Towards the in silico identification of class II restricted T-cell epitopes: a partial least squares iterative self-consistent algorithm for affinity prediction

MOTIVATION: The immunogenicity of peptides depends on their ability to bind to MHC molecules. MHC binding affinity prediction methods can save significant amounts of experimental work. The class II MHC binding site is open at both ends, making epitope prediction difficult because of the multiple binding ability of long peptides. RESULTS: An iterative self-consistent partial least squares (PLS)-based additive method was applied to a set of 66 peptides no longer than 16 amino acids, binding to DRB1*0401. A regression equation containing the quantitative contributions of the amino acids at each of the nine positions was generated. Its predictability was tested using two external test sets which gave r(pred) = 0.593 and r(pred) = 0.655, respectively. Furthermore, it was benchmarked using 25 known T-cell epitopes restricted by DRB1*0401 and we compared our results with four other online predictive methods. The additive method showed the best result finding 24 of the 25 T-cell epitopes. AVAILABILITY: Peptides used in the study are available from http://www.jenner.ac.uk/JenPep. The PLS method is available commercially in the SYBYL molecular modelling software package. The final model for affinity prediction of peptides binding to DRB1*0401 molecule is available at http://www.jenner.ac.uk/MHCPred. Models developed for DRB1*0101 and DRB1*0701 also are available in MHCPred.     

Cooperativity of hydrophobic anchor interactions: evidence for epitope selection by MHC class II as a folding process

Peptide binding to MHC class II (MHCII) molecules is stabilized by hydrophobic anchoring and hydrogen bond formation. We view peptide binding as a process in which the peptide folds into the binding groove and to some extent the groove folds around the peptide. Our previous observation of cooperativity when analyzing binding properties of peptides modified at side chains with medium to high solvent accessibility is compatible with such a view. However, a large component of peptide binding is mediated by residues with strong hydrophobic interactions that bind to their respective pockets. If these reflect initial nucleation events they may be upstream of the folding process and not show cooperativity. To test whether the folding hypothesis extends to these anchor interactions, we measured dissociation and affinity to HLA-DR1 of an influenza hemagglutinin-derived peptide with multiple substitutions at major anchor residues. Our results show both negative and positive cooperative effects between hydrophobic pocket interactions. Cooperativity was also observed between hydrophobic pockets and positions with intermediate solvent accessibility, indicating that hydrophobic interactions participate in the overall folding process. These findings point out that predicting the binding potential of epitopes cannot assume additive and independent contributions of the interactions between major MHCII pockets and corresponding peptide side chains.     

Quantitative approaches to computational vaccinology

This article reviews the newly released JenPep database and two new powerful techniques for T-cell epitope prediction: (i) the additive method; and (ii) a 3D-Quantitative Structure Activity Relationships (3D-QSAR) method, based on Comparative Molecular Similarity Indices Analysis (CoMSIA). The JenPep database is a family of relational databases supporting the growing need of immunoinformaticians for quantitative data on peptide binding to major histocompatibility complexes and to the Transporters associated with Antigen Processing (TAP). It also contains an annotated list of T-cell epitopes. The database is available free via the Internet (http://www.jenner.ac.uk/JenPep). The additive prediction method is based on the assumption that the binding affinity of a peptide depends on the contributions from each amino acid as well as on the interactions between the adjacent and every second side-chain. In the 3D-QSAR approach, the influence of five physicochemical properties (steric bulk, electrostatic potential, local hydrophobicity, hydrogen-bond donor and hydrogen-bond acceptor abilities) on the affinity of peptides binding to MHC molecules were considered. Both methods were exemplified through their application to the well-studied problem of peptides binding to the human class I MHC molecule HLA-A*0201.     

