With the emerging field of antibody treatment for neurodegenerative diseases, reliable tools are needed to evaluate new therapeutics, diagnose and select patients, monitor disease progression, and assess therapy response. Immuno-PET combines the high affinity and exceptional specificity of monoclonal antibodies with the non-invasive PET imaging technique. The application in neurodegenerative disease brain imaging is limited, most probably due to the limited uptake across the BBB. In recent years, the emergence of brain-shuttle antibodies with enhanced uptake across the blood–brain barrier extended the use of 124I-immuno-PET to brain imaging [24]. We established 89Zr-immuno-PET to pair these BBB-shuttle constructs with a PET radionuclide with superior imaging qualities and better availability than 124I [16, 17, 25, 26]. These studies revealed that the optimal timepoint for PET imaging with 89Zr-labeled bispecific aducanumab (Adu-8D3), targeting amyloid-beta plaques and mTfR1 for BBB shuttling, was 7 days p.i.. To investigate if a better target-to-background ratio can be achieved earlier, a faster clearing BBB-shuttle antibody having a His310Ala mutation (AduH310A-8D3) to reduce FcRn binding was evaluated in APP/PS1 TG mice.

Evaluation of the blood kinetics revealed a faster clearance for [89Zr]Zr-AduH310A-8D3 compared to the recently reported [89Zr]Zr-Adu-8D3 [16]. The faster blood clearance resulted, at an earlier time point, in a significantly higher brain uptake in APP/PS1 TG mice compared to WT control mice. For [89Zr]Zr-AduH310A-8D3, this was at 24 h p.i. while for [ 89Zr]Zr-Adu-8D3, this was at 72 h p.i. as previously reported [16, 17]. However, in both cases, the contrast enhanced further over time, leading to a more favorable target-to-background ratio at 72 h p.i. for [89Zr]Zr-AduH310A-8D3 in APP/PS1 TG mice compared to WT control mice. Since the same trend was reported for [89Zr]Zr-Adu-8D3 (168 h versus 72 h p.i.), it can be concluded that in addition to blood activity levels, the slower efflux of amyloid-beta-related radiotracer uptake also plays a key role in reaching optimal imaging contrast. The isotype control B12H310A-8D3 in APP/PS1 TG mice showed a similar trend as AduH310A-8D3 in WT control mice but was retained longer in the blood (Figure S7). Therefore, brain uptake levels were overall higher, and B12H310A-8D3 was not suitable to be directly compared to AduH310A-8D3. The specific amyloid-beta binding of AduH310A-8D3 was confirmed by ex vivo immunofluorescence and autoradiography. None or negligible antibody was detectable for B12H310A-8D3 in APP/PS1 TG mice, which is in line with what was previously reported for [89Zr]Zr-B12-8D3 in APP/PS1 TG mice (Fig. 5) [17]. A comparable biodistribution pattern (except the brain) was observed for [89Zr]Zr-AduH310A-8D3 and [89Zr]Zr-Adu-8D3 in APP/PS1 TG and WT control mice. High spleen uptake can be related to the high TfR expression in this organ and its function to salvage red blood cells [27]. Uptake in the kidney and liver is related to their function as catabolic organs. At the same time, the uptake in bone-related tissue can be explained by the tropism of 89Zr to bones and the TfR-expressing erythrocyte progenitor cells in the bone marrow [28].

Although an earlier amyloid-beta-related imaging time point through a reduced plasma-half life was achieved, a substantial downside is the overall lower brain uptake of [89Zr]Zr-AduH310A-8D3 with only ~ 1%ID/g at 24 and 72 h p.i., which is about half the brain uptake of [89Zr]Zr-Adu-8D3. Furthermore, the difference in amyloid-beta driven uptake in APP/PS1 TG versus WT control mice is smaller: only 1.2 fold higher for [89Zr]Zr-AduH310A-8D3 versus 2.2 fold higher for [89Zr]Zr-Adu-8D3 as reported previously [16]. This makes the H310A mutation less favorable in 89Zr-immuno-PET applications for CNS targets since the inferior capability to differentiate between disease and healthy control group could translate to limited visualization of specific target uptake in earlier stages of neurodegenerative diseases, which is a crucial aspect in current CNS PET tracer development. Putting these findings in context to similar work, smaller bispecific constructs with mTfR1-mediated brain shuttling and no Fc region also achieved a faster blood clearance [10, 11, 13]. In these studies, more rapid blood clearance resulted in lower amyloid-beta-specific brain uptake. This aligns with past findings, which demonstrated that a longer serum half-life with more prolonged circulation would increase the interactions with the TfR receptor on the BBB, resulting in better brain drug delivery [29]. Notable, in some cases, altered in vitro affinities to mTfR1 or amyloid-beta were observed compared to the native construct. In this work, the H310A mutation and the DFO* modification with the subsequent 89Zr-radiolabeling did not alter the in vitro affinity to mTfR1 or amyloid-beta of [89Zr]Zr-AduH310A-8D3 when compared to Adu-8D3 (Figure S1; Table S2).

During the time these studies were conducted, it was considered that transcytosis across the BBB is independent of direct FcRn interactions, as was shown by Selin et al. by comparing 125I-labelled 8D3 and a Fab fragment of 8D3 (Fab-8D3), which lacks the Fc fragment [24]. These findings align with earlier studies for regular IgGs, stating that the limited antibody uptake across the BBB occurs independent of FcRn interactions [30,31,32]. In contrast, several groups have concluded that the Fc-FcRn interaction does influence mAb uptake across the BBB and the efflux of mAbs out of the brain [33,34,35]. However, the experimental setups of these studies differ significantly (e.g., animal models, route of mAb delivery, methods of uptake quantification, and dosing) from this study and the one from Selin et al. Therefore, we conclude that while the Fc-FcRn interaction may play a role in mAb brain uptake and efflux, the impact of mTfR binding potentially outweighs these effects.

Despite the less favorable results regarding the effect of Fc-FcRn mutations on 89Zr-immuno-PET in CNS applications, these mutations could be promising for applying other radionuclides with less or non-residualizing behavior. For 124I (t1/2 = 4.18 d), a non-residualizing PET radionuclide, earlier time points 3 days post-injection were achieved with an intact Fc-FcRn interaction in the past due to significantly lower non-amyloid beta related brain uptake [24]. However, there are limitations when using 124I, such as availability, price, and imaging quality [25, 26]. The clinically relevant SPECT radionuclide 123I (t1/2 = 13.2 h) could be a suitable alternative. Despite the lower spatial resolution of this imaging technique in clinical settings, a small molecule anti-amyloid-beta SPECT radiotracer, 123I-ABC577, was able to differentiate Alzheimer’s disease patients from healthy controls [36]. For immuno-SPECT applications, significant brain retention of RmAb158-scFv8D3 was shown in tg-ArcSwe mice at day 3 p.i. even with the unfavorable SPECT radionuclide 125I [37]. These considerations suggest that brain-penetrating antibodies with faster blood clearance and the favorable SPECT radionuclide 123I could be a promising approach for CNS applications. In addition, using the H310A mutation could also enable the use of other routine PET radionuclides with a shorter half-life for immuno-PET, like Copper-64 (t1/2 = 12.7 h). For Fluorine-18 (t1/2 = 109 min) and Gallium-68 (t1/2 = 68 min), the clearance rates are probably still too low.