Search

BACKGROUND: alpha-Tocopherol plays an important role in protecting LDL against oxidation. However, additional effects of alpha-tocopherol at the intracellular level may contribute to the clinical outcome of intervention studies. OBJECTIVE: We investigated whether alpha-tocopherol influences the inflammatory responses of immune cells in normolipidemic and hypertriglyceridemic subjects. DESIGN: RRR-alpha-Tocopherol was administered for 6 wk at a dose of 600 IU (402 mg)/d to 12 primary hypertriglyceridemic and 8 normolipidemic (fasting triacylglycerol >3.0 and <2.0 mmol/L, respectively) subjects. Cytokine production [tumor necrosis factor alpha (TNF-alpha), interleukin (IL)-1beta, and IL-8] by mononuclear cells and superoxide production by polymorphonuclear cells and in diluted whole blood were determined before and after the intervention. RESULTS: Cytokine and superoxide production did not differ significantly between hypertriglyceridemic and normolipidemic subjects. alpha-Tocopherol supplementation resulted in a 2- to 3-fold increase in the concentration of alpha-tocopherol in plasma and LDL. Whereas superoxide production in response to phorbol 12-myristate 13-acetate decreased in all subjects, response to oxidized LDL increased in 19 of 20 subjects. Response to opsonized zymosan before alpha-tocopherol supplementation was not significantly different from that after supplementation. Lipopolysaccharide-induced cytokine production by mononuclear cells decreased after supplementation with alpha-tocopherol. CONCLUSIONS: alpha-Tocopherol differentially influences inflammatory responses of immune cells. These effects of alpha-tocopherol may be relevant in chronic inflammatory processes such as atherogenesis. Copyright © 2000 American Society for Clinical Nutrition [ABSTRACT FROM AUTHOR]

Objective: High-density lipoprotein (HDL) antiatherogenic functions seem to be diminished during inflammatory conditions such as rheumatoid arthritis (RA). The aim of this study was to investigate the effects of tumour necrosis factor (TNF) inhibition on the antioxidative capacity of HDL in RA., Methods: Plasma lipids and paraoxonase (PON-1) activity were investigated in 45 RA patients, before and during 6 months of anti-TNF therapy. In addition, HDL was isolated and tested for its ability to inhibit copper-induced oxidation of low-density lipoprotein in vitro., Results: Plasma HDL concentrations did not change considerably after 6 months of therapy. However, stable increases of PON-1 activities were observed throughout the same period (p<0.03). The increases were more obvious when related to HDL or apolipoprotein AI concentrations. HDL total antioxidative capacity significantly improved 6 months after the initiation of anti-TNF therapy (p = 0.015). The initial improvement of PON-1 activity paralleled a decrease in the inflammatory status, whereas specific TNF blockade was likely to be responsible for the long-term effects., Conclusions: Anti-TNF therapy with infliximab has beneficial effects on lipids through changes in HDL antioxidative capacity, which might be clinically relevant and contribute to the reported protective effect of anti-TNF on cardiovascular morbidity in RA. This emphasises the importance of HDL antiatherogenic capacity for cardiovascular risk in chronic inflammatory conditions.

Adiponectin is secreted from adipocytes in different multimers, of which the high molecular weight (HMW) form is supposed to mediate favorable metabolic and anti-atherogenic effects. We determined adiponectin multimers in 29 female and 22 male patients with familial combined hyperlipidemia (FCH) and 51 age-, gender-, and BMI-matched controls in relation to cardiovascular disease (CVD). We observed a clear sexual dimorphism of total adiponectin and its multimers. Female, but not male, FCH patients had significant lower total adiponectin and both HMW and low molecular weight (LMW) adiponectin than controls. The adiponectin sensitivity index (ASI), reflected by HMW/total adiponectin, and the LMW/HMW adiponectin ratio did not differ significantly between FCH females and control females. However, FCH females with CVD exhibited significantly lower ASI (34.2+/-10.1% vs 46.0+/-7.1%) and higher LMW/HMW ratio (1.5+/-0.8 vs 0.7+/-0.3) compared to FCH females without CVD, reflecting a more atherogenic adiponectin multimer distribution.

