Method, computing routine, device for predicting properties of MHC/peptide complexes, and data and peptides produced therefrom.

Abstract

The present invention relates to a method for structure-based prediction of properties of peptides and peptide analogs in complex with major histocompatibility (MHC) class I and class II molecules. The said properties mainly relate to the three-dimensional structure of an MHC/peptide complex and the binding affinity of a peptide for an MHC receptor. The invention further relates to a computer program and a device therefor. The invention further relates to data produced by a method of the invention. The invention further relates to peptides and peptide analogs predicted to bind to target-MHC molecules. The present invention thus relates to the field of immunology, with possible applications in manufacture of vaccinates, de-immunization of proteins, and manufacture of therapeutic agents, especially immunotherapeutic agents.

1 - 15 . (canceled) 
   
   
       16 . A method for producing an immunogenic peptide comprising an MHC class I or class II restricted T cell epitope that binds to an MHC class I or class II molecule and induces an MHC class I or II-restricted cytotoxic T cell response, said method comprising the steps of:
 (a) providing an amino acid sequence of a polypeptide of interest;   (b) preparing one or more overlapping putative immunogenic peptide fragments of said polypeptide of interest;   (c) receiving a representation of a complete or partial three-dimensional structure of said MHC class I or class II molecule;   (d) obtaining an ensemble of conformational representations of peptide backbone structures of said putative immunogenic peptides, said conformational representations located within the binding site of said MHC molecule;   (e) modeling the side chains of said peptide backbone structures of said ensemble in relation to said MHC molecule, thereby obtaining an ensemble of modeled MHC/peptide complexes;   (f) evaluating the binding properties of said putative immunogenic peptides for said MHC molecule, by using a scoring function which combines at least:
 f1) average binding energy component obtained by evaluating one or more components of the potential energy of each complex of the ensemble of step (e); 
 f2) conformational entropy component obtained by evaluating the conformational entropy for the complete ensemble of each modeled MHC/peptide complex of step (e); 
   (g) inferring from the results obtained in (f), one or more putative immunogenic peptides that bind to said MHC molecule;   (h) optionally preparing one or more of said putative immunogenic peptides of said polypeptide of interest;   (i) optionally testing complexes of said one or more putative immunogenic peptides said MHC molecule for an ability to be recognized by a MHC cytotoxic T cells, and to thereby induce a cytotoxic T cell response to the epitope; and   (e) selecting said one or more putative immunogenic fragments comprising an MHC class I or class II binding site that induce an MHC class I or class II cytotoxic T cell response to the epitope.   
   
   
       17 . The method according to  claim 16  wherein said conformational representation of step (c) is obtained from one of the following:
 one or more experimentally determined structures obtained by for example X-ray crystallography, nuclear magnetic resonance spectroscopy, scanning microscopy; or   one or more models derived from an experimentally determined structure, whereby said experimentally determined structure has a high sequence identity to said MHC molecule.   
   
   
       18 . The method according to  claim 16  wherein said conformational representation of step (d) is generated by a computer modeling method, said method being able to generate multiple energetically favorable backbone configurations in relation to said MHC molecule. 
   
   
       19 . The method according to  claim 16  wherein said conformational representation of step (d) is retrieved from a library of peptide structures pre-oriented in relation to said MHC molecule. 
   
   
       20 . The method according to  claim 16  wherein a complex within said ensemble of step (e) is obtained from a side-chain placement algorithm. 
   
   
       21 . The method according to  claim 16  wherein the side-chain placement of step (e) not only involves placing the side-chains of the peptide itself, but also involves placing at least one side-chain of said MHC molecule that are in contact with said peptide. 
   
   
       22 . The method according to  claim 16  wherein a complex within said ensemble of step (e) is obtained from a side-chain placement algorithm suited for global side-chain optimization. 
   
   
       23 . The method according to  claim 20  wherein the side-chain placement algorithm is a dead-end elimination (DEE) algorithm, characterized in that said DEE algorithm eliminates rotameric conformations on the basis of a mathematical criterion that allows the detection of conformations that are not compatible with the globally optimal conformation. 
   
