Regulation of astrocyte lipid metabolism and ApoE secretionby the microglial oxysterol, 25-hydroxycholesterol

Abstract

Neuroinflammation, a major hallmark of Alzheimer's disease and several other neurological and psychiatric disorders, is often associated with dysregulated cholesterol metabolism. Relative to homeostatic microglia, activated microglia express higher levels of Ch25h, an enzyme that hydroxylates cholesterol to produce 25-hydroxycholesterol (25HC). 25HC is an oxysterol with interesting immune roles stemming from its ability to regulate cholesterol metabolism. Since astrocytes synthesize cholesterol in the brain and transport it to other cells via ApoE-containing lipoproteins, we hypothesized that secreted 25HC from microglia may influence lipid metabolism as well as extracellular ApoE derived from astrocytes. Here, we show that astrocytes take up externally added 25HC and respond with altered lipid metabolism. Extracellular levels of ApoE lipoprotein particles increased after treatment of astrocytes with 25HC without an increase in Apoe mRNA expression. In mouse astrocytes-expressing human ApoE3 or ApoE4, 25HC promoted extracellular ApoE3 better than ApoE4. Increased extracellular ApoE was due to elevated efflux from increased Abca1 expression via LXRs as well as decreased lipoprotein reuptake from suppressed Ldlr expression via inhibition of SREBP. 25HC also suppressed expression of Srebf2, but not Srebf1, leading to reduced cholesterol synthesis in astrocytes without affecting fatty acid levels. We further show that 25HC promoted the activity of sterol-o-acyl transferase that led to a doubling of the amount of cholesteryl esters and their concomitant storage in lipid droplets. Our results demonstrate an important role for 25HC in regulating astrocyte lipid metabolism.

Keywords: 25-hydroxycholesterol; Alzheimer disease; apolipoprotein E; astrocyte; cholesterol metabolism; microglia; neuroinflammation; oxysterols.

Fig. 1

Microglia express Ch25h and secrete 25HC in response to LPS treatment. A: Comparison of Ch25h expression between mouse microglia and astrocytes by qPCR in response treatment with 100 ng/ml LPS for 24 h. Data are shown absolute copy numbers of Ch25h mRNA per 200,000 cells. A plot of CT values versus copy number is provided in supplemental Fig. S1. (P-values, ∗ < 0.05; two-way ANOVA) (B) Measurement of 25HC in conditioned culture media from microglia or astrocytes with or without treatment with 100 ng/ml LPS, as indicated. (P-values, ∗∗∗ <0.0002, Two-way ANOVA with Tukey test for multiple comparisons). C: 25HC production by wildtype (WT) and ch25h knockout (KO) microglia was measured in conditioned culture media after treatment with 100ng/ml LPS for various times as indicated. D: Comparison of amount of 25HC in cells versus media at 6 or 24 h after treatment of WT microglia with vehicle (control), 10 ng/ml LPS, or 100 ng/ml LPS. For direct comparison of the distribution of 25HC, cell extracts were prepared in a volume proportional to the volume of media used for cell culture (200,000 cells in 1 ml media). 25HC, 25-hydroxycholesterol; LPS, lipopolysaccharide; qPCR, quantitative qPCR.

Fig. 2

Uptake of 25HC by astrocytes. A: Fluorescence microscopy of astrocytes treated with 2 μM 25HC (containing 25HC (1.8 μM) and TopFluor-25HC (0.2 μM)) for various times. Time-dependent increase in fluorescence intensity of cells within intracellular vesicles was observed within 1 day B: On the left is a schematic of the experiment showing the addition of 25HC to a final concentration of 1 μM or 2 μM to astrocytes. Uptake of 25HC by astrocytes was monitored using LCMS to measure 25HC remaining (detected) in the culture media after 1 day of incubation on the right. 25HC, 25-hydroxycholesterol.

Fig. 3

25HC increases extracellular ApoE in astrocytes without altering Apoe mRNA expression. A: Astrocytes were incubated with 2 μM final concentration of 25HC, 7α25diHC, cholesterol, T0901317, or with an appropriate amount of vehicle (ethanol). ApoE was measured by ELISA in the conditioned media after 2 days (P-values ∗∗∗∗<0.0001, Two-way ANOVA with Sidak’s multiple comparisons test) (B) Effect of 25HC concentration and time on the accumulation of extracellular ApoE from astrocytes. C: Expression of Apoe mRNA was measured by qPCR in astrocytes treated with 2 μM each of cholesterol, 25HC or T0901317 or an appropriate amount of vehicle (ethanol) or without any treatment (none). Data was normalized to mouse Actb (actin B) as the endogenous control gene and shown as relative expression compared to “None.” (∗0.05–0.01; One-way ANOVA with Dunnett’s multiple comparisons test). D: Immunoblots of cell lysates from astrocytes treated with 0, 1, 2, or 5 μM 25HC for 48 h. Antibodies for ApoE, ApoJ, or actin are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction normalized for actin band intensity. E: Immunoblots of conditioned media from astrocytes treated with 0, 1, 2, or 5 μM 25HC for 48 h. Antibodies for ApoE or ApoJ are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction. F: Immunoblot for ApoE of conditioned media from astrocytes treated with 0, 1, 2, or 5 μM 25HC for 48 h and samples were run on native polyacrylamide gel electrophoresis. G: Extracellular cholesterol in serum-free conditioned media from astrocytes treated with vehicle (black bars) or 2 μM 25HC (grey bars). H: Extracellular ApoE in mouse astrocytes from knockin mice expressing ApoE3 or ApoE4. I: Expression of APOE was measured by qPCR in astrocytes treated with vehicle (ethanol) or 2 μM 25HC and normalized to mouse Actb (actin B) (P-values ∗ 0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Two-way ANOVA with Dunnett’s multiple comparisons test). 25HC, 25-hydroxycholesterol; qPCR, quantitative qPCR.

