Fatty acid composition in native and cultured human endothelial cells

Summary Endothelial cells from human umbilical veins were isolated by collagenase treatment. Cells were cultured in the presence of either 20% fetal bovine serum (FBS) or 20% human serum (HS). At confluency, endothelial cell lipids were labeled with tracer concentrations of tritiated arachidonic acid, then extracted and separated into lipid subclasses by thin layer chromatography. The fatty acid composition of each lipid class was determined by glass capillary gas-liquid chromatography analysis and compared to that of cells freshly isolated from the cord (NC cells). The fatty acid compositions differed only in phospholipids. Polyunsaturated fatty acids (PFAs), arachidonic, and linoleic acids were depleted in FBS cell phospholipids and replaced by both stearic and oleic acids. No significant difference could be observed between NC cell and HS cell phospholipids. We conclude that PFAs might be decreased in FBS cells because of the relative paucity of PFAs in FBS as compared to HS. It seems therefore more convenient to cultivate endothelial cells in the presence of HS, especially in respect to their phospholipid content of arachidonic acid, which is the physiological reservoir for prostacyclin synthesis. This work was supported by a grant from the Délégation Générale à la Recherche Scientifique et Technique, Paris, France (79.7.0091).     

Membrane fatty acid composition and endothelial cell functional properties

In order to study the influence of endothelial cell fatty acid composition on various membrane related parameters, several in vitro methods were developed for manipulating the fatty acid content of human endothelial cell membranes. Changes in membrane fatty acid profile were induced by using fatty acid modified lipoproteins or free fatty acids. The largest changes in endothelial fatty acid composition were obtained by culturing the cells in media supplemented with specific free fatty acids. An increase in arachidonic acid content of endothelial phospholipids was induced by supplementation with saturated fatty acids or with arachidonic acid itself. A decrease in arachidonic acid content was obtained by supplementation with other unsaturated fatty acids. Under the experimental conditions used endothelial cells showed a low desaturase activity and a high elongase activity. Considerable alterations in membrane fatty acid composition did not greatly influence certain membrane related parameters such as polymorphonuclear leukocyte adherence and endothelial cell procoagulant activity. In general, for fatty acid modified endothelial cells an association between endogenous arachidonic acid content and total production of eicosanoids was found. This study demonstrates that considerable changes in membrane fatty acid profile affect endothelial cell arachidonic acid metabolism, but it also illustrates homeostasis at the level of endothelial cell functional activity.     

Selective internalization of arachidonic acid by endothelial cells.

The asymmetric distribution of phospholipids in bovine endothelial-cell membranes was probed with 2,4,6-trinitrobenzenesulphonate and purified phospholipase A2. The data suggest that phosphotidylethanolamine is primarily located in the inner lipid bilayer, as reported for other cell types. Stearic acid is taken up by the endothelial cells and is randomly distributed among the membrane phospholipids. In contrast, the polyunsaturated fatty acids (arachidonic, eicosatrienoic and eicosapentaenoic acids) have initial incorporation into the phosphatidylcholine fraction. These fatty acids then undergo a time-dependent transfer from phosphatidylcholine to phosphatidylethanolamine. Thus we propose that endothelial cells possess a mechanism for the selective internalization of polyunsaturated fatty acids.     

Methylmalonic and propionic acidemias: lipid profiles of normal and affected human skin fibroblasts incubated with [1-14C]propionate

Normal human skin fibroblasts and those from methylmalonic acidemia and propionic acidemia patients were grown in culture. Following incubation with [1-14C]propionate, the major lipid classes in the cells were separated by thin layer chromatography and isolated fractions analyzed by radio gas chromatography for the presence of odd-numbered long-chain fatty acids; the pattern of even-numbered long-chain fatty acids was obtained also. Normal fibroblasts incorporated a small percentage of propionate into odd-numbered fatty acids which were present in all lipids studied. The abnormal cells incorporated a larger amount while maintaining the characteristic ratios of odd-numbered fatty acids found in the normal line. Most of the radioactivity was associated with phospholipids which are the predominant constituents of cell membranes. A characteristic C15/C17 ratio was found for different phospholipids and the triglyceride fraction; pentadecanoic acid was the principal odd-numbered fatty acid utilized in the assembly of complex lipids. Compared to even-numbered long-chain fatty acids the absolute amount of odd-numbered fatty acids was low (1-2%), even in affected cells. An unusual polar lipid fraction was isolated in the course of the study. In the normal cell it contained several unlabeled eicosanoids which were missing from the same fraction of both affected cell lines.     

