Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Immunocamouflage : the biophysical and biological basis of immunoprotection by grafted methoxypoly(ethylene… Le, Yevgeniya 2010

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

Item Metadata

Download

Media
24-ubc_2010_spring_le_yevgeniya.pdf [ 27.3MB ]
Metadata
JSON: 24-1.0069210.json
JSON-LD: 24-1.0069210-ld.json
RDF/XML (Pretty): 24-1.0069210-rdf.xml
RDF/JSON: 24-1.0069210-rdf.json
Turtle: 24-1.0069210-turtle.txt
N-Triples: 24-1.0069210-rdf-ntriples.txt
Original Record: 24-1.0069210-source.json
Full Text
24-1.0069210-fulltext.txt
Citation
24-1.0069210.ris

Full Text

IMMUNOCAMOUFLAGE: THE BIOPHYSICAL AND BIOLOGICAL BASIS OF IMMUNOPROTECTION BY GRAFTED METHOXYPOLY(ETHYLENE GLYCOL)  by YEVGENIYA LE B. Sc., The University of British Columbia, 2004  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Pathology and Laboratory Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  February 2010 ! Yevgeniya Le, 2010  ABSTRACT Development of novel approaches for the direct immunomodulation of allogeneic donor cells would have significant utility in tissue transplantation. Immunocamouflage of cell surfaces by covalently grafted methoxypoly(ethylene glycol) (mPEG; PEGylation) has emerged as a promising approach. While previous studies demonstrated the in vitro and in vivo efficacy of immunocamouflaged allogeneic cells and viruses, the biophysical mechanisms of immunoprotection have not been well defined due to the labile nature of biological samples. To overcome this limitation, polystyrene latex particles (1.2 and 8.0 !m) were used to elucidate the biophysical mechanisms of immunocamouflage via the effects of the chemical and physical properties of the polymer, as well as the consequences of target size. These findings were correlated with the biological studies utilizing human red blood cells and lymphocytes. It was demonstrated that the two biophysical mechanisms were responsible for the immunocamouflage of PEGylated surfaces: 1) hydrodynamic shielding of surface charge; and 2) steric exclusion of macromolecules from the surface. Surface charge camouflage of latex particles and erythrocytes was best achieved with long polymer chains, regardless of the target size. However, inhibition of surface-macromolecule interactions indicated a target size dependence. The biophysical latex model demonstrated that short chain polymers (2 kDa) were more effective at preventing protein adsorption to small beads (1.2 !m), while long chain polymers (20 kDa) exhibited increased efficacy on large particles (8.0 !m). Consistent with the biophysical model, immunocamouflage of lymphocytes (~10 !m) was best achieved using long chain polymers as measured by: 1) inhibition of antigen-antibody binding (CD3, CD4 and CD28); and 2) allorecognition in a 2-way mixed lymphocyte reaction. The biological model also demonstrated that cell surface topography and antigen localization were critical in selecting the optimal polymer size. Importantly, PEGylation did not result in any cellular toxicity at immunoprotective levels that  ii  rendered modified surfaces more biocompatible. Thus, these studies delineated the biophysical mechanisms of immunocamouflage defined by the chemical and physical parameters of the polymer and influenced by target size and surface complexity. Cell or tissue specific optimization of these factors will be critical for the efficient immunocamouflage of allogeneic cells in transfusion and transplantation medicine.  iii  TABLE OF CONTENTS ABSTRACT .................................................................................................................................... ii TABLE OF CONTENTS .............................................................................................................. iv LIST OF TABLES ....................................................................................................................... vii LIST OF FIGURES.................................................................................................................... viii LIST OF ACRONYMS AND ABBREVIATIONS....................................................................... xi ACKNOWLEDGMENTS ........................................................................................................... xiv DEDICATION ..............................................................................................................................xv I. INTRODUCTION ................................................................................................................... 1 1.0  OVERVIEW ...................................................................................................................................1  1.1  HISTORY OF TISSUE TRANSPLANTATION ........................................................................2  1.2  THE BASIS OF ANTIGENICITY IN HUMANS.......................................................................8  1.2.1 1.2.2  1.3  ALLORECOGNITION IN TISSUE TRANSPLANTATION .................................................19  1.3.1 1.3.2 1.3.3 1.3.4  1.4  METHOXYPOLY(ETHYLENE GLYCOL) AND PEGYLATION .................................................... 40 PHYSICAL PROPERTIES OF PEG AND PEG CONJUGATES........................................................ 43 CHEMISTRY OF PEGYLATION........................................................................................................ 47 PEG METABOLISM, TOXICITY AND IMMUNOGENICITY......................................................... 50 PEGYLATED BLOOD CELLS AND TISSUES ................................................................................. 52 PEGYLATED VIRUSES AND HOST CELLS.................................................................................... 56 SUMMARY OF PEGYLATION TECHNOLOGY.............................................................................. 59  HYPOTHESIS AND SPECIFIC AIMS.....................................................................................60  1.6.1 1.6.2 1.6.3  II.  CLINICAL INTERVENTIONS ............................................................................................................ 33 IMMUNOSUPPRESSIVE THERAPY ................................................................................................. 36  IMMUNOCAMOUFLAGE OF ALLOGENEIC TISSUES ....................................................40  1.5.1 1.5.2 1.5.3 1.5.4 1.5.5 1.5.6 1.5.7  1.6  ANTIBODY–MEDIATED ALLORECOGNITION............................................................................. 21 EFFECTS OF ANTIBODY–MEDIATED ALLORECOGNITION..................................................... 25 CELL–MEDIATED ALLORECOGNITION ....................................................................................... 26 EFFECTS OF CELL–MEDIATED ALLORECOGNITION................................................................ 30  CURRENT STRATEGIES FOR OVERCOMING ALLORECOGNITION ........................33  1.4.1 1.4.2  1.5  BLOOD GROUP ANTIGENS ................................................................................................................ 8 MHC CLASS I AND CLASS II ANTIGENS....................................................................................... 15  EXPERIMENTAL HYPOTHESIS ....................................................................................................... 60 SPECIFIC AIMS ................................................................................................................................... 60 SIGNIFICANCE OF CELLULAR PEGYLATION ............................................................................. 62  MATERIALS AND METHODS ......................................................................................... 65  2.0 COMMON METHODOLOGY: BIOPHYSICAL LATEX MODEL AND BIOLOGICAL CELL MODEL ......................................................................................................................................65 2.0.1 2.0.2  2.1  STATISTICAL ANALYSIS ................................................................................................................. 65 FLOW CYTOMETRY .......................................................................................................................... 65  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX MODEL .......................66  2.1.1 2.1.2 2.1.3 2.1.4 2.1.5  ALIPHATIC AMINE POLYSTYRENE LATEX................................................................................. 66 POLYSTYRENE LATEX DERIVATIZATION .................................................................................. 67 MPEG QUANTIFICATION WITH ANTI-PEG ANTIBODY ............................................................ 71 LATEX MICROAGGREGATION STUDIES...................................................................................... 72 PARTICLE ELECTROPHORETIC MOBILITY STUDIES................................................................ 73  iv  2.1.6 HUMAN PLASMA PROTEIN ADSORPTION................................................................................... 75 2.1.7 DESORBED PROTEIN ANALYSIS.................................................................................................... 77 2.1.7.1 HUMAN PLASMA PROTEIN DESORPTION ........................................................................... 77 2.1.7.2 COLORIMETRIC PROTEIN ASSAY ......................................................................................... 78 2.1.7.3 SDS-PAGE ANALYSIS................................................................................................................ 79 2.1.7.4 ITRAQ LABELING AND MASS SPECTROMETRY ANALYSIS ........................................... 80 2.1.8 ADSORBED PROTEIN ANALYSIS ................................................................................................... 82 2.1.8.1 FLOW CYTOMETRY .................................................................................................................. 82 2.1.8.2 QUANTITATIVE MICROSCOPY............................................................................................... 82  2.2  BIOLOGICAL PBMC AND RBC MODEL .............................................................................84  2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7  PBMC AND RBC EXTRACTION ....................................................................................................... 84 CELL DERIVATIZATION................................................................................................................... 84 CELL ELECTROPHORETIC MOBILITY .......................................................................................... 85 CD SURFACE MARKER ANALYSIS................................................................................................ 85 CFSE CELL PROLIFERATION ASSAY ............................................................................................ 86 2-WAY MIXED LYMPHOCYTE REACTION................................................................................... 88 PI AND 7-AAD VIABILITY ASSAYS................................................................................................ 89  III. RESULTS ........................................................................................................................... 91 3.0  OVERVIEW ................................................................................................................................91  3.1  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX MODEL .......................92  3.1.1 FLUORESCEIN-SVA-MPEG GRAFTING TO POLYSTYRENE LATEX PARTICLES ................ 92 3.1.2 QUANTIFICATION OF LATEX-GRAFTED MPEG BY ANTI-PEG ANTIBODY ........................ 94 3.1.3 MICROSCOPIC OBSERVATION OF LATEX AGGREGATION.................................................... 96 3.1.4 MPEG-MEDIATED SURFACE CHARGE CAMOUFLAGE OF LATEX PARTICLES ................. 97 3.1.4.1 PARTICLE ELECTROPHORETIC MOBILITY: EFFECTS OF LINKER CHEMISTRY, GRAFTING CONCENTRATION, POLYMER AND LATEX SIZE.......................................................... 97 3.1.4.2 PARTICLE ELECTROPHORETIC MOBILITY: EFFECT OF POLYMER ARCHITECTURE ... ...................................................................................................................................................... 100 3.1.5 MPEG-MEDIATED INHIBITION OF SURFACE-MACROMOLECULE INTERACTIONS ....... 103 3.1.5.1 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: QUANTITATIVE COLORIMETRIC PROTEIN ASSAY ....................................................................................................... 103 3.1.5.2 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: SDS-PAGE ANALYSIS ...... 105 3.1.5.3 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: ITRAQ AND MASS SPECTROMETRY ANALYSIS ................................................................................................................. 106 3.1.5.4 FLUORESCENT HUMAN PLASMA PROTEIN ADSORPTION: EFFECTS OF LINKER CHEMISTRY, GRAFTING CONCENTRATION AND POLYMER SIZE.............................................. 112 3.1.5.5 FLUORESCENT PLASMA PROTEIN ADSORPTION: EFFECTS OF POLYMER ARCHITECTURE....................................................................................................................................... 114 3.1.5.6 FLUORESCENT PLASMA PROTEIN ADSORPTION: EFFECTS OF TARGET SIZE......... 116 3.1.5.7 FLUORESCENT PLASMA PROTEIN DESORPTION ............................................................ 121 3.1.6 SUMMARY OF RESULTS FOR THE BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX MODEL.............................................................................................................................................. 123  3.2 BIOLOGICAL RED BLOOD CELL AND PERIPHERAL BLOOD MONONUCLEAR CELL MODEL ....................................................................................................................................125 3.2.1 MPEG-MEDIATED SURFACE CHARGE CAMOUFLAGE OF ERYTHROCYTES: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION............................................................................ 125 3.2.2 CAMOUFLAGE OF CD SURFACE ANTIGENS: EFFECTS OF POLYMER SIZE, GRAFTING CONCENTRATION AND TOPOGRAPHY OF THE CELL SURFACE...................................................... 126 3.2.3 PREVENTION OF ALLORECOGNITION IN A 2-WAY MIXED LYMPHOCYTE REACTION: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION .................................................... 132 3.2.4 VIABILITY OF PEGYLATED PERIPHERAL BLOOD MONONUCLEAR CELLS: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION............................................................................ 134 3.2.5 SUMMARY OF RESULTS FOR THE BIOLOGICAL RED BLOOD CELL AND PERIPHERAL BLOOD MONONUCLEAR CELL MODEL .................................................................................................. 136  IV. DISCUSSION ................................................................................................................... 138 v  4.0  OVERVIEW ...............................................................................................................................138  4.1  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX PARTICLE MODEL139  4.2  SUMMARY OF DISCUSSION ON THE BIOPHYSICAL MODEL...................................152  4.3 BIOLOGICAL RED BLOOD CELL AND PERIPHERAL BLOOD MONONUCLEAR CELL MODEL ....................................................................................................................................154 4.4  SUMMARY OF DISCUSSION ON BIOLOGICAL MODEL ..............................................161  4.5  FUTURE DIRECTIONS...........................................................................................................162  4.6  CONCLUSIONS AND OVERALL SIGNIFICANCE ...........................................................169  REFERENCES .......................................................................................................................... 172  vi  LIST OF TABLES Table 1-1: Types and numbers of transplants performed in Canada in 2008. ................................ 7 Table 1-2: Human blood groups. .................................................................................................... 9 Table 1-3: Components carrying blood group antigens that are involved in pathogen invasion. 15 Table 1-4: The properties of human immunoglobulin isotypes.................................................... 23 Table 1-5: The generation of antibody response........................................................................... 24 Table 1-6: PEG pharmaceuticals that are currently approved or undergoing clinical trials......... 42 Table 1-7: Toxicity of PEG (LD50) by molecular weight and animal species. ............................. 51 Table 1-8: Classes of experimental viruses used for PEGylation studies..................................... 58 Table 3-1: Summary of iTRAQ/MS analysis listing the numbers and the types of proteins demonstrating decreased, increased or unchanged relative abundances in the protein layer adsorbed to PEGylated surface in comparison to the bare latex control. ........ 109 Table 3-2: Proteins exhibiting decreased relative abundance on PEGylated latex as identified by iTRAQ/MS analysis.................................................................................................. 110 Table 3-3: Proteins exhibiting increased relative abundance on PEGylated latex as identified by iTRAQ/MS analysis.................................................................................................. 111 Table 3-4: Proteins exhibiting unchanged relative abundance on PEGylated latex, as identified by iTRAQ/MS analysis............................................................................................. 111  vii  LIST OF FIGURES Figure 1-1: Types of integral cell membrane proteins carrying blood group antigens................. 12 Figure 1-2: A schematic diagram of the MHC complex............................................................... 16 Figure 1-3: MHC I complex presents endogenously derived peptides......................................... 17 Figure 1-4: MHC II complex presents exogenously derived peptides. ........................................ 18 Figure 1-5: Antibody and cell-mediated effects of allorecognition.............................................. 20 Figure 1-6: Structure of an antibody molecule. ............................................................................ 21 Figure 1-7: Non-covalent forces enabling antigen-antibody interactions..................................... 22 Figure 1-8: The course of an antibody response........................................................................... 24 Figure 1-9: Structure of the T cell receptor and its specificity for the antigen:MHC complex.... 27 Figure 1-10: Structure of the T cell receptor complex.................................................................. 28 Figure 1-11: T cell activation, proliferation and differentiation require multiple signals from the same antigen presenting cell. ...................................................................................... 29 Figure 1-12: Immunosuppressive drugs act at various stages during T cell activation and proliferation................................................................................................................. 37 Figure 1-13: Exclusion zone of a covalently attached mPEG chain............................................. 45 Figure 1-14: Conformations of surface bound mPEG chains....................................................... 46 Figure 1-15: Immunoprotection zone of grafted mPEG is influenced by surface topography..... 47 Figure 1-16: Lysine ionization states............................................................................................ 49 Figure 1-17: PEGylated RBC in vivo mouse survival studies. ..................................................... 54 Figure 1-18: Immunocamouflage of human lymphocytes............................................................ 55 Figure 1-19: mPEG-mediated viral prophylaxis........................................................................... 58 Figure 1-20: The flowchart of the project..................................................................................... 62 Figure 2-1: Aliphatic amine polystyrene latex particle is the biophysical artificial cell model. .. 67 Figure 2-2: Chemical reaction of SC-, SVA- and SPA-mPEG with primary amines. ................. 68 Figure 2-3: Hydrolysis of SC-, SVA- and SPA-mPEG. ............................................................... 69 Figure 2-4: NHS-Polyglycerol structure and chemical reactivity................................................. 71 Figure 2-5: DELSA electrophoretic mobility measurements. ...................................................... 74 Figure 2-6: Zeta-meter electrophoretic mobility measurements................................................... 75 Figure 2-7: Alexa-488 fluorescent plasma protein adsorption. .................................................... 77 Figure 2-8: Flow chart for methodology in biophysical latex model. .......................................... 78 Figure 2-9: iTRAQ labeling and mass spectrometry analysis. ..................................................... 81 Figure 2-10: CFSE T cell proliferation assay. .............................................................................. 87 Figure 3-1: Modification of 1.2 µm (A) and 8.0 µm (B) particles with 5 kDa fluor-SVA-mPEG demonstrated a dose-dependent increase of the covalently bound polymer as measured by flow cytometry....................................................................................... 93 Figure 3-2: Commercial anti-PEG antibody demonstrated a poor mPEG dose-dependence profile for both 1.2 (A) and 8.0 µm (B) particles. .................................................................. 95 Figure 3-3: PEGylation of 1.2 µm latex with either SC- or SVA-mPEG monofunctional polymer did not result in latex aggregation............................................................................... 97 Figure 3-4: PEGylation of 1.2 µm (A) and 8.0 µm (B) latex particles very efficiently camouflaged latex surface charge as evidenced by decreased electrophoretic mobility. ................................................................................................................................... 100 Figure 3-5: Latex modification with linear monofunctional SVA-mPEG (A) resulted in enhanced surface charge camouflage in comparison to branched multifunctional NHS-PG (B). ................................................................................................................................... 102 viii  Figure 3-6: PEGylation of 1.2 µm particles with SC-mPEG resulted in inhibition of surfacemacromolecule interactions as indicated by decreased unlabeled plasma protein adsorption.................................................................................................................. 104 Figure 3-7: PEGylation, in addition to dramatically decreasing the total protein adsorption, resulted in the differential adsorption of specific plasma proteins. .......................... 106 Figure 3-8: iTRAQ/MS analysis utilized 100 µg of total protein and required substantially more PEGylated latex beads to yield the same amount of desorbed protein as that obtained from the bare latex control (1x). ............................................................................... 107 Figure 3-9: PEGylation of 1.2 µm particles with either SC- (A) or SVA-mPEG (B) resulted in inhibition of surface-macromolecule interactions as indicated by decreased fluorescent plasma protein adsorption. ..................................................................... 114 Figure 3-10: PEGylation of 1.2 µm particles with the linear SVA-mPEG (A) resulted in improved inhibition of fluorescent protein adsorption in comparison to the branched NHS-PG (B).............................................................................................................. 116 Figure 3-11: PEGylation of 1.2 (A) and 8.0 µm (B) particles with SVA-mPEG resulted in a dose, as well as a target size dependent decrease in fluorescent protein adsorption, measured by flow cytometry..................................................................................... 119 Figure 3-12: Prevention of fluorescently labeled protein adsorption is readily observed via microscopic analysis of SVA-mPEG modified 1.2 µm (A, B) and 8.0 µm (C, D) latex particles. .................................................................................................................... 120 Figure 3-13: The amount of fluorescent protein left on the latex particles following protein desorption decreased dramatically in comparison to the amount of protein that was initially adsorbed....................................................................................................... 123 Figure 3-14: PEGylation resulted in the efficient surface charge camouflage of human erythrocytes as indicated by decreased electrophoretic mobility of modified RBCs. ................................................................................................................................... 126 Figure 3-15: Heights of the surface molecules targeted for camouflage, their distribution on the surface and overall surface topography affected the efficacy of immunocamouflage. ................................................................................................................................... 130 Figure 3-16: Covalent modification of human peripheral blood mononuclear cells resulted in a dose-dependent decrease of mean cell fluorescence due to camouflage of CD3 (A), CD4 (B) and CD28 (C) surface antigens. ................................................................. 131 Figure 3-17: Covalent modification of human peripheral blood mononuclear cells resulted in the dose-dependent decrease in percent gated cells due to camouflage of CD3 (A), CD4 (B) and CD28 (C) surface antigens........................................................................... 132 Figure 3-18: PEGylation of one population of human PBMCs resulted in a dose- and polymer size-dependent decrease in allorecognition and T-cell proliferation in a 2-way MLR. ................................................................................................................................... 134 Figure 3-19: Viability of PEGylated human PBMCs was only compromised at high mPEG grafting concentrations and with extensive in vitro cell incubation as measured by PI (A) and 7-AAD (B) viability assays. ........................................................................ 136 Figure 4-1: Surface charge camouflage is primarily driven by polymer-mediated extension of the shear plane (SP) of the surface towards a region of decreased surface potential (Surface Potential Gradient), whereas a prevention of plasma protein adsorption is achieved due to steric exclusion by the surface-grafted polymer. ............................ 142 Figure 4-2: SVA-mPEG resulted in improved biophysical camouflage of surfaces due to its longer hydrolysis half-life......................................................................................... 143 Figure 4-3: The size of the PEGylation target influences which polymer size most effectively imparts immunocamouflage...................................................................................... 148 ix  Figure 4-4: Differences in the efficacy of surface charge camouflage on latex particles vs. RBC by grafted mPEG....................................................................................................... 155 Figure 4-5: Effects of hydrodynamic charge shielding and steric repulsion by grafted mPEG on receptor-ligand interactions. ..................................................................................... 157 Figure 4-6: Inhibition of T cell activation and proliferation by PEGylation vs. immunosuppressive therapy...................................................................................... 159 Figure 4-7: Blood type availability at CBS................................................................................. 164 Figure 4-8: Broad-spectrum antiviral prophylaxis...................................................................... 167 Figure 4-9: The economical considerations for cellular PEGylation.......................................... 168  x  LIST OF ACRONYMS AND ABBREVIATIONS 7-AAD  7-amino-actinomycin D  1º  Primary  2º  Secondary  Å  Angstrom  ANOVA  Analysis of variance  APC  Antigen presenting cell  BSA  Bovine serum albumin  BTC  Benzotriazol carbonate  ºC  Degrees Celsius  CBS  Canadian Blood Services  CD  Cluster of differentiation  CFSE  Carboxyfluorescein diacetate succinimidyl ester  cm  Centimeters  DNA  Deoxyribonucleic acid  EDTA  Ethylenediamine tetraacetic acid  ER  Endoplasmic reticulum  FDA  Food and Drug Administration  FITC  Fluorescein isothiocyanate  Fluor  Fluorescein  HLA  Human leukocyte antigen  Ig  Immunoglobulin  IL  Interleukin  IFN  Interferon  xi  iTRAQ  Isobaric tags for relative and absolute quantification  K  Lysine  kDa  Kilodalton  kg  Kilogram  MCF  Mean cell fluorescence  !g  Microgram  !l  Microliter  !m  Micrometer  m"  Macrophage  mg  Milligram  MHC  Major histocompatibility complex  min  Minutes  MLR  Mixed lymphocyte reaction  mM  Millimolar  mPEG  Methoxypoly(ethylene glycol)  MS  Mass spectrometry  MW  Molecular weight  ng  Nanogram  NHS  N-hydroxy succinimide  NK cell  Natural killer cell  nm  Nanometers  OD  Optical density  PBMC  Peripheral blood mononuclear cell  PEG  Polyethylene glycol  xii  PG  Polyglycerol  PGC  Percent gated cells  PI  Propidium iodide  RBC  Red blood cell  RF  Flory radius  RNA  Ribonucleic acid  RSV  Respiratory syncytial virus  RT  Room temperature  SEM  Standard error mean  SC  Succinimidyl carbonate  SDS-PAGE  Sodium dodecyl sulfate polyacrylamide gel electrophoresis  SP  Shear plane  SPA  Succinimidyl propionate  SV40  Simian virus 40  SVA  Succinimidyl valerate  TA-GVHD  Transfusion associated graft versus host disease  Tc  Cytotoxic T cell  TCR  T cell receptor  Th  Helper T cell  TNF  Tumor necrosis factor  TRALI  Transfusion related acute lung injury  WBC  White blood cell  WHO  World Health Organization  xiii  ACKNOWLEDGMENTS There are many people who have deeply touched my heart and greatly contributed to my growth as a scientist and, most importantly, as a person. First and foremost, I would like to thank my family and friends for being an endless source of love and inspiration. I am indebted to my supervisor, Dr. Mark D. Scott, for his support and guidance in my scientific aspirations. I am forever grateful for his generosity and understanding in my personal matters and for his open-mindedness and non-judgment. Thank you, Mark, for your great sense of humor and for uplifting my spirit during the hard times. I would also like to thank my undergraduate research supervisor, Dr. Donald E. Brooks, for encouraging me to enter the graduate school, supporting me on this journey and believing in my success. I have been blessed to work with a group of wonderful people, who have truly made my research experience fulfilling. I would like to thank Wendy Toyofuku and Dr. Dan Wang for their technical expertise and many helpful scientific discussions. Wendy, thank you for all the hard work that you do in keeping our laboratory functional and organized. I am also grateful for the friendships and support of Janet Tong, Troy C. Sutton (recently a Dr.) and Ibrahim Mustafa (soon to be a Dr.). Without your laughter, jokes and occasional teasing my days in the lab would be very sad and lonely. Finally, I would like to thank my research supervisory committee members, Dr. David Walker, Dr. Elisabeth Maurer and Dr. Juergen Kast. Only through their kind support, expert advice and helpful suggestions was I able to succeed in this project.  xiv  DEDICATION To Mama and Ira  Thank you, for taking care of my heart, my body and soul throughout all these years and for being my true friends and the best teachers. Without your unconditional love, wisdom and support none of this work would be possible.  xv  I.  INTRODUCTION  1.0  OVERVIEW History of tissue transplantation began with blood transfusion and dates back to more  than 400 hundred years ago. In countless efforts at transferring tissues, first from animal, later from human to human, it was slowly recognized that all cells in the human body express specific surface molecules that are unique to an individual and are encoded by his/her DNA. Allorecognition of cell surface antigens by antibodies and lymphocytes of the recipient’s immune system results in health complications and graft rejection. Unfortunately, due to the extremely high level of polymorphism of surface antigens and the complexity of the immune response, current clinical and immunosuppressive means of preventing the severe physiological consequences of allorecognition are often ineffective. Furthermore, the use of toxic pharmacological immunosuppressive drugs results in increased morbidity and mortality of organ transplant recipients. Thus, new, more effective and less toxic methods of overcoming immunological recognition and rejection of foreign tissues are desperately needed. Recently, modification of cell surfaces with non-toxic, non-immunogenic methoxypoly(ethylene glycol) polymer emerged as a promising approach. Due to its chemical and physical properties, this polymer can be directly attached to the surfaces of cells conferring global immunocamouflage. Most importantly, this technology has the potential of inhibiting cell-cell and cell-macromolecule interactions that are the initial and critical steps in antigen recognition and tissue rejection.  1  1.1  HISTORY OF TISSUE TRANSPLANTATION The transfer of cells, tissues and organs from one individual (animal) to another is termed  transplantation, and the transferred material is called a transplant or a graft. The history of transplantation unfolds with the earliest and most common type of transplantation – blood transfusion. In 1665 an English physician, Richard Lower, performed the first recorded successful blood transfusion in dogs. This suggested the possibility of blood transfusions in humans and in 1667 Jean-Baptiste Denis (France) and Richard Lower (England) independently reported the transfusion of blood from lambs to humans. While the outcomes were not desirable, these “experiments” were repeated by others. Over the next 10 years transfusions from different animals were tried and all were deemed unsuccessful due to the mortality of most transfused patients. Thus, in 1678 blood transfusions were prohibited in England and France and the field of transfusion science lay dormant for 140 years. It was not until the 1800’s that doctors revisited blood transfusion as a possible means of therapy and in 1818 James Blundell stated that only transfusions from human to human should be performed. This discovery improved the outcomes of blood transfusions, but still resulted in a high risk of morbidity and mortality. Between 1825 and 1830 Blundell performed 10 obstetric blood transfusions with only five being beneficial to the patient. The reasons for the mortality and morbidity due to transfusions were unknown but were understood to be caused by the blood itself. In an attempt to avoid using blood, American physicians started transfusing human and animal milk based on the erroneous notion that fat particles in milk could be converted to blood cells. This also resulted in adverse reactions and in 1884 milk was substituted by saline infusions. The first breakthrough in transfusion medicine came in 1901 when an Austrian physician, Karl Landsteiner, discovered the ABO blood group system, which earned him a Nobel Prize in 2  medicine in 1930 (1, 2). He asked whether there were small but detectable differences in blood from different individuals within the same species that resulted in incompatibility. His experimental setup consisted of mixing serum and erythrocytes from different individuals and determining if they reacted with one another. Most combinations of erythrocytes and serum resulted in no alterations, but some caused, what we now term, agglutination – serum mediated clumping of erythrocytes. After extensive studies of blood and serum sampled from many different individuals, Landsteiner and his colleagues, Decastrello and Sturli, concluded that red blood cells (RBCs) had “agglutinogens” (antigens) on their surface that differed among individuals. The presence of these surface molecules defined four types of blood: A (A only), B (B only), AB (A and B) and O (no A or B). Moreover, Landsteiner proposed that there were active substances in the serum that bound to antigens and caused agglutination. He called them isoagglutinins and they are now referred to as antibodies. In light of Landsteiner’s discovery of molecules conferring individual blood differences in humans, Hektoen suggested that the blood types of the recipient and the donor should be matched (cross-matched). In 1907 Ottenberg performed the first ABO cross-matched transfusion (3). Cross-matching greatly improved the outcomes of blood transfusions and most efforts became targeted at developing better devices and methodology for performing direct transfusions, i.e., with the blood from the donor being directly infused into the recipient. Just before and during World War I (WWI), Lewinsohn discovered sodium citrate as a blood anticoagulant (4) and Rous and Turner introduced a citrate-glucose solution for storing blood (5). Both of these discoveries allowed for blood storage and indirect transfusions during WWI. With the extensive use of transfusions during the war, it became apparent that there were still differences in individual blood types not overcome by ABO cross-matching. A large portion of transfused patients still developed reactions to foreign blood. In 1940 a second major breakthrough was made by Landsteiner, Wiener, Levine and Stentson. They discovered the Rh 3  blood group system that was soon recognized as the cause of the majority of non-ABO transfusion associated complications (6). In the post World War II era blood banks were setup all across the United States and several countries in Europe. Further discoveries in science and technology, especially the development of synthetic polymers and plastics, used for blood storage, greatly aided the practice of blood transfusion. Cross-matching for both the ABO and the Rh blood groups was routinely performed and transfusion complications became rare. Nevertheless, the search for more molecular “self” markers continued and so far, more than 280 blood group antigens have been identified. The detailed structures and functional roles for many of them are still being elucidated. With the success in blood transfusions, the possibility of organ transfer became an attractive solution for defective organ replacement. The venture into solid organ transplantations began at the turn of the twentieth century with the futile attempts in organ transplantations (e.g., kidney) from animal to human, thus replicating the history of early blood transfusions. Around the same time, the German scientist Loeb was studying the skin transplant models in mice. He performed a series of experiments, where the skin grafts were either from the same mouse (autograft), from a genetically different mouse (allograft) or from a different species (xenograft) (7, 8). He observed that the success of transplantations was directly related to the genetic similarity of the two animals. Analogously to Landsteiner, Loeb concluded that there must be biochemical characteristics that differed in the tissues of separate individuals. The natural assumption was made that the blood antigens responsible for transfusion reactions were also involved in the tissue rejection. This was later shown to be untrue but, nevertheless, suggested that transplantations within species might prove to be more successful. In 1933 a Russian surgeon, Voronoy, performed the first human-to-human kidney transplant, but it remained functional for only 2 days. The next 15 years were fraught with 4  failure until three discoveries revolutionized the world of transplantation. First, in 1948 Snell discovered that a set of genes in mice was responsible for encoding cell surface molecules that were specific to each individual animal. He demonstrated that these genes were responsible for rejection/acceptance of transplanted tumors in mice and called these genes major histocompatibility complex (MHC) (9). Secondly, in 1952 the equivalent of MHC in humans was found by Dausset and Benacerraf. They studied the patterns of skin allograft rejections in pilots and bomb victims during World War II (WWII) and discovered that certain inherited proteins were expressed on human leukocytes and resulted in disparity between individuals. They named these proteins as human leukocyte antigens (HLAs) (10). Thirdly, in the 1950’s Billingham and Medawar, through the use of mouse skin graft models, showed that the basis for failed transplants was in immunological rejection mediated by cells recognizing the foreign MHC surface-molecules on transplanted tissues (11). They have demonstrated that the immune system has the capability of distinguishing between “self” and “non-self” and that this immunity is acquired and is largely shaped during the early years of development. All three discoveries marked the beginnings of transplantation immunology and each was awarded the Nobel Prize. The inherited basis of disparity among cell surface proteins led scientist to try transplantations between closely related donors and recipients. Not surprisingly, the first successful kidney transplant was performed in identical twins by Moore et al., in 1954. In an attempt to overcome the organ rejection in unrelated individuals the earliest achievements in transplantation medicine were made in 1959. Schwartz and Dameshek discovered the immunosuppressive properties of 6-mercaptopurine and Calne demonstrated that azathioprine (AZA) prevented rejection of canine kidney allografts. In 1960 the first kidney transplantation was performed between non-identical twins under immunosuppression by Merrill et al., (12). The use of immunosuppressive therapy allowed for transplantations between unrelated crossmatched individuals and greatly increased graft survival. The availability of 5  immunosuppressive agents marked the beginning of a new era in transplantation medicine and the transplantations of other organs, such as lung (1963 by J. Hardy), liver (1967 by T. Starzl) and heart (1967 by C. Barnard) followed soon after. Over the past decades, the practice of blood and organ transplantation increased dramatically. With blood banks, transplantation centers and organ donor registries set up in most countries of the world, tissue transplantation has become the cornerstone of modern medicine. Approximately 900,000 erythrocyte, 300,000 platelet and 30,043 aphaeresis platelet units are transfused annually in Canada (Canadian Blood Services annual report 2006-2007; http://www.bloodservices.ca/). These transfusions are critical to patients recovering from trauma, undergoing surgery, cancer therapy, transplantation and chronically dependent individuals with haemoglobinopathies, such as thalassemia and sickle cell anemia. Presently, over 2,000 organ transplantations are performed every year in Canada (Table 1-1) in the patients with severe conditions of organ failure due to trauma or acquired (e.g., liver cirrhosis) and genetic disorders (e.g., diabetes type I and leukemia) (http://www.cst-transplant.ca/). The importance of tissue transplantation as the means of therapy for critically ill patients is undeniable. Unfortunately, with the use of foreign human cells and tissues for therapeutic purposes the risk remains in recognition of naturally occurring disparity among molecular markers present on all cells of a human body. This is further complicated by the diversity of the surface molecules among individuals, thus making complete tissue matching of unrelated individuals extremely hard to achieve. In addition, the use of immunosuppressive therapy is only partially successful and the drugs themselves contribute to a considerable degree of morbidity and mortality. Thus, the knowledge gained over centuries of trial and error is still continuing to expand. The field of transplantation immunology that began with the discoveries by Medawar and colleagues remains an active area of research and continues to increase our  6  understanding of the mechanisms responsible for the recognition and elimination of foreign antigens. Table 1-1: Types and numbers of transplants performed in Canada in 2008. Kidney  Enbloc Liver Split Whole Pancreas Intestine Heart Single Bilateral Combination Kidney Liver Pancreas Islet Lung Lung Organs 1,177 27 434 11 21 35 3 162 31 100 82 Source: Canadian Organ Replacement Register (CORR), Canadian Institute for Health Information (CIHI) (http://secure.cihi.ca/cihiweb/splash.html).  