Increased oxidation resistance of atherogenic plasma lipoproteins at high vitamin E levels in non-vitamin E supplemented men

atherosclerosis Atherosclerosis 124 (1996) 83-94

Increased oxidation resistance of atherogenic plasma lipoproteins at high vitamin E levels in non-vitamin E supplemented men Elina Porkkala-Sarataho, Kristiina Nyyssiinen, Jukka T. Salonen* Reasearch Institute

of Public Health, University of Kuopio, P.O. Box 1627, 70211 Kuopio, Finland

Received 14 July 1995; revised 30 January 1996; accepted 8 February 1996

Abstract The oxidative modification of human low density lipoprotein (LDL) has been widely investigated. However, there are no data concerning the oxidation susceptibility of combined very low density lipoprotein (VLDL), intermediate density lipoprotein (IDL) and low density lipoprotein fraction, although all of them are atherogenic and contain antioxidants such as a-tocopherol. We investigated the oxidation susceptibility and oxidation resistance of VLDL + LDL (including IDL) fraction by induction with CuCl, and its relation to plasma a-tocopherol concentration and lipid standardised a-tocopherol concentration in 406 non-vitamin E-supplemented men from eastern Finland. Even thought we did not give oral vitamin E or any other antioxidant supplementation to our study participants, we observed a significant, consistent relationship between measurements of oxidation resistance and plasma content of vitamin E. In the multivariate regression model, a high plasma content of vitamin E or lipid standardised vitamin E concentration were the most important determinants of lag time to maximal oxidation rate (standardised regression coefficient = 0.244, P < 0.0001 for vitamin E and 0.211, P < 0.0001 for lipid standardised vitamin E). After statistical adjustment for age, use of cigarettes, hypolipidemic medication (yes vs. no), month of the measurements, plasma concentrations of total ascorbic acid (ascorbic acid + dehydroascorbic acid), a-carotene and phospholipids, serum concentrations of LDL cholesterol and triglycerides and dietary intake of linoleic acid, the lag time to maximal oxidation rate was 10% (95% C.I. 6.0-13.5%) longer in men in the highest fifth than in the lowest fifth of plasma vitamin E content (P < 0.0001 for trend). When the fifths of lipid standardised vitamin E were compared, the lag time to maximal oxidation rate was 6% (95% C.I. 1.8-10.1%) longer in men in the highest than in the lowest fifth (P < 0.0001 for trend). Our data suggest that a-tocopherol is an important antioxidant preventing the in vitro oxidation of VLDL + LDL fraction even in non-supplemented subjects. Keywords:

Low density lipoprotein; Very low density lipoprotein; Lipid peroxidation;

Vitamin E

* Corresponding author, fax: + 358 71 162936. 0021-9150/96/$15.000 1996 Elsevier Science Ireland Ltd. All rights reserved PII SOO21-9150(96)05821-2

84

E. Porkkala-Sarataho

et al. / Atherosclerosis

There is increasing evidence that the initiation of atherosclerosis is related to lipid peroxidation and oxidative modification of low density lipoproteins (LDL) [l-3]. Native LDL does not appear to be atherogenic [l]. Lipid peroxidation is a free radical mediated oxidation of membrane lipids proceeded by a chain reaction which is influenced by antioxidants and pro-oxidants. It is suggested that radical-mediated lipid oxidation proceeds via similar mechanism in isolated LDL and very low density lipoprotein (VLDL) [4] even though the oxidation of VLDL + LDL has not been studied as much as the oxidation of plain LDL. The best known pro-oxidants are the reactive free radicals present in the gas and tar phases of cigarette smoke [5] and catalytic transition metals such as copper, iron and mercury [6]. LDL and VLDL contain several antioxidant vitamins which can protect atherogenic lipoproteins against oxidative modification. The quantitatively most important antioxidants present in LDL are cc-tocopherol followed by retinyl stearate, y-tocopherol, p-carotene, and lycopene [7]. It is assumed that the ratio of a-tocopherol to coenzyme QlO in VLDL is close to that of LDL [4]. As measured by the capability to prevent oxidative modification of LDL, the most important antioxidant is thought to be a-tocopherol [B]. Studies in WHHL rabbits and cholesterol-fed rabbits have shown the effect of the synthetic antioxbutylated probucol idants E9,lOl and hydroxytoluene (BHT) [l l] in the prevention of lipid peroxidation and of atherogenesis. However, these compounds possessside effects and are not suitable for humans [12,13]. Several studies have been reported that antioxidant supplementation increased the resistance of LDL to oxidative modification [5,14-191. Many of these reports suggest that there is no significant association between the susceptibility to Cu* + -induced oxidation of LDL and the a-tocopherol content of LDL and therefore it has been thought that the a-tocopherol-to-protein ration of LDL from donors not receiving vitamin E supplements could not be used to predict LDL resistance to oxidation induced by metal ions [8,16,20-221. However, there are no previous population studies concerning the oxidation resistance and oxida-

124 (1996) 83-94

tion susceptibility of lipoproteins. The purpose of this population study was to examine the role of vitamin E with regard to the oxidation resistance and oxidation susceptibility of VLDL + LDL in non-vitamin E supplemented men. 2. Subjects and methods 2.1. Study subjects

