Age-Related Impairments in Immune Cell Efferocytosis and Autophagy Hinder Atherosclerosis Regression

Abstract

Background: Aging is a well-established risk factor for the development and progression of atherosclerosis, but the molecular mechanisms underlying this relationship remain poorly defined, and its role in atherosclerosis regression is unknown. To uncover age-related alterations that may impair atherosclerosis regression, we investigated the response of young and old macrophages to atherogenic lipoproteins in vitro and in vivo.

Methods: Metabolic and proteomic studies were performed in vitro using macrophages differentiated from the bone marrow of young or old mice. To test the role of immune cell aging in atherosclerosis regression, bone marrow from young and old donors was transplanted into irradiated young recipient mice expressing gain-of-function AAV-PCSK9 (adeno-associated virus-proprotein convertase subtilisin/kexin type 9). Following 14 weeks of Western diet feeding, atherosclerosis regression was induced by switching to a standard laboratory diet for 4 weeks.

Results: Compared with young macrophages, old macrophages accumulated more lipid droplets upon lipid loading with the pro-atherogenic lipoprotein aggregated LDL (low-density lipoprotein), accompanied by a failure to proportionally induce autophagy and cholesterol efflux. Proteomic analysis of bone marrow-derived macrophages revealed that pathways related to endocytosis, engulfment, and phagocytosis were downregulated in old lipid-loaded macrophages. Functional studies confirmed a reduction in efferocytic capacity in old macrophages. In recipient mice transplanted with old bone marrow, atherosclerosis regression was impaired, as evidenced by inefficient resolution of circulating inflammatory cell levels, reduced activation of plaque autophagy and apoptotic cell clearance, and persistent plaque CD45⁺ and CD68⁺ content.

Conclusions: Aging impairs macrophage function through reduced efferocytosis and autophagy activation, limiting atherosclerosis regression. These results highlight the need to better define the mechanisms linking aging to atherosclerosis to develop targeted therapies for the aging population.

Keywords: atherosclerosis; cellular senescence; lipoproteins, LDL; macrophages; phagocytosis; plaque, atherosclerotic.

Figure 1.

Old bone marrow–derived macrophages (BMDMs) accumulate more lipid droplets with aggregated low-density lipoprotein (agLDL) loading. A, Fluorescence microscopy of nonloaded and agLDL-loaded BMDMs from young and old mice. Cells were fixed and stained for BODIPY and DAPI (4′,6-diamidino-2-phenylindole; n=5, BMDMs collected from different mice). Scale bar, 20 µm. B, BODIPY quantification from images in A. Data are mean±SEM. C through G, Proteomics analysis of young vs old BMDMs (n=10 per age group), with or without agLDL loading (n=5 agLDL-loaded, n=5 nonloaded per age group). C and D, Volcano plot of differentially abundant proteins (P<0.1) in (C) agLDL-loaded vs nonloaded young BMDMs and (D) agLDL-loaded vs nonloaded old BMDMs. Proteins more abundant in agLDL-loaded BMDMs are represented in red, and proteins more abundant in nonloaded BMDMs are represented in blue. E, Ingenuity pathway analysis (IPA) protein network of the function quantity of lipid droplets in old BMDMs, agLDL-loaded vs nonloaded (P<0.1; Z score, 1.671). F, IPA pathways of lipid metabolism, relating to the term accumulation (P<0.1; |Z score| >0.5). G, Z score of lipid metabolism–related IPA pathways, comparing agLDL-loaded vs nonloaded young BMDMs (left column) to agLDL-loaded vs nonloaded old BMDMs (right column). Included are pathways with P<0.1, |Z score| >0.5 in at least 1 age group. Boxes with an X signify that this pathway was not detected in that age group.

Figure 2.

