Effect of human plasma lipoproteins on prostacyclin production by cultured endothelial cells

Prostacyclin (PGI2) production by bovine aortic or human umbilical vein endothelial cells increased when either human high density lipoproteins3 (HDL3) or low density lipoproteins (LDL) were added to a serum-free culture medium. At low concentrations and short incubation times, HDL3 produced more PGI2 than LDL, but LDL was just as effective as HDL3 in 18-hr incubations with high concentrations of lipoproteins. Neither lipoprotein was toxic to the cultures as assessed by [3H]leucine incorporation into cell protein. The stimulatory effect of HDL3 and LDL on PGI2 production decreased as growing cultures became confluent. Incubation with lipoproteins neither enhanced arachidonic acid release nor increased PGI2 formation when the cells were stimulated subsequently with ionophore A23187, indicating that the lipoproteins do not affect the intracellular processes involved in PGI2 production. The addition of albumin reduced the amount of PGI2 formation elicited by HDL3 or LDL. As compared with albumin-bound arachidonic acid, from 6- to 13-fold less PGI2 was produced during incubation with the lipoproteins. Furthermore, the amount of PGI2 formation elicited by the lipoproteins in 18 hr was 4-fold less than that produced during incubation with a fatty acid mixture containing only 5% arachidonic acid, and 3-fold less than when the cells were stimulated with the ionophore A23187 for 20 min. Taken together, our results indicate that human HDL and LDL contribute to endothelial PGI2 production only in a modest way and suggest that this process is not specific for either of these two plasma lipoproteins. In view of the greater participation of albumin-bound arachidonic acid in PGI2 production, plasma lipoproteins may not play as important a role in endothelial prostaglandin formation as has been suggested.     

Lipoprotein-dependent synthesis of prostacyclin in smooth muscle cells

Low (LDL) and high density lipoproteins (HDL) stimulated prostacycline (PGI2) synthesis in rabbit and human aorta smooth muscle cells growing in culture. The lipoproteins were added to the cells in concentrations equal to that of cholesterol. It was shown that HDL exerted a stronger stimulating effect as compared to LDL. The maximal effect was observed with HDL3. HDL3 isolated from blood serum of healthy volunteers appeared to be more active in PGI2 synthesis promotion than those of CDH patients with documented coronary atherosclerosis. Purified Apo A-1 stimulated the transformation of [14C]arachidonic acid into the products of its metabolism with increased accumulation of 6-keto-PGF1 alpha among labeled metabolites. Estradiol (1.10(-7) M) showed a stimulating effect; norepinephrine (1.10(-6) M) and progesterone (1.10(-7) M) showed an inhibiting effect, whereas corticosterone (1.10(-6) M) and deoxycorticosterone (1.10(-6) M) did not influence the rate of LDL-dependent PGI2 synthesis.     

Role of lipid peroxidation in the inhibition of mononuclear cell proliferation by normal lipoproteins

Stimulated peripheral blood mononuclear cells (PBMC) can oxidize normal lipoproteins, and sufficiently oxidized lipoproteins are cytotoxic. However, the role of lipid peroxidation in the inhibition of mitogen-stimulated PBMC proliferation by physiologic concentrations of normal lipoproteins is unclear. In the present investigation, normal low density lipoprotein (LDL) and very low density lipoprotein (VLDL) suppressed [3H]thymidine incorporation and gamma interferon production in concanavalin A-stimulated PBMC without causing cell death. This suppression was accompanied by parallel increases in lipid peroxidation products measured as thiobarbituric acid reactive substances (TBARS). In contrast, high density lipoprotein (HDL) failed to inhibit PBMC and TBARS remains low. Differences between the PBMC suppression from LDL, VLDL, and HDL were best accounted for by normalizing the lipoprotein concentrations by their total lipid content. Moreover, the antioxidants superoxide dismutase and butylated hydroxytoluene each substantially ameliorated the inhibition of PBMC caused by LDL, and reduced the levels of lipid peroxidation products that were generated. Altogether, these results suggest that reactive oxygen species generated by stimulated PMBC may cause oxidative alterations of normal lipoproteins that may, in turn, account for much of the previously reported inhibition of PBMC by normal lipoproteins.     

