- Email: [email protected]
Docosahexaenoic, arachidonic, palmitic, and oleic acids are differentially esterified into phospholipids of frog retina
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3),105^111 & 2002 Elsevier Science Ltd. All rights reserved. doi:10.1054/plef.406, available online at http://www.idealibrary.com on
Docosahexaenoic, arachidonic, palmitic, and oleic acids are differentially esterified into phospholipids of frog retina R. E. Martin,1,2,3 S. A. Hopkins,2,3 R. Steven Brush,2,3 C. R.Williamson,2,3 H. Chen,2,3 R. E. Anderson1,2,3,4 1
Department of Cell Biology, Oklahoma City, OK, USA Department of Ophthalmology, Oklahoma City, OK, USA 3 University of Oklahoma Health Sciences Center and the Dean McGee Eye Institute, Oklahoma City, OK, USA 4 Department of Biochemistry and Molecular Biology, Oklahoma City, OK, USA 2
Summary Docosahexaenoic acid (22:6n-3) is highly enriched in the retina.To determine if retinal cells take up and metabolize fatty acids in a specific manner, retinas from Rana pipiens were incubated for 3 h with an equimolar mixture of tritiated 22:6n-3, arachidonic acid (20:4n-6), palmitic acid, and oleic acid.The radiolabeling of retinal lipids was determined and compared to the endogenous fatty acid content of the lipids.The results showed that in most, but not all, cases, the relative labeling with the four precursor fatty acids was similar to their relative abundance in each glycerolipid.Thus, during retinal glycerolipid synthesis, either through de novo or acyl exchange reactions, fatty acids are incorporated in proportions reflecting their steady-state mass levels. Since other studies with labeled glycerol have shown greater differences between early labeling patterns and molecular species mass, the final incorporation we report may be due primarily to acyl exchange reactions. & 2002 Elsevier Science Ltd. All rights reserved.
INTRODUCTION Belonging to the omega-3 (n-3) family of essential polyunsaturated fatty acids, docosahexaenoic acid (22:6n-3) is greatly enriched in phosphatidylserine (PS), phosphatidylethanolamine (PE) and phosphatidylcholine (PC) of the highly specialized photoreceptor outer segments of the retina.1–5 22:6n-3 is important in the function of the retina; the transduction of light into an electrical signal is impaired when the content of this fatty Received 15 October 2001 Accepted 3 May 2002 Correspondence to: R. E. Martin, Dean A. McGee Eye Institute, 608 Stanton L. Young Blvd., Rm. 401, Oklahoma City, OK, USA.Tel.: +1-405-271-7366; Fax: +1405-271-8128
Grants: The Foundation Fighting Blindness; Research to Prevent Blindness, Inc.;The Presbyterian Health Foundation;The Dean A. McGee Eye Institute; The National Institutes of Health grants EY04149, EY00871 & EY12190.
& 2002 Elsevier Science Ltd. All rights reserved.
acid is diminished in the photoreceptor cell membrane.6–9 Therefore, maintenance of high retinal levels of 22:6n-3 is important to support optimal function of the retina. Most of the metabolic steps in fatty acid chain elongation and glycerolipid biosynthesis are well characterized.10–18 In glycerolipid synthesis, pre-formed fatty acids are incorporated into their respective sn-1 and sn-2 positions during de novo synthesis or by transacylation reactions after de novo synthesis (retailoring). In de novo glycerolipid synthesis, the 22:6n-3 is first incorporated into phosphatidic acid (PA), which is sequentially converted into diglyceride (DG) and PC or PE. In retailoring reactions, glycerolipids acquire new fatty acids through the consecutive enzymatic actions of phospholipase A2, fatty acylCoA synthase, and an acyl-transferase which act collectively to rearrange the fatty acid distribution within individual glycerolipid classes.13,19 There have been a number of in vivo and in vitro studies concerning the metabolism of 22:6n-3 in the
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
106
Martin et al.
retina. In vivo studies are complicated by the participation (metabolism) of other tissues prior to the delivery of n-3 fatty acids to the retina although in vitro approaches obviate some of these complexities by using cultured retinal cells, intact retina, or eye cups.19–25 We have used intact frog retinas to study the uptake and incorporation of four fatty acids into retinal glycerolipids. A unique aspect to these experiments is that the substrates were presented together as an equimolar mixture with each fatty acid having the same specific activity. Four common fatty acids of different families were utilized: docosahexaenoic acid (22:6n-3), arachidonic acid (20:4n-6), palmitic acid (16:0), and oleic acid (18:1n-9). After 3 h of continuous incubation, the esterification of each fatty acid was compared with that of the ‘normal’ (final) endogenous fatty acid content. Differences between the esterification of the radiolabeled fatty acids (made in part by de novo synthesis) and the endogenous fatty acids hypothetically reflects the retailoring of a given glycerolipid’s fatty acid content.
METHODS
remove unabsorbed fatty acids and this wash procedure was repeated 4 times. Samples of 4 and 16 retinas were kept for processing of total retinal lipids and rod outer segments. Three independent sets of retinas were examined at each substrate concentration (n=3 for each concentration).
Lipid extraction Lipids were extracted in chloroform:methanol (2:1 v:v) following the methods of Folch et al.26 Proteins in the crude lipid extracts were pelleted by centrifugation at 1000 g for 5 min and washed with 2 volumes of chloroform:methanol. The wash was combined with the crude lipid extract. Water containing 0.33 mM diethylenetriaminepentaacetic acid (DTPA) was added and the organic layer was removed. The aqueous layer was then re-extracted with Folch theoretical upper phase (C:M:DTPA; 3:48:47). This two-phase system was capped under N2, vortexed and centrifuged for 10 min at 1000 g. The upper aqueous phase was discarded leaving the purified lipid extract which was stored under N2 in a known volume of C:M (2:1, v:v).
