Epitope Prediction Research Articles

Production of monoclonal antibody of heat-labile toxin A subunit to identify enterotoxigenic Escherichia coli by epitope mapping using synthetic peptides.

Enterotoxigenic Escherichia coli (ETEC) is a major cause of diarrhea through two enterotoxins, a heat-labile toxin and a heat-stable toxin. These toxins alter the cellular signaling pathways, ultimately triggering an increase in chloride secretion and watery diarrhea. For the development of an ETEC vaccine, we attempted to construct a peptide-specific monoclonal antibody library against heat-labile enterotoxin A subunit (LT-A) by epitope mapping using synthetic peptides. Sera produced by five mice immunized with recombinant LT-A protein were examined for specific recognition with synthetic 15-mer and 34-mer peptides of LT-A proteins using enzyme-linked immunosorbent assay. The analysis revealed that the synthetic peptides number 8, 16, 24, 33, 36, 38, and 39 reacted with an anti-LT-A polyclonal antibody. For the possible prediction of LT-A epitopes, each full-length protein sequence was subjected to BCPreds analysis and three-dimensional protein structure analysis. The data showed that three peptides (synthetic peptide numbers: 33, 36, and 38-39) have identical antigenic specificities with LT-A protein, suggesting the usefulness of these linear peptide epitopes. Based on these peptides, we produced monoclonal antibodies to improve the specificity of LT-A detection. Monoclonal antibodies produced from two peptides (numbers 33 and 36) showed affinity for an LT-A recombinant antigen. Moreover, peptide epitope prediction analysis showed that the sites of the three peptides were identical to those exhibiting actual antigenicity. Also, it was confirmed that the amino acid sequence that actually showed antigenicity was included in the peptide predicted only by ETEC-LT-A-33. Also, the specificity of the antibody for ETEC-LT-A-33 was validated using bacterial cells, and the neutralizing effect of the antibody was determined by assessing cytokine release in infected HCT-8 cells. The monoclonal antibodies produced in this study are useful toolsfor vaccine production against ETEC and can be used to identify peptide antigencandidates.

Idiotope-Driven T-Cell/B-Cell Collaboration-Based T-Cell Epitope Prediction Using B-Cell Receptor Repertoire Sequences in Infectious Diseases.

T-cell recognition of antigen epitopes is a crucial step for the induction of adaptive immune responses, and the identification of such T-cell epitopes is, therefore, important for understanding diverse immune responses and controlling T-cell immunity. A number of bioinformatic tools exist that predict T-cell epitopes; however, many of these methods highly rely on evaluating conventional peptide presentation by major histocompatibility complex (MHC) molecules, but they ignore epitope sequences recognized by T-cell receptor (TCR). Immunogenic determinant idiotopes are present on the variable regions of immunoglobulin molecules expressed on and secreted by B-cells. In idiotope-driven T-cell/B-cell collaboration, B-cells present the idiotopes on MHC molecules for recognition by idiotope-specific T-cells. According to the idiotype network theory formulated by Niels Jerne, such idiotopes found on anti-idiotypic antibodies exhibit molecular mimicry of antigens. Here, by combining these concepts and defining the patterns of TCR-recognized epitope motifs (TREMs), we developed a T-cell epitope prediction method that identifies T-cell epitopes derived from antigen proteins by analyzing B-cell receptor (BCR) sequences. This method allowed us to identify T-cell epitopes that contain the same TREM patterns between BCR and viral antigen sequences in two different infectious diseases caused by dengue virus and SARS-CoV-2 infection. The identified epitopes were among the T-cell epitopes detected in previous studies, and T-cell stimulatory immunogenicity was confirmed. Thus, our data support this method as a powerful tool for the discovery of T-cell epitopes from BCR sequences.

