomes alone would have a uniform fluorescence intensity across the surface. When l ipid domain formation is induced, patches of fluorescence would be visualized due to clustering of N B D - P G . Incubating C l q with anionic liposomes followed by fluorescence microscopy would allow for the determination of whether anionic phospholipid domains are formed. A positive control for domain formation would be N B D - P G containing liposomes incubated with polylysine. In addition, C l q can be labelled with the fluorophore, acrylodan according to the method of Yang and Glaser (1995). Using labelled C l q in the liposome incubation would allow for assessment of whether C l q colocalizes with the l ipid domains. 176 Chapter 8: Summary The properties required for Clq-mediated complement activation by anionic liposomes in vitro and the modulation of this activation by poly(ethylene glycol) (PEG) incorporation into liposomes have been investigated. Fundamental measurements of the initiation of complement activation via C l q binding to anionic liposomes were made in the first two results chapters of this thesis. Electrostatics were shown to be critical in determining C l q saturation binding levels. When conditions were such that the electrostatic attraction between C l q and liposomes was enhanced, C l q saturation binding levels increased. For instance, while neutral liposomes failed to bind C l q or activate complement under the assay conditions used in this thesis, anionic liposomes bound C l q in an acidic l ipid concentration-dependent manner. In addition, Clq/ l iposome binding was found to be dependent on the p H and ionic strength of the environment. As either the p H or the ionic strength were decreased, C l q saturation binding levels increased. A n increase in the C l q saturation binding value indicates that the number of binding sites on the liposomes has increased. Since saturation levels increased when the electrostatic component of the interaction increased, it appears that the formation of binding sites is electrostatically-mediated. The first two parts of the model proposed in this thesis to describe C l q binding to anionic liposomes deal with this electrostatic component. It is proposed that C l q is recruited to the surface of the liposomes by electrostatic attraction. Then, the proximity of positively charged residues on C l q at the 177 membrane is thought to induce the movement and concentration of anionic phospholipids on the outer bilayer leaflet adjacent to C l q thereby creating binding sites. The last part of the model deals with the actual intrinsic binding of C l q to the created binding sites on the liposomes. Because the bulk or apparent association constants were independent of the electrostatic component of the Clq/ l iposome interaction, the actual intrinsic binding of C l q to the liposome binding sites is believed to be chemically-mediated. Possibly hydrophobic interactions and/or hydrogen bonding are involved. Very little C l q bound to anionic liposomes under physiologic conditions. Wi th respect to the model, this was because at neutral p H , C l q has very little i f any net positive charge. Consequently, C l q recruitment to the liposome surface by electrostatic attraction is minimal. F rom a biological perspective, only a small amount of C l q binding is required to activate the classical pathway of complement. Once activated, the pathway is amplified at the level of the C3 convertase resulting in a large amount of C3b generation. Indeed, even though low C l q binding saturation values were obtained, these same anionic liposomes activated complement in human serum. Al so from a biological perspective, it seems likely that the antibody-independent binding of C l q would be lower than C l q binding in the presence of immunoglobulins. Marjan et al. (1994) reported that anionic liposomes activated complement less in serum depleted of immunoglobulins than in normal human serum. However, experiments carried out in this study in antibody-conserved Clq-depleted serum with C l q added back showed that a similar level of C l q binding to anionic liposomes occurred in the presence of serum as in the absence of serum. 178 The aggregation state of C l q may play a biologically important role in determining the level of C l q binding to anionic liposomes. Various observations in these studies, such as the l ipid concentration and storage time effects, suggest that as C l q becomes more aggregated, the saturation level of C l q binding increases. Other investigators have speculated that aggregated rather than monomeric C l q is the ligand of biological interest (Bordin, S. et al. 1990). A s vesicle size is one of the major determinants in the fate of intravenously administered liposomes (Juliano, R. and Stamp, D . 1975; Mayer , L . et al. 1989), the effect of liposome size on Clq-mediated complement activation in human serum was also investigated. Using the C l q E L I S A and hemolytic assay in diluted human serum, 240 nm anionic liposomes were found to bind more C l q and to activate complement more readily than 100 nm vesicles. Using the C l q equilibrium binding assay with purified C l q in buffer, C l q binding to unextruded M L V s was far greater (15 times more) than C l q binding to extruded 240 nm vesicles. The facts that negatively charged liposomes activate complement in an acidic l ipid concentration-dependent way and that larger liposomes activate complement more readily than small liposomes in vitro correlate well with the trends observed in vivo regarding circulation half-lives of liposomes. Larger liposomes or liposomes with a negative surface charge are cleared far more rapidly than neutral and small liposomes (Juliano, R. and Stamp, D . 1975). The goal of the next part of these studies was to assess one highly cationic region of C l q for its involvement in Clq/l iposomes interactions. Peptides containing residues 14-179 26 of the C l q A polypeptide were capable of inhibiting C l q binding and complement activation by anionic liposomes. While the inhibitory capacity of these cationic peptides did not require sequence or conformation specificity, the amount of positive charge on the peptides was a determining factor. When present in excess, peptides with 5 cationic residues inhibited C l q binding and complement activation. C l q peptides with only 2 cationic residues did not have this capability. Consequently, results of the competitive C l q binding and complement activation experiments with peptides as well as direct measurements of peptide/liposome interactions again identified the electrostatic component of the Clq/ l iposome interaction as being crucial. These studies indicated but could not confirm that the C l q A 1 4 . 2 6 region of C l q plays a role in C l q binding to anionic liposomes. Other regions of the protein that contain cationic residues may also be involved. The goal of the final section of this thesis was to determine the effect of P E G - l i p i d incorporation on C l q binding and complement activation by anionic liposomes. A l l of the PEG-l ipids tested, P E - P E G 2 0 0 0 , C H - P E G 6 0 0 , and C H - P E G 1 0 0 0 , were capable of inhibiting C l q binding in buffer and in human serum and of inhibiting complement activation by liposomes in serum. The inhibitory capacity of the PEG-l ip ids was dependent upon both the P E G chain length ( M W ) and the P E G surface density. The ability of PEG-l ipids to inhibit complement activation and the dependence on P E G M W and concentration in vitro parallels findings in cel l culture and in animal models that show a decrease in liposome uptake by phagocytic cells and an increase in the circulation life-times of liposomes with incorporated P E G . 