Inflammation and Atherosclerosis: Direct Versus Indirect Mechanisms

Abstract

It is now widely accepted that the development of atherosclerotic lesions involves a chronic inflammatory response that includes both innate and adaptive immune mechanisms. However, it is still unclear precisely what induces the inflammatory response. Furthermore, inflammation within the blood vessel can be divided into direct mechanisms where the primary inflammatory events occur within the intima of the blood vessel and contribute to both the initiation and progression of the plaques and indirect mechanisms where inflammation at non-vascular sites can contribute to the progression of the lesions. The direct mechanisms include lipid deposition and modification, influx of lipoprotein associated factors and microparticles derived from many different cell types, and possibly bacterial and viral infection of vascular cells. Indirect mechanisms derive from inflammation related to autoimmune diseases, smoking, respiratory infection, and pollution exposure, and possibly periodontal disease and gastric infection. The mechanisms include secretion of cytokines and other inflammatory factors into the circulation with subsequent uptake into the plaques, egress and recruitment of activated inflammatory cells, formation of dysfunctional HDL and cross reactive autoantibodies.

Introduction

The connection between inflammation and atherosclerosis is not of recent origin. Microscopic observations beginning in the 19^th century described the presence of inflammatory reddening and inflammatory cells within human plaques (1, 2). Today it is acknowledged that development of atherosclerosis is a unique type of chronic inflammatory response that involves both innate and adaptive immune mechanisms (3). The development of animal models with cholesterol feeding, blood vessel injury and systemic infection have further supported the role of inflammation in the initiation and progression of atherosclerosis. The development of immunocytochemical and flow cytometric approaches with inflammatory cell type specific antibodies has now clearly established that atherosclerotic plaques at all stages of development contain multiple types of inflammatory cells including macrophages, dendritic cells, CD4 and CD8 positive T lymphocytes, B lymphocytes, mast cells and occasionally neutrophils and other granulocytes (4, 5). There is also a rapidly emerging new paradigm that suggests that several types of inflammatory cells within plaques have multiple phenotypes and can even change phenotype after entry into the plaques depending on the exposure to specific micro-environmental cues (6). In this short review the focus will be on distinguishing between direct inflammatory mechanisms that occur within the wall of the blood vessel and are thought to be responsible for the initiation of the disease process and indirect mechanisms where inflammation at non-vascular sites contributes to the progression and destabilization of the plaques.

Direct Mechanisms: Pro-inflammatory factors that initiate the disease process

Because of page restrictions, I will not review the natural history of the development of atherosclerotic plaques but refer the reader to a number of excellent recently published reviews (5, 7, 8). However, what is still unknown is precisely what enters the normal human blood vessel that sets in motion the subsequent activation of the endothelium to express adhesion molecules and leads to the recruitment of inflammatory cells into the intima. Today, lipids are still the leading candidate and the lipid retention hypothesis has been supported by much experimental evidence (9). Although Anitschkow first demonstrated that cholesterol feeding to rabbits led to the formation of lipid loaded cells within the intima (10), more definitive evidence was provided by experiments demonstrating that LDL particles that were labeled with a non-degradable radioactive probe accumulated at lesion susceptible branch points within days of initiating cholesterol feeding in rabbits (11, 12). However, those experiments did not clarify whether the simple accumulation of LDL particles trapped by interactions with matrix molecules (9) is sufficient to activate the inflammatory response or whether the LDL particles must be modified by oxidation of the fatty acids, phospholipids, and cholesterol or by other modifications. There is experimental evidence showing that oxidized lipids induce expression of adhesion molecules by endothelial cells (13, 14) and that oxidation specific epitopes can be localized in the intima coincident with endothelial expression of adhesion molecules (15, 16). Furthermore, lipoproteins with characteristics of those oxidized in vitro can be isolated from human plaques (17). However, free cholesterol itself in a non-oxidized crystalline state, can also induce a similar degree of endothelial activation (18). Furthermore, there is now evidence for a role of free cholesterol in the induction of the NLRP3 inflammasome with consequent release of IL-1B (19). It is also probable that other factors that associate with the LDL particle or with HDL or remnant particles that enter the intima can induce an inflammatory response. These include endotoxin (20) complement factors (21) or enzymes such as myeloperoxidase (22, 23) that if catalytically active can oxidize proteins that then become pro-inflammatory. Another possibility are microparticles (also referred to as microvesicles) that have been reported to enter the artery wall (24, 25). These microparticles are derived from platelets, neutrophils, macrophages, lymphocytes, endothelial cells and other cell types. The microparticles bear the signature of the cell type of origin including membrane proteins and bioactive lipids (24, 25). Microparticles have been shown to activate endothelial expression of adhesion molecules and thus could play a role in the initiation of lesions (26).

