Patterns of tau, amyloid and synuclein pathology in ageing, Alzheimer’s disease and synucleinopathies

Alzheimer's disease (AD) is neuropathologically defined by deposits of misfolded hyperphosphorylated tau (HP-tau) and amyloid-β. Lewy body (LB) dementia, which includes dementia with Lewy bodies (DLB) and Parkinson's disease dementia (PDD), is characterized pathologically by α-synuclein aggregates. HP-tau and amyloid-β can also occur as co-pathologies in LB dementia, and a diagnosis of mixedAD/DLB can be made if present in sufficient quantities. We hypothesized that the spread of these abnormal proteins selectively affects vulnerable areas, resulting in pathologic regional covariance that differentially associates with pre-mortem clinical characteristics. Our aims were to map regional quantitative pathology (HP-tau, amyloid-β, α-synuclein) and investigate the spatial distributions from tissue microarray post-mortem samples across healthy aging, AD and LB dementia. The study involved 159 clinico-pathologically diagnosed human post-mortem brains (48 controls, 47 AD, 25 DLB, 20 mixedAD/DLB, 19 PDD). The burden of HP-tau, amyloid-β and α-synuclein was quantitatively assessed in cortical and subcortical areas. Principal components (PC) analysis was applied across all cases to determine the pattern nature of HP-tau, amyloid-β and α-synuclein. Further analyses explored the relationships of these pathological patterns with cognitive and symptom variables. Cortical (tauPC1) and temporo-limbic (tauPC2) patterns were observed for HP-tau. For amyloid-β, a cortical-subcortical pattern (amylPC1) was identified. For α-synuclein, four patterns emerged: 'posterior temporal-occipital' (synPC1), 'anterior temporal-frontal' (synPC2), 'parieto-cingulate-insula' (synPC3), and 'frontostriatal-amygdala' (synPC4). Distinct synPC scores were apparent among DLB, mixedAD/DLB and PDD, and may relate to different spreading patterns of α-synuclein pathology. In dementia, cognitive measures correlated with tauPC1,tauPC2 and amylPC1 pattern scores (P ≤ 0.02), whereas such variables did not relate to α-synuclein parameters in these or combined LB dementia cases. Mediation analysis then revealed that in the presence of amylPC1, tauPC1 had a direct effect on global cognition in dementia (n = 65, P = 0.04), while tauPC1 mediated the relationship between amylPC1 and cognition through the indirect pathway (amylPC1 → tauPC1 → global cognition) (P < 0.05). Last, in synucleinopathies, synPC1 and synPC4 pattern scores were associated with visual hallucinations and motor impairment, respectively (P = 0.02). In conclusion, distinct patterns of α-synuclein pathology were apparent in LB dementia, which could explain some of the disease heterogeneity and differing spreading patterns among these conditions. Visual hallucinations and motor severity were associated with specific α-synuclein topographies in LB dementia that may be important to the clinical phenotype and could, after necessary testing/validation, be integrated into semiquantitative routine pathological assessment.