To cross the endothelium, apoB-containing lipoproteins can bind to activin receptor-like kinase 1 and scavenger receptor B1 on the surface of the endothelial cells (ECs) lining the vessels after which they are transported by transcytosis into the intima [reviewed in (19)]. In the intima, the lipoproteins bind via ionic interactions to the negatively charged proteoglycans of the extracellular matrix (5–7) and can be attacked by various proteases, lipases, and oxidizing agents and enzymes, as evidenced by lipoproteins isolated from atherosclerotic lesions having oxidized epitopes and showing signs of proteolysis and lipolysis [reviewed in (8, 10)]. Of note, the oxidatively modified particles from human atherosclerotic lesions are more extensively modified than in the corresponding plasma, raising the possibility that the changes have been initiated already in the plasma (20). Also, each of the above-listed types of modification has been shown to induce lipoprotein aggregation in vitro (8). Aggregation and fusion of lipoprotein particles increase their affinities of their affinities to proteoglycans its and induce foam cell formation (8). Moreover, the aggregation-susceptibility of LDL particles has been linked with future cardiovascular events in patients having established atherosclerosis (21, 22). The relative importance of the various types of lipoprotein modification in atherogenesis remains to be determined, but regarding the inflammatory potential of the modified lipoproteins, the products generated during lipoprotein modification (Figure 1) may be particularly relevant, as will be discussed in this review.

Atherosclerotic lesions contain both neutral and acidic proteases and lipases that can modify lipoprotein particles. Among the proteases are serine proteases, cathepsins, and metalloproteinases that have been shown to degrade apoB-100 (23–26). Of note, these proteases may also degrade the structural proteins like collagens in the extracellular matrix and may thus lead to cap-thinning of the plaques [reviewed in (27)]. The extent of proteolytic degradation of the protein components of the lipoproteins depends on the protease: plasmin, for example, can induce fragmentation of apoB-100 of LDL and does not induce LDL aggregation and fusion, while cathepsins proteolyze apoB-100 of LDL into small fragments and induce the generation of large, fused particles (24, 28, 29). Proteolysis can also promote lipolytic modifications of the lipoproteins (30). An example of the combination of proteolysis and lipolysis is enzymatically modified LDL (E-LDL), which is generated by treating LDL with a protease, such as matrix metalloproteinase 9 or cathepsin H, and cholesteryl esterase (31–33).

Extensive oxidation can result in the fragmentation of apoB (34). Oxidation also leads to changes in the lipoprotein lipids, leads to the generation of malondialdehyde adducts in the lysine residues of the proteins, and induces the generation of oxidation-specific epitopes on the lipoproteins. Unlike proteolytic and lipolytic modifications, oxidation decreases the interaction of lipoproteins with proteoglycans. However, oxidized LDL (oxLDL) has been shown to bind to decorin-coated collagen via lipoprotein lipase thereby leading to extracellular accumulation of oxLDL (35). OxLDL also induces foam cell formation via interaction with various scavenger receptors on macrophages (36, 37).

Phospholipase A₂ (PLA₂) is an enzyme that hydrolyzes phospholipids generating non-esterified fatty acids (NEFAs) and lysophospholipids. PLA₂ comprises several subtypes and can be found extracellularly as secreted PLA₂s, intracellularly as cytosolic PLA₂s, and in the bloodstream as lipoprotein-associated PLA₂ (Lp-PLA₂) (38). Lp-PLA₂ is produced by hepatocytes, and also by inflammatory cells such as macrophages, foam cells, and mast cells, as well as T-lymphocytes in atherosclerotic plaques, and in the circulation it is mainly associated with LDL (39, 40). The various PLA₂s have distinct specificities, group IIa, group V, and group X PLA₂s being able to avidly hydrolyze phosphatidylcholine on LDL particles, while Lp-PLA₂ preferably hydrolyzes oxidized phospholipids. Lp-PLA₂ and several types of secretory PLA₂s have been found in atherosclerotic plaques and in the normal arterial intima (38, 41–43). Modification of LDL, IDL, and VLDL with PLA₂ increases their affinities to proteoglycans in vitro (44, 45), and thereby potentially enhances the retention of LDL particles in the vessel wall. PLA₂-modified LDL also promotes macrophage foam cell formation (46, 47).

