Educational and programmatic components of effective school-based food and nutrition education: an umbrella review.

Alzheimer’s disease (AD) remains the most prevalent cause of dementia and a major unmet medical challenge. Its pathological progression unfolds over decades — from amyloid-β (Aβ) deposition and astrocytic reactivity to tau aggregation and neuronal loss — yet the molecular relationships linking these processes remain incompletely understood. Positron-emission tomography (PET) has transformed our ability to visualize these proteinopathies in vivo, but existing tracers capture only fragments of the disease spectrum. Amyloid PET provides early detection but correlates weakly with cognitive decline, whereas modern tau tracers such as F-MK6240 (ref.), F-PI2620 (ref.) and F-RO-948 (ref.) excel at detecting neuronal tau aggregates only at later stages. Because astrocytic reactivity often precedes tau accumulation, new imaging approaches are required to resolve glial and neuronal contributions within the same pathological continuum. Reactive astrocytes markedly upregulate monoamine oxidase-B (MAO-B), leading to excessive γ-aminobutyric acid (GABA) release and oxidative stress that exacerbate neurodegeneration. However, reactive astrocytes often cluster near tau deposits, producing overlapping PET signals that confound the interpretation of tracer specificity. Early-generation MAO-B tracers such as C-L-deprenyl-D (C-DED) and its F-labelled analogue F-DED have been used to visualize astrocytic reactivity in vivo; however, both exhibit significant limitations. Quantification is challenging owing to their irreversible binding kinetics, whereas the formation of radiolabelled metabolites capable of crossing the blood–brain barrier and binding to monoamine transporters further complicates signal interpretation. In addition, their low selectivity for MAO-B over monoamine oxidase A (MAO-A) reduces target specificity. Moreover, C-labelled tracers such as C-DED, C-SL25.1188 and C-BU99008 suffer from the short physical half-life of carbon-11 (≈ 20 min), requiring an on-site cyclotron and radiochemistry facility, which limits their clinical utility. These constraints have motivated the development of F-labelled, reversible and more selective MAO-B tracers that enable reliable imaging of reactive astrogliosis. Among these, F-THK5351 remains a unique and controversial case. Initially developed as a tau tracer, it was later shown to bind MAO-B with high affinity, blurring the interpretation of its PET signals. Subsequent preclinical and clinical studies — including large-scale human imaging — revealed distinct retention patterns, even in amyloid-negative AD cases, suggesting that F-THK5351 may capture aspects of astrocytic pathology that are invisible to conventional tau tracers. Autoradiography and pharmacological blocking studies have supported this dual-binding profile, showing that MAO-B inhibitors markedly reduce F-THK5351 binding, yet complete displacement is rarely achieved, implying residual affinity for tau aggregates. However, even in cognitively normal individuals without tauopathy, the degree of tracer displacement is similar to that seen in patients with AD. More importantly, the residual affinity attributed to tau aggregates in AD is difficult to appreciate on visual analysis, and therefore even presumed tau binding is unlikely to be clinically significant. Conversely, in vitro studies using recombinant tau fibrils or tau-rich tissues have reported detectable but weaker interactions relative to MAO-B, although the strength, stoichiometry and binding kinetics of these interactions remain poorly defined. Thus, despite extensive clinical use and interpretive debate, the precise molecular determinants and relative contributions of MAO-B and tau binding to F-THK5351 PET signals have to our knowledge never been quantitatively and systematically characterized.