7  1.2  THE BASIS OF ANTIGENICITY IN HUMANS All tissue transplantations performed in humans, unless from the same individual  (autograft) or identical twin (syngeneic graft), are allogeneic, i.e., coming from an unrelated individual of the same species. The graft in this case is an allograft and its recognition and subsequent rejection is mediated by allorecognition of foreign donor antigens – alloantigens. All cells in the body carry surface antigens that are unique to an individual and are dictated by his or her genetic makeup. These cell surface markers are highly polymorphic [poly, meaning many and morphe, meaning shape or structure (Greek)] and are thus defined by the withinspecies variation at a gene locus that results in the production of proteins of varying structure. The two most important classes of polymorphic surface molecules in the context of transfusion and transplantation medicine are blood group antigens and human leukocyte antigens (HLAs).  1.2.1  BLOOD GROUP ANTIGENS Currently, the International Society for Blood Transfusion (ISBT) recognizes more than  280 blood group antigens constituting 29 blood groups (13-15). Blood group antigens vary in structure, level of polymorphism, extent of antigenicity and relative abundance on the cell surface. They are not only expressed on erythrocytes but are found throughout the human body, thus contributing to the recognition of foreign antigens in other non-blood cell and tissue types, Table 1-2.  8  Table 1-2: Human blood groups. Name  Function  Structure  Component Name  Antigens  Gerbich  Structure  Type I singlepass  Glycophorins C and D  MNS  Structure  Type I singlepass  Glycophorins A and B  Globoside  Structure  Glycolipid  Globotetraosylceramide  Ge2, Ge3, Ge4, Wb, Lsa, Ana, Dha M, N, S, s, U, He, Mia, Vw, +35 more P  Diego  Structure and Transport  Multi-pass  Band 3  Kx  Structure and Transport  Multi-pass  Kx glycoprotein  Rh  Structure and Transport  Multi-pass  RhCE, RhD  Kidd  Transport  Multi-pass  Gil Colton  Transport Transport  Duffy  Indian  Copy Number/RBC GPC: 135,000 GPD: 50,000  Presence in Other Tissues Fetal liver, renal endothelium  GPA: 1,000,000 GPB: 200,000  Renal endothelium and epithelium Mesodermally derived tissues Granulocytes, kidney  Dia, Dib, Wra, Wrb, Wda, Rba, +14 more Kx  1,000,000  RhCE+RhD: 100,000-200,000  Kidd glycoprotein  D, C, E, d, e, f, Cw, V, G, +36 more Jka, Jkb, Jk3  Multi-pass Multi-pass  Aquaporin 3 Aquaporin I  GIL Coa, Cob, Co3  Receptor  Multi-pass  Fy glycoprotein  Fya, Fyb, Fy3, Fy4, Fy5, Fy6  13,000-14,000  Adhesion  Type I singlepass  CD44  Ina, Inb  2,000-5,000  1,000  Fetal liver, adult skeletal muscle, brain, pancreas, heart  14,000  Vasa recta endothelium, renal medulla vascular supply  120,000-160,000  Kidney, liver, gallbladder, eye, capillary endothelium Endothelial, epithelial cells, Purkinje cells, colon, lung, spleen, thyroid, thymus, kidney Wide tissue distribution  9  Name  Function  Structure  Component Name  Antigens  Lutheran  Receptor and Adhesion  Type I singlepass  Lutheran glycoprotein, B-CAM – IgSF  Lua, Lub, Lu3, Lu4, Aua, Aub, +12 more  LandsteinerWiener  Receptor and Adhesion  Type I singlepass  LW glycoprotein – IgSF  LWa, LWab, LWb  Ok  Receptor and Adhesion Receptor and Adhesion  Type I singlepass GPI-linked  CD147 – IgSF  Oka  CDw108, semaphorin  JMH1  Xg  Receptor and Adhesion  Type I singlepass  Xga glycoprotein  Xga  9,000  Kell  Enzyme  Type II singlepass  Kell glycoprotein  K, k, Kpa, Kpb, Ku, Jsa, Jsb, +16 more  3,500-17,000  Yt  Enzyme  GPI-linked  Acetylcholinesterase  Yta, Ytb  10,000  Chido/Rodgers  Complement component Complement component  Adsorbed to blood cells GPI-linked  C4A/C4B  CH1, CH2, Rg1, +6 more Cra, Tca, Tcb, Tcc, Dra, Esa, IFC, WESa, WESb, UMC  Not determined  John Milton Hagen (JMH)  Cromer  Decay accelerating factor (CD55)  Copy Number/RBC 1,500-4,000  D+: 4,400 (adult); 5,150 (cord) D-: 2,835(adult); 3,620 (cord) Not determined  20,000  Presence in Other Tissues Fetal liver, placenta, arterial walls, BM epithelium  All cells tested Lymphocytes, peripheral blood cells, thymus Fibroblasts, fetal liver, spleen, thymus, adrenal, adult BM BM, fetal liver, testis, brain, lymphoid tissue, heart Granulocytes, innervated tissue incl. brain and muscle Plasma Vascular endothelium, epithelia of GI, GU, CNS. Soluble form in plasma, urine  10  Name  Function  Structure  Component Name  Antigens  Copy Number/RBC 20-1,500  Knops  Complement component  Type I singlepass  Complement receptor I (CD35)  Kna, Knb, McCa, SIa, Yka  ABO  Glycocalyx  Carbohydrate on GL  Carbohydrate  A, B, AB, A1  250,000>1,000,000  Hh  Glycocalyx  Carbohydrate on GL  Carbohydrate  H  Group O adult: 1,700,000, lower on other phenotypes  Lewis  Glycocalyx  Carbohydrate  Lea, Leb, Leab, LebH, ALeb, BLeb  Not determined  P  Glycocalyx  Carbohydrate on GL, adsorbed onto RBC Carbohydrate on GL  Carbohydrate  P1  500,000  I  Glycocalyx  Carbohydrate  I1  Scianna Dombrock  Unknown Unknown  Carbohydrate on GL and glycoproteins Unknown Unknown  Sc glycoprotein Do glycoprotein  RAPH  Unknown  Unknown  Glycoprotein  Sc1, Sc2, Sc3 Doa, Dob, Gya, Hy, Joa MER2  Presence in Other Tissues Blood cells, glomerular podocytes, follicular dendritic cells Epithelial cells, secretions, ectoderm and endoderm Broad distribution, soluble-all fluids except CSF in secretors Blood cells, GI, skeletal muscle, kidney, adrenal Blood cells, soluble form in cyst fluid Most human cells and glycoproteins in fluids  Not determined Not determined Not determined  Fibroblasts  IgSF = imunoglobulin superfamily, BM = bone marrow, GP = glycophorin, GL = glycolipid, GI = gastrointestinal, GU = genitourinary, CNS = central nervous system, RBC = red blood cell, CD = cluster of differentiation, GPI = glycosylphosphatidylinositol. Modified from Reid 2000 (14) and Daniels, 2002 (15).  11  Blood group antigens that were described so far have either a carbohydrate or a protein origin. Carbohydrate antigens can be found on the cell surface glycolipids or glycoproteins and define antigens of ABO, Hh, Lewis, I and P blood groups. The remaining 24 blood groups are of polypeptide nature and are found on integral membrane proteins. Structurally, these proteins can be divided into 3 broad classes: single pass (type I and II), multi-pass and glycosylphosphatidylinositol (GPI) anchored, Figure 1-1. Functionally, proteins carrying blood group antigens are divided into structural proteins, transport proteins, receptors and adhesion molecules, enzymes, complement elements/proteins, and components of the glycocalyx (glycoprotein and polysaccharide coating of cell surfaces), Table 1-2 (15).  Figure 1-1: Types of integral cell membrane proteins carrying blood group antigens. Blood group antigens can be found on integral membrane proteins that are structurally divided into single pass (A and B), multi-pass (C and D) and glycosylphosphatidylinositol (GPI) anchored proteins (E). A) Type I single pass proteins span the cell membrane once and have their N-terminus in the extracellular space. B) Type II single pass proteins also span the cell membrane once, but have their N-terminus in the intracellular space. C) Multi-pass proteins can span the cell membrane an odd number of times, in which case they have their C-terminus in the intracellular space or D) an even number of times, in which case they have both the N- and the C-termini in the intracellular space. E) Some proteins do not span the membrane and instead are anchored to the membrane via the covalent attachment to a glycolipid (GPI). Modified from Pathak 2005 (16). 12  Despite the significant progress in cloning and structural studies of blood group antigens, little is known about their function. This is partially due to the fact that polymorphism (difference in amino acids or sugars) of these antigens usually does not result in any functional alterations to the component (glycolipid or glycoprotein) that carries them. This is further complicated by the fact that most studies were performed on RBC, a relatively simple cell, biologically speaking, and the data on the functional roles of blood group antigens of other cell types is limited. In particular, very little is known about carbohydrate antigens, apart that they contribute to the glycocalyx and might protect the cell from mechanical damage and pathogen invasion. However, the available information on protein antigens is briefly summarized in Table 12, based on the recent reviews (14, 15, 17). Structural proteins and proteins with transport functions, that carry blood group antigens, are most easily understood. Structural proteins typically span the membrane once, whereas transport proteins span the membrane multiple times. Gerbich blood group is found on glycophorins C and D. These proteins serve as anchors between the lipid bilayer and the membrane skeleton, thus maintaining the mechanical stability and morphology of the cell. Band 3 of Diego blood group is an anion transporter, Kidd glycoprotein is a urea transporter and the Colton glycoprotein is a water channel. The proteins carrying Rh blood group antigen are presumed to be ammonium transporters. Several blood groups are found on receptors and adhesion proteins. Duffy glycoprotein or Duffy antigen receptor for chemokines (DARC) is a chemokine receptor that binds proinflammatory cytokines and is involved in immune functions. CD44 carries Indian antigen and is the major receptor for hyaluronan. Lutheran, LW, Ok and Xg blood group systems are found on immunoglobulin superfamily (IgSF) type proteins. Lutheran protein and B-CAM, that carry Lutheran blood group, are receptors for laminin. The functional roles for LW, Ok and Xg blood groups are presently unknown. 13  Enzymes can also be components for blood group antigens. For example, Kell blood group resides on Kell glycoprotein that converts endothelin-3 to its bioactive form. Yt antigens are found on acetylcholinesterases. Antigens of Chido/Rogers, Cromer and Knops blood groups are located on complement proteins. C4a and C4b carry Chido/Rodgers groups. Cromer is found on decay-accelerating factor (DAF) and Knops is carried on complement receptor I (CRI). Both DAF and CRI, inhibit the classical and alternative complement cascades and protect the cells from complementmediated lysis. Some adhesion and receptor proteins carrying blood group antigens have been implicated in pathogen invasion, Table 1-3. In fact, most blood group antigens are expressed in tissues other than red cells, for example, epithelium of the gastrointestinal and urogenital tracts and lungs, Table 1-2. Therefore in those tissues, blood group antigens can mediate pathogen entry and establishment of infection. For example, Duffy glycoprotein is the receptor for Plasmodium vivax merozoites and is essential for the invasion of RBC by this parasite. Similarly, decayaccelerating factor (DAF) carrying the Cromer blood group system is the receptor for Escherichia coli and enterovirus. There is a variable degree of polymorphism among blood group antigens. In some cases the whole carrier glycoprotein or glycolipid is missing and confers the null phenotype, in others single amino acid or sugar differences generate a new antigen. For example, the Gerbich phenotype Ge:-2-3-4 results in elliptocytosis of RBC with glycophorins C and D being completely absent. On the other hand, antigens of the Diego blood group are carried on Band 3 protein and single amino acid substitutions of Lys658Glu and Leu854Pro result in Wra/Wrb and Dia/Dib antigens, respectively. The biological significance of polymorphism among blood group antigens is not known. However, since many glycoproteins and glycolipids carrying these  14  antigens are utilized by pathogens to gain access into the cell, it is hypothesized that polymorphism evolved as a selective advantage for evading infection.  Table 1-3: Components carrying blood group antigens that are involved in pathogen invasion. Component/Blood group antigen Protein Blood Groups Duffy glycoprotein/Duffy Glycophorin A/MNS CD44/Indian (AnWj) Decay accelerating factor/Cromer Complement receptor I/Knops (SIa) Carbohydrate Blood Groups Lewis P  Infectious Agent Plasmodium vivax Plasmodium falciparum Haemophilus influenzae Escherichia coli, enterovirus Plasmodium falciparum Helicobacter pylori Escherichia coli, B19 parvovirus  Modified from Reid 2000 (14).  1.2.2  MHC CLASS I AND CLASS II ANTIGENS The most polymorphic antigen system in humans is encoded by the major  histocompatibility complex or human leukocyte antigen (HLA) genes. These genes code for 2 classes of cell surface proteins: MHC class I and MHC class II. The class I DNA region consists of HLA-A, HLA-B and HLA-C genes encoding a heavy (!) chain of class I surface proteins. The class II DNA region consists of DR, DP, DQ gene families encoding ! and " chains of class II surface proteins (18). The major function of MHC molecules is to bind peptides derived from within or outside the cell and present them to circulating T lymphocytes (T cells). This antigen presentation is crucial to immune surveillance for detection of foreign “invaders”. The MHC I protein complex consists of a heavy ! chain, made up of three subunits !1-!3, and a non-covalently associated invariable chain - "2 microglobulin, Figure 1-2A. The !1 and !2 domains make up the binding  15  site of MHC I that accommodates peptides of 8-10 amino acid residues. The MHC II protein complex consists of heterodimers of two non-covalently associated polypeptide chains: ! and ", Figure 1-2B. The !1 and "1 domains form an antigen-binding groove that accommodates peptides of 13-17 amino acids. The HLA polymorphism is clustered in the antigen binding site so that amino acid variations change the shape of the groove and alter the antigen-binding specificity of the whole MHC complex. Thus, the polymorphic nature of MHC proteins allows for the efficient binding of peptides of an enormous range of sequences.  Figure 1-2: A schematic diagram of the MHC complex. A) The MHC I complex is a heterodimer of an ! chain and a non-covalently associated "2microglobulin. An ! chain spans the membrane and consists of 3 subunits: !1, !2 and !3. The peptide binding cleft of MHC I is formed by !1 and !2 subunits. B) The MHC II complex is a heterodimer of ! and " chains, each folded into two subunits, designated 1 and 2. The peptide-binding cleft is wider than that of the MHC I complex and is formed by the !1 and "1 domains. Both the ! and " chains span the membrane.  The functional differences between class I and II molecules are derived from the nature and the source of peptides being bound by these protein complexes (19, 20). MHC I molecules present peptides derived from inside the cell, Figure 1-3. In the cytosol all “self” and foreign proteins are constantly degraded into short peptides by the proteosome complex. Processed peptides are transported into the endoplasmic reticulum (ER). At the same time, newly synthesized MHC I polypeptides are translocated into the ER to be folded and transported to the surface of the cell. Consequently, inside the ER, processed “self” and “non-self” peptides can 16  become bound by the newly folded MHC I chains. If the complementary peptide:MHC pair is found the whole assembly is exported to the membrane for T cell presentation. The unpaired MHC complex is retained in the ER until it binds a peptide of the correct sequence.  Figure 1-3: MHC I complex presents endogenously derived peptides. (1) Peptides are degraded in the cytosol and transported into the endoplasmic reticulum (ER). Newly synthesized MHC I complex also becomes translocated into ER. (2) Inside ER, processed peptides become bound by MHC proteins of correct specificity. (3) Paired MHC:peptide complex is transported to the membrane for presentation to T cells.  The MHC class II complex presents exogenously derived peptides, Figure 1-4. This is achieved by antigen presenting cells (APCs) constantly taking up extracellular material by phagocytosis (cellular “eating”) or pinocytosis (cellular “drinking”). Inside the endosomes engulfed matter is degraded into short peptides by proteases. Concurrently, newly synthesized MHC II polypeptides are transported from ER to the cell membrane inside the vesicles. These intracellular vesicles fuse with endosomes and processed peptides can become bound by the MHC II complex with the correct antigen specificity. The antigen:MHC II complex is delivered to the cell surface for presentation to T cells. Those MHC complexes that did not bind a complimentary peptide become degraded inside the endosomes.  17  Figure 1-4: MHC II complex presents exogenously derived peptides. (1) Peptides are taken up and transported into the cell in endosomes. (2) Inside the cell, endosomes combine with vesicles containing newly synthesized MHC II proteins. Processed peptides become bound by the complementary MHC molecules. (3) Assembled MHC:peptide complex is delivered to the surface for presentation to T cells.  The class I and II proteins are differentially expressed on cells. Class I proteins are expressed on almost all nucleated cells, thus enabling them to present pathogenic antigens derived from intracellular “invaders” (e.g., viruses). In contrast, class II molecules present peptides derived from extracellular pathogens (e.g., bacteria and parasites) and mediate activation of the effector cells of the immune system. As a result, MHC II is only present on specialized antigen presenting cells such as B lymphocytes, dendritic cells and macrophages. Human platelets and RBC contain none or very little MHC I; and no MHC II molecules (18).  18  1.3  ALLORECOGNITION IN TISSUE TRANSPLANTATION Allorecognition is the process by which the human immune system recognizes and  removes alloantigens, whether in the form of blood group or HLA type surface molecules. Antigen recognition can be mediated by soluble factors in the plasma (e.g., antibodies, Ab) or cells of the immune system (e.g., T and B cells). Although the initial stages of each type of allorecognition process are distinct, the two pathways are tightly connected reinforcing each other and sharing a number of soluble and cellular mediators, Figure 1-5. There is substantial cross-talk between the two pathways and both mechanisms of allorecognition converge at the same result – elimination of the foreign antigen. Unfortunately, alloantigen elimination is accompanied by severe health complications and tissue damage for the individual that mounted the immune response with the consequent allograft failure and rejection. Currently, these effects of allorecognition are the major impediment to the safe practice of transfusion and organ transplantation resulting in a considerable degree of morbidity and mortality of tissue transplant recipients. Although transfusion-associated complications are mostly mediated by antibody responses and organ rejection is largely sustained via cellular immunity, both the antibody and the cell-mediated allorecognition processes play important roles in transfusion and transplantation immunobiology, Figure 1-5.  19  Figure 1-5: Antibody and cell-mediated effects of allorecognition. The primary step of allorecognition is the binding of foreign antigens on the surface of transfused RBCs or transplanted tissues. This is achieved by antibody- (Ab) [1A] and/or the T cell-mediated [1C] recognition of the alloantigens. The next stage of allorecognition involves effector functions of the Abs [2A] or the T cells [2C]. [2A] Antibodies opsonize foreign cells and tissues for recognition by complement, attract cells of innate immunity e.g., macrophages (m#), and stimulate the release of inflammatory mediators and toxic substances. [2C] T cells directly (cytotoxic T cells) damage foreign cells by releasing cytotoxic mediators. T cells also stimulate B cells to produce Ab and m# to produce toxic substances. T cell stimulated production of Ab is an important consequence of alloimmunization. There is cross-talk between the Ab and the cell-mediated effects. [3] Allorecognition results in cell/organ death and the recipient’s tissue damage. 20  1.3.1  ANTIBODY–MEDIATED ALLORECOGNITION The first and the most critical step of humoral (antibody-mediated) immunity is the  binding of an antigen by an antibody of correct specificity. An antibody consists of 4 polypeptide chains: 2 identical light chains (L) and 2 identical heavy chains (H); held together by disulfide bonds, Figure 1-6. Each of the chains consists of a variable domain (VH and VL) and a constant domain (CH, CL).  Figure 1-6: Structure of an antibody molecule. An antibody consists of four chains: two identical heavy chains (H) and two identical light chains (L) held together by disulfide bonds. There is a constant region (Fc; CL+CH), which determines the isotype of an antibody and a variable region (Fab; VH+VL), which determines the specificity for an antigen. There are two identical antigen binding sites formed by two pairs of variable domains from one light (VL) and one heavy chain (VH). Modified from Janeway 2005, p. 104 (21).  The variable domains of each chain (VH+VL) make up one variable region (Fab) or a site where the antigen binds. Small variations of amino acid sequence in a variable region result in a different structure of the binding groove and determine the antibody specificity. There are approximately 108 different antibody specificities in humans (22). Because the two heavy and the two light chains are identical, each molecule of an antibody will have two identical antigen binding sites. The strength and success of antigen-antibody binding depends on the amino acid contacts inside the binding groove and the complementarity in shape and size of the epitopes or antigenic determinants – conformational shapes on the surface of antigen. Importantly, the  21  antigen-antibody interactions involve non-covalent forces: electrostatic, hydrogen bonds, Van der Waals, and hydrophobic, Figure 1-7.  Figure 1-7: Non-covalent forces enabling antigen-antibody interactions. Partial charges of electric dipoles are designated by $. Electrostatic forces decrease as 1/r2, where r is the distance between the two atoms. Van der Waals forces are much weaker than electrostatic forces, because the strength of the interaction diminishes as 1/r6. Covalent bonds never occur between an antigen and an antibody. Modified from Janeway 2005, p. 114 (21). The constant domains of each chain (CH+CL) make up a constant region (Fc) of an antibody molecule, which determines the class of the antibody (isotype) and its consequent effector functions (e.g., complement activation, immune cell activation, placental crossover). There are five antibody isotypes: immunoglobulin M (IgM), immunoglobulin D (IgD), immunoglobulin G (IgG), immunoglobulin A (IgA), and immunoglobulin E (IgE). IgG is further subdivided into 4 subclasses, IgG1-IgG4. IgA and IgM are capable of forming multimers, thus increasing their efficacy of antigen binding and removal, Table 1-4.  22  Table 1-4: The properties of human immunoglobulin isotypes. Molecular weight (kDa) Serum level in adult (mg/ml) Half-life in serum (days) Complement/classical pathway Complement/alternative pathway Placental transfer Binding to m# and Fc receptors Binding to mast cells/basophils  IgG1 146 9 21 ++ -  IgG2 146 3 20 + -  IgG3 165 1 7 +++ -  IgG4 146 0.5 21 -  IgM 190 1.5 10 +++ -  IgA1 160 3.0 6 +  IgA2 160 0.5 6 -  IgD 184 0.03 3 -  IgE 188 5*10-5 2 -  +++ + -  + -  ++ + -  -/+ -/+ -  -  + -  + -  -  + +++  All molecular weight values are for monomeric antibodies. IgM forms a pentamer of 970 kDa, while IgA forms a dimer of 390 kDa. M# – macrophage, Fc receptor – phagocyte receptor for Fc region of an antibody. Modified from Janeway 2005, p. 156 (21).  Antibodies are produced by the specialized immune cells – B cells. On their surface, B cells express receptors that are analogous in structure to an antibody, but instead of the constant region have a transmembrane domain inserted into the lipid bilayer. These membrane bound immunoglobulins (mIg) are capable of binding extracellular antigens. Upon recognition of an alloantigen, B cells become activated through the help of T helper cells that had previously detected the same antigen. B cell activation occurs through the expression of the cluster of differentiation (CD) 154 (CD40 ligand) on a T cell, which interacts with CD40 on the B cell. Additionally, T cells produce and release the B cell activating interleukins (IL) 4, IL-5 and IL-6 cytokines. These events result in B cell stimulation and differentiation into an antibody secreting plasma cell that produces soluble forms of the same membrane immunoglobulin that initially detected the antigen. The first encounter (priming) by a naïve B cell of an antigen results in antibody production following 4-5 days (lag phase), Figure 1-8. This is the amount of time it takes for a B cell to become stimulated by a T helper cell, replicate and differentiate into a plasma cell. The initial set of antibodies generated is usually of the IgM isotype and of a low affinity, Table 1-5. A subset of the activated B cells that produced the antibody differentiates into memory B cells. These cells are long-lived and persist independent of the presence of the antigen that originally 23  induced them. A re-stimulation with a previously encountered antigen generates a secondary antibody response. It is carried out by already differentiated and primed memory B cells and, therefore occurs more rapidly, within 1-2 days, Figure 1-8. The frequency of B cells that respond to an antigen in the secondary response increases 100-fold and the antibody that is generated is of higher affinity and, mostly, of the IgG isotype, Table 1-5. This heightened ability of lymphocytes to respond to previously encountered antigens forms the basis of immunological memory and allows for a rapid and a more efficient antigen removal.  Figure 1-8: The course of an antibody response. The first encounter of antigen A (Day 0) or antigen B (Day 60) by naïve B lymphocytes results in a primary antibody response. It takes 4-5 days to produce the antibody against the antigen. The antibody concentration in plasma gradually increases, plateaus and subsides after ~20 days. The restimulation with the previously encountered antigen A (Day 60) results in a secondary antibody response. In this case the antibody is produced in 1-2 days by differentiated memory B cells. The concentration of the antibody in plasma increases dramatically and is maintained at a high level to allow for the efficient clearance of the antigen. Modified from Janeway 2005, p. 21 (21).  Table 1-5: The generation of antibody response. Frequency of specific B cells Isotype of antibody produced Affinity of antibody  Primary antibody response 1:104 – 1:105 IgM > IgG Low, Ka = 106  Secondary antibody response 1:102 – 1:103 IgG, IgA High, Ka = 107  Modified from Janeway 2005, p. 448 (21).  24  1.3.2  EFFECTS OF ANTIBODY–MEDIATED ALLORECOGNITION The most detrimental effector functions of antibodies in the context of transfusion and  transplantation medicine are the activation of the complement and the coagulation pathways and the recruitment of phagocytic (e.g., macrophages) and cytotoxic (e.g., natural killer; NK) immune cells. Preformed antibodies (e.g., anti-AB/RhD or anti-HLA) bind to the donor alloantigens either on the surface of RBCs or on the cells of a transplanted organ. This results in complement activation and the release of C3a and C5a proteins into plasma to act as potent anaphylatoxins. Activation of C5b complement protein results in the downstream formation of the membrane attack complex (MAC) that is responsible for cell lysis and damage to the transplanted tissue. In addition to this, complement activation results in the release of thromboplastic substances that may lead to disseminated intravascular coagulation (DIC) resulting in vasoconstriction, renal failure (transfusion) and organ death (transplantation). The cumulative effect is the production of inflammatory cytokines and chemotactic agents. Cytokines and chemokines attract phagocytic and cytotoxic cells, that also mediate tissue and organ damage by releasing toxic substances, such as nitric oxide (NO), superoxide anion (O2-), hydrogen peroxide (H2O2) and perforins (23). In addition to having naturally occurring antibodies, humans can become primed to develop antibodies against newly encountered antigens, Figure 1-8. Such an immunization of a patient to alloantigens by previous exposure is termed alloimmunization. The risk of developing new antibodies increases with the number of times the patient is exposed to allogeneic material. For example, 25-30% of leukemic patients and up to 80% of aplastic anemia patients develop new anti-HLA antibodies due to previous transfusions (18). In addition, 17-26% of women with !3 pregnancies become immunized to paternal HLA antigens. Alloimmunization to HLA and blood group antigens is the leading cause of clinically significant delayed hemolytic transfusion reactions (DHTR), platelet refractoriness, transfusion-related acute lung injury (TRALI), 25  transfusion associated graft versus host disease (TA-GVHD), and febrile non-hemolytic transfusion reactions (FNHTR) (18). In transplantation, antibody-mediated effects contribute to hyperacute and acute rejection of the graft (24). Many histopathological changes observed during acute rejection occur due to antibody binding to the endothelial cells of the graft causing activation of complement (25). Hyperacute rejection occurs in 80% of matched transplantations and 40-60% of them have decreased graft survival even following a thorough screening of the recipient for the presence of anti-donor antibodies. Depending on the organ transplanted, acute rejection occurs in 20-50% of patients and is the leading cause of graft loss during the first year after transplantation (26). In renal transplantations, 30% of kidneys undergo acute rejection with 10% resulting in a complete loss of the graft. In addition, acute rejection is associated with poor prognosis for long-term graft survival and an increased risk of chronic rejection.  1.3.3  CELL–MEDIATED ALLORECOGNITION Cell-mediated allorecognition is dependent on the T cell’s recognition of the antigen  presented in the context of MHC I and MHC II complexes. This is achieved through the binding of antigen:MHC complex by the T cell receptor (TCR) with the correct antigen specificity. The interactions between the TCR and the antigen:MHC complex are analogous to antigen-antibody binding and are maintained by the same non-covalent forces, Figure 1-7. A T cell receptor is structurally similar to the Fab region of an antibody molecule, Figure 1-9. It consists of ! and " chains (% and & in rare cases) bridged by a disulfide bond. Just like an antibody molecule, the TCR has a variable and a constant region. The variable region is formed by the variable domains of the ! and " chains. The small amino acid differences in the variable regions of each chain are responsible for generating the diversity of the T cell receptors and their ability to recognize a wide range of antigenic determinants. In contrast to an antibody, a TCR 26  has only one antigen binding site and recognizes both the antigenic epitope and the MHC molecule (antigen:MHC complex) (27), Figure 1-9.  Figure 1-9: Structure of the T cell receptor and its specificity for the antigen:MHC complex. TCR is a heterodimer of two transmembrane chains: ! and ". The two chains are held together by a disulfide bond (S-S). The variable region determines the TCR antigen specificity and is formed by the variable domains of ! and " chains (V!+V"). The constant region is formed by constant domains of ! and " chains (C!+C"). In order for the antigen binding and recognition to occur, TCR has to be complementary in structure not only to the antigen (Ag), but also to the MHC molecule itself.  The interaction between the TCR on the T cells and the peptide:MHC complex on the antigen presenting cells (APCs) is enhanced by surface co-receptor molecules. There are two co-receptors that define two different populations of T cells with distinct effector functions. The first population expresses a surface protein – CD4 and the second population expresses CD8. CD4 and CD8 co-receptors bind to MHC II and MHC I complexes, respectively (28). In addition to co-receptors, the TCR utilizes the help of a CD3 surface complex and CD247, which is present as mostly an intracytoplasmic homodimer of ' chains, Figure 1-10. The CD3 complex is composed of two heterodimers of 3 invariant chains: %, $ and (. Together, the CD3 complex and CD247 are responsible for the propagation of the cell signaling pathways inside a T  27  cell via the actions of the cellular kinases (29, 30). The CD3 complex and CD247 are closely associated with the TCR and cumulatively are referred to as a TCR complex, Figure 1-10.  Figure 1-10: Structure of the T cell receptor complex. The T cell receptor complex consists of the T cell receptor (TCR), the CD3 complex and CD247. The CD3 complex consists of two heterodimers of three chains: %, $ and (. The CD3 complex and CD247 are necessary for T cell activation as they propagate the T cell activating signal into the cell and initiate a signal transduction pathway utilizing intracellular kinases. Modified from Janeway 2005, p. 214 (21).  The initial interaction between a T cell and an APC is mediated by cell-adhesion molecules. Lymphocyte function-associated antigen-1 (LFA-1 or CD11a/CD18), intercellular adhesion molecule-3 (ICAM-3, CD50) and CD2 on T cells bind to ICAM-1/2 (CD54/CD102), DC-SIGN (CD209) and CD58 on APCs, respectively. This interaction holds the two cells together so that TCR:CD4/CD8 and antigen:MHC I/II binding can take place (31, 32), Figure 111. This is the first T cell activating signal that is sent through the CD3 complex. A second signal is required from the same APC in order for a T cell to become fully activated and to undergo proliferation and clonal expansion, or to produce daughter T cells with the same antigen specificity. This second, co-stimulatory, signal is delivered by the interaction of B7 molecules (B7.1=CD80 and B7.2=CD86) on an APC with CD28 on a T cell (33), Figure 1-11. Once a T cell becomes activated, it expresses additional proteins to sustain a co-stimulatory signal and to contribute to further activation and differentiation, or to downregulate these 28  processes. For example, CD40 ligand of a T cell (CD40L or CD154) that binds to CD40 on an APC leads to the expression of B7 molecules on an APC and further T cell activation (34) (Figure 1-11). In contrast, the expression of CTLA-4 on a T cell (CD152) and its binding to B7 molecules on an APC results in the inhibition of the T cell activation.  Figure 1-11: T cell activation, proliferation and differentiation require multiple signals from the same antigen presenting cell. [1] The first step in T cell activation is the adhesion between an antigen presenting cell (APC) and a T cell via adhesion molecules, e.g., CD54 (ICAM-1):CD11a/CD18 (LFA1). [2] The second step is the recognition of antigen:MHC II complex by the T cell receptor (TCR). [3] The TCR:MHC II interaction is aided by the co-receptor, CD4, binding to MHC II. The two initial receptor binding steps [2] and [3] send the first T cell activating signal. [4] In order for the T cell to become completely activated and start proliferating, a co-stimulation signal is required from the same APC that sent the first signal. This is provided by costimulatory molecules, e.g., CD80/86 (B7):CD28. If the activation of a T cell occurred via a TCR:CD8 to antigen:MHC I complex binding these cells differentiate into cytotoxic T cells (Tc) that mediate destruction of cells presenting foreign peptides on their surface. If the activation of a T cell occurred via a TCR:CD4 interaction with an antigen:MHC II complex, these cells differentiate into T helper cells (Th) that participate in the humoral immune response. T helper cells fall into 3 subgroups: Th1, Th2 and the recently discovered Th17 cells (35). Th1 cells are responsible for cell-mediated immunity by activating the microbicidal properties of macrophages, and inducing B cells to produce antibodies (e.g., IgG) for opsonization and removal of extracellular antigens. Th1 cells 29  produce IL-2, interferon-% (IFN-%) and tumor necrosis factor (TNF-!) and are predominantly responsible for acute organ rejection. Th2 cells are responsible for antibody-mediated immunity by stimulating the production of different isotypes of neutralizing antibodies (e.g., IgM). These cells secrete IL-4, IL-5, IL-6, IL-10 and IL-13 and are largely involved in chronic organ rejection. The Th2 pathway is less damaging than the Th1 pathway with respect to organ transplantation because of its less severe inflammatory and cytotoxic effects (36). Finally, Th17 cells are developmentally distinct from Th1 or Th2. These cells secrete IL-17 and have been implicated in autoimmune diseases, such as multiple sclerosis, where they play a key role in inflammation and tissue injury. Th17 cells also secrete IL-22 and serve antimicrobial functions on epithelial and mucosal barriers by stimulating these surfaces to produce microbicidal proteins. Most importantly, with regards to this thesis, the initiation of cell-mediated immunity requires multiple interactions between a T cell and an APC in order for T cell activation, expansion and effector functions to take place. Firstly, all APC – T cell interactions occur via surface molecules of complementary shape/geometry and this binding must be strengthened by multiple non-covalent forces. Secondly, the same cell that presented an antigen has to provide a co-stimulatory signal ensuring that there is no accidental T cell activation. In the absence of costimulation, T cells go into a state of unresponsiveness to the antigen – anergy; and become refractory to further activation even in the presence of all the required signals (37). Furthermore, weak stimulation may also result in apoptosis and, potentially, a clonal deletion of the responsive cells.  1.3.4  EFFECTS OF CELL–MEDIATED ALLORECOGNITION The most important role that cell-mediated immunity plays in initiating adverse effects in  response to blood transfusion, is the activation of B cells to produce antibodies for destruction of antigenic material. The Th cells activate B cells to produce an opsonizing antibody that initiates 30  complement and coagulation cascades, which result in cell lysis and tissue damage (described in 1.3.2). Cell-mediated immunity also includes the effects of NK cells and macrophages. Destruction of cells and tissues by NK cells is termed antibody-dependent cell-mediated cytotoxicity (ADCC). It is triggered by NK cells that recognize the Fc domains of antibodies bound to foreign antigens. NK cells release cytotoxic granules containing perforin and granzymes that cause the destruction of allogeneic cells (e.g., red blood cells and platelets) and local damage to the recipient’s own tissues. In addition to this, armed Th1 cells activate macrophages by producing interferon-% (IFN-%) and expressing CD40 surface molecules, which interact with CD40L on macrophages. A combination of these, results in macrophages secreting pro-inflammatory cytokines and producing oxygen radicals, antimicrobial peptides and proteases that result in the foreign cell destruction and the host’s tissue damage (38). In transplantation, cell-mediated immunity causes acute and chronic graft rejection. Rejection occurs via two pathways: direct and indirect (39-42). In the direct pathway, the recipient’s T cells recognize intact allogeneic donor MHC I/II molecules presented on the surface of donor APCs (e.g., dendritic cells). This occurs during the first several weeks after the transplantation and requires donor APCs entering the recipient’s lymphoid tissues to present donor MHC molecules to the recipient’s T cells. In the indirect pathway, the recipient’s APCs present the processed allogeneic peptides derived from donor’s alloantigens and MHC I/II proteins in the context of self-MHC complex. This requires the recipient’s lymphocytes traveling into the donor tissue, taking up, processing and presenting donor-derived alloantigens. Direct recognition is thought to be responsible for acute rejection, whereas indirect recognition is responsible for chronic rejection (43, 44). Ultimately, the graft becomes a target of cytotoxic T cells that recognize alloantigens in the context of MHC class I. Cytotoxic T cells release perforin and granzymes that lead to direct tissue damage or activate apoptotic pathways (45). Release of cytokines by T cells recruits macrophages that produce reactive nitrogen 31  species and oxygen intermediates that result in graft damage. Certain cytokines themselves could be damaging, e.g., TNF-!, resulting in atherosclerotic and fibrotic changes associated with graft failure. In addition, Th cells promote the production of antibodies against alloantigens, which result in tissue damage through complement fixation. The severity of the cell-mediated effects of allorecognition is exemplified by the difficulty of overcoming chronic graft rejection. The half-life of a cadaveric renal graft is only 8-9 years (25). The primary target of immune response is the endothelium (46). Therefore, heart and lung allografts are often characterized by coronary arteriosclerosis and bronchiolitis obliterans, respectively, which are responsible for the 50% of graft failure and the 25% mortality rate seen 5 years post transplantation.  32  1.4  CURRENT STRATEGIES FOR OVERCOMING  ALLORECOGNITION Allorecognition evolved as a very effective defense mechanism for detecting and eliminating foreign matter. It is essential to our survival and to the ability to fight infections. As such, allorecognition is an evolutionarily selected and conserved process, employing multiple complex mechanisms that are still being elucidated. Currently, it is impossible to inhibit the allorecognition of specific antigens without compromising the whole immune system of an individual. Therefore, in the clinical setting, the most effective strategy for avoiding allorecognition is tissue typing, donor and recipient cross-matching and, in some cases, leukoreduction. In many instances allorecognition is unavoidable, even with correct matching and leukodepletion, resulting in a cascade of immune reactions leading to inflammation, activation of the complement and the coagulation pathways, release of cytotoxic mediators and activation of immune cells. Once the early stages of allorecognition are initiated, it becomes extremely difficult to control or inhibit the later stages of this process. Therefore, when tissue cross-matching is not sufficient, the only option left is to reduce the damaging effects of allorecognition through the non-specific global suppression of the host’s immune system.  