Modification of the fatty acid composition of cultured human fibroblasts.

The fatty acid composition of human skin fibroblasts grown in 10% dialyzed fetal calf serum can be modified considerably by adding supplemental fatty acids to the culture medium. The degree of modification was dependent on the concentration of added fatty acid over the range tested, 2.5 X 10(-5) to 1 X 10(-4) M. At the higher concentration, the extent of the modifications was as those which can be produced in nonhuman or malignant cell lines. Although the greatest changes were produced in the neutral lipid fraction, the cellular phospholipids also exhibited appreciable modifications. The phospholipids isolated from a microsomal fraction prepared from the cell homogenate exhibited similar changes in fatty acyl composition. These findings indicate that the human fibroblast can tolerate considerable variability in fatty acid composition, even in membrane phospholipids. The triglyceride content of the cells increased when they were grown in the presence of added fatty acids, but the phospholipid and cholesterol content remained unchanged. Growth was not affected by either oleic or linoleic acids, but it was reduced up to 50% when palmitic linolenic, or arachidonic acid was added in concentrations of 5 X 10(-5) M or above. Extensive modifications in phospholipid fatty acid composition also were produced in confluent monolayers of these fibroblasts. This suggest that some membrane lipid turnover occurs even when the cultures are not rapidly growing. Fatty acid modifications also were produced in the commercially available IMR-90 strain of human lung fibroblasts, suggesting that the ability to tolerate considerable differences in fatty acid composition is not a special property of the skin fibroblast line that was isolated locally.     

Preferential incorporation of eicosanoid precursor fatty acids into human umbilical vein endothelial cell phospholipids

We have examined the preferential incorporation of specific fatty acids into phospholipid classes of cultured human umbilical vein endothelial cells. Pulse-labeling of human umbilical vein endothelial cell phospholipids with radiolabeled fatty acids and inhibition of radiolabeled fatty acid incorporation by competition with excess, unlabeled fatty acids in pair-wise combinations revealed two distinct classes of esterification systems into human umbilical vein endothelial cell phospholipids. The eicosanoid precursor fatty acids, including arachidonate, 8,11,14-eicosatrienoate (ETA) and 5,8,11,14,17-eicosapentaenoate (EPA), exhibited high affinity incorporation into total phospholipids, whereas other fatty acids, including docosahexaenoate and monohydroxy eicosatetraenoates, showed low affinity incorporation. The relative degree of incorporation of eicosanoid precursor fatty acids into phospholipid classes was phosphatidylcholine (PC) greater than phosphatidylethanolamine (PE) greater than phosphatidylinositol (PI) greater than phosphatidylserine (PS). The specific activity of [14C]arachidonic acid-labeled PI was two times higher than that of any other radiolabeled phospholipids. When competitive incorporation of eicosanoid precursor fatty acids into phospholipid classes was studied, they were found to be acylated into different phospholipid classes at different rates. Although eicosanoid precursor fatty acids were not preferentially incorporated into PC, arachidonic acid was preferentially incorporated into the other phospholipids and exhibited particular selectivity in comparison with the other eicosanoid precursor fatty acids for incorporation into PI. These results demonstrate that human umbilical vein endothelial cells possess selective incorporation mechanisms for specific fatty acids into various phospholipids via the deacylation-reacylation pathway.     

