Influence of Alzheimer’s associated Aβ oligomers and oxidative stress on blood–brain barrier dysfunction

ABSTRACT

Blood–brain barrier (BBB) dysfunction is an early event observed in Alzheimer’s disease (AD). Two characteristics of AD brain and brain vasculature contribute to BBB dysfunction: the accumulation of aggregated amyloid-β protein (Aβ) and an increase in oxidative stress. This work uses a BBB model of primary human brain microvascular endothelial cells to investigate the individual and synergistic influence of both pathogenic Aβ oligomers and oxidative stress on BBB transendothelial electrical resistance (TEER), an indicator of barrier integrity. Results indicate that nontoxic, physiological concentrations of Aβ oligomers reduce TEER, while Aβ monomer remains inert. Moreover, introducing mild oxidative stress, which alone does not influence monolayer integrity, exacerbates the effect of Aβ oligomers on TEER within this BBB model. These findings advance the understanding of BBB dysfunction in AD and point toward therapeutic strategies targeting this early event that contributes to a currently irreversible disease.

Introduction

Endothelial dysfunction within the blood–brain barrier (BBB) is observed in a multitude of neurological disorders, including multiple sclerosis, Alzheimer’s disease (AD), Parkinson’s disease, stroke, and epilepsy.¹ A healthy BBB involves junctional proteins that impart low permeability and render a barrier 50- to 100-fold tighter than peripheral capillaries. Endothelial dysfunction leads to breakdown of the BBB characterized by reduction of tight and adherents junctions, changes in capillary basement membrane, and amplification of bulk flow transcytosis.^2,3 In parallel, neuronal function is compromised, as increased vascular permeability results in an influx of blood-borne molecules into the central nervous system, causing brain inflammation and altered homeostasis.^4,5

In AD, endothelial dysfunction and BBB disruption are among the earliest events in disease pathogenesis.⁴ Two key attributes of AD are known to affect BBB integrity: accumulation of aggregated amyloid-β protein (Aβ) and increased oxidative stress.⁶ Aggregated Aβ deposits as insoluble plaques in the brain as well as around the brain vasculature, where they contribute to breakdown of the BBB.^2,3 Aβ aggregates disrupt BBB integrity by affecting the expression and localization of tight junction proteins, including zonula occluden-1 protein (ZO-1),⁷ zonula occluden-2 protein (ZO-2),⁸ claudin-5,⁹ and occludin.^8,9 In parallel, AD-associated increases in oxidative stress impair mitochondrial function and activate proteolytic enzymes that degrade components of the extracellular matrix (ECM) and basement membrane, resulting in BBB dysfunction.¹⁰

These two key attributes of AD are also interconnected. In murine models that exhibit AD-like pathologies, including amyloid plaque deposition, enhanced oxidative stress is an early development of the disease state.¹¹ Similarly, the overexpression of Aβ_1–42 in yeast cultures is linked to cellular stress response, which includes oxidative stress.¹² Cultures of human neuroblastoma cells incubated with Aβ_1–42 oligomers also exhibit elevated intracellular reactive oxygen species (ROS) levels.¹³ In human brains, when levels of Aβ are elevated, a parallel increase in indicators of oxidative stress is observed.^14,15 Spatial variations also reflect this association. Both Aβ and oxidation products are increased within the hippocampus and cortex, while brain regions with low Aβ levels also present low concentrations of oxidative stress markers.¹⁶ One target of oxidation in AD is the low-density lipoprotein receptor-related protein (LPR1), an Aβ transporter, and its oxidation results in impaired clearance of Aβ from the brain.¹⁷ Conversely, binding of aggregated Aβ by metal ions, specifically zinc and copper, leads to ROS production,¹⁸ including hydrogen peroxide (H₂O₂),¹⁴ and causes oxidative damage, including lipid peroxidation.¹⁰