Finally, there is evidence that a number of different infectious agents populate human atherosclerotic lesions (27) and that infection of mice and rabbits can accelerate lesion development (27). T cell clones have been isolated from human plaques that proliferate in response to C. pneumoniae antigens (28, 29). Whether infectious agents play a role in the initiation of lesions is not clear as we have reported that C. pneumoniae infection of non-hyperlipidemic mice does not initiate development of lesions but requires prior or simultaneous hyperlipidemia (30). It is also unclear whether these infectious agents directly infect cells within the blood vessel or whether the bacteria are carried by cells that have been infected at other locations. For example, there is evidence that monocytes or macrophages that are infected by C. pneumoniae in the lungs can migrate to the aorta (31). It is also unclear whether there is active infection within atherosclerotic plaques. In most cases, it has not been possible to re-isolate viable organisms from human or experimental plaques (27). However, active infection is not a requirement for eliciting an immune or inflammatory response as even heat killed organisms can activate toll-like receptors on the plasma membrane of endothelial cells, macrophages and dendritic cells (32). Inflammatory cells within human plaques express a number of different toll-like and NOD receptors and deficiency of TLR-2 or TLR-4 impairs the development of lesions in hyperlipidemic mice and reduces the accelerated development of lesions in response to C. pneumoniae infection (33–35). Engagement of toll-like receptors activates a number of different signal transduction pathways leading to the activation of NFkB and other transcription factors that regulate the expression of pro-inflammatory cytokines (36, 37).

As noted, there appear to be multiple phenotypes of inflammatory cells that populate the atherosclerotic plaque. In hyperlipidemic mice there are monocytes expressing both high and low levels of the Ly6C antigen coupled with the CCR2+ chemokine receptor and low levels of the CX3CR1 receptor that are recruited into the plaque. The monocytes expressing high levels of Ly6C are thought to be precursors of inflammatory macrophages that secrete pro-inflammatory cytokines (38–40). There are also subsets of monocytes and dendritic cells in humans such as monocytes expressing high levels of CD14 and low levels of CD16 that may be the source of pro-inflammatory macrophages and dendritic cells in human plaques (41). There are also both immature circulating blood DC antigen positive (BDCA 1+) myeloid versus (BDCA 2+) plasmacytoid dendritic cells that are not of monocyte origin (42). The sources of both the monocytes and dendritic cells is controversial as it is not clear that they are all bone marrow derived. Some cells may also be recruited from lymph nodes and the spleen following some degree of maturation (43–45). Mature dendritic cells and macrophages can also be recruited from the lesions back to lymph nodes and spleen during regression of plaques (46, 47) and are thought to present plaque derived autoantigens to naïve T and B cells (48). In fact, this may be the primary mechanism for how mature plaques regress as studies of plaque composition of regressed plaques where regression was induced by a variety of different approaches and in a number of different species show consistent reductions in the numbers of macrophages and macrophage-derived foam cells (46, 49–52).

Based primarily on in vitro observations, it is now thought that recruited monocytes can differentiate into both pro-inflammatory M1 and anti-inflammatory M2 phenotypes dependent on the combination of cytokines that they encounter (53) and cells expressing markers of both M1 and M2 macrophages are found in both human and mouse lesions (54, 55). Furthermore, these M1 and M2 macrophages can have a similar impact on the phenotype of T lymphocytes by stimulating formation of CD4+ T helper 1 versus T helper 2 cells or T regulatory cells from naïve T cells again dependent on antigen presentation and the types of cytokines they secrete (56). A similar paradigm is thought to exist for dendritic cells and may be mediated by infection with C. pneumoniae (57). However, this is an over-simplification of what is likely to occur in human plaques at different stages of lesion development as there are probably additional phenotypes of macrophages and dendritic cells than just pro- and anti-inflammatory (58, 59). Furthermore, it is still unclear precisely what induces macrophage and lymphocyte polarization within the plaques as factors other than cytokines can also have an impact on the phenotypes of these cells (58, 60–64). There is also emerging evidence that supports additional roles for other types of lymphocytes including CD8+ T cells, T regulatory cells and NK cells (5, 65, 66). A critical question to be addressed by future research is how these different phenotypes play roles in stimulating or preventing the progression of human plaques to clinically relevant stages of the disease.

Indirect Mechanisms: Inflammation at non-vascular sites that contributes to the progression of atherosclerotic lesions

It is well established that people with autoimmune diseases such as rheumatoid arthritis, systemic lupus erythematosus, psoriatic arthritis, and ankylosing spondylitis have an increased risk of mortality from cardiovascular disease (67). This gave the first hint that inflammatory disorders at sites other than the blood vessel could contribute in a number of ways to the chronic inflammation within the artery wall. Smoking is also a well established risk factor for CVD mortality as are respiratory infection, air pollution exposure and possibly gastric and periodontal infections (27, 68, 69). The outstanding question is what the mechanisms are by which all of these non-vascular inflammatory conditions contribute to the ongoing inflammation in atherosclerotic lesions.

