Peptide binding to MHC class I is determined by individual pockets in the binding groove

Precision neoantigen discovery using large-scale immunopeptidomes and composite modeling of MHC peptide presentation

Major histocompatibility complex (MHC)-bound peptides that originate from tumor-specific genetic alterations, known as neoantigens, are an important class of anti-cancer therapeutic targets. Accurately predicting peptide presentation by MHC complexes is a key aspect of discovering therapeutically relevant neoantigens. Technological improvements in mass-spectrometry-based immunopeptidomics and advanced modeling techniques have vastly improved MHC presentation prediction over the past two decades. However, improvement in the sensitivity and specificity of prediction algorithms is needed for clinical applications such as the development of personalized cancer vaccines, the discovery of biomarkers for response to checkpoint blockade and the quantification of autoimmune risk in gene therapies. Toward this end, we generated allele-specific immunopeptidomics data using 25 mono-allelic cell lines and created Systematic HLA Epitope Ranking Pan Algorithm (SHERPA™), a pan-allelic MHC-peptide algorithm for predicting MHC-peptide binding and presentation. In contrast to previously published large-scale mono-allelic data, we used an HLA-null K562 parental cell line and a stable transfection of HLA alleles to better emulate native presentation. Our dataset includes five previously unprofiled alleles that expand MHC binding pocket diversity in the training data and extend allelic coverage in underprofiled populations. To improve generalizability, SHERPA systematically integrates 128 mono-allelic and 384 multi-allelic samples with publicly available immunoproteomics data and binding assay data. Using this dataset, we developed two features that empirically estimate the propensities of genes and specific regions within gene bodies to engender immunopeptides to represent antigen processing. Using a composite model constructed with gradient boosting decision trees, multi-allelic deconvolution and 2.15 million peptides encompassing 167 alleles, we achieved a 1.44 fold improvement of positive predictive value compared to existing tools when evaluated on independent mono-allelic datasets and a 1.17 fold improvement when evaluating on tumor samples. With a high degree of accuracy, SHERPA has the potential to enable precision neoantigen discovery for future clinical applications.

Charge-based interactions through peptide position 4 drive diversity of antigen presentation by human leukocyte antigen class I molecules

Human leukocyte antigen class I (HLA-I) molecules bind and present peptides at the cell surface to facilitate the induction of appropriate CD8+ T cell-mediated immune responses to pathogen- and self-derived proteins. The HLA-I peptide-binding cleft contains dominant anchor sites in the B and F pockets that interact primarily with amino acids at peptide position 2 and the C-terminus, respectively. Nonpocket peptide-HLA interactions also contribute to peptide binding and stability, but these secondary interactions are thought to be unique to individual HLA allotypes or to specific peptide antigens. Here, we show that two positively charged residues located near the top of peptide-binding cleft facilitate interactions with negatively charged residues at position 4 of presented peptides, which occur at elevated frequencies across most HLA-I allotypes. Loss of these interactions was shown to impair HLA-I/peptide binding and complex stability, as demonstrated by both in vitro and in silico experiments. Furthermore, mutation of these Arginine-65 (R65) and/or Lysine-66 (K66) residues in HLA-A*02:01 and A*24:02 significantly reduced HLA-I cell surface expression while also reducing the diversity of the presented peptide repertoire by up to 5-fold. The impact of the R65 mutation demonstrates that nonpocket HLA-I/peptide interactions can constitute anchor motifs that exert an unexpectedly broad influence on HLA-I-mediated antigen presentation. These findings provide fundamental insights into peptide antigen binding that could broadly inform epitope discovery in the context of viral vaccine development and cancer immunotherapy.

