Cholesterol, inflammation and innate immunity

Abstract

Hypercholesterolaemia leads to cholesterol accumulation in macrophages and other immune cells, which promotes inflammatory responses, including augmentation of Toll-like receptor (TLR) signalling, inflammasome activation, and the production of monocytes and neutrophils in the bone marrow and spleen. On a cellular level, activation of TLR signalling leads to decreased cholesterol efflux, which results in further cholesterol accumulation and the amplification of inflammatory responses. Although cholesterol accumulation through the promotion of inflammatory responses probably has beneficial effects in the response to infections, it worsens diseases that are associated with chronic metabolic inflammation, including atherosclerosis and obesity. Therapeutic interventions such as increased production or infusion of high-density lipoproteins may sever the links between cholesterol accumulation and inflammation, and have beneficial effects in patients with metabolic diseases.

In industrialized societies, the consumption of high-fat, high-cholesterol diets — known as Western-type diets (WTDs) — can lead to hypercholesterolaemia and atherosclerosis, especially in genetically predisposed individuals. The principal atherogenic lipoprotein in the blood is the low-density lipoprotein (LDL), and increased levels of LDL promote cholesterol accumulation and an inflammatory response in the artery wall, which drives the process of atherosclerosis (BOX 1). By promoting the cellular efflux of cholesterol, high-density lipoprotein (HDL) opposes this process and reduces inflammation. Increased levels of LDL lead to its entry into and retention in the arterial wall, where it may be modified by various processes such as oxidation and aggregation¹. This has two key adverse consequences: first, modified LDL functions as a ligand for macrophage pattern recognition receptors, including Toll-like receptors (TLRs), and can thereby directly trigger pro-inflammatory signalling pathways; and second, modified LDL is engulfed by macrophages, causing cellular cholesterol accumulation, which in turn amplifies TLR signalling^1–6. Increased TLR activity leads to augmented production of cytokines and chemokines, amplification of the inflammatory process and, when combined with the uptake or intracellular formation of cholesterol crystals, may lead to NLRP3 (NOD-, LRR- and pyrin domain-containing 3) inflammasome activation^7,8.

Box 1. Atherosclerosis and inflammation.

Atherosclerosis is a chronic disease of large and medium arteries in which cholesterol deposition incites a progressive macrophage-dominated inflammatory response. In atherosclerosis, the recognition and uptake of cholesterol-rich apolipoprotein B (APOB)-containing lipoproteins (mainly low-density lipoprotein (LDL) but also cholesterol-rich, partially lipolysed remnants of the triglyceride-transporting lipoproteins — that is, very low-density lipoprotein (VLDL) and chylomicrons — sometimes referred to as ‘non-HDL cholesterol’) by macrophages, especially when combined with defective high-density lipoprotein (HDL)-mediated cholesterol efflux, leads to a chronic inflammatory response involving both innate and adaptive immune responses^1,137–139. After binding to the subendothelial arterial matrix, LDL is modified by oxidation or aggregation, leading to its cellular recognition and uptake by pattern recognition receptors. Macrophages in atherosclerotic plaques may be derived from blood-borne monocytes, which are produced in the bone marrow and the spleen. Hypercholesterolaemia and cholesterol accumulation in haematopoietic stem cells (HSCs) promotes the overproduction of monocytes, which leads to their accumulation in atherosclerotic plaques; this process is opposed by HDL and cholesterol efflux pathways. In the bone marrow, cholesterol accumulation in the plasma membrane of HSCs increases the expression levels and signalling of growth factor receptors, causing the expansion of these populations and the increased production of monocytes, neutrophils and platelets^97,139,140. In mouse models of hypercholesterolaemia, HSC mobilization from the bone marrow leads to extramedullary haematopoiesis in the spleen¹⁰³, which is an important reservoir for the production of monocytes¹⁴¹ that may help to heal the heart after myocardial infarction¹⁴² but that may also contribute to atherogenesis¹⁴³.

Counter-regulatory mechanisms oppose cholesterol accumulation and inflammation in macrophages. In particular, accumulating levels of cellular cholesterol lead to the formation of specific sterols that activate the liver X receptor (LXR)–retinoid X receptor (RXR) heterodimeric transcription factors. The LXR–RXR heterodimers have a range of anti-inflammatory activities — including upregulating the expression of ATP-binding cassette transporters (ABC transporters) ABC subfamily A member 1 (ABCA1) and ABCG1, and promoting the efflux of cholesterol from macrophages — and thus may counter the amplification of TLR signalling by cellular cholesterol accumulation. ABCA1 and ABCG1 promote efflux of cholesterol onto HDL particles or onto the lipid-poor form of the main HDL protein, apolipoprotein A1 (APOA1), and initiate the process of reverse cholesterol transport (RCT), in which cholesterol is transported from peripheral tissues back to the liver via the lymphatics and bloodstream, followed by its excretion into bile and faeces^9,10. Of note, TLR activation suppresses LXR activity on its target genes, causing decreased macrophage cholesterol efflux¹¹, which probably results in an amplification of TLR signalling.

Thus, there is a feedforward mechanism in which the acute phase response effects changes in cellular cholesterol homeostasis, which amplifies the inflammatory response. More generally, the acute phase response downregulates the RCT pathway¹², which suggests that the innate immune system modifies cholesterol homeostasis as a way to amplify the inflammatory response. This may be beneficial as part of the overall immune response to infections or wound healing; however, excessive or prolonged cholesterol-facilitated immune responses can become associated with disease, notably atherosclerosis. Accordingly, chronic infections, such as with HIV-1 (REFS 13,14), and autoimmune disorders, such as systemic lupus erythematosus, rheumatoid arthritis and psoriasis¹⁵, are often associated with reduced levels of HDL, increased levels of atherogenic lipoproteins and accelerated atherosclerosis.

Although the links between cholesterol and inflammation are best exemplified by atherosclerosis, similar mechanisms may also contribute to other metabolic disorders such as obesity or to autoimmune diseases. For example, in obesity, activation of TLRs and NOD-like receptors (NLRs) on adipose macrophages in response to lipids, such as ceramides or saturated fatty acids, may lead to chronic inflammation, insulin resistance and fatty liver disease^16,17. These processes seem to be enhanced by cholesterol accumulation in adipose tissue and are reversed by the activation of RCT^18,19. In autoimmune diseases such as systemic lupus erythematosus and rheumatoid arthritis, defects in cholesterol efflux pathways seem to worsen the underlying condition^20,21. The emphasis of this Review is on alterations in immune signalling in response to cholesterol accumulation in cells that are relevant to atherosclerosis. In particular, we emphasize the relationships between inflammation and RCT, as well as the role of LXRs in counter-regulating cellular cholesterol accumulation and TLR-induced inflammatory responses.