Treatment of LDL with an enzyme secreted by different cell types within the arterial intima, namely the sphingomyelinase (SMase), leads to conformational changes in apoB-100 and induces lipoprotein particle aggregation with ensuing increased proteoglycan binding (44, 48, 49). Secretory SMase implicated in atherogenesis has an acidic pH optimum, but can hydrolyze oxidized, PLA₂-treated, or proteolyzed LDL at neutral pH (30, 49). In addition, VLDL particles appear to be very susceptible to hydrolysis and aggregation by SMase (50, 51). Interestingly, ceramide, the lipolytic product of SMase action, is found exclusively in very large aggregates (49), such as those formed in vitro by modification of LDL by SMase at pH 5.5-6 (48). Generally, the sizes of LDL aggregates after in vitro modification differ substantially depending on the type of modification and the conditions used (48, 52). Although the aggregate size can affect aggregate uptake by, e.g., leukocytes (53), no studies have been performed to systematically study the effect of aggregate size on their pro-inflammatory properties either in vitro or in vivo.

Hydrolysis of the core lipids (cholesteryl esters and triglycerides) of lipoproteins has the potential to induce the generation of large amounts of NEFAs. The lipolytic hydrolysis is enhanced if the surface of the particles is also modified, e.g., by proteolysis, which allows penetration of some of the core lipids into the surface membrane of the lipoproteins (54). The E-LDL is an example of such multiply modified lipoprotein and antibodies generated against E-LDL bind to extracellular lipids in atherosclerotic lesions (55). Hydrolysis of cholesteryl esters has the potential to generate large amounts of cholesterol and lead to the generation of cholesterol crystals (56), which are a typical feature of very advanced atherosclerotic lesions (14). More recently, small cholesterol crystals have been identified also in early atherosclerotic lesions (57).

The complement system is a tightly regulated proteolytic cascade, a key component of innate immunity comprising over 30 soluble and membrane-bound proteins that responds rapidly to clear invading pathogens and damaged host cells, as well as to limit tissue destruction and initiate tissue healing. Complement activation is triggered by exposure to endogenous danger molecules that are released from damaged or dying cells and called danger-associated molecular patterns (DAMPs) (58). Depending on the initiation, the complement cascade follows either the lectin, classical, or alternative pathway. While these pathways have different activation cascades, they all converge downstream with the formation of their respective Complement (C) 3 convertases and the generation of C3a and C3b. If the complement activation is of sufficient magnitude, the membrane attack complex C5b-9 is formed. This important innate immune effector has pore-forming, lytic properties designed to destroy pathogens and damaged cells (59–61).

Circulating C3 levels were found to be significantly elevated in patients having familial hypercholesterolemia with subclinical coronary atherosclerosis, however, there was no correlation between circulating C3 levels and increased plaque burden, indicating a local regulation of the C3 in atherosclerotic arteries (62). Indeed, active components of the C3 complement are found within the extracellular matrix of human arteries (63). Complement components, including C3 were also identified in extracellular lipoproteins isolated from human carotid atherosclerotic lesions (56), and E-LDL was shown to colocalize with C5b-9 in human atherosclerotic lesions (55). Potential triggers of complement activation in the arterial wall include modified lipoproteins, cholesterol crystals, antigen-antibody immune complexes, C-reactive protein (CRP), and apoptotic cells [reviewed in (59, 63–65)]. Importantly, aggregated and fused lipoproteins isolated from human atherosclerotic lesions possess complement-activating properties (66). These particles do not contain immunoreactive apoB-100 but contain both esterified and unesterified cholesterol, and have sizes ranging from 100 to 500 nm (33). E-LDL with or without bound CRP, oxLDL, malondialdehyde-LDL, and oxLDL-immune complexes have been shown to induce complement activation [reviewed in (67)]. E-LDL, but not oxLDL, can trigger C1 activation and bind complement components (68, 69). Factor H, a major complement inhibitor, colocalizes in human atherosclerotic lesions with C3d in the subendothelial proteoglycan-rich layer, while C5b-9 is found in deeper areas of atherosclerotic lesions suggesting that in different areas of the lesions the components may differ in their abilities to regulate complement activation (63).

