Altered mitochondrial dynamics and function in APOE4-expressing astrocytes

Abstract

APOE4 is a major risk factor for sporadic Alzheimer's disease; however, it is unclear how it exerts its pathological effects. Others and we have previously shown that autophagy is impaired in APOE4 compared to APOE3 astrocytes, and demonstrated differences in the expression of mitochondrial dynamics proteins in brains of APOE3 and APOE4 transgenic mice. Here, we investigated the effect of APOE4 expression on several aspects of mitochondrial function and network dynamics, including fusion, fission, and mitophagy, specifically in astrocytes. We found that APOE3 and APOE4 astrocytes differ in their mitochondrial dynamics, suggesting that the mitochondria of APOE4 astrocytes exhibit reduced fission and mitophagy. APOE4 astrocytes also show impaired mitochondrial function. Importantly, the autophagy inducer rapamycin enhanced mitophagy and improved mitochondrial functioning in APOE4 astrocytes. Collectively, the results demonstrate that APOE4 expression is associated with altered mitochondrial dynamics, which might lead to impaired mitochondrial function in astrocytes. This, in turn, may contribute to the pathological effects of APOE4 in Alzheimer's disease.

Fig. 1. Levels of mitochondrial dynamics proteins in APOE4 astrocytes.

a APOE3/APOE4 astrocytes were subjected to Immunoblot, using the indicated antibodies. Left panel, representative results; right panel, densitometric analysis of APOE4 cells as fold of APOE3 (dashed line; n ≥ 3) b Primary astrocytes generated from APOE3/APOE4 targeted-replacement mice were cultured for the indicated time period and subjected to Immunoblot using the indicated antibodies. Upper panel, representative results. Lower panel, quantification of the results as described in (a). c Brain sections of APOE3/APOE4 mice (5 months old; n = 3 mice/group) were co-stained using anti-GFAP (astrocytic marker) with anti-parkin or anti-Mfn1 antibodies. Right panel, co-expression of parkin (red) and GFAP (green) in the cortex area; Mfn1 (green) and GFAP (red) in the hippocampus CA3 area. (scale bars, 50 μm). Magnified inset: merged GFAP with parkin/Mfn1 in APOE4 astrocytes. Left panel, fluorescence intensity was quantified using ImageJ only in GFAP-positive cells, (fold of APOE3, dashed line). a–c Means ± SE; *p < 0.05 and **p < 0.01.

Fig. 2. Mitochondrial fraction and mitochondrial morphology in APOE4 astrocytes.

a APOE3/APOE4 astrocytes were treated with 10 μM chloroquine (CQ) for 24 h. Mitochondrial and cytosolic fractions were subjected to Immunoblot using the indicated antibodies. AIF and GAPDH were used as markers for cytosol and mitochondria, respectively. Upper panel, representative results; lower panel, densitometric analysis of proteins in the mitochondrial fraction (normalized to AIF; fold of APOE3; n ≥ 3). b APOE3/APOE4 astrocytes expressing Mito-GFP were stained with anti-apoE antibodies (scale bars 25 μm). Left panel, representative results. Right panel, co-localization analysis of Mito-GFP (green) and apoE (red) using the Pearson’s R value above threshold as calculated by the ImageJ Coloc 2 plugin after thresholding for apoE; n ≥ 50 cells; means ± SE; **p < 0.01. c APOE3/APOE4 astrocytes were subjected to real-time PCR for APOE expression (fold of APOE3; n ≥ 3); means ± SE; **p < 0.01. d, e APOE3/APOE4 astrocytes were incubated with MitoTracker Deep Red (MTDR, d) or infected with Mito-GFP (e). Upper panels, representative images and morphological skeleton analysis as generated by the MiNa tool (Scale bars, 25 μm); Lower panels, quantification of branches per network (mean network size) and individual mitochondria relative to the mitochondrial network footprint (n ≥ 50 cells). a–e Means ± SE; *p < 0.05 and **p < 0.01, APOE3-expressing compared with APOE4-expressing cells; ^p < 0.05, CQ-treated compared to untreated cells.

Fig. 3. CCCP-induced mitophagy in APOE3/APOE4 astrocytes.

