LDL resistance to oxidation: Effects of lipid phenotype, autologous HDL and alanine

Clinica Chimica Acta 379 (2007) 95 – 100 www.elsevier.com/locate/clinchim

LDL resistance to oxidation: Effects of lipid phenotype, autologous HDL and alanine Simona Baldi ⁎, Silvia Frascerra, Ele Ferrannini, Andrea Natali Department of Internal Medicine and C N R Institute of Clinical Physiology, University of Pisa School of Medicine, Pisa, Italy Received 17 November 2006; received in revised form 19 December 2006; accepted 19 December 2006 Available online 5 January 2007

Abstract Background: Although LDL resistance to copper-induced oxidation is a time-honoured method, how it is modulated by the physiologic variability of lipid phenotype and what influences the protective action of homologous HDL and exogenous alanine is still unclear. Methods: In 159 subjects without severe dyslipidemias, LDL resistance to copper-induced oxidation (lag phase) was measured under standardised conditions, with alanine and with autologous HDL. Results: Lag phase was normally distributed and averaged 68 ±10 min (range: 40–105 min). Both VLDL-triglycerides (37 ±5, 52 ±7, 59 ±7, 53 ±5 mg/dl, p b 0.05) and LDL-triglycerides (27 ±2, 27±1, 30±2, 35± 3 mg/dl, pb 0.01) increased across quartiles of lag phase. The relative LDL enrichment in triglycerides (triglycerides percent or triglycerides/cholesterol ratio) was strongly related to lag phase (r =0.29 and r= 0.31, p b 0.0005 for both) independently of age, gender, BMI, and presence of diabetes or hypertension. The protective effect of HDL was variable (+42 ±18 min) and largely dependent on the capacity of HDL to resist oxidation (r=0.69, p b 0.0001). Alanine induced a rather constant lag phase prolongation (+32±7 min) that was weakly related only to baseline lag phase (r= 0.17, pb 0.05). Conclusions: Relative triglyceride abundance protects LDL from ex-vivo oxidation, HDL particles protect LDL mainly through substrate dilution and alanine probably through a direct anti-oxidant effect. © 2007 Elsevier B.V. All rights reserved. Keywords: LDL oxidation; Lag phase; HDL oxidation; Anti-oxidants

1. Introduction The chemical modification of low-density lipoproteins (LDL) induced by oxidative reactions is a crucial step in the atherosclerotic process [1–3]. During its lifespan, and especially in their traffic within the arterial wall, LDL particles are exposed to a wide range of chemical insults, the net effect of which depends on the balance between the intensity of the challenge and the efficiency of the defence mechanisms. When oxygen radicals are overproduced – and/or lipoprotein particles are more vulnerable to this injury – the oxidative process induces modifications of the Apo-B protein that make the LDL particle less efficiently handled by the physiologic B-E

⁎ Corresponding author. Dipartimento di Medicina Interna, Via Roma 67, 56100 Pisa, Italy. Tel.:+39 050 992698; fax: +39 050 553235. E-mail address: [email protected] (S. Baldi). 0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2006.12.019

receptor-mediated pathway and more prone to accumulate into macrophages through the scavenger pathway [4]. LDL particles are physiologically protected from oxidation by the presence of anti-oxidants in the aqueous ambient or within their lipid layer. The vulnerability to oxidative modification has traditionally been estimated ex vivo by measuring the generation of conjugated dienes upon challenging LDL particles with a strong pro-oxidant such as copper or iron [5]. By using this technique, studies have confirmed the relevance of anti-oxidants in LDL susceptibility to oxidative damage [6]. Despite the relative constancy of vitamin intake and also of LDL vitamin content in homogeneous populations, LDL particles still show a wide range of susceptibility degrees to oxidative stress [7]. The fatty acids composition of triglycerides (TG, monounsaturated vs polyunsaturated or saturated) [8] and the particle size/density (small dense vs large buoyant) [9] have been proposed as factors responsible for the variable LDL resistance to oxidation. However, in quantitative terms the impact of either aspect

96

S. Baldi et al. / Clinica Chimica Acta 379 (2007) 95–100

Fig. 1. Scheme of the kinetics of conjugated dienes formation from LDL particles incubated with copper. Absmin indicates the absorbance at time 0, A slope and B slope indicate the slope of the first and second linear portion of the curve, respectively, ΔAbs indicates the absolute change in the absorbance relative to time 0.

