Deciphering HLA-I motifs across HLA peptidomes improves neo-antigen predictions and identifies allostery regulating HLA specificity

Immunoinformatics: Predicting Peptide-MHC Binding

Immunoinformatics is a discipline that applies methods of computer science to study and model the immune system. A fundamental question addressed by immunoinformatics is how to understand the rules of antigen presentation by MHC molecules to T cells, a process that is central to adaptive immune responses to infections and cancer. In the modern era of personalized medicine, the ability to model and predict which antigens can be presented by MHC is key to manipulating the immune system and designing strategies for therapeutic intervention. Since the MHC is both polygenic and extremely polymorphic, each individual possesses a personalized set of MHC molecules with different peptide-binding specificities, and collectively they present a unique individualized peptide imprint of the ongoing protein metabolism. Mapping all MHC allotypes is an enormous undertaking that cannot be achieved without a strong bioinformatics component. Computational tools for the prediction of peptide-MHC binding have thus become essential in most pipelines for T cell epitope discovery and an inescapable component of vaccine and cancer research. Here, we describe the development of several such tools, from pioneering efforts to the current state-of-the-art methods, that have allowed for accurate predictions of peptide binding of all MHC molecules, even including those that have not yet been characterized experimentally.

Multiplexed relative and absolute quantitative immunopeptidomics reveals MHC I repertoire alterations induced by CDK4/6 inhibition

Peptides bound to class I major histocompatibility complexes (MHC) play a critical role in immune cell recognition and can trigger an antitumor immune response in cancer. Surface MHC levels can be modulated by anticancer agents, altering immunity. However, understanding the peptide repertoire's response to treatment remains challenging and is limited by quantitative mass spectrometry-based strategies lacking normalization controls. We describe an experimental platform that leverages recombinant heavy isotope-coded peptide MHCs (hipMHCs) and multiplex isotope tagging to quantify peptide repertoire alterations using low sample input. HipMHCs improve quantitative accuracy of peptide repertoire changes by normalizing for variation across analyses and enable absolute quantification using internal calibrants to determine copies per cell of MHC antigens, which can inform immunotherapy design. Applying this platform in melanoma cell lines to profile the immunopeptidome response to CDK4/6 inhibition and interferon-γ - known modulators of antigen presentation - uncovers treatment-specific alterations, connecting the intracellular response to extracellular immune presentation.

Cathepsin S Regulates Antigen Processing and T Cell Activity in Non-Hodgkin Lymphoma

Genomic alterations in cancer cells can influence the immune system to favor tumor growth. In non-Hodgkin lymphoma, physiological interactions between B cells and the germinal center microenvironment are coopted to sustain cancer cell proliferation. We found that follicular lymphoma patients harbor a recurrent hotspot mutation targeting tyrosine 132 (Y132D) in cathepsin S (CTSS) that enhances protein activity. CTSS regulates antigen processing and CD4⁺ and CD8⁺ T cell-mediated immune responses. Loss of CTSS activity reduces lymphoma growth by limiting communication with CD4⁺ T follicular helper cells while inducing antigen diversification and activation of CD8⁺ T cells. Overall, our results suggest that CTSS inhibition has non-redundant therapeutic potential to enhance anti-tumor immune responses in indolent and aggressive lymphomas.

In-depth mining of the immunopeptidome of an acute myeloid leukemia cell line using complementary ligand enrichment and data acquisition strategies

The identification of T cell epitopes derived from tumour specific antigens remains a significant challenge for the development of peptide-based vaccines and immunotherapies. The use of mass spectrometry-based approaches (immunopeptidomics) can provide powerful new avenues for the identification of such epitopes. In this study we report the use of complementary peptide antigen enrichment methods and a comprehensive mass spectrometric acquisition strategy to provide in-depth immunopeptidome data for the THP-1 cell line, a cell line used widely as a model of human leukaemia. To accomplish this, we combined robust experimental workflows that incorporated ultrafiltration or off-line reversed phase chromatography to enrich peptide ligand as well as a multifaceted data acquisition strategy using an Orbitrap Fusion LC-MS instrument. Using the combined datasets from the two ligand enrichment methods we gained significant depth in immunopeptidome coverage by identifying a total of 41,816 HLA class I peptides from THP-1 cells, including a significant number of peptides derived from different oncogenes or over expressed proteins associated with cancer. The physicochemical properties of the HLA-bound peptides dictated their recovery using the two ligand enrichment approaches and their distribution across the different precursor charge states considered in the data acquisition strategy. The data highlight the complementarity of the two enrichment procedures, and in cases where sample is not limiting, suggest that the combination of both approaches will yield the most comprehensive immunopeptidome information.

A comprehensive review and performance evaluation of bioinformatics tools for HLA class I peptide-binding prediction

Human leukocyte antigen class I (HLA-I) molecules are encoded by major histocompatibility complex (MHC) class I loci in humans. The binding and interaction between HLA-I molecules and intracellular peptides derived from a variety of proteolytic mechanisms play a crucial role in subsequent T-cell recognition of target cells and the specificity of the immune response. In this context, tools that predict the likelihood for a peptide to bind to specific HLA class I allotypes are important for selecting the most promising antigenic targets for immunotherapy. In this article, we comprehensively review a variety of currently available tools for predicting the binding of peptides to a selection of HLA-I allomorphs. Specifically, we compare their calculation methods for the prediction score, employed algorithms, evaluation strategies and software functionalities. In addition, we have evaluated the prediction performance of the reviewed tools based on an independent validation data set, containing 21 101 experimentally verified ligands across 19 HLA-I allotypes. The benchmarking results show that MixMHCpred 2.0.1 achieves the best performance for predicting peptides binding to most of the HLA-I allomorphs studied, while NetMHCpan 4.0 and NetMHCcons 1.1 outperform the other machine learning-based and consensus-based tools, respectively. Importantly, it should be noted that a peptide predicted with a higher binding score for a specific HLA allotype does not necessarily imply it will be immunogenic. That said, peptide-binding predictors are still very useful in that they can help to significantly reduce the large number of epitope candidates that need to be experimentally verified. Several other factors, including susceptibility to proteasome cleavage, peptide transport into the endoplasmic reticulum and T-cell receptor repertoire, also contribute to the immunogenicity of peptide antigens, and some of them can be considered by some predictors. Therefore, integrating features derived from these additional factors together with HLA-binding properties by using machine-learning algorithms may increase the prediction accuracy of immunogenic peptides. As such, we anticipate that this review and benchmarking survey will assist researchers in selecting appropriate prediction tools that best suit their purposes and provide useful guidelines for the development of improved antigen predictors in the future.

Identification of the Targets of T-cell Receptor Therapeutic Agents and Cells by Use of a High-Throughput Genetic Platform

T-cell receptor (TCR)-based therapeutic cells and agents have emerged as a new class of effective cancer therapies. These therapies work on cells that express intracellular cancer-associated proteins by targeting peptides displayed on MHC receptors. However, cross-reactivities of these agents to off-target cells and tissues have resulted in serious, sometimes fatal, adverse events. We have developed a high-throughput genetic platform (termed "PresentER") that encodes MHC-I peptide minigenes for functional immunologic assays and determines the reactivities of TCR-like therapeutic agents against large libraries of MHC-I ligands. In this article, we demonstrated that PresentER could be used to identify the on-and-off targets of T cells and TCR-mimic (TCRm) antibodies using in vitro coculture assays or binding assays. We found dozens of MHC-I ligands that were cross-reactive with two TCRm antibodies and two native TCRs and that were not easily predictable by other methods.

