Atherosclerosis—Multiple Pathways to Lesional Macrophages

See allHide authors and affiliations

Science Translational Medicine  04 Jun 2014:
Vol. 6, Issue 239, pp. 239ps2
DOI: 10.1126/scitranslmed.3008922


Atherosclerosis is a chronic inflammatory disease that manifests in multiple vascular beds and frequently culminates in ischemic events, including myocardial infarction. Blood monocytes that are recruited to the inflamed vascular wall develop into inflammatory macrophages and foam cells, which contribute to pathogenesis at many stages of this disease and, therefore, represent a target for therapeutic interventions. Recently, alternate sources of macrophages have been identified. Here, we discuss the origin and molecular regulation of macrophages and highlight recent conceptual changes that may shape the development of effective treatments.

Cardiovascular disease is a leading cause of morbidity and mortality worldwide. It is driven by atherosclerosis, a chronic inflammatory process of the vascular wall that is initiated by alterations in cholesterol metabolism and, together with endothelial cell dysfunction, leads to the subendothelial retention of low-density lipoproteins (LDL). This vascular lipid deposition in turn alerts the immune system—a network of cells and humoral factors that operates continuously to protect the host not only from a diverse array of pathogens, but also from sterile hazards. We now know that the consecutive local congregation of both adaptive and innate immune cells is a key step in atherogenesis, orchestrating the formation and progression of lipid-rich lesions or plaques (1).

These highly inflammatory atherosclerotic lesions can grow slowly over years to eventually impede blood flow, leading to the clinical manifestation of stable angina pectoris. However, atherosclerotic lesions may also be complicated by abrupt rupture, resulting in local accumulation and activation of platelets and culminating in rapid thrombotic vascular occlusion and life-threatening conditions, such as acute myocardial infarction and stroke. Inflammatory cells, particularly myeloid leukocytes, not only lead to atherosclerotic lesion formation and destabilization, but also directly support thrombosis after plaque rupture in a process termed “immunothrombosis,” which involves the release of tissue factor, the major initiator of coagulation, or the deposition of highly thrombogenic neutrophil extracellular traps (NETs) (2). Hence, atherosclerosis can be considered a condition in which the immune system causes severe harm to the host.

A key feature of atherosclerosis is the accumulation within lesions of macrophages, which drive the local inflammatory process and promote plaque rupture and thrombosis (3). Targeting macrophages could therefore represent an efficient strategy for the treatment and prevention of atherosclerosis. However, such efforts will rely on a detailed understanding of the origin of plaque macrophages as well as their molecular regulation. The long-held view that bone marrow (BM)–derived monocytes are the one and only source of macrophages in tissues, and that mature differentiated macrophages lack proliferative capacity, has been challenged by recent publications. Here, we will give a short overview on the diversity of the monocyte-macrophage lineage with respect to their developmental origin and highlight recent advances in the understanding of macrophage biology in the context of atherosclerosis (Fig. 1).

Fig. 1 Multiple pathways to lesional macrophages.

During atherosclerosis, blood monocytes, which arise from HSPCs in the BM, are recruited to the inflamed vascular wall, where they develop into inflammatory macrophages and foam cells. However, recent reports suggest that lesional macrophages might also arise from differentiated fetal macrophages that persist from early embryonic development and self-renew throughout adulthood. In addition, lesional macrophages might also derive from circulating and resident arterial HSPCs.



The recruitment of leukocytes from the blood is a key mechanism in the immune response to tissue inflammation and infection but is also considered an essential step in atherosclerosis. During atherogenesis, an inflammatory response in the arterial vascular wall triggers the entry of blood monocytes and their local differentiation into mononuclear phagocytes and lipid-laden foam cells. Macrophage numbers in atherosclerotic mouse aortae increase up to 20-fold in the course of atherogenesis (3); this increase is determined not only by monocyte recruitment, but also by local monocyte/macrophage proliferation and survival.