results modest and a large proportion of the biological variability remains unexplained. In vivo clearly several factors modulate LDL oxidability and among these HDL appears to play a relevant role showing a strong protective effect as documented by the prolongation of the lag phase in experiments in which the two substrates were simultaneously exposed to oxidation[10–13]. Although this effect has been traditionally ascribed to the presence of antioxidant enzymes (such as paraoxonase or platelet-activating factor acetylhydrolase), no study has directly verified this hypothesis nor other mechanisms have been searched. Finally, rather than LDL susceptibility to oxidation the degree of spontaneously occurring LDL oxidation/modification would probably better reflect the extent of in-vivo oxidative stress. Unfortunately, the direct determination of these minimally modified LDL particles is hampered by technical difficulties. The protective effect of alanine on LDL oxidisability has been proposed as a valid alternative to estimate the amount of lipid peroxides present in native LDL particles [14]. However, whether this specific characteristic of alanine, which has been inferred from its chemical properties (its antioxidant action being less effective when lipid peroxides are abundant) also depends on lipoprotein composition has not been investigated. This study was set forth to establish whether LDL susceptibility to oxidation is influenced by differences in lipid and/ or protein composition and to describe what affects the protective effect of either HDL and alanine. 2. Methods

2.2. Measurement of LDL and HDL oxidation For the measurement of in-vitro LDL and HDL oxidation, we followed the experimental protocol described by Esterbauer et al. [5] with some modifications. Briefly, blood (10 ml) was collected in tubes containing EDTA (1 mg/ml) and immediately centrifuged at 3000 rpm at 4 °C. Very-low density and intermediate-density (VLDL-IDL), low-density (LDL), and highdensity (HDL) lipoproteins (d = 1.006–1.019, 1.019–1.063 and 1.063–1.210, respectively) were isolated by sequential ultracentrifugation at 42,000 rpm at 4 °C (OPTIMA L-90 K, Beckman, CA, USA) in NaBr solutions with EDTA to avoid oxidation during isolation. All solutions were degassed and samples were kept at 4 °C in a dark environment. To remove EDTA, lipoprotein fractions were gel-filtered on a Sephadex column (Econo-Pac 10 DG columns; Bio-Rad), and eluted in 1M PBS buffer, pH 7.4. LDL particle concentration was adjusted at 50 μg of protein per ml by measuring total protein with the bicinchoninic acid method before a freshly prepared aqueous copper solution (CuSO4, final concentration 1.0 μM) was added. The kinetics of conjugated dienes formation were followed spectrophotometrically at 234 nm at 37 °C every 1.5 min for 4 h and expressed as optical density (OD) units. A typical copper-induced LDL oxidation curve shows three consecutive phases: the lag phase, the propagation phase and the decomposition phase. The lag phase is defined as the time at which the first phase (A) and second phase (B) cross each other (Fig. 1). Once the experimental conditions are maintained constant, this time length depends on the intrinsic properties of the LDL particle (e.g., antioxidant content, size, pre-formed lipid peroxides). The y-axis intercept of the A slope was defined as minimal absorbance (Absmin) and the maximal increment above this value reached during the experiment was defined as ΔAbs. The antioxidant effect of alanine on in-vitro LDL oxidation was tested in separate experiments by measuring the lag phase in the presence of 0.05 mM alanine. For the measurement of HDL lag phase we followed the same protocol as for LDL but the HDL protein concentration was set at 68 μg/ml. The effect of HDL on LDL oxidation was measured as the lag phase of LDL oxidation in a

Table 1 Characteristics of the study subjects

2.1. Study subjects Plasma samples were taken after an overnight (10–14 h) fast from a total of 159 consecutive patients attending our outpatient clinic in whom a complete clinical work-up had excluded significant liver, kidney and thyroid dysfunction, severe hypertriglyceridaemia (serum TG N250 mg/dl) or hypercholesterolaemia (serum total cholesterol N250 mg/dl), systemic inflammatory diseases or current treatment with anti-oxidants, statins, and fibrates. Patients with impaired fasting glucose (≥5.6 mM and b7 mM) underwent a standard 75 g oral glucose tolerance test to establish the diagnosis.

