294
Biochimica et Biophysica Acta, 5 5 2 ( 1 9 7 9 ) 2 9 4 - - 3 0 6 © E l s e v i e r / N o r t h - H o l l a n d B i o m e d i c a l Press
BBA 78324
DISATURATED AND DIPOLYUNSATURATED PHOSPHOLIPIDS IN THE BOVINE RETINAL ROD OUTER SEGMENT DISK MEMBRANE
G E O R G E P, M I L J A N I C H *, L A R R Y A. D R A T Z
A. S K L A R
**, D R I N A L. W H I T E and E D W A R D
Division of Natural Sciences, University of California, Santa Cruz, C A 95064 (U.S.A.) (Received October 12th, 1978) Key words: Phospholipid; Acyl chain distribution;Rhodopsin; (Rod outer segment membrane)
Summary Thin-layer chromatography was used to separate the major phospholipid headgroup classes of the rod outer segment disk membrane into subfractions which differ markedly in fatty acid composition. At least 18% of the rod outer segment phosphatidylcholine must contain two saturated fatty acids. Furthermore, two unsaturated fatty acids are found in at least 43% of the phosphatidylserine, 24% of the phosphatidylcholine, and 24% of the phosphatidylethanolamine. The unsaturated acids are predominantly polyunsaturated in all cases. A similar separation, but with less resolution, was achieved with silicic acid column chromatography. The temperature dependence of the polarization of the fluorescence of trans.parinaric acid (9,11,13,15-all-trans-octadecatetraenoic acid) showed that the thermal behavior of aqueous dispersions of the phosphatidylcholine subfractions was consistent with their fatty acid compositions.
Introduction The large degree of molecular heterogeneity of phospholipids in biological membranes is well known. However, the detailed combinations of two fatty acids with each other and of these pairs with particular phospholipid heedgroups are poorly documented. Fractionation and characterizationof these combinations might provide clues for understanding the functional role(s)of * Present address: Department of Biochemistry and Biophysics, University of California, San Francisco, CA 94143, U.S.A. * * Present address:Department of Medicine, Methodist Hospital, Baylor College of Medicine, Houston, TX 77030, U,S.A,
295 this molecular polydispersity and should also suggest more relevant model membrane systems for physical studies. In addition, preparative fractionation of various phospholipid species with different fatty acid compositions from natural sources could provide material for model studies [ 1 ]. In this paper we report the separation, by a one-dimensional thin-layer chromatography (TLC) method, of the major phospholipid classes of the retinal rod outer segment membrane into subfractions which differ in fatty acid composition. Fatty acid analysis reveals the presence of saturated-saturated, saturated-polyunsaturated, and polyunsaturated-polyunsaturated species of outer segment phosphatidylcholine and phosphatidylethanolamine and saturated-polyunsaturated and polyunsaturated-polyunsaturated species of outer segment phosphatidylserine. * A similar separation can be achieved, with somewhat less resolution by silicic acid column chromatography. As a further characterization of this separation, we also show that the temperature dependence of the fluorescence polarization of trans-parinaric acid in aqueous dispersions of outer segment phosphatidylcholine subfractions is consistent with their fatty acid compositions. Materials and Methods
Preparation of rod outer segments. Highly purified bovine outer segments were isolated by the method of Raubach et al. [2] except that the second sucrose density gradient was centrifuged for 90 min and all solutions contained 0.1 mM EDTA and 0.15 mM CaCI~ to minimize oxidative damage [3]. Lipid extraction. Under the conditions employed for column chromatography we have found that retinal forms a Schiff's base with phosphatidylethanolamine and this complexation prevents complete separation of these two components. Conversion of retinal to retinaloxime eliminates this problem. Prior to lipid extraction for column chromatography, the outer segment membrane suspension was mixed with enough 0.1 M hydroxylamine (pH 7) to give a 100-fold molar excess of hydroxylamine over retinal and then exposed to an unfiltered 300 W floodlight while the sample tube was immersed in a large beaker of cold tap water. Retinaloxime appears to be unstable, so column fractionation was commenced immediately subsequent to lipid extraction. If the lipids were to be used solely for TLC the oximation step was unnecessary and was omitted. Total outer segment lipids were extracted by a modification of the procedure of Folch et al. [4]. The chloroform and methanol contained 50 rag/1 butylated hydroxytoluene. In this and in the following methods, the chloroform and methanol were redistilled and were bubbled with argon immediately before use. Whenever feasible, manipulations were performed under an argon atmosphere. Column chromatographic fractionation of phospholipid headgroup classes. The total lipids were first applied to a diethylaminoethyl-cellulose column prepared and eluted essentially by the procedure of Rouser et al. [5]. Elution with chloroform removed butylated hydroxytoluene, cholesterol, retinal, * P o s i t i o n a l a n a l y s e s of the fatty acids in these s u b f r a e t i o n s were n o t p e r f o r m e d . T h e r e f o r e , the order o f f a t t y acid t y p e s used in naming these p h o s p h o l i p l d species is a r b i t r a r y .