An In Vivo Screen to Identify Short Peptide Mimotopes with Enhanced Antitumor Immunogenicity

Tumor-associated self-antigens are potential cancer vaccine targets but suffer from limited immunogenicity. There are examples of mutated, short self-peptides inducing epitope-specific CD8⁺ T cells more efficiently than the wild-type epitope, but current approaches cannot yet reliably identify such epitopes, which are referred to as enhanced mimotopes ("e-mimotopes"). Here, we present a generalized strategy to develop e-mimotopes, using the tyrosinase-related protein 2 (Trp2) peptide Trp2180-188, which is a murine major histocompatibility complex class I (MHC-I) epitope, as a test case. Using a vaccine adjuvant that induces peptide particle formation and strong cellular responses with nanogram antigen doses, a two-step method systematically identified e-mimotope candidates with murine immunization. First, position-scanning peptide micro libraries were generated in which each position of the wild-type epitope sequence was randomized. Randomization of only one specific residue of the Trp2 epitope increased antitumor immunogenicity. Second, all 20 amino acids were individually substituted and tested at that position, enabling the identification of two e-mimotopes with single amino-acid mutations. Despite similar MHC-I affinity compared to the wild-type epitope, e-mimotope immunization elicited improved Trp2-specific cytotoxic T-cell phenotypes and improved T-cell receptor affinity for both the e-mimotopes and the native epitope, resulting in better outcomes in multiple prophylactic and therapeutic tumor models. The screening method was also applied to other targets with other murine MHC-I restriction elements, including epitopes within glycoprotein 70 and Wilms' Tumor Gene 1, to identify additional e-mimotopes with enhanced potency.

Rapid peptide exchange on MHC class I by small molecules elucidates dynamics of bound peptide

Complexes of peptides with recombinant major histocompatibility complex class I molecules (rpMHCs) are an important tool for T cell detection, isolation, and activation in cancer immunotherapy. The rapid preparation of rpMHCs is aided by peptide exchange, for which several technologies exist. Here, we show peptide exchange with small-molecule alcohols and demonstrate that they accelerate the dissociation of pre-bound peptides, creating a novel method for rapid production of rpMHCs and increasing the understanding of the conformational flexibility of the MHC-bound peptides.

•

Small alcohols can catalyze peptide exchange on MHC-I.

•

C-terminal peptide binding plays an important role in ethanol mediated exchange.

•

MHC-I ethanol peptide exchange is allotype dependent.

Withdrawn: Precision Neoantigen Discovery Using Large-scale Immunopeptidomes and Composite Modeling of MHC Peptide Presentation

This article has been withdrawn by the authors. A publication of the manuscript with the correct figures and tables has been approved and the authors state the conclusions of the manuscript remain unaffected. Specifically, errors are in Figure 6A, Supplementary Figure 10B, Supplementary Figure 10C, and Supplementary Table 5. The details of the errors are as follows: the HLA types for one sample were incorrectly assigned because of a tumor/normal mislabeling from the biobank vendor. Due to the differing HLA types between the tumor and normal sample, the sequence analysis established that the HLA alleles for this patient had been deleted (HLA LOH). The authors conclude that this was an artifact caused by the normal sample mislabeling. The corrected version can be accessed (Pyke, R.M., Mellacheruvu, D., Dea, S., Abbott, C.W., Zhang, S.V., Philips, N.A., Harris, J., Bartha, G., Desai, S., McClory, R., West, J., Snyder, M,P., Chen, R., Boyle, S.M. (2022) Precision Neoantigen Discovery Using Large-Scale Immunopeptidomics and Composite Modeling of MHC Peptide Presentation. Mol. Cell. Proteomics 22, 100506

Informatics investigations into anti-thyroid drug induced agranulocytosis associated with multiple HLA-B alleles

INTRODUCTION

Adverse drug reactions have been linked with HLA alleles in different studies. These HLA proteins play an essential role in the adaptive immune response for the presentation of self and non-self peptides. Anti-thyroid drugs methimazole and propylthiouracil have been associated with drug induced agranulocytosis (severe lower white blood cell count) in patients with B*27:05, B*38:02 and DRB1*08:03 alleles in different populations: Taiwanese, Vietnamese, Han Chinese and Caucasian.

METHODS

In this study, informatics methods were used to investigate if any sequence or structural similarities exist between the two associated HLA-B alleles, compared with a set of "control" alleles assumed not be associated, which could help explain the molecular basis of the adverse drug reaction. We demonstrated using MHC Motif Viewer and MHCcluster that the two alleles do not have a propensity to bind similar peptides, and thus at a gross level the structure of the antigen presentation region of the two alleles are not similar. We also performed multiple sequence alignment to identify polymorphisms shared by the risk but not by the control alleles and molecular docking to compare the predicted binding poses of the drug-allele combinations.

