Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The interaction of liposomes with complement proteins and protein s Marjan, Jihan Mohammed Jamil 1994

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

Item Metadata


831-ubc_1995-983205.pdf [ 2.67MB ]
JSON: 831-1.0088875.json
JSON-LD: 831-1.0088875-ld.json
RDF/XML (Pretty): 831-1.0088875-rdf.xml
RDF/JSON: 831-1.0088875-rdf.json
Turtle: 831-1.0088875-turtle.txt
N-Triples: 831-1.0088875-rdf-ntriples.txt
Original Record: 831-1.0088875-source.json
Full Text

Full Text

THE INTERACTION OF LIPOSOMES WITHCOMPLEMENT PROTEINS AND PROTEIN SbyJIHAN MOHAMMED JAMIL MARJANA.B. BRYN MAWR COLLEGE (cum laude), 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF PATHOLOGYWe accept this thesis as conformingto the required standardThe University of British ColumbiaDecember 1994© Jihan Mohammed Jamil Marjan, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________Department of f\Tti D LCD & -The University of British ColumbiaVancouver, CanadaDate CE C*hDE-6 (2/88)ABSTRACTThe use of liposomes as drug delivery units necessitates that they come intocontact with a number of plasma proteins and cells. This dissertation investigatesthe interaction of liposomes with two plasma reaction cascades: the complementand the coagulant cascade.The first section of the studies involves the investigation of the contribution ofimmunoglobulin to the liposome-induced activation of complement in human serum.Liposomes containing the negatively charged phospholipids cardiolipin,phosphatidyiglycerol or phosphatidylinositol, in addition to the neutral charged lipidsphosphatidylcholine and cholesterol, were used to activate complement in a wholeserum system. The contribution of immunoglobulin was studied by comparingnormal human serum to serum depleted of lgG and 1gM. Using hemolytic assays ofcomplement function, greater concentrations of phospholipids were required toactivate complement in the absence of immunoglobulins. Activation of the classicalpathway was confirmed using a Clq consumption assay which showed thatactivation was dependent on the presence of Clq and confirmed that greaterconcentrations of phospholipids were required to activate complement in theabsence of immunoglobulins. Complement activation was also assessed usingcrossed immunoelectrophoresis of C3 activation fragments. Using immunoblotanalysis, iC3b was detected on the surface of liposomes exposed to normal humanserum or immunoglobulin-depleted serum.Having demonstrated that specific antibody was not required, the secondsection of the studies involves the investigation of the contribution of mannosebinding protein to the liposome-induced activation of complement in human serum.The ligand to which mannose binding protein binds was also identified. Mannosebinding protein is an activator of the classical complement pathway in the absence ofimmunoglobulin and Clq. Using immunoaffinity chromatography it wasdemonstrated that mannose binding protein copurified with beta-2-glycoprotein- 1when beta-2-glycoprotein-1 was immobilized by specific antibody. Usingimmunoblot analysis mannose binding protein was shown to bind to PC:CH:CL(35:45:20 mol%) through beta-2-glycoprotein-1 which in itself binds avidly to anionicliposomes. This binding was specific as it was inhibited using N-acetylmannosamine, an inhibitor of mannose binding protein. Using complementhemolytic assays in the presence of mannose binding protein inhibitors, it wasdemonstrated that the route of activation of the classical pathway was primarilythrough Clq in the absence or presence of immunoglobulins. Preliminarycompetitive binding studies confirmed the presence of the competition between Clqand mannose binding protein that was previously observed using the functionalcomplement hemolytic assays. The competitive binding study demonstrated thatvarious concentrations of mannose binding protein, including physiologicalconcentration, could compete with and inhibit binding of 125l-labeled Clq toPC:CH:CL (35:45:20 mol%). There was no binding of 125l-labeled Clq detectedwith the neutral composition PC:CH (55:45 mol%).During the course of the immunoaffinity chromatography studies conductedin section two, it was demonstrated that a number of other proteins copurified withbeta-2-glycoprotein-1. Protein S was one of five proteins that copurified withbeta-2-glycoprotein-1 when the latter was immobilized using immunoaffinitychromatography. The final section of the studies deals with the investigation of thecontribution of protein S on the anticoagulant ability of plasma as the removal of free—III—protein S from circulation may disrupt the hemostatic balance. Using Western blotanalysis, protein S was detected on the surface of anionic liposomes. Furthermore,free protein S was deposited on the surface of increasing concentrations of anionicliposomes, quantitated using immunological assays. This effect translated into alower functional activity of protein S for plasma exposed to anionic liposomes,demonstrated using a commercial assay to measure the functional activity of proteinS. The removal of free protein S from the plasma was due to the depletion ofbeta-2-glycoprotein-1.These studies suggest that the incorporation of glycosides andpolyethyleneglycol into the liposomes would be beneficial since they may reduce theobserved interactions demonstrated in this study. This in turn could potentiallyincrease the half-life of the circulating liposomes.-iv-TABLE OF CONTENTSAbstract iiTable of Contents vlist of Tables ixList of Figures xAcknowledgments xiiiDedication xivAbbreviations xvChapter 1. Introduction 11.1. Classical pathway of complement 31.1.1. Formation of activated Cl 31.1.2. Deposition of 04 and formation of C3 convertase 61.1.3. Cleavage of 03, C5 and formation ofthe terminal complex 71.2. Alternative pathway of complement 81.3. Regulation of the complement pathway 91.4. Activation of the classical complement pathwayin the absence of immunoglobulins 101.4.1. C-Reactive Protein 101.4.2. Serum Amyloid P 121.4.3. Mannose Binding Protein 131.4.4. Polyanionic Substances 161.5. Coagulation: an overview 161.5.1. Regulation of the coagulant pathway 201.6. Liposomes 221.6.1. Physical aspect of liposomes 221.6.2. Classification of liposomes 241.7. The interaction between liposomes and complement 251.8. The interaction between liposomesand beta-2-glycoprotein-1 281.9. Overall Objective 30Chapter 2. Materials and Methods 322.1. Reagents 322.1.1. Antisera and Antibodies 322.1.2. Other Reagents 322.1.3. Kits Purchased 332.2. Preparation of reagents for experiments 342.2.1. Preparation of normal human serum and immunoglobulindepleted serum 342.2.2. Preparation of normal human plasma 352.2.3. Preparation of beta-2-glycoprotein- 1depleted plasma 352.2.4. Preparation of liposomes 362.3. Analytic Techniques 372.3.1. Functional complement assay 372.3.2. C3 crossed immunoelectrophoresis 382.3.3. Measurement of alternative pathway activation 392.3.4. Analysis of Clq binding to liposomes 402.3.5. Immunoblot analysis of proteinsassociated with liposomes 41-vi-2.3.6. Competitive binding study betweenClq and Mannose Binding Protein for anionicliposomes 442.3.7. Detection of serum mannose binding proteinbinding to beta-2-glycoprotein-1using an enzyme linked immunosorbent assay 452.3.8. Detection of beta-2-glycoprotein- 1 usingan enzyme linked immunosorbent assay 452.3.9. Immunological free protein S quantitation 462.3.10. Functional protein S Assay 472.3.11. Staining Procedures 482.4. Statistical Analysis 49Chapter 3. Results 503.1. Characterization of normal human serum and immunoglobulindepleted serum 503.2. Functional complement assay 573.3. C3 crossed immunoelectrophoresis 623.4. Measurement of alternative pathway activation 663.5. Analysis of Cl q binding to liposomes 663.6. Immunoblot analysis of 03 fragments on liposomes 723.7. Identification of proteins that copurifywith beta-2-glycoprotein-1 763.8. Binding of mannose binding protein on the surface ofPC:CH:CL (35:45:20 mol%) liposomes 84-vii-3.9. Measurement of complement hemolytic activity in thepresence of mannose binding protein inhibitors 913.10. Competitive binding study betweenClq and mannose binding protein 983.11. Binding of protein S on the surfaceof anionic liposomes 1003.12. Immunological protein S levels in plasma exposed toliposomes or anti-beta-2-glycoprotein- 1 Sepharose 1053.13. Characterization of the functional activity of protein Sin the plasma exposed to liposomesor anti-beta-2-glycoprotein- 1 Sepharose 109Chapter 4. Discussion 113Chapter 5. Summary 126Bibliography 128-viii-LIST OF TABLESTABLE 1. Characterization of immunoglobulin depleted serum 54Table 1A Quantitation of IgG and 1gM levelsusing a QM 300 Nephlometer 54Table 1 B CH5O and APH5O values 54TABLE 2. Activation of complement detected by curve shift inCrossed immunoelectrophoresis 65TABLE 3. Levels of the alternative pathway activation fragment Bbin normal human serum and immunoglobulin depleted serum 69TABLE 4. Densitometer readings of the elution profileafter exposure to anti-beta-2-glycoprotein-1 Sepharose 80TABLE 5. Immunological protein S levels in plasmaexposed to liposomes 107TABLE 6. Immunological protein S levels inbeta-2-glycoprotein-1 depleted plasma 108TABLE 7. Functional protein S levels in plasmaexposed to Iiposomes 111TABLE 8. Functional protein S levels inbeta-2-glycoprotein-1 depleted plasma 112-ix-LIST OF FIGURESFIGURE 1. The complement pathway 5FIGURE 2. Recognized pathways leading to the activationof the complement system 15FIGURE 3. Physiologic pathways of blood coagulation 19FIGURE 4. Protein profile of normal human serumand immunoglobulin depleted serum 51FIGURE 5. Immunoblot analysis of the depletion of IgGfrom immunoglobulin depleted serum 52FIGURE 6. Immunoblot analysis of the depletion of 1gMfrom immunoglobulin depleted serum 53FIGURE 7. Immunoblot analysis of Cl q content in normal human serumand immunoglobulin depleted serum 56FIGURE 8. Complement consumption in human serum byCL-containing liposomes 58FIGURE 9. Complement consumption in human serum byPG-containing liposomes 60FIGURE 10. Complement consumption in human serum byP1-containing liposomes 64FIGURE 11. Alternative pathway complement consumption in humanserum by PG, P1, CL and SA liposomes 68FIGURE 12. Effect of immunoglobulin on Clq binding to liposomescontaining anionic phospholipid 70FIGURE 13. Immunoblot analysis of iC3b bound to negatively chargedliposomes exposed to normal human serumor immunoglobulin depleted serum 74FIGURE 14. Immunoblot analysis of C9 bound to negatively chargedliposomes exposed to normal human serum orimmunoglobulin depleted serum 75FIGURE 15A. Protein profile of eluate from ananti-beta-2-glycoprotein-1 column 77FIGURE 15B. Immunoblot analysis of mannose binding protein content inthe eluate from plasma exposed to 77anti-beta-2-glycoprotein-1 SepharoseFIGURE 16A. Protein profile and immunoblot analysis of the eluatefrom plasma exposed to anti-beta-2-glycoprotein- 1Sepharose 82FIGURE 16B. Immunoblot analysis of the eluate from plasmaexposed to rabbit IgG Sepharose 82FIGURE 17. Immunoblot analysis for beta-2-glycoprotein-1on the surface of liposomes 85FIGURE 18A. Binding of mannose binding protein to PC:CH:CL(35:45:20 mol%) liposomes 88FIGURE 18B. Inhibition of binding of mannose binding protein to PC:CH:CL(35:45:20 mol%) liposomes 88FIGURE 19. Detection of serum mannose binding protein binding tobeta-2-glycoprotein-1 using an enzyme linked immunosorbentassay 93FIGURE 20A. Complement consumption in human serum by PC:CH:CL(45:45:10 mol%) containing liposomes in the presenceof mannose binding protein inhibitors 94-xi-FIGURE 20B. Complement consumption in human serum by PC:CH:PG(35:45:20 mol%) containing liposomes in the presenceof mannose binding protein inhibitors 94FIGURE 21A. Complement consumption in immunoglobulin depletedserum by PC:CH:PG (45:45:10 mol%) containingliposomes in the presence of mannose bindingprotein inhibitors 96FIGURE 21 B. Complement consumption in immunoglobulin depletedserum by CH:PG (45:55 mol%) containingliposomes in the presence of mannose bindingprotein inhibitors 96FIGURE 22. Competitive binding study between Cl q andmannose binding protein in the presence ofPC:CH:CL (35:45:20 mol%) 99FIGURE 23. Identification of beta-2-glycoprotein-1, protein Sand C4 binding protein on PC:CH:CL(35:45:20 mol%) liposomes 102FIGURE 24. Identification of protein S on PC:CH:PG(1 5:45:40 mol%) liposomes 103-xii-ACKNOWLEDGMENTThis dissertation could not have been completed without the warmth,support, help and encouragement of a very special group of people, “the Deviners”:Haiming Chen, Maria Issa, Charles Xie, Kati Zay, Vicky Monsalve, KatherineSerrano, Glenn Edin, Gordon Homer, Keith Milton, and Pam Giberson. Thanks toBrenda Roeck and the entire Clinical Laboratory at The University Hospital for theirhelp and Quaker generosity in plasma and serum pools and allowing me access toand the use of clinical laboratory equipment. My gratitude to Brad Spiller for hisphotographic expertise, to Amanda Bradley for always being there for me whichincluded her companionship, compassion, editing skills and binding experimenttechniques and to Tim Black for riding the “graduate roller coaster” with me andproviding much support and understanding.Thanks to the “Cullis” group for their wisdom and advice in the liposomefield, Sean Semple for teaching me the spin column techniques and B2GP1 ELISASand to Jeff, Sandy and Sean for many a memorable Tuesday.I would like to acknowledge Dr. Cedric Carter whose discussions havehelped me become a better person and Dr. Alison Ccx who made me realize thatthere is life after graduate school.Finally, I am grateful to my supervisor: Dr. Dana Devine who had thecourage to take on an organic chemist and convert me into a biologist and whose“open door” policy was extended not only to discussions, guidance and support inscience but also everyday life.-XIII-DEDICATIONTo Dana Mohammed Jamil MarjanThank you-xiv-ABBREVIATIONSAbs. AbsorbanceACM N-acetyl-beta-D-mannosamineAPA Antiphospholipid antibodyAPC Activated protein CBIk BlankB2GP1 Beta-2-glycoprotein 1B2GP1 DEPL Beta-2-glycoprotein 1 depleted serum or plasmaBSA Bovine serum albuminC4BP C4-binding proteinC8BP C8-binding proteinCH Cholesterol*CH Tritiated cholesteryl hexadecyl etherCL Card iolipin derived from bovine heartCR1 Complement receptor 1CRP C-Reactive ProteinDAF Decay accelerating factorDCP Dicetyl phosphateDDS Serum depleted of lgG and 1gMECL Enhanced Chemiluminescent Detection SystemEDTA Ethylenediaminetetraacetic acidEGTA Ethyleneglycol-bis- (beta-aminoethylether) tetraacetic acidELISA Enzyme-linked immunosorbent assayLPS Lipopolysaccharides-xv-LTA Lipoteichoic acidLUV Large unilamellar vesiclesMAC Membrane attack complexMann D(+) mannoseMBP Mannose binding proteinMCP Membrane cofactor proteinMLV Multilamellar vesiclesNHP Normal human plasma*NHP Handling control of NHPNHS Normal human serumPBS Phosphate-buffered SalinePC L-o<phosphatidylcholine derived from eggPE PhosphatidylethanolaminePEG PolyethyleneglycolPG L,xphosphatidylglycerol derived from eggP1 Phosphatidylinositol derived from plant soybeanRES Reticuloendothelial SystemSA SterylamineSAP Serum Amyloid PSDS-PAGE Sodium dodecylsulphate-polyacrylamide gel electrophoresisSM SphingomyelinVBS Veronal-buffered Saline-xvi-1. INTRODUCTIONThe overall purpose of the studies conducted within this dissertation hasbeen to delineate the interaction of liposomes with proteins that may affect either thesurvival of liposomes in vivo or the response of the recipient to infused liposomes.Liposomes are known to bind various plasma proteins that may contribute to theirclearance by the reticuloendothelial system, including proteins of the complementpathway. The physiological effects of protein binding and the concomitantgeneration of bioactive molecules has been shown in studies reporting complementactivation in humans receiving liposomal drug formulations (Coune, A. et al 1983) aswell as in studies demonstrating anaphylactic reactions in animals receivingintravenous infusions of liposomes (Wassef, N. et al 1989). The in vivo behaviour ofliposomes and amount of protein they bind is influenced by physical characteristicsof the phospholipid vesicle including size and surface charge (Chonn, A. et al.,1992). Thus, the studies reported herein were designed to examine the interactionof liposomes of well-defined character with the complement system and the proteinC anticoagulant pathway, two plasma protein reaction cascades that generatephysiologically important activated protein products.The first objective was to determine the effect of immunoglobulin on theactivation of complement by negatively charged liposomes. These studies led to theinvestigation of the role of mannose binding protein on the activation of complementin the absence of immunoglobulin as this protein has been shown to activate theclassical pathway in the absence of activating immunoglobulins and Clq. During thecourse of investigating a physiological ligand for mannose binding protein, thebinding of an anticoagulant protein, protein S, to immobilized beta-2-glycoprotein-1was observed. The final objective was to determine the effect of the binding of—1—protein S to immobilized beta-2-glycoprotein-1. This is of importance as thedepletion of circulating protein S from plasma by binding to immobilizedbeta-2-glycoprotein-1 found on negatively charged liposomes could affect thehemostatic balance.In recent years, there has been a great surge in the use of particulate drugcarriers, such as liposomes, in experimental and clinical medicine. The rate and siteof uptake of particles by the mononuclear phagocyte system are influenced by theinjected dose, blood flow, local tissue damage and interaction of particles with serumproteins (Senior, J. 1987, Bradfield, J. 1974, van Oss, C. 1978, Absolom, D. 1986).The interaction of serum proteins with the particles depends on the physicochemicalproperties of the particles such as charge, size, hydrophobicity and fluidity of theparticle surface. For example, the behaviour of the colloidal particles injectedintravenously is strongly dependent on their size and the number of particles injected(Bergqvista, L. et al 1983, Nagai, K. et al 1982, Strand, S. et al 1977, Richards, R. etal 1986). Particles with small diameters are exchanged mostly through the bloodcapillaries. While large particles with diameters <100 nm are drained into thelymphatic system, particles with diameters >100 nm are trapped in the interstitialspace. The interstitially injected particles are in contact with plasma and may interactwith its proteins. This in turn influences their lymphatic drainage and may result intheir sequestration within the lymph nodes. In the lymph nodes, macrophagesrecognize the particles and ingest them (Bergqvist, L. et al 1983, Frier, M. 1981).Receptors on macrophages recognize the particles through the opsonins that havecoated them or by the nature of the surface itself (Bergqvist, L. et al 1983, Frier, M.1981, Patel, H. 1988).-2-With regard to liposomes, it is estimated that approximately 70-80 % ofintravenously injected liposomes are cleared into the Kupifer cells of the liver, 5-8 %are cleared into macrophages in the spleen and the remaining vesicles are clearedinto phagocytic cells in bone marrow (Poste, G. 1983, Senior, J. 1987, Bradfield, J.1984). The rapid clearance of liposomes from the circulation and their uptake byreticuloendothelial (RE) organs is analogous to the behavior of other particulatematerials such as colloids, immune complexes and pathogens. The role of plasmaproteins in removal of effete matter and pathogens has been extensively studied(Altura, B. 1980, Bradfield, J. 1974) but it is not yet clear to what extent variousplasma proteins influence the removal of liposomes from the circulation bymacrophages. The studies described in this dissertation were designed to gaininsight into two plasma protein reaction cascades that may influence the survival ofliposomes: the complement pathway and the coagulant pathway.The remainder of this Introduction provides an overview of the two enzymaticcascades leading to the overall objectives outlined for this study.1.1. Classical pathway of complement (Figure 1)1.1.1. Formation of activated ClMacromolecular Cl in serum is composed of a Clq molecule and a tetramercomplex of two Cir and two Cis molecules which are linked together by calciumions. Clq has a structure described as a “bunch of tulips” (Reid, K. 1989) where theClq, itself a multimeric protein, is made up of 18 polypeptide chains arranged into 6collagen triple helices and 6 globular heads (Reid, K. et al 1976).The Cl q binding site within immunoglobulin is located in the Fc portion of theantibody molecule (Augener, W. et al 1971). Immunoglobulins that are able toactivate the classical pathway include the lgG and 1gM isotypes (Ishizaka, T. et al-3-1966). Data suggest that one antigen-bound 1gM is sufficient to fix one Cl moleculewhile at least two lgG molecules in a close proximity “doublet” on the cell surface arerequired. Binding of Clq to immunoglobulin occurs via the globular region of Clq.For Cl q/lgG binding, the residues involved are located in the CH2 domain of the lgGand in the globular COOH terminal “heads” of Clq (Duncan, A. et al 1988). Uponbinding of two or more globular heads, a conformational change occurs in the Clqmolecule which is transmitted to the C1r2/C s complex. Conformational changeswithin the Cl q molecule have been demonstrated by the formation of neoantigenswhich were detected by different monoclonal anti-Clq antibodies (Golan, M. et al1982). The binding of Clq to 1gM requires that 1gM convert from a planar form to a“staple” form where the pentameric structure folds upon itself to form a staple-likeshape. When 1gM is in the “staple” conformation at least two of the five subunits ofthe 1gM molecules must be adjacent and intact (Feinstein, A. et al 1971, Augener, W.etall97l).-4-Figure 1. The complement pathway. Zymogen forms of complement proteins arecleaved by activated complement proteins that have serine protease activity. Oncecleaved they act on the next protein in the pathway. The membrane attack complexis produced by nonenzymatic protein-protein interactions (figure after Devine, D.1991)Classical PathwayC5b9(membrane attack complex)-5-1.1.2. Deposition of C4 and formation of C3 convertaseC4 migration in SDS-polyacrylamide gels under nonreducing conditionssuggests a complex of approximately 195 kD and reducing conditions reveal thiscomplex is composed of three components, an alpha, beta and gamma chain. Bycleavage of a single peptide bond in native C4, Cl esterase is able to release a 8.7kD fragment referred to as C4a from the alpha chain. The remainder of the moleculeis termed C4b. C4b acquires at least four different binding sites once the cleavageoccurs. Three of these sites are stable and specific and the fourth is nonspecific andhighly labile. The three stable sites include one for C4b receptors expressed oncertain cells, one for native and activated C2 and one for C4b-binding protein. Thefourth unstable site consists of an internal thioester in the alpha chain of C4. Innative C4, this ester is protected from hydrolysis by virtue of its position in ahydrophobic pocket of the protein. The removal of C4a leads to a conformationalchange and exposure of the thioester region. This ester may either hydrolyze withsurrounding water leading to the inactivation of the binding site or the glutamylmoiety within C4b may react with hydroxyl or amine residues by transesterification(amide formation). In the latter case C4b becomes covalently and irreversibly linkedto the hydroxyl (amino) group on a target surface (Law, S. et al 1980 a&b).Native C2 has an affinity for C4b and spontaneously forms a reversiblestoichiometric complex (C4b C2) which is dependent on magnesium ions (Sitomer,G. et al 1966). Once bound and in the vicinity of the Cl complex, C2 is cleaved byCis into C2a (73 Kd) and a smaller peptide C2b (34Kd). The complex formed,C4bC2a, is referred to as C3 convertase whose function, as its name suggests, is tocleave and activate the third and thereafter the fifth component of complement.Unlike the precursor complex C4bC2 which can form and dissociate reversibly, the-6-C4bC2a complex is labile. Its two partners dissociate spontaneously andirreversibly; the half life at 370C has been reported as 15 seconds (Kerr, M. 1980)and 3 minutes (Fishelson, Z. et al 1982). Once dissociated, C2a cannot recombinewith C4b.1.1.3. Cleavage of C3, C5 and formation of the terminal complexC4b2a cleaves a single peptide bond in the alpha chain of C3 (Bokisch, V. etal 1975, Tack, B. et a! 1979), releasing a 9 kD peptide fragment called C3a. Themajor fragment, C3b, migrates in SDS-PAGE under nonreducing conditions atapproximately 180 kD. The conformational change leads to the exposure of aninternal thioester bond which has the ability to react with a hydroxyl group on atarget surface (Pangburn, M. et al 1980, Tack, B. et al 1980). If not fixed, theexposed labile ester is hydrolyzed and C3b will be inactivated by a control protein,factor I, which cleaves C3b to its inactive form iC3b.The activation of C3 and its nearby fixation permits the formation of the C5convertase, C4b2aC3b, which functions by binding C5 to C3b, facilitating itscleavage by C4b2a (DiScipio, R. 1981, Vogt, W. et a! 1978) into C5a and C5b. C5bis then able to bind C6 and C7. This C5b-7 complex becomes membraneassociated owing to the exposure of hydrophobic amino acid residues. Oncemembrane bound C8 incorporates into the complex and the C5b-8 expresses abinding site for C9 which polymerizes 1-18 molecules at the C5b-8 complex to formthe typical cylindrical lesion which forms a pore across the membrane. Cell deathcan result due to electrochemical imbalance within the cell when enough numbers ofC5b-9, the membrane attack complex (MAC), are deposited onto the surface of atarget cell (Bhakdi, S. et al 1986, DiSciplo, R. 1993).-7-1.2. Alternative pathway activation cascadeComplement can also be activated by the alternative pathway; so namedbecause it was discovered after the classical pathway. The alternative pathway iscontinuously undergoing controlled initiation in the fluid phase due to thespontaneous hydrolysis of C3 to form a C3b-like molecule referred to as C3(H20)(Pangburn, M. et al 1981, Lachmann, P. et al. 1984). Once formed this moleculebinds factor B, which itself is cleaved by factor D to form C3(H20)Bb. C3(H20)Bb isthe initiating C3 convertase of the alternative pathway. If the C3 convertase is boundto a surface, it then serves as an anchor for the assembly of the surface associatedconvertases of the alternative pathway. Once C3bBb is formed, it is able to bind C5which is then cleaved by Bb (Isenman, D. et al 1980). The larger fragment, C5b, isable to bind to C6 through a metastable binding site and thus initiates the generationof the membrane attack complex (MAC) of the complement system and theformation of polymeric C9 as described previously. C3bBb can be stabilized byincorporation of the complement protein known as properdin (Minta, J. et al 1974,Smith, C. 1984), hence the alternative pathway is also known as the properdinpathway.The alternative pathway cascade is generally viewed as a nonspecific systemof natural defence because of the mode of its triggering and its ability to function inthe absence of specific immunoglobulins. Whether C3b is generated by the classicalor the alternative pathway a feedback loop is created where C3b will continue tointeract with factor B. The alternative pathway appears to be the phylogeneticallyolder of the two main recruitment systems for complement. Factor B-like activitiesare found in invertebrates such as starfish and sea urchins which lack specific-8-antibody molecules (Ballow, M. 1977). Factor B is a protein exclusive to thealternative pathway. The biological importance of the alternative pathway may bededuced from the variety of pathogenic microorganisms, virus-infected andtransformed cells and types of parasites which are sensitive to the neutralizing orkilling capacity of this pathway in the absence of the specific antibodies.1.3. Regulation of the complement pathwayAll enzyme systems must be regulated in vivo or substrate depletion will soonoccur. Complement activation on endogenous cell membranes is controlled by anumber of proteins, both plasma derived and cell membrane anchored. Plasmaderived regulatory proteins include C4 binding protein (C4BP) which regulates theclassical pathway at C4b and factor H which binds C3b. These proteins arecofactors for the serine protease factor I which cleaves and inactivates both C4b andC3b. The C3 convertases of both pathways are also regulated by decayaccelerating factor (DAF) which prevents the association of C3b with factor B and ofC4b with C2 or dissociates these complexes once formed (Lublin, D. et al 1989).Cells also protect themselves from complement through another two cell membraneanchored proteins: membrane cofactor protein (MCP) and complement receptor 1(CR1). Both proteins act as cofactors for factor I-mediated inactivation of C3b andC4b (Seya, T. et al. 1986).Regulation also occurs at the steps of the terminal pathway. CD59, a lowmolecular mass membrane protein, is a cell-membrane anchored regulatory proteinwith the ability to restrict homologous lysis. It was first described in 1989 by Sugita(Sugita, Y. et al. 1989) and has been called HRF2O, Protectin and MIRL by differentlaboratories. The mechanism of action of CD59 has not been determined but it hasbeen suggested that CD59 binds to C8 and C9 in the assembling MAC (Men, S. et-9-al. 1990). CD59 bound to C8 is believed to prevent the unfolding of the first moleculeof C9 bound to the MAC that is necessary for membrane insertion. There is then nofurther binding of C9 or polymerization to form a torus. Another cell-membraneanchored regulatory protein named C8-binding protein (C8BP) was isolated from adetergent extract of red cells by affinity chromatography on either C8 or C9 columns(Schonermark, S. et al. 1986). C8BP has also been referred to as MAC-inhibitingprotein (Watts, M. et al 1990) and homologous restriction factor (HRF) (Zalman, L. etal. 1986). Plasma proteins can also regulate the formation of C5b-9. These proteinsinclude vitronectin (S-protein) and clusterin (SP-40-40, apolipoprotein J) (Murphy, al 1988, Fritz, 1. 