Gert M. Kostner

Lipoprotein Lp(a) and the risk for myocardial infarction

Abstract The serum lipoprotein Lp(a) concentration was measured in 76 male post-myocardial infarction (MI) patients aged between 40 and 60 years, and in 107 control subjects of the same age and sex. Quantitation was performed by the Laurell technique. It was sensitive in the range 1–60 mg/dl with a day to day C. V. of less than 4%. A considerable variation of Lp(a) concentration was noticed in the whole population with a frequency distribution of higher order. Conventional statistical methods could therefore not be applied to evaluate a possible MI risk of Lp(a). In addition to Lp(a), several other risk and anti-risk factors for atherosclerosis were assayed. The whole population was divided into normolipemics (NL) and Type IIa, IIb and IV phenotypes. In addition, the subjects were grouped into two or three Lp(a)-types, selecting several different cut-off points for Lp(a) concentrations. The results can be summarized as follows: (1) NL-controls had significantly lower total cholesterol and low density lipoprotein (LDL)-cholesterol but higher high density lipoprotein (HDL)-cholesterol values compared to the MI-patients. The HDL-cholesterol concentration was also significantly different between Type IIa controls and MI-patients. (2) Eleven per cent of the NL-controls but 25% of the NL-MI-patients exhibited Lp(a) values exceeding 50 mg/dl. Thus Lp(a) concentration above this value represents a 2.3-fold relative risk for MI. Similar findings were obtained in the Type IIa and IIb populations but not in Type IV hyperlipemics. (3) Taking 30 mg/dl as the cut-off point, Lp(a) represents a relative risk of 1.75 for MI in the NL population. (4) Hyperlipemics in general (controls + MI) exhibited higher Lp(a) values compared to NL. (5) Statistical evaluation of all the data failed to reveal any correlation between Lp(a) levels and other risk or anti-risk factors for atherosclerosis. (6) It is concluded that Lp(a) represents an independent addi tional risk factor for MI with a possible threshold value of approx. 30 mg/dl in NL.

HMG CoA reductase inhibitors lower LDL cholesterol without reducing Lp(a) levels.

Lp(a) is a plasma lipoprotein particle consisting of a plasminogenlike protein [apo(a)] disulfide bonded to the apo B moiety of low-density lipoprotein (LDL). Increased plasma levels of Lp(a), either independently or interactively with LDL levels, have been shown to be a risk factor for atherosclerosis. Recently, a new class of lipid-lowering drugs, HMG CoA reductase inhibitors, have been introduced. These drugs act by decreasing liver cholesterol synthesis resulting in up-regulation of LDL receptors, increased clearance of LDL from plasma, and diminution of plasma LDL levels. In this study, we examined the effect of HMG CoA reductase inhibitors on Lp(a) levels in three groups of subjects, five volunteers and two groups of five and 14 patients. In all 24 subjects, mean decreases were observed in total cholesterol (43 +/- 5%), total triglyceride (35 +/- 8%), very low-density lipoprotein (45 +/- 9%), and LDL cholesterol (43 +/- 5%). The mean change in high-density lipoprotein cholesterol was an increase of 7 +/- 8%. Despite the very significant decrease in LDL cholesterol levels (p less than 0.001), Lp(a) levels increased by 33 +/- 12% (p less than 0.005). This was not associated with a measurable change in the chemical composition or size of the Lp(a) particle. This emphatically suggests that Lp(a) particles, despite consisting principally of LDL, are cleared from plasma differently than LDL. The surprising finding of an increase in Lp(a) levels suggests this class of drugs may have a direct effect on Lp(a) synthesis or clearance independent of its effect on LDL receptors.