RESULTS

Two residues, Cys67 and Thr80, were identified from the multiple sequence alignments to be unique to these risk alleles alone. The molecular docking showed the poses of the risk alleles to favour the F-pocket of the peptide binding groove, close to the Thr80 residue, with the control alleles generally favouring a different pocket. The data are thus suggestive that Thr80 may be a critical residue in HLA-mediated anti-thyroid drug induced agranulocytosis, and thus can guide future research and risk assessment.

Critical assessment of approaches for molecular docking to elucidate associations of HLA alleles with adverse drug reactions

•

All software assessed could dock Abacavir back into the risk allele structure but not always predict the exact binding mode.

•

Most docking software assessed can distinguish between risk and control alleles.

•

Docking performance can be degraded by using a homology model.

•

Receptor flexibility can negatively affect the docking performance for complex HLA examples.

•

Using AutoDockFR cannot compensate for the added difficulty of docking to the unbound target.

Adverse drug reactions have been linked with genetic polymorphisms in HLA genes in numerous different studies. HLA proteins have an essential role in the presentation of self and non-self peptides, as part of the adaptive immune response. Amongst the associated drugs-allele combinations, anti-HIV drug Abacavir has been shown to be associated with the HLA-B*57:01 allele, and anti-epilepsy drug Carbamazepine with B*15:02, in both cases likely following the altered peptide repertoire model of interaction. Under this model, the drug binds directly to the antigen presentation region, causing different self peptides to be presented, which trigger an unwanted immune response. There is growing interest in searching for evidence supporting this model for other ADRs using bioinformatics techniques. In this study, in silico docking was used to assess the utility and reliability of well-known docking programs when addressing these challenging HLA-drug situations. The overall aim was to address the uncertainty of docking programs giving different results by completing a detailed comparative study of docking software, grounded in the MHC-ligand experimental structural data – for Abacavir and to a lesser extent Carbamazepine - in order to assess their performance. Four docking programs: SwissDock, ROSIE, AutoDock Vina and AutoDockFR, were used to investigate if each software could accurately dock the Abacavir back into the crystal structure for the protein arising from the known risk allele, and if they were able to distinguish between the HLA-associated and non-HLA-associated (control) alleles. The impact of using homology models on the docking performance and how using different parameters, such as including receptor flexibility, affected the docking performance were also investigated to simulate the approach where a crystal structure for a given HLA allele may be unavailable. The programs that were best able to predict the binding position of Abacavir were then used to recreate the docking seen for Carbamazepine with B*15:02 and controls alleles.

It was found that the programs investigated were sometimes able to correctly predict the binding mode of Abacavir with B*57:01 but not always. Each of the software packages that were assessed could predict the binding of Abacavir and Carbamazepine within the correct sub-pocket and, with the exception of ROSIE, was able to correctly distinguish between risk and control alleles. We found that docking to homology models could produce poorer quality predictions, especially when sequence differences impact the architecture of predicted binding pockets. Caution must therefore be used as inaccurate structures may lead to erroneous docking predictions. Incorporating receptor flexibility was found to negatively affect the docking performance for the examples investigated. Taken together, our findings help characterise the potential but also the limitations of computational prediction of drug-HLA interactions. These docking techniques should therefore always be used with care and alongside other methods of investigation, in order to be able to draw strong conclusions from the given results.

The Optimization of Bioorthogonal Epitope Ligation within MHC-I Complexes

Antigen recognition followed by the activation of cytotoxic T-cells (CTLs) is a key step in adaptive immunity, resulting in clearance of viruses and cancers. The repertoire of peptides that have the ability to bind to the major histocompatibility type-I (MHC-I) is enormous, but the approaches available for studying the diversity of the peptide repertoire on a cell are limited. Here, we explore the use of bioorthogonal chemistry to quantify specific peptide-MHC-I complexes (pMHC-I) on cells. We show that modifying epitope peptides with bioorthogonal groups in surface accessible positions allows wild-type-like MHC-I binding and bioorthogonal ligation using fluorogenic chromophores in combination with a Cu(I)-catalyzed Huisgen cycloaddition reaction. We expect that this approach will make a powerful addition to the antigen presentation toolkit as for the first time it allows quantification of antigenic peptides for which no detection tools exist.