Empirical calculation of the relative free energies of peptide binding to the molecular chaperone DnaK

We describe a methodology to calculate the relative free energies of protein-peptide complex formation. The interaction energy was decomposed into nonpolar, electrostatic and entropic contributions. A free energy-surface area relationship served to calculate the nonpolar free energy term. The electrostatic free energy was calculated with the finite difference Poisson-Boltzmann method and the entropic contribution was estimated from the loss in the conformational entropy of the peptide side chains. We applied this methodology to a series of DnaK*peptide complexes. On the basis of the single known crystal structure of the peptide-binding domain of DnaK with a bound heptapeptide, we modeled ten other DnaK*heptapeptide complexes with experimentally measured K(d) values from 0.06 microM to 11 microM, using molecular dynamics to refine the structures of the complexes. Molecular dynamic trajectories, after equilibration, were used for calculating the energies with greater accuracy. The calculated relative binding free energies were compared with the experimentally determined free energies. Linear scaling of the calculated terms was applied to fit them to the experimental values. The calculated binding free energies were between -7.1 kcal/mol and - 9.4 kcal/mol with a correlation coefficient of 0.86. The calculated nonpolar contributions are mainly due to the central hydrophobic binding pocket of DnaK for three amino acid residues. Negative electrostatic fields generated by the protein increase the binding affinity for basic residues flanking the hydrophobic core of the peptide ligand. Analysis of the individual energy contributions indicated that the nonpolar contributions are predominant compared to the other energy terms even for peptides with low affinity and that inclusion of the change in conformational entropy of the peptide side chains does not improve the discriminative power of the calculation. The method seems to be useful for predicting relative binding energies of peptide ligands of DnaK and might be applicable to other protein-peptide systems, particularly if only the structure of one protein-ligand complex is available.     

Quantitative structure-activity relationships and the prediction of MHC supermotifs

The underlying assumption in quantitative structure-activity relationship (QSAR) methodology is that related chemical structures exhibit related biological activities. We review here two QSAR methods in terms of their applicability for human MHC supermotif definition. Supermotifs are motifs that characterise binding to more than one allele. Supermotif definition is the initial in silico step of epitope-based vaccine design. The first QSAR method we review here--the additive method--is based on the assumption that the binding affinity of a peptide depends on contributions from both amino acids and the interactions between them. The second method is a 3D-QSAR method: comparative molecular similarity indices analysis (CoMSIA). Both methods were applied to 771 peptides binding to 9 HLA alleles. Five of the alleles (A*0201, A*0202, A*0203, A*0206 and A*6802) belong to the HLA-A2 superfamily and the other four (A*0301, A*1101, A*3101 and A*6801) to the HLA-A3 superfamily. For each superfamily, supermotifs defined by the two QSAR methods agree closely and are supported by many experimental data.     

Coupling in silico and in vitro analysis of peptide-MHC binding: a bioinformatic approach enabling prediction of superbinding peptides and anchorless epitopes

The ability to define and manipulate the interaction of peptides with MHC molecules has immense immunological utility, with applications in epitope identification, vaccine design, and immunomodulation. However, the methods currently available for prediction of peptide-MHC binding are far from ideal. We recently described the application of a bioinformatic prediction method based on quantitative structure-affinity relationship methods to peptide-MHC binding. In this study we demonstrate the predictivity and utility of this approach. We determined the binding affinities of a set of 90 nonamer peptides for the MHC class I allele HLA-A*0201 using an in-house, FACS-based, MHC stabilization assay, and from these data we derived an additive quantitative structure-affinity relationship model for peptide interaction with the HLA-A*0201 molecule. Using this model we then designed a series of high affinity HLA-A2-binding peptides. Experimental analysis revealed that all these peptides showed high binding affinities to the HLA-A*0201 molecule, significantly higher than the highest previously recorded. In addition, by the use of systematic substitution at principal anchor positions 2 and 9, we showed that high binding peptides are tolerant to a wide range of nonpreferred amino acids. Our results support a model in which the affinity of peptide binding to MHC is determined by the interactions of amino acids at multiple positions with the MHC molecule and may be enhanced by enthalpic cooperativity between these component interactions.     