   
       24 . The method according to  claim 20  wherein the side-chain placement algorithm is a FASTER algorithm, said algorithm being characterized by a repeated perturbation, relaxation and evaluation step. 
   
   
       25 . The method according to  claim 16  wherein the binding affinity of step (f) is represented by a single scoring value for the whole ensemble of MHC/peptide complexes, said scoring value comprising the sum of the conformational entropy for the complete ensemble of MHC/peptide complexes, and the average of the said energetical components of each of the complexes of said ensemble. 
   
   
       26 . The method according to  claim 16  wherein the binding affinity of step (f) is evaluated for the global complex, thereby accounting for interactions between pairs of residues from the peptide, the MHC molecule and both the peptide and the MHC molecule. 
   
   
       27 . The method according to  claim 16  wherein the entropical component reflects the overall conformational flexibility of the peptide. 
   
   
       28 . The method according to  claim 16  wherein the conformational representations of said peptide contained in said library are derived from experimentally determined structures. 
   
   
       29 . The method according to  claim 18  wherein the conformational representations of said peptide contained in said library are derived from computer-generated structures, said structures generated by said computer modeling method. 
   
   
       30 . The method according to  claim 16  wherein said peptide comprises one or more non-naturally occurring amino acids. 
   
   
       31 . The method according to  claim 16  wherein said MHC class I molecule comprises an HLA antigen selected from any of the HLA-A, HLA-B, HLA-C. HLA-E, HLA-F and HLA-G alleles. 
   
   
       32 . The method according to  claim 16  wherein said MHC class II molecule comprises an HLA antigen selected from any of the HLA.-DR, HLA-DQ and HLA-DP gene products. 
   
   
       33 . Data comprising
 conformational representations of one or more peptide backbone structures, each peptide demonstrating an interaction with an MHC class I or class II molecule, and   an indication of the MHC molecule associated with said conformational representation.   
   
   
       34 . A computer program comprising computing routines, stored on a computer readable medium for evaluating the binding affinity of a peptide for an MHC class I or class II molecule, said routines comprising:
 receiving an ensemble of conformational representations of structures of the complex between said MHC molecule and said peptide,   evaluating one or more components of the potential energy of each complex of the ensemble,   evaluating the conformational entropy for the complete ensemble.   
   
   
       35 . The computer program according to  claim 34  further comprising modeling for each peptide backbone structure of said ensemble in relation to said MHC molecule, at least the side-chains of said peptide. 
   
   
       36 . The computer program according to  claim 34  wherein said peptide backbone structures are obtained by computer modeling or by retrieval from a database. 
   
   
       37 . A device for evaluating the binding affinity of a peptide for an MHC class I or class II molecule, comprising:
 receiving an ensemble of conformational representations of structures of the complex between said MHC molecule and said peptide,   evaluating one or more components of the potential energy of each complex of the ensemble,   evaluating the conformational entropy for the complete ensemble.   
   
   
       38 . A peptide which binds MHC class I or class II molecules, said peptide being obtainable by using the method of  claim 16 . 
   
   
       39 . (canceled) 
   
   
       40 . A nucleic acid encoding a peptide as defined in  claim 38 . 
   
   
       41 . A nucleic acid of at least 15 nucleotides in length specifically hybridizing with the nucleic acid of  claim 40 . 
   
   
       42 . An antibody specifically recognizing a peptide according to  claim 38 . 
   
   
       43 . An antibody specifically recognizing a nucleic acid according to  claim 40 . 
   
   
       44 . A method for producing a peptide according to  claim 38  comprising:
 (a) culturing host cells comprising a nucleic acid encoding the peptide under conditions allowing the expression of the peptide, and,   (b) recovering the produced peptide from the culture.   
   
   
       45 . The peptide according to  claim 38  for use as a medicament. 
   
   
       46 . The nucleic acid according to  claim 40  for use as a medicament.