Foam cells are the histological hallmark of atherosclerotic lesions from the very beginning of the lesion development, and their contribution to atherogenesis remains throughout the lesion progression (13, 15). They are derived mostly from macrophages that take up modified and aggregated lipoproteins and lipoprotein-immune complexes via several different mechanisms (36). In addition to the macrophages, also vascular SMCs take up modified lipoproteins and can become foam cells (70, 71). The various uptake mechanisms include scavenger receptors that are important particularly in the uptake of oxLDL, Fc-gamma receptors interacting with immune complexes, cell surface proteoglycan-aided uptake of PLA₂-modified LDL, and phagocytosis of aggregated LDL and even cholesterol crystals (46, 72–75). Native VLDL remnant particles have been shown to induce foam cell formation through the VLDL receptor pathway in macrophages, and modification of VLDL particles by acetylation leads to macrophage foam cell formation via scavenger receptor-mediated uptake of the modified particles (76).

A specific uptake mechanism in macrophages has been reported for large LDL aggregates, such as SMase-modified LDL particles. Thus, aggregated LDL triggers the formation of extracellular surface-connected compartments into which the macrophages (77) and dendritic cells (78) secrete lysosomal enzymes, such as lysosomal acid lipase (LAL) (78–80). The process leads to a partial hydrolysis of the aggregated LDL outside the cells and to internalization of the partially hydrolyzed lipoproteins into lysosomes, where they are fully hydrolyzed. Actin polymerization and Toll-like receptor (TLR) 4 activation drive the intracellular catabolism of the aggregates (80, 81). In addition, treatment of LDL with macrophage-conditioned media containing catalytically active LAL induces LDL fusion and lipid droplet formation in VSMCs (54). LAL activity may be particularly prominent in the macrophage-rich acidic microenvironments observed in human atherosclerotic lesions (82).

Once lipoproteins have reached the lysosomes, their components are completely hydrolyzed (Figure 2). Importantly, cholesteryl ester hydrolysis by LAL results in the formation of unesterified cholesterol, which will then be re-esterified in the cytoplasm and packaged into intracellular lipid droplets characteristic of foam cells (83). Uptake of aggregated LDL has been shown to lead to accumulation of ceroid, a product of lipid oxidation consisting of protein-lipid complexes, in lysosomes (84). Lysosomal oxidation is likely mediated by iron and can be inhibited by cysteamine (85). On the other hand, uptake of oxLDL can result in lipid accumulation in the lysosomes (86, 87) and lead to the formation of cholesterol crystals in the lysosomes (88). However, the transformation of macrophages into foam cells may reduce their atheroinflammatory potential (89).

In vascular SMCs, foam cell formation due to uptake of native and modified VLDL was shown to depend on the activity of lipoprotein lipase (90), while the LDL receptor-related protein 1 plays a role in the uptake of aggregated LDL (91). Aggregated LDL particles increase the release of the soluble form of LDL receptor-related protein 1 from cultured vascular SMCs, and from human atherosclerotic plaques in an ex vivo model (92). Interestingly, the level of the soluble LDL receptor-related protein 1 in the circulating blood is increased in patients with coronary artery disease (92, 93). When compared with cultured human macrophages, cultured human SMCs express relatively low levels of LAL (94). Accordingly, in SMC foam cells, the ingested neutral lipids tend to accumulate within lysosomes, while in macrophages they appear as cytosolic droplets.