a APOE3/APOE4 astrocytes expressing Mito-GFP were treated with 15 μM CCCP for 1 h and stained with anti-LC3 antibodies (scale bars 25 μm). Left panel, representative results. Right panel, co-localization analysis of Mito-GFP (green) and LC3 puncta (red) using the Pearson’s R value above threshold as calculated by the ImageJ Coloc 2 plugin after thresholding for LC3 puncta; n ≥ 50 cells/treatment; means ± SE; **p < 0.01, APOE3 compared with APOE4; ^^p < 0.01, CCCP-treated compared to untreated cells. b APOE3/APOE4 astrocytes were treated with 25 μM CCCP for 4 h and analyzed by electron microscopy. Left panel, representative micrographs; right panel, the number of autophagosomes containing/interacting (orange arrows) or not (blue arrows) with mitochondria was measured per 10 μm fields (scale bars, 500 nm); n ≥ 10 fields/treatment; means ± SE; *p < 0.05 and **p < 0.01, APOE3 compared with APOE4; ^^p < 0.01, CCCP-treated compared to untreated cells.; white asterisks, spheroid/ring-shaped mitochondria (presumably mitochondria interacting with autophagosome/lysosome) c APOE3/APOE4 astrocytes expressing LC3-EGFP-mRFP were treated with 10 μM CCCP for 4 h. Left panel, representative results (scale bars 25 μm); right panel, ImageJ analysis of %area of LC3-EGFP-mRFP (arrows) or LC3-mRFP (arrow heads) positive puncta; n ≥ 50 cells/treatment; means ± SE; *p < 0.05 and **p < 0.01, APOE3- compared with APOE4; ^^p < 0.01, CCCP-treated compared to untreated cells. d APOE3/APOE4 astrocytes were treated with 10 μM chloroquine (CQ) for 48 h, with or without 7.5 μM CCCP for the last 24 h. Levels of LC3 were determined by Immunoblot. Left panel, representative results; right panel, densitometric analysis of the results is presented as fold of APOE3 (left graph) and as the difference between measured values of LC3 levels with or without CQ (right graph, Δ ± CQ); n = 3; means ± SE; *p < 0.05 and **, APOE3 compared with APOE4; ^p < 0.05 and ^^p < 0.01, treated compared with untreated cells, ##p < 0.01, combined CCCP and CQ treatment compared to each treatment alone.

Fig. 4. Synthesis/degradation of Mfn1 (a-c) and Parkin (d-g) in APOE3/APOE4 astrocytes.

a, d APOE3/APOE4 astrocytes were subjected to real-time PCR of MFN1 (a) or PARK2/PRKN (parkin) (d) expression (fold of APOE3; n ≥ 3). b Cells were treated with 1 μM MG-132 (MG) for the indicated times and the levels of Mfn1 were determined by Immunoblot. Upper panel, representative results; lower panel, densitometric analysis of Mfn1 levels is presented as fold of APOE3 (left graph) (n ≥ 3). c, g Cells were treated with 1 μM MG-132 for 8 h and cell lysates were subjected to immunoprecipitation with anti-ubiquitin antibodies. Left panel, representative results. Right panel, densitometric analysis of immunoprecipitated ubiquitinated Mfn1 (c) or ubiquitinated parkin (g) (fold of APOE3; n ≥ 3). e, f Cells were treated with 10 μM chloroquine (CQ; e) or 1 μM MG-132 (MG; f) for 24 h or 8 h, respectively, and the parkin levels were determined by Immunoblot. Upper panel, representative results; lower panel, densitometric analysis of parkin levels (fold of APOE3, n ≥ 3). a–g Means ± SE; *p < 0.05 and **p < 0.01, APOE3 compared to APOE4; ^p < 0.05 and ^^p < 0.01, CQ/MG-132 treated compared to untreated cells.

Fig. 5. Mitochondrial function in APOE3/APOE4.

a Cells were treated with 1 μM MG-132 (MG) for 8 h and PINK1 levels were determined by Immunoblot. Left panel, representative results; Right panels, densitometric analysis of PINK1 levels (fold of APOE3; n ≥ 3). FL-PINK1, full-length PINK1; 52kDa-PINK1, cleaved PINK1; *p < 0.05 and **p < 0.01, APOE3 compared to APOE4; ^p < 0.05 and ^^p < 0.01, treated compared to untreated cells. b Cells were stained with MitoTracker Red (MTR) or MitoTracker Deep Red (MTDR), and fluorescence was measured (fold of APOE3; n ≥ 3); *p < 0.05 and **p < 0.01, APOE3-expressing compared to APOE4 cells; ^p < 0.05, MTDR compared with MTR. c, g Cells were subjected to MTT assay normalized to cell number at the indicated time points (c) (n = 6/treatment). d Cells were processed and analyzed by electron microscopy. Left panels, representative results; right panel, analysis of mitochondrial density was measured using ImageJ (fold of APOE3; n ≥ 20 mitochondria; scale bars, 500 nm); means ± SE; **p < 0.01. e, f Cells were cultured for 48 h, after which, media pH (e) or ATP levels (f) were measured (n ≥ 3); means ± SE; **p < 0.01, APOE3-expressing compared to APOE4-expressing cells.

Fig. 6. Effect of rapamycin on APOE3/APOE4 astrocytes.

a APOE3/APOE4 astrocytes were treated with 150 nM rapamycin (Rapa) for 48 h and subjected to Immunoblot. Left panels, representative results. Right panels, densitometric analysis of protein levels (fold of APOE3; n ≥ 3). b Cells were treated with 150 nM rapamycin for 24 h and subjected to real-time PCR (fold of APOE3; n ≥ 3). c Cells were treated with 150 nM rapamycin for the last 24 h (72 h culture) and subjected to MTT assay normalized to cell number (fold of APOE3; n = 6/treatment). d Cells were treated with 150 nM rapamycin for 72 h, followed by staining with MitoTracker Red (MTR). Fluorescence intensity of MTR was measured using flow cytometry (fold of APOE3; n ≥ 3/treatment). e Cells were treated with 150 nM rapamycin for 72 h, after which, media pH was measured (n = 3/treatment). a–e Means ± SE; *p < 0.05 and **p < 0.01, APOE3 compared with APOE4; ^p < 0.05 and ^^p < 0.01, treated compared with untreated cells.