UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

In vivo targeting of liposomal drug carriers Longman, Shane A. 1994

You are currently on our download blacklist and unable to view media. You will be unbanned within an hour.
To un-ban yourself please visit the following link and solve the reCAPTCHA, we will then redirect you back here.

Item Metadata


831-ubc_1994-953586.pdf [ 3.34MB ]
JSON: 831-1.0088209.json
JSON-LD: 831-1.0088209-ld.json
RDF/XML (Pretty): 831-1.0088209-rdf.xml
RDF/JSON: 831-1.0088209-rdf.json
Turtle: 831-1.0088209-turtle.txt
N-Triples: 831-1.0088209-rdf-ntriples.txt
Original Record: 831-1.0088209-source.json
Full Text

Full Text

IN VIVO TARGETING OF LIPOSOMAL DRUG CARRIERSbySHANE A. LONGMANB.Sc. University of Guelph, 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENT FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of BiochemistryWe accept this thesis conformingto the required standardzQ.......;.THE UNIVERSITY OF BRITISH COLUMBIAAugust, 1994(J Shane A. Longman, 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 C)(LQArL \The University of British ColumbiaVancouver, CanadaDate ODE-6 (2/88)ABSTRACTStudies in this thesis were aimed at optimizing the in vivo targeting properties ofliposomal drug carriers. A major obstacle to targeted delivery of liposomal carriersconcerns access to the target site and subsequent binding to the target cell population invivo. This thesis presents strategies to enhance circulation lifetimes of liposomes withsurface-associated targeting proteins as a means to improve access of these carriers to atarget site. Further, a model approach to investigate binding of liposomal drugs to targets invivo is developed.This research effort was initiated following an investigation focused ondemonstrating in vivo targeting of drug (doxorubicin)-loaded large unilamellar liposomeswith covalently attached streptavidin (SA-LUV5) to a murine lymphocytic leukaemia cellline (P388) residing in the peritoneal cavity of mice. Streptavidin, a protein derived fromStreptomyces avidinii, has 4 high-affinity binding sites for biotin. A two-step targetingprocedure was used in which target cells were pre-labeled with a biotinylated monoclonalantibody, specific for the Thy 1.2 antigen expressed on P388 cells, prior to systemicadministration of SA-LUVs. In vitro studies based on this two-step procedure resulted in a30-fold and 20-fold increase in cell-associated lipid and doxorubicin, respectively, fordoxorubicin-loaded SA-LUVs compared to protein-free liposomes or to conditions wherethe target cells were not pre-labeled with biotinylated antibody. Under optimal conditionsapproximately 6,000 SA-LUVs were specifically bound per cell using this procedure. Invivo targeting of drug-loaded SA-LUVs (injected intraperitoneally or intravenously) to11P388 target cells was also achieved. However, the specificity and the efficiency of targetingwas significantly less than expected on the basis of in vitro results. One explanation for thereduced efficiency concerns the fact that only a small amount (<3%) of the injected lipiddose reached the peritoneal cavity following intravenous administration. Surprisingly,however, under conditions where high local concentration of liposomes was achieved (i.p.injection of SA-LUVs), the specificity of the targeting was also less than that achieved invitro. Non-specific binding of liposomes was largely a consequence of liposome uptake bytumor-associated macrophages.Observations suggesting that SA-LUVs have limited access to the target site led tostudies designed to improve passive targeting to the peritoneal cavity. These studies werebased on the hypothesis that procedures that increased circulating blood levels of injected(i.v.) liposomes would increase the total amount of lipid that accessed an extravascular site.Two procedures were developed to increase the circulation lifetimes of intravenouslyinjected SA-LUVs. The first procedure involved blockade of liposome uptake byphagocytic cells in the liver with a low pre-dose (2 mg/kg drug) of liposomal doxorubicin.The second involved the incorporation of a polyethylene glycol-modified phospholipid(PEG2000-DSPE) in SA-LUVs. It was shown that incorporation of up to 2 mol% PEG-PE inliposomes resulted in an improved protein-coated liposome that exhibited optimal sizecharacteristics as well as efficient binding to target cells in vitro. It was established thateach of these procedures prolonged circulation lifetimes, decreased uptake in liver andincreased accumulation in the peritoneal cavity following intravenously administration ofSA-LUVs. Combining the strategies of liver blockade and incorporation of PEG-PE further111increased circulation lifetimes and decreased liver uptake of SA-LUVs, however there wasno further increase in passive targeting to the peritoneal cavity. The presence of anestablished P388 tumor in the peritoneal cavity markedly increased the passive targeting ofSA-LUVs to that location. These results are of interest in terms of developing anunderstanding of the mechanism by which liposomes leave the vascular compartment.In order to investigate why there was reduced in vivo binding of SA-LUVs to targetcells, a model approach was developed based on using biotin-labeled multilamellar vesiclesas a target. SA-LUVs incorporating 2 mol% PEG-PE were found to bind optimally tomultilamellar vesicles that incorporated biotinoylaniinohexanoyl DSPE (BAH-MLV). Thisbinding was not reduced in the presence of normal mouse serum and SA-LUVs isolatedfrom the blood of mice previously injected (i.v.) with the liposomes, exhibit no change inbinding characteristics. In vivo studies based on SA-LUVs injected intraperitoneaflydemonstrated a 17-fold and 8-fold increase in binding to BAH-MLVs in the peritonealcavity of tumor-free and tumor-bearing animals, respectively, compared to non-targetedsystems. The extent of targeting achieved under these conditions was comparable to thatobserved in vitro. SA-LUVs injected intravenously demonstrated a 5-fold increase inbinding compared to both tumor-free and tumor-bearing animals. These studies wereextended to a solid tumor model where it was shown that the presence of intratumorallyinjected BAH-MLVs promoted the accumulation of i.v. administered SA-LLJVs. SA-LUVsinjected intravenously into mice bearing subcutaneous Lewis lung tumors accumulated 2-fold greater in tumors that had been injected with BAH-MLVs than tumors injected withMLVs.‘VTABLE OF CONTENTSABSTRACTTABLE OF CONTENTS vLIST OF FIGURES viiiLIST OF TABLES xiABBREVIATIONS xiiiACKNOWLEDGEMENTS xvDEDICATION xviChapter 1. INTRODUCTION 11.1 Targeted Drug Carriers: A Perspective 11.2 Liposomes 101.2.1 Chemical and Physical Properties of Lipids Commonly Used in LiposomalDrug Carriers Phospholipids Cholesterol 161.2.2 Liposome Preparation Multilamellar Vesicles (MLVs) Unilamellar Vesicles (SUV and LUV) 211.2.3 Drug Encapsulation Passive Trapping Techniques Active Trapping Techniques 261.2.4 Coupling of Targeting Ligands to Liposomes Covalently Coupled Antibodies 311.2.4.2 Non-Covalently Associated Ligands 351.3 Factors Controlling the Phannacodynamic Behavior of Liposomal Drugs 371.3.1 Interactions with Plasma Proteins 381.3.2 Interactions with the Reticuloendothial System (RES) 411.3.2.1 Strategies to Avoid the RES 441.3.3 Passive Targeting of Liposomal Drug Carriers 461.3.3.1 Normal Vascular Structure 461.3.3.2 Vascular Structure Within a Diseased Site 491.4 Ligand-Directed Targeting of Liposomal Drugs 51V1.4.1 Targeting Strategies .521.4.2 In Vivo Targeting of Liposomal Drugs 561.5 Objectives 56Chapter 2. MATERIALS AND METHODS 582.1 Materials 582.2 Preparation of Liposomes 602.3 SPDP Modification of Streptavidin 612.4 Coupling of SPDP-Streptavidin to Liposomes 632.5 Doxorubicin Uptake into Streptavidin-Coated Vesicles 652.6 Biotinylated Anti-Thy 1.2 Antibody Binding to P388 Cells 662.7 In Vitro Targeting of Streptavidin-Coated Liposomes to P388 Cells 662.8 Animals 672.9 Flow Cytometric Analysis 692.10 Fluorescent Microscopy 702.11 Cytotoxicity Assays 702.12 Separation of MLVs and LUVs 71Chapter 3. IN VIVO TARGETING OF STREPTAVIDIN LWOSOMES TO P388 CELLS ... 723.1 Introduction 723.2 Results 743.2.1 Characterisation of Doxorubicin-Loaded Streptavidin Liposomes 743.2.2 In Vivo Labeling of P388 Cells with Biotinylated Anti-Thy 1.2 Ab 803.2.3 In Vitro Targeting of Doxorubicin-Loaded Streptavidin Liposomesto Pre-labeled P388 Cells 843.2.4 In Vivo Targeting of Doxorubicin-Loaded Streptavidin Liposomesto Pre-labeled P388 Cells 903.3 Discussion 92Chapter 4. INCREASED EXTRAVASATION OF STREPTAVIDIN LIPOSOMES 974.1 Introduction 974.2 Results 1004.2.1 Influence of Blockade on Clearance and Biodistribution 1004.2.2 Influence of PEG on Liposome Aggregation, Targeting to P388Cells, and Clearance and Biodistribution 1044.2.3 Influence of Blockade and PEG on Clearance and Biodistribution 1074.2.4 Influence of Tumor Presence on Clearance and Biodistribution 1124.3 Discussion 114Chapter 5. TARGETING OF STREPTAVIDN LIPOSOMES TO BIOTIN-MLVs 1195.1 Introduction 119vi5.2 Results 1225.2.1 In Vitro Characterization of Binding of Streptavidin Liposomesto Various Biotin-MLVs 1225.2.2 In Vitro Binding of Streptavidin Liposomes (recovered fromplasma) to Biotin-MLVs 1265.2.3 In Vivo Binding of Streptavidin Liposomes to Biotin-MLVs 1325.3 Discussion 136Chapter 6. SUMMARIZING DISCUSSION 141REFERENCES 147VI’LIST OF FIGURESFigure 1.1Amphipathic Lipids in Bilayer Configuration 11Figure 1.2General Structure of a Phospholipid Showing Commonly Occurring Headgroupsand Fatty Acid Moieties 14Figure 1.3Polymorphic Phases Available to Lipids on Hydration 17Figure 1.4Multilamellar and Unilamellar Vesicles 19Figure 1.5Illustration of Different Drug Encapsulation Protocols with an Indication as to DrugDistribution 25Figure 1.6Redistribution of Weak Bases in Response to Transmembrane pH Gradients 30Figure 1.7Schematic Diagram of the Structure of Three Classes of Blood Capillaries 48Figure 1.8Targeting of Protein-Coated Liposomes to Cells 54Figure 2.1Structures of Various Modified Lipids 59Figure 2.2SPDP Modification of Streptavidin 62Figure 2.3Coupling of SPDP-Modified Streptavidin to Liposomes 64Figure 3.1Plasma Clearance of Streptavidin-Coated and Antibody-Coated Liposomes 76Figure 3.2Effect of Encapsulated Doxorubicin on Circulation Lifetimes of StreptavidinLiposomes 79VI”Figure 3.3Biotinylated anti-Thy 1.2 Antibody Binding to P388 Cells 81Figure 3.4In Vivo Labeling of P388 Cells with Antibody Injected Intraperitoneally 82Figure 3.5Phase Contrast and Fluorescent Micrographs of P388 Cells Labeled In Vivo withanti-Thy 1.2 Antibody 83Figure 3.6Phase Contrast and Fluorescent Micrographs of Targeted Liposomes In Vitro 85Figure 3.7Targeting of Streptavidin Liposomes to P388 Cells In Vitro 86Figure 3.8Quantification of Cell-Associated Doxorubicin and Lipid After Targeting ofDoxorubicin-Loaded Streptavidin Liposomes to P388 Cells In Vitro 88Figure 3.9In Vivo Targeting of P388 Cells with Doxorubicin-Loaded Streptavidin LiposomesInjected Intraperitoneally 91Figure 3.10In Vivo Targeting of P388 Cells with Doxorubicin-Loaded Streptavidin LiposomesInjected Intravenously 93Figure 4.1Biodistribution of SA-LUVs Following Liver Blockade 101Figure 4.2Accumulation of SA-LUVs in the Peritoneal Cavity Following Liver Blockade 103Figure 4.3Biodistribution of SA-LUVs Incorporating Various Mol% PEG2000-DSPE 106Figure 4.4Quantification of Cell-Associated Lipid after Targeting of SA-LUVs IncorporatingVarious mol% PEG2000-DSPE to P388 Cells In Vitro 108Figure 4.5Biodistribution of SA-LUVs Incorporating 2 mol% PEG2000-DSPE Following LiverBlockade 110lxFigure 4.6Accumulation of SA-LUVs Incorporating 2 mol% PEG2000-DSPE in the PeritonealCavity Following Liver Blockade 111Figure 4.7Influence of Blockade, Incorporation of 2 mol% PEG2000-DSPE on CirculationLifetimes and Liver Accumulation of SA-LUVs given (i.v.) to P388 Tumor-Bearing Animals 113Figure 4.8Influence of Blockade, Incorporation of 2 mol% PEG2000-DSPE on theAccumulation of SA-LUVs in the Peritoneal Cavity of P388 Tumor-BearingAnimals 115Figure 5.1Influence of Serum on the Binding of SA-LUVs to MLVs Incorporating DifferentBiotin-Labeled Phospholipids 125Figure 5.2Binding of SA-LUVs to BAH-MLVs 127Figure 5.3Competitive Inhibition of SA-LUV Binding to BAH-MLVs with Free Biotin orSA-LUVs 128Figure 5.4Influence of In Vivo Incubation of SA-LUVs in the Mouse Blood Compartment onBinding to BAH-MLVs In Vitro 130Figure 5.5Targeting of SA-LUVs (injected i.v.) to BAH-LUVs at a Subcutaneous Site 137xLIST OF TABLESTable 1.1Comparison of Different Drug Carriers Used for Systemic Delivery 4Table 1.2Liposome-Based Drug Formulations Undergoing Clinical Evaluation in Humans 5Table 1.3Progression of Liposome Drug Carrier Technology 7Table 1.4Effect of Acyl Chain Length, Degree of Saturation and Head Group on PhaseTransition Temperature 15Table 1.5Passive Trapping Properties of Liposomal Preparations 27Table 1.6Uptake and Retention of Lipophilic Amine Drugs that Accumulate Inside Liposomesin Response to Transmembrane pH Gradients 29Table 1.7Alternatives to Antibody-Based Targeting Ligands That Have Been Used to TargetLiposomes 32Table 1.8Summary of Covalent Coupling Techniques 33Table 3.1Doxorubicin Loading and Retention in Liposomes and Streptavidin-CoatedLiposomes 78Table 3.2IC50 Values for Liposomal and Free Doxorubicin 89Table 4.1Effect of PEG-PE on Coupling of Streptavidin to MPB-PE LUVs 105Table 5.1Separation of LUVs and MLVs 123Table 5.2Retention of Vesicles Injected in the Peritoneal Cavity 131xlTable 5.3In Vivo Targeting of SA-LUVs to BAH-LUVs 134xl’ABBREVIATIONSABTS 2,2’-azino-di-(3-ethylbenzthiazdine sulfonate)ApoA-1 apolipoprotein A-iBSA bovine serum albuminCHOL cholesterolDNP dinitrophenylDOX doxorubicinDTI’ dithiothreitolEDTA ethylenediaminetetraacetic acidFCS fetal calf serumGM1 monosialoganglioside GHAMA human anti-mouse antibodiesHBS HEPES-buffered salineHBSS Hanks buffered saline solutionHDL high density lipoproteinsHEPES N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acidi.p. intraperitoneali.v. intravenousIgG immunoglobulin GliposomesBAH-MLVs BAH-DSPE-labeled multilamellar vesiclesFATMLVs frozen-and-thawed multilamellar vesiclesLUVs large unilamellar vesiclesMLVs multilamellar vesiclesMoAb-SA-LUVs SA-LUVs with bound biotinylated antibodySA-LUVs streptavidin liposomesSUVs small unilamellar vesiclesMoAb monoclonal antibodyNEM N-ethylmaleimidexl”NGF nerve growth factorPBS phosphate-buffered salinep.c. peritoneal cavityphospholipidsB-DOPE N-biotinoyl dioleoyl phosphatidylethanolamineB-DSPE N-biotinoyl distearoyl phosphatidylethanolamineBAH-DSPE N-biotinoylaminohexanoyl distearoylphosphatidylethanolamineDPPC dipalmitoyl phosphatidyicholineDSPC distearoyl phosphatidylcholineDSPE distearoyl phosphatidylethanolamineEPC egg phosphatidyicholineMPB-DPPE N-(4-(P-maleimidophenyl)butyryl) dipalmitoylphosphatidylethanolaminePA phosphatidic acidPC phosphatidylcholinePE phosphatidylethanolaminePEG-DSPE polyethylene glycol-modified distearoylphosphatidylethanolaminePG phosphatidylglycerolPT phosphatidylinositolQELS quasi-elastic light scatteringRES reticuloendothelial systemRT room temperatureSA streptavidinSAMSA 5-acetylmercaptosuccinic anhydrideSATA N-succinimidyl S-acetylthioacetateSMPB N- succinimidyl 4-(p-maleimidophenyl)butyrateSPDP N-succinimidyl 3-(2-pyridyldithio) propionic acidVPF vascular permeability factorVVO vesiculo-vacuolar organellesxivACKNOWLEDGEMENTSSo that I don’t miss anyone, I’ve decided to use algebra. I feel a need to give apaean (my thesaurus is smokin’) to the people of:(A) U (B) U (C) U (D)where:(A) includes all the people present in Pieter’ s lab in the basement of the Biochembuilding. I wasn’t there long or very much, because when courses ended I moved tolocation (B).(B) consists of CLC-associated entities. The people there made CLC a fun, organized andclean place to work. Alas, all good things(C) is composed of persons past and present in Pieter’ s lab in Pharmacology. I drew uponthe hands and minds of people there many times (luckily it washed off) (double-punintended).(D) constitutes the people of the Lower Main Floor of the Cancer Agency. Thanks for thefriendly work environment, the air conditioning and the big lab bench (or two).Thanks, Pieter and Marcel, for introducing me to your individual ideologies of sciencethat have helped shape my perspective on the scientific lifestyle. So long and thanks forall the beer.Thanks also to the Leafs and the Canucks (in that order), Ebay and Sammy, Bob andDanielle, the Slow Fox, Latin Babe, Na Pali, Calvin and Hobbes, Led Zep, Black Tusk(x8), Gorgeous, the Blue Jays, Northern Pikes (and generally all fish of the world, exceptwalleye), Paul and Chris and?, Mamaliliculla, Gala Ball ‘90, the parties at 3235, pickledonions (not sausages), the blue Maverick, Ren and Stimpy, Saturday soccer, Tolkien,Doug’s library (Smartdrive), Hi-test, Renée, the cabin on Galiano, Dvorak (#9), Scruff,West Coast Trail, Empire, Robert Frost, the Saskatchewan Ursus americanus with a tastefor white fuel, the Micra (for all the repairs), STNG, my 2nd lower lumbar and both knees,Manning (or was that Maiming) Park, Jaggermeister, the Ballys, Hallowe’en on Bowen,Ivar, Mick’s 40th1, Marcel’s 40th (thanks in advance), Phillip, Wings (the sloop), “Life IsSweet”, my tent, Brian (the pilot), Brian (the trekker), the Green Hat, Cathy’s plants, JohnValjohn, Lipex, and a special thanks to Mark, Cory and Brynley for letting me be theoldest brother.xvDEDICATIONI’ve known since about Grade 11 that someday I’d be dedicating a Ph.D. thesis to myparents. In all that time I still haven’t thought of anything witty or profound which ismore suitable than a simple‘Thank You, Mum and Dad’.I suppose I should’ve taken more English courses.What I didn’t know was that I’d be co-dedicating this to a girl I first saw across acrowded dance class, resplendid in her full length red track pants.Thanks, Cat-Tea.I owe you two more toes.With LUV and MLV,Shane“I will accept as payment”Smiled Shylock the Jew,“Two pounds of flesh,Or a Thesis will do.”—The Merchant of Vancouver—ShanespearexviCHAPTER 1INTRODUCTION1.1 Targeted Drug Carriers: A PerspectiveDevelopment and application of target-specific phannaceuticals will significantlyalter the management and care of patients with life threatening diseases such as cancerand it is possible that targeted drugs would revolutionize the practice of health care. Theoriginal premise for developing targeted chemotherapy (“magic bullets”) was firstproposed by Paul Ehrlich in 1906 (Ehrlich, 1906). Almost 50 years later Pressman andKeighley (1948) described the use of radiolabeled polyclonal antibodies for imaging ofcancer, an observation that initiated four decades of research on the use of unique cellsurface markers to impart specificity to drugs. Applications of antibodies in thedevelopment of target-specific pharmaceuticals progressed significantly following thepivotal 1975 report of Köhler and Milstein describing the technique for monoclonalantibody production (Köhler and Milstein, 1975). The therapeutic potential of drugs andtoxins attached to monoclonal antibodies specific for antigens expressed on diseased cellshas been studied extensively for more than 20 years.To date no commercially successful pharmaceuticals have been developed on thebasis on antibody-directed targeting. This lack of success can be explained. Numerousstudies have shown both theoretically andlor experimentally how antibody properties(binding affinity and associated changes that occur following chemical modification of1antibodies) (Sung et al, 1992; Thomas et al, 1989), disease properties (target cell densityand localization, blood flow and vascular permeability to macromolecules) (Jam, 1990;Seymour et al, 1990; Dvorak et al, 1991; Blakey, 1992; Pietersz and McKenzie, 1992)and antigen properties (cellular location, intracellular recycling, degradation pathwaysand shedding) (Metezeau et al, 1984; Sung et al, 1992; Weinstein and van Osdol, 1992)can restrict antibody binding and localization in diseased sites. In addition, biologicalimmune responses to injected antibodies have resulted in: 1) limited circulation longevity(hence availability to disease site) (Rostaing-Capaillon and Casellas, 1990); 2) productionof human anti-mouse antibodies (HAMA) (Schroff et al, 1985; Lobuglio et al, 1988) and3) unexpected toxicities due to unforeseen reactions with normal cells (Schlom, 1986).Limitations of monoclonal antibodies, however, are being overcome. Generationof human and “humanized” (i.e. human Pc domain attached to murine F(ab’)2)monoclonals antibodies, for example, results in drug or toxin immunoconjugates that areless likely to illicit an immune response in humans (Schroff et al, 1985; Lobuglio et al,1988; Hale et a!, 1988). Further, specificity is being designed into the immunoconjugatesnot only through the choice of the target antigen but also in terms of the chemical linkertechnology that is employed to attach pharmaceutical agents to the antibody. Successfulapplication of “designer” antibodies is best illustrated by the recent studies fromHellström’ s laboratory (Hellström et al, 1990; Garrigues et a!, 1993; Trail et al, 1993) onthe use of a doxorubicin immunoconjugate for treatment of epithelial cell derived humancarcinomas. The antibody used (BR96) is specific for a tumor-associated antigen, relatedto Lewis Y antigen, that is expressed at a level of greater than 200,000 copies per cell2(Hellström et al, 1990). Further studies established that the antibody, once bound to thetarget cancer cells, is rapidly internalized (Garrigues et al, 1993). This property wasutilized in the design of the drug conjugate, where the anticancer drug was attached to theantibody through an acid sensitive chemical linker (Braslawsky et al, 1990). Theresulting conjugate exhibited remarkable antitumor activity, curing greater than 70% ofmice bearing human xenographs. This antibody-linked anticancer drug is now beingtested clinically in several Phase I and Phase Jill clinical studies using a humanizedversion of the BR96 antibody.As indicated above, antibodies can provide specificity to selected drugs, however,as drug carriers these proteins are severely limited by the quantity of drug that can beattached through chemical modification (Ghose et al, 1985; Ghose and Blair, 1987).Other drug carrier technologies (summarized in Table 1.1) can provide well definedsources of drugs that, in combination with monoclonal antibodies, have the potential ofdeveloping into therapeutically useful target-specific pharmaceuticals. The focus of thisthesis concerns the development of liposome drug carrier systems. Liposomes aremicroscopic spheres with an aqueous core surrounded by an outer shell consisting oflipids arranged in a bilayer configuration. The potential use of liposomes as drug carrierswas recognized over 25 years ago (Sessa and Weissmann, 1968) and since that timeliposomes have been used in a broad range of pharmaceutical applications (see Table1.2). For example, untargeted liposomes are used experimentally and clinically:1. For delivery of anticancer drugs (Rahman et al, 1990; Pestalozzi et al,1992; Cowens et al, 1993; Money-Kyrle, 1993), antimicrobial agents3TABLE 1.1Comparison of Different Drug Carriers Used for Systemic DeliveryDrug Carrier System Advantages DisadvantagesAntibodies Highly specific targeting; Low drug-carrying capacity;relatively low Mr immunogenicity; antigenicincreases access to heterogeneity; nonextravascular sites specific immune reactionsPolymer-based Improved drug stability; low Heterogeneous size; RESimmunogenicity; uptake; limited access tocontrolled release; extravascular sites; notrelatively high drug- amenable tocarrying capacity pharmaceuticalmanufacturing; oftenrequires chemicalmodification of drugLiposomes Improved drug stability; high RES uptake; limited accessdrug-carrying capacity; to extravascular sitescontrolled release;biodegradability; lowimmunogenicity;amenable topharmaceuticalmanufacturing4TABLE 1.2Liposome-Based Drug Formulations Undergoing Clinical Evaluation in HumansTherapeutic Agent Disease Treated Anticipated Liposomal BenefitAmphotericin B Systemic fungal Significantly reduced nephrotoxicity;and Nystatin infections enhanced efficacyDoxorubicin Cancer Reduced cardiotoxicity and alopecia;enhanced efficacyMuramyltripeptide Cancer- Targets to macrophagesimmunotherapyGentamicin Gram-negative Reduced nephrotoxicity; enhancedinfections efficacySalbutamol aerosol Asthma Reduced tachycardia; prolonged release“Indium Tumor imaging Preferential accumulation in tumorDNA (HLA antigen) Cancer Clinically acceptable gene transferimmunotherapy technologyMitoxantrone Cancer Reduced acute toxicity; reducedmyelosuppressionSubunit peptide Malaria Well tolerated adjuvantvaccine5(Lopez-Berestein, 1987; Swenson et al, 1990; Janoff, 1992; Vincent eta!, 1992; Alving and Swartz, 1984; Duzgunes et al, 1988), DNA andRNA (Rose et al, 1991; Hyde et al, 1993; Singhal and Huang, 1994),and antisense molecules (Bennett et al, 1992; Leserman et al, 1994);2. As adjuvants to elicit humoral (Gregoriadis, 1990; Alving, 1991; Frieset al, 1992) or cytotoxic T lymphocyte responses (Collins et al, 1992);and3. As diagnostic imaging agents (Oku et al, 1993; Ogihara-Umeda et al,1994).Clearly all these applications would benefit from strategies designed to target theliposomal carrier more effectively to the disease site.The use of liposomes as carriers of anticancer drugs has advanced significantly(see Table 1.3). For example, liposomes loaded with the antineoplastic agent doxorubicinhave demonstrated enhanced efficacy and altered in vivo tissue distribution leading todecreased acute and chronic toxic side effects (e.g. cardiotoxicity, myelosuppression, hairloss (alopecia) and ulceration of the mouth and throat (mucositis, stomatitis) (Balazsovitset al, 1989; Mayer et al, 1990; Gabizon et al, 1985; Rabman et al, 1990; Cowens et a!,1993)). Similarly, vincristine encapsulated in liposomes appears to be less toxic andmore efficacious than free vincristine (Mayer et al, 1990c; Boman et al, 1994). Thesenon-targeted liposomal pharmaceuticals represent the first generation of liposome-baseddrug delivery systems.6TABLE 1.3Progression Of Liposome Drug Carrier TechnologyTYPE DESCRIPTION UTILITY STAGES OFDEVELOPMENTFirst generation Liposomes prepared Reduced toxicity, Phase 1,11 studiesof natural and enhanced activity, completed, Phase ifisynthetic accumulation within initiated, productsphospholipids and an region of disease anticipated in 1995encapsulated drugSecond generation Conventional Improved access to Preclinical testing,liposome that region of disease, Phase I clinical trialsincorporate lipids that reduced interaction initiatedengender long with phagocytic cellscirculation lifetimes of the RESThird generation Conventional Improved therapeutic Experimentalliposome with lipids index, target cellthat engender long specific deliverycirculation lifetimeand surface-associated targetinginformation7A significant advance in the development of these first generation liposomal drugscomes from the use of specialized lipids, such as the monosialoganglioside GM1 orpolyethylene glycol-modified phosphatidylethanolamine (PEG-PE), that engender longcirculation lifetimes when incorporated in liposomes (Allen and Chonn, 1987; Klibanovet al, 1990; Blume and Cevc, 1990). It has been demonstrated that increased circulationlifetimes enhance the opportunity for liposomes, administered systemically, to leave thevascular compartment and enter certain extravascular regions (Gabizon andPapahadjopoulos, 1988; Bakker-Woudenberg et al, 1992; Wu et al, 1993). Tumors, forexample, exhibit leaky blood vessels that have a reduced tendency to retain circulatingmacromolecules (Gerlowski and Jam, 1986; Dvorak et al, 1988; Nagy et al, 1989; Kohnet al, 1992). Liposomes can extravasate in these regions, thus leading to preferentialaccumulation within tumors. Studies have now clearly demonstrated that liposomescontaining PEG-PE accumulate within these sites preferentially over conventionalliposomes (Gabizon and Papahadjopoulos, 1988; Papahadjopoulos et al, 1991; Wu et al,1993). The mechanism(s) responsible for PEG-PE-induced increases in liposomecirculation lifetimes will be discussed in detail within this chapter (see Section 1.3).Application of these lipids in liposome-based drug carrier systems has resulted in thedevelopment of second generation liposomal pharmaceuticals, some of which are beingtested extensively for the delivery of anticancer drugs (Maruyama et al, 1994; Working etal, 1994).It is envisioned that the third generation of liposomal pharmaceuticals will consistof drug-loaded liposomes with surface-associated targeting information. Site-directing8targeting ligands, such as monoclonal antibodies, can be attached to liposomes by eithercovalent or non-covalent methods (reviewed in Leserman and Machy, 1987). Asillustrated in a recent study from the laboratory of Allen (Ahmad et al, 1993), atherapeutically improved