Age dependence of retinal vascular plexus attenuation in the triple transgenic mouse model of Alzheimer's disease

Abstract

The influence of Alzheimer's disease (AD) progression and severity on the structural and functional integrity of the cerebral vasculature is well recognized. The retina is an extension of the brain; thus, changes in retinal vascular features may serve as markers of AD cerebrovascular pathologies. However, differentiating normal aging-versus AD-induced retinal vascular changes is unresolved. Therefore, we compared and quantified changes in superficial (SVP), intermediate (IVP), and deep (DVP) retinal vascular plexuses in young, middle-age, and old triple transgenic mouse model of AD (3xT-AD) to the changes that occur in age-matched controls (C57BL/6j). We used immunostaining combined with a novel tissue optical clearing approach along with a computational tool for quantitative analysis of vascular network alterations (vessel length and density) in SVP, IVP, and DVP. All three layers had comparable structural features and densities in young 3xTg-AD and control animals. In controls, IVP and DVP densities decreased with aging (-14% to -32% change from young to old, p < 0.05), while no changes were observed in SVP. In contrast, vascular parameters in the transgenic group decreased in all three layers with aging (-12% to -49% change from young to old, p < 0.05). Furthermore, in the old group, SVP and DVP vascular parameters were lower in the transgenics compared to age-matched controls (p < 0.05). Our analysis demonstrates that normal aging and progression of AD lead to various degrees of vascular alterations in the retina. Specifically, compared to normal aging, changes in vascular features of SVP and DVP regions of the retina are accelerated during AD progression. Considering recent advances in the field of depth-resolved imaging of retinal capillary network and microangiography, noninvasive quantitative monitoring of changes in retinal vascular network parameters of SVP and DVP may serve as markers for diagnosis and staging of Alzheimer's disease and discriminating AD-induced vascular attenuation from age-related vasculopathy.

Keywords: 3xTg-AD; Alzheimer's disease; Deep vascular plexus; Intermediate vascular plexus; Retina; Retinal vascular density; Superficial vascular plexus.

Fig. 1.. Amyloid-β and phosphorylated-tau deposition in brain sections.

Brain cortex, CA1, CA3, and dentate gyrus (DG) areas of young and old 3xTg-AD and old C57Bl/6j are immunostained for amyloid-β and phosphorylated tau (p-tau). Panel A demonstrates amyloid-β immunoreactivity (green) in old 3xTg-AD brain sections (middle column). Similarly, panel B demonstrates phosphorylated tau immunoreactivity (green) in old 3xTg-AD brain sections. Young 3xTg-AD and old C57Bl/6j sections did not stain with anti-amyloid-β (6E10) or anti-phosphorylated tau (AT180) antibodies (panels A and B). Cell nuclei are stained with DAPI (blue). Magnification: x20. White scale bar: 100 μm.

Fig. 2.. Image analysis of retinal vascular plexus.

(A) Retinal flat-mounts were stained for collagen IV. Eight non-overlapping regions of interest (ROI) from the midperipheral retina were Imaged (left). Tissue optical clearing allowed better visualization of intermediate and deep vascular plexuses for AngioTool analysis. A magnified view of all three layers from a representative ROI is shown on the right (10x, scale bar 1000 μm). Flat projections of each superficial, intermediate, and deep vascular plexuses were created with ZEN Lite, and the layers were analyzed using AngiTool. (B) Representative superficial (SVP), intermediate (IVP), and deep (DVP) vascular plexus views before (left) and after (right) AngioTool analysis. Vessel outlines and intersection points were highlighted. Only vessels with 5–29-pixel diameter were filtered for detection. (C) Intersections (thick arrow) and endpoints (thin arrow) were detected by AngioTool. Total vessel length is the sum of vessel lengths in each field.

Fig. 3.. Superficial vascular plexus (SVP) analysis.

(A) Representative images from SVP in young, middle-age, and old transgenic and control mice (20x, scale bar 50 μm). Bar graphs (B), (C), and (D) represent age-dependent changes within the transgenic and control animals. Bar graphs (E), (F), and (G) represent the total number of intersections, total vascular length, and the total number of endpoints comparison between young, middle age, and old transgenic and control animals. A significant difference (p < 0.05) is marked by a horizontal bar over the graphs. Data in graphs are mean ± SEM.

Fig. 4.. Intermediate vascular plexus (IVP) analysis.

(A) Representative images from IVP in young, middle-age, and old transgenic and control mice (20x, scale bar 50 μm). Bar graphs (B), (C), and (D) represent age-dependent changes within the transgenic compared to control animals. Bar graphs (E), (F), and (G) represent the total number of intersections, total vascular length, and the total number of endpoints comparison between young, middle age, and old transgenic and control animals. A significant difference (p < 0.05) is marked by a horizontal bar over the graphs. Data in graphs are mean ± SEM.

Fig. 5.. Deep vascular plexus (DVP) analysis.

(A) Representative images from DVP in young, middle-age, and old transgenic and control mice (20x, scale bar 50 μm). Bar graphs (B), (C), and (D) represent age-dependent changes within the transgenic and control animals. Bar graphs (E), (F), and (G) represent the total number of intersections, total vascular length, and the total number of endpoints comparison between young, middle age, and old transgenic and control animals. A significant difference (p < 0.05) is marked by a horizontal bar over the graphs. Data in graphs are mean ± SEM.