Blood-brain barrier leakage in Alzheimer's disease: From discovery to clinical relevance

Abstract

Alzheimer's disease (AD) is the most common form of dementia. AD brain pathology starts decades before the onset of clinical symptoms. One early pathological hallmark is blood-brain barrier dysfunction characterized by barrier leakage and associated with cognitive decline. In this review, we summarize the existing literature on the extent and clinical relevance of barrier leakage in AD. First, we focus on AD animal models and their susceptibility to barrier leakage based on age and genetic background. Second, we re-examine barrier dysfunction in clinical and postmortem studies, summarize changes that lead to barrier leakage in patients and highlight the clinical relevance of barrier leakage in AD. Third, we summarize signaling mechanisms that link barrier leakage to neurodegeneration and cognitive decline in AD. Finally, we discuss clinical relevance and potential therapeutic strategies and provide future perspectives on investigating barrier leakage in AD. Identifying mechanistic steps underlying barrier leakage has the potential to unravel new targets that can be used to develop novel therapeutic strategies to repair barrier leakage and slow cognitive decline in AD and AD-related dementias.

Keywords: Alzheimer’s disease; Barrier dysfunction; Barrier leakage; Blood-brain barrier; Cerebrovasculature; Neurovasculature.

Figure 1.. Pathological Changes in the Central Nervous System in Alzheimer’s Disease (AD).

Amyloid-beta (Aβ) and tau have been proposed to be the two key players that drive pathological progression in AD. A) Aβ, formed by the abnormal and sequential cleavage of amyloid precursor protein (APP), binds to other Aβ subunits to form oligomers, protofibrils, fibrils, diffused Aβ plaques (surrounded by glial inflammation), and Aβ plaques with a neuritic core causing neurodegeneration. B) In contrast, microtubule-associated tau proteins get abnormally hyperphosphorylated to form tau oligomers, paired helical filaments (PHF), and neurofibrillary tangles that drive AD pathology. Together, Aβ, tau, and their protein aggregates affect both neuronal and glial cells in the brain leading to C) elevated calcium influx and disruption in synaptic transmission, D) astrocytic and microglial inflammation, E) autophagy, F) cerebral amyloid angiopathy, and G) breakdown of the myelin sheath around neurons.

Figure 2.. Blood-Brain Barrier in Health.

A healthy neurovascular unit at the capillary endothelium forms the structural basis of the blood-brain barrier. A) Brain capillary endothelial cells are covered with a 200–2,000 nm thick glycoprotein-rich glycocalyx layer on the luminal side and B) a 50–100 nm thick proteoglycan-rich vascular basement membrane on the abluminal side. Proteoglycans in the vascular basement membrane anchor C) astrocytes and D) pericytes around capillary endothelial cells. E) Adhesion molecules further seal the connections among and between astrocytes and pericytes. F-G) At the endothelial interface, interendothelial junctions are further sealed by proteins such as claudin-5, occludin, zonula occludens-1, 2, 3, platelet endothelial cell adhesion molecule-1 (PECAM-1), vascular cell adhesion molecule-1 (VCAM-1), vascular endothelial (VE) – cadherins and form an intact blood-brain barrier.

Figure 3.. Blood-Brain Barrier in Alzheimer’s Disease.

Pathological changes at the neurovascular unit in AD include A) increased coagulation and thrombosis, B) structural changes in endothelial cells and disruption of interendothelial junctions, C) altered proteoglycan and glycoprotein levels in the basement membrane, D) structural changes in astrocytes such as swollen astrocytic end-feet, loss of astrocytic coverage around endothelial cells, E) clasmatodendrosis, F) in addition to a loss of dendritic connections and G) pericytes around endothelial cells.

Figure 4.. Blood-Brain Barrier Leakage & Pathological Changes in Transgenic Mouse Models of AD.

Overview of eleven of the most frequently used transgenic mouse models of AD and the onset (in months) of different pathological changes in each model. Pathological changes include cognitive dysfunction (beige), Aβ pathology (grey), cerebral amyloid angiopathy (tan), cerebral blood flow reduction (light red), and microhemorrhages (light blue). An exhaustive literature search (1970 – 2020) revealed that barrier leakage has been previously characterized in these mice using tracers like albumin (dark red, 70 kDa), dextrans (4–150 kDa), IgG (150 kDa), Evans Blue dye (0.96 kDa) and Gadolinium (0.33 kDa). The extent of leakage for these tracers is depicted as log (fold-change) of the leakage value for transgenic mice compared to that of age-matched non-transgenic mice.

Figure 5.. Mechanisms leading to Blood-Brain Barrier Leakage in Alzheimer’s Disease.

A) Schematic overview of transmembrane receptors at the neurovascular unit and their ligands that have been previously shown to induce blood-brain barrier leakage in vitro or in vivo. B) Many of these ligand-receptor pairs lead to one or more downstream signaling pathways such as PI3K/Akt, MAPK, or RhoA/ROCK that cause physiological changes in and around capillary endothelial cells. C) Some of these changes include downregulation of tight junction protein expression, delocalization of tight junction proteins from the interendothelial junctions, degradation of tight junction proteins, or changes in the cytoskeletal architecture of endothelial cells.

Figure 6.. Therapeutic Strategies for Blood-Brain Barrier Leakage in Alzheimer’s Disease.

Summary of selected studies that show the relevance of therapeutic strategies towards barrier leakage in transgenic AD mice and rats. The therapeutic strategies depicted here have either blood-brain barrier opening effects (red), blood-brain barrier sealing effects (blue), or no effect on the blood-brain barrier (dark gray). The numbers in each square refer to published studies shown in Table 4 for each strategy/model combination. Note that the number of white spaces indicates unstudied/unknown strategy/model combinations.