Design, expression, and analysis of antibody-based blood-brain barrier shuttles

Abstract: Antibody therapeutics, with their strong and highly selective target binding, are now used to treat various diseases. However, to enable their use to treat brain disorders, they must be delivered across the blood-brain barrier (BBB), as without active transport, only around 0.01% of intravenously injected doses reach the brain. Brain delivery can be done by BBB shuttles capable of binding receptors that naturally transport proteins, e.g., the Transferrin receptor (TfR). This thesis has studied strategies for designing TfR-binding shuttles and how to enhance the protein expression of antibody therapeutics. In Paper I, we shared our updated transient gene expression (TGE) protocol and developed a small-scale version to surmount the cost limitations of testing many conditions. Large variations of protein expression were observed for both protocols, prompting future studies investigating its cause(s). In paper II, we investigated if binding to the glycosaminoglycan heparan sulfate (HS) present at the BBB could improve brain delivery. Our results indicate that the BBB shuttle scFv8D3 is not dependent on the HS-binding sites identified, and adding new HS-binding sites did not enhance delivery. However, further studies are required due to HS's complexity and heterogeneity. Decreasing the TfR affinity of BBB shuttles has been shown to boost the delivery of therapeutic doses of high affinity anti-TfR antibodies, e.g., bivalent 8D3 antibodies. In Paper III, we applied the strategy to a monovalent single-chain fragment variable (scFv) of 8D3 (scFv8D3) based BBB shuttle. Our affinity mutants exhibited lowered TfR affinity, longer blood half-life, and higher brain concentration. Using our In-Cell BBB Trans assay, we concluded that the increased brain concentration is likely due to extended blood half-life. In paper IV, we fused the TfR ligand holo-transferrin to the TfR binding arms of the partly bivalent RmAb158-scFv8D3 antibody. Our results indicate that the TfR binding shifted from partly to fully bivalent, resulting in markedly decreased in vitro transcytosis. The potential transcytosis-promoting effect of the fused holoTf was absent and/or counteracted by the bivalent binding of the design. However, the strategy may still prove useful for monovalent TfR binders. In conclusion, monovalent and low-to-moderate affinity are likely beneficial binding properties for TfR-mediated brain delivery at therapeutic doses. However, whether it is possible to enhance brain delivery with HS-binding or holoTf-fusion requires further study.