ABSTRACT Enhancing antibody-dependent cellular cytotoxicity (ADCC) via N-glycan afucosylation of Asn-297 is a validated strategy to improve the clinical efficacy of therapeutic antibodies. However, the impact of distinct glycoengineering approaches on the function of antibodies has not been systematically elucidated. Here, we experimentally compared two type II anti-CD20 antibodies, Obinutuzumab (Gazyva®) and Obinutuzumab beta (MIL62, Bejescin®), which share identical amino acid sequences but exhibit divergent glycosylation profiles. MIL62 was engineered with complete core afucosylation (fucose < 0.1%) lacking bisecting N-acetylglucosamine (GlcNAc) via knockout of the GDP-fucose transporter (GFT) in Chinese hamster ovary (CHO) cells used to produce the antibody. Conversely, Gazyva was produced in CHO cells that overexpress β-1,4-N-acetylglucosaminyltransferase III (GnT-III) and α-mannosidase II (α-ManII), resulting in ~50% fucose content and >80% bisecting GlcNAc occupancy. These distinct glycoengineering strategies led to disparate functional outcomes: afucosylated MIL62 showed improved FcγRIIIA binding and ADCC potency, while Gazyva with bisecting GlcNAc modifications exhibited higher glycoform heterogeneity and reduced thermal stability. Both antibodies displayed comparable FcRn binding, mannosylation, sialylation, and murine pharmacodynamics, mediating complete depletion of B cells in blood, lymph nodes, and spleen. Upon antigen rechallenge, MIL62 suppressed specific antibody titers, indicating memory B-cell eradication and profound potential to prevent autoimmune relapse. This study demonstrates that distinct glycoengineering strategies fundamentally reshape the antibody glycan profile beyond merely reducing fucose. These structural differences not only fine-tune ADCC potency, but also impact antibody stability. Particularly, complete afucosylation of MIL62 led to optimized ADCC potency and effectively suppressed pathogenic B cells repopulation in a delta-like ligand 3 (DLL3) re-challenge model, providing a critical framework for designing next-generation antibodies with superior therapeutic efficacy.

ABSTRACT The complementarity-determining regions (CDRs) of antibodies are loop structures that are key to their interactions with antigens and are of high importance to the design of novel biologics. Existing approaches for characterizing the diversity of CDRs have limited coverage and cannot be readily incorporated into protein foundation models. Here we introduce ImmunoGlobulin LOOp Tokenizer, Igloo, a multimodal antibody loop tokenizer that encodes backbone dihedral angles and sequence. Igloo is trained using a contrastive learning objective to map loops with similar backbone dihedral angles closer together in latent space. Compared to state-of-the-art protein encoding approaches, Igloo can efficiently retrieve the closest matching loop structures from a structural antibody database, outperforming the existing methods on identifying similar H3 loops by 6.1%. Igloo assigns tokens to all loops, addressing the limited coverage issue of canonical clusters, while retaining the ability to recover canonical loop conformations. To demonstrate the versatility of Igloo tokens, we show that they can be incorporated into protein language models with IglooLM and IglooALM. On predicting binding affinity of heavy chain variants, IglooLM outperforms the base protein language model on 8 out of 10 antibody-antigen targets. Additionally, it is on par with existing state-of-the-art sequence-based and multimodal protein language models, performing comparably to models with 7 × more parameters. IglooALM samples antibody loops which are diverse in sequence and more consistent in structure than state-of-the-art antibody inverse folding models. We show that Igloo can rapidly and scalably prioritize functional antibody variants from large mutagenesis libraries, achieving a 1.9 × enrichment of experimentally validated HER2 binders in a zero-shot setting. Igloo demonstrates the benefit of introducing multimodal tokens for antibody loops for encoding their diverse landscape, improving protein foundation models, and for antibody CDR design.

