Efficient discovery of immune response targets by cyclical refinement of QSAR models of peptide binding

Peptides that induce and recall T-cell responses are called T-cell epitopes. T-cell epitopes may be useful in a subunit vaccine against malaria. Computer models that simulate peptide binding to MHC are useful for selecting candidate T-cell epitopes since they minimize the number of experiments required for their identification. We applied a combination of computational and immunological strategies to select candidate T-cell epitopes. A total of 86 experimental binding assays were performed in three rounds of identification of HLA-All binding peptides from the six preerythrocytic malaria antigens. Thirty-six peptides were experimentally confirmed as binders. We show that the cyclical refinement of the ANN models results in a significant improvement of the efficiency of identifying potential T-cell epitopes. (C) 2001 by Elsevier Science Inc.

Methods for prediction of peptide binding to MHC molecules: A comparative study

Background: A variety of methods for prediction of peptide binding to major histocompatibility complex (MHC) have been proposed. These methods are based on binding motifs, binding matrices, hidden Markov models (HMM), or artificial neural networks (ANN). There has been little prior work on the comparative analysis of these methods. Materials and Methods: We performed a comparison of the performance of six methods applied to the prediction of two human MHC class I molecules, including binding matrices and motifs, ANNs, and HMMs. Results: 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). When little or no peptide data are available, binding motifs are the most useful alternative to random guessing or use of a complete overlapping set of peptides for selection of candidate binders. As the number of known peptide binders increases, binding matrices and HMM become more useful predictors. ANN and HMM are the predictive methods of choice for MHC alleles with more than 100 known binding peptides. Conclusion: The ability of bioinformatic methods to reliably predict MHC binding peptides, and thereby potential T-cell epitopes, has major implications for clinical immunology, particularly in the area of vaccine design.

PREDBALB/c: a system for the prediction of peptide binding to H2(d) molecules, a haplotype of the BALB/c mouse

PREDBALB/c is a computational system that predicts peptides binding to the major histocompatibility complex-2 (H2(d)) of the BALB/c mouse, an important laboratory model organism. The predictions include the complete set of H2(d) class I ( H2-K-d, H2-L-d and H2-D-d) and class II (I-E-d and I-A(d)) molecules. The prediction system utilizes quantitative matrices, which were rigorously validated using experimentally determined binders and non-binders and also by in vivo studies using viral proteins. The prediction performance of PREDBALB/c is of very high accuracy. To our knowledge, this is the first online server for the prediction of peptides binding to a complete set of major histocompatibility complex molecules in a model organism (H2(d) haplotype). PREDBALB/c is available at http://antigen.i2r.a-star.edu.sg/predBalbc/.

A peptide-binding motif for I-A(g7), the class II major histocompatibility complex (MHC) molecule of NOD and Biozzi AB/H mice

The class II major histocompatibility complex molecule I-A(g7) is strongly linked to the development of spontaneous insulin-dependent diabetes mellitus (IDDM) in non obese diabetic mice and to the induction of experimental allergic encephalomyelitis in Biozzi AB/H mice. Structurally, it resembles the HLA-DQ molecules associated with human IDDM, in having a non-Asp residue at position 57 in its beta chain. To identify the requirements for peptide binding to I-A(g7) and thereby potentially pathogenic T cell epitopes, we analyzed a known I-A(g7)-restricted T cell epitope, hen egg white lysozyme (HEL) amino acids 9-27. NH2- and COOH-terminal truncations demonstrated that the minimal epitope for activation of the T cell hybridoma 2D12.1 was M12-R21 and the minimum sequence for direct binding to purified I-A(g7) M12-Y20/K13-R21. Alanine (A) scanning revealed two primary anchors for binding at relative positions (p) 6 (L) and 9 (Y) in the HEL epitope. The critical role of both anchors was demonstrated by incorporating L and Y in poly(A) backbones at the same relative positions as in the HEL epitope. Well-tolerated, weakly tolerated, and nontolerated residues were identified by analyzing the binding of peptides containing multiple substitutions at individual positions. Optimally, p6 was a large, hydrophobic residue (L, I, V, M), whereas p9 was aromatic and hydrophobic (Y or F) or positively charged (K, R). Specific residues were not tolerated at these and some other positions. A motif for binding to I-A(g7) deduced from analysis of the model HEL epitope was present in 27/30 (90%) of peptides reported to be I-A(g7)-restricted T cell epitopes or eluted from I-A(g7). Scanning a set of overlapping peptides encompassing human proinsulin revealed the motif in 6/6 good binders (sensitivity = 100%) and 4/13 weak or non-binders (specificity = 70%). This motif should facilitate identification of autoantigenic epitopes relevant to the pathogenesis and immunotherapy of IDDM.