Background: Excess body iron is associated with increased cardiovascular disease risk, possibly via non-transferrin-bound iron (NTBI)-mediated enhancement of inflammation and oxidation of low-density lipoprotein (LDL)., Methods: We assessed this proposed atherosclerotic mechanism of body iron by determining the relationship of levels of serum iron parameters, including NTBI, with plasma markers of inflammation and LDL oxidation in 232 subjects who visited the outpatient clinic for hemochromatosis family screening., Results: Plasma level of soluble intercellular adhesion molecule-1 (sICAM-1) was positively related to ferritin (standardized beta coefficient 0.16) and to NTBI (0.185) and negatively to total iron-binding capacity (TIBC, -0.166). Significant higher levels of sICAM-1 were found for subjects in the highest quartile of NTBI compared to the lowest quartile of NTBI (122 microg/L (107-141) and 106 microg/L (89-125), median (interquartile range), p<0.001). Odds ratio of subjects having sICAM-1 level above 134 microg/L (75th percentile) in the highest and lowest quartile of NTBI amounted 2.3. White blood cell count was positively related to ferritin (0.149). High-sensitivity C-reactive protein, interleukin-6, interleukin-8, oxidized LDL, oxidized LDL/apolipoprotein B and IgG and IgM antibodies to oxidized LDL were not related to any of the markers of iron status., Conclusion: Excess body iron, reflected by elevated serum ferritin and NTBI and decreased TIBC, is associated with increased plasma level of sICAM-1 but not with markers of in vivo LDL oxidation.

CXCL16 acts as a scavenger receptor for oxLDL in its membrane-bound form and induces migration of activated T cells in its soluble form. Due to these properties, CXCL16 has been suggested to play a role in both atherosclerosis and rheumatoid arthritis (RA). Our aim was to evaluate the contribution of soluble CXCL16 to the scavenging of oxLDL and its potential as a marker for cardiovascular disease (CVD) in patients with RA. We found that circulating CXCL16 was not correlated with plasma oxLDL or ApoB and was not related to the presence of CVD in RA patients. Moreover, CXCL16 did not bind and scavenge oxLDL in an in vitro setting. These data suggest that binding of oxLDL by soluble CXCL16 does not play a role in atherosclerosis and, although confirmation in larger studies is needed, that circulating CXCL16 is not related to the presence of CVD in patients with RA.

Unlabelled: Non-transferrin-bound iron (NTBI) is implicated in lipid peroxidation but the relation with oxidative modification of low-density lipoprotein (LDL) is not known. We assessed variables reflecting in vitro and in vivo LDL oxidation in two age- and sex-matched groups (n=23) of hereditary hemochromatosis heterozygotes (C282Y), characterized by a clear difference in mean serum NTBI (1.55+/-0.57 micromol/L vs 3.70+/-0.96 micromol/L). Plasma level of oxidized LDL (absolute and relative to plasma apolipoprotein B), and IgG and IgM antibodies to oxidized LDL, markers of in vivo LDL oxidation, did not differ between the groups with low and high serum NTBI. Mean lag-phase of in vitro LDL oxidation was also not significantly different between both study groups., Conclusion: these findings do not support the hypothesis that NTBI promotes oxidative modification of plasma LDL.

Objective: Circulating oxidized low-density lipoprotein (LDL) has been shown to be a useful marker for identifying patients with coronary heart disease (CHD) and persons at high cardiovascular risk. The effect of cholesterol-lowering therapy on plasma level of oxidized LDL is not clear., Methods and Results: We investigated effects of cholesterol lowering by therapeutic intervention (2 years) with atorvastatin (80 mg daily) and simvastatin (40 mg daily) on circulating oxidized LDL (absolute level and in proportion to plasma apolipoprotein B) in relation to atherosclerosis progression (carotid intima-media thickness, carotid IMT) and to inflammation (high-sensitivity C-reactive protein, hsCRP) in 115 stable patients with heterozygous familial hypercholesterolemia (FH). Atorvastatin and simvastatin reduced plasma-oxidized LDL (-43 and -35%, respectively) in proportion to the decrease in plasma apolipoprotein B. Neither absolute nor relative level of oxidized LDL correlated with carotid IMT or hsCRP at baseline. Also changes in levels of circulating oxidized LDL were not related to changes in carotid IMT and hsCRP., Conclusions: In familial hypercholesterolemia-oxidized LDL carried in plasma is strongly associated with apolipoprotein B but not with inflammation nor with carotid IMT, and statin treatment does not reduce oxidized LDL relative to apolipoprotein B.

