The use of low-EPA fish oil for long-chain polyunsaturated fatty acid supplementation of preterm infants.

Because docosahexaenoic acid (DHA) may be an essential nutrient for the visual and early cognitive development of preterm infants, DHA enrichment of preterm formulas has been recommended. This randomized trial was designed to study the n-6 and n-3 fatty acid status of healthy preterm infants fed a formula enriched with a low eicosapentaenoic-fish oil until 4 mo corrected age compared with that of infants fed a standard formula. A reference group of breast-fed infants was studied concurrently. The fatty acid content of red blood cell (RBC) phospholipid was assessed at enrollment, hospital discharge, expected term, and 3 and 6 mo postterm. The DHA content of RBC phospholipid was higher in infants fed the enriched versus the standard formula at hospital discharge, expected term, and 3 and 6 mo postterm. However, compared with infants fed the standard formula, infants fed the enriched formula had also higher RBC phospholipid eicosapentaenoic content (0.69 +/- 0.15% versus 0.25 +/- 0.12%, p < 0.001), and lower RBC phospholipid arachidonic acid content (15.1 +/- 0.93% versus 18.8 +/- 0.89%; p < 0.001). We conclude that supplementing preterm infants with low-eicosapentaenoic fish oil is effective in improving DHA status, but results in worsening of n-6 fatty acid status. We speculate that preterm infants may require a dietary supply of arachidonic acid as well as DHA if the same fatty acid status as that of breast-fed infants is to be achieved.

Effects of liver transplantation on long-chain polyunsaturated fatty acid status in infants with biliary atresia.

BACKGROUND: The long-chain polyunsaturated fatty acid (LC-PUFA) status of infants with untreated biliary atresia (BA) is known to be poor and is correlated to the severity of the liver disease. Liver transplantation (LT) markedly increases survival of patients with BA but the extent to which this reverses poor LC-PUFA status is not known. METHODS: To explore this question, the erythrocyte (red blood cell, RBC) phospholipid content of eight infants with BA who underwent LT was determined 2 months after an initial portoenterostomy, immediately before LT, and 6 and 12 months after LT. Before LT, all infants were fed a protein hydrolysate formula containing medium-chain triglycerides and essential fatty acids. Afterward, they were fed a normal diet for age. The RBC phospholipid content at each time point was compared with that of 28 age-matched control infants. RESULTS: Just before LT, median RBC phospholipid content of C20:4n-6, C20:5n-3, and C22:6n-3 was 25%, 48%, and 30% lower, respectively, than that observed in age-matched control infants. After LT, the RBC phospholipid content of most fatty acids reached normal values by 6 months. However, that of C20:4n-6 and C22:6n-3 contents remained 5% and 15% lower, respectively, than in normal control infants. Twelve months after LT, C20:4n-6 content remained lower than in normal children, but that of C22:6n-3 did not differ. The ratio of C20:3n-6/C20:4n-6, a reflection of delta-5 desaturase activity, was abnormal compared with normal children before LT (0.17 vs. 0.10, P < 0.009) but normalized by 6 months after LT (0.11 vs. 0.10, not significant). CONCLUSIONS: These data show that the abnormal LC-PUFA status of children with BA improves after LT but is not entirely reversed within a year after surgery. They suggest that the abnormal status before LT may be secondary, in part, to low delta-5 desaturase activity. The extent to which a different pre- and/or post-LT diet can prevent PUFA deficiency and/or hasten recovery of PUFA status remains to be determined.

Metabolic fate of an oral tracer dose of [13C]docosahexaenoic acid triglycerides in the rat.