Lipid transfer inhibitor protein defines the participation of high density lipoprotein subfractions in lipid transfer reactions mediated by cholesterol ester transfer protein (CETP)

Cholesterol ester transfer protein (CETP) moves triglyceride (TG) and cholesteryl ester (CE) between lipoproteins. CETP has no apparent preference for high (HDL) or low (LDL) density lipoprotein as lipid donor to very low density lipoprotein (VLDL), and the preference for HDL observed in plasma is due to suppression of LDL transfers by lipid transfer inhibitor protein (LTIP). Given the heterogeneity of HDL, and a demonstrated ability of HDL subfractions to bind LTIP, we examined whether LTIP might also control CETP-facilitated lipid flux among HDL subfractions. CETP-mediated CE transfers from [3H]CE VLDL to various lipoproteins, combined on an equal phospholipid basis, ranged 2-fold and followed the order: HDL3 > LDL > HDL2. LTIP inhibited VLDL to HDL2 transfer at one-half the rate of VLDL to LDL. In contrast, VLDL to HDL3 transfer was stimulated, resulting in a CETP preference for HDL3 that was 3-fold greater than that for LDL or HDL2. Long-term mass transfer experiments confirmed these findings and further established that the previously observed stimulation of CETP activity on HDL by LTIP is due solely to its stimulation of transfer activity on HDL3. TG enrichment of HDL2, which occurs during the HDL cycle, inhibited CETP activity by approximately 2-fold and LTIP activity was blocked almost completely. This suggests that LTIP keeps lipid transfer activity on HDL2 low and constant regardless of its TG enrichment status. Overall, these results show that LTIP tailors CETP-mediated remodeling of HDL3 and HDL2 particles in subclass-specific ways, strongly implicating LTIP as a regulator of HDL metabolism.     

Imipramine associations with plasma components and its uptake by cultured human cells

It has been proposed that in vivo variability in response to certain hydrophobic chemicals or drugs, such as imipramine, may be due in part to the varying plasma lipid levels in patients. The distribution of [3H]imipramine into the lipoproteins of human plasma was therefore studied. Differential density centrifugation of plasma containing [3H]imipramine resulted in flotation of very low density, low density and high density lipoproteins (VLDL, LDL, HDL) and approximately one-third of the total 3H radioactivity. Twelve percent of the radioactivity was present in the sedimented fraction which included most of the plasma proteins. There appeared to be little specific binding of [3H]imipramine to VLDL or LDL, as shown by ultracentrifugation, dialysis and column chromatography. [3H]Imipramine was readily incorporated into cultured human fibroblasts;o no differences were observed in cellular uptake whether it was added to the medium in plasma, LDL or HDL. Also, no differences in uptake of [3H]imipramine by LDL-receptor positive and receptor negative cells were noted. These experiments indicate that LDL is not a major vehicle for the transport of this drug and that both the bound and free fractions are available for cellular uptake.     

Influence of human low density and high density lipoprotein cholesterol on the in vitro prostaglandin I2 synthetase activity

We investigated in vitro the influence of low density lipoprotein (LDL) cholesterol and high density lioprotein (HDL) cholesterol separated from human serum on prostaglandin I2 synthetase activity studied by the conversion of prostaglandin H2 to prostaglandin I2 by the microsomal fraction of pig aorta. 6-Oxo-prostaglandin F1 alpha was analyzed by gas-liquid chromatography using prostaglandin F1 alpha as internal standard. We found a linear negative correlation (P less than 0.001) between the amount of LDL cholesterol in the incubation solution and prostaglandin I2 synthetase activity, whereas there was a positive correlation (P less than 0.01) between HDL cholesterol and prostaglandin I2 synthesis. A very low concentration of LDL cholesterol and a high concentration of HDL cholesterol stimulated prostaglandin I2 synthesis, whereas a high LDL cholesterol concentration inhibited prostaglandin I2 biosynthesis by 64%. The concentration range of LDL and HDL cholesterol was representative of physiologically low, normal or elevated levels of lipoproteins.     