1992). If the C5b-7 complex is unable to bind rapidly to amembrane surface, it will interact with either of these plasma proteins therebyblocking the membrane binding site and rendering the complex lytically inactive.Subsequent binding of C8 and C9 results in the formation of the soluble terminalcomplement complex, SC5b-9 (Jenne, D. et al 1992).1.4. Activation of the classical complement pathway in the absenceof immunoglobulins1.4.1 C-Reactive Protein (CRP)CRP, so named because it reacts with the somatic C-polysaccharide ofStreptoccus pneumoniae, was first discovered in 1930 by Tillet and Frances. Likemany other plasma proteins, it is an acute phase reactant which increases its plasmaconcentration markedly in response to a variety of acute and chronic stimuliincluding infection, burns, surgery, major trauma and other inflammatory conditions.While the plasma concentrations of most acute phase proteins may double with aninflammatory response, the concentration of CRP may be increased (from 3 ug/ml)-10-(KoIb-Bachofen, V. 1991) by as much as 100-to a 1000 fold (Stuart, J. et al 1988,Claus, D. et al 1976, Whicher, J. 1990, Ballou, S. 1992).CRP is classified as a pentraxin based on its composition of five identicalnoncovalently linked subunits arranged in a flat pentameric disc (Osmand, A. et al1977). Each of the subunits migrate at 24 kD under reducing conditions in SDSpolyacrylamide gels and the subunits can form aggregates having molecular weightsas high as 114 kD (Gotschlich, E. et al 1967). The defining characteristic for CRP isits Ca2 dependent binding to phosphorylcholine which is present in abundance inthe cell wall of the Streptoccus pneumoniae. The biological activities of CRPincludes complement activation which is triggered upon the cross-linking of themolecule whether it is aggregated CRP, polycations or multivalent PC ligands (Jiang,H. et al 1991). CRP also binds 1 phosphorylcholine molecule within each subunitwith an association constant of 1.6 x 10 M1 (Volanakis, J. et al 1971, Anderson, al 1978) and two free Ca2 per subunit. Appropriate binding then induces aconformational change that is initiated at a site(s) separate from the singlephosphoryicholine-binding site in each subunit (Mullenix, M. et al 1994, Kilpatrick, al 1982).Activation of the complement system by CRP was first demonstrated forcomplexes of the protein with phosphorylcholine (Kaplan, M. et al 1974). Analysis ofcomplement component depletion indicated that complement activation proceededthrough the classical pathway. It was shown subsequently that complexes of CRPwith a variety of other ligands including polycations (Siegel, J. et al 1975), positivelycharged liposomes (Richards, R. et al 1979) and nuclear DNA (Robey, F. et al 1985)could also activate the classical pathway. This was demonstrated by efficientdepletion of complement components C1-C5 (Kaplan, M. et al 1974) as well as a—11—requirement for human Clq (Volanakis, J. et al 1974). In essence, it is thought thatCRP and immunoglobulins are similar in that both molecules react with Cl in asimilar manner (Claus, D. et al 1977a). Recently Jiang et at (1991, 1992) havedemonstrated the binding of CRP trimers to the collagen-like region of the A chain ofClq.1.4.2. Serum Amyloid P (SAP)Human SAP is a glycoprotein present in plasma at levels of about 40 ug/mI.This 250 kD molecular weight glycoprotein is made of 10 identical face-to facepentagonal structures (Pinteric, L. et al 1979). The similarity in structure betweenSAP and CRP led to its classification as a pentraxin. In humans, SAP, unlike CRP, isnot an acute-phase reactant (Pepys, M. et al 1978). In the presence of calcium ionsSAP binds to substrates such as amyloid fibrils of any type (Pepys, M. et at 1979),agarose via pyruvate acetyl residues (Painter, R. et al 1982, Hind, C. et al 1984),heparin and dermatan sulfate (Hamazaki, H. 1987), zymosan (Potempa, L. et al1985), mannose-terminated glycoproteins (Kubak, B et at 1988), 6-phosphorylatedmannose, and certain 3-sulfated saccharides including gatactose,N-acetylgalactosamine, and glucuronic acid (Loveless, R. et al 1992), fibronectin(deBeer, F. et at 1981), Clq (Bristow, C. et at 1986), C4BP (Schwalbe, R. et al 1990,1991), iC3b (Hutchcraft, C. et al 1981), DNA (Pepys, M. et at 1987), chromatin(Breathnach, S. et al 1989) and histones (Hicks, P. et at 1992). The pathophysiologicrole of SAP, including its importance in amyloid formation is not clear.The role of SAP in complement activation has been found to occur throughits ability to bind Clq via the collagen-like region (Ying, S. et al 1992). Recently, the-12-specific regions of binding have been identified as residues 14-26 and 76-92 of theClq A chain which are within the collagen-like region (Ying, S. et al 1993).1.4.3. Mannose Binding Protein (MBP)Mannose binding protein is an animal lectin (ie, a molecule with the ability tobind specifically to certain carbohydrate structures). MBP is part of a family ofproteins called collectins which also include conglutinin, Clq and pulmonarysurfactant associated proteins A and D (Malhotra, R. et al 1992). These collectinsare now known to contain collagen-like amino acid sequences connected to C-typecarbohydrate domains (Reid, K. et al 1992). The mature MBP consists of fourregions: a) an NH2 terminal segment of 21 amino acids with 3 cysteines, b) acollagen-like domain, C) a “neck” region and d) a COOH terminal C-typecarbohydrate recognition domain. The polypeptide chains of MBP appear to belinked together by disulfide bonds and migrate in nonreduced form as a large (>200kD) molecule and in reduced form as a 31 kD polypeptide chain (Lu, J. et al 1990).The hexameric form of MBP is very similar in structure to Clq with both moleculescomposed of globular heads each joined by connecting collagen strands to a centralstem structure.With regards to its function, Ikeda et al (1987) reported that serum purifiedMBP could mediate a dose-dependent lysis of sheep erythrocytes coated withmannan. This was true regardless of the source of MBP, i.e., whether it was isolatedfrom man, rabbit or rat. The lysis was dependent on the presence of 04 and it couldbe effectively inhibited by relevant monosaccharides. Further studies indicated thatCis binds to MBP only when Cir is also present (Ohta, M. et al 1990). In anotherfunctional study, the ligand for MBP was zymosan and the activation of complement-13-was followed by measurement of the activation of the proenzyme Cis in theClr2s complex (Lu, J. et al 1990). It was shown by electron microscopy thatonly the pentamer/hexamer forms of MBP, and not smaller forms of the moleculecould activate Cis. Another approach to the study of complement activation byMBP was taken by Super et al (1990). Using microtiter plates coated with mannanas ligand for MBP, sera were analysed for their ability to deposit complement factorsonto the plate surface. Analysis of 176 healthy blood donors revealed that the levelsof complement factors bound to the mannan-coated surface showed strongcorrelation with the amount of MBP in sera. No such correlation was found for thelevels of anti-mannan antibodies in the sera. MBP, when bound to bacteria, has alsobeen shown to mediate a complement-dependent bactericidal activity (Kawasaki, al 1989, Ji Y. et al 1988 and Ihara S. et al 1991). It was these observedcharacteristics of the glycoprotein that confirmed its ability to activate the classicalpathway in the absence of immunoglobulins and Cl q.In 1993, the International Complement workshop held in Kyoto formallyestablished the presence of a new pathway referred to as the lectin pathwaywhereby activation of the classical pathway is triggered via a binding of MBP tooligosaccharides in the absence of immunoglobulins and Clq (Figure 2).The reported isolation of Cis-like MBP-associated serine protease and theevidence of its activation of MBP have proposed a potential new mechanism ofactivation of the classical complement pathway (Matsushita, M. et al 1992). In thismechanism MBP has the ability to activate the pathway in the absence ofimmunoglobulins and Cl.-14-Figure 2. Recognized pathways leading to the activation of the complementsystem.r ALTERNATIVE CLASSICAL I r LECTIN I PENTRAXINPATHWAY PATHWAY PATHWAY PATHWAYC3(H20) Clq MBP CRPISAP/ClC4C2C3Effector Functions-15-1.4.4. Polyanionic SubstancesAmong the group of polyanions acting on the complement system are somenaturally occuring glycosaminoglycans such as heparin, chondroitin sulfate,dermatan sulfate and hyaluronic acid. These substances consist of repeatingdisaccharide units composed of a hexosamine residue linked to hexuronic acid.Heparin, the best studied polyanion, is best known to modulate both the classicaland the alternative pathways of complement. With regards to the classical pathway,heparin inhibits Cl activity by directly interacting with Clq (Almeda, S. et al 1983). Italso interferes with the binding of Cis to C4 and C2 and inhibits binding of C2 to C4b(Raepple, E. et al 1976). Another polyanionic substance is DNA which has the abilityto activate the classical complement pathway in combination with lysozyme(Willoughby, W. et al. 1973), CRP and polycations (Claus, D. et al 1977b). Othercharged particles that are known to activate complement in the absence ofimmunoglobulins include monosodium urate crystal (Giclas, P. et al 1979),retroviruses (Cooper, N. et al 1976) and heparin protamine (Rent, R. et al 1975). Afurther group of polyanionic substances containing phosphodiester linkages are thelipoteichoic acids (LTA). LTA are the major cell wall constituent of gram positivebacteria (Loos, M. 1986) and are also known to activate the classical pathway. TheSalmonella r-type mutant Iipopolysaccharides (LOS, lipooligosaccharide) also hasthe ability to activate the classical complement pathway by strongly binding Clq andactivating complement.1.5. Coagulation: an overview (Figure 3)Blood clotting is a host defense mechanism that, in parallel with theinflammatory and repair responses, helps protect the integrity of the vascular system-16-after tissue injury.The blood coagulation cascade consists of proenzymes which containcatalytic domains that express protease activity. These catalytic domains arelocated on the C-terminal ends of the blood clotting enzymes. The N-terminal halvesof the proteins are unique and confer the specialized biochemical properties of eachprotein (Furie, B. et al 1988). Some of the blood-clotting proenzymes require vitaminK for their complete synthesis; these proteins include factors IX, X and VII andprothrombin, all of which contain 10 to 12 gamma-carboxylic glutamic acid residues(Stenflo, J. et al 1974, Nelsestuen, G. et al 1974). These unique amino acid residuesof gamma-carboxyglutamic acid are thought to be critical for calcium ion binding andare required for the interaction of these vitamin K dependent proteins with cellmembranes (Esmon, C. et al 1975). There are also some EGF (epidermal growthfactor) like domains found in factors VII, IX, X, XII and kringle domains found inprothrombin and factor XII. This suggests that the proenzymes involved inhemostasis have evolved into a family of closely related proteins with numerousdomains that are most likely derived from a common ancesteral gene (Doolittle, R.,1993).The sequence of events that leads to the generation of a fibrin clot is usuallydescribed as containing two distinct pathways. In the intrinsic pathway, factor XII isactivated during contact phase clotting, followed by the sequential activation offactors XI, IX, X and prothrombin. In the extrinsic pathway, a complex betweentissue factor and factor VII forms, followed by the sequential activation of factors VII,X and prothrombin. However, the physiologic pathway relevant to blood coagulationin vivo is clearly different due to the following findings: first, patients with a hereditarydeficiency of factor XII, prekallikrein or high-molecular weight kininogen have a-17-markedly prolonged partial thromboplastin time but no bleeding problems. Second,tissue factor, a normal constituent of the surface of nonvascular cells and stimulatedmonocytes, activates blood coagulation. Third, the tissue factor-factor Vlla complexactivates not only factor X but also factor IX suggesting a central role for factors VIIIand IX in coagulation initiated by the tissue-factor pathway (Osterud, B. et al 1977).This becomes more apparent as individuals with hereditary factor XI deficiencyusually have a mild hemostatic disorder while some are reported to have a moresevere hemostatic disorder (Rapaport, S. et al 1961, Schmairer, A. et al 1987).The key to initiation is currently accepted as tissue factor. The exposure ofcell surfaces expressing tissue factor to the plasma proteins results in the binding offactor VII to tissue factor. The tissue factor-factor VII complex itself may have low butfinite coagulant activity (Zur, M. et al 1982) or else the complex formed by tissuefactor and factor Vila, activates factors IX and X (Rao, L. et aI 1988). The proteaseresponsible for the initial activation of factor VII is unknown. However, once clottingis activated, several proteases further along the pathway can activate factor VII.Since factors Xa and VIIa both catalyze the activation of factor VII, there is a potentialmechanism for the acceleration of factor VII activation (Radcliffe, R. 1976).Furthermore, in the presence of negatively charged surfaces, thrombin and factorXIa catalyze the activation of factor Xl (Naito, K. et al 1991). Once factor Xla isgenerated, there is an additional mechanism for augmenting the activation of factorIX. Factor IXa in complex with factor VIlla activates factor X on membrane surfaces.Factor Xa in complex with factor Va activates prothrombin on membrane surfaces.Thrombin then cleaves fibrinogen, yielding monomeric fibrin, which then polymerizesto form the fibrin clot.-18-Figure 3. Physiologic pathways of blood coagulation. Blood coagulation is initiatedby tissue factor to which factor VII (EVIl) will bind. The complex of activated factorVII (FVIIa) and tissue factor activates factors IX (FIX) and X (FX). The proteolyticactivation of factor VII to Vila, factor XI (FXI) to XIa (FXIa), factor VIII (FVIII) to Villa(FVIlIa) and factor V (FV) to Va (FVa) through feedback mechanisms such as theconversion of prothrombin (PT) to thrombin (T) or other enzymes (dashed arrows),greatly accelerates blood clotting. The process culminates with the generation offibrin and its polymerization to form a fibrin clot. The open arrows indicate the actionof enzymes on substrates, and the narrow solid arrows indicate the conversion of aprotein from one functional state to another after the cleavage of one or morepeptide bonds (figure after Furie, B et al 1992)ClotForrrtat;or-19-Given the properties of the extrinsic pathway there does not appear to be aneed for an alternative pathway. However, an alternative pathway (the intrinsicpathway) does exist. This is mainly due to the presence of a natural inhibitor of theextrinsic pathway. This inhibitor is referred to as tissue factor pathway inhibitor(TFPI) also called lipoprotein-associated coagulation inhibitor (LAG. TFPIneutralizes the factor Vll/tissue factor complexes (Rapaport, S. 1991). Recentreports have shown that the inhibition and the binding of factor Vila/tissue factorrequires the presence of factor Xa (Girard, T. et al 1989). The physiologic role of thisinhibition, in the presence of Xa, has yet to be ascertained.1.5.1. Regulation of the coagulation pathwayThere are two main regulators of the coagulation cascade: the fibrinolyticpathway and the procoagulant inhibitors. The fibrinolytic pathway is closelycontrolled. However, the mechanism for maintaining the balance betweencoagulation and fibrinolysis is not known (Wiman, B. et al 1990). The procoagulantinhibitors are divided into two groups, those that control the serine proteases of thecoagulant pathway and those that control the cofactors V and VIII. Antithrombin Ill isone of the plasma serine protease inhibitors belonging to a family of proteinsreferred to as serpins. Antithrombin Ill serves as a protease scavenger whereby anyof the blood clotting enzymes that move away from the growing clot rapidly form acomplex with Antithrombin Ill and their activities are neutralized (Damus, P. et al1973). The protein C system is involved in regulating cofactors V and VIII. Activatedprotein C, a vitamin K-dependent plasma protein, inactivates the active forms offactors VIII and V and thus rapidly slows the blood clotting process (Walker, F. et aI1987, Fulcher, C. et al 1984). Protein S is an important cofactor for the anticoagulant-20-activity of activated protein C (Walker, F. 1980).The molecular mechanism for the activated protein C (APC) cofactor functionof protein S is not completely understood. According to the current acceptedhypothesis, protein S and APC form a 1:1 stoichiometric complex on the surface ofnegatively charged phospholipid membranes. The presence of protein S increasesthe affinity of APC for the membrane (Walker, F. 1981a &b, 1988). Protein S exists intwo distinct forms in human plasma (Dahiback, B. et al 1981). Approximately 40 %occurs as free protein S and the remainder is bound to a high molecular weight (570Kd) protein identified as C4 binding protein (C4BP). As discussed above, C4BP is aregulator of the classical complement pathway (Law, S. et at 1988). It is found inplasma at a concentration of approximately 150 mg/L and is comprised of sevenalpha chains and one beta chain. Protein S and C4BP form a 1:1 stoichiometriccomplex (Dahlback, B. 1984). The interaction between the two proteins isnon-covalent and the binding site for protein S on the C4BP molecule is thought tobe distinct from those for the complement protein C4b as demonstrated by electronmicroscopy (Dahlback, B. et at 1983a). This is also supported by the presence oftwo forms of C4BP which can be isolated from plasma exposed to barium citrate:C4BP-high and C4BP low. The C4BP-low is isolated from the supernatant and hasbeen shown to lack the beta chain and thus the molecule lacks the ability to bindprotein S (Hillarp, A. et at 1989). However, studies using proteolytic fragments ofC4BP are not in agreement with the binding sites observed by electron microscopy(Chung, L. et al 1985). With regard to C4BP function, there has been a documentedin vitro study to show that protein S is neither involved in, nor does it affect, theinteraction between C4BP and C4b (Dahlback, B. et al 1983b). However, there hasbeen a study by Fujita et al (1985) which has demonstrated that monoclonal-21-antibodies raised towards the beta chain of C4BP inhibits C4b-binding. Theconflicting studies mentioned above suggest that the binding site for C4b on C4BP isnot that clear and that the binding of protein S to C4BP may affect complementregulation.Calcium appears to increase the rate of association between protein S andC4BP at least 100-fold whereas the rate of dissociation of protein S from C4BP isvery slow both in the presence of calcium and EDTA (Schwalbe, R. et al 1990 a andb).The function of protein S as a cofactor is lost when protein S binds to C4BP(Dahlback, B 1986, Bertina, R. et al 1985). The pathophysiological importance ofprotein S has been demonstrated by findings in cases of familial selective deficiencyof free protein S where patients have increased risk of thrombosis (Comp, P. et al1984, Kemkes-Matthes, B. 1992).1.6. LiposomesLiposomes were originally described by Bangham and Home (Bangham, al. 1964) following electron microscopy of a suspension of phospholipids ofcellular origin after isolation and purification. These vesicles form spontaneously intomultilamellar bilayers and enclose an aqueous space. The structures thus formedare referred to as liposomes and are the result of amphiphilic molecules arrangingthemselves in an aqueous solution which is present in excess.1.6.1. Physical aspect of liposomesThe primary constituent of a liposome is lipid: a molecule consisting of apolar, hydrophilic “head” attached to a long nonpolar, hydrophobic “tail”. The-22-hydrophilic head typically consists of a phosphate group (hence phospholipid)whereas the hydrophobic tail is made of two long hydrocarbon chains. Becauselipid molecules have one part that is water-soluble and another part that is not, theytend to aggregate in ordered structures that sequester the hydrophobic tails from thewater molecules.Studies have shown that liposomes used for in vitro studies may responddifferently to the environmental milieu they are exposed to depending on their size,surface charge, phospholipid structure and cholesterol content. This section willreview the literature on each of the above mentioned factors.Liposomes are usually made of a mixture of phosphatidylcholine, cholesteroland some electrically charged lipid. The inclusion of charged lipids improves thephysical stability of the liposomes because electrostatic forces prevent the vesiclesfrom making contact and agglutinating.In bilayer systems, the presence of cholesterol exhibits a well characterizedability to inhibit formation of the crystalline gel state and to decrease the permeabilityof liquid crystalline lamellar systems (Demel, R. et al 1976). Cholesterol has a widevariety of effects on the physical properties of membranes. In one of the earlieststudies of the effect of cholesterol on lipid properties, it was demonstrated thatcholesterol causes the phase transition of sphingomyelin (SM) to “disappear”.Therefore, it was suggested that cholesterol can induce a liquid-crystal state in lipidsthat would otherwise occupy a gel state (Oldfield, E. et al 1972). It has also beendemonstrated that cholesterol has a special ordering effect on lipids in theliquid-crystalline state and that cholesterol may be unique in its ability to cause suchan effect.-23-1.6.2. Classification of LiposomesOnce the lipids are hydrated, multilamellar vesicles (MLVs) are formed. MLVsare heterogeneous in size (0.2-10 urn) and lamellarity. Small unilamellar vesicles(SUVs) can be made from MLVs by subjecting the MLVs to ultrasonic irradiation(Huang, C. 1969) or by passage through a French press (Barenholz, Y. et al 1979).However, their small size limits their use in model membrane studies. Their sizesrange from 25-40 nm in diameter and the radius of curvature experienced by thebilayer in SUVs is so small that the ratio of lipid in the outer monolayer to lipid in theinner monolayer can be as large as 2:1. As a result, small unilamellar vesicles(SUVs) are highly strained which can lead to an increased tendency to fuse(Lichtenberg, D. et al 1981, Wong, M. et al. 1982, Parente, R. 1984). Further, SUVsare more susceptible to attack by phospholipases and high density lipoporoteinsdue to their highly curved surfaces (Scherphof, C. et al 1984). Lastly, the SUVmembrane is often too small to allow studies of permeability or ion distributionsbetween internal and external aqueous compartments.A more useful membrane model is the large unilamellar vesicle system(LUVs) where the mean diameter is between 50-500 nm and the distribution of lipidbetween the outer and inner monolayers is closer to 1:1. LUVs may be preparedwith the use of detergents or organic solvents.Procedures that employ detergents vary depending upon the type ofdetergent, however the principle is the same. Lipids are solubilized by the detergentof choice (such as cholate or octylgiucoside) then the detergent is removed eitherrapidly by dilution or gel filtration or slowly by dialysis. As the concentration ofdetergent decreases the lipids adopt unilamellar vesicle structures. The size of theliposomes can be controlled by the rate at which the detergent is removed (Kagawa,-24-Y. et al. 1971, Mimms, L. et al. 1981). LUVs may also be prepared by firstsolubilizing the lipid in an organic solvent such as diethyl ether; buffer is added to thelipid solution and a stable emulsion is formed by sonicating the mixture; the organicsolvent is then removed by evaporation under vacuum. The resulting dispersion is aheterogenous mixture of oligolamellar and unilamellar vesicles which requires lowpressure extrusion through polycarbonate filters to achieve relatively homogeneouspreparations of unilamellar vesicles. This method is referred to as reverse phaseevaporation (Szoka, F. et al. 1978) and the LUVs produced are commonly referredto as reverse phase evaporation vesicles (REVS).Another technique involves the direct extrusion under moderate pressuresless than and equal to 500 psi of MLVs through polycarbonate filters of defined poresize. This technique can generate LUVs with size distributions in the range of 50-200nm depending on the pore size of the filter employed. Using this extrusion method,LUVs composed of a wide variety of lipid species can be readily produced at highconcentration and in the absence of contaminating detergents or organic solvents(Hope, M. et al 1985, Nayar, R. et al 1989).1.7. The interaction between liposomes and complement.The versatile nature of liposomes was established by membrane biologistswho used liposomes as a model for cell membrane studies (Bangham, A. et al 1974)and who promoted the development of another concept: the use of liposomes indrug delivery. One of the basic aims of chemical therapeutics is to deliver themedicinal substance efficiently and specifically to the site of the disease or disorder.In some instances this can be administered using free drug. In many cases the sideeffects of the administered drug may outweigh its administrative benefits. However,-25-the effectiveness of a drug can be improved by encapsulating it in a carrier such asliposomes.Proteins of the complement pathway are one group of many proteins that willinteract with liposomes. This group of proteins can control the half-life of theliposomes by allowing for the interaction of opsonins such as immunoglobulins andC3b with the liposomes. This may mediate the reticuloendothelial system clearanceof the liposomes.There have been numerous studies examining the effect of acyl chain lengthand unsaturation, effect of cholesterol, charge and size of liposomes on activation ofthe rat, mouse and human complement systems. Studies by Shin, M. et al (1978)have shown that increasing the acyl chain length of lecithin liposomes containingcholesterol decreases the release of trapped 86Rb (Rubidium) by antibody andcomplement. The studies by Devine et al have also shown that increasing theamount of cholesterol within a negatively charged composition also increasedcomplement activation. (Devine, D. et al 1994). Studies have also shown that largerliposomes are better activators of complement (Devine, D. et al 1994) than theirsmaller counterparts and that they are cleared more rapidly to the