The appearance of 13C in rat lipoprotein, blood cells, and brain lipids was followed as a function of time after the ingestion of triglycerides (TG) containing [13C]22:6n-3. The time course of 13C abundance in 22:6n-3 of various lipid pools, measured by gas chromatography combustion-isotope mass spectrometry, established precursor-product relationships within lipids. The [13C]22:6n-3 was rapidly incorporated into very low density lipoprotein-chylomicron-TG and unesterified fatty acids bound to albumin, with a concomitant maximal appearance at 3 h and further decline. Lysophosphatidylcholines (lysoPC) bound to albumin were also enriched in [13C]22:6n-3, and their labeling appeared to be mainly due to hepatic secretion at the earliest time points. From 12 h postingestion, the synthesis of [13C]22:6n-3-lysoPC was twice as high as that of unesterified [13C]22:6n-3, making lysoPC a potential source of 22:6n-3 supply for tissues. The labeling of platelets, red blood cells, and brain phospholipids presented different kinetics, presumably involving the two lipid forms of [13C]22:6n-3 bound to albumin, to different extents. We conclude that [13C]22:6n-3 esterified in TG is rapidly redistributed within blood lipoproteins and the albumin fraction and that its incorporation in lipid species bound to albumin influences its uptake by target tissues.

In vivo compartmental metabolism of 13C-docosahexaenoic acid, studied by gas chromatography-combustion isotope ratio mass spectrometry.

The exchange of docosahexaenoic acid (22:6n-3) within lipid pools in rat and human has been followed as a function of time after the ingestion of triglycerides (TG) containing 22:6n-3 labeled with 13C(13C 22:6n-3). The 13C abundance in the fatty acid was measured by gas-chromatography-combustion isotope ratio mass spectrometry which allowed the detection of 0.001 atom 13C percent 12C. The 13C 22:6n-3 appearance was rapid in the TG of very low density lipoprotein plus chylomicron fraction, in which the maximal labeling was observed at 3 and 2 h after ingestion in rat and human, respectively. Concomitant with the TG utilization of this fraction by lipoprotein lipase from tissues, unesterified 13C 22:6n-3 appeared in the plasma albumin. 13C 22:6n-3 bound to albumin was mostly present in unesterified form before 12 h post-ingestion while after that period, lysophosphatidylcholine (lysoPC) bound to albumin carried higher 13C 22:6n-3 concentrations. These lyso-PC were mostly from hepatic origin and might represent a potential source of 22:6n-3 redistribution to tissues. The 13C 22:6n-3 uptake into rat brain PC and phosphatidylethanolamine was still increasing when the concentration of plasma unesterified 13C 22:6n-3 had already dropped to a minimal plateau value and during the period of maximal plasma circulation of 13C 22:6n-3-lysoPC bound to albumin. In contrast, the uptake of 13C 22:6n-3 into blood platelet PC occurred during the phase of important circulation of 13C-22:6n-3 bound to albumin, suggesting the in vivo efficiency of the Lands pathway for this fatty acid. It is concluded that 13C 22:6n-3 esterified in TG is rapidly absorbed and redistributed within plasma lipoproteins and that its redistribution within the two lipid species bound to albumin might influence its uptake by platelets and rat brain.

Selective modifications of the phospholipid fatty acid composition in human platelet membranes using nonspecific and specific lipid transfer proteins.