Among Aβ aggregates, oligomers are proposed to be the primary pathogenic culprit. Oligomeric Aβ has been shown to selectively disrupt calcium release in neuronal cells, leading to synaptic dysfunction¹⁹ and to induce significant inflammatory responses in endothelial cells.²⁰ Our lab has previously reported that Aβ_1–42 oligomers also increase intracellular ROS in human neuroblastoma cells.¹³ We have additionally observed Aβ compromise of the BBB in a manner inversely correlated with the aggregate size.⁷ The current study extends our prior work as well as evidence for the role of Aβ oligomers in AD pathology by demonstrating that isolated Aβ oligomers, but not monomers, reduce BBB integrity, as evidenced by a decrease in transendothelial electrical resistance (TEER) of monolayers composed of primary human brain microvascular endothelial cells (HBMVECs). Moreover, these changes occur at nontoxic, physiological concentrations of Aβ. We additionally illustrate the interplay between Aβ oligomers and oxidative stress by demonstrating that TEER is further reduced when Aβ oligomer treatment of HBMVEC monolayers occurs simultaneously with low concentrations of oxidative stress inducer H₂O₂, which alone do not influence monolayer integrity. This work validates that Aβ oligomers alter barrier integrity, and this alteration can be further exacerbated in a stressed oxidative state.

Materials and methods

Materials

Aβ_1–40 was purchased from Peptide 2.0 (Chantilly, VA). Aβ_1–42 was purchased from AnaSpec (Fremont, CA). Bovine plasma fibronectin, Complete Endothelial Cell Medium, penicillin, streptomycin, fetal bovine serum (FBS), and endothelial cell growth supplement were acquired from ScienCell (Carlsbad, CA). Hydrocortisone, 1,1,1,3,3,3-hexafluro-2-propanol (HFIP), H₂O₂, tumor necrosis factor-α (TNF-α), and dimethyl sulfoxide (DMSO) were purchased from Sigma Aldrich (St. Louis, MO). 2,3-Bis-(2-methoxy-4-nitro-5-sulfophenyl)-2 H-tetrazolium-5-carboxanilide (XTT) cell proliferation assay kit was attained from Cayman Chemical (Ann Arbor, MI). Laemilli buffer, Precision Plus Western C standard, Western C horseradish peroxidase (HRP), and 6E10 monoclonal antibody were purchased from Biolegend (San Diego, CA). HRP anti-mouse IgG antibody was procured from Rockland Chemical (Boyertown, PA). Chemiluminescent substrates and Tween 20 detergent solution were acquired from Fisher Scientific (Waltham, MA). All other chemicals were purchased from VWR (Radnor, PA).

Aβ_1–40 monomer preparation

Aβ_1–40 was selected for monomer preparations, as this isoform, which is the most abundant in vivo,²¹ aggregates slower than longer isoforms.²² Lyophilized synthetic, crude Aβ_1–40 was stored at −20°C until reconstitution and preparation for experiments. To remove preexisting aggregates, Aβ_1–40 was dissolved in 50 mM NaOH at a concentration of 2–4 mg/mL and purified via fast protein liquid chromatography (FPLC). Purification utilized size-exclusion chromatography (SEC) on a Superdex 75 HR10/30 column (GE Healthcare, Piscataway, NJ) with elution in 12 mM phosphate (pH 7.4). The column was pretreated with 2 mg/mL of bovine serum albumin (BSA) to reduce nonspecific Aβ interaction with the column matrix. Concentrations of isolated monomer were determined by measuring UV absorbance at 277 nm on a Beckman Coulter UV/Vis Spectrophotometer and using an extinction coefficient of 1450 1/(M·cm). Samples were further diluted and immediately used for experiments. All manipulation of Aβ monomer was conducted using low-protein-binding tubes and pipette tips.

Aβ_1–42 oligomer preparation

Oligomers were prepared using Aβ_1–42, as this isoform comprises the earliest aggregates in vivo and forms more stable oligomeric structures than Aβ_1–40.²² Lyophilized synthetic Aβ_1–42, ≥95% purity, was stored at −80°C until reconstitution. Aβ_1–42 was reconstituted in cold HFIP to a concentration of 4 mg/mL and incubated on ice for 60 min. Following incubation, Aβ_1–42 was aliquoted and left uncovered overnight at RT to allow HFIP to evaporate, resulting in a protein film. Protein films were stored at −80°C until use. For experiments, fresh oligomers were prepared by reconstituting the protein film with DMSO to a concentration of 1.5 mM and then further diluting to 15 µM with 12 mM phosphate buffer (pH 7.4) containing 1 µM NaCl to initiate oligomerization. Following 30 min oligomerization at RT, oligomers were further diluted and immediately used for experiments. All manipulation of Aβ oligomer was conducted using low-protein-binding tubes and pipette tips.

