Inhibition of cholesterol crystal formation by apolipoproteins in supersaturated model bile.

Apolipoproteins A-1 and A-2 were purified from human plasma. At concentrations present in human bile these proteins prolonged the nucleation time of cholesterol monohydrate crystals when added to model systems of supersaturated bile. In contrast, apolipoprotein C-3 and other serum proteins did not have this effect. Also, when human gallbladder bile was fractionated by gel filtration chromatography, apolipoproteins A-1 and A-2 were among the proteins present in a fraction of bile enriched in potent inhibitors of cholesterol crystal nucleation. These findings suggest that apolipoproteins A-1 and A-2 in supersaturated human gallbladder bile could inhibit the rate of formation of solid cholesterol crystals and thus help to prevent spontaneous cholesterol gallstone formation in humans.

Antioxidant nature of bovine milk beta-lactoglobulin.

Heating is necessary for processing milk in the dairy industry, which evidently produces a conformational change in beta-lactoglobulin (beta-LG). beta-Lactoglobulin, a major protein that accounts for approximately 10 to 15% of total milk proteins, is a globular protein consisting of 162 AA with a relative molecular mass of 18.4 kDa. The purpose of the present study was to determine the antioxidant role of beta-LG in milk and the possible mechanism involved. We showed that beta-LG is a mild antioxidant whose potency is less than that of vitamin E and probucol (the latter being an antioxidant used for clinical therapy). The conversion of the beta-LG monomer to dimer was responsible, in part, for the mode of action in protecting low-density lipoproteins against copper-induced oxidation. Cross-linking the free thiol groups of beta-LG by heating (100 degrees C for 2 min), or chemically modifying the beta-LG by carboxymethylation to block the thiol groups resulted in a substantial loss of antioxidant activity. The data suggest that Cys-121 plays an essential role in the antioxidant nature of beta-LG. By using an anti-LG antibody affinity column to deplete the beta-LG from milk, we observed from the lost antioxidant activity that beta-LG contributes approximately 50% of the total activity. Because beta-LG is extremely sensitive to thermal denaturation, to maintain its antioxidant nature, dairy products consumed daily should not be overheated in order to maintain its antioxidant nature.

Ion-pair reversed-phase HPLC: assay validation of sodium tanshinone IIA sulfonate in mouse plasma.

Sodium tanshinone IIA sulfonate (STS), a hydrophilic ionic substance, is used as a cardiovascular drug. An ion-pair reversed-phase high-performance liquid chromatography (IP-RP-HPLC) method for the determination of STS in mouse plasma was initially developed. The assay involved a rapid and simple extraction process and subsequent detection at 271 nm. The retention time for STS was 7.5 min. Based on extracted STS standard mouse plasma at 1.5,10 and 50 microg/ml, the assay precision were 2.7, 2.1 and 1.7% with a mean accuracy of 96.7, 98.5 and 99.4%, respectively. At plasma concentration of 1.5, 50 and 75 microg/ml, the mean recovery of STS were 93.1, 96.3 and 97.5%. The limit of detection (LOD) and limit of quantification (LOQ) for STS was 0.1 microg/ml and 0.5 microg/ml, respectively. Linear responses were observed over a wide concentration range (0.5-100 microg/ml) for STS in mouse plasma. STS can be detected after intravenous administration. This method was performed for the first time in pharmacokinetic studies of STS in the mouse.

A novel conformation-dependent monoclonal antibody specific to the native structure of beta-lactoglobulin and its application.

Molten globules are thought to be general intermediates in protein folding and unfolding. beta-lactoglobulin (beta-LG) is one of the major bovine whey proteins, constituting approximately 10 to 15% of total milk proteins. We have recently identified beta-LG as a superior marker for evaluating thermally processed milk. Strand D of beta-LG participates in irreversible thermal unfolding as probed by a monoclonal antibody (mAb) specific to thermally denatured beta-LG. In the present study, we used native beta-LG as an immunogen to test the hypothesis that a specific mAb against the native beta-LG could be established. As result, a mAb (4H11E8) directed against the native structure of beta-LG was made. The antibody did not recognize the heat-denatured form of beta-LG, such as its dimer and aggregates. Immunoassay using this "native" mAb showed that the stability of beta-LG was at temperatures < or =70 degrees C. beta-Lactoglobulin began to deteriorate between 70 and 80 degrees C over time. The denaturation was correlated with the transition temperature of beta-LG. Further chemical modification of Cys (carboxymethylation) or positively charged residues (acetylation) of beta-LG totally abolished its immunoreactivity, confirming the conformation-dependent nature of this mAb. Using competitive ELISA, the 4H11E8 mAb could determine the native beta-LG content in commercially processed milks. Concentrations of native beta-LG varied significantly among the local brands tested. From a technological standpoint, the mAb prepared in this study is relevant to the design and operation of appropriate processes for thermal sanitation of milk and of other dairy products.