Animals/retinal incubations Animal experimentation adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Medium-sized Rana pipiens purchased from J.M. Hazen (Alburg, VT) were housed at room temperature with running water and on a 12 h dark/12 h light diurnal cycle. Once a week, frogs were fed crickets (Fluker ’s Cricket Farm, Baton Rouge, LA). The animals were maintained under these conditions for 1–2 months before use. Eye-cups from 12 h dark-adapted frogs were prepared under dim red light and placed in Ca2+-free Ringer’s solution for 10–20 min at room temperature. Retinas (20) were dissected free and placed in 5 ml of preoxygenated, Ca2+-containing (2.5 mM) Ringer’s solution for 15 min. The 3 h incubations started with the addition of the fatty acid mix to the Ringer’s solution and placement of the tube in a 251C shaking incubator. Radiolabeled fatty acids (NEN Life Sciences; Boston, MA) were conjugated with bovine serum albumin at a ratio of 2 moles of fatty acid/mole BSA and diluted into 50 mM NaHCO3. [4, 5-3H]22:6n-3, [9, 10-3H]16:0, [9, 10-3H]18:1, and [5, 6, 8, 9, 11, 12, 14, 15-3H]20:4n-6 were added in an equimolar, equal specific activity mixture. Each fatty acid was 0.8 mM for one set of incubations and 8.0 mM for another set. The specific activity for each fatty acid was 5 mCi/nmol for the 0.8 mM incubations and 0.5 mCi/nmol for the 8.0 mM incubations. Placing the flasks on ice for 10 min terminated the incubations. The retinas were then transferred to a Petri dish containing 0.1 M BSA in icecold Ringers solution. The retinas were gently swirled to
Thin-layer chromatography (TLC) Radioactivity incorporated into specific lipids was determined using HL-high performance thin-layer chromatography plates (HPTLC; Analtech, Newark, DE) and the two-dimensional, 3-solvent system method described previously.27 The plates were saturated with 3% magnesium acetate and activated for 2 h at 1101C. Lipid spots were localized with iodine vapors for the determination of radioactivity in each lipid class and with dichlorofluorescein for HPLC analysis. Iodine-stained plates were destained and the silica gel from each spot was analyzed for radioactivity with scintillation counting. Dichlorofluorescein-stained plates were scraped and fatty acids from individual lipid spots were derivitized for highpressure liquid chromatography (HPLC).
Fatty acid derivatization and high-pressure liquid chromatography Fatty acids were saponified and subsequently converted to phenacyl esters.28 These were analyzed by reversephase HPLC with in-line radiochemical detection (IN/US Systems Inc., Tampa, FL) and UV detection (242 nm) using a 5 mm, 5 mm 250 mm, Supelcosils C-18 reverse-phase column (Supelco, Bellefonte, PA). Elution with 80:20acetonitrile (AcN):H2O was isocratic for 30 min, followed by an increase to 100% by 70 min. Elution was again isocratic for 10 min and then returned to 80:20-AcN:H2O by 85 min for re-equilibration of the column.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 107
RESULTS
Fatty acid accumulation in lipids of frog retina Continuous 3 h incubations of frog retinas at either of the substrate concentrations (8 or 0.8 mM) led to similar incorporation of each fatty acid into the retinal glycerolipids. There was, however, relatively less incorporation of [3H]20:4n-6 and [3H]22:6n-3, especially at the lower substrate concentration (Fig. 1, inset). The combined relative esterification of the four radiolabeled fatty acids into specific lipids was unevenly distributed (Fig. 1). Most of the radioactivity was in PC and PE at 37 and 27%, respectively, followed by PS, DG, triglycerides (TG), and free fatty acids (FFA) at 5–10% each. The balance of radioactivity was distributed between lysophospholipids, cholesterol esters, and sphingomyelin. Labeling of PS was higher in 0.8 mM incubations than in 8.0 mM incubations. The inverse was true for TG labeling. HPLC analysis of
radiolabeled fatty acids showed that the substrate fatty acids were metabolized to some degree. Radiolabeled peaks for 20:3n-3, 22:5n-3, 18:2n-6, 18:3n-6, 14:0, and 18:0 were identified (data not shown). The distribution of the four fatty acid substrates into individual retinal glycerolipids was determined by HPLC analysis of the individual lipid spots on TLC plates. The results were found to be unique for each lipid class. These data are expressed in three ways (Figs 2–4). In the first analysis, the relative percentage contribution of each lipid class to the total esterification of each substrate fatty acid was calculated (Fig. 2). Most of the [3H]22:6n-3 that was taken up and esterified into retinal lipids was found in PE, PC, and PS, with minimal esterification into PI, TG, and DG. A similar distribution was seen for [3H]20:4n-6, with the exception of a tendency for increased esterification into PI relative to that of [3H]22:6n-3. The profile for [3H]16:0 was quite different; most of the label (78%) was
Fig. 1 Concentration dependence of lipid total labeling. Intact frog retinas were dissected free of pigment epithelium and incubated for 3 h with equimolar mixtures of [3H]22:6n-3, [3H]20:4, [3H]16:0, and [3H]18:1n-9 (0.8 mM, 5 mCi/nmol or 8.0 mM, 0.5 mCi/nmol final concentration).The radioactivity for all four fatty acids was summed for each lipid class and the relative lipid labeling was determined.Values are mean7standard deviation, n=3. Lysophosphatidylinositol (LPI), phosphatidylserine (PS), phosphatidylinositol (PI), lysophosphatidylethanolamine (LPE), lysophosphatidylcholine (LPC), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), monoglyceride (MG), diglyceride (DG), triglyceride (TG), cholesterol ester (CE).
Fig. 2 Distribution of radiolabeled fattyacidsbetween retinallipids.Intact frogretinaswere treated as described for Figure1.The relative labeling of each precursor fatty acid among all lipid classes was determined.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mMincubations).Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
& 2002 Elsevier Science Ltd. All rights reserved.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
108
Martin et al.