Characterization of Aβ_1–42 oligomers

Oligomer size was assessed using atomic force microscopy (AFM), dynamic light scattering (DLS), and SDS-PAGE with western blotting. For AFM, freshly cleaved ruby muscovite mica was modified with 0.01% v/v (3-aminopropyl)triethoxysilane to enhance protein attachment to the surface. After surface silanization, 50 μL of freshly prepared Aβ_1–42 oligomers (15 μM) were incubated on the mica surface for 10 min. Samples were rinsed with filtered deionized water, allowed to air dry, and stored desiccated until imaging. AFM images were acquired using a Table Top (TT-2) AFM (AFMWorkshop, Hilton Head, SC) under standard laboratory conditions without temperature or environmental control. For the AFM probe, a new cantilever was used in tapping mode. Sections were measured as 2 μm × 2 μm (512 pixels × 512 pixels) sections at 1 Hz frequency. Particle heights were obtained as a probability density function using the open-source Gwyddion software (version 2.59). Probability density functions from nine images were combined to provide a composite profile for each oligomer preparation.

For DLS, Aβ_1–42 oligomer preparations (15 µM) were assessed via hydrodynamic radius (R_H) measurements using a Zetasizer Nano ZS system (Malvern Instruments Ltd., Worcestershire, UK). Light scattering intensity at a 90° angle was evaluated at RT using a 60-s acquisition time. Intensity fluctuations generated by Brownian motion yield translational diffusion coefficients, and R_H was determined using the Stokes–Einstein relationship. The ZS Xplorer Software was used to generate histograms of the particle number percentage versus R_H. Histograms represent an average of three acquisitions per sample.

For SDS-PAGE with western blotting, Aβ_1–42 oligomers (15 μM) mixed 1:1 with Laemilli buffer were separated (120 V, 1.5 h) on a 4–20% Tris-glycine gel (Bio-Rad, Hercules, CA). Size determination was facilitated using Precision Plus Western C standard. The samples were transferred (14 V, 15 min) to a nitrocellulose membrane (Bio-Rad) using semi-dry transfer methods. The membrane was then blocked overnight with blocking buffer. Blots were incubated with primary 6E10 monoclonal antibody (1:2000) for 1 h followed by the secondary HRP anti-mouse IgG antibody (1:2000) and Western C HRP for 45 min. Each incubation was followed with three washes with phosphate buffered saline containing tween-20. Blots were incubated with chemiluminescent reagents for 2 min then immediately imaged using chemiluminescence on an iBright Imaging System (Thermo Fisher Scientific, Waltham, MA).

Cell culture and treatment

Endothelial monolayers were comprised of primary HBMVECs. Cells were purchased from ScienCell at passage 1. Cell cultures were maintained on surfaces coated with 12 μg/mL bovine plasma fibronectin and sustained in Complete Endothelial Cell Medium containing penicillin (100 Units/mL) and streptomycin (100 mg/mL), 5% FBS, and endothelial cell growth supplement (1x) (referred to hereafter as CECM) at 37°C in a humid atmosphere of 5% CO₂/95% air. HBMVECs were used for experiments between passages 6 and 10.

For experiments that probe cellular barrier properties, HBMVECs were seeded at a density of 50,000 cells/well into a 6.5 mm diameter polycarbonate membrane Transwell insert with 8.0 μm pore size (Costar, Corning, Inc., Corning, NY) assembled into 24-well plates. For cell viability assays, HBMVECs were seeded at 29,000 cells/well into a polystyrene 96 well plate (Costar, Corning, Inc.). Prior to seeding, both surfaces were coated with 12 µg/mL bovine plasma fibronectin and incubated for at least 2 h in a humid atmosphere of 5% CO₂/95% air. For the former, wells were wetted with 250 µL CECM supplemented with 550 nM hydrocortisone to promote barrier properties.²³ HBMVEC monolayers were sustained for 3 d (5% CO₂/95% air) in CECM supplemented with 550 nM hydrocortisone. Beginning on day 3, the medium serum content was reduced to 1% FBS, and endothelial cell growth supplement was removed.