Effects of chylomicron remnants and beta-VLDL on the class and composition of newly secreted lipoproteins by HepG2 cells

The regulation of lipoprotein secretion in the cell line HepG2 was studied. HepG2 cells were preincubated with chylomicron remnants (triglyceride- and cholesterol-rich) or with beta very low density lipoproteins (beta-VLDL) (cholesterol-rich). The medium was removed and the cells were incubated for and additional 24 hr in a lipoprotein-free medium that contained either [2-3H]glycerol or DL-[2-3H]mevalonate. Cells and media were harvested, and lipoproteins were separated and fractionated. The mass and radioactivity of the lipids in cells and in the lipoproteins were measured. The activities of cellular acyl-CoA:cholesterol acyltransferase (ACAT) and 3-hydroxy-3-methylglutaryl CoA (HMG-CoA) reductase were also determined. Preincubation with chylomicron remnants induced an increase in cellular triglyceride and stimulated both HMG-CoA reductase and ACAT. Preincubation with beta-VLDL induced an increase in cellular free and esterified cholesterol, inhibited HMG-CoA reductase and stimulated ACAT. Although the absolute amount of VLDL is small, chylomicron remnants induced large relative increases in the amount of triglyceride and phospholipid secreted in VLDL and decreases in the amount of triglyceride secreted in low density (LDL) and high density (HDL) lipoproteins as well as a decrease in the amount of phospholipid secreted in HDL. In contrast, preincubation with beta-VLDL did not affect triglyceride secretion, but markedly stimulated the amount of phospholipid secreted in HDL. Comparison of the mass of glycerolipid actually secreted with that calculated from the cellular specific activity suggested that glycerolipids are secreted from single, rapidly equilibrating pools. Cholesterol and cholesteryl ester secretion were affected differently. Preincubation with chylomicron remnants increased the amount of free cholesterol secreted in both VLDL and LDL, but did not alter cholesteryl ester secretion. Preincubation with beta-VLDL increased free cholesterol secretion in all lipoprotein fractions and increased cholesteryl ester secretion in VLDL and LDL, but not HDL. Comparison of isotope and mass data suggested that the cholesteryl ester secreted came primarily from a preformed, rather than an newly synthesized, pool. In summary, these data provide insight to the mechanism whereby a liver cell regulates the deposition of exogenous lipid.     

Effect of lipoproteins on macrophage superoxide generation

Superoxide production by macrophages and leukocytes may have an important role in atherogenesis. Whether lipoproteins modulate the superoxide production of these cells is not clear. Therefore, the effect of lipoproteins on the production of superoxide by rat peritoneal macrophages was tested. VLDL and LDL inhibited digitonin-stimulated superoxide production in a dose-dependent manner. Maximum inhibition was observed at 10 μg ml^?1 of VLDL protein and 50 μg ml^?1 of LDL protein respectively. In contrast, HDL (40 μg protein ml^?1) enhanced digitoninstimulated superoxide production (by 47 per cent). Macrophage superoxide production induced by arachidonic acid was enhanced by both VLDL (130 per cent) and HDL (84 per cent), whereas LDL had no effect. The lipoproteins had no effect on macrophage superoxide stimulated by other agonists such as phorbol myristate 13-acetate, sodium fluoride or the calcium ionophore, A23187. The effect of lipoproteins was also tested on human polymorphonuclear leukocyte superoxide generation, stimulated by digitonin and PMA. Ten μg of VLDL, 50 μg of LDL and 50 μg of HDL proteins ml^?1, inhibited digitonin-induced superoxide production by 50, 100 and 33 per cent respectively. Lipoproteins had no effect on PMA stimulated superoxide generation by human polymorphonuclear leukocytes. The stimulatory and inhibitory effects of lipoproteins on macrophage and neutrophil superoxide generation could be important in the understanding of oxidation-mediated development of atherosclerosis.     