After entering the tissue, monocytes are thought to differentiate into macrophages or migratory dendritic cells (DCs). In the mouse, blood monocytes are divided into two subsets that play distinct roles in lesion formation and progression: a major subset of Ly6Chigh CCR2+ CX3CR1low “inflammatory” monocytes (CD14+ in humans) and a less frequent subset of Ly6Clow CCR2 CX3CR1high monocytes (CD14dim CD16+ in humans) (4). Ly6Chigh inflammatory monocytes seem to preferentially give rise to lesional CD11b+ F4/80+ macrophages. Although the exact lineage relationships between circulating monocytes and lesional macrophages still remain incompletely understood, CD11b+ monocytes/macrophages are thought to be critical to atherogenesis. Their depletion in CD11b diphtheria toxin receptor (DTR) transgenic mice markedly reduced lesion development (5). Surprisingly though, CD11b+ depletion did not alter the extent or composition of already established atherosclerotic plaques, which indicates that they may play a minor role for plaque maintenance. However, this has to be taken cum grano salis because it remains to be established whether the model of CD11b DTR transgenic mice is indeed suitable to discriminate between the contribution of CD11b+ blood monocytes and that of “resident” CD11blow F4/80high macrophages.

The role of the less frequent subset of Ly6Clow monocytes still remains cryptic (6). They might develop preferentially into CD11c+ DCs (7); however, recent work has challenged this concept and suggested alternative functions: Ly6Clow monocytes patrol the vessel wall, sensing danger signals and viruses. In response to local threats, they recruit neutrophils and modulate endothelial damage (8). Absence of patrolling monocytes in Nr4a1-deficient mice is associated with enhanced atherosclerosis, suggesting that they might even have a protective function (9).

Atherosclerosis has been described as a state of nonresolving inflammation (10). Nevertheless, blood monocytes not only continuously migrate into atherosclerotic lesions to give rise to macrophage subsets, but monocytes and their lesional progeny also have the potential to again emigrate from the plaque (11). These emigrating cells predominantly display characteristics of classical DCs that travel toward the local draining lymph nodes to activate cells of the adaptive immune system, particularly antigen-specific T lymphocytes. Importantly, atherogenic T cells can also be primed locally in the arterial wall by DCs and macrophage populations, resulting in T cell activation and production of inflammatory cytokines, which in turn supports foam cell formation (1). In addition to classical DCs, monocyte-derived macrophages also are able to leave the plaque under certain circumstances—a process that is mostly observed from regressing lesions and might be essential to reduce plaque burden. Their emigration is facilitated by high-density lipoprotein (HDL) and involves the liver x receptor (LXR) (12). Notably, these macrophages display increased expression of anti-inflammatory genes such as arginase I, CD163, and C-lectin receptor (12). This suggests that fostering egress of monocyte-derived cells, in addition to blocking their recruitment, might be a promising strategy to treat and prevent atherosclerosis in the future.


Blood monocytes arise from hematopoietic stem and progenitor cells (HSPCs) and their progeny in the BM compartment. However, HSPCs not only reside in BM but are also found circulating through blood, from where they traffic to multiple peripheral organs. When they encounter tissue damage, these migratory HSPCs are retained locally, start to proliferate, and give rise to tissue-resident myeloid cells including macrophages (13). Hence, HSPCs survey peripheral organs and can act as effectors of the innate immune system providing mature leukocytes in peripheral tissues as needed. Migration of HSPCs is controlled by gradients of chemokines (such as SDF-1) or lipid mediators (such as the sphingolipid S1P). Further, ligation of Toll-like receptors (TLRs), important initiators of innate immune responses, directly trigger cell-cycle entry and myeloid differentiation of HSPCs (14). The latter have also been identified to modulate the atherosclerotic disease process. Indeed, TLRs on macrophages sense proatherogenic stimuli (such as oxidized LDL), triggering inflammatory signaling (15). Whether migratory HSPCs also replenish the macrophage pool in atherosclerotic lesions is still unclear. Nevertheless, several recent observations support the involvement of HSPCs in atherosclerotic lesion formation and progression. It has been estimated that between 100 and 400 long-term HSCs are circulating at a time (16). In response to stress or injury, HSPCs are increasingly mobilized from BM, which is controlled by sympathetic nervous system signaling. The progenitors then seed extramedullary organs such as the spleen to boost production of inflammatory monocytes. In apolipoprotein E–deficient mice, the gain in inflammatory monocytes after the induction of myocardial infarction enforced lesional monocyte recruitment and resulted in larger and more advanced atherosclerotic plaque (17).