296 retinaloxime, vitamin E, and free fatty acids. Elution with 1 : 1 chloroform/ methanol removed phosphatidylcholine and phosphatidylethanolamine and glacial acetic acid removed pure phosphatidylserine. The phosphatidylcholinephosphatidylethanolamine mixture was reduced to dryness by rotary evaporation, redissolved in chloroform and further fractionated on a silicic acid column (approx. 1 g silicic acid per 5 mg phospholipid). Thin-layer chromatography. TLC of outer segment total lipid extract was performed on 5 cm by 20 cm plates coated with a 0.25 mm thick layer of silica gel H (EM Laboratories, Inc.). For identification of the various lipid species, authentic standards were cochromatographed with a few pg of outer segment lipids in chloroform/methanol/water (65 : 25 : 5, v/v) and 50 mg/1 butylated hydroxytoluene. Standards used included bovine brain sphingomyelin (Applied Sciences Laboratories); lysophosphatidylcholine and lysophosphatidylethanol. amine (Miles Laboratories); soybean phosphatidylinositol, dimyristoyl phosphatidylcholine, dipalmitoyl phosphatidylcholine, distearoyl phosphatidylcholine, dipalmitoylethanolamine, N,N-dimethyldipalmitoyl phosphatidylcholine, vitamin E, cholesterol, retinal, retinol, triolein and oleic acid (Sigma Chemical Co.); and highly purified bovine brain phosphatidylserine (a generous gift from Dr. D. Papahadjopoulos, University of California, San Francisco). 1-Palmitoyl-2-docosahexaenoyl phosphatidylcholine and didocosahexaenoyl phosphatidylcholine, synthesized as described below, were also used as standards. Chromatograms were visualized by either exposure to I2 vapor, charring with 9 M H2SO4, or with 0.2% ninhydrin in ethanol. The same TLC system was also used to isolate phospholipid subfractions for fatty acid and phosphorus analyses. Approx. 1.0--1.5 mg of outer segment total lipid in approx. 75 pl chloroform was applied to the TLC plate as a thin band 3 cm long and was developed in an argon-saturated jar. Phosphorus and fatty acid analysis of the subfractions. The plate was dried quickly in a stream of argon a n d sprayed with 0.4% dichlorofluorescein in methanol under argon pressure. Lipid bands were visualized under ultraviolet light, marked with a pencil and quickly scraped with a razor blade into 10-ml screw-capped tubes containing 5/~g butylated hydroxytoluene and 14.7/~g heneicosanoic acid (21 : 0) (Supelco, Inc.) in 100 pl chloroform. Fatty acid methyl esters were formed under argon by a modification of the method of Morrison and Smith [6] and centrifuged at 3000 X g for 10 min to remove suspended silica gel. The fatty acid methyl esters (in the upper layer) were analyzed by gas-liquid chromatography (GLC) using a 6 ft glass column containing 10% SP-2330 on 100/120 Supelcoport (Supelco, Inc.} which was temperature programmed from 130 to 250°C at 4°C/min. Identification of peaks was made by comparison with reference methyl ester mixtures (Supelco, Inc.). The integrated area (obtained with an Autolab model 6300 digital integrator) of each peak was divided by the molecular weight of the corresponding methyl ester to yield relative molar concentrations. The lower phases from the transmethylation procedure were heated at l l 0 ° C for several hours until dry and subjected to phosphorus analysis by the method of Chen et al. [ 7 ] modified for smaller sample sizes. Synthesis of palmitoyl-docosahexaenoyl phosphatidylcholine. 1-Palmitoyl2-docosahexaenoyl phosphatidylcholine was synthesized by the method of
