Macrophage-specific apoE gene repair reduces diet-induced hyperlipidemia and atherosclerosis in hypomorphic Apoe mice

Abstract

Background: Apolipoprotein (apo) E is best known for its ability to lower plasma cholesterol and protect against atherosclerosis. Although the liver is the major source of plasma apoE, extra-hepatic sources of apoE, including from macrophages, account for up to 10% of plasma apoE levels. This study examined the contribution of macrophage-derived apoE expression levels in diet-induced hyperlipidemia and atherosclerosis.

Methodology/principal findings: Hypomorphic apoE (Apoe(h/h)) mice expressing wildtype mouse apoE at ∼2-5% of physiological levels in all tissues were derived by gene targeting in embryonic stem cells. Cre-mediated gene repair of the Apoe(h/h) allele in Apoe(h/h)LysM-Cre mice raised apoE expression levels by 26 fold in freshly isolated peritoneal macrophages, restoring it to 37% of levels seen in wildtype mice. Chow-fed Apoe(h/h)LysM-Cre and Apoe(h/h) mice displayed similar plasma apoE and cholesterol levels (55.53±2.90 mg/dl versus 62.70±2.77 mg/dl, n = 12). When fed a high-cholesterol diet (HCD) for 16 weeks, Apoe(h/h)LysM-Cre mice displayed a 3-fold increase in plasma apoE and a concomitant 32% decrease in plasma cholesterol when compared to Apoe(h/h) mice (602.20±22.30 mg/dl versus 888.80±24.99 mg/dl, n = 7). On HCD, Apoe(h/h)LysM-Cre mice showed increased apoE immunoreactivity in lesional macrophages and liver-associated Kupffer cells but not hepatocytes. In addition, Apoe(h/h)LysM-Cre mice developed 35% less atherosclerotic lesions in the aortic root than Apoe(h/h) mice (167×10(3)±16×10(3) µm(2) versus 259×10(3)±56×10(3) µm(2), n = 7). This difference in atherosclerosis lesions size was proportional to the observed reduction in plasma cholesterol.

Conclusions/significance: Macrophage-derived apoE raises plasma apoE levels in response to diet-induced hyperlipidemia and by such reduces atherosclerosis proportionally to the extent to which it lowers plasma cholesterol levels.

Figure 1. Conditional repair of the hypomorphic Apoe ^h/h allele.

Hypomorphic Apoe gene expression and gene repair strategies (A). Relative Apoe gene expression levels in peritoneal macrophages collected from both groups of mice (n = 4; B). ApoE secreted by peritoneal macrophages in the presence and absence of LXR agonist pre-treatment was quantified (n = 3; C) from Western Blot (D), mean±sem, **p<0.01, ***p<0.001; control is media without macrophages.

Figure 2. Plasma apoE and lipid levels.

Representative Western blot (A) and quantification (B; n = 3) of the band intensity of plasma apoE from wild type (WT), Apoe ^−/− (E^−/−), Apoe ^h/h and Apoe ^h/hLysM-Cre mice either fed a chow diet or a high cholesterol diet (HCD); mean±sem, two-way ANOVA Bonferonni post-tests, **p<0.01 and t-test #p<0.05. Plasma cholesterol (n = 7; mean±sem, one-way ANOVA post-test, ***p<0.001; C) and lipoprotein cholesterol distribution (pooled of 7 mice; D) are shown; inset: Western blot of apoB and quantification (n = 3); mean±sem, two-way ANOVA Bonferonni post-tests, **p<0.01, ***p<0.001. Representative Western Blot of apoE distribution in lipoprotein fractions (E).

Figure 3. ApoE expression by liver and associated Kupffer cells.

Relative CD68 (a macrophage marker) and Apoe gene expression were compared from whole liver extract of Apoe ^h/h and Apoe ^h/hLysM-Cre mice (n = 4 for baseline and n = 7 for 16-week HCD, mean±sem, *p<0.05, **p<0.01; A). Relative apoE protein was also quantified and compared from immunofluorescently labeled liver cross-sections (10 µm thick) after 16 weeks of HCD (mean fluorescence intensity (MFI); n = 7, mean±sem; B). Representative images of whole liver cross-sections show apoE (red) surrounding hepatic sinusoidal surfaces (identified with an endothelial cell marker: VE-cadherin in green), C; scale bar = 100 µm). Higher resolution images demonstrate higher apoE MFI associated with Kupffer cells in Apoe ^h/hLysM-Cre than in Apoe ^h/h mice (D; Mac-2, a macrophage marker, green; scale bar = 30 µm; white arrows point at individual kupffer cells). The apoE MFI per Kupffer cells and hepatocytes were quantified for both groups of mice and the relative ratio calculated (**p<0.01, ****p<0.0001; E).

Figure 4. Macrophage-derived apoE levels in atherosclerosis.

Histological sections of oil-red-O stained aortic roots from Apoe ^h/hLysM-Cre (A) and Apoe ^h/h mice (B; scale bar = 250 µm). Quantification of oil-red-O positive area (n = 7; mean±sem, **P<0.01; C), aortic wall area (D), and % of oil-red-O (E). Correlation analysis between oil-red-O area and plasma cholesterol (n = 14; **P<0.01; F).

Figure 5. Lesional macrophage-derived apoE.

Immunofluorescent images of Apoe ^h/hLysM-Cre (A,C) and Apoe ^h/h mice (B,D; scale bar = 250 µm) aortic roots; anti-Mac-2 (green), and apoE (red). Quantification of lesion area (E), macrophage positive and % of area (F–G); n = 7. Quantification of apoE fluorescence intensity (FI), (H); n = 7, mean±sem, *P<0.05, **P<0.01, ***P<0.001.

Figure 6. Atherosclerotic plaque composition.

Immunofluorescent images of Apoe ^h/hLysM-Cre (A) and Apoe ^h/h mice (B; scale bar = 50 µm) aortic roots; anti-smooth-muscle-cell-α-actin (SMCs, red), anti-Mac-2 (Macrophages, green) and nuclei (blue). Sirius Red stained aortic roots (brightfield (C–D) and polarized light (E–F)) from Apoe ^h/hLysM-Cre (C,E) and Apoe ^h/h mice (D,F); (scale bar = 250 µm). Collagen quantification (G–H, n = 7, mean±sem).