Utilizing structural biology tools by neurosnap for affinity maturation of single domain antibodies (VHH) targeting immune checkpoints

Abstract: The field of antibody-based therapeutics has consistently sought advancements in methodologies for candidate discovery and engineering. Historically,the discovery and development of biologics heavily relied on laboratory-based
approaches; however, computational informatics methods have recently started exerting a notable influence. Specifically, deep learning, a subset of machine learning, has gained significant traction in biomedical research. Recent strides in big data mining through next-generation sequencing have not only revolutionized the landscape of therapeutic antibody discovery but have also yielded copious antibody repertoire sequencing data, thereby offering novel avenues for employing deep learning methodologies. This study introduces a novel model designed to simulate the binding process between multi-specific ligands and membrane receptors on cell surfaces. The investigation involved the generation of complementarity-determining region 3 (CDR3) mutated libraries comprising 32 sequences of FDA-approved anti-PD1 antibodies sourced from the TABS database. Leveraging artificial intelligence-based software tools such as Protein MPNN, ESM-1F1, and MIF-ST developed by Neurosnap.ai, we employed variant generation and conducted modeling in conjunction with molecular docking and molecular dynamics (MD) simulations to discover and redesign single-domain antibodies (VHH) against the same target with enhanced affinity and specificity. Strategically, CDRs targeting PD1 were integrated into structurally compatible antibody scaffolds. Subsequently, three single-domain antibodies (VHHs) targeting Programmed Death 1 (PD1/PDL1) antigens were designed and subjected to in silico testing. Computational characterization revealed the high stability of all designs, showcasing sub-nanomolar binding affinities (KDs) with PD1/PDL1. This methodology holds promise in facilitating the discovery of stable VHHs exhibiting sub-nanomolar KDs without the need for in vitro affinity maturation.

Keywords: Immune Checkpoint Inhibitor, single domain antibodies (VHH), homology
modeling, molecular docking, MD simulation