The discovery of MHC restriction, the resolution of peptide-MHC structure, and the identification of the T cell receptor gene quickly centered epitope-level specificity as a core concern of T cell biology. Epitope mapping culminated in functional peptide-based stimulation assays and tetramer staining, which together provide exquisite resolution of individual responses. Despite these tools, quantitative assessment of epitope-resolved responses in complex human infections and tumors has rarely been applied at scale. As a result, no epitope-aware correlates of protection or risk have been established for major infections, and most human T cell profiling still occurs at the level of representative epitopes or bulk assays. In this review, I propose that variation in specificity, defined as distinct epitope-level responses across individuals, represents a largely untested critical element in our assessment of anti-pathogen and antitumor immunity. Advances in TCR repertoire profiling and reverse epitope discovery now make formal tests of this hypothesis tractable, and I discuss approaches to determine the extent to which epitope variation underlies the noise and variation of human immune phenotypes.

Pancreatic beta cells are active participants in the loss of immune tolerance because their high secretory demand creates exceptional vulnerability to inflammatory exposure, endoplasmic reticulum (ER) stress, calcium dysregulation, and redox imbalance. These conditions can promote enzymatic and nonenzymatic changes in self-proteins, generating post-translationally modified (PTM) antigens and neoepitopes that were absent or poorly represented during immune tolerance. This review integrates the contributions with the broader type 1 diabetes (T1D) field to examine how beta cell stress, transglutaminase-2 dependent reactions, peptidyl-arginine deiminase mediated citrullination, oxidative modifications, and post-translational peptide fusion reshape the antigenic landscape. We consider how these products alter antigen presentation and recognition by autoreactive T cells and how an evolving neoepitope repertoire may promote antigenic diversification and epitope spreading. Importantly, clonal recruitment of new determinants must be distinguished from apparent broadening caused by T cell receptor (TCR) cross-reactivity, dual-TCR expression, bystander activation, or tissue redistribution. Defining when modified epitopes arise, which are naturally presented, and which responses contribute to pathogenesis could improve biomarker development, patient stratification, and therapeutic strategies designed to reduce beta cell stress, thus enhancing survival and identity maintenance, while restoring immune tolerance.

This review discusses evidence indicating that the human nasopharyngeal mucosa includes a heterogeneous subset of memory B cells expressing IgD but not IgM. Such IgD+IgM- memory (IgD-ME) B cells differentiate from germinal center B cell precursors that have undergone class switching from IgM to IgD in tonsillar lymphoid follicles. This DNA recombination process is associated with a mutation-intensive program aimed at generating broadly reactive IgD antibodies that target a vast array of microbial and non-microbial antigens. We review data indicating that many IgD-ME B cells express atypical-like B cell properties, accumulate polyreactivity and some autoreactivity in a somatic hypermutation-dependent manner, and include precursors of IgD-secreting IgD+IgM- plasma cells (IgD-PCs). We further discuss the complex innate and adaptive signaling networks underlying the induction of IgD-ME B cells and their clonal differentiation into IgD-PCs. Furthermore, we consider the broad reactivity and functional effects of the IgD antibodies expressed by IgD-ME B cells and released by IgD-PCs. Moreover, we delve into the mechanisms whereby secreted IgD antibodies may contribute to mucosal tolerance and homeostasis. Lastly, we examine human disorders associated with abnormalities of the IgD response.

Interleukin (IL)-10-producing B cells represent an archetype for the regulatory functions of B cells. Here, we review the cell-surface receptors and transcriptional regulators that control IL-10 expression in B cells and enable their regulatory function in vivo. To integrate available data, we propose a three-layer competence-execution-stabilization (CES) framework for IL-10-mediated B cell regulation. We propose that B cell receptor signaling during B cell development and at homeostasis contributes to establishing high IL-10 competence in selected B cell subsets at steady state. Execution of IL-10 expression and regulatory function then requires the integration of canonical B cell activation pathways with environmental signals and involves transcriptional programs that are largely intertwined with those controlling plasma cell differentiation. Notably, chemically induced IL-10-producing B cells progressively differentiate into IL-10-producing plasma cells and acquire homogeneous and stable IL-10 expression. Together with LAG-3[+]CD138[hi] plasma cells, these cells represent the only B cell populations described so far in which IL-10 expression approaches population-wide penetrance. We therefore propose that plasma cell differentiation may constitute an attractor state for stabilizing IL-10-producing regulatory programs. The CES framework provides a conceptual basis for investigating the diversity and regulatory functions of IL-10-expressing B cell states in vivo.