1.4.1  CLINICAL INTERVENTIONS  Before donor cells, tissues or organs are used for transplantation, tissue and blood types are determined. All blood units undergo ABO testing with anti-A and anti-B antibodies and Rh testing with anti-D. There are four major blood groups: A, B, O, AB and each one can be either Rh positive (contain D antigen) or Rh negative (no D antigen). For transplantation, in addition to ABO and Rh matching, tissue typing is performed. The HLA antigens of the donor and recipient are determined. There are 6 HLA antigens: HLA-A, -B, -C from MHC class I and HLA-DR, -  33  DP, -DQ from MHC class II. The donors and recipients are only typed for HLA-A, -B and –DR. These antigens are the most immunogenic (generate immune response) and have a strong impact on transplantation outcomes. For the recipients of renal allografts, the routine pre-transplant cross-match is performed to identify preformed antibodies in the recipient’s plasma against donor HLAs. In addition, a panel reactive antibody test (PRA) can be used for individuals having an increased risk of developing anti-HLA antibodies due to pregnancy, previous blood transfusion or organ transplantation. The screen is similar to the cross-match test, and utilizes a panel of cells, which express a wide variety of known HLA antigens (25). Unfortunately, clinical tissue typing is performed for only a few, most immunogenic antigens, regardless of the hundreds being expressed on any particular cell (23). This is further exacerbated by the limited “storage time” of biological tissues. For example, platelet units can be stored for only 5-7 days to avoid excessive bacterial growth and allowable cold ischemic times for organ transplants (time between procurement and transplantation) are less than 12 hours and optimally around 3-6 hours. As a result it is impossible to test for all known antigens and perform donor-recipient cross-matches without compromising the availability and quality of the blood products and organ transplants. Therefore, extensive testing is performed only for selected patients: the chronically transfused, the immunocompromised and individuals with a known history of alloreactivity. Following tissue typing and cross-matching, the second strategy often employed in the clinic for decreasing allorecognition is leukodepletion – removal of leukocytes. With respect to transfusion, currently, the most effective means of preventing alloimmunization is the leukoreduction by filtration or gamma irradiation of platelet and red blood cell components (performed only for immue-incompetent patients). All blood products distributed by Canadian Blood Services (CBS) are leukoreduced. Unfortunately, leukoreduction does not decrease antiHLA antibody formation by patients with previous pregnancies and transplantations. 34  Leukoreduction and irradiation do not affect the development of antibodies against platelet HLA, which could still cause platelet refractoriness. In addition, leukocyte-removing treatments will not attenuate the passive transfer of antibody present in plasma; and this may be important in transfusion-related acute lung injury (TRALI) (47). Recent studies also suggest that extreme leukodepletion of HLA class II-positive B cells results in enhanced alloimmunization. For example, platelet immunization by products lacking MHC II class cells could be due to the absence of potentially tolerogenic co-stimulatory molecule-deficient APCs that result in tolerance via clonal anergy (48). Similarly, with respect to organ transplantation, the depletion of T cells from bone marrow transplants in order to prevent graft versus host disease (GVHD) is associated with the loss of the potential benefits. T cell depletion results in increased graft failure, graft rejection, disease relapse and loss of the graft versus leukemia (GVL) effect (49, 50). In addition, the recipient suffers from an impaired immunity due to loss of T cells until the new cells can develop (51). This is especially of concern for adults, where insufficient immunity results in increased risk of opportunistic infections. Thus, clinical means of tissue typing, cross-matching and leukoreduction are not failproof strategies and still result in allorecognition and alloimmunization by the patient. In addition, the effects of leukodepletion, while beneficial in some cases, are also detrimental in others. Hence, leukoreduction is performed at the discretion of individual countries, centers and hospitals, without a uniform consensus. Because of these factors, several laboratories are now actively investigating the acceptable transfusion triggers in order to minimize the number of patients recommended for transfusion. This is especially of concern in adult, intensive care unit, trauma, and surgical patients, where allogeneic blood transfusions were shown to result in increased morbidity and mortality (52). While avoiding transfusion could be an option for some patients, for the critically ill and individuals with haemoglobinopathies, it is unavoidable. 35  Similarly, in the case of organ transplantation, there could be no other alternative than the replacement of a failed organ. In the instances where transfusion and organ transplantation are necessary and allorecognition is not preventable by clinical means, the focus is on the maintenance therapy involving anti-inflammatory and immunosuppressive drugs.  1.4.2  IMMUNOSUPPRESSIVE THERAPY Presently, the most effective means of minimizing the damage mediated by the  immunological response to transplanted tissues (once allorecognition has occurred) is the use of immunosuppressive therapy. Currently used immunosuppressive agents include pharmacological agents and immunosuppressive antibodies (anti-lymphocyte globulin). Pharmacological agents fall into 3 broad categories: 1) fungal and bacterial derivatives (e.g., cyclosporin A); 2) cytotoxic agents (e.g., azathioprine); and 3) anti-inflammatory drugs (e.g., corticosteroids). These pharmaceuticals inhibit various stages during the T cell activation, proliferation and differentiation pathways, Figure 1-12. Fungal and bacterial derivatives include some of the most potent of immunosuppressive drugs - cyclosporin A (CyA), tacrolimus (FK506) and sirolimus (rapamycin) (53, 54). CyA and tacrolimus block the transcription of interleukin-2 (IL-2) cytokine responsible for T cell activation and maintenance of the T cell activated state. Sirolimus acts downstream of CyA and tacrolimus and blocks T cell proliferation in response to IL-2 and other growth factors, Figure 112. The family of cytotoxic drugs consists of azathioprine (AZA), micophenolate mofetil (MMF) and cyclophosphamide (CPM). These agents damage DNA and RNA or inhibit their synthesis preventing transcription and cell proliferation. Corticosteroids are anti-inflammatory agents that include hydrocortisone, prednisone, prednisolone, and methylprednisolone. They act through the inhibition of histamine and bradykinin release, leukocyte adhesion to endothelium;  36  suppression of prostaglandin synthesis; decrease of capillary permeability, lymphocyte and macrophage traffic; and reduction of the numbers of eosinophils and neutrophils.  Figure 1-12: Immunosuppressive drugs act at various stages during T cell activation and proliferation. All pharmacological agents act downstream of the initial T cell and antigen presenting cell (APC) interactions and antigen:MHC complex recognition. Cyclosporin A (CyA) and tacrolimus inhibit early stages of T cell activation. Sirolimus, azathioprine (AZA) and micophenolate mofetil (MMF) inhibit later stages of T-cell activation and proliferation. Immunosuppressive antibodies, such as anti-IL-2 receptor (anti-CD25) antibody, inhibit T cell activation in response to activating cytokines and growth factors. Modified from Torpey 2002 (53). All immunosuppressive drugs suppress the immune system non-specifically and exert significant cellular toxicity. Consequent to this, immunosuppressed patients exhibit an increased incidence of infection and malignancy (53, 54). For example, CyA results in nephrotoxicity, which affects 15-40% of patients. CyA also affects native kidneys resulting in dialysis dependent renal failure. Tacrolimus causes nephrotoxicity, neurotoxicity and an increased incidence of diabetes mellitus. Cytotoxic agents result in bone marrow suppression, pancytopenia, thrombocytopenia with megaloblastic anemia. These agents are mutagenic and cause hair loss, nausea and vomiting. CPM leads to infertility and cardiac dysfunction in 15% of 37  patients. Finally, corticosteroids, are perhaps the most toxic agents causing considerable morbidity, poor wound healing, cardiovascular and bone disease, growth retardation of children, and psychiatric disorders. The long-term effects of immunosuppressive drugs are demonstrated by an increased risk of acquiring opportunistic infections: 12% of liver transplant patients die from infection (55). Cytomegalovirus (CMV) is the most common infection that affects 20-50% of transplant recipients (56). Incidence of certain cancers, e.g., skin cancer, is 100 times higher in immunosuppressed patients in comparison to the general population. Infection by oncogenic viruses is responsible for skin, uterine cervical and anal cancers. Cardiovascular mortality following transplantation is also increased by 10 times in comparison to the general population. In some renal transplant studies, cardiovascular disease is the leading cause of death (57). The group of immunosuppressive antibodies consists of polyclonal anti-lymphocyte antibodies (ATG), monoclonal anti-lymphocyte antibodies (OKT3) and anti-IL-2 receptor antibodies (Anti-CD25) (53). ATG is produced by immunizing animals (rabbits, goats and horses) with either human thymocytes or lymphocytes. They are reactive against many surface molecules on T and B cells and lead to opsonization and cell lysis by complement. These have been shown to not significantly improve long-term patient and graft survival and have severe side effects. Due to their animal source these antibodies can elicit sensitization in some patients causing neutropenia and thrombocytopenia. Due to being non-specific, ATG increases the risk of early CMV infection and exhibits a 2-3 fold increase of posttransplant lymphoproliferative disease (PTLD). In the clinic, preference is always given to pharmacological agents e.g., tacrolimus and sirolimus. OKT3 is a murine monoclonal antibody directed against CD3 on T cells. It is a nonspecific agent whose effectiveness is limited by sensitization and formation of neutralizing antibodies upon repeated treatment. The use of OKT3 declined sharply due to the increased 38  incidence of viral infections and lymphoproliferative disorders (58). Finally, Anti-CD25 antibodies that prevent T cell activation by binding to the IL-2 receptor, were found to have little effect on cytotoxic effector cells and hence, were found ineffective in treating acute rejections. Currently all immunosuppressive therapies (drugs and antibodies) suffer from severe toxic effects and long-term complications. This is exacerbated by the extensive and prolonged exposure of patients to these toxic agents. Following organ transplantation, the immediate risk is acute graft rejection, therefore patients are maintained on high doses of immunosuppressive treatments. With time, therapy is decreased, but never completely removed and continues indefinitely, often throughout the lifetime of a transplant recipient or a graft (53). In addition, pharmacological immunosuppression while effective for preventing acute rejection, has very little effect in chronic rejection, which accounts for 10-80% of renal graft loss (25). Unfortunately, at present virtually all of the efforts in the medical community are directed towards improving and refining already existing pharmacological agents. Therefore, new, effective, low toxicity methods of preventing allorecognition are desperately needed. The ideal goal of any approach to prevent allorecognition would be to achieve one or all of the following: 1) to inhibit allorecognition before it occurs or at the initial stages, prior to the multitude of biochemical reactions taking place and culminating in the activation of the immune system; and 2) to gain graft specific tolerance in overcoming graft rejection and avoiding non-specific immunosuppression. None of the currently used clinical or pharmacological methods achieve this.  39  1.5  IMMUNOCAMOUFLAGE OF ALLOGENEIC TISSUES In the recent years, there has been a dramatic increase in the use of biological products as  a means of therapy, e.g., recombinant proteins for enzyme therapy and allogeneic tissues for transplantation. Therefore, camouflage of allogeneic epitopes on proteins and tissues from recognition by the components (antibodies and cells) of the recipient’s immune system has been explored. The term immunocamouflage, coined by Dr. M. D. Scott, refers to the biophysical camouflage of cell surfaces from interaction with other cells, macromolecules and viruses. Immunocamouflage aims to inhibit allorecognition during the initial stages of cell-cell and cellmacromolecule interactions via the global camouflage of cell surface molecules. This approach should result in stealth (i.e., undetectable) allogeneic tissues, eliciting graft specific tolerance while maintaining a normal host’s immune system. Historically, immunocamouflage had been achieved by covalent attachment of polyethylene glycol (PEG) derived polymers, such as methoxypoly(ethylene glycol) (mPEG), to the surface of allogeneic or xenogeneic proteins or cells using a chemical reaction. This is termed PEGylation.  1.5.1  METHOXYPOLY(ETHYLENE GLYCOL) AND PEGYLATION The advent of immunocamouflage of biological molecules began with work on  recombinant and xenogeneic enzymes and proteins. Historically, the use of purified biological products for treatment of human protein and enzyme deficiencies was impeded by the activation of the immune system against antigenic determinants on these foreign molecules. This resulted in a rapid clearance of foreign substances through complement and antibody binding and via ingestion by phagocytes. In 1977 Abuchowski and his colleagues discovered that the covalent attachment of methoxypoly(ethylene glycol) polymer to bovine liver catalase and bovine serum albumin (BSA) increased the protein’s blood circulation half-life in mice by rendering the  40  modified proteins non-immunogenic, non-antigenic, and resistant to degradation by proteases (59, 60). In the years following this discovery, PEGylation of drugs, liposome drug carriers and oligonucleotides were explored. In all these systems, similar pharmacokinetic improvements were noted and resulted in a substantially increased efficacy of the PEGylated molecules in comparison to their native counterparts. PEG-adenosine deaminase was the first approved (1990) PEG-conjugate and has been utilized for over 20 years to treat patients with severe combined immunodeficiency disease (SCID). To date few, if any, adverse effects in these patients have been reported (61-63). Consequent to the success of early PEGylation products, PEGylation technology has been used extensively in drug formulations with many drugs being approved and several pending approval for use in humans (Table 1-6) (64, 64-68). Existing PEGylated therapeutics include PEG-bioactive proteins, PEG-drugs and PEG-oligonucleotides.  41  Table 1-6: PEG pharmaceuticals that are currently approved or undergoing clinical trials. Product  Name  PEG MW kDa  Protein MW kDa  Company  Status  Indication  PEG-adenosine deaminase FDA-1990 PEG-asparaginase FDA-1994  Pegademase Adagen®  5  40  Enzon Pharmaceuticals  Launched  SCID  Pegaspargase Oncaspar®  5  135  Enzon Pharmaceuticals  Launched  Acute lymphoblastic leukemia  PEG-IFN!2a  Pegasys®  40  19.2  Hoffmann-La Roche  Launched Phase I/II  PEG-IFN!2b FDA-2000  PEG-Intron  12  19  Shering-Plough  Launched Phase I/II  PEG-G-CSF  Neulasta®  20  19  Amgen  Launched  PEG-arginine deaminase  Hepacid®  20  46.4  Phoenix Pharmacologics Shering-Plough  Phase I Phase I/II  Hepatitis C Melanoma, multiple myeloma, renal cell carcinoma Hepatitis C Melanoma, multiple myeloma, renal cell carcinoma Neutropenia from cancer therapy Hepatocellular Melanoma  PEG-glutaminase + glutamine antimetabolite DON PEG-hGH  GlutaDON  N/D  64.8  Medical Enzymes AG  Phase I/II  Various cancers  Pegvisomant Somavert®  5  22.1  Pfizer  Launched  Acromegaly  PEG-anti-TNF Fab  CD870 CimiziaTM  20  50.0  Celltech Pharmaceuticals  Phase III  Crohn’s disease, rheumatoid arthritis  5  68.0  Sangart  Phase II  Blood transfusion substitute  59.7  NIMH, Bethesda, MD, USA Bio-Technology General Corporation  Phase I  Gaucher’s disease  Phase I/II  Gout, hyperuricemia  PEG-protein  PEG-haemoglobin PEGglucocerebrosidase  Lysodase®  PEG-uricase  Puricase®  20  33.4  Prothecan®  40  0.35  Enzon Pharmaceuticals  Phase II  Cancer  ND  0.85  Enzon Pharmaceuticals  Phase I  Cancer  PEG-drug PEG-camptothecin PEG-paclitaxel PEG-oligonucleotide PEG-anti-VEGF Pegaptanib 40 !1.0 Pfizer/Eyetech Launched Age-related macular aptamer Macugen® Parmaceuticals degeneration DON – 6-diazo-5-oxo-L-norleucine; G-CSF – granulocyte colony-stimulating factor; Hgh – human growth hormone; IF – interferon; ND – no data; PEG – polyethylene glycol; SCID – severe combined immunodeficiency disease; TNF – tumor necrosis factor; VEGF – vascular endothelial growth factor; MW – molecular weight. Modified from Parveen 2006 (69).  42  Due to its extensive use in the pharmaceutical industry, poly(ethylene glycol) is the most widely utilized synthetic polymer for conjugation of biological molecules. It is synthesized by polymerization (repetitive addition of the same chemical unit) of ethylene oxide (– CH2 – CH2 – O –) initiated by a nucleophilic attack of a hydroxide ion (OH-) on an epoxide ring (  )  (69, 70). The general chemical structure of PEG is OH – CH2 – CH2 – (O – CH2 – CH2)n – OH, where n refers to the number of ethylene oxide units in the polymer chain. PEG is a non-toxic, highly water soluble polymer of low immunogenicity and antigenicity. It is approved by the U.S. Food and Drug Administration (FDA) for parenteral, oral, and topical uses in humans. The derivative of PEG that is used most commonly for modification of biological molecules is methoxypoly(ethylene glycol) (mPEG). mPEG is synthesized analogously to PEG, except that instead of using a hydroxide ion for the epoxide ring opening, a methoxide ion (CH3O-) is utilized (70). The structure of mPEG is CH3 – O – (CH2 – CH2 – O)n – CH2 – CH2 – OH. The methoxyl group acts as a “cap” that renders one end of the PEG molecule unreactive. The terminal OH group of mPEG is usually functionalized with a reactive group or a linker, which enables covalent attachment of the polymer chain to the various substrates.  1.5.2  PHYSICAL PROPERTIES OF PEG AND PEG CONJUGATES PEG is a neutral, water soluble, hydrophilic polymer binding 3 water molecules per  ethylene oxide unit (60). In addition, PEG is highly flexible with free chain rotation every 4-5 ethoxy units. Due to its extensive water binding capacity and chain mobility, polymers possess a large exclusion volume. In fact, a PEG molecule is ~10 times larger in volume (3D) than a soluble protein of the same molecular weight (70). When terminally attached to a surface (synthetic or biological) the PEG polymer exerts entropic repulsion and volume exclusion effects (71, 72), essentially acting as a mobile, heavily  43  hydrated elastic sphere, Figure 1-13. Consequently, the polymer creates a physical barrier against the approach and binding of molecules to the grafted surface (72-74). This non-fouling nature of PEG surfaces was extensively studied for use with blood contacting materials (75-81). In addition, PEG polymer forms a viscous, neutral layer and results in changes of the electrical properties of surfaces. This phenomenon was applied to capillary electrophoresis for the reduction of electroosmotic flow (82, 83, 83-85). The unique physical properties of PEG have resulted in the enhanced pharmacokinetic and therapeutic properties of PEGylated drugs, drug carriers, and therapeutic enzymes, proteins and oligonucleotides. The following improvements to modified proteins and drugs have been reported (67, 69, 70, 86): •  Increased plasma half-life due to increased size, steric interference in receptor binding, decreased antigenicity and proteolysis  •  Reduced renal clearance due to increased molecular weight (MW)  •  Reduced cellular and immunoclearance due to steric inhibition of surface interactions  •  Reduced proteolysis due to steric inhibition of enzyme binding  •  Reduced immunogenicity and antigenicity due to surface camouflage and steric inhibition of surface interactions  •  Increased solubility due to PEG hydrophilicity and solubility  44  Figure 1-13: Exclusion zone of a covalently attached mPEG chain. Due to its extensive water binding capacity and chain flexibility, terminally attached mPEG has a large exclusion volume. Polymer chains move in 3D space, constantly folding and unfolding creating a non-easily penetrable zone (sphere) of surface camouflage. This behaviour lies at the base of the surface camouflaging effects of mPEG. The gradient depicts the probability of finding the polymer at a particular point of the 3D space occupied by the polymer chain. Covalently bound mPEG can assume two main configurations: coiled mushroom or extended brush (87, 88), Figure 1-14. The preference for either configuration is dependent upon the density of polymer chains on the surface. At low chain densities, the polymer assumes a mushroom shape, Figure 1-14A. The radius of the sphere that defines the limits of polymermediated surface camouflage is defined as a Flory radius (RF) or radius of gyration (Rg). RF is calculated using RF=aN0.6, where a is the extended length of one ethylene oxide unit (3.5 Å) and N is the number of ethylene oxide units in a polymer chain (87). Based on this formula, 2, 5 and 20 kDa mPEG chains possess the following RF: 3.4 nm, 6.0 nm and 13.7 nm, respectively (Figure 1-15). If mPEG is grafted at higher densities so that polymer chains are constrained against one another, the polymer is forced to extend from the surface and the brush regime is achieved with RF=aN(a/D)2/3, where D is the separation between the chains, Figure 1-14B. The RF calculations are only theoretical treatments of highly complex polymer-surface interfaces and variations of experimentally measured values are reported. For example, for mushroom regimes Jeppesen (89) found that the extension length of the polymer is better reflected by RF=aN0.64, whereas Kenworthy (90) reported RF=6.5 and 11.5 nm for the 2 and 5 kDa mPEG, respectively. Importantly, on a macromolecular scale, polymers are very dynamic, 45  constantly folding and unfolding, thus potentially assuming fully extended conformations, e.g., ~160 nm for a 20 kDa polymer (73, 91).  Figure 1-14: Conformations of surface bound mPEG chains. A) When chains are grafted at low densities, with a large separation between chains (D), polymers conform to mushroom regime. Chains in this configuration create zones of immunoprotection with the radius defined as the Flory radius (RF). There exists a gradient of surface camouflage (shown as a color gradient) that is the highest in the middle and decreases at a distance very close to the surface and far from the surface. B) When chains are grafted at high density, with a small separation between the chains (D), polymers extend from the surface and form polymer brushes. The surface camouflage is constant and decreases at a distance far away from the surface where the chains are more mobile. The gradients in (A) and (B) depict the probability of finding the polymer at a particular point of the 3D space occupied by the polymer chain. When mPEG is grafted to biological surfaces it is unlikely that the brush border would be formed. Cell surface bound polymers are likely to assume a mushroom configuration and the location of the grafting site in relation to the overall surface topography would influence the area being protected by the polymer, Figure 1-15. Usually, direct binding results in the efficient camouflage of a surface molecule. The efficacy of indirect binding depends on the height of the moiety, to which the polymer is attached, the distance between the two molecules, and the RF of the polymer chain. Importantly, PEG not only camouflages the molecule that it is directly attached to, but also the surface features within the immunoprotection zone of the polymer, Figure 1-15.  46  Figure 1-15: Immunoprotection zone of grafted mPEG is influenced by surface topography. Grafted mPEG exerts its immunoprotective effects depending on the size of the chains (RF values on the right hand side), the point of attachment (K denotes lysine residues on surface proteins) and the topography of the surface (cell surface proteins). The grafting of mPEG occurs randomly. If it is bound not directly to a surface protein targeted for camouflage (dark colored proteins), but to a hypothetical nearby protein X, then the extent of camouflage would depend on the height of X in comparison to the targeted protein and the distance between the two proteins. Diagram modified from Le and Scott, manuscript in press.  1.5.3  CHEMISTRY OF PEGYLATION The grafting properties of mPEG are governed by the reactive end group (i.e., linker  group) bound to the polymer molecule at the -OH site. There are several commercially available types of functionalized mPEGs (92). Most linkers target nucleophilic residues of protein amino acids, although some oligosaccharide selective chemistries also exist. The order of reactivity for commonly targeted nucleophiles is S- > !-NH2 > (-NH2 > COO- " O-. However, the thiol group is rarely present in proteins and is mostly involved in the active sites of enzymes. The carboxyl group is difficult to activate and couple without crosslinking it with nearby amine residues, 47  therefore this linker chemistry is rarely utilized. On the other hand, the primary amine groups, such as the !-NH2 group of amino acids and the (-NH2 group of lysines (K), are very abundant and rarely present in the active sites. In addition, lysine is one of the most abundant amino acids in proteins constituting approximately 6% of all amino acids (93). Therefore, the residues most often targeted for PEGylation are the terminal !-NH2 and (-NH2 of lysines (69, 94). Because a given polypeptide chain contains only one terminal !-NH2 group, polymer attachment (grafting) mostly occurs at the (-NH2 sites of the protein. The chemical reactivity of PEG is dependent on the structure of the protein and the pKa of the amino acid residue targeted for modification. Protein folding is critical in PEGylation as some potentially reactive residues are masked and buried inside the protein. Environmental factors, such as the presence of charged or hydrophobic entities can change the propensity of the residue to act as a nucleophile (66). The pKa of the amino acid determines the charge state of the residue targeted by activated mPEG. The rate of the reaction is optimal when amine groups of lysine are in their deprotonated neutral form - NH2. All ionization states of lysine exist in equilibrium with each other with a particular species dominating depending on the pH of the environment, Figure 1-16A. The pKa values for ionizable groups of lysine are as follows: pKa1 (!-COOH)=2.2, pKa2 (!-NH2)=9.0, pKaR ((-NH2)=10.5 (95). Therefore, the optimal pH for the PEGylation reaction would be at or above pH 9.0. All PEGylation reactions performed in this thesis utilized pH 8.0 to maintain the environment more physiologically relevant. At pH 8.0 the relative proportion of (-NH3+:(-NH2 ionized forms is approximately 99:1, Figure 1-16B. The choice of the linker chemistry of the functionalized mPEG molecule is also important as it determines the reactivity of the activated mPEG polymer towards primary amine groups vs. water molecules in aqueous solution. The reaction with water results in linker hydrolysis rendering mPEG polymer chain incapable of covalent binding to the substrate. The  48  chemical properties of linkers used in this thesis will be discussed more fully in the Materials and Methods chapter.  Figure 1-16: Lysine ionization states. A) Lysine possesses 3 ionizable groups: !-COOH (pK1), !-NH3+ (pK2), (-NH3+ (pKR) that become deprotonated at their corresponding pH values. In aqueous solution various ionization states are in constant equilibrium and one species predominates depending on the pH of the environment. The pI (isoelectric point) of lysine is 9.7. B) PEGylation reaction is performed at pH 8.0 where most primary amine groups are protonated. Consumption of deprotonated NH2 groups due to mPEG grafting does not deplete these species as Le Chatelier’s principle maintains their constant concentration at a given pH value. 49  1.5.4  PEG METABOLISM, TOXICITY AND IMMUNOGENICITY The major excretion route for PEG is through urine. The threshold for glomerular  filtration for PEG is approximately 40 kDa, therefore polymers of less than 40 kDa are almost exclusively excreted in urine, while higher molecular weight chains (>40 kDa) are more slowly cleared via both the urine and the feces (86, 96). However, hepatobiliary clearance of PEG in humans is expected to constitute a minor route as studies in mice indicated that PEG clearance through the liver is significantly lower ("100 times) than through kidneys for all molecular weight polymers (96). Short PEG chains ("1 kDa) can become metabolized to carboxylic acid in vivo by cytochrome P450 (97), alcohol dehydrogenase (98) and aldehyde dehydrogenase (99). Metabolism of soluble PEG in humans decreases with increased molecular weight and for the 6 kDa PEG constitutes only 4% (97). Toxicological studies show very safe profiles for PEG, Table 1-7. In fact, PEG is widely used as an excipient in medicines and is routinely administered to humans via intravenous, oral, rectal and topical routes. It is also found in formulations of toothpaste, shampoo, lotions, colorants, foods, drinks and deodorants (100). Absorption of PEG orally and from intestine decreases with molecular weight; for example, oral absorption of 1 kDa PEG is only 10% and for 6 kDa PEG it is barely detectable. Small PEG chains of less than 400 Da are excluded from use in humans due to their potential of being oxidized in vivo to toxic metabolites (101). High molecular weight PEGs (>1 kDa) are considered to be safe (102, 103). PEG was shown to have no adverse reproductive and teratogenic effects and it is not mutagenic or carcinogenic (104). In 1980, the World Health Organization set the acceptable oral daily intake of PEG for humans to 10 mg/kg human/day. Thus, an adult person weighing 80 kg can consume 800 mg of PEG per day. This is far below the LD50 values reported for oral acute toxicity in animal studies, Table 17.  50  Table 1-7: Toxicity of PEG (LD50) by molecular weight and animal species. Parameter measured Oral acute toxicity  Dermal acute toxicity Intravenous acute toxicity  Compound-Da PEG-300 PEG-300 PEG-400 PEG-400 PEG-1600 PEG-4000 PEG-4000 PEG-8000 PEG-300 PEG-4000 PEG-8000  Species Rat Rabbit Rat Dog Rat Rat Rabbit Rat Rabbit Rabbit Rabbit  LD50 (per/kg body weight) 31.7 g 17.3 g 32.8 g 1.0 g 51.2 g, 50% aq. solution 59.0 g, 50% aq. solution 76.0 g, 50% aq. solution > 50.0 g > 20 ml > 20 ml, 40% >10.0 g, as 5% solution  LD50 – median lethal dose, the amount of a tested substance that results in a death of half of the tested subjects. Modified from Fruijtier-Polloth 2005 (104).  There are few, if any, reports in which antibodies to PEG have been shown to cause adverse immunological consequences (e.g., anaphylaxis) regardless of the extensive clinical use (>20 years) of PEG-drugs and routine human exposure to non-pharmacological products containing PEG. PEG was shown to be only weakly immunogenic. However, antibodies can be generated in rabbits and mice when PEG is attached to highly immunogenic proteins and introduced in combination with Freund’s complete adjuvant (FCA) (105, 106). Richter et al., have shown that 10 and 100 kDa PEG in FCA was non-immunogenic in rabbits and weakly immunogenic in mice. In the absence of FCA PEG-ovalbumin did not elicit any anti-PEG response in rabbits and only a weak response in mice (105). Furthermore, a clinical human study (106) indicated that anti-PEG antibodies were seen in only 0.2% of healthy individuals. Following the first hyposensitization of allergic patients with PEG-ragweed extract and PEGhoney bee venom, 50% of people showed an anti-PEG response that caused no adverse effects even with repeated PEG administration. Moreover, after 2 years, the titers declined to 28.5% and were deemed to be of no clinical significance. Anti-PEG antibodies were predominantly of the IgM isotype.  51  Although some recent studies have indicated the presence of anti-PEG antibodies in the context of PEG-liposome drug delivery systems (107) and PEG-asparaginase therapy (108), the 30-years of pre-clinical and clinical history of the safe use of PEGylated pharmaceuticals and non-pharmacological PEG-containing products is suggestive of PEG’s insignificant immunogenicity. In fact, the use of PEG for modification of highly antigenic recombinant proteins has resulted in decreased antigenicity, immunogenicity and clearance of these PEGconjugates (59, 60, 109-111). Indeed, the morbidity and mortality associated with enzymopathies would be greatly elevated if PEGylated enzymes did not exist. Therefore, it remains to be established whether the occurrence of PEG antibodies is protein, drug and formulation specific and whether any real or theoretical risk is outweighed by the beneficial consequences imparted by PEGylation.  1.5.5  PEGYLATED BLOOD CELLS AND TISSUES The initial interest in PEGylation of cells arose due to the high rate of alloimmunization  of chronically transfused patients with sickle cell anemia and thalassemia. The pioneering work on cellular PEGylation began in late 1990’s with the covalent modification of red blood cells in an attempt to camouflage red blood cell antigens from recognition by the recipient’s immune system (112-115). Scott et al., have shown that PEGylation of RBCs resulted in a dosedependent decrease of RBC agglutination with anti-ABO antisera and inhibition of phagocytosis of PEGylated sheep RBCs by human peripheral blood mononuclear cells (PBMCs). These experiments indicated decreased recognition of antigenic sites on RBCs. In support of these observations, murine studies demonstrated that PEGylated sheep RBCs had >5 times increased survival relative to the control non-PEGylated xenogeneic erythrocytes. Moreover, the mPEGmediated surface charge camouflage was demonstrated as a decreased electrophoretic mobility of PEGylated RBCs in particle electrophoresis analysis. Thus, the initial PEGylation studies on 52  erythrocytes have shown that covalently bound mPEG resulted in the camouflage of cell surface antigens and inherent RBC charge, thus protecting modified cells from recognition and destruction by the host’s immune system. In the subsequent experiments, the effects of membrane PEGylation on RBC function and morphology were evaluated. These studies demonstrated that modified RBCs maintained normal osmotic fragility and oxygen affinity with mPEG grafting concentrations up to 5 mM. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of RBC membrane proteins revealed that the mPEG was covalently bound to the surface proteins as evidenced by the molecular weight shifts in the area of Bands 3 and 4.1 proteins. Covalent modification of membrane proteins did not elicit any changes in Na+, K+, Cl- transport and SO4- influx at grafting concentrations of <5 mM. Normal functional features of PEGylated erythrocytes were also supported by the observation that repeated transfusions (e.g., tertiary) of PEGylated RBCs in mouse over a 50-day period resulted in normal mouse mPEG-RBC in vivo survival, Figure 1-17. Finally, PEGylated RBCs exhibited safe profiles in hypertransfused mice (>80% of RBC consisting of PEG-RBC) resulting in no toxicity and normal RBC survival kinetics, Figure 1-17 Insert. Gross examination of the mouse organs indicated no differences in liver, kidney, heart or brain in comparison to control animals. In addition, the increased size of spleen seen in hypertransfused mice with control RBCs was not seen in mice receiving mPEG-RBC.  53  Figure 1-17: PEGylated RBC in vivo mouse survival studies. The primary and tertiary transfusions of control (dashed lines) and 2 mM, 20 kDa benzotriazol carbonate (BTC)-mPEG modified mouse RBCs (solid lines) demonstrated normal in vivo mouse survival. Insert: Hypertransfusion studies, where 80% of mouse RBC mass consisted of mPEG derivatized cells, indicated no adverse effects on viability of mice. In addition, PEGylated RBCs demonstrated the same kinetic profiles as unmodified control cells. RBCs were labeled with PKH-26 fluorescent fatty acid marker and injected i.p. into BALB/C mice. Each transfusion was commenced following a complete clearance of the previously infused cells. Blood samples were obtained by tail-cuts and analyzed using flow cytometry. Modified from Scott 1998 (113). As a result of the successful initial work with PEGylated RBC that demonstrated decreased immunogenicity of allogeneic erythrocytes in vitro and in vivo animal models, PEGylation of white blood cells was studied as a means of preventing transfusion associated graft versus host disease (TA-GVHD) (116-120). These studies utilized a 1-way mixed lymphocyte reaction (MLR), in which two disparate populations of human peripheral blood mononuclear cells (PBMCs) were mixed together. The allorecognition of foreign HLA molecules resulted in a vast T cell activation and proliferation that was measured as an increase of 3H-Thymidine incorporation into the DNA of dividing cells. Scott et al., have demonstrated 54  that PEGylation of either population of human PBMCs resulted in inhibition of allorecognition as shown by decreased T cell proliferation and normal lymphocyte morphology, Figure 1-18. Furthermore, inhibition of lymphocyte proliferation was not overcome by addition of exogenous IL-2, which suggested that the initial cell-cell interactions and antigen recognition were inhibited. The camouflage of antigenic molecules was further investigated by looking at the effects of PEGylation on the camouflage of T cell surface antigens. These studies demonstrated efficient camouflage of CD markers, such as CD3, CD4, CD28 and CD80 as indicated by the decreased antibody binding to these surface molecules, Figure 1-18.  Figure 1-18: Immunocamouflage of human lymphocytes. PEGylation of human peripheral blood mononuclear cells resulted in inhibition of T cell activation and proliferation in a 1-way mixed lymphocyte reaction. PEGylation of responder cells led to the dose-dependent decrease of 3H-thymidine incorporation into the DNA of dividing cells as measured by scintillation counting. Decreased cell proliferation was also evident from lymphocyte photomicrographs showing normal cell morphology. Attenuation of T cell proliferation was, most likely, the effect of the camouflage of surface antigens important in T cell-antigen presenting cell interactions, e.g., co-stimulation pathway via CD28-CD80 binding. In fact, flow cytometric studies demonstrated decreased fluorescent antibody binding to CD28 indicating immunocamouflage of the T cell surface. Modified from Murad 1999 (119).  55  Consequent to the demonstrated mPEG-mediated prevention of cell-cell and cellantibody interactions, application of the PEGylation technology to transplantation medicine was explored (117, 120). The efficacy of control and modified pancreatic islets was compared for the treatment of type I diabetes using a diabetic rat model. These studies demonstrated that the transplant engraftment and function were unaffected by PEGylation as evidenced by achievement and maintenance of normoglycemia. Importantly, these studies also indicated that PEGylation of islets did not interfere with the complex biological signaling events necessary for sensing and maintaining glucose levels. Although one of the concerns was that the initial immunocamouflage could become lost with time due to normal homeostatic membrane regeneration and remodeling, recent studies in rats demonstrated prolonged transplant survival. PEGylated islet transplants in combination with low doses of cyclosporin A exhibited increased survival, normal function and no evidence of immune cell infiltration for at least 1 year in comparison to 2 weeks for non-modified cells (121, 122). Thus, PEGylation studies on blood cells and pancreatic islets indicated that in addition to sufficiently camouflaging the cell surface charge and the antigenic proteins, mPEG modification resulted in the inhibition of complex cellmacromolecule and cell-cell interactions.  1.5.6  PEGYLATED VIRUSES AND HOST CELLS More recently, the research of Scott et al., has examined the use of PEGylation in a  broad-spectrum antiviral prophylaxis. This approach was initially suggested by the effect of grafted mPEG on receptor-ligand interactions of lymphocytes, which is analogous to the receptor-virus binding. Thus, the effects of PEGylation on inhibition of cell-virus interactions were investigated (123-129), Figure 1-19. The initial studies by Mizouni et al., looked at the effects of direct PEGylation of the SV40 virus. These experiments have shown that the VP1 capsid proteins on the virus particle were directly modified by the polymer as indicated by 56  increased molecular weights of the protein bands visualized in SDS-PAGE and Western Blot analyses. Consequent to derivatization, the mPEG-SV40 virus demonstrated decreased propensity to infect CV-1 cells in a dose-dependent manner; for example at 2 mM grafting concentration a 95% decrease in the number of T antigen expressing cells was observed, 24 hours post infection. Later studies evaluated the efficacy of CV-1 host cell PEGylation and demonstrated that SV40 infection was successfully attenuated in a dose-dependent manner without any adverse effects on CV-1 cell viability. By 72 hours 93% of control cells became infected with the virus in comparison to 27% of 0.6 mM, 5 kDa cyanuric chloride mPEG (CmPEG) modified cells. A similar effective prophylaxis was also achieved for a rat coronavirus. PEGylation of L2 rat lung epithelial cells at low 0.2 mM grafting concentrations of benzotriazol carbonate (BTC)-mPEG resulted in an almost complete abrogation of the infection. Further experimentation with a wide variety of viruses demonstrated that PEGylation of host cells resulted in protection against adeno, picorna, papova and corona family viruses, Table 1-8. This also indicated that both receptor-mediated endocytosis and membrane fusion entries employed by different classes of viruses were successfully attenuated. Moreover, PEGylation was shown to inhibit the propagation of the infection either as a consequence of the presence of PEGylated host proteins on the viral particle during virus shedding or a protective PEG layer on the surrounding cells. Specifically, consequent to viral propagation, the infection of >90% of unmodified neighboring cells was observed in response to cytomegalovirus infection. This rate of infectivity was decreased to 6% when the host cells were PEGylated with 10 mM BTC-mPEG. Thus, these studies not only demonstrated the efficacy of virus particle and host cell PEGylation in attenuating viral infection but also suggested that this technology could be successfully utilized as a means of broad-spectrum antiviral prophylaxis, Table 1-8.  57  Figure 1-19: mPEG-mediated viral prophylaxis. A) Viral pathogenesis involves virus-receptor interactions (1), which mediate the entry of the virus into a cell (2). There, the virus can undergo multiple rounds of replication (3) followed by new viral particle “packaging” (4) and shedding of progeny virus into the extracellular space (5). Newly synthesized and released virus can undergo the same steps of pathogenesis inside a neighboring cell. The surface grafting of mPEG to either a virus particle directly (B) or to its host cells (C) can interfere with the initial virus-receptor interactions via immunocamouflage of virus surface proteins or their complementary binding receptors on the host cells. Modified from McCoy 2005 (123).  