Characterization of Duffy Binding Protein II-specific CD4+T cell responses in Plasmodium vivax patients

Plasmodium vivax Duffy Binding Protein region II (PvDBPII) is a leading vaccine candidate against blood-stage vivax malaria. Anti-PvDBPII antibodies potentially block parasite invasion by inhibition of erythrocyte binding. However, knowledge of PvDBPII-specific T cell responses is limited. Here, to assess the responses of PvDBPII-specific CD4+T cells in natural P. vivax infection, three cross-sectional studies were conducted in recovered subjects. In silico analysis was used for potential T cell epitope prediction and selection. PBMCs from P. vivax subjects were stimulated with selected peptides and examined for cytokine production by ELISPOT or intracellular cytokine staining. Six dominant T cell epitopes were identified. Peptide-driven T cell responses showed effector memory CD4+T cell phenotype, secreting both IFN-γ and TNF-α cytokines. Single amino acid substitutions in three T cell epitopes altered levels of IFN-γ memory T cell responses. Seropositivity of anti-PvDBPII antibodies were detected during acute malaria (62%) and persisted up to 12 months (11%) following P. vivax infection. Further correlation analysis showed four out of eighteen subjects had positive antibody and CD4+T cell responses to PvDBPII. Altogether, PvDBPII-specific CD4+T cells were developed in natural P. vivax infections. Data on their antigenicity could facilitate development of an efficacious vivax malaria vaccine.

Computational identification of a multi-peptide vaccine candidate in E2 glycoprotein against diverse Hepatitis C virus genotypes

Hepatitis C Virus (HCV) is estimated to affect nearly 180 million people worldwide, culminating in ∼0.7 million yearly casualties. However, a safe vaccine against HCV is not yet available. This study endeavored to identify a multi-genotypic, multi-epitopic, safe, and globally competent HCV vaccine candidate. We employed a consensus epitope prediction strategy to identify multi-epitopic peptides in all known envelope glycoprotein (E2) sequences, belonging to diverse HCV genotypes. The obtained peptides were screened for toxicity, allergenicity, autoimmunity and antigenicity, resulting in two favorable peptides viz., P2 (VYCFTPSPVVVG) and P3 (YRLWHYPCTV). Evolutionary conservation analysis indicated that P2 and P3 are highly conserved, supporting their use as part of a designed multi-genotypic vaccine. Population coverage analysis revealed that P2 and P3 are likely to be presented by >89% Human Leukocyte Antigen (HLA) molecules from six geographical regions. Indeed, molecular docking predicted the physical binding of P2 and P3 to various representative HLAs. We designed a vaccine construct using these peptides and assessed its binding to toll-like receptor 4 (TLR-4) by molecular docking and simulation. Subsequent analysis by energy-based and machine learning tools predicted high binding affinity and pinpointed the key binding residues (i.e. hotspots) in P2 and P3. Also, a favorable immunogenic profile of the construct was predicted by immune simulations. We encourage the scientific community to validate our vaccine construct in vitro and in vivo. Communicated by Ramaswamy H. Sarma

A computational quest for identifying potential vaccine candidates against Moraxella lacunata: a multi-pronged approach

Moraxella lacunata is an emerging gram-negative bacterium that is responsible for multiple nosocomial infections. The bacterium is evolving resistance to several antibiotics, and currently, no effective licensed vaccines are available, which warrants the search for new therapeutics. A multi-epitope-based vaccine has been designed for M. lacunata. The complete proteome of M. lacunata contains 10,110 core proteins. Subcellular localization analysis revealed the presence of five proteins in the extracellular matrix, while 19 proteins were predicted to be located in the outer membrane, and 21 proteins were predicted to be located in the periplasmic region. Only two proteins, the type VI secretion system tube protein (Hcp) and the transporter substrate-binding domain-containing protein, were selected for epitope prediction as they fulfilled all the criteria for being potential vaccine candidates. Shortlisted epitopes from the selected proteins were fused together using “GPGPG” linkers to overcome the limitations of single-epitope vaccines. Next, the cholera toxin-B adjuvant was attached to the peptide epitope using an EAAAK linker. Docking analysis was performed to examine the interaction between the vaccine and immune cell receptors, revealing robust intermolecular interactions and a stable binding conformation. Molecular dynamics simulation findings revealed no drastic changes in the binding conformation of complexes during the simulation period. The net binding free energy of vaccine-receptor complexes was estimated using the molecular mechanics energies combined with the Poisson–Boltzmann and surface area continuum solvation (MM-PBSA) method. The reported values were −586.38 kcal/mol, −283.74 kcal/mol, and −296.88 kcal/mol for the TLR-4-vaccine complex, MHC-I-vaccine complex, and MHC-II-vaccine complex, respectively. Furthermore, the molecular mechanics energies combined with the generalized Born and surface area continuum solvation (MM-GBSA) analysis predicted binding free energies of −596.69 kcal/mol, −287.39 kcal/mol, and −298.28 kcal/mol for the TLR-4-vaccine complex, MHC-I-vaccine complex, and MHC-II-vaccine complex, respectively. The theoretical vaccine design proposed in the study could potentially serve as a powerful therapeutic against targeted pathogens, subject to validation through experimental studies. Communicated by Ramaswamy H. Sarma