Modification of the fatty acid composition of individual phospholipids and neutral lipids after infection of the simian erythrocyte by Plasmodium knowlesi

Using capillary gas-liquid chromatography, we have analyzed the alteration in the total fatty acid, phospholipid and neutral lipid compositions of the monkey erythrocyte, after infection by the malarial parasite Plasmodium knowlesi. Data based on fatty acid quantitation show that the phospholipid composition is altered, with particularly large increases in phosphatidylcholine (PC) and phosphatidylethanolamine (PE), the most abundant phospholipids in normal and P. knowlesi-schizont-infected cells. Unesterified fatty acids were found to be less abundant in infected cells. The total fatty acid content of the cell is increased 6-fold during infection, and total fatty acid composition is also changed: the infected cells are richer in palmitate (+23%), oleate (+29%) and linoleate (+89%), but contained less stearate (-27%) and arachidonate (-40%). The determination of the fatty acid composition of individual phospholipids, neutral lipids and unesterified fatty acids showed that choline-containing phospholipids (PC and sphingomyelin) were not as altered in their fatty acid pattern as anionic phospholipids (PE, phosphatidylserine (PS) and phosphatidylinositol (PI) and lysophosphatidylcholine (lysoPC). Specific alterations in the fatty acid compositions of individual phospholipids were detected, whereas the rise in linoleic acid was the only change during infection that was recovered in each phospholipid (except PC), neutral lipid and unesterified fatty acids. The fatty acid composition of the neutral lipids and unesterified fatty acids was particularly modified: the only rise in arachidonic acid level was observed in these lipid classes after infection. The total plasmalogen level of the erythrocyte is decreased in infected cells (-60%), but their level is increased in PI.     

Characterization of arachidonic acid-induced apoptosis

Tumor necrosis factor (TNF) can induce apoptosis in a number of different cell types. This response often depends on the activity of cytosolic phospholipase A₂ (cPLA₂), which catalyzes the release of arachidonic acid from the sn-2 position of membrane phospholipids. In this study, we investigate the ability of arachidonic acid itself to cause cell death. We show that in assays with 10% fetal bovine serum (FBS) arachidonic acid will not kill, nor does act synergistically with TNF. In contrast, by lowering the concentration of FBS to 2% it is possible to use arachidonic acid to induce cell death. Arachidonic acid-induced cell death was judged to be apoptotic based on morphology and the cleavage of poly (ADP) ribose polymerase. Arachidonic acid was able to kill all cell lines tested including two human melanoma-derived cell lines, and susceptibility to arachidonic acid was not influenced by adenovirus gene products that control susceptibility to TNF. Finally, we show that arachidonic acid is unique among 20 carbon fatty acids for its ability to induce apoptosis and that several other unsaturated, but not saturated fatty acids can also induce apoptosis.     

Docosahexaenoic acid metabolism and effect on prostacyclin production in endothelial cells

Bovine aortic endothelial cultures readily take up docosahexaenoic acid (DHA). Most of the DHA was incorporated into phospholipids, primarily in ethanolamine and choline phosphoglycerides, and plasmalogens accounted for 34% of the DHA contained in the ethanolamine fraction after a 24-h incubation. The retention of DHA in endothelial phospholipids was not greater than other polyunsaturated fatty acids and unlike arachidonic and eicosapentaenoic acids, DHA did not continue to accumulate in the ethanolamine phosphoglycerides after the initial incorporation. About 15% of the [14C(U)]DHA uptake was retroconverted to docosapentaenoic and eicosapentaenoic acids in 24 h. Some of the newly incorporated [14C(U)]DHA was released when the cells were incubated subsequently in a medium containing serum and albumin. The released radioactivity was in the form of free fatty acid and phospholipids and after 24 h, 11% was retroconverted to docosapentaenoic and eicosapentaenoic acids. Total DHA uptake was decreased only 10% by the presence of a 100 microM mixture of physiologic fatty acids, but as little as 10 microM docosatetraenoic acid reduced DHA incorporation into phospholipids by 25%. DHA was not converted to prostaglandins or lipoxygenase products by the endothelial cultures. When DHA was available, however, less arachidonic acid was incorporated into endothelial phospholipids, and less was converted to prostacyclin (PGI2). Enrichment of the endothelial cells with DHA also reduced their capacity to subsequently produce PGI2. These findings indicate that endothelial cells can play a role in DHA metabolism and like eicosapentaenoic acid, DHA can inhibit endothelial PGI2 production when it is available in elevated amounts.     

Specificity of phospholipases in methylcholanthrene-transformed mouse fibroblasts activated by bradykinin, thrombin, serum, and ionophore A23187.