297 Evans and Tinoco [8]. TLC of the synthesized material showed two closely migrating bands. GLC analysis demonstrated that the main band, with a slightly smaller RF, contained a 1 : 1 mol : mol ratio of palmitic and docosahexaenoic acids and the minor band, with a larger Rr, contained all docosahexaenoic acid. Phosphorus analysis showed this minor band to represent approx. 4 mol% of the total phospholipid. This component is didocosahexaenoyl phosphatidylcholine. Polarization of fluorescence of trans-Parinaric acid. The temperature dependence of the polarization of the fluorescence of trans-parinaric acid (9,11,13,15all-trans-octadecatetraenoic acid) in phospholipid dispersions was measured using a Hitachi-Perkin-Elmer MPF-2A spectrofluorometer as described previously [9]. Results
Fig. 1 shows a thin-layer chromatogram of outer segment total lipids cochromatographed with four authentic standards. The phosphatidylserine region is separated into two spots, PS-1 and PS-2; the phosphatidylcholine region into five spots, PC-1 through PC-5; and the phosphatidylethanolamine region into three spots, PE-1, PE-2, and PE-3. Several pieces of evidence rule out significant contributions to any of these resolved spots by minor phospholipids. Treatment of an identical plate with ninhydrin showed that all spots attributed to phosphatidylserine and phosphatidylethanolamine contained primary amino groups. Authentic dipalmitoyl phosphatidylcholine, dipalmitoyl phosphatidylethanolamine and bovine brain phosphatidylserine co-migrated with PE-3, PC-5 and PS-2, respectively. Upon exposure to 9 M H:SO4 and heat, PC-5, PC-4 and PE-3 charred only weakly, in contrast to the dense black charring of the other outer segment phospholipid spots. Both authentic dipalmitoyl phosphatidylcholine and dipalmitoyl phosphatidylethanolamine also exhibited weak charring which is characteristic of very saturated compounds. No spot attributable to highly saturated phosphatidylserine was detected. Synthetic palmitoyl
Biochimica et Biophysica Acta, 5 5 2 ( 1 9 7 9 ) 2 9 4 - - 3 0 6 © E l s e v i e r / N o r t h - H o l l a n d B i o m e d i c a l Press
BBA 78324
DISATURATED AND DIPOLYUNSATURATED PHOSPHOLIPIDS IN THE BOVINE RETINAL ROD OUTER SEGMENT DISK MEMBRANE
G E O R G E P, M I L J A N I C H *, L A R R Y A. D R A T Z
A. S K L A R
**, D R I N A L. W H I T E and E D W A R D
Division of Natural Sciences, University of California, Santa Cruz, C A 95064 (U.S.A.) (Received October 12th, 1978) Key words: Phospholipid; Acyl chain distribution;Rhodopsin; (Rod outer segment membrane)
Summary Thin-layer chromatography was used to separate the major phospholipid headgroup classes of the rod outer segment disk membrane into subfractions which differ markedly in fatty acid composition. At least 18% of the rod outer segment phosphatidylcholine must contain two saturated fatty acids. Furthermore, two unsaturated fatty acids are found in at least 43% of the phosphatidylserine, 24% of the phosphatidylcholine, and 24% of the phosphatidylethanolamine. The unsaturated acids are predominantly polyunsaturated in all cases. A similar separation, but with less resolution, was achieved with silicic acid column chromatography. The temperature dependence of the polarization of the fluorescence of trans.parinaric acid (9,11,13,15-all-trans-octadecatetraenoic acid) showed that the thermal behavior of aqueous dispersions of the phosphatidylcholine subfractions was consistent with their fatty acid compositions.