Influence of n-3 fatty acids on the growth of human breast cancer cellsin vitro: Relationship to peroxides and Vitamin-E

Epidemiological studies suggest a causal relationship of dietary polyunsaturated fatty acids (PUFAs) with the morbidity and mortality from breast cancer. In order to reveal possible underlying mechanisms of these findings, we studied the influence of n-3 and n-6 PUFAs in comparison to oleic acid on the proliferation of well characterized estrogen dependent (MCF-7, ZR-75, T-47-D) and estrogen independent (MDA-MB-231, HBL-100) breast cancer cells in culture. The cell growth inhibitory effect was related to the formation of lipid peroxidation products. Normal human skin fibroblasts served as a control. In fibroblasts, the addition of 20 µg/ml of exogenous fatty acids either had no effect or caused an insignificant increase of proliferation. Similar results were obtained with MCF-7 cells. In all other breast cancer cell types, n-3 long-chain PUFAs, eicosapentaenoic and docosahexaenoic acids, were the most effective fatty acids in arresting the cell growth. Alpha-linolenic and gamma-linolenic acid exerted a variable effect on cell proliferation depending on the cell line investigated. Oleic acid significantly stimulated the proliferation of hormone-independent breast cancer cells while it had no effect on the proliferation of hormone-dependent cells. Viability studies by trypan blue excretion indicated that the arrest in cell growth was not due to major cytotoxic effects.The addition of PUFAs to breast cancer cells caused a significant increase in the formation of conjugated dienes and lipid hydroperoxides in the cellular lipids; their content was significantly correlated with the capacity of arresting cell growth. In contrast, the addition of PUFAs to fibroblasts did not increase lipid hydroperoxide formation. The addition of Vitamin E to cancer cells at a concentration of 10 µM to the PUFA-supplemented medium almost completely restored cell growth.Our data indicate that PUFAs significantly interfere with cell proliferation of breast cancer cellsin vitro due to the formation of oxidation products. In addition to that, there must be other factors involved, most probably related to the differential metabolism of PUFAs in tumor cells. Our findings may have some impact on treatment and prevention of breast cancer.

Studies on the role of specific cell surface receptors in the removal of lipoprotein (a) in man.

The binding of 125I-lipoprotein (a) [Lp(a)] to cell surface receptors was studied on cultured human fibroblasts. The results were compared with corresponding data obtained with 125I-low density lipoproteins (LDL). Equilibrium binding studies showed that Lp(a) is bound with high affinity by the cell surface receptors. The maximum binding capacity for Lp(a) was 37% lower than for LDL. For Lp(a) and LDL, the Scatchard plots displayed linearity, indicating a single category of binding sites. Half-maximal saturation occurred at a concentration of 9.52 +/- 1.04 nM for Lp(a) and 7.76 +/- 1.29 nM for LDL. Competition binding experiments revealed that Lp(a) and LDL are nearly equally potent in competing each other for the binding sites. Binding of Lp(a) and LDL were followed by suppression of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity. Cyclohexanedione treatment of Lp(a) and LDL completely abolished receptor binding. Neither Lp(a) nor LDL were specifically bound by fibroblasts obtained from a patient with homozygous familial hypercholesterolemia (FH). The removal mechanisms for Lp(a) and LDL were further compared by in vivo studies. Radioiodinated Lp(a) and LDL were injected intravenously into 12 normolipemic individuals to measure kinetic parameters of these two lipoproteins simultaneously in each subject. Mean fractional catabolic rate (FCR) of Lp(a) was 0.260 +/- 0.060 and mean FCR of LDL was 0.377 +/- 0.077 (mean +/- SD). In each subject, FCR of Lp(a) was lower than the FCR of LDL; the mean difference was 31%. The absolute synthetic rate of Lp(a) was significantly lower than the corresponding value of LDL. In each individual, the percentage of total Lp(a) that was contained in the intravascular space was higher than the corresponding value of LDL; the mean difference was 19%. A highly significant positive correlation was found between FCR of LDL and FCR of Lp(a) (r = 0.853, P less than 0.01). No relationship was found between the serum concentration of LDL-apolipoprotein B and Lp(a). The serum level of Lp(a) was positively related to the absolute rate of Lp(a) synthesis (r = 0.979, P less than 0.01). The serum level of LDL-apolipoprotein B was inversely related to FCR of LDL (r = 0.613, P less than 0.05). In a patient with homozygous FH, FCR of LDL was 0.205 and FCR of Lp(a) was 0.210. The results of these studies show that Lp(a) is specifically bound with high affinity to the same receptors of human fibroblasts as LDL. The affinity and maximum binding capacity are slightly lower for Lp(a) than for LDL. The results of the turnover studies are consistent with the assumption that Lp(a) is removed from the plasma by similar mechanisms as LDL.

