Foamy monocytes and atherogenesis in mice with combined hyperlipidemia and effects of antisense knockdown of apoCIII

Abstract

Hypertriglyceridemia (HTG), particularly in combined hyperlipidemia, increases risk for atherosclerotic cardiovascular disease, but the underlying mechanisms remain incompletely understood. We sought to determine contributions of circulating monocytes to atherosclerosis associated with HTG in combined hyperlipidemia, created by transgenic expression of human apoCIII in Ldlr^-/- mice (Ldlr^-/-ApoCIIItg) fed Western high-fat diet (WD). Tissue culture with THP-1 and primary human monocytes was used to examine effects of triglyceride (TG)-rich lipoproteins on monocytes. Ldlr^-/-ApoCIIItg mice were also treated with apoCIII antisense oligonucleotide (ASO) and examined for foamy monocytes and atherosclerosis. Compared to Ldlr^-/- mice, Ldlr^-/-ApoCIIItg mice fed WD had early and persistent increases in lipid accumulation within monocytes and enhanced atherosclerosis. Ldlr^-/-ApoCIIItg mice versus Ldlr^-/- mice had higher levels of CD11c, CD36, and cytokines in foamy monocytes, with increases in foamy monocyte adhesion to vascular cell adhesion molecule-1 and oxidized LDL uptake. Monocytes took up TG-rich lipoprotein in vivo and in vitro and changed phenotypes. Foamy monocytes infiltrated into atherosclerotic lesions, and specific and sustained depletion of CD11c⁺ (foamy) monocytes profoundly reduced atherosclerosis in Ldlr^-/-ApoCIIItg mice on WD. Treatment with apoCIII ASO lowered plasma TG and cholesterol levels, improved foamy monocyte phenotypes, and reduced atherosclerosis in Ldlr^-/-ApoCIIItg mice. In conclusion, HTG in combined hyperlipidemia accelerates atherosclerosis, in part, by increasing foamy monocyte formation and infiltration into atherosclerotic plaques. Treatment with apoCIII ASO is a potential new therapy for improving monocyte phenotypes and reducing atherosclerosis in combined hyperlipidemia.

Keywords: atherosclerosis; cholesterol; foam cells; inflammation; lipoproteins; monocytes; triglyceride.

Figure 1

Increased plasma TGs and atherosclerosis in Ldlr^−/− mice with transgenic expression of human apoCIII (Ldlr^−/−apoCIIItg). A: Ldlr^−/−ApoCIIItg and Ldlr^−/− mice were fed WD for 6 weeks. Plasma total TG and total cholesterol levels (n = 4–9 samples/group). B: TG and cholesterol distributions in lipoprotein fractions fractionated by size-exclusion chromatography. Data are presented as TG and cholesterol levels and percentage in each lipoprotein fraction (n = 3–4 samples/group). C: Representative photographs and quantitative analysis of en face Oil Red O staining for atherosclerotic lesions in whole aorta and aortic sinus in mice (n = 6–8 mice/group). D: Staining for CD11c and human apoCIII (hApoCIII) in aortic sinus lesions in mice. ∗P < 0.05 and ∗∗∗P < 0.001 compared to Ldlr^−/− group.

Figure 2

Changes in monocyte and subset proportions and lipid accumulation in Ldlr^−/−apoCIIItg mice. All samples were collected from Ldlr^−/−ApoCIIItg and Ldlr^−/− mice fed WD for 6 weeks or as indicated otherwise. A: Percentage of total monocytes in total leukocytes and percentage of CD11c⁺ monocytes in total monocytes in blood of mice. B: Representative flow cytometric (FACS) examples of monocytes showing SSC, representing cell granularity and intracellular lipid droplet accumulation, and quantification of monocyte and subset SSC. Data were collected from three independent experiments with 6–8 mice/group in each experiment (A and B). C: Changes in SSC of CD11c⁺ monocytes in mice during WD (n = 4–9 samples/group). D: FACS examples and quantification of Nile Red staining for lipids in CD11c⁺ monocytes in mice (n = 5–9 mice/group). E: Quantification of total cholesterol, cholesteryl ester, TG, and free glycerol in monocytes (n = 3 mice/group). Results were normalized to the total cell number. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 compared to Ldlr^−/− group.

Figure 3

Changes in monocyte and subset phenotypes in Ldlr^−/−apoCIIItg mice. Ldlr^−/−ApoCIIItg and Ldlr^−/− mice were fed WD for 6 weeks or as indicated otherwise. A: Change in CD11c MFI of foamy (CD11c⁺) monocytes (n = 8–10 mice/group), CD11c MFI of foamy (CD11c⁺) monocytes, and monocyte adhesion to VCAM-1 examined by ex vivo flow adhesion assay. B: Monocyte CD36 MFI and OxLDL uptake ex vivo. Data show representative FACS examples and quantification of OxLDL uptake by CD11c⁺ monocytes. C: Representative FACS examples and quantification of intracellular TNFα and IL-1β in CD11c⁺ and CD11c^– monocytes in Ldlr^−/−ApoCIIItg mice. D: Intracellular TNFα and IL-1β in CD11c⁺ monocytes in Ldlr^−/−ApoCIIItg and Ldlr^−/− mice. E: CX3CR1 MFI on CD11c⁺ and CD11c^– monocytes in mice. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 compared to Ldlr^−/− group (A).

