.1 .f 3.. a a. ”(n5 . I. . .3 ¢ .N. am. an it! :1... e 35.x...1. H3... kg...) . . .:. Iv: a ‘ V , V. , , . . . a. $35k“ .fl .5. V . . , . 4 £1! . wry» .. 2. L iv: 0.. 1.. Jr 3. .. .2 .1 .9 .an a??? 7.1 J“ 11: § um . .c .r 19., 3! am .4 . 1.4!... u.” . ' 5.....- 1h: J.) z . a.‘ vi . v . .14....umhfia . “ was. .5 naivgw: y .2 o“) This is to certify that the dissertation entitled DEVELOPMENT, VALIDATION, AND APPLICATION OF A METHOD TO CHARACTERIZE MAJOR HISTOCOMPATIBILITY COMPLEX ASSOCIATED PEPTIDES FROM EQUIVALENT T0 1 x 108 CELLS presented by Leann Michelle Hopkins has been accepted towards fulfillment of the requirements for Ph.D. degreein Pathobiology and Diagnostic I Investigation i fora/jor professor Date 12/ 2A a ’ I 012771 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/CIRC/DaieDuo.965~p. 15 DEVELOPMENT, VALIDATION, AND APPLICATION OF A METHOD TO CHARACTERIZE MAJOR HISTOCOMPATIBILITY COMPLEX ASSOCIATED PEPTIDES FROM EQUIVALENT TO 1 X 108 CELLS BY Leann Michelle Hopkins A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pathobiology and Diagnostic Investigation 2002 ABSTRACT DEVELOPMENT, VALIDATION, AND APPLICATION OF A METHOD TO CHARACTERIZE MAJOR HISTOCOMPATIBILITY COMPLEX ASSOCIATED PEPTIDES FROM EQUIVALENT TO 1 X 108 CELLS BY Leann Michelle Hopkins The class I major histocompatibility complex (MHC) presents peptides on almost all cells within the human body. Intracellular proteins are degraded into peptides of 8—11 amino acids allowing them to fit in the groove of an empty MHC class I molecule. Characterization of these MHC associated peptides can be challenging, a major difficulty being the need to obtain peptides in an adequate concentration for detection. Published protocols require a sample size that is impractically large to be clinically realistic. The objective of this research was to establish a method able to characterize MHC associated peptides from a sample of reasonable clinical availability. Based on extrapolations, it should be possible to characterize these peptides from 1 X 108 cells (or 20—30 ml of whole blood). Key to making this work, was a citric acid wash of whole cells release the peptides followed by sample clean-up using RP-HPLC and a peptide trap. Peptides were analyzed by LC—MS/MS. MHC associated peptides from several cellular samples were characterized to validate this method. Peptides from proteins with general cellular function (ubiquitin, talin, actin, and unnamed protein), platelet specific function (thrombospondin, platelet factor 4, and GPIb), and an immune response (MRP—14), were all identified. The most remarkable finding, when the method was applied to a clinical problem, was the purification of GPIb peptide from individuals with Idiopathic Thrombocytopenic Purpura (ITP). ITP is characterized by the premature immune destruction of platelets, and is associated with the production of antiplatelet autoantibodies, most often targeting platelet membrane GPIIb/IIIa and/or GPIb/IX. The GPIb peptide was identified in 4 of the 5 ITP patients (who coincidently all had HLA- B7) and in none of the random controls. To determine if the peptide was specific for the ITP patients or HLA-B7, 3 non-ITP individuals with HLA—B7 were selected as controls. GPIb (4-12) was isolated from 2 of these 3 controls. In conclusion, platelets can present GPIb peptides associated with MHC class I. Since both HLA—B7 ITP and controls present GPIb peptides, development of autoimmunity may not reside with peptide presentation, but with other mechanisms, such as loss of self-tolerance. Copyright by LEANN MICHELLE HOPKINS 2 O 02 To my mother and father For teaching me what I need to know, For giving me what I need to have, And for loving me with all their heart So I could become who I am To my sister A love that cannot be broken ACKNOWLEDGEMENTS I would like to thank Dr. John Gerlach and Dr. Robert Bull for their generosity, support, and guidance over the years. They have provided me a basis for my scientific career and opened the doors for many opportunities yet to come. For this I cannot thank them enough. I would also like to thank committee members Dr. Kenneth Schwartz for access to his ITP patient population, Joe Leykam for his expertise in the fields of HPLC and mass spectrometry, and Dr. Doug Estry, Dr. Charles Mackenzie, Dr. Tim Day and the entire committee for the time they have dedicated to my dissertation. The Immunohematology and Serology Laboratory at Michigan State University has been my home away from home for over 10 years. From the lab, I would like to acknowledge Sue Forney and Ann Gobeski for their assistance and Peggy Bull for teaching me everything I needed to know in (and out) of the laboratory. I would also like to thank Kim Thompson, a summer student in the lab, who purified some of the antibodies used in this research. I am grateful to Dr. Doug Gage, Dr. John Allison and the Mass Spectrometry Facility at Michigan State University for allowing me access to their facility. From the vi facility, I am extremely thankful to Dr. Michael Schall for his invaluable assistance in methodology development for this research. I would like to thank John Davis and the Platelet Antibody Laboratory at Michigan State University for their platelet assays and assistance provided. A special thank you to Dr. Jeff Shabanowitz at the University of Virginia whose instruction in de novo peptide sequencing was instrumental in the data analysis of this project. Lastly, I would like to thank the individuals who graciously donated blood to my dissertation project. Without their willingness to donate, this project could not have been completed. vii TABLE OF CONTENTS LIST OF TABLES ........................................... xii LIST OF FIGURES ........................................... XV LIST OF ABBREVIATIONS ................................... xvii INTRODUCTION ............................................... 1 Literature Review .......................................... 2 PART 1: Antigen Processing and Presentation .............. 3 Cytotoxic T Cells ...................................... 4 Helper T Cells: Th1 and Th2 ........................... 5 CD4+ and CD8+ T Cell Immune Response ................... 7 Role of the TCR ........................................ 8 Antigen Processing ..................................... 9 Endogenous Pathway .................................. 10 Proteasomes ....................................... 11 Transporter Associated with Antigen Processing....18 MHC Peptide Loading ............................... 2O MHC Class I Molecules ............................. 21 Exogenous Pathway ................................... 23 Antigen Presentation .................................. 27 Part 2: Applications of High Performance Liquid Chromatography and Mass Spectrometry to MHC Associated Peptide Identification ................................. 28 Review of Previous Peptide Purification Protocols....28 Acid Wash of Cells ................................... 32 Reverse Phase—High Performance Liquid Chromatography.34 Liquid Chromatography Tandem Mass Spectrometry ....... 35 Part 3: Platelets ..................................... 39 Platelet Development and Function .................... 4O Platelet Proteins .................................... 41 Expression of MHC - A, B, C on Platelets ............. 43 Part 4: Autoimmunity .................................. 44 Part 5: Idiopathic Thrombocytopenia Purpura ........... 46 Objectives and Hypothesis .............................. 51 MATERIALS ................................................ 53 viii Monoclonal Antibodies .................................. 53 Epstein Barr Virus B Cells ............................. 54 ITP and Control Samples ................................ 55 METHODS: Antibody and Sample Preparation ................. 57 1. Preparation of Monoclonal Antibodies and Samples....57 1.1. MAb Preparation ................................ 57 1.1.1. Antibody Production — MAb Hybridoma Cell Culture .................................... 57 1.1.2. Antibody Purification — Protein A Affinity Chromatography of MAb (W6/32, H8116, and L243) ...................................... 58 1.1.3. Validation of Antibody by Complement Dependent Microcytotoxicity Testing ........ 62 1.2. Cells from Cultured Cell Lines ................. 65 1.2.1. Cell Culture of EBV B Cell Line ............ 65 1.2.2. EBV B cell Sample Preparation for MHC purification ............................... 67 1.3. Cells from Whole Blood from Donors ............. 67 1.3.1. Platelet Sample Preparation ................ 68 1.3.2. White Blood Cell Sample Preparation ........ 69 1.3.3. Red Blood Cell Sample Preparation .......... 71 1.3.4 Molecular MHC Class I Typing ............... 71 METHODS: MHC Associated Peptide Characterization ........ 75 2. Affi-gel®-10 Affinity Chromatography ............... 75 2.1. Affi—gel®-1O HLA Column Preparation ............ 75 2.2. Affi-gel®-1O Affinity Chromatography - Purification of HLA ............................ 76 2.3. SDS-PAGE ....................................... 78 2.3.1. Tricine—SDS-PAGE ............................ 78 2.2.2. Silver Staining ............................. 81 3. Immunomagnetic Chromatography ...................... 83 3.1. Preparation of Magnetic Beads — Binding MAb to Beads ........................................... 83 3.2. Magnetic Bead Immunoaffinity Purification of MHC .......................................... 84 3.3. Purification of MHC associated Peptides ......... 85 3.3.1. Purification by of Peptides by Molecular Weight Filtration ........................... 86 3.3.2. Purification of Peptides by Analytical RP-HPLC ..................................... 89 3.4. Characterization of Peptides by LC-MS/MS ........ 91 3.4.1. Capillary Liquid Chromatography System ...... 92 ix 3.4.2. Mass Spectrometry .......................... 94 3.4.3. MS Data Analysis ........................... 96 4. Mild Acid Wash to purify peptides from MHC .......... 97 4.1. Extraction of MHC Associated Peptides by a Mild Acid Wash Procedure ........................ 97 4.2. Purification of Peptides Using a Trap Column....98 5. Peptide Verification ................................ 99 RESULTS ................................................. 101 1. Antibody Purification ............................. 101 1.1. Affinity Chromatography ....................... 101 1.1.1. W6/32 Antibody Purification ............... 101 1.1.2. HB116 Antibody Purification ............... 105 1.1.3. L243 Antibody Purification ................ 107 1.2. Microcytotoxicity Testing ..................... 108 1.2.1. W6/32 ..................................... 108 1.2.2. HB116 ..................................... 110 1.2.3. L243 ...................................... 110 1.2.4. H58A, SPA822, and SPA850 .................. 111 2. MHC Purification by Affi—gel®-1O Affinity Chromatography ..................................... 111 2.1 MHC Purification Using W6/32 .................. 111 2.2 MHC Purification Using HB116 .................. 113 2.3 MHC Purification Albumin Control Column ....... 116 2.4 MHC Purification W6/32 and Albumin Column ..... 116 3. MHC Purification by Immunomagnetic Chromatography.118 3.1. Magnetic Bead Cross Linking Protocol .......... 118 3.2. Immunomagnetic MHC and HSP Recovery ........... 119 3.3. Peptide Purification .......................... 119 3.3.1. Peptide Purification by MW Filtration ...... 119 3.3.2. MHC Associated Peptide Purification by Analytical RP-HPLC ........................ 125 4. MHC Associated Peptide Purification by Mild Acid Wash ......................................... 129 5. Identification of Theoretical MHC Class I Peptides ........................................ 163 DISCUSSION .............................................. 173 1. Protein A Antibody Purification ................... 173 2. Affi—gel®-10 columns .............................. 173 3. Immunomagnetic Chromatography ..................... 175 3.1. Peptide Purification by MW filtration ......... 176 3.2. Peptide Purification by Analytical RP—HPLC....181 4. Peptide Purification by Mild Acid Wash with Trap Column ............................................ 183 4.1. Control Platelets ............................. 183 4.2. Control WBCs .................................. 188 4.3. Identified Peptides are Characteristic of MHC Class I Peptides .......................... 189 4.4. ITP Platelets ................................. 190 4.5. Additional Platelet Controls (HLA-B7) ......... 198 SUMMARY AND CONCLUSIONS ................................. 202 FUTURE RECOMMENDATIONS .................................. 207 APPENDIX ................................................ 210 REFERENCES .............................................. 211 xi Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 8a. 8b. 8c. 8d. 8e. 8f. 8g. 8h. 81. 8. LIST OF TABLES Monoclonal Antibodies used in Studies ........... 53 Antibody Purification Yield .................... 103 Complement Mediated Cytotoxicity Testing ....... 109 Yield of MHC class I Purified by Affi-gel®-10..112 Amount of Platelets and WBCs Used .............. 131 HLA type of Individuals ........................ 132 Platelet Autoantibody Status ................... 133 HLA Associated Peptides Identified from Control #1 using citric acid wash ............. 135 HLA Associated Peptides Identified from Control #2 using citric acid wash ............. 145 HLA Associated Peptides Identified from Control #3 using citric acid wash ............. 147 HLA Associated Peptides Identified from ITP #1 using citric acid wash ................. 150 HLA Associated Peptides Identified from ITP #2 using citric acid wash ................. 154 HLA Associated Peptides Identified from ITP #3 using Citric acid wash ................. 156 HLA Associated Peptides Identified from ITP #4 using citric acid wash ................. 157 HLA Associated Peptides Identified from ITP #5 using citric acid wash ................. 158 HLA Associated Peptides Identified from Control #4-6 using citric acid wash ........... 161 HLA Associated Peptides Identified from ITP and controls using citric acid wash ........ 162 xii Table 9. Theoretical Ubiquitin Peptides ................. 164 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 10. ll. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Theoretical Talin Peptides .................... 165 Theoretical Thrombospondin Peptides ........... 166 Theoretical GPIB beta Peptides ................ 167 Theoretical GPIB alpha Peptides ............... 168 Theoretical GPIX Peptides ..................... 169 Theoretical GPIIB alpha Peptides .............. 170 Theoretical GPIIIA Peptides ................... 171 Theoretical Platelet Factor 4 Peptides ........ 172 Individual #1 MHC Associated Peptide Anchor Motifs ........................................ 178 EBV B cell MHC Associated Peptide Anchor Motifs ........................................ 180 EBV B Cell 9035 MHC Associated Peptide Anchor Motifs ........................................ 180 Control #1 MHC Associated Peptide Anchor Motifs ........................................ 185 Control #2 MHC Associated Peptide Anchor Motifs ........................................ 186 Control #3 MHC Associated Peptide Anchor Motifs ........................................ 187 ITP #1 MHC Associated Peptide Anchor Motifs...192 ITP #2 MHC Associated Peptide Anchor Motifs...193 ITP #3 MHC Associated Peptide Anchor Motifs...194 ITP #4 MHC Associated Peptide Anchor Motifs...195 ITP #5 MHC Associated Peptide Anchor Motifs...196 xiii Table 29. Table 30. Table 31. Control #4 MHC Associated Peptide Anchor Motifs ........................................ 199 Control #5 MHC Associated Peptide Anchor Motifs ........................................ 200 Control #6 MHC Associated Peptide Anchor Motifs ........................................ 201 xiv Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10. ll. 12. 13. 14. 15. l6. l7. 18. LIST OF FIGURES Map of MHC Region on Chromosome 6 .............. 16 W6/32 MAb Purified from Ascites ............... 102 W6/32 MAb Purification from Cell Culture ...... 104 SDS-PAGE of commercial vs. cell culture purified MAb .................................. 105 SDS-PAGE of Ascites Purified HB116 MAb ........ 106 SDS—PAGE of Ascites Purified L243 MAb ......... 107 MHC Class I Purification by Affi-gel®-10 ...... 114 SDS-PAGE of HLA class I isolated by Affi-gel®-10 .................................. 115 Albumin Negative Control Column ............... 117 SDS—PAGE of HLA class I Purified by Magnetic Beads ........................................ 121 Purified MHC class II using Magnetic Beads...122 HPLC Chromatograph of MW fractioned MHC class I peptides ............................. 124 RP-HPLC Chromatograph of MW filtered negative control ............................. 126 RP-HPLC Chromatograph of washed MW filtered negative control ............................. 127 Analytical RP-HPLC Chromatograph ............. 128 RP-HPLC Chromatograph with a peptide trap column to purify MHC from peptides ........... 130 CAD spectra of KESTLHLVL, Ubiquitin 63-71....136 CAD spectra of KEVPDACF, thrombospondin 593-600 ...................................... 137 XV Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 19. 20. 21. 22. 23. 24. 25. 26. 27. 28 29. CAD spectra of ASRPGLLF, platelet factor 4, 9-16 ............................... 138 CAD spectra of ALNELLQHV, talin 777—785 ...... 139 CAD Spectra of KESTLHLVL and synthetic peptide ..................................... 142 CAD Spectra of KEVPDACF and synthetic peptide ..................................... 144 CAD spectra of RVAPEEHPVL, y—actin 95-104....146 CAD spectra of ARVEHPFR, unnamed protein 285-293 ...................................... 148 CAD spectra of GPRGALSLL 4—12 GPIb ........... 151 CAD spectra of DTNADKQLSF MRP—14 67—76 ....... 152 CAD spectra of LDTNADKQLSF MRP-14 66—76 ...... 153 CAD spectra of XXX(K/Q)EA(L/I)ERF ............. 155 GPRGALSLL GPIb 4-12 and synthetic ............ 160 xvi LIST OF ABBREVIATIONS American Society for Histocompatibility and Immunogenetics ............................... ASHI American Type Culture Collection ................. ATCC Antibody ......................................... Ab Antigen Presenting Cell .......................... APC ATP Binding Cassette ............................. ABC Bleomycin Hydrolase .............................. BH Bovine Serum Albumin ............................. BSA Chinese Hamster Ovary ............................ CHO Class II Associated Invariant Chain Peptide ...... CLIP Cluster Differentiation .......................... CD Collision Activated Dissociation ................. CAD Cross Correlation Score .......................... Xcorr Cytotoxic T Cell ................................. Tc Cytotoxic T Lymphocyte ........................... CTL Defective Ribosomal Products ..................... DRiPs Deoxyribonucleic Acid ............................ DNA Dulbecco’s Phosphate Buffered Saline ............. DPBS Electrospray Ionization .......................... ESI Enzyme Linked Immunosorbent Assay ................ ELISA Epstein Barr Virus ............................... EBV Ethylenediaminetetracetic acid ................... EDTA Glucose Regulate Protein ......................... GRP Glycoprotein ..................................... GP Heat Shock Cognate ............................... HSC Heat Shock Protein ............................... HSP Helper T Cell .................................... Th Hepatitis B Surface Antigen ...................... HBsAg High Performance Liquid Chromatography ........... HPLC Human Leukocyte Antigen .......................... HLA Idiopathic Thrombocytopenia Purpura .............. ITP Immunoglobulin Binding Protein ................... BiP Immunoglobulin ................................... Ig Insulin Dependent Diabetes Mellitus .............. IDDM Interferon ....................................... IFN Interleukin ...................................... IL Invariant Chain .................................. Ii Leucine Aminopeptidase ........................... LAP Liquid Chromatography Tandem Mass Spectrometry... LC-MS/MS Low Molecular Mass Protease ...................... LMP Lymphotoxin ...................................... LT Macrophage Inflammatory Protein .................. MIP Major Histocompatibility Complex ................. MHC xvii Mass Spectrometry ................................ Mass to Charge Ratio ............................. Matrix Assisted Laser Desorption Ionization ...... MHC Class II Compartment ......................... Molecular Weight Cut Off ......................... Monoclonal Antibody .............................. Nominal Molecular Weight Limit ................... Non—Obese Diabetic .............................. Optical Density .................................. Peripheral Blood Mononuclear Cells .............. Phenylmethyl Sulfonyl Fluoride ................... Phosphate Buffered Saline ........................ Platelet-Endothelial Cell Adhesion Molecule-1.... Polyacrylamide Gel Electrophoresis ............... Polyetheretherketone Polymer ..................... Polymerase Chain Reaction ........................ Polymorphonuclear ................................ Prostaglandin E1 ................................. Puromycin Sensitive Aminopeptidase ............... Red Blood Cell ................................... Reverse Phase .................................... Rheumatoid Arthritis ............................. Sequence Specific Primer ......................... Sodium Dodecyl Sulphate .......................... T Cell Receptor .................................. T Helper Type 1 Cell ............................. T Helper Type 2 Cell ............................. Tandem Mass Spectrometry ......................... Thimet Oligopeptidase ............................ Thromboxane A2 ................................... Transporter Associated with Antigen Processing... Trifluroacetic Acid .............................. Tumor Necrosis Factor ............................ Ultra Violet ..................................... von Willebrand Factor ............................ White Blood Cell ................................. 8-2 Microglobulin ................................ xviii MS m/z MALDI MIIC MWCO MAb NMWL NOD OD PBMC PMSF PBS PECAM-l PAGE PEEK PCR PMN PGE1 PSA RBC RP RA SSP SDS TCR Th1 Th2 MS/MS TOP TXAZ TAP TFA TNF UV vWF WBC 82M Introduction The adaptive immune system communicates via the interaction of the Major Histocompatibility Complex (MHC), an associated peptide, and the T cell receptor (TCR). To determine the specificity and to ultimately control the immune response, it is essential to identify the antigenic peptides involved. If the antigenic peptide in an immune response is going to be characterized, it is important to be able to identify that antigen from a clinically available sample size. This research develops and validates a method for characterizing MHC class I associated peptides from a reasonably sized clinical sample (1 x 108 cells) and applies this technique to the identification of MHC class I associated peptides from an Idiopathic Thrombocytopenia Purpura (ITP) population and controls. Literature Review Our bodies are under constant attack by bacteria, parasites, virus, and other pathogens. We survive this barrage of invaders by having a line of defense, our immune system, which consists of innate and adaptive immunity. Innate immunity is a non—specific defense against pathogens or foreign material. This first line of defense includes physical and structural barriers such as the skin, acidic environments in the stomach and mucosal surface linings. The innate immune system also includes cellular defenses. The purpose of these cells, the polymorphonuclear (PMN) cells (neutrophils, eosinophils, and basophils) and mononuclear cells (macrophages and monocytes), is to nonspecifically phagocytize and break down pathogens or to release granules to destroy larger pathogens unable to be phagocytized. However, pathogens have devised ways to get around this first line of defense (Janeway et al., 2001). The next line of defense our bodies have is a specific immune response directed against the particular offending pathogen. The advantage of the more specific immune response is that it can be a more aggressive and destructive, long lasting immunity against the pathogen. This specific immune response is a complicated pathway involving T and B lymphocytes (Janeway et al., 2001). When the host is infected with a pathogen, the adaptive immune response can specifically recognize the pathogen, produce antibodies to that pathogen, and/or lyse the infected cells. Because this response is more destructive, it must be more specific in its target. The specificity and regulation of this response occurs through the interaction of a processed protein from the pathogen associated with a MHC molecule (or human leukocyte antigen (HLA) in humans) interacting with the T cell receptor (TCR) on the T lymphocyte. One of the keys to the specificity of the adaptive immune response is the MHC and its associated peptide from the processed pathogen. The MHC is a key component, but it is only one component in the interaction. What should be emphasized as the key to the adaptive immune response is the interaction of the MHC and peptide with the TCR (Janeway et al., 2001). Literature Review Part 1: Antigen Processing and Presentation There are two types of MHC molecules, MHC class I and MHC class II. MHC class I is found on nucleated cells as well as anuclear platelets. The main function of MHC class I molecules is to present peptides from degraded intracellular proteins. MHC class II is found on antigen presenting cells (APC) such as macrophages, B cells, and dendritic cells. The purpose of MHC class II molecules is to present peptides derived from degraded extracellular proteins. The MHC interacts with the TCR on T cells in a specific manner (Janeway et al., 2001). Generally, there are two types of T cells, cytotoxic T cells (Tc) and helper T cells (Th). Tc cells lyse the target cells they interact with, whereas Th cells provide cytokine help to the immune response. Two distinct membrane proteins differentiate these T cells. These membrane proteins are cluster differentiation (CD) 8, which is found on To cells, and CD4, which is found on Th cells. These T cells interact with a MHC-peptide complex via the TCR and their respective CD protein. The CD8 molecule on CD8+ T cells interact with the MHC class I molecules. The CD4 molecule on CD4+ T cells interacts with the MHC class II molecule (Janeway et al., 2001). Cytotoxic T cells When To cells interact with MHC class I and peptide, it can initiate an immune response. This response is called cell mediated immunity, which is an adaptive immunological response. In cell mediated immunity, the APC presents antigen complexed with a MHC class I molecule. The naive CD8+ Tc cell will associate via the TCR and the CD8 molecule with the MHC-peptide complex on the APC. If provided with an adequate co-stimulatory signal from the APC, Tc cells are activated, produce interleukin-2 (IL-2), and proliferate. The Tc cells release cytotoxic factors such as perforin, lymphotoxins, and serine dependent proteases that lyse their target cell (Janeway et al., 2001). Tc cells also release the cytokines y-interferon (IFN), tumor necrosis factor (TNF)-a, and TNF-B (Janeway et al., 2001). Helper T cells: Th1 and Th2 Helper T cells are CD4+ T cells that release cytokines that help direct the development of the immune response. Th cells are divided into two groups based on the type of help they give to the immune response. They are called T helper type 1 (Th1) T cells and T helper type 2 (Th2) T cells. Th1 T cells stimulate a type 1 immunity, which is characterized by macrophage activation, CD8+ T cell mediated immunity, and the production of complement binding and opsonizing antibodies (Cotran et al., 1994; Andersson, 1998; Hedlund-Treutiger et al., 1998; Singh et al., 1999; Janeway et al., 2001; Spellberg et al., 2001). Th1 cells can also contribute to humoral immunity by inducing B cells to produce strong opsonizing antibodies of immunoglobulin (Ig)G1 and IgG3 subclass in humans (Janeway et al., 2001). Th1 cells produce IL—2, y-IFN, IL—12, TNF-B, and lymphotoxin (LT)-alpha (Cotran et al., 1994; Singh et al., 1999; Janeway et al., 2001; Spellberg et al., 2001). y—IFN stimulates phagocytosis, the oxidative burst, and intracellular killing of microbes in the macrophage. y—IFN also up regulates the expression of MHC class I and class II proteins on the surface of a variety of cells stimulating antigen presentation to T cells (Spellberg et al., 2001). y—IFN and LT—alpha induce other cells to secrete proinflammatory cytokines, which leads to endothelial cell retraction and vascular smooth muscle relaxation. This allows the accumulation of blood in the vessels, diapedesis of leukocytes, and recruitment of additional cells to the area of infection (Cotran et al., 1994; Spellberg et al., 2001). Th2 T cells stimulate a type 2 immunity characterized by high antibody titers and down regulate the effects of a type 1 response (Andersson, 1998; Hedlund-Treutiger et al., 1998; Singh et al., 1999; Spellberg et al., 2001). The Th2 subset of CD4+ T cells initiates the humoral immune response, which includes aiding in class switching of antibodies to IgG2, IgG4, IgA, and IgE isotypes (Singh et al., 1999; Janeway et al., 2001). Th2 cells produce IL—4, IL—5, IL-6, IL-9, IL-10 and IL-13 cytokines (Cotran et al., 1994; Singh et al., 1999; Janeway et al., 2001). IL-4 activates B cell proliferation, antibody production, and class switching while inhibiting macrophage activation and suppressing Th1 cells. IL-10 inhibits the Th1 T cell response as well as macrophage activation (Janeway et al., 2001; Spellberg et al., 2001). CD4+ and CD8+ T cell Immune Response Both CD4+ and CD8+ T cells can contribute during an immune response. These cells may respond to different antigens, but both types of T cells can respond to the same immunological stimulus. For instance, during pregnancy, the paternal HLA antigens of the offspring can stimulate in the mother, antibody formation to inherited paternal HLA antigens as well as generating a cytotoxic T lymphocyte (CTL) response to inherited paternal HLA antigens (Bouma et al., 1996). Studies in the non-obese diabetic (NOD) mouse model of type I diabetes have shown that peptides from insulin are recognized by pathogenic CD8+ cells, substantiating the role of insulin as one of the autoantigens in type I diabetes. In addition CD4+ T cells are also involved in type I insulin dependent diabetes mellitus (IDDM) (Janeway et al., 2001) with a possible role for autoantibody formation. Role of the TCR Through the TCR, the T cell recognizes peptide bound within the groove of an MHC molecule and signaling can occur. Analysis of the TCR-MHC—peptide complex shows that a majority (greater than two-thirds) of the interactions between the TCR-MHC-peptide complex occur between the TCR and the region on the MHC molecule surrounding its bound peptide (Wang et al., 2002). However, the peptide plays a major role in TCR activation. Peptides in the MHC class I and class II complexes are identified by the TCR in a similar manner (Wang et al., 2002). The MHC-peptide complex is identified in a diagonal orientation by the TCR on the MHC-peptide complex. This allows for complete coverage and recognition of the peptide with the TCR (Wang et al., 2002). Antigen Processing In adaptive immunity, the targeted specific antigenic peptides are derived from processed bacterial, viral, foreign or abnormal proteins. However, normal cellular protein turnover is also important in MHC peptide presentation for self-recognition. Peptides are derived from many sources including the endocytosis pathway, intracellular protein degradation that includes degradation of endoplasmic reticulum components, and extracellular protein degradation (Kessler et al., 2002). These proteins are processed/degraded and complexed with the MHC. The immune system uses two distinct pathways common to all eukaryotic cells involved in the process of protein degradation. The ubiquitin-proteasome degradation pathway is used to generate 8-11 amino acid long peptides from intracellular proteins to load in MHC class I molecules (the endogenous pathway). However, not all class I peptides are necessarily processed by the ubiquitin pathway (Goldberg et al., 2002). The lysosomal-endosomal pathway is utilized for processing exogenous proteins into peptides for loading onto MHC class II molecules (the exogenous pathway) (Goldberg et al., 2002). Endogenous Pathway Antigen processing and presentation for MHC class I cells begins within the cytosol of cells. As viruses within the cell, tumors, or foreign cells present their unique proteins, immune recognition occurs. This method of immunological response is directed toward quiescent vs. stimulated status of all cells. Proteins within the cytosol are ultimately degraded into oligopeptides by proteasomes and enzymes, which will be described in more detail later. These antigenic oligopeptides are chaperoned by heat shock proteins (HSP). Within the cytosol, hsp70 and hsp9O bind and release these peptides in an ATP-dependent manner, to relay them to the other heat shock proteins and ultimately to the transporter associated with antigen processing (TAP) proteins. These oligopeptides are transported into the endoplasmic reticulum in an ATP-dependent manner by the TAP1 and TAP2 transmembrane heterodimer (Srivastava et al., 1994; Janeway et al., 2001). Newly synthesized MHC class I alpha chains and 82 microglobulin (82M) are associated with a chaperone protein within the endoplasmic reticulum. This chaperone protein is an 88 kiloDalton (kDa) membrane bound molecule called calnexin. HSP glycoprotein (gp)96 transfers the peptide to 10 the MHC class I molecule (Srivastava et al., 1994). When the oligopeptide binds to the alpha chain of the MHC molecule, calnexin is released and the alpha chain, 82M, and antigen complex can now make its way to the cell surface and be presented to CD8+ T cells. The oligopeptide bound in the cleft of the MHC class I molecule, uses two anchor amino acid residues usually located toward the