Exchangeability and rate of flip-flop of phosphatidylcholine in large unilamellar vesicles, cholate dialysis vesicles, and cytochrome oxidase vesicles.

Three model membrane systems have been characterized in terms of their interaction with phospholipid exchange proteins. Large unilamellar vesicles of phosphatidylcholine prepared by ether vaporization are shown to be homogeneous by gel filtration. Phospholipid exchange proteins from three sources are capable of catalyzing the rapid exchange of approximately half of the phospholipid from these vesicles. The remaining pool of radioactive phospholipid is virtually nonexchangeable (t1/2 of several days). Small unilamellar vesicles of phosphatidylcholine prepared by cholate dialysis also exhibit two pools of phospholipid (65% rapidly exchangable, 35% very slowly exchangeable) when incubated with beef liver phospholipid exchange protein. Cytochrome oxidase vesicles prepared both by a cholate dialysis method and by a direct incorporation method have been fractionated on a Ficoll discontinuous gradient, and tested for interaction with beef heart exchange protein. Two pools of phospholipid are once again observed (70% rapidly exchangable, 30% nonexchangeable), even for vesicles which have incorporated the transmembranous enzyme at a phospholipid to protein weight ratio of 2. The size of the rapidly exchangeable pool of phosphatidylcholine for each of the vesicle systems is consistent with the calculated fraction of phospholipid in the outer monolayer. The extremely slow rate of exchange of the second pool of the second pool of phospholipid reflects the virtual nonexistence of phospholipid flip-flop in any of these model membranes.

A new rat liver phospholipid exchange protein.

The soluble fraction from several mammalian tissue homogenates is known to stimulate phospholipid exchange between cell membrane fractions or artificial vesicles. All phospholipid exchange proteins purified to data exhibit an acidic isoelectric point. Using an assay that measures the transfer of [32P] phosphatidylcholine from liposomes to beef heart mitochondria, we report the presence of a new phospholipid exchange protein with a basic isoelectric point (8.4) in rat liver cytosol. A purification procedure, consisting of pH adjustment to 5.1, gel filtrations on Sephadex G 75 and DE 52 cellulose, isoelectric focusing between a pH of 5 and 10, and gel filtration on Sephadex G-50, yielded a fraction with high phosphatidylcholine exchange activity per mg of protein. This fraction exhibits a major band and two minor bands on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The molecular weight of the major band (18 700) is close to that for basic exchange protein fraction obtained by gel filtration (17 000). The distribution of basic and acidic exchange proteins differs markedly in various tissues and animal species. About 50 and 35% of phosphatidylcholine exchange activity from rat liver and rat intestine respectively are due to basic phospholipid exchange proteins. In contrast, no basic exchange protein was found in beef heart and only a small amount in beef liver. In the latter organ, less than 10% of phosphatidylcholine exchange activity was due to a basic phospholipid exchange protein fraction.

Catalytic properties of phospholipid exchange protein from bovine heart.

A protein which catalyzes the exchange of phosphatidylcholine between membranes has been purified from heart tissue homogenates up to 300-fold by acidic pH precipitation, (NH4)2SO4 precipitation, gel filtration, and ion-exchange chromatography. Binding of the protein to phosphatidylcholine liposomes as measured by Sepharose chromatography was nondetectable. However, isoelectric focusing experiments showed that individual molecules of phosphatidylcholine were transferred from liposomes to the soluble, partially purified protein. Exchange of phospholipid between liposomes and mitochondria was not affected by the presence of moderate amounts of cholesterol in liposomes. A search for competitive inhibitors among moieties similar to phosphatidylcholine failed to show strong binding sites in the hydrophilic part of the substrate. High concentrations of Na+, Ca2+ and Mg2+ impaired the exchange activity.

Comparison of various methods for in vitro cholesteryl ester labeling of lipoproteins from hypercholesterolemic rabbits.

