Antibodies are a class of proteins that recognize and neutralize pathogens by binding to their antigens. They are the most significant category of biopharmaceuticals for both diagnostic and therapeutic applications. Understanding how antibodies interact with their antigens plays a fundamental role in drug and vaccine design and helps to comprise the complex antigen binding mechanisms. Computational methods for predicting interaction sites of antibody-antigen are of great value due to the overall cost of experimental methods. Machine learning methods and deep learning techniques obtained promising results.In this work, we predict antibody interaction interface sites by applying HSS-PPI, a hybrid method defined to predict the interface sites of general proteins. The approach abstracts the proteins in terms of hierarchical representation and uses a graph convolutional network to classify the amino acids between interface and non-interface. Moreover, we also equipped the amino acids with different sets of physicochemical features together with structural ones to describe the residues. Analyzing the results, we observe that the structural features play a fundamental role in the amino acid descriptions. We compare the obtained performances, evaluated using standard metrics, with the ones obtained with SVM with 3D Zernike descriptors, Parapred, Paratome, and Antibody i-Patch.

Large language models trained on sequence information alone can learn high-level principles of protein design. However, beyond sequence, the three-dimensional structures of proteins determine their specific function, activity, and evolvability. Here, we show that a general protein language model augmented with protein structure backbone coordinates can guide evolution for diverse proteins without the need to model individual functional tasks. We also demonstrate that ESM-IF1, which was only trained on single-chain structures, can be extended to engineer protein complexes. Using this approach, we screened about 30 variants of two therapeutic clinical antibodies used to treat severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. We achieved up to 25-fold improvement in neutralization and 37-fold improvement in affinity against antibody-escaped viral variants of concern BQ.1.1 and XBB.1.5, respectively. These findings highlight the advantage of integrating structural information to identify efficient protein evolution trajectories without requiring any task-specific training data.

The optimization of therapeutic antibodies through traditional techniques, such as candidate screening via hybridoma or phage display, is resource-intensive and time-consuming. In recent years, computational and artificial intelligence-based methods have been actively developed to accelerate and improve the development of therapeutic antibodies. In this study, we developed an end-to-end sequence-based deep learning model, termed AttABseq, for the predictions of the antigen-antibody binding affinity changes connected with antibody mutations. AttABseq is a highly efficient and generic attention-based model by utilizing diverse antigen-antibody complex sequences as the input to predict the binding affinity changes of residue mutations. The assessment on the three benchmark datasets illustrates that AttABseq is 120% more accurate than other sequence-based models in terms of the Pearson correlation coefficient between the predicted and experimental binding affinity changes. Moreover, AttABseq also either outperforms or competes favorably with the structure-based approaches. Furthermore, AttABseq consistently demonstrates robust predictive capabilities across a diverse array of conditions, underscoring its remarkable capacity for generalization across a wide spectrum of antigen-antibody complexes. It imposes no constraints on the quantity of altered residues, rendering it particularly applicable in scenarios where crystallographic structures remain unavailable. The attention-based interpretability analysis indicates that the causal effects of point mutations on antibody-antigen binding affinity changes can be visualized at the residue level, which might assist automated antibody sequence optimization. We believe that AttABseq provides a fiercely competitive answer to therapeutic antibody optimization.

Autoantibodies to nuclear antigens are hallmarks of systemic lupus erythematosus (SLE) where they contribute to pathogenesis. However, there remains a gap in our knowledge regarding how different isotypes of autoantibodies contribute to this autoimmune disease, including the production of the critical type I interferon (IFN) cytokines by plasmacytoid dendritic cells (pDCs) in response to immune complexes (ICs). We focused on IgA, which is the second-most prevalent isotype in serum and, along with IgG, is deposited in glomeruli in individuals with lupus nephritis. We show that individuals with SLE have serum IgA autoantibodies against most nuclear antigens, correlating with IgG against the same antigen. We investigated whether IgA autoantibodies against a major SLE autoantigen, Smith ribonucleoprotein (Sm/RNP), played a role in IC activation of pDCs. We found that pDCs expressed the IgA-specific Fc receptor, FcαR, and IgA1 autoantibodies synergized with IgG in RNA-containing ICs to generate robust primary blood pDC IFN-α responses in vitro. pDC responses to these ICs required both FcαR and FcγRIIa, showing synergy between these Fc receptors. Sm/RNP IC binding to and internalization by pDCs were greater when ICs contained both IgA1 and IgG. Circulating pDCs from individuals with SLE had higher binding of IgA1-containing ICs and higher expression of FcαR than pDCs from healthy control individuals. Although pDC FcαR expression correlated with the blood IFN-stimulated gene signature in SLE, Toll-like receptor 7 agonists, but not IFN-α, up-regulated pDC FcαR expression in vitro. Together, we show a mechanism by which IgA1 autoantibodies contribute to SLE pathogenesis.

The underlying contribution of immune complexes in modulating adaptive immunity in mucosal tissues remains poorly understood. In this report, we examined, in mice, the proinflammatory response elicited by intranasal delivery of the biothreat agent ricin toxin (RT) in association with two toxin-neutralizing mAbs, SylH3 and PB10. We previously demonstrated that ricin-immune complexes (RICs) induce the rapid onset of high-titer toxin-neutralizing Abs that persist for months. We now demonstrate that such responses are dependent on CD4+ T cell help, because treatment of mice with an anti-CD4 mAb abrogated the onset of RT-specific Abs following intranasal RICs exposure. To define the inflammatory environment associated with RIC exposure, we collected bronchoalveolar lavage fluid (BALF) and sera from mice 6, 12, and 18 h after they had received RT or RICs by the intranasal route. A 32-plex cytometric bead array revealed an inflammatory profile elicited by RT that was dominated by IL-6 (>1500-fold increase in BALF) and secondarily by KC (CXCL1), G-CSF, GM-CSF, and MCP-1. RICs induced inflammatory profiles in both BALF and serum response that were similar to RT, albeit at markedly reduced levels. These results demonstrate that RICs retain the capacity to induce local and systemic inflammatory cytokines/chemokines that, in turn, may influence Ag sampling and presentation in the lung mucosa and draining lymph nodes. A better understanding of the fate of immune complexes following intranasal delivery has implications for the development of mucosal vaccines for biothreats and emerging infectious diseases.