n Gender (M/F) Age (years) BMI (kg/m2) Total serum cholesterol (mg/dl) Serum LDL-cholesterol (mg/dl) Serum HDL-cholesterol (mg/dl) Serum triglycerides (mg/dl)

Mean ± SD

Range

159 83/76 47 ± 12 26.4 ± 4.6 187 ± 36 117 ± 32 52 ± 16 90 ± 46

24–71 18.0–41.0 82–250 31–189 20–96 34–250

S. Baldi et al. / Clinica Chimica Acta 379 (2007) 95–100

97

Table 3 Cholesterol and triglycerides content of lipoproteins by quartile of lag phase II

III

IV

p

Cholesterol (mg/dl) VLDL + IDL 7±1 LDL 124 ± 5 HDL 60 ± 3 Total 191 ± 5

I

11 ± 2 122 ± 5 51 ± 3 184 ± 7

13 ± 2 123 ± 5 47 ± 3 183 ± 5

10 ± 1 130 ± 6 51 ± 2 191 ± 5

0.018 n.s. 0.013 n.s

Triglycerides (mg/dl) VLDL + IDL 37 ± 5 LDL 27 ± 2 HDL 10 ± 1 Total 74 ± 5

52 ± 7 27 ± 1 10 ± 1 89 ± 7

59 ± 7 30 ± 2 10 ± 1 99 ± 9

53 ± 5 35 ± 3 11 ± 1 99 ± 6

0.049 0.008 n.s. 0.002

Fig. 2. Histogram representing the distribution of lag phase values in the study population. The curve represents a normal distribution. mixture of LDL (50 μg/ml) and HDL (50 μg/ml) at 37 °C in the presence of 1 μM CuSO4.

2.3. Measurement of lipid parameters Total cholesterol, TG and protein concentration in serum and lipoprotein fractions, and serum HDL-cholesterol levels were determined by enzymatic-colorimetric methods using a Beckman Synchron CX4 Analyser. Serum LDL-cholesterol concentration was calculated using Friedewald's formula.

2.4. Statistical analysis Results are expressed as mean ± SE M unless otherwise specified. Data were analysed by using analysis of variance, simple and multiple linear regression.

3. Results The study population covered wide ranges of age and BMI (Table 1). As expected of subjects attending a metabolic clinic, several patients had diabetes (19%), impaired glucose tolerance (9%), hypertension (8%) or subclinical hypothyroidism (15%). In the whole cohort, LDL resistance to oxidation was normally distributed, with lag phase values ranging from 40 to 105 min, (Fig. 2). When subjects were grouped into quartiles of lag phase, age (39 ± 2, 48 ± 2, 46 ± 2 and 53 ± 2 years, p b 0.0001 by ANOVA) and BMI (23.9 ± 0.6, 26.3 ± 0.7, 26.8 ± 0.7 and 28.3 ± 0.7, p b 0.0002 by ANOVA) increased progressively, whereas gender distribution was similar (percent male: 43, 48, 48 and 36, p = n.s.).

Table 2 Parameters of in-vitro LDL oxidisability by quartile of lag phase I Lag phase (min) A slope (OD · min− 1 · 103) B slope (OD · min− 1 · 103) ΔAbs (OD) Absmin (OD)

II

III

IV

p

56.5 ± 0.8 64.6 ± 0.3 70.4 ± 0.3 81.5 ± 1.2 b0.0001 2.29 ± 0.06 2.18 ± 0.05 1.91 ± 0.04 1.81 ± 0.06 b0.0001 12.1 ± 0.3

12.1 ± 0.3

11.8 ± 0.2

11.2 ± 0.3

0.043

0.86 ± 0.02 0.88 ± 0.02 0.85 ± 0.01 0.89 ± 0.02 n.s. 0.36 ± 0.01 0.36 ± 0.01 0.35 ± 0.01 0.36 ± 0.01 n.s