Little or no information is available on biologically valid labeling of hypercholesterolemic plasma lipoproteins with cholesteryl ester. The esterification of labeled unesterified cholesterol in hypercholesterolemic rabbit plasma by the lecithin: cholesterol acyltransferase reaction is inefficient. The use of the d greater than 1.063 plasma fraction for this reaction greatly improves the efficiency, but some labeled unesterified cholesterol remains in the end products. The latter disadvantage can be avoided by the addition to whole plasma of labeled cholesteryl ester dissolved in DMSO or acetone. However, in hypercholesterolemic rabbit plasma only a small fraction of the added cholesteryl ester was associated with lipoproteins. When phosphatidylcholine/cholesteryl ester liposomes were incubated with hypercholesterolemic rabbit plasma for 18-24 h at 37 degrees C the labeled cholesteryl ester was quantitatively incorporated into lipoproteins. Chylomicron-like, cholesteryl ester-rich particles were removed by centrifugation (10(6) g X min) and the subsequently isolated d less than 1.019 and d = 1.019-1.063 (LDL) fractions were injected intravenously into normal and hypercholesterolemic rabbits. The disappearance of d less than 1.019 and LDL cholesteryl ester and the appearance of cholesteryl ester in other lipoprotein fractions was indistinguishable from that of in vivo-labeled lipoproteins. In vivo and in vitro cholesteryl ester-labeled lipoproteins were also compared by measuring the exchangeability of their cholesteryl ester with HDL cholesteryl ester in vitro. Equal exchangeability of the two labels was observed in the d less than 1.019 fraction from which the chylomicron-like particles had been removed. These findings demonstrate that when cholesteryl ester is incorporated by the liposome procedure, the distribution of labeled cholesteryl ester within the lipoprotein complex corresponds closely to that of the in vivo-incorporated labeled cholesteryl ester.

Purification and characterization of human plasma proteins that inhibit lipid transfer activities.

A protein which inhibits cholesteryl ester and triacylglycerol transfer activities was purified from human lipoprotein-deficient plasma by chromatography on phenyl-Sepharose CL-4B, chromatofocusing, Bio-Gel A-0.5m and hydroxylapatite. The inhibitor is a sialoglycoprotein with molecular weight 32 000 and a relatively broad isoelectric region of 3.9-4.3. The inhibitor suppressed triacylglycerol and cholesteryl ester transfer activities to a similar extent. Apolipoprotein A-I, which was separated from the inhibitor by chromatofocusing chromatography, suppressed triacyglycerol transfer more than cholesteryl ester transfer. The percentage reduction of lipid transfer between lipoproteins by the inhibitor was independent of the concentration of transfer protein but was decreased at higher lipoprotein concentrations. The inhibition was not observed during lipid transfer between liposomes. These results indicate that the inhibitor interacts with substrates rather than with the transfer protein.

Phosphatidylinositol distribution and translocation in sonicated vesicles. A study with exchange protein and phospholipase C.

The distribution of phosphatidylinositol and phosphatidylcholine in sonicated phospholipid vesicles (phosphatidylcholine : diphosphatidylglycerol : phosphatidylinositol, 90 : 5 : 5 mol%) has been determined by the use of exchange protein from beef heart and phosphatidylinositol-specific phospholipase C from Staphylococcus aureus. Approximately 70% of the phosphatidylinositol in the sonicated vesicles was accessible to the exchange protein and 70--75% was accessible to the phospholipase C. A similar proportion (65%) of the phosphatidylcholine was accessible to the exchange protein suggesting that phosphatidylinositol was not preferentially located in either surface of the phospholipid bilayer. The rate of translocation of both phospholipids was very slow but the rate for phosphatidylcholine (t 1/2 = 4--7 days) appeared to be greater than that for phosphatidylinositol (t 1/2 = 8--60 days). Production of asymmetric vesicles by removing phosphatidylinositol from the outer surface with either exchange protein or phospholipase C did not induce rapid phospholipid translocation.

Use of phospholipid exchange protein to measure inside-outside transposition in phosphatidylcholine liposomes.

