Assessing the feasibility of statistical inference using synthetic antibody-antigen datasets.

Simulation frameworks are useful to stress-test predictive models when data is scarce, or to assert model sensitivity to specific data distributions. Such frameworks often need to recapitulate several layers of data complexity, including emergent properties that arise implicitly from the interaction between simulation components. Antibody-antigen binding is a complex mechanism by which an antibody sequence wraps itself around an antigen with high affinity. In this study, we use a synthetic simulation framework for antibody-antigen folding and binding on a 3D lattice that include full details on the spatial conformation of both molecules. We investigate how emergent properties arise in this framework, in particular the physical proximity of amino acids, their presence on the binding interface, or the binding status of a sequence, and relate that to the individual and pairwise contributions of amino acids in statistical models for binding prediction. We show that weights learnt from a simple logistic regression model align with some but not all features of amino acids involved in the binding, and that predictive sequence binding patterns can be enriched. In particular, main effects correlated with the capacity of a sequence to bind any antigen, while statistical interactions were related to sequence specificity.

Spatial dysregulation of T follicular helper cells impairs vaccine responses in aging.

The magnitude and quality of the germinal center (GC) response decline with age, resulting in poor vaccine-induced immunity in older individuals. A functional GC requires the co-ordination of multiple cell types across time and space, in particular across its two functionally distinct compartments: the light and dark zones. In aged mice, there is CXCR4-mediated mislocalization of T follicular helper (TFH) cells to the dark zone and a compressed network of follicular dendritic cells (FDCs) in the light zone. Here we show that TFH cell localization is critical for the quality of the antibody response and for the expansion of the FDC network upon immunization. The smaller GC and compressed FDC network in aged mice were corrected by provision of TFH cells that colocalize with FDCs using CXCR5. This demonstrates that the age-dependent defects in the GC response are reversible and shows that TFH cells support stromal cell responses to vaccines.

Toward real-world automated antibody design with combinatorial Bayesian optimization.

Antibodies are multimeric proteins capable of highly specific molecular recognition. The complementarity determining region 3 of the antibody variable heavy chain (CDRH3) often dominates antigen-binding specificity. Hence, it is a priority to design optimal antigen-specific CDRH3 to develop therapeutic antibodies. The combinatorial structure of CDRH3 sequences makes it impossible to query binding-affinity oracles exhaustively. Moreover, antibodies are expected to have high target specificity and developability. Here, we present AntBO, a combinatorial Bayesian optimization framework utilizing a CDRH3 trust region for an in silico design of antibodies with favorable developability scores. The in silico experiments on 159 antigens demonstrate that AntBO is a step toward practically viable in vitro antibody design. In under 200 calls to the oracle, AntBO suggests antibodies outperforming the best binding sequence from 6.9 million experimentally obtained CDRH3s. Additionally, AntBO finds very-high-affinity CDRH3 in only 38 protein designs while requiring no domain knowledge.

Germinal centers are permissive to subdominant antibody responses.

Introduction: A protective humoral response to pathogens requires the development of high affinity antibodies in germinal centers (GC). The combination of antigens available during immunization has a strong impact on the strength and breadth of the antibody response. Antigens can display various levels of immunogenicity, and a hierarchy of immunodominance arises when the GC response to an antigen dampens the response to other antigens. Immunodominance is a challenge for the development of vaccines to mutating viruses, and for the development of broadly neutralizing antibodies. The extent by which antigens with different levels of immunogenicity compete for the induction of high affinity antibodies and therefore contribute to immunodominance is not known. Methods: Here, we perform in silico simulations of the GC response, using a structural representation of antigens with complex surface amino acid composition and topology. We generate antigens with complex domains of different levels of immunogenicity and perform simulations with combinations of these domains. Results: We found that GC dynamics were driven by the most immunogenic domain and immunodominance arose as affinity maturation to less immunogenic domain was inhibited. However, this inhibition was moderate since the less immunogenic domain exhibited a weak GC response in the absence of the most immunogenic domain. Less immunogenic domains reduced the dominance of GC responses to more immunogenic domains, albeit at a later time point. Discussion: The simulations suggest that increased vaccine valency may decrease immunodominance of the GC response to strongly immunogenic domains and therefore, act as a potential strategy for the natural induction of broadly neutralizing antibodies in GC reactions.

