Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Computational prediction of peptide-MHC class I binding interactions
(USC Thesis Other)
Computational prediction of peptide-MHC class I binding interactions
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
COMPUTATIONAL PREDICTION OF PEPTIDE-MHC CLASS I BINDING INTERACTIONS by Huynh-Hoa Thi Bui A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements of the Degree DOCTOR OF PHILOSOPHY (PHARMACEUTICAL SCIENCES) May 2003 Copyright 2003 Huynh-Hoa Bui R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 3103862 Copyright 2003 by Bui, Huynh-Hoa Thi All rights reserved. ® UMI UMI Microform 3103862 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90089-1695 This dissertation, written by HUYNH - HOA 'THI &Ut under the direction o f h 6-A dissertation committee, and approved by all its members, has been presented to and accepted by the Director o f Graduate and Professional Programs, in partial fulfillment of the requirements fo r the degree of DOCTOR OF PHILOSOPHY Director Date May 16. 2003 Dissertqtion Cmnmittei Chair R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. DEDICATION To My parents Hai Nhu Bui Huu-Phuc Thi Huynh My husband Jin Yang My brother Henry Huu-Hien Bui R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ACKNOWLEDGEMENTS Dr. Ian S. Haworth For academic advice, project guidance, and financial support to make my graduate study possible Dr. Hermann von Grafensteine For research advice Dr. Eric J. Lien Dr. Michael B. Bolger Dr. Ching-An Peng For serving on my Ph.D. committee Charles and Charlotte Krown Fellowship For financial support Members of the research group of Dr. Ian S. Haworth Members of MIDDL R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. TABLE OF CONTENTS Dedication......................................................................................................................... ii Acknowledgements.........................................................................................................iii List of Tables................................................................................................................... vi List of Figures...............................................................................................................viii Abbreviations................................................................................................................... x Abstract........................................................................................................................... xi Chapter 1. Introduction....................................................................................................1 Chapter 2. Molecular Dynamics Simulations of Peptide-HLA-A2 Complexes 13 2.1. Introduction 2.2. Methods 2.3. Results 2.4. Discussion Chapter 3. WATGEN: An Algorithm for Modeling Water Networks at Protein-Protein Binding Interfaces.......................................................... 22 3.1. Introduction 3.2. Methods 3.3. Results 3.4. Discussion Chapter 4. Peptide Backbone Library........................................................................ 56 4.1. Introduction 4.2. Methods 4.3. Results 4.4. Discussion Chapter 5. Conformational Flexibility of HLA-A2 Peptide Binding Site.............. 64 5.1. Introduction 5.2. Methods 5.3. Results 5.4. Discussion R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. V Chapter 6. MHC: A Program for Predicting Conformations of Peptide-MHC Complexes........................................................................... 73 6.1. Introduction 6.2. Methods 6.3. Results 6.4. Discussion Chapter 7. Predicting peptide-HLA-A2 Binding Interactions................................... 99 7.1. Introduction 7.2. Methods 7.3. Results 7.4. Discussion Chapter 8. Summary.....................................................................................................133 References......................................................................................................................135 Appendix A ................................................................................................................... 148 Appendix B................................................................................................................... 160 Appendix C................................................................................................................... 162 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. LIST OF TABLES Table 2.1. Experimental binding data of peptides with HLA-A2............................... 15 Table 3.1. Definition of acceptor hydrogen centers.....................................................26 Table 3.2. Reference atoms and Hydrogen centers...................................................... 28 Table 3.3. Definition of water sites............................................................................... 29 Table 3.4. Agreement between experimental and predicted water sites................... 35 Table 4.1. Coordinates of peptide positions 1 and 9.................................................... 59 Table 4.2. Peptide CA Box............................................................................................60 Table 4.3. Peptide backbone library.............................................................................. 61 Table 5.1. Conformations of HLA-A2.......................................................................... 65 Table 5.2. Conformations of peptide / HLA-A2 complexes....................................... 67 Table 5.3. Dihedral angles of HLA-A2 protein side chains.........................................69 Table 6.1. Flexible HLA-A2 residues........................................................................... 76 Table 6.2. VDW contact energy score for non-bonded atoms.................................... 76 Table 6.3. Rotatable torsion angles of amino acid sidechains.................................... 78 Table 6.4. Peptide-HLA-A2 complexes used in the MHC prediction........................83 Table 7.1. Peptide binding affinities to HLA-A2........................................................101 Table 7.2. Binding interactions of selected peptide-HLA-A2 complexes............... 127 Table 7.3. Statistics of binding classification using Kohonen self organizing feature map for MHC-binding to peptides................................................ 131 Table 7.4. Kohonen self organizing map binding classification............................... 131 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. vii Table A.l. Atom van der Waals radius....................................................................... 148 Table A.2. Amino acid parameters..............................................................................148 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. viii LIST OF FIGURES Figure 1.1. Three-dimensional structure of class IMHC molecules............................5 Figure 1.2. MHC class I peptide binding site.................................................................6 Figure 2.1. Graphical representation of the peptide-HLA-A2 interface.....................14 Figure 2.2. Motion of the y \ torsion angle of the F5 side chain................................ 18 Figure 2.3. Motion of the GlhSGFVFTL peptide backbone.......................................19 Figure 2.4. A structure of the GlhSGFVFTL / HLA-A2 complex.............................20 Figure 3.1. Computational procedure for predicting water networks........................ 24 Figure 3.2. Internal coordinate system for computing water sites..............................27 Figure 3.3. Hydrogen bonding site geometry...............................................................31 Figure 3.4. Paired water molecules at the 1I7U:C interface....................................... 42 Figure 3.5. Proportion of paired water.......................................................................... 43 Figure 3.6. Temperature factor distributions for experimental water molecules 45 Figure 3.7. Distribution of polar contacts for experimental water molecules...........47 Figure 3.8. Mean separation distances for paired water molecules............................48 Figure 3.9. Proportion of paired water molecules categorized by the number of polar contacts...........................................................................................49 Figure 3.10. Correlation between the calculated (Equation 3) and WATGEN predicted square root of the number of water sites at peptide-protein binding interfaces....................................................... 52 Figure 4.1. Backbone conformational variation of four nonameric peptides bound to HLA-A2........................................................................................57 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 4.2. Peptide backbone parameters.....................................................................59 Figure 4.3. Computed peptide backbone conformations.............................................62 Figure 5.1. Conservation of HLA-A2 backbone atoms...............................................70 Figure 5.2. Distribution of root mean squared deviation of HLA-A2 Ca atoms...... 71 Figure 6.1. Computational procedure for predicting binding conformations of peptide-MHC complexes........................................................................... 77 Figure 6.2. Computational procedure for predicting sidechain conformations.........79 Figure 6.3. Predicted peptide conformations................................................................84 Figure 7.1. Peptide composition..................................................................................117 Figure 7.2. Peptide conformational flexibility............................................................119 Figure 7.3. Peptide-MHC direct intermolecular interactions................................... 121 Figure 7.4. Peptide-Water-MHC conformations........................................................123 Figure 7.5. Water-mediated peptide-MHC interactions............................................125 Figure 7.6. Correlation between predicted and experimental peptide-HLA-A2 binding affinity......................................................................................... 127 Figure 7.7. Distribution of predicted binding affinity............................................... 128 Figure 7.8. A 3-D scatter plot of peptide-HLA-A2 binding properties....................129 Figure 7.9. Stereo view of peptide-HLA-A2 binding properties.............................. 130 Figure A .l. Hydrogen bonding geometry...................................................................159 Figure B.l. WATGEN program interface.................................................................. 160 Figure B.2. MHC program interface............................................................................161 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. ABBREVIATIONS /32m Beta-2 microglobulin CTL Cytotoxic T lymphocyte HLA Human Leukocyte Antigen MHC Major Histocompatibility Complex PDB Protein Data Bank RMSD Root mean squared deviation SQL Structured Query Language TCR T Cell Receptor VDW van der Waals R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. xi ABSTRACT Class I major histocompatibility complex (MHC) molecules are cell surface glycoproteins that bind short peptide fragments and present them to cytotoxic T lymphocytes. This antigen presentation mechanism operates in most nucleated cells and provides an important means for the immune system to detect and eliminate virally infected or tumor cells. The ability to predict peptide sequences that are capable of binding to MHC class I molecules is a crucial step toward establishing CTL-based immunotherapies and rational vaccine design. The binding of peptides to MHC molecule is both specific and promiscuous. It is specific in the sense that an MHC molecule encoded by a specific allele will bind certain peptides but not others, and it is promiscuous in the sense that different peptide sequences might bind to the same MHC molecule. It was shown that sequence diversity in the central region of the peptide is accommodated by a combination of flexibility of polymorphic MHC residues at the bottom of the groove and variation of bound water molecules underneath the peptide. Hence, determining the precise binding interactions of peptide-MHC complexes requires consideration of the structural water molecules and the flexibility of MHC sidechains in the peptide binding site. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In this research, a computer algorithm was developed to predict binding interactions of peptide-MHC class I complexes. The success of the MHC algorithm in predicting structures of peptide-HLA-A2 complexes was shown to depend on three critical factors: (1) adequate sampling of peptide backbone conformations, (2) flexibility of MHC sidechains, and (3) modeling of explicit interface water molecules. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 Chapter 1 Introduction 1.1. Immunological roles of MHC class I molecules Class I major histocompatibility complex (MHC) molecules are cell surface glycoproteins that bind short antigen-derived peptide fragments and present them to cytotoxic T lymphocytes (CTL) (Doherty & Zinkemagel, 1975; Zinkemagel & Doherty, 1979; Townsend & Bodmer, 1989; Bjorkman & Parham, 1990; Sherman & Chattopadhyay, 1993). Antigenic peptides are derived mainly from cytosolic proteins and are loaded onto MHC molecules after being transported into the endoplasmic recticulum (Yewdell & Bennink, 1992; Germain, 1994; York & Rock, 1996). This antigen presentation mechanism operates in most nucleated cells and provides an important means for the immune system to detect and eliminate virally infected or tumor cells (Germain & Margulies, 1993; Boon et al., 1994; Coulie, 1996; Yewdell & Bennink, 1999). Unlike B cells, T cells only recognize antigenic peptides when they are bound to MHC molecules, a phenomenon known as “MHC restriction” (Zinkemagel & Doherty, 1974; Zinkemagel & Doherty, 1997). Thus, the ability to predict peptide sequences that are capable of binding to MHC class I molecules is a cmcial step toward establishing CTL-based immunotherapies (Gaur & Fathman, 1994; Stauss & Dahl, 1996; Ridgway et al., 1999; Stauss, 1999) and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 vaccine development (Vajda et al., 1990; Chicz & Urban, 1994; Apostolopoulos et al., 2000; Schirle et al., 2001; Meng & Butterfield, 2002; Doytchinova & Flower, 2002a). The ability to predict the structure is also important for a better characterization of the Tcell receptor ligand and a better understanding of T cell activation, antagonism, selection and cross-reactivity (molecular mimicry). 1.2. MHC polymorphism One major characteristic of MHC molecules is their extreme polymorphism (Klein et al., 1993) (Takahata, 1995)[for a compilation see (Robinson et al., 2000)]. As of October 2000, more than 700 human MHC, also called human leukocyte antigen (HLA), class I alleles have been identified. Despite the extensive polymorphism of the MHC class I molecules, each individual can express only up to six such variants (2 HLA-A, 2 HLA-B, and 2 HLA-C). This small set is responsible for presenting peptide antigens from all potential pathogens to CD8+ T cells (Thorsby, 1999). MHC molecules therefore should be able to bind many different peptides. It has been estimated that each MHC allotype binds about 0.5% of the universe of peptides (> 108 epitopes, assuming 8-mer peptides) (Yewdell & Bennink, 1999). Nevertheless, the particular set of peptides bound by an MHC molecule is specific for each allelic variant (Rothbard & Gefter, 1991). Thus one allelic MHC product will recognize one part of the universe of peptides, whereas another allelic MHC product will recognize a different part of this universe (Buus et al., 1987). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 MHC polymorphism, from a practical point of view, therefore imposes a huge challenge to the determination of MHC binding specificities. 1.3. Structures of peptide-MHC complexes The tertiary structures of an increasing number of MHC class I molecules have been determined by X-ray diffraction (Bjorkman et al., 1987a; Bjorkman et al., 1987b; Fremont et al., 1992; Madden et al., 1992; Silver et al., 1992; Zhang et al., 1992; Madden et al., 1993; Collins et al., 1995; Wang et al., 1995; Reid et al., 1996; Smith et al., 1996a; Smith et al., 1996b; Speir et al., 1999; Glithero et al., 1999; Torino et al., 1999; Maenaka et al., 2000; Hillig et al., 2001; Speir et al., 2001). The October 2001 release of the Protein Data Bank (PDB) (Berman et al., 2000), contains 68 entries for MHC class I molecules, comprising nine allele types from human, four from mouse and one from rat. Class I MHC molecules are crystallized as ternary complexes composed of a polymorphic heavy chain, a conserved serum protein /32-microglobulin (/32m), and an endogenously derived antigenic peptide (Figure 1.1) [for reviews see (Engelhard, 1994a; Engelhard, 1994b; Young et al., 1995; Madden, 1995; Jones, 1997)]. The heavy chain is composed of three extracellular domains, a transmembrane segment and a variable cytoplasmic domain. The a l and 02 domains of the extracellular portion together form the peptide-binding site. Each domain contributes a long a-helix and four strands of an eight stranded /3 - sheet that forms the floor of the peptide-binding site. The /3-sheet is relatively flat R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 with a small propeller twist. The antigenic peptides bind in the groove formed by the /3-sheet and flanking a-helices (Figure 1.2) (Bjorkman et al., 1987a; Bjorkman et al., 1987b; Garrett et al., 1989; Saper et al., 1991). MHC class I molecules bind peptides in an extended conformation (Madden et al., 1991; Matsumura et al., 1992; Fremont et al., 1992; Madden et al., 1993), leaving the main chain atoms and termini of the peptides open for direct interactions with residues of the MHC molecule. The amino and carboxy-terminal residues of the peptide are invariantly confined to conserved pockets (referred to as pockets A and F (Garrett et al., 1989; Saper et al., 1991)) at opposite ends of the groove, while the central region bulges away from the cleft toward the solvent. Most peptides bound to MHC class I are between 8 and 10 amino acids long (Hunt et al., 1992). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 a 2 al a3 (32 m Figure 1.1. Three-dimensional structure of class I MHC molecules. The heavy and light (P2m) chains are colored in blue and green, respectively. The peptide is displayed in yellow. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 al al Figure 1.2. MHC class I peptide binding site. The peptide is shown in yellow with the N- and C-terminii indicated. The MHC peptide binding site is shown in blue with the a helices indicated. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 1.4. Peptide-MHC binding interactions The majority of peptide-MHC hydrogen bonds involve peptide backbone atoms (Matsumura et al., 1992; Fremont et al., 1992; Madden et al., 1992; Bouvier & Wiley, 1994; Bouvier & Wiley, 1996; Bouvier et al., 1998). Since backbone atoms are common to all peptides, this explains how one MHC molecule can perform high affinity binding of a large and diverse repertoire of peptides. Recently, the Kangueane’s group has analyzed and shown quantitatively that MHC sidechain- peptide backbone interactions are prevalently dominant at the binding interface, suggesting the importance of the peptide backbone conformation during peptide- MHC binding (Kangueane et al., 2001; Adrian et al., 2002). Only a small amount of the binding energy involves peptide sidechain atoms. These interactions, however, are believed to explain the specificity of MHC molecules (Matsumura et al., 1992). The peptide binding groove in the MHC class I molecules has been described as featuring six pockets (A-F) (Bjorkman et al., 1987a; Bjorkman et al., 1987b; Garrett et al., 1989; Saper et al., 1991) which have been classified on the basis of their selective recognition of specific amino acid types in residues of the bound peptide (Zhang et al., 1998). Most MHC polymorphic residues are concentrated along the peptide binding site and hence determine the binding specificity of MHC class I alleles (Falk et al., 1991). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1.5. Water at the peptide-MHC binding interface The interactions of water molecules with the MHC proteins have also received attention in relation to peptide binding. They have been analyzed in X-ray structures of peptide-MHC class I complexes (Madden et al., 1992; Fremont et al., 1995; Smith et al., 1996a; Silz et al., 2001; Hillig et al., 2001) and in the context of molecular dynamics simulations (Meng et al., 1997; Meng et al., 2000a). Molecular dynamics simulations of peptide-MHC complexes have shown that water molecules are flexible in their positions and adjust in response to different peptide sequences (Meng et al., 1997; Meng et al., 2000a; Meng et al., 2000b). The flexibility of a water network at the peptide-MHC binding interface therefore allows the MHC sidechains to modify the binding surface to accommodate different peptide sequences, while still maintaining the binding specificity. This presence of interface water hence adds another level of complication to the mechanism of how peptides can bind to MHC class I molecules. Conserved water positions and their structural and functional roles in modulating peptide-MHC interactions have recently been studied by Ogata et al. (Ogata & Wodak, 2002). The information provided should be useful for the prediction and modeling of peptide-MHC interactions. 1.6. Predicting peptide binding to MHC molecules A number of computational methods have been developed for the prediction of MHC-peptide binding [for reviews see (Hammer, 1995; Buus, 1999; Lauemoller R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 et al., 2001)]. These methods can broadly be divided into two groups: sequence based and structure based methods. Sequence based methods Using data from allele specific binding experiments, sequence based methods include sequence binding motif analysis (Sette et al., 1989; Stevanovic & Rammensee, 1994; Meister et al., 1995; Rammensee et al., 1995), weighted matrices (Parker et al., 1992; Parker et al., 1994; Parker et al., 1995; Brusic et al., 1997; Schafer et al., 1998; Rammensee et al., 1999; Udaka et al., 2000; Lauemoller et al., 2000; Reche et al., 2002), artificial neural networks (ANN) (Adams & Koziol, 1995; Honeyman et al., 1998; Milik et al., 1998; Brusic et al., 1998a), hidden Markov models (HMM) (Mamitsuka, 1998; Brusic et al., 2002), and recently support vector machines (SVMHC) (Donnes & Elofsson, 2002). A comparison of predictions generated using binding motifs, weight matrices and ANNs shows that the statistical matrix is superior in eliminating false negatives and that the ANN is superior in eliminating false positives (Gulukota et al., 1997). However, the selection of the optimal prediction method depends on the amount of available data (the number of peptides of known binding affinity to the MHC molecule of interest), the biases in the data set and the intended purpose of the prediction (screening of a single protein versus mass screening) (Yu et al., 2002). When little or no peptide data are available, binding motifs are the most useful alternative to random guessing or use of R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 10 a complete overlapping set of peptides for selection of candidate binders. As the number of known peptide binders increases, weighted matrices and HMM become more useful predictors. Because binding data is not available for many alleles, sequence based methods have been used to predict peptide binding to very few MHC molecules. Structure based methods Structural approaches for prediction evaluate how well a peptide fits into the binding groove of an MHC molecule. Protein threading (Altuvia et al., 1995; Altuvia et al., 1997; Schueler-Furman et al., 1998; Schueler-Furman et al., 2000) and sidechain packing (Lee & McConnel, 1995; Rognan et al., 1999; Kangueane et a l, 2000; Doytchinova & Flower, 2001) techniques have been applied in molecular mechanics based peptide-MHC binding predictions. The molecular mechanics based binding prediction approach can be extrapolated to a wide range of MHC molecules. Obviously, a structural approach is limited to MHC types with a known structure. However, the advantage of a structural approach is that one known structure alone might be sufficient for creating a prediction model. Currently, no comparisons of the performance between structure and sequence based methods have been published. Other methods including molecular dynamics simulations (Rognan et al., 1992; Rognan et al., 1994; Meng et al., 1997; Meng et al., 2000a), a 3D quantitative structure-activity relationship (Doytchinova & Flower, 2002b), and a computational R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 combinatorial approach (Zen et al., 2001) have also been used to predict peptide binding affinity to MHC class I molecules. 1.7. Dissertation overview It was shown that sequence diversity in the central region of the peptide is accommodated by a combination of flexibility of polymorphic MHC residues at the bottom of the groove and variation of bound water molecules underneath the peptide (Smith et al., 1996a; Smith et al., 1996b; Meng et al., 1997; Meng et al., 2000a; Ogata & Wodak, 2002). Determining the precise binding interactions of peptide- MHC complexes undoubtedly requires consideration of the structural water molecules and the flexibility of MHC sidechains in the peptide binding site. This key issue is the main obstacle standing in the way of accurate predictions of the structure of peptide-MHC class I complexes for structure based peptide-MHC binding prediction methods. The aim of this research is to develop a structure based prediction method that can account for the water-mediated binding flexibility of peptide-MHC interactions. The specific class I MHC molecule considered in this work is HLA-A2. The dissertation is organized as follows. In chapter 2, evidence for the presence of water networks at the peptide-MHC binding interface is shown computationally by molecular dynamics simulations of different peptide-HLA-A2 complexes. An R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 12 algorithm for predicting water networks at the protein-protein binding interface is described in chapter 3. Chapter 4 describes the construction of a peptide backbone library. Conformational flexibility of the HLA-A2 peptide binding site is analyzed in chapter 5. In chapter 6, a method for predicting structures of peptide-MHC complexes is described. An attempt to predict peptide-MHC class I binding is shown in chapter 7. Finally, chapter 8 gives a brief summary of the dissertation work. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 13 Chapter 2 Molecular Dynamics Simulations of Peptide-HLA-A2 Complexes 2.1. Introduction Class I major histocompatibility complex (MHC) molecules bind short peptides with high affinity and sequence specificity, but a single MHC molecule is also able to accommodate a diverse range of peptide sequences. It has been proposed that water molecules trapped between the peptide and the MHC molecule are able to modulate the interaction, and may be responsible, in part, for allowing the MHC molecule to modify its binding face to accommodate different peptide sequences (Meng et al., 1997; Meng et al., 2000a). In computer simulations of the GILGFVFTL-HLA-A2 complex, Meng et al. have shown that water molecules must be present in the peptide-MHC binding groove for the peptide conformation to be maintained (Meng et al., 1997). In these simulations, the leucine side chain at position 3 of the peptide was seen to make contact with this water network (Figure 2.1). To provide evidence for the existence of the water network, the stability of the HLA-A2 complex with mutants of the peptide GILGFVFTL in which the position 3 sidechain contains a hydroxyl group was measured (Table 2.1) (Gurlo et al., 1999). The experimental results showed that the homoserine position 3 mutant peptide R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 14 GlhSGFVFTL forms a complex of higher stability with HLA-A2 than GILGFVFTL. Conversely, both of the mutants GITFGVFTL and GISGFVFTL form complexes with HLA-A2 with much lower stability. In this chapter, the results from molecular dynamics simulations of these peptide-HLA-A2 complexes are presented and shown to be consistent with the experimental data. Figure 2.1. Graphical representation of the peptide-HLA-A2 interface. Peptide- residues are labeled G1 through L9. Residues D77, Y99, R97 and Y116 are from HLA-A2. Water molecules as well as backbone atoms of both the peptide and the HLA-A2 molecule are displayed in space fill, side chains of both peptide and HLA- A2 residues in stick representation. Water molecules are distinguished from peptide and HLA-A2 backbone-residues by a darker shade. A space fill representation of the L3 leucine sidechain is superimposed, in stipple shade, on the stick representation of F5 V6 Y116 c L9 L3. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 15 Table 2.1. Experimental binding data for the interaction of peptides with HLA-A2. O.D. (574 nm) at peptide concentrationsa Peptide Sequence Position 3 sidechain 5/iM 50juM GlhSGFVFTL -CH2-CH2OH 1.25 1.54 GILGFVFTL (wt) -CH-CH(CH3 )2 1.21 1.42 GITGFVFTL -CH2OH 1.06 1.39 GISGFVFTL -CH(CH3 )OH 0.64 0.87 a Peptide-dependent surface expression of HLA-A2 molecules on T2 cells detected by BB7.2 (O.D. is optical density) for the indicated peptide concentrations (Gurlo et al., 1999). 2.2. Methods All calculations were performed using the AMBER 4.0 all atom force field on a Silicon Graphics Indigo2 workstation. The starting structure for all simulations was the X-ray crystal structure of the class I molecule HLA-A2 complex with a peptide, GILGFVFTL derived from the influenza matrix protein (Brookhaven entry Ihhi). Coordinates of the peptide and the a l and a2 were used in the calculations. The protocol for the solvation and simulation of the HLA-A2 / GILGFVFTL complex has been previously described in detail (Meng et al., 1997; Meng et al., 2000a). Using a similar protocol, molecular dynamics simulations for HLA-A2 complexed with three modified peptides in which the leucine at position three of GILGFVFTL has been mutated to homoserine (hS), threonine, and serine were performed. The charges and all other parameters used for homoserine have previously been calculated (Meng et al., 2000b). Each HLA-A2/peptide complex R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 16 was first placed in a 25 A radius TIP3P water sphere centered upon the center of mass of the peptide. Any water molecule within 2.0A of any solute atom was discarded from the calculation. Eleven water molecules were placed in the MHC peptide binding groove, based on molecular dynamics of the parent GILGFVFTL peptide complexed with HLA-A2. The groove water and the solvent shell were energy minimized for 1000 steps, followed by 26 ps of molecular dynamics equilibration (including an initial heating phase from 0 K to 298 K for 6 ps), and finally energy minimized for 1000 steps. Following water equilibration, each complex was energy minimized for 500 steps and then subjected to a 200 ps molecular dynamics simulation, including an initial heating phase from 0 K to 298 K in lOps. Coordinates were saved every 0.4 ps and a residue-based non-bonded cutoff of 6 A was used in all calculations. The protein backbone was loosely restrained to the X- ray coordinates, and three hydrogen bonds between terminal amino acids of the peptide and conserved residues of the MHC molecule were constrained to normal hydrogen bond distances. 2.3. Results Meng et al. have previously shown that the F5 xl dihedral angle is sensitive to the maintenance of the water-mediated interaction between the F5 carbonyl group and R97 of the MHC molecule (Meng et a l, 1997). For GlhSGFVFTL, xl (F5) retained the gauche+ conformation (Figure 2.2(a)) seen in the GILGFVFTL/HLA-A2 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 17 X-ray structure, and in general the peptide conformation remained essentially unchanged throughout the simulation (Figure 2.3). Hence, the leucine to homoserine substitution does not exert a strong influence on the peptide conformation. A snapshot of the GhSGFVFTL-HLA-A2 complex taken from the simulation (Figure 2.4) shows that the hydroxyl group of the homoserine sidechain does interact with the water network. This interaction is not disruptive to the water-mediated F5-R97 interaction, however. It is believed that it is the additional interaction of the hydroxyl group with the groove water networks that gives the GlhSGVFTL complex the higher stability. In contrast, for GITGFVFTL (Figure 2.2(b)) and GISGFVFTL (Figure 2.2(c)), xl (F5) rotated from a gauche to gauche' conformation after 80 ps and 40 ps respectively, and the conformation of the peptide moves away from that in the x-ray structure. This is caused by the ineffective interaction with the water network of the threonine and serine hydroxyl groups. The results shown here are consistent with the experimental data, which show that both peptides form complexes with HLA-A2 with much lower stability, compared to GILGFVFTL. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 18 A 180 1 2 0 < ® o i c < r e k_ ■ o -60 a -1 2 0 -180 0 50 100 150 200 Time (ps) B 180 120 < a » o > c < -60 a -120 -180 0 50 100 150 200 Time (ps) C 180 1 2 0 < a > o > c < r e ■ o o £ -60 Q -120 -180 0 100 150 50 200 Time (ps) Figure 2.2. Motion of the % 1 torsion angle of the F5 side chain (N-CarCfi-Cy) in molecular dynamics simulations of the HLA-A2 complexes of (a) GlhSGFVFTL, (b) GITGFVFTL and (c) GISGFVFTL. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 19 Phi lie 2 180 120 60 0 - -60 -120 -180 Psi Hse 3 180 120 60 - 0 -60 -120 -180 100 150 200 Phi Phe 5 180 120 60 -60 -120 -180 100 150 200 0 50 Psi V al6 180 120 60 0 -60 -120 -180 50 100 150 200 0 Phi Thr 8 -60 - 50 100 150 200 Psi lie 2 1 8 0 120 -60 -120 -180 100 150 200 100 150 200 Phi Gly 4 180 120 60 -60 -120 -180 0 50 100 150 200 Psi Phe 5 180 -1 2 0 - 100 150 200 Phi Phe 7 120 - I 1.1 III . 1. . . i l l ! Psi Thr 8 180 0 -60 - -120 -180 Phi Hse 200 Psi Gly 4 180 120 -60 -1 2 0 - -180 0 50 100 150 200 Phi Val 6 180 T 1 2 0 60 -60 -1 2 0 - -180 0 50 100 150 200 Psi Phe 7 180 50 100 150 200 1 2 0 - 60 - 0 -60 -1 2 0 -180 Phi Leu 9 180 100 150 200 120 60 -120 -180 100 150 200 50 100 150 200 Figure 2.3. Motion of the peptide backbone in a molecular dynamics simulations of the peptide GlhSGFVFTL complexed with HLA-A2. In each plot the indicated dihedral angle is plotted (in degrees) against time (ps) for 200ps of simulation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 20 hS side chain Figure 2.4. A structure of the GlhSGFVFTL / HLA-A2 complex taken from the molecular dynamics simulation (100 ps). The complex is viewed from the side showing only the peptide, Y99, R97 and D77 of the MHC molecule, and the water network between the peptide and the protein. The homoserine sidechain is identified, and it is apparent that the hydroxyl group of the sidechain interacts with the water molecules that form the peptide - MHC interface. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 21 2.4. Discussion Water mediation of ligand-receptor interactions is a well known phenomenon, and disordered water molecules within proteins have been identified (Ernst et al., 1995). Water can play a fundamental role in the interactions of proteins with other molecules (Janin, 1999). Water molecules can be displaced from a protein binding site by the interacting molecule, or be trapped at the protein-ligand interface following binding. In each case, the water molecules contribute (either entropically or enthalpically) to the ligand binding free energy (Ladbury, 1996). In the case of the peptide-MHC complex, it appears that water molecules are ‘trapped’ at the peptide-MHC interface. Hence, the affinity of ligands that could either interact with, or replace, these water molecules might be increased over that of the natural peptide (Meng et al., 2000b). Water molecules trapped at the peptide-MHC interface add a further level of complication to the mechanism of peptide-MHC interactions. A flexible network of water molecules allows protein sidechains to adjust their conformation in response to different peptides. In addition, changes in the chemical nature of the bound peptide impose a different pattern of water sites. Hence, the presence of interface water networks plays an important role in determining the affinity of specific peptide-MHC interactions. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 22 Chapter 3 WATGEN: An Algorithm for Modeling Water Networks at Protein-Protein Binding Interfaces 3.1. Introduction Water is known to play an important role in the recognition and stabilization of biomolecular complexes (Levitt & Park, 1993; Ladbury, 1996; Janin, 1999). Water at the interface of a complex can increase the promiscuity of an interaction, but it can also provide increased specificity and affinity. Structural data on protein- protein recognition sites indicate that water is present in abundance at interfaces (Lo et al., 1999). Interface water molecules contribute to the close-packing of atoms that ensures complementarity between the two protein surfaces, as well as mediating polar interactions between the two proteins. These water molecules form a complex network of interconnecting hydrogen bonds that are crucial for the stability of the protein-protein binding interfaces. The ability to model them will provide a better understanding of the structural basis of biomolecular interactions. While a variety of experimental techniques have been successfully used for solvent structural studies, the most detailed picture of protein hydration at the molecular level has come from high resolution, well-refined protein structures R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 23 obtained from X-ray and neutron diffraction. However, interpretation of the resulting solvent density map is usually complicated due to solvent mobility and disorder. As a result, it is possible to determine only a limited number of well- ordered solvent sites (Savage & Wlodawer, 1986). The more usual method for predicting water molecular sites around a protein surface is that of energy minimization or molecular dynamics simulations (Goodford, 1985; Phillips, Jr. & Pettitt, 1995; Henchman & McCammon, 2002). The disadvantage of this approach is that it is dependent on the accuracy of the potential energy functions and may also be a relatively long calculation for a macromolecule. Therefore, several computational approaches, mostly aimed at predicting the first hydration shell of the protein, have been developed (Pitt & Goodfellow, 1991; Roe & Teeter, 1993; Makarov et al., 1998) based on distributions of solvent molecules around polar residues in proteins (Thanki et al., 1988). The above computational approaches for predicting protein surface hydration, however, are not appropriate for predicting hydration sites at binding interfaces where hydration shells of both binding surfaces must be optimally positioned. In this chapter, the WATGEN algorithm, which predicts locations for interface water molecules given the atomic coordinates of a protein-protein complex, is described. The WATGEN algorithm has been tested on 140 protein-peptide interfaces and the results will be discussed. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 24 3.2. Methods WATGEN algorithm The computational procedure for predicting water sites is shown in Figure 3.1. The protein complex input consists of two user defined protein chains, which are designated either as the ligand or the receptor. In the preparation step, hydrogen atoms are added to the protein, then the prediction of water networks follows four sequential steps: (1) distribution, (2) scoring, (3) selection, and (4) optimization. The first three steps determine coordinates of the oxygen atoms of water molecules. The hydrogen atoms of water molecules are added in the fourth step. 4. Hydrogen bond optimization 1. Water distribution 3. Water selection Protein complex input 2. Water interaction Addition of hydrogen atoms Preparation V Water addition Figure 3.1. Computational procedure for predicting water networks. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 25 Preparation step: addition o f hydrogen atoms Hydrogen atoms are added to the protein based on standard amino acid geometry (see Appendix for amino acid parameters). Positions of polar hydrogen atoms of SER, THR, and TYR are determined by torsion angles at which the hydrogen atoms form the maximal number of hydrogen bonds (see Appendix for hydrogen bonding criteria). The torsion angles of SER HOG and THR HOG atoms are either 60°, 180°, or 300°. Similarly, the torsion angle of TYR HOH atom is either 0° or 180°. Histidine residues are protonated at either the 5 (HID) or e (HIE) positions, whichever forms the most hydrogen bonds. Step 1. Distribution In this step, water sites (W), defined by the oxygen atoms of water molecules, are distributed around hydrogen centers (H). There are two types of hydrogen centers, donor and acceptor. Donor hydrogen centers are hydrogen atoms directly bonded to protein hydrogen bond donor atoms, and they are precomputed in the preparation step. Acceptor hydrogen centers are not real atoms and they are defined just for the purpose of computing water networks. Table 3.1 shows the definition of acceptor hydrogen centers. The internal coordinate system for computing a water site is shown in Figure 3.2. The definition of reference atoms and the parameters for computing water sites are shown in Tables 3.2 and 3.3. Only water sites located within a user defined water box are computed . Water sites that make van der Waals R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 26 (VDW) clashes, interatomic distances less than 75% of total VDW radii (see Appendix for atom van der Waals radius), with protein atoms are excluded. Table 3.1. Definition of acceptor hydrogen centers Residue Fixed atoms Angles (°) Torsion (°) Hydrogen centers5 ASN CB, CG, OD1 120 0; 180 HOD1, HOD2 ASP CB, CG, OD1 120 0; 180 HOD11, HOD22 CB, CG, OD2 120 0; 180 HOD21, HOD22 GLN CG, CD, OE1 120 0; 180 HOE1, HOE2 GLU CG, CD, OE1 120 0; 180 HOE11, HOE12 CG, CD, OE2 120 0; 180 HOE21, HOE22 SER CA, CB, OG 109.5 60; 180; 300 HOG1, HOG2, HOG3 THR CA, CG, OG1 109.5 60; 180; 300 HOG11, HOG12, HOG13 TYR CE1, CZ, OH 120 0; 180 HOH1, HOH2 MC-Ob CA, C, 0 120 0; 180 HOI, H02 a Hydrogen centers are computed based on the positions of fixed atoms with the bond length of 1.010 A and the specified angle and torsion values b MC-0 represents main-chain oxygen atoms R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 27 ® . \ e pp Figure 3.2. Internal coordinate system for computing water sites. H is a hydrogen center around which water sites (W) are distributed, P is polar atom directly connected to H, and PP is the preceding atom of P, r is the distance between W and H, 0 is the angle between W-H-P, and x is the torsion angle (see Tables 2 and 3 for atom definitions and parameters). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 28 Table 3.2. Reference atoms and Hydrogen centers Residue Precedent atoms (PP) Polar Atoms (P) Hydrogen centers (H) Hydrogen center type ARG CD NE HNE Donor CZ NH1 HN11, HN12 Donor CZ NH2 HN21, HN22 Donor ASN CG OD1 HOD11, HOD12 Acceptor CG ND2 HND1, HND2 Donor ASP CG OD1 HOD11, HOD12 Acceptor CG OD2 HOD21, HOD22 Acceptor GLN CD OE1 HOE1, HOE2 Acceptor CD NE2 HNE1, HNE2 Donor GLU CD OE1 HOE11, HOE12 Acceptor CD OE2 HOE21, HOE22 Acceptor HIS3 CG ND1 HND Donor/Acceptor CG NE2 HNE Donor/Acceptor LYS CE NZ HNZ1, HNZ2, HNZ3 Donor SER3 CB OG HOG Donor CB OG HOG1, HOG2, HOG3 Acceptor (2) THR3 CB OG1 HOG Donor CB OG1 HOG11, HOG12, HOG13 Acceptor (2) TYR3 CZ OH HOH Donor CZ OH HOH1, HOH2 Acceptor(1) MC-0 C 0 HOI, H02 Acceptor TRP CD1 NE1 HNE Donor MC-N CA N HN Donor a If a donor hydrogen and an acceptor hydrogen center occupy the same position, the acceptor hydrogen center is not defined. Number of maximum acceptor hydrogen centers is indicated in parentheses. b M C-0 and MC-N represent main-chain oxygen and nitrogen atoms respectively R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 29 Table 3.3. Definition of water sites Type R(A )a Theta (°)a Phi (°)a Hydrogen center typeb 1 1.6; 1.9; 2.2 180 - Donor 2 1.6; 1.9; 2.2 135; 150; 165 60; 120; 240; 300 Donor, (not MC-HN) 3 1.6 1.9; 2.2 135; 150; 165 90; 270 MC-HN 4 1.6 1.9; 2.2 135;150;165 0; 180 Donor 5 1.8 2.0; 2.2 180 - Acceptor 6 1.6 1.9; 2.2 105; 120; 135; 150; 165 60; 120; 240; 300 Acceptor 7 1.6 1.9; 2.2 105;120; 135; 150; 165 0; 180 Acceptor a R, Theta, and Phi refers to the r, 6, and r of the internal coordinate system for computing water sites (see Figure 3.2) b Hydrogen center type refers to hydrogen centers (H) defined in Table 3.2. MC-HN is the hydrogen atom bonded to a main-chain nitrogen atom. Step 2. Interaction Water sites are categorized into two groups, ligand or receptor. Ligand and receptor water sites are those that were computed based on ligand and receptor hydrogen centers respectively. A ligand water site and a receptor water site are interacting if their separation distance is either < 0.5 A or between 2.6 - 3.9 A. Similary, two ligand water sites are interacting if their separation distance meets the same criteria. Non interacting water sites are eliminated at this step. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 30 Step 3. Selection The selection of water sites is based on a scoring system. Each water site is given a score equal to the number of its interacting water sites determined in step 2 and the number of polar atoms (nitrogen and oxygen) within a 2.6 - 3.2 A distance range. An additional 5 points are given to water sites of type 1, 4, 5, and 7 (see Table 3.3) for preferred hydrogen bonding geometry. These computed scores set the priority for selecting water sites. Starting from high to low scores, a water site is selected if it does not make any VDW clashes with the previously selected water sites. The VDW cut off distance between water sites is 2.6 A. Step 4. Optimization There are four hydrogen bonding sites (see Figure 3.3) at which two hydrogen atoms can be defined for a water molecule. Each hydrogen bonding site is given a score equal to the number of hydrogen bonds in which the water hydrogen is a donor hydrogen. The HI bonding site is given an additional 5 points if H is an acceptor hydrogen center. The top two scored hydrogen bonding sites defined the water hydrogen atoms. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 31 / ~ N i H3 N i H 1 i H2 Figure 3.3. Hydrogen bonding site geometry. W is a computed water site based on the hydrogen center H. HI, H2, H3, and H4 are possible hydrogen bonding sites for W. They are at 1.010 A away from W and are tetrahedrally spaced. W, HI, and H are collinear. The WATGEN algorithm was written in Java programming language and compiled with Microsoft Visual J++ 6.0. The computing time for a protein-ligand (< 30 aa) complex is less than 1 minute on a PC Pentium IV 2.0 GHz. WA TGEN prediction The WATGEN algorithm was used to predict locations of water molecules at 140 protein-peptide binding interfaces (Table 3.4) extracted from the Protein Data Bank (PDB) (Bernstein et al., 1977). An interface was selected based on three criteria: (1) there are only protein atoms in the PDB file, (2) there are at least two chains and the peptide chain has less than 30 amino acid residues, and (3) there is at least one water molecule that forms water bridges at the interface. The peptide and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. protein chains were designated as the ligand and receptor respectively. A water molecule can bridge the ligand and the receptor in two ways, (1) ligand-water- receptor or (2) ligand-water-water-receptor, and the bridge distance must be shorter than 3.5 A. A water box that extends 6 A from the minimum and maximum ligand coordinates was used. Each experimental interface water was mapped to the closest predicted water site. The predictive power of the WATGEN algorithm is measured in terms of the proportion of experimental waters with a predicted water site close by. An experimental water is paired if its assigned predicted water site is within a certain distance. Five such maximum separation criteria are used, viz. 0.6, 1.0, 1.4, 1.8 and 2.2 A. Interface Analysis Interface atoms (nAtom) are composed of ligand and receptor atoms within 5 A of each other. The number of polar atoms (nPolar) is the number of nitrogen and oxygen atoms of the interface atoms, carbon atoms of the peptide or protein that are not within 5 A of another carbon of the protein or the peptide respectively are defined as non-interacting carbon atoms (nonCi). The fraction of non-interacting carbon (nonCi_f) is the ratio of nonCi/nAtom. The number of polar contacts (nP) is the total number of contacts between the peptide and protein polar atoms (nPolar) within 3.5 A. The number of polar interactions per interface atom (nP_f) is the ratio of nP/nAtom. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 33 3.3. Results Predicted water sites at protein-peptide binding interfaces The WATGEN algorithm was used to predict water sites at 140 protein- peptide binding interfaces (Table 3.4). The atomic coordinates of these interfaces were taken from the PDB with x-ray resolutions between 1.7 - 3.2 A, R-factors between 0.156 - 0.294, and numbers of peptide residues between 1- 27 amino acids. There is a total of 847 experimental interface water molecules; Each binding interface has between 1 - 2 4 water molecules. An interface water molecule must bridge the peptide and the protein in one or two ways, (1) peptide-water-protein or (2) peptide-water-water-protein, and the bridge distance is < 3.5 A. Each experimental interface water molecule was mapped to its closest predicted water site. An experimental water molecule is paired if its assigned predicted water site is within a certain distance. Examples of paired water molecules for the 1I7U:C interface are shown in Figure 3.4. Ten experimental water molecules (white) were mapped to their closest predicted water sites (green) with the separation distances (A) indicated. Five maximum separation criteria are used, viz. 0.6, 1.0, 1.4, 1.8 and 2.2 A. Figure 3.5 shows the proportion of paired water molecules at different maximum separation distance. As expected, increasing the search volume around an experimental water increases the number of pairings between experimental and predicted water sites. At 1.4 A, 1.8 A, and 2.2 A maximum R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 34 separation distances, 58.68%, 76.86%, and 87.13% of experimental water positions are located respectively. The proportion of paired experimental water molecules for each binding interface is summarized in Table 3.4. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3.4. Agreement between experimental a n d predicted water sites a t peptide-protein binding interfaces _____________ "N o Interface3 NRESb Resolution R - NHOHc NINTd NPRED6 % 0.6f % 1.01 % 1 .4 f % 1 .8 f %2.2 o o C 3 P h ° < 3 '0 0 0 h ~ 0 0 00 ~0\00 0 0 ^ 0 0 0 o o JO o o o in o o o r- in in m o o o o o o o o v o c n i n v o o t - ' - o i n c n S S o o S i n m « o o h ' o o o h O \ h h H i j i n i r i i ' o o f ' c O O i n i O c n i n O O i o 2 2 i n O O ^ O cocnincncocnOiOO><,-.in,~,~,cNin o o © o o o o o o o o o o o ro v o o ,— . o ^ O o g c n o ^ m , — ,,— , 0 0m o vo > < o in vo m o o X i o o ^ ^ o o o I S S o © cn m o i o o r f X . 