Abstract:
Introduction: Alzheimer’s disease (AD) is the most common type of dementia. Increasing evidence suggests that AD pathogenesis is not restricted to neuronal loss, but includes the involvement of the non-neuronal cells of the neurovascular unit. This study aims to characterise non-neuronal cells and their potential interaction with the AD neuropathology in the AD human brain by examining a range of proteins involved in the neurovascular unit in the middle temporal gyrus (MTG) using tissue microarrays (TMAs) containing samples of post-mortem human tissue. Methods: Immunohistochemistry was carried out on paraffin-embedded TMAs – comprised of up to 56 control and 53 AD cases. Antibodies were used to protein target proteins associated with AD neuropathology (amyloid-β, tau, ubiquitin), astrocytes (GFAP), microglia (Iba-1, HLA-DR), endothelial cells (CD31, lectin UEA-I), smooth muscle cells (αSMA) and pericytes (PDGFRβ) were used. The V-slide scanner automated imaging system was used to image the TMA’s. Densitometric analysis was carried out on the acquired images using Metamorph. Statistical analysis was conducted using GraphPad Prism to compare the mean densitometric measurements carried out for the AD and control cohorts. Correlations were conducted to investigate potential interactions between various components of the neurovascular unit, the degree of AD neuropathology and the case variables. Results and Conclusion: There was a differential degree in protein expression changes of the different components of the neurovascular unit in AD human brains. These results suggest the involvement of these cells in the neuroinflammatory and vascular changes associated with AD. Changes found to be associated with neuroinflammation (astrocytes, microglia, endothelial cells) and angiogenesis (endothelial cells) correlated with AD neuropathology, notably with tau, highlighting the interaction between neuropathology and the neuroinflammatory response. In comparison, vascular changes, specifically in perivascular cells (smooth muscle cells and pericytes) reveal no significant correlation with the extent of AD neuropathology, suggesting the independence of these changes from the neuropathology. Furthermore, although vascular changes were not associated with age, neuroinflammation-related changes were weakly associated with age. This study sheds light into the potential power of a TMA approach to study AD pathogenesis.