Chemical Modification of Influenza CD8+ T-Cell Epitopes Enhances Their Immunogenicity Regardless of Immunodominance

T cells are essential players in the defense against infection. By targeting the MHC class I antigen-presenting pathway with peptide-based vaccines, antigen-specific T cells can be induced. However, low immunogenicity of peptides poses a challenge. Here, we set out to increase immunogenicity of influenza-specific CD8+ T cell epitopes. By substituting amino acids in wild type sequences with non-proteogenic amino acids, affinity for MHC can be increased, which may ultimately enhance cytotoxic CD8+ T cell responses. Since preventive vaccines against viruses should induce a broad immune response, we used this method to optimize influenza-specific epitopes of varying dominance. For this purpose, HLA-A*0201 epitopes GILGFVFTL, FMYSDFHFI and NMLSTVLGV were selected in order of decreasing MHC-affinity and dominance. For all epitopes, we designed chemically enhanced altered peptide ligands (CPLs) that exhibited greater binding affinity than their WT counterparts; even binding scores of the high affinity GILGFVFTL epitope could be improved. When HLA-A*0201 transgenic mice were vaccinated with selected CPLs, at least 2 out of 4 CPLs of each epitope showed an increase in IFN-γ responses of splenocytes. Moreover, modification of the low affinity epitope NMLSTVLGV led to an increase in the number of mice that responded. By optimizing three additional influenza epitopes specific for HLA-A*0301, we show that this strategy can be extended to other alleles. Thus, enhancing binding affinity of peptides provides a valuable tool to improve the immunogenicity and range of preventive T cell-targeted peptide vaccines.

The Carboxy Terminus of the Ligand Peptide Determines the Stability of the MHC Class I Molecule H-2Kb: A Combined Molecular Dynamics and Experimental Study

Major histocompatibility complex (MHC) class I molecules (proteins) bind peptides of eight to ten amino acids to present them at the cell surface to cytotoxic T cells. The class I binding groove binds the peptide via hydrogen bonds with the peptide termini and via diverse interactions with the anchor residue side chains of the peptide. To elucidate which of these interactions is most important for the thermodynamic and kinetic stability of the peptide-bound state, we have combined molecular dynamics simulations and experimental approaches in an investigation of the conformational dynamics and binding parameters of a murine class I molecule (H-2K^b) with optimal and truncated natural peptide epitopes. We show that the F pocket region dominates the conformational and thermodynamic properties of the binding groove, and that therefore the binding of the C terminus of the peptide to the F pocket region plays a crucial role in bringing about the peptide-bound state of MHC class I.

Chronic beryllium disease, HLA-DPB1, and the DP peptide binding groove

Multiple epidemiologic studies demonstrate associations between chronic beryllium disease (CBD), beryllium sensitization (BeS), and HLA-DPB1 alleles with a glutamic acid residue at position 69 (E69). Results suggest that the less-frequent E69 variants (non-*0201/*0202 alleles) might be associated with greater risk of CBD. In this study, we sought to define specific E69-carrying alleles and their amino acid sequences in the DP peptide binding groove, as well as their relationship to CBD and BeS risk, using the largest case control study to date. We enrolled 502 BeS/CBD subjects and 653 beryllium-exposed controls from three beryllium industries who gave informed consent for participation. Non-Hispanic white cases and controls were frequency-matched by industry. HLA-DPB1 genotypes were determined using sequence-specific primer PCR. The E69 alleles were tested for association with disease individually and grouped by amino acid structure using logistic regression. The results show that CBD cases were more likely than controls to carry a non-*02 E69 allele than an *02 E69, with odds ratios (95% confidence interval) ranging from 3.1 (2.1-4.5) to 3.9 (2.6-5.9) (p < 0.0001). Polymorphic amino acids at positions 84 and 11 were associated with CBD: DD versus GG, 2.8 (1.8-4.6), p < 0.0001; GD versus GG, 2.1 (1.5-2.8), p < 0.0001; LL versus GG, 3.2 (1.8-5.6), p < 0.0001; GL versus GG, 2.8 (2.1-3.8), p < 0.0001. Similar results were found within the BeS group and CBD/BeS combined group. We conclude that the less frequent E69 alleles confer more risk for CBD than does *0201. Recent studies examining how the composition and structure of the binding pockets influence peptide binding in MHC genes, as well of studies showing the topology of the TCR to likely bind DPB1 preferentially, give plausible biological rationale for these findings.