Reference-based comparison of adaptive immune receptor repertoires.

B and T cell receptor (immune) repertoires can represent an individual's immune history. While current repertoire analysis methods aim to discriminate between health and disease states, they are typically based on only a limited number of parameters. Here, we introduce immuneREF: a quantitative multidimensional measure of adaptive immune repertoire (and transcriptome) similarity that allows interpretation of immune repertoire variation by relying on both repertoire features and cross-referencing of simulated and experimental datasets. To quantify immune repertoire similarity landscapes across health and disease, we applied immuneREF to >2,400 datasets from individuals with varying immune states (healthy, [autoimmune] disease, and infection). We discovered, in contrast to the current paradigm, that blood-derived immune repertoires of healthy and diseased individuals are highly similar for certain immune states, suggesting that repertoire changes to immune perturbations are less pronounced than previously thought. In conclusion, immuneREF enables the population-wide study of adaptive immune response similarity across immune states.

In silico proof of principle of machine learning-based antibody design at unconstrained scale.

Generative machine learning (ML) has been postulated to become a major driver in the computational design of antigen-specific monoclonal antibodies (mAb). However, efforts to confirm this hypothesis have been hindered by the infeasibility of testing arbitrarily large numbers of antibody sequences for their most critical design parameters: paratope, epitope, affinity, and developability. To address this challenge, we leveraged a lattice-based antibody-antigen binding simulation framework, which incorporates a wide range of physiological antibody-binding parameters. The simulation framework enables the computation of synthetic antibody-antigen 3D-structures, and it functions as an oracle for unrestricted prospective evaluation and benchmarking of antibody design parameters of ML-generated antibody sequences. We found that a deep generative model, trained exclusively on antibody sequence (one dimensional: 1D) data can be used to design conformational (three dimensional: 3D) epitope-specific antibodies, matching, or exceeding the training dataset in affinity and developability parameter value variety. Furthermore, we established a lower threshold of sequence diversity necessary for high-accuracy generative antibody ML and demonstrated that this lower threshold also holds on experimental real-world data. Finally, we show that transfer learning enables the generation of high-affinity antibody sequences from low-N training data. Our work establishes a priori feasibility and the theoretical foundation of high-throughput ML-based mAb design.

Progress and challenges for the machine learning-based design of fit-for-purpose monoclonal antibodies.

Although the therapeutic efficacy and commercial success of monoclonal antibodies (mAbs) are tremendous, the design and discovery of new candidates remain a time and cost-intensive endeavor. In this regard, progress in the generation of data describing antigen binding and developability, computational methodology, and artificial intelligence may pave the way for a new era of in silico on-demand immunotherapeutics design and discovery. Here, we argue that the main necessary machine learning (ML) components for an in silico mAb sequence generator are: understanding of the rules of mAb-antigen binding, capacity to modularly combine mAb design parameters, and algorithms for unconstrained parameter-driven in silico mAb sequence synthesis. We review the current progress toward the realization of these necessary components and discuss the challenges that must be overcome to allow the on-demand ML-based discovery and design of fit-for-purpose mAb therapeutic candidates.

Unconstrained generation of synthetic antibody-antigen structures to guide machine learning methodology for antibody specificity prediction.

Machine learning (ML) is a key technology for accurate prediction of antibody-antigen binding. Two orthogonal problems hinder the application of ML to antibody-specificity prediction and the benchmarking thereof: the lack of a unified ML formalization of immunological antibody-specificity prediction problems and the unavailability of large-scale synthetic datasets to benchmark real-world relevant ML methods and dataset design. Here we developed the Absolut! software suite that enables parameter-based unconstrained generation of synthetic lattice-based three-dimensional antibody-antigen-binding structures with ground-truth access to conformational paratope, epitope and affinity. We formalized common immunological antibody-specificity prediction problems as ML tasks and confirmed that for both sequence- and structure-based tasks, accuracy-based rankings of ML methods trained on experimental data hold for ML methods trained on Absolut!-generated data. The Absolut! framework has the potential to enable real-world relevant development and benchmarking of ML strategies for biotherapeutics design.

Individualized VDJ recombination predisposes the available Ig sequence space.