ITCep: a deep learning framework for identification of T cell epitopes by harnessing fusion features.

Neoantigens recognized by cytotoxic T cells are effective targets for tumor-specific immune responses for personalized cancer immunotherapy. Quite a few neoantigen identification pipelines and computational strategies have been developed to improve the accuracy of the peptide selection process. However, these methods mainly consider the neoantigen end and ignore the interaction between peptide-TCR and the preference of each residue in TCRs, resulting in the filtered peptides often fail to truly elicit an immune response. Here, we propose a novel encoding approach for peptide-TCR representation. Subsequently, a deep learning framework, namely iTCep, was developed to predict the interactions between peptides and TCRs using fusion features derived from a feature-level fusion strategy. The iTCep achieved high predictive performance with AUC up to 0.96 on the testing dataset and above 0.86 on independent datasets, presenting better prediction performance compared with other predictors. Our results provided strong evidence that model iTCep can be a reliable and robust method for predicting TCR binding specificities of given antigen peptides. One can access the iTCep through a user-friendly web server at http://biostatistics.online/iTCep/, which supports prediction modes of peptide-TCR pairs and peptide-only. A stand-alone software program for T cell epitope prediction is also available for convenient installing at https://github.com/kbvstmd/iTCep/.

In silico design and validation of a novel multi-epitope vaccine candidate against structural proteins of Chikungunya virus using comprehensive immunoinformatics analyses.

Chikungunya virus (CHIKV) is an emerging viral infectious agent with the potential of causing pandemic. There is neither a protective vaccine nor an approved drug against the virus. The aim of this study was design of a novel multi-epitope vaccine (MEV) candidate against the CHIKV structural proteins using comprehensive immunoinformatics and immune simulation analyses. In this study, using comprehensive immunoinformatics approaches, we developed a novel MEV candidate using the CHIKV structural proteins (E1, E2, 6 K, and E3). The polyprotein sequence was obtained from the UniProt Knowledgebase and saved in FASTA format. The helper and cytotoxic T lymphocytes (HTLs and CTLs respectively) and B cell epitopes were predicted. The toll-like receptor 4 (TLR4) agonist RS09 and PADRE epitope were employed as promising immunostimulatory adjuvant proteins. All vaccine components were fused using proper linkers. The MEV construct was checked in terms of antigenicity, allergenicity, immunogenicity, and physicochemical features. The docking of the MEV construct and the TLR4 and molecular dynamics (MD) simulation were also performed to assess the binding stability. The designed construct was non-allergen and was immunogen which efficiently stimulated immune responses using the proper synthetic adjuvant. The MEV candidate exhibited acceptable physicochemical features. Immune provocation included prediction of HTL, B cell, and CTL epitopes. The docking and MD simulation confirmed the stability of the docked TLR4-MEV complex. The high-level protein expression in the Escherichia coli (E. coli) host was observed through in silico cloning. The in vitro, in vivo, and clinical trial investigations are required to verify the findings of the current study.

Pressured cytotoxic T cell epitope strength among SARS-CoV-2 variants correlates with COVID-19 severity.