Lipoprotein secretion in lean and obese Zucker female rats in vivo and in a single-pass-perfused liver preparation

The plasma lipoprotein composition as well as lipoprotein synthesis and secretion were studied in vivo and in a single-pass-perfused liver preparation in lean and obese Zucker rats. Compared with their lean littermates the levels in the plasma of very low density lipoprotein (VLDL), intermediate density lipoprotein (IDL) + low density lipoprotein (LDL) and high density lipoprotein (HDL) were increased 4-, 2- and 2.5 fold, respectively, in obese rats. In these rats both VLDL and IDL + LDL were enriched in triglycerides, while the HDL were enriched in cholesterol. Although the VLDL and IDL + LDL protein concentrations were the same in lean and obese rats, the HDL protein concentration was 3-fold greater in the obese rats. Both the lean and obese rats incorporated similar amounts of [14C]leucine into total liver protein. However, obese rats incorporated 2.5-fold and 6-fold more [14C]leucine into VLDL and HDL in vivo, 2.7-fold and 1.7 fold more [35S]methionine in VLDL and HDL present in the perfusate, than did lean rats. The perfusate [35S]S-labelled apoproteins (apo-B100, B48; apo-E, apo-AI, apo-AIV and apo-C) were separated by gel electrophoresis and identified by autoradiography. Incorporation of [3H]glycerol into liver, VLDL, IDL + LDL and HDL triglycerides was 2-, 48-, 13- and 1.5-fold higher in obese than in lean rats, respectively. The [3H]-labelled triglycerides in VLDL and IDL + LDL present in the perfusate was 5.4-fold and 4.4-fold more in obese rat. There was no difference in the incorporation of [3H]glycerol into triglycerides of perfusate HDL between the two genotypes of rats. Thus, the hypertriglyceridaemia observed in obese Zucker rats results from very high synthetic rates of both the lipid and protein moieties of plasma lipoproteins. Before this study, no report of the simultaneous triglycerides and protein synthesis in vivo and in a single-pass-perfused liver preparations had been reported.     

Very low and low density lipoprotein synthesis and secretion by the human hepatoma cell line Hep-G2: effects of free fatty acid

The liver is a major source of the plasma lipoproteins; however, direct studies of the regulation of lipoprotein synthesis and secretion by human liver are lacking. Dense monolayers of Hep-G2 cells incorporated radiolabeled precursors into protein ([35S]methionine), cholesterol ([3H]mevalonate and [14C]acetate), triacylglycerol, and phospholipid ([3H]glycerol), and secreted them as lipoproteins. In the absence of free fatty acid in the media, the principal lipoprotein secretory product that accumulated had a density maximum of 1.039 g/ml, similar to serum low density lipoprotein (LDL). ApoB-100 represented greater than 95% of the radiolabeled apoprotein of these particles, with only traces of apoproteins A and E present. Inclusion of 0.8 mM oleic acid in the media resulted in a 54% reduction in radiolabeled triacylglycerol in the LDL fraction and a 324% increase in triacylglycerol in the very low density lipoprotein (VLDL) fraction. Similar changes occurred in the secretion of newly synthesized apoB-100. The VLDL contained apoB-100 as well as apoE. In the absence of exogenous free fatty acid, the radiolabeled cholesterol was recovered in both the LDL and the high density lipoprotein (HDL) regions. Oleic acid caused a 50% decrease in HDL radiolabeled cholesterol and increases of radiolabeled cholesterol in VLDL and LDL. In general, less than 15% of the radiolabeled cholesterol was esterified, despite the presence of cholesteryl ester in the cell. Incubation with oleic acid did not cause an increase in the total amount of radiolabeled lipid or protein secreted. We conclude that human liver-derived cells can secrete distinct VLDL and LDL-like particles, and the relative amounts of these lipoproteins are determined, at least in part, by the availability of free fatty acid.     

Decreased arachidonate metabolism in mouse peritoneal macrophages after foam cell transformation with oxidized low-density lipoproteins.