Background: Markers of lipid peroxidation are commonly used to assess oxidative stress in preeclampsia. The aim of this study was to assess the concentration of oxidized low density lipoprotein (oxLDL), a novel marker for lipid peroxidation, and that of the thiobarbituric acid reactive substances (TBARS) in the pathogenesis of severe preeclampsia and to investigate the influence of gestational age on these parameters., Method: Plasma levels of oxLDL and TBARS were assayed in women with severe preeclampsia (n = 40), normotensive pregnant controls matched for gestational age (n = 24) and normotensive pregnant controls at full term (n = 16)., Results: Women with preeclampsia showed lower oxLDL levels (mean +/- SE) than matched controls (181 +/- 12 vs. 219 +/- 14; p = 0.027), whereas no differences were found for the TBARS concentration (3.8 +/- 0.6 vs. 3.7 +/- 0.4). When women with preeclampsia were compared to control women at full term, TBARS were elevated (3.8 +/- 0.6 vs. 1.5 +/- 0.2; p = 0.01). However, in women with normotensive pregnancy TBARS were also lower in full-term control pregnancy compared to early third-trimester values (p < 0.0001)., Conclusion: Plasma TBARS decreased during the third trimester of pregnancy, underlining the importance of matching for gestational age when studying markers of lipid peroxidation in pregnant women. Women with preeclampsia had lower plasma levels of oxLDL compared to gestational age-matched controls, indicating that oxLDL could be a marker for preeclampsia.

Methionine loading seems to be accompanied by increased oxidative stress and damage. However, it is not known how this oxidative stress is generated. We performed the present crossover study to further elucidate the effects of methionine loading on oxidative stress in the blood of healthy volunteers, and to examine possible preventative effects of N -acetylcysteine (NAC) administration. A total of 18 healthy subjects were given two oral methionine loads of 100 mg/kg body weight, 4 weeks apart, one without NAC (Met group), and one in combination with supplementation with 2x900 mg doses of NAC (Met+NAC group). Blood samples were collected before and 2, 4, 8 and 24 h after methionine loading for measurements of thiol levels, protein carbonyls, lipid peroxidation, cellular fibronectin and ferric reducing ability of plasma (FRAP; i.e. antioxidant capacity). After methionine loading, whole-blood levels of free and oxidized cysteine and homocysteine were increased in both groups. Furthermore, the total plasma levels of homocysteine were higher, whereas those of cysteine were lower, after methionine loading in both groups. Lower levels of oxidized homocysteine and a higher free/oxidized ratio were found in the Met+NAC group compared with the Met group. Although the antioxidant capacity decreased after methionine loading, no major changes over time were found for protein carbonyls or cellular fibronectin in either group. Our results suggest that methionine loading may initiate the generation of reactive oxygen species by the (auto)-oxidation of homocysteine. In addition, supplementation with NAC seems to be able to partially prevent excessive increases in the levels of homocysteine in plasma and of oxidized homocysteine in whole blood, and might thereby contribute to the prevention of oxidative stress.