liposomal drug can be achieved through the use of antibody-based targeting. This study used antibody-coated, doxorubicin-loaded liposomal carriersthat specifically recognize an antigen expressed on mammalian squamous carcinoma(Samuel et al, 1989). In a murine lung cancer model, it was shown that this formulationexhibited remarkable antitumor activity, resulting in cures in over 50% of the treatedanimals (Ahmad et a!, 1993).Studies have shown that antibody-coated and other protein-coated liposomes areremoved from the circulation more rapidly than protein-free liposomes (Aragnol andLeserman, 1986; Longman et al, 1994, and Chapter 3 of this thesis). The ability of thesecarriers to escape the vascular compartment and reach an extravascular target is thereforecompromised. This problem has been partly overcome through incorporation of PEG-modified lipids (Klibanov et al, 1990; Blume and Cevc, 1990; Allen et al, 1991; also seeChapter 4 of this thesis), however, the presence of these lipids can cause a significantreduction in ligand-mediated binding of liposomes to target antigens (Klibanov et al,1991; see Chapter 4). Therefore, in terms of developing target-specific liposomalpharmaceuticals, there are significant hurdles that must be overcome. Studies describedin this thesis critically evaluate issues related to how targeting can be achieved (Chapter3), the design of protein-coated carriers that can access extravascular sites (Chapter 4),9and a model approach that can be used to promote liposome delivery to a target residingin an extravascular site (Chapter 5).It is important at this point in the thesis to review technologies related topreparation and characterization of targeted and non-targeted liposomal drug carriers.This introduction will also summarize how the physical and chemical characteristics ofliposomes influence the in vivo behavior of liposomal drug carriers. In particular, factorsthat are known to play a role in the ability of liposomal drug carriers to access a target cellpopulation will be discussed.1.2 LiposomesThe structures of aqueous dispersions of lipids, or liposomes, was firstdocumented by A. Bangham (Bangham et al, 1965). He reported that the lipids adopt abilayer configuration resulting in spherical, onion skinned structures that contain aqueouscompartments separated by bilayer lipid membranes. The bilayer structure arose as aconsequence of the amphipathic character of membrane lipids that exhibit a hydrophilic“water loving” portion as well as a hydrophobic “water fearing” portion. The hydrophilicregions tend to orient toward the aqueous phase while the hydrophobic component of themolecules orient themselves away from the water (see Figure 1.1). The resulting closedmembrane structures, referred to as liposomes (Sessa and Weissmann, 1968), areextremely useful as model membrane systems to assess the structure and functional rolesof lipids in isolation or in well defined mixed lipid systems. These studies have helped to10Figure 1.1Amphipathic Lipids In Bilayer ConfigurationHydrophilicHydrophobicAqueous bufferMembrane lipid isolatedfrom natural sourceBilayer structure11elucidate the role of membrane lipids in providing a permeability barrier (Fettiplace andHaydon, 1980; Deamer and Bramhall, 1986), as mediators of membrane fusion (Cullis etal, 1986; Wilschut and Hoekstra, 1986) and as specific triggers of biological responsessuch as complement activation (Alving and Richards, 1983; Muller-Eberhard, 1988). Inaddition, liposomes have played a critical role in characterizing how membrane proteinsinteract with lipids (Jost and Griffith, 1982) and how these proteins function in isolation(Racker, 1973; Vik and Capaldi, 1977).It is important to emphasize that a critical understanding of the chemicalcomponents of liposomes (incorporated lipids) and the structure they assume is necessaryin order to develop optimized “designer” liposomes for targeted drug delivery. Therefore,this section will provide a brief summary of the types of lipids that can be used to formliposomes. This will be followed by a description of methodologies used to prepareliposomes, including a summary of the different types of liposomes that can be obtained.Finally, techniques used to attach targeting ligands to the surface of liposomes will bediscussed.121.2.1 Chemical and Physical Properties of Lipids Commonly Used in Liposomal DrugCarriers1.2.1.1 PhospholipidsLiposomes are usually composed of phospholipids. As indicated in Figure 1.2,these lipids can exhibit different headgroups (hydrophilic regions) and different acylchain compositions (hydrophobic regions) which dictate the physical properties of thelipid (for reviews, see Chapman, 1975; Seelig, 1978; Cullis and de Kruijff, 1979).Increased unsaturation of the acyl chains decreases the order in the hydrocarbon matrix,which has been correlated to increased membrane permeability (Demel et al, 1972;Papahadjopoulos et al, 1973) and decrease in the gel to liquid-crystalline transitiontemperature (Tc) (Chapman, 1968; Chapman, 1975) (see Table 1.4).The presence of negatively charged phospholipids will control the membranesurface charge of the resulting liposome. At physiological pH, liposomes prepared ofphosphatidylserine, phosphatidylglycerol, phosphatidylinositol and phosphadic acid willbe negatively charged. This charge influences binding of mono- and divalent cations (e.g.Ca2, Mg2) (Papahadjopoulos, 1978; Moghimi and Patel, 1993) and is an importantfactor governing the binding of exogenous proteins (de Kruijff et al, 1984) and nonspecific binding of liposomes to cell membranes (Juliano and Layton, 1980). These lattertwo properties will be discussed in greater detail in Sections 1.3.13Figure1.2GeneralStructureOfAPhospholipidShowingCommonlyOccurringHeadgroupsAndFattyAcidMoietiesIo=—o0CH2—H—CHIIc=oc=oCH21H2CH2CH2CH2CH2CH2CH2CH2CH2CH2CH2CH2CH2Cl2CHIIICH2CHCH2CH2CH2CH2Cl2CH2CH2cjH2CH2CH2CH2CH2CH2CH2Cl3CH3Structures-CH2CHN(CH3) -CH2CHN’H3 -CH2CH—NH3Coo-CH2CH(OH)CHOHHOH——HeadgroupGlycerolbackbone—Acylchain—NeutralphospholipidsCholine(Phosphatidyicholine)Ethanolamine(Phosphatidylethanolamine)NegativephospholipidsSerine(Phosphatidylserine)Glycerol(Phosphatidylglycerol)Inositol—(Phosphatidylinositol)SaturatedfattyacidsLauricMyristicPalmiticStearicArachidicLignocericUnsaturatedfattyacidsPalmitoleicOleicLinoleicLinolenicHHCH3(CH2)10C00HCH3(CH2)12C00HCH3(CH2)14COOHCH3(CH2)16C00HCH3(CH2)18C00HCH3(CH2)2C00HCH3(CH2)5CH=CH(CH7COOHCH3(CH2)7CH=CH(CHCOOHCH3(CH2)5CH=CHCHCH=CH(CH7OOHCH32CH=CHCHCH=CHCHCH=CH(CH)7OOHCH3(CH2)4(CH=CHCHCH=CH(CHOOHArachidonicTABLE 1.4Effect of Acyl Chain Length, Degree of Saturation and Head Group on PhaseTransition TemperatureLipid Species (Acyl Chains) Transition Temperature (±2°C)dilauroyl PC (12:0, 12:0) -1dimyristoyl PC (14:0, 14:0) 24dipalmitoyl PC (16:0, 16:0) 41distearoyl PC (18:0, 18:0) 55stearoyl,oleoylPC(18:0, 18:1) 6stearoyl, linoleoyl PC (18:0, 18:2) -16stearoyl, linolenoyl PC (18:0, 18:3) -13stearoyl, arachidonyl PC (18:0, 20:4) -13dipalmitoyl PE (16:0, 16:0) 63dipalmitoyl PS (16:0, 16:0) 55dipalmitoyl PG (16:0, 16:0) 41dipalmitoyl PA (16:0, 16:0) 6715In addition to defining surface charge, the headgroup of the phospholipid can playa critical role in controlling whether an isolated phospholipid can assume a bilayerconfiguration after hydration. It is known, for example, that phosphatidylethanolamineswith two unsaturated acyl chains will not adopt a bilayer structure (Cullis and de Kruijff,1979). Rather, these lipids assume an alternate structure following hydration that isreferred to as the hexagonal (H11) phase (see Figure 1.3). This phase is characterized bycylinders of lipids in an “inverted” (headgroups oriented inwardly towards aqueouschannels) organization. Depending on physiological conditions such as temperature, pHand the presence of divalent cations, other phospholipids can also adopt non-bilayerstructures (for reviews, see Cullis and de Kruijff, 1979; Cullis et al, 1983). In thepresence of low pH, for example, liposomes composed of phosphatidylserine (withunsaturated acyl chains) will adopt the H11 phase (Hope and Cullis, 1980). The role oflipids that can assume non-bilayer structures in biological membranes has primarily beenbased on regulating membrane fusion (Kachar and Reese, 1982; de Kruijff et al, 1984). CholesterolIn addition to phospholipids, another commonly employed lipid used in thepreparation of liposomes is cholesterol. Cholesterol is the major neutral lipid componentin the plasma membrane of eukaryotic cells. This steroid also possess amphipathiccharacteristics due to the presence of the polar 3-13-hydroxyl group at one end of themolecule. Within a phospholipid bilayer, this hydroxyl group will become oriented such16Figure 1.3Polymorphic Phases Available to Lipids on HydrationBilayer4.— SSSI/Hexagonal (H0)17that it is adjacent to the fatty acyl carbonyls of the phospholipid (Huang, 1977). The rigidsteroid nucleus is associated with the acyl chains. The consequences of incorporatingcholesterol in a phospholipid bilayer include: 1) a condensing effect where the areaoccupied per phospholipid molecule is reduced (Rance et al, 1982; Hyslop et al, 1990); 2)a decrease in the enthalpy of the gel to liquid-crystalline transition (Ladbroke et al, 1968;Hubbel and McConnel, 1971); and 3) decreasing membrane permeability (Demel and deKruijff, 1976). In addition, it is clear that incorporation of cholesterol helps to stabilizeliposomes used for systemic delivery of drugs (Papahadjopoulos et al, 1973; Weinstein etal, 1981). This may be a consequence of the effects of cholesterol on the phospholipidbilayer, since it is well established that the presence of cholesterol at levels greater than30 mol% decrease binding of serum proteins that play a role in opsonizing liposomes(Finkeistein and Weissman, 1979; Kirby et al, 1980; Weinstein et al, 1981; also seeSection 1.3.1).1.2.2 Liposome PreparationIn order to form a liposome, the component lipids must be able to adopt a bilayerstructure on hydration. The characteristics of the resulting liposome, such as size, charge,aqueous trapped volume and lamellarity depend on the lipids used as well as thepreparation technique. In general liposomes are classified in three categories(summarized in Figure 1.4): multilamellar liposomes (MLVs), large unilamellarliposomes (LUVs) and small unilamellar liposomes (SUVs). These classifications are18Figure 1.4Multilamellar and Unilamellar VesiclesEgg phosphatidyicholine multilamellar vesicles (MLV), large unilamellar vesicles (LUV) andsmall unilamellar vesicles (SUV) visualized by freeze-fracture electron microscopy. The freeze-fracture technique relies on the fracture plane proceeding through the hydrophobic interfacebetween bilayer leaflets in a frozen sample which contains membrane structures. Once fracturedthe sample is coated with platinum and carbon, the sample itself is digested away and theremaining replica is examined using an electron microscope. The bar represents 200 nm.M LV S IUV’s BUVsDiameter: i —10 pm 005—1 pm 002 — 005 pm19primarily assigned on the basis of size and lamellarity (number of bilayers present withina liposome) characteristics that can be readily assessed by a number of physicaltechniques (NMR, X-ray crystallography and electron microscopy). The technique offreeze-fracture electron microscopy (described in the Figure 1.4 legend), however,provides a graphic illustration of liposome structure. The following sections describehow these different categories of liposomes can be prepared as well as a brief summary ofrelevant characteristics of the resulting liposomes. Multilamellar Vesicles (MLVs)MLVs can be formed simply by mechanically dispersing lipid, as a powder or drylipid film, in an aqueous solution (Bangham et al, 1965). The resulting MLVs have anappearance similar to an onion, in that they consist of numerous concentric phospholipidbilayers (see Figure 1.4). MLV systems are usually heterogeneous with respect tolamellarity and size, typically exhibiting diameters ranging from 1 to 10 jim with as littleas 5% of the associated lipids comprising the outermost lamellae (Mayer et al, 1985;Hope et al, 1986). The amount of aqueous volume enclosed within a MLV composed ofneutral lipids, such as egg phosphatidylcholine (EPC), is relatively low (—0.5 jil/jimollipid) due to close packing of the bilayers (Mayer et al, 1985). This volume, typicallyreferred to as trapped volume, can be significantly increased (up to 7 jil/jimol) throughthe incorporation of charged lipids (such as phosphatidylserine) (Hope et al, 1986), thatpromote electrostatic repulsion between lamellae, resulting in swelling and enlarged20interlamellar spaces. The trapped volumes of MLVs can also be increased by successivefreeze-thaw cycles involving freezing MLVs in liquid nitrogen followed by thawing inwarm water (FATMLVs) (Mayer et al, 1985). This leads to formation of structures thatappear by freeze-fracture microscopy to consist of vesicles within vesicles and vesiclesbetween lamellae. In addition, there is a significant increase in interlamellar spacing,resulting in a large increase in trapped volumes. These systems can encapsulate up to 10jil/jimol lipid.An alternate technique to generate MLVs, referred to as ‘reverse phase’procedures, involves hydration of lipids from an organic solvent (Szoka andPapahadjopoulos, 1978; Szoka and Papahadjopoulos, 1980; Gruner et al, 1985). Briefly,as the organic solvent is diluted or evaporated in the presence of an aqueous buffer, thelipids move from the organic phase to the aqueous phase, forming vesicles in the process.While the resulting MLVs exhibit high trapped volumes (up to 10 jil4tmol lipid), thisprocedure is limited by lipid solubility in the organic phase and difficulties in removingresidual organic solvent from the final liposome preparation (Gruner et al, 1985). Unilamellar Vesicles (SUV and LUV)As illustrated in Figure 1.4, unilamellar vesicles can be small (25-50 nm indiameter) or large (50-200 nm in diameter). SUVs can be generated directly from MLVsby sonication (Huang, 1969), French press techniques (Barenholz et al, 1979) orhomogenization techniques (Hirota and Kibuchi, 1985). The size of the resulting21liposome is the limit size, in that the area occupied by the lipids in the inner monolayer ofthe bilayer cannot be physically accommodated into a smaller radius. Further processingof these liposomes results in no further reduction in size. Vesicles of this size typicallyexhibit a 2:1 excess of lipid in the outer monolayer compared with the inner monolayerand have extremely low trapped volumes (—0.2 .tl I imol lipid) (Huang, 1969; Barenholzet al, 1979). The acute radius of curvature of these vesicles can create instability,resulting in spontaneous fusion to form larger structures (Parente and Lentz, 1984), andpoor retention characteristics. For these reasons LUVs are typically more useful thanSUVs for studies relying on liposomes as model membrane systems or as drug carriers.A gentle procedure for generating LUVs, detergent dialysis (Enoch andStrittmatter, 1979; Mimins et al, 1981), involves solubilizing dried lipid or pre-formedvesicles in an aqueous buffer that contains detergents. As the detergent is removed bydialysis, the micelles coalesce and the phospholipids adopt a bilayer configuration. Thephysical characteristics (size, trapped volumes, trapping efficiencies) of the resultingvesicles are dependent on the type of detergent employed, the rate and method ofdetergent removal, as well as the type of lipid mixture being used (Madden, 1986; Allen,1984). While this technique is valuable for reconstitution of integral membrane proteinsin model membrane systems (Kagawa and Packer, 1971; Helenius et al, 1977), lowtrapping efficiencies and difficulties in removing residual detergent from the lipid bilayerlimit the usefulness of this technique to produce drug carrier systems.A convenient method for preparation of LUVs was initially developed in thelaboratory of Papahadjopoulos (Szoka and Papahadjopoulos, 1978; Szoka and22Papahadjopoulos, 1980). This procedure consisted of forcing MLVs, prepared by reversephase evaporation procedure (see Section 1.2.2), sequentially through polycarbonatefilters exhibiting decreasing pore sizes from 1.0 urn to 0.2 jim. This technique has beenoptimized (Hope et al, 1985; Mayer et al, 1986b; Nayar et al, 1989), such thathomogeneous liposomes exhibiting well defined mean diameters (equivalent to the poresize of the filter used) can be prepared at lipid concentrations as high as 400 mg/mi, usingany lipid composition that adopts a bilayer configuration upon hydration. Whenemploying lipid compositions consisting of phospholipids with saturated acyl chainsequal to or greater than 16 carbons in length, it is necessary to extrude the MLVprecursors at a temperature above the Tc of the phospholipid (Nayar et al, 1989). Further,the extrusion must be done repeatedly (> 5 times) through stacked membranes in order toachieve a uniform size distribution. Liposomes prepared in this fashion will largely beunilamellar provided the pore size of the filter used is equal to or less than 200 nm. Thisextrusion technique has significant advantages over previous techniques since it is simple,rapid and reproducible. Further, LUVs can be prepared with a wide range of lipidcompositions, in the absence of added organic solvents and detergents. The resultingliposomes prepared through 100 nrn filters can have aqueous trapped volumes between 1and 3 jii per j.tmol lipid and, depending on lipid concentration, can result in aqueoustrapping efficiencies of over 80% (Mayer et al, 1986). Finally, this extrusion technique isamenable to pharmaceutical production. It is readily scaled up to volumes in excess of 1L and can be used under conditions that result in sterile, pyrogen-free liposomes.231.2.3 Drug EncapsulationTwo techniques are available to encapsulate drugs in liposomes: passive trappingduring liposome formation and active trapping after the liposomes have been formed.Regarding the first technique, passive entrapment of hydrophobic drugs is governed bydrug-lipid interactions, while passive entrapment of hydrophilic drugs is related to theaqueous trapped volume and lipid concentration of the liposomes during vesicleformation. The second encapsulation technique, active entrapment, relies on the ability ofcertain amphipathic drugs to redistribute across the lipid bilayer in response to atransmembrane ion gradient. The differences between these techniques are illustrated inFigure Passive Trapping TechniquesHydrophobic drugs, such as amphotericin B, can be directly incorporated intoliposomes during vesicle formation (Janoff, 1992; Lopez-Berestein et al, 1983) (seeFigure 1.5). These drugs are mixed homogeneously with the lipids in organic solventsprior to preparation of the dried lipid film. Trapping efficiencies of 100% are achievable,but this is dependent on the packing constraints imposed by lipids in the membrane.Therefore, lipid composition, the type of liposome employed and the nature of thehydrophobic drug to be loaded (Hopfer et al, 1984; Lopez-Berestein and Juliano, 1987)will all have an impact on the resulting drug-liposome complex. It should be noted that24Figure 1.5Illustration Of Different Drug Encapsulation Protocols With An Indication As To DrugDistributionTypePassive encapsulation ofhydrophobic drugPassive encapsulation ofhydrophilic drugActive encapsulation ofdrugs exhibiting aprotonizable aminefunctionAdded to solvents used tosolubilize lipids or added toaqueous bufferAdded to buffer used tohydrate lipidAdded to pre-formedliposomesMethod DistributionAp+ pH gradient25drugs associated with liposomes using these techniques can be exchangeable into othermembranes such as the plasma membrane of cells or into other lipid-rich bodies such aslipoproteins (Krupp et al, 1976).Passive encapsulation of water-soluble drugs relies on the ability of liposomes totrap a volume of aqueous buffer, with a dissolved drug, during vesicle formation.Trapping efficiencies are limited by the characteristics of the liposomes (see Table 1.5)and drug solubility. For example, the low trapped volume of SUVs (0.2 jil/iimol lipid)leads to trapping efficiencies of 1% or less (Szoka and Papahadjopoulos, 1978), while thehigher trapped volumes of FATMLVs and LUVs (1-10 il/imol lipid) prepared at lipidconcentrations up to 400 mg lipid/mi can lead to trapping efficiencies as high as 88%(Mayer et al, 1985). The retention of a passively trapped drug will depend upon drugmembrane permeability, which depends on the drug itself and the lipid composition of theliposome. For example, impermeable drugs such as methotrexate and cytosinearabinoside have retention half-times of 50 and 18 hr, respectively, in 100 nm liposomescomposed of EPC, while drugs such as doxorubicin have retention half-times of 1 hr inidentical vesicles (Bally et al, 1988). The development of ‘active’ trapping procedures,described below, enabled stable and efficient encapsulation of many of these permeabledrugs. Active Trapping TechniquesA relatively new drug trapping technique, referred to as active trapping, (Mayer etal, 1986; Madden et al, 1990; Mayer et al, 1993) is based on loading drugs into pre26TABLE 1.5Passive Trapping Properties of Liposomal PreparationsType Preparation Vesicle Entrapped agent Trapping Referencesprocedure diameter efficiency(aim) (%)aSUV Sonication 0.025-0.040 Cytosine arabinoside, 1-5 Szoka andmethotrexate, Papahadjopoulos,carboxyfluorescein 1980SUV French 0.020-0.050 Carboxyfluorescein, 5-25 Lelkes, 1984press inulin, trypsinLUV Detergent 0.1-10.0 Carboxyfluorescein, 12-42 Mimms et al,removal inulin, cytochrome c 1981; Weder andZumbuehl, 1984LUV Reverse 0.2-1.0 Cytosine arabinoside, 28-45 Szoka andphase carboxyfluorescein, Papahadjopoulos,DNA, insulin 1980LUV Extrusion 0.056-0.2 Cytosine arabinoside, 15-60 Hope et al, 1985;methotrexate, inulin, Mayer et al, 1986bDNA, 22NaMLV Mechanical 0.4-3.5 22Na, DNA, 1-8.5 Kirby andmixing carboxyfluorescein, Gregoriadis, 1984MLV Freeze- 0.5-5.0 22Na, inulin 35-88 Mayer et al, 1985thawMLV Sonicate- 0.3-2.0 Carboxyfluorescein, 27-54 Ohsawa et al, 1985dehydrate- ATP, vincristine,rehydrate meiphalan, factor ifiMLV Solvent 0.3-2.0 Streptomycin sulfate, 6.3-38 Gruner et al, 1985evaporation chioramphenicol,inulin,aTrapping efficiencies dependent on lipid concentration and lipid type, see references.27fonned liposomes that exhibit a transmembrane pH gradient (see Table 1.6). Lipophilicdrugs that have protonizable amine functions will redistribute across the membrane inresponse to the pH gradient (see Figure 1.6). The unprotonated (neutral) species of thedrug crosses the membrane and accumulates in the vesicle interior in the protonated form,until equilibrium levels corresponding to [AH}r/[AFf’]ouT = [H]N/[H]ou-r are reached,where M1 indicates the protonated form of the drug. Thus for a pH gradient of threeunits, interior concentrations of drugs 103-fold higher than exterior concentrations areachievable. This technique also allows trapping efficiencies that approach 100% (seeTable 1.6). It has been shown that the presence of a pH gradient can also decrease therate of drug efflux from the vesicles as much as 30-fold (Mayer et al, 1986). Since thisentrapment procedure does not depend on any specific drug-lipid interaction, thistechnique can be used with virtually any liposome formulation capable of maintaining astable transmembrane pH gradient (Mayer et al, 1993).1.2.4 Coupling of Targeting Ligands to LiposomesAs indicated in Section 1.1, active targeting of liposomes to specific cell types invivo can be mediated by the use of site-specific ligands, such as monoclonal antibodies,that selectively interact with target cell-surface molecules. Various targeting ligands,either covalently or non-covalently associated with the surface of liposomes, have beenused in liposome targeting. In addition to antibodies (Huang et al, 1980; Martin et al,1981; Wolff and Gregoriadis, 1984), glycolipids and proteins that have defined cell28TABLE 1.6Uptake and Retention of Lipophilic Amine Drugs that Accumulate Inside Liposomesin Response to Transmembrane pH GradientsDRUG CLASS Uptake 15 mm Uptake 2 hr(nmol/pmol lipid) (nmol4imol lipid)Daunorubicin Antineoplastic 200 204Doxorubicin Antineoplastic 202 203Epirubicin Antineoplastic 201 200Mitoxanthrone Antineoplastic 200 198Vinbiastine Antineoplastic 175a 127Vincristine Antineoplastic 178 130Chiorpromazine Local anaesthetic 98 96Dibucaine Local anaesthetic 194 176Lidocaine Local anaesthetic 87 87Quinidine Antiarrhythnic agent 203 74Dopaniine Biogenic amine 190b 177Seratonin Biogenic amine 80’ 78Imipramine Antidepressant 182 188Diphenhydraniine Antihistamine 176a 87Quinine Antimalarial 148a 81Chioroquine Antimalarial 104a 88aMimum uptake taken at 5 mm.bMimum uptake taken at 30 mm.cMaxmium uptake taken at 90 mm.From Madden, T.D. et al., Chem. Phys. Lipids, 53, 37-46, 1990. With permission.29Figure 1.6Schematic Representation of Equilibria Involved in Redistribution of Lipophiic AmineDrugs Across Bilayer Membranes in Response to Transmembrane pH GradientsOUTSDEpH 7.5INSIDEpH 4.0AHIThe concentration of the neutral (unprotonated) form of the drug is denoted as [A], theconcentration of the positively charged (protonated) form of the drug is denoted as [AHj, andthe concentration of hydrogen ions is denoted as [Hj. The association constant (Ka) for the drugcan be described as:Ka = [Hi [A] / [AHiAssuming that the Ka of the drug is the same in the intra- and extravesicular aqueouscompartment:Ka = [Hiin[A]in I [AHiin = [Ft4iout[A]out I [AHioutThe neutral form of lipophilic drugs that contain titratable amine moieties can readily equilibrateacross membranes (Rottenberg, 1979). Thus the above relationship may be simplified to:[AHj / [AHi00 = [WIn / [H]0This relationship predicts a drug concentration gradient of 1000 for a unilamellar vesicleexhibiting a transmembrane pH gradient of three units (inside acidic).AH+30surface receptors have been used (see Table 1.7). An example of the latter is transferrin,which can be attached to liposomes and subsequently used to label cells exhibiting thetransferrin receptor (Stavridis et al, 1986; Vidal et al, 1987). By far the most commonligands employed, however, are monoclonal antibodies, and therefore the followingsection will focus on methods used for the attachment of antibodies to liposomes. It isimportant to note that the techniques described are useful in attaching a variety of protein-based targeting ligands. Covalently Coupled AntibodiesThere are two basic strategies for the covalent coupling of antibodies to liposomes(see Table 1.8) (for reviews, see Heath and Martin, 1986; Leserman and Machy, 1987).In the first, antibodies are covalently modified with a hydrophobic lipid anchor, usually aderivative of the phospholipid phosphatidylethanolamine (Sinha and Karush, 1979;Jansons and Mallet, 1981; Mon and Huang, 1993) or palmitic acid (Huang et al, 1980).The resulting amphipathic antibody-conjugated lipids are then incorporated into theliposome bilayer. Detergent dialysis is used instead of organic solvents to introduce theantibody-lipid complex into liposomes in order to limit protein denaturation (Lesermanand Machy, 1987). The advantages of this strategy are that the technique involves noharsh treatment of the protein, and the lipophilic nature of the antibody-lipid complexresults in efficient incorporation of the ligand into the liposomes (85-90%, Huang et al,1980). The disadvantages include the presence of residual detergent which can result in31TABLE 1.7Alternatives to Antibody-Based Targeting Ligands that have been Used to TargetLiposomesLigand Associated Target Cell ReferenceWith LiposomeGalactose-terminated HeLa Cells Bussian andLigands Wriston, 1977Hepatocytes van Berkel et al,1993Mannose-terminated Macrophages Barratt andLigands (alveolar, peritoneal, Schuber, 1993spleen, Kupffer)Fibronectin Peritoneal Macrophages Hsu and Juliano,1982Transferrin Lymphocytes Vidal et al, 198732TABLE1.8SummaryofCovalentCouplingTechniquesStrategyCouplingAgentTypeofProteinCoupledReference(ReactiveGroup)ProteincoupledtoliposomeN-(5-dimethylamino-Fab’fragmentSinhaandKarush,1979componentbeforeliposomenapthylene-1-sulfonyl)-L-(suithydrylgroups)formationlysinecoupledtoPEN-hydroxysuccinimide(NHS)ImmunoglobulinsHuangetal,1980esterofpalmiticacid(aminogroups)CarbodiimidederivatizedPECitraconylatedF(ab’)2JansonsandMallet,1981(carboxyl groups)ProteincoupledtoliposomeCarbodiimidederivatizedPEImmunoglobulinsEndol etal,1981componentafterliposome(carboxyl groups)formationGlutaraldehydeordimethyl-ImmunoglobulinsTorchilinetal,1979;suberimidatederivatizedPE(aminogroups)Torchilinetal,1980LactosylceramideoxidizedbyF(ab’)2Heathetal,1981periodate(aminogroups)SPDP-modifiedPESPDP-modifiedantibodiesLesermanetal,1980Fab’Martinetal, 1981(sulthydrylgroups)SMPB-modifiedPBSPDP-modifiedantibodiesMatthayetal,1984SATA-modifiedantibodiesDerksenandScherphof,1985Fab’MartinandPapahadjopoulos,(sulfhydrylgroups)1982increased vesicle permeability (Leserman and