Is the exquisite specificity of lymphocytes generated by thymic selection or due to evolution?

We have previously argued that the antigen receptors of T and B lymphocytes evolved to be sufficiently specific to avoid massive deletion of clonotypes by negative selection. Their optimal 'specificity' level, i.e., probability of binding any particular epitope, was shown to be inversely related to the number of self-antigens that the cells have to be tolerant to. Experiments have demonstrated that T lymphocytes also become more specific during negative selection in the thymus, because cells expressing the most crossreactive receptors have the highest likelihood of binding a self-antigen, and hence to be tolerized (i.e., deleted, anergized, or diverted into a regulatory T cell phenotype). Thus, there are two -not mutually exclusive- explanations for the exquisite specificity of T cells, one involving evolution and the other thymic selection. To better understand the impact of both, we extend a previously developed mathematical model by allowing for T cells with very different binding probabilities in the pre-selection repertoire. We confirm that negative selection tends to tolerize the most crossreactive clonotypes. As a result, the average level of specificity in the functional post-selection repertoire depends on the number of self-antigens, even if there is no evolutionary optimization of binding probabilities. However, the evolutionary optimal range of binding probabilities in the pre-selection repertoire also depends on the number of self-antigens. Species with more self antigens need more specific pre-selection repertoires to avoid excessive loss of T cells during thymic selection, and hence mount protective immune responses. We conclude that both evolution and negative selection are responsible for the high level of specificity of lymphocytes.

Role of T cells in severe COVID-19 disease, protection, and long term immunity

Infection with SARS-CoV-2 causes wide range of disease severities from asymptomatic to life-threatening disease. Understanding the contribution of immunological traits in immunity against SARS-CoV-2 and in protection against severe COVID-19 could result in effective measures to prevent development of severe disease. While the role of cytokines and antibodies has been thoroughly studied, this is not the case for T cells. In this review, the association between T cells and COVID-19 disease severity and protection upon reexposure is discussed. While infiltration of overactivated cytotoxic T cells might be harmful in the infected tissue, fast responding T cells are important in the protection against severe COVID-19. This protection could even be viable in the long term as long-living memory T cells seem to be stabilized and mutations do not appear to have a large impact on T cell responses. Thus, after vaccination and infections, memory T cells should be able to help prevent onset of severe disease for most cases. Considering this, it would be useful to add N or M proteins in vaccinations, alongside the S protein which is currently used, as this results in a broader T cell response.

VDJdb in 2019: database extension, new analysis infrastructure and a T-cell receptor motif compendium

Here, we report an update of the VDJdb database with a substantial increase in the number of T-cell receptor (TCR) sequences and their cognate antigens. The update further provides a new database infrastructure featuring two additional analysis modes that facilitate database querying and real-world data analysis. The increased yield of TCR specificity identification methods and the overall increase in the number of studies in the field has allowed us to expand the database more than 5-fold. Furthermore, several new analysis methods are included. For example, batch annotation of TCR repertoire sequencing samples allows for annotating large datasets on-line. Using recently developed bioinformatic methods for TCR motif mining, we have built a reduced set of high-quality TCR motifs that can be used for both training TCR specificity predictors and matching against TCRs of interest. These additions enhance the versatility of the VDJdb in the task of exploring T-cell antigen specificities. The database is available at https://vdjdb.cdr3.net.

Revealing factors determining immunodominant responses against dominant epitopes

Upon recognition of peptide-MHC complexes by T cell receptors (TCR), the cognate T cells expand and differentiate into effector T cells to generate protective immunity. Despite the fact that any immune response generates a diverse set of TCR clones against a particular epitope, only a few clones are highly expanded in any immune response. Previous studies observed that the highest frequency clones usually control viral infections better than subdominant clones, but the reasons for this dominance among T cell clones are still unclear. Here, we used publicly available TCR amino acid sequences to study which factors determine whether a response becomes immunodominance (ID) per donor; we classified the largest T cell clone as the epitope-specific dominant clone and all the other clones as subdominant responses (SD). We observed a distinctively hydrophobic CDR3 in ID responses against a dominant epitope from influenza A virus, compared to the SD responses. The common V-J combinations were shared between ID and SD responses, suggesting that the biased V-J recombination events are restricted by epitope specificity; thus, the immunodominance is not directly determined by a bias combination of V and J genetic segments. Our findings reveal a close similarity of global sequence properties between dominant and subdominant clones of epitope-specific responses but detectable distinctive amino acid enrichments in ID. Taken together, we believe this first comparative study of immunodominant and subdominant TCR sequences can guide further studies to resolve factors determining the immunodominance of antiviral as well as tumor-specific T cell responses.

