Abstract Antibody-mediated protection plays a crucial role in post-exposure prophylaxis against rabies infection, while engineered antibodies with novel functional mechanisms offer enhanced preventive efficacy. In this study, we investigated the in vitro and in vivo activities as well as the protective efficacy of CR4008 antibodies with different conformational structures. We demonstrated that under passive immunization with antibodies of equivalent neutralizing efficacy, IgM antibodies exhibited superior protection against viral challenge compared with other antibodies. The underlying mechanism involves antibody barriers along the sarcolemma by IgM antibodies, which effectively inhibits viral transport in the infectious microenvironment within muscle tissues. Following high-affinity binding and neutralizing, barrier antibodies prevent the virus from retrogradely spreading to the central nervous system and prevent rabies development. Our study suggests that, in addition to neutralizing activity, effective inhibition of viral spread caused by the unique distribution of IgM antibodies in exposed tissues is critical and valuable for passive immunization against rabies infection.
Abstract Background The bispecific antibody ivonescimab targets programmed cell death protein 1 (PD-1) and vascular endothelial growth factor (VEGF). Recent clinical trials have shown it has greater efficacy against PD-L1 positive non-small cell lung cancer than pembrolizumab (Keytruda), a frequently prescribed anti-PD-1 monoclonal antibody. Ivonescimab binds to two VEGF and two PD-1 molecules, with complex formation through higher-order structure formation (or ‘daisy chain’ binding). However, the binding stoichiometries and interaction dynamics of ivonescimab with VEGF and PD-1 have not been characterized in depth. Methods We used mass photometry (MP) and kinetic modeling to analyze these interactions, quantifying the complexes formed and their affinities. Dissociation constants (KD) for ivonescimab’s binding to VEGF and PD-1 were calculated from equilibrium counts and real-time measurements, respectively. Results VEGF drove oligomerization of ivonescimab, which bound VEGF predominantly in a 2:2 stoichiometry, with KD = 0.08 nM. Higher-order oligomeric complexes, present only at low abundance, displayed markedly weaker affinities (3.17 nM; 1.29 nM). Ternary complexes of ivonescimab with its two targets consistently presented two PD-1 antigens for each ivonescimab molecule, with a 1.66 nM KD for the binding of the first PD-1 and the slightly stronger 0.89 nM for the second PD-1 molecule. Conclusions MP confirmed VEGF-induced ivonescimab oligomerization and revealed that dimers, not higher-order structures, were the most stable stoichiometry. MP enables detailed analysis of antibody–antigen interactions, even for bispecific antibodies that interact with antigens with complex stoichiometries.
Abstract Background One key step of protein drug development is the screening of protein-protein binding (PPB) affinity. Various deep learning methods have been proposed to predict PPB affinity, but they are often challenged by data leakage, data scarcity, data heterogeneity, and structure-sequence gap. Methods To address these challenges in PPB affinity prediction, we propose a comprehensive framework by developing prediction models, PPBind-3D and PPBind-1D, to predict PPB affinity. PPBind-3D leverages structural information near the protein-protein binding interface to make its predictions. Additionally, PPBind-1D was developed using sequence data to address the lack of structural data in practical applications. Results PPBind-3D, equipped with monotonic neural network-constrained multi-task learning and Self-and-Mutual Attention, demonstrated state-of-the-art performance in PPB prediction across diverse datasets under both strict and random data partition settings. PPBind-1D, trained by aligning to PPBind-3D, overcame structural data scarcity to deliver robust sequence-based predictions while maintaining high accuracy, enabling practical deployment where structural data is unavailable. To facilitate real-world application, PPBind-3D was integrated into an online tool, which is available at https://rddc.tsinghua-gd.org/tool/dg-prediction. Finally, we demonstrated three application cases of our AI models in the virtual screening of protein drugs, further confirming the framework's utility and scalability. Conclusions Comparative and ablation studies demonstrate that PPBind-3D surpasses conventional benchmark approaches. Utilizing solely sequence information, PPBind-1D closely approximates the performance of PPBind-3D, thereby providing a viable alternative when structural data are inaccessible. Validation across three virtual screening case studies confirms the efficacy of both models, illustrating that our models can significantly facilitate high-throughput screening.
Abstract Background Accurate prediction of anti-drug antibody responses remains a challenge in biologics development, as current tools primarily rely on major histocompatibility complex binding affinity, which does not capture the full complexity of immune activation. Methods We developed EPITOP™, a protein language model trained on epitopes experimentally verified via cytokine release assays, enabling direct immunogenicity prediction. This was combined with IMMUNOTOP™, which aggregates EPITOP™ outputs into protein-level scores. Results EPITOP™ achieves an Area Under the Curve (AUC) of 0.73 on the independent test set, demonstrating performance comparable to existing methods while providing biologically grounded and interpretable outputs for immunogenicity risk assessment. It also enables identification of immunogenic hotspots. Conclusions Beyond hotspot identification, the model can serve as a surrogate for peptide-level interferon-γ-associated immunogenicity when experimental measurement is impractical and may be useful for modalities in which experimental cytokine testing is difficult, although systemic cytokine release depends on additional biological factors not modeled here.
Abstract Background Besides required biochemical and biological functions, suitable biophysical characteristics or developability are critical for the success of monoclonal antibody (mAb) therapeutics. Aiming to develop disease-modifying therapeutics for diabetic peripheral neuropathy (DPN), this study applied protein engineering strategies to improve the developability of an matrix metalloproteinase-9 inhibitory mAb. Methods A miniature diversity was introduced to replace two hydrophobic residues in CDR-H3. Following mAb production, comprehensive panels of biochemical and biophysical properties were measured. The optimized mutant was further validated by pharmacokinetics and in vivo efficacy studies. Results The optimized mutant exhibited favorable developability properties while retaining its biochemical activities, including high production yield, colloidal stability, nanomolar binding affinity and inhibition potency, exclusive selectivity, no nonspecific binding, proteolytic resistance, cross-species activity, and thermostability. Furthermore, the optimized mAb demonstrated remarkably improved pharmacokinetics and significantly prolonged therapeutic effects in a mouse model of DPN. Conclusions Given the critical role of developability in mAb therapeutics, this study highlights a practical strategy for improving both developability and in vivo efficacy through rational site-directed mutagenesis.