Tween-20 increases the immunoreactivity of apolipoprotein A-I in plasma.

Tween-20 (polyoxyethylene 20), a trademark for a sorbitan polyoxyalkalene derivative is used as an emulsifier and detergent. In present studies, we report that Tween-20 is able to reduce the nonspecific binding of radiolabeled apolipoprotein A-I to test tubes. It also enhances the specific binding of 125I-labeled apolipoprotein A-I to anti-apolipoprotein A-I antibodies. Maximal enhancement was achieved at a Tween-20 concentration greater than 0.08%. The results were confirmed by using two different antisera raised in rabbits and goats. Since the immunoreactivity of apolipoprotein A-I in plasma cannot be fully detected by a conventional radioimmunoassay procedure, we have, therefore, studied the effect of Tween-20 on the immunoreactivity of apolipoprotein A-I in plasma. As determined by inhibition radioimmunoassay, we found that Tween-20 drastically increased the immunoreactivity of apolipoprotein A-I using both rabbit and goat antisera. The maximal reactivity was achieved at a Tween-20 concentration of 0.32%. However, Tween-20 did not enhance the immunoreactivity of delipidated plasma. Thus, the results indicate that the plasma lipids may hamper apolipoprotein A-I immunoreactivity and suggest that the nonionic detergent, Tween-20, somehow interacted with HDL lipids or the apolipoproteins and then facilitated the reaction of antigenic reactive site(s) of apolipoprotein A-I with their antibodies.

Physical, chemical, and immunochemical studies of apolipoprotein A-I from pigeon plasma high density lipoproteins.

Pigeon plasma high density lipoproteins (HDL) were isolated by ultracentrifugation between the densities of 1.063 and 1.21 g/ml. Gel filtration of delipidated HDL in 5 M guanidine-HCl on Sephadex G-150 yielded a major fraction which eluted at the same position as human apolipoprotein A-I isolated from HDL. In SDS-gel electrophoresis, the isolated apolipoprotein co-migrated with human apolipoprotein A-I with a molecular weight of approx. 28 000. The amino acid composition was similar to the apolipoprotein A-I isolated from human and hen plasma. The isolated apolipoprotein from pigeon plasma had therefore been designated as apolipoprotein A-I. As judged by circular dichroism (CD), the apolipoprotein A-I displayed a maximum mean residue ellipticity of approx. -3 000 at 222 nm while at concentrations greater than 0.2 mg/ml. Calculations of alpha-helicial content gave values of 85%. Lowering the concentration of apolipoprotein A-I was found to concomitantly decrease the ellipticity (absolute value) suggesting that there was some conformational change when the apolipoprotein A-I concentration varied. The isolated pigeon apolipoprotein A-I was found bound to the phospholipid (dimyristoyl phosphatidylcholine) and there was no significant conformational change upon lipid binding as judged by CD. Under the same experimental conditions, human apolipoprotein A-I exhibited a drastic conformational change by increasing its helicity in the presence of phospholipid. The helical content of human apolipoprotein A-I was increased from 48 to 85%. This finding suggests that the apolipoprotein may not necessarily increase its helical content during lipid binding. Moreover, immunochemical studies showed that rabbit antiserum prepared against pigeon apolipoprotein A-I could partially react with human apolipoprotein A-I determined by quantitative radioimmunoassay.

Design, synthesis and antithrombin activity for conformationally restricted analogs of peptide anticoagulants based on the C-terminal region of the leech peptide, hirudin.