The present study was carried out in the participants of the ‘Kuopio Atherosclerosis Prevention Study’ (KAPS), a population-based randomised double-masked 3-year trial concerning the effect of pravastatin on progression and regression of atherosclerosis in carotid and femoral arteries in men with elevated LDL cholesterol [23]. Fourhundred-and-forty-seven men, aged 44-65 at the baseline examination, whose serum LDL cholesterol was 2 4.0 mmol/l, serum total cholesterol < 7.5 mmol/l, liver enzymes alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) did not exceed 1.5-fold the laboratory upper normal limit and who had no specific contraindications, were entered in the KAPS. Fourhundred-and-ten men (92%) out of 447 completed the study. The study protocol was approved by an international policy advisory board and by the Research Ethic Committee of the University of Kuopio. Vitamin E content of plasma and the oxidation resistance and oxidation susceptibility of VLDL + LDL were measured from fresh samples from 406 subjects at the final (36-month) visit of the KAPS. Subjects came to the 36-month visits between January and October 1993. 2.2. Plasma samples

Blood samples were obtained by venipuncture after 12 h fasting, collected into vacuum tubes (Venoject, Terumo, Leuven, Belgium). Blood for lipoprotein fractionation was drawn into tubes, which contained ethylenediamine tetraacetic acid (EDTA) and placed on an ice-bath. Plasma was separated by centrifugation for 15 min at + 4°C (3500 g) and was kept in an ice-bath before further handling. Lipoprotein fractionation was

E. Porkkala-Sarataho

et al. i Atherosclerosis

started on the same day when blood was collected. Blood for vitamin E determination was taken to lithium heparine tubes and the plasma samples were stored at - 80°C until use. 2.3. Isolation of VLDL + LDL Combined VLDL and LDL were isolated from fresh EDTA-plasma by ultracentrifugation based on density-gradient ultracentrifugation (KBr gradient). Three millilitres of fresh EDTA-plasma and 625 ,ul of a NaCl + KBr (1.335 g/ml density) solution were mixed, before adding 1.17 ml of a solution (1.063 g/ml density) diluted from the stock solution. All solutions were made in water free from metals (Super Q Plus, Millipore, Bedford, MA). The tubes were centrifuged in a Beckman XL-90 ultracentrifuge (Beckman, Palo Alto, CA, USA) by using Beckman 50.4 Ti rotor at 32000 rpm for 23 h at 4°C. The top layer from the tube ( < 1.063 g/ml density) was collected with a Pasteur pipette and assayed immediately. 2.4. VLDL + LDL oxidation by CuCl, The susceptibility of VLDL + LDL to in vitro oxidation was based on the technique described by Esterbauer et al. [24]. Before oxidation EDTA was removed from VLDL + LDL by using gel filtration technique (PD-10 columns, Pharmacia, Uppsala, Sweden). After desalting the VLDL + LDL was diluted to 1.5 ml with PBS, poured into the column and eluted with phosphate buffered saline (PBS). The first 0.8 ml was discarded and the next 1.6 ml was collected. EDTA-free fraction was diluted in oxygen-saturated PBS to a protein concentration of 0.1 mg/2 ml. The protein content of the fraction was determined by a sensitive pyrogallol reagent (Labport, Tampere, Finland). Oxidation was started by adding 33.5 ~1 of 0.1 mM of freshly made CuCl, solution (final concentration in the cuvette 1.68 ,uM) to 2 ml of diluted VLDL + LDL fraction. VLDL + LDL oxidation kinetics were monitored by the change in 234 nm absorbance at 37°C every 5 min for 4.6 h by a temperature controlled spectrophotometer with an enzyme kinetics data system (Hitachi U-2000, Tokyo, Japan).

85

124 (1996) 83-94

The phases of the oxidation process were defined on the basis of the changes in absorbance at 234 nm: the lag phase, the propagation phase and the decomposition phase. To describe the oxidizability of the VLDL + LDL fraction by CuCl, mediated oxidation we used four parameters from the reaction curve: lag time to maximal oxidation rate (lag time), maximum reaction velocity (V,,,), maximum absorbance (A,,,) and the time to maximal oxidation (Amaxtime)(Fig. 1). Lag time was defined as the time from the start of the reaction to the beginning of the steepestslope and was computed by means of least squares regression (LSR) equation. V,,, was computed also by using LSR method from the slope of the absorbance curve during the propagation phase. A,,, was defined as the absorbance from the decomposition state of curve when the slope decreasesbelow 10 - ’ abs/min. The fourth parameter was defined as the time needed to reach the maximal oxidation (maximum absorbance). The between-batch coefficient of variation (CV) was determined by pooling EDTA-plasma and keeping aliquots at - 80°C for different periods of time (maximum freezing time for plasma was 1 month). After VLDL + LDL-fraction separation, the oxidation susceptibility was measured immediately. The CV of a frozen plasma pool was 11.2% for the lag time, 10.9% for the V,,,, 7.8% for the

Time (min)

Fig. 1. C&l,-induced oxidation of VLDL + LDL fraction. Kinetic absorbance measurement at 234 nm. The vertical line in the left side of the curve indicates the start of the steepest slope. The lag time is the time from the start to this line. The vertical line in the right side of the curve indicates the maximum absorbance. Time to maximal oxidation is the time needed to reach maximum absorbance.