Heterogeneity in susceptibility among individuals to COVID-19 has been evident through the pandemic worldwide. Cytotoxic T lymphocyte (CTL) responses generated against pathogens in certain individuals are known to impose selection pressure on the pathogen, thus driving emergence of new variants. In this study, we probe the role played by host genetic heterogeneity in terms of HLA-genotypes in determining differential COVID-19 severity in patients. We use bioinformatic tools for CTL epitope prediction to identify epitopes under immune pressure. Using HLA-genotype data of COVID-19 patients from a local cohort, we observe that the recognition of pressured epitopes from the parent strain Wuhan-Hu-1 correlates with COVID-19 severity. We also identify and rank list HLA-alleles and epitopes that offer protectivity against severe disease in infected individuals. Finally, we shortlist a set of 6 pressured and protective epitopes that represent regions in the viral proteome that are under high immune pressure across SARS-CoV-2 variants. Identification of such epitopes, defined by the distribution of HLA-genotypes among members of a population, could potentially aid in prediction of indigenous variants of SARS-CoV-2 and other pathogens.

The antigen presentation of MCMV specific Qa-1-restricted CD8 +T cells

Abstract Murine cytomegalovirus (MCMV) is a well-established model system for human CMV (HCMV), a ubiquitous herpesvirus that affects over 70% of people worldwide. Although the classical CD8 +T cell response to MCMV has been well characterized, little is known about the non-classical CD8 +T cell response. Our laboratory recently demonstrated that Qa-1-restricted CD8 +T cells are sufficient to protect mice from MCMV induced lethality. Our recent data shows that this protective T cell population has limited αβ TCR diversity. To identify the MCMV peptide(s) presented by Qa-1, we utilized bioinformatic tools that predict MHC binding, including NetMHC 4.0 and T Cell Epitope Prediction tools available on the Immune Epitope Database and Analysis Resource. Using these tools, we queried the entire MCMV coding sequence to identify potential nonamers predicted to bind to Qa-1. Furthermore, these candidates were also queried to identify protease cleavage sites to ensure that they are predicted to exist in vivo. Approximately 40 potential peptides were selected and tested in an in vitroscreen using Qa-1 expressing cell lines. Our results suggest that several candidate peptides bind to Qa-1 as strongly as well-defined canonical binding partners. Taken together, these data suggest that an MCMV-responsive T cell population with limited αβ TCR diversity recognizes MCMV-derived peptides presented by Qa-1. This population is likely to be of importance for viral protection when classical CD8 +T cell immunity is compromised. This work is supported by the NIH Research Grant AI046709. Shanelle Reilly is supported by research supplement R01 3AI046709-18S1 to promote diversity.

Cloning and Molecular Characterization of the Recombinant CVB4E2 Immunogenic Viral Protein (rVP1), as a Potential Subunit Protein for Vaccine and Immunodiagnostic Reagent Candidate.

The aim of the present study was, first, to clone the VP1 gene of the human coxsackievirus B4 strain E2 (CVB4E2) in the prokaryotic pUC19 plasmid expression vector then to compare it with the structural capsid proteins of the same strain using bioinformatic tools. PCR colony amplification followed through a restriction digestion analysis and sequencing process which affirmed the success of the cloning process. SDS-PAGE and Western Blotting were used to characterize the purified recombinant viral protein expressed in bacteria cells. The BLASTN tool revealed that the nucleotide sequence of the recombinant VP1 (rVP1) expressed by pUC19 highly matched the target nucleotide sequence of the diabetogenic CVB4E2 strain. Secondary structure and three-dimension structure prediction suggested that rVP1, such as wild-type VP1, is chiefly composed of random coils and a high percentage of exposed amino acids. Linear B-cell epitope prediction showed that several antigenic epitopes are likely present in rVP1 and CVB4E2 VP1 capsid protein. Additionally, phosphorylation site prediction revealed that both proteins may affect the signal transduction of host cells and can be involved in virus virulence. The present work highlights the usefulness of cloning and bioinformatics characterizations for gene investigation. Furthermore, the collected data are helpful for future experimental research related to the development of immunodiagnostic reagents and subunit vaccines based on the expression of immunogenic viral capsid proteins.