As shown in Table 2, higher lag phase values were associated with progressively lower A and B slope values while both Absmin and ΔAbs were similar. These associations were continuous, as confirmed by univariate regression analysis. In a step-wise regression model, the A slope, Absmin and ΔAbs were significant independent predictors of lag phase, with partial r values of − 0.45, +0.12, − 0.07, respectively, together explaining 42% of lag phase variability. The serum lipid profile and lipoprotein composition are given in Table 3. Subjects with longer LDL lag phases were characterised by increasing serum TG – mainly vehicled by VLDL + IDL and LDL particles – and by a re-distribution of cholesterol from HDL to VLDL + IDL (and, marginally, to LDL). LDL-TG were more closely associated with lag phase (r = 0.29, p b 0.0002) than were total serum TG (r = 0.20, p = 0.05); even stronger was the association of lag phase with the TG/Chol ratio in LDL (Fig. 3), which remained significant also after accounting for gender, age, obesity, diabetes, hypertension and total serum TG and cholesterol concentrations (partial r = 0.29, p b 0.005). Adding either HDL (50 μg of protein/ml) or alanine (0.05 M) to LDL particles (50 μg of protein/ml) resulted in a substantial and quantitatively similar percent increment in lag phase by 62 ± 30 and 48 ± 11% respectively, (mean ± SD) (Fig. 4). In absolute terms, however, the effect of alanine was rather constant (mean ± SD: + 32 ± 7 min, CV = 21%) while the lag phase prolongation induced by HDL showed a greater interindividual variability (+ 42 ± 18 min, CV = 43%). Such variability was, at least in part, determined by the heterogeneity of the intrinsic HDL ability to resist copper-induced oxidation (Fig. 5), which, on average, was also better than that of LDL (87 ± 20 vs 68 ± 10 min, p b 0.0001). This lower HDL susceptibility to oxidation was evident also when the two lipoproteins were compared at identical (68 μg /ml) protein concentration (94 ± 27 vs 80 ± 13 min, p b 0.05, n = 65). On the other hand, no relationship was observed between the protective effect of HDL and age, BMI, diabetes, gender, serum lipid concentration or lipoprotein composition. The protective effect of alanine on LDL oxidability was only modestly dependent on baseline LDL lag phase (r = 0.17, p b 0.05) – while the intercept of the model (25 ± 4 min) was highly statistically significant ( p b 0.0001) – once this

98

S. Baldi et al. / Clinica Chimica Acta 379 (2007) 95–100

Fig. 3. Scatterplot of lag phase against LDL triglyceride content expressed as percent of total LDL mass (left) or normalised to the cholesterol LDL mass (right). The dotted lines represent the 95% confidence intervals for the estimation of the mean.

influence was taken into account in a multiple regression model, the effect of alanine was related neither to clinical nor biochemical variables. 4. Discussion In our relatively unselected population of subjects attending an outpatient clinic for metabolic diseases – from which patients with clinically manifest endocrine, liver, kidney and inflammatory diseases were excluded – LDL resistance to in-vitro oxidation showed a normal distribution and a large variability. Somewhat to our surprise, we observed that resistance to oxidation improved with age and obesity (as the BMI). The associations between age and BMI with lag phase were continuous (r = 0.30 and r = 0.30, respectively, p b 0.01 for both) and additive, with estimated independent contributions of 3 ± 1 min per decade of age ( p b 0.001) and 5 ± 2 min per 10 units of BMI ( p b 0.01), together explaining 20% of overall lag phase variability. The strength of these associations was not influenced when gender, diabetes, hypertension or serum TG were entered in the model as covariates. This finding is in contrast with the common notion, actually based on limited data

[15–19], that obesity and ageing are characterised by a progressive shortening of the lag phase. Although our study was neither designed nor powered to specifically test the influence of obesity or age on LDL oxidisability, the strength of the associations and the consistency of the data make it unlikely that either a selection bias or methodology might explain the discrepancy. In addition, other studies have reported either unchanged [20] or prolonged lag phase [21,22] in the elderly. Other factors related to ageing (e.g., dietary habits, concomitant disease, increased LDL plasma residency time) may influence LDL oxidisability and explain differences among studies. With regard to the effect of obesity, two studies (from the same laboratory) that reported a negative effect of obesity on lipoprotein oxidisability [15,16] actually measured the lag phase of non-HDL particles (i.e., VLDL + IDL + LDL), while the other study [17] involved a very selected group of nondiabetic, normotensive patients with morbid obesity, which makes a comparison with our results problematic. In our experiments, resistance to copper-induced oxidation was clearly associated with aspects of the kinetics of conjugated dienes formation. The results of multivariate analysis suggest that a higher resistance is largely a function of a slower initiation

Fig. 4. Individual (and mean) lag phase values of LDL alone (50 μg/ml) and in the presence of HDL (50 μg/ml) (left) or alanine (0.05 mM) (right).