86

E. Porkkala-Sarataho

et al. / Atherosclerosis

A,,, and 10.3% for Amaxtime (n = 9). The withinbatch CV was 3.3% for the lag time, 2.9% for the V max,1.0% for the A,,, and 8.5% for Amaxtime(n = 6). 2.5. Determination of vitamin E Vitamin E was measured by using a modification of the HPLC method of Thurnham et al. [19,25]. Vitamin E was extracted from 200 ~1 plasma with 5 ml of hexane and 1 ml of ethane. After centrifugation (5 min at 3000 rpm), the top layer was separated and evaporated to dryness under nitrogen and the residue was dissolved in 200 ~1 of mobile phase (acetonitrile-methanolchloroform. 47 + 47 + 6 by vol). Samples were injected in a C,,-column (Pharmacia Super Pat Pep-S 5 pm, 4 x 250 mm) and detected at 294 nm. The HPLC system comprised a Kontron Instrument 420 pump (Milan, Italy), a Beckman 507 autosampler, a Beckman System Gold Diode Array Detector Module 168 and Beckman System Gold integration software (Beckman Instruments, San Ramon, CA, USA). 2.6. Other measurements Serum LDL cholesterol was precipitated using PVS (Polyvinyl sulphate, Boehringer Mannheim, Mannheim, Germany) and calculated as the difference between total and supernatant cholesterol. Serum total cholesterol and triglyceride concentrations were determined enzymatically (Kone Diagnostics, Espoo, Finland) with an autoanalyzer (Kone Specific, Kone Ltd., Espoo, Finland). Plasma phospholipids were assessedwith an enzymatic calorimetric method (Biomerieux, Marcyl’Etoile, France). Plasma ascorbic acid and dehydroascorbic acid (DHA) [26] as well as pcarotene [25] concentrations were determined by high-performance liquid chromatographic methods as described previously. All the measurements were done at the final (36-month) visit. Dietary intake of linoleic acid and use of vitamin supplements were measured at 12 month study visit by using a 12-month food frequency questionnaire, which was completed in an interview by a dietitian [27].

124 (1996) 83-94

2.7. Statistics The statistical analyses were carried out by using the SPSS statistical software. To separate the effect of vitamin E of those of serum lipids, lipid-standardised vitamin E concentration was used in the statistical analysis. The standardisation procedure has been described in detail earlier [28]. Multiple regression analysis was used to estimate the independent contributions of lipid standardised a-tocopherol concentration, plasma p-carotene, plasma total ascorbic acid (ascorbic acid + dehydroascorbic acid), serum LDL and triglycerides, plasma phopholipids, use of pravastatin, age, use of cigarettes, the month of the 36-month study visit and dietary intake of linoleic acid to the variation of the oxidation susceptibility and oxidation resistance of VLDL + LDL. Because the month of the 36-month study visit affected the oxidation susceptibility and oxidation resistance parameters, this effect was eliminated by entering the month of the 36-month study visit in all multiple regression analyses and covariance analyses as dummy variables. For analysis of variance and covariance, plasma vitamin E content and lipid standardised vitamin E were stratified into fifths. Serum concentrations of LDL and triglycerides, plasma concentration of total ascorbic acid, p-carotene, phospholipids, age, use of pravastatin, use of cigarettes and month of the 36-month study visit and dietary intake of linoleic acid were used as covariates. Serum concentrations of LDL cholesterol and triglycerides were not used as covariates when lipid standardised vitamin E was analysed. Twoway analysis of variance and covariance were used to analyse the modification of the association between lipid standardised vitamin E concentration with the oxidation susceptibility and oxidation resistance by pravastatin treatment and between smoking status. 3. Results

The main characteristics of the study subjects are presented in Table 1. One eligibility criterion at the entry of the study was mild to moderate

E. Porkkala-Sarataho

et al. 1 Atherosclerosis

124 (1996) 83694

87

Table 1 Lipid and antioxidant concentrations of plasma indices of oxidation susceptibility and oxidation resistance of VLDL + LDL in the study subjects (n = 406) at the final (36-month) study visit

Age (years) Serum cholesterol (mmolJ) Serum LDL-cholesterol (mmolil) Serum triglycerides (mmol,‘l) Plasma phospholipids (mmol/l) Plasma c-tocopherol (pmol!l) Plasma ascorbate (pmol/l) Plasma dehydroascorbic acid (pmol/l) Plasma B-carotene (/Lmol,‘l) Lag time (min) V mdX(mabs/min) A,,, @W AmaxtLme bin)

Mean

S.D.

Minimum

Maximum

61 5.87 4.20 1.64 4.20 34.1 52.3 3.6 0.35 86 10.2 0.61 143

4.4 1.08 I .0x 0.84 1.08 8.5 20.8 5.0 0.33 II 1.7 0.09 17

47.2 3.49 1.69 0.33 1.69 16.4 2.0 0.0 0.10 50 6.6 0.381 105

67.0 9.02 1.59 7.72 7.59 76.4 124.4 45.85 4.09 150 16.9 1.050 230

hypercholesterolemia (S-LDL-chol 2 4.0 mmol/l, S-chol < 7.5 mmol/l), and consequently the mean serum LDL cholesterol was higher than usually found among Finnish men. Also the mean plasma a-tocopherol level was reasonably high and had a wide range. The main plasma antioxidant which associated with indicators of oxidation susceptibility and oxidation resistance of VLDL + LDL was lipid standardised a-tocopherol (Table 2 and Table 3). The regression model also included other plasma antioxidants such as total ascorbic acid and pcarotene. However, neither plasma total ascorbic acid nor p-carotene had significant association with any of the parameters of oxidation susceptibility or oxidation resistance of VLDL + LDL that were measured. In the multivariate regression model a high plasma content of vitamin E or lipid standardised vitamin E concentration were the most important determinants of lagtime to maximal oxidation rate (standardised regression coefficient = 0.244, P < 0.0001 for vitamin E and 0.211, P < 0.0001 for lipid standardised vitamin El and Amaxtime(standardised regression coefficient = 0.194, P < 0.0001 lipid standardised vitamin E and 0.237, P < 0.0001 for vitamin E). The hypolipidemic medication (pravastatin) had an association with A,,, (standardised regression coefficient = -0.141, P < 0.0024) and V,,, (standardised regression coefficient = - 0.161, P

