Sialic acid content of low density lipoprotein and its relation to lipid concentrations and metabolism of low density lipoprotein and cholesterol

A low sialic acid content in low density lipoprotein (LDL) has been associated with atherogenicity and coronary artery disease (CAD) in many but not all studies. We investigated associations of the sialic acid-to-apolipoprotein B (apoB) ratio of LDL with lipoprotein lipid concentrations, kinetics of LDL, metabolism of cholesterol, and the presence of CAD in 98 subjects (CAD(+), n = 56; CAD(-), n = 42). The sialic acid ratios of total, dense, and very dense LDL were lower in the CAD(+) than CAD(-) subjects, especially at high sialic acid ratios. The LDL sialic acid ratio was inversely associated with respective lipid and apoB concentrations and positively with lipid-to-apoB ratios of LDL. The transport rates (TRs) for total and dense LDL apoB were negatively associated with their sialic acid ratios. The sialic acid ratio of dense LDL, but not that of total LDL, was inversely correlated with serum levels of cholesterol precursor sterols, indicators of cholesterol synthesis, and positively with serum levels of plant sterols, indicators of cholesterol absorption. In addition, the TR for dense LDL was positively correlated with cholesterol synthesis.In conclusion, a low LDL sialic acid ratio was associated with CAD, high numbers of small LDL particles, and a high TR for LDL apoB, and in dense LDL also with high synthesis and low absorption of cholesterol.     

Lipoprotein aggregation as an essential condition of intracellular lipid accumulation caused by modified low density lipoproteins

We have tested a hypothesis that aggregates of modified low density lipoproteins (LDL) play the key role in the accumulation of lipids by cells of unaffected aortic intima. It was demonstrated using analysis of relative dispersion of light transmission fluctuations as well as gel filtration on Sepharose CL-2B that LDL modified by oxidation, glycosylation, desialylation and malondialdehyde treatment form aggregates under the conditions of culture. Native LDL failed to aggregate under the same conditions. It was demonstrated that modified LDL, unlike native LDL, bring about a 2- to 3-fold rise in cholesteryl ester levels of cultured cells. Moreover, direct and strong correlation (r = 0.86) was observed between the degree of lipoprotein aggregation and the amount of cholesteryl esters accumulated. Removal of modified LDL aggregates by filtration through a 0.1 micron filter or gel filtration completely prevented the intracellular accumulation of cholesteryl esters. These findings indicate that LDL aggregates play an essential, if not the decisive, role in the intracellular accumulation of lipids in vitro.     

Discrete subspecies of human low density lipoproteins are heterogeneous in their interaction with the cellular LDL receptor.

The low density lipoproteins (LDL) of human plasma consist of a series of discrete particle subspecies of distinct physicochemical, immunological, and hydrodynamic properties. Such structural differences are intimately linked to the metabolic heterogeneity of circulating LDL in vivo. The current studies were designed to evaluate and compare the interaction of discrete LDL subspecies from normolipidemic subjects with the LDL receptor. Plasma LDL of d 1.019-1.063 g/ml from healthy males were fractionated into 15 subspecies of defined physicochemical characteristics by isopycnic density gradient ultracentrifugation as described earlier (Chapman et al., J. Lipid Res. 1988. 29: 442-458). The major LDL subspecies, LDL-5 to LDL-10, exhibited an overall range in density from 1.0244 to 1.0435 g/ml; individual subspecies increased in density by increments of 0.027 (LDL-5), 0.026 (LDL-6), 0.030 (LDL-7), 0.031 (LDL-8), 0.035 (LDL-9), and 0.042 g/ml (LDL-10), respectively. Taken together, these subspecies accounted for approximately 70% of the total mass of LDL of d 1.019-1.063 g/ml; their cholesterol: protein ratios decreased from 1.70 to 1.12 and particle size from 275 to 260 A with increase in density. ApoB-100 was the unique protein component in subspecies 5-8, with trace amounts (less than 0.2% of apoLDL) of both apoA-I and apoE in subspecies 9 and 10. The interaction of individual LDL subspecies with the LDL receptor on cultured human U-937 monocyte-like cells was compared by determining receptor binding affinities at 4 degrees C. Scatchard analysis of specific binding curves demonstrated a single class of binding site for each subspecies. The lowest dissociation constants were displayed by LDL subspecies 6 (Kd 5.71 nM), 7 (Kd 5.24 nM) and 8 (Kd 4.67 nM), while subspecies 5, 9, and 10 displayed significantly higher Kd values (8.35, 7.20, and 6.87 nM, respectively). Competitive displacement studies at 4 degrees C, in which unlabeled subspecies from the same gradient series competed for binding with 125I-labeled LDL subspecies, confirmed the relative binding affinities of these subspecies. As the hydrophobic lipid core of LDL undergoes a thermotropic transition at approximately 37 degrees C, which may in turn influence the surface structure of the particle, internalization and degradation studies were performed at 37 degrees C. No effect of temperature was detectable; again, LDL subspecies at each extreme of the density distribution (LDL-5 and LDL-10) displayed significantly lower binding affinities for the LDL receptor than that from the peak region (LDL-7).(ABSTRACT TRUNCATED AT 400 WORDS)     