RCT and the inflammatory response

The acute phase response inhibits RCT

The efflux of cholesterol from cholesterol-loaded macrophages in the arterial wall is primarily mediated by ABCA1 and ABCG1 (FIG. 1). These transporters promote unidirectional cholesterol efflux from macrophages onto lipid-poor APOA1 (in the case of ABCA1) or HDL (in the case of ABCG1), which initiates the process of RCT^9,10,22–25 (FIG. 1). HDL is the major lipoprotein in lymph and recent studies have shown that the lymphatic system is crucial for RCT from multiple tissues, including from the atherosclerotic aortic wall^26,27. HDL-free cholesterol is esterified by the lecithin–cholesterol acyltransferase (LCAT) enzyme. Free cholesterol or cholesteryl ester in HDL may be directly cleared in the liver via scavenger receptor B1 (SRB1; also known as SCARB1), which mediates a process of selective cholesterol or cholesteryl ester uptake without degradation of HDL²⁸. In humans, cholesteryl ester transfer protein (CETP) in the plasma mediates the transfer of cholesteryl ester from HDL to triglyceride-rich lipoproteins, such as very low-density lipoprotein (VLDL) and LDL, in exchange for triglyceride²⁹ and thus facilitates RCT. The triglycerides in triglyceride-rich lipoproteins are hydrolysed by lipoprotein lipase and hepatic lipase, and the cholesteryl ester-rich remnant particles are either cleared in the liver or further metabolized to LDL. Cholesterol deposited in the liver by RCT can be recycled in the form of secreted lipoproteins or can undergo net excretion into the bile by ABCG5 and ABCG8 (REF. 30) (FIG. 1).

Figure 1. RCT and its regulation by innate immune responses.

The process of reverse cholesterol transport (RCT) is depicted and how inflammation impairs this process is described in the red boxes. Under physiological conditions, apolipoprotein A1 (APOA1), which is the major protein component of high-density lipoprotein (HDL), is secreted by the liver and the intestines, and is assembled into a pre-βHDL particle as a result of its interaction with the ATP-binding cassette transporter ABC subfamily A member 1 (ABCA1) on hepatocytes and enterocytes. ABCA1 on macrophages promotes cholesterol and phospholipid efflux onto these relatively lipid-poor pre-βHDL particles, initiating the process of RCT. ABCG1 promotes further cholesterol efflux onto HDL particles. Free cholesterol in HDL is esterified by the enzyme lecithin–cholesterol acyltransferase (LCAT), which gives rise to cholesteryl esters. Free cholesterol or cholesteryl esters in HDL may be directly cleared in the liver via scavenger receptor B1 (SRB1), which mediates a process of selective free cholesterol or cholesteryl ester uptake in which the lipid moiety of HDL is mostly removed and the protein portion is recycled into the circulation (not shown). Cholesterol deposited in the liver by RCT can either be recycled in the form of secreted triglyceride-rich, very low-density lipoproteins (VLDLs; the main protein component of which is APOB) or can undergo net excretion into the bile via ABCG5 and ABCG8. In humans, plasma cholesteryl ester transfer protein (CETP) mediates the exchange of cholesteryl esters in HDL with triglyceride in VLDL. A lipolytic cascade mediated by lipoprotein lipase and hepatic lipase causes hydrolysis of triglycerides and results in the formation of cholesterol-rich and cholesteryl ester-rich LDL. Although most LDL is cleared in the liver, LDL may supply cholesterol to peripheral tissues and a small proportion is taken up into the arterial wall, where it is modified by oxidation or aggregation, leading to its uptake by macrophages. Modified LDL in the artery wall promotes Toll-like receptor (TLR) signalling in macrophages and it is taken up by these cells, leading to the formation of macrophage foam cells, the production of myeloperoxidase (MPO) and inflammation. IDL, intermediate-density lipoprotein; LDLR, LDL receptor.

The acute phase response induced following lipopolysaccharide (LPS) injection of mice blocks RCT at multiple points¹² (FIG. 1). LPS induces the expression of the microRNA miR-33, which causes a reduction in the levels of ABCA1 and ABCG1, and reduces cholesterol efflux from macrophages³¹. LPS treatment also causes decreased production of APOA1 in the liver¹², suppresses hepatic CETP gene expression³² and reduces excretion of cholesterol from the liver via the downregulation of ABCG5, ABCG8 and cholesterol 7α-hydroxylase (CYP7A1) expression^12,33. Taken together, these studies show that LPS injection results in impaired RCT, with the block occurring primarily in cholesterol mobilization from the liver to the intestines³³. Furthermore, in humans, HDL binds to LPS and thus blocks its ability to stimulate TLR4 signalling in monocytes and macrophages³⁴; thus, the suppression of APOA1 production and HDL levels by LPS¹² is pro-inflammatory.

In addition to blocking RCT, LPS causes accumulation of triglycerides within VLDL, which has been attributed to increased hepatic VLDL production and a reduction in the clearance of VLDL from the bloodstream by lipoprotein lipase, a phenomenon that has been designated the ‘lipaemia of sepsis’ (REF. 35). VLDL might sequester LPS, viruses and other toxic compounds to improve host survival, and it is possible that the increase in VLDL might help to maintain lipid levels in peripheral tissues to help suppress infection and to allow for tissue repair.

HDL becomes dysfunctional and pro-inflammatory

During acute infections, there are pro-inflammatory changes in HDL that parallel decreases in the ability of HDL to mediate cholesterol efflux from macrophages. HDL is converted from an anti-inflammatory lipoprotein that suppresses monocyte adhesion to the endothelium into a pro-inflammatory form that does not suppress monocyte adhesion during the acute phase response, for example, during that induced by influenza virus infection in humans^36,37. Furthermore, in patients with acute sepsis, cholesterol efflux from cultured macrophages to plasma or to HDL is greatly decreased³⁸. Studies using HDL isolated from healthy individuals following low-dose LPS challenge have shown a decreased cholesterol efflux capacity associated with accumulation of serum amyloid A in HDL, decreased LCAT expression and a decrease in HDL phospholipid content, even without major changes in HDL cholesterol or APOA1 levels^33,39. The anti-inflammatory and anti-oxidative activities of HDL are also impaired as a result of the loss of antioxidant proteins and the accumulation of oxidized phospholipids in HDL³⁷.

