Differences and overlaps in TDP-43 pathology of ‘pure’ LATE-NC compared to LATE-NC coexisting with Alzheimer’s disease

Abstract

Limbic-predominant age-related TDP-43 encephalopathy (LATE) is a common substrate of dementia in the elderly. LATE and Alzheimer’s disease (AD) share similar clinical features, and their underlying neuropathological changes—LATE-NC and ADNC—commonly co-occur. However, the histomorphological and molecular features of TDP-43 pathology in LATE-NC with or without coexisting ADNC are not yet well understood. We performed immunohistochemistry in paraffin-embedded tissue from the hippocampus, amygdala, and temporal and frontal cortices of 108 human autopsy cases including 20 cognitively unimpaired controls, 20 AD dementia cases with moderate-high severity of ADNC without LATE-NC (ADNC group), 34 AD dementia cases with LATE-NC (ADNC + LATE-NC group), 17 dementia cases with LATE-NC but no/low ADNC (pure LATE-NC group), and 17 FTLD-TDP Type A cases. We assessed TDP-43 aggregate morphology and composition using antibodies against different TDP-43 epitopes: pS409/410, pS403/pS404, and C- and N-terminal TDP-43. We also investigated nuclear clearance of physiological TDP-43 and cytoplasmic colocalization of TDP-43 and tau proteins. Pure LATE-NC cases were on average 10 years older at death than ADNC + LATE-NC, had less cognitive impairment, higher prevalence of argyrophilic grain disease (AGD) pathology, aging-related tau astrogliopathy (ARTAG), and APOEε2 allele. They also tended to show lower APOEε4 frequencies, but similar frequencies of hippocampal sclerosis and LATE-NC stages. Importantly, LATE-NC predominantly displayed a mesh-like neuritic TDP-43 pattern in the hippocampus, extending from CA1/2 to subiculum. This mesh-like pattern was present in 81% of pure LATE-NC cases and only in 18% of ADNC + LATE-NC. This pattern was also observed in 53% of FTLD-TDP Type A cases. Moreover, the aggregate composition differed in pure LATE-NC and ADNC + LATE-NC, with LATE-NC cases exhibiting increased burdens of several phosphorylated and non-phosphorylated TDP-43 species, while only the pS409/pS410 epitope was significantly associated with ADNC + LATE-NC in the amygdala. Nuclear clearance patterns also tended to differ between pure LATE-NC and ADNC + LATE-NC. Similar to ADNC + LATE-NC, TDP-43 and tau proteinopathies colocalized in pure LATE-NC with comorbid primary age-related tauopathy (PART) or low ADNC. These data suggest that LATE-NC tends to be modified in the presence of moderate–high ADNC. These differences may reflect upstream influences (age, genetics, and environmental risk factors), direct protein–protein interactions, and/or other impacts of ADNC-related mechanisms on TDP-43 proteinopathy, potentially relevant for clinical trial design and future therapeutic applications.