Changing the antigen binding specificity by single point mutations of an anti-p24 (HIV-1) antibody

The murine mAb CB4-1 raised against p24 (HIV-1) recognizes a linear epitope of the HIV-1 capsid protein. Additionally, CB4-1 exhibits cross-reactive binding to epitope-homologous peptides and polyspecific reactions to epitope nonhomologous peptides. Crystal structures demonstrate that the epitope peptide (e-pep) and the nonhomologous peptides adopt different conformations within the binding region of CB4-1. Site-directed mutagenesis of the fragment variable (Fv) region was performed using a single-chain (sc)Fv construct of CB4-1 to analyze binding contributions of single amino acid side chains toward the e-pep and toward one epitope nonhomologous peptide. The mutations of Ab amino acid side chains, which are in direct contact with the Ag, show opposite influences on the binding of the two peptides. Whereas the affinity of the e-pep to the CB4-1 scFv mutant heavy chain variable region Tyr(32)Ala is decreased 250-fold, the binding of the nonhomologous peptide remains unchanged. In contrast, the mutation light chain variable region Phe(94)Ala reduces the affinity of the nonhomologous peptide 10-fold more than it does for the e-pep. Thus, substantial changes in the specificity can be observed by single amino acid exchanges. Further characterization of the scFv mutants by substitutional analysis of the peptides demonstrates that the effect of a mutation is not restricted to contact residues. This method also reveals an inverse compensatory amino acid exchange for the nonhomologous peptide which increases the affinity to the scFv mutant light chain variable region Phe(94)Ala up to the level of the e-pep affinity to the wild-type scFv.     

Toward the quantitative prediction of T-cell epitopes: QSAR studies on peptides having affinity with the class I MHC molecular HLA-A*0201.

It would be useful for vaccine development to develop a method of rapidly identifying peptide epitopes. In this paper, the empirical three-dimensional quantitative structure-affinity relationship (3D-QSAR) methods were used to study the relationship between the three dimensional structural parameters (the isotropic surface area, ISA, and the electronic charge index, ECI) of the HLA-A*0201 binding peptide and the HLA-A*0201/peptide binding affinities. A set of 102 peptides having affinity with the class I MHC HLA-A*0201 molecule was used as training set. A test set of 40 peptides was used to determine the predictive value of the models. The 3D-QSAR models yielded a q2 = 0.5724 and a high rpred2 = 0.6955. The standard regression coefficients indicated that the hydrophobic interactions played an important role in peptide-MHC molecule binding and predicted the specific amino acid residue essential at a certain position of the peptide. The approach tested in the current paper is highly complementary to many of the methods described in references and possesses good predictability. It is a rapid and convenient method to detect high affinity peptide epitopes.     

Molecular dynamics,free energy,and SPR analyses of the interactions between the SH2 domain of Grb2 and ErbB phosphotyrosyl peptides

We studied the interactions between the SH2 domain of growth factor receptor binding protein 2 (Grb2) and ErbB receptor-derived phosphotyrosyl peptides using molecular dynamics, free energy calculations, and surface plasmon resonance (SPR) analysis. Binding free energies for nine phosphotyrosyl peptides were calculated using the MM-PBSA continuum solvent method, and excellent qualitative agreement with the SPR experimental data, with a correlation coefficient of 0.92, was obtained. Consistent with previous experimental findings, phosphotyrosyl peptides with the consensus sequence pYXNX showed favorable binding affinity for the Grb2. Unexpectedly, phosphotyrosyl peptides with the consensus sequence pYQQD, which had not shown any specific binding affinity for the Grb2 in earlier studies, also showed favorable binding affinity for the Grb2 in our experimental and computational analyses. Component analysis of the calculated binding free energies revealed that van der Waals interaction between the Grb2 and the phosphotyrosyl peptide was the dominant factor for specificity and binding affinity. These results indicate that current methods of estimating binding free energies are efficient for obtaining important information about protein-protein interactions, which are essential for the transmission of signals in cellular signaling pathways.     