S. Baldi et al. / Clinica Chimica Acta 379 (2007) 95–100

Fig. 5. Scatterplot of the absolute increment in LDL lag phase due to the presence of HDL vs the lag phase of HDL alone.

of the oxidation process (A slope), with ΔAbs – a crude measure of all available substrates prone to oxidation – and Absmin – a crude estimate of basal LDL oxidation (to the extent 234 nM absorbance is determined by the sample concentration of conjugated dienes) – playing minor roles. Interestingly, the speed of the propagation phase (B slope) does not appear to contribute to the overall variability of lag phase. The lipoprotein phenotype of the subjects with more resistant LDL was characterised by a relative redistribution of cholesterol from HDL to both VLDL + IDL and LDL particles coupled with an absolute increment in serum TG vehicled both by LDL and VLDL + IDL particles. Since TG concentration in LDL was more strongly associated with lag phase than the total serum TG concentration, we evaluated which aspect of LDL composition influenced more their vulnerability to oxidation by estimating the TG content in LDL as percent of its total mass or normalising it for either LDL protein (r = 0.28, p = 0.06 data not shown) or LDL cholesterol content (Fig. 3). This analysis indicated that ‘resistant’ LDL particles carry more TG and are relatively depleted of cholesterol, suggesting that partially delipidated LDL (i.e., smaller and more dense LDL) are less represented. However, the total lipid-to-protein ratio of these particles did not change across quartiles of lag phase (2.12 ± 0.47, 2.15 ± 0.37, 2.09 ± 0.29, 2.15 ± 0.39, mean ± SD, p = n.s.), and to the extent that the lipid-to-protein ratio reflects lipoprotein size, it appears unlikely that ‘resistant’ LDL have a larger size. Our results indicate that whenever LDL particles are relatively enriched with TG (e.g., reduced lipoprotein lipase activity [23]) they became more resistant to oxidation as measured by the kinetics of conjugated diene formation after exposure to copper. This observation is qualitatively and quantitatively consistent with the enhanced lipoprotein resistance to oxidation that has been observed in patients with hypertriglyceridaemia and its normalisation following bezafibrate treatment[24]. In that study, the LDL TG-to-cholesterol ratio was 0.288 in the patients and 0.157 in the control subjects and the observed mean lag phase of the patients was ∼14% higher than in controls; for such a difference in LDL composition, our regression data (Fig. 3) predict a similar (∼10%) increment in lag phase. A consistent increment in LDL resistance to copper-induced oxidation has also been

99

induced by acutely manipulating the TG content of LDL in-vivo (by means of a prolonged infusion of a lipid emulsion) [25]. In that study, the LDL TG-to-cholesterol ratio was raised from 0.143 to 0.212 and lag phase increased by approximately 35%. This higher-than-expected rise was probably due to the αtocopherol present in the emulsion. Finally, when the triglyceride content of LDL was increased in vitro by reconstituting delipidated LDL particles with trilinolein their resistance to oxidation was markedly improved [26]. Thus, our findings extend into the physiological domain what has been observed in pathologic states or under experimental conditions. With regard to the mechanisms, since this study was not designed for this purpose, we can only hypothesise that TG, having on average a higher prevalence of saturated fatty acids with respect to the other oxidisable lipids present into the lipoproteins (cholesteryl ester and phospholipids), probably represent a less favourable substrate for oxidation and in addition they are the main vehicle for vitamin E, one of the most efficient antioxidant defense. Although we cannot establish whether this protective effect of LDL-TG on ex-vivo oxidability has biological relevance to atherosclerosis, the current observations may contribute to explain the uncertainties surrounding the role of total serum TG as cardiovascular risk factor [27]. According to our data, and similarly to previous studies [28], HDL particles appear to protect LDL from being oxidised (as indicated by the longer lag phase of the mixture). In an attempt to establish whether this effect is determined by a peculiar interaction between the two particles or is the consequence of the increment in substrate that is made available to react with copper and/or the intrinsic ability of HDL to resist to oxidation, we estimated the impact of substrate concentration on lag phase. As shown in Fig. 6, by increasing the substrate concentration we observed a progressive – though less than proportional – prolongation of lag phase of both LDL and HDL. It is therefore likely that the protective effect of HDL has been overestimated, being determined, at least in part, by the quenching effect of the added substrate — in analogy with what is observed when the same amount of LDL is exposed to a lower copper concentration.