Fig. 3 Enrichment of radiolabeled fatty acids in individual retinal lipids. Intact frog retinas were treated as described for Figure1.The relative labeling ofthe four precursor fattyacidsin eachlipid classwas determined.Values are mean7standard deviation of sixexperiments (three 0.8 mM and three 8.0 mM incubations). Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
Fig. 4 Fattyacid abundance in lipidsrelative to fattyacid abundance in the FFA pool.Intact frogretinaswere treated as described for Figure1.The radioactivity for each of the four fatty acids in each lipid class was divided by the radioactivity of that fatty acid in the free fatty acid pool.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mM incubations). Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
found in PC, with 14% in PI. Esterification of [3H]18:1 was also unique in that 48% was esterified into PC and 30% in PE. Of the four substrate fatty acids in the retinal FFA pool, [3H]22:6n-3 was least abundant and [3H]18:1 was most abundant (1 and 10%, respectively, of the total for each fatty acid). The second approach (Fig. 3) shows the relative distribution of radioactivity between the four fatty acids within each lipid class. This presentation is different from that given in Figure 1 because it demonstrates a selectivity of fatty acid incorporation that is not influenced by the relative abundance of a given lipid in the cell. The results show that when retinal cells were presented with equimolar mixtures of the four fatty acids, certain fatty acids were preferentially enriched into each glycerolipid class. For example, [3H]22:6n-3 was most
enriched in PS (42%) and least enriched in PC (10%). The relative distribution of [3H]16:0 was quite different than that of [3H]20:4n-6 and [3H]22:6n-3. [3H]16:0 was most enriched in PC (55%), followed by PI (39%), DG (35%), TG (16%), and PE (10%). The esterification of [3H]18:1n-9 and [3H]16:0 was similar. Both were scarce in PS relative to [3H]22:6n-3 and [3H]20:4n-6, but comparatively enriched in PC and DG. [3H]18:1n-9 was unique among the four in that it was enriched in PE, DG, and TG (47, 45, and 52%, respectively). Our finding that the distribution of the four substrates differed in individual glycerolipid classes was expected, but it was not hypothesized that the substrate concentrations in the FFA pool would be different from those in the media. Rather, we expected that the intracellular pool would equilibrate with the substrate pool in
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 109
the media. This clearly did not occur, as evidenced by the relative radioactivity distributions presented in Figures 2 and 3. Since the FFA pool size ultimately determines the availability of substrate for esterification, we calculated the esterification into glycerolipids as a function of the substrate available in the FFA pool. To generate the results presented in Figure 4, the radioactivity of each fatty acid in each lipid class was divided by the radioactivity of that fatty acid in the FFA pool. This calculation demonstrates the relative sizes of the esterified and unesterified pools for each of the four fatty acids. Compared to the three major phospholipids, there was less esterification of the four substrates into PI, TG, and DG. Relative to the FFA pool, there was 15 times more [3H]22:6n-3 in PE and 10 times more in PC, compared to 3 times more [3H]16:0 in PE and 3 times more [3H]20:4n-6 in PC. The abundance of [3H]22:6n-3 in PS and [3H]20:4n-6 in PS was
approximately 6–7 times that of the corresponding substrate in the FFA pool.
Radiolabeling vs endogenous fatty acid content of frog retinal lipids Figure 5 compares the relative incorporation of radiolabeled fatty acids with their steady-state mass levels. The values were calculated as in Figure 3, where the labeling and/or molecular mass of the four substrate fatty acids was determined for each glycerolipid class and the percent contribution of each was determined. With a few exceptions, the relative incorporation of radioactivity was similar to the relative mass levels of the fatty acids in the glycerolipids. The [3H]22:6n-3 content was less than the endogenous levels in PI and DG, whereas it was greater than endogenous in PS and PE. The 20:4n-6 contents differed only in PS, where the endogenous level
Fig. 5 Radiolabeled fatty acid enrichment in lipids relative to endogenous fatty acid enrichment.Intact frog retinas were treated as described for Figure1.The relative radioactivity for each of the four fatty acids in each lipid (as in Fig. 3) and the relative mole percent of each fatty acid was determined.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mM incubations).
& 2002 Elsevier Science Ltd. All rights reserved.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
110
Martin et al.
was significantly less than the [3H]20:4n-6. The relative mass level of 16:0 was greater than the [3H]16:0 level in PI, TG, and DG. Conversely, the [3H]18:1n-9 content was more enriched in all glycerolipid classes except PS. DISCUSSION AND CONCLUSIONS In this study, we determine the relative incorporation of four major retinal fatty acids into retinal glyceral lipid classes, using frog retinas dissected and incubated with an equal molar mixture of the precursors. As one might predict, most of the radioactivity from all fatty acids was incorporated into phosphatidylcholine and phosphatidylethanolamine, followed by phosphatidylserine; these are the three major glycerolipid classes in the retina. Our purpose in incubating the mixture of fatty acids was to determine if the relative incorporation of label mimicked that of the relative mass percentage of these fatty acids. The comparison presented in Figure 5 shows some differences but there is still a remarkable similarity between relative incorporation and steady-state mass levels. This is in contrast to what we had previously observed when rat retinas were incubated with radiolabeled glycerol, or when frogs were intravitreally injected with radioactive glycerol and incorporation into rod outer segment phospholipids was determined as a function of time.23,29 In the two glycerol studies, the labeling pattern did not resemble that of the molecular species composition. These studies suggested that the distinct molecular species pattern observed in retinal phospholipids was not established at the time of de novo synthesis, but rather resulted from deacylation–reacylation (tailoring) reactions. Thus, the labeling pattern we observed in the present study might suggest that the bulk of the precursor fatty acids were incorporated into pre-formed lipids in a retailoring reaction, rather than via de novo synthesis. Unlike the media, the steady-state labeling of the cellular-free fatty acid pool reflected significant differences in the abundance of the four fatty acid precursors. This could result from differential uptake of fatty acids from the media or from different rates of incorporation into retinal glycerol lipids. Since the labeling patterns were different for the four precursors, we compared their incorporation into retinal phospholipids by three different means. In the first, the relative distribution of radioactivity in each individual precursor was determined for each lipid class. Using this approach, the two polyunsaturated fatty acids (22:6n-3 and 20:4n-6) were incorporated primarily into the major phospholipid classes (PS, PC, and PE), while palmitic acid was incorporated to a great extent (almost 80%) into PC, with very small incorporation into PE and PS. This distribution of radioactivity reflects the relative mass distribution of
palmitic acid in these glycerolipid classes. When the relative incorporation of radioactivity of the four fatty acids within each glycerolipid class was determined (Fig. 3), the two polyunsaturated fatty acids were incorporated to the greatest extent into PS. The high levels of palmitic acid incorporation were in PI and PC. Interestingly, larger relative amounts of 18:1n-9 were incorporated into most of the lipid classes than predicted on the basis of its relative mass amount (see Fig. 5). We have noticed this in other studies which utilized radioactive glycerol as lipid precursor.23,29 In rat retinas incubated with labeled glycerol for 2 h, 10% of the radioactivity was in 18:122:6 molecular species, even though the mass level of the species was 2%.23 These results suggest that oleic acid is rapidly incorporated into retinal glycerolipids. Since its steady-state mass is relatively small compared to 22:6n16:0, oleic acid must turn over more rapidly in these glycerolipids in order to maintain low steady-state composition. In summary, we have found that incubation of a mixture of four common fatty acids found in retinal glycerolipids leads to labeling patterns that resemble the steady-state mass composition. This suggests that the enzymes responsible for the glycerolipid biosynthesis show great specificity for substrate fatty acids as well as the acceptor glycerolipid.