ends of the peptide. This leaves the rest of the peptide free to interact with the TCR. Self—proteins as well as pathogenic cytosolic proteins are presented in this manner (Janeway et al., 2001). Proteasomes The proteasome degrades proteins within the cytosol of a cell into oligopeptides. Proteasomes, or multicatalytic protease complexes, are large molecular weight proteolytic complexes that are located in the cytosol and nucleus of all eukaryotic cells. The 20S proteasome is about 700 kDa and is composed of 28 subunits. The subunits are divided into two alpha and beta type subunits. The 208 proteasome is cylindrical in shape having a length of 15 nm and a diameter of 11 nm with a hollow core (Lupas et al., 1993). The proteasome structure is composed of four hollow rings arranged alpha ring, beta ring, beta ring, alpha ring. 11 Each ring has seven subunits. The beta subunits are thought to have multiple enzymatic activities. Three of the constitutively expressed beta subunits are X, Y, and Z (Goldberg et al., 2002; Gromme et al., 2002). The 268 proteasome is composed of the 20S proteasome with two 19S regulatory caps. These caps allow the degradation of ubiquitin conjugated proteins in an ATP dependent manner (Lupas et al., 1993). In the ubiquitin-conjugation process, multiple ubiquitin molecules are covalently attached to the amino group of lysine residues in the protein that is to be degraded. This polyubiquitin chain marks the protein for rapid degradation by the proteasome. Proteins containing PEST sequences, which consist of segments rich in proline (P), glutamic acid (E), serine (S), and threonine (T) amino acids, are also degraded by the proteasome (Gromme et al., 2002) . In the synthesis of proteins, the opportunity for errors can occur in the translation of information from the cellular deoxyribonucleic acid (DNA) to the functional protein. Proteins that are made which contain errors are called defective ribosomal products (DRiPs). Since DRiPs are not used by the cell and may interfere with other cellular functions, they are rapidly destroyed (Yewdell, 12 2002). Thirty percent or more of newly synthesized proteins are destroyed by proteasomes within 10 minutes of their synthesis (Yewdell, 2002). The generation of MHC class I—associated peptides is mainly dependant on the degradation of proteins by the 26S proteasome (Kessler et al., 2002). To degrade proteins used for MHC class I presentation, there are three steps in the degradation process for generating MHC class I peptides. First, proteins must be degraded by the 26S proteasome into oligopeptides that are either the correct size for MHC class I presentation (8-11 amino acids) or longer but extended only on their amino terminal end. Proteaseomes generate oligopeptides ranging in size from 2-30 amino acids in length. Only about 30% of these peptides are 8 amino acids in length or longer. The proteasome will either generate a peptide of optimal length or an N— terminal extended peptide. N-terminal but not C—terminal extended peptides are generated because the proteasome is required to make the C-terminal cleavage on MHC class I presented peptides (Goldberg et al., 2002). Second, aminopeptidases in the cytosol or endoplasmic reticulum must trim the oligopeptides that are extended at the amino terminus into peptides of the correct length, 8- 13 11 amino acids. Peptides must be of the correct length in order to bind tightly in the groove of the MHC class I molecule. Peptides longer than the optimal length tend not to bind as tightly to the MHC class I molecule (Goldberg et al., 2002). Aminopeptidases that trim the N-terminal end of the oligopeptides can both create antigenic peptides and destroy them unless the peptide is bound tightly to MHC class I molecules (Goldberg et al., 2002). Some of the known aminopeptidases are puromycin—sensitive aminopeptidase (PSA) and bleomycin hydrolase (BH). These well characterized cytosolic aminopeptidases are present in all cells. Another peptidase, which is located in the cytosol and nucleus, is tripeptidyl peptidase II. It is a very large multisubunit enzyme that cleaves tripeptides off from the N-terminal end of polypeptides. Tripeptidyl peptidase II has not been proven to be in the proteasome pathway (Goldberg et al., 2002). Aminopeptidases may also be present in the lumen of the endoplasmic reticulum. It is not known what percent of the peptides are trimmed in the cytosol and what percent are trimmed in the endoplasmic reticulum. (Goldberg et al., 2002). If aminopeptidases are located in the endoplasmic reticulum, peptides with an N-terminus extension can be trimmed to the optimal length after transportation by TAP. Since TAP translocates l4 peptides with a proline at position 2 poorly, this pathway may be needed for the class I molecules containing a proline at position 2 as an anchor residue (Gromme et al., 2002). It has been calculated that from 104 proteins degraded, only 1 peptide will end up binding to an MHC class I molecule (Gromme et al., 2002). The third process in generating peptides is the breakdown of the rest of the proteasome products by endopeptidases and exopeptidases. Most of the peptides generated are broken down into their amino acid components and recycled (Goldberg et al., 2002). Even though they are the majority of the peptides produced, peptides less than 8 amino acids cannot be detected free in the cytosol because of their rapid hydrolysis to amino acids. The shorter proteasome products presumably are degraded directly by various aminopeptidases to amino acids. The enzymes responsible for this rapid hydrolysis of proteasome products have not been identified. However, a few enzymes have been implicated including thimet Oligopeptidase (TOP). Larger peptide fragments must be broken down in two steps. First, the peptides are broken down to smaller fragments by enzymes such as TOP or other cytosolic endopeptidases. These products are further degraded by bestatin-sensitive di— or tripeptidases to single amino acids. This suggests 15 peptides that could bind to MHC could be destroyed in the cytosol or may be protected by cytosolic binding proteins (such as heat shock proteins) (Goldberg et al., 2002). The activity of the ubiquitin-proteasome degradation pathway can be modified in a number of ways to enhance or alter peptide generation. y-interferon induced cells contain three alternate proteins, low molecular mass protease (LMP)2, LMP7, and MECLl, in the 205 proteasome replacing the constitutively expressed X, Y, and Z subunits (Goldberg et al., 2002; Gromme et al., 2002). LMP2 and LMP7 also map within the MHC region on chromosome 6 (see Figure 1). These y—interferon induced subunits alter the peptidase activity. Only these induced subunits have N- terminal threonine residues. These residues are the catalytically active residues in the active site of the proteasome. DP D DR Ba LMP/T AP 3? 50 A l:ll::::::_:-::i:CL 4 t—--— Class II ———fl- Class Ill—TL— Class I ——1 Figure 1. Map of MHC Region on Chromosome 6. MHC class I, class II, and class III regions are indicated by arrows on the map. The A, B, and C loci map within the MHC class I region. DR, DO, and DP loci as well as the TAP and LMP genes map within the MHC class II region. 16 The 20S core particle has six active sites, two that preferentially cleave after hydrophobic amino acids (chymotrypsin—like), two that cleave after basic residues (trypsin-like) and two that cleave after acidic ones (peptidyl glutamyl peptide hydrolyzing activity or caspase- like). Proteasomes containing these subunits are called immunoproteasomes (Goldberg et al., 2002). Immunoproteasomes have the potential to alter the specificity or quantity of peptides that are generated. For example, using the protein ovalbumin, the 26S proteasome generates the antigenic peptide SIINFEKL (see Appendix 1 for amino acid codes) or the N—terminal extension of this peptide from ovalbumin about 6% of the time. However the 26S immunoproteasome generates these SIINFEKL containing peptides 11% of the time. This shows that the immunoproteasome enhances the production of immunodominant peptides (Goldberg et al., 2002). In addition to the 20S and 26S proteasomes (208 plus 19S ATP end cap), there is a hybrid 26S proteasome that is made of the 208 core plus a 19S end cap and a PA28 (a proteasome activator that enhances peptidase activity) end cap. PA28 can bind to the end of the 20S proteasome and dramatically increase its ability to hydrolyze small peptides. PA28 is composed of two homologous subunits, 17 PA28d and PA28B, both of which are induced by v-IFN. The PA28 subunits form a ring around the central opening of the 208 core through which substrates may enter or exit the 208 core. A large variety of biochemical activities have been proposed for the PA28 complex, including activation of the active sites of the proteasome, stimulation of peptide entry into the 20S particle, stimulation of peptide exit, and facilitating the binding of proteasomes to chaperones or to components of the endoplasmic reticulum (Goldberg et al., 2002). y—IFN also induces the major cytoplasmic trimming enzyme, leucine aminopeptidase (LAP), as well as other aminopeptidases in the endoplasmic reticulum capable of trimming N-extended precursors (Goldberg et al., 2002). Transporter Associated with Antigen Processing Once oligopeptides are generated in the cytosol, they must be transported into the endoplasmic reticulum. The endoplasmic reticulum is where empty MHC class I molecules are located. The transport of the peptides into the endoplasmic reticulum is via the TAP genes. The TAP genes map within the MHC class II region (see Figure 1) and form a heterodimer with an alpha and beta subunit coded for by TAP1 and TAP2. TAP1 and TAP2 genes transverse the membrane of the endoplasmic reticulum and transport, by an ATP 18 dependent mechanism, degraded oligopeptides into the endoplasmic reticulum. These peptides can complex with the MHC class I and 82 microglobulin dimmer and be expressed on the surface of the cell for detection by CD8+ T cells (Spies et al., 1990; Powis et al., 1992a; Powis et al., 1992b; Gromme et al., 2002). TAP1 and TAP2 are members of the ATP binding cassette (ABC) family of transport proteins. The initial step in moving the peptide from the cytosol to the endoplasmic reticulum is binding the peptide to TAP. The peptide binding site is formed by both TAP1 and TAP2 at a site between the putative pore (formed by the TAP1 and TAP2 proteins) and the ATP binding site. The translocation of peptides from the cytosol into the endoplasmic reticulum lumen by TAP requires hydrolysis of ATP. Hydrolysis of ATP causes a conformational change in the TAP proteins where the peptide can be translocated into the endoplasmic reticulum. TAP most efficiently translocates peptides of 8-16 amino acids in length. In addition to the size of the peptide, its sequence also affects the efficiency of translocation by TAP. The C-terminal residue of a peptide is a key determinant for efficient translocation by TAP with hydrophobic or basic C-termini selected (Gromme et al., 2002). 19 N-terminal extended peptides are generally bound by the TAP transporter with a higher affinity and more efficiently translocated into the endoplasmic reticulum. It is also thought that these peptides are less susceptible to destruction by peptidases in the cytosol as N-extended peptides were up to six times more stable in the cytosol than peptides of optimal MHC class I binding. TAP binds both extended and optimal length peptides (Goldberg et al., 2002). Certain TAP alleles have been found to alter the transport specificity of the degraded peptides into the endoplasmic reticulum of rats, varying the types of peptides that can complex with the MHC molecule (van Endert et al., 1992). MHC Peptide Loading Once the peptide is inside the lumen of the endoplasmic reticulum, it is associated with an empty class I MHC molecule. The alpha chain of the MHC is inserted into the membrane of the endoplasmic reticulum where it binds calnexin. The immunoglobulin binding protein (BiP), also binds the alpha chain of MHC. Both proteins may assist in the folding of the alpha chain of MHC and associating it with 82M. When the MHC dissociates from calnexin, the alpha chain and 82M binds to calreticulin. 20 The calreticulin alpha chain 82M complex is able to interact with TAP. This interaction is mediated by tapasin (TAP—associated glycoprotein). Upon binding of peptide, the MHC class I complex is released from TAP. Following peptide binding and release from the endoplasmic reticulum, MHC class I-82M-peptide complexes are transported through the Golgi apparatus and the trans—golgi network to the cell surface (Gromme et al., 2002). Peptides associated with MHC and presented on the cell surface are controlled by many factors, such as precursor protein expression and stability, efficiency of peptide generation by proteases, and rate of peptide transport into the endoplasmic reticulum (Engelhard et al., 2002). MHC class I Molecules MHC class I molecules at the cell surface consist of a glycosylated transmembrane alpha chain (45 kDa) noncovalently associated with the 11.6 kDa soluble protein, 82M (Parham et al., 1977), and a short peptide usually of 8-11 amino acids (Bjorkman et al., 1987b; Bjorkman et al., 1987a). In humans, the genes for the classical MHC are located on chromosome 6 (see Figure 1) and encoded by the three classical designated human leukocyte antigens (HLA)- A, HLA-B and HLA—C. The HLA class I loci are very 21 polymorphic with more than 263 HLA-A alleles, 501 HLA-B alleles and 125 HLA-C alleles identified to date (October 2002). This information is available on the website of the Anthony Nolan Research Institute (http://www.anthonynolan.com/HIG/index.html) (Robinson et al., 2000; Robinson et al., 2001). The alpha chain of MHC class I is composed of three domains (alnx3). The al and a2 domains are located on the top of the protein structure and make up the peptide—binding cleft. The peptide groove in the MHC is composed of two a—helices sitting on top of eight B-pleated sheets. Most of the polymorphisms in the HLA molecules are located in the alpha helix or B—pleated sheets in the peptide binding groove (Bjorkman et al., 1987b; Bjorkman et al., 1987a). The peptide binding groove has a pocket on either end of the groove that accommodates certain amino acid side chains depending on the MHC type. These are called the anchor residues (usuallylz‘1d and 9th amino acid), which hold the peptide in the groove. Once anchored on either end, the rest of the peptide is free to rotate above the MHC groove and interact with the TCR. The area in the groove that accommodates the anchor residues vary in depth and amino acid composition between MHC alleles which determines the set of peptides that can be bound by a particular class I allele. Since the shape of 22 the pockets is dependent on the MHC class I allele, there are allele-specific peptide binding motifs. These motifs only allow a few of the amino acids to fit into the pocket and act as anchor residues. The interaction with the anchor amino acid of the peptide and the MHC binding groove are essential for stable association. Because of the need for this stable association, the length of bound peptides is constrained to 8'10 residues (Gromme et al., 2002). Since the length of the MHC class I associated peptides are similar and the anchor motifs (usually the 2nd and 9fi‘andno acid) are known and can be obtained from the SYFPEITHI database (http://www.uni-tuebingen.de/uni/kxi/) (Rammensee et al., 1999), it is possible to predict potential MHC binding peptides from a protein sequence. In addition, there are several different algorithms available to predict proteasome cleavage sites (Nussbaum et al., 2001) as well as MHC class I binding peptides (Rammensee et al., 1999; Reche et al., 2002). These may be useful tools in MHC associated peptide binding prediction as well as data interpretation. Exogenous Pathway In addition to the endogenous pathway that presents intracellularly derived peptides, there is an exogenous 23 pathway for peptide presentation. The exogenous pathway presents peptides from extracellular proteins that are internalized by receptor-mediated endocytosis, phagocytosis, macropinocytosis, or endogenously expressed proteins from the secretory pathway by MHC class II molecules (Busch et al., 2000; Janeway et al., 2001). Proteins that are endocytosed become enclosed within endosomes and eventually fuse with lysosomes (Janeway et al., 2001). Within these endosomal compartments, degradation of endocytosed proteins occurs. Endocytosed proteins are cleaved into oligopeptides of 13-18 amino acids (Monaco, 1995) or 12—24 amino acids (Busch et al., 2000). This occurs by degradation with endosomal proteases such as asparaginyl endopeptidase, acid proteases that are activated at low pH, and cleavage of disulphide bonds by unknown mechanisms (Busch et al., 2000; Janeway et al., 2001). The MHC class II mechanism of antigen presentation to Th cells is similar to the MHC class I method. In the endoplasmic reticulum, three MHC class II molecules associates with its trimer chaperone molecule, the invariant chain (Ii). The MHC class II/Ii complex is exported through the Golgi apparatus and is targeted for the endocytic pathway (Busch et al., 2000). In the late 24 endocytic compartment called MHC class II compartment (MIIC), Ii is cleaved and 20—24 amino acids of the Ii (residue 81—104), called class II—associated invariant- chain peptide (CLIP), is left sitting in the cleft of the MHC class II molecule (Ghosh et al., 1995; Busch et al., 2000; Janeway et al., 2001). Ii is degraded in a stepwise manner by aspartyl proteases, cathepsin S, cathepsin L, or other proteases (Busch et al., 2000). The Ii/CLIP prevents MHC class II molecules from binding cytosolic proteins while it is being directed from the Golgi toward the endosomal compartments. In order for the MHC class II molecule to complex with the antigenic oligopeptide, DM0 and DM8 genes are required to remove the CLIP molecule. HLA—DM also stabilizes empty MHC class II molecules (Janeway et al., 2001). The oligopeptide can now associate with the MHC class II molecule and go on to be presented on the surface of the cell (Ghosh et al., 1995; Janeway et al., 2001). MHC class II molecules at the cell surface consist of a transmembrane alpha chain (34 kDa) noncovalently associated with a transmembrane beta chain (29 kDa) (Gorga et al., 1987), and a short peptide usually of 12—24 amino acids (Busch et al., 2000). In humans, the genes for the MHC class II molecules are located on chromosome 6 (see 25 Figure 1) and encoded by the three designated human leukocyte antigens (HLA)—DR, HLA-DO and HLA—DP. The HLA class II loci are also very polymorphic with more than 397 HLA-DRB alleles, 53 HLA—DQBI alleles and 100 HLA—DPB1 alleles identified to date (October 2002). This information is available on the website of the Anthony Nolan Research Institute (http://www.anthonynolan.com/HIG/index.html) (Robinson et al., 2000; Robinson et al., 2001). MHC class II molecules are composed of an alpha and beta chain. The a1 and 81 regions are located on the top of the protein structure and these are the domains that make up the peptide binding cleft. The peptide groove in the MHC class II is also composed of two a—helices sitting on top of eight B—pleated sheets with most of the polymorphisms located in the alpha helix or B-pleated sheets in the peptide binding groove (Janeway et al., 2001). The peptide binding groove of a class II molecule is more open and less restrictive in length than a class I molecule. This allows longer peptides to hang over either end of the binding groove and adds to the diversity in antigen presentation (Janeway et al., 2001). Since the anchor motifs of the MHC class II associated peptides can be obtained from the SYFPEITHI database (http://www.uni- 26 tuebingen.de/uni/kxi/) (Rammensee et al., 1999), it is also possible to predict potential MHC binding peptides from a protein sequence. Antigen Presentation The average copy number of peptides presented with MHC on the surface of a cell is 10—100 copies per cell with abundant self peptides ranging from 500 — 10,000 copies per cell. Peptides presented on the surface of the cell at 1,000 to 10,000 copies per cell represent only 1% of the total peptide pool (de Jong, 1998). The number of peptide- MHC complexes per target cell required for recognition and cytolysis by CTL has been shown to vary from several thousand to as few as one (Engelhard et al., 2002), from 100-500 complexes (de Jong, 1998), or from 1 to 20,000 complexes (Kageyama et al., 1995). A large number of peptides may be needed when the affinity of the TCR is low and a low number when the affinity of the TCR is high (Kageyama et al., 1995). 27 Literature Review Part 2: Applications of High Perfommance Liquid Chromatography (HPLC) and Mass Spectrometry (MS) to MHC Associated Peptide Identification Peptides from intracellular proteins are expressed with MHC class I on the surface of cells while MHC class II presents peptides from extracellular proteins. Because of the importance of the antigenic peptide (and the source of the antigenic peptide) in the immune response, it would be useful to characterize the MHC associated peptides to allow for immunomodulation by T cell clonal deletion or expansion (Bielekova et al., 2001) or synthetic peptide vaccine (Sundaram et al., 2002). Review of Previous Peptide Purification Protocols Initial studies characterizing MHC associated peptides typically used 109-10n‘cells or 40 grams or more of tissue. Most studies focus on cultured, transformed, or tumor cells where sample size can be 1 x.109 to 1 x 10n'cells (de Jong, 1998; Bonner et al., 2002). These protocols typically isolated the MHC molecule and associated peptide by immunoaffinity chromatography using either a generic or allele specific antibody (sometimes the cell lysate was passed over 3 columns to purify each MHC allele). Peptides were typically released 28 from the MHC by acid and heat denaturation. The peptides were purified from the larger MHC proteins by centrifugation using a molecular weight cut off filtration unit. The low molecular weight fraction, which contained the pool of peptides, was separated by one to several rounds of reverse phase (RP)-HLPC. A peptide in a selected HPLC fraction was identified by Edman degradation protein sequencing or mass spectrometry. Based on this protocol, MHC associated peptides have been identified as reported in the literature. For example, from 1-2 x 10m leukopheresed polymorphonuclear cells from a patient with chronic myeloid leukemia, defensin derived peptides were purified from HLA-DR molecules. This showed these non-APC cells with granulocyte macrophage colony-stimulating factor induced MHC class II expression presented defensin derived peptides as 25% of their high copy number peptides. This amount of peptide and peptide length (29—35 vs. 13-18 amino acids) were different than MHC class II peptides presented on APCs (Halder et al., 2000). From 40 grams of liver tissue from a Hepatitis B surface antigen (HBsAg) seropositive individual or 109 cells of Hepatitis B transfected Epstein Barr Virus (EBV) B cells, a peptide from the Hepatitis B virus nucleocapsid 29 antigen protein was purified from MHC class I molecules. This peptide also sensitized target cells for CD8+ T cell lysis. This provides evidence that antigenic Hepatitis B virus antigenic peptides were processed by hepatocytes, presented with MHC class I molecules and recognized by CD8+ T cells (Tsai et al., 1996). From 1010 EBV transformed B cells, the human equivalent of murine macrophage inflammatory protein (MIP) was purified from MHC class II molecules. In addition to this cytokine associated with lymphocyte activation, peptides from Ig proteins were also identified. This is consistent with the proteins expressed in B cells (Harris et al., 1993) . These protocols using immunoaffinity chromatography, molecular weight fractionation, and peptide sequencing to identify MHC associated peptides, are sufficient for 109- lon'cells. However, if this technique is to be applied to samples of realistic clinical availability (1 x 108 cells, 10-20 milliliters (ml) of whole blood or 0.1 grams (cw?) of cells), the sample volume must be reduced. The equivalent of 1 x 109 to 1 x lon'cells is equal to 1-100 grams (CNS) of tissue or loo—20,000 ml of whole blood. The normal volume of blood in an adult male is 66 ml/kilogram (kg) (Walker, 1993) and no more than 500 ml of blood is recommended to be 30 removed during voluntary blood donation (Walker, 1993). This makes 1 x 109 to 1 x IOll cells difficult to realistically obtain from a single clinical source. It can also be difficult to detect MHC associated peptides as MHC class I can bind 2,000 to 10,000 different peptides, which makes the abundance of a particular peptide 0.0001% to 1.0% of the total peptide pool (1 to 10,000 copies/cell). This correlates to a maximum yield of 0.16 to 1600 femtomole (fmol) per peptide per 1 x 108 cells. Due to the several manipulations used in the typical purification procedure, recovery can be 12-18% of the maximum yield (de Jong, 1998). To study MHC associated peptides from a realistic clinical sample, typical sample size needs to be reduced 10-1,000 times. In addition, Edman degradation protein sequencing and matrix assisted laser desorption ionization (MALDI) mass spectrometry protein sequencing techniques do not have the sensitivity needed to detect the majority of peptides from 1 x 108 cells based on the maximum theoretical yield from this cell number (de Jong, 1998). This makes characterizing MHC associated peptides from a clinically available sample size (1 x 108 cells) difficult. To maximize the yield of MHC associated peptide, the number of steps used in the isolation procedure should be 31 minimized. To this end, a mild acid wash of whole cells was adapted (Storkus et al., 1993; Skipper et al., 1999). This was combined with peptide purification on a C18 peptide trap. A few protocols use the mild acid wash technique but couple it with alternative peptide purification steps. These other peptide purification protocols include a SepPak C18 concentration step, 3-10 kDa membrane filtration step, and RP—HPLC separation (Storkus et al., 1993; Skipper et al., 1999), a SepPak C18 concentration step followed by RP-HPLC (Herr et al., 1999), or filtration, ultrasonication, and cation exchange chromatography (Clark et al., 2001). However, peptide purification on a C18 peptide trap purifies peptides from small molecules like salts and detergents and large proteins in one step vs. the several step typically used. The peptide trap reduces the number of manipulations and therefore increases the amount of peptide available for characterization. Acid Wash of Cells In previous studies, melanoma cells were treated with citrate phosphate buffer at a pH less than or equal to 3.6. This resulted in denaturation of class I complexes in less than one minute (15 seconds). A pH above 3.8 had no affect 32 on expression of class I complexes (Storkus et al., 1993). The denaturation of class I complexes was measured by indirect immunofluorescence by flow cytometry with the monoclonal antibody W6/32. Acid treatment appeared to destroy cellular activity with the W6/32 monoclonal antibody that recognizes a monomorphic combinatorial determinant requiring class I heavy chain and 82M association (Barnstable et al., 1978). Using a fluorescence cytometric assay, BBM—l (antiBZM), HC-10 (anti-class I heavy chain without 82M), and L243 (anti-HLA- DR) monoclonal antibodies, it was shown that melanoma cells treated with citrate—phosphate buffer at pH 3.3 lost reactivity with both the W6/32 and BBM-l antibodies but HC- 10 and L243 antibodies still reacted. This demonstrates that the class I heavy chains remain cell surface associated, but the 52M is lost (Storkus et al., 1993). At a pH as low as 3.0, class II and other non-MHC antigens were not affected by this acid treatment (Sugawara et al., 1987; Storkus et al., 1993). Citric acid pH 3.3 treatment for 1 minute decreased 85-95% of the W6/32 reactivity in a panel of diverse tumor cell lines (with the exception of a squamos cell carcinoma cell line, which reduced 74% of the W6/32 reactivity. (This variance was cell line dependent 33 and not HLA allele dependent. The reason for this variance is unknown) (Storkus et al., 1993). Human peripheral blood mononuclear cells (PBMC) were treated with glycine or citric acid buffered solutions. The solutions with a pH less than 3.3 could completely eliminate the reactivity of class I antigens within 1 minute. The citric acid buffers with 1% bovine serum albumin (BSA) were less toxic to the cells (Sugawara et al., 1987). In addition, influenza infected melanoma cells, treated with citric acid buffer pH 3.3 for 1 minute, lost their ability to be recognized by Tc cells. This ability was restored within 18 hours (optimal time to regenerate MHC class I complexes) (Storkus et al., 1993). Treatment of platelets with citric acid buffer pH 3.0 with 1% BSA for 10 minutes significantly reduced the antigenicity of HLA class I antigens as examined by enzyme linked immunosorbent assay (ELISA) using W6/32. There were no significant differences in the ultrastructure of the platelet as compared to PBS treated platelets (Kurata et al., 1989; Kurata et al., 1990). Reverse Phase-High Performance Liquid Chromatography Using RP-HPLC, the crude pool of peptides eluted from MHC molecules can be cleaned up, MHC associated peptides 34 can be separated by hydrophobic properties, and/or the complexity of the peptide pool can be reduced. RP—HPLC separates molecules based on interactions of the molecule with the hydrophobic stationary phase of the column and the mobile phase. Polypeptides adsorb to the hydrophobic surface of the reversed phase column and remain adsorbed until the concentration of the organic modifier (mobile phase) reaches the critical concentration to cause desorption (Gooding et al., 1990). Once the peptides are separated from the MHC molecules and the complexity of the peptide pool is reduced, the peptides can be identified. Liquid Chromatography Tandem.Mass Spectrometry (LC-MS/MS) LC-MS/MS can be a useful tool in the characterization of MHC associated peptides. LC-MS/MS couples a capillary liquid chromatography system to a mass spectrometer. Capillary HPLC allow for higher sensitivity detection. Also, there needs to be a method to distinguish one particular peptide from a pool of a possible 10,000 peptides. This can be accommodated by HPLC. Sequence information can be obtained by a number of methods. However, to be able to characterize the MHC associated peptides, the sensitivity of the method needs to 35 be in the low fmol range (Bonner et al., 2002). This can be accomplished by LC-MS/MS. There are three phases or steps in performing mass spectrometry. These steps include generating gas phase ions, analyzing the ions, and detecting the ions. Ions are generated using low pH solvents. With low pH, peptides become positively charged on their amino terminus. If the peptide sequence contains basic residues, these also become charged. Basic amino acids in the peptide sequence can lead to a multiple charge state (+2, +3) of the peptide. Peptides in a +2 charge state are easiest to sequence as both b and y ions have the potential to be charged (instead of just the b ions). To be able to analyze these ions, they must also be in the gas phase. There are many ways to get ions into the gas phase, electrospray ionization (ESI) is one of them. High voltage (2.5 kilovolts (kV)) is applied to the peptides in a liquid phase. This creates a Taylor cone or droplets of peptides when the liquid sprays out from the tip of the HPLC column. These droplets eventually evaporate and the peptides are left in the gas phase. Once the peptides are in the gas phase, they can enter the mass spectrometer. In the mass spectrometer, the ions need to be analyzed. There are many mechanisms to analyze 36 ions, an ion trap mass spectrometer is one of them. An ion trap traps all the ions in a chamber within a given time. The radio frequency (RF) and direct current (DC) voltages are changed so that ions with increasing mass are ejected from the trap and detected. A photo multiplier detects the ions and the data is recorded. The ion trap is now ready to trap the next set of ions. If sequence information for a particular ion is needed, an additional scan can be done on this ion. This is called tandem mass spectrometry (MS/MS). If an ion of interest is selected (usually an ion was selected because it was above background noise), all other ions will be ejected from the trap except the ion of interest. This ion is bombarded with helium gas causing the peptide to fragment along the peptide backbone at the peptide bond. These fragments become trapped within the ion trap and sequentially ejected and recorded. A peptide sequence can be obtained on this MS/MS information. Mass spectrometry measures the mass to charge ratio (m/z) of an ionized peptide. These peptides can be fragmented to generate overlapping m/z of the peptide fragments. When a peptide fragments, the amino terminus half is called the b ion and the carboxy terminus half is referred to as the y ion. From these overlapping b and y 37 ion series, sequence information of the original peptide can be obtained. Peptides can be characterized by LC-MS/MS. However, some types of peptides are easier to characterize than others. Peptides from tryptic digests have either an arginine or lysine at the carboxy terminus. This aids analysis in two ways. First, these peptides should at least have a +2 charge, having charges on the amino terminus and a charge on the arginine or lysine carboxy terminus. A charge on both ends of the peptide theoretically allows for a more complete series of b and y ions available for spectra interpretation. Without the charge at the carboxy terminus, some or all of the y ions may not be available for analysis. Second, MHC associated peptides have the potential to have any of the 20 amino acids at the carboxy terminus. This makes identifying the ions belonging to the b series vs. the y series more difficult. However, if the carboxy terminus ends in arginine or lysine, as would a tryptic digest, the initial ions in the y series can be more easily identified (Shabanowitz, 2001). An example of the difficulty characterizing MHC associated peptides is illustrated in the following study. In this study, MHC class I associated peptides 38 characterized by LC-MS/MS reported detecting 200 peptides. Of these 200 peptides, 10% were 150—600 fmol per 108 cells. The other 90% were present at 30-150 fmol. From this data, complete sequence could be obtained for only 8 of the peptides (Hunt et al., 1992). Despite these limitations, it has been shown that MHC associated peptides can be characterized by this method (Hunt et al., 1992). Literature Review Part 3: Platelets There are many different clinically available cell types that can be used to characterize MHC associated peptides. However, using platelets over other cells makes characterizing MHC associated peptides easier for several reasons. First, platelets are easily and readily obtained by venipuncture. Second, using the minimally invasive collection method of venipuncture, an ample number of platelets can be obtained from a reasonable clinical sample to characterize MHC associated peptides. Third, platelets can be easily isolated from other cell types in the blood. This allows for MHC associated peptide characterization from a single population of cells. Also, platelets normally express MHC class I molecules without expressing MHC class II molecules (Kunicki, 1989). 