T Cell Epitope Predictions

Throughout the body, T cells monitor MHC-bound ligands expressed on the surface of essentially all cell types. MHC ligands that trigger a T cell immune response are referred to as T cell epitopes. Identifying such epitopes enables tracking, phenotyping, and stimulating T cells involved in immune responses in infectious disease, allergy, autoimmunity, transplantation, and cancer. The specific T cell epitopes recognized in an individual are determined by genetic factors such as the MHC molecules the individual expresses, in parallel to the individual's environmental exposure history. The complexity and importance of T cell epitope mapping have motivated the development of computational approaches that predict what T cell epitopes are likely to be recognized in a given individual or in a broader population. Such predictions guide experimental epitope mapping studies and enable computational analysis of the immunogenic potential of a given protein sequence region.

ProGeo-neo: a customized proteogenomic workflow for neoantigen prediction and selection

BACKGROUND

Neoantigens can be differentially recognized by T cell receptor (TCR) as these sequences are derived from mutant proteins and are unique to the tumor. The discovery of neoantigens is the first key step for tumor-specific antigen (TSA) based immunotherapy. Based on high-throughput tumor genomic analysis, each missense mutation can potentially give rise to multiple neopeptides, resulting in a vast total number, but only a small percentage of these peptides may achieve immune-dominant status with a given major histocompatibility complex (MHC) class I allele. Specific identification of immunogenic candidate neoantigens is consequently a major challenge. Currently almost all neoantigen prediction tools are based on genomics data.

RESULTS

Here we report the construction of proteogenomics prediction of neoantigen (ProGeo-neo) pipeline, which incorporates the following modules: mining tumor specific antigens from next-generation sequencing genomic and mRNA expression data, predicting the binding mutant peptides to class I MHC molecules by latest netMHCpan (v.4.0), verifying MHC-peptides by MaxQuant with mass spectrometry proteomics data searched against customized protein database, and checking potential immunogenicity of T-cell-recognization by additional screening methods. ProGeo-neo pipeline achieves proteogenomics strategy and the neopeptides identified were of much higher quality as compared to those identified using genomic data only.

CONCLUSIONS

The pipeline was constructed based on the genomics and proteomics data of Jurkat leukemia cell line but is generally applicable to other solid cancer research. With massively parallel sequencing and proteomics profiling increasing, this proteogenomics workflow should be useful for neoantigen oriented research and immunotherapy.

Integrated proteogenomic deep sequencing and analytics accurately identify non-canonical peptides in tumor immunopeptidomes

Efforts to precisely identify tumor human leukocyte antigen (HLA) bound peptides capable of mediating T cell-based tumor rejection still face important challenges. Recent studies suggest that non-canonical tumor-specific HLA peptides derived from annotated non-coding regions could elicit anti-tumor immune responses. However, sensitive and accurate mass spectrometry (MS)-based proteogenomics approaches are required to robustly identify these non-canonical peptides. We present an MS-based analytical approach that characterizes the non-canonical tumor HLA peptide repertoire, by incorporating whole exome sequencing, bulk and single-cell transcriptomics, ribosome profiling, and two MS/MS search tools in combination. This approach results in the accurate identification of hundreds of shared and tumor-specific non-canonical HLA peptides, including an immunogenic peptide derived from an open reading frame downstream of the melanoma stem cell marker gene ABCB5. These findings hold great promise for the discovery of previously unknown tumor antigens for cancer immunotherapy.

Mass Spectrometry Based Immunopeptidomics Leads to Robust Predictions of Phosphorylated HLA Class I Ligands

The presentation of peptides on class I human leukocyte antigen (HLA-I) molecules plays a central role in immune recognition of infected or malignant cells. In cancer, non-self HLA-I ligands can arise from many different alterations, including non-synonymous mutations, gene fusion, cancer-specific alternative mRNA splicing or aberrant post-translational modifications. Identifying HLA-I ligands remains a challenging task that requires either heavy experimental work for in vivo identification or optimized bioinformatics tools for accurate predictions. To date, no HLA-I ligand predictor includes post-translational modifications. To fill this gap, we curated phosphorylated HLA-I ligands from several immunopeptidomics studies (including six newly measured samples) covering 72 HLA-I alleles and retrieved a total of 2,066 unique phosphorylated peptides. We then expanded our motif deconvolution tool to identify precise binding motifs of phosphorylated HLA-I ligands. Our results reveal a clear enrichment of phosphorylated peptides among HLA-C ligands and demonstrate a prevalent role of both HLA-I motifs and kinase motifs on the presentation of phosphorylated peptides. These data further enabled us to develop and validate the first predictor of interactions between HLA-I molecules and phosphorylated peptides.

High-Throughput Prediction of MHC Class I and II Neoantigens with MHCnuggets

Computational prediction of binding between neoantigen peptides and major histocompatibility complex (MHC) proteins can be used to predict patient response to cancer immunotherapy. Current neoantigen predictors focus on in silico estimation of MHC binding affinity and are limited by low predictive value for actual peptide presentation, inadequate support for rare MHC alleles, and poor scalability to high-throughput data sets. To address these limitations, we developed MHCnuggets, a deep neural network method that predicts peptide-MHC binding. MHCnuggets can predict binding for common or rare alleles of MHC class I or II with a single neural network architecture. Using a long short-term memory network (LSTM), MHCnuggets accepts peptides of variable length and is faster than other methods. When compared with methods that integrate binding affinity and MHC-bound peptide (HLAp) data from mass spectrometry, MHCnuggets yields a 4-fold increase in positive predictive value on independent HLAp data. We applied MHCnuggets to 26 cancer types in The Cancer Genome Atlas, processing 26.3 million allele-peptide comparisons in under 2.3 hours, yielding 101,326 unique predicted immunogenic missense mutations (IMM). Predicted IMM hotspots occurred in 38 genes, including 24 driver genes. Predicted IMM load was significantly associated with increased immune cell infiltration (P < 2 × 10^-16), including CD8⁺ T cells. Only 0.16% of predicted IMMs were observed in more than 2 patients, with 61.7% of these derived from driver mutations. Thus, we describe a method for neoantigen prediction and its performance characteristics and demonstrate its utility in data sets representing multiple human cancers.

Analysis of Secondary Structure Biases in Naturally Presented HLA-I Ligands

Recent clinical developments in antitumor immunotherapy involving T-cell related therapeutics have led to a renewed interest for human leukocyte antigen class I (HLA-I) binding peptides, given their potential use as peptide vaccines. Databases of HLA-I binding peptides hold therefore information on therapeutic targets essential for understanding immunity. In this work, we use in depth and accurate HLA-I peptidomics datasets determined by mass-spectrometry (MS) and analyze properties of the HLA-I binding peptides with structure-based computational approaches. HLA-I binding peptides are studied grouping all alleles together or in allotype-specific contexts. We capitalize on the increasing number of structurally determined proteins to (1) map the 3D structure of HLA-I binding peptides into the source proteins for analyzing their secondary structure and solvent accessibility in the protein context, and (2) search for potential differences between these properties in HLA-I binding peptides and in a reference dataset of HLA-I motif-like peptides. This is performed by an in-house developed heuristic search that considers peptides across all the human proteome and converges to a collection of peptides that exhibit exactly the same motif as the HLA-I peptides. Our results, based on 9-mers matched to protein 3D structures, clearly show enriched sampling for HLA-I presentation of helical fragments in the source proteins. This enrichment is significant, as compared to 9-mer HLA-I motif-like peptides, and is not entirely explained by the helical propensity of the preferred residues in the HLA-I motifs. We give possible hypothesis for the secondary structure biases observed in HLA-I peptides. This contribution is of potential interest for researchers working in the field of antigen presentation and proteolysis. This knowledge refines the understanding of the rules governing antigen presentation and could be added to the parameters of the current peptide-MHC class I binding predictors to increase their antigen predictive ability.