180 Measuring C l q binding and complement activation by liposomes has provided insight into the requirements for direct C l q binding and complement activation on a model cell . In addition, these data support the view that complement activation is an important factor in the clearance of anionic liposomes from the circulation and that modulating complement activation by liposomes decreases liposome opsonization and allows for increased liposome circulation times. 181 Bibliography Abuchowski, A . , V a n Es, T . , Palczuk, N . C . and Davis, F . F . Alterations of immunological properties of bovine serum albumin by covalent attachment of polyethylene glycol . J.Biol.Chem. 252:3578-3581, 1977. Ahkong, Q . F . , Fisher, D . , Tampion, W . and Lucy , J . A . Mechanisms of cell fusion. Nature 253:194-195, 1975. Al l en , T . M . , Austin, G . A . , Chonn, A . , L i n , L . and Lee, K . C . Uptake of liposomes by cultured mouse bone marrow macrophages: influence of liposome composition and size. Biochim.Biophys.Acta 1061:56-64, 1991a. A l l e n , T . M . and Everest, J . M . Effect of liposome size and drug release properties on pharmacokinetics of encapsulated drug in rat. J.Pharm.Exp. Therapeutics 226:539-544, 1983. Al l en , T . M . , Hansen, C , Mart in , F . , Redemann, C . and Yau-Young, A . Liposomes containing synthetic l ipid derivatives of poly(ethylene glycol) show prolonged circulation half-lives in vivo. Biochim.Biophys.Acta 1066:29-36, 1991b. Al l en , T . M . , Hansen, C , and Rutledge, J . Liposomes with prolonged circulation times: factors affecting uptake by reticuloendothelial and other tissues. Biochim.Biophys.Acta 981:27-35, 1989. A l v i n g , C . R . , Richards, R . L . and Guirguis, A . A . Cholesterol-dependent human complement activation resulting in damage to liposomal model membranes. J. Immunol. 118:342-347, 1977. Antimisiaris, S . G . , Jayasekera, P . and Gregoriadis, G . Liposomes as vaccine carriers: Incorporation of soluble and particulate antigens in giant vesicles. J.Immunol.Meth. 166:271-280, 1993. Arlaud, G . J . , Thielens, N . M . and Illy, C . Arrangement of the C I complex of complement. Biochem.Soc.Trans. 18:1148-1151, 1990. Atha, D . H . and Ingham, K . C . Mechanisms of precipitation of proteins by polyethylene glycols: Analysis in terms of excluded volume theory. J.Biol.Chem. 258:5710-5715, 1983. Bakker-Woudenberg, I . A . J . M . , Lokerse, A . F . , Ten Cate, M . T . , Mouton, J . W . , Woodle, M . C . , and Storm, G . Liposomes with prolonged blood circulation and selective localization in Klebsiella pneumoniae infected lungs. J. Infect. Dis., 168: 164-171, 1993. 182 Bally, M . B . , Nayar, R . , Masin , D . , Hope, M . J . , Cul l i s , P . R . , and Mayer , L . D . Liposomes with entrapped doxorubicin exhibit extended blood residence times. Biochim.Biophys.Acta 1023:133-139, 1990. Bangham, A . D . and Home, R . W . Negative staining of phospholipids and their structural modification by surface active agents as observed in the electron microscope. J.Mol.Biol 8:660-668, 1964. Barenholz, Y . , Amsselem, S. and Lichtenberg, D . A new method for preparation of phospholipid vesicles (liposomes)-French press. FEBS Lett. 99:210-214, 1979. Beauchamp, C O . , Gonias, S . L . , Menapace, D . P . and Pizzo, S . V . A new procedure for the synthesis of polyethylene glycol-protein adducts: effects on function, receptor recognition, and clearance of superoxide dismutase, lactoferrin, and a2-macroglobulin. Anal.Biochem. 131:25-33, 1983. Bental, M . , Wilschut, J . , Scholma, J . , and N i r , S. Ca 2 + - induced fusion of large unilamellar phosphatidylserine/cholesterol vesicles. Biochim.Biophys.Acta 898:239-247, 1987. Bergers, J . J . , Vingerhoeds, M . H . , van Bloois, L . , Herron, J . N . , Janssen, L . H . M . , Fischer, M . J . E . and Crommelin, D . J . A . The role of protein charge in protein-lipid interactions. pH-dependent changes of the electrophoretic mobility of liposomes through adsorption of water-soluble, globular proteins. Biochemistry 2:4641-4649, 1993. Beschiaschvili, G . and Seelig, J . Meli t t in binding to mixed phosphatidylglycerol/phosphatidylcholine membranes Biochemistry. 29:52-58. 1990. Bing, D . H . , Almeda, S., Isliker, H . , Lahav, J . and Hynes, R . O . Fibronectin binds to the C l q component of complement. Proc.Natl.Acad.Sci. USA 79:4198-4201, 1982. Blume, G . and Cevc, G . Liposomes for the sustained drug release in vivo. Biochim.Biophys.Acta 1029:91-97, 1990. Blume, G . and Cevc, G . Molecular mechanism of the l ipid vesicle longevity in vivo. Biochim.Biophys.Acta 1146:157-168, 1993. Bobak, D . A . , Gaither, T . A . , Frank, M . M . and Tenner, A . J . Modulation of F c R function by complement: subcomponent C l q enhances the phagocytosis of IgG opsonized targets by human monocytes and culture derived macrophages. J.Immunol. 138:1150-1156, 1987. 183 Bohnsack, J . F . , Tenner, A . J . , Laurie, G . W . , Kleinman, H . K . , Mart in , G . R . and Brown, E . J . The C l q subunit of the first component of complement binds to laminin: a mechanism for the deposition and retention of immune complexes in basement membrane. Proc.Natl.Acad.Sci.USA 82:3824-3828, 1985. Bordin, S., Ghebrehiwet, B . and Page, R . C . Participation of C l q and its receptor in adherence of human diploid fibroblast. J.Immunol. 145:2520-2526, 1990. Bordin, S. and Page, R . C . Detection of a high-affinity binding site for the globular head regions of the C l q complement protein on a human diploid fibroblast subtype. Mol.Immunol. 26:677-685, 1989. Bowden, D . W . , Ris ing, M . , Akots, G . , Myles , A . and Broeze, R . J . Homogeneous, liposome-based assay for total complement activity in serum. Clin. Chem. 32:275-278, 1986. Budzko, D . B . , Lachmann, P . J . and McConne l l , I. Activation of the alternative complement pathway by lymphoblastoid cell lines derived from patients with Burkitt 's lymphoma and infectious mononucleosis. Cell Immunology 22:98-109, 1976. Burton, D . R . Immunoglobulin G : functional sites. Mol.Immunol. 22:161-206, 1985. Burton, D . R . , Boyd , J . , Brampton, A . D . , Easterbrook-Smith, S .B . , Emanuel, E . J . , Novotny, J . , Rademacher, T . W . , van Schravendijk, M . R . , Sternberg, M . J . E . and Dwek, R . A . The C l q receptor site on immunoglobulin G . Nature 288:338-344, 1980. Calcott, M . A . and Muller-Eberhard, H . J . C l q Protein of human complement. Biochemistry 11:3443-3450, 1972. Chapman, D . Protein-lipid interactions in model and natural membranes. In: Biological Membranes. Volume 4, edited by Chapman, D . London: Academic Press, 1982, p. 179-224. Chen, A . , Gaddipati, S., Hong, Y . , Volkman, D . J . , Peerschke, E . I . B . and Ghebrehiwet, B . Human T cells express specific binding sites for C l q : Role in T cell activation and proliferation. J.Immunol. 153:1430-1440, 1994. Chen, C . Y . and Schullery, S .E . Ge l filtration of egg phosphatidylcholine vesicles. J.Biochem.Biophys.Meth. 1:189-192, 1979. Chonn, A . and Cul l i s , P . R . Recent advances in liposomal drug-delivery systems. Current Opinion in Biotech. 6:698-708, 1995. 184 Chonn, A . , Cul l i s , P . R . and Devine, D . V . The role of surface charge in the activation of the classical and alternative pathways of complement by liposomes. J.Immunol. 