Perhaps the most obvious answer is that cytokines and other factors that are generated at non-vascular sites can be released into the circulation and taken up into the wall of the blood vessel. There are increases in plasma cytokines associated with the above autoimmune diseases (67) and we have shown that respiratory infection of mice with C. pneumoniae stimulates an acute phase response with measureable increases in plasma cytokines (70). Exposure of mice to concentrated diesel exhaust or ambient particles derived from vehicular emissions and other sources of pollution also stimulates increases in cytokines (71–73). Air pollution exposure as well as smoking can lead to the oxidation of lung surfactant phospholipids (74) that may be picked up by circulating lipoproteins and delivered to the blood vessel. As noted, activated inflammatory cells such as C. pneumoniae infected monocytes and macrophages may also exit non-vascular sites of inflammation and subsequently be recruited to the atherosclerotic plaque (31). This may also be the case with inflamed peri-adventitial adipose tissue (75). There is emerging evidence that obese people with increased visceral fat have increased numbers of macrophages and lymphocytes in the stromal vascular compartment of the adipose tissue. These cells are also sources of cytokines such as TNF-alpha that are thought to contribute to insulin resistance in liver and muscle (76). It is quite conceivable that these activated inflammatory cells and/or secreted cytokines can enter the adventitia and vasa vasorum and be delivered to the plaque (75).

Another indirect mechanism that has gained recent acceptance is the inflammation associated formation of dysfunctional HDL (77). For decades, HDL cholesterol has been thought of as the “good” cholesterol as epidemiologic data consistently demonstrated a strong association between plasma HDL cholesterol levels and a reduced risk of CVD (78). It is now becoming clear that HDL cholesterol may not be an adequate marker of reduced risk of CVD as recent clinical trials with agents that increase HDL cholesterol coupled with genome wide association studies have shown that there is no reduction in risk with elevation of HDL cholesterol (77). HDL is thought to play a protective role in CVD primarily by picking up cholesterol from the plaque and transporting it to the liver to be converted to bile and then excreted, a process referred to as reverse cholesterol transport. However, HDL also has anti-inflammatory and anti-oxidant properties that can be protective of atherosclerosis. These properties are most likely dependent on the proteins carried on the HDL particle such as the anti-oxidant enzyme paraoxonase (79, 80). Patients with rheumatoid arthritis and systemic lupus erythematosus have HDL with a reduced capacity to inhibit endothelial expression of adhesion molecules and the oxidation of LDL (61). In fact, patients with established coronary artery disease also appear to have dysfunctional HDL (81). Furthermore, mice infected with influenza A (82) or with C. pneumoniae (70) or exposed to concentrated ambient particles derived primarily from vehicular emissions also have dysfunctional HDL (83).

Finally, there are autoantibodies that recognize antigens within atherosclerotic plaques but that are thought to be generated in response to antigens generated elsewhere. For example, autoantibodies that recognize oxidized phospholipids within the plaques are known to be induced in response to oxidized phospholipids within bacterial cell walls or the membranes of apoptotic cells (84, 85). It is also conceivable that oxidized phospholipids generated from lung surfactant following exposure to air pollution may also contribute to the generation of these auotantibodies (70). There are also autoantibodies that recognize cholesterol and oxidized cholesterol (86, 87) and bacterial heat shock proteins and that cross react with human heat proteins (88). Thus, these autoantibodies could form antibody:antigen complexes within the developing plaque that could further contribute to the chronic inflammatory response of atherosclerosis.

Conclusions

It is now widely accepted that the development of atherosclerosis involves a chronic inflammatory response that includes both innate and adaptive immune mechanisms. However, there continue to be questions regarding the factors that directly stimulate these mechanisms within the intima of the blood vessel and set in motion the initiation and progression of the disease process. There are multiple possibilities that include lipid deposition and oxidation, accumulation of cell derived microparticles, and non-lipid lipoprotein associated factors. It is now clear that inflammation at non-vascular sites can contribute to the progression of atherosclerosis. These include multiple autoimmune diseases, smoking, respiratory infection and pollution exposure. There are still outstanding questions concerning the mechanisms by which non-vascular inflammation contributes to the chronic inflammation in the plaque. Possibilities include increases in plasma cytokines and other pro-inflammatory factors generated at non-vascular sites, migration and recruitment of already activated inflammatory cells and the formation of dysfunctional HDL. The mandate for the future is to verify that these direct and indirect mechanisms contribute to human atherosclerosis and to develop appropriate preventive and therapeutic approaches.

Figure 1. Potential Mechanisms of How Inflammation at Non-vascular Sites Contributes to Inflammation in Atherosclerotic Lesions.

This figure depicts how inflammation in the lungs induced by infection, smoking or air pollution can activate inflammatory cells that exit the lungs and home to the blood vessel (top) or how oxidation of lung surfactant phospholipids (Ox-pL) can induce autoantibodies that react with oxidized low density lipoproteins (Ox-LDL) forming pro-inflammatory antibody-antigen complexes (Ab-Complex). Activated inflammatory cells may also be recruited from inflamed peri-adventitial adipose tissues (bottom). M1 = pro-inflammatory macrophages, M2 = anti-inflammatory macrophages, Th-1 = CD4+ T helper 1 lymphocytes, Th-2 = CD4+ T helper 2 lymphocytes.

Highlights.

• Inflammation in atherosclerosis can be divided into direct and indirect mechanisms

• Direct mechanisms are inflammatory events occurring within the blood vessel

• Direct mechanisms include lipid deposition, microparticles, and infectious agents

• Indirect mechanisms derive from inflammatory events the occur at non-vascular sites

• Non-vascular events include respiratory infection and air pollution exposure