High D-Glucose–Induced Changes in Endothelial Ca2+/EDRF Signaling are Due to Generation of Superoxide Anions

Pretreatment of porcine aortic endothelial cells with high D-glucose results in enhanced endothelium-derived relaxing factor (EDRF) formation (39%) due to increased endothelial Ca2+ release (57%) and Ca2+ entry (97%) to bradykinin. This study was designed to investigate the intracellular mechanisms by which high D-glucose affects endothelial Ca2+/EDRF response. The aldose-reductase inhibitors, sorbinil and zopolrestat, failed to diminish high D-glucose-mediated alterations in Ca2+/EDRF response, suggesting that aldose-reductase does not contribute to high D-glucose-initiated changes in Ca2+/EDRF signaling. Pretreatment of cells with the nonmetabolizing D-glucose analog, 3-O-methylglucopyranose (3-OMG), mimicked the effect of high D-glucose on Ca2+ release (41%) and Ca2+ entry (114%) to bradykinin, associated with elevated EDRF formation (26%). High D-glucose and 3-OMG increased superoxide anion (O2−) formation (133 and 293%, respectively), which was insensitive to inhibitors of cyclooxygenase (5,8,11,14-eicosatetraynoic acid [ETYA], indomethacin), lipoxygenase (ETYA, gossypol, nordihydroguaiaretic acid [NDGA]), cytochrome P450 (NDGA, econazole, miconazole), and nitric oxide (NO) synthase (L-omega N-nitroarginine), while it was diminished by desferal, a metal chelator. The gamma-glutamyl-cysteine-synthase inhibitor, buthioninesulfoximine (BSO), also increased formation of O2− by 365% and mimicked the effect of high D-glucose on Ca2+/EDRF signaling. The effects of high D-glucose, 3-OMG, and BSO were abolished by co-incubation with superoxide dismutase. Like high D-glucose, pretreatment with the O2−-generating system, xanthine oxidase/hypoxanthine, elevated bradykinin-stimulated Ca2+ release (+10%), Ca2+ entry (+75%), and EDRF (+73%). We suggest that prolonged exposure to pathologically high D-glucose concentration results in enhanced formation of O2−, possibly due to metal-mediated oxidation of D-glucose within the cells. This overshoot of O2− enhances agonist-stimulated Ca2+/EDRF signaling via a yet unknown mechanism.

Oxidation of lipoprotein Lp(a). A comparison with low-density lipoproteins.