Synthetic peptides cyclized via disulfide linkages have been synthesized as conformationally restricted analogs of a novel class of antithrombotic peptides that inhibit fibrinogen cleavage by binding to a non-enzymatic site on thrombin. Several conformational models for these inhibitors have been considered and cyclic analogs were synthesized to test their validity. Compounds designed on an alpha-helical model yielded several cyclic analogs that retained antithrombin activity. [D-Cys58, Cys61]-hirudin54-65, 5, and [D-Cys60, Cys63]-hirudin54-65, 6, had IC50 values of 26 and 30 microM, respectively, in an in vitro clot assay compared with a value of 3.7 microM for the linear hirudin54-65.

Immunochemistry of human plasma apolipoprotein C-II as studied by radioimmunoassay.

The immunoreactivity of human plasma apolipoprotein C-II was investigated using a specific radioimmunoassay. In whole plasma, the mean value quantitated was 2.21 +/- 0.415 mg/dl, while in delipidated plasma, a mean value of 3.84 +/- 1.186 mg/dl was obtained, suggesting that the antigenic sites of the apolipoprotein were not fully detected in unmodified plasma by our antibody preparation. Two detergents, Tween-20 and Triton X-100, were studied to determine if they could enhance the immunoreactivity of apolipoprotein C-II in whole plasma. At concentrations of 0.012-0.06%, Tween-20 markedly increased the immunoreactivity of whole plasma, but not of delipidated plasma, indicating that antigenic sites of plasma apolipoprotein C-II has been exposed by Tween-20. In contrast, Triton X-100 had no effect on the immunoreactivity of whole plasma apolipoprotein C-II. A radioimmunoassay conducted in the presence of 0.06% Tween-20, resulted in a mean value in whole plasma (3.39 +/- 1.11 mg/dl) that was not significantly different from that obtained when the assay was done on delipidated samples. The immunoreactivity of VLDL apolipoprotein C-II was also drastically enhanced following lipolysis by bovine milk lipoprotein lipase, supporting the hypothesis that antigenic sites are masked by the lipids. Finally, the mechanism responsible for the effect of Tween-20 on apolipoprotein C-II immunoreactivity was investigated. The results obtained from circular dichroism and ultracentrifugation suggest that the detergent may dissociate the apolipoprotein from lipoprotein particles, thus fully exposing the antigenic sites for reaction with antibodies.

Immunochemical properties of human low density lipoproteins as explored by monoclonal antibodies. Binding characteristics distinct from those of conventional serum antibodies.

Spleen cells obtained from mice immunized with human plasma low-density lipoproteins (LDL) were fused with mouse myeloma cells. The resulting hybridoma cells secreting immunoglobulin specific for LDL were screened and scored by radioimmunoassay and cloned by multiple limiting dilutions. Immunochemical properties of the monoclonal antibodies were compared with convential mouse serum antibodies. It was found that conventional antibodies precipitated LDL and bound more than 95% of 125I-labeled LDL and the maximal binding was independent of temperature. The monoclonal antibodies were incapable of precipitating LDL and bound a maximum of only 20% of the total 125I-labeled LDL. The maximal binding between monoclonal antibodies and LDL was extremely temperature-dependent. An optimal degree of binding was observed at 4 degrees C, whereas binding at 37 degrees C was only 30% of that achieved at 4 degrees C. Although the binding at 37 degrees C was low, the maximal binding could be re-established following a subsequent incubation at 4 degrees C, suggesting that the antigenic structure of LDL is reversibly modulated at temperatures between 4 and 37 degrees C. Since the orientation of apolipoprotein B in LDL is known to be dynamic at different temperatures, this result suggests that monoclonal antibodies, but not conventional antibodies, are capable of detecting subtle conformational changes in LDL. In addition, we have determined the binding affinity of LDL to monoclonal antibodies and to conventional antibodies. Only monoclonal antibodies showed a linear Scatchard plot, suggesting that the binding was to a single site with a single affinity. The monoclonal antibodies also possessed high specificity and failed to react with porcine LDL, while serum antibodies could recognize both human and porcine LDL.

Complete nucleotide sequence of pig (Sus scrofa) mitochondrial genome and dating evolutionary divergence within Artiodactyla.

