Search

The sterol binding agent 2-hydroxypropyl-beta-cyclodextrin is shown to be a convenient and useful experimental tool to probe intracellular pathways of cholesterol transport. Biochemical and cytochemical studies reveal that cyclodextrin specifically removes plasma membrane cholesterol. Depletion of plasma membrane sphingomyelin greatly accelerated cyclodextrin-mediated cholesterol removal. Cholesterol arriving at the plasma membrane from lysosomes and the endoplasmic reticulum was also removed by cyclodextrin. Cellular cholesterol esterification linked to the mobilization of cholesterol from lysosomes was strongly attenuated by cyclodextrin, suggesting that the major portion of endocytosed cholesterol is delivered from lysosomes to the endoplasmic reticulum via the plasma membrane. Evidence for translocation of lysosomal cholesterol to the endoplasmic reticulum by a plasma membrane-independent pathway is provided by the finding that cyclodextrin loses its ability to suppress esterification when plasma membrane sphingomyelin is depleted. The Golgi apparatus appears to play an active role in directing the relocation of lysosomal cholesterol to the plasma membrane since brefeldin A also abrogated cyclodextrin-mediated suppression of cholesterol esterification. Using cyclodextrin we further show that attenuated esterification of lysosomal cholesterol in Niemann-Pick C cells reflects defective translocation of cholesterol to the plasma membrane that may be linked to abnormal Golgi trafficking.

Elucidation of the pathways for intracellular transport of cholesterol is an important yet elusive goal in cell biology. Analysis of the cellular defects in the human disease Niemann-Pick C (NP-C) is providing insights into this problem. Cholesterol derived from low-density lipoprotein accumulates in lysosomes of NP-C cells, apparently because intracellular movement of such cholesterol is blocked. Identification of the NP-C gene should provide crucial molecular clues to the mechanism of cholesterol transport within cells.

The endothelial cell lining of blood vessels is now recognized as an active interface between blood and the underlying tissue. Modulation of cholesterol levels in several cell types has resulted in altered cell function. We have removed cholesterol from the endothelial cell membrane and have observed corresponding alterations in endothelial cell function. Following depletion of cholesterol from the endothelial cells, polymorphonuclear leukocyte adhesion to the cells was decreased. Angiotensin-converting enzyme activity of the endothelial cells was increased following removal of cholesterol from the endothelial cell membranes. The results of fluorescence polarization measurements suggest that these changes may be partially explained by altered membrane order.

We have examined the initial events in phagocytosis by Acanthamoeba castellanii in order to understand this process at the molecular level and have determined that phagocytosis in this organism is mediated by a receptor which recognizes mannose-rich elements in the particle to be phagocytosed. We demonstrate that the binding and internalization of yeast particles can be inhibited by the sugars (D(+)-mannose and D(-)-fructose in a stereospecific, concentration-dependent manner. This inhibition is specific; these sugars did not inhibit the uptake of latex beads by this organism. Using mannosylated neoglycoproteins, which are much more potent inhibitors of particle binding as compared with the free sugar, we demonstrate the presence of a receptor on the amoeba cell surface which is necessary for the binding of yeast as the initial event of phagocytosis. The Acanthamoeba mannose receptor also appears to be able to mediate the delivery of soluble mannose-rich molecules to a degradative compartment such as the lysosome. Knowledge of this receptor will allow a better understanding of the molecular events of phagocytosis.

The generation of second messengers during phagocytosis of yeast by Acanthamoeba castellanii was examined. The kinetics of binding and internalization of yeast by Acanthamoeba were measured and this was compared with the generation of known second messengers. We observed stimulated degradation of PI-4, 5-P2 to 1,4,5 IP3 with kinetics similar to that observed for the binding of yeast to amoeba. Similar production of IP3 could be induced upon treatment with a soluble mannosylated glycoprotein. We propose that the Acanthamoeba mannose receptor stimulates the degradation of PI-4, 5-P2 to 1,4,5 IP3 as an initial event in phagocytosis.

Greater than 90% of the cholesterol in small unilamellar vesicles composed of egg lecithin and cholesterol (molar ratio 1:0.7) was oxidized by a cholesterol oxidase from Brevibacterium sp., with a single time constant and a half-time of 1 min at 37 degrees C. The enzyme preparation used was at least 95% pure and possessed no detectable phospholipase C activity. Since cholesterol is present in both halves of the bilayer, it was concluded that transmembrane movement of cholesterol in these vesicles occurs with a half-time of 1 min or less at 37 degrees C.