Aimed at identifying possible mechanisms of the suggested high atherogenicity of Lp(a), its susceptibility for Cu(II)-induced oxidation was studied and compared with that of LDL. Since the content of antioxidants as well as the fatty acid pattern of a lipoprotein greatly affects its oxidizability, Lp(a) and LDL were characterized first with respect to these substances. Paired samples of low-density lipoproteins (LDL) and Lp(a) were isolated from seven individual donors and compared with each other. This study showed that LDL and Lp(a) are very similar with respect to their fatty acid and antioxidant composition. LDL contains approx. 1132 nmol of total fatty acids/mg lipoprotein and LDL 1466 nmol total fatty acids/mg lipoprotein. Analysis of the fatty acid composition of individual lipid classes (cholesteryl esters, phospholipids and triacylglycerols) revealed also a high similarity in the composition of these lipid classes between the two lipoproteins. A comparison of the antioxidant composition showed that Lp(a) contains less alpha-tocopherol than LDL (1.6 +/- 0.35 nmol/mg vs. 2.1 +/- 0.25 nmol/mg LDL). In copper(II)-induced lipid peroxidation experiments we found a striking difference in the susceptibility of individual lipoprotein classes between all donors. In addition, Lp(a) exhibited a 1.2 to 2.4 longer lag-phase than the corresponding LDL preparation from the same blood donor. Treatment of Lp(a) with neuraminidase resulted in a drastic decrease of the lag-phase of Lp(a). Neuraminidase treatment of LDL on the other hand had no significant effects on its susceptibility to oxidation. Supplementation of neuraminidase-treated Lp(a) with N-acetylneuraminic acid (NANA) at concentrations comparable to the naturally occurring amounts of NANA in the Lp(a) protein moiety led to an increase of the lag-phase yielding values which were comparable to those observed with native Lp(a). These results demonstrate that the fatty acid composition as well as the antioxidant concentrations of Lp(a) and LDL are quite similar; despite this fact, Cu2(+)-mediated oxidation of Lp(a) is retarded in comparison to LDL which might be due to the higher content of NANA in Lp(a).

Paraoxonase PON1 Polymorphism Leu-Met54 Is Associated With Carotid Atherosclerosis Results of the Austrian Stroke Prevention Study

BACKGROUND AND PURPOSE Genetic polymorphism at the paraoxonase locus is associated with serum concentration and activity of paraoxonase and with increased risk for coronary heart disease. Two frequent polymorphisms present at the paraoxonase gene are the methionine (M allele) leucine (L allele) interchange at position 54 and the arginine (B allele) glutamine (A allele) interchange at position 191. This is the first study to determine the effect of these polymorphisms on carotid atherosclerosis. METHODS The paraoxonase genotypes at positions 54 and 191 of 316 randomly selected individuals aged 44 to 75 years were determined by polymerase chain reaction-based restriction enzyme digestion. Carotid atherosclerosis was assessed by color-coded Duplex scanning and was graded on a 5-point scale ranging from 0 (normal) to 5 (complete luminal obstruction). RESULTS The LL, LM, and MM genotypes at position 54 were noted in 137 (43.4%), 132 (41.8%), and 47 (14.9%) subjects; the AA, AB, and BB genotypes at position 191 occurred in 172 (54.4%), 124 (39.2%), and 20 (6.3%) individuals. The LL genotype was significantly associated with the presence and severity of carotid disease (P=0.022), whereas the 191 polymorphism had no effect. Logistic regression analysis with age and sex forced into the model demonstrated plasma fibrinogen (odds ratio [OR], 1.005 per mg/dL), LDL cholesterol (OR, 1.01 per mg/dL), cardiac disease (OR, 1.75), and the paraoxonase LL genotype to be significant predictors of carotid atherosclerosis. The ORs for the associations with age and sex were 1.09 (P=0.0003) and 1.66 (P=0.052) per year. CONCLUSIONS These data suggest that the paraoxonase LL genotype may represent a genetic risk factor for carotid atherosclerosis.

Apolipoprotein E Polymorphism and Silent Microangiopathy-Related Cerebral Damage Results of the Austrian Stroke Prevention Study

