Impaired Glymphatic Function and Pulsation Alterations in a Mouse Model of Vascular Cognitive Impairment

Abstract

Large vessel disease and carotid stenosis are key mechanisms contributing to vascular cognitive impairment (VCI) and dementia. Our previous work, and that of others, using rodent models, demonstrated that bilateral common carotid stenosis (BCAS) leads to cognitive impairment via gradual deterioration of the neuro-glial-vascular unit and accumulation of amyloid-β (Aβ) protein. Since brain-wide drainage pathways (glymphatic) for waste clearance, including Aβ removal, have been implicated in the pathophysiology of VCI via glial mechanisms, we hypothesized that glymphatic function would be impaired in a BCAS model and exacerbated in the presence of Aβ. Male wild-type and Tg-SwDI (model of microvascular amyloid) mice were subjected to BCAS or sham surgery which led to a reduction in cerebral perfusion and impaired spatial learning acquisition and cognitive flexibility. After 3 months survival, glymphatic function was evaluated by cerebrospinal fluid (CSF) fluorescent tracer influx. We demonstrated that BCAS caused a marked regional reduction of CSF tracer influx in the dorsolateral cortex and CA1-DG molecular layer. In parallel to these changes increased reactive astrogliosis was observed post-BCAS. To further investigate the mechanisms that may lead to these changes, we measured the pulsation of cortical vessels. BCAS impaired vascular pulsation in pial arteries in WT and Tg-SwDI mice. Our findings show that BCAS influences VCI and that this is paralleled by impaired glymphatic drainage and reduced vascular pulsation. We propose that these additional targets need to be considered when treating VCI.

Keywords: amyloid-β (Aβ); carotid stenosis; cerebral amyloid angiopathy (CAA); glymphatic function; vascular cognitive impairment; vascular pulsation.

FIGURE 1

BCAS causes a decline in spatial learning acquisition and cognitive flexibility. Spatial learning and memory and cognitive flexibility were assessed using a Barnes maze at 3 months post-BCAS in WT and Tg-SwDI mice. (A) In the acquisition phase one hole (indicated by green) was designated as the target hole with an escape box. A probe test was performed 72 h after the last acquisition training session, in which the escape box was removed. In the reversal phase, the target hole was moved 180 degrees to the original target hole. 1 day after the probe test. A reversal probe test was performed 72 h after the last training session. (B) Motor ability assessment in acquisition training. The speed or velocity was measured at the outset to assess whether motor function was affected by the genotype or surgery across different groups. There was a significant effect of genotype between WT and Tg-SwDI mice. Post hoc tests showed significant difference between WT BCAS and Tg-SwDI BCAS mice at day 1 (^**p < 0.01), 3, 4 and 5 (*p < 0.05), respectively. (C) Spatial learning was assessed by comparing escape latency over 6 days with 2 sessions per day. There was a significant effect of BCAS surgery but not genotype across groups. Post hoc tests showed significant effect of BCAS in WT at day 5, 6 (*p < 0.05) and in Tg-SwDI at day 2 (^#p < 0.05) compared to their sham counterparts. (D) Pathlength measure was also used to evaluate spatial learning function. There was a significant effect of BCAS surgery but not genotype across groups. Post hoc tests showed significant effect of BCAS in WT mice when compared to their sham counterparts at day 1, 2, 5 (*p < 0.05) and 6 (^**p < 0.01). (E) In the acquisition 72 h probe test all mice performed above chance except WT BCAS mice (one sample t-test). No significant effect of BCAS and genotype were detected (Two-way ANOVA). To enhance the detection of spatial learning ability, reversal trials were taken to evaluate the ability of mice to learn a new location using Barnes maze (F–H). In the reversal tests, spatial learning was assessed by comparing escape latency and pathlength over 3 days with 2 sessions per day training across all groups. (F) There was a significant effect of BCAS surgery but not genotype by comparing escape latency. Post hoc tests showed significant effect of BCAS in WT mice when compared to their sham counterparts at day 2 (*p < 0.05). (G) By comparing the pathlength in the test, there was a significant effect of BCAS surgery but not genotype. Post hoc tests showed significant effect of BCAS in WT mice when compared to their sham counterparts at day 2 (p < 0.05). (H) In the reversal probe test only WT sham mice performed above chance (one sample t-test). There was no significant effect of either genotype or surgery on the percentage time spent in the correct quadrant (Two-way ANOVA). Data are mean ± SEM, n = 6–10 per group.