0 0 0 0 0 0 0 r f ' o \ o m v i T t i t H h i n a i O \ h ' O i i o o i o o oi i> c -~ o m co N- it in o\ IN oo vo to O VV O< N00 < NC N< Nin in in O o o 00 i n T-H N 00 o vo T — 1 o o i n in in O v O v O No o o i— H m 1 C OC OC OC OC O t-H i— h '— i » — i < NC NC NC NC NC N if If V O H m cn oo cn in t-H r-H T — H N OV OV OV Oc n IN IN 0 0 0 0 0 0 in in oo V O o o N O ^ H O N 0 0 0 0 t > oo o t -H t-H i-H O v O v O v t-H r-H t— ( V OV O in C N< N'— 1C Nr— Ht -H ^ H t-H ’— i < N< N< N(N i— i * — i > — i C N< n (N r-H t -H C NC N O o o o o d d O o o o o o o d d o o o o d d d cn co m in M fi fi d in o cn C N in vq m N ; in in nr vq vq vq C NC NC Nin in N ; 00 c n c n c-i C Nc n C NC N c n ri C NC NC NC NC NC NC NC N oo oo 00 00 oo O v m in in t-H C N C N C N C N 22 u U U O y o P p I — i P P QP P h p m U Q R E j H - j Q P P U U X u Q Q o o O O o a O' c/ X X <3 < <3 <3 i— H <3 t — H <3 r-H <3 < r-H < ^ H < t— H P Q »-H C N C N V O P Q m cNm-^-invor-'OOCTv ( N fO T fin 'O h o o o v O r tN fn t — i » — i i — i 1 — H i — i » — i t — i t — i C N C N C N C N 35 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3.4. Continued _______ _ _ Interfacea NRESb Resoiutio n rT NHOHc NINTd NPREDC % 0.61 % 1.0f % 1 .4 % 1 .8 * ' % 2.2 t- o - 4 — » w O O O 0 0 ^ 0 o X i n m oo oo i> o § o 2 m o CN ^ 2 O O 2 t"- 00 N O O o ° o ° 0 4 10 — 2 u - > id c n 0 0 0 ^ 0 ^ ^ * 0 0 0 0 ^ 0 ^ 0 0 0 0 0 ^ , 0 ^ ^ 0 0 o x u-1 oo o OXrn -H XX mO OX Or H 0 0 0 j , 0 ! n [ T 0 0 in^lOC'-^^OOOOC'-^mCN o o o o o ^ ^ o o m o o g o o ^ 2 2 2 2^0^1‘ v c ^ 0 rr>v 0 l^ > 2 l0 0 0 0 0 0 ^ 0 0 > < C N C N C N X © © , ~ 11~ 4 r - - C N O © c n O O N O i n c n c N in i n 0 0 cN-'frcNincncNCNNOcNcnTf- o o o o ^ J ^ o o ^ o o o o "T o NO NO f - . o o 4-C CN * —* ^ lo o o o cn ■ '1 - ■'t cn CN CN m CN 00 4-4 Tj- -^•C^OINOOO'^- 0\H0\OWf'00 0\(fl'H>OTt ^H-'+cnTj-cncncncncNcNin''t 00 O O 00 CN CN CN in vo vo in ID N- CN o o t"- 1 —< i-H on O n c n on VO t - I D c n i n m 0 0 0 0 o r - O n o \ 4 — I c n o n C \ o o C N o o c n m C N C N N " N " I D I D c n N " N " i - H i—H ND NO CN CN N" ■*t CN CN CN O n CN ID c n CN CN 00 0 0 O n • n • n CN CN >n N - r - OO c n c n c n \ l CN N CN i n N - O O n ON O n O n O n O n N" N - CN CN CN CN CN CN CN CN CN > —i CN CN i-H T “H 4 —1 CN CN CN CN o © © © © © O o O O © O © © © © © © © © © © © OONoJ2," ' rncnONtCininONON CN CN ^ CN CN CN no ^ CN <N CN c n cn © CN CN ^ ^ o o o o c n o oIO JC? CN CN CN VOhhOOOOOsOiOlhN i n i n CN CN c n cn Q Q Q Q Q Q Q W W W W W W O ' Q02?^0' 0' ' h w B B S ^ O ^ X < 1 < ; w w S S D w p Q p q m o o o u o u o ■^■in'Of-'Oooo^HCNcn-^-in'OO-oooo^cNcn^inio (NcNcNNdMcncnncncncncncncncn^^N't^Tf't 36 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3.4. Continued _______ _ I n t e r f a c e a N R E S b R e s o l u t j on r T N H O H c N I N T d N P R E D C % 0 .6 * ' % 1 . 0 f % 1 .4 * ' % 1 . 8 f % 2 . 2 ) h O h - > 1 3 P h O X < N v/v vo oo x vo o r-' O 0000,^00000 0 0 0 0 0 0 0 X ^ 0 0 0 0 0 _ l — I _ _ i 'v^h _ l _ H o o o o o o o o o o O I— I VO O O V Tf O O W in V O C T \ o o o o o o o o g ^ f^ o o o o o o o o o vo2<n 2 voocoO \ © ^ co o o o O O O O O ^ J ^ O O O O O o o o CN§©VOcO(-5'^-,/_,voVOO„,-.CO§COOOOO,-5c-5 2 o O ' t 2 m ' O r o ° i n ^ M T ( - i n 0 0 f n 2 fbN(v|0 0 2inin ( v | i n M r o o r , o o n n tO(o o o ^ o o ^ j ^ o o o ^ o t - 3 o o ^ ! o o V O m CO C N O O C N t ' ~ O V C O C N C N ^ ^ l - O O O O I ~ ' - C V C N © t ' - O O ^ H O O V O ' v t ^■N 't(SO \TfcoTt^, coco(Ocorfininincoco^, T f^ f - 00 CO VO VO C" CO CN CO CN CO 2 » CO CN 'N' o O O O o o o o CO co co CO VO VO VO VO N" ■Vt N" CN CN CN CN d d o o r-H CN CN CN CN C N i n c o c o a, H H N N (N(N(NH 0 0 0 0 0 - Tt,rt' ^ - 1 - 1 — 1 c t \ W VH fyr. N " N " '—i CN CN CN CN 00 ov oo co i — > ov ov O O v O v _ _ ' t CO OV CN CN 1 o d d vn i n N N O VO OV Ov CN t -h t -h *-h d o d o VO CO CN VO VO VO VO VO OO OO c n N" ■vt CO CO CO 1 — H t- H i - H CN CN CN CN CN CN CN CN o d d d d d © © VO VO VO VO VO 1 /N vq vq l> I> i> V ) 0 0 m CN (N CN CN CN CN 1-3 N N N Ov ^ N" N" CN CN CN CN d o d o vq cn cn vo CN CN CN CN Ov ©\ ©\ Ov OV OV VO O o v v o v o o v o v o v o v o v o o o v 00 ov ov vo UQWfefflUAH g 2 2 2 £ U Q U K o u O Ph Ph *-H Ph i— H P h H Ph C N d w M-Ti n Xl NOf f i l OUt NMCNU wwWWWt LO'v tov nn iLf eO ^OOi Jw i-H i-H i-H »-H i—I ,-H i— ( ,-h y U f t P h w * > * " co O O O O v O - —i C N C O T t - v O V O I > O O O V O ^ H C N c O ' ^ - V O V o r ~ O O O V N" N" ol" VO VO VO VO VO VO VO VO VO VO VO O VO -O vo vo VO vO vO vO 37 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3.4. Continued _______ _ Interfacea NREgb Resolution ^ NHOH° NINTd NPRED® % 0.6* % 1.0* % 1.4* %1.8*' % 2.2 5 - H o 4 - > o P h o © o o o o o o 0 0 , ^ 0 C N C N O V O in O O V 00 V O t— 00 0000 , -0000,-00 0 0 0 0 ^ 0 0 0 0 ^ 0 0 u i X X m c -- 00 c - 0 0 ^ 0 m X o o cn no o oo cl o m m n o n o O O O O ^ ^ - O O O ^ , O O O O J o ^ O O O ^ J O O O ^ X ^ o o o o o o N o X o i n i n o o h “ 0 ' n ^ i n i n o o “ ' n ^ h < = > o o o o o o o § © i> © c n o o o o 0 © ( -) 0 incNS§§oe -,c -5ON- 2b'inin' nH(Siotn0 'n0 0 r H i o 2 S 2 ' n 0 0 (S^ O O O C N O O r - , , - 1 0 0 O N o ^ ^ o o o o i n ^ o o o X . i o o m N O in C N C N C N o i n H H i n ' O H ^ M ' t i o o i ' t i n h H o o o o H h o i o o cncnenN'enenenencniOinininN'incNininininininen N OO NO N 85 o o O O0 0 N O n - C N o m c n o N O N O N O V O N " N " N " N- N " N " tT Tt c n n o C N oo oo cn cn — .. — ^ C N C N C N 00 H- N O C N in C N N O O C N 00 c n n o o o o C N C N C N O N r - H 0 0 n o c n e' en C N t"- cn C N N O in C N C" no o C N N O O C N V O o C N N O o C N cn C N cn cN O N O C N Ov o C N in Ov t—H C N Ov C N C N t -H t -H C N 0 0 C N C N C N C N C N in Ov t—H Ov O C N t -H C N C N r-H C N © © d d d d d d d d d d d O d o d d d d d d 0 0 0 0 N; 0 0 C N C N C N C N 0 0 0 0 0 0 oo m 0 0 t-H C N cn in in IT) vo cn t-H 1 C N t -H C N C N C N C N C N C N C N C N C N cN C N C N C N C N C N C N C N C N 38 ov ov O V 1 2 1 2 Ov O V ov ov ov ov O V ov On m in cn t -H oo O V Ph o U y O O y O < U U Q y y O h P h P h Q P4 i> i — i t -H y y H H oo H H r-H cn cn cn cn Ov ov w H in X o Q in U r-H go u r-H d O cn H H t -H t " i — i T — H 00 H H t -H t — H r-H ^H 1 a H H t -H a i — < o H ”S T — H K H -i T — H Ph * — > Ph 1 — S t -H Ph T — H 00 3 t ^ H o -H (N c n itin iO h o o o \O H C N c c i'tic iio h o o o \O H N t ' ^ r ' - t - - - i ' - - C ' ' C ' - C ' ' - t ' - - t ' - t ' ' 0 0 o o o o o o o o o o o o o o o o o o o v a v o v R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. CN CN O O H h o V O © sP ■ O £ K O Pi '"C T 3 U "3 o u Tf cn v mm .o a H CD O C n — I t-H CD O £ o o o o o o o o c n o o o o VO u~> oo o vo in oo in o o o o CN o o ^ OO IN IN 00 o o o o o o o o o o o o o o ‘r > o o ^ - ) , r ‘ 'O 0 0 o r r i '- l 0 0 0 0 C ' o o , r i o o rn V O V O C O IN O O V O C O O vo m vo ro co O O CO CO O - t CO CO CO CO o o o o o o CO CO O' o CO O' lo co io co O CO O' (N (N o o o o O CN ( N O C O V D I O O O - O iciin^co^-ovocNooioin VO CO Ov '— < CO CN CN O CO cn o m O CN CO co o o o CO CN 0-0 0 Ov cn n o m n O 00 VO ° IO CO ° Tf CN —I ocfcnocoviciO vH H O vf COCOCOCNCNCOCNCOCOO-OO OCNfClCO^CNtNCO't^^j ■y-.O'^J’ CN'OOOOvOVCOVOCNCO ’ (NooovoincoHiovficvoo V D O O CN CN -— ' CN co •'3- co O' vo 00 00 00 00 IN r - in in in in in in n C N t"- 00 C N Ov (N t-H cn cn cn cn V O V O V O V O r- o in in r - r - 'O Ov (N vo cn (N <N C N cs t — H 1 T-H C N i— H T — H r^* r-H C N T — H r-H in r-' in t-- i '' t " n r-' 0 0 oo oo oo vo o oo o Ov Ov r - H ov r - H ov r - H O V T -H Ov IN - 0 0 IN 0 0 C N C N C N C N o C N C N C N r — i r-H i - H r-H ■ ,1 , . ’ , ! C N C N ' , ^H l-H r - H r - H ^H ^H C N C N C N o d o o © o o o o o o o o o o d o o o d d o o o o o o o o o o o c o c o n : CN ^ ,— h ""C h — h CN o o o v o v o v o v o v o v o v o v CN O V N N IN N IN IN CN IN IN N CN t - H CN CN t — H ,-P CN 00 Ov 00 CN rH CN CN rH tvj 00 cywfeOffi cn vt in vo o oo ov O v ov C N i— H I N r-H 20 vo 00 oo T — H 0 0 r-H 0 0 t -H o vo 0 0 r H O O ov 0 w U 0 Ph Q P h P h Ph P P C N C N d d a i Pi 00 d o 0 0 d o I N P < n v b h - 3 H -j H H D 0 P P o ffl P P 3 3 s f c c i * 4 J J 1 — H 1 — 1 ^ H T — H 1 — 1 t-H T — H r— H r-H T — H O C N C N C N vo I N 0 0 ov O r-H C N c n • ^ T O o O O o O o o r-H o r-H o t-H T — H t - H t -H t — H t ^H t -H t -H 39 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 3.4. Continued _______ _ Interfacea nRggb Resolution NHOHc NINTd NPREDC % 0.6f % 1.0r % 1.41 % 1 .8 t l % 2.2 V — H V J — , iy ~ , O . O ^ 0 -y -1 I— ^ i*. lO O . 1 * -. o i g © £ > o o o g o o o © £ ; £ o ! C © o o £ £ o © in cn © i— i m m i n h oo Ov i n ~ C -~ m o o o o o o o o o o o o o or-mXXr-HCNXin 0 0 ^ ^ 0 ^ * 0 0 0 oot^-oo2S ~ooTt;ri> O i n cn i n t-- oo © ■ s j - i n m 2 v o c N 2 i n 2 © i n , ~ , i n © v o o i t ( » h 2 i n ^ “ h 2 i n h 0 M i n o i n m 0 t s § o § O N H i n ^ m (N<n'“''st-~vo~VOCNI>CN;icN c o o o .—. i n c o c o c o i n i n cn c o c o o o o ^ o g o o ^ ° o ^ ^ o o o o o ® o o o ^ S o o i n c i h i f o o o i h i j i O i r f O i c o c o i f m t N M c o oo o- co K.OlOcOOOOl^'t'OCO 1 c i N c o c o ^ i n c o o m CO CN C in in ov t - co vo CN ^H no vo C O _ _ . O in ov no N O cn cn O N On f — s Hi Hi vo vo vo N O N O 00 C N 2 ^ CN o o o 00 00 — N V O N V O V O ov Ov C O < N C N T — H in T-H t-H t -H t - H T-H cn cn T — 1 t -H N co C O T-H 40 ! h o c 3 (X i CN m Ov CN V O in o 00 cn 00 00 o o <N CN 00 * — H T — H t - H oo 00 H CN C OO N" t-H H- 0 0 o On O N o O oo CN CN CN o o t> CN CN CN CN CN CN r-H . t -H t -H T— H I . T — H ^ H t -H CN <N CN r-H ^ H t -H © © O © O © o o o o © o © d o d d o d d o d CN CN CO CN CN CN m ov 'tOOOOC't^in^^O\0\^^CO^t'^-CNCN(NcO CN CN CN CN CN CN CN CO CN CN CN Q WmP H f e y y ^ ^ ^ L) Q > < > H Ph < £ >>>: p q O n cy(?i (?i o'*?i O' ^ Ei Pq id pi < £ o s ^ 8 s § £ CN CN CN CN CN CN vo r- 00 O N o T — H CN cn in vo c-~ o oo o r H CN cn N * un N O r- 00 t - H t- H t - H t - H CN (N CN CN (N CN CN CN CN CN cn cn cn cn cn cn cn cn cn t- H t_ h H t - H t - H H 1 1 - 1 T — H t- H r -H t- H t - H T — H t - H R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.4. Continued No Interfacea NRESb Resolution (A) R- Factor NHOHc NINTd NPRED6 % 0.6* % 1.0* % 1.4* % 1.8* % 2.2 139 2VAB:P 9 2.5 0.159 117 10 49 20 60 70 90 90 140 3F58:P 11 2.8 0.2 1 1 36 0 0 100 100 100 a Filename in Brookhaven Protein Databank (PDBID) and the peptide chain '’ Number of residues of the peptide chain c Number of experimentally determined water molecules d The number of experimental water molecules at the interface e The number of predicted water sites at the interface fThe percentage of paired experimental waters (within a maximum separation distance of 0.6, 1.0, 1.4, 1.8, and 2.2 A respectively) Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 3.4. Paired water molecules at the 1I7U:C interface. Only oxygen atoms of interface water molecules and the ligand chain C are shown. Experimental water molecules are shown in white. Paired predicted water sites are shown in green, and other predicted water sites are shown in red. Separation distances (A) between experimental and predicted water sites are shown for paired water molecules. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 100 (0 a > 3 a a > a > + -> (0 ■o a > re CL c a > o o CL 60 - 20 0 76.86 58.68 33.65 14.64 87.13 0.6 1.0 1.4 1.8 2.2 Maximum Separation Distance (A) Figure 3.5. Proportion of paired water molecules at different maximum separation distances. 44 The temperature factors o f experimental waters It is possible that one of the reasons why some experimental water sites cannot be correlated with predicted sites is because of their high mobility. Such an effect could be seen by examining the experimentally derived temperature factor (B- value) of the paired and unpaired water molecules. Figure 3.6 shows the B-value distribution for paired water molecules at different maximum separation distances (0.6, 1.0, 1.4, 1.8 A) and for unpaired water molecules (maximum separation distance >1.8 A). As the maximum separation distance is increased from 0.6 to 1.8 A, the B-value distribution for the paired waters is shifted to the higher range. This shows water sites with low mobility were better predicted. Unpaired water molecules having maximum separation distance > 1.8 A have a higher proportion of water molecules in the high temperature factor range than paired water molecules do. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 10 20 30 40 50 60 70 Temperature factor (B-values) 80 □ d < 0.6 A ■ d < 1.0 A □ d < 1.4 A □ d < 1.8 A ■ d >= 1.8 A Figure 3.6. Temperature factor distributions for experimental water molecules paired at different maximum separation distances. 4 ^ L /i 46 The polar contacts o f water sites The number of polar contacts is measured by the number of protein polar atoms (nitrogen and oxygen) within 3.5 A of a water molecule. Figure 3.7 shows the distribution of the number of polar contacts (# NO) for experimental interface water molecules. Only 2.24 % of water molecules have no polar contact. About 93 % of water molecules have between 1 - 4 polar contacts. Figure 3.8 shows the mean separation distance between experimental and predicted water sites having different number of polar contacts. As the number of polar contacts increases, the mean separation distance decreases. The mean separation distance for water sites with 0, 1, 2, and > 2 polar contacts are 2.66 A, 1.76 A, 1.28 A, and < 1.16 A respectively. This indicates water sites with high number of polar contacts were better predicted. The proportion of paired experimental water molecules having different numbers of polar contacts at different maximum separation distance is shown in Figure 3.9. At 1.8 A maximum separation distance, 10.53%, 55.24%, 83.33%, and 88.02% of experimental water sites having 0, 1, 2, and > 2 polar contacts respectively are located. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. < /> © 3 O © © © S M — o c © o © C L 30 27.63 26.92 24.79 25 20 13.58 15 10 4.84 ■ 5 2.24 0 0 1 2 3 4 >4 Number of Polar Contacts (# NO) Figure 3.7. Distribution of polar contacts for experimental water molecules. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 3 T — 2.5 o o c 5 2 U ) £' I 1-5 (0 O S Q. 0) (/> 1 - - I 0.5 2.66 1.76 I S m W B m + + 1.16 + 1.06 + + 0 1 2 3 4 >4 Number of Polar Contacts (# NO) Figure 3.8. Mean separation distances for paired water molecules with respect to the number of polar contacts. 4 ^ OO Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 100 1 W 0) 3 O 0) o < D 4^ (0 TJ (0 Q _ C < D O l_ 0) a. 60 - £ 4 0 - 20 - o io£j c o oco < = > co C O ° ° ° l I ■ NO = 0 H NO = 1 □ NO = 2 □ NO > 2 0.6 1.0 1.4 1.8 Maximum Separation Distance (A) 2.2 Figure 3.9. Proportion of paired water molecules categorized by the number of polar contacts at different maximum separation distances. 50 Predicting number o f water sites There are more predicted water sites than experimentally determined ones at protein binding interfaces. Here, the number of predicted water sites are correlated with three characteristics of protein-peptide binding interfaces using data from the 140 interfaces listed in Table 2. Firstly, the number of polar atoms (nPolar) is a measure of the interface hydrophilicity. Secondly, the fraction of non-interacting carbon atoms (nonCi_f) is a measure of the interface hydrophobicity. Finally, the strength of intermolecular interactions at a protein-peptide binding interface is measured by the fraction of polar contacts per interface atom (nP_f). Equations 1 - 3 show the relationships between the square root of the number of water sites (SqWat) with three variables nPolar, nonCi_f, and nP_f. The number of water sites is positively correlated with the interface hydrophilicity and is negatively correlated with the interface hydrophobicity and the strength of intermolecular interactions. Figure 3.10 shows the plot of calculated numbers of water sites using equation 3 versus predicted values. Using equation 3, there is a 74.6 % agreement between the calculated and WATGEN predicted number of water sites at binding interfaces. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 51 Equation 1: SqWat = 3.251 + 0.0484*nPolar (R2 = 0.656) Equation 2: SqWat = 3.817 + 0.0477*nPolar - 2.997*nonCi_f (R2 = 0.717) Equation 3: SqWat = 4.435 + 0.0481*nPolar - 3.703*nonCi_f- 7.287*nP_f (R2 = 0.746) SqWat: square-root of the number of water at the interface nPolar: number of nitrogen and oxygen at the interface nonCi f: fraction of carbon not interacting with another carbon at the interface nP_f: number of polar interactions per interface atom R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 12 i 10 - < /) a) CO ro 3 O C 5 o y = 0.7459x+ 1.6053 R2 = 0.7459 12 10 8 6 0 2 4 WATGEN Predicted Water Sites Figure 3.10. Correlation between the calculated (Equation 3) and WATGEN predicted square root of the number of water sites (sqWat) at peptide-protein binding interfaces. to 53 3.4. Discussion Unlike other algorithms aimed at predicting the locations of a protein’s first hydration shell (Thanki et al., 1988; Pitt & Goodfellow, 1991; Roe & Teeter, 1993), WATGEN was developed specifically to predict water hydration sites at protein- protein / protein-peptide binding interfaces. Results from the prediction of water sites at 140 protein-peptide interfaces are promising with 59% and 77% of experimental water sites being successfully predicted within maximum separation distances of 1.4 A and 1.8 A respectively. A separation distance of 1.8 A is considered here because the x-ray resolution of 72 % of the interfaces is greater than 2.0 A. Almost all experimental interface water molecules (98%) form at least one polar contact with the protein-peptide complex. WATGEN aims at predicting water sites that can form multiple hydrogen bonds bridging the binding interface. Therefore, water sites with high number of polar contacts were better predicted. For water sites with 2 and more than 2 polar contacts, the proportions of successful prediction within 1.8 A maximum separation distance are 83% and 88% respectively. WATGEN may be further improved by considering distributing water molecules around apolar amino acid residues. Also, a possible reason why some experimental water sites cannot be correlated with predicted sites is because of their high mobility. The incorporation of water molecules at a binding interface is generally believed to be energetically unfavorable because of the larger entropic cost of trapping highly mobile water molecules. Therefore, it is assumed that an interface R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. that leaves no space between the interacting molecules will give a higher binding affinity than one in which the interface contains gaps filled by water. Yet in some cases the enthalpy gain that results from making extra water-mediated hydrogen bonds is greater than the entropic penalty that must be paid for immobilizing the water involved. To optimize the water-mediated protein-protein interactions, it is the aim of WATGEN to predict such water sites that form multiple hydrogen bonds. The presence of interface water networks can structurally or functionally play a role in conformational stability, dynamics, plasticity, specificity and selectivity of protein-protein interactions. Nevertheless, extensive interface water networks may imply weak direct interactions between the protein molecules. Indeed, the number of predicted interface water sites was shown to negatively correlate with the strength of intermolecular protein-peptide interactions. Surveys of structural data on protein-protein and protein-DNA recognition sites (Jones & Thornton, 1995; Nadassy et al., 1999; Lo et al., 1999) indicate that water is present in abundance at interfaces. Generally, crystallographic structural data with a resolution better than 2.0 A are required to permit reliable discrimination of water molecules. However, there is still no established practice among crystallographers to report solvent positions, and their number is probably underestimated, even at high resolution. To model solvent positions, the WATGEN program provides a method for predicting the likely positions of water molecules at binding interfaces. Besides protein-protein binding interfaces, the WATGEN has R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 55 recently been extended to predict water networks at DNA and RNA binding interfaces. Future work could aim to further extend the WATGEN algorithm to recognize additional interfaces including carbohydrates and user-defined molecules. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 56 Chapter 4 Peptide Backbone Library 4.1. Introduction MHC class I molecules bind peptides in an extended conformation [for reviews, see (Engelhard, 1994a; Madden, 1995)]. Hydrogen bonds anchor both peptide termini to conserved MHC polar residues at each end of the binding groove. The conserved nature of the electrostatic interactions fixes the direction and optimal length of bound peptides. In most cases these peptides have a length determined by the MHC allele, such as nonamers for HLA-A2 (Falk et al., 1991). Some MHC class I molecules (e.g. HLA-Aw68 (Guo et al., 1992)), can accommodate longer peptides and still have the peptides amino and carboxyl termini fixed in their groove; the extra length is accommodated by a bulging of the middle part of the peptide. There are only a few exceptions to this mode of binding (Collins et al., 1994; Wang et al., 1995; Smith et al., 1996b). The conformation of the first three N-terminal peptide residues and the two C-terminal peptide residues is very nearly identical in the many different peptide-MHC structures, while the central region shows much greater variability (Figure 4.1). In this chapter, a library of backbone conformations for nonameric peptides is constructed based on the available structural knowledge. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 57 Figure 4.1. Backbone conformational variation of four nonameric peptides bound to HLA-A2. PDB codes of peptides are lhhi, lhhj, lhhk, and lhhg. Peptide residues are colored by positions in the sequence and labeled accordingly, with PI at the N- terminus of the peptides and P9 at the C-terminus. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 58 4.2. Methods Figure 4.2 shows the parameters for peptide backbone calculations. The positions 1 and 9 of a peptide were fixed to the coordinates listed in Table 4.1. These coordinates were obtained from the mean positions of four structurally aligned peptides (PDBID codes lhhi, lhhj, lhhk, and lhhg). Positions 1 and 9 have rmsd < 0.3 A and therefore are structurally conserved. The Phi and Psi torsions were incremented at 30° intervals. Both trans and cis peptide bonds were considered and the Omega torsions were 180° and 0° respectively. To make the computation feasible, each CA atoms of each an acceptable peptide backbone conformation was restricted to a position specific conformational space called a “CA box” (Table 4.2). Values for Phi, Psi, and Omega torsions were obtained for each computed peptide backbone conformation. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 59 CA 1.335 A/J21.9>s1-449 A 1.552 A p h i \ i n 1 PS| \ - j 16.6°/om ega CA 120.5' 1.229 A Figure 4.2. Peptide backbone parameters. Table 4.1. Coordinates of peptide positions 1 and 9 Position Atom RMSD (A)a X Y Z 1 N 0.000 0 0 0 1 CA 0.271 0.656 0.662 1.056 1 C 0.221 2.123 0.320 0.918 9 N 0.153 21.761 0.724 0.116 9 CA 0.169 22.719 -0.056 -0.657 9 C 0.126 24.105 0.000 0.000 a Obtained from structurally aligned lhhi, lhhj, lhhk, and lhhg peptides R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 60 Table 4.2. Peptide CA Box Position RMSD Volume Dimensions (A) Box centers (A)a (A3 ) X Y Z X Y Z 1 0.271 38.720 3.272 3.441 3.439 0.656 0.662 1.056 2 0.314 41.815 3.353 3.554 3.509 4.442 1.081 1.073 3 0.325 40.687 3.471 3.175 3.692 7.029 -0.990 2.859 4 0.604 58.140 4.061 3.572 4.008 8.245 0.337 6.115 5 0.690 59.289 3.764 3.443 4.575 11.691 -0.201 6.349 6 1.155 100.266 4.504 4.371 5.093 14.525 1.293 4.821 7 0.523 48.395 3.789 3.182 4.014 17.567 -0.297 3.834 8 0.184 34.697 3.283 3.334 3.170 20.458 1.377 2.075 9 0.169 34.124 3.351 3.302 3.084 22.719 -0.056 -0.657 a Obtained from structurally aligned lhhi, lhhj, lhhk, and lhhg peptides 4.3. Results Architecture o f peptide backbone library There are a total of 454774 computed peptide backbone conformations. These conformations were divided into different sets and stored in a Microsoft Access database (Table 4.3). Table T contains conformations with all trans peptide bonds. Tables C2 - C9 contains peptide conformations with one cis peptide bond at one position between 2 - 9 . Table All contains all computed peptide backbone conformations. The database was designed this way to allow a faster search for peptide backbone conformations using structured query language (SQL) commands. Table T provides peptide conformations for all peptide sequences. Tables C2 - C9 provide additional conformations for peptides with one proline residue at a position 2 - 9. Table All is only used when the peptide sequence has two or more pro line R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 61 residues. To minimize the storage space, values for Phi and Psi torsions were stored as indices between 0-11 corresponding to 30° incremented values between 0 - 330°. Similarly, values for Omega torsions were stored as 0 or 1 corresponding to 0° or 180° respectively. Table 4.3. Peptide backbone library Tables Number of backbone conformations Cis-peptide position All 454774 All T 30597 None C2 0 2 C3 36223 3 C4 20666 4 C5 30074 5 C6 18562 6 C l 8148 7 C8 5668 8 C9 0 9 Assessment o f the peptide backbone library To ascertain that the library contains all possible peptide conformations, four x-ray peptide backbone conformations (PDBID codes lhhi, lhhj, lhhk, and lhhg) were used to search for their closest matches. Figure 4.3 shows pairs of x-ray and computed peptide conformations for the four peptides. All pairs have an rmsd value between 0.657 - 0.854 A and these peptides have very different conformations and sequences. These results indicate that the peptide backbone library provides a sufficient set of diverse conformations. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 4.4. Discussion The constructed peptide backbone library contains an exhaustive set of nonameric peptide backbone conformations that can fit in the MHC class I peptide binding site. Using these backbone conformations as structural templates for predicting peptide conformations provides an adequate sampling of the peptide conformational space. Ab initio peptide structure prediction therefore can be achieved. The computational approach developed here can be used to construct other peptide backbone libraries of different lengths. The inclusion of cis-peptide bonds is preferable for predicting low energy conformations of proline residues. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 64 Chapter 5 Conformational Flexibility of HLA-A2 Peptide Binding Site 5.1. Introduction MHC amino acid sidechains have been observed to adopt distinct conformations to accommodate binding to different peptides (Fremont et al., 1992; Madden et al., 1993; Collins et al., 1994; Fremont et al., 1995). In comparing three H2-Kb structures, four MHC sidechains (LYS66, GLU152, ARG155, and TRP167) have been shown to take on peptide-specific conformations (Fremont et al., 1995). Similarly, in the comparison of five HLA-A2 complexes, three residues (ARG97, TYR116, and TRP167) also change conformation in response to peptide binding (Madden et al., 1993). Residues ARG97 and TYR116 have been shown to undergo a concerted conformational change upon the binding of different subsets of peptides. This conformational change seems to depend on the size of the peptide C-terminal sidechain as well as the packing of the peptide sidechain two residues before the C- terminus. The binding of peptides that extend out of the C-terminal region of the peptide-binding site also requires the conformational adjustments of TYR84 and LYS146 (Collins et al., 1994). The other conformational changes occur at, or near, the surface of the MHC molecule (LYS166, TYR84, LYS146, GLU152, ARG155, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 65 and TRP167). In this chapter, conformational flexibility of the HLA-A2 peptide binding site is analyzed. 5.2. Methods Analysis o f HLA-A2 peptide binding site Thirty eight X-ray crystallographic conformations of the HLA-A2 molecule (Table 5.1) were obtained from the PDB. Dihedral angles of protein residues within the peptide binding sites were computed for assessment of structural conservation. Table 5.1. Conformations of HLA-A2 No PDBID: Chain No PDBID: Chain 1 1AKJ:A 20 1I1F:A 2 1A07:A 21 1I1F:D 3 1B0G:A 22 1I1Y:A 4 1B0G:D 23 1I1Y:D 5 1B0R:A 24 1I4F:A 6 1BD2:A 25 1IM3:A 7 1DUY.-A 26 1IM3:E 8 1DUY:D 27 1IM3:I 9 1DUZ:A 28 1IM3:M 10 1DUZ:D 29 1JF1:A 11 1HHG:A 30 1JHT:A 12 1HHG:D 31 1QR1:A 13 1HHH:A 32 1QR1:D 14 1HHI:A 33 1QRN:A 15 1HHI:D 34 1QSE:A 16 1HHJ:A 35 1QSF:A 17 1HHJ:D 36 2CLR:A 18 1HHK:A 37 2CLR:D 19 1HHK:D 38 3HLA:A R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 66 A generic HLA-A2 peptide binding site Twenty nine conformations of peptide / HLA-A2 complexes (Table 5.2) were obtained from the PDB. The peptide / HLA-A2 complexes were aligned so that the N l, C3, and C9 atoms of the peptides have the XYZ coordinates of (0, 0, 0), (X, 0, Z), and (X, 0, 0) respectively. The backbone coordinates of aligned HLA-A2 molecules were averaged. Using the averaged backbone structure as a scaffold, protein side chains were computed using the average bond lengths, angles, and torsion angles of protein atoms from the 29 conformations. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 67 Table 5.2. Conformations of peptide / HLA-A2 complexes No PDBID Protein chain Peptide chain 1 1AKJ A C 2 1A07 A C 3 1B0G A C 4 1B0G D F 5 1BD2 A C 6 1DUZ A C 7 1DUZ D F 8 1HHG A C 9 1HHG D F 10 1HHI A C 11 1HHI D F 12 1HHJ A C 13 1HHJ D F 14 1HHK A C 15 1HHK D F 16 1I1F A C 17 1I1F D F 18 1I1Y A C 19 1I1Y D F 20 1IM3 A C 21 1IM3 E G 22 1IM3 I K 23 1IM3 M 0 24 1JHT A C 25 1QR1 A c 26 1QR1 D F 27 1QRN A C 28 1QSE A c 29 1QSF A c R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 68 5.3. Results Structural conservation o f HLA-A2 peptide binding site Table 5.3 shows dihedral angles of HLA-A2 protein side chains in the peptide binding site. These are average values obtained from 38 crystallographic conformations of the HLA-A2 molecule (Table 5.1). There are a total of 36 protein side chains in the peptide binding site. Out of this total, there are 12 side chains (HIS70, THR73, HIS74, LEU81, VAL95, ARG97, HIS 114, TYR116, LYS146, GLN155, LEU156, and THR163) that are not structurally conserved. A side chain is not structurally conserved if one or more of its chi angles have standard deviations > 30°. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 69 Table 5.3. Dihedral angles of HLA-A2 protein side chains in the peptide binding site No Position Name 5C la (°) X2a (°) X3a (°) X4a (°) Structurally Conserved1 5 1 5 MET 184 (18) 174 (20) 68 (29) - Y 2 7 TYR -78 (3) 97(6) - - Y 3 9 PHE -51 (3) 91 (6) - - Y 4 33 PHE 65 (3) 103 (6) - - Y 5 34 VAL -72 (8) - - - Y 6 45 MET 185 (4) 69 (6) 103 (17) - Y 7 59 TYR 179 (4) 86 (6) - - Y 8 63 GLU -65 (6) -54 (7) 104 (12) - Y 9 66 LYS -68 (7) 185 (11) -73 (28) 171 (20) Y 10 67 VAL 81(5) - - - Y 11 70 HIS -76 (5) 20 (67) - - N 12 73 THR 204 (84) 20 (28) - - N 13 74 HIS -77 (6) 60 (57) - - N 14 76 VAL 176 (23) - - - Y 15 77 ASP -69 (5) 163 (7) - - Y 16 80 THR 182 (4) 57 (10) - - Y 17 81 LEU -70 (9) 186(60) - - N 18 84 TYR -71 (3) 167 (6) - - Y 19 95 VAL 229 (92) - - - N 20 97 ARG 180 (7) -166 (7) 106 (55) 217 (40) N 21 99 TYR 62 (3) 74(4) - - Y 22 114 HIS 181 (4) 11 (83) - - N 23 116 TYR -127 (41) 87(16) - - N 24 123 TYR 180 (3) 72(4) - - Y 25 124 ILE -170 (3) 164(10) - - Y 26 143 THR 176 (4) - - - Y 27 146 LYS 224 (44) 181(12) 208 (57) 175 (32) N 28 147 TRP -71 (5) 3(4) - - Y 29 152 VAL 174 (6) - - - Y 30 155 GLN -93 (67) 163 (53) -14 (65) - N 31 156 LEU -93 (23) 114 (82) - - N 32 159 TYR 169 (4) 78 (6) - - Y 33 160 LEU -57 (6) 168 (6) - - Y 34 163 THR 149 (71) - - - N 35 167 TRP -79 (3) 87(11) - Y 36 171 TYR -64 (4) 152 (5) - - Y a Average values of 38 structures with standard deviations in parentheses b A residue is structurally conserved if all standard deviations are <30°, Y = Yes, N = No R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 70 A generic HLA-A2 peptide binding site Figure 5.1 shows the computed generic HLA-A2 backbone framework and 29 superimposed peptide-HLA-A2 conformations. The two a-helices and P-strand regions are very conserved while more deviations are seen in the loop regions. Figure 5.3 shows the distribution of RMSD for Ca atoms. About 65% of Ca atoms have RMSD < 1.0 A and more than 91 % have RMSD < 1.4 A. Figure 5.1. Conservation of HLA-A2 backbone atoms. The superimposed Ca traces of 29 peptide-HLA-A2 conformations are shown (HLA-A2 is blue, peptide is yellow). The computed generic HLA-A2 backbone atoms are in stick form and magenta color. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. o f th e copyright owner. Further reproduction prohibited without permission. 0.50.60.70.80.91.01.11.21.31.41.51.61.71.81.9 2.0 RMSD (A) □ Percent Cumulative Percent Figure 5.2. Distribution of root mean squared deviation of HLA-A2 Ca atoms. Cumulative Percent 72 5.4. Discussion The computed generic HLA-A2 peptide binding site was oriented relative to the coordinates of the peptide backbone conformations in the peptide backbone library (chapter 4). This is necessary for the correct docking of the peptide N- and C-termini and positioning of the peptide in the binding groove. Since the backbone framework of MHC class I molecules is highly conserved, this generic HLA-A2 peptide binding site can be used as a structural backbone template for threading different MHC class I sequences. Analyses of the HLA-A2 protein sidechains show that most of the sidechains in the peptide binding site are structurally conserved in response to binding of different peptide sequences. For flexible docking of peptides, keeping these structurally conserved sidechains fixed should therefore be acceptable. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 73 Chapter 6 MHC: A Program for Predicting Binding Conformations of Peptide - MHC Complexes 6.1. Introduction Structure based prediction of peptide-MHC binding interactions requires a method that can reliably and accurately predict the conformation of both the peptide and MHC molecule. To date, methods of various degrees of accuracy have been developed to model the peptide structure in the groove (Sezerman et al., 1993; Rosenfeld et al., 1993; Rosenfeld et al., 1995; Sezerman et al., 1996; Desmet et al., 1997; Schueler-Furman et al., 1998; Schueler-Furman et al., 2000), to evalulate the compatibility between a peptide and an MHC molecule (Altuvia et al., 1995; Vasmatzis et al., 1996; Altuvia et al., 1997), and to estimate the free energy of binding of a peptide to an MHC molecule (Altuvia et al., 1995; Sezerman et al., 1996; Altuvia et al., 1997; Froloff et al., 1997; Rognan et al., 1999). For all these methods, the prediction scheme is based on a fixed MHC structure and a relatively few peptide backbone conformations. In addition, no water molecules were included in the structure prediction schemes. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 74 However, it was shown that sequence diversity in the central region of the peptide is accommodated by a combination of flexibility of polymorphic MHC residues at the bottom of the groove and variation of bound water molecules underneath the peptide (Smith et al., 1996a; Smith et al., 1996b; Meng et al., 1997; Meng et al., 2000a; Ogata & Wodak, 2002). Determining the precise peptide conformation undoubtedly requires consideration of the structural water molecules and the flexibility of MHC sidechains in the peptide binding site. Inclusion of water and protein sidechain flexibility is therefore the main obstacle standing in the way of accurate predictions of the structure of class I bound peptides using the above methods. It is the aim of this chapter to develop a better method for predicting structures of peptide-MHC complexes. To achieve an accurate model of peptide- MHC binding interactions, the MHC program developed in this chapter considers three important key factors: (1) adequate sampling of peptide backbone conformations, (2) flexibility of MHC sidechains, and (3) modeling of explicit interface water molecules. The peptide backbone library developed in chapter 4, analysis of MHC side conformational flexibility in chapter 5, and the WATGEN algorithm for solvent modeling developed in chapter 3 therefore are of great importance for the development of the MHC program. To achieve a robust calculation, a system of contact energy scores is developed to efficiently rank and select peptide-MHC conformations. For a given peptide sequence, the MHC R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 75 program provides a small set of ranked predicted peptide-water-MHC conformations as output. The MHC program is shown to predict structures of peptide-HLA-A2 complexes with a good accuracy. 6.2. Methods Computational approach The structure of a peptide-MHC complex was divided into two parts: the constant frame and the flexible part. The constant frame consists of the MHC backbone and fixed amino acid sidechains. The flexible part consists of the entire peptide and flexible MHC sidechains. The constant MHC frame was previously determined in chapter 5. In the case of the HLA-A2 molecule, there are 12 sidechains (Table 6.1) that show significant conformational differences when binding different peptides. These sidechains, therefore, are considered flexible and were predicted based on the MHC constant frame. Peptide backbone conformations were previously determined in chapter 4. Based on the MHC constant frame and peptide backbone conformations, the peptide sidechains were predicted according to a specified peptide sequence. The compatibility of peptide-MHC conformations was evaluated based on a VDW contact energy score (Table 6.2). Water molecules were then added to the binding interfaces of peptide-MHC structures with low VDW contact energy scores using the WATGEN algorithm developed in chapter 3. A scoring function was developed based on the number of hydrogen bonds and VDW R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 76 contact energy score to select the most probable peptide-MHC binding conformations for output. A summary scheme for the approach is shown in Figure 6 . 1. Table 6.1. Flexible HLA-A2 residues No Residue Number Residue Name 1 70 HIS 2 73 THR 3 74 HIS 4 81 LEU 5 95 VAL 6 97 ARG 7 114 HIS 8 116 TYR 9 146 LYS 10 155 GLN 11 156 LEU 12 163 THR Table 6.2. VDW contact energy score for non-bonded atoms Atom 1 Atom 2 Interatomic distance / total VDW radii3 Energy All All < 50 % 100 All All 50 % to < 65 % 3 All All 65 % to < 85 % 1 All All 85 % to < 95 % 0 N N 85 % to < 95 % 1 0 0 85 % to < 95 % 1 C C 95 % to < 105 % -1 0 N 95 % to < 105 % -2 a Distance between atoms 1 and 2 divided by the sum of the atoms’ VDW radii R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 77 MHC residues MHC backbone MHC conformations Output conformations Peptide sequence Peptide backbone Peptide conformations Peptide-Water-MHC conformations Peptide-MHC conformations Figure 6.1. Computational procedure for predicting binding conformations of peptide-MHC complexes. General sidechain conformation prediction scheme Sidechain conformations or rotamers of an amino acid residue were computed based on the combination values of rotational torsion angles (Table 6.3). Each rotational torsion angle was incremented by a 120° interval starting at 60°. Based on the backbone positions, sidechain rotamers corresponding to the amino acid sequence were added. A VDW contact energy score was calculated for each rotamer using the atomic pair wise energy scores listed in Table 6.2. Rotamers having a VDW contact energy score >100 were excluded. Sidechain rotamers of the whole sequence were randomly combined and their total VDW contact energy scores R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 78 were computed. Combinations with a VDW contact energy score >100 were excluded. This prediction scheme is shown in Figure 6.2. Table 6.3. Rotatable torsion angles of amino acid sidechains No Name x ia X2a X3a X4a nRotate 1 ALA 0 2 GLY - - - - 0 3 SER OG - - - 1 4 THR OG1 - - - 1 5 LEU CG CD1 - - 2 6 ILE CGI CD1 - - 2 7 VAL CGI - - - 1 8 ASN CG OD1 - - 2 9 GLN CG CD OE1 - 3 10 ARG CG CD NE c z 4 11 HIS, HID, HIE CG ND1 - - 2 12 TRP CG CD1 - - 2 13 PHE CG CD1 - - 2 14 TYR CG CD1 - - 2 15 GLU CG CD OE1 - 3 16 ASP CG OD1 - - 2 17 LYS CG CD CE NZ 3 18 PRO - - - - 0 19 CYS SG - - - 1 20 MET CG SD - - 2 a Torsion angles define the locations of the specified atoms b Number of rotational torsions R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 79 Single sidechain conformations Random combinations Backbone conformation Contact energy score Joint sidechain conformations Figure 6.2. Computational procedure for predicting sidechain conformations. HLA-A2 conformations Conformations of the HLA-A2 peptide binding site with 12 flexible protein sidechains (Table 6.1) were predicted based on the MHC constant frame. Using the sidechain conformation prediction scheme described above, a maximum of 10 HLA- A2 conformations were selected from 50 randomized sidechain combinations. The sidechain combination selection was performed sequentially and the first 10 combinations with a total VDW contact energy score <50 were selected. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 80 Peptide conformations Based on a peptide backbone conformation obtained from the peptide backbone library developed in chapter 4 and the MHC constant frame, peptide sidechain conformations were predicted in the same manner as were the HLA-A2 sidechains. The number of randomized sidechain combinations was the highest number of peptide sidechain rotamers in the sequence or 20, whichever was higher. A maximum of 10 sidechain conformations per one peptide backbone conformation was obtained. The total VDW contact energy score for all predicted peptide conformations was <50. Peptide-MHC conformations For each peptide backbone conformation, the predicted peptide conformations were evaluated for compatibility with the predicted MHC (HLA-A2) conformations. The number of random combinations of the peptide and MHC conformations was set equal to the product of their conformations. To save computer storage space, only the most compatible peptide-MHC conformations, having a total VDW contact energy score less than a cutoff value, were kept. The cutoff value, initially 75, was dynamically set equal to 10 plus the minimum VDW contact energy score of already selected peptide-MHC conformations. This prediction process was repeated for all peptide backbone conformations available for R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 81 the specified peptide sequence (see chapter 4 for the peptide backbone conformation searching strategy). Peptide- Water-MHC Conformations Peptide-MHC conformations were ranked based on their VDW contact energy score. The top 1000 conformations with VDW contact energy scores less than the minimum VDW contact energy score plus 10 were selected. Water networks at the binding interfaces of the selected peptide-MHC conformations were modeled using the WATGEN algorithm developed in chapter 3. A water box with the minimum and maximum XYZ coordinates (in A unit) of (-6.000, -10.000, - 8.000) and (30.000, 10.000, 20.000) respectively was used. Output conformations Each peptide-water-MHC conformation was given a final score based on the equation: 10*Hbond - 5*VDW where Hbond is the number of hydrogen bonds (see Appendix) and VDW is the VDW contact energy score. Conformations with a final score greater than 500 or 85% of the maximum final score were selected for output. A summary of MHC output conformations is shown in Appendix C. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 82 6.3. Results The MHC program was used to predict binding conformations of 13 peptide- HLA-A2 complexes with known X-ray structures (Table 6.4). To assess the accuracy of structure predictions, RMSD values were calculated for all output peptide conformations. For visual comparison, top scored peptide conformations (cyan) and peptide conformations with the lowest RMSD values (green) are shown in Figure 6.3. Most calculated RMSD values (lowest ones from the outputs), ranging between 1.590 - 2.421 A, are within the structure X-ray resolutions. The only two exceptions are the 1DUZ and 1IM3 structures whose calculated RMSD values are higher than the X-ray resolutions. One of the best predicted structure is the 1QR1 (Figure 6.3(d)). Virtually, all peptide sidechains were correctly positioned and the peptide backbone conformation is in very good agreement with the X-ray structure (RMSD = 1.590 A). The worst predicted structures is 1IM3 (RMSD = 2.421 A). However, for the peptide LLFGYPVYV, five different peptide-HLA-A2 X-ray conformations were obtained, 1A07, 1BD2, 1DUZ, 1HHK, and 1IM3. For 1A07 and 1BD2, the peptide LLFGYVYV forms complexes with both the HLA-A2 and TCR molecules. Two distinct peptide X-ray conformations were observed: (1) in complexes with both HLA-A2 and the TCR (Figures 6.3 (g,h)), and (2) in complexes with HLA-A2 (Figures 6.3 (i, j, k)). A conformational difference was seen for the P6 residue which points downward for peptide-HLA-A2-TCR complexes and points upward for peptide-HLA-A2 complexes. It is interesting that the P6 residue of the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. predicted peptide conformation (which is the same one for all five complexes) points downward and therefore is in good agreement with the peptide-HLA-A2-TCR structures. In general, the conformations of the first three and the last two amino acid residues of the peptides are very well predicted while the central region is quite varied. The predicted conformational differences of the central regions could possibly reflect other preferred binding conformations as was seen in the case of the peptide LLFGYPVYV. Table 6.4. Peptide-HLA-A2 complexes used in the MHC prediction No PDBID Peptide Sequence RESa (A) RMSD (A) References 1 1BOG ALWGFFPVL 2.50 2.316 (Zhao et al., 1999) 2 1I1F FLKEPVHGV 2.80 2.112 (Kirksey etal., 1999) 3 1HHI GILGFVFTL 2.50 1.974 (Madden et al., 1993) 4 1QR1 IISAVVGIL 2.40 1.590 (Kuhns et al., 1999) 5 1AKJ ILKEPVHGV 2.65 1.801 (Gao et al., 1997) 6 1HHJ ILKEPVHGV 2.50 1.830 (Madden et al., 1993) 7 1A07 LLFGYPVYV 2.60 2.298 (Garboczi et al., 1996) 8 1BD2 LLFGYPVYV 2.50 2.346 (Ding et al., 1998) 9 1DUZ LLFGYPVYV 1.80 2.390 (Khan et al., 2000) 10 1HHK LLFGYPVYV 2.50 2.338 (Madden et al., 1993) 11 1IM3 LLFGYPVYV 2.20 2.421 (Gewurz et al., 2001) 12 1HHG TLTSCNTSV 2.60 2.052 (Madden et al., 1993) 13 1I1Y YLKEPVHGV 2.20 2.121 (Kirksey et al., 1999) a X-ray Resolution b The lowest all atom RMSD of output peptide conformations R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. (a) ALWGFFPVL (1B0G) Figure 6.3. Predicted peptide conformations. Predicted structures are shown in green (best matched) and cyan (top scored), X-ray crystal structures are shown in white. Peptide N-terminal is to the left and C-terminal is to the right. (b ) FLKEPVHGV (1I1F) 85 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (c ) GILGFVFTL (1HHI) 86 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (d ) IISAWGIL (1QR1) 87 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (e ) ILKEPVHGV (1AKJ) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (f) ILKEPVHGV (1HHJ) 89 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (g ) LLFGYPVYV (1A07) 90 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (h ) LLFGYPVYV (1BD2) 91 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (i) LLFGYPVYV (1DUZ) 92 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (j) LLFGYPVYV (1HHK) 93 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (k ) LLFGYPVYV (1IM3) 94 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued ( 1 ) TLTSCNTSV (1HHG) 95 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued (m ) YLKEPVHGV (1I1Y) 96 \ - * i \\ J - . . . \ \ \ ' I R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 6.3. Continued 97 6.4. Discussion The ability of the MHC program to predict bound conformations of MHC class I peptides with a fairly high accuracy is a remarkable success given the fact that no crystal structure template for the peptide backbone conformation was used. In addition, the prediction of peptide and MHC sidechain conformations did not rely on a backbone-dependent rotamer library (Dunbrack, Jr. & Karplus, 1993) or on an MHC-specific sidechain rotamer library (Schueler-Furman et al., 1998; Schueler- Furman et al., 2000) for sidechain placements. The unavoidable combinatorial problem associated with the docking of fully flexible peptides and selected flexible MHC sidechains was efficiently tackled by the MHC program’s hierarchical approach of eliminating incompatible conformations. Despite its simplicity, the VDW contact energy score was shown to work well in evaluating and selecting highly compatible conformations of peptide-MHC complexes. Because the association of peptide-MHC complexes is not surface complementary, the modeling of explicit water molecules at the peptide-MHC binding interface was extremely important in determining the correct conformations of bound peptides. It is evident from the results that not all peptide sidechain and backbone conformations were in good agreement with the X-ray structures. The majority of conformational variability was seen in the central region of the peptide. As pointed out in the results section, it is possible that a peptide sequence can adopt different bound conformations as seen in the case of the peptide LLFGYPVYV. Therefore, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 98 predicted bound peptide conformations, although different from the observed crystal structures, might still possibly be correct conformations in a different environment (e.g. binding to the TCR). Because the inherent fluidity of a water binding surface may hypothetically enable a peptide to adopt different interchangeable bound conformations, the MHC program was designed to predict a set of possible conformations of peptide-MHC complexes rather than a single one. Variations in the predicted structures should provide a better insight into the dynamics of water - mediated peptide-MHC interactions. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 99 Chapter 7 Predicting Peptide-HLA-A2 Binding Interactions 7.1. Introduction Structure based prediction of peptide-MHC interactions requires not only a method to reliably model the structure of peptide-MHC complexes but also a predictive function to relate structural properties to binding affinities. In this chapter, structural properties of peptide-HLA-A2 complexes are analyzed for different binding affinity groups. A predictive function is developed that can be applied to estimate the binding propensity of peptide-HLA-A2 complexes. 7.2. Methods The MHC program (developed in chapter 6) was used to predict the binding conformations of 342 peptide-HLA-A2 complexes (Table 7.1). The experimental binding affinity of these peptides to HLA-A2 was obtained from the MHCPEP database (v. 1.3) at the website (http://wehih.wehi.edu.au/mhcnepA (Brusic et al., 1998b) and there are 133, 123, and 86 peptides having high, medium, and low binding affinities respectively. To characterize binding interactions, sequence and structural properties of the predicted peptide-HLA-A2 binding conformations were R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 100 analyzed for different peptide binding affinity groups. For each peptide binding affinity group, the mean values of the sequence and structural binding properties and the 95% confidence interval of the means were determined. The difference in means of any two groups was statistically compared using t-tests. 7.3. Results Table 7.1 summarizes the sequence and structural properties of predicted binding conformations of 342 peptide-HLA-A2 complexes. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 7.1. Peptide binding affinities to HLA-A2 No Sequence A ff nAb NOc BBd Pepe Comp. Wats C onf H 201 Energy1 HBk INT1 1 AAAKAAAAV 1 51 19 30597 305970 305966 456 100 70.77 -0.63 126.92 4.06 2 AAGIGILTV m 56 20 29456 294560 290275 961 100 71.33 -0.60 137.01 5.37 3 AAPTPAAPA m 53 20 1068 9895 10287 195 100 61.37 6.81 108.45 5.39 4 AAVDLSHFL 1 68 26 27607 274732 270965 138 100 69.95 -6.79 128.96 6.00 5 AIAKAAAAV m 54 19 30046 300460 300458 184 100 68.66 -7.62 126.90 4.80 6 AILGFVFTL h 69 20 27996 266770 256753 30 30 65.53 -14.87 123.17 5.43 7 ALACAAAAV m 53 18 30371 303710 303539 672 100 61.25 5.43 114.42 4.49 8 ALADAAAAV m 53 20 30371 303710 303708 17 17 56.47 -0.29 96.47 4.29 9 ALAIPQCRL m 69 23 10899 103365 108533 150 100 67.71 8.09 123.10 5.65 10 ALAKAAAAA 1 52 19 30371 303710 303654 31 31 65.03 0.97 119.00 3.61 11 ALAKAAAAI m 55 19 30371 303710 303532 76 76 62.83 -3.57 110.41 3.74 12 ALAKAAAAL h 55 19 30371 303710 303625 1001 100 77.27 -17.60 150.07 6.45 13 ALAKAAAAM m 57 19 30371 303710 302425 246 100 73.02 -0.68 133.03 5.50 14 ALAKAAAAN 1 55 21 30371 303710 303697 138 100 69.72 -1.66 124.80 5.42 15 ALAKAAAAR 1 58 22 30371 303707 299355 1001 100 76.17 1.95 147.59 5.74 16 ALAKAAAAT 1 54 21 30371 303710 303653 305 100 65.53 8.35 118.63 4.39 17 ALAKAAAAV m 54 19 30371 303708 303689 588 100 71.56 -4.96 131.95 4.44 18 ALAKAAAEV 1 58 21 30370 303700 302806 141 100 70.93 5.24 124.08 3.91 19 ALAKAAAFV m 60 19 29145 291450 290555 532 100 75.52 -8.51 141.31 4.92 20 ALAKAAAGV m 53 19 30371 303707 303680 153 100 68.55 2.27 120.09 3.89 21 ALAKAAALV m 57 19 30323 303230 292113 581 100 74.60 4.56 145.74 5.90 22 ALAKAAAPV m 56 19 8428 84280 80853 723 100 74.85 -3.28 146.38 6.05 23 ALAKAAEAV 1 58 21 30371 303710 303462 819 100 84.63 -1.02 160.21 5.55 24 ALAKAALAV m 57 19 30314 303138 303111 568 100 79.89 -2.44 159.60 5.44 25 ALAKAANAV m 57 21 30362 303620 303569 667 100 76.24 -6.15 148.53 5.99 O Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 7.1. Continued No Sequence A ff nAb NOc BBd Pep6 — — --------------j— Comp. Wat8 Confk H 201 Energy1 HBk INT1 25 ALAKAANAV m 57 21 30362 303620 303569 667 100 76.24 -6.15 148.53 5.99 26 ALAKAAPAV m 56 19 11901 119010 118999 605 100 75.27 -2.68 144.56 6.13 27 ALAKAAYAV m 61 21 27951 279453 265119 158 100 72.48 -9.67 131.88 5.35 28 ALAKAIAAV m 57 19 30204 302038 301969 579 100 74.06 -0.33 141.65 6.00 29 ALAKAPAAV m 56 19 12197 121969 121915 609 100 67.63 6.47 131.08 6.17 30 ALAKARAAV 1 60 22 30355 303550 303306 830 100 77.86 -8.21 149.51 6.19 31 ALAKAYAAV m 61 21 29816 298160 295019 121 100 71.48 -7.91 131.98 6.24 32 ALAKEAAAV m 58 21 30335 303350 302848 423 100 78.26 -6.08 150.56 5.97 33 ALAKGAAAV m 53 19 30371 303710 303700 102 100 67.04 6.67 116.95 3.86 34 ALAKLAAAV m 57 19 30096 300960 300783 940 100 74.55 -5.48 145.90 6.35 35 ALAKNAAAV m 57 21 30339 303389 303205 274 100 74.41 1.53 144.60 5.97 36 ALAKYAAAV m 61 21 29353 293530 293261 1001 100 81.96 -6.02 162.59 6.60 37 ALAPAAAAV m 52 18 13292 131526 132738 877 100 60.64 -0.54 109.83 3.84 38 ALATAAAAV m 52 20 30333 303330 302849 726 100 64.99 3.04 121.61 4.23 39 ALAVAAAAV m 52 18 30325 303250 303231 304 100 58.58 -0.54 103.23 4.21 40 ALCRWGLLL 1 74 22 29900 298248 268705 56 56 69.59 -7.89 123.68 5.09 41 ALEKAAAAV m 58 21 30327 303270 303200 899 100 82.96 0.03 156.94 5.66 42 ALFAAAAAV h 56 18 29663 296630 296074 1001 100 69.35 -17.81 131.72 5.85 43 ALFKAAAAV m 60 19 29636 296359 293841 1001 100 81.19 -8.79 156.18 5.46 44 ALIHHNTHL m 74 34 28670 285728 279697 67 67 75.27 -9.75 137.49 5.33 45 ALKKAAAAV m 58 20 30267 302670 302614 674 100 76.81 7.02 146.01 5.56 46 ALMDKSLHV h 71 27 30123 301172 297175 205 100 75.49 2.47 139.92 5.43 47 ALMKAAAAV m 59 19 30345 303450 303432 363 100 72.52 7.02 132.75 3.88 48 ALMPLYACI h 71 20 12969 126946 128271 178 100 68.23 -2.60 124.08 5.57 O to Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 7.1. Continued No Sequence A ff nAb NOc BB Pep6 A * Comp. Watg Confh H 201 Energy1 HB INT1 49 ALNKFMCQL m 76 23 30181 301436 295096 53 53 73.09 2.70 130.57 5.13 50 ALNKMFCQL h 76 23 30014 300052 294822 2 2 73.00 -18.00 150.00 6.00 51 ALNKMFYKL 1 80 24 27594 273633 249470 15 15 72.40 -10.73 128.47 5.47 52 ALNKMLCQL 1 73 23 30250 302446 298023 283 100 77.39 0.08 148.86 5.47 53 ALSKAAAAV m 55 21 30369 303686 303678 75 75 64.17 0.53 110.97 4.04 54 ALSRKVAEL m 68 26 30169 301690 293705 161 100 83.30 -8.82 153.55 5.39 55 ALSTGLIHL m 64 26 30049 300478 294934 106 100 69.32 -7.07 124.22 5.34 56 ALWGFFPVL h 75 19 11047 93640 103420 26 26 62.50 -15.08 111.62 4.54 57 AMAIHKQSQ 1 71 29 30323 303216 292846 63 63 75.03 -2.81 135.98 6.03 58 AMAKAAAAV m 56 19 30520 305200 305155 1001 100 74.69 6.87 145.09 4.28 59 AMFQDPQER 1 79 29 11987 116352 113958 74 74 73.65 -7.09 130.11 6.14 60 ATAKAAAAV 1 53 21 30597 305970 305970 401 100 71.39 -1.78 132.99 4.25 61 CLFGYPVYV h 76 22 10515 96508 101165 34 34 64.18 -11.26 111.35 5.76 62 CLGGLLTMV h 64 20 30010 300086 298947 560 100 71.97 0.10 137.04 5.44 63 CLTSTVQLV h 67 26 28545 285449 281565 19 19 63.37 -12.53 113.05 6.11 64 DLFGIWSKV 1 75 24 28001 279063 274377 162 100 80.56 -13.37 152.36 5.44 65 DLMGYIPLV m 72 22 11065 100066 104461 115 100 69.14 -0.17 122.60 5.02 66 DLMLSPDDI 1 71 26 12174 121486 121344 60 60 69.95 -2.95 123.57 5.52 67 DLVHFASPL 1 70 26 8052 80112 78354 25 25 72.64 -7.24 129.12 5.24 68 DPKVKQWPL 1 78 25 2842 25453 24537 49 49 71.08 10.73 125.22 5.31 69 ELIRVEGNL 1 72 27 27835 278317 275758 52 52 77.46 -7.35 140.60 6.27 70 ELVSEVSKV 1 68 27 28943 289425 289113 44 44 76.70 -5.86 135.05 5.41 71 EMFRELNEA m 80 29 29690 296038 290373 38 38 81.82 -6.63 144.16 5.68 72 EVAPPLLFV 1 69 20 3764 37101 37335 47 47 67.17 -1.38 115.77 5.70 O Table 7.1. Continued ________ N o Sequence A ff nAb NOc BBd Pep6 Comp/ Watg Conf*1 H2Q' Energy* HB1 O N O N N O Tf- C O r H o 0 0 C O N O C N C N O) cn n ; C N N O cn p~ © in N O 0 0 ID C O V D V D Ov oo Dp C N Ov C N cn 0 0 r -H cn in in in SO in vd in in in in in in in vd in in V D V D V D V D V D in in N O C N <N <N 0 0 0 0 0 0 o ov V O in cn r H r H C N cn o O N N O cn cn p ' vd 0 0 in 0 0 C N C N V O V D D p (N vq V O C N Dp p ' cn o p O 0 0 C O <n cn cn N O C N r H o cn r H vd cd r H in r— H r H r“ H T f r cn cn C N C N cn N O cn r -H dp cn r -H C N C N r H 4 “H r H ov o vd oo o Ov r H p^ o r H in r H r -H C N cn 1 — H o v VO VD OV Ov r H 0 0 O Dp VD r“ H CN in ( - — V p " 0 0 VO VO VO r - H Dp o o O CD VD v o o Dp OO N in s© cn CN r > CN CN CN p" VO * — 1 CD CN VO P ' VO o CN VO oo o CN Dp o ON r- O Dp ■^t o VD VD O VD T — 1 VD VD r - H CN VO 00 CD Ov i— 1 CN cn CN cn VO CD C ^ - 'H Dp 0 0 Dp o P ' CD O o o 0 0 l - H 0 0 t - H O o t'' N i i r - H ON o o 0 0 I > N o D p p' O 0 0 VO P ' o 0 0 p' VO p ' CN 1 - 1 1 - 1 CD CN 's i r - H CN CN CN CN CN CN r H CN CN CN CN CN CN <N CN CD V O VO VO P ' r - c n CD vd i o d 1 ■ i D P D p Ov C N D P vq D P cn CD P 4 d in P ' P ' p ' r - OV vo o VD VD \ m m S i > O V VO 157 r -H VD VD p '- o o p" cd C N VO o o < N p" r- ■ ■ i V O V O rn M h O n r - ~ C ~ ~ P~ P~ P ' CN VO CN Ov VO ~ o o q p~ ; d p 't v ri ri o l > O ' O ' VO o © O ' o o g v d V O o o O ' cd o- cd © o- o O ' O ' 1-H 0 0 0 4 O ' 0 0 0 4 0 4 D p D p i 0 4 D p Tt t-H O ' in p 4 O' o o cn d p 0 4 o V O Tt 00 o P ' r*o Ov C N ” 1 C N p^ VD C N v .N C N DP 1 N O i i ■ P'' O V 1 d C N i vd i 00 O O DP C N VD C N P " p ~ o 1 — 1 C N o P ' O V C N oci C N N O d p ^ d VD C N o o V O N O v o v o v o VO V O v o o V O o VO V O 00 o o o o oo cn v o ^ g o o o o v N © rr, o o Dp Ov D P ov vo o V O O V V D I — in o r H On in C N N O N O ov O O r H 00 N“ C N N O r H 00 C N P ' in m 00 o cn N- cn r H C N 00 P~ C D 00 ov O m O N p ' r H O P ' V O m N O so 00 N O ,< 3- O o cn cn C N V O O V vo r H SO cn cn vo C N O C D vo I ^ in H > cn so 00 cn cn H i > r - 00 V D C N r H Ov in r H 00 in Ov r H O O V O N o Os O N O N o in 00 o P" p ' 00 Ov Os r- N O 00 C N 1 C D C N N r H C N C N C N C N C N C N r H C N C N C N C N C N C N C N C N Tf N O VD ^H O N H > V O oo rH C D o o Ov V O O V P ' Ov ov C N p~ vo V O O N O N O V 00 N H > p ' V O oo C N ov p" Dp O oo V O C N CD O V vo C N VD VD o N O 00 C D cn N O N ID vo C D p" ov DP DP CD O P ' C N P" vo vo ov VD VD T — 1 vo O N rH rH o O N vu T-H O V o ov O V ov O P ' oo T —1 00 oo oo ov O V oo oo 00 C N r— H cn C N 1 —1 C N 1 C N C N C N 1 C N C N 1 —1 C N C N C N C N C N C N C N C N cn DP 00 On m C N o C N VD C D Dp VD C N C N V O V O o C N o 00 O r-H O o C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N 1 C N C N C N C N IN Ov in O rH VD Dp Dp C N V O VD C N C D VD O V VD C N Ov Ov V O r - P' SO 00 V O V O V O P ' P" P~ p ' P ' oo oo P " V O V O V O VD V O vo V O v o S S S ^ f i S ^ ^ g ,£3 ,£5 J 3 .c ! g » d ,£3 , d ,£3 — 1 ,£3 ,£3 .£3 u y o . . h -l . . . . fi| pH pH [X , (JL | Ph b Ut fa IX X > a > i- > S S ' I f e u ^ a o £ q d h -l S h s tL| ft, pL , O l ~ I h - l d H h O > £ § o § 2 S o o o h -l H d 8 o d H H M M H i o o o o h 4 H IX > 3 r n ^ i o v o t ' 0 0 0 \ O H O ) r o ^ , i o v o o o o o \ O H 0 4 m v t i o ' C P '- P 'P 'P 'P 'P 'P 'O O O O O O O O O O O O O O O O O O O O O V O V O V O v O v O v O v 104 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. v o o 0 0 O cn O N l> CO wo r- vq vo Ov CN co 0 0 o N O W O CO rH 0 0 CO q r-- Ov O N CO e' en 00 wo 00 o o CN q CN q in H" wn W n wo in v d iri W O wo wo W O wo wo wo wo in W O i / i wo wo N O wo wo i > o -sf C O r * * * - r- W O 00 O N N" cn ov V O CN c n C O C O 0 0 0 0 O N C O N - wo r H wo o N O 0 0 C O 0 0 O N m ov C O r H o O n o o o o wo O r H (N wn O N 0 0 r H K r H C N C N r H O V Ov 5 N O r H r H r H r H WN o o r-H o r H l> C N C O C N r H w n r-H Ov o r H 0 0 rH wn C O O N 00 wo C N rH wo r H M < 3 o C N K a o 0 s O u C l o ' CP P Q ffl o t 3 ( D 3 .B " 3 o O r- tu 3 e s H < L > I 3 al < u C/2 N- N; o q wo C N C N r-H N; C O 00 o r - o C N N" r H C O o r - c ^ r - in vo o o o N- cn vo o in c> q r - O N o o W N in r~ wo 1 wo 1 C O 1 V O ■ ■ wo 1 wo 1 C N ■ ov 1 1 ■ C N i vd i o 1 C O 1 ■ C N i OV ■ wo i r H vd (N ■ O i C N W n cn o r H O N V O 00 00 in o On C N O N OV cn vo r H cn C N C N N; V O 00 O t " C N O V r H ov N" O N C N cn O N o O V C N N; vo r - O o r-H C N C N vq 00 wo O N N O wo wo vd VO vd vo N O N O W O r H vo N O C N V O vo r H vo o V O cn vo Ov vo O N N O C N V O N O vo wS N O o r - 00 in r H r - cn vo 1 0 0 w o 4 4 C N n- r H 00 1 0 0 3 9 vo vo W O 00 o o r H 3 3 7 8 1 0 0 3 7 1 0 0 3 5 0 9 1 0 0 6 2 5 2 o o r H r H 4 7 o o r H m cn m H- H- C N H - 00 3£avvomJCjcnoo£! ^2envooo£2cnC'-^ N- cn C N in O cn vo C N C N C N C N V O 1 /1 vo C N 00 ov _ _ C N vo vo m t^ r - cn O V V O cn N- O V 00 W ^ 00 o r H ov o V O cn t"- N" T — 1 vo cn 1"- n- ■ cf vo C N ov Ov O N Tt C N W"> r - vo C" m o vo vo C N in m m in cn C N ov Ov cn t-' N O N O 00 o V O in ov vo vo C" in 00 ov r H 00 oo V O r H ■ v f r N- C N (N N O cn cn o r - r - v o vo vo oo V O V O i> vo vo o 00 r H 00 N O r H r H N O C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N i-H < N C N l-H i-H C N 00 V O vo C N 00 o in o en r - N" C N o cn C N O n C N O V o cn 00 r H C N o cn 00 o oo r H cn Ov cn r H cn o m cn V O O C N r H r H O V "'t oo r - r - cn o C N N" cn <N in vo cn in N- ov oo cn cn in in r H o oo t - ' o Ov oo O oo 00 oo oo cn t— cn C N in ov Ov in oo oo r - vo t - '- Ov V O vo t - - l > vo V O vo 00 00 r —o 00 vo o O t - - C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N r H C N C N r H C N cn cn C N oo 00 cn cn C N t-- VO ov 'O Ov C N 00 in o r - V D wn o o cn — H vo vo in cn r H C N O n- r- vo C N o vo rH C N C N O m in cn C N C N C N C N cn C N C N r H r H i r H C N vo cn n- n- 00 00 oo OO 00 ov OO 00 oo 00 00 oo 00 ov 00 00 t -H 00 r" r H r-H 00 C N C N C N C N C N C N C N C N C N C N <N C N C N C N C N C N 1 — 1 C N C N r H C N O C N O C N O 00 o O V o r H O OO r H r H r H o O o o o C N C N C N C N C N C N r H C N r H C N <N C N 1 —1 C N C N C N C N C N C N C N C N C N C N vo V O o o C N vo O o in O n 00 oo VO V O cn 1 — H r H r - wn wn O V V O vo vo C" r - V O vo e' r - V O v o vo vo V O vo r - C ~ - V O vo V O V O vo o N " VO ■ 't r m in o 00 ov cn cn 't N " O v OV cn cn oo C N o o o o ■ 't N " W o o o h-l h-l h-l h-l H h-l p p h-l H P p W p H O U b h -i h -i h -i 3 « s ! £ £ £ Pp P P o o o h-l h-l - < 3 P 4 H H Pp P p > > P p P p o o h-l h-l h-l hJ > Ph Ph > > Pp P P 0 0 h-l h-l h P h-i h h r- ^ pp > > ; > pp M ' 0 0 h-l h-l h-l > H O ^ pq 9 £ s? > pp o £ > H h-l P — 5 H M H W hI ^ ffi K a a pp > pp i i p m i r i i i p i i i i i p i p i i i r " 1 i p i i i i— h p i i i i 1 h p i i i i— — i o o o o o o o o o o o o o o o o o o o o o o o o O H O i c n t i o o h o o o v O f - O O O v O O O O O O O O O © ' — ' O N O V O N ^ H r H r H r -H r -H r H r -H r H r -H r H r H 105 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 7.1. Continued O N r H o in C O < N 0 0 0 0 in 0 0 O O S o r -H 0 0 r- in C N O N 0 0 in cn N O T — H 0 0 <N 'f 'f N O O r H C O < N p N O P o N O O N in in in in in in in N O in in iri N" in in N O in in in in in in mi Tf o < N o r f in 0 0 c^ 0 0 C O O N Tt; C O SO O S C N O N in N O in N O o o r H N O C O in o O N t " - O N T 3_ C N p N O P t > in <N m o T J- r H C O <N (N r H in C O r H d I H in (N C O T — H N O cn r H (N in O N O r -H On C N N O o O N C N < N r H n i in r H N O O N O r H o r H r H 0 0 C N cn r H 0 0 d C O M < 3 C 3 W o cn K a o U £ o U a < u Ph « ffl O in r H m ^ f O N in o C N cn cn O N C N in ■f C N cn N O O O N C N OO o o 0 0 o o p C N cn O N in ' f l > O in O N P 0 0 l > N" N O N O N O N O cN r~ C N N O N O o o N O OO N O 0 0 cN 0 0 © cn N O O N N O cn N O o o o N- 0 0 cn N O H N O OO N O © r- l > N O cn e - C D O 8 Si (D C / D O £ 00 00 On ■ h f 00 cn o cn O N m l> On N O O C N C N i 00 1 ■ cn r i C N N O i C N O i oo r H ' O v i 7 ' t m O ' o a \ H h ' j 0 ^ m o . o \ o q a ; ( S f D i n ' 0 0 5 ? i o ^ i n ^ o o H o d ^ p i m H - i o ■ i ■ ■ • ■ ■ i < ■ v • o o o o o o o o o o o o 00 C O 00 00 r H o 00 C N C N O o «n r H 00 r-* C N r H m in r H m V O W r H 'f o o o o o 00 o 00 O N ON o o "f o o o o cn ' f ■ < f 00 cn - — i o o o o o o O O N h C N m N O t - h T f C N O 'f r- n o o n- t-- n o N O cn 00 00 cn O N 00 Cn r - r H " f N O O N in e ' T f o C N r—, o t" " N O t " - C N O N in r H r H O N O r - OO r- C N in en N O C N o C N C N cn o On o V O in cn 00 00 o N O r - o O N O N " f r H cn i in N O r-— cn H i > 00 m n - o m T l- cn in in C N o 00 C N l > r - N O cn ' f O N 00 O N O N o o C N On C N Cn C N N O N O On T — 1 O N 00 r H rH 00 t -H C N C N C N C N C N C N C N C N C N C N C N C N C N 1 —1 C N C N i-H rH C N I t — h 1 o N O IN cn C N O N cn On O N On O O C N in C "- r H ,— i ,— i <n O S 76292 m C N cn O N i o in r H O N * —i C N O cn T — H r H o N O m m <n cn C N in in N O N O 00 N O cn cn N O cn cn N O O N n- o On N O in O N 00 00 cn N O N O IN - N O N O m N O m N O 00 N O m 00 t - H " f so O o On O N On O N 00 O N On O N O N O N N O N O O N o O N O N r H r H 00 C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N r H C N C N r H r H C N r H i vo o C N r H r H C N o cn cn r H o m cn C N O 00 N" ' f H * C "- 00 m so so SO On N O N" cn N O cn r~ r H N" 00 m O cn cn N O N cn t"- r-— N O l > 00 cn vo so 1 " - N O N O m N O in N O cn t > cn On C N On O N N O r H t - H O S Ov os os oo On O N On On O N NO 0 0 os r H On On T —H r H OO C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N r H C N C N 1 — 1 r H C N L 1 — 1 ’ * cn OO 00 o oo OO OO 00 O N O N O o N- cn cn 00 o O O N O m t"- C N C N r -H 11 — 1 r H 1 —1 1 —1 C N C N C N C N C N C N C N C N C N C N C N C N cn r - m IN o cn cn m 00 O 00 O S in cn C N m in C N so in m m in in m m in m IN vo in N O N O r~ - N O N O m N O r^- l > g - H ^ I g g . d . P g g o o S | o o o S o o o o o o o o o ^ ^ ^ o o o o > o o o > o o o o o o C L . 1 - L j C L i h - l H-l J o o o M h P h Ph P h o o o h - l h J h - l o o o h - l h - l o o o o H(Nf0^ln«hOOO\OHMnNtVl'OM)OONOH(S _ _ i _ j _ i __d <_d _ i _ d — j R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 1 4 3 GMNKRPILT m 7 2 2 6 12160 121423 107687 1 6 0 1 0 0 77.20 -4.17 143.87 5.90 1 4 4 GMNRHPILT m 7 3 2 9 12106 120409 116246 2 7 2 7 78.63 -7.96 140.81 5.81 Table 7.1. Continued r H v o o in C N o C N 00 v o o in N 00 O r H IN N in Ov 00 cn cn (N vq C N vq C N in n r H C O r t in <N vq r H 00 r H in t> oo C N in in C N so IT) vd in in in in in »n in in in in in in in vd in in in in id in in N- in in in C N N- m vo vo N- o o 00 vo 00 in cn cn r H o i t Ov r - in On 1 — H * in cn cn vq C N in n r H n - O cn cn o in o o V O c-- in cn Os v d N- 0 0 in in IN 0 0 C N v d H o Os in N O s C N N" vd i > C N oo C N o r H C N N- r H N " cn vo m cn cn cn N" cn C N C N C N n- cn i t C N C N cn r H r H r H T “ H r H r H r H r H r H r H r_ H O f l J-H < D d m o < N N d o U c i £ o U d. 5T Ph ■ o ffl C Q O 2: O s § 1 < D G O o 2: o v o cn c - - V O C O v o in in cn in n n - C N C N V D o vq O v oo cn 0 0 0 0 oo cn C N cn in O S in o 0 0 IN C N C O r H O v in in C N O O v C N O V r H cn r H in r H cn in i i vd ■ cd i i r H cn i d i vd i in d i r H 1 cn i id vd i 00 1 d i i t vd cn i C N O S cn v o in O V C N t - t - o r - r- O S in q n- in r H N ^r os m in O v V O r H 00 cn i t i t O v n oo 00 in oo O v C N cn T ^- cn C N V O o in cn 0 0 K V O o 0 0 vd r - Os 0 0 r H 0 0 Tt n IN in in in in IN cd 0 0 d in 0 0 v o cd C -- ^ : in r H t > vd t " C N r-- O S v o it 100 48 34 29 100 73 100 100 C N in C N v o 100 22 48 21 o o r H 35 o o o o r H o o r H r H O V 30 o o r H co Jg oo i t ov ^ vo m oo it t- it C N C N in vo ( N C N 00 r — I C N i t C N in cn 0 in V O C O C O 0 0 > — I o i t C N co I-H O O V C O 00 in r - /• r . •n- vo r - C N V O i t I t cn cn C N t -H cn m r H 00 cn O V 00 I t 00 r - I t n - r H ( w > in in Os r - C N I t o itv C" C N cn O r-h t - cn C N m O s Ov r- Ov o Ov vo m i t Ov V D cn vo O vo C N i t o C N in ^ r r - r H o O cn C N 00 vo o I T ) i t i— < oo C N i t oo 00 r — V O o vo V O r - in r H C N i n V O C N O vo o r - e- Ov oo t"- C" ov o r - V O o r H ov O V o Ov O 00 o r H C N C N C N C N C N C N C N C N i C N > — i C N C N r H r H C N C N l— l C N r H C N r H o c n 79995 r H I t l> cn O V r H V O 82679 r H ,— i o I t r- I t V O o O oo i n cn C N Ov r -H o vo o o t-- cn vo vo C N m V O cn i n o ov '■ 't i t r H i n i t *—h c n c n vo o t - " o vo o o t ~ ' i t i n o Ov O V i n C N ov ov o E - ~ 0 0 0 0 0 0 m i > cn i t vo m I t cn oo c - ~ o 0 0 \o C N t " 0 0 t - o ov t'' t " - ov 00 o 00 00 o r H Ov ov o Ov o 00 o C N C N C N cn C N C N C N C N C N r H C N C N 1 h— 1 (N C N r -H C N r H <N r H CN t " i n o c n m i n cn n - CN as CN 0 0 O V o c n i t Ov en v o r- o v u m 0 0 i n c n c n c n Os W * ( H N o r H CN o o v o v m o v o v f'' r- 0 0 m N /* r. o O S o N 0 0 0 0 0 0 c n 0 0 i n n- m o m C-- cn T — 1 o m 0 0 t - CN 0 0 0 0 Os o o O s IN N 0 0 r H 0 0 0 0 r H * — i O V Ov r— 1 O V o 0 0 o i-H CN CN CN c n CN CN CN CN CN CN CN 1 CN CN h— 1 CN 1 — 1 CN r H 0 0 CN O N " r H i n 0 0 N Os IN 0 0 o c n CN N " i t O CN o I t CN i n CN CN <N CN c n <N CN CN CN CN CN CN c n c n c n CN CN CN CN CN CN CN CN CN i t O V r H O N " o r H c n r H O s n - IN o o r H V O i n 0 0 cn r H r H O S v o so r- r- IN r- IN N v o S O N IN IN 0 0 i > v o r - ' m v o V O e ' o so w £ O * & 5 > P f* x ® a co w w >H >H H H W W J o o H H H H H H H H H H h-L| hH HH H H H H £ X O O O O O K E f f i invoiN-ooovO^cNco^i-invor-'OoovO'— icNco^twovoooo i f i t i t i t ' ^ ' i n i n i n i n i n i n i n i n i n i n v o v o v o v o v o v o i o v o i o 107 R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. Table 7.1. Continued H ffl M < o W o CN s o U ci £ B o U < D P h "O ffl ffl o <H o i n c o ON NO CN ON i n o CN o o CN *n r - cn c n no ON TT NO r - NO C h CO r - o ON Os O 0 0 m t -H 0 0 o H t q O in m CO c o CO NO i n ON N ; i n m ’ 1 H O in i n in i n in ' un I/S NO N T ) i n i n NO i n NO NO ir i i n NO ic i vS i n o o NO ON ON r - as 0 0 c n 00 t — h CN q CN 00 00 CN o t > NO i n NO i—< c n O CO ON q CO r - r— H c n O) c n as t^- ° P ON 0 0 o q H q r-H c n T -H NO CN t -H H t r-H t — H c n c n ?-H t — H t -H N 6 c n r — H c n c n r-H c n Os CN r-n CO H oo CN ON CN T -H c o n t r — i t — H CN i n r— H o CN c n CN ON i-H i— H NO CO 1 — H CN r-H c n O-s NO r - ON T t 0 0 CN 0 0 NO NO c n 00 r - on ON CN H t t“H NO NO ON o c n ON NO o m m ON t -H c n ON c n r - q 00 0 0 i n N - CN • CN 1 1 00 00* 1 CN CN i ON I C'-' 1 c n 1 ■ t — H 1 CN i H t CN i CN i i CN CN i o 1 o r ■ t • c n c n c o r-H c n c n NO i n t > CN 00 c n in c n ^H o NO CN NO Os t — H o o m q oo CO - NO NO ON VO r“ H CN i n r ~ o CN c n 00 c n t-* o o NO CN VO r-h r - c n NO r - c n r - © CN r - o o NO t — H r - ON IN- i n NO NO NO 00* C"- oo' NO NO NO NO NO c n NO c n NO NO i n NO O r - 3 0 ON t -H 00 o o t -H 4 0 t — H 5 0 1 0 0 3 2 4 3 o o r-H 1 0 0 O o t — H 4 2 1 0 0 1 0 0 4 0 c n t -h 3 6 1 4 2 3 o CN CN i n 1 0 0 < u o § s al < D o n o £ rH rH ro C\ oo C N O N o w N rr oo oo o N h 1 ^ O C N c n m CN o N " cn N O rH C O CO t~- CN CN (N i n ON CN i n T — * 4 < -* N o o o NO c n 0 0 NO N - CN ON 0 0 ON NO o ^ r c n <N r— H ■ H f m i n c n T — I 'w ' NO o c n 9 5 7 5 1 NO T — 1 c n c n NO Os ON o o i n m i n c n N - ON ON CS >n y— i i n CN CN c n i n ON OO r - 0 0 NO NO c s C n IN- c n NO CN o ON o I ^ ON ON 0 0 CN ''3' o CN 1 — M V OO Hfr O 0 0 0 0 o 0 0 < s o 0 0 o v-H o i-H O o o r - i n r - NO CN i CN 1 — t N r H CN 1-^ i-H * — I i-H 1 — H i-H i-H <N <N CN (N CN CN 0 0 CN r - CN ON CN i n v 4 r NO NO O 9 3 6 6 8 i n ( — N ON O O C"- o NO o r-- 0 0 c n c n CN 0 0 Os NO i n c n o c n o Os c n c n N - ON i - H NO o c n O i n m 1-^ 0 0 r * ^ o (N i n i n r-* CN r - i n 0 0 NO r - c n r - i n r - Tt 00 o 0 0 c n 0 0 Os 0 0 m 0 0 CN o i - H NO 1— H o m m c n i > o 0 0 o \l v 4 r c n o 00 ON o O o o o O o o 0 0 0 0 r- 00 CN l — H CN N t - H l- H CN r - H 1 l - H r - H l—H r-H CN CN CN CN CN CN t -H ON o —4 « CN CN O c n l o 0 0 o l o r ' m c n c n e n CN Hi- CN CN CN m ON C n NO m ^ r s r - CN i n r H CN r H Cs o t -H ON o c n t -H NO C s Os r ^ NO C " i n CN \ n t-H CN c n ^ H CN CN c n CN o c n c n ON C" C s o T — H t -H i n 0 0 o o o r — H >L/ c n t -H 00 o O t -H r H t — H t -H t -H t -H O o r - 0 0 0 0 0 0 OO 0 0 CN 1 CN l-H M l— H CN T — H T — 1 t— H r H t -H t — H r-H r — H CN CN CN CN CN CN r - i n 00 CN o c n O m NO H t ■'t NO NO i n NO H t t -H r-H CN r-H CN CN CN r-H CN CN CN CN CN CN CN CN CN CN (N CN CN CN CN CN CN CN CN CN CN r t NO i n c n c n i n ON m ■^r CN c n l > O c n CN 0 0 NO i f o o o Tj- c n * > NO r-~ r- i"- NO NO NO r- r~ NO l > o r- v o e' t~- I > C" r - r - r - ^ s - s s s - s s a r £ 3 r £ 3 r i = l r £ 3 r £ 3 r £ = l I s ^ w > P h C/3 a 9 m H P h tH S 'S S 3 l-H I > I — 1 O K 3 > ^ fe O H ^ h P i — 1 ^ t- 1 P h a > C O W H H i J J ! ! ! ! ! » ! ^ H is P h W 3 P h P h P h P h P h P h P h H 0 < ! Q P hS c / 3 ^ > hO W h-H h H HH l-H I — t I — I I — I f — I 5 H H P 3 E I I 1 /3 p H p H W o o o o ■ hJ J J ^ O H N t n H t i n i o h o o a i O H N c n r t i n i o ^ o o o i O H N Noc~-r-'r^c~-t^-r^i>r'i>i>ooooooc3oooooc3oooooooONONON ^m— 4 « — ^ vwd ^ vmm4 Pttd d ^ d 1B^ H 1 0 8 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 7.1. Continued V O cn 00 e' 00 00 r - cn O cn On C N 00 O N O n o o o o t - H 00 cn e n w n w n 00 o C N t - h q q (N t > q q o n t - H 00 N; C N q W O on v d Tt w n w n w n v d v d w n w n w n v d w n w n w n w n v d w n w n w n OO r- o n O s w n C N r- w n ON VO 00 cn cn o o N" "n - cn VO T-h cn V O t - H l > r- w n cn r - o O ) o cn 00 r - o w n c n 00 C N 00 C N c K vd w n t> 00 t - H cn l> t - H N" o On v d N1 cn 00 v d v d cn C N in r H C N o cn cn Tt N" cn (N cn T-H (N t - H w n N" (N cn ^ H C N 1 — H t- H t- H r -H 1 -H r -H T - H t — H t - H '"H r -H T-H t- H r H '"H t— H t — H t — H >> Ofl ( H < u O © a P ) o O £ O u Ph ffl ffl o £ t 5h < © © © © © cn © © t- H V I © O S o o V I ON © (S S © o o vi © t — H O S i/1 c n © O s c n On c n c n © © s © 00 © © © © © t-H © © r H © t - h 00 a © © © © © © © c n © a © ON © © © ON © s © V I © © © © © © 00 ON © © © © © © © 00 v o u 0 § s 01 < u oo o a .— . ^+. 00 00 ® S S I S 3 5 o m h h 0 1 vo vo vo C N r - q C N 00 t —H vo N- o 00 o S ' © q C N O N C N t - H cn o wn c n o q © © wn i N- i wn i C N t - H i nr 1 t — H 1 00 1 i r— 1 1 © 1 c n ■ 0 1 O N © o o o o o o © 00 S ' O N r - n n © S ' S ' © r-H CS 0 0 © r— H r-H r— H © © © © g £ £ H - © 0 0 © © t — r — t — - s- o n cn cs o © cn © © © © S ' cs © © cn s - © © © cs 00 O N 0 1 © cn s - © cs 00 O s oo O S w n © t — H cn V O cn cn cn 00 r— . cn m © S ' c n s © oo © © © t — H vo o 00 © O N C N 00 t -H (S t -H c n © O s m cs s in O s r- o w n t -H S ' 00 cn V O w n © © © t > © © 00 rH cs r-H s o 00 cn cn © w n O n cn © cs O s s cs O n © O N © O n S ' w n 00 o O S © O N O N l> r- V O C S © r-H O S © 00 C S m O n © © cs C N C N cn cs cs C N C N C N C N r— 1 t — H r-H C S t — H cs C S C S C S C S r-H o 0 0 S' © © © 0 0 © t —H © 0 0 © OS t - H r- O w n w n 0 0 c n © © O N © t - H t- h c n 0 0 C S c n © cn 0 0 C N t - h w n w n C N 0 0 c n © s 0 0 t-" o r- c n S' © c n © © O N o C N OO t - h C N cn C N O N cs cs O N C -- C S cn © O n © © 0 0 © t - H v o ^ H w n w n r H © 0 0 0 0 © C " - O N o O n © O N O N O N O N K - C N o C N O N o O N w n C " - o 0 0 © C S cs cs cn C S C S C S C S C S C S i n * t— H t - H C N r -H C N C N cs c n cs r -H oo t" cs t — H vo C N en cn © © 00 vo © s © © O N © © s - © r - O N © r - N- r - © cn © 00 o rH © On © © o © O S s © o c n cs cn wn o © wn 00 © U s J i r - C S © © © ^H t -H © 00 O N © 00 O s o O N 00 o s O v O S O N V ) I n . <N o cs O s © O S rn © © 00 © cs cs cs cn C N C N C S C N C S C S rH t — H 1 C S 1 cs cs © © © t - h t -H cs S ' o t — H O S O N © r - wn © © s © © O N © s - 00 C N cs C S cs C N C N cs rH t — H C S cs cn C N cs C S © cs cs © © © t — H C N c n s 00 N- © w n oo O s o 00 C N c n O N © o s © © On © s cn i-~ r - t-- w n e' wn © © 00 r - o © e - oo 00 © © © V O g X P P g g g - H j 3 - H g r - , ^ 3 g ^ ^ h S 9 cyQ^f^ o o a h - i > 0 > E K o o g u o g i f e r v r v ^ u o o o s s d Ph Ph Ph Ph O O a EldSid O h O' PP a O h H Slp!