NNAlign_MA; MHC Peptidome Deconvolution for Accurate MHC Binding Motif Characterization and Improved T-cell Epitope Predictions

The set of peptides presented on a cell's surface by MHC molecules is known as the immunopeptidome. Current mass spectrometry technologies allow for identification of large peptidomes, and studies have proven these data to be a rich source of information for learning the rules of MHC-mediated antigen presentation. Immunopeptidomes are usually poly-specific, containing multiple sequence motifs matching the MHC molecules expressed in the system under investigation. Motif deconvolution -the process of associating each ligand to its presenting MHC molecule(s)- is therefore a critical and challenging step in the analysis of MS-eluted MHC ligand data. Here, we describe NNAlign_MA, a computational method designed to address this challenge and fully benefit from large, poly-specific data sets of MS-eluted ligands. NNAlign_MA simultaneously performs the tasks of (1) clustering peptides into individual specificities; (2) automatic annotation of each cluster to an MHC molecule; and (3) training of a prediction model covering all MHCs present in the training set. NNAlign_MA was benchmarked on large and diverse data sets, covering class I and class II data. In all cases, the method was demonstrated to outperform state-of-the-art methods, effectively expanding the coverage of alleles for which accurate predictions can be made, resulting in improved identification of both eluted ligands and T-cell epitopes. Given its high flexibility and ease of use, we expect NNAlign_MA to serve as an effective tool to increase our understanding of the rules of MHC antigen presentation and guide the development of novel T-cell-based therapeutics.

Machine Learning for Cancer Immunotherapies Based on Epitope Recognition by T Cell Receptors

In the last years, immunotherapies have shown tremendous success as treatments for multiple types of cancer. However, there are still many obstacles to overcome in order to increase response rates and identify effective therapies for every individual patient. Since there are many possibilities to boost a patient's immune response against a tumor and not all can be covered, this review is focused on T cell receptor-mediated therapies. CD8+ T cells can detect and destroy malignant cells by binding to peptides presented on cell surfaces by MHC (major histocompatibility complex) class I molecules. CD4+ T cells can also mediate powerful immune responses but their peptide recognition by MHC class II molecules is more complex, which is why the attention has been focused on CD8+ T cells. Therapies based on the power of T cells can, on the one hand, enhance T cell recognition by introducing TCRs that preferentially direct T cells to tumor sites (so called TCR-T therapy) or through vaccination to induce T cells in vivo. On the other hand, T cell activity can be improved by immune checkpoint inhibition or other means that help create a microenvironment favorable for cytotoxic T cell activity. The manifold ways in which the immune system and cancer interact with each other require not only the use of large omics datasets from gene, to transcript, to protein, and to peptide but also make the application of machine learning methods inevitable. Currently, discovering and selecting suitable TCRs is a very costly and work intensive in vitro process. To facilitate this process and to additionally allow for highly personalized therapies that can simultaneously target multiple patient-specific antigens, especially neoepitopes, breakthrough computational methods for predicting antigen presentation and TCR binding are urgently required. Particularly, potential cross-reactivity is a major consideration since off-target toxicity can pose a major threat to patient safety. The current speed at which not only datasets grow and are made available to the public, but also at which new machine learning methods evolve, is assuring that computational approaches will be able to help to solve problems that immunotherapies are still facing.

DeepHLApan: A Deep Learning Approach for Neoantigen Prediction Considering Both HLA-Peptide Binding and Immunogenicity

Neoantigens play important roles in cancer immunotherapy. Current methods used for neoantigen prediction focus on the binding between human leukocyte antigens (HLAs) and peptides, which is insufficient for high-confidence neoantigen prediction. In this study, we apply deep learning techniques to predict neoantigens considering both the possibility of HLA-peptide binding (binding model) and the potential immunogenicity (immunogenicity model) of the peptide-HLA complex (pHLA). The binding model achieves comparable performance with other well-acknowledged tools on the latest Immune Epitope Database (IEDB) benchmark datasets and an independent mass spectrometry (MS) dataset. The immunogenicity model could significantly improve the prediction precision of neoantigens. The further application of our method to the mutations with pre-existing T-cell responses indicating its feasibility in clinical application. DeepHLApan is freely available at https://github.com/jiujiezz/deephlapan and http://biopharm.zju.edu.cn/deephlapan.

Predicting HLA class II antigen presentation through integrated deep learning

Accurate prediction of antigen presentation by human leukocyte antigen (HLA) class II molecules would be valuable for vaccine development and cancer immunotherapies. Current computational methods trained on in vitro binding data are limited by insufficient training data and algorithmic constraints. Here we describe MARIA (major histocompatibility complex analysis with recurrent integrated architecture; https://maria.stanford.edu/ ), a multimodal recurrent neural network for predicting the likelihood of antigen presentation from a gene of interest in the context of specific HLA class II alleles. In addition to in vitro binding measurements, MARIA is trained on peptide HLA ligand sequences identified by mass spectrometry, expression levels of antigen genes and protease cleavage signatures. Because it leverages these diverse training data and our improved machine learning framework, MARIA (area under the curve = 0.89-0.92) outperformed existing methods in validation datasets. Across independent cancer neoantigen studies, peptides with high MARIA scores are more likely to elicit strong CD4+ T cell responses. MARIA allows identification of immunogenic epitopes in diverse cancers and autoimmune disease.

Robust prediction of HLA class II epitopes by deep motif deconvolution of immunopeptidomes

Predictions of epitopes presented by class II human leukocyte antigen molecules (HLA-II) have limited accuracy, restricting vaccine and therapy design. Here we combined unbiased mass spectrometry with a motif deconvolution algorithm to profile and analyze a total of 99,265 unique peptides eluted from HLA-II molecules. We then trained an epitope prediction algorithm with these data and improved prediction of pathogen and tumor-associated class II neoepitopes.

Next-generation computational tools for interrogating cancer immunity

The remarkable success of cancer therapies with immune checkpoint blockers is revolutionizing oncology and has sparked intensive basic and translational research into the mechanisms of cancer-immune cell interactions. In parallel, numerous novel cutting-edge technologies for comprehensive molecular and cellular characterization of cancer immunity have been developed, including single-cell sequencing, mass cytometry and multiplexed spatial cellular phenotyping. In order to process, analyse and visualize multidimensional data sets generated by these technologies, computational methods and software tools are required. Here, we review computational tools for interrogating cancer immunity, discuss advantages and limitations of the various methods and provide guidelines to assist in method selection.