The process of recombination between variable (V), diversity (D), and joining (J) immunoglobulin (Ig) gene segments determines an individual's naive Ig repertoire and, consequently, (auto)antigen recognition. VDJ recombination follows probabilistic rules that can be modeled statistically. So far, it remains unknown whether VDJ recombination rules differ between individuals. If these rules differed, identical (auto)antigen-specific Ig sequences would be generated with individual-specific probabilities, signifying that the available Ig sequence space is individual specific. We devised a sensitivity-tested distance measure that enables inter-individual comparison of VDJ recombination models. We discovered, accounting for several sources of noise as well as allelic variation in Ig sequencing data, that not only unrelated individuals but also human monozygotic twins and even inbred mice possess statistically distinguishable immunoglobulin recombination models. This suggests that, in addition to genetic, there is also nongenetic modulation of VDJ recombination. We demonstrate that population-wide individualized VDJ recombination can result in orders of magnitude of difference in the probability to generate (auto)antigen-specific Ig sequences. Our findings have implications for immune receptor-based individualized medicine approaches relevant to vaccination, infection, and autoimmunity.

Characterization of mechanisms positioning costimulatory complexes in immune synapses.

Small immunoglobulin superfamily (sIGSF) adhesion complexes form a corolla of microdomains around an integrin ring and secretory core during immunological synapse (IS) formation. The corolla recruits and retains major costimulatory/checkpoint complexes, such as CD28, making forces that govern corolla formation of particular interest. Here, we investigated the mechanisms underlying molecular reorganization of CD2, an adhesion and costimulatory molecule of the sIGSF family during IS formation. Computer simulations showed passive distal exclusion of CD2 complexes under weak interactions with the ramified F-actin transport network. Attractive forces between CD2 and CD28 complexes relocate CD28 from the IS center to the corolla. Size-based sorting interactions with large glycocalyx components, such as CD45, or short-range CD2 self-attraction successfully explain the corolla 'petals.' This establishes a general simulation framework for complex pattern formation observed in cell-bilayer and cell-cell interfaces, and the suggestion of new therapeutic targets, where boosting or impairing characteristic pattern formation can be pivotal.

Nitric oxide controls proliferation of Leishmania major by inhibiting the recruitment of permissive host cells.

Nitric oxide (NO) is an important antimicrobial effector but also prevents unnecessary tissue damage by shutting down the recruitment of monocyte-derived phagocytes. Intracellular pathogens such as Leishmania major can hijack these cells as a niche for replication. Thus, NO might exert containment by restricting the availability of the cellular niche required for efficient pathogen proliferation. However, such indirect modes of action remain to be established. By combining mathematical modeling with intravital 2-photon biosensors of pathogen viability and proliferation, we show that low L. major proliferation results not from direct NO impact on the pathogen but from reduced availability of proliferation-permissive host cells. Although inhibiting NO production increases recruitment of these cells, and thus pathogen proliferation, blocking cell recruitment uncouples the NO effect from pathogen proliferation. Therefore, NO fulfills two distinct functions for L. major containment: permitting direct killing and restricting the supply of proliferation-permissive host cells.

Coupled Antigen and BLIMP1 Asymmetric Division With a Large Segregation Between Daughter Cells Recapitulates the Temporal Transition From Memory B Cells to Plasma Cells and a DZ-to-LZ Ratio in the Germinal Center.

Memory B cells and antibody-secreting plasma cells are generated within germinal centers during affinity maturation in which B-cell proliferation, selection, differentiation, and self-renewal play important roles. The mechanisms behind memory B cell and plasma cell differentiation in germinal centers are not well understood. However, it has been suggested that cell fate is (partially) determined by asymmetric cell division, which involves the unequal distribution of cellular components to both daughter cells. To investigate what level and/or probability of asymmetric segregation of several fate determinant molecules, such as the antigen and transcription factors (BCL6, IRF4, and BLIMP1) recapitulates the temporal switch and DZ-to-LZ ratio in the germinal center, we implemented a multiscale model that combines a core gene regulatory network for plasma cell differentiation with a model describing the cellular interactions and dynamics in the germinal center. Our simulations show that BLIMP1 driven plasma cell differentiation together with coupled asymmetric division of antigen and BLIMP1 with a large segregation between the daughter cells results in a germinal center DZ-to-LZ ratio and a temporal switch from memory B cells to plasma cells that have been observed in experiments.