ABSTRACT The use of animal models that can reliably predict drug performance in human patients is critical to antibody therapeutic development. Along with assessing toxicity and efficacy, determining the pharmacokinetic (PK) properties of therapeutics in Tg32 and Tg276 mice is essential to preclinical characterization. While Tg32 mice have been well established as indispensable in their ability to model the PK properties of antibody therapeutics, their intact immunity leaves them capable of mounting anti-drug antibody responses that interfere with PK interpretation. Here, we demonstrate the negative impact anti-drug responses can have on PK parameters derived from Tg32 mice, and provide strong evidence to support the use of immunodeficient Tg32 SCID mice as an equivalent means to model human PK. In addition, we investigate one possible cause for reduced FcRn function observed in Tg276 mice when compared to Tg32 mice in spite of evidence that their FcRn protein levels are actually higher. We also introduce NSG Tg32 mice and our attempts to block off-target binding of human IgGs to their high-affinity Fc gamma receptors, which failed to recover FcRn function similar to that observed from Tg32 mouse controls, dramatically limiting their utility for PK analysis. Taken together, our results provide a comparison of these preclinical animal models, so they can be used to improve human PK predictions of antibody therapeutic candidates in development.

ABSTRACT Controlling glycosylation, a critical quality attribute of biopharmaceuticals such as monoclonal antibodies, is essential, as it significantly influences biological activity and therapeutic efficacy. Although numerous studies have examined the impact of process parameters (PP, e.g. temperature, pH, dissolved oxygen) on glycosylation, the lack of standardized reporting makes cross-study comparisons challenging and prevents clear conclusions. Here, we systematically reviewed the literature and applied a normalized quantitative framework, the Glycan Indices approach, as a standardized quantitative criterion to evaluate the impact of process parameters on glycoform distribution in IgG-producing CHO cell systems objectively. This methodology enabled the integration and reinterpretation of large, heterogeneous datasets, validating some well-known patterns while providing novel perspectives about process parameters. Our analysis revealed that PP manipulations of pH, dissolved oxygen or CO2 partial pressure rarely resulted in meaningful shifts in glycosylation, with changes <5% observed for galactose, fucose, or N-acetylneuraminic acid content. In contrast, for several cases temperature and osmolality changes notably affected galactosylation (>10%) and fucosylation (1–10%), variations that may have significant biological consequences. To our knowledge, this is the first comprehensive quantitative assessment of process parameters effects on glycosylation, showing that such influences are consistently limited, independent of CHO cell line or culture mode. Based in our observations we strongly recommend reporting both glycan distribution and glycan indices when performing glycan analysis. Dual reporting facilitates inter-study comparisons and prevents subtle shifts in sugar moieties from being masked by glycan redistribution.

ABSTRACT Claudin-6 (CLDN6) is an oncofetal tight junction protein with minimal to no expression in healthy adult tissues but aberrant upregulation in ovarian malignancies, making it an attractive tumor-selective antigen for T cell-based immunotherapy. The development of CLDN6-targeting antibodies, however, has been challenged by its high homology to CLDN9, which is expressed in normal tissues and differs by only three amino acids within the extracellular domains. Here, we describe the discovery and preclinical development of ARC101, a bispecific CLDN6×CD3 antibody featuring a naturally derived, highly potent CLDN6 binder with no cross-reactivity to CLDN9 or other human membrane proteins. The stringent specificity of ARC101 eliminates off-target binding and distinguishes it from other CLDN6-targeting antibodies in development. The effector arm of ARC101 incorporates a novel conformational CD3 binder, enabling potent T cell-mediated cytotoxicity against CLDN6-expressing tumor cells in vitro and in vivo. ARC101 also demonstrated a favorable pharmacokinetic profile in cynomolgus monkeys, low immunostimulatory responses in ex vivo human donor assays, and robust biophysical properties compatible with standard antibody manufacturing. Collectively, these findings support the clinical advancement of ARC101 as a differentiated, CLDN6-specific bispecific immunotherapy with exceptional tumor selectivity and optimized T cell activity for the treatment of solid tumors.

ABSTRACT Studying the interactions between antibodies and antigens is fundamental to the development of novel therapeutic biologics. Predictions of such interactions start with data collection. Though there exist reliable resources to identify antibody structures in the Protein Data Bank (PDB), such data still requires substantial processing to be usable in predictive tasks. Redundancy in sequences needs to be removed to avoid data leakages between train, test, and validation sets. Descriptors such as surface accessibility, secondary structure, and antibody region information need to be additionally annotated. Information on inter- and intra-molecular contacts, which is crucial to studying paratope/epitope information, needs to be collected. The specialized immunoglobulin format of Nanobodies® requires a separate dataset mirroring that of antibodies, given that their structure contains only a single VHH chain. Because antibody–antigen structures account for a small amount of all protein–protein contacts, having a molecular contact reference from other proteins is also desired. To address these issues, we introduce NAStructuralDB (https://naturalantibody.com/na-structural/), a dataset of processed structures of antibodies, Nanobodies®, proteins, and their complexes with molecular contact information and associated annotations. We use the opportunity of having collected the contact data to provide a reference of binding propensities of different residues across distinct contact types.

