The effects of dose, valency, and affinity on TfR-mediated brain delivery in vivo.

Abstract

Background: Monovalent binding to the transferrin receptor (TfR) is considered the most efficient mode for high delivery of protein constructs across the blood-brain barrier via TfR-mediated transcytosis at therapeutic doses. However, growing evidence suggests this is not the case at lower, diagnostic doses. There is also a lack of data on how valency and affinity to TfR affect brain uptake independently since previous studies have not compared monovalent and bivalent antibodies with similar affinities regardless of valency (i.e. apparent affinity). Therefore, the aim was to evaluate the independent effects of valency and affinity on TfR-mediated brain delivery at different doses.

Methods: Affinity variants of antibody 8D3 were produced by introducing alanine point mutations into the complementarity-determining regions. Eleven Fab fragments and 29 IgGs were affinity screened against mouse TfR (mTfR). Six of each were chosen for production with a knob-into-hole design to have monovalent and bivalent TfR binders in full-length antibody format. The apparent affinity of these 12 antibodies were tested in an Sp2/0-Ag14 cell assay. The 10 nM apparent affinity set and the bivalent wild-type antibody were radiolabelled and injected into wild-type mice at a low (0.22 ± 0.03 mg/kg) or high (7.5 ± 0.43 mg/kg) dose. The biodistribution was measured in brain, blood and peripheral organs 4 h post-injection.

Results: Two sets of monovalent and bivalent 8D3 formats with similar mTfR apparent affinities were identified. Brain and tissue uptake was higher at the low dose than the high dose for all antibodies. At the low dose, the higher apparent affinity, bivalent antibody had higher brain uptake than either of the two lower apparent affinity antibodies. At the high dose, the monovalent antibody had higher brain uptake than the two bivalent antibodies. The peripheral distribution of the three antibodies were similar to the brain distribution at both doses.

Conclusions: Valency and apparent affinity affect brain uptake in a dose-dependent manner such that: brain uptake was affected more by apparent affinity at the low dose and by valency at the high dose. Thus, when designing constructs for TfR-mediated brain delivery, the application, and consequently the dose, are critical to consider.