Amyloid β oligomer induces cerebral vasculopathy via pericyte-mediated endothelial dysfunction

Abstract

Background: Although abnormal accumulation of amyloid beta (Aβ) protein is thought to be the main cause of Alzheimer's disease (AD), emerging evidence suggests a pivotal vascular contribution to AD. Aberrant amyloid β induces neurovascular dysfunction, leading to changes in the morphology and function of the microvasculature. However, little is known about the underlying mechanisms between Aβ deposition and vascular injuries. Recent studies have revealed that pericytes play a substantial role in the vasculopathy of AD. Additional research is imperative to attain a more comprehensive understanding.

Methods: Two-photon microscopy and laser speckle imaging were used to examine cerebrovascular dysfunction. Aβ oligomer stereotactic injection model was established to explain the relationship between Aβ and vasculopathy. Immunofluorescence staining, western blot, and real-time PCR were applied to detect the morphological and molecular alternations of pericytes. Primary cultured pericytes and bEnd.3 cells were employed to explore the underlying mechanisms.

Results: Vasculopathy including BBB damage, hypoperfusion, and low vessel density were found in the cortex of 8 to 10-month-old 5xFAD mice. A similar phenomenon accompanied by pericyte degeneration appeared in an Aβ-injected model, suggesting a direct relationship between Aβ and vascular dysfunction. Pericytes showed impaired features including low PDGFRβ expression and increased pro-inflammatory chemokines secretion under the administration of Aβ in vitro, of which supernatant cultured with bEND.3 cells led to significant endothelial dysfunction characterized by TJ protein deficiency.

Conclusions: Our results provide new insights into the pathogenic mechanism underlying Aβ-induced vasculopathy. Targeting pericyte therapies are promising to ameliorate vascular dysfunction in AD.

Keywords: Alzheimer’s disease; Aβ oligomer; Blood–brain barrier (BBB); Pericytes; Tight junction proteins.

Fig. 1

Vasculopathy was observed in the brains of 8-10mo 5xFAD mice. A Representative two-photon images showed cerebral vasculature and BBB breakdown in the somatosensory cortex of 8-10mo 5xFAD mice. 70kD Rhodamine-dextran was shown in red and 4kD FITC-dextran dye was in green. Scale bar: 100 μm B. Variation curve of the total extravascular 4kD FITC-dextran signal intensity after retro-orbital injection. C Volumetric quantification of FITC-dextran extravasation 30 min after injection. D-E Quantitative analysis of total vessel coverage area and length in the cortex of both groups. F Representative laser speckle images demonstrated the situation of cerebral perfusion. G Quantification of cerebral perfusion in ROI. All the data were presented as mean ± SEM and analyzed using unpaired t test. n = 6 for each group. * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001

Fig. 2

Vascular PDGFRβ deficiency was found in the brains of 8-10mo 5xFAD mice. A Representative immunofluorescent staining images of PDGFRβ (red), Aβ (green) and lectin (grey) co-labeling in the hippocampus (upper panel) and cortex (low panel) of 8-10mo wild type (n = 6) and 5xFAD (n = 8) mice. Scale bar: 50 μm B. Quantification of Aβ-positive area in both groups. C Quantification of PDGFRβ-positive coverage area. D Pearson’s coefficient (r) correlation analysis between the percentage of PDGFRβ⁺ area and Aβ.⁺ area in the cortex of AD mice. E Quantitative analysis of PDGFRβ mRNA levels. n = 4 for each group. F-G Western blotting analysis of PDGFRβ protein expression level in the hippocampus (upper panel) and cortex (low panel) of wild type (n = 4) and 5xFAD mice (n = 6). Data are shown as mean ± SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001

Fig. 3

Aβ oligomer stereotactic injection induced BBB leakage and local cortex hypoperfusion in 2mo WT mice. A Schematic diagram of the experiment design. B Representative in vivo two-photon images of the cortex of mice in control (n = 4), vehicle (2%DMSO) (n = 6), and Aβo-injected (n = 7) groups. 2–3 stacks were acquired to obtain the mean for each animal. Scale bar: 100 μm C. Variation curve of extravasated FITC-dextran dye intensity in these three groups after injection. D Quantitative analysis of FITC-dextran extravasation 30 min after injection. E–F Quantification of cortical vascular length and area from the three experimental groups. G Representative laser speckle images of DMSO (n = 8) and Aβo (n = 11) groups. H Quantification of D-value of cerebral perfusion between injection region (ROI 1) and contralateral cortex (ROI 2) in DMSO and Aβo groups. D-value was defined as the ROI 1 quantitative perfusion value minus ROI 2. All the data were presented as mean ± SEM. * p < 0.05, ** p < 0.01, and *** p < 0.001

Fig. 4

The level of NG2 expression decreased after Aβo injection in NG2-DsRed mice. A Representative images of coronal brain section of the vehicle (2%DMSO) and Aβo-injected NG2-DsRed mice. The area 1 mm around the injection point (the arrow points) was defined as the injection region and the ipsilateral cortex area 1-4 mm away from the injection point was defined as the para-injection region. Scale bar: 250 μm B Representative immunofluorescent staining images of NG2 (red) and lectin (grey) in the injection region of both groups. Scale bar: 50 μm C Quantification of NG2-positive coverage length (upper panel) and area (low panel) in the injection region of both groups. D Representative immunofluorescent staining images of NG2 (red) and lectin (grey) in the para-injection region of both groups. Scale bar: 50 μm E Quantification of NG2-positive coverage length (upper panel) and area (low panel) in the para-injection region of both groups. All the data were presented as mean ± SEM and analyzed using an unpaired t-test. n = 6 for each group. * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001

Fig. 5

Aβo injection inhibited PDGFRβ expression and induced an inflammatory response in PDGFRβ-cre: Ai9 mice. A Representative immunofluorescent staining images of PDGFRβ (red) and lectin (grey) in the injection region of vehicle (2%DMSO) and Aβo-injected NG2-DsRed mice. Scale bar: 20 μm B Quantification of the total length of lectin⁺ area. C-D Quantification of PDGFRβ.⁺ coverage length (left panel) and area (right panel) in the injection region of both groups. E–F Western blotting analysis of PDGFRβ protein expression level in brain tissues after 2%DMSO and Aβo injection. G QPCR analysis of PDGFRβ mRNA expression level in brain tissues after vehicle and Aβo injection. H QPCR analysis of mRNA levels of some classic inflammatory factors (IL6, IL10, TNFα, CCL2, CXCL10) in brain tissues. All the data were presented as mean ± SEM and analyzed using an unpaired t-test. n = 6 for each group. * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001

Fig. 6

Aβo-incubated pericytes demonstrated a proinflammatory profile and decreased tight-junction proteins in endothelial cells in vitro. A-B Western blotting analysis of PDGFRβ protein expression level in primary cultured pericytes incubated with Aβo for 24 h and 72 h. C Heat map summary of Aβo-stimulated genes in primary pericytes incubated with Aβo for 24 h and 72 h. D Schematic diagram of bEnd3 cells incubated with primary pericytes supernatant. E–F Western blotting analysis of tight junction protein (ZO-1, Claudin-5, and Occludin) expression levels in bEnd3 cells incubated with Aβo and primary pericytes supernatant. All the data were presented as mean ± SEM. * p < 0.05, ** p < 0.01