Bispecific brain-penetrant antibodies for treatment of Alzheimer's disease

Abstract

The emerging class of bispecific antibodies represents a significant advancement in Alzheimer's disease (AD) immunotherapy by addressing the limited brain concentrations achieved with conventional monoclonal antibodies. The majority of bispecific antibodies developed for AD treatment utilize transferrin receptor (TfR1)-mediated transcytosis to enhance blood-brain barrier (BBB) penetration, resulting in higher and more uniform brain concentrations compared to conventional antibodies. This improved delivery has demonstrated superior efficacy in reducing brain amyloid-beta (Aβ) burden. Additionally, TfR1-mediated delivery may help mitigate adverse effects such as amyloid-related imaging abnormalities (ARIA). This is likely achieved by a reduction in antibody accumulation at vascular Aβ deposits, resulting from the combined effects of lower dosing and a different brain entry route when using bispecific antibodies. Besides targeting Aβ, bispecific antibodies have been engineered to address other key pathological features of AD, including tau pathology and neuroinflammatory targets, which are critical drivers of disease progression. These antibodies also show promise in diagnostic applications, particularly as radioligands for antibody-based positron emission tomography (immunoPET), leveraging their rapid brain delivery and efficient and specific target engagement. Moreover, the principles of bispecific antibody technology have been adapted for use beyond immunotherapy. The incorporation of TfR1-binding domains into enzymes, antisense oligonucleotides, or viral vectors such as adeno-associated viruses broadens their therapeutic potential. These approaches may enable more efficient treatment strategies, not only for AD but also for other neurological disorders, by facilitating the delivery of diverse therapeutic agents across the BBB.

Keywords: Alzheimer’s disease; Bispecific antibody; Immunotherapy; The blood-brain barrier.

Fig. 1

Schematic illustration of transferrin receptor (TfR)-mediated transport across the blood-brain barrier (BBB). A. A bispecific antibody, capable of binding both the TfR and amyloid-beta (Aβ), is administered and circulates in the brain vasculature. B. The bispecific antibody binds to TfR on the luminal surface of the endothelial cells of BBB. C. The TfR facilitates transcytosis, transporting the antibody across the BBB. D. Once in the brain parenchyma, the antibody binds to Aβ aggregates.

Fig. 2

Bispecific antibody formats. A. A full-sized TfR1 antibody fused at the C-terminal ends of its heavy chains to single-chain variable fragments (scFvs) of an anti-Aβ antibody. B. A bispecific antibody comprising one heavy and light chain derived from an antibody targeting an intrabrain antigen, while the other heavy and light chain are derived from a TfR1 antibody. C. A full-sized anti-Aβ antibody fused at one of the C-terminal ends of its heavy chain to a fragment antigen-binding (Fab) region derived from a TfR1 antibody. D. A full-sized anti-Aβ antibody fused at both C-terminal ends of its light chains to scFvs derived from a TfR1 antibody. E. A tetravalent tandem bispecific antibody format with variable domains of a TfR binding antibody expressed between the constant and variable domains of a full-sized Aβ antibody. F. A bispecific antibody format (ATV) utilizing an engineered Fc domain, where one of the two heavy chains contains an amino acid sequence that binds to TfR1. G. A full-sized anti-Aβ antibody fused at one of the C-terminal ends of its heavy chain to a scFv derived from a TfR1 antibody. Blue/purple domains are directed towards Aβ, while orange/red domains, indicated by arrows, are directed towards TfR1.

Fig. 3

Brain distribution of regular IgG and bispecific antibodies. A Representative images from AI-segmented, 3D-reconstructed brains of wild-type mice following whole-body tissue clearing. Mice were dosed with AF647-conjugated control IgG (left) or bispecific ATV (right). One day post-administration, the bispecific ATV exhibited widespread distribution throughout the brain, whereas the control IgG was primarily localized to the brain surface, corresponding to leptomeningeal tissue and associated blood vessels, with weaker signal also observed in the lateral ventricles. Image adapted from Khoury et al., [17]. B. Brain distribution in an Alzheimer's disease (AD) mouse model dosed with radiolabeled control anti-Aβ RmAb158 (left, murine version of lecanemab) or bispecific RmAb158-scFv8D3 (right). At six days post-administration, brain distribution was visualized using ex vivo autoradiography and compared with Aβ40 immunostaining. RmAb158-scFv8D3 was widely distributed throughout the brain, clearly overlapping regions with Aβ pathology. In contrast, RmAb158 was mostly confined to central brain regions, likely the lateral ventricles, and appeared as hotspots, probably corresponding to large blood vessels. Image adapted from Syvänen et al., [37].

Fig. 4

Difference between the therapeutic and the diagnostic target. A. Amyloid-β (Aβ) is visualized in green, while a bispecific version of the anti-Aβ antibody lecanemab is shown as white “grains”. The merged image (right) shows how the antibody distributes to the halo of diffuse Aβ aggregates, surrounding the amyloid core of the plaque, which remains largely inaccessible to the antibody. B. Sagittal PET images of Alzheimer’s disease model mice using a bispecific antibody-based radioligand derived from RmAb158 (the murine version of lecanemab) and the small molecule amyloid PET radioligand [¹¹C]PiB, imaged at baseline and after two months of treatment with the BACE1 inhibitor NB-360 or vehicle control. The antibody-based radioligand clearly visualizes the treatment effect, showing reduced PET signal in NB-360–treated mice compared to vehicle-treated mice, indicating lower levels of Aβ pathology. In contrast, imaging with the standard small molecule radioligand [¹¹C]PiB does not detect this reduction in Aβ levels. C. Post-mortem measurements of brain Aβ₄₀ and Aβ₄₂ concentrations in the three groups confirm the PET findings with the antibody-based radioligand, showing significantly lower Aβ levels in NB-360–treated mice compared to vehicle controls. Images B and C are adapted from Meier et al., [78].