Table 1-8: Classes of experimental viruses used for PEGylation studies. Virus Picorna; TMEV Adenovirus; MAV  Coronavirus; RCV Herpesviridae; CMV  Target cell type Hamster kidney fibroblasts Mouse embryonic fibroblasts Rat lung epithelial cells Human lung fibroblasts  Paramixoviridae; Epithelial cells RSV  Receptor Unknown Murine homologue of Coxsackie and Adenovirus receptor (mCAR) Unknown Epidermal growth factor receptor (EGFR) Unknown  Mode of entry Receptor mediated endocytosis Receptor mediated endocytosis  Particle size 20 nm  Genome type ssRNA  70-90 nm  Linear dsDNA  Fusion  80-160 nm  ssRNA  Fusion  200 nm  Linear dsDNA  Receptor mediated endocytosis  120-300 nm  Circular dsDNA  TMEV – Theiler’s murine encephalomyelitis virus, MAV – mouse adenovirus, RCV – rat coronavirus, CMV – cytomegalovirus, RSV – respiratory syncytial virus, ss – single stranded, ds – double stranded. Table courtesy of M. D. Scott.  58  1.5.7  SUMMARY OF PEGYLATION TECHNOLOGY In summary, soluble PEG is an FDA approved compound for use in humans and has been  extensively utilized in the food, cosmetic and pharmaceutical industry. Moreover, the use of covalently grafted mPEG has been extended to medical applications. PEGylation of proteins, oligonucleotides, drugs and drug carriers was shown to result in a dramatically increased efficacy of these pharmacological agents. Due to mPEG’s favorable physical properties such as large exclusion volume, hydrophilicity and extensive intrachain flexibility, PEGylation resulted in enhanced pharmacokinetic and “stealth” properties of the modified substrates. Currently, several PEGylated pharmacological agents are used in the clinic to treat patients with enzymopathies, such as severe combined immunodeficiency syndrome (SCID, PEG-adenosine deaminase); various cancers, for example, acute lymphoblastic leukemia (PEG-L-asparaginase) and viral infections, e.g., hepatitis C (PEG-IFN!2a/b). Thus, more than 30 years of pre-clinical and clinical history of safe use of PEGylated agents suggests a significant therapeutic potential of PEGylated cellular products. The research in our laboratory has shown that PEGylated blood cells exhibited decreased antigenicity due to the immunocamouflage of antigenic cell surface proteins and the inhibition of cell-macromolecule and cell-cell interactions. In addition, grafting of mPEG to either a virus or a host cell resulted in a broad-spectrum antiviral prophylaxis via attenuation of viral infection. Thus, PEGylation of cells and tissues could have a significant potential in transplantation medicine by inhibiting the antibody and cell-mediated pathways of allorecognition through either camouflage of surface antigens or prevention of T cell-antigen presenting cell interactions.  59  1.6  HYPOTHESIS AND SPECIFIC AIMS  1.6.1  EXPERIMENTAL HYPOTHESIS  Based on previous studies with blood cells and viruses (112-114, 116-120, 123, 130-133), it became evident that PEGylation of cell surface proteins imparted significant immunocamouflage which resulted in decreased surface-macromolecule interactions (e.g., cell-antibody and cellvirus) and changes in the electrical properties of surfaces (e.g., decreased apparent surface charge of RBCs). In addition, PEGylation did not result in adverse morphological or functional alterations of the cells. While these proof-of-concept experiments have indicated the potential utility of PEGylation for overcoming allorecognition in tissue transplantation, the in-depth understanding of the biophysical mechanisms that govern mPEG-mediated immunocamouflage of intact cells is lacking. This is in part due to the highly labile nature of biological cells, which limits extensive biophysical experimentation. Therefore, in this thesis the biophysical mechanisms of immunocamouflage were investigated using a robust aliphatic amine polystyrene latex particle (biophysical model). Biophysical findings were correlated to the biological consequences of PEGylation by assessing immunocamouflage of erythrocytes and peripheral blood mononuclear cells (biological model). It was my hypothesis that the biophysical basis of immunocamouflage imparted by surface PEGylation is governed by physical (polymer length and architecture) and chemical (linker reactivity and grafting concentration) properties of the polymer as well as the size of the target. Consequent to these factors, surface charge and surface-macromolecule interactions would be affected.  1.6.2  SPECIFIC AIMS The goal of the study was two-fold (Figure 1-20): 1) Biophysical latex model – to study  the biophysical mechanisms of immunocamouflage imparted by surface PEGylation and to 60  evaluate the effects of the polymer and the target properties on the efficacy of immunocamouflage; and 2) Biological RBC/PBMC model – to apply the knowledge gained from the biophysical latex model for the efficient immunocamouflage of RBCs and PBMCs and establish a correlation between the biophysical mechanisms and the biological consequences of PEGylation. To achieve these goals, the following specific aims were defined: A)  Biophysical aliphatic amine polystyrene latex model. Efficacy of immunocamouflage evaluated with respect to: i) surface charge camouflage; and ii) prevention of surface-macromolecule interactions 1)  To determine whether mPEG covalently binds to the surface  2)  To determine whether mPEG linker chemistry has an effect on the efficacy of immunocamouflage  3)  To determine whether mPEG polymer molecular weight and polymer grafting concentration have an effect on the efficacy of immunocamouflage  4)  To determine whether polymer architecture has an effect on the efficacy of immunocamouflage  5) B)  To determine whether target size has an effect on the efficacy of immunocamouflage Biological red blood cell and peripheral blood mononuclear cell model. Efficacy of immunocamouflage evaluated with respect to: i) camouflage of surface charge; ii) camouflage of surface CD markers; and iii) prevention of allorecognition in a 2-way mixed lymphocyte reaction  1)  To determine whether mPEG polymer molecular weight and polymer grafting concentration have an effect on the efficacy of immunocamouflage  2)  To determine whether PEGylation adversely affects the viability of PBMCs  3)  To correlate the biophysical mechanisms with the biological consequences of PEGylation 61  Figure 1-20: The flowchart of the project. The biophysical latex model was used to study the biophysical mechanisms of immunocamouflage imparted by PEGylation. The hypothesis was tested by investigating the effects of polymer physical and chemical properties as well as the target size on the efficacy of surface charge camouflage and prevention of surface-macromolecule interactions. These biophysical effects were correlated with the biological consequences of PEGylation with respect to the cell viability, charge and surface marker camouflage and prevention of allorecognition in the mixed lymphocyte reaction.  1.6.3  SIGNIFICANCE OF CELLULAR PEGYLATION As discussed earlier in this chapter, allorecognition of foreign antigens is a major  impediment to successful transfusion and transplantation and results in alloimmunization, transfusion associated complications and acute and chronic graft rejection. The shortage of donors and currently utilized serological typing technology do not allow for a more thorough cell and tissue typing and cross-matching. This is further complicated by the number of existing alloantigens and the inherent genetic polymorphisms in a human population. Unfortunately, with respect to transfusion medicine, blood typing and leukoreduction are the only clinically used means of decreasing the severity of immune complications. This is of special concern for chronically transfused patients with hemoglobinopathies who are being routinely exposed to unmatched (HLA and non-ABO/RhD blood group antigens) blood products, 62  and for individuals requiring multiple transfusions, as in the case of cancer therapy, organ transplantation, trauma and surgery. In transplantation medicine, in addition to problems with organ availability and tissue cross-matching, the situation is further exacerbated by the use of highly toxic immunosuppressive drugs. This therapy, while alleviating immediate consequences (e.g., acute graft rejection) of tissue and organ transplantation, does not protect from chronic complications and results in a considerable degree of morbidity and mortality due to toxicity, increased incidences of infection, cancer and systemic complications. Importantly, all patients receiving allogeneic organ transplantations are severely immunocompromised and, depending on the organ transplanted, may remain on immunosuppressive regimens for life. It is my belief that cellular PEGylation could offer an improved means of decreasing immune complications associated with tissue transplantation by attenuating or inhibiting the allorecognition process itself. As described earlier in this chapter, allorecognition is highly dependent on the cell surface interactions with either soluble factors, such as antibodies, or cellular components, such as T and B cells. Moreover, cell surface interactions are maintained by a multitude of weak non-covalent forces and require a close approach and precise “fitting” of the two interacting molecules, e.g., a ligand and a receptor. None of the currently used clinical approaches target these initial interactions. For example, in the case of pharmacological drugs, inhibition of T cell activation and proliferation occurs either downstream of the initial T cell:APC interactions or targets intracellular pathways of DNA, RNA and protein syntheses that are common to many cell types in the human body. In contrast to this, donor tissue PEGylation may result in the efficient global camouflage of cell surface proteins and surface charge, thus interfering with the initial cell-antibody or cellcell binding events responsible for cell activation, proliferation and differentiation. This approach could also offer superior efficacy in preventing allorecognition as it will only interfere with the surface binding events on transplanted cells and would have no effect on other tissues or 63  intracellular biochemical pathways. Thus, cellular PEGylation will not exert any systemic cytotoxic effects and could, potentially, achieve graft-specific tolerance.  64  II.  MATERIALS AND METHODS  2.0  COMMON METHODOLOGY: BIOPHYSICAL LATEX MODEL  AND BIOLOGICAL CELL MODEL 2.0.1  STATISTICAL ANALYSIS All results were expressed as a mean ± standard error mean (SEM). A minimum of 3  replicates were performed for all studies, though in some cases (e.g., microscopy) representative samples were presented. Statistical analyses were done using SPSS v. 16.0 statistical software (Statistical Products and Services Solutions, SPSS Inc., Chicago, IL, USA). For comparison of 3 or more mean values a one-way analysis of variance (ANOVA) was performed followed by a Tukey post-hoc test for pair-wise comparisons of means. For comparison of 2 mean values, an independent variable T-test was used. Significance was determined by a two-tailed pvalue<0.05.  2.0.2  FLOW CYTOMETRY The flow cytometer (BD FACSCalibur, BD Biosciences, San Jose, CA, USA) was  calibrated against standardized samples of fluorescent beads provided by manufacturer (BD FACSCalibur, BD Biosciences, San Jose, CA, USA). Data acquisition and analysis were performed using CellQuest software (BD Biosciences, San Jose, CA, USA). In total, 20,000 events were collected per latex sample. For the lymphocyte CD camouflage analysis and peripheral blood mononuclear cell (PBMC) viability studies, 10,000 events were collected per sample. At least 50,000 events were collected for samples from the mixed lymphocyte reaction (MLR) experiments.  65  2.1  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX  MODEL 2.1.1  ALIPHATIC AMINE POLYSTYRENE LATEX In order to circumvent the use of the highly labile cells for the in-depth biophysical  studies, aliphatic amine polystyrene latex particles were utilized. The core of these beads was a hydrophobic polystyrene polymer and was negatively charged due to the presence of a small number of unused carboxyl groups (COO-) during the polymerization process, Figure 2-1. The surface of the latex particles was functionalized with primary amine (NH3+) groups covalently attached to the latex core via a six-carbon aliphatic arm. Not only did the primary amine groups allow for the covalent modification of the latex surface with an activated mPEG molecule, but they also mimicked the primary amine groups found on cell surface proteins. Thus, the latex bead model was considered as an artificial cell model. Latex particles were purchased in a water suspension (Molecular Probes, Invitrogen, Carlsbad, CA, USA) and stored at 4ºC to minimize bacterial growth. Two sizes of particles, 1.2 #m and 8.0 #m in diameter, were used in these studies in order to evaluate the effect of target size on the efficacy of immunocamouflage and to correlate the biophysical findings with the results of the biological model. For example, immunocamouflage of 8.0 #m beads was compared to that achieved on RBCs (8.0 #m) and WBCs (10 #m).  66  Figure 2-1: Aliphatic amine polystyrene latex particle is the biophysical artificial cell model. The core of the particle is slightly negatively charged and made up of hydrophobic polystyrene. The surface of the latex is functionalized with aliphatic primary amine groups. The number of positively charged amine groups is in large excess to the number of negatively charged carboxyl groups in the latex core. The particle possesses an overall positive charge. There are 3.5*107 and 7.6*109 amine groups per 1.2 and 8.0 #m particle, respectively.  2.1.2  POLYSTYRENE LATEX DERIVATIZATION The mPEG linker chemistries used in this thesis were succinimidyl carbonate (SC),  succinimidyl valerate (SVA) and succinimidyl propionate (SPA). All linkers reacted with the primary amines in an SN1 substitution reaction with the nucleophilic nitrogen atom of the amine group attacking an electrophilic carbonyl group of the succinimide ester, Figure 2-2. The chemical reaction culminated in the formation of a covalent bond and the release of the free linker group – N-hydroxy succinimide (NHS). SC-mPEG resulted in the formation of a urethane or a carbamate bond, whereas SVA- and SPA-mPEG formed an amide bond. All of these linkages are known to be very stable under physiological conditions (pH-7.4) and resistant to the enzyme breakdown (134).  67  Figure 2-2: Chemical reaction of SC-, SVA- and SPA-mPEG with primary amines. Primary amine groups on cell surface proteins and latex particles act as nucleophiles in SN1 substitution reactions releasing the free N-hydroxy succinimide linker group and forming a covalent bond with the polymer. SC-mPEG forms a carbamate or urethane bond and SPA- and SVA-mPEG form an amide bond. The chemical reactivity of the linker is determined by the chemical properties of the activating group. An important consideration is the reactivity of the activating group towards amine targets vs. water molecules. In an aqueous solution water molecules compete with amine groups to act as nucleophiles in the SN1 substitution reaction, Figure 2-3. If the water molecule reacts with the activated polymer, linker hydrolysis results rendering an mPEG molecule incapable of covalent attachment to the substrate. The hydrolysis half-life (t1/2) of the activated mPEG is, therefore, defined as the time (minutes) it takes for half of the activated mPEG molecules to lose their linkers and become inactive (pH 8.0 and 25 ºC). The hydrolysis halflives of SC, SVA and SPA linkers are 20.4, 30.6 and 16.5 minutes, respectively.  68  Figure 2-3: Hydrolysis of SC-, SVA- and SPA-mPEG. Hydrolysis of activated mPEG species is mediated by water molecules acting as nucleophiles in SN1 substitution reactions. The linker group (N-hydroxy succinimide) becomes released and the polymer chain is no longer capable of being covalently attached to the primary amines. All activated mPEG species used in this thesis (Laysan Bio, Inc., Arab, AL, USA) were linear, i.e., non-branched and monofunctional (one linker per polymer molecule) polymers. However, for investigations of the effects of polymer architecture on the efficacy of PEGylation, highly branched, multi-functional polyglycerol (PG) polymer was also examined in some studies, Figure 2-4A. The individual 3, 8 and 25 kDa PG polymer chains contained 3, 5 and 6 linkers, respectively. PG used an NHS linker of the same chemical reactivity as SVA-mPEG (Figure 2-4B) and therefore, direct comparisons were made between the two polymers. Three different molecular weights of polymers were tested in order to determine the effects of polymer size on the efficacy of immunocamouflage. PEGylation studies utilized 2, 5 and 20 kDa species and PG studies used the 3, 8 and 25 kDa polymer chains. The covalent grafting of polymers to the latex particles was performed by dissolving activated polymers in PEG buffer (50 mM K2HPO4, 105 mM NaCl, pH 8.0) and adding it to a 2.0 weight percent (% w/v) aliphatic amine polystyrene latex particle suspension (1.2 #m or 8.0 #m in diameter) to reach the desired final mPEG or NHS-PG grafting concentration. Samples were incubated for 60 minutes at room temperature (RT) with constant mixing. Following 69  derivatization, the latex samples were washed 5 times with water to remove unbound/hydrolyzed polymer, made up to 2.0% w/v suspension and stored at 4ºC for the subsequent evaluation. All derivatization reactions included two controls: a negative control consisting of a latex sample to which PEG buffer alone was added; and a soluble mPEG/PG control consisting of a latex sample to which non-activated soluble 2.0 mM (1.2 #m latex) or 5 mM (8.0 #m latex) 5 kDa mPEG or 2 mM, 8 kDa PG (1.2 #m latex) was added. Additionally, a fluorescent fluorescein-SVA-mPEG (fluor-SVA-mPEG) (Laysan Bio, Inc., Arab, AL, USA) polymer was utilized in order to measure the extent of latex modification. The fluorescent polymer was diluted to 5% (1.2 #m latex) or 1% (8.0 #m latex) with nonfluorescent polymer of the same linker chemistry and molecular weight. Latex particles were modified with the 5 kDa fluorescent polymer at 0-2 (1.2 #m latex) and 0-20 mM (8.0 #m latex) grafting concentrations. For the soluble control sample, non-activated fluor-mPEG was obtained by hydrolyzing activated fluor-SVA-mPEG in PEG buffer at pH 8.0 for 24 hours at room temperature, protected from light. Modified particles were analyzed via flow cytometry in order to quantify the amount of the covalently bound fluorescent polymer. Mean particle fluorescence values were reported, corrected for the fluorescence of the bare latex control sample.  70  Figure 2-4: NHS-Polyglycerol structure and chemical reactivity. A) Polyglycerol is a highly branched polymer possessing several reactive linkers per individual polymer chain. B) Polyglycerol utilizes an NHS linker group that has the same chemical reactivity as an SVA linker of the mPEG polymers. NHS-PG reacts with primary amine groups analogously to SVA-mPEG and forms a stable covalent amide bond with the release of a free NHS group.  2.1.3  MPEG QUANTIFICATION WITH ANTI-PEG ANTIBODY The quantification of the amount of surface bound mPEG was performed by estimating  the amount of specific anti-PEG antibody bound to modified latex particles. In these experiments 33 and 100 #l of 2.0% w/v differentially modified (2, 5, 20 kDa SVA-mPEG) 1.2 #m (0-2 mM) and 8.0 #m particles (0-10 mM) were utilized, respectively. Latex suspensions were first reacted with 20 #l of primary (1º) goat-anti-PEG antibody (IgG1 isotype; Biodesign 71  International, Saco, ME, USA) followed by incubation with the 20 #l of the secondary (2º) fluorescein isothiocyanate (FITC) conjugated goat-anti-mouse antibody (IgG isotype, SigmaAldrich, Saint Louis, MO, USA). Each antibody incubation was performed for 30 minutes at room temperature followed by 3 washes in 100 #l of 0.02 M azide phosphate buffer (0.02 M phosphate buffer with 3 mM NaN3, pH 7.4). Both, the primary and the secondary antibodies were diluted 40 (1.2 #m latex study) and 80 times (8.0 #m latex study) with 0.02 M azide phosphate buffer. The following controls were utilized in each experiment to ensure the absence of nonspecific primary and secondary antibody binding to either bare latex or to mPEG itself. Control samples in the 1.2 #m study included: 1) negative control – bare latex alone; 2) bare latex control – bare latex incubated with 1º and 2º antibodies; 3) grafted mPEG control – 2 mM, 20 kDa SVAmPEG modified latex incubated with 2º antibody; 4) soluble mPEG control – latex reacted with 2 mM, 5 kDa soluble, unactivated mPEG during PEGylation reaction and incubated with 1º and 2º antibodies. In the 8.0 #m study, the same control samples were utilized, except no soluble mPEG control was used and the grafted mPEG control consisted of 10 mM, 20 kDa SVA-mPEG modified latex. All latex samples were analyzed using flow cytometry to measure the amount of secondary fluorescent antibody binding. Mean particle fluorescence values were reported, corrected for the fluorescence of the bare latex sample without any bound antibody.  2.1.4  LATEX MICROAGGREGATION STUDIES Following each covalent modification reaction, latex was microscopically (visually)  evaluated for signs of latex aggregation. This served three purposes. Firstly, it indicated whether the reaction conditions (salt concentration, pH) were suitable to maintain the latex particles in the colloidal suspension. Secondly, it ensured that the polymer did not aggregate the latex due to polymer crosslinking or entanglement. Thirdly, it verified that the subsequent 72  experiments were performed with monodisperse latex suspension for uniform exposure to solute molecules (e.g., polymers, plasma proteins and antibodies). Latex particles (1.2 #m) were differentially modified with SC-, SVA-mPEG (2, 5, 20 kDa) and NHS-PG (3, 8, 25 kDa) at 0-2 mM grafting concentrations. A 50 times diluted 2.0% w/v latex suspension was placed on a glass slide and covered with a cover slip. The laboratory microscope setup consisting of an Olympus CK40 microscope (Olympus America Inc., Melville, NY, USA) fitted with a Q-Imaging camera (QICAM FAST, Qimaging Corporation, Surrey, BC, Canada) was used. Transmitted light images were taken at 40X magnification with Q-Capture imaging software (v.2.8.1, Qimaging Corporation, Surrey, BC, Canada).  2.1.5  PARTICLE ELECTROPHORETIC MOBILITY STUDIES The efficacy of mPEG-mediated surface charge camouflage was assessed by subjecting  particles to electrophoretic mobility measurements. The mobility of particles in an electric field was directly proportional to the surface charge of the latex according to (135): ! = U/E = $/)* where, # is the electrophoretic mobility, U – velocity of a particle, E – strength of electric field, $ – surface charge density, ) – viscosity of the medium and 1/* –thickness of the electrical double layer. Latex particles were derivatized with either SC-mPEG (2, 5, 20 kDa), SVA-mPEG (5, 20 kDa) or NHS-PG (3, 8, 25 kDa) at 0-2 mM (1.2 #m latex) and 0-20 mM (8.0 #m latex) grafting concentrations. Electrophoretic mobility measurements for the 1.2 #m particles were performed in triplicate at the negative stationary level of a Doppler Electrophoretic Light Scattering Analyzer (DELSA440SX) (Beckman Coulter, Fullerton, CA, USA) with the following settings: 25±0.02 ºC, 0.7 mA, 500 Hz and a 60 second run time with a 2.5 second on time and 0.5 second off time, Figure 2-5. A 2.0% w/v modified latex suspension was diluted 1000 times in 10 mM 73  NaCl solution prepared in Ultrapure distilled water (Invitrogen, Carlsbad, CA, USA). The electrophoretic mobility cell was calibrated against a Traceable® conductivity standard (Fisher Scientific, Pittsburgh, PA, USA) and a Zeta potential/Mobility control (Photal, Otsuka Electronics, Osaka, Japan). Electrophoretic mobility measurements for the 8.0 #m latex particles were performed using a Zeta-Meter 3.0+ (Zeta Meter, Inc., Staunton, VA, USA). The electrophoretic cell was filled with the 100 times diluted 2.0% w/v latex solution in 10 mM NaCl prepared in distilled water. An ocular micrometer and a stop watch were used to measure the time required for a particle to traverse a single division of 160 #m at the stationary level of the electrophoretic cell. At least 10 distinct particles were analyzed per sample. Measurements were performed using a molybdenum electrode at 21ºC and 250 V. The electrophoretic mobility (EM) was calculated based on the formula in Figure 2-6. All mobility values were expressed as a percent of mobility of unmodified latex particles.  Figure 2-5: DELSA electrophoretic mobility measurements. A) The DELSA apparatus consists of the laser source, the electrophoretic cell and 4-angle detectors. Positively charged bare latex particles move in the electric field applied between the cathode (-) and the anode (+), towards the cathode. The laser emits the light that is scattered off the moving particles and is captured by the detectors. B) Variations in the scattered light detected give the estimate of the particle’s mobility. This is exported as a mobility plot, where location of the peak indicates the absolute mobility value of the particles.  74  Figure 2-6: Zeta-meter electrophoretic mobility measurements. The zeta-meter apparatus consists of an electrophoretic cell where positively charged bare latex particles move towards the cathode when the electric field is applied between the cathode (-) and the anode (+). The ocular micrometer is used to trace the particles at the stationary level of the electrophoretic cell and measure the time it takes for a particle to traverse a single division of 160 #m. The formula is used to calculate the electrophoretic mobility (EM) based on the time measured and the voltage applied.  2.1.6  HUMAN PLASMA PROTEIN ADSORPTION Synthetic – “plastic” surfaces, such as polystyrene latex particles, are known to adsorb  vast amounts of proteins immediately upon exposure to blood or blood plasma (136-139). This phenomenon was used to estimate the mPEG-mediated inhibition of surface-macromolecule interactions, i.e., protein adsorption. Following informed consent, whole blood was collected from healthy volunteers into 10 ml ethylenediaminetetraacetic acid (EDTA) anticoagulant containing tubes. Human plasma was obtained by centrifugation at 1,000xg, 25ºC for 5 minutes. Plasma protein adsorption studies were carried out using both unlabeled and fluorescently labeled plasma. Unlabeled plasma protein adsorption experiments were conducted on differentially modified (0-5 mM, 5 kDa SC-mPEG) 1.2 #m latex particles. These experiments were used for quantitative estimation of the total amount of the surface adsorbed protein and for the identification of these proteins by a mass spectrometry analysis, as will be discussed later in this chapter. Unlabeled protein adsorption was performed on 500 cm2 of latex surface in a 50% plasma-latex suspension (2.0% w/v).  75  Fluorescent plasma protein adsorption studies were performed using Alexa 488 (Molecular Probes, Invitrogen, Carlsbad, CA, USA) labeled plasma, Figure 2-7. These experiments were aimed at investigating the effects of polymer linker chemistry, molecular weight, surface density and target size on the efficacy of surface camouflage. The fluorescent labeling of plasma proteins was performed according to the manufacturer’s protocol without implementing any changes. Fluorescent protein adsorption was conducted on differentially modified 1.2 #m particles (0-2 mM; 2, 5, 20 kDa SC- and SVA-mPEG and 3, 8 and 25 kDa NHS-PG) and 8.0 #m particles (0-20 mM; 2, 5, 20 kDa SVA-mPEG). The total latex surface areas of the 1.2 and 8.0 #m particles used in this study were 83.3 cm2 and 70.0 cm2, respectively. The 1.2 and 8.0 #m particles utilized 83.3% (diluted with unlabeled plasma) and 100% fluorescent protein. Beads were incubated with unlabeled or fluorescent plasma for 60 minutes at room temperature, with constant mixing. Following adsorption, latex particles were washed five times in 0.02 M phosphate buffer with 3 mM NaN3, pH 7.4 to remove unbound protein. Latex samples to which unlabeled human plasma proteins were adsorbed were further subjected to protein desorption and desorbed protein analysis as described in section 2.1.7. Latex particles with adsorbed fluorescent proteins were subjected to adsorbed protein analysis as described in section 2.1.8.  76  Figure 2-7: Alexa-488 fluorescent plasma protein adsorption. Human plasma proteins are labeled with Alexa-488 fluorescent dye. Size exclusion gel chromatography is used to separate the free dye from the labeled proteins. Fluorescent protein adsorption is performed on differentially modified latex particles. The amount of fluorescent protein adsorbed to latex is assessed by measuring mean particle fluorescence via flow cytometry.  2.1.7  DESORBED PROTEIN ANALYSIS  Refer to Figure 2-8 for a flow chart 2.1.7.1 HUMAN PLASMA PROTEIN DESORPTION In order to quantitatively and qualitatively study the effects of PEGylation on surfacemacromolecule interactions, latex adsorbed proteins were removed from the bead surface using a protein desorption procedure. Latex bound proteins were desorbed from the bare and differentially modified PEGylated beads by the addition of solubilization buffer (2% sodium dodecyl sulfate (SDS) and 80 mM dithiothreitol (DTT) in water) and incubation for 5 minutes in a 95ºC water bath. Latex samples were centrifuged for 5 minutes at 16,000xg and the supernatants containing desorbed proteins were collected and stored at -20ºC for subsequent evaluation.  77  Figure 2-8: Flow chart for methodology in biophysical latex model. Unlabeled or fluorescent plasma protein adsorption is performed on differentially modified latex particles. Latex to which fluorescent protein was adsorbed is subjected to flow cytometry or quantitative fluorescent microscopy. A portion of latex with adsorbed fluorescent protein is also subjected to desorption followed by flow cytometry. Latex to which unlabeled protein was adsorbed is subjected to desorption. Desorbed unlabeled proteins are further analyzed via SDS-PAGE, colorimetric protein assay and iTRAQ/MS analysis.  2.1.7.2 COLORIMETRIC PROTEIN ASSAY The quantitative measurement of the amount of latex adsorbed protein was performed by utilizing a reducing agent/detergent-compatible (RC DC) colorimetric protein assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA). This assay is based on the reaction of proteins with an alkaline copper tartarate solution and sodium 1,2-naphthoquinone-4-sulfonate (Folin) reagent as initially described by Lowry (140). Unlabeled proteins adsorbed to 500 cm2 of 1.2 #m latex, modified with SC-mPEG (0-2 mM; 2, 5, 20 kDa), were desorbed with 70 #l of solubilization buffer (section 2.1.7.1) and the duplicate protein samples of 25 #l were subjected to protein quantification. The assay was conducted according to the manufacturer’s instructions with the only exception: one extra wash was implemented prior to adding the color reagent. A standard curve was constructed concurrent with each assay using a known amount of standard bovine serum albumin (BSA) solution (2.0 mg/ml, Thermo Fisher Scientific, Inc., Rockford, IL, USA). 78  All values were corrected for bare latex control not exposed to plasma and expressed as the amount of protein (ng) adsorbed to 1 cm2 of latex surface.  2.1.7.3 SDS-PAGE ANALYSIS In a more detailed analysis of desorbed proteins the pattern of protein adsorption was investigated. Unlabeled proteins were desorbed (as described in section 2.1.7.1) from 500 cm2 of differentially modified (0-2 mM; 5 kDa SC-mPEG) 1.2 #m particles with 60 #l solubilization buffer and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (141). The same volume of desorbed protein solutions (33 µl) was loaded onto a 7%, 1 mm thick polyacrylamide gel. Electrophoresis was performed at room temperature, with 25 mA current, for 60 minutes. Each gel included lanes containing: sample buffer only (10 µl); molecular weight standards (7 µl; Kaleidoscope precision plus protein standards; Bio-Rad Laboratories, Inc., Hercules, CA, USA); and 100 times diluted human plasma in sample buffer (20 µl). Following SDS-PAGE, the gel was stained with Pierce ImperialTM protein stain (Thermo Fisher Scientific, Inc., Rockford, IL, USA) according to the manufacturer’s instructions. A picture of the gel was taken with an Epson 2450 scanner (Toronto, ON, Canada) and kept for reference and further analysis. A semi-quantitative estimation of the amount of total protein adsorbed to latex particles was performed using densitometry. An optical density measurements utilized the Gene Tools software (v. 3.08; SynGene, Cambridge, UK) and a gel image adjusted to +20/+20 for brightness/contrast. Total OD of each lane was measured and normalized to the 50 kDa band in the molecular weight standard to address any variations among the gels. Furthermore, each value was expressed as a percent OD of the bare latex control lane to address any variations within the gel.  79  2.1.7.4 ITRAQ LABELING AND MASS SPECTROMETRY ANALYSIS Further to SDS-PAGE analysis, the composition of the adsorbed protein layer was studied. Based on the protein assay data, 200 #g of unlabeled protein was desorbed from an estimated amount of latex using 200 #l of desorption buffer. The supernatants containing desorbed protein were submitted to the University of Victoria Proteomics Centre (www.proteincentre.com) for isobaric tags for relative and absolute quantification (iTRAQ) labeling and mass spectrometry (MS) analysis (142), Figure 2-9. Briefly, four protein samples desorbed from four differentially modified latex particles (0, 0.5, 2.0 and 5.0 mM, 5 kDa SCmPEG), were digested with Trypsin and the peptides were labeled with the iTRAQTM reagent (Applied Biosystems, Foster City, CA, USA). Labeled peptides were subjected to liquid chromatography followed by tandem mass spectrometry (LC-MS/MS) using an integrated Famos autosampler, SwitchosII switching pump and UltiMate micro pump (LC Packings a Dionex Company, Sunnyvale, CA, USA). The system was equipped with a Hybrid Quadrupole-Time Of Flight (TOF) LC/MS/MS mass spectrometer (QStar, Mary Esther, FL, USA), a nanoelectrospray ionization source (Proxeon, Odense, Denmark) and a 10 #m fused-silica emitter tip (New Objective, Inc., Woburn, MA, USA). Mass spectrometry data was processed using Protein Pilot software (v.2.0, Applied Biosystems, Foster City, CA, USA). Three stringent criteria were applied to the analysis: 99.0% confidence in peptide identification, p!0.01 to render changes in peptide amounts significant, and identification of each protein was based on at least four distinct peptides. All specific protein amounts were measured per 100 #g total protein adsorbed to latex and expressed relative to the bare latex control sample, which was set to equal 100%.  80  Figure 2-9: iTRAQ labeling and mass spectrometry analysis. Four protein samples, containing 100 #g total protein, were desorbed from four differentially modified latex particles: 0, 0.5, 2 and 5 mM, 5 kDa SC-mPEG. Proteins were enzymatically digested to peptides and all peptides coming from the same sample were labeled with the unique isobaric tag. Each tag was covalently bound to peptides via a peptide reactive group that had the same molecular weight (m) of 145 g/mol. The uniqueness of each tag came from the differentially sized reporter groups. The balance group balanced the molecular weight of the reporter group and the peptide reactive group to give each tag the same molecular weight. Following iTRAQ labeling all peptides coming from different protein samples were pooled and subjected to MS analysis that fragmented the labeled peptides and the isobaric tags (dashed lines). Peptide fragmentation allowed for protein identification and the amount of each reporter group allowed for the quantitative estimation of the identified protein in each latex desorbed sample.  81  2.1.8  ADSORBED PROTEIN ANALYSIS  Refer to Figure 2-8 for a flow chart 2.1.8.1 FLOW CYTOMETRY The effects of polymer chemical and physical properties as well as the target size on the prevention of surface-macromolecule interactions were studied on derivatized 1.2 #m (2, 5, 20 kDa SC/SVA-mPEG and 3, 8, 25 kDa NHS-PG; 0-2 mM) and 8.0 #m (2, 5, 20 kDa SVAmPEG; 0-20 mM) latex particles. Modified particles were subjected to fluorescent protein adsorption followed by flow cytometric analysis in order to measure the amount of protein adsorbed to the latex surface. Latex samples, from which fluorescent proteins were desorbed, were also evaluated using flow cytometry to measure the residual fluorescence due to the fluorescent protein that was not successfully removed. All fluorescent intensity values were corrected for the negative control sample of bare latex with no fluorescent protein adsorbed. Mean particle fluorescence was expressed as a percent of fluorescence of bare latex particle with adsorbed fluorescent protein (100%).  2.1.8.2 QUANTITATIVE MICROSCOPY Following fluorescent plasma protein adsorption, latex particles (1.2 #m and 8.0 #m) were subjected to quantitative fluorescence microscopy. These experiments were utilized to further measure and visualize the extent of fluorescent protein adsorption and to study the effects of target size on the optimal length of the polymer required for surface camouflage. Quantitative microscopy with 1.2 #m modified latex (0-2 mM; 2, 5, 20 kDa SVA-mPEG) was performed at the University of British Columbia Bioimaging Facility (Vancouver, BC, Canada) using Quorum Wave FX spinning disk confocal microscope equipped with water immersion 63X objective (Yokogawa Electric Corporation, Tokyo, Japan). Image acquisition and instrument control were performed by Velocity software (Improvision a Perkin Elmer Company, Coventry, England). 82  Quantification of fluorescence associated with fluorescent plasma protein adsorbed to the 1.2 #m latex was performed using Open Lab5 software (v.5.5.0, Improvision Open Lab, Improvision a Perkin Elmer Company, Coventry, England). For each sample, three fields of view were captured, with at least 10 particles per field and for each field both transmitted light and fluorescent light images were taken. The quantitative microscopy studies with 8.0 #m modified latex (0-20 mM; 2, 5, 20 kDa SVA-mPEG) were performed using the laboratory setup (described in 2.1.4). Fluorescent and transmitted light images of latex particles were taken at 20X and 10X magnification, respectively. The Open Lab5 software was used for the acquisition and image analysis. Four independent fields of view were considered with at least 7 particles per field. For each field, both transmitted light and fluorescent light images were taken. For both the 1.2 and the 8.0 #m studies, background values were appropriately subtracted from the mean particle fluorescent intensity that was expressed as a percent of bare latex with adsorbed fluorescent protein (100%).  83  2.2  BIOLOGICAL PBMC AND RBC MODEL  2.2.1  PBMC AND RBC EXTRACTION Following informed consent, whole blood was collected from healthy volunteers into 8  ml sodium heparin or EDTA containing tubes for peripheral blood mononuclear cell (PBMC) and red blood cell (RBC) extractions, respectively. PBMCs were obtained using Ficoll-Paque Premium (GE Healthcare, Piscataway, NJ, USA) according to the manufacturer’s protocol without any modifications. Fresh packed RBCs were collected by washing whole blood three times in isotonic saline (150 mM NaCl).  2.2.2  CELL DERIVATIZATION Extracted PBMCs and RBCs were made up to 4*106 cell/ml and 12% hematocrit  (1.5*109 cells/ml), respectively, in PEG buffer (50 mM K2HPO4, 105 mM NaCl, pH 8.0). A known volume of the PBMC suspension was added to 15 ml round bottom tubes containing a calculated amount of SVA-mPEG (2 and 20 kDa) to reach the desired grafting concentration (0-4 mM). In the case of RBCs the PEGylation reaction was performed using SPA-mPEG (2, 5, 20 kDa; 0-6 mM). Both the RBC and the PBMC cell suspensions were mixed by gentle inversion until all of the mPEG was dissolved. The PEGylation reaction was performed for 60 minutes at room temperature. Following PEGylation, PBMCs and RBCs were washed three times with RPMI 1640 medium (Invitrogen, Carlsbad, CA, USA) and saline, respectively. The washes were performed by successive centrifugation at 400xg for 10 minutes and resuspension in twice the volume of the initial PEGylation reaction. Cell counting was performed for PBMCs using a hemacytometer following Trypan Blue staining (Sigma-Aldrich, Oakville, ON, Canada). A cell suspension of white blood cells was added to the trypan blue stain at a 1:1 ratio and incubated at room temperature for 5 minutes. Trypan blue stain is an exclusion dye that stains dead cells and allows discrimination between dead and live cells during cell counting. 84  2.2.3  CELL ELECTROPHORETIC MOBILITY In order to assess the degree of surface charge camouflage imparted by PEGylation of  RBCs, cells were subjected to particle electrophoresis using the Rank Mark I electrophoresis apparatus. This particle analyzer was equipped with a horizontal microscope and a water immersion lens (143). Electrophoretic mobility measurements of unmodified control and SPAmPEG modified RBCs (2, 5, 20 kDa; 0-6 mM) were performed on 10 randomly chosen cells as described in section 2.1.5 (Figure 2-6).  2.2.4  CD SURFACE MARKER ANALYSIS For the evaluation of the efficacy of immunocamouflage on human PBMCs, CD marker  analysis was performed. If efficiently camouflaged, CD surface markers would not be bound by fluorescently labeled anti-CD antibodies and therefore, cells would not generate a positive signal in the flow cytometric analysis. Three CD markers were specifically chosen to reflect differential stages during T cell activation as well as the complexity of cell topography. CD3 is a component of the TCR complex and is implicated in the initial stages of antigen recognition and signal transduction (Figure 1-10). It is the shortest antigen tested in our studies (4 nm) (144, 144). CD4 is a TCR co-receptor. It binds to the MHC II complex following initial antigen recognition by the TCR complex and therefore functions downstream of CD3 (Figure 1-11). CD4 binding completes the first set of binding events between an APC and a T cell and sends the first activating signal to a T cell. CD4 is a fairly tall antigen of 13 nm in height (145). CD28 is a co-stimulatory molecule that binds downstream of CD3 and CD4 and is responsible for generation of the second T cell activating signal, which results in complete T cell activation and proliferation. CD28 is of intermediate height and is 7 nm tall (146). CD surface marker analysis was performed following PBMC derivatization (0-4 mM SVA-mPEG; 2 and 20 kDa) and cell counting. A cell suspension (100 µl) containing a total of 85  1*106 cells was placed into flow tubes and washed once by centrifugation at 400xg for 5 minutes and resuspension in 100 #l PBS (pH 7.2). Immunofluorescence staining was performed by adding 20 #l of either fluorescent antibody or the appropriate isotype control and incubating samples on ice for 30 minutes in the dark. The following antibodies and isotype controls were used, respectively: anti-CD3 PerCP and IgG1 PerCP; anti-CD4 PE and IgG1 PE; anti-CD28 FITC and IgG1 FITC (BD Biosciences, San Jose, CA, USA). The immunoglobulins of the same specificity (IgG) and fluorochrome as the antibodies, were used to verify that there was no fluorescence due to non-specific adsorption of antibodies to either the grafted mPEG or the cell surface. Following staining, the cells were washed twice with 2 ml of ice-cold wash buffer (PBS, 1% bovine serum albumin, 0.1% NaN3) by successive centrifugation (400xg at 4ºC, 5 minutes) and resuspension. During the last wash, cells were resuspended in 0.5 ml wash buffer and subjected to flow cytometry. Analysis was performed based on the mean cell fluorescence (MCF) and the percent gated cell (PGC) values. The mean cell fluorescence was expressed as a percent of unmodified control cells incubated with fluorescent antibodies. Mean fluorescence values for the negative control, unmodified PBMCs not incubated with antibodies, were subtracted from all the readings.  