Upon confluence, exemplified by a plateau in TEER (8–10 d) (Figure 1), monolayers were treated apically (48 h) with 1 pM–2 µM Aβ preparations (isolated Aβ_1–40 monomer or Aβ_1–42 oligomers), 0.01–1.0 mM H₂O₂, or 1–10 pM Aβ_1–42 oligomer in the presence of 0.01 mM H₂O₂. Parallel treatment with an equivalent dilution of buffer in media served as a negative control, while treatment with 20 Units/well TNF-α served as a positive control.

Figure 1.

Confluent monolayers exhibit a plateau in TEER. HBMVECs seeded onto Transwell inserts were monitored via TEER every 24 h post-seeding. Error bars represent SEM from three independent experiments performed with 7–10 replicates.

Cellular barrier properties

Cellular barrier properties were monitored via TEER, which indicates the resistance to ion diffusion across a cell monolayer. Measurements were acquired using an Endohm chamber (World Precision Instruments, Sarasota, FL), wherein a pair of concentric electrodes incorporating a voltage-sensing Ag/AgCl pellet in the center and an annular current electrode are positioned on both sides of the cells grown on a Transwell insert.²⁴ A ±10 µA square wave alternating current and a 12.5 Hz signal were applied to the monolayer using an epithelial voltohmmeter (EVOM2, World Precision Instruments, Sarasota, FL). The current and voltage across the cell layer were measured and output as the total resistance (R_total).²⁵ To obtain the resistance of the monolayer (R_cells), the resistance across a Transwell absent of cells (R_insert) was subtracted from R_total. To calculate an intrinsic value for monolayer resistance in Ω·cm², the resistance measurement was multiplied by the area of the Transwell insert (0.336 cm²). Values are reported as the fraction of TEER observed pretreatment (TEER retained), and results are presented as the mean ± SEM.

XTT

Cell viability was assessed via the reduction of XTT by nicotinamide adenine dinucleotide (NADH) produced in the mitochondria via trans-plasma membrane electron transport and an electron mediator. Following treatment, cells were incubated with a 1:1 mixture of electron mediator solution and XTT developer reagent for 2 h at 37°C (5% CO₂/95% air). The absorbance (λ = 450 nm) of each sample was evaluated using a BioTek Synergy 2 microplate reader (Agilent, Santa Clara, CA). Cell viability is expressed as a percentage of the negative control.

Statistical analysis

Statistical analysis was performed via one-way ANOVA or two-way ANOVA using GraphPad Prism 9.0 software (San Diego, CA). For treatments with Aβ or H₂O₂ alone, statistical significance was analyzed via one-way ANOVA, and comparisons vs. the negative control were performed using Dunnett’s multiple comparisons. For simultaneous H₂O₂ and Aβ_1–42 oligomer treatments, the statistical significance was analyzed via two-way ANOVA, and multiple comparisons between treatment types and vs. 0 pM Aβ_1–42 oligomers were performed using Šídák’s multiple comparisons.

Results

Aβ_1–42 oligomer preparations reproducibly contain polydisperse oligomers

The size and distribution profile of freshly prepared Aβ_1–42 oligomers at a concentration of 15 µM were evaluated using AFM, DLS, and SDS-PAGE with western blotting. AFM measurements illustrate the distribution of oligomer heights ranging from 0.2 nm to 14.6 nm, with average oligomer heights of 3.8 nm and 3.0 nm in two independent oligomer preparations (Figure 2(A,B)). DLS reflects the presence of oligomers with a polydispersity in R_H from 27.6 nm to 229 nm (Figure 2(C)) and an average R_H of 100 nm and 148 nm in two independent oligomer preparations. SDS-PAGE with western blot demonstrates the presence of monomer, trimer, and tetramer as well as oligomers ranging in molecular weight from 37 kDa to 250 kDa (Figure 2(D)). These size ranges are in agreement with other studies that have used AFM,²⁶ DLS,²⁷ and SDS-PAGE with western blot²⁸ to characterize oligomer size. Together, these characterizations illustrate reproducible oligomer preparations containing some residual monomer.

Figure 2.

Aβ oligomers exhibit a reproducible, polydisperse population. Aβ_1–42 oligomers (15 µM) were prepared using a 30-min oligomerization in 12 mM phosphate buffer containing 1 µM NaCl. Oligomer size was assessed via AFM (panels A,B), DLS (panel C), and SDS-PAGE with western blot (panel D).

