Alterations of endocannabinoid signaling and microglia reactivity in the retinas of AD-like mice precede the onset of hippocampal β-amyloid plaques

Abstract

Extra-cerebral manifestations of Alzheimer's disease (AD) develop in the retina, which is, therefore, considered a "window to the brain". Recent studies demonstrated the dysregulation of the endocannabinoid (eCB) system (ECS) in AD brain. Here, we explored the possible alterations of ECS and the onset of gliosis in the retina of AD-like mice. Tg2576 (TG) mice overexpressing the amyloid precursor protein (APP) were used at the age of 12 months, when hippocampal β-amyloid plaques had not been developed yet. Analysis of retinal gliosis showed a significant increase in the number of IBA1 (+) microglia cells in TG versus wild type (WT). Gliosis was not associated with retinal β-amyloid plaques, evident retinal degenerative signatures, or excitotoxicity; instead, oxidative stress burden was observed as increased acrolein levels. Analysis of the ECS (receptors/metabolic enzymes) through western blotting (WB) revealed the up-regulation of cannabinoid receptor 2 (CB₂) and monoacylglycerol lipase (MAGL), the enzyme responsible for the degradation of 2-arachidonoylglycerol (2-AG), in TG retinas. Fluorescence intensity analysis of anti-CB₂ and anti-MAGL immuno-stained cryosections was consistent with WB, showing their up-regulation throughout the retinal layers. No statistically significant differences were found for the other enzymes/receptors of the ECS under study. However, linear regression analysis for individual animals showed a significant correlation between CB₂ and fatty acid amide hydrolase (FAAH), diacylglycerol lipase α/β (DAGLα/β), and APP; instead, a significant negative correlation was found between MAGL and APP. Finally, ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) demonstrated a significant reduction of 2-AG in TG retinas (~0.34 ng/mg) compared to WT (~1.70 ng/mg), while a trend toward increase was found for the other eCB anandamide (AEA). Overall, our data indicate that gliosis and ECS dysregulation-in particular of CB₂, MAGL and 2-AG-occur in the retina of AD-like mice before retinal degeneration and development of hippocampal β-amyloid plaques.

Keywords: Alzheimer's disease; bioactive lipids; endocannabinoids; neuroinflammation; retina.

FIGURE 1

Immunohistochemistry of β‐amyloid in WT and TG brains. Representative brain cryosections of WT and TG mice stained with anti‐β‐amyloid immunohistochemistry (IHC). (a) 12‐month‐old WT, (b) 12‐month‐old TG, (b’) high magnification of cortex, (b”) high magnification of hippocampus. (c) 24‐month‐old WT, (d) 24‐month‐old TG, (d’) high magnification of cortex, (d”) high magnification of hippocampus. Scale bars: 900 and 300 μm. The black arrows indicate the β‐amyloid plaques. (e, f) Representative eye cryosections stained with anti‐β‐amyloid IHC in WT and TG mice respectively. Scale bar: 400 μm. (g) Schematic representation of ocular tissues visible in images e and f.

FIGURE 2

Analysis of gliosis. (a) Representative confocal images of WT and TG retinal cryosections immunostained with anti‐IBA‐1 to identify microglia cells. The white arrows indicate the IBA‐1 positive cells. Scale bar: 50 μm. (b) Representative confocal images of WT and TG retinal cryosections immunostained with anti‐GFAP to identify astrocytes and Müller cells reactivity, with corresponding plot profile graphs showing GFAP fluorescence throughout the retinal layers. Scale bar: 50 μm. (c) Quantification of IBA‐1 (+) cells counted on retinal cryosections of all experimental groups. Data are shown as mean ± SEM. The black dots indicate individual sample values (n = 4–5). (d) Quantification of GFAP fluorescence intensity on retinal cryosections of all experimental groups. Data are shown as mean ± SEM. The black dots indicate individual sample values(n = 4–5). (e) Representative confocal images of WT and TG retinal cryosections immunostained with anti‐Acrolein to investigate oxidative stress. The black box highlights the Acrolein chemical structure. Scale bar: 50 μm. (e) Quantification of Acrolein fluorescence intensity on retinal cryosections of all experimental groups. Data are shown as mean ± SEM. The black dots indicate individual sample values. Scale bar: 50 μm. Details of the statistics are reported in Table S1. For abbreviations see the abbreviations list.

FIGURE 3

Retinal thickness. Analysis of ONL (a) and INL (b) thickness from ventral to dorsal direction on retinal cryosections crossing the optic nerve (ON). (c) Representative fluorescence images of WT and TG retinal cryosections stained with bisbenzimide nuclear dye (blue). Measurements are expressed as the ratio of ONL/total retina thickness and INL/total retinal thickness calculated along the vertical meridian. Scale bar: 50 μm. Data are shown as mean ± SEM (n = 4–5). For abbreviations see the abbreviations list. Details of the statistical analyses are reported in Table S1.

FIGURE 4

Retinal expression of ECS components. Western blot analysis of CB₁ (a), CB₂ (b), TRPV1 (c), DAGLα (d), DAGLβ (e), MAGL (f), NAPE‐PLD (g), and FAAH (h) in the retinas of the two experimental groups. Molecular weights (MW) are shown on the right‐hand side. Data are shown as mean ± SEM (n = 4–10). The black dots indicate individual samples. Details of the statistics are reported in Table S1. Original whole western blot bands are reported in Figures S1–S8. Some proteins were probed on the same membranes and, therefore, the GAPDH housekeeping band is similar within each respective pair of figures (Table S4). For abbreviations see the abbreviations list.

FIGURE 5

CB₂ and MAGL immunostaining. (a) Confocal images of cryosections immunolabeled for CB₂ (green) from WT and TG retinas with relative plot profile graphs showing fluorescence intensity through the retinal layers. (b) Quantification of CB₂ fluorescence intensity. (c) Confocal images of cryosections immunolabeled for MAGL (green) from WT and TG retinas with relative plot profile graphs showing fluorescence intensity through the retinal layers. (d) Quantification of MAGL fluorescence intensity. Scale bar: 50 μm. Data are shown as mean ± SEM (n = 3–5). The black dots indicate individual samples. Details of the statistics are reported in Table S1. For abbreviations see the abbreviations list.

FIGURE 6

Endogenous levels of AEA and 2‐AG in the retinas of WT and TG mice. (a) The figure shows UPLC‐MS/MS analysis of AEA and 2‐AG in 12‐month‐old TG and WT mice. The dots indicate individual samples (n = 3–4). (b) Chemical structures of AEA and 2‐AG. (c) Heatmap showing AEA levels in 3‐, 6‐ and 12‐month‐old retinas of WT and TG; (d) heatmap showing 2‐AG levels in 3‐, 6‐, and 12‐month‐old retinas of WT and TG. The numbers on the heatmaps represent mean ± SEM (n = 3–4). Details of the statistics are reported in Table S1. For abbreviations see the abbreviations list.

FIGURE 7

Scatter plot of ECS receptors/enzymes and APP levels in individual animals. (a) DAGLα versus CB₂, (b) DAGLβ versus CB₂, (c) FAAH versus CB₂, (d) CB₂ versus APP, (e) MAGL versus APP. Data points indicate individual animals; the black line indicates the linear regression line. For abbreviations see the abbreviations list.