< 0.0004). As shown in Table 2, a high serum concentration of LDL cholesterol was significantly related to a decreased oxidation resistance, whereas, as seen in Tables 2 and 3 serum concentration of triglycerides had an inverse association with V,,, and a direct association with Amaxtime. These findings on serum concentration of triglycerides may reflect the protective effect of a-tocopherol, as triglycerides and a-tocopherol concentrations correlated strongly (0.446, P < 0.0001). In this study the VLDL + LDL fraction was used in the oxidation susceptibility measurements as VLDL is especially rich in both triglycerides and n-tocopherol. In the multivariate step-up regression model dietary intake of linoleic acid was the second most important determinant of the V,,, (standardised regression coefficient = 0.235, P < 0.0001) [27]. Fig. 2 shows the study subjects in fifths according to plasma concentration of a-tocopherol, and means of oxidation resistance are shown both with and without statistical adjustment for age, medication (pravastatin yes vs. no), smoking of cigarettes, the month of the 36- month study visit, dietary intake of linoleic acid, plasma concentration of total ascorbic acid, p-carotene and phospholipids, as well as serum concentrations of triglycerides and LDL cholesterol. The lag time to maximal oxidation rate was 10% (P < 0.0001) longer in men in the highest than in the lowest

88

E. Porkkala-Sarataho

et al. 1 Atherosclerosis

124 (1996) 83-94

Table 2 Determinants of the lag time and time to maximal oxidation (oxidation resistance) based on a multivariate regression model Regression coefficient

Lag time (min) Lipid standard&d a-tocopherol Plasma total ascorbic acid bmol/l) Plasma b-carotene (~mol/U Serum LDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Plasma phospholipids (mmol/l) Linolic acid intake (g/d) A maxtime(min) Lipid standardised a-tocopherol Plasma total ascorbic acid @mol/l) Plasma p-carotene (wW1) Serum LDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Plasma phospholipids (mmol/l) Linolic acid intake (g/d)

95% C.I.

Standardised regression Statistical coefficient significance

11.884

6.609,

17.159

0.211

0.0001

0.010

-0.035,

0.054

0.020

0.6614

3.074

0.161,

5.988

0.096

0.0387

- 1.093

-2.703,

0.516

-0.111

0.1825

0.846

-0.433

2.124

0.067

0.1941

3.196

- 1.180,

7.573

0.113

0.1519

-0.022

-0.218,

0.173

-0.010

0.8247

17.652

10.552,

24.751

0.194

0.0001

0.028

-0.032,

0.088

0.035

0.3552

4.070

0.148,

7.992

0.079

0.0420

-4.394

-6.561,

- 2.228

- 0.277

0.0001

11.289

9.568,

13.009

0.551

0.0001

5.455

- 0.435,

11.346

0.119

0.0694

0.066

-0.197.

0.329

0.019

0.6200

The multiple squared correlations for the models including also the month of the 36-month study visit, age, use of cigarettes and use of pravastatin (yes vs. no) were 0.227 (P i 0.0001) for lag time and 0.462 (P < 0.0001) for Arnaxtirne

fifth. When we used fifths of lipid standardised vitamin E concentration (Fig. 3), the lag time to maximal oxidation rate was 6% (95% C.I. 1.810.1%) longer in men in the highest than in the lowest fifth (P < 0.0001 for trend). The reason for using lipid standardised vitamin E was the strong association between plasma vitamin E and serum concentrations of triglycerides and LDL cholesterol. The correlation between vitamin E and triglycerides was 0.45 (P < 0.001) and the correlation between vitamin E

and LDL cholesterol was 0.53 (P < 0.001). Bcarotene and plasma total ascorbic acid did not correlate significantly either with LDL-cholesterol (0.10, P = 0.054 for D-carotene, 0.06, P = 0.231 for total ascorbic acid) or with triglycerides ( - 0.06, P = 0.259 for p-carotene and - 0.06, P = 0.201 for total ascorbic acid). Association of lipid standardised vitamin E levels with oxidation resistance and oxidation susceptibility did not differ significantly either between the pravastatin treated and placebo group

E. Porkkala-Sarataho

et al. 1 Atherosclerosis

89

124 (1996) 83-94

Table 3 Determinants of the maximum reaction velocity and maximum absorbance (oxidation susceptibility) based on a multivariate regression model Regression coefficient

95% C.I.