39 Platelet Development and Function Platelets, or thrombocytes, are small disc shaped (2-5 micrometer (um) in diameter) anuclear cell-like structures. Platelets are produced in the bone marrow by part of the megakaryocyte fragmenting off, which is called platelet budding. About two thirds of the platelets in the body are found circulating in the blood while one—third are located in the spleen. The lifespan of a platelet is 8-12 days. Platelets are destroyed by macrophages mainly in the spleen. Platelets play a major role in the formation of blood clots. The formation of thrombi involves trapping platelets, red blood cells, and white blood cells within a fibrin network. Platelets are important in hemostasis and help in the formation of a hemostatic plug at the site of vascular injury. Platelets adhere to the injured endothelial wall by attaching themselves via glycoprotein (GP)Ib/IX to exposed collagen via the associated von Willebrand factor (vWF). Once platelets are attached to the endothelial layer, they undergo a shape change. Among the various compounds that activate platelets, thrombin, ADP, and attachment to collagen likely serve as the most important physiological regulators. Activated platelets release the contents of their alpha granules, 4O which contain, among other things, platelet associated coagulation proteins. They also release the contents of their dense bodies containing ADP, Ca”, and vasoactive amines. Thromboxane A2 (TxA2) is released by activated platelets, which induces aggregation in other platelets. Once activated, platelets aggregate at the site of injury via platelet-platelet adhesion through fibrinogen interaction with the GPIIb/IIIa receptor. The build up of platelets, fibrinogen, and other cells forms the hemostatic plug. Platelet Proteins GPIIb/IIIa, a transmembrane protein, is a fibrinogen receptor unique for the megakaryocyte lineage. There are approximately 50,000 molecules of GPIIb/IIIa on the surface of resting platelets which accounts for 3% of the total platelet protein and 17% of platelet membrane protein mass. GPIIb is composed of alpha (Mr=115,000) and beta (Mr=22,000) subunits that are linked together by disulfide bonds. GPIIb and GPIIIa are noncovalently linked (Wadenvik et al., 1998). GPIb/IX is the second most abundant platelet membrane protein have approximately 20,000-30,000 molecules of GPIb/IX on the surface. It is the receptor for vWF and is 41 unique to platelets and megakaryocytes (Wadenvik et al., 1998). GPIb is composed of an alpha (Mr=145,000) and beta (Mr=22,000) subunits that are linked together by a disulfide bond(s) and transverse the plasma membrane. The binding sites for vWF and thrombin are located on the alpha chain while the phosphorylation sites are on the beta chain. The carboxy terminus of GPIb beta subunit is located in the cytoplasmic region of the platelet. There is a 25 amino acid signal peptide at the amino terminus of GPIb beta. The signal peptide is removed by a signal peptidase cleavage of the peptide bond. GPIb is noncovalently associated to GPIX. The GPIb-GPIX complex is involved in binding actin filaments of the cytoskeleton to the platelet membrane (Lopez et al., 1988). The GPIb/IX complex is composed of two molecules of GPIb alpha, two molecules of GPIb beta, two molecules of GPIX and one molecule of GPV (Wadenvik et al., 1998; Bussel et al., 2000). Like some other GPs, GPIb beta is polymorphic with the two alleles of GPIb arising from a Gly/Glu substitution at amino acid 15. The GlYu allele has a gene frequency of 99% (Bussel et al., 2000). Other platelet membrane expressed proteins include GPIV at 20,000 copies per platelet, the collagen receptor, GPIa/IIa has 1,000 copies per platelet, and platelet- 42 endothelial cell adhesion molecule—1 (PECAM-l) having 8,000 copies on the surface of each platelet (Wadenvik et al., 1998). Major cytoplasmic proteins include actin, which constitutes 20—30% of the total platelet protein content. Actin binding protein, talin and myosin heavy chain each make up 2-8% of the total protein content. Other abundant platelet proteins include vinculin, alpha—actinin, gelsolin, caldesmon, tropomyosin and profilin. IgG is estimated to be on the platelet surface, an average of 100 molecules, and intracellularly in alpha granules at 20,000 molecules per platelet (Wadenvik et al., 1998). Expression of MHC — A, B, C on platelets The number of HLA molecules on platelets have been reported to be between 8,000 to 15,000 per platelet (Botto et al., 1990). The presence of HLA-C on platelets was demonstrated using an anti—HLA—Cwl antibody. There appears to be variable allelic expression of HLA-C as well as HLA-A and HLA-B molecules on the surface of platelets and other cells (Kunicki, 1989). A study found equal number of HLA- A2 and HLA—Cwl molecules on the surface of platelets (Datema et al., 2000). However, other studies have found that the HLA—C antigens are greatly reduced from the expression of HLA-A and HLA-B molecules (Mueller-Eckhardt 43 et al., 1980). Some of these more abundant proteins found in platelets maybe processed and found associated with MHC class I on the surface of platelets. Literature Review Part 4: Autoimmunity It is important to characterize MHC associated peptides not only in normal cells, but in cells involved in an immunological response. There are many different immunological responses that can be characterized. One area of interest is the autoimmune diseases where MHC associated peptides from the cells of a normal individual can be compared to the peptides presented from cells of an individual with the autoimmune disease. It is not known what triggers an autoimmune response. A multitude of genetic as well as environmental factors have been implicated. It is believed that autoimmunity is initiated by a response involving T cells (Hedlund- Treutiger et al., 1998). Autoimmunity has been defined as a sustained and persistent T cell dependent immunological response toward self tissues that produces long term tissue damage. Autoimmunity occurs when self tolerance has been lost. The cause of this loss of tolerance is unknown. The only genetic markers strongly associated with autoimmunity have 44 been the MHC molecules, however, their contributions were determined to not be the sole factor in disease (Theofilopoulos, 1995). The complexity of the cause of autoimmunity can be attributed to its polygenic traits, phenotypic/genetic heterogeneity, incomplete penetrance, environmental factors, and chronic stimulation with common pathogens. The combination of these factors plays a role in the abnormal immunological response against self (Theofilopoulos, 1995). The recognition of self and nonself to the immune system has had a number of proposed models. The most recent model is called the danger model (Matzinger, 2002). What this model proposes is the immune system does not necessarily recognize self and nonself. The danger model proposes that the immune system recognizes danger or alarm signals from injured or necrotic cells. This model stresses that a foreign protein does not guarantee an immune response nor do self proteins guarantee tolerance. The basic assumption about what initiates immunity in the danger model is the signals sent by distressed cells. Signals such as exposure to pathogens, toxins, mechanical damage, necrosis, etc. are signals not sent by normal healthy cells (Matzinger, 2002). 45 Tolerance to self is acquired by clonal deletion of autoreactive T cells within the thymus or by clonal inactivation of mature peripheral lymphocytes (Hedlund- Treutiger et al., 1998). In spite of this regulation, a Th1 helper T cell response is predominant in organ specific autoimmune diseases such as IDDM, multiple sclerosis, and rheumatoid arthritis (RA), acute allograft rejection, and some chronic inflammatory disorders (Singh et al., 1999). Literature Review Part 5: Idiopathic Thrombocytopenia Purpura Idiopathic thrombocytopenia purpura (ITP) is an autoimmune disease characterized by the destruction of platelets by an unknown cause. In 80-90% of individuals with ITP, an antiplatelet antibody is evident. Common causes of thrombocytopenia such as chemotherapy and destruction secondary to other diseases are excluded when considering a diagnosis of ITP. Detection of platelet autoantibodies in addition to a low platelet count, normal bone marrow, as well as excluding other etiologies of thrombocytopenia help to confirm the diagnosis of ITP (George et al., 1995). The spleen is the major site for platelet removal as well as a site for antiplatelet antibody production in 46 chronic ITP (Wadenvik et al., 1998). Platelet loss is due to the increased destruction of these platelets by the macrophage (Schwartz, 1998). ITP has been shown to involve autoreactive B cells as well as an elevated number of 75 TCR T cells (Andersson, 1998). Ig bound to the platelet surface causes opsonization of the platelets by activated macrophages (Andersson, 1998). Most autoimmune diseases, including ITP, are associated with a Th1 type of cytokine response (Andersson, 1998). To support the association of ITP with a Th1 type of response, B cells from individuals with chronic ITP were stimulated to produce antiplatelet autoantibodies when incubated with IL-2, v—IFN and autologous platelets. This is associated with the Th1 type of activation pattern (Semple, 1998). In about 75% of ITP cases, the autoantibody produced is against the platelet membrane receptor GPIIb/IIIa or GPIb/IX (Andersson, 1998; Kuwana et al., 2001; McMillan et al., 2001). From studies with transfected CHO cell lines, it was determined that most of the antiGPIIb/IIIa antibodies bind epitopes located on GPIIb—alpha (McMillan et al., 2001). In addition, the presence of platelet reactive T cells in ITP individuals have been found. GPIIb/IIIa was one of 47 the major target antigens recognized by platelet reactive CD4+ T cells (Kuwana et al., 2001). It was found that GPIIb/IIIa autoreactive T cells do not respond to the intact GPIIb/IIIa protein, but to chemically modified and recombinant GPIIb/Illa fragments expressed in bacteria (Kuwana et al., 2001). This suggests a role for antigen processing of GPIIb/Illa and presentation with MHC. Fragments spanning the length of GPIIb and GPIIIa were used as antigens in a T cell proliferative response assay. Fragments GPIIb—alpha 18-259 (from the 871 amino acid protein) or GPIIIa 22-262 (from the 762 amino acid protein) induced a T cell proliferative response in 24 of the 25 ITP patients in the study with some patients responding to two or more fragments (Kuwana et al., 2001). This suggests that there are several T cell epitopes from GPIIb/Illa, but the most immunodominant T cell epitope(s) are found in the amino terminal of the processed GPIIb or GPIIIa protein (Kuwana et al., 2001). These immunodominant T cell epitopes appear to also be involved in the production of antiGPIIb/IIIa antibodies (Kuwana et al., 2001). GPIIb/IIIa reactive CD4+ T cells in ITP individuals have a helper activity that promotes the production of anti-GPIIb/IIIa antibodies capable of binding normal platelets (Kuwana et al., 2001). This indicates that these 48 autoreactive CD4+ T cells are involved in the production of antiplatelet antibodies in ITP (Kuwana et al., 2001). In addition, abnormal levels of GPIb were observed in one third of children with ITP (Semple et al., 1996). There appears to be little or inconsistent HLA class I association found with ITP. Most studies found no association or no significant association after a correction for the number of antigens tested (Mohanty et al., 1979; Mueller-Eckhardt et al., 1979; Veenhoven et al., 1979; Mitchell et al., 1981; Helmerhorst et al., 1982; Gratama et al., 1984; Stanworth et al., 2002). However, a few studies did find an HLA association with ITP. An increased frequency of HLA-B8 and B12 was found in adult patients with chronic ITP (Goebel et al., 1977). An increase in the frequency of Aw32 (26.9% of patients vs. 0.8% of the controls) in children with acute ITP was found (Evers et al., 1978). An increased frequency of HLA—A28 was found in chronic ITP patients (50% of patients vs. 18.7% of the controls). However, the author states that significance may be lost when corrected for the low sample size of 16 patients (el-Khateeb et al., 1986). In a study of three children with chronic ITP and their immediate family members, the HLA antigens A3 and B7 were identified in all three families and most of the family 49 members (Stuart et al., 1978). In another study of HLA association with ITP, the most significant finding was the presence of HLA-DRw2 (75% of patients vs. 23% of controls). The co-occurrence of either A3 and B7 (known to be in linkage disequilibrium with DRw2) or A26 and Bw38 was also significantly increased as compared with the control population (Karpatkin et al., 1979). While there was little to no HLA class II association found with ITP in a Japanese population, there was a strong association with HLA class II genes and the specificity of platelet antiglycoprotein autoantibody produced (Kuwana et al., 2000). There was a strong association with DRBl*0405 and DQB1*0401 with the production of antiGPIIb/IIIa and a strong association with DRBl*0803 and DQB1*0601 with the production of antiGPIb/IX (Kuwana et al., 2000). This study suggests that HLA genes do not necessarily influence the development of ITP but do influence the specificity of the antiplatelet autoantibody produced (Kuwana et al., 2000). This may be taken as evidence implicating the importance of the role of antigen processing and MHC— peptide presentation in the immune response that leads to autoantibody formation in ITP. 50 Objectives and Hypothesis To be able to characterize the MHC associated peptides from cells in quiescent and diseased states could provide key information on the immune response to pathogens and lead to better vaccine development. However, most studies have only characterized MHC associated peptides from tumor cells or cultured cells because a method to detect MHC associated peptides from a reasonable clinical sample has not been reported in the literature. Based on calculations, it was hypothesized that the most abundant MHC associated peptides presented during an immune response could be characterized from a reasonable clinical sample. The three objectives for this research were to establish a method to detect and characterize MHC associated peptides from a sample of reasonable clinical availability (1 x 108 cells). Once developed, this method would be validated by comparing the MHC associated peptides eluted from different cell populations. Finally, this method would be applied to characterize the MHC associated peptides being presented during an immunologicial reaction. The first objective was to establish the best method to isolate MHC associated peptides from a small volume of cells. Once these peptides were isolated, an optimal method to detect these peptides also had to be determined. 51 This was to be accomplished with a reasonable clinical sample size. The second objective was to apply this methodology to characterize the MHC associated peptides being presented on normal cellular samples. Peptides from platelets, white blood cells and red blood cells were compared to determine the most abundant peptides presented on these cells. This study focused on the comparison of these populations within an individual and between individuals of similar or different MHC types. The last objective was to characterize the MHC associated peptides being presented on platelets of individuals with ITP to determine if the antigen processing and presentation pathway has a role in ITP development. Our objective was to determine if there is a peptide presented on platelets specific for ITP. 52 MATERIALS Monoclonal Antibodies Hybridoma W6/32 (which produces the cytotoxic anti— MHC-A,—B,-C monoclonal antibodies (MAb)) (Parham et al., 1979) and L243 (which produces the cytotoxic anti-MHC—DR MAb) (Lampson et al., 1980) were obtained from the American Type Culture Collection (ATCCCU (Manassas, VA) (see Table 1). Archived ascites fluid was obtained for HB116 (which produces the non—cytotoxic anti—MHC—A,-B,—C MAb) (Brodsky et al., 1982) (Table 1), W6/32, and L243. Table 1. Monoclonal Antibodies used in Studies Listed in the table are the names of the MAbs (including hybridoma designation if applicable), isotype of antibody and species obtained from, specificity of the Ab, and the source the Ab was obtained. Antibody Isotype Specificity Source of Antibody H58A Mouse IgG2a MHC-A,B,C Purified MAb from VMRD, Inc. HB-116 Mouse IgGl MHC-A,B,C Hybridoma cell (MB40.5) line, ATCC. L243(HB-55) Mouse IgG2a MHC-DR Hybridoma cell line, ATCC. SPA-822 Mouse IgG2a Hsp70/Hsc70 Purified MAb from StressGen SPA—850 Rat IgG2a Grp94 Purified MAb from StressGen W6/32(HB—95) Mouse IgG2a MHC-A,B,C Hybridoma cell line, ATCC. 53 The monoclonal antibody H58A, an anti-MHC class I mouse IgG2a antibody (VMRD, Inc., Pullman, WA) (Davis et al., 1987) (see Table 1), came resuspended in phosphate buffered saline (PBS) [137 mM sodium chloride (NaCl) (J.T. Baker Inc., Phillipsburg, NJ), 2.7 mM potassium chloride (KCl) (Columbus Chemical Industries, Inc., Columbus, WI), 9.6 mM sodium phosphate monobasic (NaHéPO4) (Sigma, St. Louis, MO), and 1.5 mM potassium phosphate monobasic UGbPO4) (J.T. Baker Inc., Phillipsburg, NJ, pH 7.4] at a concentration of 1.0 milligram (mg)/ml. SPA—822 an anti- heat shock protein 70/heat shock cognate 70 (Hsp70/Hsc70) mouse IgG2a monoclonal antibody (StressGen, Victoria, B.C., Canada) (Smith et al., 1993) and SPA—850, an anti-glucose regulated protein 94 (Grp94) rat IgG2a monoclonal antibody (StressGen, Victoria, B.C., Canada) (Kulomaa et al., 1986) (Table 1) were purchased as lyophilized reagents. These antibodies were resuspended in PBS to a concentration of 1.0 mg/ml. All antibodies were divided into working aliquots (0.1 — 1.0 ml) and stored at —20 degrees Celsius (°C) until use. Epstein Barr Virus B Cells EBV transformed B cells from the lou‘cnrljfh.American Society of Histocompatibility and Immunogenetics (ASHI) 54 workshop were used as a source of HLA molecules. EBV B cells that were obtained from the ASHI Cell Bank and Repository (Bringham and Women's Hospital, Boston, MA) had been frozen in liquid nitrogen and stored by the Immunohematology and Serology Laboratory at Michigan State University. ITP and Control Samples Control/normal samples were obtained from non ITP individuals with informed consent who have been serological (MHC class I typing trays: One Lambda Inc., Canoga Park, CA; Pel Freez, Brown Deer, WI; or Genetic Testing Institute, Brookfield, WI) or molecular [Micro SSP” DNA Typing Class I ABC generic (One Lambda Inc., Canoga Park, CA)] MHC typed by the Tissue Typing Laboratory at Michigan State University (East Lansing, MI). ITP samples were obtained from the Hematology and Oncology Clinic at Michigan State University with informed consent. ITP patients selected for this study were clinically diagnosed with ITP. Patients were diagnosed with ITP based on the detection of platelet autoantibodies in addition to a low platelet count, normal bone marrow, and no other etiology for thrombocytopenia (George et al., 1995). A current platelet autoantibody level was obtained 55 on the samples by a radiometric micro technique (Schwartz et al., 1990) by the Platelet Antibody Laboratory at Michigan State University, East Lansing, MI. 56 METHODS METHODS: Antibody and Sample Preparation 1. Preparation of Monoclonal Antibodies and Samples 1.1. MAb Preparation 1.1.1. Antibody Production - MAb Hybridoma Cell Culture The hybridomas were grown in the following cell culture media. The media used for W6/32 was 86.8 percent (%) volume/volume (v/v) Dulbecco’s Modified Eagle's Medium with 4500 mg glucose/Liter (L), L-glutamate, sodium bicarbonate (NaHCO3), and pyridoxine hydrochloride (HCl) (Sigma, St. Louis, MO). The media used for L243 was 86.8% (v/v) Dulbecco’s Modified Eagle’s Medium with 1000 mg glucose/L, L-glutamate, NaHCO3, and pyridoxine HCl (Sigma, St. Louis, MO). Both media were supplemented with 10% (v/v) heat inactivated (56°C for 30 minutes to inactivate complement) fetal bovine serum (Sigma, St. Louis, MO), 1 millimolar (mM) sodium pyruvate (Sigma, St. Louis, MO), 12 mM [N-2—Hydroxyethyl]piperazine-N’-[ethanesulfonic acid] pH 7.2 (HEPES) (Sigma, St. Louis, MO), 10,000 Units (U) penicillin (Sigma, St. Louis, MO), 10 mg streptomycin (Sigma, St. Louis, MO), and 0.3 mg of gentamicin sulphate (Boehringer Mannheim Co., Indianapolis, IN). 57 Cell media was sterilized using 0.2 micrometer (um) filter unit (Nalgene® Disposable Filterware, Nalge Nunc International, Rochester, NY) and stored at 4°C until use. Before use, culture media was warmed to 37°C. Approximately 4 ml of media was added to 1 x 106 to 1 x 107 cells in 25-75 square centimeters (cu?) polystyrene tissue culture treated flasks with plug seal cap (Corning, Inc., Corning, NY) and the cells were allowed to grow. Hybridoma cells were incubated in the culture media at 37°C and 5% carbon dioxide (CO2) in a water—jacketed incubator (Forma Scientific, Inc., Marietta, OH). When the cells were overgrown, as estimated by visualization, cells were divided and additional media was added to the culture until an ample volume of cells and culture media was obtained (20-100 ml). Then, the cells were cultured and allowed to overgrow for 1-2 weeks. Cells were pelleted and the supernatant that contained the secreted MAb from the hybridoma cell line was frozen at — 70°C. 1.1.2. Antibody Purification - Protein A Affinity Chromatography of MAb (W6/32, HB116, and L243) Either cell culture supernatant (20 ml at a time) or archived frozen ascites fluid (1-5 ml) was thawed. The ascites fluid was centrifuged at 850 times gravity (x g) 58 for 10 minutes. If the ascites supernatant was not clear, then equal volumes of PBS (for IgG2a) or 1.0 Molar (M) NaCl /1 M glycine (Sigma, St. Louis, MO) pH 8.0 (for IgGl) was added. The ascites fluid was then filtered through a sterile 0.45 um filter unit (Millex-HA filter unit, Millipore Corp., Bedford, MA). Protein A sepharose beads (binding capacity 20 mg of human IgG/ml; Sigma, St. Louis, M0) were pulse spun (100 x g for 15 seconds). The supernatant was removed and either the cell culture media or ascites fluid was added to the sepharose beads and incubated overnight, rotating at 4°C, with 2.5 — 4.0 ml of Protein A sepharose beads. The Protein A beads were pelleted by centrifugation (100 x g for 1 minute) and the supernatant was removed and saved. The sepharose beads were put into a column packed with enough glass wool to cover the bottom opening of the column. The column was filled with buffer (20 m1) (PBS pH 7.4 for IgG2a or 0.5 M NaCl/ 0.5 M glycine pH 8.0 for IgG1) and allowed to drip out of the column until low, but not allowing the beads to dry. Additional buffer was added to the beads in the column and the beads were allowed to settle. Fresh buffer was allowed to continuously flow by gravity into the column to remove any unbound proteins. The optical density (OD) of the flow-through was monitored 59 at 280 nanometers (nm) (Model 111 Absorbance Monitor, Gilson, Middletown, WI). The column was washed with buffer specific for the isotype of immunoglobulin that was going to be isolated until all unbound proteins were removed as indicated by an OD reading at 280 nm. Once the beads were washed and the 280 nm absorbance was at zero, the flow was stopped and the buffer above the top of the beads in the column was removed. Elution buffer (PBS pH 3.0 was used for IgG2a antibodies and PBS pH 6.0 was used for IgG1 antibodies) was then added to the column and allowed to continuously flow as before. The flow- through was monitored by ultra violet (UV) absorbance at 280 nm and the fraction that absorbed at 280nm was collected. The pH of the solution was neutralized to pH 7.0 with 1.0 normal (N) sodium hydroxide (NaOH) (J. T. Baker Inc., Phillipsburg, NJ). Residual glycine from the buffer was dialyzed out of the IgGl (HB116) sample. Dialysis tubing (Spectra/Por® molecular porous membrane tubing, Spectrum Medical Industries Inc., Laguna Hills, CA) with a molecular weight cut off (MWCO) of 12,000-14,000 Daltons (Da), flat width of 32 millimeters (mm), and length of 12 inches was soaked in distilled water for 5 minutes. At one end of the dialysis tubing, 2-3 knots were tied. The sample was added to the 60 tubing and the top of the tubing was tied with 2-3 knots. The closed tube was placed into a 4 L flask filled with PBS and allowed to dialyze for 2 days at 4°C while stirring. The liquid in the flask was changed once during the two days. After two days, the dialyzed liquid was stored at — 20°C or immediately concentrated. The dialyzed eluate was concentrated with a Centriprep®-10 concentrator (MWCO of 10,000 Da) according to manufacturer's directions (Amicon, Beverly, MA). Briefly, Centriprep®-10 columns were prerinsed with distilled 18 mega ohm water (ddHOH) for 5 minutes. Up to 15 ml of sample was added to the column at a time and was spun at 2000 x g for 20-30 minutes. The rest of the sample was added in 15 ml aliquots and centrifuged until the total volume of the sample was reduced to approximately 5 ml. The concentration and amount of antibody recovered was determined by UV absorbance at 280 nm (Lambda 2 UV/VIS Spectrometer, Perkin Elmer, Foster City, CA) using the extinction coefficient for IgG of 1.4 = 1.0 mg/ml. The formula used to determine the concentration was [(1.4)(X mg/ml) = (OD at 280 nm)(1.0 mg/ml)] and the formula to determine the amount of the protein was (X mg/ml)(X ml) = X mg. The antibody was stored at -20°C until further use. 61 1.1.3. Validation of Antibody by Complement Dependent Microcytotoxicity Testing To test the ascites and cell culture obtained MHC class I MAb for biological activity, control T cells were obtained from the Immunohematology and Serology Laboratory at Michigan State University. To a 72 well tray (MicroWell® Mini Tray, Nalge Nunc International, Rochester, NY) oiled with light white mineral oil (Sigma, St. Louis, MO), MAb and controls were added. Undiluted anti-human lymphocyte sera (MAb obtained from the Tissue Typing Laboratory at Michigan State University, East Lansing, MI) was used as a positive control; normal serum from human male blood type AB+ non transfused donors (RJO Biologicals, Overland Park, KS) was used as a negative control both undiluted and diluted 1/2 in RPMI-1640 with L—glutamine and sodium bicarbonate (Sigma, St. Louis, MO) supplemented with 5% heat inactivated (56°C for 30 minutes) fetal bovine serum (RPMI with 5% bovine serum); and the W6/32 or HB116 MAb diluted 1/10, l/lOO, 1/1000, and 1/5000 in RPMI with 5% bovine serum were the test samples. Tests were run in duplicate. To each well, 1 microliter (ul) of the antisera and 1 ul of the T cells (2.0 x 106/ml to 3.0 x 106/ml) were added using either a 50 ul Hamilton syringe (Hamilton Microliter® 62 syringe model 705 with PB600-1 repeating dispenser, Hamilton Co., Reno, NV) or cell dispenser (Lambda JET” cell dispenser model LJ-60/72, Soken Inc., Tokyo, Japan). The cells and antibody were incubated at room temperature for 30 minutes while rotating. After incubation, 5 ul of rabbit complement (Pel-Freez, Brown Deer, WI) were added to each well and incubated at room temperature for 60 minutes while rotating. Next, 10ul of STAIN—FIX” (used to stain and fix the cells) (One Lambda, Canoga Park, CA) was added to each well and the tray was covered with a cover slide (Terasaki Insta-Seal”, One Lambda, Canoga Park, CA) and incubated for 15 minutes while rotating at room temperature. After 15 minutes, the tray was placed under an inverted contrasting light microscope at 100 times (X) magnification (Leitz DM IL, Leica Mikroskopie und Systeme GmbH, Wetzlar, Germany) and each well was visualized and scored. If the cells were 81-100% dead (as determined by a visual estimate), they received a score of 8. If the cells were 51-80% dead, they received a score of 6. A score of 4 indicates the cells were 21-50% dead and a score of 2 indicates only 11—20% of the cells were dead. A score of 1 indicates that 90-100% of the cells were alive (Hopkins, 2000). 63 To test ascites and cell culture obtained class II MAb for biological activity, control B and T cells were obtained from the Immunohematology and Serology Laboratory at Michigan State University. To a 72 well tray oiled with light white mineral oil, a set of MAb and controls were added for B cells and a set of MAb and controls were added for T cells. Undiluted anti-human lymphocyte sera (positive control); normal serum from human male blood group AB+ non transfused donors (negative control) undiluted and diluted 1/2 in RPMI with 5% bovine serum, and the L243 MAb diluted 1/2, 1/100, and 1/1000 in RPMI with 5% bovine serum were added to the tray. Tests were run in duplicate for B cells and for T cells. To each well, 1 ul of the antisera and 1 ul of the B cells (2.0 x 106/ml to 3.0 x 106/ml) or 1 ul of the T cells (2.0 x 106/ml to 3.0 x 106/ml) were added using either a 50 ul Hamilton syringe or cell dispenser. The cells and antibody were incubated at room temperature for 30 minutes while rotating. After incubation, 5 ul of class II rabbit complement (One Lambda, Canoga Park, CA) were added to each well and incubated at room temperature for 60 minutes while rotating. Next, 10ul of STAIN-FIX“ was added to each well and the tray was covered with a cover slide and incubated 64 for 15 minutes while rotating at room temperature. After 15 minutes, the tray was read as described earlier. The commercially obtained antibodies H58A, SPA—822, and SPA—850 were validated by the class I microcytotoxicity test as described using 1/10, 1/100, and 1/1000 dilutions of the antibody with T cells. Refer to Table 1 for a list of antibodies used. 1.2. Cells from Cultured Cell Lines 1.2.1. Cell Culture of EBV B Cell Line EBV B cells were thawed in a 37°C VWR® brand water bath (VWR Scientific Inc., Willard, OH) prior to use. The 1 x 106 to 1 x 107 EBV B cells in 1 ml of thawed media were added to a Falcon® 14 ml round bottom polystyrene tube (Becton Dickinson and Co., Franklin Lakes, NJ) and culture media was slowly added until the tube was full. Culture media consisted of 78.77% (v/v) RPMI-1640 media with L- glutamine and NaHCO3 (Sigma, St. Louis, MO) supplemented with 15.0 % (v/v) heat inactivated fetal bovine serum, 12 mM HEPES, 0.15% [weight/volume (w/v)] NaHCO3 (J.T. Baker Inc., Phillipsburg, NJ), 0.3 mg of gentamicin sulphate (Boehringer Mannheim Co., Indianapolis, IN), 10,000 U penicillin (Sigma, St. Louis, MO), and 10 mg streptomycin (Sigma, St. Louis, MO). Culture media was sterilized using a 0.2 um filter unit and stored at 4°C until use. Before 65 use, culture media was warmed to 37°C. After thawing the cells, the media was immediately mixed with the cells and centrifuged at 125 x g for 10 minutes. The supernatant was poured off and the cells were resuspended in 4 ml of media. Cells were grown upright in 25-75 cm? polystyrene tissue culture treated flasks with a plug seal cap. EBV B cells were incubated in the culture media at 37°C and 5% CO2 in a water—jacketed incubator. When the cells were overgrown, cells were divided and additional fresh media was added to the culture until an ample number of cells were obtained. From a sterile 100 ul aliquot of the culture, a cell count was obtained. A cell count was calculated by counting the number of cells in the 4 corner and center red cell counting squares (dimensions of a red cell square is 0.2 mm x 0.2 mm x 0.1 mm) on a hemocytometer (Neubauer style SPotlite® hemocytometer, Baxter Healthcare Corp., McGraw Park, IL). A cell count was obtained by using the formula: (total number of cells counted in the red cell squares)(dilution factor)/ (length of red cell square in mm) (width of red cell square in mm)(depth of red cell square in mm) (number of red cell squares counted) = cells/ul. Cells were pooled as needed, pelleted by 66 centrifugation, and stored at —70°C. Approximately 1 x 108 cells were used per experiment. 1.2.2. EBV B cell Sample Preparation for MHC purification EBV B cells (1 x 10°) were lysed in 1-5 ml of cell lysis buffer rotating at 4°C for 30 minutes (Sidney et al., 1995). Cell lysis buffer contained 50 mM Tris [Hydroxymethyl]amino methane hydrochloride (Tris HCl) (Sigma, St. Louis, MO) pH 8.5, 1% (v/v) 1% tert— Octylphenoxy poly(oxyethylene) ethanol (IGEPAL CA-630) (equivalent to Nonidet P-40) (Sigma, St. Louis, MO), 150 mM NaCl, 5 mM ethlyelediaminetetracetic acid (EDTA) (J.T. Baker Inc., Phillipsburg, NJ), and 2 mM phenylmethyl sulfonyl fluoride (PMSF) (Sigma, St. Louis, MO). PMSF was made as a 100 X stock in denatured ethyl alcohol 200 proof (EtOH) (Pharmaco Products Inc., Brookfield, CT) immediately before use. Since the cellular lysate could not be passed through a sterile Millex-HA 0.45 um filter unit, the lysate was centrifuged at 10,000 x g for 20 minutes to clear cellular debris (Sidney et al., 1995). The supernatant was removed and used immediately for HLA isolation. 