VDJdb: a curated database of T-cell receptor sequences with known antigen specificity

The ability to decode antigen specificities encapsulated in the sequences of rearranged T-cell receptor (TCR) genes is critical for our understanding of the adaptive immune system and promises significant advances in the field of translational medicine. Recent developments in high-throughput sequencing methods (immune repertoire sequencing technology, or RepSeq) and single-cell RNA sequencing technology have allowed us to obtain huge numbers of TCR sequences from donor samples and link them to T-cell phenotypes. However, our ability to annotate these TCR sequences still lags behind, owing to the enormous diversity of the TCR repertoire and the scarcity of available data on T-cell specificities. In this paper, we present VDJdb, a database that stores and aggregates the results of published T-cell specificity assays and provides a universal platform that couples antigen specificities with TCR sequences. We demonstrate that VDJdb is a versatile instrument for the annotation of TCR repertoire data, enabling a concatenated view of antigen-specific TCR sequence motifs. VDJdb can be accessed at https://vdjdb.cdr3.net and https://github.com/antigenomics/vdjdb-db.

Abstract B022: Properties of T-cell-recognized neoantigens

Over the past years we have learned that the T-cell-based immune system frequently responds to the neoantigens that arise as a consequence of the accumulated DNA damage causing the malignant transformation. Furthermore, recognition of neoantigens appears an important driver of the clinical activity of both T-cell checkpoint blockade and adoptive T-cell therapy as cancer immunotherapies. From the efforts dissecting the neoantigen-specific T-cell response it has become clear that only a very minor fraction of the accumulated mutations is recognized by the immune system, and the challenge to unravel the neoantigen-specific T-cell response lies in identifying which neoantigens are more likely to be true T-cell epitopes. We have analyzed neoantigen-specific T-cell reactivity in 12 melanoma patients using an in silico epitope prediction pipeline based on RNA expression, predicted HLA binding affinity, proteasomal processing and self-similarity to predict potential neoepitopes. We screened for T-cell recognition of 7000 epitopes from these 12 patients (average ~550 epitopes per patient, range: 96-1902) using our pMHC multimer combinatorial encoding technology and found 19 epitopes to be recognized by T-cells (hits) and 6981 to be “non-hits.” Based on these data we have examined the properties of T-cell recognized neoantigens. An intriguing observation is an enrichment within T-cell recognized epitopes of epitopes with the mutation positioned within the last 4 amino acids (C-terminal end of the peptide) compared to the screened set of epitopes. Fifteen out of 19 hits (approximately 80%) harbored a mutation within the last 4 amino acids of the peptide, whereas within the full set of screen epitopes it is 43%. While it is currently unclear what the reason is for this, this could reflect a biologic importance in T-cell recognition of the C-terminal part of the epitope. Furthermore, RNA expression and predicted binding affinity to HLA are important informative parameters for selecting T-cell recognized epitopes. A striking observation is that predicted binding affinity not only correlates with likelihood of observing a T-cell response but also the magnitude of this T-cell response, suggesting a hierarchy within neoantigens, and that not all neoantigens are of equal immunologic quality. In summary, our findings indicate that T-cell recognized neoantigens may differ from the neoantigen pool not recognized. In particular regarding position of the mutation with the epitope, RNA abundance and predicted HLA binding affinity. Importantly, our data reveal a hierarchy within neoantigens comparable to immunodominance known from viral infections. This hierarchy appears to depend mostly on binding affinity. These observations are likely to be highly relevant when selecting neoantigens for therapeutic manipulation such as vaccines.