Oxidized low density lipoproteins (LDL) are now considered to be one of the atherogenic lipoproteins in vivo and to play an important role in the pathogenesis of atherosclerosis. We previously demonstrated in mouse peritoneal macrophages that oxidized LDL stimulated prostaglandin (PG) E2 synthesis when incorporated into the cells [Yokode, M. et al. (1988) J. Clin. Invest. 81, 720-729]. In this study, we investigated arachidonate metabolism in macrophages after foam cell transformation. The cells were incubated with 100 micrograms/ml of oxidized LDL for 18 h, then stimulated with zymosan. Lipid-enriched macrophages which had taken up oxidized LDL produced much less eicosanoids, such as PGE2, 6-keto-PGF1 alpha, and leukotriene C4 than control cells. After labeling of the cells with [14C]arachidonic acid, they were stimulated with zymosan and the phospholipase activity was determined. The activity of lipid-enriched cells was about two-thirds of that of control cells. Then we investigated the fatty acid composition of their phospholipid fraction to clarify arachidonic acid content and mobilization. Percent of arachidonic acid of lipid-enriched cells decreased and less arachidonic acid mobilization was observed after stimulation with zymosan. These data suggest that impaired arachidonate metabolism in lipid-enriched macrophages can be explained by their decreased phospholipase activity and changes in their fatty acid composition.     

Preferential enrichment of large-sized very low density lipoprotein populations with transferred cholesteryl esters

The effect of lipid transfer proteins on the exchange and transfer of cholesteryl esters from rat plasma HDL2 to human very low (VLDL) and low density (LDL) lipoprotein populations was studied. The use of a combination of radiochemical and chemical methods allowed separate assessment of [3H]cholesteryl ester exchange and of cholesteryl ester transfer. VLDL-I was the preferred acceptor for transferred cholesteryl esters, followed by VLDL-II and VLDL-III. LDL did not acquire cholesteryl esters. The contribution of exchange of [3H]cholesteryl esters to total transfer was highest for LDL and decreased in reverse order along the VLDL density range. Inactivation of lecithin: cholesterol acyltransferase (LCAT) and heating the HDL2 for 60 min at 56 degrees C accelerated transfer and exchange of [3H]cholesteryl esters. Addition of lipid transfer proteins increased cholesterol esterification in all systems. The data demonstrate that large-sized, triglyceride-rich VLDL particles are preferred acceptors for transferred cholesteryl esters. It is suggested that enrichment of very low density lipoproteins with cholesteryl esters reflects the triglyceride content of the particles.     

Lipoprotein metabolism in the suckling rat: characterization of plasma and lymphatic lipoproteins

Suckling rat plasma contains (in mg/dl): chylomicrons (85 +/- 12); VLDL (50 +/- 6); LDL (200 +/- 23); HDL1 (125 +/- 20); and HDL2 (220 +/- 10), while lymph contains (in mg/dl): chylomicrons (9650 +/- 850) and VLDL (4570 +/- 435) and smaller amounts of LDL and HDL. The lipid composition of plasma and lymph lipoproteins are similar to those reported for adults, except that LDL and HDL1 have a somewhat higher lipid content. The apoprotein compositions of plasma lipoproteins are similar to those of adult lipoproteins except for the LDL fraction, which contains appreciable quantities of apoproteins other than apoB. Although the LDL fraction was homogeneous by analytical ultracentrifugation and electrophoresis, the apoprotein composition suggests the presence of another class of lipoproteins, perhaps a lipid-rich HDL1. The lipoproteins of lymph showed low levels of apoproteins E and C. The triacylglycerols in chylomicrons and VLDL of both lymph and plasma are rich in medium-chain-length fatty acids, whereas those in LDL and HDL have little or none. Phospholipids in all lipoproteins lack medium-chain-length fatty acids. The cholesteryl esters of the high density lipoproteins are enriched in arachidonic acid, whereas those in chylomicrons, VLDL, and LDL are enriched in linoleic acid, suggesting little or no exchange of cholesteryl esters between these classes of lipoproteins. The fatty acid composition of phosphatidylcholine, sphingomyelin, and lysophosphatidylcholine were relatively constant in all lipoprotein fractions, suggesting ready exchange of these phospholipids. However, the fatty acid composition of phosphatidylethanolamine in plasma chylomicrons and VLDL differed from that in plasma LDL, HDL1, and HDL2. LDL, HDL1, and HDL2 were characterized by analytical ultracentrifugation and shown to have properties similar to that reported for adult lipoproteins. The much higher concentration of triacylglycerol-rich lipoproteins in lymph, compared to plasma, suggests rapid clearance of these lipoproteins from the circulation.     