1.3. Cells from Whole Blood from Donors From normal controls, 30 to 50 ml of whole blood or eighty to one hundred milliliters from ITP individuals 67 (except ITP #3 which only 30 ml was collected) was collected via venipuncture into acid citrate dextrose (ACD) or EDTA anticoagulant VacutainersTM (Becton Dickinson, Franklin Lakes, NJ). Cell populations were isolated immediately by the following centrifugation methods and counted with a minimum cell count of 1 x 10° white blood cell (WBC) equivalents. WBC equivalents was used to normalize platelet HLA expression to 1 x 108 WBCs as the expression of MHC per platelet is less than the expression of MHC per WBC (Kunicki, 1989) (normalized to: 1 x 108 WBC, 1 x 10° red blood cells (RBC), and 3 x 109 platelets). Samples were further processed or stored at —70°C. 1.3.1. Platelet Sample Preparation To obtain platelet rich plasma, whole blood was centrifuged at 200 x g for 10 minutes. Platelet rich plasma (PRP) was carefully removed and placed into 5 ml 75 x 12 millimeter (mm) round bottom polystyrene tube (Sarstedt Inc., Newton, NJ). Prostaglandin E1 (PGEl) (Sigma, St. Louis, MO) was added to the PRP for a final concentration of 1 micromolar (uM) to inhibit platelet activation. A 500 ul aliquot of this PRP was given to the Platelet Antibody Laboratory at Michigan State University for a platelet antibody test by radiometric micro technique (Schwartz et al., 1990). From the remaining sample, 68 platelets were pelleted by centrifugation at 2000 x g for 10 minutes followed by two washes of 4 ml per tube with PBS, 5mM EDTA, 3% (w/v) BSA (Sigma, St. Louis, MO), 0.025% (w/v) sodium azide (Sigma, St. Louis, MO) pH 7.1. Platelets were combined into one tube and washed once with 4 ml of EDTA phosphate buffered saline (EPS) (154 mM NaCl, 10 mM EDTA-trisodium (Nag) (Sigma, St. Louis, MO), 8 mM KHfiXL, pH 7.4). Platelets were resuspended by careful mixing with a pipette in EPS and used immediately. A 20-30 ul aliquot of the resuspended platelets was diluted in saline and counted by a particle counter (Particle Data Lab, Elmhurst, IL or Beckman-Coulter Inc., Miami, FL) or using a hemocytometer as before. 1.3.2. White Blood Cell Sample Preparation Either whole blood or whole blood with the PRP removed (the volume of the PRP was replaced with EPS) collected in ETDA or ACD Vacutainers” was centrifuge at 750 x g for ten minutes. The white blood cell layer was aspirated off and placed into a 15 ml polypropylene conical centrifuge tube. The tube was filled with tris ammonium chloride (140 mM ammonium chloride (NHAII) (J.T. Baker Inc., Phillipsburg, NJ), 6mM Tris—HCl, pH 7.2) and incubated at 37°C for 5-7 minutes or until all the red blood cells were lysed as determined by visual observation. The sample was 69 centrifuged for 10 minutes at 200 x g and the supernatant was aspirated off and discarded. The white blood cell pellet was washed once with Dulbecco’s PBS (DPBS) (Sigma, St. Louis, MO) and centrifuged 10 minutes at 200 x g. The supernatant was aspirated off and discarded. The white blood cell pellet was resuspended in 10 ml of DPBS or EPS by gently pipetting. A small aliquot (10 ul) was used to make a 1/10 dilution. This dilution of WBC was used to obtain a cell count using a hemocytometer as before. WBCs were pelleted at 200 x g for 10 minutes and the supernatant removed. The white blood cells were used immediately or stored at —70°C. For Affi—Gel®-10 immunoaffinity MHC purification (cell lysis step is not used for citric acid mild acid wash purification), WBCs (1 x 10°) were lysed in 1-5 ml of cell lysis buffer rotating at 4°C for 30 minutes as previously described. Briefly, cell lysis buffer contained 50 mM Tris pH 8.5, 1% (v/v) IGEPAL CA-640, 150 mM NaCl, 5 mM EDTA, and 2 mM PMSF. The lysate was centrifuged at 10,000 x g for 20 minutes (Sidney et al., 1995). The supernatant was removed and used for HLA isolation. For magnetic bead immunoaffinity MHC purification (cell lysis step is not used for citric acid mild acid wash purification), WBCs (approximately 1 x 10°) were lysed in 1- 7O 5 ml of cold PBS/1% IGEPAL CA-630 with 2 mM PMSF (200 mM dissolved in ethanol made immediately before use) for 15 minutes at 4°C. The lysate was then centrifuged at 2000 x g for 3 minutes. 1.3.3. Red Blood Cell Sample Preparation After the platelets and/or WBC were removed, approximately 100 ul of packed RBCs was removed from the blood collected in the Vacutainerm and washed 2 times with 15 ml of EPS (200 x g for 10 minutes) in a 15 ml polypropylene conical centrifuge tube. The RBCs were then resuspended in 10 ml of EPS by gently pipetting. A small aliquot (1-10 ul) was used to make a 1/10 or 1/100 dilution. This dilution of RBCs was used to count the RBCs on a hemocytometer as before. The RBC count was adjusted in order to obtain 1 x 108 cells. 1.3.4. Molecular MHC Class I Typing From the whole blood sample of the ITP samples (with the PRP removed and replaced with EPS) spun at 750 x g for 10 minutes, 200 ul of the white blood cell layer was used to isolate DNA with a commercially available DNA isolation kit (QIAamp® DNA Blood Mini Kit, QIAGEN Inc., Valencia, CA). Following manufacturers instructions, 400 ug of protease (QIAGEN Inc., Valencia, CA) was placed into a 1.5 ml polypropylene microfuge tube (Brinkmann Instruments 71 Inc., Westbury, NY and PGC Scientific, Frederick, MD). To the tube, 200 ul of the white blood cell layer was added along with 200 ul of Buffer AL (used to lyse cells). The sample was mixed by vortexing for 15 seconds and incubated at 56°C for 10 minutes. The sample was then centrifuged briefly (13,000 x g for 5 seconds) and 200 ul of 100% EtOH was added. The sample was then mixed again by vortexing for 15 seconds, centrifuged briefly, and added to the QIAamp® spin column where it was centrifuged at 6,000 x g for 1 minute. The supernatant was discarded. To the column, 500 ul of Buffer AWl was added to wash the column and centrifuged 6,000 x g for 1 minute discarding the supernatant. Another wash of 500 ul of Buffer AW2 was applied to the column and spun at 20,000 x g for 3 minutes. The supernatant was discarded. With the QIAamp® spin column in a clean ultra violet (UV) light treated (90 seconds at 254 nm, GS Gene Linker” UV Chamber, Bio-Rad, Richmond, CA) 1.5 ml polypropylene microfuge tube, 200 ul of Buffer AE (elution buffer) was added to the column and allowed to incubate at room temperature for 1 minute. The column was spun at 6,000 x g for 1 minute and the supernatant with the eluted DNA was either used immediately or stored at —20°C. From 200 ul of buffy coat, this protocol should elute 15-25 micrograms (ug) of DNA in 200 72 ul of elution buffer with an OD of 260 nm/280 nm ratio in the range of 1.7 to 1.9. By sequence specific primer (SSP) based methodology (One Lambda Micro SSP“ DNA Typing Class I ABC generic kits, One Lambda, Canoga Park, CA), class I molecular typing was assigned. Following manufactures instructions, the primer set tray and master mix were thawed at room temperature. To the master mix (which includes salts, buffers, and each deoxyribonucleotide triphosphate (dNTP) provided by One Lambda optimized for use with their Micro SSP“ DNA typing kit), 28 units (U) of Thermus aquaticus (Taq) polymerase (Perkin Elmer, Foster City, CA) was added and mixed by vortexing. The master mix was spun at 13,000 x g for 5 seconds and 9 ul of the mix along with 1 ul of sterile ddHOH was added to the negative control well. Next, 111 ul of DNA isolated using the QIAamp® DNA Blood Mini Kit (optimal concentration of DNA for the Micro SSP" DNA typing kit is 25-200 nanograms (ng)/ul and an OD 260 nm/280 nm ratio of 1.65 — 1.8) was added to the D-mix and mixed by vortexing. The master mix was spun at 13,000 x g for 5 seconds and 10 ul of the mix was added to the rest of the 95 wells in the HLA class I SSP typing tray. The polymerase chain reaction (PCR) was carried out in a 9600 or 9700 (emulated to a 9600) thermocycler (Perkin 73 Elmer, Foster City, CA). Cycling parameters were: 1 cycle of 96°C for 130 seconds and 63°C for 60 seconds; 9 cycles of 96°C for 10 seconds and 63°C for 60 seconds; and 20 cycles of 96°C for 10 seconds, 59°C for 50 seconds and 72°C for 30 seconds, and held at 4°C. The products were electrophoresed on a 2.5 % agarose gel (Pel Freez, Brown Deer, WI) with 0.5 x TBE buffer {44.6 mM Tris[hydroxymethyl]amino methane (Tris) (Sigma, St. Louis, MO), 44 mM boric acid, (J. T. Baker, Phillipsburg, NJ), 5 mM EDTA, pH 8.0] using an electrophoresis tank (HYBAID Electro-4 Gel Tank, Hybaid Ltd., Ashford, Middlesex, UK) and power supply (Model ZOO/2.0, Bio—Rad, Richmond, CA). The gels were placed into 100-500 ml of ddHOH with 0.5-2.5 mg of ethidium bromide (Calbiochem—Behring Corp., La Jolla, CA) and incubated with gentle rotation at room temperature for 10 minutes. DNA bands were visualized using the transilluminator (Chromato-Vue Model 75-36, UVP Inc., San Gabriel, CA) at the wavelength of 254 nm. The gels were photographed for documentation with black and white film (Type 667 Polaroid® film, Polaroid Corp., Cambridge, MA) using a camera (Fotodyne PCR—10, Fotodyne Inc., Hartland, WI). The film was exposed for one second at f—stop (f)=8 and developed for one minute. Molecular types were converted to serological equivalents. 74 METHODS: MHC Associated Peptide Characterization To meet the first objective, a method to characterize MHC associated peptides from a small volume of cells was adapted from the original published protocol of immunoaffinity chromatography, molecular weight fractionation, and peptide sequencing for 109 to 10n'cells. From this original method, three methods to characterize MHC associated peptides evolved. This would lead to the ability to characterize MHC associated peptides from a small volume of cells. 2. Affi-gel®-10 Affinity Chromatography 2.1. Affi-gel®—10 HLA Column Preparation Affi-gel®-10 active ester agarose (Bio-Rad, Richmond, CA) was resuspended and 1.5 to 2 ml of gel was transferred to a 15 ml polypropylene conical centrifuge tube (Corning Inc., Corning, NY). Added to the Affi—gel®-10 was 4 times the volume of cold (4°C) ddHOH. The gel was mixed, pulse spun (100 x g for 15 seconds), and the supernatant was removed. This was repeated 3 times or until all the alcohol the Affi-gel®—10 was stored in was removed. Washing the column matrix and coupling the antibody must be completed within 20 minutes without letting the column 75 matrix dry out. To the 1.5 to 2 ml of gel, 1 ml of W6/32 MAb (1.76 mg/ml) or 1 ml of HB116 (0.77 mg/ml) was added to the column matrix. As a control column to check for nonspecific binding, 1 ml of bovine albumin (Sigma, St. Louis, MO) (1 mg/ml) was added. The column matrix was rotated at 4°C overnight to allow the antibody to couple to the column matrix. To the column matrix, ethanolamine HCl (Sigma, St. Louis, MO) pH 8.0 was added for a final concentration of 167 mM and incubated for one hour to block the remaining active sites on the Affi-gel®—10. The column matrix was transferred to a column with a glass wool plug and washed with ddHOH until the column matrix was free of reactants as detected by the UV absorbance at 280 nm. 2.2. Affi—gel®-10 Affinity Chromatography - Purification of HLA The Affi-gel®-10 with the W6/32 antibody, HB116 antibody, or bovine albumin coupled to the column matrix was washed with elution buffer [50 mM glycine pH 11.5 adapted from (Gorga et al., 1987)] or 50 mM glycine/0.1 % (w/v) deoxycholate (Sigma, St. Louis, MO), pH 11.5 (Gorga et al., 1987)], 20 mM HEPES/140 mM NaCl pH 7.8, cell lysis buffer, and 20 mM HEPES/140 mM NaCl pH 7.8 again. The cellular lysate from either lysed WBC obtained from whole 76 blood or lysed EBV B cells was added to the W6/32 bound to the column matrix and incubated rotating overnight at 4°C. The W6/32—column matrix was layered in a column filled with PBS and a glass wool plug in the bottom of the tube. The column matrix was allowed to settle. The detergent from the cell lysis buffer was washed out by letting 20 mM HEPES/140 mM NaCl pH 7.8 flow through the column until the absorbance at 280 nm was zero. Once the absorbance was zero, the residual HEPES/NaCl solution was removed (without letting the column matrix dry) and the elution buffer was added and allowed to flow by gravity through the column. The fraction that absorbed at 280 nm was collected. The extinction coefficient of 1.5 mg/ml was used for HLA molecules to calculate the amount of HLA purified (Parham, 1979) . Residual glycine from the buffer was dialyzed out of the MHC sample. Dialysis tubing with a molecular weight cut off (MWCO) of 12,000-14,000 Daltons (Da), flat width of 32 millimeters (mm), and length of 12 inches was prepared as before to dialyze glycine from immunoglobulins. The tube with the MHC solution was placed into a 4 L flask filled with distilled water and allowed to dialyze for 2 days at 4°C while stirring. The liquid in the flask was changed once during the two days. After 2 days, the tubing 77 was cut and the liquid was stored at —20°C or immediately dried by lyophilization (Virtis® Model IO—MR—TR with Freezemobile II, The Virtis Co., Inc., Gardiner, NY). The sample was resuspended in PBS at a concentration of 1 ug/ul. Purification of MHC was checked by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS—PAGE). Because MHC could not be purified using this protocol, MHC associated peptides were not characterized from samples using this purification procedure. 2 . 3 SDS-PAGE 2.3.1. Tricine-SDS-PAGE Tricine—SDS-PAGE is useful for separating small proteins and peptides from 1 to 100 kDa (Schagger et al., 1987; Schagger, 1994). A 10% polyacrylamide running gel [9.7% (w/v) acrylamide (Boehringer Mannheim Co., Indianapolis, IN), 0.3% (w/v) N,N’-methylene—bis-acrylamide (BIS) (Bio-Rad, Richmond, CA), 1M Tris[hydroxymethyl]amino methane (Tris) (Sigma, St. Louis, M0), 0.1% sodium dodecyl sulfate (SDS) (Boehringer Mannheim Corp., Indianapolis, IN), pH adjusted to 8.45 with HCl (Mallinckrodt Chemical Co., Paris, KY), and 10% (v/v) glycerol (Mallinckrodt Chemical Co., Paris, KY)] with the dimensions 102 mm width x 50 mm length x 0.75 mm thick overlaid by a 102 mm width x 23 mm length x 0.75 mm thick 4% polyacrylamide stacking gel 78 [3.9% (w/v) acrylamide, 0.1% (w/v) BIS, 1M Tris, 0.1% SDS, pH adjusted to 8.45 with HCl] was used. The glass plates were cleaned and rinsed with 100% methanol (MeOH) (Mallinckrodt Chemical Co., Paris, KY) and the gel casting apparatus (Bio—Rad, Richmond, CA) was assembled. To start the polymerization of the running gel, 0.05% (w/v) ammonium persulfate (J. T. Baker Inc., Phillipsburg, NJ) and 0.05% (v/v) N,N,N’,N’-Tetramethylethylenediamine (TEMED) (Boehringer Mannheim Corp., Indianapolis, IN) was added and gently mixed. The gel was immediately poured between the casting glass plates to 23 mm below the top of the smaller glass plate. Distilled water was carefully overlaid on top of the unpolymerized gel to create an even edge during polymerization. After polymerization, the water was removed and the plates and were carefully dried with cotton gauze. To start the polymerization of the running gel, 0.075% (w/v) ammonium persulfate and 0.075% (v/v) TEMED was added and gently mixed and was poured between the glass plates. The comb was inserted into the stacking gel to create a 1 cm long stacking gel. The gel was allowed to polymerize. The anode and cathode buffers were poured into the electrophoresis chamber. The anode buffer consisted of 0.1 M Tris pH 8.9 adjusted with HCl and the cathode buffer 79 consisted of 0.1 M Tris, 0.1 M N- tris[Hydroxymethyl]methylglycine; N-[2—Hydroxy-1,1- bis(hydroxy—methyl)ethyl]glycine (Tricine) (Sigma, St. Louis, MO), and 0.1% SDS pH 8.25. The sample (5—25 ul) and protein standards (0.1-0.2 ul) (Electrophoresis Calibration Kit, Pharmacia Inc., Piscataway, NJ) (protein standards were diluted with 100 ul of non-reducing buffer) were mixed with an equal volume of either reducing loading buffer or non-reduced loading buffer. The reducing loading buffer will reduce disulphide bonds. The non-reducing buffer was composed of 12% (w/v) SDS, 30% (v/v) glycerol, 0.05% (w/v) bromophenol blue (Sigma, St. Louis, MO), and 150 mM Tris—HCl pH 7.0. To obtain the reducing buffer, 6% (v/v) 2-mercaptoethanol (Sigma, St. Louis, MO) was added immediately prior to use to the non-reducing buffer. Loading buffer was added to the sample or protein standard and the sample was placed in a boiling water bath for 5 minutes and then immediately placed on ice and loaded into the wells of the gel. One lane of protein standards was run on every gel. Samples were electrophoresed in an electrophoresis chamber (Mini-PROTEAN® II Cell electrophoresis chamber, Bio-Rad, Richmond, CA) using a power supply. To the gel, 30 volts (V) were applied until the sample entered the 80 running gel. Once the sample was in the running gel, 130- 200 V was applied until the bromophenol blue dye front was at the bottom of the gel. When the dye was at the bottom of the gel, the voltage was turned off and the gel was immediately fixed and silver stained. 2.2.2 Silver Staining All silver staining solutions were made immediately prior to use. The gel was immediately fixed in 100 ml of 50% (v/v) MeOH/5% (v/v) formaldehyde [37.3% (v/v) formaldehyde, Mallinckrodt Baker Inc, Paris, KY] in water for 1 hour or 50% (v/v) MeOH in water overnight rotating (40 rpm) at room temperature in a polystyrene disposable sterile 150 x 15 mm petri dish (VWR Scientific Products, Willard, OH). After fixing, the gel was rinsed in 500—1000 ml of ddHOH for 30-60 minutes rotating (40 rpm) at room temperature. The gel was immersed in 0.1% (w/v) silver nitrate (Sigma, St. Louis, MO) for 30 minutes rotating (40 rpm) at room temperature. The gel was rinsed with three changes of ddHOH in 10 minutes rotating (40 rpm) at room temperature. The gel was immersed in the developing solution [3% (w/v) sodium carbonate (Sigma, St. Louis, MO)/0.02% (v/v) formaldehyde] rotating (40 rpm) at room temperature. When the protein bands became dark (less than 10 minutes), the reaction was stopped by immersing the gel 81 in 1% (v/v) acetic acid for 10 minutes rotating (40 rpm) at room temperature (Gorg et al., 1985). The gel was washed in ddHOH and archived. If protein bands did not stain dark enough or the background was too dark, the silver stain in the gel was “recycled” in order to improve the contrast (Gorg et al., 1985). To recycle the stain, the gel was then placed in Farmer’s Reducer [1% (w/v) potassium ferricynide (Sigma, St. Louis, MO)/1.6% (w/v) sodium thiosulfate (Fisher Scientific Co., Fair Lawn, NJ)] rotating (40 rpm) at room temperature until the silver stain disappears (about 5 minutes). The gel was rinsed for one minute with tap water and immersed in ddHOH water with several changes of water while rotating (40 rpm) at room temperature until the yellow color is removed from the gel. The gel is again immersed in 0.01% (w/v) silver nitrate and the process is repeated (Gorg et al., 1985). To archive the gel, the gel was rinsed in ddHOH for 10 minutes and was placed between two layers of cellophane that had been immersed in water. The cellophane was dried to the gel in a GelAir Drying System (Bio-Rad, Richmond, CA) for 1-3 hours or until dry. 82 3. Immunomagnetic Chromatography 3.1 Preparation of Magnetic Beads — Binding MAb to Earls Magnetic beads, Dynabeads® M-450 Rat anti-Mouse IgG2a (DYNAL, Oslo, Norway) or Dynabeads® M-450 Sheep anti-Rat IgG (DYNAL, Oslo, Norway), were used. The beads were resuspended to obtain a homogenous mixture and 25 ul of beads were transferred to a 1.5 m1 polypropylene microfuge tube. The microfuge tube was placed in a magnetic holder (DYNAL MPC®-E Magnetic Particle Concentrator for Microtubes of Eppendorf Type 1.5 ml, DYNAL, 0150, Norway) for one minute allowing the magnetic beads to stick to the wall of the microfuge tube and the liquid was removed by aspiration. The microfuge tube was removed from the magnet and 500 ul of cold (4°C) PBS/0.1%BSA or cold PBS/ 1% IGEPAL CA-630 was added to resuspend and wash the beads. The microfuge tube was again placed on the magnet, supernatant aspirated off and the beads were washed with 500 ul of cold PBS/0.1% BSA or cold PBS/ 1% IGEPAL CA-630 again using the magnet. After the final wash, the liquid was removed by aspiration and 0.5 — 2.0 ug of MAb in 25-500 ul of cold PBS/0.1%BSA or cold PBS/ IGEPAL CA-630 was added to the magnetic beads. The beads were incubated for 30 minutes at 4°C or room temperature with bi-directional mixing. The 83 magnetic beads were collected with the magnet for 2-3 minutes, supernatant removed by aspiration, and washed with 500 ul of cold PBS/0.1% BSA or cold PBS/ 1% IGEPAL CA-630 three times. To cross link the MAb to the magnetic beads, the beads were washed once in 10 ml of 0.2 M triethanolamine hydrochloride (Sigma, St. Louis, MO) pH 9.0 using the magnet and then the beads were resuspended in 10 ml of 0.2 M triethanolamine hydrochloride pH 9.0 and 52 mg of dimethyl pimelimidate dihydrochloride (Sigma, St. Louis, MO) was added. The beads were incubated with bi- directional mixing for 45 minutes at room temperature. After incubation, the beads were collected with the magnet and washed with 10 ml of 0.2 M ethanolamine pH 8.0 and resuspended in another 10 ml of 0.2 M ethanolamine pH 8.0 for 2 hours at room temperature. The beads were washed three times using the magnet with cold PBS/1% IGEPAL CA- 630. The magnetic beads were immediately used. 3.2. Magnetic Bead Immunoaffinity Purification of MHC The cellular lysate from either lysed WBC obtained from whole blood or lysed EBV B cells were used. The supernatant from the cell lysate was precleared by incubation with rat anti-mouse IgG2a magnetic beads for one hour at 4°C (25 ul of beads washed 3 times in PBS/1%IGEPAL 84 CA-630 using the magnet). (Preclearing of the cellular lysate was eventually dropped as it was shown not to improve the purification procedure.) The precleared lysate was then incubated with either W6/32 rat anti-mouse IgG2a magnetic beads, H58A rat anti-mouse IgG2a magnetic beads, or L243 rat anti-mouse IgG2a magnetic beads for 30 minutes to one hour at 4°C. The beads were washed once with 1 ml of cold (4°C) PBS/1% IGEPAL CA—630 and twice with cold PBS using the magnet. MHC was eluted by the addition of 25 ul of 0.2N or 25—250 ul of 0.5 M acetic acid (EM Science, Gibbstown, NJ). After a five minute incubation with acetic acid, the beads were collected by the magnet and the supernatant was removed and either run on a SDS- polyacrylamide gel immediately or the MHC associated peptides were released and then an SDS—polyacrylamide gel was run. If the beads were to be reused, the beads were washed with PBS/1% IGEPAL CA-630, eluted again with acetic acid, and the supernatant was checked for residual protein. The beads were stored in PBS with sodium azide and washed with PBS/1% IGEPAL CA-630 before being used again. 3.3. Purification of MHC Associated Peptides Two separate protocols were evaluated to obtain an optimal protocol to purify the associated peptides from the MHC molecules isolated by magnetic bead immunomagnetic 85 chromatography. The molecular weight fractionation protocol was based on previous published protocols (Tomlinson et al., 1996). 3.3.1. Peptide Purification by Molecular Weight Filtration Peptides were purified from larger molecules by releasing them from the MHC molecule and filtering them through a molecular weight filter. This was accomplished by denaturing MHC in 25 ul to 1 ml of 10% (v/v) acetic acid and boiled for five minutes. This was achieved by either adding glacial acetic acid to a final concentration of 10% (v/v) or drying the sample in a vacuum centrifuge (ThermoSavant, Holbrook, NY) with no heat until almost dry and adding 10% (v/v) acetic acid. A portion of the sample was immediately loaded onto an SDS—polyacrylamide gel. The remaining sample was dried in a vacuum centrifuge with no heat to 5-40 ul and resuspended in 160-200 ul of 0.1% (v/v) trifluroacetic acid (TFA) (Pierce, Rockford, IL) filtered through a 200 ul 5,000 nominal molecular weight limit (NMWL) filter (Ultrafree®-MC low binding cellulose filter, Millipore Corp., Bedford, MA) (prerinsed 2 times with 0.1% (v/v) TFA at (3000 rpm for 1—2 hours). To remove contaminants from the filter membrane, the filter was further pre-rinsed with HPLC water (J. T. Baker, 86 Phillipsburg, NJ), four times with 0.1% (v/v) TFA, once with 10% (v/v) HPLC grade MeOH (J. T. Baker, Phillipsburg, NJ), once with 50% (v/v) sequence grade acetonitrile (EM Science, Gibbstown, NJ), and twice with 0.1% (v/v) TFA. Alternatively, the sample was first filtered through a 2 ml, 5,000 NMWL filter (Ultrafree®-CL low binding cellulose filter, Millipore Corp., Bedford, MA) (3000 rpm for 1-2 hours). An additional 1 ml of 10% acetic acid was added to this sample and the sample was centrifuged again. The large molecular weight fraction was dried by a vacuum centrifuge and saved for analysis by SDS—PAGE. The low molecular weight fraction was dried by a vacuum centrifuge to 2—5 ul. The sample was diluted with 1 ml of water and dried again. This step was repeated with another 1 ml of water and dried to 5-40 ul. Samples were stored at -70°C until separation of the peptide pool could be done by HPLC. The peptides were separated by RP—HPLC on a Micro Gradient System (Model 130, Applied Biosystems, Foster City, CA) in order to reduce the number of predominant peptides in a fraction so that they may be sequenced by Edman degradation. Absorbance at 214 nm was monitored (Spectroflow 783 Programmable Absorbance Detector, Kratos Analytical Instruments, Chestnut Ridge, NY) and recorded (Linear 1200 Chart Recorder with the settings 0.077 range, 87 5 mV, and 15 cm/hour chart speed, Alltech Associates, Deerfield, IL). Peptides were separated using a 0.8 mm x 25.0 centimeter (cm) Vydac® C-18 reverse phase column (L.C. Packings, San Francisco, CA). The column was equilibrated for at least 20 minutes with a 2% solution B [90% sequence grade acetonitrile (EM Science, Gibbstown, NJ), 0.087 to 0.098% TFA (Pierce, Rockford, IL) in HPLC grade water (J. T. Baker, Phillipsburg, NJ)] in solution A (0.1% TFA in water) with a flow rate of 40 ul/minute at room temperature. The sample was mixed with 160-200 ul of solution A, spun at 10,000 x g for 20 minutes, and then 200 ul of sample was injected into the Micro Gradient System with a 200 ul injection loop. The following elution gradient was used: the column was held at 2% solution B for 5 minutes. Solution B was linearly increased from 2% to 37% in 60 minutes. Solution B was then linearly increased from 37% to 75% in 30 minutes. The gradient was continued from 75% to 95% solution B in 15 minutes. Finally, the gradient was held at 95% solution B for 10 minutes (modified from (Tomlinson et al., 1996). Fractions were collected into disposable borosilicate glass 12 x 75 mm culture tubes (Fisherbrand®, Fisher Scientific, Pittsburgh, PA), dried in a vacuum centrifuge, and stored at -20°C to - 88 70°C. Selected fractions were sequenced by N—terminal sequencing. HPLC fractions purified by magnet bead immunoaffinity chromatography, molecular weight filtration, and then separation by HPLC were selected for N-terminal protein sequencing (Edman degradation reaction) (Procisem Protein Sequencer, Applied Biosystems, Foster City, CA). Only the first 9 to 10 amino acids were identified from the peptides. Samples were sequenced at the Genomics Technology and Support Facility at Michigan State University, East Lansing, MI. Human Hsp70/Hsc70 and Grp94 were also isolated by magnetic bead immunoaffinity purification. The antibodies used are listed in Table 1. The purification of the peptides was by molecular weight filtration. Peptides were sequenced by the N-terminal sequencing Edman degradation reaction. Samples were sequenced at the Genomics Technology and Support Facility at Michigan State University, East Lansing, MI. 3.3.2. Purification of Peptides by Analytical RP- HPLC Peptides were also purified from the larger molecular weight proteins by an alternative method using analytical scale RP-HPLC separation. This was accomplished by 89 denaturing the MHC peptide complex supernatant collected by magnetic bead immunoaffinity purification with the addition of glacial acetic acid to a final concentration of 10% (v/v) acetic acid and placed in boiling water for five minutes. The samples were dried to approximately 5 ul using a vacuum centrifuge. Peptides were separated using a 4.6 mm x 25.0 cm protein/peptide C-18 reverse phase analytical column (Vydac®, Hesperia, CA). The column was equilibrated for 20 minutes with a 2% solution B (90% acetonitrile, 0.098% TFA in water) in solution A (0.1% TFA in water) at a flow rate of 1 ml/minute operating at room temperature. The sample was mixed with 110 ul of solution A and 100 ul of sample was autoinjected into the column using two (one for solvent A and one for solvent B) Solvent Delivery Systems (Waters Model 510, Waters, Milford, MA), an Automated Gradient Controller (Waters, Milford, MA), and the auto sampler (Waters 717 plus, Waters, Milford, MA). The following elution gradient was used: the column was held at 2% solution B for 3 minutes. Solution B was linearly increased from 2% to 37% in 57 minutes. Solution B was then linearly increased from 37% to 75% in 35 minutes. The gradient was continued from 75% to 95% solution B in 5 minutes. The gradient was finally held at 95% solution B for 5 minutes. Absorbance was monitored at 90 214 nm (Waters 2487 Dual Lambda Absorbance Detector, Waters, Milford, MA). Data was recorded using computer software (System Gold“ Chromatography Software, Beckmann Instruments Inc., San Ramon, CA). Fractions were collected (without visualization of actual peaks as the detection limit of this system was above estimated recovery for most peptides) into disposable borosilicate glass 12 x 75 mm culture tubes, dried in a vacuum centrifuge with no heat, and stored at —20°C to —70°C. Fractions with similar elution times as bradykinin (a 9 amino acid long peptide which is similar in length to an MHC class I associated peptide of 8-11 amino acids) (Sigma, St. Louis, M0) were selected for identification by liquid chromatography tandem mass spectrometry (LC-MS/MS). 3.4. Characterization of Peptides by LC-MS/MS The pool of low molecular weight peptides (from either the magnet bead immunoaffinity analytical RP-HPLC protocol or the mild acid wash (protocol to follow)) were separated and directly identified using a capillary liquid chromatography system (Waters CapLC”, Waters Corp., Milford, MA) coupled to electrospray ionization (ESI) mass spectrometer (LCQmDeca, ThermoQuest Finnigan, San Jose, CA). 91 3.4.1. Capillary Liquid Chromatography System Peptides were separated by RP—HPLC using a capillary liquid chromatography system. The system was run with computer software (MassLynx Control v 3.5 and CapLC” Diagnostics computer software, Waters, Milford, MA). Using 125 um inner diameter (ID) or 65um ID polyetheretherketone polymer (PEEKW) tubing (UpChurch Scientific, Oak Harbor, WA), the HPLC solvents A and B were mixed with an external gradient mixer and the flow was split from 7—20 ul/minute to 500 nanoliter (nl)/minute using a PEEK” Tee (UpChurch Scientific, Oak Harbor, WA) for the flow split. Fused silica tubing (360 um outer diameter (OD) x 20—75 um ID) (Polymicro Technologies LLC, Phoenix, AZ) was used for the flow split with the ratio of the length of the tubing vs. backpressure to obtain the correct flow rate of 500 nl/minute. PEEK” tubing (65 um ID) connected the rest of the gradient from the flow split to the injector. The injector was part of the mounting platform (PicoViewm, New Objective Inc., Cambridge, MA) that included the injection valve with a 500 nl loop (Rheodyne Model 7520, Rheodyne, Rohnert Park, CA), High Voltage Precolumn Module (HV), a stage holder to position the needle and adjust it in the x, y, and 2 planes, and a camera with a light source to monitor the spray. The voltage (1.7-2.5 kilovolts (kV)) was 92 applied to the sample before the column. The column was a 75 um ID column with a 15 +/- 1.5 um uncoated fused silica tip (PicoFritm, New Objective Inc., Cambridge, MA). The column was self packed using a stirred-slurry column packer pump (Model 705, Micromeritics Instrument Corp., Norcross, GA) pressurized with nitrogen gas (The BOC Group Inc., Murray Hill, NJ). The packing material was either 5 um, 300 A pore reverse phase C-18 packing (Vydac®, Hesperia, CA) or 10 um reversed phase packing (Poros R2, PerSeptive Biosystems, Framingham, MA) particles. Reverse phase particle are used to separate peptides based on the interaction between the hydrophobic stationary phase and the mobile phase. HPLC grade MeOH (J. T. Baker, Phillipsburg, NJ) was added to the packing material in a 1 ml glass vial with internal cone (Reacti—Vial”, Pierce Biotechnology Inc., Rockford, IL) to create a slurry. The slurry in MeOH was sonicated. The MeOH was decanted and replaced with fresh methanol twice. The cleared slurry was stirred with a magnetic stir bar in the nitrogen pressurized pump (200-300 psi) and the flow of the packing material was watched under a light microscope at 40 X magnification until the column was packed to the desired length (~3 cm for the Vydac® column or ~12 cm for the Poros R2 column). The slurry was 93 removed and final packing of the material in the column was finished in the pressurized pump. The column was installed on the platform after the HV module. Sample volume injected was 0.5 ul to 1.0 ul along with one of the peptide standards (a standard peptide, angiotensin (100 fm) (Sigma, St. Louis, MO) or bradykinin (250 fm) was added as a control). Peptides were eluted from the column into the mass spectrometer with a linear gradient of 2—30% solvent B [95% acetonitrile (EM Science, Gibbstown, NJ)/0.1% formic acid (J. T. Baker, Phillipsburg, NJ) in HPLC grade water (J. T. Baker, Phillipsburg, NJ)] in solvent A (0.1% formic acid in HPLC grade water) in 25 minutes. The gradient was held at 30% solvent B for 10 minute. The linear gradient was continued from 30 to 80% within 25 minutes. Mass spectrometry (MS) data was collected during the entire 60 minute run. 3.4.2. Mass Spectrometry The tip of the column was positioned in front of the orbit of the mass spectrometer. The mass spectrometer was tuned with angiotensin and a 2.5 kV spray voltage. The instrument was calibrated using the E81 Calibration Solution that included UltraMark 1621, a peptide with the sequence MRFA, and caffeine (Sigma, St. Louis, MO). The mass spectrometer was operated using the XcaliburTM software 94 package (ThermoQuest Finnigan, San Jose, CA). The Qual Browser in the Xcalibur” software was used to visualize data. The mass spectrometer was operated in one of two modes, which were the Triple Play or Big Three mode. The details of these modes will be described in the following paragraphs Using triple play mode (full, zoom, and MS/MS scans), full scans were acquired from a range of 250-2000 mass to charge ratio (m/z). A data dependent zoom scan was obtained to determine the charge state of the ion. Collision activated dissociation (CAD) spectra were obtained on the most intense ion above background. Following convention, on the fly, Dynamic Exclusion” was turned on (in order to not repeat multiple scans of the same ion) with a mass width of +/-1.0 Da (limit of the mass range to determine by mass if an ion has been repeated) and a repeat count of 2 (meaning that CAD spectra will be obtained on the same ion twice before it is excluded). The exclusion duration was for 15 minutes. The Normalized Collision Energy” for fragmentation was 35%. Normalized Collision Energy” ensures that the optimum energy is applied to each ion, independent of m/z. Full scans were acquired from 400-2000 m/z using the Big 3 mode. Following convention, data dependent CAD spectra were obtained on the 3 most intense ions above 95 background. This maximizes the number of CAD. On the fly, Dynamic Exclusion” was turned on with a mass width of 1.5 and repeat count of 1. The exclusion duration was for 1.5 minutes. The Normalized Collision EnergyTM for fragmentation was 35%. 