The uncharged surface features surrounding the active site of Escherichia coli DsbA are conserved and are implicated in peptide binding

DsbA is a protein-folding catalyst from the periplasm of Escherichia coli that interacts with newly translocated polypeptide substrate and catalyzes the formation of disulfide bonds in these secreted proteins. The precise nature of the interaction between DsbA and unfolded substrate is not known. Here, we give a detailed analysis of the DsbA crystal structure, now refined to 1.7 Angstrom, and present a proposal for its interaction with peptide. The crystal structure of DsbA implies flexibility between the thioredoxin and helical domains that may be an important feature for the disulfide transfer reaction. A hinge point for domain motion is identified-the typo IV beta-turn Phe 63-Met 64-Gly 65-Gly 66, which connects the two domains. Three unique features on the active site surface of the DsbA molecule-a groove, hydrophobic pocket, and hydrophobic patch-form an extensive uncharged surface surrounding the active-sits disulfide. Residues that contribute to these surface features are shown to be generally conserved in eight DsbA homologues. Furthermore, the residues immediately surrounding the active-site disulfide are uncharged in all nine DsbA proteins. A model for DsbA-peptide interaction has been derived from the structure of a human thioredoxin:peptide complex. This shows that peptide could interact with DsbA in a manner similar to that with thioredoxin. The active-site disulfide and all three surrounding uncharged surface features of DsbA could, in principle, participate in the binding or stabilization of peptide.

Extreme learning machine for predicting HLA-peptide binding

Machine learning techniques have been recognized as powerful tools for learning from data. One of the most popular learning techniques, the Back-Propagation (BP) Artificial Neural Networks, can be used as a computer model to predict peptides binding to the Human Leukocyte Antigens (HLA). The major advantage of computational screening is that it reduces the number of wet-lab experiments that need to be performed, significantly reducing the cost and time. A recently developed method, Extreme Learning Machine (ELM), which has superior properties over BP has been investigated to accomplish such tasks. In our work, we found that the ELM is as good as, if not better than, the BP in term of time complexity, accuracy deviations across experiments, and most importantly - prevention from over-fitting for prediction of peptide binding to HLA.

Prediction of MHC class II-binding peptides using an evolutionary algorithm and artificial neural network

Motivation: Prediction methods for identifying binding peptides could minimize the number of peptides required to be synthesized and assayed, and thereby facilitate the identification of potential T-cell epitopes. We developed a bioinformatic method for the prediction of peptide binding to MHC class II molecules. Results: Experimental binding data and expert knowledge of anchor positions and binding motifs were combined with an evolutionary algorithm (EA) and an artificial neural network (ANN): binding data extraction --> peptide alignment --> ANN training and classification. This method, termed PERUN, was implemented for the prediction of peptides that bind to HLA-DR4(B1*0401). The respective positive predictive values of PERUN predictions of high-, moderate-, low- and zero-affinity binder-a were assessed as 0.8, 0.7, 0.5 and 0.8 by cross-validation, and 1.0, 0.8, 0.3 and 0.7 by experimental binding. This illustrates the synergy between experimentation and computer modeling, and its application to the identification of potential immunotheraaeutic peptides.

Conformations of cyclic octapeptides and the influence of heterocyclic ring constraints upon calcium binding

A comparison is made between the structures and calcium binding properties of four cyclic octapeptides that differ in the number of heterocyclic thiazole and oxazoline ring constraints. The conformations of the naturally occurring cyclic octapeptides ascidiacyclamide 1 and patellamide D 2, which each contain two oxazoline and two thiazole rings, are compared by H-1 NMR spectroscopy with the analogues cyclo(Thr-D-Val(Thz)-Ile)(2) 3 with just two thiazoles, and cyclo(Thr-D-Val-alpha Abu-Ile)(2) 4, with no 5-membered rings. The conformations observed in the solid state for ascidiacyclamide (saddle) and patellamide D (twisted figure of eight) were retained in solution, whilst peptide 3 was found to have a chair shape and peptide 4 displayed a range of conformations. The solid state structure of 4 revealed that the peptide takes a relatively planar conformation with a number of transannular hydrogen bonds, which are apparently retained in solution. Complexation studies utilising H-1 NMR and CD spectroscopy yielded 1∶1 calcium-peptide binding constants (log K) for the four peptides (2.9 (1), 2.8 (2), 4.0 (3) and 5.5 (4)) as well as a 1 : 2 metal-peptide binding constant for 3 (log K = 4.5). The affinity for Ca2+ thus decreases with increasing number of 5-membered ring constraints in the macrocycle (4 > 3 > 2 approximate to 1).