Citation Format: Pia Kvistborg, Marit M. van Buuren, Daisy Philips, Nienke van Rooij, Arno Velds, Sam Behjati, Marlous van den Braber, Mireille Toebes, Lorenzo Fanchi, Maarten Slagter, Marie Stentoft Svane, Patrick Hwu, Joost van den Berg, Michael Stratton, Christian Blank, John B.A.G. Haanen, Can Kesmir, Ton N.M. Schumacher. Properties of T-cell-recognized neoantigens [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr B022.

Immune biomarkers for predicting response to adoptive cell transfer as cancer treatment

Adoptive cell transfer (ACT) is a form of personalised immunotherapy which has shown promising results in metastasised cancer. For this treatment, autologous T lymphocytes are selected and stimulated in vitro before re-administration in large numbers. However, only a fraction of patients benefit from ACT, and it is not yet known what biomarkers can predict treatment outcome. In this review, we describe what tumour characteristics are associated with response to ACT. Based on the current knowledge, the best candidate biomarker for a good anti-tumour response seems to be a large number of neoantigens with a homogeneous distribution across the tumour in combination with sufficient MHC-I expression level. Additionally, it is necessary to be able to isolate a diverse population of T cells reactive to these neoantigens from tumour tissue or peripheral blood. Additional promising candidate biomarkers shared with other cancer immunotherapies are a large number of tumour-infiltrating cytotoxic and memory T cells, normal levels of glycolysis, and a pro-inflammatory cytokine profile within the tumour. Intense research in this field will hopefully result in identification of more biomarkers for cancers with low mutational load.

Comprehensive Analysis of the Naturally Processed Peptide Repertoire: Differences between HLA-A and B in the Immunopeptidome

The cytotoxic T cell (CTL) response is determined by the peptide repertoire presented by the HLA class I molecules of an individual. We performed an in-depth analysis of the peptide repertoire presented by a broad panel of common HLA class I molecules on four B lymphoblastoid cell-lines (BLCL). Peptide elution and mass spectrometry analysis were utilised to investigate the number and abundance of self-peptides. Altogether, 7897 unique self-peptides, derived of 4344 proteins, were eluted. After viral infection, the number of unique self-peptides eluted significantly decreased compared to uninfected cells, paralleled by a decrease in the number of source proteins. In the overall dataset, the total number of unique self-peptides eluted from HLA-B molecules was larger than from HLA-A molecules, and they were derived from a larger number of source proteins. These results in B cells suggest that HLA-B molecules possibly present a more diverse repertoire compared to their HLA-A counterparts, which may contribute to their immunodominance. This study provides a unique data set giving new insights into the complex system of antigen presentation for a broad panel of HLA molecules, many of which were never studied this extensively before.

Can Selective MHC Downregulation Explain the Specificity and Genetic Diversity of NK Cell Receptors?

Natural killer (NK) cells express inhibiting receptors (iNKRs), which specifically bind MHC-I molecules on the surface of healthy cells. When the expression of MHC-I on the cell surface decreases, which might occur during certain viral infections and cancer, iNKRs lose inhibiting signals and the infected cells become target for NK cell activation (missing-self detection). Although the detection of MHC-I deficient cells can be achieved by conserved receptor-ligand interactions, several iNKRs are encoded by gene families with a remarkable genetic diversity, containing many haplotypes varying in gene content and allelic polymorphism. So far, the biological function of this expansion within the NKR cluster has remained poorly understood. Here, we investigate whether the evolution of diverse iNKRs genes can be driven by a specific viral immunoevasive mechanism: selective MHC downregulation. Several viruses, including EBV, CMV, and HIV, decrease the expression of MHC-I to escape from T cell responses. This downregulation does not always affect all MHC loci in the same way, as viruses target particular MHC molecules. To study the selection pressure of selective MHC downregulation on iNKRs, we have developed an agent-based model simulating an evolutionary scenario of hosts infected with herpes-like viruses, which are able to selectively downregulate the expression of MHC-I molecules on the cell surface. We show that iNKRs evolve specificity and, depending on the similarity of MHC alleles within each locus and the differences between the loci, they can specialize to a particular MHC-I locus. The easier it is to classify an MHC allele to its locus, the lower the required diversity of the NKRs. Thus, the diversification of the iNKR cluster depends on the locus specific MHC structure.