In order to specifically modify the fatty acid composition of cell membrane phospholipids, we have developed an original method based on the transfer of pure phospholipid molecular species to membranes. Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) subclasses containing 18:2n-6 and 22:6n-3 at the sn-2 position were incorporated into human platelet membranes using the endogenous phosphatidylinositol/PC transfer protein (PI/PC-TP) and the phospholipid transfer protein from maize (L-TP), respectively. PI/PC-TP was shown to catalyze a strict exchange of phospholipids between platelet membranes and unilamellar vesicles containing 1,2-diacylglycerophosphocholine (diacyl-GPC; 16:0/18:2-GPC, or 16:0/22:6-GPC). The proportions of 18:2n-6 and 22:6n-3 in diacyl-GPC of platelet membranes were gradually increased from 10.7 to 16.9% and from 0.8 to 10.1%, respectively, whereas the PE and PI fatty acid compositions were not changed. The diacyl-GPC enrichment in 22:6n-3 and 18:2n-6 did not induce changes in membrane fluidity parameters measured by electron-spin resonance of 5- and 16-nitroxy stearic acids. Similarly, 18:2n-6 and 22:6n-3 esterified in 1,2-diacylglycerophosphoethanolamine (diacyl-GPE) have been incorporated in platelet membranes by an apparent exchange process under conditions where donor vesicles had a phospholipid composition equivalent to that of platelet membranes. The proportions of 18:2n-6 and 22:6n-3 were selectively and progressively increased from 6.0 to 21.2% and from 2.2 to 17.2%, respectively, in diacyl-GPE of platelet membranes. Thus, the L-TP- and PI/PC-TP-catalyzed enrichment can be used for studying the modulation of membrane biological activities by defined changes of fatty acid composition of specific phospholipid classes or subclasses.

Effect of specific phospholipid molecular species incorporated in human platelet membranes on thromboxane A2/prostaglandin H2 receptors.

The incorporation of albumin-bound docosahexaenoic acid (22:6n-3), but not linoleic acid (18:2n-6), into cellular phospholipids inhibits platelet aggregation induced by the thromboxane analogue U46619. [3H]U46619 specific binding to thromboxane A2/prostaglandin H2 (TXA2/PGH2) receptors, as well as specific binding of the antagonist [3H]SQ29548 to these sites were also decreased in these modified cells (P. G., Swann et al. 1990. J. Biol. Chem. 265: 21692-21697). More than 80% of the 22:6n-3 incorporated in these cells was esterified in the various endogenous phospholipid classes and the remaining was found in neutral lipids and in the unesterified fatty acid pool. In this study, we determined whether the effects observed could be attributed to the esterification of 22:6n-3 in phospholipids and whether the 22:6n-3 biological activity might depend on its esterification in specific phospholipid classes. Therefore, pure phosphatidylcholine (PC) and phosphatidylethanolamine (PE) molecular species were transferred to platelet membranes, using lipid transfer proteins. PC and PE containing palmitate (16:0) and 22:6n-3 or 16:0 and 18:2n-6 at position sn-1 and sn-2, respectively, were incorporated into membranes only at the expense of the corresponding endogenous phospholipid class, by an apparent exchange process. When such modified membranes were tested for specific binding of U46619 and SQ29548, a significant decrease of the receptor site affinity was only observed in membranes highly enriched with 1-palmitoyl-2-docosahexaenoyl-glycerophosphocholine (16:0/22:6-GPC). Fluidity parameters measured by electron spin resonance of 5- and 16-nitroxy-stearic acids were not significantly different in membranes enriched with 16:0/22:6-GPC relative to those enriched with 16:0/18:2n-6-GPC, arguing against a generalized perturbation of the membrane due to 22:6n-3 incorporation. We conclude that molecular species of PC with 22:6n-3 at the sn-2 position can affect TXA2/PGH2 receptors. The selectivity of the inhibitory effect of PC containing 22:6n-3 is discussed.

Red blood cell fatty acid composition in low-birth-weight infants fed either human milk or formula during the first months of life.

The fatty acid composition of red blood cell (RBC) phospholipids in low-birth-weight infants was determined immediately after delivery and during the first 3 months of life. In the first study, infants were fed either human milk or two formulas with different fatty acid compositions but no long chain polyunsaturated fatty acids (LCPUFA). Both groups of formula-fed infants had significantly lower levels of docosahexaenoic acid (DHA) in RBC phospholipids compared with breast-fed infants. RBC phospholipid DHA was similar in the two formula groups at all ages. In the second study, infants received either a non-supplemented or a LCPUFA-supplemented formula. DHA remained stable in RBC phospholipids of infants supplemented with LCPUFA, whereas DHA decreased in RBC phospholipids of unsupplemented infants. These results confirm that adding DHA to formulas is more effective than increasing 18:3 n-3 content, in maintaining RBC phospholipid DHA levels.