The complete nucleotide sequence of the pig (Sus scrofa) mitochondrial genome, containing 16613bp, is presented in this report. The genome is not a specific length because of the presence of the variable numbers of tandem repeats, 5'-CGTGCGTACA in the displacement loop (D-loop). Genes responsible for 12S and 16S rRNAs, 22 tRNAs, and 13 protein-coding regions are found. The genome carries very few intergenic nucleotides with several instances of overlap between protein-coding or tRNA genes, except in the D-loop region. For evaluating the possible evolutionary relationships between Artiodactyla and Cetacea, the nucleotide substitutions and amino acid sequences of 13 protein-coding genes were aligned by pairwise comparisons of the pig, cow, and fin whale. By comparing these sequences, we suggest that there is a closer relationship between the pig and cow than that between either of these species and fin whale. In addition, the accumulation of transversions and gaps in pig 12S and 16S rRNA genes was compared with that in other eutherian species, including cow, fin whale, human, horse, and harbor seal. The results also reveal a close phylogenetic relationship between pig and cow, as compared to fin whale and others. Thus, according to the sequence differences of mitochondrial rRNA genes in eutherian species, the evolutionary separation of pig and cow occurred about 53-60 million years ago.

Molecular cloning and characterization of porcine cDNA encoding a 90-kDa heat shock protein and its expression following hyperthermia.

We have isolated and sequenced cDNA clones encoding a 90-kDa heat shock protein (HSP90) from a porcine brain cDNA library. The sequence of the 2202-nucleotide coding region showed 88.6% homology with that of the human homologue. Moreover, the deduced amino acid sequence of the porcine hsp90 cDNA was 99.7% identical to that of the human counterpart, with a difference of only three amino acids in a total of 733 residues. Expression of the gene was greatly increased in cultured cells during recovery from heat shock treatment at 45 degrees C for 60 min. Three major transcripts 2.2, 3.0, and 4.1kb in size were detected by Northern blot hybridization. These transcripts were further identified in a whole-pig hyperthermia experiment. These three hsp90 transcripts were constitutively expressed in porcine tissues including kidney, liver, brain, and heart, and their levels were markedly enhanced during recovery from 30-min hyperthermia treatment at 43 degrees C. Furthermore, we found that HSP90 was preferentially expressed in pituitary gland, brain, adrenal gland, and testis, in comparison to the other tissues.

Lipoprotein(a) enhances plasma clot lysis in vitro.

Human plasma lipoprotein(a) (Lp(a)) is a macromolecular complex that contains a protein homologous to plasminogen, the precursor of plasmin. We confirmed recent reports that Lp(a) is not activated by streptokinase or tissue plasminogen activator (t-PA) to yield plasmin-like activity. In testing the hypothesis that Lp(a) can competitively inhibit plasma clot lysis mediated by plasmin, the present study shows that Lp(a) significantly enhanced plasma clot lysis mediated by streptokinase or t-PA. The enhancement, however, was not observed in a plasmin-mediated clot lysis using a purified fibrinogen system. The addition of alpha-2-antiplasmin (alpha 2-plasmin inhibitor) to this system inhibited fibrinolysis; however, no inhibition was observed in the presence of Lp(a). One potential explanation for the Lp(a)-enhanced plasma clot lysis is that Lp(a) neutralizes the activity of alpha 2-antiplasmin present in plasma, thereby restoring the activity of plasmin to lyse the clot.

Antithrombin properties of C-terminus of hirudin using synthetic unsulfated N alpha-acetyl-hirudin45-65.

Unsulfated N alpha-acetyl-hirudin45-65 (MDL 27 589), which corresponds to the C-terminus of hirudin1-65, was synthesized by solid-phase methods. The synthetic peptide was able to inhibit fibrin formation and the release of fibrinopeptide A from fibrinogen by thrombin. The catalytic site of thrombin was not perturbed by the synthetic peptide as H-D-Phe-Pip-Arg-pNA hydrolysis (amidase activity) was not affected. The binding of synthetic peptide and thrombin was assessed by isolation of the complex on gel-filtration chromatography. A single binding site with a binding affinity (Ka) of approx. 1.0 X 10(5) M-1 was observed for thrombin-hirudin45-65 interaction. The data suggest that the C-terminal residues 45-65 of hirudin contain a binding domain which recognizes thrombin and yet does not bind to the catalytic site of the enzyme.

The C-terminal binding domain of hirullin P18. Antithrombin activity and comparison to hirudin peptides.