Biomarkers in cancer immunotherapy

Antibodies against T cell checkpoint molecules have started to revolutionize cancer treatment. Nevertheless, less than half of all patients respond to these immunotherapies. Recent work supports the potential value of biomarkers that predict therapy outcome and inspires the development of assay systems that interrogate other aspects of the cancer-immunity cycle.

MR1-restricted MAIT cells display ligand discrimination and pathogen selectivity through distinct T cell receptor usage

MAIT cells can discriminate between pathogen-derived ligands in a clonotype-dependent manner, and the TCR repertoire is distinct within individuals, indicating that the MAIT cell repertoire is shaped by prior microbial exposure.

Mucosal-associated invariant T (MAIT) cells express a semi-invariant T cell receptor (TCR) that detects microbial metabolites presented by the nonpolymorphic major histocompatibility complex (MHC)–like molecule MR1. The highly conserved nature of MR1 in conjunction with biased MAIT TCRα chain usage is widely thought to indicate limited ligand presentation and discrimination within a pattern-like recognition system. Here, we evaluated the TCR repertoire of MAIT cells responsive to three classes of microbes. Substantial diversity and heterogeneity were apparent across the functional MAIT cell repertoire as a whole, especially for TCRβ chain sequences. Moreover, different pathogen-specific responses were characterized by distinct TCR usage, both between and within individuals, suggesting that MAIT cell adaptation was a direct consequence of exposure to various exogenous MR1-restricted epitopes. In line with this interpretation, MAIT cell clones with distinct TCRs responded differentially to a riboflavin metabolite. These results suggest that MAIT cells can discriminate between pathogen-derived ligands in a clonotype-dependent manner, providing a basis for adaptive memory via recruitment of specific repertoires shaped by microbial exposure.

Complementarity of Binding Motifs is a General Property of HLA-A and HLA-B Molecules and Does Not Seem to Effect HLA Haplotype Composition

Different human leukocyte antigen (HLA) haplotypes (i.e., the specific combinations of HLA-A, -B, -DR alleles inherited together from one parent) are observed in different frequencies in human populations. Some haplotypes, like HLA-A1-B8, are very frequent, reaching up to 10% in the Caucasian population, while others are very rare. Numerous studies have identified associations between HLA haplotypes and diseases, and differences in haplotype frequencies can in part be explained by these associations: the stronger the association with a severe (autoimmune) disease, the lower the expected HLA haplotype frequency. The peptide repertoires of the HLA molecules composing a haplotype can also influence the frequency of a haplotype. For example, it would seem advantageous to have HLA molecules with non-overlapping binding specificities within a haplotype, as individuals expressing such an haplotype would present a diverse set of peptides from viruses and pathogenic bacteria on the cell surface. To test this hypothesis, we collect the proteome data from a set of common viruses, and estimate the total ligand repertoire of HLA class I haplotypes (HLA-A-B) using in silico predictions. We compare the size of these repertoires to the HLA haplotype frequencies reported in the National Marrow Donor Program (NMDP). We find that in most HLA-A and HLA-B pairs have fairly distinct binding motifs, and that the observed haplotypes do not contain HLA-A and -B molecules with more distinct binding motifs than random HLA-A and HLA-B pairs. In addition, the population frequency of a haplotype is not correlated to the distinctness of its HLA-A and HLA-B peptide binding motifs. These results suggest that there is a not a strong selection pressure on the haplotype level favoring haplotypes having HLA molecules with distinct binding motifs, which would result the largest possible presented peptide repertoires in the context of infectious diseases.