Post-translational modifications (PTMs) chemically diversify the ~20,000 genomically encoded proteins into millions of functional variants, and this diversity has profound consequences for T cell-mediated immune responses. PTM-reactive T cells have been identified across numerous autoimmune diseases and, in several cases, implicated directly as pathogenic drivers. Mechanistically, PTMs promote autoimmunity through two intertwined routes: by altering peptide-MHC and TCR binding to favor recognition of self-epitopes, and by generating neoepitopes in peripheral and/or inflamed tissues that are absent from the thymus, allowing PTM-reactive T cells to escape central tolerance. This cascade is amplified by inflammatory cytokines, upregulated antigen-processing machinery, disease-associated HLA alleles, and, in some cases, exogenous triggers such as diet. We use three well-characterized examples-citrullination in multiple autoimmune disorders, hybrid insulin peptide (HIP) formation in type 1 diabetes, and gluten deamidation in celiac disease-to illustrate how distinct PTM biology, tissue specificity, and genetic encodability shape the antigenic landscape and dictate feasible discovery strategies. We then review four complementary approaches for identifying PTM-specific T cell responses: mass spectrometry-based proteomics and immunopeptidomics, computational MHC-binding prediction, antigen-directed methods, and high-throughput T cell receptor (TCR)-directed library screening methods. For each, we discuss underlying principles, strengths, and limitations, with particular attention to the challenge posed by non-canonical, PTM-bearing residues. By comparing citrullination, HIPs, and gluten deamidation across these methodologies, we highlight how biochemical alterations, tissue accessibility, and mouse-to-human conservation determine which discovery approach is most tractable. We conclude that while synthetic peptide-based methods remain the standard for validating known epitopes, emerging high-throughput, cell-based platforms are rapidly expanding the accessibility of the PTM antigenic landscape for both known and undiscovered autoimmune epitopes.

BAFF and APRIL are TNF superfamily proteins that bind to BAFFR, TACI and BCMA, members of the TNF receptor superfamily. These proteins have both unique and overlapping roles in B cell development and survival and are major therapeutic targets for antibody- and B-cell-driven pathologies. BAFF and BAFFR are required for development and survival of follicular and marginal zone (MZ) B cells, whereas BAFF and APRIL acting through TACI and BCMA support plasma cell survival. TACI and BCMA can both be cleaved to generate soluble decoy receptors binding to BAFF or APRIL, forming feedback circuits. Thus, loss of TACI leads to an increase in BAFF, resulting in B cell hyperplasia. Recent work showed that TACI is required for MZ B cell development, a finding that has implications for understanding immune dysfunction in humans. Both monoallelic and biallelic loss-of-function TACI mutations result in immunodeficiency, potentially due to impaired MZ B cell function. Paradoxically, monoallelic TACI mutations predispose to autoimmunity. We propose this may be due to increased BAFF levels which promote selection of self-reactive B cell clones into the mature B cell pool, particularly the MZ B cell compartment. A deeper understanding of this ligand-receptor system is essential for effective therapeutic targeting.