3.4.3. MS Data Analysis The MS/MS data was searched against the non—redundant (nr) database using Turbo SeQuest v.27 in the Sequest Browser (ThermoQuest Finnigan, San Jose, CA), or Mascot from Matrix Science (Matrix Science Ltd., London, UK) (Perkins et al., 1999). The entire database was searched unless the amount of data searched exceeded the limitations of the search programs and then only the human and/or mouse databases was searched. Using SEQUEST, data was searched against the nr database with no enzyme selected, fragment ion tolerance set to 1 Da, and peptide mass tolerance set to 2.5 Da. Data with a cross correlation (Xcorr) value (a measure of the quality of the assigned sequence) equal to or greater than 1.74 was selected for possible further analysis and sequence assignment/confirmation was attempted. An Xcorr of 1.74 was chosen based on the observation that spectra scoring lower than 1.74 could not be easily identified by de novo sequencing (personal observation). 96 Using Mascot, data was searched against the nr database with no enzyme selected, no modifications, up to one missed cleavages was allowed, peptide tolerance set to +/- 2.0 Da, and MS/MS tolerance set to +/- 0.8 Da. Data indicating identity or homology based on the Mowse score (a probability score based on the identified peptide calculated by Mascot) (Perkins et al., 1999) were selected for possible further analysis and sequence assignment/confirmation was attempted. Sequence assignment was also accomplished by de novo sequencing to identify or verify a sequence. Possible complementary b and y ions were identified. By either manual calculation or with the aide of MS-tag (Protein Prospector, University of California, San Francisco, Mass Spectrometry Facility) (http://prospector.ucsf.edu) (searched against the nr, OWL, and/or SwissProt database against all entries or the human and/or mouse subsets) peptide sequence was assigned. 4. Mild Acid Wash to purify peptides from MHC 4.1. Extraction of MHC Associated Peptides by a Mild Acid Wash Procedure Approximately 3 x 109 (or equivalent to 1 x 108 white blood cells by HLA density) platelets, 1 x 108 WBC, or 1 x 108 RBC are centrifuged at 200-2000 x g for 10 minutes in a 97 microcentrifuge tube and the supernatant was removed. Platelets or cells were incubated in 200—500 ul of citric acid buffer [0.067 M citric acid monohydrate (Mallinckrodt Chemical Co., Paris, KY) and 0.123 M sodium phosphate monobasic (NaHfiXh) (Mallinckrodt Chemical Co., Paris, KY) adjusted with NaOH to pH 3.1] rotating for 10 minutes at room temperature. Citric acid buffer will denature surface MHC molecules releasing MHC associated peptides into the supernatant (Skipper et al., 1999). Platelets or cells were pelleted at 200-2000 x g for 2 minutes. The supernatant containing the peptides was removed and either used immediately or stored at —70°C. 4.2. Purification of Peptides Using a Trap Column Low molecular weight peptides, separated from salts and large proteins, were concentrated using RP-HPLC applying a step gradient. Peptides were purified using a 3.2 x 15 mm reversed phase C—18 peptide trap column (MPLC® NewGuard® Brownlee column, Applied Biosystems, Foster City, CA). Using a Micro Gradient System with 200 ul injection loop, 200-500 ul of sample were injected with a flow rate of 75 ul per minute. A step gradient consisting of 3 steps was applied (5 minutes at 2%, 12 minutes at 35% and 23 minutes at 95% to wash the trap) of 95% acetonitrile, 0.098% TFA in HPLC 98 grade water (solution B) in solution A (0.1% TFA in HPLC grade water). The absorbance was monitored at 214 nm by either a Spectroflow 783 Programmable Absorbance Detector or a 785A Programmable Absorbance Detector (Applied Biosystems, Foster City, CA). Data was recorded on a chart recorder with the settings from zero to 0.077 - 1.5 range of scale for absorbance units, 10 millivolts (mV), and 1 cm/minute chart speed or by the computer software, (CHROMELEONW v. 6.40 Chromatography Management System, Dionex Corp., Sunnyvale, CA). The peptide fractions were collected and either stored at —70°C or immediately concentrated in a Vacuum centrifuge to 2—5 ul and then stored at ~70°C. Fractions with similar elution times as bradykinin (a 9 amino acid long peptide which is similar in length to an MHC class I associated peptide) (Sigma, St. Louis, M0) were selected for identification by LC-MS/MS as previously described. 5. Peptide Verification A BLAST search was used to identify the peptides. The SYFPEITHI database (http://syfpeithi.bmi- heidelberg.com/scripts/MHCServer.dll/open.htm) (Rammensee et al., 1999) was used to compare the MHC binding motif of the individual with the peptide characterized. 99 To verify the identity of select peptides eluted from the cellular extract, a synthetic peptide with the same amino acid sequence of the eluted MHC associated peptide was synthesized and the CAD spectra of the synthetic peptide was compared to the CAD spectra of the sample. Select peptide sequences were synthesized and the synthetic peptide CAD spectra were compared to the CAD spectra of the platelet peptides in order to help confirm the identity of the peptide. The synthetic peptide was also spiked into the sample and co—elution of the synthetic and sample peptide was confirmed. Synthetic peptides were synthesized at the Genomics Technology Support Facility at Michigan State University, East Lansing, MI (SYNERGYW Personal Peptide Synthesizer, Applied Biosystems, Foster City, CA). RANKPEP is a search algorithm that predicts the top 2% of peptides from a protein sequence that would bind a particular HLA allele (http://www.mifoundation.org/Tools/rankpep.html) (Reche et al., 2002). Using the RANKPEP algorithm, peptides were identified from proteins sequences that were predicted to bind certain HLA alleles based on their peptide binding motif. 100 RESULTS 1. Antibody Purification 1.1 Affinity Chromatography 1.1.1 W6/32 Antibody Purification From archived ascites fluid, W6/32 MAb was purified by Protein A affinity chromatography (Figure 2). The absorbance of the collected eluate at 280 nm was 0.566. Using the extinction coefficient for immunoglobulins (an OD at 280 nm of 1.4 equals 1 mg/ml of antibody) up to 0.404 mg of antibody/ml was recovered. The antibody was collected in a volume of 25 ml. Therefore, up to 10 mg of W6/32 antibody was purified and recovered from the ascites supernatant. After concentration of the sample, the absorbance at 280 nm was determined to be 2.470. This gives a final antibody concentration of 1.76 mg/ml as listed in Table 2. From cell culture supernatant, W6/32 MAb was also purified by Protein A affinity chromatography (Figure 3). From 20 ml of cell culture supernatant, 1.69 mg of W6/32 antibody was purified and recovered (see Table 2). SDS—PAGE was used to estimate the purity of each antibody. Protein bands at 22 kDa and 55 kDa, the molecular weight of the heavy and light chain of IgG 101 "'i IAIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIEL. I vi ii I” l. l j‘,‘i+1‘1.i.wi l .l 0.08- “first l "77? l l ‘ hi.” V X i i 1" 1"). ii i- 1...; -1- 54-1 m5 1 in ii" 007_ .l. ..-‘ mill”, I.-. I..._'. ‘ ..i.i.i-..-.i.-', DEAN—is Lilli P ;i _, ..... Ll , IX . ii) . . I . o f‘XXI)‘ X i 'E ”ii 006- XHW.LIIL i‘X u--l}.i,l, . l . 1.. l ,. ,. . ,§x ‘ i i, .Eliu‘a’t’e'Lga't .prI: 3.0. 315.7]: .. 0.05- i ;, fl) . ' 0.04- I l 0.03- X ‘ 9’ c T ....... 00L 41. . . fll‘q,;‘ {'i F.Trw.n_i 'iil. 0.2mm/minute (one square=0.2mm) Figure 2. W6/32 MAb Purified from Ascites. W6/32 MAb was isolated by protein A immunoaffinity chromatography. The flow rate was determined by gravity. The first peak, as monitored at 280 nm, of the chromatogram represents the flow thru of the unbound protein. The second peak represents the protein (W6/32) that was eluted from protein A with PBS pH 3.0. 10 mg of protein was recovered. Absorbance at 280nm S 102 Table 2. Antibody Purification Yield The source and amount of starting material is listed. The amount of antibody purified was determined by 280 nm absorbance. Antibody activity was determined by complement mediated cytotoxicity. [] = concentration. *HB116 biological activity was unable to be determined as the antibody does not activate complement. Ab Source Starting Ab Recovery Biologic amount Activity Pre[] Post[] W6/32 Ascites 1-5 ml 10 mg N/A Yes fluid W6/32 Cell culture 20 ad. 1.69img N/A Yes supernatant HB116 Ascites 1-5 ml N/A 0.77 *N/A fluid mg/ml L243 Ascites 2.5 ml 0.84 mg 0.56mg Yes fluid L243 Cell culture 20 ml 0.18 mg N/A Yes supernatant immunoglobulin, were present on the gels. The results of the SDS-PAGE for each purified antibody will be described separately. SDS-PAGE confirmed the presence of the W6/32 antibody and estimated the purity of the sample as compared to a commercially purified antibody (SPA822) (Figure 4). 103 0J0- 9 O f "FIow7Ehru 9 c T l I ,_.._ __-___..__,..,_.,_,. ,l 0.07-: . .- 0.06-~i - W Eluate‘ at pH 3.0 3 | 0.05: l ‘ .. . i .r°l 0.04-. ,. 1 ill 0032. X Liéyir (102—~) :~ ll: :g~ ODIqt"r“aa I . . i).. t ll: U: i 0-- _7 ‘7” Time—5 0.2mm/minute (one square=0.2mm) Absorbance at 280nm Figure 3. W6/32 MAb Purification from Cell Culture W6/32 MAb was isolated by protein A immunoaffinity chromatography. The flow rate was determined by gravity. The first peak, as monitored at 280 nm, of the chromatogram represents the flow thru of the unbound protein. The second peak represents the protein (W6/32) that was eluted from protein A with PBS pH 3.0. From 20 ml of cell culture, 1.69 mg of protein was recovered. 104 Ig Heavy ‘i‘chain I ‘____ I*.Ig Light )‘chain ‘X ‘— iiiiw ‘ Iii) “%WL I“ In)" ‘ MN I: III II i WWI II I III I III‘ “III) WWII)“ H \W‘Mliiliiiii) A. Ab Purified from cell culture kDa N N G) 4 D4 U1 :‘Imarker 94 ’“ . Ig Heavy 67 I‘U‘chain ' , <———- 43 “if” . N. Ig Light 30 chain 20.1 “ “I.“ I 14.4 “lwfi' II); ‘ B. Ab from Commercial Reagent Figure 4. SDS-PAGE of commercial vs. cell culture purified MAb Samples were run on a reduced 10% SDS—PAGE silver stained. (A) 1.6 ug of cell culture purified W6/32 and (B) 2 ug of commercially available SPA822 were loaded on the gel. IgG heavy chain is 55 kDa and IgG light chain is 22 kDa as indicated with arrows for each. 1.1.2. HB116 Antibody Purification From archived ascites fluid, HB116 MAb was purified by Protein A affinity chromatography. The absorbance of the eluate at 280 nm was not determined until after the volume of the sample was concentrated. was 1.075. The absorbance at 280 nm An OD at 280 of 1.075 equals 0.77 mg/ml of 105 antibody (see Table 2). Presence of the antibody and estimation of the purity was assessed by SDS-PAGE and determined to be pure enough, as compared to commercial reagent, to be used for immunoaffinity chromatography (Figure 5). $4 (U Ad 54 (U E H8116 kDa 94 67 43 3O 4———— Ig Heavy chain 4———- Ig Light chain 20.1 14.4 Figure 5. SDS-PAGE of Ascites Purified HB116 MAb Samples were run on a reduced 10% SDS-PAGE silver stained. 1.55 ug of HB116 was loaded on the gel. IgG heavy chain is 55 kDa and IgG light chain is 22 kDa as indicated with arrows. 106 1.1.3 L243 Antibody Purification From 2.5 ml of archived ascites fluid, L243 MAb was purified by Protein A affinity chromatography. The absorbance of the collected eluate at 280 nm was 0.031 in 38 ml. A total of up to 0.84 mg of L243 antibody was purified and recovered from the ascites supernatant. After concentration of the sample to 3.5 ml, the absorbance at 280 nm was determined to be 0.23. This gives a final antibody concentration of 0.16 mg/ml and recovery of 0.56 mg of antibody as listed in Table 2. Presence of the antibody and estimation of the purity was assessed by SDS— PAGE (Figure 6). L243 kDa -marker nm' ‘ ,m‘m MN. “W W” m “j 94 67 Ig Heavy chain———-F 4" 43 30 20.1 14.4 19 Light chain———->' \W’ ‘WMM‘N \ . " Ht HUN “1““ MU‘C“ f‘M‘ Figure 6. SDS—PAGE of Ascites Purified L243 MAb Samples were run on a reduced 10% SDS-PAGE silver stained. 0.96 ug of L243 was loaded on the gel. IgG heavy chain is 55 kDa and IgG light chain is 22 kDa as indicated with arrows. 107 From 20 ml of cell culture supernatant, L243 MAb was purified by Protein A affinity chromatography. The absorbance of the collected eluate at 280 nm was 0.103 in 2.5 ml. From 20 ml of cell culture supernatant, 0.184 mg of L243 antibody was purified and recovered (Table 2). SDS-PAGE verified the presence of the antibody and estimated the purity of the sample. 1.2. Microcytotoxicity Testing 1.2.1 W6/32 Microcytotoxicity testing incubates antibodies with cells in the presence of complement and tests for antibody dependent complement mediated lysis of the cells. The W6/32 MAb (1.76 mg/ml) isolated from archived ascites fluid was mixed in duplicate with control T cells at 1/10, 1/100, 1/1000, and 1/5000 dilutions of the stock antibody. After adding complement, 80-100% lysis of the cells occurred at 1/10 to 1/1000 dilutions. At the 1/5000 dilution, 51—80% of the cells were lysed (Table 3). This indicates that the purified W6/32 MAb bound its target. The W6/32 MAb (0.32 mg/ml) isolated from 20 ml of cell culture supernatant was mixed in duplicate with control T cells at 1/2, 1/10, 1/100, and 1/1000 dilutions of the 108 Table 3. Complement Mediated Cytotoxicity Testing MAb W6/32, HB116, L243, H58A, SPA822, and SPA850 were evaluated for antibody binding by microcytotoxicity testing. Source of antibody is listed. Antibodies were tested against either T or B cells as listed. Cell lysis (greater than 51%) at antibody dilutions indicated by “+”. Antibody' Source Cells Cell lysis at Ab dilution of: 1/2 1/10 1/100 1/1000 1/5000 W6/32 Ascites T N/A + + + + W6/32 Culture T + + + + N/A L243 Ascites T - N/A - - N/A L243 Ascites B + N/A + ._ N/A L243 Culture T - N/A - — N/A L243 Culture B + N/A + - N/A H8116 Ascites T N/A - - - - H58A Commercial T N/A + + + N/A SPA822 Commercial T N/A - — - N/A SPA850 Commercial T N/A - - - N/A stock antibody. After complement was added, 81-100% lysis of the cells occurred at 1/2 through 1/1000 dilutions (Table 3). This indicates that the cell culture purified W6/32 MAb also bound its target. 109 1.2.2. H8116 The HB116 MAb (0.77 mg/ml) isolated from archived ascites fluid was mixed in duplicate with control T cells at 1/10, 1/100, 1/1000, and l/SOOO dilutions. After complement was added, O-lO % lysis of the cells occurred at all dilutions (Table 3). Since the HB116 antibody is non cytotoxic, this test was unable to evaluate Ab binding. 1.2.3. L243 The anti—HLA class II DR MAb, L243 (0.16 mg/ml) isolated from archived ascites fluid, was mixed in duplicate with positive control B cells and negative control T cells at 1/2, 1/100, and 1/1000 dilutions of the stock antibody. After class II complement was added, 80- 100% lysis of the cells occurred at almost all concentrations (Table 3). This indicates that the purified L243 MAb can bind antigen. The L243 MAb (0.0736 mg/ml) isolated from 20 ml of cell culture supernatant was mixed in duplicate with positive control B cells and negative control T cells at 1/2, 1/100, and 1/1000 dilutions of the stock antibody. After class II complement was added, 81-100% lysis of the cells occurred at 1/2 and 1/100 dilutions of the B cells. At the 1:1000 dilution of the B cells and all dilutions of the T cells, 1-10% of the cells were lysed (Table 3). This 110 indicates that the purified L243 MAb was still able to bind antigen up to a 1:100 dilution and was specific for B cells. 1.2.4. H58A, SPA822, and SPA850 Microcytotoxicity testing of the commercially obtained antibodies was also performed. The H58A, SPA822, and SPA850 MAb (1 mg/ml) were mixed in duplicate with control T cells at 1/10, 1/100, and 1/1000 dilutions of the stock antibody. After complement was added, 81—100% lysis of the cells occurred at 1/10 to 1/1000 dilutions for the H58A antibody (Table 3). The SPA822 and SPA850 had 0-10% cell lysis at all dilutions (Table 3). These antibodies bind to intracellular proteins. 2. MHC Purification by Affi-gel®-10 Affinity Chromatography 2.1. MHC Purification Using W6/32 Using 1 x 108 EBV B cells, HLA molecules were isolated using the monoclonal antibody W6/32 coupled to 2 ml of Affi—gel®—10 matrix. The lysate from the EBV B cell line was incubated with an immunoaffinity column specific for HLA class I molecules. After dialysis, the absorbance at 280 nm was 0.012. Using the extinction coefficient for HLA of 1.5, up to 0.056 mg of class I HLA molecules was obtained from 1 x 108 111 cells (Table 4). SDS—PAGE was used to verify the presence of HLA class I by the appearance of bands at 45 kDa for the heavy chain and 12 kDa for 82M. The polyacrylamide gel also estimated the purity of the sample (Table 4). Table 4. Yield of MHC class I Purified by Affi—gel®-10 The source and number of cells used is listed. The amount of HLA purified was determined by 280 nm absorbance. Albumin was used as a negative control column. Interpretation of SDS-PAGE results are summarized. Cell # of Cells Affinity HLA Protein Type Column. IRecovery Recovery by in mg SDS-PAGE EBV B 1 x 108 W6/32 0.056 Proteins not cells correct size for HLA WBC#1 1.3 x 108 W6/32 0.250 No Proteins WBC#1 1 x 108 HB116 0.087 No Proteins WBC#1 6.5 x 107 Albumin 0.066 Protein Smear WBC#2 ‘From.40 Nd W6/32 0.472 Many Proteins whole blood WBC#3 From 40 ml Albumin 0.135 Many Proteins whole blood WBC#3 From 40 ml W6/32 0.060 Many Proteins whole blood after albumin column From 1.3 x 108 WBCs from individual #1, HLA molecules were isolated using the monoclonal antibody W6/32 coupled to 1.5 ml of Affi-gel®-10 matrix. The WBC lysate was run 112 over an immunoaffinity column specific for HLA class I molecules. Figure 7 illustrates the amount of HLA eluted from the column with 50 mM glycine pH 11.5. After dialysis, the recovery of HLA class I molecules was 0.25 mg from 1.3 x 108 cells (Table 4). SDS-PAGE was unable to estimate the purity of the sample (Table 4). Lysed WBCs isolated from 40 ml of whole blood from individual #2 was bound to a W6/32 column. From the W6/32 column eluted with 50 mM glycine/0.1% deoxycholate pH 11.5, 0.472 mg of protein was recovered (Table 4). Upon analysis by SDS-PAGE, there were several non-specific proteins eluting with the W6/32 column (Figure 8). 2.2. MHC Purification Using HB116 From the same individual #1, HLA molecules were isolated using another MHC class I monoclonal antibody, HB116, coupled to 1.5 ml of Affi-gel®-10 matrix. The cell lysate from 1 x 108 WBCs from individual #1 was run over an immunoaffinity column as before. HLA molecules were eluted from the column with 50 mM glycine pH 11.5. After dialysis, the recovery of HLA class I molecules was 0.0867 mg from 1 x 108 cells. SDS-PAGE was used to verified the presence of HLA class I and estimated the purity of the sample (Table 4). 113 A - -¢.— «noc- - o...—‘- - 1 1 ' ,. , . . ._.. .. , . (l3-‘ . j: (103- " .' i i ' l a ' ;. . ., . , , . H -' ‘ I i . . .L T Flow thru . . , ‘ «u . 4 . . r . . . ‘ u | -_.—— —— Wu... .-*_— ‘. . . . . . 0 I c ‘ o o . . a . _ . . a -. i . 1 O I - O - - v a t a ’ i I I I I I < __--h-—fi—¢-J Time 0.2mm/minute (one square=0.2mm) Absorbance at 280nm Figure 7. MHC Class I Purification by Affi-ge1®-10 From 1.3 x 108 white blood cells, from WBC #1, MHC class I was eluted. The flow rate was determined by gravity. The first peak, as monitored at 280 nm, of the chromatogram represents the flow thru of the unbound protein. The second peak represents the protein (HLA) that was eluted from W6/32 Affi-gel column with 50 mM glycine pH 11.5. 0.25 mg of HLA class I molecules were purified. Note the change in absorbance scale between flow thru and MHC eluate. 114 m J.) m m m u m 5 U (5 JJ H m 5 m m m H m >~ 0) >1 G H H 'H N E H m H :1 {—1 \ H ,0 m m w H U 3 U m N N m m :3: I”: it it o o o 0 kDa E m m m m 94 $WW 1% . “M“ u \“ WWW)“‘:‘{{{\)\l\\‘\‘\““" @va ‘ (f j M (11 ‘ ‘11)) M“ ~ "“ v“‘~“1\“““)i“ i“ w. . ‘:'¢(\x)=‘(((( ; 3° * ‘ (11)“ ~ (\(xil‘wl‘ (1 Figure 8. SDS—PAGE of HLA class I isolated by Affi—gel®—10 HLA class I molecules were purified, run on 10% SDS-PAGE, and silver stained. From WBC #2, cellular lysate and proteins eluted from a W6/32 column were run. From WBC #3, cellular lysate and proteins eluted from the albumin column were run. Comparison of the W6/32 and albumin (negative control) samples indicate HLA class I molecules were not purified. 115 2.3. MHC Purification Albumin Control Column As a negative control, 1 mg of albumin was bound to 1.5 ml of Affi—gel®-10 column material. Using 6.5 x 107 WBCs from individual #1, non-specific binding was estimated using albumin coupled to the Affi-gel®-10 matrix. The WBCs were lysed and the cell lysate was run over the column. Figure 9 illustrates the amount of non—specific protein eluted from the column with 50 mM glycine pH 11.5. After dialysis, the amount of eluted non-specific protein was 0.066 mg for 6.5 x 107 cells. SDS—PAGE was used to visualize the non-specific bound proteins (Table 4) 2.4. MHC Purification W6/32 and Albumin Column Using lysed WBCs isolated from 40 m1 of whole blood from individual #3, the lysate was incubated with an albumin/PBS column to reduce non specific binding and then this sample was bound to the W6/32 column. The eluted proteins from the albumin column contained 0.135 mg of protein. Upon SDS-PAGE analysis, several non—specific proteins eluted (Figure 8). When this lysate was run over a W6/32 column, 0.060 mg of protein was eluted still giving a non-specific protein pattern by SDS-PAGE (Table 4). 116 0.10- 'f'" ‘ ? H i J: z u : 0.09- g ' ‘ ” H h ': 0.08: ”j” 0.07- ; ' 0.06- f , 0.05— .. i 0.04- f 0.03-] T " ; C. ; ' '3? 8 f :1 N 0.02- g s 4 u , , £99.11. (U 1 ' 4) _i m ‘_ 1.)" LL; 2 0.01-- a .0. r6 ‘_ f “i7 .6 ~ .- 13, H , I- , 5,3 _ 11.11510, Q Time———-> 0.2mm/minute (one square=0.2mm) Figure 9. Albumin Negative Control Column Albumin bound to Affi—gel was used as a negative control for W6/32 columns. The flow rate was determined by gravity. The first peak, as monitored at 280 nm, of the chromatogram represents the flow thru of the unbound protein. The second peak represents the non-specific protein that was eluted from the albumin column at pH 11.5. As indicated, nonspecific protein eluted from this negative control albumin column. 117 3. MHC Purification by Immunomagnetic Chromatography 3.1. Magnetic Bead Cross Linking Protocol The W6/32 MAb (0.44 mg) was cross—linked to the magnetic beads. To test the effectiveness of the cross linking procedure, magnetic beads with cross linked Ab were incubated with whole WBC. The cells and beads were then viewed under a 40X light microscope and it was determined that the cells were not bound to the beads. Next cells were incubated with 0.44 mg of W6/32 antibody and then incubated with the magnetic beads (without W6/32 MAb) and viewed under a 40X light microscope. The magnetic beads were bound to the cells. Again, magnetic beads with cross linked Ab (3.52 mg of W6/32) were incubated with whole WBC. The cells and beads were then viewed under a 40X light microscope and the cells were not bound by the beads. Magnetic beads without cross linked Ab (3.52 mg of W6/32) were incubated with whole WBC. The cells and beads were then viewed under a 40X light microscope and the cells were bound by the beads. The cross linking procedure was determined to block the MAb from binding to MHC and was discontinued. 118 3.2. Immunomagnetic MHC and HSP Recovery MHC class 1, class II and Hsp70/Hsc70 proteins were isolated by magnetic bead immunoaffinity chromatography. The purity of the proteins was determined by SDS-PAGE. The polyacrylamide gel shown in Figure 10 illustrates the isolation of MHC class I proteins using the W6/32 MAb. The polyacrylamide gel was run reduced and non reduced to distinguish the IgG proteins (55 kDa heavy chain and 22 kDa light chain) from the MHC class I proteins (45 kDa heavy chain and 12 kDa 62M). MHC class II and Hsp70/Hsc70 proteins were purified using L243 and SPA822 MAb respectively. Figure 11 shows the proteins isolated and their molecular weights. 3.3. Peptide Purification MHC molecules were purified by immunomagnetic chromatography. MHC associated peptides were purified by one of two methods. These methods either separated the low molecular weight peptides from the large proteins by MW filtration, or the peptides were separated from the proteins by analytical RP-HPLC. 3.3.1. Peptide Purification by MW Filtration WBCs were isolated from whole blood from sample #1. MHC class I proteins were isolated by immunomagnetic chromatography with the W6/32 MAb and MHC class II with the 119 Figure 10. SDS-PAGE of HLA class I Purified by Magnetic Beads Illustrated is HLA class I molecules purified from W6/32 or HB116 bound to magnet beads. Figure A illustrates the reduced 10% gel with reduced IgG proteins at 55 and 22 kDa as indicated by the arrow. Figure B illustrates the non- reduced gel with the non-reduced IgG complex at 155 kDa (indicated by arrow). The molecular weight of MHC class I heavy chain is 45 kDa and 32M is 12 kDa in both the reduced and non-reduced SDS-PAGE (indicated by arrows). Both gels were silver stained. 120 kDa “ H “ “W111 111‘ ‘ ‘ 94 l‘ 1 W111 '1‘ ‘1 1‘ 1 “1 67 43 ‘ 11 ~‘ 1W,W W . Ig Heavy chain 1‘ 111HHMWWW1WLW> 11 HLA alpha chain 30 ‘ ‘ 1 ‘f11'1‘111“ 11111 111111 “‘1“ 11 1 1 1 1 ‘ “1W'W1 Ig Light chain 2 0 . l |:1. .111'1‘ W1 11" 11211 M1111 ‘ 1 , 1 ”111,111“ “11111111111“ 11‘ 1‘1““ 1““ 1111111111“ “1111““ ‘ 82M .1.11 11. B- Non-Reduced SDS-PAGE kDa lysate 1—1 1—1 a) U fl———— Ig Complex 94 67 43 f‘———— HLA alpha chain 82M 121 Marker EBV B cells Cell Lysate kDa alpha chain 1 1— beta chain Figure 11. Purified MHC class II using Magnetic Beads MHC class II proteins eluted from L243 magnetic beads were run on a reduced 10% SDS-PAGE silver stained. MHC class II alpha chain is 34 kDa and MHC class II beta chain is 29 kDa as indicated with arrows. 122 L243 MAb. Hsp70/Hsc70 with the SPA822 MAb were isolated from EBV B cells (sample #2). Associated peptides were released, separated from the large proteins, and the low molecular weight fraction was separated by RP—HPLC into an average of 30 peaks (Figure 12). The most abundant fraction(s) were selected for N-terminal protein sequencing of the first 9-10 amino acids. From fraction #10 (the most abundant HPLC fraction) of the MHC class I associated peptides, the most abundant peptide sequence was QEGNDVAKE (mw=990.0) and the second most abundant sequence was AKIKEIVDD (mw=1031.2) from this fraction. BLAST searches were not able to identify these peptides. Sample #2 peptides had the peptides identified from class II MHC proteins as ALGTVKRKET (mw=1103.3) and FDEKSGKSWE (mw=1213.3). A BLAST search was not able to identify these peptides. Sample #2 Hsp70/Hsc70 protein were purified and the most abundant fraction was selected for sequencing. The peptide identified was YIMYIRYIVR (mw=1390.7). A BLAST search was not able to identify this peptide. These peptides were compared to the SYFPEITHI database and the anchor residues of the peptide did not match the binding motif of the HLA molecule of the individual. 123 0031_ ‘ . . . ., 1; 1 «W: . ,‘ Fraction #10 I -1 11 selected for 1 1 1 *{1 1 se encin ————> 0.027- qu g 0023- - 1 1 f 11' 1" 1 0.019—- 1 .1‘ (111111 ‘11 11 1 1 1 1, 1 1‘ 1 1. 1‘ 5‘11 | . £111.11 1'" v' ' 0.015— - ‘ ‘ t" p O ._. t?) 0.004- 1 ‘ ' m , q, 1 0' “4'. Absorbance at 214nm Time 7 0.25 cm/minute (one square=0.2mm) Figure 12. HPLC Chromatograph of MW fractioned MHC class I peptides MHC class I molecules were purified from WBCs from sample #1 by magnetic beads. Associated peptides were fractioned by MW filtration and separated by RP-HPLC with the gradient 2% solution B for 5 minutes, 2—37% in 60 minutes, 37-75% in 30 minutes, 75—95% in 15 minutes, and hold at 95% for 10 minutes with a flow rate of 40ul/minute. Solution B was 90% acetonitrile, 0.087 to 0.098% TFA in solution A — 0.1% TFA. HPLC fraction #10, an abundant fraction (indicated by arrow and box) was selected for identification by Edman degradation protein sequencing. The most abundant peptide sequences from this fraction were QEGNDVAKE and AKIKEIVDD as determined by Edman degradation protein sequencing. Peptides were entered into the BLAST search engine, but the search did not identify the peptides. 124 After these peptides sequences were obtained and could not be identified, a negative control blank (HPLC solution A) was filtered through the MW filter. An HPLC profile of the blank (Figure 13) was obtained and found to appear similar to the HPLC profile of the peptide samples (Figure 12). Even with additional washing of the MW filter, contaminants were still obtained from the filter (Figure 14). An alternative RP-HPLC protocol with a 4.6 mm analytical column was used to eliminate the use of the MW filter. 3.3.2. MHC Associated Peptide Purification by Analytical RP-HPLC To eliminate the use of the MW filter while still separating low and high molecular weight proteins, a 4.6 mm analytical RP-HPLC column was used. The purified protein was denatured by acid and heat treatment and the sample was loaded onto the column. Peptides eluting within the first 60 minutes of the gradient (2—37% solution B) were collected in 5 minute intervals (Figure 15). A few of the most abundant fractions were selected for peptide sequencing by LC—MS/MS. Sequest analysis of the LC—MS/MS run could not identify peptides in sufficient quantity for analysis. 125 0.031- 0.027- 0.023- 0.019- 0.015- 0012- c c c t Absorbance at 214nm ?’ \\ L 7 Time 0.25 cm/minute (one square=0.2mm) Figure 13. RP-HPLC Chromatograph of MW filtered negative control A blank of solution A (0.1% TFA) was filtered through the 5,000 NMWL filter and used as a negative control. Contaminants in this blank were separated by RP-HPLC with the gradient 2% solution B for 5 minutes, 2—37% in 60 minutes, 37—75% in 30 minutes, 75-95% in 15 minutes, and hold at 95% for 10 minutes with a flow rate of 40ul/minute. Solution B was 90% acetonitrile, 0.087 to 0.098% TFA in solution A — 0.1% TFA. The graph illustrates contaminants that are eluted off the 5,000 NMWL filter. This MW filter is used to separate denatured MHC molecules from their associated peptide. 126 Absorbance at 214nm 0.031- 0.027- 1 0.023- 0019 ‘ :2‘pr,r-. .. 0_0]‘5_ #1,. ~- ~j~- >~n~~ ~~Av~~~n~vwr~r«--—~rs+"»_-i-.fl.l._,.u-. .1._,-uim;.-;kr_-_1_d.;-_,,h~,,, Time 7 0.25 cm/minute (one square=0.2mm) Figure 14. RP—HPLC Chromatograph of washed MW filtered negative control A blank of solution A (0.1% TFA) was filtered with a washed 5,000 NMWL filter to remove contaminants. Contaminants from this blank were separated by RP—HPLC with the gradient 2% solution B for 5 minutes, 2-37% in 60 minutes, 37-75% in 30 minutes, 75‘95% in 15 minutes, and hold at 95% for 10 minutes with a flow rate of 40ul/minute. Solution B was 90% acetonitrile, 0.087 to 0.098% TFA in solution A — 0.1% TFA. The graph illustrates contaminants are still eluted off the washed 5,000 NMWL filter. 127 020- 1 iFraction selected :for sequencing>/)" 1 1 010- Absorbance at 214nm 9’ 1o ‘? . 61 O. . N . Time in minutes 40- 60- 80- Figure 15. Analytical RP-HPLC Chromatograph Denatured MHC class I molecules and peptides were loaded directly onto a 4.6 mm RP-HPLC column. Peptides were separated from each other and from large proteins by RP— HPLC with the gradient 2% solution B for 3 minutes, 2-37% in 57 minutes, 37-75% in 35 minutes, 75-95% in 5 minutes, and hold at 95% for 5 minutes with a flow rate of 1 ml/minute. Solution B was 90% acetonitrile, 0.098% TFA in solution A - 0.1% TFA. Arrow indicated fractions selected for LC—MS/MS. 128 4. MHC Associated Peptide Purification by Mild Acid Wash Acid eluted MHC associated peptides were loaded onto the peptide trap. Fractions of peptides eluting at 35% solution B, where bradykinin had a similar retention time (bradykinin is a 9 amino acid peptide which is similar in length to an MHC class I associated peptide) (Figure 16) were analyzed by LC-MS/MS. Sequest analysis and de novo sequencing of the LC-MS/MS run could identify a few peptides in sufficient quantity for analysis. Further controls and ITP samples were obtained and analyzed. Sample size, HLA type and platelet autoantibody status are listed in Table 5, Table 6 and Table 7 respectively. 