Surface Display-Enabled Miniprotein Evolution for Brain-Targeted Therapeutics

Abstract

Protein engineering is increasingly central to drug discovery and design. Advances in computational and high-throughput methods have enabled precise tailoring of protein structure, function, and binding affinity to create biologics with improved or novel functions. In one example, engineered tissue binding domains have enabled the brain penetration of peripherally-injected antibodies by facilitating passage through tightly regulated endothelial barriers. This can also reduce off-target effects by spatially restricting activity to the target tissue. Many protein domains can serve as scaffolds for these binders, including the scFv, FN3, and various miniproteins. The known structures of such scaffolds enable rapid development of binders against diverse targets - increasingly enabling modular, targeted therapeutics. We designed and executed a directed evolution campaign to discover novel binders enabling receptor-mediated antibody transport into brain tissue. We selected a simple, well-characterized de novo miniprotein scaffold to avoid off-target functional activities, computationally analyzing it to select residues for directed evolution via Kunkel mutagenesis. Surface display-based screening with rounds of increasing selection pressure were used to isolate a focused pool of variants. These mutants were evaluated for binding, identifying variants with enriched target binding activities for further study. Our results demonstrate a viable pipeline for miniprotein-based binder discovery that is easily generalizable across multiple targets, with clear pathways for further optimization of the pipeline itself and the incorporation of selected candidates into novel drugs.