Potential of B-Cell Epitopes Protein Ag85 Complex Mycobacterium Tuberculosis as Serodiagnostic Antigen of Tuberculosis by in Silico Study

Background: The high case of tuberculosis which isn't followed by good detection becomes an urgency for the diagnostic developments. One of them with immunodiagnostic principle uses B-cell Ag85 complex epitope. The design of the diagnostic epitope was performed by mapping the B cell epitope used in silico studies.Objective: The purpose of this research is to analyze antigenicity, physicochemical which affect immunogenicity, and homology of B-cell Ag85 complex epitope with the strain which circulates in Indonesia. Methods: The samples used were taken from the NCBI protein bank with access numbers P9WQP3 for Ag85A, P9WQP1 for Ag85B, and P9WQN9 for Ag85C. The sequences were analyzed using IEDB (Bepipred) software as the epitope prediction, VaxiJen as antigenicity prediction, ProtParam as physicochemical properties prediction, and BLASTP NCBI as sequence alignment.Results: Twenty seven epitopes were antigenic with 0.4297 to 2.6007 scores and the molecular weight was from 619.59 Da to 3145.36 Da. This research also obtained eleven stable and hydrophilic epitopes. The alignment of 11 candidate epitopes with the strain which circulates in Indonesia, had a similarity percentage of 85.71%-100% and 3 epitopes had a more significant score.Conclusion: Three epitopes of Ag85 complex; Ag85A (212-235), Ag85B (209-237), and Ag85C (283-310), were universal antigens and can be developed into diagnostic antigens in Indonesia.

Biomaterial-Based In Situ Cancer Vaccines.

Cancer immunotherapies have reshaped the paradigm for cancer treatment over the past decade. Among them, therapeutic cancer vaccines that aim to modulate antigen-presenting cells and subsequent T cell priming processes are among the first FDA-approved cancer immunotherapies. However, despite showing benign safety profiles and the capability to generate antigen-specific humoral and cellular responses, cancer vaccines have been limited by the modest therapeutic efficacy, especially for immunologically cold solid tumors. One key challenge lies in the identification of tumor-specific antigens, which involves a costly and lengthy process of tumor cell isolation, DNA/RNA extraction, sequencing, mutation analysis, epitope prediction, peptide synthesis, and antigen screening. To address these issues, in situ cancer vaccines have been actively pursued to generate endogenous antigens directly from tumors and utilize the generated tumor antigens to elicit potent cytotoxic T lymphocyte (CTL) response. Biomaterials-based in situ cancer vaccines, in particular, have achieved significant progress by taking advantage of biomaterials that can synergize antigens and adjuvants, troubleshoot delivery issues, home, and manipulate immune cells in situ. This review will provide an overview of biomaterials-based in situ cancer vaccines, either living or artificial materials, under development or in the clinic, and discuss the design criteria for in situ cancer vaccines.

Potential contribution of Helicobacter pylori proteins in the pathogenesis of type 1 gastric neuroendocrine tumor and urticaria. In silico approach.

Helicobacter pylori has been linked to several diseases such as chronic urticaria, gastritis, and type 1 gastric neuroendocrine tumors (type 1 gNET). Although these diseases seem to have different mechanisms, their relationship with H. pylori suggests a common inflammatory pathway. To identify potential cross-reactive antigens between H. pylori and humans involved in chronic urticaria and type 1 gNET. Alignment was carried out among human proteins associated with urticaria (9 proteins), type 1 gNET (32 proteins), and H. pylori proteome. We performed pairwise alignment among the human and H. pylori antigens with PSI-BLAST. Modeling based on homology was done with the Swiss model server and epitope prediction with the Ellipro server. Epitopes were located on a 3D model using PYMOL software. The highest conserved sequence was found between the human HSP 60 antigen and the H. pylori chaperonin GroEL with an identity of 54% and a cover of 92%, followed by the alpha and gamma enolases and two H. pylori phosphopyruvate hydratase, both with an identity and cover of 48% and 96%, respectively. The H/K ATPase (Chain A) showed high identity with two H. pylori proteins (35.21% with both P-type ATPase), but with low cover (only 6%). We observed eight linear and three discontinuous epitopes for human HSP 60 and three lineal and one discontinuous epitope for both alpha-enolase and gamma enolase, high conserved with H. pylori sequences. Some type 1 gNET antigens shared potential cross-reactive epitopes with H. pylori proteins, suggesting that molecular mimicry could be a mechanism that explains the relationship between the infection and this disease. Studies evaluating the functional impact of this relationship are needed.