REFERENCES 1. Anderson R. E., Maude M. B. Phospholipids of bovine outer segments. Biochemistry 1970; 9: 3624–3628. 2. Aveldan˜o M. I., Bazan N. G. Fatty acid composition and level of diacylglycerols and phosphoglycerides in brain and retina. Biochim Biophys Acta 1973; 296: 1–9. 3. Aveldan˜o M. I., Bazan N. G. Molecular species of phosphatidylcholine, -ethanolamine, -serine, and -inositol in microsomal and photoreceptor membranes of bovine retina. J Lipid Res 1983; 24: 620–627. 4. Fliesler S. J., Anderson R. E. Chemistry and metabolism of lipids in the vertebrate retina. In: Holman R. T., ed. Progress in Lipid Research, Vol. 22. London: Pergamon Press, 1983; 79–131. 5. Choe H. G., Anderson R. E. Unique molecular species composition of glycerolipids of frog rod outer segments. Exp Eye Res 1990; 51: 159–165. 6. Benolken R. M., Anderson R. E., Wheeler T. G. Membrane fatty acids associated with the electrical response in visual excitation. Science 1973; 182: 1253–1254. 7. Wheeler T. G., Benolken R. M., Anderson R. E. Visual membrane: specificity of fatty acid precursors for the electrical response to illumination. Science 1975; 188: 1312–1314. 8. Bourre J., Francois M., Youyou A., Dumont O., Piciotti M., Pascal G. et al. The effects of dietary g-linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, resistance poisons and performance of learning tasks in rats. J Nutr 1989; 119: 1880–1892. 9. Watanabe I., Kato M., Aonuma H., Hishimoto A., Naito Y., Moriuchi A. et al. Effect of dietary alpha-linolenate/linoleate
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 111
10. 11.
12. 13.
14.
15.
16.
17.
18. 19.
balance on the lipid composition and electroretinographic responses in rats. In: Zrenner E., Krastel H., Goebel H. H., eds. Advances in the Biosciences, Research in Retinitis Pigmentosa, Vol. 62. Oxford: Pergamon Journals Ltd, 1987, 563–570. Trewhella M. A., Collins F. D. Pathways of phosphatidylcholine biosynthesis in rat liver. Biochim Biophys Acta 1973; 296: 51–61. Sundler R., Akesson B. Regulation of phospholipid biosynthesis in isolated rat hepatocytes. Effect of different substrates. J Biol Chem 1975; 250: 3359–3367. Bazan N. G. Metabolism of phospholipids in the retina. Vision Res 1982; 22: 1539–1548. Birkle D. L., Bazan N. G. The arachidonic acid cascade and phospholipid and docosahexaenoic acid metabolism in the retina. In: Osborne N., Chader J., eds. Progress in Retinal Research, Vol. 5. Oxford: Pergamon, 1986; 309–335. Vance D. E. Phospholipid metabolism and cell signaling in eucaryotes. In: Vance D. E., Vance J. eds. Biochemistry of Lipids, Lipoproteins and Membranes, New Comprehensive Biochemistry, Vol. 20. New York: Elsevier, 1991; 205–240. Voss A., Reinhart M., Sankarappa S., Sprecher H. The metabolism of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,14,16,19docosahexaenoic acid in rat liver is independent of a 4-desaturase. J Biol Chem 1991; 266: 19 995–20 000. Luthria D. L., Mohammed B. S., Sprecher H. Regulation of the biosynthesis of 4, 7, 10, 13, 16, 19-docosahexaenoic acid. J Biol Chem 1996; 271: 16 020–16 025. Giusto N. M., Castagnet P. I., Ilincheta M. G., Roque M. E., Pasquare S. J. Lipid metabolism in photoreceptor membranes: regulation and mechanisms. Neurochem Res 1997; 22: 445–453. Sprecher H. Metabolism of highly unsaturated n-3 and n-6 fatty acids. Biochim Biophys Acta 2000; 1486: 219–231. Chen H., Anderson R. E. Comparison of uptake and incorporation of docosahexaenoic and arachidonic acids by frog retinas. Curr Eye Res 1993; 12: 851–860.
& 2002 Elsevier Science Ltd. All rights reserved.
20. Scott B. L., Bazan N. G. Membrane docosahexaenoate is supplied to the developing brain and retina by the liver. Proc Natl Acad Sci USA 1989; 86: 2903–2907. 21. Delton-Vandenbroucke I., Grammas P., Anderson R. E. Polyunsaturated fatty acid metabolism in retinal and cerebral microvascular endothelial cells. J Lipid Res 1997; 38: 147–159. 22. Goustard-Langelier B., Alessandri J. M., Raguenez G., Durand G., Courtois Y. Phospholipid incorporation and metabolic conversion of n-3 polyunsaturated fatty acids in the Y79 retinoblastoma cell line. J Neurosci Res 2000; 60: 678–685. 23. Li F., Chen H., Anderson R. E. Biosynthesis of docosahexaenoatecontaining molecular species in the retina. J Mol Neurosci 2001; 16: 205–214. 24. Gordon W. C., Bazan N. G. Docosahexaenoic acid utilization during rod photoreceptor cell renewal. J Neurosci 1990; 10: 2190–2202. 25. Wang N., Anderson R. E. Synthesis of docosahexaenoic acid by retina and retinal pigment epithelium. Biochemistry 1993; 32: 13 703–13 709. 26. Folch J., Lees M., Sloan-Stanley G. H. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 1957; 226: 497–509. 27. Martin R. E. Docosahexaenoic acid decreases phospholipase A2 activity in the neurites/nerve growth cones of PC12 cells. J Neurosci Res 1998; 54: 805–813. 28. Chen H., Anderson R. E. Quantitation of phenacyl esters of retinal fatty acids by high-performance liquid chromatography. J Chromatogr 1992; 578: 124–129. 29. Louie K., Wiegand R. D., Anderson R. E. Docosahexaenoatecontaining molecular species of glycerophospholipids from frog retinal rod outer segments show different rates of biosynthesis and turnover. Biochemistry 1988; 27: 9014–9020.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
Docosahexaenoic, arachidonic, palmitic, and oleic acids are differentially esterified into phospholipids of frog retina R. E. Martin,1,2,3 S. A. Hopkins,2,3 R. Steven Brush,2,3 C. R.Williamson,2,3 H. Chen,2,3 R. E. Anderson1,2,3,4 1
Department of Cell Biology, Oklahoma City, OK, USA Department of Ophthalmology, Oklahoma City, OK, USA 3 University of Oklahoma Health Sciences Center and the Dean McGee Eye Institute, Oklahoma City, OK, USA 4 Department of Biochemistry and Molecular Biology, Oklahoma City, OK, USA 2
Summary Docosahexaenoic acid (22:6n-3) is highly enriched in the retina.To determine if retinal cells take up and metabolize fatty acids in a specific manner, retinas from Rana pipiens were incubated for 3 h with an equimolar mixture of tritiated 22:6n-3, arachidonic acid (20:4n-6), palmitic acid, and oleic acid.The radiolabeling of retinal lipids was determined and compared to the endogenous fatty acid content of the lipids.The results showed that in most, but not all, cases, the relative labeling with the four precursor fatty acids was similar to their relative abundance in each glycerolipid.Thus, during retinal glycerolipid synthesis, either through de novo or acyl exchange reactions, fatty acids are incorporated in proportions reflecting their steady-state mass levels. Since other studies with labeled glycerol have shown greater differences between early labeling patterns and molecular species mass, the final incorporation we report may be due primarily to acyl exchange reactions. & 2002 Elsevier Science Ltd. All rights reserved.