146:4234-4241, 1991. Chonn, A . , Semple, S .C . and Cul l i s , P . R . Separation of large unilamellar liposomes from blood components by a spin column procedure: towards identifying plasma proteins which mediate liposome clearance in vivo. Biochim.Biophys.Acta 1070:215-222, 1991. Chonn, A . , Semple, S .C . and Cul l i s , P . R . Association of blood proteins with large unilamellar liposomes in vivo. J.Biol.Chem. 267:18759-18765, 1992. Claus, D . R . , Osmand, A . P . and Gewurz, H . Radioimmunoassay of human C-reactive protein and levels in normal sera. J.Lab. Clin.Med. 87:120-128, 1976. Claus, D . R . , Siegel, J . , Petras, K . , Skor, D . , Osmand, A . P . and Gewurz, H . Complement activation by interaction of poly anions and polycations. J.Immunol. 118:83-87, 1977. Cohen, S. The requirement for the association of two adjacent rabbit gamma-G antibody molecules in the fixation of complement by immune complexes. J.Immunol. 100:407-413, 1968. Comis, A . and Easterbrook-Smith, S .B. Binding of the complement component C l q by rat adipocyte membranes. Mol.Immunol. 22:857-861, 1985. Comis, A . and Easterbrook-Smith, S .B. Inhibition of serum complement haemolytic activity by l ipid vesicles containing phosphatidylserine. FEBS Lett. 197:321-327, 1986. Comiskey, S.J. and Heath, T . D . Serum-induced leakage of negatively charged liposomes at nanomolar l ip id concentrations. Biochemistry 29:3626-3631, 1990. Conradie, J . D . , Volanakis, J . E . and Stroud, R . M . Evidence for a serum inhibitor of C l q . Immunochemistry 12:967-971, 1975. Cooper, N . R . The classical complement pathway: Activation and regulation of the first complement component. Advances in Immunology 37:151-216, 1985. Cooper, N . R . , Jensen, F . C . , Welsh, R . M . J r and Oldstone, M . B . A . Lysis of R N A tumour viruses by human serum: direct antibody-independent triggering of the classical complement pathway. J. Exp. Med. 144:970-984, 1976. Cooper, N . R . and Morr ison, D . C . Binding and activation of the first component of 185 human complement by the l ipid A region of lipopolysaccharides. J.Immunol. 120:1862-1868, 1978. Cul l is , P . R . and Hope, M . J . Physical properties and functional roles of lipids in membranes. In: Biochemistry of lipids and membranes, edited by Vance, D . E . and Vance, J . E . Menlo Park: The Benjamin/Cummings Publishing Company, Inc., 1985, p. 25-72. Cul l is , P . R . , Hope, M . J . , De Kruijff, B . , Verklei j , A . J . and Tilcock, C . P . S . Structural Properties and functional roles of phospholipids in biological membranes. In: Phospholipids and cellular regulation, edited by Kuo , J .F . Boca Raton, Florida: C R C Press, 1985, p. 1-49. Cunningham, C M . , Kingzette, M . , Richards, R . L . , A l v i n g , C . R . , Lint , T . F . and Gewurz, H . Activation of human complement by liposomes: a model for membrane activation of the alternative pathway. J.Immunol. 122:1237-1242, 1979. De Kroon, A . I . P . M . , Soekarjo, M . W . , De Gier, J . and De Kruijff, B . The role of charge and hydrophobicity in peptide-lipid interaction. Biochemistry 29:8229-8240, 1990. Deamer, D . W . and Nichols, J . W . Proton-hydroxide permeability of liposomes. Proc.Natl.Acad.Sci.USA 80:165-168, 1983. deBeer, F . , Baltz, M . , Holford, S., Feinstein, A . and Pepys, M . Fibronectin and C4-binding protein are selectively bound by aggregated amyloid P component. / . Exp. Med. 154:1134-1139, 1981. De Kruijff, B . and Cul l i s , P . R . The influence of poly-L-lysine on phospholipid polymorphism. Evidence that electrostatic polypeptide-phospholipid interactions can modulate bilayer-non-bilayer transitions. Biochim.Biophys.Acta 601:235-240, 1980. De Kruijff, B . , V a n Zoelen, E . J . J , and V a n Deenen, L . L . M . Glycophorin facilitates the transbilayer movement of phosphatidylcholine in vesicles. Biochim.Biophys.Acta 509:537-542, 1978. Devine, D . V . The regulation of complement on cell surfaces. Trans.Med.Rev. 5:123-131, 1991. Devine, D . V . and Bradley, A . J . The complement system in liposome clearance: can complement deposition by inhibited ? Adv.Drug Delivery Rev. in press 1997. Devine, D . V . and Marjan, J . M . J . The role of immunoproteins in the survival of 186 liposomes in the circulation. Critical Reviews in therapeutic drug carrier systems 14:105-131, 1997. Devine, D . V . , Wong, K . , Serrano, K . , Chonn, A . and Cul l i s , P . R . Liposome-complement interactions in rat serum: implications for liposome survival studies. Biochim.Biophys.Acta 1191:43-51, 1994. DiCamel l i , R . , Potempa, L . A . , Siegel, J . , Suyehira, L . , Petras, K . and Gewurz, H . Binding reactivity of C-reactive protein for polycations. J.Immunol. 125:1933-1938, 1980. Dijkstra, J . , van Galen, M . and Scherphof, G . Influence of liposome charge on the association of liposomes with Kupffer cells in vitro. Effects of divalent cations and competition with latex particles. Biochim.Biophys.Acta 813:287-297, 1985. Duncan, A R . and Winter, G . The binding site for C l q on IgG. Nature 332:738-740, 1988. Entwistle, R . A . and Furcht, L . T . C l q component of complement binds fibrinogen and fibrin. Biochemistry 27:507-512, 1988. Fearon, D . T . Regulation by membrane sialic acid of beta lH-dependent decay-dissociation of amplification C3 convertase of the alternative complement pathway. Proc.Natl.Acad.Sci. USA 75:1971-1975, 1978. Fidler, I .J . , Fogler, W . E . , Kleinerman, E . S . , and Saiki , I. Abrogation of species specificity for activation of tumoricidal properties in macrophages by recombinant mouse or human interferon-gamma encapsulated in liposomes. J.Immunol. 135:4289-4296, 1985. Fiske, C . H . and Subbarow, Y . The colorimetric determination of phosphorus. J.Biol.Chem. 66:375-400, 1924. Fujiwara, M . , Baldeschewieler, J . D . and Grubbs, R . H . Receptor-mediated endocytosis of poly (acrylic acid)-conjugated liposomes by macrophages. Biochim.Biophys.Acta 1278:59-67, 1996. Funato, K . , Yoda , R . and Kiwada, H . Contribution of complement system on destabilization of liposomes composed of hydrogenated egg phosphatidylcholine in rat plasma. Biochim.Biophys.Acta 1103:198-204, 1992. Gaber, M . H . , Hong, K . , Huang, S . K . and Papahadjopoulos, D . Thermosensitive sterically stabilized liposomes: formulation and in vitro studies on mechanism of doxorubicin release by bovine serum and human plasma. Pharm.Res. 187 12:1407-1416, 1995. Gabizon, A . , Isacson, R . , Libson, E . , Kaufman, B . , Uzie ly , B . , Catane, R . , Ben-Dor, C . G . , Rabello, E . , Cass, Y . , Peretz, T. et al . Cl in ica l studies of liposome-encapsulated doxorubicin. Acta Oncol. 33:779-786, 1994. Gabizon, A . and Papahadjopoulos, D . Liposome formulations with prolonged circulation time in blood and enhanced uptake by tumours. Proc.Natl.Acad.Sci. USA 85:6949-6953, 1988. Gabizon, A . and Papahadjopoulos, D . The role of surface charge and hydrophillic groups on liposome clearance in vivo. Biochim.Biophys.Acta 1103:94-99, 1992. Gambotz, W . R . PhD. Thesis, Centre for Bioengineering, University of Washington. Poly (ethylene oxide) surfaces: synthesis, characterization and biological interaction studies, 1988. Gambotz, W . R . , Guanghui, W . , Horbett, T . A . and Hoffman, A . S . Protein Adsorption to and elution from polyether surfaces. In: Biomedical Applications of Polyethylene glycol chemistry, edited by Harris, J . M . New York : Plenum Press, 1991. Gewurz, H . , Shin, H . S . and Mergenhagen, S .E . Interactions of the complement system with endotoxin lipopolysaccharide. Consumption of each of the six terminal complement components. J. Exp. Med. 128:1049-1057, 1968. Ghebrehiwet, B . and Muller-Eberhard, H . J . Lysis of Clq-coated chicken erythrocytes by human lymphoblastoid cell lines. J.Immunol. 