While complement activation is usually regarded to be pro-atherogenic, complement plays also a smoothing role in mediating the effects of foam cell formation. Thus, while the complement component C1q recognizes and opsonizes modified forms of LDL and promotes phagocytosis and efferocytosis in macrophage foam cells (95, 96), when bound to acLDL or oxLDL it reduces the release of proinflammatory cytokines from macrophage foam cells and atherosclerosis in mouse models (97). In addition, C1q downregulates both the levels and activities of apoptosis-related proteins in human and mouse macrophage foam cells, thereby leading to a measurable increase in the survival of these cells, which, again, could slow down the build-up of the necrotic core in the atherosclerotic plaques (95). Further, factor H, the main regulator of the alternative complement activation pathway, has been shown to bind to human monocytes and macrophage foam cells and has the potential to reduce complement activation and inflammation, and to increase the binding of apoE to these cells. It further increased cholesterol efflux, and cholesterol-loaded macrophages displayed reduced transcription of proinflammatory/proatherogenic factors and increased transcription of anti-inflammatory/anti-atherogenic factors, thereby indicating that apoE and factor H interact with monocytic cells in a concerted action and that this interaction tends to reduce complement activation and inflammation in atherosclerotic lesions (98).

Acetylated LDL (acLDL) and vortexed LDL (vxLDL) are models of modified lipoproteins that have been used to study intracellular lipid accumulation in macrophages, SMCs, and ECs (161–165). Vortexing of LDL leads to aggregation of the particles, but these modifications do not change the lipid moiety of LDL particles and, therefore, experiments in which they have been used provide information merely on their effects on lipid loading on macrophages. Treatment of macrophages with modified LDLs that induce foam cell formation, including acLDL, results in up-regulation of genes involved in the inflammation and immune responses (110).

VxLDL induces mitochondrial dysfunction in macrophages, a known activator of the NLRP3 inflammasome (166). Persson et al. studied the basal and inducible cytokine expression in primary human macrophages loaded with cholesterol using vxLDL and they observed that the cholesterol contained in the vxLDL did not affect IL-1β secretion, while the secretion of TNF, IL-6, and IL-8 were significantly decreased (167). However, the amount or composition of neutral lipids in the vxLDL did not affect cellular activation by exogenous TNF, making it likely that lipid loading attenuates cytokine secretion during basal conditions, and that the effects can be overruled by TNF during an acute inflammation (167). Sabeva et al., however, reported that IL-1 and IL-6 signaling can be induced by vxLDL in THP-1-derived macrophages (135). Depending on the macrophage polarization, inflammatory responses to cholesterol loading can vary vastly, as recently shown by comparing the inflammatory responses in M1-like and M2-like macrophages upon acLDL loading (89).

Electronegative LDL (LDL(-)), carbamylated LDL (cLDL) or HDL, glycosylated LDL (AGE-LDL), and mildly oxidized LDL are found in the circulation, and they can initiate foam cell formation in cell culture [reviewed in (168–170)].

LDL(-) is a naturally occurring, minor form of modified LDL in plasma, and it can be isolated based on its differing size, density, and charge. Compared to native LDL, LDL(-) has similar oxidation levels, but it differs in its lipid and protein compositions and in its apoB-100 conformation. LDL(-) also has higher phospholipolytic activities, higher aggregation levels, and higher proteoglycan binding affinity than LDL (133).

In ECs, LDL(-) increases TNF-induced inflammatory responses, such as the expression of transcription factors nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and Activator Protein 1, the expression of Vascular Cell Adhesion Molecule 1 (VCAM-1) (137), and the secretion of the chemoattractants IL-8 and Monocyte Chemotactic Protein 1 (MCP-1) (139), C-X-C Motif Chemokine Ligand 1, the cytokine IL-6, and the growth factors platelet-derived growth factor, Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF) and Vascular Endothelial Growth Factor (138, 140, 141). In monocytes and macrophages, LDL(-) directly induces priming, inflammasome activation with subsequent IL-1β release (171, 172), as well as secretion of IL-6, IL-8, and TNF (134). Puig et al. compared the effects of oxLDL, vxLDL, and acLDL with those of LDL(-) on THP1-derived macrophages and found that while oxLDL induced a more pronounced IL-1β release, LDL(-) induced a stronger release of IL-6, IL-10, and GM-CSF than did other modified LDLs (133). Interestingly, unlike other modified LDL preparations, LDL(-) induced a very strong triglyceride (TG) accumulation in lipid droplets of macrophages (133), a finding which may have clinical significance in that macrophages isolated from human atherosclerotic aorta show significant accumulation of TGs (173).