2.2.5  CFSE CELL PROLIFERATION ASSAY In order to assess the degree of the T cell proliferation during a 2-way mixed lymphocyte  reaction (MLR; section 2.2.6), CellTraceTM carboxyfluorescein diacetate succinimidyl ester (CFSE) cell proliferation assay was used (Molecular Probes, Invitrogen, Carlsbad, CA, USA). This methodology allowed for non-radioactive, fast, and accurate measurements of T cell proliferation using flow cytometry. Staining was performed on freshly isolated human PBMCs prior to mPEG derivatization and MLR, according to the manufacturer’s protocol without implementing any changes. CFSE is colorless and non-fluorescent until its acetate groups are 86  cleaved by intracellular esterases (147). Thus, inside the cell it is converted into highly fluorescent carboxyfluorescein succinimidyl ester. Succinimidyl ester reacts covalently with intracellular amines and becomes retained inside a cell throughout the cell’s lifecycle. During cellular division the fluorescent label is inherited by the daughter cells. Therefore, with each successive division the fluorescence transferred to the daughter cells becomes progressively lower, Figure 2-10. Such a decrease in mean cell fluorescence can be observed via flow cytometry and is indicative of the cell division.  Figure 2-10: CFSE T cell proliferation assay. Shown are the results of flow cytometric analysis of CFSE stained human lymphocytes. A) The resting lymphocyte population is located in the right hand side of the histogram (outside of the M2 gate) due to the high mean cell fluorescence. B) Mitogen stimulated lymphocytes exhibit a high degree of proliferation and the corresponding cell divisions can be seen on the histogram as peaks inside an M2 gate. Each successive cell division is accompanied by a progressive decrease in CFSE intensity.  87  2.2.6  2-WAY MIXED LYMPHOCYTE REACTION A mixed lymphocyte reaction (MLR) is an in vitro measure of MHC class II disparity  between individuals. It assesses a T cell response to differences in multiple HLA class II determinants – DR, DQ and DP. Before the invention of DNA-based HLA-typing techniques, MLR was used to perform donor-recipient cross-match for bone marrow transplantations (148, 149). In the clinic, 1-way MLR experiments were performed where one PBMC population was irradiated to render the cells incapable of activation and proliferation. The irradiated cell population then acted as the stimulator population. When two disparate samples of PBMCs (from two unrelated individuals) were mixed together, the cell population that was capable of dividing acted as the responder population. The outcome of MLR experiments was an expansive T cell proliferation due to recognition of foreign antigenic epitopes on APCs by T cells. This T cell proliferation was traditionally measured by 3H-Thymidine incorporation into the newly synthesized DNA of the dividing T cells. In our studies, the irradiation of one of the cell populations was not performed, therefore both populations were capable of presenting and recognizing a foreign antigen, hence a 2-way MLR. In addition, instead of using the radioisotope measurement of 3H-Thymidine incorporation, a flow cytometric CFSE fluorescent assay was performed. The PBMCs from two unrelated donors were extracted. The two populations were referred to as PBMC1 and PBMC2 and only one of the populations was PEGylated. Following CFSE staining, cells were derivatized with 2 and 20 kDa SVA-mPEG at 0-2 mM grafting concentrations. Following derivatization, PBMCs were counted using a hemacytometer and Trypan Blue staining (described above) and resuspended in AIMV medium (Gibco, Invitrogen, Carlsbad, CA, USA). Cells were seeded in 24-well plates in triplicate (3 wells per condition) at a concentration of 1*106 cells/ml/well.  88  The following controls were used: AIMV medium alone – control for contamination; PBMC1 alone and PBMC2 alone – PBMC negative controls; PBMC1 alone and incubated with PEG buffer only, during PEGylation reaction – PEG buffer negative control; PBMC1 alone and PEGylated with 2 kDa or 20 kDa, 2 mM SVA-mPEG – negative mPEG controls; PBMC1 and PBMC2 alone and stimulated with phytohaemagglutinin (PHA) – positive controls. The PBMC, PEG buffer and mPEG negative controls were used to assess the background activation levels of PBMCs following extraction, PEG buffer treatment and PEGylation at the highest grafting concentration, respectively. PHA is a mitogen that stimulates cell division. Therefore, its addition was a test for the ability of the cells to be stimulated and to undergo proliferation. The following experimental wells were tested: PBMC1 + PBMC2 – positive control MLR, PBMC1 differentially derivatized + PBMC2 – experimental MLRs. In the MLR wells 0.5*106 cells (0.5 ml of cell suspension) were added from each donor. Cells were incubated for 13 days, at 37ºC, 5% CO2. On Day 13, cells were harvested into flow tubes by centrifugation (400xg, 5 minutes) and washed once in 2 ml of cold PBS buffer with 0.2% BSA. Harvested PBMCs were stained with anti-CD3 PE/Cy5 and anti-CD4 APC antibodies (BD Biosciences, San Jose, CA, USA) to distinguish the T cell population during flow cytometric analysis. Fluorescent staining was carried out at room temperature for 1 hour in the dark followed by one wash in 2 ml of cold PBS buffer with 0.2% BSA (centrifugation at 400xg, 5 minutes). Samples were analyzed via flow cytometry to measure the extent of lymphocyte proliferation. The data was presented as percent gated values corrected for background activation in both PMBC1 and PBMC2 control wells divided by 2.  2.2.7  PI AND 7-AAD VIABILITY ASSAYS The effects of PEGylation (0-2 mM; 2, 5, 20 kDa SVA-mPEG) on cell viability were  assessed using two viability assays: Propidium Iodide (PI; Sigma-Aldrich, Oakville, ON, 89  Canada) and 7-amino-actinomycin D (7-AAD; BD Biosciences, San Jose, CA, USA). PI (150) and 7-AAD (151) are fluorescent nucleic acid dyes that cannot cross cellular membranes and are excluded from viable cells. If an entry into the cell is gained (dead or dying cells) dyes bind to nucleic acids, become fluorescent and can be detected by flow cytometry. The two assays were utilized in order to detect any discrepancies and to ensure the reproducibility of the results. Differentially PEGylated PBMCs were seeded in AIMV medium (24-well plates) at a concentration of 1*106 cells/ml/well. Cells were incubated at 37ºC, 5% CO2 and harvested on Day 0, Day 1, Day 7 and Day 13 following PEGylation. Decrease in the viability on Days 0 and 1 reflected the acute toxicity, whereas reduced viability on Days 7 and 13 was indicative of chronic toxicity. Day 13 incubation was reflective of changes during the 13-day MLR experiment. This time point was also tested to see whether prolonged in vitro cell incubations impacted the viability of the cells. PBMCs were collected into flow tubes by centrifugation at 400xg for 5 minutes at room temperature. Cells were washed twice with 1 ml of wash buffer by successive centrifugation (400xg, 5 minutes) and resuspension. During the final wash, cells were resuspended in 100 #l of wash buffer and either 5 #l of 7-AAD (0.05 mg/ml) or 1 #l of PI (0.05 mg/ml) dye was added. Cell suspensions were incubated at room temperature for 10 minutes followed by the addition of 400 #l of wash buffer. Samples were subjected to flow cytometry analysis right after staining. Percent gated values were obtained, corrected for loss of viability in stained unmodified cell samples.  90  III. RESULTS 3.0  OVERVIEW The Results chapter is subdivided into sections according to the model utilized. The  aliphatic amine polystyrene latex model was used to investigate the biophysical mechanisms involved in the immunocamouflage of surfaces imparted by PEGylation. The immunocamouflage of the latex surface was evaluated based on the masking of the inherent positive charge of the latex and the prevention of surface-macromolecule interactions, such as human plasma protein adsorption. The efficacy of PEGylation on latex particles demonstrated a strong dependence on the chemical (e.g., linker chemistry) and physical (e.g., molecular weight) properties of the polymer, as well as the target size (1.2 vs. 8.0 #m beads). The peripheral blood mononuclear cell (PBMC) model was utilized to correlate the biophysical findings of the latex model with the biological consequences of PEGylation of cells. Immunocamouflage of the cell surface was measured as the inhibition of antibody and cell mediated mechanisms of allorecognition. These effects were tested by decreased antibody binding to specific CD surface antigens on T cells and by prevention of T cell activation and proliferation in a 2-way mixed lymphocyte reaction (MLR), respectively. The results of the biological model correlated with the biophysical findings and the same polymer and target effects were noted. In addition, the efficacy of cell surface camouflage was highly dependent on the topography of the cell and the distribution of the cell proteins targeted for mPEG modification. Furthermore, PEGylation of PBMCs was found to cause no significant cell toxicity at immunoprotective levels and negatively correlated with increased cell modification and prolonged in vitro cell culturing.  91  3.1  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX  MODEL 3.1.1  FLUORESCEIN-SVA-MPEG GRAFTING TO POLYSTYRENE LATEX  PARTICLES In order to ensure that activated mPEG covalently grafted to the latex particles and to quantify the degree of surface modification, PEGylation experiments were conducted on 1.2 and 8.0 #m particles modified with the 5 kDa fluor-SVA-mPEG and analyzed by flow cytometry. Covalently bound fluor-mPEG rendered particles fluorescent and allowed the quantitative measurement of fluorescent intensity via flow cytometric analysis. Polymer grafting was performed at 0-2 mM grafting concentrations for the 1.2 #m latex and at 0-20 mM concentrations for the 8.0 #m particles. These studies exhibited a dose-dependent, biphasic increase in mPEG surface grafting, Figure 3-1. The biphasic curves demonstrated rapid polymer binding, (i.e., a steep slope, S1) at low grafting concentrations and slower polymer addition (i.e., a shallow slope, S2) at higher grafting concentrations. The change in the slope was most pronounced with the 1.2 #m beads and occurred at the 0.5 mM concentration. Thus, increase in grafting concentration from 0.2 mM to 0.5 mM resulted in a slope change of 358.4 fluorescent units (p!0.006) in comparison to only 100.9 units for the increase from 0.5 mM to 1.0 mM grafting concentration (p"0.6). For the 8.0 #m particles, change in the slope occurred at 10 mM concentration and was less pronounced suggesting that the initial rate of the mPEG grafting to large particles decreased to a lesser extent than for the 1.2 #m beads. In addition, using hydrolyzed (no SVA linker), unactivated fluor-mPEG, some nonspecific polymer adsorption to latex was detected. With 1.2 #m particles, unactivated fluormPEG adsorption was negligible, Figure 3-1A. However, for the 8.0 #m particles a more  92  pronounced non-covalent adsorption effect was observed that was subtracted from the initial fluorescence to yield values due only to covalently bound fluor-mPEG, Figure 3-1B. Thus, this study demonstrated that SVA-mPEG was capable of covalent binding to the surface of latex in a dose-dependent manner. The rate of the polymer addition demonstrated both a rapid (S1) and a slow (S2) phase on both sizes of particles. Polymer binding decreased at higher grafting concentrations due to the steric hindrance arising from the initially grafted mPEG. This effect was most pronounced for the 1.2 #m latex due to the dramatically decreased surface area in comparison to the 8.0 #m beads: 4.5 #m2 vs. 201.0 #m2 per particle, respectively.  Figure 3-1: Modification of 1.2 !m (A) and 8.0 !m (B) particles with 5 kDa fluor-SVAmPEG demonstrated a dose-dependent increase of the covalently bound polymer as measured by flow cytometry. (A) and (B) Polymer grafting demonstrated a biphasic curve with rapid binding (S1) at lower concentrations followed by slower binding (S2) at high concentrations. Also shown is the nonspecific latex adsorption of non-activated (soluble) fluor-mPEG. A) Non specific soluble fluormPEG adsorption was negligible on 1.2 #m latex. B) The absorption of non-activated fluormPEG was more pronounced with the 8.0 #m beads and the raw data was appropriately corrected for this non-specific adsorption.  93  3.1.2  QUANTIFICATION OF LATEX-GRAFTED MPEG BY ANTI-PEG ANTIBODY After demonstrating that activated mPEG species were capable of covalent binding to  latex particles, an attempt at quantifying the amount of surface-grafted mPEG was made. Differentially modified 1.2 and 8.0 #m particles were reacted with a primary IgG1 antibody against PEG (1º; mouse-anti-PEG) followed by a reaction with the secondary fluorescent IgG antibody (2º; FITC-goat-anti-mouse) and analyzed via flow cytometry, Figure 3-2. Particles were PEGylated with 2, 5 and 20 kDa SVA-mPEG at 0-2 mM (1.2 #m latex) and 0-10 mM (8.0 #m latex) grafting concentrations. No dose-dependence of anti-PEG antibody binding to increasingly PEGylated 1.2 #m particles was observed (2, 5 and 20 kDa, 0-2 mM, p"1.0), Figure 3-2A. Although, there was a slight dose-dependent increase in fluorescence with the 20 kDa polymer, the signal became saturated at 0.5 mM grafting concentration and was found to be statistically insignificant (p"1.0). No difference in the amount of fluorescently bound antibody was observed with respect to the size of the polymer (2 vs. 5 vs. 20 kDa, 0-2 mM; p"1.0). Neither primary nor secondary antibody exhibited non-specific binding to the bare surface of the latex or to the polymer chains. A clear dose-dependent increase in anti-PEG antibody binding was observed at grafting concentrations of "2 mM for the 20 kDa polymer grafted to the 8.0 #m latex particles (2-10 mM, p!0.004), Figure 3-2B. Although, there was no dose-dependent effect noted for the 2 and 5 kDa polymers, there was, however, a slight increase in the fluorescent signal at 1-2 mM (p!0.02) and 2-5 mM (p!0.02) grafting concentrations of these polymers, respectively. In addition, an increased fluorescence was observed with the 20 kDa mPEG in comparison to 2 or 5 kDa chains at the highest grafting concentration, i.e., 10 mM (p"0.001). For example, 10 mM 20 kDa polymer resulted in fluorescent antibody signal of 748.4±150.5 fluorescence units in comparison to 344.3±81.5 and 393.3±35.2 units for the 2 and 5 kDa mPEGs, respectively, at the same grafting concentration. This suggested that 20 kDa polymer could potentially bind more 94  antibody molecules due to the increased number of ethylene oxide units recognized by the IgG molecule. There was a small amount of fluorescence detected for the 0 mM condition indicating that a slight non-specific 1º and 2º antibody adsorption to the 8.0 #m bare latex surface occurred. Overall, a commercially available monoclonal IgG1 anti-PEG antibody demonstrated poor mPEG dose- and molecular weight dependent binding profiles. The use of this antibody did not allow for quantitative measurement of the amount of mPEG grafted to the 1.2 or the 8.0 #m particles for polymer sizes of less than 20 kDa. However, the 20 kDa polymer demonstrated some dose-dependent effects with the 8.0 #m beads. These results, in general, suggested that the quantification of bound mPEG is not efficiently performed by the use of an anti-mPEG antibody.  Figure 3-2: Commercial anti-PEG antibody demonstrated a poor mPEG dose-dependence profile for both 1.2 (A) and 8.0 !m (B) particles. A) No statistically significant difference was observed in the amount of the bound antibody to increasingly modified 1.2 #m particles. B) An mPEG dose-dependent increase of antibody binding was noted only for the 20 kDa polymer on 8.0 #m particles. The following controls were utilized in this study: soluble mPEG control – latex PEGylated with unactivated soluble 2 mM, 5 kDa mPEG followed by a reaction with 1º and 2º antibodies; grafted mPEG control – latex modified with 2 mM, 20 kDa (1.2 #m latex) or 10 mM 20 kDa (8.0 #m latex) SVA-mPEG, reacted with 2º antibody. All control samples indicated absence of non-specific binding to PEG polymer and only non-specific adsorption to bare 8.0 #m latex surface was observed.  95  3.1.3  MICROSCOPIC OBSERVATION OF LATEX AGGREGATION Following each PEGylation reaction, latex samples were microscopically evaluated for  the evidence of latex particle aggregation, Figure 3-3. As all PEGylation reactions and subsequent experiments took place in solution (e.g., water, buffer, plasma), monodispersity of particles ensured their uniform exposure to the solute molecules (e.g., polymers, antibodies and plasma proteins). To this end, 1.2 #m particles were modified with increasing concentrations (02 mM) of monofunctional 2, 5 and 20 kDa SC- and SVA-mPEG polymers or multifunctional 3, 8 and 25 kDa NHS-polyglycerol (NHS-PG) polymers. Particle aggregation was assessed visually via an inverted light microscopy. Representative controls and 0.1 mM modified latex samples are shown in Figure 3-3. Monofunctional (one linker group per polymer) 2 kDa SC- and SVA-mPEG polymers demonstrated no inter-latex crosslinking events and behaved analogously to the bare latex control at all grafting conditions. This suggested that the grafted polymer chains on neighboring particles did not preferentially interact with each other and did not become entangled. In contrast, a multifunctional polyglycerol polymer (3 kDa) resulted in a high degree of latex aggregation even at very low (e.g., 0.1 mM) grafting concentrations, Figure 3-3D. Due to possessing multiple linker groups on the same polymer molecule, one chain of PG could become covalently attached to more than one latex particle, thus aggregating them. These observations held true for other molecular weights and grafting concentrations of both mono and multifunctional polymers. This experiment demonstrated that the use of monofunctional activated mPEG polymers was advantageous for preventing adverse latex aggregation events.  96  Figure 3-3: PEGylation of 1.2 !m latex with either SC- or SVAmPEG monofunctional polymer did not result in latex aggregation. Shown are transmitted light microscope images of latex suspensions: A) unmodified bare latex control; B) 2 kDa, 0.1 mM SC-mPEG modified latex; and C) 2 kDa, 0.1 mM SVA-mPEG modified latex. Shown in D) is a 3 kDa, 0.1 mM multifunctional (several reactive groups) polyglycerol polymer demonstrating enhanced aggregation due to cross binding to multiple latex particles.  3.1.4  MPEG-MEDIATED SURFACE CHARGE CAMOUFLAGE OF LATEX  PARTICLES 3.1.4.1 PARTICLE ELECTROPHORETIC MOBILITY: EFFECTS OF LINKER CHEMISTRY, GRAFTING CONCENTRATION, POLYMER AND LATEX SIZE To assess the effects of the linker chemistry (SC vs. SVA), polymer size (2, 5 and 20 kDa) and target size (1.2 and 8.0 #m particles), polystyrene beads were PEGylated at 0-20 mM grafting concentrations of activated mPEG and subjected to particle electrophoresis, Figure 3-4. The mobility of the highly charged bare 1.2 #m latex particles was found to be +2.2±0.2 (#m/sec)/(V/cm) with 3.5*107 charges/particle. For analysis purposes, the mobility of unmodified bare latex was normalized to 100%. The efficacy of charge camouflage on 1.2 #m latex was dose-dependent for both the SC and SVA linker chemistries. The most pronounced decrease in electrophoretic mobility was observed at grafting concentrations !0.5 mM, Figure 34A. For example, the mobility of 1.2 #m particles modified with 5 kDa SC-mPEG decreased from 93.1±19.8% to 58.7±12.4% when modified with 0.1 and 0.5 mM mPEG, respectively 97  (p!0.006). Similarly, the dramatic decrease in mobility from 46.9±3.2% to 14.2±3.8% was observed for the 5 kDa SVA-mPEG at 0.1 and 0.5 mM grafting concentrations, respectively (p!0.001). Latex PEGylation at higher grafting concentrations ("0.5 mM) yielded minimal decreases in mobility, suggesting that near maximal surface charge camouflage was rapidly achieved. This was consistent with the observations for covalent modification of the 1.2 #m beads with fluor-SVA-mPEG, Figure 3-1A. The most efficient polymer grafting (S1) occurred at !0.5 mM grafting concentrations. Significant differences were noted between linker chemistries with respect to the efficacy of surface charge camouflage on 1.2 #m latex. The SVA-mPEG was more efficient than SCmPEG at all grafting concentrations and polymer lengths (p<0.002), Figure 3-4A. This was, most likely, the result of the extended linker hydrolysis half-life of the SVA-mPEG (t1/2=30.6 min) in comparison to the SC-mPEG (t1/2=20.4 min), section 2.1.2. It was also demonstrated that the 20 kDa polymer was more effective at masking surface charge than either the 2 or the 5 kDa mPEGs at all grafting concentrations. For example, at a polymer grafting concentration of 0.5 mM, 5 kDa and 20 kDa SC-mPEG resulted in 58.7±12.4% and 26.5±4.6% mobility, respectively, relative to unmodified beads (p!0.009). A similar trend was observed with the SVA-mPEG linker. Specifically, at 0.2 mM grafting concentration relative mobility was decreased to 26.5±5.5% with 5 kDa polymers while the 20 kDa polymer further decreased the mobility to 2.5±3.4% that of the latex control (5 vs. 20 kDa, 0.1-2mM; p!0.001). Importantly, complete charge camouflage (0% mobility) of the 1.2 #m beads was achieved with the 20 kDa SVA-mPEG at >0.2 mM grafting concentration. An apparent reversal in mobility (i.e., charge reversal) of the 1.2 #m particles was also observed and likely arose due to the uncharged particles being carried in the opposite direction by the reverse osmotic flow of the solvent inside the electrophoresis chamber.  98  The mobility of the bare 8.0 #m latex particles was found to be +2.1±0.2 (#m/sec)/(V/cm) with 7.6*109 charges/particle and this was normalized to be equal to 100%. As with the 1.2 #m latex particles, the 8.0 #m beads exhibited a dose-dependent decrease in particle mobility for all molecular weight polymers (p!0.04), though the decrease in the mobility was much less than that observed for the 1.2 #m particles, Figure 3-4B. This was due to the higher surface charge density of the 8.0 #m beads vs. the 1.2 #m latex: 3.9*1015 vs. 7.9*1014 charges/cm2, respectively. The rate of the decrease in mobility was most pronounced at concentrations !10 mM. This was consistent with the fluor-SVA-mPEG grafting to 8.0 #m particles (Figure 3-1B), where change in the slope from fast (S1) to slow (S2) grafting occurred at 10 mM grafting concentration. PEGylation of 8.0 #m beads was also polymer size dependent, with longer polymers providing increased efficacy of surface charge camouflage (2 kDa vs. 20 kDa, 1-20 mM; p!0.001). For example, PEGylation of latex at 10 mM grafting concentration with the 2 and the 20 kDa SVA-mPEG resulted in 70.1±3.4% and 36.5±8.3% mobility, respectively. However, in contrast to the smaller particles, polymer grafting to the 8.0 #m beads did not result in a complete surface charge camouflage (i.e., zero mobility). The minimum 30.0±2.5% mobility was observed for the 8.0 #m particles subsequent to modification with 20 mM, 20 kDa SVA-mPEG. Importantly, both sizes of latex particles (1.2 and 8.0 #m) incubated with soluble, nonactivated mPEG incapable of covalent binding to surfaces did not exhibit any decrease in electrophoretic mobility. Thus, the covalent mPEG modification of 1.2 and 8.0 #m particles resulted in a dose-dependent surface charge camouflage as indicated by dramatically decreased electrophoretic mobility of the modified beads. SVA-mPEG resulted in improved camouflage due to its extended hydrolysis half-life. However, regardless of the linker chemistry and the size of the latex, longer polymer chains (20 kDa) were more effective at surface charge camouflage at all grafting concentrations. 99  Figure 3-4: PEGylation of 1.2 !m (A) and 8.0 !m (B) latex particles very efficiently camouflaged latex surface charge as evidenced by decreased electrophoretic mobility. A) The decrease in mobility was more pronounced with the SVA-mPEG in comparison to SCmPEG at all concentrations and molecular weights. (A) and (B) Charge camouflage was a function of both grafting concentration and polymer size with long chain mPEG exhibiting superior effects on both sizes of particles. Soluble mPEG, incapable of covalent grafting, did not result in any surface charge camouflage. Particle mobility was measured as described in the Materials and Methods. The mobilities of unmodified 1.2 and 8.0 µm latex beads were +2.2±0.2 and +2.1±0.2 (#m/sec)/(V/cm) (µm/sec)/(V/cm), respectively. The number of charges per 1.2 and 8.0 #m latex particles were 3.5*107 and 7.6*109, respectively and correlated with the efficacy of charge camouflage by the grafted polymer. Percent change in mobility was normalized to the mobility of unmodified latex.  3.1.4.2 PARTICLE ELECTROPHORETIC MOBILITY: EFFECT OF POLYMER ARCHITECTURE The effects of polymer architecture on the efficacy of surface charge camouflage were studied by subjecting differentially modified 1.2 #m beads to particle electrophoretic analysis. Latex was modified with either linear monofunctional SVA-mPEG (5 and 20 kDa) or highly branched multi-functional NHS-polyglycerol (3, 8, 25 kDa) at 0-2 mM grafting concentrations.  100  Both polymers possessed the same linker chemistry, therefore comparisons were based purely on the architecture of the polymer species. The SVA-mPEG is a linear, highly flexible molecule occupying a large 3D space on the surface of the latex, Figure 3-5A, Insert (i). In contrast, NHS-PG is densely branched and much less flexible, thereby occupying a limited 3D space, Figure 3-5B, Insert (ii). As shown in Figure 3-5 the linear mPEG polymer demonstrated superior surface charge camouflage properties in comparison to the branched PG molecule. As discussed in section 3.1.4.1, SVAmPEG demonstrated a dose- and polymer size dependent decrease in particle mobility, with the 20 kDa polymer resulting in a complete surface charge camouflage, Figure 3-5A. In contrast, NHS-PG did not demonstrate a polymer dose-dependent decrease in particle mobility (Figure 3-5B), likely, due to the extensive particle aggregation noted in the latex microscopy study (section 3.1.3). Optimal surface charge camouflage was achieved at 0.1 mM concentration and did not improve with increasing polymer grafting (p"1.0). Similar to mPEG, PG demonstrated polymer size dependence with the 8 and 25 kDa polymers exhibiting improved camouflage in comparison to the 3 kDa PG. For example, modification of particles with 2 mM, 3 and 25 kDa NHS-PG resulted in 58.0±5.4% and 20.0±1.8% mobility, respectively (p!0.001). Surprisingly, however, an increase in molecular weight of PG from 8 to 25 kDa did not generate improved surface charge camouflage. This was, most likely, due to the highly branched nature of the PG polymer, such that the increase in the molecular weight of PEG did not result in increased size of the polymer (RF) but, instead, led to an increased degree of branching. Relative to the linear mPEG, PG was less effective at surface charge camouflage at all molecular weights and grafting concentrations (p!0.001; except at 0.1 mM, 8 kDa PG), Figure 3-5B. For example, 0.5 mM 20 kDa mPEG modified beads exhibited 0.0±3.8% mobility in comparison to 50.7±6.9% for particles modified with the 25 kDa PG at the same grafting concentration. Importantly, incubation of latex particles with either the mPEG or PG incapable 101  of covalent grafting to latex did not result in any decrease in particle mobility. Thus, the linear and flexible mPEG resulted in superior surface charge camouflage in comparison to the highly branched, more rigid PG polymer. Grafted linear mPEG possesses increased camouflaging volume in comparison to the PG polymer due to its large 3D size that is directly proportional to its molecular weight (section 1.5.2).  Figure 3-5: Latex modification with linear monofunctional SVA-mPEG (A) resulted in enhanced surface charge camouflage in comparison to branched multifunctional NHS-PG (B). A) SVA-mPEG demonstrated a dose- and molecular weight dependent decrease in particle mobility, with longer polymers resulting in enhanced surface charge camouflage. B) NHS-PG demonstrated no dose dependent decrease in particle mobility. Improved surface charge camouflage was noted between 3 and 8 kDa PG, but not between 8 and 25 kDa polymer. Insert (i): Monofunctional (one green circle), linear, highly flexible mPEG occupies a large volume, therefore possesses large size, proportional to the molecular weight of the chain. Insert (ii): Multifunctional (several green circles), branched, less flexible and highly dense PG occupies a small 3D space. Molecular weight of this polymer is proportional to the density and the degree of branching, not the size of the molecule. The gradient in both Inserts (i) and (ii) depicts the probability of finding the polymer at a particular point of 3D space occupied by the polymer chain.  102  3.1.5  MPEG-MEDIATED INHIBITION OF SURFACE-MACROMOLECULE  INTERACTIONS 3.1.5.1 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: QUANTITATIVE COLORIMETRIC PROTEIN ASSAY In order to investigate whether PEGylation resulted in inhibition of surface macromolecule interactions, unlabeled human plasma protein adsorption experiments were performed on the 1.2 #m latex particles modified with 2, 5 and 20 kDa SC-mPEG (0-2 mM). The amount of the total protein adsorbed to differentially PEGylated beads was measured quantitatively by subjecting supernatants containing latex desorbed proteins to a reducing agent/detergent-compatible (RC DC) colorimetric protein assay. In this experiment, the decrease in the amount of adsorbed protein was indicative of the decrease in the surface-protein interactions. PEGylation of latex resulted in an mPEG dose- and molecular weight dependent decrease in surface-macromolecule interactions. As shown in Figure 3-6, SC-mPEG modified latex (2 mM, 2 kDa SC-mPEG) exhibited significantly less protein adsorption in comparison to the bare latex control: 15.2±2.0 vs. 159.9±6.4 ng/cm2, respectively (p!0.001). Surprisingly, the dramatic reduction in the amount of adsorbed protein occurred at grafting concentrations of !0.5 mM. For example, PEGylation with the 2 kDa SC-mPEG at 0.1 and 0.5 mM grafting concentrations resulted in 93.5±3.0 and 24.3±3.6 ng adsorbed protein per cm2 of latex, respectively (p!0.001). The protein adsorption profile was consistent with that seen for the surface charge camouflage of latex particles (Figure 3-4) in that the most dramatic decrease in both protein adsorption and electrophoretic mobility occurred at low grafting concentrations. Further increase (>0.5 mM) in grafting concentration resulted in a marginal decrease in protein adsorption, especially with the 2 kDa polymer (0.5-2 mM, p"0.7).  103  Protein adsorption was also observed to correlate strongly with the size of the polymer. Surprisingly and in contrast to surface charge camouflage (Figure 3-4), the maximum inhibition of surface-macromolecule interactions was achieved with the shorter polymer chains (e.g., 2 kDa vs. 20 kDa; 0.1-2.0 mM, p!0.04). Specifically, 0.5 mM, 2 and 20 kDa mPEG resulted in 24.3±3.6 ng/cm2 and 70.4±9.6 ng/cm2 adsorbed protein, respectively, in comparison to 159.9±6.4 ng/cm2 for the unmodified control. Finally, only surface grafted mPEG was capable of preventing protein adsorption. Latex particles incubated with soluble, non-activated mPEG showed a similar degree of protein adsorption as the bare latex control. In conclusion, PEGylation of latex particles efficiently prevented surface-macromolecule interactions as shown by dramatic decrease in the amount of adsorbed protein. Shorter polymer chains offered improved surface protection from protein adsorption in comparison to the long chain polymers on 1.2 #m particles.  Figure 3-6: PEGylation of 1.2 !m particles with SC-mPEG resulted in inhibition of surface-macromolecule interactions as indicated by decreased unlabeled plasma protein adsorption. Proteins desorbed from the latex surface were subjected to quantitative colorimetric protein assay. Latex PEGylation resulted in a doseand polymer length dependent decrease in the amount of surface-adsorbed protein (ng/cm2 of latex). Shorter (2 kDa) polymers were more effective at inhibiting protein adsorption in comparison to the high molecular weight polymer chains. Bare latex control adsorbed 159.9±6.4 ng of protein per cm2 of latex.  104  3.1.5.2 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: SDS-PAGE ANALYSIS In order to further investigate the PEGylation-mediated decrease of protein adsorption, latex desorbed protein samples were subjected to SDS-PAGE analysis. Unlabeled human plasma proteins were desorbed from 500 cm2 of 1.2 #m latex particles modified with 5 kDa SCmPEG at 0-2 mM grafting concentrations. Equal volumes of protein supernatants were loaded onto a 7% reducing gel and subjected to gel electrophoresis. Each lane of the gel represented proteins desorbed from one PEGylation condition. Consistent with the protein assay data, a dose-dependent decrease in the amount of plasma proteins adsorbed to PEGylated latex was observed, Figure 3-7. This was demonstrated by a reduced protein band intensity on the gel and decreased optical density (OD) values. Figure 3-7B represents the total OD measurements for each lane of the gel. OD values were expressed as a percent of total band intensity relative to the bare latex control lane (0 mM), which was set to equal 100%. Surprisingly, not all the proteins decreased equally with the increase in mPEG concentration, as would have been expected based on the decrease in the total protein adsorption. This was indicated by the non-uniform decrease in intensity of protein bands across increasingly PEGylated samples, Figure 3-7A. This also suggested that PEGylated beads demonstrated differential composition of the adsorbed protein layer in comparison to the bare latex control. In agreement with our previous findings, incubation of latex with unactivated soluble 5 kDa mPEG (Sol., 2.0 mM) did not prevent protein adsorption.  105  Figure 3-7: PEGylation, in addition to dramatically decreasing the total protein adsorption, resulted in the differential adsorption of specific plasma proteins. A) SDS-PAGE analysis of proteins desorbed from differentially modified latex particles (0-2 mM 5 kDa SC-mPEG) revealed a non-uniform decrease in protein band intensity. M.W. – Kaleidoscope multi-color molecular weight markers; P – 100 times diluted neat human plasma; Sol., 2.0 mM – proteins taken off the latex incubated with the soluble 2.0 mM, 5 kDa mPEG incapable of covalent binding. B) A graphical representation of the total optical density measurements for each lane in the SDS-PAGE gel. Optical density values were expressed as a percent of OD in the bare latex control lane (0 mM; 100%) using a Gene Tools software as described in Materials and Methods.  3.1.5.3 UNLABELED HUMAN PLASMA PROTEIN ADSORPTION: ITRAQ AND MASS SPECTROMETRY ANALYSIS While a substantial decrease of protein adsorption to PEGylated surfaces was demonstrated by both the colorimetric protein assay and the SDS-PAGE studies (Figures 3-6 and 3-7, respectively), the small amount of protein that was successfully adsorbed to the mPEGmodified particles was of interest to us as it could mediate unwanted immunological consequences. Therefore, a detailed analysis of the composition of modified vs. bare latex adsorbed protein layer was investigated using iTRAQ/MS analysis, Figure 3-8. The iTRAQ/MS protocol requires that all the protein samples subjected to the labeling contain the same amount of total protein (100 #g) in order to normalize the samples and to avoid 106  discrepancy during the mass spectrometry analysis. Since the total protein adsorption was dramatically decreased on the PEGylated latex surface, it required 3, 9 and 15 times more 0.5 mM, 2.0 mM and 5.0 mM modified latex particles (1.2 #m; 5 kDa SC-mPEG) in order to achieve the same amount of total desorbed protein as with the unmodified latex control. Thus, iTRAQ/MS analysis was different from the above described protein adsorption studies in that it investigated the composition of the protein layer and the relative abundances of specific protein species in the total protein mixture desorbed from the differentially modified beads, Figure 3-8. The values reported in this study were normalized to 100 #g total protein and were expressed relative to the unmodified latex control (100%).  Figure 3-8: iTRAQ/MS analysis utilized 100 !g of total protein and required substantially more PEGylated latex beads to yield the same amount of desorbed protein as that obtained from the bare latex control (1x). A) Total protein adsorption to PEGylated particles (2.0 mM, 5 kDa SC-mPEG) was dramatically decreased in comparison to unmodified latex control (0 mM), as demonstrated by SDS-PAGE analysis (also refer to Figure 3-7). B) Therefore, it required much more modified latex surface to obtain the same amount of surface desorbed protein as that obtained from the bare latex control (1x). Hypothetically, 3 bare latex beads would yield 100 µg total desorbed protein, whereas to obtain the same amount of protein from the 2.0 mM modified latex it would require 45 beads. M.W. – molecular size markers. SDS-PAGE and iTRAQ/MS analysis were performed as described in Materials and Methods. Comparative mass spectrometry analysis identified 95 protein species adsorbed to all latex samples. Because human plasma preparations contained low amounts of proteins derived 107  from tissues and residual white cells, platelets and RBCs, the true plasma protein subset was further delineated and consisted of 47 species. Composition of the total desorbed protein mixture varied across bare and PEGylated samples. Specifically, of the 47 true plasma proteins, 32 demonstrated a decrease and 5 exhibited an increase in relative protein abundance, while 10 remained relatively unchanged on the PEGylated surface in comparison to the bare latex control, Table 3-1. The plasma specific protein subset was further narrowed down to 32 species involved in immune regulation e.g., the complement and coagulation pathways, Table 3-1. These species participate in the primary events of discrimination between “self” and “non-self” and activate the immune system. Decreased relative abundance of these pro-activating proteins on foreign substrates is of primary importance and renders such surfaces “non-fouling” or “stealth”. It was concluded that only 6 immune proteins exhibited increased relative abundance on PEGylated latex in comparison to 20 species demonstrating reduced abundance. Six immunerelated proteins showed the same relative abundance on bare and modified latex surface. Table 3-2 lists the plasma proteins that demonstrated decreased relative abundance on mPEG modified surfaces. A large portion of these proteins (20/32 proteins) consisted of immune-relevant species. For example prothrombin, fibrinogen, factor V and plasminogen activator inhibitor, are important mediators of the coagulation pathway. C4b, LPS binding protein, C1s, C1r, C3, C9 and MASP-3 are potent activators of complement. Decreased abundance on PEGylated surfaces was also observed for proteins that serve inhibitory functions in the immune activation, e.g., plasminogen, C4b binding protein, vitamin K-dependent protein S and heparin cofactor II. An increased abundance on modified latex was noted for 5 plasma proteins, Table 3-3. Four of these species are involved in immune regulation: C1q (alpha and beta chains), a fragment of a constant chain of IgG immunoglobulin, thrombospondin 1 and beta-2-glycoprotein. Finally, PEGylated vs. bare surface desorbed protein solutions demonstrated the same composition with 108  respect to 10 plasma proteins, Table 3-4. Most of these species serve homeostatic functions, such as binding and transport, and include serum albumin, multimerin 1 and apolipoproteins AV, C-II, C-III. Only 4 proteins in this subset are involved in immune regulation: C1q (C-chain), Von Willebrand factor, platelet basic protein and platelet factor 4. Thus, iTRAQ/MS analysis demonstrated the differential composition of the adsorbed protein layer on PEGylated vs. bare latex surfaces. The enhanced depletion of the immunerelevant proteins was noted on modified latex particles and suggested that these surfaces became more biocompatible. In contrast, an increased relative abundance was observed for proteins making up the extracellular matrix and collagen-like species. Finally, several proteins demonstrated the same relative abundance in all latex desorbed protein samples. These species belonged to a class of proteins involved in transport, binding and maintenance of the homeostasis.  Table 3-1: Summary of iTRAQ/MS analysis listing the numbers and the types of proteins demonstrating decreased, increased or unchanged relative abundances in the protein layer adsorbed to PEGylated surface in comparison to the bare latex control.  Identified Decreased abundance on PEGylated surface Increased abundance on PEGylated surface Unchanged abundance on PEGylated surface  Total  Plasma proteins  Immune proteins  95 43 15 37  47 32 5 10  32 20 6 6  109  Table 3-2: Proteins exhibiting decreased relative abundance on PEGylated latex as identified by iTRAQ/MS analysis. Accession  pI  Protein Name  IPI:IPI00022229.1 IPI:IPI00019568.1 IPI:IPI00021885.1  MW kDa 512.8 65.3 75.5  6.6 5.2 5.1  IPI:IPI00453459.1  192.8  6.9  IPI:IPI00298497.3 IPI:IPI00019580.1 IPI:IPI00219713.1  50.8 88.4 48.5  8.0 7.1 5.2  IPI:IPI00021841.1 IPI:IPI00021727.1  28.1 61.7  5.3 6.2  IPI:IPI00298971.1  52.3  5.5  Apolipoprotein B-100 precursor Prothrombin precursor Fibrinogen alpha/alpha-E chain precursor Complement Component 4B preprotein Fibrinogen beta chain precursor Plasminogen precursor Fibrinogen gamma chain precursor Apolipoprotein A-I precursor C4b-binding protein alpha chain precursor Vitronectin precursor  IPI:IPI00218732.2 IPI:IPI00400826.1 IPI:IPI00294004.1  39.6 50.1 70.6  5.1 5.9 5.2  IPI:IPI00478809.3 IPI:IPI00021842.1 IPI:IPI00305461.2  248.7 34.2 72.4  5.7 5.5 5.8  IPI:IPI00032311.3  50.9  6.3  IPI:IPI00292530.1  71.4  6.3  IPI:IPI00025862.1  26.4  5.0  IPI:IPI00017696.1  74.9  4.9  IPI:IPI00164623.3 IPI:IPI00292950.3 IPI:IPI00022395.1  185.0 55.0 61.0  6.0 6.3 5.4  IPI:IPI00304273.1 IPI:IPI00290283.5 IPI:IPI00218192.1  43.4 81.9 101.2  5.2 5.0 6.2  IPI:IPI00007118.1  42.8  7.0  IPI:IPI00028413.3  69.4  5.0  IPI:IPI00514475.3 IPI:IPI00296165.5  41.1 78.3  5.7 5.8  IPI:IPI00554598.1  23.8  4.8  IPI:IPI00216065.1  40.4  5.3  % Decreased Adsorption 0.5mM 2mM 5mM 31.3 82.2 53.7 20.2 55.4 81.9 56.4  77.4  77.7  83.4  94.6  94.5  51.3 34.7  74.9 74.4  75.3 87.0  53.7  75.5  75.5  19.4  51.2  63.6  71.4  93.3  93.6  3.6  43.4  71.3  Serum paraoxonase/arylesterase 1 Clusterin Vitamin K-dependent protein S precursor Coagulation factor V precursor Apolipoprotein E precursor Inter-alpha-trypsin inhibitor H2 precursor Lipopolysaccharide-binding protein precursor Inter-alpha-trypsin inhibitor H1 precursor C4b-binding protein beta chain precursor Complement C1s subcomponent precursor Complement C3 precursor Heparin cofactor II precursor Complement component C9 precursor Apolipoprotein A-IV precursor Complement factor MASP-3 Inter-alpha-trypsin inhibitor H4 precursor Plasminogen activator inhibitor-1 precursor Inter-alpha inhibitor H3  0.0 0.0  24.8 21.2  54.1 58.6  0.0  93.6  95.5  14.0 23.7  42.8 43.6  33.4 57.3  73.4  88.1  79.6  70.9  85.1  54.3  78.2  94.2  84.2  80.4  96.7  97.6  0.0  31.3  57.9  58.1 82.2  75.5 94.6  75.9 82.3  71.0  89.4  78.5  29.8 0.0  51.2 59.5  34.4 86.7  50.1  64.1  57.1  13.5  23.1  51.5  0.0  81.9  78.1  OTTHUMP00000028705 Complement C1r subcomponent precursor Hypothetical protein DKFZp686N1868 Vitamin K-dependent protein Z precursor  41.5  62.3  70.9  69.9  90.7  86.8  0.0  42.1  77.9  0.0  36.0  58.0  Class Transport Coagulation Coagulation Complement Coagulation Coagulation Coagulation Transport Complement Extracellular matrix Enzyme Complement Coagulation Coagulation Transport Transport/ binding Immunity Transport/ binding Complement Complement Complement Coagulation Complement Transport Complement Transport Coagulation Transport/ binding Transport Complement Binding Coagulation  Values were expressed as a percent decrease in the relative abundance of proteins in the 100 #g of total protein desorbed from PEGylated vs. bare latex control. Percent decreased abundance labels: (__) >75%; (__) 51-75%; (__) 21-50%; (__) !20%. Bold are the immune-relevant proteins. IPI-international protein index, MW-molecular weight, kDa-kilodaltons, pI-isoelectric point.  