Reductions in HDL function through oxidation can also be mediated by macrophage myeloperoxidase (MPO), which is induced by inflammatory stimuli in atherosclerotic lesions^40,41 (FIG. 1). Although mice have much lower expression levels of macrophage MPO than humans, overexpression of human MPO in macrophages in LDL receptor (LDLR)-deficient mice promotes atherosclerosis⁴². Human APOA1 that has been oxidized by MPO has an impaired ability to promote cholesterol efflux via ABCA1 on macrophages and has a decreased ability to promote RCT and atherosclerosis regression when injected into mice⁴³. Individuals with coronary artery disease had higher levels of MPO-modified APOA1 than control individuals⁴⁴; their HDL had an impaired capacity to induce ABCA1-mediated cholesterol efflux from macrophages ex vivo. The levels of modified APOA1 in individuals with coronary artery disease were inversely related to ABCA1 efflux capacity and positively correlated with atherosclerotic disease. Although the plasma levels of MPO-oxidized APOA1 represent a small proportion of the total APOA1, in inflammatory microenvironments within atherosclerotic plaques, the amounts of damaged APOA1 might be sufficient to considerably impair macrophage cholesterol efflux. These observations suggest that a macrophage inflammatory response that is mediated by MPO leads to local inactivation of APOA1 in the arterial wall, reduced macrophage cholesterol efflux and increased atherosclerosis.

Huang et al.⁴⁵ recently developed a high-affinity monoclonal antibody that specifically recognizes both APOA1 and HDL that have been modified by MPO. An oxindolyl alanine (2-OH-Trp) moiety at tryptophan 72 of APOA1 (oxTrp72-APOA1) was found to be the functionally relevant, immunogenic epitope. APOA1 containing this modification was found at very low levels in the circulation of individuals with coronary heart disease but accounted for 20% of the APOA1 in atherosclerosis-laden arteries. OxTrp72-APOA1 that was recovered from human atheromas was almost entirely devoid of cholesterol acceptor activity and also had pro-inflammatory activity on endothelial cells⁴⁵. Elevated oxTrp72-APOA1 levels in patients presenting to a cardiology clinic were associated with increased cardiovascular disease risk, which suggests that this could be a biomarker for inflamed atherosclerotic plaques.

In summary, the acute phase response leads to blockage of the RCT process at multiple points. We speculate that this in turn leads to cholesterol accumulation in macrophages and to an enhancement of inflammatory responses. In addition, HDL levels are decreased during the acute phase response and compositional changes in HDL, including MPO-mediated modifications of APOA1, may convert HDL into a dysfunctional form that cannot efficiently mediate cholesterol efflux (or could even perhaps deliver cholesterol to cells) and that becomes pro-inflammatory. Although these changes in HDL and APOA1 are likely to be pro-atherogenic, we believe they may also have a physiological function in the setting of infection by enhancing the inflammatory response. This contention is supported by the observation that mice with macrophage-specific deletion of Abca1 are resistant to Listeria monocytogenes infection⁴⁶. Moreover, in cells involved in wound healing and tissue repair, maintaining cholesterol in the cell may facilitate enhanced cell proliferation^47,48.

In terms of the mechanisms connecting inflammation with decreases in RCT, many of the genes involved in RCT are induced by cellular cholesterol accumulation and the activation of LXR transcription factors. Activation of TLR3 or TLR4 signalling suppresses expression of LXR target genes in macrophages, including Abca1 and Abcg1, via induction of the transcription factor interferon-regulatory factor 3 (IRF3) (REF. 11). In addition, LPS suppresses expression of Lxr and Rxr in hepatocytes⁴⁹. Thus, suppression of LXR–RXR may be a general mechanism connecting TLR-mediated inflammatory responses to decreased RCT.

Opposing effects of cholesterol in macrophages

There is abundant evidence that the interaction of LDL with macrophages in atherosclerotic plaques leads to an increase in inflammatory gene expression. In hypercholesterolaemic mouse models of atherosclerosis, inflammatory signalling in macrophages and endothelial cells — via TLR2 and TLR4, and myeloid differentiation primary response protein 88 (MYD88)-dependent pathways — promotes cytokine and chemokine gene expression, and atherogenesis^50–52. Macrophage foam cells in progressive atherosclerotic plaques have increased expression of inflammatory genes compared with plaques undergoing regression⁵³. Human atherosclerotic plaques express increased levels of cytokines and chemokines that are dependent on MYD88-mediated signalling via various TLRs, especially TLR2 (REF. 54). The most important ligands for TLRs are probably modified forms of LDL. In mouse macrophages, TLR4 and TLR6 function in combination with CD36 to mediate macrophage inflammatory responses². In contrast to these pro-inflammatory effects of the interaction of modified LDL with macrophages, cholesterol efflux pathways suppress TLR signalling and inflammatory cytokine expression in atherosclerotic plaques^4,55,56.

Somewhat unexpectedly, peritoneal macrophages isolated from WTD-fed Ldlr^−/− mice had an overall reduction in inflammatory gene expression compared with macrophages from wild-type mice that were fed a control diet⁵⁷. This reduction in inflammatory gene expression was related to the accumulation of desmosterol, which is the penultimate molecule in the cholesterol biosynthetic pathway and is a known activator of LXRs⁵⁸. Desmosterol probably accumulates in peritoneal macrophage foam cells because the expression of desmosterol reductase, which is the last enzyme in the cholesterol biosynthetic pathway, is more repressed by cellular cholesterol loading than the enzymes that are involved in the earlier steps in the pathway, leading to disproportionate biosynthesis of desmosterol. These observations of suppressed inflammation in peritoneal macrophage foam cells highlight the importance of the specific environment of the arterial wall in the macrophage inflammatory response in atherosclerosis. The artery provides a unique environment for the interaction of LDL with macrophages, which results both in the activation of TLRs and in cholesterol accumulation in macrophages, leading to progressive atherosclerosis. The net effect of macrophage cholesterol accumulation on inflammation may thus depend on a balance of factors, such as the activation of TLRs, the localization of accumulating cholesterol within the cell, cholesterol crystal formation and inflammasome activation, and the phenotype of the macrophages involved.

Inflammatory effects of cholesterol accumulation

Promotion of TLR signalling

There is evidence that plasma membrane cholesterol enrichment promotes the formation of TLR4–MD2 (REF. 4) and TLR4–CD14 complexes⁵⁹, which enhances the response to TLR4 ligands such as LPS. Conversely, ABCA1 and ABCG1 promote macrophage cholesterol efflux and suppress macrophage inflammatory responses via TLR2, TLR3 and TLR4 (REF. 4). This may involve decreased formation of cholesterol-enriched lipid rafts in the plasma membrane and in the endosomal system. Mice that are deficient in ABCA1 and ABCG1 accumulate free and esterified cholesterol in peritoneal macrophages, and have enhanced inflammatory responses to TLR ligands and increased apoptosis when exposed to oxidized LDL³. This is in contrast to the observations of decreased inflammatory gene expression in peritoneal macrophage foam cells from Ldlr^−/− mice⁵⁷, which indicates that the mode of accumulating free cholesterol within the cell may be important in determining the inflammatory response. Even without cholesterol efflux, ABCA1 and ABCG1 promote the trans-bilayer movement of cholesterol at the plasma membrane⁶⁰; when this is defective in Abca1^−/−Abcg1^−/− macrophages, inflammatory changes ensue. Macrophage-specific deficiency in ABCA1 and ABCG1 resulted in increased atherosclerosis, and laser capture microdissection of macrophage-rich areas of plaques shows increased expression of the inflammatory chemokines CC-chemokine ligand 2 (CCL2) and CCL3 (REF. 56), which confirms that cholesterol efflux pathways in macrophages have anti-inflammatory activity in the context of the atherosclerotic plaque.