^ a a a j y c n © © © 00 O n © ,-H © © s © © © o o o © rH CN c n N - C\ ON ON ON ON O n O © o © © © © o © © © r-H t -H H t -H t — H T -H t — H r-H t — H r-H r-H H © © © © © © © © © © © CN CN CN CN R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 1 5 LLDVPTAAV m 6 2 2 2 10868 101315 108674 5 7 5 7 66.89 -3.44 123.00 5.42 2 1 6 LLFGYPVYV h 7 6 2 2 10489 95595 100301 2 7 2 7 69.41 -9.52 121.89 5.33 < D | " H o U t- — 2 R H >N 0 0 * H < u p3 Pi o U R H r H r H q C N N; O C N o q 0 0 o m C N cn q i> N; O to O N to q to N; N O to q c n to C N cn r- in r H CO 00 to O N to Rf o in T - H C N un to to to to to in to to to to to to to to to to N O to to in in in l/V C N cn C N C N r * - q Cn O r H o o to r t to C O to C N O N C N c o l > C O o un r H N; N O C N q C N to 0 0 O N O C N cn C N r-- N O un Un c n to r H cn to o r T to cn r H r H OO cn N - C N C N C N CN O CO C N ^ H to CO C N r H NO r H r H r H C N CO r— H N O to r H OO C N in N O o r— H &1 Ph ■ a m ffl < u o g §1 C D c / 3 C N ^ r to o o o C O C N O N co o ^H o r— o ^H C O v q C N 0 0 N O C N o N O q q C N ■R 0 0 C N O N *H q C N C O OO C N cn C N rH 0 0 q C N i rH 1 T t 1 N O i ■ C N rH 1 C N C N i © N O rH i T t 1 1 C N ^H r — H o i o i ■ in i C N ■ r— H 1 0 0 i rH C O N O to O N C N On O N l > o un 00 o oo r N- rH R o N O to r - C N 00 R; N; C N N O r N O R l > N O C O un o N O i > H C O On in rH 00 C N C O rH cn ■ ' f r C N N O C N oo in C N ov rH N* C N ON un 00 R" cn On C N O V 00 00 oo r - N O NO N O r - t " NO N O r-* - v o 00 V O N O N O v o vo NO oo N O vo un 1 0 0 1 0 0 o o r H 98 1 0 0 91 93 42 94 1 0 0 1 0 0 rH C O " ^ H C O 1 0 0 O V O On N O o m 14 vo r— H N O N O o o rH O N r— H C N vo cn cn l> O N t'- o in 00 o\ > — i c n cn O n O n R c n 0 0 C N c n cn C N N O c n O O n V O no O N- no N O W H H NO q On CN cn c n no cn N- r H 00 r H C N O ,-— s r H O N O N 00 O 00 m ( — s N O 00 H i On o On C N N O v o m C N r - C N O N /— N O N O On 00 —H C N N — > C N t- h O N C N t— i C N 00 o C N C N o N O C N u i v . N O C N N O 1-H R O N w in 00 O N C N ■ — M O N oo OO o O r H C N r H N O o o O N cn r** On O N m O N O N O r R r - - O N O N o O N cn O 00 ov N O On m oo On On o N O oo O N i O N r ~ ~ i O' O N C - 00 t - H in r H C N C N C N C N C N C N C N C N C N —1 C N C N C N C N C N C N C N — 1 C N r H o un r— H O N N O cn cn O r - N O —H r — H N O t — H 95632 95383 r* ^ un 00 vo C N cn r H U n ^ r o O cn r - o 00 00 R oo V O o N O 00 9388' ov cn un ov r - ov cn e' O N R C N C N cn 00 N O C N t-> C N in r - cn un r H cn cn cn en cn r- cn R R N O 00 C N r H o N O cn o 00 cn o o r H ov o O N o t-' o O O r- o O r- r - O oo o 00 00 C N N O r H cs cn C N C N C N C N C N — i C N C N r H C N C N C N C N cn C N C N r— H C N i H ^ H un o cn rH C N C N N O o m o in o C N O cn N O cn N O cn o oo in cn cn ^ H cn un cn cn On O r H cn O N o O cn r- O N- O cn C N On o cn ^ r cn cn tc N" On On 00 C N C N N O m in en R- r H o m r H cn N" ov o ov Ov 00 O N O N O On O o C -~ O O o o oo O o 00 oo C N N O t - H C N cn C N C N C N C N C N 1 C N C N r— H C N C N C N 1 — 1 1 C N 1 cn C N C N r— H C N 1 — 1 O ov r - 00 O ■ R - N O On C" o 00 in C N C N ■ R " O in C N o 00 O o C N C N C N C N r H C N C N C N C N C N C N C N C N C N C N C N cn C N C N C N m C N C N C N C N cn un C N in H - - R - O N cn C N n - H - ^ H 00 00 r - cn o r- m r- cn n - O r- r- r- r- N O N O N O 00 N O O 00 N O r- r - r- r- N O N O r- r- r- r- £ rC 4 3 4 3 a a — a a a 43 43 — 43 43 - - r a a 43 > S > w P3 P3 W Ph P3 Ph Ph C/3 on C/3 C/3 O 0 O hJ h-l h-l h-l h-l hJ h-l h-l C O C O > g | s g 5 H 5 " h-l h-l h-l h-l h-l h H h! h y W ^ H - l Ph h-l p., a > > hJ < & Q h-l h-l h-l h-l h-l s s s oo O N O C N cn N- in N O C -~ O O O n O r— H C N cn N- t/v v o r - 00 OV O T-H t - H t- H C N C N C N C N C N cN C N C N C N C N cn cn cn cn cn cn c n c n c n c n C N C N C N C N C N C N C N C N cN C N C N C N C N C N C N C N C N C N C N C N (N C N C N C N R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 00 is c W O cs W a o U £ O U P . P L , ■ a ffl P Q O £ O h cn 00 cn cs © © ov s^ o i n © t -H o 00 00 00 in vo C N C N cn i n o t -H m ov cn cn vq O S t -h o cn vq in vo cn t -H 00 vd vd s o © in in vd in in cn cn cn cn cn in in in d in in in n in i n O v O S O s O O s O S cs 00 N“ N" (N vo N- 00 © (N N-vo o 'sC cn tT < N cn o in 00 cn O V in vq cn vq 00 o C N 00 vo N- Os in i n c n t -H C S © © co O S cs T - h c n N" r^ vd c n T - H c n C N t -H o cn r— H in t -H C N ov o n C N T - h in C N n C N t — H t — H in ■ ^ r oo N- t — H cn N" C N vd (N d cn l> Tf T — H cs V O O S 00 o © o cn IT) vq t -H c n O V s- n 00 C N O N o o vo o in N" in © S ' C S oo O V i n t — H ^H in cn t -H © S' 1 - h i-i i C O 1 s^ ■ 00 1 ■ ov i d O s ■ vd i n i i cn i d i cn i d i S ■ vd i in i ^f- 1 C N t -H 1 t -h 1 S1 i r- co co O S t> m i-i o o C S >n o (N c n s © N" i n 00 vq C N < N N" O v t -H N; cn ov o in cn S ' 1 - 1 C N C N O O N" o N- 00 © C S s - C N 00 C O o 00 s~ O N vo 00 l> cn r - 00 vo cs s~ vo C N V O 00 vo r - 00 vo cn 00 O v O 00 S ' 00 l> vo T-H r- O V r - C N © 00 SS < D a s o U s~ a > -O 0 3 H < u o s §1 < u 0 0 CS o o o c s © _ , © C O ^ h c s ^ c s o o os o o so c s O s s - m o in h oo m o o o o ( ^ (^ 'o ^ i o i o r H o o c n ^ 0 0 oo v o so so © © © 00 o o V) O S 00 2 § © t-i © 2 2 oo © c s c s 2 «o O V 00 2 ; so cs in © S - cs cs (S O S t -h 00 m vo in C N cn m t -H t -h 00 m 00 S - © cs t ^ H © © >n O S co O S C N t -h 00 t — H cn V O Ov in vo C N t — H S O © S- so © so Os S ' S - S O S ' 00 in un m N" in in r - C N vo in r - t — H © s - C T l 00 CS S O C O Os co Os C N C N ^ r in r> in t -h Ov o O O N © r - © S 00 s O s so cs oo os t -h Ov r - t -H C N t -H 00 in ^H oo Ov ov s Os On cs CS cs CS cs t - H C N C N C N t - H t -h C N C N ^ H C N C N C N cs CS CS 0 0 cs © wH o so cs S ' CS CS O V C N cn o o o o © IS OO On © 78358 G C o © © t -H © On S ' © On r- r— H cn r- i-i CO S ' O n r- IS © On o o © Os r- © CO o CS CS V O v o in cn V O © CS o o On O s S O © cn 0 0 © Os N 0 0 0 0 os S ' o o On O n 0 0 v o C N cn t -H 0 0 CS cn so © O s S O O n © S ' t — H Os OS IS cs O n O n r— H o v 0 0 0 0 t -H C N i-i 0 0 so i-i 0 0 Os O n O n OS o cs cs CS CS CS t -H C N C N C N t -H r H i-i CS CS 1 —H <S CS CS CS cs cn in C N cs 00 S ' S ' oo t—H 00 CS S ' S ' t CS s- cs rv T oo t"- C N cn IS r - CS S ' so On o © eN O V C N v o o cs O n 1 — 1 c s fv o 0 0 v o O o O n © r H O n O n o © co S- On (N m cn m cs © O n ( ~ y * N v o t - h o r H O n C" cn O n Os cs On 0 0 0 0 t - h cn t - H 0 0 0 0 1-1 O S O n On h v o v Ov o cn t - H CS cs t-H CS cs t- h CS cs cs r H ^ H C N C N 1 cs CS cs I ^ C N C N cn N" S ' © r - © 0 0 S CS © S ' © 0 0 v o 0 0 N’ cn © s © © 0 0 0 0 C N , C N cs CS cs cs CS CS cs cs cs cs t—H C N C N C N cn CS cs cs CS C N C N <N cn wn © © CS S ' © © S O © © S ' v o cn C N ( N cn On 0 0 © CS o O cn r - 0 0 so s- s s S ' so s S ' s- l> V O v o 0 0 © s - S ' S ' r - 0 0 V O 3 < l 5 o d B * * s A S ^ J3 ^ t— i b -* c — I c /0 O C / m > rv w & h ft, 0 0 0 0 C X m H a a a P i P h PM S o S F g ^ g ' w £ a > £ q r a a a a a a c x S 2 C N cn m v o 00 O V o T-h C N cn N- in v o r - 00 o v o t— H cs © S ' N" N- N* in in in in in in in in in in V O © © © so C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N cs cs CS cs i n R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission. 0 0 T f so ro r — i in 0 0 ^ h t- H m l> o O H 30 cn C N r - 0 0 y — ( so T -H o in in in ro os in C N O S 30 so 1 — H 30 oo in N" 03 in »ri in in SO in in so in *n so s o in in 30 in in in in in in in in in in in so C O so ro o m N- o o it ( NN - O t - 03 in 00 cn r- os in o T f < N o C N in o so 00 q ro i t os 03 03 1- 30 os C N C N in 0 0 SO SO © ro Os r - H r - H C N O S 30 ro r - i> 30 30 o o in C O in ro C O SO ro ro ro ro T f (N C N ^T C N m i- h C N in m C N C N 1 — H r - H r - H r—H r - H 1 — 1 i- H 1 — H ^ H 1 —H i “ H ">» £? d > P W o C N w P o U S o U p . 5 T Ph p q p q O £ % - > P h T 3 < L > J ts o u t - — 3 C 3 H C D I P cr < u C / 3 N " © os so ^ H 3 0 0 3 T - H c- rH N - o q N - 0 0 o © ro 0 0 C - C N o l> C N in N - o 0 3 3 0 N - 0 0 0 0 q i — H m 3 0 r- r- 0 0 so 0 0 0 3 3 0 r- 3 0 d i C N i - H 1 0 0 1 in i os 1 ro i ^ H r - H 1 ro i d i ro i ro i in i o i 0 0 1 i 3 0 I 3 0 r - H 1 in i i r - H 1 in i C N r - H 1 o so S O C " - (N ro H ro 0 0 0 0 o so o 0 3 N " o o o 0 3 0 0 C N C N in so so i - H q C N r - H 0 3 0 0 0 0 q o 0 0 ^H m t- O so 3 0 l> O ) q N - O Oos in N - 1 — H in so cn ro O S C N d 0 0 < N so 0 3 O S d C N N " 00 rn 0 3 3 0 00 c- oo 00 r- o 0 0 r- r> r- o o l> 0 0 so 3 0 3 0 so 0 0 l> 3 0 so o T f os in 51 48 o o 1 — H O O 100 100 0 0 83 75 44 100 100 68 50 09 100 94 100 100 ^ H m 100 100 0 3 T — I 0 0 in V i ■ < 3 - 03 C O 00 O o o m 'zt h H h H h H o o cn in 00 r- It -t o C- o o 03 0 0 o 30 30 c n o o 30 H " H co cn h H < -) r o (N 03 C N 30 y— i (—1 1 — ( o r-H O 30 m in 3 7 9 3 2 30 C N r - S O N- C N in o O N" C N r r 30 00 in T — H 1 — H o O O 03 r^ in r-H t> O 00 cn O S o C O l> 0 3 in 30 H r^ 00 03 m m r - r - ~ I - * * . 03 m cn S O N - 1 -H C N O O H- 0 3 O o N" C N C N 0 3 30 03 o s O 30 o o N- s o oo in oo o in 03 m 03 03 O S l> oo 0 3 00 03 o i — H C ^ O S O O S O S C N 0 3 o o 03 30 o o C N C N C N C N C N C N C N » -H r - H C N C N ^ H C N C N r— H C N r — H C N C N C N e ' in m 0 3 o O / — 3 O ro C N os O 0 0 r- o t * Hi — H i— H cn in ,— H 0 3 C N en m 3 0 3 0 0 0 in 'w ' C N r - H 0 0 r- so i - H o cn r - H m m rr o 3 0 0 3 0 3 m cn ■ * 1 - 0 3 3 0 in C N g m m so r - H in n- C N 0 0 i - H OS o o m r - H n- 0 0 0 0 r- 0 0 C N o cn . ( — 3 r— H 0 0 so 0 3 0 0 (N m 0 3 Tt in 0 3 m o 0 0 0 3 i > 0 3 0 0 o 0 3 0 3 o i - H r- 0 3 o 0 3 o C N 03 o o 03 i> i-H o C N C N C N C N cn C N C N N 1-H C N C N T — 1 C N cn i-H C N 1-H 1-H C N C N i-H ^H n - n - i - H in IT- 00 m oo C- 03 c - ,— H C N r - o O cn C N t - t - cn m cn 30 o 30 in C N t.N o r - IT- i - H in i t C N i - H 30 C N cn t - 03 00 o 00 00 r - 03 C N o cn f— 3 cn 00 30 03 N- 00 C N c - O in in 03 00 1-H o 03 r - 03 00 O 03 03 o 1-H t - 03 o 03 o C N O o o 03 r - 1-H 1-H C N C N C N C N cn C N C N 1 — 1 1 C N C N 1 — 1 C N cn cn H C N C N 1 — 1 1 — 1 1 03 oo 03 03 C N in C- cn 30 r- O O 30 C N 00 C N C N C N C N r— 00 00 30 C N C N C N C N cn C N C N C N C N C N cn C N C N C N C N cn C N C N C N C N C N C N C N C N ^ ■ 3 0 ( S ( 3 i i n h o o N 'O c < i i o > 0 3 t ^ H i o ^ 't ^ , 't H i f l r t c— t— r - r — 3 0 3 0 3 o r - c — r - r — 3 0 i n 3 o r - 3 0 3 0 i n i n 3 o t — 3 0 3 0 g ,p —< g ,£1 g .3 PS ,P a * a C /3 C /3 Q P 'tf C /3 C /3 § a a s C /3 C /3 C /3 C /3 C /3 C /3 C /3 C /3 C /3 0 0 0 0 C /3 00 I pi h -' < C ; P i P P P oo m 30 IT- oo 03 o 1 — H C N cn r f in 30 r - oo 03 o 1 — H C N cn " t in 30 r - 00 30 30 30 30 30 r - r - r - r - C- r - r - r - r - r - oo 00 00 oo oo oo oo oo oo C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N C N 112 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 7.1. Continued ________ N o Sequence A ff nAb NOc BBd Pepe Comp.1 Watg Conf^ H 20‘ Energy* HB1 o o ■ ^ f 1 - H O f ro C N C N 00 oo ro V) r- C N t- O s of C N vn O s C "- N O o o 00 © N O ro C N S O ro N O o C N 00 o V N uo o o o f o C N no V O 1 — H C N r- co r- K N O of so Os so of 00 © 00 of o of t-- Os S O C N O f'- as r - 00 N O as 00 o 00 V N N O of N O C N c - Os V O C N of r~- C N N O ro N O V N V N V N (N ro to N to o Tt O S so vo in in 1 — H cn o so C N C N cn On C N C N T — l 1 ,_ H Tt T f cn O S i> ^ H s6 m i in V O h ; rH ON o o v > o o vn ro <N os vo in so ^H o 00 in co cn sq 00 cn in r-; in vo in in in mi in in cn 00 os o so o N O o oo so i-H Os N O d 00 o O S O S in cn C N cn wn 1-H r - in 1 — H 1 — H 1 1-H 1 — H C" vo o o O S C N vo Os C N C^ C N . C N i> 'w in in O N 1 N" 1 r t C N 1 C N i in ^H in i— H 1 C N O O ' O n ro vo o o i - H m o l> C N cn (N in in in os so os 0 0 cn N ; N ; in i - H i— H cn o cn cn 1— 1 1 -H r - H os O S o cn so m O S 0 0 1 vo so ■ 1 ro oo i— i Q\ V N C N l> C N C N Os © O O O 00 V O o f o f o o o f as vo as oo vo o f o f ro O o o so C N o C " - 00 C N as so © o o o r o o o o o C N O s oo ro so 2 £ o so cn so ^ Tf O <N as © o no t" ON rH 0 \ § t> O ro 2 £ © oo ro £ OS t — * Os r-* c n C n m C N . 4 . V O r-' o C n r - H C N N O 1 C N C N os in 0 0 cn o cn C N C N ? -H N C N O s / — S in i - H 0 0 O s N O o C n r - ~ N O N O m m so N- i - H O s v o m O S O o \| H « C N cn in N" V N C N N O C N 0 0 m i> 0 0 1 -H o so v o O s o 0 0 cn \| O S cn o i— H in cn H " C n O i — H H" m cn 1 — H in 00 cn cn N- cn N O oo o o r - * r - H t" C- r- cn O N o in o 0 0 r- O S o os 0 0 m C N C N r - H 1— H cn C N t - H C N C N C N i— H C N 1— H C N cn C N C N C N cn C N C N C N N- cn in OS C N 0 0 t " - r -H C N r< N N" o 97996 N" o i — H so in o cn i— H O 0 0 o 0 0 O o t - 0 0 oo ro c < N C N V N o 0 0 cn r - 1 -H 0 0 V N C N O S 0 0 oo V > C N C N V N i - H C N O O so r - so o 0 0 o N" cn C N /« “ N ’ 1 C N C - V N H - i - H O N C N V N o V N m cn o cn o ro C N r- 0 0 w «4i ^ H L C n O l> 1 -H l> C N 00 r< N C N 00 o 00 00 o O o 00 00 C N C N cn C N 1 -H C N C N C N i - H C N C N cn C N C N cn cn cn C N C N N O C O S~T\ hs. cn o o o C N C N O N H" Cs C N n- i— H o cn C N O N i— H t -H O C N C -- s j \ ^.1 00 C N r-' Tf i-H r~- O N O i— H N O O N O N o r - Os cn V N C N Cn Cn 00 C N 00 s o i > r ^ - C N C N V N O N O V N V N C * N o cn o r - cn r-- as qa V 4 1 o o r - ^ H 00 O N 00 cn C N O 00 O 00 00 o © o 00 00 C N C N N 1-H cn C N r-H C N C N C N 1 — 1 C N 1 C N r* N C N C N cn C < N cn C N C N OO 00 s o r - o 00 o O N " H " V N O cn N O o 00 © a O C N C N C N C N C N cn C N C N ro C N C N C N cn C N cn C N C N i- H C N i - H < N C N C N in o SO cn 1 -H in o o o n o M , f S i n o c s o o T i , o \ ' o ^ i n ' N o a \ ' o o ,t H f o t ^ i n s o t ' - ' S o t * ,‘' « s o o o i > C ' ^ v o s D r * * « t ^ i n s o s D i n i n t ' - - t > t > s o m £ o > 00 00 00 00 00 00 B-c S 0 S H c a 3 CX O 'fe ^ O ^ & < £ a o'5 3 a a a > > > OS o CN cn N" in so r- 00 Os o CN c n in so 00 os o CN oo OS OS Os os OS Os Os os Os Os o o o o o o o o o o r - H i — H 1— H CN CN CN CN CN CN CN CN CN CN CN cn cn cn cn cn cn cn cn cn cn cn cn cn 113 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 7.1. Continued _______ N o Sequence A ff nAb NOc BBd Pepe Comp.1 Watg C onl* H2Q' Energy1 HB' H r^> t o 00 © r - 00 e ' en VO t o ^ r t o r-H 00 t o o Os Ov cn O n CN CO o vo 00 cn t o cn r - CN ON CN Os CO CN t d t o t o t o t o t o vd vd t o t o t o t o t o vd vd t o t d t o t d NO t d t d VO t o vo CN 00 00 Os © r - cn O n o a s t o r r VO cn 00 vo CN 00 t o o o CN 00 00 t -H CO CN o t o t> CO q i n CO o s o o VO N- vd CN rH t> t o t o c n t o T “H r-H 00 CN t -H 00 O n CO vo T — H CO H CN CN T — H CO <> N " N ' t o N - r t t o r-H i n t ^H t d CN © VO V O N O C O v o 00 r- -22.30 00 C O 00 o N O O m o o > i cn 00 cn -16.90 V O i t O T — H O N Os q N O i> CN 0 0 v o i t CN CN o o td i cn r*- © o o to CN tH 1 C O CN i t> i CN Os C O 1 vd o o CN i I " - ' i 0 0 td i © 1 in cn i N O o O to On o m to C O N O CN o o 00 os tn o r - i t C O i 00 q On " — i t - H q o o CN cn m t-; T — H CN q N O I T ) as i t r> N O On N" C O i f N O CN 00 td N O N O © CN © CN cn CN I> r - v o 00 1> N O N O I"- r- r- r- N O N O r- r- 0 0 0 0 t- 0 0 0 0 N O 0 0 o o O o 100 100 CN N O 97 20 0 0 100 100 100 CN O N in NO 100 ? — H 100 100 100 30 35 100 22 oo oo c n oo Tf O n h ^ i — 1 » — 1 (N t ' O M 3 O n (S OO r- ON CN cn cn t"- o <N m On no - to . r-* ,— 1 vo OS Z2 cn CO t -H t o ^H O i r i c n cn cn CO O n 00 i t cn 00 CN NO i n CN ^H CO T-H 00 O n CN t - o o N - 00 t o 00 o i t o 00 vo t o 00 vo CN oo i t o NO t o 00 T-H T — H o C"- o o i t o 00 o rH CN CN '— I N I— l (N H CN CO v o rH r- o 00 v o ^H o v o 00 v o v o t o t o N - N - N - cn cn m on t> m on On <r> ' W N N y — s r y n i n N ’---- W \ M / '---- '-------------- - ' M i o i n o o [ '2 S 2 S f '( N |'t o o l o o o tn P ^ 9 ° ^ £ c o £ 2 ? ! O P O N N O ^ : 0 0 i t C N C N it it C " ON 0 0 C N C N C N O 00 00 CN ON C N C N < N rH CO o t o O N t o 0 0 vo vo o vo CN r - 0 0 rH CN CO Os o t o 0 0 w Os t o CO CO 0 0 T -H C^ t o CN V O CO CO r - 0 0 CO t -H 0 0 0 0 o o t o o 0 0 o t -H CN CN t -H N CN H CN CO 0 0 vo vo f — - N CN r- o t o ov 0 0 t -H o 'w ' i-yn r- 0 0 r- 0 0 T — 1 r- vo CO CN vo OS o CN 0 0 cn t -H 0 0 0 0 t -H o t -H 0 0 o 1 CN CN , " H t -H CN t — H CN cn o t o vo t o CO CO l> o o CN CN CN CN CN CN CN CN CN fN ON 00 NO N O no r- on r - » C N cn C N O N in cn it O no 0 0 O in N Oin N O N - N " it o o O N N O r^ l> N Oo N ON ON ON O T t it it it N " it it it C N C N C N C N N OO NO no o N ON ON ON O r ~ - O O o o r- i t cn cn o o C N T — H r-H o in 0 0 O o cn i t o i t cn C N N O t -H t -H cn C N O NO nO NO O NO N C N O N C N C N C N i — i C N C N t -H C N r - ~ o CN O t -H 00 VO CN CN t -H O O n v o to m t -H CN C- t -H T — H V O CN O n O N ON o OS OS CN Os CN CN CN 1 — 1 < N < N CN NO i t i t t - 00 Os CN CN CN CN <N CN CN CN m 00 00 cn O r- cn vo 00 NO NO tH r- o r - r - M h o ^ b 2 >• < d > £ £ £ w ^ b s? b ^ w x b i - i Q W 73 J h - 1 hJ o o o WWW £ |S C /3 C /3 W W w w O 0 w w w w >H >M >H cn t n V O r - 00 Os o T — H CN cn N - t o vo 00 OS o ^H CN cn t -H rH T — H t -H T ^H t -H ^H CN CN CN CN CN CN CN CN CN CN cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn cn R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 3 5 YLVAYQATV h 7 2 2 6 28020 278892 277001 6 3 6 3 72.33 -10.67 130.37 5.17 3 3 6 YLVSFGVWI h 7 7 2 3 28562 283014 277567 2 2 9 1 0 0 73.71 -5.34 136.26 5.41 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 7.1. Continued No Sequence A ff nAb 2 O o BBd Pep6 Comp. Watg Conf*1 H 201 Energy3 HB INT1 337 YLVTRHADV m 75 31 28846 288360 282058 61 61 76.69 0.54 138.18 5.38 338 YMDDVVLGA 1 69 24 29175 291721 289292 42 42 74.29 -2.50 137.14 4.81 339 YMDGTMSQV h 73 28 29546 295441 293658 205 100 79.64 -3.49 145.06 5.39 340 YMLDLQPET m 78 28 11647 113544 114625 108 100 77.75 -3.08 140.18 5.83 341 YMNGTMSQV h 73 28 29627 296209 293556 292 100 80.34 12.56 153.01 5.41 342 YTDQVPFSV m 74 28 11279 108438 111075 54 54 75.33 -1.89 135.19 5.63 a Binding affinity, 1 = low, m = medium, h = high (Brusic et al., 1998b) b Number of peptide atoms (excluding hydrogen atoms) c Number of peptide polar atoms (nitrogen and oxygen) d Number of peptide backbone conformations e Number of peptide conformations f Number of peptide-MHC conformations g Number of peptide-water-MHC conformations h Number of output conformations 1 Mean number of water molecules in output conformations 3 Mean calculated VDW contact energy score (see chapter 6 for definition) of output conformations k Mean number of hydrogen bonds in output conformations 1 Mean number of direct hydrogen bonds between the peptide and MHC molecules in output conformation 116 HLA-A2 binds hydrophobic peptides with higher affinity Peptide hydrophobicity was estimated by the number of polar atoms (nitrogen and oxygen) in the sequence. Figure 7.1b shows that high and medium affinity peptides are more hydrophobic than low affinity peptides. The mean numbers of polar atoms for high and medium affinity groups are 22.91 and 23.20 respectively, and these numbers are significantly lower than 25.14 for the low affinity group (statistically p < 0.001 for high and low affinity groups, p = 0.017 for medium and low affinity groups). Note that there is not a significant difference in the mean numbers of peptide atoms between the high and low affinity groups (p = 0.76) and between the medium and low affinity groups (p = 0.65) (Figure 7.1a). However, there is a significant difference in the mean numbers of peptide atoms between the high and medium affinity groups (p = 0.01). The mean numbers of peptide atoms are 69.08, 66.57, and 68.77 for the high, medium, and low affinity groups respectively. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 117 (a) 71 70 IA | 69 w o T3 1. 67 o £ 66 o 68 - 65 64 63 62 (b) 69.08 High 66.57 68.77 Medium Binding Affinity Low 27 -i 26 - 25 - | 24- Q. o 23 - k_ 0 n E 22 - 3 z 25.14 23.20 22.91 20 High Medium Low Binding Affinity Figure 7.1. Peptide composition. The mean values of (a) number of non-hydrogen peptide atoms and (b) number of peptide nitrogen and oxygen atoms are shown for different groups of peptide binding affinities. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 118 High binding affinity peptides are conformationally more flexible Figure 7.2a shows that the mean number of peptide backbone conformations adopted by a peptide sequence for the high binding affinity group is significantly higher than those of the medium and low binding affinity groups, 24,775 compared to 21,689 and 21,886 (p = 0.006 and p = 0.027 respectively). Figure 7.2b shows that the number of compatible peptide-MHC conformations for the high binding affinity group is also significantly higher than those of the medium and low binding affinity groups, 239,675 compared to 211,958 and 212,269 (p = 0.012 and p = 0.039 respectively). These results indicate that high binding affinity peptides are conformationally more flexible than medium and low binding affinity peptides. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 119 (a) 27000 (A C 0 1 25000 § 23000 O g 21000 ■ Q £ 19000 ■ o 1. 17000 0 Q. * 15000 (b) 270000 - i | 250000 r a E o 230000 c o o 210000 - 0 190000 2 a £ 170000 150000 21689 High Medium Binding Affinity Low 211958 21326S High Medium Low Binding Affinity Figure 7.2. Peptide conformational flexibility. The mean values of (a) peptide backbone conformations and (b) peptide-MHC conformations are shown for different groups of peptide binding affinities. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 120 Conformations o f high binding affinity peptide-HLA-A2 complexes are well packed The mean VDW contact energy scores of the high, medium and low binding affinity groups are -6.81, -2.53, and -2.40 respectively (Figure 7.3a). The p-values for the mean differences between the high binding affinity group and the medium and low binding affinity groups are less than 0.0001. VDW contact energy score is a measure of how well the conformations of the peptide and MHC flexible sidechains are packed together (see chapter 6). The results indicate that the conformations of high binding affinity peptide-HLA-A2 complexes are better packed than those of the medium and low binding affinity complexes. Figure 7.3b shows the mean numbers of direct hydrogen bonds between the peptide and HLA-A2. These numbers are 5.59, 5.43 and 5.49 for the high, medium and low binding affinity groups respectively. Although the value for the high binding affinity group is significantly different from the medium binding affinity group’s (p = 0.017), it is not significantly different from the low binding affinity group (p = 0.186). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 121 (a) 0.00 - 2.00 £ o o tn > . S ? -4.00 0 c LU 1 -6.00 o - 8.00 - 10.00 (b) 5.80 - i < A “ O C 0 CD c Q > O ) 2 T J £ O £ 5 1 o ■ o 5.60 5.40 - 5 5.20 a a > a. 5.00 2.40 High Medium Binding Affinity Low 5.59 5.49 High Medium Binding Affinity Low Figure 7.3. Peptide-MHC intermolecular interactions. The mean values of (a) VDW contact energy score and (b) number of direct peptide-MHC hydrogen bonds are shown for different groups of peptide binding affinities. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 122 Numbers o f solvated and output conformations are not related to peptide-MHC binding affinity Figure 7.4 shows the mean numbers of peptide-MHC conformations that were solvated and outputted by the MHC program. There is not a significant difference between the high (164 solvated, 67 outputted) and low (169 solvated, 75 outputted) binding affinity groups (p-values >0.1) although the high binding affinity group has fewer conformations. It is interesting, however, that the medium binding affinity group has significantly more solvated (238) and outputted (80) conformations than the high and low groups. Because of the criteria (see chapter 6) for selecting conformations for solvation and output are set dynamically and therefore are not standardized, it is difficult to find a physical meaning for these statistics. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 123 (a) v > 300 * 250 - o 200 - 150 - « 100 - 169 164 50 - High Medium Low (b) c o o 90 85 80 75 70 65 60 g 55 o * 50 45 40 Binding Affinity 67 80 75 High Medium Binding Affinity Low Figure 7.4. Peptide-water-MHC conformations. The mean values of (a) number of peptide-water-MHC conformations and (b) number of output conformations are shown for different groups of peptide binding affinities. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 124 Extensive interface water networks decreasepeptide-HLA-A2 binding affinities Figure 7.5a shows the mean number of predicted water molecules at peptide- HLA-A2 binding interfaces of the three binding groups (70.52, 72.06, and 72.96 for high, medium and low binding affinity groups respectively). The p-values for the differences between high-medium, high-low, and medium-low groups are 0.065, 0.008, and 0.604 respectively. These results indicate that there are significantly fewer water molecules at the binding interfaces of high affinity peptide-HLA-A2 complexes. Similarly, Figure 7.5b shows the mean number of direct and water- mediated hydrogen bonds (> 95% are water-mediated) at binding interfaces of the three binding groups (128.46, 131.97, and 132.91 for high, medium and low binding affinity groups respectively). The p-values for the differences between high- medium, high-low, and medium-low groups are 0.062, 0.030, and 0.478 respectively. High binding affinity peptide-HLA-A2 complexes therefore have fewer water- mediated hydrogen bonds than the medium and low affinity complexes. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 125 (a) 75 -| 74 - High Medium Low Binding Affinity (b) 138 n 136 - 134 - 132 - 130 - p 128 - 32.91 126 - 124 - 122 - 120 High Medium Low Binding Affinity Figure 7.5. Water-mediated peptide-MHC interactions. The mean values of (a) number of added water molecules and (b) number of peptide-Water-MHC hydrogen bonds are shown for different groups of peptide binding affinities. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 126 Predicting peptide-HLA-A2 binding affinity The experimental IC50 (nM) values of four peptides were published by Altuvia et al., (Altuvia et al., 1995) (Table 7.2). Based on these values, an equation was developed to predict the binding affinity of peptide-HLA-A2 complexes (Equation 1). The natural log of 1/IC50 (ln(l/IC50) is dependent on three variables: the number of interface water molecules (H20), the VDW contact energy score (Energy), and the number of direct hydrogen bonds between the peptide and MHC molecule (INT). Equation 1 shows that the binding affinity is negatively correlated with the number of interface water molecules and is positively correlated with the number of direct peptide-MHC hydrogen bonds. Certainly, low VDW contact energy score corresponds to high binding affinity. A good correlation (R = 0.96) between predicted and experimental binding affinity was obtained for the training set of four peptides listed in Table 7.2 (Figure 7.6). To test its predictive power, Equation 1 was used to predict binding affinity of the 342 peptide-HLA-A2 complexes. Figure 7.7a shows the predicted binding affinity distributions for the three different binding groups. The mean predicted binding affinities (ln(l/IC50) for the high, medium, and low binding affinity groups are -2.76, -3.13, and -3.42 respectively (Figure 7.7b). The mean predicted binding affinities are significantly different between the high and low binding affinity groups (p = 0.013), and are marginally different between the high-medium groups (p = 0.10), but are not significantly different between the medium-low groups (p = 0.97). These results R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 127 indicate that Equation 1 can be confidently used to differentiate between high and low binding affinity peptides. Table 7.2. Binding interactions of selected peptide-HLA-A2 complexes No Sequence Name IC50 (nM)a H 20b Energyc INT 1 GILGFVFTL MT 6 61.86 -1.77 5.60 2 LLFGYPVYV TX 11 69.41 -9.52 5.33 3 ILKEPVHGV RT 242 74.73 2.15 5.23 4 TLTSCNTSV GP 294 79.00 -2.79 5.86 a (Altuvia et al., 1995) b Mean number of water molecules in output conformations c Mean calculated VDW contact energy score (see chapter 6 for definition) of output conformations d Mean number of direct hydrogen bonds between the peptide and MHC molecules in output conformations Ln(l/IC50) = 3.573 - 0.203*H20 -1.62*Energy + 1.229*INT (Equation 1) w O T 3 < D T 3 a > V . 