However, HSPCs might even more directly contribute to the lesional macrophage pool. Monocyte-predisposed HSPCs have recently been identified in the adventitia of postnatal mouse aorta (18). Trafficking from BM was shown to replenish some of the hematopoietic potential of the aorta after irradiation, suggesting that both compartments communicate, although the turnover was low. Adventitial HSPCs were capable of forming hematopoietic colonies in vitro and led to early multilineage reconstitution in primary transplantation experiments. However, hematopoiesis arising from those HSPCs was skewed toward monocyte/macrophage lineages and notably up-regulated in proatherogenic mice. Thus, aortae of adult mice harbor macrophage precursors and could therefore represent a local vascular source of macrophages (18). Of note, the tunica adventitia—the outer layer of vessels—represents an important niche, containing not only resident cells but also inflammatory cells that traffic in and out of the vessel wall through the adventitial network of vasa vasorum (19). It is therefore likely that HSPCs and macrophages interact and reciprocally alter their phenotype and function.


Although recruitment of macrophages from the pool of bloodborne monocytes and eventually from circulating HSPCs is a hallmark of inflammation, macrophages are already present in all tissues in humans and most animals under steady-state conditions. They scavenge pathogens, oxidized lipids, apoptotic cells, and other debris. They also have a role in the maintenance of tissue homoeostasis, regulation of cell metabolism, and remodeling of tissues during embryonic development and after injury (20, 21).

On the basis of the mononuclear phagocyte system, tissue macrophages were long thought to exclusively derive from a common pluripotent hematopoietic precursor located in the BM that produced blood monocytes, which then constitutively replenished the pool of tissue macrophages. However, this concept has been questioned by early reports that suggest that tissue macrophage populations are self-proliferating and long-lived and thus could exist independently of blood monocytes. Indeed, BM transplantation and parabiosis models show that these populations were not replaced at steady state (22). Consistent with these observations, genetic deficiency in the chemokine receptor CCR2 in mice severely reduces the number of circulating monocytes but does not affect tissue macrophage populations (23).

In humans, macrophages are also able to exist independently from blood monocytes; many tissue macrophage populations are intact in patients with monocytopenia, which can be caused by leukemia or immune deficiency syndromes, and after tissue transplantation (22). Recently, fate-mapping analyses in mice demonstrated that primitive macrophages, presumably located in the yolk sac (YS), give rise to macrophage populations in many tissues including the brain, liver, and heart in an autonomous manner (2426). Specifically, tamoxifen induced pulse-labeling in utero of embryonic day 8.5 (E8.5) Csfr1 MER-iCre-MER;Rosa26stopfloxYFP embryos labeled resident macrophages in tissues of adult mice; however, no labeling was detected in HSC-derived cells, including blood monocytes (24). YS-derived macrophages enter their tissue of residence early during embryonic development, in which they expand and specialize in close association with their local environment and persist and self-renew in adult tissues (27). It is tempting to speculate that under steady state YS-derived macrophages give rise to the majority of tissue macrophages at least in some organs. Because of the qualitative approach of pulse-labeling experiments, the quantitative contribution of YS hematopoiesis to the macrophage pool in the adult organism remains to be determined. However, in mice conditionally deleted of Myb, definitive hematopoiesis does not develop, whereas tissue macrophages in many organs are not affected. This suggests that larger proportions of tissue macrophages in fact have an early embryonic origin, can self-renew in adult mice, and persist independently of HSCs (22). Exceptions to this concept exist for the intestine, where in response to a dynamic environment composed of commensal microflora, BM-derived monocytes continuously replace tissue macrophages (28). Further, monocyte contribution increases upon inflammation, and it is still unknown to what extent monocyte- and embryo-derived macrophages are distinct with respect to the effector functions (27).

Macrophages are also present in the vascular wall in mice and humans at steady state in the absence of atherosclerosis (3, 29). Their numbers strongly increase with atherosclerotic lesion development, and this increase correlates with plaque size. In addition to recruitment of monocyte-derived macrophages from the blood, local proliferation of macrophages (and dendritic cells) in plaques has been reported (30, 31). Proliferation of differentiated macrophages is not specific to atherosclerosis, but has been described for various tissues and inflammatory settings. Thus, the proliferation of distinct macrophage populations seems to represent a general mechanism for macrophage expansion during inflammation (20). The capacity to proliferate and self-renew is characteristic to macrophages of prenatal origin. Hence, local proliferation of YS-macrophages might have an important contribution to the pool of macrophages during inflammatory processes, including that of atherosclerosis. Interestingly though, not only YS-derived resident macrophages (22, 27) but also BM-derived macrophages are thought to proliferate in response to inflammatory stimuli (for example, modified LDL) (20). The relative contribution of YS-macrophages to inflammatory macrophages in atherosclerotic lesions therefore remains to be established.