Chikungunya virus (CHIKV) is a causative agent of a disease continuum, ranging from an acute transient chikungunya fever to chronic incapacitating viral arthralgia. The interaction between anti-CHIKV antibodies and the complement system has recently received attention. However, the contribution of complement activation in CHIKV-induced pathologies has not been fully elucidated. The present study was undertaken to delineate the possible contribution of complement activation in CHIKV-induced disease progression. In this study, using plasma specimens of chikungunya patients in the acute, chronic, and recovered phases of infection, we explicated the involvement of complement activation in CHIKV disease progression by ELISAs and Bio-Plex assays. Correlation analysis was carried out to demonstrate interrelation among C1q-binding IgG-containing circulating immune complexes (CIC-C1q), complement activation fragments (C3a, C5a, sC5b-9), and complement-modulated pro-inflammatory cytokines (IL-1β, IL-18, IL-6, and TNF-α). We detected elevated complement activation fragments, CIC-C1q, and complement-modulated cytokines in the varied patient groups compared with the healthy controls, indicating persistent activation of the complement system. Furthermore, we observed statistically significant correlations among CIC-C1q with complement activation fragments and C3a with complement modulatory cytokines IL-1β, IL-6, and IL-18 during the CHIKV disease progression. Taken together, the current data provide insight into the plausible association between CICs, complement activation, subsequent complement modulatory cytokine expression, and CHIKV etiopathology.

Although antibody-mediated lung damage is a major factor in transfusion-related acute lung injury (ALI), autoimmune lung disease (for example, coatomer subunit α [COPA] syndrome), and primary graft dysfunction following lung transplantation, the mechanism by which antigen-antibody complexes activate complement to induce lung damage remains unclear. In this issue of the JCI, Cleary and colleagues utilized several approaches to demonstrate that IgG forms hexamers with MHC class I alloantibodies. This hexamerization served as a key pathophysiological mechanism in alloimmune lung injury models and was mediated through the classical pathway of complement activation. Additionally, the authors provided avenues for exploring therapeutics for this currently hard-to-treat clinical entity that has several etiologies but a potentially focused mechanism.

Antibodies play a central role in the adaptive immune response of vertebrates through the specific recognition of exogenous or endogenous antigens. The rational design of antibodies has a wide range of biotechnological and medical applications, such as in disease diagnosis and treatment. However, there are currently no reliable methods for predicting the antibodies that recognize a specific antigen region (or epitope) and, conversely, epitopes that recognize the binding region of a given antibody (or paratope). To fill this gap, we developed ImaPEp, a machine learning-based tool for predicting the binding probability of paratope-epitope pairs, where the epitope and paratope patches were simplified into interacting two-dimensional patches, which were colored according to the values of selected features, and pixelated. The specific recognition of an epitope image by a paratope image was achieved by using a convolutional neural network-based model, which was trained on a set of two-dimensional paratope-epitope images derived from experimental structures of antibody-antigen complexes. Our method achieves good performances in terms of cross-validation with a balanced accuracy of 0.8. Finally, we showcase examples of application of ImaPep, including extensive screening of large libraries to identify paratope candidates that bind to a selected epitope, and rescoring and refining antibody-antigen docking poses.

Single particle analysis from cryogenic transmission electron microscopy (cryo-EM) is particularly attractive for complexes for which structure prediction remains intractable, such as antibody-antigen complexes. Here we obtain the detailed structure of a particularly difficult complex between human epidermal growth factor receptor 2 (HER2) and the antigen-binding fragments from two distinct therapeutic antibodies binding to distant parts of the flexible HER2, pertuzumab and trastuzumab (HTP). We highlight the strengths and limitations of current data processing software in dealing with various kinds of heterogeneities, particularly continuous conformational heterogeneity, and in describing the motions that can be extracted from our dataset. Our HTP structure provides a more detailed view than the one previously available for this ternary complex. This allowed us to pinpoint a previously overlooked loop in domain IV that may be involved both in binding of trastuzumab and in HER2 dimerization. This finding may contribute to explain the synergistic anticancer effect of the two antibodies. We further propose that the flexibility of the HTP complex, beyond the difficulties it causes for cryo-EM analysis, actually reflects regulation of HER2 signaling and its inhibition by therapeutic antibodies. Notably we obtain our best data with ultra-thin continuous carbon grids, showing that with current cameras their use to alleviate particle misdistribution is compatible with a protein complex of only 162 kDa. Perhaps most importantly, we provide here a dataset for such a smallish protein complex for further development of software accounting for continuous conformational heterogeneity in cryo-EM images.

In patients with immune thrombotic thrombocytopenic purpura (iTTP), autoantibodies against the metalloprotease ADAMTS13 lead to catastrophic microvascular thrombosis. However, the potential benefits of recombinant human ADAMTS13 (rADAMTS13) in patients with iTTP remain unknown. Here, we report the clinical use of rADAMTS13, which resulted in the rapid suppression of disease activity and complete recovery in a critically ill patient whose condition had proved to be refractory to all available treatments. We also show that rADAMTS13 causes immune complex formation, which saturates the autoantibody and may promote its clearance. Our data support the role of rADAMTS13 as a novel adjunctive therapy in patients with iTTP.