Study of a Transferrin Receptor Brain Shuttle for the Delivery of Nanobodies Against GABAA Receptors

Abstract

The blood-brain barrier (BBB) represents a major obstacle in drug development, restricting the access of drugs into the central nervous system (CNS) following systemic administration. This challenge is particularly evident for biological drug classes, such as antibody-based therapies, contributing to their limited application in the treatment of neurological disorders. Technologies have emerged to improve drug delivery by targeting receptors present at the BBB, such as transferrin receptor (TfR), hijacking endogenous trafficking systems to shuttle conjugated therapeutic cargo from the blood circulation into the brain parenchyma. Such systems could be utilised to improve the delivery of novel immunogenically-raised nanobodies (Nbs), which exhibit promising subtype-selectivity and pharmacological activity against neuroinhibitory γ-aminobutyric acid type A receptors (GABAARs). Initially, anti-GABAAR Nbs were genetically fused to the BBB shuttle, 8D3, and the impact on the Nb binding properties to GABAARs was investigated. This was to establish suitable formats for CNS delivery in the future. In addition, a range of unconjugated 8D3 formats were investigated for TfR binding properties. Using mammalian expression systems, antibody-based molecules were generated in a range of formats, purified, and then characterised. This included measuring target affinities for the BBB receptor and GABAAR in cell-based assays. For bispecific 8D3-Nb molecules, retention of Nb binding for GABAAR was orientation specific, although all formats retained binding affinities for TfR in the nanomolar range. Affinity for TfR was similar across unconjugated 8D3 formats, regardless of valency, and was reduced by targeted mutations of the complementarity-determining regions (CDRs). TfR expression at the cell surface was also investigated to better understand the mechanistic effect of 8D3 valency, indicating drug-induced changes were specific to the TfR-expressing cell line and may depend on receptor density. The brain uptake of 8D3 molecules was measured by establishing an in vivo workflow, revealing a clear TfR-specific uptake for mono- and bispecific 8D3 formats following their systemic administration in mice. To enhance brain penetration, three TfR binding properties were explored that were hypothesised to impact BBB trafficking: affinity, valency, and pH-sensitivity. Changes to mTfR affinity significantly affected the level of brain accumulation for monovalent 8D3 molecules, whereas the effects of affinity were less pronounced for bivalent 8D3 formats. Overall, monovalent reformatting of the wild-type, high affinity, version of 8D3 proved the most effective format for delivery into the brain. Novel 8D3 variants, exhibiting pH-sensitive binding to TfR, were investigated for the first time. Molecules with preferential binding at endosomal pH reduced brain uptake, revealing that pH-sensitivity is likely to influence transport at the BBB too. These studies represent a unique and comprehensive comparison between TfR binding properties that are regularly pursued in the engineering of active BBB shuttling moieties. The findings suggest that high affinity monovalent engagement of TfR, with preferential binding at physiological pH, is likely to be most effective mechanism to transport BBB-impermeable drugs to their therapeutic targets in the brain. This work provides an exciting foundation for the further development and characterisation of subtype-selective GABAAR protein modulators, potentially unlocking their therapeutic potential for a range of CNS disorders.