120:27-32, 1978. Giclas, P . C . , Ginsberg, M . H . and Cooper, N . R . Immunoglobulin G independent activation of the classical complement pathway by monosodium urate crystals. J.Clin.Invest. 63:759-764, 1979. Golander, C . G . , Jonsson, S., Vladkova, T . , Stenius, P . and Eriksson, J . C . Preparation and adsorption properties of photopolymerized hydrophillic films containing n-vinylpyrrolidone ( N V P ) , acrylic acid ( A A ) or ethyleneoxide (EO) units as studies by E S C A . Colloids and Surfaces 21:149-165, 1986. Griffin, F . M . Opsonization. In: Comprehensive Immunology, edited by Good, R . A . and Day, S .B . New York : Plenum Press, 1977, p. 85-113. Hack, C . E . , Eerenberg-Belmer, A . J . M . , Hannema, A . J . , Out, T . A . and Aalberse, R . C . Polyethylene glycol enhances the binding of C l q to circulating immune complexes. J.Immunol.meth. 44:211-221, 1981. 188 Hamazaki, H . Ca 2 + -mediated association of human serum amyloid P component with heparin sulfate and dermatan sulfate. J.Biol. Chem. 262:1456-1460, 1987. Harashima, H . , Ochi , Y . and Kiwada, H . Kinetic modelling of liposome degradation in serum: effect of size and concentration of liposomes in vitro. Biopharm.Drug Disp. 15:217-225, 1994a. Harashima, H . , Sakata, K . , Funato, K . and Kiwada, H . Enhanced hepatic uptake of liposomes through complement activation depending on the size of the liposomes. Pharm.Res. 11:402-406, 1994b. Harris, J . M . Introduction to biotechnical and biomedical applications of poly (ethylene glycol). In: Biotechnical and Biomedical Applications of Poly (Ethylene glycol) Chemistry, edited by Harris, J . M . New York : Plenum Press, 1991. Haverstock, D . M . and Glaser, M . Visualization of Ca 2 + - induced phospholipid domains. Proc. Natl. Acad. Sci. USA 84: 4475-4479, 1987. Hermans, J . Excluded volume theory of polymer-protein interactions based on polymer chain statistics. J.Chem.Phys. 77:2193-2203, 1982. Hernandez-Casselles, T . , Vi l la la in , J. and Gomez-Fernandez, J . C . Influence of liposome charge and composition on their interaction with human blood serum proteins. Mol.Cell.Biochem. 120:119-126, 1993. Hesketh, T . R . , Dourmashkin, R . R . , Payne, S . N . , Humphrey, J . H . and Lachmann, P . J . Lesions due to complement in l ipid membranes. Nature 233:620-623, 1971. Heusser, C . , Boesman, M . , Nordin, J . H . and Isliker, H . Effect of chemical and enzymatic radioiodination on in vitro human C l q activities. J.Immunol. 110:820-828, 1973. Hicks , P . S . , Saunero-Nava, L . , D u Clos, T . W . and M o l d , C . Serum amyloid P component binds to histones and activates the classical complement pathway. J.Immunol. 149:3689-3694, 1992. Holland, J . W . , H u i , C , Cul l i s , P . R . , and Madden, T . D . Polyethylene glycoi)-lipid conjugates regulate the calcium-induced fusion of liposomes composed of phosphatidylethanolamine and phosphatidylserine. Biochemistry 35:2618-2624, 1996. Hope, M . J . , Bal ly , M . B . , Webb, G . and Cul l is , P . R . Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. 189 Biochim.Biophys.Acta 812:55-65, 1985. Hsu, M . J . and Juliano, R . L . Interactions of liposomes with the reticuloendothelial system. II: Nonspecific and receptor-mediated uptake of liposomes by mouse peritoneal macrophages. Biochim.Biophys.Acta 720:411-419, 1982. Huang, C . H . Studies on phosphatidylcholine vesicles: formation and physical characteristics. Biochemistry 8:344-352, 1969. Hughes-Jones, N . C . and Gardner, B . The reaction between complement subcomponent C l q , IgG complexes and polyionic molecules. Mol.Immunol. 34:459-463, 1978. Humphries, G . M . K . and McConne l l , H . M . Antigen mobility in membranes and complement-mediated immune attack. Proc.Natl.Acad.Sci. USA 72:2483-2487, 1975. Israelachvili, J . Steric and fluctuating forces. In: Intermodular and Surface Forces, 2nd edition, edited by Israelachvili, J . Academic Press, London, 1992, p.288-311. Janatova, J . and Tack, B . F . Fourth component of human complement: studies of an amine-sensitive site comprised of a thiol component. Biochemistry 20:2394-2402, 1981. Janzen, J . , Song, X . and Brooks, D . E . Interfacial thickness of liposomes containing poly(ethylene glycol)-cholesterol from electrophoresis. Biophys.J. 70:313-320, 1996. Jiang, H . , Burdick, D . , Glabe, C . G . , Cotman, C . W . and Tenner, A . J . (3-Arhyloid activates complement by binding to a specific region of the collagen-like domain of the C l q A chain. J.Immunol. 152:5050-5059, 1994. Jiang, H . , Cooper, B . , Robey, F . A . and Gewurz, H . D N A binds and activates complement via residues 14-26 of the human C l q A chain. J.Biol.Chem. 25597-25601, 1992a. Jiang, H . , Robey, F . A . and Gewurz, H . Localization of sites through which C-reactive protein binds and activates complement to residues 14-26 and 76-92 of the human C l q A chain. J. Exp. Med. 176:1373-1379, 1992b. Joseph, B . S . , Cooper, N . R . and Oldstone, M . B . A . Immunologic injury of cultured cells infected with measles virus. I. Role of IgG antibody and the alternative complement pathway. J. Exp. Med. 141:761-774, 1975. Juliano, R . L . and Stamp, D . The effect of particle size and charge on the clearance rates 190 of liposomes and liposome encapsulated drugs. Biochem.Biophys.Res. Commun. 63:651-658, 1975. Kaplan, M . H . and Volanakis, J . E . Interactions of C-reactive protein with the complement system. I. Consumption of human complement associated with the reaction of C-reactive protein with pneumococcal polysaccharide and with choline phosphatides, lecithin, and sphingomyelin. J.Immunol. 112:2135-2147, 1974. Kenworthy, A . K . , Simon, S . A . and Mcintosh, T . J . Structure and phase behaviour of l ip id suspensions containing phospholipids with covalently attached poly (ethylene glycol). Biophys.J. 68:1903-1920, 1995. Ki rby , C , Clarke, J . and Gregoriadis, G . Cholesterol content of small unilamellar liposomes controls phospholipid loss to high density lipoproteins in the presence of serum. FEBS Lett. 111:324-328, 1980a. Ki rby , C , Clarke, J . and Gregoriadis, G . Effect of cholesterol content of small unilamellar liposomes on their stability in vivo and in vitro. Biochem. J. 186:591-598, 1980b. Klibanov, A . L . , Maruyama, K . , Beckerleg, A . M . , Torchil in, V . P . and Huang, L . Activi ty of amphipathic poly(efhylene glycol) 5000 to prolong the circulation time of liposomes depends on the liposome size and is unfavourable for immunoliposome binding to target. Biochim.Biophys.Acta 1062:142-148, 1991. Klibanov, A . L . , Maruyama, K . , Torchil in, V . P . and Huang, L . Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes. FEBS Lett. 1:235-237, 1990. Knauf, M . J . , B e l l , D . P . , Hirtzer, P . , Luo , Z . , Young, J . D . and Katre, N . V . Relationship of effective molecular size to systemic clearance in rats of recombinant interleukin-2 chemically modified with water-soluble polymers. J.Biol.Chem. 263:15064-15070, 1988. Kobayashi, K . , Izumi, Y . , Yoshizu, A . , Horinouchi, H . , Park, S.I., Sakai, H . , Takeoka, S., Nishide, H . , and Tsuchida, E . The oxygen carrying capability of hemoglobin vesicles evaluated in rat exchange transfusion models. Artif.Cells, Blood Sub.& Immobil. Biotech. 25:37-366, 1997. Kovacsovics, T . , Tschopp, J . , Kress, A . and Isliker, H . Antibody-independent activation of C I , the first component of complement, by cardiolipin. J.Immunol. 