Abstract:ABSTRACT Throughout life, humans experience repeated exposure to viral antigens through infection and vaccination, resulting in the generation of diverse, antigen-specific antibody repertoires. A paramount feature of antibodies that enables their critical contributions in counteracting recurrent and novel pathogens, and consequently fostering their utility as valuable targets for therapeutic and vaccine development, is the exquisite specificity displayed against their target antigens. Yet, there is still limited understanding of the determinants of antibody-antigen specificity, particularly as a function of antibody sequence. In recent years, experimental characterization of antibody repertoires has led to novel insights into fundamental properties of antibody sequences but has been largely decoupled from at-scale antigen specificity analysis. Here, using the LIBRA-seq technology, we generated a large data set mapping antibody sequence to antigen specificity for thousands of B cells, by screening the repertoires of a set of healthy individuals against 20 viral antigens representing diverse pathogens of biomedical significance. Analysis uncovered virus-specific patterns in variable gene usage, gene pairing, somatic hypermutation, as well as the presence of convergent antiviral signatures across multiple individuals, including the presence of public antibody clonotypes. Notably, our results showed that, for B-cell receptors originating from different individuals but leveraging an identical combination of heavy and light chain variable genes, there is a specific CDRH3 identity threshold above which B cells appear to exclusively share the same antigen specificity. This finding provides a quantifiable measure of the relationship between antibody sequence and antigen specificity and further defines experimentally grounded criteria for defining public antibody clonality. IMPORTANCE The B-cell compartment of the humoral immune system plays a critical role in the generation of antibodies upon new and repeated pathogen exposure. This study provides an unprecedented level of detail on the molecular characteristics of antibody repertoires that are specific to each of the different target pathogens studied here and provides empirical evidence in support of a 70% CDRH3 amino acid identity threshold in pairs of B cells encoded by identical IGHV:IGL(K)V genes, as a means of defining public clonality and therefore predicting B-cell antigen specificity in different individuals. This is of exceptional importance when leveraging public clonality as a method to annotate B-cell receptor data otherwise lacking antigen specificity information. Understanding the fundamental rules of antibody-antigen interactions can lead to transformative new approaches for the development of antibody therapeutics and vaccines against current and emerging viruses.

Use of Antibody Structural Information in Disease Prediction Models Reveals Antigen Specific B Cell Receptor Sequences in Bulk Repertoire Data

Public Baseline and shared response structures support the theory of antibody repertoire functional commonality

Predicting Antigen Specificity of Single T Cells Based on TCR CDR 3 Regions

Computational detection of antigen specific B cell receptors following immunization

Structural diversity of B-cell receptor repertoires along the B-cell differentiation axis in humans and mice

Benchmarking and integrating human B-cell receptor genomic and antibody proteomic profiling

Large-scale data mining of four billion human antibody variable regions reveals convergence between therapeutic and natural antibodies that constrains search space for biologics drug discovery

Computational Prediction of Conformational B-Cell Epitopes from Antigen Primary Structures by Ensemble Learning

Restricted epitope specificity determined by variable region germline segment pairing in rodent antibody repertoires.

Using the antibody-antigen binding interface to train image-based deep neural networks for antibody-epitope classification

Biochemical patterns of antibody polyreactivity revealed through a bioinformatics-based analysis of CDR loops

Learned features of antibody-antigen binding affinity

Antibody sequence determinants of viral antigen specificity

Prediction of conformational B-cell epitopes from 3D structures by random forests with a distance-based feature

Antibodies: Computer-Aided Prediction of Structure and Design of Function

Disease diagnostics using machine learning of immune receptors

A method for predicting linear and conformational B-cell epitopes in an antigen from its primary sequence

A Comparison of Antibody-Antigen Complex Sequence-to-Structure Prediction Methods and their Systematic Biases

Using a Machine-Learning Approach to Predict Discontinuous Antibody-Specific B-Cell Epitopes

Functional and epitope specific monoclonal antibody discovery directly from immune sera using cryoEM

Improved deep learning prediction of antigen-antibody interactions