Physicochemical explanation of peptide binding to HLA-A*0201 major histocompatibility complex: a three-dimensional quantitative structure-activity relationship study

A three-dimensional quantitative structure-activity relationship method for the prediction of peptide binding affinities to the MHC class I molecule HLA-A*0201 was developed by applying the CoMSIA technique on a set of 266 peptides. To increase the self consistency of the initial CoMSIA model, the poorly predicted peptides were excluded from the training set in a stepwise manner and then included in the study as a test set. The final model, based on 236 peptides and considering the steric, electrostatic, hydrophobic, hydrogen bond donor, and hydrogen bond acceptor fields, had q2 = 0.683 and r2 = 0.891. The stability of this model was proven by cross-validations in two and five groups and by a bootstrap analysis of the non-cross-validated model. The residuals between the experimental pIC50 (-logIC50) values and those calculated by "leave-one-out" cross-validation were analyzed. According to the best model, 63.2% of the peptides were predicted with /residuals/ < or = 0.5 log unit; 29.3% with 1.0 < or = /residuals/ < 0.5; and 7.5% with /residuals/ > 1.0 log unit. The mean /residual/ value was 0.489. The coefficient contour maps identify the physicochemical property requirements at each position in the peptide molecule and suggest amino acid sequences for high-affinity binding to the HLA-A*0201 molecule.     

Calculating Partition Coefficients of Peptides by the Addition Method

A new addition method is described in this study for calculating the partition coefficients of peptides. LogP and logD values of peptides are calculated by summing the contributions of the component amino acids. The final models are derived from a multivariate linear regression analysis of 219 peptides with known experimental data. The standard errors in a leave-one-out cross-validation are 0.23 and 0.24 log units for the logP and logD values, respectively. The predictive ability of the model is tested by an extra set of ten peptides, and the self-consistency of the model is further demonstrated by a new validation procedure called the evolution test. The parameters obtained in regression could be used as hydrophobicity scales for amino acids. The application of such hydrophobicity scales has also been discussed.     

Analysis and prediction of affinity of TAP binding peptides using cascade SVM

The generation of cytotoxic T lymphocyte (CTL) epitopes from an antigenic sequence involves number of intracellular processes, including production of peptide fragments by proteasome and transport of peptides to endoplasmic reticulum through transporter associated with antigen processing (TAP). In this study, 409 peptides that bind to human TAP transporter with varying affinity were analyzed to explore the selectivity and specificity of TAP transporter. The abundance of each amino acid from P1 to P9 positions in high-, intermediate-, and low-affinity TAP binders were examined. The rules for predicting TAP binding regions in an antigenic sequence were derived from the above analysis. The quantitative matrix was generated on the basis of contribution of each position and residue in binding affinity. The correlation of r = 0.65 was obtained between experimentally determined and predicted binding affinity by using a quantitative matrix. Further a support vector machine (SVM)-based method has been developed to model the TAP binding affinity of peptides. The correlation (r = 0.80) was obtained between the predicted and experimental measured values by using sequence-based SVM. The reliability of prediction was further improved by cascade SVM that uses features of amino acids along with sequence. An extremely good correlation (r = 0.88) was obtained between measured and predicted values, when the cascade SVM-based method was evaluated through jackknife testing. A Web service, TAPPred (http://www.imtech.res.in/raghava/tappred/ or http://bioinformatics.uams.edu/mirror/tappred/), has been developed based on this approach.     

From Amino Acid Landscape to Protein Landscape: Analysis of Genetic Codes in Terms of Fitness Landscape

Assigning the values of a certain physicochemical property for individual amino acids to the corresponding codons, we can make an amino acid property "landscape" on a four valued three dimensional sequence space from a genetic code table. Eleven property landscapes made from the standard genetic code (SGC) were analyzed. The evaluation of correlation for each landscape is done by theta value, which represents the ratio of the mean slope (as an additive term) to the degree of roughness (as a nonadditive term). The theta-values for hydropathy indices, polarity, specific heat, and beta-sheet propensity were considerably large with respect to SGC. This implies that the additivity of the contribution from each letter holds for these properties. To clarify the meaning of the so-called mutational robustness of SGC, we next examined correlations between the amino acid property and the actual "site fitnesses" of a protein. The site fitnesses were derived from a set of binding preference scores of amino acid residues at every site in MHC class I molecule binding peptides (Udaka et al. in press). We found that the SGC's theta value for an amino acid property is correlated with the significance of the property in the protein function. Adaptive walk simulation on fitness (= affinity) landscapes in a base sequence space for these model peptides confirmed better evolvability due to the introduction of SGC.     