135:2695-2700, 1985. Krumdieck, R . , Hook, M . , Rosenberg, L . C . and Volanakis, J . E . The proteoglycan 191 decorin binds C l q and inhibits the activity of the C I complex. J.Immunol. 149:3695-3701, 1992. Kubak, B . , Potempa, L . , Anderson, B . , Mahklouf, S., Venegas, M . , Gewurz, H . and Gewurz, A . Evidence that serum amyloid P component binds to mannose-terminated sequences of polysaccharides and glycoproteins. Mol. Immunol. 25:851-858, 1988. Lachmann, P . J . , Munn , E . A . and Weissmann, G . Complement-mediated lysis of liposomes produced by the reactive lysis procedure. Mol.Immunol. 19:983-986, 1970. Lappalainen, K . , Miettinen, R . , Kel lokoski , J . , Jaaskelainen, I. and Syrjanen, S. Intracellular distribution of oligonucleotides delivered by cationic liposomes: light and electron microscopic study. J.Histochem.Cytochem. 45:265-274, 1997. Lasic, D . D . , Mart in , F . J . , Gabizon, A . , Huang, S . K . , and Papahadjopoulos, D . Sterically stabilized liposomes: a hypothesis on the molecular origin of the extended circulation times. Biochem.Biophys.Acta. 1070:187-192, 1991. Lee, K . - D . , N i r , S. and Papahadjopoulos, D . Quantitative analysis of liposome-cell interactions in vitro: rate constants of binding and endocytosis with suspension and adherent J774 cells and human monocytes. Biochemistry 32:889-899, 1993. Lee, K . - D . , Hong, K . and Papahadjopoulos, D . Recognition of liposomes by cells: in vitro binding and endocytosis mediated by specific l ipid headgroups and surface charge density. Biochim.Biophys.Acta 1103:185-197, 1992. Lee, K . - D , Pitas, R . E . and Papahadjopoulos, D . Evidence that the scavenger receptor is not involved in the uptake of negatively charged liposomes by cells. Biochim.Biophys.Acta 1111:1-6, 1992. Litzinger, D . C , Buiting, A . M . J . , Rooijen, N . and Huang, L . Effect of liposome size on the circulation time and intraorgan distribution of amphipathic poly(ethylene glycol)-containing liposomes. Biochim.Biophys.Acta 1190:99-107, 1994. Litzinger, D . C . and Huang, L . Amphipathic poly(ethylene glycol) 5000-stabilized dioleoylphosphatidylethanolamine liposomes accumulate in spleen. Biochim.Biophys.Acta 1127:249-254, 1992. L i u , D . , L i u , F . and Song, Y . K . Recognition and clearance of liposomes containing PS are mediated by serum opsonin. Biochim.Biophys.Acta 1235:140-146, 1995. L i u , Y . , Liggitt , D . , Zhong, W . , Tu , G . , Gaensler, K . , and Debs, R. Cationic liposome-192 mediated intravenous gene delivery. J.Biol.Chem. 270:24864-24870, 1995. L i u , Z . , Solow, R . and H u , V . W . Fluorescence analysis of size distribution and mode of dye release from carboxyfluorescein-loaded vesicles: application to the study of complement-membrane interactions. Biochim.Biophys.Acta 945:253-262, 1988. Loughrey, H . C . , Bal ly , M . B . , Reinish, L . W . and Cul l i s , P . R . The binding of phosphatidylglycerol liposomes to rat platelets is mediated by complement. Thromb.Haemostas. 64:172-176, 1990. L u , J . , Thiel , S., Wiedemann, F L , T impl , R. and Reid, K . B . M . Binding of the pentamer/hexamer forms of mannan-binding protein to zymosan activates the proenzyme C l r 2 C l s 2 complex, of the classical pathway of complement, without involvement of C l q . J.Immunol. 144:2287-2294, 1990. Lubl in , D . and Atkinson, J . Decay accelerating factor: biochemistry, molecular biology and function. Annual Rev.Biochem. 7:35-58, 1989. Madden, T . D . , Janoff, A . S . , and Cul l is , P .R . Incorporation of amphotericin B into large unilamellar vesicles composed of phosphatidylcholine and phosphatidylglycerol. Chem.Phys.Lipids 52:189-198, 1990. Marjan, J . , X i e , Z . and Devine, D . V . Liposome-induced activation of the classical complement pathway does not require immunoglobulin. Biochim.Biophys.Acta 1192:35-44, 1994. Maruyama, K . , Yuda, T . , Okamoto, A . , Ishikura, C , Kojima, S. and Iwatsuru, M . Effect of molecular weight in amphipathic polyethyleneglycol on prolonging the circulation time of large unilamellar liposomes. Chem.Pharm.Bull. 39:1620-1622, 1991. Maruyama, K . , Yuda, T . , Okamoto, A . , Kojima, S., Suginaka, A . and Iwatsuru, M . Prolonged circulation time in vivo of large unilamellar liposomes composed of distearoyl phosphatidylcholine and cholesterol containing amphipathic poly(ethylene glycol). Biochim.Biophys.Acta 1128:44-49, 1992. Masaki , T . , Okada, N . , Yasuda, R. and Okada, H . Assay of complement activity in human serum using large unilamellar liposomes. J.Immunol.Meth. 123:19-24, 1989. Matsuo, F L , Funato, K . , Harashima, H . and Kiwada, H . The complement- but not mannose receptor-mediated phagocytosis is involved in the hepatic uptake of cetymannoside-modified liposomes in situ. J.Drug Targeting 2:141-146, 1994. 193 Mayer, L . D . , Ta i , L . C . L . , K o , D . S . C . , Mas in , D . , Ginsberg, R . S . , Cul l i s , P .R . and Bal ly , M . B . Influence of vesicle size, l ip id composition, and drug-to-lipid ratio on the biological activity of liposomal doxorubicin in mice. Cancer Res. 49:5922-5930, 1989. Mayer, M . M . Complement and complement fixation. In: Experimental Immunochemistry, edited by Kabat, E . A . and Mayer , M . M . Springfield: Charles C . Thomas Publisher, 1961, p. 133-240. M c A l e e r , M . A . and S im, R . B . The Complement System. In: Activators and Inhibitors of Complement, edited by Sim, R . B . Dordrecht: Kluwer Academic Press, 1993, p.1-15. Mcintosh, T . J . , Kenworthy, A . K . , and Needham, D . Measurement of the range and magnitude of the repulsive pressure between PEG-coated liposomes. In: Stealth Liposomes, edited by Lasic, D . and Mart in , F . Boca Raton: C R C Press, 1995, p. 63-70. Mendez, A . J . , He , J . L . , Huang, H . S . , Wen, S.R. , and Hsia , S . L . Interaction of rabbit lipoproteins and red blood cells with liposomes of egg yolk phospholipids. Lipids 23:961-967, 1988. Menzel , E . J . , Smolen, J . and Reid, K . Interaction of collagen with C l q via its collagen-like portion. Biochim.Biophys.Acta 610:265-212>, 1981. M e r i , S., Morgan, B . P . , Davies, A . , Daniels, R . H . , Olavesen, M . G . , Waldmann, H . and Lachman, P . J . Human protectin (CD59), an 18000-20000 M W complement lysis restricting factor inhibits C5b-8 catalyzed insertion of C 9 into l ip id bilayers. Mol.Immunol. 71:1-9, 1990. M e r r i l l , E . W . Poly(ethylene oxide) and blood contact. In: Poly(Ethylene glycol) chemistry: biotechnical and biomedical applications, edited by Harris, M . J . New York : Plenum Press, 1992, p. 199-220. M i m m s , L . T . , Zampighi, G . , Nozaki , Y . , Tanford, C . and Reynolds, J . A . Phospholipid vesicle formation and transmembrane protein incorporation using octylglucoside. Biochemistry 20:833-840, 1981. Moghimi , S . M . and Patel, H . M . Opsonophagocytosis of liposomes by peritoneal macrophages and bone marrow reticuloendothelial cells. Biochim.Biophys.Acta 1135:269-274, 1992. M o l d , C . Effect of membrane phospholipids on activation of the alternative complement pathway. J.Immunol. 143:1663-1668, 1989. 194 M o l d , C . and Gewurz, A . T . Activation of human complement by liposomes: serum factor requirement for alternative pathway activation. J.Immunol. 