Remnant Lipoproteins Promote Foam Cell Formation
Atherosclerosis is characterized by accumulation of inflammatory foam cells whose formation is promoted by the subendothelial retention of ApoBcontaining lipoproteins. Plaques develop in predisposed areas of the arterial tree where blood flow is either slow or has a back and forth pattern (thus coronary arteries are particularly prone).1 In these predisposed areas endothelium displays increased susceptibility to inflammation as well as greater permeability to lipoproteins with subendothelial retention in these locations. Resident subendothelial dendritic cells may be the first cells to take up retained lipoproteins to become foam cells. Some dendritic cell subtypes suppress while others promote inflammation.2 Hyperlipidemia initiates greater endothelial expression of inflammatory adhesion molecules (by multiple mechanisms) followed by macrophage and neutrophil transmigration into the subendothelial space. Eventually, macrophages as well as activated smooth muscle cells begin to accumulate and are converted to foam cells. Surprisingly, early acquisition of cholesterol by macrophages actually suppresses inflammatory responses, leading to a reparative macrophage phenotype.3 However, continued cholesterol accumulation, particularly with excessive intracellular unesterified cholesterol combined with stimulation of innate immune receptors (such as tolllike receptors), results in predominantly inflammatory macrophages. Further accumulation of macrophages (and other inflammatory cell types) ensues, followed eventually by wholesale apoptosis and necrosis with formation of the necrotic core and an unstable plaque.4 These vulnerable plaques have a high cholesterol content, many macrophages at the shoulders, thinned fibrous caps, and are prone to rupture, leading to acute coronary events.5 The physical expansion caused by sudden cholesterol crystallization in such plaques may be a major driving force for their rupture.6
Excess cholesterol accumulation can lead to initiation, promotion, and progression of atherosclerotic lesions and may even precipitate plaque rupture and acute coronary events, but where does the cholesterol come from? In classical in vitro studies, incubation of macrophages with native LDL (low density lipoprotein) did not result in foam cell formation due to downregulation of the LDL receptor.7-9 However, after LDL were oxidized or acetylated they were avidly taken up by macrophages with conversion to foam cells. Importantly, in these same studies, triglyceride-rich remnant lipoproteins (TGRL) from cholesterol-fed rabbits or dogs (referred to as ß-VLDL) needed no modification to promote foam cell formation. Note that ß-VLDL are TGRL with abnormal composition and are not equivalent to IDL (intermediate density lipoproteins). They are composed of both intestinal (with ApoB48) and hepatic (with ApoB100) TGRL remnants. ß-VLDL have density less than 1.006 (the density of plasma) and float upon ultracentrifugation whereas IDL do not float. Unlike normal VLDL which have pre-ß mobility, ß-VLDL have ß mobility upon electrophoresis, that is, they move like LDL. Finally, ß-VLDL are abnormally enriched in cholesterol (mostly esterified) due to prolonged transit time and exchange of cholesteryl ester for triglycerides through the action of cholesterol ester transfer protein (CETP).10
The contribution of various forms of oxidized LDL (including minimally modified LDL) to foam cell formation in vivo continues to be debated.11 In the meantime, a number of additional LDL modifications that promote foam cell formation may even be more quantitatively important than oxidation. These include proteoglycan binding and aggregation, especially after exposure to various phospholipases (including LpPLA2) or sphingomyelinase, which result in so-called electronegative LDL.12 Besides ß-VLDL, several other types of TGRL have also been shown to promote foam cell formation, including human VLDL from hypertriglyceridemic subjects13, human chylomicron remnants14, and remnant-like particles (RLP) isolated by incubation with immunoaffinity gels directed against a specific epitope on ApoB and ApoA1 with the intention to remove nascent TGRL and HDL.15,16 TGRL have been directly isolated from human aortic intima.17,18 In one study, 36% of the cholesterol isolated from aortic plaque in patients undergoing aortic reconstruction was from very low density lipoprotein (VLDL) and intermediate dense lipoprotein (IDL).18
Chylomicron remnants (CR) are cholesterolrich TGRL remnants produced from the hydrolysis of chylomicrons. These ApoB48- containing particles vary greatly in size and composition, becoming denser and less negatively charged as they lose triglycerides and their associated ApoC lipoproteins while increasing their concentration of cholesteryl ester. Human CR are in the range of 50 to 150 nm in diameter.19 Small VLDL and IDL are TGRL remnants produced from the hydrolysis of triglyceride-rich VLDL. Gradient density ultracentrifugation reveals small VLDL and IDL in the Sf (Svedberg flotation rate) 20 to 60 and 12 to 20 ranges respectively.10 The diameter of LDL, small VLDL, and IDL particles are, respectively, 20 to 25 nm, 30 to 80 nm, and 25 to 35 nm. The density of IDL is greater than 1.006, but less than 1.019 g/mL with a diameter of 27.5 to 30 nm in individuals without dyslipidemia. Approximately 15% to 20% of the total cholesterol is carried in IDL and a normal plasma concentration of IDL is 5 to 15 mg/dL and a total mass of 10 to 30 mg/dL.10 Lipoproteins greater than 75 nm in diameter are thought to not enter the arterial wall.20 These considerations suggest that small CR and other TGRL remnants can enter the arterial wall and contribute to atherogenesis.