A Phase Ib Study of the Combination of Personalized Autologous Dendritic Cell Vaccine, Aspirin, and Standard of Care Adjuvant Chemotherapy Followed by Nivolumab for Resected Pancreatic Adenocarcinoma-A Proof of Antigen Discovery Feasibility in Three Patients

Despite the promising therapeutic effects of immune checkpoint blockade (ICB), most patients with solid tumors treated with anti-PD-1/PD-L1 monotherapy do not achieve objective responses, with most tumor regressions being partial rather than complete. It is hypothesized that the absence of pre-existing antitumor immunity and/or the presence of additional tumor immune suppressive factors at the tumor microenvironment are responsible for such therapeutic failures. It is therefore clear that in order to fully exploit the potential of PD-1 blockade therapy, antitumor immune response should be amplified, while tumor immune suppression should be further attenuated. Cancer vaccines may prime patients for treatments with ICB by inducing effective anti-tumor immunity, especially in patients lacking tumor-infiltrating T-cells. These "non-inflamed" non-permissive tumors that are resistant to ICB could be rendered sensitive and transformed into "inflamed" tumor by vaccination. In this article we describe a clinical study where we use pancreatic cancer as a model, and we hypothesize that effective vaccination in pancreatic cancer patients, along with interventions that can reprogram important immunosuppressive factors in the tumor microenvironment, can enhance tumor immune recognition, thus enhancing response to PD-1/PD-L1 blockade. We incorporate into the schedule of standard of care (SOC) chemotherapy adjuvant setting a vaccine platform comprised of autologous dendritic cells loaded with personalized neoantigen peptides (PEP-DC) identified through our own proteo-genomics antigen discovery pipeline. Furthermore, we add nivolumab, an antibody against PD-1, to boost and maintain the vaccine's effect. We also demonstrate the feasibility of identifying personalized neoantigens in three pancreatic ductal adenocarcinoma (PDAC) patients, and we describe their optimal incorporation into long peptides for manufacturing into vaccine products. We finally discuss the advantages as well as the scientific and logistic challenges of such an exploratory vaccine clinical trial, and we highlight its novelty.

Alternative mRNA splicing in cancer immunotherapy

Immunotherapies are yielding effective treatments for several previously untreatable cancers. Still, the identification of suitable antigens specific to the tumour that can be targets for cancer vaccines and T cell therapies is a challenge. Alternative processing of mRNA, a phenomenon that has been shown to alter the proteomic diversity of many cancers, may offer the potential of a broadened target space. Here, we discuss the promise of analysing mRNA processing events in cancer cells, with an emphasis on mRNA splicing, for the identification of potential new targets for cancer immunotherapy. Further, we highlight the challenges that must be overcome for this new avenue to have clinical applicability.

Bioinformatic methods for cancer neoantigen prediction

Tumor cells accumulate aberrations not present in normal cells, leading to presentation of neoantigens on MHC molecules on their surface. These non-self neoantigens distinguish tumor cells from normal cells to the immune system and are thus targets for cancer immunotherapy. The rapid development of molecular profiling platforms, such as next-generation sequencing, has enabled the generation of large datasets characterizing tumor cells. The simultaneous development of algorithms has enabled rapid and accurate processing of these data. Bioinformatic software tools encoding the algorithms can be strung together in a workflow to identify neoantigens. Here, with a focus on high-throughput sequencing, we review state-of-the art bioinformatic tools along with the steps and challenges involved in neoantigen identification and recognition.

Determinants for Neoantigen Identification

All tumors accumulate genetic alterations, some of which can give rise to mutated, non-self peptides presented by human leukocyte antigen (HLA) molecules and elicit T-cell responses. These immunogenic mutated peptides, or neoantigens, are foreign in nature and display exquisite tumor specificity. The correlative evidence suggesting they play an important role in the effectiveness of various cancer immunotherapies has triggered the development of vaccines and adoptive T-cell therapies targeting them. However, the systematic identification of personalized neoantigens in cancer patients, a critical requisite for the success of these therapies, remains challenging. A growing amount of evidence supports that only a small fraction of all tumor somatic non-synonymous mutations (NSM) identified represent bona fide neoantigens; mutated peptides that are processed, presented on the cell surface HLA molecules of cancer cells and are capable of triggering immune responses in patients. Here, we provide an overview of the existing strategies to identify candidate neoantigens and to evaluate their immunogenicity, two factors that impact on neoantigen identification. We will focus on their strengths and limitations to allow readers to rationally select and apply the most suitable method for their specific laboratory setting.

Mass spectrometry-based identification of MHC-bound peptides for immunopeptidomics

Peptide antigens bound to molecules encoded by the major histocompatibility complex (MHC) and presented on the cell surface form the targets of T lymphocytes. This critical arm of the adaptive immune system facilitates the eradication of pathogen-infected and cancerous cells, as well as the production of antibodies. Methods to identify these peptide antigens are critical to the development of new vaccines, for which the goal is the generation of effective adaptive immune responses and long-lasting immune memory. Here, we describe a robust protocol for the identification of MHC-bound peptides from cell lines and tissues, using nano-ultra-performance liquid chromatography coupled to high-resolution mass spectrometry (nUPLC-MS/MS) and recent improvements in methods for isolation and characterization of these peptides. The protocol starts with the immunoaffinity capture of naturally processed MHC-peptide complexes. The peptides dissociate from the class I human leukocyte antigens (HLAs) upon acid denaturation. This peptide cargo is then extracted and separated into fractions by HPLC, and the peptides in these fractions are identified using nUPLC-MS/MS. With this protocol, several thousand peptides can be identified from a wide variety of cell types, including cancerous and infected cells and those from tissues, with a turnaround time of 2-3 d.

Targeting the MHC Ligandome by Use of TCR-Like Antibodies

Monoclonal antibodies (mAbs) are valuable as research reagents, in diagnosis and in therapy. Their high specificity, the ease in production, favorable biophysical properties and the opportunity to engineer different properties make mAbs a versatile class of biologics. mAbs targeting peptide–major histocompatibility molecule (pMHC) complexes are often referred to as “TCR-like” mAbs, as pMHC complexes are generally recognized by T-cell receptors (TCRs). Presentation of self- and non-self-derived peptide fragments on MHC molecules and subsequent activation of T cells dictate immune responses in health and disease. This includes responses to infectious agents or cancer but also aberrant responses against harmless self-peptides in autoimmune diseases. The ability of TCR-like mAbs to target specific peptides presented on MHC allows for their use to study peptide presentation or for diagnosis and therapy. This extends the scope of conventional mAbs, which are generally limited to cell-surface or soluble antigens. Herein, we review the strategies used to generate TCR-like mAbs and provide a structural comparison with the analogous TCR in pMHC binding. We further discuss their applications as research tools and therapeutic reagents in preclinical models as well as challenges and limitations associated with their use.

Going Beyond the Sequences: TCR Binding Patterns at the Service of Cancer Detection

Deep sequencing of T-cell receptors enables the comprehensive profiling of lymphocyte populations and the characterization of the repertoire of T-cell responses against tumors, which could be applied to diagnose cancers. Ostmeyer and colleagues introduce a novel approach to characterize TCR patterns correlating with antigen recognition. By projecting the large TCR sequence space into a handful of biophysicochemical descriptors for key residues and seeking TCRs with similar antigen-binding capabilities even in the absence of identical amino acids, this approach presents several advantages over current methods.See related article by Ostmeyer et al., p. 1671.