Use of tumor exome analysis to reveal neo-antigen-specific T-cell reactivity in ipilimumab-responsive melanoma

9085

Background: Evidence for T cell mediated regression of human cancer in particular melanoma following immunotherapy is strong. Anti-CTLA4 treatment has been approved for treatment of metastatic melanoma and blockade of PD-1 has shown encouraging results. However, it is unknown which T cell reactivities are involved in cancer regression. Reactivity against non-mutated tumor self-antigens has been analyzed in patients treated with Ipilimumab or with autologous TILs, but the size of these responses are modest. Therefore, T cell recognition of patient-specific mutant epitopes may be a potentially important component. Animal model data recently suggested that analysis of T cell reactivity against patient-specific neo-antigens may be feasible through exploitation of cancer genome data. However, human data have thus far been lacking. Methods: To address this we have used MHC class I peptide exchange technology allowing production of very large collections of pMHC complexes, together with a pMHC "combinatorial coding" strategy for parallel detection of dozens of different T cell populations within a single sample. Results: From a melanoma patient responding to ipilimumab treatment, we identified tumor specific mutations via exome sequencing of tumor material. The exome contained 1,075 non-synonymous mutations. Possible MHC epitopes covering these mutations were predicted based on; 1) predicted to bind the patient’s MHC; 2) predicted to be cleaved by the proteasome; 3) genes of which the mutated peptides arose had evidence of RNA expression. The analysis yielded 1,952 epitopes restricted to the HLA-A and HLA-B. To screen for T cell reactivity against these epitopes we used the pMHC combinatorial coding approach. We found T cell reactivity against 2 neo-antigens, including a dominant T cell response against a mutant epitope of the ATR gene product. Analysis of PBMC samples collected before and during Ipilimumab therapy showed that this particular response increased strongly after treatment from 0.06% to 0.28% of CD8 T cells after being stable in magnitude for 10 months. Conclusions: These data provide the first demonstration of cancer exome-guided analysis to dissect the effects of melanoma immunotherapy.

A comparative study of HLA binding affinity and ligand diversity: implications for generating immunodominant CD8+ T cell responses

Conventional CD8(+) T cell responses against intracellular infectious agents are initiated upon recognition of pathogen-derived peptides presented at the cell surface of infected cells in the context of MHC class I molecules. Among the major MHC class I loci, HLA-B is the swiftest evolving and the most polymorphic locus. Additionally, responses restricted by HLA-B molecules tend to be dominant, and most associations with susceptibility or protection against infectious diseases have been assigned to HLA-B alleles. To assess whether the differences in responses mediated via two major HLA class I loci, HLA-B and HLA-A, may already begin at the Ag presentation level, we have analyzed the diversity and binding affinity of their peptide repertoire by making use of curated pathogen-derived epitope data retrieved from the Immune Epitope Database and Analysis Resource, as well as in silico predicted epitopes. In contrast to our expectations, HLA-B alleles were found to have a less diverse peptide repertoire, which points toward a more restricted binding motif, and the respective average peptide binding affinity was shown to be lower than that of HLA-A-restricted epitopes. This unexpected observation gives rise to new hypotheses concerning the mechanisms underlying immunodominance of CD8(+) T cell responses.

Amino acid similarity accounts for T cell cross-reactivity and for "holes" in the T cell repertoire

BACKGROUND

Cytotoxic T cell (CTL) cross-reactivity is believed to play a pivotal role in generating immune responses but the extent and mechanisms of CTL cross-reactivity remain largely unknown. Several studies suggest that CTL clones can recognize highly diverse peptides, some sharing no obvious sequence identity. The emerging realization in the field is that T cell receptors (TcR) recognize multiple distinct ligands.

PRINCIPAL FINDINGS

First, we analyzed peptide scans of the HIV epitope SLFNTVATL (SFL9) and found that TCR specificity is position dependent and that biochemically similar amino acid substitutions do not drastically affect recognition. Inspired by this, we developed a general model of TCR peptide recognition using amino acid similarity matrices and found that such a model was able to predict the cross-reactivity of a diverse set of CTL epitopes. With this model, we were able to demonstrate that seemingly distinct T cell epitopes, i.e., ones with low sequence identity, are in fact more biochemically similar than expected. Additionally, an analysis of HIV immunogenicity data with our model showed that CTLs have the tendency to respond mostly to peptides that do not resemble self-antigens.