RES (Juliano, R. etal 1985). Studies by Chonn et al (1991) have been able to show that charge-bearingphospholipids fall into two distinct groups with respect to complement activation:negatively charged liposomes activate the classical pathway in normal human serumwhile those that are positively charged activate the alternative pathway. Otherstudies by Funato et al (1992) have shown that multilamellar vesicles bearing dicetylphosphate (a negatively charged phospholipid) will activate the alternative pathway.Studies by Mold et al (1980) and Cunningham et al (1979) demonstrated thatliposomes containing SA (positively charged phospholipid) could activate the-26-alternative pathway only in the presence of glycolipid. The SA containing liposomeswere also demonstrated to activate the classical pathway in the presence of CRP(Mold, C. et al 1981a). Published studies have demonstrated that the CRP will bindthe phosphorylcholine moiety of PC present in the liposome and will also bind Clqthus initiating the classical pathway. However, the binding observed is to thecationic phospholipid sterylamine as demonstrated by another publication (Mold, C.etal 1981b).Whether the opsonizing complement protein is C3b or immunoglobulinremains unclear. Numerous studies have shown that phagocytosis ofimmunoglobulin opsonized liposomes is readily enhanced in comparison tonon-immunoglobulin sensitized liposomes (Geiger, B. et al 1981, Hafeman, J. et al1980, Hsu, M. et al 1982, Leserman, L. et al 1980 a&b, Lewis, J. et al 1980, Kuhn, al 1983). Opsonization by complement is readily achieved in the presence ofantibodies on the surface of liposomes (Wassef, N. et al 1987 and 1993). Thesestudies have shown that the major mechanism for liposome uptake is throughopsonization with immunoglobulin and subsequent Fc-mediated phagocytosis orendocytosis.Some studies performed in vivo and in vitro have alluded to C3 fragments asbeing the specific opsonin responsible for phagocytosis. Loughrey et al (1990)demonstrated that PG containing liposomes associated with rat platelets in plateletrich plasma and that this interaction resulted in the formation of plateletmicroaggregates and was dependent on the presence of plasma. Themicroaggregates provided a potential role for the removal of liposomes from thecirculation via the binding of iC3b, found on the surface of the liposome, to CR1 typereceptors on platelets. The removal of the microaggregates, however was a-27-transient phenomenon. The formation of microaggregates was sensitive to thetreatment of plasma with heat, purified cobra venom factor and removal of the C3using anti-rat C3. Funato et at (1992) have demonstrated the deposition of C3fragments on negatively charged liposomes containing DCP exposed to rat plasma.The C3b deposited on the liposome surface was cleaved to iC3b. The authorssuggested that the presence of the degradative products of C3 present on thesurface of the liposome may be related to the phagocytosis of the liposomes bymacrophages of the liver, Kupifer cells, that bear on their surface CR1 and CR3. Amember of the same group (Matsuo, H. et at 1994) later showed that when mannosemodified negatively charged tiposomes were treated with serum heated to 560C ortreated with anti-rat C3, the deposition of iC3b was inhibited and hepatic uptake invivo was reduced.Literature has shown that by incorporating amphipathic molecules such aspolyethylene glycol (PEG) of various molecular weights and ganglosides such asGM1 into the liposomes, the circulation half-life increases (Klibanov, A. et al 1990,1991, Allen, T. et al 1989). It is thought that these molecules provide a strong stericbarrier which prevents close contact with plasma proteins. In addition GM1 is also agood inhibitor of complement activation by increasing the binding affinity of factor Hand C3b (Michalek, M. et al 1988).1.8. The interaction between liposomes and beta-2-glycoprotein-1Human beta-2-glycoprotein-1 (B2GP1) was first isolated by Schultze et al in1961. B2GP1 is a single chain glycoprotein with 325 amino acid residues and anestimated molecular weight ranging from 43 to 50 kD (Lee, N. et at 1983, Lozier, J. etal 1984, Finlayson, J. et at 1967). It contains five glucosamine-attached-28-oligosaccharide side chains that account for approximately 19% of its molecularweight and are composed of galactose, mannose, N-acetylglucosamine, fucose andN-acetylneuraminic acid. The physiological concentration of B2GP1 is approximately0.2 mg/mI; however, genetic studies have shown that variation has been observedand attributed to the occurrence of a deficiency allele (Kamboh, M. et al 1988).Little is known about the definitive physiological role of this glycoprotein;however, this glycoprotein is a constituent of chylomicrons, very low densitylipoproteins and high density lipoprotein (Polz, E. et al 1979). B2GP1 has the abilityto activate lipoprotein lipase (Nakaya, Y. et al 1980), and is required for theprecipitation of lipoproteins by anionic detergents (Burstein, M. et al 1977). B2GP1has also been shown to inhibit the contact activation system of blood coagulation(Schousboe, I. et al 1988). As well, B2GP1 is capable of binding to anionic chargedphospholipids (Wurm, H. 1984). The study by Wurm showed that B2GP1 did notbind to neutral phospholipids such as PC, PE and SM. He also showed thatnegatively charged phospholipids such as PS and P1 could interact strongly withB2GP1 and that the binding capacity was influenced by several parameters such asthe molarity of buffer and the presence of divalent cations as well as EDTA. Theability of B2GP1 to bind to negatively charged phospholipids was reduced slightlywith increasing molarity of the divalent cations, the effect being more pronounced forcalcium than magnesium. In the case of EDTA, a sharp decrease in binding wasobserved at very low concentration of EDTA in the presence of negatively chargedphospholipid and the binding remained constant at increasing concentrations ofEDTA.B2GP1 has also been recently implicated as a putative cofactor ofantiphospholipid antibodies (Costing, J. et al 1992, Matsuura, E. et al 1992, McNeil,-29-H. et al 1990). Over the past decade, there has been an increasing interest in therelation between the presence of antiphospholipid antibodies (detected by lupusanticoagulant clotting tests and anticardiolipin enzyme linked immunosorbentassays) and the occurrence of thrombotic complications. Patients withantiphospholipid antibodies appear to be predisposed to thrombotic complications(Harris, E.N. et al 1983, Derksen, R. et al 1985, Feinstein, D. 1985, Haire, W. et al.1990, Parke, A. et al 1992). The in vivo antigenic specificity of antiphospholipidantibodies has not been defined. Some investigators believe that it is thecombination of lipid and B2GP1 (McNeil, H. et al 1990, Shi, W. et al 1993). Othersbelieve the antigenic component is simply B2GP1 (Keeling, D. et al 1992, Pierangeli,S. et al. 1992). There has been a reported study stating that the antibodiesrecognize the complex of lipid and prothrombin (Bevers, E. et al 1991). There havealso been two other proteins identified as antiphospholipid antibody antigens-protein C and protein S (Costing, J. et al 1993, Vermylen, J. et al 1992). Despite theheterogeneity of the proteins that are recognized by APA, it is not known whypatients with high APA titres are prone to thrombosis.1.9. Overall ObjectiveThe overall objective of this dissertation research is to examine theinteractions of liposomes with the physiologically important plasma protein cascadesof complement and the coagulation pathway. The protein components of thesepathways can markedly affect the in vivo survival of foreign bodies and are thereforereasonable candidates to study in investigations of the mechanisms of clearance ofinfused liposomes. The experiments described in this dissertation were designed to-30-answer three specific questions:(1) The activation of human complement by liposomes has been reportedby other investigators. However, the precise mechanisms of activation of theclassical pathway by negatively charged liposomes had not been determined. Ichose a model liposome system using 100 nm unilamellar vesicles (which arecommonly used in therapeutic applications) composed of negatively chargedphospholipids mixed with cholesterol and neutral phosphatidylcholine. Using thismodel, I set out to determine whether naturally occurring antibodies againstphospholipids could account for the complement activation seen when liposomeswere reacted with human serum in vitro.(2) Having demonstrated that specific antibody was not required, the nextquestion to be addressed was to determine the role, if any, of ancillary complementactivation pathways in the antibody-independent activation of the classical pathway.In particular, these studies focused on mannose-binding protein as that moleculehas been identified by others as a direct activator of the classical pathway on thesurface of some foreign particles.(3) In studies conducted for part (2), I found a previously unrecognizedassociation between surface-bound or antibody-immobilized beta-2-glycoprotein 1and the anticoagulant protein, protein S. Since the depletion of protein S from thecirculation by binding to beta-2-glycoprotein-1 -rich liposomes could markedly upsetthe hemostatic balance, the final section of this dissertation was directed todetermining whether this association was significant.-31 -2. MATERIALS AND METHODS2.1 Reagents2.1 .1 Antisera and AntibodiesPolyclonal antisera and antibodies: Polyclonal antisera to human Clq, C3,C7, C9 and factor B were purchased from Quidel, San Diego, CA. Polyclonalantibody to human lgG and to human albumin was purchased from AtlanticAntibodies, Scarborough, Maine. Polyclonal antibody to human B2GP1 andperoxidase-conjugated antibody to human protein S were purchased from EnzymeResearch, South Bend, IN. Polycional antibody to human C4BP was purchasedfrom The Binding Site, Birmingham, U.K. Polyclonal antisera to human SAP waspurchased from DAKO, Denmark. Polyclonal antisera to human CRP waspurchased from Accurate Chemical and Scientific Corp, Westbury, New York.Polyclonal antisera to human fibrinogen was purchased from Behring Diagnostics,Germany. Peroxidase-conjugated rabbit anti-human 1gM was purchased fromZymed Laboratories, San Francisco, CA. The remaining peroxidase conjugatedantibodies which included rabbit anti-goat, mouse anti-rabbit, rabbit anti-sheep andgoat anti-mouse were purchased from Jackson lmmunoresearch, West Grove, PA.Polyclonal antibody to MBP was a gift from Dr. Steffen Thiel, Aarhus University,Denmark.Monoclonal antibodies: Monoclonal antibody (IgG) to MBP was a gift fromDr. Steffen Thiel, Aarhus University, Denmark.2.1.2. Other reagentsPure MBP was the gift of Dr. Steffen Thiel, Aarhus University, Denmark.Coomassie R250 was purchased from Pharmacia Biotech Inc. Baie d’Urfe, P0. The-32-prestained and broad molecular markers were purchased from BioRad,Mississauga, ONT. The Rainbow molecular weight markers and the ECL detectionsystem were from Amersham, UK. The protein determination reagents were eitherpurchased from BioRad, Richmond, CA for the BioRad Protein Assay or Pierce,Rockford, IL for the Bicinchoninic Acid Protein Assay. The lipids used in the studiesherein were purchased from Avanti Polar Lipids, Alabaster, Alabama with theexception of cholesterol which was purchased from Sigma, St.Louis, MO and*tritiated cholesteryl hexadecyl ether (CH ) from Dupont, Boston, MA.N-acetyl-beta-D-mannosamine (monohydrate) and D(+) mannose (mixed anomers)were purchased from Sigma, St.Louis, MO. Pure B2GP1 (2.93 mg/mI) waspurchased from Crystal Chem, Chicago, IL. Pure CRP (1.6 mg/mI) was purchasedfrom DiaMed Inc., Windham, ME. Whole lgG was purchased from AccurateChemical and Scientific Corp, Wesbury, NY. Any microtitre plates used for ELISAexperiments were of high binding quality and were purchased from DynatechLaboratories Chantilly, VA. The substrate used to detect the boundperoxidase-conjugated antibody in the ELISA technique was 3,3’,5,5’tetramethylbenzidine or 2,2-azino-di-ethylbenzthiazoline-6-sulfonic acid and waspurchased from Sigma, St.Louis, MO. 10 Kd and 30 Kd cutoff microconcentratorswere purchased from Filtron, Nothborough, MA and Amicon, Danvers, MArespectively. YM1O cuttoff filters for the miniconcentrator was also purchased fromAmicon, Danvers, MA. Polyallomer centrifuge tubes 5 x 20 mm for the competitivebinding studies were purchased from Beckman, PaloAlto, CA.2.1.3. Kits PurchasedThe ELISA kit used to measure the activation fragment Bb was purchasedfrom Quidel, San Diego, CA. The kit used to measure immunological free Protein S-33-(Liatest) and the functional Protein S (Staclot) were both purchased from DiagnosticaStago, France. The preparation of the materials were carried out at 4 °C and thefunctional assays were performed at 37 °C.2.2. Preparation of reagents for Experiments2.2.1. Preparation of normal human serum and immunoglobulin depletedserum: The normal human serum pool (NHS) was prepared from venous bloodfrom at least 20 healthy individuals distributed evenly over males and females. Theblood was allowed to clot at room temperature for 30 mm and centrifuged at 275 x gat 4 °C for 10 mm to remove red and white cell contaminants prior to storage at -700C. To prepare serum depleted of immunoglobulin classes IgG and 1gM (doubledepleted serum; DDS), serum was passed through an anti-lgM-Sepharose affinitycolumn; the anti-lgM affinity column was prepared by coupling anti-human 1gMantibody to CNBr-Sepharose 4B according to manufacturer’s instructions. Briefly,0.3 g of CNBr-activated Sepharose was allowed to swell at room temperature for 30mm in 1 mM of HCI. The mixture was washed and reswelled on a sintered glass filterusing 1 mM HCI (200 ml/g). The anti-human 1gM antibody was dissolved in 1 ml ofcoupling buffer (0.1 M NaHCO3,pH 8.3, containing 0.5 M NaCI). The proteinsolution was placed into a 15 ml Corning tube and mixed with the gel suspensionand rotated at room temperature for 2 h. The gel was filtered and blocked with 0.2 Mglycine pH 8.0 and left rotating at 4 °C overnight. The gel was then washed withcoupling buffer followed by acetate buffer (0.1M, pH4) containing 0.5 M NaClfollowed by coupling buffer and then stored in the eluting buffer containing 0.1 M TrispH, 8.0. The eluate obtained was subsequently passed though a ProteinG-Sepharose column to remove all subclasses of lgG. The levels of IgG and 1gM-34-were quantitated using a QM 300 Nephelometer (Sanofi Diagnostics, Chaska, MN).In all experiments described below, DDS was concentrated in 10 Kd cutoffmicroconcentrators to 12 mg/mI for all assays except the Clq ELISA in which it wasused at 6 mg/mI. NHS was used at equivalent protein concentrations as determinedby Bio-Rad Protein Assay.Along with Cl q, CaCI2 and MgCI2were added to DDS after all manipulationsof the serum were completed, to a final concentration of 0.15 mM and 1 mM,respectively. The Clq was repleted as the immunoglobulin affinity columns depletedClq from the serum samples (see results 3.5). This depletion is expected as Clqhas a high affinity for immunoglobulins as previously described.2.2.2. Preparation of normal human plasma: The normal human plasma pool(NHP) was prepared from venous blood collected from at least 20 healthy individualsdistributed evenly over males and females. The blood was drawn into siliconizedVacutainer tubes containing 3.8 % buffered sodium citrate. Blood:anticoagulant ratiowas 9:1. The samples were centrifuged at 2000 x g for 15 mins to remove red andwhite cell contaminants prior to storage at -70 CC.2.2.3. Preparation of B2GP1-depleted plasma: Plasma depleted of B2GP1was prepared by immunoaffinity chromatography. Anti-human B2GP1-Sepharosewas prepared by coupling anti-human B2GP1 lgG to CNBr-Sepharose 4B asdescribed above. 300 UI of NHP was then mixed and allowed to rotate with 2ml ofthe anti-B2GP1-Sepharose for 90 mins at 4 °C. Depletion of B2GP1 was confirmedby Western blot analysis; the 50 Kd band reactive with anti-human B2GP1 in NHPwas no longer present in B2GP1 depleted plasma but was present in plasma-35-samples mixed with rabbit lgG Sepharose used as a control. The amount of B2GP1present in the normal human plasma and in B2GP1 depleted plasma was quantitatedusing an ELISA with a known standard curve. The amount of B2GP1 present in NHSand in B2GP1-depleted serum, when the serum sample was passed throughanti-human B2GP1 columns, was 150 +1- 12 ug/mI and 10.7 +/- 0.5 ug/mIrespectively (n=2 in triplicates) as determined by an ELISA for the quantitation ofB2GP1 (see section 2.3.8.). The controls used consisted of rabbit IgG coupled to*CNBr-Sepharose 4B and the handling control (temperature control, NHP ) at 4 Cwhich consisted of appropriately diluted plasma. The NHP* consisted of NHPdiluted to the appropriate concentration required for the experiment and left to rotateat 4 0C for the length of time that the plasma samples were exposed to theimmunoaffinity Sepharose. When the plasma samples were exposed to theimmunoaffinity Sepharose, the levels of B2GP1 were 230 ug/mI for the handlingcontrol sample, 116 ug/mI for the plasma exposed to rabbit lgG Sepharose and 20ug/mi for the plasma exposed to anti-human B2GP1-Sepharose as determined byan ELISA for the quantitation of B2GP1 (in triplicate, see section 2.3.8.).2.2.4. Preparation of liposomes: In this study cholesterol was maintained at45 mol% as it has been observed that this minimizes the interactions with highdensity lipoproteins (Kirby, C. et al 1 980a &b) and prevents complement activationby naturally occuring antibodies to cholesterol (Alving, C. 1977, Alving, C. et al1989). Liposomes were produced by extrusion through 100 nm or 400 nm poresized filters (LUV). These well defined liposomal systems are ideal for the studiesinvolving protein/liposome interactions described within this dissertation.Multilamellar vesicles were prepared according to established methods (Olson, F. et-36-al 1979). Large unilamellar vesicles were prepared by extrusion of multilamellarvesicles through 100 nm or 400 nm polycarbonate filters (Nucleopore, Pleasanton,CA) using an extrusion device (Lipex Biomembranes, Vancouver, Canada) (Hope,M.J. et al 1985). The liposomes were resuspended in isotonic VBS buffer (10 mMsodium barbital, 145 mM NaCl, pH 7.4) or in a 7.5 % sucrose buffer. The averagediameters of the liposomes extruded through the 100 nm filters were 105 + /- 19 nm(n=6) for PC:CH:CL (35:45:20 mol%), 106 +1- 16 nm (n=6) for PC:CH:PG(15:45:40 mol%), 98 ÷1- 7 nm (n=6) for PC:CH (55:45 mol%), 101 ÷1- 10 nm (n6)for PC:CH:Pl (15:45:40 mol%) and 110 +/- 10 nm (n=6) for CH:PG (45:55 mol%)as determined by quasielastic light scattering and NICOMP (model 270) analysisusing a vesicle Gaussian unimodal distribution mode. The average diameters of theliposomes extruded through the 400 nm filters were 244 +/- 22 nm (n=6) forPC:CH:CL (35:45:20 mol%), 284 +/-20 nm (n=5) for PC:CH (55:45 mol%) and244+/-13 nm (n6) for PC:CH:PG (15:45:40 mol%) as determined by quasielasticlight scattering and NICOMP (model 270) analysis using a Gaussian unimodaldistribution mode. The phospholipid content was measured using a colorimetricphosphate assay (Fiske, C. and Subbarow, Y., 1924).2.3.Analytic Techniques2.3.1. Functional complement assay (CH5O): The functional complementassay (CH 50) was slightly modified to measure residual complement activity inserum that had been exposed to liposomes. Liposomes were titrated serially in VBScontaining 0.15 mM CaCl2, 1 mM MgCl2, 5 % D-glucose and 0.1 % gelatin(DGVB2)and 100 ul aliquots of NHS or DDS at 12 mg/mI was added to eachdilution. The liposome-lysis control (color blank) consisted of liposomes incubated-37-in the absence of serum and was analyzed in parallel with the test samples at anequivalent liposome dilution; the absorbance of these samples at 414 nm wasalways < 0.004. The amount of complement activity remaining in each tube wascompared with that of serum incubated in the absence of liposomes (100 %). Thecontribution of serum alone to the absorbance values of the test samples was <2 %of the maximum lysis. Two hundred uI of each sample were then incubated for 30mm at 37 °C, then diluted up to 400 ul using ice-cold DGVB2 and kept on ice. Thetotal residual complement content of liposome-treated serum was measured byhemolytic assay (CH5O) using established methods as described by Whaley, 1985.Briefly, sheep erythrocytes, were sensitized with rabbit 1gM anti-sheep RBC antibody(EA) and suspended at a concentration of 1 cells/mI in DGVB2 Fifty ul of EAwere incubated with 50 ul of liposome-treated serum for 30 mm at 37 °C. Sampleswere then diluted by adding 1 ml VBS containing 0.1 % gelatin and 20 mM EDTA(GVB-EDTA). Unlysed EA cells were pelleted by centrifugation and the amount ofhemoglobin released into the supernatant was quantitated spectrophometrically at414 nm. The percent lysis of EA targets was calculated for each liposome dilution asfollows:(mean test - color blank at OD 414nm/100%- color blank at OD 414nm) X 100%The percent complement consumption by the liposomes, that is, the reduction instarting complement activity, was then calculated as 100- % EA lysis.2.3.2. C3 Crossed Immunoelectrophoresis: In this assay, 1 % agarose in VBSwas poured onto 5 X 5 cm glass plates. Undiluted samples (10 ul) were placed intosample wells cut 2 cm from the bottom of the glass plate. The plates wereelectrophoresed at 35 mA for 9 h. The gel was cut 0.25 cm above the sample well-38-and an agarose layer containing polyclonal goat anti-human C3 antibody was addedat 1:200 in VBS. Electrophoresis was then carried out at right angles to the originalelectrophoresis at 30 mA for 6-7 h. The plate was then dried using absorbent paperand the C3-anti-C3 precipitate was visualized using Coomassie Blue R-250 and thendestained with a 30 % methanol solution. Negative controls consisted of NHS orDDS incubated with non-activating liposomes (PC:CH at 45:55 mol%); all liposomecompositions were used at a final concentration of 25 mM. The positive control, inwhich maximum conversion of C3 is achieved, consisted of serum incubated withinulin, a potent activator of complement. For each of the liposome preparationstested, immunoprecipitation patterns were compared with that of the negativecontrol. The treatment of serum with inulin results in the cleavage of C3 to C3b andiC3b. These smaller fragments migrate further from the origin than native C3 in thefirst dimension of electrophoresis and subsequently appear as a distinct peakseparated from native C3 in the precipitin arc. The areas under the shifted curvewere calculated using the Bioquant program (R and M Biometrics Inc., Nashville, TN)from eight successive readings.2.3.3. Measurement of alternative pathway activation: To assess activation ofthe alternative pathway, two different methods were employed. First, the alternativepathway hemolytic activity was assessed using an alternative pathway CH5O assay(APH5O). Briefly, serum exposed to liposomes [PC:CH:CL (35:45:20 mol%) at 12.5mM final concentration, PC:CH:PG (15:45:40 mol%) and PC:CH:Pl (15:45:40 mol%)both at 25 mM] was incubated with rabbit erythrocytes suspended inveronal-buffered saline containing 3.75 % D-glucose, 0.03 % gelatin, 7 mM MgCl2and 20 mM EGTA. In the presence of EGTA, rabbit erythrocytes activate the-39-alternative pathway, but not the classical pathway; the degree of hemolysis isproportional to the functional alternative pathway activity in the sample. Alternativepathway activation was also measured using an ELISA for the activation fragmentBb. This fragment is generated by the cleavage of factor B during alternativepathway activation. Microtitre plates coated with monoclonal antibody specific forthe Bb fragment capture Bb from the serum sample. After washing, the bound Bb isdetected using a peroxidase-conjugated goat-anti-Bb antiserum. The amount ofcolored reaction product was compared to that generated by standards of purifiedBb including low (0 mg/L), medium (0.103 mg/L) and high (0.267 mg/L) standards.Several normal serum pools were used to establish a normal range; the Bb valuesfor undiluted serum from healthy individuals were between 0-9 mg/L. Samplesconsisting of PC:CH:CL (35:45:20 mol%), PC:CH:PG (15:45:40 mol%) and PC:CH:PI(1 5:45:40 mol%) were exposed to NHS and DDS at a final concentration of 50 mMlipid and run in duplicate. A sample containing NHS was exposed to inulin in thepresence of 1 mM magnesium chloride; similar samples with DDS were prepared.The samples were incubated at 37 °C for 1 h. Positive controls were prepared byincubating NHS or DDS with inulin to activate the alternative pathway and generateBb fragments.2.3.4. Analysis of Cig binding to liposomes: This assay was based on theaffinity of Clq for aggregated IgG. Whole human lgG was aggregated by heating at62 0C for 30 mm. One hundred and thirty ul of aggregated human lgG at 20 ug/mlwas then adsorbed to a 96-well Immunolon II microtitre plate at 4 °C overnight.Unreacted protein binding sites were blocked with PBS containing 3 % BSA.Liposomes were titrated serially in VBS and reacted with an equal volume of NHS or-40-DDS at a protein concentration of 6 mg/mI at 37 0C for 30 mm. After this period themicrotitre plate was washed three times with PBS containing 0.5 % BSA and 0.5 %Tween 20. The liposome treated serum mixtures (130 UI) were then added to themicrotitre plate wells and incubated for 37 °C for 1 h. Maximum levels of Clq weredetected in wells containing either NHS or DDS but no liposomes (100 %): thenegative controls consisted of liposomes incubated in the absence of serum for eachdilution used. The residual binding of Clq to aggregated lgG was proportional to thelevel of free Clq in the serum. Clq binding was detected by incubating with aprimary goat anti-human Clq antisera and subsequently with peroxidase-labeledrabbit anti-goat IgG. Both antibodies were diluted to the appropriate titre in PBScontaining 3% w/v powdered skim milk. Bound peroxidase-conjugated antibodywas detected using 3,3’,5,5’ tetramethylbenzidine and color was measured at 620nm using a spectrophotometer. The percent Clq binding to lgG after incubationwith liposomes was calculated for each liposome dilution as follows:(mean test - color blank at OD 620 nm /100% - color blank at OD 620 nm) X 100%The percent Clq consumption; that is, the amount bound to the liposomes, wasthen calculated as 100- % Clq bound to lgG-coated plates.2.3.5. lmmunoblot analysis of proteins associated with liposomes: Liposomes*were trace labeled with OH , and incubated with undiluted NHS or DDS (1:4 vol:vol).Incubations proceeded at 3700 for 30 mm after which the mixtures were passedthrough a gel filtration column (BioGel Al 5m, 200-400 mesh) to separate liposomesfrom free serum components as described by Chonn et al. 1992. The radioactivityassociated with each liposome composition was counted in a beta scintillationcounter (Philips PW 4700 liquid scintillation counter, Fisher, Ottawa, Canada) and-41-standardized before running on a gel. Protein separation was performed bySDS-PAGE under reducing or non-reducing conditions, as described in the figurelegends. Prestained SDS-PAGE standards were used to estimate the molecularweight of the proteins separated. Proteins were transferred onto a nitrocellulosemembrane using a mini trans blot electrophoretic transfer cell (BioRad). Unreactedprotein binding sites were blocked overnight at room temperature using 5% w/vpowdered skim milk in PBS-Tween (0.13 M NaCI, 2.7 mM KCl, 1.47 mM KH2PO4,3.7 mM Na2HPO4and 1% Tween 20). The nitrocellulose paper was then incubatedwith either goat anti-human C3, anti-human factor B, anti-human C7, anti-human C9or non-immune goat serum all diluted to (1:500) in the same buffer for 1 h at roomtemperature. Anti-human lgG and peroxidase-conjugated rabbit anti-human 1gMwere diluted to 1:3000 while the goat anti-human Clq was diluted to 1:1000. Afterwashing 3 times for five mm in PBS containing 0.5 % Tween 20, nitrocellulose blotswere then incubated with a peroxidase-conjugated rabbit anti-goat antibody dilutedto 1:5000 for all blots except Clq where the antibody was diluted to 1:1000. Thedilution buffer contained PBS with 5 % w/v powdered skim milk and 1 % Tween 20.Bound antibody was visualized using the Amersham ECL Western blot detectionsystem.Liposomes that were incubated with plasma at 37 °C for 60 minutes werepassed through a Sepharose CL2B column. Certain experiments used exogenouspurified protein. In these experiments the plasma was allowed to equilibrate with theexogenous protein at 37 °C for 15 mm before the liposomes were added. Theradioactivity of each fraction was counted as described above and the proteinconcentration of each fraction was determined using a BCA protein determinationassay. The protein recovered from the liposomes were assayed as described by-42-Wessel and Flugge (1984). The liposome free plasma protein fractions were pooledand concentrated.Protein separation was performed by SDS-PAGE under reducing ornon-reducing conditions. Prestained SDS-PAGE standards were used to estimatethe molecular weight of the proteins separated. The different liposome fractionswere standardized such that all liposomes were at the same total phospholipidconcentration. The plasma samples were equated at 4 mg/mI unless otherwiseindicated within this text before separating on the gel. Proteins were transferred ontoa nitrocellulose membrane (BioRad, Hercules, CA) using a mini trans blotelectrophoretic transfer cell (BioRad) and unreacted protein binding sites wereblocked as described above. The blot that was to be probed with anti-human MBPwas blocked as described above with 10 mM CaCI2. The nitrocellulose paper wasthen incubated with a number of antibodies: rabbit anti-human B2GP1 was diluted to(1 :2500) or (1 :5000) for the liposomes and free plasma proteins respectively;peroxidase-conjugated goat-antihuman Protein S was diluted to (1:1000); sheepanti-human C4bp was diluted to (1:1000); and, mouse anti-human MBP was dilutedto 0.5 ug/ml containing 10 mM CaCl2. After the antibody incubations andappropriate washings, as described previously, the B2GP1 blots were incubated withperoxidase conjugated mouse anti-rabbit antibody diluted to (1 :5000) or (1:10000)for the liposomes and free plasma proteins respectively, peroxidase conjugatedrabbit anti-sheep was diluted to (1:2000) and peroxidase conjugated goatanti-mouse was diluted to (1:1000) or (1:3000) as described in the text. Boundantibody was visualized as described earlier. Exposure times were varied accordingto the protein probed.