Machy, 1987), the loss of bioactivity ofsome antibodies due to their orientation toward the interior of the liposome (Lesermanand Machy, 1987), and the relatively low trapping efficiencies for water-soluble solutesinherent in detergent dialysis liposome preparation techniques (see section the second strategy, antibodies are coupled to pre-formed liposomes containinga lipid that has a functional group that can be used to couple (covalently) added proteins.Symmetric bifunctional cross-linking reagents, such as glutaraldehyde and suberimidate(Torchilin et al, 1978; Torchilin et al, 1979) and carbocliimide (Endoh et al, 1981), permitbond formation between the amino groups of the antibody and phosphatidylethanolamine(PE) incorporated in the liposome. Inefficient side reactions such as homopolymerizationof antibody or liposomes, however, can result in protein aggregation and intervesicularcrosslinking, respectively (Endoh et a!, 1981). Heterobifunctional cross-linking reagents,such as N-succinimidyl 3-(2-pyridyldithio) propionic acid (SPDP) (Carlsson et al, 1978;Leserman et al, 1980) and N-succinimidyl 4-(p-maleimidophenyl)butyrate (SMPB) (Martinand Papahadjopoulos, 1982; Derksen and Scherphof, 1985), minimize problems ofhomopolymerization. These reagents are usually used to first modify the amino groups ofPE in organic solvent prior to liposome formation. Liposomes expressing thiopyridine ormaleimidophenyl butyrate groups can then be used to couple proteins that have accessiblesulffiydryl groups. Sulthydryl groups on Fab’ fragments of IgG may be used (Martin andPapahadjopoulos, 1982), or alternately a free sulthydryl group may be added to nativeantibody via modifying agents such as N-succinimidyl S-acetylthioacetate (SATA)(Hashimoto et al, 1986), S-acetylmercaptosuccinic anhydride (SAMSA) (Leserman,341981) or SPDP (Carlsson et a!, 1978). The bond between the liposome and antibody hasbeen shown to be more stable using SMPB, which is a nonreversible alkylating agent(Martin and Papahadjopoulos, 1982), as compared to the disulfide-coupling reactionusing SPDP. Coupling of antibodies to liposomes via a thioester or thioether bond resultsin a minimal loss of immunoreactivity to the antibody (Heath et al, 1984).This second strategy has several advantages over the incorporation of antibody-conjugated lipids into liposomes. In particular, employing pre-formed liposomes allowsthe use of procedures that result in homogeneously sized liposomes (see section 1.4.2;Hope et a!, 1985) and efficient encapsulation of drugs (see section 1.6.2; Mayer et al,1986; Madden et al, 1990). Non-Covalently Associated LigandsProtein-liposome conjugates may be used as a versatile approach to attach site-directing ligands directly on the surface of liposomes. Protein-liposome conjugates areformed by covalently coupling proteins, such as avidin or protein A, to the liposomesurface using techniques described above (see Section Site-directing ligands,such as antibodies, are then non-covalently bound to the liposome surface via specificinteractions with the protein coupled to the liposomes. This non-covalent bond betweenthe antibody and protein conjugated to the liposome is stable both in vitro (Leserman etal, 1980, Leserman et al, 1980b) and in vivo (Aragnol and Leserman, 1986).Furthermore, the binding affinity of non-covalently bound antibodies is not reduced35(Leserman et al, 1980). The advantage of this technique is that a single protein-liposomepreparation can be used with many different antibodies, whereas direct covalent couplingof antibody to liposomes commits those liposomes to a single antigen target (Gray et al,1988). The best studied examples of this approach are biotinylated ligands bound tostreptavidin-coated liposomes (Rosenberg et al, 1987; Loughrey et al, 1993; Longman etal, 1994) and antibody bound to protein A (or protein G)-bearing liposomes (Leserman eta!, 1980; Machy and Leserman, 1984; Gray et al, 1988).Liposomes covalently conjugated with avidin exploit the high-affinity interactionbetween biotin and avidin (Green, 1975) in order to non-covalently bind biotinylatedligands (Urdal and Hakomori, 1980). However the basic nature of avidin (Green, 1975)results in significant background binding to target cell populations (Finn et al, 1980).Streptavidin, a non-glycosylated neutral protein secreted by Streptomyces avidinii, hassimilar strong (1014 M’) affinity for biotin (Chaiet and Wolf, 1964) and can be used as asubstitute (Rosenberg et al, 1987; Loughrey et al, 1990; Loughrey et al, 1993; Longman eta!, 1994). For example, Rosenberg et a! (1987) demonstrated specific in vitro binding ofstreptavidin liposomes conjugated with biotinylated nerve growth factor (NGF) tocultured rat and human tumor cells bearing NGF receptors.Liposomes conjugated with protein A non-covalently bind antibody via theinteraction between protein A and the Fc portion of certain IgG classes (Langone, 1978;Ey et a!, 1978). These protein A-bearing liposomes bind specifically in vitro to cells preincubated with cell-specific antibodies, but not to cells pre-incubated without antibody(Leserman et a!, 1980; Machy and Leserman, 1984; Gray et a!, 1988). However, protein36A-bearing liposomes are limited to using only antibodies as site-directing ligands. Itshould also be noted that protein G, another protein that can bind Fc portions for a broadrange of IgG subclasses, has been considered for the attachment of antibodies toliposomes (Leserman et al, 1993).An alternative approach to non-covalent attachment of targeting ligands is basedon the use of liposomes that bear the hapten dinitrophenyl (DNP), which can noncovalently bind anti-DNP antibodies (Uemura and Kinsky, 1972; Kinsky and Nicolotti,1977). DNP-bearing liposomes with attached anti-DNP antibodies were targeted to cellsthat were chemically modified with the DNP hapten (Weinstein et al, 1977). Thisbinding required both the hapten modification of the cells and presence of the specificbivalent antibody, and could be inhibited by hapten in the soluble form (Weinstein et al,1977).1.3 Factors Controlling The Pharmacodynamic Behavior Of Liposomal DrugsOptimization of liposomes for delivery of biologically active agents requires adetailed understanding of liposome behavior in vivo and liposome-mediated changes inassociated drug distribution. Altered drug biodistribution can then be correlated tospecific toxic side effects and therapeutic benefits, thereby establishing apharmacodynamic profile for the liposomal drug. It is argued here that a greaterunderstanding of the specific mechanisms responsible for the therapeutic activity of firstand second generation liposomes (see Section 1.1) is essential if improved target-specific37carriers are to be designed. This section will review biological factors known to play arole in liposome stability and biodistribution following systemic administration. Inparticular it is well established that serum proteins can bind to liposomes and that thebinding of certain proteins can lead to leakage of entrapped contents as well as promoteliposome removal by phagocytic cells of the reticuloendothial system (RES). The abilityto control liposome uptake by cells of the RES has been a central objective of liposomeresearch. Liposomes that escape these phagocytic cells often distribute within sites thatexhibit blood vessel structures that are permeable to circulating macromolecules. Sincethis is a common feature of blood vessels present within diseased sites, such as cancer,inflammation and infection, liposomes (with no specific surface-associated targetinginformation) can accumulate within these regions. This passive targeting of liposomalcarriers is an important characteristic that must be optimized if ligand-directed targetingto cells outside the vascular compartment is to be achieved. For this reason potentialmechanisms leading to passive accumulation of liposomes at disease sites will also bereviewed here.1.3.1 Interactions with Plasma ProteinsThe interaction of liposomes with plasma proteins is important in terms of controlof liposome drug retention and blood clearance characteristics in vivo (for reviews, seeBonte and Juliano, 1986; Moghimi and Patel, 1993; Juliano and Meyer, 1993). Releaseof encapsulated contents can be promoted, for example, when liposomal lipids are38exchanged or transferred to lipoproteins in the plasma compartment (Krupp et al, 1976;Tall and Small, 1977). More specifically, it is well established that high-densitylipoproteins (HDL) can destabilize liposomes through a combination of mechanismsinvolving transfer of liposomal phospholipids to HDL (Krupp et al, 1976; Tall and Small,1977) as well as through a direct interaction with ApoA-l, an HDL-associatedapolipoprotein. Studies have shown that this protein is capable of penetrating into theliposomal bilayer, thus promoting drug release (Tall and Small, 1977; Wetterau andJonas, 1983; Klausner et al, 1985). In addition, studies by Finkeistein and Weissman(1979) suggested that complement proteins could also promote leakage of encapsulatedcontents. It has now been shown that complement protein binding to certain liposomescan result in complement activation and eventual formation of the membrane attackcomplex (for reviews, see Alving and Richards, 1983; Muller-Eberhard, 1986). Studiesusing model membrane systems have shown that this 10 nm diameter pore-like structurepartially lined with C9 molecules can promote solute release from LUVs (Malinski andNelsestuen, 1989).Serum proteins that engender release of liposomally encapsulated drugs willclearly have an impact on the amount of drug a liposome can deliver to a target cellpopulation. Phagocytic cells of the RES form the first line of defense against foreignpathogens and recognition of these pathogens as foreign is, in part, dependent on bindingof serum proteins, referred to as opsonins. Binding of opsonins to foreign particulates,such as liposomes, promotes binding to macrophages of the RES and subsequentinternalization and processing. Studies with bacteria identified that immunoglobulin,39fibronectin and certain complement proteins (C3, C3b) are the most common opsonins(Absolom, 1986; Bonte and Juliano, 1986). Association of these proteins leads toadherence to phagocytic cells through specific plasma membrane receptors present onmacrophages (Rossi and Wallace, 1983; Brown and Juliano, 1985; Senior et al, 1986;Chonn et al, 1992).Studies identifying the types of proteins bound to liposomal membranes havedemonstrated binding of numerous proteins, including significant amounts of IgG,fibronectin, C3 and C3b (Absolom, 1986; Bonte and Juliano, 1986; Loughrey et al,1990c). Further, it has been shown that liposomes with bound opsonins are phagocytosedmore readily by macrophages (Morisett et al, 1977; Hsu and Juliano, 1982; Aragnol andLeserman 1986). LUVs covalently coupled with rabbit IgG, for example, werephagocytosed 5-fold greater by rat liver macrophages than non-coated liposomes(Derksen et al, 1987). When liposomes were coated with purified fibronectin, there was a10-fold increase in macrophage uptake over control, protein-free liposomes (Hsu andJuliano, 1982). The extent to which immunoglobulins, fibronectin, C3b and C3 act asopsonins to enhance liposome internalization by macrophages in vivo, however, isunclear. Chonn (1991) has suggested that the amount of C3 associated with liposomescontaining anionic phospholipids is related to the clearance behavior of liposomes inmice. Further, these studies suggested that a serum protein previously not identified,apolipoprotein H, may also mediate liposome clearance.It is important to note that in terms of designing liposomes for targetingapplications, lipid composition can significantly reduce the impact of serum protein40binding. For example, the destabilizing effects of serum lipoproteins can be attenuated byincorporation of cholesterol (Sweeny and Jonas, 1985). Liposomes with greater than 30mol% cholesterol exhibit remarkably improved drug retention rates over liposomes thathave little or no cholesterol (Gregoriadis and Davis, 1979; Kirby et al, 1980b). The mostsignificant factor controlling the binding of serum proteins, as well as specific opsoninsinvolved in promoting liposomes clearance, is surface charge. Neutral liposomesprepared with cholesterol bind 10-fold less protein than similar liposomes that have anegative charge (Chonn, 1991). Targeting strategies based on surface-associatedimmunoglobulins, however, must consider the fact that the presence of these surface-associated proteins may promote liposome clearance (see Section 4).1.3.2 Interactions with the Reticuloendothial System (RES)As indicated above, removal of particulates from the circulation is a naturalfunction of the RES, which consists of fixed macrophages of the liver (Kupffer cells),spleen, lungs and bone marrow, as well as circulating monocytes (Altura, 1980). Thepronounced tendency for liposomes to localize in the cells of the RES, in particular thefixed macrophages lining the microvasculature sinusoids of the liver and the spleen, haslong been recognized (Gregoriadis and Ryman, 1972; Gregoriadis and Neerunjun, 1974).Liposome accumulation in macrophages can be an advantage if one is trying to targetspecific phannaceuticals to this cell population. Liposomes can be used effectively todeliver macrophage activators (Fidler, 1993; Talmadge et al, 1986) or drugs designed to41eliminate intracellular infections harbored within macrophages (Alving et al, 1978;Alving et a!, 1980; Alving and Swartz, 1984). However, most applications of liposomebased delivery vehicles would benefit from strategies that limit uptake by cells of theRES.Perhaps the three most important factors controlling liposome delivery to the RESare liposome size, liposome charge and lipid dose. It is well established that large (> 1jim) liposomes, such as MLVs, are cleared more rapidly from the blood compartmentthan smaller (< 200 nm) unilamellar vesicles (Juliano and Stamp, 1975). MLVscomposed of phosphatidyicholine and cholesterol, for example, demonstrate circulationhalf-lives of 5 to 15 mm following i.v. injection. This is approximately 4-fold less thanliposomes of identical composition with a median size in the range of 30 - 80 nm(Gregoriadis and Ryman, 1972; Juliano and Stamp, 1975). It is also interesting to notethat liposome size is important in determining which population of RES cells accumulatethe injected liposomes. Phagocytic KUpffer cells lining hepatic sinusoids represent themajor site of liposome accumulation in vivo, particularly of liposomes of larger size (>0.3 jim diameter) (Freise et al, 1980). Small liposomes (< 0.1 jim) have a greatertendency to be accumulated by phagocytic cells of the spleen (Poste et al, 1984).Increasing liposome dose does not necessarily reduce the amount of lipidaccumulated in liver Kupffer cells, however the percentage of injected liposomes in thisorgan decreases as the lipid dose increases (Poste et al, 1984; Hwang, 1987). This is dueto saturation of the ability of the liver to remove liposomes from the circulation. Adecrease in the rate of liver accumulation of liposomes achieved at high lipid doses42results in increased accumulation of liposomes in the spleen. While spleen phagocytesare involved in this uptake process, the spleen also serves as a nonspecific site ofaccumulation where circulating debris is filtered from the blood (Abra et al, 1980; Allenand Everest, 1983). Saturation of the ability of the spleen to retain liposomes results, inturn, in increased accumulation of liposomes in the bone marrow and lung (Poste, 1983;Poste et al, 1984). Studies by Abra et al (1980) illustrated the effect of a saturating predose of MLVs given as a single i.v. injection at a dose of 1.1 g lipid / kg body weight thatresulted in a 5-fold decrease in liver accumulation and a 3 to 4-fold increase in splenicaccumulation of identical liposomes given 1 hr after the pre-injection. Alternately, Allenet a! (1984) demonstrated that chronic administration of liposomes given at a high doseseverely impaired the ability of the RES system to function in terms of elimination ofinjected particles.The presence of a negative surface charge on liposomes increases the clearancerate and enhances macrophage uptake of intravenously administered liposomes (Julianoand Stamp, 1975; Senior et a!, 1985; Schroit et al, 1986; Gabizon and Papahadjopoulos,1992). For example, in vitro, MLVs containing 30 mol% phosphatidylserine (PS),phosphatidylglycerol (PG) or phosphatidylinositol (PT) were phagocytosed by mouseperitoneal macrophages 25-fold, 18-fold and 15-fold better, respectively, than vesiclescomposed exclusively of neutral phosphatidylcholine (PC) (Schroit et al, 1986). Asindicated in section 1.3.1, interaction of certain plasma proteins with negatively chargedliposomes serves to promote clearance by cells of the RES.431.3.2.1 Strategies to Avoid the RESAs noted previously, avoidance of phagocytic cells of the RES is a primaryobjective of investigators using liposomes as drug carriers. The main advantage inavoiding the RES concerns the increased bioavailability of the administered liposomes. Itis the contention of the work summarized in this thesis that ligand directed targeting ofliposomes to cells within a defined region will depend on increased passive targeting (seeSection 1.3.3) of liposomes to the target sites. Strategies to reduce the accumulation ofliposomes in the RES will increase the propensity of injected liposomes to gain access tothe target cells. As indicated above, reduction in RES uptake can be achieved throughuse of small (<200 nm) liposomes prepared with lipids that bind minimal amounts ofprotein (neutral, cholesterol-containing liposomes).Recent studies have identified two additional strategies that can result in evenfurther reductions in RES uptake. First, Bally et al (1990) demonstrated that liposomeclearance from the circulation by the RES is significantly reduced by the presence of anencapsulated cytotoxic drug, such as doxorubicin. More specifically, doxorubicin-loadedliposomes exhibited circulation lifetimes 4 to 5 times longer than identically preparedempty liposomes (Bally et al, 1990). Furthermore, it was shown that animals pre-injectedwith a low dose of liposomal doxorubicin (2 mg doxorubicinlkg) 24 hrs prior to injectionof empty liposomes resulted in significantly longer circulation lifetimes. It wasdemonstrated that this pre-dosing strategy also resulted in a 5-fold reduction in liveraccumulation (Bally et al, 1990; Parr et a!, 1993). It was concluded that this effect,44termed ‘RES blockade’, was due to accumulation of doxorubicin in liver KUpffer cellsand resulting cell toxicity. Studies have shown that the phagocytic cell capacity of theliver does not recover fully for a period of 14 days after injection of liposomaldoxorubicin. This is consistent with the fact that Kupffer cells are eliminated since theturnover time for these cells in mice has been estimated to be 8 days (van Furth, 1988).The second, and more therapeutically interesting, strategy to reduce RES uptakeof liposomes is based on inclusion of certain lipids, such as ganglioside GM1 (Allen andChonn, 1987, Gabizon and Papahadjopoulos, 1988; Allen et a!, 1989) or phospholipidsmodified to exhibit polyethylene glycol (PEG) headgroups (K!ibanov et al, 1990; Allenand Hansen, 1991; Papahadjopoulos et al, 1991; Senior et al, 1991). Addition of theselipids decreases uptake by the phagocytic cells of the RES, resulting in prolonged bloodcirculation times. For example, addition of 10 mol% GM1 in 100 nm PC/cholesterolvesicles increased circulating blood levels by as much as 4 times. There was aconcomitant decrease in uptake by the RES by 3 times 24 hrs after injection (Allen et al,1989). Similarly, incorporation of 5 mol % PEG (Mr = 2000)-derivatized DPSE(PEG2000-DSPE) in 100 nm DSPC/cholesterol vesicles can result in a 4-fold increase incirculation lifetimes and a 2-fold decrease in liver accumulation 48 hrs afteradministration (Papahadjopoulos et al, 1991). It has been proposed that the local surfaceconcentration of highly hydrated groups, such as PEG, will sterically inhibit bothelectrostatic and hydrophobic interactions of soluble plasma proteins at the liposomesurface (Allen et al, 1989; Lasic et al, 1991; Senior et al, 1991; Klibanov et al, 1991).This is supported by the fact that incorporation of negatively charged lipids (PG) into45GM1-containing PC/cholesterol LUVs does not decrease liposome circulation lifetimes at24 hrs post-injection (22.1% and 23.6% of injected dose, respectively) (Lasic et a!, 1991).It is worthwhile to note that studies using doxorubicin-loaded GMI-containingliposomes clearly demonstrate that these liposomes also promote RES blockade (Parr eta!, 1993). This suggests that incorporated lipids such as GM1 and PEG-PE will not resultin a liposome that completely avoids the RES. Rather, it has been suggested that theselipids serve to decrease the rate at which liposomes bind to and/or are internalized bymacrophages (Parr et al, 1993).1.3.3 Passive Targeting of Liposomal Drug CarriersThe primary advantage associated with liposomal carriers exhibiting a reducedtendency to accumulate in the RES system is prolonged circulation lifetimes. Increasedcirculation lifetimes have been correlated with an increased tendency of liposomes toaccess extravascular sites, particularly in regions of disease (Proffitt et al, 1983; Gabizonand Papahadjopoulos, 1988; Papahadjopoulos et al, 1991; Bakker-Woudenberg et al,1992; Wu et a!, 1993; Bally et al, 1994). This is referred to as passive targeting. Normal Vascular StructureMovement of liposomes residing in the blood compartment to an extravascularsite is dependent on blood vessel structure (Poste, 1982; Poste, 1983). Blood capillarieshave been classified, according to the architecture of their endothelium and the46underlying basement membrane, as continuous, fenestrated and sinusoidal (see Figure1.7) (Poste, 1984). Continuous capillaries, found in muscle, connective tissue, skin andmost other tissue, are composed of adjacent, tightly apposed endothelial cells forming acontinuous lining combined with an uninterrupted subendothelial basement membrane.Fenestrated capillaries, found in many glands, the gastrointestinal tract and the renalglomerulus, are composed of an endothelial layer interrupted by fenestrae varying from30 to 80 nm in diameter and sealed by a membranous diaphragm and a basementmembrane which is continuous. Sinusoidal or discontinuous capillaries are thin walledvessels found in the liver, spleen and bone marrow. The endothelium of sinusoidalcapillaries contain large openings (>200 nm in diameter) and the subendothelial basementmembrane is either absent (liver) or present in an interrupted, fragmented state (spleenand bone marrow). Clearly continuous and fenestrated capillaries represent majorphysical barriers to liposome extravasation, while the structure of sinusoidal capillariesmay allow liposomes to escape from the circulation (Wisse, 1970; Poste, 1983).Increased access to extravascular sites within tissues exhibiting a discontinuousendothelium is best illustrated by studies on liver. As indicated in Section 1.3.2,following i.v. administration of large liposomes (> 0.3 jim), the vesicles localize inphagocytic cells of the liver and, if the dose is high enough, will eventually accumulate inthe spleen and bone marrow (Poste et al, 1983). In contrast, liposomes with meandiameters of 100 nm or less will extravasate within the liver, penetrating the gaps in the47Figure 1.7Schematic Diagram Of The Structure Of Three Classes Of Blood CapillariesCONTiNUOUSEndothelial Cells çBasement Membrane ._. ••(Basa lamina)FENESTRATEDEndothelial Cells_( —. -( —Basement Membrane--,,... ..••.., ....•.•....•....•:•‘•• ••‘••--LL.DISCONTINUOUSEndothelial Cells cZZD C ZSpace of Disse_____________ ____________Parenchymal Cells t—r1fTh (ThAdapted from Poste et al, 198448endothelium lining of the hepatic sinusoids and thereby gaining access to liverparenchymal cells (hepatocytes) (Wisse, 1970). Vascular Structure Within a Diseased SiteVascular structure within a diseased region is also hyperpermeable to circulatingmacromolecules (Peterson, 1979; Gerlowski and Jam, 1986; Dvorak et al, 1988; Nagy etal, 1989; Kohn et al, 1992). It is established, for example, that tumors can exhibit uniquemicrovascular structures that are often incapable of maintaining a complete permeabilitybarrier between the vascular compartment and the growing tumor mass (Wu et al, 1993;Jam, 1987; Jam, 1988). Thus, there are potential sites where large drug carriers canescape from the circulation. In addition to sinusoidal and fenestrated vessels, bloodvessels of particular interest include blood channels that lack an endothelial cell liningallowing blood components to percolate around and between cells (Peterson, 1979;Dvorak et al, 1988), and postcapillary venules (Syrjanen, 1978; Warren, 1979; Malech,1988). The latter are characterized by vessel walls composed of endothelial cells, devoidof basement membrane, supported only by some fibrous tissue. Postcapillary venuleshave been identified as sites where circulation leukocytes can adhere to and readily egressinto surrounding tissues (Malech, 1988).In addition to the presence of unique vascular structures, secondary factorsreleased by cells within a disease site may result in the formation of hyperpermeableregions. It is known, for example, that tumors can secrete a vascular permeability factor49(VPF) that further promotes leakage of circulating macromolecules by increasingintracellular gaps and transendothelial transcytosis processes (Senger et al, 1983; Dvoraket al, 1991b; Kohn et a!, 1992; Yeo et al, 1993). Similarly, inflammatory processes,involving movement of circulating leukocytes from the blood, can lead to vascularendothelium damage that increases blood vessel permeability (Malech, 1988; Abbas et al,1991).Regardless of the mechanisms controlling passive accumulation of liposomes inregions of disease, there is substantial evidence demonstrating preferential accumulationof liposomes in sites of infection (Morgan et a!, 1981; Bakker-Woudenberg et al, 1992),sites of inflammation (Williams et al, 1986) and sites of tumor growth (Proffitt et al,1983; Gabizon and Papahadjopoulos, 1988; Papahadjopoulos et al, 1991; Wu et al, 1993;Bally et a!, 1994). This tendency for liposomes to accumulate in diseased sites has beenused to advantage. In particular, diagnostic applications of liposomal imaging agents arebased on accumulation of the carrier in sites of deep-seated infection (BakkerWoudenberg et al, 1992) as well as within tumors (Richardson et al, 1979; Ogihara et al,1986; Ogihara-Umeda et al, 1994). Further, therapeutic applications for certainantibiotics (e.g. Amphotericin B) are rationalized on the basis that liposomal carriers candeliver significantly more drug to disease sites than can be achieved with free drug(Lopez-Berestein et al, 1985; Mackaness, 1990; Janoff, 1992). As expected, it has beenshown that liposomes with long circulation lifetimes have a greater tendency toaccumulate in disease sites (Gabizon and Papahadjopoulos, 1988; Bally et al, 1990;Papahadjopoulos et al, 1991; Wu et al, 1993).50It is important to note that there is no evidence suggesting that extravasatedliposomes interact with any specific cell population other than perhaps tissue-associatedphagocytes. Bally et al (1994) demonstrated, for example, that liposomes accumulateefficiently within the peritoneal cavity of mice bearing an ascitic tumor. Analysis of thecells within this tumor suggested that liposomes interacted only with the tumor-associatedmacrophages. The bulk (> 70%) of the extravasated liposomes remained in theextracellular space.1.4 Ligand Directed Targeting Of Liposomal DrugsThe objective for developing targeted liposomal drugs has been based onachieving improved delivery of an associated drug to a defined target site. As indicatedin the previous section, this has already been achieved using conventional (firstgeneration) liposomes and optimized with second generation (polymer-coated) liposomes.Studies have demonstrated, for murine solid tumor models, 3- to 5-fold increases inanticancer drug exposure in tumors when employing optimized liposomal carriers ascompared with free drug (Mayer et al, 1993b). Recent studies suggest, however, thatmuch of the accumulated drug is not bioavailable. Drug within the tumor is encapsulatedin regionally localized liposomes and is not available to bind and selectively kill rapidlydividing tumor cells within the site (Mayer et al, 1994). It is therefore important here toindicate that the primary objective of studies outlined in this thesis is to demonstrate a51redistribution of regionally localized liposomes. A redistribution is mediated through aspecific target ligand/target molecule interaction.To date there has not been a single definitive report demonstrating liposometargeting to a defined cell population in vivo. In contrast, there have been