Post-translational modifications (PTMs) of proteins are essential to maintain homeostasis as many cellular processes rely on reversible PTMs. However, several PTMs, particularly irreversible PTMs in the extracellular space, can contribute to tissue dysfunction, inflammation, and may even trigger the development of autoimmunity against PTM-proteins. Such disease-associated PTMs are formed under conditions of metabolic stress, renal failure, smoking, or pollutants. Importantly, several of these PTMs have recently been shown to have the capacity to activate the complement system directly. These PTMs recruit C1q or Ficolin-3, thereby initiating complement activation. This induces on the one hand opsonins for clearance of (cellular) debris and on the other hand the production of inflammatory mediators. In many autoimmune diseases, patients develop a variety of autoantibodies against PTMs, further amplifying complement activation on PTM-proteins. Notably, anti-PTM antibodies are also present, albeit in lower concentrations, in diseases that are not characterized as typical autoimmune diseases. We therefore hypothesize that PTM-driven complement activation is an underlying factor for chronic low-level inflammation, thereby predisposing individuals to cardiovascular and metabolic diseases. In addition, the presence of complement activation fragments on PTM-proteins may facilitate the breakdown of immune tolerance towards PTM-proteins, further facilitating the production of anti-PTM autoantibodies. Finally, several interventions are discussed that either decrease the PTM-proteins or provide targeted complement inhibition specifically on PTM-proteins. Altogether, complement activation by PTM-proteins may represent an underestimated contributor to chronic inflammation.

Rheumatoid Arthritis (RA) is a prevalent form of inflammatory arthritis in which post-translational modifications are known to play a major role in the loss of tolerance. This review summarizes current knowledge of the role that such modifications play in the disease. Most individuals with RA have antibody responses directed against citrullinated and/or carbamylated self-antigens, but additional modifications are under active investigation. Post-translational modifications can alter the primary protein sequence, increasing binding to HLA molecules and recognition by T cells, and can generate new epitopes that are recognized by autoantibodies. They can also perturb protein structure and alter protein processing, enhancing presentation of cryptic epitopes. Finally, these modifications can alter protein function, potentially undermining cellular identity and function. Potentiation of post-translational modifications has been directly linked to environmental exposures that are epidemiologically tied to RA risk. The immunogenic effects of protein modification are associated with key variants that confer genetic risk for RA. Notably, immune assays designed to measure antibody recognition and monitor T cell responses directed against citrullinated and carbamylated antigens have emerged as useful biomarkers of disease risk and activity. The pathways of protein citrullination and carbamylation are also important targets for potential preventative and curative therapies.

This review will focus on some aspects of the biology of human peripheral B cells and how this biology might inform us as to why these cells are important therapeutic targets. Transitional B cells and the alterations in metabolism in transitional B cells as they mature may be relevant to the induction of B cell tolerance. The human extrafollicular B cell response and the activated B cell subsets that emerge both in extrafollicular responses as well as in germinal center responses will be discussed. The relatively transient nature of activated B cells in contrast to durable memory B cells will be considered and the review will also discuss how durable responses are generated both by extrafollicular responses as well as by germinal center responses.

T-cell receptor repertoires are commonly characterized through sequence diversity and antigen binding, while the cellular functions associated with individual receptors are often measured separately. In this review, I consider how functional information can be incorporated more directly into repertoire analysis by linking receptor identity with measurable effector outputs at single-cell resolution. I focus on Nanovial-based technologies that combine antigen-specific capture, stimulation, secretion measurement, viable-cell recovery, and downstream sequencing in workflows compatible with flow cytometry and single-cell sequencing platforms. Studies using conventional peptide-major histocompatibility complex ligands, the non-classical antigen-presenting molecules MR1 and CD1d, and secretion-encoded single-cell sequencing illustrate how these approaches can complement binding-based enrichment by adding information about secretory effector responses. I also discuss the extension of this framework from ligand-defined screening to dyad-resolved assays involving target cells, which can reveal how receptor identity relates to transcriptional and functional programs arising during direct cell-cell interactions. Together, these developments provide tools for studying relationships among TCR sequence, antigen recognition, and cellular response, with potential applications in therapeutic receptor discovery and in generating datasets for predictive models of T-cell function.