Ymir: A 3D structural affinity model for multi-epitope vaccine simulations.

Vaccine development is challenged by the hierarchy of immunodominance between target antigen epitopes and the emergence of antigenic variants by pathogen mutation. The strength and breadth of antibody responses relies on selection and mutation in the germinal center and on the structural similarity between antigens. Computational methods for assessing the breadth of germinal center responses to multivalent antigens are critical to speed up vaccine development. Yet, such methods have poorly reflected the 3D antigen structure and antibody breadth. Here, we present Ymir, a new 3D-lattice-based framework that calculates in silico antibody-antigen affinities. Key physiological properties naturally emerge from Ymir such as affinity jumps, cross-reactivity, and differential epitope accessibility. We validated Ymir by replicating known features of germinal center dynamics. We show that combining antigens with mutated but structurally related epitopes enhances vaccine breadth. Ymir opens a new avenue for understanding vaccine potency based on the structural relationship between vaccine antigens.

Germinal Centre Shutdown.

Germinal Centres (GCs) are transient structures in secondary lymphoid organs, where affinity maturation of B cells takes place following an infection. While GCs are responsible for protective antibody responses, dysregulated GC reactions are associated with autoimmune disease and B cell lymphoma. Typically, 'normal' GCs persist for a limited period of time and eventually undergo shutdown. In this review, we focus on an important but unanswered question - what causes the natural termination of the GC reaction? In murine experiments, lack of antigen, absence or constitutive T cell help leads to premature termination of the GC reaction. Consequently, our present understanding is limited to the idea that GCs are terminated due to a decrease in antigen access or changes in the nature of T cell help. However, there is no direct evidence on which biological signals are primarily responsible for natural termination of GCs and a mechanistic understanding is clearly lacking. We discuss the present understanding of the GC shutdown, from factors impacting GC dynamics to changes in cellular interactions/dynamics during the GC lifetime. We also address potential missing links and remaining questions in GC biology, to facilitate further studies to promote a better understanding of GC shutdown in infection and immune dysregulation.

Modeling the Dynamics of T-Cell Development in the Thymus.

The thymus hosts the development of a specific type of adaptive immune cells called T cells. T cells orchestrate the adaptive immune response through recognition of antigen by the highly variable T-cell receptor (TCR). T-cell development is a tightly coordinated process comprising lineage commitment, somatic recombination of Tcr gene loci and selection for functional, but non-self-reactive TCRs, all interspersed with massive proliferation and cell death. Thus, the thymus produces a pool of T cells throughout life capable of responding to virtually any exogenous attack while preserving the body through self-tolerance. The thymus has been of considerable interest to both immunologists and theoretical biologists due to its multi-scale quantitative properties, bridging molecular binding, population dynamics and polyclonal repertoire specificity. Here, we review experimental strategies aimed at revealing quantitative and dynamic properties of T-cell development and how they have been implemented in mathematical modeling strategies that were reported to help understand the flexible dynamics of the highly dividing and dying thymic cell populations. Furthermore, we summarize the current challenges to estimating in vivo cellular dynamics and to reaching a next-generation multi-scale picture of T-cell development.

A compact vocabulary of paratope-epitope interactions enables predictability of antibody-antigen binding.

Antibody-antigen binding relies on the specific interaction of amino acids at the paratope-epitope interface. The predictability of antibody-antigen binding is a prerequisite for de novo antibody and (neo-)epitope design. A fundamental premise for the predictability of antibody-antigen binding is the existence of paratope-epitope interaction motifs that are universally shared among antibody-antigen structures. In a dataset of non-redundant antibody-antigen structures, we identify structural interaction motifs, which together compose a commonly shared structure-based vocabulary of paratope-epitope interactions. We show that this vocabulary enables the machine learnability of antibody-antigen binding on the paratope-epitope level using generative machine learning. The vocabulary (1) is compact, less than 104 motifs; (2) distinct from non-immune protein-protein interactions; and (3) mediates specific oligo- and polyreactive interactions between paratope-epitope pairs. Our work leverages combined structure- and sequence-based learning to demonstrate that machine-learning-driven predictive paratope and epitope engineering is feasible.

Influenza A virus-induced thymus atrophy differentially affects dynamics of conventional and regulatory T-cell development in mice.