numerousstudies demonstrating target ligand dependent delivery of liposomes to cells in vitro(Leserman et al, 1981; Heath et al, 1983; Straubinger et al, 1988; Bankert et al, 1989;Ahmad and Allen, 1992). Targeted liposomes have been reported to demonstrate a 2- to28-fold increase in target cell-association over control liposomes in vitro (Leserman et a!,1981; Heath et al, 1983; Bankert et al, 1989; Ahmad and Allen, 1992). Further, selectivetoxicity of encapsulated drugs has been demonstrated when using liposomes that aretargeted to cultured cells (Heath et al, 1983; Matthay et al, 1984; Straubinger et al, 1988;Ahmad and Allen, 1992). An exhaustive review of these previous in vitro studies will notbe provided here (several excellent reviews are available elsewhere (Weinstein andLeserman, 1984; Torchilin, 1985; Connor et al, 1985; Leserman and Machy, 1987)).Rather, the focus of this section will be to summarize the strategies used to achievetargeting of liposomal carriers. Further, a brief review of the use of targeted liposomes invivo is provided.1.4.1 Targeting StrategiesLiposomes can be targeted to cells by attaching a cell-specific targeting ligand tothe surface of the liposome which is subsequently administered. This one-step targeting52approach is the most commonly employed technique. Alternatively, a two-step targetingstrategy can be used whereby the target cell is first labeled with a specific targetingligand. Subsequently, binding of the target liposome can be mediated through other welldefined binding reagents, such as avidin and biotin (Green, 1975). Target specificdelivery of protein-based imaging agents has been achieved effectively using thisapproach, where target cells are pre-labeled with a biotinylated antibody and subsequentlylabeled with radiolabeled streptavidin (Rusckowski et a!, 1992; Sung et al, 1994; Saga etal, 1994). Further, studies have demonstrated that streptavidin liposomes can bespecifically bound to target cells (in vitro) that have been pre-labeled with a biotinylatedantibody (Rosenberg et al, 1987; Loughrey et al, 1990; Longman et al, 1994). Forexample, Loughrey et a! (1990) demonstrated specific targeting of streptavidin-liposomeconjugates entrapped with carboxyfluorescein to human peripheral lymphocytes prelabeled with biotinylated monoclonal antibody in vitro.These two different targeting approaches are illustrated in Figure 1.8. Clearly, thesingle step approach, based on a targeting ligand attached directly (covalently or noncovalently) to the drug-loaded liposome, is the most straightforward. However, there arepotential disadvantages, including the fact that antibody-coated liposomes may be rapidlycleared from the circulation as a result of Fc-mediated endocytosis by macrophages(Morisett et al, 1977; Hsu and Juliano, 1982; Aragnol and Leserman, 1986). Further,target cell specificity is mediated by a single target ligand that must be modifiedextensively in order to be coupled to the liposome (see Section Loss of antibodybinding affinity is a significant problem associated with this approach. Two-stepapproaches can take advantage of the fact that several different targeting ligands can be53Figure 1.8Targeting Of Protein-coated Liposomes To CellsExample OfDirect TargetingAntibody-liposome conjugateExample Of Two-step TargetingStep 1. Pre-label cells with biotinylated antibodyBiotinylated antibodycellStep 2. Incubate streptavidin-liposome conjugateswith pre-labeled cellscellprelabeledcellStreptavidin-liposome conjugate54used in order to label a defined cell population. Further, the interaction between the prelabeled target cell and the liposome can be mediated by a well defined, high affinitychemical interaction. Biotin, for example, can bind four potential binding sites per avidinmolecule with a binding affinity of approximately 1014 (Chaiet and Wolf, 1964). Theprimary disadvantage of the two-step approach concerns the fact that the pharmaceuticalbehavior of both the cell specific targeting ligand and the liposomes must be wellcharacterized. As indicated in Section 1.1, there are, however, several examples ofantibodies that have been used clinically for both imaging and therapeutic purposes(Hellström et al, 1990; Trail et al, 1993). A final concern about the two-step procedure isthe possibility of developing immune reactions to the various target molecules employed(Schroff et al, 1985; Lobuglio et a!, 1988). It is well established that immune responses(HAMA) to injected IgG limit their utility as therapeutic agents. Additional reactionsagainst biotin on the antibody or streptavidin on the liposome, for example, wouldseverely limit the potential utility of the approach. In terms of characterizing factors thatmay influence binding of targeted liposomal drug carriers in vivo, the two-step approachis extremely versatile. Factors that govern the binding of targeting ligands to the targetcell can easily be separated from factors that govern target site access and target cellbinding of i.v. administered targeted liposomal drugs.551.4.2 In Vivo Targeting of Liposomal DrugsFew reports have demonstrated in vivo efficacy with ligand-directed liposomaldrugs. Bankert et al (1989) showed that cytosine arabinoside(Ara-C)-loaded antiidiotype-bearing liposomes suppressed the growth of a B-cell tumor (2C3) in the spleen,but failed to affect the primary 2C3 tumor in the peritoneal cavity. The laboratory ofHuang demonstrated that 125-labeled liposomes conjugated to an antibody specific for anepitope on mouse lung capillary endothelial cells located in the lung 15 times better thanliposomes conjugated to a non-specific antibody at 24 h post-injection (Hughes et al,1989; Maruyama et al, 1990). Nässander et al (1992) achieved a 28-fold increase onspecific binding of i.p. injected antibody-coated liposomes to a xenograft model of i.p.growing human ovarian carcinoma (OVCAR-3). In the first study to demonstratesuccessful efficacy, Ahmad et al (1993) were able to reduce tumor burdens or completelyeradicate tumors for a number of mice bearing murine lung squamous carcinoma (KLN205) cells following treatment with i.v. injected antibody-coupled doxorubicin-loadedliposomes.1.5 ObjectivesIn order for actively targeted drug-loaded liposomes to be considered clinicallyuseful, they must satisfy certain requirements. First, the liposomes must retain thosecharacteristics that initially defined them as useful drug carrier systems. The presence of56site-directing ligands or incorporation of molecules imparting a hydrophilic-coating onthe liposome surface must not interfere with the entrapment of drug, liposome stability,drug retention or create unacceptable levels of toxicity. Second, liposomes mustspecifically recognize and interact only with the target cell, with negligible interactionwith nontarget cells. Third, liposomes must be able to gain access to the target cellpopulation. It can be assumed, for example, that enhanced passive targeting of drug-loaded liposomes to regions which maintain the target cell population will lead toenhanced active targeting.This thesis incorporates several studies that lead to a better understanding of howto successfully target drug-loaded liposomes to cells residing within an extravascular site.This thesis also describes a novel approach in which drug targeting can be achieved.Chapter 3 deals with the in vivo targeting of doxorubicin-loaded streptavidin-coatedliposomes, injected i.v., to target cells localized in the peritoneal cavity. Chapter 4investigates strategies that increase the extravasation of streptavidin-coated liposomeswhile chapter 5 characterizes the binding of streptavidin-coated liposomes to a biotinlabeled MLV target, and explores the active targeting of streptavidin-coated liposomes tothis target in an extravascular site.57CHAPTER 2MATERIALS AND METHODS2.1 MaterialsDistearoyl phosphatidyicholine (DSPC) and N-biotinoyl dioleoylphosphatidylethanolamine (B-DOPE) were obtained from Avanti Polar Lipids and Nsuccinimidyl 3-(2-pyridyldithio) propionic acid (SPDP) was from Molecular Probes.Doxorubicin was obtained from Adria Laboratories of Canada (Mississauga, Ontario).Cholesterol, dithiothreitol (DTT), 2,2’ -azino-di-(3-ethylbenzthiazdine suphonate) (ABTS),Triton-X- 100, 13-mercaptoethanol, N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid(HEPES), N-ethylmaleimide (NEM), bovine serum albumin (BSA), streptavidin, SephadexG-50, D-biotin, sepharose CL-4B, and all salts were obtained from Sigma.[3H]-cholesterylhexadecyl ether and[‘4C]-cholesteryl hexadecyl ether were obtained from NEN and[14C]-biotin was from Amersham. Biotinylated anti-mouse Thy 1.2 antibody and normal mouseserum was purchased from Cedarlane Laboratories (Hornby, Ontario). Purified N-(4-(P-maleimidophenyl)butyryl) dipalmitoyl phosphatidylethanolamine (MPB-DPPE), Nbiotinoyl distearoyl phosphatidylethanolamine (B-DSPE), N-biotinoylaminohexanoyldistearoyl phosphatidylethanolamine (BAH-DSPE) and polyethylene glycol-modified58Figure 2.1Structures of Various Modified LipidsN-(4-(p-Maleimidophenyl)butyryl)- 1 ,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine(MPB-DPPE))NH0 0 H 0)(CH215C3ç_)“I(C2JNH HCo_Lo...)<o(cHZ)cHN-(Biotinoyl)- 1 ,2-distearoyl-sn-glycero-3-phosphoethanolamine(B-DSPE)- 00 0 0 H )(CH216C3N-((6-(Biotinoyl)amino)hexanoyl)- 1 ,2-distearoyl-sn-glycero-3-phosphoethanolamine(BAH-DSPE)59phosphatidylethanolamine (PEG-DSPE) were purchased from Northern Lipids Inc.,Vancouver, Canada.2.2 Preparation of LiposomesLarge unilamellar vesicles (LUV) were prepared using the extrusion methoddescribed by Hope et al (1985). Briefly, lipid films were prepared from a chloroformsolution by drying under a stream of nitrogen followed by vacuum evaporation for 2 h.Lipids were then hydrated at 65°C in either 300 mM citrate (pH 4.0) or 150 mlvi NaC1, 25mM HEPES, pH 7.5 (HBS) by vortex mixing such that a final lipid concentration of 50 mMwas achieved. LUVs were then prepared by extrusion (5 times) at 65°C through stackedpolycarbonate filters (100 nm pore size) (Nucleopore, Inc.) employing an Extruder (LipexBiomembranes, Vancouver). LUVs were subjected to five freeze-thaw cycles, followed byrepeated extrusion (5 times). Mean vesicle diameters were determined by quasielastic lightscattering (QELS) using a Nicomp 270 submicron particle sizer operating at a wavelengthof 632.8 nm. For lipids hydrated in 300 mM citrate (pH 4.0), the exterior pH was raised forthe coupling reaction by passing the vesicles (pH 4.0) down a Sephadex G-50 (medium)column (1.5x20 cm) pre-equilibrated with 150mM NaCJ, 25 mM HEPES, pH 7.5 (HBS).Multilamellar vesicles (MLV) were prepared by hydrating the lipid film (54 mol%DSPC, 45 mol% cholesterol, 1 mol% biotinoyl amino hexanoyl-DSPE) at 65°C in HBS byvortex mixing such that a final lipid concentration of 150 mM was achieved. MLVs werewashed twice in FIBS (3000 rpm x 10 mm) to remove any LUV or small unilamellar60vesicles (SUV) which formed during hydration. Typically, liposomes were prepared suchthat the non-exchangeable, non-metabolizable lipid[3H]-cholesteryl hexadecyl ether (forLUVs) or[14C1-cholesteryl hexadecyl ether (for MLVs) was incorporated as a liposomelabel for quantification in both in vitro and in vivo studies.2.3 SPDP Modification of StreptavidinStreptavidin (5 mg/mi in HBS) was modified with the amine reactive agent, SPDP,according to published procedures (Loughrey et a!, 1990) (see Figure 2.2). Briefly, SPDP(25 mM in methanol) was incubated at a 10-fold molar ratio to streptavidin at roomtemperature for 10 mm. SPDP-modified streptavidin was reduced with dithiothreitol (DTT)(10-fold molar excess over SPDP, 10 mm) and passed down a Sephadex G-50 columnequilibrated with HBS to remove unreacted SPDP and DTT. Reduced SPDP-modifiedstreptavidin was immediately used for coupling experiments. The extent of modification ofstreptavidin was determined by estimating the concentration of the protein at 280 nm (molarextinction coefficient, E280: 1 .66x 10) prior to the addition of DH, and the 2-thiopyridineconcentration was measured at 343 nm (E343: 7550) 10 mm after addition of DTT accordingto published procedures (Carlsson et al, 1978).61Figure 2.2SPDP Modification of Streptavidin_____ _____-sStreptavidin SPDPJLH22HN(.._)(CH2)1s-sPDP-StreptavidinDTTJH22UJO(CH2)-S -HReduced PDP-Streptavidin622.4 Coupling of SPDP-Streptavidin to LiposomesThe coupling of SPDP-modified streptavidin to liposomes was performed in amodified version of the method used by Loughrey et al (Loughrey et al, 1990) (see Figure2.3). Briefly, SPDP-streptavidin was incubated with liposomes at a ratio of 75 jigprotein/jimol lipid (10 mM final lipid concentration). The coupling reaction was stoppedafter the desired incubation time by the addition of dithiothreitol (DTT) (10 times molarexcess over MPB-PE) followed 2 mm later by the addition of N-ethyl maleimide (NEM)(10 times molar excess over DTT + SPDP), and unassociated protein was removed by gelfiltration on sepharose CL-4B equilibrated with fiBS or 300 mM citrate, pH 4.0 (if reextrusion was employed). In the absence of PEG-PE incorporation, crosslinking of vesiclesduring coupling resulted in liposome aggregation (300-400 nm in diameter as determined byQELS). Where indicated, aggregated protein-vesicle conjugates were re-extruded throughstacked polycarbonate filters (100 nm pore size) to generate a defined size population, (130-170 nm in diameter) (Loughrey et al, 1990). Vesicle interior (300 mM citrate, pH 4.0) wasmaintained during re-extrusion by, as indicated above, passing the liposomes down acolumn equilibrated with citrate buffer. A transmembrane pH gradient was established bypassing the re-extruded streptavidin-coated vesicles down a Sephadex G-50 column preequilibrated with HBS (pH 7.5). The extent of streptavidin coupling to liposomes wasdetermined using a functional assay that measured binding of[‘4C]-biotin to streptavidin.Briefly, SA-LUVs (0.50 jimol lipid in 0.5 ml) were incubated with[14C]-biotin (7.31 nmoladded, 46.9 nmollmCi) for 10 mm and unbound biotin was removed by gel filtration on63Figure 2.3Coupling of SPDP-Modilied Streptavidin to Liposomes0II—--c -(cl-i2)3iiøStreptavidin-MPB-PE Liposome Conjugate2fIl4TAVIDN 0HIReduced PDP-Streptavidin2HN€EPTAVIDIN 0HILiposome containing MPB-PE0IIH‘Sri(CH2)64Sephadex G-50 (medium) equilibrated with fiBS. The extent of [‘4C}-biotin binding toSPDP-modified streptavidin obtained after gel filtration was used as a standard to calculateprotein-to-lipid ratios.2.5 Doxorubicm Uptake into Streptavidin-Coated VesiclesStreptavidin liposomes (5-10 mM lipid) exhibiting a transmembrane pH gradient(interior acidic) were incubated with doxorubicin at a drug-to-lipid ratio of 0.2:1 (mol:mol)in HBS at 65°C for 10 mm. Free doxorubicin was separated from doxorubicin entrapped inSA-LUVs by column chromatography using Sephadex G-50 pre-equilibrated with HBS. Invitro drug retention was evaluated by placing 2 mM lipid in dialysis tubing, (6-8,000 mwco,Spectrum Medical Industries, Inc., LA), 10% fetal calf serum (Gibco Labs), and dialysingagainst 25 ml 10% fetal calf serum at 37°C. In vivo drug retention was evaluated byinjection of drug-loaded liposomes via lateral tail vein at a dose of 3.29 jimol lipid/mouse(100 mg lipid/kg). Blood was collected via cardiac puncture and placed in EDTA-treatedmicrotainers (Becton-Dickinson, Canada). Plasma was prepared by centrifuging (200 x g)blood samples for 10 mm in a clinical centrifuge and lipid and doxorubicin was quantifiedas described in Section 2.8.652.6 Biotinylated Anti-Thy 1.2 Antibody Binding to P388 CellsMurine leukaemia P388 cells grown in culture (37°C in 5% C02) were aliquoted(2x107 cells/mi) into conical bottomed polypropylene tubes, and incubated with biotinylatedanti-Thy 1.2 antibody (0 to 3.125 nM) in PBS containing 2% BSA (w/v) for 30 mm at 4°C.Cells were washed three times with PBS and then incubated with streptavidin peroxidase(10 ji.g/ml) for 30 mm at 4°C. After washing three times with PBS, cells were assayed forperoxidase activity according to the procedure of Porstmann et al (1985). Briefly, 100 il offreshly prepared ABTS solution (30 jig ABTS in 0.2 M glycine buffer, pH 6.0, and 0.2%hydrogen peroxide) was added to cells (1O) in 400 jil of acetate buffer (0.1 M, pH 4.0).The reaction was quenched at 1 hour by the addition of 10 mM sodium azide. Cells werepelleted by centrifugation and supernatants were read at 410 nm on a Bausch and Lombspectronic 2000 spectrophotometer.2.7 In Vitro Targeting of Streptavidin-Coated Liposomes to P388 CellsThe murine lymphocytic leukaemia cell line P388 was obtained from NCI TumorRepository and grown in RPMI 1640 (Flow Laboratories) supplemented with 10% fetal calfserum (FCS, Flow Laboratories). Murine leukaemia P388 cells grown in culture (37°C in5% CO) were aliquoted (10 cells/mi) into conical bottomed polypropylene tubes, andincubated with biotinylated anti-mouse Thy 1.2 antibody or alone in PBS containing 2%BSA (w/v) for 30 mm at 4°C. The fmal antibody concentration for these studies was 12.566nM. After washing twice with PBS, cells were incubated with doxorubicin-loaded SALUVs (2 mM final lipid concentration) for 30 mm at 4°C. The cells were then washed threetimes with PBS and cell-associated lipid and doxorubicin were assayed as described inSection 2.8. Cells were also analysed by flow cytometry (Section 2.9) and fluorescentmicroscopy (Section 2.10). Controls included cells incubated with doxorubicin-loadedliposomes not conjugated with streptavidin.2.8 AnimalsFemale CD1 mice (20-25 g) and female BDF1 mice (18-22 g) were obtained fromCharles River (Canada). Groups of four mice per experimental point were given thespecified treatment in either an i.p. dose (500 il volume) or a single i.v. dose via the lateraltail vein (200 t1 volume, via the lateral tail vein). Blood was collected via cardiac punctureand placed in EDTA-treated microtainers (Becton-Dickinson, Canada). Plasma wasprepared by centrifuging (200 x g) blood samples for 10 mm in a clinical centrifuge. Totalplasma volume per animal was taken to be 4.55% of mean body weight.Biodistribution studies were performed on the same mice used for plasma clearancestudies. Peritoneal cavities were lavaged with 5 ml of indicator free and Ca2,Mg2 freeHanks buffered saline solution (I{BSS) that was injected i.p. The abdomen was gentlymassaged and the peritoneal fluid was removed with a syringe equipped with a 22 g needle.Peritoneal fluid was assayed for liposomal lipid and doxorubicin as indicated below. BDF1mice bearing P388 tumors were prepared by inoculating mice i.p. with lx 106 P388 cells.67Tumors progressed for 4 days prior to i.v. injection of antibody or the specified liposomalpreparation. Four days after cell administration there was no measurable ascites fluid.Lavage fluid from either tumor-bearing or tumor-free animals that contained red cellcontamination was discarded. BDF1 mice bearing Lewis lung tumors were prepared byinoculating mice subcutaneously with 3x105Lewis lung cells. Tumors progressed for 10-13days prior to injection of liposomes.Liposomal lipid was measured by incorporation of tracer quantities of the nonexchangeable, non-metabolizable radiolabeled lipid marker[3H]-cholesteryl hexadecyl ether(Bally et al, 1990). Cell-associated lipid was then determined by liquid scintillationcounting. Samples in PBS were mixed with 5 ml Pico-Fluor 40 scintillation cocktail(Packard, Canada) prior to counting on a Packard 1900 TR scintillation counter.Cell-associated doxorubicin was measured using a fluorescent assay procedure asdescribed in Baily et al (1990). Briefly, samples to be assayed were diluted to 800 jil indH2O. Subsequently 100 ji.l 10% SDS and 100 j.il 10 mM H2S04 were added prior toaddition of 2 ml isopropanol:chloroform (1:1 vlv). This two phase system was vigorouslymixed and frozen at -20°C to promote protein precipitation. After thawing, the sampleswere centrifuged (500 x g for 5 mm) and the organic phase was carefully removed.Doxorubicin in the organic phase was measured by fluorescence (excitation wavelength,500 nm; emission wavelength, 550 nm) using a Perkin-Elmer LS5O-B spectrofluorometer.A standard doxorubicin curve was prepared using a similar extraction procedure.Following peritoneal lavage the liver and spleen were removed from each animaland weighed. Organs were homogenized using a Polytron homogenizer (Brinkmann68Instruments) in distilled H20 to produce a 10% (w/v) homogenate for the spleen and 20%(w/v) homogenate for the liver. Tissue homogenates (200 jil for spleen, 100 jil for liver)were digested with 500 p1 of Solvable (DuPont Canada, Inc., Mississauga, Ont.) for 3 h at50°C. Samples were allowed to cool to room temperature and 50 p1 of 200 mM EDTA wasadded to prevent foaming during decolourising with 200 p1 of 30% hydrogen peroxide. 25jil of iON HC1 was added to reduce chemiluminescence. Subsequently, samples weremixed with 5 ml Ultima Gold scintillation cocktail (Packard, Canada) prior to counting on aPackard 1900 TR scintillation counter.2.9 Flow Cytometric AnalysisCell-associated doxorubicin fluorescence was measured on an EPICS 753 flowcytometer (Coulter) with an excitation wavelength of 488 nm at 200 mW power with anargon ion laser. A 488 nm dichroic mirror and 488 nm bandpass filter were used tomeasure the side scatter signal. A 515 nm interference filter and a 488 nm laser blockingfilter were used to block the excitation light from the fluorescence detectors. A 515 nmlongpass filter was used to measure doxorubicin. Gates were set around the P388 cellsusing forward scatter and side scatter signals to exclude cell clumps and debris.692.10 Fluorescent MicroscopyLiving cells were viewed with a Leitz Dialux fluorescence microscope with phaseobjectives and epifluorescence illumination. Original photographs were made with Fuji 400colour slide film, with exposure time determined automatically by a photometer in thecamera.2.11 Cytotoxicity AssaysTypically, P388 cells (1 .Ox107/ml) were first incubated with biotinylated anti-Thy1.2 antibody (12.5 nM) for 30 mm at 4°C, and washed twice in RPMI media 1640 with10% (vlv) FBS. 106 cells were aliquoted into 24 well tissue culture plates and exposed toeither free drug or liposomal drug and incubated (37°C, 5% C02) for 24 hours at theindicated time, an aliquot of each culture was counted on a Coulter Counter ZM todetermine total cell count. Viability was determined using the flow cytometric method ofRoss et a! (1989). Briefly, propidium iodide (20 mg/mI final concentration) was incubatedwith cells for 30 mm, after which fluorescence was assessed on an EPICS 753 flowcytometer. An argon laser at 488 nm was used for excitation of red fluorescence greaterthan 630 nm. Relative % viability was determined by multiplying % viability by total cellcount for each culture.702.12 Separation of MLVs and LUVsMLVs (20 j.tmol lipid) and LUVs (0.5 pmol lipid) were incubated together at RT for30 mm (1.0 ml final volume). Samples were spun at 3000 rpm (1600 g) for 10 mm and thesupematant was collected. After 3 washes with FIBS (1600 g x 10 mm.) the pellet wastaken up in 1.0 ml FIBS. Both supematant and pellet were assayed for lipid as described inSection 2.8.71CHAPTER 3IN VIVO TARGETING OF STREPTAVIDIN LIPOSOMES TO P388 CELLS3.1 IntroductionAs mentioned in Section 1.1, the therapeutic index of certain drugs can beimproved, sometimes dramatically, when given in association with a liposomal carrier.This has been best characterized for the anticancer drug doxorubicin where it has beenshown that liposomally encapsulated drug is less toxic and more efficacious than free drug(Balazsovits et a!, 1989; Mayer et al, 1990; Gabizon et al, 1985; Rahman et al, 1990;Cowens et al, 1993). Reduced toxicity and enhanced anti-tumor activity are thought to be aconsequence of liposome-mediated changes in pharmacokinetics and biodistribution. Forexample, the chronic dose limiting toxicity of doxorubicin is cardiotoxicity, a toxicity that isreduced significantly when the drug is given in liposomes (Olson et al, 1982; Balazsovits etal, 1989; Gabizon et al, 1986; van Hoesel et al, 1984). Since liposomes do not accumulatein cardiac tissue, this reduction may be attributed to a decreased availability of free drug inthis tissue. Conversely, increased anti-tumor activity can be attributed to the tendency of(small) liposomes to accumulate to significant levels at sites of tumor growth (Proffitt et al,1983; Gabizon and Papahadjopoulos; 1988; Papahadjopoulos et a!, 1991; Wu et al, 1993;Bally et al, 1994).A significant advance in the development of liposomal anticancer agents hasconcerned the use of liposomes, containing GM1 or polyethylene glycol-modified72phosophatidylethanolamine (PEG-PE), that exhibit long circulation lifetimes (Allen andChonn, 1987; Gabizon and Papahadjopoulos, 1988; Klibanov et al, 1990; Allen andHansen, 1991; Huang et al, 1992) (see Section 1.3.2). Importantly, it has beendemonstrated that such liposomes readily access sites of tumor growth. In terms ofachieving cell-specific targeting this passive tendency to accumulate in sites of tumorgrowth is a fundamental requirement for targeting of liposomes. One possible limitation ofantibody-coated liposomes is that they are immunologically better recognized than protein-free liposomes and consequently exhibit shortened circulation lifetimes following i.v.administration (Aragnol and Leserman, 1986; Debs et al, 1987). Elimination of i.v.administered antibody-coated liposomes may be due to Fc-mediated clearance (Aragnol andLeserman, 1986) or to the fact that protein-liposome coupling technology promotes theaggregation of liposomes (Loughrey et al, 1990b; Loughrey et a!, 1993). Liposomeaggregates will exhibit reduced circulation lifetimes due to their increased size (Senior et al,1985). Regardless of the mechanisms involved, a reduction in circulation longevity willreduce the tendency of these protein-coated liposomes to access target cell populationresiding in an extravascular site.An alternative approach has been developed to achieve targeting in vivo thatspecifically addresses issues regarding in vivo circulation and targeted liposomeextravasation (see Section 1.3). The targeting procedure relies on the use of streptavidinliposome conjugates (SA-LUVs) that bind specifically to target cells pre-labeled with abiotinylated antibody (Loughrey et a!, 1990; Loughrey et al, 1990b; Longman et a!, 1994).These liposomes have been used to label defined cell populations in vitro and exhibit in73vivo circulation lifetimes comparable to control liposomes that lack surface-associatedprotein. The studies presented in this chapter extend these previous studies bycharacterizing SA-LUV targeting to murine lymphocytic leukaemia cells (P3 88) in vivo.Targeting is achieved by pre-labeling P388 cells with biotinylated anti-Thy 1.2 antibody.The cytotoxic advantage of targeting SA-LUVs loaded with the anticancer drug doxorubicincompared to control liposomal doxorubicin systems is demonstrated in vitro. Finally, it isdemonstrated that this two-step targeting approach can effectively target P388 cells(residing in the peritoneal cavity) following i.v. administration.3.2 Results3.2.1 Characterisation of Doxorubicin-Loaded Streptavidin LiposomesPrevious reports demonstrated that liposomes with covalently or non-covalentlyattached streptavidin can be targeted to specific cells via biotinylated antibodies (Loughreyet al, 1987; Loughrey et al, 1990, Loughrey et al, 1993b). Targeting can be achieved in twoways. First, biotinylated antibodies can be bound to streptavidin liposomes, forming animmunoliposome that can then be targeted to a selected cell population (Loughrey et al,1987; Loughrey et a!, 1993). A similar procedure has been used successfully by Ahmad etalto prepare immunoliposomes with specificity to lung tumor cells in vitro and in vivo(Ahmad and Allen, 1992; Ahmad et al, 1993). Target cells can also be pre-labeled withbiotinylated antibody prior to addition of the streptavidin liposomes (Loughrey et al, 1990;74Loughrey et al, 1993). This two-step approach has been further developed here for deliveryof liposomes containing doxorubicin to murine P388 cells in vivo. One reason for pursuinga two-step approach addresses the fact that antibody-coated liposomes can be cleared fromthe circulation much faster than streptavidin-coated liposomes. This is illustrated for themurine model system employed here by the data in Figure 3.1. This comparison of thecirculation lifetimes between SA-LUVs and SA-LUVs with bound biotinylated antibody(MoAb-SA-LUVs) shows that the antibody-coated liposomes are cleared rapidly afteradministration. One hour after i.v. injection, 52% of the SA-LUVs injected dose was foundin the circulation compared to 1.5% obtained after administration of MoAb-SA-LUVs.Increased circulation lifetimes of SA-LUVs will increase the probability that the drug-loaded carrier will access extravascular target cells.In these studies the procedure used to prepare SA-LUVs has been modified so thatthe protein-liposome conjugates can be employed for