Inflammasome activation

Accumulating cholesterol in atherosclerotic plaques may give rise to cholesterol crystal formation. Cholesterol crystal uptake or formation in macrophages has been shown to promote inflammasome activation and atherogenesis. Inflammasome activation requires two signals — a priming signal that is typically mediated by TLR4 activation and a second signal involving potassium ion efflux, lysosomal damage or reactive oxygen species (ROS) generation⁶¹ (FIG. 2). The priming signal may result from pattern recognition receptor activation — for example, the induction of TLR4–TLR6–CD36 signalling by oxidized LDL⁸ — whereas the second signal can be mediated by cholesterol crystals in lysosomes, either as a result of phagocytosis of extracellular cholesterol crystals or via CD36-mediated uptake of modified LDL and free cholesterol release from the LDL⁸. Inflammasome activation leads to the secretion of the pro-inflammatory cytokines interleukin-1β (IL-1β) and IL-18.

Figure 2. Inflammasome activation by sterols.

The NLRP3 (NOD-, LRR- and pyrin domain-containing 3) inflammasome is activated by many signals, including infectious agents and stress- or injury-induced host factors, leading to caspase 1 activation, cleavage of pro-interleukin-1β (pro-IL-1β) and pro-IL-18, and the secretion of the active cytokines, as well as in some cases resulting in a form of cell death termed pyroptosis. A priming stimulus (signal 1), acting through nuclear factor-κB (NF-κB), induces the expression of Il1b, Il18 and Nlrp3, and precedes assembly of the inflammasome complex. In atherosclerosis, priming may result from pattern recognition receptor activation; for example, combinatorial Toll-like receptor 4 (TLR4)–TLR6–CD36 signalling induced by oxidized low-density lipoprotein (LDL). The second stimulus (signal 2) may arise from various stressors. In atherosclerosis, one such signal is lysosomal damage or dysfunction, which may result from phagocytosis of extracellular cholesterol crystals via the CD36-mediated uptake of modified LDL (not shown) and free cholesterol release and crystallization in lysosomes. Other signals may result from mitochondrial damage, oxidative stress-induced production of reactive oxygen species (ROS) or ATP release from dying cells. 25-hydroxycholesterol (25-OH cholesterol) suppresses inflammasome activation by reduced sterol regulatory element-binding protein 2 (SREBP2) cleavage mediated by SREBP cleavage-activating protein (SCAP) in the endoplasmic reticulum (ER). INSIG1, insulin-induced gene 1; MYD88, myeloid differentiation primary response protein 88; TRIF, TIR domain-containing adaptor protein inducing IFNβ.

Duewell et al.⁷ used confocal reflectance microscopy to show refractile material in early mouse atherosclerotic lesions, interpreted as small free cholesterol crystals. They showed decreased lesion formation in WTD-fed Ldlr^−/− mice following irradiation and transplantation of bone marrow that was deficient in key components of the NLRP3 inflammasome or in Il1a and Il1b. These data suggest that cholesterol crystals and minimally oxidized LDL could induce inflammasome activation in macrophages and the authors proposed a prominent role for cholesterol crystals in activating the inflammasome, even in early atherosclerotic lesions. In addition, increased atherosclerosis in Apoe^−/− mice in which immune cells were deficient in autophagy protein 5 (ATG5) was proposed to reflect inflammasome activation⁶². This could be due to the ability of autophagy to suppress the inflammasome^63,64 or to promote cholesterol efflux and thus to oppose cholesterol crystal formation⁶⁵. Some but not all studies have shown that IL-1β signalling is pro-atherogenic in mouse models of atherosclerosis⁶⁶. In addition, IL-18 levels in the blood predict cardiovascular death in patients with coronary heart disease⁶⁷. Administration of IL-18 to Apoe^−/− mice increased the size of atherosclerotic lesions, and their T cell content and MHC class II expression, in an interferon-γ (IFNγ)-dependent manner⁶⁸, whereas antagonism of IL-18 reduced atherosclerosis⁶⁹. Thus, several lines of evidence support an important role of cholesterol crystal-induced inflammasome activation and inflammasome products in mouse models of atherosclerosis and possibly in human coronary heart disease.

In contrast with the above studies, Menu et al.⁷⁰ found no effect on atherosclerosis in Apoe^−/− mice deficient for components of the NLRP3 inflammasome that had been fed a WTD. Nevertheless, more recent reports have confirmed an anti-atherogenic effect of caspase 1 (REFS 71,72) or NLRP3 deficiency⁷³ in Apoe^−/− mice. The lack of effect on atherosclerosis reported by Menu et al.⁷⁰ may have been due to the analysis of very advanced atherosclerotic lesions⁶⁶. Although numerous reports support the activation of the NLRP3 inflammation by cholesterol crystals taken up by macrophages, some literature seems to be inconsistent with the presence of free cholesterol crystals in early atherosclerotic lesions⁷. Small et al.^74,75 analysed the composition of early human atherosclerotic plaques and, on the basis of the physical state of lipids that have been described in model systems, did not predict that a separate free cholesterol crystalline phase occurred in humans. Furthermore, a cholesterol crystalline phase was not observed using a polarized light microscope with a heating and cooling stage. During these studies, it was noted that plaque cholesteryl esters that were always liquid or liquid crystalline at body temperature readily formed small crystals upon cooling, raising the issue that cholesteryl ester crystal formation could be an artefact of refrigerated samples of atherosclerotic arteries. However, Rothblat and colleagues^76,77 produced authentic, highly elongated free cholesterol crystals within lysosomes of peritoneal macrophages and the formation of these crystals was reversed by addition of APOA1 or APOE to the cultures. The authors found that substantial crystal formation required extensive cholesterol loading using acetylated LDL and the presence of an inhibitor of cholesterol esterification.

The weight of evidence supports an important role for cholesterol- or oxidized LDL-induced inflammasome activation in atherosclerosis; however, the effect of inflammasome activation at different stages of the disease and the underlying mechanisms deserve further investigation. Although the major focus of these studies has been on the links between inflammasome activation in cholesterol-loaded macrophages and atherosclerosis, we speculate that cholesterol accumulation in macrophages as a result of innate immune responses may contribute to inflammasome activation in the spleen during microbial infection, which could have beneficial effects on the outcome.