INTRODUCTION Belonging to the omega-3 (n-3) family of essential polyunsaturated fatty acids, docosahexaenoic acid (22:6n-3) is greatly enriched in phosphatidylserine (PS), phosphatidylethanolamine (PE) and phosphatidylcholine (PC) of the highly specialized photoreceptor outer segments of the retina.1–5 22:6n-3 is important in the function of the retina; the transduction of light into an electrical signal is impaired when the content of this fatty Received 15 October 2001 Accepted 3 May 2002 Correspondence to: R. E. Martin, Dean A. McGee Eye Institute, 608 Stanton L. Young Blvd., Rm. 401, Oklahoma City, OK, USA.Tel.: +1-405-271-7366; Fax: +1405-271-8128
Grants: The Foundation Fighting Blindness; Research to Prevent Blindness, Inc.;The Presbyterian Health Foundation;The Dean A. McGee Eye Institute; The National Institutes of Health grants EY04149, EY00871 & EY12190.
& 2002 Elsevier Science Ltd. All rights reserved.
acid is diminished in the photoreceptor cell membrane.6–9 Therefore, maintenance of high retinal levels of 22:6n-3 is important to support optimal function of the retina. Most of the metabolic steps in fatty acid chain elongation and glycerolipid biosynthesis are well characterized.10–18 In glycerolipid synthesis, pre-formed fatty acids are incorporated into their respective sn-1 and sn-2 positions during de novo synthesis or by transacylation reactions after de novo synthesis (retailoring). In de novo glycerolipid synthesis, the 22:6n-3 is first incorporated into phosphatidic acid (PA), which is sequentially converted into diglyceride (DG) and PC or PE. In retailoring reactions, glycerolipids acquire new fatty acids through the consecutive enzymatic actions of phospholipase A2, fatty acylCoA synthase, and an acyl-transferase which act collectively to rearrange the fatty acid distribution within individual glycerolipid classes.13,19 There have been a number of in vivo and in vitro studies concerning the metabolism of 22:6n-3 in the
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
106
Martin et al.
retina. In vivo studies are complicated by the participation (metabolism) of other tissues prior to the delivery of n-3 fatty acids to the retina although in vitro approaches obviate some of these complexities by using cultured retinal cells, intact retina, or eye cups.19–25 We have used intact frog retinas to study the uptake and incorporation of four fatty acids into retinal glycerolipids. A unique aspect to these experiments is that the substrates were presented together as an equimolar mixture with each fatty acid having the same specific activity. Four common fatty acids of different families were utilized: docosahexaenoic acid (22:6n-3), arachidonic acid (20:4n-6), palmitic acid (16:0), and oleic acid (18:1n-9). After 3 h of continuous incubation, the esterification of each fatty acid was compared with that of the ‘normal’ (final) endogenous fatty acid content. Differences between the esterification of the radiolabeled fatty acids (made in part by de novo synthesis) and the endogenous fatty acids hypothetically reflects the retailoring of a given glycerolipid’s fatty acid content.
METHODS
remove unabsorbed fatty acids and this wash procedure was repeated 4 times. Samples of 4 and 16 retinas were kept for processing of total retinal lipids and rod outer segments. Three independent sets of retinas were examined at each substrate concentration (n=3 for each concentration).
Lipid extraction Lipids were extracted in chloroform:methanol (2:1 v:v) following the methods of Folch et al.26 Proteins in the crude lipid extracts were pelleted by centrifugation at 1000 g for 5 min and washed with 2 volumes of chloroform:methanol. The wash was combined with the crude lipid extract. Water containing 0.33 mM diethylenetriaminepentaacetic acid (DTPA) was added and the organic layer was removed. The aqueous layer was then re-extracted with Folch theoretical upper phase (C:M:DTPA; 3:48:47). This two-phase system was capped under N2, vortexed and centrifuged for 10 min at 1000 g. The upper aqueous phase was discarded leaving the purified lipid extract which was stored under N2 in a known volume of C:M (2:1, v:v).