129 Absorbance at 214nm 0.152 ' :y 1 . 0J01 ‘ 0.05:7 12.Brad kinin _ ' “ ' 000 " Y" "' “'“' ' *TTT“Bradykinin 50 pmol €1.1TP #6 peptides ‘TTT‘Fraction selected L for LC-MS/MS 0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 Time in minutes Figure 16. RP—HPLC Chromatograph with a peptide trap column to purify MHC from peptides. Denatured MHC class I molecules and peptides isolated from platelets were loaded directly onto a RP-HPLC peptide trap, Line 1. The low mw peptides were eluted by a change in the gradient (2% to 35% solution B). The high mw proteins were eluted from the trap by changing the solvent to 95% solution B. Solution B was 95% acetonitrile, 0.098% TFA in solution A — 0.1% TFA. The flow rate was 75 ul/minute. Line 2 represents 50 pmol of bradykinin. The fraction collected from minute 19.5 to 21.1 (similar elution time to bradykinin) was analyzed by LC-MS/MS. Gray box indicates fraction selected for LC—MS/MS. 130 Table 5. Amount of Platelets and WBCs Used. Whole blood was collected in EDTA or ACD anticoagulant. Cells were isolated from 28-93.5 ml of whole blood. Cells were counted on a particle counter or a hemocytometer and used immediately for MHC associated peptide purification. Sample Platelets obtained WBCs obtained Control #1 3.9 x 109 1.0 x 108 Control #1 2.7 x 109 1.2 x 108 Control #2 5.0 x 109 9.0 x 107 Control #3 2.7 x 109 1.0 x 108 Control #4 3.4 x 109 9.0 x 107 Control #5 9.0 x 109 1.4 x 108 Control #6 6.5 x 109 1.3 x 108 ITP #1 5.8 x 109 1.5 x 108 ITP #2 1.6 x 109 7.5 x 108 ITP #3 1.2 x 109 5.0 x 108 ITP #4 5.3 x 109 2.4 x 108 ITP #5 8.7 x 109 1.8 x 108 131 Table 6. HLA type of Individuals Control samples were typed by serological and/or molecular methods. ITP samples were typed by molecular methods. All samples were typed at the Immunohematology and Serology Laboratory at Michigan State University. An “X” denotes a blank. All molecular types were converted to serological equivalents. 3 denotes HLA A*0302 as determined by molecular typing. Sample HLA type of individual Control #1 A2,A29,B57,B60,Cw3,Cw7 Control #2 A2,A30,B60,B63,Cw2,Cw3 Control #3 A3, AX,Bl4, B8,Cw7,Cw8 Control #4 A11,A29, B7,B51,CwX,CwX Control #5 A1, A3, B7, B8,Cw7,CwX Control #6 A2, AX, B7, BX,Cw7,CwX ITP #1 A2, AX, B7,BZ7,Cw2,Cw7 ITP #2 A3, AX, B7, BX,Cw7,CwX ITP #3 A2,A24, B8,B40,Cw3,Cw7 ITP #4 1133,1124, B7,B40,Cw7,Cw2/15 ITP #5 A1,All, B7, BB,Cw7,CwX 132 Table 7. Platelet Autoantibody Status Platelet autoantibody levels were determined by the radiometric micro technique at the time of sample acquisition by the Platelet Antibody Laboratory at Michigan State University, East Lansing, MI. Historic platelet antibody levels were available from the Platelet Antibody Laboratory where given. * notes these samples were obtained prior to the ITP focus of the study and therefore platelet autoantibody levels were not obtained. Sample Current Historic Platelet Platelet Auto Ab Auto Ab Control #1 *N/A N/A Control #2 *N/A N/A Control #3 *N/A N/A Control #4 No N/A Control #5 No N/A Control #6 No N/A ITP #1 Yes Yes ITP #2 Yes Yes ITP #3 Yes Yes ITP #4 No Yes ITP #5 No Yes 133 Four peptides listed in Table 8a were identified from the platelets of Control #1 with the MHC type: HLA A2, A29, B57, B60, Cw3, Cw7. The sequences of these peptides were determined by SEQUEST searched against the nr database and/or de novo sequencing to be (K/Q)EST(L/I)H(L/I)V(L/I) (Figure 17), (K/Q)EVPDACF (Figure 18), ASRPG(L/I) (L/I)F (Figure 19), and A(L/I)NE(L/I)(L/I)(Q/K)HV (Figure 20). Amino acids in parenthesis indicates two possible amino acids could be at that position. The isobaric amino acids L and I and amino acids K and Q can not be distinguished by this method. One peptide was identified from the WBCs of Control #1 (Table 8a). The sequences of this peptides was determined to be (K/Q)EST(L/I)H(L/I)V(L/I) (Figure 17). A BLAST search of these sequences identifies them as ubiquitin 63-71 (KESTLHLVL) Accession number P02248 (Wiborg et al., 1985), thrombospondin 593-600 (KEVPDACF) Accession number NP_003237 (Dixit et al., 1986), platelet factor-4 9- 16 (ASRPGLLF) Accession number NP_002610 (Poncz et al., 1987), and talin 777—785 (ALNELLQHV) Accession number AAD13152 (Mao et al., 1998). The SYFPEITHI database (http://www.uni-tuebingen.de/uni/kxi) was used to compare the MHC binding motif of the individual with the peptide (Rammensee et al., 1999). 134 Table 8a. HLA Associated Peptides Identified from Control #1 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets, WBC, and RBCs from Control #1 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. Sample Peptide AA. Protein Cell Pos Control#1 KESTLHLVL 63-Tl ubiquitin platelet ALNELLQHV 777-785 talin platelet ASRPGLLF 946 platelet factor 4 platelet KEVPDACF SEPGOO thrombospondin platelet KESTLHLVL 53-71 ubiquitin WBC None RBC 135 Odo 0:» Ge pmHQOH mum mcofl h cam Q Um>Hmeo .Umumfla mum mcofl quEmmHu s osm n 80L mmnnme oouosomno .Hs-mo ssossosns no q>qmqsmms nonmeusmos smsqm AH\AV>AH\AVmLH\AVBmmAO\¥L we mcHUcmsvmm 0>OE mp paw ummsvmm >3 mocmsqmm Umuoflpmum one .mm.ommnAN\Ev ofiumn mmhmso Cu mmmE m0 Esuuommm Odo one H6-mo senesosno .q>qmqemms Lo nanomem oso .sH musmflm NE. com com com com com oov oom 8N .LL..L.L.LL1LLLIL1ELIE1LLL 11?. “11.11 ...IL .L.... .rL_1.ILL_ L..t LL..1L1L...L.L1%L. ...)HLHFLLtrLL1sL1LIhLLILLL1tsy1LLL1tLLar ..n .rLML...1.r.1LM....L11_F-—.1mL_..L,.Ir.L _.+L.1F_- trL .rL1..LLLL L 11L... [1.1L1LLL11LLLLL1FLHL1TI O _ m3 1 L. 63.. L . 1 . L , m mom 3% L N+on - ov NA“ 0 sex. w mu m 1100 m . + H L , ow o3... L o _ s n N. o moomousfmév .s N2 :8 is as 48 666 8L :6 82 3 H3 .3 HI HSH :9 H02 mmow mom wow wow 0mm 03 mvm wmm mg on eouepunqv 911119193 136 .Esuuomam ado msu EN Umaman mum mcofl > 6cm Q Um>ummno .Umumfla mum mcofl ucmsmmnw % Ucm Q new mmmmmE Umuoflvmum .oom1mmm CHUCOQmOQEounu mm mUQ¢m>mM mmflwflucmvfi Hmdqm .mu¢0m>mAO\Mv mfl mcflocmsvmm D>ou mb >3 mucmsvmm UmpuaUmHQ mzh .m.¢mvuLN\Ev ONLMH wmumno Ou mmmE Mo Esnuommm Odo mnH oom1mmm cflvcommoaeounu .mu4om>mx Lo muuummm GNU .mN mLsmNm &5 25 com 8m 8m 8... 8m _ 8N HMQW1L111LL$111-L111Q11awfi%Ljr411L311114L31311-1N:4LJL%;%N1.L 11MMMWLMWMM o «mow L1138 $8 $3 1_ L 1.1_ 1 SN” L, SN m. x: L N.Nmo m3 :1 1N Em L P» NSN a n «.26 m .1L m N m m > _ > mmYVLL Q $63” a 1.- CV Wu. L W W L 1.11 cm W e U 1 _ W 11 N48 _ 8 ad: m o> 1 33 1 2: QEHLINEV c> mm: mmN ovm mmv Nmm me own mom LHOHSQEHZM H02 mom mvn ovm mmm Vmw 5mm mmN mg as 137 .Eduuommm ado map GH UmHQOH mum mnofi h UGm n Um>Hmeo .Umumfla mum mcofl ucmEmmum x 6cm Q MOM mmmmmE UmuUvaHm .mH-m JV HOuumm umamumHQ mm mddwmmmd mmflwflucmvfl .Hmfim ....ZHEHV AH\Avwmmm< m...“ mcflocmfivwm 0>OE mU >3 mucmsvmm Umuoflvmum mSB .m.om¢uAN\Ev oflumu mmumno on mmmE mo Esnuommm ado $38 ma-m .¢ Monomm umamumfim .mqummmm mo muuommm a¢u .mH musmfim NE. cow can con 0mm com com com omv oov omm com 0mm CON om? cow :L. _ C r _ _L_ C P». C Fp F CF rkpxk _ E._.L%F_ _L _ C r f37;i1.....:_._.j_.___fi..f__ i L TfiiLL ink [[E w b H , ..1 O Namzmfivafi 3% n A 38 wwwvvmemidov J omrmoflmoévm 5.8% E H mi 38 . 3» 32 moms Fmvm mywmmMXu I B. on E» S 1 3w. _ 38 m m mdmm m. 00W y u T nv _ Howw _ m m _ woe m.mm® E momfiaA 12>; ; 9: EN Nam 9% 2% NE 2: 5w b; a. 9 a mg m? row mam Nwm mow NSV 3m mm— NA :3 138 .Edhuommm adv mnu CH UmHQOH mum mcofl x Unm Q Um>kmmno .Umumfla mam mcofi ucmEmmuw h Ucm Q HOw mmmmmfi Umuoflvmum mmfiMNquUfl Bmdflm >3 mocmdvmm UmuUHUmHQ mnH .mmbubbb GflHMu mm >mOAQm244 .>mAM\OVAH\AVAH\AszAH\AV< mfl mcflucmzvmm o>om mu Ucm ummsvmm .H.mamnAN\Ev ofiumu mmnmno Ou mmmE mo Esuuomam 040 mSB mmN-N>N cflamu .>moqqmzq¢ No mguummm adv .ON musmflm NE. 82 com com o2 08 8m 8w 08 SN {brLLNIrtfiL:tbttt‘tLL‘tLtbitttrerttirLJ:N+fit¢1b$TtkrrgfiawfflkyuV waLrtLttLtbtttr¢tttttt_ O __ __ u _ ego :05, é____vfi_:{ : N _ # fl Nwa _ on? m m . L _. g _ _ mNommmmmNommN 38 ON . _ NSN . . . _ MEN“. “ r 938“ E m> $8 _ 3E . .: mp N> Ne w m N.Nmm n, ma . o 8m _,. 8m." _ * 38 P» t a S Ndoo " N v m> * Homm _ W m I Vloww i m a fiomv .1oor i temuafNév 5» m: mmm mmm omv moo mmn wa new one > HI :0 H3 HE E 4.26:: mmor m5 an vmo Sum wNv mam mmw N5 ,5 139 The KESTLHLVL (Figure 21) and KEVPDACF (Figure 22) peptide sequences were synthesized and the synthetic peptide CAD spectra were compared to the CAD spectra of the platelet peptides and found to be identical. The identity of the peptide was also confirmed by spiking both of the synthetic peptides into the platelet sample and co-elution of the corresponding platelet peptide and synthetic peptide was observed. 140 Figure 21. CAD Spectra of KESTLHLVL and synthetic peptide A. KESTLHLVL peptide identified from platelet sample. Predicted masses for b and y fragment ions are listed observed ions are labeled in the CAD spectra. B. Synthetic peptide, KESTLHLVL, was synthesized and analyzed by LC-MS/MS. The CAD spectrum is similar to CAD spectra from the platelet sample in Figure 21A. Predicted masses for b and y fragment ions are listed and the and observed ions are labeled in the CAD spectra. b8 and y2 ions were probably not detected in both spectra as they are low abundance ions. 141 NE. com cow 2: com 8m 8.. _ 8m 98 , E A: ET memrrgmmmfiwmoiiflwo «mm awn. .AFqu... TEL A+w,.+imN.i .FmN mFm...mo FFm A m> m A_ A 8m as N N> AwON w 85> .08 D A F3. _ o n A- a > Na 0 mmA «L. D H m. D W+ . .i av M. .moF. A- w n H o. A N. g :8 u A : m F... $3 Nm. OleafNéo 25qu 28593 .m NE. 8m 8m 02 80 8m 8.. 8m 8N -...t..it.fFTLTrrr: F A T44.A.‘A_.-L:.l.....LLZF..,.;F i r tTiTJFT ....t....F-,.TTI..A.T.._. L.},._T L T [ArrLLTLJtLAFFn , r L14... 0 mFmAFmeom .oFmAT .me waA ““ch .23 m8 $.qu A: A A. .FflmAAwnm. .omN .NFN. EFT w an A 5% A Fat or A _. .F_?” ND noN % > .mom an A A __ o N M n .08 A A A N+ Q _- 9. w... n 03 CV? W. W A A v8 m .vmv A r m c> N ma .mov . cor NmF me gm 5? vmm mam man I‘m mmor N N. o B. ONmuafNév .3 H53 H1 .23. 5.9 H02 mmor mom mow 0mm wow 03» mvm me mNF an @238 8.9—NE .< 142 Figure 22. CAD Spectra of KEVPDACF and synthetic peptide A. KEVPDACF peptide identified from platelet sample. Predicted masses for b and y fragment ions are listed and observed ions are labeled in the CAD spectra. B. Synthetic peptide, KEVPDACF, was synthesized and analyzed by LC-MS/MS. The CAD spectrum is similar to the CAD spectra from the platelet sample in Figure 22A. Predicted masses for b and y fragment ions are listed and observed ions are labeled in the CAD spectra. b1,b2 and b6 ions were probably not detected in both spectra as they are low abundance ions. 143 o8 o8 o8 SF 2:. So So 8m. So SF. SF. 8m 8..” SN 8N 8F .. ATFLT. T... .AL.T_ F .F. T FIAT. ...TATTT...-... TTATJFTLT T F. TAFT . L.- .F T A: Fara...» ...T. .LTA___ .1 AT Tr L. rT. _TFT 0 $3 LLFLmFL>.N.oE mcmowANmm Nam F...m.3.ATN.oFM. A AF.me F.oF.NFFmF 9me m o o.N . . NAFmNA F .78 N 2.8. mm D m> ma F. 8» NW m mm A ......I O¢W A F 8 m A m A ..Ll cm «A... m. vafiafNév we. as maroon oAchFc>m .m com 8m 2: 08 com one o8 8N _ .LL. TT LLFLL- TL. ..FTLLF. .LTTTF . ..FL.L.L- L.. A! AT..L....T...A..F...Tm..-...L...FLF..t._LLL...:..L_.L_LLTr...FFFTF...~... .F_..._L.LTT. ..F.».L.L..FL.. ......tL... TLIL.F-..-.FL.TLLAF....ALFLT.L..FL_LLT... o L.FN8 A FAA A38. A A AA A. .031; ..AAALfiFFAA F A A. .. N.mNN m. NF 38 A «ANS. :8 A . 3mm F.ANmmmomm 3A A AAA. Nwwmn FFF.N D 8 ”FE A NNmm on A AANSN m> NmmN m. N 08¢ A m . _ A A m N F. l > A, > 33A A D $3 9 AL. SW A N Am A , ..L.- 8 w A > . A L m 38 A A W 8 $6 A L ..> SF 8N as we. Nmm F8 8F 8m .. A F. a. $8 “_FoFSSnAFZm .62 ......L.... .12.. .5 325% 6.9%. .< moo 9R ovo mom vmv nmm me mNF 144 One peptide listed in Table 8b was identified from the platelets of Control #2 with the MHC type: HLA A2, A30, B60, B63, Cw2, Cw3. The sequence of this peptide was determined by SEQUEST, searched against the nr database and/or de novo sequencing to be (K/Q)EST(L/I)H(L/I)V(L/I) (Figure 17). One peptide was identified from the WBCs of Control #2 (Table 8b). The sequences of this peptides was determined to be RVAPEEHPV(L/I) (Figure 23). A BLAST search of these sequences identifies them as ubiquitin 63- 71 (KESTLHLVL) Accession number P02248 (Wiborg et al., 1985) and y—actin 95-104 (RVAPEEHPVL) Accession number NP_OOl605 (Gunning et al., 1983). Table 8b. HLA Associated Peptides Identified from Control #2 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets, WBC, and RBCs from Control #2 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. Sample Peptide AA Protein Cell Pos Control#2 KESTLHLVL 63-T1 ubiquitin platelet RVAPEEHPVL 95434 actin WBC None RBC 145 .ESHQUQO Odo mQu CH UmHQOH mum mcofl > cam Q Um>hmeO .Umumwa mum mcofl uQmEmMHm > paw Q MOM mmmmmE Umuoflpmum .AH\Qv>mmmmm¢>m ma mcflocmdvmm O>OE mp >Q mocmsvmm Umuoflpmhm mQB .o.¢bmnAN\Ev Oflumu mmHMQU ou mmmE m0 ESHuoQO Qfiu mQB vOHme CHQUMn? .A>mmmmm¢>m MO whuommm QflU .mm mhdmflm EE com CON com com oov com OON ooF .--..F .TA _TAFLFTFA .A_. ...F _L....T_TA . _L-...LFWAFAL AWL ...AAL..-A...A ...T_ A...A.T_ .. ..TA. ..._|_I A.....ALA-.. AT. _ ...A._ A .....L_._......... A ..Ty .... .AALQLFT. TALL F4.-..F..LTFL.L...L.LAL.. .TAFITL-._..TLF+A.F ..F .FLTLLTLFTLALTF TrA ..TL ....T; AFT L..TL-F.rTL..ATL A -L O . A ._ . _ . _ A . . . _ .. cavN A EFF .- . A . _ . . F ...- ...... A A . _ A .. osNN .L m . N o; A .. F an A ... e. F P» F- 2. m m N8 m A N+ D L w A an o A N+ o w. m «CNN A A 8 m F A . FL ... A A :2. ALL 8 A $8 L. 35 N+mn 8F 5 o.EmLLNFIN+_>_V ..F NmF FNN wNm we. 3m NNN ONN F8 F8 FFFF SH >.A. n: IA. m: m; n_.A. A.m_ Du: 90F 05 m5 Nwo mmm «NF» NNm mmm KB. .5 146 From Control #3, one peptide listed in Table BC was identified from the platelets with the MHC type: HLA A3, AX, BB, 814, Cw7, Cw8. The sequence of this peptide was determined by SEQUEST searched against the nr database and/or de novo sequencing to be ARVEHPFR (Figure 24). No peptides were identified from the WBCs of Control #3 (Table 8c). A BLAST search of the peptide sequence purified from the platelets identifies is as an unnamed protein 285-293 (ARVEHPFR) Accession number BAB15497 (Watanabe et al., 2000) . Table 8c. HLA Associated Peptides Identified from Control #3 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets, WBC, and RBCs from Control #3 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. Sample Peptide AA. Protein Cell Pos Control#3 .ARVEHPFR 2my292 unnamed protein. platelet 147 .ESHuommm Odo mQu CH UmeQmH mum mcofi h paw Q Um>hmeo .Umumfla mum mcofl ucmEmmnm h paw Q Hem mmmmmE Umuoflpmum .mmmmm>mm ma mafiocmdqmm o>o= mp >Q mucmsqmm Umuoficmum mQH .m.momuAN\Ev ofiumn mmMMQo Cu mmmE m0 Esnuommm adv mQH mmNmem cfimuoum UmEmccs .mmmmm>m¢ mo whuummm adv .vm mhsmfim NE. 82 com com 2K 80 8m 09. com com o rFrALLLtAererterLtttrrertPrLLthrtiL A .. AitlitL ...... .. . .. all..rr..£_Lerrrt+.tLLLrCLttLLLLttALt.rrtthizi . . A . . A . . . . AA. A . . _.- 58 $5 . F .2 . . . . . A A . 3R 3: 22. 28A .803} «EA «328 N S .8 a E on 3% A mm? m m+ n ..._ m mm A ~+ n > fie. m. . .8 v A m A new W A A... w A _L . .8; 3% msomuaffié ...» mt «mm as. 36 3o «E :a 3: mA. ”AT; 5 m: >H a? Nwow nmm omm mam omv mum wNN N5 :9 148 From the three control samples, MHC associated peptides from RBCs were isolated and analyzed by mass spectrometry as listed in Table 8a, Table 8b and Table 8c. No peptides consistent with the individual’s MHC binding motif were found in the RBC sample. Peptides identified from RBCs appeared to be fragments from random protein degradation. RBCs were used as a negative control as they do not express MHC. From the ITP population, two peptides listed in Table 8d were identified from the platelets of ITP #1 with the MHC type: HLA A2, AX, B7, B27, Cw2, Cw7. The sequence of these peptides were determined by SEQUEST, searched against the nr database and/or de novo sequencing to be GPRGA(L/I)S(L/I)(L/I) (Figure 25) and ARVEHPFR (Figure 24). A BLAST search of these sequences identifies them as GPIb 4-12 (GPRGALSLL) Accession number NP_OOO398 (Koike et al., 1988), and as an unnamed protein 285-293 (ARVEHPFR) Accession number BAB15497 (Watanabe et al., 2000). Two additional peptides were identified from the WBCs of ITP #1 (Table 8d). The sequences of these peptides were determined to be DTNAD(K/Q)(Q/K)(L/I)SF (Figure 26) and (L/I)DTNAD(K/Q)(Q/K)(L/I)SF (Figure 27). A BLAST search of these sequences identifies them as MRP—14 67-76 149 (DTNADKQLSF) and MFR-14 66-76 number NP_002956 (Odink et al., (LDTNADKQLSF) 1987) respectively. Accession Table 8d. HLA Associated Peptides Identified from ITP #1 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets and WBCs from ITP #1 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Appendix 1. Amino acid codes are listed in Sample Peptide AA. Protein Cell Pos ITP#1 GPRGALSLL 4'12 GPIBB platelet ARVEHPFR 285-292 unnamed protein platelet DTNADKQLSF 67 ' 75 MRP - 14 WBC LDTNADKQLSF 56 - 75 MRP — l4 WBC 150 .ESHuoQO Odo mQu CA UmHQOH mum mGOH h paw Q Um>hmeO .Umumfla mum mcofl ucmEmMHm > bum Q HOw mcflocmsqmm O>OG mp hQ mommmE pmuoflpmum .AH\QAAH\QAmAH\QAdUmmU ma musmzqmm Umuofipwnm mQH .vm.mv¢nAN\EA OHQMH mmHMQU Cu mmmE mo Ednuommm Q40 mQH QHmD NHL¢ Admdfiwmmw MO whuommm QdU .mN mhdmflm NE 088 com 88A 00A com com 888 com one cow com com omN OON omA LILLLALLLALAIALALALLAAALA ALVLFALLAtAl.AALLAAIALAALLLIALtFAIALPLL A. A A A AAL VA A rl.. .ALL_A.AL|AAE.A AA A A AA AA .A. AA AA A A _A AAA A AWA.A A AALAL O A A A A A Name A A8 NN SAN AMA wON A A A A 3 AAA £35 N.» a H A , Aw . N ..o... A A A NNmm AA New A N88 on A new A ND : W . _. .82 $2 58 b .3 AS N2 mANN Nmm 9% gm mum awn oww wwm SH .33 AAAH cam Q Uw>HmeO .Umumfla mum macs quEmme > paw Q How mcfiocmsqwm D>ou mp >Q mmmmmE Umuoflpmnm .mmAH\QAAM\OAAO\MAQ¢ZHQ ma moumdqmm Umuoflpmnm mQH .o.o>muAN\Ev OAQMH mmHMQU ou mmmE Mo ESHuommm Q40 mQH mmAOMQ¢ZBQ mbme ¢Hnmm2 m0 whuummm QfiU .mN mHDmflm NE. 8: 82 o8 08? 8A 88 08 8A... 2.8 SN 8 -. .- _A A.L A..A..A..AA.A.A4. . :AAAAALA ...AAAA A A A_A L ALA I 1 fr: -_A.A_ A908 8.2 A A A $8 firm. .8 SA .oAANc. 8 EN w. A > ANS. Nb > m w m... . NBA 8. at. 3 A ... m N Nww 0% Odom How nuuv. A A N.» m m N88 A u m m m A r b A 62% $8 an oaAmuaAINéA ..A 8A 8N 88 8A. N8 A2 88 NNm ANS 8: AAmAAAAANAaASA. a; 7AA. :8 mm: mum omw mnn mvo Cm Nov wmm EN 0: an 152 .Edhuommm Q40 ms» Ca UmHQOH mum mcofi % @nm Q Um>nmmno .vmumfia mum mcofl ucmEmmnm > Ucm Q MOM mcfiocmdvmm Q>OG mu >3 mmmmmE UmUUHUmHm .mmAH\AAAM\OAAO\MvndzHQAH\AA mfl mocmdvmm UmuUHUmHQ mnh .v.mmmuAN\EA oflumu mmnmno on mmmE mo Eduuummm adv $38 mmAOMQ¢ZBQA mbnmm wfiummz m0 whuommm QdU .bN mhdmflm N>c ooNA 00 A A 009 com com 8A com com oov com com o A A A A A A AALAA E A AIA A .FAA. A A A _r .AA A AAA AAAALAA A. AA A..i r .1...t|¥llA|Lr|.LIA.:-:|.FA A A WE A AA A A 0.9.: AA AAA. A T... .A.... AAA..A..AAA__A_ .A. . WA A N. 82 A A. mg mg A. RAW RA 3% A A m ONAAA. m2 Ammmm .omw mom a o_. _ w . m o. mmN w n A > AV wan N. Am A > 3% ~> w m A a 9» Sn ov m. A N+ m ”V A WOQMW A _ u A Tommw A Jew Ndmm an tomonafméA .S cm: mmm mom vow «mo AmA mom mum $3 mm: waA ..A.A & _AAAAvAaAAgA. AAA/AAA E: a .A. NmNA Awe oooA oww wmm omo mAm vvv 0mm mNN w: an 153 Two peptides listed in Table 8e were identified from the platelets of ITP #2 with the MHC type: HLA A3, AX, B7, BX, Cw7, CwX. The sequence of these peptides were determined by SEQUEST, searched against the nr database and/or de novo sequencing to be GPRGA(L/I)S(L/I)(L/I) (Figure 25) and XXX(K/Q)EA(L/I)ERF (Figure 28). An “x" denotes an unknown amino acid. A BLAST search identifies the sequence GPRGALSLL as GPIb 4-12. A BLAST search did not match the peptide sequence XXX(K/Q)EA(L/I)ERF to any sequence in the database. Table 8e. HLA Associated Peptides Identified from ITP #2 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets from ITP #2 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. An X denotes an unknown amino acid. Amino acid codes are listed in Appendix 1. Sample Peptide AA. Protein Cell Pos ITP#2 GPRGALSLL 4~12 GPIBfi platelet XXX(K/Q)EA(L/I)ERF platelet 154 wcofl > cam Q Um>HmeO >3 mommmE Umuoflpmhm mommdvmm Umpoflpmnm ®£H .Esnuommm Odo mnu as UmHQOH mum .pmumHH mum mcofl ucmEmmuw > pom Q HOw masocmdqmm O>oc mp .Uflom OCHEm c30cxc3 cm mmuocmp x .mmmAH\qV¢mAO\vaxx ma .m.ammuAN\EV oflumu mmhmno on mmmE mo Esuuommm adv one mmmAH\qv¢on\xvxxx No muuommm ado mN musmflm NE. SN? 8: 82 4 8w 2.: 8m 8m 8w. 8m 8N .|:.11.rl_1.t11-_.|wrl_._1; w i . _ 1%. j. 1:: LL} J .3». w _..1 ._ 1 .1Mmur111r1:1._:.u :LL‘. L : ___ __"fi __L t 1.|.H..O N mNNF «mg? m. 25 _omrmw _ _ o. Nam o mm“ 3%: _.N 8? News . m 38 TN% 9 3N. N3 5 i. _ _ o NNN m M... to? N 8m N wNm , ... v Ni: 38 . ... 8 m 8» _ w :9 .. w w . ‘ m> mom w 3% Nme. .ooFm P» o. FNouafNév i we NNN Ev wow 9% EN N8 x x NwNF burials/Am H02??? NVNF chow omm Non mum mom wt» 0mm X X an 155 From ITP #3, one peptide listed in Table 8f was identified. ITP #3 has an MHC type of: HLA A2, A24, B8, B40, Cw3, Cw7. The sequence of this peptide was determined by SEQUEST, searched against the nr database and/or de novo sequencing to be (K/Q)EST(L/I)H(L/I)V(L/I) (Figure 17). A BLAST search of this sequence identifies it as ubiquitin 63 -71 (KESTLHLVL) . Table 8f. HLA Associated Peptides Identified from ITP #3 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets from ITP #3 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1 . Sample Peptide AA. Protein Cell Pos ITP#3 KESTLHLVL 63-71 ubiquitin platelet From ITP #4, two peptides listed in Table 8g were identified. The MHC type of ITP #4 is: HLA A3, A24, B7, B40, Cw2/15, Cw7. The sequence of these peptide were determined by SEQUEST, searched against the nr database and/or de novo sequencing to be (K/Q)EST(L/I)H(L/I)V(L/I) 156 (Figure 17) and GPRGA(L/I)S(L/I)(L/I) (Figure 25). A BLAST search of this sequence identifies it as ubiquitin 63-71 (KESTLHLVL) and GPIb 4-12 (GPRGALSLL). Table 8g. HLA Associated Peptides Identified from ITP #4 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets from ITP #4 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. Sample Peptide AA. Protein Cell Pos ITP#4 GPRGALSLL 4‘12 GPIBB platelet KESTLHLVL 63-71 ubiquitin platelet ITP #5 with the MHC type: HLA A1, A11, B7, B8, Cw7, CwX, had one peptide identified listed in Table 8h. The sequence of this peptide was determined by SEQUEST, searched against the nr database and/or de novo sequencing to be GPRGA(L/I)S(L/I)(L/I) (Figure 25). A BLAST search of this sequence identifies it as GPIb 4-12 (GPRGALSLL). 157 Table 8h. HLA Associated Peptides Identified from ITP #5 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets from ITP #5 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1 . Sample Peptide AA. Protein Cell Pos ITP#S GPRGALSLL 442 GPIBB platelet The peptide GPRGA(L/I)S(L/I)(LAI) (GPIb) (Figure 25) was identified in four of the five ITP patients and none of the random controls. The GPIb (GPRGALSLL) peptide sequence was synthesized and the synthetic peptide CAD spectra were compared to the CAD spectra of the platelet peptides and found to be identical (Figure 29). The four ITP patients (ITP #1, ITP #2, ITP #4, and ITP #5) from whom the peptide GPRGALSLL was identified coincidently all have the allele HLA-B7. The GPIb peptide, which fits the binding motif of HLA-B7 (Rammensee et al., 1999), was not identified as an abundant peptide in ITP #3 who does not have the HLA-B7 allele. 158 Figure 29. GPRGALSLL GPIb 4-12 and synthetic A. GPRGALSLL peptide identified from platelet sample. Predicted masses for b and y fragment ions are listed and observed ions are labeled in the CAD spectra. B. Synthetic peptide, GPRGALSLL, was synthesized and analyzed by LC-MS/MS. The CAD spectrum is similar to the CAD spectra from the platelet sample in Figure 29A. Predicted masses for b and y fragment ions are listed and observed ions are labeled in the CAD spectra. Ions not represented in both spectra are probably due to the low abundance of the ion. 159 1L .TL1F L1LIPL1LLLIF -Ll ......111L11rL LILIF $8. $8 vdom Warm 0mm cow own 1.5. #.L IHJl L mKow, L . 5 L: mono N\E com com own CON 0mm coo omm com omv cow omm com com _LITrLIFLIrLLLICtLiELLELLIEIFLIILEF _1LIIT O Ndwv. «don N. F _5 ..L LL...)__1PL!..L Lift JILLILILFF LL L $8 is m> 3B 3» mg N.Nmmu.LL_.L+L2V $38 265,3 .m 00» 0mm ooo omm com on». oov com com com JLIL1LJ1L1:L1_1LL LJLJLLlLlplr L1L vdmv mem as N2 3N NNN ms. 05 NB aNN 0% <8 .3 _LLHmdLLHEGH E. E o vmw mm» mmo Nmm mmw mam 3m mm? mm emwwuazév mgamfigsmi .< fl VON _ _ , PIP .7411VL1L- it)... Q Nix” .. mn— ....ON T . fl '0 V - _' | fl—r'. 0 oo 62 ST (0 eouepunqv away 160 Due to the high frequency of HLA—B7 in the random ITP population, additional controls were selected with the HLA— B7 allele. The GPRGALSLL peptide was identified in 2 of the 3 HLA-B7 controls (Table 81). Table 8i. HLA Associated Peptides Identified from Control #4—6 using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets from Control #4-6 using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. Sample Peptide AA. Protein Cell Pos Control#4 None platelet Control#5 GPRGALSLL 4-12 GPIBB platelet Control#6 GPRGALSLL 443 GPIBB platelet All peptides isolated from controls and ITP individuals are collectively listed in Table 8. 161 Table 8. HLA Associated Peptides Identified from ITP and controls using citric acid wash Listed are the sequences of the MHC class I associated peptides purified from platelets, WBCs, and/or RBCs using citric acid wash. The source protein identified by BLAST is listed with the amino acid position of the peptide in the protein sequence. Amino acid codes are listed in Appendix 1. An X denotes an unknown amino acid. Sample Peptide AA. Protein Cell Pos Control#l KESTLHLVL 63-T1 ubiquitin platelet ALNELLQHV' 77F785 talin platelet ASRPGLLF 946 platelet factor 4 platelet KEVPDACF 5E¥600 thrombospondin platelet KESTLHLVL 53‘71 ubiquitin WBC None RBC Control#2 KESTLHLVL 63““- ‘ubiquitin platelet RVAPEEHPVL 95 ‘ 104 act in WBC None RBC Control#3 ARVEHPFR 2m¥292 unnamed protein. platelet Control#4 None platelet Control#S GPRGALSLL 4-12 GPIBB platelet Control#6 GPRGALSLL 4-12 GPIBB platelet ITP#l GPRGALSLL 4‘12 GPIBB platelet ARVEHPFR 285-292 unnamed protein platelet DTNADKQLSF 57 ‘75 MRP- l4 WBC LDTNADKQLSF 65 ‘ 76 MRP- l4 WBC ITP#2 GPRGALSLL 4w12 GPIBfi platelet xxx(K/Q) EA(L/I) ERF platelet ITP#3 KESTLHLVL 63-T1 ubiquitin platelet ITP#4 GPRGALSLL 4‘12 GPIBB platelet KESTLHLVL 63-T1 ubiquitin platelet ITP#S GPRGALSLL 4‘12 GPIBB platelet 162 5. Identification of Theoretical MHC Class I Peptides Listed in the following Tables 9—17 are the results of predicted peptides that could theoretically be processed and presented on the surface of the cell with MHC class I molecules. MHC class I molecules of individuals tested that were available in the search algorithm were used for analysis. The proteins analyzed were ubiquitin, talin, thrombospondin, GPIB beta, GPIB alpha, GPIX, GPIIB alpha, GPIIIA, and platelet factor 4. Analysis was done using the RANKPEP search algorithm (http://www.mifoundation.org/Tools/rankpep.html) using the listed MHC specific profiles and default values listing the top 2% of top scoring peptides. The prediction algorithm was done to provide additional support that the identified peptides were purified from a particular MHC molecule in addition to identifying additional peptides that could be present in the sample. Most of the identified peptides in the samples were identified in the top 2% of the predicted peptides. 163 Table 9. Theoretical Ubiquitin Peptides Accession #P02248. Ubiquitin is a 76 amino acid protein. Top 2% (n=1 peptide) of peptides are listed. Peptides in bold font have been identified in samples with the corresponding HLA type. HLA Pos Sequence MW Score/Optimal % Opt A1 N/A none N/A none/210 N/A A2 7 TLTGKTITL 947.115 95.0/123 77.24 96 A0301 55 TLSDYNIQK 1081.185 88.0/137 64.23 % All 55 TLSDYNIQK 1081.185 76.0/113 67.26 % A24 N/A none N/A none/222 N/A A2902 63 KESTLHLVL 1039.235 134.0/192 69.79 % B7 36 IPPDQQRLI 1079.275 96.0/198 48.48 % 88 25 NVKAKIQDK 1043.215 98.0/175 56.00 % 827 41 QRLIFAGKQ 1060.265 87.0/152 57.24 % 1351 36 IPPDQQRLI 1079.275 111.0/159 69.81 96 164 Table 10. Theoretical Talin Peptides Accession #AAD13152. Top 2% ALNELLQHV which is #25). identified in samples with the corresponding HLA type. that fits the motif for HLA-A2 The percentage of optimal talin peptide (777-785) ranks #25 with a score of 86. (n=51) or top 4 hits are listed (except for in bold font have been binding is 69.92%. Talin is a 2541 amino acid protein. Peptides The HLA Pos Sequence MW Score/Optimal % Opt A1 18 QFEPSTMVY 1101.245 153.0/210 72.86 % 562 AGDPAETDY 937.925 147.0/210 70.00 % 1845 EPEGSFVDY 1042.085 111.0/210 52.86 % 1769 AESALQLLY 1007.165 102.0/210 48.57 % A2 2127 SLLKTVKAV' 958.195 104.0/123 84.55 % 2423 KLISSAKQV' 973.175 98.0/123 79.67 % 1767 TLAESALQL 945.085 96.0/123 78.05 % 2308 ELLGAAAAI 827.985 95.0/123 77.24 % 'TT7 ALNELLQHV 1036.195 86.0/123 69.92% A0301 398 IILKKKKSK 1085.425 91.0/137 66.42 % 430 TVLQQQYNR 1149.265 88.0/137 64.23 % 908 .AIKKKLVQR 1083.375 87.0/137 63.50 % 903 .AAAQNAIKK. 914.065 84.0/137 61.31 % A11 2314 .AAIEAAAKK. 872.035 86.0/113 76.11 % 2353 .AAATSALVKL 830.975 84.0/113 74.34 % 2201 VIATANLSR. 944.095 81.0/113 71.68 % 1682 AVSQQLAPR. 969.115 80.0/113 70.80 % A24 N/A none N/A none/222 N/A A2902 1048 LEMDSALSV’ 964.105 140.0/192 72.92 % 1769 AESALQLLY 1007.165 138.0/192 71.88 % 290 SEIEAKVRY 1094.245 137.0/192 71.35 % 2248 LELLDHVLL 1064.295 133.0/192 69.27 % B7 979 SPSAQLALI 899.065 136.0/198 68.69 % 976 QPDSPSAQL 942.005 126.0/198 63.64 % 2331 KPKEADESL 1016.125 122.0/198 61.62 % 2229 APDVRLRAL 1010.215 121.0/198 61.11 % B8 293 EAKVRYVKL 1105.345 134.0/175 76.57 % 1915 HIKHRVQEL 1159.355 120.0/175 68.57 % 2517 EARKKLAQI 1056.275 116.0/175 66.29 % 318 KMKGKNKLV 1045.325 107.0/175 61.14 % B27 2056 TRLADVVKL 1014.225 115.0/152 75.66 % 1183 QRLAQVAKA. 984.165 114.0/152 75.00 % 2454 KRLQAAGNA. 928.055 108.0/152 71.05 % 915 QRLEHAAKQ 1080.215 104.0/152 68.42 % 851 979 SPSAQLALI 899.065 124.0/159 77.99 % 2073 DPETQVVLI 1013.155 107.0/159 67.30 % 1144 DPAVQAIVI. 925.095 106.0/159 66.67 % 976 QPDSPSAQL 942.005 103.0/159 64.78 % 165 Table 11. Theoretical Thrombospondin Peptides Accession #CAA32889. Thrombospondin is a 1170 amino acid protein. Top 2% (n=23) or top 4 hits are listed. Analysis for most 8-mers is not available, therefore the 8-mer KEVPDACF peptide purified from the sample could not be verified with this algorithm. were observed in the samples. No other peptides listed HLA Pos Sequence Dflv Score/Optimal % Opt A1 757 HYNPAQYDY 1170.215 107.0/210 50.95 % 160 VQEDRAQLY 1121.225 105.0/210 50.00 % 809 ERDNCQYVY 1189.275 101.0/210 48.10 % 250 SSPAIRTNY 1008.105 95.0/210 45.24 % A2 276 ELSSMVLEL 1020.215 98.0/123 79.67 % 293 TLQDSIRKV 1059.225 86.0/123 69.92 % 89 LLLASLRQM 1044.325 84.0/123 68.29 % 286 GLRTIVTTL 973.165 83.0/123 67.48 % A0301 93 SLRQMKKTR 1147.395 86.0/137 62.77 % 633 GVEHATANK, 925.985 85.0/137 62.04 % 193 SIARLRIAK 1027.285 85.0/137 62.04 % 292 TTLQDSIRK 1061.195 73.0/137 53.28 % A11 292 TTLQDSIRK 1061.195 79.0/113 69.91 % 633 GVEHATANKL 925.985 75.0/113 66.37 % 637 .ATANKQVCK. 962.115 71.0/113 62.83 % 193 SIARLRIAK 1027.285 68.0/113 60.18 % A24 N/A none N/A none/222 N/A A2902 179 LDVPIQSVF 1017.195 125.0/192 65.10 % 282 LBLRGLRTI 1070.305 99.0/192 51.56 % 177 .AELDVPIQS 971.085 91.0/192 47.40 % 306 KELANELRR 1128.305 90.0/192 46.88 % B7 53 DPSSPAFRI 989.115 111.0/198 56.06 % 1115 RPKTGFIRV 1073.305 107.0/198 54.04 % 99 KTRGTLLAL 972.185 107.0/198 54.04 % B8 96 QMKKTRGTL 1062.275 95.0/175 54.29 % 286 GLRTIVTTL 973.165 89.0/175 50.86 % 97 MKKTRGTLL 1047.305 87.0/175 49.71 % 304 ENKELANEL 1059.145 84.0/175 48.00 % B27 1033 SRFYVVMWK 1192.475 91.0/152 59.87 % 478 .ARETKACKK 1034.235 89.0/152 58.55 % 1053 TRAQGYSGL. 952.035 84.0/152 55.26 % 411. NRCEGSSVQ 979.035 79.0/152 51.97 % B51 53 DPSSPAFRI 989.115 112.0/159 70.44 % 698 WPNENLVCV 1050.225 105.0/159 66.04 % 251 SPAIRTNYI 1034.185 102.0/159 64.15 % 1115 RPKTGFIRV 1073.305 100.0/159 62.89 % 166 Table 12. Accession #P13224. (n=4) or top 4 hits are listed. font have been identified in samples. Top 2% Theoretical GPIB beta Peptides GPIB beta is a 206 amino acid protein. Peptides in bold HLA Position Sequence MW in Da Score/Optimal % Opt A1 N/A None N/A none/210 N/A A2 149 ALAAQLALL 883.105 96.0/123 78.05 96 160 GLLHALLLV 948.215 85.0/123 69.11 % 161 LLHALLLVL 1004.325 84.0/123 68.29 % 72 GLLDALPAL 882.075 82.0/123 66.67 % A0301 176 RLRARARAR 1125.365 63.0/137 45.99 % JEN) SLTDPLVAE 944.055 58.0/137 42.34 % 6O VLTGNNLTA. 901.995 52.0/137 37.96 % 79 .ALRTAHLGA 909.055 50.0/137 36.50 % All 180 RARARAAAR 998.175 55.0/113 48.67 % 176 RLRARARAR 1125.365 50.0/113 44.25 % 191 LTDPLVAER 1013.165 48.0/113 42.48 % 190 SLTDPLVAE 944.055 47.0/113 41.59 % A24 N/A None N/A None/222 N/A A2902 N/A None N/A None/192 N/A B7 4 GPRGALSLL 883.065 159.0/198 80.30 % 119 APPALRGRL 950.165 128.0/198 64.65 % 69 LPPGLLDAL 908.115 118.0/198 59.60 % 28 APCSCAGTL 821.965 117.0/198 59.09 % B8 176 RLRARARAR 1125.365 75.0/175 42.86 % 178 IUUUUUUUUX 998.175 73.0/175 41.71 % 149 ALAAQLALL 883.105 67.0/175 38.29 % 122 ALRGRLLPY 1058.305 66.0/175 37.71 % 827 181 ARARAAARL 955.145 107.0/152 70.39 % 175 RRLRARARA 1125.365 104.0/152 68.42 % 183 ARAAARLSL 928.115 87.0/152 57.24 % 179 ARARARAAA. 913.065 85.0/152 55.92 % B51 77 LPALRTAHL 991.205 123.0/159 77.36 % 53 FPVDTTEVL 1020.145 122.0/159 76.73 % 128 LPYLAEDEL 1062.205 107.0/159 67.30 % 69 LPPGLLDAL. 908.115 90.0/159 56.60 % 167 Table 13. Theoretical GPIB alpha Peptides Accession #P07359. Top 2% (n=12) or top 4 hits are listed. were observed in the samples. GPIB alpha is a 626 amino acid protein. No peptides listed HLA Positbo Sequence MW in Da Score/Optimal % Opt n A1 319 HTTPWGLFY 1098.255 118.0/210 56.19 % 236 LQDNAENVY 1065.105 117.0/210 55.71 % 613 LLSTVSIRY 1051.255 112.0/210 53.33 % A2 515 LLFASVVLI 974.255 100.0/123 81.30 % 254 AMTSNVASV 878.985 100.0/123 81.30 % 438 ILVSATSLI 916.125 92.0/123 74.80 % 461 LLESTKKTI 1032.235 82.0/123 66.67 % A0301 170 SLANNNLTE 975.015 87.0/137 63.50 % 119 TVLDVSFNR 1050.175 79.0/137 57.66 % 369 SITFSKTPKL 1008.175 76.0/137 55.47 % 77 NLDRCELTK. 1091.245 72.0/137 52.55 % All 119 TVLDVSFNR. 1050.175 85.0/113 75.22 % 369 SITFSKTPK. 