BACKGROUND AND PURPOSE Microangiopathy-related cerebral damage (MARCD) includes white matter abnormalities and lacunar infarctions and represents a common MRI observation in subjects above 50 years of age. The risk factors of such brain abnormalities are not fully determined. The goal of this study was to determine whether the genetic heterogeneity of apolipoprotein E (apoE) contributes to the occurrence of MARCD. METHODS Brain MRI (1.5 T) was performed in 280 individuals (ages 50 to 75 years) without neuropsychiatric disease randomly selected from the official register of residents of the city of Graz, Austria. All study participants underwent apoE genotyping, carotid Doppler sonography, electrocardiography, echocardiography, and a complete blood chemistry panel. MARCD was defined as evidence of early confluent and confluent white matter hyperintensities or lacunes. Carotid atherosclerosis was graded on a five-point scale ranging from not present (0) to complete occlusion (5). RESULTS MARCD occurred in 61 individuals (21%). The distribution of apoE genotypes differed significantly between subjects with and without MARCD (P = .036). Subjects with such findings more commonly had the epsilon 2/epsilon 3 genotype (24.6% versus 10%) at similar frequencies of genotypes containing the epsilon 4 allele. The epsilon 2/epsilon 3 genotype was associated with lower levels of total cholesterol (P = .0009), LDL cholesterol (P = .00001), and apolipoprotein B (P = .00001). Also, there was a nonsignificant trend toward less cardiac disease. Other major vascular risk factors and carotid abnormalities were similar among the various genotypes. Multiple logistic regression analysis created a model of significant MARCD predictors, including age (odds ratio [OR], 1.1 per year), hypertension (OR, 3.4), and the apoE epsilon 2/epsilon 3 genotype (OR, 3.0). CONCLUSIONS These data suggest an association between the apoE epsilon 2/epsilon 3 genotype and MARCD despite favorable effects on the lipid profile and cardiac disease.

Evaluation of a new HDL2/HDL3 quantitation method based on precipitation with polyethylene glycol

The inverse correlation between plasma concentrations of high density lipoproteins (HDL) or constituents thereof with the incidence for atherosclerosis and vascular diseases is documented in many investigations [l-3]. Increasing interest currently focuses towards the differential evaluation of HDL subfractions since it has been found that HDL3 remains relatively constant, whereas HDL2 largely accounts for variations in total HDL concentrations [4]. For routine determination in the clinical laboratory, fast, simple and automated methods are required. This has been hampered by the fact that reliable measurements of HDL subclasses have only been possible by analytical ultracentrifugation. These instruments are expensive and unsuitable for processing large numbers of samples. Several methods for differential HDL quantitation have been described, e.g. ultracentrifugation in swinging bucket rotors [5], the combination of precipitation with preparative ultracentrifugation [6], gradient gel electrophoresis [7] and others. Currently, only one method is known for HDL2/HDL3 determination based solely on precipitation [8]. This method uses as the first step, heparin-Mn2+ for separation of apolipoprotein B (ApoB) containing lipoproteins, a procedure which is only infrequently used today because of the interference with enzymatic reagents for cholesterol or phospholipid determination. To overcome this and other drawbacks, a new method has been developed based on differential precipitation with polyethylene glycol.

Apolipoprotein E4allele in the normal elderly: neuropsychologic and brain MRI correlates

The presence of the apolipoprotein E e4 allele has been considered to be a risk factor for Alzheimers disease and vascular dementia. We therefore used demanding neuropsychologic testing and brain MRI to determine if elderly normals with at least one e4 allele demonstrate subclinical changes in cognition and a higher frequency of brain atrophy or silent ischemic brain damage. The study population consisted of 214 randomly selected individuals aged 50 to 75 years without neuropsychiatric or general disease. There were 175 (81.8%) subjects without and 39 (18.2%) with at least one e4 allele. The two groups were comparable for age, length of education, verbal intelligence, mood and major vascular risk factors. Apolipoprotein E e4 carriers performed significantly worse than non‐carriers when assessed for learning and memory abilities, while there were no differences in test results of conceptualization, attention, speed of mental processing and visuopractical skills. There were no between‐group differences for thromboembolic and lacunar infarcts, white matter hyperintensity grading and the semiautomatically measured white matter hyperintensity area. The extent of sulcal and ventricular widening as well as hippocampal and parahippocampal volumes were also similar between the comparative subsets. We conclude that the apolipoprotein E e4 allele is associated with subtle learning and memory deficits in normal elderly persons and may therefore be suggested a marker for accelerated cognitive aging. In this group of subjects it was not associated with brain parenchymal changes as demonstrated by MRI.