Aged macrophages show a reduced ratio of autophagy activation relative to lipid loading. A, Binding of aggregated low-density lipoprotein (agLDL; radiolabeled with ³H-cholesterol) by young and old bone marrow–derived macrophages (BMDMs) at 4 °C for 1 hour, as counts per minute (CPM)/µg cell protein (n=3). B, Uptake of agLDL (radiolabeled with ³H-cholesterol) by young and old BMDMs at 37 °C for 3 and 24 hours, as CPM/µg cell protein (n=3). C, ³H-cholesterol efflux to apoAI for 24 hours in nonloaded and agLDL-loaded BMDMs (n=4). D, Volcano plot of differentially abundant proteins in old vs young agLDL-loaded BMDMs (P<0.1; protein more abundant in old in red, and more abundant in young in blue). E, Fluorescence microscopy of BODIPY, p-ATG16L1 (phosphorylated autophagy-related 16-like 1), and DAPI (4′,6-diamidino-2-phenylindole) in nonloaded and agLDL-loaded BMDMs from young and old mice (n=4, BMDMs collected from different mice). Scale bar, 20 µm. F, Quantification of p-ATG16L1 integrated density from images in E (n=4). G through I, Quantification of p-ATG16L1 (G) and total ATG16L1 (autophagy-related 16-like 1; H) by AlphaLISA SureFire Ultra assay (n=5) and percent of phosphorylated ATG16L1 (I). J, Autophagy in nonloaded vs agLDL-loaded BMDMs, ±bafilomycin A1 (baf A1). Representative Western blot of microtubule-associated protein 1A/1B light chain 3 (LC3-I)/LC3-II (n=3). Bottom, Quantification of the LC3-II:LC3-I ratio (left) and quantification of LC3-II normalized to the adjusted total band volume, in arbitrary units (AU; right), from n=3 independent experiments. K, Correlation between fold change in BODIPY mean fluorescence intensity (MFI; Figure 1B; BODIPY MFI agLDL-loaded/BODIPY MFI nonloaded) and fold change in percentage of ATG16L1 phosphorylation (Figure 2I; %p-ATG16L1 agLDL-loaded/%p-ATG16L1 nonloaded) in young and old BMDMs (n=5). Simple linear regression in young as gray dotted line, R²=0.9357 (slope significantly nonzero, P=0.0071), and in old as red dotted line, R²=0.3495 (slope significantly nonzero, P=0.29). The young and old slopes are significantly different between each other (P=0.023). A through K, Data are mean±SEM.

Figure 3.

Impaired measures of atherosclerosis regression in mice with aged bone marrow. A, Study experimental outline. B, Plasma cholesterol levels. C, Circulating immune cell subtypes: inflammatory monocytes CD11b⁺ Ly6G⁻ Ly6C^hi; patrolling monocytes CD11b⁺ Ly6G⁻ Ly6C^lo; neutrophils CD11b⁺ Ly6C⁺ Ly6G⁺. D through N, Analyses of atherosclerotic lesions in the aortic root (n=4–6 mice per sex, per group). D, Total lesion area. E, Necrotic core area. F, Neutral lipids. G, Representative images of hematoxylin and eosin (H+E) staining (scale bar, 400 µm) and oil red O (ORO) staining (scale bar, 400 µm). H, CD45-positive lesion area as a percentage of the cellular lesion area. I, CD45-negative lesion area as a percentage of the cellular lesion area. J, CD68-positive lesion area as a percentage of the cellular lesion area. K, Collagen quantified from picrosirius red (PSR)–stained sections. L, CD45 staining (scale bar, 100 µm). M, CD68 staining (scale bar, 100 µm). N, PSR staining (scale bar, 400 µm). A through N, Data are mean±SEM.

Figure 4.

Impaired autophagy activation in leukocytes from lesions of old regressing mice. A, Representative images of lesions stained for immunofluorescence microscopy using anti-CD45, anti-p-ATG16L1 (phosphorylated autophagy-related 16-like 1), MDH (monodansylpentane; neutral lipids), and DRAQ5 (deep red anthraquinone 5; nucleus). Scale bar, 50 µm. B, Quantification of p-ATG16L1 in CD45-positive area, as percentage of CD45-positive area. C, Quantification of p-ATG16L1 in CD45-negative cellular area, as percentage of CD45-negative cellular area. D, Percentage of foamy leukocytes, defined as MDH⁺ CD45⁺ cells. Data represented as percentage of total CD45 area. A through D, Data are mean±SEM.

Figure 5.

Efferocytosis is impaired in old macrophages in vitro and in vivo. A, Volcano plot of differentially abundant proteins (P<0.1) in old vs young nonloaded bone marrow–derived macrophages (BMDMs; protein more abundant in old and young in red and blue, respectively). B, Comparison of pathways relating to endocytosis, engulfment, and phagocytosis in young and old nonloaded BMDMs, where a negative Z score means lesser expression in old cells. C, Protein network relating to the engulfment of myeloid cells in old BMDMs. D, In vitro efferocytosis assay, expressed as the percentage of BMDMs that have taken up apoptotic cells (ACs). n=4, BMDMs collected from different mice. E, Representative images of in vitro efferocytosis assay. Scale bar, 100 μm. F, Representative images of lesions stained for immunofluorescence microscopy using CD68, TUNEL (terminal deoxynucleotidyl transferase dUTP nick-end labeling), and DAPI (4′,6-diamidino-2-phenylindole). Efferocytosis events were defined as macrophage-associated TUNEL-positive ACs (yellow arrows) and distinguished from free AC not associated with a macrophage (white arrows). Scale bar, 50 µm. G, TUNEL-positive cells, normalized to lesion area. H, Ratio of free AC: CD68⁺-associated AC (efferocytosis events). Values <1 indicate a larger proportion of efferocytosis events. D through H, Data are mean±SEM.