The exchange of cholesterol between two populations of small unilamellar vesicles has been investigated using a new system. Uniformly sized egg lecithin-cholesterol vesicles containing [3H]cholesterol and the glycolipid N-palmitoyl-DL-dihydrolactocerebroside were used as donors, whereas similar vesicles containing unlabelled cholesterol and no glycolipid were used as cholesterol acceptors. The two populations of vesicles were separated with the castor bean lectin Ricinus communis. It was found that greater than 90% of the cholesterol in the donor vesicle could be exchanged with a single time constant, the half-time for the completion of this exchange process being 1.5 h at 37 degrees C. Therefore, the rate of transmembrane movement or flip-flop of cholesterol in these vesicles must be at least as fast as the intermembrane exchange process. Similar results were obtained using hemoglobin-free human erythrocyte ghosts as the acceptor membrane. If the molecular-sieve chromatography step used to fractionate the vesicles was omitted, a non-exchangeable pool of cholesterol was detected which was shown not to be due to the presence of multilamellar vesicles.

Two phospholipid exchange proteins and two phospholipases C have been employed to determine the phospholipid composition of the outer surface of the membrane of influenza virus. These four protein probes have defined the same accessible and inaccessible pool for each viral phospholipid. Phospholipids which are exchangeable or hydrolyzable are located on the outer surface, whereas the inaccessible pool is located at the inner surface of the viral bilayer. The two pools are unequal in size, with ca. 30% of the total phospholipid accessible to the four proteins, and ca. 70% inaccessible. The membrane is thus highly asymmetric with regard to the amount of phospholipid on each side of the membrane. There is also a marked asymmetry of phospholipid composition. Phosphatidylcholine and phosphatidylinositol are enriched in the outer surface, and sphingomyelim is enriched in the inner surface, whereas phosphatidylethanolamine and phosphatidylserine are present in similar proportions in each surface. This distribution is qualitatively different from that previously reported for the human erythrocyte. The close agreement between results obtained with excahnge proteins and phospholipases C demonstrates that the hydrolytic action of these enzymes does not alter phospholipid asymmetry. The nonperturbing nature of the exchange proteins has permitted the rate of transmembrane movement of phospholipids (flip-flop) in the intact virion to be studied. This process could not be detected after 2 days at 37 degrees C. It was estimated that the half-time for flip-flop is indeterminately in excess of 30 days for sphingomyelin and 10 days for phosphatidylcholine at 37 degrees C. These extremely long times provide a simple explanation for the maintenance of transbilayer asymmetry in influenza virions and possibly, other membranes. Since the viral membrane is acquired by budding through the host cell plasma membrane, the transbilayer distribution of phospholipids observed in the virions presumably reflects a similar asymmetric distribution of phospholipids in the host cell surface membrane. Because animal cells in culture do not incorporate extracellular phospholipid, our results demonstrate that individual cells have the capacity to generate asymmetric membranes.

A polar lipid accounting for 12.5% of the total lipid nitrogen has been isolated from the protozoan Acanthamoeba castellanii. On the basis of thin-layer chromatography and mass spectral analysis, the lipid has been identified as diacylglyceryltrimethylhomoserine (DGTS). Fast atom bombardment (FAB) mass spectra of DGTS are reported for the first time and are compared to the FAB mass spectra of phosphatidylcholines and the electron ionization (EI) and field desorption (FD) mass spectra of DGTS. Gas-liquid chromatographic-mass spectrometric (GLC-MS) analysis of the acyl chain composition of this lipid has shown that 87.5% consists of cis-9-octadecenoic acid. Plasma membrane isolated from this organism has shown that labeled DGTS appears in the plasma membrane but is not enriched in this fraction. DGTS has been isolated previously only from a limited number of green plants and one species of fungus. Identification of this lipid in Acanthamoeba indicates that this lipid is distributed among a diverse group of lower eucaryotes.

Coated vesicles isolated from rat liver perfused with diisopropylfluorophosphate (DFP) to inactivate endogenous cholinesterase contained newly synthesized secretory cholinesterase after a 30 min recovery. The cholinesterase is found in coated vesicles of presumed endocytic origin following DFP treatment and perfusion for 3 min with galactosylated cholinesterase, a ligand for the asialoglycoprotein receptor. Highly enriched populations of endocytic and exocytic coated vesicles can be separated by use of a novel cholinesterase mediated density shift technique. The two coated vesicle classes have very similar polypeptide compositions but differ significantly in the ratio of cholesterol to phospholipid.