Protein carbamylation is a post-translational modification of proteins involved in a variety of disease states including inflammation, impaired renal function in chronic kidney disease, and, moreover, elevated levels of carbamylated proteins are present in smokers. Carbamylation has been mechanistically linked to atherosclerosis [reviewed in (174)], especially in patients with chronic renal failure who have vastly higher levels of carbamylated LDL in serum and a higher ASCVD risk (175). It has also been shown to induce endothelial dysfunction (176), increase the expression of adhesion molecules (153), the oxidative stress, and DNA damage in endothelial progenitor cells (177). Also carbamylated HDL increases monocyte adhesion to ECs via inducing NF-κB activation and the expression of VCAM-1 and the intercellular adhesion molecule-1 (ICAM-1) (154). Carbamylation also renders LDL more prone to oxidation, and LDL with both modifications has more pro-atherogenic effects on ECs and macrophages than either modification alone (178).

Non-enzymatic modifications of proteins by glucose are called advanced glycation end-products (AGEs), and they are formed at accelerated rates in hyperglycemic diabetic patients. AGEs can also form on lipids, as evidenced by the observation of lipid-linked AGEs in LDL particles derived from individuals with or without diabetes (179). The AGE-products include circulating AGE-peptides which derive from the catabolism of AGE-modified tissue proteins, and, moreover, AGE-modified LDL is formed by direct reaction between native LDL and circulating reactive AGE-peptides (179). The AGEs, among them also the AGE-LDL, damage vascular cells, one pathway being their interaction with their purported receptors on vascular ECs, SMCs, and macrophages. Among the ensuing implicated diverse pro-atherogenic processes, those leading to endothelial dysfunction have received much attention (155). The adverse endothelial effects include apoptosis (180), reduced nitric oxide synthesis, increased production of reactive oxygen species, and increased expression of (VCAM)-1 (156). Moreover, AGE-LDL particles induce IL-6 secretion by ECs and macrophages (158).

The lipolytic products generated by the action of PLA₂, fatty acids and lysophospholipids, induce proinflammatory responses in cultured human umbilical vein ECs (181). PLA₂-treated LDL induces exocytosis of the Weibel-Palade Bodies, an early-phase endothelial inflammatory response from ECs in a lysophosphatidylcholine (LPC)-dependent manner (182). LPC acts as a find-me/eat-me signal, which belongs to a group of soluble mediators released by apoptotic cells that attract phagocytes into the tissue (183). PLA₂-LDL has also been shown to upregulate the adhesion factors E-selectin, ICAM-1, and VCAM-1 in ECs (147). In primed human macrophages, PLA₂-treated LDL and especially LDL and VLDL treated with both PLA₂ and LAL have shown to elicit a strong induction of NLRP3-dependent IL-1β secretion (56).

Treatment of LDL particles with SMase, which is secreted by several cell types within the arterial wall, leads to the generation of ceramide and to conformational changes in apoB-100 that induce particle aggregation and increased proteoglycan binding (48–50). These effects could be alleviated by treatment of the LDLs with an apoA-I mimetic peptide prior to the hydrolysis (184). SMase-LDL has been shown to induce the secretion of TNF, IL-1β, IL-6, and the chemoattractant MCP-1 from human macrophages derived from cultured THP-1 monocytes (56, 84). LDL which has been lipolytically modified by secretory PLA₂ or SMase potentiates the cellular release of proinflammatory lipid mediator arachidonic acid, and triggers activation of the cytoplasmic PLA₂ in human THP-1 monocytes (185).