Animals/retinal incubations Animal experimentation adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Medium-sized Rana pipiens purchased from J.M. Hazen (Alburg, VT) were housed at room temperature with running water and on a 12 h dark/12 h light diurnal cycle. Once a week, frogs were fed crickets (Fluker ’s Cricket Farm, Baton Rouge, LA). The animals were maintained under these conditions for 1–2 months before use. Eye-cups from 12 h dark-adapted frogs were prepared under dim red light and placed in Ca2+-free Ringer’s solution for 10–20 min at room temperature. Retinas (20) were dissected free and placed in 5 ml of preoxygenated, Ca2+-containing (2.5 mM) Ringer’s solution for 15 min. The 3 h incubations started with the addition of the fatty acid mix to the Ringer’s solution and placement of the tube in a 251C shaking incubator. Radiolabeled fatty acids (NEN Life Sciences; Boston, MA) were conjugated with bovine serum albumin at a ratio of 2 moles of fatty acid/mole BSA and diluted into 50 mM NaHCO3. [4, 5-3H]22:6n-3, [9, 10-3H]16:0, [9, 10-3H]18:1, and [5, 6, 8, 9, 11, 12, 14, 15-3H]20:4n-6 were added in an equimolar, equal specific activity mixture. Each fatty acid was 0.8 mM for one set of incubations and 8.0 mM for another set. The specific activity for each fatty acid was 5 mCi/nmol for the 0.8 mM incubations and 0.5 mCi/nmol for the 8.0 mM incubations. Placing the flasks on ice for 10 min terminated the incubations. The retinas were then transferred to a Petri dish containing 0.1 M BSA in icecold Ringers solution. The retinas were gently swirled to
Thin-layer chromatography (TLC) Radioactivity incorporated into specific lipids was determined using HL-high performance thin-layer chromatography plates (HPTLC; Analtech, Newark, DE) and the two-dimensional, 3-solvent system method described previously.27 The plates were saturated with 3% magnesium acetate and activated for 2 h at 1101C. Lipid spots were localized with iodine vapors for the determination of radioactivity in each lipid class and with dichlorofluorescein for HPLC analysis. Iodine-stained plates were destained and the silica gel from each spot was analyzed for radioactivity with scintillation counting. Dichlorofluorescein-stained plates were scraped and fatty acids from individual lipid spots were derivitized for highpressure liquid chromatography (HPLC).
Fatty acid derivatization and high-pressure liquid chromatography Fatty acids were saponified and subsequently converted to phenacyl esters.28 These were analyzed by reversephase HPLC with in-line radiochemical detection (IN/US Systems Inc., Tampa, FL) and UV detection (242 nm) using a 5 mm, 5 mm 250 mm, Supelcosils C-18 reverse-phase column (Supelco, Bellefonte, PA). Elution with 80:20acetonitrile (AcN):H2O was isocratic for 30 min, followed by an increase to 100% by 70 min. Elution was again isocratic for 10 min and then returned to 80:20-AcN:H2O by 85 min for re-equilibration of the column.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 107
RESULTS
Fatty acid accumulation in lipids of frog retina Continuous 3 h incubations of frog retinas at either of the substrate concentrations (8 or 0.8 mM) led to similar incorporation of each fatty acid into the retinal glycerolipids. There was, however, relatively less incorporation of [3H]20:4n-6 and [3H]22:6n-3, especially at the lower substrate concentration (Fig. 1, inset). The combined relative esterification of the four radiolabeled fatty acids into specific lipids was unevenly distributed (Fig. 1). Most of the radioactivity was in PC and PE at 37 and 27%, respectively, followed by PS, DG, triglycerides (TG), and free fatty acids (FFA) at 5–10% each. The balance of radioactivity was distributed between lysophospholipids, cholesterol esters, and sphingomyelin. Labeling of PS was higher in 0.8 mM incubations than in 8.0 mM incubations. The inverse was true for TG labeling. HPLC analysis of
radiolabeled fatty acids showed that the substrate fatty acids were metabolized to some degree. Radiolabeled peaks for 20:3n-3, 22:5n-3, 18:2n-6, 18:3n-6, 14:0, and 18:0 were identified (data not shown). The distribution of the four fatty acid substrates into individual retinal glycerolipids was determined by HPLC analysis of the individual lipid spots on TLC plates. The results were found to be unique for each lipid class. These data are expressed in three ways (Figs 2–4). In the first analysis, the relative percentage contribution of each lipid class to the total esterification of each substrate fatty acid was calculated (Fig. 2). Most of the [3H]22:6n-3 that was taken up and esterified into retinal lipids was found in PE, PC, and PS, with minimal esterification into PI, TG, and DG. A similar distribution was seen for [3H]20:4n-6, with the exception of a tendency for increased esterification into PI relative to that of [3H]22:6n-3. The profile for [3H]16:0 was quite different; most of the label (78%) was
Fig. 1 Concentration dependence of lipid total labeling. Intact frog retinas were dissected free of pigment epithelium and incubated for 3 h with equimolar mixtures of [3H]22:6n-3, [3H]20:4, [3H]16:0, and [3H]18:1n-9 (0.8 mM, 5 mCi/nmol or 8.0 mM, 0.5 mCi/nmol final concentration).The radioactivity for all four fatty acids was summed for each lipid class and the relative lipid labeling was determined.Values are mean7standard deviation, n=3. Lysophosphatidylinositol (LPI), phosphatidylserine (PS), phosphatidylinositol (PI), lysophosphatidylethanolamine (LPE), lysophosphatidylcholine (LPC), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), monoglyceride (MG), diglyceride (DG), triglyceride (TG), cholesterol ester (CE).
Fig. 2 Distribution of radiolabeled fattyacidsbetween retinallipids.Intact frogretinaswere treated as described for Figure1.The relative labeling of each precursor fatty acid among all lipid classes was determined.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mMincubations).Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
& 2002 Elsevier Science Ltd. All rights reserved.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
108
Martin et al.
Fig. 3 Enrichment of radiolabeled fatty acids in individual retinal lipids. Intact frog retinas were treated as described for Figure1.The relative labeling ofthe four precursor fattyacidsin eachlipid classwas determined.Values are mean7standard deviation of sixexperiments (three 0.8 mM and three 8.0 mM incubations). Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
Fig. 4 Fattyacid abundance in lipidsrelative to fattyacid abundance in the FFA pool.Intact frogretinaswere treated as described for Figure1.The radioactivity for each of the four fatty acids in each lipid class was divided by the radioactivity of that fatty acid in the free fatty acid pool.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mM incubations). Phosphatidylserine (PS), phosphatidylinositol (PI), phosphatidylcholine (PC), phosphatidylethanolamine (PE), free fatty acids (FFA), triglyceride (TG), diglyceride (DG).