Predicting peptide presentation by major histocompatibility complex class I: an improved machine learning approach to the immunopeptidome

BACKGROUND

To further our understanding of immunopeptidomics, improved tools are needed to identify peptides presented by major histocompatibility complex class I (MHC-I). Many existing tools are limited by their reliance upon chemical affinity data, which is less biologically relevant than sampling by mass spectrometry, and other tools are limited by incomplete exploration of machine learning approaches. Herein, we assemble publicly available data describing human peptides discovered by sampling the MHC-I immunopeptidome with mass spectrometry and use this database to train random forest classifiers (ForestMHC) to predict presentation by MHC-I.

RESULTS

As measured by precision in the top 1% of predictions, our method outperforms NetMHC and NetMHCpan on test sets, and it outperforms both these methods and MixMHCpred on new data from an ovarian carcinoma cell line. We also find that random forest scores correlate monotonically, but not linearly, with known chemical binding affinities, and an information-based analysis of classifier features shows the importance of anchor positions for our classification. The random-forest approach also outperforms a deep neural network and a convolutional neural network trained on identical data. Finally, we use our large database to confirm that gene expression partially determines peptide presentation.

CONCLUSIONS

ForestMHC is a promising method to identify peptides bound by MHC-I. We have demonstrated the utility of random forest-based approaches in predicting peptide presentation by MHC-I, assembled the largest known database of MS binding data, and mined this database to show the effect of gene expression on peptide presentation. ForestMHC has potential applicability to basic immunology, rational vaccine design, and neoantigen binding prediction for cancer immunotherapy. This method is publicly available for applications and further validation.

Computational Methods for Identification of T Cell Neoepitopes in Tumors

Cancer immunotherapy has experienced several major breakthroughs in the past decade. Most recently, technical advances in next-generation sequencing methods have enabled discovery of tumor-specific mutations leading to protective T cell neoepitopes. Many of the successes are enabled by computational methods, which facilitate processing of raw data, mapping of mutations, and prediction of neoepitopes. In this book chapter, we provide an overview of the computational tasks related to the identification of neoepitopes, propose specific tools and best practices, and discuss strengths, weaknesses, and future challenges.

Identification of MHC Ligands and Establishing MHC Class I Peptide Motifs

MHC class I peptide motifs are used on a regular basis to identify and predict MHC class I ligands and CD8⁺ T cell epitopes. This approach is above all an invaluable tool for the identification of disease-associated epitopes ranging from pathogen associated epitopes, tumor associated natural and neoepitopes to autoimmune disease associated epitopes. As a matter of fact, the vast majority of T cell epitopes discovered during the past two decades was identified by means of epitope prediction and MHC ligand identification. Here we describe the steps which are necessary to identify MHC epitopes from monoallelic and multiallelic cells and establish MHC class I peptide motifs to compose a reliable scoring matrix for epitope prediction. As an example, the ligands of monoallelic C1R cells and multiallelic peripheral blood mononuclear cell tissue will be identified and a scoring matrix for the prediction of HLA-C*01:02-presented T cell epitopes will be developed.

Deep learning using tumor HLA peptide mass spectrometry datasets improves neoantigen identification

Neoantigens, which are expressed on tumor cells, are one of the main targets of an effective antitumor T-cell response. Cancer immunotherapies to target neoantigens are of growing interest and are in early human trials, but methods to identify neoantigens either require invasive or difficult-to-obtain clinical specimens, require the screening of hundreds to thousands of synthetic peptides or tandem minigenes, or are only relevant to specific human leukocyte antigen (HLA) alleles. We apply deep learning to a large (N = 74 patients) HLA peptide and genomic dataset from various human tumors to create a computational model of antigen presentation for neoantigen prediction. We show that our model, named EDGE, increases the positive predictive value of HLA antigen prediction by up to ninefold. We apply EDGE to enable identification of neoantigens and neoantigen-reactive T cells using routine clinical specimens and small numbers of synthetic peptides for most common HLA alleles. EDGE could enable an improved ability to develop neoantigen-targeted immunotherapies for cancer patients.

Footprints of antigen processing boost MHC class II natural ligand predictions

Background

Major histocompatibility complex class II (MHC-II) molecules present peptide fragments to T cells for immune recognition. Current predictors for peptide to MHC-II binding are trained on binding affinity data, generated in vitro and therefore lacking information about antigen processing.

Methods

We generate prediction models of peptide to MHC-II binding trained with naturally eluted ligands derived from mass spectrometry in addition to peptide binding affinity data sets.

Results

We show that integrated prediction models incorporate identifiable rules of antigen processing. In fact, we observed detectable signals of protease cleavage at defined positions of the ligands. We also hypothesize a role of the length of the terminal ligand protrusions for trimming the peptide to the MHC presented ligand.

Conclusions

The results of integrating binding affinity and eluted ligand data in a combined model demonstrate improved performance for the prediction of MHC-II ligands and T cell epitopes and foreshadow a new generation of improved peptide to MHC-II prediction tools accounting for the plurality of factors that determine natural presentation of antigens.

Electronic supplementary material

The online version of this article (10.1186/s13073-018-0594-6) contains supplementary material, which is available to authorized users.

Lost in the crowd: identifying targetable MHC class I neoepitopes for cancer immunotherapy

INTRODUCTION

The recent development of checkpoint blockade immunotherapy for cancer has led to impressive clinical results across multiple tumor types. There is mounting evidence that immune recognition of tumor derived MHC class I (MHC-I) restricted epitopes bearing cancer specific mutations and alterations is a crucial mechanism in successfully triggering immune-mediated tumor rejection. Therapeutic targeting of these cancer specific epitopes (neoepitopes) is emerging as a promising opportunity for the generation of personalized cancer vaccines and adoptive T cell therapies. However, one major obstacle limiting the broader application of neoepitope based therapies is the difficulty of selecting highly immunogenic neoepitopes among the wide array of presented non-immunogenic HLA ligands derived from self-proteins. Areas covered: In this review, we present an overview of the MHC-I processing and presentation pathway, as well as highlight key areas that contribute to the complexity of the associated MHC-I peptidome. We cover recent technological advances that simplify and optimize the identification of targetable neoepitopes for cancer immunotherapeutic applications. Expert commentary: Recent advances in computational modeling, bioinformatics, and mass spectrometry are unlocking the underlying mechanisms governing antigen processing and presentation of tumor-derived neoepitopes.

Systematically benchmarking peptide-MHC binding predictors: From synthetic to naturally processed epitopes

A number of machine learning-based predictors have been developed for identifying immunogenic T-cell epitopes based on major histocompatibility complex (MHC) class I and II binding affinities. Rationally selecting the most appropriate tool has been complicated by the evolving training data and machine learning methods. Despite the recent advances made in generating high-quality MHC-eluted, naturally processed ligandome, the reliability of new predictors on these epitopes has yet to be evaluated. This study reports the latest benchmarking on an extensive set of MHC-binding predictors by using newly available, untested data of both synthetic and naturally processed epitopes. 32 human leukocyte antigen (HLA) class I and 24 HLA class II alleles are included in the blind test set. Artificial neural network (ANN)-based approaches demonstrated better performance than regression-based machine learning and structural modeling. Among the 18 predictors benchmarked, ANN-based mhcflurry and nn_align perform the best for MHC class I 9-mer and class II 15-mer predictions, respectively, on binding/non-binding classification (Area Under Curves = 0.911). NetMHCpan4 also demonstrated comparable predictive power. Our customization of mhcflurry to a pan-HLA predictor has achieved similar accuracy to NetMHCpan. The overall accuracy of these methods are comparable between 9-mer and 10-mer testing data. However, the top methods deliver low correlations between the predicted versus the experimental affinities for strong MHC binders. When used on naturally processed MHC-ligands, tools that have been trained on elution data (NetMHCpan4 and MixMHCpred) shows better accuracy than pure binding affinity predictor. The variability of false prediction rate is considerable among HLA types and datasets. Finally, structure-based predictor of Rosetta FlexPepDock is less optimal compared to the machine learning approaches. With our benchmarking of MHC-binding and MHC-elution predictors using a comprehensive metrics, a unbiased view for establishing best practice of T-cell epitope predictions is presented, facilitating future development of methods in immunogenomics.