-0.210

-1.091,

0.003

Standardised regression coefficient

Statistical significance

0.671

- 0.023

0.6399

-0.004

0.010

0.037

0.4229

-0.015

-0.487,

0.457

- 0.003

0.9505

0.231

- 0.030

0.491

0.147

0.0828

-0.669

-0.876,

-0.462

-0.330

0.0001

- 0.556

- 1.264,

0.152

-0.123

0.1235

0.080

0.048,

0.112

0.235

0.0001

0.089

0.042,

0.136

0.187

0.0002

0.0001

-0.0003,

0.0005

0.027

0.5693

0.013

-0.012,

0.038

0.048

0.3174

0.008

-0.006,

0.022

0.096

0.2633

- 0.007

-0.018,

0.005

-0.062

0.2410

-0.031

-0.069,

0.007

-0.129

0.1115

0.004

0.002,

0.005

0.202

0.0001

V,, @Wiltin) Lipid standard&d a-tocopherol Plasma total ascorbic acid @mol/l) Plasma B-carotene (Pmm Serum LDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Plasma phospholipids (AWN) Dietary linoleic acid (g/d) A,,, W-6 Lipid standardised cc-tocopherol Plasma total ascorbic acid (,umol/l) Plasma b-carotene (vol/l) Serum LDL-cholesterol (mmol/l) Serum triglycerides (mmol/l) Plasma phospholipids (mmol/l) Dietary linoleic acid (g/d)

The multiple squared correlations for the models including also the month of the 36-month study visit, age, use of cigarettes and use of pravastatin (yes vs. no) were 0.207 (P i 0.0001) for V,,, and 0.174 for A,,, (P < 0.0001).

(for interaction, P = 0.610 for lag time, P = 0.280 for V,,,, P = 0.313 for &axtime and P = 0.659 for A,,,) or between smokers and nonsmokers (P = 0.943 for lag time and P = 0.922 for Anaxtime)except for A,,,,, (P = 0.009) and V,, (P = 0.012) for smoking status. After statistical adjustment for factors described above the A,, was 22.5% higher and V,,, was 16.7% faster in smoking men in the highest than in the lowest fifth of lipid standardised vitamin E concentration, whereas this difference was smaller in nonsmokers (for interaction, P = 0.009 for A,,, and P = 0.012 for V,,,).

4. Disarssion In this population-based randomised doublemasked study, subjects who had high plasma concentration of vitamin E had an increased oxidation resistance of atherogenic plasma lipoproteins. A number of trials have shown that supplementation of healthy subjects with pharmacological doses of vitamin E increases the resistance of LDL to oxidation [5,14- 181.Jialal and Grundy [14] showed that a-tocopherol supplementation (D,L-a-tocopherol, 800 IU = 727 mg/ day for 3 months) results in an increase in plasma

90

E. Porkkala-Sarataho

et al. / Atherosclerosis

and LDL a-tocopherol levels and in a decreased susceptibility of LDL to oxidation and DieberRotheneder et al. [16] showed that in humans the oxidation resistance of LDL can be increased by vitamin E supplementation (RRR-a-tocopherol, 150 IU = 101 mg, 225 IU = 151 mg, 800 IU = 537 mg or 1200 IU = 805 mg a day for 21 days). Jialal and Fuller [29] suggested in their dose response study that the minimum dose of a-tocopherol needed to significantly decrease the susceptibility of LDL to oxidation is 400 IV/ day = 364 mg whereas Princen et al. [30] proposed that supplementation with even 25 IU = 25 mg vitamin E per day leads to significant protection of LDL against oxidation in vitro. The latter study was a sequential study in which the authors continued to increase the level of vitamin E in the subjects from 25 mg/day to 800 mg/day for 12 weeks at 2 week intervals. What is new in our findings is that some of our

95 1

q n

Observed lag time Adjusted lag time

Alpha-tocopheral

124 (1996) 83-94

95

q n

Observed lag time Adjusted lag bme T

1

99

;

g -7

Lipid standard&d

T

27m

atpha-tocopherol

Fig. 3. Bar graph showing the mean lag time in fifths of lipid standardised cc-tocopherol concentration, unadjusted and adjusted for plasma B-carotene, plasma total ascorbic acid (ascorbic acid + dehydroascorbic acid), plasma phopholipids, use of pravastatin, age, use of cigarettes, the month of the 36-month study visit and dietary intake of linoleic acid.

TT

(wmoVI)

Fig. 2. Bar graph showing the mean lag time in fifths of plasma cc-tocopherol, unadjusted and adjusted for plasma p-carotene, plasma total ascorbic acid (ascorbic acid + dehydroascorbic acid), serum LDL and triglycerides, plasma phopholipids, use of pravastatin, age, use of cigarettes, the month of the 36month study visit and dietary intake of linoleic acid.

study subjects had an increased oxidation resistance of VLDL + LDL even though they did not receive any antioxidant supplementation. This observation is not in total agreement with other studies [16,22,30,31] which claim that the vitamin E content of LDL isolated from plasma of unsupplemented subjects would not allow one to predict the oxidation resistance of a particular LDL preparation. In the earlier non-experimental studies the study setting was different from ours. Our study was a population based study with over 400 men, aged 44-65 who were or had been hypercholesterolemic and had plasma levels of vitamin E in a wide range (Table 1). In the previous studies there were only few, relatively young and healthy study subjects: 12 clinically healthy subjects aged 20-30 years [ 161, 15 healthy subjects aged 20-59 years [22], 20 healthy volunteers aged 21-31 years [30] and an unspecified number of healthy subjects aged 20-30 years [31]. In these studies the investigators used LDL whereas we used VLDL + LDL.