These findings may support the possibility that postprandial CR contribute to atherogenesis.21,22 Recently, much more ApoB48 was reported to be present in human carotid plaque than ApoB100.23 It appears to be the cholesteryl ester component of these remnant TGRL that is atherogenic as demonstrated by ACAT2 deficiency, which almost entirely abrogated atherosclerosis in ApoE null mice. In these knockout mice, there were normal or slightly increased numbers of both ApoB48 and B100 particles having markedly reduced cholesteryl ester and increased triglyceride content.24
TGRL Remnants Can Initiate Endothelial Inflammation
Upon incubation with TGRL, endothelial cells upregulate their expression of MCP-1, ICAM-1, and VCAM-1.25,26 MCP-1 is a chemokine that stimulates monocyte integrin activation, allowing firm adherence to ICAM-1 and VCAM-1 while also promoting transendothelial migration. Incubation of monocytes with RLP also promotes their adherence to endothelial cells.27 RLP adversely affect endothelial function by directly and indirectly inhibiting endothelial nitric oxide synthase.28 Furthermore, elevated RLP has been shown to be an independent risk factor for impaired flow-mediated, endothelial-dependant dilatation in patients with coronary artery disease.29 Elevated RLP levels have been associated with impaired coronary vasomotor response and acetylcholine-induced spasm.30,31 Elevated TGRLs were further found to be cytotoxic and induce apoptosis of endothelial cells.32
Hydrolysis of TGRL May Also Activate Endothelial Cells
Hydrolysis of TGRL has been shown to induce endothelial inflammation with production of TNFα, ICAM-1, and increased reactive oxygen species.33 In this study, it was the free fatty acids derived from the hydrolysis of TGRL, not the cholesteryl ester, triglycerides, free cholesterol or phospholipids that were associated with these effects.
Free fatty acids released during hydrolysis of TGRL can also adversely affect endothelial barrier function and increase subendothelial transfer of lipoproteins. In a study with cultured endothelial cells, exposure to oleic acid resulted in an increased transfer of LDL across the endothelium.34 TGRL hydrolysis products were reported to increase endothelial permeability by promoting disruption of the zonula occludens-1 complex which is essential for tight junction formation. Increased caspase 3 activation was also seen, which can be associated with apoptosis.35 In another study, RLP were shown to induce a strong inflammatory response with vigorous NADPH oxidase activation and superoxide formation followed by apoptosis in endothelial cells through activation of the LOX1 receptor.36
Further Observations on Foam Cell Formation
In the subendothelial space, monocytes differentiate into macrophages where they ingest ApoB-containing lipoproteins. The inaugural event is the subendothelial retention of ApoB lipoproteins.37 In a study of patients undergoing elective carotid endarterectomy, although the influx of LDL cholesterol was 19 times greater than that of TGRL cholesterol, the intimal clearance and fractional loss were similar.38 In a study of heritable hyperlipidemic rabbits, lipoprotein arterial influx was linearly related to plasma concentration; however, efflux was inversely related to lipoprotein diameter,39 suggesting the potential for greater retention of TGRL remnants. The main ApoB proteoglycan binding site is between the positively charged basic amino acids on ApoB (residues 3359 to 3369) and the negatively charged sulfate groups on the glycosaminoglycan chains of proteoglycans.40 Small VLDL and IDL have less affinity for proteoglycans; however, like LDL, sphingomyelinase causes VLDL and IDL to aggregate, fuse, and enhance their binding to proteoglycans.41 It has been shown that sphingomyelinase-induced aggregation of TGRL leads to foam cell formation.42
Although there is much greater penetration of the endothelial barrier by LDL particles, TGRLs carry significantly greater cholesteryl ester molecules per particle. It has been estimated that CR-TGRL of approximately 100 nm in diameter carry 40 times more cholesteryl ester than LDL particles.43 In a study evaluating TGRL and LDL fractions removed by density gradient ultracentrifugation from thoracic and abdominal aorta tissue at autopsy, it was found that when these fractions were incubated with mouse peritoneal macrophages, TGRL increased incorporation of radioactive oleate into cholesteryl esters by 10-to-20 fold as compared to three-to-four fold for LDL.17 Similar increases in cholesteryl ester synthesis by were seen in studies with dogs over 30 years ago.7 In patients with type III or type IV hyperlipidemia, oxidized ß-VLDL or VLDL remnants were found to cause greater macrophage cholesteryl ester formation than oxidized LDL.44,45