We investigated whether long-term alpha-tocopherol therapy in chronic smoking affects superoxide generating capacity of neutrophils ex vivo. To this purpose, we randomly assigned 128 male chronic smokers (37 +/- 21 pack years of smoking) to treatment with placebo (n = 64) or alpha-tocopherol (400 IU dL-a-tocopherol daily, n = 64). After two years of therapy, we measured phorbol 12-myristate 13-acetate-induced superoxide production of isolated neutrophils and of diluted whole blood by monitoring reduction of ferricytochrome c and luminol-enhanced peroxidase-catalyzed chemiluminescence. Plasma lipids and lipoproteins were not different between the two treatment groups. As expected, concentrations of alpha-tocopherol in plasma and in low-density lipoproteins were markedly elevated in the supplemented group compared to the placebo group (+ 120%, P < 0.0001 and + 83%, P < 0.0001, respectively). Consequently, resistance to in vitro oxidation of low-density lipoproteins (reflected by lag time of conjugated diene formation) was higher in the supplemented group than in the placebo group (+ 22%, P < 0.0001). Superoxide generating capacity of neutrophils and superoxide production in diluted whole blood did not differ between alpha-tocopherol and placebo group. It is concluded that in chronic smoking long-term supranormal alpha-tocopherol intake does not reduce neutrophil superoxide-anion generating capacity, despite large increases in the concentrations of alpha-tocopherol in plasma and in low-density lipoproteins.

Antioxidants have been postulated to exert beneficial effects in atherosclerosis. Atherosclerosis is associated with raised plasma levels of soluble intercellular adhesion molecule-1 (sICAM-1) and autoantibodies against oxidized low-density lipoprotein (oxLDL). It is not known whether antioxidants affect these plasma factors in chronic smokers. In a randomized double-blind placebo-controlled study involving 128 male normolipidemic chronic smokers the effect of a 2-year alpha-tocopherol treatment (400 IU dL-alpha-tocopherol daily) on plasma levels of sICAM-1 and autoantibodies against oxLDL was evaluated. In addition, we monitored production of superoxide by leukocytes ex vivo. It was found that compared to nonsmokers (n = 33) plasma levels of IgG but not IgM autoantibodies against oxLDL and concentrations of sICAM-1 in smokers were significantly elevated (30 and 42%, respectively). After supplementation with alpha-tocopherol concentration of TBARS in plasma and in vitro oxidizability of LDL had decreased, but autoantibodies and sICAM-1 had not changed. Production of superoxide was not different between alpha-tocopherol- and placebo-treated smokers. It is concluded that in chronic smokers, long-term treatment with alpha-tocopherol does not normalize the raised levels of sICAM-1 and autoantibodies against oxLDL, both risk factors for initiation or progression of cardiovascular disease, despite a decrease in in vitro oxidizability of LDL.

Polymorphonuclear leukocytes (PMN) have been suggested to play a role in atherosclerosis, but intracellular signaling after stimulation with oxidized low-density lipoprotein (LDL) is unknown. We investigated mechanistic aspects of oxidized LDL-induced superoxide production by human PMN, with special emphasis on intracellular Ca(2+) concentration ([Ca(2+)](i)). Oxidized LDL, but not native LDL, evoked an early but sustained increase in [Ca(2+)](i) and a delayed production of superoxide. The increase in [Ca(2+)](i) could be reduced by fucoidan and completely prevented by U73122, suggesting involvement of the scavenger receptor and coupling to the phospholipase C signal transduction pathway. Furthermore, we provide evidence that the increase in [Ca(2+)](i) partly results from protein kinase C-dependent Ca(2+) influx. The relevance of this Ca(2+) entry for oxidized LDL-stimulated effects is illustrated by the finding that superoxide production was markedly reduced in the absence of external Ca(2+). Finally, inhibition of phagocytosis by cytochalasin B abolished oxidized LDL-stimulated superoxide production without affecting, however, the Ca(2+) mobilization. These effects of oxidized LDL on [Ca(2+)](i) and on respiratory burst of PMN may underlie the occurrence of elevated levels of [Ca(2+)](i) of resting PMN in hypercholesterolemia and represent a mechanism by which PMN can amplify processes in the early phase of atherosclerosis.