Estimating the Contribution of Proteasomal Spliced Peptides to the HLA-I Ligandome

It has been reported that about 30% of the HLA-I ligands are produced by proteasomal splicing of two noncontiguous fragments of a parental protein. We report that the identification of many of those spliced peptides is ambiguous. With an alternative workflow, based on de novo sequencing and subsequent verification with multiple search tools, we estimate that the upper bound for the proportion of cis-spliced peptides is 2–6%. Nevertheless, the true contribution of spliced peptides to the ligandome may be much smaller.

Spliced peptides are short protein fragments spliced together in the proteasome by peptide bond formation. True estimation of the contribution of proteasome-spliced peptides (PSPs) to the global human leukocyte antigen (HLA) ligandome is critical. A recent study suggested that PSPs contribute up to 30% of the HLA ligandome. We performed a thorough reanalysis of the reported results using multiple computational tools and various validation steps and concluded that only a fraction of the proposed PSPs passes the quality filters. To better estimate the actual number of PSPs, we present an alternative workflow. We performed de novo sequencing of the HLA-peptide spectra and discarded all de novo sequences found in the UniProt database. We checked whether the remaining de novo sequences could match spliced peptides from human proteins. The spliced sequences were appended to the UniProt fasta file, which was searched by two search tools at a false discovery rate (FDR) of 1%. We find that 2–6% of the HLA ligandome could be explained as spliced protein fragments. The majority of these potential PSPs have good peptide-spectrum match properties and are predicted to bind the respective HLA molecules. However, it remains to be shown how many of these potential PSPs actually originate from proteasomal splicing events.

Predicting T cell recognition of MHC class I restricted neoepitopes

Epitopes that arise from a somatic mutation, also called neoepitopes, are now known to play a key role in cancer immunology and immunotherapy. Recent advances in high-throughput sequencing have made it possible to identify all mutations and thereby all potential neoepitope candidates in an individual cancer. However, most of these neoepitope candidates are not recognized by T cells of cancer patients when tested in vivo or in vitro, meaning they are not immunogenic. Especially in patients with a high mutational load, usually hundreds of potential neoepitopes are detected, highlighting the need to further narrow down this candidate list. In our study, we assembled a dataset of known, naturally processed, immunogenic neoepitopes to dissect the properties that make these neoepitopes immunogenic. The tools to use and thresholds to apply for prioritizing neoepitopes have so far been largely based on experience with epitope identification in other settings such as infectious disease and allergy. Here, we performed a detailed analysis on our dataset of curated immunogenic neoepitopes to establish the appropriate tools and thresholds in the cancer setting. To this end, we evaluated different predictors for parameters that play a role in a neoepitope's immunogenicity and suggest that using binding predictions and length-rescaling yields the best performance in discriminating immunogenic neoepitopes from a background set of mutated peptides. We furthermore show that almost all neoepitopes had strong predicted binding affinities (as expected), but more surprisingly, the corresponding non-mutated peptides had nearly as high affinities. Our results provide a rational basis for parameters in neoepitope filtering approaches that are being commonly used. Abbreviations: SNV: single nucleotide variant; nsSNV: nonsynonymous single nucleotide variant; ROC: receiver operating characteristic; AUC: area under ROC curve; HLA: human leukocyte antigen; MHC: major histocompatibility complex; PD-1: Programmed cell death protein 1; PD-L1 or CTLA-4: cytotoxic T-lymphocyte associated protein 4.

In Silico Analysis of the Minor Histocompatibility Antigen Landscape Based on the 1000 Genomes Project

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is routinely used to treat hematopoietic malignancies. The eradication of residual tumor cells during engraftment is mediated by donor cytotoxic T lymphocytes reactive to alloantigens. In a HLA-matched transplantation context, alloantigens are encoded by various polymorphic genes situated outside the HLA locus, also called minor histocompatibility antigens (MiHAs). Recently, MiHAs have been recognized as promising targets for post-transplantation T-cell immunotherapy as they have several appealing advantages over tumor-associated antigens (TAAs) and neoantigens, i.e., they are more abundant than TAAs, which potentially facilitates multiple targeting; and unlike neoantigens, they are encoded by germline polymorphisms, some of which are common and thus, suitable for off-the-shelf therapy. The genetic sources of MiHAs are nonsynonymous polymorphisms that cause differences between the recipient and donor proteomes and subsequently, the immunopeptidomes. Systematic description of the alloantigen landscape in HLA-matched transplantation is still lacking as previous studies focused only on a few immunogenic and common MiHAs. Here, we perform a thorough in silico analysis of the public genomic data to classify genetic polymorphisms that lead to MiHA formation and estimate the number of potentially available MiHA mismatches. Our findings suggest that a donor/recipient pair is expected to have at least several dozen mismatched strong MHC-binding SNP-associated peptides per HLA allele (116 ± 26 and 65 ± 15 for non-related pairs and siblings respectively in European populations as predicted by two independent algorithms). Over 70% of them are encoded by relatively frequent polymorphisms (minor allele frequency > 0.1) and thus, may be targetable by off-the-shelf therapeutics. We showed that the most appealing targets (probability of mismatch over 20%) reside in the asymmetric allele frequency region, which spans from 0.15 to 0.47 and corresponds to an order of several hundred (213 ± 47) possible targets per HLA allele that can be considered for immunogenicity validation. Overall, these findings demonstrate the significant potential of MiHAs as targets for T-cell immunotherapy and emphasize the need for the systematic discovery of novel MiHAs.

Determination of a Predictive Cleavage Motif for Eluted Major Histocompatibility Complex Class II Ligands

CD4⁺ T cells have a major role in regulating immune responses. They are activated by recognition of peptides mostly generated from exogenous antigens through the major histocompatibility complex (MHC) class II pathway. Identification of epitopes is important and computational prediction of epitopes is used widely to save time and resources. Although there are algorithms to predict binding affinity of peptides to MHC II molecules, no accurate methods exist to predict which ligands are generated as a result of natural antigen processing. We utilized a dataset of around 14,000 naturally processed ligands identified by mass spectrometry of peptides eluted from MHC class II expressing cells to investigate the existence of sequence signatures potentially related to the cleavage mechanisms that liberate the presented peptides from their source antigens. This analysis revealed preferred amino acids surrounding both N- and C-terminuses of ligands, indicating sequence-specific cleavage preferences. We used these cleavage motifs to develop a method for predicting naturally processed MHC II ligands, and validated that it had predictive power to identify ligands from independent studies. We further confirmed that prediction of ligands based on cleavage motifs could be combined with predictions of MHC binding, and that the combined prediction had superior performance. However, when attempting to predict CD4⁺ T cell epitopes, either alone or in combination with MHC binding predictions, predictions based on the cleavage motifs did not show predictive power. Given that peptides identified as epitopes based on CD4⁺ T cell reactivity typically do not have well-defined termini, it is possible that motifs are present but outside of the mapped epitope. Our attempts to take that into account computationally did not show any sign of an increased presence of cleavage motifs around well-characterized CD4⁺ T cell epitopes. While it is possible that our attempts to translate the cleavage motifs in MHC II ligand elution data into T cell epitope predictions were suboptimal, other possible explanations are that the cleavage signal is too diluted to be detected, or that elution data are enriched for ligands generated through an antigen processing and presentation pathway that is less frequently utilized for T cell epitopes.