CONCLUSIONS

T cell cross-reactivity can thus, to an extent greater than earlier appreciated, be explained by amino acid similarity. The results presented in this paper will help resolving some of the long-lasting discussions in the field of T cell cross-reactivity.

Increased sequence diversity coverage improves detection of HIV-specific T cell responses

The accurate identification of HIV-specific T cell responses is important for determining the relationship between immune response, viral control, and disease progression. HIV-specific immune responses are usually measured using peptide sets based on consensus sequences, which frequently miss responses to regions where test set and infecting virus differ. In this study, we report the design of a peptide test set with significantly increased coverage of HIV sequence diversity by including alternative amino acids at variable positions during the peptide synthesis step. In an IFN-gamma ELISpot assay, these "toggled" peptides detected HIV-specific CD4(+) and CD8(+) T cell responses of significantly higher breadth and magnitude than matched consensus peptides. The observed increases were explained by a closer match of the toggled peptides to the autologous viral sequence. Toggled peptides therefore afford a cost-effective and significantly more complete view of the host immune response to HIV and are directly applicable to other variable pathogens.

Modeling the adaptive immune system: predictions and simulations

MOTIVATION

Immunological bioinformatics methods are applicable to a broad range of scientific areas. The specifics of how and where they might be implemented have recently been reviewed in the literature. However, the background and concerns for selecting between the different available methods have so far not been adequately covered.

SUMMARY

Before using predictions systems, it is necessary to not only understand how the methods are constructed but also their strength and limitations. The prediction systems in humoral epitope discovery are still in their infancy, but have reached a reasonable level of predictive strength. In cellular immunology, MHC class I binding predictions are now very strong and cover most of the known HLA specificities. These systems work well for epitope discovery, and predictions of the MHC class I pathway have been further improved by integration with state-of-the-art prediction tools for proteasomal cleavage and TAP binding. By comparison, class II MHC binding predictions have not developed to a comparable accuracy level, but new tools have emerged that deliver significantly improved predictions not only in terms of accuracy, but also in MHC specificity coverage. Simulation systems and mathematical modeling are also now beginning to reach a level where these methods will be able to answer more complex immunological questions.

Modelling the human immune system by combining bioinformatics and systems biology approaches

Over the past decade a number of bioinformatics tools have been developed that use genomic sequences as input to predict to which parts of a microbe the immune system will react, the so-called epitopes. Many predicted epitopes have later been verified experimentally, demonstrating the usefulness of such predictions. At the same time, simulation models have been developed that describe the dynamics of different immune cell populations and their interactions with microbes. These models have been used to explain experimental findings where timing is of importance, such as the time between administration of a vaccine and infection with the microbe that the vaccine is intended to protect against. In this paper, we outline a framework for integration of these two approaches. As an example, we develop a model in which HIV dynamics are correlated with genomics data. For the first time, the fitness of wild type and mutated virus are assessed by means of a sequence-dependent scoring matrix, derived from a BLOSUM matrix, that links protein sequences to growth rates of the virus in the mathematical model. A combined bioinformatics and systems biology approach can lead to a better understanding of immune system-related diseases where both timing and genomic information are of importance.

Sensitive quantitative predictions of peptide-MHC binding by a 'Query by Committee' artificial neural network approach

We have generated Artificial Neural Networks (ANN) capable of performing sensitive, quantitative predictions of peptide binding to the MHC class I molecule, HLA-A*0204. We have shown that such quantitative ANN are superior to conventional classification ANN, that have been trained to predict binding vs non-binding peptides. Furthermore, quantitative ANN allowed a straightforward application of a 'Query by Committee' (QBC) principle whereby particularly information-rich peptides could be identified and subsequently tested experimentally. Iterative training based on QBC-selected peptides considerably increased the sensitivity without compromising the efficiency of the prediction. This suggests a general, rational and unbiased approach to the development of high quality predictions of epitopes restricted to this and other HLA molecules. Due to their quantitative nature, such predictions will cover a wide range of MHC-binding affinities of immunological interest, and they can be readily integrated with predictions of other events involved in generating immunogenic epitopes. These predictions have the capacity to perform rapid proteome-wide searches for epitopes. Finally, it is an example of an iterative feedback loop whereby advanced, computational bioinformatics optimize experimental strategy, and vice versa.

