Alzheimer's amyloid β heterogeneous species differentially affect brain endothelial cell viability, blood-brain barrier integrity, and angiogenesis

Abstract

Impaired clearance in the Alzheimer's Disease (AD) brain is key in the formation of Aβ parenchymal plaques and cerebrovascular deposits known as cerebral amyloid angiopathy (CAA), present in >80% of AD patients and ~50% of non-AD elderly subjects. Aβ deposits are highly heterogeneous, containing multiple fragments mostly derived from catabolism of Aβ40/Aβ42, which exhibit dissimilar aggregation properties. Remarkably, the role of these physiologically relevant Aβ species in cerebrovascular injury and their impact in vascular pathology is unknown. We sought to understand how heterogeneous Aβ species affect cerebral endothelial health and assess whether their diverse effects are associated with the peptides aggregation propensities. We analyzed cerebral microvascular endothelial cell (CMEC) viability, blood-brain barrier (BBB) permeability, and angiogenesis, all relevant aspects of brain microvascular dysfunction. We found that Aβ peptides and fragments exerted differential effects on cerebrovascular pathology. Peptides forming mostly oligomeric structures induced CMEC apoptosis, whereas fibrillar aggregates increased BBB permeability without apoptotic effects. Interestingly, all Aβ species tested inhibited angiogenesis in vitro. These data link the biological effects of the heterogeneous Aβ peptides to their primary structure and aggregation, strongly suggesting that the composition of amyloid deposits influences clinical aspects of the AD vascular pathology. As the presence of predominant oligomeric structures in proximity of the vessel walls may lead to CMEC death and induction of microhemorrhages, fibrillar amyloid is likely responsible for increased BBB permeability and associated neurovascular dysfunction. These results have the potential to unveil more specific therapeutic targets and clarify the multifactorial nature of AD.

Keywords: Alzheimer's disease; amyloid β; angiogenesis; blood-brain barrier; cerebral amyloid angiopathy.

FIGURE 1

Biophysical and structural analysis of Aβ species. Structural properties of full‐length, N‐and C‐terminal truncated species and the Aβ40‐Q22 mutant, pre‐treated in HFIP and reconstituted in physiologic salt concentration containing buffer were monitored by native gel electrophoresis/Western blot and Thioflavin‐T binding for up to 3 days. (a) Comparative oligomerization profiles after 1‐ and 3‐days incubation assessed by 5%–20% non‐SDS electrophoresis followed by Western blot analysis probed with a 50:50 mixture of anti‐Aβ monoclonal antibodies 4G8 and 6E10. (b) Fluorescence evaluation of Thioflavin‐T binding to the respective synthetic homologues (50 μM) either freshly reconstituted (time 0) or after 1‐ and 3‐day incubation. Results are expressed in arbitrary units (A.U.) and represent the mean ± SEM of three independent experiments after subtraction of blank levels. (c) Structural assessment of Aβ1–42 oligomers and fibrils in comparison with the non‐aggregated counterpart prepared as described in Materials and Methods and visualized by EM after negative staining with uranyl acetate. Bar represents 100 nm.

FIGURE 2

Induction of CMEC apoptosis and necrosis by full length, truncated, or mutated Aβ peptides. (a) Measurement of apoptotic cell death after treatment with 10 µM of each Aβ peptide or fragment for 1 and 3 days in CMEC. Apoptotic cell death was assessed as formation of fragmented nucleosomes using the Cell Death Detection ELISA kit (Roche). (b) Necrotic cell death, measured as amount of LDH release, after treatment of CMEC with 10 µM of each Aβ peptide or fragment for 1 and 3 days. LDH activity was assessed using the Cytotoxicity Detection Kit^PLUS (Roche) as the production of red formazan from tetrazolium salt, an NADH‐dependent reaction. (c) Representative images of contrast phase microscopy of CMEC after 24‐h treatment with 10 µM of each Aβ peptide or fragment. Cnt indicates untreated control group. Graphs are representative of 3 individual experiments of 2 replicates per group. *p < 0.05, **p < 0.01, and ***p < 0.001 versus Cnt.

FIGURE 3

Effects of full length, truncated, or mutated Aβ peptides on CMEC barrier permeability. CMEC barrier was assessed by monitoring TEER with the ECIS Zθ system (Applied Biophysics) which measures membrane resistance and capacitance in real time. (a) Formation of BBB‐like cell monolayer as a function of elevated and stable resistance after about 48 h on the electrodes. (b–d) Barrier permeability, assessed for 72 h as a reduction in resistance, after treatment with 10 µM Aβ40, Aβ42, Aβ40‐Q22 (b); 1–16, 1–34, 4–34 (c); 4–40, 4–42 (d), all compared to untreated control (dashed line). Graphs represent at least 3 individual experiments of 2 replicates per group.

FIGURE 4

Differential effects of oligomeric and fibrillar Aβ42 on CMEC death and BBB permeability. (a) Thioflavin‐T binding of oligomeric and fibrillar Aβ42 revealing the stability of these species for days after solubilization in cell medium. (b) Apoptotic cell death after treatment of CMEC with 10 µM of freshly solubilized, oligomeric, and fibrillar Aβ42 for 1 and 3 days. (c) Barrier resistance, assessed for 72 h after treatment with 10 µM of oligomeric or fibrillar Aβ42 and compared to untreated control (dashed line). (d) Representative images of the TJ protein claudin‐5 (green) in CMEC monolayers after 1 day of treatment with 10 µM of freshly solubilized, oligomeric, or fibrillar Aβ42. Cell nuclei are counterstained using ToPro nuclear staining (blue). Cnt indicates untreated control monolayers. Images and graphs are representative of 3 individual experiments of 2 replicates per group. *p < 0.05, **p < 0.01, and ***p < 0.001 versus Cnt.

FIGURE 5

Inhibition of angiogenesis by Aβ peptides and fragments. Inhibition of angiogenesis was assessed using the Millipore's Millicell μ‐Angiogenesis Inhibition Assay. CMECs were treated for 24 h with the different amyloid species and monitored by fluorescent staining with 50 μM Calcein AM. (a) Representative images of the effects of Aβ on new‐vessel formation. (b) Number of capillary branches, indicative of angiogenesis, expressed as percent change from untreated controls in each individual experiment. For each experiment capillary branches were counted from 4 randomized images for each treatment. Cnt indicates the untreated control group. Sulforaphane (Sul, 5 or 10 μM) was used as positive control for angiogenesis inhibition. Graphs represent 3 individual experiments of 2 replicates per group. *p < 0.05, **p < 0.01, and ***p < 0.001 versus Cnt.

FIGURE 6

Schematic of proposed model. A healthy brain capillary is represented at the top. The presence of oligomeric Aβ species induces EC death (left arrow), while fibrillary Aβ species increase BBB permeability (right arrow). Damage to the brain capillary due to EC death or BBB dysfunction would need the activation of angiogenesis as a repair mechanism. However, due to the presence of Aβ (all aggregation species and fragments) angiogenesis is inhibited (central red line).