Physiological concentrations of Aβ_1–42 oligomers decrease TEER

To ascertain the effect of Aβ_1–42 oligomers on endothelial permeability, confluent HBMVEC monolayers were treated with Aβ_1–42 oligomers at concentrations of 1–1000 nM. Following 48-h treatment, TEER decreased in the presence of all oligomer concentrations (Figure 3). These changes were previously shown to parallel an increase in permeability of BSA and a re-localization of tight junction proteins away from cell borders.⁷ The observed decrease was less pronounced than that induced by 20 Units/mL TNF-α (Figure 3, dashed line) and was not accompanied by cellular toxicity (Figure 4(B)). In contrast, TEER values remained constant for confluent HBMVEC monolayers treated with Aβ_1–40 monomer at concentrations of 0–2000 nM (Figure 3). These results demonstrate that Aβ oligomers compromise the barrier properties of HBMVEC monolayers, while Aβ monomer is inert.

Figure 3.

Aβ_1–42 oligomers decrease TEER, while Aβ monomer renders TEER unchanged. Confluent HBMVEC monolayers were incubated with Aβ_1–42 oligomers (closed circle) or SEC-isolated Aβ_1–40 monomer (open triangle) at concentrations of 1–2000 nM. Monolayers maintained in the absence of Aβ served as a negative control (dotted line), while incubation with 20 Units/mL TNF-α served as a positive control (dashed line). TEER was measured 48 h after treatment. Values are reported as the fraction of TEER observed prior to treatment. Error bars represent SEM from 3–6 independent experiments performed with 3–4 replicates; some error bars lie within symbols. ***p < 0.001; **p < 0.005; *p < 0.05 vs. negative control.

Figure 4.

Picomolar concentrations of Aβ oligomers compromise TEER without altering cell viability. A) confluent HBMVEC monolayers were incubated alone (0 pM, negative control), with 1–10⁶ pM Aβ_1–42 oligomers, or with 20 Units/mL of TNF-α (positive control). TEER was measured (panel A) and viability was assessed (panel B) following 48 h. Results are reported as a fraction of TEER retained prior to treatment (panel A) and percent viability relative to the negative control, represented by 100% viability (panel B). Error bars indicate SEM from 3–6 independent experiments performed with 3–4 replicates. ****p < 0.0001; ***p < 0.0005; **p < 0.01.

To explore whether Aβ_1–42 oligomers can induce changes in endothelial permeability at physiological concentrations, confluent HBMVEC monolayers were treated with picomolar concentrations Aβ_1–42 oligomers for 48 h. TEER was significantly reduced following treatment with 1–1000 pM Aβ_1–42 oligomers (Figure 4(A)). In fact, these decreases were slightly more pronounced than those observed at nanomolar concentrations, but still less pronounced than that induced by 20 Units/mL TNF-α. Again, these changes were not accompanied by cellular toxicity (Figure 4(B)). These results demonstrate oligomer-induced barrier breakdown at physiological Aβ oligomer concentrations.

Oxidative stress exacerbates the effect of Aβ_1–42 oligomers on TEER

To examine whether low levels of oxidative stress can augment the reduction of TEER by Aβ oligomers, H₂O₂ treatment, documented to induce oxidative stress in endothelial cells,^29–31 was introduced. An H₂O₂ concentration of 0.01 mM was selected for this treatment, as this H₂O₂ concentration did not alter either TEER (Figure 5(A)) or cell survival (Figure 5(B)) following 24 h incubation. In contrast, H₂O₂ treatment induced a significant, concentration-dependent decrease in TEER at 0.1 mM and 1 mM H₂O₂, with a 10% and 30% reduction, respectively (Figure 5(A)); although, these changes were not paralleled by a decrease in cellular toxicity (Figure 5(B)).

Figure 5.

Low concentrations of H₂O₂ do not compromise TEER. Confluent HBMVEC monolayers were incubated alone (0 mM, negative control) or with 0.01, 0.1, or 1 mM H₂O₂. TEER was measured (panel A) and viability was assessed (panel B) following 24 h. Results are reported as a fraction of TEER retained prior to treatment (panel A) and percent viability relative to the negative control, represented by 100% viability (panel B). Error bars indicate SEM from 2–3 independent experiments performed with 3–4 replicates. ****p < 0.0001; **p < 0.005.