110  Table 3-3: Proteins exhibiting increased relative abundance on PEGylated latex as identified by iTRAQ/MS analysis. Accession  pI  Protein Name  IPI:IPI00296099.3  MW kDa 127.5  4.7  Thrombospondin-1 precursor  IPI:IPI00549769.1 IPI:IPI00022392.1  36.1 23.7  8.5 9.3  IPI:IPI00477992.1  23.7  8.9  IPI:IPI00298828.1  36.3  8.4  Ig gamma-1 chain C region Complement C1q subcomponent, A chain precursor Complement component 1, q subcomponent, beta polypeptide Beta-2-glycoprotein I precursor  % Increased adsorption 0.5mM 2mM 5mM 5.2 18.3 26.3  Class  20.0 0.0  0.0 20.2  89.9 34.4  Extracellular matrix Immune Complement  0.0  28.1  60.4  Complement  94.9  55.2  64.7  Coagulation  Values were expressed as a percent increase in the relative abundance of proteins in the 100 #g of total protein desorbed from PEGylated vs. bare latex control. Percent increased abundance labels: (__) >75%; (__) 51-75%; (__) 21-50%; (__) !20%. Bold are the immune-relevant proteins. IPI-international protein index, MW-molecular weight, kDa-kilodaltons, pI-isoelectric point.  Table 3-4: Proteins exhibiting unchanged relative abundance on PEGylated latex, as identified by iTRAQ/MS analysis. Accession  pI  Protein Name  Class  IPI:IPI00022434.1 IPI:IPI00303283.1 IPI:IPI00418392.1 IPI:IPI00022394.2  MW kDa 66.5 84.4 136.1 22.8  5.7 5.0 7.9 8.3  Transport Receptor Transport Complement  IPI:IPI00023014.1 IPI:IPI00465378.1 IPI:IPI00021857.1 IPI:IPI00022445.1 IPI:IPI00022446.1 IPI:IPI00021856.3  225.7 38.9 8.8 10.3 7.8 8.9  5.4 6.0 4.7 9.0 8.8 4.7  Serum albumin precursor Integrin beta-3 precursor Multimerin 1 Complement C1q subcomponent, C chain precursor Von Willebrand factor precursor Apolipoprotein A-V precursor Apolipoprotein C-III precursor Platelet basic protein precursor Platelet factor 4 precursor Apolipoprotein C-II precursor  Coagulation Transport Transport Immune Coagulation Transport  Although, there was no difference in the relative abundance of the above-listed species in 100 µg total desorbed protein solution from either bare or modified latex, the total adsorption to PEGylated surfaces decreased dramatically in comparison to unmodified beads, e.g., 18.3±4.1 (2 mM, 5 kDa SC-mPEG) vs. 159.9±6.4 ng/cm2 (bare latex). IPI-international protein index, MW-molecular weight, kDa-kilodaltons, pI-isoelectric point.  111  3.1.5.4 FLUORESCENT HUMAN PLASMA PROTEIN ADSORPTION: EFFECTS OF LINKER CHEMISTRY, GRAFTING CONCENTRATION AND POLYMER SIZE Consequent to the unlabeled plasma protein adsorption experiments, the effects of polymer linker, size and grafting concentration on the inhibition of surface-macromolecule interactions were studied using Alexa 488 fluorescently labeled human plasma. The use of the fluorescently labeled plasma allowed for the direct assessment of protein binding to control and PEGylated beads. Latex particles of 1.2 #m in diameter were modified with either SC- or SVAmPEG (2, 5, 20 kDa; 0-2 mM) and subjected to fluorescently labeled human plasma protein adsorption followed by flow cytometry. The amount of surface-adsorbed protein was measured as the mean fluorescent intensity of latex particles expressed as a percent of non-modified latex control with adsorbed fluorescent protein (100%). Consistent with the results of unlabeled plasma protein adsorption (Figure 3-6) and SDSPAGE analysis (Figure 3-7) PEGylation of 1.2 #m latex resulted in a dose-dependent decrease in total fluorescent protein adsorption for both SC- and SVA-mPEG modified particles. The most pronounced decrease in protein adsorption occurred at grafting concentrations !1 mM for SC-mPEG and !0.5 mM for SVA-mPEG, Figure 3-9. The 2 kDa SC-mPEG resulted in 47.9±3.7% and 12.0±1.2% mean fluorescence for 0.1 and 0.5 mM grafting concentrations, respectively (2, 5 and 20 kDa, 0.1-0.5 mM; p!0.04). Similarly, 2 kDa SVA-mPEG exhibited 47.7±0.3% and 10.2±0.4% mean fluorescence for 0.1 and 0.5 mM grafting concentrations, respectively (2, 5 and 20 kDa, 0.1-0.5 mM; p!0.02). Latex PEGylation at higher grafting concentrations yielded a minimal decrease in protein adsorption, suggesting that near maximal surface camouflage was rapidly achieved. This observation was consistent with the findings for the covalent modification of 1.2 #m beads with fluor-SVA-mPEG, Figure 3-1A. The efficient polymer grafting to latex particles occurred at !0.5 mM concentration followed by slower polymer addition. 112  Similar to the surface charge camouflage, SVA-mPEG was more efficient at inhibiting surface-macromolecule interactions than SC-mPEG (5 kDa, 0.1-1 mM; 20 kDa, 0.1-2 mM; p!0.05) due its extended linker hydrolysis half-life, Figure 3-9. At 0.1 mM grafting concentration the 20 kDa SC-mPEG resulted in 83.7±1.0% relative protein adsorption in comparison to 52.3±3.5% with SVA-mPEG of the same molecular weight. However, the 2 kDa polymers of both linker chemistries were equally effective (0.1, 0.5, 2 mM, p"0.1). Specifically, 0.5 mM, 2 kDa polymers resulted in protein adsorption of 12.0±1.2% and 10.2±0.4% for SCmPEG and SVA-mPEG, respectively, relative to the bare latex control (100%). Surprisingly and in contrast to surface charge camouflage, short chain polymers were significantly more effective at preventing fluorescent plasma adsorption than long chain polymers (0.1-1 mM, 2 vs. 20 kDa; p < 0.001 for SC-mPEG). This was exemplified at a grafting concentration of 0.5 mM SC-mPEG with 2 and 20 kDa polymers resulting in 12.0±1.2% and 39.5±3.9% adsorption, respectively, relative to the bare latex control. The effects for the SVAmPEG were much less pronounced but still significant (2 vs. 20 kDa, 0.2-2 mM; p!0.05) demonstrating decreased mean fluorescence of particles from 21.1±2.9% to 10.2±0.4% when PEGylated with 0.5 mM, 20 kDa and 2 kDa polymers, respectively. However, for both linker chemistries and for all polymer sizes, a maximal decrease of protein adsorption was noted at 2 mM grafting concentration. Importantly, only covalently grafted mPEG inhibited protein adsorption. Latex particles incubated with soluble, non-activated mPEG showed a similar level of protein adsorption as that of the bare latex control. In conclusion, fluorescent protein adsorption paralleled the results of unlabeled protein adsorption and SDS-PAGE analysis demonstrated a significant decrease of the total amount of fluorescent protein adsorbed to PEGylated latex beads. The inhibition of surface-macromolecule interactions was dose-, linker- and polymer size-dependent with shorter polymer chains providing improved surface protection. 113  Figure 3-9: PEGylation of 1.2 !m particles with either SC- (A) or SVA-mPEG (B) resulted in inhibition of surface-macromolecule interactions as indicated by decreased fluorescent plasma protein adsorption. Both linker chemistries demonstrated a dose- and polymer length dependent decrease in total protein adsorption. SVA-mPEG was more effective at inhibiting surface-macromolecule interactions. For both linkers, shorter polymers demonstrated enhanced efficacy of surface camouflage. Mean particle fluorescence was expressed as a percent of fluorescence of bare latex with adsorbed fluorescent protein (100%).  3.1.5.5 FLUORESCENT PLASMA PROTEIN ADSORPTION: EFFECTS OF POLYMER ARCHITECTURE To further investigate the factors affecting inhibition of surface-macromolecule interactions, the effect of polymer architecture was studied. Latex particles (1.2 #m) were modified with linear SVA-mPEG (2, 5, 20 kDa; 0-2 mM) or branched NHS-PG (3, 8, 25 kDa; 02 mM) of the same linker reactivity, so that a direct comparison between the two polymers could be made. Alexa 488 fluorescently labeled human plasma was used for protein adsorption and particles were analyzed by flow cytometry. Mean particle fluorescence was expressed as a percent of non-modified latex control with adsorbed fluorescent protein (100%). Consistent with the results of the electrophoretic mobility study, branched PG resulted in decreased efficacy of preventing fluorescent protein adsorption in comparison to linear SVA114  mPEG, at 0.5-2 mM grafting concentrations and with all molecular weight polymers (p"0.04), Figure 3-10. In addition, no dose or polymer molecular weight effect was noted for PG, whereas for the SVA-mPEG short chain polymers demonstrated improved surface camouflage in comparison to long chain mPEG, Figure 3-10A. Surprisingly, the maximal inhibition of protein adsorption to PG-modified particles was achieved at the very lowest grafting concentration tested for all molecular weight polymers. Indeed, the 0.1 mM grafting concentration of PG resulted in superior inhibition of protein adsorption in comparison to the linear mPEG. This could be attributed to the decreased size of the dense, highly branched PG molecule in comparison to the large and flexible mPEG. Decreased exclusion volume of PG resulted in reduced steric repulsion by surface grafted polymer chains and thus, increased rate of surface grafting of small PG in comparison to the large mPEG molecule. However, surface camouflage imparted by PG did not improve with further latex modification and remained at the level achieved with the 0.1 mM grafting concentration (3, 8, 25 kDa, 0.1-2 mM; p"0.09), Figure 3-10B. Specifically, 3 kDa PG resulted in 38.5±13.0% and 34.6±4.9% relative protein adsorption at 0.1 and 2 mM grafting concentrations, respectively. The optimal inhibition of surface-macromolecule interactions was achieved with the 8 kDa, 2 mM PG and resulted in 15.3±0.8% protein adsorption relative to the bare latex control. This was significantly higher than 2.7±0.5% protein adsorption achieved with the similar molecular weight (5 kDa) 2 mM SVA-mPEG (p!0.001). Thus, multifunctional branched PG exhibited no dose or molecular weight dependence in preventing fluorescent protein adsorption to modified latex particles and resulted in diminished camouflage of surfaces in comparison to monofunctional linear SVA-mPEG of the same chemical reactivity.  115  Figure 3-10: PEGylation of 1.2 !m particles with the linear SVA-mPEG (A) resulted in improved inhibition of fluorescent protein adsorption in comparison to the branched NHSPG (B). A) SVA-mPEG resulted in a dose- and molecular weight-dependent decrease in the amount of fluorescent protein adsorbed to the surface of latex. Short chain polymers demonstrated improved surface camouflage in comparison to the long chain polymers. B) PG demonstrated no dose or molecular weight dependent inhibition in surface-macromolecule interactions. Only 8 kDa PG resulted in comparable surface camouflage to that of SVA-mPEG. Mean particle fluorescence was expressed as a percent of non-modified latex control with adsorbed fluorescent protein (100%).  3.1.5.6 FLUORESCENT PLASMA PROTEIN ADSORPTION: EFFECTS OF TARGET SIZE After delineating the chemical and physical properties of polymers affecting the efficacy of surface camouflage, the effects of target latex size were investigated. Inhibition of fluorescent plasma protein adsorption was studied on 1.2 and 8.0 #m differentially modified latex particles. The polystyrene beads were PEGylated with 0-2 mM (1.2 #m) or 0-20 mM (8.0 #m) 2, 5 and 20 kDa SVA-mPEG and subjected to fluorescent human plasma adsorption experiments, followed by either flow cytometry or quantitative fluorescence microscopy. As expected, PEGylation of both sizes of particles resulted in a dose-dependent decrease of the amount of fluorescent protein adsorbed, measured via flow cytometry (Figure 3-11) and quantitative fluorescent microscopy (Figure 3-12). As discussed in the preceding section (3.1.5.4), a dramatic decrease in protein adsorption was observed consequent to PEGylation of 116  1.2 #m particles with SVA-mPEG at !0.5 mM grafting concentration, e.g., 0.1 mM and 0.5 mM 2 kDa polymer resulted in 47.7±0.3% and 10.2±0.4% relative protein adsorption, respectively, as measured by flow cytometry (p!0.001) (Figure 3-11A). A similar trend was observed for the 1.2 #m particles in a confocal microscopy study: 61.6±11.2% and 22.2±2.5% relative adsorption was measured for 0.1 and 0.5 mM, 2 kDa SVA-mPEG, respectively (p!0.001), Figure 3-12B. PEGylation of 8.0 #m particles exhibited a more linear dose-dependency. Protein adsorption decreased gradually over the range of grafting concentrations used and was observed with both flow cytometry (Figure 3-11B) and inverted light fluorescent microscopy (Figure 312D). Flow cytometric measurements for the 20 kDa SVA-mPEG at 5 and 20 mM grafting concentrations were 75.6±17.1% and 18.7±2.5% adsorption relative to bare latex, respectively (p!0.001). Similarly, the results from the quantitative microscopy yielded 70.4±13.9% and 13.2±1.8% relative adsorption, for 5 and 20 mM, 20 kDa SVA-mPEG, respectively (p!0.001). Surprisingly, a strong correlation was observed between the target size and the optimal polymer size required for the inhibition of protein adsorption. Both flow cytometry (Figure 311) and quantitative microscopy (Figure 3-12) studies indicated that the 20 kDa polymer chains were more effective at preventing adsorption to the 8.0 #m particles while the short chain polymers (e.g., 2 kDa) provided better camouflage of the 1.2 #m latex beads. As shown in Figure 3-12A-B, protein adsorption to the 1.2 #m, particles was decreased to 20.2±3.2% of the latex control when modified with 1 mM, 2 kDa polymer. In contrast, the 20 kDa, polymer grafted at the same 1 mM concentration reduced the adsorption only to 40.4±4.8% of the unmodified control (p<0.001). However, the opposite effect was observed using the 8.0 #m particles (Figure 3-12C-D). The 2 kDa polymer at a grafting concentration of 10 mM, only reduced protein adsorption to 74.9±7.4% while the same grafting concentration of the 20 kDa polymer decreased protein adsorption to 35.1±6.9% of the unmodified latex (p<0.001). This target size dependence was specifically depicted in Figures 3-12A and C for the 1.2 and 8.0 #m 117  particles, respectively, utilizing side-by-side image comparisons of the amount of fluorescent protein adsorbed to particles modified with 2 vs. 20 kDa polymers. As before, incubation of latex with unactivated soluble mPEG did not result in reduced protein adsorption for either 1.2 or 8.0 #m particles (noted in Figures 3-11 and 3-12B-D). In conclusion, PEGylation of 1.2 and 8.0 #m particles with SVA-mPEG resulted in the substantially decreased surface-macromolecule interactions measured with both, flow cytometry and quantitative fluorescent microscopy. The inhibition of protein adsorption was best achieved at the high grafting concentration with short polymer chains (2 kDa) on small targets (1.2 #m latex) and with long polymer chains (20 kDa) on large targets (8.0 #m latex). Thus, target size dictated the choice for the optimal polymer length.  118  Figure 3-11: PEGylation of 1.2 (A) and 8.0 !m (B) particles with SVA-mPEG resulted in a dose-, as well as a target size-dependent decrease in fluorescent protein adsorption, measured by flow cytometry. A) Inhibition of protein adsorption on 1.2 #m particles was readily observed even at low, 0.2 mM grafting concentration. Short, 2 kDa, polymers were shown to provide improved surface camouflage in comparison to long chain polymers e.g., 20 kDa, at intermediate grafting concentrations. B) A gradual decrease in protein adsorption was observed consequent mPEG modification of 8.0 #m particles. Long polymers provided better surface camouflage in comparison to short chain polymers at intermediated grafting concentrations, e.g., 10 mM. Mean particle fluorescence was expressed as a percent of the non-modified latex control with adsorbed fluorescent protein, normalized to 100%.  119  Figure 3-12: Prevention of fluorescently labeled protein adsorption is readily observed via microscopic analysis of SVAmPEG modified 1.2 !m (A, B) and 8.0 !m (C, D) latex particles. A difference between the 1.2 and 8.0 !m beads was noted relative to the optimal polymer molecular weight. While small (2 kDa) polymers provided optimal prevention of protein adsorption for the 1.2 !m latex (A, B), the 20 kDa polymer provided optimal protection for the 8.0 !m beads (C, D) demonstrating that the target size may dictate the optimal polymer size. Images were obtained using confocal (1.2 !m latex) or inverted light microscopy (8.0 !m latex). Enlarged inserts are provided of the 1.2 !m beads at 1 mM grafting concentration of 2 and 20 kDa polymers. Relative intensity of particles was expressed as a percent of fluorescence of bare latex particles with adsorbed fluorescent protein, normalized to 100%.  120  3.1.5.7 FLUORESCENT PLASMA PROTEIN DESORPTION To evaluate the efficacy of the protein desorption process used in SDS-PAGE and iTRAQ/MS analyses and colorimetric protein assay, latex particles were subjected to fluorescent plasma adsorption followed by protein desorption and flow cytometry. This study was important to ensure that the protein samples obtained following the desorption process were representative of the actual latex adsorbed proteins. The amount of residual fluorescence was indicative of the amount of adsorbed fluorescent protein not removed by the desorption process. Fluorescent intensity values were expressed as a percent of the initial adsorption to unmodified latex control, set equal to 100%. Although SDS-PAGE, iTRAQ/MS and the protein assay utilized proteins desorbed from SC-mPEG modified 1.2 !m particles, this study was expanded to include all sizes of the beads and all polymer chemistries used in this thesis. Thus, 1.2 !m particles were modified with SC- and SVA-mPEG (2, 5, 20 kDa) and NHS-PG (3, 8, 25 kDa), and 8.0 !m particles were modified with SVA-mPEG (2, 5, 20 kDa). As demonstrated in Figure 3-13, the amount of left-over protein was significantly lower than the amount of protein initially adsorbed to latex particles, except at very high grafting concentrations, where protein adsorption was almost completely inhibited (e.g., 1.2 !m latex, 2 mM, 2 kDa polymer), Figure 3-13A-B. Specifically, the following fluorescence values were obtained for the 5 kDa SC-mPEG at 0.2 mM and 2 mM concentrations, before vs. after desorption: 26.8±1.7% vs. 4.5±1.3% and 3.3±0.3% vs. 1.6±0.3%, respectively (p"0.002). Similar results were obtained for 1.2 !m latex modified with SVA-mPEG and NHS-PG, Figures 3-13B and C, respectively. Protein desorption from SVA-mPEG modified 8.0 !m latex was highly efficient, except at intermediate concentrations of 5 kDa polymer, Figure 3-13D. The decrease in fluorescence for this molecular weight mPEG was still significant but not as dramatic as with other polymer sizes and grafting concentrations. For example, protein desorption from 10 mM, 5 and 20 kDa SVA-mPEG modified particles resulted in 25.1±15.4% and 2.3±2.3% 121  fluorescence in comparison to the original protein adsorption of 57.5±2.2% and 33.9±3.8%, respectively (p"0.02). Residual fluorescence was higher for the bare latex in comparison to the modified beads (Figure 3-13A-B). Fluorescent intensity of bare 1.2 !m particles following protein desorption was 9.7±1.5% in comparison to 1.8±0.9% for the 2 kDa, 2 mM SVA-mPEG modified latex (p"0.001). Furthermore, latex intensity values following protein desorption from either bare or modified particles never reached zero, suggesting that desorption was still incomplete. In summary, protein desorption was successful at removing #90% of the adsorbed protein, therefore proteins desorbed from SC-mPEG modified 1.2 !m latex that were used in SDS-PAGE, iTRAQ/MS analysis and the colorimetric protein assay were an accurate representation of the proteins adsorbed to these particles. A little amount of left-over fluorescence was, most likely, contributed by surface denatured, permanently adsorbed fluorescent proteins.  122  Figure 3-13: The amount of fluorescent protein left on the latex particles following protein desorption decreased dramatically in comparison to the amount of protein that was initially adsorbed. The fluorescent protein desorption from 1.2 !m particles modified with SC-mPEG (A), SVAmPEG (B) and NHS-PG (C) resulted in #90% removal of total adsorbed protein. D) Protein desorption from the 8.0 !m SVA-mPEG modified latex also resulted in #90% protein removal, except at the intermediated grafting concentrations of 5 kDa mPEG. The little amount of leftover protein, most likely, consisted of permanently adsorbed, denatured plasma proteins. Relative intensity of particles was expressed as a percent of fluorescence of bare latex with adsorbed fluorescent protein, normalized to 100%.  3.1.6  SUMMARY OF RESULTS FOR THE BIOPHYSICAL ALIPHATIC AMINE  POLYSTYRENE LATEX MODEL The use of a robust polystyrene latex model allowed for an in-depth biophysical study of the mechanisms governing polymer-mediated immunocamouflage of surfaces, not easily performed on labile cells. The study evaluated the effects of polymer chemical and physical properties as well as the target size on the efficacy of surface charge camouflage and inhibition of surface-macromolecule interactions. As demonstrated by our biophysical studies, all activated mPEG polymers were capable of covalent binding to the surface of latex in a dose-dependent manner. Unfortunately, due to the 123  poor quality of the commercially available anti-PEG antibody, a direct antibody-mediated quantitative measurement of covalently bound mPEG was not efficient for either 1.2 or 8.0 !m particles. PEGylation resulted in surface charge camouflage and prevention of surfacemacromolecule interactions. Both effects were dependent on the chemical and physical properties of mPEG as well as the target size. The SVA-mPEG polymer was more efficient in surface charge camouflage and prevention of plasma protein adsorption than SC-mPEG due to its extended hydrolysis half-life. Moreover, linear, flexible, monofunctional mPEG species demonstrated improved efficacy in comparison to the branched, dense, multifunctional PG polymers due to the absence of the latex aggregating property and the increased zone if immunoprotection. Regardless of the target size, long chain polymers resulted in improved surface charge camouflage in comparison to the short chain polymers. In contrast, target size dictated the optimal polymer length for the inhibition of surface-macromolecule interactions. Thus, prevention of plasma protein adsorption on small targets (1.2 !m) was best achieved with short chain polymers (2 kDa) and surface camouflage of large targets (8.0 !m) was best achieved with long chain polymers (20 kDa). Furthermore, detailed analysis of the surface adsorbed protein layer demonstrated that in addition to the dramatic decrease of the total protein adsorption to PEGylated surfaces, the immune-related subset of proteins was preferentially depleted from the modified particles relative to the bare latex control.  124  3.2  BIOLOGICAL RED BLOOD CELL AND PERIPHERAL BLOOD  MONONUCLEAR CELL MODEL 3.2.1  MPEG-MEDIATED SURFACE CHARGE CAMOUFLAGE OF  ERYTHROCYTES: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION Consequent to the observed decrease in electrophoretic mobility of PEGylated latex particles, the effects of polymer molecular weight and grafting concentration on surface charge camouflage of human red blood cells were studied. Human erythrocytes were PEGylated with 2, 5 and 20 kDa SPA-mPEG at 0-6.5 mM grafting concentrations and subjected to electrophoretic mobility studies. The electrophoretic mobility of the cells was expressed as a percent of the mobility of unmodified RBCs (100%). Unmodified erythrocytes are negatively charged and carry approximately 1.0*107 charges/cell (152). Non-derivatized RBCs traveled in the electric field with the mobility of – 1.2±0.1 (µm/sec)/(V/cm). As shown in Figure 3-14, PEGylation of the RBCs resulted in a dose dependent camouflage of surface charge as indicated by their decreased electrophoretic mobility. Importantly, the electrophoretic mobility data for the ~8.0 !m erythrocytes correlated well with the electrophoretic mobility measurements of the similarly sized latex particles (Figure 3-14 Insert and Figure 3-4). For both human red blood cells and latex particles, surface modification was most effective with longer polymer chains (20 kDa vs. 2 kDa; p<0.0001). In fact, grafting of larger polymers (20 kDa) to the red blood cell surface resulted in a near complete charge camouflage with ~ 0% mobility at grafting concentrations #1.6 mM SPA-mPEG. The observed difference in the extent of surface charge camouflage between 8.0 !m polystyrene latex particles and 8.0 !m red blood cells was possibly the effect of initial surface charge (beads are much more  125  highly charged) and localization of charges on the surface as will be discussed in more detail in the Discussion.  Figure 3-14: PEGylation resulted in the efficient surface charge camouflage of human erythrocytes as indicated by decreased electrophoretic mobility of modified RBCs. Covalent grafting of SPA-mPEG to human RBCs (~8.0 !m in diameter) resulted in a dose- and molecular weightdependent decrease of electrophoretic mobility. Long mPEG polymers exhibited increased efficacy of surface charge camouflage. These observations were consistent with the studies on the similarly sized 8.0 !m latex particles (Insert). The difference in the extent of surface charge camouflage between the RBCs and the latex particles was due to the initial surface charge density: 7.6*109 charges per particle vs. 1.0*107 charges per cell.  3.2.2  CAMOUFLAGE OF CD SURFACE ANTIGENS: EFFECTS OF POLYMER SIZE,  GRAFTING CONCENTRATION AND TOPOGRAPHY OF THE CELL SURFACE As a consequence of the successful biophysical camouflage of latex surfaces, we examined the effects of cellular PEGylation on biologically important receptor-ligand interactions. To this end, the mPEG-mediated inhibition of antigen-antibody binding was studied by reacting differentially modified human peripheral blood mononuclear cells with monoclonal fluorescent antibodies against CD3, CD4 and CD28 T cell surface antigens. Antigens of varying heights were selected in order to evaluate the effects of surface topography, Figure 3-15. Following antibody incubation, the cells were subjected to flow cytometry to quantify the amount of surface bound antibody based on either mean cell fluorescence (MCF) or 126  percent gated cell (PGC) analyses. T cells were PEGylated with 2 and 20 kDa SVA-mPEG at 04 mM grafting concentrations. Mean cell fluorescence was normalized to the value for unmodified cells incubated with fluorescent antibodies and set to equal 100%. PEGylation resulted in a dose-dependent decrease of antibody binding to all the tested antigens as demonstrated by both MCF (Figure 3-16) and PGC (Figure 3-17) analyses. However, substantial differences were noted between the rates of decrease in percent mean fluorescence and percent gated values. Mean cell fluorescence values exhibited a steady decline with a substantial reduction in fluorescence even at low (e.g., 0.5 mM) grafting concentrations, Figure 3-16. For example, PEGylation with 1 mM, 2 kDa SVA-mPEG resulted in 23.5±3.6%, 9.5±1.3% and 37.8±2.7% fluorescence for CD3, CD4 and CD28 antigens, respectively, in comparison to the unmodified control (100%; p"0.001). In contrast, the decrease in percent gated values (Figure 3-17) occurred more abruptly and at higher grafting concentrations, e.g., 2 mM. Hence, PEGylation of cells with 2 mM, 20 kDa SVA-mPEG resulted in percent gated values of 12.9±4.9%, 5.7±1.9% and 6.6±1.9% for CD3, CD4 and CD28 antigens respectively, in comparison to the unmodified control (100%; p"0.001). The apparent differences in the rates of decreased antibody binding were the result of the nature of the MCF vs. the PGC analyses. The MCF measures the amount of fluorescently bound antibody per cell, whereas the PGC represents the percent of total analyzed cells that exhibit any level of antibody binding. Furthermore PGC, but not the MCF analysis, indicated a polymer length-dependence in the prevention of the antibody binding. PGC analysis was consistent with the biophysical results for the 8.0 !m latex beads in that inhibition of antigen-antibody interactions on large PBMCs (10 !m diameter) was best achieved using long chain polymers (e.g., 20 kDa). For example, CD3 camouflage using a 2 mM grafting concentration resulted in 68.4±9.7 and 12.9±4.9% antibody positive cells with the 2 and the 20 kDa SVA-mPEG, respectively (2 vs. 20 kDa; 1-4 mM,  127  p"0.029). No mPEG molecular weight effect was observed with the MCF analysis and both the 2 and the 20 kDa polymers resulted in the efficient CD antigen camouflage. PGC analysis also demonstrated that immunocamouflage of cells was more complex than that of latex beads due to the inherent complexity (e.g., topography and protein distribution) of the cell surface. This was most readily exemplified by the height of the CD antigen to be camouflaged as well as the grafting site of the mPEG itself. This was modeled in Figure 3-15 and demonstrates that those CD markers (CD3 and CD28; 4 and 7 nm in height, respectively) close to the cell surface were readily camouflaged by short chain polymers while the largest CD marker (CD4; 13 nm in height) was most effectively camouflaged by long chain polymers (2 vs. 20 kDa at 2 mM grafting concentration; p"0.001). However, the efficacy of CD camouflage by polymers of increasing lengths was not only related to the height of the protein. For example, CD28 antigen (intermediate height, 7 nm) was equally well camouflaged by all molecular weight mPEGs (2 vs. 5 vs. 20 kDa; p#0.1), whereas a short CD3 antigen (3 nm in height) exhibited a polymer size dependence. Specifically, at intermediate concentrations, e.g., 1 and 2 mM, 20 kDa polymer resulted in improved CD3 camouflage in comparison to the 2 kDa SVA-mPEG (p"0.03). This could be due to the surface localization of this protein. As discussed in section 1.3.3, Figure 1-10, CD3 protein is a part of a multi-subunit T cell receptor (TCR) complex. CD3 is located very close (~12 Å) to the TCR that is 7 nm in height (145, 153). Therefore, polymer size dependency in the efficacy of CD3 camouflage could have arisen due to its proximity to the taller surface protein. Importantly, as the grafting concentration increased, the short chain polymers were able to effectively camouflage even a tall CD4. Another factor that could confer the apparent differential degree of camouflage of the surface antigens is the location of the monoclonal antibody binding site on the protein in relation to the mPEG grafting site, Figure 3-15. This could explain the results of the MCF analysis in 128  Figure 3-16 where a near-complete inhibition of antibody binding was observed for the shortest and the tallest CD3 and CD4 antigens, respectively, while the maximum camouflage of CD28 (medium height) still resulted in approximately 20% MCF, even at the highest mPEG grafting concentration, e.g., 2 mM. In conclusion, PEGylation of human PBMCs resulted in a dose- and polymer lengthdependent inhibition of allorecognition in the context of antigen-antibody interactions. Consistent with the biophysical model, large 20 kDa polymers were more effective at camouflaging CD antigens on large 10 !m PBMC targets. The level of antigen camouflage was dependent not only on the height of the targeted surface molecule but also on the location of this protein in relation to other cellular structures. Moreover, the relative camouflage of an antigen could be influenced by the location of the monoclonal antibody binding site relative to the mPEG grafting site.  129  Figure 3-15: Heights of the surface molecules targeted for camouflage, their distribution on the surface and overall surface topography affected the efficacy of immunocamouflage. A schematic diagram represents the external surface of a T cell and the heights (in nm) of the CD4, CD3, and CD28 surface antigens. The Flory radii (RF: root mean square of end to end length of the polymer chain) of the covalently bound polymers are demonstrated by the circles shown bound to protein X. The RF (in nm) values for 2 , 5 and 20 kDa polymers are shown in the centre of the shaded area. As noted by “protein X”, the RF of a chain may actually extend well above the expected maximal height from the lipid membrane. Importantly, SVA-mPEG non-specifically targets protein lysine residues (K) resulting in both direct and indirect modification of the CD markers. The location of the modified K will dictate the point of origin for the RF. The location of a monoclonal antibody binding site (A1 vs. A2) in relation to the mPEG grafting site (K) will affect the efficacy of polymer-mediated inhibition of antibody binding. The A1 site is much better camouflaged than A2 if the mPEG is attached to K of CD28. The RF values were calculated using the formula: RF = aN(3/5) (a=3.5 Å, N=number of monomers) (87). The diagram is drawn approximately to scale on the vertical axis. Note that the RF values given are for the mushroom configuration of grafted mPEG and are significantly lower than the actual linear length of a given polymer. Le and Scott, manuscript in press.  130  Figure 3-16: Covalent modification of human peripheral blood mononuclear cells resulted in a dose-dependent decrease of mean cell fluorescence due to camouflage of CD3 (A), CD4 (B) and CD28 (C) surface antigens. Mean cell fluorescence analysis did not exhibit a substantial difference in the efficacy of immunocamouflage with respect to the molecular weight of the polymer or the height of the targeted antigen. Efficient camouflage of antigens was observed with all mPEG species even at low grafting concentrations. Mean cell fluorescence was expressed as a percent of the mean cell fluorescence of unmodified cells incubated with antibodies and set to equal 100%.  131  Figure 3-17: Covalent modification of human peripheral blood mononuclear cells resulted in the dose-dependent decrease in percent gated cells due to camouflage of CD3 (A), CD4 (B) and CD28 (C) surface antigens. Percent gated analysis indicated that the inhibition of antigen-antibody interactions was dependent on the size of the polymer as well as on the CD antigen. Consistent with the results of the biophysical latex model, long polymer chains (20 kDa) were more effective at camouflaging the large surface of 10 !m PBMCs. In contrast to the uniform surface characteristics of latex particles, cell surfaces were more complex and this surface topography was reflected in the efficacy of immunocamouflage of cell surface antigens.  3.2.3  PREVENTION OF ALLORECOGNITION IN A 2-WAY MIXED LYMPHOCYTE  REACTION: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION Following the successful inhibition of antigen-antibody binding, the effects of PEGylation on the prevention of cell-cell interactions were investigated. A two-way mixed lymphocyte reaction was performed with two disparate (HLA unmatched) populations of human 132  PBMCs. In these studies, one of the PBMC populations was modified with either the 2 or the 20 kDa SVA-mPEG at 0-2 mM grafting concentrations. Cells were stained with CFSE cell proliferation dye and the extent of T-cell proliferation was estimated by a decrease in the mean cell fluorescence measured using flow cytometry on day 13 of the MLR. Analysis was performed on the CD3+/CD4+ population of T cells and values were expressed as a percent of unmodified T cell activation, set to equal 100%. PEGylation of human PBMCs resulted in a dose-dependent decrease of T cell proliferation, Figure 3-18. A dramatic decrease in lymphocyte activation was observed even at low grafting concentrations, e.g., modification of cells with 0.2 mM, 20 kDa SVA-mPEG resulted in only 28.7±9.9% T cell activation relative to the unmodified control (100%; p"0.001). Consistent with the biophysical camouflage of 8.0 !m latex particles and inhibition of antigenantibody interactions on T cells, prevention of allorecognition in an MLR was dependent on the molecular weight of the polymer and was best achieved with long chain mPEGs. Thus, the 20 kDa polymer resulted in improved efficacy of prevention of T cell-antigen presenting cell interactions at all grafting concentrations (p"0.04). Specifically, PEGylation of T cells with 0.2 mM, 2 and 20 kDa polymers resulted in 64.1±7.6% and 28.7±9.9% T cell activation, respectively, in comparison to the unmodified control MLR (100%; p"0.001). Thus, PEGylation of human PBMCs resulted in a dose- and polymer length-dependent decrease in the allorecognition of foreign antigens in the context of T cell-antigen presenting cell (MHC II) interactions. Consistent with the biophysical model, large 20 kDa polymers were more effective at camouflaging CD antigens on large 10 !m PBMC targets.  133  Figure 3-18: PEGylation of one population of human PBMCs resulted in a dose- and polymer size-dependent decrease in allorecognition and T-cell proliferation in a 2-way MLR. Flow cytometry analysis demonstrated that the 20 kDa polymer was more effective at preventing T cell (10 !m in diameter) activation than the 2 kDa polymer at all grafting concentrations tested. This was in agreement with the biophysical studies on 8.0 !m latex particles and surface CD camouflage of T cells. Values for T cell activation were expressed as a percent of the activation of unmodified T cells (100%).  3.2.4  VIABILITY OF PEGYLATED PERIPHERAL BLOOD MONONUCLEAR  CELLS: EFFECTS OF POLYMER SIZE AND GRAFTING CONCENTRATION The effects of PEGylation on the viability of peripheral mononuclear blood cells were investigated using propidium iodide (PI) and 7-amino-actinomycin D (7-AAD) membrane exclusion dyes. Human PBMCs were PEGylated with 2 and 20 kDa SVA-mPEG at 0-2 mM grafting concentrations, incubated with viability dyes and subjected to flow cytometry. Viability was expressed as a percent of the viability of unmodified cells. Viability measurements were taken on days 0, 1, 7 and 13 following PEGylation. As shown in Figure 3-19, both the PI and 7-AAD viability assays yielded similar results, therefore only the percent viability values obtained in the PI assay were reported below. The viability of the cells was compromised ("20% viability) only when human PBMCs were extensively PEGylated, e.g., 2 mM, and cultured in vitro for >6 days, e.g., 13 days. The viability decrease was mPEG dose-, mPEG molecular weight- and incubation time-dependent. PEGylation of cells with the 2 mM polymers resulted in decreased viability, in comparison to  134  lower grafting concentrations, for all polymer sizes on days 7-13 (p"0.001). For example, PEGylation with the 20 kDa polymer at 0.2 and 2 mM grafting concentrations yielded 99.7±0.2% and 11.4±11.1% cell viability, respectively on day 13 (p"0.001). Long polymers (20 kDa) were shown to decrease viability in comparison to 2 kDa mPEG at 1 mM grafting concentration on day 13 (p"0.001). On the other hand, 2 kDa polymers resulted in decreased viability in comparison to 20 kDa mPEG at 2 mM grafting concentration on day 1. Thus, on day 1 and at 2 mM grafting concentration, 2 kDa modified cells exhibited 81.6±2.0% viability in comparison to 95.5±0.3% cell viability with the 20 kDa mPEG (p"0.02). Prolonged culture of PEGylated cells also affected PBMC viability, especially evident on days 713 for 2 mM, 2 and 20 kDa polymers (p"0.001). Specifically, day 0 vs. day 13 culture of 2 mM 2 kDa SVA-mPEG modified cells resulted in 99.2±0.1% vs. 6.3±6.1% viability. While the cause for the enhanced long-term toxicity is unknown, it was probably the result of mPEG-mediated obfuscation of the cell surface receptors required for the binding of essential auto-, para- and endocrine growth factors. It is unlikely that the decreased viability of PEGylated PBMCs observed on day 13 of the cell culture was responsible for the decreased T cell proliferation measured on day 13 of the MLR, Figure 3-18. The following factors substantiate this conclusion. First of all, the viability studies demonstrated that the unmodified cells maintained >95% viability on day 13 of the cell culture and hence, these unmodified PBMCs would be able to proliferate in response to foreign antigens. Furthermore, the dead PBMCs would still be capable of stimulating allogeneic T cells. In fact, in the standard clinical methodology for 1-way MLR, one of the two human PBMC populations is rendered non-viable by gamma irradiation (section 2.2.6). Thus, the dead PEGylated cells would still be able to stimulate the activation and proliferation of nonPEGylated healthy cells if the antigens on the non-viable modified PBMCs were not sufficiently camouflaged. Finally, the onset of T cell activation and proliferation occurs within 48-72 hours 135  of the MLR. During this period of cell culture PEGylated PBMCs were shown to possess #60% viability, Figure 3-19. Thus, unmodified PBMCs stimulated to divide by live PEGylated cells, during the early stages of the MLR, would be detected in the T cell proliferation assay on day 13.  Figure 3-19: Viability of PEGylated human PBMCs was only compromised at high mPEG grafting concentrations and with extensive in vitro cell incubation as measured by PI (A) and 7-AAD (B) viability assays. Both exclusion dye assays indicated that the viability of PEGylated PBMCs remained at #85% of unmodified control PBMCs at "7 days post grafting and at immunoprotective levels ("1 mM). Cell viability was shown to negatively correlate with the extent of surface modification and invitro cell culturing. Values represented are the percent viability relative to the unmodified control PBMCs (100%).  3.2.5  SUMMARY OF RESULTS FOR THE BIOLOGICAL RED BLOOD CELL AND  PERIPHERAL BLOOD MONONUCLEAR CELL MODEL Consistent with the mPEG-mediated biophysical camouflage of latex surfaces (surface charge camouflage and prevention of protein adsorption), PEGylation of human PBMCs resulted in inhibition of both antibody and cell mediated mechanisms of allorecognition. Decrease of  136  fluorescent antibody binding to CD surface antigens and inhibition of T cell activation in a 2way MLR were readily observed following low levels of cell surface PEGylation. In agreement with the biophysical surface camouflage of 8.0 !m particles, inhibition of allorecognition on 10 !m PBMCs was best achieved with the long, 20 kDa polymer chains. In contrast to the biophysical latex model, efficacy of cell surface camouflage was dependent on the surface topography, the height of the surface protein targeted for PEGylation and perhaps, the location of an antibody binding site. Furthermore, PEGylation of cells did not result in any acute toxic effects and did not adversely affect cell viability at immunoprotective levels of surface modification. Viability of the cells was reduced only at very high grafting concentrations, e.g., 2 mM and with prolonged in vitro cell incubation, e.g., 13 days.  137  IV. DISCUSSION 4.0  OVERVIEW The biophysical mechanisms of immunocamouflage of labile biological targets were  studied by utilizing a robust polystyrene latex model. In-depth biophysical studies on PEGylated latex particles identified two main mechanisms responsible for surface immunocamouflage. Surface charge camouflage was achieved via an mPEG-mediated hydrodynamic charge shielding mechanism dependent primarily on the molecular weight of the polymer and to a lesser extent, on the surface density. Inhibition of surface-macromolecule interactions was achieved via a thermodynamic effect of mPEG polymer steric repulsion that was primarily dependent on the polymer surface grafting densities and to a lesser extent on the size of the polymer. Both biophysical mechanisms were further affected by polymer linker chemistry and architecture, as well as the target size. The same biophysical mechanisms were shown to be important for the immunocamouflage of human erythrocytes and PBMCs. Surface charge camouflage and inhibition of antigen-antibody and T cell-APC interactions were readily demonstrated upon low levels of cell PEGylation without adverse effects on cell viability. Biological surface camouflage was also influenced by chemical and physical parameters of polymers and target size and, additionally, was shown to be dependent on surface topography and surface protein distribution. Most importantly, it was shown that hydrodynamic shielding and steric repulsion mechanisms were capable of preventing antibody and cell mediated pathways of allorecognition at the early stages of receptor-ligand binding.  