1008.175 71.0/113 62.83 % 441 SATSLITPK 917.065 70.0/113 61.95 % 366 HMESITFSKL 1079.235 64.0/113 56.64 % A24 N/A none N/A :none/222 N/A A2902 52 LHLSENLLY 1101.275 112.0/192 58.33 % 20 CEVSKVASH 959.085 102.0/192 53.12 % 187 LENLDTLLL 1043.225 98.0/192 51.04 % 554 LELQRGRQV' 1098.275 94.0/192 48.96 % B7 588 RPNGRVGPL 965.125 130.0/198 65.66 % 564 VPRAWLLFL 1091.405 129.0/198 65.15 % 394 EPAPNMTTL 973.105 125.0/198 63.13 % 179 LPAGLLNGL 867.055 124.0/198 62.63 % B8 163 TPKLEKLSL 1028.255 92.0/175 52.57 % 464 STKKTIPEL 1016.195 88.0/175 50.29 % 144 ELYLKGNEL 1078.235 81.0/175 46.29 % 251 DVKAMTSNV 964.085 80.0/175 45.71 % B27 588 RPNGRVGPL 965.125 82.0/152 53.95 % 564 VPRAWLLFL 1091.405 81.0/152 53.29 % 394 EPAPNMTTL 973.105 75.0/152 49.34 % 179 LPAGLLNGL 867.055 74.0/152 48.68 % B51 115 LPALTVLDV 940.145 130.0/159 81.76 % 68 MPYTRLTQL 1122.345 123.0/159 77.36 % 1 MPLLLLLLL 1038.445 123.0/159 77.36 % 394 EPAPNMTTL 973.105 111.0/159 69.81 % 168 Table 14. Accession #NP_000165. Theoretical GPIX Peptides Top 2% (n=3) were observed in the samples. or top 4 hits are listed. GPIX is a 177 amino acid protein. No peptides listed HLA Position Sequence MW in Da Score/Optimal %Opt A1 N/A none N/A none/210 N/A A2 168 LLCATTEAL 934.115 78.0/123 63.41 % 163 ALLAGLLCA 844.085 68.0/123 55.28 % 148 VLWDVALVA 962.185 66.0/123 53.66 % A0301 57 LLANNSLQS 959.065 72.0/137 52.55 % 107 ALLQVRCAS 960.165 54.0/137 39.42 % 76 QLQTLDVTQ 1045.145 54.0/137 39.42 % All 44 TALPALPAR 909.105 55.0/113 48.67 % 10 LWATAEATK 967.115 45.0/113 39.82 % 57 LLANNSLQS 959.065 43.0/113 38.05 % A24 N/A none N/A none/222 N/A A2902 100 LEDRTPEAL 1043.155 93.0/192 48.44 % B7 49 LPARTRHLL 1076.315 118.0/198 59.60 % 66 VPPGAFDHL 952.085 98.0/198 49.49 % l MPAWGALFL 982.245 96.0/198 48.48 % B8 50 PARTRHLLL 1076.315 89.0/175 50.86 % 48 ALPARTRHL 1034.235 79.0/175 45.14 % 110 QVRCASPSL 960.125 64.0/175 36.57 % B27 38 CRGHGLTAL 927.085 69.0/152 45.39 % B51 74 LPQLQTLDV 1026.195 123.0/159 77.36 % 49 LPARTRHLL 1076.315 121.0/159 76.10 % 1 MPAWGALFL 982.245 100.0/159 62.89 % 169 Table 15. Theoretical GPIIB alpha Peptides Accession #P08514. GPIIB alpha is a 1039 amino acid protein. Top 2% (n=21) or top 4 hits are listed. No peptides listed were observed in the samples. HLA Pos Sequence MW Score/Optimal %Opt A1 837 QSQPSDLLY 1050.145 146.0/210 69.52 % 33 NLDPVQLTF 1046.185 129.0/210 61.43 % 253 SFDSSNPEY 1045.045 120.0/210 57.14 % 297 AVEILDSYY 1072.195 112.0/210 53.33 % A2 592 KLSPIVLSL 969.235 97.0/123 78.86 % 989 ALEERAIPI 1011.205 81.0/123 65.85 % 1008 LLLLTILVL 1010.365 79.0/123 64.23 % 932 AMVTVLAFL 964.225 79.0/123 64.23 % A0301 944 SLYQRPLDQ 1119.255 80.0/137 58.39 % 1017 AMWKVGFFK 1090.375 69.0/137 50.36 % 299 EILDSYYQR 1186.305 67.0/137 48.91 % 753 QIRSKNSQN 1074.155 63.0/137 45.99 % A11 299 EILDSYYQR 1186.305 70.0/113 61.95 % 969 AVPPLSLPR, 949.175 69.0/113 61.06 % 1012 TILVLAMWK 1051.375 67.0/113 59.29 % 113 QTLQTFKAR 1092.255 63.0/113 55.75 % A24 N/A none N/A none/222 N/A A2902 354 AEVGRVYLF 1053.235 123.0/192 64.06 % 709 KENETRVVL 1087.235 119.0/192 61.98 % 678 AELAVHLPQ 977.135 114.0/192 59.38 % 662 LELQMDAAN 1004.125 111.0/192 57.81 % B7 545 KPRQGRRVL 1109.345 144.0/198 72.73 % 761 NPNSKIVLL 997.195 126.0/198 63.64 % 413 GPSGRGQVL 869.975 122.0/198 61.62 % 26 APPAWALNL 929.125 121.0/198 61.11 % B8 484 VVKASVQLL 956.185 98.0/175 56.00 % 588 DFRDKLSPI 1090.255 95.0/175 54.29 % 761 NPNSKIVLL 997.195 92.0/175 52.57 % 989 ALEERAIPI 1011.205 89.0/175 50.86 % B27 351 RKLAEVGRV 1027.235 95.0/152 62.50 % 550 RRVLLLGSQ 1041.265 90.0/152 59.21 % 901 SRLQDPVLV 1026.205 89.0/152 58.55 % 62 GRVAIVVGA 841.015 89.0/152 58.55 % B51 771 VPVRAEAQV 968.125 122.0/159 76.73 % 872 LPIPSPSPI 920.135 120.0/159 75.47 % 594 SPIVLSLNV 941.135 112.0/159 70.44 % 858 FPQPPVNPL 1008.195 110.0/159 69.18 % 170 Table 16. Accession #P05106. 2% (n=16) were observed in the samples. or top 4 hits are listed. Theoretical GPIIIA Peptides GPIIIA is a 788 amino acid protein. No peptides listed Top HLA Pos Sequence IMW Score/Optimal %Opt A1 336 VTENVVNLY 1050.165 134.0/210 63.81 % 652 MTENTCNRY 1131.235 123.0/210 58.57 % 182 FVDKPVSPY 1051.215 123.0/210 58.57 % 675 GKDAVNCTY 970.055 119.0/210 56.67 % A2 722 VLLSVMGAI 902.155 92.0/123 74.80 % 342 NLYQNYSEL 1143.225 88.0/123 71.54 % 163 KLATQMRKL 1088.365 86.0/123 69.92 % 739 LIWKLLITI 1089.455 83.0/123 67.48 % A0301 232 EVKKQSVSR 1060.215 91.0/137 66.42 % 694 QYYEDSSGK 1076.095 75.0/137 54.74 % 320 LMTEKLSQK 1077.295 74.0/137 54.01 % 319 GLMTEKLSQ 1006.175 71.0/137 51.82 % All 778 STFTNITYR 1102.205 75.0/113 66.37 % 232 EVKKQSVSR 1060.215 75.0/113 66.37 % 105 QVTQVSPQR 1042.155 74.0/113 65.49 % 420 TVSFSIEAK 981.115 71.0/113 62.83 % A24 N/A none N/A none/222 N/A A2902 80 IEFPVSEAR 1047.195 108./1920 56.25 % 663 DEIESVKEL 1061.165 102.0/192 53.12 % 415 LKIGDTVSF 979.135 101.0/192 52.60 % 276 TDAKTHIAL 969.095 97.0/192 50.52 % B7 110 SPQRIALRL 1053.285 135.0/198 68.18 % 6 RPRPLWVTV 1100.365 127.0/198 64.14 % 713 CPKGPDILV 941.155 121.0/198 61.11 % 244 APEGGFDAI 875.945 121.0/198 61.11 % B8 434 QEKEKSFTI 1109.245 120.0/175 68.57 % 324 KLSQKNINL 1057.245 103.0/175 58.86 % 149 SMKDDLWSI 1071.245 100.0/175 57.14 % 230 NEEVKKQSV 1060.165 92.0/175 52.57 % B27 227 TRFNEEVKK 1150.295 107.0/152 70.39 % 286 GRLAGIVQP 910.085 96.0/152 63.16 % 305 NHYSASTTM 1011.065 74.0/152 48.68 % 169 RKLTSNLRI 1100.325 73.0/152 48.03 % B51 82 FPVSEARVL 1017.205 122.0/157 76.73 % 185 KPVSPYMYI 1097.345 116.0/157 72.96 % 388 LPEELSLSF 1034.195 115.0/157 72.33 % 211 LPMFGYKHV 1091.335 114.0/157 71.70 % 171 Table 17. Theoretical Platelet Factor 4 Peptides Accession #NP_002610. Platelet Factor 4 is a 101 amino acid protein. Top 2% (n=2 peptides) are listed. The HLA alleles B57, Cw3, or Cw7 that could bind platelet factor 4 8—mer peptide ASRPGLLF (which was purified from control #1) was not available for analysis. No other peptides listed were observed in the samples. HLA Pos Sequence MW Score/Opt % Opt A1 83 CLDLQAPLY 1035.235 104.0/210 49.52 % A2 20 LLLPLVVAF 984.295 71.0/123 57.72 % 14 LLFLGLLLL 1014.365 70.0/123 56.91 % A0301 89 PLYKKIIKK 1130.475 94.0/137 68.61 % 54 HITSLEVIK 1039.235 64.0/137 46.72 % All 43 CVKTTSQVR 1021.185 73.0/113 64.60 % 45 KTTSQVRPR 1072.225 61.0/113 53.98 % A24 N/A none N/A. none/222 N/A A2902 58 LEVIKAGPH 963.145 86.0/192 44.79 % 86 LQAPLYKKI 1073.345 85.0/192 44.27 % B7 64 GPHCPTAQL 923.055 113.0/198 57.07 % 51 RPRHITSLE 1108.275 103.0/198 52.02 % B8 90 LYKKIIKKL 1146.515 68.0/175 38.86 % B27 50 VRPRHITSL 1078.285 71.0/152 46.71 % B51 22 LPLVVAFAS 916.135 93.0/159 58.49 % 67 CPTAQLIAT 917.085 92.0/159 57.86 % 172 DISCUSSION 1. Protein A Antibody Purification The MAbs, W6/32, HB116, and L243, were purified from 20 ml of cell culture supernatant and/or 1—5 ml of ascities fluid (Figure 2, Figure 3, and Table 2). The observation from the SDS—PAGE is that the MAbs were relatively pure as compared with a commercially prepared reagents (Figure 4). It was shown by complement mediated cytotoxicity (Table 3) that the MAbs, W6/32, L243, and H58A were able to bind their target after they have been purified. The binding capability of the HB116 antibody was unable to be determined, as HB116 does not bind complement (Table 3). The SPA822 and SPA850 MAbs were used as additional negative controls as the MAb bind intracellular proteins and are unable to be tested by the microcytotoxicity method (Table 3). It has been demonstrated that by protein A immunoaffinity purification, SDS-PAGE, and microcytotoxicity testing, MAbs were purified and could bind their appropriate target. 2. Affi—gel®-10 columns To determine the most efficient antibody to purify MHC class I molecules, Affi-gel®-10 columns coupled to either 173 the MAb W6/32 or HB116 were tested. To evaluate the ability of these columns to purify the target protein, a negative control column using albumin instead of a MHC class I MAb was used (Figure 7 and Figure 9). Using WBC from sample #1, the HB116 Affi-gel®-10 column precipitated a similar amount of protein as the albumin Affi-gel®—10 column (0.087 mg vs. 0.066 and 0.135 mg respectively as listed in Table 4). Although it was difficult to evaluate the specificity of the precipitated proteins by SDS-PAGE, as no bands were seen on the polyacrylamide gel, recovery yield by absorbance at 280 nm indicates that the HB116 column eluted about as much protein as the negative control albumin column. The non- specific protein eluted from the HB116 column indicated that the HB116 is not the optimal column to use for immunoprecipitation of MHC molecules. Using WBCs from individual #1 and #2, the W6/32 column precipitated more protein than the albumin column (0.250 and 0.472 mg vs. 0.066 and 0.135 mg respectively as listed in Table 4). Proteins precipitated from the W6/32 column with lysed EBV B cells did not precipitate as much protein as individual #1 (In Table 4, 0.056 mg vs. 0.250 and 0.472 mg respectively). However, by SDS-PAGE visualization, W6/32 column did not precipitate pure MHC protein. The 174 amount of nonspecific proteins precipitated from MHC class I specific column resembled the nonspecific proteins precipitated from the albumin column. Since Affi-gel®-10 binds primary amines, it is possible for any protein with a primary amine to bind to this support. This could lead to nonspecific binding of proteins and/or a decrease in the optimal concentration of IgG (an therefore MHC class I) bound to the column. Because of the large amount of nonspecific protein binding and eluting from the Affi-gel®- 10 column, another method, immunomagnetic chromatography, was evaluated. 3. Immunomagnetic Chromatography The Dynabeads® magnetic beads have an isotype specific monoclonal antibody covalently attached to the bead. These beads will bind their target protein (IgG) with high efficiency and specificity (Worlock et al., 1991). This increased the specificity of protein purification as seen by SDS—PAGE in Figure 10, however, sufficient recovery was not obtained to characterize the MHC associated peptide. The magnetic beads were prepared by cross-linking a MAb to the covalently attached isotype specific MAb on the bead. Intact white blood cells, which express MHC class I molecules, did not bind to W6/32 cross linked to the 175 magnetic beads, but did bind to W6/32 not cross linked to the magnetic beads. From this experiment, it was determined that the cross linking procedure was blocking the MAb from binding to MHC and cross-linking of the antibody to the beads was discontinued. MHC class I, class II and Hsp70/Hsc70 proteins were purified by immunomagnetic chromatography (Figure 10 and Figure 11). Compared to the Affi-gel®-10 immunoprecipitation procedure, immunomagnetic chromatography purified proteins with minimal to no non- specific protein contamination. MHC class I proteins were purified using the W6/32 or H58A MAb. The polyacrylamide gel was run reduced and non-reduced to distinguish the IgG proteins (55 kDa heavy chain and 22 kDa light chain) from the MHC class I proteins (45 kDa heavy chain and 12 kDa 62M) (Figure 10) and provide further evidence MHC class I proteins were purified. 3.1. Peptide Purification by MW filtration The MHC class I, MHC class II and Hsp70/Hsc70 proteins were purified by immunomagnetic chromatography and the associated peptides were released from the MHC or Hsp by denaturation of the protein. The small molecular weight peptides were separated from the larger proteins by molecular weight filtration. This small molecular weight 176 fraction potentially contained up to 10,000 different peptides (de Jong, 1998). RP—HPLC using a microbore column was used to reduce the complexity of this mixture before protein sequencing and to remove unwanted salts and detergent. By RP—HPLC, this small molecular weight pool of peptides was separated into approximately 30 fractions. It was thought that in each of these 30 fractions, there could be a predominant peptide(s) at a concentration large enough to obtain sequence or anchor motif information by N—terminal protein sequencing. From individual #1, the most abundant fraction (#10) was selected for N-terminal protein sequencing (shown in Figure 12). Protein sequencing identified the peptides QEGNDVAKE and AKIKEIVDD. The MHC type of this individual is: HLA A3, A30, B60, B62, Cw3, CwX. These peptides did not match any of the predicted MHC class I anchor binding motifs consistent with the HLA type of the individual (Table 18). In addition, these peptides were not identified using the BLAST search algorithm. 177 Table 18. Individual #1 MHC Associated Peptide Anchor Motifs The MHC type of individual #1 is HLA A3, A30, B60, B62, Cw3, CwX. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Individual #1 HLA Allele 1 2 3 4 5 6 7 8 9 HLA-A*03XX(A3) L/V/M K/Y/F HLA—A*3001(A30) Y/F L HLA-B*40012 (B60) E L HLA—B*15XX(B62) Q/L F/Y HLA—Cw*0301(Cw3) L/F/M/I Peptide from Q E G N D V A K E fraction #10 Peptide from A. K I K E I V D D fraction #10 From the EBV B cell line 9035, the most abundant RP- HPLC fraction from MHC class II and Hsp70/Hsc70 associated peptides were selected for N—terminal protein sequencing. The first 10 amino acids of the peptides ALGTVKRKET and FDEKSGKSWE were identified from MHC class II associated peptides. It is more difficult to predict if these peptides fit the binding motifs associated with the MHC type as the relative positions on the peptide of the anchor 178 motifs for MHC class II molecules are not as defined as class I. The position of the anchor motifs is constant relative to each other. However, because the class II peptide is longer, the anchor positions can shift relative to the peptide from being located more at the amino terminal end to being more at the carboxyl terminal end. In addition, the length of the peptide ranges from 13 to 18 amino acids, which means that only partial sequence, was obtained. The anchor motifs located at the carboxyl terminal end may not have been sequenced and identified (Table 19). The class II MHC type of this cell line is: HLA DRll, DRX, DQ7, DQX. An “X” indicates a homozygous allele or that the allele could not be determined. The peptide, YIMYIRYIVR, was identified from Hsp70/Hsc70 associated peptides from EBV B cell line 9035. It is assumed that this peptide is present in the MHC class I processing pathway as Hsp70 is a chaperone molecule in this pathway and is in the process of being presented with MHC class I on the surface of the cell. However it is not possible to determine if the anchor motifs of this peptide would fit the HLA type of the individual from which it was purified (HLA A32, AX, B38, BX, CwX, CwX). The Hsp peptides can range up to 30 amino acids before being trimmed on the amino terminal and only the first 10 amino 179 Table 19. EBV B cell MHC Associated Peptide Anchor Motifs The MHC type of the EBV B cell 9035 is HLA DRll, DRX, DQ7, DQX. Listed 1-9 are the amino acid positions in the class II MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. The two different DQ7 motifs are due to differences in the DQAl gene. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for the EBV B cell 9035 HLA 1 2 3 4 5 6 7 8 9 Allele DRBl*llOl W/Y/F L/V/M/A/F/Y R/K/H A/G/S/P (DR11) DQB1*O301 W/Y/A/V/M A A/I/V/T/S Q/N (D07) DQBl*O30l D/E/W A/G/s/T A/C/L/M (D07) peptide A L G T V K R K ET peptide F D E K S G K S WE Table 20. EBV B Cell 9035 MHC Associated Peptide Anchor Motifs The MHC type of 9035 is HLA A32, AX, B38, BX, CwX, CwX. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Individual #1 HLA Allele 1 2 3 4 5 6 7 8 9 HLA—A*32XX(A32) HLA-B*3801(B38) F/L Peptide Y I M Y I R Y I VR 180 acids were sequenced (the peptide length on the carboxyl terminal is unknown). However, the theoretical anchor positions of this peptide did not match the binding motif (Table 20). As a negative control, HPLC solution A (used as a blank) was filtered through the MW filter and analyzed by RP-HPLC. Many contaminants were eluted from the MW filter (Figure 13). These contaminants could interfere with data interpretation of MHC associated peptides. In order to reduce the contamination, the MW filters were rinsed several times and then solution A was filtered through the filter again. Contaminants were still eluting off of the filter (Figure 14). Since a low level of sensitivity is needed to analyze MHC associated peptides from 1 x 108 cells, the contamination observed from the filters could interfere with data interpretation. An alternate method that eliminated separating the low molecular weight peptides from the larger proteins without using the molecular weight filter was investigated. 3.2. Peptide Purification by Analytical RP-HPLC Separation of low molecular weight peptides from large proteins, salts, and detergent was done using a 4.6 mm reverse phase analytical column instead of using molecular 181 weight filtration. This modification to the protocol separated low molecular weight peptides from high molecular weight proteins as well as reduced the complexity of the peptide mixture in one step. A larger capacity column (4.6 mm x 25.0 cm vs. 0.8 mm x 25.0 cm) was needed in order to not over load the column with large molecular weight proteins. MHC purified proteins were denatured, releasing the peptide from the MHC, and loaded onto the column. Peptides eluting within the first 60 minutes of the gradient (2—37% solution B) were collected (Figure 15). A few of the most abundant fractions as determined by UV absorbance were selected for peptide sequencing by LC-MS/MS. Sequest analysis of the LC-MS/MS run could not identify peptides in sufficient quantity for analysis. Based on calculations, it should be possible to characterize MHC associated peptides from 1 x 108 cells. However, due to the several manipulations used in the purification procedure, recovery can be 12—18% of the maximum yield (de Jong, 1998). Therefore, the number of steps in the purification procedure needs to be minimized in order to increase recovery so it can be detected. A one step mild acid wash procedure to release MHC associated peptides. This was coupled to sample clean up and 182 concentration on a RP trap column, and separation and identification of the peptides by in line LC-MS/MS was tested. To test this method, a cell population was needed that expresses MHC class I. These cells should also be relatively easy to obtain from individuals. Platelets were selected because they express MHC class I molecules but not MHC class II molecules and can be readily obtained by venipuncture. Enough platelets can be isolation from a small volume of whole blood (about 30 mls). Thirty milliliters of whole blood is a reasonably available clinical sample size (1 x 108 cells or 3 x 109 platelets) under most circumstances. 4. Peptide Purification by Mild Acid Wash with Trap Column 4.1. Control Platelets By using a mild acid wash of platelets, purification on a trap column and detection by LC—MS/MS four peptides were detected by LC—MS/MS from platelets from Control #1. Two of the peptides were of platelet specific function and two of the peptides were of general cellular function. Thrombospondin (Figure 18) is a protein involved in platelet aggregation and cell adhesion. Platelet factor-4 (Figure 19) is a chemokine released from activated 183 ... up an..." .— -c- -~ platelets. Ubiquitin (Figure 17), a general cellular function protein used to identify proteins for degradation, is expressed in many cells including megakaryocytes (platelet precursor) (Zhang et al., 1998), and RBCs (Galluzzi et al., 2001). Talin (Figure 20), another general cellular function protein that binds actin, is also expressed in many cells including platelets (Hagmann et al., 1992), and RBCs (Katoh et al., 1996). The peptides identified possess binding motifs for HLA molecules consistent with the individuals MHC type (Table 21). The ubiquitin peptide possesses MHC anchor binding motifs consistent with those of the HLA molecule B60, Cw*0301 (Cw3) and Cw*0702 (Cw7), the thrombospondin peptide possesses motifs consistent with Cw*0301 (Cw3) and Cw*0702 (Cw7), the platelet factor-4 peptide possesses motifs for B57, Cw*0301 (Cw3) and Cw*0702 (Cw7), and the talin peptide possesses motifs for A2. Previously published work has shown that ubiquitin 63-71 has been isolated from HLA A*2902 (A29) cells (Boisgerault et al., 1996; Rammensee et al., 1999) and B*40012 (B60) cells (Falk et al., 1995; Rammensee et al., 1999). Talin 777-785 has been previously published to be isolated from HLA A*0201 (A2) cells (Diehl et al., 1996; Rammensee et al., 1999). 184 ‘--A— ’H.‘ aQ.‘—b‘-—D*“‘ k ‘ ‘ Table 21. Control #1 MHC Associated Peptide Anchor Motifs The MHC type of control #1 is HLA A2, A29, B57, B60, Cw3, Cw7. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #1 HLA Allele l 2 3 4 5 6 7 8 9 HLA-A*0201(A2) L/M V/L HLA-A*2902 (A29) E Y HLA-B*57XX(BS7) A/T/S F/W/Y HLA-B*40012(B60) E L HLA-Cw*0301(Cw3) L/F/M/I HLA-Cw*O7XX(Cw7) Y/F/L Platelet and WBC I< E S T L H L V L peptide Platelet peptide A. L N E L L Q H V Platelet peptide A. S R P G L L F Platelet peptide K E V P D A C F From control #2, MHC associated peptides from platelets were characterized. The peptide characterized was ubiquitin (Figure 17), a general cellular function protein. The peptide identified possesses binding motifs for HLA molecules consistent with the individuals MHC type (Table 22). The ubiquitin peptide identified from control 185 #2 possesses MHC anchor binding motifs consistent with those of the HLA molecule B60 and Cw*0301 (Cw3). Table 22. Control #2 MHC Associated Peptide Anchor Motifs The MHC type of control #2 is HLA A2, A30, B60, B63, Cw2, Cw3. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. HLA- Cw2 anchor motifs are not available. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #2 HLA Allele 1 2 3 4 5 6 7 8 9 HLA-A*0201(A2) L/M V/L HLA-A*3001(A30) Y/F L HLA—B*400l2(B60) E L HLA—B*63XX (B63) HLA—Cw*O2XX(Cw2) HLA-Cw*0301(Cw3) L/F/M/I Platelet peptide 1( E S T L H L V L WBC peptide R v A P E E H P VL From control #3, a peptide of apparent general cellular function was characterized from platelets. ARVEHPFR is an unnamed human protein product (Figure 24) (Watanabe et al., 2000). From control #3, the peptide 186 ARVEHPFR possesses binding motifs partially consistent with HLA—B14 from the individual (Table 23). The peptide ARVEHPFR fits the binding motif of the first two anchor residues, but not the last position. However, not all peptides identified and listed in the SYFPEITHI Database possess all anchor motifs. Table 23. Control #3 MHC Associated Peptide Anchor Motifs The MHC type of control #3 is HLA A3, AX, B8, B14, Cw7, Cw8. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. HLA- Cw8 anchor motifs are not available. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #3 HLA Allele l 2 3 4 5 6 7 8 9 HLA-A*O3XX(A3) L/V/M K/Y/F HLA-B*08XX(BB) K K/R L HLA-B*l4XX(Bl4) R/K R/H L HLA-Cw*O7XX(Cw7) Y/F/L HLA-Cw*O8XX(Cw8) Platelet peptide A. R V E H P F R 187 . _ i. “ ~ . c .H ... A .. .... is: ., t. 4.2 Control WBCs From these three controls, MHC associated peptides were also characterized from WBCs. One peptide was detected from WBCs from Control #1. Ubiquitin (Figure 17), the general cellular function protein was identified from these WBCs. The ubiquitin peptide possesses MHC anchor binding motifs consistent with those of the HLA molecule B60, Cw*0301 (Cw3) and Cw*0702 (Cw7) (Table 21). Ubiquitin was also purified from MHC class I molecules from platelets from this individual. The peptide identified from WBCs from control #2 was from v—actin (Figure 23). Beta and gamma actin co-exist in most cell types as components of the cytoskeleton. Actin is a mediator of internal cell motility (Gunning et al., 1983). The y-actin peptide possesses MHC anchor binding motifs consistent with those of the HLA molecule Cw*0301 (Cw3) (Table 22). The binding motif for HLA—Cw*02 is unknown. Therefore, it is unknown if this peptide could bind to this HLA allele. No peptides were identified from the WBCs of control #3. This could be attributed to these peptides are below the limits of detection. Also, the high copy number peptides may not have been detected by this method because they were either not sufficiently charged or were multiply 188 charged which complicated analysis. Because these peptides were not tryptic digests, analysis was more difficult as computer programs to aide sequence analysis are optimized for tryptic peptides. In addition, common protein modifications, such as phosphorylation, to these peptides can make interpretation more difficult. 4.3 Identified Peptides are Characteristic of MHC Class I Peptides Evidence to support that MHC associated peptides have been isolated from the platelets and WBC is that the peptides identified were of the correct length and the anchor residues are consistent with MHC anchor motif of the individual. In addition, some of the peptides identified in this report have previously been isolated using other methods and reported in the literature (Rammensee et al., 1999) . To fit in the groove of an MHC class I molecule, peptides must be 8-11 amino acids long. We have identified three 8—mers, two 9-mers, and 1 lO-mer from the cells of the three controls, all of which are of correct length to fit within a MHC class I molecule. To minimize the possibility that the origin of the peptides were contaminants from the supernatant or extracellular/secreted proteins, cells were washed immediately prior to acid 189 elution. If proteins happened to contaminate the acid elution, these proteins would be removed by the peptide trap step. If the peptides were degradation products of a protein, more than one fragment should be detected in the data. To further provide support that MHC associated peptides were characterized, none of the peptides found in the platelet sample were found in the RBC control. Red blood cells were used as a control as they minimally express MHC (132 molecules of HLA per erythrocyte (Botto et al., 1990) which is approximately 1,000—6,000 times less than WBCs). Even though RBCs express the proteins ubiquitin, talin, and y-actin (Nakashima et al., 1979) and the peptide processing machinery, a proteasome (Ustrell et al., 1995), no processed peptides characteristic of the MHC class I endogenous processing pathway were detected, as RBCs essentially do not express the MHC class I molecules. 4.4 ITP Platelets Based on the results from the normal platelets, WBCs and RBCs, it is possible to identify some of the more abundant peptides on platelets and WBCs isolated from whole blood using a mild acid wash of whole intact cells. It was also possible to identify MHC class I associated peptides presented on cells from a sample volume compatible with 190 clinical availability. This technique was then applied to determine if a portion of the protein GPIb, GPIX, GPIIb, and/or GPIIIa (autoantibody targets) or other platelet proteins are presented on the platelets of individuals with ITP and if these peptides differed from random controls. From ITP #1, MHC associated peptides from platelets were characterized. The peptides identified were GPIb (Figure 25), a platelet membrane protein involved in clotting, and an unnamed protein (Figure 24), whose function is unknown. The ARVEHPRF unnamed peptide was previously identified from the platelets of control #3 (see Table 8). The peptides identified possess binding motifs for HLA molecules consistent with the individuals MHC type (Table 24). The GPIb peptide identified from ITP #1 possesses MHC anchor binding motifs consistent with those of the HLA molecule B7 and Cw7. The unnamed peptide possesses MHC anchor binding motifs consistent with those of the HLA molecule B27 and possibly HLA-CW2 (the binding motif of HLA-CW2 is unknown). From ITP #2, MHC associated peptides from platelets were characterized. The peptides identified were GPIb (Figure 25), a platelet membrane protein involved in clotting, and partially identified peptide (Figure 28). The peptides identified possess binding motifs for HLA 191 molecules consistent with the individuals MHC type (Table 25). The GPIb peptide identified from ITP #1 possesses MHC anchor binding motifs consistent with those of the HLA molecule B7 and Cw7. The unnamed peptide could possibly bind to A3, B7 and Cw7 at the 9u‘anchor position. Table 24. ITP #1 MHC Associated Peptide Anchor Motifs The MHC type of ITP #l is HLA A2, AX, B7, 827, Cw2, Cw7. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. HLA—Cw2 anchor motifs are not available. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for ITP #1 HLA Allele l 2 3 4 5 6 7 8 9 HLA-A*0201(A2) L/M V/L HLA-B*O7XX(B7) P L/F HLA-B*27XX(B27) R HLA-Cw*02XX(Cw2) HLA-Cw*O7XX(Cw7) Y/F/L Platelet peptide (3 P R G A L S L L Platelet peptide A R V E H P F R WBC peptide D T N A D K Q L SF WBC peptide l; D T N A D K Q LSF 192 Table 25. ITP #2 MHC Associated Peptide Anchor Motifs The MHC type of ITP #2 is HLA A3, AX, B7, BX, Cw7, CwX. Listed 1—9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. An “X” in the peptide sequence notes an unknown amino acid. MHC Associated Peptide Anchor Position Amino Acid Preference for ITP #2 HLA Allele 1 2 3 4 5 6 7 8 9 HLA-A*O3XX (A3) L/V/M K/Y/F HLA—B*O7XX(B7) P L/F HLA-Cw*O'7XX(Cw7) Y/F/L Platelet peptide G P R G A L S L L Platelet peptide X X X K/Q E A L/I E RF From ITP #3, an MHC associated peptide from platelets was identified. The peptide identified was ubiquitin (Figure 17), a general cellular function protein used to identify proteins for degradation. This ubiquitin peptide was also identified from control #1 and control #2 platelets and control #1 WBCs (Table 8). The peptide possesses binding motifs for HLA molecules consistent with the individuals MHC type (Table 26). The ubiquitin peptide identified from ITP #3 possesses MHC anchor binding 193 motifs consistent with those of the HLA molecule B40, Cw*0301 (CW3) and Cw7. Table 26. ITP #3 MHC Associated Peptide Anchor Motifs The MHC type of ITP #3 is HLA A2, A24, B8, B40, CW3, Cw7. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for ITP #3 HLA Allele 1 2 3 14 5 6 7 8 9 HLA-A*0201(A2) L/M V/L HLA—A*24XX(A24) Y I/F/L HLA-B*08XX(B8) K K/R L HLA—B*4OXX(B40) E L/W/M/A/T/R HLA-Cw*0301(Cw3) L/F/M/I HLA-Cw*O7XX(Cw7) Y/F/L Platelet peptide I< E S T I4 H I. V L From ITP #4, MHC associated peptides from platelets were characterized. The peptides identified were GPIb (Figure 25), a platelet membrane protein involved in clotting, and ubiquitin, a general cellular function protein used to identify proteins for degradation (Figure 17). The ubiquitin peptide was also identified from 194 control #1 and control #2 platelets and control #1 WBCs (Table 8). The peptides identified possess binding motifs for HLA molecules consistent with the individuals MHC type (Table 27). The GPIb peptide identified from ITP #4 possesses MHC anchor binding motifs consistent with those of the HLA molecule B7 and CW? (and possibly HLA—Cw2/15). The ubiquitin peptide could bind to B40 and Cw7 (and possibly HLA-Cw2/15). Table 27. ITP #4 MHC Associated Peptide Anchor Motifs The MHC type of ITP #4 is HLA A3, A24, B7, B40, Cw2/15, Cw7. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. HLA- Cw2 and HLA-CwlS anchor motifs are not available. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for ITP #4 HLA Allele 1 2 3 4 S 6 7 8 9 HLA-A*O3XX(A3) L/V/M K/Y/F HLA-A*24XX(A24) Y I/F/L HLA-B*O7XX(B7) P L/F HLA-B*4OXX(B40) E L/W/M/A/T/R HLA-Cw*02XX(Cw2) or Cw*15?? HLA-Cw*O7XX(Cw7) Y/F/L Platelet peptide G P R G A L S L L Platelet peptide 1K E S T L II I; V L 195 From ITP #5, an MHC associated peptide from platelets was characterized. The peptide identified was GPIb (Figure 25). The peptides identified possess binding motifs for HLA molecules consistent with the individuals MHC type (Table 28). The GPIb peptide identified from ITP #5 possesses MHC anchor binding motifs consistent with those of the HLA molecule B7 and Cw7. Table 28. ITP #5 MHC Associated Peptide Anchor Motifs The MHC type of ITP #5 is HLA A1, A11, B7, B8, Cw7, CwX. Listed 1—9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. HLA-CW8 anchor motifs are not available. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for ITP #5 HLA Allele l 2 3 ‘4 5 6 7 8 9 HLA-A*01XX(A1) D/E Y HLA-A*1101(A11) K/R HLA-B*O7XX(B7) P L/F HLA—B*08XX(BB) K K/R L HLA-Cw*O7XX(Cw7) Y/F/L HLA-Cw*08XX(Cw8) Platelet peptide G P R G A L S L L 196 From the ITP population, the peptide GPRGALSLL, GPIb (Figure 25) was detected as an abundant peptide by LC-MS/MS from platelets from ITP#l, ITP#2, ITP#4 and ITP#S. GPIb is a platelet specific protein that is part of the receptor for von Willebrand factor (Wadenvik et al., 1998). The anchor residues of the GPRGALSLL peptide (2“ and 9* position) that hold the peptide into the cleft of the MHC are P and L respectively. The four ITP patients from whom the peptide GPRGALSLL was identified coincidently all have the allele HLA-B7. HLA-B7 molecules have been found to bind peptides with the anchor residues P in the12nd position and L or F in the last (Table 24, Table 25, Table 27, and Table 28). The GPIb peptide (4-12), which fits the binding motif of HLA—B7, was not identified as an abundant peptide in the ITP patient who does not have the HLA-B7 allele. From the WBCs of ITP #1, two peptides whose sequences are identical except for an additional amino terminal amino acid on one of the peptides were identified, DTNADKQLSF and LDTNADKQLSF (Figure 26 and Figure 27 respectively). These peptides come from the migration inhibitory factor—related protein 14 (MRP-14) protein. MRP—14 is specific for cells of myeloid origin such as granulocytes, monocytes and macrophages. In acute and chronic inflammation, macrophages 197 can express MRP—l4 (Odink et al., 1987). These peptides possess MHC anchor binding motifs consistent with HLA molecule Cw*0702 (Cw7) (and possibly HLA—Cw2) (Table 23). This peptide was purified from the WBCs of ITP #1 who currently is producing antiplatelet autoantibodies (refer to the patient antiplatelet autoantibody status on Table 7). This peptide isolated from the WBCs of an ITP individual is of interest because it was identified from an individual during an active immune response. This peptide was only identified from the WBCs of ITP#l. It was not identified from individuals who are not actively producing antiplatelet antibodies. 