-43-2.3.6. Competitive Binding Study between Cig and Mannose Binding Proteinfor anionic liposomes : Liposomes were trace labeled with 14C-dipalmitoyl-phosphatidytcholine and were extruded through 400 nm filters and resuspended inbuffer containing 11 mM Tris, 85 mM sodium chloride and 220 mM sucrose (7.5 %),pH 7.4. The liposomes were incubated with a mixture of 125l-labeled Clq andunlabeled Clq at a concentration of 0.5 mg/mI and various concentrations of MBP(80 ug/mI, 8 ug/mI, 0.8 ug/mI and 0.08 ug/mI) under isotonic conditions at roomtemperature for 20 mins. To separate the free (unbound protein) from theliposome-bound protein, 40 ul of the reaction mixtures were layered on an isotonicintermediate density separating buffer (containing 5 mM Tris, 155 mM sodiumchloride and 115 mM sucrose (3.9 %), pH 7.4. The polyallomer centrifuge tubeswere spun at 166,300 X g for 30 minutes in a Beckman TL-100 table-topultracentrifuge using a TLS-55 swinging bucket rotor. This resulted in the pelleting of95 + /- 2 % of the liposomes. Tubes were then frozen and sliced into two pieces: thepellet containing liposome which bound protein and the supernatant which provideda measure of the equilibrium protein concentration. Tube slices were counted in agamma counter. Control tubes were run in which no liposomes were present in thereaction. The amount of Cl q spun down from the control tube was subtracted fromthe liposome-treated Clq tubes. The amount of Clq bound to the liposomes wascalculated as follows:ug of Clq bound= (cPmpellet - cpmcontroi)/specific activity of Clq mixture.Factor B was used, at the same concentrations as MBP, as a control for nonspecificprotein effect.-44-2.3.7. Detection of serum mannose binding protein binding tobeta-2-glycoprotein-l using an enzyme linked immunosorbent assay: This assayused NHS as the source for MBP. High binding microtitre plates were aliquoted with50u1 of B2GP1 (0.1 mg/mI), BSA (0.1 mg/mI) or B2GP1-depleted sera (0.1 mg/mI)at 4 0C overnight. Unreacted protein binding sites were blocked with PBScontaining 3% BSA. Fifty ul of NHS was added (source of MBP) to each well andincubated at 37 °C for 1 hr. Maximum and minimum (negative control) levels of MBPconsisted of wells coated with B2GP1 and BSA respectively. MBP binding wasdetected by incubating the plates with 50 ul of a 2 ug/mI stock solution of rabbitanti-human MBP at 25 0C for 2 h and subsequently with a peroxidase-labeled goatanti-rabbit lgG. Both antibodies were diluted to the appropriate titre in PBScontaining 3 % w/v powdered skim milk. Bound peroxidase-conjugated antibodywas detected using 3,3’,5,5’ tetramethylbenzidine and color was measured at 620nm using a spectrophotometer. Binding was reported as Absorbance at 620 nm.2.3.8. Detection of beta-2-glycoprotein-1 using an enzyme linkedimmunosorbent assay: B2GP1 concentrations in plasma samples were determinedby a competitive ELISA. Briefly, 200 ul of a 5 ug/mI solution of B2GP1 in PBS wascoated in individual wells at 50 °C (lug/well). The plate was washed three timeswith distilled water and blocked with PBS containing 3 % BSA for 2 h at roomtemperature. The wells were washed three times with PBS before adding samples.During the blocking procedure, samples and standard curves were prepared asfollows. First, a standard curve for B2GP1 was constructed over the range 0-10000ng/ml (CD. 1.156 to 0.251 respectively). Standards were diluted in PBS containing 1% NP4O in glass 13 X 100 mm test tubes and were incubated at room temperature-45-for 2 h with an equal volume of a 1:4000 dilution of rabbit anti-human B2GP1, dilutedin PBS containing 1 % NP4O. Similarly, samples were serially diluted in PBScontaining 1 % NP4O and incubated at room temperature for 2 h with an equalvolume of a 1:4000 dilution of rabbit anti-human B2GP1. At the conclusion of the 2 hincubations, 200 ul of sample was loaded into the blocked ELISA plates, in triplicate,and allowed to incubate at room temperature for 1 h. The wells were then washedthree times with PBS containing 0.05 % Tween 20 to reduce non specific binding ofprimary antibody. Following this, a peroxidase-conjugated mouse anti-rabbit (1:5000in PBS containing 0.1 % BSA), was added (200 uI/well) to the plate and allowed toincubate for 1 h at room temperature. The plate was again washed with PBScontaining 0.05 % Tween 20, followed by several washes with PBS, to remove alltraces of Tween. Peroxidase conjugated antibody was detected using2,2-azino-di-3-ethylbenzthiazoline-6- sulfonic acid (1 mg of ABTS per ml, 0.005 %H20 in Mcllvaine buffer pH 4.6). Color development was monitored at anabsorbance of 414 nm.2.3.9. Immunological free protein S quantitation: A Liatest protein S kit wasused to quantitate the amount of free protein S in plasma available after exposure of*anionic liposomes containing CH to NHP (1:3 vol:vol) or in B2GP1-depletedplasma. Briefly, to remove any contaminating lipid and bound protein S, 170 ulsamples were incubated with 30 ul PEG for 30 mins at 4 0C and the samples werethen centrifuged at 11,641 X g for 2-4 mins (Abbot Diagnostics, Mississauga, Ont.).An aliquot of the supernatants of the plasma exposed to liposomes were counted ina beta scintillation counter to detect any contaminating lipid. All samples used forthe study contained <10 % lipid contamination as specified by the manufacturer’s-46-instructions. Fifty ul of the plasma samples were then incubated with 500 ul of latexbeads coated with anti-human protein S as specified by the kit directions. Anyavailable protein S agglutinated the latex beads and the absorbance was measuredat 590 nm using a Cobas Fara (Roche Diagnostica, Mississauga, Ont). There is adirect relationship between the observed absorbance value and the concentration ofthe antigen being measured.Free protein S was quantitated in the B2GP1 depleted plasma samples andthe appropriate controls: plasma exposed to rabbit IgG Sepharose and the NHPcontrol. Briefly, 300 ul plasma samples were concentrated in 30 Kd cutoffmicroconcentrators by spinning at 5000 X g (IEC floor centrifuge, Model M25,Boston, MA) for 7 mins to a concentration of 50 mg/mI, 50 ul of the plasma was thenused as previously described for the assay. The remaining reagents of the assaywere appropriately diluted to reflect the dilution in the starting sample.2.3.10. Functional Protein S Assay: A Staclot(R) protein S kit was used toascertain the functional capability of the remaining free protein S in the plasmaexposed to trace labeled liposomes (1 vol:3 vol) or in B2GP1-depleted plasma.Briefly, for the plasma exposed to the liposomes, the samples were incubated withPEG to remove any liposome-bound or C4BP-complexed protein S and sampleswere centrifuged as described above. This treatment also removes anycontaminating lipid. The supernatants of the plasma exposed to liposomescontained <10 % lipid contamination as specified by the manufacturer’s instructions.An aliquot of the plasma sample was counted in a beta scintillation counter todetermine the amount of liposome contamination in the plasma after incubation withPEG. Fifty ul of the plasma samples exposed to the liposomes were then diluted 1:9-47-(vol:vol) with the Owen Koller buffer containing 0.03 M sodium barbitone and 0.1Msodium acetate, pH7.4. The mixture was then incubated at 37 °C for 2 mm with 50 ulof the following reagents: protein S-deficient human plasma, human activated proteinC and bovine factor Va. The time to form a clot was measured after the addition of50 ul of 0.025 M CaCI2. The clotting time was measured on a ST4 BlO machine(Stago, France) and results obtained were reported as a percentage of activitydetected in normal human plasma which was generated from a standard curvewhich comprised of 100% (1 vol NHP:9 vol Buffer), 75 %, 50 % and 25 %.The B2GP1-depleted plasma samples and the appropriate controls werequantitated for functional Protein S. The freshly thawed plasma samples wereexposed to anti-B2GP1-Sepharose and rabbit IgG-Sepharose as described above.The samples were spun at 1000 X g for 1 mm to remove any Sepharose beads. Fiftyul of the plasma samples were incubated with the appropriate reagents as describedabove. The assay requires the samples to be ideally at a concentration equivalent to10 mg/mI. As the assay is extremely sensitive the protein content of the sample wasquantitated after the Staclot experiment was performed. The experiments in thissection used plasma that was at a protein concentration of 13 mg/mI.2.3.11. Staining Procedures: Gels were either silver stained using anoptimized silver stain procedure (Heukeshoven, J. and Dernick, R. 1988) or stainedwith Coomassie R250. Briefly, the silver stain required the gel to be left overnight atroom temperature in a solution containing 10 % acetic acid and 30 % ethanol. Thiswas followed by 3x20 mm washes with 30 % ethanol and 2x5 mm washes withdistilled water. The gel was then mixed with 250 mg/mI Na2SO4solution for 1 mmfollowed by 3x5min washes with distilled water. A silver stain solution (0.6 g AgNO3,-48-22.7 ul HCHO in 300m1 water) was mixed with the gel for 30 mm. This was followedby 2 x 5 mm washes in water. A developing solution (24 g K2C03,272.4 ul HCHQ,0.00190 g Na2SO3in 600ml distilled water) was applied until full development,followed by fixation with lOml of acetic acid. For the Coomassie stain the gel wasmixed with the Coomassie R250 at 25 0C for 1 h and then destained using a solutioncontaining 20 % methanol and 10 % acetic acid.2.4. Statistical AnalysisThe 3-way ANOVA used in the study was performed using the statisticalpackage SYSTAT, Evanston, IL. Multiple comparison tests, for the data in tables 5-8,were performed using GraphPAD lnstat Software version 1 .lOa. The calculations ofmeans and standard deviation were carried out with a Sharp EL-506A scientificcalculator.-49-3. RESULTS3.1. Characterization of normal human serum and immunoglobulin depleted serum:Normal human serum (NHS) and immunoglobulin depleted serum (DDS) weresubjected to SDS-PAGE using a 4-15 % gradient gel under reducing andnonreducing conditions. Specific depletion of lgG was confirmed as all bandsdetected in NHS were also detected in DDS with the exception of the bandcorresponding to lgG (Figure 4). lgG migrates at 150 kD under nonreducing and 50(heavy chain) and 25 kD (light chain) under reducing conditions. A sample of purelgG was run on the same gel to confirm that the depleted protein corresponded tothat of lgG (Figure 4).The depletion of lgG was also confirmed by Western blot analysis undernon-reducing conditions; a 150 kD band reactive with anti-human lgG in NHS was nolonger present in DDS (Figure 5). Similarly, the depletion of 1gM was confirmed byWestern blot; the 970 kD band reactive with peroxidase-conjugated anti-human 1gMin NHS was no longer present in DDS (Figure 6).To further confirm that depletion was successful, the levels of lgG and 1gMwere quantitated using a QM 300 Nephelometer. Table 1A shows that with proteinconcentrations of 12 mg/mI, the levels of lgG and 1gM in NHS were detectable andthe undiluted samples were well within the normal range, while in DDS, the levels ofboth lgG and 1gM were below the lower limit of detection of the assay.-50-Figure 4. Protein profile of normal human serum and immunoglobulindepleted serum. NHS and DDS were prepared as described in Materials andMethods. The protein profiles were analyzed on a premade 4-15% gradient phastgel under nonreducing (Figure 4a) and reducing (Figure 4b) conditions. Two uI/laneof a 0.5 mg/mI protein solution followed by silver staining. Lane 1 contains bovineserum albumin, lane 2 human lgG, lane 3 normal human serum, lane 4 DDS, lane 5running buffer and lane 6 molecular weight marker 1:32.B1 2 3 4 5 6• 1 2 3 4 5 6. 200 Kd.4116 Kd97.4 lCd466 Kd45 Kd‘431 Kd-51-Figure 5. Immunoblot analysis of the depletion of lgG from immunoglobulindepleted serum. NHS and DDS were prepared as described in Materials andMethods. Two UI of each of the samples were subjected to SDS-PAGE using a4-15% gradient phast gel (Pharmacia) under nonreducing conditions, transferredonto nitrocellulose paper and probed with goat anti-human lgG. Lanes 1 and 3contain DDS at 6.62 mg/mI and 3.31 mg/mI respectively. Lanes 5 and 7 containNHS at 6.62 and 3.31 mg/mI respectively. Lanes 2, 4, 6 and 8 contain runningbuffer. No bands were detected when the blot was incubated with normal goatserum (data not shown)—kd12 34 5 57 S20066445-52-Figure 6. Immunoblot analysis of the depletion of 1gM from immunoglobulindepleted serum. NHS and DDS were prepared as described in Materials andMethods. Fifteen ul of each of the samples were subjected to SDS-PAGE using a4-20 % gradient minigel under nonreducing conditions, transferred ontonitrocellulose paper and probed with peroxidase rabbit anti-human 1gM. Lane 1contains DDS at 6.62 mg/mI and lane 2 contains NHS at 6.62 mg/mI. No bandswere detected when the blot was incubated with normal rabbit serum (data notshown).kd42004116497661-53-Characterization of immunoglobulin depleted serumTABLE 1A. Quantitation of IgG and 1gM levels using a OM 300 Nephlometer.NHS (g/L) DDS (çi/L)1gM 1gMUndiluted 11.3 1.15 NA NA12 mg/mI 3.3 0.323 <0.069 <0.082*= Lower limit of detection of this assay.NA= Not ApplicableCoefficient of variance for IgG measurement= 6%Coefficient of variance for 1gM measurement= 2%TABLE 1 B. CH5O and APH5O values.NHS (U/mI) DDS (U/mI)CH5O APH5O CH5O APH5ONormal range 114-202 99-173 NA NA50mg/mi 150 162 200 125NA= Not ApplicableCoefficient of variance for CH5O measurement = 8%Coefficient of variance for APH5O measurement= 9%-54-Using the Clq ELISA described in Materials and Methods section, it wasfound that the Clq level was markedly reduced in DDS compared to NHS(0D620nm = 0.02 and 0.27, respectively). Immunoblot analysis of Clq content inNHS and DDS indicated complete depletion of Clq from DDS (Figure 7, lane 3,5)during removal of immunoglobulins from NHS (Figure 7, Lanes 6-9). These datasuggest that upon depletion of immunoglobulins, the Clq is also depleted due to itsaffinity for bound immunoglobulin. Therefore, DDS was repleted with Clq to normalserum concentrations of 0.063 mg/mI. Manufacturers’s analysis of Cl q indicatesthat the purified material is greater than 95 % pure and is fully functional. The finalpool of highly purified Clq was tested for the presence of contaminatingimmunoglobulin proteins by double immunodiffusion, immunoelectrophoresis andELISA procedures and was found to be free of immunoglobulins.Lastly, to confirm that the addition of cations to DDS did not alter its hemolyticactivity in either the alternative or classical complement pathway, functional assayswere performed. Table 1 B illustrates that the classical and alternative complementactivity for both NHS and DDS are within the normal clinical range for complementfunction. Thus, analysis of the reagents used in the studies confirms the absence ofimmunoglobulin lgG and 1gM from DDS as well as the adequate replacement of Clqand required divalent cations.Although serum was passed through immunoaffinity columns in the presenceof normal serum calcium levels, only Clq was removed. Levels of Cir and Cis wereassumed to be near normal as repletion of only Clq resulted in full hemolyticfunction (Table 1 B).-55-Figure 7. Immunoblot analysis of Clq content in normal humanserum and immunoglobulin depleted serum. NHS and DDS were prepared asdescribed in Materials and Methods. Fifteen UI of sera were subjected to SDS-PAGEusing a 10% minigel, transferred onto nitrocellulose paper and probed for thepresence of Cl q using goat anti-human Cl q antibodies. Lane 3 and 5 contain DDS(without the repleted amount of Clq) at 13.24 mg/mI. Lane 3 contains DDS inreducing buffer while lane 5 contains DDS in nonreducing buffer. Lane 6 and 7contain NHS in nonreducing buffer at 13.24 mg/mI and 3.31 mg/mI respectively.Lane 9 and 10 contain NHS in reducing buffer at 13.24 mg/mI and 3.31 mg/mIrespectively. Lanes 2,4 and 8 contain running buffer. No bands were detected whenthe blot was incubated with normal goat serum.2 3 4 5 6 7 8 9 10 kd— 200_;_ i166454 31-56-3.2. Functional Complement Assay (CH5O): As a first determination of thecomplement activating properties of the liposome compositions chosen, a functionalcomplement assay, the CH5O, was performed. The liposomes used in this studycontained the anionic phospholipids CL, P1 and PG in a background of PC and 45mol% CR Figure 8 demonstrates the effect of concentration of CL in the liposomecomposition on the activation of complement in NHS and DDS. Complementconsumption in NHS was observed with increasing mol% of CL and therefore,increasing charge (Figure 8A). The liposome composition containing 20 mol% CLwas diluted to 0.0001 to confirm that the complement consumption started at 0 %(data not shown). In DDS, complement consumption was only observed when CLwas present in at least 12.5 mol% CL (Figure 8B). Figure 9 demonstrates the effectof concentration of PG in the liposome on the activation of complement.Complement consumption was evident in PG 20 mol% and higher (Figure 9A);however, complement consumption in DDS was not evident until PG was present at40 mol% (Figure 9B).-57-Figure 8. Complement consumption in human serum by CL-containingliposomes. Large unilamellar vesicles (100 nm) containing varying amounts of CLwere incubated with NHS (panel A) or DDS (panel B). Their effect on complementactivity compared to the absence of liposomes was determined by complementhemolytic assays (see Methods). The phospholipid vesicles were composed ofPC:CH (55:45 mol%) ), PC:CH:CL (50:45:5 mol%) (o) PC:CH:CL (42.5:45:12.5mol%) ), or PC:CH:CL (35:45:20 mol%) (J. Each point represents the mean of 4experiments done in duplicate; bars represent 1 SEM.-58-.-UI1—WComplementConsumption(%)-L01-000100100 0 0 -hI-•U0 Co 00B(7’ CD-CDCD0o0oC) CD•h—-‘-Ip)o0IIComplementConsumption(%)cri—iooCcxi0p 0 0 p 0 -L pI—--II----——I-II——I--4F—-H41--II——I1---I—I1—I--II.-00•Figure 9. Complement consumption in human serum by PG-containingliposomes. Large unhlamellar vesicles (100 nm) containing varying amounts of PGwere incubated with NHS (panel A) or DDS (panel B). Their effect on complementactivity compared to the absence of liposomes was determined by complementhemolytic assays (see Methods). The phospholipid vesicles were composed ofPC:CH (55:45 mol%) (4, PC:CH:PG (35:45:20 mol%) (0), PC:CH:PG (15:45:40mol%) (7), or CH:PG (45:55 mol%) (j). Each point represents the mean of 3experiments each done in duplicate; bars represent 1 SEM.-60-Dflw (DC—(DComplementConsumption(%)ComplementConsumption(%)01000010p 0 0-0(j)0-03 CDCD-‘0-C,C)CDCD-&-I.003.-_-0_-000 0 0 -L1’)(31—40(710(310-L 0 0-L 0 0The experiment was performed with a third anionic phospholipid, Pt, and againcomplement consumption was evident in DDS when the anionic phospholipid waspresent at 40 mol% (Figure 10). With all three anionic phospholipids, more lipid wasrequired in DDS to observe the same level of complement consumption seen inNHS. The neutral-charged composition PC:CH (55:45 mol%) did not consumecomplement and was used as a negative control. A 3-way ANOVA comparinganionic phospholipid mol%, lipid concentration and immunoglobulin presence orabsence (SYSTAT, Evanston, IL) was used. Each of the following liposomecompositions PC:CH:CL (35:45:20 mol%), PC:CH:PG (15:45:40 mol%) andPC:CH:Pl (1 5:45:40 mol%) showed a statistically significant lipid concentration effect(p < 0.01), a significant increase in complement activation compared to net neutralcompositions (p < 0.001) and a significant difference in the degree of activationwhen exposed to NHS versus DDS (p <0.01).3.3. C3 crossed Immunoelectrophoresis: In order to verify that the CH5O assay wasmeasuring complement activation-induced loss of activity rather than passiveadsorption of complement proteins onto the liposomes, complement activation wasalso monitored using crossed immunoelectrophoresis to measure C3 degradationproducts. Inulin was used to generate the maximum conversion of C3 todegradation products C3b and iC3b as a positive control for the crossedimmunoelectrophoresis. Results showed that liposomes exposed toimmunoglobulin-depleted serum produced C3 degradation fragments indicating thatanionic phospholipids are able to activate complement in the absence ofimmunoglobulin (Table 2). Confirming the results of CH5O analyses, lessdegradation was measured in DDS than in NHS for a given amount of lipid. Crossed-62-immunoelectrophoresis does not distinguish between C3b and iC3b; therefore, therelative proportions of C3 degradation products cannot be assessed by this method.-63-C04-0.E:3U)C0C.)4-CEE0C)Figure 10. Complement consumption in human serum by P1-containingliposomes. Large unhlamellar vesicles (100 nm) containing 40 mol% of P1 wereincubated with NHS (open symbols) or DDS (closed symbols). Their effect oncomplement activity compared to the absence of liposomes was determined bycomplement hemolytic assays (see Methods). The phospholipid vesicles werecomposed of PC:CH:PI (15:45:40) (triangles) or PC:CH (55:45 mol%) (circles).Complement consumption by PC:CH vesicles in DDS was indistinguishable fromNHS. Each point represents the mean of 3 experiments each done in duplicate; barsrepresent 1 SEM.10075502500.001 0.01 0.1 1 10 100Liposome Concentration (mM)-64-TABLE 2. Activation of complement detected by curve shift in crossedimmunoelectrophoresis. Area measured represents the complement fragments C3band iC3b generated from a precipitin curve.Area under the curve (mm2) +1-1 S.D of 8 values.Treatment NEIS DDSInulin 3.075 ± 0.077 3.279 ± 0.086CH: PG (45:55) 3.052 ± 0.137 1.280 ± 0.091PC:CH:CL (35:45:20) 2.317 ± 0.086 0.651 ± 0.124Untreated 0.0 0.0-65-3.4. Measurement of Alternative Pathway Activation: When alternative pathwayactivation by liposomes was assessed, no significant complement consumption wasobserved with any of the anionic liposome compositions tested. However, thepositive control used sterylamine at 20 mol% which showed the expectedcomplement consumption via the alternative pathway (Figure 11). The SAcontaining liposomes were diluted to 0.0005 mM to confirm that complementconsumption began at 0 mol% (data not shown).To further confirm that thealternative pathway was not activated, factor Bb fragment levels were determined.Bb is an activation peptide generated during alternative pathway activation. Therewas no significant increase of Bb, (>9 mg/L), in any of the liposome compositionswhen exposed either to NHS or DDS at a final lipid concentration of 50 mM (Table 3).These data are also supported by the absence of native factor B or factor Bbfragments on the surface of these liposomes using Western blotting techniques(data not shown).3.5. Analysis of Cig Binding to Liposomes: To further confirm that the activation inNHS and DDS proceeded via the classical pathway, the Clq ELISA was used tomonitor the consumption of Clq from NHS or DDS exposed to liposomes. Clq wasconsumed in a concentration- dependent manner following the incubation ofliposomes containing CL 20 mol%, PG 55 mol% and P1 at 40 mol% (Figure 12) withNHS or DDS. Consumption did not occur in DDS that had not been repleted withClq confirming the removal of Clq during immunoaffinity chromatography nor did itoccur with the neutral composition PC:CH (55:45 mol%). In order to control forliposome adherence to the microtitre wells, liposomes containing P1 at 40 mol% weretrace labeled with CH* prior to incubation with serum in the plate as described in themethods. None of the wells exceeded background counts of 38 cpm + /- 2%,-66-indicating no adsorption of liposomes to the microtitre plates. Negative controlswhich included plates coated with BSA alone showed no Clq consumption.-67-Figure 11. Alternative pathway complement consumption (APH5O)in human serum by PG, P1, CL and SA containing liposomes. Large unilamellarvesicles (100 nm) containing PC:CH:PG (15:45:40 mol%) (, PC:CH:CL (35:45:20mol%) (0), PC:CH:Pl (15:45:40 mol%) (+3 and PC:CH:SA (35:45:20 mol%) () wereincubated with NHS. Their effect on complement consumption compared to theabsence of liposomes was determined by alternative pathway hemolytic assay.Each point represents the mean of 2 experiments done in duplicate.__100C0.. 750E(I,Co 50C-)C0.001 0.01 0.1Liposome Concentration (mM)100-68-TABLE 3. Levels of the alternative pathway activation fragment BbIn normal human serum or immunoglobulin depleted serum exposedto liposomes.Bb levels (mg/L)(Mean of 4 values ± 1S.D.)NHS DDSHigh serum control 13.80 ± 6.86Low serum control 0.0PC:CH:CL (35:45:20 mol%) 2.47 ± 2.47 5.20 ± 2.54PC:CH:PG (15:45:40 mol%) 5.75 ± 1.25 4.35 ± 2.78PC:CH:PI (15:45:40 mol%) 5.81 ± 0.89 6.00 ± 0.5Inulin >13.8 >13.8* Highest level accurately read from standard curve.Bb levels from healthy individuals = 0-9 mg/L-69-Figure 12. Effect of immunoglobulin on Clq binding to liposomescontaining anionic phospholipids. Liposomes composed of PC:CH:CL (35:45:20mol%) (panel A), CH:PG (45:55 mol%) (panel B) or PC:CH:Pl (15:45:40 mol%)(panel C), were incubated with NHS (A), with DDS (0) or with DDS lacking Clq (I).Binding of Clq to PC:CH (55:45 mol%) liposomes incubated with NHS is shown ineach graph (U). An ELISA was used to measure the residual Clq binding toaggregated lgG (see Methods). Each point represents the mean of 4 experimentseach done in triplicate; bars represent 1 SEM.-70-C750 4- 0250A100BCLiposomeConcentration(mM)3.6. Immunoblot Analysis of C3 Fragments on Liposomes: Figure 13 shows thedetection of iC3b on the surface of negatively-charged liposomes exposed to NHS orDDS. Exposure of liposomes composed of PC:CH:CL 35:45:20 mol%, PC:CH:Pl at15:45:40 mol% and PC:CH:PG at 15:45:40 mol% to DDS followed by Western blotanalysis resulted in the detection by anti-C3 of three bands at 81 kD, 75 kD and 50kD. These molecular weights correspond to the beta chain (81 kD) and the twofragments of the alpha chain of iC3b (75 and 50 kD). Liposomes of the samecompositions exposed to NHS also had iC3b on their surface. Exposure ofliposomes composed of PC:CH (55:45 mol%) to NHS or DDS showed no bands at50, 75 or 81 kD. These calculated molecular weights are within 10 % of the reportedmolecular weights for these C3 fragments. NHS was used as a positive control byexposing to inulin to achieve maximum conversion from C3 to iC3b and producedthe same bands that migrated at 50, 75 and 81 kD. Blots probed with normal goatserum as the primary antibody failed to yield any bands. Densitometric analysis (HP3392A recording integrator with BIORAD Model 620 video densitometer) of theWestern blots suggests that iC3b deposition occurs to a greater degree withcompositions containing PG or CL than in those containing P1 as the source ofanionic phospholipid (Figure 13). Thus, complement was activated at least to the C3step by the anionic liposomes and the opsonizing complement component iC3b wasdeposited on the liposome surface. Western blots of liposomes probed withpolyclonal antibodies to C7 failed to detect this complement protein on the surface ofthe liposomes. This is likely due to the limit of detection of these assays as it hasbeen previously demonstrated that C9 has been associated with CL-containingliposomes in these studies (Figure 14) and PG- containing liposomes in previouswork (Chonn, A. et al, 1991).