25-hydroxycholesterol and the inflammasome

In macrophages and dendritic cells, the enzyme that synthesizes 25-hydroxycholesterol (25-OH cholesterol) — cholesterol 25-hydroxylase — is induced by type I IFNs downstream of TLR3 or TLR4 activation^78–80. 25-OH cholesterol has broad activity against enveloped viruses — including HIV, herpes simplex virus and Ebola virus — by preventing fusion of the viral membrane with cells⁸¹, but it can also lead to detrimental tissue damage in some settings, such as during infection with influenza virus⁸². A recent report shows that 25-OH cholesterol decreases inflammasome activation in macrophages⁸³. Accordingly, cholesterol 25-hydroxylase-deficient mice showed increased sensitivity to septic shock and exaggerated autoimmune encephalomyelitis but showed a stronger ability to repress bacterial growth after bacterial infection⁸³. Decreased cholesterol and/or sterol accumulation following 25-cholesterol hydroxylase induction is associated with decreased expression of Il1b mRNA and decreased caspase 1 activation. The underlying mechanism probably involves the suppression of cholesterol biosynthetic genes by 25-OH cholesterol. The expression of these genes is controlled by the nuclear active form of the transcription factor sterol regulatory element-binding protein 2 (SREBP2). 25-OH cholesterol binds to insulin-induced gene 1 (INSIG1) protein in the endoplasmic reticulum (ER), which prevents the SREBP cleavage-activating protein (SCAP)-mediated transport of SREBP2 from the ER to the Golgi and its subsequent cleavage into the nuclear active form⁸³. This study implies that following TLR activation and the induction of cholesterol 25-hydroxylase, decreased active SREBP2 and decreased accumulation of cholesterol or other sterols leads to inflammasome suppression. This is in contrast to the idea presented above that the acute phase response promotes macrophage cholesterol accumulation, leading to an enhanced inflammasome response. However, these ideas can be reconciled by proposing that the effects occur sequentially and that the immune system uses these various changes in macrophage cholesterol homeostasis to activate the inflammasome and then to turn it off.

Although these studies imply that increased SREBP2 activity may promote caspase 1 activation⁸³, SREBPs may in turn be cleaved and activated by caspase 1 — for example, when pore-forming bacterial toxins activate the inflammasome⁸⁴ — which suggests that there may be a positive feedback loop that promotes inflammasome activation and lipid synthesis for membrane repair. In endothelial cells, oscillatory blood flow induces inflammasome activation through the SREBP2-induced expression of Nlrp3 and Casp1 (the gene encoding caspase 1) via direct promoter activation; this induction seems to be independent of the effects of SREBP2 on cellular sterols⁸⁵.

Oscillatory or turbulent blood flow occurs in regions of the aorta such as the lesser curvature of the aortic arch and at vessel branch points. These sites are susceptible to atherosclerosis and, because of their altered blood flow, may also be adapted for the capture of pathogens by antigen-presenting cells⁸⁶. We speculate that inflammasome activation at sites of disturbed arterial blood flow, perhaps involving both endothelial cells and resident arterial macrophages⁸⁷, may have evolved to promote the clearance of pathogens; however, under conditions of hypercholesterolaemia, these same responses may promote atherogenesis.

Counteracting cholesterol-mediated inflammation

As mentioned above, LXR transcription factors exert potent anti-inflammatory effects and thus provide a mechanism that counter-regulates modified LDL-induced TLR activation and macrophage cholesterol accumulation (FIG. 3). Multiple mechanisms are likely to be involved.

Figure 3. Molecular mechanisms underlying anti-inflammatory effects of LXR activation.

Liver X receptors (LXRs) promote macrophage cholesterol efflux via the induction of ABC subfamily A member 1 (ABCA1) and ABCG1 expression, which suppresses Toll-like receptor (TLR)-mediated inflammatory responses, possibly by disrupting membrane lipid rafts (labelled 1 in the figure). LXRs induce the expression of genes mediating elongation and unsaturation of fatty acids, leading to the synthesis of long-chain polyunsaturated fatty acids (PUFAs) including omega 3 fatty acids, as well as specialized pro-resolving lipid mediators. Long-chain PUFAs mediate decreased transcriptional responses of nuclear factor-κB (NF-κB) target genes as a result of altered histone acetylation in their enhancer and/or promoter regions, without changes in nuclear p65 levels (labelled 2 in the figure). Activation of LXRs by desmosterol and oxysterols causes sumoylation (Su) of specific residues in the ligand-binding pocket of LXR, leading to binding of LXR (without retinoid X receptor (RXR)) to NF-κB and AP-1 response elements, blunting the inflammatory responses that are mediated by these transcription factors (labelled 3 in the figure). LXRs increase expression of the tyrosine protein kinase MER (MERTK), which enhances the uptake of apoptotic cells by macrophages (in a process known as efferocytosis; labelled 4 in the figure) and this leads to a suppression of TLR4-mediated inflammatory responses (not shown). Efferocytosis also causes marked LXR-dependent upregulation of ABCA1 and ABCG1, which is probably an important contributor to the anti-inflammatory effect. LXRE, LXR response element; RXRE, RXR response element.

First, LXRs promote cholesterol efflux from macrophages via the induction of ABCA1 and ABCG1, and given that cholesterol efflux via these transporters suppresses TLR-mediated inflammatory responses^3,4, this seems to be a key anti-inflammatory effect of LXR activation.

Second, LXRs induce the expression of several genes that mediate the elongation and the unsaturation of fatty acids, which leads to synthesis of long-chain polyunsaturated fatty acids (PUFAs) including omega 3 fatty acids⁸⁸. The increase in long-chain PUFAs was shown to result in decreased transcriptional responses of nuclear factor-κB (NF-κB) target genes as a result of altered histone acetylation in their enhancer and/or promoter regions, without changes in nuclear p65 levels⁸⁸. Long-chain PUFAs may also function as substrates for the enzymes that synthesize eicosanoids and specialized pro-resolving lipid mediators, such as resolvins and protectins, that promote the resolution of inflammation⁸⁹. Moreover, LXRα in the liver induces lysophospholipid acyltransferase 5 (LPCAT3), which is an enzyme that mediates the synthesis of phospholipids containing long-chain PUFAs, leading to decreased ER stress and inflammatory responses⁹⁰.

Third, activation of LXRs by desmosterol and other oxysterols causes the sumoylation of specific residues in their ligand-binding pocket, leading to the binding of LXR (without RXR) to NF-κB and AP-1 response elements, which reduces the inflammatory responses that are mediated by these transcription factors⁹¹.

Fourth, LXRs increase expression of the tyrosine protein kinase MER (MERTK), which enhances the uptake of apoptotic cells by macrophages (a process termed efferocytosis) and this leads to a suppression of TLR4-mediated inflammatory responses⁹². Efferocytosis also causes marked LXR-dependent upregulation of ABCA1 and ABCG1 (REF. 93), which is probably an important contributor to the anti-inflammatory effect.