E. Porkkala-Sarataho

et al. / Atherosclerosis

We measured x-tocopherol content from plasma whereas in other studies a-tocopherol from LDL was measured. We used VLDL + LDL instead of only LDL for three reasons. Firstly, the separation procedure for the combined fraction in large quantities was shorter than the traditional sequential methods and at the time of our study, the fast direct separation of LDL was not available. Secondly, even though based on a different oxidation assay, Mohr and Stocker have shown that radical mediated lipid oxidation proceeds via a similar mechanism in isolated LDL and VLDL and they are oxidized simultaneously [4]. They have also observed that the ratio of a-tocopherol to coenzyme Q10in VLDL is close to that of LDL [4]. Thirdly, as not only modified LDL but also modified IDL and VLDL are atherogenic [32], the combined fraction represents a more physiological lipoprotein fraction that also has relevance in atherogenesis. In our earlier supplementation study the correlation between a-tocopherol in plasma and in this fraction was 0.90 (P < 0.001) in 40 male subjects. Also the correlation between changes in plasma and VLDL + LDL a-tocopherol was very high [19]. In the present study we used a CuCl,-induced method to assess the oxidizability of atherogenic lipoprotein samples. We modified the original technique [24] to make it suitable for VLDL + LDL but did not change the nature of the assay. The method is accurate and reproducible but relatively slow if one wants to assess the time to reach maximal oxidation. The method of Esterbauer et al. [24] has been used in various studies on oxidation susceptibility and oxidation resistance of LDL [14-16,33,34] but using different parameters from ours to describe the oxidation susceptibility and oxidation resistance. Consequently the values reported in these studies are not directly comparable to ours. The predictive validity of the method has been established by its strong association with atherosclerosis [34]. The parameters which we have used to determine oxidation susceptibility seemedto be useful and informative. The lag time, V,,, and Amaxtime were unambiguous; lag time was prolonged, whereas V,,, was decreased and Amaxtimeprolonged in proportion to serum triglycerides. This

I24 (1996) 83-94

91

is to be expected because triglycerides mainly consist of fatty acids which are substrates for the conjugated diene formation in the copper induced oxidation. These findings about serum concentration of triglycerides may also reflect the protective effect of a-tocopherol, as triglycerides and cr-tocopherol concentrations correlate strongly. In this study the VLDL + LDL fraction was used in the oxidation susceptibility and oxidation resistance measurements and especially VLDL is rich in both triglycerides and a-tocopherol. A,,, behaved illogically; the more a-tocopherol there was in the sample, the higher maximum absorbance was attained. Because of the uncertainty of the factors which affect A,,, we have not used it as a parameter to assessthe oxidation susceptibility of VLDL + LDL. It has been established that HMGCoA-reductase inhibitors such as lovastatin, simvastatin and pravastatin have antioxidant effect and therefore they can decreasein vitro oxidizability of LDL by altering its composition [35-381. However, when LDL oxidation by continuous monitoring of in vitro with copper has been used, studies have shown that HMGCoA-reductase inhibitors affect only the maximal rate of diene production, total diene production [36] and the oxidation rate [38] and have no effect on lag time [36]. In the present study the result that pravastatin decreased the oxidation susceptibility (V,,, and A,,,) is in agreement with studies mentioned above. In our earlier report concerning the same population we noticed that subjects who received pravastatin had higher antioxidative capacity of LDL and reduced oxidation rate of VLDL + LDL in vitro but also longer oxidation lag when VLDL + LDL oxidation was measured by oxidation induced by hemin + H,O, [39]. Our finding that plasma p-carotene had no association with the oxidation resistance of VLDL + LDL is in accordance with the few supplementation trials, none of which has found any effect on the copper-induced oxidation resistance [40,41]. Our present results suggest that plasma vitamin C is not, in physiological concentrations, related to the oxidation resistance of the atherogenie lipoproteins in vitro. This is probably due to the hydrophilic character of ascorbic acid. be-

92

E. Porkkala-Sarataho

et al. 1 Atherosclerosis

cause of which ascorbic acid remains in plasma phase during VLDL + LDL separation and is missing from the oxidation environment. There are few trials concerning the effect of supplementation with vitamin C on oxidation resistance and their results are similar to ours. Rifici et al. [42] supplemented vitamin C and E together and alone and concluded that vitamin C supplementation (250 mg 4 times/day) alone caused only a small decrease in lipoprotein oxidation during incubation with 5 ,uM Cu for 4 h. Inhibition was not observed when incubated with a higher concentration of Cu or for a longer time. They also suggest that lipid soluble vitamin E remains associated with lipoproteins, whereas vitamin C would be removed during lipoprotein isolation. However, ascorbic acid has also proved to behave as a strong antioxidant with a sparing effect on a-tocopherol after in vitro addition to plasma or LDL [8,431. In conclusion, our data support the notion that a-tocopherol is an effective antioxidant in the protection of atherogenic lipoproteins against oxidation and that it is possible to estimate oxidation resistance of susceptibility and oxidation VLDL + LDL without oral vitamin E supplementation. In our study the subjects were hyperlipidemic and as most of rx-tocopherol in plasma is carried in VLDL and LDL, they also had relatively high levels of plasma vitamin E and a sufficient range to enable the study of its association with lipoprotein oxidizability. We still need more information concerning the effects of antioxidants and the oxidation susceptibility and oxidation resistance of atherogenic lipoproteins in healthy people. Acknowledgements We thank Kimmo Ronkainen for data analysis, Riitta Salonen and Juha Rummukainen for clinical examinations of the subjects and Anne Louheranta for interviews of 12-month food frequency questionnaire. The study was supported by grants from the Academy of Finland and the Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ.

