Autophagy in Atherosclerosis: Not All Foam Cells are Created Equal

Decades of research using genetically-modified hyperlipidemic mice has led to the conclusion that atherosclerosis is primarily a macrophage-driven process.¹ Smooth muscle cells (SMCs), the other main cell type in atherosclerotic lesions, have been depicted mostly as forming the protective fibrous cap over a vulnerable core of inflamed and dying or dead macrophages. What has been largely forgotten or de-emphasized is that in human atherosclerosis-prone arteries, the intima is a thick layer containing predominantly SMCs and their secreted proteoglycans, a feature absent from mouse arteries,^{2, 3} with macrophages and other leukocytes being relatively sparse until later stages of plaque development (Figure 1A-F).^{2, 4} Throughout all stages of human atherosclerosis, SMCs are the most abundant cell type, whereas in mouse lesions they are thought to contribute approximately 36% of total cells in advanced plaque,⁵ though in both cases they frequently masquerade as macrophages.^5,6

Figure 1. Diffuse intimal thickening, macrophages and smooth muscle cell foam cells in early human atherosclerosis.

A, anti-SMC actin antibody HHF35 staining of thoracic aorta in young adult male showing preponderance of SMCs in the diffusely thickened intima (I); (M), media. B, anti-macrophage antibody HAM56 of same lesion demonstrating macrophages confined largely to the immediate subendothelial region. C, Higher magnification of HHF35 staining demonstrating SMC foam cells near the base of the intima. Used with permission from reference 5. D, Grade 2 fatty streak in right coronary artery of a 44-year old male showing diffuse thickening in intima (I) relative to media (M). Arrowheads indicate the internal elastic lamina. E, same tissue stained with Sudan IV showing lipid staining primarily in deeper intima. F, Macrophage staining with anti-CD68. Used with permission from reference 6. G, Atherosclerotic plaque depicting differences between a macrophage and a SMC in the fate of hydrolyzed cholesteryl esters (CE) from uptake of modified lipids such as aggregated LDL (agLDL) or oxidized LDL (oxLDL). Free cholesterol (FC) generated by Lysosomal Acid Lipase (LAL) is re-esterified in endoplasmic reticulum (ER) by Acyl CoA Acyl Transferase (ACAT) to CE accumulating in cytosolic lipid droplets. Lipophagy then delivers these CE to lysosomes where LAL can re-hydrolyze to FC for efflux by ABCA1 or ABCG1 transporters to ApoA-I/HDL.

Earlier studies by Gown, Stary, Wissler and colleagues (reviewed in ⁴) indicated the presence of large numbers of SMC foam cells in the deep intima in early human atherogenesis. Based on these studies and evidence of impaired expression of the key cholesterol exporter ABCA1 by intimal SMCs,⁶ the Francis group has studied the relative contribution of SMCs and macrophages to the total plaque foam cell population. Allahverdian et al. found that SMCs are the source of at least 50% of all foam cells in human coronary plaque,⁶ while Wang et al. reported that in both nonlineage-tracing and SMC lineage-tracing ApoE-null mice, 65–70% of foam cells from mouse aortic atheromas are SMC-derived.³ The latter study also demonstrated aortic SMC foam cells are different from macrophage foam cells in that they lack cytoplasmic lipid droplet granularity as evidenced by low side scatter in flow cytometry analysis, despite similar levels of cholesteryl esters.³ This study and others suggest the definition of a bona fide foam cell does not require the presence of cytoplasmic lipid droplets. In 2021 Dubland et al. reported that vascular SMCs in humans and mice have markedly lower baseline expression of Lipa/Lysosomal acid lipase (LAL) than macrophages, and that SMC foam cells sequester lipoprotein-derived cholesteryl esters within their lysosomes, providing an explanation for the difference in sites of cholesteryl ester storage between SMC and macrophage foam cells in vivo.⁷

Understanding differences and similarities in lipid uptake, storage, and efflux between plaque SMCs and macrophages will be critical in determining how cholesterol metabolic dysfunction occurs in these cells and how novel therapeutics can be leveraged to mitigate atherosclerosis. In the current issue of Circulation Research, Robichaud et al. contribute to this phenotyping by assessing the relative impairment in autophagy of SMCs and macrophages, and whether this might provide further clues into mobilizing cholesterol from foam cells in atherosclerotic plaque.⁸ The importance of autophagy dysfunction in foam cell formation has been increasingly recognized in the last decade, first described in plaque macrophages.⁹ Although it remains unclear whether neutral lipid stores are targeted for autophagic degradation or delivered to lysosomes by bulk autophagy (i.e., selective versus non-selective autophagy), derangements in lipophagy are clear contributors to excess lipid storage of plaque cells.¹⁰ In this regard, attempts at parsing the differences in autophagy/lipophagy dysfunction between plaque SMCs and macrophages is important given SMCs appear to contribute the majority of foam cells in humans⁶ and now in more than one mouse model of atherosclerosis.^{3, 8}