Sequence characteristics of Rhipicephalus microplus Enolase gene and prediction of structure and antigenic epitopes of its encoding protein

To analyze the sequence characteristics of Rhipicephalus microplus Enolase gene, and to predict the secondary and tertiary structure and antigenic epitopes of the Enolase protein. Sixty-two engorged female R. microplus were sampled from a yellow cattle breeding farm in Zhijiang County, Huaihua City, Hunan Province in June 25, 2022. Genomic DNA was isolated from R. microplus, and the Enolase gene was amplified using PCR assay, followed by cloning, sequencing and expression of the amplification product. The sequence characteristics of the Enolase gene were analyzed using the software Clustal X, and the gene sequence was translated into amino acid sequences. The secondary and tertiary structures of the Enolase protein were deduced using the software PRABI, and the physicochemical properties of the Enolase protein were analyzed using the software PRABI. In addition, the B- and T-cell epitopes of the Enolase protein were predicted using the software ABCpred Prediction, Scratch, IEDB and NetCTL. The R. microplus Enolase gene sequence was 1 323 bp in size, and the contents of A, T, G and C bases were 24.5%, 22.5%, 27.0% and 26.0%,with 47.0% of A + T content and 53.0% of G + C content. The R. microplus Enolase gene encoded 434 amino acids, and the Enolase protein had a molecular weight of 47.12 kDa. The secondary structure of the Enolase protein contained 186 α-helixes (42.86%), 32 β-turns (7.37%), 144 random coils (33.18%) and 72 extended strands (16.59%). The Enolase protein was most probably present in cytoplasm (76.7%), followed by in mitochondrion (39.1%) and nucleus (21.7%), and the Enolase protein had no signal peptide or transmembrane domain. In addition, the Enolase protein had 14 B-cell dominant epitopes and 8 T-cell dominant epitopes. The R. microplus Enolase gene sequence exhibits a GC preference, and its encoding Enolase protein is an acidic and hydrophilic protein, with α-helixes and random coils as its primary structure, and presenting B- and T-cell dominant epitopes, which is a potential target for development of vaccines against R. microplus.

Genetic Analysis and Epitope Prediction of SARS-CoV-2 Genome in Bahia, Brazil: An In Silico Analysis of First and Second Wave Genomics Diversity

COVID-19 is an infectious disease caused by SARS-CoV-2. This virus presents high levels of mutation and transmissibility, which contributed to the emergence of the pandemic. Our study aimed to analyze, in silico, the genomic diversity of SARS-CoV-2 strains in Bahia State by comparing patterns in variability of strains circulating in Brazil with the first isolated strain NC_045512 (reference sequence). Genomes were collected using GISAID, and subsequently aligned and compared using structural and functional genomic annotation. A total of 744 genomes were selected, and 20,773 mutations were found, most of which were of the SNP type. Most of the samples presented low mutational impact, and of the samples, the P.1 (360) lineage possessed the highest prevalence. The most prevalent epitopes were associated with the ORF1ab protein, and in addition to P.1, twenty-one other lineages were also detected during the study period, notably B.1.1.33 (78). The phylogenetic tree revealed that SARS-CoV-2 variants isolated from Bahia were clustered closely together. It is expected that the data collected will help provide a better epidemiological understanding of the COVID-19 pandemic (especially in Bahia), as well as helping to develop more effective vaccines that allow less immunogenic escape.

A comparative 'omics' approach for prediction of candidate Strongyloides stercoralis diagnostic coproantigens.