Predicting Antigen Presentation-What Could We Learn From a Million Peptides?

Antigen presentation lies at the heart of immune recognition of infected or malignant cells. For this reason, important efforts have been made to predict which peptides are more likely to bind and be presented by the human leukocyte antigen (HLA) complex at the surface of cells. These predictions have become even more important with the advent of next-generation sequencing technologies that enable researchers and clinicians to rapidly determine the sequences of pathogens (and their multiple variants) or identify non-synonymous genetic alterations in cancer cells. Here, we review recent advances in predicting HLA binding and antigen presentation in human cells. We argue that the very large amount of high-quality mass spectrometry data of eluted (mainly self) HLA ligands generated in the last few years provides unprecedented opportunities to improve our ability to predict antigen presentation and learn new properties of HLA molecules, as demonstrated in many recent studies of naturally presented HLA-I ligands. Although major challenges still lie on the road toward the ultimate goal of predicting immunogenicity, these experimental and computational developments will facilitate screening of putative epitopes, which may eventually help decipher the rules governing T cell recognition.

Computational Tools for the Identification and Interpretation of Sequence Motifs in Immunopeptidomes

Recent advances in proteomics and mass-spectrometry have widely expanded the detectable peptide repertoire presented by major histocompatibility complex (MHC) molecules on the cell surface, collectively known as the immunopeptidome. Finely characterizing the immunopeptidome brings about important basic insights into the mechanisms of antigen presentation, but can also reveal promising targets for vaccine development and cancer immunotherapy. This report describes a number of practical and efficient approaches to analyze immunopeptidomics data, discussing the identification of meaningful sequence motifs in various scenarios and considering current limitations. Guidelines are provided for the filtering of false hits and contaminants, and to address the problem of motif deconvolution in cell lines expressing multiple MHC alleles, both for the MHC class I and class II systems. Finally, it is demonstrated how machine learning can be readily employed by non-expert users to generate accurate prediction models directly from mass-spectrometry eluted ligand data sets.

The Role of Mass Spectrometry and Proteogenomics in the Advancement of HLA Epitope Prediction

A challenge in developing personalized cancer immunotherapies is the prediction of putative cancer-specific antigens. Currently, predictive algorithms are used to infer binding of peptides to human leukocyte antigen (HLA) heterodimers to aid in the selection of putative epitope targets. One drawback of current epitope prediction algorithms is that they are trained on datasets containing biochemical HLA-peptide binding data that may not completely capture the rules associated with endogenous processing and presentation. The field of MS has made great improvements in instrumentation speed and sensitivity, chromatographic resolution, and proteogenomic database search strategies to facilitate the identification of HLA-ligands from a variety of cell types and tumor tissues. As such, these advances have enabled MS profiling of HLA-binding peptides to be a tractable, orthogonal approach to lower throughput biochemical assays for generating comprehensive datasets to train epitope prediction algorithms. In this review, we will highlight the progress made in the field of HLA-ligand profiling enabled by MS and its impact on current and future epitope prediction strategies.

Polymorphic sites preferentially avoid co-evolving residues in MHC class I proteins

Major histocompatibility complex class I (MHC-I) molecules are critical to adaptive immune defence mechanisms in vertebrate species and are encoded by highly polymorphic genes. Polymorphic sites are located close to the ligand-binding groove and entail MHC-I alleles with distinct binding specificities. Some efforts have been made to investigate the relationship between polymorphism and protein stability. However, less is known about the relationship between polymorphism and MHC-I co-evolutionary constraints. Using Direct Coupling Analysis (DCA) we found that co-evolution analysis accurately pinpoints structural contacts, although the protein family is restricted to vertebrates and comprises less than five hundred species, and that the co-evolutionary signal is mainly driven by inter-species changes, and not intra-species polymorphism. Moreover, we show that polymorphic sites in human preferentially avoid co-evolving residues, as well as residues involved in protein stability. These results suggest that sites displaying high polymorphism may have been selected during vertebrates’ evolution to avoid co-evolutionary constraints and thereby maximize their mutability.

Amino acid co-evolution represents cases of simultaneous substitution of amino acids at distinct positions in protein sequences. In the MHC-I protein family, such co-evolution could result from either amino acid changes across species or changes within species due to the high polymorphism of MHC-I molecules. Here we show that signals captured by global methods such as Direct Coupling Analysis (DCA) to estimate co-evolution primarily result from changes across species. Moreover, our results indicate that polymorphic sites in MHC-I molecules tend to be decoupled from co-evolving ones. This could suggest that they have been selected to maximize their mutability, which is known to be functionally important to entail MHC-I molecules with a wide repertoire of binding specificities for antigen presentation.

Mapping the tumour human leukocyte antigen (HLA) ligandome by mass spectrometry

The entirety of human leukocyte antigen (HLA)-presented peptides is referred to as the HLA ligandome of a cell or tissue, in tumours often termed immunopeptidome. Mapping the tumour immunopeptidome by mass spectrometry (MS) comprehensively views the pathophysiologically relevant antigenic signature of human malignancies. MS is an unbiased approach stringently filtering the candidates to be tested as opposed to epitope prediction algorithms. In the setting of peptide-specific immunotherapies, MS-based strategies significantly diminish the risk of lacking clinical benefit, as they yield highly enriched amounts of truly presented peptides. Early immunopeptidomic efforts were severely limited by technical sensitivity and manual spectra interpretation. The technological progress with development of orbitrap mass analysers and enhanced chromatographic performance led to vast improvements in mass accuracy, sensitivity, resolution, and speed. Concomitantly, bioinformatic tools were developed to process MS data, integrate sequencing results, and deconvolute multi-allelic datasets. This enabled the immense advancement of tumour immunopeptidomics. Studying the HLA-presented peptide repertoire bears high potential for both answering basic scientific questions and translational application. Mapping the tumour HLA ligandome has started to significantly contribute to target identification for the design of peptide-specific cancer immunotherapies in clinical trials and compassionate need treatments. In contrast to prediction algorithms, rare HLA allotypes and HLA class II can be adequately addressed when choosing MS-guided target identification platforms. Herein, we review the identification of tumour HLA ligands focusing on sources, methods, bioinformatic data analysis, translational application, and provide an outlook on future developments.

The C-terminal extension landscape of naturally presented HLA-I ligands

HLA-I molecules play a central role in antigen presentation. They typically bind 9- to 12-mer peptides, and their canonical binding mode involves anchor residues at the second and last positions of their ligands. To investigate potential noncanonical binding modes, we collected in-depth and accurate HLA peptidomics datasets covering 54 HLA-I alleles and developed algorithms to analyze these data. Our results reveal frequent (442 unique peptides) and statistically significant C-terminal extensions for at least eight alleles, including the common HLA-A03:01, HLA-A31:01, and HLA-A68:01. High resolution crystal structure of HLA-A68:01 with such a ligand uncovers structural changes taking place to accommodate C-terminal extensions and helps unraveling sequence and structural properties predictive of the presence of these extensions. Scanning viral proteomes with the C-terminal extension motifs identifies many putative epitopes and we demonstrate direct recognition by human CD8⁺ T cells of a 10-mer epitope from cytomegalovirus predicted to follow the C-terminal extension binding mode.