found in PC, with 14% in PI. Esterification of [3H]18:1 was also unique in that 48% was esterified into PC and 30% in PE. Of the four substrate fatty acids in the retinal FFA pool, [3H]22:6n-3 was least abundant and [3H]18:1 was most abundant (1 and 10%, respectively, of the total for each fatty acid). The second approach (Fig. 3) shows the relative distribution of radioactivity between the four fatty acids within each lipid class. This presentation is different from that given in Figure 1 because it demonstrates a selectivity of fatty acid incorporation that is not influenced by the relative abundance of a given lipid in the cell. The results show that when retinal cells were presented with equimolar mixtures of the four fatty acids, certain fatty acids were preferentially enriched into each glycerolipid class. For example, [3H]22:6n-3 was most
enriched in PS (42%) and least enriched in PC (10%). The relative distribution of [3H]16:0 was quite different than that of [3H]20:4n-6 and [3H]22:6n-3. [3H]16:0 was most enriched in PC (55%), followed by PI (39%), DG (35%), TG (16%), and PE (10%). The esterification of [3H]18:1n-9 and [3H]16:0 was similar. Both were scarce in PS relative to [3H]22:6n-3 and [3H]20:4n-6, but comparatively enriched in PC and DG. [3H]18:1n-9 was unique among the four in that it was enriched in PE, DG, and TG (47, 45, and 52%, respectively). Our finding that the distribution of the four substrates differed in individual glycerolipid classes was expected, but it was not hypothesized that the substrate concentrations in the FFA pool would be different from those in the media. Rather, we expected that the intracellular pool would equilibrate with the substrate pool in
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 109
the media. This clearly did not occur, as evidenced by the relative radioactivity distributions presented in Figures 2 and 3. Since the FFA pool size ultimately determines the availability of substrate for esterification, we calculated the esterification into glycerolipids as a function of the substrate available in the FFA pool. To generate the results presented in Figure 4, the radioactivity of each fatty acid in each lipid class was divided by the radioactivity of that fatty acid in the FFA pool. This calculation demonstrates the relative sizes of the esterified and unesterified pools for each of the four fatty acids. Compared to the three major phospholipids, there was less esterification of the four substrates into PI, TG, and DG. Relative to the FFA pool, there was 15 times more [3H]22:6n-3 in PE and 10 times more in PC, compared to 3 times more [3H]16:0 in PE and 3 times more [3H]20:4n-6 in PC. The abundance of [3H]22:6n-3 in PS and [3H]20:4n-6 in PS was
approximately 6–7 times that of the corresponding substrate in the FFA pool.
Radiolabeling vs endogenous fatty acid content of frog retinal lipids Figure 5 compares the relative incorporation of radiolabeled fatty acids with their steady-state mass levels. The values were calculated as in Figure 3, where the labeling and/or molecular mass of the four substrate fatty acids was determined for each glycerolipid class and the percent contribution of each was determined. With a few exceptions, the relative incorporation of radioactivity was similar to the relative mass levels of the fatty acids in the glycerolipids. The [3H]22:6n-3 content was less than the endogenous levels in PI and DG, whereas it was greater than endogenous in PS and PE. The 20:4n-6 contents differed only in PS, where the endogenous level
Fig. 5 Radiolabeled fatty acid enrichment in lipids relative to endogenous fatty acid enrichment.Intact frog retinas were treated as described for Figure1.The relative radioactivity for each of the four fatty acids in each lipid (as in Fig. 3) and the relative mole percent of each fatty acid was determined.Values are mean7standard deviation of six experiments (three 0.8 mM and three 8.0 mM incubations).
& 2002 Elsevier Science Ltd. All rights reserved.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
110
Martin et al.
was significantly less than the [3H]20:4n-6. The relative mass level of 16:0 was greater than the [3H]16:0 level in PI, TG, and DG. Conversely, the [3H]18:1n-9 content was more enriched in all glycerolipid classes except PS. DISCUSSION AND CONCLUSIONS In this study, we determine the relative incorporation of four major retinal fatty acids into retinal glyceral lipid classes, using frog retinas dissected and incubated with an equal molar mixture of the precursors. As one might predict, most of the radioactivity from all fatty acids was incorporated into phosphatidylcholine and phosphatidylethanolamine, followed by phosphatidylserine; these are the three major glycerolipid classes in the retina. Our purpose in incubating the mixture of fatty acids was to determine if the relative incorporation of label mimicked that of the relative mass percentage of these fatty acids. The comparison presented in Figure 5 shows some differences but there is still a remarkable similarity between relative incorporation and steady-state mass levels. This is in contrast to what we had previously observed when rat retinas were incubated with radiolabeled glycerol, or when frogs were intravitreally injected with radioactive glycerol and incorporation into rod outer segment phospholipids was determined as a function of time.23,29 In the two glycerol studies, the labeling pattern did not resemble that of the molecular species composition. These studies suggested that the distinct molecular species pattern observed in retinal phospholipids was not established at the time of de novo synthesis, but rather resulted from deacylation–reacylation (tailoring) reactions. Thus, the labeling pattern we observed in the present study might suggest that the bulk of the precursor fatty acids were incorporated into pre-formed lipids in a retailoring reaction, rather than via de novo synthesis. Unlike the media, the steady-state labeling of the cellular-free fatty acid pool reflected significant differences in the abundance of the four fatty acid precursors. This could result from differential uptake of fatty acids from the media or from different rates of incorporation into retinal glycerol lipids. Since the labeling patterns were different for the four precursors, we compared their incorporation into retinal phospholipids by three different means. In the first, the relative distribution of radioactivity in each individual precursor was determined for each lipid class. Using this approach, the two polyunsaturated fatty acids (22:6n-3 and 20:4n-6) were incorporated primarily into the major phospholipid classes (PS, PC, and PE), while palmitic acid was incorporated to a great extent (almost 80%) into PC, with very small incorporation into PE and PS. This distribution of radioactivity reflects the relative mass distribution of
palmitic acid in these glycerolipid classes. When the relative incorporation of radioactivity of the four fatty acids within each glycerolipid class was determined (Fig. 3), the two polyunsaturated fatty acids were incorporated to the greatest extent into PS. The high levels of palmitic acid incorporation were in PI and PC. Interestingly, larger relative amounts of 18:1n-9 were incorporated into most of the lipid classes than predicted on the basis of its relative mass amount (see Fig. 5). We have noticed this in other studies which utilized radioactive glycerol as lipid precursor.23,29 In rat retinas incubated with labeled glycerol for 2 h, 10% of the radioactivity was in 18:122:6 molecular species, even though the mass level of the species was 2%.23 These results suggest that oleic acid is rapidly incorporated into retinal glycerolipids. Since its steady-state mass is relatively small compared to 22:6n16:0, oleic acid must turn over more rapidly in these glycerolipids in order to maintain low steady-state composition. In summary, we have found that incubation of a mixture of four common fatty acids found in retinal glycerolipids leads to labeling patterns that resemble the steady-state mass composition. This suggests that the enzymes responsible for the glycerolipid biosynthesis show great specificity for substrate fatty acids as well as the acceptor glycerolipid.