delivery of a cytotoxic drug. Thetransmembrane pH gradient loading procedure, developed for encapsulation of theanticancer drugs doxorubicin and vincristine, was used (Bally et al, 1988) (see Section 2.5).Briefly, 100 mn liposomes, composed of DSPC/cholIMPB-PE (54:45:1), were prepared in300 mM citrate buffer (pH 4.0). The external pH was adjusted to pH 7.5 prior tostreptavidin attachment to MPB-PE. SPDP-modified streptavidin (2-3 sulthydryl groupsper protein) was incubated with the liposomes as indicated in Section 2.4. As shown inprevious reports, protein coupling promoted vesicle aggregation (Loughrey et al, 1990;Loughrey et al, 1993). Hence, the crosslinked streptavidin-coated liposomes were extrudedthrough 100 nm pore size filters after the coupling reaction was quenched by addition of75Cl)00Figure 3.1Plasma Clearance of Streptavidin-Coated and Antibody-Coated LiposomesStreptavidin-coated liposomes ( 0 ) and antibody-coated liposomes ( • ) were injected vialateral tail vein at a dose of 3.29 pmol lipid/mouse (100 mg lipid/kg). The mice were terminatedat the indicated times and the level of liposomal lipid was determined (see Section 2.8). Valuesshown represent the mean of results from at least 4 animals ± S.E. of the mean (p <0.05).1 201101 0090807060504030201000 4 8 12 16 20 24TIME (hours)76NEM. In order for the interior pH to remain at 4.0, the aggregated liposomes were passeddown a sepharose 4B-CL column equilibrated with 300 mM citrate buffer (pH 4.0) prior toextrusion. After the second sizing step, the resulting liposomes typically had 35-45 igstreptavidin per .imo1 lipid and exhibited a mean diameter of 130-170 nm as determined byQELS measurements.The encapsulation and retention characteristics of SA-LUVs loaded withdoxorubicin via use of a transmembrane pH gradient (interior acidic) are summarized inTable 3.1. Greater than 98% doxorubicin encapsulation could be achieved within 2 minutesat a incubation temperature of 65°C. The extent of drug uptake was not influenced byliposome-bound protein and the doxorubicin loading procedure did not affect the biotinbinding activity of SA-LUVs. The resulting drug-loaded SA-LUVs exhibited drug retentioncharacteristics similar to control liposomes, showing no significant drug release (over 4hours) in vitro (10% serum) or in vivo after i.v. injection.An important characteristic of the doxorubicin-loaded liposomes, illustrated by datain Figure 3.2, concerns their increased circulation lifetime. Similar to protein-freeliposomes (Bally et al, 1990), encapsulation of doxorubicin in SA-LUVs promotes theircirculation lifetime. At 24 hours, there was 3-fold more lipid in the circulation compared toempty SA-LUVs. This increase in circulating blood levels is associated with a significantdecrease in liver accumulation for the doxorubicin-loaded liposomes.77TABLE 3.1Doxorubicin Loading and Retention in Liposomes and Streptavidin-LiposomesRATE OF TRAPPING IN VITRO, 4 hrs IN VIVO, 4 hrsLIPOSOME UPTAKE EFFICIENCY DRUG-TO-LIPID DRUG-TO-LIPID(mm) (%) RATIO (mol%)1 RATIO (mol%)2LUVs <5 98 0.203 0.200SA-LUVs <5 98 0.201 0.203Doxorubicin uptake, trapping efficiency and drug retentions were determined as in Section 2.51. Drug retention of doxorubicin-loaded liposomes was evaluated by dialysing 2 mMlipid in 10% fetal calf serum against 25 ml of 10% fetal calf serum at 37°C.2. Drug retention was evaluated by injected of drug-loaded liposomes via lateral tailvein at a dose of 3.29 jimol lipid/mouse (100 mg lipid/kg). Blood was collected viacardiac puncture and placed in EDTA-treated microtainers. Plasma was prepared bycentrifuging (200 x g) blood samples for 10 minutes in a clinical centrifuge and lipidand doxorubicin was quantified as in Section 2.5.78Cf0‘IFigure 3.2Effect of Encapsulated Doxorubicin on Circulation Lifetime of Streptavidin LiposomesDoxorubicin-loaded (solid bar) or empty (empty bar) SA-LUVs were injected via lateral tail veinat a dose of 3.29 jimol lipid/mouse (100 mg lipid/kg). The mice were terminated 24 hours afterliposome injection and the level of liposomal lipid was determined (see Section 2.8). Valuesshown represent the mean of results from at least 4 animals ± S.E. of mean (p <0.05).302520151050EmptySA—LUVDOX—loadedSA—LUV7.3.2.2 In Vivo Labeling of P388 Cells with Biotinylated Anti-Thy 1.2 AbThe targeting potential of doxorubicin-loaded SA-LUVs was assessed using P388cells labeled with biotinylated anti-Thy 1.2 antibody as a model target cell population. Themurine Thy 1.2 antigen is a highly expressed T cell antigen with a molecular weight ofapproximately 25,000 (Ledbetter and Herzenberg, 1979). The extent of biotinylated anti-Thy 1.2 antibody binding to P388 cells is shown in Figure 3.3. These data demonstrate thatapproximately 60,000 antibodies bind per P388 cell. In addition, the data indicates thatonce bound these antibodies are not internalized (Figure 3.3 insert). The apparent bindingconstant for this antibody, as estimated from the data in Figure 3.3, is 5.5x i0 M’.In order to develop the use of P388 cells as a model target site in vivo, it isimportant to demonstrate that these cells can also be labeled with the anti-Thy 1.2 Ab invivo. The ability to stably label P388 cells in vivo with biotinylated anti-Thy 1.2 antibodyis illustrated by flow cytometric data shown in Figure 3.4. Briefly, biotinylated anti-Thy 1.2Ab was injected (i.p.) at a dose of 10 ig (Figure 3.4A) or 100 ig (Figure 3.4B) into micewith established P388 tumors. At 1, 4 and 24 hours the peritoneal cavity was lavaged andantibody labeling of peritoneal cells was assessed. P388 cells within the peritoneal cavitywere gated for on the basis of size and granularity characteristics of cultured P388 cells.The presence of cell-associated antibody was detennined by addition of FITC-conjugatedavidin. The flow cytometric data in Figure 3.4 demonstrates that the percentage ofantibody-labeled cells is constant for periods of at least 24 hours for animals injected with100 jig antibody. Fluorescent microscopy data supports this conclusion (see Figure 3.5). In80Figure 3.3Biotinylated Anti-Thy 1.2 Antibody Binding to P388 CellsMurine leukaemia P388 cells (2x107)were incubated with biotinylated anti-Thy 1.2 antibody atthe indicated concentrations. Bound antibody was determined by a colorimetric assay usingstreptavidin-peroxidase and the substrate ABTS, which was read at 410 nm, as described inSection 2.6. The level of antibody bound to the cell surface was monitored over time at 4°C and37°C (insert).(-)C0-o60000500004000030000200001 00000 I0.0 0.5 1 .0 2.5 3.0= 70000V0 60000500004000030000200001000000 1 2 3 4TIME ( hours1.5 2.0Ab Added ( nM )3.581Figure 3.4In Vivo Labeling of P388 Cells with Ab Injected i.p.Mice with established P388 tumours were injected i.p. with 10 jig (A) or 100 jig (B) ofbiotinylated anti-Thy 1.2 Ab in 0.5 ml HESS (see Section 2.8). Controls received 0 jig Ab in 0.5m1HBSS(—). Atl( ),4and()and24(---)hours,theaniinalswereterminatedand the peritoneal cavity was lavaged with 5.0 ml HBSS. Lavage cells (106) were incubatedwith F1TC-conjugated avidin (1 jig) for 30 minutes at 4°C and examined for cell-associatedfluorescence by flow cytometry (see Section 2.9).NzL)10 108 1008FluorescenceABFluorescence[ . control FITClhr, FITCjq : flilill 4hr,S .‘ ii 24hr, FITC1 4i/1 AFVrii • I,’I, •1 .!d,’dI •I?, Ijj; ‘AM#W..•.:82Figure 3.5Phase Contrast and fluorescent Micrographs of P388 Cells Labeled In Vivo with Anti-Thy1.2 AntibodyBiotinylated anti-Thy 1.2 Ab was injected i.p. into mice bearing P388 tumors as indicated. 24hrs later cells were recovered by lavage and incubated in vitro with F1TC-conjugated avidin (seeSection 3.2.2). (A) Phase contrast view of P388 cells (1O cells/mi) preincubated with 0 jigbiotinylated anti-Thy 1.2 Ab. (B) Fluorescence image of the field in A. (C) Phase contrast viewof P388 cells as in A but with preincubation with 10 jig Ab. (D) Fluorescence image of the fieldin C. (E) Phase contrast view of P388 cells preincubated with 100 jig Ab. (F) Fluorescenceimage of the field in E.vivo targeting studies using streptavidin liposomes were, therefore, initiated in animalsinjected with 100 jig biotinylated antibody.3.2.3 Tn Vitro Targeting of Doxorubicin-Loaded Streptavidin Liposomes to Pre-labeledP388 CellsPrior to initiation of in vivo targeting studies, in vitro liposome targetingexperiments were completed using P388 cells that had been pre-incubated (4°C) with anexcess of biotinylated antibody. Doxorubicin-loaded liposome targeting to P388 cells wasassessed qualitatively using flow cytometry and fluorescent microscopy, where the presenceof cell-associated doxorubicin was detected by fluorescence. Figure 3.6 shows fluorescentmicrographs of P388 cells labeled in vitro with doxorubicin-loaded SA-LUVs. Virtually allcells appear fluorescent when they had been pre-labeled with biotinylated anti-Thy 1.2 Ab(Figure 3.6B). In the absence of antibody or surface-associated streptavidin (Figure 3.6Dand F, respectively) little or no fluorescence was observed. Although not clearly shown inthe photomicrographs, labeling was restricted to the outer membrane and there was noindication of capping.Flow cytometric analysis of labeled P388 cells confirmed these results. As shown inFigure 3.7, specific cell labeling was only observed under situations where drug-loaded SALUVs were mixed with cells pre-labeled with biotinylated anti-Thy 1.2 antibody (Figure3.7A). These data indicate that, in vitro, approximately 90% of the P388 cells were labeledwith doxorubicin-loaded liposomes. This result is consistent with data obtained for cells84Figure 3.6Phase Contrast and fluorescent Micrographs of Targeted Liposomes In VitroLiposomes contained doxorubicin which fluoresces under an FITC filter (see Section 2.10). (A)Phase contrast view of P388 cells (10 cells/mi) preincubated with biotinylated anti-Thy 1.2 Ab(12.5 nM) and subsequently incubated with doxorubicin-loaded SA-LUVs (2 mM lipid) for 30minutes in vitro at 4°C. (B) Fluorescence image of the field in A. (C) Phase contrast view ofP388 cells as in A but without preincubation with Ab. (D) Fluorescence image of the field in C.(B) Phase contrast view of P388 cells preincubated with Ab and subsequently incubated withdoxorubicin-loaded control liposomes (no streptavidin). (F) Fluorescence image of the field inE.,.—85________DOX-SA-t..UU,+Øb________jI(JI IC ICC__ O - -t..LJ-b_______BI ii Iii I 1.11111I IC iee ieee__OOXLUV,+Ab_________I I Ilii I 1111111 II. ie aee ieee__D --Ab_________0I II III I 1111111 I I 11111i. ae aee ieeeP368 C.LJ.._________C38I 1111111 I I 11111.jee ieeeDOXFigure 3.7Targeting of Streptavidin Liposomes to P388 Cells In VitroDoxorubicin-loaded SA-LUVs and LUVs were prepared as described in Section 2.5. P388 cells(i07 cells/mi) were incubated with biotinylated anti-Thy 1.2 Ab (12.5 nM) where indicated for30 minutes at 4°C. After two washes with PBS, doxorubicin-loaded SA-LUVs or LUVs (2 mMlipid) were added as indicated, incubated for 30 minutes at 4°C and washed. Samples weresubsequently examined for cell-associated fluorescence by flow cytometry (see Section 2.9).‘I00C30(34JC300C0I.86labeled with F1TC-labeled streptavidin where 95% of all pre-labeled P388 cells appeared aspositive.Quantitative determinations of cell-associated doxorubicin and cell-associated lipidare shown in Figure 3.8. These data were obtained for P388 cells that had been incubated invitro at 4°C with or without biotinylated anti-Thy 1.2 antibody (10 cells suspended in 1 mlmedia at 12.5 nM Ab) (see Section 2.7). The cells were washed to remove unboundantibody and resuspended in media with drug-loaded SA-LUVs or drug-loaded LUVs(without streptavidin). The final lipid concentration was 2 mM. Liposomal lipid anddoxorubicin were assayed as described in Section 2.8. Using cells pre-labeled withbiotinylated anti-Thy 1.2 antibody there was a 20-fold increase in cell-associated drug and a30-fold increase in cell-associated lipid obtained when cells incubated with drug-loadedSA-LUVs were compared with protein-free liposomal doxorubicin. Non-specific bindingof SA-LUVs to P388 cells that had not been incubated with antibody was typically less than10% of the value obtained for targeted liposomes. Assuming 7x109 liposomes per nmollipid (calculated on the basis of a mean diameter of 100 nm) it can be estimated under theconditions described here that approximately 6,000 liposomes were bound per P388 cell.As indicated previously (Figure 3.3), about 60,000 antibodies bind per P388 cell when asaturating concentration of antibody is present.The cytotoxic activity of doxorubicin-loaded targeted SA-LUVs was also assessedfor cells exposed to drug for 24 hours (see Section 2.11). The results, summarized in Table3.2, indicate that free doxorubicin has an IC50 value almost 20-fold less than the bestliposomal drug. This was not unexpected considering the liposomes used were87Figure 3.8U)0NCIQuantffication of Cell-Associated Doxorubicin and Lipid after Targeting of DoxorubicinLoaded Streptavidin Liposomes to P388 Cells In VitroDoxorubicin-loaded SA-LUVs (40.0 ig streptavidinlpmol lipid) and LUVs were prepared asdescribed in Section 2.5. P388 cells (10) were incubated with or without biotinylated anti-Thy1.2 Ab (12.5 nM) as indicated for 30 minutes at 4°C. After two washes with PBS, SA-LUVs orLUVs (2 mM lipid) were added, incubated for 30 minutes at 4°C and washed three times. Cell-associated doxorubicin (A) and lipid (B) were determined as detailed in Section 2.8 (p <0.005).2.2 -2.0___° AN 1.8C1.61.4z1.2100><o 0.60.400.2 T0.010987654320B+Ab —AbSA—LUV+Ab —AbLUV88Table 3.2IC50 Values for Liposomal and Free DoxorubicmSAMPLE IC50 (jiM)Liposomal Dox, +Ab 3.8 ± 2.0Liposomal Dox, -Ab 10.1 ± 1.3Free Doxorubicin 0.13 ± 0.02IC50 values obtained from P388 cells exposed for 24 hours to doxorubicin-loadedstreptavidin liposomes. Cytotoxicity assays were performed as in Section 2.11 (p <0.05).89prepared from DSPC/chol, a composition known to retain drug for extended periods in vitroand in vivo (Table 3.1). Further, where targeting was achieved it was mediated by bindingto an antigen that is not readily internalized (Figure 3.3). Given these constraints it issurprising that drug-loaded SA-LUVs were 2- to 3-fold more active than doxorubicinentrapped in control liposomes (Table 3.2). These results suggest that the therapeuticactivity of a liposomal drug can be improved through targeting even when the targetingligand used is not internalized.3.2.4 Tn Vivo Targeting of Doxorubicin-Loaded Streptavidin Liposomes to Pre-labeledP388 CellsIn vivo targeting of doxorubicin-loaded SA-LUVs to P388 cells is shown in Figures3.9 and 3.10. A 100 jig dose of biotinylated anti-Thy 1.2 Ab was injected i.p. into micewith established P388 tumors (see Figure 3.4). Tn vivo targeting of liposomes given by i.p.or i.v. administration was evaluated. For i.p. targeting studies, mice were pre-injected withbiotinylated anti-Thy 1.2 Ab, and then injected i.p. with drug-loaded SA-LUVs one dayafter antibody injection. At 1 hour, the peritoneal cavities were lavaged and targeting toP388 cells was assayed as described in Section 2.8. As shown in Figure 3.9, targetedliposomes demonstrated a 2.5 fold increase in the level of cell-associated lipid and drug.Flow cytometric analysis of targeted P388 cells suggested that doxorubicin delivery wasspecific for a defined population of cells. These results, shown in the insert in Figure 3.9,were obtained by gating for P388 cells present in the isolated peritoneal cells. A population90Figure 3.9C/)C)NCfD0SIn Vivo Targeting of P388 Cells with Doxorubicin-Loaded Streptavidin LiposomesInjected IntraperitoneallyMice were injected i.p. with 100 .tg Ab in 0.5 ml HBSS (solid bar) or just 0.5 ml HBSS (emptybar). Twenty-four hours after Ab was injected, doxorubicin-loaded SA-LUVs were injected i.p.at a dose of 3.29 i.tmol lipid/mouse (100 mg lipid/kg). One hour after liposome injection theperitoneal cavities were lavaged and cell-associated doxorubicin (A) and lipid (B) weredetermined as detailed in Section 2.8. Samples were also examined for cell-associatedfluorescence by flow cytometry (A, insert) (p <0.01).C.)NC><C0S1. .5I .00.50.0—Ab +Ab91of cells (20% of gated cells) was found to be highly fluorescent. These cells were labeledwith liposomes only in animals that were pre-injected with anti-Thy 1.2 Ab.Figure 3.10 demonstrates in vivo targeting to P388 cells following i.v.administration of SA-LUVs. Briefly, this data was obtained by allowing i.v. administeredSA-LUVs to accumulate in the peritoneal cavity for 24 hrs (Bally et al, 1994) prior toinjection (i.p.) of 100 jig anti-Thy 1.2 antibody. As expected from the data in Figure 3.9,animals that did not receive antibody showed significant labeling of cells. This wasinterpreted to be a consequence of non-specific labeling of certain cells associated with theascitic tumor, in particular peritoneal macrophages. For this reason, the data presented inFigure 3.10 shows the increase in cell-associated labeling relative to control cells that didnot receive antibody. The data suggests that in the presence of injected anti-Thy 1.2antibody, a 16-fold increase in cell-specific labeling is achieved using SA-LUVs.3.3 DiscussionPrevious reports have described the preparation and characterization of a versatileliposome targeting approach that relies on the high affmity binding of biotin to streptavidin(Loughrey et al, 1990; Loughrey et al, 1990b; Loughrey et al, 1993). The studies describedin this chapter demonstrate that this targeting approach can be used to target doxorubicinloaded liposomes to murine tumor cells both in vitro and in vivo. Further, we show using invitro cytotoxicity assays that doxorubicin entrapped in these targeted liposomes is morecytotoxic than doxorubicin encapsulated in non-targeted liposomes. The results lead to920)C-)0Cl)Cl)0)00)Cl)0)C-)Figure 3.10In Vivo Targeting of P388 Cells with Doxorubicin-Loaded Streptavidin LiposomesInjected IntravenouslyEither doxorubicin-loaded SA-LUVs (solid bar) or doxorubicin-loaded LUVs (empty bar) wereinjected via lateral tail vein at a dose of 3.29 jimol lipid/mouse (100 mg lipid/kg). Twenty-fourhours after liposome injection, animals were injected i.p. with 100 jig Ab in 0.5 ml HBSS. Onehour after Ab injection the peritoneal cavities were lavaged and cell-associated lipid wasdetermined as detailed in Section 2.8. The control for streptavidin liposomes (+Ab) wasstreptavidin liposomes without Ab, and the control for liposomes (-i-Ab) was liposomes withoutAb. Asterisk indicates significant difference (p<O.O5) compared with control.*1605550454035302520151050LUV+AbSA—LUV+Ab93several conclusions regarding the design of liposomes for in vivo targeting and these arediscussed below.The first point concerns the use of two-step targeting approaches for development ofliposomal anticancer agents given by i.v. administration. This thesis is based on thepremise that in order to achieve targeting to cells within an extravascular site, the liposomesmust first be able to efficiently move from the blood compartment to an interstitial spacewithin the target site. Studies using murine models suggest that the circulation lifetimes ofthe carrier must be maximized in order to achieve efficient delivery of entrapped contents totumors (Gabizon and Papahadjopoulos, 1988; Bally et al, 1990; Mayer et al, 1990b;Papahadjopoulos et al, 1991; Gabizon, 1992; Wu et a!, 1993). An advantage of the SALUVs used here is that they exhibit circulation lifetimes comparable to liposomes that donot have associated protein (Loughrey et al, 1990b). In the case of antibody-coatedliposomes, rapid clearance from the plasma compartment has been observed following i.v.administration (Aragnol and Leserman, 1986; Debs et al, 1987). Similar results arepresented here for MoAb-SA-LUVs prepared using the avidin-biotin bridge technique (seeFigure 3.1). Clearance of these immunoliposomes may be partly attributed to liposomeaggregation; however, the data presented in Figure 3.1 were obtained using liposomesexhibiting identical mean diameters (approximately 170 nm). The presence of surfaceassociated antibody clearly decreased the circulation lifetime of the liposomes.It is important to demonstrate, in vitro, that the two-step targeting procedure can beas efficient as direct approaches involving antibody-coated liposomes. Previous reportshave indicated that antibody-coated liposomes can result in a 2- to 28-fold increase in cell94associated liposomes compared to non-targeted control liposomes (Ahmad and Allen, 1992;Bankert et al, 1989; Heath et al, 1983; Leserman et al, 1981; Matthay et al, 1984; Nassanderet al, 1992). The data presented here demonstrates that the two-step targeting procedure,based on binding SA-LUVs to P388 cells pre-labeled with biotinylated anti-Thy 1.2antibody, results in a 30-fold increase in liposomal lipid association over controls. Alimitation of the two-step procedure, however, concerns the fact that optimal targetingrequires use of a target antigen that is not internalized. The Thy 1.2 antigen was chosen as atarget ligand for this reason. Thy 1.2 is a glycolipid-anchored small molecular weight (25Kdalton) protein that is highly expressed on P388 cells. The biotinylated antibody used forthese studies exhibits a relatively high binding affinity (5.5x 10 M1) and, once bound, therate of antibody loss from the cell surface is low (Figure 3.3 insert). This retention ofsurface-associated antibody was also maintained in vivo (Figure 3.4).In vivo targeting to pre-labeled target cells was also demonstrated for doxorubicinloaded SA-LUVs injected i.p. and i.v. Interperitoneally injected liposomes exhibited a 2.5fold increase in cell-associated lipid and drug when the cells were pre-labeled with Ab(Figure 3.9). A significant portion of the non-specific cell-associated lipid and drug canprobably be attributed to liposome uptake by macrophages in the peritoneum of P388 tumorbearing animals (see Section 1.3.2). This conclusion is supported by flow cytometric results(Figure 3.9, insert) where P388 cells within the peritoneal cavity were tentatively identifiedbased on the size and granularity characteristics of cultured P388 cells. There wasnegligible non-specific cell-associated fluorescence in this defined cell population.However, in animals that received antibody, 20% of the gated cell population was shown to95be positive for liposome-associated doxorubicin. When SA-LUVs were injectedintravenously, only 3% of the injected dose reached the peritoneal cavity at 24 hours.However, of the drug-loaded SA-LUVs that did access this extravascular site, there was a40% increase in cell-associated lipid and drug compared to animals with tumors that hadnot been pre-labeled with anti-Thy 1.2 Ab (Figure 3.10).In conclusion the results presented in this chapter suggest that two-step targeting ofstreptavidin liposomes has potential for increasing the therapeutic activity of encapsulateddrug. Doxorubicin-loaded targeted liposomes bind to target cells pre-labeled withbiotinylated antibody more effectively than control liposomes in vitro and in vivo. Further,preliminary in vitro drug potency studies suggest that doxorubicin encapsulated in targetedliposomes bound to a non-internalized antigen can be 2 to 3 times more active than drugentrapped in control liposomes. However, the passive targeting of SA-LUVs to theperitoneal cavity was limited due to uptake of liposomes by the phagocytic cells of the RES(see Figure 3.2). Further, in vivo targeting was not as efficient as in vitro targeting.In the following chapters, strategies to surmount these obstacles will be described.The passive targeting of SA-LUVs will be achieved via employment of second generationliposome compositions and suppression of liver phagocytes (Chapter 4). Variablesconnected to the target will be controlled by employing biotin-labeled MLVs as modeltargets with defined binding moieties (biotin-modified phospholipids) in an effort toimprove in vivo targeting of SA-LUVs (Chapter 5).96CHAPTER 4INCREASED EXTRAVASATION OF STREPTAVIDIN LIPOSOMES4.1 IntroductionTargeting liposomes to specific cells or tissues in vivo has been a major objective ofresearch aimed at improving the therapeutic index of liposomal anticancer drugs (Nassanderet a!, 1992; Ahmad et al, 1993; Bankert et a!, 1989; Straubinger et al, 1988; Mori et al,1993). Liposome characteristics that must be optimized for in vivo targeting include targetsite specificity, binding affinity and, most importantly, physical access to the target cellpopulation. Studies using monoclonal antibody-based drug carrier systems suggest thatextravasation, the process whereby material within the blood compartment gains access toan extravascular site, will be dependent on several factors including vascular structure(Kohn et al, 1992; Jam, 1990; Badger et al, 1985) (see Section 1.3.3), antibody circulationlifetimes (Rostaing-Capaillon and Casellas, 1990) as well as antigen expressed on tumorcells (Sung et al, 1992; Thomas et al, 1989). Not unexpectedly, the rate of transport ofmacromolecules from the blood compartment to an extravascular site is dependent on size(Nagy et al, 1989).It is now well documented that liposomes, with no specific surface-associatedtargeting information, can access extravascular sites in diseased tissues such as tumors(Papahadjopoulos et a!, 1991; Bally et al, 1994). This passive targeting process is known tobe dependent on liposome size and circulation longevity (Proffitt et a!, 1983; Gabizon and97Papahadjopoulos, 1988; Gabizon, 1992). It can be suggested on the basis of these data thatany liposomal carrier formulated for in vivo ligand-mediated targeting applications wouldhave to exhibit characteristics of uniform small size and enhanced circulation lifetime.Limitations of liposomal-based targeted delivery systems include the fact that proceduresused to attach targeting ligands to the liposome surface promote aggregation (Loughrey etal, 1990b; Loughrey et al, 1993) and that the targeting ligand itself can result in immunerecognition of the carrier (Aragnol and Leserman, 1986; Longman et al, 1994). Both ofthese attributes promote clearance of intravenously administered targeted liposomes. It iswell established, for example, that aggregated liposomes formed during coupling reactionsused to attach targeting proteins to liposomes are removed rapidly by phagocytic cells of thereticuloendothelial system (RES) (Lougbrey et al, 1990b; Loughrey et a!, 1993). Further,when the targeting ligand employed is a whole antibody, the presence of surface-associatedFc portions increase liposome clearance (Aragnol and Leserman, 1986).Phagocytic cells of the RES are known to be responsible for recognition andremoval of foreign particles, including liposomes (Altura, 1980; Gregoriadis and Ryman,1972; Gregoriadis and Neerunjun, 1974) (see Section 1.3.2). Strategies that attempt to limitRES uptake of targeted liposomes must be developed. One approach involvesincorporation of lipids, such as the ganglioside GMI or polyethylene glycol-modifiedphospholipids, that are known to increase the circulation lifetime of liposomes (Allen andChonn, 1987; Gabizon and Papahadjopoulos, 1988; Klibanov et a!, 1990; Allen andHansen, 1991; Huang et a!, 1992; Papahadjopoulos et a!, 1991) (see Section 1.3.2).However, a balance must be struck between RES avoidance and targeting, because higher98(i.e. 5 mol%) incorporation of PEG-modified lipids can interfere with targeting (Klibanov etal, 1991). A second procedure involves the use of agents that specifically “block”phagocytic cells of the liver, a primary site of accumulation following intravenousadministration of liposomes (Proffitt et a!, 1983; Bally et al, 1990) (see Section 1.3.2).Previous studies by Bally et al (1990), for example, have demonstrated that pre-treatingmice with low dose liposomal doxorubicin results in a 90% reduction in liver accumulationof subsequently administered (i.v.) liposomes. Significant increases in blood levels ofliposomal lipid are observed in these pre-treated animals.Chapter 3 discussed an in vivo liposome targeting approach that involves bindingSA-LUVs to target cells pre-labeled with biotinylated antibody (Loughrey et al, 1990;Loughrey et al, 1993; Longman et a!, 1994). Results in chapter 3 showed that this two-steptargeting procedure was as efficient as approaches involving antibody-coated liposomeswhen labeling cells in vitro (see Section 3.2.3). Further, doxorubicin entrapped in thesetargeted liposomes was shown to be 2 to 3 times more active than non-targeted liposomes(see Section 3.2.3). The studies presented in this chapter examine the use of RES blockadeand/or incorporation of PEG-modified lipids to maximize the circulation lifetimes of SALUVs. These two procedures resulted in substantially increased circulation lifetimes and anincreased tendency for SA-LUVs to escape the blood compartment and achieve enhancedpassive targeting.994.2 Results4.2.1 Influence of Blockade on Clearance and BiodistributionThe studies reported in this chapter investigate two procedures to increase thepassive targeting of protein-coated liposomes, one based on inhibiting liposome uptake byphagocytic cells of the liver and the second based on incorporation of polyethylene glycolmodified phospholipids. It has been shown that uptake of liposomes by cells within theliver can be significantly reduced by pre-treating animals with a low pre-dose of liposomaldoxorubicin (2 mg drug/kg) (Bally et al, 1990) (see Section 1.3.2). This liver blockade isthought to be mediated by accumulation of the cytotoxic drug doxorubicin in liver KUpffercells. Results in Figure 4.1 illustrate how liver blockade promotes increased circulationlifetime and inhibits liver uptake of SA-LUVs. In this study, liposomes with 44 igstreptavidin4tmol lipid, exhibiting a mean diameter of 160 nm (achieved after extrusion ofcrosslinked protein-liposome aggregates) were administered at a lipid dose of 20 mg/kg.Mice were pre-treated with either DSPC/Chol “empty” liposomes (10 mg/kg lipid) orDSPC/Chol liposomal doxorubicin (10 mg/kg lipid; 2 mg/kg drug). These groups arereferred to herein as control and pre-treated animals respectively. At this low lipid pre-dose,“control” animals yielded results identical to animals receiving a pre-dose of saline. Fourhours after administration of SA-LUVs there was greater than 3 times the level of liposomallipid in the plasma of pre-treated animals (Figure 4. 