Introduction The large degree of molecular heterogeneity of phospholipids in biological membranes is well known. However, the detailed combinations of two fatty acids with each other and of these pairs with particular phospholipid heedgroups are poorly documented. Fractionation and characterizationof these combinations might provide clues for understanding the functional role(s)of * Present address: Department of Biochemistry and Biophysics, University of California, San Francisco, CA 94143, U.S.A. * * Present address:Department of Medicine, Methodist Hospital, Baylor College of Medicine, Houston, TX 77030, U,S.A,
295 this molecular polydispersity and should also suggest more relevant model membrane systems for physical studies. In addition, preparative fractionation of various phospholipid species with different fatty acid compositions from natural sources could provide material for model studies [ 1 ]. In this paper we report the separation, by a one-dimensional thin-layer chromatography (TLC) method, of the major phospholipid classes of the retinal rod outer segment membrane into subfractions which differ in fatty acid composition. Fatty acid analysis reveals the presence of saturated-saturated, saturated-polyunsaturated, and polyunsaturated-polyunsaturated species of outer segment phosphatidylcholine and phosphatidylethanolamine and saturated-polyunsaturated and polyunsaturated-polyunsaturated species of outer segment phosphatidylserine. * A similar separation can be achieved, with somewhat less resolution by silicic acid column chromatography. As a further characterization of this separation, we also show that the temperature dependence of the fluorescence polarization of trans-parinaric acid in aqueous dispersions of outer segment phosphatidylcholine subfractions is consistent with their fatty acid compositions. Materials and Methods
Preparation of rod outer segments. Highly purified bovine outer segments were isolated by the method of Raubach et al. [2] except that the second sucrose density gradient was centrifuged for 90 min and all solutions contained 0.1 mM EDTA and 0.15 mM CaCI~ to minimize oxidative damage [3]. Lipid extraction. Under the conditions employed for column chromatography we have found that retinal forms a Schiff's base with phosphatidylethanolamine and this complexation prevents complete separation of these two components. Conversion of retinal to retinaloxime eliminates this problem. Prior to lipid extraction for column chromatography, the outer segment membrane suspension was mixed with enough 0.1 M hydroxylamine (pH 7) to give a 100-fold molar excess of hydroxylamine over retinal and then exposed to an unfiltered 300 W floodlight while the sample tube was immersed in a large beaker of cold tap water. Retinaloxime appears to be unstable, so column fractionation was commenced immediately subsequent to lipid extraction. If the lipids were to be used solely for TLC the oximation step was unnecessary and was omitted. Total outer segment lipids were extracted by a modification of the procedure of Folch et al. [4]. The chloroform and methanol contained 50 rag/1 butylated hydroxytoluene. In this and in the following methods, the chloroform and methanol were redistilled and were bubbled with argon immediately before use. Whenever feasible, manipulations were performed under an argon atmosphere. Column chromatographic fractionation of phospholipid headgroup classes. The total lipids were first applied to a diethylaminoethyl-cellulose column prepared and eluted essentially by the procedure of Rouser et al. [5]. Elution with chloroform removed butylated hydroxytoluene, cholesterol, retinal, * P o s i t i o n a l a n a l y s e s of the fatty acids in these s u b f r a e t i o n s were n o t p e r f o r m e d . T h e r e f o r e , the order o f f a t t y acid t y p e s used in naming these p h o s p h o l i p l d species is a r b i t r a r y .
296 retinaloxime, vitamin E, and free fatty acids. Elution with 1 : 1 chloroform/ methanol removed phosphatidylcholine and phosphatidylethanolamine and glacial acetic acid removed pure phosphatidylserine. The phosphatidylcholinephosphatidylethanolamine mixture was reduced to dryness by rotary evaporation, redissolved in chloroform and further fractionated on a silicic acid column (approx. 