The carbohydrate content of apolipoprotein E from human very low density lipoproteins.

The carbohydrate content of the E protein of human very low density lipoprotein (VLDL) was evaluated both by colorimetric methods and by gas liquid chromatography of the trifluoroacetylated 0-methyl glycosides. The major unmodified hexose was noted to be galactose with a mole ratio with respect to protein which ranged from 0.81 to 1.54. N-acetyl glucosamine (molar ratios from 0.52 to 1.76) and N-acetyl galactosamine (molar ratios from 0.73 to 1.59) and the respective unacetylated amino sugars were noted for all of the apoproteins evaluated. Sialic acid (molar ratios from 0.79 to 1.69) was a prominent carbohydrate for each of the E protein preparations. When the apoprotein was exposed to neuraminidase with a resultant loss of two-thirds of the sialic acid, the isoelectric focus behavior was found to be unchanged. The E protein isolated from the very low density lipoproteins of Type III patients (dysbetalipoproteinemia) revealed a carbohydrate content similar to the normals or Type IV patients.     

Influence of prostaglandins on the lipid transfer between human high density and low density lipoproteins

Prostaglandin (PG) E₁ was demonstrated to stimulate the transfer of phosphatidylcholine and cholesterol esters from human high density lipoproteins (HDL₃) to low density lipoproteins (LDL). The enhancement effect of PGE₁, on the interlipoprotein lipid transfer was seen at low PG concentrations under conditions of spontaneous exchange as well as in the presence of lipoprotein-depleted plasma, or partly purified plasma lipid exchange protein. PGE₂ and PGF₂ showed no significant influence on the interlipoprotein lipid transfer. Evidence is presented suggesting that the PGE₁-induced stimulation of interlipoprotein lipid exchange results in enhancement of LCAT-catalyzed cholesterol esterification in plasma. It is proposed that the effect of PGE₁ is due to the previously described PGE₁-induced reorganization of the HDL surface [(1984) FEBS Lett. 173, 291-293] and that PG-lipoprotein interaction may be a factor regulating cholesterol homeostasis.     

Differences in carbohydrate content of low density lipoproteins associated with low density lipoprotein subclass patterns

The neutral carbohydrate content of both the protein (apoB) and lipid fractions of low density lipoproteins (LDL) from subjects with a predominance of small, dense LDL (subclass pattern B) was found to be lower than in subjects with larger LDL (subclass pattern A): 45 +/- 12 versus 64 +/- 13 mg/g apoLDL, and 58 +/- 8 versus 71 +/- 8 mg/g apoLDL (P less than 0.0005 for both). Sialic acid content of LDL lipids, but not apoB, was also reduced in subclass pattern B. ApoB and glycolipid carbohydrate content of total LDL and LDL density subfractions declined with increasing LDL density and decreasing particle diameter. Moreover, in LDL subfractions from pattern B subjects, carbohydrate content of LDL apoB, but not LDL glycolipid, was significantly lower in comparison with particles of similar size from pattern A subjects. Thus, in LDL subclass pattern B, reductions in LDL carbohydrate content are associated both with reduced concentrations of larger carbohydrate-enriched LDL subclasses, and with reduced glycosylation of apoB in all LDL particles. LDL glycolipids may vary with overall lipid content of LDL particles, but variation in apoB glycosylation may indicate differences in pathways for LDL production, and reduced apoB glycosylation may reflect the altered metabolic state responsible for LDL subclass pattern B.     

