New prospects for Zr-immuno-PET in brain applications - Alpha-synucleinopathies.

BACKGROUND

Recently, Zr-immuno-PET imaging of therapeutic antibodies, actively transported over the blood-brain-barrier via transferrin-mediated transcytosis, was demonstrated using the chelator DFO*. In these studies, aducanumab targeting amyloid-beta was fused with a transferrin binding unit: a single-chain Fab fragment derived from 8D3 (scFab8D3). Alpha-synuclein is a hallmark protein of several neurodegenerative diseases such as Parkinson's Disease, Lewy-Body-Dementia, and Multiple System Atrophy. Zr-immuno-PET imaging of alpha-synuclein can be a valuable tool for image-guided drug development and assessment of target engagement. The goal of this study was to compare two currently used constructs of 8D3 for targeting potential, namely a single moiety of scFab8D3 fused to the alpha-synuclein antibody HLu-3 (HLu-3-scFab8D3) versus HLu-3 fused with two 8D3 single-chain variable fragments (HLu-3-(scFv8D3)), by Zr-immuno-PET in an alpha-synuclein pre-formed fibril (PFF) deposition model. HLu-3 and the HIV-targeting B12-scFab8D3 were used as controls. The best-performing compound was further investigated in an animal model with predominantly intraneural target aggregation.

METHODS

Antibodies were conjugated with DFO* using DFO*-NCS and subsequently radiolabeled with Zr. Assessment of binding affinity was done by alpha-synuclein ELISA and with FACS analysis using mTfR1 expressing CHO-S cells. Radioimmunoconjugates were first evaluated in an extracellular alpha-synuclein deposition model established by intracranial injection of non-sonicated PFFs into the left striatum of C57Bl/6 WT mice, whereas saline was injected into the contralateral site as control. PET imaging was performed 1, 3, and 7 days post-injection, followed by ex vivo biodistribution, autoradiography and immunofluorescence analysis. Based on the results from these studies, the better-performing antibody candidate was tested similarly in an alpha-synuclein seeding model. The seeding model has intraneural alpha-synuclein aggregation and was established by intracranial injection of sonicated PFFs into both striata of F28tg mice, which overexpress human wild-type alpha-synuclein. Untreated F28tg and C57Bl/6 WT mice served as controls.

RESULTS

The radioimmunoconjugate was produced in sufficient radiochemical yields and purity. There was no impairment of binding affinity towards alpha-synuclein, and acceptable binding with negligible losses to mTfR1. PET imaging with [Zr]Zr-HLu-3-scFab8D3 and [Zr]Zr-HLu-3-(scFv8D3) in the deposition model showed uptake at the site of alpha-synuclein deposits. However, uptake was variable between mice. Reliable PET quantification was hampered due to the small deposition volume (~2 μL). Immunofluorescence revealed specific alpha-synuclein target engagement of both constructs with PFF deposits in the striatum, in contrast to the [Zr]Zr-B12-scFab8D3 control. Unexpectedly, ex vivo autoradiography showed uptake in some controls ([Zr]Zr-B12-scFab8D3 in the contralateral striatum without PFFs), potentially related to astrocyte activation at the injection sites. Ex vivo and PET brain uptake was higher for [Zr]Zr-HLu-3-scFab8D3 when compared to [Zr]Zr-HLu-3-(scFv8D3) and was therefore selected for further testing in the alpha-synuclein seeding model. No significant difference in in vivo and ex vivo brain uptake of [Zr]Zr-HLu-3-scFab8D3 between PFF-injected F28tg, F28tg and C57Bl/6 mice was observed. Furthermore, ex vivo immunofluorescence and autoradiography showed no specific alpha-synuclein target engagement.

CONCLUSIONS

Successful target engagement of [Zr]Zr-HLu-3-scFab8D3 and [Zr]Zr-HLu-3-(scFv8D3) with alpha-synuclein was shown in a PFF deposition model. PET imaging showed variable results, and in vivo detection of the depositions was possible in some cases. Due to the better performance in the deposition model, [Zr]Zr-HLu-3-scFab8D3 was further investigated in an alpha-synuclein seeding model with intraneural Lewy-body pathology, showing no difference between the control groups and PFF-seeded mice. Furthermore, immunostaining of seeded F28tg mice manifested sufficient intraneural alpha-synuclein pathology but no corresponding antibody accumulation. These results underscore the ongoing challenge of imaging intraneural inclusions via immuno-PET.