Regarding SMase, another point to consider is that it is expressed by macrophages. Endogenous sphingomyelins on the cell membrane can be hydrolyzed, and the generated ceramides can accumulate or be converted into numerous metabolites [reviewed in (186)]. These metabolites are involved in a variety of processes connected to cardiovascular health, they modulate the signaling and metabolic pathways driving insulin resistance, TG production, apoptosis, and fibrosis [reviewed in (187)]. Cellular uptake of SMase-modified LDL brings substantial amounts of ceramides into lysosomes, and, moreover, as LDL contains significant amounts of SM, uptake of modified LDL can also in other ways affect the sphingomyelin pool in macrophages and thus the concentrations of ceramides in them. Kinscherf et al. showed in their early work that loading of macrophages with oxLDL and acLDL delivers a substantial amounts of sphingomyelin into the cells and also stimulates ceramide formation in them (188).

Modifications of the lipoprotein core at neutral pH are more difficult to perform due to the protective surface monolayer of the lipoproteins. Therefore, the core-modifying treatments are usually preceded by a modification of the surface. To generate enzymatically modified LDL (E-LDL), LDL is first treated with a protease and subsequently with cholesterol esterase. Such multiply-modified LDL has been detected by immunohistochemical means in human atherosclerotic lesions at all stages of lesion development (55). Treatment of ECs with E-LDL leads to the expression of IL-8 (143), induces the adhesion of monocytes and T lymphocytes to EC monolayers, and stimulates the upregulation of the adhesion factors ICAM-1, platelet-endothelial cell adhesion molecule-1 (PECAM-1), P-selectin, and E-selectin (144). In aortic SMCs, E-LDL upregulates the expression of ICAM-1, which correlates with increased adhesion of T lymphocytes (144). It also strongly induces SMC foam cell formation and upregulates the expression of lectin-like oxLDL receptor 1 (LOX-1) with ensuing increase in the uptake of oxLDL (189).

In macrophages, E-LDL induces strong secretion of MCP-1 and mild release of IL-6 (145). It also promotes monocytic differentiation into dendritic cells, and, when dendritic cells are treated with CRP and/or E-LDL, the production of IL-12 and TNF is significantly increased, and a strong Th1 reaction and T cell proliferation are elicited (142).

LAL hydrolyzes TGs and cholesteryl esters in the core of the lipoprotein particles. LAL is an essential lysosomal enzyme and its deficiency leads to cholesteryl ester storage disease (190). Expression of LAL has been shown to become downregulated in several cell types when the cells are loaded with modified LDL (191, 192). LAL is also secreted from activated macrophages and can then act on lipoproteins also extracellularly (54). Treatment of macrophages with LAL-LDL activates the transcription factors p38 MAPK and NF-κB, induces the secretion of IL-8 (146), and may also promote vascular dysfunction and atherogenesis [reviewed in (193)]. LDL and VLDL, when treated first with enzymes that hydrolyze the particle surface before treatment with LAL are able to trigger a robust, NLRP3-mediated IL-1β secretion in human macrophages (56).

VLDL remnants treated with lipoprotein lipase to hydrolyze the TGs have been shown to induce increased expression of TNF, IL-1β, and IL-8 over native VLDL or lipoprotein lipase, with concurrent activation of NF-κB and activator protein 1 in peripheral blood mononuclear cells and THP-1 monocytes (194).

Oxidation is the most studied form of lipoprotein modification, and the pro-inflammatory effects of oxidized phospholipids in inflammation and various inflammatory diseases are well established (195).