Finally, LXRs are highly expressed by haematopoietic stem cells (HSCs) and myeloid progenitor cells, in which they promote cholesterol efflux via upregulating the expression of ABCA1, ABCG1 and APOE, and decrease the proliferative responses of these cells to IL-3 and granulocyte–macrophage colony-stimulating factor (GM-CSF), thus reducing the production of inflammatory cells⁹⁴.

In summary, although TLR activation reduces cholesterol efflux and promotes cholesterol accumulation in macrophages, enhancing the inflammatory response, cholesterol accumulation also leads to LXR activation and the eventual suppression of TLR-mediated inflammatory responses. This may be analogous to the role of the TLR-mediated induction of cholesterol 25-hydroxylase in eventually turning off the inflammasome response, as suggested above. Thus, the innate immune system uses changes in cholesterol metabolism to amplify the inflammatory response and then to restore homeostasis.

Increased production of inflammatory cells

The discussion so far has mainly focused on the links between cholesterol accumulation in macrophages and inflammatory responses in atherosclerotic plaques; however, another important connection occurs at the level of the bone marrow. In both cross-sectional and prospective human population studies, blood monocyte and neutrophil numbers are strongly associated with atherosclerotic cardiovascular disease⁹⁵, and studies in animal models indicate a causal relationship^87,96, which suggests that the excessive production of inflammatory cells in the bone marrow and spleen under hypercholesterolaemic conditions is important in the atherogenic process (FIG. 4).

Figure 4. Hypercholesterolaemia and defective cholesterol efflux promote myelopoiesis and atherosclerosis.

In the bone marrow, increased plasma membrane cholesterol content in haematopoietic stem cells (HSCs) and myeloid progenitor cells as a result of defective cholesterol efflux promotes increased cell surface levels of the common β-subunit (CBS) of the interleukin-3 (IL-3), IL-5 and granulocyte–macrophage colony-stimulating factor (GM-CSF) receptors and increased proliferation in response to these growth factors. Extramedullary haematopoiesis can also occur after HSCs progressively relocate from the bone marrow to the splenic red pulp. Efferocytosis in the setting of defective cholesterol efflux in macrophages fails to suppress the production of IL-17, IL-23 and granulocyte colony-stimulating factor (G-CSF), and these cytokines can promote HSC relocation to the spleen. In the spleen, the number of GM-CSF-producing innate response activator B cells (IRA B cells) increases in mice with hypercholesterolaemia, which causes increased production of monocytes and neutrophils from HSCs. In addition, defective cholesterol efflux in splenic macrophages can promote the development of monocytes from HSCs and myeloid progenitor cells in the bone marrow via macrophage colony-stimulating factor (M-CSF) and CC-chemokine ligand 2 (CCL2). Monocytes, especially LY6C^hi monocytes produced in the bone marrow and the spleen, enter the bloodstream and accumulate in atherosclerotic lesions. LDL, low-density lipoprotein; ROS, reactive oxygen species.

Increased bone marrow myelopoiesis promotes monocytosis

In mice, cholesterol efflux pathways mediated by APOE, ABCA1 and ABCG1 suppress the production of inflammatory cells in the bone marrow and the spleen; this is observed in chow-fed mice and becomes more prominent in the setting of hypercholesterolaemia^94,97. Transplantation of Abca1^−/−Abcg1^−/− bone marrow cells into Ldlr^−/− mice led to dramatic monocytosis and neutrophilia, infiltration of myeloid cells into multiple organs and accelerated atherosclerosis⁹⁷. This leukocytosis reflected markedly increased proliferation and expansion of the HSC population in the bone marrow. Increased HSC proliferation, myelopoiesis and atherogenesis were reversed by overexpression of transgenic human APOA1, which probably reflects increased cholesterol efflux from HSCs, possibly by SRB1-facilitated passive efflux⁹⁸. Subsequent studies showed similar expansion and proliferation of the HSC population in Apoe^−/− mice and to a lesser extent in Ldlr^−/− mice⁹⁴. In Ldlr^−/− mice, reductions in APOA1 and HDL levels because of Apoa1 haploinsufficiency promoted HSC population expansion and monocytosis⁹⁹. LDL or oxidized LDL seem to promote HSC proliferation and this proliferation is reversed by HDL^99,100. Similarly, children with familial hypercholesterolaemia have an inverse relationship between HDL cholesterol levels and blood monocyte numbers, which suggests that these mouse studies are relevant to humans⁹⁹.

On a mechanistic level, HSCs from Apoe^−/− or Abca1^−/−Abcg1^−/− mice showed evidence of increased plasma membrane cholesterol, increased cell surface levels of the common β-subunit of the IL-3, IL-5 and GM-CSF receptors (known as cytokine receptor common β-subunit (CBS); encoded by Csf2rb) and increased proliferation in response to these cytokines^94,97. The transfer of bone marrow cells from Apoe^−/− mice deficient in CBS into Ldlr^−/− mice led to a reduction in the number of HSCs and myeloid progenitor cells in the bone marrow and the spleen, a reduction in monocyte and neutrophil numbers in the blood, and reduced macrophage numbers in atherosclerotic lesions¹⁰¹. Thus, in the setting of dietary or genetic hypercholesterolaemia, cholesterol efflux pathways that are mediated by APOE, ABCA1 and ABCG1 in HSCs prevent increases in membrane cholesterol content and in cell-surface expression levels of CBS, and thus prevent the hyperproliferation of HSCs and excessive myelopoiesis.

Recent evidence suggests that WTDs may also induce epigenetic changes in HSCs or myeloid progenitor cells that lead to increased myelopoiesis¹⁰². In bone marrow transplantation experiments using Ldlr^−/− mice on a chow diet as recipient animals, bone marrow from mice fed a WTD caused increased myeloid cell proliferation compared with bone marrow from chow-fed mice. Bone marrow cells from mice fed a WTD showed hypomethylation of CpG regions in the genes encoding PU.1 and IRF8, which are key regulators of monocyte proliferation and macrophage differentiation¹⁰². In the blood of recipients of bone marrow from WTD-fed mice, there was increased monocytosis, their spleens were enlarged and the percentage of F4/80⁺CD86⁺ macrophages was increased¹⁰². Interestingly, mice reconstituted with bone marrow from WTD-fed mice showed an increase in aortic root plaque size in the absence of changes in serum cholesterol. Although the underlying mechanisms have not yet been established, these findings suggest that a WTD induces transplantable epigenetic changes in cells in the bone marrow that promote the production of inflammatory cells and increase susceptibility to atherosclerosis.