FIGURE 2

Decreased resting CBF following BCAS. MRI arterial spin labeling (ASL) was used to measure regional alterations in CBF. (A) Representative images of arterial spin labeling (ASL) from sham and BCAS WT and Tg-SwDI mice at 3 months following surgery. (B,C) A significant reduction of CBF in the brain cortex and hippocampus was determined post-BCAS but there was no genotype effect. * and ^** indicate p < 0.05 and 0.01, respectively. Data are presented as individual data points, mean ± SEM, n = 6–10 per group.

FIGURE 3

CSF tracer influx in the perivascular space. (A) Evans blue dye was injected into cisterna magna of a normal mouse. At the surface of the brain, dyes were found distributed along blood vessels, the middle cerebral artery (MCA) and its branches, along the superior sagittal sinus, inferior cerebral vein, and transverse sinus. (B) (Arteriole, in cortical area) and (D) Co-labeling of sections with the vascular basement membrane marker COL4 revealed the localization of CSF fluorescent tracer to the adjacent space. (C) (Capillary, in subcortical area) and (E) Tracer colocalized with the basement membrane. Representative images showing spatial location of tracer soluble lysine fixable dextran 3 kDa (D-3) (green). Scale bar: top (space) = 10 μm, bottom (lumen) = 5 μm.

FIGURE 4

Regional CSF tracer influx is altered in BCAS and Tg-SwDI mice. Representative images of fluorescent tracer influx (D-3) (green) in the (A) DL CTX and (B) hippocampus (CA1-DG molecular layer) of WT and Tg-SwDI mice sham and post-BCAS. (C,D) Quantification of D-3 tracer distribution in the DL CTX and CA1-DG molecular layer. * and ^** indicate p < 0.05 and 0.01, respectively. Data are shown as individual data points, mean ± SEM, n = 6–10 per group. Scale bar = 500 μm.

FIGURE 5

BCAS exacerbates amyloid deposition in Tg-SwDI mice. (A) Representative images of amyloid (green) and COL4 as a marker of vascular basement membranes (red) in the superficial brain cortex in Tg-SwDI sham and BCAS mice. (B) Total amyloid and (C) vascular amyloid were increased post-BCAS. (D) No significant changes of basement membrane were found. * indicates p < 0.05. Data are shown as individual data points, mean ± SEM, n = 6–9 per group. Scale bar = 50 μm.

FIGURE 6

Increased astrogliosis post-BCAS in cortex. Representative images of GFAP immunostaining to assess the degree of astrogliosis in the superficial brain cortex of WT and Tg-SwDI, sham and BCAS mice. (A,C) In the superficial cortex, BCAS caused increased astrogliosis but was unaffected in Tg-SwDI mice. (B,D) In the hippocampus, there was increased astrogliosis in Tg-SwDI mice but not post-BCAS. * and ^** indicate p < 0.05 and 0.01, respectively. Data are shown as individual data points, mean ± SEM, n = 6–10 per group. Scale bar = 100 μm.

FIGURE 7

Vascular pulsatility is reduced in Tg-SwDI animals (A) Two-photon microscopy was used to assess vessel pulsation in sham and post-BCAS WT and Tg-SwDI mice. Four categories of blood vessels were investigated: 1. Pial arteries; 2. Penetrating arteries; 3. Ascending veins; 4. Pial veins. (B) There was a significant effect of genotype on vascular pulsatility in all types of vessels measured, and significant effect of surgery on pial arteries. Post hoc test revealed a significant decrease in vascular pulsation between wild type and Tg-SwDI sham animals in pial veins (p = 0.000), pial arteries (p = 0.041), penetrating arteries (p = 0.001) as well as ascending veins (p = 0.002) Vascular pulsation was also reduced in Tg-SwDI when compared to wild type BCAS mice in pial veins (p = 0.000), penetrating arteries (p = 0.000), ascending veins (p = 0.000) but not in pial arteries (p > 0.05). Finally, vascular pulsation was decreased in wild type BCAS animals when compared to shams (p = 0.008). Data are presented as mean ± SEM, n = 6–7 per group. ^# indicates significant difference between sham and BCAS; * indicates significant difference between WT and Tg-SwDI. *p < 0.05, **, ^##p < 0.01; ***p < 0.001.

FIGURE 8

Proposed model by which BCAS and amyloid may impact on glymphatic function and predispose to VCI. In the long-term response to BCAS, reduced arterial pulsatility may impede CSF influx along the periarterial space and contribute in part to amyloid accumulation within the parenchyma and vasculature. Independently amyloid (Aβ40) has a profound impact on arterial pulsatility and impedes CSF influx. Downstream these events may influence glial cell responses including reactive astrogliosis and inflammation that can contribute to VCI.