125:696-700, 1980. Morgan, B . P . The biological effects of complement activation. In: Complement. Clinical aspects and relevance to disease, London: Academic Press, 1990, p. 36-56. M o r i , A . , Klibanov, A . L . , Torchil in, V . P . and Huang, L . Influence of the steric barrier activity of amphipathic poly(ethyleneglycol) and ganglioside G M 1 on the circulation time of liposomes and on the target binding of immunoliposomes in vivo. FEBS Lett. 284:263-266, 1991. Morikawa, K . , Nayar, R. and Fidler, I.J. In vitro activation of tumoricidal properties in mouse macrophages using the chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine ( F M L P ) incorporated in liposomes. Cancer Immunol.Immunotherapy 27:1-6, 1988. Morrison, D . C . and Kl ine , L . F . Activation of the classical and properdin pathways of complement by bacterial lipopolysaccharides (LPS) . J.Immunol. 118:362-368, 1977. Mosier , M . and McLaughl in , S. Binding of basic peptides to acidic lipids in membranes: effects of inserting alanine(s) between the basic residues. Biochemistry 31:1767-1773, 1992. Murphy, B . , Kirszbaum, L . , Walker, I. and d-Apice, J . SP40-40 a newly identified normal human serum protein found in the SC5b-6 complex of complement and in immune deposits in glomerulonephritis. J.Clin.Invest. 81:1858-1864, 1988. Nagaoka, S., M o r i , Y . , Takiuchi, H . , Yokota, K . , Tanzawa, H . and Nishiumi, S. Interaction between blood components and hydrogels with poly(oxyethylene) chains. In: Polymers as Biomaterials, edited by Shalaby, S .W. , Hoffman, A . S . , Ratner, B . D . and Horbett, T . A . New York : Plenum Press, 1985, p. 361-374. N i i , A . , Utsugi, T . , Fan, D . , Denkins, Y . , Pak, C . , Brown, D . , van Hoogevest, P . , and Filder, I.J. Optimization of the liposomes encapsulating a new lipopeptide C G P 31362 for efficient activation of tumoricidal properties in monocytes and macrophages. J.Immunotherapy 10:236-246, 1991. Ohkuro, M . , Kobayashi, K . , Takahashi, K . and Nagasawa, S. Effect of C l q on the processing of immune complexes by human neutrophils. Mol.Immunol. 83:507-511, 1994. Ohta, M . , Okada, M . , Yamashina, I. and Kawasaki, T . The mechanism of 195 carbohydrate-mediated complement activation by the serum mannan-binding protein. J.Biol.Chem. 265:1980-1984, 1990. Ola l , A . D . PhD. Thesis, Department of Chemistry, University of British Columbia. A chemical study of the interaction of proteins with PSL, 1990. Papahadjopoulos, D . Stealth Liposomes: From steric stabilization to targeting. In: Stealth Liposomes, edited by Lasic, D . and Mart in , F . Boca Raton: C R C Press, 1995, p. 1-6. Papahadjopoulos, D . , A l l e n , T . M . , Gabizon, A . , Mayhew, E . , Matthay, K . , Huang, S . K . , Lee, K . D . , Woodle, M . C . , Lasic, D . D . , Redemann, C . and Mart in , F . J . Sterically stabilized liposomes: Improvements in pharmacokinetics and antitumour efficacy. Proc.Natl.Acad.Sci. USA 88:11460-11464, 1991. Parce, J . W . , Kel ley , D . and Heinzelmann, K . Measurement of antibody-dependent binding, proteolysis, and turnover of C l s on liposomal antigens localizes the fluidity-dependent step in C I activation. Biochim.Biophys.Acta 736:92-98, 1983. Parce, J . W . , McConne l l , H . M . , Bartholemew, R . M . and Esser, A . F . Kinetics of antibody-independent activation of the first component of complement on l ipid bilayer membranes. Biochem.Biophys.Res.Commun. 93:235-242, 1980. Parr, M . J . , Anse l l , S . M . , Choi , L . S . and Cul l is , P .R . Factors influencing the retention and chemical stability of poly(ethylene glycol)-lipid conjugates incorporated into large unilamellar vesicles. Biochim.Biophys.Acta 1195:21-30, 1994. Patel, K . R . , L i , M . P . , Schuh, J .R. and Baldeschwieler, J . D . The pharmacological efficacy of a rigid non-phospholipid liposome drug delivery system. Biochim.Biophys.Acta 797:20-26, 1984. Peerschke, E . I . B . and Ghebrehiwet, B . Human blood platelets possess specific binding sites for C l q . J.Immunol. 138:1537-1541, 1987. Peerschke, E . I . B . and Ghebrehiwet, B . Platelet C l q receptor interactions with collagen- and Clq-coated surfaces. J.Immunol. 145:2984-2988, 1990. Pepys, M . B . , Dash, A . C . , Markham, R . E . , Thomas, H . C . , Wil l iams, B . D . and Petrie, A . Comparative clinical study of protein S A P (amyloid P component) and C-reactive protein in serum. Clin. Exp. Immunol. 32:119-124, 1978. Poon, P . H . , Phil l ips, M . L . and Schumaker, V . N . Immunoglobulin M possesses two binding sites for complement subcomponent C l q and soluble 1:1 and 1:2 196 complexes are formed in solution in reduced ionic strength. J.Biol. Chem. 260:9357-9365, 1985. Porter, R . R . and R e i d , K . B . The biochemistry of complement. Nature 275:699-704, 1978. Rahman, A . , Carmichael, D . , Harris, M . , and Roh, J . K . Comparative pharmacokinetics of free doxorubicin and doxorubicin entrapped in cardiolipin liposomes. Cancer.Res. 46:2295-2299, 1986a. Rahman, A . , Ganjei, A . , Neefe, J .R. Comparative immunotoxicity of free doxorubicin and doxorubicin encapsulated in cardiolipin liposomes. Cancer Chemo.&Pharmacol. 16:28-34, 1986b. Rahman, A . , White, G . , More , N . , and Schein, P .S . Pharmacological, toxicological, and therapeutic evaluation in mice of doxorubicin entrapped in cardiolipin liposomes. Cancer.Res. 45:796-803, 1985. Rebecchi, M . , Peterson, A . and McLaughl in , S. Phosphoinositide-specific phospholipase C-5 , binds with high affinity to phospholipid vesicles containing phosphatidylinositol 4,5-bisphosphate. Biochemistry 31:12742-12747, 1992. Reid, K . , Reid , B . M . and Porter, R . R . Subunit composition and structure of subcomponent C l q of the first component of human complement. Biochem.J. 155:19-23, 1976. Reid, K . B . M . , Gagnon, J. and Frampton, J . Completion of the amino acid sequences of the A and B chains of subcomponent C l q of the first component of human complement. Biochem.J. 203:559-569; 1982. Rent, R . , Ertel , N . , Eidenstein, R. and Gewurz, H . Complement activation by interactions of polyanions and poly cations. J.Immunol. 114:120-124, 1975. Richards, R . L . , Gewurz, H . , Siegel, J . and A l v i n g , C R . Interactions of C-reactive protein and complement with liposomes. J.Immunol. 122:1185-1189, 1979. Richards, R . L . , Habbersett, R . C , Scher, I., Janoff, A . S . , Schieren, H . P . , Mayer , L . D . , Cul l i s , P . R . and A l v i n g , C R . Influence of vesicle size on complement-dependent immune damage to liposomes. Biochim.Biophys.Acta 855:223-230, 1986. Robey, F . A . , Jones, K . D . and Steinberg, A . D . C-reactive protein mediates the solubilization of nuclear D N A by complement in vitro. / . Exp. Med. 161:1344-1356, 1985. 197 Rodrigueza, W . V . PhD Thesis. Studies on the antiatherogenic properties of liposomes. Dept. of Biochemistry, University of British Columbia, 1994. Roerdink, F . , Wassef, N . M . , Richardson, E . C . and A l v i n g , C R . Effects of negatively charged lipids on phagocytosis of liposomes by complement. Biochim.Biophys.Acta 734:33-39, 1983. Rosano, C L . and Hurwitz, C . Separation of hydroxyproline-containing protein from C l q (a subcomponent of complement) in serum. Clin. Chem. 23:1335-1337, 1977. Satake-Ishikawa, R . , Ishikawa, M . , Okada, Y . , Kakitani, M . , Kawagishi, M . , Matsuki, S. and Asano, K . Chemical Modification of recombinant human granulocyte colony-stimulating factor by polyethylene glycol increases its biological activity in vivo. Cell Structure and Function 17:157-160, 1992. Scherphof, G . , Roerdink, F . , Waite, M . and Parks, J . Disintegration of phosphatidylcholine liposomes in plasma as a result of interaction with high-density lipoproteins. Biochim.Biophys.Acta 542:296-307, 1978. Schreiber, R . D . The chemistry and biology of complement receptors. Springer Semin.Immunopath. 