Fig. 4

25HC increases cholesterol efflux and decreases lipoprotein reuptake. A, B, Expression of Abca1 (A) and Ldlr (B) mRNA in response to treatment of astrocytes with vehicle, 25HC (2 μM) or T0901317 (2 μM) was assessed by qPCR. Data is normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to “vehicle.” (∗0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). C: Immunoblots of cell lysates from astrocytes treated with 0, 1, 2, or 5 μM 25HC for 48 h. Antibodies for Abca1, Ldlr or actin are shown on the left of corresponding blots. Bar graphs below the blots show band quantitation after baseline correction and normalized for actin band intensity. D: Schematic diagram showing the action of inhibitors to block the effect of 25HC on LXR and SREBP pathways. (E, F, and G) Expression of Abca1 (E), Ldlr (F), and Apoe (G) mRNA in response to treatment of astrocytes with vehicle or 25HC (2 μM) in the presence of vehicle (DMSO), 5 μM LY295427 or 5 μM GSK2033 was assessed by qPCR. Data is normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to “DMSO.” (∗0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). H: ApoE in the conditioned media of astrocytes treated with vehicle or 25HC (2 μM) in the presence of vehicle (DMSO), 5 μM LY295427 or 5 μM GSK2033 was measured by ELISA in the conditioned media after 2 days I: Fluorescence of BODIPY-cholesterol within astrocytes quantified by FACS at the uptake step and the efflux steps after treatment with vehicle (ethanol) or 2 μM 25HC for 24h. Lower fluorescence within cells indicate higher efflux. J: The amount of TopFluor-LDL uptaken by astrocytes treated with vehicle (ethanol) or 2 μM 25HC for 24h was measured by microscopy as the fluorescence sum area in the field and normalized to the total number of cells in the field. (∗0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). 25HC, 25-hydroxycholesterol; qPCR, quantitative qPCR.

Fig. 5

25HC suppresses sterol biosynthesis genes in astrocytes. Expression of Srebf2 (A), Insig1 (B), Srebf1 (C), and Fasn (D) mRNA in response to treatment of astrocytes with vehicle, 25HC (2 μM) or T0901317 (2 μM) for 1 day was assessed by qPCR. Data was normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to “vehicle”. E: Schematic showing selected steps of cholesterol biosynthesis and the enzyme involved. F: Expression of Acat2, Hmgcr, and Dhcr24 in response to treatment of astrocytes with vehicle (black bars) or 2 μM 25HC (grey bars) for 1 day was quantified by qPCR. Data was normalized to Actb (actin B) as the endogenous control gene and shown as relative expression compared to “vehicle”. Quantitation of total free fatty acids (G) and free cholesterol (H) in mouse astrocytes treated with vehicle (black bars) or with 2 μM 25HC (grey bars).(∗0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Two-way ANOVA with Sidak’s multiple comparisons test). 25HC, 25-hydroxycholesterol; qPCR, quantitative qPCR.

Fig. 6

25HC increases cholesteryl esters and lipid droplet accumulation in astrocytes. Quantitation of total cholesterol esters (A) in mouse astrocytes treated with vehicle (black bars) or with 2 μM 25HC (grey bars). Detailed analyses of cholesteryl esters (B) are shown. Expression of Soat1 (C) and Soat2 (D) in astrocytes treated with vehicle (black bars) or 2 μM 25HC (grey bars) was measured by qPCR and normalized to Actb as before. E: Mouse astrocytes treated with vehicle (ethanol), 2 μM 25HC, 30 μM OA or OA+25HC were stained for lipid droplets with BODIPY (green) and DAPI (blue). Top row of images is without the SOAT/ACAT inhibitor, avasimibe, and bottom row had 1 μM avasimibe. Bar = 100 μm. F, G, Quantification of the microscopy data in E and expressed as total LD counts/total nuclei. H, Extracellular ApoE levels quantified by ELISA. Light grey bars are without avasimibe and dark grey bars are with avasimibe. (∗0.05–0.01; ∗∗0.01–0.001; ∗∗∗<0.001; ∗∗∗∗<0.0001; Data compared with unpaired t tests in A. For all others two-way ANOVA with Sidak’s multiple comparisons test). SOAT, sterol-O-acyl transferase; LD, lipid droplet; 25HC, 25-hydroxycholesterol; qPCR, quantitative qPCR.

Fig. 7

Lipid metabolism and ApoE secretion in astrocytes is modulated by 25HC, a cholesterol metabolite secreted by activated microglia. Top: Overexpression of cholesterol 25-hydroxylase (Ch25h) in microglia activated by inflammatory stimuli converts cholesterol to 25-hydroxycholesterol (25HC). Paracrine effects of secreted 25HC affects astrocyte lipid metabolism and ApoE secretion. Bottom: Schematic showing different ways 25HC alters cholesterol metabolism in astrocytes. 25HC inhibits the SREBP2 pathway to suppress cholesterol biosynthesis by reducing the expression of cholesterol biosynthetic enzymes such as Hmgcr. By suppressing LDL-receptor (Ldlr), 25HC also reduces cholesterol reuptake. Activation of LXRs by 25HC promotes expression of Abca1 that increase cholesterol efflux. Decreased reuptake of lipoproteins as well as increased secretion of lipoproteins result in higher levels of extracellular ApoE. Finally, 25HC enhances the activity of the cholesterol acyl transferase enzyme, Acat1 to promote cholesterol esterification. Thus, 25HC effectively reduces free cholesterol in astrocytes. This figure was made using BioRender (biorender.com).