Cholesterol loading in the spleen promotes myelopoiesis in the bone marrow

In addition to the cell autonomous effects of cholesterol accumulation in HSCs, cholesterol build-up in splenic macrophages and dendritic cells promotes bone marrow myelopoiesis and HSC mobilization. In Apoe^−/− and Abca1^−/− Abcg1^−/− mice, macrophages and dendritic cells in the spleen accumulate cholesterol and secrete inflammatory cytokines, which stimulates the production of inflammatory cells in the bone marrow, as well as HSC mobilization and extramedullary haematopoiesis¹⁰³. Studies in LXR-deficient mice show that the activation of LXRs in splenic macrophages by the uptake of senescent neutrophils represses the production of the pro-inflammatory cytokines IL-17, IL-23 and granulocyte colony-stimulating factor (G-CSF), which controls neutrophil production in the bone marrow¹⁰⁴. Similar findings in mice that have a knockout of Abca1 and Abcg1 in dendritic cells or macrophages indicate that these effects of LXRs are probably mediated, at least partly, by upregulation of these transporters and by the promotion of cholesterol efflux¹⁰³.

In addition to effects on neutrophils, ABCA1- and ABCG1-mediated cholesterol efflux in splenic dendritic cells and macrophages leads to the suppression of monocyte numbers apparently via reduced production of macrophage colony-stimulating factor (M-CSF) and CCL2, and diminished bone marrow myelopoiesis⁵⁶. The expression of Abca1 and Abcg1 is strongly upregulated in macrophages during efferocytosis⁹³, which probably leads to suppression of TLR-dependent inflammatory responses, including the production of IL-23. In Abca1^−/−Abcg1^−/− and Apoe^−/− mice, activation of the IL-17, IL-23 and G-CSF signalling axis in the spleen also leads to the release of HSCs from the bone marrow and to extramedullary haematopoiesis¹⁰³. Thus, hypercholesterolaemia and defective cholesterol efflux promote monocyte and neutrophil production through both cell-autonomous and cell-extrinsic mechanisms in the bone marrow and the spleen, involving HSC and myeloid progenitor cell proliferation, HSC mobilization and extramedullary haematopoiesis. Although these pathways probably enhance the response to infections, genetic alterations and dietary challenge lead to aberrant responses that promote atherogenesis.

The spleen as a site of extramedullary haematopoiesis

Robbins et al.¹⁰⁵ first showed that in Apoe^−/− mice fed a WTD, HSCs progressively relocate from the bone marrow to the splenic red pulp, where they encounter GM-CSF and IL-3, which promote myelopoiesis and the production of LY6C^hi monocytes. Monocytes produced in the spleen enter the bloodstream, accumulate in atherosclerotic lesions and secrete pro-inflammatory cytokines, ROS and proteases¹⁰⁵. A minor population of B cells in the mouse spleen, termed innate response activator B cells (IRA B cells), which are derived from peritoneal B1a cells, is mainly responsible for splenic GM-CSF production¹⁰⁶. IRA B cells have an essential role in supporting the increased production of monocytes and neutrophils, and in the survival of mice with polymicrobial sepsis¹⁰⁶. The splenic IRA B cell population is also expanded in WTD-fed Apoe^−/− mice^101,107 and genetic depletion of this population showed a pro-atherogenic role for these cells¹⁰⁷. This was attributed to a decrease in the production of conventional dendritic cells — a process that is known to be dependent on GM-CSF — and to the reduced differentiation of IFNγ-secreting T helper 1 (T_H1) cells¹⁰⁷.

In Apoe^−/− mice, the IRA B cell population is dependent on MYD88 signalling¹⁰⁷ and has increased levels of CBS on their cell surface, which is probably secondary to defective cholesterol efflux¹⁰¹; CBS deficiency reduced the size of the B1a and IRA B cell populations¹⁰¹. Thus, in Apoe^−/− mice, increased signalling of IL-3 and GM-CSF, mediated through increased levels of CBS, has a key role both in HSCs and in IRA B cells, promoting expansion of these populations, increased production of monocytes and neutrophils in the bone marrow and spleen, and probably also influencing dendritic cell maturation and T_H cell differentiation. Apoe gene expression is increased by LXR activation¹⁰⁸ and is suppressed by LPS¹⁰⁹, and thus may be decreased in IRA B cells during the acute phase response, leading to decreased cholesterol efflux, increased CBS expression and enhanced GM-CSF production. As IRA B cells have an essential role in the clearance of microorganisms¹⁰⁶, as a result of their production of GM-CSF, these data represent another example of how cellular cholesterol accumulation may lead to an enhanced immune response with beneficial effects in the setting of acute infection.

Targeting cholesterol and inflammation

LXR activators reduce atherosclerosis in mouse models¹¹⁰, which is probably a result of both cholesterol efflux and anti-inflammatory effects^111,112. Although molecules that activate LXRs have been developed as anti-atherogenic drugs, their progress in the clinic has been hampered by unwanted side effects such as fatty liver disease and LDL elevation¹¹³. Other potential therapeutic applications of LXR activators involving their anti-inflammatory effects are being evaluated, for example, in autoimmune diseases⁹² and skin diseases¹¹⁴. Recent studies have uncovered specific roles of cholesterol derivatives such as 25-OH cholesterol and 7α,25-OH cholesterol in the immune response to infections, providing new insights into the intimate links between sterol metabolism and immunity, and opening the possibility of sterol-targeted interventions as a means of immunosuppression^83,115,116.

LDL cholesterol can be lowered by statins and these drugs remain the mainstay of treatment in atherosclerosis. However, there is a large burden of residual disease in individuals treated with statins, which indicates the need for new therapies. Increasing the production or infusion of APOA1-containing HDL consistently reduces atherosclerosis in animal models^117–119 and infusion of APOA1–phosphatidylcholine complexes seems to reduce coronary atherosclerosis in humans^120,121. Moreover, the cholesterol efflux capacity of human plasma HDL is inversely correlated with atherosclerotic burden in the coronary and carotid arteries¹²². These anti-atherogenic effects are probably at least partly related to the anti-inflammatory effects of HDL. In addition to suppressing the production of inflammatory cells, HDL suppresses the expression of tumour necrosis factor (TNF)-induced cell adhesion molecules on endothelial cells that promote the entry of inflammatory cells into plaques¹²³, and HDL pretreatment reduces LPS or oxidized LDL-induced inflammatory responses in macrophages^124–126. Although part of the effect of HDL on inflammatory responses may be mediated through cholesterol efflux and the disruption of TLR-mediated signalling, there may be additional mechanisms such as the induction of the transcription factor ATF3, which function to suppress NF-κB signalling¹²⁶ and atherogenesis¹²⁷.

On a practical level, it has so far proved difficult to either increase endogenous APOA1 production or to infuse HDL in sufficient amounts to achieve a therapeutic benefit in humans. Infusions of fairly small amounts of sphingomyelin–APOA1 complexes did not reduce the volume of coronary atherosclerotic plaques¹²⁸. Although larger amounts of phosphatidylcholine–APOA1 complexes reduced plaque volume compared with baseline levels, there were side effects that may have been related to the presence of bile salts in the preparations or to excessive removal of cholesterol from tissues¹²¹.