OxLDL acts as a DAMP and triggers sterile inflammatory responses (195). In ECs it causes endothelial dysfunction and induces a pro-inflammatory and pro-thrombotic state (196). It also enhances monocyte chemotaxis and adhesion to EC in culture (197). The uptake of oxLDL by macrophages is mediated by scavenger receptors such as CD36, and it induces the assembly of a TLR4/TLR6 heterodimer, and results in NF-κB signaling and priming of these cells for inflammasome activation (198). OxLDL uptake by CD36 additionally results in intracellular cholesterol crystallization and can thereby directly activate the NLRP3 inflammasome in macrophages (88). Both minimally oxidized LDL and more highly oxidized LDL cause the secretion of pro-inflammatory cytokines by macrophages (70, 71) by activating TLR4 (199, 200). Li et al. showed in THP-1 monocytes that oxLDL and oxHDL increases the expression of NLRP3 and activates caspase-1 and induces IL-1β and IL-18 secretion in a dose-dependent manner (201). However, other studies have shown that oxLDL inhibits the production of inflammatory cytokines by macrophages in response to inflammatory stimuli, such as LPS (202). More recently, it was shown that oxLDL, as well as oxHDL3, reduces the secretion of mature IL-1β by inhibiting the activation of the NLRP3 inflammasome induced by known inflammasome activators including the acute-phase serum amyloid A protein, ATP, nigericin, and monosodium urate crystals (203). By binding to LOX-1 on SMCs, oxLDL can also induce SMC transition to an inflammatory phenotype. Thus, treating cultured SMCs with oxLDL stimulates the expression of the chemokines MCP-1 and C-X-C Motif Chemokine Ligand 1, the cytokine TNF, and the cell adhesion molecule VCAM-1 [reviewed in (204)].

OxLDL can also induce pyroptosis in macrophages (205) and ECs (206), and oxLDL-induced pyroptosis has also been implicated in endothelial dysfunction and vascular SMC foam cell formation [reviewed in (109)].

OxLDL is also able to induce a novel form of cell death coined ferroptosis, which is linked to atherogenesis (207, 208). Bai et al. showed that some of the key morphological features of ferroptosis, such as mitochondrial shrinkage, increased iron content, increased release of lactate dehydrogenase, and reduced expression of the solute carrier family 7 member 11 and glutathione peroxidase 4 were induced by oxLDL in mouse aortic ECs to a comparable amount achieved with the ferroptosis inducer erastin; the phenotype was rescued by co-stimulation with ferroptosis inhibitor ferrostatin-1 (209). Further, oxLDL- and erastin-treated cells had similarly increased levels of markers of intracellular lipid peroxidation, such as total ROS, lipid ROS, lipid peroxide, and malondialdehyde, while treatment with ferrostatin-1 downregulated the generation of these lipid peroxidation products (209). Similar induction of ferroptosis by oxLDL was observed in human coronary artery ECs (210).

While minimally oxidized LDL is found in circulation, the vast majority of circulating oxLDL is bound to antibodies and found as oxLDL immune complexes, which can prime dendritic cells and macrophages for inflammasome activation (211). While the circulating immune complexes may be too large to cross an intact endothelium, immune complexes can also be formed within the arterial intima (212). Interestingly, immune complexes composed of oxLDL and anti-LDL IgG antibodies can trigger activation of cultured human mast cells, an innate immune type of cell present in human atherosclerotic lesions, and, when activated, capable of proteolytically modifying LDL and HDL particles (213, 214).

In ECs, oxidized VLDL initiates the expression of the fibrinolysis inhibitor plasminogen activator inhibitor-1 and stress response protein heat shock factor 1 compared to the effects obtained with VLDL or LDL (215), and induces endothelial apoptosis (216). It also induces expression of IL-15, matrix metalloproteinase-2, macrophage migration inhibitory factor, as well as downregulated expression of transforming growth factor beta (TGF-β) (150). In macrophages, both oxidized VLDL and its remnants induce foam cell formation that can be attenuated by pre-incubation with TGF-β (217). In both monocytes and macrophages, oxidized VLDL induces the expression of MCP-1 (151, 152).

Oxidized HDL (oxHDL) plasma levels have been shown to correlate with ASCVD and coronary artery calcification [recently reviewed in (218)]. OxHDL induces endothelial dysfunction in ECs by a positive feedback mechanism in which an oxidative burst generates oxHDL from native HDL, activates LOX-1 which in turn increases the expression of NADPH oxidase 2, TNF, and LOX-1 receptor at the endothelial plasma membrane (219). It has also been shown to decrease the expression of CD63 in macrophages, thus attenuating the ability of the cell to ingest modified lipoproteins (220, 221). OxHDL also binds to CD36 on platelets and induces proinflammatory and procoagulant effects (222).