Methylcholanthrene-transformed mouse fibroblasts synthesize prostaglandins in response to bradykinin, thrombin, serum, and the ionophore A23187. These agents activate phospholipases, thereby releasing fatty acids from phospholipids. To examine the phospholipid specificity of the phospholipases activated by bradykinin, thrombin, serum, and A23187, cells were labeled with [14C]arachidonic acid and stimulated with these agents in the presence of delipidated bovine serum albumin. Phospholipid classes were resolved by two-dimensional chromatography on silica gel-coated paper. Only phosphatidylinositol and phosphatidylcholine lost radioactivity upon stimulation. To characterize the fatty acid specificity of the phospholipases, cells were incubated with 14C-labeled stearic, oleic, linoleic, eicosatrienoic, or arachidonic acid and then exposed to the stimuli. Bradykinin, thrombin, and serum caused specific release of radioactivity into the medium only from cells labeled with arachidonic acid or eicosatrienoic acid, whereas A23187 caused release from cells labeled with any one of the five fatty acids. We conclude that bradykinin, thrombin, and serum activate phospholipases that specifically hydrolyze arachidonyl and eicosatrienoyl phosphatidylinositol and phosphatidylcholine, whereas A23187 is less specific activator of phospholipases.     

The role of arachidonic acid in vasogenic brain edema

Arachidonic acid is released rapidly from cellular membrane phospholipids after pathological insults associated with the delayed development of brain edema. Intracerebral injection of arachidonic acid caused significant increases in brain water and sodium content with decreases in potassium content and Na+,K+-ATPase activity. The 125I-labeled bovine serum albumin spaces in brain (a measure of blood-brain barrier permeability) rose threefold 24 h after arachidonic acid injection. There was gross and microscopic evidence of edema. Saturated fatty acids and monounsaturated fatty acids were not effective. These data indicate that the endothelial cells of the blood-brain barrier are target sites for the action of arachidonic acid. It is hypothesized that the increased permeability of endothelial cells to macromolecules and water results from alterations of membrane phospholipids and increased vesicular transport, changes that are responsible for the delayed development of vasogenic edema.     

Effect of phospholipid fatty acid composition of endothelial cells on cholesterol efflux rates.

Human endothelial cells (EA.hy 926 line) were loaded with cationized low density lipoprotein (LDL) and subsequently incubated with fatty acid/bovine serum albumin complexes. The fatty acids were palmitic, oleic, linoleic, arachidonic, and eicosapentaenoic acids. The preincubations resulted in extensively modified fatty acid profiles in cell membrane phospholipids and in cellular cholesteryl esters. The cholesterol efflux from these fatty acid-modified cells was measured using 0.2 mg high density lipoprotein3 (HDL3)/ml medium. The efflux was significantly higher for the palmitic acid-treated cells, compared to all other fatty acid treatments. These differences in efflux rates were not caused by changes in the binding of HDL3 to high affinity receptors on the EA.hy 926 cells. Efflux mediated by dimethyl suberimidate-treated HDL3, which does not interact with high affinity HDL receptors, was similar to efflux induced by native HDL3 after all fatty acid treatments. Our results indicate that high affinity HDL receptors are not important for HDL-mediated efflux of cell cholesterol. The fatty acid composition of the cell membrane phospholipids may be an important determinant.     

Changes in the fatty acid profiles of red blood cell membrane phospholipids in human neonates during the first month of life

We have studied the changes in the fatty acid profiles of red blood cell membrane phospholipids in 47 infants who were exclusively fed human milk from birth to 1 month of life. Twenty blood samples were obtained from cord, 15 at 7 days and 12 at 30 days after birth. Membrane phospholipids were obtained from erythrocyte ghosts by thin-layer chromatography and fatty acid composition was determined by gas liquid chromatography. Phosphatidylcholine showed the most important changes during early life; stearic, w6 eicosatrienoic and arachidonic acids decreased whereas oleic and linoleic acids increased. In phosphatidylethanolamine, palmitic and stearic acid declined and oleic, linoleic and docosahexenoic acids increased with advancing age. Small changes were noted for individual fatty acids in phosphatidylserine. In sphingomyelin stearic acid increased from birth to 1 month and linoleic, arachidonic and nervonic acids decreased. Total polyunsaturated fatty acids of the w6 series greater than 18 carbon atoms increased with advancing age in phosphatidylethanolamine and decreased in choline and serine phosphoglycerides and in sphingomyelin. Long chain fatty acids derived from linoleic acid decreased in phosphatidylcholine but increased in ethanolamine and serine phosphoglycerides. The different behavior in the changes observed in fatty acid patterns for each erythrocyte membrane phospholipid may be a consequence of its different location in the cell membrane bilayer and specific exchange with plasma lipid fractions.     

