Harnessing the molecular Trojan horse : Evaluating properties of preclinical Aβ immunoPET radioligands for optimized brain delivery via the transferrin receptor

Abstract: With high specificity and selectivity to targets, antibodies are prime candidates for positron emission tomography (PET) radioligands. They do not passively cross the blood-brain barrier which has hindered their development for imaging intrabrain targets, like amyloid-β (Aβ) in Alzheimer’s disease. The molecular Trojan horse strategy with antibodies that bind to both the transferrin receptor (TfR) and an intrabrain target improves brain delivery of therapeutic antibodies. However, therapeutic antibodies are typically dosed substantially higher than antibody-based PET (immunoPET) radioligands.This thesis evaluated the effects of affinity, valency, and dose on the brain delivery of preclinical Aβ immunoPET radioligands via the TfR.Paper I investigated whether immunoPET with TfR-mediated brain delivery could image Aβ with similar sensitivity in rats as it has in mice. To our knowledge, this was the first time TfR-hijacking to deliver a radioligand to image Aβ was successfully demonstrated in rats; suggesting this strategy could eventually be translated to clinics.Affinity to TfR influences therapeutic delivery to the brain. In Paper II, we compared four Biacore setups and one on-cell assay for determining apparent affinities to the TfR. Absolute affinity determination was challenging since several assay conditions impacted the kinetic parameters. A directional TfR capture in Biacore may be optimal since it determined kinetic parameters while mimicking in vivo receptor conditions. Papers I and III investigated how antibody affinity affects brain delivery at tracer doses and indicated that stronger TfR affinity yielded higher brain delivery. The antibodies in Paper III lacked effector function. The resulting pharmacokinetic profiles in Aβ pathology-presenting mice indicated this may have improved target accumulation of the immunoPET radioligand.In Paper IV, we screened a novel library of monovalent and bivalent affinity variants of the anti-mouse TfR antibody, 8D3. A pair of monovalent and bivalent antibodies with an apparent affinity of 10 nM was identified and evaluated in vivo. Monovalent binding yielded higher brain uptake at a tracer dose but whether bivalent binding steered the antibody towards lysosomal degradation was unclear.In conclusion, monovalency, high affinity binding, and ablated effector function are likely beneficial properties for TfR-mediated brain delivery of an immunoPET radioligand at a tracer dose.