Different strategies adopted by K(b) and L(d) to generate T cell specificity directed against their respective bound peptides

Mouse T cell clone 2C recognizes two different major histocompatibility (MHC) ligands, the self MHC K(b) and the allogeneic MHC L(d). Two distinct peptides, SIY (SIYRYYGL) and QL9 (QLSPFPFDL), act as strong and specific agonists when bound to K(b) and L(d), respectively. To explore further the mechanisms involved in peptide potency and specificity, here we examined a collection of single amino acid peptide variants of SIY and QL9 for 1) T cell activity, 2) binding to their respective MHC, and 3) binding to the 2C T cell receptor (TCR) and high affinity TCR mutants. Characterization of SIY binding to MHC K(b) revealed significant effects of three SIY residues that were clearly embedded within the K(b) molecule. In contrast, QL9 binding to MHC L(d) was influenced by the majority of peptide side chains, distributed across the entire length of the peptide. Binding of the SIY-K(b) complex to the TCR involved three SIY residues that were pointed toward the TCR, whereas again the majority of QL9 residues influenced binding of TCRs, and thus the QL9 residues had impacts on both L(d) and TCR binding. In general, the magnitude of T cell activity mediated by a peptide variant was influenced more by peptide binding to MHC than by binding the TCR, especially for higher affinity TCRs. Findings with both systems, but QL9-L(d) in particular, suggest that many single-residue substitutions, introduced into peptides to improve their binding to MHC and thus their vaccine potential, could impair T cell reactivity due to their dual impact on TCR binding.

Reference strand-mediated conformation analysis-based typing of multiple alleles in the rhesus macaque MHC class I Mamu-A and Mamu-B loci

The rhesus macaque exhibits individual differences in susceptibility and resistance to infectious agents such as simian immunodeficiency virus (SIV) under experimental conditions, and these may be genetically determined at least in part by major histocompatibility complex (MHC) class I polymorphism. Although the importance of defining MHC class I polymorphism is well recognized, development of a generic and comprehensive molecular typing method of MHC class I alleles of the rhesus macaque has been hampered because, during the evolution of this species, multiple copies of similar DNA sequences have been generated by duplication events including the coding sequences of Mamu-A and Mamu-B loci. We report here a newly developed reference strand-mediated conformation analysis (RSCA)-based typing method of multiple Mamu-A and Mamu-B cDNAs that allowed us to estimate the number of expressed alleles. This technique detected 1-7 Mamu-A signals and 2-12 Mamu-B signals in a single sample, indicating that the number of functional alleles may vary. By comparing the data from the parents with those from the descendants in the breeding colony, several MHC class I haplotypes consisting of variable numbers of functional Mamu-A and Mamu-B alleles could be assigned.

T cell epitope identification for bovine vaccines: an epitope mapping method for BoLA A-11

T cell responses play an important role in immunity to parasites and other microbial agents of infectious diseases, therefore a number of T cell-directed vaccines are in development. Computer-driven algorithms that facilitate the discovery of T cell epitopes from protein and genome sequences are now being used to accelerate preclinical studies of human vaccines. Similar tools are not yet available for predicting T cell epitopes for animal vaccines, but there may be sufficient data available to begin the process of compiling the algorithms. We describe the construction of a novel mathematical 'matrix' that describes the properties of bovine major histocompatibility complex (BoLA) system antigen (BoLA) A-11 peptide ligands, developed for use with EpiMatrix, an existing T cell epitope-mapping algorithm. An alternative means of developing BoLA matrices, using the pocket profile method, is also discussed. Matrices such as the one described here may be used to develop T cell epitope-mapping tools for cattle and other ruminants. Epitope-mapping algorithms offer a significant advantage over other methods of epitope selection, such as the screening of synthetic overlapping peptides, because high throughput screening can be performed in silico, followed by ex vivo confirmatory studies. Furthermore, using epitope-mapping algorithms, putative T cell epitopes can be derived directly from genomic sequences, allowing researchers to circumvent labor-intensive cloning steps in the genome-to-vaccine discovery pathway.