Another innate immune pathway contributing to vascular inflammation is neutrophil extracellular trap (NET) formation, also called NETosis (223). NETs are structures of chromatin filaments coated with histones, proteases, as well as granular and cytosolic proteins, which are ejected by activated neutrophils to immobilize and eliminate pathogens as a part of the first-line immune defense.

Recent research has revealed that NETosis is associated with the initiation and progression of various noninfectious diseases (224), and NETosis has also been shown to promote thrombosis (225). NETs have been found in vascular lesions such as atherosclerotic plaques (226), and infiltration of neutrophils into arteries during the early stages of atherosclerosis has been observed in hypercholesterolemic mice (227). One theory about the reason for NETs in plaques is that periodontal pathogens enter the bloodstream from severely infected periodontal tissues, then reach the inflamed atherosclerotic lesions where they can directly contribute to the development of the lesions [reviewed in (228)].

Non-infectious triggers such as LDL-derived cholesterol crystals might also induce NETosis in atherosclerosis. NETs prime macrophages, which in turn start to produce pro-IL-1β and IL-8 (117). Cholesterol crystals can bind to CD36 on the macrophages and thereby trigger inflammasome activation with the ensuing secretion of mature IL-1β. Thus, NETs may contribute to atherogenesis by initially activating macrophages (229, 230). Studies in apoE-/- mice with deletions of two neutrophil proteases showed attenuated NETosis as well as smaller atherosclerotic lesions, and lower systemic IL-1β levels, when compared to ApoE–/– mice expressing these proteases (230), indicating a potential role for NETs and/or these proteases in this process.

Kato et al. demonstrated the appearance of oxLDL in vascular tissues and circulation even prior to atherosclerotic lesion development in apoE-KO mice (231). OxLDL as well as its various oxidized phospholipid components, especially oxPCs and LPC, mediate NETs formation and subsequent endothelial inflammatory responses in both HL-60-derived neutrophils and human polymorphonuclear neutrophils (232, 233).

Natural antibodies are antibodies that spontaneously arise without prior infection or defined immune exposure. Most serum antibodies at birth are IgM natural antibodies and, as such, they represent a primitive innate-like “layer” of the adaptive immune system (252, 253). Antigens that are specifically recognized in atherosclerosis include microbial antigens, endogenous heat shock proteins, β-2 glycoprotein I, and modified LDL, especially oxLDL (253). The degree of oxidation of LDL in humans is variable, and two major modifications prepared in vitro—copper-oxidized LDL and MDA-LDL—are recognized by human autoantibodies. Both forms share MDA-lysine as their main oxidation-specific epitope but MDA-LDL contains roughly a 10-fold higher amount of the epitope (254).

IgG and IgM antibodies against oxidation-specific epitopes of oxLDL such as MDA-modified lysine residues or the phosphorylcholine headgroups of oxidized phospholipids have also been found in the plasma, and oxLDL-immune complexes are present in plasma (255) as well as in atherosclerotic lesions of humans and animals (212). Several clinical studies investigated the association of oxLDL autoantibodies with atherosclerosis progression, and in most, atherosclerosis or its progression was directly associated with oxLDL IgG levels, and inversely associated with IgM (256, 257) [reviewed in (240)]. While oxidation-specific epitopes are also commonly found on surfaces of apoptotic cells, microvesicles, and bacteria (195), antibodies against the epitopes were shown to have the capacity to block oxLDL uptake (258), the case for their role in atherogenesis being strong.

Antibodies against carbamylated LDL have also been found in individuals with uremia or tobacco smoking, and these antibodies also cross-react with MDA- and malondialdehyde acetaldehyde-LDL (259, 260). Further, also IgG against AGE-LDL has been found in human serum and to be associated with diabetes mellitus (261).