138  4.1  BIOPHYSICAL ALIPHATIC AMINE POLYSTYRENE LATEX  PARTICLE MODEL The biophysical latex particle model, in contrast to previous PEGylation studies on labile cells and viruses (112-114, 116-120, 123, 130-133), has allowed for a more careful delineation of the biophysical mechanisms underlying the biological immunocamouflage of cell surfaces. In support of the hypothesis of this study it was demonstrated that covalently attached mPEG polymer chains imparted significant biophysical camouflage of surfaces via masking of surface charge and inhibition of surface-macromolecule interactions. The two biophysical mechanisms mediating surface camouflage of latex were determined to be as: 1) hydrodynamic shielding; and 2) steric rejection. I will briefly explain the actions of both mechanisms below. In aqueous solution, every charged surface, in our case positively charged latex, accumulates around it counter-ions (e.g., negatively charged salt ions) coming from the bulk medium in which the surface is suspended (e.g., buffer, water or plasma), Figure 4-1A. These counter-ions neutralize the surface charge, thereby creating an electric potential gradient. The electric potential is the highest at the charged surface (X0) and decreases gradually with the distance (X) away from the surface. There is a region around the charged surface where counter-ions behave as if they were physically adsorbed to the surface. This region is termed shear plane (SP). The location of SP along the surface potential gradient defines the apparent electric potential and the charge of the surface and, consequently the particle’s electrophoretic mobility. A hydrodynamic shielding effect arises due to the ability of a hydrophilic uncharged mPEG to extend the shear plane of the charged surface to a region of decreased surface electric potential (82, 83, 85), Figure 4-1B. The degree to which the shear plane is extended depends on the hydrodynamic thickness (HT) of the polymer  139  layer attached to the surface, which equals the RF of the folded polymer chain. Thus, the hydrodynamic shielding effect is primarily dependent on the molecular weight of the polymer. Steric rejection, on the other hand, is predominantly governed by the polymer chain lateral surface density, Figure 4-1C. As discussed in the Introduction (section 1.5.2) PEG is a highly flexible, hydrophilic polymer occupying a large 3D space, i.e., possessing a large exclusion volume, Figure 1-13. When mPEG polymer chains become covalently bound to the latex surface, their conformational freedom is decreased resulting in a decreased entropy of the polymers. When macromolecules adsorb to mPEG-modified surfaces, e.g., protein adsorption, they contact and compress the polymer chains resulting in a further decrease in entropy of the polymer chains. This highly thermodynamically unfavorable state is reduced by the rejection of molecules from the polymer contact points, preventing protein adsorption and increasing the entropy and the conformational freedom of the polymer chains (74, 76). This effect is most pronounced at higher surface polymer densities, Figure 4-1C. In this regime, chains are more restricted in conformational freedom and protein adsorption resulting in more dramatic steric consequences (71, 72). The experimental results of this thesis indicated that the hydrodynamic and the steric mechanisms of biophysical surface camouflage were highly dependent on the polymer chemical and physical properties as well as the target size. Chemical determinants of immunocamouflage encompassed the polymer linker chemistry and grafting concentration while the physical determinants were related to the polymer architecture, the target and the polymer size, as well as the inherent properties of mPEG, such as high level of hydrophilicity and intrachain flexibility.  140  141  Figure 4-1: Surface charge camouflage is primarily driven by polymer-mediated extension of the shear plane (SP) of the surface towards a region of decreased surface potential (Surface Potential Gradient), whereas a prevention of plasma protein adsorption is achieved due to steric exclusion by the surface-grafted polymer. A) The surface of aliphatic amine polystyrene latex is positively charged due to the presence of primary amines. The negatively charged counter-ions from a bulk aqueous solution migrate and interact with the surface to neutralize the surface charge. This creates an electric potential gradient, with the electric potential being the highest at the surface (X0) and decreasing with the distance away from the surface (X). Shear plane (SP) is defined as the region around the surface, where counter-ions behave as if being physically attached to it. Location of the SP along the electric potential gradient defines the surface charge and the mobility of latex particles. The extension of SP is proportional to the hydrodynamic thickness of the polymer layer, which in turn is governed by the length of the polymer. Thus, 20 kDa polymers (large RF; B) provide improved charge camouflage over 2 kDa polymers (small RF; C). Delta (!) is the difference in the surface potential at the shear plane of a particle modified with the short (!1) vs. the long polymer (!2). While the surface potential gradient begins at the surface, the inherent shear plane (SP) is typically located 1-3 nm above the surface. In contrast, the steric effect is maximized when chains are grafted at higher density (C), i.e., with a small separation between the chains (d). High density grafting is difficult to achieve with polymers possessing a large Flory radius (RF; e.g., 20 kDa mPEG; B) as the initially bound chains sterically inhibit the approach and binding of additional polymers. Modified from Le and Scott, manuscript in press. Chemically, the choice of linker chemistry is of importance for efficient grafting in aqueous solutions. While multiple linker chemistries for polymer grafting have been developed, not all are of biological utility. For example, T-mPEG (O-tresyl-O’-methoxypolyethylene glycol) previously demonstrated very high toxicity and very poor grafting efficacy to biological cells (115, 118). In contrast, other linker chemistries are more biocompatible and non-toxic. In this study we utilized SC, SVA and SPA linker chemistries that react with primary amines (i.e., lysine) on cell surface proteins or particles as discussed in section 2.1.2. The linkages formed by these polymer linkers are highly stable in aqueous solution at physiological pH. As described, SVA-mPEG provided improved biophysical camouflage in comparison to SC-mPEG linker chemistry with respect to surface charge camouflage and the prevention of protein adsorption (Figures 3-4 and 3-9, respectively). This could be largely attributed to the longer hydrolysis half-life of SVA-mPEG (33.6 minutes) versus SC-mPEG (20.4 minutes) in aqueous solution allowing for a higher grafting efficiency at a constant polymer concentration over the 60-minute 142  reaction time, Figure 4-2. In the case of SVA-mPEG, both hydrodynamic and steric mechanisms were enhanced because of the increased amount of surface-grafted mPEG.  Figure 4-2: SVA-mPEG resulted in improved biophysical camouflage of surfaces due to its longer hydrolysis half-life. Hydrolysis half-life is the time it takes for 50% of activated mPEG species to lose their linkers and become incapable of covalent surface grafting. The half-life of SVA-mPEG is 30.6 minutes in comparison to 20.4 minutes for SCmPEG. Thus, in aqueous solution, there will be more activated SVAmPEG available for surface modification during a 60-minute PEGylation reaction.  The second chemical parameter tested was the grafting concentration of mPEG. All experiments performed for this thesis exhibited an mPEG dose-dependent profile with the exception of the commercial IgG anti-PEG antibody binding studies, Figure 3-2. Fluor-SVAmPEG surface grafting demonstrated a dose-dependent increase in the bound mPEG on both 1.2 and 8.0 !m particles, Figure 3-1. This experiment also demonstrated the dose-dependent increase in steric exclusion nature of the surface-bound polymers. Grafting of mPEG to the surface of 1.2 and 8.0 particles exhibited a biphasic curve with the fast and efficient polymer addition at low grafting concentrations followed by slower polymer grafting at high grafting concentrations, Figure 3-1. This indicated that at high grafting concentrations sufficient amount of surface-bound polymer exerted steric hindrance on the approach and binding of additional polymers. This steric effect was diminished on large particles as a larger surface area could accommodate more polymer without exerting much hindrance. Therefore, the change from fast 143  to slow polymer grafting occurred at 10 mM grafting concentration for 8.0 !m particles in comparison to 0.5 mM for 1.2 !m beads. It was also demonstrated that in order to achieve the same degree of surface coverage as seen with the 1.2 !m particle, a 10-fold increase in the amount of polymer was necessary to camouflage the larger surface area of an 8.0 !m bead. A dose-dependent effect was also noted for the steric exclusion mechanism of mPEG that was responsible for the inhibition of unlabeled (Figures 3-6 and 3-7) and fluorescent protein adsorption (Figures 3-9, 3-11 and 3-12) to 1.2 and 8.0 !m particles. Higher grafting concentrations resulted in the greater surface density of polymers and hence enhanced steric repulsion of plasma proteins by the polymer chains. An interesting observation was that while the maximum mPEG binding capacity was not reached, protein adsorption was maximally decreased between 0.5 and 1.0 mM grafting concentrations for the 1.2 !m latex particles. This finding suggested that a complete saturation of binding sites was not necessary to achieve an efficient inhibition of protein adsorption. A decrease in the electrophoretic mobility of PEGylated 1.2 and 8.0 !m latex particles was also mPEG dose-dependent (Figure 3-4). The hydrodynamic shielding effect was enhanced due to improved polymer surface coverage at high grafting concentrations. A high surface density of polymer resulted in a global extension of the shear plane towards a region of decreased surface potential (Figure 4-1A-B). Effective charge camouflage, as will be discussed, is of significant biological importance. The physical architecture of the polymer was an important contributor to the hydrodynamic and steric mechanisms of immunocamouflage. Both SVA-mPEG and SC-mPEG are monofunctional in that they contained only one linker group per polymer chain. Studies on red blood cells demonstrated that use of bifunctional linear chains resulted in red blood cell aggregation (130). To more fully explore this finding, the effects of monofunctional and multifunctional polymers on latex aggregation, surface charge camouflage and the inhibition of 144  fluorescent protein adsorption were examined. As shown in Figure 3-3, significant aggregation was noted with the multifunctional polymer due to crosslinking of the latex particles while this effect was not seen with the monofunctional mPEG. Indeed, the monofunctional chemistries produced a steric hindrance effect that maintained particles in monodisperse suspensions that assured a higher homogeneity of derivatization. While gross aggregation of cells is likely to be a biologically adverse event, protein-protein crosslinking on a single cell may also have significant deleterious effects by potentially triggering signal transduction pathways leading to activation or apoptosis of a cell. Indeed, studies on human and murine lymphocytes suggest that this is of importance (154-156). Similarly, the surface charge camouflage (Figure 3-5) and inhibition of fluorescent protein adsorption (Figure 3-10) were best achieved with the linear monofunctional mPEGs. Due to the highly aggregating nature of the multi-functional PG polymer, the surface modification of latex was non-homogeneous. This resulted in a low density of polymer and incomplete surface coverage of the latex, which greatly reduced the effectiveness of the hydrodynamic and steric effects. This was further supported by the poor dose- and molecular weight-dependence of the surface camouflage by the multifunctional PG polymer. Because of its branched nature, increases in molecular weight of PG did not result in the increased size of the molecule but rather increased density of branching. Thus, a PG polymer had a smaller exclusion volume and RF size in comparison to the highly flexible mPEG chain of equivalent molecular weight. Therefore, PG polymers exhibited decreased steric and hydrodynamic surface camouflaging effects in comparison to mPEG. Bridging the chemical and physical bases of immunocamouflage was the molecular weight of the linear (but not the highly branched) polymers. As demonstrated, the biophysical camouflage of surface charge was directly proportional to the molecular weight (bigger is better) of the polymer. This was expected, as the hydrodynamic shielding of surface charge is 145  proportional to the hydrodynamic thickness of the polymer layer, and was best achieved with long polymers, regardless of the particle size or the linker chemistry (Figures 3-4). Indeed, at very high grafting concentrations, a complete surface charge camouflage of the 1.2 !m particles was achieved with large polymers (SVA-mPEG, 20 kDa). The biological effects of charge camouflage are multiple and underlie the immunocamouflage-mediated inhibition of antigenantibody recognition, allorecognition, viral invasion, and cell-cell interactions noted in previous studies (112-114, 116-120, 123, 130-133). In contrast to charge camouflage, prevention of protein adsorption was achieved primarily due to the steric exclusion effect of the grafted polymer chains. In contrast to surface charge camouflage, inhibition of surface-macromolecule interactions was dependent on the surface density of the polymer. Due to their large size, long chain polymers exhibited stronger self exclusion effects in comparison to the short polymer chains. Hence, dense polymer grafting was best achieved with short chain polymers. The protein adsorption data was consistent with this observation and demonstrated improved surface camouflage with short 2 kDa polymers on 1.2 µm latex particles (Figures 3-6 and 3-9). This, however, was overcome by increased grafting concentration of 20 kDa mPEG. The dependence of immunocamouflage on the relative surface area targeted (i.e., target size) for modification was also investigated. Fluorescent plasma adsorption experiments on the 1.2 and 8.0 !m diameter latex beads demonstrated that the target size dictated the optimal size of the polymer necessary for inhibition of protein adsorption. Interestingly, it was discovered that the larger polymers (20 kDa) were more effective at preventing protein adsorption on bigger 8.0 !m latex particles, while shorter polymers (2 kDa) were more effective at preventing protein adsorption on smaller, 1.2 !m particles (Figures 3-11 and 3-12). As explained previously, steric rejection of proteins is best achieved when polymers are grafted at high density or provide sufficient surface coverage. PEGylation at high surface density on 1.2 !m particles was better 146  accomplished with shorter polymers (grafted at high concentration) as the grafting of long chains at high density would be prevented via steric restriction by the initial surface-bound mPEG. On the other hand, the coverage of the large surface area of 8.0 !m particles is best achieved with the large polymer chains due to their increased RF in comparison to small polymers chains. This finding is especially important when camouflaging biological targets of varying sizes, e.g., cells vs. viruses. A shown in Figure 4-3, recent biological studies on the prevention of respiratory syncytial virus (RSV) infection supported the target size effect demonstrated by the biophysical latex model. RSV infection was best prevented when small virus particles were modified with short polymer chains or when large host cells were PEGylated with long mPEGs.  147  Figure 4-3: The size of the PEGylation target influences which polymer size most effectively imparts immunocamouflage. A) Inhibition of surface-macromolecule interactions on small latex particles (1.2 !m) was best achieved with short chain polymers (2 kDa) grafted at high surface density. B) Similarly inhibition of virus-receptor interactions was best achieved when small virus particles (0.12-0.3 !m) were PEGylated with short mPEG polymers (2 kDa). C) In contrast, inhibition of surfacemacromolecule interactions on large 8.0 !m latex beads was best achieved with large chain polymers (20 kDa). D) In agreement with this finding, prevention of virus-receptor interactions was most effective when large host cells (10-20 !m) were PEGylated with large polymer chains (20 kDa). Inhibition of surface-macromolecule interactions was measured as a decrease in the amount of latex adsorbed fluorescent human plasma proteins. Attenuation of virus-receptor interactions and consequent viral entry was measured as a decrease in expression of green fluorescent protein respiratory syncytial virus (GFP-RSV) by host cells (HeLa). Modified from Le and Scott (manuscript submitted) and Sutton, T.C. (129).  148  The protein adsorption (unlabeled and fluorescent) and SDS-PAGE studies demonstrated (Figures 3-6, 3-7 and 3-9) a significant decrease in total protein adsorption to PEGylated surfaces in comparison to the unmodified latex control. In order to study the composition of the latex adsorbed protein layer, the supernatants containing the total protein desorbed from differentially modified particles were subjected to iTRAQ labeling and MS analysis. The values in this study were normalized to 100 !g total protein desorbed from all latex samples and the specific protein abundance on PEGylated latex was expressed relative to that on the bare latex control (100%). As shown in Table 3-1, iTRAQ/MS analysis demonstrated that 69% of all the identified plasma proteins exhibited decreased relative abundance on PEGylated particles while only ~11% exhibited an increased relative abundance. More importantly, as shown in Tables 31 and 3-2, plasma proteins that are crucial mediators of recognition and rejection of allogeneic donor cells/tissue and implantable materials (138, 157-159) were almost universally depleted on the PEGylated surfaces. Specifically, this subset of proteins consisted of immune-activating species such as coagulation proteins, complement factors and immune regulatory components. In addition, the decreased relative abundance was noted for proteins playing inhibitory roles in the complement and coagulation pathways. This was not surprising as these proteins are the key regulators of the immune responses and maintain the overall homeostasis by associating with and inhibiting the complement/coagulation pro-activating proteins and preventing their potentially damaging actions in vivo. Our previous studies on biological cells showed similar mPEG-mediated “anti-fouling” effects in antibody induced agglutination reactions and red blood cell Rouleaux formation (112, 113, 115, 117, 130). Both steric and hydrodynamic mechanisms were involved in rendering PEGylated surfaces more biocompatible. Consistent with the reports of others (79, 137, 160, 161) and the results of this thesis (Figures 3-6 and 3-7), steric rejection of proteins was a crucial mediator of surface camouflage. The hydrodynamic charge shielding mechanism of PEGylation 149  was supported by the observation that camouflage of the positive surface charge of the latex resulted in specific depletion of the negatively charged proteins on the PEGylated surfaces, Table 3-2. As shown in Table 3-4, there were 10 proteins that exhibited no change in their abundance on either bare or PEGylated surfaces. As expected, these proteins serve homeostatic functions (e.g., binding and transport) and their adsorption would not be affected by surface characteristics. Only 4 proteins from this subset have immunological roles: C1q (c-chain), Von Willebrand factor, platelet basic protein and platelet factor 4. However, none of these proteins are directly implicated in immune activation. For example, the c-chain of C1q is responsible for the binding of C1q complex, and Von Willebrand factor acts as a chaperone for coagulation factor VIII. Thus, both proteins fall into the subset of binding and transport proteins. Platelet factor 4 and platelet basic protein belong to the family of chemokines responsible for downregulation of immunological response. Platelet factor 4 inhibits proliferation of lymphocytes and cytokine release by T cells (162). Platelet basic protein was shown to decrease chemotaxis and function of neutrophils in pro-inflammatory response (163). In addition, chemokines are signaling proteins and unless specifically bound by their receptor would, most likely, not exert any biological effect. Thus, none of these proteins are implicated in the activation of the immune system and their adsorption should not result in adverse immune reactions. The increased degree of biocompatibility of PEGylated surfaces noted in the iTRAQ/MS analysis could also be attributed to the inherent physical properties of mPEG, such as the high degree of chain flexibility and the spatial structure of the folded polymer chain. As suggested by L. Vroman, the reason for the biological incompatibility of synthetic materials could be a lack of the architectural complexity that is common to all biological surfaces (e.g., cells, tissues, vasculature) (164). Thus, decreased relative abundance of pro-inflammatory and pro-thrombotic proteins on PEGylated surfaces could be the result of the increased architectural complexity 150  resembling epithelial surfaces and, most importantly, the blood vessels. Interestingly, as demonstrated in Table 3-3, the majority of proteins that exhibited increased relative abundance on PEGylated surfaces constituted extracellular matrix, collagen or fibrin-like proteins, or proteins involved in the adhesion and binding of collagen-like proteins. This observation may also underlie the increased abundance of 4 proteins involved in immune functions: complement C1q (alpha and beta chains), Ig gamma-1 chain (C region), beta-2-glycoprotein, and thrombospondin-1 (Table 3-3). The C1q (alpha and beta chains) protein interacts with the constant chain of IgG (165, 166) and both could become bound to the PEGylated surfaces due to C1q’s collagen-like structure (167). The increased relative abundance of thrombospondin 1 was, most likely, due to its complex multi-domain structure that allows it to interact with celladhesive receptors and glycoproteins, including fibronectin and collagen (168). Finally, the binding of beta-2-glycoprotein I precursor and thrombospondin could potentially serve beneficial functions in cell and tissue transplantation. Beta-2-glycoprotein binds to negatively charged phospholipids and prevents initiation of the coagulation cascade (169, 170), while thrombospondin-1 is a potent inhibitor of T-cell and dendritic cell activation (171, 172). The biocompatible nature of PEGylated surfaces was, in part, derived from the increased hydrophilicity imparted by mPEG. PEG is a highly hydrophilic polymer that binds 3 molecules of water per each ethylene oxide unit. A bare latex surface, on the other hand, is highly hydrophobic and therefore, considered to be non-biocompatible (173, 174). The propensity of hydrophobic surfaces to be thrombogenic is, partially, due to the increased adsorption and surface denaturation of plasma proteins that are then bound with increased avidity. Easily denatured proteins (“soft”) are capable of unfolding and binding to hydrophobic surfaces via multiple contacts contributed by the exposure of internal hydrophobic residues. This leads to a very strong adhesion through additional H-bonds, hydrophobic, charge-charge and Van der Waals interactions (175, 176). In addition, it results in the exposure of charged residues and 151  internal antigenic sites of proteins that could elicit an immune response. In our desorption studies, it was demonstrated that not all of the adsorbed proteins could be successfully desorbed from the hydrophobic bare latex surface as some proteins formed a denatured protein layer, Figure 3-13. In contrast, fewer proteins were resistant to desorption from the PEGylated latex beads. Indeed, the amount of permanently adsorbed proteins decreased with increased mPEG modification as the hydrophobic surface area available for protein denaturation became reduced and the hydrophilic polymer modified surface became dominant.  4.2  SUMMARY OF DISCUSSION ON THE BIOPHYSICAL MODEL In summary, PEGylation of latex surfaces imparted significant immunocamouflage via  two main biophysical mechanisms: hydrodynamic shielding of surface charge and steric inhibition of surface-macromolecule interactions. Both the chemical and physical properties of the polymer and the size of the target had a significant impact on the efficacy of surface camouflage. The hydrodynamic charge shielding effect was primarily dependent on the height of the polymer layer and the overall surface coverage. Hence, surface charge camouflage was best achieved with large mPEG polymers grafted at higher density. Steric inhibition of surfacemacromolecule interactions was largely dependent on the surface density of the polymer chains and the overall coverage of the surface. Thus, inhibition of protein adsorption was best accomplished by grafting small polymers (2 kDa) to small targets (1.2 !m) and large polymers (20 kDa) to large targets (8.0 !m). Both the hydrodynamic and steric effects were improved when a monofunctional, non-branched polymer possessing extended linker hydrolysis half-life was used. PEGylated surfaces demonstrated not only a decreased total human plasma protein adsorption but also a substantial depletion of specific pro-activating immune proteins. Thus, the hydrodynamic and steric effects as well as the inherent physical properties of the polymer such  152  as hydrophilicity and architectural complexity of the folded polymer chain, all contributed to rendering PEGylated surfaces more biocompatible.  153  4.3  BIOLOGICAL RED BLOOD CELL AND PERIPHERAL BLOOD  MONONUCLEAR CELL MODEL The biophysical effects of a grafted polymer have great significance for the understanding of the biological effects of cellular PEGylation. Antigen-antibody and cell-cell interactions occur at the cell surfaces and involve non-specific interactions, such as chargecharge and non-covalent binding/adhesion, as well as specific interactions dependent on the structure and geometry of the proteins and the binding grooves. The biophysical studies of the polystyrene latex model demonstrated that non-specific interactions were readily prevented by hydrodynamic shielding and steric exclusion mechanisms mediated by mPEG. Therefore, the effects of these camouflaging mechanisms were studied with respect to biological interactions. The biological consequences of PEGylation were exemplified by the charge camouflage of intact erythrocytes (Figure 3-14). In agreement with the surface charge camouflage of 8.0 !m latex particles, hydrodynamic shielding of the negatively charged 8.0 !m RBCs was best achieved with long chain polymers (20 kDa) due to their large RF and hence, increased degree of shear plane extension, Figure 4-1A-B. In contrast to surface charge camouflage of latex surfaces, PEGylation of cells resulted in the complete camouflage of charge as indicated by zero mobility of the modified RBCs in an electric field. The apparent difference in the efficacy of the surface charge camouflage of latex and RBCs was, most likely, the result of the differences in the initial charges of the beads vs. the cells. Latex particles had greater surface charge density (number of charges per surface area) than RBCs: 7.6*109 charges per particle vs. 1.0*107 charges per cell, respectively (152). In addition, the charges on latex particles were distributed relatively uniformly, whereas 95-100% of the negative charges on RBCs are associated with sialic acid residues clustered on surface glycoproteins (177). Since PEGylation targeted residues  154  on the surface proteins, negatively charged sialic acid residues would have a great potential of being efficiently camouflaged as they are also located on surface proteins, Figure 4-4.  Figure 4-4: Differences in the efficacy of surface charge camouflage on latex particles vs. RBC by grafted mPEG. The surface density of positively charged NH3+ groups on a latex bead is approximately 800 times higher than the surface density of negatively charged COO- groups on a red blood cell. Moreover, the primary amine groups on the beads are themselves the sites for mPEG grafting and are evenly distributed on the surface. On the other hand, carboxylic groups on the RBC are from the sialic acid residues and are clustered together on surface glycoproteins, which donate lysine or N-terminal primary amine groups for the grafting of an mPEG chain. Thus, the number and the distribution of surface charges on latex vs. RBC are the main factors responsible for the decreased efficacy of surface charge camouflage of the 8.0 !m bead vs. the 8.0 !m RBC. Consistent with the biophysical latex model, PEGylation of PBMCs also resulted in inhibition of surface-macromolecule interactions. This was exemplified by the inhibition of fluorescent antibody binding to CD antigens on the surface of the cells and the inhibition of T cell proliferation in a 2-way MLR, Figures 3-16 – 3-18. Both antigen-antibody and T cell-APC interactions involve receptor-ligand binding. There are at least two requirements for the proper 155  binding to occur: 1) the receptor and the ligand must be of complementary size and geometrical shape to allow for a tight “fit”; and 2) the grooves of the receptor and the surface of the ligand fitting into the groove, must contain amino acid residues of complementary physico-chemical properties for efficient non-covalent interactions to take place, Figure 4-5A-B. For example, charge-charge interactions occur between oppositely charged atoms, H-bonds are formed between an electron acceptor and an electron donor, and van der Waals interactions are maintained by the fluctuating electron clouds of closely located atoms. When both of the above mentioned criteria are met, a ligand fits tightly into the receptor groove and the binding is further strengthened by a number of non-covalent forces. Importantly, non-covalent forces are weak and are only efficient at close distances (r). For example the strength of the charge-charge interaction decreases with 1/r2, whereas, the reduction in van der Waals forces occurs even more drastically, with 1/r6. Correlating the biophysical results with the inhibition of ligand-receptor interactions indicated that both hydrodynamic and steric mechanisms of PEGylation, identified using the polystyrene latex model, played an important role in the immunocamouflage of biological targets. The mPEG-mediated hydrodynamic surface effects resulted in a charge camouflage and therefore a decreased propensity for electrostatic interactions, such as charge-charge, van der Waals and H-bonds. Moreover, steric effects of the grafted mPEG prevented the close approach of a receptor and a ligand, thus inhibiting any non-electrostatic interactions and further contributing to a loss of electrostatic binding due to the increased separation between the two molecules, Figure 4-5C. These effects would have dramatic consequences on T cell-APC interactions. As discussed in the Introduction (section 1.3.3), several sequential binding steps involving co-TCR proteins, adhesion, and co-stimulatory molecules, must take place in order for the efficient T cell mediated allorecognition to occur. If any of those binding interactions should fail, T cells could become anergic and unresponsive to further or de novo stimulation by 156  the same antigen or be forced to undergo apoptosis. In fact, our previous studies with the TAGVHD model (116) demonstrated that a responder population of lymphocytes that has been previously exposed to a PEGylated stimulator cell population entered a state of unresponsiveness indicated by a failure to proliferate due to the subsequent introduction of exogenous IL-2 or unmodified stimulator lymphocytes. Furthermore, mPEG-lymphocytes were shown to undergo enhanced apoptosis in comparison to unmodified control cells, most likely, as a result of clonal deletion of alloresponsive cells due to a weak allogeneic stimulation, e.g., weak co-stimulation through the CD28:CD80 interaction (section 1.3.3).  Figure 4-5: Effects of hydrodynamic charge shielding and steric repulsion by grafted mPEG on receptor-ligand interactions. A) Receptor ligand interactions occur only between molecules of complementary size and geometrical shape. B) These interactions also require complementarity in the physicochemical properties of amino acids forming the contacts between the ligand and the receptor. C) Grafted mPEG polymer hydrodynamically camouflages electrical charges, inhibiting any electrical interactions. mPEG chains also sterically inhibit the close approach of the ligand and the receptor, further contributing to inhibition of the ligand-receptor binding.  Importantly, consistent with the biophysical results of the 8.0 !m latex, both hydrodynamic charge shielding and steric repulsion effects on the 10 !m PBMCs contributed to  157  inhibition of antibody and cell mediated pathways of allorecognition, and were best achieved with long polymer chains (20 kDa), Figures 3-17 and 3-18. However, in contrast to the simple latex bead, which had a relatively uniform/flat surface, the efficacy of immunocamouflage of the cells was dramatically affected by the topography (e.g., antigen height relative to the membrane) of the cell and the antigen’s location with respect to other cellular structures (Figure 3-15). Indeed, as demonstrated (Figure 3-17) the relatively low height CD3 (4 nm) and CD28 (7 nm) antigens were well camouflaged with all polymer lengths. In contrast, CD4 (13 nm in height) was only effectively camouflaged by the 20 kDa polymer at intermediate grafting concentrations (e.g., 2 mM). However, at very high grafting concentrations (e.g., 4 mM) all polymers demonstrated excellent immunocamouflage (>70% reduction in antibody positive PBMC) consequent to their very high grafting density thereby significantly lessening potential antigenicity in vivo (116). On the other hand, localization of the protein targeted for camouflage was also shown to be an important factor in defining efficacy of immunocamouflage. For example, the CD3 antigen (3 nm) was smaller than CD28 (7 nm) but its camouflage was dependent on polymer molecular weight, whereas CD28 was readily camouflaged by all mPEGs. This could be attributed to the fact that CD3 protein is part of a multi-subunit complex – T cell receptor complex, Figure 1-10. Two copies of CD3 are non-covalently associated with the TCR and CD247. While CD247 is much shorter than CD3 and should not affect its camouflage, TCR is 7 nm tall and is located only 12 Å away from CD3 (153). Thus, close localization of CD3 to other protein structures and its proximity to a taller TCR resulted in long polymer chains (20 kDa) being more efficient at CD3 camouflage in comparison to short chain mPEGs, Figure 3-17A. Furthermore, in these studies various CD antigens were selected to reflect the differential stages in the T cell allorecognition and activation as described in section 1.3.3, Figure 1-11. In support of the Significance of Cellular PEGylation (section 1.7.2), it was shown that mPEG 158  modification of PBMCs resulted in a global camouflage of the cell surface. Specifically, it was demonstrated that PEGylation efficiently inhibited the initial stages of T cell activation prior to the myriad biochemical intracellular reactions that take place, Figure 4-6. Thus, camouflage of CD3 inhibited the first step in T cell activation that results from TCR binding to a peptide:MHC complex. A further event of co-receptor binding was also prevented via camouflage of CD4 antigen. Potential inhibition of the co-stimulatory pathway was demonstrated by masking of the CD28 molecule. Importantly, PEGylation blocked receptor-ligand interactions at the surface level of the allorecognition pathway, which was in contrast to the stages targeted by pharmacological inhibitors that result in systemic toxicity and long-term complication in patients undergoing tissue transplantation, Figure 4-6.  Figure 4-6: Inhibition of T cell activation and proliferation by PEGylation vs. immunosuppressive therapy. Immunosuppressive drug/antibody mediated inhibition (purple X) of T cell activation (star1, star2, star3), proliferation [6] and differentiation [7] occurs due to suppression of intracellular biochemical pathways. Pharmaceutical drugs target late stages of T cell activation (star 2) and T cell proliferation [6] and immunosuppressive antibodies inhibit signaling through cell surface receptors [5], which also occur in the late stages of T cell activation. On the other hand, global, nonspecific PEGylation (shaded areas) of cell surface proteins on either APC or T cell, inhibit (red X) cell-cell interactions: cell adhesion [1], antigen recognition [2], coreceptor binding [3] and co-stimulation [4]. All of these are necessary for, and upstream, of the cell signaling events that activate T cells to undergo division and differentiation.  159  Finally, the effects of PEGylation on acute (days 0 and 1) and chronic (days 7 and 13) cellular toxicity were investigated. As shown in Figure 3-19, none of the mPEG conditions resulted in acute cell toxicity. Chronic toxicity was only observed at the highest (2 mM) grafting concentration and on day 13 of in vitro cell culture. Additionally, the difference in acute and chronic toxicity was shown with respect to molecular weights of the polymers. For example, a slight decrease in the viability was noted with 2 kDa mPEG in comparison to 20 kDa polymer at 2 mM grafting concentration on days 1-7. On the other hand, 20 kDa mPEG resulted in chronic toxicity on day 13 at 1 mM mPEG concentration in comparison to no decrease in viability with the 2 kDa mPEG at the same grafting concentration. Importantly, as shown in Figures 3-16 and 3-18 efficient inhibition of allorecognition was achieved at "0.5 mM grafting concentrations, the level of PEGylation at which viability of cells remained at #80% in comparison to the unmodified cells (Figure 3-19).  160  4.4  SUMMARY OF DISCUSSION ON BIOLOGICAL MODEL In summary, the biological PBMC model supported the findings of the biophysical latex  model and demonstrated that hydrodynamic surface charge camouflage and steric prevention of non-specific surface-macromolecule interactions resulted in immunocamouflage of biological surfaces. A complete surface charge camouflage was achieved with RBCs. Both the antibodyand cell-mediated mechanisms of allorecognition were readily inhibited at low grafting concentrations of mPEG, without adverse effects on cell viability. No acute cell toxicity was noted even with the increased degree of PEGylation. In addition, various stages during T cell activation were prevented due to camouflage of accessory and co-stimulatory surface molecules. Consistent with the biophysical latex model, immunocamouflage of large, cellular targets was best achieved with high molecular weight mPEG. In contrast to surface modification of synthetic beads, PEGylation of cellular membranes was affected by cell surface topography and surface protein distribution.  161  4.5  FUTURE DIRECTIONS Future PEGylation studies should aim at the practical applications of this technology to  overcome or to decrease the allorecognition of allogeneic tissues in transfusion and transplantation medicine and to design safe and effective means of broad-spectrum antiviral prophylaxis. While there are multiple clinical situations where PEGylation of allogeneic tissues could be of benefit, I will discuss three potentially high impact examples of utility. These are: 1) increase of donor blood availability in transfusion medicine; 2) prevention of allorecognition in pancreatic islet transplantation; and 3) design of a broad-spectrum antiviral prophylaxis to prevent respiratory infections. A serious concern in the blood banking system (e.g., CBS) is the shortage of blood, especially of the “universal” donor blood type – O negative (O, RhD-), Figure 4-7. The demand for this blood type is substantiated by the phenotype of the “universal” red blood cell that is devoid of any major blood group antigens such as A, B and RhD. Therefore, O negative red blood cells are considered to be “universal” donor cells as they can be transfused to potentially any blood group patient, given that he/she is not sensitized to any of the minor blood group antigens present on the “universal” cell. While it would be desirable for CBS to achieve the immunocamouflage of major blood group antigens and, thus, create a single universal RBC pool, camouflage of A and B blood groups may be problematic due to their carbohydrate nature, size and copy number. Furthermore, a defect in “production” of A and B antigen camouflage could have serious clinical consequences. On the other hand, if camouflage of the proteinacious RhD antigen is achieved it would result in “near-universal” erythrocytes through conversion of A, B and AB positive cells into A, B and AB negative RBCs. But most importantly, camouflage of the RhD antigen of O positive cells, which make up 37% of the donor pool, would result in O negative “universal” RBCs and, thus, a dramatic increase of potential blood donors from 6.6% to 43.6%, Figure 4-7. 162  In order to test the efficacy of immunocamouflage of the RhD antigen, human RhD positive erythrocytes could be isolated and derivatized with activated mPEG to impart surface immunocamouflage. The camouflage of RhD antigens could be tested in vitro by subjecting modified red blood cells to anti-serum agglutination reactions. The degree of RBC agglutination could be either estimated directly or measured via aggregometry. Further in vitro experiments could look at the decreased specific recognition and binding of the RhD antigen by fluorescent anti-RhD antibody, followed by flow cytometry. Furthermore, complement and antibody mediated cell lysis of PEGylated RhD positive RBCs could be performed by incubating these cells with anti-RhD antibody followed by the addition of complement. The degree of RBC lysis could be estimated spectrophotometrically as an increase in hemoglobin at 540 nm. Finally, in vivo studies could be performed in transgenic mice expressing RhD antigen on the surface of the mouse erythrocytes. RhD positive RBCs could then be PEGylated and transfused into nontransgenic mice of the same strain. Since mice do not naturally express RhD antigens, the animals would be expected to mount a humoral as well as an adaptive immune responses to nonPEGylated RhD+ RBCs. This could be measured as a decrease in circulation half-life of fluorescently labeled RhD+ RBCs in normal mice. A more relevant human model would be to perform an RBC agglutination assay utilizing the serum of RhD- women who have given birth to RhD+ babies. RhD- women produce anti-RhD antibody in response to the maternal-infant blood mixing during the childbirth. These experiments will predict the efficacy of mPEG-mediated camouflage of RhD antigens on human RBCs and the extent to which RhD unmatched RBCs would become recognized by naturally produced human antibodies. The mixing of non-PEGylated RhD+ human erythrocytes with the serum of pre-sensitized RhD- women would result in the extensive RBC agglutination due to anti-RhD antibody binding to the cell surface of RhD+ erythrocytes.  