7:221-249, 1984. Schreiber, R . D . , Morr ison, D . C , Podack, E . R . and Muller-Eberhard, H . J . Bactericidal activity of the alternative complement pathway generated from eleven isolated plasma proteins. / . Exp. Med. 149:870-882, 1979. Schumaker, V . N . , Calcott, M . , Spiegelberg, H . L . and Muller-Eberhard, H . J . Ultracentrifugation studies of the binding of IgG of different subclasses to the C l q subunit of the first component of complement. Biochemistry 15:5175-5181, 1976. Scieszka, J .F . , Maggiora, L . L . , Wright, S .D . and Cho, M . J . Role of Complements C3 and C5 in the phagocytosis of liposomes by human neutrophils. Pharm.Res. 8:65-69, 1991. Seaman, G . V . F . Electrokinetic behaviour of red cells. In: The red blood cell, edited by Surgenor, D . M . New York : Academic Press, 1975, p. 1135-1229. Senior, J . , Crawley, J . C . W . and Gregoriadis, G . Tissue distribution of liposomes exhibiting long half-lives in the circulation after intravenous injection. Biochim.Biophys.Acta 839:1-8, 1985. Senior, J . , Delgado, C , Fisher, D . , Tilcock, C . and Gregoriadis, G . Influence of surface 198 hydrophilicity of liposomes on their interaction with plasma protein and clearance from the circulation: studies with poly(ethylene glycol)-coated vesicles. Biochim.Biophys.Acta 1062:77-82, 1991. Senior, J . and Gregoriadis, G . Stability of small unilamellar liposomes in serum and clearance from the circulation: the effect of the phospholipid and cholesterol components. Life Sci. 30:2123-2136, 1982. Shin, M . L . , Paznekas, W . A . , Abramovitz, A . S . and Mayer , M . M . O n the mechanism of membrane damage by C : exposure of hydrophobic sites on activated C proteins. J.Immunol. 119:1358-1364, 1977. Shin, M . L . , Paznekas, W . A . and Mayer, M . M . On the mechanism of membrane damage by complement: the effect of length and unsaturation of the acyl chains in liposomal bilayers and the effect of cholesterol concentration in sheep erythrocyte and liposomal membranes. J.Immunol. 120:1996-2002, 1978. Silversmith, R . E . and Nelsestuen, G . L . Assembly of the membrane attack complex of complement on small unilamellar phospholipid vesicles. Biochemistry 25:852-860, 1986. Silvestri, L . , Baker, J .R . , Roden, L . and Stroud, R . M . The C l q inhibitor in serum is a chondroitin 4-sulfate proteoglycan. J.Biol.Chem. 256:7383-7387, 1981. Silvius, J .R. and Zuckermann, M . J . Interbilayer transfer of phospholipid-anchored macromolecules via monomer diffusion. Biochemistry 32:3153-3161, 1993. Sim, R . B . , Ar laud, G . J . , and Colomb, M . G . C I inhibitor dependent dissociation of human complement component C I bound to immune complexes. Biochem.J. 179:449-454, 1 9 7 9 . Sim, R . B . , Twose, T . M . , Paterson, D . S . and Sim, E . The covalent-binding reaction of complement component C 3 . Biochem.J. 193:115-127, 1981. Smith, K . F . , Haris, P . I . , Chapman, D . , Reid, K . B . M . and Perkins, S.J. Beta-sheet secondary structure of the trimeric globular domain of C l q of complement and collagen types VIII and X by Fourier-transform infrared spectroscopy and averaged structure predictions. Biochem.J. 301:249-256, 1994. Sorvil lo, J . , G i g l i , I. and Pearlstein, E . Fibronectin binding to complement subcomponent C l q : localization of their respective binding sites. Biochem.J. 226:207-215, 1985. Shaw, D . Charged interfaces. In: Introduction to Colloid and Surface Chemistry, edited 199 by Shaw, D . Butterworth & Co . (Publishers) L td . London, 1970, p. 133-166. Strang, C . J . , Siegel, R . C . , Phillips, M . L . , Poon, P . H . and Schumaker, V . N . Ultrastructure of the first component of human complement: electron microscopy of the crosslinked complex. Proc.Natl.Acad.Sci. USA 79:586-590, 1982. Stuart, M . A . C . , Scheutjens, J . M . H . M . and Fleer, G . J . Polydispersity effects and the interpretation of polymer adsorption isotherms. J.Poly.Sci.Poly.Phys.Ed. 18:559-573, 1980. Super, M . , Levinsky, R . J . and Turner, M . W . The level of mannan-binding protein regulates the binding of complement-derived opsonins to manna and zymosan at low serum concentrations. Clin. Exp. Immunol. 79:144-150, 1990. Swartz, G . M . , Gentry, M . K . , Amende, L . M . , Blanchette-Mackie, E . J . and A l v i n g , C R . Antibodies to cholesterol. Proc.Natl.Acad.Sci.USA 85:1902-1906, 1988. Szoka, F . and Papahadjopoulos, D . Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc.Natl.Acad.Sci.USA 75:4194-4198, 1978. Szoka, F . , J r . and Papahadjopoulos, D . Liposomes: Preparation and characterization. In: Liposomes: from physical structure to therapeutic applications, edited by Knight, Elsevier/North-Holland, 1981, p. 51-82. Ta l l , A . R . and Small , D . M . Solubilisation of phospholipid membranes by human plasma high density lipoproteins. Nature 265:163-164, 1977. Taylor, B . , Wright, J .F . , A r y a , S., Isenman, D . E . , Shuiman, M . J . and Painter, R . H . C l q binding properties of monomer and polymer forms of mouse I g M /x-chain variants. J.Immunol. 153:5303-5313, 1994. Tenner, A . J . and Cooper, N . R . Stimulation of a human polymorphonuclear leukocyte oxidative response by the C l q subunit of the first complement component. J.Immunol. 128:2547-2552, 1982. Tenner, A . J . , Lesavre, P . H . and Cooper, N . R . Purification and radiolabeling of human C l q . J.Immunol. 127:648-653, 1981. Tenner, A . J . , Ziccardi , R . J . and Cooper, N . R . Antibody-independent C I activation by E . C o l i . J.Immunol. 133:886-891, 1984. Thiel , S., Vorup-Jensen, T . , Stover, C M . , Schwaeble, W . , Laursen, S .B . , Poulsen, K . , W i l l s , A . C , Eggleton, P . , Hansen, S., Holmskov, U . , Reid , K . B . M . , and 200 Jensenlius, J . C . A second serine protease associated with mannan-binding that activates complement. Nature 386: 506-510, 1997. Thielens, N . M . , Bal ly , I . M . , Ebenbichler, C . F . , Dier ich, M . P . and Arlaud, G . J . Further characterization of the interaction between the C l q subcomponent of human C I and the transmembrane envelope glycoprotein gp41 of H I V . J.Immunol. 151:6583-6592, 1993. Ti lcock, C , Ahkong, Q . F . and Fisher, D . Polymer-derivitized technetium labeled liposomal blood pool agents for nuclear medicine applications. Biochim.Biophys.Acta 1148:77-84, 1993. Torchil in, V . P . , Kl ibanov, A . L . , Huang, L . , O 'Donnel l , S., Nossiff, N . D . and Khaw, B . A . Targeted accumulation of polyethylene glycol-coated immunoliposomes in infarcted rabbit myocardium. FASEB 6:2716-2719, 1992. Tschopp, J . , Schulthess, T . , Engel, J . and Jaton, J . C . Antigen-independent activation of the first component of complement C I by chemically crosslinked rabbit IgG-oligomers. FEBS Lett. 112:152-154, 1980. Uster, P . S . , A l l e n , T . M . , Daniel , B . E . , Mendez, C . J . , Newman, M . S . , and Zhu, G . Z . Insertion of poly(ethylene glycol) derivatized phospholipid into pre-formed liposomes results in prolonged in vivo circulation time. FEBS Lett. 386:243-246, 1996. Uwatoko, S. and Mannik, M . The location of binding sites on C l q for D N A . J.Immunol. 144:3484-3488, 1990. Vanguri , P . , K o s k i , C . L . , Silverman, B . and Shin, M . L . Complement activation by isolated myelin: activation of the classical pathway in the absence of myelin-specific antibodies. Proc.Natl.Acad.Sci. USA 79:3290-3294, 1982. Verma, J . N . , Wassef, N . M . , Wir tz , R . A . , Atkinson, C . T . , Aikawa, M . , Loomis, L . D . , and A l v i n g , C R . Phagocytosis of liposomes by macrophages: intracellular fate of liposomal malaria antigen. Biochim.Biophys.Acta 1066:229-238, 1991. Vertut-Doi, A . , Ishiwata, H . and Miyaj ima, K . Binding and uptake of liposomes containing a poly(ethylene glycol) derivative of cholesterol (stealth liposomes) by the macrophage cell line J774: influence of P E G content and its molecular weight. Biochim.Biophys.Acta 1278:19-28, 1996. Volanakis, J . E . and Narkates, A . J . Interaction of C-reactive protein with artificial phosphatidylcholine bilayers and complement. J.Immunol. 