Familial combined hyperlipidemia (FCH) is characterized by a familial occurrence of a multiple-type hyperlipidemia, associated with coronary risk. The latter may be related to increased levels of small, dense LDL particles that have been found to be more prone to oxidative modification. We isolated total LDL as fresh as possible from 12 normolipidemic relatives with a buoyant LDL subfraction profile (group 1), 7 normolipidemic subjects with a dense LDL subfraction profile (group 2), and 16 hyperlipidemic FCH subjects with a dense LDL subfraction profile (group 3). In these nonobese and normotensive men, we studied the resistance of total LDL against Cu(2+)-oxidation in vitro. In addition, we analyzed the alpha-tocopherol and the coenzyme Q10 contents of LDL and determined their relation to LDL oxidizability. LDL isolated from group 3 subjects was more susceptible to oxidative modification than LDL from group 1 subjects (lag time: 60.4 +/- 8.1 versus 70.4 +/- 11.4 minutes; P < .05). For the combined groups, the ratio of ubiquinol-10 to polyunsaturated fatty acids in LDL, together with the basal amount of dienes in LDL, were good predictors of the rate of LDL oxidation (R2 = .73, P = .0001). In groups 2 and 3, the redox status of coenzyme Q10 (ubiquinol-10/ubiquinone-10) and the ratio of ubiquinol-10 to alpha-tocopherol in LDL were reduced compared with group 1 (P < .05). The K-value a measure of the LDL density, correlated with the the redox status (r = .37, P < .05). We conclude that in subjects with FCH total LDL is more prone to oxidation, due to the predominance of dense LDL particles. In addition, the decreased redox status of coenzyme Q10 in LDL from subjects with a dense LDL subfraction profile suggests that the LDL in the circulation has already undergone some oxidation.

The pathobiochemical mechanism of arteriosclerosis in hyperhomocysteinaemia has not yet been elucidated. In vitro studies have shown that the cytotoxic properties of homocysteine can be ascribed to its generation of reactive oxygen species. We studied lipid peroxidation, both in vivo and in vitro, in 10 homozygous cystathionine synthase-deficient (CSD) patients and in a control group of 10 healthy subjects of comparable age and sex. The susceptibility of low-density lipoprotein (LDL) from hyperhomocysteinaemic patients to oxidation was determined in vitro by continuously measuring the conjugated diene production induced by incubation with copper ions. Oxidation resistance (expressed as lag time), maximal oxidation rate, and extent of oxidation (expressed as total diene production) of LDL from CSD patients were not significantly different from those of LDL from controls. Furthermore, the time needed to reach maximal diene production, i.e. t(max), was similar for LDL from patients and controls. In addition, the vitamin E concentrations in LDL of CSD patients and controls were similar. The mean concentration (+/- SD) of plasma thiobarbituric acid reactive substances (TBARS), an indicator of in vivo lipid peroxidation, was 2.2 +/- 0.7 mumol L-1 in CSD patients, a lower value than that measured in the matched controls (5.0 +/- 2.0 mumol L-1). Investigation of in vivo and in vitro parameters of lipid peroxidation shows that the increased risk of arteriosclerosis in hyperhomocysteinaemia is unlikely to be due to increased lipid peroxidation.

We investigated the effect of different interventions on aortic atherosclerosis in Watanabe rabbits. Four groups of rabbits were fed either an oleic acid-enriched diet (80% of total fat intake) with or without vitamin E supplementation (250 IU/kg) or a diet enriched in linoleic acid with or without vitamin E supplementation for 6 months. At the start of the study, plasma cholesterol concentration was 21.4 +/- 3.6 mmol/L (n = 32). The diets did not influence the mean plasma lipids and lipoprotein concentrations except for HDL cholesterol, which was increased more on the oleic acid-enriched diets than on the linoleic acid-enriched diets. Vitamin E levels in plasma and LDL were increased on the oleic acid diet and reduced on the linoleic acid diet. On the latter diet, supplementation of vitamin E was quantitatively less effective in raising plasma or LDL vitamin E levels. The susceptibility of LDL to oxidation was determined in vitro. Both oleic acid-enriched diets increased the lag time by 140% from baseline. The linoleic acid diet supplemented with vitamin E increased lag time by 59%. Linoleic acid alone, however, decreased the lag time by 30%. Similar but inverse effects were seen on LDL oxidation rate. Thus, intervention protected LDL to oxidation in the following order: oleic acid plus vitamin E > oleic acid > linoleic acid plus vitamin E > linoleic acid. Despite the differences in LDL oxidizability induced by the four experimental diets, assessment of aortic atherosclerosis at the end of the 6-month dietary study period revealed no differences among the four study groups.(ABSTRACT TRUNCATED AT 250 WORDS)