Improved preparations that are efficacious in mediating cholesterol efflux from macrophages are undergoing clinical evaluation¹²⁹. For example, pegylation of the HDL particle (as opposed to free APOA1) results in preservation of the ability of APOA1 to promote cholesterol efflux via ABCA1, reduced catabolism of APOA1 in vivo and improved anti-atherogenic efficacy, with the potential to allow lower doses of HDL to be infused but retaining efficacy¹³⁰. RVX-208 is a small molecule that functions as a bromodomain and extraterminal (BET) domain inhibitor, which displaces BET domains from chromatin and increases transcription of the human APOA1 gene, and thus increases APOA1 and HDL-cholesterol levels in humans^131–133. Although the magnitude of these responses is too small to be clinically relevant^131–133, there is the potential to identify more potent compounds that function through a similar epigenetic mechanism. Moreover, recent studies have identified a long non-coding RNA at the APOA1–APOC3–APOA4 locus that suppresses gene expression through an epigenetic mechanism¹³⁴. Targeting this locus resulted in increased APOA1 production in monkey and human hepatocytes. Future studies involving transcriptional upregulation of APOA1, improved versions of reconstituted HDL particles or perhaps APOA1-mimetic peptides^135,136 will better define their potential therapeutic benefits in patients with atherosclerosis, insulin resistance and autoimmunity.

Summary and perspective

The disruptions of cellular or organismal cholesterol homeostasis that occur as part of innate immune responses may lead to an augmentation of inflammatory responses via enhanced TLR signalling or inflammasome activation. This physiological adaptation, as exemplified by the process of HDL-mediated cholesterol efflux and RCT, becomes dysfunctional in chronic metabolic diseases such as obesity or atherosclerosis. Increasing the production of APOA1-containing HDL or activation of LXRs represent therapeutic intervention strategies that could disrupt this cycle with a potential benefit for patients with atherosclerosis, obesity, insulin resistance and autoimmune diseases.

Glossary

Low-density lipoprotein (LDL)

A 20–25 nm low-density (1.016–1.063 g ml⁻¹) lipoprotein with ~45% cholesterol, 20% phospholipids, 10% triglycerides and 25% protein (with apolipoprotein B (APOB) as the major apolipoprotein).

High-density lipoprotein (HDL)

An 8–11 nm high-density (1.063–1.210 g ml⁻¹) lipoprotein with 40–55% protein (with apolipoprotein A1 (APOA1) as the major apolipoprotein), 25% phospholipids, 15% cholesterol and 5% triglycerides. HDL particles carry cholesterol from peripheral tissues to the liver.

Liver X receptor (LXR)

LXRα and LXRβ are transcription factors that function as heterodimeric partners with retinoid X receptors (RXRs) on the promoters of many genes that are involved in cholesterol metabolism and lipogenesis. LXRs are activated by cholesterol biosynthetic intermediates, such as desmosterol, and by oxysterols derived from cholesterol. LXRs are key regulators of cellular cholesterol efflux and reverse cholesterol transport and also block the cellular uptake of low-density lipoprotein (LDL) cholesterol through the LDL receptor.

ABC transporters

A family of membrane transport proteins that use the energy of ATP hydrolysis to transport various molecules, including cholesterol and other lipids, across the membrane.

Apolipoprotein A1 (APOA1)

The liver and the intestine secrete lipid-poor APOA1, the major protein component of high-density lipoprotein (HDL) particles. APOA1 functions as an acceptor for phospholipids and cholesterol on hepatocytes, enterocytes and macrophages. Thus, it may be involved in HDL formation as well as in the efflux of cholesterol from cells.

Reverse cholesterol transport (RCT)

A multistep process that results in the net movement of cholesterol from peripheral tissues back to the liver via the blood. Cholesterol from peripheral tissues is transferred to apolipoprotein A1 (APOA1) and high-density lipoprotein (HDL) by the ATP-binding cassette transporters ABCA1 and ABCG1, respectively. The cholesteryl esters present within HDL can then be transferred, with the help of cholesteryl ester transfer protein in exchange for triglycerides, to APOB-rich lipoproteins (such as low-density lipoprotein and very low-density lipoprotein) or can be taken up in the liver by scavenger receptor B1 (SRB1). In the liver, cholesterol can be converted into bile acids for elimination.

Acute phase response

The early immune response to infection, which results in the production of cytokines and other mediators, and in an increase in the number of peripheral leukocytes.

Chylomicrons

50–200 nm diameter lowest density (<1.006 g ml⁻¹) lipoproteins that are composed of 85% triglycerides, 9% phospholipids, 4% cholesterol, and 2% protein (with apolipoprotein B48 (APOB48) as the major apolipoprotein).

Very low-density lipoprotein (VLDL)

A 30–70 nm very low-density (0.95–1.006 g ml⁻¹) lipoprotein, with ~50% triglycerides, 20% cholesterol, 20% phospholipids and 10% protein (with apolipoprotein B100 (APOB100) as the major apolipoprotein).

MicroRNA

Small RNA molecules that regulate the expression of genes by binding to the 3′-untranslated regions of specific mRNAs.

NLRP3 inflammasome

The NLRP3 (NOD-, LRR- and pyrin domain-containing 3) inflammasome consists of the NOD-like receptor NLRP3, caspase 1 and the adaptor protein ASC. It is activated by many signals, including microbial products, and stress- and injury-induced host factors, leading to caspase 1 activation, cleavage of pro-interleukin-1β (pro-IL-1β) and pro-IL-18, secretion of IL-1β and IL-18 and, in some cases, pyroptosis, which is a pro-inflammatory and lytic form of cell death.

25-hydroxycholesterol (25-OH cholesterol)

An oxysterol formed from cholesterol by the enzyme cholesterol 25-hydroxylase, which is present in the endoplasmic reticulum.

Sterol regulatory element-binding protein 2 (SREBP2)

A transcription factor that begins as a multi-transmembrane endoplasmic reticulum protein and is cleaved in the Golgi to release the basic helix–loop–helix leucine zipper transcription factor domain that binds to sterol regulatory elements in DNA.

Sumoylation

The post-translational modification of proteins that involves the covalent attachment of a small ubiquitin-related modifier (SUMO) and that regulates the interactions of those proteins with other macromolecules.

Innate response activator B cells (IRA B cells)

An effector B cell population and a transitional B1a-derived inflammatory subset that control IgM production and protect against microbial sepsis.

Statins

A family of inhibitors of hydroxymethylglutaryl-coenzyme A reductase (HMG-CoA reductase), which is an enzyme that catalyses the conversion of HMG-CoA to l-mevalonate. These molecules are mainly used as cholesterol-lowering drugs but they also have immunoregulatory and anti-inflammatory properties. l-Mevalonate and its metabolites are implicated in cholesterol synthesis and other intracellular pathways.