Cell triacylglycerol accumulation from very low density lipoproteins isolated from normal and hypertriglyceridemic human sera.

Human fibroblast cells in culture increased their intracellular triacylglycerol levels when exposed to very low density lipoproteins (VLDL) isolated from human plasma. This response was dependent on the amount of VLDL added. VLDL from normal, type IV or type V sera gave similar results. Lipoprotein lipase enhanced this intracellular triacylglycerol accumulation. It was concluded that human fibroblast cells in culture have at least two mechanisms for triacylglycerol uptake from VLDL: (1) uptake from intact lipoprotein either by surface transfer of lipoprotein lipid or internalization of the entire lipoprotein particle, and (2) re-esterification of lower glyceride and fatty acids released by lipoprotein lipase degradation of VLDL.     

Dependence on heparin chain-length of the interaction of heparin with human plasma low density lipoproteins

Successive rechromatography of commercial bovine lung heparin on human plasma low density lipoproteins (LDL) immobilized to AffiGel-10 yielded four high reactive heparin (HRH-I to IV) fractions and an unreactive fraction (URH). HRH-I was the most sulphated HRH fraction whereas URH had the least sulphation. In the presence of 10 mM Ca2+, LDL were precipitated by these heparins in the following order: HRH-II greater than HRH-III greater than HRH-IV greater than HRH-I greater than URH. The average molecular weight of HRH-I to IV was 8600, 11400, 10,100, and 10,000, respectively. A plot of log molecular weight versus the concentration of HRH required to give half-maximal precipitation of LDL showed a negative correlation (r = -0.880). These results indicate that heparin chain length is an important determinant of heparin binding to LDL in solution and may have relevance to the binding and precipitation of LDL in the arterial wall.     

Effect of human Apo AIV against lipid peroxidation of very low density lipoproteins

Recent studies have demonstrated that Apo AIV exerts a protective effect against atherosclerosis. Moreover, Qin et al. (Am. J. Physiol. 274 (1998) H1836) have demonstrated that Apo AIV, isolated from rat plasma, exerts an inhibitory effect against Cu(2+)-induced lipid peroxidation of intestinal lymph and LDL. The aim of the study was to investigate whether human Apo AIV exerts a protective effect against Cu(2+)-induced lipid peroxidation. Our results demonstrated that human Apo AIV exerted an inhibitory effect against Cu(2+) and AAPH induced lipid peroxidation of VLDL, as shown by the lower increase in the levels of TBARS and conjugated dienes in lipoproteins preincubated with Apo AIV. In addition, the tryptophan (Trp) and probe 2-(dimethylamino)-6-lauroylnaphthalene (Laurdan) fluorescence studies demonstrated that the modifications of spectral properties in both lipoproteins preincubated with Apo AIV were lower with respect to ox-lipoproteins, suggesting that Apo AIV prevents the modification of physico-chemical properties due to peroxidation.     

Combined effects of desialylated and glycosylated low density lipoproteins on lipid content of human aortic intima cells in vitro]

We have studied a combined effect of glycosylated low density lipoproteins (LDL) on the cholesterol content of cells cultured from unaffected human aortic intima. Native LDL did not alter the intracellular cholesterol level while glycosylated LDL taken in the concentration of 50 and 100 mg/ml increased the cell cholesterol content by 30 and 70 percent, respectively. The effect of the same concentrations of glycosylated LDL treated with neuraminidase (desialylated-glycosylated LDL) was twice as powerful. Desialylated LDL in the concentration of 50 and 100 mg/ml raised the cholesterol level by 1.4- and 2.1-fold, respectively. Simultaneous incubation of cells with glycosylated (50 mg/ml) and desialylated (50 mg/ml) LDL brought about a 3.4-fold increase in intracellular cholesterol. The obtained data suggest that intensive development of atherosclerosis in diabetes mellitus may be partially explained by synergic effects of desialylated and glycosylated lipoproteins as well as LDL with both types of modification.     

Interactions of native and modified human low density lipoproteins with human skin fibroblasts.