Low-density-lipoprotein (LDL) oxidation may provide the crucial link between plasma LDL and atherosclerotic-lesion formation. Oxidation can be induced in vitro by incubating LDL with cells or metal ions and can be measured by continuously monitoring conjugated-diene absorbance at 234 nm. Measurement of LDL oxidizability was improved by performing the assay with 0.05 g of LDL-protein per liter of phosphate buffer containing 1 mumol of EDTA, by initiating oxidation by adding CuCl2 (5 mumol/L) at 30 degrees C, and by using a short-run ultracentrifugation method for isolating LDL, which reduced the time needed for obtaining purified LDL and thus reduced in vitro oxidation. LDL apolipoprotein analysis and oxidizability determination showed that this method is better than the longer sequential-isolation procedure. Adding butylated hydroxytoluene (BHT) to plasma as an antioxidant unpredictably increased the LDL oxidation lag time, making BHT unsuitable as an antioxidant. Adding EDTA appeared to be sufficient to prevent in vitro oxidation. Additionally, the diene production correlated highly with the concentration of thiobarbituric acid-reactive substances (r = 0.97). No relation between the vitamin E content of LDL and the oxidation lag time was found.

Oxidative modification of low density lipoprotein (LDL) has been implicated as a factor in the generation of macrophage-derived foam cells, the hallmark of atherosclerotic plaques. Because LDL consists of discrete subfractions with different physicochemical characteristics, the question arises as to whether these LDL subfractions differ in their susceptibility to oxidative modification. To answer this question, three LDL subfractions, LDL1, LDL2, and LDL3, were isolated from the plasmas of 11 healthy volunteers by density gradient ultracentrifugation. The LDL subfractions were oxidatively modified by incubation with copper ions. Differences in the subfractions' susceptibilities to lipid peroxidation were studied by measuring the formation of the 234-nm-absorbing oxidation products every 3 minutes on an ultraviolet spectrophotometer. A significant inverse linear relation was found between LDL subfractions and lag time (regression coefficient = -8.50, p less than 0.001), indicating that both the dense LDL3 and the light LDL2 were less well protected against oxidative modification than the very light LDL1. The LDL subfractions showed a positive linear relation with the rate of oxidation (regression coefficient = 0.46, p less than 0.001) and the amount of conjugated dienes formed in the LDL subfractions after 4 hours of oxidation (regression coefficient = 9.24, p less than 0.001). Thus, both LDL3 and LDL2 were more extensively modified in time than LDL1, which may be explained by the significantly higher concentration of polyunsaturated fatty acids in LDL3 (micromoles per gram LDL cholesterol) compared with LDL1 (Tukey's test, p less than 0.05). These results indicate that the more dense LDL subfractions, that is, LDL2 and LDL3, are more susceptible to oxidative modification and therefore may contribute more to foam cell formation than the less dense LDL subfraction LDL1.

Two low-density lipoprotein subfractions, LDL-1 and LDL-2, with density ranges of respectively 1.023-1.034 and 1.036-1.041 g/ml, were isolated by aspiration after density gradient ultracentrifugation of human pooled serum. In vitro interactions of both LDL subfractions with the LDL receptor of human cultured fibroblasts, human hepatoma cell line Hep G2 and human hepatocytes were compared. No difference in association (binding and internalization) nor in degradation between LDL-1 and LDL-2 by these cells was found. However, kinetic studies in guinea pigs showed that LDL-2 disappeared faster from the circulation and accumulated to a greater extent in the liver, compared to LDL-1. Thus, we were unable to show a difference in the LDL receptor-mediated uptake of both LDL subfractions by various cells in vitro. The results obtained in vivo suggest that LDL-1 is more atherogenic than LDL-2, because its longer half-life renders the particle more susceptible to uptake by the scavenger LDL receptor on macrophages.