Stable isotope tracer and gas-chromatography combustion isotope ratio mass spectrometry to study the in vivo compartmental metabolism of docosahexaenoic acid.

A gas-chromatography combustion isotope ratio mass spectrometry (GCC-IRMS) method using carbon 13 (13C)-stable isotope to trace n-3 polyunsaturated fatty acids (PUFA) turnover in vivo is presented. Natural 13C abundance of commercial n-3 PUFA was measured from 100 to 300 ng of fatty acids and was -27.58, -27.83, and -28.16 for 22:6n-3, 22:5n-3, and 20:5n-3, expressed as delta 13C /1000 versus Pee Dee Belemnite (PDB), respectively. Precision of delta 13C /1000 values was comparable for the three PUFA and gave relative standard deviations of 0.95-0.97%. Isotope enrichment of 0.0010 at.% could be detected. Triglycerides enriched in [13C]22:6n-3 ([13C]22:6-TG) were synthesized by growing a microalgae on [1-13C]glucose. [13C]22:6n-3 represented 36 wt.% of total triglyceride fatty acids and had an isotope enrichment of 2.0420 at.%, which was the double of natural abundance. The isotope enrichment of 22:6n-3 in lipids from rat lipoproteins and red cells could be followed as a function of time after ingestion of 3 mg [13C]22:6-TG and showed specific patterns according to the lipid compartments. The retroconversion of [13C]22:6n-3 was also detected in HDL phosphatidylcholine by the appearance of [13C]22:5n-3 and [13C]20:5n-3. On the other hand, 22:6n-3 natural 13C abundance in human lipid classes of lipoproteins and blood cells has been measured using 10 ml plasma, even for the more limiting lipid compartments in terms of 22:6n-3 dose size.(ABSTRACT TRUNCATED AT 250 WORDS)

Phospholipid molecular species from human placenta lipids.

The phospholipid molecular species from a large-scale preparation of human placenta lipids were analyzed. The major placental phospholipids were choline glycerophospholipids (CPL) (53.2 wt%), sphingomyelin (21.7 wt%) and ethanolamine glycerophospholipids (EPL) (14.6 wt%). 1,2-Diacyl-glycerophosphocholine was the most abundant subclass of CPL (91.7 mol%), while EPL contained 1,2-diacyl (54.6 mol%) and 1-alk-1'-enyl-2-acyl (43.8 mol%) subclasses. The level of polyunsaturated fatty acids (PUFA) in total phospholipids was remarkably constant (38.4-39.9 mol%) within all placental batches tested. The long-chain PUFA, mainly 20:4n-6 and 22:6n-3 of the n-6 and n-3 series, respectively, were found in high proportion in all phospholipid classes, especially in EPL (46.7 mol%) and in inositol glycerophospholipids (IPL) (39.9 mol%). CPL and serine glycerophospholipids were much richer in 18:1n-9 and 18:2n-6. High levels of molecular species with arachidonic acid in the sn-2 position were found particularly in 1-alk-1'-enyl-2-acyl-glycerophosphoethanolamine (with 24.0 mol% 16:0 and 22.0 mol% 18:0 in sn-1 position) and in 1,2-diacyl glycerophosphoinositol with 42.6 mol% 18:0 in sn-1 position. EPL subclasses were rich in 22:6n-3, which occurs mainly as 16:0/22:6n-3 (11.7 mol%) in the plasmalogen form and as 18:0/22:6n-3, 16:0/22:6n-3 and 18:1/22:6n-3 in the diacyl forms. Based on their availability and composition, placental phospholipids could be of interest, for example, for supplementing artificial milk preparations with n-3 and n-6 long-chain PUFA for newborn infants with insufficiently developed 18:2n-6 and 18:3n-3 desaturation/elongation.