-72-The work described in sections 3.1-3.6 was published in 1994 as Marjan, J.Xie, Z. and Devine, D.V. “Liposome-induced activation of the classical pathway doesnot require immunoglobulin”, Biochim. Biophys.Acta 1192:35-44, 1994.-73-Figure 13. Immunoblot analysis of iC3b bound to negatively chargedliposomes exposed to NHS or DDS. Liposomes (100 nm) at 100 mM were exposedto serum and separated from free serum components by chromatography.Liposome samples were made equivalent for lipid content at 3.6 mM and were thenrun on a 4-20% gradient SDS-PAGE minigel, transferred onto nitrocellulose paperand probed with goat anti-human C3 antibodies. Lanes 1, 3, and 5 containliposomes exposed to NHS while lanes 2, 4, and 6 contain liposomes exposed toDDS. Liposomes were composed of PC:CH:CL (35:45:20 mol%) (lanes 1 and 2),PC:CH:Pl (15:45:40 mol%) (lanes 3 and 4), or PC:CH:PG (15:45:40 mol%) (lanes 5and 6). Bands at 50, 75 and 81 kD corresponding to the two halves of the alphachain and the beta chain of iC3b were detected. Densitometer readings of theWestern blot are given in the table as absorbance units. The background levelswere recorded as 0.29. No bands were detected when the blot was incubated withnormal goat serum.-74-Figure 14. Immunoblot analysis of C9 bound to negatively chargedliposomes exposed to NHS or DDS. Liposomes at 100 mM were exposed to serumand separated from free serum components by chromatography. Liposomesamples were made equivalent for lipid content at 3.6 mM and were then subjectedto SDS-PAGE using a 4-20 % gradient gel under non-reducing conditions,transferred onto nitrocellulose paper and probed with goat anti-human C9antibodies. Lanes 2 and 3 contain PC:CH:PG liposomes (15:45:40 mol%) exposedto NHS and DDS, respectively, lanes 4 and 5 contain PC:CH:CL liposomes (35:45:20mol%) exposed to NHS and DDS respectively, lanes 6 and 7 contain PC:CH:Plliposomes (15:45:40 mol%) exposed to NHS and DDS respectively and lane 8contains NHS at 3.31 mg/mI. No bands were detected when the blot was incubatedwith normal goat serum.kd20097664531>J:-..-75-3.7. Identification of proteins that copurify with beta-2-glycoprotein-1: The next set ofexperiments tested the hypothesis that MBP may bind to B2GP1 which may lead tothe activation of the classical complement pathway in the absence ofimmunoglobulin. As discussed in the introduction, B2GP1 is a glycoprotein rich inspecific carbohydrate residues namely N-acetylglucosamine and mannose with ahigh affinity for anionic phospholipid. MBP, on the other hand, is an activator of theclassical pathway in the absence of immunoglobulin and Clq and binds to targetsthat are rich in N-acetylglucosamine, N-acetylmannosamine and mannose. In orderto demonstrate that MBP may bind to B2GP1, the method of immunoaffinitychromatography was used. However, it appeared that a number of proteinscopurified with B2GP1. The proteins bound to B2GP1 were specific as none of theproteins were found to bind to rabbit lgG. This experiment used the anti-humanB2GP1-Sepharose affinity chromatography in order to deplete plasma of B2GP1and identify the proteins that elute using Western blot analysis and specificantibodies. Figure 15A shows a silver stain of the column eluate (Lane 1) and pureMBP (Lane 2). Both protein profiles show a band that migrated under reducingconditions to 31 kD. The 31 kD band was immunoreactive with the monoclonalanti-human MBP (Figure 15B). It appears from the blot that the monoclonal antibodyrecognizes a number of bands, indicating that not all the protein is reduced in thepresence of beta-2-mercaptoethanol. This is further confirmed in the antibodybinding pattern in the Western blot of pure MBP (Figure 15B lane 1). It is known thatfor Clq, a molecule similar in structure and function to MBP, both 8.5 M urea and 20mM iodoacetamide are required in the reducing buffer to completely reduce theprotein to its three chains: A, B, and C (Kolb, W. et al. 1979). However, despiteaddition of the above chemicals, the antibody binding pattern in the Western blot ofpure MBP (Figure 15B lane 1) was not altered.-76-Figure 15A. Protein profile of eluate from anti-beta-2-glycoprotein-1column. Panel A. Lane 1 contains 40 ul of a 0.2 mg/mI eluate solution subjected toSDS-PAGE using a 10 % minigel under reducing conditions. Lane 2 contains 10 ul ofa 20 ug/mI pure MBP solution run under reducing conditions. The gel was silver*stained as described in the Materials and Methods (see page 45).Figure 15B. Immunoblot analysis of the MBP content in the eluate fromthe plasma exposed to anti-beta-2-glycoprotein-1 column for MBP.Panel B. Lane 1 contains 15 ul of 20 ug/mI pure MBP. Lanes 4 and 6 contain 60 ulof 0.2 mg/mI eluate solution containing 10 mM CaCl2. Lanes 2, 3 and 5 containrunning buffer. The proteins were subjected to SDS-PAGE using a 10 % minigelunder reducing conditions. The gel was blocked and probed with monoclonalanti-human MBP and the appropriate peroxidase conjugated antibody (1:1000) asdescribed in the materials and methods. During detection, the film was exposed tothe blot for 40 mins before development.*Eluate protein concentration from exposure to 2 ml anti-beta-2-glycoprotein-1Sepharose was 0.434 mg/mI as determined by the BCA assay. Eluate proteinconcentration from exposure to 2 ml rabbit lgG Sepharose was 0.08 mg/mI asdetermined by BCA assay.-77-9,VL.VVVV-I—1C,)N‘.1 C m Cli-‘ UIbAAAC,).IQ(O014The protein profile was analyzed for intensity of bands using a densitometer(HP 3392A recording integrator with BIORAD Model 620 video densitometer). MBPaccounted for 1 % of the total protein concentration which represents 0.006 mg/mIof the total protein (0.434 mg/mI) eiuted (Table 4). The reported concentration ofeach protein in the profile is a rough estimate for the theoretical concentration. Thereported concentrations were calculated using the following equation:(Abs. of specific protein/Total Abs. of all Proteins) X Conc.Eluate (mg/mi).The calculated concentration of MBP in the protein profile is almost forty-fold greaterthan that reported for physiological concentrations of 0.16 ug/mI. Normal fiuctuationin the levels of MBP from a series of blood samples from a panel of healthylaboratory workers were shown to range between 35 ng/ml and 3000 ng/mI (Thiel,S. et al 1992a). MBP levels have also been shown to vary between ethnic groups(Garred, P. et al 1992). The plasma used in this study was pooled from 27 donors tonormalize the MBP levels, yet there is an increase in the concentration of MBPlocalized on the immunoaffinity column. The enrichment of MBP in the columneluate indicates that MBP binds to B2GP1 when it is immobilized on the anti-humanB2GP1-Sepharose. When a similar aliquot of plasma was mixed with the normalrabbit lgG-Sepharose, the 31 kD band was not present as observed by silver stain orWestern blot analysis. This indicates that MBP is not isolated through direct bindingto Sepharose, or non-specific interactions with rabbit immunoglobulin. SAP andCRP were also found to co-elute with B2GP1 from the immunoaffinity column. Theseproteins were identified using polyclonal antiserum to human CRP or SAP in Westernblot.-79-TABLE4.Densitometerreadingsoftheelutionprofileafterexposuretoanti-beta-2-glycoprotein-1Sepharose.******CalculatedMolecularAbsmg/mi%TotalTentativePlasmaConc.Weight(kD)identificationasreportedintheliterature97.40.060.049MBPtrimer0-0.16ug/mI800.130.0818ProteinS20-25ug/ml700.170.123C4BP0.2-0.4mg/ml500.190.1125B2GP1/C4BP0.15-0.3rng/310.010.0061MBP0-0.16ug/mI250.080.0511SAP40ug/ml240.090.0512CRP1ug/mI*Totalabsorbanceofeluate=0.73**Totalconcentrationofeluate0.434mg/mlCalculatedvalues=Absorbanceofspecificprotein/TotalabsorbanceXConcentrationofEluate**%Total=Calculatedconcentrationofspecificprotein/ConcentrationofeluateX100%Figure 16A shows a silver stain of the proteins that copurify with B2GP1under reducing conditions and Western blots to identify the remaining proteins thatcopurify with B2GP1. The bands that were immunoreactive with the variousantibodies that were used to probe the nitrocellulose paper. The 50 kD band wasfound to be immunoreactive with anti-human B2GP1. The 80 kD band was found tobe immunoreactive with peroxidase-conjugated anti-human Protein S. Protein Smigrates under reducing conditions as a doublet which is partially detected in thefigure. Lastly the two bands which migrated at 70 and 50 kD were found to beimmunoreactive with anti-human C4BP. These represent the 7 alpha subunits of the70 kD fragment and the 1 beta subunit of the 50 kD fragment of C4BP. The bandthat migrated at 97.4 kD on the silver stained gel likely represented a nonreducedtrimer of MBP as it was immunoreactive with the monoclonal anti-human MBP(Figure 15B). The control column containing rabbit lgG conjugated to theSepharose 4B bound some nonspecific B2GP1 as identified by Western blotanalysis. However, no bands were immunoreactive with peroxidase conjugatedanti-human Protein S or anti-human C4bp (Figure 16B). Table 4 shows that there isalso an apparent increase (four fold) in the concentration of protein S localized onthe immunoaffinity column. This indicates that protein S is another protein that isenriched when B2GP1 is localized on the anti-human B2GP1 column. Thenitrocellulose blot was also probed with anti-human fibrinogen, anti-human C3 andanti-human albumin , as these three proteins (fibrinogen, C3 and albumin) are foundin large abundance in normal plasma, however, no signal was detected.-81 -Figure 16A. Protein profile and immunoblot analysis of the eluate fromplasma exposed to anti-beta-2-glycoprotein-1 Sepharose. Panel A shows 40 ul of a0.2 mg/mI eluate solution from the anti-B2GP1 column subjected to SDS-PAGEusing a 10 % minigel under reducing conditions and silver stained as previouslydescribed. The gel was transferred and probed with anti-human B2GP1, peroxidaseconjugated Protein S and anti-human C4bp as described in the Materials and*Methods (see page 39)Figure 16B. lmmunoblot analysis of the eluate from plasma exposed torabbit IgG Sepharose. Panel B shows 40 ul of 0.2 mg/mI eluate solution from thecontrol rabbit lgG column subjected to SDS-PAGE using a 12 % minigel underreducing conditions. The gel was transferred and probed with anti-human B2GP1,peroxidase conjugated protein S and anti-human C4BP as described in the Materials**and Methods . The left lane was exposed 10 mins before any bands were visible.No bands were visible in the nitrocellulose blot exposed to anti-protein S andanti-human C4BP when the film was exposed for 10 mins.*Eluate protein concentration from exposure to 2 ml anti-beta-2-glycoprotein-1Sepharose was 0.4 mg/mI as determined by BCA assay.**Eluate protein concentration from exposure to 2 ml rabbit lgG Sepharose0.08 mg/mI as determined by BOA assay.-82-FIGURE 16Kd20011697.4—* -66—* —45 —,31X B2GP1a ProtS cxC4bpAAg StainB97.4 —Kd20011666453121.514.5Ia B2GP1 a Protein S aC4bp-83-3.8. Binding of MBP on the surface of PC:CH:CL (35:45:20 mol%) liposomes: Theexperiments described in the previous section demonstrated that MBP can bindB2GP1 when B2GP1 was immobilized on the surface of antibodies; however, thequestion remained as to whether MBP can bind to the surface of anionic liposomeswhich immobilize B2GP1. The experiments in this section were designed to test,using Western blot analysis, the hypothesis that MBP binds to PC:CH:CL (35:45:20mol%) via B2GP1 and that the binding is through the carbohydrate residues. Inorder to investigate whether the binding of MBP may be localized on the surface ofthe liposomes by binding to B2GP1; the association of B2GP1 with liposomes wasfirst verified. As discussed in the Introduction, B2GP1 has a high affinity fornegatively charged surfaces as can be seen in Figure 17. Plasma exposed toliposomes was separated from the liposomes using the Sepharose CL-2B column(as discussed in Materials and Methods). B2GP1 bound only to the anionicliposomes composed of PC:CH:CL (35:45:20 mol%) and PC:CH:PG (15:45:40mol%) (Lane 2 and 3). The B2GP1 was depleted from the plasma samples exposedto these anionic liposomes (Lane 5 and 6). Binding of B2GP1 was not observed onthe surface of PC:CH (55:45 mol%) (Lane 1) and the intensity of the bandimmunoreactive with anti-human B2GP1 in the plasma exposed to PC:CH (Lane 4)was greater than that exposed to the anionic liposomes. These results demonstratethat the anionic liposomes provide a surface upon which B2GP1 is significantlyenriched. This has been previously alluded to in studies conducted by Sommerman,E. (1986) and Chonn, A. (1991). Studies by Chonn et al (1992) have also shown this50 Kd band to be associated with rapidly cleared liposomes.-84-Figure 17. Immunoblot analysis for beta-2-glycoprotein-1 on the*surface of liposomes. Liposomes were trace labeled with CH and were exposed toplasma and passed through the Sepharose CL2B (as described in the Materials andMethods). Liposomes containing 2.5 mM total lipid were loaded on a 10 %SDS-PAGE. Twenty ul of iiposome-free plasma samples were loaded per lane (4mg/mi). The proteins were subjected to SDS-PAGE using a 10 % minigei underreducing conditions, transferred to nitrocellulose and probed with anti-human B2GP1as previously described. Panel A identifies a 50 Kd band that was immunoreactivewith anti-human B2GP1. Lanes 1, 2 and 3 contain PC:CH (55:45 mol%), PC:CH:CL(35:45:20 moi%) and PC:CH:PG (15:45:40 mol%), liposomes respectively. Theimmunoblot shown in Panel B contains the liposome free plasma from the samesamples. Lanes 4, 5 and 6 contain the liposome-free plasma of PC:CH, PC:CH:CLand PC:CH:PG at the mole percentages indicated above. Western blots reportedthe depletion of B2GP1 from plasma exposed to anionic iiposomes and theconcomitant increase in B2GP1 deposited on the surface of the negatively chargediiposomes. -Kd20011697.431A B-85-In order to determine whether MBP will be localized on the anionic surface ofthe liposome, the nitrocellulose paper was probed with monoclonal anti-humanMBP. When NHP was incubated with PC:CH:CL (35:45:20 mol%) and PC:CH (55:45mol%) (as described in Materials and Methods), MBP was not detected on thesurface of the liposomes using Western blot analysis (data not shown). This is likelydue to the relative insensitivity of this method. However, when MBP concentrationwas increased in pooled NHP by the addition of exogenous pure MBP (to 40 ug/mI)and the plasma exposed to the above mentioned liposome compositions, bindingwas observed on the PC:CH:CL (35:45:20 mol%) liposomes (Figure 18A). Theconcentration of MBP used in these experiments (40 ug/mI) was above the normalphysiological concentration which can be as high as 300 % of normal levels asreported in one study looking at the effect of malaria on MBP levels (Thiel, S. et al1992b).Binding of MBP to its ligand can be inhibited by a number of carbohydrates ofwhich N-acetylmannosamine (ACM) and N-acetylglucosamine have been shown tobe the most effective. Kawasaki, et al (1983) has shown that certain carbohydratesof the D-configuration are more effective than others at preventing the binding ofMBP to its ligand. This study also showed that maximum inhibition (100 %) wasobserved when 1 ug of purified MBP was incubated in the presence of a variety of125-labeled carbohydrate ligands and 0.3 M unlabeled glycoprotein inhibitor.Inhibition remained at a 100 % for all the concentrations of inhibitor tested up to 0.5M. Kawasaki, et al (1983) determined, using Scatchard analysis, a different bindingratio for the different carbohydrates used. Since there have been no Scatchardanalysis to examine the binding of the carbohydrates to MBP in the presence ofserum/plasma proteins, it was thought appropriate to simply use the data that lug of-86-MBP was inhibited by 0.3 M of acetylmannosamine. Based on these data, thequestion arose as to whether the binding of MBP to the anionic liposome surfacecould be inhibited by adding the appropriate concentration of inhibitor, 2 M ACM ofD-configuration. Figure 18B shows the inhibition of binding of MBP to cardiolipincontaining liposomes using 2 M ACM; as no bands were immunoreactive with themonoclonal anti-human MBP. Bands that migrated at 31 kD and wereimmunoreactive with the monoclonal antibody were found in the liposome-freeplasma fractions of both PC:CH:CL (35:45:20 mol%) and PC:CH (55:45 mol%).These studies demonstrate that MBP binds to B2GP1 when B2GP1 isimmobilized on the surface of the anti-human B2GP1-Sepharose. Another approachwas used to demonstrate the binding of MBP to B2GP1. Here NHS was used as thesource of MBP. This study showed that when B2GP1 was depleted from serum, thelevel of MBP detected using an ELISA (see Materials and Methods) was alsoreduced by 22.2 % (Figure 19). This percentage may in fact be lower as MBP maybind to other unknown protein(s) in the B2GP1-depleted serum that immobilized onthe surface of the polystyrene plate. Despite the presence of a large number ofproteins in NHS that may bind to immobilized B2GP1 (as demonstrated above), theamount of MBP bound to B2GP1 was significantly different than that bound to BSA(i.e. specific binding of MBP to B2GP1 in comparison to non specific binding).Evidence for the binding of MBP to B2GP1 was obtained by the immunoaffinityexperiment and by ELISA. It is therefore possible that the binding on the surface ofthe liposomes may also be occuring via B2GP1.-87-Figure 18. Binding of mannose binding protein to PC:CH:CL (35:45:20 mol%)liposomes and the inhibition of binding.Panel A:Binding of Mannose Binding Protein to PC:CH:CL (35:45:20 mol%)liposomes. Two hundred ul of plasma was equilibrated with 8 ug of MBP, incubatedwith trace labeled liposomes and passed through the Sepharose CL2B as describedin the Materials and Methods. The peak liposome fractions were concentrated to 2.3mM total lipid. Forty ul of the liposome samples were loaded per lane and subjectedto SDS-PAGE using a 10 % minigel and run under reducing conditions. Lane 1contains 8 ul of pure MBP at 16 ug/mI, lanes 3 and 4 contains 40 ul of plasma thathad been exposed to PC:CH:CL (35:45:20 mol%) and PC:CH (55:45 mol%)respectively; lanes 7 and 8 contain 2.3 mM PC:CH:CL (35:45:20 mol%) and PC:CH(55:45 mol%) respectively. Lanes 2 and 5 contain running buffer alone. The gelswere transferred to nitrocellulose and blocked as described previously with blockingsolution containing 10 mM CaCl2, probed with monoclonal anti-human MBP andperoxidase conjugated anti-mouse IgG (1:1000). Bound antibody was detectedusing chemiluminescence and the film was exposed to the blot for 5 mm beforedevelopment.Panel B:Inhibition of binding of mannose binding protein to PC:CH:CL(35:45:20 mol%) liposomes. Two hundred ul of plasma was equilibrated with 8 ug ofMBP and 2M N-acetylmannosamine, incubated with trace labeled liposomes used asa marker to follow the liposomes, and passed through a Sepharose CL2B asdescribed in the Materials and Methods. The peak liposome fractions wereconcentrated to 2.3 mM and forty ul of the liposome samples were loaded per laneand subjected to SDS-PAGE using a 12 % minigel and run under reducing-88-conditions. Lane 7 contains 8 ul of pure MBP at 16 ug/mI, lanes 1 and 2 contains 40ul of plasma that had been exposed to PC:CH:CL (35:45:20 mol%) and PC:CH(55:45 mol%), respectively; lanes 4 and 5 contain 2.3 mM PC:CH:CL (35:45:20mol%) and PC:CH (55:45 mol%) respectively. Lanes 3 and 6 contain running buffer.The gels were transferred onto nitrocellulose and blocked as described previouslywith blocking solution containing 10 mM CaCI2 and probed with monoclonalanti-human MBP and peroxidase conjugated anti-mouse (1:3000). The film wasexposed to the blot for 20 mm before development.-89-(3o)D-OO)%J‘‘V•1-n 0 C m-I UiI9-IICQ)%Jv—vvvvwcbM C,)IIC;’0 F’) () . (71 0:10) •b43.9. Measurement of complement hemolytic activity in the presence of mannosebinding protein inhibitors: As discussed in the introduction, MBP, in apentamer/hexamer configuration (Lu, J. et al 1990), can directly activate theclassical pathway of complement by its interaction with Cl. In order to determine therole of MBP in complement activation by liposomes, CH5O assays were used asdescribed in the Materials and Methods section. Upon exposure to NHS,compositions containing insufficient anionic phospholipid to activate complement,such as PC:CH:CL (45:45:10 mol%), were able to consume complement in thepresence of an inhibitor of MBP such as ACM (Figure 20A). This phenomenon wasalso observed with a weakly activating composition such as PC:CH:PG (35:45:20mol%) (Figure 20B) where complement consumption was enhanced in the presenceof ACM. Furthermore, in the absence of activating immunoglobulins, the liposomecompositions which include PC:CH:PG (45:45:10 mol%) and CH:PG (45:55 mol%)consumed more complement in the presence of ACM (Figure 21A and B,respectively) than in its absence. In parallel liposome-free controls where ACM andmannose were serially diluted in DGVB2 and incubated with NHS followed by thesensitized sheep red cells, complement was not activated, indicating that theobserved consumption of complement was due to the presence of the liposomesand that the presence of a high molarity of inhibitor does not affect complementconsumption. These results confirm the observations of Kawasaki et al (1983) thatmannose is a weak inhibitor of MBP and that ACM is a strong inhibitor of MBP.While the addition of mannose did not alter complement consumption, the additionof ACM resulted in increased complement consumption curves. These results alsodemonstrate that at physiological concentrations, MBP does not play a role incomplement activation by liposomes and the primary route of activation is through-91-Clq either in the presence (Figure 20A and B) or the absence (Figure 21A and B) ofactivating immunoglobulins. It is evident that complement consumption increaseswhen MBP is removed from the system. This suggests that some competition forbinding to the anionic surface occurs when both Clq and MBP are present underphysiological concentrations.-92-Figure 19. Detection of serum mannose binding protein binding tobeta-2-glycoprotein-1 using an enzyme linked immunosorbent assay. NHS was usedas a source of MBP and allowed to bind to wells containing pure B2GP1, BSA orB2GP1 depleted serum. An ELISA was used to measure the amount of binding ofMBP in the NHS to the different coated wells. Each bar represents the mean of 2experiments +/- 1 S.D., each done in triplicate.0.14______0.13 I0.120.11 I0.100.09___________E0.0800. BSA B2GP1 DEPL...93..Figure 20A. Complement Consumption in human serum by PC:CH:CL(45:45:10 mol%) liposomes in the presence of mannose bindingprotein inhibitors. Large unilamellar vesicles (100 nm) composed of PC:CH:CL(45:45:10 mol%) were serially titrated and incubated with NHS (circles), or NHS inthe presence of mannose 0.5 M (triangles) or N-acetylmannosamine 0.5 M(squares). Their effect on complement activity compared to the absence ofliposomes was determined by complement hemolytic assays (as described in theMaterials and Methods). Each point represents the mean of 2 experiments eachdone in duplicate.Figure 20B. Complement Consumption in human serum by PC:CH:PG(35:45:20 mol%) liposomes in the presence of mannose bindingprotein inhibitors. Large unilamellar vesicles (100 nm) composed of PC:CH:PG(45:45:20 mol%) were serially titrated and incubated with NHS (circles), or NHS inthe presence of mannose 0.5 M (triangles) or N-acetylmannosamine 0.5 M(squares). Their effect on complement activity compared to the absence ofliposomes was determined by complement hemolytic assays (as described in theMaterials and Methods). Each point represents the mean of 2 experiments eachdone in duplicate.-94-FIGURE 20Panel ALiposome Concentration (mM)c0.001 0.01 0.1 1 10 100Liposome Concentration (mM)00C04-.Q.EDC’)C004-Ca,Ea,0E00:10075.502501007550250Panel B0.001 0.01 0.1 10 100-95-Figure 21A. Complement Consumption in immunoglobulin depleted serum byPC:CH:PG (45:45:10 mol%) liposomes in the presence of mannose bindingprotein inhibitors. Large unilamellar vesicles (100 nm) composed of PC:CH:PG(45:45:10 mol%) were serially titrated and incubated with DDS (circles) or DDS in thepresence of 2 M N-acetylmannosamine (squares). Their effect on complementactivity compared to the absence of liposomes was determined by complementhemolytic assays (as described in the Materials and Methods). Each pointrepresents the mean of 2 experiments each done in duplicate.Figure 21 B. Complement Consumption in immunoglobulin depleted serumby CH:PG (45:55 mol%) liposomes in the presence of mannose bindingprotein inhibitors. Large unilamellar vesicles (100 nm) composed of CH:PG (45:55mol%) were serially titrated and incubated with DDS (circles) or DDS in the presenceof 2 M N-acetylmannosamine (squares). Their effect on complement activitycompared to the absence of liposomes was determined by complement hemolyticassays (as described in the Materials and Methods). Each point represents 1experiment in triplicate.-96-FIGURE 21Panel A00‘— 100C04-0ED ‘U)c0.001 0.01 0.1 1 10 100Liposome Concentration (mM)Panel BI..• I0.001 001 0.1 1 10 100Liposome Concentration (mM)-97-3.10. Competitive binding studies between Cig and mannose binding protein: Inorder to ascertain whether the experiments described in the previous section can beaccurately interpreted as binding competition between MBP and Clq for the anionicsurface, competitive binding studies were carried out. Figure 22 demonstrates thatBOug/ml and 8 ug/mI of MBP added to the reaction mixture can effectively competewith Clq and inhibit the binding of Clq to the CL containing liposomes by 100%.The concentration of MBP required to observe binding on CL containing liposomesusing immunoblot analysis was 40 ug/mI, however, no binding was observed atphysiological concentrations using the immunoblot technique. This could be due tothe presence of a number of proteins in plasma competing for the anionic surface.In this binding study the competition is limited to two proteins: MBP and Clq. In thiscompetitive binding study, when the concentration of MBP was diluted to 0.8 ug/mI,it appeared that some of the binding of Clq to the CL containing liposomes wasinhibited; however, when the MBP concentration added was diluted further to 0.08ug/mI the binding of Clq fell within the nonspecific binding range as shown withfactor B. The physiological protein concentration ratio of MBP:Clq is 1:1000. Thiswas taken into account when doing the binding study and it was demonstrated thatat this physiological protein concentration (which occurs when 0.8 ug/mI of MBP isadded), some of the binding of Clq was inhibited. As this was only a preliminarystudy, the effect of exogenous competitive ligand such as mannose and ACM in thereaction mixture is yet to be determined. However, if the data demonstratedpreviously is correct in its assumption that B2GP1 will mediate the binding of MBP tothe liposomes, then the amount of Clq that would bind to the CL containingliposomes should be decreased further. The effect of B2GP1 on the competitivebinding between Clq and MBP to CL containing liposomes remains to bedetermined. The binding of Clq to CL liposomes was completely inhibited at loxthe physiological protein concentration (8 ug/mI).