0 1 MT -2 TX ■ 3 -4 GP y = 0.868x - 0.3065 R2 = 0.9601 ■ 5 RT ■ 6 1 0 6 ■ 4 ■ 3 2 ■ 5 Experim ental ln(1/IC50 (nM)) Figure 7.6. Correlation between predicted and experimental peptide-HLA-A2 binding affinity. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 128 (a) 20-. 15- o a Predicted ln(1/IC50 (nM)) M High M Medium B Low (b) - 2.0 n ss -2.5 - £ c ~ -3.5 - u < 5 o -4.0 - ' ■ 5 £ “■ -4.5 - -5.0 - High Medium Low Binding Affinity Figure 7.7. Distribution of predicted binding affinity, (a) Distribution of Ln(l/IC50 (nM)) for different peptide binding affinity groups, (b) The mean predicted Ln(l/IC50 (nM)) values for different peptide binding affinity groups. The error bars show the 95% confident intervals of the indicated means. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 129 Visualizing binding properties o f peptide-HLA-A2 complexes Figures 7.8 and 7.9 show the 3-D (Microcal Software, 1999) and the stereo view (Bolger, 2003) of the binding properties (INT, Energy, H20 (see Table 7.2 for definitions)) of the peptide-HLA-A2 complexes. No clear separation was observed for the high, medium, and low binding affinity groups. -25 95 -30 Figure 7.8. A 3-D scatter plot of peptide-HLA-A2 binding properties. High (square), medium (circle) and low (triangle) binding affinity peptides are shown. The plot was made using the microcal software (Microcal Software, 1999). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. H |4 D D ata Mining 4D D ata Mining k < 3 , R o H d o : 3 3 f B s w t t a i : 3 4 9 * Figure 7.9. Stereo view of peptide-HLA-A2 binding properties. The graph was generated using the QMPRPlus software. © 131 Kohonen Self Organizing Feature Map for MHC - binding to peptides (Michael B. Bolger, 3/16/03) The result presented in this section was done by Dr. Michael B. Bolger. Using Kohonen self organzing features map, approximate 50% of peptides were correctly classified into high, medium, and low binding affinity groups (Tables 7.3 and 7.4). However, about 32-38% of the peptides were unclassified. The percentage of correct classification for the classified peptides therefore is ~ 80%. This classification scheme was based on the three peptide-MHC binding properties— INT, H20, and Energy described in Table 7.2. More binding properties perhaps could be used and might improve the classification power and reduce the number of unclassified peptides. Table 7.3. Statistics of binding classification using Kohonen self organizing feature map for MHC-binding to peptides Low Med High Total 86 123 133 Correct 43 61 74 Wrong 10 21 16 Unknown 33 41 43 Correct(%) 50 49.6 55.6 Wrong(%) 11.63 17.1 12 Unknown(%) 38.37 33.3 32.3 Table 7.4. Kohonen self organizing map binding classification Low3 Medium3 High3 Lowb 43 11 11 Mediumb 6 61 5 Highb 4 10 74 a Experimental binding affinity b Predicted binding affinity R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 132 7.4. Discussion The goal of this chapter is to correlate structural properties to binding affinities of peptide-HLA-A2 complexes. Since the available experimental binding affinities are categorical, it was not possible to obtain a quantitative structure-activity relationship. Nevertheless, qualitative relationships were obtained by comparing population means. Based on the observed qualitative relationships, a predictive function was developed and optimized using a limited set of available continuous binding data. It was shown that hydrophobic interactions (VDW energy score) and direct hydrogen bonds between the peptide and HLA-A2 contribute positively to the binding affinity of peptide-HLA-A2 complexes. Trapping of water molecules at the binding interface, however, has a negative effect on the peptide-HLA-A2 binding interactions. This is probably due to the relatively hydrophobic binding surface of HLA-A2 and the high entropic cost of immobilizing water molecules. Although peptide conformational flexibility was not incorporated in the predictive function, it was shown that high binding affinity peptides can adopt a significantly higher number of backbone conformations. Using Kohonen self organizing feature map, an unsupervised learning neural network method, Dr. Michael B. Bolger was able to classify peptides with a 50% accuracy. This method could be explored further to include more structural properties and to arrive at a better classification scheme. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 133 Chapter 8 Summary The focus of this work is to develop a structure based method for predicting peptide-MHC binding interactions. This requires (1) a fast and accurate modeling of peptide-MHC binding conformations and (2) a scoring function that can estimate the complex binding affinity. Three critical factors: (1) adequate sampling of peptide backbone conformations, (2) flexibility of MHC sidechains, and (3) modeling of explicit interface water molecules have been considered and contributed to the success of the MHC program in predicting structures of peptide-HLA-A2 complexes. The binding affinities of these complexes have been estimated based on a scoring function consisting of three variables: (1) VDW contact energy score, (2) number of direct hydrogen bonds between the peptide and HLA-A2, and (3) number of interface water molecules. It was shown that this scoring function can be used to differentiate high and low peptide binders with a reasonable confidence. The presented method can be extended to predicting binding interaction of not only other peptide-MHC complexes but also complexes of peptide-MHC with the TCR. The ability to predict the structure is also important for a better R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 134 characterization of the Tcell receptor ligand and a better understanding of T cell activation, antagonism, selection and cross-reactivity (molecular mimicry). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 135 REFERENCES Adams, H. P. & Koziol, J. A. (1995). Prediction of binding to MHC class I molecules. J. Immunol. Methods 185, 181-190. Adrian, P., Rajaseger, G., Mathura, V., Sakharkar, M., & Kangueane, P. (2002). Types of inter-atomic interactions at the MHC-peptide interface: Identifying commonality from accumulated data. BMC. Struct. Biol. 2,2. Altuvia, Y., Schueler, O., & Margalit, H. (1995). Ranking potential binding peptides to MHC molecules by a computational threading approach. J. Mol. Biol. 249, 244- 250. Altuvia, Y., Sette, A., Sidney, J., Southwood, S., & Margalit, H. (1997). A structure- based algorithm to predict potential binding peptides to MHC molecules with hydrophobic binding pockets. Hum. Immunol. 58, 1-11. Apostolopoulos, V., Yu, M., McKenzie, I. F., & Wilson, I. A. (2000). Structural implications for the design of molecular vaccines. Curr. Opin. Mol. Ther. 2, 29-36. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., & Bourne, P. E. (2000). The Protein Data Bank. Nucleic Acids Res. 28, 242. Bernstein, F. C., Koetzle, T. F., Williams, G. J. B., Meyer, E. F. Jr., Brice, M. D., Rodgers, J. R., Kennard, O., Shimanouchi, T., & Tasumi, M. (1977). The protein data bank: A computer-based archival file for macromolecular structures. J. Mol. Biol. 112, 535-542. Bjorkman, P. J. & Parham, P. (1990). Structure, function, and diversity of class I major histocompatibility complex molecules. Annu. Rev. Biochem. 59, 253-288. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., & Wiley, D. C. (1987a). Structure of the human class I histocompatibility antigen, HLA-A2. Nature 329, 506-512. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., & Wiley, D. C. (1987b). The foreign antigen binding site and T-cell recognition regions of class I histocompatibility antigens. Nature 329, 512-518. Bolger, M. B. QMPRPlus. 2003. Ref Type: Computer Program R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 136 Boon, T., Cerottini, J. C., van den Eynde, B., van der Bruggen, P., & van Pel, A. (1994). Tumor antigens recognized by T lymphocytes. Annu. Rev. Immunol. 12, 337- 365. Bouvier, M., Guo, H. C., Smith, K. J., & Wiley, D. C. (1998). Crystal structures of HLA-A*0201 complexed with antigenic peptides with either the a. Proteins 33, 97- 106. Bouvier, M. & Wiley, D. C. (1994). Importance of peptide amino and carboxyl termini to the stability of MHC class I molecules. Science 265, 398-402. Bouvier, M. & Wiley, D. C. (1996). Antigenic peptides containing large PEG loops designed to extend out of the HLA-A2 binding site form stable complexes with class I major histocompatibility complex molecules. Proceedings o f the National Academy o f Sciences o f the United States o f America 93, 4583-4588. Bmsic, V., Petrovsky, N., Zhang, G., & Bajic, V. (2002). Prediction of promiscuous peptides that bind HLA class I molecules. Immunol. Cell Biol. 80, 280-285. Bmsic, V., Rudy, G., & Harrison, L. C. (1998b). MHCPEP, a database of MHC- binding peptides: update 1997. Nucleic Acids Res. 26, 368-371. Bmsic, V., Rudy, G., Honeyman, G., Hammer, J., & Harrison, L. (1998a). Prediction of MHC class II-binding peptides using an evolutionary algorithm and artificial neural network. Bioinformatics. 14, 121-130. Bmsic, V., Schonbach, C., Takiguchi, M., Ciesielski, V., & Harrison, L. C. (1997). Application of genetic search in derivation of matrix models of peptide binding to MHC molecules. Ismb. 5, 75-83. Buus, S. (1999). Description and prediction of peptide-MHC binding: the 'human MHC project'. Curr. Opin. Immunol. 11, 209-213. Buus, S., Sette, A., Colon, S. M., Miles, C., & Grey, H. M. (1987). The relation between major histocompatibility complex restriction and the capacity of la to bind immunogenic peptides. Science 235,1353-1358. Chicz, R. M. & Urban, R. G. (1994). Analysis of MHC-presented peptides: applications in autoimmunity and vaccine development. Immunol. Today 15, 155- 160. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 137 Collins, E. J., Garboczi, D. N., Karpusas, M. N., & Wiley, D. C. (1995). The three- dimensional structure of a class I major histocompatibility complex molecule missing the alpha 3 domain of the heavy chain. Proc. Natl. Acad. Sci. U. S. A 92, 1218-1221. Collins, E. J., Garboczi, D. N., & Wiley, D. C. (1994). Three-dimensional structure of a peptide extending from one end of a class IMHC binding site. Nature 371, 626- 629. Coulie,P.G. (1996). Human tumor antigens recognized by cytotoxic T lymphocytes. In Tumor Immunology: Immunotherapy and Cancer Vaccines., A.G.Dalgleish and M.J.Browning, eds. (Cambridge: Cambridge University Press), pp. 95-125. Desmet, J., Wilson, I. A., Joniau, M., De Maeyer, M., & Lasters, I. (1997). Computation of the binding of fully flexible peptides to proteins with flexible side chains. FASEB J. 11, 164-172. Ding, Y. H., Smith, K. J., Garboczi, D. N., Utz, U., Biddison, W. E., & Wiley, D. C. (1998). Two human T cell receptors bind in a similar diagonal mode to the HLA- A2/TAX peptide complex using different TCR amino acids. Immunity 8, 403. Doherty, P. C. & Zinkemagel, R. M. (1975). A biological role for the major histocompatibility antigens. Lancet 1,1406-1409. Donnes, P. & Elofsson, A. (2002). Prediction of MHC class I binding peptides, using SVMHC. BMC. Bioinformatics. 3, 25. Doytchinova, I. A. & Flower, D. R. (2001). Toward the quantitative prediction of T- cell epitopes: coMFA and coMSIA studies of peptides with affinity for the class I MHC molecule HLA-A*0201. J. Med. Chem. 44, 3572-3581. Doytchinova, I. A. & Flower, D. R. (2002b). Physicochemical explanation of peptide binding to HLA-A*0201 major histocompatibility complex: a three-dimensional quantitative structure- activity relationship study. Proteins 48, 505-518. Doytchinova, I. A. & Flower, D. R. (2002a). Quantitative approaches to computational vaccinology. Immunol. Cell Biol. 80, 270-279. Dunbrack, R. L., Jr. & Karplus, M. (1993). Backbone-dependent rotamer library for proteins. Application to side-chain prediction. J. Mol. Biol. 230, 543-574. Engelhard, V. H. (1994b). Structure of peptides associated with class I and class II MHC molecules. Annu. Rev. Immunol. 12, 181-207. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 138 Engelhard, V. H. (1994a). Structure of peptides associated with MHC class I molecules. Curr. Opin. Immunol. 6, 13-23. Ernst, J. A., Clubb, R. T., Zhou, H. X., Gronenbom, A. M., & Clore, G. M. (1995). Demonstration of positionally disordered water within a protein hydrophobic cavity byNM R [see comments]. Science 267, 1813-1817. Falk, K., Rotzschke, O., Stevanovic, S., Jung, G., & Rammensee, H. G. (1991). Allele-specific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 351 , 290-296. Fremont, D. H., Matsumura, M., Stura, E. A., Peterson, P. A., & Wilson, I. A. (1992). Crystal structures of two viral peptides in complex with murine MHC class I H-2Kb. Science 257, 919-927. Fremont, D. H., Stura, E. A., Matsumura, M., Peterson, P. A., & Wilson, I. A. (1995). Crystal structure of an H-2Kb-ovalbumin peptide complex reveals the interplay of primary and secondary anchor positions in the major histocompatibility complex binding groove. Proc. Natl. Acad. Sci. U . S. A 92, 2479-2483. Froloff, N., Windemuth, A., & Honig, B. (1997). On the calculation of binding free energies using continuum methods: application to MHC class I protein-peptide interactions. Protein Sci. 6, 1293-1301. Gao, G. F., Tormo, J., Gerth, U. C., Wyer, J. R., McMichael, A. J., Stuart, D. I., Bell, J. I., Jones, E. Y., & Jakobsen, B. K. (1997). Crystal structure of the complex between human CD8alpha(alpha) and HLA-A2. Nature 387, 630. Garboczi, D. N., Ghosh, P., Utz, U., Fan, Q. R., Biddison, W. E., & Wiley, D. C. (1996). Structure of the complex between human T-cell receptor, viral peptide and HLA-A2. Nature 384, 134. Garrett, T. P., Saper, M. A., Bjorkman, P. J., Strominger, J. L., & Wiley, D. C. (1989). Specificity pockets for the side chains of peptide antigens in HLA-Aw68. Nature 342, 692-696. Gaur, A. & Fathman, C. G. (1994). Immunotherapeutic strategies directed at the trimolecular complex. Advances in Immunology 56, 219-265. Germain, R. (1994). MHC-dependent antigen processing and peptide presentation: providing ligands for T lymphocyte. Cell 76, 287-299. Germain, R. N. & Margulies, D. H. (1993). The biochemistry and cell biology of antigen processing and presentation. Annu. Rev. Immunol. 11, 403-450. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 139 Gewurz, B. E., Gaudet, R., Tortorella, D., Wang, E. W., Ploegh, H. L., & Wiley, D. C. (2001). Antigen presentation subverted: structure of the human cytomegalovirus protein US2 bound to the class I molecule HLA-A2. Proc. Natl. Acad. Sci. U. S. A 98, 6794. Glithero, A., Tormo, J., Haurum, J. S., Arsequell, G., Valencia, G., Edwards, J., Springer, S., Townsend, A., Pao, Y. L., Wormald, M., Dwek, R. A., Jones, E. Y., & Elliott, T. (1999). Crystal structures of two H-2Db/glycopeptide complexes suggest a molecular basis for CTL cross-reactivity. Immunity 10, 63-74. Goodford, P. J. (1985). A computational procedure for determining energetically favorable binding sites on biologically important macromolecules. J. Med. Chem. 28, 849-857. Gulukota, K., Sidney, J., Sette, A., & DeLisi, C. (1997). Two complementary methods for predicting peptides binding major histocompatibility complex molecules. J. Mol. Biol. 267, 1258-1267. Guo, H. C., Jardetzky, T. S., Garrett, T. P., Lane, W. S., Strominger, J. L., & Wiley, D. C. (1992). Different length peptides bind to HLA-Aw68 similarly at their ends but bulge out in the middle. Nature 360, 364-366. Gurlo, T., Meng, W. S., Bui, H. H., Haworth, I. S., & von Grafenstein, H. (1999). Experimental evidence for the presence of a water network at the peptide-MHC interface. Immunol. Lett. 70, 139-141. Hammer, J. (1995). New methods to predict MHC-binding sequences within protein antigens. Curr. Opin. Immunol. 7, 263-269. Henchman, R. H. & McCammon, J. A. (2002). Extracting hydration sites around proteins from explicit water simulations. J. Comput. Chem. 23, 861-869. Hillig, R. C., Coulie, P. G., Stroobant, V., Saenger, W., Ziegler, A., & Hulsmeyer, M. (2001). High-resolution structure of HLA-A*0201 in complex with a tumor- specific antigenic peptide encoded by the MAGE-A4 gene. J. Mol. Biol. 310, 1167- 1176. Honeyman, M. C., Brusic, V., Stone, N. L., & Harrison, L. C. (1998). Neural network-based prediction of candidate T-cell epitopes. Nat. Biotechnol. 16, 966-969. Hunt, D. F., Michel, H., Dickinsion, T. A., Shabanowitz, J., Cox, A. L., Sakaguchi, K., Appella, E., Grey, H. M., & Sette, A. (1992). Characterization of peptides bound to class I MHC molecule HLA-A2.1 by mass spectrometry. Science 256,1817-1820. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 140 Janin, J. (1999). Wet and dry interfaces: the role of solvent in protein-protein and protein-DNA recognition. Structure Fold. Des 7, R277-R279. Jones, E. Y. (1997). MHC class I and class II structures. Curr. Opin. Immunol. 9, 75- 79. Jones, S. & Thornton, J. M. (1995). Protein-protein interactions: a review of protein dimer structures. Prog. Biophys. Mol. Biol. 63, 31-65. Kangueane, P., Sakharkar, M., Lim, K. S., Hao, H., Lin, K., Ren, E. C., & Kolatkar, P. R. (2000). Knowledge based grouping of modeled HLA peptide complexes. Hum. Immunol. 61, 460-466. Kangueane, P., Sakharkar, M. K., Kolatkar, P. R , & Ren, E. C. (2001). Towards the MHC-Peptide combinatorics. Hum. Immunol. 62, 539-556. Khan, A. R., Baker, B. M., Ghosh, P., Biddison, W. E., & Wiley, D. C. (2000). HLA-A*0201/Octameric Tax peptide complex with an empty conserved peptide-N- terminal binidng site. J. Immunol. 164, 6398. Kirksey, T. J., Pogue-Caley, R. R., Frelinger, J. A., & Collins, E. J. (1999). The structural basis for the increased immunogenicity of two Hiv-reverse transcriptase peptide variant/classi major histocompatibility complexes. J. Biol. Chem. 274, 37259. Klein, J., Satta, Y., O'hUigin, C., & Takahata, N. (1993). The molecular descent of the major histocompatibility complex. Annu. Rev. Immunol. 11, 269-295. Kuhns, J. J., Batalia, M. A., Yan, S., & Collins, E. J. (1999). Poor binding of a HER- 2/neu epitope (GP2) to HLA-A2.1 is due to a lack of interactions with the center of the peptide. J. Biol. Chem. 274, 36422-36427. Ladbury, J. E. (1996). Just add water! The effect of water on the specificity of protein-ligand binding sites and its potential application to drug design. Chem. Biol. 3, 973-980. Lauemoller, S. L., Holm, A., Hilden, J., Brunak, S., Holst Nissen, M., Stryhn, A., Ostergaard Pedersen, L., & Buus, S. (2000). Quantitative predictions of peptide binding to MHC class I molecules using specificity matrices and anchor-stratified calibrations. Tissue Antigens 57, 405-414. Lauemoller, S. L., Kesmir, C., Corbet, S. L., Fomsgaard, A., Holm, A., Claesson, M. H., Brunak, S., & Buus, S. (2001). Identifying cytotoxic T cell epitopes form genomic and proteomic information. Rev. Immunogenet. 3, 00-00. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 141 Lee, C. & McConnel, H. M. (1995). A general model of invariant chan association with class II major histocompatibility complex proteins. Proc. Natl. Acad. Sci. U. S. A 92, 8269-8273. Levitt, M. & Park, B. H. (1993). Water: now you see it, now you don't. Structure. 1, 223-226. Lo, C. L., Chothia, C., & Janin, J. (1999). The atomic structure of protein-protein recognition sites. J. Mol. Biol. 285, 2177-2198. Madden, D. R. (1995). The three-dimensional structure of peptide-MHC complexes. Annu. Rev. Immunol. 13, 587-622. Madden, D. R., Garboczi, D. N., & Wiley, D. C. (1993). The antigenic identity of peptide-MHC complexes: a comparison of the conformations of five viral peptides presented by HLA-A2 [published erratum appears in Cell 1994 Jan 28;76(2):following 410], Cell 75, 693-708. Madden, D. R., Gorga, J. C., Strominger, J. L., & Wiley, D. C. (1991). The structure of HLA-B27 reveals nonamer self-peptides bound in an extended conformation. Nature 353, 321-325. Madden, D. R., Gorga, J. C., Strominger, J. L., & Wiley, D. C. (1992). The three- dimensional structure of HLA-B27 at 2.1 A resolution suggests a general mechanism for tight peptide binding to MHC. Cell 70, 1035-1048. Maenaka, K., Maenaka, T., Tomiyama, H., Takiguchi, M., Stuart, D. I., & Jones, E. Y. (2000). Nonstandard peptide binding revealed by crystal structures of HLA- B*5101 complexed with HIV immunodominant epitopes. J. Immunol. 165, 3260- 3267. Makarov, V. A., Andrews, B. K., & Pettitt, B. M. (1998). Reconstructing the protein- water interface. Biopolymers 45, 469-478. Mamitsuka, H. (1998). Predicting peptides that bind to MHC molecules using supervised learning of hidden Markov models. Proteins 33, 460-474. Matsumura, M., Fremont, D. H., Peterson, P. A., & Wilson, I. A. (1992). Emerging principles for the recognition of peptide antigens by MHC class I molecules. Science 257, 927-934. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 142 Meister, G. E., Roberts, C. G., Berzofsky, J. A., & De Groot, A. S. (1995). Two novel T cell epitope prediction algorithms based on MHC-binding motifs; comparison of predicted and published epitopes from Mycobacterium tuberculosis and HIV protein sequences. Vaccine 13, 581-591. Meng, W. S., Bui, H. H., & Haworth, I. S. (2000b). Exploiting the peptide-MHC water inteface in the computer-aided design of non-natural peptides that bind to the class I MHC molecule HLA-A2. Mol. Sim. 24, 215-228. Meng, W. S. & Butterfield, L. H. (2002). Rational design of peptide-based tumor vaccines. Pharm. Res. 19, 926-932. Meng, W. S., von Grafenstein, H., & Haworth, I. S. (1997). A model of water structure inside the HLA-A2 peptide binding groove. Int. Immunol. 9, 1339-1346. Meng, W. S., von Grafenstein, H., & Haworth, I. S. (2000a). Water dynamics at the binding interface of four different HLA-A2-peptide complexes. Int. Immunol. 12, 949-957. Microcal Software. Microcal. [6.0]. 1999. Microcal Software, Inc. 1991. Ref Type: Computer Program Milik, M., Sauer, D., Brunmark, A. P., Yuan, L., Vitiello, A., Jackson, M. R., Peterson, P. A., Skolnick, J., & Glass, C. A. (1998). Application of an artificial neural network to predict specific class I MHC binding peptide sequences [see comments]. Nat. Biotechnol. 16, 753-756. Nadassy, K., Wodak, S. J., & Janin, J. (1999). Structural features of protein-nucleic acid recognition sites. Biochemistry 38,1999-2017. Ogata, K. & Wodak, S. J. (2002). Conserved water molecules in MHC class-I molecules and their putative structural and functional roles. Protein Eng 15, 697- 705. Parker, K. C., Bednarek, M. A., & Coligan, J. E. (1994). Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains. J. Immunol. 152, 163-175. Parker, K. C., Bednarek, M. A., Hull, L. K., Utz, U., Cunningham, B., Zweerink, H. J., Biddison, W. E., & Coligan, J. E. (1992). Sequence motifs important for peptide binding to the human MHC class I molecule, HLA-A2. J. Immunol. 149, 3580-3587. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 143 Parker, K. C., Shields, M., DiBrino, M., Brooks, A., & Coligan, J. E. (1995). Peptide binding to MHC class I molecules: implications for antigenic peptide prediction. Immunol. Res. 14, 34-57. Phillips, G. N., Jr. & Pettitt, B. M. (1995). Structure and dynamics of the water around myoglobin. Protein Sci. 4,149-158. Pitt, W. R. & Goodfellow, J. M. (1991). Modelling of solvent positions around polar groups in proteins. Protein Eng 4, 531-537. 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. Rammensee, H. G., Friede, T., & Stevanoviic, S. (1995). MHC ligands and peptide motifs: first listing. Immunogenetics 41, 178-228. Reche, P., Glutting, J., & Reinherz, E. (2002). Prediction of MHC class I binding peptides using profile motifs. Hum. Immunol. 63, 701. Reid, S. W., Mcadam, S., Smith, K. J., Klenerman, P., O'Callaghan, C. A., Harlos, K., Jakobsen, B. K., McMichael, A. J., Bell, J. I., Stuart, D., & Jones, E. Y. (1996). Antagonist HIV-1 gag peptides induce structural changes in HLA-B8. J. Exp. Med. 184, 2286. Ridgway, W. M., Fasso, M., & Fathman, C. G. (1999). A new look at MHC and autoimmune disease. Science 284, 749,751. Robinson, J., Malik, A., Parham, P., Bodmer, J. G., & Marsh, S. G. (2000). IMGT/HLA database-a sequence database for the human major histocompatibility complex. Tissue Antigens 55, 280-287. Roe, S. M. & Teeter, M. M. (1993). Patterns for prediction of hydration around polar residues in proteins. J. Mol. Biol. 229, 419-427. Rognan, D., Lauemoller, S. L., Holm, A., Buus, S., & Tschinke, V. (1999). Predicting binding affinities of protein ligands from three-dimensional models: application to peptide binding to class I major histocompatibility proteins. J. Med. Chem. 42, 4650-4658. Rognan, D., Reddehase, M. J., Koszinowski, U. H., & Folkers, G. (1992). Molecular modeling of an antigenic complex between a viral peptide and a class I major histocompatibility glycoprotein. Proteins 13, 70-85. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 144 Rognan, D., Scapozza, L., Folkers, G., & Daser, A. (1994). Molecular dynamics simulation of MHC-peptide complexes as a tool for predicting potential T cell epitopes. Biochemistry 33, 11476-11485. Rosenfeld, R., Zheng, Q., Vajda, S., & DeLisi, C. (1993). Computing the structure of bound peptides. Application to antigen recognition by class I major histocompatibility complex receptors. Journal o f Molecular Biology 234, 515-521. Rosenfeld, R., Zheng, Q., Vajda, S., & DeLisi, C. (1995). Flexible docking of peptides to class I major-histocompatibility- complex receptors. Genet. Anal. 12, 1- 21. Rothbard, J. B. & Gefter, M. L. (1991). Interactions between immunogenic peptides and MHC proteins. Annu. Rev. Immunol. 9, 527-565. Saper, M. A., Bjorkman, P. J., & Wiley, D. C. (1991). Refined structure of the human histocompatibility antigein HLA-A2 at 2.6 A resolution. J. Mol. Biol. 219, 277-319. Savage, H. & Wlodawer, A. (1986). Determination of water structure around biomolecules using X-ray and neutron diffraction methods. Methods Enzymol. 