The exchange of phosphatidylcholine between (32P)phosphatidylcholine lipososomes and unlabeled mitochondria was catalyzed by a purified phospholipid exchange protein from bovine heart cytosol. The loss of (23P)phosphatidylcholine from the liposomes appeared to proceed in two stages: with 100 units of phospholipid exchange protein per ml the half-time of the initial stage was about 10 min and that of the final stage 4 days or greater. Agarose-gel chromatography of the liposomes showed an elution compatible with a homofwnwoua pool od amLL single walled vesicles. Treatment of phosphatidyl (14C) choline liposomes with phospholipase D (phosphatidylcholine phosphantidohydrolase) showed that labeled phospholipid removable during the rapid exchange phase was subject to hydrolysis by the phospholipase, but that the labeled phospholipid left after the rapid exchange was completed cound not by hydrolyzed by phospholipase D. It is proposed that the rapidly exchanging phosphatidylcholine constitutes the outer layer of the liposome bilayer. The long half-lives of 4 days or more probably represent the transposition of phosphatidylcholine from the inner to the outer layer of the liposome bilayer.

Stimulation by acidic phospholipids of protein-catalyzed phosphatidylcholine transfer.

1. The catalyzed transfer of phosphatidylcholine from unilamellar liposomes to mitochondria by phospholipids exchange protein from beef heart or from beef liver is stimulated by the presence of up to 20 mol% acidic phospholipid (phosphatidylinositol or phosphatidic acid) in the liposome. Co-sedimentation of liposomes with mitochondria increases with increasing mol% acidic phospholipid. 2. The catalyzed transfer of phosphatidylcholine from unilamellar liposomes to multilamellar vesicles by beef heart or beef liver exchange proteins is also stimulated by the presence of acidic phospholipid. No co-sedimentation of negatively charged transfer of phosphatidylcholine from multilamellar vesicles to unilamellar liposomes by phospholipid exchange protein from beef heart or beef liver reaches a maximum at 7.5% phosphatidylinositol in the liposomes. Inhibition of phosphatidylcholine transfer was observed at levels of liposome phosphatiylinositol of greater than 15 mol% only in the presence of beef liver exchange protein. 4. Changes in the surface charge of liposomes by the addition of acidic phospholipid were verified by a novel application of polyvinylchloride block electrophoresis that allows the direct measurement of the relative electrophoretic mobility of sonicated vesicles.

Phospholipid-exchange proteins as membrane probes.

Phospholipid-exchange proteins, first discovered in rat liver, have now been identified in a variety of animal and plant tissues. In rat liver one protein transfers phosphatidylcholine with a high degree of specificity. Another protein fraction is capable of transferring all major types of phospholipids as well as cholesterol. Beef heart is a good source of phospholipid-exchange protein, which transfers phosphatidylinositol and phosphatidylcholine. Studies with phospholipid-exchange protein from beef heart, beef liver and rat liver show that these proteins are useful for the study of membrane structure. When unilamellar vesicles of isotopically labeled phosphatidylcholine are incubated with nonlabeled mitochondria in the presence of phospholipid-exchange protein, only the outer portion of the phosphatidylcholine bilayer is exchangeable and translocation of lipids between inner and outer parts of the bilayer is exceedingly slow. Resealed red blood cell ghosts show an equilibration of phosphatidylcholine between the inner and outer portions of the bilayer with a t1/2 of two hours. An even faster equilibration of various phospholipids in the microsomal mmebrane appears to take place in rat-liver microsomes.

Phospholipid exchange proteins in rat intestine.

The 105,000 g supernatant and pH 5.1 supernatant fractions from rat intestinal homogenates stimulate phosphatidylcholine exchange between [32P] phosphatidylcholine liposomes and beef heart mitochondria. This active fraction shows the characteristics of a protein. Isoelectric focusing of the intestinal pH 5.1 fraction shows two peaks of phosphatidylcholine exchange activity: one at an acidic pH (4.5-5.3), the other in a basic pH range (8-9). The second peak of activity appears to be a new phospholipid exchange protein. The anatomic distribution of phosphatidylcholine exchange activity in intestine has been investigated. Expressed per mg of protein, phosphatidylcholine exchange activity is higher in mucosa than in the intestinal wall. No significant differences have been found between villi and crypts cells or between jejunal and ileal villi. Futhermore, exchange activity per mg of protein in mucosa is unaffected by fasting or by feeding a high fat or high cholesterol diet. This suggests that phospholipid exchange activity in the absorptive cells is not a rate limiting step in the process of fat absorption.

