One of the hallmarks of Alzheimer’s disease (AD) are deposits of amyloid-beta (Aβ) protein in amyloid plaques in the brain. The Aβ peptide exists in several forms, including full-length Aβ1-42 and Aβ1-40 – and the N-truncated species, pyroglutamate Aβ3-42 and Aβ4-42, which appear to play a major role in neurodegeneration. We previously identified a murine antibody (TAP01), which binds specifically to soluble, non-plaque N-truncated Aβ species. By solving crystal structures for TAP01 family antibodies bound to pyroglutamate Aβ3-14, we identified a novel pseudo β-hairpin structure in the N-terminal region of Aβ and show that this underpins its unique binding properties. We engineered a stabilised cyclic form of Aβ1-14 (N-Truncated Amyloid Peptide AntibodieS; the ‘TAPAS’ vaccine) and showed that this adopts the same 3-dimensional conformation as the native sequence when bound to TAP01. Active immunisation of two mouse models of AD with the TAPAS vaccine led to a striking reduction in amyloid-plaque formation, a rescue of brain glucose metabolism, a stabilisation in neuron loss, and a rescue of memory deficiencies. Treating both models with the humanised version of the TAP01 antibody had similar positive effects.
Here we report the discovery of a unique conformational epitope in the N-terminal region of Aβ, which offers new routes for active and passive immunisation against AD.