127, 162-183. Schafer, J. R., Jesdale, B. M., George, J. A., Kouttab, N. M., & De Groot, A. S. (1998). Prediction of well-conserved HIV-1 ligand using a matrix-based algorithm, Epimatrix. Vaccine 16, 1880-1884. Schirle, M., Weinschenk, T., & Stevanovic, S. (2001). Combining computer algorithms with experimental approaches permits the rapid and accurate identification of T cell epitopes from defined antigens. J. Immunol. Methods 257, 1- 16. Schueler-Furman, O., Altuvia, Y., Sette, A., & Margalit, H. (2000). Structure-based prediction of binding peptides to MHC class I molecules: application to a broad range of MHC alleles. Protein Sci. 9, 1838-1846. Schueler-Furman, O., Elber, R., & Margalit, H. (1998). Knowledge-based structure prediction of MHC class I bound peptides: a study of 23 complexes. Fold. Des 3, 549-564. Sette, A., Buus, S., Appella, E., Smith, J. A., Chesnut, R., Miles, C., Colon, S. M., & Grey, H. M. (1989). Prediction of major histocompatibility complex binding regions of protein antigens by sequence pattern analysis. Proc. Natl. Acad. Sci. U . S. A 86, 3296-3300. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 145 Sezerman, U., Vajda, S., Comette, J., & DeLisi, C. (1993). Toward computational determination of peptide-receptor structure. Protein Sci. 2, 1827-1843. Sezerman, U., Vajda, S., & DeLisi, C. (1996). Free energy mapping of class I MHC molecules and structural determination of bound peptides. Protein Sci. 5, 1272-1281. Sherman, L. A. & Chattopadhyay, S. (1993). The molecular basis of allorecognition. Annu. Rev. Immunol. 11, 385-402. Silver, M. L., Guo, H. C., Strominger, J. L., & Wiley, D. C. (1992). Atomic structure of a human MHC molecule presenting an influenza virus peptide. Nature 360, 367- 369. Silz, P., Michielin, O., Cerottini, J.-C., Luescher, I., & Romero, P. (2001). Crystal structures of two closely related but antigenically distinct HLA-A2/Melanocyte- Melanoma tumor-antigen peptide complexes. J. Immunol. 167, 3276-3284. Smith, K. J., Reid, S. W., Harlos, K., McMichael, A. J., Stuart, D. I., Bell, J. I., & Jones, E. Y. (1996a). Bound water structure and polymorphic amino acids act together to allow the binding of different peptides to MHC class I HLA-B53. Immunity. 4, 215-228. Smith, K. J., Reid, S. W., Stuart, D. I., McMichael, A. J., Jones, E. Y., & Bell, J. I. (1996b). An altered position of the alpha 2 helix of MHC class I is revealed by the crystal structure of HLA-B*3501. Immunity. 4, 203-213. Speir, J. A., Abdel-Motal, U. M., Jondal, M., & Wilson, I. A. (1999). Crystal structure of an MHC class I presented glycopeptide that generates carbohydrate- specific CTL. Immunity. 10, 51-61. Speir, J. A., Stevens, J., Joly, E., Butcher, G. W., & Wilson, I. A. (2001). Two different, highly exposed, bulged structures for an unusually long peptide bound to rat MHC class I RTl-Aa. Immunity 14, 81-92. Stauss, H. J. (1999). Immunotherapy with CTLs restricted by nonself MHC. Immunol. Today 20, 180-183. Stauss,H.J. and Dahl,A.M. (1996). Cellular oncogenes for tumour immunity: immunotherapy. In Tumor Immunology: Immunotherapy and Cancer vaccines., A.G.Dalgleish and M.J.Browning, eds. (Cambridge: Cambridge University Press), pp. 153-184. Stevanovic, S. & Rammensee, H. G. (1994). Identification of T-cell epitopes using allele-specific ligand motifs. Behring Inst. Mitt. 7-13. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 146 Takahata, N. (1995). MHC diversity and selection. Immunol. Rev. 143, 225-247. Thanki, N., Thornton, J. M., & Goodfellow, J. M. (1988). Distributions of water around amino acid residues in proteins. J. Mol. Biol. 202, 637-657. Thorsby, E. (1999). MHC structure and function. Transplant. Proc. 31, 713-716. Torino, J., Natarajan, K., Margulies, D. H., & Mariuzza, R. A. (1999). Crystal structure of a lectin-like natural killer cell receptor bound to its MHC class I ligand. Nature 402, 623-631. Townsend, A. & Bodmer, H. (1989). Antigen recognition by class I-restricted T lymphocytes. Annu. Rev. Immunol. 7, 601-624. Udaka, K., Wiesmuller, K. H., Kienle, S., Jung, G., Tamamura, H., Yamagishi, H., Okumura, K., Walden, P., Suto, T., & Kawasaki, T. (2000). An automated prediction of MHC class I-binding peptides based on positional scanning with peptide libraries. Immunogenetics 51, 816-828. Vajda, S., Kataoka, R., DeLisi, C., Margalit, H., Berzofsky, J. A., & Comette, J. L. (1990). Molecular structure and vaccine design. Annu. Rev. Biophys. Biophys. Chem. 19, 69-82. Vasmatzis, G., Zhang, C., Comette, J. L., & DeLisi, C. (1996). Computational determination of side chain specificity for pockets in class I MHC molecules. Mol. Immunol. 33, 1231-1239. Wang, C. R., Castano, A. R., Peterson, P. A., Slaughter, C., Lindahl, K. F., & Deisenhofer, J. (1995). Nonclassical binding of formylated peptide in crystal stmcture of the MHC class lb molecule H2-M3. Cell 82, 655-664. Yewdell, J. W. & Bennink, J. R. (1992). Cell biology of antigen processing and presentation to major histocompatibility complex class I molecule-restricted T lymphocytes. [Review] [358 refs]. Advances in Immunology 52, 1-123. Yewdell, J. W. & Bennink, J. R. (1999). Immunodominance in MHC class I restricted T lymphocyte responses. Annu. Rev. Immunol. 17, 51-88. York, I. A. & Rock, K. L. (1996). Antigen processing and presentation by the class I major histocompatibility complex. Annu. Rev. Immunol. 14, 369-96. Young, A. C., Nathenson, S. G., & Sacchettini, J. C. (1995). Structural studies of class I major histocompatibility complex proteins: insights into antigen presentation. FASEB J. 9, 26-36. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 147 Yu, K., Petrovsky, N., Schonbach, C., Koh, J. Y., & Brusic, V. (2002). Methods for prediction of Peptide binding to MHC molecules: a comparative study. Mol. Med. 8, 137-148. Zen, J., Treutlein, H. R., & Rudy, G. B. (2001). Predicting sequences and structures of MHC-binding peptides: a computational combinatorial approach. J. Comput. Aided Mol. Des 15 , 573-586. Zhang, C., Anderson, A., & DeLisi, C. (1998). Structural principles that govern the peptide-binding motifs of class I MHC molecules. J. Mol. Biol. 281, 929-947. Zhang, W., Young, A. C. M., Imarai, M., Nathenson, S. G., & Sacchettini, J. C. (1992). Crystal structure of the major histocompatibility complex clas I H-2Kb molecule containing a single viral peptide: implications for peptide binding and T- cell receptor recognition. Proc. Natl. Acad. Sci. U. S. A 89, 8403-8407. Zhao, R., Loftus, D. J., Appella, E., & Collins, E. J. (1999). Structural evidence of T cell xeno-reactivity in the absence of molecular mimicry. J. Exp. Med. 189. Zinkemagel, R. M. & Doherty, P. C. (1974). Restriction of in vitro T cell-mediated cytotoxicity in lymphocytic choriomeningitis within a syngeneic or semiallogeneic system. Nature 248, 701-702. Zinkemagel, R. M. & Doherty, P. C. (1979). MHC restricted cytotoxic T cells: studies on the biological role of polymorphic major transplantation antigens in determining T cell restriction specificity, function and responsiveness. Advances in Immunology 27, 51-177. Zinkemagel, R. M. & Doherty, P. C. (1997). The discovery of MHC restriction. Immunol. Today 18,14-17. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 148 Appendix A Table A .l. Atom van der Waals radius Atom Radius (A) C 1.70 0 1.52 N 1.55 S 1.80 H 1.20 Table A.2. Amino acid parameters Residue Index Atom Z-matrix Bond(A) Angle (°) Torsion (°) ALA 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB1 4 2 0 1.090 109.50 60 6 HB2 4 2 0 1.090 109.50 180 7 HB3 4 2 0 1.090 109.50 300 8 C 2 0 -1 1.522 111.10 180 9 0 8 2 0 1.229 120.50 0 GLY 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA2 2 0 -1 1.090 109.50 60 4 HA3 2 0 -1 1.090 109.50 300 5 C 2 0 -1 1.522 110.40 180 6 0 5 2 0 1.229 120.50 0 SER 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 OG 4 2 0 1.430 109.47 180 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 149 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 8 HOG 7 4 2 0.960 109.47 180 9 C 2 0 -1 1.522 111.10 180 10 0 9 2 0 1.229 120.50 0 THR 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB 4 2 0 1.090 109.50 180 6 CG2 4 2 0 1.525 109.47 300 7 HG1 6 4 2 1.090 109.50 60 8 HG2 6 4 2 1.090 109.50 180 9 HG3 6 4 2 1.090 109.50 300 10 OG1 4 2 0 1.430 109.47 60 11 HOG 10 4 2 0.960 109.47 180 12 C 2 0 -1 1.522 111.10 180 13 0 12 2 0 1.229 120.50 0 LEU 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.525 109.47 180 8 HG 7 4 2 1.090 109.50 300 9 CD1 7 4 2 1.525 109.47 60 10 HD11 9 7 4 1.090 109.50 60 11 HD12 9 7 4 1.090 109.50 180 12 HD13 9 7 4 1.090 109.50 300 13 CD2 7 4 2 1.525 109.47 180 14 HD21 13 7 4 1.090 109.50 60 15 HD22 13 7 4 1.090 109.50 180 16 HD23 13 7 4 1.090 109.50 300 17 C 2 0 -1 1.522 111.10 180 18 0 17 2 0 1.229 120.50 0 ILE 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 150 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 109.47 60 5 HB 4 2 0 1.090 109.50 300 6 CG2 4 2 0 1.525 109.47 60 7 HG21 6 4 2 1.090 109.50 60 8 HG22 6 4 2 1.090 109.50 180 9 HG23 6 4 2 1.090 109.50 300 10 CGI 4 2 0 1.525 109.47 180 11 HG12 10 4 2 1.090 109.50 60 12 HG13 10 4 2 1.090 109.50 300 13 CD1 10 4 2 1.525 109.47 180 14 HD1 13 10 4 1.090 109.50 60 15 HD2 13 10 4 1.090 109.50 180 16 HD3 13 10 4 1.090 109.50 300 17 C 2 0 -1 1.522 111.10 180 18 0 17 2 0 1.229 120.50 0 VAL 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB 4 2 0 1.090 109.50 300 6 CGI 4 2 0 1.525 109.47 60 7 HG11 6 4 2 1.090 109.50 60 8 HG12 6 4 2 1.090 109.50 180 9 HG13 6 4 2 1.090 109.50 300 10 CG2 4 2 0 1.525 109.47 180 11 HG21 10 4 2 1.090 109.50 60 12 HG22 10 4 2 1.090 109.50 180 13 HG23 10 4 2 1.090 109.50 300 14 C 2 0 -1 1.522 111.10 180 15 0 14 2 0 1.229 120.50 0 ASN 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 151 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.522 111.10 180 8 OD1 7 4 2 1.229 120.50 0 9 ND2 7 4 2 1.335 116.60 180 10 HND1 9 7 4 1.010 119.80 0 11 HND2 9 7 4 1.010 119.80 180 12 C 2 0 -1 1.522 111.10 180 13 0 12 2 0 1.229 120.50 0 GLN 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.525 109.47 180 8 HG2 7 4 2 1.090 109.50 60 9 HG3 7 4 2 1.090 109.50 300 10 CD 7 4 2 1.522 111.10 180 11 OE1 10 7 4 1.229 120.50 0 12 NE2 10 7 4 1.335 116.60 180 13 HNE1 12 10 7 1.010 119.80 0 14 HNE2 12 10 7 1.010 119.80 180 15 C 2 0 -1 1.522 111.10 180 16 0 15 2 0 1.229 120.50 0 ARG 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.525 109.47 180 8 HG2 7 4 2 1.090 109.50 60 9 HG3 7 4 2 1.090 109.50 300 10 CD 7 4 2 1.525 109.47 180 11 HD2 10 7 4 1.090 109.50 60 12 HD3 10 7 4 1.090 109.50 300 13 NE 10 7 4 1.480 111.00 180 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 152 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 14 HNE 13 10 7 1.010 118.50 0 15 CZ 13 10 7 1.330 123.00 180 16 NH1 15 13 10 1.330 122.00 0 17 HN11 16 15 13 1.010 119.80 0 18 HN12 16 15 13 1.010 119.80 180 19 NH2 15 13 10 1.330 118.00 180 20 HN21 19 15 13 1.010 119.80 0 21 HN22 19 15 13 1.010 119.80 180 22 C 2 0 -1 1.522 111.10 180 23 0 22 2 0 1.229 120.50 0 HID 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 115.00 180 8 ND1 7 4 2 1.390 122.00 180 9 HND 8 7 4 1.010 126.00 0 10 CE1 8 7 4 1.320 108.00 180 11 HE 10 8 7 1.090 120.00 180 12 NE2 10 8 7 1.310 109.00 0 13 CD2 12 10 8 1.360 110.00 0 14 HD 13 12 10 1.090 120.00 180 15 C 2 0 -1 1.522 111.10 180 16 0 15 2 0 1.229 120.50 0 HIE 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 115.00 180 8 ND1 7 4 2 1.390 122.00 180 9 CE1 8 7 4 1.320 108.00 180 10 HE 9 8 7 1.090 120.00 180 11 NE2 9 8 7 1.310 109.00 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 153 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 12 HNE 11 9 8 1.010 125.00 180 13 CD2 11 9 8 1.360 110.00 0 14 HD 13 11 9 1.090 120.00 180 15 C 2 0 -1 1.522 111.10 180 16 0 15 2 0 1.229 120.50 0 HIP 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 115.00 180 8 ND1 7 4 2 1.390 122.00 180 9 HND 8 7 4 1.010 126.00 0 10 CE1 8 7 4 1.320 108.00 180 11 HE 10 8 7 1.090 120.00 180 12 NE2 10 8 7 1.310 109.00 0 13 HNE 12 10 8 1.010 125.00 180 14 CD2 12 10 8 1.360 110.00 0 15 HD 14 12 10 1.090 120.00 180 16 C 2 0 -1 1.522 111.10 180 17 0 16 2 0 1.229 120.50 0 TRP 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 115.00 180 8 CD1 7 4 2 1.340 127.00 180 9 HD 8 7 4 1.090 120.00 0 10 NE1 8 7 4 1.430 107.00 180 11 HNE 10 8 7 1.010 125.50 180 12 CE2 10 8 7 1.310 109.00 0 13 CZ2 12 10 8 1.400 128.00 180 14 HZ1 13 12 10 1.090 120.00 0 15 CH2 13 12 10 1.390 116.00 180 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 154 Table A.2. Continued Residue Index Atom Z-matrix Bond(A) Angle (°) Torsion (°) 16 HH 15 13 12 1.090 120.00 180 17 CZ3 15 13 12 1.350 121.00 0 18 HZ2 17 15 13 1.090 120.00 180 19 CE3 17 15 13 1.410 122.00 0 20 HE 19 17 15 1.090 120.00 180 21 CD2 19 17 15 1.400 117.00 0 22 C 2 0 -1 1.522 111.10 180 23 0 22 2 0 1.229 120.50 0 PHE 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 115.00 180 8 CD1 7 4 2 1.400 120.00 180 9 HD1 8 7 4 1.090 120.00 0 10 CE1 8 7 4 1.400 120.00 180 11 HE1 10 8 7 1.090 120.00 180 12 CZ 10 8 7 1.400 120.00 0 13 HZ 12 10 8 1.090 120.00 180 14 CE2 12 10 8 1.400 120.00 0 15 HE2 14 12 10 1.090 120.00 180 16 CD2 14 12 10 1.400 120.00 0 17 HD2 16 14 12 1.090 120.00 180 18 C 2 0 -1 1.522 111.10 180 19 0 18 2 0 1.229 120.50 0 TYR 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 109.47 180 8 CD1 7 4 2 1.400 120.00 180 9 HD1 8 7 4 1.090 120.00 0 10 CE1 8 7 4 1.400 120.00 180 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 155 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) GLU ASP LYS 11 HE1 10 8 7 1.090 120.00 180 12 CZ 10 8 7 1.400 120.00 0 13 OH 12 10 8 1.360 120.00 180 14 HOH 13 12 10 0.960 113.00 0 15 CE2 12 10 8 1.400 120.00 0 16 HE2 15 12 10 1.090 120.00 180 17 CD2 15 12 10 1.400 120.00 0 18 HD2 17 15 12 1.090 120.00 180 19 C 2 0 -1 1.522 111.10 180 20 0 19 2 0 1.229 120.50 0 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.510 109.47 180 8 HG2 7 4 2 1.090 109.50 60 9 HG3 7 4 2 1.090 109.50 300 10 CD 7 4 2 1.527 109.47 180 11 OE1 10 7 4 1.260 117.20 90 12 OE2 10 7 4 1.260 117.20 270 13 C 2 0 -1 1.522 111.10 180 14 0 13 2 0 1.229 120.50 0 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.527 109.47 180 8 OD1 7 4 2 1.260 117.20 90 9 OD2 7 4 2 1.260 117.20 270 10 C 2 0 -1 1.522 111.10 180 11 0 10 2 0 1.229 120.50 0 0 N -1 -2 -3 1.335 116.60 180 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 156 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.525 109.47 180 8 HG2 7 4 2 1.090 109.50 60 9 HG3 7 4 2 1.090 109.50 300 10 CD 7 4 2 1.525 109.47 180 11 HD2 10 7 4 1.090 109.50 60 12 HD3 10 7 4 1.090 109.50 300 13 CE 10 7 4 1.525 109.47 180 14 HE2 13 10 7 1.090 109.50 60 15 HE3 13 10 7 1.090 109.50 300 16 NZ 13 10 7 1.470 109.47 180 17 HNZ1 16 13 10 1.010 109.47 60 18 HNZ2 16 13 10 1.010 109.47 180 19 HNZ3 16 13 10 1.010 109.47 300 20 C 2 0 -1 1.522 111.10 180 21 0 20 2 0 1.229 120.50 0 PRO 0 N -1 -2 -3 1.337 117.00 180 1 CD 0 -1 -2 1.458 126.10 356.1 2 HD3 1 0 -1 1.090 109.50 80 3 HD2 1 0 -1 1.090 109.50 320 4 CG 1 0 -1 1.500 103.20 200.1 5 HG3 4 1 0 1.090 109.50 98 6 HG2 4 1 0 1.090 109.50 218 7 CB 4 1 0 1.510 106.00 338.3 8 HB3 7 4 1 1.090 109.50 136.3 9 HB2 7 4 1 1.090 109.50 256.3 10 CA 0 -1 -2 1.451 120.60 175.2 11 HA 10 0 -1 1.090 109.50 81.1 12 C 10 0 -1 1.522 111.10 0 13 0 12 10 0 1.229 120.50 0 CYS 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 157 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 SG 4 2 0 1.810 116.00 180 8 HSG 7 4 2 1.330 96.00 180 9 LP1 7 4 2 0.679 96.70 60 10 LP2 7 4 2 0.679 96.70 300 11 C 2 0 -1 1.522 111.10 180 12 0 11 2 0 1.229 120.50 0 CYX 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 SG 4 2 0 1.810 116.00 180 8 LP1 7 4 2 0.679 96.70 60 9 LP2 7 4 2 0.679 96.70 300 10 C 2 0 -1 1.522 111.10 180 11 0 10 2 0 1.229 120.50 0 MET 0 N -1 -2 -3 1.335 116.60 180 1 HN 0 -1 -2 1.010 119.80 0 2 CA 0 -1 -2 1.449 121.90 180 3 HA 2 0 -1 1.090 109.50 300 4 CB 2 0 -1 1.525 111.10 60 5 HB2 4 2 0 1.090 109.50 60 6 HB3 4 2 0 1.090 109.50 300 7 CG 4 2 0 1.525 109.47 180 8 HG2 7 4 2 1.090 109.50 60 9 HG3 7 4 2 1.090 109.50 300 10 SD 7 4 2 1.810 110.00 180 11 LP1 10 7 4 0.679 96.70 60 12 LP2 10 7 4 0.679 96.70 300 13 CE 10 7 4 1.780 100.00 180 14 HE1 13 10 7 1.090 109.50 60 15 HE2 13 10 7 1.090 109.50 180 16 HE3 13 10 7 1.090 109.50 300 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 158 Table A.2. Continued Residue Index Atom Z-matrix Bond (A) Angle (°) Torsion (°) 17 C 2 0 - 1 1.522 111.10 180 18 0 17 2 0 1.229 120.50 0 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 159 <2.5 A >90°' >90° >90 < 3.9 A Figure A.l. Hydrogen bond geometry. D is a hydrogen bond donor, H is a donor hydrogen, A is a hydrogen acceptor, and P is the precedent atom of A. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Appendix B 160 L T ! U. } o o L U I Q . O S s m 5 o eel 3 2 < 0 q 8 * “= < 1 l y *319 *52! '3 8 *38 g D 6 o o i o CSi O r ^ . CD CO CD O CO CO CD CD LD CD CD r-- LD O r*^ in CD CD CD CD ( O O O C D O O O C D C LD OJ i s s o ° w □ m □ ^ ffj o ^1 - LD C D C\J T~ 1 i —- r-- i CO D" CD LD m O D“ CO LD CO CO CD LD CD CO T“ “ CD oo CO CD c\i CD r-- LD CO LD T * “ “ LD 1 CD CsJ m co o o < < O O O O O O O O c e g a : > LU d > - UJ _ i l u CC ■ 0 £ S £ £ ! I — OJ C O C O csj co t m cd co ( t 1 1 1 S D D D C X X X e R > > * 9 ( E J j : u LI!. Eli J S C J jb 3 © 3 > < S b 13 3* S 3 = i o ^ <3 " 3 3 E * “ m U 3 U EC % H. 31 c e J J u o 3 " P t I $ -3 CC •o "O < o CQ u - r* i 3 1X5- Q S < £ © 1 6 3 L_ X > i s J .£ .£ .£ 3 E 2 2 3 D o CD CO o cn OD CD o o CD T— o V T ■ f r 1 ) 0 C D CC R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure B.l. WATGEN program interface. 161 Help Options Compute Peptide Sequence: JllSAWGIlj Peptide ResName: IL E IL E SER ALA VAL VAL G L V IL E LEU Number of Conformations: j Max Score: -1.01 xl 1 0 0 Number of Peptide Backbone Conformations; | 28792 N umber of Peptide Conformations: j 287918 Nurtber of MHC Conformations: J 10 Number of Peptide-MHC Conformations: | 284935 Number of Peptide-MHC-Wat Conformations: J 147 2355 Figure B.2. MHC program interface. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 162 Appendix C MHC Program Output Pepide Sequence: IISAWGIL Statistics: # Protein Conf: 10 # Pep BB Conf: 28792 # Pep SC Conf: 287918 # Pep-Pro Conf: 284935 # Pep-Pro-Wat Conf: 147 Max Score: 23 55 Min Score: 1970 # Output Conf: 100 Output: No. File Name Score nWat VDW Hbond Hpro Hpep Hwat MHbond 1 01_I ISAWGIL. pdb 2355 75 33 252 41 25 33 4 2 02_IISAWGIL.pdb 2330 73 34 250 37 28 31 4 3 0 3_I ISAWGIL . pdb 2330 70 30 248 37 26 31 5 4 04_I I SAWGIL. pdb 2290 73 34 246 37 26 29 5 5 05_I ISAWGIL. pdb 2290 72 30 244 33 27 27 8 6 06_IISAWGIL. pdb 2290 70 30 244 38 25 25 7 7 07_IISAWGIL. pdb 2280 64 30 243 36 24 26 8 8 0 8_I ISAWGIL. pdb 2280 73 28 242 32 26 31 5 9 0 9_I ISAWGIL. pdb 2275 73 27 241 37 26 24 6 10 10_IISAWGIL. pdb 2270 72 38 246 40 27 25 4 11 11_I I SAWGIL. pdb 2270 72 38 246 40 27 25 4 12 12_I ISAWGIL. pdb 2255 70 33 242 34 29 22 8 13 13_IISAWGIL .pdb 2255 75 33 242 38 26 25 4 14 14_IISAWGIL .pdb 2255 71 33 242 35 27 24 7 15 15_IISAWGIL .pdb 2250 71 28 239 33 23 30 4 16 16_IISAWGIL .pdb 2245 70 31 240 36 21 29 5 17 17 I ISAWGIL. pdb 2240 70 32 240 41 19 27 4 18 18_I ISAWGIL. pdb 2240 70 36 242 36 28 26 4 19 19_IISAWGIL. pdb 2240 70 28 238 35 22 27 5 20 20_IISAWGIL.pdb 2240 70 28 238 36 25 22 6 21 21_IISAWGIL .pdb 2235 74 31 239 35 20 27 8 22 2 2_I I SAWGIL. pdb 2235 69 31 239 37 22 23 9 23 23_IISAWGIL.pdb 2230 64 26 236 30 24 27 7 24 2 4_I ISAWGIL. pdb 2230 68 30 238 30 25 25 9 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 163 MHC Program Output Continued 25 2 5_I ISAWGIL. pdb 2230 68 26 2 6_"iISAWGIL .pdb 2225 69 27 2 7_IISAWGIL. pdb 2225 66 28 2 8_I ISAWGIL. pdb 2225 68 29 2 9_11 SAWGIL. pdb 2225 71 30 3 0_IISAWGIL. pdb 2220 66 31 3 i_11 SAWGIL. pdb 2220 67 32 32_I ISAWGIL. pdb 2220 66 33 33_I ISAWGIL. pdb 2220 67 34 34_I ISAWGIL. pdb 2220 68 35 3 5_I ISAWGIL. pdb 2215 67 36 3 6_I ISAWGIL. pdb 2215 64 37 37_IISAWGIL. pdb 2215 70 38 3 8_IISAWGIL. pdb 2215 73 39 3 9_IISAWGIL. pdb 2215 67 40 4 0_IISAWGIL. pdb 2210 71 41 4 i_IISAWGIL. pdb 2210 74 42 42_I ISAWGIL. pdb 2210 72 43 43_IISAWGIL. pdb 2210 64 44 44_IISAWGIL. pdb 2205 66 45 45_11 SAWGIL. pdb 2205 69 46 46_I ISAWGIL. pdb 2205 66 47 47_I ISAWGIL. pdb 2205 72 48 4 8_IISAWGIL. pdb 2205 71 49 4 9_'iISAWGIL. pdb 2200 63 50 5 0_I ISAWGIL. pdb 2200 63 51 51_11 SAWGIL. pdb 2200 64 52 52_11 SAWGIL. pdb 2195 66 53 53_I ISAWGIL. pdb 2195 67 54 54_IISAWGIL. pdb 2195 66 55 55_11 SAWGIL. pdb 2195 74 56 56_11 SAWGIL. pdb 2195 64 57 57_'iISAWGIL. pdb 2195 67 58 5 8_"i I SAWGIL. pdb 2190 73 59 59_IISAWGIL. pdb 2190 66 60 60_I ISAWGIL. pdb 2185 64 61 6i_I ISAWGIL. pdb 2185 67 62 62_I ISAWGIL. pdb 2180 63 63 63_IISAWGIL. pdb 2180 70 64 64_11 SAWGIL. pdb 2180 62 65 65_11 SAWGIL. pdb 2180 64 66 6 6_ I ISAWGIL. pdb 2180 67 67 67_I ISAWGIL. pdb 2175 66 68 6 8_I ISAWGIL. pdb 2175 67 69 69_11 SAWGIL. pdb 2175 67 70 7 0_'iISAWGIL. pdb 2175 65 71 7i_IISAWGIL. pdb 2170 67 72 72_IISAWGIL. pdb 2170 72 73 7 3_11 SAWGIL. pdb 2170 69 34 240 31 25 29 6 29 237 33 23 26 6 27 236 30 24 25 9 29 237 33 25 23 7 29 237 33 19 30 7 32 238 38 22 26 4 30 237 36 22 21 9 26 235 29 23 28 6 32 238 37 23 22 7 32 238 36 24 24 5 29 236 32 24 27 5 29 236 32 24 23 9 29 236 36 24 23 4 31 237 31 28 23 6 29 236 32 24 27 5 36 239 35 24 27 4 30 236 37 20 24 6 30 236 40 25 19 4 30 236 39 21 22 5 25 233 30 24 25 6 31 236 34 19 27 6 29 235 33 20 25 8 31 236 35 23 24 5 25 233 34 22 22 7 26 233 27 25 26 7 28 234 29 21 26 9 34 237 32 26 24 6 25 232 33 21 24 6 33 236 39 19 23 6 25 232 29 19 30 5 33 236 33 24 25 5 27 233 35 22 23 5 31 235 34 22 24 6 24 231 32 25 21 5 30 234 31 27 21 6 33 235 34 24 24 4 31 234 35 17 25 7 32 234 33 24 23 5 26 231 37 20 20 6 28 232 29 25 24 6 24 230 31 25 19 7 34 235 35 24 20 6 27 231 34 22 19 8 27 231 28 24 26 5 37 236 40 21 20 6 29 232 29 24 25 6 32 233 34 22 23 5 30 232 29 22 26 7 34 234 38 21 21 5 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 164 MHC Program Output Continued 74 74_IISAWGIL. pdb 2170 74 36 235 40 20 22 75 75_IISAWGIL. pdb 2170 67 30 232 32 24 21 76 76_IISAWGIL. pdb 2165 67 29 231 35 25 15 77 7 7_I ISAWGIL. pdb 2165 65 33 233 32 22 23 78 7 8_I ISAWGIL. pdb 2165 65 33 233 32 22 23 79 7 9_I ISAWGIL. pdb 2165 64 31 232 29 22 26 80 8 0_I ISAWGIL. pdb 2165 66 27 230 33 21 23 81 81_I ISAWGIL. pdb 2165 71 31 232 34 21 19 82 82_IISAWGIL.pdb 2160 64 26 229 35 20 21 83 83_IISAWGIL.pdb 2160 65 26 229 28 24 23 84 8 4_I I SAWGIL. pdb 2160 64 24 228 31 22 21 85 85_IISAWGIL. pdb 2160 68 30 231 31 27 20 86 86_IISAWGIL .pdb 2155 65 31 231 33 19 23 87 8 7_I ISAWGIL. pdb 2155 72 33 232 29 22 28 88 8 8_I ISAWGIL. pdb 2155 65 31 231 33 19 23 89 8 9_I ISAWGIL. pdb 2155 71 33 232 33 25 21 90 9 0_I I SAWGIL. pdb 2150 67 30 230 37 23 17 91 91_I I SAWGIL. pdb 2150 71 26 228 27 22 22 92 92_IISAWGIL. pdb 2150 68 32 231 27 26 24 93 93_IISAWGIL .pdb 2150 67 34 232 35 18 22 94 94_I ISAWGIL. pdb 2150 63 26 228 31 21 22 95 95_IISAWGIL. pdb 2150 64 26 228 31 24 19 96 96_IISAWGIL. pdb 2145 64 31 230 35 18 23 97 97_IISAWGIL. pdb 2145 64 37 233 33 25 20 98 98_IISAWGIL. pdb 2145 66 29 229 28 23 23 99 9 9_I I SAWGIL. pdb 2145 58 43 236 39 23 22 100 100_IISAWGIL . pdb 2145 68 33 231 34 21 23 Notes: nWat Number of added H20 VDW VDW contact energy score Hbond Number of hydrogen bonds Score 10*Hbond - 5*VDW HPro Number of hydrogen bonds formed with the MHC protein HPep Number of hydrogen bonds formed with the Peptide HWat Number of hydrogen bonds formed between water molecules MHbond Number of hydrogen bonds formed between the MHC protein and peptide 4 6 7 6 6 6 5 9 5 5 6 4 7 4 7 4 4 9 6 7 6 5 5 6 6 3 4 the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Computer prediction of peptide binding to class I MHC molecules
PDF
Discovery of novel HIV-1 integrase inhibitors
PDF
Design, synthesis, biological testing and QSAR analysis of new Schiff bases of N-hydroxysemicarbazide as inhibitors of tumor cells
PDF
Conjunctival epithelial uptake of biodegradable nanoparticles: Mechanism, intracellular distribution, and absorption enhancement
PDF
DNA structures associated with the Fragile X triplet repeat sequences
PDF
Computational analysis and modeling of protein -RNA complexes and nucleic acid structure
PDF
Cytochrome P450 3A metabolism in the rabbit lacrimal gland and conjunctiva
PDF
Basolateral transport of nucleosides in rabbit tracheal epithelial cells: An equilibrative nucleoside transporter outcome
PDF
Expression and regulation of B7 on antigen presenting cells
PDF
Endogenous regulatory factors in the inhibition and down-regulation of immune responses
PDF
Dissecting the role of the catalytic core domain of the human immunodeficiency virus integrase and its role in binding to the viral DNA substrate
PDF
Drug transport mechanisms in the rabbit corneal epithelium: In vitro-in vivo correlation
PDF
Assessing the cost implications of combined pharmacotherapy in the long term management of asthma: Theory and application of methods to control selection bias
PDF
Drug efflux pumps in rabbit conjunctival epithelial cells
PDF
Antiviral activities and cytotoxicities of heterocyclic schiff bases on aminohydroxyguanidine
PDF
High mobility group A2 (HMGA2): Molecular dissections of its functions and regulation
PDF
Pharmacokinetic And Cytotoxicity Studies Of Oxazaphophorine Metabolites
PDF
Drug delivery to the posterior region of the eye: Characterization of organic cation transporters
PDF
A methodology to identify high -risk patients with diabetes in the California Medicaid populations (Medi -Cal)
PDF
A new paradigm to evaluate quality-adjusted life years (QALY) from secondary database: Transforming health status instrument scores to health preference
Asset Metadata
Creator
Bui, Huynh-Hoa Thi
(author)
Core Title
Computational prediction of peptide-MHC class I binding interactions
School
Graduate School
Degree
Doctor of Philosophy
Degree Program
Pharmaceutical Sciences
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Chemistry, pharmaceutical,health sciences, immunology,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Haworth, Ian S. (
committee chair
), Bolger, Michael B. (
committee member
), Lien, Eric J. (
committee member
), Peng, Ching-An (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-357262
Unique identifier
UC11334804
Identifier
3103862.pdf (filename),usctheses-c16-357262 (legacy record id)
Legacy Identifier
3103862.pdf
Dmrecord
357262
Document Type
Dissertation
Rights
Bui, Huynh-Hoa Thi
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
health sciences, immunology