Free cholesterol is a potent regulator of lipid transfer protein function

This study investigates the effect of altered lipoprotein free cholesterol (FC) content on the transfer of cholesteryl ester (CE) and triglyceride (TG) from very low- (VLDL), low- (LDL), and high-(HDL) density lipoproteins by the plasma-derived lipid transfer protein (LTP). The FC content of VLDL and HDL was selectively altered by incubating these lipoproteins with FC/phospholipid dispersions of varying composition. FC-modified lipoproteins were then equilibrated with [3H] TG, [14C]CE-labeled lipoproteins of another class to facilitate the subsequent modification of the radiolabeled donor lipoproteins. LTP was added and the extent of radiolabeled TG and CE transfer determined after 1 h. With either LDL or VLDL as lipid donor, an increase in the FC content of these lipoproteins caused a concentration-dependent inhibition (up to 50%) of CE transfer from these particles, without any significant effect on TG transfer. In contrast, with HDL as donor, increasing the HDL FC content had little effect on CE transfer from HDL, but markedly stimulated (up to 2.5-fold) the transfer of TG. This differential effect of FC on the unidirectional transfer of radiolabeled lipids from VLDL and HDL led to marked effects on LTP-facilitated net mass transfer of lipids. During long-term incubation of a constant amount of LTP with FC-modified VLDL and HDL, the extent of net mass transfer was linearly related to lipoprotein FC content; a 4-fold increase in FC content resulted in a 3-fold stimulation of the CE mass transferred to VLDL, which was coupled to an equimolar, reciprocal transfer of TG mass to HDL. Since lipid transfer between lipoproteins is integral to the process of reverse cholesterol transport, we conclude that lipoprotein FC levels are a potent, positive regulator of the pathways involved in sterol clearance. FC may modulate lipid transfer by altering the availability of CE and TG to LTP at the lipoprotein surface.     

Effects of prostaglandins on the production of collagenase by rabbit uterine cervical fibroblasts

Effects of prostaglandins on the production of collagenase by rabbit uterine cervical fibroblasts were investigated. Exogenous prostaglandin E2 (PGE2) and PGF2 alpha significantly stimulated the production of collagenase in a dose dependent manner, whereas PGI2 did not. Addition of arachidonic acid in the presence of absence of indomethacin to the cell culture did not show any increase in collagenase production. Recombinant human interleukin-1 (rhIL-1) also promoted the production of cervical collagenase independently of endogenous prostaglandin(s). Furthermore both exogenous PGE2 and PGF2 alpha enhanced the rhIL-1-induced collagenase production whereas PGI2 and/or indomethacin did not. These results suggested that exogenous PGE2 and PGF2 alpha but not endogenous prostaglandin(s) participate in cervical ripening and dilation by enhancing collagenase production by rabbit uterine cervical cells.     

The conversion of apoB100 low density lipoprotein/high density lipoprotein particles to apoB100 very low density lipoproteins in response to oleic acid occurs in the endoplasmic reticulum and not in the Golgi in McA RH7777 cells

The site where bulk lipid is added to apoB100 low density lipoproteins (LDL)/high density lipoproteins (HDL) particles to form triglyceride-enriched very low density lipoproteins (VLDL) has not been identified definitively. We employed several strategies to address this question. First, McA RH7777 cells were pulse-labeled for 20 min with [35S]methionine/cysteine and chased for 1 h (Chase I) to allow study of newly synthesized apoB100 LDL/HDL remaining in the endoplasmic reticulum (ER). After Chase I, cells were incubated for another hour (C2) with/without brefeldin A (BFA) and nocodazole (Noc) (to block ER to Golgi trafficking) and with/without oleic acid (OA). OA treatment alone during C2 increased VLDL secretion. This was prevented by the addition of BFA/Noc in C2. When C2 media were replaced by control media for another 1-h chase (C3), VLDL formed during OA treatment in C2 were secreted into C3 medium. Thus, OA-induced conversion of apoB100 LDL/HDL to VLDL during C2 occurred in the ER. Next, newly synthesized apoB100 lipoproteins were trapped in the Golgi by treatment with Noc and monensin during Chase I (C1), and C2 was carried out in the presence of BFA/Noc with/without OA and without monensin. Under these conditions, OA treatment during C2 did not stimulate VLDL secretion. The same pulse/chase protocols were followed by iodixanol subcellular fractionation, extraction of lipoproteins from ER and Golgi, and sucrose gradient separation of extracted lipoproteins. Cells treated with BFA/Noc and OA in C2 had VLDL in the ER. In the absence of OA, only LDL/HDL were present in the ER. The density of Golgi lipoproteins in these cells was not affected by OA. Similar results were obtained when ER were immuno-isolated with anti-calnexin antibodies. In conclusion, apoB100 bulk lipidation, resulting in conversion of LDL/HDL to VLDL, can occur in the ER, but not in the Golgi, in McA RH7777 cells.     