This review does not aspire to provide comprehensive coverage of any particular area. Rather, it is an attempt to reconsider and summarize the results produced by our team in the field of TCR repertoires over the past 15 years, and to correct some of the errors along the way. As intended, it is unavoidably subjective, one-sided, and full of citations to the author's publications. Nevertheless, I hope it will be useful to the immunological community. Someone, having read this relatively short text, will realize that there is no need to spend time studying our previous publications. And one might discover the only useful point of our collective effort hidden in a Supplementary Figure X of a previously unnoticed publication Y. The key points are formulated below as a set of short chapters. Their order does not follow the chronology of the cited works and only intermittently resembles a coherent narrative. In an attempt to summarize, we will speak here about the TCR repertoire preferences of T cell subsets, the plasticity and stubbornness of clonal memory, the price we pay for the privilege of possessing adaptive immunity, some secrets of healthy longevity, about repertoire-based diagnostics, and about immunotherapeutic approaches considered as a toolbox for correcting mistakes made by the immune system.

Human infection with or vaccination against influenza A virus has been one of the most informative systems for understanding how human antibody repertoires recognize viral glycoproteins. Decades of antibody isolation, repertoire sequencing, lineage tracing, serology, viral antigenic evolution, and structural biology have shown that antibody recognition is not simply a matter of "which epitope is targeted." Instead, each surface on the viral hemagglutinin (HA) presents a distinct structural problem for the immune system. Repeated exposures through infection and vaccination coupled with continuous antigenic drift generate complex immune histories and reveal which viral surfaces evolve under antibody pressure. Structural studies have transformed the field by showing how antibody repertoires solve epitope-specific recognition problems. Some epitopes, such as the HA central stem, recruit highly stereotyped genetic and structural solutions. Others, such as the receptor-binding site, can be approached by genetically diverse antibodies that converge on common structural solutions of receptor mimicry. Additional epitopes, including the lateral patch, anchor, head interface, and head-stem junction, reveal intermediate patterns of repertoire constraint. Together, these examples show that what antibody repertoires "see" is determined by the interplay between viral glycoprotein structure, B cell precursor availability, somatic evolution, and population-level immunoglobulin diversity.

In healthy individuals, autoreactive B cells are normally silenced via central and peripheral tolerance mechanisms. In the periphery, when autoreactive B cells recognize self-antigen via their B cell receptor (BCR) (signal 1) but do not receive T cell help (signal 2), they undergo the tolerance mechanism called anergy. Human anergic B cells are termed BND (Naïve IgD+ B cells). Previous studies have shown that BND cells are decreased in the peripheral blood of individuals with autoimmunity, suggesting they may have become activated and contributed to the development of disease. In line with this, recently it has been found that a portion of the remaining anergic B cells present in the blood of autoimmune individuals exhibit an activated phenotype. In this review, we will summarize what is known and unknown about these cells, including their relation to other recognized autoreactive B cell subsets, how they become activated in the first place, and what their pathogenic functions are. Understanding how BND cells break tolerance and contribute to disease will advance the development of new therapeutic approaches and highlight novel cell populations amenable to depletion.

As B cells are activated and differentiate into plasma cells, they execute migration programs that allow them to traffic within secondary lymphoid tissues and reach the bone marrow, where they can mature as long-lived plasma cells. Recent studies have revealed mechanisms underlying not only affinity-driven but also isotype-driven plasma cell selection. In addition, the importance of migration programs in plasma cell longevity has become increasingly appreciated. This review focuses on mechanisms that regulate plasma cell numbers and their migration to the bone marrow, with emphasis on our findings and recent advances in related fields.

The development of highly accurate deep learning models for protein structure prediction has transformed the landscape of T-cell receptor (TCR) structure data, which can now be accessed at repertoire scale. We provide a perspective on the growing field of structural TCR immunoinformatics, summarizing core principles of TCR structural biology and highlighting existing resources and tools. We outline computational methods for TCR structure prediction, and discuss outstanding challenges faced by current tools, as well as potential avenues to address these. We expand on the research enabled by the availability of predicted TCR structures, exploring the utility of TCR structure predictions for computationally inferring TCR specificity, as well as summarizing opportunities emerging from the adjacent field of antibody research. Finally, we provide a forward-looking perspective on the advances in deep learning research which have recently enabled computational design of TCRs and TCR-like binders.