Soluble TCR engagers enable high-affinity targeting of disease-relevant antigens and off-the-shelf use, thus representing a promising therapeutic modality with applications in oncology, autoimmunity, and infectious disease. Despite their therapeutic potential, affinity maturation of soluble TCRs, which typically requires a 1 million-fold improvement, is complicated by specificity challenges. Here we describe machine learning–guided protein engineering as an effective approach to rapidly identify picomolar affinity TCRs with high levels of specificity.

Etcembly is the first company to leverage generative AI to discover and engineer a T cell receptor (TCR) to picomolar affinity. We formatted our lead PRAME targeting molecule, ETC-101, into a trispecific T cell engager and demonstrated that ETC-101 specifically redirected T cell killing of PRAME-positive cancer cells only, while demonstrating a promising safety profile with no detectable off-target effects. Our data highlights the efficacy of ETC-101 as a novel drug candidate for the treatment of a wide range of PRAME-positive malignancies.

LabGenius Therapeutics’ platform leverages avidity-driven selectivity to overcome common T-cell engager (TCE) challenges, including on-target, off-tumour toxicity. In this talk, we describe how the closed-loop integration of high-throughput experimentation with machine learning has facilitated the discovery and optimisation of multispecifics for function and developability. Specifically, how we’ve developed a pipeline of VHH-based TCEs that exhibit on/off killing selectivity for TAA targets with minimal expression differences.

Genotype-specific therapeutics of leukaemia require targeting hard to drug molecules like transcription factors or mutant RAS. Intracellular antibodies are starting points as inhibitors blocking protein-protein interactions and can be engineered with effector functions such E3 ligases to create biodegraders. The antibody fragment binding site (the paratope) can also be used to screen for small molecule surrogates that can form the basis of conventional drug development. These approaches will be discussed targeting transcription factors and mutant RAS.

The pathogenicity of autoreactive antibodies has been demonstrated for many autoimmune diseases and the isotype/subclass profile can potentially influence the disease pathophysiology. Although often overlooked, IgA autoantibodies are increasingly recognized in different autoimmune indications. Here, we describe the development of anti-IgA monoclonal antibodies that can actively remove IgA from the circulation and block binding of IgA to its main Fc receptor FcαRI. Given the abundancy of IgA in human serum (1-3 mg/mL), both Fab and Fc engineering were optimised to design a monoclonal antibody with the desired properties.

Overactivation of Fα?RI by immune complexes (ICs) has been implicated in various autoimmune disorders and neuropathy. To date, there are no effective FcαRI-specific blocking antibodies available. Here we report two first-in-class anti-FcαRI antibodies, with high affinity Fab-mediated binding within the IgG binding site on extracellular domain 2 of FcαRI. They effectively block IgG and IC binding in models for ITP and RA, and displace pre-bound ICs without activation of the receptor.

Complement is a powerful part of innate immunity, yet conventional antibody therapeutics exploit it only incompletely. Complement activation typically demands high antigen density and is structurally constrained, limiting efficacy against many clinically relevant targets. Commit has developed a bispecific complement engager (BiCE) platform that through C1q recruitment effectively activates complement in a target-dependent manner. Data will be presented demonstrating preclinical efficacy, safety, and broad applicability of the BiCE platform.

In many autoimmune diseases IgG autoantibodies plays a crucial role by mistakenly targeting the body's own tissues. Different therapeutic approaches with focus on IgG reduction have emerged. Highly potent, Hansa's IgG cleaving enzymes (imlifidase and HNSA-5487) offer potential of altering disease progression and improving patient’s outcome. HNSA-5487 is in silico optimised, highly efficacious second-generation IgG cleaving enzyme with reduced immunogenicity. Both molecules are currently under clinical evaluation in several disease areas and indications. Imlifidase has already shown promising results in Guillain-Barré Syndrome (GBS) and anti-glomerular basement membrane disease (anti-GBM) and a phase III trial is ongoing in anti-GBM. With HNSA-5487 we are focusing on indications in the neuro-autoimmune space, including Myasthenia gravis.

Brain-shuttles can deliver oligonucleotides to the brain. But conjugated oligonucleotides with modifications can affect biophysical properties and functionality of brain shuttles. These effects can be addressed by engineering complexes of antibody-based shuttles that harbour additional binders to cloak the ASO payload. Such entities show improved TfR-specific transcytosis in cellular blood-brain barrier models and improved PK and brain delivery in animal models.