REFERENCES 1. Anderson R. E., Maude M. B. Phospholipids of bovine outer segments. Biochemistry 1970; 9: 3624–3628. 2. Aveldan˜o M. I., Bazan N. G. Fatty acid composition and level of diacylglycerols and phosphoglycerides in brain and retina. Biochim Biophys Acta 1973; 296: 1–9. 3. Aveldan˜o M. I., Bazan N. G. Molecular species of phosphatidylcholine, -ethanolamine, -serine, and -inositol in microsomal and photoreceptor membranes of bovine retina. J Lipid Res 1983; 24: 620–627. 4. Fliesler S. J., Anderson R. E. Chemistry and metabolism of lipids in the vertebrate retina. In: Holman R. T., ed. Progress in Lipid Research, Vol. 22. London: Pergamon Press, 1983; 79–131. 5. Choe H. G., Anderson R. E. Unique molecular species composition of glycerolipids of frog rod outer segments. Exp Eye Res 1990; 51: 159–165. 6. Benolken R. M., Anderson R. E., Wheeler T. G. Membrane fatty acids associated with the electrical response in visual excitation. Science 1973; 182: 1253–1254. 7. Wheeler T. G., Benolken R. M., Anderson R. E. Visual membrane: specificity of fatty acid precursors for the electrical response to illumination. Science 1975; 188: 1312–1314. 8. Bourre J., Francois M., Youyou A., Dumont O., Piciotti M., Pascal G. et al. The effects of dietary g-linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, resistance poisons and performance of learning tasks in rats. J Nutr 1989; 119: 1880–1892. 9. Watanabe I., Kato M., Aonuma H., Hishimoto A., Naito Y., Moriuchi A. et al. Effect of dietary alpha-linolenate/linoleate
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111
& 2002 Elsevier Science Ltd. All rights reserved.
Phospholipids of frog retina 111
10. 11.
12. 13.
14.
15.
16.
17.
18. 19.
balance on the lipid composition and electroretinographic responses in rats. In: Zrenner E., Krastel H., Goebel H. H., eds. Advances in the Biosciences, Research in Retinitis Pigmentosa, Vol. 62. Oxford: Pergamon Journals Ltd, 1987, 563–570. Trewhella M. A., Collins F. D. Pathways of phosphatidylcholine biosynthesis in rat liver. Biochim Biophys Acta 1973; 296: 51–61. Sundler R., Akesson B. Regulation of phospholipid biosynthesis in isolated rat hepatocytes. Effect of different substrates. J Biol Chem 1975; 250: 3359–3367. Bazan N. G. Metabolism of phospholipids in the retina. Vision Res 1982; 22: 1539–1548. Birkle D. L., Bazan N. G. The arachidonic acid cascade and phospholipid and docosahexaenoic acid metabolism in the retina. In: Osborne N., Chader J., eds. Progress in Retinal Research, Vol. 5. Oxford: Pergamon, 1986; 309–335. Vance D. E. Phospholipid metabolism and cell signaling in eucaryotes. In: Vance D. E., Vance J. eds. Biochemistry of Lipids, Lipoproteins and Membranes, New Comprehensive Biochemistry, Vol. 20. New York: Elsevier, 1991; 205–240. Voss A., Reinhart M., Sankarappa S., Sprecher H. The metabolism of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,14,16,19docosahexaenoic acid in rat liver is independent of a 4-desaturase. J Biol Chem 1991; 266: 19 995–20 000. Luthria D. L., Mohammed B. S., Sprecher H. Regulation of the biosynthesis of 4, 7, 10, 13, 16, 19-docosahexaenoic acid. J Biol Chem 1996; 271: 16 020–16 025. Giusto N. M., Castagnet P. I., Ilincheta M. G., Roque M. E., Pasquare S. J. Lipid metabolism in photoreceptor membranes: regulation and mechanisms. Neurochem Res 1997; 22: 445–453. Sprecher H. Metabolism of highly unsaturated n-3 and n-6 fatty acids. Biochim Biophys Acta 2000; 1486: 219–231. Chen H., Anderson R. E. Comparison of uptake and incorporation of docosahexaenoic and arachidonic acids by frog retinas. Curr Eye Res 1993; 12: 851–860.
& 2002 Elsevier Science Ltd. All rights reserved.
20. Scott B. L., Bazan N. G. Membrane docosahexaenoate is supplied to the developing brain and retina by the liver. Proc Natl Acad Sci USA 1989; 86: 2903–2907. 21. Delton-Vandenbroucke I., Grammas P., Anderson R. E. Polyunsaturated fatty acid metabolism in retinal and cerebral microvascular endothelial cells. J Lipid Res 1997; 38: 147–159. 22. Goustard-Langelier B., Alessandri J. M., Raguenez G., Durand G., Courtois Y. Phospholipid incorporation and metabolic conversion of n-3 polyunsaturated fatty acids in the Y79 retinoblastoma cell line. J Neurosci Res 2000; 60: 678–685. 23. Li F., Chen H., Anderson R. E. Biosynthesis of docosahexaenoatecontaining molecular species in the retina. J Mol Neurosci 2001; 16: 205–214. 24. Gordon W. C., Bazan N. G. Docosahexaenoic acid utilization during rod photoreceptor cell renewal. J Neurosci 1990; 10: 2190–2202. 25. Wang N., Anderson R. E. Synthesis of docosahexaenoic acid by retina and retinal pigment epithelium. Biochemistry 1993; 32: 13 703–13 709. 26. Folch J., Lees M., Sloan-Stanley G. H. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 1957; 226: 497–509. 27. Martin R. E. Docosahexaenoic acid decreases phospholipase A2 activity in the neurites/nerve growth cones of PC12 cells. J Neurosci Res 1998; 54: 805–813. 28. Chen H., Anderson R. E. Quantitation of phenacyl esters of retinal fatty acids by high-performance liquid chromatography. J Chromatogr 1992; 578: 124–129. 29. Louie K., Wiegand R. D., Anderson R. E. Docosahexaenoatecontaining molecular species of glycerophospholipids from frog retinal rod outer segments show different rates of biosynthesis and turnover. Biochemistry 1988; 27: 9014–9020.
Prostaglandins, Leukotrienes and Essential FattyAcids (2002) 67(2^3), 105^111