A single spin density gradient ultracentrifugation method in a swinging bucket rotor has been applied for the detection and isolation of low density lipoprotein (LDL) subfractions. The visualization of the LDL heterogeneity was facilitated by prestaining the serum with Coomassie Brilliant Blue R prior to density gradient ultracentrifugation for 19.5 hr. A total of 13 human serum pools was analyzed. In each pool, two LDL subfractions, a lighter LDL1 subfraction, occasionally showing a subdivision into two bands, LDL1A and LDL1B, and a heavier LDL2 could be clearly distinguished by the banding pattern in the density gradient. Physicochemical characteristics of the isolated LDL subfractions were determined. The simple method for detection and isolation of these subfractions presented here may facilitate future studies on LDL heterogeneity.

Previously we found that in the dual precipitation method for determination of HDL2- and HDL3-cholesterol the optimal concentration of dextran sulfate was 0.87 g/l instead of 1.3 or 1.5 g/l. In the present study, we performed the dual precipitation method both in serum and in EDTA-plasma samples in order to investigate whether the method is sensitive for plasma-serum differences. It was found, that in the analysis of EDTA-plasma samples a twice higher final concentration of dextran sulfate was needed. Addition of the dextran sulfate resulted in the appearance of fibrin structures in the HDL-fraction which may have diminished the effective dextran sulfate concentration in the precipitation reaction. Although the use of plasma is advised in the analysis of lipoproteins, the use of EDTA-plasma in precipitation methods for determination of HDL-cholesterol and its subfractions cannot be recommended.

We compared the dual-precipitation method for measurement of cholesterol in high-density lipoprotein subfractions HDL2 and HDL3 (Gidez et al., J Lipid Res 1982;23:1206-33) with density-gradient ultracentrifugation in a swinging-bucket rotor (Demacker et al., Clin Chem 1983;29:656-63). The concentration of dextran sulfate 15,000 (DS) needed for optimal accuracy of the HDL2-chol and HDL3-chol values was established empirically. At a DS concentration of 0.87 g/L, the values for HDL2-chol as well as for HDL3-chol in 88 sera did not differ significantly from those obtained by ultracentrifugation. The precision of the method was satisfactory and was related to the concentration. Nevertheless, the dual-precipitation method lacks specificity inasmuch as it produces no fractions that contain only one HDL subfraction. HDL2 and HDL3 each contained an equivalent amount of cholesterol from the other. At increasing DS concentrations, some radiolabeled HDL3 appeared to have precipitated prior to complete precipitation of HDL2. This lack of specificity can be tolerated in large-scale epidemiological studies for screening, but not in small-scale intervention studies or in assay of clinical samples, where better accuracy is needed and ultracentrifugation is preferred.

Feline serum lipoproteins were fractionated into four distinct classes by density gradient ultracentrifugation and characterized with respect to physical and chemical properties. The distribution of serum lipids, lipoproteins and apolipoproteins was quite unlike that in man, the cat having five times as much high density lipoproteins (HDL) as low density lipoproteins (LDL). The lipoproteins in the d less than 1.019 g/ml fraction of cats were larger and were richer in triglycerides than their human counterparts and contained a considerable amount of beta-migrating particles. The low density lipoproteins of cats and man had similar chemical composition, but cat LDL had a higher negative charge, were smaller and contained apoprotein A-I. Cat HDL consisted of two distinct subfractions HDL2 and HDL3 with similar density boundaries and particle size as in man. In cat serum and HDL fraction apoprotein A-II was a minor component. Like human serum, fasting cat serum contained only the larger species of apoprotein B, apo B-100, whereas intestinal lymph contained exclusively the smaller apo B-48. Post heparin feline and human plasma possessed both lipoprotein lipase and hepatic lipase. Chylomicrons formed after a fat load in cats were removed from the circulation as rapidly as in man. It is concluded, that the cat is another animal model of potential interest for the study of lipoprotein metabolism.