Human infection with the intestinal nematode Strongyloides stercoralis is persistent unless effectively treated, and potentially fatal in immunosuppressed individuals. Epidemiological data are lacking, partially due to inadequate diagnosis. A rapid antigen detection test is a priority for population surveillance, validating cure after treatment, and for screening prior to immunosuppression. We used a targeted analysis of open access 'omics' data sets and used online predictors to identify S. stercoralis proteins that are predicted to be present in infected stool, Strongyloides-specific, and antigenic. Transcriptomic data from gut and non-gut dwelling life cycle stages of S. stercoralis revealed 328 proteins that are differentially expressed. Strongyloides ratti proteomic data for excreted and secreted (E/S) proteins were matched to S. stercoralis, giving 1,057 orthologues. Five parasitism-associated protein families (SCP/TAPS, prolyl oligopeptidase, transthyretin-like, aspartic peptidase, acetylcholinesterase) were compared phylogenetically between S. stercoralis and outgroups, and proteins with least homology to the outgroups were selected. Proteins that overlapped between the transcriptomic and proteomic datasets were analysed by multiple sequence alignment, epitope prediction and 3D structure modelling to reveal S. stercoralis candidate peptide/protein coproantigens. We describe 22 candidates from seven genes, across all five protein families for further investigation as potential S. stercoralis diagnostic coproantigens, identified using open access data and freely-available protein analysis tools. This powerful approach can be applied to many parasitic infections with 'omic' data to accelerate development of specific diagnostic assays for laboratory or point-of-care field application.

Epitope1D: accurate taxonomy-aware B-cell linear epitope prediction.

The ability to identify B-cell epitopes is an essential step in vaccine design, immunodiagnostic tests and antibody production. Several computational approaches have been proposed to identify, from an antigen protein or peptide sequence, which residues are more likely to be part of an epitope, but have limited performance on relatively homogeneous data sets and lack interpretability, limiting biological insights that could otherwise be obtained. To address these limitations, we have developed epitope1D, an explainable machine learning method capable of accurately identifying linear B-cell epitopes, leveraging two new descriptors: a graph-based signature representation of protein sequences, based on our well-established Cutoff Scanning Matrix algorithm and Organism Ontology information. Our model achieved Areas Under the ROC curve of up to 0.935 on cross-validation and blind tests, demonstrating robust performance. A comprehensive comparison to alternative methods using distinct benchmark data sets was also employed, with our model outperforming state-of-the-art tools. epitope1D represents not only a significant advance in predictive performance, but also allows biologically meaningful features to be combined and used for model interpretation. epitope1D has been made available as a user-friendly web server interface and application programming interface at https://biosig.lab.uq.edu.au/epitope1d/.

An immunoinformatic approach to assessing the immunogenic capacity of alpha-neurotoxins in elapid snake venoms.

Introduction: Most elapid snakes produce venoms that contain alpha-neurotoxins (α-NTXs), which are proteins that cause post-synaptic blockade and paralysis in snakebite envenoming. However, existing elapid antivenoms are known for their low potency in neutralizing the neurotoxic activity of α-NTXs, while the immunological basis has not been elucidated. Methods: In this study, a structure-based major histocompatibility complex II (MHCII) epitope predictor of horse (Equus caballus), complemented with DM-editing determinant screening algorithm was adopted to assess the immunogenicity of α-NTXs in the venoms of major Asiatic elapids (Naja kaouthia, Ophiophagus hannah, Laticauda colubrina, Hydrophis schistosus, Hydrophis curtus). Results: The scoring metric M2R, representing the relative immunogenic performance of respective α-NTXs, showed all α-NTXs have an overall low M2R of <0.3, and most of the predicted binders feature non-optimal P1 anchor residues. The M2R scores correlate strongly (R2 = 0.82) with the potency scores (p-score) generated based on the relative abundances of α-NTXs and the neutralization potency of commercial antivenoms. Discussion: The immunoinformatic analysis indicates that the inferior antigenicity of α-NTXs is not only due to their small molecular size but also the subpar immunogenicity affected by their amino acid composition. Structural modification with conjugation and synthetic epitope as immunogen may potentially enhance the immunogenicity for improved antivenom potency against α-NTXs of elapid snakes.