In order to discriminate SMC and macrophage populations in the plaque with regard to both autophagosome formation and lysosome acidity/function, Robichaud et al took advantage of the often-used GFP-LC3 mouse model together with LysoTracker staining.⁸ They observed clear differences between these populations isolated from lesions, which they further characterized in vitro using cultured cells of each type. However, given the dynamic nature of autophagy in cells, its accurate evaluation is challenging in cell culture and more so in vivo, where classic autophagic markers cannot be easily studied in real-time and in inducible fashion.¹¹ At the level of tissues including atherosclerotic plaques, an evaluation of autophagy at snapshots in time are often the only way possible. Albeit challenging, assessing autophagy flux in vivo can be performed in mouse models by systemic administration of lysosome inhibitors such as chloroquine with or without autophagy induction by starvation.¹¹ Other tools including use of dually fluorescent GFP-RFP-LC3 or a combination of GFP-LC3 and mCherry-LC3 transgenic mice coupled with optical imaging can provide a more nuanced assessment of autophagy flux.¹² Whether these options have enhanced utility for a more complete assessment of autophagy in the atherosclerotic plaque should be the focus of future studies. Another important consideration is that conventional autophagy markers such as LC3 are designed for the assessment of bulk macroautophagy, leaving open the possibility of differences in selective autophagy pathways (e.g., mitophagy, aggrephagy, and lipophagy).¹⁰ The Mito-Keima and MitoQC systems have been leveraged for evaluation of mitophagy in vivo while co-localization of p62/SQSTM1 with polyubiquitinated protein aggregates has been used in assessment of aggrephagy in tissues.^{13, 14} Due to our rudimentary understanding of the machinery involved in lipophagy and a dearth of effective tools for its assessment in vivo,¹⁰ characterizing differences in the lipid homeostasis of foam cells in atherosclerotic plaques remains a challenge.

Another important issue in direct comparison of SMCs and macrophage foam cells is how comparable the cell types are in the various assays. In their first-pass analysis, Robichaud et al assess autophagy markers as well as autophagy flux differences in lipid-loaded bone marrow-derived macrophages (BMDMs) and SMCs isolated from thoracic aortas of mice. Given inherent differences between these BMDM and SMC lineages, they astutely buttress their findings using macrophages and SMCs differentiated from a common embryonic stem cell lineage. In either cell culture model, inherent differences in lipid content, the progression of autophagy/lipophagy, and cholesterol efflux lead to the conclusion that SMC foam cells have a lower capacity to hydrolyze neutral lipid stores than macrophage foam cells. The applicability of these findings to SMCs and macrophages of the plaque remains unclear and will need to be determined in future studies. Furthermore, given the known elaboration of macrophage markers by SMCs during atherogenesis, any efforts in this area will need to take advantage of the several mouse lines currently available to trace SMC lineages in vivo.

Finally, it is important to discuss lipid loading methods used to evaluate lipid homeostasis in vitro and their applicability to the in vivo setting. To study lipophagy in cultured cells, the authors utilize aggregated LDL (agLDL) to load cholesterol into macrophages but resort to methyl-β cyclodextrin (MβCD) to load free cholesterol into SMCs. This difference is needed to facilitate generation of a cytosolic cholesteryl ester substrate pool in SMCs for the lipophagy studies. While that pool forms readily in macrophages loaded with agLDL, the most likely vehicle for forming foam cells in vivo, it does not form in SMCs, presumably on the basis of the intrinsic low LAL in vascular SMCs and resulting inability to deliver free cholesterol derived from lipoprotein cholesteryl esters to the endoplasmic reticulum for reforming into cytoplasmic cholesteryl esters (Figure 1G).⁷ While this method of cholesterol loading allows assessment of autophagy/lipophagy, it raises questions about the in vivo relevance of the findings. It is possible that SMCs of the plaque might take up free cholesterol generated by hydrolysis of extracellular LDL or agLDL by LAL exocytosed by macrophages. However, the low cytosolic lipid droplets of SMC foam cells isolated from mouse lesions and the recent finding that these cells do not mimic the gene profile produced by MβCD-cholesterol loading in vitro,¹⁵ suggest plaque SMC foam cells might not have a large cytosolic lipid droplet pool available for lipophagy in vivo.

Overall, Robichaud et al perform elegant work characterizing the foam cell landscape of atherosclerotic plaques and determine that derangements in autophagy/lipophagy might explain critical differences between SMC and macrophage foam cells. Their work raises several interesting possibilities that will need to be dissected further in vivo to determine whether lipophagy defects are related to or separate from the low LAL in SMCs. Their work adds to the notion that foam cells of the plaque cannot be viewed as a monolith. Our understanding of critical differences between SMC and macrophage foam cells will serve as the basis for future therapies aimed at reducing the residual cholesterol burden of atherosclerosis not removed by existing therapies.