Definition of supertypes for HLA molecules using clustering of specificity matrices

Major histocompatibility complex (MHC) proteins are encoded by extremely polymorphic genes and play a crucial role in immunity. However, not all genetically different MHC molecules are functionally different. Sette and Sidney (1999) have defined nine HLA class I supertypes and showed that with only nine main functional binding specificities it is possible to cover the binding properties of almost all known HLA class I molecules. Here we present a comprehensive study of the functional relationship between all HLA molecules with known specificities in a uniform and automated way. We have developed a novel method for clustering sequence motifs. We construct hidden Markov models for HLA class I molecules using a Gibbs sampling procedure and use the similarities among these to define clusters of specificities. These clusters are extensions of the previously suggested ones. We suggest splitting some of the alleles in the A1 supertype into a new A26 supertype, and some of the alleles in the B27 supertype into a new B39 supertype. Furthermore the B8 alleles may define their own supertype. We also use the published specificities for a number of HLA-DR types to define clusters with similar specificities. We report that the previously observed specificities of these class II molecules can be clustered into nine classes, which only partly correspond to the serological classification. We show that classification of HLA molecules may be done in a uniform and automated way. The definition of clusters allows for selection of representative HLA molecules that can cover the HLA specificity space better. This makes it possible to target most of the known HLA alleles with known specificities using only a few peptides, and may be used in construction of vaccines. Supplementary material is available at http://www.cbs.dtu.dk/researchgroups/immunology/supertypes.html.

Bioinformatic analysis of functional differences between the immunoproteasome and the constitutive proteasome

Intracellular proteins are degraded largely by proteasomes. In cells stimulated with gamma interferon, the active proteasome subunits are replaced by "immuno" subunits that form immunoproteasomes. Phylogenetic analysis of the immunosubunits has revealed that they evolve faster than their constitutive counterparts. This suggests that the immunoproteasome has evolved a function that differs from that of the constitutive proteasome. Accumulating experimental degradation data demonstrate, indeed, that the specificity of the immunoproteasome and the constitutive proteasome differs. However, it has not yet been quantified how different the specificity of two forms of the proteasome are. The main question, which still lacks direct evidence, is whether the immunoproteasome generates more MHC ligands. Here we use bioinformatics tools to quantify these differences and show that the immunoproteasome is a more specific enzyme than the constitutive proteasome. Additionally, we predict the degradation of pathogen proteomes and find that the immunoproteasome generates peptides that are better ligands for MHC binding than peptides generated by the constitutive proteasome. Thus, our analysis provides evidence that the immunoproteasome has co-evolved with the major histocompatibility complex to optimize antigen presentation in vertebrate cells.

Identifying cytotoxic T cell epitopes from genomic and proteomic information: "The human MHC project."

Complete genomes of many species including pathogenic microorganisms are rapidly becoming available and with them the encoded proteins, or proteomes. Proteomes are extremely diverse and constitute unique imprints of the originating organisms allowing positive identification and accurate discrimination, even at the peptide level. It is not surprising that peptides are key targets of the immune system. It follows that proteomes can be translated into immunogens once it is known how the immune system generates and handles peptides. Recent advances have identified many of the basic principles involved. The single most selective event is that of peptide binding to MHC, making it particularly important to establish accurate descriptions and predictions of peptide binding for the most common MHC variants. These predictions should be integrated with those of other steps involved in antigen processing, as these become available. The ability to translate the accumulating primary sequence databases in terms of immune recognition should enable scientists and clinicians to analyze any protein of interest for the presence of potentially immunogenic epitopes. The computational tools to scan entire proteomes should also be developed, as this would enable a rational approach to vaccine development and immunotherapy. Thus, candidate vaccine epitopes might be predicted from the various microbial genome projects, tumor vaccine candidates from mRNA expression profiling of tumors ("transcriptomes") and auto-antigens from the human genome.