-98-Figure 22. Competitive binding study between Clq and mannose bindingprotein in the presence of PC:CH:CL (35:45:20 mol) Large unilamellar vesicles (400nm) composed of PC:CH:CL (35:45:20 mol%) and trace labeled with14C-dipalmitoylphosphatidylcholine were incubated with a mixture of 125-labeledand unlabeled Clq and various concentrations of MBP. The bound protein to theliposomes was separated from the free protein as described in the Materials andMethods. The empty square represents the amount of Clq binding at 0.5 mg/mI.The dark circles represent the nonspecific binding observed by Factor B at 80 ug/mIand 8 ug/mI. The empty triangles, from right to left, are the concentrations of MBPadded 0.08 ug/mI, 0.8 ug/mI, 8 ug/mI and 80 ug/mI. Each point represents themean of an experiment run in duplicate.0.500.400.30Cc’)LJ0.20a. .0.10 AAE 0.00 A0.21 0.23 0.25 0.28 0.30 0.32equilibrium [Clq] uM-99-3.11. Binding of protein S on the surface of the anionic liposomes: As has beenshown in section 3.7, a number of proteins copurify with B2GP1. The remainder ofthe dissertation focuses on one of these proteins: protein S. This protein was ofinterest in light of the current research into antiphospholipid antibodies and theirreputed cofactors. The results obtained from these studies could provide someinsight as to the thrombotic complications experienced by patients withantiphospholipid syndrome (APA). Furthermore, the depletion of protein S byliposome infusion could place patients receiving liposomal drugs at increased risk ofthrombosis. As indicated above, B2GP1 binds avidly to the surface of anionicliposomes (Figure 16). Based on this finding, the next experiment was designed toanswer the question whether the liposome surface upon which B2GP1 is enrichedcan also enrich protein S as it does when immobilized on an affinity column. Figure23 shows that when PC:CH:CL (35:45:20 mol%) liposomes are exposed to NHP andseparated using a Sepharose CL2B column, the same proteins that copurified withaffinity-isolated B2GP1 were also found on the surface of these anionic liposomes.In contrast, when another anionic composition was tested, PC:CH:PG (15:45:40mol%) exposed to NHP; protein S could not be detected using the Western blotanalysis technique. However, when protein S was added exogenously, to aconcentration of 60 ug/ml, the protein S could then be detected on the surface of PGliposomes when the liposomes were concentrated to 4mM. This concentrationrepresents a three-fold excess of physiological concentrations of protein S. As atechnique, the use of immunoblot analysis for the detection of protein S on theliposomes was not easy (Figure 24). This was perhaps due to the use of a directlyperoxidase-conjugated anti-protein S rather than the two-stage antibody bindingassay used for the other Western blots. When a complex using a-100-peroxidase-antiperoxidase was used to amplify the signal, the background levelswere high. We were unable to obtain unlabeled anti-human protein S. It is alsopossible that the low plasma concentrations of protein S (20-25 ug/mI) also madedetection difficult. All Western blots failed to detect Protein S, C4BP or B2GP1 on theneutral liposome composition PC:CH (55:45 mol%).-101-Figure 23. Identification beta-2-glycoprotein-1, protein S and C4 bindingprotein on PC:CH:CL (35:45:20 mol%) liposomes. Large unilamellar vesicles (400*nm) were trace labeled with CH and exposed to plasma and passed through aSepharose CL2B column as previously described. Liposomes were madeequivalent at 5 mM total phospholipid concentration and subjected to SDS-PAGEusing a 10 % minigel under reducing conditions. The proteins were transferred on anitrocellulose blot and probed with the indicated antibodies as described in theMaterials and Methods.-J0z0C)08OKd —,? 0z zC?C? 000 C 66D_ 0 CL Qr75Kd..J1 I5OKd. 50Kd,L4;aProt.S cC4bp a B2GP1-102-Figure 24. Identification of protein S on PC:CH:PG (15:45:40 mol%) liposomes.Three hundred ul of plasma was equilibrated with 1.8 ug of pure Protein S and*incubated with CH containing liposomes (400 nm) and passed through theSepharose CL2B as described in the Materials and Methods. The peak liposomefractions were made equivalent for total phospholipid by concentrating PG andPC:CH containing liposomes to 4 mM total phospholipid. Sixty ul of the liposomesamples were loaded per lane and subjected to SDS-PAGE using a 10 % minigel andrun under reducing conditions. 20 ul of liposome-free plasma samples were loadedper lane (4 mg/mI).Panel A. Lanes 1 contains 15 ul of pure protein S at 0.63 mg/mI. Lane 2 containsrunning buffer. Lanes 3 and 5 contain the liposome-free plasma exposed to PC:CH(55:45 mol%) and PC:CH:PG (1 5:45:40 mol%) respectively at 4mg/mI.Panel B. Lanes 1 and 3 contain PC:CH (55:45 mol%) and PC:CH:PG (15:45:40mol%). The gel was transferred and probed with peroxidase-conjugated sheepanti-human protein S as described in the materials and methods. The film wasexposed to the blot for 2 mm and 15 mm for Panel A and Panel B respectively beforedevelopment.-103-C”LUDC,LI.toU)0AAAAr%CØU)O)coe‘)AA3.12. Immunological protein S levels in plasma exposed to liposomes or anti-B2GP1Sepharose column: The detection of protein S using the Western blot technique didnot prove to be sensitive enough to detect subtle changes. Not only is the sensitivityof the technique an issue, but the Western blot technique cannot differentiatebetween free and bound protein S. This is important because only free protein S canact as an anticoagulant, as discussed in the Introduction. In order to obtain betterquantitative results and in order to differentiate between free and bound, animmunological test to detect free protein S was used. Table 5 shows that withincreasing concentration of liposomes, more protein S is deposited on the surface ofanionic liposomes. This is particularly evident for the PG containing liposomes. TheNHP was diluted (3 vol:1 vol) in Tris HCI buffer and used as a control for each of theliposome experiments. These results do not appear to be consistent with theWestern blot analysis as no protein S was detected on the surface of the PGliposomes using Western blot analysis unless exogenous protein S was added. Theimmunological test for detection of agglutination appeared to be a more sensitivemethod of detection and hence the subtle changes in levels were measurable.In order to control for the effects of incubation time and handling, protein Slevels were measured in the handling control. Table 6 shows that the handlingcontrol (temperature control) which should be at approximately 0.5 S.l. Units hasbeen reduced to 0.36 S.l. Units implying, that even at 4 °C, the antigenic levels ofprotein S may be deteriorating. There also appears to be some nonspecific bindingthat was not detected using Western blot analysis (due to the sensitivity of thetechnique) as demonstrated by the 14 % reduction in protein S levels observed afterincubation with rabbit lgG-Sepharose. The depletion of protein S from the plasmaexposed to the anti-human B2GP1 column was statistically significant compared tothat of the control column and was reported as 0.20 +1- 0.04 S.l. Units; however,-105-since the standard curve was not linear below 0.25 S.l. Units, the level was reportedas <0.25 S.l. Units. This corresponded to a 31 % reduction in comparison to thehandling control.The data in Table 5 and 6 demonstrated that the depletion of Protein Sappears to be associated with the depletion of B2GP1 either by immunoaffinitychromatography or on the surface of an anionic liposome. The levels of Protein Shave been previously shown to be reduced in patients with APA for which B2GP1 is areputed cofactor (Freyssinet, J. et al 1987, Lubbe, W. et al 1983, Costing, J. et al,1993, Santoro, S. 1994). Patients with high antiphospholipid antibody titers areknown to have a disposition to thrombosis.-106-TABLE 5. Immunological protein S levels in plasma exposed to liposomes.Plasma was exposed to liposomes at various concentrations (Table 5) anddelipidated (as described in the Materials and Methods). The samples were thenimmunologically quantitated for free protein S using the Liatest Assay. Values shownare means ± 1 standard deviation of the mean. Undiluted reference plasma levelwas 1.0 by definition.Final Liposome Free Protein S % reduction pConcentration Levels cf PC:CH value25 mM (n=4) S.I. units LiposomesPC:CH (55:45 mol%) 0.69 ± 0.06 0PC:CH:PG (15:45:40 mol%) 0.51 ± 0.19 26 <0.1NHP 0.74 ± 0.01 NAFinal Liposome Free Protein S % reduction pConcentration Levels Cf PC:CH value7 mM (n=4 in triplicate) S.I. Units LiposomesPC:CH (55:45 mol%) 0.72 ± 0.007 0PC:CH:PG (15:45:40 mol%) 0.65 ± 0.02 10 <0.05PC:CH:CL (35:45:20 mol%) 0.70 ± 0.04 3 NSNHP 0.76 ± 0.02 NANA= Not applicableNS= Not significant-107-TABLE 6 Immunological protein S levels in beta-2-glycoprotein-1-depleted plasma.Plasma was exposed to anti-human B2GP1-Sepharose as described in theMaterials and Methods. Duplicate plasma samples were then immunologicallyquantitated for free protein S using the Liatest Assay.Protein Concentration Free Protein S % reduction50 mg/mi (n = 6) Levels cf NHPS.I._UnitsAnti-human B2GP1 <0.25 31(0.20 ±_0.04)Rabbit IgG 0.31 ± 0.03 14NHY 0.36 ± 0.02*NHP : This NHP sample was handled as the chromatographed specimen withrespect to exposure to surfaces, time and temperature. In the text this is refered toas the handling or temperature control.-108-3.13. Characterization of the functional activity of protein S in the plasma exposed toliposomes or anti-beta-2-glycoprotein-1 Sepharose: The experiments described inthe previous section addressed the antigenic levels of protein S; however, thisimmunological assay cannot assess the functional status of protein S. To assess thefunctional effect of protein S interaction with anionic liposomes, the functional activityof protein S was determined using a commercial assay which measures the proteinS-dependent prolongation of factor Va-mediated coagulation. Table 7 shows thatthe time for clotting was shortened in the plasma exposed to the anionic liposomesindicating that free protein S had been removed from the system which is reflectedas a decrease in the anticoagulant activity of the remaining protein S. The timebetween addition of calcium chloride and clot formation is translated into apercentage of the NHP based on a generated standard curve which is linearbetween 25-120 % (see Materials and Methods). Plasma exposed to PG liposomescontained 32 % of protein S functional activity while plasma exposed to CLliposomes retained 50 % activity. These data show a similar trend to theimmunological levels of protein S where less protein S was found in the plasmaexposed to the PG liposomes. The specimens exposed to PG liposomes werepredicted to have a shorter clotting time as only free protein S acts as ananticoagulant.In order to determine whether the diminished activity of protein S isdependent on the depletion of B2GP1, the activity of protein S was measured inplasma depleted of B2GP1 (Table 8). Percentages for the sample plasmas (ieplasma exposed to rabbit lgG-Sepharose and the handling control) were above 120% due to the greater protein concentration (13.5 mg/mI) of these samples comparedwith the concentration of the NHP used to generate the standard curve (9.7 mg/mI).As mentioned in the Materials and Methods, the optimum protein concentration of-109-plasma samples for this experiment is 10 mg/mI. Functional protein S levels in thethawed plasma samples decreased with increasing time, therefore, the protein assaywas determined after the Staclot assay was performed. The temperature control(average 148 %) fell within 6 % of the theoretical value of the number of secondstaken to clot the plasma (118 seconds actual vs 126 seconds theoretical). Thereforethe temperature control, despite registering above the linearity of the curve (25-120%) is well within the sensitivity of the assay. The results indicate that the depletion offree protein S was dependent on the depletion of B2GP1.-110-TABLE 7. Functional protein S levels in plasma exposed to liposomes.Plasma was exposed to Jiposomes and delipidated (as described in theMaterials and Methods). The delipidated plasma samples were then tested, intriplicate, for functional activity of Protein S using the Staclot Assay.Final Liposome Clotting Time % of NHPConcentration (secs)7 mM_(n=2)PC:CH (55:45) 107 ± 5 78PC:CH:PG (15:45:40) 83.5 ± 6 32PC:CH:CL (35:45:20) 89.2 ± 2 50NHP 106 75-111-TABLE 8. Functional protein S levels in beta-2-glycoprotein-1 depleted plasma.Plasma was exposed to anti-human B2GP1-Sepharose as described in theMaterials and Methods. Duplicate plasma samples were then tested for thefunctional activity of protein S using the Staclot Assay. Values reported are theaverage of the individual results.Protein Concentration % of NHP % of NHP13.5 mg/mi Experiment #1 Experiment #2Anti-human B2GP1 68 51Rabbit IgG 125 103NHP 156 141NHP: This NHP sample was handled as the chromatographed specimen withrespect to exposure to surfaces, time and temperature. In the text this is refered toas the handling or temperature control.-112-4. DISCUSSIONThe studies within this dissertation have investigated the in vitro interaction ofnegatively charged liposomes with the classical complement pathway and aparticular aspect of the coagulation cascade. These pathways, as explained in thisIntroduction, may play a role in the half-life of circulating liposomes.There have been a number of mechanisms suggested to explain the in vivoclearance of phospholipid vesicles. One postulated mechanism is the scavengerreceptor found on macrophages that are able to recruit and internalize anionicphospholipids (Nishikawa, K. et al 1990). Using the term ‘receptor-mediated”liposome-cell interaction requires the existence of cell surface proteins that bind aspecific chemical structure (ligand) on the liposome surface and mediate endosomeformation which contains the ligand-receptor complex (Pastan, I. et al 1981, Heleius,A. et al 1983). This theory is controversial in the literature and there have not beenany studies to isolate the putative scavenger receptor for liposomes.Another mechanism which allows for the removal of liposomes from thecirculation involves the binding to liposomes of particular adhesive proteins such asfibronectin (Hsu, M. et al 1982) that will allow for nonspecific uptake by macrophagesthrough receptors that specifically recognize the amino acid sequence RGD(arginine-glycine- aspartate) present within the adhesive protein (Hynes, R. 1987).Macrophages bear receptors for certain apolipoproteins (Williams, K. et al 1987,Bisgaier, C. et al 1989) thus certain apolipoproteins bound to the surface of theliposome may also serve as opsonins for the clearance of the liposome.The last mechanism is the opsonization by lgG or components of thecomplement system; these proteins facilitate uptake using the Fc receptors (Petty,-113-H. et al 1980) or the complement-receptors present on macrophages (Roerdink, al 1983, Wassef, N., 1987). The main complement receptors include CR1, CR3and CR4 found on phagocytic cells, all of which have an affinity for degradationproducts of C3 (Fearon, D. et al 1983 and 1984, Ross, G. et al 1985). This becamethe focus of the first part of my studies.The data reported in the first section of the study demonstrate that unilamellarphospholipid vesicles bearing a net negative charge activate the classical pathway inan antibody-independent manner. This is a novel finding as no studies havepreviously demonstrated, in a whole serum system, that activation of the classicalcomplement pathway by liposomes occurs in the absence of immunoglobulins. Theabsence of complement-activating immunoglobulins from an otherwise intact serumsystem did not block the ability of negatively charged liposomes to activatecomplement. In the studies reported herein, evidence exists that the classicalpathway activation is independent of the presence of anti-phospholipid antibodies orother complement-activating immunoglobulins and that negatively chargedliposomes can deplete Clq from serum while liposomes lacking a net surfacecharge do not.The observation of Clq-mediated activation of complement bycardiolipin-containing liposomes in the absence of antibody is supported by studiesof Kovacsovics et al (1985) which reported the activation of purified 125l-C1 byliposomes containing cardiolipin. However, in those studies, cardiolipin was the onlyphospholipid which bound Clq; the other negatively charged phospholipids, P1,showed no activation. In our experimental system, all negatively chargedphospholipids tested induced complement activation by liposomes. Theinconsistency between these two studies may be due to methodological differences.-114-In the Kovacsovics study, phospholipids were used at 30-70 uM concentrations,depending on the composition. The functional hemolytic assays within thisdissertation have used a range of concentration from 0-60 mM total phospholipid.The assay also used liposomes that contained the same theoretical net surfacecharge. Thus, both the relative concentrations and the theoretical net surfacecharge differed in the studies conducted herein and by Kovacsovics. The datapresented in this dissertation indicates that when the overall charge is the same,negatively charged phospholipids are similar in their Clq binding ability.Most of the substances thusfar reported to activate complement by anantibody-independent Cl activation do so by virtue of their high surface density ofnegative charges such as phosphate or sulphate groups (Loos, M., 1982). Thestructure of CL consists of four acyl chains, buried within the plane of the membraneand two negatively charged phosphate groups, inserted between three derivatives ofglycerol. These negatively charged phosphate groups are exposed on the outsideof the membrane and may provide a possible site for Clq binding (Kovacsovics, T.,1985). Our data suggest that other negatively charged liposomes such as PG and P1may, in fact, act in a similar manner where the phosphate groups are also exposedon the outside of the membrane. It is important to note that the amount ofcomplement fragments such as iC3b deposited on cardiolipin containing liposomesexceeded that found when Pt was present but was comparable to liposomescontaining PG. Thus, it is possible that the chemical structure of the anionicphospholipid may affect the amount of complement deposition. Furthermore, directcomparisons between the results of the hemolytic or Clq assays and immunoblotanalysis must be made with caution. The former assays measure overallcomplement activation in the fluid phase of the reaction; the latter quantitates-115-complement activation fragments that bind to the liposome surface - itself highlydependent on the availability of acceptor sites. Hence while CL and PG bear anuexposedui negative charge, the negative charge of Pt is hidden within the inositol ringhence the charge maybe “shielded”. The idea of shielded charge of Pt has beenproposed earlier by Gabizon, A. et at (1988). Previous studies have shown that P1containing liposomes exposed to serum may in fact have a suppressive effect withregard to liposome uptake by macrophages (Allen, T. et at 1989, Wassef, N., et at1987, Wassef, N. et at 1991, Wassef, N. et at 1993). The studies mentioned aboveused P1 at low mol% (< 10 mol%) which even in the studies herein did notdemonstrate complement consumption as complement consumption was notobserved until P1 was present within the liposome at 40 mol%. In our studies, theshielding of the charge may account for a decrease in complement consumptionboth in normal human serum and in immunoglobulin depleted serum in comparisonto PG and CL containing tiposomes. Furthermore, the amount of iC3b deposited onthe surface of Pt liposomes was observed to be less than that on PG and CLliposomes. The direct binding of purified C3b to inositol has been previouslydemonstrated (Law, S.A. et al 1981) hence the presence of iC3b on the surface of Pttiposomes is not surprising.The dissection of the molecular interactions occurring between complementproteins and liposomes is most readily carried out in purified protein systems.However, in the absence of other serum proteins, protein-lipid interactions may takeplace that are not seen in a whole serum system. Therefore this model was selectedto analyze complement activation in a system in which all other serum proteins werepresent and in which a competitive environment exists.The pathway by which complement is activated in serum by liposomes has-116-not been completely defined and may vary by experimental model. Largemultilamellar vesicles (MLV) composed of hydrogenated egg PC, OH and dicetylphosphate, which bear a net negative charge, have recently been shown to activatethe alternative pathway in rat plasma (Funato, K. et al, 1992). The studies reportedhere cannot be directly compared to those of Funato et al owing to significantdifferences in liposome composition with respect to charge density, size, fluidity andphospholipids as well as the ratio of protein to liposomal lipid. It has been found thatliposome size, charge density and cholesterol content have a definite effect oncomplement consumption in rat serum (Devine, D. et al, 1994). In addition, thestudies reported here used human complement; and complement activation hasbeen shown to vary considerably among species (Houle, J. et al, 1984). Importantly,we found no rise in the level of the activation fragment Bb as measured by ELISA norwas any fragment Bb detected on the liposome surface using Western blottechniques. These studies did, however, detect iC3b on the surface of negativelycharged liposomes that had been incubated in human serum. This observation isconsistent with enzymatic complement activation rather than adsorption ofcomplement factors onto the liposome surface occurred. As previously discussed inthis Introduction, C3b is bound to a surface only if the C3 convertase, C4b2a, isformed. C3b can then be cleaved by factor I to form iC3b. If C3b is formed in thefluid phase and cleaved by factor I to form iC3b it can no longer bind to a surface.The extent of complement activation by liposomes remains to be determined. Thesestudies failed to detect C7 bound to the liposome surface using immunoblottingmethods and were able to visualize only a small amount of C9. Whethercomplement activation does not proceed efficiently past the C3b step on 100 nmvesicles or whether the detection system was insensitive remains to be determined.-117-The next section of the studies examined the contribution of MBP towardscomplement activation in the absence of immunoglobulins and the nature of thephysiological ligand to which MBP may bind in normal serum or plasma. Thehypothesis tested was that the ligand was B2GP1. This glycoprotein avidly binds toanionic liposomes and is found in abundance in plasma (0.2 mg/mI) and contains anumber of carbohydrate residues specific for MBP.When using normal human serum as a source of MBP it was observed thatthe absorbance (i.e. the levels of MBP) were higher when bound to B2GP1 than to anon specific protein such as bovine serum albumin. This result must be interpretedwith caution as there could be a number of proteins present in NHS that will bind toimmobilized B2GP1 as observed by the immunoaffinity chromatography. Therefore,the levels of MBP detected in this assay using absorbance readings are expected tobe lower due to the competition. More specific and direct evidence of the binding ofMBP to B2GP1 was demonstrated using the immunoaffinity column upon whichB2GP1 was immobilized. The binding of MBP to B2GP1 was also demonstrated onthe surlace of CL containing liposomes. This binding was specific as the bindingwas inhibited in the presence of acetylmannosamine. The preliminary results whichdemonstrate the competitive binding between MBP and Clq show that MBP canpartially inhibit Clq binding at physiological concentrations. It is plausible thatbinding of MBP to the anionic liposomes could not be detected at physiologicalconcentrations using Western blot analysis as other plasma proteins would alsocompete for the exposed anionic surface.The binding of MBP to carbohydrate residues has been investigated byKawasaki et al (1983) with the largest affinity being to N-acetylmannosamine andN-acetylglucosamine residues. Whether Clq possesses a similar affinity forcarbohydrate residues remains controversial. There has been a study that has-118-reported the ability of Clq to precipitate polysaccharides (Uhlenbruck, G. et al 1979)and certain lipopolysaccharides have been shown to bind to Clq and mediateclassical pathway activation (Cooper, N. et al 1978). It has been suggested that thebinding of Clq to lgG2 molecules involve lectin binding as monoclonal antibodiesshowed activation of the classical pathway mediated by an antibody-antigencomplex which was shown to be abolished if the antibodies were non-glycosylated(Nose, M. et al 1983). However, to refer to Clq as a lectin remains controversial inlight of a report by Duncan et al (1988) which showed that a short peptide sequencederived from the Fc region of lgG and which was free of carbohydrate served as aninhibitor of Clq binding. However the role of MBP as having lectin bindingproperties to the above mentioned carbohydrate residues has been observed inother complement activators, CRP and SAP. It has been demonstrated that purifiedhuman CRP immobilized onto polystyrene surfaces or onto latex beads will binddistinct glycoproteins including lgG, IgA, 1gM, asialobeta-2-glycoprotein 1 andasialofetuin (Kottgen, E. et al 1992). Kubak et al (1988) demonstrated the binding ofSAP to mannose residues and to glycoproteins which contain mannose terminatedsequences such as iC3b. Therefore, it appears that MBP is similar to othercomplement activators in its glycoprotein binding ability providing the glycoproteincontains specific carbohydrate residues. The studies within this dissertation haveshown that this glycoprotein is B2GP1.With regard to the contribution of MBP towards complement consumption, itwas demonstrated within this study that regardless of the presence or absence ofimmunoglobulins, complement consumption by liposomes appeared to occurpredominantly through the Clq pathway. This was observed in the presence of avarying amount of anionic phospholipid content and in the presence of a varyingconcentration of MBP inhibitor (Figures 20A and B, 21A and B). This finding is-119-interesting as literature demonstrates that the activation of the classical pathwaydoes occur in the presence of MBP and the absence of Clq as judged by theconsumption of C4 and formation of the heavy chain of Cis as detected byautoradiography (Ohta, M. et al 1990, Lu, J. et al 1990). It is difficult to make anaccurate comparison of the relative efficiency with which MBP and Clq can activateClr2Cls2 as studies have used different ligands, i.e., zymosan for MBP andaggregated lgG for Clq. The preliminary study of competitive binding (Figure 22)indicates that even at physiological concentrations MBP binds to CL containingliposomes, while the complement consumption data do not confirm that complementconsumption is significantly affected at physiological concentrations. Inhibition ofMBP only gave rise to a modest increase in complement consumption by theliposomes which implies that Clq is the more efficient activator of the classicalpathway in this model. This suggests that there may be other factors present in NHSthat contribute to the binding affinities of MBP and Clq. The nature of these factorsremains to be determined.No complement consumption assays were carried out when exogenous MBPwas added but it is tempting to speculate that the recognition of the carbohydratestructures, immobilized on a surface by MBP (e.g. B2GP1 on liposomes), couldbecome important if MBP truly acts as an acute phase reactant as suggested byEzekowitz et al (1988). Upon binding to its carbohydrate ligand MBP can trigger thecomplement cascade. This leads to the activation of a number of effectormechanisms in the blood which are important in defense through the generation ofinflammation and through the lysis and clearance of pathogens (Thiel, S. et al 1989,Law, S. et al 1988).In the process of depleting B2GP1 from plasma it was observed that anumber of proteins copurify with B2GP1; two of which included protein S and C4BP.