124 (1996) 83-94

References [l] Steinberg, D., Parthasarathy, S., Carew, T., Khoo, J., Witztum, J. Beyond cholesterol: modifications of lowdensity lipoprotein that increase its atherogenicity. N Engl J Med 1989;320:915. [2] Steinbrecher, U., Zhang, H., Lougheed, M. Role of oxidatively modified LDL in atherosclerosis. Free Radic Biol Med 1990;9:155. [3] Salonen, J., Y11-Herttuala, S., Yamamoto, R., Butler, S., Korpela, H., Salonen, R., Nyyssiinen, K., Palinski, W., Witztum, J. Autoantibody against oxidised LDL and carotid progression of atherosclerosis. Lancet 1992;339:883. [4] Mohr, D., Stocker, R. Radical-mediated oxidation of isolated human very-low density lipoprotein. Arterioscler Thromb 1994;14:1186. [5] Brown, K., Morrice, P., Duthie, G. Vitamin E supplementation supresses indexes of lipid peroxidation and platelet counts in blood of smokers and non-smokers but plasma lipoprotein concentrations remain unchanged. Am J Clin Nutr 1994;60:383. [6] Halliwell, B., Gutteridge, J. (eds.) Free radicals in biology and medicine, 2nd edn., Oxford, Clarendon Press, 1989. [7] Esterbauer, H., Dieber-Rotheneder, M., Waeg, G., Striegl, G., Jiirgens, G. Biochemical, structural and functional properties of oxidized low-density lipoprotein. Chem Res Toxic01 1990;3:77. [8] Esterbauer, H., Gebicki, J., Puhl, H., Jiirgens, G. The role of lipid peroxidation and antioxidants in oxidative modification of LDL. Free Radic Biol Med 1992;13:341. [9] Kita, T., Nagano, Y., Yokode, M. Probucol prevents the progression of atherosclerosis in the WHHL rabbit, an animal model for familial hypercholesterolemia. Proc Nat1 Acad Sci USA 1987;84:5928. [lo] Carew, T., Schwenke, D., Steinberg, D. An atherogenic effect of probucol unrelated to its hypercholesterolemic effect: evidence that antioxidants in vivo can selectively inhibit low density lipoprotein degradation in macrophage-rich fatty streaks slowing down the progression of atherosclerosis in WHHL rabbits. Proc Nat1 Acad Sci USA 1987;84:7725. [l l] BjGrkheim, I., Henricksson-Freyshuss, A., Breuer, O., Diczfalusy, L., Berglund, L., Henrikson, P. The antioxidant butylated hydroxytoluene protects against atherosclerosis. Arterioscler Thromb 1991;ll: 15. [12] Buckley, M., Goa, K., Price, A., Brogden, E. Probucol: a reappraisal of its pharmacological properties and therapeutic use in hypercholesterolemia. Drugs 1989;37:761. [13] Hirose, M., Shibata, A., Hagiwara, K., Imaida, K., Ito, N. Chronic toxicity of butylated hydroxytoluene in Wistar rats. Food Cosmet Toxic01 1981;19:147. 1141Jialal, I., Grundy, S. Effect of dietary supplementation with a-tocopherol on the oxidation modification of low density lipoprtein. J Lipid Res 1992;33:899. [15] Jialal, I., Grundy, S. Effect of combined supplementation with a-tocopherol, ascorbate, and p-carotene on low-density lipoprotein oxidation. Circulation 1993;88:2780.

E. Porkkala-Sarataho

et al. I Atherosclerosis

[16] Dieber-Rotheneder, M., Puhl, H., Waeg, G., Striegel, G., Esterbauer, H. Effect of oral supplementation with D-atocopherol on the vitamin E content of human low density lipoproteins and resistance to oxidation. J Lipid Res 1991;32:1325. [17] Kleinveld, H., Demacker, P., Stalenhoef, A. Comparative study on the effect of low-dose vitamin E and probucol on the susceptibility of LDL to oxidation and the progression of atherosclerosis in Watanable heritable hyperlipidemic rabbits. Arterioscler Thromb 1994;14:1386. [18] Reaven, P., Witztum, J. Comparison of supplementation of RRR-sc-tocopherol and racemic cc-tocopherol in humans. Effects on lipid levels and lipoprotein susceptibility to oxidation. Arterioscler Thromb 1993;13:601. [19] Nyyssiinen, K., Porkkala, E., Salonen, R., Korpela, H., Salonen, J.T. Increase in oxidation resistance of atherogenie serum lipoproteins following antioxidant supplementation: a randomized double-blind placebo-controlled clinical trial. Eur J Clin Nutr 1994;48:633. [20] Frei, B., Gaziano, J. Content of antioxidant, performed lipid hydroperoxides, and cholesterol as predictors of the susceptibility of human LDL to metal ion-dependent and -independent oxidation. J Lipid Res 1993;34:2135. [21] Kleinveld, H.. Hak-Lemmers, H., Stalenhoef, A., Demacker, P. Improved measurement of low-density-lipoprotein susceptibility to copper-induced oxidation: application of a short procedure for isolating low-density lipoprotein. Chn Chem 1992;38:2066. [22] Babiy, A., Gebicki, J., Sullivan, D. Vitamin E content and low density lipoprotein oxidizability induced by free radicals. Atherosclerosis 1990;81:175. [23] Salonen, R., Nyyssiinen, K., Porkkala, E., Rummukainen, J., Belder, R., Park, J.S., Salonen, J.T. Kuopio atherosclerosis prevention study (KAPS): a populationbased primary preventive trial of the effect of LDL lowering on atherosclerotic progression in carotid and femoral arteries. Circulation 1995;92:1758. [24] Esterbauer, H., Striegl, G., Puhl, H., Rotheneder, M. Continuous monitoring of in vitro oxidation of human low density lipoprotein. Free Radic Res Commun 1989;6:67. [25] Thurnham, D., Smith, E., Flora, P. Concurrent liquidchromatographic assay of retinol, I-tocopherol, /?carotene, x-carotene, lycopene and /?-cryptoxanthin in plasma, with tocopherol acetate as internal standard. Clin Chem 1988;34:377. [26] Nyyssonen, K., Pikkarainen, S., Parviainen, M., Heinonen, K., Mononen, I. Quantitative estimation of dehydroascorbic acid and ascorbic acid by high-performance liquid qromatography-application to human milk, plasma and leucosytes. J Liq Chromatogr 1988;11:1717. [27] Louheranta, A., Porkkala-Sarataho, E., Nyyssonen, K., Salonen, R., Salonen, J.T. Linoleic acid intake and oxidation susceptibility of atherogenic lipoproteins in men. Am J Clin Nutr (in press). [28] Salonen, J.T., Nyyssiinen, K., Tuomainen, T-P., Mlenpaa, P.H., Korpela, H., Kaplan, G.A., Lynch, J., Helm-