Abstract 5376: Protein structure-based modeling to improve MHC class I epitope predictions

Abstract The ability to accurately identify peptide ligands for a given major histocompatibility complex class I (MHC-I) molecule has immense value for targeted anticancer therapeutics. However, the highly polymorphic nature of the MHC-I protein makes universal prediction of peptide ligands challenging due to lack of experimental data describing most MHC-I variants. To address this challenge, we have developed a deep convolutional neural network, HLA-Inception, capable of predicting MHC-I peptide binding motifs using electrostatic properties of the MHC-I binding pocket. By approaching this immunological issue using molecular biophysics, we measure the impact of sidechain arrangement and topology on peptide binding, feature not captured by sequence-based MHC-I prediction methods. Through a combination of molecular modeling and simulation, 5821 MHC-I alleles were modeled, providing extensive coverage across human populations. Predicted peptide binding motifs fell into distinct clusters, each defined with different degrees of submotif heterogeneity. Peptide binding scores generated by HLA-Inception are strongly correlated with quantitative MHC-I binding data, indicating predicted peptides can be ranked, both within and between alleles. HLA-inception also showed high precision when predicting naturally presented peptides and can be used for rapid proteome-scale MHC-I peptide binding predictions. Finally, we show that the binding pocket diversity measured by HLA inception predicts response to checkpoint blockade. Citation Format: Eric A. Wilson, John Kevin Cava, Diego Chowell, Abhishek Singharoy, Karen S. Anderson. Protein structure-based modeling to improve MHC class I epitope predictions. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5376.

Abstract 5364: Language modeling of peptide-HLA interactions achieves state-of-the-art performance on prediction of peptide presentation by HLA Class II

Abstract Precise and sensitive prediction of neoantigen presentation to the immune system via human leukocyte antigen (HLA) class II molecules remains a challenge despite the early success of neural networks applied to HLA class I. However, it is necessary to address this modeling challenge because presentation of a neoantigen epitope by both classes of HLA molecules may be valuable to induce a sustained immune response with therapeutic cancer vaccines. Previously we have developed a machine-learning based platform, EDGETM, that provides a state-of-the-art model to predict presentation of peptides by HLA Class I. Here we propose a new addition to our EDGETM platform: a model that leverages structural information of putative epitopes and HLA class II alleles from their in-situ context to predict presentation of peptides by HLA class II. Our model achieves this by leveraging the Evolutionary Scale Model pre-trained protein language model (LM), which has been demonstrated to embed protein sequences with rich structural information. The input to the model is a linear peptide consisting of an epitope and its flanking amino acids, concatenated with structurally relevant amino acids from each HLA allele. This allows our model to treat the modeling problem entirely as a natural language processing task, which minimizes imputation of covariates found in prior approaches when performing inference in the context of vaccine design, while maximizing the richness of the LM embeddings on longer linear peptides. Crucially, this also allows our model to generalize to any allele that has a known sequence. Additionally, DR-, DP-, and DQ-specific immunoaffinity purified mass spectrometry multi-allelic (MA) presentation data were generated per tumor or cell line sample, spanning 89 alleles in aggregate. We demonstrate that incrementally decreasing HLA class II allele MA resolution during training results in substantially improved predictions for situations where MA presentation data has completely ambiguous epitope presentation across DR/DP/DQ alleles. Our model achieves an Average Precision (AP) of 0.92 and ROC-AUC of 0.98 on the same benchmark validation data as the current state-of-the-art model BERTMHC, which achieved an AP of 0.81 and ROC-AUC of 0.95. These are the best AP and ROC-AUC for an HLA Class II presentation model on this benchmark dataset to the best of our knowledge. Our model is a significant advancement in HLA class II epitope prediction that allows our EDGETM platform to bring neoantigen vaccine design optimized for both class I and class II presentation within reach. Citation Format: Daniel Sprague, Joshua Klein, Italo Faria do Valle, Olivia Petrillo, Melissa Rotunno, Matthew Davis, Monica Lane, Karin Jooss, Ankur Dhanik. Language modeling of peptide-HLA interactions achieves state-of-the-art performance on prediction of peptide presentation by HLA Class II. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5364.