-120-This proved interesting as C4BP is a complement regulator and both protein S andB2GP1 have been demonstrated to be potential cofactors for antiphospholipidantibodies as previously discussed. The presence of C4BP was also demonstratedon the surface of anionic liposomes discussed (data not shown) and is a regulator ofprotein S. However, the copurification of protein S with B2GP1 posed the questionas to whether immobilized B2GP1 can regulate the activity of protein S. Theemphasis of the remainder of the project focused on protein S and the effect of itsdepletion from plasma. As previously discussed, whilst bound protein S does notfunction as an anticoagulant, free protein S is known to be a cofactor for activatedprotein C.The depletion of B2GP1 from plasma using an immunoaffinity columnefficiently removes close to 90 % of circulating B2GP1, however, the immobilizationof B2GP1 from plasma onto the surface of the anionic liposome is not as efficient.This probably is due to the affinity differences between antibody and its ligand versusthe affinity for anionic surfaces. The studies within this dissertation demonstratedthat protein S was colocalized when B2GP1 was mobilized on the surface of anaffinity column or when localized on the surface of the anionic liposome of a certainconcentration. However the immunoblot analysis detections (Figures 23 and 24) donot elucidate the form of protein S detected: hence the need for the immunologicalassay. Free protein S was removed from plasma and was bound to B2GP1 whenthe latter was immobilized on a column (Table 6) or on the surface of an ionicphospholipid (Table 5). The results sumarized in table 5 demonstrated that there is arelationship between the amount of total phospholipid and amount of protein S. Itappeared that the greater the amount of total phospholipid containing an anioniccomponent, the greater the amount of protein S deposited. It is possible that withincreasing concentration of total phospholipid, there is more anionic surface-121-available for B2GP1 and hence more protein S is removed when B2GP1 isimmobilized on the anionic surface. These preliminary data, however, provided agood indication that a certain and substantial amount of free protein S is beingdepleted from the system such that the functionality of this protein is affected (Table7 and 8).The discrepancy between the CL and PG containing liposomes found usingthe immunoblot analysis , immunological and functional assays may be due tostructural differences between the two anionic phospholipids. Phospholipids bind allcomponents of the coagulation factor activating complexes, thus facilitating theinteractions and reactions between the participating proteins. Knowledge of thechemical and physical properties of procoagulant surfaces is still based on earlyobservations of Bangham (1961) and Papahadjopoulos et al (1962), who proposedthat procoagulant membranes should have a net negative surface charge (Bangham1961) and that the procoagulant activity of membranes depends mainly on the valueof the surface charge. The chemical nature of the anionic phospholipids appearedto be much less important (Papahadjopoulos 1962). However, in a study conductedby Rosing et al (1988), it was suggested that a net negative surface charge was notthe criteria for a procoagulant surface. This became evident as PS-containingvesicles changed from a negative to a positive net charge with the insertion ofincreasing amounts of sterylamine (a cationic phospholipid). The change in surfacecharge did not affect the vesicles’ activity in prothrombin activation. In this case, thechanges in the ability of a surface to act as a procoagulant surface were aftributed tothe polar head group chemistry. A second feature of lipid structure that caninfluence the cofactor activity of protein S is the structure of the fatty acid side chain.Increasing chain lengths tend to decrease the effectiveness of protein S. In theseries going from C12-C18 a decrease in the effectiveness of the lipids was-122-observed in the inactivation of factor Va. When the chain length was fixed and thedegree of unsaturation was varied, activity tended to increase with unsaturation. Thespecies of negatively charged phospholipids used in this study include cardiolipinwhich has an 18 carbon fatty acid chain with 2 unsaturated sites on each chain. Thephosphatidyiglycerol molecule has two predominant fatty acid chains: a 16:0 and an18:1. Other features of membrane structure can also alter the activity of factor Vainactivation complex such as cholesterol content. As the amount of cholesterol wasfixed at 45 mol% within these vesicles, this is probably not a contributing factor(Walker, F. 1988). These may explain the differences between the plasma exposedto CL liposomes and the plasma exposed to the PG liposomes.Smirnov et al (1994) have also recently demonstrated that a neutralphospholipid such as phosphatidylethanolamine can enhance factor Va inactivationby activated protein C which indicates that there may be factors, other than simply anegative charge, involved. This may include overall phospholipid structure and mayexplain the difference in the data acquired for PG and CL liposomes (Table 5 and 7).The patho-physiological importance of the protein S-C4BP binding has beendemonstrated by findings in cases of familial selective deficiency of free protein S(Comp, P. et al 1984, Schwarz, H. et al 1984). Affected family members have normalconcentrations of total protein S in plasma but most of it is complexed with C4BPand thus not active as an anticoagulant cofactor. Family members do not have highconcentrations of C4BP, which might otherwise explain their low level of free proteinS concentrations. The molecular mechanism responsible for the protein S-C4BPdisequilibrium in these patients is not known and has not to date been determined.The studies within this dissertation provide a third component that may regulate thisinteraction: B2GP1 when it is immobilized on the surface of the exposed anionic-123-phospholipid of the cell. However, this statement can not be completely justifiedwithout making certain that the C4BP levels do not increase in the eluate. Theimmunoaffinity chromatography data (Table 4) show that the levels of C4BP are notenriched by the removal of B2GP1 as demonstrated for the protein S levels.The studies which compare levels of free protein S and B2GP1 conducted inpatients with antiphospholipid antibody syndrome who have antibodies directedtowards anionic phospholipids must be interpreted with caution. It has beendemonstrated that there exists a subpopulation of antiphospholipid antibodies thatrecognize protein S/C as cofactors (Costing, J. et al 1993).Based on the studies herein, B2GP1 is an important glycoprotein whenimmobilized on a surface. However, it is hard to imagine this glycoprotein circulatesin vivo attached to all the proteins identified in Table 4 as the circulating molecularweight would be greater than 1 million daltons (1000 kD). This circulating complexwould probably be signaled for immune clearance.Reported studies have shown that negatively charged phospholipids have abenefit in the entrapment of solutes especially enzymes (Gregoriadis, G. et al, 1980),in the use of liposomes as adjuvants (Allison, A. et all 1974), negatively chargedPS-containing liposomes are used in targetting to the lung after i.v. administration ofvesicles (Poste, G. et al 1982), incorporation of gangliosides can reduce leakage ofaqueous space markers for lipsoomes in plasma (Allen, T et al 1985). These dataalso demonstrate the need for surface-modified liposomes as the interactionsdescribed above would lead to the altered half-life of the circulating liposomes.Recent reports of several studies with surface-modified liposomes containingPEG-PE have demonstrated enhanced stability in plasma and provided reducedleakage of contents (Blume, G. et al 1990, Senior, J. et al 1991). There has been an-124-implication as to reduced protein adsorption on liposomes containing PEG-PE andthat is accompanied by reduced uptake by macrophages (for a review Allen, T.1992). This implies that there is reduced protein adsorption on the liposomes whichresults in reduced uptake by the macrophages. There have also been extensivestudies as to the targetting of liposomes. A targeting ligand in this context is a groupor molecule expressed on the surface of liposomes intended to allow liposomes tointeract with the corresponding receptor as a means of directing liposomes and theircontents away from the reticuloendothelial system and towards a disease site e.g.tumor cell or to deliever drugs more efficiently to sites with thin the reticuloendothelialsystem. Liposome targeting is demonstrated in vivo through carbohydrate-recognition systems of the liver which have been used as target cells forgalactose-bearing liposomes or the coating of liposomes with antibodies andtargeting to specific cells that bear the antigen. The potential relvance of suchtargeting to cancer cells has been reviewed by Poste et al (Poste, G. 1983). Theresearch of targeting the liposome to the site of disease or disorder is growing andby the same token so is the field investigating the metabolism of the drug into thetarget cell. However, the answer to the question raised as to what is the optimummodification remains to be determined since standard liposome compositions do notexist. Nevertheless, further advances are required in the understanding of themechanisms involved which would permit even better control of biologicalinteractions and further expand the success for therapeutic applications.-125-5. SummaryThe overall objective of this project was to investigate the interaction betweenliposomes, specifically negatively charged liposomes and proteins of thecomplement and coagulation reaction cascades as these proteins may influence thestability and circulating half-life of liposomes if they are to act as therapeutic drugdelivery units.The overall objective had three main hypotheses to be tested. The first wasbased on the reported studies that LUVs that possess an overall negative chargecan activate the classical complement pathway; the mechanism of activation,however, remained to be determined. The results within this dissertation show thatnegatively charged liposomes do not require activating immunoglobulins to activatethe classical pathway of complement. This was shown via complement hemolyticassays and direct Clq consumption assays. The deposition of iC3b was alsopresent on all three liposome compositions used containing CL, PG and P1. Thisobservation is novel as no study conducted to date has carried out the experimentsin a whole serum system.The second hypothesis to be tested was the role of other plasma proteins,specifically that of mannose binding protein, on the activation of the complementpathway in the absence of immunoglobulins. This particular protein was chosen as ithas been demonstrated that it can directly activate the classical pathway in theabsence of not only immunoglobulins but also Clq. The studies conducted hereinshow that despite the binding of MBP at physiological concentrations toCL-containing liposomes, the ability to consume complement is not as effective asliterature claims for MBP. The primary route of activation of the classical pathway by-126-negatively charged liposomes is through Clq. The observation that MBP can bindto B2GP1 is specific and inhibitable. However, the data suggest that this is not theligand through which MBP, at physiological concentrations, seems to be effective atactivating complement.In the attempt to find other physiological ligands for MBP, it was observedthat other proteins were found to bind to B2GP1. One such protein was protein S.This lead to the investigation to determine whether this could affect the hemostaticbalance. The data presented herein show that free protein S was bound toimmobilized B2GP1. This is true whether the B2GP1 was immobilized on the surfaceof a column or on the surface of negatively charged liposomes. The removal of freeprotein S from plasma was translated as a loss in functional activity in the protein S.Furthermore, the data from the immunoaffinity chromatography indicates that C4BPwas not enriched implying that the removal of protein S from the system waspredominantly due to the immobilized B2GP1. However, further investigation toquantitate the levels of C4BP (using immunological assays) and to determine howthe binding of B2GP1 affects the regulatory activity of C4BP will allow for completeconfirmation that immobilized B2GP1 is solely responsible for the depletion of freeprotein S from the plasma system.-127-BIBLIOGRAPHYAbsolom, D.R.Opsonins and dysopsonins: an overview. Meth.Enzymol. 132:281-318, 1986.Allen, T.M., Ryan, J.L., Papahadjopoulos, D.Gasngliosides reduce leakage of aqueous-space markers from liposomes inthe presence of human plasma. Biochim.Biophys.Acta. 818:205-210, 1985.Allen, T.M., Hansen, C., Rutledge, J.Liposomes with prolonged circulation times: factors affecting uptake byreticuloendothelial and other tissues. Biochim. Biophys.Acta. 981:27-35,1989.Allen, T.M.Stealth liposomes: five years on. J.Liposome Res. 2:289-306, 1992.Almeda, S., Rosenberg, R.D., Bing, D.H.The binding properties of human complement component Clq. Interactionwith mucopolysaccharides. J. Biol.Chem. 258:785-791, 1983.Altura, B.Reticuloendothelial cells and host defense. Ad. Microcirc. 9:252-294, 1980.Alving, C.R., Richards, R., Guirguis, A.A.Cholesterol-dependent human complement activation resulting in damage toliposomal model membranes. J.lmmunol. 118:342-347, 1977.Alving, C., Swartz, G. Jr., Wassef, N.Naturally occuring autoantibodies to cholesterol in humans.Biochem.Soc.Trans. 17:637-639, 1989.Anderson, J., Stroud, R., Volanakis, J.Studies on the binding specificity of human C-Reactive Protein forphosphorylcholine. Fed.Proc. 37:1495, (Abstr.), 1978.Aspegren, K., Strand, S.E., Persson, B.R.Quantitative lymph scintigraphy for detection of metastases to the internalmammary lymph nodes. Biokinetics of [99m]Tc-sulphur colloids uptake andcorrelation with microscopy. Acta Radiol:Oncol. Rad. Phys. Bid. 17:17-26,1978.Augener, W., Grey, H.M. Cooper, N.R., Muller-Eberhard, H.J.The reaction of monomeric and aggregated immunoglobulins with Cl.Immunochemistry 8:1011-1020, 1971.-128-Ballou, S.P., Kushner, I.C-reactive Protein and the acute phase response. Adv.Int.Med. 37:313-336.,1992.Ballow, M.Phylogenetics and Ontogenetics of the complement systems. InComprehensive Immunology 2. Biological Amplification systems inImmunology. Day, N.K. Good, R.A. (eds), Plenum, New York, p.183-204,1977.Bangham, A.D.Correlation between surface charge and coagulant action of phospholipids.Nature 192:1197-1198, 1961.Bangham, A.D., Home, R.W.Negative staining of phospholipids and their structural modification bysurface active agents as observed in the electron microscope. J.Mol.Biol.8:660-668, 1964.Bangham, A.D., Hill, M.W., Miller, N.G.A.In Methods in Membrane Biology. Korn, E.D. (ed), Plenum, New York,p.1-68, 1974.Barenholzt, Y., Amsselem, S. and Lichtenberg, D.A new method for preparation of phospholipid vesicles (liposomes)-FrenchPress. FEBS.Lett. 99:210-214, 1979.Bergqvist, L., Strand, S.E., and Persson, B.R.Particle sizing and biokinetics of interstitial lymphoscintigraphic agents,Semin.Nucl.Med. 13:9-19, 1983.Bertina, R.M., van Wijngaarden, A., Reinalda-Poot, J., Poort, S., Born, V.Determination of plasma protein S. The protein cofactor of activated proteinC. Thromb.Haemost. 53:268-272, 1985.Bevers, E.M., Galli, M., Barbui, T., Comfurius, P., Zwaal, R.F.Lupus anticoagulant IgG’s (LA) are not directed to phospholipids only but toa complex of lipid-bound human prothrombin. Thromb. and Haemost.66:629-632, 1991.Bhakdi, S., Tranum-Jensen, J.C5b-9 assembly:average binding of one C9 molecule to C5b-8 without polyC9 formation generates a stable transmembrane pore. J.Immunol.136:2999-3005,1986.Bisgaier, C.L., Siebenkas, M.V., Williams, K.J.Effects of apolipoprotein AIV and Al on the uptake of phospholipid liposomesby hepatocytes. J.BioI.Chem. 264:862-866, 1989.Black, C.D.V., Gregoriadis, G.Interaction of liposomes with blood plasma proteins. Biochem.Soc.Trans.4:253, 256, 1976.-129-Blume, G., Cevc, G.Liposomes for the sustained drug release in vivo. Biochim.Biophys.Acta.1029:91-97, 1990.Bokisch, V.A., Dierich, M.P., Muller-Eberhard, H.J.Third component of complement (C3):structural properties in relation tofunctions. Proc.NatI.Acad.Sci. USA 72:1989-1993, 1975.Bradfield, J.W.B.Control of spillover, the importance of Kupifer-cell function in clinicalmedicine. Lancet 2:883-886, 1974.Breathnach, S., Kofler, H., Sepp, N., Ashworth, J., Woodrow, D., Pepys, M.,Hintner, H.Serum amyloid P component binds to cell nuclei in vitro and to in vivodeposits of extracellular chromatin in systemic lupus erythematosus.J.Exp.Med. 170:1433-1438, 1989.Bristow, C., Boackle, R.Evidence for the binding of human serum amyloid P component to Clq andFab. Mol.lmmunol. 23:1045-1052, 1986.Burstein, M., Legmann, P.Precipitation of triglyceride rich lipoprotein by synthetic anionic detergents.Role ofB2Glycoprotein 1. Prot.Biol.Fluids.Proc.Colloq. 25:407-410, 1977.Chonn, A.,Role of blood proteins in liposome clearance. PhD Thesis, The University ofBritish Columbia, Vancouver, Canada, 1991.Chonn, A., Cullis, P.R., Devine, D.V.The role of surface charge in the activation of the classical and alternativepathways of complement by liposomes. J.lmmunol. 146:4234-4241, 1991.Chonn, A., Semple, S.C., Cullis, P.R.Association of blood proteins with large unilamellar liposomes in vivo.Relation to circulation lifetimes. J. Biol.Chem. 267:18759-18765, 1992.Chung, L.P., Reid, K.B.M.Structural and functional studies on C4b-binding protein, a regulatorycomponent of the human complement system. Biosci.Rep. 5:855-865, 1985.Claus, D.R., Osmand, A.P., Gewurz, H.Radioimmunoassay of human C-Reactive Protein and levels in normal sera.J.Lab.Clin.Med. 87:120-128, 1976.Claus, D., Siegel, J., Petras, K., Osmand, A.P., Gewurz, H.Interaction of C-Reactive Protein with the first component of humancomplement. J.lmmunol. 119:187-192, 1 977a.-130-Claus, D.R., Siegal, J., Petras, K., Skor, D., Osmand, A.P., Gewurz, H.Complement activation by interaction of polyanions and polycations. Ill.Complement activation by interaction of multiple polyanions and polycationsin the presence of C-reactive protein. J.lmmunol. 118:83-87, 1977b.Comp, P., Nixon, R., Cooper, M., Esmon, C.T.Familial Protein S deficiency is associated with recurrent thrombosis.J.Clin.lnvest. 74:2082-2088, 1984.Cooper, N.R., Jensen, F., Welsh, R., Oldstone, M.B.Lysis of RNA tumor viruses by human serum: direct antibody independenttriggering of the classical complement pathway. J.Exp.Med. 144:970-984,1976.Coune, A., Sculier, J.P., Fruhling, J., Stryckmans, P., Brassine, C., Ghanem, G.,Laduron, C., Atassi, G., Ruysschaert, J.M., Hildebrand, J.l.V. Administration of water insoluble antimitotic compound entrapped inliposomes. Preliminary report on infusion of large volumes of liposomes toman. Cancer.Treat.Rep. 67:1031-1033, 1983.Cunningham, C., Kingzette, M., Richards, R., Alving, C., Lint, T., Gewurz, H.Activation of human complement by liposomes: A model for membraneactivation of the alternative pathway. J. Immunol. 122:1237-1242, 1979.Dahiback, B and Stenflo, J.High molecular weight complex in human plasma between vitaminK-dependent Protein S and complement component C4b binding protein.Proc. Natl.Acad.Sci. 78:2512-2516, 1981.Dahiback, B, Smith, C., Muller-Eberhard, H.Visualization of human C4b-binding protein and its complexes with vitamin kdependent protein S and complement protein C4b. Proc.Natl.Acad.Sci.USA80:3461-3465, 1 983a.Dahiback, B., Hilderbrand, B.Degradation of human complement component C4b in the presence of C4bbinding protein-protein S complex. Biochem.J. 209:857-863, 1 983b.Dahlback, B.Purification of human C4b-binding protein and formation of its complex withvitamin K dependent protein S. Biochem. J. 209:847-856, 1983c.Dahlback, B.Interaction between vitamin K dependent protein S and the complementprotein C4b-binding protein. A link between coagulation and thecomplement system. Sem.Thromb.Hemostas. 10:139-148, 1984.-131-Dahiback, B.Inhibition of protein Ca cofactor function of human and bovine proteins byC4b binding protein. J.BioI.Chem. 261:12022-12027, 1986.Damus, P., Hicks, M., Rosenberg, R.Anticoagulant action of heparin. Nature. 246:355-357, Beer, F., Baltz, M., Holford, S., Feinstein, A., Pepys, MFibronectin and C4-binding protein are selectively bound by aggregatedamyloid P component. J.Exp. Med. 154:1134-1139, 1981.Demel, R.A., De Kruijff, B.The function of sterols in membranes. Biochim.Biophys.Acta. 457:109-132,1976.Derksen, R., Kater, L.Lupus anticoagulant:Revival of an old-phenomenon. Clin.Exp.Rheumatol.3:349-357, 1985.Devine, D.The regulation of complement on cell surfaces. Transf.Med.Rev.Vol.V:123-131, 1991.Devine, D., Wong, K., Serrano, K., Chonn, A., Cullis, P.Liposome-complement interactions in rat serum. Implications for liposomesurvival studies. Biochim.Biophys.Acta 1191 :43-51, 1994.DiScipio, R.G.The conversion of human complement C5 into fragment C5b by thealternative pathway C5 convertase. Biochem.J. 199:497-504, 1981.DiScipio, R.G.The size, shape and stability of complement component C9. Mol.lmmunol.30:1097-1106, 1993.Doolittle, R.F.The evolution of vertebrate blood coagulation:a case of yin and yang.Thromb.Haemost. 70:24-28, 1993.Duncan, A. and Winter, G.The binding site for Clq on lgG. Nature 332:738-740, 1988.Esmon, C., Suttie, J., Jackson, C.The functional significance of vitamin K action:difference in phospholipidbinding between normal and abnormal prothrombin. J.Bio.Chem.250:4095-4099, 1975.Ezekowitz, R.A.B., Day, L.E., Herman, G.A.A human mannose-binding protein is an acute phase reactant that sharessequence homology with other vertebrate lectins. J.Exp.Med.167:1034-1046, 1988.-132-Fearon, D.T., Wang, W.W.Complement ligand-receptor interactions that mediate biological responses.Ann.Rev.lmmunol. 1:243-271, 1983.Fearon, Di.Cellular receptors for fragments of the third component of complement.lmmunol.Today 5:105-110, 1984.Feinstein, A., Munn, E.A., Richardson, N.E.The three-dimensional conformation of gamma M and gamma A globulinmolecules. Ann.N.Y.Acad.Sci. 190:104-122, 1971.Feinstein, D.J.Lupus anticoagulant, thrombosis and fetal loss. N.Eng.J.Med.313:1348-1350, 1985.Finlayson, J.S., Mushinski, J.F.Separation of subfractions of humanB2glycoprotein 1.Biochim.Biophys.Acta 147:413-420, 1967.Fishelson, Z., Muller-Eberhard, H.J.C3 convertase of human complement:enhanced formation and stability of theenzyme generated with nickel instead of magnesium. J.lmmunol.129:2603-2607, 1983.Fiske, C.H. and Subbarow, Y.The colorimetric determination of phosphorus. J.Biol.Chem. 66:375-400,1924.Freyssinet, J., Cazenave, J.Lupus-like anticoagulants. Modulation of the Protein C pathway andthrombosis. Thromb.Haemost. 58:678-681, 1987.Frier, M.Phagocytosis, in Progress in Radiopharmacology, (I), Cox, P.M. (ed.),Elsiever/North-Holland, Amsterdam, p.243, 1981.Fritz, I.What is clusterin? Clin.Exp.Immunol. 88:375, 1992.Fujita, T., Kamato, T., Tamura, N.Characterization of functional properties of C4-binding protein by monoclonalantibodies. J.lmmunol. 134:3320-3324, 1985.Fulcher, C., Gardiner, J., Griffin, J., Zimmerman, T.Proteolytic inactivation of human factor VIII procoagulant protein by activatedprotein C and its analogy with factor V. Blood 63:486-489, 1984.Funato, K., Yoda, R., Kiwada, H.Contribution of complement system on destabilization of hydrogenated eggphosphatidylcholine in rat fresh plasma. Biochim.Biophys.Acta.1103:198-204, 1992.-133-Furie, B. and Furie, B.Molecular and Cellular Biology of Blood Coagulation. The New Eng.J.Med.326:800-806, 1992.Gabizon, A. and Papahadjopoulos, D.Liposome formulations with prolonged circulation time in blood andenhanced uptake by tumors. Proc.Natl.Acad.Sci. USA 85:6949-6953, 1988.Galli, M., Comfurius, P., Maassen, C., Hemker, H., de Baets, M., vanBreda-Vriesman, P., Barbuo, T., Zwaal, R. and Bevers, E.Anticardiolipin antibodies (ACA) directed not to cardiolipin but to a plasmaprotein cofactor. Lancet 335:1544-1 547, 1990.Garred, P., Thiel, S., Madsen, H., Ryder, L., Jensenius, J.C., Svejgaard, A.Gene frequency and partial protein characterization of an allelic variant ofMBP associated with low serum concentrations. Clin.Exp.lmmunol.90:517-521, 1992.Geiger, B., Gitler, C., Calef, E., Arnon, R.Dynamics of antibody and lectin-mediated endocytosis of hapten-containingliposomes by murine macrophages. Eur.J.lmmunol 11:710-716, 1981.Giclas, P.C., Ginsberg, M., Cooper, N.R.Immunoglobulin G independent activation of the classical complementpathway by monosodium urate crystals. J.CIin.Invest. 63:759-764, 1979.Girard, T., Warren, L., Novotony, W., Likert, K., Brown, S., Miletich, J., Broze, G.Jr.Functional significane of the Kunitz-type inhibitory domain oflipoprotein-associated coagulation inhibitor. Nature 338:518-520, 1989.Golan, M.D., Burger, R., Loos, M.Conformational changes in Clq after binding to immune complexes:detection of neoantigens with monoclonal antibodies. J.lmmunol.129:445-447, 1982.Gotschlich, E., Edelman, G.Binding properties and specificity of C-Reactive Protein. Biochem.57:706-712, 1967.Gregoriadis, G., Neerunjun, D., Meade, T., Goolamali, S., Weereraatne, H., Bull, G.Experiences after long-term treatment of a type 1 Gaucher disease patientwith liposome-entrapped glucocerebroside: Beta-glucosidase, in EnzymeTherapy in Genetic Disseases: 2, p. 383. Desnick, R.J. (Ed), A.R.Liss, NewYork, 1980.Hafeman, D.G., Lewis, J., McConnell, H.M.Triggering of the macrophages and neutrophil respiratory burst by antibodybound to a spin-label phospholipid hapten in model lipid bilayer membranes.Biochemistry 19:5387-5394, 1980.Haire, W., Newland, J.Protein C deficiency and anticardiolipin antibodies in a family with prematurestroke. Am J.Hematol. 33:61-63, 1990.-134-Hamazaki, H.Ca2+-mediated association of human serum amyloid P component withheparan sulfate and dermatan sulfate. J.Biol.Chem. 262:1456-1460, 1987.Harris, E.N., Gharavi, A., Boey, M., Patel, B., Mackworth-Young, C., Loizou, S.,Hughes, G.R.Anticardiolipin antibodies: Detectiion by radioimmunoassay and associationwith thrombosis in systemic lupus erythematosus. Lancet 2:1211-1214,1983.Haselbeck, A.E., Schickaneder, E., von der Eltz, H., Hosel, W.Structural characterization of glycoprotein carbohydrate chains by usingdigoxigenin-labeled lectins on blots. AnaL Biochem. 191:25-30, 1990.Heleius, A., Mellman, I., Wall, D., and Hubbard, A.Endosomes. Trends. Biochem.Sci. 8:245-250, 1983.Hicks, P.S., Sauinero-Nava, L., Du Cl, T.W., Mold, C.Serum Amyloid P component binds to histones and activates the classicalcomplement pathway. J.Immunol. 149:3689-3694, 1992.Hillarp, A., Hessing, M., Dahlback, B.Protein S binding in relation to the subunit composition of human C4-bindingprotein. FEBS. Lett. 259:53-56, 1989.Hind, C.R., Collins, P., Renn, D., Cook, R., Caspi, D., Baltz, M., Pepys, M.Binding specificity of serum amyloid P component for pyruvate acetal ofgalactose. J.Exp.Med. 159:1058-1069, 1984.Hope, M.J., Bally, M.B., Webb, G., and Cullis, P.R.Production of large unilamellar vesicles by a rapid extrusion procedure.Characterization of size, trapped volume and ability to maintain a membranepotential. Biochim.Biophys.Acta 81 2:55-65, 1985.Houle, J., Hoffmann, E.Evidence for restriction of the ability of complement to lyse homologouserythrocytes. J.Immunol. 133:1444-1452, 1984.Hsu, M.J., Juliano, R.L.Interactions of liposomes with the reticuloendothelial system. II: Nonspecificand receptor mediated uptake of liposomes by mouse peritonealmacrophages. Biochim.Biophys.Acta. 720:411-419, 1982.Huang, C.H.Studies on phosphatidylcholine vesicles:formation and physicalcharacteristics. Biochem. 8:344-352, 1969.Hutchcraft, C.L., Gewurz, H., Hansen, B., Dyck, R., Pepys, M.B.Agglutination of complement-coated erythrocytes by serum amyloidP-component. J.Immunol. 126:1217-1219, 1981.-135-Hynes, R.O.Integrins: a family of cell surface receptors. Cell. 48:549-554, 1987.Ihara, S., Takahashi, A., Hatsuse, H., Sumitomo, K., Dci, K, Kawakami, M.Major component of Ra-reactive factor, a complement-activating bactericidalprotein, in mouse serum. J.lmmunol. 146:1874-1879, 1991.Ikeda, K., Sannoh, T., Kawasaki, N., Kawasaki, T., Yamashina, I.Serum lectin with known structure activates complement through theclassical pathway. J. Biol.Chem. 262:7451-7454, 1987.Isenman, D.E., Podack, E.R., Cooper, N.R.The interaction of C5 with C3b in free solution:a sufficient condition forcleavage by a fluid phase C3/C5 convertase. J.lmmunol. 124:326-331, 1980.Ishizaka, T., Ishizaka, K., Borsos, R., Rapp, H.Complement fixation of C’l by G and M but not by A antibody. J.lmmunol.97:716-726, 1966.Jenne, D., Tschopp, J.Clusterin: the intriguing guises of a widely expressed glycoprotein. TIBS17:154-159, 1992.Ji, Y., Matsushita, M., Okada, H., Fujita, T., Kawakami, M.The C4 and C2 but not Cl components of complement are responsible forthe complement activation triggered by the Ra-reactive factor. J.lmmunol.141:4271-4275, 1988.Jiang, H.X., Siegel, J., Gewurz, H.Binding and complement activation by CRP via the collagen region of Clqand inhibition of these reactions by monoclonal antibodies to CRP and Clq.J.lmmunol. 