125I-labeled low density lipoprotein (LDL) covalently bonded to Sepharose beads was not degraded by normal human fibroblasts nor did it trigger inhibition of sterol synthesis. The Sepharose beads loaded with LDL bound very tightly to the surface both of normal fibroblasts and fibroblasts from a subject with homozygous familial hypercholesterolemia; control Sepharose beads (activated sites covered with glycine) did not adhere to either cell type. LDL was extracted by a modification of the method of Gustafson (Gustafson, A. (1965) J. Lipid Res. 6, 512-517), so as to remove essentially all cholesterol, cholesterol ester and triglyceride. This modified LDL was bound, internalized and degraded as well as or better than native LDL. However, it failed to suppress sterol synthesis. These results provide additional evidence that the sterol moiety of the LDL is the key component affecting sterol synthesis. They also imply that the neutral lipids of LDL play a minor role in the binding of LDL to cell membranes and that the apoprotein rather than molecular size and shape is the critical factor.     

Provision of cholesterol to lymphocytes by high density and low density lipoproteins. Requirement for low density lipoprotein receptors

The capacity of lipoprotein fractions to provide cholesterol necessary for human lymphocyte proliferation was examined. When endogenous synthesis of cholesterol was blocked, proliferation of mitogen-stimulated normal human lymphocytes was markedly inhibited unless an exogenous source of sterol was supplied. All lipoprotein fractions with the exception of high density lipoprotein subclass 3 were able to provide cholesterol for lymphocyte proliferation. Each of the lipoprotein subfractions capable of providing cholesterol was also able to regulate endogenous sterol synthesis in cultured human lymphocytes. Provision of cholesterol by lipoproteins required the interaction of apolipoprotein B or apolipoprotein E with specific receptors on normal lymphocytes. Apolipoprotein modification by acetylation or methylation, which markedly reduced the ability to regulate sterol biosynthesis, also diminished the capacity of lipoproteins to provide cholesterol. In addition, depletion of apolipoprotein B- and apolipoprotein E-containing particles from high density lipoprotein decreased its ability to suppress cholesterol synthesis and prevented it from providing cholesterol to proliferating lymphocytes. Monoclonal antibodies directed against the receptor-recognition sites on apolipoprotein B and apolipoprotein E were used to define the specific apolipoproteins required for the provision of cholesterol to lymphocytes by the various lipoprotein fractions. The antibody to apolipoprotein B inhibited cholesterol provision by both low density lipoprotein (LDL) and other lipoprotein fractions. The antibody to apolipoprotein E did not decrease provision of cholesterol by LDL but did inhibit the capacity of other fractions to provide cholesterol. In addition, a monoclonal antibody against the ligand binding site on the LDL receptor inhibited provision of cholesterol to normal lymphocytes by all lipoproteins. Finally, lymphocytes lacking LDL receptors were unable to obtain cholesterol from any lipoprotein fraction. These studies demonstrate that LDL receptor-mediated interaction with apolipoprotein B or apolipoprotein E is essential for the provision of cholesterol to normal human lymphocytes from all lipoprotein sources.     

Interactions of high density lipoproteins with very low and low density lipoproteins during lipolysis

Interactions of high density lipoproteins (HDL) with very low (VLDL) and low (LDL) density lipoproteins were investigated during in vitro lipolysis in the presence of limited free fatty acid acceptor. Previous studies had shown that lipid products accumulating on lipoproteins under these conditions promote the formation of physical complexes between apolipoprotein B-containing particles (Biochim. Biophys. Acta, 1987. 919: 97-110). The presence of increasing concentrations of HDL or delipidated HDL progressively diminished VLDL-LDL complex formation. At the same time, association of HDL-derived apolipoprotein (apo) A-I with both VLDL and LDL could be demonstrated by autoradiography of gradient gel electrophoretic blots, immunoblotting, and apolipoprotein analyses of reisolated lipoproteins. The LDL increased in buoyancy and particle diameter, and became enriched in glycerides relative to cholesterol. Both HDL2 and HDL3 increased in particle diameter, buoyancy, and relative glyceride content, and small amounts of apoA-I appeared in newly formed particles of less than 75 A diameter. Association of apoA-I with VLDL or LDL could be reproduced by addition of lipid extracts of lipolyzed VLDL or purified free fatty acids in the absence of lipolysis, and was progressively inhibited by the presence of increasing amounts of albumin. We conclude that lipolysis products promote multiple interactions at the surface of triglyceride-rich lipoproteins undergoing lipolysis, including physical complex formation with other lipoprotein particles and transfers of lipids and apolipoproteins. These processes may facilitate remodeling of lipoproteins in the course of their intravascular metabolism.     