163  Based on the results described in this thesis, I would expect the immunocamouflage of large 8.0 !m RBCs to increase with the grafting concentration and the molecular weight of the mPEG. Thus aggregometry, antibody binding and complement-mediated cell lysis experiments should demonstrate a decreased recognition of the RhD antigen on the surface of PEGylated RBCs in comparison to an unmodified control. In addition, in vivo studies should demonstrate an increased circulation half-life of modified cells due to their decreased immune recognition and clearance.  Figure 4-7: Blood type availability at CBS. “Universal” red blood cells, O negative, are highly desirable in the blood banking system but make up only 6.6% of the total blood donor pool. A successful camouflage of RhD antigens by mPEG could result in conversion of O positive blood cells, which make up a large percentage of donated blood, into O negative universal donor cells. Figure courtesy of M. D. Scott. With respect to transplantation medicine, the efficacy of PEGylation could be tested in pancreatic islet transplantations to treat diabetes mellitus (DM). DM is characterized by hyperglycemia due to abnormal secretion and/or utilization of insulin. Replacement of insulin producing beta cells derived from healthy donors is the only effective means of re-establishing and maintaining long-term glucose homeostasis in patients with DM. Recently, pancreatic islet transplantations were explored as a superior alternative to the historical whole pancreas transplantations due to decreased antigen load and technical simplicity (117, 178). But the problem with the islet transplantation remains, as with other types of transplantations, in the allorecognition of foreign antigens on the surface of islets or beta cells. Thus, PEGylation of 164  pancreatic islets could be beneficial to treating DM due to decreased allorecognition and rejection of immunocamouflaged transplanted tissues. The efficacy of PEGylation in prevention of allorecognition of pancreatic islet transplants could be first assessed using an in vitro study. Both, rat (117) and mouse (179) diabetes models exist, I will focus on using a murine model. Mouse islet cells could be isolated and derivatized with the activated mPEG. Inhibition of anti-islet antibody binding to the proteins located either on the pancreatic capsule (e.g., collagen and fibronectin) or on beta cells themselves (e.g., Glut2 glucose transporter), could be investigated using commercially available primary and secondary fluorescent antibodies. Next, in vivo studies utilizing a diabetic mouse model could be performed. Mice would be induced to become diabetic using a chemical, streptozotocin. PEGylated healthy donor islet cells would be transplanted into an allogeneic diabetic mouse and the success of the engraftment and maintenance of the transplant function would be monitored based on the achievement of normoglycemia in the diabetic mice. Based on our previous work (117) and the findings in this thesis, antibody binding to PEGylated pancreatic islets would be expected to decrease, dependent on the dose and the molecular weight of the polymer. Similar to PBMC camouflage, immunocamouflage of islets would be best achieved with long polymers grafted at high density. It is expected that PEGylated pancreatic islets would demonstrate decreased allorecognition and hence, increased efficacy of establishing normal glucose levels. I would expect the mPEG grafting concentration to be critical in in vivo studies as the viability of PBMCs was decreased at very high grafting concentrations, Figure 3-19. So it would be important to find the optimal grafting density of polymers, at which functional properties of islets are not compromised but the allorecognition is sufficiently attenuated to prevent rejection. Finally, the potential of utilizing PEGylation technology to prevent common respiratory viral infections could be explored. Currently, there are no effective means of preventing 165  respiratory viral infections and this is especially of concern due to recent outbreaks of severe acute respiratory syndrome (SARS) and the H1N1 pandemic. Over-the-counter drugs only treat the consequences of the host’s immune response to the virus and do not inactivate or remove the pathogen itself. Thus, the design of a broad-spectrum antiviral prophylactic agent would be highly beneficial. As shown in previous studies (123-129), viral infections are successfully prevented via PEGylation of host cells and the consequent immunocamouflage of cell surface receptors responsible for the entry of the virus. Thus, if PEGylation of epithelial surface layer, through which environmental pathogens gain their access inside the cell, e.g., nasal cavity and nasopharynx, is PEGylated it could potentially result in the attenuation of the viral infection. Moreover, the activated mPEG solution can be easily formulated as an aqueous nasal gel or a throat spray applicator, Figure 4-8. The efficacy of PEGylation of epithelial surfaces could be tested in vivo using a guinea pig respiratory syncytial virus (RSV) model. Guinea pigs will undergo the treatment – PEGylation of nasal and nasopharyngeal surfaces followed by inoculation with the virus. The degree of infectivity could be assessed via histological analysis of nasopharyngeal and lung tissue and plaque assays of epithelial cells derived from the same tissues. Furthermore, in vivo human studies could utilize healthy volunteers infected with the rhinovirus (“common cold”) via nasal inoculation as described by others (180). This would provide an advantage of studying the efficacy of the host cell PEGylation in humans and follow the progression or the decrease in the severity of the infection. Based on the previously published data, I would expect PEGylation of epithelial cells to be highly effective in attenuating RSV infection. These effects would be optimized at higher grafting concentrations and with the use of longer polymer chains as shown in Figure 4-3D.  166  Figure 4-8: Broad-spectrum antiviral prophylaxis. PEGylation of epithelial cells of the nasal cavity and the nasopharynx could potentially inhibit cell receptor-virus interactions and prevent viral infections that cause common respiratory diseases. Figure courtesy of M. D. Scott.  An important consideration in applying any new technology to be utilized in the clinic is the cost and the ease of preparation of the proposed compound. As demonstrated in Figure 4-9, and discussed in Introduction, polyethylene glycol is widely available in a pure preparation (e.g., PEG-400 and PEG-1000) or as a constituent in pharmaceutical (e.g., Miralax – laxative), cosmetic and food products. Of interest to the PEGylation methodology are activated mPEG formulations that are now commercially available from a number of companies and with the wide variety of linker chemistries and polymer sizes. While industrial grade PEG products (e.g., Union Carbide) are very cheap, activated mPEG is quite expensive. On the other hand, most of the expense in synthesizing activated PEG stems from its scarce utilization restricted to only research and clinical laboratories. Importantly, activated mPEGs are synthesized from very cheap industrial grade PEGs via a multi-step chemical reaction that attaches a linker group to one end of the PEG molecule followed by purification processes, Figure 4-9. Thus, if a large scale production of activated polymers was undertaken to provide these compounds for use in PEGylation of red blood cells, transplanted tissues or as an over-the-counter applicator, this could result in a substantial decrease in the cost of the activated mPEG. As an example, if the cost of 1 g of activated mPEG was decreased to $1, which is 100 times lower than its current price but 1000 times higher than the cost of the industrial grade PEG, PEGylation of one unit of red blood cells would only cost $18! 167  Figure 4-9: The economical considerations for cellular PEGylation. The cost of industrial and pharmaceutical grade PEG compounds is much lower than the cost of the currently available research grade activated mPEG molecules. Importantly, activated mPEG is synthesized from the cheap industrial grade PEGs via a chemical process that adds a linker group to one terminal end of the polymer chain (e.g., SVA). If the current price of the activated mPEG compound is lowered to reflect the prices of raw materials the cost for PEGylation of one unit of red blood cells would only be $18. Figure courtesy of M. D. Scott. Thus, based on the extensive basic research and clinical studies on PEGylated cells and proteins, there is a great potential for this technology to be utilized in the clinic. Owing to the wide availability and affordability of raw compounds for the synthesis of activated mPEG polymers, this technology could be successfully utilized in a number of clinical and pharmaceutical applications. Future studies on PEGylation should aim at practical applications of the direct immunocamouflage of cells in tissue transplantation and in the broad-spectrum antiviral prophylaxis. The initial studies should test the efficacy of PEGylation using in vitro and in vivo models and evaluate the safety and the efficacy of modified products in order to enter the clinical trials.  168  4.6  CONCLUSIONS AND OVERALL SIGNIFICANCE The covalent grafting of mPEG to aliphatic amine polystyrene latex particles imparted a  significant surface immunocamouflaging effect that has helped delineate the mechanisms of biological camouflage seen with labile PEGylated red blood cells, lymphocytes and viruses. The biophysical model identified hydrodynamic shielding and steric repulsion mechanisms as being responsible for surface charge camouflage and inhibition of surface-macromolecule interactions, respectively. The same mechanisms were important in immunocamouflage of cell surfaces as was indicated by surface charge camouflage of RBCs and inhibition of antibody and cell-cell interactions in the PBMC model. Antibody and cell-mediated mechanisms of allorecognition are highly dependent on the receptor-ligand interactions that involve multiple non-covalent forces and non-specific adsorption as well as specific shape “fitting”. Importantly these interactions were perturbed due to: 1) camouflage of surface charge, essential in maintaining non-covalent binding forces; and 2) steric repulsion by surface grafted mPEG resulting in inhibition of close approach and “fitting” of the receptor and the ligand. Moreover, the biophysical latex model defined the chemical and physical properties of the polymers and targets for the efficient surface camouflage. It was shown that the use of monofunctional, linear, flexible chains of mPEG polymer possessing linker chemistries of extended hydrolysis half-life were desirable for the efficient and uniform surface coverage and polymer grafting. It was also demonstrated that the surface charge camouflage of latex particles was best achieved with long chain polymers at sufficient surface coverage irrespective of the size of the target. These results correlated with the observations that the surface charge camouflage of RBCs was best achieved with long chain polymers at high grafting concentrations. Surprisingly to us, the biophysical latex studies indicated that the size of the target dictates the optimal polymer length for the inhibition of surface-macromolecule interactions, i.e., the efficiency of the steric exclusion effect. Thus, it was demonstrated that the inhibition of 169  plasma protein adsorption on small 1.2 !m particles was best achieved with the short chain polymers, whereas the inhibition of protein adsorption on large 8.0 !m particles was best achieved with the long mPEG polymers. Importantly, this finding was in agreement with the biological studies utilizing a virus model (RSV). Efficient inhibition of virus-receptor interactions was achieved when small viruses were PEGylated with short polymers or large host cells were PEGylated with large polymers, Figure 4-3B, D. Furthermore, the same polymer-target relationships were also seen with the PBMCs of 10 !m in diameter. Inhibition of antibody binding to CD antigens on PBMCs and T cell:antigen presenting cell interactions were best achieved with long chain mPEGs. In contrast to the immunocamouflage of latex particles and the simplified surface of the virus, the efficacy of CD antigen camouflage was dependent on the height of the antigen, the overall topography of the cell surface and the localization of the antigen in relation to other cellular structures, e.g., surface proteins or carbohydrate moieties. Importantly, these studies indicated that the surface receptors involved in the adhesion, antigen recognition and binding, and co-stimulatory events of the T cell and APC interactions during the allorecognition process could be efficiently camouflaged. Moreover, PBMC viability studies demonstrated that PEGylation did not result in any differential acute or chronic cellular toxicity at immunoprotective levels. The detailed analysis of the latex surface adsorbed proteins, in iTRAQ/MS studies, also indicated that PEGylated surfaces became more biocompatible owing to surface charge camouflage, sterical inhibition of protein adsorption and increased hydrophilicity and architectural complexity of modified surfaces. This is of special significance for the camouflage of allogeneic tissues in transplantation as PEGylation may not only inhibit antibody and cellmediated pathways of allorecognition but could also prevent the passive binding of complement and coagulation factors important in eliciting protective host immune responses.  170  In support of my hypothesis, both biophysical and biological studies demonstrated that hydrodynamic and steric effects imparted by covalently grafted mPEG chains, resulted in the efficient surface charge camouflage and prevention of surface-macromolecule interactions. These effects were dependent on physical and chemical properties of the polymer and the size of the target. Importantly, hydrodynamic charge shielding and steric repulsion mechanisms were involved in the inhibition of both antibody and cell mediated mechanisms of allorecognition. Thus, this work helped define the biophysical mechanisms responsible for the immunocamouflage of PEGylated surfaces and factors affecting its efficacy. Moreover, this study found that the target size is a crucial determinant in optimizing PEGylation protocols for biological samples. I believe this knowledge is essential in designing a suitable PEGylation technology for cellular camouflage with the aim of preventing unwanted immune consequences in response to transplanted allogeneic tissues.  171  REFERENCES 1.  Landsteiner K. Über Agglutinationserscheinungen normalen menschlichen Blutes. Wien Klin Wochenschr. 1901;14:1132-1134.  2.  Landsteiner K. On agglutination of normal human blood. Transfusion. 1961;1:5-8.  3.  Ottenberg R. Studies in isoagglutination: I. Transfusion and the question of intravascular agglutination. Journal of Experimental Medicine. 1911;13:425-438.  4.  Lewinsohn R. Blood transfusion by the citrate method. Surgical Gynecology and Obstetrics. 1915;21:37-47.  5.  Rous P, Turner JR. Preservation of living red blood corpuscles in vitro. II. The transfusion of kept cells. Journal of Experimental Medicine. 1916;23:219.  6.  Landsteiner K, Wiener AS. An agglutinable factor in human blood recognized by immune sera for rhesus blood. Proc Soc Exp Biol Med. 1940;43:223.  7.  Loeb L. Transplantation and individuality. Biological Bulletin. 1921;40:143-180.  8.  Boyd E. Skin transplantation in the mouse and its effect on pigmentation. Experimental Physiology. 1932;21:337.  9.  Snell GD. Methods for the study of histocompatibility genes. J Genet. 1948;49:87-108.  10.  Dausset J, Nenna A. Presence of leuko-agglutinin in the serum of a case of chronic agranulocytosis. C R Seances Soc Biol Fil. 1952;146:1539-1541.  11.  Billingham RE, Medawar PB. The technique of free skin grafting in mammals. Journal of Experimental Biology. 1951;28:385.  12.  Merrill JP, Murray JE, Harrison JH, Friedman EA, Dealy Jr JB, Dammin GJ. Successful homotransplantation of the kidney between nonidentical twins. N Engl J Med. 1960;262:1251-1260.  172  13.  Daniels GL, Fletcher A, Garratty G et al. Blood group terminology 2004: from the International Society of Blood Transfusion committee on terminology for red cell surface antigens. Vox Sang. 2004;87:304-316.  14.  Reid ME, Yahalom V. Blood groups and their function. Baillieres Best Pract Res Clin Haematol. 2000;13:485-509.  15.  Daniels G. Human blood groups: introduction, terminology, and function. Oxford, UK: Blackwell Science Ltd.; 2002:1.  16.  Pathak S, Palan U. Immunology: Essential and fundamental. New Hampshire: Science Publishers, Inc.; 2005:288.  17.  Daniels G. Functions of red cell surface proteins. Vox Sang. 2007;93:331-340.  18.  Choo SY. The HLA system: genetics, immunology, clinical testing, and clinical implications. Yonsei Med J. 2007;48:11-23.  19.  Williams A, Peh CA, Elliott T. The cell biology of MHC class I antigen presentation. Tissue Antigens. 2002;59:3-17.  20.  Villadangos JA. Presentation of antigens by MHC class II molecules: getting the most out of them. Mol Immunol. 2001;38:329-346.  21.  Janeway CA, Travers P, Walport M, Shlomchik MJ. Immunobiology: The immune system in health and disease. New York: Garland Science Publishing; 2005  22.  Janeway CA, Travers P, Walport M, Shlomchik MJ. Immunobiology: The immune system in health and disease. New York: Garland Science Publishing; 2005:p. 18.  23.  Gibbs F. Adverse effects of transfusion. In: D. QE, editor. Immunohematology. Principles and practice. Philadelphia, Pennsylvania: J. B. Lippincott Company; 1993. p. 235-246.  24.  Dallman MJ. Immunobiology of graft rejection. In: Thiru S, Waldmann H, editors. Pathology and immunology of transplantation and rejection. Oxford, UK: Blackwell Science Ltd.; 2001. p. 1-19. 173  25.  Morrissey PE, Gohh RY, Monaco AP. Allorecognition and tissue typing in organ transplantation. In: Zbar AP, Guillou PJ, Bland KI, Syrigos KN, editors. Immunology for Surgeons. London: Springer-Verlag; 2002. p. 95-125.  26.  Fung JJ, Bradley JA. Clinical transplantation and the immunology or organ rejection. In: Zbar AP, Guillou PJ, Bland KI, Syrigos KN, editors. Immunology for surgeons. London: Springer-Verlag; 2002. p. 155-164.  27.  Wang JH, Reinherz EL. Structural basis of T cell recognition of peptides bound to MHC molecules. Mol Immunol. 2002;38:1039-1049.  28.  Zamoyska R. CD4 and CD8: modulators of T-cell receptor recognition of antigen and of immune responses? Curr Opin Immunol. 1998;10:82-87.  29.  Malissen B, Ardouin L, Lin SY, Gillet A, Malissen M. Function of the CD3 subunits of the pre-TCR and TCR complexes during T cell development. Adv Immunol. 1999;72:103-148.  30.  Pitcher LA, van Oers NS. T-cell receptor signal transmission: who gives an ITAM? Trends Immunol. 2003;24:554-560.  31.  Bromley SK, Burack WR, Johnson KG et al. The immunological synapse. Annu Rev Immunol. 2001;19:375-396.  32.  Montoya MC, Sancho D, Vicente-Manzanares M, Sanchez-Madrid F. Cell adhesion and polarity during immune interactions. Immunol Rev. 2002;186:68-82.  33.  Bour-Jordan H, Blueston JA. CD28 function: a balance of costimulatory and regulatory signals. J Clin Immunol. 2002;22:1-7.  34.  Andreasen SO, Christensen JE, Marker O, Thomsen AR. Role of CD40 ligand and CD28 in induction and maintenance of antiviral CD8+ effector T cell responses. J Immunol. 2000;164:3689-3697.  174  35.  Harrington LE, Hatton RD, Mangan PR et al. Interleukin 17-producing CD4+ effector T cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat Immunol. 2005;6:1123-1132.  36.  Mosmann TR, Coffman RL. TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol. 1989;7:145-173.  37.  Schwartz RH. T cell anergy. Annu Rev Immunol. 2003;21:305-334.  38.  Duffield JS. The inflammatory macrophage: a story of Jekyll and Hyde. Clin Sci (Lond). 2003;104:27-38.  39.  Benichou G, Thomson AW. Direct versus indirect allorecognition pathways: on the right track. Am J Transplant. 2009;9:655-656.  40.  Cote I, Rogers NJ, Lechler RI. Allorecognition. Transfus Clin Biol. 2001;8:318-323.  41.  Caballero A, Fernandez N, Lavado R, Bravo MJ, Miranda JM, Alonso A. Tolerogenic response: allorecognition pathways. Transpl Immunol. 2006;17:3-6.  42.  Game DS, Lechler RI. Pathways of allorecognition: implications for transplantation tolerance. Transpl Immunol. 2002;10:101-108.  43.  Lechler RI, Batchelor JR. Restoration of immunogenicity to passenger cell-depleted kidney allografts by the addition of donor strain dendritic cells. J Exp Med. 1982;155:31-41.  44.  Cramer DV, Qian SQ, Harnaha J et al. Cardiac transplantation in the rat. I. The effect of histocompatibility differences on graft arteriosclerosis. Transplantation. 1989;47:414-419.  45.  Wever PC, Boonstra JG, Laterveer JC et al. Mechanisms of lymphocyte-mediated cytotoxicity in acute renal allograft rejection. Transplantation. 1998;66:259-264.  46.  Dvorak HF, Mihm MCJ, Dvorak AM, Barnes BA, Manseau EJ, Galli SJ. Rejection of firstset skin allografts in man. the microvasculature is the critical target of the immune response. J Exp Med. 1979;150:322-337.  175  47.  Luban NL. Transfusion safety: Where are we today? Ann N Y Acad Sci. 2005;1054:325341.  48.  Semple JW, Speck ER, Cosgrave D, Lazarus AH, Blanchette VS, Freedman J. Extreme leukoreduction of major histocompatibility complex class II positive B cells enhances allogeneic platelet immunity. Blood. 1999;93:713-720.  49.  Martin PJ. The role of donor lymphoid cells in allogeneic marrow engraftment. Bone Marrow Transplant. 1990;6:283-289.  50.  Cornelissen JJ, Lowenberg B. Developments in T-cell depletion of allogeneic stem cell grafts. Curr Opin Hematol. 2000;7:348-352.  51.  Small TN. Immunologic reconstitution following stem cell transplantation. Curr Opin Hematol. 1996;3:461-465.  52.  Marik PE, Corwin HL. Efficacy of red blood cell transfusion in the critically ill: a systematic review of the literature. Crit Care Med. 2008;36:2667-2674.  53.  Torpey N, Bradley JA, Fung JJ. Immunosuppressive therapy in solid organ transplantation. In: Zbar AP, Guillou PJ, Bland KI, Syrigos KN, editors. Immunology for surgeons. London, UK: Springer-Verlag London Limited; 2002. p. 127-154.  54.  Hopkins PMA. Pharmacological manipulation of the rejection response. In: Hornick P, Rose M, editors. Transplantation immunology. Methods and Protocols. Totowa, New Jersey: Humana Press; p. 375-391.  55.  Jain A, Reyes J, Kashyap R et al. What have we learned about primary liver transplantation under tacrolimus immunosuppression? Long-term follow-up of the first 1000 patients. Ann Surg. 1999;230:441-8; discussion 448-9.  56.  Abu-Nader R, Patel R. Current management strategies for the treatment and prevention of cytomegalovirus infection in solid organ transplant recipients. BioDrugs. 2000;13:159-175.  176  57.  Ojo AO, Hanson JA, Wolfe RA, Leichtman AB, Agodoa LY, Port FK. Long-term survival in renal transplant recipients with graft function. Kidney Int. 2000;57:307-313.  58.  Cockfield SM, Preiksaitis J, Harvey E et al. Is sequential use of ALG and OKT3 in renal transplants associated with an increased incidence of fulminant posttransplant lymphoproliferative disorder? Transplant Proc. 1991;23:1106-1107.  59.  Abuchowski A, McCoy JR, Palczuk NC, van Es T, Davis FF. Effect of covalent attachment of polyethylene glycol on immunogenicity and circulating life of bovine liver catalase. J Biol Chem. 1977;252:3582-3586.  60.  Abuchowski A, van Es T, Palczuk NC, Davis FF. Alteration of immunological properties of bovine serum albumin by covalent attachment of polyethylene glycol. J Biol Chem. 1977;252:3578-3581.  61.  Hershfield MS, Buckley RH, Greenberg ML et al. Treatment of adenosine deaminase deficiency with polyethylene glycol-modified adenosine deaminase. N Engl J Med. 1987;316:589-596.  62.  Hershfield MS. Biochemistry and Immunology of Poly(ethylene glycol)-Modified Adenosine Deaminase (PEG-ADA). In: Harris MJ, Zlipsky S, editors. Poly(ethylene glycol). Chemistry and Biological Applications. Washington, DC: American Chemical Society; 1997. p. 145.  63.  Booth C, Gaspar HB. Pegademase bovine (PEG-ADA) for the treatment of infants and children with severe combined immunodeficiency (SCID). Biologics. 2009;3:349-358.  64.  Greenwald RB, Choe YH, McGuire J, Conover CD. Effective drug delivery by PEGylated drug conjugates. Adv Drug Deliv Rev. 2003;55:217-250.  65.  Greenwald RB. PEG drugs: an overview. J Control Release. 2001;74:159-171.  177  66.  Hooftman G, Herman S, Schacht E. Review: Poly(ethylene glycol)s with reactive endgroups. II. Practical consideration for the preparation of protein-PEG conjugates. Journal of Bioactive and Compativle Polymers. 1996;11:135-156.  67.  Delgado C, Francis GE, Fisher D. The Uses and Properties of PEG-Linked Proteins. Critical Reviews in Therapeutic Drug Carrier Systems. 1992;9:249-304.  68.  Kinstler O, Molineux G, Treuheit M, Ladd D, Gegg C. Mono-N-terminal poly(ethylene glycol)-protein conjugates. Adv Drug Deliv Rev. 2002;54:477-485.  69.  Parveen S, Sahoo SK. Nanomedicine: clinical applications of polyethylene glycol conjugated proteins and drugs. Clin Pharmacokinet. 2006;45:965-988.  70.  Roberts MJ, Bentley MD, Harris JM. Chemistry for peptide and protein PEGylation. Adv Drug Deliv Rev. 2002;54:459-476.  71.  Brooks DE, Haynes CA, Hritcu D, Steels BM, Muller W. Size exclusion chromatography does not require pores. Proc Natl Acad Sci U S A. 2000;97:7064-7067.  72.  Hermans J. Excluded-volume theory of polymer-protein interactions based on polymer chain statistics. J Chem Phys. 1982;77:2193-2203.  73.  Nagaoka S, Mori Y, Yokota K, Tanzawa H, Nishiumi S. In: Shalaby SW, Hoffman AS, Ratner BD, Horbett TA, editors. Polymers as biomaterials. New York: Plenum Press; 1985. p. 361.  74.  Szleifer I. Protein adsorption on surfaces with grafted polymers: a theoretical approach. Biophys J. 1997;72:595-612.  75.  Mori Y, Nagaoka S, Takiuchi H et al. A new antithrombogenic material with long polyethyleneoxide chains. Trans Am Soc Artif Intern Organs. 1982;28:459-463.  76.  Satulovsky J, Carignano MA, Szleifer I. Kinetic and thermodynamic control of protein adsorption. Proc Natl Acad Sci U S A. 2000;97:9037-9041.  178  77.  Alcantar NA, Aydil ES, Israelachvili JN. Polyethylene glycol-coated biocompatible surfaces. J Biomed Mater Res. 2000;51:343-351.  78.  Price ME, Cornelius RM, Brash JL. Protein adsorption to polyethylene glycol modified liposomes from fibrinogen solution and from plasma. Biochim Biophys Acta. 2001;1512:191-205.  79.  Ratner BD, Bryant SJ. Biomaterials: where we have been and where we are going. Annu Rev Biomed Eng. 2004;6:41-75.  80.  Merrill EW. Poly(Ethylene Oxide) and Blood Contact. A Chronicle of One Laboratory. In: Harris MJ, editor. Poly(ethylene glycol) chemistry. Biotechnical and Biomedical Applications. New York: Plenum Press; 1992. p. 199.  81.  Park KD, Kim SW. PEO-Modified Surfaces-In Vitro, Ex Vivo, and In Vivo Blood Compatibility. In: Harris MJ, editor. Poly(ethylene glycol) chemistry. Biotechnical and Biomedical Applications. New York: Plenum Press; 1992. p. 283.  82.  Emoto K, Harris JM, Van Alstine JM. Electrokinetic analysis of poly(ethylene glycol) coating chemistry. In: Harris JM, Zalipsky S, editors. ACS Symposium series 680. Washington, DC, USA: American Chemical Society; 1997. p. 374-399.  83.  Van Alstine JM, Harris JM, Shafer S, Snyder RS, Herren B. Polymer-coated surfaces to control surface zeta potential. Universities Space Research Association. 1987;USA  84.  Burns NL, Van Alstine JM, Harris JM. Poly (ethylene glycol) grafted to quartz: Analysis in terms of a site-dissociation model of electroosmotic fluid flow. Langmuir. 1995;11:27682776.  85.  Van Alstine JM, Burns NL, Riggs JA, Holmberg K, Harris JM. Electrokinetic characterization of hydrophilic polymer coatings of biotechnical significance. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 1993;77:149-158.  179  86.  Hamidi M, Azadi A, Rafiei P. Pharmacokinetic consequences of pegylation. Drug Deliv. 2006;13:399-409.  87.  de Gennes PG. Conformation of Polymers Attached to an Interface. Macromolecules. 1980;13:1069-1075.  88.  Milner ST. Polymer Brushes. Science. 1991;251:905-914.  89.  Jeppesen C, Wong JY, Kuhl TL et al. Impact of polymer tether length on multiple ligandreceptor bond formation. Science. 2001;293:465-468.  90.  Kenworthy AK, Hristova K, Needham D, McIntosh TJ. Range and magnitude of the steric pressure between bilayers containing phospholipids with covalently attached poly(ethylene glycol). Biophys J. 1995;68:1921-1936.  91.  Kubetzko S, Sarkar CA, Pluckthun A. Protein PEGylation decreases observed target association rates via a dual blocking mechanism. Mol Pharmacol. 2005;68:1439-1454.  92.  Harris JM, Case MG, Paley MS. Synthesis and Characterization of Poly(ethylene glycol) derivatives. Journal of Polymer Science: Polymer Chemistry Edition. 1984;22:341-352.  93.  Doolittle RF. Redundancies in protein sequences. In: Fasman GD, editor. Prediction of protein structure and the principles of protein conformation. N.Y.: Plenum Press; 1989. p. 599-623.  94.  Veronese FM. Peptide and protein PEGylation: a review of problems and solutions. Biomaterials. 2001;22:405-417.  95.  Nelson DL, Cox MM. Lehninger principles of biochemistry. N.Y.: Worth Publishers; 2000:115.  96.  Yamaoka T, Tabata Y, Ikada Y. Distribution and tissue uptake of poly(ethylene glycol) with different molecular weights after intravenous administration to mice. J Pharm Sci. 1994;83:601-606.  180  97.  Friman S, Egestad B, Sjovall J, Svanvik J. Hepatic excretion and metabolism of polyethylene glycols and mannitol in the cat. J Hepatol. 1993;17:48-55.  98.  Kawai F. Microbial degradation of polyethers. Appl Microbiol Biotechnol. 2002;58:30-38.  99.  Mehvar R. Modulation of the pharmacokinetics and pharmacodynamics of proteins by polyethylene glycol conjugation. J Pharm Pharm Sci. 2000;3:125-136.  100. Webster R, Didier E, Harris P et al. PEGylated proteins: evaluation of their safety in the absence of definitive metabolism studies. Drug Metab Dispos. 2007;35:9-16. 101. Herold DA, Keil K, Bruns DE. Oxidation of polyethylene glycols by alcohol dehydrogenase. Biochem Pharmacol. 1989;38:73-76. 102. Smyth HF, Carpenter CP, Weil CS. The toxicology of the polyethylene glycols. J Am Pharm Assoc Am Pharm Assoc. 1950;39:349-354. 103. Carpenter CP, Woodside MD, Kinkead ER, King JM, Sullivan LJ. Response of dogs to repeated intravenous injection of polyethylene glycol 4000 with notes on excretion and sensitization. Toxicol Appl Pharmacol. 1971;18:35-40. 104. Fruijtier-Polloth C. Safety assessment on polyethylene glycols (PEGs) and their derivatives as used in cosmetic products. Toxicology. 2005;214:1-38. 105. Richter AW, Akerblom E. Antibodies against polyethylene glycol produced in animals by immunization with monomethoxy polyethylene glycol modified proteins. Int Arch Allergy Appl Immunol. 1983;70:124-131. 106. Richter AW, Akerblom E. Polyethylene glycol reactive antibodies in man: titer distribution in allergic patients treated with monomethoxy polyethylene glycol modified allergens or placebo, and in healthy blood donors. Int Arch Allergy Appl Immunol. 1984;74:36-39. 107. Sroda K, Rydlewski J, Langner M, Kozubek A, Grzybek M, Sikorski AF. Repeated injections of PEG-PE liposomes generate anti-PEG antibodies. Cell Mol Biol Lett. 2005;10:37-47. 181  108. Armstrong JK, Hempel G, Koling S et al. Antibody against poly(ethylene glycol) adversely affects PEG-asparaginase therapy in acute lymphoblastic leukemia patients. Cancer. 2007;110:103-111. 109. Beckman JS, Minor RLJ, White CW, Repine JE, Rosen GM, Freeman BA. Superoxide dismutase and catalase conjugated to polyethylene glycol increases endothelial enzyme activity and oxidant resistance. J Biol Chem. 1988;263:6884-6892. 110. DeFrees S, Wang ZG, Xing R et al. GlycoPEGylation of recombinant therapeutic proteins produced in Escherichia coli. Glycobiology. 2006;16:833-843. 111. Ramon J, Saez V, Baez R, Aldana R, Hardy E. PEGylated interferon-alpha2b: a branched 40K polyethylene glycol derivative. Pharm Res. 2005;22:1374-1386. 112. Scott MD, Murad KL, Koumpouras F, Talbot M, Eaton JW. Chemical camouflage of antigenic determinants: stealth erythrocytes. Proc Natl Acad Sci U S A. 1997;94:75667571. 113. Scott MD, Murad KL. Cellular camouflage: fooling the immune system with polymers. Curr Pharm Des. 1998;4:423-438. 114. Murad KL, Mahany KL, Brugnara C, Kuypers FA, Eaton JW, Scott MD. Structural and functional consequences of antigenic modulation of red blood cells with methoxypoly(ethylene glycol). Blood. 1999;93:2121-2127. 115. Scott MD, Bradley AJ, Murad KL. Stealth erythrocytes: effects of polymer grafting on biophysical, biological and immunological parameters. Blood Transfusion. 2003;1:244265. 116. Chen AM, Scott MD. Immunocamouflage: prevention of transfusion-induced graft-versushost disease via polymer grafting of donor cells. J Biomed Mater Res A. 2003;67:626-636. 117. Chen AM, Scott MD. Current and future applications of immunological attenuation via pegylation of cells and tissue. BioDrugs. 2001;15:833-847. 182  118. Chen AM, Scott MD. Comparative analysis of polymer and linker chemistries on the efficacy of immunocamouflage of murine leukocytes. Artif Cells Blood Substit Immobil Biotechnol. 2006;34:305-322. 119. Murad KL, Gosselin EJ, Eaton JW, Scott MD. Stealth cells: prevention of major histocompatibility complex class II-mediated T-cell activation by cell surface modification. Blood. 1999;94:2135-2141. 120. Scott MD, Chen AM. Beyond the red cell: pegylation of other blood cells and tissues. Transfus Clin Biol. 2004;11:40-46. 121. Lee DY, Park SJ, Nam JH, Byun Y. A combination therapy of PEGylation and immunosuppressive agent for successful islet transplantation. J Control Release. 2006;110:290-295. 122. Yun Lee D, Hee Nam J, Byun Y. Functional and histological evaluation of transplanted pancreatic islets immunoprotected by PEGylation and cyclosporine for 1 year. Biomaterials. 2007;28:1957-1966. 123. McCoy LL, Scott MD. Broad-Spectrum Antiviral Porphylaxis: Inhibition of Viral Infection by Polymer Grafting with Methoxypoly(ethylene glycol). In: Torrence PF, editor. Antiviral Drug Discovery for Emerging Diseases and Bioterrorism Threats. John Wiley & Sons, Inc.; 2005. p. 379-395. 124. McCoy LL. Prevention of Viral Infection via Modification of Virus or Cells with Methoxypoly(Ethylene Glycol) [dissertation]. Vancouver: UBC; 2005. 125. Mizouni SK, Lehman JM, Cohen B, Scott MD. Viral modification with methoxypoly(ethylene glycol): Implications for gene therapy and viral inactivation. Blood. 1998;92:4627. 126. McCoy JR, Scott MD. Prevention of viral invasion by immunocamouflage of target cells. Antiviral Research. 1998;62:67. 183  127. McCoy L, Scott MD. Prevention of viral invasion by immunocamouflage of target cells. Antiviral Research. 2004;62:67. 128. Sutton TC, Scott MD. Methoxypoly(ethylene glycol) modification of viruses or host cells: A broad spectrum antiviral prophylaxis. Transfusion. 2008;48:106A. 129. Sutton TC. Prevention of respiratory syncytial virus infection via methoxypoly(ethylene glycol)-modification of the virus or its host cell. [dissertation]. Vancouver, BC: University of British Columbia; 2009. 130. Bradley AJ, Murad KL, Regan KL, Scott MD. Biophysical consequences of linker chemistry and polymer size on stealth erythrocytes: size does matter. Biochim Biophys Acta. 2002;1561:147-158. 131. Bradley AJ, Test ST, Murad KL, Mitsuyoshi J, Scott MD. Interactions of IgM ABO antibodies and complement with methoxy-PEG-modified human RBCs. Transfusion. 2001;41:1225-1233. 132. Bradley AJ, Scott MD. Separation and purification of methoxypoly(ethylene glycol) grafted red blood cells via two-phase partitioning. J Chromatogr B Analyt Technol Biomed Life Sci. 2004;807:163-168. 133. Bradley AJ, Scott MD. Immune complex binding by immunocamouflaged [poly(ethylene glycol)-grafted] erythrocytes. Am J Hematol. 2007;82:970-975. 134. Zalipsky S. Chemistry of polyethylene glycol conjugates with biologically active molecules. Adv Drug Deliv Rev. 1995;16:157-182. 135. Levine S, Levine M, Sharp KA, Brooks DE. Theory of the electrokinetic behavior of human erythrocytes. Biophys J. 1983;42:127-135. 136. Vroman L, Leonard EF. The behavior of blood and its components at interfaces. Ann NY Acud Sci. 1977;283:6.  184  137. Brash JL. Exploiting the current paradigm of blood-material interactions for the rational design of blood-compatible materials. J Biomater Sci Polym Ed. 2000;11:1135-1146. 138. Ratner BD. Blood compatibility-a perspective. J Biomater Sci Polym Ed. 2000;11:11071119. 139. Anderson JM. Biological responses to materials. Annual Review of Materials Research. 2001;31:81-110. 140. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265-275. 141. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680-685. 142. Ross PL, Huang YN, Marchese JN et al. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol Cell Proteomics. 2004;3:1154-1169. 143. Janzen J, Song X, Brooks DE. Interfacial thickness of liposomes containing poly (ethylene glycol)-cholesterol from electrophoresis. Biophysical journal. 1996;70:313-320. 144. Sun ZYJ, Kim ST, Kim IC, Fahmy A, Reinherz EL, Wagner G. Solution structure of the CD3 ectodomain and comparison with CD3 as a basis for modeling T cell receptor topology and signaling. PNAS. 2004;101:16867-16872. 145. Barclay AN, Brown MH, Law SKA, McKnight AJ, Tomlinson MG, Van der Merwe PA. The leucocyte antigen. San Diego: Academic Press; 1997:101. 146. Shaw AS, Dustin ML. Making the T cell receptor go the distance: a topological view of T cell activation. Immunity. 1997;6:361-370. 147. Bronner-Fraser M. Alterations in neural crest migration by a monoclonal antibody that affects cell adhesion. J Cell Biol. 1985;101:610-617.  185  148. Hayry P, Defendi V. Mixed lymphocyte cultures produce effector cells: model in vitro for allograft rejection. Science. 1970;168:133-135. 149. Reinsmoen NL, Kaufman D, Matas A, Sutherland DE, Najarian JS, Bach FH. A new in vitro approach to determine acquired tolerance in long-term kidney allograft recipients. Transplantation. 1990;50:783-790. 150. Martin SJ, Reutelingsperger CP, McGahon AJ et al. Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl. J Exp Med. 1995;182:1545-1556. 151. Schmid I, Krall WJ, Uittenbogaart CH, Braun J, Giorgi JV. Dead cell discrimination with 7-amino-actinomycin D in combination with dual color immunofluorescence in single laser flow cytometry. Cytometry. 1992;13:204-208. 152. Seaman GV, Knox RJ, Nordt FJ, Regan DH. Red cell agins. I. Surface charge density and sialic acid content of density-fractionated human erythrocytes. Blood. 1977;50:1001-1011. 153. Goldsby RA, Kindt TJ, Osborne BA, Kuby J. Immunology. New York: W. H. Freeman and Company; 2003:212. 154. Nemazee DA, Burki K. Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes. Nature. 1989;337:562-566. 155. Russell DM, Dembic Z, Morahan G, Miller JF, Burki K, Nemazee D. Peripheral deletion of self-reactive B cells. Nature. 1991;354:308-311. 156. Hartley SB, Crosbie J, Brink R, Kantor AB, Basten A, Goodnow CC. Elimination from peripheral lymphoid tissues of self-reactive B lymphocytes recognizing membrane-bound antigens. Nature. 1991;353:765-769. 157. Luck M, Paulke BR, Schroder W, Blunk T, Muller RH. Analysis of plasma protein adsorption on polymeric nanoparticles with different surface characteristics. J Biomed Mater Res. 1998;39:478-485. 186  158. Parzer S, Balcke P, Mannhalter C. Plasma protein adsorption to hemodialysis membranes: studies in an in vitro model. J Biomed Mater Res. 1993;27:455-463. 159. Babensee JE, Cornelius RM, Brash JL, Sefton MV. Immunoblot analysis of proteins associated with HEMA-MMA microcapsules: human serum proteins in vitro and rat proteins following implantation. Biomaterials. 1998;19:839-849. 160. Meng F, Engbers GH, Feijen J. Polyethylene glycol-grafted polystyrene particles. J Biomed Mater Res A. 2004;70:49-58. 161. Lee JH. Plasma protein adsorption and platelet adhesion onto comb-like PEO gradient surfaces. Journal of Biomedical Materials Research. 1997;34:105-114. 162. Fleischer J, Grage-Griebenow E, Kasper B et al. Platelet factor 4 inhibits proliferation and cytokine release of activated human T cells. J Immunol. 2002;169:770-777. 163. Ehlert JE, Ludwig A, Grimm TA, Lindner B, Flad HD, Brandt E. Down-regulation of neutrophil functions by the ELR(+) CXC chemokine platelet basic protein. Blood. 2000;96:2965-2972. 164. Vroman L. Methods of investigating protein interactions on artificial and natural surfaces. Ann N Y Acad Sci. 1987;516:300-305. 165. Yasmeen D, Ellerson JR, Dorrington KJ, Painter RH. The structure and function of immunoglobulin domains. IV. The distribution of some effector functions among the Cgamma2 and Cgamma3 homology regions of human immunoglobulin G1. J Immunol. 1976;116:518-526. 166. Colomb M, Porter RR. Characterization of a plasmin-digest fragment of rabbit immunoglobulin gamma that binds antigen and complement. Biochem J. 1975;145:177183.  187  167. Reid KB. Complete amino acid sequences of the three collagen-like regions present in subcomponent C1q of the first component of human complement. Biochem J. 1979;179:367-371. 168. Lawler J, Duquette M, Urry L, McHenry K, Smith TF. The evolution of the thrombospondin gene family. J Mol Evol. 1993;36:509-516. 169. Miyakis S, Giannakopoulos B, Krilis SA. Beta 2 glycoprotein I--function in health and disease. Thromb Res. 2004;114:335-346. 170. Brighton TA, Hogg PJ, Dai YP, Murray BH, Chong BH, Chesterman CN. Beta 2glycoprotein I in thrombosis: evidence for a role as a natural anticoagulant. Br J Haematol. 1996;93:185-194. 171. Doyen V, Rubio M, Braun D et al. Thrombospondin 1 is an autocrine negative regulator of human dendritic cell activation. J Exp Med. 2003;198:1277-1283. 172. Li Z, He L, Wilson K, Roberts D. Thrombospondin-1 inhibits TCR-mediated T lymphocyte early activation. J Immunol. 2001;166:2427-2436. 173. Harris JM. Introduction to bioetchnical and biomedical applications of poly(ethylene glycol). In: Harris MJ, editor. Topics in applied chemistry. New York, N. Y., USA: Plenum Press; 1992. p. 1-14. 174. Andrade JD. Interfacial phenomena and biomaterials. Med Instrum. 1973;7:110-119. 175. Vroman L. Effect of absorbed proteins on the wettability of hydrophilic and hydrophobic solids. Nature. 1962;196:476-477. 176. Horbett TA. Mass action effects on competitive adsorption of fibrinogen from hemoglobin solutions and from plasma. Thromb Haemost. 1984;51:174-181. 177. Eylar EH, Madoff MA, Brody OV, Oncley JL. The contribution of sialic acid to the surface charge of the erythrocyte. J Biol Chem. 1962;237:1992-2000.  188  178. Lakey JRT, Mirbolooki M, Shapiro AMJ. Current status of clinical islet cell transplantation. In: Hornick P, Rose M, editors. Transplantation immunology: Methods and protocols. Totowa, NJ: Humana Press Inc.; 2006. p. 47-103. 179. Baribault H. Mouse Models of Type II Diabetes Mellitus in Drug Discovery. Methods Mol Biol. 2010;602:135-155. 180. Doyle WJ, McBride TP, Skoner DP, Maddern BR, Gwaltney JMJ, Uhrin M. A doubleblind, placebo-controlled clinical trial of the effect of chlorpheniramine on the response of the nasal airway, middle ear and eustachian tube to provocative rhinovirus challenge. Pediatr Infect Dis J. 1988;7:229-238.  189  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0069210/manifest

Comment

Related Items