124 (1996) 83-94

93

rich, S.P., Salonen, R. Increased risk of non-insulin dependent diabetes mellitus at low plasma vitamin E concentrations: a 4-year follow-up study in men. Br Med J 1995;311:1124. [29] Jialal, I., Fuller, C. The effect of a-tocopherol supplementation on LDL oxidation, a dose-response study. Arterioscler Thromb 1995;15:190. [30] Princen, H., van Duyvenvoorde, W., Buytenhek, R., van der Laarse, A., van Poppel, G., Gevers leuven, J., van Hinsbergh, W. Supplementation with low doses of vitamin E protects LDL from lipid peroxidation in men and women. Arterioscler Thromb 1995;15:325. [31] Esterbauer. H., Dieber-Rotheneder, M., Striegel, C., Waeg, G. Role of vitamin E in preventing the oxidation of low-density-lipoprotein, Am J Clin Nutr 1991;53:3148. [32] Krauss, R., Williams, P., Brensike, J., Detre, K., Lindgren, F., Kelsey, S., Vranizan, K., Leby, R. Intermediatedensity lipoproteins and progression of coronary artery disease in hypercholesterolemic men. Lancet 1987;62. [33] Gieseg, S., Esterbauer, H. Low density lipoprotein is saturable by pro-oxidant copper. FEBS Lett 1994;343:188. [34] Regnstrom. J., Nilsson, J., Tornvall, P., Landou, C., Hamsten, A. Susceptibility to low-density lipoprotein oxidation and coronary atherosclerosis in man. Lancet 1992;339:1183. [35] Aviram, M., Dankner, G., Cogan, U., Hochgraf, E., Brook, J.G. Lovastatin inhibits low-density oxidation and alters its fluidity and uptake by macrophages: in vitro and in vivo studies. Metabolism 1992;41:229. [36] Kleinveld, H.A., Demacker, P.N., De-Haan, A.F., Stalenhoef, A.F. Decreased in vitro oxidizability of low-density lipoprotein in hypercholesterolaemic patients treated with 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors. Eur J Clin Invest 1993;23:289. [37] Hoffman, R., Brook, G.J., Aviram, M. Hypolipidemic drugs reduce lipoprotein susceptibility to undergo lipid peroxidation: in vitro and ex vivo studies. Atherosclerosis 1992;93:105. [38] Bredie, S.J., de-Bruin, T.W., Demacker. P.N., Kastelein, J.J., Stalenhoef, A.F. Comparison of gemfibrozil vs. simvastatin in familial combined hyperlipidemia and effects on apolipoprotein-B-containing lipoproteins, low-density lipoprotein subfraction profile, and low-density lipoprotein oxidizability. Am J Cardiol 1995;75:348-353. [39] Salonen, R., Nyyssonen, K., Porkkala-Sarataho, E., Salonen, J.T. The Kuopio Atherosclerosis Prevention Study (KAPS): effect of pravastatin treatment on lipids, oxidation resistance of lipoproteins, and atherosclerotic progression Am J Cardiol 1995:76:34C. [40] Princen. H.M.G., van Poppel, G., Vogelezang. C., Buytenhek, R., Kok, F.J. Supplementation with vitamin E but not b-carotene in vivo protects low density lipoprotein from lipid peroxidation in vitro. Effect of cigarette smoking. Arterioscler Thromb 1992:12:554. [41] Reaven, P.D., Khouw, A., Beltz, W.F., Parthasarathy, S., Witztum. J.L. Effect of dietary antioxidant combinations

94

E. Porkkala-Sarataho

et al. 1 Atherosclerosis

in humans. Protection of LDL by vitamin E but not by B-carotene. Arterioscler Thromb 1993;13:590. [42] Rifici, V.A., Khachadurian, A.K. Dietary supplementation with vitamins C and E inhibits in vitro oxidation of lipoproteins. Am Co11Nutr 1993;12:631.

124 (1996) 83-94

[43] Jialal, I., Vega, G.L., Grundy, S.M. Physiologic levels of ascorbate inhibit the oxidative modification of low density lipoprotein. Atherosclerosis 1990;82:185.