Redistribution of cholesterol and beta-sitosterol between intralipid and rat plasma lipoproteins and red blood cells in vivo

Male Wistar rats were injected intravenously with 2 mL of Intralipid containing 7.5 x 10(5) counts per minute (cpm) [14C]cholesterol and 7.5 x 10(5) cpm beta-[3H]sitosterol. Blood was withdrawn immediately and at 5, 10, 20, 60, 120, and 1440 min after injection from different animals. Plasma and red cells were separated and washed by conventional centrifugation, while lipoprotein density classes corresponding to chylomicrons, very low (VLDL), low (LDL), and high density lipoproteins (HDL) were isolated by ultracentrifugation. Total lipid and sterol compositions were determined by thin-layer chromatography in combination with gas-liquid chromatography, whereas radioactivity was measured by scintillation counting. The ratio of [14C]cholesterol/beta-[3H]sitosterol rose from 1 to 3.65 in the plasma VLDL fraction, whereas that in the LDL and HDL fractions were equilibrated at about 2, following an initial transient increase in favour of cholesterol. The appearance and disappearance of the radioactivity from LDL and HDL fractions exhibited precursor-product relationship owing probably to the conversion of the Intralipid into an intermediate lipoprotein-X-like particle, which possesses a density similar to that of LDL. The radioactive cholesterol and beta-sitosterol were incorporated into the red blood cell membranes at nearly similar initial rates, while at later times the incorporation of cholesterol was much preferred.     

Ca2(+)-dependent synthesis of prostaglandin I2 and mobilization of arachidonic acid from phospholipids in cultured endothelial cells permeabilized with saponin

The feasibility of using saponin as a permeabilization agent to study the effect of free Ca2+ concentration ([Ca2+]f) on prostaglandin I2 (PGI2) synthesis and mobilization of arachidonic acid from membrane phospholipids was investigated in cultured bovine pulmonary artery endothelial cells (BPAEC). Treatment of BPAEC with 20 micrograms/ml saponin caused selective permeabilization of the plasma membrane as determined by measurements of the release of lactate dehydrogenase and beta-hexosaminidase. In cells prelabeled with [3H]arachidonic acid for 22 h, permeabilization with 20 micrograms/ml saponin induced PGI2 synthesis and release of [3H]arachidonic acid from membrane phospholipids. These effects were dependent upon [Ca2+]f in the range 72 nM to 5 microM. Release of [3H]arachidonic acid from phospholipid classes was determined in suspensions of BPAEC prelabeled with [3H]arachidonic acid and permeabilized with 20 micrograms/ml saponin. At [Ca2+]f optimal for PGI2 synthesis, 16.2% of the total incorporated [3H]arachidonic acid was released from phosphatidylinositol (3.4%), phosphatidylethanolamine (3.5%) and phosphatidylcholine (9.3%). The time course and dependence upon [Ca2+]f of [3H]arachidonic acid release from phospholipids correlated with PGI2 synthesis. The amount of PGI2 synthesized in permeabilized BPAEC was similar to that in cell cultures treated with the calcium ionophore A23187. In comparison, however, PGI2 synthesis induced by A23187 was associated with less release of [3H]arachidonic acid from membrane phospholipids, e.g., 2.3% versus 16.2%. The greater loss of [3H]arachidonic acid from phospholipids in saponin-permeabilized BPAEC was most likely due to the loss of cell integrity and/or nonspecific effects of the detergent on phospholipases. Despite these limitations, the Ca2+ dependence observed for PGI2 synthesis and [3H]arachidonic acid mobilization suggest that saponin-permeabilization may provide a useful system for studies of the intracellular events triggered by the rise in intracellular Ca2+ which culminate in PGI2 synthesis.     

