主要会議1日目 2026年10月20日（火） - 日本時間（GMT+09:00)

• Switch-Ig technology designed to exploit the unique features of the tumour microenvironment
• An antibody engineering platform integrating AI and laboratory automation
• A first-in-class multispecific antibody for celiac disease targeting the gluten peptide/HLA complex

• Emerging systems for rapid immunization and redesign that enable the generation of clinical-grade human antibodies in under 90 days.
• Utilizing TC-mAb mice for rapid high-titre immunization and 60-day redesign cycles using CHO-based mammalian display
• Identifying antibodies against non-immunized viral variants.

Peptide-HLA (pHLA)-targeting therapeutics can potently and specifically target otherwise inaccessible intracellular antigens but require both high affinity and high specificity for clinical use. To address these challenges, we developed a high-throughput yeast-based platform to rapidly discover and engineer highly selective high-affinity soluble T cell receptors (TCRs)- and TCR mimetic (TCRm) antibodies, accelerating the development of potent and specific pHLA-targeting therapeutics.

BsAbs offer strong therapeutic potential, but complex formats can introduce developability risks, such as aggregation, non-specific binding, and instability. This session introduces early-stage developability assessment that evaluates PK (self-association, non-specific binding, FcRn affinity, and serum stability), biophysical properties, stability, and immunogenicity, highlighting how modality-driven assessment can guide lead optimization, reduce downstream risks, and accelerate bsAb R&D.

AI is transforming antibody discovery, but its impact depends on integrating high-quality data into routine workflows. Here we present an AI-native framework unifying experimental, processed, and in silico antibody data. The Genedata platform enables curated datasets, traceable insights, and agentic analysis, accelerating discovery of bispecifics, multispecifics, and ADCs while improving robustness, speed, and decision-making across antibody pipelines.

• EMLy™ as a unique, fine-tuned AI platform that integrates deep learning with sequence-structure intelligence
• Leveraging EMLy™ for the discovery of next-generation therapeutics ready for successful advancement into preclinical and clinical development, through optimized binding, stability, and manufacturability across diverse therapeutic formats.

Monoclonal antibodies have become indispensable in today's society, operating in diverse fields including treatment and diagnosis. While classical methods such as mice hybridoma and phage display often provide sufficient performance against easily identifiable target antigens, there is a growing demand for high-performance monoclonal antibodies with higher affinity and specificity. We have established an antibody generation technology employing single-cell and cell-free antibody expression methods, successfully generating high-performance antibodies against various targets from human and immunized animal samples.

In this technology, single memory B cells specific to the target antigen are selected by cell sorter, and the monoclonal antibody gene is obtained by single-cell RT-PCR. The obtained antibody gene is used as a template for cell-free coupled transcription/translation reactions without being incorporated into a plasmid. Two patented technologies-the SKIK tag, which increases expression levels, and the LZ tag, which enhances binding ability by promoting H- and L-chain association-enable efficient in vitro production of Fab proteins.

To date, we have administered antigens to rabbits, which are known to produce antibodies with high affinity and specificity in their bodies, and used our technology to create antibodies specific to various challenging antigens from the lymphocytes of immune individuals. For example, our cAMP-specific antibody has the specificity to strictly distinguish between cGMP and ATP and cAMP. As antibodies that can distinguish even slight amino acid sequences, we have created antibodies specific to activated GIP, one of the incretins, and antibodies specific to antibody drugs (anti-idiotype antibodies), which are actually being used as diagnostic reagents and pharmacokinetic analysis tools.

Furthermore, in obtaining antibodies that actually function within the human body, we identify broad-spectrum neutralizing antibodies from samples provided by individuals with a history of infection with pathogenic viruses, including COVID-19, and identify cancer-specific antibodies from antibody-producing cells infiltrating solid tumors.

Recent developments in targeted protein degradation have provided great opportunities to eliminate extracellular protein targets using potential therapies with unique mechanisms of action and pharmacology. Here we have developed a novel site-specific antibody conjugation approach to generate antibody mannose 6-phosphate (M6P) conjugates. The method uses a high affinity synthetic M6P glycan, bisM6P, that is coupled to an Fc-engineered antibody NNAS. This mutant was generated by switching the native glycosylation site from position 297 to 298 converting non-sialylated structures to highly sialylated N-glycans. The sialic acid of the glycans attached to Asn298 in the engineered antibody was selectively conjugated to bisM6P without chemoenzymatic modification, which is often used for site-specific antibody conjugation through glycans. The M6P-conjugated antibody against a protein of interest (POI) efficiently internalized and degraded targeted soluble proteins, such as human tumor necrosis factor (TNF) through the endo-lysosomal pathway as demonstrated by confocal microscopy and flow cytometry. Our platform provides a unique method for producing biologics-based degraders that may be used to treat diseases through event-driven pharmacology, thereby addressing unmet medical needs.

Chiome Bioscience has developed DoppeLib, a recently established bispecific antibody platform for the discovery and engineering of next-generation antibody therapeutics. Building on its experience in antibody generation, functional screening, and affinity maturation, Chiome continues to expand its antibody discovery capabilities. This presentation will focus on how DoppeLib can address key challenges in bispecific antibody discovery, including antibody-arm pairing, molecular design, functional screening, and downstream optimization.

DoppeLib is designed to enable systematic generation, screening, and optimization of bispecific antibody candidates by leveraging the DT40 cell-based platform that Chiome has developed and refined over the years. By supporting flexible antibody pairing and molecular design, the platform may expand opportunities to create differentiated therapeutic antibodies beyond conventional bispecific approaches. Representative use cases will include biparatopic antibody concepts with a particular focus on functional activity.

The presentation will also briefly discuss how DoppeLib can be integrated with Chiome’s broader antibody generation, functional screening, affinity maturation, and Integrated Drug Discovery capabilities to support the progression from early screening to candidate selection.

Ion channels represent pivotal devices capable to modulate different cancer hallmarks. Among ion channels which are relevant in cancer, the “human ether à-go-go-related gene” potassium channel hERG1 has been deeply studied by our group (Arcangeli A et al., Expert Opin Ther Targets. 2024. doi: 10.1080/14728222.2024.2318449).

Based on our studies, our group has developed and patented (WO2019/015936) a single chain diabody that targets and harnesses the hERG1/β1 integrin (HB) complex which is selectively expressed in several types of cancers: scDb-HB.

Based on this backbone, we have developed novel multispecific antibodies targeting e.g. the receptors for the Tumor Necrosis Factor-Related Apoptosis-Inducing Ligand (TRAIL) or voltage dependent sodium channels, like the naV 1.5.

The efficacy of such novel multispecific antibodies for treatment of Triple Negative Breast Cancer (TNBCa) has benn proven both in vitro and in vivo, in preclinical mouse models.

Traditional antibody optimization pipelines optimize affinity and developability in iterative workflows that are inherently slow, recursive, and unpredictable. We combine large-scale, quantitative affinity measurements with a fine-tuned optimization model to accurately predict >20 mutations in a parental antibody. This integrated wet-lab/dry-lab platform enables simultaneous optimization of affinity, specificity, cross-reactivity, and developability, delivering optimized therapeutic leads in less than 8 weeks.

We have investigated for logical design of biparatopic antibodies with desired mechanisms of action. We focused on how stoichiometry of immunocomplex affects the biological behavior of biparatopic antibodies. Herein, I present our recent observations from basic research of how the control of immunocomplex affected signal transduction and pharmacokinetics.

• A clinical showcase of global Innovation:
• Multispecifics
• Next-gen ADCs.
• Pipeline Technical Breakthroughs
• Clinical Progress