Hirullin P18 is a 61-amino acid hirudin-related protein having potent antithrombin activity. Similar to hirudin, it contains a highly acidic C-terminus, but has a significantly different sequence from any other known hirudin variant. The present study demonstrates that the C-terminal fragment acetyl-hirullin P18(41-62) [corrected] possesses an antithrombin potency similar to that of acetyl-desulfatohirudin(45-65). Additionally, like the hirudin fragment analog, it inhibits fibrin-clot formation by binding to a non-catalytic site on thrombin. Sequential shortening of the hirullin P18 C-terminal fragment demonstrates the critical nature of Phe51, which corresponds to the important Phe56 residue of hirudin. Although the sequences of hirullin P18(54-61) and hirudin(59-65) have substantial differences, the C-terminal functional domain represented by hirullin P18(50-61) appears to be comparable to hirudin(55-65) in terms of its functional role in antithrombin activity.

A protective role for nitric oxide in the oxidative modification of low density lipoproteins by mouse macrophages.

Low density lipoproteins (LDL) oxidatively modified by macrophages have been shown to be atherogenic in ex vivo studies. We studied the potential role of nitric oxide (NO), a free radical produced by macrophages, in LDL modification. Human LDL (1 mg/ml) were incubated with mouse peritoneal macrophages in Ham's F-10 medium. The cells were then stimulated by interferon-gamma and tumor necrosis factor-alpha to increase their production of NO from 1.3 to 12.2 microM in 24 h, as measured by nitrite. Lipid peroxidation of LDL, as measured by thiobarbituric acid-reactive materials (TBARS), was reduced in stimulated cells in a time-dependent manner. At 24 h, the decrease was about 27%. In the presence of an NO synthase inhibitor (NG-aminophomoarginine), the generation of NO was diminished and the protection against LDL lipid peroxidation was reversed. The extent of LDL protein modification was also assessed by examining its electrophoretic mobility. It was found that macrophage NO reduced the change in LDL electromobility. These data indicate that the production of NO may inhibit the oxidative modification of LDL with cytokine-stimulated macrophages. We suggest that NO plays a protective role in limiting macrophage-induced LDL modification.

Changes of plasma levels of apolipoproteins A-I, A-II, and B and their isoforms in patients with intestinal failure receiving long-term parenteral nutrition.

We have studied the plasma lipid and apolipoprotein profiles of 19 patients with intestinal failure who are receiving long-term total parenteral nutrition (TPN). These patients had significantly reduced levels of total and HDL cholesterol and normal levels of triglycerides. Radioimmunoassay determination of apolipoproteins showed a 30% and 50% reduction in apo A-I and A-II levels, respectively. Apolipoprotein B was normal in all but three patients. Isoelectric focusing showed two major isoforms of apo A-I in patients as compared with four isoforms observed in normal subjects and one major isoform of apo A-II compared with multiple isoforms. Recent epidemiologic studies indicate that an increased apo B:apo A-I ratio may be an important factor in atherogenesis. We suggest that patients with small-bowel syndrome who are currently on TPN may be at greater risk for atherosclerosis. Since TPN has restored a reasonably normal life expectancy for these patients, long-term follow-up will likely provide answers.

Postprandial exchange of apolipoprotein C-III between plasma lipoproteins.

Three healthy male subjects were given a high fat meal after fasting for 12 h. Blood samples were drawn at hourly intervals over 6 h. The plasma triglyceride levels reached peak values within 3 to 6 h postprandially. Plasma cholesterol concentration, however, remained constant in the three subjects as well as in a fasting normal subject. Apolipoprotein C-III (apoC-III), a major apoprotein in triglyceride-rich lipoprotein particles and known to be exchangeable between plasma high density lipoproteins (HDL) and triglyceride-rich particles, was studied in regard to its net transfer between lipoproteins during meal absorption and postabsorptive lipolysis. ApoC-III levels in total plasma as quantitated by radioimmunoassay were stable regardless of the increase of plasma triglycerides. When triglyceride levels increased after a meal, apoC-III in the d. less than 1.063 lipoprotein fraction increased concomitantly, while apoC-III in HDL decreased. The converse was observed during lipolysis as plasma triglycerides fell. In any case, apoC-III levels in d. less than 1.063 lipoproteins were positively correlated with plasma triglycerides (r = 0.82, p less than 0.01) after the meal. The finding suggests that the apoC-III concentrations in HDL are very dynamic in vivo. ApoC-III can transfer from HDL to triglyceride-rich particles during meal absorption and can transfer from triglyceride-rich particles to HDL during postabsorptive lipolysis.

