To characterize the blood-brain barrier-crossing properties of 1,000 antibodies using a novel high-throughput, multiplexed in vivo screening platform.

Targeting the central nervous system with protein therapeutics has long been hindered by the restrictive properties of the blood-brain barrier (BBB). Recent advances in antibody-based therapies have shown promise in traversing the BBB via target-mediated transcytosis. However, the full potential of this approach remains underexplored due to limitations in predicting in vivo transcytotic capabilities and the scope of in vivo screening methods for drug discovery.

Our findings reveal a diverse array of phenotypic characteristics and pharmacokinetic profiles among these antibodies, with some demonstrating exceptional brain biodistribution. We find examples of antibody targets and formats affecting shuttle persistence and distribution within the brain parenchyma.  Machine learning analysis of this extensive dataset enabled further optimization of these antibodies, enhancing their brain uptake and residence time. Validation in both murine and primate models confirmed pharmacokinetics of selected antibody candidates.

This work not only expands the pool of potential molecules for effective BBB transcytosis but also represents a significant paradigm shift in utilizing in vivo models during the early stages of drug discovery.