Alteration of the fatty acid composition of membrane phospholipids in mouse lymphoid cells

A simple method is described for introducing exogenous fatty acids into the membrane phospholipids of the murine leukemia cell EL-4, and into the membrane phospholipids of resting mouse lymphocytes. The method involves culturing of the cells with free or methylated fatty acids at concentrations up to 50 microgram/ml. The presence of serum in the culture medium does not interfere with fatty acid uptake, but does increase the growth rate and viability of the cells. Membrane lipid composition returns to normal after the cells are grown in medium without exogenous fatty acid. Fractionation of the cell membranes confirmed that exogenous fatty acids were incorporated into the phospholipids of the plasma membrane.     

Role of triglycerides in endothelial cell arachidonic acid metabolism

Arachidonic acid was incorporated into triglycerides by cultured bovine endothelial cells in a time- and concentration-dependent manner. At 75 microM or higher, more arachidonic acid was incorporated into triglycerides than into phospholipids. The triglyceride content of the cells increased as much as 5.5-fold, cytoplasmic inclusions appeared, and arachidonic acid comprised 22% of the triglyceride fatty acids. Triglyceride turnover occurred during subsequent maintenance culture; there was a 60% decrease in the radioactive arachidonic acid contained in triglycerides and a 40% decrease in triglyceride content in 6 hr. Most of the radioactivity was released into the medium as free fatty acid. The turnover of arachidonic acid, but not oleic acid in cellular triglycerides, decreased when supplemental fatty acid was added to the maintenance medium. Incorporation and turnover of radioactive arachidonic acid in triglycerides also was observed in human skin fibroblasts, 3T3-L1 cells, and MDCK cells. Other fatty acids were incorporated into triglycerides by the endothelial cells; the amounts after a 16-hr incubation with 50 microM fatty acid were 20:3 greater than 20:4 greater than 18:1 greater than 18:2 greater than 22:6 greater than 16:0 greater than 20:5. These findings indicate that triglyceride formation and turnover can play a role in the fatty acid metabolism of endothelial cells and that arachidonic acid can be stored in endothelial cell triglycerides.     

Manipulation of fatty acid composition of membrane phospholipid and its effects on cell growth in mouse LM cells

Fatty acid composition of the phospholipids of mouse LM cells grown in suspension culture in serum-free chemically defined medium was modified by supplementing the medium with various fatty acids bound to bovine serum albumin.Following supplementation with saturated fatty acids of longer than 15 carbons (100 μM) profound inhibition of cell growth occurred; this inhibitory effect was completely abolished when unsaturated fatty acids were added at the same concentration. Supplementing with unsaturated fatty acids such as linoleic acid, linolenic acid or arachidonic acid had no effect on the cell growth.Fatty acid composition of membrane phospholipids could be manipulated by addition of different fatty acids. The normal percentage of unsaturated fatty acids in LM cell membrane phospholipids (63%) was reduced to 35–41% following incorporation of saturated fatty acids longer than 15 carbon atoms and increased to 72–82% after addition of unsaturated fatty acids.A good correlation was found between the unsaturated fatty acid content of membrane phospholipids and cell growth. When incorporated saturated fatty acids reduced the percentage of unsaturated fatty acids in membrane phospholipids to less than 50%, severe inhibition of the cell growth was found. Simultaneous addition of an unsaturated fatty acid completely abolished this effect of saturated fatty acids.     

Fatty acid composition of serum lipids in bilharzial hepatic fibrosis and chronic active hepatitis.