Plasma lipids and apolipoprotein A-I and A-II levels in alcoholic patients.

This prospective study compared total plasma lipids, high-density lipoprotein cholesterol (HDL-C), and apolipoproteins A-I (apo A-I) and A-II (apo A-II) in 72 alcoholic patients and in 285 nonalcoholic controls. The HDL-C in the alcoholic group was not significantly different from that in the nonalcoholic controls. Alcoholic men had significantly higher levels of cholesterol and triglycerides and lower levels of apo A-I when compared with nonalcoholic controls. Alcoholic women had significantly higher levels of cholesterol and apo A-II when compared with nonalcoholic controls. Serial measurements in 25 alcoholic patients showed a significant decline in HDL-C, apo A-I, and apo A-II levels during the 4-wk hospital stay. HDL-C demonstrated its expected inverse relationship with plasma triglyceride level and its direct relationship with apo A-I, apo A-II, and the hepatic enzyme aspartate aminotransferase.

Quantification and immunolocalization of apolipoprotein E in experimental atherosclerosis.

We developed a radioimmunoassay for rabbit apolipoprotein E (apo E) for studying their plasma apo E levels and its accumulation in the aorta of rabbits fed a cholesterol diet. Delipidation of plasma did not increase the apo E immunoreactivity and this immunoreactivity was indistinguishable from that in an apo E-phospholipid complex. The antigenic determinants of apo E in lipoprotein particles were therefore fully reacted with our goat anti-apo E antibodies. In our assay system, the non-ionic detergent Tween-20 was found to be necessary to significantly reduce the non-specific binding of 125I-labeled apo E to polystyrene tubes, and yet not interfere with the assay. In rabbits (n = 6) fed a high cholesterol (1%) diet, plasma apo E increased at least 10-fold above baseline levels and reached maximal levels within 17-20 days after the onset of cho-diet feeding. These levels were sharply reduced only 10 days after resuming a normal diet. Plasma total cholesterol levels went through a similar pattern. Thus, the plasma cholesterol concentration can simply be used to monitor the increase of apo E in cholesterol-fed rabbits. All the cholesterol-fed rabbits developed atherosclerotic fatty streak lesions and apo E located mostly in the thoracic region and was significantly correlated with the accumulation of lipids in the areas of lesion. In addition, the apo E deposition was limited to the aortic areas where lipids were present. On the other hand, apo A-I was not detectable in any lesion area. Our data suggest that apo E or apo E-containing lipoproteins, may be involved in the development of atherosclerosis in cholesterol-fed rabbits.

The effect of lipoproteins on the synthesis of prostacyclin, von Willebrand factor and apolipoproteins A-I and A-II in cultured human endothelial cells.

Primary cultures of confluent human endothelial cells (ECM) were grown in media containing the major lipoproteins (LP) and lipoprotein deficient serum (LDS). The release of 6-keto-PGF1 alpha, von Willebrand factor (VIII RAg) and apolipoproteins (apo) A-I and A-II were investigated by radioimmunoassay. The cell-associated VIII RAg, apo A-I and apo A-II were also confirmed by fluorescein antibodies, and the synthesis of the apolipoproteins was examined by incorporation of [3H]leucine. Apo A-I and apo A-II were located and synthesized in ECM, yet only apo A-I was released into the medium. Very low density (VLDL) and low density lipoproteins (LDL) in concentrations of 50-600 micrograms/ml stimulated release of apo A-I. Stimulation of ECM for 5 min with thrombin (T) or arachidonic acid (A) did not induce apo A-I release. VIII RAg was always released into the media from ECM. The release was not affected by the lipoproteins. VIII RAg was also localized on the cell surface (VIII RAgC) and approximately 80% was released by trypsin. LDL stimulated the occurrence of factor VIII RAg on the cell surface. 6-Keto PGF1 alpha was always released into the medium and the production was stimulated by T and AA. The main lipoproteins (50-600 micrograms/ml) and apo A-I and A-II did not affect the release of 6-keto-PGF1 alpha. This study shows that endothelial cells synthesize and release proteins important for thrombogenesis and atherosclerosis. The release of apolipoproteins A-I was stimulated by VLDL and LDL, and the concentration of cell-related factor VIII RAg was stimulated by LDL.