1 g silicic acid per 5 mg phospholipid). Thin-layer chromatography. TLC of outer segment total lipid extract was performed on 5 cm by 20 cm plates coated with a 0.25 mm thick layer of silica gel H (EM Laboratories, Inc.). For identification of the various lipid species, authentic standards were cochromatographed with a few pg of outer segment lipids in chloroform/methanol/water (65 : 25 : 5, v/v) and 50 mg/1 butylated hydroxytoluene. Standards used included bovine brain sphingomyelin (Applied Sciences Laboratories); lysophosphatidylcholine and lysophosphatidylethanol. amine (Miles Laboratories); soybean phosphatidylinositol, dimyristoyl phosphatidylcholine, dipalmitoyl phosphatidylcholine, distearoyl phosphatidylcholine, dipalmitoylethanolamine, N,N-dimethyldipalmitoyl phosphatidylcholine, vitamin E, cholesterol, retinal, retinol, triolein and oleic acid (Sigma Chemical Co.); and highly purified bovine brain phosphatidylserine (a generous gift from Dr. D. Papahadjopoulos, University of California, San Francisco). 1-Palmitoyl-2-docosahexaenoyl phosphatidylcholine and didocosahexaenoyl phosphatidylcholine, synthesized as described below, were also used as standards. Chromatograms were visualized by either exposure to I2 vapor, charring with 9 M H2SO4, or with 0.2% ninhydrin in ethanol. The same TLC system was also used to isolate phospholipid subfractions for fatty acid and phosphorus analyses. Approx. 1.0--1.5 mg of outer segment total lipid in approx. 75 pl chloroform was applied to the TLC plate as a thin band 3 cm long and was developed in an argon-saturated jar. Phosphorus and fatty acid analysis of the subfractions. The plate was dried quickly in a stream of argon a n d sprayed with 0.4% dichlorofluorescein in methanol under argon pressure. Lipid bands were visualized under ultraviolet light, marked with a pencil and quickly scraped with a razor blade into 10-ml screw-capped tubes containing 5/~g butylated hydroxytoluene and 14.7/~g heneicosanoic acid (21 : 0) (Supelco, Inc.) in 100 pl chloroform. Fatty acid methyl esters were formed under argon by a modification of the method of Morrison and Smith [6] and centrifuged at 3000 X g for 10 min to remove suspended silica gel. The fatty acid methyl esters (in the upper layer) were analyzed by gas-liquid chromatography (GLC) using a 6 ft glass column containing 10% SP-2330 on 100/120 Supelcoport (Supelco, Inc.} which was temperature programmed from 130 to 250°C at 4°C/min. Identification of peaks was made by comparison with reference methyl ester mixtures (Supelco, Inc.). The integrated area (obtained with an Autolab model 6300 digital integrator) of each peak was divided by the molecular weight of the corresponding methyl ester to yield relative molar concentrations. The lower phases from the transmethylation procedure were heated at l l 0 ° C for several hours until dry and subjected to phosphorus analysis by the method of Chen et al. [ 7 ] modified for smaller sample sizes. Synthesis of palmitoyl-docosahexaenoyl phosphatidylcholine. 1-Palmitoyl2-docosahexaenoyl phosphatidylcholine was synthesized by the method of
297 Evans and Tinoco [8]. TLC of the synthesized material showed two closely migrating bands. GLC analysis demonstrated that the main band, with a slightly smaller RF, contained a 1 : 1 mol : mol ratio of palmitic and docosahexaenoic acids and the minor band, with a larger Rr, contained all docosahexaenoic acid. Phosphorus analysis showed this minor band to represent approx. 4 mol% of the total phospholipid. This component is didocosahexaenoyl phosphatidylcholine. Polarization of fluorescence of trans-Parinaric acid. The temperature dependence of the polarization of the fluorescence of trans-parinaric acid (9,11,13,15all-trans-octadecatetraenoic acid) in phospholipid dispersions was measured using a Hitachi-Perkin-Elmer MPF-2A spectrofluorometer as described previously [9]. Results
Fig. 1 shows a thin-layer chromatogram of outer segment total lipids cochromatographed with four authentic standards. The phosphatidylserine region is separated into two spots, PS-1 and PS-2; the phosphatidylcholine region into five spots, PC-1 through PC-5; and the phosphatidylethanolamine region into three spots, PE-1, PE-2, and PE-3. Several pieces of evidence rule out significant contributions to any of these resolved spots by minor phospholipids. Treatment of an identical plate with ninhydrin showed that all spots attributed to phosphatidylserine and phosphatidylethanolamine contained primary amino groups. Authentic dipalmitoyl phosphatidylcholine, dipalmitoyl phosphatidylethanolamine and bovine brain phosphatidylserine co-migrated with PE-3, PC-5 and PS-2, respectively. Upon exposure to 9 M H:SO4 and heat, PC-5, PC-4 and PE-3 charred only weakly, in contrast to the dense black charring of the other outer segment phospholipid spots. Both authentic dipalmitoyl phosphatidylcholine and dipalmitoyl phosphatidylethanolamine also exhibited weak charring which is characteristic of very saturated compounds. No spot attributable to highly saturated phosphatidylserine was detected. Synthetic palmitoyl