Foxp3+ Treg cells, which are crucial for maintenance of self-tolerance, mainly develop within the thymus, where they arise from CD25+ Foxp3- or CD25- Foxp3+ Treg cell precursors. Although it is known that infections can cause transient thymic involution, the impact of infection-induced thymus atrophy on thymic Treg (tTreg) cell development is unknown. Here, we infected mice with influenza A virus (IAV) and studied thymocyte population dynamics post infection. IAV infection caused a massive, but transient thymic involution, dominated by a loss of CD4+ CD8+ double-positive (DP) thymocytes, which was accompanied by a significant increase in the frequency of CD25+ Foxp3+ tTreg cells. Differential apoptosis susceptibility could be experimentally excluded as a reason for the relative tTreg cell increase, and mathematical modeling suggested that enhanced tTreg cell generation cannot explain the increased frequency of tTreg cells. Yet, an increased death of DP thymocytes and augmented exit of single-positive (SP) thymocytes was suggested to be causative. Interestingly, IAV-induced thymus atrophy resulted in a significantly reduced T-cell receptor (TCR) repertoire diversity of newly produced tTreg cells. Taken together, IAV-induced thymus atrophy is substantially altering the dynamics of major thymocyte populations, finally resulting in a relative increase of tTreg cells with an altered TCR repertoire.

The immuneML ecosystem for machine learning analysis of adaptive immune receptor repertoires.

Adaptive immune receptor repertoires (AIRR) are key targets for biomedical research as they record past and ongoing adaptive immune responses. The capacity of machine learning (ML) to identify complex discriminative sequence patterns renders it an ideal approach for AIRR-based diagnostic and therapeutic discovery. To date, widespread adoption of AIRR ML has been inhibited by a lack of reproducibility, transparency, and interoperability. immuneML (immuneml.uio.no) addresses these concerns by implementing each step of the AIRR ML process in an extensible, open-source software ecosystem that is based on fully specified and shareable workflows. To facilitate widespread user adoption, immuneML is available as a command-line tool and through an intuitive Galaxy web interface, and extensive documentation of workflows is provided. We demonstrate the broad applicability of immuneML by (i) reproducing a large-scale study on immune state prediction, (ii) developing, integrating, and applying a novel deep learning method for antigen specificity prediction, and (iii) showcasing streamlined interpretability-focused benchmarking of AIRR ML.

Evaluation of CD8 T cell killing models with computer simulations of 2-photon imaging experiments.

In vivo imaging of cytotoxic T lymphocyte (CTL) killing activity revealed that infected cells have a higher observed probability of dying after multiple contacts with CTLs. We developed a three-dimensional agent-based model to discriminate different hypotheses about how infected cells get killed based on quantitative 2-photon in vivo observations. We compared a constant CTL killing probability with mechanisms of signal integration in CTL or infected cells. The most likely scenario implied increased susceptibility of infected cells with increasing number of CTL contacts where the total number of contacts was a critical factor. However, when allowing in silico T cells to initiate new interactions with apoptotic target cells (zombie contacts), a contact history independent killing mechanism was also in agreement with experimental datasets. The comparison of observed datasets to simulation results, revealed limitations in interpreting 2-photon data, and provided readouts to distinguish CTL killing models.

Agent-Based Modeling of T Cell Receptor Cooperativity.

Immunological synapse (IS) formation is a key event during antigen recognition by T cells. Recent experimental evidence suggests that the affinity between T cell receptors (TCRs) and antigen is actively modulated during the early steps of TCR signaling. In this work, we used an agent-based model to study possible mechanisms for affinity modulation during IS formation. We show that, without any specific active mechanism, the observed affinity between receptors and ligands evolves over time and depends on the density of ligands of the antigen peptide presented by major histocompatibility complexes (pMHC) and TCR molecules. A comparison between the presence or absence of TCR-pMHC centrally directed flow due to F-actin coupling suggests that centripetal transport is a potential mechanism for affinity modulation. The model further suggests that the time point of affinity measurement during immune synapse formation is critical. Finally, a mathematical model of F-actin foci formation incorporated in the agent-based model shows that TCR affinity can potentially be actively modulated by positive/negative feedback of the F-actin foci on the TCR-pMHC association rate kon.