1A). At this same time point there wasa 5-fold reduction in the amount of lipid recovered in the liver of pre-treated animals100Figure 4.1Biodistribution of Streptavidin Liposomes Following Liver Blockade24 h after injection of a predose of liposomal doxorubicin (10 mg/kg lipid, drug/lipid ratio of 0.2[mo]Jmol]) ( • ) or empty liposomes ( 0 ), SA-LUVs were injected via lateral tail vein at a doseof 0.66 imol lipid/mouse (20 mg/kg lipid). The mice were terminated at the indicated times andthe level of streptavidin liposomal lipid was determined (see Section 2.8) for the tissuesindicated. Values shown represent the mean of results from at least 4 animals ± S.E. of the mean(p <0.05).30A225200 TB175150112514001200100080020000 4 8 12 16 20 24TIME (hours)C101(Figure 4. 1B). Consistent with previous results (Bally et al, 1990), accumulation of SALUVs in spleen of pre-treated animals was significantly greater than that obtained forcontrols (Figure 4. 1C).In addition to measuring plasma levels of SA-LUVs, the tendency of theseliposomes to leave the blood compartment and enter an extravascular site was assessed bymeasuring the time dependent accumulation of liposomal lipid in the peritoneal cavity. Thevascular structure lining the peritoneal cavity of control animals has been shown to berelatively impermeable to circulating macromolecules and therefore movement ofliposomes into this compartment represents a stringent test of extravasation. Transport ofmacromolecules (fluorescently labeled dextran) from the blood into the peritoneal cavity inmice has been shown to be dependent on macromolecular size and circulating blood levels(Nagy et al, 1989). The influence of pre-treatment with liposomal doxorubicin (to increasecirculation blood levels) on the transfer of SA-LUVs from the blood compartment to theperitoneal cavity is shown in Figure 4.2. Liposomal lipid, as measured using the nonexchangeable, non-metabolizable lipid label([3H]-cholesteryl hexadecyl ether), could berecovered from the peritoneal cavity by lavage of animals given (i.v.) SA-LUVs. The levelof lipid recovered was 2 to 5 times greater in pre-treated animals than in controls.102Figure 4.2I:I.Accumulation of Streptavidin Liposomes in the Peritoneal Cavity Following LiverBlockadePredoses of liposomal doxorubicin ( • ) and empty liposomes (C) ), and the subsequent SALUV dose were as described in Figure 4.1. Values shown represent the mean of results from atleast 4 animals ± S.E. of the mean (p <0.05).1.751 .501.251 .000.750.500.250.00 I I I0 4 8 12 16 20 24TIME (hours)1034.2.2 Influence of PEG on Liposome Aggregation, Targeting to P388 Cells, and Clearanceand BiodistributionThe second approach used to increase the circulation lifetimes of SA-LUVsinvolved incorporation of polyethylene glycol (PEG)-modified phospholipids. It is wellestablished that PEG-modified lipids, when incorporated into protein-free liposomes,dramatically increase circulation lifetimes (Papahadjopoulos et al, 1991; Allen et al, 1989;Mayhew et al, 1992) (see Section 1.3.2). Preliminary data suggest that incorporation ofPEG-modified lipids may interfere with protein coupling reactions (Klibanov et al, 1991).Therefore, by incorporating PEG2000-DSPE at various mol% in SA-LUVs, these studieswere intended to balance increased circulation lifetimes with optimal targeting. Initialstudies focused on defining the reaction conditions required to achieve 35 to 45 igstreptavidin per jimol lipid. As shown in Table 4.1, protein coupling rates weresignificantly reduced when using liposomes with 5 mol% PEG-PE. However, satisfactory(35-45 .ig) levels of coupling could be achieved with longer incubation times. Remarkably,protein coupling to liposomes with 2 and 5 mol% PEG-PE did not result in liposomeaggregation associated with the coupling procedure. Even at values as low as 1 mol% therewas a significant reduction in the tendency of these liposomes to aggregate during proteincoupling.Since it is well established that decreases in liposome size promote increasedcirculation lifetimes, it was expected that incorporation of PEG-PE would also significantlyincrease retention of SA-LUVs in the plasma compartment. As shown in Figure 4.3A, the104TABLE 4.1Effect of PEG-PE on Coupling of Streptavidm to MPB-PE LUV1PEG-PE INCUBATION SIZE jig STREPTAVIDIN(mol%) TIME (mm) (nm) / jimol LIPID0 5 180 35-451 5 140 35-452 5 120 35-455 50 120 35-45a All incubations were done at 10 mM lipid and at an initial protein-to-lipid ratio of 75 jigstreptavidinljimol lipid. Liposome size was determined by QELS. Extent of coupling wasdetermined as described in Section 2.4.105Cf00>Figure 4.3Biodistribution of Streptavidin Liposomes Incorporating Various Mol % PEG-DSPESA-LUVs containing PEG2000-DSPE at 0, 1,2 and 5 mol% were injected via lateral tail vein at adose of 0.66 jimol lipid/mouse (20 mg/kg lipid). The mice were sacrificed at 24 h and lipidlevels were determined for the tissues indicated (see Section 2.8). Values shown represent themean of results from at least 4 animals ± S.E. of the mean (p <0.05).543202502001 501 00500BO 12% 5106level of SA-LUVs in plasma 24 hours after intravenous administration increases withincreasing PEG-PE levels. Differences in plasma levels observed for liposomes with 0, 1and 2 mol% PEG-PE can be attributed to changes in SA-LUV size. The size of SA-LUVswith 2 and 5 mol% PEG-PE (120 nm diameter) were comparable as judged by QELSmeasurements. Therefore, increased plasma levels observed for liposomes with 5 mol%PEG-PE (Figure 4.3B) are likely due to a direct effect of this lipid on reducing RESaccumulation (Papahadjopoulos et al, 1991; Mayhew et al, 1992).The effect of various mol% PEG-PE on the binding of SA-LUVs to P388 cells prelabeled with a biotinylated antibody is shown in Figure 4.4. These results were obtainedusing a two-step targeting approach (Loughrey et al, 1993; Longman et al, 1994) (seeSection 2.7) in which P388 cells were first incubated (4°C, 30 mm) with and withoutbiotinylated anti-Thy 1.2 antibodies. After extensive washing, the cells were resuspended inmedia and SA-LUVs were added. Binding of liposomal lipid to these cells was detenninedafter a 30 mm incubation and extensive washing, as described in Section 2.7. The datademonstrates that incorporation of 5 mol% PEG-PE significantly reduced SA-LUVsassociation with P388 cells pre-labeled with biotinylated anti-Thy 1.2 Ab.4.2.3 Influence of Blockade and PEG on Clearance and BiodistributionThe characterisation studies summarized above in Section 4.2.2 suggest thatincorporation of 2 mol% PEG2-DSPE (52 mol% DSPC, 45 mol% Chol and 1 mol%MPB-PE) in liposomes used for covalent attachment of streptavidin was optimal with107Figure 4.4Quantification of Cell-Associated Lipid after Targeting of Streptavidin LiposomesIncorporating Various Mol% PEG2000-DSPE to P388 Cells In VitroSA-LUVs (35-45 jig streptavidin/jimol lipid) incorporating PEG2000-DSPE were prepared asdescribed in Section 2.2-2.4. P388 cells (l0) were incubated with (solid bars) or without(empty bars) biotinylated anti-mouse Thy 1.2 Ab (10 jig) for 30 minutes at 4°C. Cells werewashed with PBS prior to addition (2 mM fmal concentration) of SA-LUVs with various mol%PEG2oo-DSPE. After a 30 minute incubation at 4°C, the cells were washed and cell associatedlipid was determined as detailed in Section 2.8.12(f)_J 11-JLii 100o 875004—30E210I108respect to achieving efficient protein coupling with no increase in vesicle size and noreduction in target cell association as mediated through binding to a biotinylated antibody.Using this liposomal formulation, studies combining the influence of PEG-PE incorporationand blockade of liver phagocytic cells were initiated. The results, shown in Figure 4.5,indicate that a combination of these strategies resulted in remarkably improved circulationlifetimes (Fig. 4.5A) and significant reductions in liver accumulation (Fig. 4.5B). At 24hours, for example, there was 15 times more liposomal lipid present in the plasma of micegiven SA-LUVs incorporating 2 mol% PEG-PE when those mice were pre-dosed withliposomal doxorubicin (2 mg/kg drug) compared to liposomes incorporating 0 mol% PEGPE in control mice. Further, the lowest level of lipid accumulation in liver (less than 6% ofthe injected dose at 24 hrs) was achieved when using 2 mol% PEG-PE containing SALUVs in combination with blockade of liver phagocytes.The impact of combining both RES blockade and incorporation of PEG-PB on thelevel of liposomal lipid accumulation within the peritoneal cavity 24 hrs after i.v.administration is shown in Figure 4.6. The results demonstrate that the greatest effect onliposome movement from the plasma compartment to this extravascular site occurs only asa result of blockade of liver phagocytes. Tn both the presence and absence of 2 mol% PEGPB there was a 4 to 5 fold increase in delivery of SA-LUVs to this site when animals werepre-treated with liposomal doxorubicin.109Figure 4.5Biodistribution of Streptavidin Liposomes Incorporating 2 Mo! % PEG-DSPEFollowing Liver BlockadeAt 24 h after injection of a low dose liposomal doxorubicin (10 mg/kg lipid, drug/lipid ratio of0.2 [mol/mol] (solid symbols) or saline (empty symbols), SA-LUVs containing 0 mol% (•with blockade; 0 , without blockade) or 2 mol% (• , with blockade; D , without blockade)PEG2-DSPE were injected via lateral tail vein at a dose of 0.66 jamol lipid/mouse (20 mg/kg).The mice were sacrificed at the indicated times and lipid levels of this subsequent dose weredetermined for the tissues indicated (see Section 2.8). Values shown represent the mean from atleast 4 animals ± S.E. of the mean (p <0.05).3025..200o 15.110502001751501 251000 4 8 12 16 20 24TIME (hours).A110Figure 4.6(2>0.000% PEG—PE 2% PEG—FEAccumulation of Streptavidin Liposomes Incorporatmg 2 Mo) % PEG200-DSPE in thePeritoneal Cavity Following Liver BlockadePredoses (solid fills, with blockade; empty fills, without blockade) and subsequent doses of SALUVs incorporating 0 mol% or 2 mol% PEG2000-DSPE were as described in Figure 4.5 (p <0.05).1 .751 .501 .251 .000.750.500.25III1114.2.4 Influence of Tumor Presence on Clearance and BiodistributionThe tendency of SA-LUVs to exit the blood compartment was also evaluated inmice bearing P388 cells grown in the peritoneal cavity of BDF1 mice. It should be notedthat the clearance and biodistribution characteristics of SA-LUVs in CD1 mice (used fordata in Figures 4.1-4.6) and BDF1 mice were comparable. The results of this study, whereSA-LUVs were given (i.v.) to BDF1 mice with and without a tumor burden (see Section2.8), are shown in Figures 4.7 and 4.8. The data in Figure 4.7 illustrates three importantpoints. First, in mice with P388 tumors, 24 hrs after i.v. administration the level ofliposomal lipid in blood (Figure 4.7A) is highest and the level in the liver (Figure 4.7B) islowest when animals have been given SA-LUVs with 2 mol% PEG-PE. In comparison toanimals that received SA-LUVs which were prepared in the absence of PEG-PE andinjected with no additional attempts to reduce the size of the aggregated liposomes, therewas more than a 100-fold increase in the level of liposomal lipid in the plasma at 24 hours.Second, consistent with results in Figure 4.6, in both the presence and absence of tumor thelevel of liposomal lipid in the blood was increased and the level in the liver decreased whenanimals had been pre-treated with a low dose liposomal doxorubicin. Third, the level ofliposomal lipid in the plasma at 24 hours was significantly less when those animals hadestablished P388 tumors. This is consistent with results from other laboratories that suggestthat tumor cells secrete vascular permeability factors that promote macromolecule transportacross blood vessels (Senger et al, 1983; Kohn et a!, 1992). This would be expected toincrease liposomal extravasation rates.112Cl)C0>SAFigure 4.7Influence of Blockade, Incorporation of 2 Mol% PEG2000-DSPE on Circulation Lifetimesand Liver Accumulation of Streptavidin Liposomes Given (i.v.) to P388 Tumour-BearingAnimalsThree days after i.p. injection of 106 P388 cells or saline, animals were given (i.v.) a predose ofliposomal doxorubicin (10 mg/kg lipid, drug/lipid ratio of 0.2 [mollmol] (solid fills) or saline(empty fills). SA-LUVs containing 0 mol% or 2 mol% PEG2000-DSPC were injected (i.v.) 24 hlater at a dose of 0.66 jimol lipid/mouse (20 mg/kg). The mice were sacrificed at 24 h and lipidlevels of the subsequent injection were determined for the blood and liver, as described in theSection 2.8. Values shown represent the mean of results from at least 4 animals ± S.E. of themean (p <0.05).17.515.012.510. 501 00500B+tumor +tumor —tumor0% PEG 2% PEG 2% PEG113The presence of tumor cells in the peritoneal cavity significantly increased the levelof SA-LUV accumulation in the peritoneal cavity (Figure 4.8). There was a 50-foldincrease in the amount of liposomal lipid recovered from the peritoneal cavity of micebearing an established P388 tumor compared to non-tumor-bearing animals. In both thepresence or absence of blockade of liver phagocytes, as much as 7% of the injected dose ofSA-LUVs was isolated following peritoneal lavage. Overall, delivery of SA-LUVs to thisextravascular site was increased more than 90 fold by incorporation of PEG2000-DSPE (2mol%) and utilising animals with established tumors.4.3 DiscussionPrevious studies have focused on developing a liposome targeting approachwhereby SA-LUVs bind to target cells pre-labeled with a biotinylated antibody (Loughrey etal, 1990; Loughrey et al, 1993; Longman et al, 1994). An advantage of this two-stepprocedure is that a single liposomal carrier can be developed for targeting to a variety ofpre-targeted biotinylated ligands. In chapter 3 it was shown that the two-step targetingapproach was as efficient as single-step approaches using antibody-coated liposomes (seeSection 3.2.3). Further, the therapeutic advantage of targeting SA-LUVs with entrappeddoxorubicin has been demonstrated (in vitro) relative to a non-targeted liposomaldoxorubicin (see Section 3.2.3). Finally, the in vivo targeting of streptavidin liposomes to atarget in an extravascular site was achieved (see Section 3.2.4).114Figure 4.8Influence of Liver Blockade, Incorporation of 2 Mol % PEGzooo-DSPE on theAccumulation of Streptavidin Liposomes in the Peritoneal Cavity of P388 Tumour-Bearing AnimalsThree days after i.p. injection of 106 P388 cells or saline, animals were given (i.v.) a predose ofliposomal doxorubicin (solid fills) or saline (empty fills), as described in Figure 4.7. SA-LUVscontaining 2 mol% PEG-PE were injected (i.v.) 24 h as described in Figure 4.7.>302520151050+tumor —tumor115The studies presented in this chapter were aimed at optimizing the pharmacokineticproperties of SA-LUVs with respect to maximizing circulation lifetimes as well as thetendency of these liposomes to exit the blood compartment and access an extravascular site(passive targeting). The results demonstrate that increased circulation lifetimes can beachieved using procedures that restrict liposome uptake by the RES. Improved delivery to adefined extravascular compartment, the peritoneal cavity, is achieved as a consequence ofthe increase blood levels. It can be concluded on the basis of these data that SA-LUVs canbe designed effectively to access potential target cells outside the blood compartment.Design characteristics that are considered most relevant are discussed here.Studies examining the natural tendency of liposomes to accumulate in sites of tumorgrowth (passive targeting) suggest that long circulation lifetimes are required for optimaldelivery (Proffitt et al, 1983; Gabizon and Papahadjopoulos, 1988; Gabizon, 1992). Chapter4 demonstrates that the circulation lifetime of SA-LUVs can be increased by use ofprocedures that minimize liposome uptake by phagocytic cells of the RES. SA-LUVsprepared with 2 mol% PEG-PE exhibit increased circulation longevity (see Figure 4.3).Furthermore, pre-administration of liposomal doxorubicin blocks the ability of phagocyticcells in the liver to accumulate liposomes. Reduced liver uptake resulted in increasedcirculation lifetimes for SA-LUVs (Figure 4.1) and increased the extent to which theyaccumulated in a defined extravascular space (Figure 4.2).A second design characteristic to increase liposome access to an extravascular site issize. It is believed that movement of liposomes from the blood compartment into tumorsoccurs as a consequence of tumor-mediated changes in blood vessel structure (see Section1161.3.3). These blood vessels are more permeable to circulating macromolecules thancomparable vessels in normal tissue. Leakage may occur through pores or fenestrationswithin the endothelial cells lining the blood vessels (Senger et al, 1983; Jam, 1990; Kohn etal, 1992). Alternatively, transport may involve trancytosis through endothelial cells viacytoplasmic vesicles (Senger et al, 1983; Jam, 1990; Kohn et al, 1992). Liposome transportthat is mediated by either of these processes will be dependent on particle size. It has beenshown here (see Figure 4.3) and elsewhere (Nagy et al, 1989) that liposome movement fromthe blood compartment to an extravascular site is increased as vesicle size is decreased.Section 4.2.2 demonstrates that liposomes size can be easily controlled during the covalentattaching of targeting proteins to liposomes by incorporating small quantities (2 mol%) ofPEG-modified PE. At this level, the presence of PEG-PE did not effect the rate ofstreptavidin coupling (Table 4.1) or the extent to which the resulting liposomes could bindto biotinylated target cells (Figure 4.4).In conclusion, the results presented in chapter 4 demonstrate that SA-LUVs havepotential for delivery of therapeutic agents to specific cells present in tumors. IncorporationofPEG2000-DSPE (2 mol%) prevents liposome aggregation during coupling. The resultingprotein-coated liposomes exhibit increased circulation lifetimes that can be further increasedby use of pre-treatments that incapacitate phagocytic cells present in the liver. Incombination, the approaches used here to maximize the circulation lifetime of SA-LUVsresulted in a 90 fold increase in the level of liposome accumulation in a definedextravascular site, the peritoneal cavity. In animals bearing an intraperitoneal tumor, almost1177% of the SA-LUV dose could be recovered from the peritoneal cavity 24 Irs after i.v.administration.118CHAPTER 5TARGETING OF STREPTAVIDIN LIPOSOMES TO BIOTIN-MLVs5.1 IntroductionTargeting of liposomes to specific targets in vivo has been the aim of many studiesdirected at improving the therapeutic index of liposomal drugs (see Section 1.1). Thetargeting of liposomes to cells has been achieved using many targeting molecules, such asmonoclonal antibodies (Huang et al, 1980; Martin et al, 1981; Wolff and Gregoriadis,1984), avidin and streptavidin (Rosenberg et al, 1987; Loughrey et al, 1993; Longman Ctal, 1994), transferrin (Vidal et al, 1987; Stavridis et al; 1986), as well as other ligands thatbind to a specific cell surface (see Section 1.2.4). In vitro liposome studies have establishedthat the high-affinity interaction between targeting ligand and receptor must be maintainedfor optimal targeting. This involves several factors such as the type of chemicalmodification required and the nature of the coupling techniques used, such that the bindingof the targeting ligand to its receptor is not compromised. In addition, liposome surfacecharacteristics must be such that non-specific binding to cell surfaces is minimized andtarget ligand/receptor interaction is not inhibited. This may be achieved, for example, byincorporating PEG-modified lipids into the ligand-coated liposome bilayer (see Section4.2.2). The binding of liposomes to the target cell population in vitro can be increased byaddition of more targeting ligands to the liposomal surface (Huang et a!, 1980; Maruyama etal, 1990; Nässander et al, 1992). Further, if the targeted liposome is to be internalized along119with the target receptor, then vesicle size must be such that normal internalization processescan occur. Transferrin bound to the transferrin receptor, for example, is internalized byclatherin coated endocytic vesicles that exhibit a mean diameter of 100 nm (Goldstein et al,1979; Steinman et a!, 1983). Liposomes targeted to this receptor should not exceed thissize.In addition to maintaining cell surface specificity, liposomes used for in vivotargeting must be able to access the target cell population. Successful in vivo targeting ofliposomes has been achieved using monoclonal antibodies specific for lung endothelial cells(Maruyama et al, 1990b) as well as for tumor cells that seed within the lung (Ahmad et al,1993) (see Section 1.4.2). A therapeutic advantage has been demonstrated for the latterwhen employing doxorubicin-loaded antibody-coated liposomes (Ahmad et al, 1993). If thetarget cell population resides outside the vascular compartment, then the in vivo targetedliposomal formulations must exhibit characteristics that promote extravasation to thedisease site. Although the mechanisms responsible for this extravasation process are notwell understood, it is established that liposome size and circulation lifetime are criticaldeterminants (Proffitt et al, 1983; Gabizon and Papahadjopoulos, 1988; Gabizon, 1992). Ithas been documented that liposomes lacking surface-associated targeting information canaccess extravascular sites in diseased tissues such as tumors (Proffitt et al, 1983; Gabizonand Papahadjopoulos, 1988; Papahadjopoulos et a!, 1991; Wu et a!, 1993; Bally et a!,1994). Optimal delivery of liposomes and associated contents requires use of liposomesthat: 1) exhibit mean diameters of less than 200 nm; 2) are retained in the plasma120compartment for extended periods and 3) are capable of retaining associated contentsfollowing i.v. administration.Chapter 3 demonstrated an in vivo liposome targeting approach that involvesbinding SA-LUVs (injected intravenously) to target cells pre-labeled with biotinylatedantibody (residing in the peritoneal cavity) (Longman et al, 1994). Results in Chapter 3indicated that this two-step approach was as efficient as approaches involving antibody-coated or receptor-coated liposomes when labeling cells in vitro. In Chapter 4, the use ofRES blockade and incorporation of PEG-modified lipids resulted in substantially increasedcirculation lifetimes and an increased tendency for SA-LUV to escape the bloodcompartment. Section 3.2.4 demonstrated that SA-LUVs injected intravenously couldeffectively target to P388 cells residing in the peritoneal cavity, provided the cells were prelabeled, in vivo, with a biotinylated cell-specific antibody. The specificity observed overcontrol (protein free liposomes) was, however, only 3-fold in comparison to in vitro studiesthat demonstrated almost a 20-fold increase in specificity (see Section 3.2.3). The decreasein cell labeling was a consequence of several factors including the presence of macrophagesand other host-derived inflammatory cells within the peritoneal cavity that non-specificallyinteract with the liposomes. Finally, it was not clear from Chapter 3 whether in vivobinding of the biotinylated antibody to the target cell population was as efficient as thatobserved in vitro.The many complicating factors described above make it difficult to optimize andinterpret in vivo targeting of liposomes. In Chapter 5, a model target system has beendeveloped based on the targeting of SA-LUVs to biotin-labeled MLVs. By substituting121biotin-labeled MLVs for a target cell population, we have been able to eliminate variablesrelated to antigen density, antibody/antigen turnover and cell number. The studies presentedhere demonstrate that biotin-labeled MLVs are stable in vivo, retained well within theinjection sites, can effectively bind intravenously administered SA-LUVs and are useful as amodel target for which to characterize SA-LUV binding.5.2 Results5.2.1 In Vitro Characterization of Binding of Streptavidin Liposomes to Various BiotinMLVsThe studies described here are based on the use of biotin-labeled MLVs as a modeltarget that can be specifically labeled with SA-LUVs. SA-LUV binding to biotin-MLVswas assayed using a procedure that separates MLVs from LUVs by low speedcentrifugation, as shown in Table 5.1. Briefly,[3H]-labeled SA-LUVs, spun at 1600 g for10 mm, do not pellet. In contrast,[14C]-labeled biotin-MLVs pellet efficiently with greaterthan 98% of the lipid recovered in the pellet after 3 washes. When protein-free[3H]-LUVswere mixed with[14C]-labeled biotin-MLVs or, conversely, when[3H]-SA-LUVs weremixed with [‘4C]-MLVs (no biotin) there was complete separation of the MLVs (98%recovered in the pellet) and LUVs (98% recovered in the supernatant). Specificity ofbinding is demonstrated when SA-LUVs are mixed with biotin-MLVs, where greater than98% of both the MLVs and LUVs were found in the washed pellet.122TABLE 5.1Separation of LUVs and MLVSaSAMPLE SUPERNATANT (pmol lipid) PELLET (pmol lipid)MLVs SA-LUV MLVs SA-LUV(14C) (3H) (14C) (3H)SA-LUVs 0.000 0.49 1 0.000 0.000(0.5 j.imol lipid)biotin-IvlLVs 0.009 0.000 19.7 13 0.000(20 j.tmol lipid)LUVs 0.000 0.493 19.801 0.000(0.5 jimol) +biotin-MLVs(20 jimol)SA-LUVs 0.000 0.489 19.752 0.000(0.5 gmol) +MLVs(20 i.imol)SA-LUV 0.000 0.009 19.846 0.490(0.5 j.tmol) +biotin-MLVs(20 j.tmol)aMl incubations were done at RT for 30 mm in a 1.0 ml volume, at which time sampleswere spun at 3000 rpm (1600g) for 10 mm. MLV and LUV lipid present in the pellet andsupernatant were determined as described in Section 2.12.123Using the centrifugation assay described above, initial studies assessed the effect ofaltering the biotin-labeled phospholipid that was incorporated into the MLVs. The bindingof SA-LUVs to biotin-MLVs was determined in the presence and absence of 10% serum.Three different biotinylated lipids were used including: 1) an unsaturated biotin-labeledphospholipid derived from DOPE (B-DOPE); 2) a saturated biotin-labeled phospholipidderived from DSPE (B-DSPE) and 3) a saturated biotin-labeled phospholipid where thebiotin was coupled to DSPE via a six carbon hexanoyl spacer arm (BAH-DSPE) (forstructures, see Figure 2.1). The biotinylated lipid was always incorporated at the level of 1mol% in the MLVs and SA-LUV binding was determined as a function of MLVconcentration (Figure 5.1). Three points can be made on the basis of the data in Figure 5.1.First, SA-LUV binding to MLVs containing B-DOPE is significantly inhibited by thepresence of serum. This may be due to enzymatic cleavage of the biotin headgroup orexchange of the biotinylated lipid. Second, in the absence of serum the binding of SALUVs to MLVs containing B-DSPE is significantly greater than results obtained usingMLVs with B-DOPE. Third, binding of SA-LUVs to biotin-MLVs was most efficientwhen the MLVs incorporated biotinoylaminohexanoyl DSPE (BAH-DSPE). Theconcentration of MLVs required to obtain 50% binding of the SA-LUVs was decreased 3-fold when the biotin was attached via the 6 carbon spacer arm in BAH-DSPE. Targeting toeither of the DSPE-modified biotin lipids was not effected by the addition of serum.Since the data in Figure 5.1 suggests that the most efficient binding of SA-LUVoccurs when using MLVs with BAH-DSPE, the remaining studies focused on targeting to124Figure 5.10z0C)0Influence of Serum on the Binding of Streptavidin Liposomes to MLVs IncorporatingDifferent Biotin-Labeled PhospholipidsSA-LUVs (0.2 mM) were incubated at RT for 30 mm with MLVs incorporating B-DOPE (v),B-DSPE (•) or BAH-DSPE (B) in the presence (solid symbols) or absence (empty symbols) of10% normal mouse senim. Quantification of bound SA-LUV was determined as in Section2.12.2001801 60140120100806040200.