Binding of cationic pentapeptides with modified side chain lengths to negatively charged lipid membranes: Complex interplay of electrostatic and hydrophobic interactions

Basic amino acids play a key role in the binding of membrane associated proteins to negatively charged membranes. However, side chains of basic amino acids like lysine do not only provide a positive charge, but also a flexible hydrocarbon spacer that enables hydrophobic interactions. We studied the influence of hydrophobic contributions to the binding by varying the side chain length of pentapeptides with ammonium groups starting with lysine to lysine analogs with shorter side chains, namely ornithine (Orn), α,γ-diaminobutyric acid (Dab) and α, β-diaminopropionic acid (Dap). The binding to negatively charged phosphatidylglycerol (PG) membranes was investigated by calorimetry, FT-infrared spectroscopy (FT-IR) and monolayer techniques. The binding was influenced by counteracting and sometimes compensating contributions. The influence of the bound peptides on the lipid phase behavior depends on the length of the peptide side chains. Isothermal titration calorimetry (ITC) experiments showed exothermic and endothermic effects compensating to a different extent as a function of side chain length. The increase in lipid phase transition temperature was more significant for peptides with shorter side chains. FTIR-spectroscopy revealed changes in hydration of the lipid bilayer interface after peptide binding. Using monolayer techniques, the contributions of electrostatic and hydrophobic effects could clearly be observed. Peptides with short side chains induced a pronounced decrease in surface pressure of PG monolayers whereas peptides with additional hydrophobic interactions decreased the surface pressure much less or even lead to an increase, indicating insertion of the hydrophobic part of the side chain into the lipid monolayer.     

Non-additivity in protein-protein interactions

The energy of binding between proteins may be seen as the sum of the contributions of the individual amino acid residues. These contributions are additive when the binding energy, due to different amino acid residues, is independent of the interactions between amino acids in the same polypeptide chain. A measure of non-additivity is the coupling free energy. In this communication it is shown that: (1) the coupling free energy is the sum of intramolecular and intermolecular contributions; and (2), when additivity exists, experimentally determined values for the free energy of transfer of amino acids from water to the hydrophobic protein-protein interface are a very good approximation of their contribution to the energy of binding. Additivity cycles can be useful in determining the precise conditions where this approximation holds.     

Recapitulation and design of protein binding peptide structures and sequences

An important objective of computational protein design is the generation of high affinity peptide inhibitors of protein-peptide interactions, both as a precursor to the development of therapeutics aimed at disrupting disease causing complexes, and as a tool to aid investigators in understanding the role of specific complexes in the cell. We have developed a computational approach to increase the affinity of a protein-peptide complex by designing N or C-terminal extensions which interact with the protein outside the canonical peptide binding pocket. In a first in silico test, we show that by simultaneously optimizing the sequence and structure of three to nine residue peptide extensions starting from short (1-6 residue) peptide stubs in the binding pocket of a peptide binding protein, the approach can recover both the conformations and the sequences of known binding peptides. Comparison with phage display and other experimental data suggests that the peptide extension approach recapitulates naturally occurring peptide binding specificity better than fixed backbone design, and that it should be useful for predicting peptide binding specificities from crystal structures. We then experimentally test the approach by designing extensions for p53 and dystroglycan-based peptides predicted to bind with increased affinity to the Mdm2 oncoprotein and to dystrophin, respectively. The measured increases in affinity are modest, revealing some limitations of the method. Based on these in silico and experimental results, we discuss future applications of the approach to the prediction and design of protein-peptide interactions.     