The active synthesis of phosphatidylcholine is required for very low density lipoprotein secretion from rat hepatocytes

Hepatocytes obtained from rats fed a choline-deficient diet for 3 days were cultured in a medium +/- choline (100 microM) or methionine (200 microM). We investigated how choline deficiency affected hepatic lipogenesis, apolipoprotein synthesis, and lipoprotein secretion. The mass of triacylglycerol and phosphatidylcholine secreted was increased about 3-fold and 2-fold, respectively, by the addition of either choline or methionine to the cultured cells. Similarly, a 3-fold stimulation in the secretion of [3H]triacylglycerol and [3H]phosphatidylcholine derived from [3H]oleate was observed after the addition of choline or methionine. Fractionation of secreted lipoproteins by ultracentrifugation revealed that the reduced secretion of triacylglycerol and phosphatidylcholine from choline-deficient cells was mainly due to impaired secretion of very low density lipoproteins (VLDL) (but not high density lipoproteins (HDL)). Fluorography of L-[4,5-3H]leucine-labeled lipoproteins showed a remarkable inhibition of VLDL secretion by choline deficiency. The addition of choline or methionine stimulated the synthesis of phosphatidylcholine and increased the cellular phosphatidylcholine levels to that in normal cells. While there was little effect of choline on the synthesis and amount of cellular phosphatidylethanolamine, the addition of methionine diminished cellular phosphatidylethanolamine levels. Choline deficiency did not change the rate of incorporation of L-[4,5-3H]leucine into cellular VLDL apolipoproteins, nor the rate of disappearance of radioactivity from L-[4,5-3H]leucine-labeled cellular apoB, apoE, and apoC. These results suggest that hepatic secretion of VLDL, but not HDL, requires active phosphatidylcholine biosynthesis. Secondly, the inhibitory effect of choline deficiency on VLDL secretion can be compensated by the methylation of phosphatidylethanolamine.     

Endothelial inositol phosphate generation and prostacyclin production in response to G-protein activation by AlF4-.

In order to elucidate the role of guanine-nucleotide-binding proteins (G-proteins) in endothelial prostacyclin (PGI2) production, human umbilical vein endothelial cells, prelabelled with either [3H]inositol or [3H]arachidonic acid, were stimulated with the non-specific G-protein activator aluminium fluoride (AlF4-). AlF4- caused a dose- and time-dependent generation of inositol phosphates, release of arachidonic acid and production of PGI2. The curves for the three events were similar. When the cells were stimulated in low extracellular calcium (60 nM), they released [3H]arachidonic acid and produced PGI2, but depleting the intracellular Ca2+ stores by pretreatment with the Ca2+ ionophore A23187 totally inhibited both events, although the cells still responded when extracellular Ca2+ was added. The Ca2+ ionophore did not inhibit the generation of inositol phosphates in cells maintained at low extracellular Ca2+. Pertussis toxin pretreatment (14 h) altered neither inositol phosphate nor PGI2 production in response to AlF4-. To investigate the functional role of the diacylglycerol/protein kinase C arm of the phosphoinositide system, the cells were pretreated with the protein kinase C activator 12-O-tetradecanoylphorbol 13-acetate (TPA) or the protein kinase C inhibitor 1-(5-isoquinolinylsulphonyl)-2-methylpiperazine (H7). TPA inhibited the AlF4(-)-induced inositol phosphate generation but stimulated both the release of arachidonic acid and the production of PGI2. H7 had opposite effects both on inositol phosphate generation and on PGI2 production. These results suggest that AlF4(-)-induced PGI2 production is mediated by a pertussis-toxin-insensitive G-protein which activates the phosphoinositide second messenger system. This production of PGI2 can be modulated by protein kinase C activation, both at the level of inositol phosphate generation and at the level of arachidonic acid release.