Clustering patterns of cytotoxic T-lymphocyte epitopes in human immunodeficiency virus type 1 (HIV-1) proteins reveal imprints of immune evasion on HIV-1 global variation

The human cytotoxic T-lymphocyte (CTL) response to human immunodeficiency virus type 1 (HIV-1) has been intensely studied, and hundreds of CTL epitopes have been experimentally defined, published, and compiled in the HIV Molecular Immunology Database. Maps of CTL epitopes on HIV-1 protein sequences reveal that defined epitopes tend to cluster. Here we integrate the global sequence and immunology databases to systematically explore the relationship between HIV-1 amino acid sequences and CTL epitope distributions. CTL responses to five HIV-1 proteins, Gag p17, Gag p24, reverse transcriptase (RT), Env, and Nef, have been particularly well characterized in the literature to date. Through comparing CTL epitope distributions in these five proteins to global protein sequence alignments, we identified distinct characteristics of HIV amino acid sequences that correlate with CTL epitope localization. First, experimentally defined HIV CTL epitopes are concentrated in relatively conserved regions. Second, the highly variable regions that lack epitopes bear cumulative evidence of past immune escape that may make them relatively refractive to CTLs: a paucity of predicted proteasome processing sites and an enrichment for amino acids that do not serve as C-terminal anchor residues. Finally, CTL epitopes are more highly concentrated in alpha-helical regions of proteins. Based on amino acid sequence characteristics, in a blinded fashion, we predicted regions in HIV regulatory and accessory proteins that would be likely to contain CTL epitopes; these predictions were then validated by comparison to new sets of experimentally defined epitopes in HIV-1 Rev, Tat, Vif, and Vpr.

Prediction of human protein function from post-translational modifications and localization features

We have developed an entirely sequence-based method that identifies and integrates relevant features that can be used to assign proteins of unknown function to functional classes, and enzyme categories for enzymes. We show that strategies for the elucidation of protein function may benefit from a number of functional attributes that are more directly related to the linear sequence of amino acids, and hence easier to predict, than protein structure. These attributes include features associated with post-translational modifications and protein sorting, but also much simpler aspects such as the length, isoelectric point and composition of the polypeptide chain.

Evolutionary and immunological implications of contemporary HIV-1 variation

Evolutionary modelling studies indicate less than a century has passed since the most recent common ancestor of the HIV-1 pandemic strains and, in that time frame, an extraordinarily diverse viral population has developed. HIV-1 employs a multitude of schemes to generate variants: accumulation of base substitutions, insertions and deletions, addition and loss of glycosylation sites in the envelope protein, and recombination. A comparison between HIV and influenza virus illustrates the extraordinary scale of HIV variation, and underscores the importance of exploring innovative HIV vaccine strategies. Deeper understanding of the implications of variation for both antibody and T-cell responses may help in the effort to rationally design vaccines that stimulate broad cross-reactivity. The impact of HIV-1 variation on host immune response is reviewed in this context.

Identification of wheat varieties using matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry and an artificial neural network

A novel tool for variety identification of wheat (Triticum aestivum L.) has been developed: an artificial neural network (ANN) is used to classify the gliadin fraction analysed by matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOFMS). The robustness of this novel method with respect to various experimental parameters has been tested. The results can be summarised: (i) With this approach 97% of the wheat varieties can be classified correctly with a corresponding correlation coefficient of 1.0, (ii) The method is fast since the time of extracting gliadins from flour can be reduced to 20 min without significant decrease in overall performance, (iii) The storage of flour or extracts under standard conditions does not influence the classification ability (i. e. the generalisation ability) of the method, and (iv) The classification obtained is not influenced by the identity of the operator making the analysis. This study demonstrates that a combination of an ANN and MALDI-TOFMS analysis of the gliadin fraction provides a fast and reliable tool for the variety identification of wheat. Copyright 1999 John Wiley & Sons, Ltd.