Catabolism of human low density lipoproteins by human hepatoma cell line HepG2

The mechanism of hepatic catabolism of human low density lipoproteins (LDL) by human-derived hepatoma cell line HepG2 was studied. The binding of 125I-labeled LDL to HepG2 cells at 4 degrees C was time dependent and inhibited by excess unlabeled LDL. The specific binding was predominant at low concentrations of 125I-labeled LDL (less than 50 micrograms protein/ml), whereas the nonsaturable binding prevailed at higher concentrations of substrate. The cellular uptake and degradation of 125I-labeled LDL were curvilinear functions of substrate concentration. Preincubation of HepG2 cells with unlabeled LDL caused a 56% inhibition in the degradation of 125I-labeled LDL. Reductive methylation of unlabeled LDL abolished its ability to compete with 125I-labeled LDL for uptake and degradation. Chloroquine (50 microM) and colchicine (1 microM) inhibited the degradation of 125I-labeled LDL by 64% and 30%, respectively. The LDL catabolism by HepG2 cells suppressed de novo synthesis of cholesterol and enhanced cholesterol esterification; this stimulation was abolished by chloroquine. When tested at a similar content of apolipoprotein B, very low density lipoproteins (VLDL), LDL and high density lipoproteins (HDL) inhibited the catabolism of 125I-labeled LDL to the same degree, indicating that in HepG2 cells normal LDL are most probably recognized by the receptor via apolipoprotein B. The current study thus demonstrates that the catabolism of human LDL by HepG2 cells proceeds in part through a receptor-mediated mechanism.     

Selective accumulation of low density lipoproteins in damaged arterial wall

To determine whether damaged arterial wall selectively accumulates lipoproteins, normocholesterolemic rabbits were injected with human radiolabeled low density lipoproteins, high density lipoproteins, and/or albumin 24 hr to 12 weeks after balloon-catheter de-endothelialization of the abdominal aorta. When 125I-labeled low density lipoproteins and 99mTc-labeled albumin were injected simultaneously, the amount of 125I-low density lipoprotein present 24 hr later in abdominal aortas increased steadily, for several weeks, above the amount present at 24 hr in control animals. The increase correlated closely with the degree of re-endothelialization and correlated closely with the degree of re-endothelialization and reached an average maximum for the whole abdominal aorta of three times control when re-endothelialization was between 75 and 85% complete. By contrast, the amounts of 99mTc-albumin or 125I-labeled high density lipoprotein in balloon-damaged abdominal aortas, and the amounts of 125I-low density lipoprotein, 125I-high density lipoprotein, or 99mTc-albumin in undamaged thoracic aortas of injured animals showed no such increase. As early as 2 weeks after de-endothelialization, en face radioautographs made following injection of 125I-labeled low density lipoproteins revealed localized areas of greatest radioactivity around the leading edges of regenerating endothelial islands, broad areas of intermediate radioactivity corresponding to the de-endothelialized areas, and very like radioactivity in the re-endothelialized areas. This pattern occurred rarely with 125I-labeled high density lipoproteins and not at all with 125I-labeled albumin. The results suggest that low density lipoproteins are selectively accumulated by the healing rabbit aorta and that the accumulation is greatest in regions where the endothelium is actively regenerating.     

Effect of homocysteinylation of low density lipoproteins on lipid peroxidation of human endothelial cells

