Post-mortem assessment of hypoperfusion of cerebral cortex in Alzheimer's disease and vascular dementia

By measuring perfusion-sensitive and disease-related proteins in post-mortem Alzheimer's disease brain, Thomas et al. reveal pathological hypoperfusion of neocortex that exceeds the decline in metabolic demand. The hypoperfusion is independent of the severity of structural vascular disease in most cases, but is associated with elevated levels of the vasoconstrictor endothelin-1.Perfusion is reduced in the cerebral neocortex in Alzheimer's disease. We have explored some of the mechanisms, by measurement of perfusion-sensitive and disease-related proteins in post-mortem tissue from Alzheimer's disease, vascular dementia and age-matched control brains. To distinguish physiological from pathological reduction in perfusion (i.e. reduction exceeding the decline in metabolic demand), we measured the concentration of vascular endothelial growth factor (VEGF), a protein induced under conditions of tissue hypoxia through the actions of hypoxia-inducible factors, and the myelin associated glycoprotein to proteolipid protein 1 (MAG:PLP1) ratio, which declines in chronically hypoperfused brain tissue. To evaluate possible mechanisms of hypoperfusion, we also measured the levels of amyloid-beta(40), amyloid-beta(42), von Willebrand factor (VWF; a measure of microvascular density) and the potent vasoconstrictor endothelin 1 (EDN1); we assayed the activity of angiotensin I converting enzyme (ACE), which catalyses the production of another potent vasoconstrictor, angiotensin II; and we scored the severity of arteriolosclerotic small vessel disease and cerebral amyloid angiopathy, and determined the Braak tangle stage. VEGF was markedly increased in frontal and parahippocampal cortex in Alzheimer's disease but only slightly and not significantly in vascular dementia. In frontal cortex the MAG:PLP1 ratio was significantly reduced in Alzheimer's disease and even more so in vascular dementia. VEGF but not MAG:PLP1 increased with Alzheimer's disease severity, as measured by Braak tangle stage, and correlated with amyloid-beta(42) and amyloid-beta(42): amyloid-beta(40) but not amyloid-beta(40). Although MAG:PLP1 tended to be lowest in cortex from patients with severe small vessel disease or cerebral amyloid angiopathy, neither VEGF nor MAG:PLP1 correlated significantly with the severity of structural vascular pathology (small vessel disease, cerebral amyloid angiopathy or VWF). However, MAG:PLP1 showed a significant negative correlation with the level of EDN1, which we previously showed to be elevated in the cerebral cortex Alzheimer's disease. These finding are in contrast with the previously demonstrated reduction in EDN1, and positive correlation with MAG:PLP1, in the hypoperfused white matter in Alzheimer's disease. The decline in MAG:PLP1 strongly suggests pathological hypoperfusion of the frontal cortex in Alzheimer's disease. Although severe small vessel disease or cerebral amyloid angiopathy may contribute in some cases, abnormal vascular contractility mediated by EDN1 is likely to be a more important overall contributor. Both amyloid-beta accumulation and hypoperfusion are likely to cause the upregulation of VEGF.