A novel method has been developed for the study of phospholipid exchange and fusion of phospholipid vesicles. Two homogeneous populations of single bilayer phosphatidylcholine vesicles of similar size but markedly different density have been prepared. "Dense" vesicles were made from brominated dioleoyl phosphatidylcholine. "Light" vesicles were prepared from dioleoyl phosphatidylcholine. The two populations were easily separated by density gradient centrifugation. Phosphatidylcholine exchange protein from beef liver was used to promote lecithin exchange between the vesicle populations. Only the lecithin of the external monolayers of the vesicles was available for exchange by exchange protein, implying that flip-flop of vesicle phosphatidylcholine did not take place at a detectable frequency. No spontaneous intervesicle phosphatidylcholine exchange was observed. However, the dense and light vesicles did spontaneously fuse, over several hours, to produce particles of hybrid density.

Purified phosphatidylcholine exchange protein was used to exchange phosphatidylcholine between homogeneous single-walled phosphatidylcholine vesicles and human erythrocyte ghosts. When excess ghosts were present, it was found that only 70% of the vesicle phosphatidylcholine was available for exchange. This fraction corresponds closely to the amount of phosphatidycholine in the outer monolayer of these vesicles, indicating that only the outer surface of the vesicle is accessible to the exchange protein. Also, it was found that all phosphatidylcholine introduced into vesicles by the exchange protein was available for subsequent exchange. Using the exchange protein, asymmetrical vesicles were prepared in which the outer monolayer was either enriched or depleted in radioactive phosphatidylcholine as compared to the inner monolayer. Re-equilibration of the radioactivity between the two surfaces of the vesicle (flip-flop) could not be detected, even after 5 days at 37degrees. It is estimated that the half-time for flip-flop is in excess of 11 days at 37degrees. These results indicate that the properties of the exchange protein can be expolited to measure phosphatidylcholine flip-flop rates and possible phosphatidylcholine asymmetry in biological and model membranes, without altering the structure of the membrane.

The kinetics of cholesterol exchange between two populations of small unilamellar vesicles has been investigated. There is no change in the initial rate of this exchange process over a 100-fold change in the acceptor vesicle concentration at a constant donor concentration. These results are not consistent with a collision-dependent exchange mechanism. In support of transfer via the aqueous phase, the inclusion of a negatively charged lipid into the vesicles did not affect the exchange rate. Evidence for a water-soluble pool of cholesterol that had partitioned ut of the vesicle was obtained. Finally, cholesterol exchange was observed when donor and acceptor membranes were separated by a barrier through which neither could pass. These data together support our contention that the exchange of cholesterol between these vesicles involves a water-soluble intermediate.

Cholesterol is a major component of biological membranes, yet there is very little information concerning its distribution across the membrane. Recent experiments in our laboratory, using cholesterol oxidase, have demonstrated that cholesterol can undergo a rapid transbilayer movement in lecithin-cholesterol vesicles in a half-time of 1 min or less at 37 degrees C. In order to support this conclusion, we have sought other approaches to the measurement of this process. We now report our finding that the transbilayer movement of thiocholesterol in phospholipid vesicles occurs in a half-time of 1 min or less at 20 degrees C.

The endoplasmic reticulum is the principal site of synthesis and initial incorporation of membrane lipids in eukaryotic cells; the enzymes of glycerolipid biosynthesis are exclusively located on its cytoplasmic surface. To maintain a phospholipid bilayer in this organelle, newly synthesized phospholipids must be translocated to the lumenal surface. Consistent with this are measurements indicating that movement of phospholipids across microsomal membranes is rapid, with a half-time less than 5 min (refs 3 and 4). Rapid movement of phospholipids has also been detected across the plasma membrane of Bacillus megaterium, another site of de novo lipid biosynthesis. The rapid transmembrane movement of phosphatidylcholine has not been detected, however, in vesicles prepared from microsomal lipids. These latter data suggest involvement in the endoplasmic reticulum of a phospholipid-translocating protein, as was first proposed by Bretscher who called it 'flippase'. Here we report reconstitution of a phospholipid flippase from rat liver microsomes into lipid vesicles.

It is shown that methanolysis of dihydrosterculoyl lecithin with boron trifluoride-methanol introduces artifacts which are absent if the methyl ester is prepared by saponification of the lipid followed by treatment with diazomethane.