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0MLVs (mM LIPID)125these MLVs, hereafter referred to as BAH-MLVs. In an effort to establish the bindingcapacity of BAH-MLVs, a study assessing binding as a function of SA-LUV concentrationwas completed. The results, shown in Figure 5.2, were obtained in the presence of serumusing an MLV concentration of 20 imol. Under these conditions a maximum of 1.3imoles of SA-LUV lipid can bind to 20 i.imoles of BAH-MLV lipid in a 1 ml assayvolume. Further, 100% SA-LUV binding was observed at concentration up to 0.8 mM.The specificity of the binding reaction between SA-LUVs and BAH-MLVs isillustrated by the competitive inhibition studies summarized in Figure 5.3. The resultsdemonstrate that addition of free biotin (Figure 5.3A) completely inhibits binding atconcentrations of 1.5 mlvi. This level of biotin is such that there are approximately 2.5 freebiotins per streptavidin molecule bound to the liposomes. This result is consistent withprevious studies demonstrating that streptavidin bound to the liposome is capable of binding2 to 3 moles biotins per mole streptavidin (Loughrey et al, 1993). Further, the results inFigure 5.3B demonstrate that the binding reaction between radiolabeled SA-LUVs andBAH-MLVs is readily inhibited by addition of excess levels of non-radiolabeled SA-LUVs.A 5-fold decrease in binding is observed when the radiolabeled liposomes are diluted 5-foldwith “cold” SA-LUVs.5.2.2 In Vitro Binding of Streptavidin Liposomes (recovered from plasma) to Biotin-MLVsThe results presented to this point demonstrate that BAH-MLVs can serve as anappropriate target for SA-LUVs. The reaction is specific (biotin and/or “cold” SA-LUV126zCC?Figure 5.2Binding of Streptavidin Liposomes to BAH-MLVsBAH-MLVs (20 mM) were incubated at RT for 30 mm with various concentrations of SALUVs. Quantification of bound SA-LUV was determined as in Section 2.12.20001 8001 6001 4001 2001 00080060040020000 400 800 1200 1600PEG—SA---LUV ADDED (,uM)2000127Figure 5.3Competitive Inhibition of Streptavidin Liposome Binding to BAH-MLVs with Free Biotinor Streptavidin LiposomesVarious concentrations of free biotin or cold SA-LUV were incubated with BAH-MLVs (20mM) and [H3j-SA-LUV (0.5 mM) at RT for 30 mm. Quantification of bound SA-LUV wasdetermined as in Section o 0.075__:& 0.0500.0250.0000.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75BIOTIN ADDED (nmol)0.2500.225> 0.2000.175>0.150II 0.125I 0.1000.075.. 0.050CD0.0250.0000.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0COLD PEG—SA—LUV ADDED (umol)128addition blocks binding), there is negligible background binding (binding to biotin-freeMLVs), is not inhibited by addition of serum proteins, and can be controlled by the natureof the biotinylated phospholipid incorporated into the MLV. The following studies addresswhether this model targeting system can be used in vivo. First, it is important todemonstrate that SA-LUVs recovered from the plasma compartment following i.v.administration can still bind BAH-MLVs. This study, illustrated in Figure 5.4, used SALUVs that were retained in the circulation of mice following i.v. administration of SALUVs at a dose of 100 mg lipid/kg. Plasma was isolated at 1 and 24 hrs. Subsequently, 200jimol of isolated SA-LUVs were incubated with various concentrations of BA}{-MLVs.The binding of SA-LUVs isolated 1 hr after administration was identical to controlliposomes that had not been injected. There was a slight, yet significant, decrease in SALUVs binding when the liposomes were recovered 24 hrs after administration. As indicatedin Figure 5.4, about 2-fold more MLV lipid was required to achieve 50% binding of SALUVs that were isolated from plasma 24 hr after administration when compared withcontrols. It is important to note that the binding to BAH-MLVs of SA-LUVs isolated fromanimals maintained in an avidin-enriched, biotin-depleted diet for 5 days was essentiallyidentical to the data shown in Figure 5.4.It is also important to demonstrate that the target, BAH-MLVs, retained bindingactivity after in vivo administration and that this target could be retained within a definedsite after administration. For these studies BAH-MLVs or control MLVs were injected i.p.and the level of liposomal MLV lipid and associated SA-LUV binding capacity wasdetermined. The results, summarized in Table 5.2, indicate that for both control MLVs and129Cz0>C?Figure 5.4Influence of In Vivo Incubation of Streptavidin Liposomes in the Mouse BloodCompartment on Binding to BAH-MLVs In VitroSA-LUV was injected via lateral tail vein at a dose of 3.29 imol lipid/mouse (100 mg/kg lipid).The mice were terminated at 1 and 24 hrs after injection and liposomal lipid in the plasma wasrecovered and measured as in Section 2.8. Recovered SA-LUVs (0.5 mM) (0, 0 hr; D, 1 hr;V, 24 hrs) were incubated with BAH-MLVs (20 mM) at RT for 30 mm. Quantification ofbound SA-LUV was determined as in Section 2.12.___2001 801 601401 201008060402000 5 10 15 20 25 30 35 40MLVs (mM LIPID)130TABLE 5.2Retention of Vesicles Injected in the Peritoneal Cavity% Recovery in peritoneal Concentration of BAH-MLV toLiposomes cavity 24 h after administration achieve 50% of maximum- tumour SA-LUV bindingMLV 85-90% N.A.BAH-MLV 85-90% 2.5 mMLUV 2.2% N.A.BAH-MLVs, MLVs or LUVs (20 imol lipid) were injected i.p. in mice. Vesicles wererecovered by lavage 24 hrs later, and assayed for their ability to bind SA-LUVs (see Section5.2.2).131BAH-MLVs greater than 85% of the injected (i.p.) dose of lipid is retained within theperitoneal cavity 24 hrs after administration. Further, the ability of SA-LUVs to bindisolated BAH-MLVs is identical to that observed for BAH-MLVs that were not injected.5.2.3 In Vivo Binding of Streptavidin Liposomes to Biotin-MLVsThe studies summarized in Figure 5.4 and Table 5.2 clearly indicate that it isfeasible to target SA-LUVs to BAH-MLVs in vivo. Several approaches, summarizedbelow, were taken to demonstrate this. In general, SA-LUVs (or control, streptavidin-free,LUVs) were administered either i.v. or i.p. at a dose of 3.29 iimoles per mouse (100 mglipid/kg) 1 hr after an i.p. injection of BAH-MLVs (or control, biotin-free, MLVs) weregiven at a lipid dose of 20 tmoles per mouse in an injection volume of 0.5 ml. The studieswere conducted using animals with or without established ascitic tumors. The presence ofan ascitic tumor has been shown to increase access of circulating macromolecules (Nagy etal, 1989; Flessner et al, 1985) and liposomes (Bally et al, 1994) to the peritoneal cavity.The assay for targeting was based on measuring: 1) the amount of LUVs recovered in theperitoneal cavity 24 hrs after administration and 2) the amount of recovered SA-LUV thatcould be pelleted at 1600 g after a 10 minute centrifugation. Centrifugation pellets bothMLVs and cells present in the peritoneal cavity, and therefore flow cytometry was utilizedto distinguish which population bound SA-LUVs. Finally, it should be noted that thesestudies used SA-LUVs or control-LUVs that were loaded with the anticancer drugdoxorubicin. The use of entrapped doxorubicin is based on the fact that: 1) doxorubicin is132a convenient fluorescent molecule that can readily be evaluated by flow cytometry and aquantitative fluorescence drug assay; 2) the presence of doxorubicin in the LUVs increasesliposome circulation longevity (Bally et al, 1990); and 3) the primary research objective isto develop procedures that result in more specific delivery of anticancer drugs to sites oftumor growth.Results from these initial in vivo targeting studies, summarized in Table 5.3,illustrate several points. First, following i.p. injection of LUVs in mice pre-injected i.p.with BAH-MLVs (see Table 5.3, Group A) one observes significant retention of injectedLUV lipid only when LUVs are coated with streptavidin. In the absence of tumor, almost40% of the injected SA-LUV dose was recovered 24 hr following administration, whereasless then 1% of the control (no streptavidin) LUV dose was recovered. Retention wasdependent on use of MLVs with surface-associated biotin. Furthermore, flow cytometricanalysis revealed that pelleted SA-LUVs were associated with MLVs and not peritonealcells. Flow cytometric analysis of the pellet identified peritoneal cells on the basis of sizeand granularity. Peritoneal macrophages were more specifically identified using a F1TCanti-Mac 1 antibody.The second point is that the presence of an established tumor in the peritoneal cavitygreatly increases the retention of i.p. administered LUVs, even with non-targeted liposomes(see Table 5.3, Group B). This is consistent with studies that suggest that establishedperitoneal tumors can block lymphatic drainage, the primary mechanism responsible forelimination of liposomes from the peritoneal cavity (Feldman, 1975; Hirano and Hunt,1985). Retention of LUVs injected i.p., for example, increased from 0% of injected dose in133TABLE5.3InVivoTargetingof SA-LUVstoBAH-MLVsThreedaysafteri.p.injection of106P388cells(whereindicated),MLVswereinjectedi.p.atadoseof20jimollipid/mouse.SA-LUVswereinjectedi.p.orvialateraltailveinatadoseof3.29.jmollipid/mouse(100mg/kglipid).Themicewereterminatedateither1or24hrs(forSA-LUVsinjectedi.p.ori.v.,respectively)andliposomallipidintheperitonealcavitywasrecoveredandmeasuredasinSection2.8.QuantificationofboundSA-LUVwasdeterminedasinSection2.12.GroupLIJVRouteofMLVTypeAsciticTotalLUVLipidTotalLUVFoldImprove-TypeAdminis-(injectedi.p.)TumourRecoveredinp.c.LipidPelletedmentOvertration(nmol)(nmol)ControlA.SA-LUVsi.p.BAH-MLVs-1138±170502±8617.9-foldSA-LUVsi.p.MLVs-32±2028±18-LUVsi.p.BAH-MLVs-0±00±0-B.SA-LUVsi.p.BAH-MLVs+1381±1001249±2078.4-foldSA-LUVsi.p.MLVs+499±66160+22-LUVsi.p.BAH-MLVs+540±25146±39-C.SA-LUVsi.v.BAH-MLVs-27.3±0.815.7±1.05.8-foldSA-LUVsi.v.MLVs-26.8±1.43.2±0.4-LUVsi.v.BAH-MLVs-26.0+0.12.2±0.8-D.SA-LUVsi.v.BAH-MLVs+121.1±28.182.8±14.96.2-foldSA-LUVsi.v.MLV+111.5±13.711.8±2.0-LUVsi.v.BAH-MLVs+157.7±6.717.8±11.1-the absence of tumor to almost 17% in the presence of tumor. Of these retained LUVs, only28% could be pelleted, and flow cytometric analysis suggested that pelleted LUVs primarilyassociated with peritoneal macrophages. In contrast, under conditions where targetingcould be achieved (i.e. SA-LUVs injected i.p. into mice pre-injected i.p. with BAH-MLV5)42% of the injected dose was retained in the peritoneal cavity in the presence of tumor,comparable to that observed in the absence of tumor (35%). Remarkably, more than 90%of the retained SA-LUVs were pelleted, and flow cytometric analysis revealed that pelletedSA-LUVs were predominantly associated with BAH-MLVs. The amount pelleted (1.25jimoles) is equivalent to the maximum amount of SA-LUVs that could be bound to 20jimoles of BAH-MLVs (based on the in vitro data shown in Figure 5.2).Two additional points worth noting concern the targeting of i.v. administered SALUVs to BAH-MLVs present in the peritoneal cavity (see groups C and D, Table 5.3). Asdocumented elsewhere (Bally et a!, 1994), the presence of an established tumor increasesthe amount of lipid recovered in the peritoneal cavity following i.v. injection of LUVs.Secondly, pre-injection of MLVs (in comparison with no injection) engenders increasedaccumulation of i.v. injected LUVs. This could be a result of saturation of the peritonealmacrophages with MLVs. Thirdly, in both the presence and absence of tumors, preinjection of BAH-MLVs (in comparison with control MLVs) did not promote delivery ofi.v. injected SA-LUVs to the peritoneal cavity, although there was approximately a 6-foldincrease in the amount of pelleted SA-LUVs when BAH-MLVs were present in the cavity.As previously noted for LUVs and SA-LUVs administered i.p., the majority of pelletedLUVs and SA-LUVs found under control (non-targeting) conditions were cell associated.135As a final test is to whether SA-LUVs can be targeted in vivo to regionally localizedBAH-MLVs, we chose to examine delivery of liposomal lipid to solid tumors derivedfollowing subcutaneous injection of Lewis Lung sarcoma cells. Two tumors wereestablished per mouse, ranging in size from 200 to 500 mg. One tumor was given anintratumoral injection of BAH-MLVs while the contralateral tumor was injected withcontrol MLVs. One hr after injection of MLVs the animals were given an i.v. injection ofeither SA-LUVs or control LUVs. One day later tumors were removed and the amount ofLUV lipid per g tumor was determined. The results, shown in Figure 5.5, clearlydemonstrate that the presence of BAH-MLVs within the solid tumor promoted delivery ofSA-LUVs. Compared to controls, consisting of tumors injected with control MLVs oranimals treated with protein-free LUVs, there was at least a 2-fold increase in tumordelivery when the targeting reaction between LUVs and MLVs was facilitated bystreptavidin and biotin.5.3 DiscussionChapter 3 demonstrated that doxorubicin-loaded SA-LUVs injected i.v. can betargeted to cells pre-labeled with a biotinylated antibody. Chapter 4 showed that throughthe use of RES blockade and incorporation of PEG-modified lipids, SA-LUVs demonstrateenhanced circulation lifetimes and an increased tendency to escape the blood compartment,resulting in increased passive targeting of SA-LUVs within a potential target site. Furtheroptimization of this targeting approach requires a better understanding of how target136Figure 5.5Targeting of Streptavidin Liposomes (Injected i.v.) to BAH-MLVs at a Subcutaneous SiteMice were injected subcutaneously in the right and left thighs with 3x105 Lewis lung cells andallowed to grow 10-13 days, at which time BAH-MLVs and MLVs (20 pmol lipid) wereinjected subcutaneously into the right and left tumour sites, respectively. One hr later, animalswere injected (i.v.) with SA-LUV or LUV (3.29 .imol lipid/mouse; 100 mgfkg lipid).Combinations included BAH-IvlLVs and SA-LUVs (CI), MLVs and SA-LUVs (•) and BARMLVs and LUVs (a). 24 hrs later the animals were terminated and lipid levels at thesubcutaneous tumour sites were determined as described in Section 2.8. Values shown representthe mean from at least 4 animals ± S.E. of the mean (p <0.05).3000250137variables such as target antigen number and target cell number influence SA-LUV deliveryand localization within a target site. In order to achieve this understanding we havedeveloped a model target system based on the high affinity binding reaction between SALUVs and biotin incorporated on the surface of a large multilamellar vesicle. Initial studiescharacterizing this model system have been described in this chapter and the resultsobtained lead to several general conclusions regarding the potential for targeting liposomaldrug carriers to extravascular sites.The model approach described here used SA-LUVs as the targeted drug carrier andbiotin-labeled MLVs as the target. Chapters 3 and 4 described the optimization of SALUVs for systemic delivery of anticancer drugs and concluded that maintenance ofliposome size and drug retention characteristics was required for optimal target specificdelivery. As demonstrated here (see Table 5.2) and elsewhere (Weinstein, 1987) large (>2jim) multilamellar liposomes are retained well within the injection site after regionaladministration. The target MLVs were optimized for binding efficiency through theselection of appropriate biotinylated lipids incorporated in the MLVs. Figure 5.1demonstrated that biotin, linked to a saturated phospholipid via a six carbon spacer arm,was optimal in terms of SA-LUV binding to MLVs. Further, the binding reaction wasshown to be specific (Figure 5.3) and was not influenced by prior in vivo exposure (Figure5.4 and Table 5.2).The in vivo targeting studies based on SA-LUV binding to BAH-MLVsdemonstrate that liposome targeting can be achieved. As expected, the most efficient targetspecific delivery was achieved following i.p. injection of SA-LUVs in animals with BAH138MLVs localized in the peritoneal cavity. The results of the i.p./i.p. targeting studiesdemonstrate that targeting can be close to 100% efficient. In vitro studies, for example,suggest that a maximum of 1.3 jimoles SA-LUV lipid can bind per 20 tmoles BAH-MLVlipid (see Section 5.2). In the in vivo studies in the presence of tumor, the amount ofpelleted SA-LUV was 1.25 jimoles in animals given an i.p. injection of 20 iimoles BAHMLVs (Table 5.3).Clearly the most challenging targeting to be achieved is based on i.v. administeredSA-LUVs that must access a target outside the vascular compartment. The model approachdescribed here should provide an ideal way in which to assess targeted delivery systems.The number of targets, BAH-MLVs, can be estimated at between 1 and 30 x i09 within theperitoneal cavity, assuming that MLVs exhibit mean diameters in the range of 2 to 10 jim.Further, each BAH-MLV target has 10,000 to 200,000 target ligands expressed on thesurface (calculation based on MLVs incorporating 1 mol% BAH-DSPE and having 5% ofthe lipid in the outer lamellae). Finally, the binding reaction between the target and thecarrier is mediated by the high-affinity binding between biotin and streptaviclin. Using thismodel, targeted SA-LUVs demonstrated at least a 6-fold increase in specific binding toBAH-MLVs compared to non-targeted SA-LUVs and LUVs, with as much as 90% ofavailable SA-LUVs bound to the target (see Table 5.3). However, Table 5.3 clearlyindicates that having an excess of high affinity target sites present in the peritoneal cavitydoes not promote accumulation of SA-LUVs given i.v., even under conditions where thevascular permeability of the blood vessels lining the peritoneum has been increased inresponse to factors released from a growing ascitic tumor.139In conclusion, the results presented in this chapter demonstrate that SA-LUVs canbe targeted to MLVs incorporating a biotin-labeled phospholipid at an extravascular site.Saturated biotin-labeled phospholipids containing a linker arm to connect the biotin to thephospholipid yielded the most efficient binding of SA-LUVs to target MLVs in vitro. SALUVs incubated in vitro with normal mouse serum or recovered in the plasma of micemaintained their ability to bind to BAH-MLVs. Binding of SA-LUVs to BAH-MLVs in theperitoneal cavity was increased 6-fold and 18-fold for SA-LUVs injected i.v. and i.p.respectively.140CHAPTER 6SUMMARIZING DISCUSSIONThe objective of this thesis was to demonstrate and optimize in vivo targeting ofprotein-coated liposomes to a target population. Initially, it was shown that SA-LUVsadministered both intraperitoneally and intravenously specifically bound to P388 target cellsresiding in the peritoneal cavity. However, the low access of SA-LUVs to the target sitelocation and inconsistency between in vitro and in vivo binding results suggested thatlimited passive targeting and poorly defined in vivo biotin-streptavidin interactions wereresearch areas to be addressed. Consequently, additional studies investigated strategies toimprove passive targeting to the peritoneal cavity, and developed a model target to controltarget-associated variables in order to achieve binding to a defined target.In Chapter 3, specific targeting of SA-LUVs to target cells in vivo wasdemonstrated. However, there was significantly more non-specific cell-associatedliposomes in vivo than in vitro. This can be attributed to the presence of large numbers oftumor-associated macrophages. Further, even with saturating concentrations of SA-LUVs,only 20% of target cells were specifically labeled in vivo compared to 90% in vitro. The invivo biotin-streptavidin interaction between a biotinylated antibody and streptavidin linkedto a large macromolecule (liposome) may be hindered. It has been demonstrated elsewherethat the biotin-streptavidin interaction should be maintained in vivo. 125-labeledstreptavidin, for example, specifically bound in vivo to cells pre-labeled with biotinylatedantibody (Saga et al, 1994). Further, it is shown in this thesis that: 1) the presence of141normal mouse serum did not inhibit the specific binding of SA-LUVs to cells pre-labeledwith biotinylated antibody (see Section 3.2.3); 2) SA-LUVs recovered from the plasmacompartment or peritoneal cavity bound efficiently to a biotinylated target in vitro (seeSection 5.2.2); and 3) SA-LUVs isolated from animals fed an avidin-enriched diet,designed to eliminate endogenous biotin, demonstrated identical in vitro binding to biotinlabeled MLVs compared to SA-LUVs that had not been injected (see Section 5.2.2). Theseexperiments suggest streptavidin moieties attached to liposomes should retain their bindingaffinity for biotin-labeled targets in vivo for possible use as imaging agents or delivery ofdrugs.The fact that in vivo and in vitro biotin-streptavidin interactions between abiotinylated antibody and streptavidin liposome were inconsistent led to the development inChapter 5 of a model target system in which the in vivo targeting of SA-LUVs to a definedtarget population could be achieved. These studies were initiated in order to characterizethe biotin-streptavidin interaction in vivo. Results in Chapter 5 found that the bindingefficiency of SA-LUVs to biotin-labeled MLVs (BAH-MLVs) agreed well between in vitroand in vivo results, contrary to results in Chapter 3 between SA-LUVs and target cells prelabeled with biotinylated antibody. These contradictory results may be related to a 6 carbonspacer arm that was used to link the biotin moieties to DSPE lipid in MLVs, but was notused to link biotin moieties to the antibody molecules. Employment of a long spacer arm tolink biotin to DSPE may facilitate the ability of the biotin moiety to access and bind tostreptavidin present in the PEG protective layer on the surface of SA-LUVs. It has beenpostulated that incorporation of PEG-PE imparts a hydrophilic coating and/or steric barrier142on liposomes, which can hinder the binding of ligand-directed liposomes to targets. Resultsin Chapter 4 established that 5 mol% PEG-PE inhibited binding of SA-LUVs to P388 cells.Chapter 5 demonstrated that the incorporation of a 6 carbon spacer arm to link the biotinmoiety to the DSPE lipid significantly increased the binding of SA-LUVs to biotin-labeledMLVs.Studies on the use of biotin-labeled MLVs as a model target population will allowmanipulation of target-associated variables, such as antigen density and binding siteaccessibility. The studies presented in Chapter 5 employed 1 mol% BAH-DSPE. It wouldbe interesting to determine the influence of increasing the number of biotin-binding sites onthe targeting of SA-LUVs. For example, results in Chapter 3 indicated that although 60,000anti-Thy 1.2 antibodoes bound to target P388 cells, only 6,000 liposomes bound. BAHMLVs (2 - 10 jim diameter) incorporating 1% BAH-DSPE have been estimated to expressbetween 10,000 and 200,000 biotin moieties on the outer monolayer. The maximumnumber of liposomes bound per BAH-MLV (assuming 10 pm vesicles and 200,000 biotins)was 9,100 SA-LUVs. Hence, increasing the mol% BAH-DSPE may not further increaseSA-LUV binding. Concerning binding site accessibility, as discussed above, altering thelength of the spacer arm used to link biotin to DSPE may influence the access of the biotinmoiety to streptavidin present in a PEG coating on the liposome surface. Alternately,controlling PEG-PE variables such as polymer Mr and mol% incorporated should increasesuccessful interactions between biotin and streptavidin.Chapter 3 also showed that SA-LUVs accumulate significantly in the cells of theRES, thus leading to reduced circulation lifetimes. Strategies to reduce the accumulation of143liposomes in the RES will increase the propensity of injected liposomes to gain access tothe target cells. Enhanced circulation lifetimes, leading to increased blood liposome levels,have been correlated with increased passive targeting to diseased sites (Gabizon andPapahadjopoulos, 1988). Chapter 4 explored strategies to increase the circulation lifetimesof SA-LUVs in an effort to enhance passive targeting. The strategies of liver blockade andincorporation of PEG-PE into liposomes both decreased uptake in liver and hence resultedin increased circulation lifetimes. Each strategy also resulted in increased accumulation inthe peritoneal cavity of intravenously administered SA-LUVs. Combining the strategiesfurther decreased liver uptake and increased circulation lifetimes of SA-LUVs, but did notfurther increase passive targeting to the peritoneal cavity.There are at least two possible reasons for this lack of synergism between the twostrategies. First, the mechanism by which liposomes exit the vasculature and enter theperitoneal cavity may be rate limiting. Two mechanisms have been proposed. Leakagemay occur through pores or fenestrations within the endothelial cells lining the bloodvessels, or may involve transcytosis through endothelial cells by way of cytoplasmicvesicles (Kohn et al, 1992). Tn either case, increasing the blood circulation levels of SALUVs may only increase passive targeting up to a certain lipid level. Secondly, the mannerin which liver blockade and PEG-PE increase blood circulation lifetimes of SA-LUVs mayinfluence the mechanism involved in passive targeting. Liver blockade directly inhibits thephagocytic capability of KUpffer cells, while PEG-PE bestows a hydrophilic coating andsteric barrier on the liposome which results in reduced interactions with serum proteins, andthus reduced opsonization leading to macrophage uptake. The PEG coating present on the144liposome may, for instance, limit passage through the pores or fenestrations responsible forpassive targeting.Results in Chapter 5 suggested that liposome accumulation in the peritoneal cavityis not a function of equilibrium between liposome concentrations in the blood compartmentand the peritoneal cavity. If an equilibrium situation existed, a large excess of targets in theperitoneal cavity would drive accumulation of SA-LUVs to that location. This was notobserved (see Table 5.3), as SA-LUVs administered intravenously accumulated in theperitoneal cavity to the same extent in the presence of either excess BAH-MLVs or MLVs.The presence of P388 tumor in the peritoneal cavity had the greatest influence onthe passive targeting of SA-LUVs to that location (Chapter 4). The blood vessels associatedwith tumors are known to be hyperpermeable to macromolecules, such as liposomes.Synthesis and secretion of a glycoprotein known as vascular permeability factor (VPF) insolid and ascites tumors has been implicated in this enhanced permeability. In isolation,this protein has a potency estimated to be 50,000 times that of histamine (Senger et al,1983). The presence of VPF has been reported in human and animal malignant asciticeffusions (Yeo, 1993) and has been shown to localize in blood vessels near the tumor. VPFmay induce the formation of gaps between adjacent vessel endothelial cells located neartumors. Alternately, Kohn et al (1992) proposed that the predominant pathway by whichextravasation of macromolecules through tumor vessels, but not normal vessels, occurredwas via a system of membrane-bound, interconnecting vesicles and vacuoles known asvesiculo-vacuolar organelles (VVO). VVO structures are found in the endothelial cellslining both tumor-supplying and normal blood vessels, and it has been suggested that145exposure to VPF may result in up-regulation of VVO function and hence permeability. Inany case, strategies to induce vessel hyperpermeability, such as administration of VPF orinflammatory mediators (i.e. histamine), could be employed to increase passive targeting ofliposomes. Of particular interest would be the localization of permeability-enhancingfactors to blood vessels near tumors.In conclusion, the studies presented in this thesis establish that protein-coatedliposomes are capable of specific in vivo binding to target cells pre-labeled withbiotinylated antibody. The passive targeting of intravenously administered SA-LUVs totarget sites and the in vivo biotin-streptavidin interaction were identified as essentialcomponents in achieving efficient in vivo targeting. Further investigation andunderstanding of these components will potentially lead to improved in vivo targeting ofliposomal drug carrier systems to diseased sites in an extravascular location.146REFERENCESAbbas, A.K., Lichtman, A.H. and Pober, J.S. (1991) in Cellular And MolecularImmunology. Wonsiewicz, M.J., ed. W.B. Saunders Company, Philadelphia, PA.Abra, R.M., Bosworth, M.E. and Hunt, C.A. (1980) Res. Commun. Chem. Pathol.Pharmacol. 29, 349-360.Absolom, D. (1986) Methods Enzymol. 132,281-317.Abuchowski, A., Van Es, T., Palczuk, N.C. and Davis, F.F. (1977) J. Biol. Chem. 252,3578-3581.Ahmad, land Allen, T.M. (1992). Cancer Res. 52,4817-4820.Ahmad, I., Longenecker, M., Samuel, J. and Allen, T.M. (1993) Cancer Res. 53, 1484-1488.Allen, T.M. (1984) in Liposome Technology, Vol. L Gregoriadis, G., ed. CRC Press, BocaRaton, FL., 109-122.Allen, T.M. (1989) in Liposomes In The Therapy Of Infectious Diseases And Cancer.Lopez-Berestein G., Fidler I.J., eds. Alan R. Liss, mc, New York, 405-415.Allen, T.M. and Chonn, A. (1987) FEBS Lett. 223,42-46.Allen, T.M. and Everest, J.M. (1983) J. Pharmacol. Exp. Therap. 