Fig. 6. Dose-response of substrate concentration and lag phase respectively of LDL (black symbols), HDL (white symbols) or a balanced mixture of LDL + HDL particles (gray symbols). Data obtained using the same lipoproteins pooled from different patients.

100

S. Baldi et al. / Clinica Chimica Acta 379 (2007) 95–100

Only by comparing the lag phase of the single lipoproteins with the lag phase of the mixture at equal protein concentration (a situation also closer to the in-vivo conditions), can the direct effect of HDL on LDL oxidisability be estimated correctly. The data shown in Fig. 6 indicate that this anti-oxidant action of HDL, though small, is evident at physiologic substrate concentrations. However, the variability of lag phase of LDL + HDL mixtures is largely determined by the interindividual differences in HDL resistance to copper-induced oxidation (Fig. 5). The strength of this association is such that we conclude that little additional information is gained from the measurement of lag phase in HDL and LDL mixtures that is not already provided by the evaluation of HDL oxidisability. Alanine prolongs the lag phase by interfering with the reaction of copper and lipids; its effect is considered to be inversely proportional to the amount of naturally occurring lipid peroxides [14,29]. The relative constancy of the absolute increment in LDL lag phase (Fig. 4) – with a coefficient of variation (21%) approaching the intra-assay coefficient of variability (10%) of the estimate of this effect – even in a heterogeneous population such as our study group questions the possibility to use this variable as a tool to estimate the degree of ‘native’ LDL oxidation and suggests the possibility that alanine acts as a pure antioxidant substance delaying the reaction between copper and peroxides in a non specific pattern. In conclusion, the resistance of LDL particles to copperinduced oxidation is essentially related to their relative TG content. By adding HDL particles to LDL the lag phase of the mixture is prolonged mainly because of the dilution with a substrate inherently more resistant to oxidation (with little direct protective effect). Alanine exerts an anti-oxidant action that is unlikely to be determined by characteristics of LDL different from their resistance to oxidation. Acknowledgements We wish to thank Sara Burchielli for her technical assistance. References [1] Steinberg D, Parthasarathy S, Carew TE, Khoo JC, Witztum JL. Beyond cholesterol. Modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med 1989;320:915–24. [2] Stocker R, Keaney Jr JF. Role of oxidative modifications in atherosclerosis. Physiol Rev 2004;84:1381–478. [3] Nakajima K, Nakano T, Tanaka A. The oxidative modification hypothesis of atherosclerosis: the comparison of atherogenic effects on oxidized LDL and remnant lipoproteins in plasma. Clin Chim Acta 2006;367:36–47. [4] Grundy SM. Oxidized LDL and atherogenesis: relation to risk factors for coronary heart disease. Clin Cardiol 1993;16:I3–5. [5] Esterbauer H, Striegl H, Puhl H, Dieber-Rotheneder M. Continuous monitoring of in vitro oxidation of human low density lipoprotein. Free Radical Res Commun 1989;6:67–75. [6] Jialal I, Fuller CJ. Effect of vitamin E, vitamin C and beta-carotene on LDL oxidation and atherosclerosis. Can J Cardiol 1995;11:97G–103G [Suppl G]. [7] Cominacini L, Garbin U, Cenci B, et al. Predisposition to LDL oxidation during copper-catalyzed oxidative modification and its relation to alphatocopherol content in humans. Clin Chim Acta 1991;204:57–68.