When HBMVECs were treated simultaneously with Aβ_1–42 oligomers and H₂O₂, monolayers exhibited a reduction in TEER more pronounced than parallel treatments with oligomers alone (Figure 6(A)). This augmentation of barrier breakdown was observed at both 1 pM and 10 pM oligomer concentrations. Additionally, simultaneous treatment with Aβ_1–42 oligomers and H₂O₂ exhibited a decrease in TEER compared to monolayers treated with H₂O₂ alone. These TEER changes were observed in the absence of any decrease in cell survival (Figure 6(B)). These results demonstrate the ability of H₂O₂-associated oxidative stress to exacerbate decreases in TEER induced by Aβ oligomers.

Figure 6.

H₂O₂ exacerbates Aβ_1–42 oligomer-induced decreases in TEER. Confluent HBMVEC monolayers were incubated alone (0 pM, negative control) or treated with 1 or 10 pM oligomers alone (black bars) or simultaneously with 0.01 mM H₂O₂ (white bars). TEER was measured (panel A) and viability was assessed (panel B) following 48 h. Results are reported as a fraction of TEER retained prior to treatment (panel A) and percent viability relative to the negative control, represented by 100% viability (panel B). Error bars indicate SEM from 3-4 independent experiments performed with 4 replicates (panel A) and 2 replicates (panel B). ****p < 0.0001; *p < 0.05.

Discussion

It is estimated that pathogenic changes in the AD brain initiate up to 20 y before symptoms arise.³² The earliest aggregates, oligomers, are reported to disrupt and alter multiple physiological functions of cells in and around the central nervous system. While reports on the effect of Aβ oligomers on neuronal cell function are robust, the action of Aβ oligomers is less explored at physiological concentrations in cerebrovascular cells. Yet, Aβ deposition around the cerebrovasculature is observed in over 80% of AD cases.³³ Previous studies on vascular cell lines treated with higher than physiological Aβ oligomer concentrations revealed that low micromolar concentrations of Aβ oligomers trigger apoptotic pathways in endothelial cells through upregulation of nuclear factor-kappa B (NF-KB), TNF-α, interlukin-6, caspase-3, the superoxide dismutase 2, and the tumor protein p53.²⁰ At high micromolar concentrations, oligomers activate endothelial caspase-8 and −9.³⁴ The current study explored the effect of oligomeric Aβ at physiological concentrations on the endothelial cells that encompass the cerebrovasculature. Our lab previously reported that the smallest Aβ aggregates are the most damaging to the BBB by demonstrating an inverse relationship between aggregate size and increased permeability across HBMVEC monolayers.⁷ In this report, we demonstrate that physiological, picomolar concentrations of Aβ oligomers reduce TEER across HBMVEC monolayers, and this effect is enhanced in the presence of oxidative stress.

Aβ_1–42 oligomers significantly reduced TEER across HBMVEC monolayers at concentrations as low as 1 pM (Figures 3,4(A)), and this effect was observed in the absence of cell toxicity (Figure 4(B)). In contrast, monomeric Aβ_1–40 failed to elicit any change in TEER (Figure 3) even at low micromolar concentrations. These results agree with other studies that have demonstrated that Aβ aggregates impact barrier properties. Cocultures of mouse bEnd.3 endothelial cells and C8-D1 astrocytes treated with micromolar concentrations of Aβ_1–42 displayed increased permeability, reduced TEER, and diminished expression of tight junction proteins at cellular borders.³⁵ Similar results are reported in primary human brain microvascular endothelial cells treated with micromolar concentrations of Aβ_1–40 aggregates.⁷ The current study, however, is the first to report this response induced by oligomeric forms of Aβ and at physiological Aβ concentrations. Our findings also extend observations of the pathogenic effects of physiological oligomer concentrations in other cell types. For example, nanomolar concentrations of oligomeric Aβ have been shown to impair insulin signaling in human neuroblastoma cells,³⁶ to induce toxicity in mouse brain slices,³⁷ and to activate primary mouse microglia.³⁸