Dissociation between cholesterol secretion and plasma lipid transfer activity in rabbits.

Human and rabbit plasma contains a lipid transfer protein that transfers cholesteryl esters and triglycerides among the plasma lipoproteins and may also have a role in the movement of lipids into and out of cells. Little is known about the regulation of the activity of the lipid transfer protein, but in the rabbit, hypercholesterolemia is associated with increased plasma lipid transfer activity (LTA). Perfused rabbit livers secrete LTA, and hepatic cholesterol secretion is increased in rabbits with diet-induced hypercholesterolemia. Thus, experiments were performed with rabbits to determine if LTA is regulated by a concerted hepatic secretion of lipoprotein protein cholesterol and LTA. Rabbits were fed chow or chow plus coconut oil (14% wt/wt), and plasma lipids, LTA, and the rate of secretion of cholesterol into plasma were determined. Coconut oil feeding increased plasma cholesterol by 68%, LTA by 42%, and hepatic cholesterol secretion by 69%. Mevinolin (75 mg/day), an inhibitor of cholesterol biosynthesis, lowered LTA and plasma cholesterol without affecting the rate of secretion of cholesterol into plasma. These studies provide further evidence that, in the rabbit, plasma cholesterol and LTA are closely related, and the association is not likely to be caused by a concerted hepatic secretion of cholesterol and LTA.

Blockade of intestinal lipoprotein clearance in rabbits injected with Triton WR 1339-ethyl oleate.

Although Triton WR 1339 has been used to block triglyceride or cholesterol removal from plasma, no data are available on the extent to which Triton WR 1339 administered to rabbits blocks clearance of newly absorbed dietary lipids. In the present study, we have measured the efficiency of this blockade during a 24-hr interval. After the Triton WR 1339 administration, plasma Sf greater than 400 and d less than 1.019 g/ml lipoprotein lipid concentrations increased greatly, but the concentration of d greater than 1.019 g/ml lipids decreased. In the rabbits fed 0.5% cholesterol for 1 week, the increase in d less than 1.019 g/ml and the decrease in 1.019 less than d less than 1.063 g/ml lipoprotein fractions 24 hr after the Triton WR 1339 injection were much greater than in the chow-fed Tritonized rabbits. After the Triton treatment, 50% of intravenously injected LDL-125I-labeled apoB disappeared in 24 hr, but little or no apoB appeared in other lipoprotein fractions and no VLDL apoB was converted to LDL. Labeled cholesterol and retinol were fed to rabbits and 24-hr increments in plasma cholesteryl- and retinyl-ester label and mass were measured. In chow-fed Tritonized rabbits about one-half of the absorbed oral doses of both labeled lipids was recovered in plasma, indicating that Triton WR 1339 does not completely inhibit the clearance of intestinal lipoproteins. When rabbits were injected with Triton and an ethyl oleate emulsion, the blockade of dietary lipid removal from plasma was substantially improved and chylomicron cholesterol uptake by extra-hepatic tissues was completely abolished.

Cholesterol catabolism in the rabbit in fasted and fed states.

Urinary and fecal endogenous steroid excretion of fed or fasted New Zealand white rabbits was determined by the isotopic steady state method after subcutaneous implantation of radioactive cholesterol. While plasma cholesterol was increasing during a 9-day fast, fecal steroid excretion decreased to 10% of the excretion rates in the fed state. Refeeding the fasted rabbits led to a decrease in plasma cholesterol and an increase in fecal endogenous steroid excretion. Urinary steroid excretion, which represented 18% of total endogenous steroid excretion for fed animals, decreased during fasting and increased during refeeding, but these changes were relatively small. The small intestine, cecum, and colon of fed or fasted rabbits had similar endogenous steroid was acidic steroid. During attempts to alter the circulating bile acid concentration by supplying deoxycholate (200 mg/day) to fed rabbits or cholestyramine (2 g/day) to fasted rabbits, plasma cholesterol concentration did not change to the same extent as during fasting or refeeding, respectively. The decreased cholesterol catabolism and the hypercholesterolemia that are seen in the fasting rabbit may result from decreased clearance of plasma cholesterol.