Cyclic octapeptides containing thiazole. Effect of stereochemistry and degree of flexibility on calcium binding properties

Solution conformation and calcium binding properties have been investigated for the two cyclic octapeptides cyclo(-D-Thr-D-Val(Thz)-Ile-)(2) (4) and cyclo(-Thr-Gly(Thz)-Ile-Ser-Gly(Thz)-Ile-)(5) and the results are compared to those for the cyclic octapeptides previously studied; ascidiacyclamide (1), patellamide D (2), cyclo(-Thr-D-Val(Thz)-Ile-)(2) (3), and cyclo(-Thr-D-Val-alphaAbu-Ile-)2 (6). Both 4 and 5 contain two heterocyclic thiazole ring constraints but the latter has a larger degree of flexibility as a consequence of the glycine residues within the cyclic framework. The solution conformation of 4 and 5 was determined from H-1 NMR spectra and found to be a twisted figure of eight similar to that for 2. Complexation studies using H-1 NMR and CD spectroscopy yielded 1 : 1 calcium-peptide binding constants (logK) for the two peptides (2.3 (4) and 5.7 (5)). For 5 the magnitude of the binding constant was verified by a competition titration using CD. The different calcium-binding affinities of 3 (logK = 4.0) and 4 is attributed to the stereochemistry of the threonine residue. The magnitude of the binding constant for 5 compared to 3 and 4 (all peptides containing two thiazole ring constrains) demonstrates that the increase in flexibility of the cyclic peptide has a dramatic effect on the Ca2+ binding ability. The affinity for Ca2+ thus decreases in the order (6 similar to 5 > 3 > 2 similar to 1 > 4). The number of carbonyl donors available on each peptide has only a limited effect on calcium binding. The most important factor is the flexibility, which allows for a conformation of the peptide capable of binding calcium efficiently.

Potent T cell response to a class I-binding 13-mer viral epitope and the influence of HLA micropolymorphism in controlling epitope length

The BZLF1 antigen of Epstein-Barr virus includes three overlapping sequences of different lengths that conform to the binding motif of human leukocyte antigen (HLA) B*3501. These 9-mer ((56)LPOGQLTAy(64)), 11-mer ((54)EPLPQGQLTAy(64)), and 13-mer ((52)LPEPLPQGQLTAY(64)) peptides all bound well to B*3501; however, the CTL response in individuals expressing this HILA allele was directed strongly and exclusively towards the 11-mer peptide. In contrast, EBV-exposed donors expressing HLA B*3503 showed no significant CTL response to these peptides because the single amino acid difference between B*3501 and B*3503 within the F pocket inhibited HLA binding by these peptides. The extraordinarily long 13-mer peptide was the target for the CTL response in individuals expressing B*3508, which differs from B*3501 at a single position within the D pocket (B*3501, 156 Leucine; B*3508, 156 Arginine). This minor difference was shown to enhance binding of the 13-mer peptide, presumably through a stabilizing interaction between the negatively charged glutamate at position 3 of the peptide and the positively charged arginine at HLA position 156. The 13-mer epitope defined in this study represents the longest class I-binding viral epitope identified to date as a minimal determinant. Furthermore, the potency of the response indicates that peptides of this length do not present a major structural barrier to CTL recognition.

Inhibition of protein interactions with the beta(2) sliding clamp of Escherichia coli DNA polymerase III by peptides from beta(2)-binding proteins

The sliding clamp of the Escherichia coli replisome is now understood to interact with many proteins involved in DNA synthesis and repair. A universal interaction motif is proposed to be one mechanism by which those proteins bind the E. coli sliding clamp, a homodimer of the beta subunit, at a single site on the dimer. The numerous beta(2)-binding proteins have various versions of the consensus interaction motif, including a related hexameric sequence. To determine if the variants of the motif could contribute to the competition of the beta-binding proteins for the beta(2) site, synthetic peptides derived from the putative beta(2)-binding motifs were assessed for their abilities to inhibit protein-beta(2) interactions, to bind directly to beta(2), and to inhibit DNA synthesis in vitro. A hierarchy emerged, which was consistent with sequence similarity to the pentameric consensus motif, QL(S/D)LF, and peptides containing proposed hexameric motifs were shown to have activities comparable to those containing the consensus sequence. The hierarchy of peptide binding may be indicative of a competitive hierarchy for the binding of proteins to beta(2) in various stages or circumstances of DNA replication and repair.

Deriving matrix of peptide-MHC interactions in diabetic mouse by genetic algorithm

Finding motifs that can elucidate rules that govern peptide binding to medically important receptors is important for screening targets for drugs and vaccines. This paper focuses on elucidation of peptide binding to I-A(g7) molecule of the non-obese diabetic (NOD) mouse - an animal model for insulin-dependent diabetes mellitus (IDDM). A number of proposed motifs that describe peptide binding to I-A(g7) have been proposed. These motifs results from independent experimental studies carried out on small data sets. Testing with multiple data sets showed that each of the motifs at best describes only a subset of the solution space, and these motifs therefore lack generalization ability. This study focuses on seeking a motif with higher generalization ability so that it can predict binders in all A(g7) data sets with high accuracy. A binding score matrix representing peptide binding motif to A(g7) was derived using genetic algorithm (GA). The evolved score matrix significantly outperformed previously reported