[8] Kontush A, Hubner C, Finckh B, Kohlschutter A, Beisiegel U. How different constituents of low density lipoprotein determine its oxidizability by copper: a correlational approach. Free Radic Res 1996;24:135–47. [9] Chait A, Brazg RL, Tribble DL, Krauss RM. Susceptibility of small, dense, low-density lipoproteins to oxidative modification in subjects with the atherogenic lipoprotein phenotype, pattern B. Am J Med 1993;94:350–6. [10] Parthasarathy S, Barnett J, Fong LG. High-density lipoprotein inhibits the oxidative modification of low-density lipoprotein. Biochim Biophys Acta 1990;1044:275–83. [11] Sanguinetti SM, Brites FD, Fasulo V, et al. HDL oxidability and its protective effect against LDL oxidation in Type 2 diabetic patients. Diabetes Nutr Metab 2001;14:27–36. [12] Navab M, Hama SY, Anantharamaiah GM, et al. Normal high density lipoprotein inhibits three steps in the formation of mildly oxidized low density lipoprotein: steps 2 and 3. J Lipid Res 2000;41:1495–508. [13] Navab M, Hama SY, Cooke CJ, et al. Normal high density lipoprotein inhibits three steps in the formation of mildly oxidized low density lipoprotein: step 1. J Lipid Res 2000;41:1481–94. [14] Picard S, Talussot C, Serusclat A, Ambrosio N, Berthezene F. Minimally oxidised LDL as estimated by a new method increase in plasma of type 2 diabetic patients with atherosclerosis or nephropathy. Diabetes Metab 1996;22:25–30. [15] Van Gaal LF, Zhang A, Steijaert MM, De Leeuw IH. Human obesity: from lipid abnormalities to lipid oxidation. Int J Obes Relat Metab Disord 1995;19(Suppl 3):S21–6. [16] Van Gaal LF, Vertommen J, De Leeuw IH. The in vitro oxidizability of lipoprotein particles in obese and non-obese subjects. Atherosclerosis 1998;137:S39–44 [Suppl]. [17] Myara I, Alamowitch C, Michel O, et al. Lipoprotein oxidation and plasma vitamin E in nondiabetic normotensive obese patients. Obes Res 2003;11:112–20. [18] Balkan J, Kanbagli O, Mehmetcik G, Mutlu-Turkoglu U, Aykac-Toker G, Uysal M. Increased lipid peroxidation in serum and low-density lipoproteins associated with aging in humans. Int J Vitam Nutr Res 2002;72:315–20. [19] Mosinger BJ. Human low-density lipoproteins: oxidative modification and its relation to age, gender, menopausal status and cholesterol concentrations. Eur J Clin Chem Clin Biochem 1997;35:207–14. [20] Schmuck A, Fuller CJ, Devaraj S, Jialal I. Effect of aging on susceptibility of low-density lipoproteins to oxidation. Clin Chem 1995;41:1628–32. [21] Stulnig TM, Jurgens G, Chen Q, et al. Properties of low density lipoproteins relevant to oxidative modifications change paradoxically during aging. Atherosclerosis 1996;126:85–94. [22] Cherubini A, Zuliani G, Costantini F, et al. High vitamin E plasma levels and low low-density lipoprotein oxidation are associated with the absence of atherosclerosis in octogenarians. J Am Geriatr Soc 2001;49:651–4. [23] Packard CJ. Triacylglycerol-rich lipoproteins and the generation of small, dense low-density lipoprotein. Biochem Soc Trans 2003;31:1066–9. [24] de Man FH, de Beer F, van der Laarse A, et al. The hypolipidemic action of bezafibrate therapy in hypertriglyceridemia is mediated by upregulation of lipoprotein lipase: no effects on VLDL substrate affinity to lipolysis or LDL receptor binding. Atherosclerosis 2000;153:363–71. [25] Callow J, Samra JS, Frayn KN. Effect of infusion of a triacylglycerol emulsion on low-density lipoprotein composition and oxidizability. Atherosclerosis 1998;137:115–23. [26] Visioli F, Bordone R, Perugini C, Bagnati M, Cau C, Bellomo G. The kinetics of copper-induced LDL oxidation depend upon its lipid composition and antioxidant content. Biochem Biophys Res Commun 2000;268:818–22. [27] Abdel-Maksoud MF, Hokanson JE. The complex role of triglycerides in cardiovascular disease. Semin Vasc Med 2002;2:325–33. [28] Raveh O, Pinchuk I, Fainaru M, Lichtenberg D. Kinetics of lipid peroxidation in mixtures of HDL and LDL, mutual effects. Free Radic Biol Med 2001;31:1486–97. [29] Carantoni M, Abbasi F, Warmerdam F, et al. Relationship between insulin resistance and partially oxidized LDL particles in healthy, nondiabetic volunteers. Arterioscler Thromb Vasc Biol 1998;18:762–7.