The fatty acid pattern of blood serum lipids was examined by gas liquid-chromatography in 30 cases with bilharzial hepatic fibrosis, 11 cases with chronic active hepatitis accompanied by jaundice, and 28 healthy individuals as a comparison group of the same socioeconomic class of patients. In addition, the fatty acid patterns of the three major serum lipid classes, namely: cholesterol ester, phospholipids and triglycerides, were also investigated in seven cases of each group by gas liquid chromatography. The most remarkable differences were: a depression of the essential fatty acid level (linoleic and arachidonic) in both groups of patients together with a concomitant elevation of oleic acid in the bilharzial group and an elevation of oleic, palmitic, palmitoleic acids in the chronic active hepatitis group. The depression of linoleic and arachidonic acids was explained by the low fat diet intake, malnutrition, and the malabsorption factors which were frequent in all the patients studied. The elevation of monoethenoid acids was attributed to the decrease in the ability of the liver to desaturate the endogenous saturated and monounsaturated acids to polyunsaturated ones.     

Alteration of fatty acid composition of LM cells by lipid supplementation and temperature.

Alteration of the fatty acid composition of monolayer cultures of LM cells grown in chemically defined medium was achieved by supplementation with fatty acids complexed to bovine serum albumin. Phospholipids containing up to 40% linoleate were found in cells grown in medium containing 20 mu g of linoleate/ml. Incorporation of linoleate into phospholipids reached a plateau after 12-24 hr, and cells remained viable for at least 3-4 days. Although linoleic, linolenic, and arachidonic acids were incorporated into LM cells equally well, only the latter was elongated by these cells under these experimental conditions. Nonadecanoic acid was incorporated to a lesser extent than the polyunsaturated fatty acids. Phosphatidylcholine and phosphatidylethanolamine of LM cells had different fatty acid compositions; phosphatidylethanolamine contained more longer chain and unsaturated fatty acids. Cells were also grown in the absence of choline and presence of choline analogs such as N,N-dimethylethanolamine, N-methylethanolamine, 3-amino-1-propanol, and 1-2-amino-1-butanol. The analog phospholipids in these cells had fatty acid compositions which were intermediate between those of phosphatidylethanolamine and phosphatidylcholine of control cells grown in the presence of choline. Linoleate was found in both phosphatidylcholine and phosphatidylethanolamine of cells supplemented with linoleate. The sphingolipid fraction of these cells, however, did not contain significant amounts of linoleate. When linoleate was present in the phospholipids, compensatory decreases in the oleate and palmitoleate content of phospholipids were observed. Lowering of the growth temperature to 28 degrees produced an increase in unsaturate fatty acid content of the phospholipids. When linoleate was supplied to cells grown at 28 degrees, there was no further increase in the unsaturated fatty acid composition of the phospholipids. Using both fatty acid supplementation and lowered growth temperature, LM cell membranes can be produced which have phospholipids with vastly different fatty acid compositions.     

Distribution of phospholipid molecular species containing arachidonic acid and cholesterol in V79-UF cells

V79-UF cells were isolated from Chinese hamster V79 cells as a cell line that requires exogenous unsaturated fatty acids for growth. V79-UF cells incorporated arachidonic acid into phospholipids. The molecular species of diacyl phosphatidylcholine and phosphatidylethanolamine containing arachidonic acid comprised 61.4 and 70.5% of the total phospholipid molecular species in total membranes and 58.1 and 64.7% in plasma membrane, respectively. Polyunsaturated molecular species were distributed in a higher amount in the intracellular membranes than in the plasma membrane. No significant difference was seen in the diffusion coefficient between the plasma membranes from cells supplemented with oleic and arachidonic acids in spite of a distinct difference in the degree of unsaturation between the molecular species of these plasma membranes. The amount of cholesterol in the plasma membrane was higher in the cells grown in the presence of arachidonic acid than in those grown in the presence of oleic acid.     

Composition of phospholipids and of phospholipid fatty acids and aldehydes in human red cells

Improved methods for lipid analysis that have been developed recently were employed to reevaluate the phospholipid composition, the fatty acid and fatty aldehyde composition of the total phospholipid, and the fatty acid composition of the individual phospholipids of normal human red cells. Thirty-three fatty acids and five fatty aldehydes were estimated and tentatively identified in the total phospholipid of normal human red cells. Additional minor components were evident. The major individual phospholipids were isolated by silicic acid thin-layer chromatography and quantified. The fatty acid compositions of phosphatidyl ethanolamine, phosphatidyl serine, lecithin, and sphingomyelin were determined. Each of these phospholipids showed a distinctive and characteristic fatty acid pattern.