Homocysteine-thiolactone (HcyT) is a toxic product whose synthesis is directly proportional to plasma homocysteine (Hcy) levels. Previous studies demonstrated that the interaction between HcyT and low density lipoproteins (LDL) induces the formation of homocystamide-LDL adducts (Hcy-LDL). Structural and functional alterations of Hcy-LDL have been described and it has been suggested that homocysteinylation could increase atherogenicity of LDL. Oxidative damage of endothelial cells (EC) is considered to be a critical aspect of the atherosclerotic process. To further investigate the molecular mechanisms involved in the atherogenicity of homocysteinylated LDL, we studied the effect of interaction between Hcy-LDL and EC on cell oxidative damage, using human aortic endothelial cells (HAEC) as experimental model. Homocysteinylation of LDL was carried out by incubation of LDL, isolated from plasma of healthy normolipemic subjects, with HcyT (10-100 microM). In our experimental conditions, homocysteinylation treatment was not accompanied by oxidative damage of LDL. No modifications of apoprotein structure and physico-chemical properties were observed in Hcy-LDL with respect to control LDL (c-LDL), as evaluated using the intrinsic fluorescence of tryptophan and the probe Laurdan incorporated in lipoproteins. Our results demonstrated that Hcy-LDL incubated at 37 degrees C for 3 h with HAEC, induced an oxidative damage on human EC with a significant increase of lipid hydroperoxides in cells incubated with Hcy-LDL with respect to cell incubated with c-LDL. The compositional changes were associated with a significant decrease viability in cells treated with Hcy-LDL. The relationship between the levels of -SH groups of LDL and the oxidative damage of HAEC has been demonstrated. These results suggest that Hcy-LDL exert a cytotoxic effect that is likely related to an increase in lipid peroxidation and oxidative damage of EC.     

Effect of lipopolysaccharide toxin on lipid and protein composition of human serum low density lipoproteins

Complexes of lipopolysaccharide (LPS) B of Salmonella typhimurium with human low density lipoproteins (LDL) formed during in vitro coincubation via spontaneous incorporation of LPS (complex LDL-LPS) or through the incorporation stimulated by the serum protein fraction (LPS/LDL complex) were studied. The LPS/LDL complex was shown to maximally bind 0.24 mg of LPS per 1 mg of LDL protein, whereas the LDL-LPS complex contained only 0.07 mg of LPS per 1 mg of LDL protein. The observed incorporation of LPS into LDL particles was not possibly associated with a transfer of lipids or proteins from high density lipoproteins to LDL. The insertion of LPS was probably accompanied by the expulsion of a small portion of phosphatidylcholine molecules from the outer monolayer of LDL into the aqueous medium and by an increase in the phosphatidylethanolamine concentration in LDL. Simultaneously, the level of esterified cholesterol in the LPS/LDL complex decreased, and the concentrations of free cholesterol and triacylglycerols showed a rise. The level of free fatty acids in the LPS/LDL complex increased more than twofold compared with intact LDL. The enhancement of LPS incorporation did not result in the insertion of any serum proteins into LDL, in which apoB-100 remained the major apolipoprotein (ca. 90%); apoB-100 fragments made up to 5-7%, whereas apoE and apoC contained altogether ca. 3-5%. It is suggested that the LPS/LDL complex obtained can bind to three types of cell receptors, i.e., apoB/E receptors, LPS receptors and scavenger receptors of macrophages (monocytes); the increased level of free fatty acids in the LPS/LDL complex may accelerate its subsequent catabolism.     

Recycling of vitamin E in human low density lipoproteins.

Oxidative modification of low density lipoproteins (LDL) and their unrestricted scavenger receptor-dependent uptake is believed to account for cholesterol deposition in macrophage-derived foam cells. It has been suggested that vitamin E that is transported by LDL plays a critical role in protecting against LDL oxidation. We hypothesize that the maintenance of sufficiently high vitamin E concentrations in LDL can be achieved by reducing its chromanoxyl radicals, i.e., by vitamin E recycling. In this study we demonstrate that: i) chromanoxyl radicals of endogenous vitamin E and of exogenously added alpha-tocotrienol, alpha-tocopherol or its synthetic homologue with a 6-carbon side-chain, chromanol-alpha-C6, can be directly generated in human LDL by ultraviolet (UV) light, or by interaction with peroxyl radicals produced either by an enzymic oxidation system (lipoxygenase + linolenic acid) or by an azo-initiator, 2,2'-azo-bis(2,4-dimethylvaleronitrile) (AMVN; ii) ascorbate can recycle endogenous vitamin E and exogenously added chromanols by direct reduction of chromanoxyl radicals in LDL; iii) dihydrolipoic acid is not efficient in direct reduction of chromanoxyl radicals but recycles vitamin E by synergistically interacting with ascorbate (reduces dehydroascorbate thus maintaining the steady-state concentration of ascorbate); and iv) beta-carotene is not active in vitamin E recycling but may itself be protected against oxidative destruction by the reductants of chromanoxyl radicals. We suggest that the recycling of vitamin E and other phenolic antioxidants by plasma reductants may be an important mechanism for the enhanced antioxidant protection of LDL.