226, 539-544.Allen, T.M. and Hansen, C. (1991) Biochim. Biophys. Acta 1068, 133-141.Allen, T.M., Agrawal, A.K., Ahmad, I., Hansen, C.B. and Zalipsky, S. (1994) J. LiposomeRes. 4, 1-26.Allen, T.M., Hansen, C. and Rutledge, J. (1989) Biochim. Biophys. Acta 981,27-35.Allen, T.M., Hansen, C., Martin, F., Redemann, C. and Yau-Young, A. (1991) Biochim.Biophys. Acta 1066,29-36.Allen, T.M., Murray, L., MacKeigan, S. and Shah, M. (1984) J. Phamacol. Exp. Therap.229, 267-275.Altura, B.M. (1980) Adv. Microcirc. 9,252-294.Alving, C.R. (1991) J. Jinmunol. Meth. 140, 1-13.Alving, C.R. and Richards, R.L. (1983) in Liposomes. Ostro, M., ed. Marcel Dekker, NewYork, 209-287.Alving, C.R. and Swartz, G.M. (1984) in Liposome Technology, Vol II. Gregoriadis, G.,ed. CRC Press, Boca Raton, FL., 55-68.Alving, C.R., Steck, E.A., Chapman, W.L. Jr., Waits, V.B., Hendricks, L.D., Swartz, G.M.Jr. and Hanson, W.L. (1978) Proc. Nazi. Acad. Sci. USA 75,2959-2963.Alving, C.R., Steck, E.A., Chapman, W.L., Waits, V.B., Hendricks, L.D., Swartz, G.M. andHanson, W.L. (1980) Lfe Sci. 26, 223 1-2238.147Anderson, P.M., Katsanis, E., Sencer, S.F., Hasz, D., Ochoa, A.C. and Bostrom, B. (1992)J. Immunol. 12, 19-31.Aragnol, D. and Leserman, L.D. (1986) Proc. Nati. Acad. Sci. USA 83, 2699-2703.Badger, C.C., Krohn, K.A., Peterson, A.V., Shulman, H. and Bernstein, I.D. (1985) CancerRes. 45, 1536-1544.Bakker-Woudenberg, I.A.J.M., Lokerse, A.F., ten Kate, M.T. and Storm, G. (1992)Biochim. Biophys. Acta 1138, 3 18-326.Balazsovits, J.A.E., Mayer, L.D., Bally, M.B., Cullis, P.R., McDonell, M., Ginsberg, R.S.and Falk, R.E. (1989) Cancer Chemother. Pharmacol. 23: 8 1-86.Bally M.B., Masin, D., Nayar, R., Cullis, P.R. and Mayer, L.D. (1994) Cancer Chemother.Pharmacol. 34, 137-146.Bally, M.B., Hope, M.J., Mayer, L.D., Madden, T.D. and Cullis, P.R. (1988) in LiposomesAs Drug Carriers. Gregoriadis, G., ed. John Wiley and Sons Ltd., Chichester, 841-853.Bally, M.B., Nayar, R., Masin, D., Hope, M.J., Cullis, P.R. and Mayer, L.D. (1990)Biochim. Biophys. Acta 1023, 133-139.Bangham, A.D., Standish, M.M. and Watkins, J.C. (1965) J. Mol. Biol. 13, 238-252.Bankert, R.B., Yokota, S., Ghosh, S.K., Mayhew, E. and Jou, Y-H. (1989) Cancer Res. 49,301-308.Barenholz, Y., Amselem, S. and Lichtenberg, D. (1979) FEBS Lett. 99, 210-213.Barratt, G. and Schuber, F. (1993) in Liposome Technology, Vol ifi. CRC Press, BocaRaton, FL., 199-218.Beaumier, P.L. and Hwang, K.J. (1983) Biochim. Biophys. Acta 731,23-30.Bennett, C.F., Chiang, M., Chan, H., Shoemaker, J.E. and Mirabelli, C. (1992) Pharinacol.41, 1023-1033.Blakey, D.C. (1992) Acta Oncol. 31, 9 1-97.Blume, G. and Cevc, G. (1990) Biochim. Biophys. Acta 1029,91-97.Boman, N., Masin, D., Mayer, L.D., Cullis, P.R. and Bally, M.B. (1994) Cancer Res. 54,2830-2833.Bonte, F. and Juliano, R.L. (1986) Chem. Phys. Lipids 40, 359-372.Braslawsky, G.R., Edson, M.A., Pearce, W., Kaneko, T. and Greenfield, R. (1990) CancerRes. 50, 6608-66 14.Brown, P.J. and Juliano, R.L. (1985) Science 228, 1448-145 1.Bussian, R.W. and Wriston, J.C. Jr. (1977) Biochim. Biophys. Acta 471, 336-340.Carlsson, J., Drevin, H. and Axen, R. (1978) Biochem. J. 173,723-737.Chaiet, L. and Wolf, F.J. (1964) Arch. Biochem. Biophys. 106, 1-5.148Chapman, D. (1968) in Biological Membranes, Vol. 1. Chapman, D., ed. Academic Press,N.Y., 125-153.Chapman, D. (1975) Quart. Rev. Biophys. 8, 185-235.Chonn, A. (1991) Ph.D. Thesis.Chonn, A., Semple, S.C. and Cullis, P.R. (1992) J. Riot. Chem. 267, 18759-18765.Collins, D.S., Findlay, K. and Harding, C.V. (1992) J. Immunol. 148, 3336-3341.Connor, J., Sullivan, S. and Huang, L. (1985) Phar,nacol. Ther. 28, 341-365.Connor, J., Yatvin, M. and Huang, L. (1984) Proc. Natl. Acad. Sci. USA 81, 1715-1718.Cowens, J.W., Creaven, P.J., Greco, W.R., Breener, D.E., Tung, Y., Ostro, M., Pilkiewicz,F., Ginsberg, R. and Petrelli, N. (1993) Cancer Res. 53,2796-2802.Cullis, P.R. and de Kruijff, B. (1979) Biochim. Biophys. Acta 559, 399-420.Cullis, P.R., de Kruijff, B., Hope, M.J., Verkleij, A.J., Nayar, R., Farren, S.B., Tilcock,C.P.S., Madden, T.D. and Bally, M.B. (1983) in Membrane fluidity in Biology, Vol 1.Aloia, R.C., ed. Academic Press, N.Y., 40-83.Cullis, P.R., de Kruijff, B., Verkleij, A.J. and Hope, M.J. (1986) Biochem. Soc. Trans. 14,242-245.Cullis, P.R., Hope, M.J., Bally, M.B., Madden, T.D., Mayer, L.D. and Janoff, A.S. (1987) inLiposomes. Ostro, M.J., ed. Marcel Dekker, New York, 39-72.Cullis, P.R., Hope, M.J., de Kruijff, B., Verkleij, A.J. and Tilcock, C.P.S. (1985) inPhospholipids And Cellular Regulation, Vol 1. Kuo, J.F., ed. CRC Press, Boca Raton,FL., 1-59.Davis, J.H. (1983) Biochim. Biophys. Acta 737,117-172.de Kruijff, B., Cullis, P.R., Verkleij, A.J., Hope, M.J., van Echteld, C.J.A. and Toraschi,T.F. (1984) in Enzymes Of Biological Membranes. Martinosi, A., ed. Plenum, NewYork, 13 1-204.Deamer, D.W. and Bramhall, J. (1986) Chem. Phys. Lipids 40, 167-188.Debs, RJ, Heath, TD and Papahadjopoulos, D (1987) Biochim. Biophys. Acta 901, 183-190.Demel, R.A. and de Kruijff, B. (1976) Biochim. Biophys. Acta 457, 109-132.Demel, R.A., Geurts van Kessel, W.S.M. and van Deenen, L.L.M. (1972) Biochim.Biophys. Acta 266, 26-37.Derksen, J.T.P. and Scherphof, G. (1985) Biochim. Biophys. Acta 814, 151-155.Derksen, J.T.P., Morselt, H.W.M., Kalicharan, D., Hulstaert, C.E. and Scherphof, G.L.(1987) Exp. Cell Res. 168,105-115.Duncan, R., Seymour, L.W., O’Hare, K.B., flanagan, P.A., Wedge, S., Ulbrich, K.,Strohalm, J., Subr, V., Spreafico, F., Grandi, M., Ripamonti, M., Farao, M. and Suarato,A. (1992) J. Controlled Release 18, 123-132.149Duzgunes, N., Perumal, V.K., Kesavalu, L., Goldstein J.A., Debs, R.J. and Gangadharam,P.R.J. (1988) Antimicrob. Agents Chemother. 32, 1404-1411.Dvorak, H.F., Nagy, J.A. and Dvorak, A.M. (1991) Cancer Cells. 3,77-85.Dvorak, H.F., Nagy, J.A., Dvorak, J.T. and Dvorak, A.M. (1988) Am. .1. Pathol. 133, 95-109.Dvorak, H.F., Sioussat, T.M., Brown, L.F., Berse, B., Nagy, J.A., Sotrel, A., Manseau, E.J.,Van De Water, L. and Senger, D.R. (1991b) J. Exp. Med. 174, 1275-1278.Ehrlich, P. (1906) in Collected Studies on Immunology, Vol. 2. John Whiley, New York,442-447.Endoh, H., Suzuki, Y. and Hashimoto, Y. (1981) J. Immunol. Meth. 44,79-85.Enoch, H.G. and Strittmatter, P. (1979) Proc. Nati. Acad. Sd. USA 76, 145-149.By, P.L., Prowse, S.J. and Jenkin, C.R. (1978) Biochemistry 15,429-436.Faradji, A., Bohbot, A., Frost, M., Schmitt-Goguel, M., Siffert, J.C., Dufour, P., Eber, M.,Lallot, C., Wiesel, M.L., Bergerat, J.P. and Oberling, F. (1991) J. Clin. Oncol. 9, 1251-1260.Feldman, G.B. (1975) Cancer Res. 35, 325-332.Fell, H.P., Yarnold, S., Hellström, I., Hellström, K.E. and Folger, K.R. (1989) Proc. Nati.Acad. Sci. USA 86, 8507-8511.Fettiplace, R. and Haydon, D.A. (1980) Physiol. Rev. 60, 5 10-550.Fidler, U. (1993) in Liposome Technology, Vol II. Gregoriadis, G., ed. CRC Press, BocaRaton, FL., 45-64.Finkelstein, M.C. and Weissmann, G. (1979) Biochim. Biophys. Acta 587, 202-2 16.Finn, F.M., Titus, G. and Montibeller, J.A. (1980) J. Biol. Chem. 255, 5742-5746.Flessner, M.F., Dedrick, R.L. and Schultz, J.S. (1985) Am. J. Physiol. 248, H15-H25.Freise, J., MUller, W.H., Brölsch, C.H. and Schmidt, F.W. (1980) Biomedicine 32, 118-123.Fries, L.F., Gordon, D.M., Richards, R.L., Egan, J.E., Hollingdale, M.R., et al (1992) Proc.Natl. Acad. Sci. USA 89, 358-362.Gabizon, A. (1992) Cancer Res. 52, 89 1-896.Gabizon, A. and Papahadjopoulos, D. (1988) Proc. Nati. Acad. Sci. USA 85, 6949-6953.Gabizon, A. and Papahadjopoulos, D. (1992) Biochim. Biophys. Acta 1103, 94-100.Gabizon, A., Dagan, A., Goren, D., Barenholz, Y. and Fuks, Z. (1982) Cancer Res. 42,4734-4739.Gabizon, A., Goren, D., Fuks, Z., Meshorer, A. and Barenholz, Y. (1985) Br. J. Cancer 51,681-689.Gabizon, A., Meshorer, A. and Barenholz, Y. (1986) J. Nati. Cancer Inst. 77,459-467.150Gao, X. and Huang, L. (1993) J. Liposome Res. 3, 17-30.Garrigues, 3., Garrigues, U., Hellström, I. and Hellström, K.E. (1993) Am. J. Pathol. 142,607-Gerlowski, L.E. and Jam, R.K. (1986) Microvasc. Res. 31,288-305.Ghose, T. and Blair, A.H. (1987) CRC Crit. Rev. Ther. Drug Carrier Syst. 3,263-361.Ghose, T., Blair, A.H., Uadia, P., Kulkarni, P.N., Goundalkar, A., Mezei, M. and Ferrone,S. (1985) Ann. NYAcad. Sci. 448, 2 13-227.Ghosh, P., Das, P.K. and Bacchawat, B.K. (1982) Arch. Biochem. Biophys. 213,266-269.Ghosh, R. and Seelig, J. (1982) Biochim. Biophys. Acta 691, 151-158.Goldstein, J., Anderson, R. and Brown, S. (1979) Nature 279, 679-Gray, A.G., Morgan, J., Linch, D.C. and Huehns, E.R. (1988) Clin. Exp. Immunol. 72, 168-173.Green, N.M. (1975) Adv. Protein Chem. 29, 85-133.Gregoriadis, G. (1990) Immunol. Today 11, 89-97.Gregoriadis, G. (1991) J. Antimicrob. Chemother. (Suppi. B) 28, 39-48.Gregoriadis, G. and Davis, C. (1979) Biochem. Biophys. Res. Commun. 89, 1287-1293.Gregoriadis, G. and Neerunjun, E.D. (1974) Eur. J. Biochem. 47, 179-185.Gregoriadis, G. and Ryman, B.E. (1972) Eur. J. Biochem. 27,485-491.Gregoriadis, G., ed. (1993) Liposome Technology, Vol. II Gregoriadis, G., ed., CRC Press,Boca Raton, FL.Gruner, S.M., Lenk, R.P., Janoff, A.S. and Ostro, M.J. (1985) Biochemistry 24,2833-2842.Hale, G., Dyer, M.J.S., Clark, M.R., et al (1988) Lancet 2, 1394-1399.Hashimoto, K., Loader, J.E. and Kinsky, S.C. (1986) Biochim. Biophys. Acta 816, 163-168.Heath, T.D. and Martin, F.J. (1986) Chem. Phys. Lipids 40, 347-358.Heath, T.D., Fraley, R.T. and Papahadjopoulos, D. (1981) Science 210, 539-541.Heath, T.D., Fraley, R.T., Bentz, J., Voss, E.W., Herron, J.N. and Papahadjopoulos, D.(1984) Biochim. Biophys. Acta 770, 148-158.Heath, T.D., Montgomery, J.A., Piper, J.R. and Papahadjopoulos, D. (1983) Proc. Nati.Acad. Sci. USA 80, 1377-1381.Helenius, A., Fries, E. and Kartenbeck, J. (1977) J. Cell Biol. 75, 866-880.Hellström, L, Garrigues, H.J., Garrigues, U. and Hellström, K.E. (1990) Cancer Res. 50,2183-2190.Hirano, K. and Hunt, C.A. (1985) J. Pharmaceutical Sci. 74, 915-921.Hirota, S. and Kikuchi, M. (1985) Japanese Patent No. 7934-1985.151Hope, M.J. and Cullis (1980) Biochem. Biophys. Res. Commun. 92, 846-852.Hope, M.J., Bally, M.B., Mayer, L.D., Janoff, A.S. and Cullis, P.R. (1986) Chem. Phys.Lipids 40, 89-107.Hope, M.J., Bally, M.B., Webb, G. and Cullis, P.R. (1985) Biochim. Biophys. Acta 812, 55-65.Hopfer, R.L., Mills, K., Mehta, R., Lopez-Berestein, G., Fainstein, V. and Juliano, R.L.(1984) Antimicrob. Agents Chemother. 25,387-389.Hsu, M.J. and Juliano, R.L. (1982) Biochim. Biophys. Acta 720,411-419.Huang, A., Huang, L. and Kennel, S.J. (1980) .1. Biol. Chem. 255, 8015-8018.Huang, C.H. (1969) Biochemistry 8, 344-352.Huang, C.H. (1977) Lipids 12, 348-356.Huang, S.K., Mayhew, E., Gilani, S., Lasic, D.D., Martin, F.J. and Papahadjopoulos, D.(1992) Cancer Res. 52,6774-6781.Hubbel, W.L. and McConnell, H.M. (1971) J. Am. Chem. Soc. 93, 314-319.Hughes, B.J., Kennel, S., Lee, R. and Huang, L. (1989) Cancer Res. 49, 6214-6220.Hutchinson, F.J. and Jones, M.N. (1988) FEBS Lett. 234,493-496.Hwang, K.J. (1987) in Liposomes. Ostro, M.J., ed. Marcel Dekker, New York, 109-156.Hyde, S.C., Gill, D.R., Higgins, C.F., Trezide, A.E.O., MacVinish, L.J., Cuthbert, A.W.,Ratcliff, R., Evans, M.J. and Colledge, W.H. (1993) Nature (London) 362,250-255.Hyslop, P.A., Morel, B. and Sauerheber, R.D. (1990) Biochemistry 29, 1025-1038.Israelachvili, J.N., Marceija, S. and Horn, R.G. (1980) Q. Rev. Biophys. 13, 121-148.Jam, R.K. (1987) Cancer and Metastasis Rev. 6, 559-Jam, R.K. (1988) Cancer Res. 48,2641-2658.Jam, R.K. (1990) Cancer Res. (Suppi.) 50, 814s-819s.Janoff, A.S. (1992) J. Lab. Invest. 66,655-658.Jansons, V.K. and Mallet, P.L. (1981) Analyt. Biochem. III, 54-59.Jost, P.C. and Griffith, O.H., eds. (1982) Lipid-protein Interactions, Vol 1 and 2. Wiley,New York.Juliano, R.L. and Layton, D. (1980) in Drug Delivery Systems: Characteristics AndBiomedical Applications. Juliano, R.L., ed. Oxford University Press, Oxford, England,189-236.Juliano, R.L. and Meyer, M. (1993) in Liposome Technology, Vol ifi. Gregoriadis, G., ed.CRC Press, Boca Raton, FL, 15-26.Juliano, R.L. and Stamp, D. (1975) Biochem. Biophys. Res. Commun. 63, 65 1-658.152Kachar, B. and Reese, T.S. (1982) Nature (London) 296,464-466.Kagawa, Y. and Packer, E. (1971) J. Biol. Chem. 246, 5477-5487.Kinsky, S.C. and Nicolotti, R.A. (1977) Ann. Rev. Biochem. 46,49-67.Kirby, C., Clarke, J. and Gregoriadis, G. (1980) FEBS Lett. 111, 324-328.Kirby, C., Clarke, J. and Gregoriadis, G. (1980b) Biochem. J. 186, 59 1-598.Kirby, C.J. and Gregoriadis, G. (1984) in Liposome Technology, Vol. I. Gregoriadis, G.,ed. CRC Press, Boca Raton, FL., 19-27.Klausner, R.D., Blumenthal, R., Innerarity, T. and Weinstein, J.N. (1985) J. Biol. Chem.260, 13719-13727.Klibanov, A.L., Khaw, B.A., Nossiff, N., O’Donnell, S.M., Huang, L., Slinkin, M.A. andTorchilin, V.P. (1991b) Am. J. Physiol. (Suppi.) 261,60-65.Klibanov, A.L., Maruyama, K., Beckerleg, A.M., Torchilin, V.P. and Huang, L. (1991)Biochim. Biophys. Acta 1062, 142-148.Klibanov, A.L., Maruyama, K., Torchilin, V.P. and Huang, L. (1990) FEBS Lett. 268, 235-237.Koff, W.C. and Fidler, I.J. (1985) Antiviral Res. 5, 179-190.Köhler, G. and Milstein, C. (1975) Nature (London) 256,495-497.Kohn, S., Nagy, J.A., Dvorak, H.F. and Dvorak, A.M. (1992) Lab. Invest. 67, 596-607.Konno, H., Tadakuma, T., Kumai, K., Takahashi, T., Ishibiki, K., Abe, 0. and Sakaguchi,S. (1990) Jpn. J. Surg. 20,424-428.Krupp, L., Chobanian, A.V. and Brecher, P.1. (1976) Biochim. Biophys. Acta 72, 125 1-1258.Ladbroke, B.D., Williams, R.M. and Chapman, D. (1968) Biochim. Biophys. Acta 156, 333-340.Langone, J.J. (1978) J. Immunol. Methods 24,269-285.Lasic, D.D., Martin, F.J., Gabizon, A., Huang, S.K. and Papahadjopoulos, D. (1991)Biochim. Biophys. Acta 1070, 187-192.Ledbetter, J.A. and Herzenberg, L.A. (1979) Immunological Rev. 47, 63-90.Lelkes, P.1. (1984) in Liposome Technology, Vol. I. Gregoriadis, G., ed. CRC Press, BocaRaton, FL., 5 1-65.Leserman, L. and Machy, P. (1987) in Liposomes. Ostro, M.J., ed. Marcel Dekker, NewYork, 157-194.Leserman, L., Machy, P. and Zelphati, 0. (1994) J. Liposome Res. 4, 107-119.Leserman, L.D. (1981) in Liposomes, Drugs And Immunocompetant Cell Functions.Nicolau, C. and Paraf, A., eds. Academic Press, London, 109-122.153Leserman, L.D., Barbet, 3., Kourilsky, F.M. and Weinstein, J.N. (1980) Nature (London)288,602-604.Leserman, L.D., Machy, P. and Barbet, 3. (1981) Nature (London) 293,226-228.Leserman, L.D., Weinstein, J.N., Blumenthal, R. and Terry, W.D. (1980b) Proc. Nati.Acad. Sci. USA 77,4089-4093.Lobuglio, A.F., Saleh, M.N., Lee. J., Khazaeli, M.B, et al (1988) J. Nail. Cancer Inst. 80,932-936.Longman, S.A., Cullis, P.R., Choi, L., de Jong, G., and Bally, M.B. (1994) CancerChemother. Pharmacol. (In Press).Lopez-Berestein, G. and Juliano, R.L. (1987) in Liposomes. Ostro, M.J., ed. MarcelDekker, New York, 253-276.Lopez-Berestein, G., Bodey, G.P., Fronkel, L.S. and Mehta, K. (1987) J. Clin. Oncol. 5,310-317.Lopez-Berestein, G., Fainstein, V., Hopfer, R., Mehta, K., Sullivan, M.P., Keating, M.,Rosenblum, M.G., Mehta, R., Luna, M., Hersh, E.M., Reuben, J., Juliano, R.J. andBodey, G.P. (1985) J. Infect. Dis. 151,704-710.Lopez-Berestein, G., Mehta, R., Hopfer, R., Mills, K., Kasi, L., Mehta, K., Fainstein, V.,Luna, M., Hersh, E.M. and Juliano, R.J. (1983) J. Infect. Dis. 147, 939-944.Loughrey, H.C., Bally, M.B. and Cullis, P.R. (1987) Biochim. Biophys. Acta 901, 157-160.Loughrey, H.C., Bally, M.B., Reinish, L.W. and Cullis, P.R. (1990c) Thromb. Haemostas.I64, 172-176.Loughrey, H.C., Choi, L.S., Cullis, P.R. and Bafly, M.B. (1990b) J. Immunol. Methods 132,25-35.Loughrey, H.C., Choi, L.S., Wong, K.F., Cullis, P.R. and Bally, M.B. (1993) in LiposomeTechnology, Vol. ifi, Gregoriadis, G., ed., CRC Press, Boca Raton, FL., 163-178.Loughrey, H.C., Ferraretto, A., Cannon, A-M., Acerbis, G., Sudati, F., Bottiroli, G.,Masserini, M. and Soria, M.R. (1993b) FEBS Lett. 332, 183-188.Loughrey, H.C., Wong, K.F., Choi, L.S., Cullis, P.R. and Bally, M.B. (1990) Biochim.Biophys. Acta 1028,73-81.Macby, P. and Leserman, L.D. (1984) EMBO J. 3, 197 1-1977.Mackaness, G.B. (1990) J. Liposome Res. 1, 503-5 10.Madden, T.D. (1986) Chem. Phys. Lipids 40,207-222.Madden, T.D., Harrigan, P.R., Tai, L.C.L., Bally, M.B., Mayer, L.D., Redelmeier, T.E.,Loughrey, H.C., Tilcock, C.P.S., Reinish, L.W. and Cullis, P.R. (1990) Chem. Phys,Lipids 53, 37-46.Malech, H.L. (1988) in Inflammation: Basic Principles And Clinical Correlates. Gallin, J.I.,Goldstein, I.M. and Snyderman, R., eds. Raven Press, New York, 297-308.154Malinski, l.A. and Nelsestuen, G.L. (1989) Biochemistry 28,61-70.Martin, F.J. and Papahadjopoulos, D. (1982) J. Biol. Chem. 257,286-288.Martin, F.J., Hubbell, L. and Papahadjopoulos, D. (1981) Biochemistry 20,4229-4238.Maruyama, K., Holmberg, E., Kennel, S.J., Klibanov, A., Torchilin, V. and Huang, L.(1990) J. Pharm. Sci. 79,978-984.Maruyama, K., Kennel, S.J. and Huang, L. (1990b) Proc. Nat!. Acad. Sci. USA 87, 5744-5748.Maruyama, K., Okamoto, A., Ishida, 0., Kojima, S., Suginaka, A., Huang, L. and Iwatsuru,M. (1994) J. Liposome Res. 4,701-724.Matthay, K.K., Heath, T.D. and Papahadjopoulos, D. (1984) Cancer Res. 44, 1880-1886.Mayer, L.D., Bally, M.B. and Cullis, P.R. (1986) Biochim. Biophys. Acta 857, 123-126.Mayer, L.D., Bally, M.B. and Cullis, P.R. (1990) J. Liposome Res. 1,463-480.Mayer, L.D., Bally, M.B., Cullis, P.R., Wilson, S.L. and Emerman, J.T. (1990b) CancerLett. 53, 183-189.Mayer, L.D., Bally, M.B., Loughrey, H., Masin, D. and Cullis, P.R. (1990c) Cancer Res. 50,575-579.Mayer, L.D., Cullis, P.R. and Bally, M.B. (1994) J. Liposome Res. 4, 529-553.Mayer, L.D., Hope, M.J. and Cullis, P.R. (1986b) Biochim. Biophys. Acta 858, 161-168.Mayer, L.D., Hope, M.J., Cullis, P.R. and Janoff, A.S. (1985) Biochim. Biophys. Acta 817,193-196.Mayer, L.D., Madden, T.D., Bally, M.B. and Cullis, P.R. (1993) in Liposome Technology,Vol II. Gregoriadis, G., ed. CRC Press, Boca Raton, FL., 27-44.Mayer, L.D., Nayar, R., Thies, R.L., Boman, N.L., Cullis, P.R. and Bally, M.B. (1993b)Cancer Cheinother. Pharmacol. 33, 17-24.Mayhew, E.G., Lasic, D., Barbar, S. and Martin, F.J. (1992) mt. J. Cancer 51, 302-309.Metezeau, P., Elguindi, L and Goldberg, M.E. (1984) EMBO J. 3,2235-2242.Mimnis, L.T., Zampighi, G., Nozaki, Y., Tanford, C. and Reynolds, J.A. (1981)Biochemistry 20, 833-840.Moghimi, S.M. and Patel, H.M. (1988) FEBSLett. 233, 143-147.Moghimi, S.M. and Pate!, H.M. (1989) Biochim. Biophys. Acta 984, 384-387.Moghimi, S.M. and Patel, H.M. (1993) in Liposome Technology, Vol ifi. Gregoriadis, G.,ed. CRC Press, Boca Raton, FL., 43-58.Money-Kyrle, J.F., Bates, F., Ready, J., Gazzard, B.G., Phillips, R.H. and Boag, F.C. (1993)Clin. Oncol. 5, 367-37 1.Morgan, J.R., Williams, K.E., Davies et al (1981) J. Med. Microbiol. 14, 213-217.155Mori, A. and Huang, L. (1993) in Liposome Technology, Vol ifi. Gregoriadis, G., ed. CRCPress, Boca Raton, FL., 153-162.Mori, A., Kennel, S.J. and Huang, L. (1993) Pharm. Res. 10, 507-5 14.Morrisett, J.D., Jackson, R.L. and Gotto, A.M. (1977) Biochim. Biophys. Acta 472,93-133.Muller-Eberhard, H.J. (1986) Ann. Rev. Immunol. 4, 503-528.Muller-Eberhard, H.J. (1988) Ann. Rev. Biochem. 57, 321-347.Nagy, J.A., Henzberg, K.T., Masse, E.M., Zientara, G.P. and Dvorak, H.F. (1989) CancerRes. 49, 5448-5458.Nässander, U.K., Steerenberg, P.A., Poppe, H., Storm, G., Poels, L.G., De Jong, W.H. andCrommelin, D.J.A. (1992) Cancer Res. 52, 646-653.Nayar, R., Hope, M.J. and Cullis, P.R. (1989) Biochim. Biophys. Acta 986, 200-206.Ogihara, I., Kojima, S. and Jay, M. (1986) Eur. .1. NucI. Med. 11,405-411.Ogihara-Umeda, I., Sasaki, T., Toyama, H., Oda, K., Senda, M. and Nishigori, H. (1994)Cancer Res. 54,463-467.Ohsawa, T., Miura, H. and Harada, K. (1985) Chem. Pharin. Bull. 33, 2916-2923.Oku, N., Namba, Y., Takeda, A. and Okada, 5. (1993) Nuci. Med. Biol. 20,407-412.Olson, F., Hunt, C.A., Szoka, E.C., Vail, W.J. and Papahadjopoulos, D. (1979) Biochim.Biophys. Acta 557, 9-23.Olson, F., Mayhew, E., Maslow, D., Rustum, Y. and Szoka, F. (1982) Eur. J. Cancer Cliii.Oncol. 18, 167-176.Papahadjopoulos, D. (1978) in Membrane Fusion. Poste, G. and Nicolson, G.L., eds.North-Holland, Amsterdam, 77-90.Papahadjopoulos, D., Allen, T.M., Gabizon, A., Mayhew, E., Matthay, K., Huang, S.K.,Lee, K.D., Woodle, M.C., Lasic, D.D., Redemann, C. and Martin, F.J. (1991) Proc. Nati.Acad. Sci. USA 88, 11460-11464.Papahadjopoulos, D., Jacobson, K., Nir, S. and Isac, T. (1973) Biochim. Biophys. Acta 311,330-339.Parente, R.A. and Lentz, B.E. (1984) Biochemistry 23,2353-2362.Parr, M.J., Bally, M.B. and Cullis, P.R. (1993) Biochim. Biophys. Acta 1168, 249-252.Pestalozzi, B., Schwendener, R. and Sauter, C. (1992) Annals Oncol. 3,445-449.Peterson, H-I. (1979) in Tumor Blood Circulation. Peterson, H-I., ed. CRC Press, BocaRaton, FL., 77-85.Pietersz, G.A. and McKenzie, I.F.C. (1992) Immunol. Rev. 129, 57-80.Porstmann, B., Porstmann, T., Nugel, E. and Evers, U. (1985) J. Immunol. Methods 79, 27-37.156Poste, G. (1983) Biol. Cell 47, 19-38.Poste, G., Bucana, C., Raz, A., Bugeiski, P., Kirsh, R. and Fidler, I.J. (1982) Cancer Res.42, 1412-1422.Poste, G., Kirsh, R. and Koestler, T. (1984) in Liposome Technology, Vol ifi. Gregoriadis,G., ed. CRC Press, Boca Raton, FL., 1-28.Pressman, D. (1948) J. Immunol. 59, 141-146.Proffitt, R.T., Satorius, A., Chiang, S.M., Sullivan, L. and Adler-Moore, J.P. (1991) J.Antimicrob. Chemother. 28,49-61.Proffitt, R.T., Williams, L.E., Presant, C.A., Tin, G.W., Uliana, J.A., Gamble, R.C. andBaldeschwieler, J.D. (1983) Science 220, 502-505.Racker, E. (1973) J. Membr. Biol. 10,221-233.Rahman, A., Kessler, A., More, N., Sikic, B., Rowden, G., Woolley, P., and Schein, P.S.(1980) Cancer Res. 40, 1532-1537.Rahman, A., Treat, J., Roh, J.K., Potkul, L.A., Alvord W.G., Forst, D. and Woolley, P.V.(1990) J. Clin. Oncol. 8, 1093-1100.Rance, M., Jeffrey, K.R., Tullock, A.P., Butler, K.W. and Smith, I.C.P. (1982) Biochim.Biophys. Acta 688, 191-200.Richards, R.L., Gewurz, H., Siegel, J. and Alving, C.R. (1979) J. Immunol. 122, 1185-1189.Richardson, V.J., Ryman, B.E., Jewkes, R.F., Jeyasingh, K., Tattersafi, M.N.H., Newlands,E.S. and Kaye, S.B. (1979) Br. J. Cancer 40, 35-43.Rihova, B., Ulbrich, K., Strohaim, J., Vetvicka, V., Bilej, M., Duncan, R. and Kopecek, J.(1989) Biomaterials 10, 335-342.Roerdink, F., Dykstra, J., Hartman, G., Bolscher, B. and Scherphof, G. (1981) Biochim.Biophys. Acta 667,79-89.Rose, J.K., Buonocore, L. and Whitt, M.A. (1991) Biotechniques 10, 520-524.Rosenberg, M.B., Breakefield, X.O. and Hawrot, E. (1987) J. Neurochem. 48, 865-875.Ross, D.D., Joneckis, C.C., Ordonez, J.V., Sisk, A.M., Wu, R.K., Hamburger, A.W. andNora, R.E. (1989) Cancer Res. 49, 3776-3782.Rossi, J.D. and Wallace, B.A. (1983) .1. Biol. Chem. 258, 3327-333 1.Rostaing-Capaillon, 0. and Casellas, P. (1990) Cancer Res. 50,2909-2916.Rottenberg, H. (1979) Methods Enzymol. 55, 547-569.Rusckowski, M., Fritz, B. and Hnatowich, D.J. (1992) J. Nuci. Med. 33, 1810-1815.Saga, T., Weinstein, J.N., Jeong, J.M., Heya, T., Lee, J.T., Le, N., Paik, C.H., Sung, C. andNeumann, R.D. (1994) Cancer Res. 54,2160-2165.157Samuel, J., Noujaim, A.A., Willans, D.J., Brzezinska, G.S., Haines, D.M. and Longenecker,B.M. (1989) Cancer Res. 49,2465-2470.Scherphof, G.L., Damen, J. and Wilschut, J. (1984) in Liposome Technology, Vol ifi.Gregoriadis, G., ed. CRC Press, Boca Raton, FL, 205-224.Schiom, J. (1986) Cancer Res. 46, 3225-3238.Schroff, R.W., Foon, K.A., Beatty, S., Oldham, R.K. and Morgan, A.C. (1985) Cancer Res.45, 879-885.Schroit, A.J., Madsen, J. and Nayer, R. (1986) Chem. Phys. Lipids 40,373-393.Seelig, 3. (1978) Biochim. Biophys. Acta 515, 105-140.Senger, D.R., Galli, S.J., Dvorak, A.M., Perruzzi, C.A., Harvey, V.S. and Dvorak, H.F.(1983) Science 219, 983-985.Senior, J. and Gregoriadis, G. (1982) Lfe Sci. 30,2123-2136.Senior, J., Crowley, J.C.W. and Gregoriadis, G. (1985) Biochim. Biophys. Acta 839, 1-8.Senior, J., Delgado, C., Fisher, D., Tilcock, C. and Gregoriaclis, G. (1991) Biochim.Biophys. Acta 1062,77-82.Senior, J., Waters, J.A. and Gregoriadis, G. (1986) FEBS Lett. 196, 54-58.Sessa, G. and Weissmann, G. (1968) J. Lipid Res. 9, 3 10-318.Seymour, L.W., Ulbrich, K., Strohaim, J., Kopecek, J. and Duncan, R. (1990) Biochem.Phar,nacol. 39, 1125-1 131.Silvius, J.R. (1972) in Lipid-Proteins Interactions, Vol. 2. lost, P.R. and Griffith, O.H., eds.Wiley and Sons, New York, Chapter 7.Singhal, A. and Huang, L. (1994) J. Liposome Res. 4, 289-299.Sinha, D. and Karush, F. (1979) Biochem. Biophys. Res. Commun. 90, 554-560.Stavridis, J.C., Deliconstantinos, G., Psallidopoulos, M.C., Armenakas, N.A., Hadjiminas,D.J. and Hadjiminas, J. (1986) Exp. Cell Res. 164, 568-572.Steinman, R.M., Meilman, I.S., Muller, W.A. and Cohn, Z.A. (1983) J. Cell Biol. 96, 1-27.Storm, G., Oussoren, C., Peeters, P.A.M. and Barenholz, Y. (1993) in LiposomeTechnology, Vol ifi. Gregoriadis, G., ed. CRC Press, Boca Raton, FL, 345-383.Straubinger, R.M., Lopez, N.G., Debs, R.J., Hong, K. and Papahadjopoulos, D. (1988)Cancer Res. 48, 5237-5245.Sung, C., Schockley, T.R., Morrison, P.F., Dvorak, H.F., Yarmush, M.L. and Dedrick, R.L.(1992) Cancer Res. 52, 377-384.Sweeney, S.A. and Jonas, A. (1985) Biochim. Biophys. Acta 835,279-290.Swenson, C.E., Stewart, K.A., Hammett, J.L., Fitzsimmons, W.E. and Ginsberg, R.S.(1990) Antimicrob. Agents Chemother. 34, 235-240.158Syrjinen, K.J. (1977) Exp. Path. 15, 348-354.Szoka, F. and Papahadjopoulos, D. (1978) Proc. Nati. Acad. Sci. USA 75,4194-4198.Szoka, F. and Papahadjopoulos, D. (1980) Annu. Rev. Biophys. Bioeng. 9,467-508.Tall, A.R. and Small, D.M. (1977) Nature (London) 265, 163-164.Talmadge, J.E., Lenz, B.F., Klabansky, R., Simon, R., Riggs, C., Guo, S., Oldham, R.K.and Fidler, I.J. (1986) Cancer Res. 46, 1160-1163.Thomas, G.D., Chappell, M.J., Dykes, P.W., Ramsden, D.B., Godfrey, K.R., Ellis, J.R.M.and Bradwell, A.R. (1989) Cancer Res. 49, 3290-3296.Tilcock, C.P.S. (1986) Chem. Phys. Lipids 40, 109-125.Tilcock, C.P.S., Bally, M.B., Farren, S.B. and Cullis, P.R. (1982) Biochemistry 21,4596-4601.Torchilin, V.P. (1985) CRC Crit. Rev. Ther. Drug Carrier Syst. 2,65-115.Torchilin, V.P., Goidmacher, V.S. and Smirnov, V.N. (1978) Biochem. Biophys. Res.Commun. 85,983-990.Torchilin, V.P., Khaw, B.A., Smirnov, V.N. and Haber, E. (1979) Biochem. Biophys. Res.Commun. 89, 1114-1119.Torchilin, V.P., Klibanov, A.L., Huang, L., O’Donnell, S., Nossiff, N.D. and Khaw, B.A.(1992) FASEB J. 6, 2716-2719.Trail, P.A., Willner, D., Lasch, S.J., Henderson, A.J., Hofstead, S., Casazza, A.M.,Firestone, R.A., Hellström, I. and Hellström, K.E. (1993) Science 261, 212-215.Uemura, K. and Kinsky, S.C. (1972) Biochemistry 11,4985-4094.Urdal, D.L. and Hakomori, S. (1980) J. Biol. Chem. 255, 10509-10516.van Berkel, T.J.C., Kruijt, J.K., Spanjer, H.H., Kempen, H.J.M. and Scherphof, G.L. (1993)in Liposome Technology, Vol ifi. CRC Press, Boca Raton, FL., 219-230.van Furth, R. (1988) in Inflammation: Basic Principles And Clinical Correlates. Gallin, 3.1.,Goldstein, I.M. and Snyderman, R., eds. Raven Press, New York, 28 1-295.van Hossel, Q.G.C.M., Stearenberg, P.A., Crommelin, D.J.A., van Dijk, A., van Oost, W.,Klein, S., Douze, J.M.C., de Wildt, D.J. and Hillen, F.C. (1984) Cancer Res. 44, 3698-3705.Vidal, M., Sainte-Marie, J., Philippot, J.R. and Bienvenue, A. (1987) FEBS Lett. 216, 159-163.Vik, S.B. and Capaldi, R.A. (1977) Biochemistry 16, 5755-5762.Vincent, M., Webster, M.H. and Pether, J.V.S. (1992) Lancet 339, 374-375.Warren, B.A. (1979) in Tumor Blood Circulation. Peterson, H-I., ed. CRC Press, BocaRaton, FL., 1-47.159Weder, H.G. and Zumbuehi, 0. (1984) in Liposome Technology, Vol. I. Gregoriadis, G.,ed. CRC Press, Boca Raton, FL., 79-107.Weinstein, J.N. (1987) in Liposomes. Ostro, M.J., ed. Marcel Dekker, New York, 277-332.Weinstein, J.N. and Leserman, L.D. (1984) Pharmacol. Ther. 24,207-233.Weinstein, J.N. and van Osdol, W. (1992) Cancer Res (Suppi.). 52, 2747s-2751s.Weinstein, J.N., Klausner, R.D., Innerarity, T.L., Ralston, E. and Blumenthal, R. (1981)Biochim. Biophys. Acta 647, 270-284.Weinstein, J.N., Yoshikami, S., Henkart, P., Blumenthal, R. and Hagins, W.A. (1977)Science 195,489-492.Wetterau, J.R. and Jonas, A. (1983) .1. Biol. Chem. 258, 2637-2643.Wileman, T.E. (1991) Adv. Drug Delivery Rev. 6, 167-180.Williams, B.D., O’Sullivan, M.M., Saggu, G.S., Williams, K.E., Williams, L.A. andMorgan, J.R. (1986) Br. .1. Rheum. 25, 98-103.Wilschut, J. and Hoekstra, D. (1986) Chem. Phys. Lipids 40, 145-166.Wisse, E. (1970) J. Ultrastruct. Res. 31, 125-142.Wolff, B. and Gregoriadis, G. (1984) Biochim. Biophys. Acta 802, 259-273.Working, P.K., Newman, M.S., Huang, S.K., Mayhew, B., Vaage, I. and Lasic, D.D. (1994)J. Liposome Res. 4, 667-688.Wu, N.Z., Da, D., Rudoll, T.L., Needham, D., Whorton, A.R. and Dewhirst, M.W. (1993)Cancer Res. 53, 3765-3770.Yeo, K-T., Wang, H.H., Nagy, J.A., Sioussat, T.M., Ledbetter, S.R., Hoogewerf, A.J.,Zhou, Y., Masse, E.M., Senger, D.R., Dvorak, H.F. and Yeo, T-K. (1993) Cancer Res.53, 2912-29 18.160


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