The observed Aβ oligomer-induced reduction in TEER is likely caused by increases in both paracellular permeability and transmembrane ionic flux. Aβ aggregates can alter the expression and localization of tight junction proteins, including zonula occludens proteins,^7,8 claudins,⁹ and occludin,^8,9 which increases paracellular permeability. Aβ aggregates can also induce NF-κB activation and nuclear translocation,³⁹ which is known to disrupt the expression of endothelial cell surface adhesion molecules³⁹ and mediate increases in monolayer permeability.³⁹ Additionally, Aβ oligomers activate caspase-8 signaling, which engages mitochondrial pathways inducing caspase-9 activation resulting in vascular injury.^34,40 Aβ oligomers can also alter transmembrane permeability by forming pores in bilayer membranes.⁴¹ Computational simulations align with these results and further suggest that Aβ tetramer and octamer structures induce membrane disruption, in which water permeation occurs through lipid stabilized pores mediated by the β-sheet edges of oligomers.⁴² In neuronal cells, Aβ oligomers affect calcium release in cortical neurons.¹⁹ Membrane pores formed by Aβ oligomers can be calcium permeable and can thus disrupt Ca²⁺ homeostasis by rapidly elevating intracellular Ca²⁺ concentrations.^43,44 In fact, in the bovine endothelial cell line KOM-1, nanomolar concentrations of Aβ_1–42 increased intracellular concentrations of Ca^2+,45 indicative of a rise in BBB permeability.

Oxidative stress can also damage the endothelial cells of the BBB, leading to activation of inflammatory pathways and cell death, further exacerbating brain damage.^46–48 Elevated ROS contribute to endothelial dysfunction⁴⁹ and can lead to increased permeability of the BBB, increasing the neurodegenerative process.⁵⁰ Human-induced pluripotent stem cell-derived brain microvascular endothelial-like cells exposed to acute or chronic doses of 1 and 10 mM H₂O₂ displayed concentration-dependent local, discrete structural changes with defects such as delaminations, focal leaks, and combined focal leaks/delaminations, with the latter being the most prevalent during acute exposure.³⁰ Thus, the incorporation of oxidative stress within cell culture studies exploring the effect of Aβ on the BBB would better represent the physiological disease state.

A low level of oxidative stress introduced by the treatment of HBMVEC monolayers with 0.01 mM H₂O₂ led to a more pronounced oligomer-induced reduction of TEER (Figure 5(A)) in the absence of cell toxicity (Figure 6(A)). This effect was observed at an H₂O₂ concentration that alone had no effect on TEER (Figure 5(A)), in agreement with other studies reporting no disruption of endothelial cell tight junction proteins at low concentration of H₂O₂.⁴⁶ These results highlight the interplay among oxidative stress, Aβ, and endothelial cell function in AD. Such interplay was also observed when HUVEC cultures treated with aggregated Aβ displayed an increase in semicarbazide-sensitive amine oxidase, an enzyme that metabolizes primary amines to generate ROS, leading to vascular dysfunction.⁵¹ A similar interplay has also been observed for neuronal cell function. Aβ_1–42 treatment of primary mouse and hippocampal neurons provoked a dampening in the oscillation of metabolic ATP levels as well as an effect on mitochondrial respiration, leading to an increased oxidized state.⁵² Furthermore, incubating neurons with Aβ_1–42 leads to lipid peroxidation resulting from H₂O₂ as marked by protein-bound 4-hydroxy-2-trans-nonenal.⁵³

Conclusion

This work demonstrates that Aβ oligomers reduce the electrical resistance across HBMVEC monolayers, while monomer is inert. Moreover, this response is observed at picomolar Aβ concentrations, which reflect physiological levels. The impact of Aβ oligomers on barrier function is exacerbated in the presence of low concentrations of H₂O₂. Thus, AD-related BBB breakdown may be caused by a culmination of multiple pathogenic effects, and understanding the interplay between oxidative stress and Aβ is crucial in elucidating AD pathogenesis. These findings further imply that strategies targeting Aβ oligomers as well as ROS production may be potential therapeutic approaches for AD.

Funding Statement

This work was supported by the Support to Promote Advancement of Research and Creativity (SPARC) Graduate Research Grant from the University of South Carolina to Brittany E. Watson

Disclosure statement

No potential conflict of interest was reported by the author(s).