Membranal and Blood-Soluble HLA Class II Peptidome Analyses Using Data-Dependent and Independent Acquisition

The interaction between HLA class II peptide complexes on antigen-presenting cells and CD4⁺ T cells is of fundamental importance for anticancer and antipathogen immunity as well as for the maintenance of immunological tolerance. To study CD4⁺ T cell reactivities, detailed knowledge of the presented peptides is necessary. In recent years, dramatic advances in the characterization of membranal and soluble HLA class I peptidomes could be observed. However, the same is not true for HLA class II peptidomes, where only few studies identify more than hundred peptides. Here we describe a MS-based workflow for the characterization of membranal and soluble HLA class II DR and DQ peptidomes. Using this workflow, we identify a total of 8595 and 3727 HLA class II peptides from Maver-1 and DOHH2 cells, respectively. Based on this data, a motif-based binding predictor is developed and compared to NetMHCIIpan 3.1. We then apply the workflow to human plasma, resulting in the identification of between 34 and 152 HLA-DR and between 100 and 180 HLA-DQ peptides, respectively. Finally, we implement a data-independent acquisition workflow to increase reproducibility and sensitivity of HLA class II peptidome characterizations.

High-throughput and Sensitive Immunopeptidomics Platform Reveals Profound Interferonγ-Mediated Remodeling of the Human Leukocyte Antigen (HLA) Ligandome

Comprehensive knowledge of the human leukocyte antigen (HLA) class-I and class-II peptides presented to T-cells is crucial for designing innovative therapeutics against cancer and other diseases. However methodologies for their purification for mass-spectrometry analysis have been a major limitation. We designed a novel high-throughput, reproducible and sensitive method for sequential immuno-affinity purification of HLA-I and -II peptides from up to 96 samples in a plate format, suitable for both cell lines and tissues. Our methodology drastically reduces sample-handling and can be completed within five hours. We challenged our methodology by extracting HLA peptides from multiple replicates of tissues (n = 7) and cell lines (n = 21, 108 cells per replicate), which resulted in unprecedented depth, sensitivity and high reproducibility (Pearson correlations up to 0.98 and 0.97 for HLA-I and HLA-II). Because of the method's achieved sensitivity, even single measurements of peptides purified from 107 B-cells resulted in the identification of more than 1700 HLA-I and 2200 HLA-II peptides. We demonstrate the feasibility of performing drug-screening by using ovarian cancer cells treated with interferon gamma (IFNγ). Our analysis revealed an augmented presentation of chymotryptic-like and longer ligands associated with IFNγ induced changes of the antigen processing and presentation machinery. This straightforward method is applicable for basic and clinical applications.

Improved Prediction of Bovine Leucocyte Antigens (BoLA) Presented Ligands by Use of Mass-Spectrometry-Determined Ligand and in Vitro Binding Data

Peptide binding to MHC class I molecules is the single most selective step in antigen presentation and the strongest single correlate to peptide cellular immunogenicity. The cost of experimentally characterizing the rules of peptide presentation for a given MHC-I molecule is extensive, and predictors of peptide–MHC interactions constitute an attractive alternative. Recently, an increasing amount of MHC presented peptides identified by mass spectrometry (MS ligands) has been published. Handling and interpretation of MS ligand data is, in general, challenging due to the polyspecificity nature of the data. We here outline a general pipeline for dealing with this challenge and accurately annotate ligands to the relevant MHC-I molecule they were eluted from by use of GibbsClustering and binding motif information inferred from in silico models. We illustrate the approach here in the context of MHC-I molecules (BoLA) of cattle. Next, we demonstrate how such annotated BoLA MS ligand data can readily be integrated with in vitro binding affinity data in a prediction model with very high and unprecedented performance for identification of BoLA-I restricted T-cell epitopes. The prediction model is freely available at http://www.cbs.dtu.dk/services/NetMHCpan/NetBoLApan. The approach has here been applied to the BoLA-I system, but the pipeline is readily applicable to MHC systems in other species.

'Hotspots' of Antigen Presentation Revealed by Human Leukocyte Antigen Ligandomics for Neoantigen Prioritization

The remarkable clinical efficacy of the immune checkpoint blockade therapies has motivated researchers to discover immunogenic epitopes and exploit them for personalized vaccines. Human leukocyte antigen (HLA)-binding peptides derived from processing and presentation of mutated proteins are one of the leading targets for T-cell recognition of cancer cells. Currently, most studies attempt to identify neoantigens based on predicted affinity to HLA molecules, but the performance of such prediction algorithms is rather poor for rare HLA class I alleles and for HLA class II. Direct identification of neoantigens by mass spectrometry (MS) is becoming feasible; however, it is not yet applicable to most patients and lacks sensitivity. In an attempt to capitalize on existing immunopeptidomics data and extract information that could complement HLA-binding prediction, we first compiled a large HLA class I and class II immunopeptidomics database across dozens of cell types and HLA allotypes and detected hotspots that are subsequences of proteins frequently presented. About 3% of the peptidome was detected in both class I and class II. Based on the gene ontology of their source proteins and the peptide's length, we propose that their processing may partake by the cellular class II presentation machinery. Our database captures the global nature of the in vivo peptidome averaged over many HLA alleles, and therefore, reflects the propensity of peptides to be presented on HLA complexes, which is complementary to the existing neoantigen prediction features such as binding affinity and stability or RNA abundance. We further introduce two immunopeptidomics MS-based features to guide prioritization of neoantigens: the number of peptides matching a protein in our database and the overlap of the predicted wild-type peptide with other peptides in our database. We show as a proof of concept that our immunopeptidomics MS-based features improved neoantigen prioritization by up to 50%. Overall, our work shows that, in addition to providing huge training data to improve the HLA binding prediction, immunopeptidomics also captures other aspects of the natural in vivo presentation that significantly improve prediction of clinically relevant neoantigens.

NetMHCpan-4.0: Improved Peptide-MHC Class I Interaction Predictions Integrating Eluted Ligand and Peptide Binding Affinity Data

Cytotoxic T cells are of central importance in the immune system's response to disease. They recognize defective cells by binding to peptides presented on the cell surface by MHC class I molecules. Peptide binding to MHC molecules is the single most selective step in the Ag-presentation pathway. Therefore, in the quest for T cell epitopes, the prediction of peptide binding to MHC molecules has attracted widespread attention. In the past, predictors of peptide-MHC interactions have primarily been trained on binding affinity data. Recently, an increasing number of MHC-presented peptides identified by mass spectrometry have been reported containing information about peptide-processing steps in the presentation pathway and the length distribution of naturally presented peptides. In this article, we present NetMHCpan-4.0, a method trained on binding affinity and eluted ligand data leveraging the information from both data types. Large-scale benchmarking of the method demonstrates an increase in predictive performance compared with state-of-the-art methods when it comes to identification of naturally processed ligands, cancer neoantigens, and T cell epitopes.