126:1820-1825, 1981. 201 Wasan, K . M . and Lopez-Berestein, G . The past, present, and future uses of liposomes in treating infectious diseases. Immunopharm.Immunotox. H:\-15, 1995. Wassef, N . M . and A l v i n g , C . R . Complement-dependent phagocytosis of liposomes by macrophages. Methods Enzymol. 149:124-134, 1987. Wassef, N . M . , Matyas, G . R . , and A l v i n g , C . R . Complement-dependent phagocytosis of liposomes by macrophages: suppressive effects of "stealth" lipids. Biochem. Biophys. Res. Comm. 176:866-874, 1991. Wassef, N . M . and A l v i n g , C . R . Complement-dependent phagocytosis of liposomes. Chem.Phys.Lipids 64:239-248, 1993. Wassef, N . M . , Johnson, S . H . , Graeber, G . M . , Swartz, G . M . , Schultz, C . L . , Hailey, J .R . , Johnson, A . J . , Taylor, D . G . , Ridgeway, R . L . and A l v i n g , C . R . Anaphylactoid reactions mediated by autoantibodies to cholesterol in miniature pigs. J.Immunol. 143:2990-2995, 1989. Wines, B . D . and Easterbrook-Smith, S .B. Enhancement of the binding of C l q to immune complexes by polyethylene glycol . Mol.Immunol. 25:263-266, 1988. Wong, M . , Anthony, F . H . , Til lack, T . W . and Thompson, T . E . Fusion of dipalmitoylphosphatidylcholine vesicles at 4°C. Biochemistry 21:4126-4132, 1982. Woodle, M . C . Surface-modified liposomes: assessment and characterization for increased stability and prolonged blood circulation. Chem.Phys.Lipids 64:249-262, 1993. Woodle, M . C . and Lasic, D . D . Sterically stabilized liposomes. Biochim.Biophys.Acta 1113:171-199, 1992. Woodle, M . C , Matthay, K . K . , Newman, M . S . , Hidayat, J . E . , Col l ins , L . R . , Redemann, C , Mart in , F . J . and Papahadjopoulos, D . Versatility in l ipid compositions showing prolonged circulation with sterically stabilized liposomes. Biochim.Biophys.Acta 1105:193-200, 1992. Woodle, M . C , Storm, G . , Newman, M . S . , Jekot, J . J . , Coll ins , L . R . , Mart in , F . J . and Szoka, F . C . J r Prolonged systemic delivery of peptide drugs by long-circulating liposomes: illustration with Vasopressin in the brattleboro rat. Pharm.Res. 9:260-265, 1995. Wright, J . K . , Tschopp, J . , Jaton, J . C . and Engel, J . Dimeric , trimeric and tetrameric complexes of immunoglobulin G fix complement. Biochem.J. 187:775-780, 1980. 202 Yang, L . and Glaser, M . Membrane domains containing phosphatidylserine and substrate can be important for the activation of protein kinase C . Biochemistry 34:1500-1506, 1995. Yaroslavov, A . A . , K u l ' k o v , V . E . , Polinsky, A . S . , Baibakov, B . A . and Kabanov, V . A . A polycation causes migration of negatively charged phospholipids from the inner to outer leaflet of the liposomal membrane. FEBS Letters 340:121-123, 1994. Y i n g , S., Gewurz, A . T . , Jiang, H . and Gewurz, H . Human serum amyloid P component oligomers bind and activate the classical complement pathway via residues 14-26 and 76-92 of the A chain collagen-like region of C l q . J.Immunol. 150:169-176, 1993. Yonemasu, K . , Stroud, R . M . , Neidermeier, W . and Butler, W . T . Chemical studies on C l q ; A modulator of immunoglobulin biology. Biochem.Biophys.Res. Commun. 43:1388-1394, 1971. Yoshioka, H . Surface modification of haemoglobin-containing liposomes with polyethylene glycol prevents liposome aggregation in blood plasma. Biomaterials 12:861-864, 1991. Zalipsky, S. Polyethylene glycol-l ipid conjugates. In: Stealth Liposomes, edited by Lasic , D . and Mart in , F . Boca Raton: C R C Press, 1995, p. 93-102. Ziccardi , R . J . A new role for Cl- inhibi tor in homeostasis: control of activation of the first component of human complement. J.Immunol. 128:2505-2508, 1982. Ziccardi , R . J . and Cooper, N . R . Active Disassembly of the first component of complement, C I , by Cl-inactivator. J.Immunol. 123:788-792, 1979. Zohair, A . , Chesne, S., Wade, R . H . and Colomb, M . G . Interaction between complement subcomponent C l q and bacterial lipopolysaccharides. Biochem.J. 257:865-873, 1989. 203 Appendix 1: Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes at pH 7.2 in a Purified Protein System: Effect of Lipid Concentrations. The Scatchard plots for the binding of C l q to C L 2 0 liposomes where the total l ipid concentration was 0.5 m M in the reaction are shown in Panel A and B . Scatchard plots for C l q binding where the total l ipid was 2 m M are shown in Panel C and Panel D . The curvi-linear Scatchard plots (Panel A and Panel C) was derived from the complete set of binding data for C l q binding to C L 2 0 liposomes from Figure 19. The straight-line plots (Panel B and Panel D) were derived from the steep portion of the binding plots. 204 205 Appendix 2 Scatchard Plots for C l q Binding to 30 and 40 mole % Cardiolipin-containing Liposomes at pH 7.2 in a Purified Protein System. The Scatchard plots derived from the C l q binding data in Figure 10 for C l q binding to C L 3 0 and C L 4 0 liposomes is shown in Panel A and B , respectively. Binding assays were carried out with a total l ipid concentration of 0.5 m M in the reaction. A 0.40 0.80 v (umol C1q bound/umol total lipid 206 Appendix 3: Scatchard Plots for C l q Binding to Cardiolipin-containing Liposomes at pH 4 in a Purified Protein System. Scatchard plots were derived from the C l q binding data in Figure 17 for C l q binding to C L 2 0 liposomes (Panel A) and to C L 4 0 liposomes (Panel B and C) at p H 4. Panel C shows the steep portion of the Scatchard plot; the complete set of binding data was used to generate the curvi-linear Scatchard plot shown in Panel B . A 0.12 0.30 v (Mmol C1q bound/nmol total lipid) 207 Appendix 4: Scatchard Plots for C l q Binding to Phosphatidylglycerol-containing Liposomes at pH 4 in a Purified Protein System. Scatchard plots were derived from the C l q binding data in Figure 18 for C l q binding to PG40 liposomes (Panel A) and to P G 55 liposomes (Panel B) at p H 4. A 1.00 30 v (umol C1q bound/umol total lipid) 208 Appendix 5: Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes at pH 7.2 in a Purified Protein System at 100 mM NaCl. Scatchard plots were derived from binding data in Figure 21 for the binding of C l q to 240 nm C L 2 0 liposomes (Panel A and Panel B) and to C L 2 0 M L V s (Panel C) at 100 m M N a C l . While the complete set of C l q binding data for 240 nm C L 2 0 liposomes was used to generate the curvi-linear Scatchard plot shown in Panel A , the steep portion of the Scatchard plot is shown in Panel B . 209 Appendix 6: Scatchard Plots for C l q Binding to 20 mole % Cardiolipin-containing Liposomes in Clq-Depleted serum (1/60): Effect of Lipid Concentration. Scatchard plots for the binding of C l q to C L 2 0 liposomes in serum (1/60) where the total l ipid concentration was 0.6, 2, and 6 m M in the reaction are shown in Panel A , B , and C , respectively. Scatchard plots were derived from the complete set of binding data for C l q binding to C L 2 0 liposomes in Clq-depleted serum from Figure 22. A 0.50 v (pmol C1q boundypmol total lipid) 210 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088736/manifest

Comment

Related Items