ABSTRACT The Ginkgo Datapoints Antibody Developability (AbDev) Competition, a blinded benchmark for developability prediction characterized entirely on a single, industrial-scale experimental platform, was conducted from September 8 to November 18, 2025. We benchmarked predictors across five biophysical properties – hydrophobicity, thermostability, self-association, expression titer, and polyreactivity – using a public training set of 246 clinical antibodies and a blinded, held-out test set of 80 antibodies. We received submissions from 113 teams spanning 25 countries, 38 companies, and 39 universities. Winning submissions differed by assay. Top Spearman’s ρ values on the test set reached 0.708 (hydrophobicity), 0.392 (thermostability), 0.356 (polyreactivity), 0.337 (self-association), and 0.310 (titer). Cross-validation scores from the public training set consistently exceeded held-out test performance, indicating overfitting and limited out-of-distribution generalization. Together, these results provide a standardized snapshot of current antibody developability modeling capabilities and highlight a key bottleneck: available datasets are too small and heterogeneous to support robust, assay-spanning prediction. Meaningful progress will require larger, standardized, and diverse experimental datasets – with harmonized protocols and rich metadata – to train and validate models that generalize reliably for future antibody discovery campaigns.

ABSTRACT The development of computational models addressing therapeutic antibodies faces significant challenges. Particularly, the prediction of binding affinity across a diverse set of measurements, due to the scarcity of data. A critical data element is the set of antibody–antigen interaction pairs associated with sequences. To address this issue, we developed the Antigen Specific Antibody Database (ASD, https://naturalantibody.com/agab/), a database aggregating antibody-antigen interaction data from multiple studies with standardized formatting and annotations. Our dataset compilation strategy resulted in data from 15 distinct sources, resulting in 1,097,946 unique antibody–antigen interactions (with 9575 unique antigens). The ASD captures diverse affinity measures and qualitative binding assessment, along with metadata including UniProt and PDB identifiers, target protein names, confidence levels, and experimental conditions such as type of measured affinity, source organism, and germline genes. Through this integration drive, we make available an ample resource of interaction data gathered from the public domain to act as a foundation for model development and further data generation.

ABSTRACT Antibody therapeutics are a rapidly growing class of biopharmaceuticals, but concerns regarding potential developability issues persist. While complementarity-determining region (CDR) loops are imperative for antigen specificity and mutations are challenging, the framework regions can be exchanged to align with developability attributes such as aggregation, clearance, and viscosity, all governed by physicochemical characteristics. In this study, we systematically analyze the electrostatic and hydrophobic surface properties of germline-encoded antibody frameworks to assess their role in modulating Fv developability. Using structure prediction and surface patch analysis, we identify differences between kappa and lambda light-chain frameworks, characterize outlier germlines with extreme surface properties, and demonstrate using hydrophobic interaction chromatography and a heparin column that framework selection can compensate for CDR loop physicochemical characteristics. Our findings reveal that rational framework selection can serve as a systematic tool for optimizing antibody developability. This study provides a toolbox for antibody design, enhancing therapeutic candidate selection by leveraging inherent germline properties.

ABSTRACT Driven by the substantial limitations of first generation 4–1BB agonists urelumab and utomilumab, the field has shifted toward engineering next-generation molecules with improved therapeutic windows. This review provides a comprehensive analysis of this evolution, detailing how key molecular design strategies are used to restrict 4–1BB activation to the tumor microenvironment. We summarize available clinical data, highlighting that 4–1BB bispecific antibodies exhibit superior antitumor efficacy and more favorable safety profiles compared with their monospecific predecessors. Furthermore, we discuss strong rationale for combination strategies, emphasizing how 4–1BB signaling provides the crucial costimulatory signal necessary to sustain durable anti-tumor responses. In summary, this review elucidates the scientific basis of antibody engineering aimed at improving safety and tumor-selective activation of 4–1BB agonists and outlines future directions for optimizing their clinical application in cancer immunotherapy.