Figure 4

Monocyte uptake of TGRL and phenotypic changes. A: Representative FACS examples of mouse monocyte uptake of TGRL in vivo and ex vivo. Left panel (in vivo): DiI-mTGRL isolated from Ldlr^−/−ApoCIIItg mice and labeled with DiI were intravenously injected into WT mice (recipients), and monocyte uptake of DiI-mTGRL was examined in recipient blood 24 h later after staining for CD204 and CD11c. Right panel (ex vivo): Blood from WT mice was incubated with or without DiI-mTGRL for 3 h, and monocyte uptake of DiI-mTGRL was examined after staining for CD204 and CD11c. Data shown were gated monocytes (CD204⁺). B: Representative FACS examples of THP-1 monocyte uptake of TGRL (indicated by DiI signal after incubation with DiI-TGRL for 4 h) and lipid accumulation (indicated by elevated SSC and Nile Red staining after incubation with TGRL for 48 h). C: Monocyte uptake of TGRL and lipid accumulation with lipoprotein lipase (LPL) inhibition or addition. THP-1 monocytes or blood from healthy humans were incubated with DiI-labeled or unlabeled postprandial TGRL in the presence or the absence of orlistat (orli) for 4 h (upper panels for DiI-TGRL uptake) or 24 h (lower left and middle panels for lipid accumulation). Lower right panel: representative FACS examples of lipid accumulation in THP-1 monocytes treated with TGRL for 48 h in the presence or the absence of exogenous LPL. D: Representative FACS examples and quantification of effects of TGRL treatment on THP-1 CD36 expression and uptake of OxLDL. THP-1 monocytes were treated with TGRL for 48 h and then, after being washed with PBS to remove TGRL, incubated with DiI-OxLDL in the presence or the absence of a CD36 mAb for an additional 4 h. THP-1 expression of CD36 was examined after incubation with TGRL and before incubation with DiI-OxLDL, and THP-1 uptake of DiI-OxLDL was examined after incubation with DiI-OxLDL. E: Effects of TGRL on THP-1 expression of IL-6, MCP-1, and IL-1β examined by quantitative RT-PCR. All representative examples were from ≥3 independent experiments with similar results. Data are shown as mean ± SEM.

Figure 5

Role of foamy monocytes in atherosclerosis development in Ldlr^−/−apoCIIItg mice. A: Representative FACS example showing labeling of circulating CD11c⁺ (foamy) monocytes with microbeads and histology of aortic sinus showing infiltration of microbead (green)-labeled CD11c⁺ monocytes. Ldlr^−/−ApoCIIItg mice fed WD for 6 weeks received intravenous injection of fluorescent microbeads; labeling of circulating monocytes was examined at 24 h, and infiltration of labeled monocytes was examined at 48 h after microbead injection. B: Representative FACS examples showing DiI-TGRL-labeled monocytes in donor (left) and recipient (right) blood and histology of recipient aortic sinus showing infiltration of DiI-TGRL (red)-labeled monocytes (CD11c⁺, Fig. 4A). DiI-mTGRL was intravenously injected into WT mice (donors). After 24 h, blood was analyzed by FACS, and mononuclear cells were isolated from donors and injected intravenously into recipient Ldlr^−/−ApoCIIItg mice fed WD (12 weeks). Twenty-four hours later, the same recipients received another injection of labeled mononuclear cells from WT and were examined for labeled monocytes in the circulation at 24 h and infiltration of labeled monocytes at 48 h after the second injection of mononuclear cells. C–G: Effects of depletion of CD11c⁺ (foamy) monocytes on atherogenesis in HTG. Ldlr^−/−ApoCIIItg mice fed WD received intravenous injection of low-dose clodrosome (clod) or saline (and Ldlr^−/− mice received saline) daily for 6 weeks (n = 5–7 mice/group). C: Representative FACS examples showing specific depletion of CD11c⁺ monocytes with clodrosome injection (1 week) and quantification of CD11c⁺ and CD11c^– monocytes. D: Plasma total TG and cholesterol levels. E: Representative photographs and quantification of en face Oil Red O staining of whole aorta. F: Representative photographs and quantification of Oil Red O staining in aortic sinus. G: Representative photographs and quantification of CD11c staining in aortic sinus lesions. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 compared to Ldlr^−/− mice with saline (D); ##P < 0.01, ###P < 0.001 compared to Ldlr^−/−ApoCIIItg mice with saline (C).

Figure 6

Effects of apoCIII ASO treatment on plasma lipids, monocyte phenotypes, and atherosclerosis in Ldlr^−/−apoCIIItg mice. Ldlr^−/−ApoCIIItg mice fed WD were treated with a GalNac-conjugated ASO against human apoCIII or a GalNac-conjugated CO (and Ldlr^−/− mice were treated with CO) at 10 mg/kg body weight weekly for 12 weeks. A: Plasma total TG and cholesterol levels (n = 14–19 mice/group from three independent experiments with 4–7 mice/group in each experiment). B: Representative FACS examples showing SSC and CD11c of CD36⁺ and CD36^– monocytes (treatment for 4 weeks) and quantification of changes in proportions, SSC, and CD11c MFI of CD36⁺ monocytes with treatment (n = 11–14 mice/group). C: Representative photographs and quantification of en face Oil Red O staining of whole aorta. D: Representative photographs and quantification of Oil Red O staining in aortic sinus. E: Representative photographs and quantification of CD11c staining in aortic sinus lesions. Data are shown as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 compared to Ldlr^−/− mice with CO (A and B); ##P < 0.01, ###P < 0.001 compared to Ldlr^−/−ApoCIIItg mice with CO (A and B).