4.5 Additional Platelet Controls (HLA—B7) Some of the identified peptides from the control and ITP platelets are from proteins with general cellular functions while others are from proteins with platelet specific function. Of interest is the platelet specific function peptide, GPIb (GPRGALSLL), because antiplatelet autoantibodies most often target the GPIIb/IIIa and/or GPIb/IX complex. The four ITP patients from whom the peptide GPRGALSLL (GPIb) was identified coincidently all had the allele HLA- B7. The ITP #3 as well as the 3 random controls did not have the HLA-B7 allele nor was the peptide GPRGALSLL 198 identified. To determine if the peptide was specific for the ITP patients or the HLA-B7 molecule, we selected 3 non- ITP individuals with HLA-B7 as additional controls. Platelets from control #4, although HLA-B7 positive with binding motif for the peptide (Table 29), did not appear to present GPRGALSLL as a major peptide on the surface of the platelets associated with MHC class I molecules. This could be due to a difference in antigen processing or the GRPGALSLL peptide could have been presented but not at a level sufficient for detection. No other peptide was detected from this sample. Table 29. Control #4 MHC Associated Peptide Anchor Motifs The MHC type of Control #4 is HLA A11, A29, B7, B51, CwX, CwX. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions are the anchor residues from the SYFPEITHI Database. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #4 HLA Allele 1 2 3 4 5 6 7 8 9 HLA-A*1101(All) K/R HLA-A*2902 (A29) E Y HLA—B*O7XX(B7) P L/F HLA-B*5101(B51) A/P/G V/I From control #5 and control #6, GPRGALSLL (GPIb peptide) was purified from the platelets. The GPIb peptide 199 possessed binding motifs consistent with the HLA type of these individuals (Table 30 and Table 31). Table 30. Control #5 MHC Associated Peptide Anchor Motifs The MHC type of Control #5 is HLA A1, A3, B7, B8, Cw7, CwX. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions is the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #5 HLA Allele l 2 3 4 5 6 7 8 9 HLA-A*01XX(A1) D/E Y HLA-A*O3XX(A3) L/V/M K/Y/F HLA—B*O7XX(B7) P L/F HLA—B*08XX(BB) K K/R L HLA-Cw*O7XX(Cw7) Y/F/L Platelet peptide G P R G A L S L L Based on the data reported, it is possible, using a citric acid wash, to identify some of the more abundant peptides on platelets and WBCs isolated from whole blood. It has been shown that the characterization of MHC class I associated peptides presented on cells from a sample volume compatible with clinical availability can be achieved. This method was applied to identify peptides from platelets 200 Table 31. Control #6 MHC Associated Peptide Anchor Motifs The MHC type of Control #6 is HLA A2, AX, B7, BX, Cw7, CwX. Listed 1-9 are the amino acid positions in the class I MHC associated peptide. Below the amino acid positions is the anchor residues from the SYFPEITHI Database. Peptides identified from this sample are listed below the HLA allele binding motifs. MHC Associated Peptide Anchor Position Amino Acid Preference for Control #6 HLA Allele 1 2 3 4 5 6 7 8 9 HLA-A*0201(A2) L/M V/L HLA-B*O7XX(B7) P L/F HLA-Cw*O7XX(Cw7) Y/F/L Platelet peptide (3 P R G A L S L L in an ITP population. It was found that GPIb (4-12) presented on platelets is not specific for ITP patients, but may be tied to the HLA-B7 binding motif. 201 SUMMARY AND CONCLUSIONS Using Protein A affinity chromatography, MAbs were purified from cell culture supernatant or ascites fluid and coupled to different solid supports in an effort to establish an optimal protocol to characterize the MHC associated peptides from a sample size compatible with clinically availability. Based on UV absorbance, SDS—PAGE, and microcytotoxicity testing, ample amount of MAbs were purified from both cell culture supernatant and ascites fluid for further immunoaffinity use. Affi-gel®-10 active ester agarose was used as a solid support coupled to a MAb with a specificity for MHC molecules to immunoprecipitate the MHC protein. Using the Affi-gel®—10 columns, multiple non-specific proteins co- eluted with the target protein from the column. This conclusion was based on the comparison of the polyacrylamide gels of the electrophoresed eluate from MHC specific Affi-gel®-10 columns and from the negative control, albumin Affi-gel®—10 columns. From this comparison, it was concluded that both columns bind and elute similar amounts of non-specific proteins. This method of MHC purification did not allow for the progression to the final step of sequencing the MHC 202 associated peptides because the MHC molecule was not purified. Therefore, if MHC associated peptides are to be purified and characterized, the solid support the MAb is attached to should reduce the co-elution of non—specific proteins. Purification of MHC molecules was attempted using magnetic beads as an alternate solid support. MHC class I, class II, and Hsp70/Hsc70 proteins were purified using immunomagnetic chromatography. Using a MAb coupled to magnetic beads, improved protein purification was obtained versus the Affi-gel®—10 column. The amount of non-specific protein elution was determined by visualization using SDS- PAGE. Since the magnetic bead protocol was able to elute relatively pure MHC and HSP proteins, purification and sequencing of the MHC and HSP associated peptides were attempted. However, when the RP-HPLC profile of the MHC sample was compared to the RP-HPLC profile of the negative control (blank) sample, it was concluded that contaminants from the 5,000 NMWL filter unit was contaminating the sample. Because MHC associated peptides are present in low concentrations [1,600 fmol maximum theoretical yield from 1 x 108 cells (de Jong, 1998)], this level of contamination was not acceptable and an alternate method to separate associated peptides from MHC was evaluated. 203 An alternate method to separate low molecular weight peptides from large proteins, salts, and detergents used analytical scale RP—HPLC without prior molecular weight filtration. However, using this method, peptides could not be detected or sequenced by LC-MS/MS. It is mathematically possible to characterize MHC associated peptides from a reasonable clinical sample size of 1 x 108 cells. However, the number of manipulations used to purify the peptide significantly reduces the amount of peptide available for the final analysis step of peptide sequencing. Therefore, the number of steps in the peptide purification protocol should be reduced. A protocol that minimizes peptide loss was evaluated. Using a citric acid wash of whole intact cells, it was possible to identify some of the more abundant peptides on platelets and white blood cells isolated from 28-94 ml of whole blood. These peptides were of correct length to fit in the groove of a MHC class I molecule. The anchor residues of the peptides were consistent with the MHC type of the individual. In addition, some of the peptides identified were in the top 2% of predicted peptides for the assumed parent protein using the RANKPEP search algorithm. This demonstrates that MHC class I associated peptides presented on cells from a sample volume compatible with 204 clinical availability can be identified. This meets the first objective of this project, which was to establish a method to characterize MHC associated peptides from a reasonable clinical sample (1 x 108 cells). Using platelets, WBCs, and RBCs, peptides were characterized from several individuals. The peptides identified were from both general cellular function proteins (ubiquitin, actin, talin, and an unnamed protein) as well as proteins from functions specific for the cells from which they were purified (thrombospondin, platelet factor 4, and GPIb). In addition, from WBCs in an active immune response (ITP #1), the peptide from MRP-14, a protein expressed by WBCs during inflammation, was identified. Some of the peptides were purified from more than one sample. The same peptide could be isolated from different individuals with the same HLA allele. This resulted in cellular specific and immune state specific peptides being identified and compared from different cell types from several individuals. This meets the goal for objective 2. Objective 3 was to characterize MHC associated peptides presented on platelets of individuals with ITP. The citric acid wash protocol was applied to identify peptides found on platelets from this population. It was 205 found that platelets can present GPIb (4—12) polypeptides associated with MHC class I. However, the presentation of GPIb (4-12) is not specific for ITP patients, but may be tied to the HLA-B7 binding motif. Since both HLA-B7 ITP patients and appropriate HLA—B7 controls are presenting GPIb peptides, it suggests that determinants of which person develops autoimmunity may not reside with peptide presentation, but with a loss of self-tolerance. This provides additional information on the possible mechanism of the cause of autoimmunity in ITP. Using tetramer analysis, further studies involving T cell recognition of the HLA—B7/GPIb (4-12) peptide should be evaluated. 206 ' I" FUTURE RECOMMENDATINS An optimal protocol to characterized MHC associated peptides from a sample size compatible with clinical availability (1 x 108 cells) was established. A citric acid wash of intact cells with separation of peptides from large proteins by an RP—HPLC trap column was utilized. This was followed by peptide identification by LC-MS/MS. Using this protocol it was possible to characterize the most abundant peptides from two different cell populations (platelets and WBCs). From this study, it was found that general cellular function proteins such as ubiquitin, talin, and y-actin as well as cell specific proteins such as GPIb, and thrombospondin were presented with MHC class I molecules on platelets and/or WBCs. Further studies should include platelets and WBCs from additional individuals with HLA types and binding motifs different from those reported here. This will give a general idea of the overall source of proteins used in the non-immune mediated (or quiescent) endogenous MHC associated peptide presentation pathway. Further studies should also characterize peptides from additional cell types. This would be important in establishing a base knowledge of the quiescent state of 207 antigen processing and presentation that could be compared to an immunologically active state. Using ITP as an immunologically active model, it was found that GPIb peptide could be presented on platelets from individuals with ITP as well as healthy controls. The anchor residues on the GPIb polypeptide, GPRGALSLL (4—12), are consistent with binding motif of the HLA-B7 molecule. Further studies involving ITP should focus on identifying other peptides from GPIb and other abundant platelet proteins from individuals with similar and different HLA types in ITP. In addition to studying antigen processing and presentation in ITP, identifying an antigen specific T cell to the GPRGALSLL (GPIb) peptide/HLA-B7 complex would be an important step in identifying autoreactive T cells in the periphery of ITP individuals. This could be accomplished by tetramer analysis. Tetramer analysis would determine if HLA B-7 individuals with ITP have CD8+ T cells that could bind this HLA-B7 and GPIb peptide complex. This is accomplished by staining a soluble GPIb peptide/HLA-B7 complex bound to avidin at a 4:1 ratio and incubating this with T cells stained with an antibody (anti-CD8 is used to evaluate MHC class I molecules). These tetrameric MHC- peptide and T cell complexes are analyzed by flow 208 cytometric analysis (Altman et al., 1996). Tetramer analysis would provide further support for the role of the GPIb (4-12) peptide in ITP and the loss of tolerance. This research has demonstrated a valid method to characterize the most abundant MHC associated peptides from a realistic clinical sample size (1 x 108 cells). This protocol was then used to characterize MHC associated peptides from platelet as well as WBCs. This information was applied to the characterization of peptides on platelets from an ITP population. This research has established a valid method to characterize MHC associated peptides from a reasonable clinical sample and applied it to characterize MHC peptide presentation during an immune response . 209 Appendix 1 Amino Acid Code Symbol Mass Alanine Ala A 71 Arginine Arg R 156 Asparagine Asn N 114 Aspartic Acid .Asp D 115 Cysteine Cys C 103 Glutamic Acid Glu E 129 Glutamine Gln Q 128 Glycine Gly G 57 Histidine His H 137 Isoleucine Ile I 113 Leucine Leu L 113 Lysine Lys K 128 Methionine Met M 131 Phenylalanine Phe F 147 Proline Pro P 97 Serine Ser S 87 Threonine Thr T 101 Tryptophan Trp W 186 Tyrosine Tyr Y 163 Valine Val V 99 210 REFERENCES Altman, J. D., Moss, P. A., Goulder, P. J., Barouch, D. H., McHeyzer-Williams, M. G., Bell, J. I., McMichael, A. J., Davis, M. M., 1996. Phenotypic analysis of antigen-specific T lymphocytes. Science 274, 94-96. Andersson, J., 1998. Cytokines in idiopathic thrombocytopenic purpura (ITP). Acta Paediatr Suppl 424, 61-64. Barnstable, C. J., Bodmer, W. F., Brown, G., Galfre, G., Milstein, C., Williams, A. F., Ziegler, A., 1978. Production of monoclonal antibodies to group A erythrocytes, HLA and other human cell surface antigens—new tools for genetic analysis. Cell 14, 9- 20. Bielekova, B., Martin, R., 2001. Antigen-specific immunomodulation via altered peptide ligands. J Mol Med 79, 552-565. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., Wiley, D. C., 1987a. The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature 329, 512-518. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., Wiley, D. C., 1987b. Structure of the human class I histocompatibility antigen, HLA- A2. Nature 329, 506-512. Boisgerault, F., Khalil, I., Tieng, V., Connan, F., Tabary, T., Cohen, J. H., Choppin, J., Charron, D., Toubert, A., 1996. Definition of the HLA-A29 peptide ligand motif allows prediction of potential T-cell epitopes from the retinal soluble antigen, a candidate autoantigen in birdshot retinopathy. Proc Natl Acad Sci U S A 93, 3466—3470. Bonner, P. L., Lill, J. R., Hill, S., Creaser, C. S., Rees, R. C., 2002. Electrospray mass spectrometry for the identification of MHC class I-associated peptides expressed on cancer cells. J Immunol Methods 262, 5- 19. 211 Botto, M., So, A. K., Giles, C. M., Mason, P. D., Walport, M. J., 1990. HLA class I expression on erythrocytes and platelets from patients with systemic lupus erythematosus, rheumatoid arthritis and from normal subjects. Br J Haematol 75, 106-111. Bouma, G. J., van Caubergh, P., van Bree, S. P., Castelli- Visser, R. M., Witvliet, M. D., van der Meer-Prins, E. M., van Rood, J. J., Claas, F. H., 1996. Pregnancy can induce priming of cytotoxic T lymphocytes specific for paternal HLA antigens that is associated with antibody formation. Transplantation 62, 672-678. Brodsky, F. M., Parham, P., 1982. Monomorphic anti—HLA- A,B,C monoclonal antibodies detecting molecular subunits and combinatorial determinants. J Immunol 128, 129-135. Busch, R., Doebele, R. C., Patil, N. S., Pashine, A., Mellins, E. D., 2000. Accessory molecules for MHC class II peptide loading. Curr Opin Immunol 12, 99— 106 . Bussel, J. B., Kunicki, T. J., Michelson, A. D., 2000. Platelets: New Understanding of Platelet Glycoproteins and Their Role in Disease. Hematology (Am Soc Hematol Educ Program), 222-240. Clark, R. E., Dodi, I. A., Hill, S. C., Lill, J. R., Aubert, G., Macintyre, A. R., Rojas, J., Bourdon, A., Bonner, P. L., Wang, L., Christmas, S. E., Travers, P. J., Creaser, C. S., Rees, R. C., Madrigal, J. A., 2001. Direct evidence that leukemic cells present HLA- associated immunogenic peptides derived from the BCR- ABL b3a2 fusion protein. Blood 98, 2887-2893. Cotran, R. S., Kumar, V., Robbins, S. L. (Eds.). 1994. Robbins Pathologic Basis of Disease, W. B. Saunders Co., Philadelphia. Datema, G., Stein, S., Eijsink, C., Mulder, A., Claas, F. H., Doxiadis, II, 2000. HLA-C expression on platelets: studies with an HLA-Cwl-specific human monoclonal antibody. Vox Sang 79, 108-111. Davis, W. C., Marusic, S., Lewin, H. A., Splitter, G. A., Perryman, L. E., McGuire, T. C., Gorham, J. R., 1987. 212 The development and analysis of species specific and cross reactive monoclonal antibodies to leukocyte differentiation antigens and antigens of the major histocompatibility complex for use in the study of the immune system in cattle and other species. Vet Immunol Immunopathol 15, 337-376. de Jong, A., 1998. Contribution of mass spectrometry to contemporary immunology. Mass Spectrom Rev 17, 311- 335. Diehl, M., Munz, C., Keilholz, W., Stevanovic, S., Holmes, N., Loke, Y. W., Rammensee, H. G., 1996. Nonclassical HLA—G molecules are classical peptide presenters. Curr Biol 6, 305-314. Dixit, V. M., Hennessy, S. W., Grant, G. A., Rotwein, P., Frazier, W. A., 1986. Characterization of a cDNA encoding the heparin and collagen binding domains of human thrombospondin. Proc Natl Acad Sci U S A 83, 5449-5453. el—Khateeb, M. S., Awidi, A. S., Tarawneh, M. S., Abu- Khalaf, M., 1986. HLA antigens, blood groups and immunoglobulin levels in idiopathic thrombocytopenic purpura. Acta Haematol 76, 110-114. Engelhard, V., Brickner, A., Zarling, A., 2002. Insights into antigen processing gained by direct analysis of the naturally processed class I MHC associated peptide repertoire. Mol Immunol 39, 127. Evers, K. G., Thouet, R., Haase, W., Kruger, J., 1978. HLA frequencies and haplotypes in children with idiopathic thrombocytopenic purpura (ITP). Eur J Pediatr 129, 267-272. Falk, K., Rotzschke, O., Takiguchi, M., Gnau, V., Stevanovic, S., Jung, G., Rammensee, H. G., 1995. Peptide motifs of HLA-B58, B60, B61, and B62 molecules. Immunogenetics 41, 165-168. Galluzzi, L., Paiardini, M., Lecomte, M. C., Magnani, M., 2001. Identification of the main ubiquitination site in human erythroid alpha-spectrin. FEBS Lett 489, 254- 258. 213 George, J. N., El-Harake, M. A., Aster, R. H., 1995. Thrombocytopenia due to enhanced platelet destruction by immunologic mechanisms. In: Beutler, E., Lichtman, M. A., Coller, B. S. and Kipps, T. J. (Eds), Williams Hematology. McGraw-Hill Inc., New York. Ghosh, P., Amaya, M., Mellins, E., Wiley, D. C., 1995. The structure of an intermediate in class II MHC maturation: CLIP bound to HLA-DR3. Nature 378, 457- 462. Goebel, K. M., Hahn, E., Havemann, K., 1977. HLA matching in autoimmune thrombocytopenic purpura. Br J Haematol 35, 341—342. Goldberg, A., Cascio, P., Saric, T., Rock, K., 2002. The importance of the proteasome and subsequent proteolytic steps in the generation of antigenic peptides. Mol Immunol 39, 147. Gooding, K. M., Regnier, F. E. (Eds.). 1990. HPLC of biological molecules: methods and applications, Marcel Dekker, Inc., New York. Gorg, A., Postel, W., Weser, J., Schiwara, H. W., Boesken, W. H., 1985. Horizontal SDS electrophoresis in ultrathin pore-gradient gels for the analysis of urinary proteins. Science Tools 32, 5-9. Gorga, J. C., Horejsi, V., Johnson, D. R., Raghupathy, R., Strominger, J. L., 1987. Purification and characterization of class II histocompatibility antigens from a homozygous human B cell line. J Biol Chem 262, 16087-16094 . Gratama, J. W., D'Amaro, J., de Koning, J., den Ottolander, G. J., 1984. The HLA-system in immune thrombocytopenic purpura: its relation to the outcome of therapy. Br J Haematol 56, 287-293. Gromme, M., Neefjes, J., 2002. Antigen degradation or presentation by MHC class I molecules via classical and non-classical pathways. Mol Immunol 39, 181. Gunning, P., Ponte, P., Okayama, H., Engel, J., Blau, H., Kedes, L., 1983. Isolation and characterization of full-length cDNA clones for human alpha-, beta-, and 214 gamma-actin mRNAs: skeletal but not cytoplasmic actins have an amino-terminal cysteine that is subsequently removed. Mol Cell Biol 3, 787-795. Hagmann, J., Grob, M., Burger, M. M., 1992. The cytoskeletal protein talin is O—glycosylated. J Biol Chem 267, 14424-14428. Halder, T. M., Bluggel, M., Heinzel, S., Pawelec, G., Meyer, H. E., Kalbacher, H., 2000. Defensins are dominant HLA-DR-associated self-peptides from CD34(-) peripheral blood mononuclear cells of different tumor patients (plasmacytoma, chronic myeloid leukemia). Blood 95, 2890-2896. Harris, P. E., Maffei, A., Liu, Z., Colovai, I., Reed, E. F., Inghirami, G., Suciu-Foca, N., 1993. Naturally processed cytokine-derived peptide bound to HLA-class II molecules. J Immunol 151, 5975-5983. Hedlund-Treutiger, I., Wahlstrom, J., Elinder, G., 1998. Role of the T cell receptor in idiopathic thrombocytopenic purpura (ITP). Acta Paediatr Suppl 424, 46-50. Helmerhorst, F. M., Nijenhuis, L. E., de Lange, G. G., van den Berg-Loonen, P. M., Jansen, M. F., von dem Borne, A. E., Engelfriet, C. P., 1982. HLA antigens in idiopathic thrombocytopenic purpura. Tissue Antigens 20, 372-379. Herr, W., Ranieri, E., Gambotto, A., Kierstead, L. S., Amoscato, A. A., Gesualdo, L., Storkus, W. J., 1999. Identification of naturally processed and HLA- presented Epstein-Barr virus peptides recognized by CD4(+) or CD8(+) T lymphocytes from human blood. Proc Natl Acad Sci U S A 96, 12033-12038. Hopkins, K. A., 2000. HLA Cytotoxicity Testing. In: Hahn, A. B., Land, G. A. and Strothman, R. M. (Eds), ASHI Laboratory Manuel, Vol. I. American Society of Histocompatibility and Immunogenetics, Lenexa, pp. 1.C.1 1-7. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A. L., Appella, E., Engelhard, V. H., 1992. Characterization of peptides 215 bound to the class I MHC molecule HLA-A2.1 by mass spectrometry. Science 255, 1261-1263. Janeway, C. A., Travers, P., Walport, M., Shlomchik, M., 2001. Immunobiology: the immune system in health and disease. Garland Publishing, New York. Kageyama, S., Tsomides, T. J., Sykulev, Y., Eisen, H. N., 1995. Variations in the number of peptide-MHC class I complexes required to activate cytotoxic T cell responses. J Immunol 154, 567-576. Karpatkin, S., Fotino, M., Gibofsky, A., Winchester, R. J., 1979. Association of HLA-DRw2 with autoimmune thrombocytopenic purpura. J Clin Invest 63, 1085-1088. Katoh, K., Kano, Y., Masuda, M., Fujiwara, K., 1996. Mutually exclusive distribution of the focal adhesion associated proteins and the erythrocyte membrane skeleton proteins in the human fibroblast plasma membrane undercoat. Cell Struct Funct 21, 27-39. Kessler, B., Glas, R., Ploegh, H., 2002. MHC class I antigen processing regulated by cytosolic proteolysis- short cuts that alter peptide generation. Mol Immunol 39, 171. Koike, K., Ohta, S., Urata, Y., Kagawa, Y., Koike, M., 1988. Cloning and sequencing of cDNAs encoding alpha and beta subunits of human pyruvate dehydrogenase. Proc Natl Acad Sci U S A 85, 41-45. Kulomaa, M. S., Weigel, N. L., Kleinsek, D. A., Beattie, W. G., Conneely, O. M., March, C., Zarucki-Schulz, T., Schrader, W. T., O'Malley, B. W., 1986. Amino acid sequence of a chicken heat shock protein derived from the complementary DNA nucleotide sequence. Biochemistry 25, 6244-6251. Kunicki, T., 1989. Biochemistry of Platelet-Associated isoantigens and alloantigens. In: Kunicki, T. J. and George, J. N. (Eds), Platelet Immunobiology: molecular and clinical aspects. J. B. Lippincott Co., Philadelphia. Kurata, Y., Oshida, M., Take, H., Furubayashi, T., Mizutani, H., Tomiyama, Y., Yonezawa, T., Tarui, S., 216 1990. Acid treatment of platelets as a simple procedure for distinguishing platelet-specific antibodies from anti-HLA antibodies: comparison with chloroquine treatment. Vox Sang 59, 106-111. Kurata, Y., Oshida, M., Take, H., Furubayashi, T., Nakao, H., Tomiyama, Y., Kanayama, Y., Nagao, N., Okubo, Y., Yonezawa, T., et al., 1989. New approach to eliminate HLA class I antigens from platelet surface without cell damage: acid treatment at pH 3.0. Vox Sang 57, 199-204. Kuwana, M., Kaburaki, J., Kitasato, H., Kato, M., Kawai, S., Kawakami, Y., Ikeda, Y., 2001. Immunodominant epitopes on glycoprotein IIb-IIIa recognized by autoreactive T cells in patients with immune thrombocytopenic purpura. Blood 98, 130-139. Kuwana, M., Kaburaki, J., Pandey, J. P., Murata, M., Kawakami, Y., Inoko, H., Ikeda, Y., 2000. HLA class II alleles in Japanese patients with immune thrombocytopenic purpura. Associations with anti- platelet glycoprotein autoantibodies and responses to splenectomy. Tissue Antigens 56, 337-343. Lampson, L. A., Levy, R., 1980. Two populations of Ia-like molecules on a human B cell line. J Immunol 125, 293- 299. Lopez, J. A., Chung, D. W., Fujikawa, K., Hagen, F. S., Davie, E. W., Roth, G. J., 1988. The alpha and beta chains of human platelet glycoprotein Ib are both transmembrane proteins containing a leucine-rich amino acid sequence. Proc Natl Acad Sci U S A 85, 2135-2139. Lupas, A., Koster, A. J., Baumeister, W., 1993. Structural features of 26S and 208 proteasomes. Enzyme Protein 47, 252-273. Mao, L., Fan, Y. H., 1998. Complete cDNA sequence of human talin. Direct Submission, Accession Number AAD13152. Matzinger, P., 2002. The danger model: a renewed sense of self. Science 296, 301-305. McMillan, R., Lopez-Dee, J., Loftus, J. C., 2001. Autoantibodies to alpha(IIb)beta(3) in patients with 217 chronic immune thrombocytopenic purpura bind primarily to epitopes on alpha(IIb). Blood 97, 2171-2172. Mitchell, E. A., Birch, H. P., Woodfield, D. G., 1981. HLA and idiopathic thrombocytopenic purpura. N Z Med J 94, 258-259. Mohanty, D., Quadri, M. I., Das, K. C., Sengupta, S., 1979. HLA and idiopathic thrombocytopenic purpura (ITP). Br J Haematol 42, 329-330. Monaco, J. J., 1995. Pathways for the processing and presentation of antigens to T cells. J Leukoc Biol 57, 543-547. Mueller-Eckhardt, C., Mayr, W., Lechner, K., Mueller- Eckhardt, G., Niessner, H., Pralle, H., 1979. HLA antigens in immunologic thrombocytopenic purpura (ITP). Scand J Haematol 23, 348-352. Mueller-Eckhardt, G., Hauck, M., Kayser, W., Mueller- Eckhardt, C., 1980. HLA-C antigens on platelets. Tissue Antigens 16, 91-94. Nakashima, K., Beutler, E., 1979. Comparison of structure and function of human erythrocyte and human muscle actin. Proc Natl Acad Sci U S A 76, 935-938. Nussbaum, A. K., Kuttler, C., Hadeler, K. P., Rammensee, H. G., Schild, H., 2001. PAProC: a prediction algorithm for proteasomal cleavages available on the WWW. Immunogenetics 53, 87-94. Odink, K., Cerletti, N., Bruggen, J., Clerc, R. G., Tarcsay, L., Zwadlo, G., Gerhards, G., Schlegel, R., Sorg, C., 1987. Two calcium-binding proteins in infiltrate macrophages of rheumatoid arthritis. Nature 330, 80-82. Parham, P., 1979. Purification of immunologically active HLA-A and -B antigens by a series of monoclonal antibody columns. J Biol Chem 254, 8709-8712. Parham, P., Alpert, B. N., Orr, H. T., Strominger, J. L., 1977. Carbohydrate moiety of HLA antigens. Antigenic properties and amino acid sequences around the site of glycosylation. J Biol Chem 252, 7555-7567. 218 Parham, P., Barnstable, C. J., Bodmer, W. F., 1979. Use of a monoclonal antibody (W6/32) in structural studies of HLA-A,B,C, antigens. J Immunol 123, 342-349. Perkins, D. N., Pappin, D. J., Creasy, D. M., Cottrell, J. S., 1999. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551-3567. Poncz, M., Surrey, S., LaRocco, P., Weiss, M. J., Rappaport, E. F., Conway, T. M., Schwartz, E., 1987. Cloning and characterization of platelet factor 4 cDNA derived from a human erythroleukemic cell line. Blood 69, 219-223 . Powis, S. H., Mockridge, I., Kelly, A., Kerr, L. A., Glynne, R., Gileadi, U., Beck, S., Trowsdale, J., 1992a. Polymorphism in a second ABC transporter gene located within the class II region of the human major histocompatibility complex. Proc Natl Acad Sci U S A 89, 1463-1467. Powis, S. J., Deverson, E. V., Coadwell, W. J., Ciruela, A., Huskisson, N. S., Smith, H., Butcher, G. W., Howard, J. C., 1992b. Effect of polymorphism of an MHC-linked transporter on the peptides assembled in a class I molecule. Nature 357, 211-215. Rammensee, H., Bachmann, J., Emmerich, N. P., Bachor, O. A., Stevanovic, S., 1999. SYFPEITHI: database for MHC ligands and peptide motifs. Immunogenetics 50, 213- 219. Reche, P., Glutting, J., Reinherz, E., 2002. Prediction of MHC class I binding peptides using profile motifs. Hum Immunol 63, 701. Robinson, J., Malik, A., Parham, P., Bodmer, J., Marsh, S., 2000. IMGT/HLA database - a sequence database for the human major histocompatibility complex. Tissue Antigens 55, 280-287. Robinson, J., Waller, M., Parham, P., Bodmer, J., Marsh, 8., 2001. IMGT/HLA Database - a sequence database for the human major histocompatibility complex. Nucleic Acids Research 29, 210-213. 219 Schagger, H., 1994. Denaturing Electrophoretic Techniques. In: Jagow, G. v. and Schagger, H. (Eds), A Pratical Guide to Membrane Protein Purification. Academic Press, San Diago. Schagger, H., von Jagow, G., 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166, 368-379. Schwartz, K. A., 1998. Immune Thrombocytopenia in Adults. Biomedical Progress 11, 23-28. Schwartz, K. A., Gauger, J. A., Davis, J. M., 1990. Precise quantitation of PAIgG: a new radiometric microtechnique. Am J Hematol 33, 167-176. Semple, J. W., 1998. Immunobiology of T helper cells and antigen-presenting cells in autoimmune thrombocytopenic purpura (ITP). Acta Paediatr Suppl 424, 41-45. Semple, J. W., Milev, Y., Cosgrave, D., Mody, M., Hornstein, A., Blanchette, V., Freedman, J., 1996. Differences in serum cytokine levels in acute and chronic autoimmune thrombocytopenic purpura: relationship to platelet phenotype and antiplatelet T- cell reactivity. Blood 87, 4245-4254. Shabanowitz, J., 2001. Personal Communication. Sidney, J., del Guercio, M. F., Southwood, S., Engelhard, V. H., Appella, E., Rammensee, H. G., Falk, K., Rotzschke, O., Takiguchi, M., Kubo, R. T., et al., 1995. Several HLA alleles share overlapping peptide specificities. J Immunol 154, 247-259. Singh, V. K., Mehrotra, S., Agarwal, S. S., 1999. The paradigm of Th1 and Th2 cytokines: its relevance to autoimmunity and allergy. Immunol Res 20, 147-161. Skipper, J. C., Gulden, P. H., Hendrickson, R. C., Harthun, N., Caldwell, J. A., Shabanowitz, J., Engelhard, V. H., Hunt, D. F., Slingluff, C. L., Jr., 1999. Mass- spectrometric evaluation of HLA-A*0201-associated peptides identifies dominant naturally processed forms 220 of CTL epitopes from MART-1 and gp100. Int J Cancer 82, 669-677. Smith, D. F., Sullivan, W. P., Marion, T. N., Zaitsu, K., Madden, B., McCormick, D. J., Toft, D. O., 1993. Identification of a 60-kilodalton stress-related protein, p60, which interacts with hsp90 and hsp70. Mol Cell Biol 13, 869-876. Spellberg, B., Edwards, J. E., Jr., 2001. Type 1/Type 2 immunity in infectious diseases. Clin Infect Dis 32, 76—102. Spies, T., Bresnahan, M., Bahram, S., Arnold, D., Blanck, G., Mellins, E., Pious, D., DeMars, R., 1990. A gene in the human major histocompatibility complex class II region controlling the class I antigen presentation pathway. Nature 348, 744-747. Srivastava, P. K., Udono, H., Blachere, N. E., Li, Z., 1994. Heat shock proteins transfer peptides during antigen processing and CTL priming. Immunogenetics 39, 93-98. Stanworth, S. J., Turner, D. M., Brown, J., McCloskey, D., Brown, C., Provan, D., Navarrete, C. V., Newland, A. C., 2002. Major histocompatibility complex susceptibility genes and immune thrombocytopenic purpura in caucasian adults. Hematol 7, 119-121. Storkus, W. J., Zeh, H. J., Salter, R. D., Lotze, M. T., 1993. Identification of T-cell epitopes: rapid isolation of class I-presented peptides from viable cells by mild acid elution. J Immunother 14, 94—103. Stuart, M. J., Tomar, R. H., Miller, M. L., Davey, F. R., 1978. Chronic idiopathic thrombocytopenic purpura. A familial immunodeficiency syndrome. Jama 239, 939-942. Sugawara, S., Abo, T., Kumagai, K., 1987. A simple method to eliminate the antigenicity of surface class I MHC molecules from the membrane of viable cells by acid treatment at pH 3. J Immunol Methods 100, 83—90. Sundaram, R., Dakappagari, N. K., Kaumaya, P. T., 2002. Synthetic peptides as cancer vaccines. Biopolymers 66, 200-216. 221 Theofilopoulos, A. N., 1995. The basis of autoimmunity: Part II. Genetic predisposition. Immunol Today 16, 150-159. Tomlinson, A. J., Jameson, S., Naylor, S., 1996. Strategy for isolating and sequencing biologically derived MHC class I peptides. J Chromatogr A 744, 273—278. Tsai, S. L., Chen, M. H., Yeh, C. T., Chu, C. M., Lin, A. N., Chiou, F. H., Chang, T. H., Liaw, Y. F., 1996. Purification and characterization of a naturally processed hepatitis B virus peptide recognized by CD8+ cytotoxic T lymphocytes. J Clin Invest 97, 577-584. Ustrell, V., Realini, C., Pratt, G., Rechsteiner, M., 1995. Human lymphoblast and erythrocyte multicatalytic proteases: differential peptidase activities and responses to the 118 regulator. FEBS Lett 376, 155- 158. van Endert, P. M., Lopez, M. T., Patel, S. D., Monaco, J. J., McDevitt, H. O., 1992. Genomic polymorphism, recombination, and linkage disequilibrium in human major histocompatibility complex-encoded antigen- processing genes. Proc Natl Acad Sci U S A 89, 11594- 11597. Veenhoven, W. A., Sijpesteijn, J. A., van der Schans, G. S., 1979. HLA antigens in idiopathic thrombocytopenic purpura. Acta Haematol 62, 153-158. Wadenvik, H., Stockelberg, D., Hou, M., 1998. Platelet proteins as autoantibody targets in idiopathic thrombocytopenic purpura. Acta Paediatr Suppl 424, 26- 36. Walker, R. H., 1993. American Association of Blood Banks Technical Manual. American Association of Blood Banks, Bethesda. Wang, J. H., Reinherz, E. L., 2002. Structural basis of T cell recognition of peptides bound to MHC molecules. Mol Immunol 38, 1039-1049. Watanabe, K., Kumagai,A., Itakura,S., Yamazaki,M., Tashiro,H.,, Ota, T., Suzuki,Y., Obayashi,M., 222 Nishi,T., Shibahara,T., Tanaka,T.,, Nakamura, Y., Isogai,T. and Sugano,S, 2000. Accession Number BABlS497. Wiborg, O., Pedersen, M. S., Wind, A., Berglund, L. E., Marcker, K. A., Vuust, J., 1985. The human ubiquitin multigene family: some genes contain multiple directly repeated ubiquitin coding sequences. Embo J 4, 755- 759. Worlock, A. J., Sidgwick, A., Horsburgh, T., Bell, P. R., 1991. The use of paramagnetic beads for the detection of major histocompatibility complex class I and class II antigens. Biotechniques 10, 310-315. Yewdell, J., 2002. To DRiP or not to DRiP: generating peptide ligands for MHC class I molecules from biosynthesized proteins. Mol Immunol 39, 139. Zhang, Y., Wang, Z., Liu, D. X., Pagano, M., Ravid, K., 1998. Ubiquitin-dependent degradation of cyclin B is accelerated in polyploid megakaryocytes. J Biol Chem 273, 1387-1392. 223