Proliferation of macrophages, including that of resident tissue macrophages, is controlled by growth factors such as macrophage colony-stimulating factor (M-CSF) and granulocyte/macrophage colony-stimulating factor (GM-CSF), and also by cytokines such as interleukin 4 (IL-4) (20). M-CSF not only promotes myeloid cell differentiation and survival (32), it also enhances macrophage scavenger receptor expression and lipoprotein uptake (33). In line with this, absence of M-CSF in mouse models of atherosclerosis reduces lesion formation (34). However, the response of macrophages to molecular signals depends on the inflammatory condition and probably a distinct microenvironment. In different disease models, tissue macrophages proliferate in response to M-CSF and GM-CSF, whereas IL-4 induces alternative macrophage activation in the setting of T helper 2 cell–mediated inflammation (23, 35). Interestingly, genetic deficiency in IL-4 also reduces lesion development in mouse models of atherosclerosis (36).

In the future, it will be important to dissect the cellular determinants as well as the local molecular cues that regulate macrophage proliferation in the scenario of atherosclerosis. These mechanisms as well as the consequences of macrophage accumulation by proliferation (as opposed to recruitment) are yet to be explored.


The modulation of macrophage accumulation and function represents an attractive therapeutic target for the treatment and prevention of vascular inflammation. However, the development of such strategies depends on a detailed understanding of the origin and regulation of lesional macrophages. Macrophages can derive from primitive hematopoiesis in the early embryo as well as from BM-derived blood monocytes. While the latter are continuously replaced and make only minor contributions to tissue macrophage populations under homeostatic conditions, YS-derived macrophages persist in many adult tissues independently of definitive HSPCs. Thus, these macrophage populations may not only have different developmental origins; rather, their functional phenotype and response to environmental cues is probably distinct and could lead to different responses in arterial inflammation (24, 37). Whether YS-derived macrophages exist in the vasculature, and whether they play a role in atherogenesis, is to date unknown; however, their presence would be in line with recent reports on the proliferative capacity of mature macrophages in the aorta (31). Defining the role of these cells in human arteries in both health and disease may provide new therapeutic opportunities to modulate atherosclerosis and other types of vascular inflammation. Strategies that specifically target individual cell types will be essential to define the contribution of distinct macrophage populations to vascular pathologies.

The role of macrophages during vascular inflammation is not necessarily deleterious. Resident macrophages have many physiological functions in tissues (20). In atherosclerosis, macrophages clear dead cells and other debris, which potentially reduces the inflammatory burden in the vessel wall. A critical issue in the inflammatory process of the vascular wall is apoptosis. Although induced apoptosis can suppress lesion formation, loss of macrophages can result in reduced scavenging of apoptotic bodies and formation of a necrotic core (10). This in turn may trigger lesion progression as well as plaque instability and rupture. Hence, therapeutic strategies targeting macrophage recruitment and/or proliferation also have to account for their potentially beneficial housekeeping functions in vascular tissues.

The recruitment of circulating monocytes to the inflamed arterial wall can be inhibited by blocking chemokines that mediate their adhesion and migration (such as macrophage migration inhibitory factor) (38). In contrast, targeting BM-independent resident macrophages within specific tissues could prove difficult. Extensive depletion of phagocytes by using liposomes would deprive the host of physiologically important macrophage populations such as liver Kupffer cells. Resident macrophages might instead be targeted through the alteration of proliferation signals, such as IL4 or M-CSF (35), correction of functional defects by restoring cholesterol efflux through LXR agonists (12), or by inhibition of apoptosis by targeting the transcription factor MafB (39). Nondepleting liposome-based therapies using the encapsulation of drugs or nanoparticle-delivered small interfering RNA could thereby help to reach cell-specific targets (40).

Modulation of macrophage function represents an attractive therapeutic strategy. This will depend on a deeper understanding of the heterogeneity of macrophage subsets and their developmental origin and molecular regulation. In addition, the regulatory signals derived from the local microenvironment that potentially control phenotype and functional specification of macrophages in response to physiological and inflammatory stimuli require further investigation.


  1. Acknowledgments: This work was supported by the SFB 914, the Deutsche Forschungsgemeinschaft–Forschergruppe 923, the FP7 program (PRESTIGE), DZHK, and the Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (German Ministry of Education and Research).
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article