146:2324-2330, 1991.Jiang, H.X., Robey, F.A., Gewurz, H.Localization of sites through which C-Reactive Protein binds and activatescomplement to residues 14-26 and 76-92 of the human Clq A chain.J.Exp.Med. 175:1373-1379, 1992.Juliano, R.L., Stamp, D.The effect of particle size on the clearance rates of liposome and liposomeencapsulated drugs. Biochem.Biophys.Res.Communic. 63:651-658, 1975.Kamboh, M., Ferrell, R., Sepehrnia, B.Genetic studies of human apolipoproteins IV. Structural heterogeniety ofapolipoprotein H (B2 glycoprotein 1). Am.J.Hum.Genet. 42:452-457, 1988.Kaplan, M.H. and Volanakis, J.Interaction of CRP complexes with the complement system. I. Consumptionof human complement associated with the reaction of C-Reactive Protein withpneumococcal C-polysaccharide and with choline phosphatides, lecithin andsphingomyelin. J.Immunol. 112:2135-2147, 1974.-136-Kawaga, Y. and Pecker, E.Partial resolution of the enzymes catalyzing oxidative phosphorylation.J. BioLChem. 246:5477-5487, 1971.Kawasaki, N., Kawasaki, T., Yamashina, I.Isolation and characterization of a Mannan Binding Protein from HumanSerum. J.Biochem. 94:937-947, 1983.Kawasaki, N., Kawasaki, T., Yamashina, I.A serum lectin (mannan binding protein) has complement-dependentbactercidal activity. J. Biochem.Tokyo. 106:483-489, 1989.Keeling, D.M., Wilson, A.J.G., Mackie, l.J., Machin, S.J., lsenberg, M.Some antiphospholipid antibodies bind to B-Glycoprotein-1 in the absenceof phospholipid. Brit.J.Haematol. 82:571-574’ 1992.Kemkes-Matthes, B.Acquired protein S deficiency. Clin.Investg. 70:529-534, 1992.Kerr, M.A.The human complement system:assembly of the classical pathway C3convertase. Biochem.J. 189:173-181, 1980.Kilpatrick, J.M., Kearney, J., Volanakis, J.Demonstration of calcium induced conformational change(s) in CRP by usingmonoclonal antibody. MoI.lmmunol. 19:1159-1165, 1982.Kirby, C.J., Clarke, J. and Gregoriadis, G.Effect of cholesterol content of small unilamellar liposomes on their stability invivo and in vitro. Biochem.J. 186:591-598, 1980a.Kirby, C.J., Clarke, J., Gregoriadis, G.Cholesterol content of small unilamellar liposomes controls phospholipid tohigh density liposportiens in the presence of serum. FEBS.Lett. 111:324-328,1980b.Klibanov, A.L., Maruyama, K., Torchilin, V., Huang, L.Amphipathic polyethyleneglycols effectively prolong the circulation time ofliposomes. FEBS 268:235-237, 1990.Klibanov, A., Maruyama, K., Beckerleg, A., Torchilin, V., Huang, L.Activity of amphipathic poly(ethylene glycol) 5000 to prolong the circulationtime of liposomes depends on the liposome size and is unfavorable forimmunoliposome binding to target. Biochim. Biophys.Acta. 1062:142-148,1991.KoIb, W.P., KoIb, L.M. and Podack, E.R.Clq:Isolation from human serum in high yield by affinity chromatography anddevelopment of a highly sensitive hemolytic assay. J.Immunol.122:2103-2111, 1979.-137-KoIb-Bachofen, V.A review on the biological properties of C reactive Protein. Immunobiol.183:133-145, 1991.Kottgen, E., Hell, B., Kage, A., Tauber, R.Lectin specificity and binding characteristics of human CRP. J.lmmunol.149:445-453, 1992.Kovacsovics, T., Tschopp, J., Kress, A., Isliker, H.Antibody-independent activation of Cl the first component of complement bycardiolipin. J.lmmunol.135:2695-2700, 1985.Kubak, B., Potempa, L, Anderson, B., Mahklouf, S., Venegas, M., Gewurz, H.,Gewurz, A.Evidence that serum amyloid P component binds to mannose-terminatedsequences of polysaccharides and glycoproteins. Mol.lmmunol. 25:851 -858,1988.Kuhn, S.H., Gemperli, B., Shephard, E., Joubert, J., Weidemann, P., Weissmann, G.,Finkelstein, M.Interaction of liposomes with human leukocytes in whole blood.Biochim.Biophys.Acta. 762:119-127, 1983.Lachmann, P.J. and Hughes-Jones, N.C.Initiation of complement activation. Semin.lmmunopathol. 7:143-162, 1984.Law, S.K.A., Lichtenberg, N.A., Holcombe, F.H., Levine, R.P.Interaction between the labile binding sites of the fourth (C4) and fifth (C5)human complement proteins and erythrocyte cell membranes. J.lmmunol.125:634-639, 1980(a).Law, S.K.A., Lichtenberg, N.A., Holcombe, F.H., Levine, R.P.Covalent binding and hemolytic activity of complement proteins.Proc.Natl.Acad.Sci.USA 77:7194-7198, 1980(b).Law, S.K.A. and Reid, K.Activation and control of the complement. ln:Complement. Male, D (ed.).IRL Press, Oxford, UK, p.9-27, 1988.Lee, N.S., Brewer, H.B.B., Osborne, J.,B-2-Glycoprotein 1:molecular properties of an unusual apolipoprotein H.J.Biol.Chem. 258:4765-4770, 1983.Leserman, L., Weinstein, “Liposomes and lmunobiology: Tom.B., Six, H. Eds. p.241 Elsevier/NorthHolland, New York, 1980a.Leserman, L., Weinstein, J., Blumenthal, R., Terry, W.Receptor-mediated endocytosis of antibody-opsonized liposomes by tumorcells. Proc. Natl.Acad. Sci. USA 77:4089-4093, 1 980b.-138-Lewis, J., Hafeman, D., McConnell, H.Kinetics of antibody-dependent binding of haptenated phsophollpid vesiclesto a macrophage related cell line. Biochemistry 19:5376-5386, 1980.Lichtenberg, D., Freire, E., Schmidt, C.F., Barenholz, Y., FeIgner, P.E. andThompson, T.E.Effect of surface curvature on stability, thermodynamic behavior and osmoticactivity of dipalmitoylphosphatidylcholine single lamellar vesicles.Biochem.20:3462-3467, 1981.Loos, M.Antibody independent activation of Cl the first component of complement.Ann. Immunol. (Paris) 1 33C:1 65-179, 1982.Loos, M., Clas, F., Fischer, W.Interaction of purified lipoteichoic acid with the classical complementpathway. Infect. Immun. 53:595-599, 1986.Loughrey, H.C., Bally, M.B., Reinish, L.W., Cullis, P.R.The binding of phosphatidylglycerol liposomes to rat platelets is mediated bycomplement. Thromb. and Haemost. 64:172-176, 1990.Loveless, R., Sullivan, G., Raynes, J., Yuen, C., Feizi, T.Human serum amyloid P is a multispecific adhesive protein whose ligandsinclude 6-phosphotylated mannose and the 3-sulphated saccharidesgalactose, N-acetylga!actosamine and glucuronic acid. Embo.J. 11:813-819,1992.Lozier, J., Takahashi, N., Putnam, F.Complete amino acid sequence of human plasmaB2glycoprotein 1.Proc. Nat.Acad. Sci. USA 81:3640-3644, 1984.Lu, J., Thiel, S., Wiedemann, H., Timple, R., Reid, K.Binding of the pentamer/hexamer forms of mannan-binding protein tozymosan activates the proenzyme Clr2Cls2 complex of the classicalpathway of complement, without involvement of Clq. J.Immunol.144:2287-2294, 1990.Lubbe, W., Butler, W., Palmer, S., Liggins, G.Fetal survival after prednisone suppression of maternal lupus anticoagulant.Lancet 1:1361-1363, 1983.Lublin, D., Atkinson, J.Decay accelerating factor:Biochemistry, molecular biology and function.Ann.Rev.Immunol. 7:35-58, 1989.Malhotra, R., Haurum, J., Thiel, S., Sim, R.Interaction of Clq receptor with lung surlactant Protein A. Eur.J.Immunol.22:1437-1445, 1992.-139-Marjan, J., Xie, Z., Devine, D.V.Liposome-induced activation of the classical complement pathway does notrequire immunoglobulin. Biochim.Biophys.Acta. 1192:35-44, 1994.Matsuo, H., Funato, K., Harashima, H., Kiwada, H.Complement mediated phagocytosis of mannoside modified liposomes. (B2.Abst.) Liposome Research Days Conference, Vancouver, B.C. 1994.Matsushita, M., Fujita, T.Activation of the classical complement pathway by mannose-binding proteinin association iwth a novel Cis like serine protease. J.Exp.Med.176:1497-1502, 1992.Matsuura, E., Igarashi, Y., Fujimotot, M., Ichikawa, K., Suzuki, T., Sumida, T.,Yasuda, T., Koike, T.Heterogeniety of anticardiolipin antibodies defined by the anticardiolipincofactor. J.lmmunol. 148:3885-3891, 1992.McNeil, H.P., Simpson, R.J., Chesterman, C.N. and Krilis, S.A.Antiphospholipid antibodies are directed against a complex antigen thatincludes a lipid-binding inhibitor of coagulation:B-Glycoprotein-1 (apoH).Proc.Natl.Acad.Sci.USA 87:4120-4124, 1990.Men, S., Morgan, B.P., Davies, A., Daniels, R.H., Olavesen, M.G., Waldmann, H.,Lachman, P.J.Human protectin (CD59), an 18000-20000 MW complement lysis restrictingfactor inhibits C5b-8 catalysed insertion of C9 into lipid bilayers. Immunology71:1-9, 1990.Michalek, M.T., Bremer, E.G., Mold, C.Effect of gangliosides on activation of the alternative pathway of humancomplement. J.lmmunol. 140:1581-1587, 1988.Mimms, L.T., Zampighi, G., Nozaki, Y., Tanford, C. and Reynolds, J.A.Phospholipid vesicle formation and transmembrane protein incorporationusing octylglucoside. Biochem. 20:833-840, 1981.Minta, J.O., and Lepow, l.H.Studies on the subunit structure of human properdin. Immunochemistry11:361-368, 1974.Mold, C., Gewurz, H.Activation of human complement by liposomes:serum factor requirement foralternative pathway activation. J. I mmunol. 125:696-700, 1980.Mold, C., Gewurz, H.Inhibitory effect of C-Reactive protein on alternative pathway activation byliposomes and Streptococcus Pneumoniae. J.lmmunol. 127:2089-2092,1981a.-140-Mold, C., Rodgers, C.P., Richards, R.L., Alving, C., Gewurz, H.Interactions of CRP with Iiposomes.lll. Membrane requirements for binding.J.lmmunol. 126:856-860, 1981b.Mullenix, M. and Mortensen, R.Calcium ion binding regions in C-Reactive protein: localization and regulationof conformational changes. MoLimmunol. 31:615-622, 1994.Murphy, B., Kirszbaum., L, Walker, L, d’Apice, J.SP-40-40 a newly identified normal human serum protein found in the SC5b-6complex of complement and in immune deposits in glomerulonephritis.J.Clin.Invest. 81:1858-1864, 1988.Nagai, K., Ito, Y. and Otsuka, N., Muranaka, A.Deposition of small [99m]Tc labeled colloids in bone marrow and lymphnodes. Eur.J.Nucl.Med. 7:66-70, 1982.Naito, K., Fujikawa, K.Activation of human blood coagulation factor XI independent of factor XII:factor XI is activated by thrombin and factor XIa in the presence of negativelycharged surfaces. J.Biol.Chem. 266:7353-7358, 1991.Nakaya, Y., Schaefer, E., Brewer, H.Activation of human post heparin lipoprotein lipase by apolipoprotein H(B2Glycoprotein 1). Biochem.Biophys.Res.Comm. 95:1168-1172, 1980.Nayar, R., Hope, M.J., and Cullis, PR.Generation of large unilamellar vesicles from long chain saturatedphosphatidyicholines by extrusion technique. Biochim.Biophys.Acta.986:200-206, 1989.Nelsestuen, G., Zytkovicz, T., Howard, J.The mode of action of vitamin K: identification of gamma-carboxyglutamicacid as a component of prothrombin. J.Biol.Chem. 249:6347-6350, 1974.Nishikawa, K., Arai, H., Inoue, K.Scavenger receptor-mediated uptake and metabolism of lipid vesiclescontaining acidic phospholipids by mouse peritoneal macrophages.J.Biol.Chem. 265:5226-5231, 1990.Nose, M., Wigzell, H.Biological significance of carbohydrate chains on monoclonal antibodies.Proc. Natl.Acad. Sci. USA 80:6632-6636, 1983.Ohta, M., Okada, M., Yamashina, I., Kawasaki, T.The mechanism of carbohydrate-mediated complement activation by theserum mannan-binding protein. J.BioI.Chem. 265:1980-1984, 1990.-141-Oldfield, E., Chapman, D.Molecular dynamics of cerebroside-cholesterol andsphingomyelin-cholesterol interactions:implication for myelin membranestructure. FEBS LetI. 21:303-306, 1972.Olson, F., Hunt, C.A., Szoka, E.G., Vail, W.J. and Papahadjopoulos,D.Preparation of liposomes of defined size distribution by extrusion throughpolycarbonate membranes. Biochim. Biophys.Acta 557:9-23, 1979.Oosting, J.D., Derksen, R.H.W.M., Entjes, H.T.I., Bouma, B.N., deGroot, P.G.Lupus anticoagulant is frequently dependent on the precence ofB2-glycoprotein-i. 67:499-502, 1992.Oosting, J.D., Derksen, R., Bobbink, I., Hackeng, T., Bouma, B., deGroot, P.Antiphospholipid antibodies directed against a combination of phospholipidwith prothrombin, Protein C or Protein S: An explanation for their pathogenicmechanism. Blood 81:2618-2625, 1993.Osmand, A.P., Friedneson, B, Gewurz, H, Painter, R., Hofman, T., Shelton, E.The characterization of C-Reactive Protein and Cit as pentraxins:homologous proteins displaying cyclic pentameric symmetry.Proc.Natl.Acad.Sci. USA 74:739-745, 1977.Osterud, B., Rapaport, S.Activation of factor IX by the reaction product of tissue factor and factor VII:additional pathway for initiating blood coagulation. Proc. Natl.Acad . Sci. USA74:5260-5264, 1977.Painter, R, De Escallon, I., Massey, A., Pinteric, L., Stern, S.The structure and binding characteristics of serum amyloid protein (9.5Salphal -glycoprotein). Ann.NY.Acad.Sci. 389:199-215, 1982.Pangburn, M.K., Muller-Eberhard, H.J.Relation of a putative thioester bond in C3 to activation of the alternativepathway and the binding of C3b to biological targets of complement.J.Exp.Med. 152:1102-1114, 1980.Pangburn, M.K., Schreiber, R.D., Muller-Eberhard, H.J.Formation of the initial C3 convertase of the alternative complementpathway. Acquisition of C3b-like activities by spontaneous hydrolysis of theputative thiolester in native C3. J.Exp.Med. 154:856-867, 1981.Papahadjopoulos, D., Hougie, C, Hanahan, D.J.Influence of surface charge of phospholipids on their clot-promoting activity.Proc.Soc.Exp.Med. 111:412-416, 1962.Parente, R.A. and Lentz, B.E.Phase behavior of large unilamellar vesicles composed of syntheticphospholipids. Biochem.23:2353-2362, 1984.-142-Park, Y.S., Maruyama, K., Huang, L.Lipid microsphere containing lipophilic home preparation and oxygentransportation under physiological conditions. Biochim.Biophys.Acta1108:257-260, 1992.Parke, A.L., Bona, R., Maier, D., Walker, F.J.The thrombotic diathesis associated with the presence of phospholipidantibodies may be due to low levels of free Protein S. Am.J.Med. 93:49-56,1992.Pastan, l.H. and Willingham, M.C.Journey to the centre of the cell: role of the receptosome. Science214:504-409, 1981.Patel, H.M.Fate of liposomes in lymphatics in Liposomes as Drug Carriers, RecentTrends and Progress, Gregoriadis, G. (ed.) John Wiley and Sons; New York,51, 1988.Pepys, M., Dash, A., Markham, Thomas, H., Williams, B., Petrie, A.Comparative clinical sstudy of protein SAP (amyloid P component) andC-reactive protein in serum. Clin.Exp.lmmunol. 32:119-124, 1978.Pepys, M., Dyck, R., de Beer, F., Skinner, M., Cohen, A.Binding of serum amyloid P-component by amyloid fibrils. Clin.Exp.lmmunol.38:284-293, 1979.Pepys, M. and Butler, P.Serum amyloid P component is the major calcium dependent specific DNAbinding protein of the serum. Biochem.Biophys.Res.Commun. 148:308-313,1987.Petty, H.R., Hafeman, D.G., McConnell, H.M.Specific antibody dependent phagocytosis of lipid vesicles by RAW264macrophages results in the loss of cell surface Fc but not C3b receptoractivity. J.lmmunol. 125:2391-2396, 1980.Pharmacia in :Affinity Chromatography Principles and Methods.p.15 Ljungforetagen AB, Orebro Sweden, 1979.Pierangeli, S.S., Harris, E.N., Davis, S.A., DeLorenzo, G.B-2-Glycoprotein-1 enhances cardiolipin binding activity but is not the antigenfor antiphospholipid antibodies. Brit.J. Heamatol. 82:565-570, 1992.Pinteric, L. and Painter, R.Electron microscopy of SAP in the presence of calcium: alternative forms ofassembly of pentagonal molecules in 2 dimensional lattices. Can.J.Biochem.57:727-736, 1979.-143-Podack, E.R., Esser, A.F., Biesecker, C., and Muller-Eberhard, H.J.Membrane attack complex of complement. J.Exp.Med. 151:301-313, 1980.Podack. E.R. and Tschopp, J.Circular polymerization of the ninth component of complement.J.Biol.Chem., 257:15204-1 521 2, 1982.Polz, E., Kostner, G.The binding of B-2-Glycoprotein 1 to human serum lipoproteins. FEBS Lefl.102:183-186, 1979.Poste, G., Bucana, C., Raz, A., Bugeiski, P., Kirsh, R., Fidler, l.J.Analysis of the fate of systemically administered liposomes and implicationsfor their use in drug delivery. Cancer. Res. 42:1412-1422, 1982.Poste, G.Liposome targeting in vivo:problem and opportunities. Biol.Cell. 47:19-38,1983.Potempa, L., Kubak, B., Gewurz, H.Effect of divalent metal ions and pH upon the binding reactivity of humanserum amyloid P component, a C-reactive protein homologue, for zymosan.J.BioLChem. 260:12142-12147, 1985.Radcliffe, R., Nemerson, Y.Mechanisms of activation of bovine factor VII: products of cleavage by factorXa. J.Biol.Chem. 251:4749-4802, 1976.Raepple, E., Hill, H.U., Loos M.Mode of interaction of different polyanions with the first (Cl), the second (C2)and the fourth (C4) component of complement.l.effect on fluid phase Cl andon Cl bound to EA or to EAC4. Immunochemistry. 13:251-255, 1976.Rao, L., Rapaport, S.Activation of factor VII bound to tissue factor: a key in early step in the tissuefactor pathway of blood coagulation. Proc.Natl.Acad.Sci. 85:6687-6691,1988.Rapaport, S., Proctor, P., Patch, M., Yettra, M.The mode of inheritance of PTA deficiency:evidence of major PTA deficiencyand minor PTA deficiency. Blood 18:149-165, 1961.Rapaport, S.The extrinsic pathway inhibitor: the regulation of tissue factor dependentblood coagulation. Thromb. Haemost. 66:6-15, 1991.Reid, K.B.M. and Porter, R.R.Subunit composition and structure of Clq. Biochem.J. 155:19-23, 1976.-144-Reid, K.B.M.Chemistry and molecular Genetics of Clq. Behring lnst.Mitt. 84:8-19, 1989Reid, K.B. and Thiel, S.Clq and related molecules in defence. In: Natural lmmunity-Humoralfactor.Kerr, M. and Sim, E. (eds.) Oxford:IRL Press, Oxford University Press,1992.Rent, R., Ertel, N., Eisenstein, R. and Gewurz, H.Complement activation by interaction of polyanions andpolycations.l. Heparin-protamine induced consumption of complement.J.lmmunol. 114:120-124, 1975.Richards, R., Gewurz, H., Siegel, J., Alving, C.Interactions of C-reactive protein and complement with liposomes. II.Influence of membrane composition. J.lmmunol. 122:1185-1189, 1979.Robey, F., Jones, K., Steinberg, A.C-reactive protein mediates the solubilization of nuclear DNA by complementin vitro. J.Exp.Med. 161:1344-1356, 1985.Roerdink, F., Wassef, N., Richardson, E.C., Alving, C.R.Effects of negatively charged lipids on phagocytosis of liposomes opsonizedby complement. Biochim.Biophys.Acta 734:33-39, 1983.Rosing, J., Speijer, H and Zwaal, R.Prothrombin activation on phospholipid membranes with positiveelectrostatic potential. Biochem. 27:8-11, 1988.Ross, G.D., Medof, M.E.Membrane complement receptors specific for bound fragments of C3.Adv.Immunol. 37:21 7-267, 1985.Santoro, S.Antiphospholipid antibodies and thrombotic predisposition: Underlyingpathogenetic mechanisms. Blood 83:2389-2391, 1994.Scherphof, G. and Morselt, H.On the size-dependent disintegration of small unilamellar phosphatidylcholinevesicles in rat plasma: evidence of complete loss of vesicle structure.Biochem.J. 221:423-429, 1984.Schmairer, A., Silverberg, M., Kaplan, A., Colman, R.Contact activation and its abnormalities ln:Hemostasis and Thrombosis BasicPrinciples and Clinical Practice 2nd Edition. Colman, R., Hirsh, J., Marder, Vand Salzman, E. (eds.)Phi?adelphia, J.B.Lippincott, p.18-38, 1987.-145-Schonermark, S., Rauterberg, E.W., Shin, M.L., Loke, S., Roelocke, D.,Hansch, G.M.Homologous species restriction in lysis of human erythrocytes: a membranederived protein with C8-binding capacity functions as an inhibitor.J.lmmunol. 136-1772-1776, 1986.Schousboe, I., Rasmussenm, M.The effect of B-2-Glycoprotein 1 on the dextran sulphate and sulphatideactivation of the contact system (Hageman factor system) in bloodcoagulation. lnt.J.Biochem. 20:787-792, 1988.Schultze, H.E., Heide, K., Haupt, H.Uber em bisher unbekanntes niedermolekulares B2-Globulin desHumanserums. Naturwissenschaften 23:719, 1961.Schwalbe, R., Dahlback, B., Hillarp, A., Nelsestuen, G.Assembly of protein S and C4b binding protein on membranes. J.Biol.Chem.265:16074-16081, 1990a.Schwalbe, R., Dahlback, B., Nelsestuen, G.Independent association of serum amyloid P component, protein S andcomplement C4b with complement C4b-binding protein and subsequentassociation of the complex with membranes. J.Biol.Chern. 265:21749-21757,1990b.Schwalbe, R., Dahlback, B., Nelsestuen, G.Heparin influence on the complex of serum amyloid P component andcomplement C4b-binding protein. J.Biol.Chem. 266:12896-12901, 1991.Schwarz, H., Fischer, M., Hopmeier, P., Batard, M., Griffen, J.Plasma protein S deficiency in familial thrombotic disease. Blood:64:1297-1300, 1984.Senior, J.Fate and behavior of liposomes in vivo:a review of controlling factors.Crit. Rev.Ther. Drug Carrier Syst. 3:123-193, 1987.Senior, J., Delgado, D., Fisher, D., Tilcock, C., Gregoriadis, G.Influence of surface hydropholicity of liposomes on their interaction withplasma protein and clearance from the circulation: studies with poly(ethyleneglycol)-coated vesicles. Biochim.Biophys.Acta. 1062:77-82, 1991.Seya, T., Turner, J.R. and Atkinson, J.P.Purification and characterization of a membrane protein (gp 45-70) that is acofactor for cleavage of C3b and C4b. J.Exp.Med. 163:837-855, 1986.Shi, W., Chong, B.H., Chesterman, C.N.B-2-Glycoprotein-1 is a requirement for anticardiolipin antibodies binding toactivated platelets:differences with lupus anticoagulants. Blood81 :1255-1261, 1993.-146-Shin, M., Paznekas, W., Mayer, M.On the mechanism of membrane damage by complement:the effect of lengthand unsaturation of the acyl chain in liposomal bilayers and the effect ofcholesterol concentration in sheep erythrocyte and liposomal membranes.J.lmmunol.120:1996-2002, 1978.Siegel, J., Osmand, A., Wilson, M., Gewurz, H.Interactions of C-reactive protein with the complement system.ll.CRP-mediated consumption of complement by poly-L-lysine polymers andother polycations. J.Exp.Med. 142:709-721, 1975.Smirnov, M.D., Esmin, C.T.Phosphatidylethanolamine incorporation into vesicles selectively enhancesfactor Va inactivation by activated Protein C. J.Biol.Chem. 269:816-819,1994.Sitomer, G., Stroud, R.M., Mayer, M.M. 2+Reversible adsorption of C’2 by EAC’4:role of Mg , enumeration ofcompetent SAC’4, two-step nature of C’2 fixation and estimation of itsefficiency. Immunochemistry 3:57-69, 1966.Smith, C.A., Pangburn, M.K., Vogel, C-W, Muller-Eberhard, H.J.Molecular architecture of human properdin, a positive regulator of thealternative pathway of complement. J. Biol. Chem. 259:4582-4588, 1984.Sommerman, E.F.Factors influencing the biodistribution of liposomal systems, PhD Thesis, TheUniversity of British Columbia, Vancouver, Canada, 1986.Stenflo, J., Ferlund, P., Egan, W., Roepsstorff, P.Vitamin K dependent modification of glutamic acid residues in prothrombin.Proc.Natl.Acad.Sci.USA 71:2730-2733, 1974.Stuart,J., Whicher,J.Tests for detecting and monitoring the acute phase response.Arch.Dis.Childhood 63:115-117, 1988.Sugita, Y., Tobe, T. Oda, E., Tomita, M., Yasukawa, K., Yamaji, N., Takemoto, T.,Furuichi, K., Takayama, M., Yano, S.Molecular cloning and characterization of MACIF, an inhibitor of membranechannel formation of complement. J.Biochem. 106:555-557, 1989.Super, M., Levinsky, R., Turner, M.The level of mannan-binding protein regulates the binding of complementderived opsonins to mannan and zymosan at low serum concnetrations.Clin.Exp.lmmunol. 79:144-150, 1990.-147-Szoka, F. and Papahadjopoulos, D.Procedure for preparation of liposomes with large internal aqueous spaceand high capture by reverse-phase evaporation. Proc.Natl.Acad.Sci.USA75:4194-4198, 1978.Tack, B.F., Morris, S.C., Prahi, J.W.Third component of human complement:structural analysis of thepolypeptide chains of C3 and C3b. Biochem. 18:1497-1503, 1979.Tack, B.F., Harrison, R.A., Janatova, J., Thomas, M.L., Prahi, J.W.Evidence for the presence of an internal thiolester-bond in the thirdcomponent of complement. Proc.Natl.Acad.Sci.USA 77:5764-5768, 1980.Thiel, S., Holmskov, U., Hviid, L., Laursen, S., Jensenius, J.The concentration of the C-type lectin, mannan-binding protein, in humanplasma increases during an acute phase response. Clin.Exp.lmmunol90:31-35, 1992a.Thiel, S.Mannan-binding protein: a complement activating animal lectin.Immunopharmacology 24:91-99, 1 992b.Tillet, W. and Francis, T.Serological reactions in pneumonia with a non-protein somatic fraction ofpneumococcus. J.Exp.Med.52:561 -571,1930.Uhlenbruck, G., Karduck, D., Haupt, H., Schwick, H.G.C-reactive protein (CRP), 9.5s alpha-i -glycoprotein and Cl q :serum proteinswith lectin properties? Z.lmmunol. 155:262-266, 1979.van Oss, C.J.Phagocytosis as a surface phenomenon. Annu.Rev.Microbiol. 32:19-39,1978.Vermylen, J., Arnout, J.Is the antiphsopholipid syndrome caused by antibodies directed againstphysiologically relevant phospholipid-protein complexes? J.Lab.Clin.Med.120:10-12, 1992.Vogt, W., Schmidt, G., von Buttlar, B., Dieminger, L.A new function of the activated third component of complement: binding toC5 activation, an essential step for C5 activation. Immunology 34:29-40,1978.Volanakis, J. and Kaplan, M.Specificity of C-Reactive Protein for choline phosphate residues ofpneumococcal C-polysaccharide. Proc.Soc.Exp.Biol.Med. 136:612-614,1971.-148-Volanakis, J. and Kaplan, M.Interaction of C-Reactive Proetin with the complement system. II.Consumption of guinea pig complement by CRP complexes: requirement forhuman Clq. J.Immunol.113:9-17, 1974.von Zeipel, G., Hanson, H.S., von Stedingk, L.V.Purification from euglobin of the first component (Cl) of complement and itssubcomponents by heparin-Sepharose chromatography,Acta.Path.Micrtobiol.Scand.Sect.C 85:123-130, 1977.Walker, F.J.Regulation of activated protein C by a new protein. J.Biol.Chem.255:5521-5524, 1980.Walker, F.J.Regulation of bovine activated protein C by protein S: the role of the cofactorprotein in species specificity. Thromb.Res. 22:321-327, 1981 a.Walker, F.J.Regulation of activated Protein C by Protein S. The role of phospholipid infactor Va inactivation. J.Biol.Chem. 256:11128-11131, 1981b.Walker, F.J., Chavin, S., Fay, P.Inactivation of factor VIII by activated protein C and protein S.Arch. Biochem. Biophys. 252:322-328, 1987.Walker, F.J.Interactions of Protein S with membranes. Sem.Thromb.Hemostas.14:216-221, 1988.Wassef, N., Alving, C.Complement dependent phagocytosis of Iiposomes by macrophages.Methods in Enzymology. 149:124-134, 1987.Wassef, N., Johnson, S., Graeber, G., Swartz, G., Schultz, C., Hailey, J., Johnson,A., Taylor, D., Ridgway, R., Alvind, C.Anaphylactoid reactions mediated by autoantibodies to cholesterol inminiature pigs. J.Immunol. 143:2990-2995, 1989.Wassef, N. and Alving C.Complement-dependent phagocytosis of liposomes. Chem.Phys.Lipids.64:239-248, 1993.Watts, M.J., Dankert, J.R. and Morgan. B.P.Isolation and characterization of a membrane-attack-complex-inhibitingprotein present in human serum and other biological fluids. Biochem.J.265:471-477, 1990.-149-Wessel, D., Flugge, U.J.A method for the quantitative recovery of protein in dilute solution in thepresence of detergents and lipids. Anal.Biochem. 138:141-143, 1984.Whaley, K.Measurement of complement . In: Methods in complement for clinicalimmunologists. p.77 K.Whaley, (ed.) Churchill Livingstone, Edinburgh, 1985.Whicher, J.Acute phase proteins, physiology and clinical uses. Clin.Biochem.Revs.11:4-9, 1990.Williams, K.J., Tall, A.R., Bisgaier, C., Brocia, R.Phospholipid liposomes acquire apolipoprotein E in atherogenic plasma andblock cholesterol loading of cultured macrophages. J.Clin.Invest.79:1466-1472, 1987.Willoughby, W.F., Ford, R.T., Shin, H.S.Complement activation by lysozyme-DNA complexes (Abstr.). J.lmmunol.111:296-297, 1973.Wiman, B., Hamsten, A.The fibrinolytic enzyme system and its role in the etiology of thromboembolicdisease. Sem.Throm.Hemost. 16:207-216, 1990.Wong, M., Anthony, F.H., Tillack, T.W. and Thompson, T.E.Fusion of dipalmitoylphosphatidylcholine vesicles at 4 C. Biochem.21 :41 26-4132, 1982.Woodle, M.C. and Lasic, D.D.Sterically stabilized liposomes. Biochim.Biophys.Acta 1113:171-199, 1992.Woodle, M.C., Matthay, K., Newman, M.S., Hidayat, J.E., Collins, L.R.,Redemann, C. Martin, F.J., Papahadjopoulos, D.Versitilty in lipids compositions showing prolonged circulation with stericallystabilized Iiposomes. Biochim.Biophys.Acta. 1105:193-200, 1992.Wurm, H.B-2-Glycoprotein 1 (Apolipoprotein H) interactions with phospholipid vesicles.Int.J.Biochem. 16:511-514, 1984.Ying, S., Jiang, H., Gewurz, A., Gewurz, H.Human serum amyloid P component binds and activates the classicalcomplement pathway via collagen-like region of Clq. FASEB.J.(Abst.)6:A1451, 1992.Ying, S., Gewurz, A., Jiang, H., Gewurz, H.Human serum amyloid P component oligomers bind and activate theclassical complement pathway via residues 14-26 and 76-92 of the A chaincollagen-like region of Clq. J.lmmunol. 150:169-176, 1993.-150-Zalman, L., Wood, L., Muller-Eberhard, H.Isolation of a human erythrocyte membrane protein capable of inhibitingexpression of homologous complement transmembrane channels.Proc.Nat.Acad.Sci.USA. 82:7711-7715, 1986.Zur, M., Radcliffe, R., Oberdick, J., Nemerson, Y.The dual role of factor VII in blood coagulation: initiation and inhibition of aproteolytic system by a zymogen. J.Biol.Chem. 257:5623-5631, 1982.-151-


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items