Influx of cholesterol into plasma in rabbits with fasting hyperbetalipoproteinemia.

New Zealand white rabbits exhibited as much as a threefold increase in plasma cholesterol but no change in hepatic cholesterol when fasted for 7-9 days. Agarose electrophoresis and ultracentrifugation of plasma samples showed that only low density lipoprotein increased during fasting. Fasting changed the composition of the low density lipoprotein by increasing the percentage of cholesterol and decreasing the percentage of triglyceride while protein and phospholipid remained the same. Rates of cholesterol secretion into plasma, measured by Triton WR 1339 injection, and rates of plasma cholesteryl ester synthesis, determined by [2-(14)C]mevalonate injection, were similar for fed and fasted rabbits. These findings suggest that fasting hypercholesterolemia in rabbits did not result from increased production of low density lipoproteins. Triton WR 1339 was shown to inhibit plasma cholesterol esterification in vitro.

Purification and characterization of lipid transfer protein(s) from human lipoprotein-deficient plasma.

Lipid transfer activities from human plasma have been characterized to determine whether triglyceride and cholesteryl ester transfer proteins are identical. After sequential purification by phenyl-Sepharose, CM-cellulose, chromatofocusing, and gel filtration, both triglyceride and cholesteryl ester transfer activities were purified approximately 15,000-fold compared to lipoprotein-deficient plasma, with a 14% recovery of both transfer activities. The gel filtration fraction showed two bands, Mr 58,300 and 66,400, as determined by electrophoresis in sodium dodecyl sulfate. Two samples, each containing predominately one of the two bands, were obtained by selectively combining the eluates from the gel filtration column. The specific activities of triglyceride and cholesteryl ester transfer promoted by the larger protein were within 10% of those for the smaller protein. The relative rates of transfer for cholesteryl ester, triglyceride, retinyl ester, and cholesteryl ether for each fraction were the same. The transfer of triglyceride by either the large or small molecular weight component was almost completely inhibited by mercurial compounds, whereas cholesteryl ester transfer was relatively unaffected. We conclude that triglyceride and cholesteryl ester are transferred by the same plasma protein(s).

Lipid transfer proteins in the study of artificial and natural membranes.

Lipid transfer proteins, differing in their specificity for the transfer of lipids and for the surfaces on which they act, have been purified from various mammalian tissues and subsequently characterized. Several of their properties make them useful research tools. They have been used alone or with other techniques to study the distribution and mobility of phospholipids in artificial vesicles and in natural membranes, and have been used to create asymmetric phospholipid vesicles. Lipid transfer proteins are capable of altering the lipid composition of membranes by introducing new lipids or by depletion of existing lipids. Some of the transfer proteins can effect a net transfer of phospholipids, glycosphingolipids and cholesterol from one structure to another, whereas others appear to act primarily in promoting exchange. Some lipid transfer proteins are capable of introducing spin labeled and fluorescent lipid analogs into the outer surface of membranes. Because of lipid transfer proteins do not seem to alter membrane lipid asymmetry or permeability of membranes, they are useful tools for studying the effect of lipid substitution on membrane-mediated transport processes and on various membrane-bound enzyme systems.

Accelerated transfer of neutral glycosphingolipids and ganglioside GM1 by a purified lipid transfer protein.

The transfer of labeled neutral glycosphingolipids from sonicated phosphatidylcholine vesicles to erythrocyte ghosts is greatly stimulated by a nonspecific lipid transfer protein purified from beef liver. Globo-tetraglycosylceramide is transferred at a rate 40% of that for dipalmitoylphosphatidylcholine. II3-alpha-N-Acetylneuraminosyl-gangliotetraglycosylceramide is also transferred by the transfer protein, either from sonicated phosphatidylcholine vesicles or from ganglioside micelles to erythrocyte ghosts. The nonspecific lipid transfer protein catalyzes the net transfer of glycosphingolipids from brush border membrane vesicles (from rabbit intestine) to sonicated phosphatidylcholine/cholesterol vesicles.