Coronary Risk Associated with Type III Hyperlipidemia
Type III hyperlipidemia is characterized by increased accumulation of ß-VLDL in plasma. This phenotype is commonly thought to be rare, being the result of an apo E 2-2 genotype (about 1 in 100 persons) together with a genetic predisposition to excess VLDL production, such as APOA5 variants,46 or acquired overproduction of VLDL as with obesity or hypothyroidism. The prevalence of type III is frequently cited as approximately 1 in 10,000.47 However, in the Lipid Research Clinics (LRC) Prevalence Study, type III hyperlipidemia was found in 0.4% of men in the general population.48 This study represents one of the only studies to apply classic criteria to all participants to define type III, namely, the presence of a ß-VLDL band upon electrophoresis of the density <1.006 fraction isolated after ultracentrifugation of plasma.
While markedly increased risk of atherosclerotic disease has long been appreciated for patients with type III, a population-based estimate of risk was not available until our recent publication (PNH).49,50 Additional, previously unpublished analyses utilizing data from the more recent of these studies50 are presented in Figures 1-3. The study groups consisted of 1759 population-based controls and 1170 cases with onset of clinical CAD by age 60 in men and 70 in women, all with ultracentrifugation performed on plasma samples. Type III hyperlipidemia was defined as present if the ratio of measured VLDL cholesterol/total triglycerides was > 0.30 with total triglycerides > 150 mg/dL.51 The prevalence of type III (0.68%) we identified in the control population was very similar to the LRC Prevalence Study estimate, especially in consideration of the increased obesity expected in the population. The prevalence among our cases was 2.7%, almost identical to that reported by Goldstein, et al.52 In Figure 1, risk associated with the presence of type III is given with adjustment for LDLC, HDL-C, triglyceride categories (which excluded type III subjects), hypertension, diabetes, and cigarette smoking. In addition to the traditional yes/no definition of type III, we show the markedly increasing risk associated with more severe type III as defined by an algebraic estimate of plasma ß-VLDL cholesterol levels. CAD risk was increased over 40-fold in the most severe category. These severe cases represent only about 1/1000 control subjects yet most did not have any xanthomas. Perhaps those with tuberous xanthomas and/or palmar striae would be found as infrequently as 1/10,000. Elevations in triglycerides without type III were associated with increased CAD risk, but to a much lesser extent as shown in Figure 2. Interestingly, many cases of type III hyperlipidemia would have been missed if ultracentrifugation had only been performed in those with triglycerides over 300-400 as shown in figure 3. It should be noted that estimates of risk associated with remnant accumulation can vary substantially, depending on the method or parameter used.10,53-55
In summary, despite significantly lower plasma concentrations than LDL, TGRL and TGRL remnants contribute to atherosclerosis plaque formation. With increasing obesity rates, these TGRL-derived particles may play a greater role in the development of atherosclerotic burden. Non-HDL cholesterol goals therefore may become even more important in the management of the dyslipidemic patient.
Disclosure statement: Dr. Hopkins has received honoraria from Merck & Co. Dr. Hopkins has received research grants from Regeneron and Takeda Pharmaceuticals. Dr. Nelson has received honoraria from Abbott Laboraties, Amarin Corp., AstraZeneca, Atherotech, Bristol-Myers Squibb, Daiichi Sankyo Inc., GlaxoSmithKline, Gilead Pharmaceuticals, Kowa Pharmaceuticals America, Merck & Co., Pfizer Inc., and Novartis Pharmaceuticals.
Download Lipid Spin Spring 2013