Tailored amino acid diversity for the evolution of antibody affinity

Antibodies are a unique class of proteins with the ability to adapt their binding sites for high affinity and high specificity to a multitude of antigens. Many analyses have been performed on antibody sequences and structures to elucidate which amino acids have a predominant role in antibody interactions with antigens. These studies have generally not distinguished between amino acids selected for broad antigen specificity in the primary immune response and those selected for high affinity in the secondary immune response. By studying a large data set of affinity matured antibodies derived from in vitro directed evolution experiments, we were able to specifically highlight a subset of amino acids associated with affinity improvements. In a comparison of affinity maturations using either tailored or full amino acid diversification, the tailored approach was found to be at least as effective at improving affinity while requiring fewer mutagenesis libraries than the traditional method. The resulting sequence data also highlight the potential for further reducing amino acid diversity for high affinity binding interactions.     

Tailored amino acid diversity for the evolution of antibody affinity

Antibodies are a unique class of proteins with the ability to adapt their binding sites for high affinity and high specificity to a multitude of antigens. Many analyses have been performed on antibody sequences and structures to elucidate which amino acids have a predominant role in antibody interactions with antigens. These studies have generally not distinguished between amino acids selected for broad antigen specificity in the primary immune response and those selected for high affinity in the secondary immune response. By studying a large data set of affinity matured antibodies derived from in vitro directed evolution experiments, we were able to specifically highlight a subset of amino acids associated with affinity improvements. In a comparison of affinity maturations using either tailored or full amino acid diversification, the tailored approach was found to be at least as effective at improving affinity while requiring fewer mutagenesis libraries than the traditional method. The resulting sequence data also highlight the potential for further reducing amino acid diversity for high affinity binding interactions.     

Arsenite binding to synthetic peptides: the effect of increasing length between two cysteines

We utilized radioactive 73As-labeled arsenite and vacuum filtration methodology to determine the binding affinity of arsenite to eight synthetic peptides ranging from 13 to 24 amino acids long and containing one or two cysteines separated by 0-17 intervening amino acids. Six of the eight peptides were highly similar in amino acid sequence and were based on cysteine containing regions of the hormone-binding site of the human estrogen receptor-alpha (e.g., the sequence of peptide 28 is LEGAWCGKGVEGTEHLYSMKCKNV). The peptides with 0-14 intervening amino acids between two cysteines bound arsenite with Kd values of 2.7-20.1 uM and with Bmax values from 36 to 103 nmol/mg protein (from 0.083 to 0.19 nmol/nmol of protein). Thus, increasing the number of intervening amino acids from 0 to 14 made very little difference in the observed Kd values for arsenite, a surprising finding. Therefore, these peptides are flexible in solution and effectively contain a dithiol high affinity binding site for arsenite. Peptide 17 with two C separated by 19 amino acids bound arsenite with a Kd of 123 uM and a Bmax of 41.8 nmol/mg. The monothiol peptide 19 bound arsenite with a Kd of 124 uM and a Bmax of 26 nmol/mg protein. All experimental binding curves fit well to a one site binding model.     

A rapid method for quantitative prediction of high affinity CTL epitopes: QSAR studies on peptides having affinity with the class I MHC molecular HLA-A*0201

Summary It would be useful to develop a method to rapidly identify peptide epitopes for vaccine development. In this paper, empirical three-dimensional quantitative structure-affinity relationship (3D-QSAR) methods were used to study the relationship between the three dimensional structural parameters (the isotropic surface area, ISA, and the electronic charge index, ECI) of the HLA-A^*0201 binding peptide and the HLA-A^*0201/peptide binding affinities. A set of 102 peptides having affinity with the class I MHC HLA-A^*0201 molecule was used as training set. A test set of 40 peptides was used to determine the predictive value of the models. The 3D-QSAR models gave aq ²=0.5724 and highr _pred ² =0.6955. According to the standard regression coefficients, it is known that the hydrophobic interactions (in these studies, the ISA is highly correlative with the hydrophobic property) play a dominant role in peptide-MHC molecule binding, and also which amino acid residue with what property is needed at specific position of the peptide. The approach we have taken is highly complementary to the many excellent methods described in references and appears highly predictive. It is a rapid and convenient method for detecting high affinity peptide epitopes.